Proceedings of the RILEM International Symposium on Bituminous Materials: ISBM Lyon 2020 (RILEM Bookseries, 27) 3030464547, 9783030464547
This volume highlights the latest advances, innovations, and applications in bituminous materials and structures and asp
102 16 60MB
English Pages 1995 [1806] Year 2021
Preface
Logos and Sponsors
Contents
RILEM Publications
Technical Committee 264-RAP (“Asphalt Pavement Recycling”)
Effect of Aging on the Rheological Properties of Blends of Virgin and Rejuvenated RA Binders
1 Introduction
2 Experimentation
2.1 Materials
2.2 Testing Plan
3 Results and Analysis
3.1 Penetration Value and Softening Point
3.2 Dynamic Shear Rheometer (DSR) Results
4 Conclusions
References
Experimental Investigation on Water Loss and Stiffness of CBTM Using Different RA Sources
1 Introduction
2 Materials
3 Testing Methods and Procedures
4 Results
4.1 Water Loss
4.2 ITSM
5 Conclusions
References
International Evaluation of the Performance of Warm Mix Asphalts with High Reclaimed Asphalt Content
1 Introduction
2 Inter-laboratory Testing Program
2.1 General Principle
2.2 Material
2.3 RA Characterization
2.4 Mix Design
2.5 Tests and Results
2.6 Comments on the Results
3 Conclusion
References
Technical Committee 272-PIM (“Phase and Interphase Behaviour of Bituminous Materials”)
Interlaboratory Test to Characterize the Cyclic Behavior of Bituminous Interlayers: An Overview of Testing Equipment and Protocols
1 Introduction
2 Testing Equipment and Protocols of Participating Laboratories
3 Preliminary Results
4 Conclusions
References
Complex Bituminous Binders, Are Current Test Methods Suitable for?
1 Introduction
2 Experimental Plan
3 Results
4 Conclusion
References
Technical Committee 278-CHA (“Crack-Healing of Asphalt Pavement Materials”)
Terms and Definitions on Crack-Healing and Restoration of Mechanical Properties in Bituminous Materials
1 Introduction
2 Rheological Response of Bituminous Materials to Cyclic Loading
3 Continuum Damage Mechanics Versus Fracture Mechanics
4 Restoration and Crack-Healing During Rest
5 Conclusion: Proposed Definitions
References
Testing Methods to Assess Healing Potential of Bituminous Binders
1 Introduction
2 Materials
3 Testing Methods
3.1 TS-SRP
3.2 TS-MRP
3.3 LASH
4 Preliminary Results and Concluding Remarks
References
Other Conference Papers
A Comparative Study of RTFO, PAV and UV Aging Using FT-IRS and DSR Tests
1 Introduction
2 Methodology
2.1 Asphalt Binder
2.2 UV and Oven Aging Simulation Test in the Laboratory
2.3 Fourier Transform Infrared Spectroscopy (FT-IRS) Test
2.4 Experimental Plan
3 Result and Discussion
3.1 Analysis of Carbonyl and Sulfoxide Content
3.2 C=O and S=O Indices of Aged Asphalt Binder at Similar Stiffness
4 Conclusion
References
A Critical Analysis of the Standard Used to Evaluate De-icing Damage in Asphalt Materials
1 Introduction
2 Experimental Testing
3 Simulation of Test
4 Conclusions and Recommendations for Future Work
References
A Laboratory Procedure for Improving the Design of Road Rehabilitation Actions: Study of a Real Case in a Highway Pavement
1 Introduction
2 Description of the Laboratory Procedure
3 Application Example: Study of a Real Case in a Highway Pavement
4 Conclusions
References
A Literature Review of Bitumen Aging: From Laboratory Procedures to Field Evaluation
1 Introduction
2 Accelerated Laboratory Aging Protocols
3 Aging Indexes and Relationship with Field-Aging
3.1 Chemical Indexes
3.2 Physical Indexes
3.3 Physical and Chemical Coupling Approaches
4 Conclusion
References
A Method to Reduce Occlusion While Measuring Pavement Surface Profiles Using Triangulation Based Laser Scanners
1 Introduction
2 Background and Scope
3 Experimental Work
4 Discussion
References
A New Approach to Calibration and Use of Mechanistic-Empirical Design Methods
1 Background
2 CalME Mechanistic-Empirical Design Software
3 Observations Leading to New Calibration Method
3.1 Calibration Steps
4 Conclusions
References
A New Approach Towards Scientific Evaluation and Performance-Related Design of Bituminous Joint Sealing Materials and Constructions
1 Introduction
2 Decisive Influences on Joint Sealings
2.1 Loads Induced by Weather
2.2 Loads Induced by Traffic
2.3 Other Loads
3 Design of Test Specimen
3.1 Concrete Mixture
3.2 Representative Dimensions of Specimen
4 An Approach Towards More Realistic Functional Testing
5 Conclusion and Outlook
References
A New Bituminous Ballast-Less and Sleeper-Less Thin Railway Track Structure
1 Introduction
2 The Bituminous GB5® Slab with Corkelast® ERS
3 Validation as Rail Fastening System for Heavy Rail Application
4 Conclusions
References
A New Method for Healing Quantification of Bituminous Materials
1 Introduction
1.1 Viscoplastic Strain, Damage and Healing
2 Materials
3 Experimental Investigation
4 Results and Discussion
4.1 Normalised Strain Recovery
4.2 Estimation of Viscoelastic and Viscoplastic Strain Components
4.3 Generation of Healing Curves
5 Summary
References
A Numerical Model to Predict the Thermo-Viscoelastic Behaviour of Asphalt Concrete for Electric Road System
1 Electric Road System
2 Thermo-Viscoelastic Simulations
3 Results and Analyses
4 Conclusions
References
A Numerical Study of the Influence of Aggregates Shape on Creep and Recovery Tests Behavior
1 Introduction
2 Materials and Methods
2.1 Model Construction, Calibration and Validation
3 Results and Discussions
References
A Simpler Method for Establishing an Approximate Value of Built-in Temperature Differential in Concrete Pavement
1 Introduction
2 Literature Review
3 Objective and Scope
4 Methodology Adopted
5 Observations and Discussion
6 Conclusions
References
A Study on Fatigue Behaviour of Bitumen Emulsion Mastic, Modified with Active Fillers
1 Introduction
2 Materials and Experiments
2.1 Basic Materials
2.2 Mastic Preparation
2.3 Testing Program
2.4 Data Analysis Approach
3 Results and Discussion
4 Conclusions
References
A Study on the Properties of Polyurethane-Elastomer-Modified Asphalt
1 Introduction
2 Experimental
2.1 Materials
2.2 Preparation of the Polyurethane Elastomer Modifier
2.3 Preparation of Modified Asphalt
2.4 Characterization
3 Result and Discussion
3.1 Basic Properties of Modified Asphalt
3.2 Analysis of Infrared Spectroscopy
3.3 Analysis of Fluorescence Microscopy
3.4 Analysis of Rheological Properties
4 Conclusion
References
A Wear-Based Description of Mechanisms Underlying the Bitumen Removal Process
1 Introduction
2 Experimental Program
2.1 Simulation of Wear in Laboratory
2.2 Test Method
3 Results
3.1 Evolutions of Friction and Texture
3.2 Visual Observations
3.3 Characterization by an Energy Approach
3.4 Discussions
4 Conclusions
References
Active Mitigation of Low-Temperature Cracking in Asphalt Pavements
1 Introduction
2 Thermomechanical Modeling
2.1 Thermal Formulation
2.2 Simulated Weather Conditions
2.3 Mechanical Formulation
2.4 Modeling of an Idealized Case
3 Results and Discussion
4 Conclusion
References
Ageing Behavior of Porous and Dense Asphalt Mixtures in the Field
1 Introduction
2 Test Section and Experimental Method
2.1 Asphalt Mixture Design
2.2 Overview of the Test Section
2.3 UV Reflective Glass-Plate
2.4 Cyclic Indirect Tensile Test
3 Results and Discussion
3.1 Field Ageing Behavior
3.2 UV Ageing Behavior
4 Conclusions
References
An Attempt to Characterize the RAP Material for Hot Recycled Mix Design Purposes
1 Introduction
2 Materials and Methodology
2.1 Materials
2.2 Fragmentation Test
3 Results and Discussion
4 Conclusions
References
An Attempt to Distinguish Thermal from Oxidative Ageing of Asphalt Binders by NRTFOT
1 Introduction
2 Materials and Methods
2.1 Material
2.2 Methods
3 Results and Discussion
3.1 MCR
3.2 FTIR
4 Conclusion
References
An Interlaboratory Test Program on the Extensive Use of Waste Aggregates in Asphalt Mixtures: Preliminary Steps
1 Introduction
2 Background, Objective and Experimental Approach
3 Results and Analysis
3.1 Aggregate Characterization
3.2 Mix Design
4 Summary and Conclusions
References
An Investigation on Longitudinal Unevenness Indicators and Their Potential on Surface Characterisation
1 Introduction
2 Longitudinal Profile Indicators
3 Input Data
4 Statistical Comparison
5 Correlations Between Different Indicators
6 Conclusions
References
An Overview of Black Space Evaluation of Performance and Distress Mechanisms in Asphalt Materials
1 Introduction
2 Damage and Healing
3 Rheological Assessment of Cracking Behavior
4 Rheological Characterization of Alternative Binders
5 Summary
References
Analysis of a New Approach for Characterizing the Performance of Asphalt Binders Through the Multiple Stress Creep and Recovery Test
1 Introduction
2 Materials
3 Standard MSCR test
4 Uniaxial Repeated Load test
5 The New Approach for MSCR Test
6 Conclusions
References
Analysis of High Temperature Performance of Composite Asphalt Rubber Mixture
1 Introduction
2 Materials and Experiment
2.1 Materials
2.2 Experimental Sections
3 Results and Discussion
3.1 Dynamic Stability Analysis and Deformation Law
3.2 Rebound Deformation Analysis
3.3 Shear Strength Analysis
4 Conclusions
References
Analysis of Low Temperature Cracking Behavior at Binder, Mastic and Asphalt Concrete Levels
1 Introduction
2 Material and Test Program
3 Low Temperature Testing of Bitumen and Mastic
3.1 Fracture Toughness Testing (FTT)
3.2 Double Torsion (DTT)
4 Validation Using Small and Medium Scale Performance Testing
5 Conclusions
References
Analysis of Ovalization Measurements on Flexible Airfield Pavement Under HWD Dynamic Impulse Load
1 Introduction
2 History of the Ovalization Test
3 Ovalization Campaigns for Airfield Pavement Needs
4 New Design Proposals for the Ovalization Device
5 Conclusion
References
Application of Activated Carbon on Induced Heating-Healing Characteristics of Aged Mixes
1 Introduction
2 Materials and Sample Preparation
3 Testing Methods
3.1 Laboratory Aging
3.2 Microwave Heating
3.3 Cyclic Breaking-Healing
4 Results and Discussion
4.1 Effect of Aging Level on Microwave Heating
4.2 Effect of Aging Level on Induced Healing
5 Conclusions
References
Application of the Synthetics Geo-Composites in the Arid Zones for Rehabilitation of the Flexible Pavements Road Experimental Analysis
1 Introduction
2 Application of the Géo-Synthetic in the Linear Infrastructures
3 Experimental Tests
4 Results and Discussions
4.1 Tests to the Laboratory
4.2 Tests of HWD
5 Conclusion
References
Applying the Grid Method to Study the Mechanical Behavior at Micro Scale of Asphalt Mixtures Containing Reclaimed Asphalt Pavements
1 Introduction
2 Experimental Set-Up and Conditions
2.1 Material and Design
2.2 Loading Conditions
2.3 Outline of the Grid Method
3 Results
3.1 Response at the Macroscopic Scale
3.2 Response at the Microscopical Scale
4 Conclusion
References
Asphalt Aggregate Adhesion: Study of the Influence of the Morphological, Chemical and Mineralogical Properties of Different Aggregates from Southern Brazil
1 Introduction and Background
1.1 Research Objectives
2 Materials and Test Methods
3 Discussion of Results
4 Concluding Remarks
References
Asphalt Layers Thickness Determination for Top Down Cracking Initiation and a Case Study
1 Introduction
1.1 Top Down Cracking Identification and Studies
1.2 Scope of the Current Study
2 Methodology Followed
3 Results
4 Case Study
5 Conclusions
References
Asphalt Permeability In-service: How Can It Be Evaluated Before and After Moisture Damage?
1 Background
1.1 Asphalt Performance Properties
1.2 Routine Testing for Moisture Damage
2 Development of Test Procedures
3 Experimental Methods and Findings
3.1 Experimental Strategy
3.2 Research Findings
4 Implementation of Outcomes
References
Assessing Geometric Features Impact of Laboratory Specimens on ITS Variation
1 Introduction
2 Materials
3 Methods
4 Experimental Program
4.1 Mix Design
4.2 Indirect Tensile Strength (ITS)
4.3 ANOVA Statistical Method
5 Results and Analysis
6 Conclusions
References
Assessment of Tensile Strength Reserve of Asphalt Mixtures at Low Temperatures
1 Introduction
2 Materials and Methods
3 Results and Discussion
4 Conclusions
References
Assessment of the Behavior of Emulsified Asphalt Mixes During Curing
1 Introduction
2 Methodology
3 Results
3.1 Water Loss
3.2 Oedometric Tests
3.3 Curing and Mechanical Characterization
4 Conclusion
References
Characterization of Aluminosilicate-Based Warm-Mix Asphalt Additive Using Experimental Techniques
1 Introduction
2 Materials and Methods
2.1 Reagents
2.2 Synthesis of Asphalt Foaming Additive: STEAM
2.3 Materials Characterization
3 Results and Discussion
4 Conclusions and Future Scope
References
Characterization of Layer Index in Bituminous Pavement Using Mechanistic-Empirical Approach Based on Concentration Factor in a Layered System
1 Introduction
2 Objective of Present Work
3 Flexible Pavement with Three Layered System
3.1 Odemark’s Method for Transformation of Layered System
3.2 Determination of Concentration Factor Using Vertical Interface Stress
4 Determination of Pavement Surface Deflection
5 Characterization of Deflection Bowl
6 Discussion on Test Results
7 Conclusions
References
Characterization of Rejuvenated Asphalt Binders Using FTIR and AFM Techniques
1 Introduction
2 Experimental Design
2.1 Material Collection
2.2 Chemical Characterization and Modification
2.3 Blending of Rejuvenators and Asphalt Binder
3 Results and Discussion
3.1 Effect of Rejuvenators on Asphalt Oxidative Potential
3.2 Effect of Rejuvenators on Asphalt Morphologies
4 Conclusions
References
Characterization of the Lateritic Soil of Kamboinsé (Burkina Faso)
1 Introduction
2 Materials and Methods
2.1 Materials
2.2 Methods
3 Results and Discussion
3.1 Raw Materials
3.2 Improved Cement Laterite
3.3 Litho-Stabilized Laterite
4 Conclusion
References
Circular Accelerated Surfacing Tester (CAST) for Evaluating Chip Sealing Binders
1 Introduction
2 Methodology
2.1 Circular Accelerated Surfacing Tester (CAST)
2.2 Materials and Test Standards
2.3 Stereo-Photogrammetric Image Analysis
2.4 Results and Discussion
3 Conclusion
References
Coefficient of Thermal Expansion and Thermally Induced Internal Cracking of Asphalt Mixes
1 Introduction
2 Objectives
3 Materials
4 Test Methods
5 Results and Discussion
6 Summary and Conclusions
References
Combined Recycling of RAS and RAP: Experimental and Modeling Approach
1 Introduction
2 Literature Review
3 Materials and Methodology
4 Results and Discussion
5 Conclusion and Perspectives
References
Combined Use of Natural Rubber, Biomass and Plastic Wastes in Bitumen Modification and Flexible Pavement Construction
1 Introduction
2 Materials and Methods
2.1 Materials
2.2 Preparation of Soda Lignin
2.3 Preparation of Natural Rubber-Lignin Mixture for Bitumen Modification
2.4 Modification of Bitumen and Specification Properties
2.5 Analysis of Prepared Samples and Construction of Experimental Pavement
3 Result and Discussion
4 Conclusion
References
Comparative Evaluation of Different Methods for Assessing the Glass Transition Temperature of Bituminous Binders
1 Introduction
2 Materials and Methods
3 Experimental Results
3.1 DMA
3.2 TMA
4 Conclusions
References
Comparative Study of the Mechanical Behaviour of Bitumen- and Cement-Dominated Mixtures with Reclaimed Asphalt
1 Introduction
2 Materials and Methods
3 Results and Discussion
4 Conclusions
References
Comparing Compressive and Flexural Dynamic Asphalt Modulus by Statistical and Neural Network Modelling
1 Introduction
2 Material and Methods
3 Results and Discussion
4 Conclusion
References
Comparing Laboratory Curing and In-Pavement Curing of a Foamed Bitumen Stabilized Base Course Material
1 Background
2 Method
3 Results and Discussion
4 Conclusion
References
Comparing Wet Mixed and Dry Mixed Binder Modification with Recycled Waste Plastic
1 Introduction
2 Background
2.1 Recycled Plastic as a Binder Extender and Modifier
2.2 Wet Mixing and Dry Mixing Processes
3 Methods
4 Results and Discussion
5 Conclusion
References
Comparison Between Rheological Behavior of a Neat Asphalt Binder and a Bio-Binder from Renewable Source
1 Introduction
2 Materials and Methods
3 Results and Discussion
4 Conclusions
References
Comparison of Alternate Binder Aging Methods
1 Introduction
2 Materials and Methods
3 Results and Discussion
4 Conclusions
References
Comparison of Different DSR Protocols to Characterise Asphalt Binders
1 Introduction
2 Experiment
3 Results
4 Conclusion
References
Comparison of the Effects of Hydrated Lime on the Moisture-Induced Damage of Stone Mastic Asphalt (SMA) Mixtures
1 Introduction
2 Materials and Methods
2.1 Materials
2.2 Mixture Design
2.3 Saturated Ageing Tensile Stiffness (SATS)
3 Test Results and Discussion
4 Conclusions
References
Comprehensive Model for the Prediction of the Phase Angle Master Curve of Asphalt Concrete Mixes
1 Introduction
2 Data Source, Model Specification, and Estimation
2.1 Data Sources
2.2 Lorentzian Master Curve
2.3 Predictive Model
2.4 Estimation Results
3 Conclusion
References
Crack Evolution of Bitumen Under Torsional Shear Fatigue Loads
1 Introduction
2 Modelling Fatigue Crack Evolution in Bitumen
3 Materials and Laboratory Tests
4 Results and Discussion
5 Conclusions
References
Crack Self-healing in Asphalt as a Flow Process
1 Introduction
2 Materials and Methods
2.1 Materials
2.2 Characterisation of Self-healing and Flow
3 Results
3.1 Self-healing of Asphalt Mixture with Porous Aggregates
3.2 Self-healing of Asphalt Mortar Beams and the Composition of the Drained Mixture
4 Conclusions
References
Cracked Asphalt Pavement Behaviour Under Thermal and Load Cycles – Laboratory Investigation Using Accelerated Loading System
1 Introduction
2 Experiment Description
2.1 Pavement Structure
2.2 Thermomechanical Properties of HMA and Fatigue Pavement Design
2.3 Strain Gauges
2.4 Experimental Protocol
3 Preliminary Results
3.1 TML Response Before the Creation of the Idealized Transverse Crack
3.2 TML Response After the Creation of the Idealized Transverse Crack
4 Conclusion
References
Cracking Healing Evaluation of HMA Using Indirect Tensile Stiffness Modulus Test
1 Introduction
2 Methodology
2.1 Materials and Mix Designs
2.2 Study Specimens
2.3 Healing Evaluation Procedure
3 Results and Discussion
3.1 Evaluation of Stiffness Modulus
3.2 Time and Binder Type Effect on Stiffness Recovery
4 Conclusions
References
Creep Properties of Asphalt Binder, Asphalt Mastic and Asphalt Mixture
1 Background and Objective
2 Methodology and Testing
3 Results
4 Summary and Conclusions
References
Critical Reclaimed Asphalts Binders’ Properties from Empirical and Functional Point of View
1 Introduction
2 Research Objectives
3 Materials and Methods
4 Test Results and Discussion
5 Conclusions
References
Degree of Binder Activity on 100% Recycled Mixtures and Its Linear Viscoelasticity Behavior
1 Introduction
2 Degree of Binder Activity Determination
2.1 RAP Materials
2.2 Artificial Aged Binder and Mixture
2.3 Indirect Tensile Test and Degree of Binder Activity
2.4 Results and Discussion
3 Experimental Study in the Linear Viscoelastic Domain
3.1 Samples
3.2 Complex Modulus and 2S2P1D Modelling
3.3 Results and Discussion
4 Conclusions
References
Determination of Epoxy Resin Concentration in Epoxy Modified Bitumens
1 Introduction
2 Methodology
2.1 Preparation of Standards
3 Results and Discussion
3.1 Epoxy Concentration by the FTIR Method
3.2 Epoxy Concentration by the FNAA Method
3.3 Comparison of FTIR and FNAA Methods
3.4 Precision
4 Conclusion
References
Determination of Thermal Stresses in Asphalt Layers as a Problem of Thermo-elasticity and Unsteady Heat Flow
1 Introduction
2 Theoretical and Finite Element Formulation of the Problem
3 Analyzed Pavement Structures and Material Data
4 Presentation of the Results
5 Conclusions
References
Determining the Prominence of Texture Scales on Road Skid Resistance
1 Introduction
2 Specimens of Real Road and Artificial Surfaces
3 Decomposition of the Surface Profiles and Creation of the New Profiles
4 Results and Discussion
5 Conclusion
References
Development of Permanent Deformation Models with the Inclusion of AIMS Coarse Aggregate Angularity
1 Introduction
2 Methodology
2.1 Materials, Mix Designs and Sample Preparation
2.2 Performance Testing Evaluation
3 HMA Permanent Deformation Model
3.1 Multiple Non-linear Regression Models (MNLR)
3.2 Soft Computing Models
3.3 Parametric and Sensitivity Analysis
4 Conclusions
References
Discrete Modelling of 2PB Complex Modulus Test of Hot Asphalt Mixes Using Irregular Aggregates
1 Introduction
2 Experimental Procedures
2.1 Complex Modulus Test
3 Numerical Procedures
3.1 Sample Preparation
3.2 Complex Modulus Test Modelling
4 Calibration of the Numerical Modelling
5 Conclusions
References
Effect of Ageing Kinetics on the Rheological Properties of Bituminous Binders and Mixes: Experimental Study and Multi-scale Modeling
1 Introduction
2 Materials and Mix Design
2.1 Materials Properties
2.2 Ageing of Asphalt Mixture
2.3 Ageing of Bitumen
3 Rheological Tests
3.1 Mixes Rheological Tests
3.2 Properties of Bitumen
4 Multiscale Modeling
5 Conclusion
References
Effect of Fiberglass Geogrid Reinforcement to Fatigue Resistance of Bituminous Mixtures
1 Introduction
2 Specimens Preparation and Experimental Devices
3 Results and Analysis
4 Conclusions
References
Effect of Filler Type and Content on the Rheological Properties of Asphalt Mastics
1 Introduction
2 Materials and Experimental Investigation
2.1 Material
2.2 Characterization of Fillers
2.3 Design of Asphalt Concrete Mix
2.4 Preparation and Testing of Asphalt Mastics
3 Results and Discussions
3.1 Performance of Asphalt Mastics
4 Conclusions
References
Effect of RAP and Binder Properties on Indirect Tensile Strength and Dynamic Modulus of Cold Recycled Foamed Asphalt Mixtures with High RAP Content
1 Motivation and Objective
2 Materials
2.1 RAP
2.2 Foamed Asphalt
2.3 Cold Recycled Foamed Asphalt Mixtures
3 Methodology
3.1 Indirect Tensile Strength
3.2 Dynamic Modulus
4 Results and Discussions
4.1 Indirect Tensile Strength
4.2 Dynamic Modulus
5 Conclusions
References
Effect of RAP Gradation on Fracture Tolerance of Asphalt Mixtures
1 Introduction
2 Materials
3 Experimental Design
4 Results and Discussion
5 Conclusion and Recommendation
References
Effect of Soybean Oil Derived Additives on Improved Performance of Polymer-Modified Asphalt Binder and Mix Containing 50% Fine-Graded RAP
1 Introduction
2 Materials and Experimental Methods
3 Results and Analysis
4 Conclusions and Future Recommendations
References
Effect of Temperature on Self-healing Properties of Bituminous Binders
1 Introduction
2 Materials and Methods
3 Results and Discussion
4 Conclusions
References
Effect of Thermal Ageing on the Mechanical Properties and Cracking Behaviour of Asphalt Concrete
1 Introduction
2 Rheological and Crack Growth Properties of Asphalt Concrete
2.1 Materials and Method
2.2 Laboratory Ageing Procedure and Specimen Preparation
2.3 Complex Modulus Test
2.4 Semi-Circular Bending (SCB) Test
2.5 Experimental Results and Discussion
3 Viscoelastic Crack Growth Modelling
4 Comparison Between Numerical Model and Experiments
5 Conclusion
References
Effects of Adding a Geotextile into a Pavement Structure on FWD Experimental Data and Estimated Pavement Remaining Service Life
1 Introduction
2 Methodology
3 Results and Analysis
4 Conclusion
References
Effects of Aggregate Shape Parameters and Gradation on High-Modulus Asphalt Mix Performance
1 Introduction
2 Materials and Methods
2.1 Aggregate Shape Parameters Determination Using Image Analysis
2.2 Sample Preparation and Test Methods
3 Results and Discussions
4 Conclusions
References
Effects of Binder Content, Filler Content, Air Void and Modification on Damage Characteristics of Asphalt Mix Under Fatigue Loading
1 Introduction
2 Materials
3 Test Methods
4 Methodology
5 Results and Discussion
6 Conclusion
References
Effects of Chemical and Microstructural Constituents on the Healing Characteristics of Asphalt Binders
1 Introduction
2 Materials and Methods
2.1 Binders
2.2 Test methods
3 Results
3.1 Effects of the Chemical Composition on the Healing Potential
3.2 AFM x Healing
4 Conclusions
References
Effects of Water Saturation and Freeze-Thaw Cycles on Fatigue Behavior of Bituminous Mixtures
1 Introduction
2 Specimen Preparation and Conditioning
2.1 Specimen Fabrication
2.2 Specimen Conditioning
3 Fatigue Test and Experimental Devices
3.1 Fatigue Failure Criteria
3.2 Wöhler Curve
4 Conclusions
References
Efficiency of Ultrasonic Pulse Velocity Test in the Determination of Residual Asphalt Mix Pavement Properties
1 Introduction
2 Material and Experimental Methodology
3 Results and Discussions
3.1 SNR Effects on TOF Determination and Waves Cartographies
3.2 Wave Propagation Theory and Mechanical Properties Cartographies
4 Conclusion
References
Elasto-Plastic Model for Bitumen Stabilized Materials Using Triaxial Testing and Finite Element Modelling
1 Introduction
2 Laboratory Tests
2.1 Mixtures Design
2.2 Laboratory Methodology
2.3 Triaxial Test Setup
2.4 Resilient Modulus Testing
3 Finite Element Elasto-Plastic Model
3.1 ABAQUS Model Creation
3.2 Model Calibration
3.3 Model Validation
3.4 Hardening Effect and Cohesion Variation
4 Results
4.1 ABAQUS Curves Versus Laboratory Curves
5 Conclusion
References
Empirical, Rheological and Chemical Properties of Aged Binder with Rejuvenators at Different Ageing Levels
1 Introduction
2 Materials and Methods
3 Results and Discussion
3.1 Empirical Properties
3.2 Rheological Properties
3.3 Chemical Properties
4 Conclusion
References
Enhanced Models for Vehicle Fuel Consumption for Life Cycle Assessment of Pavements
1 Introduction
2 Methodology
3 Illustrative Highway Scenarios
3.1 SLDV Scenarios
3.2 LLDV and Two Axle Truck Scenarios
3.3 Bus and Combination Truck Scenarios
4 Fuel Economy Prediction Results
4.1 SLDV Midsize Vehicles
4.2 SLDV Large Sedan Vehicles
4.3 LLDV Large SUV and Two Axle Class 5 Truck Vehicles
4.4 City Bus and Combination Truck Tractor-Trailer Vehicles
5 Summary
References
Enhanced Self-healing Properties in Stone Mastic Asphalt with Encapsulated Bitumen Rejuvenators
1 Introduction
2 Materials and Methods
2.1 Materials
2.2 Sunflower Oil Encapsulation Procedure
2.3 Fatigue Tests
2.4 Self-healing Measurements
3 Results and Discussion
4 Conclusions
References
Establishing Correlations Between Force Ductility and DSR Parameters of Asphalt Emulsion Residues
1 Introduction
1.1 Objective
2 Materials and Experimental Plan
3 Results and Discussions
4 Conclusion
References
Estimation of Complex Shear Modulus of Binder Blends Produced with RAP Binder and Rejuvenator
1 Introduction
2 Materials
3 Complex Shear Modulus Test Results
4 Estimation Methods
4.1 Approach I
4.2 Approach II
4.3 Estimation Results
5 Conclusions
References
Estimation of Fatigue Damage in Bituminous Mixes Using Digital Image Processing
1 Introduction
2 Objective
3 Experimental Investigations
3.1 Materials
3.2 Experiments
3.3 Monitoring of Fatigue Damage in ITFT Specimens
4 Results and Discussion
4.1 ITS and ITFT Results
4.2 Relationship Between Crack Area and the Loading Cycles
5 Conclusions
6 Future Work
References
Estimation of Glassy Modulus and Differences Between Analysis Methods with Consideration of Aging
1 Introduction
2 Changes of Gg and Other Parameters with Conditioning
3 Summary
References
Evaluate the Impact of the Reclaimed Asphalt Pavement (RAP) Binder Activation on its Shape Properties
1 Introduction
2 Materials and Methods
2.1 Method
3 Results
4 Conclusions
References
Evaluating the Cracking Performance of Asphalt Mixes Using Four-point Beam Fatigue and Semicircular Beam Testing
1 Introduction
2 Material and Methods
3 Test Results and Analysis
4 Conclusion
References
Evaluation of Aging Processes in Binders Stabilised from Cationic Bituminous Emulsion
1 Introduction
2 Materials and Testing Methods
3 Results and Discussion
3.1 Conventional Properties of Bituminous Binders
3.2 Rheological Properties of Bituminous Binders
4 Summary and Conclusions
References
Evaluation of Cracking Performance Indices from Disk-Shaped Compact Tension Testing
1 Introduction and Background
2 Materials and Pavement Sections
3 Performance Indexes from DCT Fracture Test
4 Field Performance
5 Results and Discussion
6 Summary and Conclusions
References
Evaluation of Engineered Crumb Rubber (ECR) Performance Characteristics, Including Warm-Mix Equivalence with Polymer, Draindown Prevention, and Release Enhancement
1 Introduction
2 Objectives
3 Materials Tested
4 Mixture Performance Testing
4.1 Mixture Performance Test Results
4.2 Draindown Test
5 Warm Mix Equivalence
6 Summary and Conclusions
References
Evaluation of Epoxy Asphalt Binders for Open-Graded Friction Course (OGFC) Application
1 Background
1.1 Literature Review
2 Objective and Scope
3 Laboratory Test Results
3.1 Curing Behavior
3.2 Performance Grading
3.3 Fatigue and Aging Resistance
4 Conclusions
References
Evaluation of Fracture Resistance of SMA Mixtures with Various Percentages of TLA-RAP Through the Semi-circular Bend (SCB) Test
1 Introduction
2 Methodology
2.1 Materials and Sample Preparation
2.2 Performance Testing Evaluation
3 Results and Discussion
3.1 Stress-Strain Behaviour
3.2 Fracture Factors Versus RAP Content
4 Conclusions
References
Evaluation of Laboratory and Field Performance of Cold In-Place Recycling (CIR) Asphalt Mixtures
1 Introduction
2 Goal and Objectives
3 Experimental Plan
4 Results and Discussion
4.1 Laboratory Results
4.2 Field Results
5 Conclusions
References
Evaluation of Long Term Performance of High Modulus Bituminous Mixes
1 Introduction
2 Scope of the Work
3 Materials and Experimental Work
3.1 Material Selection
3.2 Selection RAP Percentage in the Mix
3.3 Design of Mixes
3.4 Long Term Aging of Mix
3.5 Indirect Tensile Strength Test (ITS)
3.6 Resilient Modulus (Mr)
3.7 Indirect Tensile Fatigue Test (ITFT)
3.8 Moisture Susceptibility
4 Conclusions
References
Evaluation of Pavement Damage Through the Analysis of Asphalt Layer Modulus and Strain Evolutions During an Accelerated Pavement Test
1 Introduction
2 Presentation of the Full-Scale Test
3 Pavement Damage Evaluation from FWD Tests
3.1 Backcalculation of Asphalt Layer Moduli
3.2 Evaluation of Pavement Damage from Asphalt Layer Moduli
4 Pavement Damage Evaluation from Strain Data Analysis
4.1 Comparison of Measured and Calculated Strains in the Asphalt Layer
4.2 Evaluation of Pavement Damage from Strain Evolution
5 Conclusions
References
Evaluation of Service Level of Emulsions Used in Surface Treatments in Different Stages of Ageing
1 Introduction
1.1 Brief Considerations About Literature Review
2 Method
3 Results
3.1 Results of MSCR
4 Conclusions
References
Evaluation of the Performance of Cold In-Place Recycled Base Layer Supported with Geogrid
1 Introduction
2 Goal and Objectives
3 Materials and Methodology
4 Results and Discussion
4.1 Laboratory Results
4.2 Layered Elastic Analysis Results
5 Conclusions
References
Evaluation of the Properties of Bituminous Binders Recovered from Various Sites in Europe
1 Introduction
2 Experimental
2.1 Materials
2.2 Test Methods
3 Results
3.1 Evaluation of the Aging in the Wearing Course
3.2 Comparison to Laboratory Aging Tests
4 Conclusions
References
Evolution of the Thermo-rheological Indices of Asphalt Binders with Aging
1 Introduction
2 Materials and Testing
3 Results and Discussion
4 Summary and Conclusions
References
Examining the Effects of a Self-healing Elastomer on the Properties of Bitumen
1 Introduction
2 Methodology
3 Results and Discussion
4 Conclusion
References
Experimental and Numerical Study of Low Noise Porous Asphalt Pavement
1 Introduction
2 Material and Method
2.1 Material
2.2 Method
3 Results
4 Conclusion
References
Experimental Evaluation of Swelling and Absorption of Crumb Rubber Aggregates
1 Introduction
2 Materials, Experimental Procedures and Analysis
3 Results and Comments
4 Conclusions
References
Experimental Investigation on Ultraviolet Aging Properties of Silica Nanoparticles-Modified Bitumen
1 Introduction
2 Materials and Methods
2.1 Materials
2.2 Ageing Process
3 Experimental Results
3.1 Dynamic Viscoelastic Properties
3.2 FT-IR Analysis
4 Conclusions
References
Experimental Study on the Grading Distribution of Cold Recycled Asphalt Mixtures Produced with Bitumen Emulsion and High Strength Cement
1 Introduction
2 Experimental Plan
3 Results and Analysis
4 Conclusions
References
Experimental Tests on Diffusion Phenomenon Between Two Different Bitumens
1 Introduction
2 Objective and Scope
3 Materials and Methods
3.1 Materials
3.2 Methods
4 Results and Analysis
4.1 Experimental Rheological Results
4.2 Analysis of the Results
5 Conclusions
References
Experimental Validation of the Dual-Oxidation Routes in Bituminous Binders
1 Introduction
2 Materials and Methods
2.1 Materials and Ageing Treatment
2.2 Spectroscopic Techniques
3 Results and Discussion
4 Conclusions
References
Extending the Use of RAP in Airport Asphalt Resurfacing
1 Introduction
2 Background
3 Methods
4 Results and Discussion
4.1 Mixture Design
4.2 Recycled Asphalt Production Properties
4.3 Asphalt Production
5 Conclusion
References
Factors Affecting Mortar Thickness Distribution and Its Relationship to Cracking Resistance of Asphalt Mixtures
1 Introduction and Objectives
2 Methodology of Imaging
3 Mortar Thickness Distribution Analysis
4 Effect of Modification and Compaction Temperature
5 Effect of Mortar Thickness on Aging of Asphalt Mixtures
6 Summary of Findings
References
Fatigue and Moisture Damage Resistance of Bituminous Mixtures for Railway Trackbeds
1 Introduction
2 Fatigue and Moisture Damage of Bituminous Mixtures
3 Materials and Test Protocol
4 Test Results and Analysis
4.1 Study for Unconditioned Samples
4.2 Study for Moisture Conditioned Samples
5 Discussion
References
Fracture Characterization of Stone Mastic Asphalt (SMA) with Hydrated Lime Through the Semi-circular Bending Test Approach
1 Introduction
2 Research Methodology
2.1 Bituminous Mixture Design and Test Parameters
2.2 Fracture Determination
3 Results and Discussion: Effects of Temperature, HL Content, Aggregate Type and Conditioning on the Fracture Toughness
4 Conclusions
References
Friction Properties of Asphalt Surfacings Containing Sintered Lightweight Aggregate
1 Introduction
2 Materials
2.1 Aggregates
2.2 Mixtures
3 Friction After Polishing Test
4 Results and Discussions
4.1 Aggregate Friction
4.2 Asphalt Friction
5 Conclusions
References
From Laboratory Mixes to Full Scale Test: Rutting Evaluation of Bio-recycled Asphalt Mixes
1 Tested Materials
2 Accelerated Pavement Test
3 Representativeness of the Rutting Test
4 Results and Conclusion
References
Full-Scale Pavement Testing of a High Polymer Modified Asphalt Concrete Mixture
1 Introduction
2 Experimental Program
2.1 Characteristics and Properties of Employed Materials
2.2 Construction, Instrumentation, and Loading Protocols of Experiments
3 Analysis of Measured Pavement Responses
4 Closing Remarks
References
GLOBE: An Innovative Technical Solution to Ensure Waste Free Cold Logistic of Bituminous Binders
1 Scope of the Study
1.1 Globe Solid Bitumen Approach: Benefits and Challenges Toward Current Ecosystem
1.2 The Technical Challenges
1.3 Pellet Design
1.4 Solid Bitumen Storage Stability Evaluation
1.5 Asphalt Mix Evaluation
2 Results and Discussions
2.1 Effect of Gelling Additive on Solid Bitumen Mechanical Resistance
2.2 Effect of Process Technology on Storage Abilities of Solid Bitumen.
2.3 Effect of the Pellet Design on the Storage Abilities
2.4 Asphalt Mixture Performances
3 Conclusions and Prospects
References
Healing Characterisation of Waste-Derived Bitumen Based on Crack Length
1 Introduction
2 Model of Bitumen’s Healing Index
3 Materials and Experimental Protocol
4 Results and Discussion
4.1 Fundamental Properties of Healing Index Characterised by Crack Length
4.2 Enhancements of Healing Properties by Bio-Oil and LDPE
4.3 Healing Rate Modelling by Ramberg-Osgood Model
5 Summary and Conclusions
References
High Temperatures Characterization of Polymer Nanocomposite Modified Asphalt Binder
1 Background of the Study
2 Materials and Samples Preparation
2.1 Materials
2.2 Sample Preparation
3 Experimental Design and Methods
3.1 Rheological Properties
4 Results and Discussion
4.1 The Master Curve of Rheological Properties
4.2 Failure Temperature
4.3 Rutting Parameter
5 Conclusion
References
How to Improve the Miscibility of Asphalt Binder and Polyolefins by Phosphoric Acid
1 Introduction
2 Experimental Methods
2.1 Material and Samples
2.2 Microstructural Investigation
2.3 Rheological Measurements
References
Humidity Damage Index (HDI) of Recovered Asphalt from Reclaimed Asphalt Pavement-RAP Using Different Aggregates
1 Introduction
2 Literature Background
3 Materials and Methods
4 Results and Analysis
5 Conclusions
References
Impact of Different Fillers on the Properties of Asphalt Mixtures
1 Introduction
2 Materials and Methods
2.1 Materials
2.2 Filler Testing
2.3 Mastic Testing
2.4 Mixtures Testing
3 Discussion and Results
3.1 Stiffening Effect
3.2 Effect on Binder Ageing
3.3 Effect on the Resistance to Moisture Damage of Asphalt Mixtures
4 Conclusions
References
Impact of the Performance of the Asphalt Concrete and the Geogrid Materials on the Fatigue of Geogrid-Reinforced Asphalt Concrete—Experimental Study
1 Introduction
2 Key Points of Previous Research
3 Experimental Laboratory Test Program Design
4 Samples Preparation
5 Results and Discussion
6 Conclusion
References
Impact of Viscoelastic and Fatigue Behavior of Asphalt Mixtures Made with RAP and Asphalt Shingle on the Life Cycle Assessment Results
1 Introduction
2 Lab Results and Pavement Design
2.1 Lab Results
2.2 Pavement Design
3 Life Cycle Assessment
3.1 Introduction
3.2 LCA Results
4 Conclusion
References
Implementing Crumb Rubber Gap-Graded and Open Graded Asphalt Technologies—The Australian Experience
1 Introduction
2 Crumb Rubber Binder Blend Properties
3 Crumb Rubber Open Graded Asphalt—OGA10(CR)
4 Crumb Rubber Gap-Graded Asphalt—GGA14(CR)
5 Summary
References
Improved Testing Setup for Real-Time Monitoring of PMBs During Manufacturing and Rotational Viscosity Measurements
1 Introduction
2 Materials and Methods
2.1 Laboratory Simulation of PMBs Manufacturing
3 Results
4 Conclusions
References
Improving the Durability of Asphalt Mixtures with Hydrated Lime: Field Results from Recent French Sections
1 Introduction
2 Vosges Department Sections
2.1 Materials
2.2 Monitoring Methodology
2.3 Results
3 Highway A84 Sections
3.1 Materials
3.2 Monitoring Methodology
3.3 Results
4 Conclusions
References
Improving the Modulus Master Curve of Bituminous Mixes Using Ultrasonic Measurements
1 Introduction
2 Problem Statement and Objectives
3 Experimental Study
3.1 Materials
3.2 Complex Modulus Testing
3.3 Ultrasonic Waves Velocity Testing
4 Results and Discussions
4.1 Complex Modulus Through Ultrasonic Testing
4.2 Modulus Master Curves
5 Summary and Conclusions
References
Influence of Aggregate Gradation on the Permanent Deformation of Asphalt Mixtures
1 Introduction
2 Objectives
3 Materials and Methods
3.1 Materials
3.2 Mix Design
3.3 Laboratory Testing
4 Analysis
5 Conclusions
References
Influence of Binder Properties on Dynamic Shear Response of Asphalt Mixture
1 Introduction
2 Materials and Method
3 Results and Discussion
3.1 Dynamic Shear Test of Asphalt Mixture
3.2 DSR Test of Bitumen
3.3 Temperature Sweep of Bitumen
4 Summary and Future Perspective
References
Influence of Confinement Pressure on the Viscoelastic Response of Bituminous Mixtures
1 Introduction
2 Experimental Investigation
3 Results and Discussion
3.1 Test Temperature: 5 °C
3.2 Test Temperature: 35 °C
4 Conclusion
References
Influence of Linear Viscoelastic Behaviour of Pavement Layers Interface for Heavy Weight Deflectometer Test
1 Introduction
2 Heavy Weight Deflectometer Test Simulation
2.1 Studied Pavement Structure
2.2 Spectral Element Method Applied to Heavy Weight Deflectometer Test Simulation
2.3 Interface Viscoelastic Behaviour Characterization and Modelling
3 Results
4 Conclusion
References
Influence of Natural Aggregates’ Mineralogical Composition on the Adhesiveness and Affinity of Bitumen
1 Introduction
2 Materials
2.1 Natural Aggregates
2.2 Bituminous Binders
3 Results and Methodology
3.1 Adhesiveness
3.2 Affinity
4 Conclusions
References
In-Situ Cold Recycling Solutions for Roads Maintenance at the National Roads Network in Romania
1 Introduction
2 Materials
3 Laboratory Methods
4 Results
5 Discussion
6 Conclusion
References
In-Situ Measurements of Variations of RAP Contents in Hot Mix Asphalt by a Handheld FT-IR Spectrometer
1 Introduction
1.1 Objective
2 Materials and Experimental Plan
3 Experimental Method
3.1 Quick Binder Extraction Procedure
3.2 FT-IR Data Collection and Analysis
4 Results and Discussions
4.1 Change in Carbonyl Index with Increased RAP Content
4.2 Effect of Short-Term Aging in the Plant
4.3 Determination of RAP Content in Fresh Mix
5 Conclusion
References
Interlaminar Mode-I Fracture Characterization Underwater of Reinforced Bituminous Specimens
1 Introduction
2 Underwater WST and Specimen Characterization
3 Results and Discussion
4 Conclusion
References
Interphase Relations in the Characterisation of Bitumen Emulsion-Cement Composites
1 Introduction
2 Materials and Methods
2.1 Preparation and Testing of Specimens
3 Results
4 Conclusions
References
Introducing an Improved Testing Method to Evaluate the Fatigue Resistance of Bituminous Mortars
1 Introduction
2 Materials and Methods
2.1 Materials
2.2 Design Considerations and Modelling
2.3 Specimen Fabrication
2.4 Fatigue Testing
3 Results and Discussion
4 Conclusions
References
Investigating Fatigue Characteristics of Asphalt Binder Modified with Phase Change Materials Using Dynamic Shear Rheometer
1 Introduction
2 Sample Preparation
2.1 Asphalt Binder and PCM
2.2 Environmental Scanning Electron Microscopy
3 LAS Test
4 Result and Analysis
4.1 ESEM Observation
4.2 LAS Test Results
5 Conclusion
References
Investigating the Creep and Fatigue Properties of Recycled Asphalt Concrete Containing Waste Engine and Waste Cooking Oil
1 Introduction
2 Materials
3 Research Plan
4 Mixture Design, Fabrication of Specimens and Tests
5 Results and Discussions
6 Conclusions
References
Investigating the Nonlinear Behaviour of Neat and Modified Binders Through Large Amplitude Oscillatory Shear (LAOS) Testing
1 Introduction
2 Experimental Program
2.1 Materials Used
2.2 Linear and Nonlinear Oscillatory Shear Testing
3 Results and Discussions
3.1 Linear Viscoelastic Range Determination
3.2 Nonlinear Response Characterization
4 Conclusions
References
Investigating the Use of a Binder Cohesion Test to Evaluate Cracking Resistance of Asphalt Mixtures
1 Motivation, Background and Objective
2 Materials
2.1 Asphalt Mixture
2.2 Asphalt Binder
3 Asphalt Binder Evaluation
3.1 Poker-Chip Test
3.2 Measurement of the ΔTc Parameter
4 Asphalt Mixture Evaluation
4.1 Overlay and Indirect Tensile Strength Test
5 Results
5.1 Comparison of Binder and Asphalt Mixture Fatigue and Fracture
6 Conclusions and Discussion
References
Investigation of Cold Regions Dense Graded HMAC (EME) Sensitivity to Bitumen and Filler Content
1 Introduction
2 Experimental Program
3 Testing Samples and Methods
4 Tests Results and Analysis
4.1 Stiffness: Complex Modulus Test, E*
4.2 Low Temperature Resistance
4.3 Rutting Resistance
5 Conclusion
References
Investigation of Cracking Potential of Modified Asphalt Mixes Composed of Synthetic Fibers by Performing 4-point Bending Test
1 Introduction and Background
2 Objectives and Scope
3 Materials and Mix Design
3.1 Mix Design and Material Properties
3.2 Optimum Fiber Content
4 Performance Tests at Intermediate Temperature
4.1 Fatigue Test by Four-Point Bending Test
4.2 Fatigue Test – IDEAL CT Test
5 Comparative Analysis Between 4-Point Bending Test and IDEAL CT Index
6 Conclusions
References
Investigation of Different Design Methods for Determining the Appropriate Binder Ratio on Recycled Asphalt Mixtures
1 Introduction and Background
2 Materials and Methods
2.1 Asphalt Binders
2.2 RAP Characterization
2.3 Blending Study
3 Results and Discussion
4 Conclusions
References
Investigation of Fracture Behavior of Recycled Asphalt Mixtures Using a Discrete Element Computational Model
1 Introduction
2 Materials and Testing Methods
3 DEM Model
4 Simulation Results
5 Parametric Study
6 Conclusion
References
Investigation of the Performance of Warm Mix Asphalt Additives on 50/70 Penetrated Bitumen and Pre-cost Study
1 Introduction
2 Materials and Methods
3 Test Results
4 Pre-cost Study
5 Conclusion
References
Investigations of Electrical Conductivity of Graphite Nano-platelet (GNP)-Taconite Modified Asphalt Binders
1 Introduction
2 Experimental Investigation
3 Results and Discussion
4 Conclusion
References
Laboratory and in Situ Damage Evaluation of Geogrid Used in Asphalt Concrete Pavement
1 Introduction
2 Materials Used
2.1 Semi Coarse Asphalt Concrete Materials
2.2 Glass Fiber Grids
3 Laboratory Direct Tensile Tests on Recovered Grids from Slabs
4 Full Scale Test Section
5 Conclusion
References
Laboratory Assessment of Low Temperature Resistance of High Modulus Asphalt Concrete in Relation to Climatic Conditions in Poland
1 Introduction
2 Scope
3 Thermal Stress Restrained Specimen Test (TSRST)
4 Coefficient of Thermal Expansion
5 Indirect Tensile Strength
6 Review of the Tests Results
7 Assessment of the Risk of Pavement Cracking
8 Summary
References
Laboratory Comparison of In-Situ, Ex-Situ and Laboratory Produced Foamed Bitumen Stabilized Base
1 Introduction
2 Methods
3 Results and Discussion
4 Conclusion
References
Laboratory Evaluation of Rolling Resistance
1 Introduction
2 Test Method
2.1 Machine and Procedure
2.2 Materials and Samples
2.3 Characterizations
3 Results
3.1 Effects of Speed and Temperature
3.2 Relationship Rolling Resistance—Macrotexture
4 Modeling
5 Conclusions
References
Laboratory Investigation of Permanent Deformation of Modified Asphalt Mixes Using Nanocellulose
1 Introduction
1.1 Objectives
2 Materials and Methods
2.1 Asphalt Binder
2.2 Cellulose Nanocrystals
2.3 Research Methods
3 Results of Laboratory Testing and Discussion
3.1 Scanning Electron Microscopy and Rheology of the Asphalt Binder
3.2 Hamburg Wheel Tracking Test (Asphalt Mix)
3.3 Comparison of Rheology Test and HWD Test Results
4 Conclusions
References
Laboratory Performance of Stabilized Base with 100% Reclaimed Asphalt Pavement (RAP) Using Portland Cement, Bitumen Emulsion and Foamed-Bitumen
1 Introduction and Background
2 Materials
3 Testing, Results and Discussion
3.1 Resilient Modulus Test/MnPAVE Analysis
3.2 Semi-circular Bend Test
4 Conclusions and Future Extensions
References
Large Scale Characterization of Crumb Rubber Modified Asphalt Mixtures Using Dry Process
1 Introduction
2 Materials and Methods
3 Experimental Results
4 Conclusions
References
LCA of an Asphalt Mixture Produced in a Southern Italian Plant: Analysis and Perspectives
1 Introduction
2 Methodology and Selected Asphalt Production Plant
2.1 LCA
2.2 The Selected Plant
3 Results and Discussion
4 Conclusions
References
Life Cycle Assessment of Asphalt Pavements Using Crumb Rubber: A Comparative Analysis
1 Introduction
2 Materials
2.1 Asphalt Mixtures
2.2 Mechanical Performance of Asphalt Mixtures
3 Life Cycle Assessment
3.1 Goal and Scope Definition
3.2 Inventory Analysis
3.3 Impact Assessment
4 Results and Interpretation
5 Conclusion
References
Life Cycle Assessment of Self-healing Versus Traditional Maintenance Road Techniques
1 Introduction
1.1 Self-healing Asphalt Pavements and the Environment
2 Methodology
3 Results
4 Discussion and Conclusions
References
Long-Term Modelling of Composite Pavement Performance
1 Introduction
2 Objective
3 Reflective Cracking Data
4 Model Framework—Whole Life Long-Term Composite Pavement Model
5 Parametric Study
6 Simulations
7 Conclusions
References
Mechanical Assessment of Recycled Bituminous Mixtures
1 Introduction
2 Materials and Methods
3 Results and Discussion
4 Conclusions
References
Mechanical Behavior of Asphalt Mixtures Applied in Field in the Southern Brazil
1 Introduction
2 Materials and Experimental Plan
3 Results and Analysis
3.1 Linear Viscoelastic Characterization
3.2 Evaluation of Permanent Deformation
3.3 Fatigue Analysis
3.4 Fatigue Performance Evaluation Using LVECD Program Simulations
3.5 Global Resistance to Rutting and Fatigue
4 Conclusions and Perspectives
References
Mechanical Properties of Bitumen and Asphalt Mixture Modified with Polymer Additives
1 Introduction
2 Materials and Experimental Program
3 Test Results and Analysis
4 Conclusions
References
Methodology for Evaluating the Performance of Bituminous Binders Based on Rheological Indicators: Impact of the Use of a Rejuvenator
1 Introduction
2 Materials and Methods
2.1 Materials
2.2 Method
3 Test Results and Discussion
3.1 Discriminating Power of Rheological Indicators
3.2 Effect of Rejuvenator
4 Conclusions
References
Microwave Crack-Healing Capability on Asphalt Mixtures with Silicon Carbide
1 Introduction
2 Materials and Methods
2.1 Materials and Test Specimen Preparation
2.2 Crack-Healing of Asphalt Mixtures
3 Analysis of Results
4 Conclusions
References
Modeling of T/C Complex Stiffness Modulus Test and Non-linearity of Asphalt Concrete Mixes
1 Introduction
2 Viscoelasticity Modeling: VENoL Model
2.1 Existing Rheological Models: Advantages and Disadvantages
2.2 Proposal of a Different Design
3 Creation of the VENoL(T, ω) Model
3.1 Operating Mode of the VENoL(T, ω) Model
3.2 Example with MANGIAFICO’s Experimental Data (2014) [6]
4 Creation of the VENoL(T, ω, ε0) Model
4.1 Operating Mode of the VENoL(T, ω, ε0) Model
4.2 Example with GRAZIANI’s Experimental Data (2019) [7]
5 Conclusion
References
Modeling of T/C Fatigue Test with Boundary Element Method and Linear Fracture Mechanics
1 Introduction
2 Crack Propagation Simulation by MBEMv3.0
3 Fatigue Test on Cylindrical Samples
3.1 Model
3.2 Results
4 Conclusion
References
Modelling of Oxidative Ageing in Bitumen Using Thermodynamics of Irreversible Processes (TIP): Potential and Challenges
1 Introduction
2 Diffusion-Reaction Model for Spurt Oxidation
3 Compositions and Film Ageing Simulations
4 Results and Discussion
5 Conclusions
References
Monitoring of Railway Structures of High-Speed Line Bretagne-Pays de la Loire with Bituminous and Granular Sublayers
1 High-Speed Line BPL
2 Data Processing
3 Comparison of Response Between Granular and Bituminous Structures
3.1 Deflection Measurements
3.2 Acceleration Measurements
4 Conclusion
References
Monitoring Stiffness Evolution of Asphalt Concrete Through Modal Analysis
1 Introduction
2 Methodology
3 Results
4 Conclusions
References
Multi-criteria Framework to Evaluate the Oxidative Aging Resistance of Bitumen Binders
1 Introduction
2 Experimental Plan
3 Results and Discussion
4 Conclusion
References
Multiphysics Simulation and Validation of Field Ageing of Asphalt Pavements
1 Introduction
2 Development of Weak-Form Equation-Based Computational Models
3 Results
4 Conclusions
References
Multiscale Analysis of Pavement Texture—Effect on Skid Resistance
1 Introduction
2 Experimental Study
2.1 Polishing and Skid Resistance Measurements
2.2 Texture Measurement
3 Multiscale Analysis
4 Conclusions
References
Multi-scale Evaluation of Asphalt Binders Containing Different Proportions of Reclaimed Asphalt Pavement (RAP)
1 Introduction
2 Material
3 Traditional Method to Determine Modulus of Bulk Binders
4 Nano-scale Measurements of Binder Modulus
4.1 Linear Viscoelastic Modelling
5 Comparison of Moduli Obtained at Macro- and Nano- Scales
6 Conclusions
References
Novel Modified Recycled Mastic for Demanding and Sustainable Asphalt Mixtures
1 Introduction
2 Materials and Methods
2.1 Materials
2.2 Methods
3 Results and Discussion
3.1 Affinity Between Aggregate and Binder
3.2 Thermogravimetry Analysis (TGA)
3.3 Dynamic Shear Rheometry (DSR)
3.4 Wheel Tracking Test (WTT)
3.5 Stiffness Modulus and Phase Angles
3.6 Fatigue Resistance
3.7 Other Tests
4 Conclusions
References
Numerical Simulation of Two-Point and Four-Point Bending Fatigue Tests on Asphalt Mix Using Intrinsic Material Parameters Calibrated from Digital Image Correlation Data
1 Introduction
2 Equations of the Model
3 Simulation of 4PB and 2PB Fatigue Tests Using the Same Model Parameters
4 PB Fatigue Tests with DIC
5 Simulations Considering the Actual Crack Paths: Effect on Fatigue Curves
6 Conclusion
References
On the Assessment and Optimisation of the Processing Conditions of Tyre-Rubber Modified Bitumen
1 Introduction
2 Methodology and Properties of Materials
3 Test Results and Discussion
4 Conclusions
References
Organic Compounds Evaluation from Fumes Generated in Laboratory by Bio-recycled Asphalt Mixtures
1 Background
2 Experimental Set
2.1 Bituminous Mixture Device and Fume Generation, Sampling and Analysis
2.2 Experimental Protocol
3 Experimental Program
4 Conclusions
References
Parametric Study of Binder Properties on Thermomechanical Performance of Bituminous Mixtures
1 Introduction
2 Materials
3 Thermomechanical Performance of Bituminous Mixtures
4 Results and Discussion
5 Conclusions
References
Pavement Evaluation Method Using MMS
1 Introduction
2 Road Pavement Condition Surveys Using Point Cloud
2.1 Rutting
2.2 Flatness
2.3 Potholes
2.4 Crack Percentage
3 Accuracy Verification
3.1 Accuracy Verification of Rutting, Flatness, Crack Percentage
3.2 Accuracy Verification of Potholes
4 Installation in the System
5 Future Prospects
5.1 Damage Prediction of Road Surface Properties
5.2 Structural Health Prediction
6 Conclusions
References
Performance Testing of Hot Mix Asphalt Containing Biochar
1 Introduction
2 Materials
3 Sample Preparation and Test Method
4 Results, Statistical Analysis and Discussion
5 Conclusion
References
Predicting the Master Curve of Bituminous Mastics with Micromechanical Modelling
1 Introduction
2 Methods
3 Results and Discussion
4 Conclusions
References
Predicting the Potential Impact of Geopolymers on the Creep Recovery Properties of Asphalt Binder
1 Introduction
2 Methodology
2.1 Materials Preparation
2.2 Dynamic Shear Rheometer (DSR)
3 Results and Discussion
3.1 Effect of Geopolymer on the Asphalt Binder Properties
3.2 Effect of Geopolymer on the Recovered Strain of Asphalt Binder
4 Conclusions
References
Prediction of Fatigue Behaviour of Asphalt Mix from Tests on Asphalt Mastic
1 Introduction
2 Materials
3 Test-Method
3.1 Asphalt Fatigue—4 Point Bending Beam (4PB)
3.2 Mastic Fatigue with Hyperbolic Specimen Shape
4 Results
5 Conclusion
References
Preliminary Investigation of Using Nanocellulose in Bituminous Materials
1 Introduction
2 Methodology
2.1 Sample Preparation
2.2 Binder Characterization
3 Results and Discussion
3.1 Performance Grade Testing
3.2 Linear Amplitude Sweep Test (LAS)
3.3 Temperature-Frequency Sweep Master Curves
4 Conclusion
References
Properties at Low Temperatures of Warm Mix Containing High Content of Multi-recycled RAP
1 Introduction
2 Materials and Samples
3 Thermal Stress Restrained Specimen Test (TSRST)
3.1 Testing Device
3.2 Testing Procedure and Properties at Low Temperature
4 RAP, Recycling Cycle and Air Void Content Influence
5 Conclusion
References
Proposal for Aging Quantity Model for Estimating Binder Aging Level of HMA Mix
1 Introduction
2 Experimental Program
2.1 Materials
2.2 Methods
3 Results and Discussions
4 Conclusions
References
Quantification of Compactability of Bituminous Mixtures with Different Aggregate Gradations Using Superpave Gyratory Compactor and Shear Box Compactor
1 Introduction
2 Materials and Sample Preparation
3 Compactability Indices Based on SBC and SGC
3.1 Normalized Height Versus the Compaction Cycles
3.2 %Gmm Versus the Compaction Cycles
3.3 Construction Densification Index (CDI) and Traffic Densification Index (TDI)
4 Laboratory Rutting Results
5 Conclusions
References
Reduction of Bitumen Content and Production Temperature of Hot-Mix Asphalt Incorporating RAP Using Dune Sand and Lime
1 Introduction
2 Materials and Experimental Procedures
2.1 Material Properties
2.2 Mix Design
3 Results and Discussion
3.1 Effects on the Bitumen Content
3.2 Effect of the Decrease of the Manufacturing Temperature
4 Conclusion
References
Relationship Between Fatigue Damage of Asphalt Binders and Corresponding Mixtures
1 Introduction
2 Materials and Methods
3 Asphalt Binders and Mixtures Fatigue Damage Characterization
4 Fatigue Life Simulation of Asphalt Binders and Mixtures and Relationship Between Their Fatigue Life
5 Conclusions
References
Relationship Between Linear Viscoelastic Properties of Asphalt Binders and Corresponding Mixtures
1 Introduction
2 Materials and Methods
3 Analysis and Discussions
4 Conclusions
References
Relaxation Spectrum: Why It Matters and How to Correctly Develop One?
1 Introduction
2 Theoretical Background
3 A New Determination of Relaxation Spectrum
4 Experimental Study
5 Conclusion
References
Rheological Characterisation of Modified Bitumens with Biodiesel-Derived Biobinders
1 Introduction
2 Experimental Programme
2.1 Materials
2.2 Test Methods
3 Results and Discussion
3.1 Rheological Master Curves
3.2 Changes in Rheological Properties After Ageing
4 Conclusions
References
Rheological Investigation on the Rutting Characteristics of Nanoclay Modified Asphalt Binders
1 Introduction
2 Materials and Sample Fabrication
3 Experimental Plan
4 Results and Discussion
4.1 Discussion on Master Curve
4.2 Determination of h parameter
4.3 Determination of R and Jnr
5 Conclusion
References
Rheological Modelling of the Bitumen from Reclaimed Asphalt with Rejuvenation and Re-ageing
1 Introduction
2 Experimental Method
2.1 Materials and Specimen Preparation
2.2 Complex Modulus Measurement Procedure
2.3 Rheological Modelling and Delta-Method
3 Results and Analysis
3.1 Effect of Rejuvenation and Re-ageing on |G*| and δ Master Curves
3.2 Evolution of 2S2P1D Model Parameters with Rejuvenation and Re-ageing
3.3 Variation of AMW Distributions with Rejuvenation and Re-ageing
4 Conclusion
References
Rheological Properties and 2S2P1D Modeling of Seven Bituminous Mixtures with Different Brazilian Bitumens
1 Introduction
2 Materials and Experimental Plan
2.1 Materials
2.2 Experimental Procedures
3 Results and Analysis
3.1 Linear Viscoelastic Characterization
4 Conclusions
References
Rheological Properties and Rutting Characterization of Natural Rubber Modified Bitumen
1 Introduction
2 Materials and Methods
2.1 Materials
2.2 Bitumen Blending
2.3 Bitumen Short-Term Ageing
2.4 Testing Protocol
2.5 Test Methods
3 Results and Discussion
3.1 Effects of NR on the Rheological Properties of NRMB
3.2 Effects of NR on the Rutting Resistance of NRMB
4 Conclusions
References
Rheological Properties of Asphalt Binder Compound Modified by Bio-oil and Organic Montmorillonite
1 Introduction
2 Materials and Methods
2.1 Materials
2.2 Testing Methods
3 Results and Discussions
3.1 Linear Viscoelastic Properties
3.2 Rutting Resistance
3.3 Fatigue Performance
4 Conclusions
References
Rheological Properties of Bituminous Mastic with Different Filler Content
1 Introduction
2 Asphalt Mastic Preparation
3 Testing Results and Discussion
3.1 Penetration, Ductility and Softening Point
3.2 Brookfield Viscosity
3.3 DSR Testing Results
4 Conclusion
References
Rheological Properties of Derived Fractions Composed of Aromatics, Resins, and Asphaltenes
1 Introduction
2 Materials and Test Design
2.1 Materials
2.2 Separation and Derivation Process
2.3 Rheological Tests
3 Results
4 Conclusions
References
Rheometrics Framework Analysis to Capture Self-assemblies Organization in Bitumen Matrix
1 Introduction
2 Experimental Section
3 Results and Discussion
3.1 Liquid to Solid Transition
3.2 Thermal Hysteresis
3.3 Mechanical Properties
4 Conclusions
References
Roles of Exogenous Cationic and Endogenous Bitumen Surfactants in the Stability of Bitumen/Water Interfaces
1 Introduction
2 Materials and Experimental Methods
2.1 Materials
2.2 Experiments
3 Results and Discussion
4 Conclusion
References
Rubber-Bitumen Interaction of Plant-Blended Rubberized Bitumen Prepared Under Various Blending Conditions
1 Introduction
2 Experimental
2.1 Source of Raw Materials
2.2 Test Methodology
3 Results and Discussion
3.1 Changes in Molecular Structure of Rubberized Bitumen
3.2 Workability of Rubberized Bitumen
3.3 Storage Stability of Rubberized Bitumen
3.4 Rheological Properties of Rubberized Bitumen at High Temperatures
4 Conclusions
References
Rutting Performance Analysis for Pavements with Bituminous Stabilized Mixtures as Base Layers
1 Introduction
2 Materials and Methods
2.1 Constitutive Models for BSM
2.2 Reconstruction Structural Solutions
2.3 Multilayer Pavement Model
3 Results
4 Conclusions
References
Sensitivity Analysis for Rheological Determination of Glass Transition and Crossover Temperatures at Various Reference Frequencies
1 Introduction
2 Experimental Plan
3 Results and Discussion
4 Conclusion
References
Shear Fatigue Cracking Analysis of Pavement with Thick Asphalt Layers Induced by Traffic Load
1 Introduction
2 Pavement Structures
3 Stress Intensity Factor Analysis of Different Pavement Structures
3.1 Finite Element Model
3.2 Material Parameters
3.3 Shear Stress Intensity Factor of Different Pavement Structures
4 Shear Fatigue Cracking Parameters Based on Shear Fatigue Test
4.1 Shear Fatigue Test
4.2 Paris Formula Parameters
5 Shear Cracking Life Prediction of Thick Asphalt Layers
6 Conclusion
References
Solar and Permeable Road: A Prototypical Study
1 Introduction
2 Design of Surface Properties
3 Design of Surface Properties
4 Conclusions
References
Statistical Analysis of the AIGLE 3D Crossover Trial Campaign: Experimental Design, Methods and Results
1 Introduction
2 Crossover Trial Campaign
3 Method of Analysis
3.1 Analysis Area
3.2 Studied Data
4 Results
4.1 On Analysis Area
4.2 On Studied Degradations and Deformations
5 Conclusions
References
Study of the Shear Stiffness of Bituminous Mastics Incorporating Recycled Glass Fillers
1 Introduction
2 Methods and Materials
2.1 The Annular Shear Rheometer ASR
2.2 Materials
3 Results and Discussions
3.1 Influence of the Nature of the Filler
3.2 Influence of the Gradation Size Curve
4 Conclusions
References
Study on Aging Resistance of Bitumen Rejuvenated with Various Rejuvenators for Hot Recycling
1 Introduction
2 Materials
3 Methodology
4 Target Dosage Relations
5 Ageing Index
6 Conclusions
References
Study on Reversibility of Aging and Recycling of Bitumen Based on Rheology, Adhesion and Chemical Properties
1 Introduction
2 Materials and Methods
2.1 Materials
2.2 Methods
3 Results and Discussions
3.1 Effect of Aging and Recycling on Rheological Properties of Bitumen
3.2 Analysis of Adhesion Work Between Bitumen and Aggregate
3.3 The Results of FTIR
4 Conclusions
References
Superpave 5: Improving Asphalt Mixture Performance
1 Introduction
2 Mixtures and Materials
3 Test Methods
4 Results and Discussion
5 Summary and Conclusion
References
Surface Dressing Treatment for Applications on Solar Roads
1 Introduction
2 The Surface Dressing Treatment
3 Determination of the Optimum Glue Volume
4 Mechanical and Optical Performance
5 Conclusions and Next Steps
References
Sustainability Assessment of Novel Performance Enhancing Chemical Bitumen Additive
1 Introduction
2 Service Life Prediction
2.1 Material Characterization of Base Layers
2.2 Traffic and Climatic Loads
2.3 Results of the Service Life Prediction
3 Sustainability Assessment
3.1 Methodology
3.2 Results of the Sustainability Assessment
4 Conclusions
References
Temperature and Loading Rate Susceptibility of Bituminous Mixtures on Monotonic Testing
1 Introduction
2 Methods and Materials
2.1 Fenix Test
2.2 Materials
2.3 Test Plan
3 Results
4 Conclusions
References
Temperature Dependency of the Stiffening Effect of Hydrated Lime in Stone Mastic Asphalt (SMA) Mixtures
1 Introduction
2 Materials and Methods
2.1 Materials
2.2 SMA Mixture Design
2.3 Determination of Stiffness Modulus
3 Test Results and Discussion
4 Conclusions
References
Testing of Reclaimed Asphalt Model Systems for the Evaluation of the Effectiveness of Rejuvenators
1 Introduction
2 Materials and Methods
3 Results and Discussion
4 Conclusions
References
The Bitumen Related Problems Observed on Soft Asphalt Concrete Pavements: Case Study
1 The Peculiar Case in Southwest Finland
2 Materials and Methods
3 Results
4 Conclusions
References
The Chemistry Behind Rheological and Thermal Transitions of Oxidized Bitumen
1 Introduction
2 Experimental Design
3 Results and Discussion
3.1 Oxidative Aging of Asphalt Binders
3.2 Relationship Between Oxidation and Crossover Temperature
3.3 Overview of Thermal Transitions and Their Relationship with Rheology and Aging
4 Conclusions
References
The Effect of a Chemical Warm Mix Additive on the Self-Healing Capability of Bitumen
1 Introduction
2 Methodology
2.1 Sample Preparation and Characterization
2.2 Fatigue and Healing Characterization
3 Results and Discussion
3.1 The Effects of Aging
3.2 Fatigue and Healing Characterization
4 Conclusions
References
The Effect of Additives on Water Vapor Condensation on Bituminous Surfaces
1 Introduction
2 Materials and Methods
2.1 Bitumen Selection and Sample Preparation
2.2 Experimental Setup
2.3 Experimental Protocol
3 Results and Discussion
4 Conclusion
References
The Effect of Mix Constituents in the Permanent Deformation Resistance of Asphalt Mixtures
1 Introduction
2 Experimental Design
3 Results and Discussion
4 Summary and Conclusions
References
The Heating and Compaction of Asphalt Mixtures as a Self-healing Mechanism
1 Introduction
2 Materials and Methods
3 Results Analysis
3.1 Complex Modulus
3.2 Fatigue Life
3.3 Healing Potential Index (HPI)
4 Conclusions
References
The Impact of Binder Type on Pavement Design: A Comparative Study
1 Introduction
2 Experimental Study
3 Data Analysis and Results
4 Conclusions
References
The Influence of Cement Type on Early Properties of Cold In-Place Recycled Mixtures
1 Introduction
2 Materials and Methods
3 Results and Discussion
4 Conclusions
References
The Influence of the Mastic Coating of Untreated Reclaimed Asphalt Pavement on the Permanent and Resilient Behaviours
1 Introduction
2 Material and Test Programme
2.1 Material Characterisation
2.2 Sampling
2.3 Triaxial Apparatus
2.4 Repeated Load Triaxial (RLT) Tests
3 Experimental Results and Analysis
3.1 Conditioning Phase
3.2 Resilient Test
4 Conclusions
References
The Influence of Wax Model Compounds on the Surface Topography of Bitumen
1 Introduction
2 Experimental
2.1 Materials
2.2 AFM Setup
2.3 DSC Tests
3 Results and Discussion
4 Conclusions
References
The Low Temperature Performance of Rubberized Bituminous Mixture Using the Dry Process
1 Introduction
2 Materials and Test Methods
2.1 Material
2.2 Test Methods
3 Test Result Analysis
4 Conclusions and Recommendations
References
The Use of Direct Shear Test for Optimization of Interlayer Bonding Under a Poroelastic Layer
1 Introduction
2 Materials and Methods
3 Results and Discussion
4 Conclusions
References
Time and Storage Dependent Effects of Bitumen—Comparison of Surface and Bulk
1 Introduction
2 Materials and Methods
2.1 Material
2.2 Methods
3 Results and Discussion
3.1 FTIR Spectroscopy
3.2 DSR
4 Conclusion
References
Top-Down Cracking in Airfield Pavement Structures—An In-Situ Test Survey
1 Introduction
2 Investigation Program
3 Analysis
4 Modelling
5 Conclusion
References
Tracking Degree of Blending Between Recycled and Virgin Binder Through Asphalt Mix Phase Angle
1 Introduction
2 Materials and Methods
3 Results and Discussion
4 Conclusion
References
TSRST Heterogeneous Modeling for Understanding Failure Depending on Asphalt Mix Design and Experimental Conditions
1 Introduction
2 Numerical Approach
2.1 Material
2.2 Random Generation of Aggregate Skeleton
2.3 Thermal and Mechanical Properties
2.4 Mesh, Boundary Conditions and Loading
3 Results
4 Conclusion
References
Understanding of Structural and Surface Tension Properties of Asphalt Model Using Molecular Dynamics Simulation
1 Background
2 Molecular Dynamics and Experimental Program
2.1 Molecular Dynamics Details
2.2 Surface Tension Measurement with the Pendant Drop Method
3 Results and Discussion
3.1 Asphalt Density
3.2 Surface Tension
3.3 Molecular Structural Properties
4 Summary and Conclusions
References
Use of Modified Reclaimed Asphalt in Warm Mixtures
1 Introduction
2 Experimental Program
3 Results and Discussions
3.1 Indirect Tensile Strength (ITS) Test and Water Sensitivity
3.2 Stiffness
3.3 Rutting Performance
3.4 Fatigue Performance
4 Conclusions
References
Use of Multiple Stress Creep Recovery Test to Evaluate Viscoelastic Properties of Asphalt Mastic
1 Introduction
2 Material Preparation
3 Experiment Procedures and Rheological Model for Mastic Asphalt
3.1 Multiple Stress Creep Recovery Test
3.2 Burgers Model
4 Experimental Results and Analysis
4.1 Results of MSCR Test for Asphalt Mastics with Different Filler Contents
4.2 Results and Analysis on Model Fit Using the Burgers Model
4.3 Analysis on Correlation of MSCR Test Observations and Burgers Model Parameters
5 Conclusion
References
Use of Recycled Asphalt Pavement (RAP) in Airport Pavements
1 Introduction
2 Design and Construction of Full-Scale Test Lanes
2.1 Sequence of Asphalt Concrete Paving
3 Pavement and Material Characterization Tests
3.1 Uniformity Tests with Heavy Weight Deflectometer
3.2 Laboratory Characterization Tests
4 Summary
References
Using Linear Elastic Layer Analysis Based Method for Back-Calculating the Layer Moduli of Small Paneled Thin White Topping Pavements
1 Introduction
2 Literature
3 Objective and Scope
4 Methodology Adopted
5 Observations and Discussion
5.1 KGPBACK Software for Back-Calculation
6 Conclusions
References
Validation of the Time-Temperature Superposition Principle (TTSP) in the Non-linear Domain for Bituminous Mixtures with Reclaimed Asphalt Pavement (40% RAP)
1 Introduction
2 Material and Test Device
2.1 Description of the Tests
2.2 Material and Sample Preparation
3 Linear Domain Investigation: Linear ViscoElastic (LVE) Behaviour
3.1 Complex Modulus Test
3.2 Experimental Results
4 Non-linear Domain Investigation: Non-linear ViscoElastic (NLVE) Behaviour
4.1 Direct Tensile Test
4.2 Experimental Results
5 Conclusion
References
Viennese Aging Procedure (VAPro): Adaption for Low-Temperature Testing
1 Introduction
2 Material and Methods
2.1 Material
2.2 Methods
3 Results and Discussion
4 Summary and Outlook
References
Viscoelastic Analysis of Top-Down Crack in Geosynthetic Reinforced Asphalt Pavements Using FEM
1 Introduction
2 Material Properties and Modeling in Abaqus
3 Results and Discussion
4 Conclusion
References
Correction to: Proceedings of the RILEM International Symposium on Bituminous Materials
Correction to: H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4
Recommend Papers
![Fibre Reinforced Concrete: Improvements and Innovations: RILEM-fib International Symposium on FRC (BEFIB) in 2020 [1st ed.]
9783030584818, 9783030584825](https://ebin.pub/img/200x200/fibre-reinforced-concrete-improvements-and-innovations-rilem-fib-international-symposium-on-frc-befib-in-2020-1st-ed-9783030584818-9783030584825.jpg)
File loading please wait...
Citation preview
RILEM Bookseries
Hervé Di Benedetto · Hassan Baaj · Emmanuel Chailleux · Gabriele Tebaldi · Cédric Sauzéat · Salvatore Mangiafico Editors
Proceedings of the RILEM International Symposium on Bituminous Materials ISBM Lyon 2020
Proceedings of the RILEM International Symposium on Bituminous Materials
RILEM BOOKSERIES
Volume 27
RILEM, The International Union of Laboratories and Experts in Construction Materials, Systems and Structures, founded in 1947, is a non-governmental scientific association whose goal is to contribute to progress in the construction sciences, techniques and industries, essentially by means of the communication it fosters between research and practice. RILEM’s focus is on construction materials and their use in building and civil engineering structures, covering all phases of the building process from manufacture to use and recycling of materials. More information on RILEM and its previous publications can be found on www.RILEM.net. Indexed in SCOPUS, Google Scholar and SpringerLink.
More information about this series at https://link.springer.com/bookseries/8781
Hervé Di Benedetto · Hassan Baaj · Emmanuel Chailleux · Gabriele Tebaldi · Cédric Sauzéat · Salvatore Mangiafico Editors
Proceedings of the RILEM International Symposium on Bituminous Materials ISBM Lyon 2020
Editors Hervé Di Benedetto University of Lyon/ENTPE Vaulx-en-Velin, France
Hassan Baaj University of Waterloo Waterloo, ON, Canada
Emmanuel Chailleux Gustave Eiffel University Champs-sur-Marne, France
Gabriele Tebaldi University of Parma Parma, Italy
Cédric Sauzéat University of Lyon/ENTPE Vaulx-en-Velin, France
Salvatore Mangiafico University of Lyon/ENTPE Vaulx-en-Velin, France
ISSN 2211-0844 ISSN 2211-0852 (electronic) RILEM Bookseries ISBN 978-3-030-46454-7 ISBN 978-3-030-46455-4 (eBook) https://doi.org/10.1007/978-3-030-46455-4 © RILEM 2022, corrected publication 2022 No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Permission for use must always be obtained from the owner of the copyright: RILEM. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
The proceedings of the RILEM International Symposium on Bituminous Materials, which took place during December 14–16 in Lyon, France (ISBM Lyon 2020), gathered 246 scientific and technical papers from all over the world. In recent years, an increasing number of RILEM Technical Committees (TCs) have been established within Cluster F (“Bituminous Materials and Polymers”) in order to tackle key scientific challenges related to properties of bituminous materials and structures as well as innovative practices in road, railway and airfield infrastructures. RILEM ISBM Lyon 2020 was the first joint event of three RILEM Technical Committees of Cluster F: • 264-RAP (“Asphalt Pavement Recycling”) • 272-PIM (“Phase and Interphase behaviour of bituminous Materials”) • 278-CHA (“Crack-Healing of Asphalt Pavement Materials”) The research works, which were presented at the symposium and collected in this volume, span across a very large range of topics in the field. They offer a state-of-theart picture of the latest advancements on sustainable infrastructures. Among others, the key topics are • • • • • • • • • • • •
Recycling Phase and interphase behaviour Cracking and healing Modification and innovative materials Durability and environmental aspects Testing (laboratory, in situ, accelerated) and modelling Multi-scale properties Surface characteristics Structure performance, modelling and design Non-destructive testing Back analysis Life cycle assessment v
vi
Preface
All papers accepted for presentation at the conference and reported in these proceedings were reviewed and carefully selected by the Scientific Committee. The significant amount of contributions is a reflection of the richness of scientific and technical work within the domain. The subjects of the articles range from fundamental research on rheology and material damage to applied research on innovative materials and modern technologies in pavement engineering. This incredibly wide spectrum of academic publications demonstrates the flourishing activities of RILEM Cluster F. For several decades, RILEM has been successfully fostering international collaborations between academia and industry, resulting in cutting-edge scientific research and high-impact advancements. This vital network between the main actors of scientific and technological development is paramount for achieving environmental, economic and social sustainability of infrastructures and of civil engineering in general. Furthermore, the scientific support of three internationally renowned societies, International Society for Asphalt Pavements (ISAP), European Asphalt Technology Association (EATA) and ASTM International, helped RILEM ISBM Lyon 2020 reach its objectives in terms of scientific rigour and impact. We thank all authors, presenters, participants and members of the Scientific and Organizing Committees for their precious contribution to the success of RILEM ISBM Lyon 2020 as well as Springer Nature for the edition of these proceedings. The conference delivery and the operation of some ancillary activities had to be modified in order to comply with the challenging situation related to the COVID-19 crisis. Finally, we would like to acknowledge the great support of our sponsors. Vaulx-en-Velin, France Vaulx-en-Velin, France Champs-sur-Marne, France Vaulx-en-Velin, France Waterloo, Canada Parma, Italy
Hervé Di Benedetto Salvatore Mangiafico Emmanuel Chailleux Cédric Sauzéat Hassan Baaj Gabriele Tebaldi
Logos and Sponsors
Sponsors
vii
viii
Logos and Sponsors
Platinum Sponsors
Gold Sponsors
Logos and Sponsors
Silver Sponsors
ix
x
Logos and Sponsors
Institutional Sponsors
This symposium benefited from the financial support of Projet IDEXLYON of Université de Lyon in the framework of the Programme Investissements d’Avenir (ANR-16-IDEX-0005), as well as from the support of the Greater Lyon and of the Auvergne-Rhône-Alpes Region.
Contents
Technical Committee 264-RAP (“Asphalt Pavement Recycling”) Effect of Aging on the Rheological Properties of Blends of Virgin and Rejuvenated RA Binders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Di Wang, Augusto Cannone Falchetto, Martin Hugener, Laurent Porot, Atsushi Kawakami, Bernhard Hofko, Andrea Grilli, Emiliano Pasquini, Marco Pasetto, Hassan Tabatabaee, Huachun Zhai, Margarida Sá da Costa, Hilde Soenen, Patricia Kara De Maeijer, Wim Van den Bergh, Fabrizio Cardone, Alan Carter, Kamilla L. Vasconcelos, Xavier Carbonneau, Aurelie Lorserie, Goran Mladenovi´c, Marko Oreškovi´c, Tomas Koudelka, Pavel Coufalik, Edoardo Bocci, Runhua Zhang, Eshan V. Dave, and Gabriele Tebaldi Experimental Investigation on Water Loss and Stiffness of CBTM Using Different RA Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andrea Grilli, Simone Raschia, Daniel Perraton, Alan Carter, Amir Rahmanbeiki, Patricia Kara De Maeijer, Davide Lo Presti, Gordon Airey, Chibuike Ogbo, Eshan V. Dave, and Gabriele Tebaldi International Evaluation of the Performance of Warm Mix Asphalts with High Reclaimed Asphalt Content . . . . . . . . . . . . . . . . . . . . . Paul Marsac, Edoardo Bocci, Fabrizio Cardone, Augusto Cannone Falchetto, Xavier Carbonneau, Martins Zaumanis, Alan Carter, Ma Carmen Rubio-Gámez, Miguel Del Sol Sánchez, Eshan V. Dave, and Gabriele Tebaldi
3
11
19
xi
xii
Contents
Technical Committee 272-PIM (“Phase and Interphase Behaviour of Bituminous Materials”) Interlaboratory Test to Characterize the Cyclic Behavior of Bituminous Interlayers: An Overview of Testing Equipment and Protocols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Francesco Canestrari, Thomas Attia, Hervé Di Benedetto, Andrea Graziani, Piotr Jaskula, Youngsoo Richard Kim, Maciej Maliszewski, Jorge C. Pais, Christophe Petit, Christiane Raab, Davide Ragni, Dawid Rys, Cesare Sangiorgi, Cédric Sauzéat, and Adam Zofka Complex Bituminous Binders, Are Current Test Methods Suitable for? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Laurent Porot, Emmanuel Chailleux, Panos Apostolidis, Jiqing Zhu, Alexandros Margaritis, and Lucia Tsantilis
29
37
Technical Committee 278-CHA (“Crack-Healing of Asphalt Pavement Materials”) Terms and Definitions on Crack-Healing and Restoration of Mechanical Properties in Bituminous Materials . . . . . . . . . . . . . . . . . . . Greet Leegwater, Amir Tabokovi´c, Orazio Baglieri, Ferhat Hammoum, and Hassan Baaj Testing Methods to Assess Healing Potential of Bituminous Binders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Orazio Baglieri, Hassan Baaj, Francesco Canestrari, Chao Wang, Ferhat Hammoum, Lucia Tsantilis, and Fabrizio Cardone
47
55
Other Conference Papers A Comparative Study of RTFO, PAV and UV Aging Using FT-IRS and DSR Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Roksana Hossain, Shams Arafat, Delmar Salomon, and Nazimuddin M. Wasiuddin
65
A Critical Analysis of the Standard Used to Evaluate De-icing Damage in Asphalt Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lisa Lövqvist, Jiqing Zhu, Romain Balieu, and Nicole Kringos
73
A Laboratory Procedure for Improving the Design of Road Rehabilitation Actions: Study of a Real Case in a Highway Pavement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fernando Moreno-Navarro, J. Sierra, M. Sol-Sánchez, and Ma Carmen Rubio-Gámez
81
Contents
A Literature Review of Bitumen Aging: From Laboratory Procedures to Field Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rodrigo Siroma, Mai Lan Nguyen, Pierre Hornych, and Emmanuel Chailleux
xiii
89
A Method to Reduce Occlusion While Measuring Pavement Surface Profiles Using Triangulation Based Laser Scanners . . . . . . . . . . Subham Jain, Animesh Das, and K. S. Venkatesh
97
A New Approach to Calibration and Use of Mechanistic-Empirical Design Methods . . . . . . . . . . . . . . . . . . . . . . . . . . Rongzong Wu, John T. Harvey, and Jeremy Lea
103
A New Approach Towards Scientific Evaluation and Performance-Related Design of Bituminous Joint Sealing Materials and Constructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . J. Buchheim, Ch. Recknagel, P. Wolter, and K. Kittler-Packmor A New Bituminous Ballast-Less and Sleeper-Less Thin Railway Track Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Diego Ramirez Cardona, Simon Pouget, G. van der Houwen, M. Vlak, S. Sénéchal, and N. Calon A New Method for Healing Quantification of Bituminous Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Remya Varma, Romain Balieu, and Nicole Kringos A Numerical Model to Predict the Thermo-Viscoelastic Behaviour of Asphalt Concrete for Electric Road System . . . . . . . . . . . . . Talita de Freitas Alves, Thomas Gabet, Jean-Michel Simonin, and Ferhat Hammoum A Numerical Study of the Influence of Aggregates Shape on Creep and Recovery Tests Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Marcone de Oliveira Junior, Márcio Muniz de Farias, and Carlos Alexander Recarey Morfa A Simpler Method for Establishing an Approximate Value of Built-in Temperature Differential in Concrete Pavement . . . . . . . . . . . Venkata Joga Rao Bulusu, Sudhakar Reddy Kusam, and Amarnatha Reddy Muppireddy A Study on Fatigue Behaviour of Bitumen Emulsion Mastic, Modified with Active Fillers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ahmed Al-Mohammedawi and Konrad Mollenhauer A Study on the Properties of Polyurethane-Elastomer-Modified Asphalt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Peng Wang, Jian Xu, Yongchun Qin, Erhu Yan, Jie Wang, and Weiying Wang
111
119
127
135
143
151
159
167
xiv
Contents
A Wear-Based Description of Mechanisms Underlying the Bitumen Removal Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Minh-Tan Do, Veronique Cerezo, and Christophe Ropert
175
Active Mitigation of Low-Temperature Cracking in Asphalt Pavements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Quentin Adam, Gerald Englmair, Eyal Levenberg, and Asmus Skar
183
Ageing Behavior of Porous and Dense Asphalt Mixtures in the Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ruxin Jing, Aikaterini Varveri, Xueyan Liu, Athanasios Scarpas, and Sandra Erkens An Attempt to Characterize the RAP Material for Hot Recycled Mix Design Purposes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gurunath Guduru and Kranthi Kuna An Attempt to Distinguish Thermal from Oxidative Ageing of Asphalt Binders by NRTFOT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ingrid Camargo, Bernhard Hofko, Johannes Mirwald, and Hinrich Grothe An Interlaboratory Test Program on the Extensive Use of Waste Aggregates in Asphalt Mixtures: Preliminary Steps . . . . . . . . . . . . . . . . . . Marco Pasetto, Emiliano Pasquini, Giovanni Giacomello, Fernando Moreno-Navarro, Raul Tauste-Martinez, Augusto Cannone Falchetto, Michel Vaillancourt, Alan Carter, Nunzio Viscione, Francesca Russo, Marta Skaf, Marko Oreškovi´c, Ana Cristina Freire, Peter Mikhailenko, and Lily Poulikakos An Investigation on Longitudinal Unevenness Indicators and Their Potential on Surface Characterisation . . . . . . . . . . . . . . . . . . . . Parisa Rossel-Khavassefat, Jacques Perret, Mehdi Ould-Henia, and Marc Delaby An Overview of Black Space Evaluation of Performance and Distress Mechanisms in Asphalt Materials . . . . . . . . . . . . . . . . . . . . . . Gordon Airey, Jo E. Sias, Geoffrey M. Rowe, Hervé Di Benedetto, Cédric Sauzéat, and Eshan V. Dave Analysis of a New Approach for Characterizing the Performance of Asphalt Binders Through the Multiple Stress Creep and Recovery Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aline Vale, Juceline Bastos, and Jorge B. Soares Analysis of High Temperature Performance of Composite Asphalt Rubber Mixture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Haibin Li, Mingming Zhang, Ahmed Abdulakeem Temitope, Hongjun Jing, Guijuan Zhao, and Qingwei Ma
191
199
207
215
223
231
239
247
Contents
Analysis of Low Temperature Cracking Behavior at Binder, Mastic and Asphalt Concrete Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Christiane Raab, Martin Arraigada, and Hamza Ibrahimi Analysis of Ovalization Measurements on Flexible Airfield Pavement Under HWD Dynamic Impulse Load . . . . . . . . . . . . . . . . . . . . . Maissa Gharbi, Michaël Broutin, Ichem Boulkhemair, Mai Lan Nguyen, and Armelle Chabot Application of Activated Carbon on Induced Heating-Healing Characteristics of Aged Mixes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Saeed Amani, Amir Kavussi, and Mohammad M. Karimi Application of the Synthetics Geo-Composites in the Arid Zones for Rehabilitation of the Flexible Pavements Road Experimental Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mouloud Abdessemed, Rabéa Bazzine, and Said Kenai Applying the Grid Method to Study the Mechanical Behavior at Micro Scale of Asphalt Mixtures Containing Reclaimed Asphalt Pavements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ali Moukahal, Mathilde Morvan, Julien Van Rompu, and Evelyne Toussaint Asphalt Aggregate Adhesion: Study of the Influence of the Morphological, Chemical and Mineralogical Properties of Different Aggregates from Southern Brazil . . . . . . . . . . . . . . . . . . . . . . . Chaveli Brondani, Cléber Faccin, Karine W. Kraemer, Fernando D. Boeira, Luciano Pivoto Specht, and Andrea V. Nummer
xv
255
263
271
279
285
293
Asphalt Layers Thickness Determination for Top Down Cracking Initiation and a Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Nikolaidis and E. Manthos
301
Asphalt Permeability In-service: How Can It Be Evaluated Before and After Moisture Damage? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Venter, K. J. Jenkins, C. E. Rudman, and O. R. Grobbelaar
309
Assessing Geometric Features Impact of Laboratory Specimens on ITS Variation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nunzio Viscione, Francesca Russo, R. Veropalumbo, and C. Oreto
315
Assessment of Tensile Strength Reserve of Asphalt Mixtures at Low Temperatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Marek Pszczola and Cezary Szydlowski
323
xvi
Contents
Assessment of the Behavior of Emulsified Asphalt Mixes During Curing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Amélie Thiriet, Vincent Gaudefroy, Jean-Michel Piau, Frédéric Delfosse, Emmanuel Chailleux, and Christine Leroy Characterization of Aluminosilicate-Based Warm-Mix Asphalt Additive Using Experimental Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . Rituraj Patel, Vinay Hosahally Nanjegowda, Jagadeesh Mahimaluru, and Krishna Prapoorna Biligiri Characterization of Layer Index in Bituminous Pavement Using Mechanistic-Empirical Approach Based on Concentration Factor in a Layered System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S. Purakayastha, P. P. Biswas, M. K. Sahis, and G. C. Mondal Characterization of Rejuvenated Asphalt Binders Using FTIR and AFM Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rayhan Bin Ahmed and Kamal Hossain Characterization of the Lateritic Soil of Kamboinsé (Burkina Faso) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Marie Thérèse Marame Mbengue, Adamah Messan, Abdou Lawane, and Anne Pantet Circular Accelerated Surfacing Tester (CAST) for Evaluating Chip Sealing Binders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lia C. van den Kerkhof, Jeremy P. Wu, Shaun R. Cook, and Philip R. Herrington Coefficient of Thermal Expansion and Thermally Induced Internal Cracking of Asphalt Mixes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sang Soo Kim, Moses Akentuna, and Munir Nazzal Combined Recycling of RAS and RAP: Experimental and Modeling Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Abdeldjalil Daoudi, Anne Dony, Alan Carter, Layella Ziyani, and Daniel Perraton Combined Use of Natural Rubber, Biomass and Plastic Wastes in Bitumen Modification and Flexible Pavement Construction . . . . . . . . Swapan Kumer Ray, Riyadh Hossen Bhuiyan, Muhammad Saiful Islam, Md. Jaynal Abedin, Zahidul Islam, and Rashed Hasan Comparative Evaluation of Different Methods for Assessing the Glass Transition Temperature of Bituminous Binders . . . . . . . . . . . . Ezio Santagata, Chiara Tozzi, Orazio Baglieri, and Davide Dalmazzo
331
339
347
355
363
371
379
387
395
405
Contents
Comparative Study of the Mechanical Behaviour of Bitumenand Cement-Dominated Mixtures with Reclaimed Asphalt . . . . . . . . . . . Miomir Miljkovi´c, Bohdan Doł˙zycki, Mariusz Jaczewski, and Cezary Szydłowski
xvii
413
Comparing Compressive and Flexural Dynamic Asphalt Modulus by Statistical and Neural Network Modelling . . . . . . . . . . . . . . . Ali Jamshidi and Greg White
421
Comparing Laboratory Curing and In-Pavement Curing of a Foamed Bitumen Stabilized Base Course Material . . . . . . . . . . . . . . . Thomas Weir and Greg White
429
Comparing Wet Mixed and Dry Mixed Binder Modification with Recycled Waste Plastic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Greg White and Finn Hall
435
Comparison Between Rheological Behavior of a Neat Asphalt Binder and a Bio-Binder from Renewable Source . . . . . . . . . . . . . . . . . . . . Leidy V. Espinosa, Fernanda Gadler, Rafael V. Mota, Kamilla L. Vasconcelos, and Liedi L. B. Bernucci Comparison of Alternate Binder Aging Methods . . . . . . . . . . . . . . . . . . . . John A. Noël, M. Rezwan Quddus, Payman Pirzadeh, Pavel Kriz, and Ralph Shirts Comparison of Different DSR Protocols to Characterise Asphalt Binders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Laurent Porot, Johannes Büchner, Michael Steineder, Sjaak Damen, Bernhard Hofko, and Michael P. Wistuba Comparison of the Effects of Hydrated Lime on the Moisture-Induced Damage of Stone Mastic Asphalt (SMA) Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Amar Badal, Juan S. Carvajal-Munoz, and Gordon Airey Comprehensive Model for the Prediction of the Phase Angle Master Curve of Asphalt Concrete Mixes . . . . . . . . . . . . . . . . . . . . . . . . . . . Adrián Ricardo Archilla, José Pablo Corrales-Azofeifa, and José P. Aguiar-Moya Crack Evolution of Bitumen Under Torsional Shear Fatigue Loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yangming Gao, Linglin Li, and Yuqing Zhang Crack Self-healing in Asphalt as a Flow Process . . . . . . . . . . . . . . . . . . . . . Daniel Grossegger, Alvaro Garcia, and Gordon Airey
443
449
457
465
473
481 489
xviii
Contents
Cracked Asphalt Pavement Behaviour Under Thermal and Load Cycles – Laboratory Investigation Using Accelerated Loading System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Youness Berraha, Daniel Perraton, Guy Doré, and Michel Vaillancourt
497
Cracking Healing Evaluation of HMA Using Indirect Tensile Stiffness Modulus Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lee P. Leon
505
Creep Properties of Asphalt Binder, Asphalt Mastic and Asphalt Mixture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Johannes Büchner, Michael P. Wistuba, and Thilo Hilmer
513
Critical Reclaimed Asphalts Binders’ Properties from Empirical and Functional Point of View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tomas Koudelka, Ondrej Dasek, and Michal Varaus
521
Degree of Binder Activity on 100% Recycled Mixtures and Its Linear Viscoelasticity Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pablo Menezes Vestena, Pedro Orlando Borges de Almeida Junior, Luciano Pivoto Specht, Débora Tanise Bordin, Helena Leitão Pinheiro, Kamilla L. Vasconcelos, and Gustavo Menegusso Pires
529
Determination of Epoxy Resin Concentration in Epoxy Modified Bitumens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lia C. van den Kerkhof, David Alabaster, and Philip R. Herrington
537
Determination of Thermal Stresses in Asphalt Layers as a Problem of Thermo-elasticity and Unsteady Heat Flow . . . . . . . . . . Marcin Gajewski, Wojciech Ba´nkowski, and Beata Gajewska
545
Determining the Prominence of Texture Scales on Road Skid Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Malal Kane
553
Development of Permanent Deformation Models with the Inclusion of AIMS Coarse Aggregate Angularity . . . . . . . . . . . . Lee P. Leon
559
Discrete Modelling of 2PB Complex Modulus Test of Hot Asphalt Mixes Using Irregular Aggregates . . . . . . . . . . . . . . . . . . . . . . . . . . Juan Carlos Quezada and Cyrille Chazallon
567
Effect of Ageing Kinetics on the Rheological Properties of Bituminous Binders and Mixes: Experimental Study and Multi-scale Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Somé, J.-F. Barthélémy, V. Mouillet, and Ferhat Hammoum
575
Contents
Effect of Fiberglass Geogrid Reinforcement to Fatigue Resistance of Bituminous Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reuber Freire, Hervé Di Benedetto, Cédric Sauzéat, Simon Pouget, and Didier Lesueur Effect of Filler Type and Content on the Rheological Properties of Asphalt Mastics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jayvant Choudhary, Brind Kumar, and Ankit Gupta Effect of RAP and Binder Properties on Indirect Tensile Strength and Dynamic Modulus of Cold Recycled Foamed Asphalt Mixtures with High RAP Content . . . . . . . . . . . . . . . . . . . . . . . . . . Wangyu Ma, Di Wang, Fan Gu, Adam Taylor, and Randy West Effect of RAP Gradation on Fracture Tolerance of Asphalt Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yu Yan, David Hernando, Bongsuk Park, Gabriele Tebaldi, and Reynaldo Roque Effect of Soybean Oil Derived Additives on Improved Performance of Polymer-Modified Asphalt Binder and Mix Containing 50% Fine-Graded RAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Joseph Podolsky, Austin Hohmann, Conglin Chen, Zahra Sotoodeh-Nia, Nicholas Manke, Barrie Saw, Nacu Hernandez, Paul Ledtje, Michael Forrester, R. Christopher Williams, Eric Cochran, and Theodore Huisman Effect of Temperature on Self-healing Properties of Bituminous Binders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ezio Santagata, Fabrizio Miglietta, Orazio Baglieri, and Lucia Tsantilis Effect of Thermal Ageing on the Mechanical Properties and Cracking Behaviour of Asphalt Concrete . . . . . . . . . . . . . . . . . . . . . . . B. Kouevidjin, J.-F. Barthélémy, C. Somé, H. Ben Dhia, and A. Feeser Effects of Adding a Geotextile into a Pavement Structure on FWD Experimental Data and Estimated Pavement Remaining Service Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jean-Claude Carret and Alan Carter Effects of Aggregate Shape Parameters and Gradation on High-Modulus Asphalt Mix Performance . . . . . . . . . . . . . . . . . . . . . . . . Taher Baghaee Moghaddam and Hassan Baaj
xix
585
591
599
607
615
623
631
639
647
xx
Contents
Effects of Binder Content, Filler Content, Air Void and Modification on Damage Characteristics of Asphalt Mix Under Fatigue Loading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hamidreza Sahebzamani, Mohammad Zia Alavi, Orang Farzaneh, and Serge Krafft Effects of Chemical and Microstructural Constituents on the Healing Characteristics of Asphalt Binders . . . . . . . . . . . . . . . . . . . Leni F. M. Leite, Francisco T. S. Aragão, Thaísa F. Macedo, Patrícia H. Osmari, Margareth C. C. Cravo, Luis A. H. Nascimento, and Luciana N. Dantas Effects of Water Saturation and Freeze-Thaw Cycles on Fatigue Behavior of Bituminous Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Duc Thang Tran, Cédric Sauzéat, Hervé Di Benedetto, and Bertrand Pouteau Efficiency of Ultrasonic Pulse Velocity Test in the Determination of Residual Asphalt Mix Pavement Properties . . . . . . . . . . . . . . . . . . . . . . . S. Benaboud, M. Takarli, F. Allou, F. Dubois, A. Nicolaï, Bertrand Pouteau, and A. Beghin Elasto-Plastic Model for Bitumen Stabilized Materials Using Triaxial Testing and Finite Element Modelling . . . . . . . . . . . . . . . . . . . . . . Francesco Preti, Beatriz Chagas Silva Gouveia, Elena Romeo, Gabriele Tebaldi, Eshan V. Dave, and Jo E. Sias Empirical, Rheological and Chemical Properties of Aged Binder with Rejuvenators at Different Ageing Levels . . . . . . . . . . . . . . . . . . . . . . . Marko Oreškovi´c, Laurent Porot, Stefan Trifunovi´c, and Goran Mladenovi´c Enhanced Models for Vehicle Fuel Consumption for Life Cycle Assessment of Pavements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rami Chkaiban, Elie Y. Hajj, Gary Bailey, Muluneh Sime, Hao Xu, Seyed-Farzan Kazemi, and Peter E. Sebaaly Enhanced Self-healing Properties in Stone Mastic Asphalt with Encapsulated Bitumen Rejuvenators . . . . . . . . . . . . . . . . . . . . . . . . . . Nilo Ruiz-Riancho, Tahseen Saadoon, Alvaro Garcia, and Robin Hudson-Griffiths Establishing Correlations Between Force Ductility and DSR Parameters of Asphalt Emulsion Residues . . . . . . . . . . . . . . . . . . . . . . . . . . Nazimuddin M. Wasiuddin and Mohammad Islam
655
663
671
679
687
695
703
711
719
Contents
Estimation of Complex Shear Modulus of Binder Blends Produced with RAP Binder and Rejuvenator . . . . . . . . . . . . . . . . . . . . . . . A. Forton, Salvatore Mangiafico, Cédric Sauzéat, Hervé Di Benedetto, and Paul Marc Estimation of Fatigue Damage in Bituminous Mixes Using Digital Image Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Shubham Modi, Tejaswini Lakshmi Tavva, Priyadarshini Saha Chowdhury, and Sudhakar Reddy Kusam Estimation of Glassy Modulus and Differences Between Analysis Methods with Consideration of Aging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Geoffrey M. Rowe Evaluate the Impact of the Reclaimed Asphalt Pavement (RAP) Binder Activation on its Shape Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . Wellington L. G. Ferreira, Verônica T. F. Castelo Branco, Kamilla L. Vasconcelos, and Amit Bhasin Evaluating the Cracking Performance of Asphalt Mixes Using Four-point Beam Fatigue and Semicircular Beam Testing . . . . . . . . . . . . Liya Jiao, John T. Harvey, Mohamed Elkashef, Rongzong Wu, and David Jones Evaluation of Aging Processes in Binders Stabilised from Cationic Bituminous Emulsion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wojciech Sorociak, Katarzyna Konieczna, Jan B. Król, ˙ Dawid Zymełka, and Karol J. Kowalski
xxi
725
733
741
747
755
763
Evaluation of Cracking Performance Indices from Disk-Shaped Compact Tension Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Katie E. Haslett, Eshan V. Dave, and Jo E. Sias
771
Evaluation of Engineered Crumb Rubber (ECR) Performance Characteristics, Including Warm-Mix Equivalence with Polymer, Draindown Prevention, and Release Enhancement . . . . . Punyaslok Rath, J. Meister, B. Jahangiri, H. Majidifard, and W. Buttlar
779
Evaluation of Epoxy Asphalt Binders for Open-Graded Friction Course (OGFC) Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Raquel Moraes and Fan Yin
787
Evaluation of Fracture Resistance of SMA Mixtures with Various Percentages of TLA-RAP Through the Semi-circular Bend (SCB) Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lee P. Leon and Karla Bovell
795
xxii
Contents
Evaluation of Laboratory and Field Performance of Cold In-Place Recycling (CIR) Asphalt Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . Ahmed Saidi, Ayman Ali, Yusuf Mehta, Ben C. Cox, and Wade Lein Evaluation of Long Term Performance of High Modulus Bituminous Mixes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Anil Kumar, Ashish Kesharwani, Arun Kumar Goli, M. Amaranatha Reddy, and K. Sudhakar Reddy Evaluation of Pavement Damage Through the Analysis of Asphalt Layer Modulus and Strain Evolutions During an Accelerated Pavement Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Xuan Quy Le, Mai Lan Nguyen, Pierre Hornych, and Quang Tuan Nguyen Evaluation of Service Level of Emulsions Used in Surface Treatments in Different Stages of Ageing . . . . . . . . . . . . . . . . . . . . . . . . . . . Gledson Silva Mesquita Junior, Carla Marília Cavalcante Alecrim, Lilian Medeiros Gondim, and Suelly Helena de Araújo Barroso Evaluation of the Performance of Cold In-Place Recycled Base Layer Supported with Geogrid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ahmed Saidi, Andrae A. Francois, Ayman Ali, and Yusuf Mehta Evaluation of the Properties of Bituminous Binders Recovered from Various Sites in Europe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hilde Soenen, Xavier Carbonneau, Xiaohu Lu, Carl Robertus, and Benoit Tapin
803
811
819
825
833
841
Evolution of the Thermo-rheological Indices of Asphalt Binders with Aging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Runhua Zhang, Jo E. Sias, and Eshan V. Dave
849
Examining the Effects of a Self-healing Elastomer on the Properties of Bitumen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mike Aurilio and Hassan Baaj
857
Experimental and Numerical Study of Low Noise Porous Asphalt Pavement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Christina El Sawda, Fateh Fakhari Tehrani, Jérôme Dopeux, Philippe Reynaud, Joseph Absi, and Christophe Petit Experimental Evaluation of Swelling and Absorption of Crumb Rubber Aggregates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yasmina Mahmoudi, Salvatore Mangiafico, Cédric Sauzéat, Hervé Di Benedetto, Simon Pouget, Frédéric Loup, and Jean-Philippe Faure
865
873
Contents
Experimental Investigation on Ultraviolet Aging Properties of Silica Nanoparticles-Modified Bitumen . . . . . . . . . . . . . . . . . . . . . . . . . . . Goshtasp Cheraghian, Di Wang, Yun Su Kim, and Michael P. Wistuba Experimental Study on the Grading Distribution of Cold Recycled Asphalt Mixtures Produced with Bitumen Emulsion and High Strength Cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chiara Mignini, Fabrizio Cardone, and Andrea Graziani Experimental Tests on Diffusion Phenomenon Between Two Different Bitumens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stefano Noto, Salvatore Mangiafico, Elena Romeo, Cédric Sauzéat, Hervé Di Benedetto, and Gabriele Tebaldi Experimental Validation of the Dual-Oxidation Routes in Bituminous Binders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Georgios Pipintakos, H. Y. Vincent Ching, Uwe Mühlich, Hilde Soenen, Sabine Van Doorslaer, Peter Sjövall, Aikaterini Varveri, Christophe Vande Velde, and Xiaohu Lu Extending the Use of RAP in Airport Asphalt Resurfacing . . . . . . . . . . . Greg White and Ali Jamshidi Factors Affecting Mortar Thickness Distribution and Its Relationship to Cracking Resistance of Asphalt Mixtures . . . . . . . . . . . . Hussain U. Bahia, Jiwang Jiang, and Ying Lia Fatigue and Moisture Damage Resistance of Bituminous Mixtures for Railway Trackbeds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Diego Ramirez Cardona, Hervé Di Benedetto, Cédric Sauzéat, and N. Calon Fracture Characterization of Stone Mastic Asphalt (SMA) with Hydrated Lime Through the Semi-circular Bending Test Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Abdallah M. Ahmed, Juan S. Carvajal-Munoz, and Gordon Airey Friction Properties of Asphalt Surfacings Containing Sintered Lightweight Aggregate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ignacio Artamendi, Bob Allen, Phil Sabin, and Neil Leake From Laboratory Mixes to Full Scale Test: Rutting Evaluation of Bio-recycled Asphalt Mixes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Juliette Blanc, Emmanuel Chailleux, Pierre Hornych, Chris Williams, Zahra Sotoodeh-Nia, Laurent Porot, Simon Pouget, François Olard, Jean-Pascal Planche, Davide Lo Presti, and Ana Jimenez del Barco
xxiii
879
887
895
903
911
919
927
935
943
951
xxiv
Contents
Full-Scale Pavement Testing of a High Polymer Modified Asphalt Concrete Mixture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jhony Habbouche, Peter E. Sebaaly, Elie Y. Hajj, and Murugaiyah Piratheepan GLOBE: An Innovative Technical Solution to Ensure Waste Free Cold Logistic of Bituminous Binders . . . . . . . . . . . . . . . . . . . . . . . . . . . Mouhamad Mouazen, Yvong Hung, Olivier Moglia, Pauline Anaclet, Alice Ngo, Serge Krafft, Flavien Geisler, and Vincent Gaudefroy Healing Characterisation of Waste-Derived Bitumen Based on Crack Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Linglin Li, Yangming Gao, and Yuqing Zhang High Temperatures Characterization of Polymer Nanocomposite Modified Asphalt Binder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Shaban Ismael Albrka Ali, Rebaz Yousif Ahmed, and Peshawa Sediq How to Improve the Miscibility of Asphalt Binder and Polyolefins by Phosphoric Acid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Paolino Caputo, Valeria Loise, Francesca R. Lupi, Emanuela Lombardo, Ines Antunes, and Cesare Oliviero Rossi Humidity Damage Index (HDI) of Recovered Asphalt from Reclaimed Asphalt Pavement-RAP Using Different Aggregates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ana Figueroa and Mansour Solaimanian
959
967
975
983
991
999
Impact of Different Fillers on the Properties of Asphalt Mixtures . . . . . 1007 Adriano Teixeira and Didier Lesueur Impact of the Performance of the Asphalt Concrete and the Geogrid Materials on the Fatigue of Geogrid-Reinforced Asphalt Concrete—Experimental Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1015 Bertrand Pouteau, Antoine Martin, Kamal Berrada, Jacques-Antoine Decamps, and Patrice Diez Impact of Viscoelastic and Fatigue Behavior of Asphalt Mixtures Made with RAP and Asphalt Shingle on the Life Cycle Assessment Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1023 Nikolaos Vlasopoulos, Duc Tung Dao, Nouffou Tapsoba, Hervé Di Benedetto, Cédric Sauzéat, Mohsen Ech, and Nicolas Miravalls Implementing Crumb Rubber Gap-Graded and Open Graded Asphalt Technologies—The Australian Experience . . . . . . . . . . . . . . . . . . 1031 Laszlo Petho
Contents
xxv
Improved Testing Setup for Real-Time Monitoring of PMBs During Manufacturing and Rotational Viscosity Measurements . . . . . . . 1039 G. Giancontieri, Simon Pouget, and Davide Lo Presti Improving the Durability of Asphalt Mixtures with Hydrated Lime: Field Results from Recent French Sections . . . . . . . . . . . . . . . . . . . . 1047 Didier Lesueur, Pierre Métais, Patrick Pibis, Samyr El Bedoui, Hervé Ruat, Stéphane Bouron, and Ferhat Hammoum Improving the Modulus Master Curve of Bituminous Mixes Using Ultrasonic Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1055 Pejoohan Tavassoti, Taher H. Ameen, Hassan Baaj, and Giovanni Cascante Influence of Aggregate Gradation on the Permanent Deformation of Asphalt Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1063 Bernardita Lira, Robert Lundström, and Jonas Ekblad Influence of Binder Properties on Dynamic Shear Response of Asphalt Mixture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1071 Jiqing Zhu, Abubeker Ahmed, Xiaohu Lu, and Safwat Said Influence of Confinement Pressure on the Viscoelastic Response of Bituminous Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1079 B. S. Abhijith, S. P. Atul Narayan, and J. Murali Krishnan Influence of Linear Viscoelastic Behaviour of Pavement Layers Interface for Heavy Weight Deflectometer Test . . . . . . . . . . . . . . . . . . . . . . 1087 Jean-Marie Roussel, Hervé Di Benedetto, Cédric Sauzéat, and Michaël Broutin Influence of Natural Aggregates’ Mineralogical Composition on the Adhesiveness and Affinity of Bitumen . . . . . . . . . . . . . . . . . . . . . . . . 1095 Adelin Stirb, Paul Marc, Anda Belc, Florin Belc, and Gheorghe Lucaci In-Situ Cold Recycling Solutions for Roads Maintenance at the National Roads Network in Romania . . . . . . . . . . . . . . . . . . . . . . . . . 1103 Marian Peticila, Marius Muscalu, Ioan Stetco, and Marian Raicu In-Situ Measurements of Variations of RAP Contents in Hot Mix Asphalt by a Handheld FT-IR Spectrometer . . . . . . . . . . . . . . . . . . . . 1111 Shams Arafat, Nazimuddin M. Wasiuddin, and Delmar Salomon Interlaminar Mode-I Fracture Characterization Underwater of Reinforced Bituminous Specimens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1119 Maissa Gharbi, Armelle Chabot, Jean-Luc Geffard, and Mai Lan Nguyen
xxvi
Contents
Interphase Relations in the Characterisation of Bitumen Emulsion-Cement Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1127 Miomir Miljkovi´c, Andrea Graziani, and Chiara Mignini Introducing an Improved Testing Method to Evaluate the Fatigue Resistance of Bituminous Mortars . . . . . . . . . . . . . . . . . . . . . . . 1135 Alexandros Margaritis, Georgios Pipintakos, Li Ming Zhang, Geert Jacobs, Cedric Vuye, Johan Blom, and Wim Van den Bergh Investigating Fatigue Characteristics of Asphalt Binder Modified with Phase Change Materials Using Dynamic Shear Rheometer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1143 Haya Almutairi, Dandi Zhao, and Hassan Baaj Investigating the Creep and Fatigue Properties of Recycled Asphalt Concrete Containing Waste Engine and Waste Cooking Oil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1151 Hasan Taherkhani and Farid Noorian Investigating the Nonlinear Behaviour of Neat and Modified Binders Through Large Amplitude Oscillatory Shear (LAOS) Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1159 Pejoohan Tavassoti, Roberto M. Aurilio, Dandi Zhao, and Hassan Baaj Investigating the Use of a Binder Cohesion Test to Evaluate Cracking Resistance of Asphalt Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . 1167 Angelo Filonzi, Ramez Hajj, Satyavati Komaragiri, and Amit Bhasin Investigation of Cold Regions Dense Graded HMAC (EME) Sensitivity to Bitumen and Filler Content . . . . . . . . . . . . . . . . . . . . . . . . . . . 1175 Charles Neyret, Sébastien Lamothe, Daniel Perraton, Alan Carter, Marc Proteau, Hervé Di Benedetto, and Bertrand Pouteau Investigation of Cracking Potential of Modified Asphalt Mixes Composed of Synthetic Fibers by Performing 4-point Bending Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1181 Luis Alberto Perca Callomamani and Leila Hashemian Investigation of Different Design Methods for Determining the Appropriate Binder Ratio on Recycled Asphalt Mixtures . . . . . . . . . 1189 Matheus S. Gaspar, Gustavo S. Pinheiro, Kamilla L. Vasconcelos, and Liedi L. B. Bernucci Investigation of Fracture Behavior of Recycled Asphalt Mixtures Using a Discrete Element Computational Model . . . . . . . . . . . . . . . . . . . . . 1197 Meng Xu, Jia-Liang Le, Tianhao Yan, Mugurel Turos, Mihai Marasteanu, and Decheng Feng
Contents
xxvii
Investigation of the Performance of Warm Mix Asphalt Additives on 50/70 Penetrated Bitumen and Pre-cost Study . . . . . . . . . . . 1205 Süleyman Nurullah Adahi Sahin ¸ and Metin ˙Ipek Investigations of Electrical Conductivity of Graphite Nano-platelet (GNP)-Taconite Modified Asphalt Binders . . . . . . . . . . . . . 1213 Jia-Liang Le, Mihai Marasteanu, Jhenyffer Matias De Oliveira, and Mugurel Turos Laboratory and in Situ Damage Evaluation of Geogrid Used in Asphalt Concrete Pavement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1219 Cyrille Chazallon, Julien Van Rompu, Mai Lan Nguyen, Pierre Hornych, S. Mouhoubi, D. Doligez, L. Brissaud, Y. Le Gal, and E. Godard Laboratory Assessment of Low Temperature Resistance of High Modulus Asphalt Concrete in Relation to Climatic Conditions in Poland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1227 Ba´nkowski Wojciech, Gajewski Marcin, Horodecka Renata, and Mirski Krzysztof Laboratory Comparison of In-Situ, Ex-Situ and Laboratory Produced Foamed Bitumen Stabilized Base . . . . . . . . . . . . . . . . . . . . . . . . . 1235 Thomas Weir and Greg White Laboratory Evaluation of Rolling Resistance . . . . . . . . . . . . . . . . . . . . . . . . 1241 Ebrahim Riahi, Christophe Ropert, and Minh-Tan Do Laboratory Investigation of Permanent Deformation of Modified Asphalt Mixes Using Nanocellulose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1249 Thomas Johnson and Leila Hashemian Laboratory Performance of Stabilized Base with 100% Reclaimed Asphalt Pavement (RAP) Using Portland Cement, Bitumen Emulsion and Foamed-Bitumen . . . . . . . . . . . . . . . . . . . . . . . . . . . 1257 Chibuike Ogbo, Eshan V. Dave, and Jo E. Sias Large Scale Characterization of Crumb Rubber Modified Asphalt Mixtures Using Dry Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1265 Moises Bueno and Lily Poulikakos LCA of an Asphalt Mixture Produced in a Southern Italian Plant: Analysis and Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1273 Giuseppe Sollazzo, Sonia Longo, Maurizio Cellura, and Clara Celauro
xxviii
Contents
Life Cycle Assessment of Asphalt Pavements Using Crumb Rubber: A Comparative Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1281 Zhengyin Piao, Moises Bueno, Peter Mikhailenko, Muhammad Rafiq Kakar, Davide Biondini, Lily Poulikakos, and Stefanie Hellweg Life Cycle Assessment of Self-healing Versus Traditional Maintenance Road Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1289 Ana M. Rodríguez-Alloza, Daniel Garraín, Juan Gallego, and Federico Gulisano Long-Term Modelling of Composite Pavement Performance . . . . . . . . . . 1297 Evangelia Manola, Nick Thom, and Andy Collop Mechanical Assessment of Recycled Bituminous Mixtures . . . . . . . . . . . . 1305 Vítor Antunes, José Neves, and Ana Cristina Freire Mechanical Behavior of Asphalt Mixtures Applied in Field in the Southern Brazil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1313 Silvio Schuster, Cléber Faccin, Fernando D. Boeira, Luciano Pivoto Specht, and Deividi da Silva Pereira Mechanical Properties of Bitumen and Asphalt Mixture Modified with Polymer Additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1321 Quang Tuan Nguyen, Thi Thanh Nhan Hoang, Xuan Cay Bui, Van Cham La, Thi Kim Dang Tran, Quang Phuc Nguyen, and Nhu Hai Nguyen Methodology for Evaluating the Performance of Bituminous Binders Based on Rheological Indicators: Impact of the Use of a Rejuvenator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1327 Nathalie Piérard, Stefan Vansteenkiste, Ann Vanelstraete, and Philippe Peaureaux Microwave Crack-Healing Capability on Asphalt Mixtures with Silicon Carbide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1335 Alvaro González, Jonathan Valderrama, and José Norambuena-Contreras Modeling of T/C Complex Stiffness Modulus Test and Non-linearity of Asphalt Concrete Mixes . . . . . . . . . . . . . . . . . . . . . . . 1343 Léo Coulon, Georg Koval, Cyrille Chazallon, and Jean-Noël Roux Modeling of T/C Fatigue Test with Boundary Element Method and Linear Fracture Mechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1351 A. Dansou, S. Mouhoubi, Cyrille Chazallon, and M. Bonnet
Contents
xxix
Modelling of Oxidative Ageing in Bitumen Using Thermodynamics of Irreversible Processes (TIP): Potential and Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1359 Uwe Mühlich, Georgios Pipintakos, and Christos Tsakalidis Monitoring of Railway Structures of High-Speed Line Bretagne-Pays de la Loire with Bituminous and Granular Sublayers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1367 Diana Khairallah, Juliette Blanc, Pierre Hornych, Louis-Marie Cottineau, Olivier Chupin, Jean-Michel Piau, Diego Ramirez Cardona, Alain Ducreau, and Frederic Savin Monitoring Stiffness Evolution of Asphalt Concrete Through Modal Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1375 Anders Gudmarsson and Abubeker Ahmed Multi-criteria Framework to Evaluate the Oxidative Aging Resistance of Bitumen Binders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1383 Jeramie J. Adams, Michael D. Elwardany, Jean-Pascal Planche, Yvong Hung, Jeanne Zhu, Soenke Schroeder, and Mouhamad Mouazen Multiphysics Simulation and Validation of Field Ageing of Asphalt Pavements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1391 Eman Omairey and Yuqing Zhang Multiscale Analysis of Pavement Texture—Effect on Skid Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1399 Wiyao Edjeou, Veronique Cerezo, Hassan Zahouani, and Ferdinando Salvatore Multi-scale Evaluation of Asphalt Binders Containing Different Proportions of Reclaimed Asphalt Pavement (RAP) . . . . . . . . . . . . . . . . . . 1407 K. Lakshmi Roja, Bhaskar Vajipeyajula, and Eyad Masad Novel Modified Recycled Mastic for Demanding and Sustainable Asphalt Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1415 Fernando C. G. Martinho, Luís G. Picado Santos, and Francisco M. Silva Lemos Numerical Simulation of Two-Point and Four-Point Bending Fatigue Tests on Asphalt Mix Using Intrinsic Material Parameters Calibrated from Digital Image Correlation Data . . . . . . . . . 1423 Ibishola Santos, Paul Marsac, Olivier Chupin, and Ferhat Hammoum On the Assessment and Optimisation of the Processing Conditions of Tyre-Rubber Modified Bitumen . . . . . . . . . . . . . . . . . . . . . . . 1431 Juan S. Carvajal-Munoz, Ayad Subhy, and Davide Lo Presti
xxx
Contents
Organic Compounds Evaluation from Fumes Generated in Laboratory by Bio-recycled Asphalt Mixtures . . . . . . . . . . . . . . . . . . . . 1439 Vincent Gaudefroy, Davide Lo Presti, Laurent Porot, Simon Pouget, Jean-Pascal Planche, Chris Williams, and Emmanuel Chailleux Parametric Study of Binder Properties on Thermomechanical Performance of Bituminous Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1447 Gabriel Orozco, Cédric Sauzéat, Hervé Di Benedetto, Simon Pouget, Stéphane Dupriet, and Simon Baudouin Pavement Evaluation Method Using MMS . . . . . . . . . . . . . . . . . . . . . . . . . . 1455 Hinari Kawamura, Yoshifumi Nagata, Tomoaki Tokuno, Tetsuya Ishida, Tsukasa Mizutani, and Junko Yamashita Performance Testing of Hot Mix Asphalt Containing Biochar . . . . . . . . . 1465 Shadi Saadeh, Yazan Alzubi, Basel Zaatarah, Ellie H. Fini, and Pritam Katawal Predicting the Master Curve of Bituminous Mastics with Micromechanical Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1473 Hassan Fadil, Denis Jelagin, and Manfred N. Partl Predicting the Potential Impact of Geopolymers on the Creep Recovery Properties of Asphalt Binder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1481 Abdulrahman Hamid, Hassan Baaj, and Mohab El-Hakim Prediction of Fatigue Behaviour of Asphalt Mix from Tests on Asphalt Mastic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1489 Michael Steineder, Valentin Donev, Lukas Eberhardsteiner, and Bernhard Hofko Preliminary Investigation of Using Nanocellulose in Bituminous Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1495 Ali Qabur, Hui Liao, Dandi Zhao, and Hassan Baaj Properties at Low Temperatures of Warm Mix Containing High Content of Multi-recycled RAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1503 A. Pedraza, Hervé Di Benedetto, Cédric Sauzéat, and Simon Pouget Proposal for Aging Quantity Model for Estimating Binder Aging Level of HMA Mix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1511 Sungun Kim, Kyongae Ahn, and Kwang W. Kim Quantification of Compactability of Bituminous Mixtures with Different Aggregate Gradations Using Superpave Gyratory Compactor and Shear Box Compactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1519 V. T. Thushara and J. Murali Krishnan
Contents
xxxi
Reduction of Bitumen Content and Production Temperature of Hot-Mix Asphalt Incorporating RAP Using Dune Sand and Lime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1527 Jamel Neji, Ahmed Siala, Saloua El Euch Khay, and Amara Loulizi Relationship Between Fatigue Damage of Asphalt Binders and Corresponding Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1535 Letícia S. de Oliveira, Jorge L. O. Lucas Júnior, Lucas F. A. L. Babadopulos, and Jorge B. Soares Relationship Between Linear Viscoelastic Properties of Asphalt Binders and Corresponding Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1543 Letícia S. de Oliveira, Lucas S. V. da Silva, Lucas F. A. L. Babadopulos, and Jorge B. Soares Relaxation Spectrum: Why It Matters and How to Correctly Develop One? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1551 Kin-Ming Chan and Yuhong Wang Rheological Characterisation of Modified Bitumens with Biodiesel-Derived Biobinders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1563 Ana I. Weir, Gordon Airey, Colin Snape, and Ana Jiménez del Barco Carrión Rheological Investigation on the Rutting Characteristics of Nanoclay Modified Asphalt Binders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1571 Mayank Sukhija and Nikhil Saboo Rheological Modelling of the Bitumen from Reclaimed Asphalt with Rejuvenation and Re-ageing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1579 Edoardo Bocci, Emiliano Prosperi, and Paul Marsac Rheological Properties and 2S2P1D Modeling of Seven Bituminous Mixtures with Different Brazilian Bitumens . . . . . . . . . . . . . 1587 Évelyn Paniz Possebon, Luciano Pivoto Specht, Hervé Di Benedetto, Silvio Schuster, and Deividi da Silva Pereira Rheological Properties and Rutting Characterization of Natural Rubber Modified Bitumen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1595 Jarurat Wititanapanit, Juan S. Carvajal-Munoz, Chakree Bamrungwong, and Gordon Airey Rheological Properties of Asphalt Binder Compound Modified by Bio-oil and Organic Montmorillonite . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1603 Tingting Xie, Yifang Chen, and Chao Wang Rheological Properties of Bituminous Mastic with Different Filler Content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1611 Xi-li Yan, Peng Li, Qing-long You, and Xi-juan Xu
xxxii
Contents
Rheological Properties of Derived Fractions Composed of Aromatics, Resins, and Asphaltenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1619 Feipeng Xiao and Jiayu Wang Rheometrics Framework Analysis to Capture Self-assemblies Organization in Bitumen Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1627 Caroline Bottelin, Marc Chardonnet, Philippe Marchal, Lazaros Vozikis, and Yvong Hung Roles of Exogenous Cationic and Endogenous Bitumen Surfactants in the Stability of Bitumen/Water Interfaces . . . . . . . . . . . . . 1635 Shweta Jatav, Patrick Bouriat, Pauline Anaclet, Yvong Hung, Francis Rondelez, and Christophe Dicharry Rubber-Bitumen Interaction of Plant-Blended Rubberized Bitumen Prepared Under Various Blending Conditions . . . . . . . . . . . . . . 1641 Xiong Xu, Zhen Leng, Jingting Lan, Rui Li, Zhifei Tan, and Anand Sreeram Rutting Performance Analysis for Pavements with Bituminous Stabilized Mixtures as Base Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1649 Beatriz Chagas Silva Gouveia, Francesco Preti, Elena Romeo, Eshan V. Dave, Gabriele Tebaldi, and Jo E. Sias Sensitivity Analysis for Rheological Determination of Glass Transition and Crossover Temperatures at Various Reference Frequencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1657 Michael D. Elwardany, Jean-Pascal Planche, and Jeramie J. Adams Shear Fatigue Cracking Analysis of Pavement with Thick Asphalt Layers Induced by Traffic Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1665 Shan Jingsong, Han Lujia, Guo Zhongyin, and Li Feng Solar and Permeable Road: A Prototypical Study . . . . . . . . . . . . . . . . . . . 1675 Domenico Vizzari, Pierfabrizio Puntorieri, Filippo G. Praticò, Vincenzo Fiamma, and Giuseppe Barbaro Statistical Analysis of the AIGLE 3D Crossover Trial Campaign: Experimental Design, Methods and Results . . . . . . . . . . . . . . . . . . . . . . . . . 1681 Pierre Gayte and Marie Delaye Study of the Shear Stiffness of Bituminous Mastics Incorporating Recycled Glass Fillers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1689 Mohamed Mounir Boussabnia, Michel Vaillancourt, and Daniel Perraton Study on Aging Resistance of Bitumen Rejuvenated with Various Rejuvenators for Hot Recycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1697 Chetan Kumara, Gurunath Guduru, Bharath Gottumukkala, and Kranthi Kuna
Contents
xxxiii
Study on Reversibility of Aging and Recycling of Bitumen Based on Rheology, Adhesion and Chemical Properties . . . . . . . . . . . . . . . . . . . . 1705 Meng Guo, Haiqing Liu, Yubo Jiao, and Yiqiu Tan Superpave 5: Improving Asphalt Mixture Performance . . . . . . . . . . . . . . 1711 Reyhaneh Rahbar-Rastegar, M. Reza Pouranian, Dario Batioja-Alvarez, Mohammad Ali Notani, Miguel Montoya, and John E. Haddock Surface Dressing Treatment for Applications on Solar Roads . . . . . . . . . 1719 Domenico Vizzari, Eric Gennesseaux, Stéphane Lavaud, Stéphane Bouron, and Emmanuel Chailleux Sustainability Assessment of Novel Performance Enhancing Chemical Bitumen Additive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1727 Christian Holldorb, Amina Wachsmann, Sonja Cypra, Markus Oeser, Nicolás Héctor Carreño Gómez, and Michael Zeilinger Temperature and Loading Rate Susceptibility of Bituminous Mixtures on Monotonic Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1735 Ramon Botella, Félix E. Pérez-Jiménez, Rodrigo Miró, Adriana H. Martínez, and Teresa López-Montero Temperature Dependency of the Stiffening Effect of Hydrated Lime in Stone Mastic Asphalt (SMA) Mixtures . . . . . . . . . . . . . . . . . . . . . . 1743 Mariella Cardenas, Juan S. Carvajal-Munoz, and Gordon Airey Testing of Reclaimed Asphalt Model Systems for the Evaluation of the Effectiveness of Rejuvenators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1751 Davide Dalmazzo, L. Urbano, P. P. Riviera, and Ezio Santagata The Bitumen Related Problems Observed on Soft Asphalt Concrete Pavements: Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1759 Michalina Makowska and Katri Eskola The Chemistry Behind Rheological and Thermal Transitions of Oxidized Bitumen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1767 Alejandra Baldi, Rafael Ernesto Villegas-Villegas, José P. Aguiar-Moya, and Luis G. Loria-Salazar The Effect of a Chemical Warm Mix Additive on the Self-Healing Capability of Bitumen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1775 Roberto M. Aurilio, Mike Aurilio, and Hassan Baaj The Effect of Additives on Water Vapor Condensation on Bituminous Surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1783 F. Tarpoudi Baheri, M. Rico Luengo, T. M. Schutzius, and D. Poulikakos
xxxiv
Contents
The Effect of Mix Constituents in the Permanent Deformation Resistance of Asphalt Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1791 Otávio José de Freitas Gomes, Jorge B. Soares, and Juceline Batista S. Bastos The Heating and Compaction of Asphalt Mixtures as a Self-healing Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1799 Caio R. Santos, Marina M. Cabette, Jorge C. Pais, and Paulo A. Pereira The Impact of Binder Type on Pavement Design: A Comparative Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1807 Konstantina Georgouli, Christina Plati, and Andreas Loizos The Influence of Cement Type on Early Properties of Cold In-Place Recycled Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1815 Bohdan Doł˙zycki, Mariusz Jaczewski, and Cezary Szydłowski The Influence of the Mastic Coating of Untreated Reclaimed Asphalt Pavement on the Permanent and Resilient Behaviours . . . . . . . . 1823 Laura Gaillard, Cyrille Chazallon, Pierre Hornych, Juan Carlos Quezada, and Jean-Luc Geffard The Influence of Wax Model Compounds on the Surface Topography of Bitumen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1831 Johan Blom, Niko Van den Brande, and Hilde Soenen The Low Temperature Performance of Rubberized Bituminous Mixture Using the Dry Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1839 Dongdong Ge, Xuelian Li, and Zhanping You The Use of Direct Shear Test for Optimization of Interlayer Bonding Under a Poroelastic Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1845 Piotr Jaskula, Marcin Stienss, Cezary Szydlowski, and Dawid Rys Time and Storage Dependent Effects of Bitumen—Comparison of Surface and Bulk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1853 Johannes Mirwald, Stefan Werkovits, Ingrid Camargo, Daniel Maschauer, Bernhard Hofko, and Hinrich Grothe Top-Down Cracking in Airfield Pavement Structures—An In-Situ Test Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1861 Amir Sadoun, Michaël Broutin, Ichem Boulkhemair, Alexandre Duprey, Arnaud Mazars, and Marc Huault Tracking Degree of Blending Between Recycled and Virgin Binder Through Asphalt Mix Phase Angle . . . . . . . . . . . . . . . . . . . . . . . . . . 1869 Payman Pirzadeh and Hassan Baaj
Contents
xxxv
TSRST Heterogeneous Modeling for Understanding Failure Depending on Asphalt Mix Design and Experimental Conditions . . . . . . 1877 Fateh Fakhari Tehrani, Christophe Petit, and Joseph Absi Understanding of Structural and Surface Tension Properties of Asphalt Model Using Molecular Dynamics Simulation . . . . . . . . . . . . . 1885 Lingyun You, Theodora Spyriouni, Qingli Dai, Zhanping You, Jaroslaw W. Drelich, and Ashok Khanal Use of Modified Reclaimed Asphalt in Warm Mixtures . . . . . . . . . . . . . . . 1893 G. Ferrotti, Francesco Canestrari, J. Xiaotian, and Fabrizio Cardone Use of Multiple Stress Creep Recovery Test to Evaluate Viscoelastic Properties of Asphalt Mastic . . . . . . . . . . . . . . . . . . . . . . . . . . . 1901 Song Li, Ghim Ping Ong, and Fujian Ni Use of Recycled Asphalt Pavement (RAP) in Airport Pavements . . . . . . 1909 Navneet Garg, Hasan Kazmee, and Lia Ricalde Using Linear Elastic Layer Analysis Based Method for Back-Calculating the Layer Moduli of Small Paneled Thin White Topping Pavements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1917 Venkata Joga Rao Bulusu, Sudhakar Reddy Kusam, and Amarnatha Reddy Muppireddy Validation of the Time-Temperature Superposition Principle (TTSP) in the Non-linear Domain for Bituminous Mixtures with Reclaimed Asphalt Pavement (40% RAP) . . . . . . . . . . . . . . . . . . . . . . 1925 Mohamed Amine Bradai, Nouffou Tapsoba, Cédric Sauzéat, Hervé Di Benedetto, and Jamel Neji Viennese Aging Procedure (VAPro): Adaption for Low-Temperature Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1933 Daniel Maschauer, Johannes Mirwald, Bernhard Hofko, and Hinrich Grothe Viscoelastic Analysis of Top-Down Crack in Geosynthetic Reinforced Asphalt Pavements Using FEM . . . . . . . . . . . . . . . . . . . . . . . . . 1941 Masoud Jalali and Hasan Taherkhani Correction to: Proceedings of the RILEM International Symposium on Bituminous Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hervé Di Benedetto, Hassan Baaj, Emmanuel Chailleux, Gabriele Tebaldi, Cédric Sauzéat, and Salvatore Mangiafico
C1
RILEM Publications
The following list is presenting the global offer of RILEM Publications, sorted by series. Each publication is available in printed version and/or in online version.
RILEM Proceedings (PRO) PRO 1: Durability of High Performance Concrete (ISBN: 2-912143-03-9; e-ISBN: 2-351580-12-5; e-ISBN: 2351580125); Ed. H. Sommer PRO 2: Chloride Penetration into Concrete (ISBN: 2-912143-00-04; e-ISBN: 2912143454); Eds. L.-O. Nilsson and J.-P. Ollivier PRO 3: Evaluation and Strengthening of Existing Masonry Structures (ISBN: 2-912143-02-0; e-ISBN: 2351580141); Eds. L. Binda and C. Modena PRO 4: Concrete: From Material to Structure (ISBN: 2-912143-04-7; e-ISBN: 2351580206); Eds. J.-P. Bournazel and Y. Malier PRO 5: The Role of Admixtures in High Performance Concrete (ISBN: 2-91214305-5; e-ISBN: 2351580214); Eds. J. G. Cabrera and R. Rivera-Villarreal PRO 6: High Performance Fiber Reinforced Cement Composites—HPFRCC 3 (ISBN: 2-912143-06-3; e-ISBN: 2351580222); Eds. H. W. Reinhardt and A. E. Naaman PRO 7: 1st International RILEM Symposium on Self-Compacting Concrete (ISBN: 2-912143-09-8; e-ISBN: 2912143721); Eds. Å. Skarendahl and Ö. Petersson PRO 8: International RILEM Symposium on Timber Engineering (ISBN: 2-91214310-1; e-ISBN: 2351580230); Ed. L. Boström PRO 9: 2nd International RILEM Symposium on Adhesion between Polymers and Concrete ISAP ’99 (ISBN: 2-912143-11-X; e-ISBN: 2351580249); Eds. Y. Ohama and M. Puterman xxxvii
xxxviii
RILEM Publications
PRO 10: 3rd International RILEM Symposium on Durability of Building and Construction Sealants (ISBN: 2-912143-13-6; e-ISBN: 2351580257); Ed. A. T. Wolf PRO 11: 4th International RILEM Conference on Reflective Cracking in Pavements (ISBN: 2-912143-14-4; e-ISBN: 2351580265); Eds. A. O. Abd El Halim, D. A. Taylor and El H. H. Mohamed PRO 12: International RILEM Workshop on Historic Mortars: Characteristics and Tests (ISBN: 2-912143-15-2; e-ISBN: 2351580273); Eds. P. Bartos, C. Groot and J. J. Hughes PRO 13: 2nd International RILEM Symposium on Hydration and Setting (ISBN: 2-912143-16-0; e-ISBN: 2351580281); Ed. A. Nonat PRO 14: Integrated Life-Cycle Design of Materials and Structures—ILCDES 2000 (ISBN: 951-758-408-3; e-ISBN: 235158029X); (ISSN: 0356-9403); Ed. S. Sarja PRO 15: Fifth RILEM Symposium on Fibre-Reinforced Concretes (FRC)— BEFIB’2000 (ISBN: 2-912143-18-7; e-ISBN: 291214373X); Eds. P. Rossi and G. Chanvillard PRO 16: Life Prediction and Management of Concrete Structures (ISBN: 2-91214319-5; e-ISBN: 2351580303); Ed. D. Naus PRO 17: Shrinkage of Concrete—Shrinkage 2000 (ISBN: 2-912143-20-9; e-ISBN: 2351580311); Eds. V. Baroghel-Bouny and P.-C. Aïtcin PRO 18: Measurement and Interpretation of the On-Site Corrosion Rate (ISBN: 2-912143-21-7; e-ISBN: 235158032X); Eds. C. Andrade, C. Alonso, J. Fullea, J. Polimon and J. Rodriguez PRO 19: Testing and Modelling the Chloride Ingress into Concrete (ISBN: 2-912143-22-5; e-ISBN: 2351580338); Eds. C. Andrade and J. Kropp PRO 20: 1st International RILEM Workshop on Microbial Impacts on Building Materials (CD 02) (e-ISBN 978-2-35158-013-4); Ed. M. Ribas Silva PRO 21: International RILEM Symposium on Connections between Steel and Concrete (ISBN: 2-912143-25-X; e-ISBN: 2351580346); Ed. R. Eligehausen PRO 22: International RILEM Symposium on Joints in Timber Structures (ISBN: 2-912143-28-4; e-ISBN: 2351580354); Eds. S. Aicher and H.-W. Reinhardt PRO 23: International RILEM Conference on Early Age Cracking in Cementitious Systems (ISBN: 2-912143-29-2; e-ISBN: 2351580362); Eds. K. Kovler and A. Bentur PRO 24: 2nd International RILEM Workshop on Frost Resistance of Concrete (ISBN: 2-912143-30-6; e-ISBN: 2351580370); Eds. M. J. Setzer, R. Auberg and H.-J. Keck
RILEM Publications
xxxix
PRO 25: International RILEM Workshop on Frost Damage in Concrete (ISBN: 2912143-31-4; e-ISBN: 2351580389); Eds. D. J. Janssen, M. J. Setzer and M. B. Snyder PRO 26: International RILEM Workshop on On-Site Control and Evaluation of Masonry Structures (ISBN: 2-912143-34-9; e-ISBN: 2351580141); Eds. L. Binda and R. C. de Vekey PRO 27: International RILEM Symposium on Building Joint Sealants (CD03; eISBN: 235158015X); Ed. A. T. Wolf PRO 28: 6th International RILEM Symposium on Performance Testing and Evaluation of Bituminous Materials—PTEBM’03 (ISBN: 2-912143-35-7; e-ISBN: 978-2-912143-77-8); Ed. M. N. Partl PRO 29: 2nd International RILEM Workshop on Life Prediction and Ageing Management of Concrete Structures (ISBN: 2-912143-36-5; e-ISBN: 2912143780); Ed. D. J. Naus PRO 30: 4th International RILEM Workshop on High Performance Fiber Reinforced Cement Composites—HPFRCC 4 (ISBN: 2-912143-37-3; e-ISBN: 2912143799); Eds. A. E. Naaman and H. W. Reinhardt PRO 31: International RILEM Workshop on Test and Design Methods for Steel Fibre Reinforced Concrete: Background and Experiences (ISBN: 2-912143-38-1; e-ISBN: 2351580168); Eds. B. Schnütgen and L. Vandewalle PRO 32: International Conference on Advances in Concrete and Structures 2 vol. (ISBN (set): 2-912143-41-1; e-ISBN: 2351580176); Eds. Ying-shu Yuan, Surendra P. Shah and Heng-lin Lü PRO 33: 3rd International Symposium on Self-Compacting Concrete (ISBN: 2912143-42-X; e-ISBN: 2912143713); Eds. Ó. Wallevik and I. Níelsson PRO 34: International RILEM Conference on Microbial Impact on Building Materials (ISBN: 2-912143-43-8; e-ISBN: 2351580184); Ed. M. Ribas Silva PRO 35: International RILEM TC 186-ISA on Internal Sulfate Attack and Delayed Ettringite Formation (ISBN: 2-912143-44-6; e-ISBN: 2912143802); Eds. K. Scrivener and J. Skalny PRO 36: International RILEM Symposium on Concrete Science and Engineering— A Tribute to Arnon Bentur (ISBN: 2-912143-46-2; e-ISBN: 2912143586); Eds. K. Kovler, J. Marchand, S. Mindess and J. Weiss PRO 37: 5th International RILEM Conference on Cracking in Pavements— Mitigation, Risk Assessment and Prevention (ISBN: 2-912143-47-0; e-ISBN: 2912143764); Eds. C. Petit, I. Al-Qadi and A. Millien PRO 38: 3rd International RILEM Workshop on Testing and Modelling the Chloride Ingress into Concrete (ISBN: 2-912143-48-9; e-ISBN: 2912143578); Eds. C. Andrade and J. Kropp
xl
RILEM Publications
PRO 39: 6th International RILEM Symposium on Fibre-Reinforced Concretes— BEFIB 2004 (ISBN: 2-912143-51-9; e-ISBN: 2912143748); Eds. M. Di Prisco, R. Felicetti and G. A. Plizzari PRO 40: International RILEM Conference on the Use of Recycled Materials in Buildings and Structures (ISBN: 2-912143-52-7; e-ISBN: 2912143756); Eds. E. Vázquez, Ch. F. Hendriks and G. M. T. Janssen PRO 41: RILEM International Symposium on Environment-Conscious Materials and Systems for Sustainable Development (ISBN: 2-912143-55-1; e-ISBN: 2912143640); Eds. N. Kashino and Y. Ohama PRO 42: SCC’2005—China: 1st International Symposium on Design, Performance and Use of Self-Consolidating Concrete (ISBN: 2-912143-61-6; e-ISBN: 2912143624); Eds. Zhiwu Yu, Caijun Shi, Kamal Henri Khayat and Youjun Xie PRO 43: International RILEM Workshop on Bonded Concrete Overlays (e-ISBN: 2-912143-83-7); Eds. J. L. Granju and J. Silfwerbrand PRO 44: 2nd International RILEM Workshop on Microbial Impacts on Building Materials (CD11) (e-ISBN: 2-912143-84-5); Ed. M. Ribas Silva PRO 45: 2nd International Symposium on Nanotechnology in Construction, Bilbao (ISBN: 2-912143-87-X; e-ISBN: 2912143888); Eds. Peter J. M. Bartos, Yolanda de Miguel and Antonio Porro PRO 46: Concrete Life’06—International RILEM-JCI Seminar on Concrete Durability and Service Life Planning: Curing, Crack Control, Performance in Harsh Environments (ISBN: 2-912143-89-6; e-ISBN: 291214390X); Ed. K. Kovler PRO 47: International RILEM Workshop on Performance Based Evaluation and Indicators for Concrete Durability (ISBN: 978-2-912143-95-2; e-ISBN: 9782912143969); Eds. V. Baroghel-Bouny, C. Andrade, R. Torrent and K. Scrivener PRO 48: 1st International RILEM Symposium on Advances in Concrete through Science and Engineering (e-ISBN: 2-912143-92-6); Eds. J. Weiss, K. Kovler, J. Marchand, and S. Mindess PRO 49: International RILEM Workshop on High Performance Fiber Reinforced Cementitious Composites in Structural Applications (ISBN: 2-912143-93-4; e-ISBN: 2912143942); Eds. G. Fischer and V. C. Li PRO 50: 1st International RILEM Symposium on Textile Reinforced Concrete (ISBN: 2-912143-97-7; e-ISBN: 2351580087); Eds. Josef Hegger, Wolfgang Brameshuber and Norbert Will PRO 51: 2nd International Symposium on Advances in Concrete through Science and Engineering (ISBN: 2-35158-003-6; e-ISBN: 2-35158-002-8); Eds. J. Marchand, B. Bissonnette, R. Gagné, M. Jolin and F. Paradis PRO 52: Volume Changes of Hardening Concrete: Testing and Mitigation (ISBN: 2-35158-004-4; e-ISBN: 2-35158-005-2); Eds. O. M. Jensen, P. Lura and K. Kovler
RILEM Publications
xli
PRO 53: High Performance Fiber Reinforced Cement Composites—HPFRCC5 (ISBN: 978-2-35158-046-2; e-ISBN: 978-2-35158-089-9); Eds. H. W. Reinhardt and A. E. Naaman PRO 54: 5th International RILEM Symposium on Self-Compacting Concrete (ISBN: 978-2-35158-047-9; e-ISBN: 978-2-35158-088-2); Eds. G. De Schutter and V. Boel PRO 55: International RILEM Symposium Photocatalysis, Environment and Construction Materials (ISBN: 978-2-35158-056-1; e-ISBN: 978-2-35158-057-8); Eds. P. Baglioni and L. Cassar PRO 56: International RILEM Workshop on Integral Service Life Modelling of Concrete Structures (ISBN 978-2-35158-058-5; e-ISBN: 978-2-35158-090-5); Eds. R. M. Ferreira, J. Gulikers and C. Andrade PRO 57: RILEM Workshop on Performance of cement-based materials in aggressive aqueous environments (e-ISBN: 978-2-35158-059-2); Ed. N. De Belie PRO 58: International RILEM Symposium on Concrete Modelling—CONMOD’08 (ISBN: 978-2-35158-060-8; e-ISBN: 978-2-35158-076-9); Eds. E. Schlangen and G. De Schutter PRO 59: International RILEM Conference on On Site Assessment of Concrete, Masonry and Timber Structures—SACoMaTiS 2008 (ISBN set: 978-2-35158-061-5; e-ISBN: 978-2-35158-075-2); Eds. L. Binda, M. di Prisco and R. Felicetti PRO 60: Seventh RILEM International Symposium on Fibre Reinforced Concrete: Design and Applications—BEFIB 2008 (ISBN: 978-2-35158-064-6; e-ISBN: 9782-35158-086-8); Ed. R. Gettu PRO 61: 1st International Conference on Microstructure Related Durability of Cementitious Composites 2 vol., (ISBN: 978-2-35158-065-3; e-ISBN: 978-2-35158084-4); Eds. W. Sun, K. van Breugel, C. Miao, G. Ye and H. Chen PRO 62: NSF/ RILEM Workshop: In-situ Evaluation of Historic Wood and Masonry Structures (e-ISBN: 978-2-35158-068-4); Eds. B. Kasal, R. Anthony and M. Drdácký PRO 63: Concrete in Aggressive Aqueous Environments: Performance, Testing and Modelling, 2 vol., (ISBN: 978-2-35158-071-4; e-ISBN: 978-2-35158-082-0); Eds. M. G. Alexander and A. Bertron PRO 64: Long Term Performance of Cementitious Barriers and Reinforced Concrete in Nuclear Power Plants and Waste Management—NUCPERF 2009 (ISBN: 978-235158-072-1; e-ISBN: 978-2-35158-087-5); Eds. V. L’Hostis, R. Gens and C. Gallé PRO 65: Design Performance and Use of Self-consolidating Concrete—SCC’2009 (ISBN: 978-2-35158-073-8; e-ISBN: 978-2-35158-093-6); Eds. C. Shi, Z. Yu, K. H. Khayat and P. Yan PRO 66: 2nd International RILEM Workshop on Concrete Durability and Service Life Planning—ConcreteLife’09 (ISBN: 978-2-35158-074-5; ISBN: 978-2-35158074-5); Ed. K. Kovler
xlii
RILEM Publications
PRO 67: Repairs Mortars for Historic Masonry (e-ISBN: 978-2-35158-083-7); Ed. C. Groot PRO 68: Proceedings of the 3rd International RILEM Symposium on ‘Rheology of Cement Suspensions such as Fresh Concrete (ISBN 978-2-35158-091-2; e-ISBN: 978-2-35158-092-9); Eds. O. H. Wallevik, S. Kubens and S. Oesterheld PRO 69: 3rd International PhD Student Workshop on ‘Modelling the Durability of Reinforced Concrete (ISBN: 978-2-35158-095-0); Eds. R. M. Ferreira, J. Gulikers and C. Andrade PRO 70: 2nd International Conference on ‘Service Life Design for Infrastructure’ (ISBN set: 978-2-35158-096-7, e-ISBN: 978-2-35158-097-4); Eds. K. van Breugel, G. Ye and Y. Yuan PRO 71: Advances in Civil Engineering Materials—The 50-year Teaching Anniversary of Prof. Sun Wei’ (ISBN: 978-2-35158-098-1; e-ISBN: 978-2-35158-099-8); Eds. C. Miao, G. Ye and H. Chen PRO 72: First International Conference on ‘Advances in Chemically-Activated Materials—CAM’2010’ (2010), 264 pp., ISBN: 978-2-35158-101-8; e-ISBN: 9782-35158-115-5; Eds. Caijun Shi and Xiaodong Shen PRO 73: 2nd International Conference on ‘Waste Engineering and Management— ICWEM 2010’ (2010), 894 pp., ISBN: 978-2-35158-102-5; e-ISBN: 978-2-35158103-2, Eds. J. Zh. Xiao, Y. Zhang, M. S. Cheung and R. Chu PRO 74: International RILEM Conference on ‘Use of Superabsorsorbent Polymers and Other New Addditives in Concrete’ (2010) 374 pp., ISBN: 978-2-35158-104-9; e-ISBN: 978-2-35158-105-6; Eds. O.M. Jensen, M.T. Hasholt, and S. Laustsen PRO 75: International Conference on ‘Material Science—2nd ICTRC—Textile Reinforced Concrete—Theme 1’ (2010) 436 pp., ISBN: 978-2-35158-106-3; eISBN: 978-2-35158-107-0; Ed. W. Brameshuber PRO 76: International Conference on ‘Material Science—HetMat—Modelling of Heterogeneous Materials—Theme 2’ (2010) 255 pp., ISBN: 978-2-35158-108-7; e-ISBN: 978-2-35158-109-4; Ed. W. Brameshuber PRO 77: International Conference on ‘Material Science—AdIPoC—Additions Improving Properties of Concrete—Theme 3’ (2010) 459 pp., ISBN: 978-2-35158110-0; e-ISBN: 978-2-35158-111-7; Ed. W. Brameshuber PRO 78: 2nd Historic Mortars Conference and RILEM TC 203-RHM Final Workshop—HMC2010 (2010) 1416 pp., e-ISBN: 978-2-35158-112-4; Eds. J. Válek, C. Groot and J. J. Hughes PRO 79: International RILEM Conference on Advances in Construction Materials Through Science and Engineering (2011) 213 pp., ISBN: 978-2-35158-116-2, eISBN: 978-2-35158-117-9; Eds. Christopher Leung and K.T. Wan
RILEM Publications
xliii
PRO 80: 2nd International RILEM Conference on Concrete Spalling due to Fire Exposure (2011) 453 pp., ISBN: 978-2-35158-118-6; e-ISBN: 978-2-35158-119-3; Eds. E.A.B. Koenders and F. Dehn PRO 81: 2nd International RILEM Conference on Strain Hardening Cementitious Composites (SHCC2-Rio) (2011) 451 pp., ISBN: 978-2-35158-120-9; e-ISBN: 9782-35158-121-6; Eds. R.D. Toledo Filho, F.A. Silva, E.A.B. Koenders and E.M.R. Fairbairn PRO 82: 2nd International RILEM Conference on Progress of Recycling in the Built Environment (2011) 507 pp., e-ISBN: 978-2-35158-122-3; Eds. V.M. John, E. Vazquez, S.C. Angulo and C. Ulsen PRO 83: 2nd International Conference on Microstructural-related Durability of Cementitious Composites (2012) 250 pp., ISBN: 978-2-35158-129-2; e-ISBN: 978-2-35158-123-0; Eds. G. Ye, K. van Breugel, W. Sun and C. Miao PRO 84: CONSEC13—Seventh International Conference on Concrete under Severe Conditions—Environment and Loading (2013) 1930 pp., ISBN: 978-2-35158-1247; e-ISBN: 978-2- 35158-134-6; Eds. Z.J. Li, W. Sun, C.W. Miao, K. Sakai, O.E. Gjorv and N. Banthia PRO 85: RILEM-JCI International Workshop on Crack Control of Mass Concrete and Related issues concerning Early-Age of Concrete Structures—ConCrack 3— Control of Cracking in Concrete Structures 3 (2012) 237 pp., ISBN: 978-2-35158125-4; e-ISBN: 978-2-35158-126-1; Eds. F. Toutlemonde and J.-M. Torrenti PRO 86: International Symposium on Life Cycle Assessment and Construction (2012) 414 pp., ISBN: 978-2-35158-127-8, e-ISBN: 978-2-35158-128-5; Eds. A. Ventura and C. de la Roche PRO 87: UHPFRC 2013—RILEM-fib-AFGC International Symposium on UltraHigh Performance Fibre-Reinforced Concrete (2013), ISBN: 978-2-35158-130-8, e-ISBN: 978-2-35158-131-5; Eds. F. Toutlemonde PRO 88: 8th RILEM International Symposium on Fibre Reinforced Concrete (2012) 344 pp., ISBN: 978-2-35158-132-2; e-ISBN: 978-2-35158-133-9; Eds. Joaquim A.O. Barros PRO 89: RILEM International workshop on performance-based specification and control of concrete durability (2014) 678 pp., ISBN: 978-2-35158-135-3; e-ISBN: 978-2-35158-136-0; Eds. D. Bjegovi´c, H. Beushausen and M. Serdar PRO 90: 7th RILEM International Conference on Self-Compacting Concrete and of the 1st RILEM International Conference on Rheology and Processing of Construction Materials (2013) 396 pp., ISBN: 978-2-35158-137-7; e-ISBN: 978-2-35158-138-4; Eds. Nicolas Roussel and Hela Bessaies-Bey PRO 91: CONMOD 2014—RILEM International Symposium on Concrete Modelling (2014), ISBN: 978-2-35158-139-1; e-ISBN: 978-2-35158-140-7; Eds. Kefei Li, Peiyu Yan and Rongwei Yang
xliv
RILEM Publications
PRO 92: CAM 2014—2nd International Conference on advances in chemicallyactivated materials (2014) 392 pp., ISBN: 978-2-35158-141-4; e-ISBN: 978-235158-142-1; Eds. Caijun Shi and Xiadong Shen PRO 93: SCC 2014—3rd International Symposium on Design, Performance and Use of Self-Consolidating Concrete (2014) 438 pp., ISBN: 978-2-35158-143-8; e-ISBN: 978-2-35158-144-5; Eds. Caijun Shi, Zhihua Ou and Kamal H. Khayat PRO 94 (online version): HPFRCC-7—7th RILEM conference on High performance fiber reinforced cement composites (2015), e-ISBN: 978-2-35158-146-9; Eds. H.W. Reinhardt, G.J. Parra-Montesinos and H. Garrecht PRO 95: International RILEM Conference on Application of superabsorbent polymers and other new admixtures in concrete construction (2014), ISBN: 978-235158-147-6; e-ISBN: 978-2-35158-148-3; Eds. Viktor Mechtcherine and Christof Schroefl PRO 96 (online version): XIII DBMC: XIII International Conference on Durability of Building Materials and Components (2015), e-ISBN: 978-2-35158-149-0; Eds. M. Quattrone and V.M. John PRO 97: SHCC3—3rd International RILEM Conference on Strain Hardening Cementitious Composites (2014), ISBN: 978-2-35158-150-6; e-ISBN: 978-235158-151-3; Eds. E. Schlangen, M.G. Sierra Beltran, M. Lukovic and G. Ye PRO 98: FERRO-11—11th International Symposium on Ferrocement and 3rd ICTRC—International Conference on Textile Reinforced Concrete (2015), ISBN: 978-2-35158-152-0; e-ISBN: 978-2-35158-153-7; Ed. W. Brameshuber PRO 99 (online version): ICBBM 2015—1st International Conference on BioBased Building Materials (2015), e-ISBN: 978-2-35158-154-4; Eds. S. Amziane and M. Sonebi PRO 100: SCC16—RILEM Self-Consolidating Concrete Conference (2016), ISBN: 978-2-35158-156-8; e-ISBN: 978-2-35158-157-5; Ed. Kamal H. Kayat PRO 101 (online version): III Progress of Recycling in the Built Environment (2015), e-ISBN: 978-2-35158-158-2; Eds I. Martins, C. Ulsen and S. C. Angulo PRO 102 (online version): RILEM Conference on Microorganisms-Cementitious Materials Interactions (2016), e-ISBN: 978-2-35158-160-5; Eds. Alexandra Bertron, Henk Jonkers and Virginie Wiktor PRO 103 (online version): ACESC’16—Advances in Civil Engineering and Sustainable Construction (2016), e-ISBN: 978-2-35158-161-2; Eds. T.Ch. Madhavi, G. Prabhakar, Santhosh Ram and P.M. Rameshwaran PRO 104 (online version): SSCS’2015—Numerical Modeling—Strategies for Sustainable Concrete Structures (2015), e-ISBN: 978-2-35158-162-9 PRO 105: 1st International Conference on UHPC Materials and Structures (2016), ISBN: 978-2-35158-164-3; e-ISBN: 978-2-35158-165-0
RILEM Publications
xlv
PRO 106: AFGC-ACI-fib-RILEM International Conference on Ultra-HighPerformance Fibre-Reinforced Concrete—UHPFRC 2017 (2017), ISBN: 978-235158-166-7; e-ISBN: 978-2-35158-167-4; Eds. François Toutlemonde and Jacques Resplendino PRO 107 (online version): XIV DBMC—14th International Conference on Durability of Building Materials and Components (2017), e-ISBN: 978-2-35158-159-9; Eds. Geert De Schutter, Nele De Belie, Arnold Janssens and Nathan Van Den Bossche PRO 108: MSSCE 2016—Innovation of Teaching in Materials and Structures (2016), ISBN: 978-2-35158-178-0; e-ISBN: 978-2-35158-179-7; Ed. Per Goltermann PRO 109 (2 volumes): MSSCE 2016—Service Life of Cement-Based Materials and Structures (2016), ISBN Vol. 1: 978-2-35158-170-4; Vol. 2: 978-2-35158-171-4; Set Vol. 1&2: 978-2-35158-172-8; e-ISBN : 978-2-35158-173-5; Eds. Miguel Azenha, Ivan Gabrijel, Dirk Schlicke, Terje Kanstad and Ole Mejlhede Jensen PRO 110: MSSCE 2016—Historical Masonry (2016), ISBN: 978-2-35158-178-0; e-ISBN: 978-2-35158-179-7; Eds. Inge Rörig-Dalgaard and Ioannis Ioannou PRO 111: MSSCE 2016—Electrochemistry in Civil Engineering (2016); ISBN: 978-2-35158-176-6; e-ISBN: 978-2-35158-177-3; Ed. Lisbeth M. Ottosen PRO 112: MSSCE 2016—Moisture in Materials and Structures (2016), ISBN: 9782-35158-178-0; e-ISBN: 978-2-35158-179-7; Eds. Kurt Kielsgaard Hansen, Carsten Rode and Lars-Olof Nilsson PRO 113: MSSCE 2016—Concrete with Supplementary Cementitious Materials (2016), ISBN: 978-2-35158-178-0; e-ISBN: 978-2-35158-179-7; Eds. Ole Mejlhede Jensen, Konstantin Kovler and Nele De Belie PRO 114: MSSCE 2016—Frost Action in Concrete (2016), ISBN: 978-2-35158182-7; e-ISBN: 978-2-35158-183-4; Eds. Marianne Tange Hasholt, Katja Fridh and R. Doug Hooton PRO 115: MSSCE 2016—Fresh Concrete (2016), ISBN: 978-2-35158-184-1; eISBN: 978-2-35158-185-8; Eds. Lars N. Thrane, Claus Pade, Oldrich Svec and Nicolas Roussel PRO 116: BEFIB 2016—9th RILEM International Symposium on Fiber Reinforced Concrete (2016), ISBN: 978-2-35158-187-2; e-ISBN: 978-2-35158-186-5; Eds. N. Banthia, M. di Prisco and S. Soleimani-Dashtaki PRO 117: 3rd International RILEM Conference on Microstructure Related Durability of Cementitious Composites (2016), ISBN: 978-2-35158-188-9; e-ISBN: 9782-35158-189-6; Eds. Changwen Miao, Wei Sun, Jiaping Liu, Huisu Chen, Guang Ye and Klaas van Breugel
xlvi
RILEM Publications
PRO 118 (4 volumes): International Conference on Advances in Construction Materials and Systems (2017), ISBN Set: 978-2-35158-190-2; Vol. 1: 978-235158-193-3; Vol. 2: 978-2-35158-194-0; Vol. 3: ISBN:978-2-35158-195-7; Vol. 4: ISBN:978-2-35158-196-4; e-ISBN: 978-2-35158-191-9; Ed. Manu Santhanam PRO 119 (online version): ICBBM 2017—Second International RILEM Conference on Bio-based Building Materials, (2017), e-ISBN: 978-2-35158-192-6; Ed. Sofiane Amziane PRO 120 (2 volumes): EAC-02—2nd International RILEM/COST Conference on Early Age Cracking and Serviceability in Cement-based Materials and Structures, (2017), Vol. 1: 978-2-35158-199-5, Vol. 2: 978-2-35158-200-8, Set: 978-2-35158197-1, e-ISBN: 978-2-35158-198-8; Eds. Stéphanie Staquet and Dimitrios Aggelis PRO 121 (2 volumes): SynerCrete18: Interdisciplinary Approaches for Cementbased Materials and Structural Concrete: Synergizing Expertise and Bridging Scales of Space and Time, (2018), Set: 978-2-35158-202-2, Vol.1: 978-2-35158-211-4, Vol.2: 978-2-35158-212-1, e-ISBN: 978-2-35158-203-9; Eds. Miguel Azenha, Dirk Schlicke, Farid Benboudjema, Agnieszka Knoppik PRO 122: SCC’2018 China—Fourth International Symposium on Design, Performance and Use of Self-Consolidating Concrete, (2018), ISBN: 978-2-35158-204-6, e-ISBN: 978-2-35158-205-3; Eds. C. Shi, Z. Zhang, K. H. Khayat PRO 123: Final Conference of RILEM TC 253-MCI: Microorganisms-Cementitious Materials Interactions (2018), Set: 978-2-35158-207-7, Vol.1: 978-2-35158-209-1, Vol.2: 978-2-35158-210-7, e-ISBN: 978-2-35158-206-0; Ed. Alexandra Bertron PRO 124 (online version): Fourth International Conference Progress of Recycling in the Built Environment (2018), e-ISBN: 978-2-35158-208-4; Eds. Isabel M. Martins, Carina Ulsen, Yury Villagran PRO 125 (online version): SLD4—4th International Conference on Service Life Design for Infrastructures (2018), e-ISBN: 978-2-35158-213-8; Eds. Guang Ye, Yong Yuan, Claudia Romero Rodriguez, Hongzhi Zhang, Branko Savija PRO 126: Workshop on Concrete Modelling and Material Behaviour in honor of Professor Klaas van Breugel (2018), ISBN: 978-2-35158-214-5, e-ISBN: 978-235158-215-2; Ed. Guang Ye PRO 127 (online version): CONMOD2018—Symposium on Concrete Modelling (2018), e-ISBN: 978-2-35158-216-9; Eds. Erik Schlangen, Geert de Schutter, Branko Savija, Hongzhi Zhang, Claudia Romero Rodriguez PRO 128: SMSS2019—International Conference on Sustainable Materials, Systems and Structures (2019), ISBN: 978-2-35158-217-6, e-ISBN: 978-2-35158-218-3 PRO 129: 2nd International Conference on UHPC Materials and Structures (UHPC2018-China), ISBN: 978-2-35158-219-0, e-ISBN: 978-2-35158-220-6
RILEM Publications
xlvii
PRO 130: 5th Historic Mortars Conference (2019), ISBN: 978-2-35158-221-3, eISBN: 978-2-35158-222-0; Eds. José Ignacio Álvarez, José María Fernández, Íñigo Navarro, Adrián Durán, Rafael Sirera PRO 131 (online version): 3rd International Conference on Bio-Based Building Materials (ICBBM2019), e-ISBN: 978-2-35158-229-9; Eds. Mohammed Sonebi, Sofiane Amziane, Jonathan Page PRO 132: IRWRMC’18—International RILEM Workshop on Rheological Measurements of Cement-based Materials (2018), ISBN: 978-2-35158-230-5, e-ISBN: 978-2-35158-231-2; Eds. Chafika Djelal, Yannick Vanhove PRO 133 (online version): CO2STO2019—International Workshop CO2 Storage in Concrete (2019), e-ISBN: 978-2-35158-232-9; Eds. Assia Djerbi, Othman OmikrineMetalssi, Teddy Fen-Chong PRO 134: 3rd ACF/HNU International Conference on UHPC Materials and Structures—UHPC’2020, ISBN: 978-2-35158-233-6, e-ISBN: 978-2-35158-234-3; Eds. Caijun Shi and Jiaping Liu
RILEM Reports (REP) Report 19: Considerations for Use in Managing the Aging of Nuclear Power Plant Concrete Structures (ISBN: 2-912143-07-1); Ed. D. J. Naus Report 20: Engineering and Transport Properties of the Interfacial Transition Zone in Cementitious Composites (ISBN: 2-912143-08-X); Eds. M. G. Alexander, G. Arliguie, G. Ballivy, A. Bentur and J. Marchand Report 21: Durability of Building Sealants (ISBN: 2-912143-12-8); Ed. A. T. Wolf Report 22: Sustainable Raw Materials—Construction and Demolition Waste (ISBN: 2-912143-17-9); Eds. C. F. Hendriks and H. S. Pietersen Report 23: Self-Compacting Concrete state-of-the-art report (ISBN: 2-912143-233); Eds. Å. Skarendahl and Ö. Petersson Report 24: Workability and Rheology of Fresh Concrete: Compendium of Tests (ISBN: 2-912143-32-2); Eds. P. J. M. Bartos, M. Sonebi and A. K. Tamimi Report 25: Early Age Cracking in Cementitious Systems (ISBN: 2-912143-33-0); Ed. A. Bentur Report 26: Towards Sustainable Roofing (Joint Committee CIB/RILEM) (CD 07) (e-ISBN 978-2-912143-65-5); Eds. Thomas W. Hutchinson and Keith Roberts Report 27: Condition Assessment of Roofs (Joint Committee CIB/RILEM) (CD 08) (e-ISBN 978-2-912143-66-2); Ed. CIB W 83/RILEM TC166-RMS
xlviii
RILEM Publications
Report 28: Final report of RILEM TC 167-COM ‘Characterisation of Old Mortars with Respect to Their Repair (ISBN: 978-2-912143-56-3); Eds. C. Groot, G. Ashall and J. Hughes Report 29: Pavement Performance Prediction and Evaluation (PPPE): Interlaboratory Tests (e-ISBN: 2-912143-68-3); Eds. M. Partl and H. Piber Report 30: Final Report of RILEM TC 198-URM ‘Use of Recycled Materials’ (ISBN: 2-912143-82-9; e-ISBN: 2-912143-69-1); Eds. Ch. F. Hendriks, G. M. T. Janssen and E. Vázquez Report 31: Final Report of RILEM TC 185-ATC ‘Advanced testing of cement-based materials during setting and hardening’ (ISBN: 2-912143-81-0; e-ISBN: 2-91214370-5); Eds. H. W. Reinhardt and C. U. Grosse Report 32: Probabilistic Assessment of Existing Structures. A JCSS publication (ISBN 2-912143-24-1); Ed. D. Diamantidis Report 33: State-of-the-Art Report of RILEM Technical Committee TC 184-IFE ‘Industrial Floors’ (ISBN 2-35158-006-0); Ed. P. Seidler Report 34: Report of RILEM Technical Committee TC 147-FMB ‘Fracture mechanics applications to anchorage and bond’ Tension of Reinforced Concrete Prisms—Round Robin Analysis and Tests on Bond (e-ISBN 2-912143-91-8); Eds. L. Elfgren and K. Noghabai Report 35: Final Report of RILEM Technical Committee TC 188-CSC ‘Casting of Self Compacting Concrete’ (ISBN 2-35158-001-X; e-ISBN: 2-912143-98-5); Eds. Å. Skarendahl and P. Billberg Report 36: State-of-the-Art Report of RILEM Technical Committee TC 201-TRC ‘Textile Reinforced Concrete’ (ISBN 2-912143-99-3); Ed. W. Brameshuber Report 37: State-of-the-Art Report of RILEM Technical Committee TC 192-ECM ‘Environment-conscious construction materials and systems’ (ISBN: 978-2-35158053-0); Eds. N. Kashino, D. Van Gemert and K. Imamoto Report 38: State-of-the-Art Report of RILEM Technical Committee TC 205-DSC ‘Durability of Self-Compacting Concrete’ (ISBN: 978-2-35158-048-6); Eds. G. De Schutter and K. Audenaert Report 39: Final Report of RILEM Technical Committee TC 187-SOC ‘Experimental determination of the stress-crack opening curve for concrete in tension’ (ISBN 978-2-35158-049-3); Ed. J. Planas Report 40: State-of-the-Art Report of RILEM Technical Committee TC 189-NEC ‘Non-Destructive Evaluation of the Penetrability and Thickness of the Concrete Cover’ (ISBN 978-2-35158-054-7); Eds. R. Torrent and L. Fernández Luco Report 41: State-of-the-Art Report of RILEM Technical Committee TC 196-ICC ‘Internal Curing of Concrete’ (ISBN 978-2-35158-009-7); Eds. K. Kovler and O. M. Jensen
RILEM Publications
xlix
Report 42: ‘Acoustic Emission and Related Non-destructive Evaluation Techniques for Crack Detection and Damage Evaluation in Concrete’—Final Report of RILEM Technical Committee 212-ACD (e-ISBN: 978-2-35158-100-1); Ed. M. Ohtsu Report 45: Repair Mortars for Historic Masonry—State-of-the-Art Report of RILEM Technical Committee TC 203-RHM (e-ISBN: 978-2-35158-163-6); Eds. Paul Maurenbrecher and Caspar Groot Report 46: Surface delamination of concrete industrial ffioors and other durability related aspects guide—Report of RILEM Technical Committee TC 268-SIF (e-ISBN: 978-2-35158-201-5); Ed. Valerie Pollet
Technical Committee 264-RAP (“Asphalt Pavement Recycling”)
Effect of Aging on the Rheological Properties of Blends of Virgin and Rejuvenated RA Binders Di Wang, Augusto Cannone Falchetto, Martin Hugener, Laurent Porot, Atsushi Kawakami, Bernhard Hofko, Andrea Grilli, Emiliano Pasquini, Marco Pasetto, Hassan Tabatabaee, Huachun Zhai, Margarida Sá da Costa, Hilde Soenen, Patricia Kara De Maeijer, Wim Van den Bergh, Fabrizio Cardone, Alan Carter, Kamilla L. Vasconcelos, Xavier Carbonneau, Aurelie Lorserie, Goran Mladenovi´c, Marko Oreškovi´c, Tomas Koudelka, Pavel Coufalik, Edoardo Bocci, Runhua Zhang, Eshan V. Dave, and Gabriele Tebaldi Abstract The use of rejuvenators has seen a consistent increase over the years in the asphalt pavement industry. This is due to the need for maximizing the demand for incorporating a higher amount of Reclaimed Asphalt (RA) in the pavement asphalt mixtures. In order to tackle this challenge, the Task Group 3, focusing on asphalt D. Wang Technische Universität Braunschweig, Braunschweig, Germany A. Cannone Falchetto (B) University of Alaska Fairbanks, Fairbanks, AK, USA e-mail: [email protected] M. Hugener Empa—Materials Science and Technology, Dübendorf, Switzerland L. Porot Kraton Chemical B.V., Almere, The Netherlands A. Kawakami Public Works Research Institute, Tsukuba, Japan B. Hofko Institute of Transportation, TU Vienna, Vienna, Austria A. Grilli University of the Republic of San Marino, San Marino, Republic of San Marino E. Pasquini · M. Pasetto University of Padova, Padova, Italy H. Tabatabaee Cargill, Wayzata, MN, USA H. Zhai Idaho Asphalt Supply Inc., Idaho Falls, ID, USA M. S. da Costa National Laboratory of Civil Engineering (LNEC), Lisbon, Portugal © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_1
3
4
D. Wang et al.
binders and additives, of the RILEM TC RAP conducted an interlaboratory activity to evaluate the effect of aging on blends of virgin and rejuvenated RA binders. A set of conventional tests including penetration value at 25 °C, softening point temperature and rheological measurements by means of Dynamic Shear Rheometer (DSR) were selected to involve a large number of participants. A binder, recovered from field RA, was treated with a bio-based rejuvenator and blended with a virgin binder to simulate recycling at three different percentages 60, 80, and 100%. These blends, as well as the pure virgin binder, were next short and long term aged and tested to evaluate the changes in the corresponding properties. Relatively consistent results were obtained for the entire set of blends. The aging of combined rejuvenated RA and virgin binders was comparable to that experienced for the pure virgin binder. DSR data provided a more precise evolution of the impact of aging on the materials. Keywords Asphalt binder · Reclaimed asphalt (RA) · Rejuvenator · DSR
1 Introduction In the considerations of both environmental and economic benefits, different types of recycled, industrial by-products materials have been used for the asphalt pavement H. Soenen NYNAS, Antwerp, Belgium P. K. De Maeijer · W. Van den Bergh University of Antwerp, Antwerp, Belgium F. Cardone Polytechnic University of Marche, Ancona, Italy A. Carter LCMB-ETS Montreal, Montreal, Canada K. L. Vasconcelos Laboratory of Pavement Technology—University of São Paulo, São Paulo, Brazil X. Carbonneau · A. Lorserie CST COLAS, Magny-les-Hameaux, France G. Mladenovi´c · M. Oreškovi´c Faculty of Civil Engineering, University of Belgrade, Belgrade, Serbia T. Koudelka · P. Coufalik Brno University of Technology, Brno, Czech Republic E. Bocci eCampus University, Novedrate, Italy R. Zhang · E. V. Dave University of New Hampshire, Durham, NH, USA G. Tebaldi University of Parma, Parma, Italy
Effect of Aging on the Rheological Properties …
5
construction for several decades [1] making reclaimed asphalt pavement materials one of the most reused construction materials [2, 3]. Different technologies can be selected to maximize the use of the Reclaimed Asphalt (RA) within the Hot Mix Asphalt (HMA) in a laboratory environment. However, due to the aged asphalt binder with higher stiffness and brittleness, the reuse of a high amount of RA can lead to potentially poorer performance of asphalt pavements against cracking [2, 4]. Hence, the use of RA in HMA is commonly limited to 25–50% for construction purposes [4]. In the upcoming decades, more and more RA have to be recycled and reused due to environmental and economic reasons. The current degree of recycling undoubtedly limits the design possibilities of pavement engineers and road authorities. Therefore, novel technologies are needed to maximize the value of RA up to a higher level. The use of rejuvenators represents an alternative solution to increase the rate of RA or the use of very hard RA, minimizing the impact of the recycled material on the performance properties of the mixtures and pavements [5]. In several previous studies [6–8], it has been successfully demonstrated that some rejuvenators can substantially restore the properties of aged asphalt binder with limited impact on its temperature susceptibility. With respect to the use of a high level of RA in asphalt pavement, RILEM established a specific Technical Committee (TC): 264-RAP (Asphalt Pavement Recycling). Within the framework of this TC, Task Group 3 on Asphalt Binder for Recycled Asphalt Mixtures is devoting its activities on the investigation of the effect of restoring the lost properties of aged RA binders by using rejuvenators. A total of 20 laboratories have committed to participate in this group; some outcomes have been already reported [9, 10]. This paper mainly focuses on presenting and analyzing the results from the rheological tests.
2 Experimentation 2.1 Materials The RA binder was extracted with Trichloroethene and recovered according to EN 12697-1 and EN 12697-3 standard by one single laboratory, from an RA material source collected in Germany. The selected rejuvenator was a bio-based additive; its specific amphipathic chemical structure, disperses the highly polar fractions limiting the agglomeration of asphaltenes [11]. A 50/70 paving grade binder (EN 12591) was used as reference material. The proposed procedure ensures that a single source of material can be used by all the laboratories throughout the entire research activity. Additional information on the selected materials can be found elsewhere [10, 12]. After extraction RA binder was melted at 160 °C and blended with 8% of rejuvenator manually mixing the blend at 160 °C for one minute. The rejuvenated binder was tested as it is (100 RA) or with a 50/70 paving grade bitumen. The blends consisted
6
D. Wang et al.
of 60% and 80% of the rejuvenated binder and a complementary part of 50/70 which were named 60 RA and 80 RA, respectively.
2.2 Testing Plan The testing plan consisted of evaluating different blends of RA binder which were treated with an optimal amount of rejuvenator and virgin binder in a ratio of 60, 80, and 100%. The optimal rejuvenator content was determined in a preliminary study [9]. An initial analysis of basic properties confirmed that through the optimal rejuvenator dosage it is possible to recover the penetration value of a target 50/70 fresh blend material as a function of the three different blend ratios (60 RA, 80 RA, and 100 RA) [10]. In addition, to better evaluate the aging effect on the performance properties of virgin binder and blended binders, artificial aging procedures were conducted in the laboratory environment. Different studies have begun evaluating binder durability at extended aging times (40 h of PAV aging or longer) as potentially more indicative of binder long term performance of binders modified with additives [13]. However, in this research, standard short and long-term aging was performed for each material with the Rolling Thin Film Oven Test (RTFOT) and Pressure Aging Vessel (PAV) procedures based on EN 12607-1 and EN 14769, respectively. For each material under different aging levels, the penetration values at 25 °C (EN 1426) and the softening point temperature (EN 1427) were measured. Temperaturefrequency-sweep (T-f-sweep) tests were performed on each binder by using the Dynamic Shear Rheometer (DSR) (EN 14770). Testing was performed at temperatures, ranging between −6 and +82 °C every 6 °C, with the 4, 8 and 25 mm plates. A frequency sweep from 0.1 to 10 Hz was used. Since 20 different laboratories from different regions were involved in this research, each laboratory selected only a single blending ratio among the three proposed. Among the participants, 16 laboratories provided softening point temperature and DSR results and 11 the penetration values.
3 Results and Analysis 3.1 Penetration Value and Softening Point As mentioned before, the use of the rejuvenator can restore the fresh blend asphalt binder to the target properties (penetration); however, this capability may be mitigated over time. The experimental results of the penetration value and softening point under different aging conditions are analyzed and discussed in this section to evaluate the aging effect of rejuvenators. As an example, the trend of the virgin and 80% blend ratio from four laboratories under different aging conditions is shown in Fig. 1.
Effect of Aging on the Rheological Properties …
7
Fig. 1 Comparison among 50/70, 80 RA and RA binders under different aging conditions
Only limited differences can be observed among the four laboratories, while very similar trends are found for virgin and blend materials under different aging conditions. Hence, the relative change in properties is of the same order of magnitude. This trend can be also observed in the case of 60 and 100% blend ratios. Overall, no remarkable difference in basic properties was found between the three blend ratios at the different aging levels highlighting the role of the rejuvenator in restoring the consistency of blended RA binders. Results confirmed that the selected rejuvenator dosage allows restoring the consistency of RA binder at 25 °C and high service temperatures. It can be affirmed that the effect of aging on 80 RA is similar to that on 50/70.
3.2 Dynamic Shear Rheometer (DSR) Results The evolution of DSR results is reported as a function of the different aging conditions in Figs. 2 and 3 for two different representative temperatures, 28 and 58 ºC. Results refer to the single frequency of 1.59 Hz. As in the previous section, the 80% blend ratio is presented as an example. The plot of complex shear modulus |G*| versus phase angle δ under different aging conditions is exhibited. The average value and the results from each laboratory are displayed. From Fig. 2, it can be observed that the participating laboratories had relatively similar measurements at both temperatures (28 and 58 ºC). For the different aging levels. Linear relationships are found between the logarithm of the complex shear modulus |G*| and phase angle δ. Overall, the effect of the rejuvenator in the blend ratio of 80% was preserved also at the different aging levels, as supported by the experimental measurements of the four laboratories. Similar results can be found in the blend ratio of 60 and 100%.
8
D. Wang et al.
Fig. 2 Complex shear modulus |G*| versus phase angle δ at a 28 ºC and b 58 ºC
A comparison among the different blends (50/70, 60 RA, 80 RA, and 100 RA) across the three aging levels at 28 and 58 ºC was performed reporting the average value of |G*| at 1.59 Hz in the bar charts of Fig. 3. A non-univocal trend is observed for |G*| at the two selected temperatures. While at 28 ºC the shear modulus of the rejuvenated material is generally lower than the virgin one regardless of aging conditions, an opposite trend is exhibited at 58 ºC. Further studies will investigate the rejuvenated binder properties in comparison with 50/70 over a wide range of temperatures through master curve and black diagram analysis.
4 Conclusions In the present study, the possibility of using rejuvenators to reuse, for the mix design a high amount of binder obtained from RA was experimentally investigated with the penetration grading system and DSR measurements. Based on the experimental results and analyses the following conclusions can be drawn:
Effect of Aging on the Rheological Properties … 1.E+07
Fig. 3 Complex shear modulus |G*| for virgin binder and blends at a 28 ºC and b 58 ºC
9
60RA
80RA
100RA
50/70
G* [Pa]
1.E+06 1.E+05 1.E+04 1.E+03 virgin
1.E+07
60RA
RTFOT
80RA
100RA
PAV
a)
PAV
b)
50/70
G* [Pa]
1.E+06 1.E+05 1.E+04 1.E+03 virgin
RTFOT
• The optimal rejuvenator dosage does not change when using different proportions of RA and 50/70 binder. A single optimal rejuvenator content can be used to partially restore the rheological properties at different blending ratios. • The use of rejuvenators can contribute to the recovery of the rheological properties of a blend of RA and virgin binders at intermediate temperatures while preserving the benefit at high temperatures. • Based on the standard aging procedure, the proposed DSR-graph with G* and phase angle can be used as a general method to test the performance of a rejuvenator in combination with an RA-binder, a virgin binder, and other additives. Although the results obtained in this study are promising, additional experimental support is needed by extending the present research effort to the investigation of more types of RA binders and different rejuvenator categories, as well as potentially extending the aging time. Acknowledgments The RILEM Technical Committee on Asphalt Pavement Recycling (RAP) 264 and the contributions of Nynas AB, Kraton Chemical B.V. and TU Braunschweig ISBS pavement laboratory staff to the RILEM interlaboratory activities are gratefully acknowledged.
10
D. Wang et al.
References 1. EAPA: Arguments to Stimulate the Government to Promote Asphalt Reuse and Recycling. https://eapa.org/wp-content/uploads/2018/07/arguments_stimulate_asphalt_May2008. pdf. Last accessed 2019/11/28 (2008) 2. McDaniel, R.S., Anderson, R.M.: Recommended Use of Reclaimed Asphalt Pavement in the Superpave Mix Design Method: Technician’s Manual (No. Project D9-12 FY’97). National Research Council (US). Transportation Research Board (2001) 3. EAPA, European Asphalt Pavement Association: Asphalt in Figures. https://eapa.org/wp-con tent/uploads/2018/12/AIF_2017.pdf. Last accessed 2019/10/15 (2017) 4. TL Asphalt-StB: Technische Lieferbedingung für Asphaltmischgut für den Bau von Verkehrsflächenbefestigung. FGSV Verlag (2013). (In German) 5. EAPA, European Asphalt Pavement Association: Recommendations for the Use of Rejuvenators in Hot and Warm Asphalt Production, EAPA, Brussels, August 2018 6. Mogawer, W.S., Booshehrian, A., Vahidi, S., Austerman, A.J.: Evaluating the effect of rejuvenators on the degree of blending and performance of high RAP, RAS, and RAP/RAS mixtures. Road Mater. Pavement Des. 14(sup2), 193–213 (2013) 7. Zaumanis, M., Mallick, R.B., Poulikakos, L., Frank, R.: Influence of six rejuvenators on the performance properties of reclaimed asphalt pavement (RAP) binder and 100% recycled asphalt mixtures. Constr. Build. Mater. 71, 538–550 (2014) 8. Grilli, A., Bocci, E., Bocci, M.: Hot Recycling of Reclaimed Asphalt using a Bio-based Additive, 8th International RILEM Symposium on Testing and Characterization of Sustainable and Innovative Bituminous Materials, October 7–9. Ancona, Italy (2015) 9. Cannone Falchetto, A., Porot, L., Riccardi, C., Hugener, M., Tebaldi, G., Dave, E.: Effects of Rejuvenator on Reclaimed Asphalt Binder: An Exploratory Study of the RILEM TC 264-RAP Task Group 3. In RILEM 252-CMB-Symposium (pp. 195–200). Springer, Cham (2018) 10. Porot, L., Hugener, M., Cannone Falchetto, A., Wang, D., Kawakami, A., Tebaldi, G.: Aging of Rejuvenated RAP Binder—A RILEM Inter-Laboratory Study. In: 7th Eurasphalt and Eurobitume Congress, 12–14 May 2020, Madrid 11. Porot, L., Broere, D., Wistuba, M., Gronniger, M.: Asphalt and Binder Evaluation of Asphalt Mix with 70 % Reclaimed Asphalt. European Asphalt Technology Association (EATA) 2017, Dubendorf (2017) 12. Wang, D., Koziel, M., Cannone Falchetto, A., Riccardi, C., Hugener, M., Porot, L., Kim, Y.S., Cheraghian, G., Wistuba, M.: Experimental investigation on the effect of rejuvenator on the use of a high amount of recycled asphalt binder. In: 9th International Conference on Maintenance and Rehabilitation of Pavements (MAIREPAV9), 1–3 July 2020, Zurich 13. Asphalt Institute Technical Advisory Committee, State-of-the-Knowledge: Use of the Delta T c Parameter to Characterize Asphalt Binder Behavior (2019). https://www.asphaltinstitute. org/download/8139/. Last accessed 2020/02/06
Experimental Investigation on Water Loss and Stiffness of CBTM Using Different RA Sources Andrea Grilli , Simone Raschia , Daniel Perraton , Alan Carter , Amir Rahmanbeiki, Patricia Kara De Maeijer , Davide Lo Presti , Gordon Airey , Chibuike Ogbo , Eshan V. Dave , and Gabriele Tebaldi Abstract Cold recycling of reclaimed asphalt (RA) is a promising technique to build or to maintain roads, combining performance and environmental advantages. Although this technique has been extensively used worldwide, there is no unique and internationally-shared method to characterize cold recycled mixtures. The previous work of the RILEM TC 237-SIB TG6 successfully attempted to characterize different RA sources with both traditional parameters (gradation, bitumen content and geometrical properties) and non-conventional properties (fragmentation and strength testing). The current RILEM TC 264-RAP TG1 mainly focuses on the influence of different RA sources on physical and mechanical characteristics of cement-bitumen treated materials (CBTM) using foam or emulsified bitumen, taking into consideration compaction and curing methods. This paper presents results from the first step of the inter-laboratory project in which foamed bitumen and cement were A. Grilli (B) Department of Economics, Science and Law, University of the Republic of San Marino, Via Consiglio dei Sessanta 99, 47891 San Marino, Republic of San Marino e-mail: [email protected] S. Raschia · D. Perraton · A. Carter · A. Rahmanbeiki Department of Construction Engineering, École de Technologie Supérieure (ÉTS), 1100 Notre-Dame, Ouest, Montreal, Canada P. Kara De Maeijer EMIB, Faculty of Applied Engineering, University of Antwerp, Groenenborgerlaan 171, 2020 Antwerp, Belgium D. Lo Presti Department of Engineering, University of Palermo, Viale Delle Scienze, 90128 Palermo, Italy D. Lo Presti · G. Airey Nottingham Transportation Engineering Centre, University of Nottingham, Nottingham NG7 2RD, UK C. Ogbo · E. V. Dave Department of Civil and Environmental Engineering, University of New Hampshire, 33 Academic Way, Durham, NH 03824, USA G. Tebaldi Department of Engineering and Architecture, University of Parma, Parco Area delle Scienze 181/a, 43124 Parma, Italy © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_2
11
12
A. Grilli et al.
used as binders. The influence of two RA sources, one from Alabama (USA) and one from San Marino, were investigated through the collaboration of several laboratories. Specimens were manufactured with the same diameter by means of both Marshall and gyratory compactors and then cured following two procedures: free surface drying (FSD) and partially-surface drying (PSD). A preliminary study allowed obtaining specimens with similar volumetric properties. Along with compactability and water loss, the indirect tensile stiffness modulus was measured and analyzed. The results have shown that the RA source and curing procedure influence the CBTM mechanical properties. Keywords Cold recycling · Foamed bitumen · Reclaimed asphalt · Strength · Water loss
1 Introduction Cement-bitumen treated materials (CBTM) are produced with a significant amount of reclaimed asphalt (RA), bitumen emulsion or foamed bitumen, cement and water. CBTM is a hybrid material which inherits properties from both asphalt concrete (AC), since the bituminous component plays a significant role, and cement-treated materials (CTM), since the cement influences stiffness and curing of the mixture [1, 2]. The RILEM TC 237-SIB TG6 on cold recycling worked in 2012–2018 on the characterization of reclaimed asphalt to be used for cold recycling [3]. This characterization differs from the commonly used RA characterization when used in hot recycling where the aged bitumen softens, and the aggregate gradation combines with the virgin materials that usually are the dominant component. In cold recycling, the RA behaves as it is and for instance black gradation, fragmentation resistance and geometric properties of particles have a significant influence on the performance of the final product. At the end of the TG6 mandate, the RILEM TC 264-RAP launched a new TG, so-called TG1, on cold recycling with the purposes of sharing mix design and curing procedures that can better support the selection of the optimum mixture in relationship with RA characteristics and product performance. A new inter-laboratory testing program has been proposed to the TG1 members to evaluate the influence of different RA sources, appropriately selected and characterized, on strength and stiffness of cold recycled mixtures. The objectives of this paper were to assess both the influence of the RA source and curing conditions on properties of CBTM, such as water loss and stiffness.
Experimental Investigation on Water Loss …
13
2 Materials Two RA sources, one from the Republic of San Marino so-called RA1 and another one from NCAT-USA so-called RA2, were preliminary characterized by means of the fragmentation test which measures the amount of produced fines (passing material to 1.7 mm sieve size) after modified Proctor compaction [4]. The RA1 and RA2 properties are reported in Table 1. The two RA sources mainly differ each other by fragmentation resistance, gradation and maximum size. A mineral filler characterized by 91% of passing material at 0.008 mm sieve size, 50.1% voids of dry compacted filler measured by means of a Rigden apparatus (EN 1097–4) and a stiffening power R&B of 11.5 °C (EN 13179–1) was selected to adjust both RA gradations following the respective maximum density curves. A GU type cement (CSA A3000) with compressive strength at 28 days of 43.9 MPa (ASTM C109) and Blaine surface area of 410 cm2 /g was chosen to improve the bituminous mastic consistency and short-term resistance. The bitumen has a penetration value of 64 mm·10–1 (EN 1426) and a softening point of 52 °C (EN 1427). A preliminary study established 1.9% of water by bitumen weight to achieve at 170 °C an expansion ratio of 10 and half-life of 6 s considered suitable for foam applications. The two CBTM mixtures consisted of 4.0% of added water, 3.0% of bitumen, 1.5% of cement, 5.5% and 10.0% of filler and 93.0% and 88.5% of RA, for RA1 and RA2 respectively. Dosages refer to the total solid weight (RA, filler and cement). The dosages of water and bitumen were chosen according to common values typically used for cold recycled material [5, 6]. Mixtures were coded as CBTM16RA1 and CBTM10RA2 using RA1 and RA2 respectively. CBTM16RA1 and CBTM10RA2 gradation follow the Fuller distribution with Dmax = 16 mm and Dmax = 10 mm, respectively. Table 1 Properties of RA1 and RA2 Property
Standard
Unit
RA1
RA2
Bitumen content
ASTM D6307
%
5.51
5.49
Nominal maximum particle size
ASTM D448-03
mm
16
10
Maximum specific gravity
ASTM C127-128
–
2.482
2.498
Water absorption
ASTM C127-128
%
1.10
1.08
Fragmentation @5 °C
ASTM D1557
%
7.6
–
Fragmentation @20 °C
ASTM D1557
%
6.7
13.9
Flakiness index
EN 933-3
%
5
–
Shape index
EN 933-4
%
3
–
14
A. Grilli et al.
3 Testing Methods and Procedures According to a preliminary study, the two mixtures were compacted to get a representative air void content of 14.3% ± 1% using both Marshall and Shear Gyratory Compactor (SGC). The compaction with Marshall was performed employing around 1100 g of mixture placed in the mould and compacted with 50 blows on each side. The same amount of material was used for specimens compacted with the SGC. In this case, the compaction was carried out at fixed specimen height, 67.7 mm, which was reached with an average of 40 gyrations for mixtures. Compacted specimens were demoulded and divided into two subseries following two protocols: free surface drying (FSD) and partially sealed drying (PSD). FSD required specimens to be cured in the air so they can dry from all sides, whereas PSD required to seal the cylinder sides of the specimen by a plastic film to allow drying only from the specimen top surface. Specimens were cured in a climatic chamber for 14 days at 40 ± 2 °C and 55 ± 5% of relative humidity and weighted every day to monitor the water loss over time. Both mixtures were characterized in terms of the indirect tensile stiffness modulus (ITSM) test at 2, 10 and 20 °C (EN 12697-26). ITSM was measured after 3 h conditioning at the testing temperature. At least 3 test repetitions for each series were performed.
4 Results 4.1 Water Loss Figure 1 shows the average values of water loss as a function of curing days at 40 °C for both CBTM16RA1 and CBTM10RA2. The results shown are related only to Lab.1 and each curve is an average of 10 specimens. Water loss evolution appears to follow a bilinear trend changing rate significantly after 3-day curing period. As can
Fig. 1 Experimental data of water loss for the studied mixtures
Experimental Investigation on Water Loss …
15
Table 2 Parameters obtained by a bilinear regression on water loss results Series
Regression factors 0-3d (phase 1)
Regression factors 3-14d (phase 2)
Phase1 versus phase2
PSD versus FSD
s1
k1
R2
s2
k2
R2
s2 /s1 (%)
days
CBTM16RA1 Marshall FSD
−0.823
3.845
0.934
−0.025
1.629
0.858
3.094
–
CBTM10RA2 Marshall FSD
−0.856
3.724
0.952
−0.010
1.381
0.779
1.116
–
CBTM16RA1 Marshall PSD
−0.499
3.909
0.962
−0.085
2.731
0.977
16.988
5
CBTM10RA2 Marshall PSD
−0.563
3.798
0.951
−0.051
2.341
0.871
9.027
10
CBTM16RA1 −0.516 SGC 100 FSD
3.977
0.956
−0.024
2.494
0.854
4.695
–
CBTM10RA2 −0.507 SGC 100 FSD
3.525
0.739
−0.016
2.222
0.692
3.119
–
CBTM16RA1 −0.417 SGC 100 PSD
4.001
0.973
−0.074
2.885
0.967
17.720
–
CBTM10RA2 −0.406 SGC 100 PSD
3.635
0.786
−0.058
2.684
0.944
14.264
–
be noticed in Fig. 1, the general trend does not seem influenced by the compaction method and mix type, whereas the curing method influenced the water loss evolution, considerably. PSD method showed a lower water loss than FDS method. At the same time, Marshall compacted specimens showed higher water loss than SGC compacted specimens, probably caused by the different intergranular voids dispersion and size. The experimental data were analyzed by a bilinear regression considering the intersection point at a 3-day curing time and Table 2 shows the regression factors (s = slope; k = intercept value and R2 = regression coefficient) for all series. FSD implied a fast decrease of water content until 3 days (phase 1) then a rather flat evolution (phase 2). This evolution was less remarkable for PSD series. FSD showed a higher rate difference between phase 1 and phase 2 than PSD. By extending the regression equation, PSD required more than 5 additional days to reach the FSD final water content.
4.2 ITSM Figure 2 depicts the ITSM values at 2, 10 and 20 °C for all series. It can be affirmed that PSD allows higher ITSM values than FSD. Moreover, CBTM10RA2 seems stiffer than CBTM16RA1 as well as SGC specimens seems to have higher stiffness
16
A. Grilli et al.
Fig. 2 ITSM results for: a CBTM16RA1, b CBTM10RA2
than Marshall specimens. Slow water evaporation and a higher water content allows more effective cement hydration to be achieved with direct effect on stiffness values. However, the results from different laboratories showed a notable scattering of values. Probably 3 repetitions for each series are not enough and the reproducibility of the testing method has to be improved.
5 Conclusions This paper shows the results from the first phase of the RILEM 264-RAP TG1 interlaboratory project focused on cold recycling. Particularly, the influence of two RA sources in two mixtures using foamed bitumen and cement as binders were investigated. The results showed that even if RA2 has fragmentation value poorer than RA1, the correction of the gradation curve through mineral filler following the respective maximum density gradation (Fuller distribution) allows CBTM10RA2 to behave similarly to CBTM16RA1. Water loss evolution followed a bilinear trend changing rate after a 3-day curing period, significantly. FSD induced a fast decrease of water content until 3 days (phase 1) then a rather flat evolution (phase 2) with the highest rate difference between phase 1 and phase 2. PSD prevented water evaporation and potentially required more than 5 additional days to reach the FSD final water content. The higher water content probably allows a more effective cement hydration in PSD entailing higher ITSM values than FSD. Moreover, SGC specimens seems to show
Experimental Investigation on Water Loss …
17
higher stiffness values than Marshall specimens. However, the results from different laboratories showed a remarkable scattering of values and reproducibility of the testing method has to be improved. Acknowledgement The RILEM Technical Committee on Asphalt Pavement Recycling (RAP) 264 would like to thank CBR (Cooperativa Braccianti Riminese, Italy), NYNAS AB and Willemen Infra NV (Belgium) for their collaboration.
References 1. Cardone, F., Grilli, A., Bocci, M., Graziani, A.: Curing and temperature sensitivity of cement– bitumen treated materials. Int. J. Pavement Eng. 16(10), 868–880 (2014). https://doi.org/10. 1080/10298436.2014.966710 2. Grilli, A., Graziani, A., Bocci, M.: Compactability and thermal sensitivity of cement–bitumentreated materials. Road Mater. Pavement Des. 13(4), 599–617 (2012). https://doi.org/10.1080/ 14680629.2012.742624 3. Tebaldi, G., Dave, E.V., Falchetto, A.C., Hugener, M., Perraton, D., Grilli, A., Lo Presti, D., Pasetto, M., Loizos, A., Jenkins, K.: Recommendation of RILEM TC237-SIB on fragmentation test for recycled asphalt. Mater. Struct. 52(4), 82 (2019) 4. Perraton, D., Tebaldi, G., Dave, E.V., Bilodeau, F., Giacomello, G., Grilli, A., Graziani, A., Bocci, M., Grenfell, J., Muraya, P.: Tests campaign analysis to evaluate the capability of fragmentation test to characterize recycled asphalt pavement (RAP) material. In: Presented at 8th RILEM International Symposium on Testing and Characterization of Sustainable and Innovative Bituminous Materials, 2016. Springer 5. Fu, P., Jones, D., Harvey, J.T., Bukhari, S.A.: Laboratory test methods for foamed asphalt mix resilient modulus. Road Mater. Pavement Des. 10(1), 188–212 (2009) 6. Grilli, A., Graziani, A., Bocci, E., Bocci, M.: Volumetric properties and influence of water content on the compactability of cold recycled mixtures. Mater. Struct. 49(10), 4349–4362 (2016). https://doi.org/10.1617/s11527-016-0792-x
International Evaluation of the Performance of Warm Mix Asphalts with High Reclaimed Asphalt Content Paul Marsac, Edoardo Bocci, Fabrizio Cardone, Augusto Cannone Falchetto, Xavier Carbonneau, Martins Zaumanis, Alan Carter, Ma Carmen Rubio-Gámez, Miguel Del Sol Sánchez, Eshan V. Dave, and Gabriele Tebaldi Abstract The Task Group 2 of the RILEM Technical Committee 264-RAP on non-cold recycling identified the need to gather some insights about possible durability issues associated with the combination of Warm Mix Asphalt (WMA) and of Reclaimed Asphalt (RA) in order to prioritize the characterization or research needs for relevant damage types. For this purpose, an international inter-laboratory testing program was launched in 5 different laboratories. This was developed on the basis of the challenges which a road material designer commonly faces: reduce the temperature, add RA, but preserve the same properties. Each laboratory characterized, according to its internal protocol, two variations of the same type of mixture P. Marsac (B) IFSTTAR, Nantes, France e-mail: [email protected] E. Bocci Università eCampus, Novedrate, Italy F. Cardone Università Politecnica delle Marche, Ancona, Italy A. C. Falchetto University of Alaska Fairbanks, Fairbanks, AK, USA X. Carbonneau CST Colas, Magny-les-Hameaux, France M. Zaumanis EMPA, Dübendorf, Switzerland A. Carter Ecole de Technologie Supérieure, Montreal, Canada M. C. Rubio-Gámez · M. Del Sol Sánchez University of Granada, Granada, Spain E. V. Dave University of New Hampshire, Durham, NH, USA G. Tebaldi University of Parma, Parma, Italy © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_3
19
20
P. Marsac et al.
(similar grading curve, binder content, etc.): a Hot Mix Asphalt (HMA) without RA taken as reference and a WMA with 40% RA. The performances of all the WMA with 40% RA are then compared in order to see if common trends emerge from the different characterization methods and reveal property issues specific to WMA + RA combinations. Keywords Reclaimed asphalt · Warm mix asphalt · Recycling · Inter-laboratory testing · Performance evaluation
1 Introduction The combination of Warm Mix Asphalt (WMA) processes with the recycling of Reclaimed Asphalt (RA) is a promising solution to mitigate the environmental impact of bituminous materials used in road construction and maintenance. It could undoubtedly bring immediate benefits through energy savings and resource preservation during material production. However, the durability of WMA with RA is of paramount importance in the context of life-cycle cost and global warming potential assessment, as premature failures could largely overshadow the initial production savings. This major durability aspect is much more challenging to assess than production aspects. This is because it encompasses many time-related distresses such as fatigue, aging and moisture-related deterioration which can only be estimated or simulated in questionable ways and will actually be known only at the end of the service life of the material in the road structure. In this context, the Task Group 2 of the RILEM Technical Committee 264-RAP on non-cold recycling identified the need to gather better insights on various durability issues associated with WMA+RA combination in order to prioritize the characterization or research needs for relevant damage types. For this purpose, an international inter-laboratory testing program was launched.
2 Inter-laboratory Testing Program 2.1 General Principle As the objective of the Task Group 2 is not focused on the properties of a particular bituminous material but on the effect of the combination of WMA+RA on the overall properties of the bituminous material, it was initially decided not to choose a specific common material for the inter-laboratory testing as usual practice for a round-robin test. By contrast, the use of a WMA process and the recycling of RA are required. This implies that the results of the laboratories on different materials are not comparable in terms of absolute values. For this reason, parallel measurements on a hot mix asphalt with the same design parameters taken as local reference are also required. In this
International Evaluation of the Performance …
21
way, the relative effect of the combination of WMA +RA on the mix properties can be evaluated and compared among the different laboratories. However, for the effects to be comparable, the causes must also be similar. Consequently, for a proper comparison purpose it was also decided to use the same RA material, virgin binders and RA amount fixed at 40%.
2.2 Material The RA samples were delivered in 25 kg bags. RA material was sampled from the same RA stockpile on a mixing plant with an industrial quartering machine to assure a good homogeneity among the different bags. The virgin binders provided by Nynas (50/70 for the reference mixes and 160/220 for the mixes with 40% RA) were distributed with the RA. The 160/220 was chosen in order to obtain a 40/50 binder when blended with 40% RA binder according to the blending law of EN 13108-1 [1].
2.3 RA Characterization In the first step, RA was characterized in the different laboratories. As RA is the same for all, this step is similar to a round-robin test. The results are presented in Table 1. The results show some differences between the laboratories. Notably, the grading curves are remarkably different from one laboratory to another based on the reproducibility value reported in EN 12697-2 (R = 1.7%) [2]. This reflects either a difference in the samples or laboratory methods. Though an industrial quartering machine was used for the RA sampling, it is not possible to attribute the differences to the first or the second cause. In the framework of the testing program, the RA characterization Table 1 RA characterization results Lab
Binder content (%)
Penetration (1/10mm)
Softening point (°C)
Passing @0.063mm (%)
Passing @2 mm(%)
Passing @10 mm(%)
Lab#1
4.60
11
77.1
6
34
94
Lab#2
4.50
11
74.6
9
37
97
Lab#3
4.39
12
76.2
6
31
90(1)
Lab#4
4.44
7
76.6
10
43
95
Lab#5
4.45
–
–
8(2)
34(1)
90
(1) interpolated (2) @0.08 mm
22
P. Marsac et al.
Table 2 Types of mixes, processes and mixing temperatures chosen by the laboratories Lab
Mixture designation (EN 13108-1) [1]
Binder content (%)
Processes and mixing temperatures (°C) HMA (0% RA) WMA foam (40% RA)
WMA additive (40% RA)
Lab#1
AC 16 bin 50/70 5.0
150
–
130
Lab#2
AC 10 surf/bin 50/70
5.4
150
130
130
Lab#3
AC 8 surf 50/70 5.9
160
–
140
Lab#4
AC 14 base 50/70
4.1
155
–
130
Lab#5
AC 10 surf 50/70
4.5
160
–
140
was originally intended as a routine preparation step for the mix design. The results show that this is not necessarily a straightforward process. Yet it is a key element for the mix design in the recycling process as it determines the relevance of all the following stages.
2.4 Mix Design The different types of mixes as well as the WMA process chosen by the 5 participating laboratories are presented in Table 2. Overall, the panel encompasses a range of mix types from base to surface courses and the 2 main types of warm processes: WMA with foam and WMA with additive.
2.5 Tests and Results The different tests performed by the participating laboratories are summarized in Table 3 while the test results are presented in Tables 4, 5, 6, 7 and 8. According to the type of test, the relative effect is calculated with the ratio WMA/HMA if higher values reflect better properties (water sensitivity and modulus) and with the ratio HMA/WMA if lower values reflect better properties (gyratory compactor and rutting).
International Evaluation of the Performance …
23
Table 3 Testing plan performed by the laboratories Test
HMA
WMA foam (40% RA)
WMA additive (40% RA)
Gyratory compactor (EN 12697-31) [3]
Lab#1, 2
Lab#2
Lab#1, 2
Water sensitivity (EN 12697-12 A: ITSR) [4]
Lab#1, 2, 4
Lab#2
Lab#1, 2, 4
Water sensitivity (EN 12697-12 B: i/C)
Lab#2, 4
Lab#2
Lab#2, 4
Rutting (EN 12697-22 large size device) [5]
Lab#2, 3
Lab#2
Lab#2, 3
Stiffness Modulus-cyclic compression test (ASTM D3497) [6]
Lab#1
Stiffness Modulus-2 points bending (EN 12697-26) [7]
Lab#2
Lab#1
–
Lab#2
Table 4 Gyratory compaction test results: air void content (EN 12697-31) Lab
HMA (%)
WMA foam (40% RA)
WMA additive (40% RA)
Absolute (%)
Relative effect (HMA/WMA)
Absolute (%)
Relative effect (HMA/WMA)
Lab#1 @100 gyrations
4.2
–
–
4.0
1.05
Lab#2 @60 gyrations
9.2
8.4
1.10
8.9
1.03
Table 5 Water sensitivity test results (EN 12697-12 A) Lab
HMA (%)
WMA foam (40% RA)
WMA additive (40% RA)
Absolute (%)
Relative effect (WMA/HMA)
Absolute (%)
Relative effect (WMA/HMA)
–
99.6
1.13
1.00
91.9
1.06
–
88.6
1.02
Lab#1
88.4
–
Lab#2
86.6
86.3
Lab#4
87.0
–
Table 6 Water sensitivity test results (EN 12697-12 B) Lab
HMA (%)
WMA foam (40% RA)
WMA additive (40% RA)
Absolute (%)
Relative effect (WMA/HMA)
Absolute (%)
Relative effect (WMA/HMA)
1.11
98.5
1.14
–
88.1
0.99
Lab#2
86.7
95.9
Lab#4
89.3
–
24
P. Marsac et al.
Table 7 Rutting test results (EN 12697-22 large size device) Lab
HMA (%)
WMA foam (40% RA)
WMA additive (40% RA)
Absolute (%) Relative effect (HMA/WMA)
Absolute (%)
Relative effect (HMA/WMA)
Lab#2 @30,000c
4.6
4.6
1.00
4.8
0.96
Lab#3 @10,000c
5.9(1)
–
–
1.8
3.31
(1) binder content 5.9% for HMA and 4.85% for WMA
Table 8 Stiffness modulus |E*| test results (ASTM D3497 and EN 12697-26) Lab
HMA (MPa)
WMA additive (40% RA) Absolute (MPa)
Relative effect (WMA/HMA)
Lab#1 (ASTM D3497) @50°C, 0.1Hz
144
127
0.87
Lab#1 (ASTM D3497) @20°C, 1Hz
5380
4981
0.93
Lab#1(ASTM D3497) @5°C, 20Hz
28,175
23,077
0.82
Lab#2 (EN 12697–26) @15°C, 10Hz
9103
7726
0.85
2.6 Comments on the Results For all the tests but one the relative effects are greater than or equal to 1. This means that the mixes with 40% RA obtained with both WMA processes perform equal to or better than the reference HMA. Only the stiffness modulus values were higher for the reference HMA. This suggests that the soft binder (160/220) used for the WMA mixtures overcompensated the hard RA binder or/and that the lower mixing temperatures of the WMA processes reduce the aging of the binder. Though these results were not really expected at the initiation of the inter-laboratory testing, they are consistent with the conclusions of a study on 40% multi-recycling of HMA [8]. This does not mean that no durability issues are associated with combinations of WMA and 40% RA on real job sites. The 5 laboratory tests used in this testing program may not be pertinent for their detection. The potential effect on other properties will be assessed further as the Task Group inter-laboratory testing program will be completed with fatigue, cracking and thermal stress restrained specimen tests. Furthermore, the RA heating process is a key parameter for the determination of the recycling capability of the different types of mixing plants. Therefore, particular attention has to be paid to this aspect considering that the laboratory RA heating process is not representative of all the actual plant processes.
International Evaluation of the Performance …
25
3 Conclusion The international inter-laboratory testing program designed by the Task Group 2 of the RILEM Technical Committee 264-RAP on non-cold recycling was intended to identify performance issues associated with the combination of WMA technologies and RA recycling. The first step of the program was dedicated to the RA characterization according to the round-robin test practice. The second step was designed to assess the relative performances of WMA mixes containing 40% RA compared to HMA taken as reference in terms of compactability, water sensitivity, rutting resistance, and stiffness modulus. In the first step, significant differences were observed among the laboratories’ results for aggregates grading curve (white curve). This highlights that careful consideration must be given to RA characterization as it determines the relevance of all the following stages of the recycling process. In the second step, the 6 different types of WMA mixes tested, from base to surface courses, performed equal to or better than the reference HMA for all the properties except for the stiffness. Thus, no specific issue associated with WMA with 40% RA was detected. To ascertain these conclusions the study will be extended to other properties.
References 1. EN 13108-1: Bituminous Mixtures—Material Specifications—Part 1: Asphalt concrete 2. EN 12697-2+A1: Test Methods for Hot Mix Asphalt—Part 2: Determination of Particle Size Distribution 3. EN 12697-31: Test Methods for Hot Mix Asphalt—Part 31: Specimen Preparation by Gyratory Compactor 4. EN 12697-12: Test Methods for Hot Mix Asphalt—Part 12: Determination of the Water Sensitivity 5. EN 12697-22+A1: Test Methods for Hot Mix Asphalt—Part 22: Wheel Tracking 6. ASTM D3497: Standard Test Method for Dynamic Modulus of Asphalt Mixtures 7. EN 12697-26: Test Methods for Hot Mix Asphalt—Part 26: Stiffness 8. Hugener, M., Kawakami, A.: Simulating repeated recycling of hot mix asphalt. Road Mater. Pavement Des. 18(sup2), 76–90 (2017). https://doi.org/10.1080/14680629.2017.1304263
Technical Committee 272-PIM (“Phase and Interphase Behaviour of Bituminous Materials”)
Interlaboratory Test to Characterize the Cyclic Behavior of Bituminous Interlayers: An Overview of Testing Equipment and Protocols Francesco Canestrari, Thomas Attia, Hervé Di Benedetto, Andrea Graziani, Piotr Jaskula, Youngsoo Richard Kim, Maciej Maliszewski, Jorge C. Pais, Christophe Petit, Christiane Raab, Davide Ragni, Dawid Rys, Cesare Sangiorgi, Cédric Sauzéat, and Adam Zofka Abstract The performance assessment of multi-layered pavements strongly depends on the mechanical behavior of the interface between bituminous layers. So far, comprehensive studies have been carried out mainly using quasi-static laboratory tests focusing on the interlayer shear strength at failure. However, it is generally recognized that cyclic shear testing will lead to the determination of parameters which are more closely linked to the performance of pavements under traffic loading than the quasi-static shear tests. This paper outlines the research work that has been carried out within the Task Group 3 “Pavement multilayer system” of the RILEM TC 272-PIM. The activities focused on cyclic shear testing of interfaces in bituminous pavements involve an interlaboratory test with nine participating laboratories. The interface behavior was investigated through both direct shear and torque tests on F. Canestrari (B) · A. Graziani · D. Ragni Polytechnic University of Marche, Ancona, Italy e-mail: [email protected] T. Attia · H. Di Benedetto · C. Sauzéat University of Lyon ENTPE, Vaulx-en-Velin, France P. Jaskula · D. Rys Gdansk University of Technology, Gda´nsk, Poland Y. R. Kim North Carolina State University, Raleigh, USA M. Maliszewski · A. Zofka Road and Bridge Research Institute IBDiM, Warsaw, Poland J. C. Pais University of Minho, Braga, Portugal C. Petit University of Limoges, Limoges, France C. Raab Swiss Federal Laboratories for Materials Science and Technology Empa, Dübendorf, Switzerland C. Sangiorgi University of Bologna, Bologna, Italy © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_4
29
30
F. Canestrari et al.
double-layered specimens extracted from lab compacted slabs prepared by one of the laboratories. The different testing equipment and protocols used by the participating laboratories are presented, highlighting the variety of geometries, loading modes, and testing parameters. Keywords Interlayer bonding · Bituminous interlayer · Cyclic loading · Stiffness · Damage accumulation · Interlaboratory test
1 Introduction Traffic loading on pavement structures induces horizontal stresses that may become critical when poor interlayer bonding conditions exist, leading to premature pavement failure. Static shear testing devices have been largely applied to analyze the interlayer shear properties, mainly expressed in terms of the interlayer shear strength (ISS) [1]. Since traffic loads applied to the pavement are cyclic in nature, cyclic interlayer testing should lead to more realistic predictions of pavement performance. In particular, it would be helpful to understand the interlayer behavior in terms of cyclic shear stiffness for better predicting the bearing capacity of multi-layered pavements. Moreover, the fatigue shear resistance at the interface should be investigated to establish, for a given pavement structure, which is predominant between bottom-up cracking (depending on tensile stress/strain) and slippage failure (strictly linked to the bonding between pavement layers). In order to fill this gap, an interlaboratory test on “Cyclic Interlayer Shear Testing” was promoted by Task Group 3 “Pavement Multilayer System” of RILEM Technical Committee 272-PIM “Phase and Inter-phase behavior of bituminous Materials”. Within the RILEM interlaboratory test, different cyclic shear testing devices currently available were used by the participating laboratories with the aim of identifying appropriate testing protocols and meaningful parameters. Several tests were performed on double-layered specimens extracted from lab compacted slabs prepared by one of the laboratories using a single bituminous mixture [2]. The objective of this paper is to present different testing devices and protocols used by the participating laboratories for performing cyclic interface testing.
2 Testing Equipment and Protocols of Participating Laboratories The Cyclic-Ancona Shear Testing Research and Analysis (ASTRA) device (Fig. 1a) was used at Polytechnic University of Marche (UNIVPM). It consists of two halfboxes separated by an adjustable gap [2]. One half-box moves vertically allowing the application of a vertical load (parallel to the interface). The other half-box can slide
Interlaboratory Test to Characterize the Cyclic …
31
horizontally (perpendicular to the interface) allowing shear dilatancy. The device also allows applying a prefixed normal load (perpendicular to the interface), by means of a pneumatic actuator. Cyclic-ASTRA tests were carried out in haversine stress control mode at 10 and 20 °C with a frequency of 5 Hz, without the application of a normal load. Four stress amplitudes were used, based on the average interlayer shear strength (ISS) of the double-layered specimens measured with the static test. The ratio between the applied amplitudes and the measured ISS were 0.29, 0.37, 0.44 and 0.51. For each cycle, the data acquisition system recorded the shear load (used to calculate the shear stress τ0 ) and the relative interface displacement u 0 . Results show typical three stages evolution of the permanent displacement at the layer interface as a function of the number of cycles until failure. The Advanced Shear Tester (AST) device (Fig. 1b), was used both at Gdansk University of Technology (GUT) and at Road and Bridge Research Institute (IBDiM). It consists primarily of two parts, stationary holder and moving collar, with a variable gap size between them [3]. AST device can apply a normal confining stiffness to the interface by means of four die springs. GUT carried out cyclic AST tests at 20 °C by adopting a gap of 2 mm; the loading cycle was characterized by a minimum (50 kPa) and maximum shear stress (330 and 450 kPa) set as 20% of the Leutner interlayer shear strength (ISS). Each cycle consisted of 0.05 s of loading time, 0.05 s of unloading time and 0.1 s of rest time. Several parameters were obtained as a function of the cycle number such as permanent strain, shear stiffness (permanent and resilient), total accumulated dissipated energy (from hysteresis loops) and equivalent energy ratio. IBDiM carried out Cyclic AST tests with a gap of 5 mm at 13 °C, both without and with initial normal stress (400 kPa). The testing protocol included a frequency sweep (1, 2, 5, 10 Hz) in sinusoidal control stress mode followed by a haversine fatigue test at 10 Hz. The Modified Advanced Shear Tester (MAST) device (Fig. 1c) was used at North Carolina State University (NCSU). It is composed of a fixed side platen and a movable side platen separated by a gap of 8 mm. The movable side platen is free to move vertically (parallel to the interface) as well as horizontally (perpendicular to the interface); MAST device allows also the application of a normal confining stiffness to the specimen by means of a bolt and spring system [4]. Cyclic MAST tests were carried out in haversine displacement control at a temperature of 20 °C, with a frequency of 5 Hz and without applying any normal (confining) stress at the interface. Displacement and shear stress amplitudes, u 0 and τ0 respectively, were collected during the test in order to calculate the interlayer shear stiffness K = τ0 /u 0 . Typical results of constant displacement amplitude cyclic (CDAC) tests consider as output the number of cycles to failure (Nf ) associated with the 50% reduction of interlayer shear stiffness. The Cyclic Compressed Shear Bond (CCSB) device was used by Empa, in collaboration with the University of Dresden. Double-layered cylindrical specimens are glued into half-shells mounted on aluminum supports (Fig. 1d). The shear load is dynamically displacement-controlled using a sinusoidal signal and applied by a vertically movable support, while the horizontally movable support enables the application of normal load. Different shear displacement amplitudes were selected
32
F. Canestrari et al.
at different temperatures (0.03 mm @−10 °C, 0.07 mm @10 °C, 0.11 mm @30 °C and 0.15 mm @50 °C) in order to run the tests within the linear viscoelastic range. For each temperature with corresponding displacement amplitude, 4 normal loads (0.0, 0.3, 0.6 and 0.9 MPa) and 5 frequencies (0.1, 0.3, 1, 3 and 10 Hz) were chosen. The Repeated Impulse Leutner (RIL) device was designed at the University of Bologna and enables the evaluation of shear strengths under cyclic shear loading when subject also to a normal load. The aim of the proposed test is to evaluate the shear strength of the interlayer under cyclic shear loading when subject also to a normal load. The device consists of a traditional Leutner device fixed on a frame equipped with a pneumatic piston able to carry out a constant and normal pressure to the upper face of the double-layered cylindrical specimen (Fig. 1e). RIL device applied cyclic shear load according to a haversine and repeated loading pulse, with the peak value fixed at 250 kPa, a pulse repetition period of 1500 ms and a rise time of 130 ms. The University of Minho developed the Dynamic Interlayer Shear Testing (DIST) device to assess the interlayer bond of asphalt pavements by applying monotonic or cyclic shear loads on double-layered cylindrical specimens. The device is composed of two circular crowns to fix each part of the specimen (Fig. 1f) and, being equipped with two servo-hydraulic actuators, allows applying also normal forces at the interface. The interlayer shear behavior was evaluated at 20 °C, without normal load, in three modes: (1) frequency sweep; (2) displacement sweep; (3) fatigue. Frequency sweep tests were carried out by applying a cyclic shear displacement of 0.1 mm at frequencies from 0.1 to 2 Hz. Displacement sweep tests were carried out at 1 Hz applying cyclic displacements from 0.05 to 0.30 mm. The number of cycles was very low to avoid damaging the specimen. Fatigue tests were carried out at 5 Hz applying a cyclic displacement of 0.05, 0.10 and 0.15 mm. The Interlayer Shear-Torque (IST) device (Fig. 1g) was used to perform fatigue tests at University of Limoges by means of a servo-hydraulic testing equipment [5]. IST fatigue tests were carried out in controlled stress mode at a frequency of 10 Hz and a temperature of 20 °C. A sinusoidal torque and a constant axial compression load N of 0.05 kN were applied to the double-layered cylindrical specimen glued with an epoxy resin to the loading plates. IST fatigue test returns, for each level of torque amplitude, a dataset including the evolution of torque, axial load, torsional rotation angle and phase angle as a function of the number of cycles. The interlayer behavior was characterized through the complex shear modulus G ∗ . The evolution of the norm of the complex shear modulus |G ∗ | as a function of the number of loading cycles was analyzed at the end of each IST fatigue test. Classical 50% modulus reduction criterion was adopted to determine the number of cycles to failure (Nf ) associated with the corresponding shear stress amplitude. The 2T3C Hollow Cylinder Apparatus (2T3C HCA) was developed at the University of Lyon/ENTPE to study the mechanical behavior of pavement interlayers [6] (Fig. 1h). Torsion and tension/compression are applied independently to a hollow cylinder specimen placed inside a thermal chamber (from −20 to 60 °C). The dimensions of the double-layered hollow cylinder specimen were 74 mm total height, 172 mm external diameter and 122 mm internal diameter (thickness of 25 mm). Two
Interlaboratory Test to Characterize the Cyclic …
33
pairs of non-contact sensors measure the vertical displacement and the displacement due to the rotation between the top and the bottom of the specimen. 3D Digital Image Correlation was used to obtain local displacements on the surface of the specimen. The corresponding strain tensor components εzz , εθz in both layers. The displacement gaps at the interface in the axial direction uz and in rotation uθ were determined by means of a specific analysis. Furthermore, a special acquisition system allowed to calculate the associated vertical stresses σzz and shear stresses τθz . Cyclic sinusoidal loadings at small strain amplitudes were applied at four temperatures (10, 20, 30, 40 °C) and, for each temperature, at four frequencies (0.01, 0.03, 0.1, 0.3 Hz). For each couple of temperature and frequency, several mechanical parameters were obtained and, among others, the complex interface shear modulus K θ∗z .
3 Preliminary Results The cyclic testing procedures applied by the participating laboratories focused on stiffness and fatigue resistance (Fig. 2) of the double-layered specimens. Stiffness results showed a time and temperature dependent behavior and allowed the construction of master curves. Fatigue results showed that the interlayer behavior can be analyzed following the same approaches normally adopted for bituminous mixtures (i.e. stiffness reduction with increasing loading cycles). In all cases, the gap imposed by the testing device had a relevant influence on the results.
4 Conclusions Nine laboratories applied their own different cyclic testing equipment and protocols to evaluate interlayer shear stiffness and/or the fatigue resistance of double-layered asphalt specimens. Very different geometries, loading modes, and testing parameters were compared. Although these differences did not allow the direct comparison of the measurements, results clearly showed that the influence of testing frequency and temperature can be analyzed using the time-temperature superposition principle typically adopted for bituminous mixtures. Moreover, some key issues were identified for improving the testing standardization such as the specimen clamping (glued, not glued), the interlayer gap, the testing mode (stress- or strain-controlled) and the applied loading amplitude (either stress or strain).
34
F. Canestrari et al.
a)
b)
c)
d)
e)
f)
g)
h)
Fig. 1 Laboratory testing equipment used by the participating laboratories: a cyclic-ASTRA; b AST; c MAST; d CCSB; e RIL; f DIST; g IST; h 2T3C HCA
Interlaboratory Test to Characterize the Cyclic …
35
Fig. 2 Stiffness master curves (left); stiffness evolution in fatigue tests with different displacement amplitudes u 0 (right)
References 1. Canestrari, F., Ferrotti, G., Lu, X., Millien, A., Partl, M.N., Petit, C., Phelipot-Mardelé, A., Piber, H., Raab, C.: Mechanical testing of interlayer bonding in asphalt pavements. In: Partl M. et al. (eds.) Advances in Interlaboratory Testing and Evaluation of Bituminous Materials. RILEM State-of-the-Art Reports 9, 303–360. Springer, Dordrecht (2013) 2. Ragni, D., Graziani, A., Canestrari, F.: Cyclic interlayer testing in bituminous pavements. In: Proceedings of 7th International Conference on Bituminous Mixtures and Pavements (7ICONFBMP), Thessaloniki, CRC Press (2019) 3. Zofka, A., Maliszewski, M., Bernier, A., Josen, R., Vaitkus, A., Kleizien˙e, R.: Advanced shear tester for evaluation of asphalt concrete under constant normal stiffness conditions. Road Mater. Pavement Des. 16, 187–210 (2015) 4. Cho, S.H., Kim, Y.R.: Verification of time–temperature superposition principle for shear bond failure of interlayers in asphalt pavements. Transp. Res. Rec. J. Transp. Res. Board 2590, 18–27 (2016) 5. Ragni, D., Takarli, M., Petit, C., Graziani, A., Canestrari, F.: Use of acoustic techniques to analyse interlayer shear-torque fatigue test in asphalt mixtures. Int. J. Fatigue 131, 105356 (2019)
36
F. Canestrari et al.
6. Attia, T., Di Benedetto, H., Sauzéat, C., Pouget, S.: Experimental investigation of the mechanical behaviour of interfaces between pavement layers. In: Proceedings of GeoShanghai 2018 International Conference: Transportation Geotechnics and Pavement Engineering, 344–352. Singapore: Springer Singapore (2018)
Complex Bituminous Binders, Are Current Test Methods Suitable for? Laurent Porot, Emmanuel Chailleux, Panos Apostolidis, Jiqing Zhu, Alexandros Margaritis, and Lucia Tsantilis
Abstract The asphalt industry is constantly working to enhance the performances of asphalt materials, introducing innovative and more sustainable solutions. In this context, the incorporation of materials, such as additives, polymers, is more and more used to improve the properties of neat bitumen. This leads to even more complex bituminous binders, raising the question, are the current specifications and test methods appropriate for complex materials? To deal with this, the RILEM Technical Committee 272-PIM ‘Phase and Interphase behaviour of innovative bituminous Materials’ with its Task Group TG1 is looking at the efficiency of various test methods for complex binders with an extensive inter-laboratory program with 17 laboratories. It includes seven different binders, two neat bitumen, two polymer modified bitumen and three binders with liquid additives, emphasising on compositional and physical changes at different conditions. The focus is low temperature; while a complementary experimental program encompasses as well as testing at intermediate and high temperatures. The outcomes of the work will provide indications on how robust the current binder characterisation techniques are and establish technical recommendations for future test methods specially designed for complex binders. Some first results are presented hereby.
L. Porot (B) Kraton Polymers B.V., Almere, The Netherlands e-mail: [email protected] E. Chailleux Université Gustave Eiffel, Champs-sur-Marne, France P. Apostolidis TU Delft, Delft, The Netherlands J. Zhu VTI, Linköping, Sweden A. Margaritis University of Antwerp, Antwerp, Belgium L. Tsantilis Politecnico Di Torino, Turin, Italy © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_5
37
38
L. Porot et al.
Keywords Bitumen · Polymer modified Bitumen · Asphalt additive · BBR · Rheology · Chemistry · Test protocol
1 Introduction The worldwide demand for long-lasting pavement is leading the asphalt industry to incorporate various chemical substances, such as polymers or additives, in order to enhance the sustainability of pavement materials. Thus, performance-based binders need, together with the implementation of new modification technologies in bitumen, accurate standardised characterisation protocols and in-depth understanding of phase and inter-phase phenomena of these binders. Within this framework, RILEM initiated the Technical Committee 272-PIM ‘Phase and Interphase behaviour of innovative bituminous Materials’ on assessing the performance of complex binders. It aims to provide recommendations for the experimental tools and protocols used by the asphalt research and engineering community on bituminous binders. For this purpose, the Task Group 1 (TG1) of RILEM 272-PIM is specifically focusing on binders, while TG2 is looking at asphalt mix and TG3 on pavement structure. Herein, some preliminary results of TG1 activities are presented. The current characterisation protocols have been initially developed for conventional binders and applied for modified bitumen. Penetration value and softening point temperature, amongst others, are empirical properties used in specifications [1]; while, since the initiation of the Strategic Highway Research Program (SHRP) Superpave program, more fundamental properties using Dynamic Shear Rheometer (DSR) were introduced [2]. There was already some fundamental work, within the RILEM, on characterising the physico-chemical properties of bituminous binders [3, 4]. Furthermore, the activities of TG1 are extending to complex binders. In this research, focus was given on elucidating the low temperature performance of fresh and aged complex binders formulated with polymers and liquid additives by using amongst others, Fraass breaking point, Bending Beam Rheometer (BBR), 4mm plate DSR, Differential Scanning Calorimetry (DSC). A complementary testing program focuses on intermediate and high temperature domains as well. A total of 17 laboratories expressed interest in this experiment, 13 from Europe and 4 from the US. The experiment started earlier in 2019 and about 50% of the results have been delivered by the end of 2019. Examples of preliminary analyses are presented in this paper.
2 Experimental Plan The way to consider complex binders is wide and includes various materials, liquid or viscous additives either bio-based or hydrocarbon-based, polymers, solid particles.
Complex Bituminous Binders …
39
The amount used can vary between dopes from 0.1 to 1.5%, or additives from 2 to 7–10%, or extenders from 20 to 40%, or replacement above 50%, all per weight of bitumen. In order to restrain the scope, only two groups of complex binders were considered, polymer modified binders and additive blends, and compared with neat bitumen. Group 1 is for assessing Polymer modified Bitumen with: • Bit1, a standard 35/50 pen grade bitumen used as a reference • PmB1, a standard commercial Polymer modified Bitumen • PmB2, a highly polymer modification with 7.5% of high vinyl SBS in Bit2 Group 2 is to assess liquid additives in bitumen with: • • • •
Bit2, a standard 70/100 pen grade bitumen used as a reference Blend1, 4% of commercial bio-based asphalt recycling additive in Bit1 Blend2, 8% of Refined Engine Oil Bottom (REOB) in Bit1 Blend3, 4% of paraffinic oil, typically not used as paving additive, in Bit1.
For Group 1, the reference bitumen and PmBs were selected having the same consistency at intermediate temperature; PmB1 was industry produced and composition is proprietary information; PmB2 was lab produced. For Group 2, the dosage per blend was determined with the aim at resulting in properties as close as possible to the properties of the reference bitumen, Bit2; all blends being lab produced. As the goal of the TG was not to run a round robin on testing, the experimental program was agreed and defined as a toolbox, where each laboratory could select the most appropriate way to assess the binders and depending on the available equipment of each laboratory. Table 1 shows the experimental matrix with, for each binder and test, the number of laboratories, and, in brackets, the numbers of results provided by end of 2019.
3 Results As shown in Table 1, the experiment is still ongoing, with 50% of data available, therefore, only partial analysis is presented here. As a first attempt, the different binders, from Group 1 or Group 2, were analysed towards current specifications. Figure 1 displays them in the pen grade system, as used in Europe (EN 12591, EN 14023), with penetration value at 25 °C and softening point temperature. Figure 2 displays them in the Performance Grade (PG) system, as used in the US (AASHTO M320) with the exact low, medium and high temperature criteria. For each figure, the points are for the mean value and the error bars representing the maximum and minimum values. With pen grade system, depending on the binder, the analysed data set was from one to maximum three laboratories. The Group 1 binders, Bit1, PmB1 and PmB2 were displaying similar penetration value at 25 °C, while the softening point temperatures were much more discriminant. For the Group 2, Bit2, Blend 1–3 were all in the 70/100
40
L. Porot et al.
Table 1 Rilem PIM TG1—experimental matrix Sample
Bit 1
PmB 1
PmB 2
Bit 2
Blend 1
Blend 2
Blend 3
RTFOT
7 (5)
8 (5)
9 (6)
9 (8)
7 (6)
8 (7)
7 (6)
PAV
7 (5)
8 (5)
9 (6)
9 (8)
7 (6)
8 (7)
7 (6)
Fraass
3 (2)
4 (2)
5 (3)
3 (3)
2 (2)
2 (2)
2 (2)
BBR
7 (2)
8 (2)
9 (3)
9 (5)
8 (4)
8 (4)
8 (4)
ABC
1 (0)
1 (0)
1 (0)
1 (0)
3 notch beam
2 (0)
2 (0)
2 (0)
2 (0)
2 (0)
2 (0)
2 (0)
4mm DSR
6 (3)
7 (3)
7 (3)
6 (3)
5 (3)
6 (3)
5 (3)
DSC
5 (2)
5 (2)
5 (2)
5 (2)
4 (1)
5 (2)
4 (1)
Aging conditioning
Low temperature testing
1 (0)
Intermediate and high temperature testing Penetration
8 (4)
9 (4)
10 (5)
9 (6)
8 (5)
8 (5)
8 (5)
Softening Point
7 (3)
8 (3)
9 (4)
8 (5)
7 (4)
7 (4)
7 (4)
Force ductility
2 (1)
2 (1)
3 (2)
Elastic recovery
2 (1)
2 (1)
3 (1)
DSR
12 (5)
13 (5)
14 (6)
14 (7)
12 (6)
13 (6)
12 (6)
MSCR
7 (0)
7 (0)
8 (1)
10 (3)
8 (2)
9 (2)
8 (2)
2 (1)
3 (2)
4 (2)
3 (1)
2 (1)
2 (1)
2 (1)
Other testing Storage stability FTIR
8 (6)
8 (6)
9 (7)
10 (9)
8 (7)
9 (8)
8 (7)
Microscopy
3 (0)
3 (0)
4 (0)
4 (0)
3 (0)
3 (0)
3 (0)
pen box and, considering the error bars, there was no significant difference among them. For all binders, the variability between labs was relatively low except for the PmBs with high value of softening point temperature showing variation between 2 and 3 °C. However, the number of data points is still limited. Depending on the binder, the PG grading data set analysed was from one (Bit1, PmB1) to maximum four laboratories (Bit2). With Group 1, as compared to the neat bitumen, the temperature span for PmB was extended, especially at high temperatures. The PmB2, using Bit2 as base bitumen, maintained the low and intermediate temperatures and increased significantly the high temperature. For Group 2, all blends used the Bit1 as base bitumen and displayed similar shift to the left either for low, intermediate and high temperatures. Overall, the properties were not significantly different from the reference bitumen Bit2. The variability between labs, as shown with error bars, was relatively low except for PmB2. However, more data is needed here. Between the pen and PG grade system, the binders from the different groups disclosed the same grading. However, considering penetration value at 25 °C and PG intermediate critical temperature, addressing similar temperature domain, there were
Complex Bituminous Binders …
41
Fig. 1 The seven binders in the pen grading system
Fig. 2 The seven binders in the PG grading system
interesting differences. The pen values for Group 1 binders were similar and lower than Group 2, while for the PG intermediate T, PmB2 value was lower by almost 5 °C than PmB1 and Bit1 and in the same magnitude of range as the Group 2 binders. The PG grading system helps to better distinguish between complex bituminous binders. Further analysis is ongoing including ductility, DSR data analysis with broader parameters and low temperature properties. Herein, a preliminary analysis of low temperature testing results is provided as an example. Figure 3 displays, in a bar chart, the Fraass breaking point temperature with the BBR temperatures when the creep stiffness S ≤ 300 MPa and the relaxation m-value ≥ 0.300, including the delta between them, known as Tc. Depending on the binder, the sets of data were from between two and four laboratories. As expected, the trend was similar at least, the lower the Fraass temperature, the lower the BBR critical temperatures were, although for the Blend1, the Fraass seems to underestimate the low temperature performance as indicated by BBR. A comparison by pair is made in Fig. 4 between (a) the two BBR temperatures and (b) the Fraass breaking point and the BBR continuous (as the max of TS and Tm-value ) temperatures; the diagonal line is for equi-temperature. For the BBR, considering
42
L. Porot et al.
Fig. 3 Low temperature Fraass and BBR for the seven binders
Fig. 4 Comparison of a BBR S and m-value critical temperatures, b Fraass and critical BBR
variability within the labs, there is a good balance between both criteria except for PmB1. At comparing the Fraass and BBR temperatures, for the Group 2, the Fraass is significantly higher than the BBR and all ranking in the same magnitude of range. The BBR continuous temperature seems to better discriminate among them, within the variability of the set of data. With more data including 4 mm plate DSR and DSC, further analysis will be presented in a future study.
4 Conclusion The RILEM 272-PIM has set up a specific Task Group, TG1, looking at phase and interphase of complex bitumen. Two groups of binders, Polymer modified Bitumen and liquid additive blends, are included in an exhaustive experimental program encompassing physical and chemical properties. Seventeen laboratories are involved, providing a unique data set and characterisation approach on complex binders. The activities are still ongoing, and some preliminary results are presented in this paper, as example.
Complex Bituminous Binders …
43
In terms of standard specifications, PmB binders clearly display very different properties, and the additive group shows all binders within a similar grading, in either pen grade or PG grade systems. Thanks to the PG system and by recording properties at low, intermediate and high temperature, some binders were already distinguished more efficiently at intermediate temperatures as compared with pen grade system. Low temperature properties were further analysed based on empirical testing with Fraass breaking point and more fundamental rheology characterisation with BBR testing. While Fraass can distinguish between the two groups, within each group such as between additive blends, it does not enable to differentiate binders. By combining both creep stiffness and m-value critical temperatures, BBR provides a more efficient diagnosis of different complex bituminous binders. Acknowledgements This work has been conducted under the framework of the RILEM Technical Committee TC 272-PIM on “Phase and Interphase behaviour of innovative bituminous Materials” with the great contribution of Kraton (NL), Western Research Institute (US), University of New Hampshire (US), IFSTTAR (FR), TU Delft (NL), Institute of Transportation TU Wien (AU), FHWA (US), CEREMA mediterranee (FR), Colas CST (FR), Strukton (NL), Politecnico di Torino (IT), Belgium Road Research Centre (BE), Swedish National Road and Transport Research Institute (SE), Nynas and University of Antwerp (BE), Nebraska DoT (US), Aalto University (FI), Institut für Straßenwesen, Technische Universität Braunschweig (DE).
References 1. BitVal: Analysis of Available Data for Validation of Bitumen Tests. Report on Phase 1 of the BitVal Project (2006) 2. Anderson, D., Christensen, D., Bahia, H., Dongre, R., Sharma, M., Antle, C., Button, J.: Binder Characterization and Evaluation, Volume 3: Physical Characterization. Washington (1994) 3. Soenen, H., Besamusca, J., Fischer, H., Poulikakos, L., Planche, J., Das, P., Kringos, N., Grenfell, J., Lu, X., Chailleux, E.: Laboratory investigation of bitumen based on round robin DSC and AFM tests. Mater. Struct. 47(7), 1205–1220 (2013) 4. Poulikakos, L., Hofko, B., Cannone Falchetto, A., Porot, L., Ferrotti, G., Grenfell, J.: Recommendations of RILEM TC 252-CMB: relationship between laboratory short-term aging and performance of asphalt binder. Mater. Struct. 52(69) (2019)
Technical Committee 278-CHA (“Crack-Healing of Asphalt Pavement Materials”)
Terms and Definitions on Crack-Healing and Restoration of Mechanical Properties in Bituminous Materials Greet Leegwater, Amir Tabokovi´c, Orazio Baglieri, Ferhat Hammoum, and Hassan Baaj
Abstract Fatigue damage is one of the main modes of deterioration of bituminous materials that leads to the failure of flexible pavements. However, it is well known that bituminous materials have some capability of recovering inflicted damage to some extent due to their viscoelastic nature. This process has been referred to as “selfhealing”. Recently, research on crack-healing of bituminous materials has gained more importance and popularity due to the development of self-healing materials by materials scientists and engineers, that have been used in several applications. The Rilem Technical Committee TC 278-CHA (Crack-Healing of Asphalt Pavement Materials) was created in 2016 to tackle the challenges related to testing and evaluating the crack-healing properties of bituminous materials. However, the committee members noticed that there is no clear agreement within the scientific community about some basic terms to use when it comes to this new area of interest. Hence, this paper aims to clarify some of the concepts related to crack-healing and propose some common terms to the researchers and pavement engineers. Keywords Crack-healing · Restoration · Recovery · Bituminous materials · Self-healing materials · Terminology
G. Leegwater · A. Tabokovi´c Delft University of Technology, Delft, The Netherlands G. Leegwater TNO, Delft, The Netherlands O. Baglieri Politecnico di Torino, Turin, Italy F. Hammoum IFSTTAR, Nantes, France H. Baaj (B) University of Waterloo, Waterloo, ON, Canada e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_6
47
48
G. Leegwater et al.
1 Introduction During (cyclic) mechanical loading a reduction of strength and stiffness can be observed in bituminous materials [1]. Interestingly, in many cases, this loss in strength and/or stiffness of bituminous materials can be totally or partially restored [2]. When first discovered, the restoration of mechanical properties in asphalt was called “Healing” or “Self-healing”. As this effect is significantly large, several pavement design methods address this effect by using an empirical healing-factor [3, 4]. Even though the restoration of mechanical properties is regularly observed, there is no consensus or clear agreement regarding the mechanism that drives this restoration. The reason why it is challenging to understand the restoration, it that there are many mechanisms occurring at the same time that are hard to untangle. Therefore, in order to facilitate the scientific discussions and improve the understanding within the scientific community, we should start by having a clear and concise terminology. In this paper, an introduction to the different phenomena is given and a framework of definitions is proposed. This set of terms and definitions for restoration and crack-healing in bituminous materials will also assist interaction with other research domains. In the field of materials science, in the beginning of this century, a research movement started aiming at designing materials that are able to restore damage autonomously, creating the domain of “self-healing materials”. Since then, the research on self-healing materials has moved into the field of bituminous materials and several researchers have started investigating different types of solutions to improve self-healing capabilities of asphalt mixes and binders, Fig. 1. As a result, terminology of the self-healing world has entered to bituminous materials field. Creating one set of definitions will make pavement engineering publications and discussions more comprehendible for people outside the pavement engineering community.
Fig. 1 Different approaches to introduce additional self-healing performance (extrinsic) to bituminous materials
Terms and Definitions on Crack-Healing …
49
2 Rheological Response of Bituminous Materials to Cyclic Loading Especially in cyclic loading (typically imposed in so-called fatigue tests), there are several mechanisms that come into play with respect to the loss and consequently the restoration of mechanical performance [5–8]. Therefore, the purpose of this section is to provide a small introduction to the response of bituminous materials subjected to cycling loadings. This response depends on several mechanisms related to the time- and temperature-dependant nature of bituminous materials. When interpreting the response, it is important to discriminate loss in mechanical properties due to formation of micro-cracks from loss due to other undesired biasing effects, more specifically local heating, thixotropy and non-linearity. There is still debate about the relative impact of each these phenomena, however the main phenomena that may occur during cyclic loading tests are briefly introduced below. Non-linear viscoelasticity (NLVE)—Bituminous materials start exhibiting a strain dependent visco-elastic behaviour, commonly referred to as NLVE, when applied strain exceeds a characteristic threshold value (LVE limit). Determination of LVE limit for bituminous materials is generally not trivial. Moreover, it must be taken into account that binder located within inter-aggregate spaces in mixtures is subjected to strain levels significantly higher than the overall mix strain, resulting in non-linearity when an asphalt mix is subjected to relatively small strain levels [9]. As a consequence of non-linearity, proportionality between stresses and strains is no longer valid resulting in a difference in observed stiffness, however this does not imply any damage and is therefore purely biasing effects in the test. Self- or Local Heating—Application of repeated loading to bituminous materials may cause an increase of internal temperature of tested sample due to energy dissipated by visco-elastic effects. As a consequence of this higher temperature, the material shows a stiffness reduction. This phenomenon occurs mainly in the initial stage of loading at a relatively high rate [5]. Self-heating has been proven by several authors measuring temperature variations on bituminous binders [10, 11] and mixtures [12]. Thixotropy—Thixotropy is associated to shear-induced reorganizions of the material’s microstructure resulting in a viscosity decrease [13]. When left to rest, due to Brownian motion, the elements of the microstructure are able to slowly regain their more favourable energy positions and the microstructure is rebuilt recovering the original viscosity. The whole process is completely reversible. Evidences of thixotropy occurrence in bituminous materials have been shown [14] and methods to separate thixotropy from fatigue process have been proposed [8, 15]. Micro-cracking—Fatigue cracking is traditionally described as a two-stage process, including the (I) initiation phase, in which micro-cracks appear in a diffuse way to form a network and the (II) propagation phase, in which micro-cracks coalesce and grow until macro-cracking occurs [1]. Beginning of cracks has been related to micro-compositional flaws in the binder medium [16] that grow slightly in each load cycle. Cracks can grow as both adhesive fractures (between the binder and
50
G. Leegwater et al.
the aggregates) and cohesive fractures (within the thin film of binders). In this document, deleterious structural changes related to micro-crack initiation and propagation are referred to as damage. This definition of damage implies the use of continuum damage mechanics.
3 Continuum Damage Mechanics Versus Fracture Mechanics In the analyses of cracking damage, different approaches can be taken. An important facet of the damage analyses is whether principles from fracture mechanics are used or continuum damage mechanics [17]. As seen in the previous section where microcracks are assumed to be smeared over a wide region, which more typical in fatigue cracking, the use of continuum damage mechanics is more appropriate. However, fracture mechanics is well suited for the analysis of structures and materials that contain a single crack, which is the case when micro cracks coalesce to one macro crack in fatigue or in the case of thermal cracking. The perspective taken when analysing the damage, should also be considered when evaluating healing, as this is by definition revering he consequences of the damage.
4 Restoration and Crack-Healing During Rest As mentioned before, the restoration of the mechanical properties can be readily observed in bituminous materials if the mechanical loading is interrupted and the specimen is left to rest. When the cyclic loading is stopped during a fatigue test, the effect of local heating and thixotropy immediately starts to diminish. This part of the restoration is not interpreted as a change in material behaviour or performance, but is regarded as the return to the original state of the material. This part of the restoration is therefore by definition limited to the reversible experimental biasing effects and “recovery” is seen as a suitable term for this part of the restoration. Next to this, the closing and repair of cracks due to the viscous nature of bitumen is also expected to contribute to the restoration [18–21]. We propose to call this part of the restoration “self-healing”. The contribution of recovery and self-healing to the restoration is schematically shown in Fig. 2. The relative contribution of each of these phenomena to the restoration is still an active research topic and is shown to be test dependent [8]. Based on current insights, it is expected that the restoration due to self-healing is slower compared to restoration due to recovery. Next to this, the scientific community assesses that it is likely that the self-healing of cracks affects performance in the
Terms and Definitions on Crack-Healing …
51
Fig. 2 Schematic overview of contributions to the restoration of performance of different phenomena
field, while the biasing effects in fatigue tests are very much related to the unrealistic loading, which is needed for accelerated laboratory testing.
5 Conclusion: Proposed Definitions The main conclusions of this paper are the definitions on crack-healing that aim to help the community to have a clear discussion on restoration of mechanical properties and crack-healing. These definitions are formulated for the use in this specific domain, as for instance other domains e.g. aging of bituminous materials will have their own definition of damage. By separating restoration in recovery and self-healing these definitions are also applicable to healing of cracks that are not the result of fatigue loading, but caused by e.g. temperature cracking. Damage: Loss of original mechanical properties, i.e. the strength or the stiffness, due to the initiation, coalescence and propagation of micro-cracks within the material. Restoration: The total observed change in mechanical properties after a period of rest. Recovery: Component of the restoration that can be attributed to changes in response resulting from cyclic loading, more specifically heating and thixotropy. Self-healing: Component of the restoration that can be attributed to the closure and repair of (micro) cracks. Intrinsic (Self-)healing: Aspect of the self-healing behaviour that is inherent to the material used.
52
G. Leegwater et al.
Extrinsic (Self-)healing: Aspect of the self-healing behaviour that can be attributed to an added phase or action that is specifically used to improve self-healing capabilities, to trigger crack-healing or prevent crack propagation. Acknowledgments The authors would like to acknowledge the members of the Rilem TC 278CHA (Crack Healing of Asphalt) and the members of Cluster F (Bituminous Materials and Polymers) for their comments, suggestions and contributions to the very constructive discussions during the different TC and Cluster meetings.
References 1. Di Benedetto, H., De La Roche, C., Baaj, H., Pronk, A., Lundström, R.: Fatigue of bituminous mixtures. Mater. Struct. 37(3), 202–216 (2004) 2. Baaj, H., et al.: Recovery of asphalt mixture stiffness during fatigue loading rest periods. Constr. Build. Mater. 158, 591–600 (2018) 3. Bazin, P., Saunier, J.: Deformability, fatigue and healing properties of asphalt mixes. Int. Conf. Struct. Des. Asphalt Pvmts (1967) 4. Bonnaure, F., Huibers, A., Boonders, A.: A laboratory investigation of the influence of rest periods on the fatigue characteristics of bituminous mixes. J. Assoc. Asphalt Paving Technol. 51, 104–128 (1982) 5. Di Benedetto, H., Nguyen, Q.T., Sauzéat, C.: Nonlinearity, heating, fatigue and thixotropy during cyclic loading of asphalt mixtures. Road Mater. Pavement Des. 12(1), 129–158 (2011) 6. Di Benedetto, H., Ashayer Soltani, A., Chaverot, P.: Fatigue damage for bituminous mixtures: a pertinent approach. J. Assoc. Asphalt Paving Technol. 65 (1996) 7. Baaj, H., Di Benedetto, H., Chaverot, P.: Effect of binder characteristics on fatigue of asphalt pavement using an intrinsic damage approach. Road Mater. Pavement Des. 6(2), 147–174 (2005) 8. Mangiafico, S., et al.: Quantification of biasing effects during fatigue tests on asphalt mixes: non-linearity, self-heating and thixotropy. Road Mater. Pavement Des. 16(sup2), 73–99 (2015) 9. Kose, S., et al.: Distribution of strains within hot-mix asphalt binders: applying imaging and finite-element techniques. Transp. Res. Rec. 1728(1), 21–27 (2000) 10. Bodin, D., Soenen, H., de La Roche, C.: Temperature effects in binder fatigue and healing tests. In: Eurasphalt & Eurobitume Congress, Vienna (2004) 11. Riahi, E., et al.: Quantification of self-heating and its effects under cyclic tests on a bituminous binder. Int. J. Fatigue 104, 334–341 (2017) 12. Lundström, R., Ekblad, J., Isacsson, U.: Influence of hysteretic heating on asphalt fatigue characterization. J. Test. Eval. 32(6), 484–493 (2004) 13. Mewis, J., Wagner, N.J.: Thixotropy. Adv. Coll. Interface. Sci. 147, 214–227 (2009) 14. Mouillet, V., et al.: Thixotropic behavior of paving-grade bitumens under dynamic shear. J. Mater. Civ. Eng. 24(1), 23–31 (2011) 15. Shan, L., et al.: Separation of thixotropy from fatigue process of asphalt binder. Transp. Res. Rec. 2207(1), 89–98 (2011) 16. Lytton, R. et al.: The fatigue cracking of asphalt mixtures in tension and compression. Adv. Asphalt Mater. 243–272 (2015) 17. Tabakovi´c, A.: The Influence of Reclaimed Asphalt Pavement (RAP) on Binder Course Mix Performance. University College Dublin (2007) 18. Kim, Y.R., Little, D.N., Benson, F.C.: Chemical and mechanical evaluation on healing mechanism of asphalt concrete. J. Assoc. Asphalt Paving Technol. 59 (1990)
Terms and Definitions on Crack-Healing …
53
19. Phillips, M.: Multi-step models for fatigue and healing, and binder properties involved in healing. In: Eurobitume Workshop on Performance Related Properties for Bituminous Binders, Luxembourg (1998) 20. Bhasin, A., et al.: A framework to quantify the effect of healing in bituminous materials using material properties. Road Mater. Pavement Des. 9(sup1), 219–242 (2008) 21. Leegwater, G., Scarpas, A., Erkens, S.: The influence of boundary conditions on the healing of bitumen. Road Mater. Pavement Des. 19(3), 571–580 (2018)
Testing Methods to Assess Healing Potential of Bituminous Binders Orazio Baglieri, Hassan Baaj, Francesco Canestrari, Chao Wang, Ferhat Hammoum, Lucia Tsantilis, and Fabrizio Cardone
Abstract Despite the fact that ability of bituminous materials to heal damage has been largely studied and that a wide literature is available on this topic, there is still no consensus in the research community on actual factors governing healing process and methods to be used to quantitatively asses healing performance of bituminous materials. In such a context, one of the main objectives of the RILEM Technical Committee 278-CHA is to evaluate the different test methods proposed by various research teams to characterize healing in bituminous materials and come up with clear and concise recommendations regarding the evaluation of healing behavior of bituminous materials in the laboratory. To this regard, an inter-laboratory experimental program has been recently started with the specific purpose of testing a common set of selected bituminous binders by using three different protocols. This paper describes the protocols adopted in the experimental program and reports some preliminary results obtained to date. Keywords Bituminous binders · Healing · Damage · Dynamic shear rheometer
O. Baglieri (B) · L. Tsantilis Politecnico di Torino, Turin, Italy e-mail: [email protected] H. Baaj · F. Cardone University of Waterloo, Waterloo, ON, Canada F. Canestrari Università Politecnica delle Marche, Ancona, Italy C. Wang Beijing University of Technology, Beijing, China F. Hammoum Université Gustave Eiffel, Champs-sur-Marne, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_7
55
56
O. Baglieri et al.
1 Introduction Bituminous binders are recognized to have the interesting ability to restore their strength or stiffness lost due to external loading when allowed to rest for a given time interval [1, 2]. This ability is generally referred to as healing. Despite such a phenomenon has been widely studied for long time and a number of testing methods have been proposed, there is no consensus among the scientific community on a possible standardized protocol to be used for the evaluation of binders healing potential. In fact, the development of a reliable testing procedure is not a trivial task due to many effects coming into play in laboratory measurements, including reversible materials effects (non-linearity, local heating, thixotropy) not related to micro-damage [3, 4]. However, the research community would benefit from the adoption of a single recommended methodology for material’s evaluation and classification purposes. This is the main challenge of the Task Group TG2 of the Rilem Technical Committee TC 278-CHA (Crack-Healing of Asphalt Pavement Materials), that launched an inter-laboratory testing program in which a common array of bituminous binders has been evaluated by means of different protocols with the purpose of comparing the aptitude of such protocols in discriminating and ranking material healing response. This paper describes the protocols selected for the interlaboratory experimentation highlighting the different rationales underlying various testing approaches. Moreover, some preliminary results are presented and discussed.
2 Materials A set of three bituminous binders of different type and provenience was used in the inter-laboratory experimentation, including two neat bitumen and one SBS polymer modified binder (PMB). Description of investigated materials is given in Table 1.
3 Testing Methods The three selected protocols used by different laboratories involved in the activities of TG2 to evaluate healing potential of the bituminous binders are based respectively on: Table 1 Bituminous binders used in the experimental investigation
Binder
Type
Provenience
Pen grade
B1
Neat
Italy
70/100
B2
Neat
Ireland
40/60
B3
SBS PMB
France
30/45
Testing Methods to Assess Healing Potential …
(1) (2) (3)
57
Time sweep tests with single rest period (TS-SRP) Time sweep tests with multiple rest periods (TS-MRP) Linear amplitude sweep test with single rest period (LASH).
For all protocols, measurements were carried out by means of DSR equipped with 8 mm parallel plates and 1 mm gap. For comparison purposes, a common temperature of 20 °C and frequency of 10 Hz were adopted by all laboratories. Three replicates were performed for each type of test and average results were considered in the subsequent analysis.
3.1 TS-SRP TS-SRP test consists of two oscillatory shear loading phases separated by the interposition of a single long rest period (2 h), with the purpose of evaluating the kinetic of recovery process over time [5]. The first loading phase, preceded by a conditioning phase of 30 min, is interrupted at specific levels of damage reached by the material, corresponding to the reduction of the norm of the complex modulus |G*| equal to 15, 35 and 50%. Oscillatory loading is applied in the strain-controlled mode at a strain amplitude belonging to the non-linear viscoelastic domain of the material. Based on preliminary strain amplitude sweep tests performed on all selected binders, a strain amplitude value of 3% was chosen. During both the conditioning phase and the rest period, a low strain amplitude (≤0.1%) was also applied to monitor the evolution of G* as a function of time. With the aim of evaluating and subtracting reversable material effects occurring during rest period and not related to micro-damage repair, binders were also subjected to TS-SRP tests in which the first loading phase is substituted with a dummy phase, characterized by a few loading cycles during which a negligible damage is induced in specimens. A schematic representation of TS-SRP tests with effective and dummy first loading phase is depicted in Fig. 1.
Fig. 1 Schematic representation of TS-SRP test with effective (a) and dummy (b) first loading phase
58
O. Baglieri et al.
The self-healing response of binders was assessed by introducing a healing index HI TS-SRP , calculated as indicated in Eq. 1: H I T S−S R P =
|G ∗ | R−F R F − |G ∗ | R−R F · 100 |G ∗ | F−F R F
(1)
where |G ∗ | R−F R F is the difference between the final value |G ∗ | R, f −F R F and the initial value |G ∗ | R,i−F R F of the norm of the complex modulus recorded in the rest period of effective TS-SPR tests, |G ∗ | R−R F is difference between the final value |G ∗ | R, f −R F and the initial value |G ∗ | R,i−R F of the norm of the complex modulus recorded in the rest period of dummy TS-SRP tests, and |G ∗ | F−F R F is the modulus loss imposed during the first loading phase of effective TS-SRP tests, given by the difference between the value of the norm of the complex modulus at the end of the conditioning phase |G ∗ |C, f −F R F and the value of the norm of the complex modulus at the beginning of rest period |G ∗ | R,i−F R F .
3.2 TS-MRP TS-MRP tests consist of time sweeps interrupted by multiple rest periods, aimed at mimicking real loading conditions occurring in the field [6, 7]. The evolution of the loss modulus (|G ∗ | · sinδ) is continuously monitored during the entire test as representative of damage accumulation. Each loading phase is interrupted when the |G ∗ | · sinδ drops down to 65% of its initial value (selected damage level) and is resumed after 30 min. A strain amplitude belonging to the non-linear viscoelastic domain of the material was applied. During each rest-period a low strain amplitude (≤ 0.1%) is applied in order to monitor the recovery of loss modulus. A minimum number of 12 load-rest phases are recommended in all test conditions. The test data analysis provides the identification of numerical parameters to quantify the overall fatigue endurance limit of material as follows: N Fat = N0 + N f H ; N f H = lim N H (n); N H (n) = n→∞
Ni = N H (i) + N∞
n
N H (i);
i=1
(2)
where N Fat is the effective number of loading cycles that the material withstands before failure considering the “n” rest periods, N 0 is the number of loading cycles in the first loading phase, N H is the overall self-healing contribution after “n” rest periods, N i is the number of loading cycles after each “i-th” rest period which are necessary to reach the selected level of damage, N H (i) is the self-healing contribution (variable depending on the “i-th” rest-period considered) and N ∞ is the thixotropy contribution (constant for each rest/load phase). It is observed that
Testing Methods to Assess Healing Potential …
59
N H (i) decreases with the number of rest periods applied and vanishes after a certain number of rest periods. Thus, after a high number of rest periods (virtually infinite) N H , defined as the self-healing potential, converges to a finite asymptotic value NfH that represents the cumulative self-healing capability for infinite number of rest periods (Fig. 2). A healing index HITS-MRP is then introduced, calculated with Eq. 3: H I T S−M R P =
NfH N Fat − N0 · 100 = · 100 N0 N0
(3)
3.3 LASH LASH test consists in a Linear Amplitude Sweep (AASHTO TP101–12) with the inclusion of a rest period at a fixed level of damage [8]. LASH test data can be interpreted with the Simplified-Viscoelastic Continuum Damage (S-VECD) model approach to identify the damage characteristic curve, which is described in terms of pseudo stiffness (C) versus damage intensity (S). The details of the S-VECD model application on bituminous binder are available elsewhere [9]. The key concept of LASH procedure is to take the S f as a critical damage parameter, which represents the damage intensity of S up to failure occurrence. The analyzed damage states are defined at 25%Sf , 50%Sf , 75%Sf , and 125%Sf and include values above and below the critical damage parameter. For each Sf -based damage level, multiple rest periods of 60, 300, 900, and 1800 s are introduced to characterize the healing behavior. An example of results obtained from LASH test is given in Fig. 3, which shows typical damage characteristic curves of the material before and after rest period. A healing index HILASH is defined based on the damage recovery as expressed in Eq. 4, in which the S 1 and S 2 represent the measured S values immediately preceding and after the rest period, respectively: H I L AS H =
S1 − S2 · 100 S1
(4)
4 Preliminary Results and Concluding Remarks With the only purpose of providing a preliminary comparison of the healing ranking obtained adopting different experimental approaches (TS-SRP, TS-MRP, and LASH protocols), a summary of healing indicators is displayed in Fig. 4 for the three investigated materials. In the case of TS-SRP tests the healing indicator is represented by HITS-SRP (Eq. 1), calculated at a level of damage corresponding to a reduction of
N0
65%(G*·sinδ)0
3
ΔN1 ΔN2
2
N. of cycles
1
ΔN∞
NH(n)
NH(i) Model
Experimental data
N. of rest-period
Fig. 2 Loading-recovery sequence used in TS-MRP tests (a). Modelling of self-healing parameter N H (n) ()
G*·sinδ
(G*·sinδ)0
NfH
60 O. Baglieri et al.
Testing Methods to Assess Healing Potential …
61
1.2
Fig. 3 Damage characteristic curve before and after rest period from LASH test
LAS Test
Rest Period with Healing
1.0
C
LASH Test
0.8
0.6
0.4
S2 0.0
S1
Sf
0.5
1.0
1.5
S
the norm of the complex modulus |G*| of 35% and a single rest period of 2 h; for TS-SRP tests the healing indicator is HITS-MRP (Eq. 3) corresponding to a 35% drop in dissipated energy and multiple rest periods of 1800 s; for LASH tests the indicator is HILASH (Eq. 4), obtained at a level of damage of 50%Sf and a single rest time of 1800 s. It is observed that, even if all the tests were performed in oscillatory shear strain mode at common values of temperature and frequency (set at 20 °C and 10 Hz, respectively), a univocal ranking was not achieved. When comparing neat binders B1 and B2, a similar ranking was obtained according to TS-SRP and LASH methods, opposite to that highlighted using TS-MRP. On the other hand, the polymer-modified bitumen B3 revealed the highest healing capability when tested using the TS-MRP method, it showed a similar healing capability to B1 when TS-SRP was used, while it exhibited the worst healing properties with the adoption of LASH protocol. Such unexpected findings bring to light the non-negligible role played by the choice of testing parameters and analytical approach used to assess healing. From these preliminary results, it can be stated that there is an urgent need to converge towards a standardized laboratory protocol capable of providing accurate Fig. 4 Healing indicators according to TS-SRP, TS-MRP, and LASH methods
B1 52%
B2
B3
50% 43% 37%
34% 21% 15%
14% 3%
TS-SRP HITS-SRP
TS-MRP HITS-MRP
LASH HI LASH
62
O. Baglieri et al.
and reliable description of the healing ability of bituminous materials, representative of a wide spectrum of prevalent environmental and loading conditions in the field.
References 1. Bazin, P., Saunier, J.P.: Deformability, fatigue and healing properties of asphalt mixes. In: Proceedings of the 2nd International Conference on the Structural Design of Asphalt Pavements, Ann Arbor, Michigan, pp. 553–69 (1967) 2. Francken, L.: Fatigue performance of as bituminous road mix under realistic best conditions. Transp. Res. Rec. 712, 30–37 (1979) 3. Di Benedetto, H., Nguyen, Q.T., Sauzéat, C.: Nonlinearity, heating, fatigue and thixotropy during cyclic loading of asphalt mixtures. Road Mater. Pavement Des. 12(1), 129–158 (2011) 4. Baaj, H., et al.: Recovery of asphalt mixture stiffness during fatigue loading rest periods. Constr. Build. Mater. 158, 591–600 (2018) 5. Santagata, E., Baglieri, O., Tsantilis, L., Dalmazzo, D.: Evaluation of self-healing properties of bituminous binders taking into account steric hardening effects. Constr. Build. Mater. 41, 60–67 (2013) 6. Canestrari, F., Virgili, A., Graziani, A., Stimilli, A.: Modeling and assessment of self-healing and thixotropy properties for modified binders. Int. J. Fatigue 70, 351–360 (2015) 7. Mazzoni, G., Stimilli, A., Canestrari, F.: Self-healing capability and thixotropy of bituminous mastics. Int. J. Fatigue 92(1), 8–17 (2016) 8. Xie, W., Castorena, C., Wang, C., Kim, Y.R.: A Framework to characterize the healing potential of asphalt binder using the linear amplitude sweep test. Constr. Build. Mater. 154, 771–779 (2017) 9. Wang, C., Castorena, C., Zhang, J., Kim, Y.R.: Unified failure criterion for asphalt binder under cyclic fatigue loading. Road Mater. Pavement Des. 16(S2), 125–148 (2015)
Other Conference Papers
A Comparative Study of RTFO, PAV and UV Aging Using FT-IRS and DSR Tests Roksana Hossain, Shams Arafat, Delmar Salomon, and Nazimuddin M. Wasiuddin
Abstract The objective of this study was to investigate RTRO (Rolling Thing Film Oven), PAV (Pressure Aging Vessel), UV (Ultraviolet), and forced draft oven aging behavior of original binder through dynamic shear rheometer (DSR) and FTIRS (Fourier Transform Infrared Spectrometer). Different aging equipment, namely RTFO, PAV, UV chamber, and forced draft oven, were used to age the binder until equal stiffness (same G*/sinδ) was obtained. A comparison has been performed among these four aging procedures based on their differences in chemical and rheological properties. FT-IRS was used to envisage the chemical alteration, whereas DSR was used to perceive the rheological changes. Samples were prepared in RTFO at 163 °C and were collected at different degrees of RTFO aging. PAV, UV, and oven aged samples were prepared considering conditions without RTFO aging. Results showed that carbonyl (C=O) and sulfoxide (S=O) content of different degrees of RTFO and PAV aged samples increased with duration. On the other hand, C=O and S=O content in UV and oven aged samples decreased after 36 h of aging. At similar stiffness of RTFO and PAV aged binders, C=O content was similar, but S=O content was higher in PAV aged binders. DSR results also showed a resemblance with the FT-IRS results, which indicated that these different aging procedures with different equipment could be correlated with each other. Keywords RTFO · PAV · UV · Forced draft oven · Aging · FT-IRS · DSR
1 Introduction Aging reduces the long-term performance of asphalt binders as aging makes the asphalt binder brittle and stiff. It is crucial to understand the aging mechanism of
R. Hossain · S. Arafat · N. M. Wasiuddin (B) Louisiana Tech University, Ruston, LA, USA e-mail: [email protected] D. Salomon Pavement Preservation Systems, LLC, Garden City, ID, USA © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_8
65
66
R. Hossain et al.
asphalt binder to improve the service life of asphalt pavement. Several researchers [1– 3] studied the aging behavior and tried to find the relationship between the chemical and rheological properties of the aged asphalt binder. According to Li et al., aging degraded the chemical, physical and rheological performance of the neat asphalt binder [3]. The rheological properties of original and aged PG binders are mainly determined by the dynamic shear rheometer (DSR). FT-IRS can detect the carbonyl and sulfoxide content in asphalt binders. Carbonyl content is an excellent measurement for aging in asphalt binder [1]. A linear relation was obtained between the carbonyl and viscosity by Liang et al. [4]. A universal model was developed by Liu et al. to predict the carbonyl content after long-term aging behavior [2]. However, the chemical and rheological behavior of similarly stiff aged binders was not studied in the past. Also, there is a knowledge gap between the RTFO, PAV, UV, and oven aging at equal stiffness. Therefore, this study was initiated to understand the aging mechanism of asphalt binder at similar stiffness by comparing the chemical bondings and rheological behavior after four different types of aging (RTFO, PAV, UV, and forced draft oven).
2 Methodology 2.1 Asphalt Binder A neat unmodified PG 64-22 asphalt binder was selected to conduct the RTFO, PAV, UV, and forced draft oven aging.
2.2 UV and Oven Aging Simulation Test in the Laboratory In this study, UV and force draft oven aging was performed following Hossain and Wasiuddin [5]. Two grams of binder was taken in a small can and mixed with dichloromethane (DCM) to make the binder liquid enough to spread on a 14.1 cm circular PAV plate uniformly. The UV chamber reached at a constant temperature of 70 °C after about 45 min. The PAV plate with asphalt binder and DCM solvent was kept under the hood for drying. The dried PAV plate was placed inside the UV chamber when the UV chamber had reached at 70 °C as shown in Fig. 1a. Similarly, dried PAV plate was placed inside the forced draft oven at 70 °C as shown in Fig. 1b. The thickness of the asphalt binder film was calculated to be 32μ, and intensity of the UV aging chamber was 100 mW/m2 .
A Comparative Study of RTFO, PAV and UV Aging …
67
2.3 Fourier Transform Infrared Spectroscopy (FT-IRS) Test A 4300 handheld ATR (attenuated total reflection) FT-IRS from Agilent was used in this study. Each spectrum was collected in 4000–400 cm−1 wavenumber (fingerprint region) with Microlab PC software. No separate sample preparation method was required to collect the spectrum. The sample was directly placed with a spatula over the FT-IRS crystal, as shown in Fig. 1c. The carbonyl and sulfoxide indices were calculated based on Eqs. (1) and (2). IC=O =
Peak height at 1695 cm−1 Peak height at 2920 cm−1 or Peak height at 1456 cm−1 or Peak height at 1375 cm−1
(1)
IS=O =
Peak height
at 2920 cm−1 or
Peak height at 1031 cm−1 Peak height at 1456 cm−1 or Peak height at 1375 cm−1
(2)
2.4 Experimental Plan In this study, different degrees of RTFO and PAV aging were performed to simulate extreme aging. RTFO aging was performed by following the AASHTO T240 standard. PAV aging was performed by following the AASHTO R28 standard except performing RTFO aging before placing the binder in the PAV chamber. Table 1 represents the experimental plan for the data collection and the result analysis procedure of this study. Table 1 Experimental plan for data collection and result analysis Binder type
Binder condition
PG64-22 (original unmodified binder)
Original RTFO aging
PAV aging (W/O RTFO)
Aging duration 1RTFO
1 h 25 min
2RTFO
2 h 50 min
3RTFO
4 h 15 min
4RTFO
5 h 40 min
6RTFO
8 h 30 min
1PAV
20, 40 and 60 h
2PAV 3PAV
UV aging (W/O RTFO)
24, 36, 48 and 60 h
Forced draft oven aging (W/O RTFO)
24, 36, 48 and 60 h
Test DSR (For rheological characterization) and RT-IRS (For chemical bonding analysis)
68
R. Hossain et al.
3 Result and Discussion 3.1 Analysis of Carbonyl and Sulfoxide Content Figures 2 and 3 show the carbonyl (C=O) and sulfoxide (S=O) peak height indices of original, RTFO, PAV, UV, and oven aged PG 64-22 binder. Due to oxidation, oxygencontaining functional groups such as C=O and S=O increase in asphalt binder. After 1RTFO, 2RTFO, 3RTFO, 4RTFO, and 6RTFO aging, the C=O index of PG 64-22 binder was increased by 23%, 36%, 54%, 65%, and 81%, respectively (Fig. 2). In the case of 1PAV, 2PAV, and 3PAV aging, the C=O index was increased by 59%, 70%, and 72%, respectively (Fig. 3). It was observed that, after 40 h of PAV aging, change in the C=O index was not significant. UV and oven aged binder showed an increment of the C=O index up to 36 h of aging. After 36 h of aging, the C=O index was reduced in both UV and oven aged binders. As shown in Fig. 3, the S=O index was increased by 1.27, 1.61, 1.91, 2.09, and 2.20 times of original binder in the case of 1RTFO, 2RTFO, 3RTFO, 4RTFO, and a
b
c
Fig. 1 a UV aging of PG 64-22 binder, b Oven aging of PG 64-22 binder, c Sample placed over the FT-IRS crystal for data collection
Fig. 2 C=O peak height index (1695 cm−1 /1456 cm−1 ) for PG 64-22 original and aged binder
A Comparative Study of RTFO, PAV and UV Aging …
69
Fig. 3 S=O peak height index (1031 cm−1 /1456 cm−1 ) for PG 64-22 original and aged binder
6RTFO aging, respectively. In the case of 1PAV, 2PAV, and 3PAV aging, the S=O index was increased by 2.83, 3.38, and 3.96 times. In the case of UV aged binders, the S=O index was increased by 1.62, 1.73, 1.43, and 1.39 times after 24 h, 36 h, 48 h, and 60 h of aging. Again, after 24 h, 36 h, 48 h, and 60 h of oven aging, the S=O index was increased by 1.66, 2.07, 2.18, and 1.96 times. An increment of the functional groups (C=O and S=O indices) can make the asphalt binder more viscous leading to brittle and harder asphalt binder.
3.2 C=O and S=O Indices of Aged Asphalt Binder at Similar Stiffness From the DSR temperature sweep test, stiffness (G*/sinδ) and phase angle (δ) values of different RTFO and PAV aged PG 64-22 binders were obtained. It was observed that 3RTFO and 1PAV, 4RTFO and 2PAV, and 6RTFO and 3PAV aged binders have similar stiffness and similar phase angle at 64 °C as shown in Fig. 4. Base binder’s
a
Fig. 4 RTFO and PAV aged (W/O RTFO) PG 64-22 binder a stiffness b phase angle
b
70
R. Hossain et al.
stiffness increased by 12 times after 3RTFO and 1PAV, 7 times after 4RTFO and 2PAV and 10 times after 6RTFO and 3PAV aging (Fig. 4a). Again, the base binder’s phase angle decreased by 11% after 3RTFO and 1PAV, 18% after 4RTFO and 2PAV, and 25% after 6RTFO and 3PAV aging (Fig. 4b). It can be concluded that extreme aging increased stiffness and reduced phase angle of the base binder, which indicated asphalt binder became harder during the aging process. It can be observed from Fig. 5 that at similar stiffness and phase angle, the C=O index of RTFO and PAV aged binder was the same for two different types of indices calculations. Figure 6 represents two different types of S=O indices of RTFO and
Fig. 5 Comparison of C=O index of RTFO and PAV aged PG 64-22 binder at similar stiffness
Fig. 6 Comparison of S=O index of RTFO and PAV aged PG 64-22 binder at similar stiffness
A Comparative Study of RTFO, PAV and UV Aging …
71
PAV aged binders at similar stiffness and phase angle. The S=O index of PAV aged samples was higher than the S=O index of RTFO aged samples in the case of both index calculations. It can be concluded that, at similar stiffness of unmodified aged binder, the C=O index was similar. In contrast, the S=O index was not similar, which indicated the C=O index has a direct correlation with the stiffness of the binder.
4 Conclusion A comparative study between original and four different types of aged (RTFO, PAV, UV and forced draft oven) binder were performed to understand the rheological and chemical behaviors and to find out the relation between the rheological properties and chemical bondings. Oxygen containing functional groups increased after different degrees of RTFO and PAV aging. On the other hand, after UV and oven aging, these compounds increased up to a certain time and then decreased. Stiffness and phase angle values were similar in case of higher degrees of RTFO aging and higher degrees of PAV aging. The C=O content of the aged binders was the same at similar stiffness which indicates C=O content correlates with the stiffness of the binder. In contrast, at similar stiffness, the S=O content of the PAV aged binder was higher than the RTFO aged binder. In further analysis, the relationship between the C=O index and stiffness values would be studied in detail.
References 1. Martin, K.L., Davison, R.R., Glover, C.J., Bullin, J.A.: Asphalt aging in Texas roads and test sections. Transp. Res. Rec. 1269, 9–19 (1990) 2. Liu, F., Zhou, Z., Wang, Y.: Predict the rheological properties of aged asphalt binders using a universal kinetic model. Constr. Build. Mater. 195, 283–291 (2019) 3. Li, Y., Wu, S., Liu, Q., Dai, Y., Li, C., Li, H., Nie, S., Song, W.: Aging degradation of asphalt binder by narrow-band UV radiations with a range of dominant wavelengths. Constr. Build. Mater. 220, 637–650 (2019) 4. Liang, Y., Wu, R., Harvey, J.T., Jones, D., Alavi, M.Z.: Investigation into the oxidative aging of asphalt binders. Transp. Res. Rec. 2673(6), 368–378 (2019) 5. Hossain, R., Wasiuddin, N.M.: Evaluation of degradation of SBS modified asphalt binder because of RTFO, PAV, and UV aging using a novel extensional deformation test. Transp. Res. Rec. 2673(6), 447–457 (2019)
A Critical Analysis of the Standard Used to Evaluate De-icing Damage in Asphalt Materials Lisa Lövqvist , Jiqing Zhu , Romain Balieu , and Nicole Kringos
Abstract De-icing fluids are known to have a potential to negatively affect infrastructure materials such as asphalt. In order to evaluate the resistance of asphalt materials to de-icing fluids, the European standard method EN 12697-41 is commonly used. There are however a number of issues related to the method, such as a high variability of the results and poor correlation between test results and behavior in the field, which may be caused by some of the parameters of the test. This paper aims to identify and investigate some parameters that are of importance to the relevancy of the test. To do this, experimental tests of asphalt mastic and mixture are performed with two concentrations of a deicer and water. Additionally, to gain a better understanding of what occurs in the tested samples during the conditioning and loading, finite element simulations using a microscale model with a mesh based on an X-ray CT scan of a real sample, are performed. The results show that both the geometry and the set-up of the conditioning and mechanical testing can cause misleading results. Based on this, recommendations are made for areas for future studies to improve the test method. Keywords De-icing · Asphalt · Damage · Finite element model · Surface tensile strength
1 Introduction The standard method EN 12697-41 [1] used for evaluating the resistance of bituminous materials to de-icing fluids is associated with a number of problems such as a high variability of surface tensile strengths and failure types (failure in the glue, failure in the asphalt or a combination of both) and the inability to detect differences L. Lövqvist (B) · R. Balieu · N. Kringos Department of Civil and Architectural Engineering, KTH Royal Institute of Technology, Stockholm, Sweden e-mail: [email protected] J. Zhu Swedish National Road and Transport Research Institute, Linköping, Sweden © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_9
73
74
L. Lövqvist et al.
between fluids [2]. Additionally, the results from the laboratory tests are not always consistent with the found behavior in the field [3]. The heterogeneity of the asphalt material and the complex set-up of the test complicate the process of identifying which measures are required to improve the accuracy and applicability of the test. To gain a better understanding of this, it is important to understand what occurs inside the asphalt sample during the different stages of the test. As a first step towards this, this paper aims to identify some of the parameters that may affect the relevancy of the test. In addition to conducting experimental tests, a microscale model [4] is used to investigate how these parameters affect the result and how the test could be improved.
2 Experimental Testing Two types of experimental tests were conducted in this study: testing the infiltration of two concentrations (25 and 50% by weight) of potassium formate (PF) and water in mastic, and testing the resistance of asphalt mixture to the same fluids. In order to test the difference in infiltration properties between the two concentrations of de-icing fluid and water, mastic cylinders consisting of 27 weight% bitumen of paving grade 160/220 and 73 weight% filler were prepared, seen in Fig. 1a. After recording the initial weight of the samples and placing them in the different conditioning fluids, weighing of the samples in saturated surface dry state was performed with regular intervals, from 2 h in the beginning to 11 days by the end of the measuring period of 30 days. Considering the low air void content in the mastic and that no movement of the fluids occur during the conditioning, it can be assumed that any infiltration of the fluids will be due to diffusion. The difference in diffusion between the fluids can therefore be determined by measuring the change in weight of the samples in each group. From these measurements, it was concluded that water diffuses much faster
Fig. 1 a) The mastic samples, b) the recorded weight change for the conditioning groups
A Critical Analysis of the Standard Used to Evaluate De-icing …
75
into the mastic than the de-icing fluids, as shown in Fig. 1b. This could stem from the fact that at high concentrations, a reverse osmosis may occur where the PF attracts the water and thus prevents it from infiltrating the asphalt. This effect should be investigated further as it could explain why asphalt samples conditioned in some deicing fluids have performed better than water in the EN 12697-41 [3]. Additionally, due to this effect, it is vital that the exact concentration of the de-icing fluid used in the field is tested in the lab. The resistance of a dense asphalt mix (ABT16) to the fluids was assessed by following the standard method EN 12697-41 [1]. The asphalt consisted of granite aggregates of maximum 16 mm and 5.7% by mass bitumen of paving grade 160/220. Cylindrical samples with a diameter of 10 cm were compacted and cut to a height of 3 cm. On the cut surface a test surface was defined by drilling a 0.5 cm deep circular groove with a width of 0.45 ± 0.05 cm, 5 cm from the center, on which a metal cylinder was glued. After conditioning the samples in the different fluids for 70 days, the surface tensile strength was tested by applying a force of 200 N/s on the test surface (i.e. the metal cylinder) while fixing the sides of the samples to the bottom of the test machine. Reference samples which were not conditioned were also tested. While this experimental study only uses new materials, the standard method allows both new and old material to be used. A more comprehensive study, including old field material, is therefore recommended to be performed. The recorded surface tensile stresses at failure are given in Table 1. It is seen that the samples conditioned in the de-icing fluid required a 22–27% lower stress to fail than both the reference samples and the water conditioned samples. Additionally, the values recorded for the conditioned samples are more dispersed than for the reference samples. This indicates that the conditioning procedure is one of the main causes of the variability of the results in this method. Adding to this conclusion is the type of failure for each group, presented in Table 2 through the percentage of failure that occurred inside the asphalt (i.e., not in the glue or the interface between the glue and the asphalt). In the reference group all failures occurred inside the asphalt, while in all the conditioned groups the type of failure varied between the asphalt and the glue. This is a strong indication that the glue is also damaged during the conditioning. The recorded tensile strength will then depend Table 1 The recorded surface tensile stress at failure for the different conditioning groups Average failure stress (kPa) Standard deviation
Reference
Water
50% PF
25% PF
1675
1692
1215
1299
123
253
205
400
Table 2 The maximum and minimum percentage of failure occurring in the asphalt for each group Reference (%)
Water (%)
50% PF
25% PF
Max
100
90
100
80
Min
100
25
30
25
76
L. Lövqvist et al.
on the strength of the glue or interface and not on the strength of the asphalt, thus leading to a misleading result.
3 Simulation of Test For the majority of the simulations a micromechanical model was used. The details about the model can be found in [4]. A realistic FE mesh, including mastic, aggregates and air voids, was obtained from an X-ray CT scan of an asphalt sample. A loading cylinder and a plate representing the fixtures during the mechanical loading were added to the meshed geometry. Only one fourth of the geometry was needed in the simulations due to reasonable symmetry of the sample. Figure 2 shows the real sample and the constructed FE geometry. In the simulations, the effect of the conditioning in the desiccator was represented by assuming that all air voids are saturated with moisture at the beginning of the conditioning period. Additionally, no diffusion is assumed to occur in the glue or the metal. The first parameter investigated is the time for conditioning. While it may be enough for one combination of asphalt and deicer, it may not be for another. To show this effect, two different diffusion coefficients were used for the mastic and the aggregates [5]: 1.3 × 10−13 m2 /s and 3.1 × 10−12 m2 /s for the mastic and 1.2 × 10−10 m2 /s and 2.0 × 10−10 m2 /s for the aggregates. As shown in Fig. 3a and b, while there is a difference between the materials, both samples show a significant moisture content gradient between the center and the outer edge of the test surface. This shows that depending on the combination of material and deicer, the already relatively long conditioning period of 70 days may still not be enough to obtain a homogenous moisture content in the sample at the point of testing. The existence of moisture in asphalt is known to deteriorate its properties [6]. It is thus likely that the outer edge of the test surface, which has a higher moisture content and has been subjected to moisture for a longer period, will be more damaged than
Fig. 2 Real sample a) without metal cylinder, b) with metal cylinder. c) The FE geometry
A Critical Analysis of the Standard Used to Evaluate De-icing …
77
Fig. 3 Concentration profile along the y-axis in a material with a) poor diffusive properties, and b) good diffusive properties. c Stress gradient along the y-axis
the inner part. This moisture (and possible damage) gradient is even more important when considering the vertical stress field caused by the simulated mechanical test, shown in Fig. 3c for a homogenous material. It is found that a stress gradient exists on the test surface with an approximately 9 times higher value on the outer edge of the test surface than in the center. This is due to the fact that the test is geometry dependent, thus making even a qualitative ranking of the materials based on the test results impossible. This could explain the poor correlation between laboratory and field results. While these simulations have assumed that no moisture enters the glue or the interface between the glue and the asphalt, the experimental results presented earlier in this paper indicate that it does. This means that the outer parts of the interface and the glue will also have the highest damage due to the moisture in addition to experiencing the highest stress during the testing, thus increasing the risk of failure occurring there. One approach to improve the test by decreasing the moisture gradient is to place the samples on an elevated lattice during the conditioning, increasing the chance for moisture to diffuse into the sample from the bottom, as seen in Fig. 4a. In addition to the improvement of the moisture content profile along the test surface, the moisture
Fig. 4 Concentration profiles obtained from: a) increasing the diffusion from the bottom, and b) the alternative test geometry after 70 and 32 days
78
L. Lövqvist et al.
content is more homogenous through the thickness of the sample which may increase the chance of the fracture occurring inside the sample and not in the interface. Additionally, the glued metal cylinder on the top test surface impairs the possibility of both a homogenous moisture content in the sample and failure occurring in the asphalt during the mechanical test. If, instead, another test was specified which would not require any external parts which hinders the conditioning, for example a direct tension or compression test, the moisture profile could be improved, which is shown in Fig. 4b. This indicates that a much shorter conditioning time of 32 days is required to reach an acceptable moisture content homogeneity. While this simulation is run for a thinner sample (3 cm) than specified for any of the tests, it can still be assumed that the time required to reach a homogenous moisture content for a sample with a 5 cm thickness can be reduced from 70 days.
4 Conclusions and Recommendations for Future Work This paper investigated which parameters that may cause the issues related to the standard method EN 12697-41, using both experimental tests and FE modeling. It can be concluded that there is a high risk that the set-up of the test affects the usefulness of the results. Firstly, it was shown that it is vital to use the same concentration of deicing fluid in the lab as in the field since this affects the diffusive properties. Secondly, the gluing is damaged during the conditioning and will therefore likely fail during the mechanical test instead of the asphalt, thus causing misleading conclusions. It is therefore recommended to investigate whether there is a glue which is not affected by the conditioning process. Thirdly, it was concluded that despite the relatively long time period (70 days) required for the conditioning, the moisture may not reach the center of the test surface, causing a moisture content gradient. This gradient can be slightly improved by elevating the samples on a grid during the conditioning. In summary, the complex, inhomogeneous and geometry dependent stress field in combination with the inhomogeneous moisture content and arbitrary conditioning give results which are not representative of what occurs in the field. This therefore indicates some of the reasons as to why results from the test, including relative measures, may not correspond to the results in the field. From these results it can be recommended that the set-up and the time required for the conditioning are reviewed to ensure a homogenous and realistic moisture content. Alternatively, as it is the gluing and the geometry of the sample that cause most of the issues discussed in this paper, a new mechanical test, such as direct tension or compression, which removes the need for an external glued part and ensures a homogenous stress field, could be considered. Acknowledgements The authors gratefully acknowledge the support of the Swedish Transport Administration within the BVFF program and the KTH Road2Science Center. The authors would also like to thank Henrik Hellman and Andreas Waldemarson at the Swedish National Road and Transport Research Institute for conducting the de-icing resistance tests.
A Critical Analysis of the Standard Used to Evaluate De-icing …
79
References 1. European Committee for Standardization CEN: EN 12697-41 Bituminous mixtures—test methods for hot mix asphalt—part 41: resistance to de-icing fluids (2013) 2. Ekblad, J., Edwards, Y.: Precision of method for determining resistance of bituminous mixtures to de-icing fluids. Mater. Struct. 41(9), 1551–1562 (2008) 3. Waldemarson, A.: Utveckling av provningsmetod för test av avisningsmedel. Presented at Metoddagen 2017. https://asfaltskolan.se/wpcontent/uploads/Metoddagen/MET17/MET17_ Andreas_Waldemarson_Test_avisningsmedel.pdf 4. Lövqvist, L., Balieu, R., Kringos, N.: A micromechanical model of freeze-thaw damage in asphalt mixtures. Int. J. Pavement Eng. (2019) 5. Kringos, N., Scarpas, A., DeBondt, A.: Determination of moisture susceptibility of mastic-stone bond strength and comparison to thermodynamical properties. J. Assoc. Asphalt Paving Technol. 77, 435–478 (2008) 6. Varveri, A., Zhu, J., Kringos, N.: Moisture damage in asphaltic mixtures. In: Huang, S., Di Benedetto, H. (eds.) Advances in Asphalt Materials: Road and Pavement Construction, vol. 1, pp. 303–344. Woodhead Publishing Limited (2015)
A Laboratory Procedure for Improving the Design of Road Rehabilitation Actions: Study of a Real Case in a Highway Pavement Fernando Moreno-Navarro, J. Sierra, M. Sol-Sánchez, and Ma Carmen Rubio-Gámez Abstract Pavement rehabilitation has become an issue of major concern for road administrations. In the coming years, the development of more effective strategies for improving these actions will be crucial. On some occasions, traditional methods used to define pavement rehabilitation solutions are not able to select the best alternative from a sustainable, technical and economical point of view. Given these challenges, a new methodology was developed for use as a complementary tool for optimizing road pavement rehabilitation actions. This methodology is based on a comparative analysis of the structural behaviour of different types of materials and the subsequent selection of the best alternative based on the results obtained. The present paper describes the methodology and summarises its application during the rehabilitation of a highway pavement. The results of this study indicate that the use of this procedure could optimize the consumption of natural and economical resources in rehabilitation actions, offering more competitive alternatives when the available budget cannot cover the costs associated with traditional solutions. Keywords Pavement · Asphalt · UGR-FACT · Bituminous mixtures · Road
1 Introduction In the last decades, pavement rehabilitation has become one of the main concerns for road administrations. Nonetheless, the majority of the tools available for selecting rehabilitation solutions were developed on the basis of experimental observations made several decades ago [1–3] or are based on linear-elastic mechanics [4, 5], and they do not take into account the advantages offered by the latest developments in asphalt materials (high performance modified bitumens, fibres, rejuvenators, interlayer structural reinforcement systems, etc.) and they are not able to integrate F. Moreno-Navarro (B) · M. Sol-Sánchez · M. C. Rubio-Gámez University of Granada, Granada, Spain e-mail: [email protected] J. Sierra Consejería de Fomento y Vivienda de la Junta de Andalucía, Granada, Spain © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_10
81
82
F. Moreno-Navarro et al.
the accumulation and propagation of damage suffered by the materials during their service life (thus some discrepancies could appear between the stresses and strains calculated and those that actually occur in the pavement). Consequently, in many cases the solutions selected to conduct rehabilitations are not the optimal alternative available (in terms of costs and mechanical performance). In addition, the use of these tools limits the application of non-traditional materials (crumb rubber modified binders, high RAP content, warm-mix asphalts, etc.), since they consider these to be structurally more expensive. In this regard, this paper presents a novel methodology (based on structural performance) to select the optimal thickness and type of materials to be used in pavement rehabilitation actions.
2 Description of the Laboratory Procedure The test procedure can be conducted within the time frame of executing the rehabilitation action (it can be implemented easily and rapidly) and at a laboratory scale (with reduced costs). As data input (advisable, but not strictly necessary), it uses the deflection of the pavement to be rehabilitated, as well as the traffic and climatic conditions to withstand. Then, based on the solution proposed by traditional tools (reference solution), other alternatives can be studied at laboratory level using UGR-FACT (University of Granada-Fatigue Asphalt Cracking Test) [6]. Unlike other common tests that evaluate bituminous materials individually, UGR-FACT allows for evaluating specimens composed of more than one layer (including or not include interlayer reinforcement systems) and different qualities of foundations (or defects) can be simulated by varying the stiffness of its spring (Fig. 1). Specimens of at least 3 different thicknesses are manufactured with all the solutions to be studied (reference and the different alternatives to be considered). Then, they are tested using stresses comparable to those to be suffered during their service life (and adapting the stiffness of the spring to simulate the characteristics of the existing structure) and similar environmental conditions (test temperature and specimen moisture contents). Based on the results obtained, design laws (which relates thicknesses and deflections produced in the layer) and fatigue laws (which relates deflections and number of load cycles that cause a complete crack propagation through the layer) are defined, and the equivalent thicknesses of the studied alternatives with respect to that used by the reference solution are determined. For this purpose, horizontal displacements in the bottom of the layer and vertical displacements in the top of the layer (initial displacement) are related to the number of load cycles supported by the specimens (fatigue laws), and to the thickness of the layer (design laws). The displacements produced with the thickness of the reference solution (Fig. 2a) are introducing into the fatigue law to determine its life span (Fig. 2b). Using an inverse process, the associated displacements to this life span in the alternative solutions are defined (Fig. 2c) and finally, the equivalent thicknesses calculated (Fig. 2d).
A Laboratory Procedure for Improving the Design of Road …
83
Fig. 1 Sketch of the test procedure presented
Fig. 2 Summary of the process carried out to determine the equivalent thickness to the solution propose
84
F. Moreno-Navarro et al.
3 Application Example: Study of a Real Case in a Highway Pavement As an example, this section summarises an application of the presented methodology to a rehabilitation action on the A-92 highway in south Spain. Based on the study of deflections carried out in the existing pavement, as well as the Spanish catalogue for pavement rehabilitation design [3], the proposed reference solutions were: a regrowth of the pavement using an overlay with a thickness of 3 cm (in the areas where the deflections measured were less than 40 mm/100); partial removal of the deteriorated layers and subsequent replacement with 10 cm of new asphalt mixtures (in the areas where the deflections measured were between 40 and 51 mm/100). The bituminous mixtures selected for the reference rehabilitation were: a BBTM 11 (EN 13108-2) manufactured with a modified binder PMB 45/80-65 in the case of the regrowth with an overlay of 3 cm; and an AC 22 (EN 13108-1) manufactured with a conventional binder B 35/50 and the same BBTM 11 manufactured with PMB 45/80-65 in the case of the partial removal and replacement with 10 cm of new mixtures (7 cm of AC 22 + 3 cm of BBTM 11). As an alternative to these two mixtures, additional BBTM 11 and AC 22 mixtures were designed using a crumb rubber (CR) modified binder, in order to evaluate the efficiency of substituting traditional mixture with other more sustainable mixtures manufactured with crumb rubber from end-of-use tyres. Table 1 summarises the main characteristics of the mixtures studied. Figure 3a and b shows the fatigue and design laws obtained for the BBTM 11 mixtures tested at 20 °C. Taking the thickness defined for the reference solution in the Spanish design catalogue (3 cm, Step 1), the displacement associated with the reference material BBTM 11 was determined (Step 2). Following this, the number of load cycles supported (life span) by the reference material under this displacement was calculated (Steps 3 and 4). After that, the displacement associated with this life span in the alternative material BBTM 11 CR is determined (Steps 5 and 6) and using it, the equivalent thickness for the reference materials is calculated (Steps 7 and 8). Based on the results obtained, it was observed that using the alternative material BBTM 11 CR it is possible to reduce the thickness of the surface layer by approximately 7%. Following a similar process, it was observed that using the Table 1 Properties of the mixtures studied Mixture ID
BBTM 11
BBTM 11 CR
AC 22
AC 22 CR
Apparent density (mg/m3 ), EN 12697-6
2146
2134
2478
2450
Voids in mixture (%), EN 12697-8 ITS at 15 °C (kPa), EN 12697-23 ITSR (%), EN 12697-12 WTS
(mm/103
cycles), EN 12697-22
16.5 1287 91.9 0.056
18.0 1161 90.7 0.054
4.3 1720 87.3 0.072
4.0 1389 96.9 0.063
A Laboratory Procedure for Improving the Design of Road …
85
Fig. 3 Results obtained during the study of BBTM 11 and AC 22 mixtures: a, c design laws; b, d fatigue laws
alternative material AC 22 CR it is possible to reduce the thickness of the binder layer by approximately 35% (Fig. 3c and d). Based on the information obtained in the laboratory, a cost analysis was conducted for the different solutions (Table 2). It was observed that using the alternative materials (BBTM 11 CR and AC 22 CR) and taking into consideration the reduction of the thicknesses of the layers (from 3 to 2.8 cm in surface layer; and from 7 to 5 cm in binder layer) it is possible to produce an overall saving of around 40,000 e (more than 11% of the original budget). Table 2 Properties of the mixtures studied Regrowth with overlay
Removal and replacement
BBTM11
BBTM11 CR
AC22
AC22 CR
Bituminous mixture cost (e/t)
28.05
28.52
15.44
20.64
Bituminous mixture density (kg/m3 )
2146
2134
2478
2450
Thickness of the layer (m)
0.030
0.028
0.07
0.05
Bituminous mixture per m2 (kg)
64.4
59.8
173.4
122.5
Cost per m2 of rehabilitation (e)
1.81
1.70
2.68
2.53
Surface (m2 )
135,000
Solution cost (e)
243,769.24
230,077.32
443,908.12
419,034.97
Saving (e)
–
13,691.92
–
24,873.16
Saving (%)
–
5.62
–
5.60
99,000
86
F. Moreno-Navarro et al.
Fig. 4 Detail of the trial section before opening to traffic
Based on the results obtained in this study, a trial section was constructed during the rehabilitation action conducted in the A-92 highway in September 2017 (Fig. 4). Currently, this trial section is under evaluation to assess the development of these materials under real environmental and service conditions (and to compare the results with those obtained in the laboratory during the solution design).
4 Conclusions This article presents a novel methodology developed to optimize the resources used in road rehabilitation actions, and to select the most competitive solutions in terms of costs. This laboratory procedure can be carried out at both the project and construction stages (which can be helpful for decision-making). The results obtained in real experiences where this procedure has been applied have demonstrated that it is sensitive to the changes produced in the composition of the materials (such as type and content of binder, and the presence of interlayer reinforcements) in terms of the design of the pavement (thickness of the layers) and the external conditions (load stress applied and temperatures).
References 1. AASHTO: Guide for Design of Pavement Structures. American Association of State Highway and Transportation Officials, Washington, DC (1993) 2. Ministerio de Fomento. Instrucción de Carreteras, Norma 6.3 IC, Rehabilitación de Firmes. Ministerio de Fomento, Madrid, Spain (2003) 3. Department of Transport: Structural Design of Flexible Pavements for Interurban and Rural Roads. Technical Recommendations for Highways, Pretoria, South Africa (1996)
A Laboratory Procedure for Improving the Design of Road …
87
4. AASHTO: Guide for Mechanistic-empirical Design of New and Rehabilitated Pavement Design Structures. American Association of State Highway and Transportation Officials, NCHRP 137A, National Cooperative Highway Research Program, Washington, DC (2004) 5. ICAFIR: Instrucción para el diseño de firmes de la red de carreteras de Andalucía. JAOP/GIASA47-2007, Consejería de Obras Públicas y Transportes de la Junta de Andalucía, Spain (2007) 6. Moreno-Navarro, F., Rubio-Gámez, M.C.: A review of fatigue damage in bituminous mixtures: understanding the phenomenon from a new perspective. Constr. Build. Mater. 113, 927–938 (2016)
A Literature Review of Bitumen Aging: From Laboratory Procedures to Field Evaluation Rodrigo Siroma, Mai Lan Nguyen, Pierre Hornych, and Emmanuel Chailleux
Abstract Bitumen aging increases the risk of pavement cracking. A good understanding of this phenomenon is crucial for designing long-lasting pavements. In this context, this paper presents a review of the relevant literature on neat bitumen aging. A broad set of indexes often used to estimate the aging level and end-of-life criteria of bitumen are analyzed and discussed. Keywords Bitumen · Aging · Laboratory aging · Field aging · End-of-life criteria
1 Introduction Despite being the main binding agent in road construction, bitumen is highly prone to age-hardening, which increases pavement cracking potential. During asphalt mix production and compaction, elevated rate of oxidation and lighter components evaporation occurs due to high-temperatures exposition [1]. From then on, progressive oxidation, steric hardening, and oil exudation take place throughout pavements’ service life. Hence, this paper presents an in-depth literature review of neat bitumen aging.
2 Accelerated Laboratory Aging Protocols Of several laboratory protocols [2] proposed to reproduce field aging effects, most consist of extended heating procedures as it accelerates the rate of bituminous oxidation, which is desirable. However, the use of excessively high temperature may diverge the aging effect in laboratory samples from that of the field [3, 4]. For bitumen, usual procedures are based on relatively thin-film samples exposed to high temperatures to simulate the short- and long-term aging effects, such as Rolling R. Siroma (B) · M. L. Nguyen · P. Hornych · E. Chailleux IFSTTAR, Nantes, France e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_11
89
90
R. Siroma et al.
Thin-Film Oven Test (RTFOT) and Pressure Aging Vessel (PAV), respectively. The Universal Simple Aging Test (USAT) is a 300 μm film test proposed as an optimized alternative to standard procedures [5]. Although it is a static thin-film test, the sample thickness is comparable to that of the bitumen film on the aggregates, in which oxidation can be considered as kinetics-controlled, i.e., absence of oxidation gradient [4]. For asphalt mix, procedures can be classified according to the sample state (loose or compacted), accelerator factor (forced-draft oven, modified pressure, etc.), and aging factor (oxygen, enriched gas, actinic light, etc.). Compacted specimen aging can produce suitable samples for performance tests, whereas loose mix aging, such as [6], yields more homogeneous aging with a high oxidation rate [7]. In the literature, some drawbacks are often stated: aging compacted samples is time-consuming [8], generates oxidation gradient, and can change the void content and geometry of the sample [7]; difficulty in ensuring sufficient adhesion and cohesion when compacting loose aged mix [7, 8]; and elevated conditioning temperature (above 100 °C) may alter aging mechanisms [3, 8]. An interesting approach to aging compacted samples, the Viennese Aging Procedure (VAPro) [8] uses a triaxial cell with compressed air enriched with ozone and nitric oxides (highly oxidative gases present in the atmosphere) at 45–75 °C, usually reached on surface layers in summer. A similar procedure for bitumen aging can be found in [9].
3 Aging Indexes and Relationship with Field-Aging 3.1 Chemical Indexes Fourier Transform Infrared (FTIR) Spectroscopy is widely used to track the oxygen uptake through two main functional groups: carbonyl (C=O), as it has a well-known linear relationship with the log viscosity over time [3, 10]; and sulfoxide (S=O), which is the principal factor that leads to increased viscosity in high sulfur content bitumen [10]. The first stage of oxidation is marked by a rapid formation of both described indexes, where the growth rate of S=O is higher than that of C=O; then, S=O appears to stabilize while C=O amount increases at a constant rate [3]. Despite being a handy and effective tool, the lack of standardization results in divergence of procedures as well as calculation methods between laboratories [11, 12]. Bitumen can be separated into four generic chemical groups—according to increasing polarity—known as SARA (Saturate, Aromatic, Resin, and Asphaltene) fraction. While nonpolar saturate content remains practically unchanged, the components of the other fractions tend to shift from aromatic through resin to asphaltene with aging [3, 13]. The colloidal instability index (Ic) is the ratio between the sum of asphaltene and saturate contents and the sum of aromatic and resin contents. The increase in Ic value over time represents a bitumen more prone to cracking. Bitumens
A Literature Review of Bitumen Aging …
91
extracted from cracked surface layers of European roads indicate as end-of-life criterion the limit value of 0.6 [14]. However, the bitumens studied were softer than those currently used, so this limit value should be adjusted based on bitumen consistency. Differential Scanning Calorimetry (DSC) determines the glass transition temperature (Tg) and the amount of crystallizable fraction. The Tg value can be related to bitumen brittleness [15]. At temperatures above Tg, the crystallization that occurs may represent an undesirable cracking resistance performance [16]. After laboratory aging, an increase in the crystallizable fraction was observed in [13]. In [17], bitumens with higher crystallizable fraction content display poorer low-temperature performance. The Size Exclusion Chromatography (SEC) method classifies bitumens based upon their molecular size distribution. Laboratory aging [13, 18] leads to a remarkable increase in the large molecular size (LMS) region. In high-speed condition [19], it is possible to highlight asphaltene agglomeration [13]. The ratio between its value after and before aging [18] is often used as an aging index. Extracted bitumens from cracked sections of a test road in Montana (US) displayed a higher LMS population [20].
3.2 Physical Indexes Empirical Tests Classical tests such as Penetration and Softening Point have been used for decades as indexes to track neat bitumen aging. Bitumen aging leads to a decrease in penetration and an increase in softening point value over time. For the Zaca-Wigmore Test Road (US), a limit value of 30 was stated for penetration (originally 100–200) as a pavement end-of-life criterion, although lower values were found in extracted bitumen from other test roads with little or no cracks [20]. Further, values lower than 30 can be found in hard bitumen currently used for high-performance asphalt mixes. Mechanical Intrinsic Test Complex modulus (G* = G + iG = |G*|eiϕ ) of bitumen is measured by loading frequency sweep tests at multiple temperatures. The aging level can be tracked through either directly from test measurements or parameters of linear viscoelastic models. Laboratory aged bitumen [21] and bitumen extracted from loose aged asphalt mixes [11, 22] display the following evolution of the analogical Huet-Such parameters with aging: the glassy modulus (spring G∞ ) and parabolic element (k) remain unchanged; the parabolic element (h) decreases in [11, 22], but remains unchanged in [21]; the dashpot (β) considerably increases; the δ parameter tends to increase; and the τ parameter considerably increases. The Christensen-Anderson-Marasteanu (CAM) model is based on analytical equations whose parameters have physical meanings. During aging [21, 23], the
92
R. Siroma et al.
crossover frequency (ωc ) and modulus (G*C ) reduce; the v value decreases; the glassy modulus (G∞ ) remains unchanged; and the R-value increases, indicating a flattening of the master curve. The evolution of the crack resistance at low temperatures can be outlined by graphs, such as the ωc versus R-value [24] and the crossover modulus (G*C ) versus crossover temperature (Tδ=45° ) [25]. The risk of thermal cracking of asphalt mix may be evaluated by temperatures corresponding to the phase angle of 27º (Tδ=27° ) or 45º (Tδ=45° ), both at 7.8 Hz [26, 27]. In [27], for 35/50 pen bitumens extracted from cracked road sections in France, Tδ=27° = 12 °C and Tδ=45° = 35 °C. The GloverRowe (G-R) parameter has been largely used to assess the cracking potential of asphalt mixes. Based on previous studies, Rowe [28] determined the following two criteria: Damage onset at 180 kPa and significant cracking at 600 kPa. The G-R parameter showed good accuracy in predicting the cracking susceptibility of aged sections [29]. Creep modulus can be measured by the Bending Beam Rheometer (BBR) test for evaluating the bituminous material’s capacity to resist low-temperature cracking. At 60 s loading time, the Ts is the temperature at which the stiffness (S) is ≤300 MPa and the Tm, related to relaxation properties, is the temperature at which the slope m of the log-stiffness versus log-time curve presents a value ≥0.3 [30]. The Tc parameter is the difference between TS and Tm [31]. As bituminous material ages, the curve flattens, and Tc becomes increasingly negative [24]. A limit value of Tc ≤ −2.5 °C indicates the damage onset, while Tc ≤ −5 °C implies the need for more urgent remediation. The Christensen-Anderson (CA) model can be used as a surrogate of BBR [32], whose advantage is to reduce the amount of bitumen required.
3.3 Physical and Chemical Coupling Approaches Several attempts and approaches have been proposed to build relationships between the chemical properties and physical performance of the bitumen. The δ-Method was proposed by Themeli et al. [22] to estimate the Apparent Molecular Weight Distribution (AMWD). It is based on the relationship between applied frequency and molecular relaxation time. According to the δ-Method, aging results in shifting from the lower molecular weight fraction (intermicellar environment) to a high weight population (asphaltene agglomeration) [11, 22]. The crossover frequency corresponds to the AMWD median value and evolves toward higher weight populations, which is in accordance with the literature. The δ-Method appears promising to connect rheological properties with the colloidal structure of bitumen. Due to the good correlation stated in the evolution of chemical indexes of laboratory- and field-aged material, Qin et al. [33] proposed an approach to predict the evolution of the rheological properties that occur in field pavements. A satisfactory relation was showed between chemical, such as FTIR carbonyl + sulfoxide and the total pericondensed aromaticity (TPA) determined by SAR-AD fraction, and
A Literature Review of Bitumen Aging …
93
Table 1 End-of-life criteria found in the literature Aging index
Estimated end-of-life criterion
References
Ic
Higher than 0.6
[14]
Penetration
Lower than 30
[20]
Tδ=27°
Higher than 12 °C at 7.8 Hz
[27]
Tδ=45°
Higher than 35 °C at 7.8 Hz
[27]
G-R value
Damage onset = 180 kPa; significant cracking = 600 kPa
[24]
Tc
Damage onset ≤ −2.5 °C; significant cracking ≤ −5 °C
[31]
mechanical indexes, such as the crossover modulus (G*c) and the characteristic relaxation time (λ). These latter parameters can be used to determinate the R-value and the ωc for different aging stages of bitumen, therefore, making possible the building of master curves through the Christensen-Andersen (CA) model. This procedure shows a good correlation with the results of bitumens extracted from the field. Similar to the previous study, Marsac et al. [11] stated that the linear relationship between the four Huet modified model parameters (h, δ, β, and log(τ)) and the carbonyl index makes possible to estimate the evolution of G* during aging. However, no validation was performed in this study. An alternative approach was proposed by Elwardany et al. [34] to determining Tc through the Viscoelastic Temperature Range (TVE ), represented by the temperature at the glassy-viscoelastic behavior transition, and the crossover temperature (viscoelastic-viscous behavior transition), both determined by the 4 mm Dynamic Shear Rheometer test. TVE was identified as the temperature corresponding to the maximum loss modulus at low temperature and display a satisfying correlation with the modulated DSC glass transition temperature (Tg). For some bitumen the peak is unclear. Despite the good correlation, this approach still needs to be further validated.
4 Conclusion As bitumen aging makes bituminous material more prone to cracking, several techniques that better reproduce the field aging effects through increasingly optimized laboratory procedures have been proposed. Despite using different approaches, the discussed indexes display a general coherence of the evolution of the bituminous material’s behavior. According to the incidence and positive evaluation in the literature, mechanical and some chemical indexes—especially FTIR—are the most recommended, due to their sensitivity to track the aging evolution. Some end-of-life criteria proposed in the literature, summarized in Table 1, are no longer suitable to assess currently used bitumens, especially the harder ones for high-performance asphalt mixes. Finally, attempts at physical and chemical coupling proposed by recent studies may be potential tools toward a better understanding of this phenomenon, thus enabling the construction of longer-lasting pavements.
94
R. Siroma et al.
Acknowledgements The authors gratefully acknowledge the French National Research Agency (ANR) for the financial support of the MoveDVDC project (ANR-17-CE22-0014).
References 1. Hunter, R., Self, A., Read, J.: The Shell Bitumen Handbook, 6th edn. ICE, London (2015) 2. Airey, G.D.: State of the art report on ageing test methods for bituminous pavement materials. Int. J. Pavement Eng. 4(3), 165–176 (2003) 3. Petersen, J.C.: A Review of the Fundamentals of Asphalt Oxidation: Chemical, Physicochemical, Physical Property, and Durability Relationships. TRC, No. E-C140 (2009) 4. Rad, F.Y., et al.: Investigation of proper long-term laboratory aging temperature for performance testing of asphalt concrete. Constr. Build. Mat. 147, 616–629 (2017) 5. Farrar, M., et al.: The universal simple aging test (USAT) and low-temperature performance grading using small-plate dynamic shear rheometry: an alternative to standard RTFO, PAV, and BBR for HMA and WMA. FHWA, Fundamental properties of asphalts and modified asphalts III product: FP 14 (2015) 6. de la Roche, C., et al.: Development of a laboratory bituminous mixtures ageing protocol. Advanced Testing and Characterization of Bituminous Materials, vol. 2 (2009) 7. Elwardany, M.D., et al.: Evaluation of asphalt mixture laboratory long-term ageing methods for performance testing and prediction. RMPD 18, 28–61 (2017) 8. Steiner, D., et al.: Towards an optimised lab procedure for long-term oxidative ageing of asphalt mix specimen. Int. J. Pavement Eng. 17(6), 471–477 (2016) 9. Mirwald, J., et al.: Impact of reactive oxygen species on bitumen aging – The Viennese binder aging method. Cons. Build. Mat. 257, 119495 (2020) 10. Petersen, J.C., Glaser, R.: Asphalt oxidation mechanisms and the role of oxidation products on age hardening revisited. RMPD 12(4), 795–819 (2011) 11. Marsac, P., et al.: Recycling. In: Testing and characterization of sustainable innovative bituminous materials and systems: state-of-the-art report of the RILEM technical committee 237-SIB. Springer (2018) 12. Dony, A., et al.: MURE National Project: FTIR spectroscopy study to assess ageing of asphalt mixtures. In: 6th Eurasphalt & Eurobitume Congress, Prague (2016) 13. Le Guern, M., et al.: Physico-chemical analysis of 5 hard bitumens: identification of chemical species and molecular organization before and after artificial aging. Fuel 89(11), 3330–3339 (2010) 14. Huet, J.: Spectrographie infrarouge et composition chimique globale des bitumes en cours d’évolution dans les sections routières expérimentales. In: 4th Eurobitume, Madrid (1989) 15. Kriz, P., Stastna, J., Zanzotto, L.: Physical Aging in Semi-crystalline Asphalt Binders. Asphalt Paving Technology: AAPT 77, pp. 795–826, Philadelphia (2008) 16. Beghin, A.: Apport de mesures rhéologiques et de pelage à l’analyse de la rupture de liants bitumineux. Université Pierre et Marie Curie-Paris VI, Paris (2003) 17. Poulikakos, L.D.: RILEM 252-CMB Symposium: Chemo-mechanical Characterization of Bituminous Materials. vol. 20, pp. 77–83. Springer (2019) 18. Lee, S.-J., et al.: Short-term aging characterization of asphalt binders using gel permeation chromatography and selected Superpave binder tests. CBM 22(11), 2220–2227 (2008) 19. Brûlé, B., Ramond, G., Such, C.: Relationships between composition, structure, and properties of road asphalts: state of research at the French public works central laboratory. In: 65th Annual Meeting of the TRB (1986) 20. Finn, F.N., et al.: Asphalt Properties and Relationship to Pavement Performance: Literature Review. SHRP Task 1.4, CA (1990)
A Literature Review of Bitumen Aging …
95
21. Yusoff, N.I.M., Airey, G.D., Hainin, M.R.: Predictability of complex modulus using rheological model. Asian J. Sci. Res. 3(1), 18–30 (2010) 22. Themeli, A., et al.: Molecular weight distribution of asphaltic paving binders from phase-angle measurements. RMPD 16, 228–244 (2015) 23. Anderson, D.A., Marasteanu, M.: Continuous models for characterizing linear viscoelastic behavior of asphalt binders. In: ISAP Workshop, Madison (2010) 24. Rowe, G.M.: Tc—some thoughts on the historical development. In: Binder ETG (2016) 25. Widyatmoko, D., Elliott, R.: Mapping crack susceptibility of bituminous materials with binder durability. In: 5th RILEM International Conference, Limoges (2004) 26. Ramond, G., Such, C.: Le module complexe des liants bitumineux. LCPC, Paris (2003) 27. Ballié, M., Migliori, F.: Étude de la fissuration par le haut des bétons bitumineux. LCPC, Paris (1999) 28. Rowe, G.M.: Prepared discussion for the AAPT paper by Anderson et al.: Evaluation of the relationship between asphalt binder properties and non-load related cracking. Journ. AAPT 80, 649–662 (2011) 29. Rowe, G.M., King, G., Anderson, M.: The influence of binder rheology on the cracking of asphalt mixes in airport and highway projects. J. Test. Eval. 42(5), 1063–1072 (2014) 30. Glover, C.J., et al.: Development of a new method for assessing asphalt binder durability with field validation. Texas Transportation Institute, College Station (2005) 31. Anderson, R.M., et al.: Evaluation of the Relationship between Asphalt Binder Properties and Non-Load Related Cracking. AATP, Tampa (2011) 32. Rowe, D.G.M.: New methods for assessing rheology data such as Tc and G-R parameter and their relationship to performance of REOB in asphalt binders and other materials. In: Asphalt Mix and Binder ETG Meeting, Ames (2017) 33. Qin, Q., et al.: Field aging effect on chemistry and rheology of asphalt binders and rheological predictions for field aging. Fuel 121, 86–94 (2014) 34. Elwardany, M.D., Planche, J.-P., Adams, J.J.: Determination of binder glass transition and crossover temperatures using four-mm plates on a dynamic shear rheometer: significance and applications. TRB 2019 Annual Meeting, Washington (2019)
A Method to Reduce Occlusion While Measuring Pavement Surface Profiles Using Triangulation Based Laser Scanners Subham Jain, Animesh Das, and K. S. Venkatesh
Abstract One of the considerations, in the characterization of pavement surface, includes an accurate measurement of the pavement surface profile. With the advent of modern technologies, automated methods (such as laser scanning) are being preferred over manual methods (such as the sand patch test). The laser scanning measurement method is based on the principle of active triangulation, where a camera (laser detector), a laser emitter, and the illuminated laser spot form a triangle. However, due to obstruction from other nearby parts of the profile, this triangulation may not get completed for certain points of the surface profile, leading to occlusion of some data-points. In such a situation, generally, as per the current practice for pavement surfaces, interpolation is used to estimate the height of the occluded data points. In this study, an approach is suggested where two laser emitters are used (rather than one) to reduce this occlusion. Thus, even if the camera cannot see a point illuminated by one laser emitter, it may be able to capture the same point when it is illuminated by the other laser emitter. This minimizes (in some cases, altogether eliminates) occlusion. Experimental results are presented to support this approach. Keywords Pavement surface · Laser profiling · Triangulation · Occlusion
1 Introduction For pavement surface profile measurement, modern automated methods are preferred over manual ones because of reduced subjectivity, greater accuracy, and faster measurement. Triangulation based laser scanning is one such automated measurement method [1–3]. In this method, a camera (laser detector), a laser emitter, and the point illuminated by the laser emitter (laser spot) together form a triangle. Subsequently, the principle of triangulation is used to measure the height of that illuminated point from its image.
S. Jain · A. Das (B) · K. S. Venkatesh IIT Kanpur, Kanpur, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_12
97
98
S. Jain et al.
Fig. 1 Occurrence of occlusion
Occlusion happens when the triangle between the laser, camera, and the illuminated point does not get completed (refer Fig. 1). As seen in Fig. 1, the point marked ‘A’ is not being lit by the laser source and hence getting occluded.
2 Background and Scope Triangulation based laser scanning is often used for the measurement of pavement surface undulations [1–3]. In the triangulation process, there is always a possibility of occlusion happening at some points in the profile, due to the obstruction caused by the other nearby points. In one study, two commercial laser scanners, consisting of two self-contained source-sensor assemblies were used, ostensibly, to reduce the data occlusion [4]. A few studies have compared various interpolation techniques for replacing the occluded data points [4, 5]. In the present work, an alternative approach is suggested to reduce occlusion, where two laser emitters are used (rather than one). Further, a suitable data processing algorithm is suggested to process the surface profile data generated by the two laser emitters.
3 Experimental Work As mentioned in the previous section, in the present study, two laser emitters (one of 532 nm wavelength (green) and the other of 648 nm wavelength (red), and both having 50 mW power) and one camera (of focal length 5–50 mm) are used. The lasers chosen are of different colours—green and red so that the laser source of an illuminated point can be readily identified in the images. The camera and the pair of laser emitters are attached to a linear motion guide so that the entire set-up can be made to move along a line in order to scan a given pavement surface. Figure 2
A Method to Reduce Occlusion While Measuring Pavement Surface …
99
Fig. 2 Illustration of the position of an illuminated point
shows the position of one such point illuminated in the entire scanning process. Further details of the design and development of equipment and the calibration of data acquired have been discussed elsewhere [6]. In an ideal case, when there is no obstruction (that is, the surface is perfectly planar), each point on this surface would get illuminated twice, and therefore, would be measured twice. In order to illuminate each point (on this perfectly planar surface) twice, the linear motion guide must traverse a certain fixed distance (X0 , as seen in Fig. 3a). In a case when the surface is not perfectly planar (that is, if all the points do not have the same height), then this distance would generally not always be constant at X0 throughout, in order to illuminate each point twice (as seen in Fig. 3b), but would vary depending on the height of the point being illuminated. Thus, for a road surface that has different heights at different points (due to its natural undulations), X would not be the same for each point. Estimation of X for each point would be cumbersome (as there can be thousands of points). To reduce complexity, this distance has been kept fixed all the points in this study. This fixed distance is taken to be equal to the distance for a perfect plane located at the mean height of the pavement surface under consideration (X1 as seen in Fig. 4).
a. Illustration of the distance to be moved by the camera and laser emitter pair to illuminate a given point twice, when the surface is perfectly plane.
b . Illustration of the distance to be moved by the camera and laser emitter pair to illuminate a given point twice, when the surface is not perfectly plane.
Fig. 3 Illustration of the distance to be moved by the camera and laser emitter pair to illuminate a given point twice
100
S. Jain et al.
Fig. 4 Schematic diagram explaining data acquisition of pavement surface
Subsequently, scanning was performed on a pavement core sample, and images of the laser-lit surface were captured. Appropriate image processing steps were applied to these images to obtain the height information of the individual surface profile points. As seen in Fig. 5a and b, some points got occluded (due to the reasons explained in the section titled ‘Introduction’). In order to reduce the occlusion in these two measurements, the following method is proposed: • If a point gets illuminated by the green laser emitter but not by the red laser, then the measurement made due to the green laser emitter is taken. • If a point gets illuminated by the red laser emitter but not by the green laser, then the measurement made due to the red laser emitter is taken. • If a point gets illuminated by both the green laser and the red laser, then the average of the measurement made due to both the laser emitters is taken. • If, however, some points get missed by both the laser emitter, then that point remains occluded and may be addressed by conventional interpolation. The results of the sample data are shown in Fig. 6. In Fig. 5a, two data-points were occluded, and in Fig. 5b, four data-points were occluded, whereas in Fig. 6,
a Surface profile acquired by the green laser emitter
Fig. 5 Surface profile acquired by both the laser emitters
b Surface profile acquired by the red laser emitter
A Method to Reduce Occlusion While Measuring Pavement Surface …
101
Fig. 6 Surface profile obtained by replacing the missed data points
no data-points were occluded. Thus, the data collected on the ten different lines are summarized in Table 1. Therefore, the occlusion (which would have occurred if only one laser emitter is used) is minimized by using the method discussed in this study. Table 1 Data collected on ten lines and the occlusion-free data obtained Sample No.
Percentage of occlusion-free data obtained due to surface lit by green laser
Percentage of occlusion-free data obtained due to surface lit by red laser
Percentage of occlusion-free data obtained after following the suggested approach
1
95.77
92.96
100
2
94.37
100
100
3
92.96
95.77
100
4
97.18
94.37
100
5
98.59
95.77
100
6
98.59
91.55
100
7
98.59
92.96
100
8
95.77
94.37
100
9
91.55
85.92
100
10
85.92
84.51
100
102
S. Jain et al.
4 Discussion The present study suggests an approach to minimize occlusion of data in triangulation-based laser scanning. Instead of using only one laser emitter, two were used to minimize the occlusion and improve the data quality. In the present work, surface profile is obtained by averaging the known points (that is, measured by both the laser emitters) and replacing the missed points. In Fig. 6, in which the surface profile obtained by replacing missed data points is plotted, it can be seen that the X coordinate of some of the matching points is not same. That is, the X coordinate of a point which is once estimated by the red laser and then again by the green laser, is different. This error in X coordinate estimation could be attributed to using a fixed value of X, that is X1 (refer Fig. 4), for matching each point on the profile. As discussed earlier, X cannot be same for all the points on the pavement profile, rather it is dependent on the height of each point. Further, due to the error in height measurement made by the laser device, the mean height of the pavement surface could have been measured erroneously. Since, X1 is estimated from mean height of the profile, this error could have then carried forward. In future work, an improved algorithm can be developed, which would be more accurate by avoiding this approximation of keeping the X fixed. Further, the algorithm could remove propagation of error in height estimation to error in position estimation. Acknowledgements Professor Partha Chakroborty, Department of Civil Engineering, IIT Kanpur, for the useful discussion.
References 1. Bitelli, G., Simone, A., Girardi, F., Lantieri, C.: Laser scanning on road pavements: a new approach for characterizing surface texture. Sensors 12(7), 9110–9128 (2012) 2. Zuniga-Garcia, N., Prozzi, J.A.: High-definition field texture measurements for predicting pavement friction. Transp. Res. Rec. 2673(1), 246–260 (2019) 3. Li, Q., Yang, G., Wang, K.C.P., Zhan, Y., Wang, C.: Novel macro- and microtexture indicators for pavement friction by using high-resolution three-dimensional surface data. Transp. Res. Rec. 2641(1), 164–176 (2017) 4. Cigada, A., Mancosu, F., Manzoni, S., Zappa, E.: Laser-triangulation device for in-line measurement of road texture at medium and high speed. Mech. Syst. Signal Process. 24(7), 2225–2234 (2010) 5. Vilaça, J.L., Fonseca, J.C., Pinho, A.C.M., Freitas, E.: 3D surface profile equipment for the characterization of the pavement texture—TexScan. Mechatronics 20(6), 674–685 (2010) 6. Jain, S., Venkatesh, K.S., Das, A.: Design of an active triangulation based measurement device for pavement surfaces, Int. J. Pavement Eng. (2021). https://doi.org/10.1080/10298436.2021. 1873329
A New Approach to Calibration and Use of Mechanistic-Empirical Design Methods Rongzong Wu, John T. Harvey, and Jeremy Lea
Abstract Mechanistic empirical (M-E) pavement design methods have been available for practice since the late 1970s, and many agencies use them for design and analysis of new pavement and rehabilitation projects. The mechanistic part has typically been based on mechanics and laboratory testing, sometimes validated with data from well-instrumented accelerated pavement test (APT) sections. The empirical transfer functions have traditionally been calibrated with small numbers of field sections from which materials properties have been sampled. This paper presents a new approach for calibration, that provides better compatibility with pavement management systems and future use of performance related specifications. Keywords Mechanistic-empirical design · Calibration · Performance related specification · Materials library · Pavement management system
1 Background Mechanistic empirical (M-E) pavement design methods have been available for practice since the late 1970s, and many agencies use them for design and analysis of new pavement and rehabilitation projects. The mechanistic part has typically been based on mechanics and laboratory testing, sometimes validated with data from wellinstrumented accelerated pavement test (APT) sections. The mechanics models for fatigue damage, fracture and permanent deformation have been improved considerably over the years, however, the results must still be translated using transfer functions to pavement management units of distress such as percent of wheelpath cracked and rut depth. The calibration of the empirical transfer functions has typically relied on data from a small set (10–50) of APT sections, with their limitations of replicates, loading and climate conditions, or on sampling from a small set of field sections (10–100) from which materials are sampled and tested. R. Wu · J. T. Harvey (B) · J. Lea University of California Pavement Research Center, Department of Civil and Environmental Engineering, Davis, CA 95616, USA e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_13
103
104
R. Wu et al.
The results of these calibration efforts have not been fully satisfactory. First, the performance models are found to be difficult to calibrate in the sense that it is common to see large differences between model predictions and observations, which can be attributed to limits of the models. Second, there is a disconnect between the calibration, where materials from the test section are paired with the test section performance, and design practice in a “low-bid” environment, where the project is won by the contractor who will perform the work at the lowest cost. Low-bid project delivery is also called “design-bid-build” project delivery because the design is fully completed without knowledge of the materials other than that they meet minimum standard specifications, and then put out for bid. In typical calibration the “reliability” factors are based on the differences between the observed performance and the predicted performance, without accounting for the fact that the calibration was performed with full knowledge of the materials used on each calibration section. These calibrations do not distinguish between the variability of performance within a project from construction variability for a single contractor, and the variability between projects that comes from not knowing which contractor will have the winning bid and what material they will bring to the construction site. The betweenproject variability can be reduced if there are performance related specifications (PRS) in the contract. In this paper, a framework is proposed to address these obstacles and difficulties of calibration. This new approach uses network level performance data from the pavement management system (PMS) with orders of magnitude more observations and length of pavement than are used in the traditional approach. The framework does not require sampling of materials from specific sections in the network, but rather uses mechanistic testing from a representative sample of materials across the network, or regions within the network to perform the calibration. Variability of performance and reliability of design (probability that the design will meet or exceed the design life) is accounted for through separate consideration of within-project and between-project variability.
2 CalME Mechanistic-Empirical Design Software CalME is a software program developed for mechanistic-empirical design using incremental-recursive damage models and Monte Carlo simulation to account for within-project variability, that was initially calibrated using extensive APT and some field sections [1–4]. A new web based version of CalME (v3.0) has been developed in the last two years to replace the previous (v2.0) desktop version. CalME was incorporated into the California Department of Transportation (Caltrans) Highway Design Manual in the early 2010s. Caltrans requested in early 2019 that CalME v3 be recalibrated and its results rechecked for validity and reliability. The University of California Pavement Research Center (UCPRC) proposed that the calibration be done using a new approach that would take advantage of Caltrans’ investments in its pavement management system databases for as-builts,
A New Approach to Calibration and Use of Mechanistic-Empirical … Table 1 Pavement management system performance data for calibration
Pavement type New asphalt pavement, aggregate base New asphalt pavement, other bases Asphalt overlays on asphalt Asphalt overlays on concrete
105 Observations 8530
Lane-km 1701
3292
672
147,837
32,723
9331
2656
its collection of condition survey data since 1978, and extensive use of PRS tests on the various types of asphalt mixes it uses. The data being used for calibration of asphalt pavements are summarized in Table 1. In contrast, to this “big data”, the total number of all types of pavement sections in the US Long-Term Pavement Performance (LTPP) General Pavement experiment is 934 [5].
3 Observations Leading to New Calibration Method Two types of contributions to variability and therefore reliability had been identified by the UCPRC in previous work, defined as “within-project variability” and “between-projects variability”. Within project variability comes from variation of the natural subgrade, and variability of construction using the given set of materials that a contractor brings to a single project. Within project variability considers the rate of development of distress within the project, i.e. the time for distress to appear starting from the weakest sections and finishing with the strongest sections (Fig. 1). If there is no within-project variability then theoretically the entire project would fail at exactly the same time. This of course does not happen in practice. For all other conditions constant, if two contractors have the same median materials properties Fig. 1 Two different within-project variabilities
106
R. Wu et al.
Fig. 2 Between-projects variability for two projects
and construction quality, and one contractor has greater within-project variability, the project with the more variable materials and construction will reach failure sooner. Monte Carlo simulation is used in CalME to simulate within-project variability. Between-projects variability is the uncertainty regarding the materials that a contractor will bring to the project in a low-bid environment. As shown in Fig. 2, if two contractors have the same within-project variability, but one contractor brings an asphalt material with a better combination of stiffness and fatigue properties, then the project built by the first contractor will reach failure later. Without PRS, asphalt materials only need to pass binder specifications and volumetric mix design requirements and the stiffness and fatigue properties can vary widely despite the mix passing the volumetric specifications. The performance related properties of the materials used for calibration with PMS data in the new approach are not known. Instead, data from representative mixes for the time period of performance data in the PMS, tested in the 1990s and 2000s, were used to represent median properties for that time period for all sections. The bias of different within-project variabilities when considering different shift factors for between-projects variability can be eliminated by shifting all sections to the median performance (50% of the wheelpath cracked) because the within-project variability does not affect this inflection point. Asphalt mix was found to be by far the most important variable and almost a dominant controlling performance in CalME simulations. This indicates that mixes representative of the time period of the performance data should be used in the calibration, and that mixes in the design materials library need to be updated as mix design technology changes. It was assumed that calibration using very large amounts of data available in the PMS performance data and representative mix data for that time period would provide a more comprehensive calibration than using detailed sampling and testing of materials from a few projects. Once calibrated, comparisons with sections with detailed
A New Approach to Calibration and Use of Mechanistic-Empirical …
107
sampling and testing can provide additional validation. The weighted average performance of a set of mixes was used from the UCPRC databases that is representative of the time periods present in the cracking and rutting performance data. It was observed that performance data tend to become biased over time, showing less average cracking at later years as pavements that have cracked are maintained or rehabilitated, typically when 10–50% of the wheelpath was cracked. Fitting of the existing data for each segment on the network and extrapolating the trend of cracking versus time provided a reasonable set of curves of crack initiation and increasing extent up to and beyond 50% of the wheelpaths cracked.
3.1 Calibration Steps The steps in the calibration approach used for CalME are: 1.
2.
3.
4.
5.
Identify and assemble information. This includes pavement management system data, which consists of performance data from condition surveys on projects segmented to an average of about 0.06 km (wheelpath cracking, rutting), and as-built information (date of construction, treatment type, other explanatory variables), organized into time series with explanatory variables. This information is used to evaluate the sensitivity of performance data to explanatory variables and develop a matrix of the most important variables. Performance data and explanatory variables are assembled for each cell of the matrix. Calibration focuses on those cells that have greater quantities of performance data. Explanatory variables to be included are those within the designer’s control or knowledge and for which the agency also has data: truck traffic, materials types, thicknesses, and climate regions. Explanatory variables that are outside the control or knowledge of the designer and for which there is no information in the data available are: materials properties for a given type and construction quality. Find or create a mix with properties representative of the period in which the projects in the performance data were built. Run CalME for each cell of the matrix and plot damage versus time using the representative mixes and take the weighted average to reflect previous Caltrans practice. For each project in a cell fit a cracking versus time curve to the observed cracking data from the pavement management system and extrapolate it to 100% wheelpath cracking. This is shown in Fig. 3. For each project in a cell, where necessary, extrapolate the observed cracking data versus time curve to find the damage at which cracking is equal to 50% extent. This is the damage to median cracking (50% probability of cracking) on a project. This is shown in Fig. 3. Shift all observed cracking time histories to pass through the single inflection point at 50% cracking of the wheelpath, as shown in Fig. 3. Calibrate the CalME
108
R. Wu et al.
Fig. 3 Project extrapolated time histories with between-projects shift to median time to 50% wheelpath cracking, resulting in between project calibration
6.
7.
8.
cracking model by shifting the CalME simulations to pass through the same point. Fit the within-project variability. Review information regarding construction variability for thickness and compaction and use it as the starting point for simulating within-project variability using Monte Carlo simulation, sampling between runs for thickness, and material stiffness and cracking performance. These results indicated that the literature results for variability of construction used to calibrate CalME v2 were satisfactory. Validate the models now that they are calibrated for both between-projects and within-project variability by rerunning CalME for each cell of the matrix and comparing the calculated results with the project results in the cell. For forward design using CalME, reliability of the designs using CalME can potentially focus on a mix properties, with different low percentile properties used for designs for commensurate percentile risks of failure for non-PRS projects. Alternatively, the variance of the shift factor to bring all projects with same structure, traffic, and climate, to the same point of 50% of the wheelpath cracked, combined with median within-project variability, can be considered as a statistically sound alternative to calculate reliability (probability of failure prior to the intended design life). This decision will be finalized once the calibration is completed.
A New Approach to Calibration and Use of Mechanistic-Empirical …
109
4 Conclusions A new approach has been developed for calibration of mechanistic-empirical pavement design methods. The method explicitly and separately considers betweenprojects and within-project variability. The new approach permits use of big data from pavement management systems by assuming median properties for materials, and calibrating for different structures, material types, traffic, and climate. This approach offers better integration with pavement managements systems and performance related specifications. It is also more statistically sound than the traditional method of calibration.
References 1. Ullidtz, P., Harvey, J., Basheer, I., Jones, D., Wu, R., Lea, J., Lu, Q.: CalME, a mechanisticempirical program to analyze and design flexible pavement rehabilitation. TRR 2153, 143–152 (2010) 2. Ullidtz, P., Harvey, J., Kannekanti, V., Tsai, B.-W., Monismith, C.: Calibration of mechanisticempirical models using the california heavy vehicle simulators. In: Proceedings, 10th International Conference on Asphalt Pavements, Quebec City, Quebec (2006) 3. Ullidtz, P., Harvey, J., Tsai, B.-W., Monismith, C.: Calibration of mechanistic empirical models for flexible pavements using the WesTrack experiment. AAPT 77, 591–630 (2008) 4. Wu, R., Harvey, J.: Calibration of asphalt concrete cracking models for california mechanisticempirical design (CalME). In: 7th International Conference on Cracking in Pavements, RILEM, Delft, Netherlands, pp. 537–547 (2012) 5. LTPP FAQ page. https://highways.dot.gov/pavements-and-materials/long-term-pavement-per formance/frequently-asked-questions. Accessed 15 Nov 2019
A New Approach Towards Scientific Evaluation and Performance-Related Design of Bituminous Joint Sealing Materials and Constructions J. Buchheim, Ch. Recknagel, P. Wolter, and K. Kittler-Packmor
Abstract While there is a permanent improvement of concrete pavement mixtures and pavement construction types over the last decades, the state-of-the-art joint sealing materials and joint constructions seem to stagnate on an antiquated empirical level. This status has been reaffirmed in the latest European standard. The consequences in the motorway network due to unsatisfying capability and durability of joint sealing systems are unacceptable. In addition, inadequate traffic performance (noise emissions, roll-over comfort) and traffic safety losses in the joint area of concrete pavements are existing challenges. These deficits and weaknesses reflect a demand for joint sealing materials and constructions whose approval requirements take functional aspects into account. Furthermore a sufficient analysis of decisive loads and a practice-oriented method to evaluate the requirements towards performance and durability is still missed. In this contribution decisive loads to German highways are analyzed. The design of test specimen for representative functional testing of joint sealing systems is discussed. The focus is on the geometry of the test specimens and the used concrete mixture. Finally, a new approach for a functionorientated test concept that considers representative load functions is presented. The potential of this approach to validate the durability and capability of various joint sealing systems is also presented using an example. Keywords Joint sealing · Expansion joint · Performance testing · Capability · Durability
J. Buchheim (B) · Ch. Recknagel · P. Wolter · K. Kittler-Packmor Federal Institute for Materials Research and Testing (BAM), Berlin, Germany e-mail: [email protected] Ch. Recknagel e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_14
111
112
J. Buchheim et al.
1 Introduction Due to the increasing volume of heavy duty traffic [1], the demands on the durability and capability of highways and especially concrete roads are growing in equal measure. The innovation in concrete pavement technology and the associated technical rules and approval guidelines keep pace with this trend [2]. As far as materials and products for joint sealings are concerned, the development is much slower. Furthermore the existing European rules are not taking into account, that the joint sealing is an interacting system. The whole system needs to be seen and rated as a functional group mastering the sealing task, see Fig. 1. Instead, European guidelines contain different tests for each system component. Even for different sealing constructions, different tests to determine physical properties are defined and compared to empirically specified limit values [3]. The application of this method leads to declining and inconsistent product quality of the sealants. The durability and functionality of many joint sealing systems is therefore not guaranteed. As a solution to existing problems of joint sealing systems in concrete pavements, we propose a function-orientated test routine. Towards introducing a new test method, an analyse of the influences and loads on the joint sealing system is presented at first. The influence of test specimen design to mechanical behaviour of the joint sealing follows. As a result of this analysis, the authors present a first approach for performance and durability testing of joint sealing systems. The test procedure according to this approach allows two statements: the evaluation of the overall performance of the joint filling system in order to withstand the effective loads and a statement on durability of the system due to endurance tests.
Fig. 1 Essential components of the joint sealing system
A New Approach Towards Scientific Evaluation …
113
2 Decisive Influences on Joint Sealings Concrete roads are exposed to a wide range of influences. None of these influences occurs in isolation, on the contrary, they occur simultaneously in different combination. This leads to differing loading effects. For performance assessment and valid durability testing the simultaneity and variety of loads has to be considered. An overview of different influences and their loading effects follows.
2.1 Loads Induced by Weather Daily and seasonal variation in temperature leads to slow expansion and contraction of the concrete slabs resulting in mechanical stress and deformation of the joint filling. Simultaneously, UV radiation, rain, and humidity can occur. Superposition of these influences might increase single loadings on the sealing system, as aging effects of UV radiation is promoted by increasing temperature.
2.2 Loads Induced by Traffic Every truck that passes concrete highways, causes dynamic deformation of every joint in three dimensions. The deformation leads to compression, tension and shear in the filling material. According to [4], deformation of the joint sealant, normal to the bonding zone of concrete and joint filling, moves up to the millimeter range. The database to quantify the movement of concrete slabs and joint deformations in 3D is not complete and goes back to earlier pavement constructions [4]. It is conceivable that old and modern pavements differ in their deformation behaviour. The new sensor system “Senso Joint” [5] offers an approach for permanent 3D in situ measurement. Placed in German concrete highways, monitoring campaigns will contribute to close the leak of deformation data. Further traffic-related damage to joint sealings is caused by the abrasion of the joint filling material by wheels, especially in summer. A result is insufficient filling of the joint, which may affect the tightness of the sealing. Traffic is a decisive load on roads and its dynamic effects bear high damage potential—consideration of the various effects is essential for performance testing.
114
J. Buchheim et al.
2.3 Other Loads More over, chemical and other moisture-inducing loads occur. The application of de-icing salt solution in winter season, for instance, causes effects in the concrete, the joint filling and the bonding zone.
3 Design of Test Specimen 3.1 Concrete Mixture
80 60 40 20 0
0%
0%
22%
21%
43%
12% 29%
17%
6% 5% 8% 8% 8% 7%
6% 7% 9% 9%
10%
15%
19%
21%
EAC
TSC
BLC
(a)
19%
8,0 / 22,4 5,0 / 8,0 2,0 / 5,0 1,0 / 2,0 0,5 / 1,0 0,1 / 0,5 water cement
6
water consumtion [kg/m²]
mass proportion
100
grain size of aggregate [mm]
The decisive influence of the concrete mixture towards the joint sealing results from the specific water absorption of the concrete. Its porous structure leads to capillary water transport in the concrete. During rainfall, for example, moisture can accumulate in the contact area of the concrete slabs and the joint filling. In the worst case, this leads to the separation of the adhesive compound. In order to create a bond between the concrete and the filling that is comparable to that in concrete roads, it is important to use a concrete mix that equals the mix of common concretes pavements. Common concrete mixtures for pavement on German heavy-duty roads are mixtures for bottom-layer concrete (BLC) and, in recent years, mainly exposed aggregate concrete (EAC)—both mixtures according to [6]. A mixture used for bond testing with joint filling products with respect to [3] is defined in [7]. Further it is called “test specimen concrete” (TSC). Each composition of the presented concretes is illustrated in Fig. 2a. TSC contains approximately 31% cement stone by mass— 21% cement and 10% water. The proportion of cement stone is much higher than in EAC (28%) and BLC (22%). The different composition leads to a clearly different capillary water transfer. In laboratory tests we determined significant differences in
BLC EAC TSC
5 4 3 2 1 0 0
time [d]
2
4
6
8
(b)
Fig. 2 Characteristics of different concretes: mass proportion of concrete mixtures (a); consumption of water over time (b)
A New Approach Towards Scientific Evaluation …
115
Table 1 Results of FE-simulation comparing relative peak stress level and influence range of transversal contraction depending on the joint length—hot sealing compound in −20 °C use case Joint length (mm)
Relative peak stress σmax_length /σmax_800 mm (%)
Relative length with σ ≥ 0.9 σmax_length (%)
200
95
73
500
100
90
800
100
94
capillary absorption for the three types of concrete (see Fig. 2b). When using TSC for test specimens, unrealistically intensive moisture influences on the adhesive bond are expected. Thus a pavement concrete such as EAC has to be considered for functional testing of sealing systems.
3.2 Representative Dimensions of Specimen To simulate realistic stress conditions inside of joint fillings in laboratory conditions, the joint’s geometry is decisive. The unhindered transversal contraction in the edge areas of the joint has large influence on the stress distribution over the joint length, especially for short joint lengths. This is also what results of a FE-simulation of a joint for varying length, see Table 1. Joint width, height and boundary conditions are kept constant. The bituminous filling is modeled to be a hot sealing compound, assumed as linear-elastic material and extended to a constant width. The currently used joint length of 200 mm for approval testing of sealing products in Germany [3] is unsuitable. To apply maximum stress and its realistic distribution over 90% joint length, a length of at least 500 mm is recommended.
4 An Approach Towards More Realistic Functional Testing A first attempt for a new simplified performance and durability test was realised at BAM in 2016. The aim is to identify the functional long-term behaviour of joint sealing systems. The focus is on the durability und capability of the system towards more realistic loads and load combination, regardless of the chosen filling construction type. All types of sealing systems are exposed to the same test procedure. For this approach, only two approximated external influences are applied simultaneously: • an annual temperature curve and • quasi-static and dynamic mechanical deformation, normal to the bonding zone. The mechanical structure and load functions to simulate one annual cycle are shown in Fig. 3. The first load cycle characterises the performance of the tested
116
J. Buchheim et al.
T [°C]
specimen with F joint filling system
50 15 -20 5
s [mm]
mechanical actuaƟon
primarily induced degradation mechanical fatigue ageing
climaƟc T chamber
s
summer
spring
0,1 -0,1
autumn
winter
f = 1 Hz
0 -1 18
time [h]
82
122
163
Fig. 3 Scheme of the test setup and loads of an annual cycle (thermal and mechanical)
system, the ability to withstand the applied loads. One year of real exposure is experimentally simulated in one week—168 h. To prove the durability of the system, this cycle is repeated ten times for ten years of exposure. The specimen consist of two exposed aggregate concrete slabs connected by a joint filling of 200 mm length. An example for a tested joint filling system with adhesive failure after six load cycles in the test is shown in Fig. 4. The evaluation of the joint filling system after the test with ten cycles is based on the function of the system: impermeability to water, state of the adhesive and cohesive bond, and the change of position of the filling. This approach allows to evaluate the functional behaviour of the entire joint sealing system and the behaviour of the materials used, in particular. The functional test identified commonly used sealing systems (category N2) as unsuitable for use in German motorways. Their fatigue behaviour and ageing resistance did not correspond to the required functionality. As one consequence of this outcome, a new product category of more efficient bituminous sealants was introduced: N2+. 2
cycle 1
cycle 6
F Fmax
2
0 Fmin
s [mm]
4
1
force F [kN]
6
0
-1
-2
-2 0
16
32
48
64
80
96
112 128 t [h] 160
Fig. 4 Force curve for two load cycles and test specimen before the test and after six cycles with failure of the adhesive bond
A New Approach Towards Scientific Evaluation …
117
5 Conclusion and Outlook A new approach for a function test of joint sealing systems in roadways was presented, based on the decisive loads of the system’s application. It was shown that the functional test with permanent superposition of mechanical and thermal loads leads to a more realistic response of the functional system behaviour. The results, regarding tested joint filling systems, correlate with the experiences in road construction practice with these products. Thus, the results and the development of the product category N2+ further validate the presented idea of functional testing. The aim is to learn more about the functional behaviour of joint sealing systems in order to create the basis for new sealing designs and materials. In addition, a better lifetime prognosis and adapted planning of maintenance for joint sealings are aimed to ensure quality and availability of concrete roads. Therefore the significance of the function test has to be increased. The conclusions regarding the size of test specimen has to be taken into account. Furthermore, investigations on the influence of other real loads on the functionality of sealing systems must be conducted. The basis for these investigations will be laid at BAM: a new test stand is being developed, which will contribute to solving the open questions.
References 1. BASt Homepage. https://www.bast.de/BASt_2017/DE/Verkehrstechnik/Fachthemen/v2-verkeh rszaehlung/zaehl_node.html. Accessed 28 Oct 2019 2. FGSV—Road and Transportation Research Association: Guidelines for the Standardization of Pavement Structures of Traffic Areas (RStO 12), Cologne (2012) 3. FGSV—Road and Transportation Research Association: Zusätzliche Technische Vertragsbedingungen und Richtlinien für Fugen in Verkehrsflächen (ZTV Fug-StB 15), Cologne (2015) 4. Hean, S., Partl, M.N.: Long-term behavior of polymer bitumen joint sealants on a trial road section. In: Proceedings of the 3rd Eurasphalt and Eurobitume Congress, pp. 546–559. Vienna (2004) 5. Recknagel, C., Spitzer, S., Hoppe, J., Wenzel, N., Pirskawetz, S.: SENSO JOINT—an innovative sensor system for a sustainable joint design of concrete pavements. In: Proceedings of the 9th Conference on Maintenance and Rehabilitation of Pavements, Zürich (2020) 6. BMVI—Federal Ministry of Transport and Digital Infrastructure: Allgemeines Rundschreiben Straßenbau Nr. 4/2013, Bonn (2013) 7. EN 13880-12: Hot applied joint sealants—Part 12: Test method for the manufacture of concrete test blocks for bond testing
A New Bituminous Ballast-Less and Sleeper-Less Thin Railway Track Structure Diego Ramirez Cardona, Simon Pouget, G. van der Houwen, M. Vlak, S. Sénéchal, and N. Calon
Abstract This paper presents a new structure design for a ballast-less and sleeperless bituminous railway track. The structure is thinner than the conventional French tracks made with granular materials. This makes it especially suited for track renovation works in tunnels and other infrastructure configurations where the respect of the accurate and durable railway track geometry is essential. The proposed structure is a system associating a high-performance bituminous mixture, called GB5® , and the edilon)(sedra Embedded Rail System (ERS) using the Corkelast® resin as rail fastening system. This paper presents details of the components of the proposed structure and of its construction methodology. The system was submitted to mechanical tests to assess its aptitude to serve as rail fastening system in Europe, according to EN 13481 part 5 and EN 13146 part 9 test methods for non-ballasted rail fastening systems. The sample manufacturing and test protocols are described in this paper. The results from two tests are presented and analyzed. The system was found to be resilient and to meet the requirements of the European standard. Further analysis and large scale tests will give more insight into the behavior of the system, especially in terms of permanent deformation of the bituminous mixture and of the supporting soil. Keywords Ballast-less track · Sleeper-less track · Bituminous mixture · Embedded rail system (ERS) · Railway · Track · Tunnel
D. Ramirez Cardona (B) · S. Pouget Eiffage Infrastructures, Direction of Research and Innovation, Corbas, France e-mail: [email protected] G. van der Houwen · M. Vlak · S. Sénéchal edilon)(sedra Direction R&D, Haarlem, The Netherlands N. Calon SNCF Réseau, Direction of Industrial Production, Paris, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_15
119
120
D. Ramirez Cardona et al.
1 Introduction The development of the railway transportation sector in France imposes the rehabilitation of the existing infrastructure, in order to cope with current and future needs such as promoting railway freight, augmenting circulation speeds and increasing the train frequencies. One of the challenges to overcome is the outdatedness of the existing infrastructure. The French railway network comprises a high number of tunnels that date from the late nineteenth century. Since then, these structures have not been updated and many of them present an unsuitable size with respect to modern rolling stock, especially freight trains. To this date, 15% of the registered railway tunnels in France are not operational. The restoration and updating works needed for railway tunnels are very costly and time-consuming, making them incompatible with the infrastructure managers’ budget and with the line exploitation constraints. Another challenge is the enhancement of the mechanical properties of the tracks. Rehabilitation works include increasing the thickness of the ballast and subgrade layers in order to provide more bearing capacity and to allow the installation of modern concrete sleepers. According to the SNCF design guideline IG90260, the minimal track thickness under the sleeper is 70 cm (ballast + subballast) for an UIC 3-4 line on a formation of quality S1 (GTR classification). This leads to a global track height of over 85 cm, which is higher than most existing old tracks. However, in tunnels, increasing the thickness of the track reduces their gauge. Where possible, the tunnel floor is then excavated to compensate for this gauge loss but this implies costly and time consuming earthworks. On most cases, however, excavation works could endanger the stability of the structure making it impossible to compensate the gauge loss. Such tunnels become then assets that need frequent track and platform maintenance works (Fig. 1). There is therefore a fundamental need to develop an innovative track system for tunnels that allows increasing the line capacity while respecting the gauges and the integrity of the existing structures. The system also needs to provide a high security level for train circulations, limit traffic disturbance during construction works, be cost-competitive and present a long service life and low maintenance needs. These
Fig. 1 Tunnel track subgrades in bad state and common rehabilitation works requiring excavation and replacement of the subgrade materials
A New Bituminous Ballast-Less and Sleeper-Less Thin …
121
challenges are addressed in the REVES project, led by the SNCF. This project aims to develop a track system of maximum 36 cm in thickness, achieving a reduction of 50 cm in track height, compared to a regular ballasted track. One identified solution to this challenge is to use the edilon)(sedra Corkelast® ERS (Embedded Rail System) design, which is a high-performance and maintenance-free rail fastening system that was developed as a solution for thin tracks on concrete slabs. The system has worldwide approvals and is in operation for heavy rail application since the beginning of the ‘70 s. The design allows suppressing the ballast and sleepers of the conventional ballasted track design. However, concrete slabs require long construction and curing periods which are incompatible with the constraints of track rehabilitation works. Bituminous mixtures, on the contrary, can be quickly laid and used almost immediately, making it an interesting alternative to concrete for track rehabilitation. The potential of using bituminous mixtures in ERS was assessed by Eiffage in Spain in 2013. In the framework of the Bitutran project, a 20 m long test section of the Alicante (Spain) tram-train system was built with a bituminous ERS track structure. The project aimed to reduce the vibrations generated by train circulations by using a bituminous mixture with enhanced damping properties. The bituminous layer is 45.5 cm thick. After five years of service, the Bitutran structure is in perfect state and no maintenance has been required to rectify the track geometry. This is a good indicator of the capacity of the bituminous ERS to maintain the track gauge and to resist permanent deformation due to rutting. In this paper, a thinner bituminous slab with ERS structure is proposed and the results on the first laboratory tests are presented.
2 The Bituminous GB5® Slab with Corkelast® ERS The solution presented in this paper consists on an ERS where the rail fastening system results from enclosing and bonding the rails with the edilon)(sedra Corkelast® resin, in an a bituminous mixture channel (Fig. 2). The Corkelast® ERS associates well with the bituminous mixture and serves as a continuous support for the rail to evenly transfer the load from the rail to the supporting bituminous slab. The bituminous layer is made with a GB5® mixture [1, 2], developed by Eiffage. This high-performance mixture combines an optimized aggregate packing, which increases its density, with a Biprène® polymer-modified binder, which enhances its elasticity. The GB5® is then a material with high modulus, high resistance to rutting, high fatigue resistance properties and low susceptibility to moisture damage. It is a suitable material for heavy load traffic, without requiring excessive layer thickness. The GB5® is also very easy to lay and to compact using common road construction engines. Low compaction effort is needed for optimal density and wheel compactors can be used, which is practical for operations within tunnels. A high bearing capacity support is however necessary to ensure the thinness of the layer.
122
D. Ramirez Cardona et al.
Fig. 2 Scheme of the bituminous edilon)(sedra Corkelast® ERS using GB5® channel
The construction of the bituminous ERS begins with laying the GB5® in two layers. The layers’ thickness depends on the rail height and on the design results (load, traffic, etc.). For normal conditions, the total GB5® thickness is not greater than 30 cm, with minimum thickness below the channel of 10 cm. The bituminous mixture can be loaded almost immediately after being laid, which contributes to the rapidity of the construction works. The channels are then sawn, followed by the installation of the rails using shims or spacers. Finally, the Corkelast® is poured to fix the rails to the GB5® .
3 Validation as Rail Fastening System for Heavy Rail Application Two real-scale samples were tested following the EN 13481 part 5 and EN 13146 part 9 test protocols for non-ballasted rail fastening systems. Each sample used a different rail profile: 50E6 and 60E1, respectively. The GB5® slabs were fabricated at the Center for Studies and Research of Eiffage in Lyon, France. The samples were assembled and the tests were carried out at the edilon)(sedra laboratories in Haarlem, Netherlands. Table 1 and Fig. 3 describe the two samples. In order to prepare the laboratory samples, the channel vertical walls were cut out from the top GB5® layer, then glued to the bottom GB5® layer using a resin. Table 1 ERS samples description
Rail profile
ERS 50 (cm)
ERS 60 (cm)
Channel depth
14
16
Rail height
15.3
17.2
Top GB5® layer thickness
14
16
Bottom GB5® layer thickness
10
12
Total
GB5®
width
40
Total GB5® length
60
Rail length
80
A New Bituminous Ballast-Less and Sleeper-Less Thin …
123
Fig. 3 Scheme of a sample
Consequently, the bottom of the channel is located at the interface between the top and bottom GB5® layers. The rail and PVC tubes (material saving items) were then placed using shims. Finally, the Corkelast® was poured to fix them to the bituminous channel. Due to the macrotexture of the GB5® mixture, the bottom and sides of the sample were covered with a resin to assure a flat and conformable contact between the sample and the support (see Fig. 4). It is important to note that in a real use case, the bottom of the channel will be located inside the bottom bituminous layer, and that the bonding between the two layers will be assured by a bond coat bituminous emulsion. The EN13481 part 5 and EN 13146 part 9 test protocols and the results of the each sample are presented in Table 2. The test measures the static and dynamic vertical stiffness of the assembly before and after applying a repeated loading that simulates the effect of traffic. The measures are done at two different load application angles (α—angle between the load line and the line normal to the running surface of the rails): 0° and 26°. The samples were tested for fastening categories C and D according to the norm. The repeated loading consisted of 3 million cycles at constant
Fig. 4 Sample mounted in the press for testing at α = 0° and α = 26°, respectively
124
D. Ramirez Cardona et al.
Table 2 Fastening system test results—EN 13481-5 test protocol Test stage - category 1
Static test - CD
α
Vertical stiffness [MN/m] ERS 50 ERS 60
0°
21.2
22.5 28.8
Variation after repeated loading ERS 50 ERS 60 Limit
2
Dynamic test (10Hz) - D
0°
27.0
3
Static test – CD
26°
23.2
24.1
4
Dynamic test (10Hz) - D
26°
32.6
32.6
5
Repeated loading (RL)
33°
6
Static test - CD (after RL)
26°
23.9
24.1
3%
0%
7
Dynamic test (10Hz) - D (after RL)
26°
36.5
36.0
12%
10%
8
Static test - CD (after RL)
0°
22.1
22.3
4%
1%
9
Dynamic test (10Hz) - D (after RL)
0°
30.9
29.1
14%
1%
Go and GE > SF > CE > BE > LS. Despite there are slight differences between the fatigue laws for the same mastic, the ranking of fatigue resistance of mastics was the same for each approach. In general, each filler led to a unique improvement of fatigue performance as compared to the CBE binder, knowing that the used fillers increase significantly the complex shear modulus of CBE as found in previous work [7]. Therefore, the high CBE mastic complex shear modulus does not significantly affect the capability of the mastic to dissipate the applied energy by relaxation, this finding is lined up with Little and Petersen findings [8]. Hence, it can be concluded that the ET, GE and Go blended fillers show substantially improvement in fatigue life. Especially the ET (Ettringite), as it has a complex crystalline structure in CBE, which improves the elastic behaviour of CBE and offers higher plastic flow to arise before crack initiation, results in higher ability to accommodate more damage levels. This is perhaps due to the complex interaction resulted from the physiochemical effect and the way in which ET influences the CBE microstructure. Another explanation could be offered here,
164
A. Al-Mohammedawi and K. Mollenhauer
Fig. 2 Fatigue laws: a G50% , b DPA, c DER, d RDEC, e ESR
the binder-filler affinity might enhance the bonding force at the interface. Therefore, the bond force formed within the filler-bitumen interface could be the reason for having an interactive effect on the fatigue resistance.
4 Conclusions In this study, the fatigue properties of CBE binder and mastics with different fillers were studied by time sweep controlled strain tests under three unique strain levels. The fatigue resistance of the CBE mastics is enhanced significantly by incorporating active fillers. However, the ET filler causes more improvement than that of the rest. The ability of mastic modified with ET to resist higher damage is probably due to the physiochemical interaction between the CBE binder and the co binder (the hydraulic binder), which in turn, interrupts the crack initiation in the mastic. Moreover, the GE, GO and SF mastics have generally similar fatigue lives. While the incorporation of CE and LD have nearly comparable effect on fatigue life of CBE. In addition to that, as inactive filler, LS filler has the lowest fatigue life, this early fatigue failure might be due to the poor interaction between LS particles and CBE binder (low adhesive force at the interface). All used fatigue approaches offer a relatively similar rank for the tested mastics and comparable fatigue life results. However, the traditional
A Study on Fatigue Behaviour …
165
approach defines the fatigue failure in an arbitrary way, thus inaccurate analyses could be resulted. Whereas, RDEC method is considered interesting concept as it focuses on the change in the dissipated energy but the experimental raw data, in some cases, difficult to examine.
References 1. Nikolaides, A.F.: Bituminous mixtures and pavements VI. In: Bituminous Mixtures and Pavements VI—Proceedings of the 6th International Conference on Bituminous Mixtures and Pavements, ICONFBMP (2015) 2. Dai, Q., Sadd, M.H.: Parametric model study of microstructure effects on damage behavior of asphalt samples. Int. J. Pavement Eng. (2004) 3. Pronk, A.C.: Comparison of 2 and 4 point fatigue tests and healing in 4 point dynamic bending test based on the dissipated energy concept. In: 8th International Conference on Asphalt Pavements (1997) 4. Rowe, G.M.: Application of the dissipated energy concept to fatigue cracking in asphalt pavements (1996) 5. Kim, Y.R., Little, D.N., Lytton, R.L.: Fatigue and healing characterization of asphalt mixtures. J. Mater. Civ. Eng. (2003) 6. Mitchell, D., Witczak, M., Mamlouk, M., Kaloush, M., Kaloush, K.: Validation of initial and failure stiffness definitions in flexure fatigue test for hot mix asphalt. J. Test. Eval. (2007) 7. Al-Mohammedawi, A., Mollenhauer, K.: Viscoelastic response of bitumen emulsion mastic with various active fillers. In: 9th International Conference on Maintenance and Rehabilitation of Pavements, MAIREPAV (2020) 8. Little, D.N., Petersen, J.C.: Unique effects of hydrated lime filler on the performance-related properties of asphalt cements: physical and chemical interactions. J. Mater. Civ. Eng. (2005)
A Study on the Properties of Polyurethane-Elastomer-Modified Asphalt Peng Wang, Jian Xu, Yongchun Qin, Erhu Yan, Jie Wang, and Weiying Wang
Abstract A new type of elastomer-modified asphalt, polyurethane (PU)-elastomermodified asphalt, was synthesized with polytetramethylene ether glycol (PTMEG) and diphenylmethane diisocyanate (MDI) as the main agents. The R value (– NCO/–OH) was 1.15. The polyurethane-elastomer-modified asphalt was prepared by adding different amounts of a polyurethane elastomer modifier to the asphalt using a high-speed emulsification shearing machine, and the properties of modified asphalt were characterized. The experimental results indicate that the performance of polyurethane-elastomer-modified asphalt has been significantly improved. Fluorescence microscopy indicated that the modifier could disperse well in asphalt. The dynamic shear rheological test (DSR), the ductility test and the multiple stress creep recovery test (MSCR) were used to analyze the asphalt samples. The results indicate that the high- and low-temperature performance of the asphalt was significantly improved. Keywords Modified asphalt · Polyurethane · Elastomer · Road engineering
1 Introduction Polyurethane (PU) is a general designation of macromolecular compounds containing a repeated carbamate group (–NHCOO–) on the main chain. PU was mainly synthesized by the reaction of isocyanate and polyol. PU materials exhibit diversity due to differences in isocyanates, polyols, additives and processes, which makes the PU materials widely used in various fields of national economy [1]. The PU elastomer exhibits not only excellent durability, strength, wear resistance, low-temperature flexibility but also advantages of both rubber and plastic [2]. PU, as a kind of an asphalt modifier, has been found to improve high-temperature stability and low-temperature flexibility of asphalt and has a great application value P. Wang · J. Xu · Y. Qin (B) · E. Yan · J. Wang · W. Wang Research Institute of Highway Ministry of Transport, Beijing, China e-mail: [email protected] Key Laboratory of Transport Industry of Road Structure and Material, Beijing, China © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_21
167
168 Table 1 Basic properties of 70# matrix asphalt
P. Wang et al. Parameter
Results
Parameter
Results
Penetration (25 °C)/0.1 mm
64
Viscosity (135 °C)/Pa·s
0.327
Softening point/°C
47
Density/(g/cm3 )
1.034
Ductility (5 °C)/cm
0
on road materials. At present, there are few research reports on PU-modified asphalt compared to those on SBS-modified asphalt. Terry et al. [3] disclosed a method for synthesizing PU-modified asphalt by matrix asphalt, hydroxyl-terminated polybutadiene and isocyanate reaction. Cuadri [4] used renewable castor oil instead of polyol to react with isocyanate, and synthetic PU was used for asphalt modification. The results indicate that the high-temperature performance of asphalt was improved. The use of PU prepolymers to improve asphalt performance has been reported in many studies. In this work, a thermoplastic PU elastomer was synthesized as an asphalt modifier. As the PU elastomer improves the elasticity and toughness of the asphalt, the road surface can be highly resistant to plastic deformation caused by vehicles and the environment [5].
2 Experimental 2.1 Materials PTMEG with a hydroxyl value of 54–58 mg KOH/g was supplied by Wanhua Chemical Co., Ltd. MDI with a 29–31% content of (–NCO) was supplied by the same supplier. The matrix asphalt is 70#, and the specific properties are presented in Table 1.
2.2 Preparation of the Polyurethane Elastomer Modifier The PTMEG was vacuum dehydrated under 110 °C and 0.1 MPa for 2–3 h, and then cooled to 60 °C. A certain amount of MDI was added to prepare a prepolymer. The chain extender was added to the prepolymer, and these were fully reacted to prepare a PU elastomer modifier.
A Study on the Properties …
169
2.3 Preparation of Modified Asphalt The matrix asphalt was heated at 165 °C for 1 h, and the PU elastomer modifier was slowly added to the matrix asphalt in batches. The modifier was sheared by a highspeed emulsification shearer for 1 h, and the asphalt was stirred for approximately 4 h to prepare modified asphalt.
2.4 Characterization Basic properties test. The properties of modified asphalt were tested by the “Standard Test Methods of Bitumen and Bituminous Mixtures for Highway Engineering” [6]. Dispersion test. The dispersion of the modified asphalt was observed by Axio Imager Z2m (Zeiss) [7]. Fourier transform infrared spectroscopy (FTIR). FTIR was tested by a Nicoet 560, and the sample was evenly spread on a KBr sheet with a scan range of 4000–500 cm−1 . Rheological properties. The dynamic shear rheometer (DSR), Ultra+ made by Malvern, was used to test the rutting factor (G*/sinδ) of the asphalt [8]. Test of multiple stress creep recovery (MSCR). The MSCR was carried out by a DSR [9]. The creep recovery (R) and the unrecoverable creep compliance (Jnr) were obtained.
3 Result and Discussion 3.1 Basic Properties of Modified Asphalt PU-elastomer-modified asphalt was prepared using 70# matrix asphalt with 4, 6, 8 and 10% modifiers. Table 2 indicates that the high- and low-temperature properties of the asphalt were significantly improved by the elastomer. As the amount of the addition increases, the softening point gradually increases, and the ductility (5 °C) also increases compared with the matrix asphalt. The performance of asphalt was Table 2 Basic properties of modified asphalt Parameter
Matrix asphalt
4%
6%
8%
10%
Softening point/°C
47
53.2
57.6
61.3
62.4
Ductility (5 °C)/cm Penetration (25 °C)/0.1 mm
0 64
18.6
24.2
33.9
34.5
60
58
57
57
170
P. Wang et al.
Fig. 1 Infrared spectrum of modified asphalt
improved by PU elastomer dispersed in asphalt. The PU molecular chain contains soft and hard segments. The hard segment contains a benzene ring and a urethane bond, which can be combined with aromatic phenol and resin (aromatics) in asphalt. The crosslinks may be formed by hydrogen bonding and chemical bonding. The soft segment molecular chain acts as a plasticizer that enhances the flexibility of the asphalt.
3.2 Analysis of Infrared Spectroscopy An infrared spectrum of modified asphalt is illustrated in Fig. 1. The infrared absorption peaks of the modified asphalts are similar. The absorption peak at 3334 cm−1 is a carbamate structure (–NH–CO–) stretching vibration peak. As the amount of addition increased, the characteristic absorption peak intensity increased. The results indicate that the PU elastomer modifier has been dispersed into the asphalt.
3.3 Analysis of Fluorescence Microscopy Figure 2 shows that the PU elastomer modifier can be well dispersed in the asphalt without agglomeration of modifier. As the amount of modifier increased, the fluorescent area of the asphalt gradually increased. When the content was 8%, the elastomer modifier was uniformly dispersed in the asphalt. When the content was 10%,
A Study on the Properties …
171
Matrix asphalt
6%
4%
8%
10%
Fig. 2 Dispersion of the PU elastomer modifier
the modifier could still be well dispersed, but the particle size of the modifier was increased. When 10% of the modifier was added, the increase of the modified asphalt basic properties was slowed down. Thus, the economical optimum content was 8%.
3.4 Analysis of Rheological Properties The rheological properties were tested using a dynamic shear rheometer (DSR). The sample was placed in a parallel plate of a 25 mm diameter, and the gap was 1 mm between the two parallel plates. The temperature sweep was a linear scan with an angular velocity of 10 rad/s. The results are illustrated in Fig. 3.
Fig. 3 G*/sinδ of modified asphalt (unaged and aged)
172
P. Wang et al.
Fig. 4 Jnr of modified asphalt (64 °C)
Asphalt is a viscoelastic material, and viscoelastic properties change with temperature. The asphalt turns to flow dynamics with temperature increase, which increases the viscosity of the asphalt, thus reducing the elasticity, increasing the phase angle and increasing the asphalt fluidity. The larger the phase angle is at the same complex modulus, the more unrecoverable deformation of the asphalt is, and the permanent deformation is more likely to be caused. The larger the asphalt G*/sinδ is, the stronger asphalt’s anti-rutting ability is. The G*/sinδ of unaged asphalt is 2.52 kPa (8%, 82 °C), and the G*/sinδ of aged asphalt (TFOT) is 2.86 kPa (8%, 82 °C). At 8%, the modified asphalt can reach a high-temperature grade of 82 °C with 8% content. The high-temperature range of China’s summer hot area is 50–70 °C; thus, the modified asphalt can meet this season’s temperature requirements. To further study the rheological properties of the modified asphalt, the R and Jnr of the modified asphalt were tested by MSCR. The results are illustrated in Figs. 4 and 5. Figures 4 and 5 indicates that when the content of the modifier increases, Jnr decreases, R increases, and asphalt R0.1 increases from 6.84 to 82.45%. The shear deformation of the modified asphalt is reduced, and the elastic recoverable component is improved. Therefore, the modified asphalt has high resistance to rutting.
4 Conclusion As a kind of an asphalt modifier, PU elastomer can be well dispersed in asphalt through a high-speed emulsification shearing machine. As the content of the modifier increases, the dispersion effect decreases, but there is no significant aggregation. In addition, when the amount of the modifier increases, the asphalt anti-rutting factor
A Study on the Properties …
173
Fig. 5 R of modified asphalt (64 °C)
increases, Jnr decreases, R increases, and modified asphalt’s high-temperature antirutting performance is enhanced.
References 1. Liu, Y.: Polyurethane Resin and its Usage, 2nd edn. Chemical Industry Press, Beijing (2011) 2. Drobny, J.G.: Handbook of Thermoplastic Elastomers, 2nd edn. Elsevier, New York (2015) 3. Terry, C.E., Berard, R.A.: Polyurethane-modified bitumen coating composition. US5981010 (1999) 4. Cuadri, A.A., García-Morales, M., Navarro, F.J.: Processing of bitumens modified by a bio-oilderived polyurethane. Fuel 118, 83–90 (2014) 5. Jin, X., Naisheng, G., You, Z.: Research and Development Trends of Polyurethane Modified Asphalt. Mater. Rev. (2019) 6. JTG E20-2011: Standard test methods of bitumen and bituminous mixtures for highway engineering 7. Xia, L., Cao, D., Zhang, H.: Study on the classical and rheological properties of castor oilpolyurethane pre polymer (C-PU) modified asphalt. Constr. Build. Mater. 112, 949–955 (2016) 8. AASHTO T315-08: Determining the rheological properties of asphalt binder using a dynamic shear rheometer (DSR) 9. AASHTO TP 70-11: Standard method of test for multiple stress creep recovery (MSCR) test of asphalt binder using a dynamic shear rheometer (DSR)
A Wear-Based Description of Mechanisms Underlying the Bitumen Removal Process Minh-Tan Do, Veronique Cerezo, and Christophe Ropert
Abstract The paper presents a study aiming at understanding the wear of a dense asphalt concrete, especially the behavior of its bitumen layer. Tests are conducted in laboratory to simulate mechanical actions induced by traffic. Friction and texture analyses are completed by microscopic observations to identify involved processes. Studying the variation of wear volume with energy dissipated by friction helps to estimate wear rates. Explanations are provided in terms of damage mechanisms occurred in the bitumen layer. Keywords Skid resistance · Bitumen removal · Wear · Dissipated energy
1 Introduction It has been accepted that the skid resistance evolution of asphalt concrete due to traffic involves a removal of the bitumen layer followed by a polishing of the aggregates. Whereas extensive research has been devoted to the polishing stage, mechanisms underlying the bitumen removal are not well understood. There is a need to complete this lack of knowledge in order to predict the duration of the so-called early-age period of an asphalt concrete, during which the tire traction is lower than expected. Wear of road surfaces has been studied through the analysis of friction and texture characteristics. This approach assumes implicitly that wear involves only a smoothing of the surface asperities. In the case of coated systems, like bituminous concretes, wear of the coating can induce processes such as cracking and debris formation. Additional characterizations and physical parameters are then required. In this paper, through the example of a dense asphalt concrete (widely used on French secondary roads), attempts are made to introduce visual observations and energy considerations to improve the understanding of the evolution of the bitumen layer when the material is subjected to wear.
M.-T. Do (B) · V. Cerezo · C. Ropert Université Gustave Eiffel, AME/EASE, Bouguenais, France e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_22
175
176
M.-T. Do et al. Water / abrasives Rubber cone 0 ×
(a)
5 mm
(b)
30 mm
Fig. 1 Wehner/Schulze machine a polishing head; b friction-measuring head
2 Experimental Program 2.1 Simulation of Wear in Laboratory Tests are performed by the Wehner/Schulze machine, which is equipped with two rotary heads. The first head (Fig. 1a) is equipped with three rubber-covered cones rolling on the test surface (core of 225 mm in diameter) at 500 rotations per minute. For a given point of the surface, one rotation represents three passes of the cones. The wear track is a circular band of 60 mm in width (Fig. 1a). To accelerate the wear process, a mix of water and abrasive is projected on the test surface during the rotations. The second head (Fig. 1b) is equipped with three rubber pads sliding on the test surface (contact pressure of 0.2 N/mm2 ) under wet conditions from 100 km/h to complete stop. Value of the recorded friction coefficient at 60 km/h is used for analyses.
2.2 Test Method Friction and texture are measured on the test surface before polishing and at 1; 2; 3; 5; 7; 10; 15; 20; 50; 90; 180 (× 103 ) passes. For the texture, 3D topographical maps are measured at three zones of 5 mm × 5 mm in size. Careful placement of the core makes it possible to follow the evolution of the texture of each zone. A second core is subjected to the same polishing process and aggregates (one per stage) are extracted from the surface before polishing and at 2; 7; 20; 50; 180 (× 103 ) passes. They are embedded in an epoxy matrix (Fig. 2) and their cross-section is analyzed by means of a scanning electron microscope (SEM).
A Wear-Based Description of Mechanisms Underlying …
Extracted aggregate
177
Bitumen / aggregate interface
Epoxy
Polished zone
0,45
2,00
0,40
1,50
Sdq
Friction coefficient
Fig. 2 Specimen subjected to polishing and aggregates extracted for SEM observations
0,35
0,30
1,00
0
50 000
100 000
150 000
200 000
0,50
0
Number of passes 19015
50 000
100 000
150 000
200 000
Number of passes Zone 1
Zone 2
Zone 3
Fig. 3 Evolution of the friction coefficient (left) and the surface root-mean-square slope (right)
3 Results 3.1 Evolutions of Friction and Texture Figure 3 (left) shows an increase of friction up to 5.103 passes then a decrease until the end of the test. The root-mean-square slope Sdq (related to surface asperities’ shape) of the three measurement zones does not evolve in the same direction (Fig. 3b) making it difficult to explain the underlying phenomena.
3.2 Visual Observations Topographical map of zone 2 (Fig. 4 top) shows a progressive loss of roughness due to polishing. In contrast, zone 3 (Fig. 4 bottom) shows that the material removal reveals new asperities and, as a matter of fact, an increase of roughness. These examples explain the evolution of Sdq (Fig. 3) but information about the material removed by wear is still missing.
178
M.-T. Do et al.
Fig. 4 Topographical maps of zone 2 (top) and zone 3 (bottom)
SEM photos (Fig. 5) show that the so-called “bitumen layer” is actually a mix of bitumen and very fine inclusions (probably debris from friction between the aggregates during the mix process). After 7000 passes, the bitumen layer is still present but cracks and decohesion between the bitumen and the inclusions can be seen and some inclusions are exposed (these inclusions would correspond to the “new” asperities in zone 3, Fig. 4). At later polishing stages, the reduction of the layer thickness is visible but the bitumen does not disappear completely. At 180,000 passes, aggregates are polished but some residual bitumen (more porous, due to cracks and detachment of inclusions) can still be seen. The increasing (resp. decreasing) part of the friction evolution with wear has been attributed to the removal of the bitumen layer (resp. the polishing). Visual observations show that this separation is not as clear and the two processes (bitumen removal and polishing) can exist until the end of the test. They provide thus experimental evidences confirming results issued from modeling [1]. They also reveal that the bitumen removal involves cracking and material detachment. This finding confirms preliminary results obtained on another asphalt mix with high bitumen content [2].
A Wear-Based Description of Mechanisms Underlying …
179
Fig. 5 SEM photos of the aggregates at different polishing stages (lower magnification is used for the last stage to show a polished part and some residual bitumen)
3.3 Characterization by an Energy Approach Results from the previous section have shown that the bitumen removal involves complex processes. As it is difficult to isolate an individual process, a global characterization seems to be more suitable. The approach adopted in this work is inspired from that developed by Fouvry [3] based on the study of the variation of wear volume with energy dissipated by friction. As Fouvry’s equations are specific to fretting experiments, the following formulae proposed by Ramalho [4] are used instead: E = F.V t = F.x
(1)
where ΔE is the energy dissipated during the time interval Δt; F is the friction force exerted during Δt; V is the relative speed between the contacting bodies; and Δx is the travelled distance during Δt. Equation (1) is applied to friction forces recorded during polishing tests. For each friction value F i , the travelled distance Δx i is obtained by counting the number of passes Δni elapsed from the previous friction value F i−1 and using the following equation:
M.-T. Do et al. 30
30
25
25
Δ (Sq) (μm)
Δ (Sq) (μm)
180
20 15 10
15
y = 0,012x + 6,8668
10 5
5 0
20
0
200
400
600
800
1 000
0
y = 0,0885x 0
200
Average of 3 zones
400
600
800
1 000
Dissipated energy (kJ)
Dissipated energy (kJ)
0 to 7000 passes Linéaire (0 to 7000 passes)
7000 to end Linéaire (7000 to end)
Fig. 6 Variation of the wear volume with the dissipated energy (left) Mean trend with scatter due to the measurement zones; (right) Regression lines on the two linear parts
xi = (n i /3).2π.r
(2)
where: r is the radius of the circle described by a point representative of the cones (this point is chosen arbitrarily as being at the middle of the cones’ length, that means r = 0.082 m); and Δni is divided by 3 to obtain de number of rotations. Unlike tribological tests like cylinder/cylinder where the wear track is limited to a precise location, determination of the wear volume in Wehner/Schulze tests is a real issue. After different attempts, it was found that the decrease of Sq , i.e., the difference of Sq between the before-polishing state and the considered polishing stage, is the most relevant way. Variation of wear volume with dissipated energy is nonlinear (Fig. 6 left). This trend is similar to that observed in [5] whereas studies, for example in [3, 4], have shown a linear variation. The difference can be attributed to the presence of a coating in the present study and in [5].
3.4 Discussions If one separates “by eye” the variation of energy into two linear parts (Fig. 6 right), it can be said that the wear rate is rapid up to 5–7.103 passes then it decreases and remains constant until the end of the test. According to Kato [5], the roughness of the substrate (on which the coating is laid) has a strong effect on wear of the coating. As the roughness of the aggregates is initially high, there would be a rapid wear of the bitumen (for covered aggregates like zone 3 in Fig. 4) in the beginning of the test together with the polishing of exposed aggregates (like zone 2 in Fig. 4); these combined mechanisms contribute to the slope of the first linear part. As the test continues, the thickness of the bitumen layer is reduced (as assessed by SEM photos in Fig. 5) together with a reduction of materials detached from the aggregates
A Wear-Based Description of Mechanisms Underlying …
181
(because protruded asperities already vanish); as a consequence, the wear volume decreases and so the slope of the second linear part. It can be seen that visual observations (topographical maps and SEM photos) provide a better insight into the structure of the materials and constitute a good complement to texture analysis. On the other hand, considering the complexity of involved processes, a wear characterization based on dissipated energy can reflect in a quantitative way the mechanisms identified from visualizations. The proposed approach is promising and needs to be confirmed by further experiments. Consideration of other factors affecting wear such as the abrasives would be relevant as well.
4 Conclusions Using microscopic observations and energy-based characterization to describe wear of a dense asphalt concrete, it was possible to identify new processes underlying the evolution of its skid resistance under traffic. Wear is a combination of the damage of the bitumen layer and the polishing of exposed aggregates. The wear rate is not constant due to the bitumen acting as a coating. The roughness of the aggregates has a significant effect on the removal of the bitumen. Further experiments should confirm this finding and bring to light other characteristics like the hardness of the bitumen or the cohesion at the bitumen/aggregate interface.
References 1. Do, M.T., Kane, M., Tang, Z., de Larrard, F.: Physical model for the prediction of pavement polishing. Wear 267, 81–85 (2009) 2. Cerezo, V., Ropert, C., Hichri, Y., Do, M.T.: Evolution of the road bitumen/aggregate interface under traffic-induced polishing. Engineering Tribology, First Published (2019). https://doi.org/ 10.1177/1350650119829370 3. Fouvry, S., Kapsa, P.: An energy description of hard coating wear mechanisms. Surf. Coat. Technol. 138, 141–148 (2001) 4. Ramalho, A., Miranda, J.C.: The relationship between wear and dissipated energy in sliding systems. Wear 260, 361–367 (2006) 5. Kato, K.: Wear in relation to friction—a review. Wear 241, 151–157 (2000)
Active Mitigation of Low-Temperature Cracking in Asphalt Pavements Quentin Adam, Gerald Englmair, Eyal Levenberg, and Asmus Skar
Abstract Asphalt pavements in cold regions are often exposed to low-temperature cracking distress. The driving mechanism for this type of damage is usually an isolated event of fast surface cooling in combination with low-temperature levels. Another common characteristic of pavements in cold regions is the need for salting or mechanical clearing operations (or both) to address ice and snow events. An emerging solution to the latter issue is embedded heating systems—comprising of electric heating elements. These are commonly installed to help melt snow or prevent the accumulation of surface ice. This paper investigated an additional potential benefit of such heating systems—the ability to mitigate the development of low-temperature cracks. Thermomechanical calculations were carried out for an idealized pavement system modeled as a linear viscoelastic half-space. First, simulated winter-weather conditions were imposed to generate a surface crack at some point in time for a pavement without heating. Then after, the simulations were repeated—but with an active heating system. For the case considered, it is found that cracking can be potentially mitigated by the heating system if activated approximately half-an-hour before the time at which crack would occur. Keywords Asphalt pavement · Low-temperature cracking · Embedded electric heating · Linear viscoelasticity
1 Introduction Asphalt pavements in cold regions are often characterized by low-temperature cracking [1]. The driving mechanism for this type of damage is usually an isolated event of fast surface cooling in combination with low-temperature levels. These conditions generate thermally-induced tensile stresses, exceeding the material’s tensile strength—resulting in cracks [2]. Another common characteristic of pavements in cold regions is the need for salting or mechanical clearing operations (or Q. Adam · G. Englmair · E. Levenberg (B) · A. Skar Department of Civil Engineering, Technical University of Denmark, Kgs. Lyngby, Denmark e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_23
183
184
Q. Adam et al.
both) to address ice and snow events [3]. An emerging solution for the latter issue is embedded heating systems—comprising of electric heating elements in the form of, e.g., ribbons [4]. These are commonly installed to help melt snow or prevent the accumulation of surface ice. This work addresses the case of an asphalt pavement with an embedded electric heating system close to the road surface. It contains an in silico investigation of the ability to mitigate low-temperature cracking conditions by switching on the system. Specifically, the required time of activation prior to a cracking event is parametrically explored. The investigation considers an idealized case of a viscoelastic half-space (representing a pavement system) hosting a thin heat-generating layer at a fixed depth (representing the heating system). It is assumed that the thermal and the mechanical models are decoupled such that the mechanical responses do not affect the temperature field [5]. Thus, the thermal analysis can be performed separately, and the results can serve as input to obtain mechanical responses. Regarding the thermal analysis, the upper boundary (i.e., the pavement surface) is subjected to synthetic heat flux sources mimicking realistic cold-weather conditions that lead to a thermal crack. Subsequently, the one-dimensional heat equation is considered and solved using the finite difference method. Regarding the mechanical analysis, a three-dimensional thermo-viscoelastic equation is used in combination with the time-temperature superposition principle—assuming thermo-rheological simplicity of the material.
2 Thermomechanical Modeling 2.1 Thermal Formulation The analysis requires solving a one-dimensional thermal problem for a half-space for which the basic governing differential equation is 1 ∂T q˙ ∂2T + = 2 ∂z k α ∂t
(1)
where T = T (z, t) is the temperature (◦ C) at depth z(m) during time instant t(s), q˙ = q(z, ˙ t) is heat generation (W · m−3 ), k is thermal conductivity (W · m−1 · K−1 ), and α is thermal diffusivity (m2 · s−1 ) defined as α = k ·c−1 ·ρ −1 wherein c is specific heat capacity (J · kg−1 · K−1 ), and ρ is density (kg · m−3 ). The solution to Eq. (1) is sought through a finite difference scheme by discretizing the depth coordinate into z (m) increments and the time into t (s) increments. Weather effects during clear-sky conditions are introduced by the imposition of heat flux at the half-space surface, taking the general form
Active Mitigation of Low-Temperature Cracking …
−k
185
∂ T (z, t) = a f S + h c (Tair − T (0, t)) + f D − σ ε S (T (0, t) + 273.15)4 ∂z z=0 Solar
Convection
Longwave radiation
(2) where a = 1 (unitless) is the absorptivity of the surface, f S = f S (t) (W · m−2 ) is the intensity of shortwave solar radiation, h c (W · m−2 · K−1 ) denotes an average convection heat transfer coefficient, Tair = Tair (t) (◦ C) is the instantaneous air temperature just above the half-space boundary, f D = f D (t) (W · m−2 ) represents downwelling longwave radiation from the atmosphere, σ is the Stefan-Boltzmann constant (W · m−2 · K−4 ), and ε S (unitless) is the surface emittance in the case of asphalt concrete. The heat generation term q˙ in Eq. (1) is zero except for the depth z = z R where heating ribbons are embedded. The heat production per unit length of an individual ribbon is w R (t)(W · m−1 ), and the deployed center-to-center spacing is denoted by R (m). Consequently, the heat generation term q˙ in Eq. (1) at depth z R for a sublayer with thickness z is q(z ˙ R , t) =
w R (t) . R z
(3)
2.2 Simulated Weather Conditions The thermal analysis is based on synthetically generated winter-weather conditions that simulate the evolution of the different effects in Eq. (2). The following formula is applied for reproducing shortwave solar radiation assuming cloudless sky conditions fS =
N
Pnswr
n=0
H (t − 86,400n) π(t − 86,400n) sin tSS −H (t − tSS − 86,400n)
(4)
where n is an integer day counter with N denoting the total number of winter days being considered in the analysis, Pnswr is the peak intensity of shortwave radiation (W·m−2 ) during the nth winter day, tSS (s) denotes the average duration of sunshine in a typical day, i.e., the time from sunrise to sunset, and H (·) is the Heaviside step function. The convection heat transfer coefficient is calculated according to ASHREA [6] with a wind speed of 10 km/h and a characteristic length of 10 m. The air temperature is assumed to linearly increase during sunshine and linearly decrease during nighttime, in a repetitive manner, given by the following formula Tair =
base Tair
−
amp Tair
amp t 2Tair t − 86,400 · + tSS 86,400
186
Q. Adam et al.
amp 172,800Tair t t − tSS − 86,400 · − 2 86,400 86,400tSS − tSS
(5)
base ◦ where Tair ( C) and Tair (◦ C) are, respectively, the base air temperature level and the amplitude at which the air temperature varies w.r.t. the base temperature. The angular brackets · denote Macaulay brackets, and the square brackets with missing top bars · denote a floor function. amp
2.3 Mechanical Formulation The mechanical problem considers an infinitesimal linear viscoelastic material element that resides at the half-space surface where z = 0 m. Initially, the half-space is considered completely stress-free at a temperature level Tr . When the temperature drops below Tr , stresses are induced due to the elements’ inability to deform in the x − y plane. Tr represents the temperature level below which such stresses can never recover in full even if infinite time is allowed; it is assumed herein that Tr = 0 ◦ C. The governing integral equation representing the thermal-induced stress σT in the x − y plane is αL σT (t) = 1−ν
τ =t
Er (ξ − ξ )dT (τ )
(6)
τ =0
where α L is the coefficient of linear thermal expansion (◦ C−1 ), ν is the Poisson’s ratio (unitless), τ is a time-like integration variable (s), Er (t) is the one-dimensional relaxation modulus (MPa) for temperature Tr , T = T (0, t) − Tr , and t =t ξ = ξ(t) = t =0
dt aT
ξ = ξ (τ ) =
τ =τ
τ =0
dτ aT
(7)
where aT = aT (T ) is the time-temperature shift factor (unitless), assumed herein to follow the WLF equation with c1 (unitless) and c2 (◦ C) as parameters. Finally, the relaxation function is taken as a four-parameter sigmoid on a double-log scale [7] E ∞ 1 + (ξ/τ D )n D Er (ξ ) = (ξ/τ D )n D + E ∞ /E 0
(8)
where E ∞ is the long-term equilibrium modulus (MPa), τ D is a relaxation time (s), n D is a shape parameter (unitless), and E 0 is the instantaneous modulus (MPa). The solution of Eq. (6) was carried out numerically utilizing the trapezoidal rule.
Active Mitigation of Low-Temperature Cracking …
187
Table 1 Parameter values for thermal and mechanical calculations Thermal formulation
Mechanical formulation αL
2 × 10−5
(◦ C−1 )
(W · m−2 )
ν
0.35
(–)
−10
(◦ C)
c1
30
(–)
4
(◦ C)
c2
300
(◦ C) (MPa)
z
0.01
(m)
Pnswr
300
(W · m−2 )
t
20
(s)
fD
200
base Tair amp Tair
ze
1.5
(m)
Te
10
(◦ C)
k
1
(W · m−1
zR
0.1
(m)
E∞
300
ρ
2200
(kg · m−3 )
R
0.2
(m)
τD
106
(s)
c
1000
(J · kg−1 · K−1 )
tSS
2 × 104
(s)
nD
0.35
(–)
E0
30,000
(MPa)
· K−1 )
2.4 Modeling of an Idealized Case Calculations were performed according to the above models with the parameter values for the thermal and mechanical formulation shown in Table 1.Initially, the thermal model was applied to calculate the evolution of the surface temperature T (0, t) of the half-space. Then, the latter was used to compute the evolution of the corresponding thermal stress σT (t). The simulation duration spanned ten days (N = 9) from t = 0 s to t = 864,000 s. For the last two simulation days, the air temperature (see Eq. (5)) was linearly reduced by an hourly rate of 0.18 °C to generate a critical cold condition at the end of the 10th day. A tensile strength of 2.5 MPa is assumed for the half-space material.
3 Results and Discussion Figure 1 shows the development of T (0, t) and σT (t) over the four last days— considering no heat generation. For the eight first days, due to periodicity of the simulated weather conditions, T (0, t) (and σT (t)) evolved in a periodical manner. At the end of the 10th day, the surface temperature T (0, t) dropped to approximately −18 °C with a cooling rate of 0.4 °C/h. With the assumed tensile strength of the material, a crack would have occurred at tcrack = 862,790 s (i.e., end of the 10th day, nighttime). Next, the required time of heat activation before the cracking th was evaluated for three different values: 600, 1800, and 3600 s. After activation, the heating intensity was constant: w R = 40 W/m. Figure 2 shows the resulting development of σT (t) for the last 5500 s of the simulation. For th = 600 s the thermal stress would still have exceeded the materials’ tensile strength, and hence a crack would have occurred. However, th = 1800 s (and therefore also th = 3600 s) was sufficient to mitigate the low-temperature cracking.
188
Q. Adam et al. 0.5
σ T (t)
−2
0
Temperature [°C]
−4 −0.5
−6
−1
−8 −10
crack
−12 −14 −16 −18 500000
Day 7 550000
−2
tensile strength
T(0,t)
Day 8 600000
650000
−1.5
Day 9 700000
Day 10
750000
800000
850000
Thermal stress [MPa]
0
−2.5 −3 900000
Time [s]
Fig. 1 Development of surface temperature and induced thermal stress
Thermal stress [MPa]
−2.2
Δ th = 3600 s
−2.3
−2.4
Δ th = 1800 s
Tensile strength −2.5
crack Δ th = 600 s −2.6
859000
860000
861000
862000
863000
864000
Time [s]
Fig. 2 Thermal stress development with heat activation before the cracking event
4 Conclusion The synthetic thermomechanical analysis of the considered idealized case shows that it is possible to mitigate low-temperature cracking. It was found that the system should be activated about half-an-hour before the time at which the cracking would occur. Thus, parametric studies of several types of pavements and weather conditions will allow for guiding engineering practice towards embedded heating systems. This work is the first step towards a necessary evaluation of the thermomechanical effects of embedded electrical heating systems in asphalt pavements. Further calculations considering real weather and pavement structure data are required.
Active Mitigation of Low-Temperature Cracking …
189
Acknowledgements This work was carried out as part of the Snowless project. Snowless has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 831120.
References 1. Teltayev, B., Radovskiy, B.: Low temperature cracking problem for asphalt pavements in Kazakhstan. In: 8th RILEM International Conference on Mechanisms of Cracking and Debonding in Pavements, RILEM Bookseries 13 (2016) 2. Dave, E.V., Buttlar, W.G.: Thermal reflective cracking of asphalt concrete overlays. Int. J. Pavement Eng. 11(6), 477–488 (2010) 3. Fay, L., Akin, M., Shi, X., Veneziano, D.: Winter operations and salt, sand and chemical management, national cooperative highway research program (NCHRP), Final-Report No. 20-07318 (2013) 4. Sangiorgi, C., Settimi, C., Tataranni, P., Lantieri, C., Adomako, S.: Thermal analysis of asphalt concrete pavements heated with amorphous metal technology. Adv. Mater. Sci. Eng. 2018, 1–8 (2018) 5. Markert, B.: Weak or strong: on coupled problems in continuum mechanics. Report No.: II-20. Universität Stuttgart, Stuttgart/Germany (2010) 6. ASHRAE Handbook, Chapter 51: Snow melting and freeze protection, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. American Society of Heating, Atlanta/GA (2015) 7. Forough, S.A., Nejad, F.M., Khodaii, A.: An investigation of different fitting functions to accurately model the compressive relaxation modulus master curve of asphalt mixes. Road Mater. Pavement Des. 16(4), 767–783 (2015)
Ageing Behavior of Porous and Dense Asphalt Mixtures in the Field Ruxin Jing, Aikaterini Varveri, Xueyan Liu, Athanasios Scarpas, and Sandra Erkens
Abstract Bitumen ageing is one of the principal factors causing the deterioration of asphalt pavements. As bitumen ages, the pavement loses its ability to relax stresses during loading/unloading and thermal cooling process, thus the risk of cracking increases. Oxidation and ultraviolet (UV) radiation are believed to be the main factors that can cause bitumen ageing during pavement service life. The aim of this study is to evaluate the mechanical behavior of porous and dense asphalt pavements during field ageing. Pavement test sections were constructed in 2014 and are being exposed to actual environmental conditions since then. To investigate the effect of UV radiation on ageing, UV reflective glass-plates were utilized to cover part of the pavement surface. To study the evolution of the pavements’ mechanical properties, asphalt cores were collected from the test sections periodically (at one-year intervals). The changes in the stiffness modulus of the mixtures were determined via cyclic indirect tensile tests. The results show that the effect of mineral aggregate packing, and hence of air-void distribution and connectivity, on the ageing sensitivity (both thermal and UV ageing sensitivity) of the pavements with time was found to be significant, as the changes of the stiffness of the porous mixtures were greater than that of dense mixtures. Keywords Field ageing · Porous mixture · Dense mixture · Stiffness · UV radiation
R. Jing (B) · A. Varveri · X. Liu · A. Scarpas · S. Erkens Delft University of Technology, Delft, The Netherlands e-mail: [email protected] A. Scarpas Khalifa University, Abu Dhabi, United Arab Emirates © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_24
191
192
R. Jing et al.
1 Introduction Ageing of asphalt mixtures occurs during production and construction and continues throughout the service life of the pavement. Because of ageing, asphalt becomes stiffer and loses its ability to relax stresses under repeated traffic loading and during the cooling process, thus the risk of fatigue and cracking increases [1]. Depending on their mechanisms, different types of ageing can be distinguished such as thermal ageing and UV ageing [2]. Thermal ageing is considered to be the primary asphalt ageing mechanism and is highly dependent on the climate conditions, i.e. temperature. On the other hand, UV ageing only takes place within first micrometers of the top pavement layer; however it should not be ignored because of its relation with raveling, which is a surface pavement distress [3]. Past studies show a clear relationship between the air voids content and the ageing sensitivity of asphalt mixtures [4]. Porous asphalt shows a higher degree of ageing than dense asphalt [5]. Although some investigations have been carried out, field data for different types of asphalt mixtures under diverse climatic conditions are still needed to understand better field ageing and to further develop a proper laboratory ageing protocol. The goal of this study is to evaluate the ageing behavior of porous and dense asphalt mixture in the field. The main objectives of this study are to (i) investigate the effect of field ageing on asphalt stiffness on the basis of Cyclic Indirect Tensile Test, (ii) correlate the changes of the stiffness with climate temperature data, (iii) determine the effect of the UV radiation on the resulting of asphalt stiffness.
2 Test Section and Experimental Method 2.1 Asphalt Mixture Design One Porous Asphalt (PA) and one Stone Mastic Asphalt (SMA) mixtures were designed using the aggregate gradations shown in Fig. 1. For both mixtures, the same type of aggregate, i.e. Norwegian sandstone, with a nominal maximum size of 16 mm was used. Norwegian sandstone is a type of crushed stone with density 2740 kg/m3 . The target air void content was 16% and 5% for the PA and SMA mixtures, respectively. The same type of PEN 70/100 bitumen was used for both mixtures. The binder content was 5.0% and 6.4% for the PA and the SMA mixtures, respectively. A factory filler which contents limestone and 25% of calcium hydroxide, i.e. Wigro 60 K filler, with density 2780 kg/m3 was used for both mixtures.
Ageing Behavior of Porous and Dense Asphalt Mixtures …
193
(b)
(a) 100
100
Binder content 6.4%
Percentage passing (%)
Percentage passing (%)
Binder content 5.0% 80 60 40 20 0
22.4 11.2 5.6
2
0.5
0.18
0.063
Grain size (mm)
80 60 40 20 0
22.4 11.2 5.6
2
0.5
0.18
0.063
Grain size (mm)
Fig. 1 Aggregate gradation of asphalt mixtures a PA mixture b SMA mixture
2.2 Overview of the Test Section The construction phase of the test sections started with the removal of the existing old pavement surface. After the milling process, a bitumen emulsion tack coat layer was sprayed on the surface. Then the new stone asphalt concrete (STAC) layer of 6 cm thickness was laid first, and the 5 cm thickness top porous (PA) and dense (SMA) asphalt layers were placed on the left and right lanes separately, as shown in Fig. 2 (left). STAC is commonly used as base layer of pavements in the Netherlands. The layers were compacted using a roller compactor. The construction of the test sections was done in October 2014. The profile of the new pavement structures is shown in Fig. 2 (right), in which the old STAC (17 cm thickness) and cement bound asphalt granulate base layers (AGRAC, 25 cm thickness) exist more than 10 years.
PA/SMA (5 cm) New STAC (6 cm) Old STAC (17cm)
AGRAC (25 cm)
Sand Subgrade (500 cm)
Fig. 2 Profile of test section
194
R. Jing et al.
Fig. 3 Schematic representing the use of UV reflective glass-plate
2.3 UV Reflective Glass-Plate UV reflective glass-plates were utilized to cover part of the pavement surface. The glass plates were capable of reflecting most of the UV radiation but they allowed thermal waves to penetrate and heat the pavement. This capability of the glass plates was verified by the UV and temperature measurement results. One measurement is given in the table in Fig. 3. The values in the table represent the temperature and UV intensity on the pavements without (Pavement) and with (Pavement_G) glass plates covered. The glass plates have been lifted up by pegs to allow for ageing of the pavement due to other factors such as water and air (oxygen), Fig. 3. The glass plates were installed on March 2016.
2.4 Cyclic Indirect Tensile Test Samples (with and without the UV effect) with a diameter of 100 mm and a thickness of 50 mm were cored from the PA and SMA layers on an annual basis since 2014. The dynamic modulus of the cores were determined by means of Cyclic Indirect Tensile Test (IT-CY) according to NEN-EN 12697-26. The tests were performed using the Universal Testing Machine (UTM) at five frequencies i.e. 0.5, 1, 2, 5 and 10 Hz and four testing temperature i.e. 0, 10, 20 and 30 °C. The conditioning time before testing was set to be 4 h to equilibrate sample’s temperature and three replicates were tested.
3 Results and Discussion 3.1 Field Ageing Behavior According to the Time-Temperature Superposition principle, the master curve of the dynamic modulus was generated at a reference temperature of 20 °C. Figure 4 shows the evolution of stiffness for the porous (PA) and dense (SMA) mixtures from 2014 to 2018. Unfortunately, no cores were taken from the SMA section right after laying.
Ageing Behavior of Porous and Dense Asphalt Mixtures …
(a)
1011
10 10
10 9
2014 2015 2016 2017 2018
10 8 10 -3
10 -1
10 1
10 3
Dynamic modulus (Pa)
Dynamic modulus (Pa)
10 11
195
(b)
1010
109 2015 2016 2017 2018
108 10 -3
10 5
10 -1
Frequency (Hz)
10 1
10 3
10 5
Frequency (Hz)
Fig. 4 Master curve of asphalt mixtures at reference temperature 20 °C a PA mixture b SMA mixture
Overall, the dynamic modulus of both PA and SMA mixtures increases with time. The rate of modulus change is higher for the PA than for SMA mixture. This is mainly because of the high void content of PA, which leads to an inherently high sensitivity to oxidative ageing. More interestingly though, it can be observed that the change in the dynamic modulus for both mixtures is significantly higher for two periods, namely from 2015 to 2016 and from 2017 to 2018. The weather data that were collected from a weather station close to the test sections show the fluctuations in the yearly mean temperature and the number of days with high temperatures (more than 20 °C) over the years from 2015 to 2018. Figure 5a shows that the yearly mean temperature in 2016 and 2018 was about 1 °C higher than in 2015 and 2017. Also Fig. 5b demonstrates that daily mean temperature exceeded 20 °C for 23 days in 2016 and 31 days in 2018, while 20 and 19 days were recorded in 2015 and 2017. In addition, 2016 and 2018 were in general warmer than 2015 and 2017 as shown by the number of days that the daily mean temperature was above 10 °C.
(b) 300 10.83
11.75
10.82
11.52
10
5
0
2015
2016
2017
2018
High temperature days
Yearly mean temperature (°C)
(a) 15
T > 20 °C T > 10 °C
217
200
200
215 186
100 20
0
2015
23
2016
19
2017
31
2018
Fig. 5 Temperature information from 2015 to 2018 a Yearly mean temperature, b high temperature days
196
R. Jing et al.
3.2 UV Ageing Behavior The effect of UV radiation on ageing is demonstrated through the ITT results plotted in Fig. 6 for the PA and SMA mixtures with and without the UV reflective plates. The UV reflecting plates were installed on 2016, therefore by 2018 the test sections were covered by glass only for a period of two and half years. In Fig. 6, the PA/SMA labels denote the samples with UV effect (not covered by glass), while the PAG/SMAG labels denote the sample from the section covered by glass (without UV effect). It is observed that UV radiation has a greater effect on PA than SMA mixtures. The difference in modulus for the PA and PAG samples increases with time. In contrast, there is no significant changes in stiffness for the SMA samples with and without the
(a)
(b)
10
PA PAG
10
10 9
10-3
10-1
101
103
2016
Dynamic modulus (Pa)
Dynamic modulus (Pa)
2016
SMA SMAG
10 10
10 9
105
10-3
10-1
Frequency (Hz)
(c)
10 9
10-1
101
103
2017
Dynamic modulus (Pa)
Dynamic modulus (Pa)
PA PAG
10 10
10 10
10 9
105
10-3
10-1
101
103
105
Frequency (Hz)
(f)
(e) PA PAG
10 10
10 9
10 -1
10 1
Frequency (Hz)
10 3
10 5
2018
Dynamic modulus (Pa)
2018
Dynamic modulus (Pa)
105
SMA SMAG
Frequency (Hz)
10 -3
103
(d)
2017
10-3
101
Frequency (Hz)
SMA SMAG
10 10
10 9
10 -3
10 -1
10 1
Frequency (Hz)
Fig. 6 Effect of UV radiation on stiffness master curve of asphalt mixtures
10 3
10 5
Ageing Behavior of Porous and Dense Asphalt Mixtures …
197
UV radiation effect. Generally, it has been found that UV ageing takes place within the top 5 µm of the exposed binder film [3]. Therefore, it is anticipated that the high void content of the PA mixture, will allow for larger surface areas of asphalt binder to be exposed to UV radiation. In addition, the high interconnected air voids in PA may allow for UV radiation to penetrate deeper in the PA layer, which is not the case for the SMA layer as the mixture is a quite dense. Because the data represent a short time period, the effect of UV radiation on dense asphalt (SMA mixture) may be more significant in the next years.
4 Conclusions Identifying the ageing behavior of asphalt mixtures in the field is important on the development of artificial laboratory ageing methods for mixture ageing that allow for reliable predictions of long-term mixture performance. For this reason, asphalt pavement sections with one open and one dense mixtures were constructed in 2014 and continuously exposed to environmental conditions. A series of stiffness tests were conducted on asphalt cores, which were yearly taken from the pavement sections. The results show that field ageing have more influence on porous mixture than dense mixture, because of the high void content of the porous mixture, which leads larger binder area directly exposed to the air and temperature variations. High temperature and the number of warm days play an important role on the evolution of field ageing. UV radiation has a substantial effect on the field ageing sensitivity of porous asphalt, but not on dense asphalt. This is probably because of the high interconnected air voids network in porous mixtures, which allows UV radiation to easily penetrate into the asphalt layer. As a continuation of this research, asphalt cores (both porous and dense asphalt) will be taken from the pavement sections on an yearly basis to monitor the changes in their stiffness. In the meantime, an experimental testing program is currently undertaken on asphalt mixture and (extracted) bitumen with the aim to develop a proper ageing protocol to simulate long-term ageing of porous and dense asphalt mixtures. Acknowledgements The authors gratefully acknowledge the Dutch Ministry of Infrastructure and Water Management (Rijkswaterstraat) for funding this project.
References 1. Isacsson, U., Zeng, H.: Relationships between bitumen chemistry and low temperature behavior of asphalt. Constr. Build. Mater. 11(2), 83–91 (1997) 2. Petersen, J., Glaser, R.: Asphalt oxidation mechanisms and the role of oxidation products on age hardening revisited. Road Mater. Pavement Des. 12(4), 795–819 (2011)
198
R. Jing et al.
3. Durrieu, F., Farcas, F., Mouillet, V.: The influence of UV ageing of a SBS modified bitumen: comparison between laboratory and on site ageing. Fuel 86(10), 1446–1451 (2007) 4. Lu, X., Redelius, P., Soenen, H., Thau, M.: Material characteristics of long lasting asphalt pavement. Road Mater. Pavement Des. 12(3), 567–585 (2011) 5. Jemere, Y. Development of a laboratory ageing method for bitumen in porous asphalt. Constr. Build. Mater. 204–210 (2010)
An Attempt to Characterize the RAP Material for Hot Recycled Mix Design Purposes Gurunath Guduru and Kranthi Kuna
Abstract The design of bituminous mixes with high Reclaimed Asphalt Pavement (RAP) content requires the determination of aggregate gradation and the properties of recovered bitumen from RAP. However, in addition to requiring sophisticated equipment, it has been stated by many researchers and practitioners that the current recovery methods have certain other limitations. Hence, the present study focuses on the simple and practical tests from which the properties of the RAP material can be reasonably estimated. The fragmentation test protocol, which was recommended by the RILEM TC264 committee, was evaluated using six sources of RAP of varying properties. Based on test results, it was observed that the fragmentation value determined at 5 °C in combination with the black aggregate curve can be used to predict the percentage fines (46
64-xx
B6
SBS
50–70
>60
76-xx
*In Brazil, the low temperature limit is not considered due to climate characteristics throughout the country
Table 3. Mixture design specifications Mixture ID
M1
M2
M3
M4
M5
M6
M7
M8
Type of binder
B1
B2
B3
B4
B5
B4
B4
B6
Binder content (%)
4.20
5.00
4.80
5.30
4.00
4.20
5.72
5.88
Bulk density (g/cm3 )
2.467
2.586
2.423
2.408
2.490
2.458
2.342
2.352
Air voids (%)
4.1
4.0
5.0
4.2
3.6
4.0
4.0
4.0
NMAS (mm)
19.0
12.5
12.5
12.5
12.5
19.0
19.0
19.0
3 Standard MSCR test Figures 1a, b show Jnr,3200 and the recovery percentage for all 6 binders using the MSCR test protocol from standards M322 [8] and T350-14 [9] at 5 different temperatures. All binders were subjected to RTFOT (Rolling Thin Film Oven Test). Parameter Jnr is presented on the high temperature of binder PG (Performance Grade).
Jnr,3200
5 Stan. 0 B1
64°C
76°C
PG temperature
B2
B3
B4
Hea. V. he. Ext. B5
B6
Recovery (%) (3200Pa)
(b) 100
(a) 10
50 0 58°C 64°C 70°C 76°C 82°C -50 B1
B2
Temperature B3
Fig. 1. Standard MSCR: a Jnr,3200 and traffic volume indications and b R(%)
B4
B5
B6
242
A. Vale et al.
Flow Number
800 600 400
Heavy traffic volume
200 Medium 0 Low M1
M2
M3
Mixtures M4 M5
M6
M7
M8
Fig. 2. Flow number values for the 8 mixtures
Test results indicate that binders B2 and B6 have higher resistance to permanent deformation. Although B2 presents lower values of Jnr,3200 , B6 presents higher percentage of recovery, indicating that B6 has a better mechanical behavior with respect to permanent deformation. The other binders have similar behavior towards the recovery percentage at different temperatures.
4 Uniaxial Repeated Load test For FN values, the test data were treated with the Francken model [13, 14]. The vertical stress was of 204 kPa at 60 °C [11]. Figure 2 shows the FN value for each mixture along with traffic volumes within Brazil [11]. Mixture M8 showed higher resistance to rutting. Although MSCR results of Jnr,3200 for B2 indicated that M2 should have similar behavior to M8, it presented proper FN values for pavements with medium traffic volume, as well as mixtures M1, M4, M5, M6 and M7, while M3 presented the lowest FN (84). Such results suggest that for permanent deformation, the binder recoverable percentage is a better indicator than Jnr,3200 . It is possible to identify that the addition of fly ash improved FN, from 84 (M3) to 188 (M4). Therefore, using such thermoelectric power plant residue may not only serve as waste destination, but also as an improvement for mixture mechanical behavior.
5 The New Approach for MSCR Test The standard MSCR test protocol is divided in 3 steps. The conditioning step applies 10 cycles of 100 Pa load for 1 s, and rest periods of 9 s. The current test applies 20 cycles, the first 10 with 100 Pa, and then other 10, with 3200 Pa during 1 s and 9 s of rest period [9]. The new test protocol [2] applies 30 cycles with 3200 Pa of load during 1 s, with rest period of 9 s. Since this paper focuses on plastic deformation, both test methodologies were performed at high temperatures (58, 64, 70, 76 and 82 °C). Figures 3a–d show graphics with Jnr (calculated from the 10 initial cycles), %R
Analysis of a New Approach for Characterizing …
0
64°C
76°C
PG Temperatures
B1
B3
B4
B5
B6
0 58°C 64°C 70°C 76°C 82°C -50 B1
B2
Temperature B3
B4
B5
B6
(d) 100
2 1 0 -1
50
50 0 58°C 64°C 70°C 76°C 82°C
15-30
Non-Recoverable Strain Rate, Δεnr (%/cycle)
B2
Average Recovery Strain, RS (%/cycle)
Jnr
5
Percent of Recovery, %R (%)
(b) 100
(a) 10
(c)
243
58°C 64°C 70°C 76°C 82°C B1
Temperature
B2
B3
B4
B5
B6
-50 B1
B2
Temperature B3
B4
B5
B6
Fig. 3. Test results: a Jnr, b %R, c εnr and d RC15–30
(total percent of recovery), εnr (Eq. 1), and RC15–30 (obtained from the average of percent of recovery from cycles 15 through 30). εnr =
εnr30 − εnr15 30 − 15
(1)
Through the analysis of %R and RC15–30 , binders B2 and B6 presented higher %R. Jnr and %R from the standard MSCR test protocol showed similar values for the new protocol suggested. However, the new approach has other parameters to be analyzed. Figures 4a–c show parameters RC (Recovery Capacity), F-index (Flow index) and T-index (Temperature index), respectively. The deformation recovery capacity is considered higher for greater values of RC. While higher values of F-index demonstrate less capacity for absorbing energy without deforming, regardless of the binders recovery capacity. For the T-index, the temperatures analyzed were 58 and 82 °C, indicating greater susceptibility to change behavior with temperature changes.
244
A. Vale et al.
2
RC
(a)
0
58°C
64°C
B1
B2
70°C
Temperature
B3
B4
(b)
60
T-index
F-index
B5
82°C B6
(c)
80 40 20 0
76°C
58°C 64°C 70°C 76°C 82°C B1
B2
Temperature B3
B4
B5
B6
140 120 100 80 60 40 20 0 -20
B1
B2
Mixtures B3 B4 B5
B6
Fig. 4. Test results: a RC, b F-index and c T-index
6 Conclusions The parameters from the new MSCR approach show similarities to the standard protocol, which means it may represent binder behavior. Jnr for binders B4 and B6 suggests better behavior for B6, which was confirmed by the FN values for mixtures M7 and M8. When it comes to binder B2, MSCR results pointed to a mixture with high resistance to rutting. However, FN for mixture M2 corresponds only to medium traffic volume, requiring further investigation of the influence of the aggregate choice for that specific mixture. For the new MSCR approach, F-index and T-index indicate that binders with better performance towards permanent deformation are usually susceptible to temperature variations, absorbs more energy without deforming, and have a higher recoverable deformation. In short, parameters proposed for the new MSCR approach are more detailed than the standard parameters, even though, for binder selection to rutting resistance, the standard Jnr and %R are still efficient.
References 1. Domingos, M.D.I., Faxina, A.L.: Susceptibility of asphalt binders to rutting: literature review. J. Mater. Civil Eng. 28(2) (2016) 2. Moreno-Navarro, F., Tauste, R., Sol-Sánchez, M., Rubio-Gámez, M.C.: New approach for characterizing the performance of asphalt binders through the multiple stress creep and recovery test. Road Mater. Pavement Des. (2019) 3. Ferreira, J.L.S., Bastos, J.B.S., Soares, J.B.: Métodos de seleção granulométrica com foco na resistência à deformação permanente. Revista Transportes 24(2), 46–52 (2016) (in Portuguese)
Analysis of a New Approach for Characterizing …
245
4. Kim, S.: Identification and assessment of the dominant aggregate size range (DASR) of asphalt mixture. Dissertation. Florida University, Gainesville, Florida (2006) 5. ASTM D7643-16: American Society for Testing and Materials. Standard Specification for Performance Graded Asphalt Binder (2016) 6. Bukowski, J., Youtcheff, J., Harman, T.: The Multiple Stress Creep Recovery (MSCR) procedure. Federal Highway Administration (2011) 7. D’angelo, J., Kluttz, R., Dongré, R., Stephens, K., Zanzotto, L.: Revision of the superpave high temperature binder specification: the multiple stress creep recovery test. J. Assoc. Asphalt Paving Technol. 76, 123–162 (2007) 8. AASHTO M 332: American Association of State Highway and Transportation Officials. Standard specification for Performance-Graded Asphalt Binder Using Multiple Stress Creep Recovery (MSCR) Test. Washington (2018a) 9. AASHTO T 350: American Association of State Highway and Transportation Officials. Multiple Stress Creep Recovery (MSCR) Test of Asphalt Binder Using a Dynamic Shear Rheometer (DSR). Washington (2018b) 10. Witczak, M.W., Kaloush, K., Pellinen, T., El-Basyouny, M., Von Quintus, H.: NCHRP Report 465: Simple Performance Test for Superpave Mix Design. Transportation Research Board, Washington, DC (2002) 11. Bastos, J.B.S., Silva, S.A.T., Soares, J.B., Nascimento, L.A.H., Kim, Y.R.: Triaxial stress sweep test protocol considerations for permanent deformation characterization of asphalt mixtures. Road Mater. Pavement Des. 19(2), 431–444 (2016) 12. Stempihar, J., Gundla, A., Underwood, B.S.: Interpreting stress sensitivity in the multiple stress creep and recovery test. J. Mater. Civil Eng. 30(2) (2017) 13. Dongre, R., D’angelo, J., Copeland, A.: Refinement of flow number as determined by the asphalt mixture performance tester for use in routine QC/QA practice. Transp. Res. Rec. J. TRB 2127, 127–136 (2009) 14. Francken, L.: Pavement deformation law of bituminous road mixes in repeated load triaxial compression. In: Proceedings of the Fourth International Conference on the Structural Design of Asphalt Pavements, vol. I, pp. 483–496. The University of Michigan, Ann Arbor, Michigan (1977)
Analysis of High Temperature Performance of Composite Asphalt Rubber Mixture Haibin Li, Mingming Zhang, Ahmed Abdulakeem Temitope, Hongjun Jing, Guijuan Zhao, and Qingwei Ma
Abstract Asphalt rubber mixture is often described as an environmentally friendly mixture due to its usage of recycled rubber gotten from used tires and the subsequent improved service life. It has been established that adding SBS to rubber asphalt can further improve the performance of rubber asphalt mixture. In order to evaluate the performance improvement of the asphalt rubber mixture, several related high temperature performance and mechanical properties tests were conducted. The influence trend was obtained and evaluated based on the analysis of dynamic stability, relative deformation, and shear strength. For both SMA-13 and SMA-16 mixture, CR/SBS composite modified asphalt had better high temperature performances than asphalt rubber, but worse than SBS asphalt. In addition, the CR/SBS composite modified asphalt had larger internal friction angle, which could give better cohesive force for aggregate gradation. Moreover, it provided a wider range of applications for asphalt rubber mixture in road construction, which realized the waste cyclic utilization and environmental protection. Keywords Road engineering · Asphalt rubber · SBS and CR composite modified asphalt · High temperature performance · Mechanical property
H. Li (B) · M. Zhang · A. A. Temitope · H. Jing · G. Zhao School of Architecture and Civil Engineering, Xi’an University of Science and Technology, Xi’an, China e-mail: [email protected] A. A. Temitope e-mail: [email protected] H. Jing e-mail: [email protected] G. Zhao e-mail: [email protected] Q. Ma Xi’an Highway Research Institute, Xi’an, China e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_31
247
248
H. Li et al.
1 Introduction Waste rubber is made by crushing recycled rubber products such as tires and rubber shoes to remove impurities. Processing waste automobile tires into rubber powder for modified asphalt not only solves the environmental pollution problem but also promotes the recycling of resources [1]. Therefore, how to effectively solve this problem is of paramount importance [2– 5] and relevant experts have done lots of work on this. Xia Wei [6] studied the road performance, high temperature performance and construction technology of asphalt rubber, and obtained the mixing time, the number of rubber powder and the amount of asphalt rubber for influence of asphalt rubber concrete. Xiao Chuan [7] determined the preparation scheme of rubber asphalt by orthogonal test method, and found through dynamic shear rheological (DSR) test: properly reducing the rubber to gel ratio of asphalt rubber cement, and using cement instead of mineral powder as filler, helps to improve the high temperature performance of rubber asphalt mixture. In this paper, the CR/SBS composite asphalt mixture was compared with SBS asphalt mixture and asphalt rubber mixture to explore the performance of composite asphalt rubber mixture and study high temperature performance evaluation index of CR/SBS asphalt mixture.
2 Materials and Experiment 2.1 Materials Caltex 90#A Matrix asphalt was used in this paper. The crumb rubber powder (15% weight of asphalt) used in this study was a kind of ordinary non-activated rubber, 30 mesh. SBS content was 2% weight of asphalt.
2.2 Experimental Sections Dynamic stability test, test of rebound deformation of asphalt mixture and shear strength test were conducted in the study.
Analysis of High Temperature Performance …
249
3 Results and Discussion 3.1 Dynamic Stability Analysis and Deformation Law (1)
Type 13 Asphalt Mixture Dynamic stability
Dynamic stability (times/mm)
The dynamic stability (DS) is shown in Fig. 1. It was found that the asphalt mixture prepared with asphalt rubber had the lowest DS values, which were 3369 cycles/mm at 60 °C and 2027 cycles/mm at 70 °C, respectively. In response to this, the ARG-13 had the highest relative deformation (4.10% at 60 °C and 5.77% at 70 °C). However, with the addition of SBS, the DS values of CR/SBS asphalt mixture at 60 and 70 °C increased obviously and reached to 4566 cycles/mm and 3751 cycles/mm. This demonstrated that the SBS incorporation increased the stiffness of asphalt mixtures, which in turn reduced its deformation under cyclic loading. Compared to the SMA13, ARG-13 and CR/SBSG-13 mixture had smaller DS values due to its gap-gradation which could provide better bearing capacity with more asphalt content. At the same time, with the temperature increasing by 10 °C, the dynamic stability of ARG-13, SMA-13 and CR/SBSG-13 decreased by 39.8%, 30.9% and 17.8% respectively, which were still much higher than the specification requirements. It indicated that the temperature had a significant impact on dynamic stability. Though the DS value of CR/SBSG-13 still had obvious decrease of 62.5 and 36.6% when compared to that of SMA-13 mixture at 60 and 70 °C. Therefore, CR/SBS composite modified asphalt has better high temperature performances than asphalt rubber, but worse than SBS asphalt. For crumb rubber is
ARG-13
CR/SBSG-13
SMA-13
Fig. 1. Dynamic stability of various mixtures at different temperatures for 60 min
H. Li et al.
Dynamic stability (times/mm)
250
ARG-16
CR/SBSG-16
SMA-16
Fig. 2. Dynamic stability of various grades at different temperatures
cheaper than SBS modifier, so the CR/SBS composite modified asphalt can improve the mixture performances while reducing the material costs. (2)
Type 16 Asphalt Mixture Dynamic stability
The rutting test results of ARG-16, SMA-16 and CR/SBSG-16 mixture, at 50, 60 and 70 °C are shown in Fig. 2. The DS value of gradation 16 had similar change, but higher than that of gradation 13 at the same temperature and asphalt binder. The DS values of CR/SBS asphalt mixture at 60 and 70 °C increased to 7412 cycles/mm and 4748 cycles/mm, which had 62.3 and 62.3% rate of increase when compared with CR/SBSG-13 mixture. It also proved the effect of SBS modifier on asphalt rubber and advantages of gradation 16. At the same time, with the temperature increasing by 10 °C, the dynamic stability of ARG-16, SMA-16 and CR/SBSG-16 decreased by 46.9, 80.7 and 56.1%, but they were still much higher than the specification requirements. The high temperature of SMA-16 had obvious decrease with temperature increasing. Based on the good properties of CR/SBS mixture, the different gradation could be selected according to the pavement grade and requirements for site construction.
3.2 Rebound Deformation Analysis Figure 3a, b showed the spring back deformation of type 13 and type 16 mixtures at different times. The rebound deformation of each mixture was very small, the maximum deformation was 0.04 mm earlier, and CR/SBSG and SMA returned to
ARG-13 SMA-13 CR/SBSG-13
0.00
-0.02
-0.04
1d
2d
3d
251 springs - back deformation (mm)
springs - back deformation (mm)
Analysis of High Temperature Performance … 0.02
0.00
-0.02
-0.04 1d
4d
2d
3d
4d
place time (day)
place time (day)
(a) Rebound Deformation of Type 13
ARG-16 SMA-16 CR/SBSG-16
(b) Rebound Deformation of Type 16
Fig. 3. Springbuck deformation under various gradations
the original level on the third day, indicating that wet rubber asphalt mixture did not have large rebound deformation under wheel load.
3.3 Shear Strength Analysis The calculated uniaxial penetration shear strength, penetration modulus and unconfined compression strength are shown in Figs. 4, 5, 6 and 7. Judging from the shear strength, SMA-16 had the highest shear strength at 60 °C, followed by CR/SBSG-16. ARG-16 had the lowest shear strength, which was consistent with the dynamic stability test result and verified the advantage of gap gradation in high temperature performance. Meanwhile, the penetration modulus of Fig. 4. Uniaxial shear strength
252 Fig. 5. Uniaxial penetration modulus
Fig. 6. Unlimited failure load
Fig. 7. Unconfined compressive strength
H. Li et al.
Analysis of High Temperature Performance …
253
SMA-16 at 60 °C was also the highest which was consistent with the shear strength evaluation. From the perspective of internal friction angle, all the angles decreased with temperature increasing, but the difference of three mixtures was very small. The CR/SBSG-16 was the largest and the SMA-16 was larger than that of ARG16. This illustrated that the compound asphalt with rubber and SBS could provide better cohesive force for aggregate gradation at 60 °C to get better high temperature performance, which also reflected better effectiveness compared to other two asphalt binders.
4 Conclusions (1)
(2)
For both SMA-13 and SMA-16 mixture, CR/SBS composite modified asphalt has better high temperature performances than asphalt rubber, but worse than SBS asphalt. As crumb rubber is cheaper than SBS modifier, so the CR/SBS composite modified asphalt can improve the mixture performances while reducing the material costs. Compared with asphalt rubber and SBS modified asphalt, the CR/SBS composite modified asphalt had larger internal friction angle. It illustrated the compound asphalt with rubber and SBS could give better cohesive force for aggregate gradation, which would provide asphalt pavement surface better service function in high temperature season.
References 1. Guo, L.P., Zhang. Z.Q., Zhang, W., et al.: Study on the mucilage performance of rubber asphalt based on high temperature properties and bond behavior. J. Wuhan Univ. Technol. (2012) 2. Niu, D., Han, S., Ouming, X.U., et al.: Pavement performance of rubber asphalt mixture with polyethylene additives. J. Jiangsu Univ. (2016) 3. Zhang, X., Zhang, B., Chen, H., et al.: Feasibility evaluation of preparing asphalt mixture with low-grade aggregate, rubber asphalt and desulphurization gypsum residues. Materials 11(8), 1481 (2018) 4. Wang, L., Xing, Y.M., Chang, C.Q.: Experimental study on temperature characteristic of compound crumb rubber modified asphalt and mixture. Adv. Mater. Res. 446–449(1), 3647–3651 (2011) 5. Xiao, C., Ling, T., Qiu, Y.: Optimization of technical measures for improving high-temperature performance of asphalt-rubber mixture. J. Mod. Transp. 21(4), 273–280 (2013) 6. Wei, X.: Study on Road Performance of Waste Rubber Powder Modified Asphalt And Asphalt Mixture. Chongqing Jiaotong University, Chongqing (2009) 7. Chuan, X.: Research on High Temperature Performance and Construction Technology of Rubber Asphalt and Mixture. Chongqing Jiaotong University, Chongqing (2009)
Analysis of Low Temperature Cracking Behavior at Binder, Mastic and Asphalt Concrete Levels Christiane Raab, Martin Arraigada, and Hamza Ibrahimi
Abstract In 2008, performance oriented methods for conformity testing of asphalt mixtures were introduced in the European standard series EN 13108. These standards establish primary performance properties like low temperature resistance, stiffness, resistance to fatigue and permanent deformation. However, these tests are complex and require enormous testing efforts making it difficult to implement them in routine testing. The research project VEGAS (Vereinfachte Ansprache des Gebrauchsverhaltens von Asphalt), a collaboration between 3 research institutions from Germany Austria and Switzerland, aimed therefore to simplify the testing system, but keeping the valuable information about the material performance. Testing effort was reduced to the determination of the volumetric properties and carrying out tests only on the binder and mastic using multiscale models. This paper focuses on some of the experimental efforts carried out by Empa in Switzerland, related to low temperature cracking of different types of bitumen and mastic and fatigue experiments on asphalt samples and slabs. Within this work a new testing procedure using the Double Torsion Testing method is proposed to evaluate low temperature crack propagation in mastic. Further, the testing on asphalt slabs was performed with a scaled traffic simulator and the results were used for validating laboratory scale test on cylindrical specimens. The focus of the paper is providing results and discussion of certain test methods, a comprehensive overview of the whole project is not intended. Keywords Bitumen · Mastic · Low temperature cracking · Double torsion test · Validation · MMLS3
C. Raab (B) · M. Arraigada Empa, Science and Technology, Duebendorf, Switzerland e-mail: [email protected] H. Ibrahimi Vinci Autoroutes, Narbonne, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_32
255
256
C. Raab et al.
1 Introduction In 2008, performance oriented methods for conformity testing of asphalt mixtures were introduced in the European standard series EN 13108. These standards establish primary performance properties like low temperature resistance, stiffness, resistance to fatigue and permanent deformation. However, these tests are complex and require enormous testing efforts making it difficult to implement them in routine testing. The research project VEGAS (Vereinfachte Ansprache des Gebrauchsverhaltens von Asphalt), a collaboration between 3 research institutions from Germany Austria and Switzerland, aims therefore to simplify the testing system, but keeping the valuable information about the material performance. It consisted of different work packages and comprised test methods for the whole temperature spectrum of bituminous materials. The overall goal was finding connections between the results from bitumen, mastic and asphalt testing. As for the Swiss part, the focus was on the validation using medium scale performance testing. Here, testing was done using a 1:3 traffic simulator for comparing the performance the different asphalt mixtures from 3 countries. Additionally, Empa was involved in the determination of low temperature performance. Here, a new testing procedure using the Double Torsion Testing (DTT) method was proposed to evaluate low temperature crack propagation in mastic.
2 Material and Test Program In the frame of the research project VEGAS different aggregates, fillers and bitumen were used. The aggregates represent country specific types used in Austria, Germany and Switzerland. 10 common types of unmodified and polymer-modified bitumen from well-known producers were chosen. For all investigated material combinations, designations consisting of 3 numbers were defined as depicted in Table 1: numbers 1 and 2 represent the type of bitumen and the producer while number 3 stands for the type of aggregate. This paper describes the results of testing carried out at Empa. Due to the limitation of paper’s length, only the results of low temperature testing of bitumen and mastic in terms of Fracture Toughness Test (FTT) and Double Torsion Testing (DTT) are presented. Further, the validation on asphalt specimens using medium scale performance testing using a 1:3 model traffic simulator MMLS3 will be shown and compared to small scale testing on cylindrical specimens.
Analysis of Low Temperature Cracking Behavior …
257
Table 1. Materials used in research project VEGAS Bitumen
Aggregate
Mastic
Asphalt
Aggregate
Name
Name
B11
M111
A111
B12
M121
A121
B13
M131
A131
Type
Producer
Name
50/70
A B C D
B14
M141
A141
A
B21
Gabbro
M211
A211
B
B22
Filler lime
M221
A221
C
B23
(Germany)
M231
A231
D
B24
M241
A241
70/100
B32
M122
A122
45/80–65
B42
M222
A222
50/70
B12
Porphyr
M322
A322
25/55–55
B22
Filler: lime
M222
A222
B32
Austria
25/55–55
70/100
M322
A322
45/80–65
B
B42
M422
A422
50/70
B12
M123
A123
25/55–55
B22
Silicat
M223
A223
70/100
B32
Filler: lime
M323
A323
45/80–65
B42
Switzerland
M423
A423
3 Low Temperature Testing of Bitumen and Mastic 3.1 Fracture Toughness Testing (FTT) The Fracture Toughness Testing FTT is described in the technical specification CEN/TS 15963 [1] and aims to evaluate the low temperature behavior of unmodified and modified bituminous binder samples. It is based on the calculation of the Fracture Toughness Temperature TFT by means of a three point bending test on a notched binder sample. In the frame of this work, asphalt mastic specimens were evaluated for the first time using this procedure. The preparation of mastic specimens proved to be a challenge due to the high viscosity of the material and the difficulty of filling the mold without holding small air bubbles that can create discontinuities. The results showed however good repeatability. Figure 1a displays the TFT for bitumen 50/70 from different producers, while Fig. 1b depicts a comparison of TFT results for bitumen 25/55–55 and mastic with same bitumen and German filler. Figure 1c presents the results for different bitumen compared to the mastic with same bitumen and different fillers.
258
C. Raab et al. 0
0 M131 M121
-10 M111
-15
a -20
B13
-5 B14
B12
M141
T FT (°C)
T FT (°C)
-5
-10
B23 B21
M211 B22
-15
50/70
B11
b -20
M221
M231 B24 M241
25/55-55
0
-15
M121 M122 M123 B22 M221 M222 M223 B23 M231 M232 M233 B24 M241 M242 M243
-10 B12
T FT (°C)
-5
c -20 Fig. 1. TFT results: a TFT for bitumen 50/70 from different producers compared to the mastic with same bitumen and German filler, b TFT for bitumen 25/55–55 compared to the mastic with same bitumen and German filler. c TFT of different bitumen compared to the mastic with same bitumen and different fillers
The results in Fig. 1a, b clearly show the influence on the bitumen producer on TFT . As visible in Fig. 1c, there is a trend in the relationship between the binder and the mastic produced with the same type of bitumen and different fillers: unmodified bitumen specimens have in all cases but one, lower TFT than the corresponding mastic. On the other hand, polymer modified bitumen specimens have higher TFT than the mastic made with the same binder.
3.2 Double Torsion (DTT) In the case of the Double Torsion Testing (DTT), there are no standards and the literature shows only a few examples of the use of the double torsion principle for characterizing low temperature crack propagation in asphalt concrete specimens [2]. Therefore, it was necessary to develop a methodology to produce the plate shaped specimens and to test them following the theory. Molds were designed and constructed, so that it was possible, to create a notch without the need of sawing. In this way, asphalt mastic plate specimens were produced without damaging the specimen while demolding. Further, the setup for applying the loading was designed and constructed as shown in Fig. 2a. The specimens were loaded at −20 °C and the results of the obtained force-deformation curves were determined (see Fig. 3b). As visible from Fig. 3b two parameters are important: the max. force and the crack length. It is also possible to calculate the stress intensity factor. Figure 3 shows the results for mastic with unmodified bitumen 50/70 in terms of maximum force and
Analysis of Low Temperature Cracking Behavior …
a
259
b
Fig. 2. Double Torsion Test (DTT). a Test set-up for mastic. b Force-deformation curves for two different mastic types (examples)
Fig. 3. DTT results for mastic with unmodified bitumen 50/70. a Max. force (N). b Stable crack propagation length (mm)
stable crack propagation length, that can be interpreted as the resistance that the material oppose to the progression of the already growing crack. According to Fig. 2b repeatability for specimens from the same material is quite good. Further, the results in terms of force and crack length clearly differ between mastic with unmodified and polymer modified bitumen with both values being greater in case of the latter. When looking at the results for unmodified bitumen the influence of different producers is visible, especially M131 seems more prone to cracking and crack propagation.
4 Validation Using Small and Medium Scale Performance Testing For validation testing different asphalt plates according to Table 1 were constructed and tested using two different methods. On the one hand performance was evaluated for cores and tested according to the German guideline AL Sp-Asphalt 09 [3]. On the other hand, validation tests were made on big sized (1.6 × 0.6 × 0.04 m) notched
260
C. Raab et al.
plates in which the load was applied by the laboratory 1:3 scaled traffic simulator MMLS3 [4] in form of rolling tires, simulating the repetitive bending produced by real traffic on a road eventually producing a crack that will propagate though the pavement. In this kind of test, crack formation is facilitated with a notch and the objective is to determine the number of MMLS3 cycles required the crack to reach the upper part of the plate. This is also not a standard tests and the set-up required correct design and the development of a procedure to evaluate crack propagation with LVDTs and the use of a Digital Imagining Correlation system as visible in Fig. 4. Figure 5 shows deformation vs. the accumulated load cycles for all asphalt plates with none modified bitumen 50/70 and with polymer modified bitumen 25/55–55. The results show that the fatigue performance of asphalt plates with 50/70 bitumen is much less than the performance of the plates with polymer modified bitumen. Fig. 4. Mobile load simulator MMLS3: test set-up
2.0 1.5 1.0
load cycles
600’000
500’000
400’000
300’000
200’000
0.0
100’000
0.5
0
deformation amplitudes [mm]
Fig. 5. Deformation measurements of MMLS3 tests for all asphalt plates
Analysis of Low Temperature Cracking Behavior …
261
5 Conclusions From the limited evaluation described in this paper the following conclusions can be drawn: • Fracture Toughness and Double Torsion Testing proved to be valuable tools for low temperature testing of bitumen and/or mastic. The results for FTT clearly show the influence on the bitumen producer on TFT , a fact that has recently led to a lot of problems with bitumen quality. • From FTT a correlation between bitumen and mastic made with the same bitumen depending on the kind of bitumen (none modified vs modified) seems to exist. • Validation testing with MMLS3 showed different performance of asphalt plates with none modified bitumen and with polymer modified bitumen.
References 1. CEN/TS 15963 Bitumen and bituminous binders—determination of the fracture toughness temperature by a three point bending test on a notched specimen. European committee for standardization (CEN), Brussels (2014) 2. Kim, H., Partl, M.N.: Development of a double-torsion fracture test for bituminous materials. In: 11th International Conference on Asphalt Pavements, ISAP, Nagoya, Japan, August (2010) 3. Sp-Asphalt, A.L.: Technical guideline for the determination of stiffness and fatigue performance of asphalt. German Research Society for Roads and Transportation (FGSV), Cologne (2009) 4. Arraigada, M., Perrotta, F., Raab, C., Tebaldi, G., Partl, M.N.: Use of APT for the validation of the efficiency of asphalt reinforcement grids. In: Aguia-Moya, J.P. et al. (eds.) The Roles of Accelerated Pavement Testing in Pavement Sustainability, pp. 509–521. Springer (2016). ISBN 978-3-319-42796-6. https://doi.org/10.1007/978-3-319-42797-3_33
Analysis of Ovalization Measurements on Flexible Airfield Pavement Under HWD Dynamic Impulse Load Maissa Gharbi , Michaël Broutin , Ichem Boulkhemair , Mai Lan Nguyen , and Armelle Chabot
Abstract Non-destructive techniques (NDT) have emerged for the last decades to in-situ evaluate the pavement materials and interface conditions. These techniques turn out to be valuable tools to evaluate pavement bearing capacity, but present limitations to provide, on their own, reliable information about the interface bonding conditions. Developed in the 70’s, the ovalization test aims at evaluating the interface bonding condition between pavement layers over the cross of a rolling wheel loads. It consists in measuring the diameter variation of a cavity at different depth levels and in three horizontal directions (longitudinal, transverse and at 45° related to the loading direction). The French civil Aviation technical center (STAC) launched a research project to evaluate this device for flexible airfield pavement submitted to dynamic loadings. This paper presents some of the first multi-height tests that have been performed using the Heavy Weight Deflectometer (HWD) impulse loading at the STAC’s full-scale airfield pavement test facility constructed on the Centre d’Expérimentations Routière (CER) site at Rouen in France. The results lead to propose the design of new ovalization devices for airfield pavements. Keywords Ovalization test · HWD · Airfield pavement · Interfaces · Modeling
1 Introduction Diagnosis step is of major concern in the frame of any study about in-service pavement. It allows assessing the pavement condition, identifying distresses, evaluating a residual life and/or anticipating potential maintenance. The interface bonding conditions between bituminous layers are a particular area of focus [1]. Non-destructive techniques (NDT), such as the Heavy Weight Deflectometer (HWD) used for airfield pavements [2, 3], have emerged for the last decades to in-situ evaluate the pavement M. Gharbi (B) · M. Broutin · I. Boulkhemair French Civil Aviation Technical Center (STAC), Bonneuil-sur-Marne, France e-mail: [email protected] M. L. Nguyen · A. Chabot MAST-LAMES, Univ Gustave Eiffel, IFSTTAR, 44344 Bouguenais, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_33
263
264
M. Gharbi et al.
materials and the interface conditions [4]. These techniques turn out to be valuable tools to evaluate pavement bearing capacity, but present limitations to provide, on their own, reliable information about interfaces conditions. One of the common options is to implement strain gauges at the bottom of pavement layers during the pavement construction. However, this method is costly and delicate. Developed in the 70’s from the regional laboratory of Saint Brieuc [5–7], the ovalization test aims at evaluating interface bond condition between pavement layers over the cross of rolling wheel loads. It consists in measuring the diameter variation of a cavity at different depth levels and in three horizontal directions (longitudinal (L), transverse (T) and 45° related to the loading direction). The French civil Aviation technical center (STAC) launched a research project to evaluate this test for flexible airfield pavement submitted to dynamic loadings. Its objective is to evaluate existing devices and to modernize or eventually to propose new systems for airfield flexible pavement interface characterization needs. The ovalization system has to enable: • Simultaneous multi-depth measurements at different levels within the bituminous surface course, including the pavement surface level; • Interface characterization linked to both rolling wheel loading and HWD stationary impulse loading at several distances from cavity including the possibility to position the HWD test plate centered right above the cavity (axisymmetric configuration). This paper describes first the ovalization test principle and some test responses obtained on a full-scale section that have been loaded by HWD during autumn 2018.
2 History of the Ovalization Test As indicated in the literature [5, 8], the variation of the diameter of a cylindrical cavity (in three directions) linked to a surface loading, provides valuable information about stresses and strains inside the structure (Fig. 1a). In the case of a multi-layer structure, measuring such diameter variation below and above the interface between two layers allows assessing its bonding condition. Based on this concept, the ovalization test
Fig. 1 The Ovalisation principle [5]: a diameter variation of the cylindrical cavity under loading; b and c the unidirectional sensor
Analysis of Ovalization Measurements …
265
Fig. 2 Ovalization devices: a the ovalization tridirectional first generation [3], b the 3 plane direction; c the CEREMA device [7]
was firstly proposed in 1974 with a simple configuration named: the unidirectional sensor and used under rolling loads [5]. It is composed of one electromagnetic sensor mounted on a bracket which ensures its holding to a set position in the hole and the setting of its measurement range (Fig. 1b, c). This unidirectional measurement system was improved and replaced with a tridirectional system (Fig. 2a) [6, 9]. This first generation version, developed in 1983 [3] contains three electronic sensors which measure simultaneously the hole diameter variation in three plane directions (L, T and 45° relatively to the loading direction) (Fig. 2b). As the setting-up of this tridimensional system (Fig. 2a) was very tedious and sensitive, the Centre d’étude et d’expertise sur les risques, l’environnement, la mobilité et l’aménagement (CEREMA) upgraded it. It leads to the 2nd generation tridirectional probe (Fig. 2c). This new device includes a six sensors system (two per direction) and involves electric motors for automated positioning. Like the aforementioned unidirectional and first generation tri-directional probe, this second generation tri-directional probe enables measuring the cavity diameter variations at only one depth at a time.
3 Ovalization Campaigns for Airfield Pavement Needs In the aim to evaluate the second generation of the tridirectional probe [7] for airfield pavement, the device has been tested under both rolling wheel loading and HWD impulse loading. Tests were conducted in 2018 with this system, designed for 107 mm cavity diameter, in partnership with the CEREMA, at the STAC’s full-scale airfield pavement test facility constructed on the Centre d’Expérimentations Routières (CER) site (Rouen, France). The pavement structure is a conventional airfield flexible pavement composed of a bituminous surface course and a humidified UGA subbase layer (Fig. 3). It comprises two areas, singled out by the subgrade water-content. The ovalization test campaign was performed in the area presenting a water-saturated subgrade. The average temperature within the bituminous layer was constant during
266
M. Gharbi et al.
Fig. 3 STAC’s full scale facility and HWD versus ovalization experimental protocol
the whole test campaign, equal to 7 °C. The maximal acquisition frequency reached during this test campaign was ~130 Hz. Figure 4a illustrates an example of the diameter variation curves measured in the three directions (T, L and 45°) under rolling loading condition, at an unidentified level within the bituminous course, some doubts existing about the tests referencing. A regular shape without significant defect is observed. It shows that the ~130 Hz acquisition frequency is high enough for those loading conditions. It can then be concluded that the current ovalization system seems to be satisfactory for tests performed under rolling loading and with a 130 Hz acquisition frequency. In a second part, the experimental program includes HWD tests at six different distances from the cavity (Fig. 3). In the cavity, the ovalization system is placed one at a time at three depths within the bituminous layer. Two HWD load levels (respective peak values F1 ~ 250 kN and F2 ~ 125 kN) were applied. For each HWD test and for each load level, three drops were performed for repeatability study. Despite the important number of tests that have been carried out under the HWD loading in this experimental program, only a very few tests are exploitable. The Fig. 4b presents one example of the diameter variation curves measured in the three directions at the bottom of the cavity, corresponding to a test performed at 60 cm from the cavity. An oscillation shape is observed for all curves which derives from mechanical vibrations
Fig. 4 CEREMA ovalization device measurements: a cavity diameter variations under rolling loading, b cavity diameter variations under HWD at the bottom of the cavity
Analysis of Ovalization Measurements …
267
of the probe. Furthermore, the acquisition frequency (~130 Hz) which is the maximal rate applicable for this prototype is not satisfying. Those experimental findings are then not reliable enough and cannot be used for comparison with FE modeling work. New devices are proposed.
4 New Design Proposals for the Ovalization Device For tests performed under HWD load, due to such dynamic load, previously mentioned results show some limitations of the current system of the ovalization test. Actually it is not possible to reach an adequate acquisition frequency for such a dynamic load. The measurements involving HWD tests require appropriate acquisition rates, at least 300 Hz and preferably 1 kHz. Furthermore it does not allow the measurement of the diameter variation at the surface of the surface AC layer, and the implementation of centered HWD tests calls for wireless system. To achieve these requirements, two prototypes were recently developed by the STAC (Fig. 5). It consists in self-bearing horizontal tray equipped, with six DD1 sensors (provided by HBM constructor) positioned in pairs for each direction (L, T, 45°). It makes possible to assess the diameter variation. Each system can easily be upgraded to a multiple tray system for simultaneous multi-depth measurements. Measurements at the surface of the bituminous course layer are possible. The mechanical and electronical systems retained are consistent with measurements under HWD dynamic impulse loads. Particularly, it enables reaching acquisition rates compatible with those solicitations. Tests were conducted at the STAC full-scale airfield pavement test facility in Bonneuil-sur-Marne [2, 10]. Figure 6 shows first results of the ovalization test performed with the two news prototypes successively placed within the cavity at the pavement surface under HWD impulse loading. Tests are firstly carried out with
Fig. 5 STAC’s Ovalization prototype: a first design, b second design
268
M. Gharbi et al.
Fig. 6 Diameter variations curves at the surface of the cavity using: a the first design, b second design
the HWD plate positioned adjacent to the cavity edge with a HWD load peak value of F ~ 170 kN and with an adequate frequency acquisition (4800 Hz). Others load peak values were applied and the results are presented in [10]. For the second prototype (Fig. 6b) an oscillation shape is observed for the diameter variations curves contrary to first prototype (Fig. 6a) which seems to be more satisfactory. The improvement of both prototypes is in progress to overcome some limitations related to the implementation step and to the curve oscillation issue.
5 Conclusion This paper presents some results from two ovalization test campaigns carried out at the STAC’s full-scale airfield pavement test facility, on the CER site in 2018. The CEREMA second generation probe of Ovalization device is evaluated for airfield pavement needs. These test campaigns include both conventional tests using rolling wheel load, and new type of loading involving Heavy Weight Deflectometer (HWD) impulse load, performed at different distances from the cavity. Those tests where. The cavity diameter variation is studied in three horizontal directions (T, L and at 45°) for both loading conditions. The conventional tests showed satisfactory results. By contrast, measurements obtained under HWD impulse load are not exploitable: the acquisition rate is too low for such solicitation and mechanical vibrations are observed. Furthermore, the current system does not enable performing diameter measurements at the pavement surface, neither carrying out HWD tests with the load plate centered right above the cavity (axisymmetric configuration). In order to overcome these current system limitations, two new ovalization systems are being developed by the authors [10] and first tests are briefly presented here. In parallel, 3D finite element dynamic modeling is developed for data analysis, more tests and results will be illustrated in [10].
Analysis of Ovalization Measurements …
269
References 1. Petit, C., Chabot, A., Destrée, A., Raab, C.: Interface debonding behaviour. In: Buttlar, W.G., Chabot, A., Dave, E.D., Petit, C., Tebaldi, G. (eds.) MCD in Asphalt and Composite Pavements. RILEM State-of-the-Art Reports, 1st ed, vol. 28, pp. 103–153. Springer, Cham (2018). https:// doi.org/10.1007/978-3-319-76849-6_3 2. Broutin, M., Sadoun, A., Duprey, A.: Comparison between HWD backcalculated subgrade dynamic moduli and in situ static bearing capacity tests. In: Al-Qadi, I., Ozer, H., Loizos, A., Murrell, S. (eds.) AHP2019, pp. 404–413. ASCE Press (2019) 3. STAC: Airfield flexible pavement assessment using HWD. Guide Technique (2014) 4. Sadoun, A., Broutin, M., Simonin, J.M.: Assessment of HWD ability to detect debonding of pavement layer interfaces. In: Chabot, A., Buttlar, W.G., Dave, E.V., Petit, C., Tebaldi, G. (eds.) Rilem Bookseries, vol. 13, pp. 763–769. Springer, Netherlands, Dordrecht (2016) 5. Kobisch, R., Peyronne, C.: L’ovalisation : une nouvelle méthode de mesure des déformations élastiques des chaussées. Bull. Liaison Labo P. et Ch. (102), 59–71 (1979) 6. Goacolou, H., Keryell, P., Kobisch, R., Poilane, J.P.: Utilisation de l’ovalisation en auscultation des chaussées. Bull. Liaison Labo P. et Ch. (128), 65–75 (1983) 7. Ruiz, O., Voisin, G.: L’essai d’ovalisation. RGRA (962) (2019) 8. Savin, G.N.: Stress Concentration Around Holes, 1st ed. Pergamon (1961) 9. Martin, J.M.: Ovalisation, exécution et exploitation des mesures : méthode d’essai LPC, vol. 41. Laboratoire Central des Ponts et Chaussées (1995) 10. Gharbi, M., Broutin, M., Schneider, T., Maindroult, S., Chabot, A.: Towards an adapted ovalization system for flexible airfield pavement interface characterization using rolling-wheel or HWD loads. In: Chabot, A., Hornych, P., Harvey, J., Loria-Salazar, L. (eds) LNCE, vol. 96, pp. 516–525. Springer, Cham. https://doi.org/10.1007/978-3-030-55236-7_53
Application of Activated Carbon on Induced Heating-Healing Characteristics of Aged Mixes Saeed Amani, Amir Kavussi, and Mohammad M. Karimi
Abstract Aging phenomenon reduces the flexibility of asphalt pavements. Excessive loading and environmental conditions make asphalt mixes more prone to cracks/micro-cracks initiation and propagation. Cracks/micro-cracks in asphalt mixes can be healed through an induced healing process. Flowing and crack filling capability of asphalt binders play an important role in induced healing efficiency. This paper investigates the effects of different aging level on induced heating-healing characteristics of asphalt mixes under microwave radiation. Asphalt mixes were modified applying activated carbon to enhance its electromagnetic sensitivity. To compare heating absorption, cubic samples were fabricated at different aging levels. In addition, breaking-healing cycles through indirect tensile (IDT) test was applied to evaluate the effect of aging on induced healing process. Results revealed that microwave heating rate of asphalt mixes decreased as the aging level increased. In addition, aging has adverse effects on induced healing efficiency of asphalt mixes. The experimental testing results indicated that the devastating influence of breaking healing cycles was increased as the aging levels increased, which was more pronounced at higher aging levels. Keywords Induced healing · Microwave heating · Laboratory aging · Cyclic Breaking-Healing · Electromagnetic radiation
1 Introduction Aging occurs in asphalt binder due to the presence of oxygen, ultraviolet radiation, and change in temperature. Asphalt aging process can alter the chemical and physical properties, which resulted in tending low performance of pavements [1]. Cracking caused by cyclic loading, environmental conditions, and aging are the main cause of pavements distresses [2]. Cracking propagation resulted in reduction of service life and an increase in failure and maintenance costs. Asphalt mixes have the ability S. Amani · A. Kavussi (B) · M. M. Karimi Department of Civil and Environmental Engineering, Tarbiat Modares University, Tehran, Iran e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_34
271
272
S. Amani et al.
to self-heal their cracks/micro-cracks when exposed to high temperatures [3]. If the asphalt binder is not aged exceedingly, asphalt mixes can heal the cracks/microcracks. By increasing asphalt binder temperature, the viscosity decreases dramatically. This means that, when asphalt mix temperature gets to Newtonian fluid temperature or higher, the melted binder moves inside the crack/micro-cracks and regain the lost strength [3–5]. Concerning to enhancement of induced healing, different heating sources such as microwave radiation [4–6], magnetic induction [7–9], and infrared radiation [10] were used to accelerate the induced healing process of asphalt mixes. The optimum temperature of asphalt mixes in the induced healing process under the electromagnetic radiation is in the range of 90–100 °C [3–5]. High temperature leads asphalt binder to exhibit efficient fluidity and enhanced induced healing. To enhance heating and healing properties under the electromagnetic field, most researchers used different types of a conductive additive, such as fiber-based [7, 9, 10], powder-based [4, 6], and particle-based [8, 10]. The temperature distribution in fiber-based and particle-based additives are usually non-uniform, which increased the risk of the aging phenomenon and overheating samples around this additive [9]. Most of the studies evaluated the induced healing process of unaged (virgin) mixes. Moreover, aging can change the chemical composition of asphalt binder and mixes, which can be expected that influence healing efficiency. Recently, GomezMeijide et al. investigated the effects of aging and RAP content on induced healing characteristics of asphalt mixes [8]. Based on the results, the researchers concluded that aging has a negative impact on the effectiveness of induced healing process. Fracture energy is a good indicator for analyzing the cracking potential of aged mixes [11]. According to RILEM TC 50-FMC, fracture energy is typically calculated as the area under the load-displacement (P-δ) curve (work of fracture) through the IDT test and then normalized for the diameter and average thickness of specimen or ligament area as expressed in Eq. (1). Wf 1 = Gf = Alig DT
δ P(δ) dδ
(1)
0
where Gf , W f , D, and T are fracture energy (J/m2 ), work of fracture (J), diameter (m), and the average thickness of IDT samples (m), respectively. Researchers have proposed several testing methods to determine the characteristics of long-term aging of pavements without proper validation of the results of field aging. AASHTO R30 is the current standard method to resemble long-term aging. In addition, the finding of previous studies [1, 11, 12] also focused on the long-term aging procedure of compacted specimens and relevant aging time in the field. The objective of the research was the evaluation of aging of mixes on their induced heating-healing characteristics. To this end, compacted samples were aged at constant temperature and various durations to simulate different aging levels. In order to compare heating absorption at different aging levels, cubic samples were prepared and were tested under microwave radiation. In addition, induced healing efficiency of
Application of Activated Carbon on Induced …
273
asphalt mixes at different aging levels was evaluated under breaking-healing cycles through the IDT test at 25 °C.
2 Materials and Sample Preparation The limestone aggregate with dense gradation and nominal maximum aggregate size (NMAS) of 12.5 mm was used according to ASTM D3515-01 standard method. An asphalt binder 85/100 Pen was used. Activated carbon, as a powder-based additive, was used as asphalt binder modifier. Based on the previous study [4, 6], optimum activated carbon content of 5% (by wt of asphalt binder) was selected to achieve enough sensitivity to electromagnetic radiation and appropriate mechanical performance. In order to prepare activated carbon modified asphalt binder, asphalt binder was heated to 150 °C and was mixed with activated carbon applying the shear rate of 4000 rpm for 30 min. Using Marshall mix design method, binder content of 4.7% of the total mix by weight was calculated to yield target air voids contents of 4.3%. It is worth to mention that neat asphalt binder was used in the Marshall mix design, and in order to avoid discrepancy in testing results, the aggregate gradation and the asphalt binder content were not changed when of activated carbon powder was introduced.
3 Testing Methods The experimental plan in this study involved testing of control asphalt mixes at the unaged condition and activated carbon modified asphalt mixes at both unaged and aged condition. Different types of asphalt mixes at different aging levels have been evaluated for induction heating rate and resistance to cracking.
3.1 Laboratory Aging Figure 1 shows correlations between laboratory and field aging. As it is shown in Fig. 1, the rate of aging evolution in asphalt mixture increases as the service time (exposure time) increases (the slope of the curve increases as the times increased). As a result of this, it can be explained that this procedure is in agreement with the fact that the aging level of asphalt mix increases dramatically as the service time increases. Therefore, exposure time of asphalt mix at a constant temperature can be depicted as a function of aging asphalt mixture at various service periods. In addition, various times of pavement service life (e.g., 5, 10, 15, and 20 years) has taken into consideration, then, enough time to induce the exposure time is calculated. Substituting interested pavement service life in formula of regression model shown in Fig. 1, required time for aging asphalt specimens were 3, 5, 7 and 9 days at 85
274
S. Amani et al.
Equivalent of aging in field (year)
20
ºC 15
10
Regression model
5
y = 0.1458x2 + 0.9042x + 1.075 R² = 0.9511 0 0
2
4
6
8
10
Aging duration in laboratory (day)
Fig. 1. Correlation between laboratory and field aging at various aged samples [1, 11, 12]
°C to reach the aging level of 5, 10, 15 and 20 years of service life, respectively that was named A5, A10, A15, A20. The aging level of the specimens at construction time was named A0 (unaged). According to the laboratory aging procedure, asphalt mix specimens were placed in an oven at a constant temperature of 85 °C at different exposure times. After these times, the oven was turned off, and samples were left to be cooled down to room temperature (i.e., 25 °C) for 16 h.
3.2 Microwave Heating Cylindrical samples of 101.6 mm diameter and 65 mm height were compacted using a Marshall hammer. These were aged in an oven at above-mentioned conditions. Afterward, samples were sawed to prepared cubic samples of 60 mm dimensions. The cubic samples were exposed to microwave radiation for six cycles of 20 s heating time in order to compare the heating absorption of samples. After each 20 s heating cycle, the microwave oven door was opened to measure the surface temperature of the samples randomly at six different points of the specimens (using a non-contact infrared gun). Then, the microwave oven door was closed to start the next heating cycle. Microwave heating rate of the asphalt samples after each heating cycle was measured. The measurements were averaged at three replicates. Extrapolating the temperature results overtimes, the microwave heating rate of the asphalt samples was calculated.
Application of Activated Carbon on Induced …
275
3.3 Cyclic Breaking-Healing The IDT test was conducted to evaluate the effects of different aging levels of asphalt mixes under breaking-healing cycles. To meet this objective, an experimental procedure was conducted to apply cycles of breaking and healing on each sample. After conditioning samples at testing temperature for 2 h, sample was loaded in the radial direction at a rate of 50.8 mm/min, following AASHTO T-322 testing procedure to obtain IDT strength. This was with the aim of getting to failure mode and determining fracture energy up to the maximum IDT strength. Afterwards, samples were placed at 25 °C for 2 h for relaxation purposes. The optimum temperature for healing of the samples under microwave radiation was determined to be between 90 and 100 °C. Therefore, samples were heated in microwave oven to reach to the corresponding temperature and heal the existing cracks, which were formed while the sample was loaded in the breaking step. Finally, samples were kept for 24 h at 25 °C in order to get prepared for the next loading cycle. The healing index (H.I) as represented in Eq. (2) is defined as the ratio of fracture energy resisted by that of the healed specimens (Ei ) to fracture energy of resisted by initially tested samples (E0 ). H.I =
Ei E0
(2)
4 Results and Discussion 4.1 Effect of Aging Level on Microwave Heating Figure 2 shows the heating behavior of asphalt mixes. At first glance, it is apparent that activated carbon can effectively enhance electromagnetic sensitivity of asphalt mixes. For example, the microwave heating rate of control samples at unaged condition (A0) for modified samples was 1.02 °C/s, while the corresponding heating rate for control samples was 0.47 °C/s (A0-control). In addition, the microwave heating rate decreased as the aging level increased. For instance, heating rate of activated carbon modified samples that were aged at different aging levels of A5, A10, A15, and A20 were 0.92, 0.83, 0.79, and 0.70 °C/s, respectively. Therefore, it can be expected that the induced healing efficiency of asphalt mix decrease as the aging level increases.
4.2 Effect of Aging Level on Induced Healing Figure 3 shows the effects of aging on induced healing index. The results revealed that at the unaged condition, the H.I of control samples (A0-control) was lower than
276
S. Amani et al. 160
Temperature Vs Heating Time 140
Aging level A0-Control
120
A0 A5
100
A10 A15
80
A20 60
40
20 0
20
40
60
80
100
Heating Time (s)
Fig. 2. Heating behavior of asphalt mixes at different aging levels
Fig. 3. Effects of aging level on induced healing index through IDT test at 25 °C
120
Application of Activated Carbon on Induced …
277
the modified sample (A0). Higher H.I of activated carbon can be attributed to an enhancement of induction heating rate under microwave radiation. In addition, the results indicated that the efficiency of induced healing to heal micro-cracks/cracks reduced dramatically as the aging level increased. For example, at the first breakinghealing cycles, the reduction of HI in unaged (A0) modified asphalt mix, when aging levels is increased to A5, A10, A15, and A20, were 7, 10, 19, and 22%, respectively. In addition, increasing the breaking-healing cycles and aging levels have adverse effects on H.I. For example, the detrimental effect of breaking-healing cycles in modified asphalt mix at unaged condition (A0) were 18, 32, 43, and 48%, respectively, while; this value for A20 specimen were 37, 43, 50, and 58%, respectively.
5 Conclusions Considering the effects of aging on induced heating-healing characteristics of asphalt mixes containing activated carbon, the following conclusions were drawn: • Under microwave heating, presence of activated carbon in mixes, increases their heating sensitivity and healing characteristics. • The rate of heating and healing in asphalt mixes were decreased as the aging level was increased. Hence, its effectiveness is more pronounced in less aged mixes. • Detrimental effects of breaking-healing cycles on asphalt mixes were increased as the aging level was increased. This was more pronounced at higher aging levels.
References 1. Islam, M.R., Hossain, M.I., Tarefder, R.A.: A study of asphalt aging using indirect tensile strength test. Constr. Build. Mater. 95, 218–223 (2015) 2. Mobasher, B., Mamlouk, M., Lin, H.: Evaluation of crack propagation properties of asphalt mixtures. J. Transp. Eng. 123(5), 405–413 (1997) 3. García, Á.: Self-healing of open cracks in asphalt mastic. Fuel 93, 264–272 (2012) 4. Karimi, M.M., Jahanbakhsh, H., Jahangiri, B., Nejad, F.M.: Induced heating-healing characterization of activated carbon modified asphalt concrete under microwave radiation. Constr. Build. Mater. 178, 254–271 (2018) 5. Gallego, J., del Val, M.A., Contreras, V., Páez, A.: Heating asphalt mixtures with microwaves to promote self-healing. Constr. Build. Mater. 42, 1–4 (2013) 6. Amani, S., Kavussi, A., Karimi, M.M.: Effects of aging level on induced heating-healing properties of asphalt mixes. Constr. Build. Mater. 263, 120105 (2020) 7. Norambuena-Contreras, J., Garcia, A.: Self-healing of asphalt mixture by microwave and induction heating. Mater. Des. 106, 404–414 (2016) 8. Gómez-Meijide, B., Ajam, H., Lastra-González, P., Garcia, A.: Effect of ageing and RAP content on the induction healing properties of asphalt mixtures. Constr. Build. Mater. 179, 468–476 (2018) 9. Pamulapati, Y., Elseifi, M.A., Cooper III, S.B., Mohammad, L.N., Elbagalati, O.: Evaluation of self-healing of asphalt concrete through induction heating and metallic fibers. Constr. Build. Mater. 146, 66–75 (2017)
278
S. Amani et al.
10. Ajam, H., Lastra-González, P., Gómez-Meijide, B., Airey, G., Garcia, A.: Self-healing of dense asphalt concrete by two different approaches: electromagnetic induction and infrared radiation. J. Test. Eval. 45(6), 1933–1940 (2017) 11. Rahbar-Rastegar, R., Daniel, J.S., Dave, E.V.: Evaluation of viscoelastic and fracture properties of asphalt mixtures with long-term laboratory conditioning. Transp. Res. Rec. 2672(28), 503– 513 (2018) 12. Bell, C.A., Wieder, A.J., Fellin, M.J.: Laboratory aging of asphalt-aggregate mixtures: field validation (1994)
Application of the Synthetics Geo-Composites in the Arid Zones for Rehabilitation of the Flexible Pavements Road Experimental Analysis Mouloud Abdessemed, Rabéa Bazzine, and Said Kenai
Abstract The zones of the south Algerian, as the zone of Ouargla, are among the hottest zones, that are characterized by a stern and elevated temperature that can reach the 50 °C and a very big thermal gradient, without forgetting the aridity and dried it that provoke the loads and solicitations on the layer of rolling, of the pavements of the linear infrastructures (roads, aircraft pavements). What gives birth to cracks and the permanent distortions to small scale. Indeed, the downward propagation of these cracks has ominous effects on the lower layers and decrease the lift of the whole considerably especially to the presence of water, that major the damages and accelerate the deteriorations of these last. This work enters in the setting of the contribution for the understanding of these phenomena and therefore to find the adequate solutions that converge toward the safeguard of the body of pavements and the structural capacity of its infrastructures. One will proceed, on the basis of a data base to establish a state of the places that allows to lead an experimental companion to the laboratory and in situ. The integration of the composites geogrides can give satisfactory results and can increase the 1–4 times lasted it of life of the road infrastructure. The results of the dynamic tests by HWD, on road sections reinforced by geo-composites, will be the subject of analysis and reading in this analysis, in order to wedge the other found experimentally results. Keywords Pavement · Flexible · Road · Geo-composite · Behavior · Analysis · Experimentation
1 Introduction Algeria has a very important road heritage, composed of freeways, roads and the local and farming paths. The last statistics [1], show the existence of a linear global of more than 116,000 km and with freeways of more than 4500 km of length, on the M. Abdessemed (B) · S. Kenai Laboratory GeoMaterials and Civil Engineering, University of Blida1, Blida, Algeria R. Bazzine University of Kasdi Merbah-Ouargla, Ouragla, Algeria © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_35
279
280
M. Abdessemed et al.
whole Algerian territory. However, these roads, being to about 32% in arid zones (temperature and elevated thermal gradients), what requires a particular attention in their conception, construction and maintenance. The aridity and dried it provoke, sometimes, of the loads and the excessive solicitations on the layer of rolling of these road pavements, what gives birth to cracks and the permanent distortions in their body of pavement. One finds in the maintenance practice and reinforcing of the road and/or airport infrastructures, several techniques, that converge toward their restoration, the rebirth of their lift and the remedies in case of obvious deteriorations. The recent research, bent on the detailed survey and the understanding of the phenomena of the cracks, especially in zones of hot climates (arid), or the thermal gradient is raised. These researches even proposed some solutions for the stop or the self-timing of the propagation of these cracks, or even to remedy the state of the flexible pavement damaged. Among the proposed solutions, the backing of the surface layers that increases the structural and reduced capacity in a considerable manner the constraints (normal and tangent) to the level of the pavement body [2]. Also, the geo-composite, as alternative solution, has been used with success for the stability of the soils of foundation of the infrastructures of bases (road, rail and airfields). These matters played main roles for the improvement of the mechanical performance of these pavements [3]. The research confirmed, that the geo-composite as geogrids of separation, are better the placed for this type of remedy. The influence of the insertion of the geogrids on the performance of the pavement made of bituminous concrete showed its efficiency in the reduction of the constraints and distortions in the body of pavement [4]. Our work, that has been elaborated, at a time to the laboratory of research (GeoMaterials of the university of Blida1/Algeria) and in situ, consists to an experimental survey, on a section of flexible road pavement, with the application of a category of the geo-composite (geogrids) as materials of separation and/or reinforced). This work initiated by an experimental companion in the laboratory knew city, with the confection of test-tubes made of bituminous concrete of reduced measurements tested in bending 4 points. These test-tubes, witnesses and/or reinforced have been prepared with the same composition of the concrete crotchet.
2 Application of the Géo-Synthetic in the Linear Infrastructures Since more of 15ans, the geo-synthetic began to be applied in the airport and road infrastructures. Indeed, the first application in Algeria (Fig. 1) was in 2004 to the main track of the airfield of Djanet (1800 km to the south of Algiers). These applications propagated themselves in several infrastructures (roads and airfields) in Algeria, especially in the zones of hot climates, to face the phenomena of ascents of cracks [5] in addition to his role of separator, the geogrids (family of the geo-textiles), as
Application of the Synthetics Geo-Composites in the Arid Zones …
281
Fig. 1. Application of geo-synthetic in the pavement Djanet airport
applied tablecloth, gave a gain in the constraints and distortions in the rolling layer (bituminous concrete). The geogrids could play the role of backing tablecloth [3].
3 Experimental Tests In order to understand the influence of the geogrids in the behavior of the reinforced flexible pavements, we conducted an experimental work, based on two shutters: • Shutter: tests to the laboratory on test-tubes to reduced scales, • Shutter: tests in situ by falling mass of plates (HWD). The tests to the laboratory by bending four points, consist in applying (to the superior tablecloth), a load (strength) repartees on two points on both sides of the middle of the trial plate and this plate, that are pushed on two supports (simple of a quoted and double of the other quoted). These trial plates, have been reinforced by grappling and application of two types of geogrids (Fig. 2). The measurements of the plates are identical: (60 × 20 × 10) cm. And the composition of the bituminous concrete is: aggregations ground 0/3, 3/8 and 8/15, with a binder made of pure asphalt 40/50.
Fig. 2. Plate witness reinforced by geogrids
282
M. Abdessemed et al.
Fig. 3. The system of falling and consistent mass of the deflection basin [8]
For the tests in situ, on a flexible pavement section, or a pseudo companion— dynamic to been achieved by application of the mass fall HWD [5]. The objective is to know the influence of the geogrids on the values of the constraints and distortions (Fig. 3). The HWD device consists in reproducing the solicitations due to the passage of a heavy vehicle or plane and to measure the reaction of the pavement while measuring the basin of deflection with the help of nine geophones, with a length of mobile stem of the order of 2.25 m, that can be positioned to the wanted places (abscissas of the axis horizontal), while analyzing the graphs here before.
4 Results and Discussions 4.1 Tests to the Laboratory The results gotten, in the sense of the values of maximal load of rupture, the maximal arrow and the ultimate constraint, for every type of plate, are illustrated in the following table (Table 1) The maximal load (of rupture) is determined, so to the total rupture of the sample (plate). It is about a global ductile rupture of the piece tested, without rip, nor decay of the geogrids inserted. The results found (Picture 1), show the effect of the geogrids on the behavior of the plates after backing. One notices a clean gain for every type of geogrids with percentages varying between 30 and 38%, for the maximal load and a reduction of Table 1. Test results of bending of the trial plates Value/Identification
Plate witness
Plate reinforced by geogrid/type1
Plate reinforced by geogrid/type1
Maximal load (kN)
16.03
20.85
22.15
Maximal distortion (mm)
0.232
0.218
0.205
Gain of load (%)
−
30.07
38.18
Gain of distortion (%)
−
−6.04
−11.64
Application of the Synthetics Geo-Composites in the Arid Zones …
283
the distortion between 6 and 11.7% the tablecloth of geogrid, in addition to his role of separation, can be used like means of reinforcing in shoed them flexible in asphalt.
4.2 Tests of HWD A comparison of the values of deflections and constraints, before and after the reinforcing of the track by geogrids, has been made (Figs. 4 and 5). In traction, the middle values of the deflections of first geophone recorded in reinforced them is equal in 686.9 µms (micrometers) against a value of 865 µms in the non reinforced areas, either a reduction of 26% of the distortions. For the constraints, one notes a value of 2310 kPa before backing, against the value of 2142 kPa, after backing, either a reduction of 8%. These results confirm the performance of the geogrids as tablecloth of reinforcing for the flexible pavements [7] (Fig. 5). Fig. 4. Measure before application geogrids
Fig. 5. Measure after application geogrids
284
M. Abdessemed et al.
5 Conclusion The results of this static assessment and the one pseudo-dynamic, by the measure of the distortions and constraints, to the layer of rolling (bituminous concrete) and in depth of the track, showed that the geogrids in addition to his role of separator, can play the role of reinforcing. Indeed, the found results, in situ, compared experimentally to those detmined to the laboratory, confirmed the reduction of the constraints to the surface of the layer, as well as the gain brought on the load of rupture. The findings to pull this survey are: • the solution to reinforce a flexible pavement by insertion of the tablecloths of geogrids, seem to be a very efficient and alternative solution to the traditional solutions (reloading, addition of matter); • The adequate choice of the type of geogrid improves the durability and the stability of the plates made of asphalt; in the sense of constraints; • The tests in situ by HWD can make themselves to value the static and dynamic behavior of a road and/or airport pavement while determining the constraints and distortions at any point of the pavement; • The geogrid, as composite separator, contributes to the improvement of the constraints and distortions and therefore to reduce the dimensionality and to delay the propagation of the downward cracks.
References 1. Ministry of the Public Works and Transportation (MTPT): Seized of statistics of the linear infrastructures in Algeria, Direction of the roads. Algiers, Algeria (2018) 2. Giroud, J.P., Ah-Line, C., Bonaparte, R.: Design of unpaved roads and trafficked areas with geogrids. In: Proceedings of Symposium Polymer Grid Reinforcement, Science and Engineering Research Council and Netlon Ltd, London, pp. 116–127 (2015) 3. Abdessemed, M., Kenai, S., Bali: Experimental and numerical analysis of the behavior of an airport pavement reinforced by geogrids. J. Constr. Build. Mater. 94, 547–554 (2015) 4. Djellali, A., Houam, A., Saghafi, B., Hamdane, A.: Static analysis of flexible pavements over expansive soils. Int. J. Civil Eng. 15(3), 391–400 (2017) 5. Service technique de l’aviation civile (2009) Les laboratoires du STAC, Déflectomètre à masse tombante (Heavy Weight Deflectometer), Rapport HWD, France (2009) 6. Siriwardane, H., et al.: Analysis of flexible pavements reinforced with geogrids. J. Geotech. Geol. Eng. 28, 287–297 (2010) 7. Broutin, M.: Évaluation des chaussées souples aéroportuaires à l’aide du déflectomètre à masse tombante (HWD). Thèse de Doctorat, Ecole des Ponts, Paris Tech, LCPC, Paris, France (2010) 8. Abdessemed, M., Kenai S.: Application of HWD non destructive tests on rein-forced airport pavement, 11èmes Rencontres Géosynthétiques – 7-9 Mars 2017, Lille- France (2017)
Applying the Grid Method to Study the Mechanical Behavior at Micro Scale of Asphalt Mixtures Containing Reclaimed Asphalt Pavements Ali Moukahal, Mathilde Morvan, Julien Van Rompu, and Evelyne Toussaint
Abstract The objective of this study is to characterize the mechanical performances of reclaimed asphalt pavements (RAP), at different percentages in asphalt mixtures. Six formulations R0, R10, R20, R30, R40 and R50 of the same asphalt mixture formula, a semi-coarse asphalt concrete, with a 35/50 pen bitumen were produced in Colas scientific and technical center, in Magny-les-Hameaux, by introducing 0, 10, 20, 20, 30, 40 and 50% RAP, respectively. Indirect tensile strength tests were carried out on three of these specimens (R0, R30 and R50). Stain fields were calculated using a full field measurement method, named the Grid Method (GM) to compare the behavior of the specimens. This method consists in determining and analyzing both displacement and strain fields on the surface of the specimen. Results highlight the difference in behavior between samples both at the micro- and macro-scales. Keywords Reclaimed asphalt pavements (RAP) · Grid method (GM) · Strain field · Displacement field
1 Introduction As the demand for public infrastructure increases, so does the need for building materials. Today, the demand for aggregates is much greater than before. Road maintenance, rehabilitation and demolition projects generate products, known as reclaimed asphalt pavement (RAP), that can be recycled and then reused in the construction of new roads. Choosing RAP is a good option because it reduces the need to excavate more virgin aggregates, and introduces bituminous binder into the recycling process [1]. RAP can be-reused by means of hot and cold recycling technologies [2]. A. Moukahal · M. Morvan (B) · E. Toussaint Institut Pascal—CNRS: UMR6602, Université Clermont Auvergne, CNRS, SIGMA Clermont, Institut Pascal, 24 avenue des Landais, 63171 Aubière Cedex, France e-mail: [email protected] J. Van Rompu Colas [Magny-les-Hameaux], COLAS, Campus scientifique et technique 4 rue Jean Mermoz, 78772 Magny-les-Hameaux, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_36
285
286
A. Moukahal et al.
The rate of RAP in asphalt mix depends on many factors. This rate varies from 30% for highway pavements to 60% for other applications [3]. Hence the interest of evaluating the mechanical performances, at the macro scale as well as at the micro scale, of mixtures containing RAP. The study of different phenomena at the microscale requires a full-field measurement technique that allows to measure displacement fields and to calculate strain fields at this scale. The Grid Method (GM), developed and used in the laboratory, well suited due to its good resolution both in displacement and strain. Recently, this method has been used to characterize the mechanical behavior of asphalt mixtures containing RAP [4] Other full field measurement methods such as the digital image correlation technique for instance can also be used for the same purpose, but generally with a lower resolution, and longer post treatment time [5]. The main objective of this paper is to evaluate the mechanical behavior of specimens made with different rates of RAP at the macro and the micro scale and to compare the behavior of virgin and recycled materials.
2 Experimental Set-Up and Conditions 2.1 Material and Design 2.1.1
Raw and Recycled Materials
The asphalt mixture formula is the same for the six formulation (a BBSG 0/10). The virgin material is constituted from diorite aggregates, limestone filler and 35/50 pen bitumen. The bitumen is derived from the Port-Jérôme refinery. The RAP were recovered from a planing operation carried out on the A28 highway site. Its binder content is 5.72% and its fines content (passing to 0.063 mm) is very high (11.2%). The penetration (25°, 0.1 mm) of the aged and virgin binder is 8 and 36, and their softening point (°C) is 71.6 and 51.6, respectively.
2.1.2
Grading Curve and Bitumen Content of the Mixtures
Six hot mixtures asphalt containing 0, 10, 20, 30, 40 and 50% of RAP were made and provided by Colas scientific and technical center, in Magny-les-Hameaux. The objective is to produce different asphalt mixtures with the same grading curve (Fig. 1). The bitumen content for each formulation is given in Table 1.
2.1.3
Specimen Preparation
The preparation mode of specimens is according to the method B of the NF EN 12697–12 [6]. The “Duriez” specimen are molded in metal cylinders of 80 mm
Applying the Grid Method to Study the Mechanical Behavior …
287
Fig. 1. Grading curves of the six formulations
Table 1. Bitumen contents in different formulations Bitumen 35/50
R0
R10
R20
R30
R40
R50
5.3
4.7
4.2
3.6
3
2.4
Fig. 2. Shape of specimens before sawing (a), after (b) sawing and after grid transfer (c)
internal diameter and their mass is 1000 g. Since the molding is carried out in an imposed force mode, the specimen’s height therefore varies (from 80 to 90 mm). The shape of the specimen before and after sawing is visible in Fig. 2a, b respectively.
2.2 Loading Conditions Indirect tensile tests were carried out on specimens equipped with grids glued on their sawn surfaces, in a temperature-controlled room at 23 °C, with a displacement
288
A. Moukahal et al.
rate of 0.01 mm/s. A rubber sheet was placed between the sample and the plates to ensure a better pressure distribution on the specimen.
2.3 Outline of the Grid Method The GM is an optical technique that consists in measuring displacements fields in order to obtain in plane deformation fields at the surface of materials. A grid of 0.2 mm pitch, printed on a polymer film, is transferred to the substrate surface using a white color adhesive [7] (Fig. 2c). If the substrate deformed, the grid transferred to its surface faithfully follows its movements. Images obtained with a camera, are processed using an algorithm based on the Windowed Fourier Transform in order to compute the phase change of the grid. A modulation and demodulation approach allows increasing the resolution of the technique [8].
3 Results 3.1 Response at the Macroscopic Scale The load-displacement curves of specimens containing 0, 30 and 50% of RAP are presented in Fig. 3. The three curves present similar shapes and are composed of 3 portions: a first concave portion illustrating the settlement phase of the specimen, a second linear portion that shows the linear force-displacement response and a nonlinear portion in which the load reaches its maximum values before the unloading phase. The rigidity of the materials in the linear zone increases with the increase in the RAP rate. Indeed, the slope of the linear part of the curve that increases with the increase of RAP rate evolves from 1.4, 1.7 to 2.5 kN/mm for the mixtures R0, R30 and R50 respectively. This gap is due to the difference in behavior between virgin Fig. 3 Load—displacement curves for R0, R30 and R50
Applying the Grid Method to Study the Mechanical Behavior …
289
and aged binder. Recycled aggregate binder is harder than the virgin binder as it was already exposed to aging cycles during its service phase.
3.2 Response at the Microscopical Scale 3.2.1
Strain Distribution Over the Sections
Some examples of horizontal and vertical strain fields obtained with the grid method for a loading level of 3 kN are presented in Fig. 4 for three RAP levels. The εxx strain maps show an overall expansion of the specimens and the εyy strain maps show compression along the loading direction. In addition, the strain distribution strongly depends on the distribution of the aggregates and the binder. It becomes almost zero in the areas of the strain map corresponding to the aggregates, which means that the strain is concentrated in the binder areas. It also observed that the strains are mainly concentrated around the loading axis because of the specimen’s geometry and the tested conditions. Comparing the behavior of three specimens at the same loading level clearly highlights that the response varies with the formulations. The difference between the rigidity of the three compositions is analyzed based on the results obtained from the strain maps. The deformation in the whole section reaches the highest value and the widest distribution for the specimen made of virgin materials. The mean deformation value of the whole section along the x-axis is 3.3 × 10−3 , 2.4 × 10−3 and 0.7 × 10−3 , and along the y-axis is −4.1 × 10−3 , −3.7 × 10−3 and − 1.3 × 10−3 for the specimen containing 0%, 30% and 50% of RAP, respectively. This means that the mixture made with virgin materials is able to deform more than the recycled mixture. Then the presence of recycled aggregate binder in the mixture increases its ability to withstand compressive stresses.
3.2.2
Behavior in the Neighborhood of Virgin and Recycled Aggregates
For a better understanding of the behavior of RAP materials, results, in the neighborhood of both virgin and recycled aggregate are compared. In the previous section, it has been shown that specimens containing a high rate of RAP, present a higher rigidity compared to the virgin specimen. To complement these findings, at the scale of aggregates, the mean deformation in the vicinity of a virgin (B) and a recycled (A) aggregates is calculated for a RAP30 (see Fig. 5). A significant difference is observed between these calculated values for A and B zones. The mean deformation along the x- and y-axis is 1.2 × 10−3 and − 5.3 × 10−3 for A and 8.6 × 10−3 and − 13.9 × 10−3 for B, respectively. This
290
A. Moukahal et al.
Fig. 4. Strain field of the microstructure of the three samples at the same load (3 kN)
Fig. 5. Value of qx superimposed to the microstructure of the sample 30% RAP
result shows that the aged binder which covers the recycled aggregates, causes lower deformation values in their vicinity compared to virgin aggregate. ε is defined. εmean is the mean Finally, a parameter q defined by the ratio: q = εmean deformation over the whole section and ε is the deformation value in the current measurement point. The parameter q allows to detect critical values that are far from the mean deformation value. In Fig. 5, one can clearly distinguish the high
Applying the Grid Method to Study the Mechanical Behavior …
291
difference between the value of q around the virgin aggregate (B) and around the recycled aggregate (A).
4 Conclusion The mechanical response of asphalt mixture with different rates of RAP has been studied using the GM. At the macroscale level, the rigidity of the material is improved by increasing the RAP rate in the mixture. At the microscale, the mean value of the deformation in the neighborhood of recycled aggregates is smaller than the one in the neighborhood of virgin aggregates. More results depending on the loading levels will be analyzed in the future.
References 1. Chowdhury, R., Apul, D., Fry, T.: A life cycle based environmental impacts assessment of construction materials used in road construction. Resour. Conserv. Recycl. 54(4), 250–255 (2010). https://doi.org/10.1016/j.resconrec.2009.08.007. 2. Xiao, F., Yao, S., Wang, J., Li, X., Amirkhanian, S.: A literature review on cold recycling technology of asphalt pavement. Constru. Build. Mater. 180, 579-604 (2018). https://doi.org/10. 1016/j.conbuildmat.2018.06.006 3. European Asphalt Pavement Association: The Asphalt Paving Industry a Global Perspective, European Asphalt Pave-ment Association (EAPA) and the National Asphalt Pavement Association (NAPA) (2011) 4. Teguedi, M.C., Blaysat, B., Toussaint, E., Moreira, S., Liandrat, S., Grédiac, M.: Applying a fullfield measurement technique for studying the local deformation in reclaimed asphalt pavements. Constr. Build. Mater. 121, 547–558 (2016). https://doi.org/10.1016/j.conbuildmat.2016.06.013 5. Teguedi, M.C.: Comportement local des enrobés recyclés: apport des mesures de champs cinématiques, PhD thesis, Université Clermont Auvergne (2018) 213 p. 6. AFNOR: Test method for hot mix asphalt. Determination of the water sensitivity of bituminous specimens. Norme francaise NF EN 12697-12 Bituminous mixtures (2008, décembre) 7. Piro, J.-L, Grediac, M.: Producing and transferring low-spatial-frequency grids for measuring displacement fields with moire and grid methods. Exp. Tech. 28(4), 23–26 (2004). https://doi. org/10.1111/j.1747-1567.2004.tb00173.x 8. Molimard, J., Surrel, Y.: Grid Method, Moiré and Deflectometry. In: Grédiac, M., Hild, F., Pineau, A. (eds.) Full-Field Measurements and Identification in Solid Mechanics, pp. 61–92. Wiley, Hoboken, NJ USA (2012)
Asphalt Aggregate Adhesion: Study of the Influence of the Morphological, Chemical and Mineralogical Properties of Different Aggregates from Southern Brazil Chaveli Brondani, Cléber Faccin, Karine W. Kraemer, Fernando D. Boeira, Luciano Pivoto Specht, and Andrea V. Nummer Abstract The proper characterization of the materials is therefore paramount for the adequate selection and use in design and execution of paving projects. One important property of the asphalt concrete is the binder-aggregate adhesion, consequence of the physical-chemical interaction between the bitumen and the aggregate surface, which keeps the elements together under traffic loads and water flow. A study was conducted to evaluate the morphological, chemical and mineralogical properties of aggregates from 12 quarries used in the production of dense asphalt mixtures in southern Brazil, as well as the interaction of the aggregates with the respective bituminous binders. The rock samples were evaluated and classified employing various methods, including Petrography, X-ray Fluorescence, Digital Image Processing (DIP) using the AIMS 2 system, traditional aggregate characterization tests, and adhesion to the bituminous binder. It was observed that acid rocks, with high Silicon content and less rough, with larger mineral size and lower AIMS values, exhibit worse performance in the visual adhesion test. The Fe/Si and Ca/K ratios were also identified as good indicators to anticipate the aggregate behavior when evaluating the adhesion to the bituminous binder. Comparative analyses based on the results from Petrography, XRF, AIMS2 and adhesion to the bituminous binder culminated in development of a scale ranking that can be used to classify the visual adhesion from terrible to excellent. Thus, it is presented as a good indicator of the behavior of the materials when evaluated the variations in the degree of adhesion allowing the correct selection of materials by the designer. Keywords Adhesion · Aggregate · Asphalt · AIMS · Petrography · XRF
C. Brondani (B) · K. W. Kraemer · F. D. Boeira · L. P. Specht · A. V. Nummer Federal University of Santa Maria/PPGEC, Santa Maria, RS, Brazil C. Faccin Centro de Ensino Superior Riograndense, Marau, RS, Brazil © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_37
293
294
C. Brondani et al.
1 Introduction and Background The adhesion between asphalt binder and aggregate is an important property for the performance of asphalt mixtures. The absence of compatibility between binder and aggregate is one of the leading causes of moisture damage in asphalt concretes. Adhesion can be assessed by visual analysis or by evaluation of the mechanical behavior of asphalt mixtures. In Brazil, the main evaluation mechanism is the empirical static immersion tests. Adhesive failure is characterized by separation of the asphalt coating from the aggregate. According to Kanitpong and Bahia [6], adhesion can be understood as the amount of energy required to break the bond between the bitumen and the aggregate. In other words, adhesion is the property that ensures the bond between the elements under the action of traffic and water. Several studies demonstrate that the properties of aggregates can strongly influence the strength of adhesive bonds [2, 3, 8, 10]. The chemical constituents, porosity, size and shape, rough-ness, surface area, and polarity are some of the main factors that can impact the adhesion between aggregate and bitumen. Therefore, adhesion is a consequence of the physicochemical interaction between the asphalt binder and the aggregate surface minerals. According to Sant’Ana [9], another factor that affects the affinity (adhesive-ness) of the binder to the aggregate is the electric charge, which is inherent to the type of matrix rock that gives rise to granulars. The balancing of electrical charges with asphalt components is hampered by higher silica contents, which make the aggregate surface more acidic, influencing the quality of the adhesive bond [11]. Acid aggregates generally exhibit poor chemical interaction with binders and consequently poor adhesion [7].
1.1 Research Objectives The experimental study presented herein evaluated the adhesion of several bitumenaggregate combinations in order to analyze the influence of morphological, chemical and mineralogical properties of different aggregates on the adhesiveness of asphalt mixtures commonly used in southern Brazil.
2 Materials and Test Methods Twelve aggregates from different regions of the state of Rio Grande do Sul, located in southern Brazil, were selected, including volcanic igneous rock, hypabyssals and plutonics, and a sample of a metamorphic rock. In order to isolate the effect of the
Asphalt Aggregate Adhesion: Study of the Influence …
295
aggregate in the analysis, the same asphalt binder, the asphalt binder CAP 50/70, was employed in all adhesion test samples. The mineralogy of the rocks evaluated was characterized by Petrography and X-ray Fluorescence techniques. The morphology was evaluated with the use of a modern image analysis system, the AIMS 2. Lastly, the adhesion was evaluated through empirical static immersion tests. Petrography is the systematic description of rock, mainly by the unaided eye and by microscopical methods. The petrographic examination of aggregates aims to describe the structure, the texture and the mineral composition of the rocks and minerals forming the various particles. The chemical composition of the rock samples was determined employing the X-ray fluorescence (XRF ED) technique with energy dispersive. To perform the test semi-quantitatively, the Bruker XRF ED S2 Ranger equipment was used in samples with approximately 40 mm in diameter and 4 mm in height, resulting in an average mass of 5 g per sample. The sample holder was inserted into the equipment and the elemental analysis option was selected in the operating system. The measurement was performed under vacuum and, at the end of the process, the equipment provided a list of the elements found in the sample. The use of Aggregate Image Measurement System 2 (AIMS2) is among the main Digital Image Processing (DIP) techniques used in the field of Transportation infrastructure to characterize rock morphology. The AIMS2 system provides angularity, sphericity, flattening characteristics, elongation, and texture information for large aggregates, while angularity and shape are provided for small aggregates, based on this work in the aggregate classification developed by Al Rousan [1]. To evaluate the adhesion between large aggregates and the asphalt binder, the static immersion adhesive test, or Brazilian visual adhesion standard test [4], was performed. The test consists of mixing 500 g of aggregate, previously cleaned and heated, with the asphalt binder. After the mixture has cooled, the container is covered with distilled water at 40 °C for 72 h. The results are deemed satisfactory if there is no bitumen film dislocation at the end of 72 h, and unstatisfactory when there is exposure of the aggregate surface.
3 Discussion of Results The rock samples were classified as basic, intermediate and acidic according to the XRF composition and its classical classification. When assessing the results from the Brazilian visual adhesion standard test, virtually all specimens presented adhesion considered unsatisfactory, but with quite different levels of unaggregated asphalt film. It is known that this test has limitations, tests using other methodologies to confirm the results would be interesting. For a better understanding of this interaction, a comparative scale was developed. This scale discretizes the qualities of the binder/aggregate coverage and is presented in Table 1.
296
C. Brondani et al.
Table 1 Adhesiveness ranking based on parameters of technological tests of mineral aggregates Rocks Classification
Sample Name
Rock name
Volcanic
P1
Basalt
Volcanic
P2
Basalt whit Olivine
Hypabyssal
P13
OlivineDiabase
Volcanic
P9
Olivine-Basalt
Basic
Volcanic
P4
Basalt
Hypabyssal
P15
Diabase
Volcanic
P14
Andesite
Volcanic
P6/P7/P8
Dacite
Intermediate
Acid
Volcanic
P3
Rhyodacite
Plutonic
P10
Syenogranite
Plutonic
P5
Syenogranite
Metamorphic
P11/P12
Marble
Texture
Texture AIMS¹
Phaneritic inequigranular to very fine-grained porphyritic Phaneritic inequigranular fine to medium grained porphyritic Phaneritic inequigranular to finegrained porphyritic Phaneritic inequigranular to finegrained porphyritic Afanitic to very fine phaneritic Phaneritic inequigranular to coarse porphyritic Phaneritic inequigranular coarse to very coarse porphyritic Very fine
Visual Adhesion
Si
Ca
K
Ratio Fe/Si
Ratio Ca/K
38,2
25,05
20,39
2,47
1,52
8,26
Excellent
41,26
24,91
17,13
3,06
1,66
5,60
Very Good
34,63
29,68
18
3,52
1,17
5,11
Very Good
624-704
36,75
27,58
22,07
4,42
1,33
4,99
Good
465-561
29,86
33,86
21,17
2,26
0,88
9,37
Good
428-532
33,87
27,11
21,36
2,43
1,25
8,79
Good
701
33,34
30,01
16,5
6,6
1,11
2,50
Extremely poor
561-465
26,32
37,91
11,72
10,42
0,69
1,12
Extremely poor
Phaneritic inequigranular to fine- 679-724 grained porphyritic Phaneritic inequigranular fine to 737-774 very fine grained porphyritic Phaneritic inequigranular very fine 661-653 to fine grained porphyritic Phaneritic inequigranular to very fine-grained porphyritic
Composition XRF (%) Fe
CAP 50/70
349-369
24,22
38,44
9,48
13,99
0,63
0,68
Poor
253-240
13,06
48,17
7
20,28
0,27
0,35
Poor
196-168
10,84
50,78
4,14
22,1
0,21
0,19
Very Poor
485-470
5
3,09
80,08
0,62
1,62
129,16
Good
As it can be observed in Table 1, the basic rocks, such as P1, P2, P9 and P4, initially presented a more favorable result, although the visual adhesiveness standard test considered their adhesiveness unsatisfactory. At the same time, the intermediate to acidic rocks, such as P3, P10, P5 and P14, exhibited less favorable and equally unsatisfactory behavior in terms of aggregate coverage by the binder. One of the factors that explains this situation is related to the silica content of the rock. The more acidic the rock and the larger the size of its minerals, as in the case of P5 and P10 syenogranites, the worse the adhesiveness. The quarry P11/P12 is displaced from Table 1 because it is classified as metamorphic rock, being presented for comparison only, once the acidity classification applies only to igneous rocks. Based on the petrography and FRX results, silica levels above 38%, with presence of quartz minerals in the sample, were understood by this study as a limiter that causes the aggregate-binder bonds to occur less effectively, resulting in low adhesion. The results of the main constituent elements of the rock obtained with the FRX test indicated samples with higher adhesiveness to have a high iron (Fe) content, as well as an iron to silica (Fe/Si) ratio greater than 1.0. In the samples evaluated, higher Ca contents (above 17%) and a Ca/K ratio above 5.0 indicated better adhesion, as these are values found in basic rocks (P1, P2, P9, P4, P13, and P15). The results indicate that the level of adhesion may also be influenced by the texture and the degree of alteration of the rock minerals. Fine textured rocks, described by petrography as Phaneritic very fine to fine and all microporphytic (samples P1, P2, P9 and P4), exhibited AIMS values in the range of 465–774 (thus classified as very
Asphalt Aggregate Adhesion: Study of the Influence …
297
rough) showed better adhesiveness. Rocks with coarser textures, such as Phaneritics medium to coarse phaneritics (syenogranites), presented lower AIMS values (between 253 and 168), were classified as soft and exhibited worse adhesiveness. Fine rocks that present altered minerals or altering edges exhibited lower AIMS values. One example is the P3 (AIMS around 369), which has a fine to very fine texture, but has altered minerals and zeolites filling the tonsils. Furthermore, it was found that roughness improves adhesion. Roughness tends to be higher in rocks with very fine to fine texture, and lower (soft) for rocks with coarse or fine texture that have altered minerals. According to Frazão [5], the mineralogical alterations lead to the removal of the bases, which makes the surface of the aggregate acidic even if it is of a basic character. Thus, adhesion is expected to be worse in more altered rocks (as observed in petrographic analysis), such as rocks P15, P14, P10 and P5, for example. Figure 1 illustrates the drop in ranking of visual adhesion CAP 50/70, represented by the bars, that occurs in the direction of basic-to-acid rocks. Most trend lines, which represent the experimental data values, also decrease in the same direction, which is the case of the Fe and Ca lines. In contrast, the Si and K trend lines exhibit opposite behavior, with increasing values in the same direction. It can also be observed that quarries with basic aggregates are located to the left of the graph, which is due to the XRF composition percentages. The intermediate aggregates should have presented better ranked visual adhesion when compared to the acidic aggregates, but this did not occur in this research. One possible explanation for the worse visual adhesion results of the P14 and P6/P7/P8 quarries is the degree of surface alterability of the aggregates, and not their chemical composition directly. In the quarries classified as acidic, at the rightmost of the graph, the visual adhesiveness ranking starts to decrease again as the percentage of Si increases.
Fig. 1 Relationship between FRX composition, rock acidity and visual adhesion classification CAP 50/70
298
C. Brondani et al.
The results from the various tests presented here indicate that the variations in the degree of adhesiveness are mostly influenced by the type of rock, due to its varied mineralogical compositions that directly influence the bond between bitumen and aggregate by means of the electrostatic attraction forces.
4 Concluding Remarks A comparative analysis of the results from petrographic slide, XRF and bituminous binder adhesion resulted in the development of a ranking for visual adhesiveness, which ranges from excellent to extremely poor. The strong correlation between the silica content (Si) and the adhesive failure was evident in the results. Rocks with high Si content were classified as intermediate and acidic, presenting worse performance in the visual adhesiveness test. When considering also the chemical relation, a high Fe/Si ratio (above 1.0), calcium values above 20%, and Ca/K ratio above 5.0 considerably improved the coverage of the film that covers the aggregate in the adhesion test. Therefore, these relationships can be good indicators to predict aggregate behavior when assessing adherence to the bituminous binder, due to the crushing process providing the characterization of the aggregates, making it easier for the designer to choose the materials to compose the asphalt mixture, thus being able to identify aggregates that may present problems and consequently optimize the project on the Southern of Brazil to these studied sampling. A significant influence of the aggregate texture was also observed, where rougher aggregates exhibited less adhesive flaws.
References 1. Al Rousan, T.M.: Characterization of aggregate shape properties using a computer automated system, p. 229f. Ph.D. Thesis—Texas A&M University, College Station, TX (2004) 2. Cui, S., Blackman, B.R., Kinloch, A.J., Taylorn, A.C.: Durability of asphalt references 53 mixtures: effect of aggregate type and adhesion promoters. Int. J. Adhes. Adhes. 100–111 (2014) 3. Curtis, C.W., Ensley, K., Epps, J.: Fundamental properties of asphalt-aggregate interactions including adhesion and absorption. Strategic Highway Research Program, National Research Council, Washington D.C. (1993) 4. DNER–ME 078–Brazilian visual adhesion standard test. https://ipr.dnit.gov.br/normas-e-man uais/normas/meetodo-de-ensaio-me/dner-me078-94.pdf (1994) 5. Frazão, E.B.: Tecnologia de rochas na construção civil. Associação Brasileira de Geologia de Engenharia Ambiental. São Paulo–SP (2002) 6. Kanitpong, K., Bahia, H.U.: Evaluation of HMA moisture damage in Wisconsin as it relates to pavement performance. Int. J. Pavement Eng. 9(1), 9–17 (2008) 7. Liberatori, L.A., Constantino, R.S.: Melhoradores de adesividade para misturas asfálticasEstado da Arte. 18° Encontro de Asfalto-Rio de Janeiro RJ (2006) 8. Nageswaran, P.D.C.: Adhesion of aggregate-binder systems. M.Sc. Thesis. Delft University of Technology (2016)
Asphalt Aggregate Adhesion: Study of the Influence …
299
9. Sant’Ana, W.C.: Estudo de Misturas de Areia-Asfalto a Quente para o Estado do Maranhão. Dissertação de Mestrado. Escola Politécnica da Universidade de São Paulo (EPUSP). São Paulo-SP (1992) 10. Tarrer, A., Wagh, V.: The effect of physical and chemical characteristics of the aggregate on bonding. Strategic Highway Research Program; National Research Council, Washington D.C. (1991) 11. Yoon, H.H., Tarrer, A.R.: Effect of aggregate properties on stripping. Transp. Res. Rec. 1171, pp. 37–43. Washington, D.C. (1988)
Asphalt Layers Thickness Determination for Top Down Cracking Initiation and a Case Study A. Nikolaidis and E. Manthos
Abstract Top-down cracking (TDC) is a deterioration mechanism where cracks initiate at the pavement surface and propagate downwards with time. It is caused by horizontal tensile strains generated at the tire-pavement contact area. This paper presents a methodology for determining the thickness of the asphalt layer above which TDC will be the predominant deterioration mechanism in comparison to the bottom up cracking. The methodology covers two scenarios: one when all asphalt layers have the same stiffness modulus and the other when the asphalt layers have two different stiffness moduli. i.e. stiffness of surface layer (40 mm thick), S1 , and stiffness of the rest of asphalt layers, S2 . The findings concluded that in flexible and semi-flexible pavements with uniform asphalt stiffness layers, when the thickness of the asphalt layers is above a value within the range 100–210 mm (for ESAL80 axle loading), the pavement will most probably fail due to top-down cracking and not due to bottom up cracking. The exact thickness above which TDC will occur depends on the magnitude of Foundation Surface Modulus (FSM) and the asphalt stiffness. Similar range of thicknesses was found (100–230 mm) above which TDC will occur in flexible pavements with non-uniform asphalt stiffness values (S1 /S2 is up to 1.13). A case study along a motorway section of 16 km long confirms the findings of the methodology developed. Keywords Top down cracking · Flexible and semi-flexible pavements
1 Introduction Cracking is one of the major pavement distress mechanisms. Cracking initiation and propagation characterize the pavement fatigue life. Almost all of the past and current flexible design methodologies determine pavement fatigue life considering the horizontal tensile strain or stress at the bottom of the asphalt layers. The above determination is made under the assumption that the maximum tensile strain occurs A. Nikolaidis · E. Manthos (B) Aristotle University of Thessaloniki, Thessaloniki, Greece e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_38
301
302
A. Nikolaidis and E. Manthos
at the bottom of the asphalt layers and hence cracking initiates and propagates from asphalt layer bottom towards the pavement surface. However, research and practice of the last decades has shown that cracking can also initiate from the pavement surface or near the pavement surface and propagate towards the bottom of the asphalt layers. This type of cracking is notified as top down cracking (TDC). The factors affecting the magnitude of the horizontal tensile strain causing TDC mainly are the: stiffness and thickness of the asphalt layers, applied axle load and tire configuration, thickness and material of base/sub-base layers, and bearing capacity of the subgrade. The tire structure also has a variable detrimental effect, certainly in the case of surface strains.
1.1 Top Down Cracking Identification and Studies An excellent study on top down cracking identification worldwide was made by the Mississippi Department of Transportation and The University of Mississippi [1]. According to the above study the Transportation Research Laboratory (TRL) was the first to identify the TDC in the late 1970s. In the 1980s TDC was also identified in South Africa [2], The Netherlands [3] and France [4]. In 1992 a detailed field investigation and numerical study on top down cracking was conducted in Japan [5]. In USA TDC was observed in the late 1990s [6]. A fundamental study for the mechanism of TDC was carried out and reported in 1998 and 2001 [7, 8]. Since then researchers in other countries have also studied and analyzed TDC failures and identified the causes of TDC. An extensive reference list can be found in a paper published in 2017 [9].
1.2 Scope of the Current Study The scope of the current study was to determine the thickness of the asphalt layer above which top-down cracking is the dominant pavement deterioration mechanism in comparison to bottom-up cracking. The results obtained were applied to a case study where almost exclusively top down study had been occurred.
2 Methodology Followed In order to determine whether top-down cracking (TDC) or bottom-up cracking (BUC) is the dominant pavement deterioration mechanism, the load induced tensile strains at the surface of the pavement and at the bottom of the asphalt layer were determined and compared.
Asphalt Layers Thickness Determination for Top Down Cracking Initiation …
303
The determination of the tensile stains was carried out with the use of BISAR 3.0 software assuming an elastic multi-layer system. For this study a 2-layer system and a 3-layer system was selected. In the first case the two distinct layers were the asphalt layer and the foundation layer (unbound layer and sub-grade combined); in the 3-layer system the distinct layers were the wearing course, the binder/asphalt base layer and the foundation layer. The software inputs needed for calculating the relevant strains were: the axle loads and tire structure, the positions at which the strains will be calculated, the asphalt layer thickness and the stiffness of the asphalt layer, and the stiffness of the foundation layer expressed in Foundation Surface Modulus (FSM). One type of axle load was considered, that which corresponds to the most common standard loading adopted by many countries and pavement design methodologies, the equivalent standard axle load (ESAL), a single axle with dual tires and axle load of 80 kN (ESAL80 ) (radius of contact area 0.105 m and smooth tire surface were considered). In a previous study [9] another three types of axle loading were also considered. The horizontal stresses and strains, in all cases, were determined at the surface (z = 0), at the bottom of the asphalt layer (z = t, where t = total variable thickness of asphalt layer), for the 2-layer system and also at the interface between wearing course and binder course (z = 40 mm), in the case of 3-layer system. The strains were determined at three positions: the axis of symmetry of the loading areas of the one set of the dual tires (i.e. middle point between one set of dual tires), at the edge of the one circular loading area and at the center of one of the of the loading area (tire). For the 2-layer system, the minimum thickness of the asphalt layer was selected to be 100 mm with uniform stiffness. The stiffness (modulus of elasticity) values were considered to range from 2.000 to 12.000 MPa which covers very low, medium, or high stiffness asphalts. For the 3-layer system, the asphalt layers were considered to be a wearing course layer of 40 mm thick and an asphalt binder/asphalt base layer of variable thickness. The stiffness of the wearing course layer varied from 3.500 to 8500 MPa and the rest of the layers from 4.000 to 12.000 MPa. The unbound layers and the subgrade were considered as one layer called the foundation layer, which stiffness was represented by the Foundation Surface Modulus (FSM). The principle of FSM was adopted from the British design methodology. The FSM considered in this study ranged from 50 to 400 MPa. This range covers the flexible pavements (Foundation classes 1 and 2) and semi-flexible pavements (Foundation classes 3 and 4). More details about the foundation classes can be found in refs. [10, 11].
304
A. Nikolaidis and E. Manthos
Table 1 Asphalt thickness above which TDC develops, 2-layer system (ESAL80 ) Foundation class Asphalt
Class 1: ≥ 50 Mpa
stiffness, MPa
Class 2: ≥ 100 Mpa
Class 3: ≥ 200 Mpa
Class 4: ≥ 400 Mpa
Foundation surface modulus (FSM), MPa 50
70
90
100
150
190
200
300
390
400
Asphalt layer thickness above which TDC is expected to develop, mm (For dual tire load of 40 kN) 2000
171
a
144
b
4000
193
185
100
100
176
172
c
100
100
100
100
100
100
100
100
100
100
100
100
c
156
d
153
100
100
100
100
100
100
100
100
100
100
f
6000
200
196
190
187
172
8000
206
200
195
193
182
173
e
171
10000
209
204
200
198
189
182
180
160
12000
211
207
203
201
193
187
186
171
a
g a
Initiation of TDC also at thickness: a ≤ 102 mm. b ≤ 131 mm. c ≤ 103 mm. d ≤ 117 mm. e ≤ 101 mm. f ≤ 123 mm. g ≤ 112 mm
3 Results After extensive number of calculations with the BISAR software (runs with increments of 1 mm of asphalt layer thickness) the results obtained have been tabulated and are as shown in Tables 1 and 2. The criterion for determining the TDC development was the value of horizontal strain developed at the surface of the pavement and at the bottom of the asphalt layer(s). When the absolute value of the strain at the surface was greater than the absolute value of the strain at the bottom, and this remained so for any further increase of asphalt layer thickness, the pavement was assumed to fail due to top-down fatigue cracking (TDC) rather than due to the bottom-up fatigue cracking. Table 1 shows the thickness of all asphalt layers with uniform stiffness, per foundation class and asphalt layer stiffness, above which top down cracking (TDC) will be the only pavement deterioration mechanism and the pavement will fail due to top down cracking and not due to bottom up cracking. Table 2 shows the total thickness values of the asphalt layers (two layers with different stiffness values) above which TDC is dominant.
4 Case Study A motorway section, 16 km long, opened to traffic in 2008 and in 2014 started to show surface cracks (longitudinal, with or without branches, block-patterned cracks, short diagonal cracks and in few cases fatigue cracks-crocodile cracks). The pavement constructed was flexible with a theoretical 22 cm total asphalt layer thickness (4 cm wearing course with 10% air voids and 18 cm with dense asphalt
Asphalt Layers Thickness Determination for Top Down Cracking Initiation …
305
Table 2 Asphalt thickness above which TDC develops, 3-layer system (ESAL80 )
Asphalt stiffness ratio, S1/S2
Foundation class Class 1: ≥ 50 MPa Class 2: ≥ 100 MPa Foundation surface modulus (FSM). MPa 50
70
90
100
150
190
Thickness above which TDC is expected. mm 0.29-0.92 0.94
100 168
100 a
b
160
100
100
100
100
100
100
100
100
1.06
226
221
216
213
203
196
1.08
227
220
214
212
200
192
1.13
229
221
213
210
196
185
1.25
BU
BU
BU
BU
232
224
1.38
BU
BU
BU
BU
235
225
BU
BU
BU
BU
BU
BU
1.42-2.13
a = Also for thickness ≤ 127 mm initiation of TDC, b = Also for thickness ≤ 139 mm initiation of TDC, BU = Bottom up cracking only
concrete with 4% air voids). The base/subbase course was 40 cm thick with unbound crushed material rested over a small or medium height embankment with selective soil material. After taking a number of cores at representative locations (30 on the west branch and 19 on the east branch) the type and length of the crack (TDC or BUC) was determined. Top down cracking was observed in forty-nine (49) out of fifty-two (52) cores taken. The results are as shown in Table 3. In three cases (cores) straight through cracks were observed due to small thickness of asphalt layers (less than 16 cm) and low FSM value. Further to coring, the stiffness of the two type of asphalts used was also determined as well as their indirect tensile strength (ITS) at 25 °C. The average stiffness value of the asphalt used in the wearing course, S1, was found to be 5138 MPa and for the asphalt in the rest of the layers, was found to be 4701 MPa, i.e. ratio S1 /S2 = 1.09. From Table 3 it can be seen that TDC developed at an average thickness of asphalt layers of 19.5 cm or 20.4 cm, west branch or east branch direction, respectively. For a Table 3 Results out of coring (average values) West branch
East branch
Thickness (cm) Asphalt layers 19.5
TDC depth (cm) Top layer 3.9
Thickness (cm) Asphalt layers
6.8
20.4
TDC depth (cm) Top layer 4.0
8.7
306
A. Nikolaidis and E. Manthos
an S1 /S2 ratio found equal to 1.09 and assuming that the FSM of the flexible pavement was within 150 MPa to 190 MPa the findings comply well with the results of Table 2.
5 Conclusions Based on the results obtained the following conclusions can be drawn: 1.
2.
3.
4.
5.
6.
7. 8. 9.
The tables derived provide an easy tool to determine the asphalt layer(s) thickness above which TDC and not BUC is the dominant tensile fatigue failure mechanism on flexible (or semi-flexible pavements). The thickness of the asphalt layer above which TDC becomes dominant depends on the stiffness of the foundation layer, expressed as foundation surface modulus (FSM), and the stiffness of the asphalt, hence of the asphalt layer. For a 2-layer system regardless of the pavement type (flexible or semi-flexible), for the same FSM, as the stiffness of the asphalt layer increases, the thickness of the asphalt layer above which TDC is dominant also increases. Similarly, regardless of the pavement type (flexible or semi-flexible), for the same asphalt stiffness, as the stiffness of the foundation layer increases, the thickness of the asphalt layer above which TDC is dominant decreases. When the flexible pavement consists of a wearing course layer of 40 mm having lower asphalt stiffness than the stiffness of the asphalt layers underneath (S1 /S2 < 1.00), a 3-layer system, TDC is more likely to occur in almost all cases. For the same pavement structure, when the ratio S1 /S2 is between 1.06 and 1.13, TDC is dominant when the total asphalt layer thickness is above a value between the range of 185–229 mm depending of the FSM value. The findings of the methodology developed with respect to TDC development comply well with the results obtained from site. Top down cracking in comparison to bottom up cracking can be detected almost at its initiation; hence it is much easier, quicker and less costly to be repaired. As a consequence, to the above statement, pavements with thick asphalt layers should always be encouraged to be designed.
References 1. Uddin, W.: A synthesis study of noncontact nondestructive evaluation of top-down cracking in asphalt pavements, Mississippi Department of Transportation, Mississippi, USA (2013) 2. Hugo, F., Kennedy, T.W.: Surface cracking of asphalt mixtures in southern Africa. In: Proceedings of the Association of Asphalt Paving Technologists, vol. 54, pp. 454–501 (1985) 3. Pronk, A.C., Buiter, R.: Aspects of the interpretation and evaluation of falling weight deflection (FWD) measurements. In: Fifth International Conference on the Structural Design of Asphalt Pavements, vol. 1, pp. 461–474. Delft, The Netherlands (1982)
Asphalt Layers Thickness Determination for Top Down Cracking Initiation …
307
4. Dauzats, M., Rampal, A.: Mechanism of surface cracking on wearing courses. In: Sixth International Conference on Structural Design of Asphalt Pavements, vol. 1, pp. 231–247. Michigan, USA (1987) 5. Matsuno, S., Nishizawa, T.: Mechanism of longitudinal surface cracking in asphalt pavements. In: Seventh International Conference on Asphalt Pavements, vol. 2, pp. 277–291. Nottingham, UK (1992) 6. Uhlmeyer, J.S., Willoughby, K., Pierce, L.M., Mahoney, J.P.: Top-down cracking in Washington State asphalt concrete wearing course. J. Transp. Res. Board 1730, 110–116 (2000). (Transport Research Record, USA) 7. Myers, L.A., Roque, R., Ruth, B.E.: Mechanisms of surface-initiated longitudinal wheel path cracks in high-type bituminous pavements. J. Assoc. Asphalt Paving Technol. 67, 401–432 (1998) 8. Myers, L.A., Roque, R., Birgisson, B.: Propagation mechanisms for surface-initiated longitudinal wheelpath cracks. J. Transp. Res. Board 1778, 113–122 (2001). (Transport Research Record, USA) 9. Nikolaides, A., Manthos, E.: Determination of asphalt layer thickness above which axle loadinduced strains initiate top-down cracking. In: 10th International Conference on Bearing Capacity of Roads, Railways and Airfields (BCRRA), pp. 1457–1464. CRC Press, Athens, Greece (2017) 10. Highways England: Design Manual for Roads and Bridges, vol. 7, Pavement design and maintenance, Section 2, Pavement design and Construction, Part 3, HD 26/06, Pavement Design, Highways England, London (2006) 11. Highways England: Design Manual for Roads and Bridges, vol. 7. Pavement design and maintenance, Section 5, Pavement design and Construction, Part 2, IAN 73/06. Design Guidance for Road Pavement Foundation, Highways England, London (2009)
Asphalt Permeability In-service: How Can It Be Evaluated Before and After Moisture Damage? C. Venter, K. J. Jenkins, C. E. Rudman, and O. R. Grobbelaar
Abstract The first bituminous road surface was purportedly placed in Paris, France, more than 150 years ago. The primary purpose of a surfacing layer was to protect the underlying pavement from moisture ingress, confirming the innate cognisance of waterproofing and low permeability to enhance pavement performance. Over the decades various methods have been implemented to evaluate asphalt permeability, including volumetric properties e.g. voids in mix, field testing e.g. using a falling head permeameter and laboratory tests on specimens or cores e.g. air or water permeability (constant or falling head). The evaluation methods each have elements of empiricism e.g. overlooking lateral versus vertical permeability or underestimating hydrostatic pressures in-service. This paper describes the development of a laboratory permeameter that evaluates vertical permeability at high pressures i.e. representative of traffic induced hydraulic pressures. The apparatus, termed “High Pressure Permeameter” HPP, is able to sustain pressures above 200 kPa. The research is based on specimens retrieved from surfacing seals and asphalt. Cores were acquired from different sources all over South Africa, including different designs and climates. Testing includes laboratory Marvil (low pressure) permeability and HPP tests at 100, 150 and 200 kPa, correlated with volumetric mix properties and CT scans. In addition, moisture damage using the Moisture Inducing Simulating Test (MIST), induces representative hydrostatic pressures over sustained periods. Findings reveal inconsistent correlation between Marvil and HPP asphalt permeability values at 100 and 150 kPa testing pressure; however, at 200 kPa higher variability is apparent. The range of mixes that have been evaluated allows for preliminary guideline limits for HPP permeability method to be developed, which can pre-empt identification of mixes with unacceptably high permeability. Keywords Asphalt permeability · High pressure permeameter · Inter-connected voids · Moisture conditioning · MIST test
C. Venter · K. J. Jenkins (B) · C. E. Rudman · O. R. Grobbelaar Stellenbosch University, Stellenbosch, South Africa e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_39
309
310
C. Venter et al.
1 Background 1.1 Asphalt Performance Properties The need for durable, functional and waterproof surfacing layers on roads was recognized in the mid-nineteenth century in France. In order to meet this objective, asphalt mixtures of various compositions were trialed. Subsequently, mix designs in many different countries were empirically selected based on in-service performance. It was only in the twentieth century that volumetric and engineering properties became part of the mix evaluation process. Construction of the asphalt remains paramount in the quality assurance of the asphalt in the chain of events. This includes homogenous distribution of the mix and its temperature, as well as adequate compaction. This will assist in reducing the permeability of HMA pavements and in turn risk management of moisture damage. Anti-stripping additives can also be added to the HMA mixture to improve the moisture resistance of the pavement, as reported by NCHRP [1].
1.2 Routine Testing for Moisture Damage High durability pavements require, amongst other attributes, low permeability asphalt layers. This also applies to seal surfacings. Moisture ingress into the base and other supporting layers can lead to cracking, rutting, increased deflections and a reduction in the carrying capacity of the pavement [2]. This highlights the need for routine evaluation of an asphalt mix’s resistance to moisture damage. There are two important components to the risk of moisture ingress: (a) (b)
Air voids: in particular inter-connected voids, and Moisture damage: related to stripping of the binder from the mix leading to increased permeability and loss in the integrity of the layer.
Representative evaluation techniques of moisture resistance in a laboratory require both moisture damage simulation and permeability testing at pore pressures representative of in-service conditions. Twagira [3] evaluated the hydrodynamic phenomenon as investigated by Land Transport New Zealand [4] and Nakajima et al. [5] The main finding of this study was that at traffic speeds of 70–120 km/h on a wet surface, the tyres generate hydraulic pressures of between 100 and 200 kPa in the surfacing layer. This was first analysed with chip seal surfacings.
Asphalt Permeability In-service: How Can It Be Evaluated Before …
311
2 Development of Test Procedures Simple permeability testing apparatus such as the Marvil permeameter i.e. falling head method at pressures below 2 kPa, is commonly used in southern Africa and elsewhere to determine the permeability of an asphalt layer in situ [9]. The test can also be adapted for laboratory testing on cores from the road. Significant variability of the results occurs however, based on leakages, lateral migration of water and local defects etc. resulting in poor repeatability. The low pressures at which the Marvil test is undertaken, do not provide insight into the interconnectivity of the voids, however. An apparatus that is capable of testing high pressure permeability i.e. up to 200 kPa was designed and developed by Grobbelaar [6] to explore the effective void structure in surfacing mixes. The apparatus is termed the High Pressure Permeameter HPP and was initially aimed at evaluating the permeability of chip seals but was expanded to include asphalt surfacings by Venter [7]. Conditioning of surfacing mixes to simulate wet trafficking in a summer rainfall region i.e. worst case scenario, could be effected using the Moisture Induced Sensitivity Test MIST apparatus developed by Twagira [3]. In order to introduce damage, the simulation is performed with water at 60 °C pulsed at 140 kPa and approximately 0.5 Hz (0.54 s load and 1.4 s unload) in a triaxial cell. The duration of the conditioning is 6 h and 2 min for asphalt and seals, which is significantly longer than the conditioning of Bitumen Stabilised Materials BSMs of less than 10 min.
3 Experimental Methods and Findings 3.1 Experimental Strategy As part of an investigation of premature failures of a range of asphalt mixes in South Africa, Venter [7] tested a range of cores extracted from the highways. In-service pavements were cored in areas with surfacings in good conditions and others in poor conditions. The cores were evaluated in terms of volumetric properties, Marvil permeability [8], HPP permeability (before and after MIST conditioning). In addition, CT scans were carried out on selected cores before and after MIST conditioning to evaluate possible changes in the void structure of the mixes.
3.2 Research Findings Two mixes are selected for analysis here, namely:
312
C. Venter et al.
Fig. 1 High pressure permeability pre-MIST of Mixes Cd and Cp (3 repeats per bar)
• Cd—surfacing mix with 19 mm maximum size and bitumen rubber binder (approximately 6.3%) and 9.5% average voids. • Cp—surfacing mix 19 mm maximum size, including steel slag and bitumen rubber binder (approximately 6.7%) and 4.0% average voids. As a benchmark, the average Marvil permeability for mix Cd was 66.2 ml/min with a range of 10th percentile = 10 ml/min and 90th percentile = 145 ml/min. Mix Cp yielded zero Marvil permeability. By comparison, the results of the HPP permeability are provided in Fig. 1. Several observations are of interest i.e. the rate of increase in permeability with pressure is significant and mixes with low permeability (Cp) yield to moisture ingress at a representative traffic-induced pressure (150 kPa). The in-service condition of the mixes correlates strongly with the permeability results i.e. Mix Cd is prone to moisture ingress and damage whereas Mix Cp is not. At test pressures of 100 and 150 kPa, there is evidence of damage to the asphalt with MIST conditioning of the cores defined by the increase in permeability, see Fig. 2. At a test pressure of 200 kPa a reduction in permeability is measured, which is counter-intuitive. This could be a result of binder migration at 60 °C and high pressure. In addition, the greater project showed the 200 kPa pressure measurements to be erratic without MIST conditioning, which could possibly be attributed working at the upper limits of the induced pore pressures in the mixes. Verification of the damage induced by hydrodynamic pulsing of water in the MIST test can be visually identified by comparing CT scans of the same specimen before and after MIST conditioning. Under pressure, flow paths are accessed, especially with interconnected voids in the mix. The active flow paths lead to erosion of the binder i.e. the stripping mechanism, apparent in the outer 10 mm of the core perimeter (Fig. 3).
Asphalt Permeability In-service: How Can It Be Evaluated Before …
313
Fig. 2 High pressure permeability pre- and post-MIST of mixes Cd (3 reps ea.)
Fig. 3 CT scans of Mix Cd Core 1 Pre-MIST and Post-MIST
4 Implementation of Outcomes Based on the findings presented here and in the greater research project [7] into asphalt permeability several outcomes are pertinent, some of which can be implemented in practice:
314 Table 1 Preliminary HPP permeability classification ranges for asphalt surfacings at 100 or 150 kPa pressure, without MIST conditioning
C. Venter et al. 100 kPa
150 kPa
Good
38.00 ml/min
• Permeability tests carried out at low pressures ( 5
(1) (2)
The modulus of binder course has been considered in this analytical study as 1700 MPa as recommended in IRC: 37-2018 [2] for use of Bituminous concrete (BC) and Dense Bituminous Macadam (DBM) with VG 30 bitumen at 35 °C. The elastic modulus of unbound granular layer has been considered as 335 MPa which corresponds to 100% CBR.
3.1 Odemark’s Method for Transformation of Layered System According to Odemark, the two layered system as shown in Fig. 1 may be transformed into a homogeneous system with an elastic modulus (E 2 ), where the equivalent thickness of the transformed section has been expressed by Eq. (3).
Characterization of Layer Index in Bituminous Pavement …
349
Fig. 1 Transformation of a two layered system in to a homogeneous system by Odemark’s method
Therefore, equivalent thickness of the homogenous layer he = f h1
3
E1 E2
(3)
where, ‘ f ’ is a correction factor usually ranges between 0.8 and 1.0.
3.2 Determination of Concentration Factor Using Vertical Interface Stress To find out the interface deflection, the concept of concentration factor has been used in present work. It has been found that the vertical stress distribution in homogeneous elastic, isotropic medium is independent of the elastic modulus of the system. However, in case of layered system, the vertical stress distribution is significantly influenced by the elastic modulus of the respective layers. In this backdrop, the vertical stress at a depth (Z ) in a two layered system with a surface load intensity (q) acting on a flexible circular loaded area with radius (a) can be expressed by Boussinesq’s theory as given in Eq. (4). σZ = q 1 −
Z
n
√ a2 + Z 2
(4)
where, the term ‘n’ is defined as concentration factor which depends on modular ratio, which is the ratio of elastic modulus of top layer (E 1 ) and bottom layer (E 2 ). The correlation between concentration factor and modular ratio has been established [3] and presented in Table 1 and Eq. (5). The maximum depth of the pavement has been considered as three times the width of the contact radius of wheel load to establish the correlation and has been used in this paper.
E1 n = 3.052 E2
−0.660
; Regression Co - efficient r 2 = 0.985
(5)
350
S. Purakayastha et al.
Table 1 Variation of concentration factor with modular ratio Modular ratio (E 1 /E 2 )
1
2.5
5
10
25
50
100
Concentration factor (n)
3.0
1.667
1.062
0.6735
0.3676
0.2321
0.1464
Modular ratio (E 1 /E 2 )
200
300
400
500
Concentration factor (n)
0.0923
0.0705
0.0582
0.0501
4 Determination of Pavement Surface Deflection In order to determine the vertical interface deflection, in a two layered system, the vertical interface stress (σ Z 0 ) on top of granular layer along the center line of wheel load may be obtained using Eq. (6) and the vertical stress (σ Zr ) at different radial distance (r ) from the load at interface level may be obtained using Eq. (7). σZ 0 =
Pn 2π Z 2
σ Zr = σ Z 0
1+
1 r 2
(6)
n2 +1 (7)
Z
where, Z = h e = Equivalent thickness in a two layered system. The modulus of binder layer (E 1 ) and granular layer (E 2 ) were considered for estimation of interface deflection on the top of the granular base. Similarly, the modulus of granular layer (E 2 ) and subgrade (E 3 ) were considered in the analysis for estimation of deflection on the top of subgrade. The summation of these deflections has been considered as pavement deflection in this analysis. The interface vertical stress transfer takes place from a stiffer layer (E 1 ) to a softer layer with an elastic modulus of (E 2 ) in a two layered system as shown in Fig. 2. It is to be noted that the stress dispersion angle (θ ) reduces with the increase in modular ratio in a two layered system and vice-versa. Therefore, the radius dispersed area (as ) of stress at the layer interface may be obtained from Eq. (8). From Fig. 2, πa 2 × q = πasi2 × σ Zri Fig. 2 Stress dispersion in a two layered system
Characterization of Layer Index in Bituminous Pavement …
351
where, ‘q’ is the surface stress intensity acting on a circular plate of radius ‘a’ on the pavement surface. Or, asi = a ×
q σ Zri
(8)
Therefore, the deflection on the top of interface unbound granular layer at a depth (Z ) at different radial distances from the center of load may be determined by Eq. (9). dsri =
1.5 σ Zri asi E i+1
(9)
where, σ Zri = Vertical stress acting at the interface of ith and i + 1th layerE i+1 = Elastic modulus of i + 1th layerasi = Radius of stress dispersed area at the interface of consecutive layer of pavement, which initiates the change in stress intensity from the ith layer to the i + 1th layer. The principle followed for determination of interface deflection at binder-granular layer interface has also been used for estimation of deflection at subgrade-granular base interface. The pavement deflections at the center of the loading plate and at a radial distance of 300–1800 mm at an interval of 300 mm were determined analytically for its comparison with other relevant analytical and experimental work.
5 Characterization of Deflection Bowl In general, detailed analysis of deflection bowl is required to evaluate probable extent of damage and distress from the condition monitoring of road pavement by FWD. In this backdrop, the deflection bowl parameters like Base Layer Index (BLI), Middle Layer Index (MLI) and Lower Layer Index (LLI), which were suggested by Horak [4] (Eqs. 10–12) have been used in present work. where, BLI = D0 −D300
(10)
MLI = D300 −D600
(11)
LLI = D600 −D900
(12)
In which, D0 , D300 , D600 , and D900 are the deflection at sensor offset 0, 300, 600 and 900 mm respectively from the load. It has been found that the BLI, MLI, and LLI have good correlations with the structural behavior of a multilayered flexible pavement.
352
S. Purakayastha et al.
From FWD output data, BLI and MLI are generally considered as the indicator of strength of bituminous layer and granular layer respectively, whereas LLI indicates the subgrade strength. The indication of fatigue failure may be represented by the increase in BLI value, whereas the LLI value indicates the probability of failure of pavement under rutting. In the present analysis, Poisson’s ratio for bituminous layer, granular layer, and soil subgrade were considered as 0.5, 0.4, and 0.4 respectively.
6 Discussion on Test Results In the present paper, BLI, MLI and LLI, which were obtained using present analytical approach were compared with field data analysis based on the data obtained on National Highway (NH-6) by Reddy et al. [5] in Table 2. It is relevant to note from Table 2 that, the BLI value has been found to get influenced with thickness and elastic modulus of binder course. Moreover, thickness of granular base also affects the BLI. However, the BLI value increases in a pavement section with thinner binder course and comparatively smaller modulus ranging between 650 and 700 MPa. But, the BLI was found to be less when the binder course was with higher elastic modulus ranging between 1500 and 2200 MPa. However, extensive study relating effect of thickness and modulus on layer index, in a multilayer system needs to be done carefully for further characterization of deflection bowl.
7 Conclusions The analytical model presented in this paper primarily emphasizes to estimate the deflection bowl of a multilayered flexible pavement using the interface deflection as design criteria with application of concentration factor. The pavement deflection obtained from analytical approach has been used to characterize the deflection bowl in terms of BLI, MLI, and LLI. The results obtained from present numerical approach have been found in good agreement with deflection bowl parameter obtained from field investigation. However, in future the correlation of extent and nature of damage of pavement in different layers with the layer index need to be studied extensively to for reliable evaluation of pavement performance.
7
6
5
4
3
2
122.5 km
1
Present model
134.8 km
Present model
124 km
Present model
123.8 km
Present model
123.3 km
Present model
123 km
Present model
122.75 km
Present model
Location on NH-6
S. No.
95
95
95
90
90
95
550
550
420
405
405
485
485
351.6
586.7
686.6
1082
949.3
986.9
990.5
E1
95
Elastic modulus (MPa)
h1
h2
Thickness (mm)
Table 2 Details of surface deflection with layer index
274.4
317.8
174.3
156.7
177.4
272.4
207.5
E2
84.4
69.8
49.5
46.0
71.5
95.8
54.7
E3
0.7607
0.3418
0.4901
0.2697
0.5134
0.4002
0.4029
0.3825
0.4452
0.3575
0.3399
0.2604
0.3487
0.3160
BLI
0.0909
0.0948
0.0846
0.0820
0.2226
0.1643
0.2466
0.2021
0.2245
0.1526
0.1257
0.0883
0.1586
0.1382
MLI
0.0501
0.0421
0.0471
0.0486
0.1187
0.0888
0.1354
0.1063
0.1084
0.0794
0.0651
0.0468
0.0905
0.0630
LLI
Characterization of Layer Index in Bituminous Pavement … 353
354
S. Purakayastha et al.
References 1. Camacho-Garita, E., et al.: Structural analysis for APT sections based on deflection parameters. Transp. Res. Rec. J. Transp. Res. Board 1–10 (2019) 2. IRC: 37: Guidelines for the Design of Flexible Pavements. Indian Road Congress, New Delhi (2018) 3. Biswas, P.P.: Mechanistic—empirical design of bituminous pavement using concentration factor in a two layered system, PhD thesis. Jadavpur University (2005) 4. Horak, E.: Benchmarking the structural condition of flexible pavements with deflection bowl parameters. J. South Afr. Inst. Civ. Eng. 50(2), Paper 652, pp. 2–9 (2008) 5. Reddy, A.M., et al.: Selection of genetic algorithm parameters for back-calculation of pavement moduli. Int. J. Pavement Eng. 5(2), 81–90 (2004)
Characterization of Rejuvenated Asphalt Binders Using FTIR and AFM Techniques Rayhan Bin Ahmed and Kamal Hossain
Abstract This study investigates the chemical and morphological behavior of rejuvenated asphalt binders. A TFOT aged PG 58-28 binder was mixed with three different types of rejuvenators that include waste cooking oil/untreated used cooking oil (UT), treated used oil (TR), and Hydrolene (HL) at 3, 6, and 9% by the weight of the total binder. To understand the characteristics and the performance of these rejuvenators, three sets of characterization tests were conducted: rheological, chemical, and morphological. This paper presents a summary of the results from our chemical and morphological studies. Gas Chromatography-Mass Spectroscopy (GC-MS) was conducted to identify the chemical composition of rejuvenators, while Fourier Transformed Infrared Spectroscopy (FTIR) was conducted for obtaining chemical functional group information. Furthermore, an Atomic Force Microscopy (AFM) was employed to obtain the micro-morphological properties of rejuvenated asphalt binders. The collected data showed that rejuvenation changes the chemical composition and alters the micro-structures of binders significantly, which impacts the overall performances of binder. In fact, this experimental study showed a good correlation between chemical compositional and morphological features with the rheological performance of the binder. The research findings are expected to contribute to the performance evaluation and characterization of rejuvenated asphalt mixes. Keywords Reclaimed asphalt pavement (RAP) · Aged binder · Rejuvenating agent · Chemical composition · Morphology
1 Introduction Over the last few decades, the necessity for developing a sustainable method of pavement construction has led to the increase use of recycled materials. In general, reclaimed asphalt pavement (RAP) is the major source of recycled materials that can be used for pavement construction. However, RAP materials are aged, which R. B. Ahmed (B) · K. Hossain Advanced Road and Transportation Engineering Lab (ARTEL), Department of Civil Engineering, Memorial University, St. John’s, NF, Canada e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_45
355
356
R. B. Ahmed and K. Hossain
may lead to pavement distress and early cracking in the pavement [1]. Therefore, it is recommended to use rejuvenators to improve the performance of the aged binder after reducing its viscosity [2]. When we use rejuvenator during recycling, one of the underlying questions is how rejuvenation helps to retrieve these properties and unfortunately the answer is still not clear. Therefore, this effort is designed to obtain an understanding of how the chemistry and microstructure of an asphalt binder changes as the rejuvenator is added with the aged binder. FTIR is the most widely used technique to obtain an understanding of chemical properties and the functional groups of asphalt binders. Recently, in the USA, Hossain et al. [3] investigated the effect of different additives on the performance of asphalt binders using FTIR. This study reports a significant correlation between the chemical compositions of asphalt binder with its rheological properties. For asphalt microstructure analysis, researchers across the world have increasingly been using the AFM technique [4]. Many studies made significant observations on how these microstructures are associated with asphalt morphology and rheological performance [5]. Most of the studies in this area found a good correlation between the formations of microstructure (bee structure) with the modification/rejuvenation of binders. Differing observations were also reported by some studies [6]. Despite these researches, still, there is a knowledge gap on how rejuvenation affects the asphalt morphology and microstructure. Therefore, this research attempts to generate a better understanding of this topic area. During the research project, the authors completed a comprehensive review to find out the existing literature gap on rejuvenated asphalt binders, which was reported in [7]. Based on the comprehensive review, the following objectives have been set for this study: • To investigate the chemical characteristics of rejuvenators that affects the properties of asphalt using GC-MS and FTIR techniques • To understand the effect of rejuvenators on the morphological features of asphalt at micro-level through AFM analysis.
2 Experimental Design 2.1 Material Collection The PG 58-28 binder is selected for this study which is generally used in the southern region of Newfoundland and Labrador, Canada. The Thin Film Oven (TFO) test was employed to prepare the laboratory simulated aged binder as per the guideline of ASTM D1754. Rejuvenators studied in this research are used cooking oil and Hydrolene (HL). Used cooking oil has been studied in two forms: (1) collected raw oil/untreated oil (UT) and (2) chemically processed oil/treated oil (TR).
Characterization of Rejuvenated Asphalt Binders Using FTIR …
357
Fig. 1 Types of rejuvenators
2.2 Chemical Characterization and Modification The chemical properties of rejuvenators are the fundamental characteristics that dominate the behavior of asphalt. HL is a commercially available material and its fundamental properties are well-known. However, UT is a used oil whose fundamental properties are still unknown, and it varies based on its source and use. One of the major properties for the used oil is the existence of free fatty acids (FFA) which weakens the adhesion with bitumen and reduces the binder adhesive properties. GCMS is one of the widely used techniques to quantify the composition of fats and oils. This study investigates the chemical composition of UT using the methylation procedure for GC-MS analysis as per ISO 12966-2:2017. Based on the GC-MS result, the molecular components of UT contain both polar (–CH2 –) and non-polar (–COO–) ends, which denotes that molecules are hydrophilic and have a strong affinity to water. This affinity has a harmful effect on the overall moisture susceptibility of asphalt. To reduce these FFA contents, the transesterification method has been employed [8] and used to prepare the TR rejuvenator used in this project. Later, the acid value test was also conducted as per ASTM D5555 to observe the change of acid value before and after transesterification process. The test found the acid value of TR reduces to 1.95 mL/g from 5.45 mL/g for the UT sample. All types of rejuvenators, including UT, TR, and HL are shown in Fig. 1.
2.3 Blending of Rejuvenators and Asphalt Binder A rejuvenated asphalt binder was prepared by mixing aged binder with three rejuvenators UT, TR, and HL at the concentrations of 3, 6, and 9% respectively by the weight of base asphalt. The blending process was done manually with a glass stirrer for 30 min at 100 °C.
358
R. B. Ahmed and K. Hossain
3 Results and Discussion 3.1 Effect of Rejuvenators on Asphalt Oxidative Potential The functional groups of rejuvenated asphalt binders were analyzed by IR spectrum table. For clarity, only the binder with 3 and 6% rejuvenation has been presented in Fig. 2. Out of the region specified, 1032 cm−1 and 1744 cm−1 bands correspond to sulfoxide (S=O) and carbonyl (C=O) respectively, which reflects the aging and rejuvenating degree of asphalt [9]. The intensity of C=O bond increases with the increase of UT dosages. However, this peak intensity decreased gradually for increasing dosages of HL and 9% HL shows the lowest peak. According to the literature review, this C=O bond is a saturated aliph and it is responsible for making the binder softer, which is more prone to rutting [10]. Also, this peak intensity was recorded 0.0176, 0.0171, and 0.0189 for 3, 6, and 9% TR rejuvenated sample. This result is consistent with our previous analysis, as 6% TR shows higher stiffness. Also, the peak intensity of Sulfoxide (S=O) is higher for TR samples compared to others. The presence of C=C aromatic bond was detected at 1452 cm−1 for all the samples which exhibit a strong bonding of the constituents. For aged asphalt, the intensity was 0.279, whereas for UT, TR and H sample, the intensity was highest for TR samples. Another C–O alcohol peak was found at 1161.15 cm−1 wavelength, which was only observed for UT rejuvenated samples, not in TR and HL samples due to the chemical modification after the transesterification process. Therefore, the conclusion drawn from the FTIR spectrum analysis is that the chemical constituents of binder can be changed with different rejuvenator types and the modification of rejuvenator. Quality improvement has a notable influence on the rejuvenation property and the overall performance of the binder as well.
Normalised Absorbance
6% HL 3% HL 6% TR 3% TR 6% UT 3% UT Aged binder 4000
3600
3200
2800
2400 2000 1600 Wavenumber cm-1
1200
800
Fig. 2 Normalized FTIR spectra of aged, UT, TR, and HL rejuvenated binder
400
Characterization of Rejuvenated Asphalt Binders Using FTIR …
359
3.2 Effect of Rejuvenators on Asphalt Morphologies Figure 3 shows the micrographs of 3 and 6% rejuvenated sample topography with their phase images. For the aged sample, a combination of different sized bee structure (small and large bees) was observed in the topographic image. In general, one can also easily notice that the aged specimen exhibited a fewer number of bees than specimens with rejuvenators. Also, some white spots are present in both topographic and perpetua domain of the phase image, which might be due to the aging of binder. For 3% UT sample (Fig. 3b), the size of the bee structures decreased compared to the aged binder. When the concentrations increased from 3 to 6%, the quantity of bee structure decreased but a larger sized bee structure was observed. In case of 6% UT, the phase contrast between the two phases changed significantly compared to 3% UT
Fig. 3 Topography (top) and phase (bottom) images of different rejuvenated binders. Image dimensions are 10 µm × 10 µm; topography image scale 155.6 nm; phase image scale 34°
360
R. B. Ahmed and K. Hossain
sample. The brighter phase represents the higher phase lag which will enhance the viscous behavior of the binder. Also, a dark region was observed for 6% UT in Fig. 3c, which might be “sal” phase mentioned by Masson et al. [11]. For the 3% TR sample (Fig. 3d), a significant increase of bee structures was observed compared to aged and UT samples, which denotes the higher stiffness of the binder. In the case of 6% TR, a reduced quantity of bee structures with reduced size was observed. However, in phase image, a clear disperse and a matrix phase was available for 3% TR samples. In topographic image, “sal” phases are visible for 6% TR which is identical to 6% UT. The HL rejuvenated samples in Fig. 3f–g exhibited the lowest bee structures. In addition, 6% HL rejuvenated asphalt shows a higher brighter phase which is responsible for the enhanced loss moduli and higher viscous property of asphalt. Based on the topographic and phase image analysis, it can be observed that the increase of rejuvenation dosage changes the microstructures and the phase contrast of the binders. This indicates a change in morphology of binders after rejuvenation.
4 Conclusions Based on the experimental study, the following conclusions can be drawn: • Free fatty acids (FFA) was observed in the UT rejuvenator by GC-MS analysis, which enhances hydrophilic characteristics and moisture susceptibility of asphalt binder. Therefore, it is required to reduce the FFA contents. • The chemical modification of UT influenced the rejuvenation behavior with aged asphalt, and higher quality of UT (lower FFA) may lead to better pavement performance. • FTIR test results exhibited a higher rate of C=O stretch at 1744 cm−1 for UT, which is responsible for softening the binder and inducing rutting failure. In contrast, this stretch was very negligible for other binders and therefore shows better stiffness. The result was also consistent with the stiffness trend found at the viscosity test. • From the AFM analysis, the decreased number of bees, with the increase of bee size, was observed for the increased dosages of rejuvenation. The higher bee structure was found for TR rejuvenated sample which was consistent with the past rheological test. • Based on the phase contrast of different samples, the higher percentages of rejuvenation show a brighter phase image which indicates the higher loss modulus. Therefore, rejuvenation of the aged binder significantly changes the morphological features. Acknowledgments The authors gratefully acknowledge the Leslie Harris Centre of Regional Policy and Developments and Multi-Material Stewardship Board of Government of Newfoundland and Labrador for the research funding.
Characterization of Rejuvenated Asphalt Binders Using FTIR …
361
References 1. Hossain, K., Karakas, A., Hossain, Z.: Effects of aging and rejuvenation on surface free energy measurements and adhesion of asphalt mixtures. J. Mater. Civ. Eng. 31(7) (2019) 2. Brown, E.R.: Preventive maintenance of asphalt concrete pavements. Transp. Res. Rec. J. Transp. Res. Board 6–11 (1988) 3. Hossain, Z., et al.: Evaluation for warm-mix additive-modified asphalt binders using spectroscopy techniques. J. Mater. Civ. Eng. 25(2), 149–159 (2013) 4. Hossain, K., Hossain, Z.: A synthesis of computational and experimental approaches of evaluating chemical, physical and mechanistic properties of asphalt binders. J. Adv. Civ. Eng. (2018) 5. Pauli, A.T., Grimes, R.W., Beemer, A.G., Turner, T.F., Branthaver, J.F.: Morphology of asphalts, asphalt fractions and model wax-doped asphalts studied by atomic force microscopy. Int. J. Pavement Eng. 12(4), 291–309 (2011) 6. Yu, X., Zaumanis, M., Dos Santos, S., Poulikakos, L.D.: Rheological, microscopic, and chemical characterization of the rejuvenating effect on asphalt binders. Fuel 135(1), 162–171 (2014) 7. Ahmed, R.B., Hossain, K.: Waste cooking oil as an asphalt rejuvenator: a state-of-the-art review. Constr. Build. Mater. 230, 116985 (2019) 8. Leung, D.Y.C., Guo, Y.: Transesterification of neat and used frying oil: optimization for biodiesel production. Fuel Process. Technol. 87(10), 883–890 (2006) 9. Herrington, P.R., Ball, G.F.A.: Dependence oxidation mechanism of asphalt. Fuel 75(9), 1129– 1131 (1996) 10. Azahar, W.N.A.W., Jaya, R.P., Hainin, M.R., Bujang, M., Ngadi, N.: Chemical modification of waste cooking oil to improve the physical and rheological properties of asphalt binder. Constr. Build. Mater. 126, 218–226 (2016) 11. Masson, J.-F., Leblond, V., Margeson, J.: Bitumen morphologies by phase-detection atomic force. J. Microsc. 221(1), 17–29 (2006)
Characterization of the Lateritic Soil of Kamboinsé (Burkina Faso) Marie Thérèse Marame Mbengue, Adamah Messan, Abdou Lawane, and Anne Pantet
Abstract Laterite is a very common material in tropical countries. Good quality lateritic deposits are becoming increasingly rare. This material is used as a base course for most roads built in tropical Africa. These materials are selected on the basis of geotechnical test results in accordance with current standards. However, these materials do not perform well in situ after a short period of time, causing premature pavement degradation. This poor behavior can be ascribed to insufficient geotechnical testing to justify the choice of these materials in road construction. In this work, in addition to traditional geotechnical tests (particle size analysis, Atterberg limits, modified Proctor, CBR Index), compression tests were performed in addition to traditional geotechnical tests to determine compressive strength and elasticity modulus. Geotechnical tests carried out on the laterite of Kamboinsé according to the lithology of the quarry reveal that the characteristics decrease with depth and that the first layer can be used as a base layer. Cement improvement for small percentages (1, 2 and 3%) shows a clear improvement in mechanical characteristics. An addition of 2% cement allows the use of the second base layer, improvement of 3% allows the use of the first layer as a base layer. An addition of granite crushed granites of class 0/31.5 improved the physical properties of materials for additive percentages varying from 20, 25, 30 and 35% by weight of dry materials. This improvement allows the use of layers 1, 2 and 3 as a base layer for low traffic, layer 4 can be used as a base layer. Keywords Laterites · Geotechnical characterization · Cement improvement · Litho-stabilization · Kamboinsé
M. T. M. Mbengue (B) · A. Messan · A. Lawane Laboratoire Eco-Matériaux et Habitats Durables (LEMHaD), Institut International d’Ingénierie de l’Eau et de l’Environnement (2iE), Ouagadougou, Burkina Faso e-mail: [email protected] M. T. M. Mbengue · A. Pantet Laboratoire Ondes et Milieux Complexe (LOMC), Université de Normandie, Le Havre, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_46
363
364
M. T. M. Mbengue et al.
1 Introduction In tropical Africa, the majority of road structures are in untimely degradation condition. This is due to several factors including material quality, poor drainage of pavement water and undersizing [5]. Our present study will involve geotechnical characterization of gravel and laterites from the quarry located in Kamboinsé by considering the different horizons that make up the quarry. The aim is on the one hand to be able to make recommendations on the ideal depth of exploitation of these materials, which layer would provide better characteristics (physical and mechanical). On the other hand, to know what dosage of cement and what percentage of crushed material would provide better mechanical characteristics.
2 Materials and Methods 2.1 Materials Samples were taken according to lithological layers encountered. Thus, the kamboinse sample has four different layers that are rated KC1, KC2, KC3, KC4. For this quarry, the topsoil was stripped to a thickness of 0.27 m. Then comes the KC1 layer which is 1.9 m thick, the materials are red-brown in color. This is followed by the 2.7 m thick KC2 layer, which is brown in color with white nodules. These first two layers are located in the B horizon of this profile. The KC3 layer extends to a depth of 3.9 m and is made of purplish red and white spotted materials. Finally, the 8.3 m thick KC4 layer in red color. Samples are stored in hermetically sealed barrels to avoid variations in the natural water content of the materials. The binder used to improve the mechanical characteristics of the materials is CimFaso CEMII/A-L 42.5R cement. Its dosage will be done according to 1–3% by weight for each layer.
2.2 Methods Identification and mechanical tests were carried out on both raw samples and samples mixed with cement. This mixing was done by pre-wetting the raw material before adding the necessary quantities of cement to prevent loss of particles from the cement. Mechanical tests were performed on wet cylindrical samples with optimal water content and compacted with Modified Proctor energy. These samples are then stored in a room at room temperature. All tests were carried out according to the AFNOR standard (Association Française de Normalisation).
Characterization of the Lateritic Soil of Kamboinsé …
365
Table 1 Mechanical properties Layer
CBR to 90% OPM
CBR to 95% OPM
CBR to 98% OPM
E (MPa)
Rc (MPa)
KC1
37
55
–
70.61
0.58
KC2
30
41
–
41.4
0.52
KC3
15
26
31
34.2
0.49
KC4
–
13
16
29.3
0.33
3 Results and Discussion 3.1 Raw Materials Table 1 shows the results of the identification and compaction tests performed on the samples from this quarry. Blue values are low for all layers, showing that the materials have a low content of swelling clays. This result is confirmed by the results of the plasticity index because according to Casagrande chart, these thin parts are not very plastic silts. Therefore, the HRB classifications shows the same trend, KC2, KC3 and KC4 layers are classified as fine soils (respectively A5, A7-6, A7-5) while materials of KC1 layer, are sandy and gravelly soils with fine (A2-5). Plasticity indices of all layers of the quarry are less than 15% which is the limit for base layer use according to CEBTP [2]. Proctor test shows that for Kamboinsé quarry, optimal dry densities decrease with depth. From the point of view of optimal dry density, KC1 layer of kamboinse laterite can be used both as a base layer and as a foundation layer because the latter is superior to 2 t/m3 [2]. KC2 layer can be used as a base layer because its density is higher than 1.8 t/m3 . First, it can be seen that CBR index decreases with depth for Kamboinsé quarry. The lower the fine content, the lower the plasticity index, the lower the optimal water content, the higher the CBR index. Mechanical properties of this soils are poor except thus of KC1 layer which shows 55% of CBR index, 70.61 MPa of modulus and a compressive stress of 0.58 MPa. This results show that only KC1 layer can be used as sub-base layer in road [2]. To improve the mechanical characteristics of lateritic soil, this material is mixed with cement. In this part, lateritic soil will be mixed with low doses of cement so that they can be used in flexible pavements.
3.2 Improved Cement Laterite Geotechnical and Mechanical Properties Table 2 shows that plasticity index decreases, water content decreases slightly and the dry density increases with the cement content. This decrease is desirable from a stabilization perspective. Similar results were obtained by other authors with Portland
366
M. T. M. Mbengue et al.
Table 2 Physical and compaction properties Identification of samples
Atterberg limits ωL
ωP
Ip
Modified Proctor ωopm (%) γd (t/m3 )
0% KC1
47.1
38.2
8.9
13.5
1% KC1
46.4
37.7
8.8
8
2% KC1
45.4
37.4
7.9
9.6
3% KC1
43.8
36.7
7.13
10.4
0% KC2
46.7
37
9.7
16.2
1% KC2
46.1
37.2
9
2% KC2
45.5
37
3% KC2
44.7
37
0% KC3
44.8
34
ICBR to 98% OPM
55
–
2.25
70
85
2.27
142
146
2.35
175
190
1.81
41
–
16.1
1.85
59
61
8.5
15.1
1.93
76
92
7.8
14.2
1.97
108
117
10.9
19.4
1.76
20
25
1% KC3
20.2
1.76
30
31
2% KC3
20.4
1.74
46
48
3% KC3
20.4
1.75
80
84
0% KC4
42.4
29.4
13
2.16
ICBR to 95% OPM
20
1.64
13
16
1% KC4
21.6
1.71
26
31
2% KC4
20.5
1.74
31
34
3% KC4
25.4
1.63
71
102
cement [4] and it has been postulated that when cement is mixed with cohesive finegrained soils, there is cation exchange, flocculation and soil agglomeration, which decreases the soil plasticity index and should reduce the optimal water content due to decrease in its specific surface. This means that cement reduces the clay activity of the material. Based on the specifications of CEBTP [2], only KC1 layer-improved cement can be used as base layer since its density is greater than 2 t/m3 . It is noted that CBR index increases with cement content, which shows that addition of cement increases the mechanical characteristics of laterites. This increase is due to the stiffening of laterite through hydration of cement which develop hydrates as hydrated calcium silicate (CSH). The CBR of KC2 improved to 3%, as well as those of KC1 layer that improved to 2 and 3% are greater than 100, therefore they can be used as a sub-base layer according to CEBTP [2]. However, only KC1 layer improved with 3% cement can be used as a base layer because it is higher than 160. Table 3 shows the evolution of Young’s modulus (according to the standard NF EN 13286-43) and the compressive strength (according to the standard NF EN 1328641) of kamboinse laterites in relation to the percentage of cement after 7 days of curing. In all cases, the modulus increases with the cement content. Only KC1 layer improved to 3% has a modulus superior to 300 MPa. This value guarantees the good mechanical performance of a lateritic gravelly material and its use as a base layer of road structure [1].
Characterization of the Lateritic Soil of Kamboinsé …
367
Table 3 Variation of Young’s modulus and compressive strength according to layers and cement contents Parameters
% Cement
E (MPa)
0% Cem
90.34
1% Cem
64.5
373.5
57.6
121.65
2% Cem
234.86
467.6
69.04
215
307.2
Rc (MPa)
KC1
KC2 62.52
KC3
KC4
52.25
56.8
3% Cem
603.75
95.12
229
0% Cem
0.36
0.43
0.25
0.34
1% Cem
0.57
0.56
0.71
0.42
2% Cem
1.57
1.75
1
0.95
3% Cem
2.42
2.34
1.07
0.6
E: Young’s Modulus Rc: Compressive strength Cem: Cement
The compressive strength increases with increasing cement content. The same trend was noted by Millogo et al. [5]. Only KC1 and KC2 layers (2.42 and 2.34 MPa respectively) improved to 3% Cement can be used as a base layer because they are between 1.8 and 3 MPa.
3.3 Litho-Stabilized Laterite Geotechnical and Mechanical Properties Apart from cement stabilization, laterites used in road construction can be improved by substituting part of their weight with sand, crushed shells, palm nut shells, vegetable fibers [3]. This last part is devoted to improving laterite by adding crushed material. To this end, improvement will be achieved by adding granites of class 0/31.5. Different percentages of crushed granites ranging from 20, 25, 30 and 35% by dry weight were added to the raw laterite. Table 4 presents the evolution of the optimal density with the percentage of granite crushed added. This is due to the use of coarse granular materials in the mixture, Table 4 Variation in density according to the percentage of crushed material added % Granites
KC1
KC2
KC3
KC4
0% G
2.16
1.81
1.76
1.64
20% G
2.17
1.96
1.86
1.77
25% G
2.23
1.89
1.95
1.81
30% G
2.15
2.05
1.92
1.81
35% G
2.04
2.32
2
1.85
368
M. T. M. Mbengue et al.
allowing a better interweaving of grains and consequently, the formation of a denser soil layer. For 35% of granites, there is a sudden decrease in density. This KC1 material had the best densities, this decrease can be explained by the excessive presence of coarse grains which eventually increase the voids index of the mixture. The best densities are obtained for a contribution of 35% for the other materials, thus bringing the densities of KC2 and KC3 into the density range of a base layer (greater than 2 t/m3 ). The optimum water content of layers KC1, KC2, KC3 and KC4 varies respectively from 13.5 to 8.03; from 16.2 to 8; from 19.4 to 12.3; from 20 to 13.6 according to the percentage of granite added. It is clear that this water content decreases with the percentage of granites, the lowest values are obtained at 35% granites. Table 5 shows the variation in CBR index according to the percentage of granite. CBR increases with the percentage of granite. This is due to the fact that the mixture becomes increasingly insensitive to water as shown by the considerable decrease in water content and increase in density [3]. The maximum values for KC1 and KC2 are given by 30% granite, and for soil layers KC3 and KC4 by 35% crushed. From a 30% improvement, all layers can be used as sub-base layer, for the same percentage, KC1 and KC2 can be used as base layer. A 35% KC3 can be used as a base layer. As for compressive strength, there is an increasing trend, but results are dispersed dû to the homogenous of samples. After improvement from 30%, all layers can be used as sub-base layer except KC4 because this resistance is between 0.5 and 1.5 MPa. Modules are presented in Table 5. These values are quite dispersed certainly due to the difficulty in fixing sensor supports due to the crumbling of laterite. But overall, Table 5 Variation in the mechanical properties of the different layers according to the percentage of crushed material added Parameters
% Granite
KC1
KC2
KC3
KC4
E (MPa)
0% G
90.34
62.52
56.8
52.25
20% G
168.35
38.83
10.63
8.53
25% G
190.034
18.18
21.57
12.36
30% G
206.71
43.8
42.36
17.76
Rc (MPa)
CBR to 95% OPM
35% G
112.586
204.54
41.96
26.52
0% G
0.36
0.43
0.25
0.34
20% G
1.72
0.76
0.32
0.31
25% G
0.51
0.51
0.32
0.35
30% G
0.57
1.114
0.57
0.32
35% G
0.91
0.73
0.55
0.26
0% G
55
41
20
13
20% G
76
74
24
16
25% G
79
84
34
18
30% G
86
101
44
56
35% G
74
70
84
58
Characterization of the Lateritic Soil of Kamboinsé …
369
we notice a clear increase in the module depending on the rate of crushed material, more pronounced on the KC1 layer. The values obtained for KC1 from 20% and KC2 to 35% are in the module range of lateritic materials used in pavement layers. They are higher than that obtained by Millogo et al. [5] for the raw laterite of Sapouy, which was around 100 MPa.
4 Conclusion From these studies, we observe that in their raw state, these materials can only be used as foundation layers. Surface materials such as KC1 layer are recommended. Cement improvement and litho-stabilization show that mechanical and physical properties of these materials can be improved and allow their use as base layer for light traffic. For cement a percentage of 2% is sufficient, for crushed cement a percentage of 30% is sufficient.
References 1. Bagarre, E.: Utilisation des graveleux latéritiques en technique routière (1990) 2. CEBTP: Guide pratique de dimensionnement des chaussées pour les pays tropicaux (1984) 3. Issiakou, M.S., Saiyouri, N., Anguy, Y., Gaborieau, C., Fabre, R.: Etude des matériaux latéritiques utilisées en construction routière au Niger: Méthode d’amélioration (2015) 4. Mengue, E., Mroueh, H., Lancelot, L., Eko, R.M.: Mechanical improvement of a fine-grained lateritic soil treated with cement for use in road construction. J. Mater. Civ. Eng. 29, 04017206 (2017) 5. Millogo, Y., Hajjaji, M., Ouedraogo, R., Gomina, M.: Cement-lateritic gravels mixtures: microstructure and strength characteristics. Constr. Build. Mater. 22, 2078–2086 (2008)
Circular Accelerated Surfacing Tester (CAST) for Evaluating Chip Sealing Binders Lia C. van den Kerkhof, Jeremy P. Wu, Shaun R. Cook, and Philip R. Herrington
Abstract Development of a laboratory scale device for testing chip seal road surfacing performance at realistic loading rates and under controlled temperatures is described. A variety of sealing binders have been tested, and the forces applied to the seal quantified. The damage produced in the seal surface by the CAST (Circular Accelerated Surfacing Tester) machine was quantified by a stereo-photographic method which captured damage arising from chip loss, displacement and reorientation. As part of a study into a performance related bitumen specification for chip seal binders, the CAST machine was used to verify seal performance against the viscosity at 45 °C. A correlation was found between viscosity and seal performance at 45 °C. The results are presented here to illustrate the operation of the CAST machine. Keywords Accelerated surfacing tester · Road surfaces · Durability · Field performance · Performance graded specification · Bituminous binder · Modified binder
1 Introduction It is difficult to assess the performance of chip seal road surfacings under conditions in the laboratory. Road trials are costly, can takes years to evaluate, are challenging to monitor and involve many external variables which are difficult to control. The Circular Accelerated Surfacing Tester (CAST) machine was developed so that preliminary testing of multiple types of surfacings could be completed on a laboratory scale and within a short timeframe. Bitumen is a viscoelastic material for which the material properties are loading rate and temperature dependent. The CAST machine is designed to simulate real field loading rates on pavement surfaces such as chip seals, thin asphalt, and pavement marking materials under controlled temperature conditions. Details of the machine and testing methods are presented in this paper.
L. C. van den Kerkhof (B) · J. P. Wu · S. R. Cook · P. R. Herrington WSP Research and Innovation, Petone, Wellington, New Zealand e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_47
371
372
L. C. van den Kerkhof et al.
To illustrate the use of the device, as part of a study into development of a performance based bitumen specification for chip seal binders, tests was undertaken to evaluate the relationship between binder viscosity and seal damage at 45 °C.
2 Methodology 2.1 Circular Accelerated Surfacing Tester (CAST) CAST (Fig. 1) consists of a rotating pair of 320 mm diameter pneumatic tyres which apply a vertical load of up to 200 kg as they travel around a 7.4 m long circular track at up to 50 km/h. The small radius of the track ensures a high level of scuffing to the test surface to provide the accelerated nature of the test. Four samples can be inserted into the track for simultaneous testing providing a consistent environment for comparative testing. It is housed within an insulated enclosure allowing tests to be undertaken at a controlled air temperature. The temperature of the test plates is also controlled independently of the air temperature by a fluid circulating system between −7 and 55 °C. Full detail of CAST is described in the publication [1]. The surface forces exerted by the CAST tyres were measured using a three-axis load cell (Measurex MLD61, 300 kg range per axis), calibrated with forces up to 2 kN (using a Shimadzu EZ-LX test machine). A rigid textured surface similar to that of a test seal was fixed to the top of the load cell in the plane of the track. Measurements were conducted at speeds from 4 to 25 km/h. The results showed a decrease in vertical load with increasing speed, presumably due to aerodynamic lift, whereas the sideways (centripetal) force remained constant. At a testing speed of 25 km/h, the vertical and sideway forces on the surface were 0.5 and 0.6 kN respectively (or 500 and 600 kPa based on the tyre footprint at that speed).
Fig. 1 Circular accelerated surfacing tester (CAST)
Circular Accelerated Surfacing Tester (CAST) for Evaluating …
373
Table 1 CAST testing conditions Test parameter
Value
Seal temperature (°C)
45 °C ± 0.5
Air temperature (°C)
20 °C ± 3
Tyre linear speed (km h−1 ) 25 Tyre pressure
36 psi (continuously measured)
Tyre dead load (kg)
187 kg (total weight of the tyre and support assembly)
Tyre temperature
Approximately equal to the air temperature (tyre temperature and pressure do not change significantly due to the short test duration)
Number of passes per test
105
Seal type
Single coat
Chip grade
NZTA M/6 grade 3 (retained 9.5 mm passing 13.0 mm, washed). Greywacke aggregate sourced from Belmont quarry in Lower Hutt
Chip application rate
1300 g (14.5 kg m−2 or 179 m2 m−3 )
Binder application rate
252 g (2.15 Lm−2 , cold)
Test seals were prepared by pouring or spraying binder onto a heated aluminium plate (307 × 405 mm). The preparation temperature was adjusted, depending on the binder type, between 130 and 165 °C to obtain an equiviscous condition. The plates are identical and were manufactured to have a rough surface texture similar to a fine chip seal (equivalent to sand circle measurement of 230 mm). Chip was applied at room temperature and initially reoriented by tamping with a small wooden block. Before the plate cooled to room temperature the seal was rolled with 2 passes of a 10 kg Vialit-test rubber roller to help dislodge any loose chip. Additional rolling is undertaken to reorientate the chip by tracking the seals under a pneumatic tyre with a 100 kg deadweight at 25 °C for 616 passes/minute (14 min). The seal is slowly traversed backwards and forwards perpendicular to the tracking tyre, to cover the entire seal area. The testing conditions used are given in Table 1.
2.2 Materials and Test Standards A wide variety of binders were selected (Table 2) including styrene-butadiene-styrene (SBS) polymer modified binders and two experimental nanoclay modified binders (details of which are in [2]). Standard penetration measurements were carried out at 25 °C per ASTM D5 [3]. The viscosity was measured using a dynamic shear rheometer (TA Instruments AR2000 Ex) at 45 °C with a 25 mm diameter 4° cone and plate geometry at shear rates from 0.0005 to 5 s−1 and reported at 1.58 s−1 . A single source of aggregate was used (Table 1).
374
L. C. van den Kerkhof et al.
2.3 Stereo-Photogrammetric Image Analysis Damage to the seals from CAST testing was quantified using stereo photography to measure changes in the void volume of the surface in a defined rectangular area in the wheel path (127 × 305 mm) (Fig. 2). The analysis was done using ImageMaster™ (Topcon). This program can create a surface model and give depth, area and volume information. An image pair was taken before and after CAST testing. The program uses this given information to establish a scale and form a triangulated irregular network to model the seal surface. The level of detail of the surface
110 mm
127 mm
50 mm
70 mm
Fig. 2 Top: damaged seal plate after cast testing, the central area used for the volume calculation is bounded by the yellow lines and white reference dots around the edge of the plate. Bottom: schematic diagram illustrating the void volume (in red), used to determine seal damage
Circular Accelerated Surfacing Tester (CAST) for Evaluating …
375
can be controlled by controlling the mesh size, 0.75 mm was used in this case. This resulted in sufficient surface detail without excessive analysis time. The void volume was calculated as the empty space below a plane defined by invariant reference marks on the lip of each plate. The known distance between the reference marks was used to set the scale in mm. The void volume change was used as a measure of seal damage as illustrated in Fig. 2, and encompasses changes in the volume because of chip “rollover”, lateral displacement (shearing) as well as simple chip loss. The greater the void volume increase, the greater the seal damage. Typically 4–8 replicates were tested in each case and the results averaged.
2.4 Results and Discussion The binders used in this study are listed in Table 2 with corresponding penetration and viscosity data. The CAST results were consistent with expectations in that highly polymer modified binders showed the least seal damage. Harder binder grades generally showed less seal damage than softer ones but the correlation with penetration was poor, a better correlation was observed between viscosity and seal damage (Fig. 3). Table 2 Penetration at 25 °C and viscosity at 45 °C (1.58 s−1 ) data for the binders used in the study Bindera
3% SBS
4% SBS-A
4% SBS-B
5% SBS-A
2% NCc
6% NCc
Penetration
96
61
83
124
130
103
Viscosity (Pa s)
2212
4002
3061
5662
1393
947
Binder type A20Eb (5% SBS)
S20Eb (5% SBS)
S25Eb (6.5% SBS)
A15Eb (6.5% SBS)
Penetration
57
66
73
60
Viscosity (Pa s)
7527
5102
6384
8745
Binder type 180-200-A
180-200- B
180-200-C
180-200-D
130-150
80-100-A
Penetration
195
200
168
171
130
78
Viscosity (Pa s)
890.3
527.6
922.3
655.3
1123
1134
Binder type 80-100-B
80-100-C
60-70-A
60-70-B
40-50-A
40-50-B
Penetration
94
83
67
63
45
43
Viscosity (Pa s)
1461
1803
2070
2205
2083
2655
a The
SBS modified binders differ in the base bitumen and amount of blending oils used provided by ARRB Group Ltd complying with Austroads specification AGPT/T190, see
b Binders
[4] c Nano-clay
modified
376
L. C. van den Kerkhof et al.
Fig. 3 Lower void volume change versus viscosity (shear rate 1.58 s−1 ) at 45 °C
There was little difference in performance between the bitumen grades commonly used for sealing (60–70 to 180–200 penetration). For two of the lower SBS concentration (3–4%) binders, damage was similar to that of unmodified binders. The coefficient of variation (sd/mean) for the CAST test results was typically about 40%. Variation could possibly be reduced in future by increasing the number of plates tested, but the large spread was found to be mainly attributable to the real variability of damage to the seals, not an artefact of the data processing or sample preparation. The chip shape and packing in a seal is irregular and the orientation of a given chip to the tyre will strongly affect the detailed stresses it experiences and the resulting damage. Damage can be highly localized especially where overall damage is minor. This models the situation observed in the field.
3 Conclusion The CAST machine allows inexpensive assessment of chip seal binders and seal designs under representative in-service loading rates and stresses. The relative field performance of different binders can thus be assessed with much greater confidence than that possible using much slower loading rate tests (e.g. sweep tests). The analysis method captures seal damage due to shearing and chip “roll-over” common at high temperatures and not just chip loss. The illustrative results presented highlight the relative benefits of high (>5%) SBS content binders in preventing high temperature
Circular Accelerated Surfacing Tester (CAST) for Evaluating …
377
seal damage and indicate that the viscosity (>45 °C) may be a useful predictor of seal damage. Further work is continuing to validate CAST test rankings against field observations. Acknowledgements The authors would like to thank the NZ Transport Agency for funding this research. The facts presented, and views expressed in this work should not be regarded as being the opinion or the official policy of the NZ Transport Agency. The authors would like to thank Austroads and ARRB Group Ltd for permission to make use of data from [4].
References 1. Neaylon, K.L., et al.: Effect of Road Seal Type on Resistance to Traffic Stresses. NZTA Research Report 634. NZ Transport Agency (2017) 2. Bagshaw, S.A., et al.: Effect of blending conditions on nano-clay bitumen nanocomposite properties. Road Mater. Pavement Des. 1–22 (2018) 3. ASTM: ASTM D5/D5M-13 Standard Test Method for Penetration of Bituminous Materials. ASTM International, West Conshohocken (2013) 4. Urquhart, R., Patrick, S.: Performance of Asphalt and Spray Grade PMBs in Sprayed Seals. Austroads Report AP-T345-19 (2019)
Coefficient of Thermal Expansion and Thermally Induced Internal Cracking of Asphalt Mixes Sang Soo Kim, Moses Akentuna, and Munir Nazzal
Abstract The coefficient of thermal expansion/contraction (CTE) is an important factor influencing the low temperature thermal cracking of asphalt pavements. The typical CTE of asphalt mixture decreases as the temperature drops, showing a sigmoidal form of gradual transition. The objectives of this study were to determine the effects of CTEs of asphalt binder and aggregate on the mixture CTE, to determine the effects of the mixture CTE on the low temperature performance of the asphalt mixture, and to investigate the transition behavior of the mixture CTE at low temperatures. Two limestone aggregates and four asphalt binders with different CTE values were used to prepare eight asphalt mixture combinations. The mixture CTE values were determined using a custom-made CTE device where the thermally induced deformations were measured in two diametral directions. The low temperature performance was determined using the Asphalt Concrete Cracking Device (ACCD), a concentric thermal stress restrained specimen test. Based on the test results, it was concluded that the measured asphalt mixture CTE values were significantly smaller than the values predicted by the current AASHTOWare Pavement ME model. The cracking temperature measured by ACCD became colder as the mixture CTE became smaller. The transition of the mixture CTE appeared to be caused by internal cracking and could be modeled with the isochronal stiffness of asphalt binder. The mixture CTE transition temperature was highly correlated with the ACCD cracking temperature. Keywords Asphalt · Coefficient of thermal expansion · Low temperature cracking · Internal cracking
S. S. Kim (B) Ohio University, Athens, OH, USA e-mail: [email protected] M. Akentuna Louisiana Transportation Research Center, Baton Rouge, LA, USA M. Nazzal University of Cincinnati, Cincinnati, OH, USA © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_48
379
380
S. S. Kim et al.
1 Introduction The coefficient of thermal expansion and contraction (CTE) is one of major factors influencing the low temperature cracking phenomenon of asphalt pavements. When the asphalt pavement temperature decreases, the pavement restrained from contraction develops thermal tensile stress. When this thermal stress exceeds the strength of the asphalt pavement, low temperature cracking occurs. Together with the cooling rate and the rheological properties, the CTE of asphalt mixture is directly related to the induced thermal stress of the asphalt pavements. However, the thermo-volumetric properties of asphalt mixture and their contribution to the low temperature performance are still largely unknown. Current form of asphalt mixture CTE model given in Eq. (1) commonly used in asphalt pavement analysis for low temperature including AASHTOWare Pavement ME is first proposed by Jones et al. [1]. αmi x =
V M A × Bb + Vagg × Bagg 3 × Vtot
(1)
where, αmi x is linear mixture CTE, V M A is voids in mineral aggregates (%), Vagg is volume of aggregates in mixture (%), Vtot is total mixture volume (100%), Bb and Bagg are binder and aggregate volumetric CTEs, respectively. There has been several attempts to relate the mixture thermo-volumetric properties including CTE to the thermo-volumetric properties of the asphalt binder and aggregate with varying degrees of success. Nam and Bahia [2] found out that the binder glass transition temperature was significantly related to the mixture CTE transition temperature and that the binder and aggregate CTE values were somewhat related to the mixture CTE values. Akentuna et al. [3] observed that the mixture CTE was more affected by the coarse aggregates than the fine aggregate in their study.
2 Objectives The objectives of this study were • To determine the effects of CTEs of asphalt binder and aggregate on the mixture CTE, • To determine the effects of the mixture CTE on the low temperature performance of the asphalt mixture, and • To investigate the transition behavior of the mixture CTE at low temperatures.
Coefficient of Thermal Expansion and Thermally Induced Internal …
381
Table 1 Aggregate blend gradation for asphalt mixtures Sieve
12.5 mm
9.5 mm
#4
#8
#16
#30
#50
#100
#200
% Passing
100.0
95.6
54.6
33.9
22.3
12.8
6.6
3.5
2.2
3 Materials Eight densely graded asphalt mixture samples were prepared using four SHRP binders (AAA-1, AAC-1, AAF-1, and AAM-1) whose thermos-volumetric properties were reported by Bahia and Anderson [4] and two local limestone aggregates. The two aggregates consisted of low (4.2 με/°C) and high (8.4 με/°C) CTE limestones from Grove City and Belle Center, Ohio, respectively. Aggregate CTEs were determined by using a strain gauge method on polished aggregate surface as described in earlier study [3]. Each limestone aggregate was all sieved and then each sieve was blended to obtain the gradation given in Table 1. The mixtures were prepared with 5.7% asphalt binder content by weight of the mix in the laboratory and then shortterm aged (135 °C for 4 h) before compaction in the Superpave gyratory compactor with 65 design number of gyrations.
4 Test Methods The BBR tests were performed on asphalt binders at −15 and −21 °C. For each binder, a creep stiffness mastercurve was constructed and fitted to the CAM rheological model. Arrhenius function was also used for temperature dependent shift factor. From the CAM mastercurve equation and the temperature shift function for each binder, isochronal creep stiffness at 60 s was determined for the temperature range from 20 to −60 °C to relate the thermo-volumetric behavior of asphalt mixtures. Asphalt mixture CTE was measured using the Ohio CTE Device (OCD). In OCD measurement, a 50 mm thick and 150 mm diameter specimen was placed in a square aluminum frame as shown in Fig. 1a. The OCD measures diametric dimensional changes with two Linear Variable Differential Transducers (LVDTs) together with temperatures of specimen and the aluminum frame during cooling. The OCD samples were cooled from 20 to −60 °C with 20 °C per hour rate. The mixture thermal strains were determined from the LVDT readings by adjusting for the shrinkage of the aluminum frame. The frame contraction was determined with four reference samples with known CTE values. Mixture strains were then fitted to the Bahia-Anderson’s [4] proposed CTE model as given in Eq. 2.
ε = C + αg T − Tg
T − Tg + R αl − αg ln 1 + exp R
(2)
382
S. S. Kim et al.
Fig. 1 Ohio CTE device (OCD) (a) and ACCD ring assembly (b)
where, ε = thermal strain; C = constant, α l = the liquid state asymptote (liquid state CTE), α g = the glassy state asymptote (glassy state CTE); T = temperature of test specimen; T g = glass transition temperature and R = parameter indicating the temperature range of the glass transition zone. Mixture CTE was determined as the first derivative of Eq. 2 or the slope of the mixture thermal strain-temperature plot as given in Eq. 3. C T E = αg + αl − αg / 1 + exp[− T − Tg /R
(3)
For a temperature range of 0 to −30 °C, the average standard deviation of CTEs measured by OCD was found to be 0.3 με/°C [3]. The thermally induced cracking temperature was measured with the Asphalt Concrete Cracking Device (ACCD). As shown in Fig. 1b, ACCD is a concentric Thermal Stress Restrained Specimen Test (TSRST), developed as a simple alternative method for the traditional TSRST. The device works on the principle that the differences in CTE between a low CTE invar ring inside and a cored and notched 150 mm diameter asphalt mixture specimen outside induces tensile stresses in the mixture and cause cracking upon cooling in an environmental chamber [5].
5 Results and Discussion The average of four CTE measurements by OCD for eight mixes are shown in Fig. 2. For all mixes, CTE values remain relatively constant from 0 to −20 °C. As temperature decreases further, mix CTE starts to decrease. It is interesting to note that the mixes with stiff binders (AAF-1 and AAM-1) begin the CTE transition at the significantly warmer temperatures than the mixes with softer binders (AAA-1 and AAC-1), suggesting that the binder stiffness may play an important role in mix
Coefficient of Thermal Expansion and Thermally Induced Internal … 25
25
10 A-GC-ST-m A-BC-ST-m
5
20
-20 Temperature, °C
10 C-GC-ST-m C-BC-ST-m
5
-20 Temperature, °C
0
25
15 10 0 -40
15
0 -40
(c) AAF-1 Mixes
F-GC-ST-m F-BC-ST-m
5
(b) AAC-1 Mixes
20
0
-20 Temperature, °C
0
CTE, με/°C
CTE, με/°C
15
0 -40
CTE, με/°C
CTE, με/°C
(a) AAA-1 Mixes
20
25
383
(d) AAM-1 Mixes
20 15 10 5
M-GC-ST-m M-BC-ST-m
0 -40
-20 Temperature, °C
0
Fig. 2 Coefficient of thermal expansion/contraction (CTE) of eight mixes determined by Ohio CTE device (OCD)
ACCD Cracking Temperature, °C
(a)
-10
AAA-1 AAC-1 AAF-1 AAM-1
-15 -20 -25 -30 GC Mix -35
BC Mix
(b)
-15
ACCD Cracking Temperature, °C
CTE and the transition. There are distinctive aggregate CTE effects on the mixture CTE. For all binders used, the high CTE limestone mixes (GC mixes) exhibited consistently higher CTE values than the low CTE limestone mixes (BC mixes) for the temperature range, from 0 to −40 °C. The thermally induced cracking temperatures of eight mixes determined by ACCD are presented in Fig. 3a. The mixes with softer and lower Tg binders (AAA-1 and AAC-1) failed at colder temperatures than the mixes with stiffer and warmer Tg binders (AAF-1 and AAM-1). The low CTE aggregate mix (GC mix) resulted in a
-20
y = 0.8715x - 0.3114 R² = 0.94
-25 -30 -35 -35
BC Mix GC Mix
-25
-15
Mix Tg , °C Fig. 3 Asphalt concrete cracking device (ACCD) cracking temperature (a) and comparison to mix thermal transition temperature, Tg (b)
384
S. S. Kim et al.
Fig. 4 Coefficient of thermal exp./contr. (CTE) by current (a) and new (b) models
colder ACCD cracking temperature than the high CTE aggregate counterpart (BC mix) for each binder. The ACCD cracking temperatures were strongly related to the CTE transition temperature measured by OCD as shown in Fig. 3b. The current AASHTOWare Pavement ME CTE model (Eq. 1) significantly overestimated the mix CTEs by as high as 100% near 0 °C and was not able to account for the temperature dependency of CTE as shown in Fig. 4a. A new Sigmoidal form of CTE model is proposed (Eq. 4) and the prediction results are shown in Fig. 4b. The temperature dependency of CTE was introduced by using the temperature dependent isochronal creep stiffness at 60 s loading time in the model. αagg · Vagg + αl · Vb C T E(T ) = 0.64 ·
S(60s,T )−1069 1 + e 602
(4)
where, S(60 s, T ) is creep stiffness at 60 s loading time and temperature T in MPa unit, the constant 1069 is in MPa unit and indicates the center of CTE transition. It is interesting to note that the measured mix CTEs near −60 °C were even lower than the aggregate CTE values. These low CTE values are only possible when internal cracking occurs where the most of thermal contractions become localized without contributing to overall contraction of asphalt mix specimen. To better understand the transition phenomenon, the thermal strain of asphalt mix was measured by cooling from 25 to −60 °C followed by four cycles of heatingcooling between −60 and −20 °C. In this experiment, the transition near −30 °C only occurred during the first cooling cycle. However, when the same sample was
Coefficient of Thermal Expansion and Thermally Induced Internal …
385
reheated to 25 °C, the specimen exhibited the same transition behavior on the very next cooling cycle. This experiment suggests that the CTE transition was most likely a cracking phenomenon, warming to 25 °C healed the cracks but warming to −20 °C was not.
6 Summary and Conclusions Based on the test results presented, the following conclusions were drawn from this study: • The measured asphalt mixture CTE values were significantly smaller than those predicted by the current AASHTOWare Pavement ME model. • The cracking temperature measured by the ACCD became colder as the mixture CTE became smaller. • The transition of the mixture CTE appeared to be caused by internal cracking and could be modeled with the isochronal stiffness of the asphalt binder. • The mixture CTE transition temperature had a strong correlation with the ACCD cracking temperature.
References 1. Jones, G.M., Darter, M.I., Littlefield, G.: Thermal expansion-contraction of asphalt concrete. J. Assoc. Asphalt Paving Technol. 37, 56–100 (1968) 2. Nam, K., Bahia, H.U.: Effect of binder and mixture variables on glass transition behavior of asphalt mixtures. J. Assoc. Asphalt Paving Technol. 73, 89–119 (2004) 3. Akentuna, M., Kim, S.S., Nazzal, M., Abbas, A.R.: Asphalt mixture CTE measurement and the determination of factors affecting CTE. J. Mater. Civ. Eng. 29(6) (2017) 4. Bahia, H.U., Anderson, D.A.: Glass transition behavior and physical hardening of asphalt binders (with discussion). J. Assoc. Asphalt Paving Technol. 62, 93–129 (1993) 5. Kim, S., Akentuna, M., Nazzal, M., Abbas, A.: Effect of asphalt binder grade and type on low temperature performance determined using asphalt concrete cracking device (ACCD). J. Assoc. Asphalt Paving Technol. 85, 641–658 (2016)
Combined Recycling of RAS and RAP: Experimental and Modeling Approach Abdeldjalil Daoudi, Anne Dony, Alan Carter, Layella Ziyani, and Daniel Perraton
Abstract The use of Reclaimed Asphalt Pavement (RAP) and Recycled Asphalt Shingles (RAS) in combination with new materials in asphalt mixes appears to be a good solution to achieve economic and environmental benefits while maintaining or increasing the performances. The main objective of this work is to verify the application of the two blending laws for the prediction of conventional, physicochemical and rheological parameters at the binder scale. After extraction and recovery operations of RAP and RAS binders, the recovered binders were mixed with a virgin bitumen (PG 58-34) in several recycled bitumen ratios (RBR) up to 40% wt. The blends were characterized by conventional (penetration and Ring and Ball softening point), rheological (shear complex modulus) and physicochemical (Infrared spectroscopy (FTIR), Differential Scanning Calorimetry (DSC)) tests to compare the rheological behavior, the oxidation levels and the glass transition temperature. The 2S2P1D model was used to fit experimental shear complex modulus data. The proposed models describe well the experimental data from conventional, rheological and physicochemical tests. Keywords Recycling · RAS and RAP · Rheology · 2S2P1D model
1 Introduction Both for environmental and economic concerns, the use of recycled materials in the pavement is more and more important for the companies. Reclaimed Asphalt Pavement (RAP) is one of the most used materials. It consists of about ~95% wt. of aggregates and ~5% wt. of binder aged during its life on the road (mechanical + thermal ageing). Post-consumer Recycled Asphalt Shingles (RAS), initially dedicated to roofing materials especially in North America, can also be reused for road A. Daoudi (B) · A. Carter · D. Perraton École de Technologie Supérieure, Montréal/ÉTS, Montréal, Canada e-mail: [email protected] A. Dony · L. Ziyani Université Paris-Est, Institut de Recherche en Constructibilité (IRC), École Spéciale des Travaux Publics (ESTP Paris), Paris, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_49
387
388
A. Daoudi et al.
construction in combination with virgin materials. These RAS can contain about 25–40% wt. of a hard bituminous binder. At the end of their life cycle, they are recovered and shredded to be combined with new materials in the manufacture of Hot Mix Asphalt (HMA). The combination of RAP and RAS seems to be a good way to reduce costs and employment of non-renewable natural resources (bitumen and aggregates), and to reduce the energy and gas emission associated with the extraction and transportation of the new material. However, it is important to consider the different natures of materials, in particular the bitumen, and so to know their influence on the final material to ensure a good behavior under traffic and climatic conditions throughout its service life.
2 Literature Review In Canada, the Quebec Ministry of Transport’s LC 4202 standard allows using up to 20% wt. of RAP with 100% of active binder, up to 3% wt. of RAS for surface layers and 5% wt. for base layers with 25% of active binder. The standard also allows to use up to 20% wt. of a combination of recycled materials, for example RAP + RAS. Several studies focusing on RAP or RAS bitumen mobilization rate and on the problem of the blend between the virgin bitumen and the recycled binders (bitumen from RAP or RAS) are proposed. Zhao et al. [1] have developed a chemical method using the GPC (Gel Permeation Chromatography) test to study the mobilization rate of RAP and RAS materials. The results show that RAP binder mobilization rate decreases with the increasing RAP content in the mixture. The mobilization rate could reach 100% at low RAP contents (10 and 20%). For RAS, the mobilization rate reaches a peak at 5% of RAS content and then starts to decrease with the increase of RAS content. Navaro et al. [2] have used the sequenced extraction associated with carbonyl indexes determination to show the importance of manufacturing conditions (time and mixing temperature) on the homogeneity of the blend between the two binders (virgin and from RAP) in a recycled asphalt. Im and Zhou [3] have dealt with the problem from a rheological point of view; the authors have proposed to investigate the variation of PG H-L temperature with addition of RAS and RAP binder. Different percentages of virgin, RAP and RAS binders were blended and then evaluated through DSR and BBR tests in terms of the high and low PG temperatures. The result shows that the increase of PG H-L temperatures is not linear, but it is possible to approximate it as linear until 30% wt. of recycled binders. Shell bitumen book proposes two equations (“blending laws” Eqs. 1 and 2) for the prediction of the conventional characteristics (penetration and Ring and Ball temperature) for the blend of two bituminous binders with the same nature (same crude oil): log(P) = a × log(P1 ) + b × log(P2 )
(1)
Combined Recycling of RAS and RAP: Experimental and Modeling …
T = a × T1 + b × T2
389
(2)
P and T are respectively the penetration and the Ring and Ball temperatures of the blend, P1 and P2 are the penetration of bitumen 1 and bitumen 2 at the same temperature, T1 and T2 are the ring and ball temperatures of bitumen 1 and bitumen 2 and a and b are the proportions by weight respectively of the bitumen 1 and bitumen 2 in the blend, both laws are widely used in Europe for the determination of the characteristics of a bitumen mixture containing RAP (NF EN 13108-1 annex A). Mangiafico et al. [4] have worked on the variation of 2S2P1D’s model parameters with the variation of RAP binder rate in the blend. The 2S2P1D is a rheological model proposed by Olard and Di Benedetto [5] consisting of two elastic (spring) elements, two parabolic creep elements and one viscous (dashpot) element, each one associated with a corresponding constant of the following Eq. (3): G ∗ (iωτ ) = G 00 +
1+
δ(iωτ )−k
G 0 − G 00 + (iωτ )−h + (iωβτ )−1
(3)
ω is the angular frequency, τ is the characteristic time, accounting for the TimeTemperature Superposition Principle, G0 and G00 are respectively the value of the shear complex modulus at very high frequencies (glassy) and at very low frequencies (static); k, h, δ and β are shape parameters, dimensionless constants. Mangiafico et al. [4] have generalized Eqs. (4) and (5) for the prediction of the parameters of 2S2P1D model. log(Bx% ) = log(B0% ) + x log(BRAP ) − log(B0% )
(4)
Ax% = A0% + x[ARAP − A0% ]
(5)
A and B are generic terms representing the different 2S2P1D parameters following linear (G0 , k, h, δ, β) and logarithmic (τ) tendencies and x is the amount by weight of RAP binder in the mix. It is noted that the equations proposed by Shell for the prediction of penetration Eq. (1) and Ring and Ball temperature Eq. (2) of bitumen mixes have the same form than the two equations proposed by Mangiafico et al. assuming that (Eq. 6): a+b=1
(6)
log(P) = log(P1 ) + b log(P2 )−log(P1 )
(7)
T = T1 + b(T2 −T1 )
(8)
Equations (1) and (2) will be:
390
A. Daoudi et al.
The main objective of this work is to verify the possibility of a generalization of the two blending laws for the prediction of the conventional, physicochemical and rheological parameters. We take as a hypothesis that the linear logarithmic equation is used for parameters depending on temperature (penetration, complex modulus …) and the classic linear equation is used for parameters independent of the temperature (Ring and Ball temperature, CO indexes …). A third term is added to describe the RAS binder in the blends Eqs. (9) and (10). log(Bx% ) = log(B0% ) + x[log(BRAP ) − log(B0% )] + y[log(BRAS ) − log(B0% )] (9) Ax% = A0% + x[ARAP − A0% ] + y[ARAS − A0% ]
(10)
A and B are generic terms representing the different parameters from conventional tests, physicochemical tests, rheological tests and the 2S2P1D model parameters. x and y are respectively the content by weight of RAP binder and RAS binder in the mix.
3 Materials and Methodology For this work, a non-modified PG 58-34 was used as virgin binder. The penetration and the Ring and Ball temperature of this bitumen are respectively 153 ± 1 1/10 mm and 43 ± 0.1 °C. The RAP material came from Montréal region and contained 5.44% wt. of bitumen (13 ± 1 1/10 mm for penetration and 80 ± 0.1 °C for Ring and Ball temperature). The RAS material was from post consummation and contained 22.9% wt. of bitumen (2 ± 1 1/10 mm for penetration and 135 ± 0.1 °C for Ring and Ball temperature). The Asphalt analyzer was used for the extraction of RAP binder and RAS binder, a specific procedure was applied for RAS because of its high binder content. Twelve blends were prepared with the three binders (virgin PG 58-34, RAP binder, RAS binder) by varying the percentage of recycled materials. Three cases were studied, given in Table 1. Conventional tests were performed according to the European standards EN 1426 and EN 1427. For the physicochemical parameters, two tests were achieved: FourierTransform Infrared Spectroscopy (FTIR) test was used to measure and compare the ageing of mixes with the determination of CO indexes (carbonyl indexes). Then, the differential scanning calorimetry (DSC) test involved studying glass transition temperature of materials. The rheological characterization tests were carried out using a Kinexus Pro+ (Netzsch) dynamic shear rheometer (DSR) at eleven temperatures from −35 to 65 °C and a frequency sweep between 0.1 and 10 Hz.
Combined Recycling of RAS and RAP: Experimental and Modeling …
391
Table 1 Bitumen mixes tested RAP binder/virgin binder
RAS binder/virgin binder
Virgin binder/RAP binder/RAS binder
RAPB/VB (%)
RAPB/VB (g)
RASB/VB (%)
RASB/VB (g)
RAPB/RAS/VB (%)
RAPB/RASB/VB (g)
15/85
10.5/59.5
3.0/97.0
2.1/67.9
13.5/1.5/85.0
9.5/1.1/59.5
20/80
14.0/56.0
4.5/95.5
3.2/66.9
17.0/3.0/80.0
11.9/2.1/56.0
30/70
21.0/49.0
6.0/94.0
4.2/65.8
25.5/4.5/70.0
17.9/3.2/49.0
40/60
28.0/42.0
–
–
30.0/5.0/65.0
21.0/3.5/45.5
–
–
–
–
34.0/6.0/60.0
23.8/4.2/42.0
4 Results and Discussion The penetration and Ring and Ball results are reported in Fig. 1. The more we add recycling materials, the more we have a hardening effect on the virgin binder. This hardening follows a linear trend with a very good R2 . The predicted points (PRED) were calculated using Eq. (9) for penetration and Eq. (10) for R&B temperature. The prediction seems to be quite good for mixes containing RAP and RAS until 40% and for all the other mix. The tendency to overestimate the needle penetration values is probably related with the degree of blending between the three bituminous binders. The results of FTIR and DSC tests are presented in Fig. 2. For FTIR the experimental points show a linear increase of the ageing C=O index with increasing the amount of recycled materials with a very good regression coefficient (0.96). The same tendency is noticed for the glassy temperature (DSC test). The predicted points were calculated using Eq. (10) and seem to describe the experimental points well. Figure 3 shows the experimental results of shear complex modulus tests and the modeling with 2S2P1D. The calibration of 2S2P1D model was performed by minimizing the error between the experimental points and the points from 2S2P1D (the blue curves). For all the mixes tested the error is ±5% for |G*| and ±3° for the phase angle.
Fig. 1 Variation of conventional parameters with the variation of recycling rate
392
A. Daoudi et al.
Fig. 2 Variation of physicochemical parameters with the variation of recycling rate
Fig. 3 DSR results in Cole-Cole and master curve representations for the blend containing 15% of RAP binder
The orange points were calculated using Eq. (9) and seem to describe the experimental points (the blue points) very well. The shape parameters (k, h, δ, β) for 2S2P1D model were predicted using Eq. (10) and the parameter τ (2S2P1D PRED) were predicted using Eq. (9). The results describe very well the experimental points and the predicted points with ±5% for |G*| and ±3° for the phase angle until 45 °C with a maximum error of |G*| of around 30%. It was very difficult to test the RAS binder especially at low temperatures and high frequency, so we tried to use the back calculation from Eqs. (9) and (10) to predict the rheological behavior of RAS binder.
5 Conclusion and Perspectives The main objective of the work was to verify the application of two blending laws for the prediction of conventional, physicochemical and rheological characteristics. Three cases of blending were studied namely: Virgin Binder + RAP Binder, Virgin Binder + RAS Binder and Virgin Binder + RAP Binder + RAS Binder. The proposed mixing laws allow:
Combined Recycling of RAS and RAP: Experimental and Modeling …
• • • •
393
Good prediction for RAP binder until 40% Acceptable prediction until 6% RAS Good prediction for VB + RAPB + RASB until 40% recycling Good prediction of the parameters of 2S2P1D model.
The next step of this work is the prediction of the linear viscoelastic (LVE) behavior of asphalt mixes containing a mix of RAP and RAS (virgin materials + RAP + RAS) from the rheological behavior of the corresponding binder and to verify the evaluation of the amount of “active” binder from recycled materials.
References 1. Zhao, S., Huang, B., Shu, X., Woods, M.E.: Quantitative characterization of binder blending: how much recycled binder is mobilized during mixing? Transp. Res. Rec. 2506(1), 72–80 (2015) 2. Navaro, J., Bruneau, D., Drouadaine, I., Colin, J., Dony, A.: Caractérisation d’un mélange de liants bitumineux par dissolution en couche mince d’un enrobé recyclé (2010) 3. Im, S., Zhou, F.: Field Performance of RAS Test Sections and Laboratory Investigation of Impact of Rejuvenators on Engineering Properties of RAP/RAS Mixes (No. FHWA/TX-14/0-6614-3). Dept. of Transportation, Research and Technology Implementation Office, Texas (2014) 4. Mangiafico, S., Di Benedetto, H., Sauzéat, C., Olard, F., Pouget, S., Planque, L.: New method to obtain viscoelastic properties of bitumen blends from pure and reclaimed asphalt pavement binder constituents. Road Mater. Pavement Des. 15(2), 312–329 (2014) 5. Olard, F., Di Benedetto, H.: General “2S2P1D” model and relation between the linear viscoelastic behaviours of bituminous binders and mixes. Road Mater. Pavement Des. 4(2), 185–224 (2003)
Combined Use of Natural Rubber, Biomass and Plastic Wastes in Bitumen Modification and Flexible Pavement Construction Swapan Kumer Ray, Riyadh Hossen Bhuiyan, Muhammad Saiful Islam, Md. Jaynal Abedin, Zahidul Islam, and Rashed Hasan Abstract A novel bitumen modification process has been developed using lower grade natural rubber-ribbed smoked sheet (NR-RSS) and soda lignin (SL) isolated from waste green coconut fiber. At first, natural rubber along with rubber additives was mixed in toluene and then soda lignin was added. Obtained mixture was characterized by rotational viscometer, Fourier transform infrared spectrometer, universal testing machine, simultaneous thermal analyzer, etc. The mixture was then blended with 60–70 penetration grade bitumen (B). Physical, spectral, microscopic and thermal properties of modified bitumen (MB) were evaluated. It was shown that the prepared natural rubber-lignin mixture significantly improved the physical properties of bitumen as required for flexible pavement construction in tropical countries. Marshal stability was improved by 25% in comparison to 60–70 penetration grade bitumen. Based on the result obtained, an experimental pavement was constructed using modified bitumen and waste polyethylene-coated aggregates (Pav-MB-PCA). For comparison, another pavement of similar size was constructed with 60–70 penetration grade bitumen and non-coated aggregates (Pav-B-NCA). More than two years observation showed better performances of the pavement made with modified bitumen and waste polyethylene coated aggregates. All findings have been discussed in detail. Keywords Bitumen · Natural rubber ribbed smoked sheet · Waste green coconut fibre · Waste polyethylene · Marshal stability · Flexible pavement
1 Introduction Most of the tropical countries have already faced the impact of climate change on the flexible pavement. A significant variation in the daily and seasonal temperature and prolonged rainfall and thus water clogging induces early development of distress condition of bituminous pavement. Adaptive maintenances, such as construction of S. K. Ray (B) · R. H. Bhuiyan · M. S. Islam · Md. J. Abedin · Z. Islam · R. Hasan Fibre and Polymer Research Division, Bangladesh Council of Scientific and Industrial Research, Dhaka 1205, Bangladesh © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_50
395
396
S. K. Ray et al.
permeable pavements, use of modified binders and improved routine maintenances have been recommended by several reports [1]. The incorporation of different polymers and rubbers into bitumen can improve some properties, e.g. improved flexibility, elasticity at low temperature, thermal and crack resistance, moisture resistance, etc. and thus increased service life, etc. [2, 3]. Typically, polymer or rubber modified bitumen are more viscous compared to unmodified bitumen and improve adhesive bonding among the aggregates. Natural rubber could be introduced into bitumen by direct fusion or in presence of suitable solvent(s). But the dispersion capability of rubber is poor in bitumen and thus special measures need to be taken obtaining homogeneous rubber modified bitumen [2]. Some investigations were also carried out on the use of lignin, a highly complex phenolic biopolymer, to prevent bitumen oxidation or hardening [4, 5]. Lignin is a wonderful aromatic biopolymer due to its abundance and relatively low cost [6]. It shows the resistance to oxidation, resistance to water, absorption of ultraviolet radiation, etc. The effect of lignin on the physicomechanical properties and thermo-oxidative degradation resistance of natural rubber has been investigated [7, 8]. The waste thermoplastics, e.g. polyethylene, polypropylene, etc. could be used in flexible pavement construction by adding them at a certain percentage with aggregates and reported it as a good strategy for utilization of these wastes [9]. So, proper management and utilization of waste thermoplastics in asphalt pavement construction can provide an important outlet for such materials to keep the environment green. The aim of this research work was to examine the properties of natural rubber-lignin composite and its application in bitumen modification. Another objective was to evaluate the moisture and heat resistant properties of flexible pavement constructed with modified bitumen and waste thermoplastic coated aggregates. The overall work might be considered as a solution of proper polymeric waste management and their effective utilization in sustainable pavement construction.
2 Materials and Methods 2.1 Materials 60–70 penetration grade bitumen (B) was used to prepare modified bitumen (MB). Natural rubber ribbed smoked sheet (NR-RSS), waste green coconut fibre, waste high density polyethylene (HDPE) shopping bags, standard aggregates, e.g. granite stone of different sizes and sand were used in this work. Other chemicals used in the experiment were of reagent grade.
Combined Use of Natural Rubber, Biomass and Plastic Wastes …
397
Table 1 Preparation of bitumen modifier (BM) Natural rubber (g)
Toluene (ml)
125
1250
Rubber additives (g)
Soda lignin (g)
Temperature (°C)
Stirring speed (rpm)
Time (h)
Name of sample
25
1800–2000
4
BM-1
0
0
15
0
BM-2
15
12.5
BM-3
2.2 Preparation of Soda Lignin The dried coconut fiber was pre-treated with dilute sulfuric acid and then digested with 3% sodium hydroxide solution at 140 °C for 2 h. After cooling, the slurry was filtered and black liquor was collected. pH of the black liquor was reduced to 2.5 using dilute sulfuric acid and precipitated lignin was collected, purified and vacuum dried (yield 35.32%). Analytically milled soda lignin (SL) powder (particle sizes 230
>230
ASTM D 92
Specific gravity@ room temperature
1.02
0.99
ASTM D70-18
Loss on heating (%)
90 °C. However, there is no general consensus about this classification among experts [2]. Statistical evaluation of binders’ properties, performed on 34 RAs in Germany [3] revealed that 12% of binders had a softening point higher than 80 °C and further 18% had a value between 70 °C and 90 °C. The testing thus found that almost one third of RAs contains binders which could potentially be problematic. The above mentioned information is related to especially empirical tests which are not necessarily sufficient for describing the behavior of asphalt binders [4]. Instead of needle penetration, binder assessment at medium temperatures can be done using the VET parameter (Viscoelastic temperature) or Tc (Critical temperature difference), while low-temperature behavior can be tested using critical temperatures calculated for Stiffness, S(60) TcS and for m-value, m(60) Tcm .
2 Research Objectives Aim of this paper is to describe and characterize empirical and some functional properties of binders recovered from RA with their subsequent classification based on degree of ageing.
3 Materials and Methods The tested RAs were obtained from a total of 5 different asphalt plants. These samples were labeled as R1–R10. Samples were obtained in accordance with EN 932-1. Binder extraction was performed in accordance with EN 12697-1 followed up by a distillation according with EN 12697-3. Empirical properties of asphalt binders were assessed by needle penetration test (EN 1426) and softening point test (EN 1427).
Critical Reclaimed Asphalts Binders’ Properties from Empirical …
523
Functional properties were determined using DSR (dynamic shear rheometer) in accordance to EN 14770 and BBR (Bending Beam Rheometer) based on EN 14771. VET as well as the corresponding complex shear modulus (GVET ) were determined at a frequency of 1.59 Hz. VET is defined as transition temperature when the phase angle (δ) value reaches 45° at a given frequency. The elastic component of the complex shear modulus (G ) is thus equal to the viscous component (G ) of the complex shear modulus. It was found that VET parameter correlated with the occurrence of the cracks in the road. The value of VET and corresponding GVET is dependent on the binder grade and production procedure of the binder [5]. The determined low-temperature characteristics in BBR were creep stiffness S(60) and relaxation rate (m-value) m(60) . Critical temperatures were calculated for S(60) (TcS = 300 MPa) and m(60) (Tcm = 0.3). Subsequently, the Tc parameter was derived. Critical threshold values for this parameter are based on the ductility test at 15 °C at 1 cm/min. The parameter is, however, always evaluated based on the BBR measurements (Tc = TcS −Tcm ). Parameter Tc was originally developed to describe the embrittlement of the binder. The rate of embrittlement characterized by Tc is believed to indicate the occurrence of block cracking but it can be linked also to other type of distresses such as fatigue cracking [6].
4 Test Results and Discussion The determined needle penetration values ranged between 12 (0.1 mm) and 25 (0.1 mm), while the softening points ranged between 57.8 °C and 79.6 °C. The calculated penetration index (PI) was therefore in the range of -0.90 and 1.37. A total of 7 binder samples could be classified as P15 (R1, R3, R4, R5, R6, R7 and R9) and only two samples as S70 (R1 and R7), see Fig. 1. The binder R7 differed significantly from all the other samples in that. It was the only one with PI > 0.7 (it can be assumed that the R7 binder was modified). In total 7 out of 10 binders could be labeled as so-called critical belonging to the P15 or S70 category. (Table 1) Figure 2 shows the relationship between GVET and VET. Upon ageing the value of VET increases while the value of GVET decreases, which increases the risk of crack formation. Figure 2 also shows classification degrees of binder ageing, adapted from [5] and also provided in Table 1. With the exception of R7 binder, which could be classified as very aged (+++), and binders R2 and R10, which in contrast were aged very little (–), all the remaining binders had a VET in the range between 21.5 and 24.4 °C. Based on the VET, these binders would be classified as medium aged (++). The VET interval was very small (2.9 °C) though. It is therefore useful to include the value of GVET in the subsequent assessment. The values GVET laid in the interval between 6259 kPa and 11849 kPa. This group could then be further divided into two subgroups based on the ageing degree: slightly (R4 and R5) (±) and medium (R1, R3, R6, R8 and R9) (++) aged. It is also important to note that R1 binder, which in terms of empirical tests was evaluated as critical, was assessed as medium-aged in terms of VET.
524
T. Koudelka et al.
Fig. 1. Empirical tests results Table 1. Binder classification matrix, VET & GVET [5]
G*VET (Pa)
VET (°C) 0 to 20
20 to 40
>107
–
±
5 × 106 to 107
+
++
+++
106 to 5 × 106
++
+++
++++
Fig. 2. Binder behavior at intermediate temperatures through VET & GVET
40 to 60
Critical Reclaimed Asphalts Binders’ Properties from Empirical …
525
Fig. 3. Critical parameters from BBR measurements
From the perspective of low critical temperatures, it was proven that the ability of relax was a decisive aspect. When assessing the behaviour based on the m-value, the binders should have a higher critical temperature compared to the assessment based on the creep stiffness, as can be seen in Fig. 3. The only exception was sample R4. Critical temperatures TcS ranged between −17.9 and −9.4 °C, while the temperatures based on Tcm were in the range of −16.1 and −7.2 °C. From the perspective of the TC parameter it was shown that worst behavior (largest difference between the critical temperatures TcS a Tcm ) was observed for the binder with the highest softening point (probably modified). Overall 7 out of the 9 assessed binders had parameter TC > –2.5 °C, one binder was in the interval −5 °C < TC < − 2.5 °C (R6) and one binder had –5 °C > TC (R7). Therefore, it cannot be said that reclaimed asphalt binders would present a higher risk of crack formation at medium temperatures in most cases with regards to the TC parameter because the difference between critical temperatures was sufficiently small.
5 Conclusions Based on authors’ experience as well as on literature survey, a table evaluating the degree of ageing was constructed, see Table 2. When assessing the binder according to penetration, most binders (7 out of 10) would be classified as the so-called critical binders (pen. < 15 (0.1 mm)). On the other hand, the assessment from the perspective of VET, was different. When assessing VET and GVET it was possible to classify only one binder as critical (+++ or ++++). Remaining binders could be classified as medium (++) or less aged (±, –). Also
526
T. Koudelka et al.
Table 2. Table for the evaluation of the degree of ageing [3, 5, 6] Parameter Pen. SP
2
[0.1 mm] ≥35 rating [-] 1 [°C]
VET &
≤60
Taken from 3
4
5
25 ≤ x 4500). The macrostructure of the mix types (a-RM, b-SRM, c-SAM, d-AM) is shown in Fig. 1b. A total of 97 cylindrical samples were fabricated for model development. The data sizes were similar to that of researchers [4], who used approximately 100 samples for the prediction of flow numbers for asphalt concrete.
Fig. 1. a Aggregate particle and AIMS angularity description b Mixtures macro-structure
562
L. P. Leon
Table 1. Design graded mix design Sieve (mm) % passing
12.5
9.5
4.75
2.36
0.6
0.3
0.15
0.075
Lower limit
80
70
50
35
18
13
8
4
Mix design
93
86
67
47
24
15
10
6
Upper imit
100
90
70
50
29
23
16
10
The underline was a means of highlighting the final mix design used in the research and showing that it was within the specified ranges of the local gradation standard
2.2 Performance Testing Evaluation Asphalt concrete mixture under loading is exposed to elastic, visco-elastic, plastic and visco-plastic deformation. The repeated load axial creep tests (RLAT) were executed in accordance to BS EN 12697-Part 25. Test measured the resistance of HMA to plastic flow and permanent deformation as cyclic axial load was applied to cylindrical samples. In the unconfined creep test, an axial stress (70 and 200 kPa) is applied for 3600 s or until failure. The test is conducted at specified temperatures of 25 and 45 °C.
3 HMA Permanent Deformation Model RLAT database included information on the coarse aggregate angularity index (CAI), density (D), height (H), air voids (V a ), voids in mineral aggregate (VMA), temperature (T ), axial loading (σ ), intercept (a), slope (b), loading application (N), and output variable permanent deformation strain (εp ). The inputs of Model 2 were similar to that of the Rutting Rate Model and Asphalt Institute Rutting Model respectively.
3.1 Multiple Non-linear Regression Models (MNLR) Multiple non-linear regression analysis was implemented for expressing the relationship between mix properties and permanent deformation. Equation (1) prediction model resulted in coefficient of determination (R2 ) of 0.891 with estimation errors of 3% and 10% for over and under predictions respectively. Equation (2) prediction model resulted in R2 of 0.771 with estimation errors of 12% for over and under predictions respectively. 0.16 0.26 0.77 1.42 0.14 MNLR − Model 1 ε p = 12.18 H σC AIa0.01 Tb0.58 N
(1)
MNLR − Model 2 ε p = 4.23C AI 0.03 a 0.90 b1.78 N 0.35
(2)
Development of Permanent Deformation Models …
563
3.2 Soft Computing Models Soft computing methods are more intelligent in recognizing the patterns and are capable of dealing with a multiplicity of factors which are required to augment the prediction models and integrate laboratory and/or field. One such technique is the Gene expression programming (GEP), which is a hunt strategy, which includes computer programs (e.g., numerical articulations, decision trees, and polynomial builds). Equation (3) GEP model calculated permanent deformation coefficients of R2 for training and testing model of 0.957 and 0.632 respectively with mean-squareerrors (MSE) of 0.055 and 0.915 respectively. Fitness was 827 and 617 respectively with estimation errors of 1.5% for over predictions of ranges 1.5–2.5% strain. Equation (4) GEP prediction model calculated permanent deformation coefficients of R2 for training and testing model were 0.871 and 0.728 respectively. The MSE were 0.162 and 0.706 respectively. Fitness was 735 and 647 respectively with estimation errors of 5% for over predictions of ranges 1.5–2.5% strain and 8% for under predictions of ranges above 3.7% strain. GEP—Model 1 ⎡ ⎤ T
10 a b(σ + 6.8) −0.98− a − 0.46 + ε p = ⎣ (H +T )HN ⎦ + 23.5 b σ +9.54
H+N σ + + b b2 + (3) C AI ± 7.55bT − 72.1b 0.86/a GEP—Model 2 ε p = 7.93b + 2a 2 + 2b2 + ln(a) √ √ √ N b a + 10b a + 2b − a + ln ln 3.22
−8.15b √ C AI − 10 + N
(4)
3.3 Parametric and Sensitivity Analysis The vigour of the predictive equations is carried out by examining how well the predicted solutions concur with the underlining physical and theoretical behaviour of asphalt to changes of mix type, aggregates, loading and environmental conditions. As shown in Fig. 2a, deformation behaviour decreases as the coarse aggregate in the mix becomes more angular. This is expected as the increase in aggregate angularity
564
L. P. Leon
Fig. 2. Permanent deformation strains behaviour due to a CAI b T c σ d N
provides an increase in aggregate interlock, which is related to resistance to deformation. Any mixture that provides restriction of aggregate particle movements will also decrease its deformation strains. When temperature in a flexible pavement increase so does the rutting potential as shown only in GEP model 1 (Eq. 3) of Fig. 2b. As temperature increased above 35 °C there is an increase in the deformation of the mix. This is because binder become viscous with increases temperature resulting in the aggregates supporting most of the traffic stress which leads to increase in the deformation (rutting). Loading stress and loading applications caused by traffic loading as shown in Fig. 2c and d has noteworthy effect on the deformation behaviour. As loading stresses and load repetitions increases the permanent deformation strain increases.
4 Conclusions Existing permanent deformation models are based on linear regression analysis. The results of the study provide insights into the contribution of imaging technique angularity indices on the estimations of permanent deformation that are interesting in their own right. Multiple non-linear regression (MNLR) and gene expression programming (GEP) were applied as methods to develop explicit equations for the
Development of Permanent Deformation Models …
565
prediction of permanent deformation strain. With some improvements the laboratory model could be used to predict and rank mixes based on their permanent deformation behaviour. Current minimum criteria for aggregate angularity in HMA pavement mixtures influences mix designs, construction costs and asphalt performance, hence, an aggregate angularity criterion in rutting estimations are important for road and highway organisations.
References 1. Allen, D.L., Robert C.D.: A computerized analysis of flexible pavement rutting behavior (PAVRUT). UKTRP-85–31. Lexington, KY, Kentucky Transprotation Center (1986) 2. Bessa, I.S., Branco, V.T.F.C., Soares, J.B., Neto, J.A.N.: Aggregate shape properties and their influence on the behaviour of hot mix asphalt. J. Mater. Civ. Eng. 27(7), 04014212 (2014) 3. Chen, J.-S., Chang, M. K., Lin, K.Y.: Influence of coarse aggregate shape on the strength of asphalt concrete mixtures. J. Eastern Asia Soc. Transport. Stud. 6, 1062–1075 (2005) 4. Gandomi, A.H., Alavi, A.H., Mirzahosseini, M.R., Nejad, F.M.: Nonlinear genetic-based models for prediction of flow number of asphalt mixtures. J. Mater. Civ. Eng. 23(3), 248–263 (2011) 5. Ghasemi, P., Mohamad, A., Derrick, K.R., Williams, R.C.: Principal component analysisbased predictive modeling and optimization of permanent deformation in asphalt pavement: elimination of correlated inputs and extrapolation in modeling. Struct. Multidiscip. Optim. 59(4), 1335–1353 (2018) 6. Hu, J., Pengfei, L., Dawei, W., Oeser, M.: Influence of aggregates’ spatial characteristics on air-voids in asphalt mixture. Road Mater. Pavement Des. 19(4), 837–855 (2018) 7. Huang, B., Xingwei, C., Xiang, S., Masad, E., Mahmoud, E.: Effects of coarse aggregate angularity and asphalt binder on laboratory- measured permanent deformation properties of HMA. Int. J. Pavement Eng. 10(1), 19–28 (2009) 8. Huang, C.-W., Al-Rub, R.K.A., Masad, E.A., Little, D.N.: Three-dimensional simulations of asphalt pavement permanent deformation using a nonlinear viscoelastic and viscoplastic model. J. Mater. Civil Eng. 23(1), 56–68 (2010) 9. Kenis, W.J.: Predictive design procedures-a design method for flexible pavements using the VESYS structural subsystem. In: Volume I of Proceedings of 4th International Conference on Structural Design of Asphalt Pavements, Michigan: University of Michigan (1977) 10. Lee, C-J., White, T.D., West, T.R.: Effect of fine aggregate angularity on asphalt mixture performance. FHWA/IN/JTRP-98/20, West Lafayette, Indiana: Purdue University Indiana Department of Transportation (2000) 11. Li, Q., Xiao, D.X., Wang, K.C., Hall, K.D., Qiu, Y.: Mechanistic-empirical pavement design guide (MEPDG): a bird’s-eye view. J. Modern Transp. 19(2), 114–133 (2011) 12. Little, D., Button, J., Jayawickrama, P., Solaimanian, M., Hudson, B.: Quantify Shape, Angularity and Surface Texture of Aggregates Using Image Analysis and Study Their Effect on Performance. Report No. 0–1707- 4. College Station, Texas: Texas Transportation Institute (2003) 13. Majidzadeh, K., Aly, M., Bayomy, F., El-Laithy, A.: Implementation of A Pavement Design System. Final Report EES 579 (1) and (2). Columbus, OH: Ohio State University (1980) 14. Masad, E., Fletcher, T.: Aggregate Imaging System (AIMS): Basics and Applications. Report 5–1707–01–1. College Station, Texas: Texas Transportation Institute (2005) 15. Radziszewski, P.: Modified asphalt mixtures resistance to permanent deformations. J. Civ. Eng. Manag. 13(4), 307–315 (2007) 16. Singh, D.: A Laboratory Investigation and Modelling of Dynamic Modulus of Asphalt Mixes for Pavement Applications. Doctoral Dissertation, The University of Oklahoma (2011)
Discrete Modelling of 2PB Complex Modulus Test of Hot Asphalt Mixes Using Irregular Aggregates Juan Carlos Quezada and Cyrille Chazallon
Abstract Asphalt mixtures are complex multiphase materials showing a viscoelastic behavior. In order to assess the mechanical response under traffic of these materials, a typical practice is to perform a complex modulus test. An alternative to laboratory characterization is to simulate numerically the complex modulus test by means a discrete approach. Classical discrete approaches are able to reproduce the mechanical performances of asphalt mixtures, but show some difficulties to model non-spherical particles. In this work, the complex modulus test is reproduced numerically for an asphalt mixture by 3D simulations carried out with the LMGC90 code. A viscoelastic contact law is used to handle the interaction between particles with irregular shapes. The numerical aggregates were generated using the particle size distribution (PSD), the flakiness index (FI), and statistic data of actual aggregates, without using complex imaging technics. Experimental and numerical testing campaigns were conducted for the complex modulus test on trapezoidal samples in a two-point bending (2PB) configuration. The numerical model was able to reproduce the mechanical performances obtained during the experimental tests, regarding the viscoelastic properties of the material. The influence on the mechanical performances of particle shape considering irregular aggregate was analyzed. The proposed model can be used to simulate the mechanical response of road structure under traffic loading concerning rutting, crack propagation and fatigue damage. Keywords Discrete modelling · 2PB complex modulus test · Irregular particles · Burgers contact model
1 Introduction Hot asphalt mixes are complex multiphase systems consisting of aggregates, mastic of bitumen and air. To quantify the viscoelastic properties of these mixtures, such as the norm of the complex modulus |E*| and the associate phase angle , it is usual to J. C. Quezada (B) · C. Chazallon ICUBE, UMR 7357, CNRS, INSA de Strasbourg, Strasbourg, France e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_72
567
568
J. C. Quezada and C. Chazallon
carry out a complex modulus test. Laboratory tests measure the macroscale behavior of asphalt mixes at large-scale, but this approach cannot identify the behavior at the particle-scale. One way around this problem is performing numerical simulations by means a discrete approach. This numerical method model the interaction of independent rigid or deformable bodies in contact. In the last years, the Discrete Element Method (DEM) has been broadly employed to simulate the mechanical performances of asphalt mixes [1–3]. In these studies, the mortar is simulated by a viscoelastic law interacting with particles at contact. Nevertheless, most of the studies have modelled the aggregates using spherical particles. In a granular material, such as asphalt mixtures, the particle shape plays a major role regarding fabric anisotropy, force transmission and friction mobilization [4–7]. In this work, laboratory 2PB complex modulus tests were conducted on trapezoidal samples to identify the viscoelastic properties of an asphalt mixture. Then, numerical 2PB tests were performed using a discrete approach with irregular particles. Finally, the results from numerical complex modulus tests are compared with the experimental data to validate the numerical approach.
2 Experimental Procedures 2.1 Complex Modulus Test The experimental campaign was conducted on trapezoidal samples, using a two point bending (2PB) configuration according to the EN 12697-26:2012 specification [8]. For this campaign, four samples were manufactured using a Semi Coarse Asphalt Concrete (BBSG) 0/10 with a bitumen 35/50 grade, saw-cut from slabs. Table 1 displays the particle size distribution (PSD) of the employed aggregates. Each trapezoidal specimen is 250 mm height, 56 mm at the bottom, 25 mm at the top and 25 mm width, with a total mass M about 0.6 kg, a bulk density of 2289 kg/m3 and a void content of 4.7%. For these experimental trials the frequency values tested were set at 3, 6, 10, 25, and 40 Hz, while the selected temperature values were − 10, 0, 10, 15, 20, and 30 °C. On each sample, a sinusoidal displacement z is applied at the target frequency, with amplitude of 63.10–6 m, corresponding to a strain of 50.10–6 . The maximum force F0 , the maximum deflection z0 and the phase angle were measured during the last ten seconds of the test. To identify the viscoelastic properties of this asphalt mixture, first, the real and imaginary part of E*, E1 and E2 , can be determined using: E 1 = γ F0 /z 0 ∗ cos() + 10−6 μω2
(1)
E 2 = γ ∗ F0 /z 0 ∗ sin()
(2)
Discrete Modelling of 2PB Complex Modulus Test …
569
Table 1 Particle size distribution for the BBSG 0/10 Bréfauchet asphalt Sieve size (mm)
12.5
10
8
6.3
4
2
1
0.5
0.25
0.125
0.063
PSD (%)
100
90
75
58
44
31
20
15
12
9
6.5
where γ is a shape factor and μ the mass factor. Finally, the viscoelastic properties can be determined as |E*|=(E1 2 + E2 2 )1/2 and = arctan(E2 /E1 ).
3 Numerical Procedures 3.1 Sample Preparation The modeling campaign of 2PB tests was carried out using the LMGC90 software, which is dedicated to multiple physics simulation of discrete material and structures [9]. In this numerical campaign, four specimens composed of irregular polyhedral aggregates were prepared, considering the same geometry and total mass of samples used in the experimental tests. The numerical specimens are composed of rigid irregular polyhedral particles. About 10,000 particles are created following the experimental PSD (Table 1) cut at 2 mm, in order to reduce the total quantity of elements in the sample, where the fines are included in the mortar phase. Then, for each particle an ellipsoid is generated, where its surface is defined by the parametric Eqs. (3), (4) and (5): x = S f r cos(θ ) cos(ϕ)
(3)
y = r cos(θ ) sin(ϕ)
(4)
z = r sin(θ ) cos(ϕ)/S f
(5)
where the angles θ and ϕ varies in the ranges –π/2 ≤ θ ≤ π/2 and –π ≤ ϕ ≤ π, r is the half dimension from the PSD and Sf is a shape factor. The Sf value was varied randomly between 1.0 and 1.5. This factor reproduces the elongation and the flakiness of experimental aggregates on the numerical particles, where the flakiness index (FI) is equals to 17%. This procedure generates different ellipsoid surfaces, where are placed randomly eight vertices. This number of vertices was identified visually from 28 experimental aggregates (Fig. 1a) where the average number is equals to 7.77, and the standard deviation is 0.916. This protocol creates convex polyhedrons, based on simplified shapes of actual aggregates, without using complex imaging technics. A snapshot of the numerically generated particles with this protocol is displayed in Fig. 1b.
570
J. C. Quezada and C. Chazallon
Afterwards, the generated particles are poured by gravity into a box composed of six walls with dimensions: 0.25 m length, 0.025 m width and 1.6 m height. The initial density of these particles is 2647 kg/m3 . Here, the coefficient of friction between each body (particles and walls) is reduced to 0.1 with the goal of generating a dense initial sample. The, the generated numerical slab is cut to obtain the same dimensions as the experimental samples. This procedure is carried-out by eliminating the particles outside the trapezoidal volume. Then, each resulting sample is placed and fixed to a bottom and a top plate. The latter simulates the mobile plate in the experimental setup. As the mortar phase is not modelled numerically, the void content on numerical samples is about 34%. Then, the density of particles is increased to 3556 kg/m3 , to reach the total sample mass of 0.6 kg, concentrating the mass of the aggregates and the mortar phase within the particles. In this stage, a viscoelastic contact law was applied between particles to simulate the mortar phase. The contact model chosen was the Burgers model [10, 11], which describes quite well the mechanical performances of asphalt mixtures, such as relaxation, creep and dynamic response [1, 12]. This viscoelastic model is composed of a Maxwell section and a Kelvin-Voigt section, in series. Due to the Maxwell dashpot, this model is not suitable for predicting the mechanical response at very lowfrequencies. Four parameters describe its mechanical behavior: the Maxwell’s stiffness and viscosity Km and Cm , and the Kelvin-Voigt’s stiffness and viscosity Kk and Ck . The Burgers contact law manages the interaction of rigid particles surrounded by a viscoelastic phase [13]. This model links each particle with the neighbouring particles. One snapshot of resulting samples composed of irregular particles is provided in Fig. 1c.
3.2 Complex Modulus Test Modelling The numerical campaign was conducted on four numerical samples, for six temperatures and five frequencies. These tests were performed by applying the same sinusoidal displacement applied on experimental test on the top plate, during five cycles. The displacement and the total reaction at the top plate are mapped throughout the test. Then, the maximum force and displacement, F0 and z0 , are measured to calculate the numerical values of |E*| and , using Eqs. (1) and (2). The time step was set to 5 × 10–5 s in all simulations to ensure numerically stability in resolution. The computational time needed to perform the loading cycles was comprises between 1.7 and 22 h.
4 Calibration of the Numerical Modelling In the aim to calibrate the contact model parameters regarding the experimental data, an analytical macroscopic approach was adopted, based on the Burgers constitutive
Discrete Modelling of 2PB Complex Modulus Test …
571
Fig. 1 a Actual aggregates used in the visual identification of the shape and the number of vertices. b Snapshot of aggregates generated numerically. c Snapshot of a resulting sample
model. For each temperature, the best fit parameters regarding the experimental values of |E*| and were obtained using an optimization algorithm. These parameters are displayed in Table 2. In the aim to validate the proposed numerical approach, Fig. 2 displays the comparison between the experimental and numerical results for the norm of the complex modulus and phase angle isotherms respectively. As expected, experimental and numerical data of |E*| collapse in the same curves. On the other hand, numerical values follow the trend of experimental ones, despite some fluctuations around the average values. Table 2 Burgers model parameters used in the numerical simulations T (°C)
Km (Pa)
Cm (Pa.s)
Kk (Pa)
Ck (Pa.s)
−10
1.20 ×
1.73 ×
2.57 ×
3.15 × 108
0
1.07 × 109
6.36 × 108
8.41 × 109
9.55 × 107
10
9.14 ×
108
1.48 ×
108
2.26 ×
109
2.47 × 107
15
7.32 ×
108
8.82 ×
107
1.42 ×
109
1.68 × 107
20
5.72 × 108
4.79 × 107
8.24 × 108
1.05 × 107
30
3.30 ×
8.55 ×
2.24 ×
3.26 × 106
109
108
109
106
1010
108
572
J. C. Quezada and C. Chazallon
Fig. 2 a Norm of complex modulus isotherms for experimental and numerical tests. b Phase angle isotherms for experimental and numerical tests
5 Conclusions In this paper discrete simulations of complex modulus tests in a 2PB configuration were presented. The numerical specimens were prepared using irregular polyhedral particles. The particle-generation procedure was based on statistic criteria to reproduce the elongation, flakiness and the number of vertices identified from actual aggregates, without using 3D scan or other imaging technics. The mortar phase was modelled using a contact law between polyhedrons based on the Burgers model. The numerical campaign reproduces the experimental tests, where four trapezoidal samples were subjected to five frequencies and six temperatures. The outcomes in this study show that the numerical simulations with irregular polyhedral are able to reproduce the complex modulus properties of bituminous materials. Regarding isotherm curves, the numerical results of these simulations were in a good agreement with experimental data. In conclusion, this preliminary work can be considered as encouraging for the validation of the proposed numerical approach for the study of asphalt mixtures features such as creep, rutting and fatigue cracking performance.
References 1. McDowell, G., Collop, A., Wu, J.: A dimensional analysis of scaling viscosity and velocity in dem of constant strain rate tests on asphalt. Geomech. Geoeng. Int. J. 4(2), 171–174 (2009) 2. Yu, H., Shen, S.: A micromechanical based three-dimensional dem approach to characterize the complex modulus of asphalt mixtures. Constr. Build. Mater. 38, 1089–1096 (2013) 3. Dondi, G., Vignali, V.: Modeling the dsr complex shear modulus of asphalt binder using 3d discrete element approach. Constr. Build. Mater. 54, 236–246 (2014) 4. Alonso-Marroquin, F., Herrmann, H.: Calculation of the incremental stress-strain relation of a polygonal packing. Phys. Rev. E 66(2), 021301 (2002) 5. Lu, M., McDowell, G.: The importance of modelling ballast particle shape in the discrete element method. Granular Matter 9(1–2), 69–80 (2007) 6. Azéma, E., Radjai, F., Dubois, F.: Packings of irregular polyhedral particles: strength, structure and effects of angularity. Phys. Rev. E 87(6), 062203 (2013)
Discrete Modelling of 2PB Complex Modulus Test …
573
7. Souza, L.T., Kim, Y.R., Souza, F.V., Castro, L.S.: Experimental testing and finite-element modeling to evaluate the effects of aggregate angularity on bituminous mixture performance. J. Mater. Civ. Eng. 24(3), 249–258 (2012) 8. En 12697-26: Bituminous mixtures. Test methods for hot mix asphalt. Part 26: Stiffness. (AFNOR ed.) (2012) 9. LMGC90, https://git-xen.lmgc.univ-montp2.fr/lmgc90/lmgc90_user/wikis/home 10. Betten, J.: Creep Mechanics. Springer Science & Business Media (2008) 11. Liu, Y., You, Z.: Determining burgers model parameters of asphalt materials using creeprecovery testing data. In: Pavements and Materials: Modeling, Testing, and Performance, pp. 26–36 (2009) 12. Cai, W., McDowell, G., Airey, G.: Discrete element visco-elastic modelling of a realistic graded asphalt mixture. Soils Found. 54(1), 12–22 (2014) 13. Quezada, J.C., Chazallon, C.: Complex modulus modeling of asphalt concrete mixes using the non-smooth contact dynamics method. Comput. Geotech. 117, 103255 (2020)
Effect of Ageing Kinetics on the Rheological Properties of Bituminous Binders and Mixes: Experimental Study and Multi-scale Modeling C. Somé, J.-F. Barthélémy, V. Mouillet, and Ferhat Hammoum Abstract The oxidative ageing of hot mix asphalt (HMA) and foamed warm mix asphalt (WMA), performed through laboratory accelerated ageing processes is investigated in this study. 50/70 pen-grade bitumen was used to produce the mixes. Four ageing durations (0, 3, 6 and 9 days) were considered to investigate the kinetics of long-term ageing based on a RILEM ageing procedure. The binders of the aged mixes were extracted and characterized. Moreover, RTFOT and PAV tests were carried out on fresh bitumen to simulate the long-term ageing of HMA. These procedures were adapted to simulate the long-term ageing of WMA for which no standardized ageing procedure exist. Rheological tests were performed to determine the stiffness modulus of binders and mixes. The results show that the stiffness modulus increases with the ageing kinetic for both binders and mixes. This increase is more pronounced for WMA and their binders than HMA. Furthermore, the equivalence between the RTFOT and RTFOT + PAV ageing procedures performed on bitumen and the RILEM procedure carried out on loose mixes was discussed for both HMA and WMA. Finally, a multi-scale model was implemented to predict stiffness moduli and phase angles of the mixes based on bitumens’ and aggregates’ properties. The results show a relatively good agreement between the multi-scale model and the experimental data curve-fitted with the well-known modified Huet-Sayegh model. Keywords Bituminous mixture · Ageing · Stiffness modulus · Multiscale modeling
C. Somé · J.-F. Barthélémy (B) Cerema IdF/ITM, Équipe de recherche DIMA, 120 rue de Paris, 77171 Sourdun, France e-mail: [email protected] V. Mouillet Cerema, Équipe DIMA, av. Albert Einstein - CS 70 499, 13593 Aix-en-Provence, France F. Hammoum Ifsttar, Laboratoire MIT, Route de Bouaye - CS 5004, 44344 Bouguenais, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_73
575
576
C. Somé et al.
1 Introduction Bituminous mixes are viscoelastic complex composite materials build up from a mixture of aggregates, bitumen, fillers and additives. As such, their end performances entirely depend on their component’s percentage and properties, their interaction with each other and the evolution of their properties other time. Many parameters have a significant influence, such as aggregate size and grading, filler size distribution, bitumen properties and the interaction between bitumen and aggregates. Understanding the influence of all these parameters is a challenging task and requires a multi-scale modeling to assess the effect of each component [1, 2]. Moreover, the ageing induced by the manufacturing conditions and the field climatic modifies the chemical composition of the binder that makes the mixes’ performances to evolve. Indeed, the ageing of mixes occurs in two main stages: The short-term ageing is caused by the manufacturing in the mix plant, the transport and the paving during which the ageing process is mainly associated to a departure of the volatile compounds of bitumen and the long-term ageing during pavement life is influenced by climatic conditions, and road traffic and is related to steric hardening, oxidation and UV rays which affects only the 12 mm upper layer of the roads. In order to simulate the ageing that has taken place over several years, accelerated ageing methods were developed in laboratory. Among these methods, one can distinguish the ageing processes on bitumen RTFOT (Rolling Thin Film Oven Test) followed by PAV (Pressure Aging Vessel) which is claimed to reproduce 7– 10 years natural ageing of a asphalt pavement [3]) and those carried out on loose mixes. The prediction of field ageing based on laboratory ageing protocols is still questionable and imprecise (from 5 to 10 years). In general, the predictions are more or less based on the analysis of one of the following indicators: carbonyl or sulfoxide indexes [4], rheological properties, mechanical performance but never all indicators together. More than twenty ageing procedures on the mixes can be found in the literature [5] but to our knowledge there is no equivalence between these latter and the RTFOT + PAV ageing methods that can be performed on bitumens. For WMA the ageing protocols are still under investigations. This study has three main objectives: Firstly it aims to compare the ageing kinetics of HMA and WMA based on mixes and the extracted binders rheological properties characterization. Secondly, it aims to implement a multi-scale model capable to predict the rheological properties of both HMA and WMA based on theirs constituents (bitumens, aggregates, air voids) properties. Incidentally, it also allows to compare the RILEM ageing protocol on loose mixes (described CEN TS 12697-52) and the RTFOT + PAV method on bitumen (described in EN 12607-1 (RTFOT) and EN 14769 (PAV)) developed for HMA and to propose an adaptation of these ageing protocols to foamed WMA.
Effect of Ageing Kinetics on the Rheological Properties of Bituminous Binders …
577
2 Materials and Mix Design 2.1 Materials Properties The mixes were produced with a paving grade bitumen, whose properties of needle penetration (EN 1426) and softening temperature (EN 1427) are respectively 57 × 0.1 mm and 49.6 ◦ C, and porphyry aggregates whose Los Angeles (EN 1097-2) and Microdeval (EN 1097-1) properties are 17 and 8%. The mixture volume composition is given in Table 1. For WMA 1% of Cecabase additive was added into the bitumen.
2.2 Ageing of Asphalt Mixture Ageing of hot mix asphalt (HMA): The hot mixes were manufactured respectively at 165 ◦ C and then subjected to different ageing durations from 0 to 9 days. After the mixing, the loose HMA mixture was cool down to 135 ◦ C and spread in thin layer (not more than 2.5 cm) into a tray and placed in a ventilated oven during 4 h to simulate the short term ageing. Then, it was cool down to 85 ◦ C and kept in the oven for 3, 6 or 9 days to reproduce the long term ageing. At the end of the ageing procedure the loose mixes were reheated at 135 ◦ C during 2 h then compacted. The reference hot mix was compacted without long-term ageing and designated H0 while those subjected to long-term ageing were labeled H3, H6, H9 corresponding to the 3, 6 or 9 d ageing. Ageing of foamed warm mix asphalt (WMA): The foamed bitumen was produced with a mixture of hot bitumen, 1% of Cecabase additive and 5.5% of added water which allows to obtain maximum expansion. The WMA were manufactured at 135 ◦ C then immediately spread in thin layer. The short and long term ageing and compaction procedures applied are similar to that of HMA. The mixes manufactured were labeled W0, W3, W6, W9 for reference, 3, 6 or 9 d ageing respectively.
Table 1 Mix composition Constituent Limestone filler 0/2 2/6.3 6.3/10 Bitumen Voids
Volume fraction (%)
Density (kg m−3 )
2.45 22.89 20.78 36.86 12.15 4.87
2700 2720 1040
578
C. Somé et al.
2.3 Ageing of Bitumen Simulation of HMA field ageing by RTFOT and PAV protocols: The RTFOT and PAV ageing protocols were carried out on fresh bitumen to simulate respectively the short-term and long-term ageing that a binder undergoes during HMA production until paving and its natural ageing during the service life. The RTFOT was carried out at 163 ◦ C during 75 min according to EN 12607-1. The long-term ageing of HMA was simulated using the pressure ageing vessel (PAV) test in accordance to the EN 14769. Simulation of WMA field ageing by RTFOT and PAV protocols: The short term ageing of WMA was simulated by the RTFOT at 140 ◦ C other parameters remaining unchanged. This temperature has been selected in order to be the closest to the manufacturing temperature of WMA. The RTFOT and RTFOT + PAV aged bitumens and those recovered from the mixes (H0 to H9, W0 to W9) were tested with dynamic shear rheometer (DSR).
3 Rheological Tests 3.1 Mixes Rheological Tests The complex modulus E ∗ of the mixes was carried out on a prismatic cantilever beam in accordance to EN 12697-26 (Annex A: 2P-PR) varying the temperature from −10 to 50 ◦ C and the frequency 3, 40 Hz apart from the hot mixes which were previously characterize between 5 and 30 ◦ C for mixes H3 and H6 or between 5 and 50 ◦ C for H0 and H9. The master curves of |E ∗ | and the phase angle δ were built by shifting horizontally the isotherms as described in reference [6, 7]. Figure 1 show that the stiffness modulus increases logically with the long term ageing duration for both HMA and WMA. This increase in the modulus induces a decrease in the phase angle as shown in Fig. 2. Moreover, the experimental data of complex modulus have been curve-fitted with the modified Huet-Sayegh model described in [8–10] at a reference temperature of 10 ◦ C. The parameters are given in Table 2 then used to validate the multiscale model presented in Sect. 4.
3.2 Properties of Bitumen Dynamic shear rheometer (DSR) test was performed to characterize the linear viscoelastic properties (l.v.e.) of the asphalt binders. It was performed over a temperature range from 10 to 80 ◦ C and over the frequency sweep from 0.681 to 21.54 Hz. Applying the same approach as before for mixes, the master curves of the norm of
Effect of Ageing Kinetics on the Rheological Properties of Bituminous Binders …
Fig. 1 Master curves of complex phase angle of mixes at Tref = 10 ◦ C
Fig. 2 Master curves of binders at Tref = 10 ◦ C
579
580
C. Somé et al.
Table 2 Modified H-S model parameters for binders and mixes curve-fitted at Tref = 10 ◦ C Mixes
E 0 (MPa)
E ∞ (MPa)
δ
τ
k
h
β
W0
88.7
2.21e+04
2.78
1.45
0.231
0.771
1000
W3
44.9
2.29e+04
2.69
5.27
0.173
0.573
1000
W6
2.10
2.72e+04
2.51
24.5
0.148
0.478
1000
W9
23.8
2.51e+04
2.45
704
0.159
0.515
1000
H0
1.53
4.85e+04
2.34
2.34e-02
0.1
0.446
7902
H3
484
5.00e+04
3.64
9.75e-02
0.107
0.480
404
H6
0.03
3.34e+04
2.30
4.65e-01
0.127
0.425
200
H9
35.5
4.24e+04
3.12
3.07e-01
0.108
0.457
1000
Bitumen
G0
G∞
δ
τ
k
h
β
B0
1.e−05
3.00e+03
10.9
5.94e−05
0.238
0.660
124
B(RT F O T 163)
1.e−05
3.63e+02
2.50
4.67e−03
0.350
0.664
64.4
B(RT F O T 163 + P AV )
1.e−05
4.32e+02
4.82
1.38e−02
0.331
0.646
160
B(H0 )
1.e−05
2.40e+03
10.3
2.42e−04
0.268
0.656
173
B(H3 )
1.e−05
5.47e+02
3.82
2.37e−03
0.296
0.643
114
B(H9 )
1.e−05
1.45e+03
9.92
1.06e−03
0.279
0.653
211
B(RT F O T 140)
1.e−05
3.00e+03
10.3
6.37e−05
0.246
0.645
237
B(RT F O T 140 + P AV )
1.e−05
1.02e+03
8.12
1.47e−03
0.263
0.626
346
B(W0 )
1.e−05
3.00e+03
17.4
4.08e-04
0.230
0.623
269
B(W3 )
1.e−05
4.50e+02
7.20
3.31e−02
0.278
0.627
193
B(W6 )
1.e−05
2.61e+03
23.7
2.61e−03
0.237
0.593
1680
the shear modulus |G ∗ | and phase angle δ were built and represented in Fig. 2. Figure 2 show that the shear modulus increases with the long term ageing duration for all binders recovered from HMA and WMA. This increase in the modulus induces a decrease in the phase angle as shown in Fig. 2. In addition, one can observe the temperature imposed in RTFOT test at 163 ◦ C test simulates quite well the short term ageing of the HMA (H 0) production as already shown by literature [11]. However, the long term ageing simulated by the RTFOT at 163 o C + PAV is more severe than the RILEM long term ageing procedure at 9 days on loose mixes. Regarding the WMA, the RILEM ageing procedure on loose mixes is more severe than the RTFOT at 140 ◦ C and RTFOT at 140 ◦ C + PAV. The |G ∗ | of the bitumen extracted from short term aged mixture W0 is similar to that of the RTFOT at 140 ◦ C+PAV which is supposed to simulate the long term ageing. In addition, it should be noted that the complex modulus of the bitumen recovered from W9 was not tested because this latter was so hard. In addition, these data were also curve-fitted with the modified Huet-Sayegh model and the obtained parameters are given in Table 2.
Effect of Ageing Kinetics on the Rheological Properties of Bituminous Binders …
581
4 Multiscale Modeling Asphalt concrete is a multiscale material that we propose to decompose into three scales (see Fig.3): • the macroscale is made of coarse aggregates coated with bitumen, a mortar phase partially filling the intergranular space and air voids (voids content estimated to be 5.6 ± 0.8% for the tested mixes), • at the mesoscale, the mortar is composed of a mastic matrix embedding sand grains, • at the microscale, the mastic is composed of a bitumen matrix embedding fillers. The aggregates are assumed spherical coated with soft bitumen layer whose thickness e is proportional to the aggregate diameter d according to the relation inspired from Duriez’ relation e = e0 × (d/d0 )0.8 , where d0 = 1 mm, e0 is a setting parameter which varies between 20 and 28 μm according to Duriez [12]. At the mastic and mortar scales, the Mori-Tanaka (matrix/inclusions) scheme was used while the generalized self consistent scheme was applied at the mixture scale as in Ref. [1]. At low frequency and high temperature, the bitumen behaves like a Newtonian fluid therefore, the model based on coated aggregate by thin bitumen layer fails to predict the complex modulus of the mix as previously mentioned [13]. Moreover, stiffness is needed to account the contact between aggregates when the binder becomes liquid. Thus, according to an idea proposed in [13], the contact is modeled by a juxtaposition of bitumen thin layer (of surface fraction χ ) and actual grain contacts (of surface fraction (1 − χ )). The grain contact does not allow interpenetration of grains but only relative sliding so that it is characterized by a linear spring model relating the local shear stress to the tangential displacement jumps thanks to a shear stiffness kt . For χ = 1, the grains are entirely surrounded by a thin layer of bitumen, which implies a total loss of stiffness at high temperature or low frequency. On the contrary if χ < 1 such a model is consistent with the idea of connexity of the granular skeleton even at high temperature or low frequency. The aggregates linear isotropic elastic properties are E = 62 GPa, ν = 0.2 (this latter was assumed in this study). The bitu3K E ∗ men bulk modulus K was assumed constant while the shear modulus μ∗ = 9K −E ∗ depends on ω. The parameters kt , χ and α have been determined by minimizing the objective function between the multiscale prediction and the data fitted with the modified H-S model whose parameters are given in Table 2. Figure 4a comparison between the multiscale and the modified H-S models. The results show a relatively good agreement for kt between 10 and 410 MPa mm−1 from softest to stiffest mixes, χ 0.1 and 0.1 ≤ e0 ≤ 0.4 µm. For readability purposes, only a few results have been represented.
582
C. Somé et al.
Fig. 3 Multiscale model for asphalt mix
Fig. 4 Comparison between multi-scale model (hom), modified H-S model and experiments (ex p)
5 Conclusion The effect of HMA and WMA long term ageing kinetic on the l.v.e. properties of binders and mixes has been investigated. Four ageing durations (0, 3, 6 and 9 d) were considered to investigate the kinetics of long-term ageing based on a RILEM procedure. The results show that the mixes stiffness modulus increases highly with the ageing duration specially for WMA. Moreover, RTFOT and PAV ageing procedure have been also performed on fresh bitumen to simulate HMA short and long term ageing through based on their l.v.e. analysis. Furthermore, RTFOT and PAV ageing procedure were adapted to reproduce ageing of WMA by reducing the RTFOT protocol temperature from 163 to 140 ◦ C. The results showed an increase of the binders’ stiffness with the ageing duration, consistent with the mixes properties. In addition, a multiscale model has been implemented to predict the l.v.e. of the mixes based on those of the constituents. The model predict relatively well the l.v.e. properties of the mixes, by means of good calibration of the contact parameters kt , χ and α. These latter can be well calibrated by performing laboratory experiments in a wide range of temperatures. Finally, in this study the bitumen Poisson’s ratio ν varies between 0.4 and 0.5 and is assumed following the modified HS model given in [13]. As a perspective, an experimental measurement of the bitumen and mix Poisson’s ratio ν can allow improving the multiscale model.
Effect of Ageing Kinetics on the Rheological Properties of Bituminous Binders …
583
References 1. Pichler, C., Lackner, R., Aigner, E.: Generalized self-consistent scheme for upscaling of viscoelastic properties of highly-filled matrix-inclusion composites—application in the context of multiscale modeling of bituminous mixtures. Compos. Part B 43, 457–464 (2012) 2. Alam, S.Y., Hammoum, F.: Viscoelastic properties of asphalt concrete using micromechanical self-consistent model. Arch. Civil Mech. Eng. 15, 272–285 (2015) 3. Xiao, F., Amirkhanian, S.N., Karakouzian, M., Khalili, M.: Rheology evaluations of WMA binders using ultraviolet and PAV aging procedures. Construct. Build. Mater. 79, 56–64 (2015) 4. Kambham, B.S., Ram, V.V., Raju, S.: Investigation of laboratory and field aging of bituminous concrete with and without anti-aging additives using FESEM and FTIR. Constr. Build. Mat. 222, 1193–202 (2019) 5. Kim, Y.R., Rad, F.Y., Elwardany, M., Underwood, S., Farrar, M.J., Glaser, R.R.: Long-term aging of asphalt mixtures for performance testing and prediction. Project NCHRP 09-54 Report (2013) 6. Chailleux, E., Ramond, G., Such, C., de La Roche, C.: A mathematical-based master-curve construction method applied to complex modulus of bituminous materials. Road Mater. Pavement Design 7(sup1), 75–92 (2006) 7. Booij, H.C., Thoone, G.P.J.M.: Generalization of Kramers-Kronig transforms and some approximations of relations between viscoelastic quantities. Rheol. Acta 21(1), 15–24 (1982) 8. Olard, F., Di Benedetto, H.: General “2S2P1D” model and relation between the linear viscoelastic behaviours of bituminous binders and mixes. Road Mater. Pavement Des. 4(2), 185–224 (2003) 9. Such, Ch.: Analyse du comportement visqueux des bitumes. Bulletin de liaison des laboratoires des ponts et chaussées 127, 25–35 (1983) 10. Sayegh, G.: Contribution aé l’étude des propriétés viscoélastiques des bitumes purs et des beétons bitumineux. Ph.D., University Paris, vol. 6 (1965) 11. Such, C., Ballié, M., Lombardi, B., Migliori, F., Ramond, G., Samanos, J., Simoncelli, J.P.: Susceptibilité au vieillissement des bitumes - Expérimentation A08 - Groupe national Bitume. Études et recherches des Laboratoires des ponts et chaussées, CR19, p. 163 (1997) 12. Duriez, M.: Traité de matériaux de construction. Dunod, Tome 2 (1950) 13. Di Benedetto, H., Olard, F.: Linear viscoelastic behaviour of bituminous materials: from binders to mixes. Road Mater. Pavement Des. 5 Sup 1, 163–202 (2004)
Effect of Fiberglass Geogrid Reinforcement to Fatigue Resistance of Bituminous Mixtures Reuber Freire, Hervé Di Benedetto, Cédric Sauzéat, Simon Pouget, and Didier Lesueur
Abstract This work aims at verifying the fiberglass geogrid contribution to the fatigue resistance of bituminous mixtures. To conduct the research, two different configurations of slabs were fabricated. The first one was composed of two mixtures layers glued by emulsion. The second one was identical to the first one but reinforced by fiberglass geogrid in the interface. This geogrid was fabricated at Afitexinov and has 100 kN/m of maximum resistance. Tests were performed on cylindrical samples horizontally cored into slabs. Interfaces were thus vertical. Cyclic tension/compression tests in controlled strain mode were carried out. Seven samples from each type of specimen were tested at frequency of 10 Hz and temperature of 10 °C. Four strain amplitudes were targeted: 80, 90, 100 and 110 µm/m. An increase in fatigue resistance was observed for high strain levels, for sample with the geogrid reinforcement. However, obtained ε6 , parameter from Wohler curves, which is used in the French design method, were similar for both specimens configuration. Keywords Fiberglass geogrid · Reinforcement · Bituminous mixtures · Fatigue
1 Introduction The use of fiberglass geogrids placed in bituminous mixtures layers has increased as a technical solution to rehabilitate pavements, extend its service life and reduce maintenance costs [1]. They could be used for both rehabilitation and construction of new bituminous pavements [2, 3]. The reinforcement by geogrid can be effective to reduce the main distresses in flexible pavements worldwide, rutting and cracking.
R. Freire (B) · H. Di Benedetto · C. Sauzéat University of Lyon/ENTPE, Lyon, France e-mail: [email protected] S. Pouget EIFFAGE Infrastructures, Research & Innovation Department, Velizy-Villacoublay, France D. Lesueur Afitexinov, Champhol, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_74
585
586
R. Freire et al.
According to some authors, fiberglass geogrids are preferable for presenting hightension resistance and flexibility at once [4]. It is also thermally and chemically stable at mixing temperatures for bituminous mixtures [5], and easily removable by milling in the case of further pavement maintenances. Some studies are found in literature investigating the geogrid reinforcement to the fatigue life of bituminous mixture. According to [1, 6, 7] and [8], the geogrid increases the fatigue life of reinforced bituminous mixtures. French design method presents a fatigue resistance verification on its procedure. Thus, an increase in a fatigue resistance due to geogrid reinforcement could lead to a great improvement in the design of new pavement structures. Therefore, this work aims at verifying the fiberglass geogrid contribution to the fatigue resistance of bituminous mixtures. The geogrid effect in the parameter ε6 used on the French pavement design method is analyzed.
2 Specimens Preparation and Experimental Devices One type of bituminous mixture was used to conduct the experimental campaign. This mixture is called Béton Bitumineux Semi-Grenu (BBSG) 0/10 according to the classification in French standards [9]. It is a semi-coarse bituminous mixture designed for wearing course and composed of up to 10 mm size aggregates. The BBSG 0/10 gradation curve is presented in Fig. 1a. It is composed of mineral aggregates with rhyodactic and rhyolitic nature, limestone filler and 20% of recycled asphalt pavement (RAP) containing 4.75% of bituminous binder. These aggregates were mixed with 4.40% of 35/50 penetration grade bituminous binder [10]. The total bituminous binder content in the mixture was 5.53%. The geogrid used to reinforce the bituminous mixture is Notex Glass®, presented in Fig. 1b, provided by the French company Afitexinov. This geogrid is composed by fiberglass yarns and polyester knitted veil, both coated with bituminous emulsion.
Fig. 1 Components of tested specimens: a bituminous mixture gradation curve and b fiberglass geogrid Notex Glass®
Effect of Fiberglass Geogrid Reinforcement …
587
The grid has a square mesh opening of 25 mm and a maximum tensile strength of 100 kN/m in both directions. Slabs of 600 × 400 × 150m m were fabricated by the French company EIFFAGE Infrastructures using a French wheel compactor [11]. Slabs were composed of two layers having approximately the same thickness. In unreinforced slab (so-called configuration B), the interface is only constituted of tack-coat. For reinforced slab (so-called configuration C), interface includes tack-coat and geogrid. Emulsion made with 160/220 pen grade bitumen was used as tack coat for both unreinforced and reinforced slabs. For the unreinforced one, an amount of 290 g/m2 of residual binder was used, while for the reinforced one, an amount of 800 g/m2 was used, divided in two applications, one on each geogrid surface. The reinforced slab fabrication is conducted by first compacting half height slab (75 mm). Then, the first tack coat application is done (400 g/m2 ), followed by the geogrid placement, then the second tack coat application is done (400 g/m2 ) on the geogrid surface and finally the second half height (75 mm) slab is compacted. From each slab fabricated, cylindrical samples with 75 mm in diameter and 140 mm in height were cored (see plan in Fig. 2a). Fatigue tests were performed using a hydraulic press (INSTRON®) having a maximum force capacity of ±25 kN. A thermal chamber was used for temperature controlling during the tests. The tests were carried out by applying axial tensioncompression sinusoidal loading with a controlled strain amplitude until the specimen failure. Four strain amplitudes were targeted: 80, 90, 100 and 110 µm/m. Seven samples from each type of specimen were tested at frequency of 10 Hz and temperature of 10 °C. The axial deformation measurements are done by four extensometers, a couple with 72.5 mm length disposed 180° from one another, and another couple with 75 mm length disposed 180° from one another (c.f. Fig 2b), both fixed in the middle height of specimens. The strain amplitude controlled during the test is calculated by the average of the two extensometers of 72.5 mm length. The temperature is measured by a thermal gauge (PT100 temperature probe) fixed on the specimen surface.
Fig. 2 Specimens preparation: a slabs coring plan with specimen’s dimensions and b instrumental set up scheme: extensometers and temperature probe location
588
R. Freire et al.
3 Results and Analysis The complex modulus (stiffness) variation was monitored during loading cycles for each tested specimen as a representation of the damage level in the material. Three phases can be identify on the complex modulus decrease due to loading in the fatigue characterization [12, 13]. Phase I: characterized by a rapid decrease in complex modulus and increase in phase angle, mainly due to biasing effects, such as self-heating or thixotropy. Phase II: quasi-linear decrease in complex modulus and increase in phase angle due to fatigue damage. Phase III: is the failure phase. The threshold between phase II and III is considered the failure point. Fig 3a presents the complex modulus (E*) divided by the initial modulus (E 0 ) obtained from fatigue tests, only at variation between phase II and III for the unreinforced specimen (B) and Fig. 3b for the reinforced specimen (C). From test results, it is noticeable that higher strain amplitudes supported less number of cycles before the deterioration. Five different failure criteria were used, three global (50% modulus decrease, change in E* versusN concavity, and change in phase angle’s slope) and two local (25% amplitude variation, and 5° phase angle variation, both on individual extensometer measurement). The full description of those failure criteria are found in [12]. Wohler curves for the reinforced and unreinforced configurations were, then, created and are presented in Fig. 4. The different configurations presented distinct susceptibility to strain amplitude variation. At high strain amplitudes, more number of cycles are necessary to lead the reinforced specimen to failure. However, the strain amplitude from which the material supports 1 million loading cycles (ε6 ) obtained for both configurations were very similar, 88 µm/m for unreinforced specimen and 89 µm/m for the reinforced one. Therefore, the geogrid contribution to the fatigue resistance considered in the French design method used nowadays is negligible, despite the fact that the specimens presented different fatigue behavior.
Fig. 3 Complex modulus (stiffness) variation during fatigue tests: a unreinforced specimens, and b reinforced specimens
Effect of Fiberglass Geogrid Reinforcement …
ε6
589
Reinforced
Unreinforced
Fig. 4 Fatigue test results in Wohler curves for reinforced and unreinforced specimens
4 Conclusions This work presented an experimental fatigue test campaign aiming at verifying the fiberglass geogrid contribution to the fatigue resistance of bituminous mixtures. The reinforced specimens presented higher fatigue damage resistance at high strain amplitudes comparing to the unreinforced ones. However, according to fatigue resistance considered in the French design method used nowadays, translated as ε6 value verification, the geogrid reinforcement effect was negligible.
References 1. De Bondt, A.H.: 20 years of research on asphalt reinforcement—achievements and future needs. In: 7th international RILEM conference on cracking in pavements, pp. 327–335. Delft (2012) 2. Geosynthetic Material Association: Handbook of Geosynthetics, USA (2002) 3. Cost Action 348: Reinforcement of pavements with steel meshes and geosynthetics (REIPAS). In: Final report of a European Concerted Research Action Designated as COST Action 348 (2006) 4. Nguyen, M.L., Blanca, J., Kerzréhoa, J.P., Hornych, P.: Review of glass fibre grid use for pavement reinforcement and APT experiments at IFSTTAR. Road Mater. Pavement Design 14(1), 287–308 (2013). https://doi.org/10.1080/14680629.2013.774763 5. Darling, J.R., Woolstencroft, J.H.: A study of fiberglass pavement reinforcement used in different climatic zones and their effectiveness in retarding reflective cracking in asphalt overlays. In: Proceedings of the 5th International RILEM Conference on Cracking in Pavements. Limoges (2004) 6. Arsenie, I.M., Chazallon, C., Themeli, A., Duchez, J.L., Doligez, D.: Measurement and prediction model of the fatigue behavior of glass fiber reinforced bituminous mixture. In: 7th International RILEM Conference on Cracking in Pavements, pp. 653–664. Delft (2012) 7. Millien, A., Dragomir, M.L., Wendling, L., Petit, C., Iliescu, M.: Geogrid interlayer performance in pavements: tensile-bending test for crack propagation. In: Scarpas, T., Kringos, N., Al-Qadi, I.L., Loizos, A. (eds.) 7th RILEM International Conference on Cracking in Pavements, vol. 2, pp. 1209–1218. Springer, Dordrecht (2012)
590
R. Freire et al.
8. Godard, E., Van Rompu, J., Brissaud, L., Gileni, F.: Renforcement des chaussées bitumineuses par grille de verre : bilan de 25 ans d’expérience, derniers développements reinforcement of bituminous pavements by glass grids: 25 years of experience, latest developments. In 12èmes Rencontres Géosynthétiques, pp. 85–94. Nancy (2019) (In French) 9. Association Française de Normalisation (AFNOR): “Spécifications des matériaux Partie 1: Enrobés bitumineux” (in French), NF EN 13108-1, Paris (2007) 10. Association Française de Normalisation (AFNOR): Bitumes et liants bitumineux—Détermination de la pénétrabilité à l’aiguille” (in French), NF EN 1426, Paris (2018) 11. Association Française de Normalisation (AFNOR): Mélange bitumineux—Méthodes d’essai pour mélange hydrocarboné à chaud—Partie 33: confection d’éprouvettes au compacteur de plaque” (in French), NF EN 12697-33+A1, Paris (2007) 12. Baaj, H.: Comportement des matériaux granulaires traits aux liants hydrocarbonés. (Doctoral dissertation). ENTPE-INSA Lyon, Lyon (2002) (In French) 13. Di Benedetto, H., de la Roche, C., Baaj, H., Pronk, A., Lundström, R.: Fatigue of bituminous mixtures. Mater. Struct. 37, 202–216 (2004). https://doi.org/10.1007/BF02481620
Effect of Filler Type and Content on the Rheological Properties of Asphalt Mastics Jayvant Choudhary, Brind Kumar, and Ankit Gupta
Abstract This study investigated the influence of various fillers on the rheology and performance of asphalt mastics. Mastics were prepared by mixing a PG70-XX binder with two waste fillers (limestone sludge waste, glass powder) and conventional stone dust filler in three different proportions. Firstly, physical and chemical properties of fillers were determined and then asphalt concrete mixes were designed. Based on their optimum bitumen contents, the filler bitumen ratios were determined and mastics were prepared. The multiple stress creep and recovery (MSCR) test and linear amplitude sweep (LAS) analyses were performed to analyze the behavior of asphalt mastics against rutting and fatigue. In general, the use of waste fillers improved the rheological behavior of mastics in comparison to conventional stone dust mastic. The stiffness of all mastics increased with the filler bitumen ratio. The glass powder mastic displayed superior stiffness at all temperatures and exhibited higher elastic recovery and lower creep compliance. Keywords Asphalt mastic · Filler · Rheology · MSCR · LAS · Sustainability
1 Introduction Asphalt mixes are conventionally made up of aggregates, bitumen and fillers. Fillers are the finest part of aggregates which pass through 75 μm sieve [1]. The homogenous mix of filler and bitumen is usually defined as the mastic. Several studies have observed that the physical and chemical properties of the filler, their physicalchemical interaction with bitumen, and their volumetric concentration in the mix significantly influence the performance of mastic and asphalt mixes [2, 3]. Stone dust, cement, and hydrated lime are being conventionally utilized in asphalt mix composition as fillers since they deliver satisfactory performances in the mix. However, several studies have observed that the solid wastes like bauxite residue [2]; carbide lime [2]; rice straw ash [2] etc. could be beneficially utilized as fillers. This study J. Choudhary (B) · B. Kumar · A. Gupta Indian Institute of Technology (Banaras Hindu University), Varanasi, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_75
591
592
J. Choudhary et al.
Table 1 Chosen gradation of asphalt concrete mixes Sieve sizes (mm)
13.2
9.5
4.75
2.36
1.18
0.6
0.3
0.15
0.075
Lower-upper 90–100 70–88 53–71 42–58 34–48 26–38 19–28 12–20 4–10 limits (%) Adopted gradation (%)
91
74
62
50
43
35
25
14
4.0, 5.5, 7.0
aimed to investigate the influence of two industrial waste fillers, glass powder (GP) and dried limestone sludge (LS) on the performance of asphalt mastics. These wastes are generated during the cutting and polishing operations of glass and limestone slabs in their respective industries. The results were compared with conventional mastics made with stone dust (SD) as filler.
2 Materials and Experimental Investigation 2.1 Material Coarse and fine dolomite aggregates and PG 70-XX bitumen were utilized in the study. The gradation used to prepare the asphalt concrete mix was chosen as per Indian guidelines [1] (Table 1). SD was utilized as the conventional filler and was collected from a stone quarry in Sonbhadra District (24.46° N, 82.99° E). LS was collected from the dump yard of dimension stone industry located in Kota City (26.91° N, 75.78°). Whereas, GP was collected from the dump yard of a glass factory located in Bhopal City (23.26° N, 77.41° E). Oven dried filler which passes through 0.075 mm Sieve Was utilized in this study.
2.2 Characterization of Fillers Specific gravities of all fillers were determined as per ASTM D854. Fineness of the fillers was determined by calculating their fineness modulus (FM) values. Particle shape and surface texture was analyzed using scanning electron microscopy (SEM) analysis. Porosity of fillers was determined as per the German filler test [4]. Prevalent minerals in the filler composition were evaluated using X-Ray Diffraction (XRD) investigation. The analysis of Methylene blue values (MBV) of fillers was done as per the EN 933–9 specification to enumerate the harmful clay content and organic matters in fillers. Various results are stated in Table 2.
Effect of Filler Type and Content on the Rheological Properties …
593
Table 2 Properties of various fillers Property
GP
LS
SD
Inferences
Specific gravity
2.370
2.650
2.698
GP and LS has lower specific gravity and thus occupies larger volume in asphalt mix
Methylene blue value 1.25 (g/kg)
3.75
3.25
All fillers have low MBV (less than 10) which indicated the presence of lower harmful clay content per unit weight of material
German filer value (g)
75
97
85
GP and LS had lowest and highest fractional voids per unit weight respectively
Fineness modulus
4.66
3.03
5.38
SD and LS are coarsest and finest filler respectively
Primary Mineralogical composition (XRD)
Quartz (SiO2 )
Calcite (CaCO3 ), Quartz (SiO2 ), Enstatite (Mg2 Si2 O6 )
Dolomite (CaMg(CO3 )2 ), Quartz (SiO2 ), Ertixite (Na2 Si4 O9 )
SD and LS has dolomite and calcite respectively which are water insoluble mineral having good asphalt adhesion. GP had quartz which is associated with poor moisture sensitivity
2.3 Design of Asphalt Concrete Mix The Marshall mix design procedure was followed as per MS-2 [5] specification to determine optimum asphalt content (OAC) of asphalt mixes. The OAC is considered as asphalt content corresponds to 4% air voids of the mixes. The effective filler bitumen ratio was then calculated and asphalt mastics were designed according to it.
594
J. Choudhary et al.
2.4 Preparation and Testing of Asphalt Mastics A total of nine types mastics were designed by mixing GP, LS, and SD fillers with bitumen in three different filler bitumen ratio, which was decided based on effective OAC of the mixes and their filler contents (4, 5.5, and 7%). The filler content was calculated as the percentage of the total weight of aggregates in mix as per the Indian pavement design guideline [1]. The OAC of the mixes are stated in Table 3 along with the corresponding filler bitumen ratio of their mastics. The mixing was done at 163 °C QUOTE using mechanical mixer operating at 2000 rpm for 30 min. The prepared mastics were short term aged using Thin Film Oven as per ASTM D1754 and were used for testing in Multiple Stress Creep and Recovery test. While, for the testing of fatigue life, short term aged mastics were further subjected to long term ageing as per the protocol suggested by [6]. The rheological properties (complex modulus and phase angles) of short term aged mastics were determined using frequency sweep test performed at intermediate temperature (25 °C) and high temperatures (46–70 °C). The spindle of 8 mm gap (with 2 mm gap) and 25 mm diameter (with 1 mm gap) was used for testing at intermediate and higher temperatures respectively. The testing is conducted in strain controlled mode with applied strain is kept equal low enough so that results falls within the linear viscoelastic region. The testing is done at the frequency range of 0.1–100 rad/s to analyze the influence of testing frequency on rheological parameters. The MSCR test was conducted on short-term aged samples at 64 °C, using a dynamic shear rheometer (DSR) with 25 mm parallel plate in diameter (1 mm gap), in accordance with the AASHTO T 350. During the tests, each sample was subjected to ten consecutive cycles at two stress levels (0.1 and 3.2 kPa) and every cycle underwent one second creep loading followed by a nine-second recovery without loading. The non-recoverable compliance (Jnr ) and the percent recovery (%R) after 10 cycles at 0.1 and 3.2 kPa were studied. The Jnr value was calculated as the ratio between the average non-recoverable strain for 10 creep and recovery cycles, and the applied stress for those cycles. The fatigue failure analysis was carried with Linear Amplitude Sweep (LAS) test as per AASHTO TP 101. This test was done with DSR with a standard geometry of 8 mm parallel plates and a 2 mm thickness gap. It measures accelerated damage of mastic using cyclic loading by linearly increasing load amplitudes (1–30%). The relationship between the number of loading cycles to failure (Nf ) and the applied initial strain amplitude (γ) can be expressed by the following equation. N f = A(γ )−B Where A and B are the fitting coefficients.
(1)
Effect of Filler Type and Content on the Rheological Properties …
595
3 Results and Discussions 3.1 Performance of Asphalt Mastics The results of frequency sweep analysis are presented in the form of black diagrams between the complex modulus and phase angle as stated in Fig. 1. Black diagrams of all mastics were found to shift upwards and towards left as the filler content in them were increased. This clearly indicated that the inclusion of filler imparts stiffness and elasticity to the mastics. Mastics prepared with GP displayed higher complex modulus and lower phase angles followed by LS and SD. This implied the higher stiffening action of GP mastics due to their volume in the mixes. It is also observed that SD and LS mastics followed a consistent trend throughout while GP mastic starting deviate from its trend at higher filler contents. This suggested that at higher filler concentration GP mastics deviate from linear viscoelastic region, even when testing is performed at the strain within the linear viscoelastic limits. The mineralogical composition of GP is also different from LS and SD fillers. LS and SD consist of dolomite and calcite in their compositions which enhance bitumen filler adhesion, while GP consisted of quartz whose adhesion behavior with bitumen is debatable. This aspect may also affect the viscoelastic behavior of mastics and needed a detailed further study. The rutting resistance of mastics was found to increase with the filler bitumen ratio as determined from their decrease in Jnr values and the increase in percentage recovery values at both stress levels (Table 3). The GP mastics displayed highest rutting resistance followed by LS and SD mixes. The higher stiffening of GP mastic might be attributed to the higher volume of glass powder at each filler level. Higher rutting resistance of GP mastics might also be due to the angular nature of particles as well as due to their higher porosity. For all the mastics the fatigue lives were Table 3 Table displaying various parameters obtained from MSCR test of mastics Filler type
Filler content (%)
Name of mastic
Effective filler Bitumen ratio
Percentage volume of filler in the mastic (%)
Jnr at 0.1 kPa (kPa−1 )
Jnr at 3.2 kPa (kPa−1 )
% R at 0.1 kPa (%)
SD
4
SD 4
0.66
19.64
1.0839
1.2343
5.15
0.78
5.5
SD 5.5
0.98
26.23
0.7132
0.8042
5.62
0.94
GP
LS
% R at 3.2 kPa (%)
7
SD 7
1.38
33.98
0.5013
0.5733
6.55
1.70
4
GP 4
0.71
23.04
0.4819
0.5491
7.86
1.00
5.5
GP 5.5
1.01
30.48
0.2056
0.2142
13.42
1.50
7
GP 7
1.46
38.26
0.0753
0.0597
28.66
6.10
4
LS 4
0.70
20.88
0.8668
1.0184
7.09
1.10
5.5
LS 5.5
1.06
28.55
0.5557
0.6414
5.96
1.07
7
LS 7
1.45
35.36
0.2976
0.3282
10.93
2.99
596
J. Choudhary et al. 1.00E+08
Complex Modulus (Pa)
Complex Modulus (Pa)
1.00E+08 1.00E+07 1.00E+06 1.00E+05 1.00E+04 1.00E+03 1.00E+02
1.00E+07 1.00E+06 1.00E+05 1.00E+04 1.00E+03 1.00E+02 1.00E+01
1.00E+01 10
20
30
40
50
60
70
80
10
90
20
30
Bitumen
SD4
SD5.5
40
50
60
70
80
90
Phase Angle (º)
Phase Angle (º) Bitumen
SD7
(a)
LS4
LS5.5
LS7
(b)
Complex Modulus (Pa)
1.00E+08 1.00E+07 1.00E+06 1.00E+05 1.00E+04 1.00E+03 1.00E+02 1.00E+01
10
20
30
40
50
60
70
80
90
Phase Angle (º) Bitumen
GP4
GP5.5
(c)
GP7
Fig. 1 a Black diagram of SD mastics. b Black diagram of LS mastics. c Black diagram of GP mastics
found to reduce with the increase in filler bitumen ratio and the increase in strain magnitude (Table 4). The decrease in fatigue lives might be due to the increase in the stiffness of the mastics. In general, the SD mastics have the highest fatigue lives followed by LS and GP mastics. However, LS mastic corresponding to 4% filler exhibited higher fatigue life than SD and GP mastics. It seemed like at lower filler concentration; LS distributed more uniformly than other fillers due to their finest particle size which improved the fatigue resistance of mastics. This effect diminished at higher concentration due to increase in stiffness. The GP mastics were found to be most strain susceptible.
Effect of Filler Type and Content on the Rheological Properties …
597
Table 4 Table displaying various parameters obtained from LAS analysis of mastics Type of mastic
A
B
SD 4
23903
2.50
SD 5.5
20671
2.45
SD 7
7719
2.73
GP 4
13959
2.59
GP 5.5
7012
2.70
GP 7
2408
2.73
LS 4
27839
2.52
LS 5.5
16440
2.51
5758
2.64
LS 7
Nf and γ relationship Nf =23903
(γ)−2.50
−2.45 Nf =20671 (γ) −2.73 Nf =7719 (γ) −2.59 Nf =13959 (γ) −2.70 Nf =7012 (γ) −2.73 Nf =2408 (γ) −2.52 Nf =27839 (γ) −2.51 Nf =16440 (γ) −2.64 Nf =5758 (γ)
Nf at different strains 0.1%
1%
2.5%
5%
7733532
23697
2368
415
5960708
20578
2156
391
4298452
7374
585
86
5554930
13904
1282
211
3589537
6754
556
84
1268580
2560
217
33
9398720
27993
2766
480
5371132
16499
1650
289
2585007
5511
477
75
4 Conclusions This study analyzed the performance of asphalt mastics made with LS and GP as fillers at three filler contents. Both wastes displayed physical and chemical properties synonymous of a good filler. Both GP and LS mastics also displayed higher stiffening and rutting resistance than conventional SD mastics at each filler contents. However, this higher stiffening of these fillers also resulted in relatively lower fatigue lives of their mastics at each filler contents.
References 1. MORTH (Ministry of Road Transport and Highways): Specifications for Road and Bridge Works (Fifth Revision). Indian Road Congress, New Delhi (2013) 2. Choudhary, J., Kumar, B., Gupta, A.: Application of waste materials as fillers in bitu-minous mixes. Waste Manage. 78, 417–425 (2018) 3. Das, A.K., Singh, D.: Investigation of rutting, fracture and thermal cracking behavior of asphalt mastic containing basalt and hydrated lime fillers. Constr. Build. Mater. 141, 442–452 (2017) 4. National Asphalt Pavement Association (NAPA): Evaluation of baghouse fines for hot mix asphalt. Information series, vol. 127, National Asphalt Pavement Association (1999) 5. Asphalt Institute: Mix Design Methods for Asphalt Concrete and Other Hot-Mix Types: Manual Series No. 2 (MS-2) 6th ed. Asphalt Institute (1997) 6. Behera, P., Singh, A., Reddy, M.: An alternative method for short- and long-term ageing for bitumen binders. Road Mater. Pavement Des. 14(2), 445–457 (2013)
Effect of RAP and Binder Properties on Indirect Tensile Strength and Dynamic Modulus of Cold Recycled Foamed Asphalt Mixtures with High RAP Content Wangyu Ma , Di Wang , Fan Gu , Adam Taylor, and Randy West Abstract Cold recycling with foamed asphalt is a sustainable pavement rehabilitation technology and becomes more widely applied in many countries. However, there are still some ongoing scientific debates on the issues such as cold recycling mix design, pavement structural design and analysis. Many popular mix design methods adopted indirect tensile strength (ITS) as the most important design criterion. Dynamic modulus (|E*|) representing the viscoelastic behavior which has been found in the cold recycled foamed asphalt mixtures also become a helpful parameter for accurate pavement structural analysis. For cold recycled foamed asphalt mixtures with only reclaimed asphalt pavement (RAP), RAP and virgin binder used for producing foamed asphalt may govern the performance test results. The objective of this study was to determine the effects of RAP and virgin binder properties on the ITS and |E*| of the recycled mixtures. Three RAP materials (i.e., fine, medium, coarse) and two virgin binders (i.e., PG 58-34 and PG 67-22) were selected to produce six recycled mixtures. The effects of RAP properties (i.e., gradation, fineness modulus, F-E ratio) and virgin binder performance grade (PG) on the two performance test results were evaluated. The ITS results of the mixtures with finer RAP or higher PG virgin binder tended to be higher. The |E*| result was remarkably affected by RAP gradation at most of the testing frequencies and temperatures. Mixtures with the same RAP material but different virgin binder can achieve similar |E*| results. Keywords Cold recycling · Foamed asphalt · Reclaimed asphalt pavement · Indirect tensile strength · Dynamic modulus
W. Ma (B) The Transtec Group, Inc, 1828 Good Hope Road, Enola, PA 17025, USA e-mail: [email protected] D. Wang Technische Universität Braunschweig, Braunschweig, Germany F. Gu · A. Taylor · R. West National Center for Asphalt Technology At Auburn University, 277 Technology Parkway, Auburn, AL 36830, USA © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_76
599
600
W. Ma et al.
1 Motivation and Objective Cold recycling with foamed asphalt is a sustainable pavement rehabilitation technology for it can save tons of virgin aggregate and reduces fuel consumption. Several laboratory testing methods have been used to investigate the effect of material properties. Kim and Lee [1] found cold recycled foamed asphalt mixtures with finer RAP had higher indirect tensile strength (ITS) but the study was only based on the same asphalt binder. Fu et al. [2] found the cold recycled mixes with higher foamed asphalt content had higher ITS results. Diefenderfer et al. [3] studied the effect of recycling agent and mineral additives on the dynamic modulus of cold recycled mixtures. However, the effects of RAP and foamed asphalt properties on the dynamic modulus were still not fully understood. This study was to investigate these effects in order to provide more guidance on the material selection in mix design and pavement analysis. The ITS was selected as one method because it is efficient and widely used in the cold recycling mix design procedure. Dynamic modulus was also selected since this viscoelastic characteristic may be implemented in the future mechanics-empirical pavement design method. The objective of this study was to investigate the effects of RAP and asphalt properties on the indirect tensile strength and dynamic modulus of cold recycled foamed asphalt mixtures with high RAP content.
2 Materials 2.1 RAP Three RAP materials, fine, medium, and coarse, from the construction site, were used in this study. The gradations of these RAP materials were tested following ASTM C136 and plotted in Fig. 1. Table 1 summarizes several basic properties of the recycled materials, including the RAP binder content, recovered binder PG, fineness modulus (FM) and F-E particles’ percentage. The recovered binder was tested by using dynamic shear rheometer (DSR) and bending beam rheometer (BBR) to determine the true performance grade (AASHTO T315, AASHTO T313, and AASHTO MP1). The fineness modulus, indicating an overall size distribution of the RAP particles, was determined based on the gradation of the RAP materials (ASTM C125). The F-E particles was measured following ASTM 4791.
Effect of RAP and Binder Properties on Indirect Tensile Strength …
601
Fig. 1 Gradations of RAP materials
Table 1 Summary of the RAP materials RAP
RAP binder content (%)
Recovered binder true PG
Fineness modulus
3:1 F-E particles (%)
Fine
4.92
103.6–12.7
4.5
1
Medium
4.85
103.5–9.3
5.4
5
Coarse
4.10
103.6–12.7
5.5
1
2.2 Foamed Asphalt Two virgin asphalt binders were used to produce foamed asphalt in laboratory. One is PG 67-22 binder from Lake City, Florida. The other is PG 58-34 binder from Alberta in Canada. These two asphalt binders were used to produce two foamed asphalt at 160 °C. The optimum foaming water content for PG 67-22 binder was 1.8%. The expansion ratio was 11 times. The half-life was 9 seconds. The optimum foaming water content for PG 58-34 binder was 2.0%. The expansion ratio was 19 times. The half-life was 10 seconds.
2.3 Cold Recycled Foamed Asphalt Mixtures Six types of cold recycled foamed asphalt mixtures were tested (i.e., each type of foamed asphalt was blended with three different RAP materials). One percent of hydraulic cement by weight of the dry RAP was added in each mixture. The composition and properties of cement met ASTM C150 standard. For ITS test, both 2
602
W. Ma et al.
and 3% foamed asphalt content were used in each of these mixtures. For dynamic modulus test, only 3% foamed asphalt content was used in these mixtures. During mixing, RAP was mixed with cement, water, and foamed asphalt in a laboratory twin-shaft pug mill [4]. All the specimens were compacted by the Superpave Gyratory Compactor (SGC) for 30 gyrations because the compacted specimens had similar density and air voids content to the cold in-place recycled pavement sections [5]. The compacted specimens were cured in oven at 40 °C for 3 days [4]. The bulk specific gravity and theoretical maximum specific gravity was measured using an automatic vacuum sealing system following AASHTO T331 and ASTM D6857, respectively.
3 Methodology 3.1 Indirect Tensile Strength The ITS specimens were molded into 100-mm diameter and 63.5-mm height. Four replicates were used for each mixture. Then dry ITS specimens were cured at 25 °C for 2 h while wet ITS specimens were immersed in water tank at 25 °C for 24 h before testing. Specimens after conditioning were tested to failure on a load frame where a diametrical load applied at a rate of 50.8 mm/min. The peak load was recorded to calculate tensile strength based on ASTM D6931.
3.2 Dynamic Modulus The current procedure for dynamic modulus in an asphalt mixture performance tester (AMPT) requires fabrication of a big sample (minimum height of 160 mm) and coring and cutting process to obtain a specimen with 100-mm diameter and 150-mm height (AASHTO TP 79-15). However, in cold recycled pavement layer, non-uniformity of density/air void distribution was found between the upper and bottom parts [6]. After coring and cutting from a big sample, density differences between the upper and lower parts of test specimen may still exist. To address this issue, small specimens with 50-mm diameter and 110-mm height were cored horizontally from the big samples. These dimensions have been used to evaluate several cold recycled pavement projects across the U.S. [3]. Tests were conducted in AMPT at three temperatures (4, 20, and 40 °C) and four frequencies (0.01, 0.1, 1, and 10 Hz).
Effect of RAP and Binder Properties on Indirect Tensile Strength …
603
4 Results and Discussions 4.1 Indirect Tensile Strength The ITS test results of the cold recycled mixes were shown in Fig. 2. The ITS results were found not significantly affected by the densities because of the weak correlation (R-squared = 0.068). Correlations between ITS results and the RAP properties were also investigated. Only the gradation-related property (i.e., fineness modulus) rather than RAP binder content or F-E particles’ percentage was found to strongly correlate with the indirect tensile strength of cold recycled mixes. Therefore, RAP gradation was considered as the dominant property to characterize the three RAP materials. Figure 2 shows the mixtures with fine RAP had highest dry ITS and the mix with coarse RAP had the lowest dry ITS, regardless of asphalt PG and foamed asphalt content. The same trend shows in the wet ITS results even though the mixes with medium and coarse RAP are similar at 2% foamed asphalt content. This may be because the foamed asphalt was distributed more uniformly among finer RAP particles, and therefore, the internal bonding provided by the asphalt increased. Also, mixtures with finer gradations had better compatibility and the asphalt mastic bonding between RAP particles may be stronger because of better compaction. This finding matched the results provided by Kim and Lee [1]. The mixtures with PG 58-34 binder tended to have lower ITS than those with PG 67-22 binder. It may be because the foamed asphalt using higher PG binder was more viscous and provided more stiffness in the asphalt-RAP bonding. Tensile strength ratio was relatively consistent above 0.8 except for two mixes.
Fig. 2 Indirect tensile strength test results of the cold recycled mixtures
604
W. Ma et al.
Figure 2 shows mixtures with 3% foamed asphalt content (FAC) tended to have higher dry ITS results than those with 2% FAC, which may be because more foamed asphalt provided more internal bonding between RAP particles.
(a) Mixtures with PG 58-34 Binder
(b) Mixtures with PG 67-22 Binder Fig. 3 Dynamic modulus master curve for the cold recycled mixtures
Effect of RAP and Binder Properties on Indirect Tensile Strength …
605
4.2 Dynamic Modulus Figure 3 shows the dynamic modulus master curves of the cold recycled foamed asphalt mixtures. Figure 3a shows the master curves of three mixtures with PG 58-34 binder had similar trends while their differences were more evident at lower frequencies. The mixture with fine RAP had the highest modulus, followed by the mixtures with medium and coarse RAP. This may be because the mixtures with finer gradations had better compactibility and lower air voids content. Figure 3b shows the master curves for the mixtures with PG 67-22 binder were in the same range as the curves in Fig. 3a except for the mixture with medium RAP. This mixture had a lower modulus than other mixtures from intermediate to high frequencies.
5 Conclusions • With the same foamed asphalt, the cold recycled foamed asphalt mixture with finer RAP gradation tended to have higher indirect tensile strength; With the same RAP, the mixtures with higher PG virgin binder (with adequate foaming properties) tended to have higher indirect tensile strength. • Mixtures with the same RAP material but different virgin binder had similar dynamic modulus. • When using PG 58-34 virgin binder, mixture with fine RAP tended to have high modulus than other two mixtures at low and intermediate temperatures (4 and 20 °C), although the difference was not statistically significant. When using PG 67-22 virgin binder, the mixtures with fine and coarse RAP had similar moduli and both were significantly higher than moduli of the mixture with medium RAP.
References 1. Kim, Y., Lee, H.D.: Development of mix design procedure for cold in-place recycling with foamed asphalt. J. Mater. Civ. Eng. 18(1), 116–124 (2006) 2. Fu, P., Jones, D., Harvey, J.T.: The effects of asphalt binder and granular material characteristics on foamed asphalt mix strength. Constr. Build. Mater. 25(2), 1093–1101 (2011) 3. Diefenderfer, B.K., Bowers, B.F., Schwartz, C.W., Farzaneh, A., Zhang, Z.: Dynamic modulus of recycled pavement mixtures. Transp. Res. Record No. 2575, 19–26 (2016) 4. GmbH, W.: Wirtgen Cold Recycling Technology. Windhagen, Germany (2012) 5. Cross, S.A.: Determination of superpave gyratory compactor design compactive effort for cold in-place recycled mixtures. Transp. Res. Record No. 1819, 152–160 (2003) 6. Apeagyei, A.K., Diefenderfer, B.K.: Evaluation of cold in-place and cold central-plant recycling methods using laboratory testing of field-cored specimens. J. Mater. Civ. Eng. 25(11), 1712–1720 (2012)
Effect of RAP Gradation on Fracture Tolerance of Asphalt Mixtures Yu Yan, David Hernando, Bongsuk Park, Gabriele Tebaldi, and Reynaldo Roque
Abstract This study investigated the effect of RAP gradation on the fracture tolerance of asphalt mixtures in terms of fracture energy density (FED). Experimental factors included two RAP sources, one virgin binder and four RAP contents of 0, 20, 30 and 40%, resulting in a total of seven mixtures. Binder fracture energy tests were conducted to determine the FED of blends of virgin and extracted RAP binders. Moreover, interstitial component direct tension (ICDT) tests were employed to measure the FED of the interstitial component, which is the fine portion of a mixture that governs cracking performance. Finally, Superpave IDT tests were performed to obtain the FED of RAP mixtures. Increasing RAP content led to reductions in FED at binder, IC and mixture levels. More importantly, the reduction in FED was almost identical for mixtures and IC, whereas a less pronounced drop was observed for binders. This finding substantiates that full blending between RAP and virgin asphalt as artificially created in binder specimens did not occur in the actual RAP mixtures. Coarse RAP yielded greater IC and mixture FED values than fine RAP despite the lower RAP binder stiffness of the latter, substantiating that RAP gradation controls the distribution of RAP binder in mixtures and consequently, high RAP binder content in IC portion reduces the fracture tolerance of a mixture. A new design guideline that considers RAP binder content and properties along with RAP gradation is proposed to determine the maximum allowable RAP content, so that cost savings and environmental benefits can be maximized by reducing the need for virgin pavement materials and landfill space.
Y. Yan (B) Tongji University, Shanghai, China e-mail: [email protected] D. Hernando University of Antwerp, Antwerp, Belgium B. Park · G. Tebaldi · R. Roque University of Florida, Gainesville, USA G. Tebaldi University of Parma, Parma, Italy © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_77
607
608
Y. Yan et al.
Keywords Reclaimed asphalt pavement · Gradation · Recycling · Fracture energy density · Interstitial component
1 Introduction The use of reclaimed asphalt pavement (RAP) is a sustainable approach that brings in significant economic and environmental benefits. AASHTO M323 Standard Specification for Superpave Volumetric Mix Design assumes that RAP binder fully blends with virgin binder in the design of new mixtures, which maximizes the contribution of RAP binder. However, conflicting conclusions have been reported with respect to the contribution of RAP binder in recycled mixes [1]. Current consensus in the asphalt community is that virgin binder selection should be based on RAP binder properties, e.g., binder performance grade (PG), and the effective RAP binder content that contributes to the total binder content in the mix. However, there is not yet a well-accepted methodology to accurately predict the effective contribution of RAP binder, which affects both mixture performance and the economic competitiveness of the recycling process. During mix design, RAP is typically treated like another stockpile that is combined with the virgin aggregate to achieve the desired mix gradation. Recognizing that full blending does not occur in real asphalt mixes, Nguyen [2] reported that RAP size tremendously affected the level of blending (the smaller the size of RAP particles, the better the distribution of RAP in the mix and, consequently, the better the interaction between RAP and virgin binder), and the stiffness of recycled mixtures. Coarse RAP particles tend to have higher binder content than fine RAP particles so when mixed with virgin aggregate, more RAP binder is transferred from the coarse RAP particles to virgin aggregates than from the fine RAP particles [3]. In a separate study, mixes with coarser RAP outperformed mixes with finer RAP, and RAP particle size was concluded to have a considerable effect on the contribution of RAP binder to total binder content, the aggregate skeleton of the mix, and ultimately the performance of the mix (rutting and cracking) [4]. Therefore, further understanding on the effect of RAP gradation on mixture performance is warranted. This study aimed to investigate the effect of RAP aggregate gradation (size distribution) on mixture fracture tolerance and to propose a design guideline that considers not only RAP binder properties and content, but also RAP aggregate gradation.
2 Materials This study employed a PG 76-22 PMA, Georgia granite and local sand, and two RAP sources. The binder content of the coarser RAP (RAP-C, 4.6%) was slightly lower than the finer RAP (RAP-F, 4.8%); however, the former had notably greater stiffness
Effect of RAP Gradation on Fracture Tolerance of Asphalt Mixtures
609
100
Percent Passing (%)
Fig. 1 Aggregate gradation of two RAP sources and prepared RAP mixtures
80 60 40
RAP-C RAP-F Mixtures
20
19
12.5
9.5
4.75
2.36
1.18
0.6 0.3 0.15 0.075
0
Table 1 Details of RAP content in mixtures RAP content (%, by weight of aggregate)
RAP binder content (%)
Virgin binder content (%)
Total binder content (%)
RAP binder replacement rate (%)
Reference
0
0.0
5.2
5.2
0
RAP-C
20
0.9
5.0
5.9
15
RAP-F
30
1.4
4.5
5.9
23
40
1.8
3.8
5.6
33
20
0.9
4.3
5.2
18
30
1.5
3.6
5.0
29
40
1.9
3.0
4.9
39
than the latter, as indicated by their absolute viscosity of 624,400 poises and 98,400 poises. A 12.5-mm nominal maximum aggregate size Superpave mixture was selected as the reference mix. Figure 1 shows the aggregate gradation of both RAP sources and the reference mixture. All RAP mixtures (20, 30 and 40% RAP by weight of aggregate) were designed to have the same aggregate gradation as the reference mixture, which was achieved by adjusting the virgin aggregate components. The Superpave mix design method was followed to determine the design asphalt content resulting in 4% air voids at 75 gyrations. Table 1 summarizes the total binder content and RAP binder replacement rate (percent of RAP binder divided by total binder content) of each mixture.
3 Experimental Design The effect of RAP gradation on fracture properties was assessed at the mixture level and then further investigated at the binder and interstitial component (IC) levels.
610
Y. Yan et al.
According to the dominant aggregate size range and interstitial component (DASRIC) model, IC is the fine portion of a mixture (passing a 1.18-mm sieve) that governs cracking performance [5]. Superpave indirect tensile (IDT) tests were conducted to measure the fracture energy density (FED) of asphalt mixtures [6]. FED describes the damage tolerance before fracture occurs. For each mixture, three replicates were tested at 10 ˚C. All specimens had an air void content within 7 ± 0.5% and were conditioned through short-term oven aging. The binder fracture energy (BFE) test, which allows to accurately determine binder FED at intermediate temperatures, was conducted on blends of virgin and recovered RAP binders. BFE tests were performed at 15˚C on rolling thin film oven plus pressure aging vessel-aged specimens [7]. Note that binder FED is independent of testing temperature in the 0–20 ˚C range, which makes the comparison among binder, IC and mixture level appropriate. IC specimens were prepared with the fine aggregate (virgin and RAP) passing a 1.18-mm sieve and following the relative volume of effective binder and IC aggregates in the corresponding mixtures. FED was determined by performing the interstitial component direct tension (ICDT) test [8]. Three replicate IC specimens were tested at 10 ˚C for each mix and the average IC FED values were reported.
4 Results and Discussion
1.0
1.0
0.8
0.8
Normalized FED
Normalized FED
Figure 2 shows the FED of binder, IC, and mixture specimens normalized with respect to the mix without RAP. Increasing RAP content reduced FED at three levels for both RAP sources. Note that the reduction in FED was almost identical for mixtures and IC, whereas a less pronounced drop was observed for binders. This finding substantiates that full blending between RAP and virgin asphalt as artificially created in binder specimens did not occur in the actual RAP mixtures. Conversely, IC specimens exhibited a level of blending similar to that of the corresponding RAP
0.6 0.4
Mixture IC Binder
0.2 0.0
0.6 0.4
Mixture IC Binder
0.2 0.0
0%
10%
20%
30%
Binder replacement rate
40%
0%
10%
20%
30%
Binder replacement rate
Fig. 2. Normalized FED of binder, IC, and mixture: a RAP-C and b RAP-F
40%
Effect of RAP Gradation on Fracture Tolerance of Asphalt Mixtures 50%
RAP content in IC fraction
6
Mixture FED (kJ/m3)
611
5 4 3 2
RAP-C
1
RAP-F
0 0%
10%
20%
30 %
Binder replacement rate
40%
RAP-C RAP-F
40% 30% 20% 10% 0% 0%
10%
20%
30%
40%
Binder replacement rate
Fig. 3 Mixture results for two RAP sources: a mixture FED and b RAP content in IC fraction
mixes. Of note, the significant reduction in normalized FED of 40% RAP-C mixture was believed to be caused by the large proportion of weak Florida limestone in the coarse fraction of the mix coming from RAP-C. Figure 3 illustrates that RAP gradation had a dominant effect on mixture fracture tolerance. Despite the higher stiffness of RAP-C binder, mixtures with RAP-C exhibited greater FED than those with RAP-F (Fig. 3a). Figure 3b, which plots the RAP content (aggregate plus binder) in the IC portion of each mix, clearly shows that mixtures with the finer RAP contained twice as much RAP in the fine fraction as compared to mixtures with the coarser RAP. In other words, since RAP aggregate gradation controls the distribution of RAP binder in recycled mixtures and full blending does not occur, finer RAP sources tend to increase the proportion of aged and brittle binder in the fine portion of the mix that governs fracture response, which results in lower fracture energy density values. These results show that the existing design guidelines for RAP mixtures based only on RAP binder content and properties are insufficient to capture the impact of RAP gradation on RAP binder distribution and, thus, on cracking performance. Figure 4 illustrates a new design approach where the maximum allowable RAP Fig. 4 New design approach to determine the maximum allowable RAP content in a mixture
612
Y. Yan et al.
content is defined by RAP binder properties and RAP gradation. Conceptually, the combination of a RAP binder of low stiffness with a coarse RAP will allow to incorporate the maximum RAP content. Conversely, the scenario of a very fine RAP with an extremely stiff binder will notably limit the amount of RAP that can be incorporated into a new mixture.
5 Conclusion and Recommendation The experimental investigation showed that, for a similar binder replacement rate, a mixture produced with a coarser RAP exhibited greater fracture energy density (FED) than a mixture incorporating a finer RAP, despite the former having a notably stiffer RAP binder. The mixtures produced with the finer RAP contained twice as much RAP in their interstitial component, the finer fraction of the mixture that is believed to drive fracture response. In other words, as full blending between RAP binder and virgin binder does not occur, RAP gradation controls the distribution of RAP binder within the mixture, which may result in recycled mixtures with high proportion of aged and brittle binder in their most critical fraction from the standpoint of cracking performance. To account for this effect, this study strongly recommends introducing RAP gradation as a key element in the determination of the maximum allowable RAP content in new mixtures. The interstitial component direct tension (ICDT) test, which was proven to represent a blending scenario similar to that in real mixtures while being less labor intensive than mixture testing, appears to be a powerful tool to help in the design process of RAP mixtures. Future research efforts should focus on performing ICDT testing on a variety of RAP sources with a broad range of RAP binder stiffness and RAP gradation to investigate effects on FED and establish recommendations on maximum allowable RAP contents.
References 1. Nguyen, H.V.: Effects of mixing procedures and RAP sizes on stiffness distribution of hot recycled asphalt mixtures. Constr. Build. Mater. 47, 728–742 (2013) 2. Saliani, S., Carter, A., Baaj, H.: Investigation of the impact of RAP gradation on the effective binder content in hot mix asphalt. In: CSCE Resilient Infrastructure Conference, London (2016) 3. Saliani, S.S., Carter, A., Baaj, H., Tavassoti, P.: Characterization of asphalt mixtures produced with coarse and fine recycled asphalt particles. Infrastructures 4(67) (2019) 4. Kim, S., Roque, R., Guarin, A., Birgisson, B.: Identification and assessment of the dominant aggregate size range (DASR) of asphalt mixture. J. Assoc. Asphalt Paving Technol. 75, 789–814 (2006) 5. Roque, R., Buttlar, W.G.: The development of a measurement and analysis system to accurately determine asphalt concrete properties using the indirect tensile mode. J. Assoc. Asphalt Paving Technol. 304–332 (1992) 6. Yan, Y., Hernando, D., Roque, R.: Fracture tolerance of asphalt binder at intermediate temperatures. ASECE J. Mater. Civil Eng. 29(9) (2017)
Effect of RAP Gradation on Fracture Tolerance of Asphalt Mixtures
613
7. Yan, Y., Preti, F., Romeo, E., Tebaldi, G., Roque, R.: Fracture energy density of interstitial component of asphalt mixtures. Mater. Struct. 51 (2018) 8. Al-Qadi, I.L., Carpenter, S. H., Roberts, G., Ozer, H., Aurangzeb, Q.: Determination of usable residual asphalt binder in RAP. Illinois Department of Transportation Research Report ICT-09031, Springfield, IL, USA (2009)
Effect of Soybean Oil Derived Additives on Improved Performance of Polymer-Modified Asphalt Binder and Mix Containing 50% Fine-Graded RAP Joseph Podolsky, Austin Hohmann, Conglin Chen, Zahra Sotoodeh-Nia, Nicholas Manke, Barrie Saw, Nacu Hernandez, Paul Ledtje, Michael Forrester, R. Christopher Williams, Eric Cochran, and Theodore Huisman Abstract Recent work has shown that epoxidized plant oil materials work well as rejuvenators in recycled asphalt pavement (RAP) and as enhancers of polymer modification of neat binders. Earlier work using a small-angle x-ray scattering (SAXS) device (Xenoncs Xuess 2.0 UHR) with asphaltenes modified and unmodified with a rejuvenator called epoxidized methyl soyate (EMS) showed a decrease in asphaltene particle size by up to 20% when the rejuvenator was used at a dosage of 2.75% by total weight of the binder for modification. In addition, past work also showed epoxidized plant oil materials when used with sulfur eliminated phase separation of SBS polymers from the asphalt matrix and promoted improved elastic recovery. For this work a series of experiments were undertaken in rheology, and mix performance with a interlayer mix design containing 50% RAP to better understand why epoxidized plant oil materials have these effects. Two epoxidized plant oil materials and a commercial rejuvenator derived from corn oil were used for comparison purposes. In this work it was shown that the two epoxidized plant oil materials made significant improvements to beam fatigue and low temperature mix performance as well as low temperature binder performance. However, only one of them sub epoxidized soybean oil (SESO) did not degrade the elastic recovery of the base polymer modified binder as shown through multiple stress creep recovery (MSCR) testing of the rejuvenated binders.
J. Podolsky (B) · A. Hohmann · Z. Sotoodeh-Nia · B. Saw · N. Hernandez · P. Ledtje · M. Forrester · R. Christopher Williams · E. Cochran · T. Huisman Iowa State University, Ames, IA, USA e-mail: [email protected] C. Chen Australian Road Research Board, Melbourne, Australia N. Manke MSA Professional Services, Baraboo, WI, USA © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_78
615
616
J. Podolsky et al.
Keywords Rheology · Epoxidized plant oil materials · Low temperature performance · Fatigue performance
1 Introduction Use of reclaimed asphalt pavement (RAP) in the production of new and rehabilitated asphalt pavements has experienced growing interest over the past several years. The primary reasons behind increased interest and use of RAP in asphalt paving is due to emission and fuel usage reductions that take place during the extraction and transporting of virgin materials [1]. The major challenge of increasing the amount of RAP in asphalt mixtures is because there is too much variability within the material’s composition due to the fact that a pile of RAP held by a contractor could have been produced from RAP collected over several years from different locations (different aggregates, and asphalt binder) [2]. Past research has shown that as long as RAP contents in asphalt mix are kept below 15% there is no significant influence to asphalt mix performance [3–5]. Once the RAP content goes above 15%, higher variability within the RAP could affect asphalt mix performance adversely [6]. One way to alleviate concerns due to high variability within RAP is through better processing of RAP before use in the production of asphalt mixes. One method to better process RAP is called fractionation, which essentially is the splitting of RAP into two or more stockpiles based on different sieve sizes. RAP fractionation allows for better quality control of RAP and allows higher amounts of RAP to be used in asphalt mix designs [7]. The focus of this study is to examine the effects of soybean oil derived additives as rejuvenators on binder performance and the mix performance of a HMA interlayer mix design using 50% fine-graded RAP materials. The requirements for an interlayer mix design to meet specifications are extremely high as the binder must have 90+% elastic recovery at 58 °C at a creep stress level of 3.2 kPa and mix must pass at least 100,000 cycles in beam fatigue testing according to most owner/agency specifications. Thin interlayers are known to be an effective treatment to relieve stresses and improve resistance to reflective cracking in asphalt mix overlays built above an interlayer. In the State of Iowa, RAP let alone, fractionated fine-graded RAP is not allowed in interlayer mix designs due to concerns about over stiffening of the mix causing resistance to reflective cracking to decrease. Thus, the focus of this research is to evaluate various soybean oil derived rejuvenators effects on an interlayer mix design containing 50% fine-graded RAP materials using the multiple stress creep recovery (MSCR) and bending beam rheometer (BBR) tests and beam fatigue and disk-shaped compact tension (DCT) tests to evaluate binder and mix performance.
Effect of Soybean Oil Derived Additives …
617
2 Materials and Experimental Methods The neat asphalt binder used for this research was a polymer modified PG 58-34E+ procured from a binder supplier located in Northwest Iowa, and the RAP was obtained from a contractor located in central Iowa. The Iowa Department of Transportation (DOT) bituminous lab determined properties of the RAP, which were then used to develop a mix design once the RAP was fractionated, based on material above and below the No. 4 Sieve to create two different stockpiles (fine and coarse RAP). The interlayer mix design was composed of 50% fine RAP and 50% manufactured sand. The job mix formula data as well as the blended mix gradation data for the interlayer mix design are shown in Table 1. The soybean oil derived rejuvenators used for this study are called epoxidized methyl soyate (EMS) and sub epoxidized soybean oil (SESO) while a commercial rejuvenator derived from corn oil (Jive) was used for comparison purposes. The dosage of the rejuvenators in the polymer modified PG 58-34E+ was 8% by weight of the binder added or 0.368% by total mix weight. EMS is produced from the methanolysis of epoxidized soybean oil (ESO), while SESO is produced from the partial epoxidation of soybean oil (ESO). For binder performance evaluation, the multiple stress creep recovery (MSCR) and bending beam rheometer (BBR) tests will be used to evaluate how the three rejuvenators affect elastic recovery at three temperatures (46, 52 and 8 °C) and estimate the critical low temperature as well as the Table 1 Job mix formula and gradation information
Property
Interlayer mix
Sieve size (mm)
Percent passing (%)
Air voids (%) 2
25.4
100.0
VMA (%)
17.7
19
100.0
VFA (%)
88.7
12.7
100.0
Film thickness (μm)
9.58
9.51
100.0
Filler to bit ratio
0.97
4.76
97.0
Gse
2.671
2.38
71.0
Pbe
6.91
1.19
45.0
Pba
0.42
0.595
27.0
Total AC (%)
7.3
0.297
15.0
Virgin AC (%)
4.6
0.149
9.7
AC from RAP (%)
2.7
0.074
6.7
Recycled binder ratio (%)
37
Pan
0.0
618
J. Podolsky et al.
Tc according to AASHTO MP 19-10, TP 70-13 and T 313-12. Due to there being such a high amount of RAP (50% fine RAP) and high amount of recycled binder (37%) in the interlayer mix design it was decided that disk compact tension (DCT) testing (ASTM D7317), with Brovold Low Temperature Fracture equipment at low temperatures would be done in addition to 4-point beam fatigue testing using an IPC Beam Fatigue Apparatus (ASTM D7460) according to Iowa DOT specifications for mix performance evaluation (Matls IM 510 and Matls IM 510 App A).
3 Results and Analysis The MSCR was performed on the control and the rejuvenated group at 58, 52, and 46 °C to examine the effects of rejuvenator addition on elastic recovery of the polymer modified PG 58-34E+ binder. The two main outputs from this test are percentage of elastic recovery (R) and non-recoverable creep compliance, Jnr , at the two creep stress levels of 0.1, and 3.2 kPa. Polymer modified binders with E+ (90+% elastic recovery) are rated based on the R value determined for the 3.2 kPa creep stress level according to Iowa DOT specifications. As shown in Table 2 the R values for Table 2 MSCR and BBR results Name
A
B
C
D
Test temp (°C)
% Recovery
Jnr (1/kPa)
Load level 0.1 kPa
Load level 3.2 kPa
Load level 0.1 kPa
Load level 3.2 kPa
58
59.06
96.86
0.6891
0.0225
52
97.32
96.29
0.0083
0.0118
46
91.13
95.16
0.0168
0.0093
58
98.48
96.22
0.0135
0.0327
52
98.84
96.26
0.0067
0.0219
46
47.80
95.57
0.3610
0.0165
58
91.6
68.85
0.1805
0.7064
52
95.11
88.58
0.0692
0.1577
46
97.16
93.64
0.0366
0.0804
58
93.70
61.87
0.4150
2.1754
52
97.55
91.73
0.1135
0.2580
46
98.40
95.25
0.0518
0.1137
Critical low temp (°C)
Stiffness (1) or m-value controlled (2)
Tc
−36.24
1
1.95
−45.22
2
−0.11
−43.23
2
−4.44
N/A
N/A
N/A
= PG 58-34E +, B = PG 58-34E + w/8% SESO, C = PG 58-34E + w/8% EMS, and D = PG 58-34E + w/8% Jive
*A
Effect of Soybean Oil Derived Additives …
619
creep stress level 3.2 kPa do not decrease with increasing temperature for group B (SESO) whereas the other two rejuvenator groups both decrease R with increased temperature. In addition to MSCR testing, BBR testing was done to examine the three rejuvenator effects on the low temperature performance grade of the PG 5834E+ binder. SESO and EMS show improvement to the critical low temperature as compared to the control PG 58-34E+, but this is not the case for Tc . In the case of Tc positive means stiffness controlled while negative means m-value controlled. As binders age they become more m-value controlled. For JIVE, the critical low temperature and Tc could not be determined as the binder was too soft at −24 and −30 °C to successfully test. The only result was gained from testing at −24 °C, which was stiffness of 1.86 MPa and m-value of 0.44 leading the authors to believe that this binder would be highly m-value controlled for Tc . Out of the three rejuvenators used, SESO was the best performing in terms of critical low temperature and Tc . The results of doing the interlayer test according to Iowa DOT specifications (10 Hz, 2000 μ strain, failure at 50% of initial flexural stiffness—established at 200th load cycle) on three specimens from each of the four interlayer groups made with 50% fine-graded are shown in Fig. 1a as average values (error bars signifying 1 standard deviation in each direction). The criteria to pass this test is that Nf must be greater than 100,000. Based on this criteria both SESO and EMS met this requirement. Effects seen in Nf are mirrored in cumulative dissipated energy as shown in Fig. 1b. To evaluate low temperature performance of the 50% RAP interlayer mix design for the control and the three rejuvenator groups, DCT testing was conducted in accordance with ASTM D7313 at −24 °C on three specimens for each group. Figure 2 presents the average DCT test results for each group (error bars signify 1 standard deviation in each direction). From these results all the groups pass the criteria of standard traffic (400 J/m2 ) at −34 °C based on the time temperature superposition principle of −10 °C lower than the test temperature −24 °C. However, the three
Fig. 1 Interlayer beam fatigue results at 2000 με and 20 °C for a load cycles (Nf ), and b cumulative dissipated energy (MJ/m3 ).
620
J. Podolsky et al.
Fig. 2 DCT results for the control and the rejuvenated mixtures
rejuvenated groups meet high traffic criteria of 460 J/m2 , but only the EMS group meets the very high traffic criteria of 690 J/m2 as established by the Iowa DOT.
4 Conclusions and Future Recommendations Based on the work done in this research, it is possible to develop an interlayer mix design with a high amount of fine-graded RAP, however this is dependent on the effects from a rejuvenator. From this work it was seen that both SESO and EMS meet fatigue criteria as established by the Iowa DOT, but only EMS meets the low temperature criteria of the DCT at −24 °C for very high traffic. This result is misleading as the binder performance from SESO is the strongest, and it could be interpreted that if there were 0.5% more of SESO in the PG 58-34E+ then the low temperature mix and binder performance would be met for −46 °C. The most interesting result from this work is that SESO does not degrade the elastic recovery of the +90% elastic recovery PG 58-34E+ at 58 °C and actually spreads the low temperature grade down further thus increasing the continuous grade range even more.
References 1. West, R.: Reclaimed asphalt pavement management: best practices, Auburn. National Center for Asphalt Technology, NCAT Draft Report, AL (2010) 2. Hajj, E., Sebaaly, P., Kandiah, P.: Use of reclaimed asphalt pavements (RAP) in airfields. HMA pavements, AAPTP Project (05-06) (2008)
Effect of Soybean Oil Derived Additives …
621
3. Al-Qadi, I.L., Elseifi, M., Carpenter, S.H.: Reclaimed asphalt pavement—a literature review (2007) 4. Huang, B., Zhang, Z., Kingery, W., Zuo, G.: Fatigue crack characteristics of HMA mixtures containing RAP. In: Proceeding 5th International Conference on Cracking in Pavements, RILEM, pp. 631–638 (2004) 5. Shah, A., McDaniel, R., Huber, G., Gallivan, V.: Investigation of properties of plant-produced reclaimed asphalt pavement mixtures. Transp. Res. Record J. Transp. Res. Board 1998, 103–111 (2007) 6. Al-Qadi, I.L., Aurangzeb, Q., Carpenter, S.H., Pine, W.J., Trepanier, J.: Impact of high RAP contents on structural and performance properties of asphalt mixtures (2012) 7. Valdés, G., Pérez-Jiménez, F., Miró, R., Martínez, A., Botella, R.: Experimental study of recycled asphalt mixtures with high percentages of reclaimed asphalt pavement (RAP). Constr. Build. Mater. 25(3), 1289–1297 (2011)
Effect of Temperature on Self-healing Properties of Bituminous Binders Ezio Santagata , Fabrizio Miglietta , Orazio Baglieri , and Lucia Tsantilis
Abstract Healing properties of bituminous binders are commonly evaluated by subjecting test specimens to oscillatory shear loading with inclusion of single or multiple rest periods. In almost all protocols adopted by researchers, during both loading and rest phases the same constant temperature is imposed to the specimen. However, it is widely recognized that temperature may have a great impact on healing mechanisms taking place in the binder. In order to address this issue, this paper investigates the healing properties of bituminous binders by means of a test procedure which consists of two time sweep tests conducted at the same temperature and by an intermediate rest period whose temperature is set at different values. Response of the material is quantitively assessed by introducing two healing indicators based on stiffness and strength recovery, respectively. Discussion of the experimental results presented in this paper underlines the relevance of temperature which needs to be taken into account to correctly rank bituminous binders in terms of their self-healing potential. Keywords Bituminous binder · Self-healing · Temperature
1 Introduction Self-healing is commonly recognized as an intrinsic property which characterizes various materials, such as bitumen, capable of restoring their functionality after being subjected to damage [1]. From an engineering point of view, self-healing counteracts damages occurring in the form of nano/microcracks, while it has no effect with respect to macro-damage since most materials do not sufficiently flow to close open cracks [2]. Starting from the theory of polymer crack healing [3], the self-healing process in bitumen has been described by referring to a three-step mechanism. This consists E. Santagata (B) · F. Miglietta · O. Baglieri · L. Tsantilis Department of Environment, Land, and Infrastructure Engineering, Politecnico di Torino, Corso Duca Degli Abruzzi, 24, 10129 Torino, Italy e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_79
623
624
E. Santagata et al.
of an initial approach of crack surfaces caused by bitumen flow, followed by microcrack closure due to wetting, and completed by final interdiffusion and randomization of bitumen molecules which eventually lead to the complete restoration of initial properties [4]. In addition to bitumen inner characteristics [5, 6], the overall process is affected by several extrinsic factors, such as loading conditions [7], rest periods [8] and temperature [9]. Moreover, the time-dependent phenomenon known as thixotropy superposes to true self-healing through the rearrangement of bitumen micro-structure [10, 11]. Current methods proposed for the assessment of the healing capability of bitumen mainly involve the use of the Dynamic Shear Rheometer (DSR) via fatigue-based tests. These are performed with interrupted [12, 13] or intermittent loading conditions [14] in a range of temperatures which are selected in order to ensure cohesive failure. Although most tests are carried out at constant temperature, it is well known that temperature changes may significantly affect the healing properties of bitumen [15]. This paper investigates the influence of rest period temperature on the healing process of bituminous binders subjected to interrupted loading conditions. Furthermore, it presents a methodology for the evaluation of self-healing properties which also takes into account thixotropic effects.
2 Materials and Methods Three bituminous binders of different origin and penetration grade were employed in the experimental investigation: two neat bitumens (B1 and B2, belonging to the 70/100 and 40/60 penetration grade, respectively), and an SBS-modified binder (B3, classified as a 25/55 penetration grade binder). As reported in Table 1, preliminary characterization of the binders at high and low temperatures included the determination of softening point (EN 1427) and glass transition temperature [16]. For the assessment of their self-healing potential, considered binders were tested with a DSR (Physica MCR 301 from Anton Paar Inc.) by following the procedure described below, indicated as FRF (Fatigue-Rest-Fatigue). By making use of the 8 mm parallel-plates geometry, binder specimens (characterized by a 2 mm height) are initially conditioned for 30 min at a selected temperature which is thereafter assumed as fatigue test temperature. Fatigue damage in the specimens is induced in two phases of strain-controlled oscillatory loading at a frequency of 10 Hz. A prolonged rest period is interposed between the two fatigue loading phases. Table 1 Results of preliminary characterization of considered binders
B1 Softening point, TR&B (°C) Glass transition, Tg (°C)
B2
B3
45.8
53.6
62.1
−17.8
−27.2
−19.1
Effect of Temperature on Self-healing Properties … Table 2 Testing conditions
625
Material
T (°C) during loading
T (°C) during rest period
B1
10
10, 20, 30
B2
10
10, 20, 30
B3
14
14, 20, 30
Fatigue test temperature was selected in order to ensure iso-stiffness conditions for all binders, with a corresponding initial value of the norm of the complex modulus equal to 34 MPa. This is comprised in the 10–50 MPa range which has been proven to be suitable for triggering the onset of cohesive fatigue microcracks rather than for causing adhesion loss or instability flow [17]. Imposed strain amplitude, required to be close to the linear viscoelastic limit, was identified by means of preliminary strain amplitude sweep tests. For the considered binders the most appropriate value was found to be equal to 2%. The first fatigue loading phase was interrupted when a 50% reduction of the initial value of the norm of the complex modulus was recorded. The second fatigue loading phase (after rest) was conducted until failure. Rest period had a fixed duration of 2 h, during which the norm of the complex modulus was monitored by applying an oscillatory shear strain of 0.1%. Temperature imposed during the rest period was varied in order to assess its impact on the healing properties of the binders. The last 30 min of the rest period were used to bring the specimen back to fatigue test temperature and to achieve stable thermal conditions. The above described FRF test procedure was supplemented with tests (indicated as RF tests) characterized by a single fatigue loading phase preceded by a prolonged rest period. These tests were intended to provide data for estimating thixotropic effects that need to be subtracted from those of FRF tests in order to assess true self-healing. A summary of adopted testing conditions is given in Table 2. At least two replicates were performed for each test and average results were considered in analyses. A schematic representation of results obtained from FRF and RF tests is provided in Fig. 1.
Fig. 1 Schematic representation of the results obtained from FRF a and RF, b tests
626
E. Santagata et al.
By referring to the results of FRF and FR tests, quantification of the healing capability of bituminous binders was carried out by introducing the following two indices, in which spurious thixotropic effects are duly removed. As indicated in Eqs. (1) and (2), these indices, denoted as IG and IN , reflect stiffness and fatigue endurance gain, respectively: IG =
|G ∗ | R−F R F − |G ∗ | R−R F · 100 |G ∗ | F−F R F
(1)
NF RF · 100 NRF
(2)
IN =
where |G ∗ | R−F R F is the difference between the norm of the complex modulus recorded at the end and at the beginning of the rest period (|G ∗ | R, f −F R F and |G ∗ | R,i−F R F , respectively) in FRF tests; |G ∗ | F−F R F is the difference between the norm of the complex modulus recorded at the end of the thermal conditioning phase and that at the beginning of rest period (|G ∗ |C, f −F R F and |G ∗ | R,i−F R F , respectively) in FRF tests; |G ∗ | R−R F is the difference between the norm of the complex modulus recorded at the end and at the beginning of the rest period (|G ∗ | R, f −R F and |G ∗ | R,i−R F , respectively) in RF tests; N F R F and N R F are the number of loading cycles applied to reach 50% reduction of the initial value of the norm of the complex modulus in the second loading phase of FRF tests and in the single loading phase of RF tests, respectively.
3 Results and Discussion Values of the abovementioned indices obtained from the processing of experimental data are reported in the IN -IG diagram displayed in Fig. 2. As indicated by the fact that all data points are above the IN = IG identity line, in all cases (i.e. for each binder at all rest period temperatures) stiffness gain (IG ) was found to be greater than endurance
IG [%]
100
B1 (10 °C)
B1 (20 °C)
B1 (30 °C)
80
B2 (10 °C)
B2 (20 °C)
B2 (30 °C)
60
B3 (14 °C)
B3 (20 °C)
B3 (30 °C)
40 20 0 0
20
40
60
80 100
IN [%] Fig. 2 Comparison between self-healing indices obtained from FRF and RF tests
Effect of Temperature on Self-healing Properties … 100
IN [%]
80 60
627 B1
T R&B (B1)
T g (B1)
B2
T R&B (B2)
T g (B2)
B3
T R&B (B3)
T g (B3)
40 20 0 -30 -20 -10 0
10 20 30 40 50 60 70
Temperature [°C] Fig. 3 Temperature dependency of fatigue endurance index
gain (IN ). Such an outcome can be explained by referring to the previously described three-step self-healing mechanism. In particular, it can be postulated that changes in stiffness measured in the low strain conditions imposed during rest periods may be affected both by microcrack closure due to wetting and by partial interdiffusion and randomization of bitumen molecules. On the contrary, it is reasonable to hypothesize that when imposing higher strains in the fatigue loading phases, the weak interactions occurring at microcrack faces do not reflect into a quantitatively equivalent endurance gain. A similar gain in stiffness and fatigue endurance is thus expected only in those cases in which the self-healing process is fully developed, with the restoration of the pristine homogeneous matrix. In the case of the experimental data gathered during the performed investigation, these conditions were found only for binder B1 when adopting a rest period temperature of 30 °C, with similar values of IG and IN . As expected, for all considered binders both indices increased with temperature. For a comparison of the different self-healing capability of the considered bituminous binders, IN was plotted as a function of temperature as shown in Fig. 3. It can be observed that data points follow a characteristic sigmoidal trend within an upper and lower boundary. For the purpose of fitting the data to the sigmoidal equation, it was assumed that the upper boundary, corresponding to complete healing, is approached at the softening point temperature (TR&B ) [15], while the lower boundary is reached in the vicinity of the glass transition temperature (Tg ) where no flow can occur at microcrack faces. It can be observed that bitumen B1 exhibited the greatest self-healing potential at all temperatures and also showed the greatest temperature sensitivity (i.e. highest slope of the fitting curve) in the intermediate temperature range corresponding to visco-elastic behavior. In terms of relative ranking, binder B1 was followed by B2 and B3, with the lowest self-healing aptitude being displayed by the SBS-modified binder.
628
E. Santagata et al.
4 Conclusions The experimental study presented in this paper investigated the effect of temperature on the healing potential of bituminous binders, quantitively assessed by means of two healing indices which refer to stiffness and fatigue endurance gain, respectively. Obtained results indicate that the proposed indices provide information which is consistent with the kinetics of self-healing and highlight the relevance of temperature conditions which occur during rest periods. Direct comparison of the healing potential of the binders considered in the investigation shows that SBS modification, which is known to reduce crack accumulation rate, does not necessarily lead to enhanced self-healing properties. On the other hand, when allowed to rest at sufficiently high temperatures, neat bitumen can exhibit remarkable self-healing properties which lead to mechanical properties similar to those of the undamaged state.
References 1. Sun, D., Sun, G., Zhu, X., Guarin, A., Li, B., Dai, Z., Ling, J.: A comprehensive review on self-healing of asphalt materials: mechanism, model, characterization and enhancement. Adv. Colloid. Interf. 256, 65–93 (2017) 2. García, Á.: Self-healing of open cracks in asphalt mastic. Fuel 93, 264–272 (2012) 3. Wool, R.P., O’Connor, K.M.: A theory crack healing in polymers. J. Appl. Phys. 52(10), 5953–5963 (1981) 4. Phillips, M.C.: Multi-step models for fatigue and healing, and binder properties involved in healing. In: Proceedings of Eurobitume Workshop on Performance Related Properties for Bituminous Binders. Luxembourg, no. 115 (1998) 5. Kim, Y.R., Little, D.N., Benson, F.C.: Chemical and mechanical evaluation of healing mechanisms in asphalt concrete. J. Assoc. Asphalt Pav. Tech. 59, 240–275 (1990) 6. Santagata, E., Baglieri, O., Dalmazzo, D., Tsantilis, L.: Rheological and chemical investigation on the damage and healing properties of bituminous binders. J. Assoc. Asphalt Pav. Tech. 68, 567–595 (2009) 7. Zhang, Z., Rogue, R., Birgisson, B., Sangpetngam, B.: Identification and verification of a suitable crack growth law. J. Assoc. Asphalt Pav. 70, 206–241 (2001) 8. Bahia, H.U., Zhai, H., Bonnetti, K., Kose, S.: Non-linear viscoelastic and fatigue properties of asphalt binders. J. Assoc. Asphalt Pav. 68, 1–34 (1999) 9. Menozzi, A., Garcia, A., Manfred, N.P., Tebaldi, G., Schuetz, P.: Induction healing of fatigue damage in asphalt test samples. Constr. Build. Mater. 74, 162–168 (2015) 10. Santagata, E., Baglieri, O., Tsantilis, L., Dalmazzo, D.: Evaluation of self-healing properties of bituminous binders taking into account steric hardening effects. Constr. Build. Mater. 41, 60–67 (2013) 11. Shan, L., Yiqiu, T., Underwood, B.S., Kim, Y.R.: Thixotropic Characteristics of Asphalt Binder. J. Mater. Civ. Eng. 23(12), 1681–1686 (2011) 12. Santagata, E., Baglieri, O., Dalmazzo, D., Tsantilis, L.: Investigating cohesive healing of asphalt binders by means of a dissipated energy approach. Int. J. Pav. Res. Technol. 10, 403–409 (2017) 13. Wang, C., Xie, W., Underwood, B.S.: Fatigue and healing performance assessment of asphalt binder from rheological and chemical characteristics. Mater Struct. 51(6), 171 (2018) 14. Canestrari, F., Virgili, A., Graziani, A., Stimilli, A.: Modeling and assessment of self-healing and thixotropy properties for modified binders. Int. J. Fatigue 70, 351–360 (2015)
Effect of Temperature on Self-healing Properties …
629
15. Tang, J., Liu, Q., Wu, S., Ye, Q., Sun, Y., Schlangen, E.: Investigation of the optimal self-healing temperatures and healing time of asphalt binders. Constr. Build. Mater. 113, 1029–1033 (2016) 16. Elwardany, M.D., Planche, J.-P., Adams, J.J.: Determination of binder glass transition and crossover temperatures using 4-mm plates on a dynamic shear rheometer. Transp. Res. Record 2673(10), 247–260 (2019) 17. Safaei, F., Hintz, C.: Investigation of the effect of temperature on asphalt binder fatigue. Asphalt pavements. In: Proceedings of the International Conference on Asphalt Pavements, ISAP, vol. 2, pp. 1491–1500. CRC Press (2014)
Effect of Thermal Ageing on the Mechanical Properties and Cracking Behaviour of Asphalt Concrete B. Kouevidjin, J.-F. Barthélémy, C. Somé, H. Ben Dhia, and A. Feeser
Abstract This article assesses the impact of ageing on the properties of asphalt concrete (AC) by combining experiments and numerical simulations. Rheological tests were performed at four thermal ageing durations (0, 3, 6 and 9 days). The results show an increase in the complex modulus and a decrease in phase angle with ageing time. For each ageing time, semi-circular bending tests were carried out under two loading rates (1 and 5 mm/min) and three temperatures (−20, 0 and 20 ◦ C). The results shown an increase of the peak load when the loading rate increases for each thermal ageing time and temperature. No meaningful variation of the peak load with respect to the ageing time was observed at low temperatures. However, it was found that at 20 ◦ C, the peak load increases with ageing duration. A finite element modelling was implemented, coupling viscoelastic behaviour and crack growth criterion. The results show a similarity between numerical and experimental results. Keywords Asphalt concrete · Viscoelasticity · Crack growth · Thermal ageing
1 Introduction Cracking which represents one of the main forms of pavement pathologies, is often accelerated by the thermal ageing phenomenon of Asphalt Concrete (AC). A better understanding of the effect of thermal ageing on the rheological and fracture properties of AC is therefore a key step in predicting the service life of pavement structures. Significant research have been conducted to investigate the mechanical and chemical aspects of AC thermal ageing [1, 2]. Recent studies have shown that, B. Kouevidjin · J.-F. Barthélémy (B) · C. Somé Cerema, Project-Team DIMA, 77171 Sourdun, France e-mail: [email protected] B. Kouevidjin · H. Ben Dhia CentraleSupelec, MSSMat Laboratory, Paris-Saclay University, 91191 Gif-sur-Yvette, France A. Feeser Cerema, Laboratory of Strasbourg, 67000 Strasbourg, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_80
631
632
B. Kouevidjin et al.
in general, at low temperatures, fracture parameters are more degraded as the ageing time increases [3, 4]. However, few studies deal with the impact of thermal ageing on cracking properties under different loading rates and different temperatures. On the numerical level, models of fracture mechanics have been developed by several authors. Most of them have been applied to AC while describing them as elastic materials. Theoretical analyses of viscoelastic fracture mechanics and their applications to viscoelastic materials such as AC are quite limited. Continuum damage models such as cohesive zone model [5] are often used to predict the fracture behaviour of AC. These models are simple to implement but very sensitive to the mesh size. Schapery et al. [6] have developed numerical methods to determine the mechanical fields at the crack tip and the energy release rate in the crack propagation process over time. Nguyen et al. [7] studied crack growth in viscoelastic materials using an energetic approach by adopting a direct calculation procedure for the energy release rate. The effect of loading rate, thermal ageing and temperature on viscoelastic crack growth are the issues addressed in this paper. In the next section, rheological and cracking tests of AC subjected to thermal ageing are presented. Sect. 3 describes an overall energetic approach to study the crack growth in AC based on the energetic criterion described in [7]. In Sect. 4, computational and experimental cracks growth were compared in order to verify the accuracy of the numerical models at different ageing times and loading rates.
2 Rheological and Crack Growth Properties of Asphalt Concrete 2.1 Materials and Method Semi-coarse asphalt concrete (AC 10) is used in this study. The bitumen used is a conventional paving grade bitumen, with a measured penetrability of 42 × 0.1 mm and a softening temperature of 53.2 ◦ C.
2.2 Laboratory Ageing Procedure and Specimen Preparation In order to simulate in the laboratory the AC thermal ageing over the life of pavements, the RILEM method reported in CEN/TS 12697-52 is adopted. In this ageing procedure, the loose bituminous mixture is manufactured then spread in thin layer (25 mm) into pans. Once the mixture is cooled down to 135 ◦ C, the pans were placed in an oven at 135 ◦ C for 8 ± 0.5h. Then, the mixture is cooled to 85 ◦ C and maintained at this temperature for three long-term ageing durations (3, 6, 9 days). Reference mix (t0) and aged mixes (named t3, t6, t9) were compacted then sliced to obtain semi-circular and prismatic samples.
Effect of Thermal Ageing on the Mechanical Properties …
633
2.3 Complex Modulus Test The complex modulus test were carried out on prismatic specimens using the twopoint bending method described in NF EN 12697-26. The tests were performed in the linear viscoelastic range at five frequencies (3 to 40 Hz) and nine temperatures (−30 to 50 ◦ C).
2.4 Semi-Circular Bending (SCB) Test In order to evaluate the fracture characteristics of AC in this study, the SCB test was used. The SCB samples have a notch of length 10 and 0.35 mm of width. Three testing temperatures (−20, 0 and 20 ◦ C) and two loading rates (1 and 5 mm/min) were selected for each thermal ageing time. For more details about the test, please refer to NF EN 12697-44.
2.5 Experimental Results and Discussion 2.5.1
Rheological Analysis
The variations in rheological properties due to ageing were examined in this paragraph. The results shown in Fig. 1a, b indicate an increase in the complex modulus and a decrease in the phase angle between 0 and 9 days of ageing. Indeed, viscoelastic properties are more affected by thermal ageing at very low frequencies. The full line curves plotted in Fig. 1a, b illustrate the calibration of the Generalised Maxwell (GM) for a reference temperature θ R = 0 ◦ C. The results of complex modulus and phase angle given by the model are nearly the same as the experimental results. The ondulations observed on the master curves are often related to the number of Maxwell chains used to model the viscoelastic behaviour of the asphalt. For our study, 19 chains were sufficient to predict the overall behaviour of our material.
2.5.2
Fracture Analysis
Fig. 2a, b show a slight increase of the peak load when the loading rate increases. However, no significant variation of the peak load with respect to ageing times was observed at these temperatures. The results obtained at 20 ◦ C (Fig. 2c) indicated a significant increase of the peak load when the loading rate increases. In contrast to the results obtained at −20 and 0 ◦ C, an increase of peak load was observed at 20 ◦ C when the ageing time increases. Finally, it can been seen that, for each ageing time, the peak load decreases with increasing temperature due to the material’s viscoelasticity.
634
B. Kouevidjin et al.
(b)
(a) Fig. 1 Master curves of: a |E ∗ | at θ R = 0 ◦ C; b ϕ at θ R = 0 ◦ C
(a)
(b)
(c)
Fig. 2 Effect of ageing and loading rate on peak load at: a −20 ◦ C; b 0 ◦ C; c 20 ◦ C
(b)
(a)
(c)
Fig. 3 Effect of ageing and loading rate on fracture strain at: a −20 ◦ C; b 0 ◦ C; c 20 ◦ C
Figure 3a–c shown a decrease in fracture strain (1) as a function of ageing time, due to the hardening of the material. ε max =
U Fmax × 100 W
(1)
where U Fmax is the displacement at the peak load and W the height of the specimen.
Effect of Thermal Ageing on the Mechanical Properties …
635
3 Viscoelastic Crack Growth Modelling In this section, the energetic approach modelling crack propagation in mode I in a viscoelastic medium is presented [7]. Let’s consider a pre-cracked domain (l) (see Fig. 4a). For clarity’s sake, a progressive kinematic loading ξ(t) is defined whose dual is noted Q(t). In the case of an isothermal and uniform transformation, the overall Clausius-Duhem’s inequality shape can be written as: .
.
D = Qξ − W ≥ 0
(2)
where W denotes the elastic strain energy stored in the system. The viscoelastic behaviour is modelled by the GM model shown in Fig. 4b. This behaviour can be formulated as follows [7, 8]:
σ (t) = Cel : ε(t) − σ r (t)
avec Cel =
∼
∼
r i=0
C
et σ r (t) =
∼i
r i=1
C : ε ϑ (t) (3) ∼i
i
where εϑ , C are respectively the viscous strain and elasticity tensor of Maxwell’s i
∼i
i-th element. σ r can be interpreted as pre-stress tensor. From (3), it can be shown that the elastic strain energy does not depends only on the load (ξ ) and the geometry (l) but also on the field of internal variables {εϑ , i = 1, r } [7, 8]: i
1 W(ξ, l, {ε }) = 2 ϑ
(l)
r ε − εϑ : C : ε − εϑ ε:C :ε+ ∼0
i
i=1
∼i
i
d(l) (4)
Introducing (4) into (2) and assuming that the dissipation related to the evolution of the crack is proportional to the length rate of the crack, the crack growth criterion can be formulated as follows [7]:
Fig. 4 a SCB notched geometry. b GM model
636
B. Kouevidjin et al. . . ∂W D − D = Gv = − l = Rl ⇒ ∂l |ξ,{εϑ } ϑ
.
Gv < R ⇒ l = 0 . Gv = R ⇒ l ≥ 0
(5)
. ∂W {εv } , Gv et R are respectively viscous dissipation, energy v ∂{ε } |ξ,l release rate and cracking resistance. A detailed description crack growth modelling can be found in [8].
where
Dϑ = −
4 Comparison Between Numerical Model and Experiments In the framework of numerical simulation, the three-point bending problem was implemented using the finite element library Getfem++ [9]. The R-values were determined on notched specimens. At 0 ◦ C, it was assumed that no energy was dissipated away from the crack tip and that all the energy produced during the extension of the crack is a surface energy. Thus the R-values were calculated by using the RILEM fracture energy formula [10] and summarized in Table 1. The results of the computational simulations were validated by comparing them with the experimental results, as shown in Fig. 5a, b. In a general way, it can be seen that, for each ageing time and loading rate, the numerical and experimental load-displacement curves are consistent for the selected specimens. However, when several specimens are tested, account should be taken of the heterogeneity between specimens which may lead to a dispersion of the test results.
Table 1 Cracking resistance values Loading rate R value for t0 (mm/min) (N/m) 1. 5.
34.43 33.92
R value for t3 (N/m)
R value for t6 (N/m)
R value for t9 (N/m)
30.34 26.49
27.04 20.53
23.12 20.29
(a) Fig. 5 Comparison of F_U curves at: a 1 mm/min. b 5 mm/min
(b)
Effect of Thermal Ageing on the Mechanical Properties …
637
5 Conclusion This study examined the impact of thermal ageing on the mechanical and fracture properties of AC. For this purpose, complex modulus and cracking tests were carried out. Experimental results have shown that thermal ageing increases the stiffness of the material. As a result, the failure deformation decreases when the ageing duration increases. The crack growth modelling approach is then used to simulate semi circular bending problem. Overall, good agreement with experimental results in terms of the force-deflexion curves is obtained for each loading rate and thermal ageing time. This indicates that the model can predict the crack propagation in AC. Aiming at refining the model, while remaining within reasonable numerical simulation times and further approximating the real material composition (heterogeneity), evolutionary multi-model formulations, articulated within the framework of multiscale Arlequin models will be conducted and implemented in the future.
References 1. Petersen, J.C.: A Review of the Fundamentals of Asphalt Oxidation: Chemical. Physical Property, and Durability Relationships. Transportation Research Board, Physicochemical (2009) 2. Kim, Y.R., Castorena, C., Elwardany, M.D., Rad F.Y., Underwood, S., Akshay, G., Gudipudi, P., Farrar, M.J., Glaser, R.R.: Long-Term Aging of Asphalt Mixtures for Performance Testing and Prediction. Transportation Research Board (2018) 3. Braham, A.F., Buttlar, W.G., Clyne, T.R., Marasteanu, M.O., Turos, M.I.: The effect of longterm laboratory aging on asphalt concrete fracture energy. J. Assoc. Asphalt Paving Technol. 78, 417–445 (2009) 4. Farhad, Y.R., Elwardany, M.D., Castorena, C., Kim, Y.R.: Investigation of proper long-term laboratory aging temperature for performance testing of asphalt concrete. Constr. Build. Mater. 147, 616–629 (2017) 5. Barenblatt, G.I.: The mathematical theory of equilibrium cracks in brittle fracture. In: Volume 7 of Advances in Applied Mechanics, pp. 55 – 129. Elsevier (1962) 6. Schapery, R.A.A.: Theory of crak initiation and growth in viscoelastic media. Theoretical development. In: Asphalt Paving Technology: Association of Asphalt Paving TechnologistsProceedings of the Technical Sessions, vol. 78, pp. 417–445 (1975) 7. Nguyen, S.T., Dormieux, L., Le Pape, Y., Sanahuja, J.: Crack propagation in viscoelastic structures: theoretical and numerical analyses. Computat. Mater. Sci. 50(1), 83–91 (2010) 8. Kouevidjin, B., Somé, C., Barthélémy, J.-F., Ben Dhia, H., Feeser, A.: Influence du vieillissement thermique sur les propriétés viscoélastiques et les paramètres de rupture des matériaux bitumineux. In: 24e Congrès Français de Mécanique, pp. 2689–2698. Brest, France (2019) 9. Pommier, J., Renard, Y.: An open source generic c++ it library for finite element methods. http://getfem.org/ 10. RILEM 1985-TC 50-FMC: Determination of the fracture energy of mortar and concrete by means of three-point bend tests on notched beams. Fract. Mech. Concrete RILEM Recommend. Mater. Struct. 18(106)
Effects of Adding a Geotextile into a Pavement Structure on FWD Experimental Data and Estimated Pavement Remaining Service Life Jean-Claude Carret and Alan Carter
Abstract Impregnated geotextiles are commonly used in road construction or rehabilitation to minimize water infiltration into the pavement structure. However, the effects of such waterproofing membranes on the mechanical behavior of the pavement structure are difficult to predict. In this paper, Falling Weight Deflectometer (FWD) measurements were used to highlight differences between pavement structures with and without geotextile. The FWD measurements were performed on two recently rehabilitated roadways where a bitumen impregnated non-woven geotextile was placed between the surface layer and the base layer for some sections. The FWD measurements were performed on both sections rehabilitated with and without the geotextile and the deflection basins were compared. The surface curvature index (SCI), the base damage index (BDI) and the base curvature index (BCI) were also compared. The measurements were then used to predict the remaining service life (RSL) of the pavement structures. The results of this study suggest that traditional approaches used to interpret FWD measurements are not suitable when a bitumen impregnated geotextile is present between two asphalt layers. Keywords Geotextile · Road rehabilitation · FWD measurement · Deflection basin parameters · Remaining service life
1 Introduction Pavement structures deteriorate under the effect of traffic and thermally induced stresses. In addition, water infiltration in pavement structures significantly increases the rate of deterioration [1]. Therefore, rehabilitation operations are essential to
J.-C. Carret (B) · A. Carter Construction Engineering Department, École de Technologie Supérieure (ÉTS), Montreal, QC, Canada e-mail: [email protected] A. Carter e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_81
639
640
J.-C. Carret and A. Carter
enhance the service life of pavement structures. A very common rehabilitation technique is to mill over a thin thickness (about 50 mm) and apply a new asphalt layer (mill and fill). However, this technique does not address the issue of water infiltration and does not bring any structural reinforcement to the structure. An alternative is to use full-depth reclamation (FDR) process and rebuild with an increased overlay thickness but this approach is rather expensive in comparison and requires important new material resources. Another possibility is to adapt the mill and fill technique by laying a geotextile before resurfacing with a new asphalt overlay. Geotextiles can act as waterproofing membranes that slow water infiltration [2, 3] and improve the resistance to crack propagation [4]. However, the impact of this type of membrane on the mechanical behavior and the life duration of the pavement structure is difficult to predict and varies depending on many factors [5]. The aim of this paper is to compare sections rehabilitated with the technique using FDR process and the mill and fill technique with an impregnated geotextile added before resurfacing. To do so, FWD measurements were performed in sections of two newly rehabilitated boulevards and in sections that were not rehabilitated. FWD deflection basin parameters of the different sections were compared. Finally, the estimated remaining service life of the sections were calculated and compared.
2 Methodology Falling Weight Deflectometer (FWD) measurements were performed two boulevards (A and B) on the south shore of Montreal (QC, Canada). These two boulevards were rehabilitated a few months prior to the measurements. Two rehabilitation techniques were used on different sections of the boulevards and are described in Fig. 1. The first rehabilitation technique consists in milling 50 mm before laying a bitumen impregnated non-woven geotextile and resurfacing with 50 mm of new asphalt. The second rehabilitation technique uses full-depth reclamation process on 300 mm before reconstructing the pavement layers. With this second technique, the new thickness of the asphalt layers is 210 mm for Boulevard A and 150 mm for Boulevard B.
Fig. 1 Rehabilitation techniques. a 50 mm milling, geotextile laying and resurfacing; b 300 mm full-depth reclamation and reconstruction of asphalt layers
Effects of Adding a Geotextile into a Pavement Structure …
641
For each boulevard, FWD measurements were performed in sections rehabilitated with the two rehabilitation techniques and in one section that was not rehabilitated. Overall, 43 measurements were recorded in 6 sections. A summary of the FWD data collected is presented in Table 1. FWD deflection basin parameters are powerful indicators to interpret FWD measurements in terms of mechanical behavior of the pavement structure [6, 7]. They give indications on the condition of the different layers composing the pavement structure. The deflection basin parameters used in this study are introduced in Table 2. The American Association of State Highway and Transportation Officials (AASHTO) introduced the notion of remaining service life (RSL) for pavement structures [8]. Many methods exist to determine the RSL [9] and most of them require to calculate the structural number of the pavement (SN). In this paper, the Rhode’s relationships [10] were used to calculate SN because they only require FWD data and the knowledge of the thickness of the asphalt layers. Details on that calculation can be found elsewhere [10], but the structural numbers are used to calculate the condition factor (CF) of the pavement according to Eq. (1): Table 1 Summary of the FWD data collected Boulevard
Rehabilitation status
Length of section (m)
A
300 mm FDR and reconstruction
103
7
50 mm milling, geotextile laying and resurfacing
178
11
No rehabilitation
350
7
300 mm FDR and reconstruction
45
4
50 mm milling, geotextile laying and resurfacing
107
8
No rehabilitation
240
6
B
Number of FWD measurements
Table 2 FWD deflection basin parameters used in this study Name of the parameter
Equation
Aim of the parameter
Central deflection (D0 )
–
Characterizing the global condition of the structure
Surface Curvature Index (SCI) SCI = D0 − D300
Characterizing the condition of the top pavement layers
Base Damage Index (BDI)
BDI = D300 − D600
Base Curvature Index (BCI)
BCI = D1500 − D900 Characterizing the condition of the subgrade
Characterizing the condition of the base pavement layers
Note D0 , D300 , D600 , D900 and D1500 are the measured deflections in micrometers at the distance of 0, 300, 600, 900 and 1500 mm from the center of the loading plate
642
J.-C. Carret and A. Carter
C F = S N c /S N i
(1)
where SNc is the current structural number of the pavement (µm) and SNi is the initial structural number of the pavement (µm). Then, the remaining service life is determine using the following AASHTO relationship: C F = 1 − 0.7e−(0.01∗RSL+0.85)
2
(2)
3 Results and Analysis The central deflections obtained on Boulevard A are displayed in Fig. 2. The central deflections of the section rehabilitated with 300 mm FDR and reconstruction of the asphalt layers are far lower than those of the other sections. Moreover, central deflections of the non-rehabilitated section and of the section rehabilitated using a geotextile are quite similar. It indicates that the global mechanical condition of the pavement structure was not improved by the rehabilitation technique using the geotextile. The same trend was observed for the other Boulevard. Figure 3 displays the averaged values of SCI, BDI and BCI obtained for the three sections of each boulevard. The same trend was observed on both boulevards. BCI is very similar for the three sections, which indicates that the condition of the subgrade is similar for all sections. This was expected because the subgrade was not modified by the rehabilitation techniques. SCI and BDI are only similar for the sections rehabilitated with the geotextile and for the non-rehabilitated sections. SCI and BDI are higher for those sections than for the section rehabilitated with the 300 mm FDR and reconstruction technique. It indicates that the mechanical condition of the asphalt layers was not improved by the rehabilitation technique using the geotextile. Because of the resurfacing, a higher SCI was expected for the sections Fig. 2 Central deflections for a 40 kN FWD load obtained on Boulevard A in the three measurement zones. Horizontal bars represent the average value of each zone
Effects of Adding a Geotextile into a Pavement Structure …
643
rehabilitated with the geotextile in comparison to the non-rehabilitated sections. However, the geotextile may act as a very thin softening layer that decreases the rigidity of the top layers. The procedure detailed earlier was used to determine the SN, CF and RSL values for all the FWD measurements. In this work, the sections newly rehabilitated using 300 mm FDR and full reconstruction of the asphalt layers were considered new and having 100% of remaining service life. Averaged values of SN, CF and RSL obtained for each section are given in Table 3. The estimated remaining service life of the sections rehabilitated with the geotextile is less than 50% for both boulevards. However, previous studies demonstrated that geotextiles could improve the life duration of pavement structures [5, 11]. Therefore, estimation of remaining service life based solely on FWD measurements may not be adapted for structures having a geotextile between asphalt layers. These results should be verified with further studies because asphalt layers thickness, which is a critical value in the calculation process, is not precisely known and can vary a lot Table 3 Averaged values of the structural number (SN), the condition factor (CF) and the remaining service life (RSL) of the pavement Boulevard
Rehabilitation status
A
300 mm FDR and reconstruction 50 mm milling, geotextile laying and resurfacing No rehabilitation
3.90
B
SN
CF
RSL (%)
4.12
1
100
3.61
0.88
48
0.95
77
300 mm FDR and reconstruction
4.12
1
50 mm milling, geotextile laying and resurfacing
3.56
0.86
100 42
No rehabilitation
3.40
0.82
32
Fig. 3 Averaged values of SCI, BCI and BDI for a 40 kN FWD load obtained on Boulevard A (left) and B (right) in the three measurement zones
644
J.-C. Carret and A. Carter
inside the sections. For example, the estimated RSL of the non-rehabilitated section of Boulevard A is 77%. It is too high considering the degradation state of this section and the estimated RSL of the same section drops to 52% if the thickness of the asphalt layers is set at 140 mm instead of 130 mm.
4 Conclusion In this study, the effect of adding a geotextile in a pavement structure was evaluated. To this end, FWD measurements were performed on non-rehabilitated sections and on sections rehabilitated with two different techniques. First, FWD deflection basin parameters were calculated to compare the condition of the different sections. It was shown that the rehabilitation technique using a geotextile does not improve the mechanical condition of the structure. Moreover, the resurfacing with 50mm of new asphalt does not increase the SCI value in comparison to the values obtained for the non-rehabilitated sections. It may be due to the presence of the geotextile that act as a very thin softening layer. Then, FWD data was used to estimate the remaining service life of each section. Sections rehabilitated with the geotextile have a very reduced estimated RSL of 48% and 42%. Since geotextiles can increase the life duration of pavement structures, the methodology used to estimate the RSL may not be adapted for pavement structures incorporating geotextiles. These results need to be confirmed with further studies. Also, it is possible that other estimation methods of the RSL that are not based on FWD data could give more accurate results.
References 1. Saevarsdottir, T., Erlingsson, S.: Water impact on the behaviour of flexible pavement structures in an accelerated test. Road Mater. Pavement Des. 14(2), 256–277 (2013) 2. De Souza, C.N., De Souza, B.B.: Effect of bituminous impregnation on nonwoven geotextiles tensile and permeability properties. Geotext. Geomembr. 29(2), 92–101 (2011) 3. Zornberg, J.G., Azevedo, M., Sikkema, M., Odgers, B.: Geosynthetics with enhanced lateral drainage capabilities in roadway systems. Transp. Geotech. 12, 85–100 (2017) 4. Solatiyan, E., Bueche, N., Vaillancourt, M., Carter, A.: Permeability and mechanical property measurements of reinforced asphalt overlay with paving fabrics using novel approaches. Mater. Struct. 53(8) (2020) 5. Solatiyan, E., Bueche, N., Carter, A.: A review on mechanical behavior and design considerations for reinforced-rehabilitated bituminous pavements. Constr. Build. Mater. 257 (2020) 6. Horak, E.: Benchmarking the structural condition of flexible pavements with deflection bowl parameters. J. South Afr. Inst. Civ. Eng. 50(2), 2–9 (2008) 7. Talvik, O., Aavik, A.: Use of FWD deflection basin parameters (SCI, BDI, BCI) for pavement condition assessment. Baltic J. Road Bridge Eng. 4(4), 196–202 (2009) 8. AASHTO. AASHTO Guide for Design of Pavement Structures. American Association of State Highway and Transportation Officials, Washington, D.C. (1993) 9. Rhode, G.T., Hartman, A.: Comparison of procedures to determine structural number from FWD deflections. In: 18th Transport Research Conference Proceedings, pp. 99–105. Christchurch, New-Zealand (1996)
Effects of Adding a Geotextile into a Pavement Structure …
645
10. Rhode, G.T.: Determining pavement structural number from FWD testing. Transport Research Record No 1448, 61–68 (1994) 11. Bilodeau, J.P., Doré, G., Perier, L.: Falling weight deflectometer analysis of flexible pavement structure built with geotextile drainage layers. Revue canadienne de génie civil 41(6), 540–549 (2014)
Effects of Aggregate Shape Parameters and Gradation on High-Modulus Asphalt Mix Performance Taher Baghaee Moghaddam and Hassan Baaj
Abstract Studying the performance of an asphalt concrete mix generally focuses on parameters such as the volumetrics of an asphalt mix, asphalt binder type, and content. However, the aggregate source and morphology can also significantly contribute to the performance of asphalt mixes. This paper investigates the influence of aggregate shape parameters (roundness, concavity, shape factor, and elongation) and gradation on the performance of low air-void high-modulus asphalt mix or Enrobé à Module Élevé (EME). Two high modulus mixes based on nominal maximum aggregate sizes of 12.5 and 19 mm were developed using two sources of mineral aggregates. The shape parameters of aggregate fractions were determined using an aggregate image analyzer. Thermo-mechanical tests were performed to evaluate performance of the developed mixes. The compactability of EME mixes was assessed using the Superpave Gyratory Compactor (SGC). The rheological properties of the mixes were evaluated by conducting complex modulus test at a range of temperatures (−10 to 54 °C) and loading frequencies (0.1–25 Hz). Furthermore, Hamburg wheel tracking and four-point bending beam fatigue tests were conducted to assess the rutting and fatigue performance of EME mixes, respectively. The aggregate image results indicated a considerable difference in roundness and concavity parameters of the aggregates. It was also concluded that the difference varies by changing the aggregate source and size. In addition, the performance test results showed that changes in the aggregate shape parameters along with the gradation can affect the performance of the mixes. Keywords Asphalt mix · Mineral aggregates · Aggregate shapes · Performance
1 Introduction Hot Mix Asphalt (HMA) is a composite material consisting of aggregate particles and asphalt cement or bitumen. Aggregate particles form more than 90% of the asphalt mix by weight and are considered as key parameters in HMA mix design. Aside from T. B. Moghaddam · H. Baaj (B) Centre for Pavement and Transportation Technology (CPATT), Department of Civil and Environmental Engineering, University of Waterloo, Waterloo, ON, Canada e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_82
647
648
T. B. Moghaddam and H. Baaj
the aggregate size and gradation; texture, morphological and shape characteristics of the particles would affect the mix performance. In HMA, asphalt cement and fine aggregates form mastic which coat the coarser aggregates and assure durability of the mix. Research studies have been performed to evaluate the effects of aggregate morphology on HMA characteristics. Wang et al., evaluated the effects of the morphological characteristics (shape, angularity and texture) of fine and coarse aggregates on high-temperature viscoelastic properties of asphalt mixes [1]. In other studies, the effects of shape parameters were investigated on the stability, flow and strength of asphalt mixes [2–4]. The role of aggregate morphological parameters (flatness and angularity) were also investigated on the performance of stone mastic asphalt (SMA) mixes [5]. The most common methods to evaluate aggregate morphology and shape parameters are based on indirect measures which sometimes contribute to inconsistency. In this regard, visual inspections of aggregates are probably the most effective approach. Nowadays, by evolving technology, image processing techniques are becoming the most accurate and effective ways in measuring aggregate morphological characteristics [6–8]. High-Modulus Asphalt or Enrobé à Module Élevé- (EME) is a type of asphalt concrete that represents high modulus/stiffness, high durability, superior rutting performance and good fatigue resistance. This type of mix was developed in France in the 1980s. EME is a very good option to be used in lower and upper binder courses in the pavement structure which are subject to the highest levels of tensile and compressive stresses [9]. EME mix is very compactable and has a dense structure, and its properties could be affected by the source of used aggregates. Therefore, the main objective of this study is investigating the effect of aggregate gradations and morphological parameters on the compaction ability and thermomechanical characteristics of EME mixes.
2 Materials and Methods In this study, EME 12.5 and EME 19 were fabricated using two different sources of aggregate materials (MRT and BL). The gradations of EME mixes are shown in Fig. 1. In addition, polymer modified PG 82-28 asphalt cement was used in this research.
2.1 Aggregate Shape Parameters Determination Using Image Analysis In order to determine aggregate shape parameters, an aggregate scanning device (OCCHIO belt aggregate image analyzer) was used. The aggregates with different
Effects of Aggregate Shape Parameters and Gradation …
649
sizes were prepared and placed on a vibration plate and then freely fell on a transparent conveyor belt where the aggregate shapes were scanned. The scanned aggregates images were then analyzed using an image processing software in order to determine the aggregate morphological parameters. The measured shape parameters were roundness, concavity, shape factor and elongation. They were determined using the following equations [10, 11]: Rd =
100 ×
(π × D0 )2 2 (P )
(1)
ACvx − A ACvx
(2)
C = 100 ×
4.π.A P2 Db E L = 100 × 1 − Da F=
(3) (4)
Passing (%)
In these equations, Rd is the mean roundness (%); C is the mean concavity (%); F is the mean shape factor; EL is mean elongation (%); Da , Db are the two-dimension diameters of an ellipse having the same area as the particle; A is the area of the particle; P is the perimeter of the particle; ACvx is the convex hull area; D0 is the diameter of equivalent inner circle at each peripheral point. 100 90 80 70 60 50 40 30 20 10 0 0.01
EME 12.5 - MRT EME 19 - MRT EME 12.5 - BL EME 19 - BL
0.1
1
Sieve size (mm) Fig. 1 Aggregate particle size gradation
10
100
650
T. B. Moghaddam and H. Baaj
2.2 Sample Preparation and Test Methods To fabricate the EME mixes, the aggregate particles and asphalt cement (bitumen) were heated in an oven to reach 165 °C (mixing temperature) and mixed for about 90 s. Thereafter, the loose mixtures were compacted at their compaction temperatures (155 °C). Superpave Gyratory Compactor (SGC) and Asphalt Vibratory Compactor (AVC) were used to fabricate cylindrical and beam shaped specimens. Dynamic modulus test was conducted according to AASHTO T342 standard procedure using Material Testing System (MTS-810) equipment. The SGC compacted samples were cored and cut to produce Ø100 × H150 mm cylindrical specimens. The test was performed at six different loading frequencies from 0.1 to 25 Hz and five different temperatures (−10 to 54 °C). The Hamburg-Wheel Track Tester (HWTT) was used to investigate the rutting performance of EME mixes according to AASHTO T324. In this test, the specimens were submerged and conditioned under water (wet condition) at 50 °C. The four-point bending beam fatigue test was performed according to AASHTO T 321. 380 × 63 × 50 mm beam specimens were prepared in the lab for the fatigue testing. The test was conducted at intermediate temperature of 21 °C using 10 Hz loading frequency. The test was conducted at different strain levels under controlled strain (displacement) mode.
3 Results and Discussions The shape parameters of the used aggregates were determined, and the results are presented in Fig. 2. As can be seen from the results, BL aggregates have high roundness values compared to MRT. It is also clear from the results that the difference between the aggregate concavity is more substantial for finer aggregates compared to the coarser ones. In the case of shape factor, MRT aggregates have higher values than BL materials for the most size ranges. In addition, the elongation values of both aggregate materials are similar and vary between 30 and 45%. The compactability test results are provided in Fig. 3a. As can be observed from this figure, the mixes are more compactable than conventional Superpave asphalt mixes where the air-void in the mix are less than 4% at the Design Number of Gyrations (Ndes = 125) for the high-volume highways. Additionally, it was observed that the amount of fine materials (passing sieve #4) had an inverse correlation with the compaction ability of the mixes. That is to say, the mixes with less amount of fine aggregates had higher compactability. Figure 3b illustrates the |E*| master-curves of the mixes. According to the dynamic modulus test results, the mixes fabricated with MRT aggregates have higher modulus values specifically at higher temperatures (lower frequencies). That might be due to the lower roundness value of this aggregate source. Rutting test results (Fig. 3c) showed that the rut depths of all mix types were very low (~0.6 mm) after 20,000 of the wheel pass, and changes in the aggregate source
Effects of Aggregate Shape Parameters and Gradation …
(a)
20 MRT
70
BL
60
Percent (%)
Percent (%)
(b)
Roundness
80
50 40 30 20
MRT
18 16
BL
14 12 10 8 6 2 0
0.075 - 0.3 0.3 - 0.6 0.6 - 1.18 1.18 - 2.36 2.36 - 4.75 4.75 - 9.5 9.5
S6, and S4 < S3 < S6 < S5, respectively. Shortterm aging results illustrated fumed silica NPs had little effect on the phase angle and complex modulus on bitumen samples. The results showed bitumen sample with silica NPs could be use as UV resistance due to it has better performance
Fig. 1 a complex modulus, phase angle, and b rutting resistance of modified bitumen samples with NPs before and after aging
882
G. Cheraghian et al.
Fig. 2 Index of viscosity aging of control base and modified bitumen
in UV aging. This means fumed silica nanoparticles had a greater effect on UV aging 12 days. In other words, silica NPs with increasing UV light absorption had better shielding condition than short-term aging. In Fig. 1b showed the permanent deformation resistance or rutting factor in bitumen samples between 30 and 70 °C before and after aging. As highlighted can be seen in Fig. 1b, after aging process, rutting results show that nanoparticles improved bitumen against rutting resistance deformation. In other words, a certain concentration of fumed silica nanoparticles have a good compatibility against destruction of UV radiation and thermal aging. Validation of FTIR results with rheological aging properties such as the index of viscosity aging (IVA) and index of complex modulus aging are usual for different modified bitumen [18]. The index of viscosity aging (IVA) at 50 °C is applied to investigate the aging properties: IVA = (viscosity of bitumen after − viscosity of bitumen before) /viscosity of bitumen before After short term and UV aging processes, the index of viscosity aging in samples modified by NPs are smaller than control samples which confirm bitumen aging resistance increase with fumed silica NPs (Fig. 2).
3.2 FT-IR Analysis The results of difference of carbonyl and sulfoxide are reported in Fig. 3. The carbonyl index had increased at UV aged samples more than bitumen control. Furthermore, carbonyl index (IC=O) and sulfoxide index (IS=O) increased with expanding aging times. The carbonyl and sulfoxide index decreased in NFS additive samples. The results shown NPs have suitable performance due to NFS have great ability as efficient UV-shielding coatings, in other words, nanoparticles with nano-diameter have an excellent capability to absorb UV light [19]; In addition, silica NP is a suitable
Experimental Investigation on Ultraviolet Aging …
883
Fig. 3 Structural index of bitumen before and after aging
material in reflect UV light. CR = (Index of bitumen after − index of bitumen before) /index of bitumen before Figure 4 shows SEM images of the dispersion of fumed silica NPs in bitumen samples. Nano fumed silica with the large surface area could coat and support a wide area. Unique structure of nano fumed silica absorbs and reflects UV light, and as a shield, it prevents destruction of the upper structure and as well as with trapping bitumen volatile compounds, decelerates the leaving process of them from bitumen. In Fig. 4. Nanoparticles dispersed uniformly between the bitumen sample.
Fig. 4 FESEM images of modified bitumen with silica nanoparticles at different magnifications
884
G. Cheraghian et al.
4 Conclusions In this work, characteristics of modified bitumen samples by fumed silica NPs were investigated. The following experimental results were obtained: (1) adding silica nanoparticles to bitumen considerably improves bitumen aging resistance due to a reduction in the index of viscosity aging, as well as rutting results show that nanoparticles improved bitumen against rutting resistance deformation. (2) The chemical indexes and FTIR of bitumen samples have been shown NPs affect the carbonyl and sulfoxide index. Silica NPs have great ability as efficient UV-shielding coatings in bitumen. (3) Compared to the base control sample, the modified bitumen sample is more stability against UV radiation and thermal aging.
References 1. Zhu, C., Zhang, H., Shi, C., Li, S.: Effect of nano-zinc oxide and organic expanded vermiculite on rheological properties of different bitumens before and after aging. Constr. Build. Mater. Production, 30–37 (2017) 2. Cheraghian, G., Falchetto, A.C., You, Z., Chen, S., Kim, Y.S., Westerhoff, J., Moon, K.H., Wistuba, M.P.: Warm mix asphalt technology: An up to date review. J. Cleaner Production, 268, 122–128 (2020) 3. ASTM, D. 2872 Standard Test Method for Effect of Heat and Air on a Moving Film of Asphalt (Rolling Thin-Film Oven Test). Annual Book of ASTM Standards, 4 (2004) 4. ASTM D6521–13. (2013). Standard practice for accelerated aging of asphalt binder using a pressurized aging vessel (PAV). USA: Annual Book of ASTM Standards 5. Hu, J., Wu, S., Liu, Q., Hernández, M.G., Zeng, W., Nie, S., Wan, J., Zhang, D., Li, Y.: The effect of ultraviolet radiation on bitumen aging depth. Materials 11(5), 747 (2018) 6. Mousavi, S.M., Hashemi, S.A., Babapoor, A., Medi, B.: Enhancement of rheological and mechanical properties of bitumen by polythiophene doped with nano Fe3 O4 . JOM 71(2), 531–540 (2019) 7. Cheraghian, G., Wistuba, M.P.: Ultraviolet aging study on bitumen modified by a composite of clay and fumed silica nanoparticles. Sci Rep 10(1), 1–17 (2020) 8. Kleizien˙e, R., Paliukait˙e, M., Vaitkus, A.: Effect of nano SiO2 , TiO2 and ZnO modification to rheological properties of neat and polymer modified bitumen. In: International Symposium on Asphalt Pavement & Environment, pp. 325–336. Springer, Cham (2019) 9. Li, J., Yu, J., Wu, S., Pang, L., Amirkhanian, S., Zhao, M.: Effect of inorganic ultraviolet resistance nanomaterials on the physical and rheological properties of bitumen. Constr. Build. Mater. 152, 832–838 (2017) 10. Karahancer, S.: Investigating the performance of cuprous oxide nano particle modified asphalt binder and hot mix asphalt. Constr. Build. Mater. 212, 698–706 (2019) 11. Jin, J., Chen, B., Liu, L., Liu, R., Qian, G., Wei, H., Zheng, J.: A study on modified bitumen with metal doped nano-TiO2 pillared montmorillonite. Materials 12(12), 1910 (2019) 12. Wu, S., Zhao, Z., Li, Y., Pang, L., Amirkhanian, S., Riara, M.: Evaluation of aging resistance of graphene oxide modified asphalt. Appl. Sci. 7(7), 702 (2017) 13. Li, R., Xiao, F., Amirkhanian, S., You, Z., Huang, J.: Developments of nano materials and technologies on asphalt materials—a review. Constr. Build. Mater. 143, 633–648 (2017) 14. Zhang, H., Zhu, C., Yu, J., Tan, B., Shi, C.: Effect of nano-zinc oxide on ultraviolet aging properties of bitumen with 60/80 penetration grade. Mater. Struct. 48(10), 3249–3257 (2015)
Experimental Investigation on Ultraviolet Aging …
885
15. Zahid, M., Heredia-Guerrero, J.A., Athanassiou, A., Bayer, I.S.: Robust water repellent treatment for woven cotton fabrics with eco-friendly polymers. Chem. Eng. J. 319, 321–332 (2017) 16. Cheraghian, G.: Evaluation of clay and fumed silica nanoparticles on adsorption of surfactant polymer during enhanced oil recovery. J. Jpn. Petrol. Inst. 60(2), 85–94 (2017) 17. Cheraghian, G., Wistuba, M. P.: Effect of fumed silica nanoparticles on ultraviolet aging resistance of bitumen. Nanomaterials, 11, 454 (2021) 18. Feng, Z.G., Wang, S.J., Bian, H.J., Guo, Q.L., Li, X.J.: FTIR and rheology analysis of aging on different ultraviolet absorber modified bitumens. Constr. Build. Mater. 115, 48–53 (2016) 19. Huang, X.J., Zeng, X.F., Wang, J.X., Zhang, L.L., Chen, J.F.: Synthesis of monodispersed ZnO@ SiO2 nanoparticles for anti-UV aging application in highly transparent polymer-based nanocomposites. J. Mater. Sci. 54(11), 8581–8590 (2019)
Experimental Study on the Grading Distribution of Cold Recycled Asphalt Mixtures Produced with Bitumen Emulsion and High Strength Cement Chiara Mignini , Fabrizio Cardone , and Andrea Graziani
Abstract Cold recycled asphalt mixtures (CRAM) can lead to real economic advantages only if their performance is comparable to that of traditional hot mix asphalt. This experimental study aims to investigate the effect of the grading distribution on the volumetric and mechanical properties of CRAM produced with high reclaimed asphalt dosage (>70%), bitumen emulsion and high strength cement. Four grading distributions were analysed: two continuously graded curve (derived from the FullerThompson maximum density curve) with nominal maximum aggregate size (NMAS) 16 and 10 mm. The third was obtained increasing the filler dosage, and the fourth grading distribution was a gap-graded type, with NMAS 16 mm. Workability and compactability of the fresh CRAM were evaluated using the gyratory compactor. CRAM long-term complex modulus was measured using cyclic compression tests. Results showed that the gap-graded and the high-filler grading distributions allowed obtaining the lower voids in the mixture. Having fixed the number of voids in the mixture, the gap-graded CRAM showed a higher stiffness modulus and lower phase angle if compared to CRAM produced with the continuously graded curve. Keywords Cold recycling · Reclaimed asphalt · Grading distribution · Bitumen emulsion · Complex modulus
1 Introduction Cold recycled asphalt mixtures (CRAM) are an effective technology in terms of sustainability and cost-effectiveness, but they typically have lower performances compared to traditional hot mix asphalt [1]. Lower mechanical properties can be related to a higher void content in CRAM. Moreover, they need a curing time before reaching their final properties [2, 3]. Study of CRAM workability and compactability is essential because of the effect of the voids on CRAM final properties [4, 5]. The proper choice of the intergranular water content is crucial to have a good balance C. Mignini (B) · F. Cardone · A. Graziani Dipartimento di Ingegneria Civile Edile e Architettura, Università Politecnica delle Marche, Ancona, Italy e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_113
887
888
C. Mignini et al.
between mixability (homogeneous mixture) and compactability (reduction of voids) of CRAM. An excess of water is discouraged as well, to avoid leakage of liquids and filler-sized particles during the compaction (with the consequent alteration of the CRAM composition) [2]. Generally, continuously graded curves are adopted in cold mixtures production. To promote the use of CRAM in upper pavement layers, their mechanical properties should be improved. Recent studies attempted evaluating the effect of different grading curves on the mechanical properties of CRAM [6, 7]. Other researches proposed the use of not conventional types of cement for improving the response of cold mixtures. Supplementary blended cementitious fillers increase the mechanical performance of cold mixtures [8], whereas rapid hardening cement has been found to improve the early strength of cold mixtures [9]. The goal of the present laboratory investigation is to evaluate the effect of grading distribution on CRAM suitable for upper pavement layers. Voids of CRAM both at fresh and cured state were evaluated through a volumetric analysis. Mechanical properties evaluation was carried out by measuring complex modulus.
2 Experimental Plan The CRAM were produced in the laboratory using reclaimed asphalt (RA) aggregate, natural limestone sand, limestone filler, bitumen emulsion, cement and additional water. Three gradations of the same RA aggregate source were used. The original RA aggregate, coming from the milling of the bituminous layer of end-of-life pavements, was RA0/16. Further RA fractions, i.e. RA0/10 and RA2/16, were derived by sieving operations. Two commercial cationic slow-setting bitumen emulsions with 60% of residual bitumen were used: a traditional and a modified emulsion (respectively C60B10 and C60BP10, according to EN 13808). Two high strength cement were employed as co-binders. The first was a sulfo-aluminous cement (CSA) with a compressive strength of 80 MPa after 28 days. The second was a Portland slag cement, type II/B-S (PSC), strength class 52.5N (EN 197-1). Four grading distributions, labelled FT-16, FT-10, Fine-16, GG-16, were considered (Table 1 and Fig. 1). The reference gradation (FT-16) was the Fuller-Thompson maximum density curve with exponent 0.45 and nominal maximum aggregate size (NMAS) of 16 mm. The second (FT-10) was obtained reducing the NMAS of the reference distribution to 10 mm, and the third (Fine-16) was obtained adding 5% Table 1 Composition of aggregate blends Grading
Type
NMAS (mm)
Composition
FT-16
Max density (exp. 0.45)
16
80% RA0/16, 17% sand, 3% filler
FT-10
Max density (exp. 0.45)
10
75% RA0/10, 17% sand, 8% filler
Fine-16
Continuous, high filler
16
80% RA0/16, 20% filler
GG-16
Gap graded
16
70% RA2/16, 20% sand, 10% filler
Experimental Study on the Grading Distribution of Cold … 100
Passing by volume (%)
Fig. 1. Grading distributions
75
889
FT-16 FT-10 Fine-16 GG-16
50 25 0 0.01
0.1
1
Sieve size (mm)
10
100
of filler to the reference distribution. The fourth grading distribution (GG-16) was a gap-graded type, with NMAS 16 mm. Emulsion dosage (with respect aggregate mass) was 5.0%, whereas cement dosage was 2.5%. Mixing was carried out in the laboratory following a procedure developed in previous studies [2]. Two specimens for each mixture were obtained with the gyratory compactor using 150 mm moulds. The experimental plan was developed in two phases (Fig. 2). During phase A, specimens were produced fixing the energy of compaction (number of gyrations equal to 100) to study the influence of water content on the compactability of the mixtures. Each CRAM was mixed considering 3 different intergranular water content wF . Compactability was assessed using a volumetric analysis [4] in terms of voids in the mixture at the fresh state V m,fresh (i.e. the percentage of the total mixture volume occupied by air and intergranular water): Vm,fresh = Vv,A,fresh + VW,F /V
(1)
where V is the volume of the specimen, V v,A,fresh and V W,F are the volume of air and intergranular water, respectively. Specimens were cured at 25 °C and RH = 70% for at least 90 days. During curing part of the water evaporates and part of the water is bounded in the cement hydration process. Nevertheless, a small amount of water can remain trapped in the specimen, in the form of not-interconnected voids. Besides, cement changes its volume because of hydration. The hydrated cement (consisting of reacted cement and the non-evaporable water) has a higher volume than the anhydrous one [10]. Thus, if the amount of trapped water is negligible, the voids of the cured mixture are: Vm,cured = Vv,A,cured /V
(2)
where V v,A,cured is the volume of air obtained by summing V v,A,fresh and the volume of evaporated water. CRAM will have different volumetric properties with respect to the fresh state, specifically V m,cured < V m,fresh . The volume of voids of cured specimens V m,cured was calculated according to EN 12697–8. The volume of non-evaporable
890
C. Mignini et al.
Fig. 2. Experimental plan
water (bounded by the cement) was considered on par of the other solid components of CRAM. In Phase B the fixed height compaction mode was selected to produce specimens with V m,fresh of 9.5%. CRAM were produced with both types of cement, modified emulsion and gradings FT-16 and GG-16. The adopted intergranular water content was 3.3% and 2.9% for the two CRAM gradings, respectively. Curing was performed for 6 months at 25 °C and RH = 70%. Specimens were cored to get cylindrical specimen with 75 mm diameter. The complex modulus was measured using the testing system AMPT PRO. The axial strain was measured in the middle of the specimen with three LVDT, placed 120° apart, with a measuring base of 70 mm. To obtain an axial strain with a target amplitude of 30 microstrains, a haversine compression loading was applied. Test temperatures and frequencies ranged from 5 to 55 °C and 0.1 to 10 Hz, respectively. 20 loading cycles were applied at each frequency.
3 Results and Analysis Figure 3 displays V m,fresh as a function of the intergranular water content on the CRAM produced with modified emulsion. CRAM with traditional emulsion exhibited similar behaviour. The adoption of grading distributions different from the reference one results in different V m,fresh . Specifically, comparing the voids of the CRAM produced with the same wF (i.e. 3.0% for CSA mixtures and 2.5% mixtures for PSC) and different grading distribution, better compactability is obtained using Fine-16 (average reduction of V m,fresh on the four CRAM equal to −1.4%) and GG-16 (−0.9% on average). The influence of the decrease of the NMAS from 16 mm to 10 mm is smaller (−0.4% on average). The effect of the type of cement and emulsion is less relevant. Figure 4 displays the reduction of the voids in the mixture due to the passage from fresh to cured state. Some differences may be highlighted between mixtures with
Experimental Study on the Grading Distribution of Cold … FT-16 FT-10 Fine-16 GG-16
10.0
9.0
8.0 1.5
a)
2.5
3.5
Intergranular water (%)
11.0
Vm,fresh (%)
Vm,fresh (%)
11.0
891 FT-16 FT-10 Fine-16 GG-16
10.0
9.0
8.0 1.5
4.5
b)
2.5
3.5
4.5
Intergranular water (%)
Fig. 3 Voids in the mixture as a function of the intergranular water of CRAM with modified emulsion and cement: a CSA, b PSC
different grading distributions. Nevertheless, the effect of the grading distribution is not clear, which may be due to the variability of the test. As regards the influence of the type of binders (i.e. cement and emulsion), the use of CSA and PSC allows reducing voids on average of 1.4% and 1.7%, respectively. Therefore, it may appear that hydrated PSC has a higher volume compared to CSA. However, the difference between the reduction measured for CRAM produced with the two cements is not considered significant and may be due to the variability of the experimental data. Further studies are needed to deepen this effect. The effect of the type of emulsion is even more negligible (average reduction of 1.5% and 1.6% with traditional and modified emulsion, respectively). It can be stated that the improvement of CRAM volumetric properties must be obtained at the fresh-state through a proper mix-design and compaction. After curing, CRAM will always have lower voids compared to the fresh state. The reduction of voids could be slightly related to the type of binders. Overall, the effect of the type of binder is less significant compared to those obtained through the mix-design optimisation. Figure 5 shows the results of complex modulus measurement on mixtures with CSA and grading curves FT-16 and GG-16. CRAM with PSC exhibited similar behaviour. Experimental data, displayed in the Cole-Cole diagram (Fig. 5a) and Black diagram (Fig. 5b), form a continuous curve. Thus, the Time-Temperature Superposition Principle is considered valid. Despite the same composition and volumetric properties, CRAM with GG-16 had higher stiffness modulus both at low and high temperatures (high and low frequencies). The phase angle is lower in CRAM with GG-16 at all frequencies.
4 Conclusions This paper evaluates the effect of different grading distribution on physical and mechanical properties of CRAM produced with bitumen emulsion and high strength
892
C. Mignini et al.
Vm,fresh - Vm,cured (%)
3.0 2.5
FT-16 Fine-16
FT-10 GG-16
2.0
1.7% 1.4%
1.5 1.0 0.5 0.0
CSA-B
CSA-BP
PSC-B
PSC-BP
CRAM Fig. 4 Reduction of voids in the mixture after curing (error bars are the standard error of mean) 1.E+05
1000
500 FT-16
0
a)
0
5000
GG-16
10000
Storage modulus (MPa)
Stiffness moduus (MPa)
Loss moduus (MPa)
1500
1.E+04
1.E+03 FT-16
1.E+02
15000
b)
0
10
GG-16
20
30
Phase angle (°)
Fig. 5 Complex modulus measurement, CSA mixtures: a Cole-Cole plan, b Black diagram
cement. The Fuller-Thompson distribution does not confer better compactability to CRAM. The increase of fines or a gap-graded distribution allow the reduction of voids. At the end of curing, voids of CRAM are further reduced because of the increase in the volume of cement. Compared to mixtures produced with FullerThompson grading, CRAM produced with a gap-graded distribution showed, in the long-term, higher stiffness modulus and smaller phase angles at all temperatures. Acknowledgements The Authors express their gratitude to Italcementi –HeidelbergCement Group for funding the research and provide the cement. The Authors thanks Valli Zabban S.p.A. and Società Cooperativa Braccianti Riminese for supply bitumen emulsion and RA.
Experimental Study on the Grading Distribution of Cold …
893
References 1. Fang, X., Garcia-Hernandez, A., Lura, P.: Overview on cold cement bitumen emulsion asphalt. RILEM Tech. Lett. 1, 116–121 (2016) 2. Graziani, A., Godenzoni, C., Cardone, F., Bocci, M.: Effect of curing on the physical and mechanical properties of cold-recycled bituminous mixtures. Mater. Des. 95, 358–369 (2016) 3. Du, S.: Effect of curing conditions on properties of cement asphalt emulsion mixture. Constr. Build. Mater. 164, 84–93 (2018) 4. Grilli, A., Graziani, A., Bocci, E., Bocci, M.: Volumetric properties and influence of water content on the compactability of cold recycled mixtures. Mater. Struct. 49(10), 4349–4362 (2016) 5. Ouyang, J., Pan, B., Xu, W., Hu, L.: Effect of Water Content on volumetric and mechanical properties of cement bitumen emulsion mixture. J. Mater. Civ. Eng. 31(6), 04019085 (2019) 6. Raschia, S., Mignini, C., Graziani, A., Carter, A., Perraton, D., Vaillancourt, M.: Effect of gradation on volumetric and mechanical properties of cold recycled mixtures (CRM). Road Mater. Pavement Design 1–15 (2019) 7. Xu, O., Wang, Z., Wang, R.: Effects of aggregate gradations and binder contents on engineering properties of cement emulsified asphalt mixtures. Constr. Build. Mater. 135, 632–640 (2017) 8. Dulaimi, A., Al Nageim, H., Ruddock, F., Seton, L.: High performance cold asphalt concrete mixture for binder course using alkali-activated binary blended cementitious filler. Constr. Build. Mater. 141, 160–170 (2017) 9. Saadoon, T., Gómez-Meijide, B., Garcia, A.: Prediction of water evaporation and stability of cold asphalt mixtures containing different types of cement. Constr. Build. Mater. 186, 751–761 (2018) 10. Brouwers, H.J.H.: The work of Powers and Brownyard revisited: part 1. Cem. Concr. Res. 34(9), 1697–1716 (2004)
Experimental Tests on Diffusion Phenomenon Between Two Different Bitumens Stefano Noto, Salvatore Mangiafico, Elena Romeo, Cédric Sauzéat, Hervé Di Benedetto, and Gabriele Tebaldi
Abstract A better understanding of the interaction between old and new bitumen is considered of key importance in the production of new hot asphalt mixtures with high percentages of reclaimed asphalt pavement. This improvement could potentially avoid shorter service life of the material and failure due to reduced performance in rutting and fatigue. This study is part of an on-going research in the framework of a PhD thesis. The aim of this paper is to present the experimental results focussed on the characterisation of the diffusion phenomenon between two different bitumens. A DSR has been used to experimentally measure the changes in shear complex modulus during time at 50 °C of a double-layer specimen. The main hypothesis is that a changing in the measurements of the shear complex modulus is stemmed from a changing in the structure of the sample: this implies that a blending phenomenon between the two bitumens is going on, mainly driven by diffusion between the two materials. Keywords Diffusion modelisation · Visco-elastic behaviour · DSR tests · Double layer specimen · RAP
1 Introduction The increasing interest on the use of a high amount of reclaimed asphalt pavement (RAP) in the production of new asphalt mixtures has led the researchers to study the chemico-physical processes to maintain high-performance level: among the other the binder film thickness, the cluster phenomenon, the interaction between different binders and the concentration of different molecules at a different molecular weight. The percentage of new bitumen in mixtures with high RAP content has to be carefully evaluated to avoid rutting and cracking failure, mainly due to RAP heterogeneity [1, S. Noto (B) · S. Mangiafico · C. Sauzéat · H. Di Benedetto ENTPE, University of Lyon, Lyon, France e-mail: [email protected] S. Noto · E. Romeo · G. Tebaldi University of Parma, Parma, Italy © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_114
895
896
S. Noto et al.
2] and RAP clusters [3]. The authors in this paper [3] quantified the presence of RAP clusters in mixtures with high RAP content with a focus in the mix production phase. Several studies focused on the interaction between Virgin and RAP bitumen and the estimation of the degree of blending [4, 5], in which they developed an analytical relationship for the blending of bitumen. They verified that the linear visco-elastic behaviour depends on the content of RAP binder. Thus, the 2S2P1D visco-elastic model parameters can be found through simple linear equation as a function of the percentages of RAP in the blend. On the other hand, Nahar [6] assessed the changing in the microstructure at the interface of two different bitumen in contact using the atomic force microscopy. He verified the presence of a diffusion phenomenon in the material. Experimental works on cyclic extraction of bitumen performed on HMA with a high amount of RAP verified the lack of homogeneity in the rheological properties between the outer and the inner layers of the bitumen in the RAP [7]. Other studies focused on the estimation of diffusion by means of molecular computer simulation, modelling the molecular structure of the two bitumen [8]. From the experimental point of view, researchers [10–12] developed a DSR experiment to assess the diffusion mechanism in a double-layer specimen. Each thin layer was obtained by different materials in order to verify the interaction between the two bitumens. Diffusion phenomenon was studied from the chemical point of view, tracing different chemical species in the material. This random movement driven the gradient of concentration of the chemical species is well represented by Fick’s laws [9, 10]. A deeper comprehension of the phenomenon could help to define reliable parameters to reach a wider awareness in the use of asphalt mixtures with a high amount of RAP.
2 Objective and Scope This paper describes the preliminary experimental investigation focused on studying the diffusion phenomenon at 50 °C between two bitumen having different viscosities. A DSR test was implemented to trace the evolution of the phenomenon. It consists of a series of frequency sweeps at 50 °C using a sample composed of two layers of different bitumens. The test is based on previous studies [10–12].
3 Materials and Methods 3.1 Materials Three bitumens were used for this study. The “Bitumen 1” (B1) is a 50/70 pure bitumen with a penetration grade at 25 °C (UNI EN 1426) of 51 dmm. The “Bitumen
Experimental Tests on Diffusion Phenomenon …
897
2” (B2) is a 170/210 pure bitumen with a penetration grade at 25 °C of 171 dmm. The TR&B are 49.6 °C and 37.2 °C, respectively (UNI EN 1427). The “Bitumen 3” (B3) is a fully blended bitumen, and it was prepared in the laboratory by mixing 50% of the total weight of Bitumen 1 and 50% of “Bitumen 2” for five minutes at 150 °C. It has a penetration grade of 99 dmm, and the TR&B is 45 °C. The scope of the Bitumen 3 is to simulate the “full diffusion condition” and to have the related performance.
3.2 Methods The blending phenomenon was experimentally studied with a dynamic shear rheometer (DSR) Anton Paar MCR101. Firstly, amplitude sweep was performed to define the limit of the linear viscoelastic region for each bitumen. A strain γ of 0.5% was chosen as a reference strain to perform all the tests. Mechanical properties were monitored every thirty minutes for a total duration of 22 h at a constant temperature of 50 °C. The temperature of 50°C was selected based on the ring and ball temperatures of the two bitumen. A series of frequency sweep tests at 1, 3 and 10 Hz were performed recording fifteen measurements for each frequency. Thirty minutes of resting period in strain control was set with γ equal to zero between each frequency sweep test. A 25 mm diameter parallel plate configuration (PP25) was used. The selected testing gap was 0.5 mm to avoid any leakage of material during the experiment. DSR is equipped with a Peltier cell in the bottom plate and a Peltier hood to cover the spindle and the sample. Secondly, Bitumen 1, 2 and 3 were tested using a 0.5 thick single layer sample to assess the trend in the rheological properties as a function of time during the test. Thirdly, the tests were performed using double-layer samples. These samples were composed of two discs of 0.25 mm of thickness each. The discs were put in contact at the beginning of the test. The procedure of the preparation of double-layer samples developed by Kriz [11] was adapted. The following paragraph describes the procedure (Fig. 1). The two bitumens were heated up until they reached the workability. Thus, 0.5 g of Bitumen 1 was poured on a transparent plastic film. The same procedure was applied to make the second disc of Bitumen 2. The disc of Bitumen 1 was placed on the bottom plate of the DSR with the transparent film upward and conditioned at 50 °C for 5 min. The target thickness of the disc was reached by setting the gap distance of the two plates of the DSR. The chamber temperature was set at 40 °C, and the disc was conditioned for 5 min. The Bitumen 2 disc was attached to the top plate, and after a conditioning time of 5 min, the target gap was reached, considering the thickness of the two transparent films. Then, the temperature was set to −10 °C, in order to peel off the two transparent film from the material. After a conditioning time of 10 min at 30 °C, the two discs were separately trimmed in order to remove the excess material and to have a regular geometry of the vertical surfaces. After a
898
a)
S. Noto et al.
b)
Fig. 1 a A bitumen disc on a transparent film ready to be placed in the DSR. b DSR test set up during the preparation of the two-layer sample
conditioning time of 10 min at 50 °C, the two discs were put in contact, and the test started. This test aims to indirectly observe the evolution of blending between two different bitumens as a function of time.
4 Results and Analysis 4.1 Experimental Rheological Results The authors present the experimental results only obtained at a frequency of 1 Hz, due to the need to summarise the results. As explained in the paragraph above, the reference values were defined, performing the test on single-layer samples. The reference values for Bitumen 1, 2 and 3 are essential to distinguish the changing in the material only due to chemical factors, such as oxidation, from other phenomena (see Fig. 2). The results for double layer samples using different bitumens can be found in the same graph (indicated as “DL” in the caption). This chart shows the measurements of the norm of the shear complex modulus during the time for a total duration of 22 h. The measurements were recorded every thirty minutes. The previous graph shows that it is not immediate to make a direct comparison between the trends of the experimental results to assess the blending phenomenon.
4.2 Analysis of the Results A normalisation was applied to get the global trends of the results. The measurements at time equal to thirty minutes were chosen to normalise the results since there is marked variability in the measurements at time equal to zero. The following normalisation was applied, as shown in the Eq. (1):
Experimental Tests on Diffusion Phenomenon …
899
Fig. 2. Evolution of the norm of the shear complex modulus during for two bitumen (B1 & B2), a perfect blend of both bitumen (B3) and a double layer sample (DL)
∗ G (t)
nor m
=
|G ∗ (t)| − |G ∗ (30 min)| |G ∗ (30 min)|
(1)
The following figure shows the normalised results (see Fig. 3). A focus on the reference curves (B1, B2, B3) reveals that there is a similar trend in the evolution of the norm of the shear complex modulus. What can be clearly seen in this chart is the difference between the reference tests and the double layer tests. Analysing this chart, it can be noticed how the trend of the results obtained by the double-layer test exactly fits the trend reference curves (B1, B2 and B3) applying a coefficient of 0.5. The graph shows that there is a marked rise in the trend of shear complex modulus of the double layer sample composed of two bitumens. This implies that a changing in the rheological properties could have occurred independently by the oxidation of the bitumens. This phenomenon could potentially be explained by the presence of a blending effect, mainly due to diffusion.
5 Conclusions This paper presents the preliminary experimental results of a series of frequency sweep test as a function of time. It is focused on the characterisation of the interaction between two bitumens having different viscosities. The central hypothesis is that a changing in the measurements of the shear complex modulus is stemmed from a changing in the structure of the sample: this implies that
900
S. Noto et al.
Fig. 3. Evolution of the normalised shear complex modulus (Eq. 1) as a function of time, for two bitumen (B1 & B2), a perfect blend of both bitumen (B3) and a double layer sample (DL)
a blending phenomenon between the two bitumens is going on, mainly driven by diffusion between the two materials. Further analysis will analytically model the diffusion phenomenon to obtain a relationship between the rheological measurements and the diffusion phenomenon, typically described by tracing the concentration of the chemical species in the material. Acknowledgement We thank Mr. Paolini Massimo from the company Valli Zabban Spa for the support and for the provision of the bitumens.
References 1. Bonaquist. R.: Laboratory evaluation of hot mix asphalt (HMA) mixtures containing recycled or waste product materials using performance testing. Publication FHWA-PA-2005-006-9832(19), Pennsylvania Department of Transportation, Office of Planning and Research (2005) 2. Jiménez del Barco Carrión, A, Lo Presti, D., Airey, G.D.: Binder design of high RAP content hot and warm asphalt mixture wearing courses. Road Mater Pavement Design 16(sup1), 460–474 (2015) 3. Bressi, S., Cavalli, M.C., Partl, M.N., Tebaldi, G.: Particle clustering phenomena in hot mix asphalt mixtures with high content of reclaimed asphalt pavements. Constr. Build. Mater. 100, 200–217 (2015) 4. Mangiafico, S., Di Benedetto, H., Sauzéat, C., Olard, F., Pouget, S., Planque, L.: Influence of reclaimed asphalt pavement content on complex modulus of asphalt binder blends and corresponding mixes: experimental results and modelling. Road Mater. Pavement Design 14, 132–148 (2013)
Experimental Tests on Diffusion Phenomenon …
901
5. Mangiafico, S., Sauzéat, C., Di Benedetto, H., Pouget, S., Olard, F., Planque, L.: Complex modulus and fatigue performances of bituminous mixtures with reclaimed asphalt pavement and a recycling agent of vegetable origin. Road Mater. Pavement Des. 18, 315–330 (2016) 6. Nahar, S.N., Mohajeri, M., Schmets, A.J.M., Scarpas, A., van de Ven, M.F.C., Schitter, G.: First observation of blending-zone morphology at interface of reclaimed asphalt binder and virgin bitumen. Transport. Res. Record: J. Transport. Res. Board 2370, 1–9 (2013) 7. Liphardt, A., Radziszewski, P., Król, J.: Binder blending estimation method in hot mix asphalt with reclaimed asphalt. Procedia Eng. 111, 502–509 (2015) 8. Cong, P., Hao, H., Zhang, Y., Luo, W., Yao, D.: Investigation of diffusion of rejuvenator in aged asphalt. Int. J. Pavement Res. Technol. 9, 280–288 (2016) 9. Malkin, A.Y., Isayev, A.: I: Rheology: Concepts, Methods and Applications, 2nd edn. ChemTec Publishing, Toronto (2012) 10. Karlsson, R., Isacsson, U., Ekblad, J.: Rheological characterisation of bitumen diffusion. J. Mater. Sci. 42, 101–108 (2007) 11. Rad, F.Y., Sefidmazgi, N.R., Bahia, H.: Application of diffusion mechanism: degree of blending between fresh and recycled asphalt pavement binder in dynamic shear rheometer. Transp. Res. Rec. 2444, 71–77 (2014) 12. Kriz, P., Grant, D.L., Veloza, B.A., Gale, M.J., Blahey, A.G., Brownie, J.H., Shirts, R.D., Maccarrone, S.: Blending and diffusion of reclaimed asphalt pavement and virgin asphalt binders. Road Mater. Pavement Design 15, 78–112 (2014)
Experimental Validation of the Dual-Oxidation Routes in Bituminous Binders Georgios Pipintakos, H. Y. Vincent Ching, Uwe Mühlich, Hilde Soenen, Sabine Van Doorslaer, Peter Sjövall, Aikaterini Varveri, Christophe Vande Velde, and Xiaohu Lu
Abstract Oxidative ageing in bituminous materials is considered to be one of the most important factors for distress types in road applications. The increasing interest in oxidative ageing has highlighted the need for a thorough understanding of the oxidation mechanisms at molecular level. This paper offers some insight in the validity of the proposed hypotheses about the oxidation routes of bitumen, the fast- and the slow-rate route, reflecting on previous studies. Fourier-Transform Infrared (FTIR) and Electron Paramagnetic Resonance (EPR) spectroscopy were utilised for this verification. To elucidate the uncertain formation of sulfoxides, an additional surface investigation with Time-of-Flight Secondary Ion Mass Spectrometry (TOF-SIMS) was performed. The findings of the aforementioned techniques reveal the existence of the oxidation products reported previously and contribute to the understanding of the oxidation mechanisms. Overall, this research strengthens experimentally the hypotheses of the dual-oxidation routes of bitumen. Keywords Oxidative ageing · Bitumen · FTIR · TOF-SIMS · EPR
G. Pipintakos (B) · U. Mühlich EMIB Research Group, University of Antwerp, Antwerp, Belgium e-mail: [email protected] H. Y. V. Ching · S. Van Doorslaer Bimef Research Group, University of Antwerp, Antwerp, Belgium H. Soenen Nynas NV, Antwerp, Belgium P. Sjövall RISE Research Institutes of Sweden, Gothenburg, Sweden A. Varveri Pavement Engineering, Delft University of Technology, Delft, Netherlands C. Vande Velde ELCAT Research Group, University of Antwerp, Antwerp, Belgium X. Lu Nynas AB, Nynäshamn, Sweden © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_115
903
904
G. Pipintakos et al.
1 Introduction Hitherto, various studies have demonstrated the link between oxidative ageing and alterations in physical and chemical properties, which eventually lead to cracking and raveling [1, 2]. Although considerable research has been devoted to the rheological consequences of oxidative ageing, less attention has been paid to the understanding and evidence of the oxidation routes at molecular scale. As articulated by various researchers two major oxidation routes may exist, namely the chemically distinct “fast-spurt” and the “slow/long-term” routes [3–6]. This idea has gained considerable support since a direct link with the asphalt production stages can be established. It appears that the initial increase of viscosity during production, transportation and compaction is controlled by the fast reaction whereas the long-term ageing during service life is attributed to a slow reaction [6]. Briefly, during the fast reaction path different products are produced with examples including hydroaromatic hydroperoxides and sulfoxides. However, as hydroperoxides might be thermally unstable they can also decompose to form free radicals, which can successively initiate the slower hydrocarbon reaction [3, 4]. Among the main products of the slow oxidation route are sulfoxides and carbonyls. Of pragmatic importance is the fact that upon sulfoxide formation during the slow reaction, alcohol chemical groups are also produced which strengthens the belief that the infrared absorption band (around 1100 cm−1 ) of alcohols and sulfoxides may coincide [4, 7]. It appears that experimental validation of the above mechanisms has been confined primarily to sulfoxide and carbonyl formation. Considerable research has been devoted to the determination of these functional groups via chemical analysis such as FTIR [8–11]. So far, little attention has been paid to the proof of free radicals as well as to the alcohol formation that accompanies sulfoxides. These factors, if experimentally proven, could validate at least the fast oxidation route and establish the basis for the understanding of the slow reaction. In this study, a number of uncertainties regarding the oxidation routes are addressed. An issue of obvious concern is the validation of the source of the increased sulfoxide indicator. To explore this aspect and unravel the possible overlap of sulfoxide and alcohol in an infrared spectrum, TOF-SIMS was utilised. The organic radical formation is finally investigated with EPR spectroscopy, underutilised for bitumen applications and appropriate for the identification of the free radicals. Links between the results of the three techniques under predefined oxidation time and temperature will finally confirm certain hypotheses of the oxidation routes.
Experimental Validation of the Dual-Oxidation …
905
Table 1 Properties of the binders Material Bitumen
Property
Binder
Test Method
A
B
Penetration 25 °C (0.1 mm)
16
189
EN1426
Softening point ( o C)
61.1
37.5
EN1427
Viscosity 135 °C (mm2 /s)
1285
203
EN12595
2 Materials and Methods 2.1 Materials and Ageing Treatment For the experimental part two bituminous binders were used as specified in Table 1, namely a hard binder A and a soft one, binder B. For a thorough investigation of the oxidation evolution of the two binders, modified thin film oven ageing (M-TFO) of approximately 1 mm thick layer was preferred for the FTIR and the TOF-SIMS measurements. That was not the case for the EPR measurements where binder films of thickness 3–5 mm were aged directly in polypropylene tubes in a ventilated oven. The temperature chosen (50 °C) was considered realistic in the framework of ageing simulations in order to accelerate the oxidation process and simultaneously capture and differentiate between the main oxidation routes. The ageing time intervals (0–2–5 and 8 days) were selected according to preliminary infrared findings for a prolonged ageing time (Fig. 1 [left]) using the same ageing treatment.
2.2 Spectroscopic Techniques FTIR The FTIR analysis was performed with a Thermo Scientific Nicolet iS10 FTIR spectrometer equipped with an Attenuated Total Reflectance (ATR) fixture and a Smart Orbit Sampling Accessory. At least three replicas were measured per ageing time interval in order to ensure repeatability, while the collected spectrum ranged from 400 cm−1 to 4000 cm−1 with a resolution of 4 cm−1 . EPR Spectra of aged and unaged bitumen samples were recorded with a Bruker Elexsys E680 spectrometer fitted with an ER 4102ST TE102 mode resonator at ~9.75 GHz (X-band). Customised 2 mL polypropylene Eppendorf tubes were used as sample holders ensuring that the total material quantity did not overfill the cavity. Preliminary power saturation measurements were performed to select the appropriate power level of 0.5 mW in order to avoid saturation effects. Two replicas were measured per ageing interval under constant instrumental settings: center field = 341 mT, sweep width
906
G. Pipintakos et al.
= 20 mT, number of sweeps = 2, resolution = 2048 points, modulation amplitude = 0.1 mT and modulation frequency = 100 kHz. The signal intensity was finally processed based on the peak-to-trough distance and normalised by the sample mass. TOF-SIMS TOF-SIMS analysis was conducted in a TOF-SIMSIV instrument (IONTOF GmbH, Germany) using 25 keV Bi3+ primary ions and low-energy electron flooding for charge compensation. The bitumen samples were deposited on silicon wafer substrates and subsequently allowed to cool according to a special protocol [12]. The analysis was carried out at a sample temperature of −80 °C over analysis areas of 200 µm × 200 µm. Positive and negative ion spectra were acquired at three locations of each sample, with the instrument optimised for high mass resolution.
3 Results and Discussion Semi-quantitative methods for analyzing the FTIR spectrum can be useful for identifying and characterising the evolution of specific oxidation products such as sulfoxides [8–11]. Consistent with previous studies, this study demonstrated an initial rapid increase of sulfoxides followed by a constant rate formation, (Fig. 1 [right]), which can be attributed to the fast and slow oxidation route respectively [3, 4]. As shown in Fig. 1 [left], binder A has almost completed the fast reaction path at about 5 days, whereas binder B reached this transition point at about 2 days. Additionally, the sulfoxide intensity, analyzed with a widely accepted quantitative method [8], documented that binder A suffered from a harsher oxidation effect, under the predefined ageing conditions, compared to binder B. From the previous analysis, it becomes apparent that at 8 days of controlled ageing the slow oxidation route has been initiated for both binders. For a fair comparison of the two binders the formation of free radicals, reported simultaneously with the
Fig. 1. Evolution of the sulfoxide intensity via FTIR [left] and indicative reaction phases [right]
Experimental Validation of the Dual-Oxidation …
907
Fig. 2. Example of X-band EPR spectra with the insets showing expansions of selected regions of organic radicals (top) and vanadyl ions (bottom) [left] and evolution of these signals [right]
sulfoxide during the spurt, was investigated in this time-frame. Fig. 2 [left] shows the time dependent evolution of the EPR spectrum of a typical binder sample and signals from organic radicals (RO· , OH· ) and a vanadyl species (VO2+ ). Comparing regions where there was only a VO2+ contribution (around g = 1.9536) with the corresponding organic radical regions (around g = 2.0016) reveals clear trends for the radical evolution (insets) [13]. What is striking about the data in Fig. 2 [right] is that the intensity of the organic radicals is constantly increasing for both binders, within the time-frame of the fast route, whereas VO2+ species appear to remain steady. The successive increments of the organic radicals, as oxidation evolved, imply that indeed the expected radical formation is evident during the fast-rate route. A comparison of the two binders indicates that the relative increase of organic radical intensity of binder A was almost doubled in comparison with binder B. On the other hand, the vanadyl signal intensities of the two binders remained unchanged, something which could be further associated with the vanadium content and used as a marker for the origin of crude oil. Having supported experimentally to some extent certain hypotheses for the mechanisms of the fast oxidation route articulated previously, a TOF-SIMS analysis for the changes of the chemical compounds upon ageing was considered crucial. Being motivated by the negligible change of the infrared spectrum during oxidation in the area 3000–3500 cm−1 related with various oxygen-containing compounds and the rise of the peak around 1100 cm−1 assigned to alcohol and sulfoxide appearance, a further exploration was performed based on previous studies [12]. It is assumed that at 8 days under the given ageing conditions both binders are in the slow reaction path. Thus, according to the underlying mechanisms sulfoxides and multifunctional alcohols as well as carbonyls are produced. The FTIR spectrum confirms partially the former and validates the latter. The results of the TOF-SIMS surface analysis unravel the riddle for the oxidation products of the fast
908
G. Pipintakos et al.
Fig. 3. Intensity of SOx -containing [left] and HOx -containing compounds via TOF-SIMS [right]
oxidation route. The findings were analyzed based on normalised signal intensity and categorised in SOx -containing, Fig. 3 [left], and HOx -containing fragments Fig. 3 [right]. It can be seen from the intensity of the fragments with the generic formula RSOx that indeed the formation of sulfoxide-related compounds is evident also with this technique. When it comes to the HOx -containing fragments, the preliminary observations with the infrared spectrum along with the rise of these compounds, indicates eventually to be alcohol-related. Moreover, the oxidation effect of binder A is significantly stronger than binder B, which is in agreement with the FTIR results. It is also worth mentioning that the amount of nitro-compound fragments was increased during oxidation, in contrast to the organic ions without heteroatoms and the ones containing S and N without oxygen which remained unchanged. In the future, this exploratory study will be exploited for the quantification of different fragments formed during the ageing process.
4 Conclusions In this investigation the mechanisms of the dual-oxidation routes of bitumen were validated experimentally. The results from FTIR supported the existence of two oxidation paths, a fast- and a slow-rate and the rapid sulfoxide formation during the fast path. EPR measurements during the fast oxidation route confirmed that organic radicals are produced and can be associated with the initiation of the slow oxidation route. This study established also a quantitative framework for detecting sulfoxide and alcohol compounds in the slow reaction path. Taken together, the findings of the FTIR and TOF-SIMS spectrometry indicated that sulfoxides, carbonyls and other oxygen-containing compounds, with high probability to be alcohols, are formed during the slow oxidation route. Overall, the insights gained from this study may be particularly interesting for improving our understanding for the underlying oxidation mechanisms which subsequently can be incorporated in the bitumen characterisation by means of the intensity of the ageing products.
Experimental Validation of the Dual-Oxidation …
909
Acknowledgments The authors gratefully acknowledge support from Nynas AB, and the E.U. for HYVC’s H2020-MSCA-IF (grant number 792946, iSPY).
References 1. Gómez, W.D.F.: Quintana, H.R., Lizcano, F.R.: A review of asphalt and asphalt mixture aging. Ing. e Investig (2013) 2. Soenen, H., Lu, X, Laukkanen, O.-V.: Oxidation of bitumen: molecular characterization and influence on rheological properties. Rheol. Acta 55(4), 315–326 (2016) 3. Petersen, J., Harnsberger, P.: Asphalt aging: dual oxidation mechanism and its interrelationships with asphalt composition and oxidative age hardening. Transp. Res. Rec. J. Transp. Res. Board 1638, 47–55 (1998) 4. Petersen, J.C.: A review of the fundamentals of asphalt oxidation (E-C140). Transp. Res. Rec. J. Transp. Res. Board (2009) 5. Herrington, P., James, B., Henning, T.F.P.: Validation of a bitumen oxidation rate model. Transp. Res. Rec. J. Transp. Res. Board 2632, 110–118 (2017) 6. Das, P.K., Balieu, R., Kringos, N., Birgisson, B.: On the oxidative ageing mechanism and its effect on asphalt mixtures morphology. Mater. Struct. 48(10), 3113–3127 (2014) 7. Bruice, P.Y.: Organic Chemistry, 8th edn. Pearson Education (2016) 8. Lamontagne, J., Dumas, P., Mouillet, V., Kister, J.: Comparison by Fourier transform infrared ( FTIR ) spectroscopy of different ageing techniques. Appl. Road Bitumens 80, 483–488 (2001) 9. Tarsi, G., Varveri, A., Lantieri, C., Scarpas, A., Sangiorgi, C.: Effects of different aging methods on chemical and rheological properties of bitumen. J. Mater. Civ. Eng. 30(3) (2018) 10. Tauste, R.: Understanding the bitumen ageing phenomenon : a review. Constr. Build. Mater. 192, 593–609 (2018) 11. Green, J., Yu, S., Pearson, C., Reynolds, J.: Analysis of sulfur compound types in asphalt. Energy Fuels 7, 119–126 (1993) 12. Lu, X., Sjövall, P., Soenen, H.: Structural and chemical analysis of bitumen using time of-flight secondary ion mass spectrometry (TOF-SIMS). Fuel 199, 206–218 (2017) 13. Espinosa, M., Campero, P.A., Salcedo, R.: Electron spin resonance and electronic structure of vanadyl−porphyrin in heavy crude oils. Inorg. Chem. 40 (2001)
Extending the Use of RAP in Airport Asphalt Resurfacing Greg White and Ali Jamshidi
Abstract Traditionally, airports in Australia have resisted the incorporation of recycled asphalt pavement (RAP) in resurfacing works. However, in 2018 an airport allowed RAP from the temporary construction ramps and the existing surface millings to be incorporated. These RAP materials were considered to be low-risk and mediumrisk sources. Statistical analysis of the mixture design, RAP production testing and asphalt production compliance testing were undertaken to determine the potential impact of extending RAP use to include these low- to medium-risk sources in airport asphalt surfacing. It was found that the minor changes associated with the incorporation of RAP were unlikely to be detrimental. However, it was concluded that the mixture design must characterise and reflect the RAP sources, and the asphalt production must monitor the RAP, similar to other raw materials such as virgin aggregate and bituminous binder. Future work should consider extending these findings to other airport surfacing projects in Australia, allowing specification changes in the future. Keywords Recycled asphalt · Airport · Aircraft · Pavement · Surface
1 Introduction Airport asphalt in Australia and many other countries is designed based on the Marshall method [1] with samples compacted by 75 blows of a Marshall hammer. Runway asphalt is generally 14 mm maximum nominal size and is typically constructed in 50–80 mm thick layers. Airport asphalt is usually densely graded with a high bitumen content and hydrated lime is often added as an anti-stripping agent. Although common in road pavements and the airports of some countries, Australian airports have generally resisted the use of RAP, particularly in the surface layer of flexible runway pavements. This resistance reflects concerns regarding G. White (B) · A. Jamshidi University of the Sunshine Coast, Sunshine Coast, QLD, Australia e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_116
911
912
G. White and A. Jamshidi
the potential impact on surface performance, generally indicated by resistance to cracking, resistance to deformation and resistance to moisture damage [2]. Furthermore, surface durability and the risk of coarse aggregate ravelling out of the surface and becoming a hazard to fragile jet engines is always a concern to airports [1]. In 2017 Australia developed a performance-based specification for dense graded airport asphalt [3] which optionally included RAP. Initial projects used low-risk RAP from the temporary construction ramps, but this was only available in low quantities, around 4–6% of the asphalt volume [4]. To extend the use of RAP in airport asphalt resurfacing to medium-risk RAP sources, another Australia airport included RAP from both the temporary ramps and the material textured from the twelve year old existing surface. This research analyses the mixture design, RAP production and asphalt production testing to determine whether the medium-risk RAP had any detrimental effect on the asphalt produced as part of the runway resurfacing work. Trends in the production results were statistically analysed and recommendations include the ongoing use of modest volumes of low- to medium-risk RAP in airport asphalt production.
2 Background In many countries, including Australia, airports generally only have one runway and airport operations can not interrupted to affect runway resurfacing works. Consequently, most airport resurfacing projects are performed in short night-time work periods, with temporary ramps constructed to provide a transition between the new and the previous surface levels, allowing aircraft operations to continue during the day. Each night, the previous ramp is removed. Because most runway resurfacings are overlays, the height of the temporary ramps is generally 40–60 mm and the ramps are typically 6–10 m long. The ramps are a potential source of RAP, which was traditionally disposed at landfill or stockpiled on site for alternate uses. The other common source of RAP during airport resurfacing is the millings from the pre-texturing of the existing surface. Because most asphalt runways in Australia are grooved, 6–10 mm of the existing surface is typically removed by cold planning machines. The millings are usually older asphalt, but are normally uniform and are often comprised of asphalt mixtures that were designed to the same airport-specific volumetric composition. As stated above, these different sources of potential RAP for airport resurfacing work are categorized by their level of variability and reliability: • Low-risk. Material from the temporary ramps that was produced to the same mixture design, using the same raw materials, just a few days or weeks prior. • Medium-risk. Millings from the generally consistent existing surface, likely to be a uniform airport asphalt mixture produced from the same or similar materials, albeit typically 10–12 years old.
Extending the Use of RAP in Airport Asphalt Resurfacing
913
• High-risk. Unknown RAP from a central stockpile, including diverse sources of different mixtures of different ages and with different raw materials. Many airports are located in remote areas, far from fixed asphalt production facilities, so high-risk RAP is often not available anyway. Consequently, the sources of RAP applicable to airport resurfacing generally include the low-risk temporary ramps, as well as the medium-risk material from milling and texturing of the existing surface.
3 Methods To determine the potential effects of extending RAP use from low-risk sources to medium-risk sources, the data from a pilot project was analysed. The airport was resurfaced in 2018 during short night-work periods and the asphalt production and construction were typical of runway resurfacing works. The airport is located on the central Queensland coast in Australia and caters for domestic (B737) and international (up to B787) aircraft on a regular basis. Although the project was performed notionally allowing RAP, the time required to accumulate adequate volumes, test the RAP and adjust the mixture design delayed its inclusion. As a result, approximately 40% of the runway was resurfaced without RAP and the other 60% of the runway included 15% RAP, 5% from the temporary ramps (low-risk) and 10% from the existing surface millings (medium-risk). The runway surface was inspected in October 2019 and was found to be in sound condition, as expected for a surface of its age. There was no noticeable difference in the condition of the surface between the areas where RAP was used and where it was not used.
4 Results and Discussion 4.1 Mixture Design The virgin mixture, containing no RAP, was typical of dense graded 14 mm sized airport asphalt (Table 1) with a proprietary plastomeric polymer modified binder known as JetBind® . To allow the RAP to be incorporated into the design, the construction trials included removal of a mock-ramp and texturing of a small area of the existing surface. The two RAP sources were characterized in triplicate and the mixture design was adjusted. As shown in Fig. 1a (grading) and Table 1 (other properties) the RAP mixture design was similar to the virgin design. The two RAP sources showed significant grading variability and some gradings were outside the mixture specification limits (Fig. 1b) for the coarse aggregate. Furthermore, the aged existing surface RAP had a lower binder content, 4.7% compared to 5.2% for the ramps, and the existing surface RAP had a higher moisture content, 1.7% compared to 0.6%.
914
G. White and A. Jamshidi
Table 1 Asphalt mixture volumetric and other properties
Property
Specification Virgin mixture RAP mixture
Binder content (%)
5.2–5.8
5.5
5.5
Maximum density (t/m3 )
Report only
2.477
2.474
Marshall air voids (%)
3.0–4.0
3.5
3.3
Marshall Stability (kN)
>10
16.5
17.5
Marshall Flow(mm)
1,000,000
>1,000,000
Tensile strength >80 ratio (%)
99
98
Indirect tensile Report only modulus (MPa)
3460
3780
Note All testing to the requirements of the airport asphalt specification [3]
(a) Virgin and RAP mixtures Fig. 1 Comparison of aggregate gradation
(b) Triplicate surface and ramp RAP
Extending the Use of RAP in Airport Asphalt Resurfacing
915
Table 2 Statistics for RAP production properties Property
Surface RAP Ave
Ramp RAP CV
Ave
T-test CV
p-value
Binder content (%)
4.8
8%
5.2
3%
0.03
Moisture content (%)
1.6
30%
0.6
80%
0.01
9.5 mm
84
6%
87
3%
0.13
6.7 mm
70
8%
74
4%
0.19
4.75 mm
59
11%
60
6%
0.75
2.36 mm
42
13%
44
7%
0.54
1.18 mm
31
11%
33
9%
0.37
0.600 mm
24
11%
25
10%
0.50
0.300 mm
15
12%
15
14%
0.89
0.150 mm
10
14%
9
22%
0.40
0.075 mm
7
17%
6
30%
0.36
Percentage passing sieve (%)
Note Ave. = average, CV = coefficient of variation
However, given the modest RAP content available for use, the mixture designs were not affected by the differences between the RAP and the virgin materials.
4.2 Recycled Asphalt Production Properties During production, the ramp RAP and the existing surface RAP were routinely tested for grading, binder content and moisture content to detect any migration of properties over time. The average moisture content and binder contents were comparable to those reported above for the mixture design and remained significantly different for the ramp and existing surface RAP (Table 2). However, the existing surface RAP and ramp RAP gradings were not significantly different, indicating that these two sources of RAP were essentially similar except for their binder content and age. Furthermore, the variability of the gradings associated with the two RAP sources were comparable or less than for common virgin aggregate sources [5] indicating the use of RAP is unlikely to introduce an asphalt production challenge due to variability.
4.3 Asphalt Production The asphalt production test results were analysed to determine the impact of RAP on asphalt variability and consistence with the mixture designs. The gradings were
916
G. White and A. Jamshidi
consistent, with and without RAP, with Student T-test p-values all greater than 0.05. The average grading and two standard deviations either side of the average grading were also all within the specified [3] production tolerances for asphalt gradation (Fig. 2), although the RAP mixture was slightly coarser, with (on average) approximately 3% less material passing the 4.75, 6.7 and 9.5 mm sieves. The other production properties were also comparable, as shown in Fig. 3. It is expected that the slightly different gradings (Fig. 2) resulted in a lower voids in the aggregate, and consequently a lower Marshall air voids and higher Marshall Stability for the mixture with the RAP, which were all significantly different with p-values below 0.01. In contrast, the binder content, maximum density and Marshall Flow were not significantly different, with p-values of 0.31, 0.36 and 0.06, respectively. These properties are not greatly affected by the asphalt mixture grading, noting that Marshall Flow is usually dominated by the binder properties in polymer modified mixtures. Although there were some significant differences between the virgin mixture and the mixture with RAP, these differences were not detrimental, as they were accounted for in the mixture design process and accommodated by verification of consistent RAP properties during production.
Fig. 2 Comparison of aggregate production gradation for virgin and RAP mixes
Extending the Use of RAP in Airport Asphalt Resurfacing
(a) Binder content, Flow and Air Voids
917
(b) Stability and Voids in Aggregate
Fig. 3 Comparison of asphalt production properties for virgin and RAP mixes
5 Conclusion It was concluded that the pilot project demonstrated that the use of RAP in airport asphalt surfacing can be extended from the low-risk source associated with temporary ramps, to the medium-risk source associated with existing surface texturing by cold planning machines. However, the mixture design must characterize and reflect the RAP sources and the production must monitor the RAP, similar to other raw materials, such as virgin aggregate and bituminous binder. In the future, other project data should be analysed to confirm the trends observed in the pilot project and then the Australian airport asphalt specification should be amended to include RAP characterization and production monitoring protocols.
References 1. Airfield Pavement Essentials, Practice Note 12, Australian Airports Association, April 2017 2. White, G.: State of the art: asphalt for airport pavement surfacing. Int. J. Pavement Res. Technol. 11(1), 77–98 (2018) 3. Performance-based Airport Asphalt Model Specification. Version 1.0, Australian Asphalt Pavement Association. Melbourne Victoria, 1 February 2018 4. White, G.: Incorporating RAP into airport asphalt resurfacing. In: 7th International Conference on Bituminous Mixtures and Pavements, Thessaloniki, Greece, pp. 543–550, 12–14 June 2019 5. White, G., Jamshidi, A.: Resetting dense graded airport asphalt production and construction tolerances. Int. J. Construct. Manage. (in Press)
Factors Affecting Mortar Thickness Distribution and Its Relationship to Cracking Resistance of Asphalt Mixtures Hussain U. Bahia, Jiwang Jiang, and Ying Lia
Abstract This study introduces a novel method to determine the asphalt mortar thickness distribution through two-dimensional image processing techniques. It is claimed that in asphalt mixture design, there exists a potential optimal mortar thickness distribution, which can be used as metrics for quality control of durability. Gradation analysis from images was performed to verify the reliability of the analysis based on known gradation curve of mixtures and potential limitations for analysis are introduced. Image-based calculated gradation of asphalt mixtures is compared with gradation calculated from sieve analysis to verify accurate propotions of mixtures from images. Mixtures produced with modified and unmodified asphalt binders were utilized to compare the mortar distribution and the analysis of results is provided. In addition, three compaction temperatures were selected to investigate the temperature influence on mortar thickness distribution. The results indicate that there could be significant effects of binder modification and compaction temperature on mortar thickness distribution. Furthermore, the relationship between the mortar thickness distribution and the cracking resistance from the Semi-CircularBend test were analyzed. The relationships obtained verify that thinner mortar thickness, with a wider distribution, could cause the asphalt mixture to be more sensitive to aging, which is detrimental to the durability of the mixtures. The method used can be utilized for better mixture design and quality control to optimize binder content and durability. Keywords Asphalt mastic · Film thickness · Imaging · Cracking resistance · Oxidative aging
H. U. Bahia (B) · J. Jiang · Y. Lia Department of Civil and Environmental Engineering, University of Wisconsin—Madison, 1415 Engineering Dr., Madison, WI 53706, USA e-mail: [email protected] J. Jiang School of Transportation, Southeast University, 2# Sipailou, 210096 Nanjing, Jiangsu, China © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_117
919
920
H. U. Bahia et al.
1 Introduction and Objectives Researchers have reported that internal structure of an asphalt mixture has a significant influence on its mechanical properties and resistance to major distresses including rutting, fatigue and thermal cracking as well as the low temperature cracking [1–4]. An important parameter of internal structure, distribution of asphalt mastic or mortar film thickness has not been the focus of these studies, although its importance in durability is widely recognized. Studies have been performed to investigate the two-dimensional image of asphalt mixture through image processing [5] and also the three-dimensional space [6]. However, the method of calculating the fine matrix/mastic thickness distribution used in these recent studies are not clear as the threshold of aggregate size to be considered in the mastic are chosen without clear justification [7–10]. This study is focused on developing a more accurate method for selecting the minimum aggregate size that can be detected from imaging, showing the effects of modification and compaction temperature on mortar film distribution in typical mixtures, and verifying the effect of mortar thickness on the durability of asphalt mixtures. The main objectives of the study are to: • Develop and implement a protocol for mortar thickness distribution analysis for asphalt mixtures and define representative indices that capture the important features of mortar thickness distribution; • Evaluate effect of modifiers and compaction temperatures on mortar film thickness distribution of asphalt mixtures; and • Verify effect of mortar thickness on the durability by correlating the results of SCB test with the mortar thickness distribution.
2 Methodology of Imaging Asphalt mixtures, mixed and compacted to a height of 150 mm and diameter of 150 mm, are cut into 4 pieces in vertical directions, as shown in Fig. 1. Six sections of the asphalt mixture are scanned by a high-resolution scanner with resolution of 1200 dpi (equal to 0.0211 mm/pixel). Colored images of asphalt mixture profile can be obtained and preprocessed into binary images which is in black and white instead of in RGB (Red, Green, Blue) format. In previous studies, image processing tools was developed named Image Processing and Analysis System (IPAS) [11]. The common digital imaging processors algorithm is utilized to transform the colored images into grayscale images. The original grayscale images are further processed by performing filtering and transformation. The median filter is selected to remove the noises as explained by Coenen [9]. For image accuracy verification purpose, gradation calculated through images is compared with actual mixture sieve analysis. To get the gradation from images of asphalt mixture, the binary images need to contain as much aggregates’ information as possible without minimum loss due to transformation and thresholding. Previous
Factors Affecting Mortar Thickness Distribution …
921
Fig. 1 Cutting and scanning process of the 2D image
research showed coarse aggregate gradation can be gathered through the equivalent diameter of aggregates from image processing technique [2]. Equivalent diameter is the diameter of a fictitious circular aggregate that has the same area as the aggregate. However, it has been claimed the fine aggregates gradation was not captured due to contact between fine aggregates [12]. To cope with this, we have utilized Surface erosion–dilation techniques to solve this problem. The erosion-dilation technique [13] can smooth the angularity of the aggregate to some extend such that some unnecessary contact points induced by transformation of colored image is reduced. Based on previous research, the gradation calculated via equivalent diameter showed very similar result compared to their sieve analysis result. In this study, equivalent diameter- based calculations were adopted [14]. The binary image can be utilized to calculate the mortar thickness distribution. After converting the colored images to binary images, all aggregates are labelled in different numbers. For each pixel in the boundary of the aggregate object, called outline pixel, a searching radius is used to perform searching for the minimum distance between this pixel to another aggregate. As shown in Fig. 2, the mortar film thickness is defined as the minimum distance between the outline pixel and the neighboring coarse aggregates. For each image, thousands of different mortar thickness values were collected and, as shown in Fig. 2, use to plot the frequency histogram of these values counted by an interval of 0.1 mm. To quantify the distribution of the mortar thickness, two indexes, including mean value of the mortar average thickness (Tm ) and the standard deviation of the mortar thickness (SDt ), are proposed.
8%
Searching radius
Zoom
6%
Frequency
Mortar film thickness
4%
2%
0%
Mortar thickness (mm)
Fig. 2 Schematic map of searching in asphalt mixture and typical sample for mortar thickness distribution frequency histogram
922
H. U. Bahia et al.
3 Mortar Thickness Distribution Analysis The mortar film thickness distributions of the same mixture design produced with different binders (N, P, H and S) as well as varying compaction temperatures at 115, 145, and 165 °C were obtained, and the summary of both mean value and standard deviation were calculated. The results indicate that the distribution is a function of the binder modification or compaction temperature. In most cases, the mean value of mortar film thickness was close to 1.00 mm in most mixtures, however, the distribution of the mortar thickness varies and shifts from the mean value (Tm ) or standard deviation (SDt ) depending on the modifier used, and the compaction temperature. Figure 3 shows three typical frequency histograms of the mortar thickness distribution from the data set, including P modified asphalt mixtures, H modified asphalt mixtures and N asphalt mixtures. For the same binder content, the mixtures with higher Tm and lower SDt are considered better from a durability point of view because thicker films with a narrow variation ensure good resistance to moisture and oxidation, and a narrow variation ensures good economics as no binders are wasted in thicker or thinner than needed films. The importance of estimating the mortar film thickness is related to cost and durability. While traditionally we consider the Voids filled with Asphalt (VFA) and in some cases we estimate the average film thickness, the results in Fig. 3 show that we also need to estimate the distribution of the film thickness. Since it is known that asphalt binder oxidation and moisture damage are diffusion-controlled phenomena, it is logical to assume that a higher mean film thickness is better for durability than a smaller mean value. In addition, since it is shown in this study that distribution of the film thickness can vary significantly between mixtures, a narrow distribution of a higher mean is also better than a higher mean value with wide distribution. 7%
145 CH
Narrower but thinner T m /S D t = 1 .4 0
6%
145 CN 145 CP
Frequency
5% 4% Narrow and thick T m /S D t = 1 .5 2
3%
Thicker but wider T m /S D t = 1 .4 2
2% 1% 0% 0
0.4
0.8
1.2
1.6
2
2.4
Mortar thickness (mm) Fig. 3 Typical frequency histogram of the mortar thickness distribution
2.8
3.2
3.6
Factors Affecting Mortar Thickness Distribution …
923
Obviously, we need to decide what constitutes an optimum mortar film thickness. In the literature and in some specifications, the minimum average film thickness estimated from surface area of aggregates of 8–10 µm are used [15], and VFA values of 60–75% are commonly used. The problem is that neither of these 2 measures truly reflect the distribution of the film (SD) and thus are not sufficient. These measures also assume that it is the binder rather than the mortar that should be controlled for durability. This assumption is not realistic since studies have shown that most of the voids are larger than 1.00 mm in diameter as detected by the MRI technologies [16]. Also, it is well documented that most of the surface area (~80%) comes from the filler portion with sizes well below 0.10 mm. It is therefore not accurate to assume that calculated VFA and average binder film thickness are related to durability since voids are not in this range, and filler is an extension of the binder in coating the larger aggregates. Therefore, since higher Tm and lower SDt is the target, the ratio of Tm and SDt (Tm /SDt ) can be proposed to quantify the effect of mortar thickness on durability.
4 Effect of Modification and Compaction Temperature Modification of asphalt binder has been shown in to influence the internal structure of aggregates in asphalt mixtures due to effect of modifiers on lubricity of binders [3]. Figure 4 illustrate the Tm /SDt of mixtures with different modification at different compaction temperatures. It is found that the highest value of Tm /SDt from P modified asphalt mixture occurs at compaction temperature of 165ºC. For both H and S modified asphalt mixtures, the highest Tm /SDt occurs at compaction temperature of 145 °C, while for neat asphalt mixtures, lower compaction temperature of 115ºC gives the highest Tm /SDt . 1.55 1.5
115 º C 145 º C 165 º C
Tm /SDt
1.45 1.4 1.35 1.3 1.25 P
H
S
N
Types of modification Fig. 4 The Tm /SDt value from asphalt mixtures with different modifications at varying compaction temperatures
924
H. U. Bahia et al.
Aging change (ST-LT)
12.00
S-34 V-28 8.00
4.00 0.60
0.70
0.80
0.90
T m (mm)
Fig. 5 Relationship between aging rate from FI and the Tm from mortar thickness distribution
5 Effect of Mortar Thickness on Aging of Asphalt Mixtures Two mixture designs and two aging conditions were chosen for this study: Shortterm oven aging (STOA) with four hours loose mix aging at 135 °C, while long-term oven aging (LTOA) followed the procedure of 12 h loose mix aging at 135 °C. Two SCB specimens of 50 ± 1 mm thick slices were obtained from the center of the SGC specimen. As for image processing, the SGC specimens were cut vertically into four pieces and six images were scanned from each SGC specimen. Finally, four replicated SCB samples were prepared for each mixture under both aging conditions and 12 images were scanned from two replicated SGC samples from each mixture. The relationship between the change in FI due to aging (ST-LT) versus the Tm of mixtures is shown in Fig. 5. The limited data verify that thinner mortar thickness contributes to higher sensitivity to aging (change in FI), The slope of aging rate versus Tm from S-34 binder and V-28 binder are extremely different, which may be caused by the higher aging resistance of the S-34 binder as compared to the V-34 binder. For further study, it is necessary to relate both the mortar thickness distribution and the durability performance of mortar materials to the durability of asphalt mixtures.
6 Summary of Findings Based on the analysis of results collected the following points provide a summary of findings: 1.
Mortar thickness distribution in mixtures of same aggregate gradations and binder contents is significantly influenced by modification of asphalt binder and compacting temperature. The mean thickness (Tm ) and distribution (SDt ) of
Factors Affecting Mortar Thickness Distribution …
2.
3.
925
mortar in mixtures can be estimated using the imaging and analysis procedure developed from a set of 2-D images of mixtures using a flat-bed scanner. The ratio of Tm /SDt is proposed as a single index for mortar film thickness effects. There exists an optimum compacting temperature for asphalt mixtures with different modified binders. Higher Tm/SDt is favorable in mixture design as it indicates better distribution of mortar thickness leading to more durable mixture. Based on the SCB- Flexibility Index results and mortar thickness distribution analysis, the concept of better resistance to aging for higher values of Tm/SDt is verified. It is found that thinner mortar thickness could contribute to a higher sensitivity to aging, which may be detrimental to the durability of the asphalt mixtures.
References 1. Masad, E., Button, J.: Relationship between aging rate from implications of experimental measurements and analyses of the internal structure of hot-mix asphalt. TRR J. Transport. Res. Board 1891:212–220 (2004) 2. Yue, Z.Q., Bekking, W., Morin, I.: Application of digital image processing to quantitative study of asphalt concrete microstructure. Transport. Res. Record 1492, 53–60 (1995) 3. Sefidmazgi, N.R., Tashman, L., Bahia, H.: Internal structure characterization of asphalt mixtures for rutting performance using imaging analysis. Road Mater. Pavement Design 13(sup1), 21–37 (2012) 4. Underwood, B.S.: Multiscale constitutive modeling of asphalt concrete (2011) 5. Jiang, J., et al.: Effect of the contact structure characteristics on rutting performance in asphalt mixtures using 2D imaging analysis. Construct. Build. Mater. 136, 426–435 (2017) 6. Bhasin, A., Izadi, A., Bedgaker, S.: Three dimensional distribution of the mastic in asphalt composites. Const. Build. Mater. 25(10), 4079–4087 (2011) 7. Kose, S., et al.: Distribution of strains within hot-mix asphalt binders: applying imaging and finite-element techniques. J. TRB 1728 (2000). 8. Jiang, J., et al.: Evaluating the mastic distribution of asphalt mixtures based on a new thickness threshold using 2D image planers. Road Mater Pavement Design 19(6), 1422–1435 9. Coenen, A.: Image analysis of aggregate structure parameters as performance indicators of rutting resistance. Ph.D. thesis, University of Wisconsin-Madison (2011) 10. Sefidmazgi, N.R.: Hot mix asphalt design to optimize construction and rutting performance properties. Dissertation, The University of Wisconsin-Madison (2014) 11. Chen, D., Sefidmazgi, N.R., Bahia, H.: Exploring the feasibility of evaluating asphalt pavement surface macro-texture using image-based texture analysis method. Road Mater. Pavement Design 16(2), 405–420 (2015) 12. Masad, E., et al.: Internal structure characterization of asphalt concrete using image analysis. J. Comput. Civ. Eng. 13(2), 88–95 (1999) 13. Masad, E., Button, J.: Unified imaging approach for measuring aggregate angularity and texture. Computer-Aided Civil and Infrastructure Engineering 15(4), 273–280 (2000) 14. Elseifi, M.A., et al.: Validity of asphalt binder film thickness concept in hot-mix asphalt. Transport. Res. Record 2057(1), 37–45 (2008) 15. Radovskiy, B.: Analytical formulas for film thickness in compacted asphalt mixture. Transport. Res. Rec. J. Transport. Res. Board 1829, 26–32 (2003)
926
H. U. Bahia et al.
16. Masad, E., Jandhyala, V.K., Dasgupta, N., Somadevan, N., Shashidhar, N.: Characterization of air void distribution in asphalt mixes using X-ray computed tomography. J. Mater. Civ. Eng. 14(2), 122–129 (2002)
Fatigue and Moisture Damage Resistance of Bituminous Mixtures for Railway Trackbeds Diego Ramirez Cardona, Hervé Di Benedetto, Cédric Sauzéat, and N. Calon
Abstract This paper focuses on the fatigue resistance properties and susceptibility to moisture damage of different bituminous mixtures used as sub-ballast layers in French railways. The analysis is made taking into account the specific working conditions and service life requirements of trackbeds. Three different road basecourse mixtures used as sub-ballast layers in France were studied. A new moistureconditioning procedure was proposed and used to induce moisture damage in the samples. Tension-compression fatigue tests were carried out on cylindrical samples at the Laboratory of Civil Engineering and Construction (LGCB) of the ENTPE, University of Lyon. All the tested mixtures were found to present satisfactory fatigue resistance properties. Low voids content and the use of polymer-modified bitumen were found to improve the resistance to moisture damage. Keywords Sub-ballast layer · Bituminous mixture · Fatigue resistance · Fatigue damage · Moisture damage · Tension-compression test · High-speed lines
1 Introduction Bituminous mixtures have been used in several countries for sub-ballast layers as an alternative to compacted soil. Some advantages of the use of bituminous subballast layers include reduced vertical accelerations in the ballast, reduced track maintenance, lower moisture content variation of the subgrade and constituting a trafficable access way during construction [1–4].
D. Ramirez Cardona (B) Eiffage Infrastructures, Direction of Research & Innovation, Corbas, France e-mail: [email protected] H. Di Benedetto · C. Sauzéat ENTPE, Laboratory of Tribology and System Dynamics (LTDS) (UMR CNRS 5513), University of Lyon, Vaulx-en-Velin, France N. Calon SNCF Réseau, Direction of Industrial Production, Paris, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_118
927
928
D. Ramirez Cardona et al.
The 3 km test section of the East-European (EE) HSL was the first track section with bituminous sub-ballast in France and is in service since 2007. In more than 10 years of commercial service, the maintenance works on this test section have been significantly lesser compared to the adjacent sections with conventional subballast [4]. The material used for this test section is a typical French base-course bituminous mixture called GB3 [5]. Since then, four major high-speed lines (HSL) in France have been totally or partially built with bituminous a sub-ballast layer, using a GB4 type mixture [5]. Even though the feedback is positive, it does not allow accounting for the long term performance of the bituminous materials in the trackbeds. As sub-ballast layers, the base-course mixtures are exposed to weather conditions, especially moisture, which differs from their usual working conditions in road pavement structures. This paper presents the effects of moisture conditioning on the fatigue resistance properties of three different base-course mixtures: GB3, GB4 and a mixture with polymer-modified binder (PMB) called GB PMB. The interest of using PMBs in bituminous railway trackbeds is addressed. Fatigue tension-compression tests were carried out on cylindrical samples at the LGCB/LTDS laboratory of the Ecole Nationale des Travaux Publics de l’Etat (ENTPE), part of the University of Lyon, France. A new moisture conditioning procedure was developed at the same laboratory for this study.
2 Fatigue and Moisture Damage of Bituminous Mixtures The repeated loading generated by traffic entails the appearance of micro-cracks which progressively degrade the material’s mechanical properties. After a certain number of applied loading cycles, micro-cracks evolve into macro-cracks that cause failure. This progressive weakening of the material properties leading to failure is known as fatigue damaging [6]. The French design method, used for roads and adapted for sub-ballast layers, seeks to limit the strain at the bottom of bituminous layers to a lower value than that identified as the fatigue resistance of the material. This value, called ε6 , is accessed through laboratory fatigue tests. Fatigue testing at different strain amplitudes (ε0 ) allows plotting a Wöhler Curve (number of cycles to failure Nf against ε0 ), described as: log ε = α − β log N
(1)
In this study, strain-controlled sinusoidal tension-compression fatigue tests were carried out on cylindrical samples. A slow and progressive stiffness (|E*|) loss is observed as the number of applied loading cycles (N) increases. Failure is considered when the |E*| loss accelerates, which indicates the initiation of a macro-crack within the sample [7]. Moisture damage of bituminous mixtures was defined by [8] as “the progressive functional deterioration of a pavement mixture by loss of the adhesive bond between
Fatigue and Moisture Damage Resistance of Bituminous …
929
the asphalt cement and the aggregate surface and/or loss of the cohesive resistance within the asphalt cement principally from the action of water”. Moisture damage mechanisms might affect the service life of the materials without necessarily affecting their short-term performances. Fatigue tests were then used to assess the degradation after exposure to moisture, in complement to complex modulus tests which results are presented in [7].
3 Materials and Test Protocol The tested GB3 and GB4 mixtures are common French road base course materials [5]. GB3 mixtures are considered basic materials and GB4 mixtures are usually intended for pavement structures subjected to high traffic. The tested GB4 mixture is the same one that constitutes the BPL HSL sub-ballast. Thanks to its optimized aggregate packing, which increases the density, it is considered as an improved alternative to the reference GB3 without the use of additives. The GB PMB mixture has the same mix design as the GB3 but it uses a SBS polymer modified bitumen (see Table 1). All three mixtures where fabricated at the Centre for Studies and Research of the French company Eiffage. In order to accelerate moisture induced damage, the samples were submerged in a water bath at 60 ± 1 °C for 168 h, then surface dried and placed in a freezer at −18 ± 1 °C for 24 h inside a hermetic plastic bag with 10 cl of water. Finally the samples were placed inside a temperature chamber at 40 ± 1 °C for 30h. This new moisture conditioning protocol was developed based on the “Duriez” and AASTHO T283 tests. It is aimed to accelerate moisture damage mechanisms without causing damage of other natures [7], such as micro-cracking that can be induced when forcing the saturation of the samples. Fatigue tests were carried out applying an axial tension-compression sinusoidal loading, centered on zero stress and strain, by a hydraulic press in strain-control mode. The monitored signal is the average of the strain signals given by three extensometers (see Fig. 1). The samples were placed inside a thermal chamber. The target temperature for all tests was 10 °C (surface temperature) and all tests were loaded at 10 Hz.
4 Test Results and Analysis The fatigue resistance properties of the materials were characterized for different moisture conditioning states. The test results are indicators of the fatigue resistance properties of the materials under the specific conditions of the laboratory tests. Therefore; the number of cycles until failure (Nf ) cannot be interpreted as the number of loading cycles that the materials withstand on a pavement structure for a specific strain amplitude. The Nf value corresponds to the average of the following failure
930
D. Ramirez Cardona et al.
Fig. 1 Instrumented sample (left) and instrumentation top-view scheme (right)
Table 1 Characteristics of the tested mixtures GB3
GB4
GB PMB
Bitumen penetrability at 25 °C
35/50
35/50
35/50
Binder type
Base bitumen
Base bitumen 2.5% Biprène® 41R
Binder content [% aggregate 4.4 mass]
4.8
4.4
Bulk specific gravity [g/cm3 ]
2.49
2.48
2.49
Coarse and fine aggregates (0/14)
Rhyolite—Baglione quarry (53 Mayenne, France)
Added filler
Limestone—Haut-Lieu quarry (59 Nord, France)
criteria: 50% loss of the initial modulus, maximum phase angle and change in the evolution of the Dissipated Energy Ratio. The samples’ voids contents are typical.
4.1 Study for Unconditioned Samples Fig. 2 presents a comparison of the three materials at a single strain level (70 μm/m) in a |E*| against N plot. Its serves as an example to illustrate the influence of the mixtures’ characteristics on the modulus loss during a fatigue test. The higher modulus values observed for the GB4 can be explained by its optimized granular packing. However, the GB4 does not present a longer fatigue life than the GB PMB. Figure 3 presents the Wöhler curves for each material at the non-conditioned state. The non-conditioned GB4 and GB3 mixtures present very similar fatigue resistance properties. Optimizing the aggregate packing of bituminous mixtures does not seem to significantly affect the fatigue resistance properties. The GB PMB, however, presents a better fatigue behavior than the GB3 for the tested strain amplitudes (see
Fatigue and Moisture Damage Resistance of Bituminous …
931
Fig. 2. |E*| versus N plots of a test at 70 μm/m of each of the three studied mixtures
Fig. 3. Wöhler curves: non-conditioned GB3 compared to GB4 (a) and GB PMB (b)
Fig. 3b). The GB PMB’s fatigue life is expected to be much higher than the GB3’s. Even though the GB4 and GB PMB present the same susceptibility to changes in strain amplitude (same slope), the latter presents higher fatigue life at all strain amplitudes. The use of PMB is then more effective in enhancing fatigue resistance of bituminous mixtures than optimizing the aggregates packing (i.e. increasing density).
4.2 Study for Moisture Conditioned Samples Figure 4 compares the Wöhler curves of the conditioned and non-conditioned GB3 and GB4 mixtures. For the GB3, the lower ε6 value and higher data scatter of the moisture conditioned results show a clear degradation of the material. As for the GB4, the fatigue parameters do not seem to have been altered by the exposure to moisture. Nonetheless, the high dispersion of both data sets (conditioned and not) makes it difficult to confirm a clear tendency. The GB4 samples subjected to moisture
932
D. Ramirez Cardona et al.
Fig. 4. Effect of moisture conditioning on GB3 (a) and GB4 (b) fatigue properties (Wöhler curves)
conditioning seem to have been affected very differently one from the other. This could be due to the presence of zones of privileged contact with water in some samples. As the GB3, the GB PMB shows clear signs of deterioration after moisture conditioning (see Fig. 5). However, the moisture-conditioned GB PMB ε6 is higher than those from the GB3 and GB4, in both states. This clearly indicates that the GB PMB performs better to fatigue than the other two studied mixtures, even after moisture damage. Figure 5b presents the saturation of the samples after water bath and surfacedrying with respect to the samples’ air voids. The GB4 and GB PMB present similar saturation levels in spite of the higher voids content of the latter. Saturation levels of the GB3 are very scatter which is an indicator of the higher susceptibility to moisture of this material.
Fig. 5. Effect of moisture conditioning on GB PMB fatigue properties (Wöhler curves) (a) and saturation of the samples against voids content (b)
Fatigue and Moisture Damage Resistance of Bituminous …
933
5 Discussion Measures from instrumented zones in French HSL show that the strain levels to which bituminous sub-ballast layers are exposed are very low (lower than 20μm/m) [4, 9]. These measurements, though mostly for high-speed train circulations, also include data from freight trains and passenger trains at low speeds. Based on the French road design method, the fatigue resistance properties of the three studied mixtures, even after moisture damage, seem suitable for their use in trackbeds. Mix design and construction operations need, however, to limit the exposure to water as much as possible. The instrumentation of classic lines (for freight and passenger trains) with bituminous trackbeds is nonetheless recommended to obtain long-term feedback for this kind of traffic. The results from this study can be used to optimize the design of bituminous sub-ballast layers. The use of optimized aggregate packing and of PMB would provide high durability. Further steps might include the establishment of correlations between the laboratory fatigue resistance properties and in-situ fatigue life, which would be possible through the analysis of the feedback from inservice structures. The influence of the typical “M” shape of a bogie loading on the bituminous trackbed should be studied through mechanical tests and modelling.
References 1. Teixeira, P.F., López Pita, A., Ferreira, P.A.: New possibilities to reduce track costs on high-speed lines using a bituminous sub-ballast layer. Int. J. Pavement Eng. 11(4), 301–307 (2010) 2. Rose, J.G., Bryson, L.S.: Hot mix asphalt railway trackbeds: trackbed materials, performance evaluations ans significant implications. In: International Conference on Perpetual Pavements, vol. 7326, p. 19 (2009) 3. E. A. P. A.: Asphalt in Railway Tracks. Breukelen, Pays Bas (2014) 4. Ramirez Cardona, D., Di Benedetto, H., Sauzéat, C., Calon, C., Saussine, G.: Use of a bituminous mixture layer in high-speed line trackbeds. Constr. Build. Mater. 125, 398–407 (2016) 5. AFNOR: Bituminous Mixtures—Material Specifications—Part 1: Asphalt Concrete, France, p. 17 (2007) 6. Di Benedetto, H., Corté, J.-F.: Matériaux routiers bitumineux 2 : constitution et propriétés thermomécaniques des mélanges, Paris (2004) 7. Ramirez Cardona, D.: Characterisation of Thermomechanical Properties of Bituminous Mixtures Used for Railway Infrastructures. Ecole Nationale des travaux publics de l’état (ENTPE), University of Lyon (2016) 8. Kiggundu, B.M., Roberts, F.L.: Stripping in HMA Mixtures: State-of-the-Art and Critical Review of Test Methods, Auburn, AL (1988) 9. Khairallah, D.: Analyse et modélisation du comportement mécanique des structures de voies ferrées avec sous-couche bitumineuse, Ecole Centrale de Nantes (2019)
Fracture Characterization of Stone Mastic Asphalt (SMA) with Hydrated Lime Through the Semi-circular Bending Test Approach Abdallah M. Ahmed, Juan S. Carvajal-Munoz , and Gordon Airey
Abstract Hydrated lime (HL) use as an active filler is due to its ability to increase the stiffness of asphalt mixtures as well as the mechanical and physicochemical performance through improved rutting resistance, reduced aging, and enhanced moistureinduced damage resistance. However, understanding of the effects of HL on fatigue or more specifically on the fundamental fracture properties of asphalt mixtures still requires further investigation. As a result, this research study assessed the fracture toughness characteristics of stone mastic asphalt (SMA) mixtures with locally available materials and testing procedures in the UK context. SMA cores were subjected to semi-circular bending (SCB) testing. The experimental matrix comprised various factors for assessing and comparing the fracture toughness of the manufactured mixtures with 0, 1, 2, and 3% hydrated lime by weight of total mixture as well as a control without hydrated lime. These factors included the effects on the fracture of (1) testing temperatures (0, 10, and 20 °C), (2) aggregate type (granite and limestone), (3) environmental conditioning (ageing combined with moisture damage), and (4) the effects of filler type and content. From the SCB experimental results, it was found that the fracture is highly factor-dependent. 2 and 3% HL improved the fracture strength of the manufactured cores, whereas 1% exhibited relatively low or negligible effects on the fracture strength. Finally, the use of granite aggregates substantiated a superior fracture resistance with a 40% increase compared to mixtures produced with limestone. Keywords Hydrated lime · Stone mastic asphalt · Fracture · Semi-circular bend · Toughness
A. M. Ahmed · G. Airey Nottingham Transportation Engineering Centre (NTEC), University of Nottingham, Nottingham, UK e-mail: [email protected] J. S. Carvajal-Munoz (B) Department of Civil and Environmental Engineering, Universidad del Norte, Barranquilla, Colombia e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_119
935
936
A. M. Ahmed et al.
1 Introduction The fatigue cracking resistance of bituminous mixtures significantly affects the durability and longevity of pavements. Various methods are currently available to study fatigue cracking from the mechanical point of view but recent efforts have been focused towards more detailed understanding of the materials’ response. Precisely, fracture is a material property that helps understanding the response to fatigue cracking from a more fundamental perspective. Moreover, the fatigue of pavements is a multidependant material response in which the materials’ properties (i.e., air voids, bitumen content, gradation, aggregate type, environmental conditioning) and the testing conditions (i.e., temperature, time) have an influence that needs to be studied for specific conditions. Based on current literature available, an interesting research gap was identified with respect to very limited/scarce studies assessing the effects of HL on the fracture properties of SMA using the Semi-circular Bend (SCB) test approach. Therefore, this research study investigates the fracture resistance of aged and unaged SMA produced with the addition of hydrated lime (HL) at various temperatures using the SCB test approach. SMA is known for its durability and rutting resistance and is a mixture type predominately used as the binder coarse in main roads subjected to heavy traffic. HL is a filler type that has been in use in the asphalt industry from the early 70 s mainly in the US. HL is an inorganic chemical compound of calcium dihydroxide (Ca(OH)2 ) that is manufactured from limestone rocks through a chemical process. Through recent research efforts, HL has demonstrated proven benefits in pavements that relate to increased resistance to moisture damage, rutting and aging. Nevertheless, the fracture resistance induced by HL is still under discussion and further research needs to be conducted to gain a deeper understanding as to the mechanisms and benefits for specific materials and testing conditions. Previous research has investigated the strength of SMA mixtures with the addition of steel slag (i.e., a byproduct of the steel manufacturing process) [1]. The investigation focused on the low-temperature cracking performance, which was found to increase with the use of steel slag aggregates in contrast to basalt aggregates. In addition to the additive’s influence on the fracture resistance, the authors also concluded that the aggregate type significantly impacts this material response. There has been significant research into the fracture properties of bitumen and the crack onset and propagation phenomena [2–5]. Multiple materials and testing approaches have been considered. For instance, [2] investigated the self-healing cracking properties of asphalt mixtures to discover the effects of temperature on the healing rates of mastics. The results indicated that the amount and duration of heat transfer to the material would affect the recovery of the damaged bitumen. Conversely, Simone and Lantieri [3] found that the fatigue and healing properties of bituminous pavements can be further improved with the addition of new materials such as rubber and fillers. The main conclusion deals with the superior performance obtained for those mixtures produced with recycled materials in contrast to the virgin mixtures. Additional efforts include the work by [4] that studied the suitability of a SCB test for
Fracture Characterization of Stone Mastic Asphalt (SMA) …
937
Table 1 Mixture characteristics and testing parameters considered Mixture type Air voids (%)
Bitumen type
Condition
Aggregate type
HL (%)
Temperature (°C)
SMA10mm
4.0
30/45
Unaged
Limestone
0, 1, 2, 3
0, 20
SMA10mm
4.0
30/45
Conditioned
Limestone
0, 1, 2, 3
0, 10, 20
SMA10mm
4.0
30/45
Conditioned
Granite
0, 1, 2, 3
0, 10, 20
the assessment of the fracture resistance of bituminous mixtures. According to the authors, this testing method allows for the comparison of various fracture properties such as crack growth, crack rate, and effective crack length. The SCB method has been shown to be simple and repeatable enough to be used as an effective indicator of the fracture properties of bituminous mixtures [5].
2 Research Methodology 2.1 Bituminous Mixture Design and Test Parameters All materials and parameters selected for the experimental programme are derived from the conducted literature review and are in accordance with the main aim of the investigation. A SMA with 10-mm NMAS was selected for a surface wearing course in consideration to the fact that it is of common use in various European countries but only scarce studies have been conducted in the UK with locally available materials and testing protocols. The influence of the following parameters on the fracture resistance of the SMA mixture were considered: (i) Temperature [6]; (ii) Aggregate type [7], and (iii) Environmental conditioning [6, 7]. This conditioning comprised combined effects of moisture damage and ageing through the use of Saturation Ageing Tensile Stiffness (SATS) protocol (2.1MPa, 85 °C, and 24 h conditioning, BS EN 12697-45:2012). The SMA mixtures were produced following the BS EN 13108-5:2016/PD 6691:2007 [8]. The specimens produced from a roller compacted slab were cored and trimmed to obtain homogeneous cylindrical specimens (100-mm diameter and 40-mm height). A total of at least three true replicates were included per HL content and testing temperature. Table 1 shows mixture characteristics and parameters considered.
2.2 Fracture Determination The fracture strength was determined using the SCB approach, following the BS EN 12697–44 Standard Method [9]. Equation 1 shows the determination of the maximum stress that the sample underwent with the addition of a geometry factor. Equation 2
938
A. M. Ahmed et al.
Fig. 1 a Actual testing rig. b Schematics of the testing rig. c Schematics of the required sample dimensions
shows the process of achieving the fracture toughness based on geometric features and the maximum stress given by Eq. 1. Previous research has shown this method is rigorous and provides with sufficient precision and good repeatability for fracture characterisation [10]. Figure 1 shows the schematics of the SCB test procedure. 4.263 × Fmax D×t α 2 α − 799.94 = σmax · (−4.995 + 155.58 W W α 3 α 4 α 5 + 2141.9 − 2709.1 + 1398.6 W W W σmax =
KIC
(1)
(2)
where σmax is the maximum stress, Fmax is the maximum force applied (Newtons), D is the sample diameter (mm) and t is the thickness (mm). Where, K IC is the fracture toughness, Wα is the geometric factor, α is the notch depth (mm) and W the height (mm). The specimens were kept at the desired testing temperature for a minimum of 4 h. After that, these were positioned on the frame (Fig. 1) in an Instrom machine with a controlled loading rate of 5-mm per minute. The samples were loaded until these reached failure and these were also checked after failure for area compliance following the standard method.
3 Results and Discussion: Effects of Temperature, HL Content, Aggregate Type and Conditioning on the Fracture Toughness A comparison of the temperature effects on the fracture toughness is presented in Figs. 2, 3 and 4. Altogether, these indicate that the increase of the SCB testing
Fracture Characterization of Stone Mastic Asphalt (SMA) …
939
Fig. 2 Fracture toughness of unaged mixtures produced with limestone aggregates and incorporated HL
temperature produced considerable reduction in the fracture toughness, equal to about 48% for all HL contents as shown in Fig. 2 for 20 °C increase. However, as HL content increased, the percent reduction decreased slightly. These conclusions are valid for both unaged and aged mixtures (Fig. 2 vs. Fig. 3), as well as for different aggregate types (Fig. 3 vs. Fig. 4). Although not shown in the results, temperature effects were also observed when the specimens were subjected to testing in terms of the time required for the sample to reach failure. As the temperature increased, longer times were required for the sample to fail as a consequence decreased stiffness of the mixtures. For these conditions, those specimens with the highest HL content reached failure quicker in contrast to the others. The effects of environmental conditioning on the fracture toughness for mixtures produced with limestone aggregates can be interpreted from Figs. 2 and 3. It is observed that environmental conditioning of the samples induced a reduction in the resistance to fracture of the mixtures (14% for 0 °C and 38% for 20 °C), which indicates that the SATS procedure could have induced detrimental effects on this mechanical response as a result of: (1) a stiffer mastic with reduced fracture resistance, and (2) the effects induced by the water exposure of the samples in confined conditions of high temperature and pressure for prolonged time. As depicted in Fig. 3, the 2% HL induced a higher level of improvement in the fracture toughness in terms of ≈6% increase at 0 °C and ≈17% increase at 20 °C. The results show that the effects are mostly seen at high temperatures as there was a 2.5× increase compared to the increase in the samples tested at 0 °C. Furthermore, the conditioned samples produced with limestone aggregates showed that 2% HL provided the most significant improvement in the fracture toughness across all testing temperatures with increases of ≈11% at 0°, and ≈24% at 20° (Fig. 4). This suggests that environmental conditioning effects on this aggregate type produced a superior
940
A. M. Ahmed et al.
Fig. 3 Fracture toughness of conditioned mixtures produced with limestone aggregates and incorporated HL
Fig. 4 Fracture toughness of conditioned mixtures produced with granite aggregates and incorporated HL
fracture performance of HL at higher testing temperatures in contrast to those of the granite aggregates. A comparison of the effects of aggregate type on the fracture toughness provides evidence on the aggregate-type-dependency of this mechanical property. In fact, as shown in Figs. 3 and 4, the fracture toughness is slightly higher for all testing temperatures and HL contents for the granite aggregate specimens. The effects of testing temperature are clear for all HL concentrations, aggregate types and environmental
Fracture Characterization of Stone Mastic Asphalt (SMA) …
941
conditioning of the specimens. Precisely, the lowest temperature showed the highest Fracture Toughness values and as temperature increased, this value was reduced significantly as a result of the lowering in the specimen stiffness modulus. In terms of testing variability, for all testing carried out the lowest test temperature (i.e., 0 °C) exhibited the highest variability as depicted in the error bars in Figs. 2, 3 and 4.
4 Conclusions This paper investigated the fracture toughness of SMA mixtures with incorporated HL. The major conclusions relate to: (i) HL has a positive influence on the fracture toughness of SMA with increases up to 24%; (ii) The addition of HL was most effective in conditioned samples with granite aggregate; (iii) Optimum HL concentration for conditioned and unaged mixtures with limestone aggregate is between 1 and 3%; (iv) the increase in HL up to 2% had an increasing effect on mixture toughness in most cases (v) HL has the most significant influence at intermediate temperature (20 °C) in contrast to low temperature (0 °C); and (vi) the effect of conditioning via SATS had an effect on the fracture toughness that was predominantly aggregate-dependent. Acknowledgements Juan S. Carvajal-Munoz thanks the support provided by Universidad del Norte and Colciencias-Colfuturo (Colombia) to conduct PhD studies at the Uni of Nottingham.
References 1. Wu, S., Xue, Y., Ye, Q., Chen, Y.: Utilization of steel slag as aggregates for stone mastic asphalt (SMA) mixtures. Build. Environ. 42(7), 2580–2585 (2007) 2. García, Á.: Self-healing of open cracks in asphalt mastic. Fuel 93, 264–272 (2012) 3. Simone, A., Lantieri, C.: Fatigue and healing properties of low environmental impact rubberized bitumen for asphalt pavement. Coatings 7(5), 66 (2017) 4. Huang, B., Shu, X., Zuo, G.: Using a notched semicircular bending fatigue test to characterize fracture resistance of asphalt mixtures. Eng. Fract. Mech. 109, 78–88 (2013) 5. Molenaar, A., Erkens, S., Scarpas, A., Liu, X.: Semi-circular bending test; simple but useful? In: Asphalt Paving Technology: Association of Asphalt Paving Technologists-Proceedings of the Technical Sessions, pp.794–815 (2002) 6. Airey, G.: Fundamental binder and practical mixture evaluation of polymer modified bituminous materials. Int. J. Pavement Eng. 5(3), 137–151 (2004) 7. Kim, K., El Hussein, M.: Variation of fracture toughness of asphalt concrete under low temperatures. Constr. Build. Mater. 11(7–8), 403–411 (1997) 8. British Standards Institution: BS EN 13108–5:2016: Bituminous Mixtures. Material Specifications. Stone Mastic Asphalt. London: BSI (2016) 9. British Standards Institute: BS EN 12697–44. Bituminous Mixtures. Testing Methods for Hot Mix Asphalt Part 44: Crack propagation by Semi-circular Bending Test. London: British Standards Institute (2010) 10. Hofman, R., Oosterbaan, B., Erkens, S., der Kooij, J.: Semi-circular bending test to assess the resistance against crack growth. In: Performance Testing and Evaluation of Bituminous Materials, pp.257–263 (2001)
Friction Properties of Asphalt Surfacings Containing Sintered Lightweight Aggregate Ignacio Artamendi, Bob Allen, Phil Sabin, and Neil Leake
Abstract Friction properties of a natural aggregate of known good polishing characteristics and a manufactured lightweight aggregate have been determined by means of the friction after polishing test. The friction of a blend of 75% w/w natural aggregate and 25% w/w lightweight aggregate was also estimated based on the friction values of the individual components and the blend ratio. Asphalt surfacings were then designed by replacing 25% w/w of the coarse natural aggregate with lightweight aggregate. Two types of thin surfacing mixtures, SMA and AC, typically used in low speed and high speed roads, were produced and evaluated. Friction characteristics of the asphalt surfacings were determined in the laboratory using the friction after polishing test. It was found that the lightweight aggregate had superior friction and polishing characteristics to that of natural aggregate. Laboratory results also indicated that thin surfacing materials containing lightweight had better friction and skid resistance than those without lightweight aggregate. Keywords Friction · Skid resistance · Lightweight aggregate · Friction after polishing · Asphalt surfacing
1 Introduction The polishing resistance of an aggregate is typically assessed in the laboratory using the polished stone value [1]. Aggregates with high PSV values are typically specified for situations where high skid resistance is required such as, approaches to junctions, roundabouts and high traffic volume roads like motorways [2]. The availability of high-PSV aggregate is, however, limited. Also, the sources of these aggregates are in many instances far from the production plant resulting in significant transport costs.
I. Artamendi (B) · B. Allen · P. Sabin · N. Leake Aggregate Industries, Hulland Ward, UK e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_120
943
944
I. Artamendi et al.
Furthermore, for high risk areas like approaches to pedestrian areas, traditional High Friction Surfaces (HFS) composed of “tough” polish-resistant, abrasionresistant aggregates like calcined bauxite bonded to the pavement surface using a resin are typically specified. Alternative aggregate sources have also been used to improve friction and skid resistance. For example, artificial lightweight aggregates like expanded clay, slate and shale have been shown to provide high levels of skid resistance when used in surface dressings or chip seals [3]. Furthermore, expanded clay lightweight aggregate has also been used in asphalt concrete mixtures as a replacement of some of the natural aggregate. It has been shown that the incorporation of expanded clay lightweight aggregate improved friction levels and reduced stopping distances [4]. The Friction after Polishing (FAP) test [5] is another laboratory test used to determine the friction properties of both aggregate and asphalt mixtures. The FAP test has been found particularly useful for predicting the friction properties of blends of aggregates of different polishing characteristics [6]. Furthermore, the test can be used to predict the in-situ skid resistance of surfacing materials [7]. In this work, the friction properties of a high PSV aggregate and a manufactured lightweight aggregate produced by the sintering process have been determined by means of the friction after polishing (FAP) test. The friction values of a blend of natural mineral aggregate and sintered lightweight aggregate (SLWA) have then been estimated theoretically based of the friction properties of the individual components and the blend ratio. This blend ratio has been used to design asphalt surfacing materials containing lightweight aggregate. Friction of the asphalt surfacing materials produced with blends of natural mineral aggregate and lightweight aggregate have then been determined in the laboratory using the friction after polishing test.
2 Materials 2.1 Aggregates In this study a natural mineral aggregate and a lightweight aggregate were used. The natural aggregate was a volcanic tuff formed by the consolidation of volcanic ash. This aggregate is highly resistant to polishing and has a declared PSV value of 68. The lightweight aggregate, on the other hand, was a manufactured aggregate produced by sintering pulverized fly ash. The fly ash is a waste legacy material from coal fired power stations for the generation of electricity. The sintered lightweight aggregate (SLWA) is made by adding controlled amounts of water to the fly ash in specially designed pelletising dish pans. The pellets formed are then heated at a temperature of around 1250 °C. The result is a hard vitrified external crust with an internal honeycombed structure of interconnecting voids within the aggregate. The particles formed are rounded in shape and range in size.
Friction Properties of Asphalt Surfacings …
945
Table 1 Aggregate properties Aggregate type
Particle density (Mg/m3 ) Oven dry
Water absorption (%) SSD
Apparent
NA
2.70
2.72
2.75
0.7
SLWA
1.49
1.71
1.90
14.7
Particle density and water absorption values for the natural aggregate (NA) and the sintered lightweight aggregate (SLWA) were determined in accordance with BS EN 1097-6 and are presented in Table 1. It can be seen that particle densities of the lightweight aggregate were considerably lower than that of the natural aggregate. Also, the high water absorption value indicates the porous nature of this type of aggregate.
2.2 Mixtures Two thin surfacing materials, SMA 10 surf 40/60 and AC 14 surf PMB, were designed and evaluated in this study. The SMA 10 is typically used in low speed roads and roundabouts whereas the AC 14 is used in high speed roads like motorways where relatively high surface texture and good noise reduction and skid resistance properties are required. First, two control asphalt surfacing materials, SMA 10 (C) and AC 14 (C) were designed in the laboratory using high PSV natural aggregate (NA). Two mixtures, namely SMA 10 (25) and AC 14 (25), were also designed by replacing 25% by weight of the coarse natural aggregate with SLWA. Single sizes 8/14 mm and 4/8 mm SLWA and 6.3/14 mm, 4/10 mm, 2/6.3 mm and 0/4 mm natural aggregate were used. Filler was also added to the mixtures at different proportions. Optimum binder content was determined using a gyratory compactor. Proportioned aggregate blends and binder were mixed at 170 °C and then compacted in 100 mm moulds using the gyratory compactor. The number of gyrations selected was 100. Three compacted specimens per mixture were manufactured. Design binder contents and, maximum density, bulk density and air voids are presented in Table 2. It can be seen that the binder contents for the mixtures containing Table 2 Mixture design parameters Property
SMA 10 (C) SMA 10 (25) AC 14 (C) AC 14 (25)
Design binder content (%)
6.3
8.8
5.1
7.6
Maximum density (Mg/m3 )
2.467
2.084
2.525
2.105
Bulk density
(Mg/m3 )
at 100 gyrations
Voids (%) at 100 gyrations
2.153 12.7
1.976 5.3
2.218 12.2
1.968 6.5
946
I. Artamendi et al.
SLWA were higher than those for the control mixture. This was to compensate for binder absorption by the porous lightweight aggregate. Results also show that both maximum density and bulk density decreased when 25% of the coarse aggregate was replaced with SLWA.
3 Friction After Polishing Test Laboratory measurements of the friction properties of the aggregates and asphalt mixtures were carried out using the Friction after Polishing (FAP) test. The FAP test is designed to carry out first accelerated polishing and then to assess their friction characteristics. The polishing process is carried out by three rubber conical rollers rotating at high speed whereas friction measurement is carried out by three small rubber slider pads attached to a rotating head. Full details of the test can be found elsewhere [6]. Aggregate specimens for FAP testing consisted of mosaics of 225 mm diameter. These specimens were prepared in a similar way to those for the PSV test. The aggregates were sieved through a 10 mm sieve and retained on a 7.2 mm grid sieve. Individual aggregate particles were then fixed in a mould using epoxy resin. Two specimens per aggregate sample were prepared for testing. Asphalt specimens, on the other hand, were prepared by taking 225 mm cores from laboratory roller compacted slabs. Two specimens per mixture were used for testing. The procedure followed to test the aggregate specimens consisted of applying different polishing cycles from 0 to 90,000, followed by friction measurements. For the asphalt specimens, the procedure develop at the Technical University of Berlin (TUB) was followed. This procedure includes grit blasting of the asphalt specimens to remove the excess bitumen. A full cycle of the test has the following stages: 1, Friction test of asphalt specimen; 2, Polish for 90,000 cycles followed by friction test; 3, Grit blasting; 4, Friction test; 5, Polish for 90,000 cycles followed by friction test; and 6, Friction test “to the limit” which is achieved when subsequent measurement of friction differ by 0.005 or less.
4 Results and Discussions 4.1 Aggregate Friction Friction values at different number of polishing cycles ranging from 0 to 90,000 for the high PSV natural aggregate (NA) and the sintered lightweight aggregate (SLWA) are presented in Fig. 1. It can be seen that, as expected, the friction decreased with increasing number of polishing passes. Also, the friction values for the SLWA were
Friction Properties of Asphalt Surfacings …
947
1.00 0.90
Fric on coefficient
0.80 0.70 0.60 0.50 0.40 SLWA
0.30
NA (PSV68)
0.20
75% NA - 25% SLWA 0.10 0.00
0
20,000
40,000
60,000
80,000
100,000
No. of passes
Fig. 1 Evolution of friction of the natural aggregate, sintered lightweight aggregate and a blend of the two aggregates
considerably higher than those for the natural aggregate (NA). This suggests better in-situ skid resistance performance of the SLWA compared to the NA. The friction characteristics of blends of aggregates can be predicted theoretically based on the simple relationship proposed by Dunford [6]. Thus, the friction of a blend of coarse aggregates from two sources, A and B, can be predicted using Eq. 1 [7]: FAPAB = mA × FAPA + mB × FAPB
(1)
where mA and mB are the proportions (ratio) by mass and, FAPA and FAPB are the friction values of the individual components. Predicted friction values for a blend of 75% by weight of natural aggregate and 25% by weight SLWA at different polishing cycles determined using Eq. 1 are presented in Fig. 1. It can be seen that the friction values increased as the as the proportion of SLWA increased. Also, friction value of this particular blend at 90,000 polishing cycles was 0.54 which is rarely achieved using natural aggregates only.
4.2 Asphalt Friction Friction test results for the asphalt surfacing materials are presented in Fig. 2. It can be seen first that, the initial friction for all the materials was relatively low as the aggregates particles were still fully coated with bitumen. Also, a slightly better initial friction was found for the mixtures containing lightweight aggregate.
948
I. Artamendi et al. 0.70 SMA 10 (C) SMA 10 (25) AC 14 (C) AC 14 (25)
Fric on coefficient
0.60
0.50
0.40
0.30
0.20
0 Passes
90000 Passes
Before grit blasƟng
0 Passes
90000 Passes
Limit
AŌer grit blasƟng
Fig. 2 Friction of asphalt surfacing materials
Friction after the first polishing stage increased as some of the aggregate particles became exposed. Also, friction value for the SMA 10 (C) was higher than that for SMA 10 (25) which may suggest that the rate of bitumen removal for this material was higher than for the others. Friction values for the AC 14 materials, on the other hand, were practically the same suggesting good adhesion of the PMB to both types of aggregates. Figure 2 also shows that the friction increased considerably after grit blasting as more bitumen was removed from the aggregate surface. Furthermore, friction values of the mixtures containing lightweight aggregate were markedly higher than those with only natural aggregate. Results also indicated better friction for the AC 14 mixtures compared to the SMA materials. These differences could be attributed to differences in surface macrotexture, the AC 14 having higher macrotexture than the SMA. Friction values after the second polishing stage decreased noticeably as the exposed aggregate particles became polished during the polishing process. Furthermore, the mixtures with the lightweight aggregate retained the highest levels of friction, particularly the AC 14. Also, experience with the equipment in Germany has suggested that the final level of friction, after both polishing stages have been carried out, simulates the state of skid resistance that occurs after four to six years of traffic.
5 Conclusions • Friction of the sintered lightweight aggregate was considerably higher than that of a high PSV natural aggregate.
Friction Properties of Asphalt Surfacings …
949
• Thin surfacing mixtures containing 25% w/w lightweight aggregate showed better friction than those mixtures with natural aggregate only. This suggests enhanced in-situ skid resistance for mixtures containing lightweight aggregate. • The AC 14 surf PMB with 25% w/w lightweight aggregate showed higher friction than the SMA 10 surf 40/60 with 25% w/w lightweight aggregate. This could be attributed to differences in macrotexture between the two asphalt surfaces. • Asphalt surfacings containing high PSV aggregate and SLWA could be used in areas reserved only to traditional high friction surfaces (HFS).
References 1. EN 1097–8: 2009. Test for mechanical and physical properties of aggregates. Determination of the polished stone value 2. CD 236 Surface course material for construction, Design Manual for Road and Bridges, Volume 7 Pavement Design and Maintenance, Section 5 Surfacing and Surfacing Materials, Part 1, London, The Stationary Office (2019) 3. Martin, B.: Ceramic aggregates cuts Kansas roads costs, Better Roads (2003) 4. Losa, M., Leandri, P., et al.: Mechanical and performance-related properties of asphalt mixes containing expanded clay aggregate. Transport. Res. Record J. Transport. Res. Board 22–30 (2008) 5. EN 12697-49: 2014: Bituminous mixtures. Test methods for hot mix asphalt. Determination of friction after polishing 6. Dunford, A.: Use of the Wehner-Schulze Machine to Explore Better Use of Aggregates with Low Polish Resistance 2: Experiments Using the Wehner-Schulze Machine, Published Project Report PPR605. Transport Research Laboratory, Wokingham, UK (2012) 7. Huschek, S.: Experience with skid resistance prediction based on traffic simulation. In: Proceedings of the 5th International Symposium of Pavement Surface Characteristics, 6–10 June, Toronto, Canada (2004)
From Laboratory Mixes to Full Scale Test: Rutting Evaluation of Bio-recycled Asphalt Mixes Juliette Blanc, Emmanuel Chailleux, Pierre Hornych, Chris Williams, Zahra Sotoodeh-Nia, Laurent Porot, Simon Pouget, François Olard, Jean-Pascal Planche, Davide Lo Presti, and Ana Jimenez del Barco Abstract The present paper describes the rutting behavior of innovative mixes incorporating 50% of Reclaimed Asphalt (RA) with bio-materials. They were assessed in the laboratory and in a full-scale accelerated experiment. The innovative mixes studied here contained bio-materials especially designed to help recycling by reactivating the aged binder from RA. Four mixes were evaluated: three of them are manufactured with bio-materials, (two bio-rejuvenators and one bio-binder) and one was a control mix, which was a high modulus asphalt mix (EME2). In this study, the rutting resistance of the four mixes was first evaluated in the laboratory with both European and US methods. The full-scale test was then performed in order to evaluate the rutting resistance of the bio-recycled asphalt mixes under heavy traffic (200,000 load cycles loaded at 65 kN) and compared with the control one. A simplified analysis leads to the conclusion that, with the Nantes climate, a daily traffic of 150 heavy vehicles per day applied over 20 years corresponds to approximately 200,000 heavy vehicle loads when the surface temperature exceeds 30 °C. Therefore, it can be considered that the rutting evaluation made on the carrousel represented almost 20 years of traffic during hot-day periods. The results obtained on the test track were consistent with the laboratory rutting tests showing good performance for J. Blanc (B) · E. Chailleux · P. Hornych Université Gustave Eiffel, Campus de Nantes, France e-mail: [email protected] C. Williams · Z. Sotoodeh-Nia Iowa State University, Ames, USA L. Porot KRATON Chemical B.V. A Subsidiary of Kraton Corporation, Almere, The Netherlands S. Pouget · F. Olard EIFFAGE Infrastructures, Paris, France J.-P. Planche Western Research Institute, Laramie, USA D. Lo Presti Universita Degli Studi Di Palermo, Palermo, Italy A. J. del Barco University of Granada, Granada, Spain © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_121
951
952
J. Blanc et al.
all the mixes. The materials presenting the best performance on the test track also presented the best performance in the laboratory. Keywords Bio-materials · Reclaimed asphalt · Accelerated pavement test · Rutting test · Laboratory test
1 Tested Materials The aim of the project was to maximise the use of RA with bio-based materials. For this purpose, three mixes were designed incorporating 50% RA content, and using three different and complementary innovative bio-materials [1]: • Mix1: designed with a bio-based rejuvenator, SYLVAROAD™ RP1000, a performance Additive, from Kraton Chemical used to treat the RA. It was specifically designed to increase RA content up to 100% or to reuse very hard, low quality RA. • Mix2: designed with a bio-binder, Biophalt® , from Eiffage Infrastructure, used for total replacement of bituminous binder in recycling techniques. • Mix3: designed with a bio-based additive from Iowa State University, an Epoxidized Methyl Soyate (EMS) aimed to compatibilise virgin binder and aged binder from RA. Aggregate grading curve and binder content were chosen using aggregate packing optimisation concepts (GB5® type material) in order to maximize mix density and particle interlock [2]. The properties of these three GB5-type mixes (Mix 1, Mix2 and Mix3) were compared with the reference high modulus asphalt mix (EME2), commonly used in France. Before construction, laboratory rutting tests were performed on mixes produced in lab following the European method (French wheel tracking test) and US method (flow number test). Results are presented in Table 1. All three innovative materials Table 1 Results of laboratory tests for resistance to rutting Mixes
WTT EU method EN 12697–22 + A1 30,000 cycles at 60 °C Rut Depth (%)
Rutting resistance AASHTO TP-79 (flow number) at 7% air void, T = 54 °C (cycles)
Requirements for a high modulus Mix
6% polymer content) may offer additional advantages in flexible pavements subjected to heavy and slow-moving traffic loads. The objective of this effort was to evaluate the structural response of an HP AC mixture using full-scale pavement testing. Two experiments were conducted: Experiment 1 consisted of a PMA AC layer on top of a crushed aggregate base (CAB) and a subgrade (SG), and Experiment 2 consisted of an HP AC layer with a reduced thickness on top of the same CAB and SG. Both pavements were subjected to dynamic loads applied at the pavement surface simulating falling weight deflectometer (FWD) loading conditions. Pavement surface deflections along with critical pavement responses at different depths and locations within the pavement layers were monitored during testing through embedded instrumentations. Field-produced mixtures were sampled during construction and evaluated for performance properties. At the end of each experiment, cores were cut from the AC layer for in-place density measurements. In general, at the centerline of the load, the reduced thickness of the HP AC layer resulted in higher vertical surface deflections and vertical stresses at the middle of the CAB layer and at 15.2 cm below the surface of the SG layer. The mechanistic analysis of the built pavement structures showed that the HP pavement would result in better AC fatigue and rutting performance, higher rutting in the unbound layers, but similar total pavement rutting. Keywords Polymer modified binder · Pavement instrumentation · Pressure cells · Dynamic loads · Mechanistic analysis J. Habbouche Virginia Transportation Research Council, Charlottesville, VA, USA e-mail: [email protected] P. E. Sebaaly (B) · E. Y. Hajj · M. Piratheepan University of Nevada, Reno, NV, USA e-mail: [email protected] E. Y. Hajj e-mail: [email protected] M. Piratheepan e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_122
959
960
J. Habbouche et al.
1 Introduction Although polymer-modified asphalt (PMA) mixtures, with 2–3% polymer content, have historically shown improved long-term performance, it is also believed that asphalt concrete (AC) mixtures with high polymer (HP) content (i.e., >6% polymer content) may offer additional advantages in flexible pavements subjected to heavy and slow-moving traffic loads. An extensive review of the literature was conducted to compile all information and findings related to the performance of HP asphalt binders and mixtures [1–3]. While several previous studies have highlighted the positive impacts of HP modification on asphalt binders and mixtures, there is still a lack of understanding for the structural value of an HP AC mix. The authors conducted a study to investigate the impact of HP modification on performance characteristics of asphalt mixtures. The main objective of this work was to design and evaluate the engineering properties and performance characteristics of conventional PMA and HP AC mixtures manufactured using locally available material in Florida [1, 2, 4]. The laboratory evaluation showed similar or less stiff engineering properties (i.e., dynamic modulus [E*]) and better performance of HP AC mixtures in terms of fatigue, rutting, top-down cracking, and reflective cracking when compared with those of the corresponding PMA control AC mixtures. Furthermore, the true impact of these improvements was evaluated through mechanistic analysis of flexible pavements incorporating the two types of mixtures (i.e., PMA and HP) [1, 2] using the 3D-MOVE analysis software [5]. Despite all these benefits and improvements, there is still a need to assess the impact of HP modification on the structural capacity of an AC mixture using full-scale pavement testing prior to implementation in the field. Therefore, the objective of this paper was to evaluate the structural capacity of an HP AC mixture using full-scale pavement testing.
2 Experimental Program Two experiments were conducted at the University of Nevada Reno (UNR) full-scale pavement testing facility using the PaveBox. Experiment 1 (i.e., PaveBox_PMA) consisted of a PMA AC layer, 10.8 cm (4.25 in.) thick, on top of 22.9 cm (9.0 in.) crushed aggregate base (CAB) and 155.0 cm (61.0 in.) subgrade (SG). Experiment 2 (i.e., PaveBox_HP) consisted of an HP AC layer with a reduced thickness of 8.9 cm (3.5 in.) on top of the same CAB and SG.
2.1 Characteristics and Properties of Employed Materials The SG material in the PaveBox experiments was classified as A-2-7 and clayey sand with gravel (group symbol: SC) according to the American Association of
Full-Scale Pavement Testing of a High Polymer Modified …
961
State Highways and Transportation Officials (AAHSTO) system and the Unified Soil Classification System (USCS), respectively. A maximum dry density (γ d,max ) of 2010 kg/m3 (125.5 pcf) was achieved at an optimum moisture content (W opt ) of 11.8% in accordance with AASHTO T99 [6]. A typical local CAB layer was sampled and used in the full-scale PaveBox experiments. Similarly, a γ d,max of 2164 kg/m3 (135.1 pcf) was achieved at W opt of 8.8% based on the laboratory-determined compaction curve. Two asphalt binders were used in this effort; a PG76-22 PMA and an HP binder. The PMA and HP binders had 3.0% and 8.0% SBS polymer by weight of binder and a percent recovery at 3.2 kPa (%R3.2 ) of 54.3% and 97.5%, respectively. The same aggregate gradation with a nominal maximum aggregate size (NMAS) of 12.5 mm (0.5 in.) was targeted for both experiments. Two AC mixtures were designed following the Superpave mix design methodology; the mixtures were compacted using the Superpave gyratory compactor (SGC) for 100 gyrations based on the targeted traffic level D. The mixtures for the PaveBox experiments were produced at the mix design optimum binder contents (OBCs): 5.6% for PaveBox_PMA and 5.7% for PaveBox_HP. Loose mixtures were collected during placement, and were evaluated for their engineering property in terms of E* and phase angle (δ) (AASHTO T378 [6]), and for performance characteristics in terms of their resistance to fatigue cracking (AASHTO T321 [6]) and rutting. The E* and rutting were evaluated for reheated loose mixtures, while fatigue cracking was evaluated on long-term oven-aged mixtures (AASHTO R30 [6]). Test specimens were compacted to the as-constructed air voids of the AC layer in the PaveBox. Overall, the asphalt binder type (i.e., PMA or HP) had an impact on the magnitude of E* and δ. Lower E* and δ values were observed for the PaveBox_HP mix at intermediate frequencies indicating a more flexible behavior under traffic loading. Moreover, better fatigue and rutting performance characteristics were observed for the PaveBox_HP AC mix when compared with those of the PaveBox_PMA AC mix which can be credited to the HP modification of the asphalt binder (refer to Fig. 1 below).
2.2 Construction, Instrumentation, and Loading Protocols of Experiments The PaveBox consists of a square box with internal dimensions of 315 by 315 by 183 cm (124 by 124 by 72 in.). The box is braced at two levels with steel beams and tension rods to act as lateral bracing system. Figure 2a shows a three-dimensional (3D) schematic of the PaveBox. The SG material was shoveled from the stockpile and mixed in a mixer for less than a minute. The moist SG material was then transported and placed within the PaveBox area. A vibratory plate compactor was used to achieve the required in-place compaction. Similar procedure was followed to construct the CAB layer with higher required compaction effort.
962
J. Habbouche et al. 10,000 Flexural Microstrain @77°F (25°C)
Fig. 1 Performance characteristics of PaveBox_PMA and PaveBox_HP mixes: a fatigue relationship at 77 °F (25 °C) and b rutting relationship at 122 °F (50 °C)
PaveBox_PMA PaveBox_HP
1,000
100 1.E+04 1.E+05 1.E+06 1.E+07 1.E+08 1.E+09 Number of Cycles to failure
(a) εp/εr @ 122°F (50°C)
100
10 PaveBox_PMA PaveBox_HP
1 10
(a)
100 1000 10000 Number of Loading Cycles
100000
(b)
Fig. 2 a 3D schematic of the PaveBox, and b on-site half-ton asphalt mixer used to mix and produce PMA and HP AC mixes for PaveBox
PMA and HP asphalt mixes for the PaveBox experiments were mixed in a half-ton asphalt mixer (refer to Fig. 2b). The aggregates were proportioned out of each bin onto a feeder belt according to the mix design percentages and transported to the mixing pug mill. The aggregates were heated in the pug mill at the mixing temperature for a duration of 15 min: temperatures of 163 °C (325 °F) and 175 °C (340 °F) were used for the PMA and HP AC mixes, respectively. The heated liquid asphalt binder was added into the pug mill on top of the hot aggregates. The mixing process continued for an additional duration of 15 min to ensure uniformity and proper coating of aggregates
Full-Scale Pavement Testing of a High Polymer Modified …
963
Fig. 3 Section view for PaveBox_PMA experiment (No. 1) with all instrumentation
within the AC mix. The produced AC mix was dumped and spread uniformly over the entire PaveBox area, leveled to a thickness of approximately 63.5 mm (2.5 in.) of un-compacted material, and then compacted to achieve a thickness of 25–38 mm (1.0 to 1.5 in.) per lift and a target in-place density of 92–96%. Both pavements were subjected to dynamic loads applied at the pavement surface simulating falling weight deflectometer (FWD) loading conditions. Pavement surface deflections along with critical pavement responses at different depths and locations within the pavement layers were monitored during testing through embedded instrumentations. Linear variable differential transformers (LVDTs) were used to record pavement surface deflections while total earth pressure cells (TEPC) were used to capture load-induced stresses in the CAB and SG layers. Both pavement structures were subjected to a series of four loading levels: 26.7; 40.0; 53.4; and 71.2 kN (6000; 9000; 12,000; and 16,000 lb). Twenty-five cycles, each consisting of a pulse duration of 0.1 s followed by a rest period of 0.9 s, were applied at each incremental dynamic load. Figure 3 shows a sketch of the PMA pavement structure with the installed instruments.
3 Analysis of Measured Pavement Responses Preprocessing steps were undertaken for all recordings to identify and separate the appropriate load-induced response signals from the recorded data. While both pavement structures were fully instrumented, only the recordings under the center of the load are summarized in this paper: the LVDT at the surface of AC layer (L0); strain gauge at the bottom of AC layer (S1); pressure cell at the middle of the CAB layer (P7); and pressure cell at 6 inch from the surface of the SG layer (P3). Figure 4a shows the measured vertical surface deflections in the PaveBox_PMA
964
J. Habbouche et al.
10
480
L0 - PaveBox_PMA L0 - PaveBox_HP
Tensile Strain at Bottom of AC Layer (microstrain))
Measured Vertical Surface Deflection (mm*10^-1)
12
8 6 4 2 0 20
40 60 Applied Load (kN)
420
S1 - PaveBox_PMA S1 - PaveBox_HP
360 300 240 180 120 60 0 20
80
40
60
80
Applied Load (kN)
(a)
(b)
Fig. 4 Measurements at the center of the loading plate: a vertical surface deflections; and b horizontal tensile strains at the bottom of the AC layer
and PaveBox_HP experiments as a function of the surface load levels. Regardless of the applied load level, a higher vertical surface deflection at the middle of the loading plate (i.e., L0) was observed in the case of the HP AC layer when compared with the PMA AC layer. This is demonstrated with vertical surface deflection measurements in the PaveBox_HP that are 22–76% higher than those observed in the PaveBox_PMA. Figure 4b shows the measured tensile strains at the bottom of AC layer in the PaveBox_PMA and PaveBox_HP experiments as a function of surface load levels. In both experiments, an increase in the tensile strain was observed with the increase in the applied surface load level. The tensile strain measurements in the PaveBox_PMA experiment were higher than or similar to the respective measurements in the PaveBox_HP experiment. The TEPC measurements for the vertical stresses in the PaveBox_PMA and PaveBox_HP experiments as a function of surface load levels at the center of the loading plate in the middle of the CAB layer (P7) and 15.2 cm (6 in.) under the surface of SG layer (P3) are presented in Fig. 5. Regardless of the surface loading
400
160
P7 - PaveBox_PMA P7 - PaveBox_HP
Measured Vertical Stress (kPa)
Measured Vertical Stress (kPa)
500
300 200 100 0 20
40
60
Applied Load (kN)
(a)
80
140 120
P3 - PaveBox_PMA P3 - PaveBox_HP
100 80 60 40 20 0 20
40
60
80
Applied Load (kN)
(b)
Fig. 5 Measured vertical stresses at the center of the loading plate: a at the middle of the CAB layer (P7); and b at 6 in. below the top of the SG layer (P3)
Full-Scale Pavement Testing of a High Polymer Modified …
965
level, higher vertical stresses in the middle of the CAB (i.e., P7) were observed in the PaveBox_HP experiment when compared with the PaveBox_PMA experiment. This is demonstrated with vertical stress measurements in the PaveBox_HP experiment that are 85–100% higher than those observed in the PaveBox_PMA experiment. Similarly, regardless of the loading level, slightly higher vertical stresses under the middle of the loading plate was observed at a distance of 15.2 cm (6 in.) below the top of the SG layer in the PaveBox_HP experiment when compared with the PaveBox_PMA experiment (P3). This is demonstrated with vertical stress measurements in the PaveBox_HP experiment that are 43–46% higher than those observed in the PaveBox_PMA experiment. Overall, at the centerline of the load, the reduced thickness of the HP AC layer resulted in higher vertical surface deflections and vertical stresses at the middle of the CAB layer and at 6 in. below the surface of the SG layer.
4 Closing Remarks A mechanistic-empirical (ME) analysis of the built pavement structures was conducted using the backcalculated layers’ moduli in conjunction with the laboratory-developed fatigue and rutting performance models for the PMA and HP AC mixes. The ME analysis resulted in a better fatigue and rutting performance for the HP AC layer when compared with the PMA AC layer. Higher rut depths were observed in the unbound layers of the HP pavement structure, especially in the CAB layer. However, similar total rut depths were determined for the PMA and HP pavement structures. In general, the overall results of both full-scale experiments and ME analyses showed a better structural capacity for HP AC mixes when appropriately designed in a pavement structure. However, the reduced HP AC layer may result in higher responses under the middle of the surface load, thus leading to an increase in deformation in the unbound layers.
References 1. Habbouche, J., Hajj, E.Y., Sebaaly, P.E.: Structural coefficient of high polymer modified asphalt mixes. Project BE321. Florida Department of Transportation, Tallahassee, https://www.fdot. gov/research/, July 2019 2. Habbouche, J.: Structural coefficients of high polymer modified asphalt mixes based on mechanistic-empirical analyses and full-scale pavement testing. Doctoral dissertation, University of Nevada, Reno (2019) 3. Habbouche, J., Hajj, E.Y., Sebaaly, P.E., Piratheepan, M.: A Critical review of high polymermodified asphalt binders and mixtures. Int. J. Pavement Eng. (2018). https://doi.org/10.1080/ 10298436.2018.1503273 4. Habbouche, J., Hajj, E.Y., Piratheepan, M., Sebaaly, P.E.: Impact of high polymer modification on performance characteristics of asphalt mixtures. Association of Asphalt Pavement Technologists, Lino Lakes, Minn., (2019) (in Press)
966
J. Habbouche et al.
5. Siddharthan, R.V., Hajj, E.Y., Sebaaly, P.E., Nitharsan, R.: Formulation and application of 3Dmove: a dynamic pavement analysis program. Publication FHWA-RD-WRSC-UNR-201506. University of Nevada, Reno (2015) 6. AASHTO: Standards and specifications. American Association of State Highway and Transportation Officials, Washington, D.C.
GLOBE: An Innovative Technical Solution to Ensure Waste Free Cold Logistic of Bituminous Binders Mouhamad Mouazen, Yvong Hung, Olivier Moglia, Pauline Anaclet, Alice Ngo, Serge Krafft, Flavien Geisler, and Vincent Gaudefroy
Abstract GLOBE is a French acronym for “Granulés pour la LOgistique des Bitumes d’Enrobage” literally meaning «Bituminous granulates for the logistic of coating binders». This project supported by ADEME (French Environment & Energy Management Agency) focuses on the development of an innovative technical solution to ensure a logistic of bituminous binders from refineries to asphalt mix plants which is cold, waste free and thus cleaner and safer. The main challenge is to modify a material such as bitumen, usually handled hot in liquid form, in order to be able to produce a granular form of it, which stays stable over time. In order to do that, it is required to overcome in-depth its typical binder characteristics, especially its creep behavior and its exceptional adhesiveness properties. This implies to modify the rheology of the binder and to take into consideration the granulation technology. The final product should avoid agglomeration phenomena, while taking into account the mechanical and thermal stresses associated with the handling, storage and transport of the pellets. Keywords Bitumen · Pellets · Rheological properties · Granulation technology
M. Mouazen (B) · O. Moglia TOTAL M&S, Bitumen Direction, 562 avenue du parc de L’Ile, Nanterre, France e-mail: [email protected] Y. Hung · P. Anaclet Centre de Recherche de Solaize, TOTAL M&S, Chemin du canal, 69360 Solaize, France A. Ngo · S. Krafft · F. Geisler EIFFAGE Infrastructures, 8 rue du Dauphiné, 69964 Corbas, France V. Gaudefroy Allée des Ponts et Chaussées, Université Gustave Eiffel, 44344 Bouguenais, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_123
967
968
M. Mouazen et al.
1 Scope of the Study 1.1 Globe Solid Bitumen Approach: Benefits and Challenges Toward Current Ecosystem Most of bitumen material is used in building and road pavement application. Bitumen is generally in the form of a black material which is highly viscous or even solid-like at ambient temperature, and which liquefies when heated. But from a rheological point of view, common paving grade bitumen is still considered as a viscoelastic liquid above ambient temperature (temperature crossover point). Generally, bitumen is stored and transported under hot conditions, in bulk, in tanker trucks or by boats at high temperatures of about 120–160 °C. However, the storage and transportation of bitumen under hot conditions has some drawbacks. Firstly, the transportation of bitumen under hot conditions in liquid form is considered dangerous and is highly restricted from a regulatory point of view. This mode of transportation does not present any major difficulties when the transportation equipment and infrastructure are in good conditions. If this is not the case, it can become problematic; bitumen delivery distances are therefore limited. Secondly, maintaining bitumen at high temperatures in tanks or in tanker trucks consumes energy. Lastly, maintaining bitumen at high temperatures for long periods can affect the properties of the bitumen and thus change the final performance levels of the asphalt mix. In order to overcome the problems of transporting and storing bitumen under hot conditions, packaged solutions for bitumen in solid have been developed. This mode of transportation of bitumen in such conditions represents only a minimal fraction of the amounts transported throughout the world but is essential for some geographic regions which are difficult and expensive to access using conventional transportation means. The aim of the present project is to develop a solid form of bitumen, which can be transported, handled and stored even at high ambient temperatures, especially at a temperature above 50° C commonly reached in containers in hot countries in order to set up a sustainable ecosystem based on handling solid bitumen under wider conditions as an opportunity for improving road pavement industry.
1.2 The Technical Challenges To overcome the technical challenge of shaping a material such as bitumen, into a granular, solid form, stable over times, it is necessary to carry out in-depth modifications of its characteristics. However, the granular shape is desired only during the time of storage and transport of bitumen, as the end use remains the same, namely: the cohesion of the granular skeleton and the transmission of mechanical stresses of the asphalt on the road. Three scientific and technological barriers are identified with different levels of difficulties: shaping bitumen granules, temporary removal of tackiness of the bitumen and recovery of the binder properties during the use.
GLOBE: An Innovative Technical Solution …
969
1.3 Pellet Design The designing of bitumen granules can be summarized in a granulation step followed by a step of encapsulation / filming to ultimately result in a core/shell system. Modification of core to develop a “solid” bitumen. According to the literature, there are several ways to solidify a bitumen matrix [1–4]. Among them, the use of gelling additive at relatively low concentrations ( 1.5, asphalt mix is highly resistant to moisture damage in field; HDI < 0.5, is highly susceptible to moisture damage, and the asphalt mix resistance to moisture damage is medium when 0.5 ≤ HDI ≥ 1.5 [4].
3 Materials and Methods This research presents partial results of the RAP fractions analysis, SARA and AFM, for three samples from Colombia: (A) Recovered(CO), (B) Recovered (CO), (C) Recovered (CO) and one sample from the USA, (Z) Recovered (US). The adhesion, cohesion and contact angle using the Sessile Drop Test to find the Surface Free Energy [5, 6], were analyzed for three original binders: (X)PG 64–22(US), (Y)PG 58–28(US) and 60–70 dmm(CO), and one recovered binder (Z) Recovered (US). The aggregates chosen were: Silica (US), Limestone (US), Sandstone (Coello Quarry) (CO), Sandstone (Alto La Laguna Quarry) (CO) and Limestone (Coello River) (CO). To determine the HDI, it was necessary to find the work of adhesion wet ). For binder-aggregate (W AS ) and the work of debonding binder-aggregate (W ASW each test, 15 measurements were made per type of binder and aggregate, to have sufficient repeatability. The test plan is observed in (Fig. 2). X(PG64-22) Y(PG58-28) Binders
BBS Contact Angle Pen 60-70dmm (CO) (Z) Recovered RAP (US) TSFE
SARA (A) Recovered RAP(CO) AFM (B) Recovered RAP(CO) (C) Recovered RAP(CO)
Fig. 2 Test plan
HDI
Aggregates
Silica(US) Limestone(US) Sandstone(Coello Quarry)(CO) Sandstone(Alto la Laguna)(CO)
BBS Contact HDI Angle TSFE
Limestone (Coello River) (CO) (CO):Colombia, (US) The USA, Recovered:it cames from RAP
Humidity Damage Index (HDI) of Recovered Asphalt …
1003
4 Results and Analysis Recovered binders from Colombia, the asphaltenes content was greater than 70%, while for the Binder Pen 60–70 dmm (original) it was less than 20%. Binder (Z) Recovered (US) had an asphaltene content of 32.56%. The high content of asphaltenes and low content of maltenes is related to the rigidity of the material, so that the only binders that have an adequate fractions composition, that is to say that they have a balance between elasticity and stiffness, are the Pen 60-70dmm and the (Z) Recovered (US). Table 1 gives a summary of the SARA results [7]. The AFM analysis of the Pen 60–70 dmm binder shows a defined bee structure and dispersion of the asphalt components. This bee structure is attributed to the interaction between the crystallization of paraffin waxes and the remaining asphalt fractions [8]. The surface topography of this asphalt has a height difference of 8 μm between the lowest and the highest part, (Fig. 3a). The (Z) recovered RAP (US) binder shows adequate colloidal dispersion of the fractions. This can possibly be Table 1 SARA results Binder
Asphaltenes (%)
Maltenes (%)
Pen 60–70 dmm (CO)
13.59
86.41
(Z) Recovered RAP (US)
32.56
67.44
(A) Recovered RAP (CO)
84.70
15.30
(B) Recovered RAP (CO)
67.31
32.69
(C) Recovered RAP (CO)
67.53
32.47
Fig. 3 AFM image of the phases present on the surface of each asphalt binder at 3 μm scale
1004
A. Figueroa and M. Solaimanian
Fig. 4 a Aggregates TSFE, b Binders TSFE and c Humidity Damage Index-HDI
attributed to the mixing technology applied. The homogeneity on the surface of this binder was verified by the small difference in height (4 μm) between the highest and the lowest phase, as well as in the smaller range of colors of the image scale. Little evidence of bee structure in this sample would indicate a low crystallization of the paraffinic components. The microstructure of this binder suggests low stiffness and good elastic recovery (see Fig. 3b). For the asphalt (A) Recovered RAP (CO), (see Fig. 3c), a rigid matrix (periphery) was observed that surrounds the almost imperceptible structure of bees (catanaphase) and the paraphase corresponding to the solvent (soft matrix). These phases coincide with: the interstitial and dispersed phases, and with the matrix respectively found with AFM in RAP asphalts by [9]. The results of Surface Free Energy of aggregates and binders are shown in (Fig. 4a, b) respectively. The higher aggregate TSFE was the Silica (US). The best Colombian aggregate was the limestone from Coello River. The TSFE of the recovery RAP (US) was higher than the original binders. The lowest results of TSFE were to Limestone (US) and the Colombian binder whose penetration is 60–70 dmm. The HDI for different combinations of aggregates and binders, are showed in (Fig. 4c).
5 Conclusions Results of the HDI for Z binder indicate its high resistant to moisture damage in the field with some aggregates. This binder shows remarkably good adhesion with SDM. Both, SUS and LUS aggregates have better HDI with X binder than Y binder. The most susceptible combination was found for the LUS aggregates with Z binder and P binder. However, the HDI of this binder with sandstone (CO) and limestone (CO) was found to be medium resistance to moisture damage. This observation indicates the important influence of aggregate mineralogy on HDI. The AFM analysis of recovered Colombian binders, indicated poor quality (i.e. lower HDI) compared to the Z binder. The observed difference could be the result of binder chemistry or simply the fact that the Z binder was softer than the other RAP binders used in this study. It is
Humidity Damage Index (HDI) of Recovered Asphalt …
1005
also remarkable that the P binder, one of the most applied in Colombia, has a high susceptibility to moisture damage in field, combined with SUS, LUS, SDM, SDA and LCO aggregates. Acknowledgments Universidad de La Salle, Larson Transportation Institute, Department of Civil & Environmental Engineering. PSU and LANAME laboratory, Costa Rica.
References 1. Noferini, L., Simone, A., Sangiorgi, C., Mazzotta, F.: Investigation on performances of asphalt mixtures made with reclaimed Asphalt pavement: effects of interaction between virgin and RAP bitumen. Int. J. Pavement Res. Technol. 10(4), 322–332 (2017) 2. FHWA: Reclaimed Asphalt pavement in Asphalt mixtures: state of the practice. Rep. No. FHWAHRT-11-021, no. FHWA, p. McLean, Virginia (2011) 3. Kennedy, T.W., Tam, W.O., Solaimanian, M.: Optimizing Use of Reclaimed Asphalt Pavement with the Superpave System (1998) 4. Development of Methods to Quantify Bitumen-Aggregate Adhesion and Loss of Adhesion Due to Water. Texas A&M University (2006) 5. AASHTO Designation: T 361-161: Standard Method of Test for Determining Asphalt Binder Bond Strength by Means of the Binder Bond Strength (BBS) Test (2016) (Online). Available: https://www.techstreet.com/ans/standards/aashto-t-361-16?product_id=1924305#jumps 6. Figueroa, A., Velasquez, R., Reyes, F., Bahia, H.: Effect of Water Conditioning for Extended Periods on the Properties of Asphalt Binders, no. 2372 (2013) 7. ASTM D4124 - 09(2018) Standard Test Method for Separation of Asphalt into Four Fractions (Online). Available: https://www.astm.org/Standards/D4124.htm. Accessed: 30-Oct-2019 8. Kim, H.H., Mazumder, M., Torres, A., Lee, S.-J., Lee, M.-S.: Characterization of CRM binders with wax additives using an Atomic Force Microscopy (AFM) and an optical microscopy. Adv. Civ. Eng. Mater. 6(1), 20160071 (2017) 9. Rashid, F., Hossain, Z., Bhasin, A.: Nanomechanistic properties of reclaimed asphalt pavement modified asphalt binders using an atomic force microscope. Int. J. Pavement Eng. 20(3), 357–365 (2019)
Impact of Different Fillers on the Properties of Asphalt Mixtures Adriano Teixeira and Didier Lesueur
Abstract Different mineral fillers from various sources in Portugal were compared as such, in mastics and in asphalt mixtures. The materials included recovered mineral fillers of varying petrographical origin (limestone, granite or rhyolite) obtained from the dedusting system of local asphalt plants, and commercial fillers, active or not, such as limestone filler, hydrated lime, natural hydraulic lime and Portland cement. All fillers were compared using standard filler tests, including dry porosity (Rigden air voids—EN 1097-4), bitumen number (EN 13179-2) and delta Ring and Ball (EN 13179-1). The impact of the filler on the chemical ageing of the binder was also studied using the Pressure Aging Vessel (PAV) test. Finally, dense and porous asphalt mixtures were manufactured with some selected fillers. Their resistance to moisture damage was evaluated using respectively the Indirect Tensile Strength Retained (ITSR - EN 12697-12) and the Cantabro test (EN 12697-11 and NLT-362). From this thorough study, the effect of the different fillers can be compared on all aspects of their potential impact on mixture properties: resistance to moisture damage, aging and mechanical properties. The benefit of using an active filler is clearly demonstrated, especially in the case of hydrated lime. Keywords Mineral Filler · Hydrated Lime · Mastic · Asphalt Mixture
1 Introduction The filler is an important component of an asphalt mixture [1–3]. However, the way it should be tested and its real influence in the final mixture are still debated, like for example the existence of an optimum filler content [4] or its impact on bitumen aging [5]. Therefore, a very systematic study was performed on this topic in Portugal A. Teixeira (B) CICCOPN, Maia and ISEP, Porto, Portugal e-mail: [email protected] D. Lesueur Lhoist, Grenoble, France e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_128
1007
1008
A. Teixeira and D. Lesueur
in order to better understand how different fillers behave and their impact in the context of Portuguese formulations and materials. The selected results presented in this paper are indeed extracted from the thorough work performed by A. Teixeira in his thesis doctoral [6]. Different mineral fillers from various sources in Portugal were compared as such, in mastics and in asphalt mixtures using standard and advanced filler tests. Mastics (filler + binder blends) were also prepared to highlight their stiffening effect but also to measure the way they potentially modify chemical ageing of the binder. Finally, dense and porous asphalts were manufactured with some selected fillers. Their resistance to moisture damage was evaluated. From this thorough study, the effect of the different fillers can be compared on all aspects of their potential impact on mixture properties: resistance to moisture damage, aging and mechanical properties.
2 Materials and Methods 2.1 Materials The tested materials included three types of fillers. First, 12 mineral were recovered from the dedusting system of local asphalt plants. They are identified by the two letters “FR” for recovered filler followed by the initial of their petrographical origin: FRG, FRR and FRL respectively for granite (6 samples), rhyolite (1 sample) and limestone (5 samples). A number was added when several sources for the same type were available. Second, 5 commercial limestone fillers were obtained from different local suppliers. They are identified by the two letters "FC" for commercial filler followed by a number. Finally, 3 active fillers were obtained from different local suppliers. A sample of calcic hydrated lime (CHDA), corresponding to CL 90-S type according to EN 459-1, a sample of hydraulic lime (CHCA), corresponding to HL 5 type according to EN 459-1 and a sample of composite Portland cement (CEM) corresponding to CEM II/B-L 32.5 N type according to EN 197-1.
2.2 Filler Testing All fillers were compared using standard and advanced filler tests. All test methods required in the European standard for mineral fillers in asphalt mixtures (EN 13043) were performed: dry porosity (Rigden air voids—EN 1097-4), delta Ring and Ball (EN 13179-1) and bitumen number (EN 13179-2). In addition to these, the density in kerosene was also measured based on Annex A of EN 1097-3. Note that delta
Impact of Different Fillers on the Properties …
1009
Ring and ball can’t be done with hydrated lime because in the conditions of the test (37.5vol.% of filler in bitumen), given the high bitumen absorption power of the product, the specimen is not rich enough in bitumen to perform the test.
2.3 Mastic Testing Some of the fillers were tested for their impact on the aging resistance of the binder. This was done using the procedure described in [5] where mastics at 20wt.% of filler in a 35/50 bitumen were aged for 25 h in the Pressure Aging vessel (PAV—EN 14769). The difference in Ring and Ball softening temperature can then be measured for each mastic before and after aging (EN 1427). The following fillers were tested: FRG1, FC1, CHDA, CHCA and CEM.
2.4 Mixtures Testing Dense and porous asphalts were manufactured with some selected fillers. The dense asphalt mixtures were AC 14 surf 35/50 according to EN 13108-1 made with granitic aggregate (same origin as FRG5) using fillers FRG5, FC1, CHDA and CHCA. The bitumen content was 5wt.% based on total mixture. The total filler content was 8wt.% in the mixture, with 2.5wt.% being the added filler to be tested, and the rest coming from the 0/6 sand. The resistance to moisture damage was evaluated using the Indirect Tensile Strength Retained (ITSR - EN 12697-12). The porous asphalt mixtures were PA 12.5 mixtures according to EN 13108-7 also made with granitic aggregate (same origin as FRG5) and fillers FRG5, FC1, CHDA and CHCA. Two binders were used: a neat 35/50 bitumen and a polymer-modified PMB 45/80–60. The binder content was 4.3wt.% based on total mixture. The total filler content was 3.3wt.%, all of it being the added filler to be tested. The resistance to moisture damage was evaluated measuring the mass loss in the Cantabro test (EN 12697-11 and NLT-362).
3 Discussion and Results 3.1 Stiffening Effect It is generally accepted that Rigden air voids, bitumen number and delta R&B are three different ways to estimate the stiffening power of a given filler. The results presented in Fig. 1 showed that Rigden air voids for all recovered fillers (FRG, FRR and FRC) felt between 33.7 and 45.1% (Fig. 1). In parallel, all FC fillers behaved in a
1010
A. Teixeira and D. Lesueur
very similar way with a mean value of 29.4% and a narrow range of 28.3–31.2%. On the contrary, active fillers were clearly differentiated, with hydrated lime having the highest value of all fillers (59.9%), and hydraulic lime (43.7%) and cement (36.3%) behaving more like the “normal” mineral fillers, whether recovered or commercial. Delta R&B, which measures the stiffening of 37.5vol.% of a filler in 70/100 bitumen, showed more or less the same trend; recovered fillers experienced a wide range (9.2–33.0 °C) while FC fillers were all behaving in a very similar way with a mean value of 11.1 °C and a narrow range of 9.0–12.8 °C. Active fillers were again clearly differentiated, with hydrated lime having too-high a stiffening power to have the test performed and hydraulic lime (21.9 °C) and cement (13.6 °C) behaving more like the “normal” mineral fillers. Bitumen number, which quantifies the amount of water needed to achieve a reference viscosity, and therefore quantifies the stiffening effect in water, was found to follow the same trend. On the contrary, density in kerosene didn’t show the same trend, with a lot of scatter in the data. Hydrated lime was still differentiated but other fillers weren’t and the data experienced a large variation. Correlations between all properties are illustrated in Fig. 1. The best correlation is found between Bitumen Number and Rigden air voids, with a correlation coefficient of 0.88. Slightly lower correlation is found for delta R&B (0.74), which is not so surprising given that no data point is present for hydrated lime, a point clearly driving the correlation for the former parameters. Finally a poor correlation is found with density in kerosene. Clearly, Rigden air voids is a very relevant parameter, that correlates well with tests where interactions with bitumen are measured. It has in addition the advantage of being fast and cheap to perform, with no need of manipulating hot bitumen. On the contrary, density in kerosene shows too much scatter to be used in a systematic way.
3.2 Effect on Binder Ageing The effect of mineral filler on binder aging is only repeated here, given that it has been described at length in a separate paper [5]. Granitic mineral filler FRG1 and commercial limestone filler FC1 were tested in comparison to all active fillers. It was observed that both “normal” mineral fillers had essentially no effect on bitumen ageing in the test method. Interestingly, both Portland cement and hydraulic lime were also seen to reduce bitumen age-hardening, but to a lesser extent that hydrated lime. Given that it is widely accepted that the anti-ageing effect of hydrated lime originates from the acid-base interactions between the bitumen and lime [7], the same mechanisms are probably at play with Portland cement and hydraulic lime, and a lower surface area together with less basic surface sites could probably justify a lower efficiency.
Impact of Different Fillers on the Properties …
1011
Fig. 1 Correlations between Rigden air voids and delta R&B, Bitumen Number and density in kerosene for all fillers: recovered granitic (FRG), rhyolitic (FRR) and limestone (FRC), commercial limestone (FC) and Active fillers (AF)
3.3 Effect on the Resistance to Moisture Damage of Asphalt Mixtures The effect of the filler on the resistance to moisture damage of AC mixtures is shown on Fig. 2. Changing the filler from granitic (FRG5) to limestone (FC1) had a limited impact with an almost unchanged Indirect Tensile Strength Retained (ITSR) of
Fig. 2 Indirect Tensile Strength Retained for the AC with neat bitumen as a function of filler type. For each mixture, the left clear bar is the strength for the dry specimens and the right dark bar, for the wet specimens. The ratio (ITSR) between both parameters is also given
1012
A. Teixeira and D. Lesueur
A/ PA with 35/50 bitumen
B/ PA with polymer-modified bitumen
Fig. 3 Mass loss in the Cantabro test for the PA with neat bitumen (a) and polymer-modified bitumen (b), as a function of filler type. D: Dry and W: wet specimens. The top arrow indicates the maximum value for each set of formulas, and the bottom arrow, the minimum
respectively 76.2 and 78.0%. On the contrary, using an active filler, whether hydrated or hydraulic lime, gave a significative improvement to the ITSR with respectively 87.6 and 86.3%. The effect of the filler on the resistance to moisture damage of Porous Asphalt mixtures is shown on Fig. 3. Changing the filler from granitic (FRG5) to limestone (FC1) had again a limited impact with the neat bitumen with very similar wet and dry losses. However, a positive effect of limestone was visible with the polymer-modified binder formula, with a wet mass loss of 21.7% as compared to 27.2% for FRG5. In all cases, using an active filler was found to be again a very effective way to limit mass losses in the Cantabro test (Fig. 3). Hydrated or hydraulic lime gave a significative improvement essentially in terms of wet mass losses for all formulas. The impact was particularly seen for the combination of unmodified bitumen and hydrated lime, with a wet mass loss better than the dry mass losses of the "normal fillers".
4 Conclusions Different mineral fillers from various sources were compared. The comparison based on standard tests showed that hydrated lime was by far the most viscosifying filler. The other active fillers behaved indeed in the same way as “normal” mineral fillers. Except for density in kerosene which showed a large scatter, all properties were well correlated to the Rigden air voids. This confirms the relevance of this cheap and simple test, that is therefore perfectly suited in order to anticipate the stiffening effect of the fillers in bitumen.
Impact of Different Fillers on the Properties …
1013
Only active fillers were seen to have an impact on the chemical ageing of the bitumen. This effect was more pronounced with hydrated lime than with the other active fillers, hydraulic lime and cement. Finally, dense and porous asphalts were manufactured with some selected fillers. Their resistance to moisture damage was seen to be improved in the presence of active filler for both Asphalt Concrete and Porous Asphalt formulas. No systematic improvement was observed when using limestone filler versus granitic filler. From this very complete study, the benefits of using an active filler are clearly demonstrated, especially in the case of hydrated lime, the only filler providing three functionalities, namely higher stiffening, reduced aging and improved resistance to moisture damage.
References 1. Tayebali, A., Malpass, G., Khosla, N.: Effect of mineral filler type and amount on design and performance of asphalt concrete mixtures. Transp. Res. Record: J. Transp. Res. Board 1609, 36–43 (1998) 2. Vansteenkiste, S., Vanelstraete, A.: Properties of fillers: relationship with laboratory performance in hot mix asphalt. J. Assoc. Asphalt Paving Technol. 77, 361–398 (2008) 3. Faheem, A., Hintz, C., Bahia, H., Al-Qadi, I., Glidden, S.: Influence of filler fractional voids on mastic and mixture performance. Transp. Res. Record: J. Transp. Res. Board 2294, 74–80 (2012) 4. Lesueur, D., Lázaro, M.M., Andaluz, D., Ruiz, A.: On the impact of the filler on the complex modulus of asphalt mixtures. Road Mater. Pavement Design 19(5), 1057–1071 (2018) 5. Lesueur, D., Teixeira, A., Lázaro, M.M., Andaluz, D., Ruiz, A.: A simple test method in order to assess the effect of mineral fillers on bitumen ageing. Constr. Build. Mater. 117, 182–189 (2016) 6. Teixeira A (2015) Avaliação do impacto de diferentes fileres no desempenho das misturas betuminosas, Ph.D. Thesis, Univ. Porto (Portugal), 338p 7. Lesueur, D., Petit, J., Ritter, H.-J.: The mechanisms of hydrated lime modification of asphalt mixtures: a state-of-the-art review. Road Mater. Pavement Des. 14(1), 1–16 (2013)
Impact of the Performance of the Asphalt Concrete and the Geogrid Materials on the Fatigue of Geogrid-Reinforced Asphalt Concrete—Experimental Study Bertrand Pouteau, Antoine Martin, Kamal Berrada, Jacques-Antoine Decamps, and Patrice Diez Abstract Geogrids are used in asphalt layers in transportation infrastructure since 1980s. Various authors describe this kind of interlayer systems that are used to reduce reflective cracking and increase fatigue life. A recent research work published by Arsenie in 2013 focused on the impact of the fatigue life of asphalt concrete reinforced with geogrid. Part of this study consists in the laboratory study of the impact through 4 points bending fatigue tests and demonstrate a better fatigue performance due to the use of geogrid. The initial aim of the work presented here was to reproduce this results and correlate increase of fatigue performance to the property of the geogrid. However, the experimental results were unexpected and leads to completely new conclusions. The paper presents the key references of the literature. Then, the authors present the careful preparation of samples. The results are presented and analyzed. Finally, the conclusion discusses and compares the new results with the literature and draw some perspective upon the evolution of the fatigue criteria to be used for further study. Keywords Geogrid · Fatigue · Reinforced · 4 points bending · Experimental
1 Introduction This paper focuses on the use of geogrid as a reinforcement interlayer for pavement to improve resistance to traffic loading. The detailed anatomy of a geogrid [1] and other uses of geogrid [2] are not covered here. From one of the first referenced use of fiber glass grid on a road jobsite by Marks in 1990 [3] to now, geogrid interlayers demonstrate a reduction of reflective cracking in many publications [4–7]. Research projects were also carried to try to understand and model the role played by the geogrid [1, 6, 8].
B. Pouteau (B) · K. Berrada · J.-A. Decamps · P. Diez Eurovia, Centre de Recherche, Mérignac, France e-mail: [email protected] A. Martin ETF, Direction technique, Mérignac, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_129
1015
1016
B. Pouteau et al.
The recent research work published by Arsenie in 2013 [6] focused on the impact of the fatigue life of asphalt concrete reinforced with geogrid. The current study follows up this work and aims to differentiate the performance of various geogrid products of the market from their impact on the fatigue life of reinforced asphalt concrete. The next section of the paper recalls the major points of Arsenie’s work concerning materials, samples manufacturing and fatigue testing. Section 3 describes the experimental campaign for this new study and Sect. 4 focuses on samples preparation. Results are discussed in Sect. 5 and the conclusion is presented in Sect. 6.
2 Key Points of Previous Research This section is dedicated to results from [6]. A test method (4PB-G) is designed to measure fatigue performance of composite beam that include geogrid interlayer. This test method is inspired by the 4 points bending test in EN 12697-24:2012 appendix D. The asphalt concrete (AC1) considered as a reference material for the previous study is a semi coarse asphaltic concrete (EN 13108:2007) with a compacity of 92.5%. Its proportions are provided in Table 1. The reference geogrid (GG1) is a commercial fiber glass product with a size of mesh of 40 × 40 mm2 . The recommended size of prismatic specimen for 4 points bending test according to EN 12697-24:2012 and EN 12697-26:2012 is: 450 × 50 × 50 mm3 . The author chose a size of 630 × 100 × 100 mm3 to comply the size of the mesh of the geogrid (at least 2 mesh patterns in the width of specimen). Two configurations were tested. Non-reinforced asphalt concrete (NRAC) and reinforced asphalt concrete (RAC). For each configuration, 18 samples were tested—i.e. 6 slabs. Samples are manufactured using the slab compactor (EN-13108:2007). For both configurations, 3 layers of 50 mm AC1 are realized. A tack coat (TC) is applied between each layer. The rates are 300 g/m2 of residual bitumen for NRAC and 600 g/m2 bitumen for RAC. For RAC, the geogrid is applied between each layer. Sawing is performed to extract 3 samples of following height profile per slab: • For NRAC: 25 mm AC1-TC-50 mm AC1-TC-25 mm AC1 • For RAC: 25 mm AC1-TC-GG1-50 mm AC1-TC-GG1-25 mm AC1. After sawing, 2 pieces of AC1 are glued to beam edges to reach a length of 630 mm. Elementary fatigue tests 4PB-G are realized at 10 °C and 25 Hz. Strain level ε are selected between 110 × 10–6 and 150 × 10–6 . For each specimen, the amount of Table 1 Mix proportion of the reference material—AC1 Granular fraction
0/4
4/6
6/10
Filler
Bitumen 35/50
Proportion (%)
40.68
9.46
42.58
1.89
5.39
Impact of the Performance of the Asphalt …
1017
Table 2 4PB-G fatigue test results of NRAC et RAC from [6] NRAC
ε6 = 115.0 (10–6 )
p = 5.29 (−)
RAC
ε6 = 127.4 (10–6 )
p = 5.97 (−)
cycles to failure Nf50 is obtained when the apparent complex modulus of the specimen is reduced to 50% of its initial value (measure at 100 cycles of loading). Test results is expressed as the couple values (ε6 , p) determined from a Wöhler curve in Eq. (1) by regression of all the 18 elementary fatigue tests. Results are presented in Table 2. N f 50 = 106 (ε/ε6 )− p
(1)
3 Experimental Laboratory Test Program Design The experimental laboratory test program designed by the authors is presented in Table 6, at the end of this section. Its parameters are geogrids characteristics and asphalt concrete fatigue performance. The impact of this parameters is measured on a fatigue test method adapted from 4PB-G, named 4PB-G , which specificities are described below. To fully characterize this impact, asphalt concrete materials are also submitted to two-point bending test on trapezoidal shaped specimens (2PB-TR) and four-point bending test on prismatic shaped specimens (4PB-PR) according to EN 1269724:2012 appendix A and D and 4PB-G’ which represent a reference value without geogrid reinforcement. Table 3 details the gradation of the two asphalts concrete material. AC2 is designed to have a high fatigue performance with 5.39% of 35/50 bitumen and AC3 is designed to have a normal fatigue performance with a proportion of 3.94% of 35/50 bitumen. All performance values are provided in Sect. 5. Table 4 provides the geogrids nomenclature with their key performances. All the tested geogrids have square shaped mesh. The mesh size and the resistance to failure (in tension) Rt are the parametric values of the study for this material. Particularity of 4PB-G (compared to 4PB-G) are presented as underlined columns headers in Table 5. Tests are performed on HYD-4PT by Cooper. Fig. 1 is extracted from a preliminary study lead in partnership with Dresden University. Above 10 Hz, 4PB results are affected by frequency and need correction to compensate the apparent Table 3 Gradation values for AC2 and AC3 (in %) Sieve mm
0.063
0.125
0.5
1
2
4
6.3
8
10
14
AC2
9
11.2
17.6
23.9
34.2
46.2
57.1
72.7
94.3
100.0
AC3
9.4
11.6
18.4
25
38.5
43.7
51.9
67.7
92.9
100.0
1018
B. Pouteau et al.
Table 4 Geogrid nomenclature Code
Mesh Shape
Mesh Size (mm2 )
Rt (kN/m)
GG1
Square
40 × 40
>100
Square
25 × 25
>100
Square
25 × 25
>50
GG4
Square
20 × 20
>100
|E*| / |E*(10Hz)|
GG2 GG3
1.2
Asphalt concrete I -10°C
0.7
Asphalt concrete II +10°C
0.2
Asphalt concrete III +15°C 0.1
1
10
100
Aliminum +20°C
Frequency (Hz)
Fig. 1 Impact of the frequency on normalized modulus norm measure with HYD-4PT device for three asphalt concrete materials (in red) and aluminum material (in black) [9]
decrease of 10% [9]. This decrease would have a great impact on Nf50 . To avoid correction procedures, the frequency of the test is set to 10Hz (Table 6). For each fatigue test, a total of 9 unit test is performed to reduce the time of the study. The length of specimen, L, is 600 mm and the distance between the clamps, A, is reduced to 182 mm which still complies to standard requirements. This choice is described on Fig. 2 in Sect. 4 dedicated to samples preparation. Table 5 4PB-G and 4PB-G testing conditions Device
f (Hz)
T (°C)
Unit tests
L (mm)
A (mm)
Failure
4PB-G
HYD-4PT
10
10
9
600
182
Nf50
4PB-G
Zwick frame
25
10
18
630
200
Nf50
Table 6 Experimental laboratory test program Asphalt concrete
Geogrid
2PB
4PB
4PB-G’
AC2
–
X
X
X
AC2
GG1
X
AC2
GG2
X
AC2
GG3
AC3
–
AC3
GG1
X
AC3
GG4
X
X X
X
Impact of the Performance of the Asphalt …
1019
Fig. 2 Sawing sequence: (1) At 40 mm from the reference edges a 40 mm saw cut is realized. (2) Remove of 25 mm on the top and at the bottom. (3) Produce the three 600 × 100 × 100mm3 samples (4) remove extra width on third beam. Clamping configuration: a Clamping configuration requires a length of 630 mm b the clamping configuration of 4PB-G fits the specimen length
4 Samples Preparation Slabs of 600 × 400 × 150 mm3 are manufactured in laboratory using the MLPC slab compactor. Geogrid are carefully prepared so that two adjacent edges are cleanly cut without any fiber off. It aims to localize the mesh position while sawing the slab. Length is unrolling direction. Tack coat is applied with a brush, before laying the geogrid on the whole surface of the slab. Sawing is performed according to the sawing sequence (Fig. 2) to produce 600 × 100 × 100 mm3 samples. For 4PB-G the clamp setup is chosen to avoid applying a load on a bonded piece of AC (see Fig. 2) resulting in a manufacture of a new clamp support.
5 Results and Discussion Table 7 presents results by using the following reference nomenclature: ACi-GGj-X, where ACi stands for the asphalt concrete, GGj (if any) stands for the geogrid and X for the test among (2PB, 4PB and 4PB-G ). Temperature is 10 °C for each test. Test results are values (ε6 , p) and mean |E*init | which is the mean of all initial modulus values for each test at 10 °C and 10 Hz (i.e. 9 to 18 specimen). The use of geogrid leads to an increase of voids content, which leads to a reduction of specimen rigidity. Mean |E*init| is strongly correlated to mean voids (with a R2 of 0.998). No indication of sliding at interfaces. The differences between 2PB and 4PB on fatigue is well known by the international community [10]. No impact of the use on fatigue of any of the geogrids is identified on AC2 mix. No impact on GG1 on AC3 mix either. A 9.7% of increase is observed between AC3-GG4 and AC3 configuration for 4PB-G . This increase is just above threshold of meaningful difference which demonstrates that GG4 may
1020
B. Pouteau et al.
Table 7 Experimental results of fatigue tests Reference
Voids (%)
Frequency (Hz)
Mean |E*init | (MPa)
ε6 (10–6 )
P (–)
AC2-2PB
3.7
25
12999
128
5.848
AC2-4PB
5.1
10
12136
146
5.263
AC2-4PB-G
7.9
10
10037
143
5.435
AC2-GG1-4PB-G
8.9
10
9368
143
5.108
AC2-GG2-4PB-G
8.1
10
9783
146
5.587
AC2-GG3-4PB-G
8.5
10
9543
144
5.882
AC3-2PB
7.4
25
92
5.814
AC3-4PB-G
9.5
10
14222
93
6.214
AC3-GG1-4PB-G
10.7
10
13264
91
5.234
AC3-GG4-4PB-G
9.5
10
13038
102
5.459
have an impact on fatigue performance: complementary tests are needed to confirm this statement.
6 Conclusion Main conclusions of this study are: (i) literature results are not confirmed. Use of geogrid is likely to have no impact on fatigue performance for asphalt concrete measured with 4PB-G method. Only one configuration may lead to an increase and requires further investigations; (ii) the decrease observed on the rigidity of the multi-layered beams is fully explained by voids evolution; (iii) since on site feedback demonstrates a positive impact of geogrids on durability of pavement; both 4PB-G and 4PB-G are not good tools to assess this behavior; (iv) further investigations should be driven to develop a new analysis criterion of failure over 50% of loss of rigidity or (v) investigate another experimental setup to assess the positive impact of geogrid on pavement durability; (vi) the impact of frequency on modulus for 4PB tests requires further works both on the calibration procedure and the metrology to identify the origin of this behavior.
References 1. Nguyen, M.L., Blanc, J., Kerzreho, J.P., Hornych, P.: Review of glass fibre grid use for pavement reinforcement and APT experiments at IFSTTAR. Road Mater. Pavement Des. 14(1), 287–308 (2013) 2. Christopher, B.R.: Geogrid in Roadway and Pavement Systems. A Design Workshop (2010) 3. Marks, V.J.: Glasgrid fabric to control reflective cracking. In: Conference Proceedings (1990) 4. Darling, J.R., Woolstencroft, J.H.: Fiberglass Geogrid performance evaluation for retarding reflective cracking. In: Proceedings of the 4th International RILEM Conference (2000)
Impact of the Performance of the Asphalt …
1021
5. Bush, A.J., Brooks, E.W.: Geosynthetic Materials in Reflective Crack Prevention (2007) 6. Arsenie, I.M.: Etude et modélisation des renforcements de chaussées à l’aide de grilles en fibre de verre sous sollicitations de fatigue. PhD Thesis (2013) 7. Kerzreho, J.P., Michaut, J.P., Hornych, P.: Enrobé armé de grille en fibre de verre : Test sur le manège de fatigue de l’IFSTTAR. Revue générale des routes, 48–51 (2010) 8. Perez, S.A.: Approche expérimentale et numérique de la fissuration réflective des chaussées: De la chaussée […] au laboratoire en passant par la modélisation. PhD thesis (2009) 9. Hanel, T.: Vergleichbarkeitsstudie ausgewählter Versuche zur Ermüdungsbeständigkeit von Asphalten nach EN 12697-24. Master thesis Technische Universität Dresden (2013) 10. Di Benedetto, H., de la Roche, C., Baaj, H., Pronk, A., Lundström R.: Fatigue of bituminous mixtures: different approaches and rilem group contribution. In: 6th RILEM Symposium PTEBM’03, Zurich (2003)
Impact of Viscoelastic and Fatigue Behavior of Asphalt Mixtures Made with RAP and Asphalt Shingle on the Life Cycle Assessment Results Nikolaos Vlasopoulos, Duc Tung Dao, Nouffou Tapsoba, Hervé Di Benedetto, Cédric Sauzéat, Mohsen Ech, and Nicolas Miravalls Abstract The use of reclaimed asphalt pavement (RAP) and recycled asphalt shingles (RAS) for the production of new asphalt mixtures has significant effect on their viscoelastic and fatigue properties. Experimental results show that addition of RAP (up to 40%) and asphalt shingle (up to 5%) in asphalt mixtures increased the stiffness and improved the fatigue properties (ε6 parameter). This paper investigates the influence of the viscoelastic and fatigue properties on the environmental performance of a secondary road operated for 20 years over its whole life cycle. Results showed that the use of asphalt mixtures with higher proportion of RAP and RAS had lower global warming potential (GWP), primarily due to the lower use of raw materials and secondarily due to changes to the use phase impacts. Longer road lifecycles and higher levels of road traffic are expected to considerably increase the contribution of the road service phase and in particular that due to pavement deflection at locations experiencing high temperature climates. Keywords Linear viscoelastic · Fatigue · Reclaimed asphalt pavement (RAP) · Shingles (RAS) · Pavement design · Life cycle assessment (LCA)
1 Introduction The use of reclaimed asphalt pavement (RAP) and recycled asphalt shingles (RAS) to produce new asphalt mixtures has had great success in asphalt paving industry and has become very common [1]. In most cases, using of RAP and RAS allows a reduction in pavement life cycle costs, improves the asphalt mixture mechanical behavior [2] and improves environmental performance [3–5]. However, the content of RAP and RAS should be carefully determined as well as the mixture design, in order to avoid premature cracking [1]. N. Vlasopoulos (B) · D. T. Dao · N. Tapsoba · M. Ech · N. Miravalls LafargeHolcim Research Centre, 95 rue du Montmurier, 38291 Saint-Quentin-Fallavier, France e-mail: [email protected] H. Di Benedetto · C. Sauzéat University of Lyon, ENTPE, Lyon, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_130
1023
1024
N. Vlasopoulos et al.
In this paper, results from lab studies published previously [6–8] have been used to design some pavement structures with asphalt mixtures containing different levels of RAP and RAS. These pavement designs where thereafter analyzed using a new life cycle assessment (LCA) software to establish their environmental performance during their whole life cycle, ranging from construction and maintenance to their service phase. This holistic approach ensures that any improvement options in the construction phase will not have detrimental results to any of the other life cycle stages.
2 Lab Results and Pavement Design 2.1 Lab Results Five bituminous mixtures with different RAP content (from 0% to 40% of the total mass) and shingles (RAS) content (from 0% to 5% of the total mass) were produced (Table 1). The viscoelastic and fatigue behavior of the 5 mixtures were tested in order to evaluate the impact of RAP and RAS. The results are summarized in Table 1. Further details on the testing procedures and results can be found in previous published papers [6–8]. The results presented in Table 1 show that the use of RAP and RAS has a significant effect on viscoelastic and fatigue properties of asphalt mixture. Addition of RAP (up to 40%) and asphalt shingles (up to 5%) increased the stiffness and improved the ε6 values. Table 1 RAP and RAS (shingles) content, linear viscoelastic properties (norm of complex modulus and Poisson’s ratio at 10 °C, 15 °C and 10 Hz) and fatigue properties from Wöhler curve (ε6 parameter and β, slope of fatigue straight line) GB3
M2
M4
M5
M6
RAP (%)
0
15
15
15
40
Shingles (%)
0
0
3
5
0
|E* |
9300
8100
9200
9800
10500
|E* | (10 °C, 10 Hz)
(15 °C, 10 Hz)
12300
10557
12236
13177
14296
Poisson’s ratio (15 °C, 10 Hz)
0.35
0.25
0.34
0.37
0.44
ε6 (fatigue parameter)
90
77.5
95.2
102
95.6
β (slope of fatigue line)
5
4.24
7.08
5.99
5.12
Impact of Viscoelastic and Fatigue Behavior of Asphalt …
1025
2.2 Pavement Design A pavement structure was designed using asphalt mixtures (M2, M4, M5, M6) as base course material, in order to study the influence of RAP and RAS content. Thicknesses and materials of the wearing course (0.025 m of very thin asphalt wearing course— BBTM in French), binder course (0.04 m of BBM in French) and of the subbase (0.10 m of unbound aggregate course—GNT in French) were fixed (see Table 3). The design was performed using the French Pavement Design Method with ALIZE software. The structure was designed for a traffic type T2 + according to the details presented in Tables 1 and 2. Then, the use of RAP or RAS influences the thickness of the base course. Results are given in Table 3. A conventional structure with bitumen treated base (GB3 in French) is also given as a reference design. Results show the interest of using good or high quantity of recycled materials (M4, M5 and M6) as they can help reduce by 20% the thickness of the base course compared to that made with GB3mix containing no recycled asphalt. Table 2 Assumptions used for pavement design T2 + / 2%
Traffic type / Traffic growth rate Average heavy trucks per day per lane
300 trucks
Number of lane per way/Number of way
2/1
Width of lane
3.5 m
Table 3 Design parameters and results according to the French design method, for a pavement using different type of asphalt mixtures for base course Parameters used and results (base course thickness)
Asphalt used for base course GB3
M2
M4
M5
M6
Epsilon 6
90
77.5
95.2
102
95.6
BETA
5
4.24
7.08
5.99
5.12
SH
0.025
0.025
0.025
0.025
0.025
SN
0.3
0.3
0.3
0.3
0.3
Kc
1.3
1.3
1.3
1.3
1.3
Ks
1
1
1
1
1
EpsiTadm
103.4
86.7
114.5
120.6
112
EpsiT
97.2
85.9
111.4
113.7
106.1
Wearing course (m)
0.025
0.025
0.025
0.025
0.025
Binder course (m)
0.04
0.04
0.04
0.04
0.04
Base course (m)
0.16
0.20
0.14
0.13
0.13
Unbound aggregate subbase (m)
0.10
0.10
0.10
0.10
0.10
1026
N. Vlasopoulos et al.
3 Life Cycle Assessment 3.1 Introduction The LCA assessment of the several designed structure pavements was done using a software developed by LafargeHolcim and critically reviewed by British Research Establishment in 2019 [9]. The life cycle inventory was calculated based on a pavement length of 1 km, total wearing course width of 7 m and a service life of 20 years and covered the impact of the construction, maintenance and use phases. The construction phase was modelled by taking into account the production of the raw materials, their transport to the jobsite and the impact of the construction process. The fuel consumption for all asphalt mixes is assumed not affected by the content of recycled materials. Maintenance operations included a renewal of the wearing course every 5 years while the wearing and binder courses were reconstructed at year 10. The service phase of a road included the evaluation of the impacts linked with the extra fuel consumption due to roughness [10], deflection [11], texture [12] and albedo [13]. Road roughness and texture were re-established after each maintenance operation, while the effects of climate (temperature and solar radiation) on deflection and albedo were assessed by considering two project locations based on Paris and Marseille (northern and southern part of France).
3.2 LCA Results Although detailed results were calculated for all 26 indicators specified in European standards [14], only global warming potential (GWP) results are shown in this paper due to space restrictions. GWP indicator was chosen as it is an important indicator for road authorities around the world. Figure 1 presents the contribution of each of the life cycle stages to the overall GWP results for each pavement design (Paris location). The results clearly show that the addition of RAP and RAS reduced the total GWP emissions due primarily to the lower thickness of the base course and thus lower requirements for production of raw materials and secondarily due to the lower material transport and construction related impacts. Figure 2 presents the GWP contributions of the individual road layers during road construction which confirms that the lower impact is due to the base course construction (materials, transport and placement). Finally, Fig. 3 presents the contribution of the use phase modules to the GWP results of each pavement design (for Paris and Marseille locations). Texture and albedo contributions were similar for all designs as they all had the same wearing course and maintenance schedules. Roughness contribution was zero for all designs as the IRI level was maintained at below 1.5 m/km due to the frequent road maintenance.
Impact of Viscoelastic and Fatigue Behavior of Asphalt …
1027
Fig. 1 Contribution of life cycle stages to the GWP results of each pavement design
Results show that pavement designs M4 to M6 have the lowest use phase impacts due to the lower contribution of the deflection phase. This decrease in deflection and thus dissipated energy was due to improved viscoelastic properties of the base course and could be further reduced if base thickness is increased. Furthermore, the impact of albedo was higher than that of the other use phase modules as it was not impacted by the low traffic levels and the frequent maintenance re-established the low albedo value of the wearing course. As expected the impacts of albedo and deflection increased as the project location changed from Paris to Marseille due to an increase in solar radiation and ambient temperature.
Fig. 2 Contribution of road layers construction to the GWP results of each pavement design
1028
N. Vlasopoulos et al.
Fig. 3 Contribution of service phase to the GWP results of each pavement design
4 Conclusion This paper examined how the viscoelastic and fatigue properties of asphalt mixtures containing different quantities of RAP and RAS can influence pavement design and the resulting road environmental performance. Results showed that pavement build with materials containing a high proportion of RAP and RAS had overall lower global warming potential (GWP), thanks to the lower use of raw materials and lower use phase impacts. Longer road lifecycles and higher levels of road traffic are expected to considerably increase the contribution of the road use phase and in particular that due to pavement deflection at locations experiencing high temperature and solar radiation climates.
References 1. Newcomb, D.E., Epps, J.A., Zhou, F.: Use of RAP & RAS in high binder replacement asphalt mixtures: a synthesis. Special Report, 213 (2016) 2. West, R., Timm, D., Willis, R., Powell, B., Tran, N., Watson, D., Sakhaeifar, M., Robbins, M., Brown R., Vargas-Nordcbeck A., Villacorta, F.L.: Phase IV NCAT pavement test track findings. National Center for Asphalt Technology, Auburn University, Report Number NCAT 12–10 (2012) 3. Vidala, R., Moliner, E., Martínez, G., Rubio, M.C.: Life cycle assessment of hot mix asphalt and zeolite-based warm mix asphalt with reclaimed asphalt pavement. Resour. Conserv. Recycl. 74, 101–114 (2013) 4. Giani, M.I., Dotelli, G., Brandini, N., Zampori, L.: Comparative life cycle assessment of asphalt pavements using reclaimed asphalt, warm mix technology and cold in-place recycling. Resour. Conserv. Recycl. 104, 224–238 (2015) 5. Farina, A., Zanetti, M.C., Santagata, E., Blengini, G.A.: Life cycle assessment applied to bituminous mixtures containing recycled materials: crumb rubber and reclaimed asphalt pavement. Resour. Conserv. Recycl. 117, 204–212 (2017)
Impact of Viscoelastic and Fatigue Behavior of Asphalt …
1029
6. Tapsoba, N., Sauzéat, C., Di Benedetto, H., Baaj, H., Ech, M.: Behaviour of asphalt mixtures containing reclaimed asphalt pavement and asphalt shingle. Road Mater. Pavement Des. 15(2), 330–347 (2014) 7. Tapsoba, N., Sauzéat, C., Di Benedetto, H., Baaj, H., Ech, M.: Three dimensional analysis of fatigue tests on bituminous mixtures. Fatigue Fract. Eng. Mater. Struct. 38(6), 730–741 (2015) 8. Baaj, H., Ech, M., Tapsoba, N., Sauzéat, C., Di Benedetto, H.: Thermomechanical characterization of asphalt mixtures modified with high contents of asphalt shingle modifier (ASM) and reclaimed asphalt pavement (RAP). Mater. Struct. 46(10), 1747–1763 (2013) 9. British Research Establishment (BRE): Critical review of LafargeHolcim’s LCA model and tool for road applications. BRE reference QUO-01583-M1Y4B6 (2019) 10. Hajj, E.Y., Xu, H., Bailey, G., Sime, M., Chkaiban, R., Kazemi, S.F., Sebaaly, P.E.: Enhanced Prediction of Vehicle Fuel Economy and other Operating Costs, Phase I: Modeling the Relationship between Vehicle Speed and Fuel. Technical Report Federal Highway Administration, Washington, DC (2017) 11. Chupin, O., Piau, J.M., Chabot, A.: Evaluation of the Structure-Induced Rolling Resistance (SRR) for pavements including viscoelastic material layers. Mater. Struct. 6(4) (Springer Netherlands) (2013) 12. Zaabar, I., Chatti, K.: NCHRP Report 720: Estimating the Effects of Pavement Condition on Vehicle Operating Costs. Transportation Research Board of the National Academies, Washington, D.C. (2012) 13. Xu, X., Gregory, J., Kirchain, R.: Evaluation of the Albedo-induced Radiative Forcing and CO2 Equivalence Savings: A Case Study on Reflective Pavements in Four Selected U.S. Urban Areas. Cambridge, MA 02139: Massachusetts Institute of Technology (2017) 14. EN 15804:2012+A1: Sustainability of construction works—Environmental product declarations—Core rules for the product category of construction products. European Committee for Standardization (CEN) (2013)
Implementing Crumb Rubber Gap-Graded and Open Graded Asphalt Technologies—The Australian Experience Laszlo Petho
Abstract Approximately 48 million tyres reach the end of their lives in Australia each year and a substantial proportion of these end up in land fill sites. Although crumbed rubber modified bitumen is used extensively worldwide, the application has been very limited in asphalt mixes in Australia. The paper summarises the performance assessment of the crumb rubber gap-graded asphalt—GGA14(CR)—and open graded asphalt—OGA10(CR) which were trialed in 2018/19. Both asphalt mixes were designed and manufactured with warm mix technology, allowing lower mixing and compaction temperatures, resulting in significant energy saving and emission reduction. The paper discusses the crumb rubber binder blend properties and its storage stability over time. The benchmarking of the rheological properties were carried out by means of the dynamic shear rheometer. Details on the volumetric mix design and performance based mix design such as moisture sensitivity, Hamburg wheel tracking and particle loss tests are provided in the paper for the relevant mix type. The paper also provides details of the manufacturing and construction and field procedures. Keywords Crumb rubber · Warm mix · Gap graded · Open graded · Performance
1 Introduction Every year millions of tyres in Australia reach their end-of-life. The end-of-life tyres are a potentially valuable resource for recycling and there is a considerable environmental benefit linked to the technology. The use of crumb rubber binder (CRB) can provide increased durability and cracking resistance. CRBs are typically used in two types of asphalt, gap graded asphalt (GGA) and open graded asphalt (OGA) and experience shows that crumb rubber modification enhances the in situ performance of asphalt mixes compared with conventional asphalt types.
L. Petho (B) Fulton Hogan Infrastructure Services, Brisbane, Australia e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_131
1031
1032
L. Petho
The OGA10(CR) and GGA14(CR) asphalt mixes were developed in sequence and trialed in a joint venture with the City of Gold Coast in 2018 and 2019. The OGA10(CR) mix was placed on a heavily trafficked arterial road in January 2018 and in June 2019. The GGA14(CR) asphalt mix was placed on three different road sections between June 2018 and June 2019. A total of 1000 tonnes of each mix was placed.
2 Crumb Rubber Binder Blend Properties In the binder blend design process, it had to be ensured that the blend of selected base binder, the rubber crumb and any other additives provided a durable and stable bituminous binder. The minimum rubber crumb content had to be 18% per total binder weight. The binder was a wet blended terminal blend; with this process it had to be ensured that the rubber crumb particles were not completely digested in the bitumen matrix. The different crumb rubber binder (CRB) manufacturing methods are discussed in detail by others [1]. It is of particular importance that the CRB maintain its properties over a fairly long period of time. Many CRB applications are limited due to the short shelf life of the product (less than 4 h). In order to meet scheduling requirements at the asphalt plant, the CRB had to maintain its properties for at least 48 h when stored at 180 °C in the asphalt plant’s bitumen tanks. The CRB properties over time are summarised in Table 1; the CRB was found to be stable over the nominated 48 h period. Given the complex nature of the CRB, its selection and assessment cannot be done by only using conventional binder test methods. For this reason, it was benchmarked on the rheological properties of different bitumen [2] and polymer modified bitumen (PMB) [3] widely available on the Australian market. In Australia plain binders are classified using their viscosity; Class 170 (C170) binder has a viscosity of 170 Pa s at 60 °C and would be roughly equivalent to a 70/100 penetration grade bitumen. A10E and A15E binders are high SBS content (>5%) PMBs extensively used in Australia. Further to the conventional binder properties (Table 1), the CRB was assessed by the Table 1 CRB properties Property
Specification
Digestion time (h) 1
4
11
48
Penetration (4 °C, 200 g, 60 s)
Min. 15
16
18
21
26
Resilience (25 °C, %)
Min. 20
37
39
42
50
Torsional recovery (25 °C, 30 s, %)
Report
23
24
30
33
Softening point (°C)
Min. 55
69
70
67
65
Viscosity (175 °C)
1.5–4.0
1.9
1.6
1.8
1.7
Implementing Crumb Rubber Gap-Graded …
1033
Table 2 General and performance based properties of various binders Test method
C170
A10E
CRB (@ 1 h)
Viscosity @ 165 °C
N/A
0.6
1.9
Torsional recovery @ 25 °C, 30 s (%)
N/A
76
23
Softening point (°C)
N/A
98
69
Viscosity @ 60 °C, 1 rad/s
176
3243
3520
Complex modulus (G*) @ 60 °C, 10 rad/s
1693
9603
16620
Phase angle (°) @ 60 °C, 10 rad/s
87
47
59
G*/sinδ(°)
1.7
13.1
19.4
dynamic shear rheometer (DSR) [4]. The complex modulus and the phase angle of the different binders were considered as key benchmarking parameters (Table 2). The complex modulus and phase angle represents a measure of the response at high-temperature of the asphalt binder and higher G*/sinδ indicates resistance to permanent deformation [5]. Preventing permanent deformation is a key consideration in sub-tropical climates [6]. It had also already been shown in multiple studies that the DSR test provides very good repeatability [7]. A general correlation between G*/sinδ and rutting susceptibility is that a binder with a higher G*/sinδ will produce an asphalt mix with a lower rutting susceptibility [5]. CRB shows high G*/sinδ value, therefore its susceptibility to rutting is considered low.
3 Crumb Rubber Open Graded Asphalt—OGA10(CR) When a warm mix additive is added in any volumetric asphalt mix design, establishing the right laboratory compaction temperature is of high importance. This ensures that the volumetric design is correct and not masked by the viscosity of the binder and provides the basis for acceptance testing in the field. For these particular projects presented in this paper, a wax based warm mix additive was used; however, it is not mandatory to use wax based warm mix additive and other works showed that a surfactant based warm mix additive provides the same benefits in terms of temperature reduction and workability. The laboratory compaction temperature for the volumetric design was established as described in AS/NZS 2891.2.2:2014; Appendix B provides the method for the determination of the equivalent compaction temperature. Test specimens are compacted without the warm mix additive at the reference temperature and with the warm mix additive at several different temperatures. Where the two compacted densities are identical on the temperature-density chart, the equivalent temperature can be established (Fig. 1). The grading, binder content and volumetric properties met the requirements of PSTS112 [8]. The binder content was optimised by using the asphalt particle loss test,
1034
L. Petho
also known as the Cantabro test. Based on all volumetric properties and the particle loss test, 6% binder content was selected. This binder content is approximately 1.2% higher than normally selected for an OGA10 asphalt mix with the conventional A15E SBS modified binder. The higher CRB content resulted in a mix that had more than 20% air voids and maintained a reasonable level of particle loss (Table 3). The binder drain –off test was carried out at 175 °C; this test method is based on the Schellenberg test method [9]. The binder drain-off was found to be 0.02% which is considered extremely low value for an OGA. At the mix design stage it was found that not all of the metered binder could be recovered in the bitumen content test using toluene, a solvent predominantly used in Australia. Therefore, the procedure described in the South African Bitumen Association Manual 19 [10] was used. It was found that a binder factor of 0.89 is valid for crumb rubber binder, i.e. 89% of the metered binder can be reclaimed with the test. The establishment of this value was critical to ensure the volumetric mix 2.180
Compacted density (t/m3)
2.140
Reference temperature @ 175°C - no WMA additive
2.120 2.100 2.080
2.160
Compaction with WMA additive
2.160
2.154
2.129
2.088 2.075
2.075
2.060
2.064
2.060 2.040 120
130
140
150
160
170
180
Compaction temperature (°C) Fig. 1 Equivalent temperature chart
Table 3 Volumetric properties of the OGA10(CR) design Binder content % Maximum density
5.0 (t/m3 )
Compacted density (t/m3 ) Binder drainage (%)
6.0
7.0
8.0
2.688
2.632
2.584
2.557
2.027
2.075
2.103
2.149
0.04
0.07
0.11
0.02
Particle loss (%)
25.6
14.5
6.8
3.7
Air voids (%)
24.6
21.2
18.6
16.0
Voids in mineral aggregate (VMA%)
33.5
32.0
31.5
31.2
Voids filled with binder (VFB%)
26.5
34.0
41.0
49.0
Binder film index (BFI)
15.3
18.0
21.5
25.1
Implementing Crumb Rubber Gap-Graded …
1035
design is correct, plant manufacturing can be kept under control and acceptance test results are interpreted correctly. The binder factor of 0.89 was found to be valid for the GGA14(CR) asphalt type as discussed in Sect. 4. CRB binders normally require high production and paving temperatures due to the extremely viscous nature of the binder [11]. The warm mix additive, which was an important aspect of the whole mix design, enabled lower production and compaction temperatures without compromising the field procedure. No fuming was observed at the asphalt plant or on site. Mixing temperatures started at 163 °C; following established site process, the mixing temperatures were lowered to 148 °C. For benchmarking purposes, a conventional OGA10(A15E) with SBS modified binder was paved first before laying the OGA10(CR) test section. While in situ air voids are normally not tested for OGA, tests were carried out in this instance to ensure design objectives were met. Average air voids for the OGA10(A15E) mix was 19.6% and for the OGA10(CR) this was 22.5%. These values were found to be satisfactory and it was concluded that the design intent was met in terms of allowing free drainage.
4 Crumb Rubber Gap-Graded Asphalt—GGA14(CR) The design concept for the GGA14(CR) asphalt is to provide a relatively high CRB content of the mix for durability, which is balanced out with the stable aggregate skeleton for stability. This technology has been successfully used in the US, however it was the first of its kind in Australia, therefore the technology transfer process had to be conducted carefully. GGA14(CR) provides a cost effective treatment for roads with extensive block cracking. The benefit of the high binder content is that because to its flexibility, the block cracking does not reflect through the asphalt layer, even if it is constructed in thinner total pavement thickness than dense graded asphalt. The GGA14(CR) asphalt mix offers superior performance to traditional asphalt while providing significant environmental benefit by using used tyres. PSTS112, in line with the Californian specification, specifies a minimum binder content by weight as 7.5%. The trial mix was designed with a CRB content of 7.8%; it allowed that all performance and volumetric requirements could be met as detailed in Table 4. The CRB type used for mix design and field trials was identical to the binder detailed in Table 1. The volumetric design, including workability test of the mix, was carried out by using the gyratory compactor [12]. The laboratory compaction temperature when using the warm mix additive was established with the same methodology explained for the OGA10(CR) asphalt mix in Sect. 3 and was found to be 145 °C. While binder drain-off is not a design requirement for GGA14(CR), it was decided to use 0.5% fibres per total mix weight, to mitigate any risk with regard to the high binder content. Subsequently, it was found during the trials that no binder was pooling on the back of the delivery trucks nor segregated behind the paver screed. Production
1036
L. Petho
Table 4 GGA14(CR) asphalt mix design parameters Property
Test method
Requirement Design mix properties
Gyratory compaction (no. of cycles) AS/NZS 2891.2.2 N = 50–150 110 VMA (%)
Q311
18.0–23.0
80%, the mixture was classified as having sufficient resistance with standard procedure but this was not true anymore with the severe one. In parallel, mixtures with hydrated lime had more than 90% retained resistance for all testing conditions [2, 6]. Thus, it was decided to construct 20 test sections in 2011–13 at elevations 300–1090 m, with traffic intensities 35–500 heavy vehicles/day [6]. Sections consisted in BBSG mixtures with 5.7–5.8% binder (50/70 pen grade bitumen except for 2 sections in 35/50 and 70/100) with or without up to 1.8% hydrated lime [6]. All mixtures were produced in an industrial batch plant, based on two local aggregate sources (granite and crushed Meuse river gravel).
2.2 Monitoring Methodology For each section, the Mean Texture Depth (MTD—NF EN 13036–1) was measured by the Vosges Department laboratory over time every 20 m with “axis” (center line) and “wheel-path” values. 20 points were measured for each lane and the average value was reported.
2.3 Results Average MTD for each section was measured as a function of age with and without lime (Fig. 1). For all new BBSGs, the MTD after construction was equal on average to 0.59 mm. The standard deviation for the 20 values for all the sections was 0.04 mm, showing great uniformity. No significant difference on construction appeared between the formulae with and without lime (Fig. 1). The differences started to appear from the 3rd year on (Fig. 1). At this age, the average MTD of the mixtures without lime reached 0.68 mm, whereas that of surfacings with lime stayed close to 0.61 mm. At the same time the standard deviation of the non-modified surfacings doubled compared to its initial value, indicating a greater heterogeneity of the surface conditions. The field observations enabled this increase to be associated with initial, and scattered, raveling/stripping. The increased standard deviation (0.07 mm) thus illustrated that surface damage was starting to develop. After the 4th year, these differences became even more accentuated (Fig. 1) with an average MTD of 0.79 mm for BBSGs without lime, compared to 0.66 mm for the treated-ones. These differences remained similar at the end of the 5th year (Fig. 1), with a number of sections without lime exceeding 0.8 mm. Degradations was now so marked that some sections required preventive maintenance (microsurfacing or surface dressing). The scatter in MTD was also highly pronounced without lime, with a standard deviation reaching 0.12 mm, well above the repeatability of the method (12%), showing that the degradations were heterogeneously distributed.
1050
D. Lesueur et al.
Fig. 1 Mean Texture Depth as a function of pavement age for the 20 sections built in 2011–13 based on BBSG with or without Asphacal® hydrated lime in the Vosges Département
After 6 years the sections with lime still had a satisfactory average value of 0.69 mm and a standard deviation of 0.09 mm, i.e. the value observed for BBSGs without lime after 3 years.
3 Highway A84 Sections 3.1 Materials 37 000 t of AC surf 0/14 (BBSG) were laid in 7-cm-thick layers on highway A84 between the cities of Caen (Normandy) and Rennes (Brittany) in 2012–13. The mixtures were made with siliceous sandstone and 4.4% 35/50 bitumen. 15% Reclaimed Asphalt Pavement (RAP) was introduced in all productions amounting to 0.7% additional binder in the mixture. Hydrated lime was added at 0, 0.5 or 1.5%. All mixtures were produced in an industrial batch plant. The sections were 1.6 to 6 km-long and where built on both directions of the highway.
3.2 Monitoring Methodology Monitoring was organized since the beginning of the construction works by Cerema Rouen and Ifsttar Nantes. Raw materials and mixtures were tested at the initial stage. Slabs were extracted after 3 and 5 years. They were cut to various dimensions with widths 400–600 mm and lengths 600–700 mm. Specimens from the wearing course could be extracted by further sawing and/or coring. The 35/50 bitumen was tested as such and after short-term (Rolling Thin Film Oven Test RTFOT EN 12607–1) and long term (Pressure Aging Vessel PAV EN 14769) aging. Binders were also recovered (EN 12697–3) from the all mixtures. All
Improving the Durability of Asphalt Mixtures with Hydrated …
1051
binders were tested for standard properties (penetration EN 1426, Ring & ball softening temperature EN1427) and advanced rheological ones (complex shear modulus EN 14770 and Bending Beam Rheometer BBR EN 14771). In parallel, chemical indicators were also measured on the binders such as the asphaltenes content (based on a test method similar to ASTM D6560) and FTIR indices [7]. The mixtures were tested for moisture damage using a test adapted from EN 12697–12, performed on the specimens cored from the slabs, then broken in Indirect Tension in dry conditions and after a conditioning of 14 days at 40 °C. Mean Texture Depth (MTD—NF EN 13036-1) were also measured as described in Sect. 2.
3.3 Results The properties measured on the binders are summarized in Fig. 2. First, binder age-hardening in the absence of hydrated lime could be observed thru the increase in softening point from an initial value of 54.4–68.4 °C after 5 years (Fig. 2). This hardening was also perceived using other indicators, as illustrated by a similar increase in the temperature at which the complex modulus G*/sin delta reaches 1 kPa at 10 rad/s or a decrease in penetration value (Fig. 2). On the low temperature side, the temperature at which the BBR modulus S(t) after 60 s reached 300 MPa (TS=300 ) or the slope m(t) reached 0.3 S(t) (Tm=0.3 ), are seen to increase form an initial value of −15.3 °C for both parameters to −12.4 and −9.4 °C respectively, showing the well-known observation that the cracking temperature increases with binder aging, hence increasing the risk of observing thermal cracking. Note that the initial binder after RTFOT and RTFOT + PAV reached properties that were quite similar to the ones from the binder extracted right after construction and after 5 years respectively (Fig. 2). In the presence of hydrated lime and after 5 years, the binder showed a reduced aging with a softening temperature of 58.8 or 58 °C for respectively 0.5 and 1.5% hydrated lime. These values were very close to that of the binder right after construction, confirming the reduction in age-hardening imparted by hydrated lime [6]. The other indicators showed the same fact, including for the low temperature fracture-criteria that remained essentially unchanged. Note that the chemical indicators, based on both asphaltenes and FTIR, confirmed that the reduced aging in the presence of lime was due to a reduction in the asphaltenes creation. Asphaltenes remained close to 16% after 5 years for the lime sectors when it was 19 after 5 years in the reference one. The slower formation of oxidized compounds such as carbonyls and sulfoxides was also recorded thru FTIR testing. The results for Indirect Tensile Strength of the mixtures are shown on Fig. 3. After construction, the reference sector showed a pretty high value of 93.8% for the ITSR, that dropped to 49.5% after 5 years (Fig. 3). In the presence of lime, the ITSR could be maintained above 85%. Note that air voids for the 5 years-old specimens with
1052
D. Lesueur et al.
Fig. 2 Rheological properties for the neat initial bitumen, the same binder after laboratory aging (RTFOT and RTFOT + PAV) and the binder extracted from A84 sections after 5 years
Fig. 3 Indirect Tensile Strength (ITS) for the initial mixtures and after 5 years. The ratio (ITSR) between dry and wet ITS is also given. Void content and Mean Texture Depth are also given
Improving the Durability of Asphalt Mixtures with Hydrated …
1053
lime were also lower (6–7.4 vs. 9.3%), thus participating to the improved resistance to moisture damage.
4 Conclusions Mixtures with and without hydrated lime were compared in two different French contexts. First, 20 sections with semi-coarse ACs (BBSG) in the Vosges department were followed for 6 years based on visual observations and measurements of Mean Texture Depth (MTD). This clearly showed that the mixtures with hyrdated lime expereicned less raveling/stripping and pavements with lime after 6 years were in a similar damage-state as the ones without lime after 3 years. The monitoring will be pursued over time in order to see the performance over a longer time frame, trying also to extract cores and test them in a similar way to what was done on highway A84. Second, the sections on highway A84, also based on BBSG-type mixtures, showed that hydrated lime slowed-down bitumen aging and allowed the mixture to maintain a good resistance to moisture damage even after 5 years. However, no surface damage was yet visible on the test sections regardless of their lime content, even if some very limited raveling/stripping could be observed on the untreated section, as maybe captured by a slightly higher MTD (Fig. 3). Clearly, monitoring must be continued in order to see if the changes recorded in the indicators materialize into premature damage for some of the sections.
References 1. 2. 3. 4. 5. 6.
Lesueur, D., Petit, J., Ritter, H.J.: Eur. Roads Rev. 20, 48–55 (2012) Lesueur, D., Denayer, C., Ritter, H.J., et al.: Proceeding 6th E&E Congress, Prague (2016) Sainton: Presented to Club d’échanges "Route et Transport" Île de France (in French) (2011) Lesueur, D., Petit, J., Ritter, H.-J.: Road Mater. Pavement Design 14(1), 1–16 (2013) Chupin, O., Piau, J.-M., Hammoum, F., et al.: Const. Build. Mater. 163, 169–178 (2018) Pibis, P., Siegel, J.-B., El-Bedoui, S., Lesueur, D.: Revue Générale des Routes et Aménagements 961, 69–73 (2019). (in French) 7. Pieri, N., Planche, J.P., Kister, J.: Analusis 4(24), 113–122 (in French) (1996)
Improving the Modulus Master Curve of Bituminous Mixes Using Ultrasonic Measurements Pejoohan Tavassoti, Taher H. Ameen, Hassan Baaj, and Giovanni Cascante
Abstract The stiffness of bituminous mixtures varies with temperature and frequency of load due to its viscoelastic nature, and this behavior is represented by the modulus master curve. This curve can be constructed using laboratory test results and models based on known material properties. This approach uses material specifications to provide a good estimation of the modulus over a range of frequencies. This estimated curve should be corrected with respect to a measured reference value; which can be obtained by non-destructive ultrasonic methods. Once the dynamic modulus at high loading frequencies is measured with the ultrasonic test, the predictive master curves can be calibrated as to represent the materials properties in the linear viscoelastic domain. In this study, Ultrasonic Waves Velocity (UWV) tests are conducted on specimens of two laboratory prepared mixes to correct the conventional modulus master curves. Moreover, the variability introduced by ultrasonic receiver type and coupling in the UWV measurements are examined, and a practical testing configuration is proposed. The measured dynamic moduli are found consistently higher than the conventional master curves; which underestimate the modulus of bituminous mixture. Finally, the method is verified by shifting the moduli measured at different frequencies and temperatures to the reduced frequencies at the reference temperature of 20 °C. Keywords Asphalt pavement · Master curve · Complex modulus · Non-destructive testing · Ultrasonic waves
1 Introduction Proper characterization of bituminous materials mandates understanding their thermo-mechanical behavior with respect to varying temperature and rate of loading. When changes in stiffness modulus is of interest, this is typically achieved through performing a series of complex modulus tests at multiple temperatures and at different P. Tavassoti · T. H. Ameen · H. Baaj (B) · G. Cascante University of Waterloo, Waterloo, ON, Canada e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_134
1055
1056
P. Tavassoti et al.
loading frequencies. The material is then treated as thermo-rheologically simple to apply the time-temperature superposition principle (TTSP). As a result, stiffness modulus of the material can be estimated over a wider range of temperatures and loading rates, as compared to those practically doable in the laboratory [1]. To this end, assumptions about the limiting maximum and minimum moduli of mixes are made, which are based on the volumetric properties and maximum modulus of asphalt cements [2, 3]. In mechanistic pavement design approaches, such as the Mechanistic Empirical Pavement Design Guide (MEPDG), the modulus values are considered important input. Therefore, improving the accuracy of modulus prediction using modulus master curve can contribute to a better pavement design. Different methods of modulus master curve construction can be found in the literature. Boz et al. [4] investigated the conventional symmetric model in MEPDG and Richards nonsymmetric sigmoid along with different shift factor functions (i.e., Arrhenius, Williams–Landel–Ferry (WLF), and 2nd order polynomial). They concluded that using Richards nonsymmetric model with a 2nd order polynomial provides a better estimate of the frequency-dependent modulus of bituminous mixes. The generalized logistic model by Richards can be described by Eq. (1). Log E ∗ = δ +
α 1 1 + λeβ+γ (Log fr ) /λ
(1)
where |E* | is the magnitude of complex modulus, fr is the reduced frequency in Hz, and δ, α, ß, γ, and λ are treated as fitting parameters. In this equation, δ stands for the minimum modulus of the mix while the value of (δ + α) represents the limiting maximum modulus. The maximum modulus is determined using Hirsch’s model as expressed by Eq. (2); which is based on presumed values for maximum modulus of asphalt cement and volumetric properties of the mix. ∗ E
mi x
VMA (V M A)(V F A) + 3G ∗ binder =Pc C 1 − 100 10, 000 1 − Pc +
MA )+ (1− V100 VMA 4,200,000
3V F A|G ∗ |binder
0.58 ∗| binder 20 + 3V F A|G VMA and Pc = 0.58 3V F A|G ∗ |binder 650 + VMA
(2)
where E∗mix is complex modulus in psi, VMA voids in mineral aggregate, is percent VFA is percent voids filled with asphalt, G∗binder is magnitude of asphalt cement complex shear modulus in psi, and C is assumed as 4,200,000. Using a value of 1 GPa for the asphalt cement modulus in Eq. (2), this model provides the maximum mixture modulus to be used when developing the modulus master curve [2].
Improving the Modulus Master Curve of Bituminous …
1057
2 Problem Statement and Objectives Complex modulus measurements at temperatures lower than −10 °C and at frequencies beyond 30 Hz is typically inconvenient and, in many cases, impractical. On the other hand, use of maximum limiting modulus in master curve construction relies on the assumption of empirical models such as Hirsch’s model as well as maximum asphalt cement modulus (typically assumed 1 GPa). Nondestructive testing (NDT) methods such as Ultrasonic Waves Velocity (UWV) measurement, used in this study, offer great benefits to improve the accuracy of predictive models by directly measuring the high frequency modulus of mixes. To this end, 54 kHz compression waves are used to directly measure the modulus of asphalt concrete specimens at different temperatures. The measurement at −10 °C can provide a more realistic estimate of limiting maximum modulus of mixes. The results of the conventional complex modulus testing with and without the moduli from UWV method are used to construct the modulus master curves.
3 Experimental Study 3.1 Materials Two different dense graded asphalt concrete mixes with different gradations (nominal maximum aggregate sizes of 19.0 mm and 12.5 mm) and two performance graded asphalt cements (i.e., PG 70-28 and PG 64-28) were prepared and tested in this study. Replicates of cylindrical specimens of approximately 100 mm in diameter by 150 mm in height were prepared after coring and trimming the originally compacted specimens from the Superpave Gyratory Compactor (SGC). The trimmed ends of the cylindrical specimens also help maintain a better contact between the transducers and the specimen surface during the ultrasonic testing.
3.2 Complex Modulus Testing Complex modulus values of the two mixes were measured in uniaxial compressive mode at five different temperatures of -10, 4, 21, 37, and 54 °C. At each of these temperatures, six loading frequencies, namely 25, 10, 5, 1, 0.5, and 0.1 Hz, were applied targeting the same strain level within the linear viscoelastic limit [5]. The testing sequence was from lowest temperature and highest frequency to highest temperature and lowest frequencies to minimize any damage to the specimens. Strain was measured using three extensometers mounted on the specimens 120° apart.
1058
P. Tavassoti et al.
3.3 Ultrasonic Waves Velocity Testing In order to avoid discrepancies due to materials variability, the same specimens were used for both complex modulus testing and ultrasonic measurements. The ultrasonic waves through transmission velocities were measured at five temperatures identical to the nominal temperatures used in complex modulus testing. To this end, a pair of 54 kHz P-wave transducers (emitter and receiver) were used. At each given temperature, multiple measurements were made to account for the testing variability. A consistent amount of coupling gel was used to enhance transmissibility of the waves during the tests. Figure 1 shows a view of the UWV test setup. The test setup included a function generator, an oscilloscope, filtering and excitation boxes, transducers, temperature-controlled chamber, and probe holders. The specimens were maintained inside the chamber all the time. The transducers were kept in place using a specifically designed holder that maintained an approximately constant and relatively small pressure on the specimens. The specimens were conditioned in the environmental chamber at the target temperature for the period of overnight to reach equilibrium before testing. The only exception was in case of 54 °C for which conditioning was limited to about 6 h to avoid high temperature induced damages. Furthermore, given that properties of bituminous mixes are strongly sensitive to changes of temperature, all of the specimens were tested inside the enclosed chamber to avoid any discrepancies due to thermal fluctuations. The same excitation voltage and frequency were used for all of the tests performed at different temperatures to ensure that the wave propagation is not affected by changes in the input signal characteristics.
Fig. 1 A view of the ultrasonic testing setup (left) and example of transmitted wave (right)
Improving the Modulus Master Curve of Bituminous …
1059
Table 1 Summary of ultrasonic measurements at different temperatures Temperature (°C)
Mix Type
T (μs)
L (mm)
Vp (m/s)
Vs (m/s)
|E* | (MPa)
−11
70–28
36.0
153.29
4264.0
2494.0
38,579.6
4
70–28
38.0
153.29
4039.2
2332.0
34,003.6
21
70–28
40.8
153.29
3755.1
2138.5
28,822.9
37
70–28
46.3
153.29
3311.2
1800.8
20,924.7
54
70–28
51.8
153.29
2959.7
1522.7
15,309.7
−11
64–28
36.5
152.28
4179.8
2444.7
36,270.0
4
64–28
38.5
152.28
3960.2
2286.4
31980.8
21
64–28
42.0
152.28
3626.8
2065.5
26306.5
37
64–28
48.1
152.28
3169.0
1723.5
18752.6
54
64–28
54.6
152.28
2791.9
1436.4
13328.9
4 Results and Discussions 4.1 Complex Modulus Through Ultrasonic Testing Compression wave propagation velocity along the length of compacted asphalt concrete cylinders was calculated based on the travelling time of first wave arrival and the average length of the specimens. Table 1 provides a summary of the average values per mix. The specimens’ diameter (i.e., 100 mm) were large enough to meet the minimum required boundary conditions when using a 54 kHz transducer. The stiffness modulus, compressive- (Vp ) and shear-wave (Vs ) velocities were calculated using Eq. (3). Temperature dependent Poisson’s ratios of 0.24, 0.25, 0.26, 0.29, and 0.32 were assumed in this study at −10, 4, 21, 37, and 54 °C, respectively, according to Tavassoti et al. [6].
∗ ρVs2 3V p2 − 4Vs2 (1 − 2υ) E = and Vs = V p V p2 − Vs2 2(1 − υ)
(3)
4.2 Modulus Master Curves The results of the complex modulus tests were used to construct the modulus master curves at a reference temperature of 20 °C. The generalized non-symmetric Richards sigmoid function was used along with a second order polynomial for the shift factor as described earlier by Eq. (1). First, only the results of the complex modulus tests were used to construct the modulus master curves. Next, the results of ultrasonic
1060
P. Tavassoti et al.
wave velocity measurements at five temperatures were combined with the complex modulus results to construct the improved curves. For this purpose, at each testing temperature an additional modulus value associated with a frequency of 54,000 Hz was appended before performing the generalized reduced gradient nonlinear optimization. Figures 2 and 3 show the results for the mixes prepared with PG 70-28 and PG 64-28, respectively. For both of the mixes, it can be seen that at reduced frequencies beyond 0.1 Hz the improved curve yields a higher stiffness modulus value as compared to the conventional master curve. This was expected as the limiting maximum modulus assumption is believed to underestimate the actual maximum modulus. This can be confirmed
Fig. 2 Modulus master curves with and without ultrasonic results for PG 70-28 mix
Fig. 3 Modulus master curves with and without ultrasonic results for PG 64-28 mix
Improving the Modulus Master Curve of Bituminous …
1061
by comparing the assumed values with those calculated based on high frequency measurements from ultrasonic testing. The strain levels for UVW testing are at very small strain level, as compared to the complex modulus testing. It should be emphasized that the differences between the two methods are significant, considering that these curves are plotted in bi-logarithmic scale. Quantitatively, a maximum difference of 31.8% and 59.5% for mixes prepared with PG 64-28 and PG 70-28 was observed, respectively. Therefore, such correction in the predicted modulus master curve can have positive impacts on modeling the bituminous materials response in a flexible pavement structure.
5 Summary and Conclusions Application of Ultrasonic Waves Velocity (UWV) measurement towards improving the accuracy of modulus master curve for bituminous mixes was successfully investigated in this study. The following conclusions can be drawn: • An improved modulus master curve can be constructed by combining the results of UWV measurements and complex modulus tests. • The conventional master curve generally underestimates the modulus values as compared to the improved curve. For the mixes investigated in this study, an underestimation as large as 59.5% was observed. • Stiffness modulus values measured using UWV tests follow the expected trend of changes with temperature. When testing at very low temperatures is not practical, UWV results at either 4 or −10 °C can provide a better estimation of the limiting maximum modulus as compared to the Hirsch’s model assumption.
References 1. Di Benedetto, H., Olard, F., Sauzéat, C., Delaporte, B.: Linear viscoelastic behaviour of bituminous materials: from binders to mixes. Road Mater. Pavement Des. 5(1), 163–202 (2004) 2. Christensen, D.W., Anderson, D.A.: Interpretation of dynamic mechanical test data for paving grade asphalt cements. J. Assoc. Asphalt Paving Technol. 61, 67–116 (1992) 3. Bonaquist, R.F.: Refining the simple performance tester for use in routine practice, vol. 614. Transportation Research Board (2008) 4. Boz, I., Tavassoti-Kheiry, P., Solaimanian, M.: The advantages of using impact resonance test in dynamic modulus master curve construction through the abbreviated test protocol. Mater. Struct. 50, 176 (2017) 5. AASHTO TP 342-11: Standard method of test for determining dynamic modulus of hot-mix asphalt concrete mixtures (2015) 6. Tavassoti-Kheiry, P., Solaimanian, M., Qiu, T.: Characterization of High RAP/RAS asphalt mixtures using resonant column tests. J. Mater. Civ. Eng. 28(11), 04016143 (2016)
Influence of Aggregate Gradation on the Permanent Deformation of Asphalt Mixtures Bernardita Lira, Robert Lundström, and Jonas Ekblad
Abstract The following study aims to measure the resistance to permanent deformation of mixtures with varying aggregate gradation and correlate these results to the performance predicted by the Marshall stability test. The experimental study is performed on 6 mixtures designed with the Marshall mix design method and measured using the wheel tracking test. The total rut depth obtained with the wheel tracking test shows a high correlation to the Marshall flow values, as higher rutting corresponds to mixtures with higher flow. The study has shown the ability of Marshall flow value to predict the resistance to permanent deformation, independent of asphalt mixture type, as well as a close correlation between aggregate gradation and stability. Keywords Asphalt mixture · Aggregate gradation · Marshall stability and flow · Permanent deformation · Wheel track test
1 Introduction Mix design of asphalt mixtures consists on determining the proportions of aggregate and binder to ensure an adequate quality. During the years, certain properties have been identified as crucial for assuring proper performance: bitumen binder type and content, and aggregates gradation, shape and texture, among others. The Marshall mix design method evolved during the period from World War II to the late 1950s at the U.S. Army, Corps of Engineers, from a need to design/manufacture asphalt mixtures that could resist heavier loads. As a part of the Marshall mix design method, the Marshall stability test has been widely used to assess the performance of asphalt mixtures. The stability value has been considered as a measure of the resistance of the specimen to develop internal shear planes and the flow value is considered to measure the plasticity of an asphalt mixture [1]. One of the primary reasons justifying the wide use of the Marshall mix design was that B. Lira (B) · R. Lundström · J. Ekblad KTH Royal Institute of Technology, Stockholm, Sweden e-mail: [email protected] NCC Industry AB, Herrjärva, Sweden © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_135
1063
1064
B. Lira et al.
the equipment required to prepare and test laboratory specimens for other methods was too expensive [2] in contrast to the simplicity of the Marshall procedure. Empirically-based design methods, such as the Marshall method, mainly select bitumen content for a certain aggregate type and gradation, meeting criteria based on experience. This means that, strictly speaking, they are valid only for those mixture types and field conditions for which correlations have been established [3]. The development during years has led to other mix design methods that are more responsive to changes in materials and traffic conditions. These methods have however a tendency to become much more complicated, requiring new test equipment and advanced calculations and input parameters, in many cases making them difficult and impractical to be used by the industry on a daily basis.
2 Objectives This study aims to find a practical and simple way to characterize widely different gradations using existing laboratory equipment and calculation procedures that may be applicable for quality assurance and control in the production of asphalt mixtures. The main objective of this study is to measure the resistance to permanent deformation of mixtures with varying aggregate gradation and correlate these results to the performance predicted by the Marshall stability test and the volumetric properties.
3 Materials and Methods 3.1 Materials The experimental study was performed with crushed aggregates classified as granite of tonalitic composition from a quarry situated in northern Sweden. To study the effect of aggregate gradation on the behavior of the tested mixtures a penetration binder 70/100 with no modifications was chosen. Basic properties of the aggregates used and of the recovered binder are presented in Lira et al. [4].
3.2 Mix Design The mixtures used in this study have been design using the Marshall mix design method starting from 6 different gradations, with maximum nominal aggregate size of 16 mm, ranging from fine to coarse graded as visualized in Fig. 1. The tested gradations have been determined in a systematic way ranging from fine dense graded to coarse SMA-type mixtures. Mixture B and C represent two
Influence of Aggregate Gradation on the Permanent …
1065
types of dense graded mixtures while mixture E is a stone mastic asphalt according to the Swedish Transport Administration’s specifications [5]. Mixtures A and F are extreme cases outside the specification, and mixture D is in between the 2 mixture types. During planning of the experimental work, it was noticed that expanding to more mixtures, more closely spaced, becomes irrelevant as such precision does not exist in the production of asphalt mixtures. According to the Marshall method, samples with 3 different binder contents were prepared for each gradation with a target air void content of 3%. The Marshall stability test was performed on each sample, and the air void content was determined. Mixtures D, E and F had fiber added at 0.3% by the mass of the mixture. The basic properties for each mixture after binder content determination are presented in Table 1. Coarseness of these mixtures is defined as fraction of material larger than 4 mm. As observed in Table 1, the results show a very similar amount of binder content for all mixtures. It is characteristic of gap graded mixtures to present high voids in mineral aggregates, as observed for mixture F which is extreme in that sense. Figure 2 presents the results from the Marshall stability test for all mixtures. There is a significant correlation between the maximal load before failure (Marshall stability) and the coarseness, with the highest stability value obtained for the finedense graded mixture (A) and decreasing stability as the amount of coarse material increases. It may also be observed that the highest value for Marshall flow, i.e. highest deformation at failure, is obtained by mixture B with 48% material larger than 4 mm
Fig. 1 Gradation curves for mixtures A to F; B typical dense graded and E typical SMA
1066
B. Lira et al.
Table 1 Marshall properties for mixtures A to F Property
Mixture A
B
C
D
E
F
Binder content (% by weight)
5.8
6.0
5.8
5.9
5.9
5.9
Air void content (% by volume)
3.6
1.5
1.2
1.2
2.6
6.4
VMA (% by volume)
15.7
15.4
14.5
14.7
15.6
17.9
VFA (% by volume)
85.4
89.6
91.7
93.2
86.5
73.7
Coarseness (%)
38
48
57
66
76
84
VMA: voids in mineral aggregate, VFA: voids filled with asphalt, Coarseness: proportion of stone material larger than 4 mm
Fig. 2 Results from the Marshall stability test with the amount of coarse material for each mixture
and decreases for increased coarseness. The lowest flow value is obtained by the stone mastic asphalt (mixture E). In the figure it may be observed that as the mixtures become coarser the response is increasingly more ductile. Mixture A, which presents the highest stability, shows a brittle behavior, as its stability decreases dramatically after reaching its maximum value. On the other hand, mixture F shows a very ductile behavior where the sample does not totally fail. It may be believed that for mixtures where the response is mostly dominated by the coarse aggregate, loading produces a re-arrangement of the aggregates preventing a complete failure. The Marshall stability test has an abstract physical meaning for failure, on which the defined parameters have a diffuse physical definition.
Influence of Aggregate Gradation on the Permanent …
1067
Table 2 Results from wheel tracking test as an average of two samples for each mixture Mixture
Air void content (% by vol.)
Deformation rate (mm/1000 cycles)
Total rut depth (mm)
A
2.3
0.13
4.7
B
1.6
0.36
10.4
C
1.2
0.20
7.9
D
1.0
0.11
5.7
E
2.1
0.09
3.2
F
4.7
0.18
5.7
3.3 Laboratory Testing To evaluate the resistance to permanent deformation, 2 slabs of each mixture were tested using the wheel tracking test (EN 12,697–22). The wheel tracking test applies a repeated load with a rolling wheel for 10,000 cycles or until the rut exceeds 20 mm. The rut depth is measured after every cycle and the final rut depth (d10000 ) and the deformation rate, determined based on the difference in rut depth between 5000 and 10,000 cycles, is recorded. Table 2 summarizes the results from the wheel tracking test. Results show a significant linear relationship between deformation rate and total rut depth. Mixture E (SMA) have the lowest rut depth while mixture B (dense graded) the highest. Mixture D and F have the same total rut depth but show a difference in deformation rate.
4 Analysis The Marshall stability test is commonly used to determine the optimal binder content for a specific mixture, but it may also provide a performance prediction. As expressed by Metcalf [6], the load-carrying ability of an asphaltic mixture is a function of the flow value as well as the stability, and Marshall stability alone is not an absolute measure of strength. Still, the typical Marshall design criteria are in practice based on a minimum stability and VMA. Also, the Marshall mix design method is limited to dense graded asphalt mixtures, while in this study a large range of mixtures have been tested. To assess resistance to permanent deformation a comparison between Marshall values and wheel track results is performed. Figure 3 presents both Marshall parameters compared with the total rut depth, as well as the corresponding linear regression. Results from the wheel tracking test show a significant correlation at the 95% confidence level between Marshall flow value and the total rut depth (R2 = 0.79). It may be observed that mixture E (SMA) has the lowest flow value as well as
1068
B. Lira et al.
Fig. 3 Marshall stability and flow value versus total rut depth for all mixtures (in parenthesis their coarseness)
lowest total rut, showing high resistance to permanent deformation. Mixture B (dense graded) shows the highest rut depth as well as the highest flow value. Regarding Marshall stability, the correlation to permanent deformation is obscured by the 2 extreme out-of-specifications mixtures (A and F), giving a coefficient of determination of 0.18 when using all mixtures. However, for mixtures between B and E the coefficient improves to 0.91. The obtained results show that, even though Marshall stability test and wheel tracking test have fundamental differences regarding loading type, compaction of the specimen and confinement, both identify mixtures designed to resist deformation and mixtures with low rutting resistance. This is specially interesting as the Marshall stability test is not designed for mixtures outside of the dense graded mixture thresholds, but still is able to recognize stable mixtures through the flow value.
5 Conclusions The present study has focused on the influence of aggregate gradation on the rutting resistance of asphalt mixtures and its correlation to the Marshall mix design parameters. This has been achieved by testing six asphalt mixtures with different gradations, varying from fine-dense graded to gap-graded, with the Marshall stability test and the wheel tracking test.
Influence of Aggregate Gradation on the Permanent …
1069
Results show that Marshall stability has a significant correlation with the amount of coarse material in a mixture, as higher stability values were achieved by dense graded mixtures. No significant general relationship was found between stability and rutting. On the other hand, permanent deformation measured with the wheel tracking test showed a significant correlation with Marshall flow, being both methods able to identify the SMA mixture as the one with highest resistance to deformation. Stone mastic asphalt is designed to maximize its deformation resistance through stone-to-stone contact. The outcomes from this study shows that the Marshall stability test gives important information about a mixture’s response to different failure modes, even when used with other mixture types than dense graded asphalt. This may be used on production quality control to validate the functional characteristics of a specific asphalt mixture.
References 1. White, T.D.: Marshall procedures for design and quality control of asphalt mixtures. In: Asphalt Paving Technology, vol. 54, San Antonio, Texas (1985) 2. Roberts, F.L., Mohammad, L.N., Wang, L.B.: History of hot mix asphalt mixture design in the United States. J. Mater. Civ. Eng. 279–293 (2002) 3. Goetz, W.H.: The evolution of asphalt concrete mix design. In: Asphalt concrete mix design: development of more rational approaches ASTM STP 1041, Philadelphia, U.S. (1989) 4. Lira, B.: Ekblad, J., Lundström, R.: Evaluation of asphalt rutting based on mixture. Road Mater. Pavement Des. (2019) 5. The Swedish Road Administration, TDOK 2013:0529 Bitumenbundna lager, The Swedish Road Administration (2017) 6. Metcalf, C.: Use of Marshall stability test in asphalt paving mix design. In: Highway Research Board Bulletin (1959)
Influence of Binder Properties on Dynamic Shear Response of Asphalt Mixture Jiqing Zhu , Abubeker Ahmed , Xiaohu Lu, and Safwat Said
Abstract For predicting the shear-related rutting potential of asphalt pavement, it is important to understand the shear response of asphalt mixture and find the link between the mixture response and component material properties. Using controlled aggregate gradation, this paper investigates the influence of bituminous binder on the response of asphalt mixture under dynamic shear loading. Four asphalt mixtures, with two different grades of bitumen, were analysed by a simple shear test apparatus with dynamic loading. The binders, both original and after short-term aging, were characterised using a dynamic shear rheometer. The modulus and phase angle of both the asphalt mixture and bitumen samples were measured. Master curves were constructed. The results revealed that the asphalt mixture phase angle maximum, where a rutting performance indicator can be obtained, appears in a relatively narrow range of shear modulus. This range corresponds to a certain level of bitumen complex shear modulus (G*) and phase angle. By temperature sweep, a temperature value can be interpolated at the specified bitumen G* level. This temperature provides a possibility to predict the frequency of phase angle maximum for a given asphalt mixture type. It is noted that the asphalt mixture phase angle maximum corresponds to a relatively high bitumen modulus level compared to other high-temperature criteria. Keywords Bitumen · Asphalt Mixture · Dynamic Shear · Rheology
J. Zhu (B) · A. Ahmed · S. Said Swedish National Road and Transport Research Institute (VTI), Olaus Magnus väg 35, SE-581 95 Linköping, Sweden e-mail: [email protected] A. Ahmed e-mail: [email protected] S. Said e-mail: [email protected] X. Lu Nynas AB, 149 82 Nynäshamn, Sweden e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_136
1071
1072
J. Zhu et al.
1 Introduction Reliable pavement structure design and performance prediction require fundamental material parameters of asphalt mixture from dynamic mechanical analysis, e.g. modulus and phase angle. The most common test protocols nowadays, including EN 12697-26 and AASHTO T 342, employ bending, axial compression-tension or indirect tension for characterising the dynamic response of asphalt mixture. However, regarding the rutting failure of asphalt pavement, a shear test method with dynamic loading is probably more preferable, because it can better reflect the complex stress state associated with the permanent deformation of asphalt mixture [1]. To analyse the dynamic shear response of asphalt mixtures, Said et al. [2] developed a simple shear test apparatus, named shear box, and investigated its use for characterising the permanent deformation potential. From the modulus master curve, the complex viscosity of asphalt mixture at maximum phase angle was proposed as a rutting performance indicator. The shear behaviour of asphalt mixture depends on both its bituminous binder and mineral aggregates. In order to predict the shear-related rutting potential of asphalt pavement, it is thus of great importance to find the link between the mixture response and component material properties. By using controlled aggregates and gradation, this paper investigates the influence of bituminous binder. This may assist in finding the relevant bitumen properties that contribute to the response of asphalt mixture under dynamic shear loading.
2 Materials and Method Two types of dense-grade asphalt mixtures, denoted as ABT11-C with crushed aggregates and ABT11-G with gravel aggregates (uncrushed), were tested with a simple shear test apparatus (the shear box). Their gradation is identical, with nominal maximum aggregate size of 11.2 mm. Two grades of unmodified bitumen, penetration 70/100 and 160/220, were used for each mixture type. Totally four asphalt mixtures were investigated. More information on the mixture design can be found in the previous publication [2]. Binders were analysed with a dynamic shear rheometer (DSR) both before and after the rolling thin film oven test (RTFOT, by EN 12607-1). By these tests, the modulus and phase angle of both the asphalt mixture and bitumen samples were measured, and master curves were constructed.
Influence of Binder Properties on Dynamic …
1073
3 Results and Discussion 3.1 Dynamic Shear Test of Asphalt Mixture The shear box test with dynamic loading gives results of shear modulus and phase angle of asphalt mixture at various temperatures and frequencies, which allows to construct the master curves. The phase angle master curve of asphalt mixture typically presents a maximum (a peak) that has been experimentally proven to be a critical point between the binder and aggregate skeleton to dominate the mixture response under dynamic shear loading [2]. Figure 1 presents the constructed master curves of the four asphalt mixtures at reference temperature 10 °C. As expected, the bitumen grade significantly affects the shear modulus of asphalt mixture. For both ABT11-C and ABT11-G, a harder bitumen results in a higher shear modulus level. As for the phase angle, a ‘shift’ effect could be observed when changing the bitumen grade for each mixture type. The phase angle curve shifts to higher frequencies when using a softer bitumen, while the phase angle level is not significantly affected. For both mixture types, the phase angle maximum moved from 0.01 to 0.1 Hz when changing the 70/100 bitumen to 160/220. As the complex viscosity (shear modulus divided by angular frequency) of asphalt mixture at the maximum phase angle has been used as a rutting performance indicator [2], the focus of this study is placed on analysing the properties at the respective phase angle maximum of each asphalt mixture. The related results are listed in Table 1. It is indicated that, at respective maximum phase angle, the asphalt mixtures have similar shear modulus. The values lie in a relatively narrow range of shear modulus (Fig. 1), approximately 500–700 MPa as indicated by the studied asphalt mixtures, although a slight difference could be seen when using different types of aggregates (600–700 MPa for ABT11-C and 500–600 MPa for ABT11-G). For both binders, the phase angle maximum of ABT11-G mixture was slightly higher than that of ABT11-C, indicating the effect of aggregate characteristics. For each mixture type, as the shear modulus is similar at the phase angle maximum, it is the frequency of phase angle maximum that determines the asphalt mixture viscosity. Thus, asphalt mixtures with 70/100 bitumen have much higher complex viscosity than the ones with 160/220.
3.2 DSR Test of Bitumen As already shown in Table 1, for both asphalt mixture types, the 70/100 bitumen locates the phase angle maximum at 0.01 Hz, while the 160/220 bitumen at 0.1 Hz at the reference temperature 10 °C. Moreover, the phase angle maximum corresponds to certain level of mixture shear modulus. This makes it interesting to investigate the binder properties at these specific temperature and frequency levels and find out if
1074
J. Zhu et al.
(a) ABT11-C asphalt mixtures
(b) ABT11-G asphalt mixtures Fig. 1 Master curves of asphalt mixtures at reference temperature 10 °C, constructed on the basis of shear box test results
Influence of Binder Properties on Dynamic …
1075
Table 1 Results at the phase angle maximum of asphalt mixture, reference temperature 10 °C Mixture type
Bitumen type
Phase angle maximum (°)
Frequency of phase angle maximum (Hz)
Shear modulus at phase angle maximum (MPa)
Complex viscosity at phase angle maximum (MPa·s)
ABT11-C
70/100
32
0.01
635
1.1 × 104
ABT11-C
160/220
32
0.1
705
1.4 × 103
ABT11-G
70/100
35
0.01
529
6.7 × 103
ABT11-G
160/220
37
0.1
558
9.7 × 102
Table 2 DSR test results of binders at 10 °C Binder type
Frequency (Hz)
Complex shear modulus G* (kPa)
Phase angle (°)
Complex viscosity (kPa·s)
70/100, original
0.01
404
70
6.4 × 103
70/100, RTFOT
0.01
584
64
9.3 × 103
160/220, original
0.1
489
73
7.8 × 102
160/220, RTFOT
0.1
774
67
1.2 × 103
certain binder properties need to reach a specific level in order to approach the phase angle maximum. For this, binders of the same penetration grades were analysed with DSR both before and after RTFOT. The test temperature was 10 °C and frequency was 0.01 Hz for 70/100 bitumen and 0.1 Hz for 160/220 bitumen. Table 2 lists the obtained results. It is indicated that the binders after short-term aging (RTFOT) had similar complex shear modulus (G*) and phase angle in the ranges of approximately 600–800 kPa and 64–67° respectively. This roughly corresponds to the original binders of 400– 500 kPa for G* and 70–73° for phase angle. The different test frequencies, as well as the aging effect, resulted in very different complex viscosity values of the binders.
3.3 Temperature Sweep of Bitumen Provided that the phase angle maximum of asphalt mixture corresponds to a certain G* level of bitumen by DSR, the frequency of phase angle maximum, and thus the mixture complex viscosity, could be predicted for a given mixture type by analysing the binder master curve. Furthermore, based on the time-temperature superposition principle, the frequency at a reference temperature can be equivalently converted to a temperature at a reference frequency. Thus, the temperature corresponding to a specified bitumen G* level at a reference frequency might be even more practical to
1076
J. Zhu et al.
predict the asphalt mixture viscosity at phase angle maximum for a given mixture type. For this, the above-mentioned binders were analysed by temperature sweep at reference frequency 10 rad/s (1.59 Hz). The results of RTFOT binders are shown in Fig. 2, while original binders in Fig. 3. By respectively locating the G* levels of RTFOT and original binders presented in Table 2, it is found that these levels correspond to 21 °C for 160/220 bitumen and 27 °C for 70/100 bitumen at 10 rad/s. These interpolated temperatures have a relation with, and thus provide a possibility to predict, the frequency of asphalt mixture phase angle maximum (0.01 Hz for mixtures with 160/220 bitumen and 0.1 Hz for mixtures with 70/100 bitumen at 10 °C), and thus the asphalt mixture rutting performance. In addition, the German BTSV temperature (at which the bitumen G* is 15 kPa, corresponding to the softening point) [3] and binder performance grading (PG) high-temperature criteria (at which the bitumen G* is approximately 1 kPa for original binder and 2.2 kPa after RTFOT, due to the high phase angle very close to 90°) are plotted in Figs. 2 and 3 as well. It is indicated that the asphalt mixture phase angle maximum corresponds to a relatively high modulus level, comparing with other approaches regarding high-temperature binder properties.
Fig. 2 Temperature sweep results of binders after RTFOT at 10 rad/s
Influence of Binder Properties on Dynamic …
1077
Fig. 3 Temperature sweep results of original binders at 10 rad/s
4 Summary and Future Perspective Considering the rutting potential, this paper focuses on analysing the shear properties of asphalt mixture at its phase angle maximum, as well as the corresponding binder properties. It is indicated that the asphalt mixture phase angle maximum appears in a relatively narrow range of shear modulus, which corresponds to a certain level of bitumen complex shear modulus and phase angle. By temperature sweep, a temperature value can be interpolated at the specified bitumen G* level and this temperature provides a possibility to predict the frequency of asphalt mixture phase angle maximum, which is noted corresponding to a relatively high bitumen modulus level compared to other high-temperature criteria. These findings are limited to unmodified bitumen so far. Further investigation is needed for polymer-modified bitumen.
1078
J. Zhu et al.
References 1. Sousa, J.B., Leahy, R.B., Harvey, J., Monismith, C.L.: Part I - Test method selection. In: Permanent Deformation Response of Asphalt Aggregate Mixes, Report No. SHRP-A-415, pp. 11–115. SHRP, Washington DC (1994) 2. Said, S.F., Hakim, H., Eriksson, O.: Rheological characterization of asphalt concrete using a shear box. J. Test. Eval. 41(4), 602–610 (2013) 3. Alisov, A., Riccardi, C., Schrader, J., Cannone Falchetto, A., Wistuba, M.P.: A novel method to characterise asphalt binder at high temperature. Road Mater. Pavement Des. (2018). https://doi. org/10.1080/14680629.2018.1483258
Influence of Confinement Pressure on the Viscoelastic Response of Bituminous Mixtures B. S. Abhijith, S. P. Atul Narayan, and J. Murali Krishnan
Abstract This investigation reports the use of stress-strain-time data for different frequencies at a given temperature for computing the retardation spectra. While the existing attempts use a master curve based approach with continuous relaxation spectra, the use of a discrete retardation spectrum at a given temperature directly computed from the model parameters can be useful and is devoid of the associated approximations. Such an approach also amplifies the difficulties related to identifying a unique Prony series parameters for a wide range of frequencies at any given temperature. A dense graded bituminous mixture prepared with an unmodified binder was used for the study. The tests were performed on cylindrical specimens with 100 mm diameter and 150 mm height. The specimens were tested at six frequencies, ranging from 25 to 1 Hz, for two temperatures, 5 and 35 °C, in unconfined condition as well as at a confinement pressure of 200 kPa. The stress-strain response at these frequencies and temperatures were used in the estimation of material parameters for a generalized Voigt-Kelvin model. Depending on the temperature, frequency of loading, and confinement pressure, a Prony series with a spectrum of retardation time was selected. The results show that at a given temperature, a single set of model parameters is not sufficient to describe the strain response at multiple frequencies. In addition, the Prony series coefficients are shown to vary significantly at 35 °C for the confined tests, while the difference between unconfined and confined tests are insignificant at 5 °C.
B. S. Abhijith Research Scholar, Department of Civil Engineering, Indian Institute of Technology Madras, Chennai 600036, India S. P. Atul Narayan Assistant Professor, Department of Civil Engineering, Indian Institute of Technology Madras, Chennai 600036, India e-mail: [email protected] J. Murali Krishnan (B) Professor, Department of Civil Engineering, Indian Institute of Technology Madras, Chennai 600036, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_137
1079
1080
B. S. Abhijith et al.
Keywords Confinement pressure · Retardation spectra · Generalized Voigt-Kelvin model
1 Introduction Bituminous mixtures exhibit different relaxation/retardation response depending on the amplitude, frequency of loading, and test temperature. In addition, the behaviour can vary when the mixtures are subjected to confinement pressure. While for the numerical application of linear viscoelastic models, discrete relaxation/retardation spectrum is preferred, it is seen that most of the current studies appeal to fitting a master curve to frequency domain data (in terms of storage modulus, loss modulus, and phase lag) and use various linear viscoelastic models to propose continuous relaxation/retardation spectra (see for instance Bhattacharjee et al. [1]; Katicha et al. [2] for the use of fractional derivative viscoelastic models and Sun et al. [3] for the use of HN model to process and generate discrete retardation time). In this investigation, it is proposed to directly use the stress-strain-time data for different frequencies at a given temperature and elicit the required spectra. For an applied load, the creep response of the bituminous mixtures can be described using a generalized Voigt-Kelvin model [4, 5]. The creep compliance for such a model is given as follows: t −t + Ji 1 − e τi , η i=1 n
J (t) = Jg +
(1)
where J(t) is the creep compliance at time t, J g is instantaneous compliance, η is the viscosity coefficient of a dashpot, J i, and τ i is the retardation strength and retardation time of the ith Kelvin unit. Further, the strain time history, ε(t), can be predicted using the stress time history, σ (t), using an integral viscoelastic constitutive relation of the form: s (t) = J (0)σ (t) +
J˙(t − s)σ (s)ds,
(2)
0
where s is an integration variable. In this study, a cylindrical sample of the bituminous mixture is subjected to a haversine compressive cyclic loading at six frequencies (25, 20, 10, 5, 2, and 1 Hz) and two temperatures (5 and 35 °C). At a specific temperature, the test is carried out from high frequency to low frequency. For each frequency, a total of 20 cycles of loading is applied in which the last ten cycles (steady-state) are chosen for modelling purpose. The load amplitude at each frequency is adjusted in such a way that a target strain of approximately 100 με is achieved. The strain level chosen is in agreement
Influence of Confinement Pressure …
1081
with the strain levels that are applied for the determination of dynamic modulus of bituminous mixtures [6]. Further, the tests are also conducted in the presence of 200 kPa confinement pressure. When the bituminous mixtures are subjected to the test conditions mentioned above, if the response of the material to applied load is linear viscoelastic, a Prony series with a set of retardation times and retardation strengths should be sufficient enough to describe the strain response at all frequencies at a specific temperature. The coefficient of determination (R2 ) statistic is used as a basis to determine the model parameters that best fit the experimental strain. With the help of R2 statistic, it is shown in this paper that, at a specific temperature, the number of parameters required to model the strain response for all the frequencies is greater than ten, while a few parameters are sufficient to describe the behaviour at a particular frequency. In addition, the Prony series coefficients are shown to vary significantly at 35 °C for the confined tests, while the difference between unconfined and confined tests are insignificant at 5 °C.
2 Experimental Investigation An unmodified binder conforming to VG-30 as per Indian standard, IS 73 [7], was used in the manufacture of bituminous mixtures. The aggregate gradation (midgradation) corresponding to bituminous concrete grade-2 with a nominal maximum particle size of 13.2 mm and a binder content of 5% was selected as per MoRTH [8] specifications to fabricate bituminous mixtures. The bituminous mixtures were compacted using a shear box compactor [9]. Before compaction, the mixtures were short term aged at a temperature of 135 °C for 4 h ± 5 min and 155 °C for 30 min [10]. Three cylindrical specimens of 100 mm diameter and 150 mm height were obtained from a compacted beam specimen. The specimens meeting the desired air void criterion of 4 ± 0.5% were used for further testing. Hereafter, this bituminous mixture is referred to as BC-VG30.
3 Results and Discussion Confined tests were conducted following the AASHTO [6] protocol. The test specimens were pre-conditioned under the confinement pressure for a time duration until the strain response reaches a constant value. The initial strain value at the start of each frequency is normalized in such a way that ε(0) is zero. Next, the material parameters are chosen for the generalized Voigt-Kelvin model for the ten test cycles consisting of 500 data points such that the error between computed and measured strain is minimized. Simulations are carried out using MATLAB R2018a. Integration is carried out using ‘trapz’ function, and the model parameters are estimated using the ‘lsqcurvefit’ tool.
1082
B. S. Abhijith et al.
3.1 Test Temperature: 5 °C As a first step, the experimental strain at 25 Hz was predicted using Eqs. (1) and (2). The R2 value greater than 0.9 was chosen as a basis for obtaining the best fit. At a frequency of 25 Hz, a four-parameter model was found to best fit the experimental strain with an R2 value of 0.9752 for the unconfined test and 0.9781 for the confined test (see Table 1 for the details of the model parameter). However, the same fourparameter model was not able to predict well as frequency reduces from 25 to 1 Hz with an R2 value at 1 Hz being 0.3487 and 0.3361 for the unconfined and confined test, respectively. Fig. 1 shows the comparison of experimental and predicted strains for the BC-VG30 mixture tested without confinement pressure at 5 °C. Further, the number of parameters in Eq. (1) was chosen in such a way that the predicted strain at all frequencies from 25 to 1 Hz will have R2 > 0.8 with the experimental strain. It was found that an 11-parameter model is required to describe the experimental strain observed from 25 to 1 Hz (see Fig. 2). The model coefficients were found to be nearly the same for both confined and unconfined tests indicating a negligible influence of confinement pressure at 5 °C. The model parameters are given in Table 2 for confined tests. For unconfined tests, at 25 and 1 Hz frequency, the R2 value was found to be 0.9888 and 0.8878, respectively. For confined tests, at 25 and 1 Hz frequency, the R2 value was found to be 0.9885 and 0.8848, respectively. Table 1 Four term Prony series—model parameters (5 °C temperature) Ji (1/MPa)
τi (s)
Jg (1/MPa)
η (MPa.s)
0.00692, 0.03910, 0.1955 and 1.0446
3.997 × 10–5
143367.19
No confinement pressure 6.092 × 10–6 , 7.008 × 10–6 , 1.046 × 10–5 and 1.904 × 10–5
(a) 25 Hz
(b) 1 Hz
Fig. 1 Comparison of experimental and predicted strains using a four-parameter model for BCVG30 mixtures at 5 ˚C (unconfined test condition)
Influence of Confinement Pressure …
1083
(a) 25 Hz
(b) 1 Hz
Fig. 2 Comparison of experimental and predicted strains using the 11-parameter model for BCVG30 mixtures at 5 °C (confined test condition).
Table 2 Eleven term Prony series—model parameters (5 ˚C temperature) Ji (1/MPa)
τi (s)
Jg (1/MPa)
η (MPa.s)
5.342 × 10–8 , 4.002 × 10–7 , 3.028 × 10–6 , 2.123 × 10–5 , 1.382 × 10–4 , 8.396 × 10–4 , 0.0047, 0.0246, 0.1292, 0.6386 and 3.1798
4.024 × 10–5
8.8419 × 107
200 kPa confinement pressure 6.145 × 10–7 , 7.588 × 10–7 , 1.082 × 10–6 , 1.534 × 10–6 , 2.166 × 10–6 , 3.060 × 10–6 , 4.141 × 10–6 , 3.204 × 10–7 , 9.494 × 10–7 , 1.673 × 10–5 and 2.553 × 10–5
3.2 Test Temperature: 35 °C At 35 °C, the experimental strain at 25 Hz was predicted using Eq. (1) and Eq. (2). A three-parameter model (see Table 3) was able to describe the strain response at 25 Hz with an R2 value being 0.9175 for the unconfined test and 0.8654 for the confined test. The viscosity coefficient from the model for the confined test is found to be significantly higher compared to the unconfined test. This shows that the model parameters and constants significantly differ between unconfined and confined tests as temperature increases. Table 3 Three term Prony series—model parameters (35 °C) Ji (1/MPa)
τi (s)
Jg (1/MPa)
η (MPa.s)
0.0124, 0.1774 and 1.595
1.3078 × 10–4
17135
0.0193, 0.3622 and 1.6153
1.2320 × 10–4
83211
No confinement pressure 1 × 10–4 , 1 × 10–5 and 3.1 × 10–3 200 kPa confinement pressure 2 × 10–4 , 1 × 10–5 and 2 × 10–3
1084
B. S. Abhijith et al.
(a) 25 Hz
(b) 10 Hz
Fig. 3 Comparison of experimental and predicted strains using a three-parameter model for BCVG30 mixtures at 35 °C (unconfined test condition)
While a three-term Prony series model predicts the experimental strain at 25 Hz with an R2 value higher than 0.9, the same Prony constants will not describe the strain response at frequencies below 20 Hz (see Fig. 3b). Thus, the number of parameters in Eq. 1 was increased to predict the experimental strain from 25 to 1 Hz. It was found that even a 15-parameter model obtained by fitting Eq. 1, to 25 Hz frequency will not be able to predict the strain response from 10 Hz frequency. This indicates that, at 35 °C, the retardation times of the material keeps changing even at closer intervals of frequency. Thus, a single set of coefficients will not be sufficient enough to predict strain across frequencies at 35 °C.
4 Conclusion In this study, the strain response of the bituminous mixtures was predicted using a generalized Voigt-Kelvin model. At a specific temperature, the number of model parameters that are sufficient to describe strain response at all frequencies was determined. The findings indicate that, at 5 °C, an 11-parameter model is required to describe the strain response at all frequencies, and confinement pressure seems to have a negligible influence on the model coefficients. Further, at 35 °C, it was observed that the number of parameters required to describe the strain response varies at each frequency. Also, the model constants change significantly in the case of confinement pressure. For instance, the viscosity coefficient at 35 °C for tests with confinement pressure is observed to be nearly five times that of the unconfined test.
Influence of Confinement Pressure …
1085
References 1. Bhattacharjee, S., Swamy, A.K., Daniel, J.S.: Continuous relaxation and retardation spectrum method for viscoelastic characterization of asphalt concrete. Mech. Time-Depend. Mater. 16(3), 287–305 (2012) 2. Katicha, S.W., Apeagyei, A.K., Flintsch, G.W., Loulizi, A.: Universal linear viscoelastic approximation property of fractional viscoelastic models with application to asphalt concrete. Mech. Time-Depend. Mater. 18(3), 555–571 (2014) 3. Sun, Y., Huang, B., Chen, J.: A unified procedure for rapidly determining asphalt concrete discrete relaxation and retardation spectra. Constr. Build. Mater. 93, 35–48 (2015) 4. Findley, W.N., Lai, J.S., Onaran, K.: Creep and Relaxation of Nonlinear Viscoelastic Materials, with an Introduction to Linear Viscoelasticity. North-Holland, New York (1976) 5. Deepa, S., Saravanan, U., Murali Krishnan, J.: On measurement of dynamic modulus for bituminous mixtures. Int. J. Pavement Eng. 20(9), 1073–1089 (2019) 6. AASHTO T-378: Standard method of test for determining the dynamic modulus and flow number for asphalt mixtures using the asphalt mixture performance tester (AMPT). American Association of State Highway and Transportation Officials, Washington, D.C (2017) 7. IS 73.: Specifications for Paving Bitumen, Fourth Revision, Bureau of Indian Standards, New Delhi, India (2013) 8. MoRTH.: Specifications for Road and Bridge Works. Section 500, Fifth Revision, Indian Roads Congress, New Delhi, India (2013) 9. ASTM D7981.: Standard Practice for Compaction of Prismatic Asphalt Specimens by Means of the Shear Box Compactor. ASTM International, West Conshohocken, PA (2015) 10. AASHTO R30.: Standard Practice for Mixture Conditioning of Hot Mix Asphalt, American Association of State Highway and Transportation Officials, Washington, D.C., USA (2015)
Influence of Linear Viscoelastic Behaviour of Pavement Layers Interface for Heavy Weight Deflectometer Test Jean-Marie Roussel , Hervé Di Benedetto , Cédric Sauzéat, and Michaël Broutin
Abstract This paper presents a study in which an advanced constitutive model the viscoelastic behaviour of pavement layer interfaces is implemented in a pavement structure simulation. In particular, the study focuses on the effects of frequency and temperature-dependent behaviour of pavement layer interface. The airfield pavement response under Heavy Weight Deflectometer (HWD) load is investigated through a simulation performed using the Spectral Element Method (SEM). The numerical model includes dynamic effects and viscoelasticity for bituminous mixtures as well as interface between them. Results show the importance of taking into account the real behaviour (viscoelastic) of interfaces rather than a simplified one (sliding or bonded). Keywords Heavy weight deflectometer · Pavement layer interface · Viscoelasticity · Spectral element method · Dynamic simulation
1 Introduction For road and airfield pavement construction, special attention is paid to pavement interlayer. In the case of flexible pavements, bitumen emulsion is applied between two asphalt layers in order to bond them together. Indeed, it has been largely shown that, when asphalt layers are bonded together, flexible pavements have a much better bearing capacity and increased lifetime. In most of design procedures, interfaces are considered to be perfectly bonded. However, interfaces behaviour is still poorly understood. It has been recently shown from laboratory and in situ measurements that interfaces should be considered as viscoelastic [1, 2]. J.-M. Roussel (B) · H. Di Benedetto · C. Sauzéat LTDS (UMR CNRS 5513), University of Lyon/ENTPE, Rue M. Audin, 69518 Vaulx-en-Velin, France e-mail: [email protected] J.-M. Roussel · M. Broutin STAC, French Civil Aviation Technical Center, 31 av. du maréchal Leclerc, 94380 Bonneuil-sur-Marne, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_138
1087
1088
J.-M. Roussel et al.
Fig. 1 Scheme of a typical HWD loading and measuring systems
In recent years, premature airfield pavements failures due to interface delamination have been observed [3]. These examples highlight the importance of pavement monitoring and evaluation. In particular, interfaces conditions should be taken into account in assessment methods. The most widely used device for airfield pavement assessment is the Heavy Weight Deflectometer (HWD). For this test, an impulse loading is applied at the surface of the pavement by the mean of a mass falling on a load plate. The load is recorded by a load cell placed above the load plate and deflections are measured by geophones located at given distances from the load centre (see Fig. 1). HWD tests are generally analysed by a backcalculation process to obtain the mechanical parameters of the pavement layers and the subgrade. This process involves a mechanical model, of which parameters are adjusted to fit the experimental data. HWD modelling is often achieved using classical numerical methods, for instance the Finite Element Method used in Broutin [4] and Roussel et al. [5]. This paper aims at evaluating the effects of considering the viscoelastic and temperature-dependent behaviour of interface on HWD test results. For this, the HWD test has been simulated using an improved Spectral Element Method (SEM) which is able to take into account the viscoelastic behaviour of bituminous materials as well as interfaces. The considered load plate radius is 45 cm, which is a typical load plate for HWD.
2 Heavy Weight Deflectometer Test Simulation 2.1 Studied Pavement Structure The studied pavement structure corresponds to the test track of the French Civil Aviation Technical Centre test facility, located in Bonneuil-sur-Marne (Paris suburbs, France). It is a flexible four-layered pavement, designed with the French rational design method for airfield pavements. Surface layers consist in two asphalt layers, called BBA and GB. These materials have been tested in the laboratory of University of Lyon/ENTPE to determine their complex modulus in small strain domain using a tension-compression complex modulus test. Their linear viscoelastic behaviours are modelled with the 2S2P1D model calibrated on experimental data, presented in Fig. 2 [6]. More details on BBA/GB interface behaviour are given in Sect. 2.3. Deep unbound material layers, called GNT and SOL, are assumed to have a linear elastic behaviour. Their properties can also be found in Table 1.
Influence of Linear Viscoelastic Behaviour of Pavement …
1089
Table 1 Structure, material properties and interfaces conditions for studied pavement Layer material
Thickness (cm)
Modulus (MPa)
Poisson’s ratio (–)
Bulk density (kg/m3 )
Airfield type bituminous mixture (BBA)
14.7
∗ E 2S2P1D (B B A)
0.30
2 500
Interface BBA/GB
Bonded, sliding or linear viscoelastic
Base type bituminous mixture (GB)
15.6
∗ E 2S2P1D (G B)
0.30
2 500
Interface GB/GNT
Bonded 275
0.30
2 000
165
0.30
1 800
Untreated graded 51.5 aggregates (GNT) Interface GNT/SOL Subgrade (SOL)
Bonded
2.2 Spectral Element Method Applied to Heavy Weight Deflectometer Test Simulation The Heavy Weight Deflectometer (HWD) test is simulated by using the 2D axisymmetric Spectral Element Method (SEM) [7]. This semi-analytical and frequencydomain method is based on the Lamé-Navier equation (see Eq. 1) resolution which links the displacement vector u to the Lamé coefficients λ, μ and the bulk density ρ. It has been shown to accurately simulate the HWD test and to be computationally efficient. This method has been set up on the software Matlab® . It is well adapted to multi-layered geometries and takes advantage of the radial symmetry of the HWD test. It also enables the direct use of complex modulus frequency expression [8]. This feature will be used for both bituminous mixtures and interfaces viscoelastic behaviour. SEM computations include calculations every 1 Hz, from 1 to 200 Hz and cover the HWD frequency range. ρ u¨ = (λ + μ)grad(div(u)) + µu
(1)
1090
J.-M. Roussel et al.
2.3 Interface Viscoelastic Behaviour Characterization and Modelling Within the SEM code for HWD simulation, interface between layers i and i + 1 is modelled by defining the complex shear modulus K ∗ relating the radial displacement ∗ ∗ − u r,i+1 (complex notation) and the shear stress τr∗z (complex notation, see gap u r,i Eq. 2). The perfectly bonded case for the same interface would correspond to an infinite value of K ∗ . ∗ ∗ − u r,i+1 ) τr∗z = K ∗ (u r,i
(2)
The shear complex modulus K ∗ of an interface made of bitumen emulsion has been obtained with the 2T3C Hollow Cylinder Apparatus (2T3C HCA) device, recently developed at the University of Lyon/ENTPE [9, 10]. For this test, cyclic axial and rotation loadings are applied to a hollow cylinder composed of two layers of bituminous mixtures bonded by a bitumen emulsion. Experimental data have been shifted according to the time-temperature superposition principle and Williams-LandelFerry equation constants C 1 and C 2 have been determined (Table 2). The 2S2P1D model, presented in Fig. 2, in terms of shear complex modulus of an interface, has been fitted on the experimental data (Fig. 3) and the values of its seven constants are in Table 2.
3 Results Several simulations were performed, considering different cases for interface between the two asphalt layers (BBA and GB): the case where layers are perfectly bonded, and the cases for which interface behaviour is linear viscoelastic with different temperatures (0, 15, 25 and 50 °C). In Fig. 4, the shear stress is presented as a function of the horizontal displacement gap for bonded and viscoelastic cases. These data are collected at the BBA/GB interface (z = −14.7 cm), at a radial distance r = 1.00 m from the symmetry axis. In the bonded case, the horizontal displacement gap is close to zero, due to the continuity imposed condition. In the viscoelastic cases, the energy dissipation is clearly seen with hysteresis. When increasing the interface temperature, the hysteresis area increases but the maximum shear stress level remains constant (~ 0.1 MPa). In Fig. 5, the maximum central deflection is plotted as a function of the BBA/GB interface temperature. For any interface temperature, the difference with the perfectly bonded case is rather small (150
Mass variation
%
0.04
Increase of softening point after RTFOT
°C
6.78
Residual penetration
%
5.5
Ductility after RTFOT
Cm
>150
3 Results and Methodology 3.1 Adhesiveness The adhesiveness of the road bitumen to the natural aggregates coming from the five different quarry sources was determined through spectrophotometry. The principle of the method consists in determining the adhesiveness of the bitumen to the natural aggregate, calculated through the absorption of a dyestuff from a solution of given concentration by the natural aggregate as such and by the filmed aggregate (the dyestuff not absorbed by the bitumen). The results of the adhesiveness tests performed during the present study (Fig. 3) emphasize the following aspects. The 50/70 road bitumen shows an improved adhesiveness to the aggregates coming from the sources „Patars” (AR), „Poiana Ilvei” (BN) and „Forotic” (CS), the obtained values being adequate (min. 80%) even in the case of non-added bitumen. Conversely, its adhesiveness to aggregates coming from the sources „Morlaca” (CJ) and „Hauzesti” (TM) is unsatisfactory. 100 90
96.52
80 ADESIVENESS [%]
86.21
84.9 79.37
79.51
70 60 50 40 30 20 10 0
Fig. 3 Adhesiveness values
Patars Morlaca Hauzesti Poiana Ilvei Forotic
Influence of Natural Aggregates’ Mineralogical …
1099
The chemical composition of the three sources with suitable adhesiveness (min. 80%) differs, it is therefore estimated that the superior adhesiveness values of these three aggregate sources is due to the petrographic nature of the aggregates or the shape of the chipping grains. According to the visual analysis of the used chipping grain shape, the rounder aggregates (Morlaca and Hauzesti) show a higher adhesiveness to bitumen. This study confirms the importance of the grain shape for the behavior of asphalt mixtures. The aggregates coming from different sources but having the same chemical composition (Morlaca and Poiana Ilvei) subjected to another crushing process present a significantly different behavior. The shape of the grains and the quantity of silicon dioxide (SiO2 ) impact more on the adhesiveness than the origin of the aggregates (only in the case of the studies aggregates). The non-coated aggregates with neuter chemical composition (that is the silicon dioxide content between 52 and 65%) present a higher concentration of the coloring solution, following the performance of the adhesiveness test. The coated aggregates from the five sources show a roughly uniform concentration of the coloring solution, all the concentrations following the value 0.8 × 10–3 . The variation of the concentrations results from the degree of bitumen covering of the aggregates used in the test, this degree is influenced by the temperature, luminosity inside the laboratory and the awareness of the user during the coating. The test showed that a good mechanical grip realized on a hard aggregate is more important than the mineralogical composition of the aggregates for the preservation of the adhesiveness between bitumen and aggregates [6, 7]. It has been noted that in the case the aggregates are covered with a thin layer of water or dust, the bitumen can cover the aggregate grains but it cannot adhere to their surface [8]. The thin water/dust layer prevents the realization of the contact between the bitumen and the surface of the aggregates [9].
3.2 Affinity The affinity of the aggregate to the bitumen was determined as the visual degree of covering with bitumen of the grains in the non-hardened asphalt mixture subjected to the procedure of water mixing of bottles rolled during a given time. The results of the laboratory tests (Fig. 4) performed on the affinity between bitumen and natural aggregates emphasize the following main aspects. The affinity of the used 50/70 bitumen to the aggregates from the source „Forotic” (CS) is higher than the one from the other sources of aggregates. The lowest affinity values were obtained on the aggregates from the sources „Patars” (AR) and „Hauzesti” (TM), the aggregates from the source „Hauzesti” (TM) presented also a lower adhesiveness than the minimum stipulated by AND
1100
A. Stirb et al.
Affinity(%)
Fig. 4 Affinity values
100 90 80 70 60 50 40 30 20 10 0
89.84%
84.26%
75.54%
69.33%
Forotic Poiana Hauzesti Morlaca Patars Ilvei
100
100
84.9
86.21 95
min. 80%
80 79.51
79.37 70
90 89.84
84.26
80.15 60
80 75.54
50
85
Affinity (%)
96.52
90
Adehesiveness (%)
80.15%
75 70
69.33
40
65 Patars
Morlaca
Hauzesti Poiana Ilvei Forotic
Fig. 5 Adhesiveness—affinity comparison
605. This was not the case for the aggregates from the „Patars” source, since in this case the adhesiveness value was higher than the minimum one (Fig. 5).
4 Conclusions The adhesiveness between the bitumen and the natural aggregates is determined by mechanical factors, the humidity degree and dust content of the aggregates, the micro-texture of the aggregates, the grading of the mineral mix and the physical and chemical nature of the aggregates (acid, base and neuter). The tests showed that a good mechanical link, realized on a hard aggregate, is much more important than the mineralogical composition of the aggregates for maintaining the adhesiveness between the bitumen and the aggregates. The affinity test takes 6, 24, 48 and 72 h, but the results begin to be relevant only after 24.
Influence of Natural Aggregates’ Mineralogical …
1101
The results emphasize the fact that the affinity is directly proportional to the wearing resistance of the aggregates (a value of the wearing resistance ranging in the quality/screening conditions stipulated by AND 605 specific to each technical class of road, leads to an adequate affinity value). In this regard, the conclusion resulting from this comparison is that the wearing resistance of the aggregates influences the affinity and its value depends on the mechanical characteristics of the aggregates used when manufacturing of the asphalt mixture. In order to increase the values of the affinity and the adhesiveness of the bitumen to the natural aggregates where the minimum values are reached, chemical additives or the use of polymer modified bitumens are needed to increase these characteristics.
References 1. Li, Y., Yang, J., Tan, T.: Study on adhesion between asphalt binders and aggregate minerals under ambient conditions using particle modified atomic force microscope probes. Const. Build. Mater. 159–165 (2015) 2. Valdès, G., MirÒ, R., Martinez, A., Calabi, A.: Effect of the physical properties of aggregates on aggregate-asphalt bond measured using UCL method. Constr. Build. Mater. 73, 399–406 (2014) 3. Pop, M., Ciont, N., Iliescu, M.: Assessing the influence of different additives on the recycling capabilities of asphalt mixtures. In: 16th International Multidisciplinary Scientific Conference SGEM, Book, (2016) 4. AND 605, Romanian Standard: Mixturi asfaltice executate la cald. Conditii tehnice privind proiectarea, prepararea si punerea in opera. AND 605 (Hot asphalt mixes. Technical conditions concerning the design, manufacturing and laying) (2016) 5. EN 12697-11: Bituminous mixtures. Test methods for hot mix asphalt. Part 11. Determination of the affinity between aggregate and bitumen (2012) 6. Fischer, H.R., Dilingh, E.C., Hermese, C.G.M.: On the interfacial interaction between bituminous binders and mineral surfaces as present in asphalt mixtures. Appl. Surf. Sci. 265, 495–499 (2013) 7. Krzysztof, B., Jacek, O., Hubert, P., Marta, W.W.: Testing of bitumen-aggregate affinity by various methods. In: 6th Euroasphalt & Eurobitume Congress, Prague (2016) 8. Porot, L., Besamusca, J., Soenen, H., Apaegyei, A., Grenfell, J., Sybiliski, D.: Bitumen/Aggregate affinity rilem round robin test of rolling bottle test. In: 8th Rilem International Symposium on Testing and Characterization of Sustainable and Innovative Bituminous Materials, RILEM Bookseries 11 (2018) 9. Bagampadde, U., Isasccon, I.: Impact of bitumen and aggregate composition on stripping in bituminous mixtures. Mater. Struct. 39, 303–315 (2006)
In-Situ Cold Recycling Solutions for Roads Maintenance at the National Roads Network in Romania Marian Peticila, Marius Muscalu, Ioan Stetco, and Marian Raicu
Abstract As a response to the negative impact of private companies on the quality of works and deadlines that carry out maintenance works for the National Roads, of the fluctuations of the economic growth in EU and Romania, with the correlation with the public funds allocated, National Company for Road Infrastructure Administration has decided to buy machinery and prepare its own staff for carrying out works of roads based on cold recycling solutions. The recycling solutions were mainly based on in place recycling with bitumen emulsion or foamed bitumen with Wirtgen recyclers type WR 240i. This paper presents the laboratory tests for the mix design and a few comparison for road layers that have been made by cold recycling in place with bitumen emulsion. Keywords Cold recycling with bitumen emulsion · Mix design · Trial section · Road maintenance
1 Introduction The National Company of National Highways and Roads from Romania has the new experience as a works performer, because until 2018 was executed the maintenance works for roads through public procurement with private companies. Next its present a complete recipe study that was applied to the rehabilitation of a secondary national road, located in the particularly severe climatic area, in the specific conditions for mountain areas in the region of Moldova. The recycling machine used was of the type Wirtgen WR 240i and we preferred the application of the recycling technology with bitumen emulsion. The choice of the type of recycling was made also from “logistical” considerations, because the company is a public institution and public procurement for bitumen emulsion is much easier in terms of the availability of delivery. M. Peticila (B) · I. Stetco · M. Raicu National Company for Road Infrastructure Administration/NCRIA, Bucharest, Romania M. Muscalu NCRIA—Road Laboratory DRDP Iasi, Iasi, Romania © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_140
1103
1104
M. Peticila et al.
The laboratory study also followed the identification of a laboratory procedure as “fast” as possible for the quality control of the recycling works. As it is known, the productivity of cold recycling works is very high, but the compacted road material goes through a slow curing process. Thus, although the control performed on the samples with material taken from site after 14 days or by drilling the layer executed after 28 days. In these conditions, if we don’t have confirmation for quality works, it is necessary to repair a large volume of works. In this respect, the results obtained by laboratory tests at 7, 14 and 28 days were compared to validate the control method at 7 days. The recycling for the rehabilitation was done on the DN 2E road (km 30 + km 505–36 + 200 and km 47 + km 300–50 + 400). The traffic value corresponding to the 2015 is a daily average of 3500 standard axles. Even if the traffic demands are low, we have very severe climatic conditions specific in Eastern Carpathian. In the summers it’s very hot and in the winters its frequent temperature below −25 °C, and the annual averages in the range of 8–10 °C. It is also noted that in winter we can have 2 cycles of freeze-thaw in one single day. The structure of the existing road was of the flexible type, with bituminous layers, a road on which in the last decades a lot off maintenance works were performed, which led to a heterogeneous structure. The milling was performed on a depth of 12 cm of the asphalt mixing layers. The milling material (Reclaimed asphalt pavement—RAP) improved the grain size curve by adding quarry aggregates (20%) and active filler (2%). By adding bitumen emulsion (2.5% residual bitumen), a material was obtained that compacted and a base course in the new road structure, with a thickness of 20 cm.
2 Materials For RAP we find the average bitumen content of 4.35%. The bitumen extracted from the RAP was analyzed and the laboratory test confirmed that is very aged. On the bitumen extracted with chloroform, a penetration its 25·0.1 mm was at 25 °C (method from EN 1426) and by chromatographic analysis TLC/FID—“Iatroscan” was obtained colloidal index IC = 0.25 (Technical norm AND 574:2002). The particle size curve, average values, for the RAP material is shown in Table 1. Due to the extreme climatic conditions in road DN2E area, high quality aggregates have been added. The particle size distribution for the recycled mix (RAP, aggregates added and filler) is shown in Table 1. We used for the correction of the granularity curve only crushing aggregates from a single quarry, respectively the following proportions: sand 0–4 mm—18%, 4–8 mm—10% and 16–31.5 mm—10% (percentage reported at 100% mass of aggregates in the mixture, without emulsion). These aggregates correspond to the EN 13043 standard of performance certification (Declaration of the performance of the essential characteristics of the construction product by the manufacturer—DoP with the system 2+). The aggregates are
In-Situ Cold Recycling Solutions for Roads Maintenance …
1105
Table 1 Grading of aggregate—sieving analysis Material
mm sieve 0.063
0.125
1.0
2.0
4.0
8.0
16
22.4
Passing (%) RAP
0.2
0.8
22
29
38
55
79
93
Recycled mix
2.6
4.5
23
31
44
63
77
91
characterized by high fragmentation resistance (Class LA20 ), high density (ρssd ~ 2.800 Mg/m3 ) and high resistance to freeze-thaw (class MS18 ). For the correction of the granularity of fine parts, but also with the role of active filler, a hydraulic road binder in proportion of 2% was used, product certified according to EN 13282-1:2013—type HRB E3 (Doroport TB® 25). For the mixture was obtained the compaction characteristics by the Proctor method (EN 13286-2) and the optimum compaction for this mix is 6.5% and the maxim volumic mass is 2150 kg/m3 . For the role of binder of the mixture with recycled material, aggregates and filler was used cationic bitumen emulsion with slow breaking type C65B2, product certified according to EN 13808 (DoP, system 2+). The main characteristics for this emulsion is: binder content by water content—65% (Class 9), breaking behavior by breaking value index >170 (Class 5). The variants of the mix design, considering the residual bitumen content in the mixture, expressed as a percentage of the total mass (b%), the following: V1—b = 2% (adding 3.17% emulsion), V2—2.5% (adding 3.96% emulsion), V3—3.0% b (adding 4.75% emulsion) and V4—b% = 3.5% (adding 5.55% emulsion).
3 Laboratory Methods The laboratory study followed the complete evaluation of the rheology of the recycled mixture, the curing of the mixture, as well as the establishment of a simple methodology for the design and control of a quickly the in place production. The aim was to use the test equipment and methods from European standards. Given the high mobility of the recycling equipment in the territory and that it is accompanied by a semi-mobile laboratory with minimal equipment, also we performed the tests with the Marshall methodology [1, 2]. To characterize the RAP, the aggregates and the recycled mixture were used the methods of the European standards: determination of the granularity—EN 9331-1 and EN 933-2, mixed composition—EN 12697-1 and EN 12697-2. For the preparation of the samples in the laboratory was used a mixer used in the preparation of hot asphalt mixtures, of capacity above 50 kg with two counterrotating shafts, at which the heating was not started. Each material was mixed with
1106
M. Peticila et al.
the quantity of water for 10 min (to aim the OMC), after which the bitumen emulsion has beed added and the mixing continued for another 3 min. For the compaction of the specimens were used in comparison with 3 standardized methods: compression with hydraulic press according to EN 13286-41 (specimen Ø 150 mm), with “hammer Marshall” according to EN 12697-30 (specimen Ø 100 mm) compaction with gyratory press according to EN 12697-31 (specimen Ø 150 mm). For all specimens made in laboratory, the mass was calculated to obtain the volumic mass 2150 Mg/m3 and the height of samples in according with test standard. For the characterization of the samples the standards were used: density determination—EN 12697-6, water content—EN 12697-28. In all cases, the water content of the specimens was determined on the samples on which was performed the mechanical resistance tests. For the determination of the mechanical strength characteristics the standards were applied: indirect tensile strength (ITS) - EN 12697-23, water sensitivity by Tensile Strength Retained (TSR = ITSwet/ ITSdry )—EN 12697-12, stability/flow -EN 12697-34 and stiffness—EN 12697-26-Annex C.
4 Results The procedure for determining resistances in 7 days was as followed: day 1— 6 samples were prepared; day 2—the specimens was extract from molds and they were putting in the climatic room (20 °C, ~ 80% humidity); day 3—all specimens are placed in a ventilated oven at 40 °C; day 6—3 specimens are immersed in water at 20 °C and the others remain in climatic room; day 7—for the 3 dry specimens was determines the density and water content, then for all 6 specimens was performed the tests for mechanical strength (ITS or S/F or stiffness). The average values are presented in Table 2. Also, tests were performed according to the Marshall methodology, according to an identical procedure as when testing the hot mix asphalt. The curing procedure in the laboratory was identical the procedure was identical to the shown in the previous paragraph. In the day 7 the specimen was immersed in water at 60 °C for 45 min and after for then was determining the stability and flow. The results are presented in the table no. 4. To evaluate the relevance of the tests performed through the 7-day procedure, the stiffness module was determined at 7 and 14 days. The specimen diameter was Ø150 mm, imposed by the value of 32.5 mm, maximum size of aggregates from recycled mix. The procedure at 7 days was identical to the one presented in the previous paragraph. The procedure for determining in 14 days was followed: day 1—8 specimens were prepared; day 2—the specimens was extract from molds and they were putting in the climatic room (20 °C, ~ 80% humidity); day 3—all specimens are placed in a ventilated oven at 40 °C; day 6—the specimens are remove in climatic room; day 14—was performed the determination of stiffness modulus. The average values are presented in Table 4.
In-Situ Cold Recycling Solutions for Roads Maintenance …
1107
Table 2 TSR (ITSwet/ITSdry), specimen diameter −Ø150 mm, testing at 7 days Trial test
Mix V1—2.0% b Mix V2—2.5%b Mix V3—3.0%b Mix V4—3.5%b
Specimen with diameter Ø150 mm, made with hydraulic press at 75 KN Water content (%)
0.91
0.96
1.04
1.12
Density (Mg/m3 )
2.117
2.136
2.131
2.123
ITSdry (N/mm2 )
0.394
0.424
0.351
0.351
ITSwet
(N/mm2 )
TSR (%)
0.339 86.03
0.324 76.33
0.313 89.06
0.338 96.23
Specimen with diameter Ø100 mm, made with „hammer” Marshall at 75 blows Water content (%)
0.5
0.78
0.85
0.94
Density (Mg/m3 )
2.142
2.115
2.067
2.087
ITSdry (N/mm2 )
0.374
0.355
0.295
0.325
ITSwet
(N/mm2 )
TSR (%)
0.296 79.10
0.325 91.57
0.204 69.30
0.284 87.60
Table 3 Marshall test at 7 days (specimen diameter −Ø100 mm) Trial test
Mix V1—2.0%b Mix V2—2.5%b Mix V3—3.0%b Mix V4—3.5%b
Density (Mg/m3 )
2.120
2.135
2.141
2.095
Water content (%)
0.50
0.78
0.85
0.94
Stability (KN)
7.18
6.33
3.92
3.15
Fluaj (mm)
2.67
2.628
2.61
2.48
Tangential flow (mm) 1.23
0.98
1.06
0.93
S/F (KN/mm)
2.45
1.50
1.27
2.69
Table 4 Stiffness (E) at 7 days and 14 days (specimen diameter—Ø150 mm) Trial test
Mix V1—2.0%b Mix V2—2.5%b Mix V3—3.0%b Mix V4—3.5%b
Density geom. (Mg/m3 )
2.176
2.182
2.179
2.198
Stiffness—7 days (MPa)
4.477
3.810
3.756
3.330
Stiffness—14 days 4.533 (MPa)
3.933
4.178
3.556
5 Discussion Although here it’s presented a small number of values, with a good approximation we can fit the data with a polynomial function of degree 2, so we can see that for ITS evolution we have two parabolas, but with opposite directions. With the data from Table 2, for the specimen with Ø150 mm, with the same measurement units,
1108
M. Peticila et al.
and added the polynomial trend line, we obtain the function and the coefficient for correlation R2 with Eqs. (1), (2) and (3). We can say that, after a certain value of the bitumen content, if the bitumen content (b) increases, ITSDry decreases, but ITSWet and TSR increase. (usual, in the literature its linear increase for ITSwet [2]). The tendency is very clear for the preparation of samples in the press, on the samples with large diameter. For samples with the diameter of 150 mm made in the hydraulic press is a good repeatability, because for the maximum diameter of the aggregates of 31.5 mm it is even required by the standard that the diameter of its samples either 150 mm. In addition, compression with hydraulic press is more controlled and slowly. I T Sdr y = −0.03b2 + 0.13b + 0.266; R 2 = 0.60
(1)
I T Swet = 0.04b2 − 0.22b + 0.629; R 2 = 0.89
(2)
T S R = 0.16b2 + 0.84b + 1.1852; R 2 = 0.81
(3)
Although the standard of making samples limits the use of specimens Ø100 mm to maximum 16 mm seize of aggregates, we consider that for preliminary determinations in the laboratory study for the elaboration of the recipe or rapid determinations in the site laboratory, 100 mm diameter specimens can be used for ITS determination or even Marshall methodology. In addition, considering the forced curing in the oven, the tests obtained at 7 days are also relevant. Regarding at the Marshall tests, it is observed that from a stability point of view, the requirements for a base layer are easily reached even for lower values of the bitumen content. As normal, as the bitumen content increases, the flow increases. Therefore, the stability-flow ratio is relevant. According to the Romanian technical norms, for a main road in a domain 1.5–2.5 KN/mm can be accepted. From Table 3 it turns out that by the Marshall test we can narrow the field in which we find the optimal bitumen content in the range 2.5–3.0%, in the case of the studied mixture.
6 Conclusion First conclusion, after the comparative analysis of the tests, we can conclude that for the recipe study a manufacturing and testing methodology can be used at 7 days, even if samples of smaller diameter than the standard one are made. The use of small specimens is practical for the control of the production in place, since a mobile laboratory has a limited space for storage in climatic rooms or in ovens. After we considered the preliminary mechanical characteristics evaluation at 28 days for ITSdry/wet and Marshall Stability, the mix design V2 (bitumen content of 2.5%) was selected for the construction of the base course. Also, the choice was made for economic reasons in order to have a lower binder content. Then, core specimens ware extracted by rotary drilling through the layers. The average density for the
In-Situ Cold Recycling Solutions for Roads Maintenance …
1109
three 150 mm diameter cores, extracted by rotary drilling, was 1.950 Mg/m3 , which implies a degree of compaction of above 90%. Also, we obtain the values ITSdry = 0.315 N/mm2 , and stiffness E = 2680 MPa (average for 3 samples). The stiffness modulus decrease by almost 30%, compare with laboratory study form mix design, due to poor compaction energy applied on site. This may be the cause of the large difference between the values established by the laboratory study and the samples from drilling. Another cause can be a much better mixing in the laboratory of the mixture, compared to the real mixing performed by the machine in place, because the recycler does not have an integrated mixer, the mixing is done with the milling, so the mixing also depends on the speed of the machine [2].
References 1. Cazacliu, B., Peticila, M., Guieysse, B., Colange, J., Leroux, C., Bonvallet, J., Blaszczyk, R.:: Effect of process parameters on foam bitumen-based road material production. Road Mater. Pavement Des. 9(3) (2008) 2. Wirtgen GmbH: Wirtgen Cold Recycicling Technology, 1st edition (2012)
In-Situ Measurements of Variations of RAP Contents in Hot Mix Asphalt by a Handheld FT-IR Spectrometer Shams Arafat , Nazimuddin M. Wasiuddin, and Delmar Salomon
Abstract Excessive use of reclaimed asphalt pavement (RAP) in hot mix asphalt (HMA) has a detrimental consequence on the performance of asphalt pavement. Continuous quality control is crucial during the production phase, but it remains challenging to do so. Handheld Fourier Transformed Infrared Spectrometer (FTIRS) has an advantage over other test methods with respect to mobility and faster processing time which makes it a potential candidate for being a quality control tool. In this study, RAP mix from ten selected projects were investigated during the production phase using a handheld FT-IRS to determine the RAP content in the fresh mix. Carbonyl index of unaged binder (varies from 0.0166 to 0.0283) and RAP binder (varies from 0.0577 to 0.0924) were measured before the mixing began. Assuming a linear increase in carbonyl index with addition of RAP, the amount of added RAP in the mix was determined. Attenuated total reflectance (ATR) FT-IR spectrum was collected utilizing a small amount of binder which can be extracted in the field in 15 min by the quick extraction method developed in this study. At least five samples from each mix were collected at different intervals to investigate the amount of RAP in the mix. Plant samples showed the variation in RAP content within ±5% range which can be attributed to either inadequate blending or addition of variable amount of RAP. This study with ten field projects indicates that handheld FT-IRS has the potential to be used as an effective quality control tool. Keywords Handheld FT-IRS · Quality control · Plant mix · Quick extraction · Carbonyl index · RAP content
S. Arafat · N. M. Wasiuddin (B) Louisiana Tech University, Ruston, LA, USA e-mail: [email protected] D. Salomon Pavement Preservation Systems, LLC, Garden City, ID, USA © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_141
1111
1112
S. Arafat et al.
1 Introduction Incorporation of excess amount of reclaimed asphalt pavement (RAP) in asphalt mixture has adverse effect on pavement performance. So, continuous quality control during the production phase is crucial. It has been a challenging task to quickly detect the amount of RAP in the hot mix asphalt (HMA) by examining the fresh mix in the plant. As RAP contains highly aged binder, there is a chance to detect the presence of RAP by investigating the aging state of the fresh mix. Aging of asphalt binder and mixes can be investigated in the laboratory by various test methods considering their contribution in assessing certain property of the aged binder and mixes [1]. Since each test has its specific feature to define aging, processing time is also an important factor for accelerated aging determination in laboratory as well as in the field. Based on this factor, FT-IR spectroscopy requires less time compared to other tests and can also provide reliable results. Besides, a portable FT-IRS has the potential to examine the aging state of the binder or mix in the field. Aging is not only a physical phenomenon of stiffening the asphalt binder, but also chemical and rheological changes those occur during this process. In response with that, studies performed with the aid of FT-IRS showed that both the carbonyl (C=O) and sulfoxide (S=O) groups can be considered as an indicator of aging in asphalt binder [2]. Different analysis methods for quantifying C=O and S=O aging indices were performed by several researchers [3, 4]. Previous study conducted by the authors concluded that only carbonyl index can effectively identify the aging level of the asphalt mix. Overlapped absorbance spectra of sulfoxide (from aged binder) and silicon oxide (from aggregate) makes it indecisive to use the sulfoxide index as an aging indicator for the asphalt mix [5]. In most of the studies FT-IR spectrometer was used as a supporting tool to show the chemical changes due to oxidation. Most importantly, chemical changes in the RAP were not addressed properly in previous studies. Chemical aging state of RAP and mix obtained from ten different sources were investigated in this study. RAP aggregate and asphalt mixture were extracted using a quick extraction procedure using DCM (Dichloromethane). With the aid of a handheld FT-IR spectrometer field measurement of carbonyl index of binder, RAP and fresh mix samples were conducted. Results obtained from those measurement were used to determine the RAP content in plant mix samples for quality assurance.
1.1 Objective Objective of this study is to determine the RAP content in the hot mix asphalt in the plant by a handheld FT-IRS as a measure of quality control.
In-Situ Measurements of Variations of RAP Contents …
1113
2 Materials and Experimental Plan Ten different types of asphalt mixes were collected from five plants in northern Louisiana, USA. Three to five samples for each mix were collected at certain time interval and total ten different mixes were collected. The liquid binder was collected from the binder tank and RAP was collected from the RAP pile for each of the ten mixes. Details information of all the mixes are available in the final report submitted by the authors [6]. The ATR-FTIRS was performed on the unaged binder, binder residue obtained from the RAP and mix samples and FT-IR spectra were recorded. Binder residues were prepared by quick extraction process. Three to five samples for each plant mix were tested. The result provided here is the average of ten spectra for each sample.
3 Experimental Method 3.1 Quick Binder Extraction Procedure A quick extraction method was developed by the authors in the previous study which can be implemented in the field in less than 15 min with minimal supplies and effort [7]. Afterward the collection and cooling of asphalt mix, approximately 100 g of loose asphalt mix was taken in a 475 ml (16-oz) mason jar. Then around 100 ml of DCM was poured into the jar. The jar was then shaken moderately and left for 5 min with closed lid. The solvent with dissolved asphalt was then filtered through an 80-micron nylon mesh filter. The filtered liquid was collected in a metal pan and was left in an open place for 10 min while all the DCM got evaporated. The binder residue was then collected using a metal spatula and was placed on the FT-IRS crystal for spectra recording. RAP was also extracted by following the same procedure.
3.2 FT-IR Data Collection and Analysis Data Collection. A portable Agilent 4300 handheld FT-IR was used for the spectroscopy. A single reflection diamond ATR (Attenuated Total Reflectance) sample interface was used. Each spectrum was collected in the spectral region of 600– 4000 cm−1 at 4 cm−1 resolution by the default Microlab Mobile software equipped with the instrument and was reported as an average of 24 spectra. Data Analysis. FT-IR spectra were analyzed in the fingerprint region (650– 1800 cm−1 ) to identify the presence of new functional group or change of existing groups in the asphalt binder as well as in extracted mix due to aging. Quantitative analysis was then performed to measure the degree of aging using area ratio [8]. This measurement was performed by measuring the area under the spectrum with peak
1114
S. Arafat et al.
at 1695 cm−1 corresponding to C=O functional group and also the area under the spectrum with peak at 1455 cm−1 corresponding to asymmetric bending vibration of CH2 and CH3. The ratio of these two area is denoted as carbonyl index (I CO ). For measuring the area under a certain peak, baseline was drawn by connecting the valleys on the both sides of the peak and the area under the spectra bounded by the baseline is considered. Baseline for C=O functional group and vibration of CH2 and CH3 groups extended from the wavenumber 1684–1718 cm−1 and 1394–1487 cm−1 respectively.
4 Results and Discussions 4.1 Change in Carbonyl Index with Increased RAP Content It was presumed that the carbonyl index of a mix varies linearly with increased amount of RAP in the mix. Six mixes were prepared in the laboratory with 0, 10, 20, 40, 80 and 100% of RAP. RAP was heated for three hours before mixing. First two hours at 110 ºC and next hour at 163 ºC. RAP was first mixed with heated aggregate thoroughly and then the binder was poured. Asphalt content of the RAP was already known from the plant and additional binder was added to the mix to make the total asphalt content of 4.6%, which was the design asphalt content. One hundred grams of mix was taken in a small jar and all the mix was used to extract the binder. Two ends of the carbonyl index of the mix are limited by the I CO of the unaged mix and the RAP respectively. Figure 1 shows the increase of I CO from 0.0166 (for unaged binder) to 0.0634 (for RAP) with increase of the percentage of RAP. It is observed from Fig. 1 that all the points lie in the close vicinity of the straight line obtained by connecting the lowest and height value of the carbonyl index. The R2 value of the trend line (not shown in the Fig. 1) was found to be 0.9935 which is an indication of very good linear correlation. Fig. 1 Linear increase in carbonyl index with increase in RAP content in the laboratory mix
In-Situ Measurements of Variations of RAP Contents …
1115
4.2 Effect of Short-Term Aging in the Plant All the mixes were collected from two different locations in the plant: from ‘drum’ and from the ‘truck’. Carbonyl index were measured for all the mixes and based on this measurement the percentage of RAP could be determined. In Fig. 2, carbonyl index of four mixes are shown where ‘mix D’ is collected from drum and remaining three were collected from the truck. The bar charts are representing the carbonyl index of the unaged binder, used RAP and predicted value of carbonyl index if designed RAP were used. For example, the carbonyl index of unaged binder and RAP in ‘mix D’ has the carbonyl index of 0.0196 and 0.0697 respectively. The predicted carbonyl index of the mix made of 16% of the RAP binder was calculated to be 0.0276 by the following equation where RBR stands for RAP to total Binder Ratio: IC O−Mi x = IC O−Binder ∗ (1 − R B R) + IC O−R A P ∗ R B R
(1)
Except ‘mix D’ all other mixes showed a significant increase in the carbonyl index than the predicted value. It can be mentioned that ‘mix D’ was collected from the drum whereas, other three mixes were collected from the truck. There is a chance that ‘mix B’, ‘mix E’ and ‘mix H’ were short-term aged when they were being stored in silo. If the carbonyl indices were used to calculate the RAP percent in the mix it would give a much higher percentage although RAP content might be in the design limit. So, it is important to choose the location from where the sample should be collected. In case of ‘mix D’ the carbonyl index was obtained around the predicted value. So, ‘mix D’ and other mixes collected from the drum were used to determine the RAP content in the mix. Further study will be conducted to investigate the effect of short-term aging of mix in the silo.
Fig. 2 Carbonyl Index of short-termed aged mixes in the plant
1116
S. Arafat et al.
Fig. 3 RAP content determined by handheld FT-IRS in the plant
4.3 Determination of RAP Content in Fresh Mix Figure 3 is the RAP content of six different plant mixes collected from the drum. Three to five samples of each mix were extracted and tested in FT-IR to determine the RAP content in the plant. Design RAP content for each mix is shown on the horizontal axis. Carbonyl index for different samples of same mix are slightly different which can be utilized to calculate the corresponding RAP binder percent in the mix. For example, unaged binder and RAP in ‘mix A’ has the carbonyl index of 0.0220 and 0.0758 respectively. Carbonyl index of one sample from ‘mix A’ was found to be 0.0285. If I CO of the unaged binder corresponds to 0% RAP binder and I CO of RAP binder corresponds to 100% RAP binder, for a given value of I CO of 0.0285, the percentage of RAP binder in the mix was found to be 12% by the following equation: R A P% =
IC O(Mi x) − IC O(U nagedbI nder) ∗ 100 IC O(R A P) − IC O(U nagedbI nder)
(2)
Percentage of RAP in the ‘mix A’, varies from 12 to 21% where the design RAP was 16% (Fig. 3). For all the mix tested here the RAP content were found within ±5% range, only exception of one sample from ‘mix F’. There might be a probability of using higher amount of RAP in the mix or the RAP was not properly blended from where the sample was collected. Whatever the reasons are, the FT-IRS can detect the variation in carbonyl index and dictates the investigation of mix production anomaly.
5 Conclusion The study was conducted to determine the RAP content in the mix using a handheld FT-IRS. The following conclusions can be made from this study:
In-Situ Measurements of Variations of RAP Contents …
1117
• The carbonyl index of a mix increases linearly with increase in the RAP content. This linear relationship can be utilized to determine the RAP content in the mix. • FT-IRS can successfully determine the RAP content in the mix. Except one sample of certain mix, all other mix was found to be in ±5% range of the design RAP content. If the RAP or RAP pile is not homogeneous, spectra of the RAP should be collected each time when the plant mix is collected • Mix sample should be collected from the outlet of the drum before it reaches to the silo to minimize the effect of short-term aging. This study with ten field projects indicates that handheld FT-IRS has the potential to be used as an effective quality control tool. Acknowledgements This study was funded by the Louisiana Transportation Research Center (LTRC).
References 1. Fernández-Gómez, W.D., Rondón Quintana, H., Reyes Lizcano, F.: A review of asphalt and asphalt mixture aging: Una revisión. Ingenieria e investigacion 33(1), 5–12 (2013) 2. Huang, S.C., Grimes, W.: Influence of aging temperature on rheological and chemical properties of asphalt binders. Transp. Res. Rec. 2179(1), 39–48 (2010) 3. Liang, Y., Wu, R., Harvey, J.T., Jones, D., Alavi, M.Z.: Investigation into the oxidative aging of asphalt binders. Transp. Res. Rec. 0361198119843096 (2019) 4. Hofko, B., Alavi, M.Z., Grothe, H., Jones, D., Harvey, J.: Repeatability and sensitivity of FTIR ATR spectral analysis methods for bituminous binders. Mater. Struct. 50(3), 187 (2017) 5. Hung, A.M., Fini, E.H.: Absorption spectroscopy to determine the extent and mechanisms of aging in bitumen and asphaltenes. Fuel 242, 408–415 (2019) 6. Wasiuddin, N.M., Noor, L., Hossain, R., Arafat, S.: Field Implementation of Handheld FTIR for Polymer Content Determination and for Quality Control of RAP Mixtures, Louisiana Transportation Research Center. LTRC), Baton Rouge, Louisiana (2020) 7. Arafat, S., Noor, L., Wasiudding, N., Salomon, D.: Implementation of handheld FT-IR spectrometer to determine the RAP content for quality control of plant mix. In: 99th Annual Meeting of the Transportation Research Board (2020) 8. Washington, D.C., Yut, I., Bernier, A., Zofka, A.: Field applications of portable infrared spectroscopy to asphalt products. Introduction Unmanned Aircraft Syst. 127 (2015)
Interlaminar Mode-I Fracture Characterization Underwater of Reinforced Bituminous Specimens Maissa Gharbi , Armelle Chabot , Jean-Luc Geffard, and Mai Lan Nguyen
Abstract In order to investigate the interlaminar mode I fracture behavior of pavements, Wedge Splitting Tests (WST) have been developed for specimens extracted directly from field sections. As part of the French SolDuGri project focusing on bituminous pavement reinforced by glass fiber grids, WST are applied under static loads and different climatic conditions. This paper highlights the impact of water on such an interlaminar bond characterization. Results derived from samples of five to ten specimens (unreinforced or reinforced with two types of grids) tested under each type of conditions are reported here. The Digital Image Correlation (DIC) technique has been used herein. Depending on the glass fiber grid used, the observation of fractured interfaces and splitting force-crack mouth opening displacement (CMOD) curves does reveal differences. The presence of water in the bituminous emulsion influences on the dispersion of results depending on some type of interfaces. These interlaminar mode-I fracture characterizations should yield greater insight into results recently obtained from accelerated pavement tests conducted on the IFSTTAR fatigue carousel. Keywords Fiber grids · Interface · Moisture · Wedge splitting test · Digital image correlation
1 Introduction In an effort to delay the reflective cracking propagation phenomenon or to reinforce top bituminous layers, one of the main ideas proposed for roads by the fiber manufacturing industry is to insert grids into bituminous pavement structures [1–3]. The durability of such composite structures depends in part on ensuring a strong bond between grids and bituminous materials. This condition is essential, especially when M. Gharbi (B) French Civil Aviation Technical Center (STAC), Bonneuil-sur-Marne, France e-mail: [email protected] A. Chabot · J.-L. Geffard · M. L. Nguyen MAST-LAMES, Univ Gustave Eiffel, IFSTTAR, 44344 Bouguenais, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_142
1119
1120
M. Gharbi et al.
using grids inside top pavement layers. Along with rolling loads and temperature variations, the presence of water near the interface must be considered in studying debonding mechanisms [4]. To better understand the interlaminar mode I fracture behavior of pavements, Wedge Splitting Tests (WST) have been developed, whether underwater or not, for specimens extracted directly from road sections [5, 6]. These tests have displayed rather interesting results, in comparison with results obtained on actual pavement test sections. As part of the French SolDuGri project (ANR-14CE22-0019), which examines bituminous pavement reinforced by glass fiber grids [7], WST are applied under static loads and different climatic conditions. Subsequent to works aimed at characterizing the interfacial bond of multilayer composite specimens [8], a water immersion WST was introduced. This paper briefly presents the WST equipment and testing campaigns before giving the main results obtained and then drawing a conclusion.
2 Underwater WST and Specimen Characterization Extracted from in situ bituminous sections and in accordance with a RILEM recommendation [4], a large number of specimens differentiated by both type and testing conditions were prepared and assessed [5]. Specimen dimensions were chosen so that the interface dimensions contained a minimum of 3 × 3 grid meshes [6]. This project’s two selected glass fiber grids, denoted G3 and G4 for P6 and P7 specimen types respectively, have a mesh size of 40 × 40 mm2 . Their tensile strength equals 100 kN/m, with the two grids differing from each other in the type of coating resins employed (stiffer for G3) [7]. They were glued with a 500 g/m2 bitumen emulsion applied on the bituminous material layers (E = 11,300 MPa at 15 °C and 10 Hz; υ = 0.35). Due to the field compaction conditions, the bituminous material displayed average void contents of 4% and 6% below and above of the grid respectively; it was composed, in accordance with French bituminous mix standard NF EN 13108-1, using an aggregate size of 0/10 and 5.6% bitumen of 35/50 penetration grade. Similarly to Chabot et al. [8], an aquarium designed for testing specimens under water was built (Fig. 1a). The specimens were first vacuum-saturated with water at a depressurization of up to 86 kPa (15 min in an air box, then 2 h in water). Next, they were stored in a water bath at around 20 °C before being tested under the static condition of a controlled 2 mm/min displacement. Considering "α" (equal to 7° here) as one-half the wedge angle, the splitting load can be written: FS = 2FH ; FH = FM /2tg(α). During testing, both the Crack Mouth Opening Displacement (CMOD) and crack propagation length were measured by means of the Digital Image Correlation (DIC) technique (Fig. 1b). From the WST load-displacement curve (splitting load (FS )—CMOD (δDIC )), Fig. 1b), the debonding propagation resistance of the interface is classically obtained by determining the specific fracture energy GF given in Eq. 1, i.e.:
Interlaminar Mode-I Fracture Characterization Underwater …
1121
(b)
(a)
Fig. 1 WST: a Specimen tested under water, b the 3 indicators of the FS —δDIC curve
GF =
SA SF
(1)
In Eq. 1, SF is the fracture area of the specimen and SA the splitting energy calculated from the area under the FS —δDIC curve (Fig. 2b). From this curve, results depending on the three values of FS , GF and δDIC can be compared as follows. From more than 100 prepared specimens, dedicated to studying the effects of water on the various types of interfaces, fracture debonding mode I tests were conducted first at ambient temperature in air and then under water at around 20 °C. Moreover, in order to obtain information on the temperature effect, the reinforced P7 specimens (i.e. with the G4 grid inserted) were tested under a colder water of around 7 °C. All the testing conditions have been repeated at least ten times. However, some of the first tests performed (not reported here) have been interrupted due to a bad geometry of the samples resulting from the sawing process. Table 1 summarizes the average moisture characteristics of each specimen type and underwater testing condition. Table 1 Moisture characteristics of specimens tested under water Specimen name
Test T (°C)
Dry mass (g)
Wet mass (g)
Absorbed water volume (cm3 )
Sr
Immersion time (h)
P8
~20°C
~15320
~15540
~157 (103–238)
~48 (31–71)
~45 (26–74)
P6 (G3)
~20 °C
~15130
~16132
~180 (118–250)
~52 (37–66)
~26 (21–28)
P7 (G4)
~20 °C
~14550
~14680
~130 (103–189)
~43 (34–61)
~41 (7–96)
P7 (G4)
~7 °C
~14551
~14900
~133 (83–265)
~42 (28–83)
~36 (18–98)
1122
M. Gharbi et al.
Each specimen was weighed both before and after being saturated with water. Its absorbed water volume could then be evaluated as a function of void content. According to European Standard NF EN 12697-12, the degree of saturation (Sr ) for each specimen was calculated and reported in Table 1.
3 Results and Discussion Figure 2 summarizes the WST results obtained both in air and under water at around 20°C. Under these two testing conditions, it is shown that regardless of the grid used to reinforce the bituminous material, i.e. G3 or G4 grid for a P6 or P7 specimen, respectively: • The maximum bonding value of FSMax is smaller for P6 and P7 specimens than for unreinforced P8 specimens, whose value reaches 100% in underwater tests (Fig. 2a); • The fracture energy GF obtained, for the same value of δDIC = 4 mm, is reduced for P6 and P7 specimens compared to unreinforced P8 specimens, whose GF value reaches 100% in air (Fig. 2b). These two results lead to concluding that compared to unreinforced P8 specimens, the interfacial bond characteristics of reinforced P6 specimens (with a G3 grid) and reinforced P7 specimens (G4 grid) are reduced by roughly 1/3 relative to the crack initiation process and by 2/3 relative to propagation resistance. The fracture energy GF values reported in Figs. 2b, c, for the same fracture area (i.e. SF = 0.021 m2 , corresponding to an average crack length of around 0.140 m, as measured by DIC), illustrate that the dispersion in results under water is greater for both P8 specimens and reinforced P6 specimens in comparison with reinforced P7 specimens. Delamination at the interface may have occurred in the tack coat for P8 and P6 specimens, while it clearly occurred in the P7 grid. Figure 3 shows that the G4 grid has broken in all parts of P7 specimen interfaces [5]. This outcome is also observed at 7 °C (Fig. 3b). From the literature, it has been shown that water and immersion times may exert a negative influence on the interlayer bond. This influence has been demonstrated
m Fig. 2 Comparison of results (normalized values): a FSMax , b GδFDIC =4 mm , c G=0.021 F
2
Interlaminar Mode-I Fracture Characterization Underwater …
1123
500 450 400 350 300 250 200 150 100 50 0
)
( P7-6a P7-6b P7-5a P7-5b P7-13b P7-13a P7-15b P7-16a average curve
0
2
(a)
GF (J/m2)
GF (J/m2)
Fig. 3 Fracture surface view of a reinforced P7 specimen tested under water: a ~20 °C, b ~7 °C
4 DIC
6
(mm)
8
10
(b)
500 450 400 350 300 250 200 150 100 50 0
)
( P7-9a P7-10a P7-8b P7-15a P7-10b average curve
0
2
4
6 DIC
8
10
(mm)
Fig. 4 GF results for reinforced P7 specimens tested under water: a ~20 °C, b ~7 °C
by means of shear fracture testing at hot temperatures (i.e. 40–60 °C) [9] as well as mixed-mode testing between cement and bituminous materials around 20 °C [8]. No clear effect of water has been found here, except on the dispersion of results. However, the dispersion of WST results, for the interfaces between such bituminous materials tested under fracture mode I at 20 and 7 °C, is not due to differences in the limited specimen storage time in water (Table 1). In fact, during the test, the water itself was responsible for altering tack coat strength among all specimens tested under water, except for P7. The G4 grid bonded to the tack coat seems to have altered bitumen resistance by further destroying the adhesion between the granular layer and the bitumen. Moreover, fracture surface observations at low temperature around 7 °C (Fig. 3b) highlight this phenomenon to a greater extent than observations at 20 °C (Fig. 3a). Indeed, the mode I fracture energy spent for debonding reinforced P7 specimen interfaces at 7 °C exceeds that spent at 20 °C (Fig. 4).
4 Conclusion In conjunction with four-point bending fatigue tests and accelerated loading tests (APT) conducted on several pavement structures throughout the course of the French
1124
M. Gharbi et al.
SolDuGri project (ANR-14-CE22-0019) [10], the WST has been developed specifically for specimens extracted from road sections [5, 6]. This approach has led to experimentally characterizing the Interlaminar Mode-I Fracture of the interlayer bond of pavement reinforced with fiberglass grids. To investigate the effect of moisture on the interlayer bond, a water immersion WST has also been proposed. In this paper, the WST campaigns were carried out in air and under water. The final fracture surface observations and the splitting force-CMOD curve results for 3 types of specimens have been presented. Two types of grids, denoted G3 and G4, were used; they only differ from each other in the coating resins applied. The resin used for the G3 grid is stiffer than the one used for the G4 grid [7]. Compared to the WST results of unreinforced specimens, the bond between bituminous layers when adding these two grids is reduced for both by roughly 1/3 relative to crack initiation and by 2/3 relative to propagation resistance. At 20°C, the presence of water in the materials exerts an influence on the dispersion of all results, except for specimens reinforced with the G4 grid. For the G4 grid, WST results under water at two temperatures (7 and 20 °C) illustrate the thermal effect on bonding behavior; also, it was shown that the G4 grid does indeed adhere to the bituminous material. The fracture mechanisms are localized inside the G4 grid itself and on the bond between aggregates and tack coat bitumen emulsion. This laboratory study yields greater understanding of some of the poor interface bond results obtained recently in the field during the APT campaign. Acknowledgements The work presented in this paper was sponsored by the French National Research Agency (ANR—SolDuGri project ANR-14-CE22-0019).
References 1. Nguyen, M.-L., Blanc, J., Kerzreho, J.-P., Hornych, P.: Review of glass fibre grid use for pavement reinforcement and APT experiments at IFSTTAR. Road Mater. Pavement Des. 14(1), 287–308 (2013) 2. Chabot, A., Buttlar, W.G., Dave, E.V., Petit, C., Tebaldi, G. (eds.): 8th RILEM International Conference on Mechanisms of Cracking and Debonding in Pavements. RILEM Bookseries, vol. 13, 1st edn, Springer Netherlands (2016). https://doi.org/10.1007/978-94-024-0867-6 3. Canestrari, F., D’Andrea, A., Ferrotti, G., Graziani, A., Partl, M.N., Petit, C., Raab, C., Sangiorgi, C.: Advanced Interface Testing of Grids in Asphalt Pavements. In: Partl, M., Porot, L., Di Benedetto, H., Canestrari, F., Marsac, P., Tebaldi, G. (eds.) RILEM State-of-the-Art Reports, vol. 24, pp. 127–202. Springer, Cham (2018) 4. Petit, C., Chabot, A., Destrée, A., Raab, C.: Recommendation of RILEM TC 241-MCD on interface debonding testing in pavements. Mater. Struct. 51(4), article 96 (2018) 5. Gharbi, M.: Caractérisation du collage des interfaces de chaussées par essais de rupture en mode I. ECN Ph.D. thesis, ED SPI (https://www.theses.fr/2018ECDN0037/document) (2018). 6. Gharbi, M., Nguyen, M.L., Trichet, S., Chabot, A.: Characterization of the bond between asphalt layers and glass fiber grid with help of a wedge splitting test. In: Loizos, A., Al-Qadi, I., Scarpas T. (eds.) BCRRA 2017, pp. 1517–1524. CRC Press, London, UK (2017). https:// doi.org/10.1201/9781315100333-201
Interlaminar Mode-I Fracture Characterization Underwater …
1125
7. Doligez, D., Godard, E., Gharbi, M., Chazallon, C., Hornych, P., Chabot, A., Nguyen, M.L.: For Sustainable reinforcements of infrastructures with glass fiber grids. In: XXVIth World Road Congress, 6–10 Oct, Abu Dhabi (2019). 8. Chabot, A., Hammoum, F., Hun, M.: A 4pt bending bond test approach to evaluate water effect in a composite beam. Eur. J. Environ. Civ. Eng. 21(sup1), 54–69 (2017) 9. Raab, C., Partl, M.N., Abd El Halim, AO.: Experimental Investigations of Moisture Damage in Asphalt. Int. J. Pavement Res. Technol. 5(3), 133–141 (2012). 10. Nguyen, M.L., Chazallon, C., Sahli, M., Koval, G., Hornych, P., Doligez, D., Chabot, A., Le Gal, Y., Brissaud, L., Godard, E. Design of reinforced pavements with glass fiber grids: from laboratory evaluation of the fatigue life and its FEM/DEM modelling to accelerated full-scale test. In: Chabot A., Hornych P., Harvey J., Loria-Salazar L. (eds.) Accelerated Pavement Testing to Transport Infrastructure Innovation, LNCE, vol. 96, pp. 329–338, Springer, Cham, (2020). https://doi.org/10.1007/978-3-030-55236-7_34.
Interphase Relations in the Characterisation of Bitumen Emulsion-Cement Composites Miomir Miljkovi´c , Andrea Graziani , and Chiara Mignini
Abstract Strategic research on bitumen emulsion-based materials is substantial for addressing numerous environmental effects of road infrastructure. The present specifications for bitumen emulsions are based on properties that are insufficient for a fundamental understanding of their relation to the physicomechanical behaviour of realistic coarse-graded materials for structural layers. The aim of the interlaboratory testing (ILT) of RILEM TC 280-CBE Task group 1 is to investigate the link between the intrinsic interphase relations and macroscopic physicomechanical behaviour of bitumen emulsion-cement composites. Thus, the objective of this individual research was creating the basis for the above ILT by experimental and analytical evaluation of mortar-based model systems, by investigating their physicomechanical behaviour, and by conducting the sensitivity analysis of the relations between phases. The experimental evaluation comprised compactability and volumetric evaluation of fresh mixtures, indirect tensile-based testing at different stages of curing, as well as the temporal measurement of the evaporation of water from specimens. Keywords Bitumen emulsion-cement composite · Mortar · Model system · Multiphase composition · Volume fractions · Physicomechanical behaviour
1 Introduction Strategic research on bitumen emulsion-based materials is substantial for addressing numerous effects of road infrastructure associated with energy consumption and M. Miljkovi´c (B) Faculty of Civil Engineering and Architecture, University of Niš, Aleksandra Medvedeva 14, Niš 18000, Serbia e-mail: [email protected] A. Graziani · C. Mignini Department of Civil and Building Engineering, and Architecture, Università Politecnica delle Marche, Via Brecce Bianche, Ancona 60131, Italy e-mail: [email protected] C. Mignini e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_143
1127
1128
M. Miljkovi´c et al.
CO2 emission, and is therefore of key importance for its sustainable future [1]. Present specifications for bitumen emulsions are based on properties that are insufficient for the fundamental understanding of their relation to the physicomechanical behaviour of realistic coarse-graded materials for structural layers. Establishing this link could be achieved by investigating the mesoscale of emulsion-based composites at the length scale of the fine aggregate matrix, which include mastic (bitumen emulsion, cement, and filler) and mortar (mastic and sand). The Task group 1 of RILEM TC 280-CBE “Multiphase characterisation of cold bitumen emulsion materials” organised an interlaboratory testing (ILT) aiming at creating a robust scientific framework for investigating the link between the intrinsic interphase relations and macroscopic physicomechanical behaviour of bitumen emulsion-cement composites. In this regard, the aim of this research was creating an experimental and analytical basis for the above ILT by defining the concept of mortar-based model systems, by investigating their physicomechanical behaviour, and by conducting the sensitivity analysis of the relations between characteristic phases. This aim is addressed by the following individual objectives: • Optimisation of the composition of the mortar model system; • Investigation of the feasibility of the adopted sequential mixing procedure; • Investigation of the feasibility of testing the mortar specimens at very early stages of curing, i.e. during the first few hours; • Sensitivity analysis of the relation between the composition of mastic and fine aggregate. The adoption of the composition of the model system for ILT considered the importance to have (1) minimum extrusion of water to minimise the difference between the designed and composition after compaction, (2) maximum compactability, (3) satisfactory mechanical strength to enable the testing at early stages of curing, and (4) fine aggregate composition as little as possible sensitive to the variations in other mixture variables.
2 Materials and Methods This research was performed using constituents with properties as close as possible to those to be used in ILT. The studied mortar mixtures were produced with cationic bitumen emulsion (C60B10) containing 60 % of base bitumen 70/100 and approximately 1 % of emulsifier, ordinary Portland cement CEM II/B-LL 32.5 R, and fine aggregate. The fine aggregate consisted of limestone filler (particles from 0 to 0.063 mm) and natural limestone sand (particles from 0.063 to 2 mm) with the water absorption of 1 % (EN 1096-6), comprising 5 and 95 % of the mineral mixture, respectively. The particle size distribution of the fine aggregate was adjusted to correspond to the distribution used in the previous research [2–4]. The comparison of the particle size distributions of the used fine aggregate and the distribution adopted for the interlaboratory testing (including the permitted range) is shown in Fig. 1. The
Interphase Relations in the Characterisation of Bitumen …
1129
Table 1 Variables defining the volumetric composition of the composite at initial stage Variable
Value A
B
C
D
A-5
B-5
Air voids, ϕ v /%
3
3
3
3
5
5
Mastic to mortar volume ratio, ϕ mast /ϕ mort
0.333
0.333
0.333
0.400
0.333
0.333
Bitumen to fines volume ratio, ϕ b /(ϕ f + ϕ c )
3
2
1.36
1.36
3
2
Cement to filler volume ratio, ϕ c /ϕ f
0.333
0.200
0.128
0.128
0.333
0.200
Fig. 1 Particle size distributions of the aggregate used in this preliminary research compared with the target line and the permitted range, as proposed for the interlaboratory testing
difference between these two distributions was considered minor enough and ensured the relevance of the results of this preliminary testing for the interlaboratory testing itself. Generally, the composition of mortar was completely defined by several variables. However, knowing the composition of emulsion, and the absorption of water by sand to be 1 % by mass, the set of these variables was reduced to four. For the optimisation and experimental evaluation, six different compositions presented in Table 1 were studied in this research. Mixtures A to C were designed by increasing the content of filler, the mixture D was designed by increasing the content of emulsion, and the mixtures A-5 and B-5 had the same composition as A and B, with the air void content increased to 5 % as the only difference.
2.1 Preparation and Testing of Specimens The mortar mixtures were produced by the sequential mixing in two steps [5]. In the first step, dry sand, filler, and water (to be adsorbed by sand) were mixed for about 1 min and stored in a sealed plastic bag for at least 12 h. In the second step,
1130
M. Miljkovi´c et al.
the obtained fine aggregate in the saturated surface-dry condition was mixed for approximately 1 min with cement and emulsion, and afterwards, stored in sealed conditions for about 20 min. On the one hand, this time interval was considered long enough for the thorough homogenisation of the mixture (primarily its bituminous part, easily observable by the mortar appearance) which allowed the development of some adhesion. On the other hand, it was low enough not to be affected by the variability in reactivity of cements. The series of three specimens with the target diameter and height of 100 and 50 mm, respectively, were compacted for each of the compositions by a gyratory compaction applying the vertical pressure of 600 kPa with 30 gyrations per minute, and the inclination of 1.25°. After compaction, the specimens were cured at the relative humidity of (70 ± 5) % and the temperature of (22 ± 3) °C on small plates. To quantify the loss of water, the mass of specimens was measured at 0.5, 1, 2, 3, 5, 24 h of curing. At 1 day of curing, the specimens were tested using a static indirect tensile setup at the temperature of 5 °C and the vertical displacement rate of (50 ± 2) mm/min.
3 Results The results of the average number of gyrations and the loss of material during the compaction are shown in Fig. 2. It was apparent that the increase in the content of filler and the consequential decrease in the emulsion to fines (cement and filler) ratio made the mixture significantly less compactable. On the contrary, the increase in the content of emulsion, as expected, drastically reduced the number of gyrations. The decrease in the content of emulsion and increase in the content of fines from the
Mass loss during compaction (%)
2.0 A
1.5 B
D A-5
1.0 B-5 C
0.5
0.0 Mortar
Fig. 2 Average number of gyrations and the loss of material during compaction
Fig. 3 Lost to total initial water mass ratio during the first day of curing
Water loss respect to total water (%)
Interphase Relations in the Characterisation of Bitumen …
1131
70 60 50 40
A B C D A-5 B-5
30 20 10 0 0.01
0.10
1.00
Curing time (Days)
mixtures A to C, as well as the increase in target void content from A to A-5 and from B to B-5 clearly reduced the content of the extruded material (assumed to be mainly consisting of water). However, the increase in the content of mastic from C to D almost doubled the extruded material. The extruded water was clear or whitish, suggesting that there were no bitumen droplets in it and that the coalescence of emulsion was mostly accomplished. The results on the water loss by evaporation over the first day of curing and the corresponding indirect tensile strengths are shown in Figs. 3 and 4, respectively. It can be observed that the relative differences in water loss between the mixtures were relatively small and the evaporation process accelerates during this very early age, contrary to what is typical for the later stages of curing [6–8]. The bitumen emulsion to fines ratio and the increase in the target void content decelerated evaporation. The indirect tensile strength at the first day of curing increased with the decrease in the bitumen emulsion to fines content (from A to C) and decreased with the volume fraction of mastic and the target void content. All obtained results we in the range expected for this type of composite [9] and seemed to be directly related to the extent of water evaporation. The relations between the bitumen to fines and the cement to filler volume ratios for different filler to fine aggregate mass ratios are shown in Fig. 5. The filler to fine aggregate volume ratio increased with the decrease in bitumen emulsion and was in the range from 6 to about 12 %. This was an important factor for the mechanical behaviour of composite and was more sensitive at lower bitumen to fines ratios.
4 Conclusions The selected compositions and the sequential mixing procedure allowed the production of homogeneous mixtures. The true initial composition was directly affected by the amount of extruded material, and thus the pore space in general. To minimise the difference between the designed and realised multiphase composition at the initial
1132 0.8
Indirect Tensile Strength (MPa)
Fig. 4 Overview of the indirect tensile strengths for the considered mortar mixtures
M. Miljkovi´c et al.
C
0.6
B A
D
0.4
A-5
B-5
0.2
0.0
Mortar
Fig. 5 Relations between the bitumen to fines and the cement to filler volume ratios for different filler to fine aggregate mass ratios, calculated for ϕ mast /ϕ mort = 0.333
stage of curing, as accurate as possible prediction of the amount of extruded material in the design of mixtures is necessary. Besides the expected effect of target void content after compaction, the extrusion was strongly dependent on the volumetric composition of mastic phase. This suggested the critical role of pore system and emulsion breaking during compaction, as well as of their relation as the function of the multiphase composition on the sub-mortar level. Although all the above phenomena took place in mastic, the obtained results indicated that the influence of the volume fraction of skeleton made up of sand particles, i.e. the thickness of the mastic around sand particles, was not to be neglected. Because of the different rates of evaporation at 1 day of curing, the compositions of residual binders differed between the mixtures, preventing a more reliable conclusion on the influence of the initial composition on mechanical properties. Based on these criteria, the composition A-5 was adopted as the most appropriate for ILT.
Interphase Relations in the Characterisation of Bitumen …
1133
References 1. Dorchies, P.T., Chappat, M., Bilal, J.: Environmental road of the future: analysis of energy consumption and greenhouse gas emissions. In Fiftieth Annual Conference of the Canadian Technical Asphalt Association (CTAA) Canadian Technical Asphalt Association (2005). 2. Miljkovi´c, M., Radenberg, M., Fang, X., Lura, P.: Influence of emulsifier content on cement hydration and mechanical performance of bitumen emulsion mortar. Mater. Struct. 50(3), 185 (2017). https://doi.org/10.1617/s11527-017-1052-4 3. Miljkovi´c, M., Radenberg, M.: Effect of compaction energy on physical and mechanical performance of bitumen emulsion mortar. Mater. Struct. 49(1–2), 193–205 (2016). https://doi.org/10. 1617/s11527-014-0488-z 4. Miljkovi´c, M., Radenberg, M.: Fracture behaviour of bitumen emulsion mortar mixtures. Constr. Build. Mater. 62, 126–134 (2014). https://doi.org/10.1016/j.conbuildmat.2014.03.034 5. Mignini, C., Cardone, F., Graziani, A.: Experimental study of bitumen emulsion–cement mortars: mechanical behaviour and relation to mixtures. Mater. Struct. 51(6), 149 (2018). https://doi.org/ 10.1617/s11527-018-1276-y 6. Miljkovi´c, M., Poulikakos, L., Piemontese, F., Shakoorioskooie, M., Lura, P.: Mechanical behaviour of bitumen emulsion-cement composites across the structural transition of the cobinder system. Constr. Build. Mater. 215, 217–232 (2019). https://doi.org/10.1016/j.conbui ldmat.2019.04.169 7. Raschia, S., Perraton, D., Graziani, A., Carter, A.: Influence of low production temperatures on compactability and mechanical properties of cold recycled mixtures. Constr. Build. Mater. 232, 117169 (2020). https://doi.org/10.1016/j.conbuildmat.2019.117169 8. Du, S.: Effect of curing conditions on properties of cement asphalt emulsion mixture. Constr. Build. Mater. 164, 84–93 (2018). https://doi.org/10.1016/j.conbuildmat.2017.12.179 9. Miljkovi´c, M., Radenberg, M.: Characterising the influence of bitumen emulsion on asphalt mixture performance. Mater. Struct. 48(7), 2195–2210 (2015). https://doi.org/10.1617/s11527014-0302-y
Introducing an Improved Testing Method to Evaluate the Fatigue Resistance of Bituminous Mortars Alexandros Margaritis, Georgios Pipintakos, Li Ming Zhang, Geert Jacobs, Cedric Vuye, Johan Blom, and Wim Van den Bergh
Abstract An improved test is proposed to determine the fatigue resistance of bituminous mortars by applying time-sweep cyclic (oscillatory) tests using a dynamic shear rheometer (DSR). To evaluate its applicability, a bituminous mortar mixture derived from a dense asphalt concrete (AC) mix was utilized. The objectives of this study were twofold; firstly, to design a new geometry that ensures higher repeatability of measurements and failure at a predefined zone. Secondly, to provide a methodology to evaluate the fatigue resistance of a mortar sample including the theoretical design of the mortar composition and the actual fatigue test protocol. The findings of this study present an efficient test procedure for the estimation of the fatigue life of bituminous mortars, with a consistent fracture zone, reproducible mortar samples and high coefficient of determination (R2 ) for the fatigue curves. Reflecting on other fatigue testing approaches, this theoretical and empirical study provides additional value towards standardization of fatigue testing in bituminous mortars. Keywords Bituminous mortar · Asphalt mortar · Fatigue resistance · FEM · DSR
1 Introduction The fatigue resistance of asphalt mixtures is a critical factor that must be determined to estimate the service life of asphalt pavements. The EN12697-24 describes different test setups for the estimation of the fatigue resistance of asphalt mixtures. Usually, hot mix asphalt fatigue tests are time- and material-consuming. For example, a 4- or 2- point-bending fatigue test (strain-controlled) typically requires sixteen working days and at least eighteen specimens, without considering the sample preparation time. Cracking can be caused by two main failure types: cohesive failure within the mesoscale phase (mortar/mastic) and adhesive failure in the interphase of aggregate and mortar/mastic [1]. Typically, the bituminous mortar phase is defined as the A. Margaritis (B) · G. Pipintakos · L. M. Zhang · G. Jacobs · C. Vuye · J. Blom · W. Van den Bergh EMIB Research Group, University of Antwerp, Antwerp, Belgium e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_144
1135
1136
A. Margaritis et al.
combination of bitumen, filler and sand, whereas mastic phase refers to the combination of bitumen and filler alone. In this study, the focus is on the fatigue resistance of bituminous mortars, as part of the potential cohesive failure zone. Previous studies have evaluated fatigue on the mesoscale and successfully attempted to relate it to the macroscale (asphalt mixture) fatigue characteristics [2, 3]. During the LOT (Life Optimization Tool) project, cylindrical mortar columns were introduced and evaluated to understand and model the ravelling mechanisms of porous asphalt [4]. Considerable research has been devoted to the improvement of fatigue testing for bituminous blends in mastic and mortar scale [5, 6]. A previous study demonstrated that the mineral aggregate skeleton is coated and bonded by mortar [7]; therefore, bituminous mortar can be considered as a more explicit mesoscale representation, than bituminous mastic. Following previous attempts to increase repeatability and predict fatigue performance accurately, this paper suggests a new column geometry to evaluate the fatigue resistance of bituminous mortars. The first objective of this research is to design a new geometry that ensures higher repeatability of measurements and promotes failure at a predefined zone. Secondly, to provide a practical methodology to fabricate and then evaluate the fatigue resistance of mortar samples using DSR oscillation tests.
2 Materials and Methods 2.1 Materials The composition of the bituminous mortar used in this research is based on a densegraded APO-B asphalt mixture (AC14 with performance requirements), a mixture used in base layers in Flanders (Belgium). In this study, the mineral skeleton introduced in the mortar is the filler together with the part of the aggregates passing the 0.5 mm sieve. The utilized composition of the mortar is presented in Table 1. Table 1 Bituminous mortar composition (% m/m)
Component
Percentage (%)
Limestone 6/14 ( 0, n η,2 < 1, λη,R−T,1 = ω2R−T,tr · λη,R−T,2 . Knowing R E and Iη , the VENoL model can be exploited by calculating the ∗ (Eq. 6) and its derivatives: modulus norm, phase complex stiffness modulus E M angle, … ∗ ∗ ∗ + i · EM = R E + i · ω · Iη = EM EM
(6)
3.2 Example with MANGIAFICO’s Experimental Data (2014) [6] In [6], T/C complex stiffness modulus tests were carried out by imposing a controlled sinusoidal deformation of amplitude 50 μm/m. The samples were cylindrical with a diameter of 75 mm and a height of 150 mm. The experimental data of the sample A.0.35-50.R4 were used here. The corresponding used bitumen binder has a grade 35/50 and composes 5.35% of the sample mass. The air gap is 3.7% of the volume. The experimental data were retrieved by graphical reading with the software “GetData Graph Digitizer”. The linear viscoelastic modeling parameters for the sample A.0.35-50.R4 are given in Table 1. The good fit of the model on the experimental data is observable in Fig. 2.
Modeling of T/C Complex Stiffness Modulus Test …
1347
Table 1 A.0.35-50.R4 sample modeling parameters WLF law
Tr e f = 14.2 ◦ C
C2 = 196.094 ◦ C
C1 = 30.860 × 10−4
Transition reduced pulsation
ω R−T,tr = 7.50
Real part R E (ω R−T )
Imaginary part I η,R−T (ω R−T )
rad/s
ω R−T ≤ ω R−T ,t r
ω R−T ≥ ω R−T ,t r
ω R−T ≤ ω R−T ,t r
ω R−T ≥ ω R−T ,t r
R E,low,1 = 4.5 MPa
R E,up,2 = 36,500 MPa
Iη,R−T,low,1 = 3.00 × 104 MPa s
Iη,R−T,up,2 = 2.40 × 106 MPa s
λ E,R−T,1 = 7.11 × 101 s/rad
λ E,R−T,2 = 2.50 × 104 s/rad
λη,R−T,1 = 9.88 × 103 s/rad
λη,R−T,2 = 1.80 × 102 s/rad
a E,1 = 0.220
a E,2 = −0.164
aη,1 = −0.206
aη,2 = 0.250
n E,1 = 2.850
n E,2 = 2.850
n η,1 = 0.090
n η,2 = −0.110
Fig. 2 Sample A.0.35-50.R4 (T/C complex stiffness modulus test, 50 μm/m)—Plot of experimental data and VENoL(T, ω) model in the graph “phase angle ϕ function of reduced pulsation ω R−T ” (left) and in the “Cole-Cole space” (right). The peak of the phase angle corresponds to the transition reduced pulsation ω R−T,tr
4 Creation of the VENoL(T, ω, ε0 ) Model 4.1 Operating Mode of the VENoL(T, ω, ε 0 ) Model It consists in applying the Time-Temperature-Amplitude Semi-Superposition Principle (TTASSP). First, the previous VENoL(T, ω) model must be fitting to a reference amplitude ε0,r e f that we choose. Then, this model is evolved in applying a new translation factor a A to the reduced pulsation ω R−T (Eq. 7) and a power coefficient b A to the upper bound Iη,R−T,up,2 (Eq. 8). It has been observed that a A and b A seem to follow a WLF-type law according to ε0 (Eqs. 9 and 10). Careful, these observations were done for the only experimental data found in literature for three complete Cole-Cole curves (part 4.2) and are valid only above the transition reduced pulsation. Further
1348
L. Coulon et al.
data will be needed later to confirm or evolve these non-linearity observations. ω R−T A = a A · ω R−T
(7)
bA Iη,R−T A,up,2 = Iη,R−T,up,2
(8)
−A1 · ε0 − ε0,r e f log a A = A2 + ε0 − ε0,r e f −B1 · ε0 − ε0,r e f log b A = B2 + ε0 − ε0,r e f
(9)
(10)
4.2 Example with GRAZIANI’s Experimental Data (2019) [7] For this study in [7], a bituminous mixture for wearing course with a bitumen binder 70/100 dosed to 5.3% of the total mass is used. T/C complex stiffness modulus tests were carried out for three different deformation amplitudes: 15, 30 and 60 μm/m. The samples were cylindrical, D-94 mm × H-120 mm. The experimental data of the sample S2 are used here, which have an air void content of 8.5% of the total volume. The non-linear viscoelastic modeling parameters for the sample S2 are given in Table 2. Figure 3, on the left, shows the good fit of the model. On the right, we observe the non-linearity direction in the Cole-Cole space for different temperatures and frequencies.
5 Conclusion The VENoL models fit the experimental data accurately. Ten parameters are needed for the VENoL(T, ω) model with ω R−T ≥ ω R−T,tr , sixteen for the complete VENoL(T, ω) model and fourteen for the VENoL(T, ω, ε0 ) model with ω R−T ≥ ω R−T,tr . Thus, the model is adaptable. The future work will consist in evolving this model to use it with DEM and in time domain.
Im. Part I η,R−T (ω R−T ) for ω R−T ≥ ω R−T ,t r Iη,R−T,up,2 = 2.63 × 106 MPa s λη,R−T,2 = 3.30 × 102 s/rad aη,2 = 0.260 and n η,2 = −0.085
Real part R E (ω R−T ) for ω R−T ≥ ω R−T ,t r
R E,up,2 = 27000 MPa
λ E,R−T,2 = 4.00 × 103 s/rad
a E,2 = −0.164 and n E,2 = 2.250
ε0,r e f = 30 μm/m
WLF law for b A ω R−T,tr,r e f = 6.10 × 10−3 rad/s
A1 = 1.020 B1 = −0.0042
ε0,r e f = 30 μm/m
WLF law for a A
Transition reduced pulsation for ε0,r e f
C1 = 30.860
Tr e f = 14.2 ◦ C
WLF law for aT
Table 2 S2 sample modeling parameters
B2 = 48 μm/m
A2 = 90 μm/m
C2 = 196.094 ◦ C
Modeling of T/C Complex Stiffness Modulus Test … 1349
1350
L. Coulon et al.
Fig. 3 Sample S2 (T/C complex stiffness modulus test, 15/30/60 μm/m)—On the left, plot of experimental data and VENoL(T, ω, ε0 ) model above the transition reduced pulsation ω R−T,tr in the Cole-Cole space. On the right, overlay in the Cole-Cole space of the non-linearity tests (0 to 100 μm/m over 50 cycles, for different temperatures at 10 Hz) and the complex stiffness modulus (50 μm/m) with the use of the VENoL(T, ω, ε0 ) model. To verify the TTSP, one test is done to (20 °C, 10 Hz) and another one to (8.1 °C, 0.1 Hz)
Acknowledgments The work presented in this article was sponsored by the French National Research Agency (ANR-MoveDVDC project, ref. ANR-17-CE22-0014-03).
References 1. Huet, C.: Étude, par une méthode d’impédance, du comportement viscoélastique des matériaux hydrocarbonés. Ph. D. thesis. Université de Paris, France (1963) 2. Sayegh, G.: Contribution à l’étude des propriétés viscoélastiques des bitumes purs et des bétons bitumineux. Ph. D. thesis. Université de Paris, France (1965) 3. Olard, F.: Comportement thermomécanique des enrobés bitumineux à basses températures. Relations entre les prop. du liant et de l’enrobé. Ph. D. thesis. ENTPE, Lyon, France (2004) 4. Carreau, P.J.: Rheological equations from molecular network theories. J. Rheol. 16(1), 99–127 (1972) 5. Yasuda, K.: Investigation of the analogies between viscometric and linear viscoelastic properties of polystyrene fluids. Ph. D. thesis. MIT, Cambridge, USA (1979) 6. Mangiafico, S.: Linear viscoelastic properties and fatigue of bituminous mixtures produced with RAP and corresponding binder blends. Ph. D. thesis. ENTPE, Lyon, France (2014) 7. Graziani, A., Cardone, F., Virgili, A., Canestrari, F.: Linear viscoelastic characterisation of bituminous mixtures using random stress excitations. Road Mater. Pavement Des. 20(sup1), 390–408 (2019)
Modeling of T/C Fatigue Test with Boundary Element Method and Linear Fracture Mechanics A. Dansou, S. Mouhoubi, Cyrille Chazallon, and M. Bonnet
Abstract This work presents the simulation of 3D crack propagation in samples in order to determine fatigue life. The modellings have been achieved by using MBEMv3.0: a fast software based on the Symmetric Galerkin Boundary Element Method (SGBEM) accelerated with the Fast Multipole Method (FMM) in 3D elasticity. Fatigue crack propagation has been simulated with Paris law. We present the simulations of a tensile/compression fatigue test on cylindrical samples of a semicoarse asphalt concrete considered as homogeneous, containing very small cracks. When the number of cycles increases, the cracks propagate, and we can observe a loss of rigidity of the sample. Parametric studies of the modelling parameters have been performed where the damage evolutions exhibit a typical shape that proves that, for asphalt concrete materials subjected to T/C fatigue test, the shape of the fatigue test curve is mainly governed by biasing effects at the beginning then by mechanical damage at the end. Keywords Fatigue life · Crack propagation · SGBEM · FMM
1 Introduction Fatigue cracks are among the major causes of structural degradation. Experimental methods have been widely used since a long time to predict their propagation. However numerical methods are very interesting because some destructive experimental tests, which can entail important costs and long delays can be avoided. This work presents the simulation of 3D crack propagation in order to determine fatigue life by using a fast-numerical software developed in our laboratory: MBEMv3.0. This software is based on two sophisticated methods of integral equations: The Boundary Element Method (BEM) and the Fast Multipole Method (FMM). A. Dansou (B) · S. Mouhoubi · C. Chazallon ICUBE, UMR 7357, INSA de Strasbourg, 24 boulevard de la Victoire, 67084 Strasbourg, France e-mail: [email protected] M. Bonnet POEMS, UMR 7231, ENSTA ParisTech, 828 boulevard des Maréchaux, 91120 Palaiseau, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_172
1351
1352
A. Dansou et al.
Although the Finite Element Method (FEM) is the best-known method in fracture mechanics, BEM has significant advantages, for example, the reduction by one of dimension. For a 3D problem, only the boundary and crack surfaces are discretized, rather than the volume. Moreover, meshes can easily be generated and design changes do not require a complete remeshing; this is very suitable for crack problems. One of the main drawbacks of BEM is that the solution matrix resulting from the formulation is unsymmetric and fully populated, whereas, the FEM matrices are usually much larger but sparse. An interesting approach of BEM is the Symmetric Galerkin BEM (SGBEM) which is based on a variational (weak) version of the integral equations. It entails double surface integrations but leads to matrix operators which exhibit symmetry and sign-definiteness. SGBEM is used in many works, see for example Gray and Paulino [1], Frangi [2], or Pham et al. [3]. Over recent decades, the performance of boundary analysis is further improved with the Fast Multipole Method (FMM) introduced by Rokhlin [4]. The classical bottlenecks of the BEMs caused by the fully populated matrix are alleviated as the FMM splits all element integrals into near-field and far-field interactions, the latter being clustered in a recursive, multilevel fashion. This process results in a lessened storage complexity, typically defined by the sparse near-field matrix, and faster solution based on iterative solvers. This makes boundary element analysis applicable to large BEM models with very good performance, see for example the works of Yoshida [5], Chaillat [6], Pham et al. [3] or Trinh et al. [7].
2 Crack Propagation Simulation by MBEMv3.0 Based on SGBEM and FMM, Pham [8] and then Trinh [9] developed a Fortran code for 3D elasticity problems, called MBEMv2.0. This software inherits a number of innovative algorithms from the boundary element community such as: the singular integration schemes, the index of severity for adjusting the Gaussian quadrature density to the interelement distance, the nested Flexible GMRES (FGMRES) which uses the near-interaction matrix as a preconditioner and the extension to multizone configurations of the SGBEM. These features and attendant performance enhancements are reported in [7, 9]. Matrix-vector operations are performed using subroutines from the BLAS library. The FGMRES algorithm used is available at www.cer facs.fr and described. A new version of the code noted MBEMv3.0, is developed by Dansou et al. in [10–12]. In this version, multiple strategies have been proposed and implemented for performance enhancement, including a data reusing technique, a shared memory parallelism using OpenMP directives and the proposal of a new sparse matrix method. The enhanced code has been run on various large-scale tests (N=O(106 )) and has proved to be very robust and of excellent accuracy. The time speedup compared to MBEMv2.0 has exceeded 50 in some crack propagation simulations. For example, ten propagation cycles of 64 circular cracks in a cube sample are simulated in two hours on a 20-core computer. The new version can deal with surface breaking cracks,
Modeling of T/C Fatigue Test with Boundary Element Method …
1353
Fig. 1 Crack front element in MBEMv3.0
complex multizone problems and proposed a direct coupling with FEM to simulate thin structures. In MBEM v3.0, the Paris law is used to simulate crack propagation. Abundant experimental evidence supports the view that the crack growth rate can be correlated with the cyclic variation in the stress intensity factors (SIFs), e.g. through: da (s) = AK m (s) dN
(1)
where s is the arc length coordinate along the crack front (Fig. 1), N is the number of loading cycles, da/dN is the fatigue crack advancement rate per cycle, ΔK = K max − K min is the SIF range for the current cycle, while A and m are parameters that depend on the material, environment, frequency, temperature and stress ratio. The SIFs are evaluated through extrapolation from the computed displacement discontinuity field. For example, K I is evaluated by: K I2
μ = 8(1 − ν)
2π 5 4φ − φ 1 · n a
(2)
where K I2 is the mode I SIF at the node 2, φ 1 , φ 5 are the nodal crack opening displacements at nodes 1 and 5, the nodes being numbered as shown in Fig. 1. The geometrical advance of the crack (Fig. 1) is described by moving points of the crack front in the local plane orthogonal to the front. The direction θ 0 of the crack advancement is assumed to be given by the maximum circumferential stress criterion: ⎤ ⎡ KIef f 2 1 ⎣ KIef f θ0 = tan − sign(K I I ) + 8⎦ 2 4 KI I KI I
(3)
where K I e f f = K I + B|K I I I | is an equivalent local mode I stress intensity factor which accounts for the tearing mode being active, B being a material parameter. The length Δa of the crack advancement is determined from the Paris law:
1354
A. Dansou et al.
m/2 a(s) = A K I2 (s) + K I2I (s) N
(4)
3 Fatigue Test on Cylindrical Samples 3.1 Model We consider cylindrical samples of a semi-coarse asphalt concrete considered as homogeneous (E = 17,000 MPa, ν = 0.3). The sample of dimensions R = 75 mm, H = 150 mm contains 64 randomly oriented cracks with initial radius of rc = 1 mm, see Fig. 2. The centers of the cracks are located on a grid of regular step in such way that they don’t intersect. The initial mesh features 3264 elements and 31,494 DOFs. The material properties for the Paris law are A = 10−8 mm/cycle and m = 4.5, see [13]. The tolerance for the iterative solver is set to 10−3 . The tensile load is σ = 1.7 MPa which corresponds to an initial strain ε = 10−4 .
(a) Model with rc=1mm Fig. 2 Model of a multi-cracked sample
(b) Cracks in propagation
Modeling of T/C Fatigue Test with Boundary Element Method …
1355
3.2 Results When the number of cycles increases, the cracks propagate, and we can observe a loss of rigidity of the sample. The decrease of the residual stiffness according to the number of loading cycles can be divided in two main stages. The first stage can correspond to the initiation of the crack propagation. In this study, very small (radius = 1 mm) initial cracks are generated. The residual rigidity at the initial state is equal to 99.5%. This version of our code does not yet include criteria to initiate the propagation. As a result, the propagation speed is very low at the beginning of the simulation, the fatigue of the material is governed by biasing effects in this part. Table 1 gives a summary of the residual stiffness, the number of cycles and the crack size during this first part. A propagation from a radius of 1–2 mm requires 5 million of loading cycles and the residual stiffness decreases from 99.54% to 99.33%. This very low speed is observed until a radius of 4 mm corresponding to a residual stiffness of 98.66%. This limit is very interesting since it can permit to determine the threshold value of stress intensity factor below which fatigue cracks do not propagate. This threshold can then be compared to values obtained with various crack propagation criteria or with experiments. For the second stage, we monitor in Fig. 3 the decrease of the residual stiffness. The final state of the first stage (cracked sample containing cracks of radius 4 mm) is taken as the initial state for this second stage (point A in Fig. 3). In Fig. 3, we clearly observe an acceleration of the rigidity loss, and can determine the fatigue life of the sample (number of cycles required to lose half of the initial rigidity). In this simulation, the fatigue life is 6.6 × 105 cycles after stage 1 (initiation of the propagation) and the cracks have radii of 12–13 mm at this state (point B in Fig. 3). At the end of this stage, we observe a very fast decrease of the residual stiffness. This can be assimilated to the third region observed in experiments when the stress intensity factor approaches the fracture toughness and the material fails. This limit can help to determine a numerical value of the fracture toughness which can be compared to values obtained with experiments. Table 1 Initiation of the crack propagation
Loading cycles
Crack radius
Residual stiffness (E/E0)
0
No crack
100%
–
1 mm
99.54%
5,185,400
2 mm
99.33%
6,419,225
3 mm
98.75%
6,910,735
4 mm
98.66%
1356
A. Dansou et al.
Fig. 3 Evolution of residual stiffness of a multi-cracked sample
4 Conclusion This paper has shown MBEMv3.0 can help to simulate T/C fatigue tests. A Multicracked cylindrical sample of asphalt concrete is simulated for the purpose. The fatigue test curve is obtained by using Paris law. The shape of the curve is governed by biasing effects in a first stage, and subsequently by mechanical damage. Acknowledgements This work is supported in part by the French National Research Agency (SolDuGri project ANR-14-CE22-0019) and in part by the region “Grand-Est, France”.
References 1. Gray, L.J., Paulino, G.H.: Symmetric Galerkin boundary integral formulation for interface and multizone problems. Int. J. Num. Meth. Eng. 40, 3085–3101 (1997) 2. Frangi, A.: Fracture propagation in 3d by the symmetric Galerkin boundary element method. Int. J. Fract. 116, 313–330 (2002) 3. Pham, D., Mouhoubi, S., Bonnet, M., Chazallon, C.: Fast multipole method applied to symmetric Galerkin boundary element method for 3D elasticity and fracture problems. Eng. Anal. Boundary Elem. 36, 1838–1847 (2012) 4. Rokhlin, V.: Rapid solution of integral equations of classical potential theory. J. Comput. Phys. 60(2), 187–207 (1985) 5. Yoshida, K.: Applications of fast multipole method to boundary integral equation method. Ph.D. thesis. Kyoto University, Japan (2001) 6. Chaillat, S.: Méthode multipôle rapide pour les équations intégrales de frontière en élastodynamique 3-D. Application à la propagation d’ondes sismiques, Ph.D. thesis. Ecole des Ponts ParisTech, Champs-sur-Marne, France (2008) 7. Trinh, Q.T., Mouhoubi, S., Chazallon, C., Bonnet, M.: Solving multizone and multicrack elastostatic problems: a fast multipole symmetric Galerkin boundary element method approach. Eng. Anal. Boundary Elem. 50, 486–495 (2015) 8. Pham, D.: Méthode multipôle rapide pour les équations intégrales variationnelles en élasticité tridimensionnelle et en mécanique de la rupture. Thèse. INSA de Strasbourg (2010) 9. Trinh, Q.T.: Modelling multizone and multicrack in three-dimensional elastostatic media: a fast multipole Galerkin boundary element method. Thèse. INSA de Strasbourg (2014)
Modeling of T/C Fatigue Test with Boundary Element Method …
1357
10. Dansou, A., Mouhoubi, S., Chazallon, C., Bonnet, M.: Modeling multi-crack propagation by the fast multipole symmetric Galerkin BEM. Eng. Anal. Bound. Elém. 106, 309–319 (2019) 11. Dansou, A., Mouhoubi, S., Chazallon, C.: Optimizations of a fast multipole symmetric Galerkin boundary element method code. Numer. Algorithms 1–22 (2019) 12. Dansou, A.: Fast multipole boundary element method approach for multicracked structures: application to road pavement reinforcement. Thèse. INSA de Strasbourg (2019) 13. Corté, J.F., Di Benedetto, H.: Matériaux routiers bitumineux: description et propriétés des constituants. Lavoisier (2004)
Modelling of Oxidative Ageing in Bitumen Using Thermodynamics of Irreversible Processes (TIP): Potential and Challenges Uwe Mühlich, Georgios Pipintakos, and Christos Tsakalidis
Abstract Diffusion-reaction models derived within the frame of thermodynamics of irreversible processes (TIP) for mixtures enjoy a number of advantages over purely phenomenological models. Potential and challenges of the TIP approach are discussed exemplary by means of a particular model focusing on spurt oxidation in bitumen. More specifically, film ageing experiments are simulated considering different bitumen compositions in terms of SARA fractions. The saturation times predicted for different compositions are contrasted, for instance, with initial asphaltene content and Gaestel index. The model predicts a decrease in saturation time for increasing asphaltene content, which is in line with experimental results. Keywords TIP · Oxidative ageing · SARA · Asphaltenes · Gaestel index
1 Introduction Ageing of bitumen is an unavoidable coupled reaction-diffusion phenomenon which results in stiffening of the material and successively in different distress types. Recently, a concept for modelling oxidative ageing which rests on the theory of thermodynamics of irreversible processes (TIP) has been proposed in [1]. It establishes direct links between continuum scale and underlying molecular structure by means of activity models expressed in terms of physically meaningful parameters like solubilities, molar volumes, etc. In addition, it can distinguish between different bitumen compositions in terms of SARA fraction. What is still missing from the aforementioned model is how model predictions are affected by variations in composition. This paper stresses the significance of TIP modelling by contrasting relationships between model predictions and experimentally reported findings. U. Mühlich (B) Facultad de Ciencias de la Ingeniería, Universidad Austral de Chile, Valdivia, Chile G. Pipintakos EMIB Research Group, University of Antwerp, Antwerp, Belgium C. Tsakalidis Aristotle University of Thessaloniki, Thessaloniki, Greece © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_173
1359
1360
U. Mühlich et al.
2 Diffusion-Reaction Model for Spurt Oxidation The model proposed in [1] considers a simplified version of the spurt mechanism proposed by [2, 3]. Briefly, the corresponding reaction can be described as follows; a dioxygen molecule comes initially close to a saturated ring, situated between two aromatic rings. Successively, two hydrogen atoms are taken off from the saturated ring, which converts the latter into an aromatic ring. Subsequent to this “aromatization”, the O2 and the two hydrogen atoms find a sulfur atom. A sulfoxide is formed and the remaining oxygen forms together with the two hydrogen atoms a water molecule. To incorporate this mechanism into a reaction-diffusion framework, the following five species are defined, numbered in the following by α = 1,…,5: dioxygen (1), water (2), aromatized compounds (ARA*) with one S=O (3), aromatizable compounds with traces of sulfur (ARA) (4) and saturates (5). Additionally, it is assumed that every ARA member can only participate once in the following simplified chemical reaction 1O2 + 1A R A → 1A R A∗ + 1H2 O
(1)
where “ARA” denotes an average molecule accounting for the effect of aromatics, resins and asphaltenes. Saturates are considered separately because they do not participate in the reaction given in Eq. (1). Due to Eq. (1), all molar fractions can be expressed by means of the molar fractions of dioxygen and water χ 1 , χ 2 and the initial molar fractions of saturates, χ50 . Assuming, that only O2 diffuses implies, that all other species remain, on average, at their spatial locations. Therefore, the model reduces to two essential mass balances which read in terms of molar concentrations ∂τ c1 +
1 div J = − M1 ∂τ c2 =
(2) (3)
with molar mass of O2 , M 1 , diffusion flux vector J and reaction rate density . The components of J with respect to a Cartesian coordinate system defined by coordinates x j , j = 1, 2, 3 read in terms of molar fractions as J1, j = L j1 χ1, j + j2 χ2, j
(4)
where χα,k indicates the gradient of χα with respect to the spatial coordinate x k and L is a diffusion parameter. The terms j1 , j2 in (2) are given by j1 = 1 +
∂ A15 ∂ A15 χ1 χ1 + χ1 , j2 = + χ1 1 − χ1 − χ2 ∂χ1 1 − χ1 − χ2 ∂χ2
(5)
Modelling of Oxidative Ageing in Bitumen Using Thermodynamics …
1361
Table 1 Parameters selected for the simulations, taken from [5, 6] Species
Density
Molecular weight
Molar volume
g cm3
g mol
cm3 mol
Solubility √ MPa
Dioxygen Water
1.114 0.988
32 18
28.0 18.2
14.0 48.0
44,272.2 151,790.4
Aromatics Resins Asphaltenes
1.066 1.054 1.200
440 990 4500
437.0 940.0 3750.0
21.0 19.0 20.0
Saturates
0.887
370
417.0
16.5
g cm s2
52,177.9
with 1 ∗ v1 [δ1 − δ]2 − v5∗ [δ5 − δ]2 (6) RT N N ∗ ∗ where δ = α=1 δα α and α = vα χα / β=1 vβ χβ . The solubility and molar volume of each species i is denoted respectively as δ i and vi∗ . To relate the reaction rate density with information about composition in terms of SARA fractions, the methodology proposed in [4] has been exploited here and the final result reads A15 =
∗ ∗ 2 2 v [δ − δ] [δ − δ] v 1 4 + 4 = B14 χ1 [1 − χ50 ][1 − χ1 − χ2 ] − χ2 exp 1 (7) RT RT where B14 is the reaction parameter, R is the ideal gas constant and T is the absolute temperature. For details regarding the derivation of Eqs. (4) and (7) the reader is referred to [1].
3 Compositions and Film Ageing Simulations To estimate the properties of the ARA-mixture with mass fractions β 4k is considered, where k = 1, 2, 3 refer to asphaltenes, resins and aromatics, respectively. The molar volume is determined by v4∗ =
3 k=1
∗ β4k v4k with β4k =
0 w4k 1 − w50
(8)
and the molar mass M 4 is computed analogously. For the solubility δ 4 , the arithmetic average δ4 = (δ41 + δ43 + δ43 )/3 is used. In addition, it can be assumed that M3 = M4 , v3∗ = v4∗ and δ3 and δ4 . The values used for all subsequent computations are given in Table 1.
1362
U. Mühlich et al.
Let consider now three groups, I, II and III, with w50 = 0.05, 0.10, 0.15, respectively. 14 virtual bitumen compositions are generated by varying the remaining reactive fractions systematically within possible ranges for the individual fractions proposed in the literature [7]. In addition, two real bitumen compositions, of bituminous binders A and B, are incorporated in this study, see Fig. 1. The remaining model parameters L and B14 (see Eqs. 2, 3 and 7) are chosen such that speed and the total duration of the spurt reaction coincide with experimental data, obtained from thin-film ageing tests performed for binders A and B. It should be noted that the same isothermal conditions (50 °C) were chosen for both the ageing treatment and the model simulations. Hence, the parameters were computed using back-calculation, by demanding a least square method between the experimental exp sim ) given in Eq. 9. spurt saturation time (t SST ) and the one from simulations (t SST sim sim exp 2 exp 2 − t SST B → min t SST − t SST A + t SST
(9)
To account for the fact, that the model captures only the spurt reaction whereas in real ageing experiments spurt and slow reaction occur, the procedure is based on the hypothesis, that the production rate of sulfoxides differs for the spurt and slow reaction. Echoing Petersen’s work [2, 8], this study demonstrated a rapid formation of sulfoxides, recorded with Fourier Transform Infrared spectroscopy (FTIR), followed by an almost constant slope. Regarding the model simulations, spurt reaction is supposed to be finished when dioxygen uptake reaches a plateau, referred to as saturation time in the following. That is, an association between the change of sulfoxide production rate and the saturation time can be speculated in Fig. 2. After back-calculating the governing parameters L and B14 for the model simulations, the partial differential Eqs. (2) and (3) were solved with the assistance of a finite element package [9]. What is striking from the performed model predictions is
Fig. 1 Compositions used in this study. Columns 6 and 12 refer to compositions of real bituminous binders A and B, respectively, whereas all other compositions are generated artificially
Modelling of Oxidative Ageing in Bitumen Using Thermodynamics …
1363
Fig. 2 Experimentally determined sulfoxide intensity and simulated O2 uptake as functions of ageing time for binder A [left] together with a schematic illustration of the ageing test setup [right]
that the so-called saturation time varied with different compositions. It comes naturally, motivated by the previous observation, to explore possible correlations of the saturation time with established experimental evidence during oxidative ageing.
4 Results and Discussion The correlation analysis demonstrates the potential of TIP modelling of oxidative ageing of bitumen, in order to distinguish between different bitumen compositions. As depicted in Fig. 3 the model predicts a linear decrease in saturation time with increasing asphaltene fraction [left] and Gaestel index I c [right], defined as the ratio of the sum of asphaltenes and saturates to the sum of resins and aromatics. There are two questions to be answered. First, why does the model predict these trends and second, do these trends reflect a real behaviour of bitumen? In order to understand the correlations in terms of TIP modelling one should keep in mind the following. All simulations have been performed with the same L and B14 which are expected to depend on pressure. Hence, pressure is kept constant. Density is kept constant too. With increasing asphaltene fraction, volume and mass of the average ARA molecule increase. Under these conditions, increasing asphaltene content implies less molecules per reference volume, i.e., less ARA molecules to be oxidized which in turn means lower saturation time. Although the increase of asphaltene percentage and I c through oxidation has been well documented, hitherto, this trend remained rooted in experimental level [10, 11]. The challenges arise when the hypotheses and predictions of the model are discussed from an experimental point of view. It appears that by varying the SARA compositions accordingly, the average ARA molecule could become more susceptible to oxidation effect as in reality this can be attributed to the generation of more reactive faces, with direct contact to O2 .
1364
U. Mühlich et al.
Fig. 3 Saturation time contrasted with the asphaltene content [left] and the Gaestel index [right]
As mentioned, the saturation time was found to decrease with increased asphaltene content. Assuming the same binder in two different ageing stages, according to experimental observations, the asphaltene content is evidenced to be increased in the extended ageing stage. Thus, it can be suspected that in the severe ageing stage a sufficient amount of O2 has at least reacted with the hypothesized average ARA molecule, indicating that the remaining time upon ideal saturation will be reduced compared with the initial ageing stage. Grouping also the compositions in terms of saturate content could at some extent support the secondary role of them with regards to the expected I c and successively the colloidal stability of the binders. However, is apparent that the saturate content, although is considered unreactive through ageing, could affect the saturation time as an increased tendency was observed with increased saturate content. Finally, the I c has been reported to increase with extended ageing time [10, 11]. Following the same rational perspective as previously for the asphaltenes, one could expect also a reduced saturation time in a more severe ageing stage than in the initial stage.
5 Conclusions This work differs from previous efforts to predict the oxidative ageing in bitumen in two key ways. Firstly, it accounts for different bitumen compositions and secondly points out the significance of the ageing mechanisms utilized in the developing model in such a way that a link with experimental observations can be finally, at least qualitatively, presumed. Overall, this study establishes a bridge between molecular scale and binder performance. The absence of the slow reaction, as well as spectroscopic techniques in order to support the underlying mechanisms in which the model is based on are still to be incorporated in a future study.
Modelling of Oxidative Ageing in Bitumen Using Thermodynamics …
1365
References 1. Mühlich, U.: Towards a continuum thermodynamics framework for mechanism-based modelling of oxidative ageing in bitumen. Int. J. Pavement Eng. 21:11, 1419–1428 (2020) 2. Petersen, J.C., Glaser, R.: Asphalt oxidation mechanisms and the role of oxidation products on age hardening revisited. Road Mater. Pavement Des. 12(4), 795–819 (2011) 3. Herrington, P.R.: Diffusion and reaction of oxygen in Bitumen films. Fuel 94, 86–92 (2012) 4. Pekar, M., Samohyl, I.: The Thermodynamics of Linear Fluids and Fluid Mixtures. Springer, Berlin (2014) 5. Powers, D.P.: Characterization and asphaltene precipitation modeling of native and reacted crude oils. Ph.D. thesis, University of Calgary (2014) 6. Akbarzadeh, K., Alboudwarej, H., Svrcek, W.Y., Yarranton, H.Y.: A generalized regular solution model for asphaltene precipitation from n-alkane diluted heavy oils and bitumens. Fluid Phase Equilib. 232(1), 159–170 (2005) 7. Hunter, R.N., Self, A., Read, J.: The Shell Bitumen Handbook, 6th edn., ICE Publishing (2015) 8. Petersen, J.C.: A review of the fundamentals of asphalt oxidation Chemical, Physicochemical, Physical Property, and Durability Relationships. TRB Transportation Research Circular EC140 (2009) 9. Alnaes, M.: The fenics project version 1.5. Arch. Numer. Softw. 3(100), 9–23 (2015) 10. Großegger, D.: Microstructural aging of bitumen contribution title. In: 6th Eurasphalt & Eurobitume Congress Proceedings, Congress, Prague (2016) 11. Mikhailenko, P., Baaj, H.: Comparison of chemical and microstructural properties of virgin and RAP binders and their SARA fractions. Energy Fuels 33(4), 2633–2640 (2019)
Monitoring of Railway Structures of High-Speed Line Bretagne-Pays de la Loire with Bituminous and Granular Sublayers Diana Khairallah, Juliette Blanc, Pierre Hornych, Louis-Marie Cottineau, Olivier Chupin, Jean-Michel Piau, Diego Ramirez Cardona, Alain Ducreau, and Frederic Savin Abstract Tamping and ballast wear, due to dynamic stresses, lead to frequent and costly maintenance operations. To mitigate this problem, an innovative track structure with an asphalt concrete layer under the ballast was built on the Bretagne-Pays de la Loire High-Speed Line (BPL HSL). It is intended, among other things, to reduce the amplitude of the accelerations produced at the passage of High-Speed Trains (HST), which are a major cause of ballast settlements. BPL HSL includes 105 km of asphalt concrete sub-layer under the ballast and 77 km of granular sub-layer. In order to study the dynamic responses of these different structures, four track sections were instrumented using, among others, accelerometers, strain gauges and temperature probes. In this paper, we present the track instrumentation, the acquisition system installed to collect all the measurements, and the comparative results of dynamic and mechanical behavior between structures since July 2017, when the BPL HSL was opened to commercial traffic. Keywords Ballasted track · Asphalt concrete · Monitoring · Sensors · Data acquisition
D. Khairallah (B) · J. Blanc · P. Hornych · L.-M. Cottineau · O. Chupin · J.-M. Piau IFSTTAR, Bouguenais, France e-mail: [email protected] D. Ramirez Cardona Eiffage Infrastructures, Corbas, France A. Ducreau SNCF Réseau, Nantes, France F. Savin Setec, Paris, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_174
1367
1368
D. Khairallah et al.
1 High-Speed Line BPL Tamping and ballast wear, due to dynamic stresses, lead to frequent and costly maintenance operations. It has been demonstrated that the settlements in the ballast are linked to the high accelerations produced in this layer by the passage of high-speed trains (HST). The solution with bituminous underlayment has been used since the 1970s in several countries like the United States, Italy, Spain, especially on hightraffic and high-speed lines (HSL). In France, the interest in this technique is rather recent. Following a first test on the East European HSL, which showed a satisfactory behavior, structures with an asphalt concrete sublayer under the ballast were built, the Bretagne-Pays de la Loire (BPL) high-speed lane. This solution is intended, among other things, to reduce the amplitude of the accelerations produced at the passage of the HST. On BPL HSL, the railway underlayment is made of granular materials on 77 km west of the track and with asphalt concrete on 105 km east of the route. The asphalt concrete is a class 4 grave-bitume (GB4) according to the NF EN 13-108-1 French standard, and the unbound granular material (UGM) is an A type with particle size 0/31.5 mm. The innovative track structure consists on a 15 cm thick UGM adjustment layer surmounted by a 12 cm thick GB4 layer. The standard track structure consists on a 35 cm thick treated soil surmounted by a 20 cm thick UGM capping layer. Both structures lay on a subgrade that was treated with lime and hydraulic binders. The different structures are shown on Fig. 1. Instrumenting several sections of the BPL line appeared to be an opportunity to study the behavior of different sections with granular and bituminous underlayment, and to compare their behavior. Accelerometers, weather stations, anchored displacement sensors, horizontal and vertical strain gauges, moisture probes and temperature sensors were installed in each section. The instrumentation is detailed in [1, 2]. The sensors are connected by cables to the data acquisition systems. The acquisition systems are based on PEGASE boards, developed by IFSTTAR [3]. In this article, only the measurements made with accelerometers and anchored displacement sensors are presented. All the measurement systems are powered by batteries and solar panels, for continuous year-round operation without interruption. All the data collected on each section is transferred continuously via a 4G network (within a few hours maximum) to a remote server. Thus, each system is autonomous in energy and data transmissions.
2 Data Processing The BPL track was first circulated, at increasing speed, during a test phase that began in November 2016 and lasted until January 2017. In order to analyze the measurements made during these test phase, different routines and functions were developed using the Scilab software. The train speed is first calculated, by calculating the time
Monitoring of Railway Structures of High-Speed Line …
1369
Fig. 1 Two structures of the BPL high-speed line
difference between signals of two accelerometers placed at different longitudinal positions, at the top of the sub-ballast layer. Then, the program performs low-pass filtering of the measurements, function of train speed, to eliminate dynamic effects [4, 5]. The adjusted filter allows to extract the peaks of all sensor signals. Once signals are filtered, the number of bogies can be determined. 4 types of high-speed trains have been identified: simple trains with 13 and 15 bogies, and double trains with 26 and 30 bogies (Fig. 2). The developed processing routines make it possible in particular to calculate the maximum and minimum values of the sensor signals (Fig. 3), which can then be used to compare the responses obtained on different sections, or for different speeds. The highest peak of deflection is determined for each train passage on each section. For accelerometer signals, a positive and a negative peak (corresponding to upward and downward accelerations respectively) are determined.
Fig. 2 Example of temporal deflection signals of single and double trains with a 15 bogies and b 30 bogies, having circulated on section 4 in July 2017
1370
D. Khairallah et al.
Fig. 3 Retained peaks from a deflection signals, b acceleration signals
3 Comparison of Response Between Granular and Bituminous Structures 3.1 Deflection Measurements The different signal treatment procedures were applied to the measurements recorded during a period of 23 months, after the opening of the line to commercial traffic. More than 16,000 HST passes were recorded during this period. All these passages were exploited and the maximum deflection values recorded by the anchored deflection sensors were determined. Measurements obtained on section 2 (with granular sublayer) and on sections 1 and 4 (with bituminous sub-layer), were compared. The soil (subgrade) on section 1 is composed of yellowish fine sand with water inflow detected at 4.5 m depth. Sections 4 and 2 are built on a yellowish clay soil with sand and scattered blocks. No water was detected on this site. Figure 4 compares the maximum deflection values U z recorded for all double 30-bogie trains that circulated on the three mentioned sections, starting from the track opening to commercial traffic (July 2017) up to May 2019. It can be seen that sections 2 and 4 have similar maximum deflection values, ranging between 25 and 42 mm/100, but with different evolutions in time. On the granular section, deflection peaks present seasonal variations: U z increases between September and May and decrease during the summer months. On bituminous section 4, the deflection evolution is quite different, the values decreasing during the first 7 months of track service, followed by a stabilization of deflection values, regardless of the season. On section 1 with bituminous sublayer and sandy soil with water infiltrations, unlike on sections 2 and 4 with clay soil, the maximum deflection values are significantly higher, of the order of 48 to 70 mm/100 (for double trains). Seasonal variations during the first 7 months of service are also more pronounced on this section.
Monitoring of Railway Structures of High-Speed Line …
1371
Fig. 4 Comparison between the maximum deflection values measured on section 2 with granular sublayer and sections 4 and 1 with bituminous sublayer
3.2 Acceleration Measurements In a second step, we compare accelerometers measurements obtained on section 2 with granular sub-layer and on section 4 with bituminous sub-layer, during the commercial phase. Figure 5 shows a comparison of the minimum and maximum accelerations for two accelerometers installed below the rail axis at the top of the sub-layer: AS2 on section 2 with granular sub-layer and A3 on section 4 with a bituminous sublayer. The results clearly show that acceleration peaks at the top of the GB4 sublayer are much lower than those measured at the top of the granular sublayer, both for upward and downward accelerations. It can be concluded that, during the commercial phase, the presence of the GB layer reduces by a factor of about 2 the accelerations measured under the ballast, as it was the case during the speed up test phase [2].
4 Conclusion The BPL HSL consists of two different structures, with granular and bituminous sublayers. It was instrumented with more than 100 sensors during its construction, to monitor the response of the two types of structures. This paper treats data collected during the speed-up test phase of the track and during the first 2 years of commercial traffic. The latter averages 23 HST per day. 4 HST types run on the track: single trains with 13 and 15 bogies, and double trains with 26 and 30 bogies. The analysis of this phase therefore required the processing of a large number of measurement files. Only a part of this data is presented in this paper, focusing on the evolution of
1372
D. Khairallah et al.
Fig. 5 Comparison of positive and negative vertical accelerations peaks measured by accelerometer AS2 on section 2 with granular sub-layer and accelerometer A3 on section 4 with bituminous sub-layer, during the commercial phase
maximum deflections and maximum and minimum accelerations over the duration of the observation period. The deflection analysis showed deflection levels higher on section 1 (sand subgrade) than on sections 2 and 4 (clay subgrade), probably due to differences in soil type and water content. Deflections’ seasonal variations were observed on sections 1 and 2, but not on section 4. This confirms the large influence of the subgrade soil on the track deflections. No change in the average level of deflection (which may reflect a degradation of the track) was observed during the 2 years follow-up, on either track section. The analysis of accelerations measured at the top of the sub-layers indicates significant reduction of accelerations on the bituminous sections, in comparison with the UGM section, by a factor of about 2. As for the deflection values, no evolution of acceleration levels was observed during the 2-year period, indicating a stable behavior of the track.
Monitoring of Railway Structures of High-Speed Line …
1373
References 1. Khairallah, D., Blanc, J., Cottineau, L.M., Hornych, P., Pouget, S., Hosseingholian, M., Ducreau, A., Savin, F.: Monitoring of railway structure with bituminous underlayment. In: 3rd International Symposium on Railway Geotechnical Engineering, Georail, Paris (2017) 2. Khairallah, D., Blanc, J., Cottineau, L.M., Hornych, P., Piau, J.-M., Pouget, S., Hosseingholian, M., Ducreau, A., Savin, F.: Monitoring of railway structures of the high-speed line BPL with bituminous and granular sublayers. Constr. Build. Mater. 211, 337–348 (2019). https://doi.org/ 10.1016/j.conbuildmat.2019.03.084 3. Le Cam, V., Lemarchand, L., Martin, W., Bonnec, N.: Improving wireless sensor behavior by means of generic strategies. In: Structural Health Monitoring, vol. 1, pp. 696–703, Sorrento, Naples, Italy (2010) 4. Khairallah, D.: Analyse et modélisation du comportement mécanique de structures de voies ferrées avec sous-couche bitumineuse [Analysis and modeling of the mechanical behavior of railway structures with bituminous underlayment]. Ecole Centrale de Nantes, Nantes, Ph.D. thesis (2019) (in French) 5. Khairallah, D., Chupin, O., Blanc, J., Hornych, P., Piau, J.-M., Ramirez Cardona, D., Ducreau, A., Savin, F.: Monitoring and modeling of railway structures of the high-speed line BPL with asphalt concrete underlayment. In: Transportation Research Board (TRB) 99th Annual Meeting, Washington, USA (2020)
Monitoring Stiffness Evolution of Asphalt Concrete Through Modal Analysis Anders Gudmarsson and Abubeker Ahmed
Abstract The stiffness of asphalt concrete, which is used for thickness design of pavements, is known to increase with time due to age hardening of the binder. Therefore, aging models to account for the stiffness evolution over time are in some applications considered with the aim to improve pavement life predictions. Modal analysis is an economic test method with a verified precision for characterizing the complex moduli of asphalt concrete. In this paper, the stiffness evolution of asphalt concrete has been monitored through modal analysis. The simplicity of the modal test setup facilitates accurate repeated testing under identical conditions. Therefore, modal testing was performed repeatedly at the temperatures of 20 and −20 °C over a period up to around 300 days. The testing was performed on samples in a controlled laboratory environment and does not intend to provide a measure of the age hardening of real pavements. The aim of this paper is to characterize the stiffness evolution of asphalt mixes with modified and unmodified binders in a controlled environment. Cyclic indirect tensile tests were performed in addition to modal analysis to characterize the stiffness evolution under different loading configurations. Results from the modal analysis clearly showed that the rate of the stiffness evolution is initially slower for the modified binder in comparison to the unmodified binder. Keywords Modal analysis · Complex modulus · Aging · Polymer modified Bitumen
1 Introduction The stiffness evolution of asphalt concrete due to age hardening can lead to premature failure of pavements [1–3]. The rate of the age hardening does not only depend on the properties of the materials, but also on important factors such as environmental A. Gudmarsson (B) Peab Asfalt AB, Stockholm, Sweden e-mail: [email protected] A. Ahmed The Swedish National Road and Transport Research Institute (VTI), Linköping, Sweden © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_175
1375
1376
A. Gudmarsson and A. Ahmed
conditions [2, 3]. This may limit a general characterization of the phenomena, since for example it has been shown that higher annual air temperature results in increased age hardening of asphalt compared to colder climates [3, 4]. However, comparative studies of asphalt concrete mixes under equal conditions can facilitate the design of pavement materials to reduce the risk of premature failure. In addition, characterizing the stiffness and the stiffness evolution directly after mixing and compaction can be useful to assess the risk of permanent deformation in newly paved asphalt concrete. Modal analysis is an accurate methodology to characterize the complex moduli of asphalt concrete [5, 6]. In this method, a miniature accelerometer (~0.6 g), which is glued to the asphalt concrete specimen with free boundary conditions, measures the vibrational response to an impact load. A small impact hammer is used to apply a transient load that excites the specimen over a wide frequency range. The measured acceleration (output) and impact force (input) provides frequency response functions (FRFs) that are used to determine the complex moduli. The frequency response depends on the stiffness, density, dimensions and the boundary conditions of the specimen. In this paper, a polymer modified and a conventional unmodified asphalt mix are measured and compared. The modal testing were performed repeatedly over a period up to around 300 days. The testing was performed on samples in a controlled laboratory environment and does not intend to provide a measure of the age hardening of real pavements. Cyclic indirect tensile testing (CITT) were performed according to EN 12697-26:2018 Annex F [7] in addition to modal analysis to characterize the stiffness evolution under different loading configurations. The aim of this paper is to characterize the stiffness evolution of asphalt mixes with modified and unmodified binder in a controlled environment.
2 Methodology Two dense graded asphalt concrete mixes (labeled Mix 1 and Mix 2) with nominal maximum aggregate size of 11 mm were laboratory produced using a gyratory compactor. Mix 1 and Mix 2 have a binder content of 5.9 and 5.8% by weight and an air void content of 1.5 and 2.1%, respectively. This may generally be considered as too low air void content. However, the air voids should be between 1.5 and 3.5% according to the Swedish specifications for this mix type (ABT 11). The difference between the two asphalt concrete mixes is that a 70/100 penetration grade binder was used in Mix 1 and a 40/100-75 SBS polymer modified binder was used in Mix 2. The modal testing were performed to disc-shaped specimens with a diameter and height of approximately 150 mm and 30 mm, respectively. The testing were performed at 20 °C and then repeated after 6 h of conditioning time at −20 °C. After the second measurements, the specimens were conditioned at 20 °C until the next measurements. Three specimens of each mix were measured up to 14 days under the same conditions inside a climate chamber. The measurements continued up to around
Monitoring Stiffness Evolution of Asphalt Concrete …
1377
Fig. 1 The increasing resonance frequency (fr) of specimen 1a at 20 °C due to age hardening
300 days on one sample of each mix. The disc-shaped specimens were measured by exciting the flexural modes of vibration [8]. The following equipment were used to measure the FRFs: an impact hammer (PCB model 086E80), an accelerometer (PCB model 352B10), a signal conditioner (PCB model 480B21), a data acquisition device (NI USB-6251 M series) and a computer. The measured impact force and acceleration in time domain were transformed to frequency domain by using the fast Fourier transform. FRFs were determined by calculating the ratio of the complex valued cross-spectrum (input and output) to the complex valued power spectrum (input) according to Eq. 1. H ( f ) = (Y ( f ) X ∗ ( f ) / X ( f )X ∗ ( f )
(1)
where H(f ) is the complex frequency response function, Y (f ) is the measured acceleration, X(f ) is the measured input force and X*(f ) is the complex conjugate of the input force. Five measurements of the impact force and response were used to determine an average of the complex FRFs for each specimen and measurement temperature. The complex moduli were characterized by fitting computed FRFs to the measurements [5, 9, 10]. FRFs were computed by using the finite element method and the Havriliak-Negami (HN) equation to model the linear viscoelastic material properties [9, 11].
1378
A. Gudmarsson and A. Ahmed
3 Results Figure 1 shows the measured fundamental resonance frequency (fr) for the specimen 1a of Mix 1 at 20 °C. The results show the normalized resonance frequency of measurements repeated up to more than 300 days. The measured resonance frequency systematically increases from day to day. The increasing resonance frequency is a result of increasing stiffness of the sample. Figure 2 shows the evolution of the fundamental resonance frequency from day 1 up to 14 days for the three specimens of Mix 1 and Mix 2. The two mixes resulted in a different evolution of the resonance frequency at especially 20 °C. Mix 1, with the conventional 70/100 binder, resulted in a larger initial increase of the stiffness for the first days after the production and compaction. Contrarily, the stiffness of Mix 2, with the polymer modified 40/100-75 binder, did not increase at 20 °C during the first 14 days. Furthermore, the increase of the stiffness for Mix 1 with the 70/100 binder were lower at −20 °C compared to at 20 °C. The stiffness of Mix 2 increased slightly at the temperature of −20 °C. These differences between the 70/100 and 40/100-75 mix were systematic for the three specimens of each mix at both temperatures. Previous work have reported that asphalt concrete with SBS polymer modified bitumen display smaller changes of the material properties with time compared to asphalt concrete with unmodified bitumen [12]. Figure 3 presents the results of the modal testing that was performed at 20 °C and −20 °C on one of the specimens of each mix up to around 300 days. Similarly, to Fig. 2, the evolution of the resonance frequency is compared to the resonance frequency measured at day 1. The results presented in Fig. 3 show that the stiffness of Mix 2 at 20 °C starts to increase after the first 14 days. In contrast to the other presented results in Fig. 3, the stiffness of Mix 2 at 20 °C increases linearly with time up to 300 days.
Fig. 2 The stiffness evolution of the three specimens of Mix 1 (●, and ) and Mix 2 (*, ▲ and ) measured by modal analysis up to 14 days
Monitoring Stiffness Evolution of Asphalt Concrete …
1379
Fig. 3 The stiffness evolution of one specimen of Mix 1 (●) and Mix 2 (*) measured by modal analysis up to around 300 days
The repeated measurements at 20 °C also show that differences in the age hardening between Mix 1 and Mix 2 may reduce with time. Likewise, the differences between Mix 1 and Mix 2 at −20 °C reduces with the repeated measurements up to around 300 days. Table 1 present the modal analysis determined dynamic moduli at a loading frequency of 10 kHz and the stiffness increase in percentage. Note that the binder classified to a 40/100-75 result in a lower mix stiffness at 10 kHz compared to the mix with the 70/100 binder. Figure 4 shows the averaged stiffness evolution in percent measured at 20 and − 5 °C by the CITT on the same three specimens of each mix. The testing does not capture the stiffness evolution directly after compaction, since the testing started at day 7 after the production and continued once a week up to one month. The testing then continued less frequently up to almost 400 days. The CITT does not result in the same systematic stiffness evolution as the results of the modal analysis and no clear trend is obvious for the first 50 days at 20 °C. However, similar to the modal Table 1 Evolution of the dynamic moduli |E*| measured by modal analysis Temperature
20 °C
−20 °C
Specimen
|E*| at 10 kHz (MPa) (increase in %) Day 1
Day 14
Day 154
Mix 1 70/100
22863 (–)
23379 (2.3)
24296 (3.9)
Mix 2 40/100-75
20897 (–)
20889 (0)
21916 (4.9)
Mix 1 70/100
37549 (–)
38083 (1.4)
38535 (1.2)
Mix 2 40/100-75
35351 (–)
35633 (0.8)
36194 (1.6)
Day 291
Day 316 24779 (2.0)
22333 (1.9)
8.5 6.9
38669 (0.3) 36348 (0.4)
Total increase (%)
3.0 2.8
1380
A. Gudmarsson and A. Ahmed
Fig. 4 The averaged stiffness evolution measured by CITT at a loading frequency of 2 Hz
analysis, the higher temperature results in higher stiffness increase after 300 days and Mix 2 show the smallest total stiffness increase at the colder temperature.
4 Conclusions The evaluation of the age hardening of Mix 1 and Mix 2 showed that the stiffness evolution depend on the temperature and binder type. In the case of Mix 2 with the modified binder, the stiffness measured by modal analysis did not increase during the first two weeks at 20 °C. Contrarily, the stiffness of Mix 1 with the unmodified binder increased with the highest rate during the first two weeks after production at 20 and −20 °C. The modal testing clearly showed differences in the initial stiffness evolution of Mix 1 and Mix 2. However, the differences in the age hardening between the mixtures reduced over a longer period. It was not possible to draw conclusions on different or similar aging characteristics between Mix 1 and Mix 2 from the CITT measurements at 20 °C. However, the modal testing of the three specimens of each mix provided repeatable results of the stiffness evolution at both 20 and −20 °C.
References 1. Airey, G.D.: State of the art report on ageing test methods for bituminous pavement materials. Int. J. Pavement Eng. 4(3), 165–176 (2003) 2. Das, P.K.: Ageing of asphalt mixtures: micro-scale and mixture morphology investigation. Ph.D. thesis. KTH Royal Institute of Technology (2014) 3. National Academies of Sciences, Engineering, and Medicine: Environmental Effects in Pavement Mix and Structural Design Systems. The National Academies Press, Washington, DC (2007). https://doi.org/10.17226/23244
Monitoring Stiffness Evolution of Asphalt Concrete …
1381
4. Dickson, E.J.: The hardening of Middle East petroleum asphalts in pavement surfacing. J. Assoc. Asphalt Paving Technol. 49, 30–63 (1980) 5. Gudmarsson, A., Ryden, N., Di Benedetto, H., Sauzéat, C.: Complex modulus and complex Poisson’s ratio from cyclic and dynamic modal testing of asphalt concrete. Constr. Build. Mater. 88, 20–31 (2015) 6. Gudmarsson, A., Carret, J.-C., Pouget, S., Nilsson, R., Ahmed, A., Di Benedetto, H., Sauzéat, C.: Precision of modal analysis to characterise the complex modulus of asphalt concrete. Road Mater. Pavement Des. 20(sup1), S217–S232 (2019). https://doi.org/10.1080/14680629.2019. 1590224 7. European Standard EN 12697-26:2018: Bituminous mixtures—test methods—Part 26: stiffness (Annex F), CEN, 26 February 2018 8. Gudmarsson, A.: Resonance testing of asphalt concrete. Ph.D. thesis. KTH Royal Institute of Technology (2014) 9. Gudmarsson, A., Ryden, N., Birgisson, B.: Characterizing the low strain complex modulus of asphalt concrete specimens through optimization of frequency response functions. J. Acoust. Soc. Am. 132(4), 2304–2312 (2012) 10. Carret, J.-C., Pedraza, A., Di Benedetto, H., Sauzéat, C.: Comparison of the 3-dim linear viscoelastic behavior of asphalt mixes determined with tension-compression and dynamic tests. Constr. Build. Mater. 174, 529–536 (2018) 11. Havriliak, S., Negami, S.: A complex plane analysis of α-dispersions in some polymer systems. J. Polym. Sci. C 14, 99–117 (1966) 12. Lu, X., Said, S., Carlsson, H., Soenen, H., Heyrman, S., Redelius, P.: Performance evaluation of polymer modified Bitumens on a heavily trafficked test road. Int. J. Pavement Res. Technol. 7(6), 381–388 (2014)
Multi-criteria Framework to Evaluate the Oxidative Aging Resistance of Bitumen Binders Jeramie J. Adams, Michael D. Elwardany, Jean-Pascal Planche, Yvong Hung, Jeanne Zhu, Soenke Schroeder, and Mouhamad Mouazen
Abstract A multi-criteria framework, using both traditional and specially-designed long-term aging protocols, were utilized in order to discriminate binder formulations with respect to aging resistance due to oxidation. Two protocols were explored: a more traditional type of extended PAV under standard temperature and pressure conditions for durations up to 72 h, and an ambient pressure precision oven test developed by the Western Research Institute (WRI) under work with the Federal Highway Administration (FHWA). This second protocol ages 100 μm thick films at 70 °C for durations up to 84 days. By utilizing thin films, various complications due to oxygen diffusion effects are nearly eliminated. The methodology uses several indicators from infrared spectroscopy and rheology to fit an advanced oxidation kinetics model, also developed by WRI with the FHWA. This oxidation model allows determination of rate constants for the first (fast) and secondary (slow/constant) reactions, and also the amount of reactive material in bitumen. In this study, the overall aging resistance methodology relied heavily on changes in rheological behavior on the whole bitumen temperature range. Special emphasis was put on the mid- to low-temperature rheology, to determine if the modifiers reduce the non-load related surface cracking potential through the determination of various rheological aging sensitive parameters including Tc , the rheological index (R), Black space parameters and crossover frequency and temperature. Overall, the impact of binder formulation on oxidative ageing resistance was clearly observed and confirmed through this multi-criteria approach which allowed clear discrimination between various formulations and the efficiency of modification with respect to given oxidation indicators. Results showed that it was possible to differentiate between additives that chemically retard oxidation from additives that merely act to soften the bitumen.
J. J. Adams (B) · M. D. Elwardany · J.-P. Planche Western Research Institute, Asphalt and Petroleum Technologies, 3474 North 3rd Street, Laramie, WY 82072, USA e-mail: [email protected] Y. Hung · J. Zhu · S. Schroeder · M. Mouazen Bitumen Research Team, Centre de Recherche de Solaize, TOTAL Marketing and Services, Solaize, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_176
1383
1384
J. J. Adams et al.
Keywords Oxidation · Ageing resistance · Additive · Modelling · Binder characterization
1 Introduction Global consciousness for sustainability is growing, and will become more important in the future, especially as the values and risks of materials are increasingly assessed by life cycle analysis—which is also occurring for the paving industry. From a more pragmatic standpoint, the longevity of the pavement is directly related to the cost of maintaining roadways. Recent strides have been made to improve pavement longevity through pavement construction, matching bitumen physical properties to climate and traffic conditions, and relating bitumen chemistry to oxidation. Oxidation, however remains a critical parameter inherent to bitumen that remains relatively uncontrolled, unlike other industries, such as antioxidants in polymers. Long-term oxidation of bitumen in pavement is a major chronic failure mechanism leading to the degradation of pavement performance, regardless how well it is constructed [1]. Certainly, more should and can be done to reduce the effects of oxidation in bitumen. TOTAL and the Western Research Institute (WRI) have been actively working separately and together, to understand and develop methods and products that can reduce the impacts of chemical oxidation in bitumen. In this paper, results are presented showing the effects of different bitumen formulations on long-term aging by two different conditioning methods and how the effects on chemical oxidation can be distinguished from mechanical softening.
2 Experimental Plan A sample set was designed to include a base bitumen from a European commercial grade (35/50 penetration) product that was modified with 4 different additives at less than 5 wt%, and the same base bitumen formulated as a soft bitumen. Within this sample set, one experimental bitumen was also included (Table 1). For brevity only the Control and Bitumen 1, 5 and 8 will be discussed. Bitumen samples were aged by two different conditioning methods to simulate accelerated long-term aging. A more rigorous conditioning was performed using extended pressure aging vessel (PAV) at 16, 24, 40 and 72 h, whereas a more inservice aging was performed using thin-film (100 μm) oven aging (TFOA) at 70 °C for 3,15, 28, 63, and 84 days [2]. No short-term aging was done to avoid viscosity related issues during RTFOT. Chemical oxidation was determined by FTIR evolution of the carbonyl (C=O) and sulfoxide (S=O) content [1–3]. C=O + S=O data were used to fit the following oxidation kinetics model developed at the WRI and others [2]:
Multi-criteria Framework to Evaluate the Oxidative …
1385
Table 1 Select data for samples used to evaluate aging resistance Sample
Type of modification
Class
PEN (dmm)
SP (°C)
Control
Commercial bitumen (35/50)
35/50
38
52.5
Bitumen 1
Control + SBS polymer additive
25/55-60
30
64
Bitumen 2
Control + SBS polymer additive
25/55-50
35
61
Bitumen 4
Control + additive
35/50
36
52.2
Bitumen 5
Softer base of the control (160/220)
160/220
185
40.2
Bitumen 6
Control + additive
35/50
38
54.2
Bitumen 8
Experimental bitumen formulation
35/50
43
57.8
[C=O + S=O] = R M × 1 −
k2 × pon2 k1 ×
pom2
1−e
−k1 × pm o ×t 2
+ k2 × pon2 × R M × t
The model accounts for a fast initial rate of spurt oxidation (k1 ) followed by a longer-term slower rate (k2 ) of oxidation, which occurs more steadily [1, 2, 4–6]. RM is the reactive material in the bitumen, which is a class of compounds most sensitive to oxidation, and is source and condition dependant. This study showed that the RM is sensitive to bitumen modification with additives. Bitumen with higher RM produce more C=O + S=O at the same relative aging conditions as another bitumen with a lower RM. In the above equation, PO2 is the oxygen partial pressure, t is aging time, m and n are the pressure exponents for the fast and slow reactions, respectively. Rheological properties were measured by DSR using 25, 8 and/or 4 mm parallel plates to determine Superpave PG temperatures, generate the Black space plots, and calculate various aging sensitive rheological parameters such as Tc (Tc = Tc (S) – Tc (m)), where S = stiffness and m-value, rheological index (R), and crossover frequency (ωc ) and temperature (Tx ), as well as other parameters not reported here [3–5]. The 4 mm parallel plate method, developed at WRI [7], was used to evaluate small bitumen samples (25 mg) under low temperature sweeps. The ability to use such a small sample size allows determination of rheological properties of TFOA samples (ca 400 mg).
3 Results and Discussion It has long been known that as bitumen oxidizes it incorporates oxygen through a variety of oxidation pathways. Ultimately, the oxygen uptake is manifested as an increase in carbonyl and sulfoxide functionalities that can be quantified using FTIR. The aging conditions are important when interpreting FTIR measurements during severe aging, such as when using temperatures above 100–110 °C which begin to favour other oxidation pathways and generate products not observed at in-service pavement conditions [1, 8]. Figure 1 shows the aging evolution of C=O and C=O + S=O absorbance for the Control and Bitumen 1, 5 and 8. Both aging
1386
J. J. Adams et al.
methods show that Bitumen 8 has a much lower susceptibility to uptake oxygen. However, PAV shows significantly less oxygen uptake for Bitumen 1 than with TFOA, when compared to the Control. This may be linked to accelerated degradation of the SBS unsaturated polymer reacting with oxygen under more rigorous PAV aging conditions. From the oxidation model the RM increased: PAV = Bitumen 1 < 8 < 5 < Control; TFOA = Bitumen 8 < 1 < 5 < Control. Oxidative aging also has a significant impact on Tc and the crossover temperature. Figure 2 shows the evolution of these parameters. Due to their sensitivity towards oxidation, stress relaxation and healing propensity, both parameters are being considered for cracking indicators. For some of the data, the soft formulation of the Control shows differences, but no others. Black space and master curve data give insight into the rheological properties of bitumen and modifications. Changes in these curves upon oxidation provide insight into the magnitude of the impact of oxidation, as shown in Fig. 3. Black space interpretation is complicated by rubbery plateau for Bitumen 1, by feathering for 8 (no polymer) and by even more discontinuities due to the soft nature of Bitumen 5, however clear distinction is made from the master curves which shows Bitumen 1 and 8 intersecting the Control before aging, but after aging having more curvature than the Control. Through developments in the Christensen-Andersen (CA) model, changes in the master curves can be quantified using the Rheological Index (R) and the crossover frequency. As R increases the curve becomes flatter and as the crossover frequency decreases the sample becomes stiffer. Figure 4 shows that for PAV data Bitumen 1 and 8 are more resistant to changes upon oxidation than the Control or Bitumen 5.
Fig. 1 Left plots show the change ( = aged − unaged) FTIR C=O (top) and C=O + S=O (bottom) absorbances for different PAV aging times. Likewise, right plots show TFOA results
Multi-criteria Framework to Evaluate the Oxidative …
1387
Fig. 2 Top plots show the change ( = aged − unaged) in Tc for the PAV aged samples (left) and the TFOA samples (right). Likewise, bottom plots show in crossover temperature
Fig. 3 Top plots show Black space plots for 72 h PAV aging (left) and for the unaged (right) samples. The bottom shows master curves after 72 h PAV aging (left) and unaged (right)
1388
J. J. Adams et al.
Fig. 4 The left shows the change in R as a function of PAV aging and the right shows the relationship between R and the crossover frequency. Note CA model is most ideal for unmodified bitumen
4 Conclusion Significant strides have been made to produce Bitumen formulations that show resistance towards oxidative aging, although some aging resistance indicators appear to be sensitive to the method of oxidation. Some aging indicators show strong evidence of lower aging during PAV aging but less of an effect under more in-service like TFOA conditions. The main conclusions are: • From PAV aging, FTIR shows that Bitumen 1 and 8 were significantly resistant to taking up oxygen, however by TFOA, Bitumen 1 displays less oxidation resistance. This may be linked to SBS degradation of Bitumen 1 under PAV aging in such conditions. • Tc and crossover temperature show that Bitumen 8 and 1 had the least detriment to these cracking parameters, however Bitumen 5 also appeared to perform well. The performance of Bitumen 5 is likely tied to its much softer nature. • Rheological shape parameters from the CA model do show that Bitumen 8 and Bitumen 1 undergo less flattening upon oxidation than the Control and Bitumen 5. • Advanced oxidation kinetics based on PAV and TFOA data do show that Bitumen 1 and 8 significantly outperform the Control and Bitumen 5. However, the ranking between Bitumen 1 and 8 is related to the aging method. Acknowledgments This work was supported by TOTAL. Appreciation is extended to M. MERCE, formerly TOTAL and P. Coles, J. Forney, and J. Loveridge, from WRI for experimental support.
Multi-criteria Framework to Evaluate the Oxidative …
1389
References 1. Petersen, J.C.: A Review of the Fundamentals of Asphalt Oxidation. Transportation Research E-Circular (E-C 140), TRB, Washington, D. C. (2009) 2. Glaser, R., Turner, T.F., Loveridge, J.L., Salmans, S.L., Planche, J.P.: Aging Mastercurve (FP 10) and Aging Rate Model (FP 11), Fundamental Properties of Asphalts and Modified Asphalt, vol. III. Technical White Paper, WRI, Laramie, WY (2015) 3. Glover, C.J., Han, R., Jin, X., Prapaitrakul, N., Cui, Y., Rose, A., Lawrence, J.J., Padigala, M., Arambula, E., Park, E.S., Martin, A.E.: Evaluation of Binder Aging and Its Influence in Aging of Hot Mix Asphalt Concrete. Publication FHWA/TX-14/0-6009-2 (2014) 4. Anderson, M., King, G., Hanson, D., Blankenship, P.: Evaluation of the relationship between asphalt binder properties and non-load related cracking. J. Assoc. Asphalt Paving Technol. 80, 615–649 (2011) 5. Garcia Cucalon, L., Kaseer, F., Arámbula-Mercado, E., Epps, A., Martin, N.M., Pournoman, S., Hajj, E.: The crossover temperature: significance and application towards engineering balanced recycled binder blends. RMPD (2018) 6. Elwardany, M.D., Yousefi Rad, F., Castorena, C., Kim, Y.R.: Climate-, depth-, and time-based laboratory aging procedure for asphalt mixtures. AAPT 87, 467–512 (2018) 7. Federal Highway Administration (FHWA): Four-mm Dynamic Shear Rheometry. FHWA TechBrief, Publication No. FHWA-HRT-15-053, US Department of Transportation 8. Rad, F.Y., Elwardany, M.D., Castorena, C., Kim, Y.R.: Investigation of proper long-term laboratory aging temperature for performance testing of asphalt concrete. Constr. Build. Mater. 147, 616–629 (2017)
Multiphysics Simulation and Validation of Field Ageing of Asphalt Pavements Eman Omairey and Yuqing Zhang
Abstract Long-term field ageing of asphalt pavement plays a vital role in limiting the pavements’ service life. Models have been established to represent the multiple physics that contribute to the ageing of asphalt pavement, including: (1) heat transfer to determine pavement temperature profile, (2) diffusion of oxygen from the air into the connected air voids of the asphalt pavement, (3) diffusion of oxygen from the air void channels to the inside of the asphalt mastic coating films, and (4) the growth of oxidation products in the asphalt binders. These four ageing-related physics were mathematically modelled individually in the literature; however, they were not effectively integrated and coupled into a comprehensive computational model. The challenge lies in that the ageing-related physics are circularly dependent and time-dependent. Another challenge results from the complexity in numerical modelling of the high nonlinearity caused by the circular dependency between the four physics. This study uses weak-form partial differential equation (PDE)-based finite element (FE) program to efficiently couple the physics into one integrated model. The model inputs include the available site-specific hourly weather data, binder oxidation kinetics, mixture design properties, and thermal characteristics of pavement materials. The model was validated using the field measurements of the oxidation products (carbonyl) from The Federal Highway Administration (FHWA) reports. Results show that the model can effectively address the circular dependency between the ageing-related multiphysics and reliably predict the oxidation products along pavement depth for asphalt pavement road section. Keywords Long-term ageing · Carbonyl growth · Oxygen diffusion · Multiphysics modelling · Oxidative ageing
E. Omairey · Y. Zhang (B) Aston Institute of Materials Research, Aston University, Birmingham B4 7ET, UK e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_177
1391
1392
E. Omairey and Y. Zhang
1 Introduction Oxidative ageing in asphalt pavements is the reaction between bitumen constituents and atmospheric oxygen [1, 2]. Previous studies indicate a multiphysics nature of the oxidative ageing of asphalt pavements [3, 4]. Specifically, the ageing of asphalt materials involves three multiphysics: (1) oxidation reaction; (2) mechanical responses (e.g., altered viscosity); and (3) environmental processes such as heat transfer and oxygen diffusions leading to varying environmental profiles within pavements. However, most of the existing studies only focused on “one-way” or “one physics” of the ageing, which ignored the interaction among these physics. The fact is that the interrelationships of these physics are circularly dependent on the dynamic ageing process. For example, oxygen distribution affects the chemical oxidation of the bitumen resulting in increased viscosity and a reduced diffusivity, which in return, affect the oxygen distribution and the oxidative ageing. Thus, the oxidative ageing of asphalt pavements should be modelled in a multiphysics perspective. However, applying the multiphysics models in ageing predictions is limited due to restrictions in the FE modelling, such as non-user-friendly interfaces which limit modelling abilities and cause restrictions on constitutive models. The weak-form PDE-based FE modelling is a tool that can achieve the objective. The heat transfer, oxygen diffusion, and oxidation kinetics are physical and chemical physics that occur simultaneously in the pavement. The constitutive equations of these physics can be represented as ordinary differential equations (ODE) or partial differential equations (PDE). Previous studies developed comprehensive models to characterise the non-uniform oxidative ageing process of bitumen in the pavement equations, which are listed in Table 1 [5–7]. To address the aforementioned research needs, this study develops a framework for modelling oxidative ageing of asphalt pavements in a multiphysics perspective using the physics listed in Table 1. In the following section, Comsol solver is employed because it provides an efficient computational platform to solve the weak-form PDEs and address the coupling effects of different physics [8, 9]. Subsequently, the integrated FE model will be used to compute the change in each physics during the oxidative ageing.
2 Development of Weak-Form Equation-Based Computational Models The proposed oxidative ageing model consisted of several components, including (1) model geometry and materials; (2) model physics; and (3) model parameters, variables, and time-dependent interpolation functions. A two-dimensional geometry was built to represent the pavement structure (Fig. 1a). The dimensions of the geometry are customised according to the pavement structure and mixture’s volumetric properties. For example, the thickness of mastic coating film (width of domain 2 in
Multiphysics Simulation and Validation of Field Ageing …
1393
Table 1 Constitutive equations of oxidative ageing Pavement Heat Transfer Model (Han et al. [7]) Fourier’s one-dimensional heat transfer equation:
∂T ∂t
=α
∂2T ∂x2
, heat diffusivity: α =
k ρ·c p
Boundary conditions of the top pavement surface: ρC p x 2
∂ Ts ∂t
= Q s − α · Q s + Q a − Qr − Q c − Q f
where T is the ageing absolute temperature; t is the ageing time; x is the pavement depth; k is the thermal conductivity of asphalt; ρ is the density of asphalt; C p is the heat capacity of asphalt; Ts
is pavement surface temperature; Q s is the heat flux due to solar radiation; α is the albedo of the pavement surface (the fraction of reflected solar radiation); Q a is the down-welling long-wave radiation heat flux from the atmosphere; Q r is the outgoing long-wave radiation heat flux from the pavement surface; Q c is the convective heat flux; Q f is the conduction from the surface into the pavement. Oxygen Transport Model (Glover et al. [5]) cT R ∂C A(x,t) ∂ ∂P Diffusion equation: ∂ P(x,t) = f c f · Do ∂t ∂x ∂x − h . ∂t Coefficient of oxygen diffusion equation:
Do T
= 5.21 × 10−12 L SV −0.55 , L SV = e(m+H S·C A)
where P is the oxygen pressure; x is the pavement depth; Do is oxygen diffusivity in pure asphalt; c is a factor that converts reaction rate of carbonyl area (CA) to rate of oxygen consumption; f c f is the field calibration factor adjusting Do ; h is the solubility constant of oxygen in asphalt; L SV is the low shear rate-limiting viscosity; H S is asphalt hardening susceptibility; m is an experimental parameter; R is the universal gas constant. Oxidative Kinetics Model (Jin et al. [6]) ∂C A ∂t
= M RT F O k f e−k f t + kc
k f = A f P ∝ e−Ea f /RT kc = Ac P ∝ e−Eac /RT where M RT F O is the limiting amount of carbonyl formation due to the first-order reaction after hot mix production; k f and kc are fast-rate and constant-rate reaction constants; A f and Ac are fast-rate and constant-rate pre-exponential factors, respectively; ∝ is the reaction order for oxygen pressure; E a f and E ac are fast-rate and constant-rate activation energies, respectively .
Fig. 1 Geometry and material properties of the integrated ageing model. a Proposed 2D geometry divided into five domains (not plotted to a scale); b Illustration of the concept of diffusion depth, c The pavement structure of road section US277 showing the optimised thermal conductivity, heat capacity, and density for each layer
1394
E. Omairey and Y. Zhang
Fig. 1a) is defined as the diffusion depth (d D ) which is the mastic film thickness that surrounds the interconnected air channels (shown in Fig. 1b) [5]. The heat transfer in the solids depends on the thermal properties of the materials. Therefore, the thermal characteristics of the pavement layers are defined. The inputs for each pavement layer (for each domain in Fig. 1a) include the thermal conductivity, density and heat capacity (the optimised thermal inputs are shown in Fig. 1c). Field data for road section US277 which is located in Lamar, South Texas, USA were obtained from FHWA reports [5, 10] and employed to test the overall ageing model. The pavement structure for road section US277 is shown in Fig. 1c. A time-dependent study was conducted by using five circularly-dependent physics that work simultaneously at certain domains of the geometry, as shown in Fig. 2. The model interfaces were built by converting the equations listed in Table 1 into heat transport physics and weak-form PDE-based physics. The heat transfer interface employed the Long-Term Pavement Performance (LTPP) Climate Tool to obtain the site-specific hourly values of air temperature, wind speed, solar radiation, albedo and emissivity. Other parameters, such as the bitumen’s kinetics and coefficient of oxygen diffusion in the air, were introduced into the ageing model in the parameters and variables components. These inputs were obtained from the literature [5] by conducting Fourier transform infrared spectroscopy (FTIR), laboratory ageing tests, and rheological tests on the virgin bitumen; whereas, mixture specific information can be obtained from X-ray CT images for field cores extracted from the road sections.
3 Results Data for road section US277 were implemented into the model and run for 3.5 years of field ageing simulations, and the results are shown in Fig. 3. The temperature variation with ageing time calculated at different pavement depths (pavement surface, 0.03 m and 0.06 m which are at the middle and bottom of the asphalt layer) are shown in Fig. 3a. It can be seen from Fig. 3b, c that the oxygen partial pressure in the air channels (P) and in the asphalt mastic (P1), respectively, are decreasing during the summer and increasing in the winter. This observation indicates that during summer season, the bitumen is oxidising, therefore consuming oxygen from the air channels and causing the oxygen concentration levels and subsequently the oxygen partial pressure to drop. This also leads to the quick increase of the ageing product carbonyl (CA1) during summer and relatively much slower or none increase of the CA1 during winter, as shown in Fig. 3d. Figures 3a–d show that temperature varies much higher on the surface than that in the middle of pavement, with least variation occuring at the bottom. The oxygen pressure in the middle layer is much lower than that on the surface and bottom of the asphalt layer, because the oxygen can diffuse into the asphalt not only from the top asphalt surface but also from the bottom via the unbound aggregate base layer. Carbonyl grows much quicker on asphalt layer surface due to higher temperature and
Multiphysics Simulation and Validation of Field Ageing …
1395
Fig. 2 Illustration of the circular-dependent and time-dependent physics of the ageing model. The highlighted domains in red colour are the active domains for each physics. The hourly outputs for these physics are: pavement temperature (T); oxygen partial pressure in the air voids (P); oxygen partial pressure in the asphalt mastic film (P1), carbonyl area at the air-mastic interface (CA) as a boundary condition (B.C.); and carbonyl area inside the asphalt mastic (CA1)
the full availability of oxygen. The carbonyl content decreases and tends to become constant with increasing the layer depth due to the limited availability of oxygen and lower temperature values within the asphalt layer. The model has been validated using field measurements of carbonyl area obtained from recovered bitumen at different depths of the core samples and recorded in FHWA reports. The results are illustrated in Fig. 3e, it can be seen that the field measurements suffer from unsystematic changes in CA content along the depth, this is somehow expected from the field data; nonetheless, the predicted CA values are in a good agreement with CA measurements from the field in terms of the growth pattern.
4 Conclusions A weak-form PDE-based multiphysics model is proposed to predict the oxidative ageing of the asphalt pavements. The model requires location-specific climate inputs, oxidation kinetics inputs obtained from ageing, rheological and chemical laboratory tests, and mixture volumetric properties obtained from the pavement mix design
1396
E. Omairey and Y. Zhang
Fig. 3 Results of the oxidative ageing model for road section US277 plotted against the ageing time in the field, including a Temperature profile of the asphalt pavement; b Oxygen pressure in the air voids (P); c Oxygen pressure inside the asphalt mastic film (P1); d Carbonyl are inside the asphalt mastic (CA1); e Predicted and measured carbonyl areas plotted against the pavement depth at different field life spans; whereas (i) Predicted and measured carbonyl areas plotted against the pavement at different depths for road section US82. The plot in 3(a–d) highlighted in light blue indicates the winter period, which recognise the different ageing patterns between summer and winter seasons
specifications. The model has been validated using the field measurements of oxidation products for road sections US277 and US82 which are located in Laredo, south of Texas, and Lubbock, north of Texas, USA, respectively. The conclusions are as follows: 1.
2.
The model can effectively model and solve the circular dependency between the ageing-related multi-physics (i.e., heat transfer, oxygen diffusions & oxidative kinetics). Weak-form PDE-based FEM can effectively solve the circular dependence and can reliably predict the annual hourly profiles of temperatures, oxygen pressures and oxidation products along with pavement depth at different climate zones.
References 1. Prapaitrakul, N., Han, R., Jin, X., Glover, C.J.: A transport model of asphalt binder oxidation in pavements. Road Mater. Pavement Des. 10(sup1), 95–113 (2009)
Multiphysics Simulation and Validation of Field Ageing …
1397
2. Omairey, E.L., Zhang, Y., Al-Malaika, S., Sheena, H., Gu, F.: Impact of anti-ageing compounds on oxidation ageing kinetics of bitumen by infrared spectroscopy analysis. Constr. Build. Mater. 223, 755–764 (2019) 3. Luo, X., Gu, F., Lytton, R.L.: Kinetics-based aging prediction of asphalt mixtures using field deflection data. Int. J. Pavement Eng. 20(3), 287–297 (2019) 4. Al-Rub, R.K.A., Darabi, M.K., Kim, S.M., Little, D.N., Glover, C.J.: Mechanistic-based constitutive modeling of oxidative aging in aging-susceptible materials and its effect on the damage potential of asphalt concrete. Constr. Build. Mater. 41, 439–454 (2013) 5. Glover, C.J., Han, R., Jin, X., Prapaitrakul, N., Cui, Y., Rose, A., Lawrence, J.J., Padigala, M., Arambula, E., Park, E.S., Martin, A.E.: Evaluation of binder aging and its influence in aging of hot mix asphalt concrete: technical report (No. FHWA/TX-14/0-6009-2). Texas. Dept. of Transportation. Research and Technology Implementation Office (2014) 6. Jin, X., Cui, Y., Glover, C.J.: Modeling asphalt oxidation in pavement with field validation. Pet. Sci. Technol. 31(13), 1398–1405 (2013) 7. Han, R., Jin, X., Glover, C.J.: Modeling pavement temperature for use in binder oxidation models and pavement performance prediction. J. Mater. Civ. Eng. 23(4), 351–359 (2011) 8. Zhang, Y., Birgisson, B., Lytton, R.L.: Weak form equation–based finite-element modeling of viscoelastic asphalt mixtures. J. Mater. Civ. Eng. 28(2), 04015115 (2015) 9. Manual, C.M.R.: COMSOL, p. 1084. France, Grenoble (2013) 10. Martin, A.E., Arambula, E., Kutay, M.E., Lawrence, J., Luo, X., Lytton, R.: Comparison of fatigue analysis approaches for hot-mix asphalt to ensure a state of good repair (No. SWUTC/13/600451-00012-1). Southwest Region University Transportation Center (US) (2013)
Multiscale Analysis of Pavement Texture—Effect on Skid Resistance Wiyao Edjeou, Veronique Cerezo, Hassan Zahouani, and Ferdinando Salvatore
Abstract The surface properties (skid resistance, rolling noise and rolling resistance) of the road are closely related to the surface texture. Several scales of texture intervene in the interaction between a tire and the road pavement. However, the influence of each scale on these properties is little or not known. During lifetime, the texture of road evolves under the wear induced by road traffic (polishing) and climate changes. This texture change affects pavement surface performances. This paper presents a laboratory study, which aims at developing a multiscale approach to identify the relevant texture scales, which can explain skid resistance values. For multiscale decomposition, a wavelet-based multiscale analysis method is selected owing to its flexible time-frequency resolution. After comparing several functions, the Morlet wavelet is used. Then, laboratory tests are performed on various bituminous mixtures. Polishing tests are done with Wehner & Schulze machine to assess skid resistance evolution under traffic and 3D-cartographies of the surface are realized at different stages of polishing to assess surface texture evolution. Keywords Surface texture · Multiscale analysis · Wavelet · Skid resistance
1 Introduction Pavement properties such as rolling noise, skid resistance and rolling resistance depend closely on the texture of the surface. Skid resistance on the road tends to decrease under the influence of traffic and climate. These phenomena are results of surface texture evolution. For a bituminous concrete surface, there are two phases of
W. Edjeou (B) · V. Cerezo French Institute of Science and Technology in Transportation, Planning and Networks, Nantes, France e-mail: [email protected] H. Zahouani · F. Salvatore Laboratory of Tribology and Dynamics of Systems, Écully, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_178
1399
1400
W. Edjeou et al.
evolution. First, removal of the bitumen layer resulting in an increase in skid resistance. Then the polishing of the aggregates causes a decrease of the skid resistance [1]. For road engineers, texture is divided into microtexture (asperities whose height is inferior to 0.5 mm in width and 0.2 mm in height) and macrotexture [2]. This separation is not sufficient to explain the phenomena involved in the generation of skid resistance (adhesion and hysteresis) and its evolution. Indeed, the texture is composed of several scales whose contribution to friction forces generation is not clear. Moreover, the standard texture parameters do not explain clearly the evolution of the skid resistance [3]. Usually, the texture signal is separated into components of roughness, undulation and shape using a filter. After filtering, the texture parameters are averaged over the entire profile. Thus, it is impossible to access the information contained in the different scales of the profile. Additional research is needed to solve this problem. Knowing the evolution of the texture at each scale would indeed help maintaining skid resistance properties. Thus, we must find a method of decomposition, which will allow us to have access to different scales of texture of road pavement. This method should have a variable resolution in frequency and in the space domain. The wavelet method has these characteristics [4]. This article proposed a method based on multiscale analysis of pavement texture to assess skid resistance during the lifetime of road pavement. The objective is to follow the evolution of each scale of texture during the polishing to be able to relate it to the evolution of the skid resistance in a second time. The paper is divided into two parts. The first part presents the experiments carried out during the study. It details the techniques of polishing, skid resistance estimation and texture acquisition used in the study. The second part details the numerical study focusing on multiscale decomposition and the relationship with friction.
2 Experimental Study The experimental study aims at reproducing the effect of polishing on various types of pavement samples in the laboratory and estimating texture and skid resistance during the process. Samples are circular discs of 225 mm in diameter. This disc is a core sample of asphalt mix. Three asphalt mixes are considered in this work [5, 6]. In the rest of this paper, the three formulas are named F1, F2 and F3. Table 1 shows their composition.
2.1 Polishing and Skid Resistance Measurements Polishing is performed with the Wehner & Schulze machine. This machine reproduces polishing due to road traffic [1] and allows measuring skid resistance at various
Multiscale Analysis of Pavement …
1401
Table 1 Bituminous mixtures compositions Compositions Aggregates
Binder
Mix composition (%) F1
F2
F3
4/6
72.8
73.2
74.7
1/2
9
16
0
Sand 0/2
9
0
17.5
Filler limestone
3.9
5.5
2.4
Polymer modified bitumen
5.3
5.3
5.4
stages of polishing, according to standard NF EN 12697-49 [7]. Polishing is done by rolling with a slight slip of three rubber cones on the sample. In the measuring unit of skid resistance, the measuring head is lowered on the test surface, with an addition of water. The measuring head is equipped with three rubberized pads and can be triggered electronically. The braking torque generated by the contact between the pads and the surface is measured continuously and recorded until the immobilization of the measuring head. The skid resistance in the laboratory is calculated from the friction torque measured at 60 km/h. The slip rate is of the order of 0.5–1%. The procedure defined by [1] was used in this study. Figure 1 shows the results of the skid resistance coefficient obtained as a function of polishing. For the three formulas, we can distinguish two phases: an increase of the coefficient of friction up to the maximum and a phase of decrease of the coefficient. In his study [1] attributes the increase of the skid resistance to the removal of the binder and the reduction of the friction to the polishing. Since the same amount of binder was
Fig. 1 Skid resistance evolution during polishing
1402
W. Edjeou et al.
used for F1 and F2, we observe almost the same changes in friction. The difference between these formulas is observed beyond 50,000 rotations of Wehner & Schulze. The decrease of the friction is stronger for F2. This can be explained by the fact that the microtexture is stronger for F1 by considering its formulation. This also confirms the observation of [1]. The values of the coefficient of friction obtained for F3 are lower than those of the first two formulas. We can attribute the fact that the maximum was reached at 6,000 rotations at the amount of binder used for this formula. Thus, it takes longer to remove the binder. The low value of the friction may be due to the composition of the formula. Indeed, this formula has more large granulate than the others. The increase of the coefficient of friction between the state of polishing of the maximum friction and 10,000 rotations is due to the washing of the sample by the machine. Indeed, after the washing per-formed by Wehner and Schulze, there is still silica used for polishing. This leads to an increase in small aggregates resulting in a slight increase in friction. One solution found to solve this problem is the complete washing of the sample outside the machine. The analysis of the texture of these three formulas will give us further information about their polishing behavior.
2.2 Texture Measurement The measurement of the texture is made at different polishing stages. It was done on each sample over four areas to be representative of the whole surface. The size of the areas is 3 mm × 3 mm. These areas will be name Z1, Z2, Z3 and Z4. Figure 2 shows the sample and the different texture measurement zones. The acquisition of the maps is done with the Alicona Infinite Focus sensor. A compromise had to be found between the time of measurement and precision. Then, the measurements are made at magnification ×10. The vertical and lateral resolutions Fig. 2 Sample with different zones of texture acquisition
Multiscale Analysis of Pavement …
1403
are respectively 753 nm and 3910 nm. To better understand the influence of pavement texture, mappings were made at the initial stage, at the maximum of friction curve (5–6000 passes), at 90,000 passes of polishing and at the end of the test (i.e. 180,000 passes). The maximum skid resistance lies between 5,000 and 10,000 passes depending on the asphalt mix. The choice of mapping stage in maximum skid resistance range has less influence on pavement texture which will be obtained. Then, the data are exported in the software MountainsMap (Version 7) for outliers’ removal, slope suppression and the removal of the form. Our study being in two dimensions, we extract profiles on each surface cartography. Three profiles are extracted at 25%, 50% and 75% of width of each cartography. These profiles are named P1, P2 and P3.
3 Multiscale Analysis In this part, the wavelet-based multiscale decomposition method is implemented under Matlab software based on the algorithm defined in [8]. In his study [9] recommends using either the Morlet wavelet or the eighth derivative of Gauss for decomposition. After testing these wavelets, we found a mean reconstruction error on the average height Ra using the Morlet wavelet of 3% with a standard deviation of Ra of 1.5 µm. The average reconstruction error for the Gauss wavelet is 1.5% with a standard deviation of 0.05 µm. We notice that reconstruction errors of both wavelets are small. Also, the calculation time using the Gauss wavelet is larger due to the multiple numerical derivatives. Thus, the Morlet wavelet was preferred. To be able to quantify the amplitude of each scale, the amplitude spectrum defined by [4] was used. This is the average height of the profile at each scale. We have drawn the spectrum of amplitude for the three formulas according to the polishing stages. Figure 3 shows the results obtained.
Fig. 3 Amplitudes spectrum, a F1, b F2, c F3
1404
W. Edjeou et al.
First, the three formulas exhibit similarities (Fig. 3). When the number of passes increases, we notice a slight reduction of small scales ( 90) for the three binders and their corresponding mixtures as observed in Fig. 2. It is to be noticed that both graphs approach the line of equality between experimental and predicted values through the model. Fig. 3 shows the application of the SHStS transformation to obtain the blend properties from the binder properties in the Cole-Cole representation, where the sequence of the four transformations used in the methodology is shown.
Fig. 1 Linear viscoelastic characterization of asphalt binders and corresponding mixtures at the reference temperature of 21.1 °C
Relationship Between Linear Viscoelastic Properties …
1547
Table 2 Binders and mixtures 2S2P1D and WLF parameters Binder
WLF
B2
B3
2.7
2.7
3.0 6.93E−05
τ0 (s)
2.5E−05
4.0E−5
k
0.25
0.26
0.28
h
0.62
0.62
0.61 3.0E + 08
G∞ (Pa)
3.7E + 08
3.0E + 08
G0 (Pa)
0
0
0
β
55
55
66
C1
10.35
12.85
14.71
C2
90.65
113.00
122.00
M1
M2
M3
Mixture τ0 (s)
0.05
0.043
0.05
E∞ (Pa)
3.80E + 10
3.97E + 10
3.70E + 10
E0 (Pa)
1.30E + 08
1.40E + 08
9.5E + 07
Fig. 2 Relationship between experimental and predicted values of absolute value of complex modulus for asphalt binders and mixtures
|G*| experimental (MPa) 0.0E+00 2.0E+01 4.0E+01 6.0E+01 8.0E+01 8.0E+01 4.0E+04
|E*| model (MPa)
Parameter
6.0E+01
3.0E+04 M1 M2 M3 B1 B2 B3 Equality
2.0E+04 1.0E+04 0.0E+00 0.E+00
1.E+04
2.E+04
3.E+04
4.0E+01 2.0E+01
|G*| model (MPa)
Parameter
B1 δ
0.0E+00 4.E+04
|E*| experimental (MPa)
4 Conclusions This paper investigated the relationship between linear viscoelastic properties of asphalt binders and their corresponding asphalt mixtures (with fixed aggregate skeleton), by means of an experimental campaign and modeling techniques. The model used for the representation of linear viscoelastic behavior (small amplitudes and few load repetitions), known in the literature as 2S2P1D, was satisfactory for modeling the behavior of mixtures and binders. Out of the 7 constants of the 2S2P1D model, associated with the 2 independent constants of the WLF equation representing the TTSP, adjusted for complex modulus of binders, 6 could be kept for their respective mixtures. All these 6 material constants are linked to time dependent properties, confirming something already reported in literature, which is that the viscoelastic
1548
L. S. de Oliveira et al.
Fig. 3 SHStS transformation to obtain asphalt mixtures’ LVE results from asphalt binders’ LVE results
behavior of the mixture is inherited from the asphalt binder. The variation of the other constants was described by the transformation known in the literature as SHStS, which presented satisfactory adjustments. Further work is undergoing considering more materials, including other aggregate skeletons and modified binders. Acknowledgements The authors thank Fundação Cearense de Apoio ao Desenvolvimento Científico e Tecnológico – FUNCAP and Conselho Nacional de Desenvolvimento Científico e Tecnológico – CNPq for the authors’ grants.
References 1. Roberts, F.L., Khandal, P.S., Brown, E.R., Lee, D.Y., Kennedy, T.W.: Hot mix asphalt materials, mixture design and construction. NAPA Education Foundation, Lanham, MD (1996) 2. Kim, Y.R.: Modeling of Asphalt Concrete. ASCE Press, McGraw-Hill Construction, Reston, VA, New York (2009) 3. Babadopulos, L.F.A.L.: Phenomena occurring during cyclic loading and fatigue tests on bituminous materials: identification and quantification. PhD Thesis, University of Lyon (ENTPE), pp. 1–303, Lyon (2017) 4. Christensen, R.M.: Theory of Viscoelasticity, 2nd edn. Academic, New York (1982) 5. Silva, H.N.: Caracterização viscoelástica linear de misturas asfálticas: Operacionalização Computacional e Análise pelo Método dos Elementos Finitos. Dissertação de Mestrado, Programa de Pós-Graduação em Engenharia de Transportes, Universidade Federal do Ceará, Fortaleza (2009) (in Portuguese). 6. Di Benedetto, H., De La Roche, C., Baaj, H., Pronk, A., Lundström, R.: Fatigue of bituminous mixtures. Mater. Struct. 37, 202–216 (2004) 7. Delaporte, B., Di Benedetto, H., Chaverot, P., Gauthier, G.: Linear viscoelastic properties of bituminous materials including new products made with ultrafine particles. Road Mater. Pavement Des. 10, 7–38 (2009)
Relationship Between Linear Viscoelastic Properties …
1549
8. Pouget, S., Sauzéat, C., Di Benedetto, H., Olard, F.: Prediction of isotropic linear viscoelastic behavior for bituminous materials—forward and inverse problems. In: 5th Eurasphalt and Eurobitume Congress, Istambul (2012) 9. ASTM D 7175.: Standard test method for determining the rheological properties of asphalt binder using a dynamic shear rheometer. West Conshohocken (2015) 10. AASHTO T 342.: Determining dynamic modulus of HotMix asphalt (HMA). Washington (2011)
Relaxation Spectrum: Why It Matters and How to Correctly Develop One? Kin-Ming Chan and Yuhong Wang
Abstract One of popular models to describe mechanical behavior of asphalt binder is the Generalized Maxwell Model (GMM), which is analogized as a large number of springs and dashpots connected in a parallel fashion. Theoretically, the continuous relaxation spectrum of GMM of a particular asphalt binder is unique and reflects its fundamental physicochemical characteristics. The commonly used approach is to search the relaxation spectrum from dynamic modulus test data. However, the calculation of the relaxation spectrum from dynamic modulus test data is an ill-posed problem. Although different approaches have been developed for the conversion of the dynamic modulus test data to relaxation spectrum, their validity remains unknown. In this research, a new determination of relaxation spectrum, which obtained by considerations of both static modulus test, such as relaxation modulus test, and dynamic modulus test is proposed and it is found the obtained spectrum is more realistic when one compare both data of relaxation modulus test and dynamic modulus test. Keywords Relaxation spectrum · Generalized Maxwell model · Aging · Microstructures
1 Introduction In modeling asphalt binders within a small range of response, linear viscoelastic theory and the Generalized Maxwell Model (GMM) are often applied. The relaxation spectrum of GMM, a positive real-value function of relaxation time τ , carries the fundamental mechanical information of a viscoelastic material. Although several methods have been developed to search the spectrum, unfortunately, not all of them K.-M. Chan (B) · Y. Wang Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hong Kong, China e-mail: [email protected] Y. Wang e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_197
1551
1552
K.-M. Chan and Y. Wang
are handy and accurate. The methods may be classified into two groups: theoretical approaches and phenomenological approaches. In theoretical approaches, e.g., the Rouse’s model, relationships between microscale parameters and the spectrum are investigated. Although this treatment leads to meaningful insights, the microscale parameters, such as molecule mobility, are often hard to be obtained. Conversely, in phenomenological approaches, the spectrum is determined from mechanical tests, which are easier as compared with the theoretical one. This study is focused on the phenomenological approach of determining relaxation spectrum for asphalt binder. In general, phenomenological approaches are a curve-fitting process to minimize errors between predicted values by the model and experimental data. Based on test methods, those approaches may be further divided into dynamic modulus test method and static modulus test method. The former refers to spectrum determination using data from dynamic modulus tests (e.g., dynamic shear modulus test, dynamic triaxial test). The latter refers to spectrum determination using data from static modulus tests (e.g., relaxation modulus test, creep test). Both approaches are often used in characterizing asphalt binders. However, previous studies are mainly focused on the development of parametric models with data obtained from either dynamic or static modulus test. The two tests are not combined to characterize asphalt materials, nor is a cross check made between the tests. In this study, a new method is developed to determine the relaxation spectrum of asphalt binder by considering both the static and dynamic modulus tests.
2 Theoretical Background Assume that a material is isotropic and under shear loading in a linear viscoelastic range. According to the Boltzmann superposition principle, the stress tensor σ (t) can be expressed by the convolution of the shear modulus function G(t) and time , in which ∗ stands for the convolution derivative of strain tensor ε(t), i.e., σ = G ∗ dε dt of functions. By the configuration of GMM and assuming numerous asphaltenes of different sizes and their interconnections, the relaxation modulus G(t) can be calculated by: τ =∞
G(t) = ∫ H (τ )e− τ d ln τ τ =0
t
(1)
where H (τ ) is the continuous relaxation spectrum, which is a function of a continuous variable τ , called relaxation time. Furthermore, if one applies a unit dynamic strain ε = eiωt (ω is angular frequency), the spectrum can be obtained from storage modulus G (ω) or loss modulus G (ω) by:
Relaxation Spectrum: Why It Matters …
(ωτ )2 d ln τ 1 + (ωτ )2 τ =0 τ =∞ ωτ d ln τ G (ω) = ∫ H (τ ) 1 + (ωτ )2 τ =0
1553
τ =∞
G (ω) = ∫ H (τ )
(2)
In the phenomenological approaches, although some studies provide inversion of the relation, i.e., to express the spectrum directly by the modulus, however, the one-toone correspondence does not generate the entire spectrum due to the limited range of test data. Therefore, modulus recovery, a process to predict and extend the modulus data with data fitting technique, is often required. As mentioned before, existing phenomenological approaches may be classified into two groups: static modulus test method and dynamic modulus test method. However, both methods have their shortages. For the static modulus test method, as compared with Lorentzian functions in Eq. (2) as a kernel, the exponential kernel in Eq. (1) decays much faster in the curve fitting. Hence, influences of the first a few data points from the relaxation modulus test on the relaxation spectrum are extremely high. However, relaxation modulus data at very beginning is not very reliable and smooth since the applied strain is yet stable. For the dynamic modulus test method, although data obtained in the dynamic modulus test is smoother, it also has some drawbacks. Firstly, for material with high fluidity, the steady state is hard to be reached and this leads to a systematic error. Consequently, some data points at high temperature become unrealistic. Secondly, spectra derived by storage modulus and loss modulus may not be the same and the mismatch has been widely reported in literature [e.g., 1]. In addition, both the static and dynamic modulus test methods have some common shortages. Firstly, Eqs. (1) and (2) are well-known as Fredholm integral equations, which are shown to be ill-posed and the solution H (τ ) is not unique and stable. As a result, the obtained spectrum may vary a lot with slight change in modulus data [2]. Secondly, test data is always noisy and thus the solution is seriously affected by the ill-posedness. Thirdly, only limit range of data from the relaxation or dynamic modulus tests can be obtained, although this can be solved by time-temperature superposition, the process adds variation to the experimental data.
3 A New Determination of Relaxation Spectrum To calculate the spectrum in the integral Eqs. (1) and (2) numerically, one can transfer the problem into a system of linear equations. In this paper, the trapezoidal rule is adopted to discretize the integral equations into linear equation forms. Equation (3) is used as an example for illustration. Assume that a set data G(t) is obtained, namely: G(t1 ), G(t2 ), …, G(tn ). The algorithm first assumes H (τ ) = 0 if τ < τ1 or τ > τn , for some very small quantity of τ1 and large quantity of τn . Then it evaluates H (τ1 ),
1554
K.-M. Chan and Y. Wang
)−ln(τ1 ) H (τ2 ),H (τ3 ),…, H (τn ), where ln(τk ) = ln(τ1 ) + (k − 1) ln(τnn−1 for 1 ≤ k ≤ n. In other words, those τi ’s is linearly distributed in the log scale. The algorithm that approximates G(ti ) is as follows:
− ti ln τn − ln τ1 = Ki j H τ j G(ti ) ≈ H τ j e τ j n−1 −
(3)
ti
n −ln τ1 where the index j = 1, . . . , n and K i j = e τ j ln τn−1 and the summation convention is used. Similar procedure can be applied to calculate H τ j ’s from G (ωi ) and G (ωi ) after the complex modulus |G ∗ (ω1 )|, |G ∗ (ω2 )|, …, |G ∗ (ωn )| and phase angle δ(ω1 ), δ(ω2 ), …, δ(ωn ) are obtained. Here this is assumed that the number of data obtained from the relaxation modulus test and dynamic modulus test are the same. To combine both data sets from the static and dynamic modulus tests, one may 2 ωτ ωi τ j ln τn −ln τ1 n −ln τ1 , Mi j = , i.e., the make use of the matrices L i j = ( i j ) 2 ln τn−1 2 n−1 1+(ωi τ j ) 1+(ωi τ j ) matrices of storage and loss moduli. Concatenating matrices K i j , L i j , Mi j and T T T T arrays G ti1 , G ωi2 , G ωi3 , i.e., let Ni j = Li j Mi j Ki j T and G i = [G(ti )]T G (ωi ) T G (ωi ) T , the system of linear equations then becomes:
G i ≈ Ni j H τ j
(4)
The new method attempts to search a spectrum satisfying data of relaxation of storage modulus (G (ωi ) ≈ L i j H τ j ) modulus (G(ti ) ≈ K i j H τ j ), and data and loss modulus (G (ωi ) ≈ Mi j H τ j ) simultaneously. The searching of H (τ1 ), H (τ2 ), …, H (τn ) is done by minimizing the squared errors of the system of liner equations. It is noted that the ill-posedness mentioned before may make the squared T Ni j nearly singular. Moreover, when performing curve fitting, issues matric Ni j such as over-fitting, smoothness of calibrated results and influence of low-pass filters have to be considered. This can be solved by using the regularization method that is widely adopted in inverse problems, especially for ill-posed ones. Honerkamp & Weese suggest one should make use of L 2 -regularization method to search the spectrum [3, 4]: 2
Ni j H τ j − G¯ i + λ H τ j 2 min [ H (τ j )]∈Rn
(5)
where ∗ stands for L 2 norm and hence the loss function between essential squared 2 error. The real-valued constant λ in regularization term λ H τ j is called regu larization parameter and this gives weight of penalty to magnitude of each H τ j ’s. The parameter can be set by several algorithms such as Restricted Maximum Likelihood Method [3], L-curve Method [4], Cross-validation Method [5]. Hence a smoother solution can be obtained as (I represents identity matrix of size n):
Relaxation Spectrum: Why It Matters …
−1 T T H τ j ≈ Ni j Ni j − 2λI Ni j Gi
1555
(6)
4 Experimental Study This section discusses the validation of the new method by experimental data. Two asphalt binders were used: A virgin asphalt binder of penetration grade 60/70 (designated as Virgin) and the same binder treated by Rolling Thin-Film Oven (RTFO) and the Pressure Aging Vessel (PAV) (designated as Aged). The purpose of the experiment is just to assess the new method; therefore, the scope of the experiment is limited. The RTFO and PAV ageing follows the standard AASHTO T240 and AASHTO R28, respectively. All the specimens are 8 mm in diameter and 2 mm in height. Two sets of binder subject to DSR’s relaxation modulus test and dynamic modulus test at 1% shear strain. The reference temperature is set to be 25 °C. In the dynamic modulus test, TTS is adopted with test temperature at 13, 19, 25, 31 and −C T −T 37 °C. The Williams–Landel–Ferry Model log aT = C +1 (T −Tr e f ) and generalized 2 ( ref ) 1 ∗ β1 +γ1 (log ω) − λ1 are used logistic sigmoidal function log|G (ω)| = υ1 + α1 1 + λ1 e in TTS to predict the shifting factor aT and |G ∗ (ω)|. Then the storage moduli and loss moduli are calculated by G (ω) = |G ∗ (ω)| cos δ(ω), G (ω) = |G ∗ (ω)| sin δ(ω) ∗ (ω)| and the Kramers–Kronig relation’s approximation δ(ω) ≈ π2 d ln|G . On the other d ln ω hand, relaxation moduli are described by the generalized logistic sigmoidal function − 1 as G(t) = υ2 + α2 1 + λ2 eβ2 +γ2 (log ω) λ2 . The raw data and fitted curves for both the virgin and aged binders are shown in Figs. 1 and 2. The comparison mainly considers the new method, the method using dynamic modulus test data, and the method using relaxation modulus test data. Data range for the relaxation moduli and dynamic moduli is −16 ≤ log t ≤ 1and − 4 ≤ log ω ≤ 20, respectively. The obtained relaxation spectra for virgin and aged binders are showed in Figs. 3 and 4. The spectra are then used to back calculate relaxation moduli, storage moduli and loss moduli, which are further used to compare with the original raw data. The comparison results are shown in Figs. 5, 6, 7, 8, 9 and 10. Figure 3 shows that spectra obtained from the relaxation modulus test and dynamic modulus test are quite different. This reveals the shortages mentioned above if one only considers one of the tests. Conversely, the spectra of the new method keep the characteristics of both spectra obtained by the two tests. Moreover, in Figs. 3, 4, 5, 6, 7, 8, 9 and 10, although the back-calculated dynamic moduli from the dynamictest-derived spectra fit the test data very well, there are large differences between the back-calculated relaxation moduli from the static-modulus-test-derived spectra and the relaxation modulus test data, and vice versa in the situation of back-calculated relaxation moduli. The new method strikes a balance between two methods.
1556
K.-M. Chan and Y. Wang
Fig. 1 Relaxation moduli
Fig. 2 Complex moduli and phase angles
5 Conclusion A new method is developed to derive relaxation spectrum by combining data from both relaxation modulus test and dynamic modulus test. Theoretically, the relaxation modulus, dynamic modulus and relaxation spectrum should uniquely characterize the mechanical behaviors of asphalt through one-to-one correspondence. However, due to the limits in test methods, relaxation spectrum determined from different set
Relaxation Spectrum: Why It Matters …
1557
Fig. 3 Spectra for virgin binder
Fig. 4 Spectra for aged binder
of data is often different. The new method incorporates information from both tests and provides a balanced approach to determining relaxation spectrum.
1558
K.-M. Chan and Y. Wang
Fig. 5 Comparison of back-calculated relaxation moduli of virgin binder
Fig. 6 Comparison of back-calculated relaxation moduli of aged binder
Relaxation Spectrum: Why It Matters …
Fig. 7 Comparison of back-calculated storage moduli of virgin binder
Fig. 8 Comparison of back-calculated storage moduli of aged binder
1559
1560
K.-M. Chan and Y. Wang
Fig. 9 Comparison of back-calculated loss moduli of virgin binder
Fig. 10 Comparison of back-calculated loss moduli of aged binder
References 1. Groetsch, C.W.: The theory of Tikhonov Regularization for Fredholm Equations, 104p. .Boston Pitman Publication (1984) 2. Honerkamp, J., Weese, J.: Determination of the relaxation spectrum by a regularization method. Macromolecules 22(11), 4372–4377 (1989) 3. Harville, D.A.: Maximum likelihood approaches to variance component estimation and to related problems. J. Am. Stat. Assoc. 72(358), 320–338 (1977)
Relaxation Spectrum: Why It Matters …
1561
4. Hansen, P.C.: Analysis of discrete ill-posed problems by means of the L-curve. SIAM Rev. 34(4), 561–580 (1992) 5. Golub, G.H., Heath, M., Wahba, G.: Generalized cross-validation as a method for choosing a good ridge parameter. Technometrics 21(2), 215–223 (1979)
Rheological Characterisation of Modified Bitumens with Biodiesel-Derived Biobinders Ana I. Weir, Gordon Airey, Colin Snape, and Ana Jiménez del Barco Carrión
Abstract Growing environmental concerns and scarcity issues of petroleum-derived bitumen have motivated the search for alternative and renewable sources to be used in asphalt mixtures. The concept of biobinders (binders manufactured from biomass) are becoming more popular in the field as a potential solution. In particular, waste biomass products are of interest due to their availability and impact on sustainability. Biobinders have shown great potential to reduce bitumen demand and have exhibited good performance in terms of resisting common distresses affecting roads. However, they are still relatively unknown and detailed characterisation in terms of their engineering properties is needed before they can be used in practice. This paper describes the bio-modification of two penetration grade bitumens with a vegetalderived biodiesel residue. Three modified bitumens were produced and the rheological characteristics were analysed using a dynamic shear rheometer (DSR). The results indicate that the addition of biobinder ‘softens’ the bitumens by decreasing the complex modulus and increasing the viscous response, except for the higher content modified blend which at high temperatures exhibits a more elastic behaviour. Similar to aged conventional binders, laboratory ageing of the blends increases the elastic response and stiffness. Keywords Bio-oil · Biodiesel · Biobinder · Dynamic mechanical analysis · Rheology
1 Introduction Most bituminous binders used for pavements are derived from fossil fuels. Their function is to fulfill the role of a binder in asphalt mixtures composed of two components: A. I. Weir (B) · G. Airey · C. Snape Nottingham Transportation Engineering Centre, Faculty of Engineering, University of Nottingham, Nottingham NG7 2RD, UK e-mail: [email protected] A. Jiménez del Barco Carrión LabIC.Ugr, University of Granada, 18071 Granada, Spain © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_198
1563
1564
A. I. Weir et al.
aggregates and binder. Aggregates provide the mixture with resistance to compression due to friction between them, while binders keep the aggregates bound together providing tensile resistance and cohesion to the mixture [1]. The binder’s rheological and adhesive properties allow these asphalt mixtures to withstand daily stresses such as traffic loads and environmental conditions without suffering excessive damage. Bitumens (conventional or modified) are the most common materials used as binders in asphalt mixtures. However, growing environmental concerns and scarcity issues of these raw materials have motivated the search for alternative and renewable sources to reduce dependency on petroleum-based binders [2, 3]. Consequently, the concept of biobinders is becoming more popular as a potential effective replacement for conventional binders. Biobinders are defined as binder alternatives made from non-petroleum-based renewable sources, which should not rival any food material, and have environmental and economic benefits [4]. They can be manufactured from a range of biomass feedstocks, including algae, wood sources, waste and vegetable oils [4]. According to the literature, they can be used as bitumen extenders (20–75% replacement), modifiers (147
>100
118
>150
86
69
G* @ 10 Hz 19 °C (kPa)
20,688
23,752
12,116
14,252
43,357
18,103
23,931
Rheological Properties and 2S2P1D Modeling …
1589
Table 2 Volumetric properties of the bituminous mixtures Mixture
Project Bitumen content (%)
Effective bitumen content (%)
VMA (%)
VBR (%)
MEA (g/cm3 )
DASR (%)
CAP 50/70-A
5.85
5.83
17.78
77.15
2.27
36.20
CAP 50/70-B
5.72
5.70
17.64
77.36
2.27
36.30
CAP 50/70-C
5.30
5.28
17.69
77.42
2.26
36.10
CAP 50/70-D
6.04
6.02
18.28
78.23
2.26
36.00
CAP 30/45
5.47
5.45
17.63
77.24
2.27
36.20
AMP 60/85
5.88
5.86
18.10
77.90
2.26
36.40
AMB 08
5.60
5.58
17.50
77.20
2.28
36.30
The properties (penetration, softening point, ductility and dynamic shear modulus—G*) of the seven bitumen are presented in Table 1. The volumetric properties (Mineral Aggregate Voids (VMA), Void Bitumen Ratio (VBR), Apparent Specific Mass (ASM) and Dominant Aggregate size range (DASR)) of the mixtures are presented in Table 2.
2.2 Experimental Procedures The test specimens were compacted to a height of 160 mm and a diameter of 100 mm using the Superpave Gyratory Compactor (SGC) based on AASHTO R 35 and M 323-17. After the molding process, the cutting and grinding of the surfaces were performed for the final diameter of 100 mm with the height of 150 mm for the dynamic modulus testing. The air void contents were set at 5.5 ± 0.5%. The influence of these bitumens on the LVE behavior of the mixtures was characterized by the compression complex modulus test based on the standard AASHTO T 342, with temperatures ranging from −10 to 54 °C and frequencies from 0.01 to 25 Hz. Axial strain amplitudes (ε0 ) from 50 to 75 μm/m were applied. From the experimental results, the 2S2P1D modeling [1] was performed.
1590
É. P. Possebon et al.
3 Results and Analysis To better understand the results obtained according to the actions detailed in the methodology, the following items show the linear viscoelastic characterization and its respective analysis.
3.1 Linear Viscoelastic Characterization The LVE behavior was obtained by average from three samples for each mixture. From the experimental data, 2S2P1D model was calibrated and the seven parameters of 2D 2S2P1D model and the constants C1 and C2 of WLF equation [3] were obtained, being presented in Table 3. When analyzing the parameters of the 2S2P1D model, some conclusions can be drawn. As E0 (asymptotic glassy modulus) and E00 (asymptotic static modulus) highly depend on mix composition [4], therefore, since aggregate and particle size characteristics are the same for all mixtures, parameters E0 and E00 should be similar or equal. According to Table 3, when checking the E00 two distinct groups are perceptible. The first three mixtures have values of close order and higher than the last four mixtures, that have very similar values. For the E0 , the values are within the same magnitude of values, but also differ from each other, with no defined pattern as to the type of bitumen. Table 3 2S2P1D model and WLF equation parameters used to model the LVE behavior. Tref = 21 °C Mixture
E*
WLF
E00 (MPa)
E0 (MPa)
K
h
δ
τE (s)
β
C1
C2 (°C)
CAP 50/70-A
68
27,886
0.240
0.651
2.48
0.047
2180
16.83
130.42
CAP 50/70-B
35
27,300
0.193
0.520
1.89
0.045
10,000
18.98
137.13
CAP 50/70-C
83
29,454
0.145
0.435
2.11
0.011
20,000
18.13
133.19
CAP 50/70-D
67
25,851
0.187
0.503
2.39
0.027
12,198
22.76
154.80
CAP 30/45
30
26,700
0.195
0.580
1.90
0.235
10,000
27.24
190.33
AMP 60/85
26
27,345
0.210
0.575
2.25
0.039
10,000
22.57
163.03
AMB 08 30
26,000
0.199
0.575
1.83
0.036
10,000
23.02
166.92
Rheological Properties and 2S2P1D Modeling …
1591
Thus, for this group of mixtures of the same granulometric composition, it is noticed that the variables E0 and E00 , although strongly linked to the stone matrix, also suffer the influence of the bitumen, either by type or bitumen content. For example, CAP50/70-C have the lowest bitumen content and it is the mixture with the highest E0 , while CAP50/70-D, with the highest bitumen content, has the lowest E0 value. Variations in model parameters may still be associated with the user’s own adjustment [2]. However, the coefficients of variation found for the data of this research in E0 and E00 are higher than those obtained in [2], reinforcing that the user variability of the model is not predominant but rather of the characteristics of the mixtures. For bitumen influence, [4] indicated that only parameters k; h; β and δ seems to depend of the bitumen. Thus, for a mixture with same bitumen, these parameters must be close. For the mixtures of this research, it can be seen in Table 3 that the different bitumens led to significant differences in these parameters. Exception is the parameter β. The value of the variations β should be considered in logarithmic scale, so it can easily be verified that a relatively small variation occurs in the corresponding simulation if the value of β is set above 1000 [2]. However, the values that led to the best calibration of the mixtures are those presented in Table 3. From the calibration of the constants of the 2S2P1D model, the LVE behavior can be graphically visualized. Time-Temperature principle independent analyzes were made in the Cole-Cole Plan and the Black Diagram, which allow visualizing and distinguishing elastic and viscous portions of the mixtures (Fig. 1). The points are the experimental results and the lines are the modeling. When analyzing the Cole-Cole and Black, the influence of the different bitumens on the viscous portion of the mixtures is verified. The phase angle response (ϕ) varies by approximately 15° with the conventional 50/70 penetration bitumen mixtures of the mixture with CAP50/70-C and CAP50/70-D showing a more elastic behavior than the others over the frequency and temperature scans tested. These two mixtures along with that of the 50/70-A were also the ones with the highest E00 . The other four mixtures (with rubber bitumen, modified SBS polymer bitumen, conventional
Fig. 1 Cole-Cole and Black diagram of Brazilian bituminous mixtures
1592
É. P. Possebon et al.
Fig. 2 Experimental data and 2S2P1D simulated master curves of dynamic modulus at 21 °C; Left) semi-log space; Right) log-log space
bitumen 50/70-B and conventional bitumen 30/4) showed very similar E00 , in addition to having a more active viscous portion with larger phase angles. To visualize the stiffness of the mixture over the full spectrum of temperatures and frequencies, the curves of the norm of complex modulus master are constructed (Fig. 2). The susceptibility of the stiffness response is verified as a consequence of bitumen type variation both for the high frequency and low temperature range (upper semi-log space region) as well as the low frequency and high temperature range (bottom region of the log-log space). The conventional 30/45 bitumen was what led to the highest stiffness between the mixtures in most of the analyzed low frequency spectrum. In contrast to this, the CAP50/70-C and CAP50/70-D mixtures have the lowest stiffness, however with the highest asymptotic static modulus. The CAP50/70-A mixture has a marked stiffness gain over the others when analyzing the low temperatures and high frequencies. The other mixtures with rubber (AB08), SBS polymer (AMP 60/85), and CAP50/70-B exhibit intermediate behavior to the others. The phase angle master curve along the temperatures and frequencies is in Fig. 3. It can be observed that the phase angle master curves of the mixtures present high variations, suggesting a strong influence of the bitumen type on the different delays between applied stresses and measured deformations, reinforcing the analysis of the Cole-Cole plane and Black Diagram of the large variation in the viscous portion of the mixtures for different bitumen types.
4 Conclusions It was verified that these mixtures had similar values of the constants E0 and E00 , however, they were not identical. For this group of mixtures of the same granulometric composition, it is noticed that the variables E0 and E00 , although strongly linked to
Rheological Properties and 2S2P1D Modeling …
1593
Fig. 3 Phase angle master curves of mixtures at 21 °C
the stone matrix, also suffer the influence of the bitumen, either by type or bitumen content. It is important to state that even mixtures made with same specification bitumens, just produced in different refineries, the LVE’s behavior changes in a ratio as large as comparing with the modified bitumens. A more complete investigation and with more samples and kinds of aggregates is needed to truly understand the behavior of Brazilian mixtures with these so different bitumens. Also, one must consider performing tests in bitumens and three-dimensionally mixtures tests.
References 1. Olard, F., Di Benedetto, H.: General, “2s2p1d” model and relation between the linear viscoelastic behaviours of bituminous bitumens and mixes. Road Mater. Pavement Des. 4(2), 185–224 (2003). https://doi.org/10.1080/14680629.2003.9689946 2. Mangiafico S., Sauzéat, C., Di Benedetto, H.: 2S2P1D model calibration error from user panel for one bitumen and one bituminous mixture. Adv. Mater. Sci. Eng. 2019, 16 pages (2019). https://doi.org/https://doi.org/10.1155/2019/6547025 3. Williams, M.L., Landel, R.F., Ferry, J.D.: The temperature dependence of relaxation mechanisms in amorphous polymers and other glass-forming liquids. J. Am. Chem. Soc. 77(14), 3701–3707 (1955) 4. Di Benedetto, H., Olard, F., Sauzéat, C., Delaporte, B.: Linear viscoelastic behaviour of bituminous materials: from bitumens to mixes. Road Mater. Pavement Des. Special Issue EATA, 163–202 (2004)
Rheological Properties and Rutting Characterization of Natural Rubber Modified Bitumen Jarurat Wititanapanit, Juan S. Carvajal-Munoz, Chakree Bamrungwong, and Gordon Airey
Abstract From the early nineties, Thailand has been the world’s largest producer and exporter of natural rubber latex (NRL). However, the current unbalances in demand/supply has led to a sharp reduction of NRL prices, which has required the Thailand government to boost NRL prices by encouraging domestic consumption and promoting its use as an alternative material for road infrastructure applications. Aiming at selecting and optimizing a suitable percentage of natural rubber latex able to improve the rheological and rutting resistance performance characteristics of natural rubber modified bitumen (NRMB), this research assesses the application of high ammonia concentrated natural rubber latex, which incorporates 60% Dry Rubber Content (DRC) and 40% water by total weight. This NRL was used as a modifier of AC60/70 bitumen at 0, 3, 5, 8% of DRC concentrations. The experimental programme comprised two stages that assessed: (1) the rheological properties by means of the Dynamic Shear Rheometer (DSR) test; and (2) the rutting performance through the Multiple Stress Creep Recovery (MSCR) test. The results indicated that NR increased the complex modulus and reduced phase angle at high temperature, while also providing significant improvement in the rutting resistance. Altogether, the main laboratory results proved that NR latex containing 8% of DRC to be the optimum NR dosage that induces improvements on the rheological properties and rutting performance of NRMB blends. Keywords Natural rubber · Modified bitumen · Rheology · Rutting performance
J. Wititanapanit (B) · C. Bamrungwong Department of Rural Roads, Ministry of Transport, Bangkok, Thailand J. S. Carvajal-Munoz · G. Airey Nottingham Transportation Engineering Centre (NTEC), University of Nottingham, Nottingham, UK J. S. Carvajal-Munoz Department of Civil and Environmental Engineering, Universidad del Norte, Barranquilla, Colombia © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_202
1595
1596
J. Wititanapanit et al.
1 Introduction Natural rubber latex (NRL) is a material produced from the rubber tree Hevea Brasiliensis, which is widely available in tropical areas such as South America and Southeast Asia. NR is generally used as manufactured product for various industry applications that include: the automobile industry, textiles, adhesives, footwear, construction applications and others. This material is recently being assessed for potential road infrastructure applications. The reason behind this interest comes from the fact that Southeast Asia currently possesses bulging stocks of NRL, along with persistent trends of decreasing trade price, which have encouraged local governments and environmental agencies to find alternative applications that help alleviating these circumstances. Furthermore, the pressing environmental needs for recycling of waste materials such has encouraged multiple investigations and practical projects intended to use rubber in its various forms, among which, recycled tyre rubber has extensively been reported in the literature for applications in pavements engineering [1–4]. In this context, the current circumstances faced by some countries in Southeast Asia with respect to NRL requires alternative applications focused on increasing domestic consumption, which also needs urgent research efforts to fully understand the suitability of this material in terms of its effects on bituminous mixtures’ performance, the cost-effectiveness and the environmental aspects involved. As a consequence, this research is encouraged by these gaps that constitute relevant needs for the society and pavement the engineering practice worldwide. It is worth considering that very limited previous studies have dealt with the use of NRL but some benefits are starting to be unveiled for road infrastructure applications. In fact, a previous research study found that NRL enhances permanent deformation resistance by means of changes of bitumen physical properties, which can be captured with various tests methods widely available nowadays [5]. Furthermore, additional studies concluded that epoxidised NR is able to alter bitumen consistency to an extent that induces enhanced durability of polymer modified asphalt (PMA) mixtures [6]. In line with current societal and environmental needs and the lack of comprehensive studies on NR, the main aim of this research study is assessing the rheological and rutting resistance natural rubber modified bitumen (NRMB). The short-term ageing effects on these responses are also considered as a secondary objective. The manuscript first discusses the effects of NRL on the rheological properties and rutting performance in order to gain further insights on the suitable use for pavement applications. The paper is completed with the most relevant findings and conclusions based on experimental data and suggestions for optimum NRL content are provided.
Rheological Properties and Rutting Characterization …
1597
Table 1 Proportions of base materials for manufacturing of the NRMB blends Blend code DRC (%) Proportion by weight for mixing AC60/70
NRL
Proportion by weight after blending (water is evaporated) AC60/70
NRL (DRC)
AC 60/70
0%
1200 g (100%)
–
1200 g (100%)
–
NRMB3
3%
1164 g (95.10%)
60 g (4.90%)
1164 g (97%)
36 g (3%)
NRMB5
5%
1140 g (91.94%)
100 g (8.06%)
1140 g (95%)
60 g (5%)
NRMB8
8%
1104 g (87.34%)
160 g (12.66%) 1104 g (92%)
96 g (8%)
2 Materials and Methods 2.1 Materials A 60/70-penetration grade bitumen from Thailand was used as a base binder (controlbinder). This base bitumen was modified with various concentrations of natural rubber latex (NRL) (3%, 5%, and 8%) as shown in Table 1. The natural rubber used contained high ammonia content, with 60% Dry Rubber Content (DRC) and 40% water by total weight.
2.2 Bitumen Blending NRMB were produced by mixing NRL with base bitumen at 0%, 3%, 5%, and 8% by weight of dry rubber content. The detailed combination of NRMB before and after production is listed in Table 1. The blending bitumen process comprised the use of high-speed shear mixer (speed range of 1500–3500 rpm) and a temperature range of 120–150 °C. The typical time required to obtain homogeneous blends was 3 h. 1200 g for each NRMB blend were produced.
2.3 Bitumen Short-Term Ageing Short term ageing for bitumen was done by means of Rolling Thin Film Oven Test (RTFOT) in accordance with ASTM D2872-19 standard.
1598
J. Wititanapanit et al.
2.4 Testing Protocol This research study considered two main laboratory tests: (1) Dynamic Shear Rheometer (DSR) test for investigating the rheological properties of bitumen and NRMB, and (2) Multiple Stress and Recovery (MSCR) test for studying permanent deformation (rutting sensitivity) (in accordance with the ASTM D7405-15 Standard).
2.5 Test Methods For the characterisation of the NRMB, rheological tests were conducted including the following: (i) Amplitude sweep test (not shown for the sake of brevity) to determine the linear viscoelastic region (LVE) of unaged and RTFO-Aged bitumen to further conduct frequency sweep tests. (ii) Frequency sweep test to study the rheological properties of bitumen at varying temperature. The data was used to produce master curves of Complex Modulus (G*) and Phase Angle (δ) and Black Space Diagrams. The Black Space Diagrams represent a simpler approach than master curves (i.e., no shifting process required) to study the relation between stiffness and viscoelastic properties of bituminous materials [7, 8]. Test parameters included: temperature from 25 °C to 85 °C with 10 °C intervals; frequency from 0.1 Hz to 10 Hz; loading mode was strain-controlled mode; limit strain was 0.5% for unaged bitumen and 0.1% for RTFO-Aged bitumen, which were within the LVE region; parallel-plate geometry of 8-mm and 25-mm spindle size with 2 mm and 1 mm gap, respectively. (iii) Multiple Stress Creep and Recovery (MSCR) test to determine the elastic recovery properties of the control bitumen and NRMB at high testing temperature. The data from MSCR was used as a measure of permanent deformation resistance (rutting sensitivity). Test parameters included: temperature of 70 °C (Thailand maximum pavement surface temperature); stress level of 0.1 kPa and 3.2 kPa; and parallel-plate geometry with 25 mm spindle size with 1 mm gap. In this study, the permanent deformation resistance of the NRMB was assessed in terms of the Average Percent Recovery (R) and the non-recoverable creep compliance (J nr ). Both R and J nr are used to measure the recovery ability (elastic behavior) of bitumen and NRMB after repeated loading.
3 Results and Discussion 3.1 Effects of NR on the Rheological Properties of NRMB The effects of NR on the rheology of NRMB are presented in this section in terms of comparisons of master curves (Complex Modulus G* and Phase Angle δ) and Black Space Diagrams. The master curves were produced from frequency sweep test conducted at various temperatures, using a reference temperature of 55 °C for
Rheological Properties and Rutting Characterization …
1599
Fig. 1 Comparison of complex modulus master curve of (a) unaged and (b) RTFO-aged bitumen
the time-temperature superposition principle. From Fig. 1 it is observed that the rheological impact of NR on the NRMB is similar for unaged and RTFO-aged blends. More precisely, the increase of NR concentration produced noticeable changes on the G* values, which are more pronounced at high loading times (lower frequencies) and high testing temperatures in contrast to the other temperatures and frequencies considered. In consequence, this stiffening effect substantiates the positive influence on the rutting response of the NRMB as discussed later in this manuscript. The results presented in Fig. 2 are consistent with the previous findings of G* master curves in terms of enhanced elastic response of the NRMB in contrast to control bitumen, which can be observed by the reduction in phase angle as the NRL concentration increases. This observation was true for both ageing conditions considered. A comparison of the Black Space Diagrams for unaged and RTFO-aged is presented in Fig. 3. The comparison indicated that increasing the NRL content produced a shifting effect of the Black Diagram curve towards higher G* and lower δ values, which was more prominent for unaged (Fig. 3a) than that for RTFO-aged materials. These results substantiate the fact that NRMB exhibits higher stiffness and a more elastic response at high temperatures which supposes better response to rutting.
3.2 Effects of NR on the Rutting Resistance of NRMB Figure 4a illustrates a sharp increase of the R-values with NRL content increase, which was true for both stress conditions considered as per standard (0.1 kPa and
1600
J. Wititanapanit et al.
Fig. 2 Comparison of Phase Angle master curve of (a) unaged and (b) RTFO-aged bitumen
3.2 kPa). These resulst justify the improvement of the elastic recovery of the modified bitumen by NRL addition and validates a higher rutting resistance of NRMB, reaching a maximum of 33% and 13% recovery at 8% NRL for 0.1 kPa and 3.2 kPa, respectively. In consistence with these results, the J nr-diff value (% difference of J nr for low and high stresses considered) increased in parallel with NRL content increase, which validates the enhanced resistance to rutting (Fig. 4b). Nevertheless, the J nr-diff suggests an impact of NRL on increased stress sensitivity when the materials undergo short term ageing, which correlated to the bituminous mixture production process. Nevertheless, high-stress sensitivity commonly occurs as a consequence of polymer modification at high concentrations and is affected by bitumen-specific responses [9, 10].
4 Conclusions This paper assessed the rheological effects and rutting performance of bitumen modification with natural rubber latex (NRL). Based on the laboratory test results, these conclusions are offered: (i) in terms of rheological effects, it was found that NRL improved the elastic response of the blends as a result of increased stiffness (i.e., reduced δ and increased G* values). This effect was more pronounced at high temperatures. (ii) The effects of NRL on rutting resistance substantiated the beneficial effect by improving the elastic recovery and rutting resistance of the base, supported by the R-values and J nr-diff . However, the stress sensitivity was affected by NRL content as commonly occurs with high-content polymeric modification of bitumen. (iii) The overall results of this research study indicated 8% NRL (DRC) as the optimum. To
Rheological Properties and Rutting Characterization …
1601
Fig. 3 Comparison of Black Space Diagrams of (a) unaged and (b) RTFO-aged bitumen
Fig. 4 Comparison of the (a) average R-values and (b) J nr-diff values for RTFO-aged
further validate these findings, testing will be conducted on aged bitumen (rheology, fatigue, fracture) and will include AC40/60 bitumen. Acknowledgements Juan S. Carvajal-Munoz thanks the support provided by Universidad del Norte and Colciencias-Colfuturo for conducting PhD studies at the University of Nottingham.
1602
J. Wititanapanit et al.
References 1. Gawel, I., Stepkowski, R., Czechowski, F.: Molecular interactions between rubber and asphalt. Ind. Eng. Chem. Res. 45(9), 3044–3049 (2006) 2. Lo Presti, D.: Recycled tyre rubber modified bitumens for road asphalt mixtures: a literature review.Constr. Build. Mater. 49, 863–881 (2013) 3. Celauro, B., Celauro, C., Lo Presti, D., Bevilacqua, A.: Definition of a laboratory optimization protocol for road bitumen improved with recycled tire rubber. Constr. Build. Mater. 37, 562–572 (2012) 4. Subhy, A., Lo Presti, D., Airey, G.: An investigation on using pre-treated tyre rubber as a replacement of synthetic polymers for bitumen modification. Road Mater. Pavement Des. 16(sup1), 245–264 (2015) 5. Al-Mansob, R., Ismail, A., Alduri, A., Azhari, C., Karim, M., Yusoff, N.: Physical and rheological properties of epoxidized natural rubber modified bitumens. Constr. Build. Mater. 63, 242–248 (2014) 6. Al-Mansob, R., Ismail, A., Yusoff, N., Albrka, S., Azhari, C., Karim, M.: Rheological characteristics of unaged and aged epoxidised natural rubber modified asphalt. Constr. Build. Mater. 102, 190–199 (2016) 7. Airey, G.: Rheological characteristics of polymer modified and aged bitumens. Ph.D. Thesis, University of Nottingham, United Kingdom (1997) 8. Airey, G.: Use of black diagrams to identify inconsistencies in rheological data. Road Mater. Pavement Des. 3(4), 403–424 (2002) 9. White, G.: Grading highly modified binders by multiple stress creep recovery. Road Mater. Pavement Des. 18(6), 1322–1337 (2016) 10. White, G.: Grading of Australian bitumen by multiple stress creep recovery. Road Transp. Res. 24(4), 30–44 (2015)
Rheological Properties of Asphalt Binder Compound Modified by Bio-oil and Organic Montmorillonite Tingting Xie, Yifang Chen, and Chao Wang
Abstract In recent years, the waste cooking oil (WCO) based bio-asphalt is developed as a potential alternative binder for the traditional petroleum asphalt. Several previous studies demonstrate that the WCO based bio-oil modification generally enhance the binder cracking resistance but has an adverse effect on high temperature performance. This paper aims to improve the rutting resistance of WCO based bioasphalt by a further compound modification of Organized Montmorillonite (OMMT). Two neat asphalt binders with penetration grades of 70 and 90 are selected in this study. The WCO based bio-oil is firstly added into the neat binders followed by further modified with the OMMT additive. The neat binders, bio-binders and Bio/OMMT binders are then characterized by rheological performance tests in this study. The frequency sweep, multiple stress creep recovery (MSCR) and linear amplitude sweep (LAS) tests are respectively conducted for linear viscoelastic properties, rutting resistance and fatigue performance evaluations. Experimental results indicate that the fatigue resistance of bio-binders at intermediate temperature are improved compared to the neat binders whereas the addition of bio-oil obviously reduces the binder rutting resistance at high temperature. Furthermore, adding the OMMT into the bio-binders is found to effectively enhance the rutting resistance when increasing the OMMT content without compromising the fatigue performance. Therefore, it can be preliminary concluded in this study that the Bio/OMMT compound modification approach can be utilized to address the rutting potential concern of WCO based bio-binders at the high temperature condition. Keywords Bio-asphalt · Organic-montmorillonite · Rheology · Fatigue · Rutting
1 Introduction Because of the limited petroleum resources, the asphalt industry is currently facing the enormous development challenges. In recent years, researchers are looking T. Xie · Y. Chen · C. Wang (B) Department of Road & Urban Railway Engineering, Beijing University of Technology, Beijing 100124, China e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_203
1603
1604
T. Xie et al.
for alternative sustainable materials to replace the asphalt binder that applied for hot mix asphalt. Bio-oil modified asphalt is widely addressed recently in asphalt industry in order to replace petroleum asphalt partly or completely with bio-asphalt. Most of previous studies examined the feasibility of utilizing various bio-oils as the asphalt modifiers. It is generally found that the bio-asphalt modified from different biomass sources have distinguished rheological behavior compared with the petroleum asphalt. The wasted cooking oil (WCO) is one of the biomasses that widely addressed for bio-asphalt application in recent years. Wang et al. found that the rutting resistance of WCO based bio-asphalt decreased whereas the binder fatigue resistance is improved compared to the control asphalt [1]. Sun et al. demonstrated that the addition of WCO based bio-oil into asphalt can reduce the softening point and viscosity but improved the permeability and ductility, meanwhile, the addition of bio-oil also decreased the dynamic modulus of asphalt mixtures [2]. In addition, Gong et al. investigated the potential of utilizing the WCO based bio-modifier as the rejuvenator for long-term aged asphalt binders [3]. It can be generally summarized from the previous research efforts that the WCO based bio-asphalt always shows reduced rutting resistance due to the bio-oil soften effect but displays improved fatigue performance. Thus, how to further address the rutting potential without compromising the fatigue performance should be a key concern for WCO based bio-asphalt application. Several studies demonstrated that adding the organic montmorillonite (OMMT) into the neat asphalt can effectively improve the its rutting resistance at high temperature conditions. From the physical properties (penetration and softening point) of the OMMT/polyethylene compound modified binders, it was reported that the binder high temperature property can be improved after the OMMT addition [4, 5]. Additionally, the aging characteristic of OMMT modified asphalt was also investigated and the OMMT additive can also help to improve the binder aging resistance [6]. The objective of this study is to investigate the possibility of improving the rutting resistance of WCO based bio-asphalt by incorporating the OMMT additives. The OMMT modification effects on the viscoelastic properties, rutting potential and fatigue resistance of bio-asphalt are to be comprehensively characterized and compared.
2 Materials and Methods 2.1 Materials Two base neat asphalt binders with penetration grades of 70 and 90 were selected in this study. Three WCO based bio-oil dosage (1, 3 and 5%) and three OMMT dosage (1, 3 and 5%) were variously utilized to prepare the bio-binders and Bio/OMMT binders. The summary of the tested asphalt binders is given in Table 1. All binders were tested in the original un-aged conditions.
Rheological Properties of Asphalt Binder … Table 1 Summary of tested asphalt binders
1605
Base binders
Binder ID
Modifiers
N70 binders
N70
None
N70 + 1%Bio
1% Bio-oil
N70 + 3%Bio
3% Bio-oil
N70 + 5%Bio
5% Bio-oil
N70 + 5%Bio + 3%OMMT
5% Bio-oil and 3%OMMT
N70 + 5%Bio + 5%OMMT
5% Bio-oil and 5%OMMT
N90 binders
N90
None
N90 + 3%Bio
3% Bio-oil
N90 + 3%Bio + 1%OMMT
3% Bio-oil and 1%OMMT
N90 + 3%Bio + 3%OMMT
3% Bio-oil and 3%OMMT
2.2 Testing Methods Frequency Sweep Test. The frequency sweep (FS) test is mainly employed to measure the linear viscoelastic parameters of asphalt binders under non-damage conditions. In this study, the FS test was conducted respectively at 35, 20 and 5 °C with the frequency of 0.1–100 rad/s. The measured dynamic modulus was horizontally shifted to construct a smooth mastercurve based on the time-temperature superposition principle. MSCR Test. The MSCR test was conducted at the same temperature of 60 °C for all binders following the MSCR procedure [7]. The percent recovery (R) and non-recoverable creep compliance (J nr ) were calculated. The J nr at 3.2 kPa (J nr3.2 ) is currently utilized in the specification to distinguish the rutting potential of different asphalt binders. LAS Test. The linear amplitude sweep (LAS) test was utilized to evaluate the fatigue properties of asphalt binder [8]. The LAS test was run at a typical intermediate temperature of 20 °C of Beijing area. The asphalt binder fatigue performance can be simulated from the LAS data interpretation based on the simplified-viscoelastic continuum damage (S-VEVD) model.
1606
T. Xie et al. 1.E+08
Dynamic Modulus (Pa)
Dynamic Modulus (Pa)
1.E+09
1.E+07
1.E+05 N70 N70+1%Bio N70+3%Bio N70+5%Bio N70+5%Bio+3%OMMT N70+5%Bio+5%OMMT
1.E+03
1.E+01 0.001
(a)
0.1
10
1000
100000
N90 N90+3%Bio N90+3%Bio+1%OMMT N90+3%Bio+3%OMMT
1.E+07
1.E+06
1.E+05
1.E+04 0.01
0.1
1
(b)
Reduced Frequency (rad/s)
10
100
1000
Frequency (rad/s) 2.0
3.0 2.0
1.0
Log ΦT
Log ΦT
1.0 0.0 -1.0
N70 N70+1%Bio N70+3%Bio N70#+5%Biol N70+5%Bio+3%OMMT N70+5%Bio+5%OMMT
-2.0 -3.0
(c)
0
10
20
0.0
N90 N90+3%Bio N90#+3%Bio+1%OMMT N90+3%Bio+3%OMMT
-1.0
30
-2.0
40
Temperature (°C)
(d)
0
10
20
30
40
Temperature (°C)
Fig. 1 Linear viscoelastic properties a dynamic modulus master curves of N70 binders, b frequency sweep at 20 °C of N90 binders, c shift factors of N70 binders, d shift factors of N90 binders
3 Results and Discussions 3.1 Linear Viscoelastic Properties The constructed dynamic modulus master curves at a reference temperature of 20 °C for all tested binders are presented in Fig. 1. It is seen that increasing the bio-oil weights dramatically decreases the stiffness of the N70 and N90 binders whereas the further OMMT addition can gradually increase the modulus of softened bio-binders. Additionally, the corresponding shift factors are also presented in Fig. 1, suggesting the temperature sensibility of the neat binders are just slightly changed due to the Bio/OMMT modification.
3.2 Rutting Resistance The time-strain curves of all tested binders during the MSCR tests are compared in Fig. 2. For both N70 and N90 binders, it is observed that the bio-oil additions
1607 100000
10000
10000
1000
1000
Strain (%)
100000
100 10 1
(a) 12
Jnr(kPa-1)
10
100
N70 N70+1%Biol N70+3%Bio N70+5%Bio N70+5%Bio+3%OMMT 70N+5%Bio+5%OMMT 0
50
100 150 Time (seconds)
200
250
1
(b) 4
0.1kPa 3.2kPa
8 6 4
0
50
100 150 Time (seconds)
200
250
0.1kPa 3.2kPa
3
2
1
2 0
(c)
N90 N90+3%Bio N90+3%Bio+1%OMMT N90+3%Bio+3%OMMT
10
Jnr(kPa-1)
Strain (%)
Rheological Properties of Asphalt Binder …
0
(d)
Fig. 2 MSCR test results a time-strain responses of N70 binders, b time-strain responses of N90 binders, c J nr results of N70 binders, d J nr results of N90 binders
gradually increase the permanent deformation, however, further OMMT modification show less rutting potential. Especially, when the weights of bio-oil and OMMT are equal, the compound modified Bio/OMMT binder displays much lower nonrecoverable strain than the neat N70 binder. The J nr is currently the specification parameter to distinguish the rutting potential of asphalt binders. The J nr under two creep stress levels are also compared in Fig. 2, in which similar trends are observed that the asphalt binder rutting resistance is decreased with the bio-oil modification but increased when the OMMT is further added.
3.3 Fatigue Performance The LAS based fatigue life prediction of the tested binders are summarized in Fig. 3. It is seen that the bio-oil are demonstrated again to improve the binder fatigue performance. Regarding to the OMMT impact, lower OMMT weight also shows positive effects on binder fatigue resistance, however, increasing the OMMT weight displays negative impact on the binder fatigue. Therefore, the Bio/OMMT compound modification still needs further investigation on the binder fatigue performance.
1608
T. Xie et al.
6,000
2.5%Strain
5,000
5%Strain
5,000
2,000 1,000
3,000 2,000 1,000
0
(a)
5%Strain
Nf
Nf
4,000 3,000
2.5%Strain
4,000
0
(b)
Fig. 3 Strain-controlled fatigue life prediction a N70 binders, b N90 binders
4 Conclusions This paper presents a study on improving the rutting resistance of WCO based bioasphalt by a further compound OMMT modification. The OMMT addition into the bio-binders was found to effectively enhance the rutting resistance without compromising the fatigue performance. Therefore, it can be preliminary concluded that the Bio/OMMT compound modification approach can be utilized to address the rutting potential concern of the WCO based bio-binders. Acknowledgments The authors would like to acknowledge the sponsorship from Beijing Municipal Education Commission (KM201810005020) and National Natural Science Foundation of China (51608018).
References 1. Wang, C., Xue, L., Xie, W., You, Z., Yang, X.: Laboratory investigation on chemical and rheological properties of bio-asphalt binders incorporating waste cooking oil. Constr. Build. Mater. 167, 348–358 (2018) 2. Sun, Z., Yi, J., Huang, Y., Feng, D., Guo, C.: Properties of asphalt binder modified by bio-oil derived from waste cooking oil. Constr. Build. Mater. 102, 496–504 (2016) 3. Gong, M., Yang, J., Zhang, J., Zhu, H., Tong. T.: Physical-chemical properties of aged asphalt rejuvenated by bio-oil derived from biodiesel residue. Constr. Build. Mater. 105, 35–45 (2016) 4. Fang, C., Yu, R., Zhang, Y., Hu, J., Zhang, M., Mi. X.: Combined modification of asphalt with polyethylene packaging waste and organophilic montmorillonite. Polymer Testing 31(2), 276–281 (2012) 5. Fang, C., Liu, X., Yu, R., Liu, P., Lei. W.: Preparation and properties of asphalt modified with a composite composed of waste package poly (vinyl chloride) and organic montmorillonite. J. Mater. Sci. Technol. 30(12), 1304–1310 (2014) 6. Zhang, H., Yu, J., Wang, H., Xue L.: Investigation of microstructures and ultraviolet aging properties of organo-montmorillonite/SBS modified bitumen. Mater. Chem. Phys. 129(3) (2011) 7. AASHTO.: Standard method of test for multiple stress creep recovery (MSCR) test of asphalt binder using a dynamic shear rheometer (DSR). AASHTO T 350-14, Washington D.C (2018)
Rheological Properties of Asphalt Binder …
1609
8. AASHTO.: Standard method of test for estimating damage tolerance of asphalt binders using the linear amplitude sweep. AASHTP TP 101, Washington D.C (2014)
Rheological Properties of Bituminous Mastic with Different Filler Content Xi-li Yan, Peng Li, Qing-long You, and Xi-juan Xu
Abstract Bituminous mixture is widely used for pavement construction. It is made of coarse aggregates, fine aggregates and mineral filler, and mixed with bitumen. The coarse aggregates form a skeleton structure in mixture and the fine gravel with bitumen form a mortar. The mineral filler with bitumen forming a mastic, which determines the rheological characteristics of bituminous mixture. This study focuses on rheological properties of bituminous mastic with different content of mineral filler and with different types of bitumen such as base bitumen, SBS modified bitumen and warm bitumen (modified by Evotherm or Sasobit). By measuring the penetration, ductility and softening point of different mastics, and by applying the Brookfield test and DSR test, the viscosity-temperature curve and viscoelastic property are analyzed. The contribution of mineral filler content in mastic is discussed. The difference of rheological property among different mastic is emphasized, and the effects of warm mixing are analyzed. The results show that the technical parameters of bituminous mastic (penetration, ductility, softening point and viscosity) change linearly with the mineral filler content in normal or semi-logarithmic coordinates. The warm bituminous mastic has a larger phase angle in dynamic viscoelastic analysis and SBS bituminous mastic has a smaller phase angle. The warm additive of type “surfactant” doesn’t change the technical parameters of bitumen, but increases dynamic shear modulus, the warm type of “bitumen viscosity reduction” changes the technical parameters of bitumen. Keywords Bituminous mastic · Mineral filler content · Modified bitumen · Warm bitumen · Rheological property
X. Yan · P. Li (B) · Q. You School of Highway, Chang’an University, Xi’an, People’s Republic of China e-mail: [email protected] X. Xu Xi’an Highway Research Institute, Xi’an, People’s Republic of China © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_204
1611
1612
X. Yan et al.
1 Introduction Asphalt mixture has been widely used for paving project all over the world. It is a typical viscoelastic material and composed of asphalt binder and mineral aggregate. Generally, the mineral aggregate used to produce asphalt mixture can be divided into coarse aggregate, fines and filler according to their size. Coarse aggregate builds the main structure of asphalt mixtures, and fines fill the voids among coarse aggregate. Asphalt binder and filler form the mastic, which works as the bounding material to hold all particles together. The property of the mastic determines the viscoelasticity of overall mixture and affects the pavement performance deeply. A great number of studies have been conducted to evaluate and model the properties of bituminous mastic from the perspective of viscoelasticity and constitutive models. Dynamic shear rheometer (DSR) has been widely used to evaluate viscoelastic properties of mastics [1, 2]. Al-Khateeb et al. proposed a simplified model to characterize the micro mechanism of mastics and estimate its mechanical behavior. Delaporte et al. [3] developed an annular shear rheometer to investigate the linear viscoelastic properties of both binders and mastics. The effects of filler characteristics and aging of the binder on the complex shear modulus have been measured and analyzed. Hesami et al. [4] proposed an empirical framework to describe the viscosity of asphalt mastic. Riccardi et al. [5] simulated rheological properties of virgin and Reclaimed Asphalt Pavement (RAP) mixed binder based on the modified Nielsen model and DSR measurements. Apeagyei et al. [6] investigated the moisture induced damage in asphalt mixture from the perspective of the aggregate-mastic interfacial properties using advanced laboratory testing methods. Harris and Stuart [7] evaluated effects of mineral filler and mastics on the performance of stone matric asphalt (SMA). This study focuses on rheological properties of asphalt mastic with different mineral filler contents and warm mix additives. By measuring the penetration, ductility and softening point of different mastics, their viscoelastic properties are analyzed.
2 Asphalt Mastic Preparation Four types of asphalt binders were used for this study, i.e. 70# neat binder, SBS modified binder, Evotherm modified binder and Sasobit modified binder. The mineral filler was made of limestone. Generally, the filler to binder ratio ranges from 0.6 to 1.6 in most commonly used asphalt mixture. The filler used in SMA could reach up to 1.6. Therefore, five filler to binder ratios were considered to prepare mastics in this study, i.e. 0.8, 1.0, 1.2, 1.4, 1.6. Mastic was manually mixed. Neat binder and warm additive modified binder were heated to 135 °C and SBS modified binder was heated to 145 °C The fillers was heated to 10 °C higher than the temperature of the binder. The total weight of each
Rheological Properties of Bituminous Mastic …
1613
batch was about 200 g. Filler was added gradually during mixing. As temperature dropped, the mixing become difficult and the bowl need to be place back into the oven. Such processes usually were repeated three to four times till no bubbles were observed. Laboratory tests, including penetration, ductility, DSR and etc., were used to evaluate the properties of binders and mastics.
3 Testing Results and Discussion 3.1 Penetration, Ductility and Softening Point Penetration, ductility and softening point are three basic parameters adopted by the Chinese specifications. Tests have been conducted following the requirements in Technical Specifications for Construction of Highway Asphalt Pavements. Measurements of penetration, ductility and softening point are presented in Fig. 1. As expected, as filler content increases, both penetration and ductility of mastics made of four types of binders decrease and their softening points increase. During the ductility test, two kinds of breaking interface were observed, the point break and the fractured edge. The point break means that, the mastic remains flexibility. On the other hand, as filler content increased, mastic became brittle and the fractured edge was observed. The filler content, at which mastic became brittle, is affected by adding SBS modifiers, and the range is from 0.8 to 1.0. However, adding warm additive does not affect this turning point. In addition, Fig. 1 also shows that, the properties of neat binder and Evotherm modified binder are almost the same. Evotherm is a surfactant based additive. During the mixing process, moisture evaporated and therefore the effect of reducing viscosity was not observed.
3.2 Brookfield Viscosity The Brookfield viscosity tests were conducted following Standard test methods of bitumen and bituminous mixtures for highway engineering. The flow characteristic of mastics was evaluated at 135 °C by measuring the shear stress and shear velocity. The viscosity versus Temperature curves were drawn for each type of mastic. Effects of filler content were analyzed. Flow Characteristic of Mastics. At 135 °C viscosities were measured under different shear velocities for mastics made of four types of binders at four filler to binder ratios, i.e. 0, 0.8, 1.2, 1.6. The velocity versus Stress relationships were drawn in Fig. 2. Generally, the relationship curves are linear and pass the original point, which means at high temperature, the mastic can be considered as Newton fluid.
1614
X. Yan et al.
80
SBS 70# Evotherm Sasobit
70
SBS 70# Evotherm Sasobit
50
Ductility/mm
Penitration/0.1mm
60
40
10
30
20 0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
110
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
Filler to Binder Ratio
Filler to Binder Ratio
SBS 70# Evotherm Sasobit
100
Softening Point /
90 80 70 60 50 40 0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
Filler to Binder Ratio
Fig. 1 Variation of technical index with the mineral filler content for bituminous mastics
The exception was found for SBS modified binder with 1.2 and 1.6 filler contents. Intercepts to Y axis are 2 and 5 Pa respectively, which means that the mastic is not Newton fluid. Viscosity versus Temperature Curve. Viscosities were measured at five temperatures, i.e. 125, 135, 145, 155, 165 °C. Testing results of asphalt binder and mastics with 1.2 filler contents are presented in Fig. 3. In the half logarithmic chart, linear relationship can be observed. The viscosity of mastic with Sasobit modified binder is significantly lower than the rest of mastics. However, such reduction of viscosity was not observed on Evotherm modified mastics. Effects of Filler to Binder Ratio. Figure 4 illustrate the relationship between viscosity and filler to binder ratio at 125 and 135 °C. At both temperatures, linear relationships were observed. Mastic made of SBS modified binder has the highest viscosity and Mastic made of Sasobit modified binder has the lowest. The effects of adding Evotherm is not significant.
Rheological Properties of Bituminous Mastic …
1615 500
250
450 400
150
70# Filler to Binder Ratio 0 0.8 1.2 1.6
100
Shear Stress/Pa
Shear Stress/Pa
200
SBS Filler to Binder Ratio 0 0.8 1.2 1.6
350 300 250 200 150 100
50
50 0
0
20
40
60
80
0
100
0
10
250
250
200
200
150
Sasobit Filler to Binder Ratio 0 0.8 1.2 1.6
100
50
0
0
20
40
60
80
Shear Velosity/Pa
Shear Stress/Pa
Shear Velosity/s-1
20
Shear Velosity/s-1
150
30
40
Evotherm Filler to Binder Ratio 0 0.8 1.2 1.6
100
50
0
100
0
20
40
60
80
100
Shear Velosity/s-1
Shear Velosity/s-1
Fig. 2 Experimental rheogram of different bituminous mastics at 135 °C
SBS 70# Evotherm Sasobit
SBS 70# Evotherm Sasobit
10
ωf=0
ωf=1.2
Viscosity/Pa·s
Viscosity/Pa·s
1
1
0.1 125
135
145
155
165
125
135
145
155
165
Fig. 3 Viscosity-temperature curves of bitumen and bituminous mastics (ω f = 0和1.2)
3.3 DSR Testing Results As discussed before, the significant effects of adding Evotherm to mastics have not been observed by measuring penetration, ductility, softening points and viscosity at high temperature. Further, DST tests were conducted at four lower temperatures, i.e. 58, 64 and 70 °C. Dynamic shear modulus and phase angles were measured and
1616
X. Yan et al.
recorded, the rutting index, G*/sinδ, was also calculated. Results are presented in Figs. 5 and 6. Increasing fillers content leads to the increases of dynamic shear modulus and therefore rutting index. At each temperature, the phase angles of different mastics remain approximately the same, which means adding Evotherm does not significantly
SBS 70# Evotherm Sasobit
10
10
135℃
Viscosity/Pa·s
Viscosity/Pa·s
125℃
1
0.0
SBS 70# Evotherm Sasobit
1
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
0.0
0.2
0.4
0.6
Filler to Binder Ratio
0.8
1.0
1.2
1.4
1.6
Filler to Binder Ratio
Fig. 4 Viscosity of bituminous mastics with the mineral filler content (125 and 135 °C) 86
20
70# Evotherm ωf=1.0
18 16
85
Phase Angle/°
14
G*/kPa
70# Evotherm ωf=1.0
12 10 8
84
83
6 4 2 58
60
62
64
66
68
82 58
70
60
62
64
66
68
70
1.0
1.2
Fig. 5 Variation of viscoelastic index with temperature for two mastics (ω f = 1.0)
58 64 70
58 64 70
70#
Evotherm
10
G*/kPa
G*/kPa
10
1 0.0
0.2
0.4
0.6
0.8
Filler to Binder Ratio
1.0
1.2
1 0.0
0.2
0.4
0.6
0.8
Filler to Binder Ratio
Fig. 6 Variation of the dynamic shear modulus with mineral filler content
Rheological Properties of Bituminous Mastic …
1617
changes of proportion of the viscosity and elasticity of the mastics. However, mastics with Evotherm has the higher shear modulus as shown in Fig. 5, leading to a higher rutting resistance.
4 Conclusion The following conclusions can be reached from this study. Generally, at high temperature, the mastic can be considered as Newton fluid. However, at the same temperature, SBS modified mastics with 1.2 and 1.6 filler contents exhibit non-Newton fluid characteristics. As increase of filler contents in mastics, softening points increases. Meanwhile, the penetration and the ductility decrease linearly in the half logarithmic chart. The filler content does not affect the phase angle. Adding Sasobit increases the softening point of asphalt binder and mastic, and reduces the rotation viscosity. The similar effects have not been found for Evotherm modified binder and mastics. However, Evotherm improves the dynamic shear modulus of mastic and therefore increases the rutting resistance. Acknowledgements This study has been partially supported by the Shaanxi Provincial Highway Research Foundation [grant number 17-11K] and the Sichuan Provincial Highway Research Foundation of China [grant number 2017-F-01].
References 1. Al-Khateeb, G.G., Irfaeya, M.F., Khedaywi, T.S.: A new simplified micromechanical model for asphalt mastic behavior. Constr. Build. Mater. 149, 587–598 (2017) 2. Shashidhar, N., Shenoy, A.: On using micromechanical models to describe dynamic mechanical behavior of asphalt mastics. Mech. Mater. 34(10), 657–669 (2002) 3. Delaporte, B., Di Benedetto, H., Chaverot, P., et al.: Linear viscoelastic properties of bituminous materials: from binders to mastics. J. Assoc. Asphalt Paving Technol. 76, 455–494 (2007) 4. Hesami, E., Jelagin, D., Kringos, N., et al.: An empirical framework for determining asphalt mastic viscosity as a function of mineral filler concentration. Constr. Build. Mater. 35, 23–29 (2012) 5. Riccardi, C., Falchetto, A.C., Losa, M., et al.: Rheological modeling of asphalt binder and asphalt mortar containing recycled asphalt material. Mater. Struct. 49(10), 4167–4183 (2016) 6. Apeagyei, A., Grenfell, J., Airey, G.: Moisture-induced strength degradation of aggregate-asphalt mastic bonds. Road Mater. Pavement Des. 15(S1), 239–262 (2014) 7. Harris, B.M., Stuart, K.D.: Analysis of mineral fillers and mastics used in stone matrix asphalt. J. Assoc. Asphalt Paving Technol. 64(1), 211–234 (1995)
Rheological Properties of Derived Fractions Composed of Aromatics, Resins, and Asphaltenes Feipeng Xiao and Jiayu Wang
Abstract The chemical makeup of the asphalt binder determines its rheological and mechanical properties. The objective of this study is to reveal the rheological properties of various fractions composed of aromatics, resins, and asphaltenes. First, asphalt binder (pen grade 60/80) was separated into four fractions and then resins and asphaltenes were respectively doped into aromatics to fabricate four derived fractions. The frequency-temperature sweep was conducted by a dynamic shear rheometer (30–80 °C, 0.01–20 Hz, and 0.01% shear strain). The complex shear modulus master curves of various fractions were developed based on measured data. The test results showed that the increase in the content of resins or asphaltenes resulted in the increased complex shear modulus, stated that an increase in polarity leaded to higher stiffness. Meanwhile, various fractions had different sensitivity to temperature. Resins showed the most sensitivity to temperature change among all the fractions. The synergy of resins and aromatics resulted in the different master curve shapes of derived fractions. Furthermore, the predicted model of the master curve based on composite material idea showed a good fitting trend within a frequency over 0.1 Hz. Future work should focus on the interaction of aromatics/resins and aromatics/asphaltenes to fully determine the effect of SARA fractions. Keywords SARA · Derived fraction · Rheological properties · DSR · Master curve
1 Introduction It has been well-known that asphalt binder is the residue from the crude oil refining process and mainly composed of H, C, N, S, O, and other metal elements. The separation study of asphalt binder started from 1836, Boussingault first separated native asphalt into a distillable fraction called petrolene and a solid fraction, asphaltene [1]. A few decades later, Richardson defined the asphaltenes as the insoluble part in F. Xiao (B) · J. Wang Tongji University, Laboratory of Road and Traffic Engineering of Ministry of Education, Shanghai 201804, China e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_205
1619
1620
F. Xiao and J. Wang
Table 1 Summarization of SARA characteristics [5] Characteristics Saturates
Aromatics
Resins
Asphaltenes
Percent wt.%
5–15
30–45
30–45
5–20
Color
Colorless
Yellow to red
Black
Black
GTT °C
−70
−20
Not clear
No transition
H/C
Close to 2
1.5
1.4
1.1
MW g/mol
470–800
570–980
780–1400
800–3500
Composition
Long aliphatic Slightly aliphatic, Complex aromatic Many condensed chains, very few condensed rings aromatic rings and aromatic rings aromatic rings metal
Density g/cm3
Around 0.9
1
1.07
1.15
Notes GTT is glass transition temperature, and the test temperature of density is 20 °C
naphtha and malthenes as the soluble part [2]. The complexity of malthenes composition made it hard to conduct in-depth studies, Corbett proposed the reference method by using elution-adsorption liquid chromatography on active alumina with solvents of increasing polarity and aromaticity to separate malthenes to into saturates (S), aromatics (A) and resins (R) [3]. The current standard test method of ASTM D-4124 is modified from Corbett’s method [4]. Table 1 shows the basic characteristics of four separations (SARA: Saturate, Aromatic, Resin, and Asphaltene). The relationship between the asphalt components and properties has been studied for many years. In 1979, Dealy mixed asphaltenes and asphalt binder at different proportions and found that the viscosity of the asphalt binder increased with the increase of asphaltenes content and decreased with the increase of maltenes content [6]. Other researchers also found similar trends of asphaltenes by adding reclaimed asphalt binder [7], polyphosphoric acid (PPA) [8], and fluid catalytic cracking slurry [9]. As agreed, the rheological properties of asphalt are predominantly determined by the physicochemical properties of asphalt binder and SARA dependence. Sultana and Bhasin produced four derivative binders from two parent binders by adding SARA fractions. They found that an increase in polar fractions content increased both stiffness and tensile strength [10]. Xu et al. also used a similar method and indicated that shape index R of the master curve, creep compliance and residual relaxation model showed a decreased trend with the increase of asphaltene or resin [11]. More detailed studies focused only on SARA fractions and their interactions. Wang et al. studied the rheological and chemical properties of SARA fractions and developed relationships between the chemical properties of SARA fractions and rheological performance of asphalt binders [12]. Wei et al. revealed the interactions between SARA fractions through TG/DSC (Thermogravimetry/Differential Scanning Calorimetry) method [13]. The former studies indicated an effective way to analyze the effect of SARA fractions on the rheological properties of asphalt binders by using derived asphalt binders with different fraction percentages. Nevertheless, the rheological properties of individual fractions, especially asphaltenes, are not yet well understood due to the
Rheological Properties of Derived Fractions …
1621
insufficient amount of SARA fractions through the traditional separation method. Therefore, the objective of this study is to reveal the rheological properties of various fractions composed of aromatics, resins, and asphaltenes.
2 Materials and Test Design 2.1 Materials One type of asphalt binder (penetration grade 60/80) was selected for SARA separation and further testing. 10 g (±1 g) of asphalt binder was separated into four fractions according to a modified ASTM D 4124-09 method [14], then, resins were doped into aromatics fraction at the weight percentages of 58.4 and 32.6% (referred AR1 and AR2, respectively). Asphaltenes were also added into aromatics fractions at the ratio of 21.5% and 4.1% (referred AAs1 and AAs2, respectively). The flow chart of the experiment design is shown in Fig. 1.
2.2 Separation and Derivation Process Separation process During the separation process, the asphalt binder was first separated into n-heptane insoluble asphaltenes and soluble maltenes. Then the maltenes were divided into saturates, aromatics, and resins using liquid chromatographic technology. The eluents were heptane for saturates, toluene and toluene/alcohol for aromatics, and finally alcohol for Reins. The collection rate of each fraction was 10–20 ml/min in the receiving funnel. Derivation process During the derivation process, two asphaltene/aromatics fractions and two resins/aromatics fractions were obtained. First, each fraction was fully dissolved by toluene using an ultrasonic oscillator (60 °C for 10 min). After that, the rotary evaporation device was used to separate solvents and solutes at 110 °C for around 30 min. To make sure the toluene is completely evaporated, the fraction was dried in a vacuum oven for 20 min at 170 °C.
2.3 Rheological Tests The rheological tests were performed by a Malvern Kinexus rheometer. Resins, aromatics, and four derived fractions were tested. All tests were conducted with an 8-mm diameter parallel plate geometry and a 2-mm gap setting due to the limited
1622
F. Xiao and J. Wang
weight of fractions. The frequency sweep tests were conducted between 0.01 and 20 Hz. Four derived fractions were tested at the temperature range of 30–80 °C, aromatics were tested from -10 to 40°C, while resins were tested from 80 to 130 °C. The controlled strain mode with an applied strain of 0.01% was selected to ensure the tests remained in the linear viscoelastic region. The Arrhenius equation was used as TTSP (Time-temperature superposition principle) to develop master curves for all the fractions. Equation (1) is the standard logistic model for complex shear modulus master curve, and Eq. (2) is the TTSP equation. Log G ∗ = θ +
α−θ 1 + eβ+γ •log fr
(1)
Where, G ∗ is complex shear modulus (kPa), θ is equilibrium modulus, α is glassy modulus, fr is reduced frequency, β and γ are two shape parameters. 1 1 Ea − log fr = log f + 2.303 × R T Tr
(2)
where, f is test frequency (Hz), Tr is reference temperature (◦ K), T is test temperature, (◦ K), E a is activation energy (J • mol−1 ), R is universal gas constant (J mol−1 K−1 ).
Fig. 1 Flow chart of experiment design
Rheological Properties of Derived Fractions …
1623
3 Results Figure2 summaries the test results from the frequency-temperature sweeps, some predicted data of aromatics and resins were calculated from their master curves from the previous study [12]. Figure 2a indicates complex shear modulus at various frequencies at 40 °C. As expected, the complex shear modulus increased with the increased frequency, the resins showed the highest complex shear modulus, while aromatics indicated the lowest complex shear modulus. In other words, for different fractions, an increase in the amount of resins increased the complex modulus. A similar trend can also be found in the derived fractions AAs1 and AAs2. These results suggested that an increase in polarity leads to higher stiffness and modulus. In Fig. 2b, the loading frequency was selected as 1.585 Hz, which was close to normal loading frequency in the pavement. The complex shear modulus of each fraction showed a decreasing trend when the temperature rose, but their descent rates were different. For example, the fractions containing aromatics showed a slight decrease in descent rate in double logarithmic coordinates. Therefore, the resins were most sensitive to temperature among all the fractions. Also, with the increased content of aromatics, the complex shear modulus decreased, which can be attributed to the less polarity. Data from various test temperatures were horizontally shifted to form a single smooth curve at 25 °C based on the Arrhenius equation, as shown in Fig. 3. Different fractions indicated the various shapes of complex shear modulus master curves. For example, the aromatics increased slightly below 0.1 Hz and rose dramatically over 0.1 Hz, while the resins showed an increasing trend below 0.1 Hz and kept unchanged over 0.1 Hz. The synergy of resins and aromatics resulted in the shapes of AR1 and AR2. Also, the complex modulus of AR1 fraction at a high frequency was even greater than that of resins because of the interaction effect between resin and aromatic. Similar trends can also be found in AAs1 and AAs2, although the complex shear modulus of asphaltenes was unknown. Furthermore, an idea from composite material was introduced into this study. Equation (3) was used to predict the complex shear modulus of AR1 and AR2. The assumption of this equation was the same shear strain during the loading. 1/G∗derivated f raction = V A × γ A /G∗ A + (1 − V A ) × γ R /G∗ R
(3)
Where, G∗derivated f raction is the complex shear modulus of derivated fraction, V A is the volume of aromatics fraction, γ A and γ R are two constant values of aromatics and resins, respectively. G∗ A and G∗ R are complex shear modulus of aromatics and resins, respectively. As shown in Fig. 3, the predicted curve showed good fitting effect (R-square ≈1.0) over the frequency of 0.1 Hz, but the gap between measured data and predicted data increased sharply as the frequency decreased, which could be attributed to the liquid aromatics. More data points would be needed to fully correlate the rheological parameters and chemical properties based on FTIR and GPC tests in the future study.
1624
F. Xiao and J. Wang Aromatics AAS2
Resins AR1
AAS1 AR2
Aromatics Predicted Aromatics Resins Predicted resins AAS1 AAS2 AR1 AR2
100000
Complex shear modulus (kPa)
Complex shear modulus (kPa)
100000
10000
1000
100
10
1
10000
1000
100
10
1
0.1
1
10
100
-20
0
20
(a) Results at various frequencies at 40
40
60
Temperature (
Frequency (Hz)
80
100
120
140
)
(b) Results at various temperatures at 1.585 Hz
Fig. 2 Frequency sweep test results for different fractions at 40 °C
Fig. 3 Complex shear modulus master curve of different fractions
4 Conclusions • It is possible to measure the rheological properties of both aromatics/resins and aromatics/asphaltenes derived from asphalt binder after a modified separation process. • An increase in the amount of resins or asphaltenes improved the complex shear modulus value, stated that an increase in polarity would lead to higher stiffness and modulus values.
Rheological Properties of Derived Fractions …
1625
• Various fractions had different sensitivity to temperature. Resins were most sensitive to temperature among all the fractions. • The synergy of resins and aromatics resulted in the different master curve shapes of AR1 and AR2. The predicted model based on the composite materials showed a good fitting trend within over the frequency of 0.1 Hz. Future research work will focus on the interactions between two fractions and their chemical effects.
References 1. Boussingault, J.: Extrait d’un mémoire sur la composition des bitumes. CR Acad. Sci. 3, 375–378 (1836) 2. Richardson, C.: The Modern Asphalt Pavement. Wiley (1905) 3. Corbett, L.W.: Composition of asphalt based on generic fractionation, using solvent deasphaltening, elution-adsorption chromatography, and densimetric characterization. Anal. Chem. 41(4), 576–579 (1969) 4. Materials, A.S.f.T.a.: Test Methods for Separation of Asphalt Into Four Fractions. ASTM, Philadelphia (2001) 5. Lesueur, D.: The colloidal structure of bitumen: consequences on the rheology and on the mechanisms of bitumen modification. Adv. Colloid Interface Sci. 145(1–2), 42–82 6. Dealy, J.M.: Rheological properties of oil sand bitumens. Can. J. Chem. Eng. 57(6), 677–683 (1979) 7. Salvatore, et al.: Effect of colloidal structure of bituminous binder blends on linear viscoelastic behaviour of mixtures containing Reclaimed Asphalt Pavement (2016) 8. Catarina, V., et al.: Influence of polyphosphoric acid on the consistency and composition of formulated bitumen: standard characterization and NMR insights. J. Anal. Methods Chem. 2016, 1–6 (2016) 9. Liang, M., et al.: Rheological and chemical characterization of deoiled asphalt modified with FCC slurry. Mater. Struct. 49(9), 3607–3617 (2016) 10. Sultana, S., Bhasin, A.: Effect of chemical composition on rheology and mechanical properties of asphalt binder. Constr. Build. Mater. 72, 293–300 (2014) 11. Xu, Y., Zhang, E., Shan, L.: Effect of SARA on rheological properties of asphalt binders. J. Mater. Civ. Eng. 31(6), 04019086 (2019) 12. Wang, J., et al.: Modelling of rheological and chemical properties of asphalt binder considering SARA fraction. Fuel (2019) 13. Wei, B., et al.: Integrative determination of the interactions between SARA fractions of an extra-heavy crude oil during combustion. Fuel 234, 850–857 (2018) 14. Wang, T., et al.: Effects of SARA fractions on low temperature properties of asphalt binders. Road Mater. Pavement Des. 1–18 (2019)
Rheometrics Framework Analysis to Capture Self-assemblies Organization in Bitumen Matrix Caroline Bottelin, Marc Chardonnet, Philippe Marchal, Lazaros Vozikis, and Yvong Hung
Abstract In this work, the liquid to solid transition of self-assemblies into bitumen matrix obtained with gelator additives is studied through rheometrics framework analysis. A very small amount of these organic compounds is able to harden bitumen and to shift the temperature range where they display solid-like behavior. Current test methods such as the softening point, the penetration needle test or the indentation depth are not sufficient to describe the gelling process of such soft matter. Rheometrics analysis are found to be more appropriate to rigorously determine the liquid to solid transition of the system as well as their mechanical properties. In opposite to the classical methods of characterization, the rheological experiments require a very small volume of sample, they are more sensitive and they give more information. Thanks to the sensitivity of rheometrics analysis, different temperatures can be identified such as the temperature of the Newtonian/viscoelastic transition and the temperature of the liquid/solid transition of pure or modified bitumen. Another advantage of rheometrical test is the characterization of the different viscoelastic behaviors of these materials with enlarged time spectrum. Keywords Bitumen · Self-assemblies · Rheometry · Gelling temperature · Viscoelasticity
1 Introduction Molecular self-assembly and network formation by a very small amount of gelators in a liquid has experienced an enormous increase in interest during the past decade, due to their ability to turn liquids into thermoreversible gels [1]. The literature studies performed on these materials have shown that the sol-gel transition C. Bottelin · P. Marchal · L. Vozikis Laboratoire Réactions et Génie des Procédés, UMR 7274, CNRS/Université de Lorraine, 1 rue Granville, BP 20451, 54001 Nancy Cedex, France M. Chardonnet · Y. Hung (B) Bitumen Research Team, Centre de Recherche de Solaize, Total Marketing and Services, Chemin du Canal, 69360 Solaize, France e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_206
1627
1628
C. Bottelin et al.
temperature can be accurately determined via rheological measurements and that the thermomechanical history, such as the cooling rate or the shear stress applied, has a strong impact on the mechanical properties of the organogels [2]. However, the use of these gelator additives to harden solid materials such as bitumen has been poorly studied in the literature [3]. The current test methods such as the softening point, the penetration needle test or the indentation depth are not able to accurately characterize the bitumen/gelator systems. Therefore, we present in this paper a reliable procedure for monitoring the gelation of modified bitumen via rheometrics analysis under a controlled thermomechanical history.
2 Experimental Section The effect of two gelling additives, called A and B, on the rheological behavior of the bitumen was studied. The pure bitumen matrix was also analyzed. Rheological experiments were performed on a stress imposed rheometer (AR2000, TA Instruments) equipped with a 40 mm diameter parallel plates geometry. The temperature of the lower plate surface was controlled within ±0.1 °C by a Peltier plate. A gap of 800–1000 μm was employed in all the measurements. Before the samples were loaded on the Peltier plate, they were all heated at 160 °C in vials for ≈30 min then manually stirred. The solution (2–3 mL) sampling was then placed on the rheometer at 160 °C and pre-sheared at a shear rate of 200 s−1 for ≈10 min to obtain a homogeneous solution. The rheometer was cooled with a cooling rate of 5 °C min−1 from 160 to 0 °C. Oscillatory measurements at small amplitude were carried out during the cooling process in order to follow the kinetics of the system as a function of temperature. The strain amplitude was fixed at 10–3 and the angular frequency at 10 rad s−1 . At T = 0 °C, after this temperature was kept for 10 min, the temperature ramp was reverted and the sample was heated back to 60 °C. Mechanical spectroscopy experiments were carried out at 60 °C to determine the effect of the gelling additives on the mechanical properties of the bitumen.
3 Results and Discussion 3.1 Liquid to Solid Transition Spectrum of the pure bitumen. The kinetics of solidification of the pure bitumen under cooling conditions is shown in Fig. 1. At high temperature, the pure bitumen behaves as a Newtonian liquid and the viscous modulus-temperature dependence follows Arrhenius-type behavior (ln G” ∝ T −1 ). By decreasing the temperature, both viscoelastic moduli increase but the elastic modulus increases faster than the
Rheometrics Framework Analysis to Capture Self-assemblies …
1629
Fig. 1 Storage (G’) modulus, loss modulus (G”) and phase angle (δ) versus temperature (T) for the pure bitumen, at a fixed frequency of 10 rad s−1 and a strain of 10−3 . The dashed line represents the fit of Arrhenius’s law to the experimental points. The straight lines materialize the limits of the three regions. The α transition temperature is defined by the deviation from the Arrhenius behavior and the cross-over transition temperature is defined by the crossover point G = G”.
viscous modulus, indicating the formation of a rigid structure. Two relaxations mechanisms can be clearly identified: (i) the first is the Newtonian/viscoelastic relaxation due to the α transition which occurs at a temperature T α and (ii) the viscoelastic liquid/solid relaxation called crossover transition at a temperature T χ . The α transition corresponds to the Brownian/non Brownian diffusion of the asphaltene particles [4] and is determined by the deviation from the Arrhenius behavior such as T α ≈ 100 °C. The cross-over point can be associated here to the formation of a network of aggregates composed of asphaltene particles [4] and its value can be evaluated by the crossover point G’ = G” or δ = 45° with δ, the phase angle, such as T χ ≈ 18 °C. Lower temperature glass solid domain (Region IV not discussed here) defined by G’ ~ 109 Pa associated to β transition or glass transition (maximum of G”) is not appeared. This β transition corresponds to maltene glass state [4]. Effect of the gelator additive. The effect of the gelling additive on the liquid to solid transition of the bitumen is shown in Fig. 2. At high temperature, the presence of the gelling agent has no impact on the G’ and G” values. Two mechanical relaxations also appear with decreasing temperature but the transition temperatures differ from those of the pure bitumen. For the bitumen modified by the gelator A, the Newtonian/viscoelastic relaxation occurs at T α’ ≈ 110 °C, instead of 100 °C for the pure bitumen (Fig. 2a). This value is slightly higher than T α , which assumes a contribution of the gelling additive to the Brownian/non Brownian transition. The second relaxation at lower temperature can be associated to a sol-gel transition due to the formation of a percolating network of micrometer-long fibers of gelling agent. The sol-gel transition temperature, determined by the crossover point G’ = G”, occurs at T sol-gel ≈ 88 °C. Below T sol-gel , G’ and G” increase progressively with a remarkable parallel trend characteristic to gel state. The addition of the gelator A leads to a shift
1630
C. Bottelin et al.
of the liquid/solid transition of the bitumen to higher temperature promoting a larger solid behavior. The experimental data collected at temperatures lower than 50 °C are not relevant due to the bad quality of the signals. For the bitumen modified by the gelator B, the α’ and sol-gel transitions occur respectively at T α’ ≈ 102 °C and at T sol-gel ≈ 70 °C. It must be emphasized that there is no crossover point in that case and that T sol-gel is evaluated by the inflexion point of the phase angle. Nonetheless, as observed above, G’ and G” change with a parallel trend exhibiting closely a gel state area. The rheological measurements highlight the efficiency of the gelling additive on the liquid/solid transition of the bitumen. They increase the range of temperature where bitumen is solid, up to 90 °C, without increasing the bitumen viscosity at typical processing temperatures (140–160 °C).
3.2 Thermal Hysteresis Once the modified bitumen reaches the “gel” state, the temperature ramp is reverted and the sample is heated up to 60 °C. The typical example given in Fig. 3 shows a marked thermal hysteresis between the cooling and heating processes. This phenomenon has already been observed for usual liquids modified by the same type of gelator and is attributed to the fact that the network formation is more difficult to achieve than its breaking [2].
3.3 Mechanical Properties Mechanical spectra. The influence of the gelator additive on the viscoelastic properties of the bitumen are clearly illustrated in Fig. 4. The pure bitumen behaves as a viscoelastic liquid at 60 °C, with a viscous modulus clearly higher than the elastic one, whereas the modified bitumen is a viscoelastic solid (gelator A) or is approaching this state (gelator B). The rising of the moduli at low frequency is due to the evolution of the strength of the network during the test. We can also underline the improvement of the mechanical properties with the additives. The elastic moduli of the additived bitumen A and B are respectively five or four orders of magnitude greater than the elastic modulus of the pure bitumen at 0.1 rad s−1 . Critical stress and critical deformation. The G’ versus σ 0 or γ 0 profiles given in Fig. 5 determine the critical stress (σ C ) and the critical strain (γ C ) defined as the mechanical threshold between linear and nonlinear regimes. Fig. 5b shows that the linear viscoelastic domain, defined as the region where the moduli do not depend on the strain amplitude, is restricted to small deformations for the system bitumen/gelator A ( γ C ≈ 0.008). For this system, σ C ≈ 1500 Pa. The drop of the
Rheometrics Framework Analysis to Capture Self-assemblies …
1631
Fig. 2 Storage (G’) modulus, loss modulus (G”) and phase angle (δ) versus temperature (T ) for the bitumen additived by the gelator A (a) and the gelator B (b), at a fixed frequency of 10 rad s−1 and a strain of 10–3 . The dashed line represents the fit of Arrhenius’s law to the experimental points. The straight lines materialize the limits of the three regions. The α’ transition temperature is defined by the deviation from the Arrhenius behavior and the sol-gel transition temperature is defined by the crossover point G = G”.
moduli is not observable for the two other systems, so γ C > 0.04 for the pure bitumen and γ C > 0.02 for the additived bitumen B and σ C > 3000 Pa.
1632
C. Bottelin et al.
Fig. 3 Storage (G’) modulus, loss modulus (G”) and phase angle (δ) versus temperature (T ) for the bitumen additived by the gelator B, for the cooling and heating processes, at a fixed frequency of 10 rad s−1 and a strain of 10–3 .
Fig. 4 Storage (G’) modulus (full symbols) and loss modulus (G”) (open symbols) versus angular frequency (ω) for pure bitumen (black squares), bitumen additived by the gelator A (pink triangles) and by the gelator B (green circles), at a fixed strain of 10–3 , at 60 °C.
4 Conclusions A bitumen, additived or not, may solidify when its temperature is lowered. The temperature of the liquid/solid transition as well as the temperature of the Newtonian/viscoelastic transition can be rigorously determined using rheological experiments. Our study has also demonstrated that the gelator additives are very efficient to increase the range of temperatures where the bitumen behaves as a solid and to
Rheometrics Framework Analysis to Capture Self-assemblies … Stress sweep test - T = 60 °C 1E+06
1E+05
G' or G" (Pa)
G' or G" (Pa)
1E+06
1E+04 1E+03 1E+02 1
(a)
10
100
σ0 (Pa)
1000
10000
1633
Stress sweep test - T = 60 °C
1E+05 1E+04 1E+03 1E+02 1E-06 1E-05 1E-04 1E-03 1E-02 1E-01 1E+00
(b)
γ0 (-)
Fig. 5 a Storage (G’) modulus (full symbols) and loss modulus (G”) (open symbols) versus shear stress amplitude (σ 0 ) for the pure bitumen (black squares), the bitumen additived by the gelator A (pink triangles) and by the gelator B (green circles), at a fixed angular frequency of 10 rad s−1 , at 60 °C. b Same data represented as a function of the strain amplitude (γ 0 ).
improve its mechanical properties at 60 °C, without increasing the viscosity at high temperature. Therefore, rheology is more sensitive and provides more information than the classical methods of characterization.
References 1. Terech, P., Weiss, R.G.: Low molecular mass gelators of organic liquids and the properties of their gels. Chem. Rev. 97(8), 3133–3160 (1997) 2. Collin, D., Covis, R., Allix, F., Jamart-Grégoire, B., Martinoty, P.: Jamming transition in solutions containing organogelator molecules of amino-acid type: Rheological and calorimetry experiments. Soft Matter 9(10), 2947–2958 (2013) 3. Isare, B., Petit, L., Bugnet, E., Vincent, R., Lapalu, L., Sautet, P., Bouteiller, L.: The weak help the strong: Low-molar-mass organogelators harden bitumen. Langmuir 25(15), 8400–8403 (2009) 4. Lesueur, D.: The colloidal structure of bitumen: consequences on the rheology and on the mechanisms of bitumen modification. Adv. Coll. Interface. Sci. 145(1–2), 42–82 (2009)
Roles of Exogenous Cationic and Endogenous Bitumen Surfactants in the Stability of Bitumen/Water Interfaces Shweta Jatav, Patrick Bouriat, Pauline Anaclet, Yvong Hung, Francis Rondelez, and Christophe Dicharry Abstract Bitumen-in-water emulsions are often used for paving applications but some of them lack long-term stability. The relative importance of the cationic surfactant added to the aqueous phase to favor emulsification and the endogenous ones contained in bitumen is investigated by measuring the visco-elastic properties of three different bitumen-water interfaces as a function of time. We observe that the long-term emulsion stability correlates well with high values of the interfacial storage (G ) and loss (G ) moduli. Since these parameters become significantly weaker when an emulsamine cationic surfactant is added to water, we emphasize that the use of exogenous surfactants may be detrimental to long-term emulsion stability from interfacial behavior point of view. Keywords Water/bitumen interface · Interfacial shear rheology · Cationic surfactant · Emulsion stability
1 Introduction Bitumen oil-in-water emulsions are widely used but not all bitumen can form an emulsion stable enough for road paving and sealing applications. It is therefore important to determine which are the surfactant molecular species responsible for behavior and the long-term stability of the bitumen-water interface. Various techniques for studying the interfacial properties of bitumen/water systems have been reported in literature. Some of the techniques are tensiometry [1–3], micropipette technique [4, 5], dilatational rheology [2, 3, 6], interfacial shear rheology [2, 7], atomic force microscopy[3, 8], thin liquid film technique [1], bottle tests [7], and chemical characterization [9]. In particular, the interfacial S. Jatav · P. Bouriat (B) · F. Rondelez · C. Dicharry Laboratoire des Fluides Complexes et Leurs Réservoirs, Université de Pau et des Pays de l’Adour, E2S UPPA, CNRS, TOTAL, UMR 5150, Pau, France e-mail: [email protected] P. Anaclet · Y. Hung · F. Rondelez Bitumen Research Team, TOTAL Marketing and Services, Centre de Recherche de Solaize, Solaize, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_207
1635
1636
S. Jatav et al.
shear rheology technique (ISR) has made it possible to demonstrate that, slowly, the adsorbed subfractions of asphaltenes constitute a network at the interface and responsible for the long-term stability of the oil-water interface [10], a phenomenon known as aging. In this work, aging of three different bitumen-water interfaces (bitumen noted A, B, and C and displaying different emulsion abilities) have been investigated by ISR, with and without a cationic surfactant added to the aqueous phase. We will show that the long-term stability is determined by the nature of the endogenous surfactants to form specific interfacial film behavior and that the cationic surfactant tends to damp the ability of endogenous surfactants to aggregate and modify oil-water interfaces.
2 Materials and Experimental Methods 2.1 Materials The bitumen samples A, B, and C, emulsamine L60 (cationic surfactant, referred to as EL60 in the following), 32% concentrated HCl, and deionized water are provided by TOTAL MS, Lyon. The solvents toluene and n-heptane of purity ≥99.5%, are purchased from Fisher Scientific. In order to reduce viscosity, the bitumen samples used in the experiments are diluted in heptol 10 (50/50 vol% mixture of n-heptane and toluene) as model systems to mimic bitumen matrix. The stability trend of bitumen emulsions (bitumen content: 60 and 69 wt%) manufactured from each bitumen A, B, and C is stated by TOTAL MS according to EN13808 standard. Bitumen A forms more stable emulsions than bitumen C, bitumen B being intermediate. The diluted bitumen-in-water emulsions, prepared in 50 ml centrifugal tubes using Ultra Turrax homogenizer (at 9500 rpm for 5 min) with bitumen concentration 10 g/L in heptol 10, follow the same emulsion stability trend. Dilutions are prepared by sonicating the mixture for 30 min in a sealed glass bottle. For ISR study, two water phases are used: one containing 0.36 mg EL60/g of water and at pH3, and the other without EL60 and at pH2.
2.2 Experiments The ISR experiments are conducted on a MCR302 rheometer of Anton Paar. The details of the method can be found in reference [10]. The interfacial rheology system consists in a bicone of 34.12 mm of radius having a very sharp edge of 5° rotating in a glass cup of 40 mm radius. The system temperature is controlled by a Peltier plate and is maintained at 20 °C for all experiments. To conduct the ISR experiment, the glass cell is first filled with a water phase up to a marked height. After that, the bicone edge is positioned at the air/water interface using an interface gap sensing
Roles of Exogenous Cationic and Endogenous Bitumen …
1637
profile in RheoCompass software, which is operated by normal force transducer of the rheometer. The oil phase is then added carefully in the glass cell up to a marker height. Finally, the glass cell is covered with a lid that contacts the oil phase. This ensures a non-slipping plane necessary for data calculation and also prevents or limits oil evaporation. At this point, the heights of the water and oil phases, density of oil and water phases are entered in the experimental profile of the software for data analysis. The physical aging of the diluted bitumen/water interface is observed by applying a small amplitude of oscillatory strain 0.3% at 0.05 Hz for 400 min. Experiments with different amplitudes have also been performed to verify that this amplitude was small enough to be in the linear regime.
3 Results and Discussion The results of the ISR experiments are shown in Fig. 1, where the rheological properties, G and G are plotted as a function of the aging time (t) for the bitumen/water interface, with and without EL60 surfactant in the water phase. The three graphs correspond to diluted bitumen A (Fig. 1a), B (Fig. 1b), and C (Fig. 1c). The G and G values for the heptol 10 (without bitumen)/water interface are also plotted in presence of EL60 to provide a baseline for comparison. Emulsamine alone. In that case, the G and G values initially decrease with time, either due to residual impurities or transitional rearrangement of the interface, and then remain constant and significantly low. This shows that the EL60 molecules form a fluid interface and do not build a 2d-network. Bitumen alone. Without EL60, on the contrary, the G and G values of the bitumen/water interfaces markedly increase with time (2–3 orders of magnitude, see Fig. 1a–c), after a short lag time depending on the bitumen (see Table 1). This behavior suggests that the surface-active molecules (endogenous surfactant) of the diluted bitumen gradually adsorb at the interface and self-assemble to form a viscoelastic interface that exhibits film with substantial mechanical behavior. As best seen in Fig. 1a, b, G increases at a faster rate than G and becomes greater than G , after 41 and 63 min respectively. This crossover marks the transition from a liquid-like to a solid-like network. Bitumen + Emulsamine. The mechanical strength of the interfaces for all three diluted bitumen are significantly weaker in the presence of EL60, with G and G values lower by at least 2 orders of magnitude (Fig. 1a, b), after a longer lag time. We surmise that the successive dips, observed when G and G values become significant, are due to the breaking of the weak viscoelastic interface under the mechanical constraint, followed by a spatial rearrangement. For Bitumen C (Fig. 1c) the G and G values cannot be distinguished from those of the pure heptol 10/water with EL60. In order to compare more finely the bitumen behaviors and distinguish the influence of the added EL60 surfactant with those of the endogenous bitumen, several
1638
S. Jatav et al.
specific data can be extracted from Fig. 1a–c: (1) the crossover-time when G = G, (2) the G and G values at this time, (3) the G and G values at 300 min, (4) the G lag time. These data are reported in Table 1. The diluted bitumen A/water interface has the lowest crossover-time, followed by bitumen B then bitumen C. The G (= G ) values at the crossover obey the same decreasing order for the three bitumen without EL60 but they are identical within the experimental accuracy with EL60 systems. The G and G values at 300 min are the highest for diluted bitumen A and the lowest for diluted bitumen C, both with and without EL60 respectively. Finally, the G and G values at 300 min are considerably larger in the absence of EL60. All these observations support that the endogenous surfactant molecules form a viscoelastic interface, provided they are given enough time to reach the interface by Brownian diffusion throughout the bitumen bulk phase. Therefore, in order to adsorb at the interface, the endogenous bitumen surfactants will have to displace, partially or totally, the cationic molecules already present. The adsorption process is thus slowed down when EL60 is present. This competition process leads to an 3–5 times increase in lag time (see Table 1) for bitumen A and B, respectively. For bitumen C, it is so long that it exceeds the experimental time. Above the lag time, G and G start increasing significantly above the base line. However, the maximum values reached after 300 min remain 40–70 times less than the comparable data measured in the absence of EL60. Since our results show that EL60 molecules do not contribute to G and G , one can conclude that the interface is far from saturated in endogenous surfactant molecules. Even after such a long time, it is still a complex mixture of both types of surfactant molecules, most probably with islands of aggregated bitumen surfactants coexisting with individual EL60 molecules. The mechanical strength is provided by the endogenous surfactants alone. Those of bitumen A, are the most able to form a strong viscoelastic interface, followed by bitumen B and C. Such differences in behavior can readily be explained if one admits that these bitumen possess different endogenous surfactants, either in concentration or in composition. Interestingly, this classification matches the stability of the bitumen/water emulsions given by TOTAL MS.
4 Conclusion The study of the shear rheology of aging diluted bitumen/water interfaces, with and without a cationic surfactant added to the water phase, proves that the surfaceactive molecules present in the bitumen are the essential contributors to the emulsion stability. In the current case study, the endogenous surfactants of bitumen A are more efficient for long-term stability than those of bitumen B and C. This effect can be quantified by measuring the G’ and G” viscoelastic moduli: high values correlate directly with high emulsion stability, and vice versa. Finally, the introduction of a cationic surfactant in the aqueous phase is useful at short times to form the emulsion
Roles of Exogenous Cationic and Endogenous Bitumen … Fig. 1 The storage (G ) and loss modulus (G ) evolution with aging time at 0.03% of oscillatory strain and 0.05 Hz frequency for 10 g/L a A, b B, and c C in heptol 10 and water interface without EL60 and with EL60 surfactant (0.36 mg of surfactant per gram of water)
1639
1640
S. Jatav et al.
Table 1 Some characteristic values associated with the viscoelastic properties of the diluted bitumen/water interface Bitumen/water G = G uncertainties ~0,02 mPa m
Crossover time (min)
G = G at crossover (mPa m)
G and G at ~ 300 min (mPa m) G
G
G lag time (min)
A/without EL60
41
8.6
67
28
8
B/without EL60
63
5.6
24
11
12
C/without EL60
130
3.5
7.7
5.8
40
A/with EL60
42
0.05
1.5
0.4
~38
B/with EL60
~53
0.06
0.5
0.3
~30
C/with EL60
~300
0.06
0.06
0.04
~350
but hinders the adsorption of the endogenous bitumen surfactants onto the bitumenwater interface at long times. The choice of the exogenous surfactant for a given bitumen is therefore critical and the present methodology should help in this selection.
References 1. Tchoukov, P., Yang, F., Xu, Z., Dabros, T., Czarnecki, J., Sjöblom, J.: Role of asphaltenes in stabilizing thin liquid emulsion films. Langmuir 30, 3024–3033 (2014) 2. Miao, L., Li, F., Sun, D., Wu, T., Li, Y.: Interfacial and electrokinetic properties of asphaltenes and alkali/surfactant/polymer in produced water system. J. Petrol. Sci. Eng. 133, 18–28 (2015) 3. Rocha, J.A., Baydak, E.N., Yarranton, H.W., Sztukowski, D.M., Ali-Marcano, V., Gong, L., Shi, C., Zeng, H.: Role of aqueous phase chemistry, interfacial film properties, and surface coverage in stabilizing water-in-bitumen emulsions. Energy Fuels 30, 5240–5252 (2016) 4. Tsamantakis, C., Masliyah, J., Yeung, A., Gentzis, T.: Investigation of the interfacial properties of water-in-diluted-bitumen emulsions using micropipette techniques. J. Colloid Interface Sci. 284, 176–183 (2005) 5. Esmaeili, P., Lin, F., Yeung, A.: Stability of emulsified heavy oil: the combined effects of deterministic DLVO forces and random surface charges. Langmuir 28, 4948–4954 (2012) 6. Neuville, M., Rondelez, F., Cagna, A., Sanchez, M.: Two-step adsorption of endogenous asphaltenic surfactants at the bitumen–water interface. Energy Fuels 26, 7236–7242 (2012) 7. Pensini, E., Tchoukov, P., Yang, F., Xu, Z.: Effect of humic acids on bitumen films at the oilwater interface and on emulsion stability: potential implications for groundwater remediation. Colloids Surf. A 544, 53–59 (2018) 8. Kuznicki, N.P., Harbottle, D., Masliyah, J.H., Xu, Z.: Probing mechanical properties of watercrude oil interfaces and colloidal interactions of petroleum emulsions using atomic force microscopy. Energy Fuels 31, 3445–3453 (2017) 9. Lalli, P.M., Jarvis, J.M., Marshall, A.G., Rodgers, R.P.: Functional isomers in petroleum emulsion interfacial material revealed by ion mobility mass spectrometry and collision-induced dissociation. Energy Fuels 31, 311–318 (2017) 10. Ligiero, L.M., Dicharry, C., Passade-Boupat, N., Bouyssiere, B., Lalli, P.M., Rodgers, R.P., Barrere-Mangote, C., Giusti, P., Bouriat, P.: Characterization of crude oil interfacial material isolated by the wet silica method. Part 2: Dilatational and shear interfacial properties. Energy Fuels 31, 1072–1081 (2017)
Rubber-Bitumen Interaction of Plant-Blended Rubberized Bitumen Prepared Under Various Blending Conditions Xiong Xu, Zhen Leng, Jingting Lan, Rui Li, Zhifei Tan, and Anand Sreeram
Abstract Rubberized bitumen produced from waste tyre rubber has been widely used as a sustainable paving material around the world. However, different production conditions may lead to rubberized bitumens with different performances. This study aims to understand the mechanism of physicochemical interaction between rubber particles and virgin bitumen of plant-blended rubberized bitumen prepared under various blending conditions. To achieve this objective, samples of plant-blended rubberized bitumen prepared under four conditions were first collected, including blending at 170 °C for 1 h, and blending at 170 °C for 1 h followed by blending at 183 °C for 1, 3, and 5 h. Then the microstructure, viscosity, storage stability, and rheological properties of these samples were characterized. Microstructural results indicate main crosslinking structures of rubber particles remain stable and undamaged with the occurrence of chemical reactions including oxidation, decarboxylation, and devulcanization during the whole production period. The storage stability results state the rubberized binder exhibits no obvious phase separation after storage at 163 °C for 48 h, and becomes more stable as the blending time prolongs. The rheological results indicate the rubber-bitumen interaction is mainly composed of early-stage absorption and swelling of rubber particles, continuous emission of volatile bitumen fractions, and late-stage partial degradation of fully swollen rubber particles. Keywords Rubberized bitumen · Plant production · Rubber-bitumen interaction · Microstructure · Rheological properties
X. Xu · Z. Leng (B) · J. Lan · R. Li · Z. Tan · A. Sreeram Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hong Kong S.A.R., China e-mail: [email protected] X. Xu School of Civil Engineering and Architecture, Wuhan Institute of Technology, Wuhan, PR China © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_208
1641
1642
X. Xu et al.
1 Introduction Scrap tyres, also known as end-of-life tyres, are extensively used as sustainable paving modifying materials, which are generally shredded to fine crumb rubber (CR) in application of asphalt industry [1]. Incorporation of CR into virgin asphalt pavement materials can make contributions on significant improvement of the performance properties, such as rutting resistance, fatigue resistance, as well as tyre-road noise [2]. However, optimum performances of rubberized bitumen can be reached depending upon the good compatibility/interaction between CR and virgin bitumen [3], which is closely related to the production conditions, such as blending time and temperature. Some publications have proved rubber-bitumen interactions at elevated temperatures in blending procedure can be developed to the stages [4, 5], including (I) rubber particles absorb the light fractions of asphalt and swell; and (II) the swollen CR get degraded into smaller pieces, thus during the mixing, the integrated binder exhibits a trend of first increase and then decrease in terms of the viscosity changes. Ragab et al. performed a study on investigating the rubber-bitumen interaction mechanism by varying blending conditions to control the swelling behaviors of CR in virgin bitumen [6], and concluded that the swollen size of CR will not be significantly diminished after 10 min-interaction at 200 °C, but the reduced swelling and large numbers of finer CR can be observed to be well dispersed in binder after 8 h-interaction at 200 °C. Despite that several studies have shown the interaction correlations between rubber and bitumen [7, 8], most of them are based on the test results of laboratory prepared samples, not dependent on the findings from analysis of the samples obatined from plant production. However, the chemical and physical changes of rubberized binders in industrial production may be considerably different from the indoor simulations. Therefore, it is of vital importance and practical significance to gain a better knowledge of how rubber-bitumen binders work in hot-mix plant production. On basis of this, it will be useful and instructive to provide a theoretical guidance in hot-mix production of high-quality rubberized binders. This study aims at analyzing the correlation between the rubber-bitumen interaction and plant-production time. The microscopic structure, viscosity, storage stability, as well as rheological properties of the rubberized bitumen binders were characterized for justifications.
2 Experimental 2.1 Source of Raw Materials CR (minus 30 mesh), grinded from waste tyre rubbers, were locally provided. Rubberized bitumens were hot mix produced in a local asphalt factory. The first sample was
Rubber-Bitumen Interaction of Plant-Blended …
1643
collected at 170 °C after 1 h, and the other three samples were obtained at 183 °C after extra 1, 3, and 5 h. To brief the descriptions in this paper, those four samples were orderly marked as CRMA0, CRMA1, CRMA3, and CRMA5, respectively.
2.2 Test Methodology 2.2.1
Fourier Transform Infrared Spectroscopy (FTIR)
Molecular structure of each binder was detected using Nicolet iS50 FTIR equipped with an attenuated total reflectance (ATR) detective system. The results were respectively collected after scan times of 32 within the wavenumber range of 4000–400 cm−1 .
2.2.2
Workability
The workability of each binder was tested using NDJ-1C Brookfield viscometer at 4 temperatures (namely 135, 150, 165 and 180 °C) according to ASTM D4402.
2.2.3
Storage Stability
Binder samples were poured into toothpaste tubes with 3.2 cm-diameter and 16 cmheight for the storage stability test at 163 °C for 48 h. The binders of top- and bottom-equals were taken for softening point tests according to ASTM D36, and their values of the softening point difference (SPD) were used to evaluate the storage stability.
2.2.4
Dynamic Shear Rheometer (DSR) Test
Rheological parameters (complex modulus and phase angle) of binders were tested at the fixed frequency of 10 rad/s and 64, 70, 76, 82, and 88 °C using MCR 702 DSR apparatus. The gap distance of 25 mm-diameter paralleled plates was set at 2 mm.
1644
X. Xu et al.
3 Results and Discussion 3.1 Changes in Molecular Structure of Rubberized Bitumen The changing molecular structures of rubberized binders subjected to various blending hours are displayed in Fig. 1. As seen that the oxidative groups, located at 1729 cm−1 (C=O) and 1030 cm−1 (S=O, C–O–C), are the main changes. Also, the amount of C=O groups exhibits an increase at the beginning of 4 h and then drops to a lower value within the next 2 h, while the groups at 1030 cm−1 firstly diminish and then a bit increase, which illustrates the changes in chemical structure of the rubberized binder can be separated into two major reaction stages, referring to the oxidation and decarboxylation. In addition, the bands located at 3622–3062 cm−1 and 1575, 1537, 1465 cm−1 , attributing to the vibrations of −OH and benzene skeleton, present no obvious changes, and only C–O–C groups at 1109 cm−1 appear a small increase. These findings indicate the rubberized binder suffers no obvious structural damages at minus 183 °C within 6 h. To achieve a better understanding of the two-stage reaction analyzed above, the characteristic group index of oxidative groups at 1729 cm−1 (C=O) and 1030 cm−1 (S=O, C–O–C) are calculated by Eqs. (1) and (2) to perform semiquantitative analysis [9], as no visible changes within the band area between 3000 and 2760 cm−1 can be observed. C=O
CRMA5
C-O-C
S=O C-O-C
CRMA3
CRMA1
CRMA0 Shoulder peak
1729 1575 1537 1465
-OH
4000
1109
3062
3622 3500
3000
2500
2000
1500 -1
Wavenumber, cm Fig. 1 FTIR spectra of various rubberized binders
1030 1000
Rubber-Bitumen Interaction of Plant-Blended …
1645
Fig. 2 Chemical changes in rubberized binder as the production time prolongs
I (1729) = I (1030) =
Band ar ea o f C = O centr ed ar ound 1729 cm−1 Band ar ea o f C − C at 3000 ∼ 2760 cm−1
(1)
Band ar ea o f S = O and C − O − C centr ed ar ound 1030 cm−1 Band ar ea o f C − C at 3000 ∼ 2760 cm−1 (2)
Therefore, the changing trends of rubberized binders in the characteristic group index are plotted in Fig. 2. As seen, it indicates that some of carboxyl-based C=O groups formed at earlier reaction stage (before nearly 4 h) will subsequently occur to have thermal pyrolysis reactions, causing decarboxylation to emit CO2 , while C–O– C groups will be formed on molecular structures of pyrolysis products. In addition to those reactions, the S=O groups from virgin bitumen will be thermally converted into SO2 emissions as well [10].
3.2 Workability of Rubberized Bitumen Figure 3 reveals the workability of the rubberized binders. The viscosity of the binder increases in large after mixing for first 2 h and drops somewhat in next 4 h mixing, which indicates that rubber particles display a limited swelling behavior at an early stage in production, causing direct viscoelastic responses without destruction of their main chain structures, while the volatilization of light fractions and agglomeration of oxidative reaction products will also bring about an increase of viscosity. With the ongoing production, the decarboxylation and desulphurization are the main reactions to release numbers of CO2 and SO2 , and the hydrogen bond and molecular weight of −COOH related molecules drops to some extent, causing the decreasing viscosity, which can be well justified in accordance with FTIR analysis. In addition, the viscosity values of those binders at 180 °C appear to be low levels less than
1646
X. Xu et al.
Fig. 3 Viscosity-temperature curves of rubberized binders.
2.0 Pa s, which demonstrates that the binders can own a good workability while mixing with aggregates.
3.3 Storage Stability of Rubberized Bitumen Figure 4 illustrates the storage stability of the rubberized binders. As blending time increases, the distance between both softening point plots of top- and bottom- binder becomes close, and the SPD values of the binders are all below 4.0 ºC which is considered as the temperature limit to determine the high-temperature stability of modified bitumen [11]. This indicates that the rubberized binder exhibits no significant phase separation at 163 ºC for 48 h-storage and gets more stable while blending as the time prolongs. This is because an increase of the blending time promotes the continuous interaction between rubber particles and virgin bitumen, leading to a full Fig. 4 Storage stability of rubberized binders in various production stages
Rubber-Bitumen Interaction of Plant-Blended …
1647
Fig. 5 Complex modulus (a) and phase angle (b) of rubberized binders
swelling and a partial dissolution of rubber particles in virgin bitumen, to reduce the phase separation.
3.4 Rheological Properties of Rubberized Bitumen at High Temperatures Figure 5 displays the rheological properties of the rubberized binders. The complex moduli of the rubberized binders show a slight increase as the blending time increases to 4 h, but the phase angles decrease dramatically. When the blending time continues to 6 h, the complex moduli and phase angles both increase a little. This states that the improved deformation resistance of rubberized binder is dependent on the early-stage self-elasticity of rubber particles, continuous hardening effects of virgin bitumen, as well as the late-stage elastic modifications of virgin bitumen by degraded rubbers during the whole production. The reasons can be basically explained following the two divisions of stage changes: (a) the volatilization of light fractions in virgin bitumen and the full swelling of rubber particles without degradation products; and (b) the small amounts of degraded rubbers dispersing into virgin bitumen.
4 Conclusions This study aims to investigate the rubber-bitumen interaction of rubberized bitumens produced in a plant at high temperatures for 6 h. The results-based conclusions are summarized as follows. (1)
Based on the FTIR analysis, the main chain structures of rubber particles in binder will be hardly broken and destructed during a long period of production, while the early formation of C=O and the late emissions of CO2 and SO2 will happen due to the oxidation, decarboxylation, and devulcanization.
1648
(2)
(3)
(4)
X. Xu et al.
In terms of viscosity, the rubberized binder shows no entire declining trend as expected but displays a change of early-stage increase and late-stage decrease during the 6-h production. The storage stability results indicate that plant-blended rubberized binders can be well stored without an obvious phase separation at high temperatures, while as the mixing time prolongs, the storage stability will be also improved in subsequence. According to the DSR results, the rubber-bitumen interaction exhibits an extremely complex physicochemical change during production at high temperatures for 6 h, which contains the early-stage absorption and swelling of rubber particles, consistent emissions of easy-volatilized bituminous fractions, and the late-stage degradation of fully swollen rubber particles in very small ratios.
References 1. Zanetti, M.C., Fiore, S., Ruffino, B., Santagata, E., Dalmazzo, D., Lanotte, M.: Characterization of crumb rubber from end-of-life tyres for paving applications. Waste Manage. 45, 161–170 (2015) 2. Yu, H., Leng, Z., Xiao, F., Gao, Z.: Rheological and chemical characteristics of rubberized binders with non-foaming warm mix additives. Constr. Build. Mater. 111, 671–678 (2016) 3. López-Moro, F.J., Moro, M.C., Hernández-Olivares, F., Witoszek-Schultz, B., AlonsoFernández, M.: Microscopic analysis of the interaction between crumb rubber and bitumen in asphalt mixtures using the dry process. Constr. Build. Mater. 48, 691–699 (2013) 4. Lo Presti, D.: Recycled tyre rubber modified bitumens for road asphalt mixtures: a literature review. Constr. Build. Mater. 49, 863–881 (2013) 5. Wang, S., Wang, Q., Wu, X., Zhang, Y.: Asphalt modified by thermoplastic elastomer based on recycled rubber. Constr. Build. Mater. 93, 678–684 (2015) 6. Ragab, M., Abdelrahman, M.: Enhancing the crumb rubber modified asphalt’s storage stability through the control of its internal network structure. Int. J. Pavement Res. Technol. 11(1), 13–27 (2018) 7. Sienkiewicz, M., Borz˛edowska-Labuda, K., Zalewski, S., Janik, H.: The effect of tyre rubber grinding method on the rubber-asphalt binder properties. Constr. Build. Mater. 154, 144–154 (2017) 8. Yu, H., Leng, Z., Zhou, Z., Shih, K., Xiao, F., Gao, Z.: Optimization of preparation procedure of liquid warm mix additive modified asphalt rubber. J. Clean. Prod. 141, 336–345 (2017) 9. Xu, S., Yu, J., Wu, W., Xue, L., Sun, Y.: Synthesis and characterization of layered double hydroxides intercalated by UV absorbents and their application in improving UV aging resistance of bitumen. Appl. Clay Sci. 114, 112–119 (2015) 10. Rubio, M.C., Moreno, F., Martínez-Echevarría, M.J., Martínez, G., Vázquez, J.M.: Comparative analysis of emissions from the manufacture and use of hot and half-warm mix asphalt. J. Clean. Prod. 41, 1–6 (2013) 11. Leng, Z., Padhan, R.K., Sreeram, A.: Production of a sustainable paving material through chemical recycling of waste pet into crumb rubber modified asphalt. J. Clean. Prod. 180, 682–688 (2018)
Rutting Performance Analysis for Pavements with Bituminous Stabilized Mixtures as Base Layers Beatriz Chagas Silva Gouveia , Francesco Preti , Elena Romeo , Eshan V. Dave , Gabriele Tebaldi , and Jo E. Sias
Abstract Cold recycling techniques are becoming more and more popular everywhere in the world, as it is well proven that bituminous stabilized mixtures (BSM) are sustainable both from an environmental and economical point of view. There have been limited studies focusing on the performance of BSMs with respect to the service lives of pavement structures with in-place recycled layers. For this reason, this research focuses on the modelling of multilayer structures with different layer combinations and thicknesses. The aim is to compare structures with different base layer solutions, such as BSM, typical granular unbound base layer, and combinations of those two. Finite element analysis software has been used to construct axisymmetric elasto-plastic 2D models of various pavement cross-sections. The modelled structures were subjected to a cyclic loading of 0.1 s with 0.9 s of rest period. The load applied on the structures has been calibrated in order to have a comparable effect to the real tire pressure experienced by the structures under traffic. Rutting evolution curves until a failure threshold value of 20 mm have been developed and the results have clearly shown how the BSM as a base layer can provide superior or comparable performance in terms of rutting as compared to granular virgin material. Keywords Cold recycling · Modeling · Finite elements · Bitumen stabilized materials · BSM
1 Introduction Considering that cold recycling of distressed flexible pavements is growing in importance due to urgent global environmental issues, it is natural that several researchers have been dedicated to this topic on the past recent years [1–3]. The benefits of cold recycling techniques are not only environmental, but also technical and economic, as B. C. S. Gouveia · F. Preti (B) · E. Romeo · G. Tebaldi Università di Parma, Parma, Italy e-mail: [email protected] F. Preti · E. V. Dave · J. E. Sias University of New Hampshire, Durham, NH, USA © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_209
1649
1650
B. C. S. Gouveia et al.
it provides a reduction of energy and natural material consumption, reuse of high-cost material, and reduction of CO2 emissions. Where an existing pavement is recycled, old seals or asphalt surfacing is usually mixed with the underlying layer and treated to form a new base or subbase layer. This new layer may be constructed using Bitumen Stabilized Materials (BSM), which are composed of a granular material (such as virgin aggregates, recycled cement treated materials, and/or up to 100% reclaimed asphalt), treated with either bitumen emulsion or foamed bitumen [4]. The objective of this study is to develop and use an axisymmetric elastoplastic 2D finite element model for a pavement structure with BSM materials to investigate its overall performance on rutting. Going even further, this research analyses the long term life cycle of BSM, as these materials are expected to enhance the properties of recycled pavement underlayers, providing equal or better service lives, at lower costs, when compared to those achievable with traditional granular base layer constructed with virgin materials. The assessment of service life was done by comparing eight different alternatives for a reconstruction project, one reference and other seven possible reconstruction or recycling solutions. These structural solutions were analyzed regarding their overall performance on rutting, through the proposed FEM axisymmetric elastoplastic 2D model, as described in the following sections.
2 Materials and Methods 2.1 Constitutive Models for BSM Bitumen Stabilized Materials can be defined as the material produced of granular or RAP recycled materials, treated at ambient temperature with foamed bitumen or emulsion binders. Previous studies [5] alert that BSM are known to have limited tensile strength since the aggregate is only partially coated by the binder (especially new virgin binder), as the new virgin binder content usually varies from 1 to 3% of total weight of mixture. As a result, the shear strength of materials based on aggregate interlock is the material property that is most closely linked to performance, a behavior that is also observed for granular materials. Granular materials are known for carrying limited tensile stress and for undergoing plastic deformations under higher compressive stresses. Thus, elasto-plastic modelling may be a useful way to simulate the behavior of BSM materials. Since the Mohr-Coulomb yield criterion can simulate plastic response, Triaxial Compression tests was conducted on the BSM samples to determine two primary material properties, namely cohesion and friction angle. The results of the Triaxial Compression test, proceeded as prescribed on the South African Technical Guideline for the design and construction of BSM [4], were used as primary inputs for the BSM materials.
Rutting Performance Analysis for Pavements …
1651
The extraction of local material properties from triaxial laboratory tests was conducted using an inverse analysis. Whereby, the global measurements from lab tests, namely the overall force on specimen and the specimen deformation, were used as the inverse analysis parameters. In other words, the constitutive properties of materials were found such that, the global response of a simulated lab test in finite elements matched the lab measured global response. The inverse analysis was conducted to first match the force-displacement curves of the triaxial test with no confining pressure. The constitutive model and local properties from the no confinement condition was thereafter validated by conducting tests and simulations with 100 and 200 kPa confining pressure. Thus, ensuring that the model and material properties yield acceptable prediction of the BSM behavior under different conditions.
2.2 Reconstruction Structural Solutions As previously pointed out, eight different alternatives for reconstruction of a deteriorated pavement were proposed. Table 1 presents the layer thicknesses for each alternative that was used in structural analysis. All the input properties for the asphalt layers, granular base layer and soil were adopted from the existing literature [6]. The input values for the BSM layer were obtained from the constitutive model developed previously and calibrated using laboratory test results, as described in previous section. Material properties are presented in Table 2. Table 1 Thicknesses (cm) of layers on proposed structural solutions Layer
ML1
ML2
ML3
ML4
ML5
ML6
ML7
ML8
HMA wearing course
2.5
5
2.5
5
2.5
2.5
2.5
2.5
HMA binder layer
5
10
5
10
5
5
5
7.5
BSM
15
15
30
30
N.A
30
15
15
Granular base
15
15
15
15
30
N.A
N.A
15
Soil
Semi-infinite
Table 2 Mechanical properties of layers Layer
E (MPa)
Poisson’s ratio
Cohesion (kPa)
Friction angle (°)
HMA
2000
0.35
–
–
Binder
2500
0.35
–
–
BSM
270
0.40
450
31
Granular base
150
0.40
75
40
Soil
70
0.40
50
15
1652
B. C. S. Gouveia et al.
The elastic modulus (E) for all the base materials has been adjusted keeping in consideration their location within the multilayer structure. The E value is indeed affected by the confining pressure that is dependent on the depth where the material is placed in the multilayer structure. This is due to the stress hardening effect, the material under increased applied stress exhibits less deformation and therefore greater stiffness or resilient modulus [7].
2.3 Multilayer Pavement Model In order to understand the behaviour of the layered solutions, an axisymmetric elastoplastic 2D model was built on the software ABAQUS. After Burmister [8] established the analytical solution for layered linear elastic half-space model, as well as after the introduction of finite element methods, the use of axisymmetric 2-D approach has been a regular solution to analyse flexible pavement and several codes were created for pavement analysis and design [9]. The developed axisymmetric elastoplastic 2D model (Fig. 1) consisted on a 4 × 4 m mesh, where each layer, for each structural solution, had their thickness and mechanical properties as shown on Tables 1 and 2. This domain size was determined using the domain extent analysis, multiple trial models were simulated in order to make sure that the boundary conditions were not affecting the results. Tangential behaviour between layers was inputted as rough and normal behaviour between layers as hard contact with no separation allowed. The modelled structures were subjected to a cyclic loading of 80 kN (most commonly used value in pavement design in the U.S. for equivalent single axle load or ESAL), with duration of 0.1 s with 0.9 s of rest period. The final surface of application of the load was considered
Fig. 1 Multilayer axisymmetric elastoplastic 2D model
Rutting Performance Analysis for Pavements …
1653
as a circle with 10 cm of radius. The failure of the structure was considered as the number of load application where the plastic deformation reached 20 mm depth. The reason for using rutting as the only criteria to determine pavement life is due to rutting being the main distress affecting base layers. In those layers there are minimal to no concerns about cracking since these layers are usually constructed with unbounded or partially bounded materials.
3 Results The results of the analysis in terms of the number of ESALs needed to reach the terminal rutting level of 20 mm is presented in Fig. 2. The results clearly show that the rutting resistance of each structure is dependent on the asphalt thickness on top of the base layer. Doubling the thickness of the asphalt wearing course and binder layer (as in structures ML2 and ML4), makes their base courses reach more than the double the service life, as it takes more than the double ESALs to reach the 20 mm deformation. For purposes of comparison in between BSM recycling technique and traditional base layer is necessary to keep in consideration just the structures where those two are same thickness and have the same thickness of asphalt layers on top. In those cases, for example in the ML5 and ML6 simulation, it is possible to see that a structure with BSM layer and that with traditional base layer have comparable rutting performance. Even more interesting is the comparison in between the ML1 simulation (reference structure) that can be imagined as a structure obtained after rehabilitation of the structure ML5. Results show that the BSM replacing 15 cm of the granular base is
Fig. 2 Multilayer 2D simulation results comparison
1654
B. C. S. Gouveia et al.
able to result in the same or increased service life of the structure, proving to be an attractive alternative for pavement rehabilitation.
4 Conclusions This paper gives an insight on how can be possible to predict rutting phenomena in pavement structures basing on elasto-plastic properties of materials, the research study also demonstrates how BSM materials, when used as base layers, can guarantee excellent mechanical performance. The purpose of this study was not to assess that a BSM layer is better than a traditional base layer from a mechanical point of view, but to show that conducting pavement rehabilitation using cold in-place recycling technique is able to yield a comparable rutting performance as compared to a full reconstruction scenario with use of traditional base material, which uses only virgin materials. Cold recycling techniques allows to avoid the usage of virgin materials and minimize the need of transportation, therefore are fast type of operation, cheaper and sustainable for an environmental point of view. Multiple studies [10] have quantified these economic and environmental advantages. The findings of the analysis conducted in this paper provide mechanistic analysis results to support use of in-place recycled BSM and also present a framework that can be adopted for future pavement analysis. The future extension of this research study is to analyse the effect of temperature on mechanical properties of materials. The idea is to implement the model in order to keep in account the different behaviour of the pavement structure under multiple climate conditions. In addition, other multilayer models with different BSM types and different layer combinations are going to be simulated.
References 1. Dave, E.V.: Cold recycling of asphalt concrete: Review of North American state of practice. In: RILEM TC-SIB: TG-6, Liverpool (2011) 2. Tebaldi, G., et al.: Cold recycling of reclaimed asphalt pavements. In: Testing and Characterization of Sustainable Innovative Bituminous Materials and Systems, pp. 239–296. Springer, Cham (2018) 3. Betti, G., et al.: Active fillers’ effect on in situ performances of foam bitumen recycled mixtures. Road Mater. Pavement Des. 18(2), 281–296 (2017) 4. Asphalt Academy: Technical guideline: Bitumen stabilised materials. In: A Guide for the Design and Construction of Bitumen Emulsion and Foamed Bitumen Stabilised Materials (2009) 5. Jenkins, K.J., et al.: New laboratory testing procedures for mix design and classification of bitumen-stabilised materials. Road Mater. Pavement Des. 13(4), 618–641 (2012) 6. Siripun, K., Jitsangiam, P., Nikraz, H.: Characterization analysis and design of hydrated cement treated crushed rock base as a road base material in Western Australia. Int. J. Pavement Res. Technol. 2(6), 257–263 (2009)
Rutting Performance Analysis for Pavements …
1655
7. Buchanan, S.: Resilient modulus: what, why, and how. Vulcan Mater. Company 8(31), 07 (2007) 8. Burmister, D.M.: The general theory of stresses and displacements in layered soil systems. III. J. Appl. Phys. 16(5), 296–302 (1945) 9. Thompson, M.R., Elliott, R.P.: ILLI-PAVE based response algorithms for design of conventional flexible pavements. Transp. Res. Rec. 1043, 50–57 (1985) 10. Haslett, K.E., Dave, E.V., Mo, W.: Realistic traffic condition informed life cycle assessment: interstate 495 maintenance and rehabilitation case study. Sustainability 11(12), 3245 (2019)
Sensitivity Analysis for Rheological Determination of Glass Transition and Crossover Temperatures at Various Reference Frequencies Michael D. Elwardany, Jean-Pascal Planche, and Jeramie J. Adams
Abstract Accurate and relevant rheological characterization is critical for the improvement of current binder specification, modification and formulation strategies, and ultimately the pavement performance. Significant changes have affected the binder quality since the Superpave specification development due to various economic, technical, and environmental motives. This paper assesses the effect of some of these changes on binder rheology. The rheological behavior of binders was investigated by plotting the data in isochronal plots. In these plots, the rheological behavior at a reference frequency is classified into three regions: (1) near glassy region at temperatures below the glass transition temperature (Tg ), (2) terminal region at temperatures above the crossover temperature (Tx ), and (3) an intermediate “transition” region in between. The intermediate temperature range (TIR ) parameter is the difference between Tx and Tg . In this study, Tg and Tx of 30 binders were obtained using 4-mm parallel plates geometry on a Dynamic Shear Rheometer (DSR), and using isochronal plots calculated at various reference frequencies. Additionally, modulated differential scanning calorimetry (MDSC) was used as a widely accepted technique to determine Tg of the binders. In the paper, DSR-calculated Tg at various reference frequencies is compared with MDSC-measured Tg. Tg obtained at a reference frequency of 10 rad/s shows promising correlations with DSC-measured Tg and low-PG temperature based on stiffness (Tc (S)), while Tx at 10 rad/s correlates with low-PG temperature based on stiffness slope, m-value (Tc (m)). Furthermore, TIR correlates with Tc , an empirical parameter considered for future specifications in the USA related to cracking. The aforementioned correlations are very promising and provide insights related to the fundamental meaning of the Tc parameter and guidelines to formulate binders that meet future specifications. In-depth analysis is ongoing to evaluate how isochronal parameters (Tg , Tx , and TIR ) are affected by the reference frequency, and how they relate to binder composition. Keywords Glass transition temperature · Crossover temperature · Intermediate region temperature range (TIR ) · Isochronal plots · Sensitivity analysis M. D. Elwardany (B) · J.-P. Planche · J. J. Adams Western Research Institute, Asphalt and Petroleum Technologies, 3474 North 3rd Street, Laramie, WY 82072, USA e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_210
1657
1658
M. D. Elwardany et al.
1 Introduction Accurate and relevant rheological characterization is critical for the improvement of current binder specification, modification and formulation strategies, and ultimately the pavement performance. Significant changes have affected the binder quality since the Superpave specification development due to various economic, technical, and environmental motives [1–5]. Binder rheological behavior was investigated by plotting the data in isochronal plots using Prony series representation for viscoelastic materials as described in Elwardany et al. [4]. The rheological behavior, evaluated using a dynamic shear rheometer (DSR), is classified at a reference frequency into three regions: (1) near glassy region at temperatures below the glass transition temperature (Tg ), (2) terminal region at temperatures above the crossover temperature (Tx ), and (3) an intermediate “transition” region in between. The intermediate temperature range (TIR ) parameter is the difference between Tx and Tg , as shown in Fig. 1 [4]. In the paper, DSR-calculated Tg at various reference frequencies is compared with MDSC-measured Tg. Tg obtained at a reference frequency of 10 rad/s shows promising correlations with DSC-measured Tg and low-PG temperature based on stiffness (Tc (S)) [4], while Tx at 10 rad/s correlates with low-PG temperature based on stiffness slope, m-value (Tc (m)) [4]. Furthermore, TIR correlates with Tc [4], an empirical parameter considered for future specifications in the USA related to unmodified binder cracking resistance [6]. The aforementioned correlations, based on 10 rad/s reference frequency, are very promising and provide insights related to the fundamental meaning of the Tc parameter and guidelines to formulate binders that meet future specifications [4]. TIR was further proven as a useful parameter to evaluate unmodified binder cracking resistance based on an independent data set from NCHRP 09–60 project [5].
Fig. 1. Isochronal plots: storage and loss moduli versus reduced temperature at 10 rad/s
Sensitivity Analysis for Rheological Determination of Glass …
1659
In this study, Tg and Tx of 30 binders were obtained using isochronal plots calculated at various reference frequencies (1, 10, and 62.83 rad/s (10 Hz)). In-depth analysis is ongoing to evaluate how isochronal parameters (Tg , Tx , and TIR ) are affected by the selection of reference frequency, and how they relate to binder composition and performance.
2 Experimental Plan A comprehensive binder matrix, of 30 binders, was considered in this sensitivity analysis. Binder matrix includes strategic highway research program (SHRP) binders, as a reference, unmodified binders from various sources including reference binders from Asphalt research consortium (ARC) project, air blown binders and blends, re-refined engine oil bottoms (REOB)-modified binder, Polymer-Modified Asphalt (PMA) with styrene-butadiene-styrene (SBS), SBS and REOB, ethylene terpolymer (Elvaloy), ethylene vinyl acetate (EVA), EVA and Fisher Tropsch paraffin. In addition, special binders or refinery bases, such as multigrade binder, and a solvent deasphalted (SDA) residue with petroleum oil blend were also considered. More info on the Asphalt Industry Research Consortium (AIRC) binder matrix and findings are presented in Elwardany et al., Adams et al., Planche et al. [1–4]. Universal simple aging test (USAT) was used to condition the binders as an alternative to standard rolling thin-film oven (RTFO) and pressure aging vessel (PAV). USAT is a thin-film (300 µm) aging method for asphalt binders, developed by the Western Research Institute (WRI), to simulate short-term aging in 50 min at 150 °C and long-term aging in 8 h at 100 °C and 2.1 MPa pressure. These aforementioned conditions are considered equivalent to standard RTFO and PAV conditions [7]. Data presented in this paper is based on USAT-aged condition equivalent to standard RTFO + PAV20h aging. Binder glass transition parameters were evaluated using modulated differential scanning calorimetry (DSC) using sample weights ranging from 8 to 14 mg [3–5, 8]. Samples were equilibrated at 165 °C for five minutes and then cooled to −90 °C at −2 °C/min with ±0.5 °C modulation every 80 s. Samples were then equilibrated for five minutes before a heating cycle to 165 °C at 2 °C/min with same modulation conditions [8]. Rheological evaluation for asphalt binders were conducted using a dynamic shear rheometer (DSR), 4-mm parallel plates geometry, developed by WRI [9, 10], was used at 0 to −30°C temperature range and 8-mm parallel plates were used at 15– 45 °C temperatures in 15 °C intervals [1–5, 9, 10]. The frequency sweeps were performed over a range of frequencies between 0.1 to 100 rad/s. Mastercurves were constructed and fitted to Prony series, among other models, using Rhea software [4]. The advantage of using 4-mm DSR geometry is the use of a small sample size 25 mg as oppose to 15 g required to run the standard bending beam rheometer (BBR) for low-temperature performance grade (PG) determination [9, 10].
1660
M. D. Elwardany et al.
3 Results and Discussion Glass transition temperature is one of the important factors that contributes to the binder resistance to age-induced surface cracks under the internal restraint damage mechanism [5]. Glass transition temperature is related to binder molecular weight distribution, average molecular number in particular [11]. DSR-determined glass transition temperature (Tg ) was determined as the average between the temperature at the peak of the loss modulus isochronal plot and the temperature at storage temperature at 575 MPa [4]. As shown in Figs. 2 and 3, the correlation between DSRdetermined Tg and DSC-determined Tg is valid at various reference frequencies (1, 10, and 62.83 rad/s (10 Hz)). Glass transition temperature was also found to correlate with the critical cracking temperature based at stiffness = 300 MPa in the BBR [4]. Figure 4 shows that this correlation works well regardless of the chosen reference frequency. Crossover temperature was also found to correlate with the critical cracking temperature based at slope m = 0.3 in the BBR [4]. Figure 5 shows that this correlation works well regardless of the chosen reference frequency. Several researchers discussed the importance of the crossover temperature as an aging index parameter that is also sensitive to m-controlled binders (with negative Tc ) [12, 13]. Consequently, the intermediate region temperature range (TIR ) correlates to the Tc parameter obtained from the BBR test [4]. A Tc value of −5 °C is considered as critical in terms of field surface cracking - unmodified binders featuring Tc below −5 °C after long term aging are often prone to cracking. -30
-20
-10
0 -40
-30
-20
-10
-20
-10
0 -40
0
-30 y = 0.790x + 3.127 R² = 0.852
DSC- Tg(H) (°C)
y = 0.800x + 7.720 R² = 0.862
DSC- Tg(H) (°C)
DSC- Tg(H) (°C)
-30
-30
-20
-10
-20
-10 y = 0.780x - 0.647 R² = 0.841 0 -40
0
DSR- Tg (°C) at ωr = 10 rad/s
DSR- Tg (°C) at ωr = 1 rad/s
-30
-20
-10
0
DSR- Tg (°C) at ωr = 62.83 rad/s
Fig. 2. Correlations between the DSC-determined glass transition temperature at half-height of the reversing capacity curve Tg (H) and DSR-determined glass transition temperature
-20
-30
-20
-10
DSR- Tg (°C) at ωr = 1 rad/s
0
y = 0.982x - 2.179 R² = 0.792
DSC- Tg(I) (°C)
DSC- Tg(I) (°C)
DSC- Tg(I) (°C)
y = 0.984x + 3.231 R² = 0.785
-30
-10 -40
-40
-40
-40
-30
-20
-10 -40
-30
-20
-10
DSR- Tg (°C) at ωr = 10 rad/s
0
-30
-20 y = 0.978x - 6.714 R² = 0.796 -10 -40
-30
-20
-10
0
DSR- Tg (°C) at ωr = 62.83 rad/s
Fig. 3. Correlations between the DSC-determined glass transition temperature at the inflection point of the reversing capacity curve Tg (I) and DSR-determined glass transition temperature
Sensitivity Analysis for Rheological Determination of Glass … -40
-40
-10 -40
-30
-20
-10
-30
Tc(S) (°C)
Tc(S) (°C)
Tc(S) (°C)
-20
-20
-10 -40
0
-40
y = 0.970x - 5.960 R² = 0.899
y = 0.984x - 0.267 R² = 0.914 -30
1661
-30
-20
-10
-30
-20
-10 -40
0
DSR- Tg (°C) at ωr = 10 rad/s
DSR- Tg (°C) at ωr = 1 rad/s
y = 0.955x - 10.634 R² = 0.883 -30
-20
-10
0
DSR- Tg (°C) at ωr = 62.83 rad/s
Fig. 4. Correlations between the critical cracking temperature Tc (S = 300 MPa) and the DSRdetermined glass transition temperature (Tg ) -40
-40
-40
y = 0.640x - 43.477 R² = 0.752
-20
-30
Tc(m) (°C)
-30
Tc(m) (°C)
Tc(m) (°C)
y = 0.657x - 38.776 R² = 0.769
-20
-30
-20 y = 0.627x - 47.222 R² = 0.736
-10
0
20
40
60
-10
0
20
40
60
-10
0
DSR- Tx (°C) at ωr = 10 rad/s
DSR- Tx (°C) at ωr = 1 rad/s
20
40
60
DSR- Tx (°C) at ωr = 62.83 rad/s
Fig. 5. Correlations between the critical cracking temperature Tc (m = 0.3) and the DSRdetermined crossover temperature (Tx ). 5
5
0
0
0
-5
-5
-5
-10 -15 20
y = -0.423x + 17.979 R² = 0.918
40
-10
60
80
-15 20
5
y = -0.404x + 18.039 R² = 0.916
40
-10
60
80
-15 20
y = -0.389x + 18.104 R² = 0.914
40
60
80
Fig. 6. Correlations between the intermediate region temperature range (TIR ) and Tc parameter
Figure 6 shows that this correlation works well regardless of the chosen reference frequency. Based on Fig. 6, the recommended TIR threshold value of 50 °C at 10 rad/s reference frequency [4, 5] is equivalent to 47.6 °C at 1 rad/s reference frequency and 52 °C at 62.83 rad/s (10 Hz) reference frequency. This threshold corresponds to − 5 °C the critical Tc value, above mentioned.
4 Conclusion The presented sensitivity analysis is based on 30 binders after long-term aging. Isochronal plots were constructed using the Prony series model at various reference frequency (1, 10, and 62.83 rad/s (10 Hz). The main conclusions are:
1662
M. D. Elwardany et al.
• The 4-mm parallel plates on a DSR is a reliable method to obtain glass transition temperatures that correlate well with those obtained from DSC. • Results showed that the critical cracking temperature based on creep stiffness (Tc (S = 300 MPa)) correlates with the glass transition temperature of the binder. • Results also showed that the critical cracking temperature based on the relaxation rate (m-value) (Tc (m) = 0.3) correlates with the binder crossover temperature. • The intermediate region temperature range (TIR ) is a shape parameter that describes the shape of the storage and loss moduli isochronal plots. TIR strongly correlates to the Tc parameter (another rheological shape parameter). Based on the preliminary analysis, binders with TIR > 50°C at 10 rad/s reference frequency could be problematic in the field in terms of cracking-resistance performance due to their poor stress relaxation ability. • The recommended TIR threshold of 50 °C at 10 rad/s is equivalent to 47.6 °C at 1 rad/s reference frequency and 52 °C at 62.83 rad/s (10 Hz) reference frequency. • All the aforementioned observed were valid regardless of chosen reference frequency. Acknowledgments This work has been conducted under the AIRC project. The authors would like to acknowledge the members of the AIRC Eurovia, Eiffage, Husky, BRRC, IFSTTAR, TOTAL, Shell, Puma Bitumen, Federal Highway Administration, RoadMat, Repsol, Surfax, and WRI for their funding, sample supply, and technical inputs during the various phases of the project. Appreciation is also extended to Ms. P. J. Coles, Mr. J. Forney, and Mr. Z. Hunter, from WRI, for running rheological and thermal analysis.
References 1. Planche, J.-P., Elwardany, M.D., Adams, J.J., Boysen, R., Rovani, J.: Linking Binder Characteristics with Performance: The Recipe to Cope with Changes in Bitumen Binder Quality. XXVI PIARC World Road Congress, Abu Dhabi, United Arab Emirates (2019) 2. Planche, J.-P., Elwardany, M.D., Adams, J.J.: Chemo-mechanical characterization of bitumen binders with the same continuous PG-grade. In: RILEM 252-CMB Symposium. RILEM Bookseries, vol. 20, Springer, Cham (2019) 3. Adams, J.J., Elwardany, M.D., Planche, J.-P., Boysen, R., Rovani, J.F.: Diagnostic techniques for various asphalt refining and modification methods. Energy Fuels (2019). https://doi.org/10. 1021/acs.energyfuels.8b03738 4. Elwardany, M.D., Planche, J.-P., Adams, J.J.: Determination of binder glass transition and crossover tempreatures using 4-mm plates on a dynamic shear rheometer. Transp. Res. Rec. (2019). https://doi.org/10.1177/0361198119849571 5. Elwardany, M.D., King G., Planche, J.-P., Rodezno, C., Christensen, D., Fertig III, R.S., Kuhn, K.T., Bhuiyan, F.H.: Internal restraint damage mechanism for age-induced pavement surface distresses: black cracking and raveling. J. Asphalt Paving Technol. (2019) (Accepted/In Press) 6. Anderson, R.M., King, G.N., Hanson, D.I., Blankenship, P.B.: Evaluation of the relationship between asphalt binder properties and non-load related cracking. J. Assoc. Asphalt Paving Technol. 80, 615–649 (2011) 7. Federal Highway Adminstration (FHWA).: The Universal Simple Aging Test. FHWA TechBrief, Publication No. FHWA-HRT-15–054, US Department of Transportation, Federal Highway Adminstration. (2016)
Sensitivity Analysis for Rheological Determination of Glass …
1663
8. Turner, T.F., Branthaver, J.F.: DSC studies of asphalt and asphalt components. In: Usmani (ed.) Asphalt Science and Technology. Marcel Dekker Inc., NY (1997), pp. 59–101 9. Federal Highway Adminstration (FHWA).: Four-mm dynamic shear rheometry. In: FHWA TechBrief, Publication No. FHWA-HRT-15–053, US Department of Transportation, Federal Highway Adminstration. (2017) 10. Farrar, M., Sui, C., Salmans, S., Qin, Q.: Determining the low temperature rheological properties of asphalt binder using a dynamic shear rheometer. White Paper (2015) 11. Fox, T., Flory, P.: Second-order transition temperatures and related properties of polystyrene. I. influence of molecular weight. J. Appl. Phys. 21, 581–591 (1950) 12. Farrar, M.J., Turner, T.F., Planche, J.-P., Schabron, J.F., Harnsberger, P.M.: Evolution of the crossover modulus with oxidative aging. Transp. Res. Rec. 2370(1), 76–83 (2013) 13. Garcia Cucalon, L., Kaseer, F., Arámbula-Mercado, E., Epps, A., Martin, N. M., Pournoman, S., Hajj, E.: The crossover temperature: Significance and application towards engineering balanced recycled binder blends. Road Mater. Pavement Des. 20, 1391–1412 (2018)
Shear Fatigue Cracking Analysis of Pavement with Thick Asphalt Layers Induced by Traffic Load Shan Jingsong, Han Lujia, Guo Zhongyin, and Li Feng
Abstract Top-Down cracking is one of the major distresses induced by traffic load for the pavement with thick asphalt layers. In this paper, an analytic model of pavement with thick asphalt layers is established in which four types of base layers are considered corresponding to the different combinations of semi-rigid and granular materials. The shear stress intensity induced by the traffic load in thick asphalt layer is calculated by the analytic model firstly. Then many beam specimens of asphalt mixture are made in laboratory and the shear fatigue test were conducted. According to the test results, the fatigue cracking parameters in Paris equation were regressed. Finally, the shear fatigue life of thick asphalt layers was evaluated basing on shear stress intensity and Paris Law. The research results show that the shear fatigue life increases with the increasing of asphalt layers thickness. The structure I (thick asphalt layer with semi-rigid base and semi-rigid subbase) has the longest fatigue life and structure IV (thick asphalt layer with granular base and granular subbase) is weakest in resisting shear fatigue cracking. Keywords Thick asphalt pavement · Top-down crack · Shear fatigue test · Fatigue life
1 Introduction According to lots of field investigations, Top-Down (TDC) crack near the Wheel track is one of the common distresses in many asphalt pavements which propagates from pavement surface to bottom. Many scholars have done lots of research on S. Jingsong · H. Lujia Shandong Key Laboratory of Civil Engineering Disaster Prevention and Mitigation, Shandong University of Science and Technology, Qingdao, China G. Zhongyin (B) College of Transportation Engineering, Tongji University, Shanghai, China e-mail: [email protected] L. Feng Shandong Huitong Construction, Jinan, China © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_211
1665
1666
S. Jingsong et al.
Top-Down shear cracking induced by traffic load considering different pavement structures. Literature [1] used the combination of fracture mechanics and multidomain mixed boundary node method to conduct physical analysis and predict stress intensity factors on a TDC pavement, and used enriched basis functions to simulate the singularity of stress at the crack tip. Literature [2] used a lot of three-dimensional finite element analysis methods to study the crack behavior of cracked asphalt layers under different traffic loads. The study [3, 4] established a simulation model to predict the development of TDC on asphalt pavement. The orthogonal test and logistic regression method were used to analyze the main factors influencing the development of TDC [5, 6] and in the literature [7] EFG and FE were coupled to analyze the influence of structural parameters on TDC expansion. In the literature [8–11] finite element method were used to analyze the influence of temperature, crack length, thickness and stiffness of base layer on TDC crack stress intensity factor. From the above literatures, the development and propagation of TDC involve many factors such as pavement structure, traffic load, environment and so on. In this study, four types of thick asphalt pavement with different base material were included and shear fatigue test were conducted in laboratory to evaluate the fatigue life of TDC of thick asphalt layers.
2 Pavement Structures In this study, four types of thick asphalt pavement structures are proposed. The thickness of asphalt layers including SMA (Stone Mastic Asphalt) and AC (Asphalt concrete) are same in the four pavement structures and ranged from 18 to 30 cm, but the base materials are different (Fig. 1). In Structure I the base layers are cement treated base and subbase (CTB), but in structure IV it is flexible unbound granular (UGM) base and subbase. Differently, the base layers of structure II and III are combination of CTB and UGM. SMA-13
4 cm
SMA-13
4 cm
SMA-13
4 cm
SMA-13
AC-20
6 cm
AC-20
6 cm
AC-20
6 cm
AC-20
6 cm
AC-25
8-20 cm
AC-25
8-20 cm
AC-25
8-20 cm
AC-25
8-20 cm
CTB
16 cm
UGM
16 cm
CTB
16 cm
UGM
16 cm
CTB
16 cm
CTB
16 cm
UGM
16 cm
UGM
16 cm
Subgrade
(a)Structure I
Subgrade
(b)Structure II
Subgrade
(c)Structure III
Fig. 1 Four pavement structures with different base materials
4 cm
Subgrade
(d)Structure IV
Shear Fatigue Cracking Analysis of Pavement …
1667
Table 1 Material parameters Layer
Thickness/cm
Modulus/MPa
Poisson’s ratio
Density/kg/m3
AC
18–30
10622
0.30
2400
CTB
36
1200
0.20
2300
UGM base
16
500
0.35
2300
UGM subbase
16
300
0.35
2300
60
0.35
1800
Soil base
3 Stress Intensity Factor Analysis of Different Pavement Structures 3.1 Finite Element Model In this study, finite element model (FEM) was used to analyze the shear fatigue cracking of pavement. The analysis model is 7 m deep by 10 m width and the traffic load was applied as eccentric load. A two-dimensional finite element cracking model of ABAQUS was established and the model element type of asphalt pavement structure fatigue analysis was CPE4R. Singular element was placed near the crack tip.
3.2 Material Parameters According to previous research and test results in lab, the dynamic modulus of asphalt concrete was taken at 21 °C and 10 Hz. The static modulus were taken for semi-rigid materials and unbound materials. All of the material parameters were listed in Table 1.
3.3 Shear Stress Intensity Factor of Different Pavement Structures The shear stress intensity factors of the four pavement structures were calculated and plotted in Fig. 2. The results showed that the shear stress intensity factor at the four structures decreased with the increase of the asphalt thickness, indicating that increasing the thickness of the asphalt layer can delay the propagating of TDC. Base course materials properties significantly influenced the shear stress intensity factors of TDC. If the base and subbase were semi-rigid materials (structure I), the stress intensity factor increased slowly with the increase of crack length at the beginning stage. When the crack extended to about 65% of the thickness of asphalt layer, the
1668
S. Jingsong et al.
0.18 0.16
0.40 0.35 0.30 0.25 0.20 18cm 0.15 24cm 0.10 30cm 36cm 0.05 42cm 0.00 -0.05 -0.05 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45
Crack Depth (m)
Type II stress intensity factor K
Stress intensity factor K
0.50 0.45
0.14 0.12 0.08 0.06 0.04 0.02 0.00
0.10
0.15
0.20
0.25
0.30
0.35
(b) Structure II 0.30
0.18 0.14 0.12 18 cm 22 cm
0.10 0.08
26 cm 30 cm 34 cm
0.06 0.04 0.02 0.05
0.10
0.15
0.20
Crack Depth(m)
(c) Structure III
0.25
0.30
Type II stress intensity factor K
0.16
Type II stress intensity factor K
0.00 0.05
Crack Depth(m)
(a) Structure I
0.00 0.00
18cm 22cm 26cm 30cm 34cm
0.10
0.25 0.20
18cm 24cm
0.15
30cm 0.10
36cm 42cm
0.05 0.00 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Crack Depth(m)
(d) Structure IV
Fig. 2 Shear stress intensity factor at different cracking length for different asphalt thickness
stress intensity factor began to increase rapidly. Significantly different from structure I, when the base layer or the Subbase contains the UGM layer (Structure II, Structure III, Structure IV), the shear stress intensity factor of TDC increases rapidly in the initial stage with the increase of the crack propagation. When the crack extended to a certain depth (about 50% of the asphalt thickness), the stress intensity factor reached a peak, and then decreased slowly.
4 Shear Fatigue Cracking Parameters Based on Shear Fatigue Test Paris formula model is currently the most widely used for the fatigue crack propagation defined as Eq. (1) in which crack propagation rate ddNa is used representing the crack propagates length per loading cycle. Under the limited test condition, the relationship between crack propagation rate and stress intensity factor can be regressed.
Shear Fatigue Cracking Analysis of Pavement …
da = C(K )n dN
1669
(1)
where K is the amplitude of the stress intensity factor under a given load; C, n are the fracture parameters of a given material; a is the length of the crack; N is number of standard axle load; h is the thickness of the pavement. According to Eq. 1, the following relationship can be derived: ln
da dN
= lnC + nlnK
(2)
Equation (2) shows that ln ddNa has a linear relationship with lnK. In this study the shear fatigue test of asphalt mixture beams were conducted to obtain the Pars formula parameters C and n.
4.1 Shear Fatigue Test The rutting slabs of SMA13, AC20 and AC25 were made and cut into 150 mm × 50 mm × 50 mm beam specimens. Then pre-cut notch with 7mm deep was made for the shear fatigue test. The test temperature was set as 20 °C. The fatigue load peak was set as 800 N. And loading frequency was 10 Hz. The test steps are as follows: (1)
(2) (3)
Before the test, the specimens was kept at a 20 °C environmental chamber for more than 5 h, and the test temperature was adjusted to 20 °C by air conditioning; Inserting the end of one specimen in straight shear test mold and fixed, then placing the test mold on the MTS 810 material test system (Fig. 3); Set the MTS loading parameters including loading frequency, loading waveform, data collection frequency and so on.
The fatigue crack growth curves of the three asphalt mixtures (AC-20, AC-25, SMA) are shown in Fig. 4. In the plot the crack propagation rate were fast at the initial stage, then the crack propagation rate decreased and maintained at a stable rate for the three mixtures. When the crack reached a certain length, the crack grew rapidly and became unstable with the increase of loading times. Overall, SMA had the best ability to resist shear cracking, followed by AC-20, and AC-25 was the worst.
4.2 Paris Formula Parameters (1)
Crack propagation rate of beam specimens
1670
S. Jingsong et al.
Fig. 3 Shear fatigue test
50
Fig. 4 Cracking propagation length with loading cycles
45 SMA13
Crack Length (mm)
40 35
AC20
30
AC25
25 20 15 10 5 0 0
0.5
1
1.5
Loading Times (
2
2.5
3
)
The relationship between crack length of beam specimen and the number of loading cycles can be got by the shear fatigue test, and then the crack propagation rate ddNa can be obtained by differentiating the fatigue crack growth curve of three asphalt mixtures. In addition, a finite element model of shear fatigue test in lab was established and the variation of stress intensity factor with crack length in asphalt beam was calculated (Fig. 5). (2)
Regression of Paris formula parameters
The relationship between ln K and crack length was got by taking logarithm of the stress intensity factor (Fig. 6). Linear regression was performed on logarithm of crack growth rate and corresponding stress intensity factor, and parameters C and n of Paris formula could be obtained, listed in Table 2.
Shear Fatigue Cracking Analysis of Pavement …
1671
Table 2 Paris formula parameters Asphalt mixtures
n
Intercept
C
Correlation
SMA13
1.925
−12.201
5.03E−6
0.9238
AC20
1.8391
−11.832
7.27E−6
0.9382
AC25
1.3968
−10.8
2.04E−5
0.9042
12
Fig. 5 Shear stress intensity factor with cracking length of beam specimen
10
K MPa
8 6 4 2 0 0
10
20
30
40
50
Crack Length (mm)
5 Shear Cracking Life Prediction of Thick Asphalt Layers Using the Paris formula parameters regressed above and the shear stress intensity factors calculated by FEM, the shear fatigue life of thick asphalt layers in different pavement structures can be integrated (Fig. 7). According to the results, the shear fatigue life of asphalt pavement increases with the increase of the thickness of asphalt pavement, and the shear fatigue life of structure I is remarkably longer than other structures. Increasing the thickness of asphalt layers can clearly increase the shear fatigue life for structure I. Structure II and structure III have close fatigue life. The fatigue life of the four pavement structures is ranked as structure I > structure II > structure III > structure IV.
6 Conclusion (1)
The shear stress intensity factor of TDC is significantly affected by the type of the base layer material. The stress intensity factor is smaller and increases slowly with the propagation of cracking when the semi-rigid base and subbase are used. Instead, when UGM layer is used as base or subbase layer, the shear stress intensity factor of TDC increases rapidly during the propagation of cracking and begins to decrease when the crack length exceeds about 50% of the asphalt thickness.
1672
S. Jingsong et al.
Fig. 6 Logarithmic relationship between K
da dN
-9.5
and
SMA13
-9
AC20 AC25
-8.5
ln
-8 -7.5 -7 1
1.25
1.5
1.75
ln K (MPa
2
2.25
2.5
)
9
Fig. 7 Shear fatigue life of thick asphalt layers
)
8
Structure I StructureII Structure III StructureIV
7
Fatigue Life (
6 5 4 3 2 1 0
0
10
20
30
40
50
Surface Thickness (cm)
(2)
(3)
Fatigue shear tests show that SMA had the better ability to resist shear cracking than AC-20 and AC-25. The parameters C and n in the Paris formula can be regressed by the fatigue shear test data and FEM results. The shear fatigue life increases with the increase of the thickness of asphalt layer, and clearly affected by the base materials. The structure I (CTB base and CTB subbase) has the longest shear fatigue life that is significantly greater than the other three structures.
References 1. Miao, Y., He, T.G., Yang, Q., Zheng, J.J.: Multi-domain hybrid boundary node method for evaluating top-down crack in Asphalt pavements. Eng. Anal. Boundary Elem. 34, 755–760 (2010) 2. Ameri, M., Mansourian, A., Heidary Khavas, M., Aliha, M.R.M., Ayatollah, M.R.: Cracked asphalt pavement under traffic loading—A 3D finite element analysis. Eng. Fract. Mechan. 78,
Shear Fatigue Cracking Analysis of Pavement …
1673
1817–1826 (2011) 3. Fan, G., Luo, X., West, R.C., Taylor, A.J., Taylor, A.J., Moore, N.D.: Energy-based crack initiation model for load-related top-down cracking in asphalt pavement. Constr. Build. Mater. 159, 587–597 (2018) 4. Ling, M., Xue Luo, Y., Chen, S.H., Lytton, R.L.: A calibrated mechanics-based model for top-down cracking of asphalt pavements. Constr. Build. Mater. 208, 102–112 (2019) 5. Yongjie, L., Shaopu, Y., Jianxi, W.: Research on pavement longitudinal crack propagation under non-uniform vehicle loading. Eng. Fail. Anal. 42, 22–31 (2014) 6. Shen, S., Zhang, W., Shen, L., Huang, H.: A statistical based framework for predicting field cracking performance of asphalt pavements: Application to top-down cracking prediction. Constr. Build. Mater. 116, 226–234 (2016) 7. Luo, H., Zhu, H., Miao, Y., Chen, C.-y.: Simulation of top-down crack propagation in asphalt pavements. J. Zhejiang Univ. Sci. A Appl. Phys. Eng. 11, 223–230 (2010) 8. Zhao, Y., Wang, S., Zhou, C., Tan, Y.: Analysis of top-down cracking of asphalt pavements based on fracture mechanics approach. J. Tongji Univ. (Nat. Sci.) (2010) 9. Aliha, M.R.M., Sarbijan, M.J.: Effects of loading, geometry and material properties on fracture parameters of a pavement containing top-down and bottom-up cracks. Eng. Fract. Mech. 166, 182–197 (2016) 10. Sun, Y., Cong, D., Zhou, C., Zhu, X., Chen, J.: Analysis of load-induced top-down cracking initiation in asphalt pavements using a two-dimensional microstructure-based multiscale finite element method. Eng. Fract. Mech. 216, 1064 (2019) 11. Ai, C., Xu, C., Ren, D., Guo, Y., Yang, E.: Characterization of vertical and horizontal propagations of double cracks in asphalt pavements under moving loads. J. Southwest Jiaotong Univ. (2018)
Solar and Permeable Road: A Prototypical Study Domenico Vizzari, Pierfabrizio Puntorieri, Filippo G. Praticò, Vincenzo Fiamma, and Giuseppe Barbaro
Abstract A solar road is a pavement infrastructure able to generate electricity exploiting the solar radiation. Notwithstanding the progress of the research, several issues still hinder these typologies from getting an outstanding role, among which their uncertain durability, surface performance, and permeability. In the light of this, a novel permeable and solar road is here proposed. An innovative top layer was designed, taking into account hydraulic and transport-related issues. The basic idea was to improve the wet friction by setting up and advancing a water drainage system. The pavement was validated in terms of hydraulic and friction properties, in order to assess its ability to perform satisfactorily in dry and wet conditions. A system to simulate rainfall was designed and constructed. The results demonstrate the feasibility of coupling energy harvesting and premium surface properties (such as permeability). The contributions of this study aim to improve from a draining point of view the typical solar pavement package. Results can benefit both practitioners and researchers. Keywords Solar pavement · Friction · Rainfall simulator · Hydroplaning · Permeable pavement
1 Introduction As is well known, a solar road usually includes a road surface layer, an electronic layer and a base plate layer. The semi-transparent road surface layer provides traction for vehicles based on surface texture and supports the traffic load. Sunlight passes
D. Vizzari (B) IFSTTAR/MAST-MIT, Nantes, France e-mail: [email protected] P. Puntorieri · V. Fiamma · G. Barbaro Mediterranea University/DICEAM, Reggio Calabria, Italy F. G. Praticò Mediterranea University/DIIES, Reggio Calabria, Italy © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_212
1675
1676
D. Vizzari et al.
through solar embedded collector cells, light-emitting diodes and a heating element [1]. The flash floods are an important issue which can be extremely destructive [2, 3]. Climate change has a direct impact on the precipitation patterns. In this scenario, it is necessary to consider how to adapt new road systems such as solar pavements to climate change. Determining the best practice approach on how to adapt infrastructure to a changing climate is an ongoing challenge for the engineering community [4]. The first main consequence of the rainfall on a solar road is the formation of the water film. The lack in drainage systems during a rainfall facilitates the formation of the water film on the surface, leading to potential hydroplaning phenomena, especially for impermeable surfaces (e.g., solar pavements). For a better comprehension of the drainage ability of a solar pavement, it is necessary to study the behavior of the water film and the effects of the pipe networks using a rain simulator. In the light of the above, the main objectives of this paper are to set up and validate a solar pavement from the perspective of road safety, in wet and dry conditions.
2 Design of Surface Properties The prototype herein proposed differs from the common structure of a solar pavement through a drainage network. This latter is included in the solar road along the entire width and length. Among the main surface requirements of pavements, the following can be listed: friction, texture, and quietness [5]. For the evaluation of the texture and of the friction [6], CIRS (Centro Interuniversitario Sperimentale di Ricerca Stradale ed Aeroportuale) specifications suggest to perform the Skid Test (using the British portable pendulum) to obtain the BPN (British Pendulum Number, now Pendulum Test Value PTV according to the EN 13036-4) and the sand patch test to obtain SH (Sand Height, mm, according to the EN 13036-1). The minimum values of PTV and SH for a DGFC (Dense Graded Friction Course) are 55 and 0.4 mm (CIRS), respectively. The Skid tester and sand patch method were used to obtain the values of PTV and SH for five types of glass: (i) smooth glass surface used for the experiment (SURF1); (ii) three rough glasses (SURF2, SURF3, SURF4); (iii) and a solar panel with a plexiglas surface (SOLAR). Results (averages) are shown in Table 1. It is noted that PTV and SH values look very far from the lower specification limits. Importantly the macro-texture of the rough glass results very low, so contributing to reduce wet friction. To this end, it is noted that: several skid-resistant glass floor were produced that have a skid resistance of about 50–80 [7]; a potential solution is to use a controlled solution of hydrofluoric acid to uniformly etch the surface of the glass. The result is a glass which is not susceptible to coloration and it is less porous than a sandblasted glass. In addition, the wet dynamic coefficient of friction (testing thought ANSI A137.1-2012 Section 9.6, DCOF-DOT-3000) is 0.56–0.67 [8];
Solar and Permeable Road: A Prototypical Study Table 1 Experimental values of PTV and SH on five different glass surfaces
1677
Type of glass
PTV
SH (mm)
SURF1
13
0.20
SURF2
20
0.20
SURF3
13
0.27
SURF4
20
0.46
SURF5
14
0.20
a further solution is to use glass aggregates bonded together using the polyurethane as binder [9]. Furthermore, the glass aggregates reduce the efficiency of the solar panel from 43 to 76% depending on the thickness of the top layer, the binder content and the presence of fine particles [10]. Based on the experiments and studies above, it follows that friction and texture properties of solar roads are key-factors and this crucial factor calls for further breakthrough innovation. To this end, it was decided to explore for permeable and solar roads.
3 Design of Surface Properties This section describes the experiments designed and carried out in order to set up a solution able to decrease the risk of safety concerns in wet conditions. A full-scale prototype of solar road capable of quickly draining the water is presented. To this end, a vital issue is to estimate the maximum volume of water to manage on a solar pavement. A research on the highest rainfalls recorded in Europe, with a 100-year return period has been carried out. Based on the international literature [11–13], the maximum intensity in Europe corresponds to about 300 mm/h, even if thresholds usually refer to durations between 1 and 100 h, and to intensities from 1 to 220 mm/h. Consequently, the value of 220 mm/h was used. The prototype idea is to place pipes in the solar pavement. Based on a previous feasibility study [14], the design for 1 m2 of pavement draining system consists of 49 pipes, having a diameter of 2 mm. To evaluate the drainage ability of the solar pavement, a device was designed to simulate rainfall and run off processes. The experimental set up is composed of (1) A lattice of droppers (one per 0.0125 m2 ) to simulate rainfalls (automated laboratory rainfall simulation system) with different rainfall intensities. (2) A glass smooth pavement (see Fig. 2b and Table 1 for the data about friction and texture), having 49 (7 × 7) pipes made of glass fibers. (3) A system to estimate the drainage capacity by calculating the mass of the water flowing through the pipes (herein termed Qp) and the mass flowing through the closing section (corresponding, in a real, section to the kerb, herein termed Qcs). (4) A system of screws that allow adjusting the surface slope from 0 to 7% in order to simulate the road cross slope (banking). The following procedure was used to estimate the need for having a given diameter of pavement pipes: (A) The rainfall simulation system was activated, for a given number of pipes, pipe diameter, and slope. To this end, a given
1678
D. Vizzari et al.
number of pipes was plugged, following a symmetrical scheme. (B) The time needed to achieve a given constant height of water was recorded (equilibrium time, Fig. 1 y-axis). (C) The corresponding height of water was recorded (equilibrium height, radius of the bubble). (D) The experiments were carried out for different slopes and for different numbers of pipes. (E) The procedure above was carried out by using different diameters, number of pipes, and slopes. (F) Experimental data were used to simulate flows and heights in a 3.5-wide lane, including the consequences deriving from having, at the same time, Qp (deriving from the sections above) and Qcs. (G) Finally, by considering the points above, a 7×7 pipes system was considered, with a final pipe diameter of 6 mm. Figure 1 refers to the results obtained in the experiments, under the hypotheses shown above (cf. Point g). A bubble chart is here used. This latter can be considered a variation of the scatter plot, in which the data points are replaced with bubbles for considering a third variable (in this case, water height). It illustrates the relationship among the three main variables: (i) the number of pipes working, that is to say not obstructed (x-axis); (ii) the equilibrium time, that is to say the time needed to reach a constant level of water (t, constant level of water, y-axis); (iii) and the corresponding equilibrium height (h, diameter of the bubble). Fig. 1a refers to a slope of 0%, while Fig. 1b addresses the case with a transverse slope of 2.5%, according to Italian road design standards. Based on results, it may be observed that considering a cross slope of 0, the equilibrium between the flow rate provided by the device and the flow rate passing through the pavement is reached in 17 min with 3 pipes and a water film of 17 mm. Increasing the pipes number, the equilibrium is reached in less time (t) and with a lower water depth (h). For example, 7 pipes reach the flow equilibrium in about 3 min with h = 4 mm. Performing the test for 2 or 1 pipes, the value of h exceeds 20 mm, overflowing the edge around the layer and consequently compromising the validity of the test (there is not an equilibrium height). When changing the slope from 0 to 2.5%, the water flow passing through the pipes decreases because a part is drained in the closing section thanks to the layer’s slope. In this case, the equilibrium between the input and the output flow is reached in 5 min with only one pipe and h is 4 mm. With 4 pipes, the layer is able to drain the water reaching the equilibrium in
Fig. 1 Number of pipes, time to reach the input-output flow equilibrium, and water film thickness
Solar and Permeable Road: A Prototypical Study
1679
Fig. 2 a Percentage of water flow passing through the pipes (Qp) and the closing section of the pavement (Qcs) for different values of cross slope; b smooth glass surface used for the test
1.5 min for a value of h = 1 mm. Figure 2a illustrates how the slope of the pavement (x-axis) influences the percentage of flow drained through the pipes (QP) or through the closing section (Qcs). (i) For a cross slope of 0% the water is drained by the pipes (Qp 100%; Qcs 0%); (ii) for a cross slope of 7% most of the water is drained thanks to the same slope (Qp 17%; Qcs 83%); (iii) the ratio between Qp and Qcs is about 1 when the slope is between 2 and 2.5%. Experiments above demonstrate that for a transverse slope of 2.5%, consistent with the majority of EU standards, the pavement is able to drain an intense rainfall, preventing the formation of the water film and reducing the hydroplaning probability. For clogging, it is noted that pipes are 6 mm wide and the tortuosity is 1. This implies that the de-clogging of pavement structure could be carried out quite easily, based on bottom-up or top-down methods [15, 16]. Overall, the proposed solution appears to reduce water depth, with benefits for texture, friction, and hydroplaning mitigation [17, 18].
4 Conclusions In this paper, moving from a typical solar road pavement, a prototype of solar and permeable pavement was designed and produced. Pipes were located at the corners of each solar slot. Overall, in 1 m2 of pavement, 49 pipes with a diameter of 6 mm were made. Friction potential in the interface between rubber and glass is a critical requirement and involves different phenomena for the diverse driving maneuvers, interface conditions, tire/pavement context, and pipes pattern. In terms of friction, plan glass surfaces are not a good solution especially in wet conditions. For permeability, hydraulic tests demonstrate that the surface set up is able to drain an intense rainfall, preventing the formation of the water film. Considering also road cross slope, a considerable percentage of the drained water passes through the closing section, increasing the drainage ability of the pavement. Clogging over time is a concurrent issue and pipe distribution and type were designed accordingly. Further researches will focus on the improvement of the friction, testing not only new glasses
1680
D. Vizzari et al.
with texture-related solutions, but also layers composed of glass aggregates mixed together using a transparent glue.
References 1. Dezfooli, A.S., Nejad, F.M., Zakeri, H., Kazemifard, S.: solar pavement: a new emerging technology. Sol. Energy 149, 272–284 (2017) 2. Abudi, I., Carmi, G., Berliner P.: Rainfall simulator for field runoff studies. J. Hydrol. 454– 455(6), 76–81 (2012) 3. Barbaro, G., Petrucci, O., Canale, C., Foti, G., Mancuso, P., Puntorieri, P.: Contemporaneity of floods and storms. A case study of metropolitan area of Reggio Calabria in Southern Italy. In: Proceedings of New Metropolitan Perspectives (NMP) (Reggio Calabria, Italy), Smart Innovation, Systems and Technologies, vol. 101, pp. 614–620 (2019) 4. Taylor, M.A.P., Philp, M.P.: Investigating the impact of maintenance regimes on the design life of road pavements in a changing climate and the implications for transport policy. Transp. Policy 41, 117–135 (2015) 5. Praticò, F.G., Vaiana, R., Iuele, T.: Macrotexture modeling and experimental validation for pavement surface treatments. Constr. Build. Mater. 95, 658–666 (2015) 6. Torbruegge, S., Wies, B.: Characterization of pavement texture by means of height difference correlation and relation to wet skid resistance. J. Traffic Transport. Eng. 2(2), 59–67 (2015) 7. Walking on glass. https://www.rogerwilde.com/slip_resistance.pdf. Last accessed 25 Aug 2020 8. Walkerglass. https://walkerglass.com/docs/productspecifications.pdf. Last accessed 1 July 2020 9. Vizzari, D., Chailleux, E., Gennesseaux, E., Lavaud, S., Vignard, N.: Viscoelastic characterization of transparent binders for application on solar roads. Road Mater. Pavement Des. (2019). https://doi.org/10.1080/14680629.2019.1588774 10. Vizzari, D., Chailleux, E., Lavaud, S., Gennesseaux, E., Bouron, S.: Fraction factorial design of a novel semi-transparent layer for applications on solar roads. Infrastructures 5, 5 (2020) 11. Guzzetti, F., Peruccacci, S., Rossi, M.: Rainfall thresholds for the initiation of landslides in central and southern Europe. Geomorphology. 290, 39–57 (2017) 12. García-Bartuala, R., Schnider, M.: Estimating maximum expected short-duration rainfall intensities from extreme convective storms. Phys. Chem. Earth Part B 26, 675–681 (2001) 13. Crosta, G.B., Frattini, P.: Distributed modelling of shallow landslides triggered by intense rainfall. Nat. Hazards Earth Syst. Sci. 3, 81–93 (2003) 14. Vizzari, D., Puntorieri, P., Praticò, F., Fiamma, V., Barbaro, G.: Energy harvesting from solar and permeable pavements: a feasibility study. Ann. Chim. Sci. Mater. 42, 503–520 (2018) 15. Razzaghmanesh, M., Beecham, S.: A review of permeable pavement clogging investigations and recommended maintenance regimes. Water 10, 337 (2018) 16. Malesi´nska, A., Rogulski, M., Puntorieri, P., Barbaro, G., Kowalska, B.: Displacements of the pipe system caused by a transient phenomenon using the dynamic forces measured in the laboratory. Measur. Control (UK). 51, 443–452 (2018) 17. Praticò, F.G., Astolfi, A.: A new and simplified approach to assess the pavement surface microand macrotexture. Constr. Build. Mater. 148, 476–483 (2017). https://doi.org/10.1016/j.conbui ldmat.2017.05.050 18. Cerezo, V., Gothie, M., Menissier, M., Gibrat, T.: Hydroplaning speed and infrastructure characteristics. In: Proceedings of the Institution of Mechanical Engineers, Part J: Journal of Engineering Tribology, SAGE Publications, vol. 224, no. 9, pp. 891–898 (2010)
Statistical Analysis of the AIGLE 3D Crossover Trial Campaign: Experimental Design, Methods and Results Pierre Gayte and Marie Delaye
Abstract These ever-increasing needs for road maintenance and the willingness to optimise its maintenance works with a long-term approach leads Cerema to develop new tools and methods for road survey. Based on Pavemetrics Laser Crack Measurement System (LCMS) sensors and a precise GNSS localization system, Cerema has designed two new inspection vehicles called AIGLE 3D. To make the measurements campaign more reliable, we implemented an optimized organization and significant improvements such as remote maintenance and a merged tool controlling all sensors in vehicles. Pavemetrics, a partner of Cerema, developed the LCMS system and new 3D functionalities allowing measurements up to 130 km/h. LCMS sensors can measure the roadway with an accuracy of 1 mm in the directions of the roadway and 0.5 mm in height up to a speed of 130 km/h. This set (LCMS sensors, localization) provide a 3D model of the road surface instead of considering it as a single wire. Thus, we obtain a true representation of the data collected with the system. The data processing system developed at Cerema makes it possible to precisely geo-locate all degradations, deformations, geometry and objects on the roadway. With this new technology, it becomes necessary to set up a new method of exploitation and analysis for the auscultation of roadway. Cerema performed a cross-testing campaign in order to evaluate the quality of the data produced by AIGLE 3D inspection vehicles. The purpose of this article is to resume the crossover trial campaign and set up an analysis method. Results of this measurement are then presented and analyzed. Keywords AIGLE 3D · Inspection vehicle · Road survey · Asset management
P. Gayte (B) · M. Delaye Cerema ITM, 110 Rue de Paris, 77171 Sourdun, France e-mail: [email protected] M. Delaye e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_213
1681
1682
P. Gayte and M. Delaye
1 Introduction Since 2015, modernization of the evaluation of French national road assets leads to the development of two new innovative inspection vehicles called “Aigle 3D” equipped with Pavemetrics Laser Crack Measurement System (LCMS) [1] and a precise GNSS localization system. They allow to measure, at high speed, degradation, deformation and object on the road surface. A data and computing center processes these data to detect and geolocate data. In order to quantify the quality of Aigle 3D data, Cerema performed a main crossover trial campaign. The purpose of this article is to (1) resume the crossover trial campaign by presenting the test plan and parameters; (2) set up an analysis method of the collected data; and finally (3) present and analyze results.
2 Crossover Trial Campaign During the crossover trial campaign, 9 tests were performed as presented in Table 1. Direction + (−) means that the run is performed along increasing (decreasing) road direction. Average length of these sections is between 1.5 and 5 km. Two Aigle 3D inspection vehicles performed the tests on these 9 sections located in two different departments near Lyon, France. Experimental conditions of runs studied in the present paper were as following (tables, figures and results of this paper do not present runs performed with other conditions of speed, trajectory and luminosity): • • • •
Speed of vehicle: constant at 80 km/h during all the run Trajectory: stable between the right and left road marks Luminosity: by day Operator: 4 operators trained and qualified to perform measurements
Table 1 List of inspected sections during the cross test campaign performed with department, road, direction and average length Section
Road
Direction
Average total length (km)
Number of runs
A
RD29
+
3
15
B
RD29
−
3
15
C
RD154
+
2
12
D
RD306
+
1.5
14
E
RD154 and RD306
−
3.5
14
F
RD518
+
2.5
21
G
RD518
−
2.5
22
H
RD518
+
5
12
I
RD518
−
5
13
Statistical Analysis of the AIGLE 3D Crossover …
Road lane
1683
Road marks
Analysis area
Physical limits Fig. 1 Analysis area definition from road marks (white, right and left sides) and two physical limits (red, at the beginning and end of the lane)
3 Method of Analysis 3.1 Analysis Area In order to compare precisely the parameters measured on each section (presented in the next section: total crack length, mean MTD, mean rut depth and mean rut width) from one run to another, analysis area are defined as the area of the road lane between the following bounders: • Road marks on right and left sides, and • Two physical limits chosen at the beginning and the end of the section (for example a noticeable road mark, crack or sign on the road surface easy to recognize on every run of the section) These rules allows defining the analysis area that ensure the analysis to be done within the same physical area for each run on a given section (Fig. 1).
3.2 Studied Data Aigle 3D inspection vehicle are able to measure and quantify a large range of degradation and deformations of the road surface [1–4]. The study present four of these data: • 2 degradations parameters: crack length and mean texture depth (MTD) • 2 deformations parameters: width and depth of rut measurements For each section (from A to I), for each run (between 12 and 22 runs per sections) these parameters are computed to obtain: • Total crack length as the sum of lengths of each cracks detected on the section
1684
P. Gayte and M. Delaye
Table 2 Statistics on analysis area of sections (A to I) computed Section
Number of runs
Average measured area (m2 )
Maximum measured area (m2 )
A
15
8116
8420
B
15
8150
8162
C
12
6679
D
14
3271
E
14
F
21
G
Minimum measured area (m2 )
Standard deviation (m2 )
Normalized standard deviation (%)
8067
89
1.09
8137
7
0.08
6719
6655
21
0.32
3277
3266
4
0.11
7653
7729
7539
52
0.68
5543
5629
5518
24
0.44
22
5658
5697
5628
17
0.31
H
12
15324
15409
15242
46
0.30
I
13
15259
15323
15178
50
0.33
• Mean MTD as the average MTD of the section • Mean rut depth as the average rut depth of the section • Mean rut width as the average rut width of the section.
4 Results 4.1 On Analysis Area Analysis area are computed following the method described in part 2.1 for each section and each run. Table 2 presents these results. Except section A (where it is equal to 1.09%), normalized standard deviation is less than 0.68% and so the range between maximum and minimum measured area is narrow. Regarding these results, analysis area computed in this step are satisfying. Lane marks on section A were in a poorer quality than on other sections. Then the boundaries were harder to detect and it may explain the higher standard deviation for this section. Analysis area allow to compare parameters on a stable area and to go further to study degradations and deformation on these analysis areas.
4.2 On Studied Degradations and Deformations Degradation and deformations parameters are computed using automatic detection algorithms. For each section from A to I, every parameters is analyzed inside the analysis area. For example, total crack length is the sum of lengths of every cracks inside the analysis area and the mean rut depth is the average rut depth measured
Statistical Analysis of the AIGLE 3D Crossover …
1685
in the analysis area. Measurement outside the analysis area are excluded from the current analysis. Figure 2 presents results for degradations parameters studied here in a box plot diagram. Results are normalized by the average value (calculated on all test runs of the section). Orange dash represents the median value. Results show a rather good repeatability as a large majority of values is centered in the 0.85–1.15 interval. Figure 3 presents results for deformations parameters studied in a box plot diagram. Results are normalized by the average value (calculated on all test runs of the section) and orange dash represents the median value. All values are between 0.9 and 1.1 which means mean rut parameters (depth and width) are measured with a good repeatability. For a given section, relative average spread of a section (RAS(S)) is defined as follow:
Fig. 2 Box plot of the normalized total crack length (left) and mean MTD (right) for sections A to I. Data normalized by the average data on the given section.
Fig. 3 Box plot of the normalized mean rut depth (left) and mean rut width (right) for sections A to I. Data normalized by the average rut depth/width on the given section.
1686
P. Gayte and M. Delaye
Table 3 Mean (average RAS(S) on sections A to I), maximum and minimum relative average spread (%) obtained on the 9 sections for the four studied data Mean relative average spread (%)
Maximum relative average spread (%)
Minimum relative average spread (%)
Total crack length
4.4
10.2
1.8
Mean MTD
4.4
7.9
2.2
Mean rut depth
1.6
2.5
0.9
Mean rut width
1.5
2.6
0.7
R AS(S) =
1 − xi − x n i
(1)
−
where n is the number of runs, x is the average value of the parameter on the section and xi are individual values of the parameters for each run from 1 to n on the section. From RAS(S) for sections A to I and for each parameter studied in this paper, Table 3 presents a summary of the results obtained for all sessions combined. Table 3 gives an overall view of the average differences observed for each parameter. Degradation parameters (cracks and MTD) have a rather good RAS with an average value of 4.4%. Results are even better with deformations parameters (rut) with an average RAS less than 1.6%.
5 Conclusions This paper is a first attempt to analyze and quantify the quality of data measured and computed using Aigle 3D inspection vehicles. The results obtained in this study are encouraging: areas and parameters (total crack length, mean MTD, mean rut depth and mean rut width) are measured with a good accuracy and reproducibility at the scale of a road section. Then these vehicles give similar results and according to studies these results are reliable with the classical field measurements [2, 5]. Future studies may focus on specific prospects. Working on different test conditions occurring during measurement (speed, trajectory, daylight, etc.) and more road degradations and deformation parameters may enhance this study. The present global quantitative analysis may hide local phenomenon. A local analysis coupled with a qualitative analysis of the measured degradations and deformations of the road surface would give us more information about reproducibility of the measurement made with Aigle 3D inspection vehicles.
Statistical Analysis of the AIGLE 3D Crossover …
1687
References 1. Laurent, J., Hébert, J.-F., Talbot, M.: Using full lane 3D Road texture data for the automated detection of sealed cracks, bleeding and raveling. In: Pavement and Asset Management: Proceedings of the World Conference on Pavement and Asset Management. ISBN 9780367209896 (2017) 2. Wang, K.C.P. et al.: Automated Survey of Pavement Distress Based on 2D and 3D Laser Images (2011) 3. Wix, R., Leschinski, R.: 3D Technology for Managing Pavements (2013) 4. Laurent, J., et al.: 3D Laser Road Profiling for the Automated Measurement of Road Surface Conditions and Geometry (2013) 5. Henning, T.F.P., Mohammad, N.U.M.: Did We Get What We Wanted? Getting Rid of Manual Condition Surveys (2013)
Study of the Shear Stiffness of Bituminous Mastics Incorporating Recycled Glass Fillers Mohamed Mounir Boussabnia, Michel Vaillancourt, and Daniel Perraton
Abstract This paper studies the effect of recycled glass (RG) filler on the shear stiffness of mastics (mixture of bitumen and filler) in the linear viscoelastic domain (LVE). The research project was carried out at the Laboratory of Roads and Bituminous Materials (LCMB) of the École de technologie supérieure (ÉTS) using a new experimental device called the annular shear rheometer (ASR). The complex shear modulus G* of bitumen and mastics were measured under various temperatures and frequency conditions (from −25 to 45 °C and from 0.03 to 10 Hz). Two types of RG fillers (crushed glass and micronized glass) and a reference limestone filler were used to manufacture the PG70-28 based mastics. The complex coefficient of reinforcement R M ∗ was used to compare the shear stiffness of RG mastics with a limestone-based mastic. Results indicate that the RG filler effect is important at high temperature and/or low frequency where no significant effect was observed at low temperature and/or high frequency. Moreover, the filler size gradation seems to have most influence on the viscoelastic properties of the mastics at high temperature. Overall, it was shown that RG fillers are suitable for improving mastic shear stiffness in a pavement structure. Keywords Bitumen · Mastic · Annular shear rheometer · Reinforcement coefficient · Glass filler · Rheology
1 Introduction Recycled glass (RG) materials are constantly being produced due to the growth of the economy and human consumption. At the same time, most of North America’s recycling and sorting centers aren’t well equipped to produce RG materials meeting the required quality by the glass industry [1]; Consequently, an important part of RG materials are stored in landfills [1]. Thereby, the reuse of RG materials in the road M. M. Boussabnia (B) · M. Vaillancourt · D. Perraton Department of Construction Engineering, École de Technologie Supérieure, Montreal, Canada e-mail: [email protected] 1100 Notre-Dame Ouest, Montreal, QC 3C1K3, Canada © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_214
1689
1690
M. M. Boussabnia et al.
industry as the substitution of raw materials seems like a good solution to reduce the use of natural materials, releases the pressure on landfill and helps aim the road construction industry towards sustainability [2]. Recently, significant efforts have been made by several researchers to improve the stiffness of bituminous materials by using RG materials. Among them, the research team of the «Laboratoire sur les chaussées et les matériaux bitumineux» (LCMB) in the École de Technologie Supérieure (ÉTS) at Montreal, attempts to assess the effects of RG materials on the rheology of bituminous mixtures. The research program studies the possibility of replacing aggregates and fillers (fine particles with a few tens of microns diameter) used in different pavement layers with aggregates and fillers from RG materials. In this paper, only the replacement of limestone filler with RG fillers in the mastic is presented. The mastic is composed of bitumen and filler, it constitutes an intermediate phase between the binder and the bituminous mix [4]. Moreover, when considering the difference in scale between the dimensions of the mastic components and those of the asphalt mix, it is possible to consider the mastic as the real binder in the asphalt mix [3, 4]. The study of the rheology of mastics is then a relevant way to understand the rheological behavior of asphalt concrete materials [4]. The objective of this paper is to assess the effect of RG fillers on the shear stiffness of bituminous mastics. In order to accomplish this task, the influence of the following parameters on the shear stiffness of mastics is analyzed: i) the nature of the filler and ii) the gradation curve characteristics. To do that, the shear stiffness of the mastic is measured through a complex shear modulus G* test using the annular shear rheometer (ASR). The complex coefficient of reinforcement R M ∗ was used to compare the shear stiffness of RG fillers based mastics with a reference limestone based mastic.
2 Methods and Materials 2.1 The Annular Shear Rheometer ASR The ASR is an experimental device developed at the National School of Public Works (ENTPE) in France in 2002. The ASR measures the complex shear modulus of bitumen and bituminous mastics [4]. It makes possible to test samples with annular geometry (hollow cylinder), with a size larger than that of the other known tests (DSR; Searle Rheometer; etc.), over a wide range of temperatures and frequencies [3]. It measures the sinusoidal shear stress and the sinusoidal shear strain (distortion) applied on bitumen or mastic sample. The sample (specimen like a hollow cylinder in Fig. 1) has a thickness of 5 mm, a diameter of 105 mm and a height of 40 mm. The test can therefore be considered quasi-homogeneous. A schematic description of the ASR is given in Fig. 1. The ASR test is carried out at eight temperatures ranging from + 45 to −25 °C with a step of 10 °C. For each temperature, the specimen is subjected to a sequence
Study of the Shear Stiffness of Bituminous …
1691
Fig. 1 Schematic description of the Annular Shear Rheometer. Adapted from Delaporte Brice [4]
of 6 frequencies: 0.03, 0.1, 0.3, 1, 3 and 10 Hz. Finally, two additional frequencies named 1 Hz return and 0.1 Hz return are repeated at the end of the test to check the damage of the specimen. The coefficient of complex reinforcement is used to quantify the stiffening potential of fillers on the mastics. R ∗M is defined as the ratio of the complex modulus of the mastic to that of the bitumen for each pair of temperature and frequency (T, f).R ∗M is a dimensionless complex number having a norm |R ∗M | and a phase angle ϕ R . Its calculation is made at the same equivalent frequency for mastic and bitumen ( f em = f eb = f e ) as shown in Eq. (1). The time-temperature principle must be verify in order to use R ∗M . |R ∗M |
=
∗ G
mastic |G ∗bitumen |
ϕ R = ϕG ∗mastic − ϕG ∗bitumen
(1)
1692
M. M. Boussabnia et al.
2.2 Materials The binder tested in this work is a PG 70-28 grade bitumen. It is modified by the addition of SBS polymers introduced during its manufacture. The addition of polymers improves the rheological properties of the bitumen at high temperature and reduces the kinetic and thermal susceptibility [3–5] of the bituminous mix. This experimental campaign includes three types of fillers: the limestone filler C (reference material), the post-consumer crushed glass filler V (filler from sorting centers) and the micronized glass filler Vx (Verrox® ). The particle size varies from 0 to 160 µm. All fillers are mechanically separated using sieves of size: 160, 80 and 38 µm. The three granular classes constructed are: 0–38 µm for fine fillers, 38– 80 µm for fillers, 80–160 µm for medium fillers. The gradation curves of these fillers, obtained from a laser particle size device at the LCMB Laboratory, are presented in Fig. 2. A filler is identified by its type (C, V, Vx ) and its granular size. For example: C 0–80, is a limestone filler having particles of size between 0, 1 and 80 µm.
Fig. 2 Synthesis of laser particle size gradation results of the fillers used in the project
Study of the Shear Stiffness of Bituminous …
1693
3 Results and Discussions 3.1 Influence of the Nature of the Filler The influence of the filler’s nature is studied by comparing two different minerals: the limestone and the amorphous silica (glass). The fillers used for making mastics are: C 0–80, V 0–38 and Vx 0–38. To assess the influence of the nature of the fillers, the complex coefficient of reinforcement R ∗M is used. It is calculated at Tr e f = 15◦ C. The PG 70–28 bitumen is considered as a reference material. Figure 3 shows the evolution of the coefficient R ∗M (norm and phase angle) of the tested mastics. To understand the interpretation of R M ∗ results, it is important to mention that each decade of equivalent frequency f e (x-axis in Fig. 3) represents a temperature, like [10–3 , 10–2 ] represents +45 °C and [104 , 105 ] represents −25 °C. The values of R ∗M vary for each mastic in a given decade of f e (or given temperature). As an example, for mastic with filler Vx 0–38 at +45 °C, |R ∗M | decreases from 7 to 5.5, which means that this filler has more stiffening potential on mastic at low frequency (0.03 Hz) than at high frequency (10 Hz). Also, a value of 7 for the |R ∗M | means that the mastic’s |G* | is 7 times greater than that of the bitumen. Finally, in order to compare RG fillers to limestone filler, if we consider for example +45 °C, the value of |R ∗M | for V 0–38 is around 6 and around 3.1 for C 0–80, which means that |G* | of V 0–38 mastic is 93% more higher than that of C 0–80 mastic. Figure 3 shows that the RG fillers increase the shear stiffness of mastics more than the limestone filler over the entire equivalent frequency range (RG mastics are in average 34% stiffer than limestone mastic over all f e ). The evolution of R ∗M for the three fillers seems uniform and varies between 5.5 and 8 for the RG fillers and between 3 and 6.5 for the limestone filler. At high temperatures (+45 °C), the filler Vx 0–38 has the highest value (|R ∗M | = 7), while the limestone filler has the lowest value (|R ∗M | = 3.1). The Vx 0–38 mastic is then 125% stiffer than the C 0–80 mastic when the equivalent frequency f e < 10−2 Hz. Furthermore, the phase angle ϕ R is close for all the mastics. It varies in a range of −4° to +4° over the entire equivalent frequency range. It should also be noted that the value of ϕ R is positive for the 3 mastics when
Fig. 3 Evolution of a the norm and b phase angle of R ∗M for the tests: RCA-PG 70 28-C-0 80-S-40, RCA-PG 70 28-V-0 38-S-40 and RCA-PG 70 28-Vx-0 38-S-40 at T = 15 °C
1694
M. M. Boussabnia et al.
the equivalent frequency is greater than 104 Hz, indicating a softening effect due to the addition of fillers at low temperature. Finally, the discontinuity in the R ∗M curve is due to the fact that the mastics don’t verify the time-temperature principle. Discussion In the literature, siliceous mineralogy is considered to have a lower affinity with bitumen than limestone [3]. This is due to the negative zeta power of silica aggregates and the negative nature of dominant polar features in bitumen. Silica aggregates are also sensitive to water which favors the stripping of the bituminous mixture [1]. However, our results show that RG fillers have a better stiffening power than limestone filler. The behaviors of mastics observed in this part can’t be explained by the difference in mineralogy alone. However, it has been noted in the literature [5] that the surface energy of glass is greater than that of limestone. Surface energy influences the amount of bitumen fixed by the filler (stiffening gradient) [5]. This effect can be seen on the phase angle curves since it is always the lowest in the case of glass, especially for the Verrox. This may partly explain the high stiffening power of RG fillers at high temperatures. The results obtained are also encouraging since the glass fillers performed better than the limestone filler.
3.2 Influence of the Gradation Size Curve The distribution of the particles size curve is studied by comparing spread size (S) RG fillers and tight size (T) RG fillers. The fillers used for making mastics are: V 0–160-S and V80–160-T. Figure 4 presents the evolution of R ∗M (in norm and in phase angle) for the materials of this part of the study. Figure 4 shows that the reinforcement of these mastics is close at low temperatures (f e varies in [101 , 105 ]). The norm of R M ∗ varies between 5.5 and 6.5 in this zone. At + 35 °C, mastic with V 80–160-T is 100% stiffer than the mastic with V 0–160-S. In
Fig. 4 Evolution of a the norm and b the phase angle of R M ∗ for the tests: RCA-PG 70 28-V-80 160-T-40 and RCA-PG 70 28-V-0 160-S-40 at T = 15 °C
Study of the Shear Stiffness of Bituminous …
1695
addition, the variation of the phase angle is small and remains less than 5° in absolute value. In general, it can be said that the V 80–160-T mastic performed better than the V 0–160-S mastic over the entire frequency range. Discussion The results of the influence of the gradation size curve on the stiffness of mastic show that the tight size gradation fillers improve the performance of the mastics better than the spread size fillers. The phase angle varies slightly for mastics and remains close to that of bitumen. Also, results revealed a close behavior at low temperature and a slightly higher stiffness in the case of the tight particle size. This result was also proposed by Lackner [6] who studied the influence of the mineralogy and the particle size of fillers on the behavior of mastics using a BBR (Bending Beam Rheometer). The variation of the phase angle seems to be not revealing and remains close to that of the bitumen. The author concluded that the particle size of fillers has a relatively low influence on the behavior of bituminous mastics at high frequencies [4]. Our results seem to confirm the results of the literature. At high temperature, a more important reinforcement can be noticed for the tight size gradation. It is possible that tight size gradation favors the creation of a small scattered granular skeleton. This can generate significant intergranular contacts in the mastic [3]. Nevertheless, the variation of rigidity remains weak to conclude.
4 Conclusions The results from this study of the shear stiffness of bituminous mastics using the annular shear rheometer have led to the following conclusions: (1) The reinforcement of mastics was close at low temperature. (2) The siliceous mineralogy of glass fillers demonstrated a greater affinity with PG 70–28 bitumen compared to limestone filler. (3) the gradation size curve is the most noticed factor regarding the reinforcement of the bitumen at high temperature. Tight size distribution lead to more reinforcement than well graded one. Overall, these results encourage the use of RG materials as filler additive in bituminous mixtures.
References 1. Lachance-Tremblay É, et al.: Linear Visco-Elastic (LVE) properties of asphalt mixtures with different glass aggregates and hydrated lime content. J. Mater. Civil Eng. (2018) 2. Huang, Y., Bird, R.N., Heidrich, O.: A review of the use of recycled solid waste materials in asphalt pavements. Resour. Conserv. Recycl. 52(1), 58–73 (2007) 3. Olard François, O.F.: Comportement thermomécanique des enrobés bitumineux à basses températures Relations entre les propriétés du liant et celle de l’enrobé. PhD, Institut National des Sciences Appliquées de Lyon, France (2003)
1696
M. M. Boussabnia et al.
4. Delaporte Brice, D.B.: Linear viscoelastic properties of bituminous materials: from binders to mastics. J. Assoc. Asphalt Paving Technol. 76, 455–494 (2007) 5. Lesueur Didier, L.D.: La Rhéologie des Bitumes: Principes et Modification. Rhéologie 2, 1–30 (2002) 6. Lackner, R., Spiegl, M., Blab, R., Eberhardsteiner, J.: Is low-temperature creep of asphalt mastic independent of filler shape and mineralogy—arguments from multiscale analysis. J. Mater. Civil Eng. (ASCE) 1561(Issue 5), 485–491 (2005)
Study on Aging Resistance of Bitumen Rejuvenated with Various Rejuvenators for Hot Recycling Chetan Kumara, Gurunath Guduru, Bharath Gottumukkala, and Kranthi Kuna
Abstract In this study, oxidized bitumen blend was rejuvenated with five rejuvenators, namely Tall oils, two fatty acids (FA 1 and FA 2), two paraffinic oil (PO 1 and PO 2). A relationship was developed between the viscosity of aged bitumen and target rejuvenator content for all the rejuvenator types considered in the study. The aging of bitumen was simulated by conditioning the VG40 bitumen in Rolling Thin Film Oven (RTFO) and Pressure Aging Vessel (PAV) for different conditioning periods. The relationship between viscosity of artificially aged bitumen and the target rejuvenator content was verified using recovered bitumen from Reclaimed Asphalt Pavement (RAP) material from two different sources. An error percent of 1–13% was observed in achieving target viscosity. To study the effectiveness of the rejuvenator w.r.t to the aging of rejuvenated bitumen blends, target rejuvenator content derived from the developed relationships was added to RAP bitumen blend, and the resulting bitumen blends were subjected to short term and long term aging. The aging indices were then calculated as the ratio of complex modulus at aged state to unaged complex modulus, which indicates the aging susceptibility of rejuvenated blends. The aged bitumen blended with Paraffin Oils category rejuvenator showed the least resistance to aging. Keywords Rejuvenator · Hot recycling · Aging resistance
1 Introduction Due to the exponential increase in the cost of the bitumen and the government’s policy to reduce the aggregate quarrying activities in a number of states in India, the road agencies are actively considering the use of high proportions of the Reclaimed Asphalt Pavement (RAP) material in the new hot mix asphalt (HMA) production. C. Kumara · G. Guduru · K. Kuna (B) Indian Institute of Technology, Kharagpur, India e-mail: [email protected] B. Gottumukkala Central Road Research Institute, New Delhi, India © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_215
1697
1698
C. Kumara et al.
Using high RAP content in HMA mixes may result in reduced crack resistance that often leads to premature pavement failures and consequently, early maintenance and increased rehabilitation costs [1]. To counter this increased cracking potential in high RAP asphalt mixes the most widely adopted approach is to use rejuvenators with or without softer grade bitumen [2]. Many recent studies identified various rejuvenators that have the potential to restore the rheological properties of aged bitumen which range from organic oils to petroleum-based rejuvenators. Based on the source and availability of these rejuvenators can be classified into five different categories [3]; Paraffinic Oils (PO), Aromatic Extracts, Napthenic Oils, Fatty Acids (FA) and Tall Oils. Description for each of the rejuvenator type can be found elsewhere [3]. Studies have shown that the addition of the rejuvenator significantly improves the fatigue resistance of the mixes with RAP [4, 5]. Research also shown that the addition of rejuvenators to mixes with RAP could improve the moisture susceptibility [6]. Due consideration should be also given to ageing resistance of the rejuvenated bitumen in the selection of the rejuvenator type as the ageing resistance of the rejuvenated bitumen depends on the type of rejuvenator [7]. However, limited studies focused on the effect of rejuvenator type on aging on the rejuvenated bitumen. Mazzoni et al. [8] studied the effect of rejuvenator type on the bitumen blend with R/T ratio of 0.5. The study was carried out on RAP bitumen blended with three commercial rejuvenators—two fatty acid categories and one tall oil, at dosages recommended by the manufacturer of the rejuvenator. It was found based on rheological studies that the rejuvenated bitumen is more aging resistant than the virgin bitumen. The dosage of a particular type of rejuvenator primarily depends on the properties of aged bitumen in the RAP. The dosage rates are chosen such that the physical properties of aged bitumen blends are restored to the properties of the target bitumen grades. Therefore, a single dosage specification without considering RAP characteristics is not possible. Using high doses of rejuvenator may soften the aged bitumen, which could be beneficial with respect to the cracking resistance of the HMA mix. However, that also may negatively impact the rut resistance of HMA mix. Therefore, the rejuvenator dosage should be optimized before used in mix production. Moreover, owing to the difference in the chemical properties, the target dosages vary for each rejuvenator type. This means that the dosage selection should be based on meeting at least one of the physical properties of bitumen such as penetration value, viscosity and softening point of the target bitumen grade. The studies that compared the efficiency of different rejuvenator types by way of aging resistance either adopted dosages recommended by the supplier without due consideration to RAP properties or un-optimized dosage without an acknowledgement to rejuvenator type. The presents study considered these shortcomings in the literature and aimed to compare the effectiveness of different rejuvenator types based on the aging index values on RAP bitumen blends with optimized dosages of rejuvenators. The approach adopted to determine optimum (or target) dosage of each rejuvenator was loosely based on the European Asphalt Pavement Association (EAPA) recommendations [9].
Study on Aging Resistance of Bitumen …
1699
2 Materials The virgin bitumen used in this study is Viscosity Grade VG40 and VG10 (IS 73: 2013) bitumen. The softening point of the VG40 bitumen tested in accordance with IS 1205 was found to be 54 ˚C while viscosity at 60 ˚C (IS:1206 (Part 2)) was found to be 399 Pa.s. Similarly, the softening point and viscosity of VG10 bitumen were found to be 43 ºC and 186 Pa.s, respectively. VG40 bitumen was used to produce the aged bitumen by conditioning it in RTFO and PAV. VG10 bitumen used as the virgin bitumen in the RAP bitumen blends. A total of five rejuvenators which can be categorized into three of five categories recommended by the National Center for Asphalt Technology (NCAT) [7] were used in the present study. Out of two paraffinic oils, one is Waste lubricating engine oil (PO 1/WEO) and other is a commercial product (PO 2) which is also waste oil engineered to use as a bitumen rejuvenator. All other rejuvenators (two Fatty Acids category and one Tall oils category) considered in the study are either by-products (Tallow (FA 1) and crude tall oil) or used oils (Waste Vegetable Oil—WVO/FA 2). Two RAP sources were used in the study. The recovered bitumen from RAP was tested for softening point value and viscosity. The absolute viscosity (at 60 ºC) of the recovered bitumen was found to be 13000 Pa.s and 7000 Pa.s respectively for Source 1 and Source 2. Due to equipment limitations, an approximate method was adopted to determine the absolute viscosity of bitumen recovered from RAP. Firstly, absolute viscosity of virgin bitumen and RAP blends with R/T ratio of 0.3, 0.5 and 0.65 was determined and plotted on viscosity blending chart with a viscosity (at 60 ºC) on the vertical axis (log scale) and R/T ratio on horizontal axis. A best-fit line connecting the points was determined and extrapolated. The ordinate value of the extrapolated line at R/T ratio of 1 (100% RAP bitumen) was adopted as RAP bitumen viscosity.
3 Methodology To achieve the objective of the study which is to evaluate the effectiveness of five different rejuvenators with respect to aging resistance of rejuvenated bitumen blends, the testing plan was divided into three phases—target dosage determination, verification of the relationship from first phase with the blends produced with recovered bitumen from RAP and aging resistance evaluation on RAP bitumen blends with rejuvenators. The optimum dosage of rejuvenator was determined based on viscosity and softening point values of the bitumen as recommended by EAPA. To simulate the aging of bitumen during the asphalt life cycle, four aging conditions were considered; standard RTFO conditioning i.e. for 85 min at 163 °C (RTFO), PAV aging of RTFO aged bitumen for 12 h (PAV—12 h), standard PAV aging of RTFO aged bitumen (PAV—20 h) and PAV aging of RTFO aged bitumen for 48 h (PAV—48 h). After aging bitumen to different aging conditions, rejuvenator was added to aged bitumen to restore its original properties of VG40. The blending of aged bitumen
1700
C. Kumara et al.
with rejuvenator was carried out using a blender. The rejuvenator dosage was varied for each aged bitumen condition until bitumen meets the target property. Before blending, aged bitumen was heated to 140 °C in a container, and then rejuvenator was added to aged bitumen and mixed it with a blender at 1000 rpm for 15 min. The rheological properties (G*) of the virgin, aged, and rejuvenated bitumen were determined using a DSR.
4 Target Dosage Relations Figure 1 presents absolute viscosity values (at 60 ºC) of base bitumen aged to different conditions with different dosages of Tallow. As expected, the viscosity of the aged bitumen decreased in all cases. The dashed lines in the figure indicates the target viscosity of base bitumen which is 399 Pa.s. The target dosage is the rejuvenator dosage at which the rejuvenated bitumen achieves target bitumen property. The target dosage for each aging condition was determined based on the regression equation (exponential relation) developed between rejuvenator dosage and viscosity that are presented in Fig. 1a. The effectiveness of the selected rejuvenators can also be evaluated based on the tests for physical properties on the rejuvenated bitumen. Figure 1b, which presents the decreasing trend of viscosity (from 3563 Pa.s) of the aged bitumen when blended with different rejuvenators depicts the effectiveness of the rejuvenators. As can be seen from the figure, both the Paraffinic oil based rejuvenators are less effective compared to other rejuvenators. To quantify the effectiveness, a term efficiency was used. Efficiency is the ratio of the percentage change in the viscosity if the target dosage of rejuvenator is added to the target dosage. The efficiency for PO 1, PO 2, FA 1, FA 2 and tall oil were found to be 5.12, 5.06, 8.45, 11.23 and 9.45 respectively when blended with aged bitumen with a viscosity of 3563 Pa.s.
b RTFO PAV-12 hrs. PAV-20 hrs. PAV-48 hrs. Target
4000 3000 2000 1000 0
0
10
Rejuvenator Dosage (%)
Viscosity @ 60°C (Pa.s)
Viscosity @ 60˚C (Pa.s)
a
Tall Oil PO1 PO2 FA1 FA2 Target
4000 3000 2000 1000 0
0
5
10
Rejuvenator Dosage (%)
Fig. 1 Change in viscosity of aged bitumen blends a for bitumen subjected to different aging conditions and blended with varying dosages of FA 1; b PAV aged bitumen blended with different rejuvenators and varying dosages
Fig. 2 Optimum (Target) dosages of rejuvenators required to achieve target base bitumen (VG40) viscosity for a given bitumen blend viscosity
Target dosage (%)
Study on Aging Resistance of Bitumen … 25
1701 PO 1 FA 2
20
PO 2 Tall Oils
FA 1
15 10 5 0
0
1000 2000 3000 4000 Viscosity of the bitumen blend (Pa.s)
5000
Figure 2 presents the relationships for dosage required to achieve the target property i.e. viscosity of 399 Pa.s for bitumen blend with certain R/T and viscosity derived from relationship plots similar to Fig. 1a. To verify the relations presented in Fig. 2, which were developed based on laboratory aged bitumen, bitumen blends were prepared from the bitumen extracted two RAP sources. Table 1 presents dosage of the rejuvenator obtained from Fig. 2 for RAP bitumen blends, absolute viscosity (at 60 ºC) obtained by blending the target dosage and % deviation (or error) from the target viscosity. The table also presents the corresponding R/T ratio and viscosity of RAP bitumen blends. The % error values from Table 1 suggests that the relations developed using the laboratory aged bitumen predicted dosage requirement close to RAP bitumen blends requirements.
5 Ageing Index To evaluate the effectiveness of the rejuvenator w.r.t to aging resistance of the bitumen blend aging index, which is the ratio of aged complex modulus values to unaged complex modulus values, was compared. Figure 3 presents, the ageing indices developed for both RTFO aged and PAV aged RAP bitumen blended (RAP source 1, R/T = 0.6, Viscosity @ 60 ˚C = 2450 Pa.s) with VG10 virgin bitumen and various rejuvenators. The target dosages for the blends are as in Table 1. As can be seen from the figure, all rejuvenated bitumen blends are more sensitive to aging compared to base bitumen. PO 1 (WEO) found to be most sensitive to the ageing compared to other rejuvenators. Accelerated ageing of the PO 1 bitumen blend may be attributed to high volatile compounds that are usually present in petroleum-derived products. The other Paraffin Oil (PO 2), as mentioned is a commercial rejuvenator which is used lubricant oil engineered to function as rejuvenator by adding a small dosage of additives that improve properties such as viscosity and stability. These additives might have limited the volatility of the compounds in PO 2 and improved ageing resistance. Relatively higher ageing index values for FA 1 (Tallow) blend compared to FA 2 blend is not in the expected lines. It is expected that the FA 2 (WVO) blend would result in more ageing because of the presence of a high concentration of unsaturated fatty acids which are more reactive than saturated fatty
13.3
13.9
8.1
7.2
7.2
PO 1
PO 2
FA 1
FA 2
Tall oil
345
420
390
381
415
13.8
5.0
2.5
4.8
3.8
9.4
5.3
5.4
5.9
10.7
392
411
380
359
377
Viscosity (Pa.s)
R/T = 0.6 (η = 1695 Pa.s) Dosage (%)
Dosage (%)
% error
R/T = 0.6 (η = 2540 Pa.s)
R/T
Viscosity (Pa.s)
RAP 2 (η = 7000 Pa.s)
RAP 1 (η = 13000 Pa.s)
Source
Table 1 Percentage deviation from target base bitumen viscosity
2.0
2.8
5.0
10.3
5.8
% error
6.9
6.9
7.7
13.3
12.6
Dosage (%)
396
390
381
382
407
Viscosity (Pa.s)
R/T = 0.7 (η = 2390 Pa.s)
RAP 2 (η = 7000 Pa.s)
1.0
2.5
4.8
4.5
1.8
% error
1702 C. Kumara et al.
Study on Aging Resistance of Bitumen …
1703 Tall oils
Tall oils
Aging index (ǀG*ǀ RTFO / ǀG*ǀUnaged)
FA 2
Aging index (ǀG*ǀ PAV / ǀG*ǀ Unaged)
FA 2
FA 1
FA 1
PO-2
PO-2
PO-1
PO-1 Base bitumen
Base bitumen 0
5 Aging index
10
0
10 20 Aging index
30
Fig. 3 Aging indices (left—RTFO aged, right—PAV aged) of RAP bitumen blends rejuvenated with different rejuvenators
acids which are present in more proportion in Tallow (FA 1). Bitumen blend with crude tall oil has also shown relatively good ageing resistance.
6 Conclusions Tests for physical properties on aged bitumen with different dosages of rejuvenator suggest that the Tall Oil and Fatty Acid categories are more effective in softening the aged bitumen compared to Paraffinic Oils. For the same reason, the dosage required to achieve target physical property if Paraffinic Oils are used is more. It was also noted that that the relationships developed to predict the target dosages based on aged bitumen blends using laboratory aged bitumen can be successfully used to predict target dosage for RAP bitumen blends with known viscosity. It was found that the RAP bitumen blends rejuvenated with WEO are least resistant to aging compared to other rejuvenated bitumen blends. In general, rejuvenated bitumen found to be more sensitive to aging compared to virgin bitumen.
References 1. McDaniel, R., Michael Anderson, R.: NCHRP REPORT 452: Recommended Use of Reclaimed Asphalt Pavement in the Superpave Mix Design Method: Technician’s Manual Transportation (2001) 2. Haghshenas, H.F., Kim, Y.R., Kommidi, S.R., Nguyen, D., Haghshenas, D.F., Morton, M.D.: Evaluation of long-term effects of rejuvenation on reclaimed binder properties based on chemical-rheological tests and analyses. Mater. Struct. Constr. (2018). https://doi.org/10.1617/ s11527-018-1262-4 3. NCAT: NCAT researchers explore multiple uses of rejuvenators. Asph. Technol. News. 26 (2014) 4. Zaumanis, M., Mallick, R.B., Poulikakos, L., Frank, R.: Influence of six rejuvenators on the performance properties of Reclaimed Asphalt Pavement (RAP) binder and 100% recycled asphalt mixtures. Comput. Chem. Eng. (2014). https://doi.org/10.1016/j.conbuildmat.2014. 08.073
1704
C. Kumara et al.
5. Tran, N.H., Taylor, A., Wills, R.: Effect of Rejuvenator on Performance Properties of HMA Mixtures with High RAP and RAS Contents (2012) 6. Zaumanis, M., Mallick, R.B., Frank, R.: Determining optimum rejuvenator dose for asphalt recycling based on superpave performance grade specifications. Constr. Build. Mater. (2014). https://doi.org/10.1016/j.conbuildmat.2014.07.035 7. Cavalli, M.C., Zaumanis, M., Mazza, E., Partl, M.N., Poulikakos, L.D.: Effect of ageing on the mechanical and chemical properties of binder from RAP treated with bio-based rejuvenators. Compos. Part B Eng. (2018). https://doi.org/10.1016/j.compositesb.2017.12.060 8. Mazzoni, G., Bocci, E., Canestrari, F.: Influence of rejuvenators on bitumen ageing in hot recycled asphalt mixtures. J. Traffic Transp. Eng. (English Ed. (2018). doi: https://doi.org/10. 1016/j.jtte.2018.01.001 9. EAPA: Recommendations for the use of rejuvenators in hot and warm asphalt production. Belgium (2018)
Study on Reversibility of Aging and Recycling of Bitumen Based on Rheology, Adhesion and Chemical Properties Meng Guo, Haiqing Liu, Yubo Jiao, and Yiqiu Tan
Abstract The reversibility of aging and recycling of bitumen directly influences the recycling efficiency and durability of recycled pavement. Rolling thin film oven (RTFO) and pressure aging vessel (PAV) were conducted to simulate the different degrees of aging process. The rheology, adhesion and chemical properties of bitumen samples at different aging and recycling stages were characterized by dynamic shear rheometer (DSR), contact angle test and Fourier transform infrared spectroscopy (FTIR). The results reveal that the complex modulus of short-term aged bitumen can be completely recovered to the level of virgin bitumen. The complex modulus of long-term aged bitumen can also be recovered, but higher than that of virgin bitumen. Aging reduced the adhesion between bitumen and limestone. The anti-stripping agent can increase the adhesion of both short-term aged bitumen and long-term aged bitumen by 19.0% and 44.8%, respectively. The carbonyl index and sulfoxide index increased by 3614.6% and 441.9% after long-term aging, and decreased by 54.0% and 40.5% after being recycled, respectively. The revelation of reversibility of aging and recycling of bitumen can help to understand the mechanism of aging and recycling and improve the recycling technology. Keywords Bitumen · Aging · Recycling · Complex modulus · Adhesion · FTIR
1 Introduction Recycling of aged bitumen has become a hot issue in road industry. The rheological and adhesive properties of aged and recycled bitumen have great impacts on the M. Guo · H. Liu · Y. Jiao (B) Beijing University of Technology, Beijing, China e-mail: [email protected] H. Liu e-mail: [email protected] Y. Tan Harbin Institute of Technology, Harbin, China e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_216
1705
1706
M. Guo et al.
performances of recycled asphalt mixture, which needs to be studied. Gómez et al. found that the complex modulus of bitumen increased, the phase angle decreased, and the mechanical properties of bitumen were close to that of solid with the increase of aging time [1]. Zhou et al. found that the logarithmic values of complex modulus and rutting factor had a good linear decreasing relationship with the content of rejuvenator [2]. Yi et al. found aging played a significant role in the reduction of adhesive force and surface energy of bitumen, and weaken the adhesion between bitumen and aggregate [3]. Furthermore, studying on the micro-changes in aging process is helpful to reveal the mechanism of recycling. Traxler et al. found that the main causes of bitumen aging were volatilization of light components, oxidation, internal structure changes, and increase of strong polar compounds [4]. Lamontagne et al. demonstrated that the active groups in the molecular chain would combine with oxygen molecules in the air and form polar molecules containing carbonyl through chemical reaction [5]. In this study, the extract oil was applied as the rejuvenator in order to restore the rheological property of aged bitumen. Considering that the adhesion property of aged bitumen had decreased, anti-stripping agent was added to the rejuvenator to improve the adhesion ability between bitumen and aggregates. Moreover, FTIR was used to characterize the microstructure of the aged bitumen before and after recycling.
2 Materials and Methods 2.1 Materials 70# bitumen was used in this experiment. The base oil of rejuvenator was extract oil. Another component was anti-stripping agent. Limestone, distilled water, glycerin and formamide were used in contact angle test.
2.2 Methods Laboratory aging of bitumen. Rolling thin film oven (RTFO) and pressure aging vessel (PAV) were conducted to simulate the short-term and long-term aging processes according to ASTM D2872 and ASTM D6521, respectively. Sample preparation. The short-term aged bitumen and the long-term aged one were both added with 5wt% extract oil and 0.3wt% anti-stripping agent. The samples with only 5wt% extract oil were also prepared. The above samples were stirred artificially with glass rods for 5 min at 130°C.
Study on Reversibility of Aging and Recycling of Bitumen …
1707
DSR test. Oscillation sweep tests were carried out at 25°C, 35°C, 45°C, 55°C and 65°C, respectively. The strain was set to be 0.015% and the frequency was 0.01Hz100Hz. The master curve of the complex modulus of bitumen was drawn according to the time temperature superposition principle. Contact angle test. Dataphysics-TP50 contact angle analyzer made in Germany was used in this experiment. Samples in each group were measured 3–5 times with an average temperature of 25°C. According to literature [6–9], the adhesion work between bitumen and aggregate was calculated. FTIR test. FTIR-ATR was used. The Origin and OMNIC software were used to analyze the spectra. Aging of bitumen in this experiment was mainly oxidation reaction, so the changes of carbonyl group (1700cm−1 ) and sulfoxide group (1030cm−1 ) were mainly considered. In order to eliminate the difference of peak area caused by different sample size, carbonyl index (I C=O ) and sulfoxide index (I S=O ) were used to quantitatively characterize the aging degrees of aged bitumen and recycled bitumen. The area of saturated C-H bond bending vibration absorption peak (1460-1376cm−1 ) was used as reference area [10].
3 Results and Discussions 3.1 Effect of Aging and Recycling on Rheological Properties of Bitumen Fig. 1a reveals that the complex modulus of bitumen increases during aging. Fig. 1b shows that the complex modulus of the short-term recycled bitumen decreases significantly at various frequencies, which is about 50% lower than that of the short-term aged bitumen. It indicates that the softening effect of rejuvenator on short-term aged bitumen is significant. Fig. 1c shows that the complex modulus of the long-term bitumen also decreases greatly at all frequencies, which is about three times lower than that of the long-term aging, but still higher than that of the virgin bitumen. It indicates that this rejuvenator has limited effect on the recovery of complex modulus of long-term aged bitumen.
3.2 Analysis of Adhesion Work Between Bitumen and Aggregate Fig. 2 shows that the adhesion work between virgin bitumen and limestone decreases by 0.21 mJ/m2 after short-term aging. After long-term aging, the adhesion work decreases by 8.56 mJ/m2 . Aging reduces the adhesion between bitumen and limestone. After adding rejuvenator containing only extract oil, the adhesion work
1708
M. Guo et al.
Fig. 1 Master curves of complex modulus for different samples: a before and after aging; b before and after recycling for short-term aging; c before and after recycling for long-term aging
Fig. 2 Adhesion work between bitumen and limestone (mJ/m2 )
increases by 0.57 mJ/m2 and 6.58 mJ/m2 , respectively. After adding rejuvenator containing both extract oil and anti-stripping agent, the increments of adhesion work for short-term and long-term aged bitumen increase by 7.7 mJ/m2 and 14.43 mJ/m2 , respectively. Extract oil and anti-stripping agent can improve adhesion work.
Study on Reversibility of Aging and Recycling of Bitumen …
1709
Fig. 3 FTIR spectra of virgin, long-term aged and recycled long-term aged bitumen
Table 1 Carbonyl index, sulfoxide index and their change rate Items
I C=O
I S=O
W C=O
Virgin bitumen
0.0031
0.0144
–
Long-term aged
0.1160
0.0781
3614.6%
Recycled long-term aged
0.0534
0.0465
− 54%
W S=O – 441.9% − 40.5%
3.3 The Results of FTIR Fig. 3 shows that the peaks around 1700 and 1030 cm−1 increase significantly after aging. This is because carbon and sulfur reacted with oxygen to produce a large number of carbonyl and sulfoxide groups. The peaks decrease after adding rejuvenator, which may be due to the dilution of rejuvenator. Table 1 shows that I C=O and I S=O of bitumen increase by 3614.6% and 441.9% respectively after long-term aging. After recycling, I C=O and I S=O decrease by 54% and 40.5%, respectively.
4 Conclusions The complex modulus, adhesion work and functional group concentration of bitumen are reversible in the process of aging and recycling. (1)
(2)
The complex modulus of bitumen increases with aging, and decreases with recycling. The effect of rejuvenator is more obvious for short-term aged bitumen. The adhesion between bitumen and limestone decreases after aging. The rejuvenator can improve the adhesion, especially when the rejuvenator contains anti-stripping agent.
1710
(3)
M. Guo et al.
A lot of carbonyl and sulfoxide groups are produced after aging. After adding rejuvenator, the functional groups related to aging decreased and the aged bitumen get certain degrees of recycling.
References 1. Fernández-Gómez, W., Quintana, H., Lizcano, F.: A review of asphalt and asphalt mixture aging: Una revisión. Ingenieria e Investigación 33(1), 5–12 (2013) 2. Zhou, Z., Yang, Y., Zhang, Q.: Recycling behavior of recycling agent on aged asphalt. J. Traffic Transp. Eng. 11(6), 10–16 (2011) 3. Yi, J., Pang, X., Feng, D.: Studies on surface energy of asphalt and aggregate at different scales and bonding property of asphalt–aggregate system. Road Mater. Pavement Design 1–24 (2017) 4. Traxler, R.N.: Relation Between Asphalt Compsition and Hardening by Volatilization and Oxidation. Proc. Assoc. Asphalt Paving Technol. 30, 359–377 (1961) 5. Lamontagne, J., Dumas, P., Mouillet, V.: Comparison by fourier transform infrared (FTIR) spectroscopy of different ageing techniques: Application to road bitumens. Fuel 80, 483–488 (2011) 6. Good, R.J.: Contact angle, wetting, and adhesion: A critical review. J. Adhes. Sci. Technol. 6(12), 1269–1302 (1993) 7. Van Oss, C.J., Good, R.J., Chaudhury, M.K.: Additive and nonadditive surface tension components and the interpretation of contact angles. Langmuir 4(4), 884–891 (1988) 8. Owens, D.K., Wendt, R.C.: Estimation of the surface free energy of polymers. J. Appl. Polym. Sci. 13(8), 1741–1747 (1969) 9. Hefer, A.W., Hefer, A.W., Herbert, B.E.: Bitumen surface energy characterization byqa inverse gas chromatography. J. Test. Eval. 35(3), 233–239 (2006) 10. Hofko, B., Porot, L., Falchetto, C.A.: FTIR spectral analysis of bituminous binders: Reproducibility and impact of ageing temperature. Mater. Struct. 51(2), 45 (2018)
Superpave 5: Improving Asphalt Mixture Performance Reyhaneh Rahbar-Rastegar, M. Reza Pouranian, Dario Batioja-Alvarez, Mohammad Ali Notani, Miguel Montoya, and John E. Haddock
Abstract The traditional Superpave mixture design method selects optimum asphalt binder content to yield 4% air voids content in asphalt mixtures, while anticipating approximately 7% air voids content at the time of construction. This approach can produce mixtures with high ageing and moisture damage potential. A recently developed mixture design method named “Superpave 5” is nearly identical to the traditional Superpave mixture design method, except that optimum binder content is selected to yield 5% air voids content with the expectation that when field compacted, the mixture will also yield 5% air voids content. The objective of this study is to assess the viscoelastic, cracking, and rutting properties of Superpave 5 mixtures and compare them with a traditionally designed Superpave mixture. Four plant-mixed Superpave 5 and one laboratory-mixed traditional Superpave mixtures were evaluated using complex modulus testing, the Illinois Flexibility Test, and the Hamburg Wheel Track Test. The results indicate that Superpave 5 mixtures exhibit improved laboratory rutting performance. However, the Flexibility Index does not seem to be an appropriate cracking indicator, as it appears unable to distinguish between air voids contents. Keywords Superpave 5 · Performance · Viscoelastic · Fracture · Rutting
1 Introduction Asphalt mixtures in the United States are mostly designed using the Superpave mixture design method, which selects optimum binder content based on 4% air voids content. When placed in the field such mixtures are typically compacted to approximately 7–8% air voids content, 92–93% of the theoretical maximum mixture density (Gmm ). The idea of field compacting mixtures to air voids contents twice the design air voids content not only seems odd, but it can produce asphalt pavements that are susceptible to moisture damage and oxidative aging. A better method would be R. Rahbar-Rastegar (B) · M. R. Pouranian · D. Batioja-Alvarez · M. A. Notani · M. Montoya · J. E. Haddock Purdue University, West Lafayette, IN, USA e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_217
1711
1712
R. Rahbar-Rastegar et al.
to design and place asphalt mixtures at the same air voids content, providing the method can help achieve higher field densities (lower air voids). Such a method was initiated in 1970 by Laboratoire Centrale de Ponts et Chausees (LCPC) in France [1]. Recently, the Superpave mixture design method has been altered slightly, based on this idea of equivalent laboratory and field air voids content, and is referred to as “Superpave 5” [2]. The Superpave 5 mixture design method selects optimum binder content based on 5% air voids content. Such mixtures are expected to achieve 5% air voids content (95% of Gmm density) when field compacted. To-date, there is limited research investigating the performance of Superpave 5 mixtures. Huber et al. (2019) compared the ageing level, cracking, and rutting performance of mixtures designed by Superpave 5 and constructed in a 2013 trial project to traditional Superpave-designed control section mixtures after 5 years in service. The performance grade and Tc of extracted and recovered binders showed a lower ageing for Superpave 5 mixtures. Additionally, visual field inspection revealed that Superpave 5 mixtures had a better performance in both cracking and rutting [3]. In another study, the cracking potential of Superpave 5 was explored using the Illinois Flexibility Index Test (I-FIT). The results show an increase in fracture energy and strength parameters for Superpave 5 mixtures, but a reduction in the Flexibility Index (FI) [4]. In 2018 the Indiana Department of Transportation (INDOT) placed several Superpave 5 mixtures, scattered throughout the six INDOT districts [5]. As part of these trials, virgin aggregates, virgin binder, recycled asphalt pavement (RAP), and plantmixed asphalt mixture were collected during construction. The objective of the study reported herein is to evaluate the viscoelastic, rutting, and cracking characteristics of the Superpave 5-designed asphalt mixtures from these trials and compare them to the properties of a conventionally designed Superpave mixture.
2 Mixtures and Materials Five asphalt mixtures, four Superpave 5 and one traditional Superpave mixture were used in this study. The plant-mixed mixtures were collected during asphalt production from the INDOT Fort Wayne (FW), Crawfordsville (CR), Vincennes (VN), and Greenfield (GR) districts. The 9.5- and 12.5-mm mixtures were placed in pavement surface layer, while the 19.0-mm mixture was used for the base layer. Superpave 5 laboratory specimens were fabricated from the plant-mixed asphalt mixtures, while the raw materials from the Greenfield project were used to design and fabricate the corresponding conventional Superpave (control) specimens. To produce the Superpave 5 specimens in the laboratory, the plant-mixed materials were placed in oven at the mixing temperature for 2 h immediately prior to compaction. The Superpave 5 specimens were fabricated using 50 gyrations of a Superpave Gyratory Compactor (SGC) to obtain 5 ± 0.5% air voids content. The control mixture was laboratory-mixed, laboratory-compacted (LMLC) and the test
Superpave 5: Improving Asphalt Mixture Performance
1713
Table 1 Mixture data Mixture ID
District
Binder Type
Binder Content
Control
Greenfield
PG 64–22
5.4%
GR
Greenfield
PG 64–22
FW
Fort Wayne
CR VN
VFA (%)
Mixture Type
Binder replacement (%)
Average Air Voids, %
–
19.0
11.8
6.8
5.4%
65.3
19.0
15.4
5.1
PG 64–22
6.5%
70.2
9.5
13.8
5.1
Crawfordsville
PG 64–22
6.1%
68.7
9.5
24.3
5.2
Vincennes
PG 64–22
6.2%
68.5
12.5
20.1
5.1
Fig. 1 Aggregate gradation of different mixtures
specimens were fabricated to have 7 ± 0.5% air voids content, as is done in conventional asphalt mixture testing. The mixtures information is summarized in Table 1, while Fig. 1 shows the aggregate gradations of the mixtures.
3 Test Methods Complex modulus testing was used to determine the viscoelastic properties of asphalt mixtures. Testing was performed on three replicates from each mixture using an Asphalt Mixture Performance Tester (AMPT), and following AASHTO TP79 protocol. The dynamic modulus (E* ) and phase angle data were then analyzed and
1714
R. Rahbar-Rastegar et al.
the master curves developed using the time-temperature superposition principle to evaluate the stiffness and relaxation capability of the asphalt mixtures. To characterize the mixture cracking properties, the I-FIT was conducted in accordance with AASHTO TP105. The cracking resistance of three replicates from each mixture was determined at the intermediate temperature of 25°C. The test data were analyzed using I-FIT software and the Fracture Energy (Gf ) and Flexibility Index (FI) parameters compared. Gf is calculated from the fracture work, the area under the load-displacement curve, and the FI is defined as the fracture energy divided by the failure slope [6]. Mixture rutting behavior was assessed using the Hamburg Wheel Track Tester (HWTT) at 50 °C, according to AASHTO T324. Two, 150 mm diameter specimens from each mixture were subjected to cyclic loading up to 20,000 passes, or a rut depth of 12.5 mm, whichever came first.
4 Results and Discussion Figure 2a shows the average dynamic modulus master curves for the five mixtures. The FW mixture has the highest stiffness, followed by CR and VN mixtures. Despite having approximately two percent difference in air voids content, there is not a significant difference between the dynamic moduli of the Superpave 5 and traditional Superpave GR mixtures. This could be due to different ageing outcomes, as the Superpave 5 mixture was plant produced and the control mixture laboratory produced. Black space diagrams for the mixtures are shown in Fig. 2b to capture the mixtures’ stiffness and relaxation capabilities. Generally, with increasing dynamic modulus and decreasing phase angle, mixtures tend to be more prone to cracking. The VN mixture shows lower phase angles compared to the other mixtures, especially at intermediate and high temperatures. The CR mixture’s phase angles indicate better cracking resistance than the other mixtures. Finally, the GR Superpave 5 and control mixtures have similar phase angles, indicating similar viscoelastic properties.
Fig. 2 a Dynamic modulus master curves, b Black space diagram
Superpave 5: Improving Asphalt Mixture Performance
1715
Fig. 3 Load-displacement curve comparisons
The fracture behavior of the asphalt mixtures is exemplified by the two typical load-displacement curves obtained from the GR I-FIT test results, as shown in Fig. 3. As can be seen, the Superpave 5 curve has a higher peak load, while the curve slope after peak load (failure slope) is much deeper. This trend applies to all Superpave 5 mixtures, resulting in very low FI values, compared to the control mixture. Figure 4a shows the mixtures’ peak load, and Fig. 4b represents the fracture energy and flexibility index values. The peak load values are higher for all Superpave 5 mixtures. Specifically, the peak load of the GR Superpave 5 mixture increased 32% in comparison to the peak load of control mixture. However, the two mixtures do not have significantly different fracture energies. Steep failure slopes (−17.8 to − 55.9) were observed for Superpave 5 mixtures, resulting in unreasonably low FI values (0.14 to 1.4). This phenomenon has been reported in other studies where a lower FI was found for stiffer mixtures having lower air voids contents [3, 7]. The finding does not agree with the field observations that indicate better cracking performance for the asphalt mixtures with lower air voids contents [2]. It appears the FI parameter is not able to distinguish the effect of air voids content. Figure 5a shows the rut depth variation during the HWTT. The final rut depths (mm) of different mixtures after 20,000 wheel passes are compared in Fig. 5b. All the mixtures meet the 12.5 mm maximum rut depth criteria. Generally, the Superpave 5 mixtures show better rut resistance. The CR, FW, and VN mixtures have finer aggregate structures than the control mixture, but experienced less rutting, which is to be expected, since these mixtures also have higher stiffnesses (dynamic modulus) compared to the control mixture. The GR Superpave 5 mixture has slight better
Fig. 4 Fracture properties: a Peak load, b Fracture energy and Flexibility Index
1716
R. Rahbar-Rastegar et al.
Fig. 5 Rutting properties of mixtures: a rutting Curves, b rut depth
rutting resistance, about 20%, than does the control mixture. The data appear to indicate the Superpave 5 mixture did not improve the stripping inflection point (SIP), an indication a mixture may be susceptible to moisture damage.
5 Summary and Conclusion The main objective of this study was to evaluate the viscoelastic, rutting, and cracking characteristics of Superpave 5 asphalt mixtures using laboratory tests. The viscoelastic properties were evaluated using AMPT complex modulus testing. In addition, I-FIT and HWTT testing were used to evaluate mixture cracking and rutting potential. Given the test results and analyses, the following conclusions are drawn: – The viscoelastic properties (dynamic modulus and phase angle) of the Superpave 5 mixtures approximate those of the conventional Superpave mixture. However, the Superpave 5 and conventional mixtures’ cracking and rutting behaviors are different. This suggests mixtures with similar viscoelastic properties do not necessarily have similar laboratory test performance. – The I-FIT results indicate that asphalt mixtures designed and compacted at 5% air voids content have improved peak loads and caused no significant difference in fracture energies, compared to the control mixture. However, the FI is not recommended as a cracking indicator for Superpave 5 mixtures, as the parameter appears unable to distinguish between air voids contents. – While Superpave 5 mixtures appear to have better rutting resistance than do conventional Superpave mixtures, Superpave 5 mixtures did not show improved moisture damage characteristics. Therefore, while Superpave 5 mixtures are recommended for improving rut resistance, such mixtures should be tested for moisture susceptibility. Since all five Superpave 5 mixtures have been placed into service during 2019, their in-service cracking and rutting performance can be monitored over time and
Superpave 5: Improving Asphalt Mixture Performance
1717
additional testing conducted to compare the laboratory test results with actual inservice performance.
References 1. LPC Bituminous Mixtures Design Guide, Laboratoire Central des Ponts et Chaussees (2007) 2. Hekmatfar, A., Shah, A., Huber, G., McDaniel, R., Haddock, J. E.: Modifying laboratory mixtures design to improve field compaction. Road Mater. Pavement Design 16-S2, 149–167 (2015) 3. Huber, G., Wielinski, J., Campbell, C., Padgett, J., Rowe, G., Beeson, M., Cho., S.: Superpave 5: Relationship of In-Place Air Voids and Asphalt Binder Aging, Presented at the Association of Asphalt Pavement Technologies (AAPT) Annual Meeting (2019) 4. Ali, U. M.: Evaluating the Cracking Resistance of a Superpave 5 mix and a Conventional Superpave Mix using the Illinois Flexibility Index Test, University of Illinois, IL 5. Haddock, J. E. et al.: Implementation the Superpave 5 Asphalt mixture design method in Indiana. Joint Transportation Research Board SPR-4211, FHWA/IN/JTRP-2020/12 (2020) 6. Al-Qadi, I., et al.: Testing Protocols to Ensure Performance of High Asphalt Binder Replacement Mixes Using RAP and RAS, Report No. FHWA-ICT-15–017. Illinois Center for Transportation, Rantoul, IL (2015) 7. Batioja-Alvarez, D., Lee, J., Haddock, J.: Understanding the Illinois Flexibility Index Test (I-FIT) using Indiana asphalt mixtures. Transp. Res. Record J. Transp. Res. Board 2673(6), 337–346 (2019)
Surface Dressing Treatment for Applications on Solar Roads Domenico Vizzari, Eric Gennesseaux, Stéphane Lavaud, Stéphane Bouron, and Emmanuel Chailleux
Abstract The solar road concept is a pavement system able to harvest energy from the solar radiation. This paper is part of a wide project for the design of a multilayer system able to harvest energy through the road. In this technology, the top layer plays a fundamental role because it has to support the traffic load, guarantee the vehicle friction and protect the solar cells. For this purpose, the authors proposed a novel approach for the construction of a semi-transparent layer made of glass aggregates bonded together through a transparent polyurethane, inspired from surface dressings. The procedure consists on spreading a first layer of polyurethane binder on the solar cells in order to obtain a uniform surface. Once the polymerization is completed, a second layer of binder is laid down and the aggregates are spread and compacted immediately. For the samples manufacturing, three different thermosetting polyurethanes and a 2/4 mm fraction of glass aggregates were used. The samples were compared in terms of optical and mechanical performances. The authors studied also the packing density of the aggregates in order to quantify the optimum amount of binder. The results show the potential of this novel top layer in terms of optical and mechanical performances, demonstrating the feasibility of the surface dressing treatment for applications in full scale solar roads. Keywords Solar road · Surface dressing · Packing density · Polyurethane
1 Introduction This paper is part of a wide project for the design of a prototype able to harvest the sunlight radiation through the road. The system proposed is a multi-layer pavement composed by: (i) a semi-transparent top layer made of glass aggregates bonded together through a transparent polyurethane; (ii) an electrical layer containing the solar cells; (iii) a porous concrete layer and (iv) a waterproof base layer [1].
D. Vizzari (B) · E. Gennesseaux · S. Lavaud · S. Bouron · E. Chailleux IFSTTAR, Nantes, France e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_218
1719
1720
D. Vizzari et al.
The top surface plays a fundamental role because it has to support the traffic load, guarantee the vehicles friction, protect the solar cells, allow the passage of the sunlight and maximize the heat transfer to the lower layers. Unfortunately, the electrical power generated by the solar cells is strongly dependent on the top layer optical properties. More in detail, binder transparency, glass aggregates shape and size, the presence of dirt and dust and the thickness of the layer can dramatically reduce the electrical output up to 80%. This sharp reduction is due to the light absorption of the material and to the diffusive reflection of the light [2], scattered in all the directions because of the irregular shape of the aggregates. For this reason, the authors focused on the reduction of the thickness, introducing a novel approach for the manufacture of the semi-transparent layer based on the surface dressing technique. The material used is recycled glass, crushed and selected according to the volume fractions of 4/6, 2/4 and 1/2 mm, and three thermosetting polyurethanes, characterized by high transparency and mechanical behavior similar to the bitumen at room temperature.
2 The Surface Dressing Treatment The surface dressing consists in a thin film of binder, which is spread on the road surface and covered with a layer of stone chippings [3]. Taking inspiration from this type of treatment, the authors proposed the following procedure for the manufacture of the semi-transparent layer: (i) stick the solar cells on the concrete mortar; (ii) lay down a first layer of glue, in order to cover the solar cells and protect it from punching, and have an uniform surface; (iii) once the first layer of glue is totally polymerized, lay down a second thin layer of glue and (iv) immediately spread out the glass aggregates and compact it to ensure a penetration of the aggregate into the fluid layer. The compaction is very low and it is carried out manually in the laboratory. The success of this procedure is affected by the amount of glue used for the second layer. The aggregates have not to be totally immerged into the binder, in order to provide skid resistance to the surface. For this reason, the packing density of the surface dressing was studied and presented in this paper using different models/techniques. To spread the glass aggregates, a special device was developed. It is composed by a tank containing the glass aggregates and a fluted roller. Pushing forward the device, the aggregates fall from the tank to the roller and are then pushed down on the surface. The result is a single layer of aggregates spread out uniformly (Fig. 1). The repeatability of the device was evaluated in terms of surface density of the aggregates. The idea was to weight the material that fall down within a delimited area (A = 23.8*19.4 cm2 ) and repeat the test 5 times for each glass fraction. The surface density is given by the ratio W/A, where W is the weight of the glass aggregates fallen into the area A. The results are summarized in the Table 1.
Surface Dressing Treatment for Applications on Solar Roads
1721
Table 1 Aggregate spreading density for different glass fractions. Legend: 1st, 2nd... 5th = i-th measurement of surface density; μ = average; σ = standard deviation; CV = coefficient of variation Surface density of the glass aggregates (kg/m2 ) 1st
2nd
3rd
4th
5th
μ
CV (%)
1/2 mm
1.76
1.52
1.51
1.62
1.58
1.6
0.1
6.30
2/4 mm
1.98
2.14
1.95
2.04
1.82
1.99
0.12
5.90
4/6 mm
2.95
3.41
3.11
3.11
3.32
3.18
0.19
5.90
The results show a good repeatability with a coefficient of variation around 6% and, as expected, the surface density increases according to the size of the glass aggregates.
3 Determination of the Optimum Glue Volume A relevant question for the manufacture of the surface dressing is to calculate the exact necessary amount of binder, paying attention to the embedment of the aggregates. Based on the literature [4] for bituminous surface dressings, an approach is to achieve a 50% aggregates immersion after rolling and finishing the embedment to the vehicles. Alternatively, the aggregates immersion could be up to 70%. In this case, the pull out of aggregates is reduced, but the road could be prone to bleeding [5]. Based on the French specifications, the amount of binder for a single layer of surface dressing (using the aggregate fraction 2/4mm) ranges between 0.8 and 1 kg/m2 [6]. It is important to ensure that those specifications are applicable and realistic for a semi-transparent layer made of glass aggregates and glue. The objective is to calculate the exact amount of binder needed to fill the voids and to have the aggregate immersion of 70%. At this scope, the packing density f of the glass aggregates and, consequently, the porosity p, has been measured according to three different methods/techniques: i) the circle/sphere packing model, ii) the de Larrard model and iii) the technique of image correlation. (i)
(ii)
A first approach was to consider the aggregates as perfect spheres and study their arrangement, with the intention of maximizing the occupied space. The spheres (or circles) can be arranged so that the center of each sphere forms a symmetric and periodic pattern [7]. This type of ordered arrangement is called lattice. In two-dimensional space (R2 ), the highest packing density is given for an hexagonal lattice (0.907), while for a square lattice R2 is equal to 0.785 [8]. In three-dimensional space (R3 ), moving from circles to spheres, the packing density is 0.605 and 0.524 respectively [9]. The packing density calculation based on de Larrard Model [10] was performed using the software BetonLabPro. The input are the grading curve of the fraction 2/4 mm, the packing density measured experimentally, the
1722
(iii)
D. Vizzari et al.
packing index defining the energy of compaction used to measure the packing density (here K = 4.75 corresponding to a vibration on a vibrating table) and the volume of the surface dressing (4 × 1000 × 1000 mm3 ). The digital image correlation consisted on taking a picture of the glass aggregates spread out in a delimited area. The picture, taken perpendicularly to the surface, is a 2D projection of the glass aggregates on the surface. For a better result, the glass aggregates has been painted in black and the picture was filtered in order to increase the contrast. Finally the packing density was measured using the software ImageJ.
Knowing the porosity, it’s possible to measure the amount of glue required to fill all the voids. Assuming: (i) an average density ρ of the glue of 1.1 gr/cm3 , (ii) an aggregate immersion I of 70% and (iii) a thickness t of 4 mm for the surface dressing; the amount of glue AG per m2 is given by: AG = p ρ I t
(1)
The results are summarized in the Table 2. The circle packing models in 2D give the lowest quantities of glue. This appears logical as it describe an ideal situation with perfectly round aggregates in only 2D. The 3D packing models are giving higher glue volume due to the third dimension taken into account. De Larrard model is 3D as well and it gives the highest glue volume since the aggregates are not considered as perfect spheres. It as to be noted that this model has been developed for concrete and may not be applicable for casting in a very thin layer as it is the case here (4mm). Finally, the image correlation, which is a 2D approach, appears as the most reliable one. This method is appropriate for a very thin layer and consist of a direct measurement of the aggregates porosity. The result appears coherent with the standard. Table 2 Surface dressing for different glass fractions. Legend: f = packing density; p = porosity; AG = amount of glue per m2 Method
Model
f
p=1−f
AG (kg/m2 )
(i)
Square lattice packing 2D
0.785
0.215
0.66
Hexagonal lattice packing 2D
0.907
0.093
0.29
Square lattice packing 3D
0.74
0.26
0.8
Hexagonal lattice packing 3D
0.605
0.395
1.22
(ii)
De Larrard Model
0.578
0.422
1.30
(iii)
Image correlation
0.706
0.294
0.91
Surface Dressing Treatment for Applications on Solar Roads
1723
4 Mechanical and Optical Performance The mechanical performances of three different thermosetting polyurethanes have been studied. The optical performances of surface dressings designed with each polyurethanes were measured. Regarding the mechanical performance of the glues, the complex modulus lE*l vs temperature T have been measured for a frequency of 1 Hz. This has been done by performing the dynamic mechanical analysis on cylindrical solid specimens (Ø 9x18 mm3 ), cured at 50 °C. The polyurethanes B and C are stiffer compared with a typical bitumen 35–50, especially for temperatures higher than 30 °C (Fig. 2). This result is encouraging, taking into account that the photovoltaic roads could be built in hot areas. Fig. 1 Aggregates spreading
Fig. 2 Complex modulus versus temperature
1724
D. Vizzari et al.
Table 3 Power loss and BPN for different types of polyurethane
Sample
A
B
C
Power loss
15.6%
13%
15.5%
BPN
67
58
56
The optical performance was evaluated studying the power loss of the solar cell because of the surface dressing treatment. The test was performed in a controlled environment, using a lamp of 400W for the simulation of the sunlight radiation. The highest efficiency of the solar cell is achieved when the solar cell delivers the maximum power Pmax . The Pmax occurs for a certain value of current and voltage [11]. Covering the solar cell with the surface dressing, the Pmax reduces. In order to describe the reduction of power, the following equation was defined: P L = 100 Pmax,unc − Pmax,c /Pmax,unc
(2)
where: PL is the power loss [%]; Pmax,unc is the maximum power point of the solar cell [mW] and Pmax,c is the maximum power point of the solar cell covered by the surface dressing [mW]. The PL ranges theoretically between 0% (top layer perfectly transparent) and 100% (top layer totally black). The power loss was measured on three samples using the polyurethanes A, B and C and with a surface dressing having 1.99 kg/m2 of glass aggregates 2/4mm and an amount of glue of 0.9 kg/m2 . The polymerization of the binder was achieved after 20–24 h, at the room temperature of 20 °C. The power loss (Table 3) of the polyurethanes A, B and C is around 15%. The results are very encouraging, considering that the previous semi-transparent layers had a power loss between 45 and 76% [12]. In terms of skid resistance (Table 3), the British Pendulum has been performed on the same samples in wet conditions, obtaining a BPN (British Pendulum Number) between 56 and 67. The results, higher than 55, are “generally satisfactory, meeting all but the most difficult conditions encountered on the roads” [13]. Anyway, further studies are requested for better-understanding the effect of the microtexture.
5 Conclusions and Next Steps The goal of this paper was to propose a novel method for the manufacture of the semi-transparent layer of a photovoltaic road based on the surface dressing treatment. The top surface is made by glass aggregates laid down on a thin layer of transparent polyurethane. The first step was to evaluate the packing density of the glass aggregates in order to calculate the optimal amount of glue. Based on the results, the surface dressing requires 1.99 kg/m2 of glass aggregates 2/4mm uniformly spread on a thin layer of polyurethane (0.9 kg/m2 ).
Surface Dressing Treatment for Applications on Solar Roads
1725
The authors tested three different types of thermosetting polyurethanes. In terms of mechanical performance, the glues were compared with a typical bitumen 35–50, studying the variation of complex modulus according to the temperature. The glues B and C are more stiff than the bitumen, especially in the range of temperature 30– 60°C. Taking into account that a solar road could be built in very hot areas, this result is encouraging. Regarding the optical performance, the power loss is between 15 and 16%. This result is due to the low thickness of the surface dressing treatment (around 0.5 cm), the absence of fine particles that could interfere with the sunlight radiation and the absence of discontinuity between the solar cell and the glue. In other terms, the solar cell is fully merged with the polyurethane. The skid resistance meets the BPN standard, but further research is needed. The next steps will be the construction of a working prototype in order to measure the electrical output and prove the feasibility of this system.
References 1. Vizzari, D., Chailleux, E., Gennesseaux, E., Lavaud, S., Vignard, N.: Viscoelastic characterization of transparent binders for application on solar roads. Road Mater. Pavement Design. (2019). https://doi.org/10.1080/14680629.2019.1588774 2. Hendy, S: Light scattering in transparent glass ceramics. Appl. Phys. Lett. (2002) 3. Overseas Road Note 3 (2nd Edition): A guide to surface dressing in tropical and sub-tropical countries. First Published 2000 ISSN:0951–8797 4. Gransberg, D., James, M.B.: Chip Seal Best Practice. NCHRP. Synthesis 342 (2005) 5. https://www.pavementinteractive.org/ 6. Guide technique pour l’utilisation des materiaux alternatifs de Bourgogne 7. Mistry, R.: Circle Packing, Sphere Packing and Kepler’s Conjecture. May 15 2016 8. Fukshansky, L.: Revisiting the hexagonal lattice: on optimal lattice circle packing (2011) 9. Weisstein, Eric W.: Sphere Packing. From MathWorld -A Wolfram Web Resource 10. Stovall, T., De Larrard, F., Buil, M.: Linear packing density model of grain mixtures 48(1), 1–12 (1986) 11. Vitorino, M. A., Hartmann, L. V., Lima, M. N., Correa B. R.: Using the model of the solar cell for determining the maximum power point of photovoltaic systems (2017) 12. Vizzari, D., Chailleux, E., Lavaud, S., Gennesseaux, E., Bouron, S.: Fraction factorial design of a novel semi-transparent layer for applications on solar roads. Infrastructures 5(1), 5 (2020) 13. ASTM E303–96. Experiment No. 6
Sustainability Assessment of Novel Performance Enhancing Chemical Bitumen Additive Christian Holldorb, Amina Wachsmann, Sonja Cypra, Markus Oeser, Nicolás Héctor Carreño Gómez, and Michael Zeilinger
Abstract Increasing traffic, changing climatic conditions and new safety requirements during construction, as well as the pursuit of sustainable road construction generate the need for new pavements solutions (materials). A novel reactive chemical bitumen additive (B2Last® ) can be such a solution, because B2Last® -modified asphalt has the capability of being produced and paved at reduced temperatures. In this investigation, data collected from field tests, together with stiffness and fatigue tests carried out in the laboratory, were compiled to study the service life of asphalt layers with and without B2Last® following the German Standards (RDO Asphalt 09). The results of these tests were combined with data gathered from past field trials in order to perform an extensive service life analysis. Based on these results, a sustainability assessment was done to point out three main aspects: (i) ecological aspects by a life cycle assessment (LCA) according to ISO 14040; (ii) economic aspects conferring to life cycle costs; (iii) social aspects like occupational exposure to vapors and aerosols. The results obtained so far are major sustainability requirements which are fulfilled by using B2Last® . This study is illustrating that better performances of pavement can still be achieved nowadays by providing innovative solutions to the asphalt industry. Keywords Modified bitumen · Performance testing · Asphalt recycling · Life cycle assessment · Sustainability assessment
C. Holldorb (B) · A. Wachsmann · S. Cypra Institute for Transport Systems and Infrastructure, Karlsruhe University of Applied Sciences, Karlsruhe, Germany e-mail: [email protected] M. Oeser · N. H. C. Gómez Institute of Highway Engineering, RWTH Aachen University, Aachen, Germany M. Zeilinger BASF SE, Ludwigshafen, Germany © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_219
1727
1728
C. Holldorb et al.
1 Introduction Due to higher traffic loads and changing climate conditions, coupled with new regulations to decrease the exposure of asphalt workers to bitumen vapors and aerosols during construction, as well as an overall pursuit of a more sustainable road construction generate the need for new materials. A novel isocyanate-based bitumen modifier (B2Last® ) has shown to increase asphalt performance [1], as well as to enable production and paving at lower temperatures. The working principle of the additive is that it creates a chemical bond between constitutive elements of the bitumen (mainly resins and asphaltenes). Since the additive reacts with the functional groups within the bitumen, it works especially better in mixtures with reclaimed asphalt. Due to oxidation, the bitumen within the mixture has more functional groups for the additive to react with. The chemical bond increases the elasticity of the modified bitumen, increasing rutting and fatigue stability. Recently, the additive has shown to enable asphalt production and paving at lower temperatures due to a reduced viscosity during the reaction process. To date, two paving trials have confirmed its temperature reduction capabilities. In both cases, the additive was dosed in an in-line blending procedure at the mixing plant. The temperature reduction is possible because the additive, a low viscous fluid, does not affect greatly the viscosity of the neat bitumen during mixing, transport and construction. But once it finally reacts, during the last stages of the construction, enhances the material performance. It also allows the use of softer bitumen (e.g. PEN 70/100 instead of 50/70), extending useful temperature interval upwards, while maintaining the low temperature properties undisturbed. The aim of this paper is to present the research results of a comprehensive sustainability assessment done with data gathered from field trials and a service life prediction analysis done in the laboratory based on the German mechanistic-empirical design method (FGSV—RDO Asphalt 09).
2 Service Life Prediction To be able to perform the sustainability assessment, firstly a prediction of the service life was required. For this purpose, the German mechanistic-empirical design guide for asphalt pavements (FGSV—RDO Asphalt 09) was followed. According to this standard, the prediction is done by calculating the resulting deformations at the bottom of the base layer, using the linear elastic multilayer theory (or the finite element method), considering traffic and temperature data. Afterwards, Miner´s Law is applied to predict the failure using the fatigue equation of the base layer, which is obtained after thorough laboratory tests. For this analysis, the material properties of an unmodified and a B2Last® -modified base layer were obtained in the laboratory, while calibrated data from FGSV—RDO Asphalt 09 was used for the surface, binder, frost protection layers, as well as for the soil. Base layers are not usually mixed with
Sustainability Assessment of Novel Performance Enhancing …
1729
modified bitumen, but since the only standardized way to predict service life is through the base layer, this method was chosen.
2.1 Material Characterization of Base Layers For this investigation, an AC 22 T S (dense graded base layer with max. Grain size of 22 mm) was chosen. Firstly, the mixture recipe had to be verified, since both reference and B2Last® -modified mixtures were mixed with 50% of reclaimed asphalt. The same PEN 50/70 bitumen was used for both mixtures; the modified variant was modified with 2.5% B2Last® by weight of bitumen content (neat and reclaimed). To recreate the warm mix capabilities of the additive in the laboratory, roller-segment-compacter plates were produced at 115 °C for the modified variant, while the reference variant plates were produced at the standard temperature of 135 °C. To predict the service life, the stiffness (master curve) and fatigue behavior of the unmodified and modified base layers were obtained following the German Standards TP Asphalt-StB Part 24 and Part 26 respectively. Both standards describe cyclic indirect tensile strength tests (CITS), but at different temperatures and frequencies. The fatigue behavior was obtained by performing CITS tests at 20 °C, at 10 Hz, and at three different stresses levels, while the stiffness master curve was obtained by performing CITS tests at −10, 5 and 20 °C, and at frequencies ranging from 0,1 to 10 Hz, also at different stress levels. The results can be found in Fig. 1. The B2Last® modified asphalt base layer shows a higher fatigue curve (Fig. 1a), which means that by an equal strain it can withstand a higher number of load cycles. On the other hand, the master curve (Fig. 1b) of the modified variant also shows a higher stiffness at low frequencies (high temperatures) and a lower stiffness at high frequencies (low temperatures), compared to the reference variant. This effect is desirable and translates to a better asphalt performance in both high and low temperature range.
2.2 Traffic and Climatic Loads Aside from the material properties, traffic and climatic data had to be considered. For the calculations, these values were the same for both layers. The traffic load chosen was an AADT of 42.500 vehicles per day, with a 21% share of heavy traffic vehicles (yearly increase 3%). The design period was set to 30 years and the load collective was taken from FGSV—RDO Asphalt 09 for a long-distance federal motorway, which translated to 129,2 M equivalent 10-t equivalent-single-axle-loads. With respect to the climatic data, the frequency distribution of the surface temperature was taken from the standard for Zone 3, which corresponds to the south of Germany.
1730
C. Holldorb et al.
Fig. 1 Fatigue curves a and |E*| master curves b of AC 22 T S variants
2.3 Results of the Service Life Prediction The last step for the prediction was to define the layer thicknesses of the flexible pavement structure. Two scenarios were analyzed (Table 1). The thicknesses for the scenario 1 were taken from Table 1 of the FGSV RStO 12 German Design Standard with design values for standardized cases, for the highest load class (Bk100). Since in scenario 1 the reference material did not comply with the design period, the second scenario was defined so that reference layer could comply with the design period. Table 1 Layer thicknesses and service life prediction results
Layer
Scenario 1 Thicknesses (mm)
Scenario 2 Thicknesses (mm)
Surface layer
40
40
Binder layer
80
80
Base layer
220
290
Reference
15,5 years
32,3 years
B2Last®
31,5 years
59,9 years
Sustainability Assessment of Novel Performance Enhancing …
1731
The results (Table 1) show that the modified layer can withstand roughly twice as long as the unmodified layer in both scenarios.
3 Sustainability Assessment 3.1 Methodology To demonstrate the advantages of an asphalt modified with B2Last® compared to conventional asphalt produced according to FGSV ZTV Asphalt, a comprehensive sustainability assessment was conducted for use in Germany [2]. Based on the approaches published in [3], the information modules shown in Fig. 2 were considered. The variants shown in Table 2 were examined. The variants differ only in the choice of binder, which affects both the lifetime calculated according to [2] and the manufacturing temperatures. Based on European standards ISO 14040, DIN EN 15643–4, DIN EN 15643–5, DIN EN 15804, DIN prEN 17392, ISO 15686–5 and [4, 5], the asphalt pavement of a fictitious 3 km long and 14.5 m wide motorway roadway was assessed for the three variants. A period of 60 years and a discount rate of 3% were considered. Over this period of time, all relevant processes like material extraction, material production, material transports, construction as well as deconstruction were taken into account in order to be able to make comparative statements about environmental effects, construction costs and workplace exposure (cf. Fig. 2).
Fig. 2 Considered information modules for the assessment of an asphalt pavement in accordance with European Standards DIN EN 15804, DIN EN 17392–1 and DIN EN 15643–5
1732
C. Holldorb et al.
Table 2 Overview of the examined variants Variant 1—conventional
Variant 2.1 - B2Last® (overall)
ADS
ATS
ADS
B
B + B2Last®
ABS
ABS
Variant 2.2 – B2Last® (only ADS & ABS) ATS
ADS
ABS
ATS
B + B2Last® B
Binder
PmB
Layer Thickness [mm]
40
80
290
40
80
290
40
80
290
Service Life [years]
10
20
30
10
30
60
10
30
30
Mixing Temperature (°C)
170
140
140
170
PmB Polymer modified Bitumen B Bitumen ATS Asphalt Base Course ABS Asphalt Binder Course ADS Asphalt Surface Course
3.2 Results of the Sustainability Assessment It could be demonstrated that the use of B2Last® not only saves energy and thus reduces the greenhouse gas emissions caused (cf. Fig. 3), but also significantly reduces the construction costs and the exposure of workers to vapors and aerosols during construction. If B2Last® is used in all three asphalt layers (Variant 2.1), a total primary energy saving of 41.3%, a reduction of greenhouse gas emissions by 46.0%, a reduction of 23.7% of the construction costs and a reduction of up to 2/3 of the exposure of workers to vapors and aerosols compared to the conventional construction method (Variant 1) can be achieved. If B2Last® is only used in ADS and ABS (Variant 2.2), savings of 12.6% (total primary energy demand), 13.8% (greenhouse gas emissions) and 7.8% (construction costs) as well as a halving of the exposure are achieved. The plausibility of these statements could be verified by the systematic variation of significant parameters (service life of B2Last® layers; production temperature of B2Last® -modified asphalt; discount rate) within the scope of sensitivity analyses. The main reasons for the positive effects when using B2Last® are the extended service lives (cf. Table 2, Variants 1 and 2.1), the possibility of reducing the asphalt mixing temperature and the fact that conventional polymer modification, which is an energy-intensive process, can be dispensed with. The reduction of the asphalt production temperature has the advantage of reducing the energy demand and the resulting greenhouse gas emissions, as well as the reduction of the workplace exposure, thus reducing the health risk for construction workers.
Sustainability Assessment of Novel Performance Enhancing …
1733
Fig. 3 Total primary energy demand [GJ] and Global Warming Potential [t CO2 -eq.] over an observation period of 60 years, differentiated by layer
4 Conclusions In this paper, the service life of a novel chemical additive was studied to perform a comprehensive sustainability evaluation. The laboratory results show that the isocyanate-based additive increased the asphalt performance allowing to double, in this case, the service life of a base layer. With this information, a sustainability assessment was performed that concluded that the novel additive allows considerable savings of primary energy (and thus greenhouse gas emissions), construction costs and helps reduce bitumen vapors and aerosols during construction as well.
References 1. Carreño, N., Renken, L., Schatz, W., Zeilinger, M., Bokern, S., Fleischel, O., Oeser, M..: New type of chemical modification of asphalt binders to enhance the performance of flexible pavements. Proceedings of the 7th Eurasphalt &Eurobitume Congress v1.0 (2020) 2. Holldorb, C., Brzuska, A., Cypra, S.: Nachhaltigkeitsbewertung zum Einsatz von B2Last® modifiziertem Asphalt. University of Applied Sciences, Karlsruhe (2020).unpublished) 3. Brzuska, A., Cypra, S., Holldorb, C., Stöckner, M.: CO2 -Emissionen bei der Straßenerhaltung. Straße und Autobahn 69, 845–853 (2018) 4. Gesprächskreis Bitumen: Expositionsbeschreibung Herstellung und Beförderung von Asphalt. https://www.bgbau.de/fileadmin/Die_BG_BAU/Gespr%C3%A4chskreis_Bitumen/ExpoHerst TranspAsphalt.pdf. Last accessed 27 May 2020 5. Gesprächskreis Bitumen: Expositionsbeschreibung Einbau von Walzasphalt im Straßenbau. https://www.bgbau.de/fileadmin/Die_BG_BAU/Gespr%C3%A4chskreis_Bi-tumen/ExpoWa lzasphalt.pdf. Last accessed 27 May 2020
Temperature and Loading Rate Susceptibility of Bituminous Mixtures on Monotonic Testing Ramon Botella , Félix E. Pérez-Jiménez , Rodrigo Miró , Adriana H. Martínez , and Teresa López-Montero
Abstract The loading rate and temperature susceptibility of bituminous materials is of great importance. Their modulus and phase angle are strongly affected by any change of these two variables. For that reason, many researchers have explored the time-temperature superposition principle that assumes a thermo-rheologically simple behavior, i.e., all curves have the same characteristic time variation law with temperature. Nowadays it is common to obtain the so-called master curve of any bituminous material (binder, mastic or mixture) using dynamic testing. A side from that, several monotonic test methods have been proposed recently to obtain the cracking resistance of asphalt mixtures at low and intermediate temperatures. Many of them are focused on extracting ductility related parameters at temperatures close to 20 °C. Since the cracking resistance of the mixtures is strongly related with their stiffness and phase angle, it is also dependent on the loading rate and the test temperature. Joining these two concepts, this paper presents the efforts to analyze the influence of loading rate and test temperature in the different parameters that can be extracted from a monotonic test, such as the Fenix test. By testing different mixtures at different temperatures and different loading rates, the influence of these three variables in the different parameters that characterize the stiffness and ductility of the mixtures was analyzed. Finally, the possibility of obtaining a sort of master curve or an absolute reference value for each mixture in terms of any of the parameters considered was explored. Keywords Pavements · Bituminous mixtures · Cracking · Master curve · Fenix test
1 Introduction Fracture oriented tests have become an important tool for asphalt mixture performance characterization in most countries recently. They were first used in areas in R. Botella (B) · F. E. Pérez-Jiménez · R. Miró · A. H. Martínez · T. López-Montero Universitat Politècnica de Catalunya—BarcelonaTech, Barcelona, Spain e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_220
1735
1736
R. Botella et al.
which low temperatures and/or high rates of temperature drop are common. The stiffness of the mixture at low temperatures propitiated a simple extrapolation of the techniques developed for stiffer materials, such as steel or concrete, in the framework of fracture mechanics [1]. More recently, the focus widened considering the possibility of evaluating cracking properties of mixtures at intermediate temperatures. In those cases, the application of fracture mechanics is more challenging, since the behavior of the mixture at temperatures between 15 and 25 °C is highly viscous leading to significant plastic deformation before cracking. For instance, obtaining the real fracture energy or fracture toughness at temperatures above 0 °C for a regular asphalt mixture is very complicated since separating the different types of energies released during the specimen breaking process is impossible without making some assumptions. However, in-field experience shows that cracking also takes place in areas in which temperatures stay in the intermediate range, making the cracking characterization of mixtures at those temperatures necessary. In those cases, analyzing the area below the load-displacement curve may be helpful. This magnitude is the total energy invested to break the specimen during the test. Additionally, the high temperature susceptibility showed by asphalt mixtures has led to many research studies about the obtention of the master curve of both binders and mixtures, in terms of the dynamic modulus and/phase angle using short cyclic tests. The objective of those techniques is obtaining a general portrait of the behavior of the material at a wide range of temperatures and loading rates. Little work has been done to try to obtain this kind of temperature/loading rate behavior in terms of other fracture parameters. In that regard, in 2009, Tussiwand et al. tried to obtain a master curve in terms of the fracture toughness for a solid rock propellant, a material that also shows a viscoelastic behavior [2]. Despite the thermal susceptibility of asphalt mixtures, some researchers and administrations are looking for procedures that can evaluate the cracking properties of these materials at intermediate temperatures with short testing time. For that reason, some methodologies have been implemented in the U.S.A. based on performing monotonic testing at only one temperature of interest [3, 4]. However, obtaining a temperature-behavior curve can improve the decision and comparison process. This research paper presents the efforts to combine two concepts; a cracking focused test at intermediate temperatures, i.e., the Fenix test [5], and the master curve analysis, to provide a cracking characterization method that considers the different range of temperatures and loading rates that the mixture is going to sustain during its service life.
Temperature and Loading Rate Susceptibility of Bituminous …
1737
Fig. 1 Fenix test set-up
2 Methods and Materials 2.1 Fenix Test The Fénix test is a tension test that uses similar specimen geometry than that of the SCB test and applies a loading configuration similar to that of the DC(T) test. The test is named after the research project FENIX that funded its development. The test is normally performed at a constant displacement rate of 1 mm/min and 20 °C, but in this research project, other loading rates and temperatures were explored, as detailed in the Test plan section. The specimens were carved from Marshall compacted cilinders and the notch was approximately 6 mm deep and 4 mm wide. For more details about the test procedure, refer to [5] (Fig. 1).
2.2 Materials The two mixtures employed in this study are detailed below: • Centered BBTM 8B (8 mm NMAS) with a 5% in mixture mass of an SBS polymer modified bitumen PMB 45/80–65. Coarse aggregate was porphyry, fine aggregate limestone. • Centered AC16S (16 mm NMAS) with a 5% in mixture mass of a 50/70 pen binder. Coarse aggregate was porphyry, fine aggregate limestone. These two mixtures represent the most common choices when designing a wearing course in Spain. These two mixtures were selected to have a broad spectrum using the minimum number of mixtures. The AC16 S represents the continuously graded
1738
R. Botella et al.
mixtures with conventional binders while the BBTM 8B represent the gap-graded mixtures with polymer-modified binders.
2.3 Test Plan The test plan consisted of applying the Fenix test to the specimens of the two mixtures at three different temperatures (0, 10 and 20 °C) and at three different loading rates (0.1, 1 and 10 mm/min). Three replicates were tested for each temperature/loading rate combination, for 27 specimens per mixture. Five replicates lead to results that are more reliable but due to materials and testing time limitations this study was conducted with three.
3 Results A side from the total energy released during the test and the peak stress, Table 1, also shows a ductility related parameter that measures the displacement between the maximum stress and the 50% the maximum stress after peaking. The trends shown by the peak stress and the total energy were very similar. Both mixtures showed maximum values of the total energy at 10 °C and 10 mm/min. This result was expected, since this parameter is minimum at high temperatures and low loading rates and also a very low temperatures and high loading rates. A maximum value should exist between those extreme conditions. The maximum total energy values were statistically the same for both mixtures: 1,209 ± 25 J/m2 and 1,179 ± 130 J/m2 . Considering that those values were obtained at the same testing conditions (10 mm/min and 10 °C) and that the binders and gradation of the two mixtures were very different this maximum value may be more related with the test itself than with the mixture characteristics. However, the evolution of the total energy with the loading rate and temperature was very different between the mixtures. The total energy values at 0 and 20 °C were shifted graphically to produce a semi-continuous master curve (Fig. 1). The shift factors required are shown in Table 2. The Arrhenius Eq. (1) was tried successfully to obtain the dependency of the shift factors with the temperature, choosing 10 °C as reference temperature. This equation was able to fit the shift factors with a mean relative error of 3.5% for the AC16 S mixture and a 2.3% for the BBTM 8B mixture, with only one parameter, i.e., the activation energy, Ea (Fig. 2). log(aT ) =
Ea · R
1 1 − T T0
(1)
Temperature and Loading Rate Susceptibility of Bituminous …
1739
Table 1 Mean values and COV of some parameters obtained in the Fenix test Total energy
Peak Stress
d0.5PPS-dPS
Mixture
Temperature (ºC)
Loading rate (mm/min)
Mean (J/m2 )
COV (%)
Mean (MPa)
COV (%)
Mean (mm)
COV (%)
AC16S
0
0.1
906
13
0.81
5
0.32
18
1
1058
8
0.89
4
0.58
15
10
1060
11
1.05
1
0.47
17
0.1
685
7
0.31
11
0.86
6
1
855
5
0.62
7
0.71
11
10
1209
2
1.22
2
0.39
0
0.1
248
5
0.11
11
1.25
15
1
402
5
0.24
2
0.97
9
10
714
10
0.47
11
0.94
6
0.1
924
9
0.49
6
0.82
6
1
676
8
0.27
9
1.99
11
10
675
8
0.27
9
1.99
11
0.1
726
5
0.35
7
1.63
16
1
725
4
0.35
6
1.63
16
10
1173
11
0.71
7
1.06
10
0.1
189
17
0.06
7
2.65
22
1
402
9
0.15
6
2.53
15
10
679
7
0.28
10
1.99
11
10
20
BBTM 8B
0
10
20
Fig. 2 Master curve attempts for the AC16S and BBTM 8B mixtures
The activation energies obtained for the two mixtures were very similar and their magnitudes were very similar to values obtained in dynamic modulus master curves for these kinds of mixtures. The results obtained indicated the existence of a maximum value for the total energy dissipated during the Fenix test. In addition, at low temperatures this parameter seems to be independent of the loading rate, for displacement velocities equal or higher than 1 mm/min. The main difference between mixtures was the total energy
1740
R. Botella et al.
Table 2 Values for the shift factors and fitting coefficients of Eq. 1 Mixture
Temperature (°C) Shift factor, aT Activation energy, Ea Gas cnst, R (J/mol ºK) (KJ/mol)
AC16 S
0
100
10
1
20
0.01
BBTM 8B 0
200
10
1
20
0.009
133
8.314
145
value at which this parameter remained constant at 0 °C and 1–10 mm/min. This value was 1060 ± 117 J/m2 for the AC16S mixture and 675 ± 50 J/m2 for the BBTM 8B mixture. Even though the maximum total energy was the same for the two mixtures, this “stabilization value” was 36% lower for the BBTM 8B mixture.
4 Conclusions This short paper presents a preliminary study about the feasibility of composing a sort of a master curve using the Fenix test at different temperatures and loading rates in terms of the total energy expended during the test. The main objective was producing a curve or a parameter able to clearly differentiate and rank mixtures by their thermal and loading rate susceptibility. The evolution of the total energy with the temperature/loading rate was graphically shifted to try to fit a continuous type of curve but the data was not enough to produce a good fit. However, some interesting results were obtained: • The maximum total energy in the Fenix test at any temperature/loading rate combination was independent of the type of mixture and was 1179 ± 130 J/m2 . • At lower temperatures (0 °C) the total energy was independent of the loading rate for displacement velocities over 1 mm/min, producing an “stabilization value”. • This “stabilization value” was significantly lower (36%) for the mixture expected to present a better cracking behavior (modified binder). Only the total energy was analyzed, but future work can be done on other parameters obtained from the Fenix test. More mixtures should be analyzed, with more replicates, to either confirm or reject the trends observed. Additionally, lower temperatures should be included in the test plan. Acknowledgement Part of the research presented on this paper belongs to the project RTI2018096224-J-I00 that has been cofounded by the Spanish Ministry of Economy, Industry and Competitiveness, inside the National Program for Fostering Excellence in Scientific and Technical Research, National Subprogram of Knowledge Generation, 2018 call, in the framework of the Spanish National Plan for Scientific and Technical Research and Innovation 2017–2020, and by the European
Temperature and Loading Rate Susceptibility of Bituminous …
1741
Union, through the European Regional Development Fund, with the main objective of Promoting technological development, innovation and quality research.
References 1. Dave, E., Botella, R., Marsac, P., Bodin, D., Sauzéat, C., Nguyen, M.L.: Cracking in asphalt materials. RILEM State-of-the-Art Reports 28, 33–102 (2018) 2. Tussiwand, G.S., Saouma, V., Terzenbach, R., De Luca, L.T.: Fracture mechanics of composite solid rocket propellant grains: material testing. J. Propul. Power 25(1), 60–73 (2009) 3. Mohammad, L.N., Kim, M., Elseifi, M.: Characterization of asphalt mixture’s fracture resistance using the semi-circular bending (SCB) test. In: 7th RILEM International Conference on Cracking Pavements, Delft, The Netherlands (2012) 4. Nemati, R., Haslett, K., Dave, E.V., Sias, J.D.: Development of a rate-dependent cumulative work and instantaneous power-based asphalt cracking performance index. Road Materials and Pavement Design (Issue sup1: EATA Granada), pp. 315–331 (2019) 5. Pérez-Jiménez, F.E., Valdés, G., Miró, R., Martínez, A.H., Botella, R. Fénix Test.: Development of a New Test Procedure for Evaluating Cracking Resistance in Bituminous Mixtures. Transp. Res. Record 2181, 36–43 (2010)
Temperature Dependency of the Stiffening Effect of Hydrated Lime in Stone Mastic Asphalt (SMA) Mixtures Mariella Cardenas, Juan S. Carvajal-Munoz , and Gordon Airey
Abstract The effect of hydrated lime (HL) on the stiffening effect of bituminous mixtures is mainly driven by a physical phenomenon derived from the higher Rigden Air Voids (RAV) of HL in contrast to other filler types commonly used in bituminous mixtures such as granite and limestone. The stiffening effects of HL at room temperature for some mixture types have been reported in the literature but the knowledge of the effects on Stone Mastic Asphalt (SMA) mixtures is still limited. Similarly, the temperature-dependency has been theorised in the literature but scarce experimental evidence is currently available. As a consequence, this research focuses on studying the effects of a wide range of temperatures on the stiffening effect of SMA mixtures with HL. Indirect Tensile Stiffness Modulus (ITSM) testing was used in order to assess the stiffness properties of SMA mixtures manufactured with limestone aggregates, a conventional 30/45-penetration grade bitumen, and the addition of HL as a partial replacement of limestone filler (LF). The bitumen to filler ratio was kept as 1:1 for all manufactured mixtures. The results indicated: (i) a clear temperature-dependency of the stiffening effect of HL, which is more prominent at intermediate temperatures (20 and 30 °C). (ii) The percentage of added HL and the average stiffness of the specimens had an approximately linear correlation. (iii) The stiffening effect is less prominent at the lower and higher temperatures considered. (iv) The partial replacement of LF by HL produced stiffer mixtures than those without replacement, which suggests improved mechanical response of these SMA mixtures. Keywords Stiffness · Hydrated Lime · Temperature Dependency · Stone Mastic Asphalt
M. Cardenas (B) · J. S. Carvajal-Munoz · G. Airey Nottingham Transportation Engineering Centre (NTEC) / University of Nottingham, Nottingham, UK e-mail: [email protected] J. S. Carvajal-Munoz Department of Civil and Environmental Engineering / Universidad del Norte, Barranquilla, Colombia © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_221
1743
1744
M. Cardenas et al.
1 Introduction Hydrated lime (HL) is a fine-white powder, produced from limestone rocks, which was originally used as a filler for bituminous mixture production in the 1950s and it was mainly used for enhanced bitumen-aggregate adhesion. In the 1970s, during the petroleum crisis in the USA, a general decrease in the quality of bitumen induced a higher susceptibility of asphalt pavements to moisture damage and frost. By then, after thorough investigation and field testing, HL was found to be the most effective additive to limit moisture damage which led to a wide use in the USA. In fact, about 10% of current bituminous mixtures produced in the USA incorporates HL. Due to extensive research, HL is now known to provide additional benefits to the mixture performance. Specifically, according to Lesueur et al. [1], HL is considered as an active filler due to its ability to: (1) reduce bitumen chemical ageing by delaying the chemical oxidation process; (2) stiffen the mastic and mortar which improves rutting resistance, and (3) increase the resistance to fatigue cracking due to a mechanism that helps delay crack propagation. From the physical point of view, the active nature of HL is partially explained by the higher Rigden Air Voids (RAV) of HL in contrast to other filler types, which are related to the porosity of the dry compacted filler and leads to increased stiffening effect of the mastic and the mixture. Thereby, the use of HL promotes increased strength of the bituminous mixtures with positive performance implications [2]. This stiffening effect has been theorised to be more pronounced at intermediate testing temperatures [1], but very limited experimental data is currently available in the literature to corroborate this theory in terms of the temperature-dependency of this HL effect. Based on a recent literature review by [3], more comprehensive studies on bituminous mixtures are required in order to understand the effects of HL and its mechanisms of action with specific materials, mixture designs and testing conditions across Europe. Furthermore, although HL has proven benefits for bituminous mixtures, particularly in France and the Netherlands, its application in other European countries is still lagging in comparison to the USA. It is therefore required that current investigations should expand knowledge about HL in the European context. In consequence, this paper presents an experimental analysis of Stone Mastic Asphalt (SMA) mixtures with incorporated HL produced with locally available materials and testing protocols in the United Kingdom. Indirect Tensile Stiffness Modulus (ITSM) testing was used in order to assess the stiffness of the SMA mixtures. To contribute to some of the gaps in current knowledge, the main aims of this investigation were to assess the temperature-dependency of the stiffening effect of HL, and to determine the correlation of the HL content and the stiffness modulus in SMA mixtures.
Temperature Dependency of the Stiffening Effect of Hydrated …
1745
2 Materials and Methods 2.1 Materials The SMA mixtures were manufactured with limestone aggregates, 30/45 penetration grade bitumen, and incorporated HL as a partial replacement of limestone filler (LF) at concentrations of 0%, 1%, 2%, and 3% by total weight of mixture. The bitumen to filler ratio was kept as 1:1 for all mixtures produced.
2.2 SMA Mixture Design SMA is a durable surfacing material suitable for use in urban streets and highways. It is composed of a high proportion of coarse aggregate that withstands permanent deformation and is typically composed by bitumen, aggregates, filler and additives (e.g., fibres). Typical SMA is composed of 70–80% coarse aggregate, 8–12% filler, 6– 7% binder and 0.3% fibre [4, 5]. SMA mixtures with a nominal maximum aggregate size of 10mm were manufactured in accordance with BS EN 13108–5:2016 Standard [6]. The manufacturing and compaction temperatures are based on bitumen viscosities and these were 165 ± 5˚C and 145 ± 5˚C, respectively. The aggregates and bitumen were heated for a minimum of 8h and 3h, respectively. HL was added to the dry aggregates in the mechanical mixing drum, where these were subjected to a mixing time of 30 s. Immediately afterwards, the bitumen was added and all components were mixed for 3 min. For compaction, a roller compactor with smooth steel surface was used to simulate the commonly used compaction method for asphalt wearing courses in the field. Roller compacted slabs were produced with 306 × 306 × 60 mm3 dimensions, designed to a target air voids content of 4%. Six cylindrical specimens were cored from the slabs, removing 10 mm from top and bottom in order to reduce the scattering in air voids content along the specimens’ height and produce smooth surfaces. The final specimen geometry was set as 100-mm diameter and 40-mm thickness.
2.3 Determination of Stiffness Modulus Stiffness is defined as the resistance to deformation of a given material when subjected to external loading and it constitutes a valuable parameter for purposes of designing and characterisation of bituminous mixtures as it provides understanding of the mechanical properties and performance-related characteristics [7]. Testing was carried out using the Nottingham Asphalt Tester (NAT), with Indirect Tensile Stiffness Modulus (ITSM) configuration, with Linear Variable Differential Transformers
1746
M. Cardenas et al.
(LVDTs) to measure the horizontal deformation at the mid-span of the circumference of the specimens. During testing, a vertical load is applied to the specimen by loading strips, which causes a relatively uniform tensile stress in the horizontal direction perpendicular to the plane of loading [8, 9]. During normal traffic conditions, asphalt acts elastically and its ITSM directly relates to its bending resistance and the load spreading ability [10]. Following the BS-EN-12697-26:2018 test method [11], Annex C, the following parameters were used for determination of stiffness modulus of the mixture cores: (1) Target rise time = 124 ± 4 ms; (2) Mean horizontal deformation = 5 ± 2 µm; (3) Poisson’s ratio = 0.35; (4) Diameter = 100mm; (5) Thickness = 40mm; Temperatures = 0–50 °C at 10 °C increments. The samples were conditioned at each temperature for at least 4 hours. A total of 6 samples per mixture type were tested for ITSM. Testing began at the lowest temperature and proceeded to the highest. The stiffness modulus was automatically computed by the NAT using Eq. 1 [11]. Sm =
F(v + 0.27) zh
(1)
Where: Sm is the indirect tensile stiffness modulus (MPa);F is the peak value of the applied vertical load in Newtons (N);z is the amplitude of the horizontal deformation obtained during the load cycle in millimetres (mm);h is the mean thickness of the test specimen in millimetres (mm);v is the Poisson’s ratio, set to 0.35 for bituminous materials.
3 Test Results and Discussion From the gain in stiffness from the control mixture (i.e., 0%HL) to the mixtures with maximum HL content (i.e., 3% HL) shown in Table 1, it is observed that the stiffening effect was more prominent at intermediate temperatures (i.e., 20 and 30 °C) in contrast to the lower- and upper-temperature spectrum. This clearly represents experimental evidence of the temperature-dependency of the stiffening effect that was previously theorised in literature [1, 3]. This observation is presented in italics in the last column of Table 1, and it refers to the increase or gain of average stiffness modulus of the mixtures between 0%HL and 3%HL. This finding is consistent with previous studies by [1], and it is explained in terms of the physical interaction of HL with the bitumen to increase the mastic stiffness at intermediate temperatures, whereas this effect is reduced with higher and lower testing temperatures. It is also observed that the lower temperatures exhibited the highest variability in terms of the standard deviation values and COV. As shown in Fig. 1 the correlation of stiffness modulus and HL content for each testing temperature considered exhibit a clear reduction of the stiffness rate of increase (slope of the linear fitting) as the temperature increases from 0 to 50 °C. Furthermore, the comparison of temperature and HL content shows that the average
Temperature Dependency of the Stiffening Effect of Hydrated …
1747
Table 1 Comparison of the stiffness moduli at six testing temperatures and basic statistics
Mixture type (%)
0 1 2 3 0 1 2 3 0 1 2 3 0 1 2 3 0 1 2 3 0 1 2 3
Test temperature ( C) 0
10
20
30
40
50
Average stiffness (MPa)
16,385 18,101 19,048 19,523 9,263 11,236 11,396 11,526 3,478 4,366 4,562 4,597 1,060 1,295 1,477 1,455 407 450 513 507 219 194 260 261
Standard deviation
1,395 1,067 940 789 464 351 140 327 75 88 93 70 83 44 37 45 25 27 36 48 1 12 9 11
COV (%) 9 6 5 4 5 3 1 3 2 2 2 2 8 3 3 3 6 6 7 9 1 6 3 4
Gain of stiffness (Control to 3% HL) (MPa)
Percent gain of stiffness (Control to 3% HL) (%)
3,138
19.15
2,263
24.43
1,119
32.16
395
37.29
100
24.05
41
18.91
stiffness approaches very low values as the testing temperature goes above 30 °C. Interestingly, as shown in Fig. 1, the lowest testing temperature (0 °C) had the highest coefficient of determination, (R2 = 0.9324), which substantiates a strong linear correlation between these variables. Notwithstanding, the strength of the correlation was lower for other temperatures due to the thermo rheological behaviour of bitumen and the typical variability associated to mechanical testing of bituminous materials. Hartman et al. [10] stated that at higher temperatures the non-linear and viscoelastic material behaviour is more pronounced which would appear to be in agreement with the results.
1748
M. Cardenas et al.
Fig. 1 Comparison of the effect of HL on the stiffening effect of bituminous mixtures at various testing temperatures (0–50 °C, at 10 °C increments)
4 Conclusions This paper assessed the temperature-dependency of the stiffening effect induced by hydrated lime (HL) on Stone Mastic Asphalt (SMA) mixtures. Based on the results and data analysis, the following conclusions are drawn: (i) The temperaturedependency of the stiffening effect induced by HL was clearly identified with experimental data; (ii) Except for the 50 °C testing temperature (R2 = 0.56), both the HL content and the average stiffness of the core specimens exhibited a linear correlation as substantiated by high correlation coefficients (R2 > 0.7); (iii) The increase in testing temperature reduced the impact of the stiffening effect induced by HL on the manufactured mixtures; (iv) At high temperatures (above 40 °C), the addition of hydrated lime to the SMA mixtures and the average stiffness follow a more nonlinear correlation; and (v) the change in stiffness of the mixtures with incorporated HL with respect to the control mixtures (0% HL) indicated that the partial replacement of natural filler (i.e., limestone filler) leads to stiffer mixtures which are expected to have less susceptibility to permanent deformation, thereby substantiating the ability of HL to act as an active filler in the SMA mixtures produced. Acknowledgements Juan S. Carvajal-Munoz thanks the support provided by Universidad del Norte and Colciencias-Colfuturo (Colombia) to conduct PhD studies at the University of Nottingham.
Temperature Dependency of the Stiffening Effect of Hydrated …
1749
References 1. Lesueur, D., Petit, J., Ritter, H.J.: The mechanisms of hydrated lime modification of asphalt mixtures: A state-of-the-art review. Road Mater. Pavement Des. 14(1), 1–16 (2013) 2. Vansteenkiste, S., Visscher, J., Denayer, C.: Influence of hydrated lime on the field performance of SMA10 mixtures containing polymer modified binder. (online) Available at: https://www. h-a-d.hr/pubfile.php?id=924 (2016). Accessed 12 Dec 2018 3. Lesueur, D., Petit, J., Puiatti, D., Ritter, H.: Hydrated lime a proven additive for durable asphalt pavements critical literature review. (online) Eula.eu. Available at: https://www.eula.eu/sites/ eula.eu/files/2011%2012%20EuLA_Asphalt_-_literature_review_UK.pdf (2018). Accessed 22 Oct 2018 4. Blazejowski, K.: Stone matrix asphalt: Theory and practice. CRC Press (2016) 5. Hunter, R. N. (ed.). Asphalts in road construction. Thomas Telford (2000) 6. BS EN 13108–5:2016. Bituminous mixtures. Material specifications. Stone Mastic Asphalt 7. Carvajal-Munoz, J.S., Kaseer, F., Arambula, E., Martin, A.E.: Use of the resilient modulus test to characterize asphalt mixtures with recycled materials and recycling agents. Transp. Res. Rec. 2506(1), 45–53 (2015) 8. Hakim, H., Nilsson, R., Vieira, J. and Said, S.: Round robin test of stiffness modulus by indirect tensile method according to En 12697–26:2004 Annex C. (online) Available at: https://www. h-a-d.hr/pubfile.php?id=635 (2012). Accessed 12 Dec 2018 9. Read, J., Whiteoak, D., & Hunter, R. N.: The shell bitumen handbook. Chapter 16: Testing of asphalts. Thomas Telford (2003) 10. Hartman, A., Gilchrist, M. and Walsh, G.: Effect of mixture compaction on indirect tensile stiffness and fatigue. (online) Available at: https://citeseerx.ist.psu.edu/viewdoc/download? doi=10.1.1.555.3060&rep=rep1&type=pdf. (2001). Accessed 12 Dec 2018) 11. BS EN 12697–26:2018 Bituminous mixtures. Test methods. Stiffness.
Testing of Reclaimed Asphalt Model Systems for the Evaluation of the Effectiveness of Rejuvenators Davide Dalmazzo, L. Urbano, P. P. Riviera, and Ezio Santagata
Abstract This paper presents the results of a preliminary study on the use of a novel approach for the evaluation of the effectiveness of rejuvenating agents used in asphalt mixtures containing reclaimed asphalt (RA). Such an approach is based on the mechanical testing of model systems constituted by a single-sized RA material, in its original state and after pretreatment with rejuvenators, compacted to a target volumetric condition. Model systems were prepared by making use of two rejuvenators and by also considering a reference combination of virgin binder and aggregates extracted from the RA. Tests were carried out for the evaluation of stiffness modulus and indirect tensile strength as a function of curing time. Experimental results showed that the proposed approach can capture the changes occurring in the aged RA binder as a consequence of the effects induced by rejuvenators. Quantification of the rejuvenating effects was carried out by analyzing recorded stiffness and strength variations, and by referring to a ductility toughness index, introduced to better describe stress-strain response after failure. Keywords Reclaimed asphalt · Rejuvenators · Model systems · Stiffness modulus · Indirect tensile strength
1 Introduction Reclaimed asphalt (RA) obtained from the milling of existing asphalt pavements is widely used as a recycled component of asphalt mixtures, thereby leading to a reduction of the depletion of nonrenewable natural resources [1, 2]. Although agencies limit the reuse of RA to prevent a loss of fatigue endurance and thermal cracking resistance [3], significant amounts of RA can be employed if rejuvenating agents are used to restore the original properties of the aged binder contained in it. In current recycling techniques, the effectiveness of such additives is one of the crucial D. Dalmazzo · L. Urbano · P. P. Riviera · E. Santagata (B) Department of Environment, Land and Infrastructure Engineering, Politecnico di Torino, Corso Duca degli Abruzzi, 24, 10129 Torino, Italy e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_222
1751
1752
D. Dalmazzo et al.
aspects to consider in order to correctly tailor the composition of the asphalt mixtures and to define a realistic maximum content of RA. The effectiveness of rejuvenators is usually evaluated by directly analyzing the performance of mixtures containing RA [4–6], or by assessing the rheological properties of the binders extracted from the mixtures by means of solvents [7, 8]. In such cases, the true effects of rejuvenating agents are partially superposed either with those of mix-dependent factors (including aggregate interlock) or with those that derive from the use of solvents (homogenization during recovery and presence of unknown residues) [5, 7, 9]. Hence, both procedures do not capture the peculiarities of different rejuvenators, which may diffuse to a different extent through the aged binder films, consequently affecting to different degrees their rheological and bonding characteristics [10]. This paper illustrates the preliminary results obtained by making use of a novel approach for the assessment of the effectiveness of rejuvenators, which relies on the testing of model systems made of a single-sized RA material. It is believed that such an approach, to be further validated and improved by means of additional testing, truly highlights the effects of rejuvenators on RA binders and may be functional for the ranking and selection of different products to be employed in paving applications.
2 Materials and Methods Model systems considered in the investigation were made of a single-sized 5/8 RA characterized by a binder content equal to 4.4% by weight of aggregates (EN 12697– 39) and a maximum density equal to 2.592 g/cm3 (EN 12697–5). This RA fraction was derived from a stockpile obtained from the milling of different bituminous layers of a national motorway. Particle size distribution of RA and of its extracted aggregates (EN 933–1) are shown in Table 1, where P is the percent passing and D is the sieve opening. Two rejuvenators, named A and B, were employed in the study. The first is a mixture of paraffinic hydrocarbon oils and the second is a non-toxic additive with a vegetable oil base. According to the manufacturers’ data sheets, A and B were characterized by a specific gravity at 20 °C in the range of 0.86–0.90 g/cm3 and of 0.98 g/cm3 , respectively; by a viscosity at 20 °C in the range of 34–45 mPa s and of approximately 500 mPa s, respectively; and by a flash point temperature greater than 220 °C and of about 150 °C, respectively. Both products, liquid at ambient temperature and sprayed cold on hot RA (preheated at 150 °C), were dosed at 8% Table 1 Particle size distributions of RA and extracted aggregates D (mm)
8
6.3
5.6
5
4
2
1
0.5
0.063
RA
P (%)
100
56
28
2
–
–
–
–
–
Aggregates
P (%)
100
76
57
39
28
20
15
12
5.5
Testing of Reclaimed Asphalt Model Systems for the Evaluation …
1753
by weight of RA binder, corresponding to 0.34% by weight of RA. A virgin neat bitumen (of the 50/70 penetration grade) was also employed in the investigation for the manufacture of reference samples as discussed in the following. Model systems were prepared with the single-sized RA material, either in its original state or pretreated with a rejuvenator, which was compacted to a target value of the voids in mineral aggregates (VMA), set at 30%; subsequent tests were carried out in the indirect tensile configuration for the assessment of stiffness and strength. The rationale of such an approach relies on the fact that the use of a single-sized RA material compacted to a very high VMA value is believed to yield model systems with a structure in which the response under indirect tensile stresses is mainly dependent upon the bonding of the binder phase, ultimately affected by employed rejuvenators, with limited effects due to aggregate interlock. The absence of agglomerations and fines in the RA material also eliminates any additional phenomena which may occur during heating and compaction, thus affecting the results of volumetric and mechanical tests. It is essential to underline that model systems do not reproduce, by any means, bituminous mixtures constituted by 100% RA, which are conceived with a continuous grading and exhibit a response under loading similar to that of standard asphalt mixtures. Model systems were prepared at 150 °C by mean of the gyratory shear compactor (EN 12697–31) with a fixed geometry (100 mm diameter, 65 mm height), suitable for mechanical tests. Target density corresponding to 30% VMA was derived from preliminary compaction tests carried out at the same temperature, with the consequent identification of the quantity of material needed for the preparation of final specimens. The investigation considered three types of model systems. The so-called “black system” (BS), constituted by RA with no rejuvenator; the “rejuvenated systems” (RSA and RS-B) prepared making use of RA pretreated with rejuvenator A or B; and the “white system” (WS), also referred to as reference system, obtained by combining the aggregates extracted from the RA with the previously mentioned virgin neat bitumen (with a dosage equal to that of the RA). The structure common to all the considered model systems (with constant and controlled volumes of solid aggregate skeleton and binder) is schematically represented in Fig. 1a, while the corresponding phase distribution, variable from one type of model system to the other, is displayed in Fig. 1b. Following their preparation, model systems were cured for 3 days at 20 °C and thereafter subjected to tests for the determination of indirect tensile stiffness modulus, ITSM (EN 12697-26C) and indirect tensile strength, ITS (EN 12697–23). In order to capture the time-dependent effect of rejuvenators, which originates from their progressive diffusion through the aged RA binder, further tests were carried out for the determination of ITSM after 7, 10, 15 and 52 days of curing, and for the determination of ITS after 52 days of curing. Experimental data recorded during ITS tests were processed in the usual fashion, for the determination of strength, and with the assessment of normalized tensile stresses (given by the actual calculated tensile stresses divided by the maximum value at failure) as a function of calculated compressive strains [11, 12]. Such a representation was functional for the evaluation of the ductility toughness index (DTI), calculated as the integral of the normalized
1754
D. Dalmazzo et al.
VMA
Aggregates
(b)
(a)
BS
RS
WS
Absorbed binder Effective BS/RS binder Rejuvenator Effective WS binder Air Voids
Fig. 1 Structure of model systems a and corresponding phase distributions b
curve in the post-peak region (considered to extend up to a strain of 2E-2 mm/mm from the peak), believed to be a parameter that directly depends upon the behavior of the binder present in the model system.
3 Results and Discussion
E (MPa)
Average results obtained from ITSM tests performed at different curing times on the model systems are shown in Fig. 2. As expected, the BS displayed the highest stiffness moduli as a result of the presence of the aged stiff RA binder. When the RA binder was replaced by the virgin neat bitumen (in the case of the WS), stiffness values decreased significantly, with an approximate reduction of 50%, regardless of curing time. It is interesting to observe that use of the two rejuvenators seemed to restore the stiffness properties of the RA binder with corresponding stiffness moduli of the RS-A and RS-B very close to those of the reference WS. It was also noticed that the two rejuvenated systems did not display a clear time-dependency of stiffness, with variations around the mean value which were similar to those exhibited by the 4500 4000 3500 3000 2500 2000 1500 1000 500 0
BS RS-A RS-B WS
3
7
10
15
Days Fig. 2 Stiffness moduli of model systems as a function of curing time
52
Testing of Reclaimed Asphalt Model Systems for the Evaluation …
1755
other two systems (BS and WS). Such an outcome was explained by considering the very small strain levels imposed during ITSM tests, which may not be sufficient to highlight the progression of diffusion phenomena within the bitumen films covering the aggregate particles in RA. As shown in Fig. 3a, average indirect tensile strength values obtained from ITS tests were consistent with those of ITSM tests in terms of the relative ranking of the model systems. In fact, the BS exhibited the highest ITS, while the WS, RS-A and RS-B displayed significantly lower values, quite close to each other. However, it was also observed that the presence of rejuvenators clearly increased the compressive strain at failure (Fig. 3b). As in the case of ITSM results, the time-dependency of the rejuvenated systems could not be appreciated. Once again, this was interpreted by considering the limited strain values reached at failure. Fig. 3c displays the DTI values of the considered model systems. As expected, the stiff and more brittle binder contained in the BS led to the lowest DTI values, while the highest values were recorded for the system containing virgin binder (WS). Values associated to the rejuvenated systems (RS-A and RS-B) were intermediate between those of the two other types of model systems. Analysis of the DTI seems to provide interesting information on the true rejuvenating capability of the employed products, thereby allowing a comparison among them. In particular, obtained results suggest that both rejuvenators had an almost immediate diffusion through the aged binder films, leading to a very high values of 1.2
4E-02
Strain (mm/mm)
ITS (MPa)
1.0 0.8 0.6 0.4 0.2 0.0
2E-02 1E-02 0E+00
3 days
(a)
3E-02
52 days
(b)
3 days
52 days
DTI (-)
1.0 0.8
BS
0.6
RS-A
0.4
RS-B
0.2
WS
0.0
(c)
3 days
52 days
Fig. 3 ITS a, compressive strain at peak b, and DTI values c of model systems as a function of curing time
1756
D. Dalmazzo et al.
DTI after 3 days of curing, which were only slightly improved in the long term (after 52 days). The outcomes synthesized above, to be considered with caution as a result of the limited data set, can be explained by referring to the high strain level imposed in post-peak conditions and with the corresponding loss of residual interlocking. In such conditions, the response of the binder contained in a model system is bound to be mainly affected by its cohesion, which should also reflect ongoing rejuvenating phenomena.
4 Conclusions Based on the results presented in this paper, it can be concluded that the proposed approach for the evaluation of the effectiveness of RA rejuvenating agents, which relies on the preparation and testing of model systems, is extremely promising. In particular, it presents the advantage of avoiding the use of solvents for the extraction of aged binders, thereby allowing the assessment of their response under loading with no alteration of their structure and composition. In such a context, it was shown that valuable information can be obtained by focusing on the post-failure response of the model systems when loaded in the indirect tensile mode. Further studies are certainly needed in order to improve and validate the proposed approach. As a first step, tests should be carried out by exploring, in a systematic fashion, the effects associated to the use of RA of different age and origin, and to the adoption of variable percentages of employed rejuvenators.
References 1. Asphalt Institute: MS-2 Asphalt Mix Design Methods. 7th edn. USA (2014) 2. Al-Qadi, I.L., Elseifi, M., Carpenter, S.H.: Reclaimed asphalt pavement—a literature review. Research Report FHWA-ICT-07–001. ICT (2007) 3. Zaumanis, M., Mallick, R.B., Frank, R.: 100% recycled hot mix asphalt: A review and analysis. Resour. Conserv. Recy. 92, 230–245 (2014) 4. Mogawer, W., Austerman, A., Roque, R., Underwood, S., Mohammad, L., Zou, J.: Ageing and rejuvenators: Evaluating their impact on high RAP mixtures fatigue cracking characteristics using advanced mechanistic models and testing methods. Road Mater Pavement 16, 1–28 (2015) 5. Shen, J., Amirkhanian, S., Aune Miller, J.: Effects of rejuvenating agents on superpave mixtures containing reclaimed asphalt pavement. J. Mater. Civil Eng. 19, 376–384 (2007) 6. Im, S., Zhou, F., Lee, R., Scullion, T.: Impacts of rejuvenators on performance and engineering properties of asphalt mixtures containing recycled materials. Constr. Build. Mater. 53, 596–603 (2014) 7. Zaumanis, M., Mallick, R.B., Poulikakos, L., Frank, R.: Influence of six rejuvenators on the performance properties of Reclaimed Asphalt Pavement (RAP) binder and 100% recycled asphalt mixtures. Constr. Build. Mater. 71, 538–550 (2014)
Testing of Reclaimed Asphalt Model Systems for the Evaluation …
1757
8. Zaumanis, M., Mallick, R.B., Frank, R.: Evaluation of different recycling agents for restoring aged asphalt binder and performance of 100 % recycled asphalt. Mater. Struct. 48, 2475–2488 (2015) 9. Tauste, R., Moreno-Navarro, F., Sol-Sánchez, M., Rubio-Gámez, M.C.: Understanding the bitumen ageing phenomenon: A review. Constr. Build. Mater. 192, 593–609 (2018) 10. Carpenter, S.H., Wolosick, J.R.: Modifier influence in the characterization of hotmix recycled material. Transp. Res. Record. 777, 15–22 (1980) 11. Huang, B., Shu, X., Li, G.: Laboratory investigation of portland cement concrete containing recycled asphalt pavements. Cement Concrete Res. 35, 2008–2013 (2005) 12. Modarres, A.: Investigating the toughness and fatigue behavior of conventional and SBS modified asphalt mixes. Constr. Build. Mater. 47, 218–222 (2013)
The Bitumen Related Problems Observed on Soft Asphalt Concrete Pavements: Case Study Michalina Makowska and Katri Eskola
Abstract Four roads paved with soft asphalt concrete during the autumn of 2017 by one contractor in a Nordic country, damaged during the following winter. The road users reported that bitumen detached from the surface and was sticking to the fur of pet animals, as well as to the surfaces of the vehicles. The article discusses aspects connected with production and paving, which may have affected the performance. In addition, the material analysis of the samples collected from the field was performed. The core samples were tested by Indirect Tensile Strength (ITS). Because of the low strength of the soft asphalt concrete the ITS is not routinely performed for this material. Therefore, the results of ITS were inconclusive due to the lack of reliable references. The gradation was analyzed after the bitumen extraction and recovery, and no abnormalities were discovered. The Zero Shear Viscosity at multiple temperatures and fractional composition was performed on both extracted and reference binders type V1500 and 650/900 (EN 12591) after evaluation of bitumen purity with infrared spectroscopy. The bitumen extracted from the most questionable pavement expressed comparable rheological behavior to references, but was characterized by the exceptionally high content of saturates. The short research project was unable to determine if the damage was a result or the reason for the altered binder composition. Nevertheless, the link between abnormal performance observed in the field and abnormal chemical characteristic of bitumen is reported. Keywords Viscosity · Fractionation · Damage · Soft Asphalt Concrete
1 The Peculiar Case in Southwest Finland The soft asphalt concrete (PAB), which replaced the oil gravel in the 1990, is especially used in the northern regions of Europe on low volume roads [1]. It is usually M. Makowska (B) Aalto University School of Engineering, 11000, 00076 Aalto, Finland e-mail: [email protected] K. Eskola Finnish Transport Infrastructure Agency, 33, 00521 Helsinki, Finland © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_223
1759
1760
M. Makowska and K. Eskola
composed of lowest abrasion resistance class aggregate and very soft bitumen, e.g. V1500 (PAB-V) or 650/900 (PAB-B). In the autumn of 2017, twenty roads in the Southwest Finland region were overlaid with soft asphalt concrete (PAB). Of those twenty roads, three were paved using PABV, seventeen—using PAB-B. In January 2018 a series of complaints from the road users surfaced and was reported in press as disturbing. The bitumen from the asphalt on road number 2410 (PAB-V mixture) seemingly detached from the road and stained cars, traffic users, windows of the houses near the road, as well as pets [2–4]. The users demonstrated how swiping the surface with a sponge resulted in transfer of the black matter onto the material [2]. After the publication of the article, additional complaints were reported to authorities about three roads paved within the same paving season within the same contract agreement, namely roads 1900 (PAB-B), 12137 (PAB-V) and 12029 (PAB-B) [5]. During and after the paving season of 2017 a significantly higher rainfall occurred. Additionally, some of the quality control information were not reported, e.g. performance of the EN 12697–12 (method C) to establish the moisture resistance of mixture. The aim of hereby research was on establishing if there was anything exceptional about the raw materials used for the construction, as well as to compare the mixture with guidelines.
2 Materials and Methods Four reported roads (Table 1) were chosen for the analysis. From each of the roads five core samples of 100 mm were collected in April 2018, approximately six months from laydown. The tests applied on the collected bituminous mixture included: the soluble binder content (SFS-EN 12697–1); binder recovery (12697–3); the determination of particle distribution (SFS-EN 12697–2); thickness of specimen (SFS-EN 12697–36); Indirect Tensile Strength (EN 12697–23) (in air, 4 h conditioning 10 °C, MultiPlex 50-E); the filler density for the fraction below 0.125 mm grain size (SFS-EN 1097– 7); the apparent density of mineral aggregate from 0,125-4mm fraction of aggregate, as well from 4 to 20 mm fraction (SFS-EN 1097–6). Unfortunately, the air void content or asphalt mixture densities were not recorded due to miscommunication, therefore the voids filled with asphalt (VFA) remain undetermined. As a reference for the binders extracted from PAB pavements, the V1500 bitumens stored for rejuvenation studies in 2015, 2017 and 2018 were used. Due to the lack of previous knowledge on the effect of extraction on the rheology of such binders, a reference was prepared. From each V1500 reference, a 100 g was dissolved in 500 ml of dichloromethane and recovered using the binder recovery procedure (SFS-EN 12697–3). The V1500 bitumens (reference and recovered) were tested for rotational viscosity measurements, (MCR 302 with H-PTD200, Anton Paar, Austria) using the coneplate (CP) geometry of 50 mm diameter (cone truncation 0.1 mm, CP angle 1° [ISO
The Bitumen Related Problems Observed on Soft Asphalt …
1761
Table 1 The information from the tender agreement about the four reported roads, where MPoverlay, LTA—overlay on top of the full depth reclamation performed on old pavement, AADT— Average annual daily traffic, AN —resistance to abrasion from studded tyres as specified in EN 13043 (category based on Nordic abrasion value), FI- shape of coarse aggregate as specified in EN 13043 (category based on flakiness index), D-very low trafficked road Road No
AADT
Ordered pavement
Aggregate properties
Quality control class for road [1]
1900
1253
PAB-B 16/MP
AN 19, FI35
D
2410
698
PAB-V 16/LTA
AN 19, FI35
D
12029
999
PAB-B 16/MP
AN 19, FI35
D
12137
495
PAB-V 16/MP
AN 19, FI35
D
3219:1995] at shear rates of 0.01–100 [1/s] between temperatures of 20–100 °C at 10 °C intervals, with additional measurement at 25 °C [6]. Thin layer chromatography-based fractionation of bitumen (IP469) [7] was performed on reference and recovered binders, using 5 rods per sample.
3 Results The gradation curves indicated that the particle distribution for the road 2410 was in fact within the boundaries set for PAB 16 (Fig. 1). The other three roads have seemingly finer composition. The binder content suggested for the PAB-B mixtures should be within 4–4.5%, while 3.2–3.7% for PAB-V mixtures. Considering those values and slightly elevated density of fines in the road 2410 (Table 2), we could postulate that the binder in the mixture may have been in an excess. The ITS of the cores was small. There was no recent experience on the PAB pavement strength. Therefore, the analysis of this property is challenging. In the
Fig. 1 Gradation curves for considered roads against the envelope of PAB 16
1762
M. Makowska and K. Eskola
studies on the use of emulsion as binder in PAB-V [8] the average ITS measured at 10 °C for fresh PAB-V samples was between 99 and 130 kPa depending on the aggregate, while the bitumen emulsion bound PAB had 44 kPa. The tests after 150 days in field resulted in the range of values between 233 and 299, and 149 kPa for bitumen emulsion bound PAB. A different study investigating the effect of fluxing oils on the PAB performance reports values between 103 and 160 kPa for one-monthold samples of PAB prepared with biofluxed bitumen [9]. Therefore, the samples collected for hereby study from PAB-V roads after approximately six months from paving operation were weaker than the old experience would suggest (Table 2). Additionally, laboratory staff reported staining of the hands. The analysis of the rotational viscosity indicated that all samples expressed the shear rate independent behavior above 40 °C. For simplicity the viscosity at 0.1 [1/s] shear rate was compared across the temperatures. The bitumen extracted from 2410 was softer than the bitumen extracted from 12137 road, yet not softer than the reference and extracted reference V1500 bitumens (Fig. 2). The reference binders have been in cold storage for 6 months to 3 years, their viscosity may have increased during storage and reheating for sampling. However, the viscosity at 60 °C for those reference materials was below 2000 mPas, which is the upper limit for the V1500 grade. The bitumens from road 2410 and 12137 were extracted six months after initial laydown. It is expected that their grade would increase by one due to the short-term aging in field. The viscosity value at 60 °C indicates values in line with V3000 grade. The fractional analysis of the reference and extracted bitumens suggests that the effect of extraction is within the repeatability limit of the IP469 procedure (Table 3). Due to the fact that the repeatability and reproducibility limits for IP469 set in the procedure description are allowing quite significant deviations between the samples the comparison is challenging. The only fraction in which the difference between the result and the average of references was beyond the reproducibility level, was in the case of saturate content for 2410 road (Fig. 3). Based on that results it is expected that the bitumen would be significantly softer. However, the area of the saturate peak in chromatogram was similar for all V1500 samples. It was the area of other fractions which was lower. Due to the fact of using similar sample mass and application of 1 µl spot on the rod, the mass of saturates should be linked to the area of the saturate peak. The reason behind the differing split Table 2 Aggregate density, indirect tensile strength, binder content as well as height of core for roads in question Density (ton/m3 )
ITS at 10 °C
Pb
H
Road No
3
Asphaltene content (wt %) Mass loss at 163 °C, 5 h (%)
>50 EN13303
40.08
The Impact of Binder Type on Pavement Design …
1809
Table 2 Aggregate gradation of the investigated asphalt mixes % passing
Sieve size (mm) 19
9.5
4.75
2.36
0.3
0.075
Pure
95
66
40
25
8.1
4.7
8% selenizza
93
65
38
24
8.9
5
12% selenizza
95
70
46
26
8.4
4.5
Table 3 Asphalt binder properties and mixes volumetric characteristics Type of asphalt
PEN
Softening point
Vα (%)
Vbeff (%)
Pure
32
56
3.3
11.6
8% selenizza
36
54
3.6
11.3
12% selenizza
32
56
3.1
11.5
The thickness of the asphalt layer ranges from 14 to 19 cm. Table 2 presents the gradation of the asphalt mixes considered, while Table 3 shows the binder characteristics (penetration index-PEN and softening point) and volumetric properties (air void percentage-Va and effective binder content by volume- Vbeff ) of the asphalt mixes investigated. The aggregate gradation is typically the same in the three mixes and within the upper and lower values of the standards. The purpose of producing the modified binders was twofold. On the one hand, an effort was made to improve the characteristics and the performance of the 50–70 PEN asphalt with the addition of a modifier. On the other hand, the amount of the selenizza additive was investigated in order to achieve the same mechanical characteristics and performance with the pure 30–45 PEN asphalt. Additives tend to decrease the penetration value and increase the softening point value. As seen in Table 3, with the addition of 12% selenizza in 50–70 PEN asphalt, the asphalt binder properties (penetration and softening point) match those of 30–45 PEN asphalt. Four specimens from each mixture were compacted with the gyratory compactor according to EN standards (EN 12697-31 2007). Initial dimensions of the specimens were 170 mm high and 150 mm diameter in accordance with related standards. The specimens were cored in order to get specimens of 150 mm high and 100 mm diameter according to the related standards. E* testing was performed according to AASHTO standards (T 342-11). A controlled sinusoidal (harvesine) compressive loading was applied to each specimen for a range of loading frequencies and temperatures. Testing was conducted at six temperatures 4, 15, 20, 25, 37 °C and six frequencies, 25, 10, 5, 1, 0.5 and 0.1 Hz. The mean value of the E* of the four replicates is considered to be the representative modulus for each mixture. According to the methodology described in the MEPDG the field calibrated model for calculating total permanent deformation is shown in Eq. (1).
1810
K. Georgouli et al.
p(HMA) = εp(HMA) × hHMA k = β1r × kz × εr(HMA) × 101r × nk2r × β2r × T3rk × β3r
(1)
where p(HMA) is the accumulated permanent or plastic vertical deformation in the HMA layer (in), εp(HMA) the accumulated permanent or plastic axial strain in the HMA layer (in/in), εr(HMA) the resilient or elastic strain calculated by the structural response model in the mid-depth of each HMA sublayer (in/in), hHMA the thickness of the HMA layer (in), n the number of axle-load repetitions, T the mix or pavement temperature (°F), k1r ,2r ,3r global field calibration parameters (k1r = −3.35412, k2r = 0.4791, k3r = 1.5606) and β1r , β2r , β3r local or mixture field calibration constants all set to 1.0, kz the depth confinement factor. Equal to kz = (C1 + C2 × D) × 0.328196D , C1 = −0.1039 × (HHMA )2 + 2.4868 × HHMA − 17.342, C2 = 0.0172 × (HHMA )2 − 1.7331 × HHMA + 27.428, where D is the depth below the surface (in) and HHMA the total HMA thickness (in). The viscoelastic analysis of the investigated pavement sections was performed utilizing the 3-D Move software program. The number of traffic repetitions used in the analysis was calculated at 2.9 × 107 ESALs for 20 years. The dynamic analysis was performed considering a moving vehicle with 50 km/h.
3 Data Analysis and Results Figure 1 presents the E* master curves with respect to the reduced frequency (fr ) for the three asphalt mixes at a reference temperature of 25 °C. It is evident that E* values of mixes with pure asphalt are higher, while those of mixes with 8% selenizza are lower. For example, for loading frequency 10 Hz (logfr 1 Hz), E* value of asphalt mixes with pure asphalt, 12% selenizza and 8% selenizza is about 10,000 MPa, 8,000 MPa and 6,000 MPa, respectively. This notification may be of great importance in the pavement design process for the determination of the asphalt
Fig. 1 E* master curves
The Impact of Binder Type on Pavement Design …
1811
Fig. 2 Vertical strains at the bottom of the asphalt layers a hHMA = 14 cm and b hHMA = 19 cm
layer thickness, depending on the limiting values of the performance criteria (i.e. rutting depth). Figure 2a, b shows the developed vertical strains at the bottom of the asphalt layer and at the center of the load for 14 cm and 19 cm asphalt layer thickness (hHMA ), respectively. Strains reduce with the increase of the asphalt layer thickness in every case. Thus, strains developed in case of the asphalt mix with 12% selenizza are lower, while in case of the asphalt with 8% selenizza are higher. This finding leads to the results of Fig. 3 that shows the estimated rutting depth (RD) for each mix versus the various asphalt layer thicknesses considered. Estimated rut depth in case of the asphalt mix with 12% selenizza is lower, while in case of the asphalt with 8% selenizza is higher. Figure 4 shows the differences of the estimated RD of the asphalt mixes with modified binder with respect to the mixes with pure asphalt, in terms of Root Mean Square Error (RMSE) (expressed in percentage). Considering the above it can be concluded that asphalt mixes with 12% selenizza, appear to have the lower rut depth values, but still close enough to mixes with pure asphalt.
4 Conclusions Analysis results showed that mixes with pure asphalt appear to have higher E* values within the margins of loading conditions considered in this study. However, strain analysis showed that the lower values of vertical strains developed within the asphalt layer, for every thickness considered, correspond to mixture with 12% selenizza.
1812
K. Georgouli et al.
Fig. 3 RD versus asphalt layer thickness
Fig. 4 RMSE (%) values
As such, asphalt mixes with 12% selenizza appear to have lower rut depth values, despite the fact that E* values are lower compared to those of pure asphalt mixes. In light of the above, selenizza additive in a percentage of 12% could be used as an alternative in case the production of pure (30–45 PEN) asphalt is not feasible. The increased cost of using modified binder could be equilibrated in terms of better pavement performance (in terms of permanent deformation) and eventually lower maintenance costs.
The Impact of Binder Type on Pavement Design …
1813
References 1. Oscarsson, E.: Mechanistic-Empirical Modeling of Permanent Deformation in Asphalt Concrete Layers. Doctoral Thesis, Department of Technology and Society, Lund University, Sweden (2011) 2. Sousa, J.B., Craus, J., Monismith, C.L.: Summary Report on Permanent Deformation in Asphalt Concrete, Report SHRP-A/IR-91-104, Strategic Highway Research Program (1994) 3. NCHRP.: Guide for Mechanistic-Empirical Design of New and Rehabilitated Pavement Structures Appendix gg-1: Calibration of Permanent Deformation Models for Flexible Pavements (2004) 4. Fontes, L.P.T.L., Trichês, G., Pais, J.C., Pereira, P.A.A.: Evaluating permanent deformation in asphalt rubber mixtures. Constr. Build. Mater. 24, 1193–2000 (2010) 5. Suo, Z., Wong, W.G.: Nonlinear properties analysis on rutting behaviour of bituminous materials with different air void contents. Constr. Build. Mater. 23, 3492–3498 (2009) 6. Huang, X.M., Zhang, Y.Q.: A new creep test method for asphalt mixtures. Road Mater. Pavement Des. 11(4), 969–991 (2010)
The Influence of Cement Type on Early Properties of Cold In-Place Recycled Mixtures Bohdan Doł˙zycki , Mariusz Jaczewski , and Cezary Szydłowski
Abstract Cold in-place recycling is a commonly used maintenance treatment in rehabilitation of low and medium volume roads in Poland. Typically, two types of binding agents are used—cement and bituminous emulsion (or foamed bitumen). Due to the harsh Polish climate with many freeze/thaw cycles and frequent occurrence of saturated conditions, the used amounts of cement are higher than those commonly used in warmer parts of Europe. While there is usually only one type of bituminous emulsion dedicated for cold recycling on the market, there are numerous types of cements, which differ in chemical composition and properties. The conducted research presents possible development of cold recycled mixture properties over curing time, taking into account the type of cement used. Two types of cement were tested in laboratory investigation—common Portland CEM I 32.5 R cement and Portland-fly ash CEM II 32.5 B-V cement with longer setting time. Cold recycled mixtures were designed with the same composition and amount of binding agents, but differed in the type of cement used. For both mixtures, indirect tensile strength and modulus were tested after 7, 28 and 90 days of curing in laboratory conditions. The laboratory tests confirmed lower values of strength and modulus for the fly ash cement after 7 and 28 days in comparison to the typical cement, but after 90 days the properties of both tested mixtures presented similar values. If the overall predicted fatigue life and long-term mechanical properties are the same, the use of slow-setting cements may result in reduction of reflective cracking on the surface of the pavement. In the case of low and medium volume roads, where there is no need for fast paving of the asphalt layers and more time may be allowed for the cold-recycled mixture to achieve the required initial strength, slow-setting cements should be considered as a viable treatment for reduction of the risk of reflective cracking.
B. Doł˙zycki (B) · M. Jaczewski · C. Szydłowski Gdansk University of Technology, Gdansk, Poland e-mail: [email protected] M. Jaczewski e-mail: [email protected] C. Szydłowski e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_230
1815
1816
B. Doł˙zycki et al.
Keywords Bitumen emulsions · Cement · Reclaimed asphalt · Indirect tensile testing · Cold recycling
1 Introduction Low and medium volume roads in Poland are often in poor condition. One of the typically used maintenance treatment in rehabilitation using cold in-place recycling [1, 2] with two binding agents—cement and bituminous emulsion (or foamed bitumen). In Poland the amount of cement used in-place recycling is higher than used in warmer parts of Europe [3, 4]. It resulted from experience with harsh climatic conditions characterized by numerous freeze/thaw cycles and frequent occurrence of saturated state. While on the Polish market there is usually only one type of bituminous emulsion dedicated for cold recycling, there are numerous types of cements, which differ in chemical composition and properties [5]. In most cases typical CEM I Portland cement is used for cold recycling [3, 4, 6–9], regardless of the region of usage. Only isolated studies [10, 11] sought to analyse the possibility of use of other types of cement, including slow-setting cement. The domination of CEM I Portland cement is commonly related to the requirements and the construction schedule considerations. The conducted research was designed to evaluate the possibility of obtaining a cold recycled mixture that can develop its properties over a longer time of curing, without overall negative effect during the construction and service phase. Classification of cement is based on its main constituents. In Poland, the most commonly used cement type for cold recycled mixtures is the Portland cement CEM I, which consists mostly of clinker (95–100%). The EN 197-1 standard also allows the use of Portland-composite cements, that may contain blastfurnace slag, silica fume, pozzolana, fly ashes or limestone. They are classified as CEM II main type. Cements with lower clinker content, such as blastfurnace cement (CEM III) or pozzolanic cement (CEM IV), may be used as well [5]. The content of various constituents affects the setting time, the rate of strength development and the final strength of the cement. The type of cement also affects the heat of hydration during setting. While the use of CEM I cements ensures higher early strength of the mixture, it may also lead to higher number of shrinkage cracks, due to higher heat of hydration and higher shrinkage, as compared to mixtures that contain CEM II type cements, which exhibits a lower rate of strength development, and, consequently, lower shrinkage and lesser tendency towards initiation of transverse cracks.
2 Materials and Methods Six different cold recycled mixtures with two different types of cement—Portland CEM I 32.5 R cement and Portland fly-ash CEM II 32.5 B/V cement (containing
The Influence of Cement Type on Early Properties …
1817
Table 1 Composition of the tested cements [5] Cement
Main constituents Clinker (%)
CEM I
95–100
CEM II B/V
65–79
Minor additional constituents (%) Siliceous fly ash (%) –
0–5
21–35
0–5
clinker and siliceous fly ash)—were tested in the study (three different mixtures for each type of cement). The composition of the tested cements is given in Table 1. The cold recycled mixtures were composed of: 70% reclaimed asphalt pavement, 15% virgin fine aggregate and 15% virgin coarse aggregate. The influence of cement was assessed for a set amount of emulsion (E = 4%) and three different amounts of cement (C = 1, 2 or 3%). The properties and compositions of the tested mixtures are presented in Table 2. Mixtures were described with abbreviations taking into account the content of binding agents (e.g. C1E4 meaning 1% cement and 4% emulsion). For the purpose of the analysis, two laboratory test were selected: Indirect Tensile Stiffness Modulus Test (ITSM) according to EN 12697-26 (load time: 124 ± 4 ms, target horizontal deformation: 5 µm) and Indirect Tensile Strength (ITS) according to EN 12697-23 (displacement rate: 50 mm/min). The temperature of the test was Table 2 Tested materials Property
Mixture designation* C1E4 (a)
C2E4 (a)
C3E4 (a)
C1E4 (b)
C2E4 (b)
C3E4 (b)
Gradation (mm), (%) 31.5
100.0
100.0
100.0
100.0
100.0
100.0
16
86.1
86.1
86.1
86.1
86.1
86.1
8
54.1
54.1
54.1
54.1
54.1
54.1
4
36.8
36.8
36.8
36.8
36.8
36.8
2
27.6
27.6
27.6
27.6
27.6
27.6
1
20.1
20.1
20.1
20.1
20.1
20.1
0.5
12.1
12.1
12.1
12.1
12.1
12.1
0.125
2.6
2.6
2.6
2.6
2.6
2.6
0.063
1.3
1.3
1.3
1.3
1.3
1.3
Cement content (%)
1
2
3
1
2
3
Emulsion content (%)
4
4
4
4
4
4
Air voids (%)
10.4
11.9
10.4
12.4
12.4
11.9
Type of cement
CEM I 32.5R
* Cold
CEM II B/V 32.5R
recycled mixtures: (a) with CEM I 32.5 R, (b) with CEM II B/V 32.5 R
1818
B. Doł˙zycki et al.
selected as +5 °C. The tests were conducted for three curing ages: 7 days, 28 days and 90 days after compaction. In each case three separate specimens were tested.
3 Results and Discussion The results of the stiffness modulus and strength tests for all the tested mixtures are presented in Table 3 as mean values with coefficients of variation (CoV), which reflect the scatter of the results. The tests were conducted 7, 28 and 90 days after compaction of the specimen. As shown in Table 3, a properly designed cold recycled base should contain at least 2% of CEM I cement or at least 3% of CEM II B/V cement. High requirements for Poland [12] regarding stiffness modulus and strength are related to harsh climatic conditions, under which from November until April the pavement is subjected to frost and numerous freeze/thaw cycles in water-saturated state. Moreover, a layer constructed from cold recycled material should exhibit mechanical properties sufficient to enable further paving works as soon as possible. The aforementioned conditions result in high amounts of required cement. While the relatively high strength requirements and usage of CEM I cement with high clinker content are forced by the climatic conditions, they consequently lead to occurrence of reflective cracking [1]. Other cements, which are less susceptible to shrinkage cracking (CEM II), do not always enable obtaining of satisfactory shortterm strength parameters (after 7 or 28 days), when used in the same quantities as CEM I. In order to gain insight into their performance in longer perspective, the third test was performed 90 days after compaction. It proved that after the longer period both cement types exhibited comparable results, despite the considerable differences after 7 and 28 days. To compare the development of mechanical properties in time for both types of cement, the results in Table 4 are presented as ratio compared to the values for 28 days of curing. In the first case, values were compared to those obtained for the same type of cement after 28 days. In the second case, the results for CEM II were compared to those obtained for CEM I cement after 28 days. The development of mechanical properties in the presented study for both types of cement is similar: after the initial rapid build-up it slows down after 28 days, reaching the maximum value after 90 days. In the case of stiffness modulus, while the development process is the same for both cements (around 60–70% after 7 days and around 130–170% after 90 days) the absolute values for CEM II B/V cement are much lower. Mixtures containing this cement needed almost 90 days to reach the stiffness modulus values that CEM I mixtures reached after 28 days of curing. As for strength, initial increase is much faster for CEM I cement, which obtains almost its final strength (80–90% of strength) as soon as after 7 days. In the case of mixtures with CEM II B/V cement, the strength values after 7 days were much lower; only after 90 days they reached levels comparable do those obtained by the CEM I mixtures.
The Influence of Cement Type on Early Properties …
1819
Table 3 Laboratory test results Property
Mixture designationa C1E4 (a)
C2E4 (a)
C3E4 (a)
C1E4 (b)
C2E4 (b)
C3E4 (b)
Requirements [1] –
ITSM, +5 °C 7 days, mean, MPa
2279
3999
5767
1769
2262
3397
7 days, CoV, %
12.7
5.9
4.6
18.7
8.0
5.2
28 days, mean, MPa
3022
6012
8792
2215
3368
6948
2000 –700
28 days, CoV, %
1.7
5.8
1.5
26.0
13.1
5.3
0
90 days, mean, MPa
5137
8177
11,912
3178
5716
10,286
–
90 days, CoV, %
5.8
1.4
9.1
20.4
13.4
8.8
7 days, mean, MPa
0.42
0.71
0.85
0.47
0.47
0.57
7 days, CoV, %
13.9
8.1
7.5
2.2
8.9
12.1
28 days, mean, MPa
0.46
0.77
1.08
0.47
0.64
0.88
28 days, CoV, %
9.7
6.9
3.2
3.8
3.1
7.2
90 days, mean, MPa
0.71
1.07
1.52
0.63
0.79
1.39
90 days, CoV, %
2.1
4.2
4.6
14.0
22.1
6.0
ITS, +5 °C
a Cold
0.5–1.0
0.7–1.6
–
recycled mixtures: (a) with CEM I 32.5 R, (b) with CEM II B/V 32.5 R
4 Conclusions Laboratory tests after 7 and 28 days confirmed lower values of strength and modulus for Portland-fly ash cement CEM II B/V in comparison to typical Portland cement CEM I, but after 90 days the properties of both tested mixtures presented similar values. The performed tests show that it is possible to obtain the same level of strength
ITS +5 °C
ITS + 5 °C
ITSM +5 °C
ITSM +5 °C
Property
1.00
1.70
28 days
90 days
1.35
CEM II/CEM I(28 days) ratio
1.32
90 days
1.40
1.00
1.00
1.02
1.39
1.00
28 days
1.53
90 days
1.00
1.02
1.00
28 days
7 days
0.91
7 days
1.05 CEM II/CEM II(28 days) ratio
0.73
90 days
CEM II/CEM I(28 days) ratio
1.43
1.00
0.80
CEM II/CEM II(28 days) ratio
C1E4 (b)
28 days
0.79
1.35
1.00
0.66
C3E4 (a)
0.59
0.92
1.36
1.00
0.67
C2E4 (a)
7 days
CEM I/CEM I(28 days) ratio
0.75
7 days
CEM I/CEM I(28 days) ratio
C1E4 (a)
Mixture designation
Table 4 Properties as ratio in comparison to properties for 28 days of curing
1.02
0.83
0.61
1.23
1.00
0.74
0.95
0.56
0.38
1.70
1.00
0.67
C2E4 (b)
1.28
0.81
0.53
1.58
1.00
0.65
1.17
0.79
0.39
1.48
1.00
0.49
C3E4 (b)
1820 B. Doł˙zycki et al.
The Influence of Cement Type on Early Properties …
1821
parameters using both types of cement. The Portland-composite cement only requires longer setting time. Taking into account the shrinkage that occurs during setting, it is advisable to use Portland-composite cements, since their shrinkage is lower and, in consequence, the risk of cracking is decreased. If the overall predicted fatigue life and long-term mechanical properties are the same, the use of slow-setting cements may result in reduction of reflective cracking on the surface of the pavement. In the case of low and medium volume roads, where there is no need for fast paving of the asphalt layers and more time may be allowed for the cold-recycled mixture to achieve the required initial strength, slow-setting cements should be considered as a viable treatment for reduction of the risk of reflective cracking. Acknowledgments Part of the research was supported by the project RID-1A (DZP/RID-I06/1/NCBR/2016) financed by the National Centre for Research and Development and the General Directorate for National Roads and Motorways under the program “Development of Road Innovations”.
References 1. Doł˙zycki, B., Jaskuła, P.: Review and evaluation of cold recycling with bitumen emulsion and cement for rehabilitation of old pavements. J. Traffic Transp. Eng. 6(4), 311–323 (2019) 2. Chomicz-Kowalska, A., Stepien, J.: Cost and eco-effective cold in-place recycled mixtures with foamed bitumen during the reconstruction of a road section under variable load bearing capacity of the subgrade. Procedia Eng. 161, 980–989 (2016) 3. Bocci, M., Grilli, A., Cardone, F., Graziani, A.: A study on the mechanical behaviour of cement–bitumen treated materials. Constr. Build. Mater. 25, 773–778 (2011) 4. Meocci, M., Grilli, A., La Torre, F., Bocci, M.: Evaluation of mechanical performance of cement–bitumen-treated materials through laboratory and in-situ testing. Road Mater. Pavement Des. 18(2), 376–389 (2017) 5. Peukert, S.: Cementy powszechnego u˙zytku i specjalne. Polski Cement, Kraków (2000) 6. Kavussi, A., Modarres, A.: A model for resilient modulus determination of recycled mixes with bitumen emulsion and cement from ITS testing results. Constr. Build. Mater. 24, 2252–2259 (2010) 7. Dolzycki, B., Jaczewski, M., Szydlowski, C.: The long-term properties of mineral-cementemulsion mixtures. Constr. Build. Mater. 156, 799–808 (2017) 8. Buczy´nski, P., Iwa´nski, M.: Complex modulus change within the linear viscoelastic region of the mineral-cement mixture with foamed bitumen. Constr. Build. Mater. 172, 52–62 (2018) 9. Raschia, S., Graziani, A., Carter, A., Perraton, D.: Laboratory mechanical characterisation of cold recycled mixtures produced with different RAP sources. Road Mater. Pavement Des. 20(sup1), 233–246 (2019) 10. Baghini M.S., Ismail A., Bin Karim, M. Rehan.: Evaluation of cement-treated mixtures with slow setting bitumen emulsion as base course material for road pavements. Constr. Build. Mater. 94, 232–336 (2015) 11. Shaowen, D.: Effect of curing conditions on properties of cement asphalt emulsion mixture. Constr. Build. Mater. 164, 84–93 (2018) 12. Doł˙zycki B.: Polish experience with cold in-place recycling, Bestinfra 2017. In: IOP Conference Series: Materials Science and Engineering 236, 012089 (2017)
The Influence of the Mastic Coating of Untreated Reclaimed Asphalt Pavement on the Permanent and Resilient Behaviours Laura Gaillard, Cyrille Chazallon, Pierre Hornych, Juan Carlos Quezada, and Jean-Luc Geffard Abstract Asphalt aggregates arise from the demolition of asphalt road layers. The ORRAP (Optimal Recycling of Reclaimed Asphalts in low-traffic Pavements) Project concerns the cold recycling of 100% Reclaimed Asphalt Pavement (RAP) without binder addition in base and subbase layers of low-traffic roads. In this context, a test program was performed with Rhine Region materials to evaluate the impact of the mastic coating of untreated RAP. A source of reclaimed asphalts with 4.4% of bitumen is tested before (RAP) and after binder extraction (RAP-BE), then is compared to an unbound granular material (UGM) with a similar particle size distribution. Repeated Load Triaxial (RLT) tests were conducted to study the permanent and resilient behaviours. The results show that the RAP before binder extraction and the UGM present similar permanent behaviours, while the aggregates without binder reveal low strains. Concerning the resilient phase, the RAP and the RAP-BE show similar levels of strains, but the resilient moduli of the UGM are significantly lower than those of the RAP. Keywords Reclaimed asphalt pavement · Mastic coating · Repeated load triaxial test · Mechanical behaviour
1 Introduction Many studies on recycling of asphalt aggregates have been carried out, in particular on the incorporation of Reclaimed Asphalt Pavement (RAP) in hot asphalt mixtures [1–3], in warm processes by means of additives or foamed asphalts [4, 5], as well as in cold recycling with foamed or emulsion bitumen [6, 7]. However, the use without any addition, developed in Sweden [8], is still not widespread. In France, the cold strategy allows to reuse RAP with high Polycyclic Aromatic Hydrocarbons (PAH) L. Gaillard (B) · C. Chazallon · J. C. Quezada ICUBE, UMR 7357, CNRS, INSA de Strasbourg, Strasbourg, France e-mail: [email protected] P. Hornych · J.-L. Geffard MAST-LAMES, Université Gustave Eiffel, IFSTTAR, Bouguenais, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_231
1823
1824
L. Gaillard et al.
Table 1 Characteristics of the RAP and the UGM RAP
RAP-BE
UGM
Grading (mm)
25 AE 0/14
0/14
0/14
Particle density (g/cm3 )
2.48
2.65
2.64
Proctor maximum dry density (g/cm3 )
2.02
2.21
2.16
Proctor optimum water content (%)
5.86
5.31
7.60
PAH content (mg/kg)
298
0
0
contents, rate up 500 mg/kg. The ORRAP (Optimal Recycling of Reclaimed Asphalts in low-traffic Pavements) project concerns the cold recycling of 100% RAP without binder addition in base and subbase layers of low-traffic roads, instead of Unbound Granular Materials (UGM). In this framework, the effect of the mastic coating is studied, to measure the benefit of this recycling. A series of Repeated Load Triaxial (RLT) Tests is carried out to analyse the permanent strains and the resilient moduli of a source of untreated RAP and to compare them with a conventional UGM.
2 Material and Test Programme 2.1 Material Characterisation The tested RAP comes from demolition of wearing courses of Rhine Region. This material is defined by 25 RAP 0/14, according to the French Standard NF EN 131088 [9], where 25 is the maximum size of particles before binder extraction and 0/14 the grading of particles after binder extraction. The RAP contains 4.4% of bitumen 10/20. This material is studied before (RAP) and after binder extraction (RAP-BE), then is compared to an unbound granular material (UGM) with a similar particle size distribution (Fig. 1). Table 1 gives the main characteristics of the RAP and the UGM. Unbound granular materials and asphalt mixtures represent the limits of the RAP behaviour. Figure 1 displays the UGM 0/14 mm and the asphalt mixture 0/14 mm grading curves of the French Standards [10, 11], where 0/14 mm is the grain fraction. The binder extraction releases fines, consequently the fines content (passing through the sieve 63 μm) is multiplied by four before and after extraction.
2.2 Sampling The specimens are compacted using a vibrocompression method in one layer, by applying a vertical load and a horizontal vibration. The cylindrical sample is 0.16 m diameter and 0.32 height. Table 2 presents the characteristics of different specimens.
The Influence of the Mastic Coating of Untreated Reclaimed …
1825
Fig. 1 Particle size distributions of the studied materials
Table 2 Characteristics of specimens
RAP Compaction density
(g/cm3 )
Void content
RAP-BE
UGM
1.96
2.14
1.96
0.27
0.24
0.35
Porosity (%)
21.1
19.29
25.85
Water content (%)
3.9
3.3
3.9
Number of specimens
3
2
2
2.3 Triaxial Apparatus The sample is placed in a triaxial cell, and a constant confining pressure is applied with air. A hydraulic actuator is used to apply a cyclic axial load. The specimen is instrumented with three Hall effect displacement transducers: two axial and one radial.
2.4 Repeated Load Triaxial (RLT) Tests The sign conventions are: • The stresses are positive in compression. • The axial and radial strains are positive in contraction. • A positive (negative) volumetric strain indicates a contracting (dilating) behaviour.
1826
L. Gaillard et al.
RLT test is one of the most reliable experiments to analyse the permanent and resilient behaviours of unbound granular materials. The test principle, defined by the French Standard EN 13286–7 [12], consists of applying a cyclic axial stress σ1 and a constant confining pressure σ3. Equations (1) and (2) define the deviator stress q and the mean stress p. During the test, the axial strains ε1 and the radial strains ε3 are measured, then the deviatoric strains εq and the volumetric strains εv can be calculated (Eqs. 3 and 4). q = σ1 − σ3
(1)
p = (σ1 + 2σ3 )/3
(2)
εq = 2(ε1 − ε3 )/3
(3)
εv = ε1 + 2ε3
(4)
The test presents two phases: • The conditioning phase: application of a high stress level (q = 340 kPa and p = 113 kPa) over 30,000 cycles at 2 Hz to stabilise the permanent strains. • The resilient test: application of several stress paths over 100 cycles at 2 Hz to study the elastic strains. Figure 2 presents the several stress paths.
Fig. 2 Stress paths
The Influence of the Mastic Coating of Untreated Reclaimed …
1827
Fig. 3 a Evolution of axial permanent strains with standard deviation at cycles 10,000 and 30,000. b Evolution of radial permanent strains with standard deviation at cycles 10,000 and 30,000
3 Experimental Results and Analysis 3.1 Conditioning Phase Figure 3a, b show the evolution of axial and radial permanent strains. The RAP before binder extraction and the UGM present similar permanent strains, particularly for the axial strains. However, the permanent strains of RAP-BE, after binder extraction, are very low. The main explication is the fines content, in addition of the tested density and water content. The RAP before and after extraction present respectively 3.0% and 12.6% of fines. According to Jing [13, 14], the quantity of fines has a significant benefit to rutting resistance, until a critical water content close to saturation. Indeed, before binder extraction the fines are trapped in the mastic coating, while after the extraction the fines are released and can be reduce the rutting.
3.2 Resilient Test For the sake of clarity, only maximum values of resilient strains are presented. Figure 4a–c display respectively the deviatoric and volumetric resilient strains as well as the resilient moduli. Equation (5) defines the resilient modulus Er . Er = σ1r /ε1r
(5)
where σ1 r is the axial resilient stress and ε1 r the axial resilient strain. Each material has a dilatant behaviour, except the RAP at higher cell pressures. Thus, the higher confining pressure, the higher is resilient modulus. We notice that the UGM presents high strains, while the RAP before and after binder extraction have similar levels of strain. Finally, if we consider RAP as an unbound granular
1828
L. Gaillard et al.
Fig. 4 Resilient deviatoric strains. b Resilient volumetric strains. c Resilient moduli
material, the resilient moduli of RAP are significant. In addition, these values are clearly higher than those of the Rhine Region UGM.
4 Conclusions This paper reports some results of the study of a source of Reclaimed Asphalt Pavement (RAP), with a high Polycyclic Aromatic Hydrocarbons (PAH) content. The difference between an unbound granular material and the RAP is the smooth contact between the rigid particles. This contact has consequences on the permanent and resilient behaviours. First, the permanent strains of RAP are high and outside specifications, according to the French Standard for Unbound Granular Materials (UGM) NF EN 13286-7 [12]. However, the RAP presents a higher level of resilient modulus than UGM. In comparison with the tested UGM, the mechanical performances of the RAP are higher both regard to the rutting and the resilient modulus. Consequently, these results confirm that the untreated RAP can be used instead of UGM, in base
The Influence of the Mastic Coating of Untreated Reclaimed …
1829
and subbase layers in low-traffic roads. This procedure is an alternative to deposit solution for contaminated RAP. Acknowledgement This work is part of the ORRAP (Optimal Recycling of Reclaimed Asphalts in low-traffic Pavements) project funded by ERDF—INTERREG V (3.1 ORRAP), which studies the cold recycling of reclaimed asphalts aggregates without binder addition.
References 1. Colbert, B., You, Z.: The determination of mechanical performance of laboratory produced hot mix asphalt mixtures using controlled RAP and virgin aggregate size fractions. Constr. Build. Mater. 26(1), 655–662 (2011) 2. Li, X., Marasteanu, M.O., Williams, R.C., Clyne, T.: Effect of reclaimed asphalt pavement (proportion and type) and binder grade on asphalt mixtures. Transp. Res. Record 2051(1), 90–97 (2008) 3. Valdés, G., Pérez-Jiménez, F., Miró, R., Martínez, A., Botella, R.: Experimental study of recycled asphalt mixtures with high percentages of reclaimed asphalt pavement (RAP). Constr. Build. Mater. 25(3), 1289–1297 (2010) 4. Dinis-Almeida, M., Castro-Gomes, J., de Lurdes Antunes, M.: Mix design considerations for warm mix recycled asphalt with bitumen emulsion. Constr. Build. Mater. 28(1), 687–693 (2011) 5. Porot, L., Eduard, P.: Laboratory evaluation of half-warm recycling with bio-based additive. In: Proceedings of 6th Eurasphalt & Eurobitume Congress. Prague (2016) 6. Guatimosim, F.V., Vasconcelos, K.L., Bernucci, L.L.B., Jenkins, K.J.: Laboratory and field evaluation of cold recycling mixture with foamed asphalt. Road Mater. Pavement Des. 19(2), 385–399 (2016) 7. Sangiorgi, C., Tataranni, P., Simone, A., Vignali, V., Lantieri, C., Dondi, G.: A laboratory and field evaluation of cold recycling mixture for base layer entirely made with reclaimed asphalt pavement. Constr. Build. Mater. 138, 232–239 (2017) 8. Jacobson, T.: Återvinning av krossad asphalt som bär- och förstärkningslager. Del 2—Erfarenheter från fältstudier.VTI notat 32-2002 (2003) 9. AFNOR.: Mélanges bitumineux—Spécifications pour le matériau—Partie 8: Agrégats d’enrobés. Norme française NF EN 13108-8 (2016) 10. AFNOR.: Graves non traitées—Spécifications. Norme française NF EN 13285 (2004) 11. AFNOR.: Mélanges bitumineux—Spécifications des matériaux—Partie 1: Enrobés bitumineux. Norme française NF EN 13108-1 (2007) 12. AFNOR.: Mélanges avec ou sans liant hydraulique—Partie 7: Essai triaxial sous charge cyclique pour mélanges sans liant hydraulique. Norme française NF EN 13286-7 (2004) 13. Jing, P., Nowamooz, H., Chazallon, C.: Permanent deformation behaviour of a granular material used in low-traffic pavements. Road Mater. Pavement Des. 19(2), 289–314 (2016) 14. Jing, P., Nowamooz, H., Chazallon, C.: Unsaturated mechanical behaviour of a granular material. Road Mater. Pavement Des. 20(6), 1429–1451 (2018)
The Influence of Wax Model Compounds on the Surface Topography of Bitumen Johan Blom, Niko Van den Brande, and Hilde Soenen
Abstract In literature, microscopy techniques including atomic force microscopy (AFM), scanning electron microscopy (SEM) and recently also confocal laser scanning microscopy (CLSM) have been used to study the microstructure of bitumen. Especially by AFM, various surface morphologies have been observed. Well-known are the so-called bee structures, related to a regular height profile of lower and higher regions. The appearance of these microstructures has been linked to the presence of crystallizing waxes. In addition to this, other regions surrounding, or in between the bee structures have been described. These have been referred to as the perpetua or peri structures, and very little is known about the origin or formation of these phases. In this study, straight, saturated hydrocarbons of specific chain lengths, were added to a wax-free naphthenic bitumen, and the air-bitumen interfaces were investigated by AFM. As expected, the unmodified bitumen did not display any topological microstructure, and by adding waxes to this binder, bee structures could be generated. However, not only bee structures were obvious, but the intermediate phases could also be generated under certain conditions. The chain length of the waxes, as well as their blending ratios were the determining parameters for the type of structures that were formed. So, in conclusion, this study demonstrates that not only the bee structures, but also the intermediate and/or surrounding structures are related to the presence of waxy hydrocarbons, and no other molecular structures are needed or involved in the formation of these structures. Keywords AFM · Bee structure · Crystallization · Wax
J. Blom (B) Faculty of Applied Engineering, EMIB Research Group, University of Antwerp, 171 Groenenborgerlaan, 2020 Antwerp, Belgium e-mail: [email protected] N. Van den Brande Physical Chemistry and Polymer Science (FYSC), Vrije Universiteit Brussel (VUB), Pleinlaan 2, 1050 Brussels, Belgium H. Soenen Nynas NV, 171 Groenenborgerlaan, 2020 Antwerp, Belgium © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_232
1831
1832
J. Blom et al.
1 Introduction Numerous papers have been devoted to studying the AFM topography of bituminous binders [1–5]. Initially, the so-called bee structures, when they were first observed, were associated to the presence of asphaltenes [1], afterwards, they were related to crystallizing wax molecules [2, 3]. However, not only bees, also other structures, surrounding or in between these bees have been observed and these are still rather unknown. They have, among other explanations, been associated to an interaction between asphaltenes and wax molecules. More recent studies have reported that all these structures, bee and surrounding phases, are only observed at the surface of air-bitumen interfaces, and not in the bulk of the material. They are also not formed when investigating the bitumen surface against another substrate [3]. So probably, these structures are less important with regard to the binder performance. The purpose of this paper is to demonstrate that all the topological structures observed in AFM and associated with bees are related to the presence of crystallizable compounds.
2 Experimental 2.1 Materials Two binders are included in this study; An unmodified, wax-free bitumen, grade 50/70, penetration of 64 dmm at 25 °C, and a binder containing natural wax, grade 40/60, penetration 42 dmm. The wax content of the binders was verified by differential scanning calorimetry (DSC). N-alkanes of different chain lengths, respectively C18, C24, C40 and C44 were acquired from TCI EUROPE NV. The physical characteristics of these n-alkanes are listed in Table 1. A commercial Fischer-Tropsch (FT) hard wax, with a melting range of 80–125 °C, was also acquired. Blends of bitumen and the alkanes were prepared as follows: the bitumen was heated to 160 °C, an amount of binder was taken, left to cool to room temperature, and the desired amount of the respective n-alkane or FT paraffin was added, afterwards this blend was reheated to 160 °C for 45 min, and homogenized by stirring the hot sample for at least 15 min. Test specimens for the AFM were prepared as follows; a drop of hot binder, approximately 20 mg, was applied to a glass plate, which was subsequently reheated to 160 °C to allow the drop to flow and form a flat surface. After preparation, specimens were stored inside a closed box at room temperature (20–22 °C) for a maximum of 24 h before testing.
The Influence of Wax Model Compounds on the Surface … Table 1 Physical characteristics of the n-alkanes used in this study
1833
Chain length Nr. C-atoms
Tm (melting temperature), °C
Tb (boiling temperature), °C
Purity (reported by TCI) (%)
C18
29
317
>98
C24
52
391
>99.0
C40
84
524
>97.0
C44
86
548
>97.0
2.2 AFM Setup The AFM measurements were performed with an Asylum Research MFP-3D AFM. All AFM scans were performed using tapping mode in air and at room temperature, 24–26 °C, and with an AC 160 TS silicon cantilever.
2.3 DSC Tests The binders were investigated by a Mettler Toledo DSC1. The bitumen binders were first heated in an oven at 160 °C, and a small amount (8–10 mg) was filled in a aluminium pan. In the DSC, the samples were homogenized at 140 °C, for 5 min, this was followed by a cooling scan from 140 to −60 °C, and by a heating scan in the same temperature interval, at a scanning rate of 10 °C/min.
3 Results and Discussion In Fig. 1, the AFM surface topography of the two binders included in this study is presented: (a) the unmodified wax-free bitumen, (b) the wax containing bitumen. As expected, the wax-free binder does not show any surface structure, while the binder containing natural wax shows small bee structures. This has been reported before in literature [3]. In addition, it is obvious that for this particular wax-containing binder each bee structure seems to be surrounded by another structure, which up to now has not been clearly defined. In this paper, it will be demonstrated that the bee structure, as well as the surrounding island observed in the second binder can be created by adding waxes to the wax-free binder. In Fig. 2, the AFM surface structure of the wax-free binder after adding different types of n-alkanes is presented. Each time 2% of wax is added. The addition of C24, Fig. 2a, does not result in the formation of bee structures. Instead two distinct phases are formed, which can be distinguished in the topography (amplitude) image, and which are even more pronounced in the phase image.
1834
J. Blom et al.
Fig. 1 AFM images of the two unmodified binders, a a wax-free binder, b a binder containing natural wax
On other hand, adding C40 to this binder, Fig. 2b, results in bee formations, and in this case the surfaces surrounding the bees display a rather rough area but without distinct extra phases. In Fig. 2c, d and e, combinations of hydrocarbons were added to the wax-free binder. In these cases, when investigating the amplitude and phase images, bees are clearly observed, and in some cases, surrounding phases can also be distinguished. Especially for Fig. 2c, the formation of small islands surrounding each bee structure is obvious. The structure of this sample resembles the structure of the binder containing natural wax, shown in Fig. 1b in this paper. The AFM images show that the combination and the chain length of the wax is important. In Fig. 2a, b, the difference in the AFM topography between adding 2% C24 or 2% C40 is clear, C24 does not result in a bee formation while C40 does. This can be related to the melting behavior of these two waxes in the wax-free binder, as is shown in the DSC measurements presented in Fig. 3. The binder with 2% C24 has a lower melting range as compared to the binder with 2% C40. Compared to the temperature in the AFM, indicated in Fig. 3, the wax in the binder containing C24 is almost fully melted, while the wax in the binder containing C40 is still at the onset of its melting range. At the AFM test temperature, which was 25–26 °C, the C40 wax is crystalline and forms a hard phase. This phase is formed during cooling, on a soft bitumen phase. It has been described that a stiff thin plate on a softer substrate upon further cooling leads to a wrinkling phenomenon. This is further explained in [5]. Therefore, in this work, the authors conclude that the crystalline phase forms the bee structures, while the wax which is at the AFM test temperature above its melting temperature stays soft and forms the island structure. The island structures were most pronounced when the two extreme waxes, with the largest difference in chain length, C18 and the FT wax, were combined. For completeness; the DSC heating
The Influence of Wax Model Compounds on the Surface …
1835
Fig. 2 AFM images of the wax-free binder after adding 2% in total of a C24, b C24 and C40, c C24 and C40, d C24 and C40 different ratio (1.5/0.5% & 1.75/0.25%), e C18 and FT wax
scans also indicate a recrystallization, directly after the glass transition temperature. This is most pronounced for the binder containing C24.
1836
J. Blom et al.
0.35
0.25 0.2
AFM- temperature
Bit + 2% C24 exotherm
Heat Flow, W/g
Bit + 2% C40 0.3
melƟng C40 melƟng C24 glass transiƟon
0.15 0.1 -55 -45 -35 -25 -15
-5
5 15 25 Temperature, °C
35
45
55
65
75
Fig. 3 DSC heating scans of the wax-free binder after adding C24, and after adding C40
4 Conclusions In this study, straight, saturated hydrocarbons of specific chain lengths, were added to a wax-free naphthenic bitumen. The bitumen-air interfaces were investigated by AFM. As expected, the unmodified bitumen did not display any topological microstructure. By adding the n-alkanes to this binder, bee structures could be generated. This has already been shown in literature. However, in this study, not only the bee structures could be created, also the intermediate structures could be induced by adding waxes or n-alkanes. To achieve this, blends of n-alkanes with high and a low melting points or short and longer chain lengths needed to be added. The chain length of the waxes, their concentration and blending ratios determine the type and ratio of phases that are formed. So, in conclusion, this study demonstrates that not only the bee structures, but also the intermediate and/or surrounding phases are related to the presence of waxy hydrocarbons, and no other molecular structures are needed or involved in the formation of these structures. This finding is in agreement to the proposed mechanisms for the bee structures, based on a wrinkling phenomenon [5], and it is also in agreement with a recent study conducted by TOF-SIMS [6] (time of flight secondary ion mass spectrometry) in which the only domains observed on the surface of a wax containing binder were related to more or less paraffinic areas, not involving asphaltene type of structures.
References 1. Loeber, L., Sutton, O., Morel, J., Valleton, J.-M., Muller, G.: New direct observations of asphalts and asphalt binders by scanning electron microscopy and atomic force microscopy. J. Microsc. 182(1), 32–39 (1996) 2. Soenen, H., Besamusca, J., Fischer, H.R., Poulikakos, L.D., Planche, J.-P., Das, P.K., Chailleux, E.: Laboratory investigation of bitumen based on round robin DSC and AFM tests, Mater. Struct.
The Influence of Wax Model Compounds on the Surface …
1837
47(7), 1205–1220 (2014). https://doi.org/https://doi.org/10.1617/s11527-013-0123-4 3. Blom, J., Soenen, H., Katsiki, A., Van den Brande, N., Rahier, H., van den Bergh, W.: Investigation of the bulk and surface microstructure of bitumen by atomic force microscopy. Constr. Build. Mater. 177, 158–169 (2018) 4. Ramm, A., Downer, M.C., Sakib, N., Bhasin, A.: Morphology and kinetics of asphalt binder microstructure at gas, liquid, and solid interfaces. J. Microsc. (2019). https://doi.org/10.1111/ jmi.12842 5. Lyne, Å.L., Wallqvist, V., Rutland, M.W., Claesson, P., Birgisson, B.: Surface wrinkling: the phenomenon causing bees in bitumen. J. Mater. Sci. 48(20), 6970–6976 (2013). https://doi.org/ https://doi.org/10.1007/s10853-013-7505-4 6. Lu, X., Sjövall, P., Soenen, H.: Structural and chemical analysis of bitumen using time-of-flight secondary ion mass spectrometry (TOF-SIMS), Fuel, 199 206–218. ISSN: 0016–2361 (2017)
The Low Temperature Performance of Rubberized Bituminous Mixture Using the Dry Process Dongdong Ge , Xuelian Li, and Zhanping You
Abstract The paper evaluated the low temperature performance of the bituminous mixtures, which were produced in the plant using the dry process and compacted in the laboratory. ElastikoR crumb rubber was chosen to modify the bituminous mixture. Four types of bituminous mixtures, including leveling layer and surface layer gradation with and without crumb rubber, were produced in a plant in Iron Mountain, MI. The volumetric design of the mixture followed the Superpave mixture design. The leveling layer and surface layer mixture used 25 and 17% recycled asphalt pavement (RAP) (by weight of bituminous mixture). To improve the moisture resistance, dosages of 0.125 and 0.08% anti-stripping agent (by weight of bitumen) were selected for the leveling layer and surface layer. The low temperature cracking properties of four types of bituminous mixtures were evaluated using the Disk-shaped Compact Tension (DCT) test. The RAP content, bitumen content, and nominal maximum aggregate size (NMAS) influenced the low temperature cracking performance of the bituminous mixture. The fracture energy of the surface layer bituminous mixture was higher than that of the leveling layer bituminous mixture. The cracking resistance of the leveling layer and surface layer bituminous mixture were improved after the addition of rubber. The maximum CMOD was linearly correlated with the fracture energy of different types of bituminous mixtures. Keywords Rubber · DCT · Dry process · Low temperature cracking · Bituminous materials
1 Introduction Crumb rubber normally included rubber particles ranged from 4.75 mm to less than 0.075 mm. Three methods are currently available to crush scrap tires to crumb rubber. D. Ge · Z. You (B) Michigan Technological University, Houghton, MI 49931, USA e-mail: [email protected] X. Li Changsha University of Science & Technology, Changsha 410004, Hunan, China © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_233
1839
1840
D. Ge et al.
The most commonly used cracker mill process could crush the tire rubber to ground crumb rubber, ranging in size from 5 to 0.5 mm. The granulator process tears apart the rubber to granulated crumb rubber particles, ranging in size from 9.5 to 0.5 mm. The micro-mill process was used to produce very fine crumb rubber in size ranging from 0.5 to 0.075 mm. Sometimes, the cryogenic technique used liquid nitrogen to make the rubber particles extreme brittle to smash into smaller particles [1]. Crumb rubber can be used in the bituminous pavement with the wet process or the dry process. In the wet process, rubber particles were mixed with bitumen first before the production of the bituminous mixture. While, in the dry process, crumb rubber was treated as part of the fine aggregate in the final bituminous mixture [2]. The dry process was firstly adopted in the US. With the PlusRive technology. The PlusRide dry process technology was adopted in different states DOTs, which proved that the strategy was not a cost-effective method [3]. Some test sections performed well, while some did not have comparable performance than conventional bituminous mixture [4]. The dry process can use batch and drum-dryer plants to incorporate rubber into the bituminous mixture, but the mixing temperature and duration need to be adjusted. Unlike conventional bituminous mixture construction, rubber bituminous mixture compaction with dry process needs a finishing roller to compact the mixture until the temperature decreased to 60 °C. The reaction between the rubber and bitumen at elevated temperature will swell the mixture. The quality of rubber modified bituminous mixture with the dry process need to be controlled by monitoring the rubber gradation, rubber percentage, rubber pretreatment, and mixing time [5]. There are still many unsolved issues regarding the application of rubber with the dry process. More field data related to the dry process are needed to evaluate the performance. The influence of the dry process on the environment and the cost-effectiveness needed to be validated. The application of the dry process under different climate conditions with different mixture designs and different raw materials can provide more reliable data for the guidance of further construction.
2 Materials and Test Methods 2.1 Material A Performance Grade (PG 58-34) bitumen was used as base bitumen in this study. ElastikoR crumb rubber (10% by weight of bitumen based on vendor’s recommendation) was chosen to modify bituminous mixtures with the dry process. Four types of bituminous mixtures, including leveling layer and surface layer gradation with and without crumb rubber, were produced in a plant in Iron Mountain, MI. The leveling layer bituminous mixture was used for the leveling course, and the surface layer bituminous mixture was used for the surface course. All the aggregates are from the local quarry station. The mixtures were produced in the plant with the dry process and compacted in the laboratory.
The Low Temperature Performance of Rubberized Bituminous …
1841
Table 1 The TSR test results of the leveling layer and surface layer bituminous mixture Sample
Average tensile strength of dry subset
Average tensile strength of conditioned subset
TSR
Leveling layer
175.08
151.44
0.86
Surface layer
165.7
149.79
0.90
The volumetric design of the mixture followed the Superpave mixture design. The leveling layer and surface layer bituminous mixture used 25 and 17% recycled asphalt pavement (RAP) (by weight of bitumen). The bitumen content in RAP is 4.15%. The new bitumen added for the leveling layer and surface layer bituminous mixture are 3.8 and 5.4%. If considered the bitumen in RAP, the final bitumen content of the leveling layer and surface layer bituminous mixture are 4.9 and 6.1%. In order to improve the moisture resistance, dosages of 0.125 and 0.08% ZT-SP (anti-stripping agent) (by weight of bitumen) were selected for the leveling layer and surface layer. The TSR results of the leveling layer and surface layer bituminous mixture design are shown in Table 1. The TSR values of the leveling layer and surface layer bituminous mixture met the standard restriction (TSR > 0.8).
2.2 Test Methods The DCT test was used to evaluate the low temperature fracture energy of the bituminous mixture. The tension load was applied on a circular specimen with a single edge notch, according to ASTM D7313. The test temperature was set as -24 °C, which followed the standard suggestion that 10 °C greater than the low temperature performance grade of the bitumen (−34 °C). In order to evaluate the influence of test temperature on the low temperature cracking test, the −18 °C was also adopted. The load during the test was controlled by the crack mouth opening displacement (CMOD) and varied by maintaining a constant CMOD rate of 0.017 mm/s. The load and CMOD were acquired during the test. The area under the load-CMOD curve was computed to calculate the fracture energy.
3 Test Result Analysis The DCT test was adopted to evaluate the low temperature cracking performance of different bituminous mixtures. Three replicates were tested for four types of bituminous mixture. The fracture energy, peak load, and maximum CMOD of different types of bituminous mixtures under different test temperatures are displayed in Fig. 1. The air voids of four types of bituminous mixtures are about 7.0%; the influence of
1842
D. Ge et al.
air void on the low temperature performance of bituminous mixtures was not considered. The fracture energy of the surface layer bituminous mixture was higher than that of the leveling layer bituminous mixture (Fig. 1a). The surface layer bituminous mixture had a smaller nominal maximum aggregate size (NMAS), lower RAP content, and higher bitumen content, which contributed to the improved low temperature cracking resistance. The cracking resistance of the leveling layer and surface layer bituminous mixture were improved after the addition of rubber. The fracture energy of the surface layer mixture with rubber was 1.52 times higher than the surface layer mixture without rubber. The fracture energy of the leveling layer mixture with rubber was 1.37 times higher than the leveling layer mixture without rubber. The fracture energy was increased with the increase of the test temperature. The bituminous mixture was less brittle with the increase of test temperature. The peak load of different bituminous mixtures was similar, except the surface layer bituminous mixture with rubber at −24 °C had a higher peak load (Fig. 1b). The bituminous mixture gradation, RAP content, and bitumen content did not have much influence on the peak load of the bituminous. The peak load was slightly increased with the decrease of the test temperature, and the bituminous mixture became more brittle as temperature decreased, thus increased the peak load. The CMOD was kept in a constant value to maintain the stability of the test during the post-peak region. The test was stopped after the load returned to 0.1 kN. The CMOD after the peak load reflected the propagation of cracking inside the sample during the DCT test. Bituminous mixture with higher maximum CMOD had higher ductility. The maximum CMOD of the surface layer and leveling layer bituminous mixture increased after the addition of rubber (Fig. 1c). The addition of rubber increased the elastic property of the bituminous mixtures. The maximum CMOD of the surface layer bituminous mixture was higher than that of the leveling layer bituminous mixture. The lower RAP content, higher bitumen content, and smaller NMAS were the main contribution of higher maximum CMOD of surface layer mixture. The maximum CMOD was increased with the increase of the test temperature. The elastic property of the bituminous mixture was improved with the increase of test temperature, thus increased the maximum CMOD during the DCT test. During the DCT test, the fracture energy and maximum CMOD followed a similar trend under different test conditions. The correlation between the max CMOD and fracture energy is shown in Fig. 2. The maximum CMOD was linearly correlated with the fracture energy of different types of bituminous mixtures (R2 = 0.851, p < 0.001). By increasing the elastic property of the bituminous mixture, both the maximum CMOD and fracture energy could be improved; the low temperature cracking resistance of the bituminous mixture could be effectively improved.
The Low Temperature Performance of Rubberized Bituminous …
(a) Fracture energy
1843
(b) Max load
(c) Max CMOD
Fig. 1 The low temperature cracking test parameters of different mixtures Fig. 2 The correlation between max CMOD and fracture energy
1844
D. Ge et al.
4 Conclusions and Recommendations In this paper, the low temperature cracking property of four types of plant mixed and laboratory compacted bituminous mixtures were evaluated with the DCT test. According to the test results, the main findings are: 1.
2.
3.
4.
The RAP content, bitumen content, and NMAS influenced the low temperature cracking performance of bituminous mixture. Bituminous mixture with lower RAP content, higher bitumen content, and smaller NMAS had better low temperature cracking performance. The addition of rubber increased the elastic property of the bituminous mixture at low temperatures, thus increased the low temperature performance of the bituminous mixture. The fracture energy and maximum CMOD were increased with the increase of the test temperature. The peak load was slightly increased with the decrease of the test temperature. The maximum CMOD was linearly correlated with the fracture energy of different types of bituminous mixtures.
Acknowledgments This study is sponsored by the Michigan Department of Environment, Great Lakes, and Energy (EGLE), the former Michigan Department of Environmental Quality (MDEQ), through the Dickinson County Road Commission of the United States. Trademark or manufacturers’ names appear in this paper only because they are considered essential to the objectives of this paper.
References 1. Chavez, F., Marcobal, J., Gallego, J.: Laboratory evaluation of the mechanical properties of asphalt mixtures with rubber incorporated by the wet, dry, and semi-wet process. Constr. Build. Mater. 205, 164–174 (2019) 2. Ge, D., You, Z., Chen, S., Liu, C., Gao, J., Lv, S.: The performance of asphalt binder with trichloroethylene: improving the efficiency of using reclaimed asphalt pavement. J. Cleaner Prod. 232, 205–212 (2019) 3. Chen, S., Gong, F., Ge, D., You, Z., Sousa, J.: Use of reacted and activated rubber in ultra-thin hot mixture asphalt overlay for wet-freeze climates. J. Cleaner Prod. 205, 369–378 (2019) 4. Zhou, H., Holikatti, S., Vacura, P.: Caltrans use of scrap tires in asphalt rubber products: a comprehensive review. J. Traffic Transp. Eng. 1(1), 39–48 (2014) 5. Oliveira, J., Silva, H., Abreu, L., Fernandes, S.: Use of a warm mix asphalt additive to reduce the production temperatures and to improve the performance of asphalt rubber mixtures. J. Cleaner Prod. 41, 15–22 (2013)
The Use of Direct Shear Test for Optimization of Interlayer Bonding Under a Poroelastic Layer Piotr Jaskula , Marcin Stienss , Cezary Szydlowski , and Dawid Rys
Abstract Poroelastic Road Surfaces (PERS) are characterised by porous structure with air void content of 20% or higher and stiffness almost 10 times lower than that of a standard asphalt course. Such properties enable noise reduction by up to 12 dB in comparison to SMA 11 mixture. However, the disadvantage of a poroelastic pavement is its low durability, which partially results from delamination from the lower layer. The paper aims to investigate the effect of type and amount of tack coat used as well as the texture of the lower layer on interlayer bonding, to better recognise the issue of bonding under the poroelastic layer. For this purpose, direct shear test with monotonic and cyclic load was performed for 44 types of layer interfaces. The effect of type and amount of bituminous emulsion used for tack coat as well as the impact of lower layer surface were distinguished. The results of the analysis indicated that tack coats made from softer residual bitumen (70/100) exhibited better fatigue resistance, while polymer modification had minor effect. Moreover, milling of the lower layer resulted in a significant increase in shear strength. Finally, several types of layer interfaces were selected to be constructed in the full-scale test sections. Keywords Poroelastic pavements · Crumb rubber · Highly modified asphalt · Direct shear testing · Interlayer bonding
P. Jaskula (B) · M. Stienss · C. Szydlowski · D. Rys Gdansk University of Technology, Gdansk, Poland e-mail: [email protected] M. Stienss e-mail: [email protected] C. Szydlowski e-mail: [email protected] D. Rys e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_234
1845
1846
P. Jaskula et al.
1 Introduction Poroelastic Road Surfaces (PERS) are asphalt mixtures with air voids content at least 20% and partial replacement of mineral aggregate by crumb rubber aggregate. The difference to Porous Asphalt consist on higher elastic deformation under wheel load. Porosity and high elasticity enable high tyre/road noise reduction. Recent test results [1, 2] indicated a reduction by up to 12 dB compared to the standard SMA 11 wearing course. However, the disadvantage of poroelastic pavements is their low durability. Ravelling and delamination from the lower layer were identified as the primary causes of insufficient durability of PERS [3, 4]. The work presented in the paper is a part of a wider research project SEPOR (Safe Eco-Friendly Poroelastic Road Surface), which aims to develop a durable poroelastic road surface and addresses the optimisation of interlayer bonding as one of the issues. At the initial stage of the program, there are several problems which need to be solved: (1) Influence of the consistency of bitumen, polymer modification, and the applied amount of tack coat both on shear strength and fatigue resistance of interlayer bond should be recognized. (2) Theoretically, the surface of the lower layer has a significant impact on interlayer bonding quality, but this hypothesis has to be verified. The effect of interlayer bonding strength was evaluated in the study [5], where inclined shear test delivered interface bonding strength between polyurethane PERS mixtures and lower layers made from AC or SMA. The results varied from 0.5 MPa to 0.7 MPa and 0.7 MPa to 1.3 MPa, for the AC and SMA lower layers respectively. The study [6] indicated that a decrease in stiffness of the lower layer (binder course) adversely influences the response of the PERS layer. The paper aims to investigate the effect of type and amount of tack coat used as well as the surface texture of the lower layer on interlayer bonding. The studies described here were performed in laboratory scale. Finally the analysis enabled selection of layer interface systems to be constructed on full-scale test sections.
2 Materials and Methods The specimens prepared for optimisation of interlayer bonding consisted of the upper poroelastic layer, the interface layer of bitumen emulsion and the lower layer made of asphalt concrete (AC) or stone mastic asphalt mixture (SMA). While the upper poroelastic layer was the same in each case, interface tack coats and lower layers varied. The mixture used for the poroelastic layer was labelled as PSMA 5 (poroelastic SMA) and consisted of mineral aggregate, crumb rubber aggregate, limestone filler and highly modified asphalt binder 45/80–80 with at least 7% content of SBS polymer. The proportions of components are given in Table 1. More detailed properties and optimisation of the poroelastic
The Use of Direct Shear Test for Optimization of Interlayer …
1847
Table 1 Composition of the poroelastic mixture PSMA 5, prepared in the laboratory Aggregate, (% of aggregate mass) Mineral aggregate
Limestone filler
5/8
2/5
0/2
0
72
6
7
Asphalt, Bitumen (% of mixture mass)
Crumb rubber
1/4
0.5/2
45/80–80
10
5
11
Air voids % (v/v)
14.5
mixture PSMA 5 are described in previous study [7]. It should be emphasised that the poroelastic mixture PSMA 5 exhibits much lower stiffness (around 200 MPa) in comparison to 2400 MPa obtained for the reference SMA 11 and 4700 MPa in case of AC 16 mixtures (IT-CY test at 25 °C, according to EN 12697-26). Analogously, internal shear strength (at 20 °C, according to EN 12697-48) results were 0.77 MPa for PSMA 2.35 MPa for the reference SMA 11 and 2.88 MPa for the AC 16 W. Two typical asphalt mixtures were used for the lower layer: 1) AC 16 with neat bitumen 35/50 and 2) SMA 11 with polymer-modified bitumen 45/80–55. Two different mixtures were considered in order to vary the surface texture and stiffness of the lower layer: denser and stiffer for AC16 and more open and less stiff for SMA 11. Moreover, the effect of grooved texture obtained by milling of the lower layer was analysed. The interface layer was applied as a tack coat over the lower layer. Cationic bituminous emulsions with the following bitumens were applied: three SBR-modified bitumens 35/50, 50/70 and 70/100 as well as one neat bitumen 70/100. The amount of residual bitumen equalled 0.1/0.15, 0.2 or 0.3 kg/m2 . Mixtures were prepared and compacted in laboratory conditions using slab roller compactor. The grooved surface was obtained with a full-scale milling machine. All lower layers were first covered with bitumen emulsion and then, after the time required for full emulsion breakdown, the poroelastic mixture was laid and compacted. The specimens for further laboratory tests were cored from the doublelayer slabs and the diameter equalled 100 or 150 mm. Figure 1 presents examples of specimens during the preparation process.
Fig. 1 Specimen preparation: a AC 16 lower layer covered with bitumen emulsion, b grooved surface of SMA 11 after milling covered with bitumen emulsion, c cylindrical specimens with lower and upper layers cored from a slab
1848
80 60 40
Kini
20 0
Full failure
100
N25%
120 N=50
b Shear stiffness K [MPa]
a
P. Jaskula et al.
25% K ini
0 50
5000 Number of cycles
10000
Fig. 2 Example results of direct shear test in the mode of: a monotonic loads—shear stress versus total deformation, b cyclic loads—shear stiffness versus number of cycles
The goal was to identify how the texture of the lower layer as well as the type and amount of bitumen at the interface impact shear strength and fatigue shear resistance of interlayer bonding. For this purpose direct shear tests were performed in two modes of load: (1) monotonic load till full shear and (2) cyclic load till full shear of the interface or till 100,000 cycles. Both tests were performed at the temperature of 20 °C. In the case of monotonic load the standard Leutner [8] test was applied, where the constant rate of deformation equals 50 mm/min. An example of test results obtained from direct shear test with monotonic load is presented in Fig. 2a. The designation in Fig. 2a are as follows: τ—shear stress, δ—deformation, k—shear stiffness, W—total dissipated energy. For the purpose of this paper, shear strength τmax was considered as a measure of interlayer bonding quality. The cyclic load tests were performed in constant stress mode and the load conditions were the same for all specimens. Minimum shear stress was set as 50 kPa, maximum equalled 150 kPa. The cycle of load consisted of 0.05 s of loading time, 0.05 s of unloading time and 0.1 s of rest time. The shear modulus K was determined according to the methodology described in [9] in each test cycle. The fatigue resistance was defined by a number of cycles N25% till the shear modulus decreased to 25% of the initial value Kini , determined in the 50th cycle of the test. The higher decrease to 25% of Kini than typically assumed 50% was selected to be closer to full slip of interlayer join. An example of decrease in shear stiffness is presented in Fig. 2b.
3 Results and Discussion Results obtained from direct shear tests are summarized in Table 2 as average values representing each considered type of layer interface. The base group contains 2 specimens with the same lower and upper layer, type and amount of bitumen used for tack coat.
The Use of Direct Shear Test for Optimization of Interlayer …
1849
Table 2 Average shear strength and average number of load repetitions till 25% of Kini Type of lower layer
AC 16
SMA 11
Type of bitumen used for tack coat
Monotonic load strength (MPa)
Cyclic loads N25% (–)
For content of bitumen (kg/m2 )
For content of bitumen (kg/m2 )
0.1/0.15a
0.2
0.3
0.1/0.15a
0.2
0.3
35/50 + SBR
0.16
0.26
0.22
1804
2386
1929
50/70 + SBR
–
0.30
–
4343
2332
3130
70/100
0.28
0.28
0.26
4155
5126
6298
35/50 + SBR
0.20
0.20
0.17
1155
2041
1358
50/70 + SBR
0.42
0.45
0.36
2773
–
3087
70/100
0.38
0.48
0.43
11,383
3988
3332
70/100 + SBR 0.30
0.32
0.26
4154
6185
1145
SMA 11 milled crosswise
70/100 + SBR 0.58
–
0.52
22,828
–
5000
SMA 11 milled lengthwise
70/100 + SBR 0.72
–
0.98
9070
–
11,894
a In
the case of milled lower layer the amount of 0.15 kg/m2 was applied
The average coefficient of variance COV for the values given in Table 2 was also calculated for results among replicates and they equal 20% for shear strength and 35% for number of cycle loads. Interlayer bond shear tests performed for specimens with poroelastic mixture yielded higher dispersion of results than analogous tests performed for standard asphalt mixtures. Moreover, it cannot be stated that a change in bitumen content has any effect on shear strength or fatigue resistance. Probably the differences in tack coat amount were too small for the case of poroelastic course. Thereby, all the tested types of layer interfaces were divided into 9 groups. A given group consists of specimens with the same type of bituminous emulsion used for tack coat and the same type of lower layer. Figure 3 shows average values and standard deviations of τmax and N25% for the considered 9 groups of layer interfaces. On the basis of Fig. 3 it can be stated that application of softer bitumen for tack coat caused an increase both in shear strength and in fatigue resistance. The modified emulsion did not improve the quality of interlayer bond. The type of lower layer (AC16 or SMA11) has a minor effect on interlayer bonding. However, groups with milled lower layer exhibited much higher shear bond strength and number of cycles till failure than the other groups of bonds. Generally, with an increase in shear strength, the number of cycles till failure increases as well. This observation can be partially associated with the changing proportion between the maximum cyclic load and the shear strength. While the shear stress in cyclic load was the same for each specimen, the shear strength differed across the tested samples. Thereby, a more adequate parameter to describe fatigue
1850
P. Jaskula et al.
Fatigue resistance N25% [-]
30000
Lower layer AC16
25000
Bitumen 35/50 + SBS 50/70 + SBS
20000
70/100 SMA11
15000
35/50 + SBS 50/70 + SBS
10000
70/100 5000 0 0.00
70/100 + SBS
0.20
0.40
0.60
0.80
1.00
Shear strength τmax [MPa]
1.20
SMA11 milled: crosswise
70/100 + SBS
lengthwise
70/100 + SBS
Fig. 3 Average values and standard deviations of shear strength τmax and cycles till failure N25% for the considered 9 groups of layer interfaces
resistance than the N25% value should be introduced, taking into account the ability to bear cyclic loading compared to the overall strength of the specimen. The results of the N25% parameter for specimens in one group have low repeatability, especially in the case of grooved surface of lower layer. It means that the test procedure or the analysis of results should be improved in further studies. The emulsion containing bitumen 70/100 + SBR with the amount of residual binder of 0.3 kg/m2 was eventually selected to be applied at the full-scale sections. Two types of lower layer will be also used: SMA11 with smooth surface and SMA11 with grooved surface. In further research, the observations of distress progress from the full-scale sections will be compared with the results from direct shear tests with monotonic and cycle load.
4 Conclusions The performed tests supplemented previous works on interlayer bond quality of poroelastic surfaces in the field of shear strength and fatigue resistance. Based on the performed tests, it may be concluded that in the case of double-layer samples including poroelastic course the best results were obtained for the softest polymer modified bitumen 70/100 used as a tack coat. A minor improvement was observed when the lower layer was less stiff (SMA asphalt course used instead of standard AC binder course). However, significant improvement in interlayer bonding was observed when the lower layer was grooved by milling. With the obtained results, it was impossible to observe any influence of different amounts of the applied tack coat on the interlayer bonding of poroelastic and standard asphalt layers. The obtained maximum shear strength under monotonic loading during the direct shearing of
The Use of Direct Shear Test for Optimization of Interlayer …
1851
double-layer sample was similar or slightly lower than the maximum shear strength obtained for internal direct shear of the poroelastic layer. Acknowledgments Research work reported in this article was sponsored by the Polish National Centre for Research and Development (NCBiR) under the project SEPOR (Grant TECHMATSTRATEG 1/347040/17/NCBR/2018).
References ´ ˙ 1. Ejsmont, J., SwieczkoZurek, B., Jaskula, P.: Low Noise Poroelastic Road Pavements Based on Bituminous, Nois-Con 2019, 1–8 (2019) ´ ˙ 2. Ejsmont, J., SwieczkoZurek, B., Owczarzak, W., Sommer, S., Ronowski, G.: Tire/road noise on poroelastic road surfaces. EuroNoise 2018, 2719–2724 (2018) 3. Cesbron, J., Skov, R.S.H., Bendtsen, H.: Lab testing of performances. Final PERSUADE seminar (2015) 4. Bendtsen, H., Stahlfest, R., Skov, H., Andersen, B., Olesen, E., Skov, R.S.H., Andersen, B., et al.: Performance of PERS at the Kalvehave test site, Rapport nr. 514 (2014) 5. Liao, G., Wang, H., Asce, A.M., Zhu, H., Sun, P., Chen, H.: Shear strength between poroelastic road surface and sublayer with different bonding agents. J. Mater. Civ. Eng. 30(3), 1–9 (2018) 6. Srirangam, S.K., Anupam, K., Casey, D., Liu, X., et al.: Evaluation of structural performance of poro-elastic road surfacing pavement subjected to rolling truck tire loads. J. Transp. Res. Board, 1–25 (2015) 7. Jaskula, P., Szydlowski, C., Stienss, M., Rys, D., Jaczewski, M.: Durable poroelastic wearing course SEPOR with highly modified bitumen. In : AIIT International Congress on Transport Infrastructure and Systems in a changing World, TIS Roma 2019 (2019) 8. Leutner, R.: Untersuchung des Schichtenverbundes Oberbau. Bitumen 3, 84–91 (1979) 9. Jaskula, P.: Sczepno´sc´ warstw asfaltowych w wielowarstwowych układach nawierzchni drogowych. Wydawnictwo Politechniki Gda´nskiej, (in Polish), Gdansk (2018)
Time and Storage Dependent Effects of Bitumen—Comparison of Surface and Bulk Johannes Mirwald, Stefan Werkovits, Ingrid Camargo, Daniel Maschauer, Bernhard Hofko, and Hinrich Grothe
Abstract Bitumen is a product obtained from crude oil refinement. Its complex nature demands for a sophisticated approach to thoroughly assess it and link chemical composition, mechanical behavior and microstructure. Besides conventional mechanical testing, performance based test methods like dynamic shear rheometer (DSR) or bending beam rheometer (BBR) are gaining importance since they describe the actual viscoelastic behavior versus time and temperature. Furthermore, FourierTransformation-Infrared (FTIR) spectroscopy has recently become one of the most used analytical methods in bitumen research. The method give information regarding chemical composition in terms of functional groups and can distinguish between different ageing states and different kinds of modification. The presented study shows time and storage dependent effects on the surface and in the bulk of bitumen. It focuses on three main questions: (a) Is the surface and bulk affected by exposition to visible light within a timespan of up to 20 days? (b) Do the changes on the surface correlate with changes in bulk properties? (c) Does the surface of bitumen passivate, preventing further atmospheric oxidation? To answer these questions, two different unmodified binders were prepared and measured with FTIR and DSR. Half of the samples were exposed to visible light, while the others were stored in the dark at a controlled laboratory atmosphere. The results of FTIR spectroscopy show that the surface, when stored in visible light, passivates significantly with increasing storage time. When stored in the dark, no relevant changes are measured. The results from bulk (DSR) show the same trends but to a much lesser extent, compared to the surface. An overall sample storage in the dark is recommended. FTIR measurements should be conducted within an hour after preparation, while DSR measurements can be done up to one day after preparation. Keywords Bitumen · Time dependency · Storage dependency · FTIR spectroscopy · DSR J. Mirwald · S. Werkovits · H. Grothe (B) Institute of Material Chemistry, TU Wien, Getreidemarkt 9/BC/01, 1060 Wien, Austria e-mail: [email protected] I. Camargo · D. Maschauer · B. Hofko Institute of Transportation, TU Wien, Gußhausstrasse 28/E230-3, 1040 Wien, Austria © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_235
1853
1854
J. Mirwald et al.
1 Introduction Characterization of bitumen has evolved over the decades of research and adapted to its industrial needs for application. While mechanical testing has been the centerpiece of bitumen classification, chemical analysis has started to show its insightful impact in the field. Especially Fourier-Transformation-Infrared (FTIR) Spectroscopy has become a popular method due to its time-efficient measurements and easy implementation. The results are valuable in regards to chemical information linked to the functional groups present in bitumen, which allows distinguishing between different states of ageing and modification. While a DSR measurement is performed according to the EN 14770 or the ASSHTO T 315, FTIR spectroscopy has not been standardized yet and is thus still varying across different laboratories, countries and continents. Previous work showed a first step towards a distinct measurement routine for analyzing bitumen with FTIR Spectroscopy [1, 2]. It suggests that after normalization at the highest band (2090 cm−1 ) integration is done according to the following parameters: • Carbonyls (AICO ): 1660–1800 cm−1 • Sulfoxides (AISO ): 1079–984 cm−1 • Reference aliphatic band (AICH3 ): 1525–1350 cm−1 . Furthermore, standard statistical evaluated is carried out to determine the ageing index AIFTIR as shown in Eq. (1). AIFTIR = (AICO + AISO )/(AICH3 )
(1)
Whether the division step is necessary or not, needs further investigation and remains an open question. Even after following the same data treatment routine, various parameters like time and temperature for sample preparation and storage affect the outcome of a spectroscopic measurement. Therefore, uniform sample preparation routine and defined parameters are required to obtain a consistent quality for bitumen analysis with FTIR Spectroscopy. This study looks at the time dependent effects of bitumen on the surface (FTIR Spectroscopy) and compares it to bulk properties changes (DSR). Therefore, two different unmodified 70/100 penetration graded binders were prepared, stored at ambient laboratory conditions and measured up to 20 days after preparation. These results can reveal critical information in regards to storage stability, possible passivation and recommendations for future measurement routines. All these factors can have a significant impact on the outcome of each measurement and need to be questioned.
Time and Storage Dependent Effects of …
1855
Fig. 1 Bitumen sample stored under glass lid (left) and metal lid (right)
2 Materials and Methods 2.1 Material This study was performed using two 70/100 penetration graded bitumen. The binders were heated up to 100–150 °C for a maximum of 2–5 min and homogenized prior to its transfer into the silicone mold using a spatula or thermometer. All samples were stored under ambient laboratory conditions between 21 and 24 °C (tracked with 2 sensors). Half of the samples were covered with a glass lid, the other half with a metal lid. By doing so, half of the samples were exposed to visible light while the others were stored in the dark, as shown in Fig. 1. The light did not contain the UV part (λ ≤ 350 nm) of the spectrum, since it is blocked by conventional window glass.
2.2 Methods 2.2.1
Fourier Transformation Infrared Spectroscopy (FTIR)
The samples inside the silicone mold were applied onto the diamond ATR crystal of the spectrometer (Perkin-Elmer Spectrum Two) and measured subsequently. Each spectrum was recorded in the range of 600–4000 cm−1 at a resolution of 2 cm−1 (4 scans). The test schedule was as follows: 0, 1, 3, 7, 11 and 18 days after preparation. For each type of binder and day, two samples were measured with four repeats. A background of the empty ATR crystal was recorded prior to each measurement. The received aging indices (AIFTIR ) were statistically evaluated as mentioned earlier and can be found in the results in Fig. 3.
1856
2.2.2
J. Mirwald et al.
Dynamic Shear Rheometer (DSR)
For rheological testing, the samples were mounted and measured according to the EN 14770. All measurements were conducted at certain days after preparation (1, 3, 6, 10, 15, 20) using a MCR 302 Anton Paar DSR. The measurement parameters were set at eight different temperatures (40–82 °C in 6 °C steps) and a frequency sweep from 0.1 to 40 Hz. To better visualize the small changes in the material, the dynamic modulus and phase angle of 46 °C and 1.6 Hz are plotted as |G*|Dayi /|G*|Day1 and δDayi − δDay1 versus storage duration (days) which will be called ageing index AI|G*| and AIδ .
3 Results and Discussion 3.1 FTIR Spectroscopy The results from FTIR spectroscopy of the two different binders follow a similar trend when comparing sample storage under light and in the dark. The respective spectra of binder A (top) and B (bottom) are shown in Fig. 2. While exposed to light (left),
Fig. 2 FTIR spectra of binder A (top) and B (bottom) stored in the light (left) and in the dark (right)
Time and Storage Dependent Effects of …
1857
both binders exhibit a substantial increase in carbonyls (1700 cm−1 ) and sulfoxides (1030 cm−1 ) with storage time. Furthermore, an increase in the lower fingerprint area between 1300 and 900 cm−1 is observable over time. This can be explained by a shift in polarity towards more polar fractions upon ageing [3]. The intensity increase in the respective part of the spectrum goes in line with an increase in asphaltene content (or n-heptane insoluble fraction) within the binder. When being stored in the dark (right), only little changes can be seen. Small bands located at 1735 and 1110 cm−1 appear, which also rises during the exposure to light. These bands can be assigned to two different vibration modes of carboxylic acids [4]. The minor changes in the sulfoxide band at 1030 cm−1 can be explained by the fact that the formation of sulfoxides is taking place at a much faster rate due to the difference in electron configuration when comparing it to carbon. Therefore sulfoxides will usually form first during bitumen oxidation [5]. Another new band is coming up at 1260 cm−1 (only binder B) which is linked to the increase in the region around 1100 cm−1 and a band at 810 cm−1 . These bands could be assigned to a functional group such as an organic sulfate ester (ROSO3 − ) [6, 7]. Their ageing indices (AIFTIR ) of both binders are plotted in Fig. 3. RTFOT (Rolling Thin Film Oven Test) level of ageing (yellow) is reached after one day of storage under light, RTFOT + PAV (Pressure Ageing Vessel Test) level (orange) after day 6. When being stored in the dark, no notable changes in these two functional groups are being measured. Whereas binder A never exceeds RTFOT level of aging when stored in the dark, binder B does so, indicating a higher ageing susceptibility. With increasing storage time a deceleration of the surface oxidation is observed, which is given by a flattening of the power law fit in Fig. 3. This can be explained by the increasing incorporation of oxygen upon longer storage time, slowly passivating the surface and slowing down any further oxygen uptake into the material. Due to increasing viscosity during storage, the oxidized species at the surface are slowly transported into the material, which does allow further oxidization, just at a much slower rate.
Fig. 3 Ageing indices (AIFTIR ) of binder A and binder B stored at light and in the dark
1858
J. Mirwald et al.
Fig. 4 Ageing indices AI|G*| (left) and AIδ (right) of both binders stored at light and in the dark at 46 °C and 1.6 Hz
3.2 DSR The rheological data of binder A and B show similar trends in ageing progress with storage time. AI|G*| increases upon longer storage time, while AIδ decreases (see Fig. 4). The complex modulus of binder A stored in the light reaches an increase of 24% from day 1 to day 20, whereas the samples stored in the dark only reach an increase of 11%. The values after 20 days of storage in the light (AI|G*| of 1.2) are still far away from a RTFOT reference level (AI|G*| of 3.4). Binder B shows the same trend but is not affected by storage conditions. The difference in the complex modulus when stored in the light reaches an increase of 29% after 20 days of storage (AI|G*| of 1.3), while when stored in the dark it reaches an increase of 28%. This again is far below the respective RTFOT reference level (AI|G*| of 2.9). Interestingly, binder B shows a similar behavior for both storage conditions, indicating that it is more ageing susceptible regardless of storage conditions. Hence, storage time (and storage conditions, depending on the binder) have an impact on the results. The significance or the change in the material has a much smaller effect compared to FTIR spectroscopy, where RTFOT level is reached after 1 day and RTFOT + PAV level just after 6 days.
4 Conclusion Sample storage time and conditions have a major impact on the results of FTIR spectroscopy. Data treatment involving normalization, integration and statistical evaluation as well as proper sample preparation and storage is required for uniform results. Significant difference between storage under light or in the dark are observable. The current state of the art combined with our tests supports the proposal of normalization at the main CH3 band at 2920 cm−1 , followed by integration and optional referencing according to literature [1]. Sample preparation (short heat up phases) and storage has
Time and Storage Dependent Effects of …
1859
a significant impact and should be chosen carefully. With increasing storage time, a deceleration in oxidation at the surface is observed, indicating a oxidation slowdown and a passivation on the surface. Rheological bulk parameters on the other hand are not as much affected by storage time and conditions, but follow the same trend as in FTIR spectroscopy. Differences between storage in the light or in the dark can only be observed in some binders, but do not reach short-term ageing level of the binder. For both analysis methods, a proper sample storage in the dark is recommended. While samples for FITR spectroscopy should be measured within a day (or even within minutes or hours) after sample preparation, DSR samples can be stored in the dark and measured up to one day after preparation. Acknowledgement The authors would like to thank Solutions 4 Science Handels GmbH for providing the Perkin Elmer Spectrum TwoTM FTIR spectrometer and BMI Group S.à.r.l. for their financial support.
References 1. Hofko, B., et al.: Repeatability and sensitivity of FTIR ATR spectral analysis methods for bituminous binders. Mater. Struct. 50(3) (2017) 2. Hofko, B., et al.: FTIR spectral analysis of bituminous binders: reproducibility and impact of ageing temperature. Mater. Struct. 51(2), 45 (2018) 3. Handle, F., et al.: The bitumen microstructure: a fluorescent approach. Mater. Struct. 49(1–2), 167–180 (2016) 4. Petersen, J., et al.: Molecular interactions of asphalt. Tentative identification of 2-quinolones in asphalt and their interaction with carboxylic acids present. Anal. Chem. 43(11), 1491–1496 (1971) 5. Petersen, J.C., Glaser, R.: Asphalt oxidation mechanisms and the role of oxidation products on age hardening revisited. Road Mater. Pavement Des. 12(4), 795–819 (2011) 6. Chihara, G.: Medical and biochemical application of infrared spectroscopy. V. Infrared absorption spectra of organic sulfate esters. Chem. Pharm. Bull. 8(11), 988–994 (1960) 7. Larkin, P.: Infrared and Raman Spectroscopy. Principles and Spectral Interpretation, pp. 129– 130. 225 Wyman Street, Waltham, MA 02451, USA: Elsevier (2011)
Top-Down Cracking in Airfield Pavement Structures—An In-Situ Test Survey Amir Sadoun , Michaël Broutin , Ichem Boulkhemair , Alexandre Duprey, Arnaud Mazars, and Marc Huault
Abstract A new type of distresses, not taken into account in the frame of the French airfield flexible pavement design method, has appeared for some years on several pavement structures: this is top-down cracking, under the pass of heavy aircraft landing gears. It is actually considered in the design methods that the cracking is supposed to initiate at the bottom of the bituminous layers, where the tensile stresses are the highest, and then opens progressively up to the surface (bottom-up cracking). This phenomenon has been observed in recent years on several flexible airfield pavements in France. Toulouse airport (ATB) noticed this type of distresses on the 14R32L runway, which is a rigid pavement reinforced with a layer of bituminous materials. An experiment has been conducted by Toulouse airport, CEREMA, Airbus and the STAC, to study the pavement behavior and the evolution of this phenomenon. Ovalization measurements, deflection basins monitoring by means of inclinometers and vertical displacement of the concrete slabs measurements were performed under an Airbus A319 trafficking. In parallel, Heavy Weight Deflectometer (HWD) tests were carried out on several test points over underlying concrete slab center, slab edge and slab corner. A mechanical model was developed, based on a 2D finite element (FE) technique. It enables to consider either static or dynamical impulse HWD loading, and includes viscoelasticity properties for the surface bituminous materials. Keywords Top-down cracking · Semi-rigid pavement · Viscoelastic behavior · Dynamic FE modeling · HWD
A. Sadoun (B) · M. Broutin · I. Boulkhemair · A. Duprey French Civil Aviation Technical Center (STAC), 31 av. Maréchal Leclerc, 94381 Bonneuil-sur-Marne, France A. Mazars The French Agency for Ecological Transition and Regional Cohesion (CEREMA), 1 Avenue du Colonel Roche, 31400 Toulouse, France M. Huault Toulouse Airport, 31703 Blagnac, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_236
1861
1862
A. Sadoun et al.
Fig. 1. ATB 14R/32L pavement structure
1 Introduction Toulouse Airport (ATB) is the fifth biggest airport in France and handles more than 10 million passengers per year. It numbers two runways. The longest one, 14R/32L, 3,500 m long, is used for test flights, and has hosted the first flights of the Concorde, the A380, and more recently the A350. This runway was built between 1968 and 1969 and consists in 40 cm thick 5 × 7.5 m concrete slabs over a 15 cm thick layer of granular materials treated with a hydraulic binder. In 2003/2004, the central 22.5 m transversal joints were postdowelled with dual-dowels positioned at the mid thickness of the slabs and sealed with a resin, and the structure was reinforced (Fig. 1) with a 7–11 cm of High modulus asphalt concrete base course and 6 cm of aeronautical bituminous materials surface course. A 2 cm tar sands layer was added at the top of the concrete slabs to prevent reflection cracking. In 2013, cracks were observed at the surface of the bituminous materials, matching exactly the positions of the underlying concrete slabs joints. Corings (Fig. 2) have demonstrated that it is top-down cracking limited to the top layer, a phenomenon not taken into the frame of the French airfield flexible pavement design method [1].
2 Investigation Program The observations made on the 14R/32L runway challenged pavement mechanics current knowledge. This is the reason why ATB, the CEREMA and the STAC started
Top-Down Cracking in Airfield Pavement …
1863
Fig. 2. Top-down crack in the joint area
an ambitious program of investigations, analysis and modeling to better understand this top-down cracking phenomenon. The first phase consisted in an experimentation using the Heavy Weight Deflectometer (HWD) and some external instrumentation (surface deflection from inclinometer measurements, ovalization tests, vertical displacement of the concrete slabs using LVDT) to study the behavior of the pavement linked to the pass of an A319neo two wheels landing gear loaded at 337.5 kN. This study was performed on a limited sample including 3 transversal joints in sound, intermediate, and distressed areas. For the HWD, two load levels, 200 and 300 kN, were chosen to study the linearity of the pavement response, and for each load level, 3 drops were performed to study the repeatability of the tests. The aircraft tests were performed with a landing gear moving longitudinally with 5 cm from the transverse joint. In parallel, Heavy Weight Deflectometer (HWD) tests were carried out on several test points over underlying concrete slab center, slab edge and slab corner [2]. The temperature was monitored during the tests for bituminous materials and concrete slabs. The average temperature of the bituminous materials was around 24 ± 2 °C and there was no temperature gradient within the concrete slabs.
1864
A. Sadoun et al.
The consistency between the results allowed launching a second phase with only HWD tests, generalized all over the runway. In this paper, only phase 1 results are presented.
3 Analysis The LVDT measurements under A319neo landing gear pass indicate vertical displacements around 0.10–0.15 mm (Fig. 3), with opposite movements between the middle of the transverse joint (driving slab) and the corner (lifting slab). The measurement collected from the sensors set up on the two sides of the joint show that the load transfer is good. The ovalization measurements [3] at different depths within the bituminous materials demonstrate that the layers are fully-bonded. The deflection measured with the inclinometers in the cracked area is 50% higher than the ones measured in the sound area, and remain under 20/100 mm. These results are consistent with the HWD deflections (Picture 4). The study of the HWD deflections allows studying both the load transfer and the structural condition of the edges and corners of the slabs. The results showed that the
Fig. 3. LVDT measurements of vertical concrete slabs displacement
Fig. 4. Examples of HWD deflections (cracked area (a) and healthy area (b))
Top-Down Cracking in Airfield Pavement …
1865
load transfer of the underlying concrete slabs is good for both cracked and sounds areas. The structural health of slabs is satisfying for the two areas even if the indicators values are slightly different. Load transfer
Structural health
IT(B)
IT(Co)
IQ(B)
IQ(Co)
Cracked areas
0.60
0.40
1.39
2.17
Sound areas
0.62
0.42
1.21
1.62
The indicators used to evaluate the load transfer and the structural health, developed by the STAC for rigid airfield pavements, are defined below.
4 Modelling A 2D finite element modeling was developed by the STAC to simulate the HWD impulse loading and try to understand the top-down cracking phenomenon. The mesh includes: two concrete slabs separated with a dummy joint, the 2003– 2004 reinforcement continuous bituminous material layers and the HWD loading plate. The interfaces between all layers are considered as perfectly bonded. A new simulation with the tar sands and the debonding between the concrete slabs and materials treated with hydraulic binder will be developed in a second time, using [4] developments. Two simulations were compared, involving respectively an elastic or viscoelastic behavior for bituminous materials: • For the elastic modeling, the moduli are set to the value of new materials at the testing temperature and a 30 Hz frequency, representative to the HWD impulse load solicitation.
1866
A. Sadoun et al.
Fig. 5. Finite elements simulation of the HWD test on the joint area
• For the viscoelastic one, a Huet-Sayegh model is implemented in the time domain. With the absence of material characterization, complex moduli from the STAC test facility are used. Picture 6 presents the stresses produced by the impulsive loading in the pavement at loading peak value and during unloading phase. The results show (by convention, tensile stresses are taken positives): • Similar compression stresses under loading plate at peak for both simulations (elastic and viscoelastic), with higher side effects for the viscoelastic simulation, • A higher tensile stresses at the top of the bituminous materials than the bottom for the viscoelastic simulation, • Appearance under the loading plate at the end of the unloading phase, of tensile stresses at the top of the bituminous materials who are significantly higher in the viscoelastic simulation than the elastic one (+200%), • A higher tensile stresses at the top of the bituminous materials than the bottom for the viscoelastic simulation. The viscoelastic effect has then been highlighted, even if other phenomena may be involved. Modeling improvements are planned, e.g. by adding dowels between the concrete slabs, or by taking into account the debonding between the concrete and bituminous layers, or upgrading the simulation to a 3D finite element modeling.
5 Conclusion This paper presents an ambitious experimentation conducted by ATB, CEREMA, Airbus and the STAC, to understand the top-down cracking appearance phenomenon. It involved HWD tests and measurements with an external instrumentation under the pass of a test aircraft landing gear. The analyses conducted permitted to better understand the encountered phenomena. Numerical modeling works demonstrated a significant effect of the
Top-Down Cracking in Airfield Pavement …
1867
Compression stresses
Tensile stresses
Compression stresses
Fig. 6. Stresses under HWD loading
viscoelasticity on the tensile stresses generated at the top of the bituminous materials, and then its contribution to the top-down cracking initiation.
References 1. Airfield flexible pavement assessment using HWD. In: Guide Technique https://www.stac.avi ation-civile.gouv.fr/(2014) 2. Sadoun, A., Broutin, M.: Advanced modelling for rigid pavement assessment using HWD. 6th Transport research arena 18–21 April 2016. Transp. Res. Procedia 14, 3572–3581 (2016) (Elsevier) 3. Gharbi, M., Broutin, M., Boulkhemair, I., Nguyen, M. L., Chabot, A.: Analysis of ovalization measurements on flexible airfield pavement under HWD dynamic impulse load. In: Submitted to the International Symposium on Bituminous Materials (ISBM) (2020) 4. Sadoun, A., Broutin, M., Simonin, J.-M.: Assessment of HWD ability to detect debonding of pavement layer interfaces. In: Chabot, A., Buttlar, W.G., Dave, E.V., Petit, C., Tebaldi, G. (eds.) 8th RILEM International Conference on Mechanisms of Cracking and Debonding in Pavements, pp. 763–769. Springer, Netherlands, Dordrecht (2016)
Tracking Degree of Blending Between Recycled and Virgin Binder Through Asphalt Mix Phase Angle Payman Pirzadeh and Hassan Baaj
Abstract Performance of Reclaimed Asphalt Pavement (RAP) in a new mix depends on the degree of blending between the aged asphalt binder of the RAP and the new virgin binder. The extent of blending determines the homogeneity of the new mix and thus its performance. A recent study by ExxonMobil and University of Waterloo has indicated that silo storage improves homogeneity of RAP-containing mixes post manufacturing. Samples were collected at selected silo storage durations. Resistance to rutting, fatigue and low temperature cracking exhibited improvement. However, the complex modulus mastercurves exhibited a reduction in |E*| values at lower frequencies up to 12 h of storage, which was unexpected. In addition to lower |E*| values, the slopes of mastercurves were changing with silo storage time. Literature suggests that the slope of a log-log plot of |E*| versus frequency is an estimate of the phase angle of the material. Therefore, deriving phase angle from the mix mastercurves, the evolution of blending between RAP and virgin binder could also be tracked. Data suggests that the derived phase angle is also sensitive enough to detect the aging that would happen as a result of a prolonged silo storage or incompatible binder chemistry. Keywords RAP · Diffusion · Phase angle · Asphalt mix · Blending · Silo storage
1 Introduction Utilization of reclaimed asphalt pavement (RAP) in manufacturing new asphalt mix has been a common practice in asphalt industry [1–5]. Availability of aggregate and binder in RAP to be recycled back into new asphalt mix has been a strong driving force in promoting sustainability of asphalt industry [5]. Recent reports from National Asphalt Pavement Association (NAPA) is pointing to an increase in average P. Pirzadeh (B) Sarnia Technology Application and Research, Imperial Oil Ltd, Sarnia, ON, Canada e-mail: [email protected] H. Baaj University of Waterloo, Waterloo, ON, Canada © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_237
1869
1870
P. Pirzadeh and H. Baaj
RAP content of asphalt mix, i.e. from 15.6% in 2009 to 20.1% in 2017 [6]. While recycling has significantly improved sustainability of asphalt industry, introduction of used aged binders into new asphalt mix has also brought new challenges [4, 7–10]. The binder in RAP is usually much stiffer than virgin binder due to suffering aging during its service life. Therefore, increase in RAP content of mix requires attention to three major concerns: (1) managing the stiffness of RAP portion, (2) achieving the appropriate final gradation of the manufactured mix, and (3) sufficient blending between aged and virgin binder to achieve the appropriate cohesion within the final resulting binder as well as adhesion between the binder and aggregates [4, 7–9]. Failure of addressing the mentioned concerns results in a heterogeneous mix with weak spots where stress accumulates over time. This stress accumulation typically causes pre-mature failure of the pavement [7–9]. Several studies have been conducted to better understand the mechanism and efficiency of blending between aged stiff binder of RAP and soft virgin binder [10– 19]. Summary of all the laboratory studies indicate that blending between the binders can be modelled in two stages: (1) mechanical spreading during which the two binders are mechanically mixed with each other and a contact surface area is established between them, (2) diffusion where material start to mingle with each other due to difference in concentration [7, 8, 10, 19]. It has been demonstrated that diffusion is the dominant process and significantly influences extent of blending between the binders [7, 8, 10]. Laboratory-scale research by Kriz et al. indicated that the diffusion phenomena between aged and virgin asphalt binders follows Arrhenius Law, i.e. higher temperatures can facilitate progress of diffusion [7]. It was also demonstrated, for the examined binders and mixes, that a typical time of 10–12 h is required for diffusion to approach completion [8]. Recently in collaboration with University of Waterloo, ExxonMobil extended the diffusion study to the field, i.e. studied impact of diffusion on plant produced RAP-containing mix, with a RAP content ranging from 15–40%. Asphalt silo was utilized as a platform to keep the manufactured mix samples at a temperature range of 140–150 °C. Mix samples were collected at selected silo storage times of 0–24 h. Performance assessment of the mix on resistance to rutting, fatigue and low temperature cracking highlighted a progressive improvement during 0–12 h of storage [9]. Complex moduli of sampled mixes were also measured in a range of frequencies and temperatures [9]. Mastercurves of sampled mixes were constructed, as demonstrated in Fig. 1. Up to 12 h of storage, all mixes exhibited a decrease in |E*| at lower frequency range. However, hardening of asphalt was noted upon storage up to 24 h, as |E*| increased. Binder examination revealed that this increased stiffness can be due to various aging processes, and in one case specifically was due to oxidation from the chemistry of the utilized binder [9]. An interesting feature of the constructed mastercurves was the evolution of the slope of log|E*| or log|G*| versus logω as a function of silo storage duration. Booij and Thoone demonstrated the importance of this slope through generalization of Kramer-Kronig transforms for both viscoelastic solids and liquids [20]:
Tracking Degree of Blending Between Recycled and Virgin …
1871
Fig. 1 Mastercurves of |E*| for two surface courses (HL-3) collected from two asphalt plants as a function of silo storage time from Plants 1 and 2 (P1 and P2).
π ∂ log G ∗ , δ≈ 2 ∂ log ω
(1)
where δ is the phase angle of the viscoelastic material. Essentially, Eq. (1) suggests that, in the absence of phase angle measurements, this quantity can be estimated from the mastercurve of the material on log-log scale (see Fig. 2a). Later, Rowe [21] and Oshone et al. [22] verified the validity of this approximation in Eq. (1) for various types of asphalt binders (G*) and mixes (E*) in both shear and extensional modes of measurement. While Oshone et al. warned about applicability of Eq. (1) on RAP-containing mixes due to unknown degree of blending between aged and virgin binders, they provided an example where reasonable agreements were observed between predicted and measured phase angles in mix sample containing 30% RAP. This potentially highlights the capability of estimating mix phase angle through slope approach if thermal histories of collected samples are well-controlled. This paper intends to demonstrate that Eq. (1) can semi-quantitatively track evolution in diffusion between binders and as a result the extent of blending between them.
2 Materials and Methods The details of sample collection and analysis, as well as the operating conditions of the asphalt plants, are available in [9]. In summary, two asphalt plants, referred to as Plant 1 and Plant 2, collaborated in manufacturing samples. Two courses were collected from each plant: surface (HL-3) and base (HL-8), for a total of four mix samples. The two base courses contained 20% and 40% RAP, and the two surface courses had 15 and 30% RAP. Samples were collected at pre-selected times of storage: 0, 1, 4, 8, 12 in both plants, and 24 h in Plant 2. Mixes were designed according to pavement specifications in the Province of Ontario, Canada. The virgin binders utilized during
1872
P. Pirzadeh and H. Baaj
(a)
(b)
Fig. 2 a A typical asphalt binder mastercurve on a Log-Log scale; the slope of this curve is an estimation of phase angle (δ) of the binder at a given frequency. b Mastercurves of surface course mix as a function of silo storage time; at lower frequencies, the slope of mastercurves are different for different storage durations.
manufacturing were PG 58–28 and PG 52–34 for the surface and base courses, respectively. The complex moduli of different HMA mixtures were determined using Material Testing System (MTS) at the University of Waterloo according to AASHTO TP 62– 07. Compacted samples were subjected to a repetitive, compressive, and sinusoidal load. Two Linear Variable Differential Transducers (LVDTs) were utilized to measure the deformation of test specimen. Triplicate Ø100 × 150H mm cylinder specimens were used in this test. All mix samples were examined at six loading frequencies (0.1, 0.5, 1.0, 5.0, 10.0 and 25 Hz) and five temperatures (−10.0, 4.4, 21.1, 37.8, and 54.4 °C). Mix mastercurves were constructed by merging the collected moduli to a reference temperature of 21.1 °C. Per Fig. 2b, the log|E*| versus logω for data f < 1 Hz were fit to a straight line to estimate the phase angles by the slope (see Fig. 2b).
3 Results and Discussion Figure 3 summarizes the calculated average slopes, i.e. the average phase angle (δ), from the mastercurves of measured |E*| for the four collected mix courses. Within the variabilities of sampling and test methods, a trend of increase in phase angle is noted in all samples between 0 to 12 h of silo storage. This trend is an indication of diffusion progress. At the beginning of blending, RAP acts as a black rock. As a result, a lower phase angle is exhibited by the mix. Due to its lower phase angle at the 0 h of storage, the mix has more elastic characteristics which can make it susceptible to stress accumulation and premature failure. As diffusion progresses by storage time, larger content of RAP binder gets activated and contributes to the binder content of the mix, as also suggested by reduction in |E*| values in Figs. 1 and 2b (and evolution of VMA and VFA as well [9]). Increased binder content and higher
Tracking Degree of Blending Between Recycled and Virgin …
1873
Fig. 3 Average phase angle of collected mix courses as estimated based on average slope from Fig. 2b as a function of silo storage time. P1 and P2 indicate Plant 1 and Plant 2, respectively; HL-3 and HL-8 indicate surface and base course mixes, respectively.
temperatures help the homogeneity of overall binder blend and stress accumulating points. In addition, better homogeneity enables the binder to relax the stresses more easily, which is reflected by the overall increase in phase angle magnitude. However, in case of Plant 2, a decrease in phase angle is noted between 12 and 24 h of storage. This decrease in phase angle is accompanied by an increase in stiffness of the mix. As investigated earlier [9], the stiffness increase was verified by increase in PG grading at both low and high temperatures of the binder. Specifically in the case of Plant 2 base course, chemistry of binder had a role in premature oxidative aging of the binder and faster stiffening of the material [9]. Further investigation on the data is underway to better understand phase angle behaviour from the presented mastercurves. It is expected that the shape and patterns in the phase angle mastercurves can further shed light on evolution of diffusion and estimating extent of blending between aged and virgin binders.
4 Conclusion Complex modulus mastercurves of both asphalt binder and mix contain information about the characteristic of material at different temperatures and frequencies of loading. While the primary objective is to understand the stiffness of material at a wide range of temperature or loading times, the slope of log|E*| − logω mastercurves provide estimates of phase angle of the material, i.e. microstructure characteristics. As demonstrated in this study, |E*| mastercurves of RAP-containing mix
1874
P. Pirzadeh and H. Baaj
samples exhibit different behaviours and evolve over time, blending between binder progresses, and even when binder ages due to long storage times. Tracking the slopes of these mastercurves provides indications on progress of blending or exposure to (premature) aging. Further study is needed to investigate patterns of evolution in phase angle to optimize blending of binders when RAP is utilized.
References 1. Bonaquist, R.: Can I run more RAP? In: Hot Mix Asphalt Technology, pp. 11–13 (2007) 2. Ozer, H., et al.: Evaluation of RAP impact on hot-mix asphalt design and performance. J. Assoc. Asphalt Paving Technol. 78, 317–351 (2009) 3. McDaniel, R.S., et al.: Effects of reclaimed asphalt pavement content and virgin binder grade on properties of plant produced mixtures. Road Mater. Pavement Des. 13(S1), 161–182 (2012) 4. Zaumanis, M., Mallick, R.B.: Review of very high-content reclaimed asphalt use in plantproduced pavements: state of the art. Int. J. Pavement Eng. 16, 39–55 (2015) 5. Liu, S., et al.: Technological, environmental and economic aspects of asphalt recycling for road construction. Renew. Sustain. Energy Rev. 75, 879–893 (2017) 6. Williams, B.A., Copeland, A., Carter Ross, T.: Asphalt pavement industry survey on recycled materials and warm-mix asphalt usage. In: 2017, 8th Annual Survey Information Series 138 (2018) 7. Kriz, P., et al.: Blending and diffusion of reclaimed asphalt pavement and virgin asphalt binders. Road Mater. Pavement Des. 15, 78–112 (2014) 8. Kriz, P., et al.: Reclaimed Asphalt Pavement—Virgin Binder Diffusion in Asphalt Mixes. Canadian Technical Asphalt Association, Winnipeg, MB, Canada (2014) 9. Pirzadeh, P. et al.: Impact of asphalt plant silo storage to enhance blending between RAP & virgin binders in hot mix asphalt. J. Road Mater. Pavement Des. (2019) https://doi.org/10.1080/ 14680629.2019.1684346 10. Zhang, K., Muhunthan, B.: Effects of production stages on blending and mechanical properties of asphalt mixtures with reclaimed asphalt pavement. Constr. Build. Mater. 149, 679–689 (2017) 11. Ma, T., et al.: Influence of preheating temperature of rap on properties of hot-mix recycled asphalt mixture. J. Test. Eval. 44, 762–769 (2016) 12. Pérez Madrigal, D., et al.: Effect of mixing time and temperature on cracking resistance of bituminous mixtures containing reclaimed asphalt pavement material. J. Mater. Civ. Eng. 29, 0417058 (2016) 13. Mogawer, W.S., et al.: Determining the influence of plant type and production parameters on performance of plant-produced reclaimed asphalt pavement mixtures. Transp. Res. Board: J. Transp. Res. Board 2268, 71–81 (2012) 14. Mogawer, W., et al.: Performance characteristics of plant produced high RAP mixtures. Road Mater. Pavement Des. 13, 183–208 (2012) 15. McDaniel, R.S., et al.: Investigation of low- and high-temperature properties of plant-produced RAP mixtures. Federal Highway Report, FHWA-HRT-11-058 (2012) 16. Zhao, S., et al.: Quantitative evaluation of blending and diffusion in high RAP and RAS mixtures. Mater. Des. 89, 1161–1170 (2016) 17. Huang, B., et al.: Laboratory investigation of mixing hot-mix asphalt with reclaimed asphalt pavement. Transp. Res. Board: J. Transp. Res. Board 1929, 37–45 (2005) 18. Zhao, S., et al.: Quantitative characterization of binder blending: how much recycled binder is mobilized during mixing. Transp. Res. Board: J. Transp. Res. Board 2506, 72–80 (2015) 19. Yousefi Rad, F., et al.: Application of diffusion mechanism: degree of blending between fresh and recycled asphalt pavement binder in dynamic shear rheometer. Transp. Res. Board: J. Transp. Res. Board 2444, 71–77
Tracking Degree of Blending Between Recycled and Virgin …
1875
20. Booij, H.C., Thoone, G.P.J.M.: Generalization of Kramers-Kronig transforms and some approximations of relations between viscoelastic quantities. Rheol. Acta 21, 15–24 (1982) 21. Rowe, G.M.: Phase angle determination and interrelationships within bituminous materials. In: Loizos, Partl, Scarpas, Al-Qadi (eds.) RILEM Symposium on Advanced Testing and Characterization of Bituminous Material, pp. 43–52. Taylor & Francis Group, London 22. Oshone, M., et al.: Prediction of phase angles from dynamic modulus data and implications for cracking performance evaluation. Road Mater. Pavement Des. 18(Sup4), 491–513 (2017)
TSRST Heterogeneous Modeling for Understanding Failure Depending on Asphalt Mix Design and Experimental Conditions Fateh Fakhari Tehrani, Christophe Petit, and Joseph Absi
Abstract The Thermal Stress Restrained Specimen Test (TSRST) is used to determine the low temperature cracking susceptibility of asphalt mixture. In this test, a specimen is subjected to a constant temperature decrease rate. Due to the prohibited thermal shrinkage, cryogenic stress is built up in the specimen. The progression of stresses imposes failure in the specimen. There is certain experimental limitation to measure the real value of the stress. This paper aims to develop the numerical approach permitting to estimate the local stress and strain value in the specimen. This approach consists of the heterogeneous multiscale modeling of the asphalt mixture. The heterogeneous scales ranging from smallest to largest are mastic, mortar and HMA (hot mix asphalt). Each scales’ length is made out of a viscoelastic matrix and an elastic aggregates skeleton. Granular skeletons are generated using a Random object modeler software or MOA (Modeleur d’Objet Aléatoire). These models correspond to a specific aggregate grading curve. Thereafter the heterogeneous geometries are imported into ABAQUS finite element code. The coupled thermal and mechanical numerical simulations are realized at different temperatures. The intensity of local stress and strain values will be compared to the ones found from the experimental measurement. The aim of this modelling is to try to understand the failure strength not depending only on the material but also on environmental parameters, cooling system, specimen size and also formulation of mixes. Keywords TSRST · Heterogeneous approach · Viscoelastic · Finite element modeling
F. F. Tehrani (B) · C. Petit Université de Limoges, GC2D Egletons, France e-mail: [email protected] F. F. Tehrani Conservatoire National des Arts Et Métiers, Paris, France J. Absi Université de Limoges, IRCER, UMR-CNRS, 7315 Limoges, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_238
1877
1878
F. F. Tehrani et al.
1 Introduction At ambient temperature, the mechanical behavior of asphalt mixes is viscoelastic. By increasing the temperature, compression stresses are created which are then dissipated by relaxation. At low temperature, asphalt concrete behaves as a rigid elastic material. Therefore, when the temperature drops, tension stresses are created without being able to dissipate. When these tensions stresses reach the limit of resistance of asphalt concrete, cracks appear. In this study, a numerical finite element model based on a heterogeneous multiscale approach is developed. In this approach, the bituminous composites at each scale are considered as a biphasic medium constituted of elastic Inclusions and a viscoelastic matrix. The main objectives of this numerical approach are: first to understand the initiation of cracking and failure in bituminous media, secondly to identify numerous effects such as climatic chamber, elaboration mixes, size and boundary conditions. This study will be associated to RILEM TC PIM in Task Group TG2 regarding low temperature performance of mixes.
2 Numerical Approach 2.1 Material This study is based on an asphalt concrete 0/10 mm (BB 0/10) with 5.4% of binder. Its aggregate grading curve is continuous, and contains diorite aggregates of 0/2, 2/6, and 6/10 mm. Moreover, 1% of limestone filler was added in the mix formula.
2.2 Random Generation of Aggregate Skeleton In general, Commercial finite element for numerical simulation is unable to create models that contain a random distribution of an aggregate gradation. Therefore, Random Object Generator (MOA) which is a tool developed in our laboratory can be used to generate three-dimensional inclusions randomly spaced within a finite volume; a parallelepiped or a cylinder in three-dimensional space. The inclusions can be as non-intersecting or intersecting. In this study, the inclusions are considered separate with a minimal distance imposed by the user. Physically, this distance is representative of the layer of viscoelastic matrix surrounding the inclusions [1, 2]. In this study, ten random distributions were generated and the closest curve to the aggregate curve of asphalt mixes has been chosen. This generation was imported into finite elements code such as ABAQUS. Three phases can then be constructed through the application of successive Boolean operations, these phases being the
TSRST Heterogeneous Modeling for Understanding Failure Depending …
2
L D
D= 60 mm
L=240 mm
L=210 mm
L=180 mm
L D
D= 60 mm
D= 60 mm
D= 60 mm
L=150 mm
L=120 mm
D= 60 mm
1879
2 .5 L D
3 L D
3.5 L D
4
a)
b)
Fig. 1 a 3D models of the asphalt concrete. b The imposed boundary conditions
inclusions, matrix and air voids. Figure 1 presents a model with 3D inclusions and imposed boundary conditions. Since the asphalt concrete is a high compact material (≈85% volume filling rate by aggregates), it is impossible to create the complete geometry using a unique scale. Therefore, a multi-scale approach must be adopted. This method considers asphalt mixture as a biphasic medium composed of elastic aggregates and a viscoelastic matrix for low strain level [3]. In this study four scales are considered as: 1. 2. 3. 4.
Bitumen scale considered as a homogeneous scale Mastic scale which contains the bitumen and the filler with a diameter of less than 100 µm. Mortar scale which contains the mastic and the sand with a diameter of less than 2 mm. HMA scale showing mastic and gravel (2 mm < d ≤ 10 mm)
2.3 Thermal and Mechanical Properties The thermal properties of aggregates and bitumen have been taken from the literature [4] (see Table 1). Thermal properties, conductivity, specific heat and density, have been calculated using the following mixing laws: 2λbitumen + λsand − 2Vsand × (λbitumen − λsand ) 2λbitumen + λsand + Vsand × (λbitumen − λsand ) = (1 − m sand ) × C P−bitumen + m sand × C P−sand = (1 − Vsand ) × ρbitumen + Vsand × ρsand
λmor tar = λbitumen × C P−mor tar ρmor tar
1880
F. F. Tehrani et al.
Table 1 Thermal properties of matrix and inclusions
Conductivity (W m−1 K−1 )
Specific heat (J kg−1 K−1 )
Density (kg m−3 )
Biyumen
0.162
1560
1030
Sand
2.59
650
2750
Mortar
0.778
791
2190
1,0E+05 1,0E+04 1,0E+03 1,0E+02 1,0E+01 1,0E+00 1,0E-01 1,0E-02 -40
-35
-30
-25
-20
-15
-10
-5
0
5
10
Temprature (C°) Fig. 2 Equivalent elastic modulus of mortar at f = 1E − 5Hz [5]
The mechanical behavior of the aggregates is considered elastic with a Young’s modulus of 70 GPa and a Poisson’s ratio of 0.15. As it is well known, the behavior of HMA at low temperature is almost elastic. Therefore, equivalent elastic modulus of mortar at f = 1E − 5 is used as an input in the numerical models. (see Fig. 2).
2.4 Mesh, Boundary Conditions and Loading The 10-node quadratic tetrahedral elements (C3D10M a 10-node modified quadratic tetrahedron in ABAQUS code) were used to meshing the 3D microstructure. For a given geometry, several meshing densities, with varying degree of refinement, have been generated to identify the optimal mesh size (see Fig. 3). To assure the continuity of the displacement, a fully cohesive state is imposed on the entire inclusion surface using the ‘tie contact’, a function predefined in ABAQUS software. [2]. The boundary conditions are: • On the bottom layer, the vertical and horizontal displacements were blocked • The initial temperature (5 °C) was imposed on the external surface.
TSRST Heterogeneous Modeling for Understanding Failure Depending …
a) Parallelepiped specimen
1881
b) Cylindrical specimen
Fig. 3 Meshing operation for 3D models
Figure 1b shows the boundary conditions imposed on the numerical models. In this work, the temperature gradient during a TSRST test is considered −10 °C/h which is sufficiently slow to obtain a low cooling rate in the sample during the test.
3 Results Figure 4 shows the numerical and experimental variation of stress of TSRST specimen in function of temperature. Experimental results are provided by 4 laboratories in RILEM TC PIM TG2. 4,5 4
Stress (MPa)
3,5
Experimental results
3
Numerical simulation results
2,5 2 1,5 1 0,5 0 -30
-25
-20
-15
-10
-5
0
5
10
Temprature (C°) Fig. 4 Numerical and experimental variation of stress in function of temperature for TSRST specimen
1882
F. F. Tehrani et al.
L= 180 mm
L= 120 mm
Figure 5 shows the distributions of the local maximum principal stress obtained for heterogeneous numerical models with different dimensions. It is observed that the maximum stress was obtained at the top and bottom sides of the models. The concentration of the local stress in these areas leads to the failure of the specimen. This phenomenon has been observed for cylindrical and rectangular experimental specimens. (Fig. 6) Based on these results and in order to avoid the platen effect, it will be interesting to make a bead of glue around the samples with a reduced angle (almost 45°).
D=60mm
D=60mm
Fig. 5 Local stress values of numerical heterogeneous models of HMA
Fig. 6 Experimental Rilem TG2 TSRST specimen. a Cylindrical, b rectangular
TSRST Heterogeneous Modeling for Understanding Failure Depending …
1883
4 Conclusion In this paper, a heterogeneous numerical approach capable of calculating the thermomechanical properties of HMA from the properties of its components are developed. Digital models have been generated thanks to “MOA”, a software written in C++ language which generates random aggregate. we will be orienting our future work towards a parametric study highlighting the influence of the specimen size and cooling system rate.
References 1. Tehrani, F.F., Absi, J., Allou, F., Petit, C.: Heterogeneous numerical modeling of asphalt concrete through use of a biphasic approach: Porous matrix/inclusions. Comput. Mater. Sci. 69, 186–196 (2013) 2. Tehrani, F.F., Absi, J., Allou, F., Petit, C.: Investigation into the impact of the use of 2D/3D digital models on the numerical calculation of the bituminous composites’ complex modulus. Comput. Mater. Sci. 79, 379–389 (2013) 3. Olard, F., Di Benedetto, H.: General “2S2P1D” model and relation between the linear viscoelastic behaviours of bituminous binders and mixes. Road Mater. Pavement Des. 4(2), 185–224 (2003) 4. Riahi, E., Allou, F., Botella, R., Dubois, F., Absi, J., Petit C.: Quantification of self-heating and its effects under cyclic tests on a bituminous binder, Int J Fatigue. 104: 334–341 (2017) 5. Delaporte, B., Di Benedetto, H., Sauzéat, C., Chaverot, P.: In: International Conference on Bearing Capacity of Roads Railways and Airfields, Trondheim, Norway (2005)
Understanding of Structural and Surface Tension Properties of Asphalt Model Using Molecular Dynamics Simulation Lingyun You, Theodora Spyriouni, Qingli Dai, Zhanping You, Jaroslaw W. Drelich, and Ashok Khanal
Abstract The objective of this study is to characterize the molecular structural and surface tension properties of asphalt model using molecular dynamics (MD) simulation. The previously proposed 12-component asphalt model (AAA-1) was used in the MD simulation to evaluate the surface tension and structural properties in the molecular level of asphalt at different temperatures. The calculated densities from the asphalt models were in good agreement with the reference data. The asphalt model was validated with this comparison. The surface tension was predicted from the MD models and compared with the with the pendant drop test methods. The results showed the reasonable comparison. The molecular structural properties of asphalt model mainly on the radial distribution function (RDF) was also investigated. The first peaks and the highest peaks between groups of molecules identified the most contributed atom pairs in the asphalt group models which have general agreement with references. Based on these, the proposed MD simulations provide insights to understand the molecular structural and surface tension properties of asphalt at different temperatures. Keywords Molecular dynamics · Rheology · Structural changes · Surface tension L. You School of Civil and Hydraulic Engineering, Huazhong University of Science and Technology, Wuhan 430074, China L. You · Q. Dai · Z. You (B) Department of Civil and Environmental Engineering, Michigan Technological University, Houghton, MI 49931, USA e-mail: [email protected] T. Spyriouni Scienomics SARL, 16 rue de l’Arcade, 75008 Paris, France J. W. Drelich Department of Materials Science and Engineering, Michigan Technological University, Houghton, MI 49931, USA A. Khanal Department of Chemistry and Biochemistry, University of the Incarnate Word, San Antonio, TX 78209, USA © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_239
1885
1886
L. You et al.
1 Background Molecular dynamics (MD) simulations represent a powerful tool for the investigation and study of asphalt mixtures at the molecular level. The MD term refers to statistical mechanics-based theoretical approaches that are used to mimic the physical and chemical properties of asphalt subjected to different conditions. Modeling asphalt using all the molecular components it contains is an overwhelming challenge due to the high number of components and the fact that the actual asphalt composition is a matter of dispute. Therefore, it is necessary to further study the MD modeling of asphalt with an accuracy component assumption.
2 Molecular Dynamics and Experimental Program 2.1 Molecular Dynamics Details The AAA-1 asphalt model proposed by Li and Greenfield [1] was adopted in this study. The type and number of molecules in this model were adjusted to fit the laboratory results for atomic composition and asphalt specifications of performance grade (PG) 58–28. The AAA-1 model consisted of 12 components from the following classes: asphaltenes, naphthene aromatics, polar aromatics and saturates [1]. Three different kinds of molecular structures were used to represent the asphaltene fraction: asphaltene-phenol, asphaltene-pyrrole, and asphaltene-thiophene. The amount of the molecules in the model is three, two, and three, respectively. Also, there are two kinds of molecular structures selected to represent the naphthene aromatics fraction: cyclohexane-naphthalene (DOCHN) and perhydrophenanthrene-naphthalene (PHPN). The amount of DOCHN and PHPN in the asphalt model are thirteen and eleven, respectively. Benzobisbenzothiophene, pyridinohopane, quinolinohopane, and thioisorenieratane were selected to represent the polar aromatic fraction (resins) in the asphalt model. Here, the amount of these molecules is fifteen, four, four, and four, respectively. Meanwhile, the amount of four hopane and four squalane molecules were used in the asphalt model to represent the saturate fraction. In this work, the partial atomic charges were calculated according to the Generalized Amber Force-Field procedure as follows. The geometry of the isolated molecules was optimized using a post-Hartree-Fock method, the Møller-Plesset perturbation theory (MP2), with the 6-31G* basis set. Subsequently, a single-point energy calculation was performed using the Hartree-Fock method and the 6-31G* basis set for calculating the charges with the Mertz-Singh-Kollman scheme [2]. The calculations were executed within the MAPS platform of Scienomics (Materials and Processes Simulations Platform, Version 4.2, Scienomics SARL, Paris, France). The initial configurations of the asphalt model were built based on Theodorou and Suter method [3] implemented in MAPS. There are three different initial configurations created at each modeling temperature to average over different initial mixing of
Understanding of Structural and Surface Tension Properties …
1887
the components. The following procedure was used for system preparation. Firstly, the modeling box dimensions were set to 45, 35, and 35 Å along the x, y, and z-axis, respectively (density 1.0 g/cm3 approximately). Secondly, the configurations were minimized by initially applying steepest descent, followed by conjugate gradient minimization. The minimized configurations were subjected to NVT runs for 50 ps at 298.15 K, followed by NPT runs at 298.15 K and ambient pressure for 200 ps. Subsequently, the configurations were compressed to high pressure 100 atm (NPT runs) for 1 ns to eliminate the holes, and then, they were subjected to NVT for 50 ps and finally, to NPT runs at ambient pressure and 298.15 K for 1 ns. The densities resulting from the final NPT runs were typically higher than those from the initial NPT runs, indicating that holes were destroyed following this procedure. The long-range electrostatic interactions were computed using the particle mesh Ewald approach with a cut-off distance of 12 Å. The MD computations were carried out using LAMMPS through the MAPS. To calculate the surface tension from MD, the conventional Kirkwood-Buff and Irving-Kirkwood theories were employed here. As illustrated in Eq. (1), where L y is the box length along the y-direction, and Pij is the ij component of the pressure tensor. Factor ½ reflects that there are two interfaces in the system due to the periodic boundary conditions [4]. τ=
Ly Pzz + Px x Pyy − 2 2
(1)
The periodic box of asphalt was extended in one dimension from 45 to 100 Å in order to create an interface with a vacuum. The resulting configuration was run in the NPT ensemble for 2 ns for achieving equilibrium with a clear interface between the asphalt and the vacuum above it. The average surface tension of asphalt was calculated from the equilibrated section of the trajectory, typically the last 1 ns.
2.2 Surface Tension Measurement with the Pendant Drop Method The surface tension of asphalt was evaluated using the pendant drop method, as shown in Fig. 1. In the pendant drop test, two parameters of the pendant drop that should be experimentally determined are the equatorial diameter D and the diameter d at the distance D from the top of the drop. The surface tension of asphalt specimen would then result from Eq. (2) [5]. γ = ρg D 2 H
(2)
H −1 = B4 S a + B3 S 3 − B2 S 2 + B1 S − B0
(3)
1888
L. You et al.
Table 1 Empirical constants [5] S
a
B4
B3
B2
B1
B0
0.401–0.46
2.56651
0.32720
0
0.97553
0.84059
0.18069
0.46–0.59
2.59725
0.31968
0
0.46898
0.50059
0.13261
0.59–0.68
2.62435
0.31522
0
0.11714
0.15756
0.05285
0.68–0.90
2.64267
0.31345
0
0.09155
0.14701
0.05877
0.90–1.00
2.84636
0.30715
– 0.69116
– 1.08315
– 0.18341
0.20970
Fig. 1 Laboratory experiments for asphalt surface tension: a set-up of the pendant drop method, b an asphalt drop at 373.15 K (100 °C), and c a typical asphalt drop and schematic illustration of the selected plane method.
where γ is the surface tension of the asphalt specimen, measured in N/m, ρ is the density difference between asphalt and air, g is the acceleration due to gravity (9.8 m/s2 ), and H is the shape parameter of asphalt drop. H depends on the value of the shape factor S, where S is defined as d/D [5], and it can be calculated using Eq. (3). In Eq. (3), Bi (i = 0, 1, 2, 3, 4) and a are empirical constants for a specific range of S as demonstrated in Table 1.
3 Results and Discussion 3.1 Asphalt Density Typically, the asphalt density is used for validating the model and force field used for the MD simulations. In this study, the density was calculated from NPT simulations for the temperatures of 298.15, 333.15, 360.15, 340.15, and 440.15 K. At each temperature, the density was calculated as an average over three different initial configurations from runs of 1.5 ns. The equilibrated part of the trajectory, usually
Understanding of Structural and Surface Tension Properties …
1889
Table 2 Asphalt density comparisons Temperature (K) Density (g/cm3 )
298.15
333.15
360.15
340.15
440.15
Li and Greenfield [1]
0.998
0.970
0.959
0.931
0.904
Current model
0.986
0.968
0.957
0.933
0.905
the last 1 ns was used for calculating the average. The calculation results were compared with the density data reported by Li and Greenfield [1]. The comparisons were demonstrated in Table 2. It can be observed that the calculation data is a good agreement with the reference data, which indicates that the current asphalt model and force field describe the asphalt system satisfactorily.
3.2 Surface Tension The pendant drop method was used to measure the surface tension of asphalt at 353.15, 363.15, 373.15, and 383.15 K. It was measured equal to 29.5, 29.0, 28.3, and 27.8 mN/m for the above temperatures, respectively, as shown in Fig. 2. Each data point of laboratory surface tension represents the average of three replicates. In addition, the surface tension was calculated with the MD simulations at 298.15, 328.15, 343.15, 363.15, 383.15, 408.15, and 438.15 K. The results are shown in Fig. 2. A comparison with the experimental results shows that the asphalt model overestimated the actual surface tensions, but all of the overestimated errors were less than 10%. Simulation and experimental results indicate that the surface tension decreases linearly with increasing temperature. 60
Surface tension (mN/m)
Fig. 2 Surface tension of asphalt at different temperatures (simulation and experiment)
50
Simulation Experiment
Y=111.07-0.22X R2=0.96
40
30
20
10 280
320
360
400
Temperature (K)
440
480
1890
L. You et al.
3.3 Molecular Structural Properties The model structure (or molecular interactions) at 374.15 K (viscoelastic liquid state) was revealed via the radial distribution function (RDF) of the molecules center-ofmass. The RDF shows the probability of finding a particle (atom, molecules, etc.) at a given distance away from another particle. The RDFs of the molecules centerof-mass have been averaged according to the group of molecules they belong to, namely asphaltenes, naphthene aromatics, polar aromatics, and saturates. In the four plots of Fig. 3, each group is depicted concerning the other groups. For asphaltenes, the first peak is observed for the interactions of asphaltenes-naphthene aromatics and asphaltenes-polar aromatics around 4 Å. The first peak of asphaltenes-polar aromatics comes from the interactions of asphaltene-thiophene (C51 H62 S) and benzobisbenzothiophene (C18 H10 S2 ), while the first peak of asphaltenes-naphthene aromatics comes from the interactions between DOCHN (C30 H46 ) and asphaltene-pyrrole (C66 H81 N). Especially for the interactions of asphaltenes-polar aromatics, after the first peak, a second much higher peak appears at a higher distance around 5.2 Å, which due to the interactions between pyridinohopane (C36 H57 N) and asphaltene-thiophene (C51 H62 S). Another high peak appears for the interactions of asphaltenes-saturates at 4.5 Å, which represents the interactions between squalane (C30 H62 ) and asphaltenethiophene (C51 H62 S). Furthermore, it is interesting to note that the highest peak of the pairs of polar aromatics-naphthene aromatics is located at a distance less than 4 Å (~3 Å), which represents the interactions of thioisorenieratane (C40 H60 S) and PHPN (C35 H44 ).
4 Summary and Conclusions The surface tension of asphalt was calculated via Kirkwood-Buff and IrvingKirkwood theories in the proposed asphalt molecular dynamic model. Also, the modeling surface tension of asphalt was compared with the pendant drop method in the laboratory at various temperatures. A comparison with the experimental results shows that the asphalt model overestimated the actual surface tensions, but all of the overestimated errors were less than 10%. Respectively, the molecular structures were investigated using the radial distribution function (RDF) of the molecules center-ofmass. The first peaks and the highest peaks between groups of molecules (asphaltenes, naphthene aromatics, polar aromatics, and saturates) were shown in Fig. 3, and the atom pairs contributing the most to them were identified. These results have general agreement with references.
Understanding of Structural and Surface Tension Properties … 40
20
Polar aromatics-polar aromatics Polar aromatics-asphaltenes Polar aromatics-naphthene aromatics Polar aromatics-saturates C36H57N - C51H62S
30
g(r)
g(r)
40
Asphaltenes-asphaltenes Asphaltenes-polar aromatics Asphaltenes-naphthene aromatics Asphlatenes-saturates
30
1891
C30H46 - C66H81N
C40H60S - C35H44
20
C51H62S - C18H10S2 10
10
0
0
2
4
6
8
10
0
12
0
2
4
r (Å)
(a) Asphaltene40
8
10
12
(b) Polar aromatic40
Naphthene aromatics-naphthene aromatics Naphthene aromatics-asphaltenes Naphthene aromatics-polar aromatics Naphthene aromatics-saturates
30
6
r (Å)
Saturates-saturates Saturates-asphaltenes Saturates-polar aromatics Saturates-naphthene aromatics
30
g(r)
g(r)
C30H62 - C51H62S
C40H60S - C35H44
20
10
0
20
10
0
2
4
6
8
10
r (Å)
(c) Naphthene aromatic-
12
0
0
2
4
6
8
10
12
r (Å)
(d) Saturate-
Fig. 3 Radial distribution function (RDF) of the molecules center-of-mass grouped per category (asphaltenes, naphthene aromatics, polar aromatics, and saturates)
References 1. Li, D.D., Greenfield, M.L.: Chemical compositions of improved model asphalt systems for molecular simulations. Fuel 115, 347–356 (2014) 2. Bayly, C.I., et al.: A second generation force field for the simulation of proteins, nucleic acids, and organic molecules. J. Am. Chem. Soc. 117(19), 5179–5197 (1995) 3. Theodorou, D.N., Suter, U.W.: Shape of unperturbed linear polymers: polypropylene. Macromolecules 18(6), 1206–1214 (1985) 4. Gray, C., Gubbins, K.: Theory of surface tension for molecular fluids. Mol. Phys. 30(1), 179–192 (1975) 5. Drelich, J., Fang, C., White, C.: Measurement of interfacial tension in fluid-fluid systems. Encycl. Surface Colloid Sci. 3, 3158–3163 (2002)
Use of Modified Reclaimed Asphalt in Warm Mixtures G. Ferrotti, Francesco Canestrari, J. Xiaotian, and Fabrizio Cardone
Abstract Warm Mix Asphalt (WMA) has the advantage over conventional Hot Mix Asphalt (HMA) of reducing pollutant emissions and energy costs. The use of WMA technologies can be fruitfully coupled with the addition of Reclaimed Asphalt (RA) improving asphalt mixture performance. However, this aspect needs to be investigated in more detail. In this regard, this paper focused on the optimization and performance investigation of a dense graded modified asphalt mixture for wearing courses, produced with WMA technology and percentages of RA up to 25%. Two WMA mixtures with two contents of RA were prepared by using a chemical additive. The results were also compared with an HMA control mixture produced with a lower RA content. Shear gyratory compacted specimens were used for a series of laboratory tests to evaluate the performance. The study of stiffness, water sensitivity, permanent deformation and fracture characteristics pointed out that appropriately designed WMA mixtures with RA can be produced without penalizing the performance with respect to the control mixture. Keywords Warm mix asphalt · Reclaimed asphalt · Stiffness · Permanent deformation · Cracking
1 Introduction Because of the great both economic and environmental benefits, the use of Reclaimed Asphalt (RA) has gained popularity. However, the stiff binder released by RA aggregates, especially when modified RA is considered, restricts its maximum amount allowed in Hot Mix Asphalt (HMA) due to the risk of mixtures with a reduced workability and more prone to some pavement distresses [1]. In this sense, Warm Mix Asphalt (WMA) technology can represent an effective technical solution for avoiding the downsides of high RA content. Indeed, the reduction of production
G. Ferrotti (B) · F. Canestrari · J. Xiaotian · F. Cardone Università Politecnica delle Marche, Ancona, Italy e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_240
1893
1894
G. Ferrotti et al.
temperatures by 20–40 °C induces lower bitumen aging, thus mitigating the stiffening effect of the aged bitumen from RA and not penalizing the workability of the mixture [2]. Despite this benefit, additional concerns are related to the presence of Polymer Modified Bitumens (PMB) especially if both virgin and RA bitumens are polymer modified. Indeed, possible interactions between virgin and aged polymer chains as well as WMA additives may modify the rheological behavior of binder phase compromising the performance of the mixtures [3–5]. Given this background, the objective of this research is to evaluate the feasibility of using percentages of modified RA up to 25%, in WMA mixtures prepared with a modified bitumen. A laboratory mechanical characterization, also carried out on a conventional recycled HMA, allowed the potentialities of WMA technology combined to recycling techniques to be evaluated.
2 Experimental Program Virgin (limestone, basalt, calcareous filler) and modified RA aggregates were used to produce the aggregate blends. The bitumen content of RA (Dmax = 5 mm) was 8.1% by aggregate weight and the recovered bitumen had a penetration of 21 dmm and a softening point of 81 °C. An SBS polymer modified bitumen classified as PmB 45/80-70 was selected as the binder. A chemical WMA additive available on the market, having an alkyl-amide polyamines-based composition and characterized by a density at 25 °C of 0.94 g/cm3 , a flash point higher than 150 °C and a viscosity at 25 °C of 150 cP, was chosen. Three dense-graded bituminous mixtures for wearing course were investigated: two WMAs and one HMA mixture as reference. The grading curves of the three asphalt mixtures were different and designed according to the Italian technical specifications for road construction [6]. The mixtures were coded as: WMA-20-3.7, containing 20% RA, 3.7% added binder and 5.1% total binder; WMA-25-3.7, containing 25% RA, 3.7% added binder and 5.5% total binder; HMA-15-4.6, containing 15% RA, 4.6% added binder and 5.6% total binder. The WMA mixtures were mixed and compacted at 130 °C and 120 °C respectively, whereas the HMA mixture were mixed and compacted at 170 °C and 160 °C, respectively. For each mixture a series of cylindrical specimens (D = 100 mm) were prepared by means of a gyratory compactor to the same target air voids content (equal to 5%). Specimens were then subjected to a mechanical characterization in terms of Indirect Tensile Strength (ITS), water sensitivity (Indirect Tensile Strength Ratio—ITSR), stiffness (Dynamic Modulus |E*| and Indirect Tensile Stiffness Modulus—ITSM), fatigue (Indirect Tensile Fatigue Test—ITFT) and rutting resistance.
Use of Modified Reclaimed Asphalt in Warm Mixtures
1895
Fig. 1 ITS and ITSR results
3 Results and Discussions 3.1 Indirect Tensile Strength (ITS) Test and Water Sensitivity ITS test was performed at 25 °C, according to EN 12697-23. Six replicates were performed for each mixture, three in dry condition and three in wet condition, where the conditioning was performed according to EN 12697-12 (Method A). The Indirect Tensile Strength Ratio (ITSR) value was calculated as ratio between the ITS values in wet and dry conditions. Figure 1 shows the average ITS and ITSR values for all the investigated mixtures. In both dry and wet conditions, the ITS values of WMA mixtures are higher than the reference HMA and satisfied the local technical specification requirements (0.95 < ITS < 1.70). However, WMAs show a slightly lower moisture resistance, especially for higher RA contents.
3.2 Stiffness Dynamic complex modulus was measured to evaluate the stiffness of mixtures, at 4 temperatures (from 5 to 50 °C) and 6 frequencies (from 0.1 to 20 Hz) by means of Asphalt Mixture Performance Tester (AMPT). Two replicates were performed for each mixture. The Huet–Sayegh (HS) model [7] was used to fit the frequency dependency of the dynamic modulus |E*| and the Williams-Landel-Ferry (WLF) equation [8] was used to model the shift factor trend with temperature. The dynamic modulus master curves of all mixtures at Tref = 20 °C are shown in Fig. 2a and the parameters of HS (E g , E e , k, h, δ, t) and WLF (C1 and C2) models are listed in Table 1. It can be observed that both WMA mixtures showed lower dynamic moduli with respect to the reference HMA, over the entire range of loading frequency. This is mainly imputable to the less oxidation effects that WMA mixtures experienced during the lower temperature production process. Therefore, compared to the reference HMA
1896
G. Ferrotti et al.
Table 1 HS and WLF model parameters HMA-15-4.6
E g (MPa)
E e (MPa)
k (−)
h (−)
δ (−)
t (s)
C1 (−)
C2 (°C)
32182
50
0.138
0.389
2.002
0.822
22.9
166.5
WMA-20-3.7
28015
109
0.184
0.503
2.405
0.322
18.3
138.5
WMA-25-3.7
25109
77
0.192
0.513
2.52
0.36
15.4
117.1
a)
b)
Fig. 2 Dynamic modulus master curves of mixtures at Tref = 20 °C (a), ITSM versus |E*| @ 2 Hz at 20 °C (b)
mixture, the WMAs showed a significantly reduced stiffening effect mainly due to aging. Moreover, both WMA mixtures showed a similar trend for the master curves, highlighting the prevalent role of aging rather than RA content on the development of stiffness properties. Stiffness properties were also measured by the ITSM at 20 °C, according to EN 12697-26. Figure 2b reports the comparison between the ITSM values and the dynamic moduli identified in the master curves at the same temperature (20 °C) and frequency (2 Hz–comparable with the rise time of 124 ms) of the ITSM tests. Figure 2b shows that ITSM results are consistent with the dynamic modulus data obtained through uniaxial compression test, confirming lower stiffness of both WMA mixtures compared to the reference mixture.
3.3 Rutting Performance Rutting resistance of mixtures was determined by performing cyclic compression tests at 60 °C, according to EN 12697-25. Two replicates were performed for each mixture. A cyclic axial stress with an amplitude of 0.3 MPa and a steady confining pressure of 0.2 MPa were adopted during the entire test (10,000 load cycles). The cumulative axial strain of the specimen was recorded and the creep curve was built. The creep rate fc (i.e. slope of the least square linear fit of the quasi linear part of the creep curve) and the permanent deformation after 1000 load cycles ε1000
Use of Modified Reclaimed Asphalt in Warm Mixtures Table 2 The rutting parameter results
1897
Mixture
fc (μstrain/cycle)
ε1000 (%)
HMA-15-4.6
0.081
0.834
WMA-20-3.7
0.044
0.53
WMA-25-3.7
0.048
0.423
(i.e. deformation calculated from the least square power fit of the quasi linear part at 1000th cycle) were used to characterize the rutting potential of the mixtures. Specifically, the lower the value of these parameters, the higher the rutting resistance of the mixture. The average values of the rutting parameters of WMA mixtures, shown in Table 2, are remarkably lower than those of the reference mixture, indicating that WMA mixtures exhibited higher rutting resistance. These results highlight that despite the mitigated stiffness due to the reduced production temperature, the WMA mixtures would not experience any permanent deformation accumulation risk in the early period of their service life.
3.4 Fatigue Performance ITFT was carried out at 20 °C according to the British standard BS DD ABF (1997). During the test, the number of loading cycles was recorded until the complete failure of the specimen was obtained. Three initial strain levels (100, 125 and 150 με) were selected for each mixture and the corresponding stresses to be applied were calculated by using the ITSM values (Fig. 2b) measured before the fatigue tests. Two specimens were tested for each strain level. The ITFT results are shown in Fig. 3, where the initial horizontal strain εh,initial is plotted as a function of the number of loading cycles to failure Nf . Test data were analyzed according to a power law regression and the fatigue curve Fig. 3 Fatigue curve of mixtures at 20 °C
1898
G. Ferrotti et al.
2 (i.e. εh = a·N−b f ) was reported for each mixture. The obtained high R values (> 0.95) for all mixtures indicate good reliability to predict fatigue life by using this fitting model. From the analysis of fatigue curves, it can be noted that, at high and medium strain levels, WMA mixtures can withstand higher numbers of loading cycles than the reference one, highlighting a better fatigue behavior for WMA mixtures; whereas, at low strain levels, comparable behaviour is expected between WMAs and HMA. These results are consistent with the stiffness analysis, which suggests that the higher stiffness of the reference mixture could result in a more brittle tendency. Moreover, the comparison between WMA mixtures shows that a higher RA content results in a slightly higher fatigue resistance, especially at higher strain level.
4 Conclusions This paper focused on the performance analysis of dense graded recycled WMA mixtures for wearing courses produced with different modified RA amounts (20 and 25%) and an SBS modified bitumen. The mixtures were analyzed in terms of strength and stiffness properties, moisture susceptibility, rutting and fatigue behavior by testing gyratory compacted specimens. Compared to a conventional recycled HMA, the investigated WMA mixtures showed: • higher strength properties in both dry and wet conditions but a higher moisture susceptibility that tends to increase with the increase in RA content; • lower stiffness properties and higher fatigue resistance. This finding highlights that lower production temperatures (i.e. WMA technologies) reduce short-term aging effects leading to stiffness mitigation as well as improved fatigue performance. The increase in RA content did not affect the stiffness of WMAs but allowed the achievement of higher fatigue resistance; • significantly higher rutting resistance despite lower stiffness properties. Besides, no RA content effect on the rutting behavior was observed. Overall, the outcomes of this experimental investigation strongly supported the practical use of percentages of modified RA, up to 25%, in WMA mixtures prepared with a modified bitumen. Acknowledgements This research was sponsored by ASFALTI s.r.l. This support is greatly acknowledged. Data analysis and opinions are those of the Authors and do not necessarily reflect those of the sponsoring agency.
Use of Modified Reclaimed Asphalt in Warm Mixtures
1899
References 1. Capitão, S.D., Picado-Santos, L.G., Martinho, F.: Pavement engineering materials: review on the use of warm-mix asphalt. Constr. Build. Mater. 36, 1016–1024 (2012) 2. Zaumanis, M., Mallick, R.B.: Review of very high-content reclaimed asphalt use in plantproduced pavements: state of the art. Int. J. Pavement Eng. 16(1), 39–55 (2015) 3. Stimilli, A., Virgili, A., Canestrari, F.: Warm recycling of flexible pavements: Effectiveness of Warm Mix Asphalt additives on modified bitumen and mixture performance. J. Clean. Prod. 156, 911–922 (2017) 4. Stimilli, A., Frigio, F., Sciolette, S., Canestrari, F.: In-plant production of warm recycled mixtures produced with SBS modified bitumen: A case study. In Proceedings of the AIIT International Congress on Transport Infrastructure and Systems, TIS 2017, pp. 143–151. Taylor and Francis Group, London (2017) 5. Ragni, D., Ferrotti, G., Lu, X., Canestrari, F.: Effect of temperature and chemical additives on the short-term ageing of polymer modified bitumen for WMA. Mater. Des. 160, 514–526 (2018) 6. ANAS S.p.a.: Capitolato Speciale di Appalto – Norme Tecniche: Pavimentazioni stradali/aeroportuali. Rev. 24/04/2009, Ente Nazionale delle Strade (2009) 7. Sayegh, G.: Viscoelastic properties of bituminous mixtures. In Proceedings of the 2nd International Conference on Structural Design of Asphalt Pavements, Washington, D.C. (1967) 8. Ferry, J.D.: Viscoelastic Properties of Polymers. Wiley, New York (1961)
Use of Multiple Stress Creep Recovery Test to Evaluate Viscoelastic Properties of Asphalt Mastic Song Li, Ghim Ping Ong, and Fujian Ni
Abstract Asphalt mastic is regarded as a bonding agent within the asphalt mixture and has a significant impact on asphalt mixture performance. Multiple stress creep recovery (MSCR) test are performed in this study to evaluate the effect of filler on high-temperature performance of asphalt mastic. In addition, the creep-recovery response of asphalt mastic with a wide range of creep stress levels was modeled using the Burgers model. It was found from the non-recovery compliance and percentage recovery results obtained from the MSCR test that the addition of filler can significantly improve the elastic and recoverable response of the asphalt mastic, and hence enhances the rutting potential of asphalt mastic. The Burgers model parameters including E m , ηm , E k and ηk were obtained for asphalt mastic with different filler contents and creep stresses. Through correlating these parameters and MSCR test indicators, it was found that the parameter ηm has the highest correlation to the MSCR test results. Keywords Asphalt mastic · Viscoelastic response · Burgers model · Multiple stress creep recovery (MSCR) test
1 Introduction Asphalt mixture is considered to be a composite mixture consisting of mastic-coated aggregates rather than binder-coated aggregates [1]. Asphalt mastic, consisting of asphalt binder and filler particles, is hence the bonding agent within the asphalt mixture [2]. As such, it is necessary to understand the performance of the asphalt mastic and how its viscoelasticity can be affected by mineral filler.
S. Li · F. Ni School of Transportation, Southeast University, Nanjing, China G. P. Ong (B) Department of Civil & Environmental Engineering, National University of Singapore, Singapore, Singapore e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_241
1901
1902
S. Li et al.
The evaluation of rutting potential of asphalt mastic on high temperature is similar with the asphalt binder in laboratory. A complex modulus (G* ) and phase angle (δ) in a linear viscoelasticity manner can be examined [3]. A rutting parameter, denoted as G* /sinδ, is often used to evaluate the rutting performance of asphalt binder and mastic [4]. Zero shear viscosity (ZSV) is also used to estimate rutting potential of asphalt binder and mastic [5]. Besides G* /sinδ and ZSV, the viscous component of creep stiffness Gv of asphalt binder and mastic could serve as an indicator of rutting resistance through statistical fitting [6]. The multiple stress creep recovery (MSCR) test is adopted by recent researchers to study asphalt mastics as it considers the nonlinear viscoelasticity of asphalt binder and mastic, and has high correlation with rutting resistance of asphalt mixture [7, 8]. It is also known that the addition of mineral filler in asphalt mastic leads to reduced susceptibility to permanent deformation with an enhanced elastic response [9]. In addition, the stiffness improvement of the asphalt mastic is a function of the filler content. Rheological models can help better understand the viscoelastic behavior of the asphalt mastic and several rheological models such as the Burgers model, the six-element solid model, the generalized Maxwell model, the generalized Kelvin-Voigt model and the power law model have been proposed in the literature [10, 11]. The Burgers model, for example, models the elastic, viscoelastic and viscous flow properties of the materials effectively, and can better explain the creep and recovery responses of asphalt binder as compared to other models. This paper therefore attempts to investigate the viscoelastic performance of asphalt mastic with different filler contents using the MSCR test. The key parameters for the Burgers model are derived based on the creep-recovery response of the asphalt mastic from the MSCR test. The viscoelastic properties development of the mastic with different contents of mineral filler are then quantified. The relationship between indicators derived from MSCR test and Burgers model parameters are also examined.
2 Material Preparation Two asphalt binders and one mineral filler were selected to prepare the asphalt mastic. The binders were both styrene–butadiene–styrene (SBS) modified asphalt binders. SBS contents were 3.5% and 4.0% of asphalt binder by mass, respectively. The hightemperature performance grading of these two asphalt binders were PG70 and PG 76, which were used to differentiate the two asphalt binders. The selected mineral filler was limestone with a specific gravity of 2.705 g/cm3 . The filler was comprised of particles passing through a 0.075 mm (#200) sieve. Asphalt mastic samples, in this study, were prepared by blending of the asphalt binder and mineral filler with filler content, which was a filler-to-binder ratio in the asphalt mastic by mass [12]. A wide range of filler content was selected to fabricate the mastic sample in this study, namely 0, 50, 80, 100, 120, 150, and 180%. For the MSCR test, the blended asphalt mastic needs to be molded into a specimen with 20 mm in diameter and 3–4 mm in thickness by a silicone mold.
Use of Multiple Stress Creep Recovery Test to Evaluate Viscoelastic Properties …
1903
3 Experiment Procedures and Rheological Model for Mastic Asphalt 3.1 Multiple Stress Creep Recovery Test The multiple stress creep recovery (MSCR) test was performed using the Dynamic Shear Rheometer (DSR) operating in constant stress mode. Following the AASHTO T350 standard [13], the test was performed at 64 °C using parallel plates geometry with a plate diameter of 25.0 mm and a gap of 1.0 mm. In this study, the final test was performed with 1 s of creep loading and 9 s of recovery at stresses of 0.1, 3.2, 6.4 and 12.8 kPa with 10 cycles per stress level. Two indicators, namely non-recovery compliance (J nr ) and percentage recovery (%R), are obtained from the MSCR test.
3.2 Burgers Model Burgers model is a combination of the Maxwell model and the Kelvin model in series. It consists of four mechanical components: two spring elements (E m , E k ) which comply with the Hookean principle and the two dashpots (ηm , ηk ) which follow the Newtonian principle. The subscripts m and k indicate the Maxwell and Kelvin elements, respectively. Under constant stress, the creep and recovery response of Burgers model can be expressed as Eq. (1) [10]. ⎧ E σ0 σ0 σ0 − kt ⎪ ⎪ 1 − e ηk + t+ (t < t1 ) ⎨ E ηm Ek ε(t) = σ m Ek E σ0 0 ⎪ − kt − (t−t1 ) ⎪ 1 − e ηk 1 e ηk t1 + (t ≥ t1 ) ⎩ ηm Ek
(1)
where σ 0 is the creep stress; t is the time of test; and t 1 is the time when the load is removed. Recovery response occurs when t > t 1 .
4 Experimental Results and Analysis 4.1 Results of MSCR Test for Asphalt Mastics with Different Filler Contents Figure 1 presents the results of J nr and %R for asphalt mastics for different filler contents and binder types. J nr and %R are found to be highly dependent on filler content. The former tends to decrease with increasing filler content while the latter
1904
S. Li et al. PG70-0.1kPa PG70-3.2kPa PG70-6.4kPa PG70-12.8kPa PG76-0.1kPa PG76-3.2kPa PG76-6.4kPa PG76-12.8kPa
Jnr (1/kPa)
0.8
0.6
Notation indicates: asphalt binderstress level
0.4
0.2
0.0
120 100 80
%R (%)
1.0
60 40 20
0
50
80 100 120
150
Filler content (%)
(a) Jnr vs. filler content
180
0
Notation indicates: asphalt binder-stress level 0
50
PG70-0.1kPa PG70-3.2kPa PG70-6.4kPa PG70-12.8kPa PG76-0.1kPa PG76-3.2kPa PG76-6.4kPa PG76-12.8kPa
80 100 120
150
180
Filler content (%)
(b) %R vs. filler content
Fig. 1 MSCR test results for asphalt mastics with different filler contents
increases with increasing filler content. This implies that the addition of filler improves recovery property and elastic response of asphalt mastic. Figure 1 also indicates that the improvements of J nr and %R are more significant at higher stress levels and at lower filler contents. Moreover, it can be seen from Fig. 1 that mastic with PG76 asphalt binder exhibited better elastic behavior and higher stiffness than that with PG70 binder.
4.2 Results and Analysis on Model Fit Using the Burgers Model The fifth cycle of the ten cycles of the MSCR test for a given stress level was chosen as a representative creep and recovery response for the asphalt mastic and was used to model the viscoelastic behavior by Burgers model. By performing multiple nonlinear regression, the four parameters of the Burgers model (i.e. E m , ηm , E k and ηk ) of asphalt mastic at different filler contents and stress levels were obtained and are presented in Fig. 2. It can be observed from the figure that E m , ηm , E k and ηk increase with increasing filler content until a filler content of around 150% where further addition of filler tends to reduce the creep and recovery moduli and viscosities. It is probable that the viscoelastic property of the asphalt mastic can no longer be improved when the filler increases to an excessively high content. In addition, the rates of increase in E m , ηm , E k and ηk with respect to filler content decrease with an increase in stress levels. This implies that the addition of filler results in better improvement at lower stress levels. Moreover, asphalt mastic using PG76 binder provides better performance as compared to that of PG70 binder.
Use of Multiple Stress Creep Recovery Test to Evaluate Viscoelastic Properties … 350000
Em-0.1kPa Em-3.2kPa Em-6.4kPa Em-12.8kPa Ek-0.1kPa Ek-3.2kPa Ek-6.4kPa Ek-12.8kPa
100000 80000 60000
Notation indicates: Elastic modulus of M axwell unit (Em) or Kelvin unit (Ek)stress level
40000 20000
250000 200000
Notation indicates: Viscosity of M axwell unit ( m) or Kelvin unit ( k)stress level
150000 100000 50000
0
0
0
50
80 100 120
150
180
0
120000
60000
Notation indicates: Elastic modulus of M axwell unit (Em) or Kelvin unit (Ek)stress level
40000 20000
150
180
ηm-0.1kPa ηm-3.2kPa ηm-6.4kPa ηm-12.8kPa ηk-0.1kPa ηk-3.2kPa ηk-6.4kPa ηk-12.8kPa
300000
Viscosity (Pa.s)
80000
80 100 120
350000
Em-0.1kPa Em-3.2kPa Em-6.4kPa Em-12.8kPa Ek-0.1kPa Ek-3.2kPa Ek-6.4kPa Ek-12.8kPa
100000
50
Filler content (%) (b) Viscosity of mastic with PG76
Filler content (%) (a) Elastic modulus of mastic with PG76
Elastic modulus (Pa)
ηm-0.1kPa ηm-3.2kPa ηm-6.4kPa ηm-12.8kPa ηk-0.1kPa ηk-3.2kPa ηk-6.4kPa ηk-12.8kPa
300000
Viscosity (Pa.s)
Elastic modulus (Pa)
120000
1905
250000 200000 150000
Notation indicates: Viscosity of M axwell unit ( m) or Kelvin unit ( k)stress level
100000 50000 0
0 0
50
80 100 120
150
180
Filler content (%) (c) Elastic modulus of mastic with PG70
0
50
80 100 120
150
180
Filler content (%)
(d) Viscosity of mastic with PG70
Fig. 2 Effect of filler content and stress level on E m , ηm , E k and ηk within Burgers model for asphalt mastic with different binder performance grades
4.3 Analysis on Correlation of MSCR Test Observations and Burgers Model Parameters The correlations of the MSCR test observations (J nr and %R) and the Burgers model parameters including E m , ηm , E k and ηk were evaluated by using the Pearson correlation coefficient (PCC). The results are presented in Fig. 3. A PCC value close to 1.0 is indicative of a high correlation between two variables of interest. It was found that J nr presents a negative correlation with the Burgers model parameters while %R indicates a positive correlation. The parameters of the Maxwell element within the Burgers model showed a higher correlation than that of the Kelvin element. Moreover, ηm showed the highest correlation with the MSCR test indicators and hence is an important parameter for the Burgers model in terms of viscoelastic performance.
1906
S. Li et al.
Fig. 3 PPC values between rheological indicators from MSCR test and Burgers model parameters
1.0 %R vs. Burgers parameters
0.8 0.5
Jnr vs. Burgers parameters
0.57 0.43
0.34
PCC
0.3
0.15
0.0 Em
ηm
Ek
ηk
-0.3 -0.5
-0.42
-0.36
-0.35
-0.47
-0.8 -1.0
5 Conclusion Based on the results of J nr and %R obtained from the MSCR test, the addition of filler can improve the elastic and recoverable responses of asphalt mastic, and hence enhances the rutting resistance of asphalt mastic. The parameters of Burgers model, E m , ηm , E k and ηk , increase with increasing filler content, decreasing creep stress levels and binder with higher performance grade. The effect of filler on the improvement of these parameters is more observable at lower stress levels. In addition, the Burgers model parameters showed a negative correlation with J nr and a positive correlation with %R, and ηm has the highest correlation to the observations found in the MSCR test.
References 1. Riccardi, C., Cannone Falchetto, A., Losa, M., Wistuba, M.P.: Development of simple relationship between asphalt binder and mastic based on rheological tests. Road Mater. Pavement Des. 19(1) 18–35 (2016) 2. Zeng, M., Wu, C.: Effects of type and content of mineral filler on viscosity of asphalt mastic and mixing and compaction temperatures of asphalt mixture. Transp. Res. Record: J. Transp. Res. Board 2051(1), 31–40 (2008) 3. Cosme, R.L., Teixeira, J.E.S.L., Calmon, J.L.: Use of frequency sweep and MSCR tests to characterize asphalt mastics containing ornamental stone residues and LD steel slag. Constr. Build. Mater. 122(2016), 556–566 (2016) 4. Reddy, C.G., Babitha, K.N., Biligiri, K.P.: Binder rheological relationships depicting wheel tracking rutting parameter for asphalt mixtures: a case study. J. Test. Eval. 45(5), 20160259 (2017) 5. Biro, S., Gandhi, T., Amirkhanian, S.: Determination of zero shear viscosity of warm asphalt binders. Constr. Build. Mater. 23(5), 2080–2086 (2009)
Use of Multiple Stress Creep Recovery Test to Evaluate Viscoelastic Properties …
1907
6. Delgadillo, R., Nam, K., Bahia, H.: Why do we need to change G*/sinδ and how? Road Mat. Pavement Des. 7(1), 7–27 (2006) 7. D’Angelo, J.A.: The Relationship of the MSCR test to rutting. Road Mat. Pavement Des. 10(sup1), 61–80 (2011) 8. White, G.: Grading highly modified binders by multiple stress creep recovery. Road Mat. Pavement Des. 18(6), 1322–1337 (2016) 9. Yin, H.M., Buttlar, W.G., Paulino, G.H., Benedetto, H.D.: Assessment of existing micromechanical models for asphalt mastics considering viscoelastic effects. Road Mat. Pavement Des. 9(1), 31–57 (2011) 10. Liu, Y., You, Z.: Determining burger’s model parameters of asphalt materials using creeprecovery testing data. In: Pavements and Materials: Modeling, Testing, and Performance, pp. 26–36 (2009) 11. Zhao, Y., Ni, Y., Zeng, W.: A consistent approach for characterising asphalt concrete based on generalised Maxwell or Kelvin model. Road Mater. Pavement Des. 15(3), 674–690 (2014) 12. Li, S., Ni, F., Dong, Q., Zhao, Z., Ma, X.: Effect of filler in asphalt mastic on rheological behaviour and susceptibility to rutting. Int. J. Pavement Eng. (2019) (in press) 13. AASHTO, T350–14(2018): Standard Method of Test for Multiple Stress Creep Recovery (MSCR) Test of Asphalt Binder Using a Dynamic Shear Rheometer (DSR) (2018)
Use of Recycled Asphalt Pavement (RAP) in Airport Pavements Navneet Garg, Hasan Kazmee, and Lia Ricalde
Abstract Federal Aviation Administration (FAA) currently does not allow use of recycled asphalt pavement (RAP) on Airport Improvement Program (AIP) funded airport runway and taxiway pavement projects, and is allowed for use only in pavement shoulders. FAA’s National Airport Pavement and Materials Research Center (NAPMRC) was established to evaluate performance of new and sustainable asphalt material technologies (such as WMA, RAP, etc.) under heavy aircraft loading at high pavement temperatures. As part of Test Cycle 2 (TC2), six test lanes were constructed – four outdoors and two indoors, each encompassing three different test sections. In two indoor lanes, RAP was added to the warm mix asphalt (WMA). Lane-1 is the control section with FAA standard P401 specification hot mix asphalt (HMA). Heavy weight deflectometer (HWD) tests were performed on the constructed test lanes to characterize the pavements. Extensive laboratory tests are planned on RAP/WMA and HMA (field cores and loose mixes). The test lanes will be subjected to accelerated pavement tests (APT) using custom designed airport heavy vehicle simulator (HVS-A) to study rutting performance (at high pavement temperatures) and fatigue behavior. This paper presents construction of test lanes, asphalt mix designs (with and without RAP), results from HWD tests on test lanes with RAP/WMA, and results from laboratory tests on pavement materials.. Keywords Airport pavements · HMA · WMA · RAP · Airport heavy vehicle simulator · HWD
N. Garg (B) Federal Aviation Administration, Airport Technology R&D Branch, William J. Hughes Technical Center, Atlantic City, NJ 08405, USA e-mail: [email protected] H. Kazmee · L. Ricalde Applied Research Associates, Inc., 2628 Fire Rd, Suite 300, Egg Harbor Township, NJ 08234, USA © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 H. Di Benedetto et al. (eds.), Proceedings of the RILEM International Symposium on Bituminous Materials, RILEM Bookseries 27, https://doi.org/10.1007/978-3-030-46455-4_242
1909
1910
N. Garg et al.
1 Introduction Airports have to comply with the Federal Aviation Administration (FAA) advisory circulars (AC) to design and construct airfield pavements to be eligible for funding granted by Airport Improvement Program (AIP). The current version of FAA’s construction specification designated as FAA AC 150/5370-10H does not allow the use of recycled materials in surface courses of airfield pavements (except for shoulder pavement). It also does not provide any guideline for warm mix asphalt (WMA) in airfield paving. Use of RAP in WMA can be beneficial since this type of asphalt concrete is produced at a lower temperature and thus, can contain the secondary aging of RAP to a minimal level [1]. In United States, WMA has seen very limited used on airports, mainly taxiways [2]. Acknowledging these lack of guidance, FAA’s 10-year research and development program opted to study several warm mix technologies and RAP extensively. In line with that effort, Garg et al. investigated the rutting performance of a warm mix treated with chemical additive during the initial test cycle (TC-1) at the National Airport Pavement and Materials Research Center (NAPMRC) [3]. An airport heavy vehicle simulator was procured for this facility to conduct the full-scale tests at controlled temperature. In TC-2, several other WMA technologies are being investigated, including RAP incorporated WMA. Traffic tests will be conducted for rutting and fatigue characterization. This paper discusses the WMA-RAP mix design, construction of the test lanes, pavement and laboratory characterization effort supplementary to the traffic tests.
2 Design and Construction of Full-Scale Test Lanes For TC2, a total of six test lanes were constructed with different asphalt concrete mixes. This paper summarizes job mix formula, pavement characterization and material characterization results of only three mixes (subject of this paper). Fig. 1 shows the layout and design of TC2 test lanes discussed in this paper. Lane 1 was constructed outdoor whereas, lanes 5 and 6 were constructed indoor. Lanes 1 and 6 were paved with a PG76-22 control hot mix and neat binder (PG64-22) warm mix treated with organic additive, latex, lime and reclaimed asphalt pavement (RAP) materials, respectively. The same RAP mix was also used to pave the bottom 15.24 cm of lane 5; whereas, the top 7.62 cm (surface lift) was surfaced with the same PG64-22 mix without the RAP. Mixes in lane 1 and 6 as well as the surface lift of lane 5 are designated as mixes A, E and F, respectively. The existing 12.7 cm thick asphalt concrete from TC1 was milled along with 10.16 cm of P-209 crushed stone base course. Finally, the P-401 type asphalt concrete was laid in three 7.62 cm thick lifts. Note that P-401, P-209 and P-154 represent the FAA designations for HMA, crushed stone base and subbase, respectively. The pavement layers were designed on the premise that the rutting failure would be limited to the asphalt concrete only. Each of the test lanes were divided in three representative test sections known as North,
Use of Recycled Asphalt Pavement (RAP) in Airport Pavements
1911
Fig. 1 Layout and design of TC2 test lanes (1 in. = 25.4 mm; 1 psi = 6.89 kPa)
Center and South sections. The job mix formulas (JMF) for TC2 mixes were developed with 75 gyrations following the FAA AC 150/5370-10H. Previous literature, industry experts and manufacturer’s instructions indicated that addition of warm mix additive should not change the optimum asphalt content [4]. Table 1 summarizes the JMF properties of the TC2 asphalt concrete mixes discussed in this paper. Asphalt pavement analyzer (APA) tests were conducted on plant produced lab compacted specimens prepared for quality assurance tests. The polymer modified control mix outperformed the neat binder mixes in terms of accumulated rut depth. Presence of RAP in the latex mix induced better rutting resistance according to the test results. Since RAP is a hydrogen bound material, a slightly higher temperature is recommended for the production of warm mix to avoid moisture related damages [5]. The same guideline was followed for the production of the RAP mix.
2.1 Sequence of Asphalt Concrete Paving Table 2 highlights the sequence of asphalt concrete paving in lanes 1, 5 and 6 along with the quality control (QC) and assurance (QA) test results. The QC tests were performed on plant compacted samples while QA verification tests were conducted on AC cores collected from the constructed test lanes. Similar to the JMF, the asphalt content in the bottom and middle lifts of lanes 5 and 6 were slightly higher than that
1912
N. Garg et al.
Table 1 Summary of job mix formula requirements for TC2 mixes Requirement
Criteria
Mix designation A—HMA PG76-22 (Control)
E—WMA PG64-22 (Organic) + Latex + RAP + Lime
F - WMA PG64-22 (Organic) + Latex + Lime
Test lanes
1
5 (Bottom 6 in.) and 6
5 (Top 3 in.)
Asphalt content
5.0%
5.5%
Warm mix additive dosage
N/A
1.5%
No. of gyrations
75
Lab mixing temp. (°F)
315–325
296–323
297–307
Lab compaction temp. (°F)
306–315
289–298
288–297
Air voids (%)
3.5%
3.4%
Tensile strength ratio
Minimum 80% at 70–80% saturation
92.5%
89.8%
92.3%
APA rut depth (AASHTO T340) (mm)