Third RILEM International Conference on Concrete and Digital Fabrication: Digital Concrete 2022 (RILEM Bookseries, 37) 3031061152, 9783031061158

Citation preview

RILEM Bookseries

Richard Buswell Ana Blanco Sergio Cavalaro Peter Kinnell   Editors

Third RILEM International Conference on Concrete and Digital Fabrication Digital Concrete 2022

Third RILEM International Conference on Concrete and Digital Fabrication

RILEM BOOKSERIES

Volume 37 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

Richard Buswell Ana Blanco Sergio Cavalaro Peter Kinnell •

•

•

Editors

Third RILEM International Conference on Concrete and Digital Fabrication Digital Concrete 2022

123

Editors Richard Buswell School of Architecture, Building and Civil Engineering Loughborough University Loughborough, Leicestershire, UK

Ana Blanco School of Architecture, Building and Civil Engineering Loughborough University Loughborough, Leicestershire, UK

Sergio Cavalaro School of Architecture, Building and Civil Engineering Loughborough University Loughborough, Leicestershire, UK

Peter Kinnell School of Mechanical, Electrical and Manufacturing Engineering Loughborough University Loughborough, Leicestershire, UK

ISSN 2211-0844 ISSN 2211-0852 (electronic) RILEM Bookseries ISBN 978-3-031-06115-8 ISBN 978-3-031-06116-5 (eBook) https://doi.org/10.1007/978-3-031-06116-5 © RILEM 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 RILEM Digital Concrete conference series was inaugurated at ETH Zurich, Switzerland, in 2018, in response to the growing interest of digital manufacturing technologies and in particular, those based on large-scale additive manufacturing. Since the first exploration of these technologies in the 2000s, there has been an exponential growth of commercial and academic activity and in 2020, TU Eindhoven, Netherlands, hosted the second in the series which was online due to the COVID-19 pandemic. The conference has become the focus for materials and process led research, with historically high interest in extrusion-based additive manufacturing, but always with solid representation of alternative fabrication methods and techniques. Now in 2022, the third conference is hosted by Loughborough University, UK, between 27 June and 29 inclusive. This is the first conference to be held in person for four years, and the response to the call for contributions was high with over 240 abstracts submitted, from 101 organisations in 27 countries. These translated into 202 full papers and extended abstracts being submitted for full peer review. These proceedings contain a selection from the best full papers submitted to the conference arranged in topic sets that reflect the parallel oral presentation sessions. Work relating to the wet material used in extrusion technologies is presented in three sessions: one on wet material property control and two on printability and set control. The progression over the three conferences is the shift from understanding the fundamentals, towards quality control and testing: at the centre of the current RILEM technical committee on performance requirements and testing of fresh printable cement-based materials (PFC). Binders and aggregates are also topical where the body of work is developing rapidly. These are split into three sessions on: aggregates, strain hardening materials and alternative binders, aligning with the parallel RILEM technical committee on Assessment of Additively Manufactured Concrete Materials and Structures (ADC). The performance of the materials and creating structural elements in particular is significantly represented, mirroring the next steps for the production technology to deliver in real applications.

v

vi

Preface

Reinforcement, structures and hardened properties have total of eight sessions: three on reinforcement, one on structural design and optimisation, one on durability and two on heterogeneities and defects. These topics all lay pathways to enable more competent design for materials that are produced through additive manufacturing. In addition, other digital fabrication approaches are explored in three sessions on: material jetting, particle bed binding and alternative processes, showcasing the continued innovation in this exciting area of research. Finally, the digital aspects of design and process control are key to a successful future for digital fabrication, and two sessions presented the latest innovations in design and digital workflow and process control, toolpath and inspection. The organising team offers a warm welcome to all delegates to the conference. To have the opportunity to invite colleagues to share in this event at Loughborough where we started our journey in 3D concrete printing 18 years ago is a real honour. To celebrate, the conference showcases these articles, alongside poster presentations and other oral presentations, framed by 12 invited and keynote speakers—all leading lights in the field. It would not be possible without the tremendous effort of the scientific committee, whose review task was huge. Everyone responded rapidly to requests, as did the contributing authors: to everyone—a big thank you. Finally, we acknowledge and give thanks our sponsors, whose support is vital to maintaining the quality of these events. These were at the time of writing: COBOD, in partnership with the NEXCON project and SIKA (Platinum), Synthomer (Silver), Elkem (Bronze). We would also gratefully acknowledge support from our workshop sponsors HAL Robotics, the enabling support of UK Research and Innovation and our partners, the Institute of Concrete Technology. We hope you enjoy the proceedings and conference. June 2022

Richard Buswell Ana Blanco Sergio Cavalaro Peter Kinnell

Organisation

Conference Chair Richard Buswell

Loughborough University, UK

International Scientific Committee Chair of the Committee Nicolas Roussel

Gustave Eiffel University, France

Co-chairs of the Committee Freek Bos Viktor Mechtcherine Dirk Lowke

Eindhoven University of Technology, The Netherlands Dresden University of Technology, Germany Technical University of Braunschweig, Germany

Members of the Committee Abdelhak Kaci Aileen Vandenberg Alexandre Pierre Ali Kazemian Arnaud Perrot Asko Fromm Behzad Nematollahi Branko Šavija Claudiane Ouellet-Plamondon Costantino Menna

CY Cergy Paris University, France Technical University of Braunschweig, Germany CY Cergy Paris University, France Louisiana State University, USA Southern Brittany University, France Hochschule Wismar University of Applied Sciences, Germany The University of Sheffield, UK Delft University of Technology, The Netherlands University of Quebec, Canada University of Naples Federico II, Italy

vii

viii

Dietmar Stephan Domenico Asprone Emmanuel Keita Geert De Schutter Gideon van Zijl Harald Kloft Jacques Kruger Jaime Mata Falcon Jay Sanjayan Jolien Van Der Putten Kim Van Tittelboom Ksenija Vasilic Liberato Ferrara Manu Santhanam Mohammed Sonebi Rob Wolfs Robert Flatt Richard Buswell Sandra Nunes Steffen Mueller Timothy Wangler Vítor Cunha Wilson Ricardo Leal da Silva Xiangming Zhou Yamei Zhang

Organisation

Technical University of Berlin, Germany University of Naples Federico II, Italy Gustave Eiffel University, France Ghent University, Belgium Stellenbosch University, South Africa Technical University of Braunschweig, Germany Stellenbosch University, South Africa ETH Zurich, Switzerland Swinburne University of Technology, Australia Ghent University, Belgium Ghent University, Belgium DBV
German Society for Concrete and Construction Technology, Germany Polytechnic University of Milan, Italy Indian Institute of Technology Madras, India Queens University Belfast, UK Eindhoven University of Technology, The Netherlands ETH Zurich, Switzerland Loughborough University, UK TU Delft, The Netherlands Dresden University of Technology, Germany ETH Zurich, Switzerland University of Minho, Portugal Danish Technological Institute (Concrete Centre), Denmark Brunel University London, UK Southeast University, China

Organising Committee Richard Buswell (Chair) Ana Blanco (Co-chair) Sergio Cavalaro (Co-chair) Peter Kinnell (Co-chair) Jie Xu (Co-chair) James Dobrzanski John Temitope Kolawole Muhammad Nura Isa Xingzi Liu Siduo Lei Liam White Renata Monte Ivvy Pedrosa-Cavalcante-P-Quintella Eduardo Quintella-Forencio

Organisation

Keywords RILEM Digital Concrete 3D Concrete Printing Cementitious Materials Shotcrete 3D Printing Powder Bed Printing Printing Technologies Structural Engineering

ix

Contents

Alternative Processes Zero-Waste Production of Lightweight Concrete Structures with Water-Soluble Sand Formwork . . . . . . . . . . . . . . . . . . . . . . . . . . . Daria Kovaleva, Maximilian Nistler, Alexander Verl, Lucio Blandini, and Werner Sobek An Early Trial on Milling 3D Printed Concrete Geometries: Observations and Insights of the Process . . . . . . . . . . . . . . . . . . . . . . . . Jie Xu, John Temitope Kolawole, John Provis, James Dobrzanski, Peter Kinnell, Sergio Cavalaro, Weiqiang Wang, and Richard Buswell Mobile Additive Manufacturing: A Case Study of Clay Formwork for Bespoke in Situ Concrete Construction . . . . . . . . . . . . . . . . . . . . . . Gido Dielemans, Lukas Lachmayer, Tobias Recker, Lidia Atanasova, Christian Maximilian Hechtl, Carla Matthäus, Annika Raatz, and Kathrin Dörfler Adaptive Foam Concrete in Digital Fabrication . . . . . . . . . . . . . . . . . . . Robert Schmid, Georg Hansemann, Michael Autischer, and Joachim Juhart

3

9

15

22

Structural Design and Optimisation Mesh Mould Prefabrication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ammar Mirjan, Jaime Mata-Falcón, Carsten Rieger, Janin Herkrath, Walter Kaufmann, Fabio Gramazio, and Matthias Kohler The Production of a Topology-Optimized 3D-Printed Concrete Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ticho Ooms, Gieljan Vantyghem, Yaxin Tao, Michiel Bekaert, Geert De Schutter, Kim Van Tittelboom, and Wouter De Corte

31

37

xi

xii

Contents

Injection 3D Concrete Printing (I3DCP) Combined with Vector-Based 3D Graphic Statics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yinan Xiao, Noor Khader, Aileen Vandenberg, Dirk Lowke, Harald Kloft, and Norman Hack 3DCP Structures: The Roadmap to Standardization . . . . . . . . . . . . . . . Jolien Van Der Putten, Maartje J. Hoogeveen, Marijn J. A. M. Bruurs, and Hans L. M. Laagland

43

50

Binders and Aggregates 1: Aggregates Mix Design for a 3D-Printable Concrete with Coarse Aggregates and Consideration of Standardisation . . . . . . . . . . . . . . . . . . . . . . . . . . Markus Taubert and Viktor Mechtcherine

59

Fresh and Hardened Properties of 3D Printable Foam Concrete Containing Porous Aggregates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kirubajiny Pasupathy, Sayanthan Ramakrishnan, and Jay Sanjayan

65

Sustainable 3D Concrete Printing with Large Aggregates . . . . . . . . . . . Wilson Ricardo Leal da Silva, Martin Kaasgaard, and Thomas J. Andersen Design and Fabrication of Spatially Graded Concrete Elements with Ice Aggregate Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vasily Sitnikov, Lena Kitani, Artemis Maneka, Ena Lloret-Fritsch, Juney Lee, and Benjamin Dillenburger

71

78

Binders and Aggregates 2: Alternative Binders Accelerating Early Age Properties of Ultra-Low Clinker Cements for Extrusion-Based 3D Printing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rutendo Rusike, Michael Sataya, Alastair T. M. Marsh, Sergio Cavalaro, Chris Goodier, Susan A. Bernal, and Samuel Adu-Amankwah

87

Developing Printable Fly Ash–Slag Geopolymer Binders with Rheology Modification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tippabhotla A. Kamakshi and Kolluru V. L. Subramaniam

93

Formulation and Characterization of a Low Carbon Impact Cementitious Ink for 3D Printing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Estelle Hynek, David Bulteel, Antoine Urquizar, and Sébastien Remond

99

Strategies for Reducing the Environmental Footprint of Additive Manufacturing via Sprayed Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Aurélie Favier and Agnès Petit Mechanical Performance of 3-D Printed Concrete Containing Fly Ash, Metakaolin and Nanoclay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Ahmed Abdalqader, Mohammed Sonebi, Marie Dedenis, Sofiane Amziane, and Arnaud Perrot

Contents

xiii

Binders and Aggregates 3: Strain Hardening Materials Incorporation and Characterization of Multi-walled Carbon Nanotube Concrete Composites for 3D Printing Applications . . . . . . . . 119 Albanela Dulaj, Monica P. M. Suijs, Theo A. M. Salet, and Sandra S. Lucas Properties of 3D-Printable Ductile Fiber-Reinforced Geopolymer Composite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 Shin Hau Bong, Behzad Nematollahi, Venkatesh Naidu Nerella, and Viktor Mechtcherine Feasibility of Using Ultra-High Ductile Concrete to Print Self-reinforced Hollow Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Junhong Ye, Yiwei Weng, Hongjian Du, Mingyang Li, Jiangtao Yu, and Md Nasir Uddin Development of Cementitious Metamaterial with Compressive Strain Hardening Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 Keisuke Nishijo, Motohiro Ohno, and Tetsuya Ishida Consistency of Mechanical Properties of 3D Printed Strain Hardening Cementitious Composites Within One Printing System . . . . . . . . . . . . . 145 Karsten Nefs, A. L. van Overmeir, Theo A. M. Salet, A. S. J. Suiker, B. Šavija, E. Schlangen, and Freek Bos Design and Digital Workflow Uncertainty Quantification of Concrete Properties at Fresh State and Stability Analysis of the 3D Printing Process by Stochastic Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 Zeinab Diab, Duc Phi Do, Sébastien Rémond, and Dashnor Hoxha Simulation of 3D Concrete Printing Using Discrete Element Method . . . 161 Knut Krenzer, Ulrich Palzer, Steffen Müller, and Viktor Mechtcherine Influence of Infill Pattern on Reactive MgO Printed Structures . . . . . . . 167 AlaEddin Douba, Palash Badjatya, and Shiho Kawashima Durability Evaluation of Durability of 3D-Printed Cementitious Materials for Potential Applications in Structures Exposed to Marine Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Fabian B. Rodriguez, Cristian Garzon Lopez, Yu Wang, Jan Olek, Pablo D. Zavattieri, Jeffrey P. Youngblood, Gabriel Falzone, and Jason Cotrell

xiv

Contents

Two Year Exposure of 3D Printed Cementitious Columns in a High Alpine Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 Timothy Wangler, Asel Maria Aguilar Sanchez, Ana Anton, Benjamin Dillenburger, and Robert J. Flatt Salt Scaling Resistance of 3D Printed Concrete . . . . . . . . . . . . . . . . . . . 188 Manu K. Mohan, A. V. Rahul, Geert De Schutter, and Kim Van Tittelboom Influence of the Print Process on the Durability of Printed Cementitious Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194 Jolien Van Der Putten, M. De Smet, P. Van den Heede, Geert De Schutter, and Kim Van Tittelboom Freeze-Thaw Performance of 3D Printed Concrete: Influence of Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200 Arnesh Das, Asel Maria Aguilar Sanchez, Timothy Wangler, and Robert J. Flatt Heterogeneities and Defects Mechanical Properties and Failure Pattern of 3D Printed Hollow Cylinders and Wall Segments Under Uniaxial Loading . . . . . . . . . . . . . 209 Shantanu Bhattacherjee, Smrati Jain, Manu Santhanam, and G. Thiruvenkatamani Impact of Drying of 3D Printed Cementitious Pastes on Their Degree of Hydration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216 Rita M. Ghantous, Yvette Valadez-Carranza, Steven R. Reese, and W. Jason Weiss The Environment’s Effect on the Interlayer Bond Strength of 3D Printed Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222 Gerrit M. Moelich, J. J. Janse van Rensburg, Jacques Kruger, and Riaan Combrinck Evaluation of the Bond Strength Between 3D Printed and Selfcompacting Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228 Michiel Bekaert, Kim Van Tittelboom, and Geert De Schutter Interlocking 3D Printed Concrete Filaments Through Surface Topology Modifications for Improved Tensile Bond Strength . . . . . . . . 235 Jean-Pierre Mostert and Jacques Kruger Digitally Fabricated Keyed Concrete Connections . . . . . . . . . . . . . . . . . 241 Patrick Bischof, Jaime Mata-Falcón, Joris Burger, and Walter Kaufmann

Contents

xv

Material Jetting A 3D Printing Platform for Reinforced Printed-Sprayed Concrete Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249 Lex Reiter, Ana Anton, Timothy Wangler, Benjamin Dillenburger, and Robert J. Flatt Influence of Material and Process Parameters on Hardened State Properties of Shotcrete 3D-Printed Elements . . . . . . . . . . . . . . . . . . . . . 255 David Böhler, Inka Mai, Niklas Freund, Lukas Lachmayer, Annika Raatz, and Dirk Lowke Shotcrete 3DCP Projection Angle and Speed Optimization: Experimental Approaches and Theoretical Modelling . . . . . . . . . . . . . . 261 Benjamin Galé, Thierry Ursenbacher, Agnès Petit, and Vincent Bourquin ARCS: Automated Robotic Concrete Spraying for the Fabrication of Variable Thickness Doubly Curved Shells . . . . . . . . . . . . . . . . . . . . . 267 Mishael Nuh, Robin Oval, and John Orr Particle Bed Binding Particle Bed Technique for Hempcrete . . . . . . . . . . . . . . . . . . . . . . . . . . 277 V. Danché, A. Pierre, K. Ndiaye, and T. T. Ngo Effect of Curing in Selective Cement Activation . . . . . . . . . . . . . . . . . . . 283 Friedrich Herding, Inka Mai, and Dirk Lowke Evaluating the Effect of Methyl Cellulose on Hardened State Properties in Selective Cement Activation . . . . . . . . . . . . . . . . . . . . . . . 289 Inka Mai, Friedrich Herding, and Dirk Lowke Selective Paste Intrusion: Stability of Cement Paste Mixtures Towards Changing Ambient Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296 A. Straßer, Carla Matthäus, D.Weger, T. Kränkel, and C. Gehlen Printability and Set Control Set-On Demand Concrete by Activating Encapsulated Accelerator for 3D Printing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305 Sasitharan Kanagasuntharam, Sayanthan Ramakrishnan, and Jay Sanjayan Using Limestone Powder as a Carrier for the Accelerator in Extrusion-Based 3D Concrete Printing . . . . . . . . . . . . . . . . . . . . . . . . 311 Yaxin Tao, Karel Lesage, Kim Van Tittelboom, Yong Yuan, and Geert De Schutter

xvi

Contents

Printability Assessment of Cement-Based Materials Using Uniaxial Compression Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318 Ilhame Harbouz, Ammar Yahia, Emmanuel Rozière, and Ahmed Loukili Monitoring Strain Using Digital Image Correlation During Compressive and Tensile Loading: Assessment of Critical Strain of Cement-Based Materials Containing VMA . . . . . . . . . . . . . . . . . . . . 324 Yohan Jacquet, Arnaud Perrot, and Vincent Picandet Temperature Impact on the Structural Build-Up of Cementitious Materials – Experimental and Modelling Study . . . . . . . . . . . . . . . . . . . 330 Alexander Mezhov, Annika Robens-Radermacher, Kun Zhang, Hans-Carsten Kühne, Jörg F. Unger, and Wolfram Schmidt Early Age Shear and Tensile Fracture Properties of 3D Printable Cementitious Mortar to Assess Printability Window . . . . . . . . . . . . . . . 337 Andrea Marcucci, Sriram K. Kompella, Francesco Lo Monte, Marinella Levi, and Liberato Ferrara A Strain-Based Constitutive Model Ensuring Aesthetic 3D Printed Concrete Structures: Limiting Differential Settlement of Filaments . . . . 343 Jacques Kruger, Jean-Pierre Mostert, and Gideon van Zijl Process Control, Toolpath and Inspection Process Control for Additive Manufacturing of Concrete Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351 Lukas Lachmayer, Robin Dörrie, Harald Kloft, and Annika Raatz Generative Structural Design: A Cross-Platform Design and Optimization Workflow for Additive Manufacturing . . . . . . . . . . . . . . . 357 Saqib Aziz, Ji-Su Kim, Dietmar Stephan, and Christoph Gengnagel A Closed-Loop Workflow for Quality Inspection and Integrated Post-processing of 3D-Printed Concrete Elements . . . . . . . . . . . . . . . . . 364 Norman Hack, Carsten Jantzen, Leon Brohmann, Markus Gerke, Karam Mawas, and Mehdi Maboudi Force Flow Compliant Robotic Path Planning Approach for Reinforced Concrete Elements Using SC3DP . . . . . . . . . . . . . . . . . . . . . 370 Robin Dörrie and Harald Kloft Reinforcement Integrating Reinforcement with 3D Concrete Printing: Experiments and Numerical Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379 Jon Spangenberg, Wilson Ricardo Leal da Silva, Md Tusher Mollah, Raphaël Comminal, Thomas Juul Andersen, and Henrik Stang

Contents

xvii

Flow-Based Pultrusion of Anisotropic Concrete: Mechanical Properties at Hardened State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385 Léo Demont, Malo Charrier, Pierre Margerit, Nicolas Ducoulombier, Romain Mesnil, and Jean-François Caron Core Winding: Force-Flow Oriented Fibre Reinforcement in Additive Manufacturing with Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391 Stefan Gantner, Philipp Rennen, Tom Rothe, Christian Hühne, and Norman Hack Integration of Mineral Impregnated Carbon Fibre (MCF) into Fine 3D-Printed Concrete Filaments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397 Tobias Neef, Steffen Müller, and Viktor Mechtcherine Flexural Behaviour of Steel-Reinforced Topology-Optimised Beams Fabricated by 3D Concrete Printing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404 Costantino Menna and Laura Esposito Fundamental Study on Automated Interlayer Reinforcing System with Metal Fiber Insertion for 3D Concrete Printer . . . . . . . . . . . . . . . . . . . . 411 Tomoya Asakawa, Tomoya Nishiwaki, Kazunori Ohno, Shigeru Yokoyama, Yoshito Okada, Shotaro Kojima, Youichi Satake, Yoshihiro Miyata, Yuki Miyazawa, Youhei Ito, and Hideyuki Kajita Robotically Placed Reinforcement Using the Automated Screwing Device – An Application Perspective for 3D Concrete Printing . . . . . . . 417 Lauri Hass and Freek Bos Proof-of-Concept: Sprayable SHCC Overlay Reinforcement Regime for Unreinforced 3D Printed Concrete Structure . . . . . . . . . . . . . . . . . . 424 Seung Cho, Marchant van den Heever, Jacques Kruger, and Gideon van Zijl Pre-installed Reinforcement for 3D Concrete Printing . . . . . . . . . . . . . . 430 Lukas Gebhard, Patrick Bischof, Ana Anton, Jaime Mata-Falcón, Benjamin Dillenburger, and Walter Kaufmann Wet Material Property Control Material Design and Rheological Behavior of Sustainable Cement-Based Materials in the Context of 3D Printing . . . . . . . . . . . . . 439 Silvia Reißig, Venkatesh Naidu Nerella, and Viktor Mechtcherine Measuring Plastic Shrinkage and Related Cracking of 3D Printed Concretes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446 Slava Markin and Viktor Mechtcherine Automated Visual Inspection of Near Nozzle Droplet Formation for Quality Control of Additive Manufacturing . . . . . . . . . . . . . . . . . . . . . . 453 Derk Bos and Rob Wolfs

xviii

Contents

Selected Test Methods for Assessing Fresh and Plastic-State 3D Concrete Printing Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460 John Temitope Kolawole, Danny De-Becker, Jie Xu, James Dobrzanski, Sergio Cavalaro, Simon Austin, Nicolas Roussel, and Richard Buswell Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467

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: 2-912143-05-5; 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: 2-912143-10-1; e-ISBN: 2351580230); Ed. L. Boström

xix

xx

RILEM Publications

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 PRO 10: 3rd International RILEM Symposium on Durability of Building and Construction Sealants (ISBN: 2-912143-13-6; e-ISBN: 2351580257); Eds. 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: 2-912143-19-5; 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

RILEM Publications

xxi

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 PRO 25: International RILEM Workshop on Frost Damage in Concrete (ISBN: 2-912143-31-4; 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; e-ISBN: 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: 2-912143-42-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

xxii

RILEM Publications

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 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: ConcreteLife’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

RILEM Publications

xxiii

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 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: 978-2-35158-086-8); 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: 978-2-35158-084-4); 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

xxiv

RILEM Publications

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: 978-2-35158-072-1; e-ISBN: 978-2-35158-087-5); Eds. V. L’Hostis, R. Gens, 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: 978-2-35158-074-5); Ed. K. Kovler 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); Ed. 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: 978-2-35158-115-5, 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: 978-2-35158-103-2, 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; e-ISBN: 978-2-35158-107-0; Ed. W. Brameshuber

RILEM Publications

xxv

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: 978-2-35158-110-0; 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, e-ISBN: 978-2-35158-117-9; Eds. Christopher Leung and K.T. Wan 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: 978-2-35158-121-6; 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: 978-2-35158-124-7; e-ISBN: 978-2- 35158-134-6; Eds. Z. J. Li, W. Sun, C. W. Miao, K. Sakai, O. E. Gjorv & 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: 978-2-35158-125-4; 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

xxvi

RILEM Publications

PRO 87: UHPFRC 2013 – RILEM-fib-AFGC International Symposium on Ultra-High 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. Bjegović, 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 PRO 92: CAM 2014 - 2nd International Conference on advances in chemically-activated materials (2014) 392 pp., ISBN: 978-2-35158-141-4; e-ISBN: 978-2-35158-142-1; 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, 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, H. Garrecht PRO 95: International RILEM Conference on Application of superabsorbent polymers and other new admixtures in concrete construction (2014), ISBN: 978-2-35158-147-6; e-ISBN: 978-2-35158-148-3; Eds. Viktor Mechtcherine, 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, V. M. John PRO 97: SHCC3 – 3rd International RILEM Conference on Strain Hardening Cementitious Composites (2014), ISBN: 978-2-35158-150-6; e-ISBN: 978-2-35158-151-3; Eds. E. Schlangen, M. G. Sierra Beltran, M. Lukovic, 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

RILEM Publications

xxvii

PRO 99 (online version): ICBBM 2015 - 1st International Conference on Bio-Based Building Materials (2015), e-ISBN: 978-2-35158-154-4; Eds. S. Amziane, 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, 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 PRO 106: AFGC-ACI-fib-RILEM International Conference on Ultra-HighPerformance Fibre-Reinforced Concrete – UHPFRC 2017 (2017), ISBN: 978-2-35158-166-7, e-ISBN: 978-2-35158-167-4; Eds. François Toutlemonde & Jacques Resplendino PRO 107 (online version): XIV DBMC – 14th International Conference on Durability of Building Materials and Components (2017), e-ISBN: 978-2-35158159-9; Eds. Geert De Schutter, Nele De Belie, Arnold Janssens, 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: 978-2-35158171-4, 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

xxviii

RILEM Publications

PRO 112: MSSCE 2016 - Moisture in Materials and Structures (2016), ISBN: 978-2-35158-178-0, 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: 978-2-35158-182-7, 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, e-ISBN: 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: 978-2-35158186-5; 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: 978-2-35158-189-6; Eds. Changwen Miao, Wei Sun, Jiaping Liu, Huisu Chen, Guang Ye and Klaas van Breugel PRO 118 (4 volumes): International Conference on Advances in Construction Materials and Systems (2017), ISBN Set: 978-2-35158-190-2, Vol. 1: 978-2-35158-193-3, 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; Eds. Manu Santhanam, Ravindra Gettu, Radhakrishna G. Pillai and Sunitha K. Nayar 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 Cement-based 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: 978-2-35158204-6, e-ISBN: 978-2-35158-205-3; Eds. C. Shi, Z. Zhang, K. H. Khayat

RILEM Publications

xxix

PRO 123: Final Conference of RILEM TC 253-MCI: MicroorganismsCementitious 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: 978-2-35158-215-2; 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: 978-2-35158218-3 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; PRO 130: 5th Historic Mortars Conference (2019), ISBN: 978-2-35158-221-3, e-ISBN: 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 and 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 Omikrine-Metalssi and 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

xxx

RILEM Publications

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: 2-91214323-3); 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 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: 2-912143-70-5); 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

RILEM Publications

xxxi

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 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 floors and other durability related aspects guide - Report of RILEM Technical Committee TC 268-SIF (e-ISBN: 978-2-35158-201-5); Ed. Valerie Pollet

Alternative Processes

Zero-Waste Production of Lightweight Concrete Structures with Water-Soluble Sand Formwork Daria Kovaleva1(B) , Maximilian Nistler2 , Alexander Verl2 , Lucio Blandini1 , and Werner Sobek1 1

2

Institute for Lightweight Structures and Conceptual Design, University of Stuttgart, Pfaffenwaldring 14, 70569 Stuttgart, Germany [email protected] Institute for Control Engineering of Machine Tools and Manufacturing Units, University of Stuttgart, Seidenstrasse 36, 70174 Stuttgart, Germany [email protected] https://www.ilek.uni-stuttgart.de/

Abstract. In the face of climate change and resource crisis, the construction industry can contribute to sustainable economic development by reducing its material consumption, emissions and waste. The application of lightweight construction principles in combination with circular production processes offers a comprehensive approach to this problem. The newly developed zero-waste formwork production method, based on the additive manufacturing of water-soluble sand mixture, enables to produce geometrically complex concrete structures and to reuse formwork material repeatedly in production cycles. This paper demonstrates the potential of this method for the production of the topologically optimised concrete single-span beam with spatial lattice structure (Fig. 1). First, a brief overview of the method will be given, followed then by a description of the design and production of the concrete component under the consideration of production parameters. Keywords: Recyclable formwork · Water-soluble sand formwork Lightweight concrete structures · Additive manufacturing

1

·

Introduction

The use of lightweight construction principles is one of the effective strategies to reduce material consumption, emissions and waste. However, the production of lightweight concrete structures is still usually labour-intensive, expensive, and often leads to solid waste generation. This is primarily due to the fabrication of formwork, which accounts for up to two-thirds of the total production budget. Additive manufacturing methods, including powder-bed-based 3D printing, are particularly well suited for the production of formworks for geometrically Supported by German Research Foundation (DFG). c The Author(s), under exclusive license to Springer Nature Switzerland AG 2022  R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 3–8, 2022. https://doi.org/10.1007/978-3-031-06116-5_1

4

D. Kovaleva et al.

Fig. 1. Water-soluble sand formwork (left), concrete beam (middle) and formwork material ingredients: sand, dextrin and water (right)

complex concrete components [3]. Improving the environmental aspects of these methods is a prerequisite for their broader use in future construction practices. One approach to this problem is additive manufacturing of recyclable materials. This would enable to produce spatial forms of almost any complexity and to reuse the formwork material repeatedly in production cycles. Such a method based on powder-bed-based printing of water-soluble sand mixture has been recently developed by the authors and will be demonstrated on the production of a singlespan beam with a complex lattice structure. First, the production method and the automated manufacturing unit will be briefly introduced, followed by the description of the design and production process steps of the concrete prototype.

2

Materials and Methods

The proposed method is a closed-loop production cycle consisting of two consecutive parts: production of the concrete structure and the recycling of the formwork material (Fig. 2a). The first part consists of fabricating the formwork geometry in a powder bed, cleaning it from unbound sand and pouring the concrete. The second part involves flushing out the formwork, filtering out excess water, drying and grinding. 2.1

Formwork Material

A mixture of silica sand and dextrin (a water-soluble industrial starch) is used as the primary formwork material. The binder supplied in powder form turns into a gel when water is added and binds the sand grains when it evaporates, resulting in reversible hardening of the mixture. This property enables the formwork to gain sufficient strength to withstand the hydrostatic pressure of concrete and be soluble in room temperature water. 2.2

Powder-Bed-Based Formwork Manufacturing Unit

In order to combine the properties of the formwork mix with the powder bed additive manufacturing, the wetting and drying process steps were converted into a layer-by-layer water-jetting procedure followed by a thermal treatment.

Zero-Waste Production of Lightweight Concrete Structures

5

Fig. 2. Overview of production method (a) and the perspective view of manufacturing unit (b)

For automatic water dosing and binder activation, an industrial DoD (Dropon-Demand) printhead with 32 nozzles of 130 µm diameter was used. By rotating the printhead in relation to the print direction, an adaptable printing resolution varying from 4.5 mm at 0◦ to 0.13 mm at 85◦ can be implemented to optimise the printing time. To ensure the curing of formwork, drying takes place immediately after printing every layer using carbon infrared (IR) heaters. These heaters operate in the mid-infrared frequency range, best suited for heating and evaporating water molecules. An average of 40–60 s are sufficient to dry a 1.2 mm deep moist sand layer at a temperature not exceeding 150 ◦ C to prevent dextrin polymeric chains from decomposition. The manufacturing unit, including the components described above, as well as the sand supply system, was designed as a 3-axis gantry system with a powder bed of 70 cm × 100 cm × 40 cm of the printing volume (Fig. 2b). The unit is modular and can be extended along X-axis that enables to increase the printing area as well as to add other modules to the production line.

3

Design and Production of Lightweight Concrete Beam

The setup described above was tested by the production of formworks for a lightweight concrete component with a complex spatial structure. 3.1

Design of the Lattice Beam

The design approach is built upon a broader study of functionally graded concrete structures carried out by the authors in recent years [1,2]. It is based on structural optimisation techniques where the material is distributed within the defined design domain according to the stress state under the given loads. The production method plays a decisive role here, as it determines how the stress field can be converted into a producible geometry with variable properties. In this

6

D. Kovaleva et al.

Fig. 3. Overview of the design-to-production flow within Rhino/Grasshopper digital environment. Structural analysis of solid beam (a), generation and structural analysis of new conformal lattice (b), modelling of concrete structure (c), generation of formwork geometry (d), arrangement of formwork parts in virtual powder bed (e), trajectory planning (f)

example, a periodic gradient lattice structure was chosen as the basic topology. This gives the possibility to materialise both the magnitude and the direction of the principal stresses. The design of a concrete structure begins with the structural analysis of the design domain filled with uniform solid material (Fig. 3a). Then, based on the obtained directions and magnitudes of principal stresses, a new conformal lattice structure of a given scale is created, followed by its structural analysis and the determination of the cross-sections of its elements (Fig. 3b). Finally, the outer surface of a closed mesh of concrete structure is created using voxel-based modelling and Boolean operations (Fig. 3c). 3.2

Preparation and Printing of Formwork

In the next step, the formwork geometry is created by offsetting the mesh surface outwards to a given distance (in the present case - 13 mm), sufficient to resist the hydrostatic pressure of concrete (Fig. 3d). Then the formwork body is split into parts and oriented in the virtual volume of the powder bed to optimize the production time (Fig. 3e). The printing resolution is defined by the user (in the present case - 1.2 mm), which results in automatic determination of printhead angle, the height of the layer and the configurations of printing trajectories (Fig. 3f). Finally, the complete structure of the G-code file, consisting of printing routines, IR-drying periods and service movements for all layers, is compiled and sent to the machine. Total production time is defined as the sum of four

Zero-Waste Production of Lightweight Concrete Structures

7

Fig. 4. Production of formwork. DoD water-jetting (a), IR-drying (b) and cleaning from unbound sand (c)

basic parameters: printing resolution, printing speed, time for drying and service movements. For the present example, the formwork geometry consisting of three parts with an overall printing volume of 45 cm × 27 cm × 12 cm, was printed in 5.5 h including 60 min for water-jetting, 190 min for drying and 80 min for service movements (Fig. 4). 3.3

Casting, Demolding and Reuse of Formwork Material

After the printing is finished, the formwork is excavated from the powder bed, cleaned of unbound sand and assembled for casting. The properties of the concrete structure, such as the smallest possible diameter of the individual lattice struts, are defined in the design phase depending on the flowability and the maximum aggregate size of the concrete mixture applied. In the present case, a self-compacting concrete premix Sika Grout 551 with a grain size of 1 mm was used. It does not require any additional shaking, which could damage the formwork, and also provides an early strength of 40 MPa, so the formwork could be removed after 24 h. After the concrete is hardened, the formwork is placed on a grid and rinsed with the running water (Fig. 5a). The total production time for the beam beyond hardening period is around 8 h, including formwork printing (5.5 h), cleaning, assembling, placing reinforcement, casting and demolding. To reuse the formwork material, the sand with the binder is separated from the excess moisture, collected in a container and placed for 6 h in an oven at 110◦ C to let all the water evaporate. The dried material is crushed, milled on a machine and is ready to use. Both new and recycled material mixtures were tested during the production series. Studies have shown that the strength of the recycled material compared to the new one is reduced by about 20–30% (4–3.5 MPa vs. 5 MPa). This is related to the partial solubility of the dextrin in water and is currently being researched.

8

D. Kovaleva et al.

Fig. 5. Washing out of formwork (a) and close-ups of formwork and concrete structures (b)

4

Conclusion

In this paper, a zero-waste production method of complex concrete structures using water-soluble sand formworks was presented and tested on the production of a lattice concrete beam. The results have shown that the use of watersoluble formworks contributes to sustainable production and enables exploring new types of producible geometries, for example, the components with integrated networks of cavities, undercuts, and bottlenecks. Compared to other production methods, this technology can offer an advantage in the manufacturing of lightweight structures with non-uniform stresses, such as slabs, beams or columnto-slab transition zones, where the preferred rheological properties of concrete and material savings can be combined with optimal integration of rebars. The scalability of the method and its application in production of other structural typologies will be addressed in future studies. Acknowledgments. This research is funded by German Research Foundation (DFG) as a part of the Priority Program SPP 2187: “Adaptive modularized construction made in flux” (grant number 423987937). The preliminary studies to this work were supported by DFG under Germany’s Excellence Strategy EXC 2120/1-390831618.

References 1. Herrmann, M., Sobek, W.: Functionally graded concrete: numerical design methods and experimental tests of mass-optimized structural components. Structu. Concr. 18(1), 54–66 (2017). https://doi.org/10.1002/suco.201600011 2. Kovaleva, D., Gericke, O., Kappes, J., Tomovic, I., Sobek, W.: Rosenstein pavilion: design and structural analysis of a functionally graded concrete shell. Structures 18, 91–101 (2019). https://doi.org/10.1016/j.istruc.2018.11.007 3. Meibodi, M.A., et al.: Smart Slab. Computational design and digital fabrication of a lightweight concrete slab (2018). http://papers.cumincad.org/cgi-bin/works/ paper/acadia18 434

An Early Trial on Milling 3D Printed Concrete Geometries: Observations and Insights of the Process Jie Xu1(B) , John Temitope Kolawole1 , John Provis2 , James Dobrzanski1 , Peter Kinnell3 , Sergio Cavalaro1 , Weiqiang Wang1 , and Richard Buswell1 1 School of Architecture, Building and Civil Engineering, Loughborough University,

Loughborough, UK [email protected] 2 Department of Materials Science and Engineering, The University of Sheffield, Sheffield, UK 3 Wolfson School of Mechanical, Electrical and Manufacturing Engineering, Loughborough University, Loughborough, UK

Abstract. As 3D Concrete Printing (3DCP) technology develops, requirements on the form and surface quality of the final products are increasing. Layer-wise deposition results in the so-called ‘staircase effect’ which can lead to limitations on the attained precision and accuracy of geometries. Applying other shaping processes with a higher manufacturing precision can be deployed to combat this and milling is one example that has been shown to yield benefits. This paper presents an early trial of a milling process applied after printing and before the final hardened state of the material. A case study of a panel component is presented and observations are reported, which include: the critical nature of the material state, the control of debris and milling path sequence and direction. Insights are formulated into a three-tier structure to help develop signpost issues for the development of the approach. Keywords: Concrete printing · Milling · Subtractive postprocess · Geometric conformity · Surface finish

1 Introduction 3D Concrete Printing (3DCP) has been the focus of significant research and development activity in architecture, construction and infrastructure in recent years [1]. It enables significant freedom to create complex building forms without formwork in a digitally driven, automated way. Like the other Additive Manufacturing processes, the implicit layer-wise deposition limits the resolution and precision of the features that can be manufactured; sharp corners and surface finish are two examples [2]. There may also be deformation in build due to the material rheology [3]. Such matters become more significant as manufacture moves away from one-off components to those that are needed to integrate with other building systems. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 9–14, 2022. https://doi.org/10.1007/978-3-031-06116-5_2

10

J. Xu et al.

Inspired by hybrid manufacturing processes found in manufacturing [4], a few researchers have proposed subtractive post-processes with higher precision such as milling, onto printed surfaces to obtain desirable tolerances [5]. The milled surface finish of printable concrete has also been explored in [6], but the milling process dynamics remain unreported. This paper provides observations of implementing milling mortars prior to their hardened state based on an early trial of manufacturing a curved, ribbed panel component at Loughborough University. Insights are discussed to support future work to develop systematic milling strategies that compliment 3DCP.

2 Milling a Printed Panel Part A test part of 950 mm long, 400 mm wide and 100 mm high was designed to be curved, ribbed and reinforced. Figure 1 shows the curved surface (Fig. 1a), the reverse surface with three ribs (Fig. 1b) and the cross-section (c), which also indicates where a sheet of alkali-resistant glass textile reinforcement, and 6 mm diameter fibre-reinforced polymer (FRP) bars, were placed.

Fig. 1. The trial panel CAD models showing the upper surface (a), ribbed sections (b) and crosssection (c).

A hybrid concrete printing (HCP) workflow was developed and implemented for the manufacture, as shown in Fig. 2. The milling toolpath needs to be generated and physically verified by moving the robot along the toolpath, prior to manufacture. Because this part was not flat, the preparation of a support surface was required, for which the milling was applied to a dense foam base to prepare a curved support in a pre-production step. Reinforcement (textile and FRP bars here) was hand placed during printing, using the printed material as the support. Optical measurement was applied on the completed component for geometric verification (details can be found in [2, 5]). The part was printed in curved layers onto the milled foam base in an ambient temperature of ~ 20 °C. One layer was printed, the process paused to allow the textile reinforcement to be placed, and then resumed, pausing again for the placement of the FRP bars, before the final layers were printed. The mix reported in [2, 6] was used: a water-binder-sand ratio of 0.255:1:1.499 with 1% superplasticizer and 0.506% retarder (relative to the mass of the binder: 70% cement, 20% fly ash and 10% silica fume). The process configuration was a 30 mm wide filament with a 20 mm stepover and 13 mm layer height.

An Early Trial on Milling 3D Printed Concrete Geometries: Observations

11

Fig. 2. HCP process workflow for trial panel manufacturing.

The material state is critical for the milling process as the material is required to be sufficiently rigid to withstand the milling force but also soft to allow cutting with milling tools [6]. In this case, the printed part (Fig. 3a – near net-shape) was allowed to cure in the ambient environment for approximately four hours to obtain an appropriately stiff surface. The milling operations used a 12 mm diameter, 283 mm long, 2-flute ball nosed cutter and produced the part shown in Fig. 3b (the net-shape).

Fig. 3. A) The printed part (near-net-shape) and b) the part post-milling (net-shape).

3 Results and Discussion A number of the trial panel parts were manufactured to explore the process parameters that would yield good quality component. Several issues that affected process performance or generated failures were observed. These are presented as a matrix in Table 1

12

J. Xu et al.

as examples arranged in three levels that improve the overall part quality by increasing degrees. Table 1. Three-level 3DCP postprocess definition supported by the results of the experiment. Process Level

Level 1:

Level 2:

Level 3:

Good Form Conformance

Good Surface Quality

Good Efficiency

Observations of counter-exPlastic debris building amples

Wrong milling sequence Comparison of rib side surup and forcing the side causing the central rib to face finish under climb millrib to collapse under the collapse, saved by manual ing and conventional milling ‘push’ of the cutter strengtheningof the rib

Key factors Debris attachment from evidence

Milling path sequence

Milling direction; Debris attachment

Correlation with material state

Medium

Medium

High

Postprocess took 10 times of the printing time (0.5h)

Curing speed; Cut efficiency; Debris attachment High

An uncontrolled curing environment (the four hours post printing) proved to be problematic. Estimates of whether the correct milling state had been achieved were based on the exposed surface of the part. For the majority of trials, once the milling cut beneath the first few millimetres of the surface, the material was too soft to cut effectively, creating some of the issues highlighted in Table 1. Controlling water loss and ensuring consistent curing through at least the milling depth is important to enable this approach to be applied effectively. ‘Good form conformance’ (Table 1: level 1) is achieved when the milling process does not cause any damage or detriment to the net-shape (the target geometric form). Two examples that prevented this from being accomplished were the build-up of debris (milling ‘chippings’) and the wrong sequence of applied passes with the milling tool. Ideally, the milled material should form ‘chips’, or something similar, but the problems experienced when the material if not set sufficiently was that the wet material stuck to the surface of the part and the tool, where it would accumulate and be ‘pushed’ around the surface by the cutter as it made its subsequent passes. This is difficult to remove and can lead to damage to weaker features (Table 1, level 1: left hand image) and so management of debris is therefore very important. Another factor that can lead to damage to the net-shape is that ability of the material to sustain the forces imposed by the milling, and this can result in a sensitivity to the sequencing of passes of the tool, particularly when creating fine features (Table 1, level 1: right hand image). Because each milling pass changes the object geometry (shifting the near net-shape a step closer to the net-shape) the rigidity and positional constraints, the

An Early Trial on Milling 3D Printed Concrete Geometries: Observations

13

path sequence needs to be carefully designed to ensure the tool forces can be sustained throughout manufacture. ‘Good surface quality’ (Table 1: level 2), is a primary motivation for adding milling to the 3DCP process. The attainable quality of the surface, and the ability to milling effectively is affected by the hardened state of the material [6], the milling direction and rotation and the manner in which the tool it applied to the surface are also all influencing factors. The material state remains constant when machining typical materials in manufacturing [7], but here the material is transitioning from a plastic state into its final hardened state. Here, the direction of rotation of the tool proved to be important. A conventional milling approach (where the direction of the rotation of the tool goes with the direction of tool travel) tends to scrape debris from the surface to create a neat surface, whilst climb milling (with opposite rotation) was found to ‘press’ the plastic debris back into the surface, leading to a poor surface finish or even damage. This was most evident on the vertical surface of the rib (see image in Table 1: level 2). In addition, two consequences of the attachment of the debris to the surface were that: the attachment of the debris clogged the milling tool reducing the cut quality; and also the unremoved debris on the part surface could set further and damage the surface finish. ‘Good efficiency’ (Table 1: level 3) highlights the issues found over manufacturing time to implement the process. It took about one hour to complete the milling plus the four hours after 3DCP to allow the part to cure sufficiently. The printing operation, even with the placement of reinforcement, was around 30 min. There is, however, great potential to reduce this with the use of admixtures. Some final observations were the trade-off between the softer material being (hypothetically) easier to cut and the resultant surface quality: it was found that several passes for the tool were needed to clean and improve the surface to achieve the required finish. In addition, removing the debris, which for this part largely fell onto the finished part because of its predominantly horizontal orientation, effectively doubled the milling time to allow stopping the machine to clean by hand.

4 Conclusions This paper reports on early trials of milling 3D printed concrete prior to its fully hardened state. Observations and issues are discussed in terms of three levels of attainment required to deliver a good quality result. The key finding, and perhaps not surprisingly, is the importance of the control of the mortar hardening state. The material state dominates the success of the process effecting the consistency of the debris created during the milling, which can stick to the surfaces of the part and tools. This also has implications for cutting sequencing and the forces that can develop while shaping features leading to potential damage of the net-shape: the target geometry. The degree of material set also influences the ability for the printed structure to sustain the forces imposed by the milling operation and hence it becomes obvious to consider a material ‘open time’ for the milling in much the same way as the readily accepted principles of printing concrete.

14

J. Xu et al.

Acknowledgements. This work was supported by the UK’s Industrial Strategy Challenge Fund for Transforming Construction, UKRI grant number EP/S031405/1 and grant number EP/SO19618/1.

References 1. Ma, G., Buswell, R.A., da Silva, W.R.L., Wang, L., Xu, J., Jones, S.Z.: Technology readiness: a global snapshot of 3D concrete printing and the frontiers for development. Cem. Concr. Res. 156, 106774 (2022) 2. Xu, J., et al.: Inspecting manufacturing precision of 3D printed concrete parts based on geometric dimensioning and tolerancing. Autom. Constr. 117, 103233 (2020) 3. Wolfs, R.J., Bos, F.P., Salet, T.A.M.: Early age mechanical behaviour of 3D printed concrete: numerical modelling and experimental testing. Cem. Concr. Res. 106, 103–116 (2018) 4. Nassehi, A., Newman, S., Dhokia, V., Zhu, Z., Asrai, R.I.: Using formal methods to model hybrid manufacturing processes. In: ElMaraghy H. (eds) Enabling Manufacturing Competitiveness and Economic Sustainability, pp. 52–56. Springer, Heidelberg (2011)https://doi.org/ 10.1007/978-3-642-23860-4_8 5. Buswell, R.A., et al.: Geometric quality assurance for 3D concrete printing and hybrid construction manufacturing using a standardised test part for benchmarking capability. Cem. Concr. Res. 156, 106773 (2022) 6. Dobrzanski, J., et al.: Milling a cement-based 3D printable mortar in its green state using a ball-nosed cutter. Cement Concr. Compos. 125, 104266 (2022) 7. Tormach, CLIMB MILLING VS. CONVENTIONAL MILLING (SNEAKY CNC TRICKS) https://blog.tormach.com/climb-milling-versus-conventional-milling-sneaky-cnctricks Accessed 31 Jan 2022

Mobile Additive Manufacturing: A Case Study of Clay Formwork for Bespoke in Situ Concrete Construction Gido Dielemans1(B) , Lukas Lachmayer2 , Tobias Recker2 , Lidia Atanasova1 , Christian Maximilian Hechtl3 , Carla Matthäus3 , Annika Raatz2 , and Kathrin Dörfler1 1 TT Professorship Digital Fabrication, Technical University of Munich, Munich, Germany

[email protected]

2 Institute of Assembly Technology, Leibniz University Hannover, Hannover, Germany 3 Chair of Materials Science and Testing, Technical University of Munich, Munich, Germany

Abstract. The in situ production of concrete building components with Additive Manufacturing (AM) provides new possibilities in design and function. Current deployable solutions are often stationary gantry systems, which need to increase in size with the constructed object. This research aims to address this issue by using mobile robotic systems for in situ AM instead, which can manufacture structures that exceed their static work range. However, where stationary AM systems inherently exhibit a high level of accuracy, mobile AM systems must be context-aware through onboard sensing and therefore pose a significant research challenge in their deployment and operation. A case study is performed with a mobile AM system using a print-drive-print approach for the sequential fabrication of a 1:1 scale clay formwork of a bespoke, reinforced, and lightweight-concrete column, on which this paper presents first results. A two-tiered system is applied and validated, with initial global localization through 2D SLAM, and a second refinement relative to the work piece through a 2D scanner fitted at the end-effector. Keywords: Additive manufacturing · Extrusion 3D printing · Clay formwork · Mobile robotics · Architecture and digital fabrication

1 Introduction This paper presents the initial findings of an experimental study designed to explore the capabilities of a mobile robot for extrusion 3D printing of a clay formwork for bespoke building components, by using a sequential print-drive-print approach. We present a newly developed collaborative mobile robot that serves as a research platform to study various architectural applications and bring AM to the construction site (Fig. 1). The accuracy and repeatability of a two-tier navigation and localization system is demonstrated and validated by the AM of a column formwork from multiple locations, where the mobile robot is navigated around a pre-fabricated reinforcement. The mobile robot prints and drives in an alternating fashion; After a segment is manufactured, the robot is © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 15–21, 2022. https://doi.org/10.1007/978-3-031-06116-5_3

16

G. Dielemans et al.

navigated to a new location to manufacture the successive segment, the diagram of this alternating building sequence is shown in Fig. 2. With this experiment, we aim to validate that the mobile robot can additively manufacture a component formwork with sufficient accuracy despite the production from several subsequent robot locations over numerous navigation- and localization maneuvers.

Fig. 1. A 9-axis mobile construction platform Fig. 2. Building sequence for the equipped with a clay extrusion end-effector and column formwork printing two-tier localization system comprising A) a Laser Range Finder on the base for rough localization via 2D SLAM in global world frame, and B), a 2D Profile Laser Scanner on the end-effector for refined localization through recordings of the work piece.

2 State of the Art The use of mobile systems for AM is currently being explored in several contexts, with various scales in application. This ranges from small robots that scale the structure under construction [1] to larger, mobile cranes [2]. Some concepts have highlighted the unbounded workspace of such systems through their mobility, where the feasibility of the system is reliant on sensing systems that can maintain a high level of accuracy [3, 4]. The potential of these systems is further explored for cooperative operation with the aim to increase the scale of printed objects, while increasing the effectiveness and reducing the construction time through parallelization. The combination of AM techniques with other fabrication strategies allows for expanding its application space. For example, due to the form-freedom of concrete in its liquid form, a combination of additive manufacturing and casting is a viable solution to obtain novel geometries and add structural requirements. Where concrete extrusion and casting was explored in the creation of several artistic columns [5], combinations of different materials for casting and extrusion have shown promising results [6–8]. Main challenges of this include the stability of the structure during printing, which is amplified

Mobile Additive Manufacturing: A Case Study of Clay Formwork

17

by the complexity of the shape, and the hydrostatic pressure applied to the structure by the casting material. The design of the material is therefore highly important, which has been highlighted in the various developments of set on demand concrete mixtures [9, 10].

3 Method 3.1 Design Principles and Mobile Manufacturing Process A column is designed such that its formwork is manufactured of several discrete polygonal segments using clay extrusion with a mobile robotic system. After a set of segments is completed, initially six segments, then every 3 segments, a layer of lightweight concrete can be cast. The formwork geometry is constructed around a reinforcement cage, which illustrates the potential for construction of custom complex reinforced concrete building components using mobile AM methods, as depicted in Fig. 3.

Fig. 3. AM sequence, displaying the initial printing poses 1–6, and the last poses, 30–32.

The column has a wide base, narrowing to a reduction of the cross section to 60% of its size at 1.5 m height, before flaring outward and reaching a total height of 2 m. Within this experiment, a single robotic system is used to construct the building component, while the segmented nature of the structure reinforces the concept to be expanded towards using multiple robots cooperatively constructing a single object in the future. 3.2 Casting Process The clay mold is built both horizontally and vertically from several hexagonally shaped segments (see Sect. 3.1). When casting concrete into the 3D-printed mold of clay, it is crucial to limit the pressure applied on the fresh mold. Due to its low density, lightweight concrete is particularly suitable for this application. In this experiment, the column is cast in sections upwards as the formwork is printed, to reduce the hydrostatic pressure

18

G. Dielemans et al.

on the formwork and fixate the geometry, while it would benefit the interlayer bond to cast concrete into the formwork at once. Contradicting requirements exist on the material regarding form-filling, compaction, hydrostatic pressure, and load-bearing behavior during casting. For a fluid material, little compaction energy is required while good form filling behavior is achieved. However, a high hydrostatic pressure acts on the mold as the material cannot bear its own weight, which could lead to the collapse of the mold. In contrast, stiff materials have a limited form filling capacity, but can support large casting heights. The setting time determined with VICAT starts at approx. 4 h and ends at 5 h, with the objective that the open time is longer than the time for printing the next set of segments, so that the formation of cold joints can be reduced. The structural build-up in the fresh concrete before setting fixates the formwork allowing fabrication of subsequent segments. Pre-investigations showed no collapse of the clay mold, but insufficient form filling and a slight deformation due to compaction energy applied. For the column, the mix design shown in Table 1 is applied. The thixotropy enhancing agent is used to enhance both, form filling and green strength at rest. Table 1. Mixture design and casting testing for lightweight concrete casting material Material

g/dm³

Cement (CEM I 52.5 R)

448.4

Lightweight aggregates: expanded glass granulate 0.1-2 mm

319.7

Limestone powder (Warsteiner Kalksteinmehl)

110.4

Silica fume (Silicol P) Water including water absorpon of lightweight aggregates

62.1 317.6

Superplascizer (BASF ACE 460)

2.8

Thixotropy enhancing agent (BASF)

0.1

3.3 Navigation, Localization, and Positioning The mobile robot is equipped with 2-tier localization system (see Fig. 4, left), its first tier being a 3D LiDAR, which allows for mapping of the environment and a rough estimation of the global robot pose in the world frame (with ±5 cm and ±3° estimated accuracy), in this case done by fusing the LiDAR in 2D with the wheel odometry. The base motion planning is based on a planar, wheeled omnidirectional model. The accuracy with which the base navigation system brings the robot to the desired position only needs to be sufficient for the robot to carry out the laser scan at approximately the right location for the refined localization, i.e., the second tier of the system. Refined robot localization is achieved by fitting two point clouds captured with a 2D laser scanner (Keyence LJV8200) mounted on the end-effector. Before changing location, the robot scans a segment of the work piece. After navigating to a new position, the robot scans the same segment, which is then aligned with the previous scan and thus used to estimate the robot’s new pose relative to the work object. This refined localization must be in the sub-centimeter range, to align the segments with extruded strands of approximately 7 mm in width.

Mobile Additive Manufacturing: A Case Study of Clay Formwork

19

Fig. 4. Diagrams depicting the two-tier mobile AM localization system (left), and a visualization of the robot’s mapping and navigation environment.

4 Case Study and Initial Results First experiments were conducted to validate the functionality of the approach for component-based localization. For this purpose, a column section was printed and measured with the presented robot. The end effector scanner was moved in a zigzag pattern over the printed component. Afterwards, the movement of the robot was simulated by moving the printed component. After scanning the same path again, the ICP algorithm provides the position correction values shown in Table 2. For verification, the displacement was measured using a 3DoF Faro Laser Tracker Vantage and three targets. Since the external measurement, and the approach used via the ICP provide the same offset values, this method could further be used to localize the robot in the subsequent printing process of large-scale components from several robot locations and validated accordingly (Fig. 5). Table 2. Measured transformation parameters from ICP and 3D reference laser measurement. Trans./Rot XYZ

X

Y

Z

Distance

X

Y

Z

Trans. ICP/[cm,°]

2.898

0.374

0.021

2.92

−0.084

−0.080

17.282

Trans. Faro Vantage [cm,°]

3.000

0.000

0.010

3.00

−0.070

−0.072

17.73

For the initial position, the relative localization was executed by manual pointmeasuring and alignment with known points situated on the work object’s base, which additionally allowed to construct a transformation between the robot map and the architectural design environment. After initial manual alignment, the final layer of each segment was scanned for further automatic alignment using the profile laser as described above; this concept could be validated on 2 printed segments.

20

G. Dielemans et al.

Fig. 5. Clay extrusion of segment 1 (right) and 2 (left)

5 Conclusion and Further Research The presented case study provides a novel method of constructing complex geometries using a mobile AM system, which are to be extended to the use of multiple robotic systems to provide scalability in terms of both size and efficiency. This can be explored for both cooperative as collaborative cases, either with the system performing the same process or providing extended capabilities with other materials or processes. With regards to localization and a print-drive-print approach, further development into real-time data recognition will be conducted. Capturing 2D-profile scans with the attached scanner during the production process will not only provide as-built data of the printed component but will also enable reacting to material deformation through print path adjustment. In the future, this should also allow for printing-while-moving maneuvers, i.e., repositioning of the robot’s base during the printing process. The effect of lightweight aggregates on the interaction between clay and concrete will be considered in more detail in subsequent studies to verify internal curing of the concrete through release of previously absorbed water by the lightweight aggregates. Additional investigations are planned on the interaction between clay and concrete with regards to the surface quality. This involves studies of the distribution of water across the cross-section of the casted concrete. Finally, further material development for mitigating hydrostatic pressure while maintaining good surface quality are targeted. Acknowledgments. This work has been executed within projects A03, B04, and B05 of the collaborative research center TRR277 – Additive Manufacturing in Construction (AMC) as funded by the German Research Foundation (DFG) - Project number 414265976 - TRR 277.

References 1. “Minibuilders.” http://robots.iaac.net/ Accessed 2 Aug 2017 2. Keating, S.J., Leland, J.C., Cai, L., Oxman, N.: Toward site-specific and self-sufficient robotic fabrication on architectural scales. Sci. Robot. 2(5), 15 (2017). https://doi.org/10.1126/scirob otics.aam8986 3. Sustarevas, J., Benjamin Tan, K.X., Gerber, D., Stuart-Smith, R., Pawar, V.M.: YouWasps: Towards Autonomous Multi-Robot Mobile Deposition for Construction. In: IEEE International Conference Intelligent Robots Systems, pp. 2320–2327 (2019) https://doi.org/10.1109/ IROS40897.2019.8967766

Mobile Additive Manufacturing: A Case Study of Clay Formwork

21

4. Zhang, X., et al.: Large-scale 3D printing by a team of mobile robots. Autom. Constr. 95(August), 98–106 (2018). https://doi.org/10.1016/j.autcon.2018.08.004 5. Anton, A., et al.: Concrete choreography. Fabr. 2020, 286–293 (2020). https://doi.org/10. 2307/J.CTV13XPSVW.41 6. Burger, J., et al.: Design and fabrication of a non-standard, structural concrete column using eggshell: ultra-thin, 3D printed formwork. In: Bos, F.P., Lucas, S.S., Wolfs, R.J.M., Salet, T.A.M. (eds.) DC 2020. RB, vol. 28, pp. 1104–1115. Springer, Cham (2020). https://doi.org/ 10.1007/978-3-030-49916-7_105 7. Wang, S., Conti, Z.X., Raspall, F.: Optimization of Clay Mould for Concrete Casting Using Design of Experiments (2019) 8. Bruce, M., Clune, G., Xie, R., Mozaffari, S., Adel, A.: Cocoon-3D Printed Clay Formwork for Concrete Casting (forthcoming) (2022) 9. Szabo, A., Reiter, L., Lloret-Fritschi, E., Gramazio, F., Kohler, M., Flatt, R.J.: Processing of set on demand solutions for digital fabrication in architecture. In: Mechtcherine, V., Khayat, K., Secrieru, E. (eds.) RheoCon/SCC -2019. RB, vol. 23, pp. 440–447. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-22566-7_51 10. Lloret-Fritschi, E., et al. Challenges of real-scale production with smart dynamic casting. In: Challenges of Real-Scale Production with Smart Dynamic Casting, pp. 299–310 (2019)

Adaptive Foam Concrete in Digital Fabrication Robert Schmid1(B) , Georg Hansemann1 , Michael Autischer2 , and Joachim Juhart2 1 Institute of Structural Design, Graz University of Technology, Graz, Austria

[email protected] 2 Institute of Technology and Testing of Construction Materials, Graz University of Technology,

Graz, Austria

Abstract. One of the process-related advantages of 3D concrete printing (3DCP) is the exact placement of concrete. It only seems logical, that the production method has to be developed further so that a specific concrete with specific properties can be actively placed where it needs to be. For this purpose, a suitable technology has been developed, which makes it possible to print foam concrete in different densities by extrusion method. (150–1000 kg/m3 ). This innovation opens up new areas of shaping and manufacturing of effective building structures. In contrast to conventional casting of concrete, construction components can now be composed of predefined zones that have different building physic, static and structural requirements due to the corresponding concrete type. (gradient concrete structures). The research at hand determines the requirements along the entire process chain through several prototypes. The results provide information about the feasibility of 3DP with adaptive foam concrete. With the possibility of grading the properties of the concrete within each building component, lighter and more sustainable buildings can be realized while conserving valuable resources. Keywords: Foam concrete · Adaptive densities · Additive fabrication · 3D printing · Digital fabrication · Resource-efficient

1 Introduction Foam concrete (FC) is a material that is slowly gaining attention in the field of Digital Fabrication with Concrete (DFC). As a lightweight cementitious material with excellent binder/volume ratio and insulating properties, FC has great potential to be an ideal addition in DFC. It seems particularly desirable to further develop FC so that it can be used in 3D concrete printing (3DCP) [1–4]. The precise and delicate handling of concrete as a material was unusual, but it did not take long before this process-related property was recognized to be a great opportunity. The build-up process makes it possible to adjust the quantity and quality of the printing material to the local required parameters [5]. The research at hand provides valuable information about the feasibility of adaptive foam concrete (AFC) in 3DCP and presents applications that use AFC to reduce CO2 equivalents in building industry. The aim of the work was to determine the technical © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 22–28, 2022. https://doi.org/10.1007/978-3-031-06116-5_4

Adaptive Foam Concrete in Digital Fabrication

23

requirements along the process chain and to gather initial experience by means of prototype production. The development of the printing system for AFC took place in two phases. After initial findings using a prototype, the experience gained was used to produce a more advanced printing system. The FC was made by the pre-foaming method [2]. The achievable densities were limited by the delivery rate of the foam generator. It was possible to produce FC in the density range of 150–1000 kg/m3 .

2 Experimental Setup The dry material is automatically mixed by the mortar mixing pump (1) and temporarily stored in a material container. The protein foam is supplied just in time by means of the foam generator (2). The two machines are interconnected so that the desired FC density can be easily changed at any time. It is defined by the ratio of the feed rates of the cement paste and protein foam. The low-destructive mixing of the two material components is then ensured by using a static mixer (3). The foam concrete (A) is conveyed into the printing nozzles mixing chamber where the setting accelerator (B) is injected. The accelerator pump (4) is responsible for metering. In the mixing chamber, homogenization is again performed by a rotating mixing shaft. A pinch valve in front of the nozzle outlet enables starting and stopping operations. The replaceable nozzle tips are responsible for shaping the deposited foam concrete layer (Fig. 1).

Fig. 1. (1) mortar pump; (2) foam generator; (3) static mixer; (4) accelerator pump; (5.1) printing nozzle; (5.2) printing nozzle with material buffer

Printing nozzle variant 5.2 has an intermediate material buffer to ensure continuous printing with FC even over long conveying distances. The feed rate is controlled by an additional pump within the printing nozzle. The connected machine components are controlled via the central robot system. The modification of the respective densities was done by hand, but can be embedded in the machine code.

24

R. Schmid et al.

3 Material Design 3.1 Raw Material As main material for the solids in the first development phase was a CEM I 52.5 N and a fine limestone powder (type I) were selected. A synthetic product and a proteinbased product were used as foaming agents. The used alkali-free accelerator had to be mixed with water before insert. The dosages of solids varied between 25% and 55%. In the second development phase, the objective was to select materials which allow a production of foam concrete (density < 200 kg/m3 ) and conventional concrete (density ≈ 2200 kg/m3 ) out of one recipe. The used cement, the protein foaming agent and the accelerator with a solids content of 50% are out the first phase. In addition, the recipe contains a latent hydraulic binder, a limestone powder (type II), sand with a grading between 0.1–0.4 mm and a proportion of a powdered superplasticiser. The values of the particle size analysis of the solids are in Fig. 2.

Fig. 2. Particle size analysis of the selected solids

3.2 Mix Design The composition of the recipe is designed for a smooth transition from normal concrete to foamed concrete during the process. No adaptations are made to the mix during the changeover from normal concrete to FC. Only the dosage of the hardening accelerator has to be adapted to the cement content. The cement content of 1 m3 concrete is 700 kg. The fast reacting cement in combination with the hardening accelerator provides the required early strength, but can be sufficiently processed by adding in the other constituents. The latent hydraulic binder ensures increased final strength. By adding the superplasticiser, the water content of the recipe can be reduced, with positive effects on the early and

Adaptive Foam Concrete in Digital Fabrication

25

final strength. The fine sand fraction in the recipe is essential for the workability of the mixture. A coarser sand grain is hardly possible for use in foamed concrete due to sedimentation problems. Fresh State Properties. The first phase of development with cement and limestone powder (type I) showed a much more stable performance of the protein-based foaming agent as the synthetic product. The FC from the protein agents achieved densities of 180– 550 kg/m3 . The FC was not capable of being printed at this stage due to the insufficient pressure device. In the second phase, on the already adapted equipment, AFC could be printed in densities of 150–1000 kg/m3 . Without adding the foam, concrete with a density of approximately 2200 kg/m3 is printed with the same recipe. Hardened Properties. The 28 d compressive strengths (fcm,28d ) of the specimens from the first phase of cement and limestone powder (type I) reached 0.1–0.4 MPa at a density of 180 kg/m3 , depending on the different compositions of cement and limestone powder. The specimens with a density of 550 kg/m3 achieved a fcm,28d of 2 MPa. The AFC printed on the adapted equipment in the second phase had 3 density levels (see Fig. 3). The density variants of 150 kg/m3 and 280 kg/m3 have only low strength values and can only be used for filling and insulation purposes and not for structural applications. The part of the AFC with a density above 1000 kg/m3 reached a fcm,28d of 1.9 MPa and can thus be statically stressed. The mix without foam achieves an fcm,28d of 62.1 MPa.

Fig. 3. Printed specimen of adaptive foam concrete (AFC) with varying density conversion during the printing process. (1) 150 kg/m3 ; (2) 280 kg/m3 ; (3) 1070 kg/m3

4 Applications Due to its material properties in wet and dry state, the current applications of FC can be found in the fields of thermal insulation, cavity filling, compensation layers, precast

26

R. Schmid et al.

concrete and block production. The intention to use printed AFC in a formwork-free construction process, where the stability of the material allows it to be stacked in an additive process, opens up a wide range of new applications. Figure 4–5 shows application concepts for printed AFC in the precast concrete industry. The total mass of those components is to be reduced by means of load-appropriate, inserted AFC parts. The required strengths are to be achieved via geometry and adaptive material density. The respective densities vary according to the loads occurring in each case due to being walked on or the concrete being poured afterwards. Also within each of the components.

Fig. 4. (1) floor slab; (2) element slab; (3) balcony slab; (4) cavity wall

Fig. 5. Prototype of an element slab with printed AFC – 150 to 280 kg/m3

Adaptive Foam Concrete in Digital Fabrication

27

5 Conclusions and Outlook It was possible to develop a printable recipe with one composition of solids and adjustable liquid components, which covers a density variation of 150–2200 kg/m3 . (Fig. 5–6) By adjusting the accelerator dosage and adding protein foam, all requirements of 3DCP are met. Depending on the density, the concrete can be used as a load bearing or as a filling or insulating component. The technology enables the manufacture of gradient concrete structures in principle. The development of the recipe is not complete and certainly still offers potential for optimisation in the areas of increasing the compressive strength and reducing the CO2 equivalent. The developed 3D printing system for AFC enables a good basis for further research. Improvements are to be made above all in the area of the print nozzle since the surface accuracy is negatively affected by material buildup within the nozzle tip and mixing shaft.

Fig. 6. Left: foam concrete - 600 kg/m3 , Right: print concrete without foam - 2200 kg/m3

Acknowledgments. The work was carried out in cooperation with the company MAI International GmbH, which specializes in industrial products in the field of 3DCP.

References 1. Markin, V., Nerella, V.N., Schröfl, C., et al.: Material design and performance evaluation of foam concrete for digital fabrication. Materials (Basel, Switzerland) 12(15), 2433 (2019) 2. Markin, V., Krause, M., Otto, J., et al.: 3D-printing with foam concrete: FROM material design and testing to application and sustainability. J. Build. Eng. 43, 102870 (2021) 3. Bedarf, P., Dutto, A., Zanini, M., et al.: Foam 3D Printing for Construction: A Review of Applications, Materials, and Processes. (2021)

28

R. Schmid et al.

4. Mechtcherine, V., Markin, V., Will, F., et al.: CONPrint3D ultralight – production of monolithic, load-bearing, heat-insulating wall structures by additive manufacturing with foam concrete. Bauingenieur 94, 405–415 (2019) 5. Bos, F., Wolfs, R., Ahmed, Z., et al.: Additive manufacturing of concrete in construction: potentials and challenges of 3D concrete printing. Virtual and Physical Prototyping 11(3), 209–225 (2016)

Structural Design and Optimisation

Mesh Mould Prefabrication Ammar Mirjan1(B) , Jaime Mata-Falcón2 , Carsten Rieger3 , Janin Herkrath4 , Walter Kaufmann2 , Fabio Gramazio1 , and Matthias Kohler1 1 Gramazio Kohler Research, ETH Zurich, Zurich, Switzerland

[email protected]

2 Chair of Structural Engineering (Concrete Structures and Bridge Design), ETH Zurich,

Zurich, Switzerland 3 Sika Services AG, Zurich, Switzerland 4 PERI SE, Weißenhorn, Germany

Abstract. The Mesh Mould technology combines formwork and structural reinforcement into a robotically fabricated construction system. This method allows for the industrial and full-scale realisation of complex curved, steel-reinforced concrete structures without the need for conventional formwork. The paper presents a new material and cost efficient industrial robotic prefabrication process of 3D reinforcement cages (mesh elements). A novel robotic wire application process makes it possible to now fabricate mesh elements with continuous reinforcement in two commonly orthogonal directions. Thanks to the implementation of an automated structural design approach, complex structures can be dimensioned and optimised following international standards for steel-reinforced concrete structures. Furthermore, the paper presents the development of a new, appropriate concrete mixture that is stable to fill the permeable meshes with. Keywords: Digital fabrication · Reinforcement · Formwork · Robotic construction

1 Introduction With the widespread adoption of computational design methods, by simultaneously reducing the environmental impact of concrete, the construction industry is increasingly focusing on curved geometries and complex concrete structures. Nevertheless, with current means of production, complex concrete structures find only limited application in the construction industry, predominantly for representative buildings, such as museums, theatres, or churches, due to the high expense of time, labour and material involved for the construction of the custom formwork. As non-standard curvilinear mould constructions are usually produced as unique elements, they cannot be reused or recycled, making them unsustainable for a wider range of building applications. Additive fabrication processes, like 3D printing with concrete, on the other hand, currently struggle with the integration of structural reinforcement, limiting the technology to non-load bearing applications. Today, no structural concrete construction system is available on the market that enables an efficient, cost and waste effective realisation of complex concrete structures. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 31–36, 2022. https://doi.org/10.1007/978-3-031-06116-5_5

32

A. Mirjan et al.

2 Methods, Materials and Techniques The Mesh Mould technology challenges established construction methods for building geometrically complex concrete structures. The research proposes a structural concrete construction system that enables the efficient realisation of such structures without creating waste. It addresses this problem by combining formwork and reinforcement into a robotically prefabricated solution [1]. An industrial robot builds a 3D mesh structure, which acts as both formwork and structural reinforcement. The mesh elements (Mesh Moulds) are assembled and filled on the construction site. The mesh is sufficiently dense to retain the concrete and shape it. After filling is completed and the concrete cover is applied, the surface can be manually trowelled or be left as exposed shotcrete surface. The mesh structure remains inside the concrete and reinforces it structurally. Guided by a computational workflow, the construction system involves digital planning, structural design, robotic fabrication, and a specifically developed concrete material mix. The system is explained in the following. 2.1 Computational Structural Design The design of geometrically complex concrete structures is typically very demanding. In order to foster the applicability of Mesh Mould, a computational workflow has been developed to enable an efficient structural design of bidimensional concrete structure with complex geometries. The first step of this workflow consists of analysing the geometry on whether the shape is feasible in relation to curvature, dimensions, etc.. The geometry is then segmented into buildable parts and subdivided into shell elements to calculate the internal forces by means of a linear-elastic finite-element analysis. Using a sandwich model approach based on the theory of plasticity, the internal forces are then split in a combination of membrane forces acting in the sandwich core and its covers, which can be dimensioned using the well-known membrane yield conditions [2]. This code compliant design method enables to automatically dimension the required reinforcement of the structure and to generate a computational 3D model with congruent reinforcement spacing and wire diameter, defining the location, curvatures and intersection points of all wires of a Mesh Mould element. Finally, all relevant fabrication related information, as for example the robotic trajectory to place, bend and weld the reinforcement wire, is integrated into the data model. 2.2 Robotic Prefabrication of Mesh Moulds Building on previous research on the Mesh Mould [3], two major alterations are introduced to the technology. First, in order to increase efficiency and production speed, the single mobile robotic system was shifted to a multi-robotic prefabrication setup, enabling parallelisation of fabrication tasks and the handling of larger reinforcement wire diameters (up to 12 mm). Second, the topology and spatial organisation of the individual wires was modified to facilitate continuous reinforcement in both directions, fostering compliance with design codes for reinforced concrete structures. The basis for the realisation of the meshes is the computational processing of the 3D mesh geometry in a series of planar sections (to define the vertical reinforcement) and horizontal projections along

Mesh Mould Prefabrication

33

these sections (to generate the horizontal reinforcement and the intersection points of the welds). The physical realisation of the mesh is congruent to the computational method. Employing an integrated design to fabrication workflow, the slices are fabricated using a conventional stirrup bending machine to pre-bend the outline of the slices. A custom developed robotic end-effector minimally corrects the geometry of the wire by attaching a series of cross wires (Fig. 1a). A second robotic end-effector with a magnetic grasping mechanism places the slices on the production platform (Fig. 1b). Once all slices are in place, the horizontal reinforcement is attached, utilising a novel robotic wire application process developed specifically for this process. The robotic end-effector rolls a wire over the slices, pushing against the structure to bend the wire (and to generate the welding force) while contact-welding it at all intersection points (Fig. 1c). Since the structure is flexible and the appropriate contact force is crucial to achieve repeatable welds, a force control algorithm is applied to control the position of the robot along a desired trajectory. Finally, the complete element is detached from the production platform and prepared for shipping (Fig. 1d).

Fig. 1. Mesh Mould robotic prefabrication process.

2.3 Concrete Infill and Finishing Materials The filling process of the infill material differs from standard concrete casting or pumping. Nevertheless, the basis for this material should be standard concrete that can be ordered from a ready-mix concrete supplier. The concrete needs to be pumpable using a conventional concrete pump, and modified in a way that the meshes can be filled but the concrete does not flow out of the meshes. This adaptation of the concrete was achieved by the addition of special additives that increase the cohesion of the mix while maintaining a pumpable mix. After filling the concrete can be still smoothened, i.e., the setting of the concrete is not affected by the additives (Table 1).

34

A. Mirjan et al. Table 1. Exemplary basic concrete. Name

Type

Amount [kg/m3]

Cement Aggregates

CEM II A-LL 0–8 mm

350 1850

Water Admixture

Sika ViscoCrete

170 3.5

To ensure good pumpability and dense filling of the meshes, the maximum aggregate size was reduced to 8 mm and the binder content was increased. This basic mix should show a flow table spread of 50 cm [4] before modification. The additives are added to the truck mixer on site to modify the mix. After filling, the surfaces of the meshes need to be covered and smoothened. This is done by dry shotcrete application. This ensures adequate compaction of the cover, hence protection of the steel, and the fine material (max 4 mm aggregate size) facilitates smoothing of the surfaces. 2.4 Material Characterisation of the Infill Material To evaluate the potential lack of a conventional compaction for the concrete used in the core of the Mesh Mould structures, the mechanical performance and density of the concrete infill in a straight wall produced with the Mesh Mould technology were compared to standard 150 mm cubes cast inside a formwork with the same concrete mix. Six cubic samples were produced compacting with a needle, while no compaction was applied in another six cubes. The Mesh Mould wall had a length of 2.0 m, a height of 2.0 m and a thickness of 250 mm (Fig. 2a). A total of 18 cores with a diameter of 100 mm and a length of 200 mm were taken from the central part of the wall at several locations and in different directions (Fig. 2b).

Fig. 2. Compression tests of the infill concrete in different coring locations and orientations: (a) test wall before shotcrete application; (b) marked cores in the finished wall.

Mesh Mould Prefabrication

35

The results of the Mesh Mould core samples showed very similar results regardless of their location and orientation. Table 2 compares the results of the Mesh Mould wall to the reference cast cube samples. The strengths measured in different geometries were converted for comparison purposes to f cm , i.e., the compressive strength of standard cylinders with a diameter of 150 mm and a length of 300 mm. Table 2. Density and compressive strength of the core samples and cast cubes. Core samples from MM wall

Cast and compacted cubes

Cast non-compacted cubes

f cm [MPa]

43.11) CoV = 0.04

48.02) CoV = 0.04

43.62) CoV = 0.03

ρ [kg/m3 ]

2307 CoV = 0.00

2333 CoV = 0.00

2301 CoV = 0.00

1) Accounting for the size effect with a factor 1/1.03 2) Using a factor 0.82 according to EN 13791 [5] to transform fcm,cube into fcm

The scatter in the three tested configurations is very similar, indicating that the process is robust and reliable. The lack of compaction reduces slightly the density (ca. 1%) and the compressive strength (ca. 10%). However, the achieved strength even without compaction is sufficient for the intended structural applications. The identical results of the Mesh Mould technology and the non-compacted samples might open the way to a simple quality control process not relying on drilled cores. Further material tests in different production conditions, as well as structural tests should be performed to study the compliance of the Mesh Mould technology with building codes.

3 Implementation A full-scale architectural building structure was realized (Fig. 3e) to validate the results of this research. The doubly-curved wall is 4 m wide and 4 m high and placed on a reinforced foundation within an outdoor environment. The structure comprises four prefabricated Mesh Mould elements. To demonstrate the applicability of the technology in a real-world context, the onsite assembly (Fig. 3a), filling (Fig. 3b, 3c) and finishing (3d) were conducted with local construction workers and concrete from a local concrete plant. The building project successfully demonstrated some key challenges of employing the technology in an out-of-the-laboratory setting. Multiple meshes were connected in two directions, making it possible to build large structures. The two meshes on the bottom are lap sliced with the reinforcement of the foundation, highlighting that the technology can be combined with traditional concreting methods. A computational workflow from the geometry design, structural design, reinforcement dimensioning and generation, as well as the robotic prefabrication of the meshes, can be merged with traditional onsite construction tasks and machinery for the filling and finishing.

36

A. Mirjan et al.

Fig. 3. Onsite assembly (a), filling (b,d), finishing process (d) and the completed demonstrator (e), a four metre tall, doubly curved, fully load-bearing concrete structure.

4 Conclusion and Outlook The research brings forward an efficient structural concrete construction system that enables the realisation of complex concrete structures without creating waste. The demonstrator project confirms the applicability of the technology for the AEC industry, making freeform building structures a viable alternative to standardised structures, offering more freedom to construction specialists. In regard to the structural capacities of the construction system, the mechanical properties of the infill concrete used in the core of Mesh Mould structures showed no anisotropy and similar scatter as conventionally cast concrete. Despite a slightly reduced compaction, density and compressive strength of the infill concrete, the achieved strength is sufficient for the intended structural applications. The positive findings on the material characterisation should be complemented with an analysis of the structural performance of the technology. Such a study should determine if the structures can be designed identically or similarly as conventionally fabricated concrete structures. Acknowledgements. The authors would like to express their gratitude to the following people: Mattis Koh and Marvin Rüppel for developing the robotic fabrication process. Nathalie Reckinger and Marius Weber for implementing the structural evaluation process and for planning, conducting and evaluating the structural experiments. Sebastian Jehle, Phillippe Fleischmann and Michael Lyrenmann for their technical support.

References 1. Hack, N.: Mesh Mould: A Robotically Fabricated Structural Stay-in-Place Formwork System. Doctoral Thesis. ETH Zurich. (2018) 2. Marti, P.: Design of concrete slabs for transverse shear. Struct. J. 87(2), 180–190 (1990) 3. Hack, N., et al.: Structural stay-in-place formwork for robotic in situ fabrication of non-standard concrete structures: a real scale architectural demonstrator. Automation in Construction 115, 103197 (2020) 4. DIN Deutsches Institut für Normung: DIN EN 12350-5: Testing fresh concrete - Flow table test (2019) 5. DIN Deutsches Institut für Normung e. V.: DIN EN 13791: Assessment of in-situ compressive strength in structures and precast concrete components (2020)

The Production of a Topology-Optimized 3D-Printed Concrete Bridge Ticho Ooms(B) , Gieljan Vantyghem , Yaxin Tao , Michiel Bekaert , Geert De Schutter , Kim Van Tittelboom , and Wouter De Corte Department of Structural Engineering and Building Materials, Ghent University, Ghent, Belgium [email protected]

Abstract. In the last few years, the development of 3D concrete printing (3DCP) technology has flourished exponentially both in academics and the construction industry. Many problems inherent to 3DCP are already being tackled on a material level. However, in the practical realization of large-scale components there are still a lot of questions to be answered. In this study, we discuss the production process of a topology-optimized 3D-printed concrete bridge structure. As the entire process is largely different compared to the manufacturing of traditional concrete structures, the problems, workarounds, and insights gathered from this project are valuable for future constructions using 3DCP. The geometry of the bridge was based on topology optimization results and further developed through the use of parametric modelling. After careful considerations, the bridge geometry was discretized into four segments and printed as integrated formwork. Several measures were taken during the printing process in order to produce the separate sections. The assembly process entailed the handling of the printed components, the placement of reinforcement and prestressing tendons, the production of the end blocks, and the handling and joining of the printed sections. For the latter, also the process of pouring self-compacting concrete in the printed formwork is discussed and more details about the post-tensioning procedure are provided. Keywords: 3D concrete printing · Topology optimization · Concrete · Bridge · Post-tensioning

1 Introduction Concrete construction is often associated with straight and geometrically simple building components. The main reasons are the time, costs and labor directly related to the construction of formwork to shape and retain fresh concrete. However, with the rise of 3D concrete printing (3DCP) in recent years, more complex structures with organic shapes are being explored, accommodating architectural freedom and optimized topologies. Some notable 3DCP examples in the context of this paper include two pedestrian bridges in the Netherlands [1, 2] and the topology-optimized post-tensioned concrete girder by Vantyghem et al. [3]. From a practical point of view, we present a new project involving the production of a structurally optimized, 3D-printed bridge structure in concrete (see Fig. 1). © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 37–42, 2022. https://doi.org/10.1007/978-3-031-06116-5_6

38

T. Ooms et al.

In this paper, each step of the project from beginning to completion is described with the main focus on the manufacturing of the bridge. The level of optimization and material reduction compared to a traditional concrete structure is out of the scope of this paper. The aim of this report is to share the showcased 3D-printing construction methodology and obtained insights with peers in academia and industry.

Fig. 1. The completed post-tensioned concrete bridge, manufactured with 3DCP.

2 Design The design of the bridge structure is a second iteration of the 3D-printed topologyoptimized (TO) post-tensioned (PT) concrete girder from Vantyghem et al. [3]. With a substantially wider top surface compared to its predecessor, the structure can be categorized as a bridge. The new design incorporates two longitudinal, TO, PT, concrete girders, which are integrated in the bridge deck as one continuous shape (see Fig. 1). The single-span bridge is simply supported at the four corners and the deck is designed for a uniformly distributed live load of 5 kN/m2 . Consequently, the PT girders are inclined outwards for approximately 20°, to add some transverse prestress.

Fig. 2. The design of the bridge: (a) topology optimization [4] and (b) extrapolated 3D model

The topology of the girders is based on a result from the work of Amir and Shakour [4]. Similarly, the resulting 2D image (see Fig. 2a) is used as a template for the base geometry and converted into a 3D shape (see Fig. 2b) by means of parametric modelling in Rhinoceros/Grasshopper. This process facilitates adjustments in the pre-design stage.

The Production of a Topology-Optimized 3D-Printed Concrete Bridge

39

The length of the new bridge is 4.75 m and the width varies between 1.4 m (deck) and 1.9 m (girders). The bridge deck has a variable thickness with a minimum of 0.12 m at the mid span to 0.22 m at the edges, while the depth of the beams is 0.6 m. The end blocks are added separately and are not included in the printing process. Apart from the quality control, the benefit of casting these end blocks is having flat and smooth surfaces as supports for the final installation of the bridge (see Fig. 2b).

3 Manufacturing In order to the manufacture the bridge structure, only the outer shell of the bridge was 3D-printed with concrete, serving as integrated formwork. The anchorage blocks were cast separately. After the assembly of all components, self-compacting concrete was used to fill the printed shell. Finally, post-tensioning was applied, and the completed structure was flipped over and placed on temporary supports. In the following sections, more details and comments are provided for each step in the manufacturing process. 3.1 3D Concrete Printing The complexity of the bridge structure lends itself to benefit from the use of 3DCP. In this project a novel 3DCP technology with a fast-setting material was used to cope with the steep overhangs present in the geometry of the bridge and to allow for increased buildability. However, still some measures were taken to facilitate the printing process and reduce the risk of premature failure, such as numerical analysis using previously developed simulation tools [5, 6] and 3D-printed spacers as temporary, lateral supports. Printing Code. The shell geometry, defining the bridge topology, was subdivided into four segments at specific locations in order to limit the amount of additional support pieces. The segments were printed without infill pattern, as the shell will be filled with concrete after the assembly. In contrast to the design of the bridge, the openings in the optimized girders are also printed (see Fig. 1 and Fig. 2b). Due to the applied 3DCP technology, it was more efficient to have one continuous print path instead of separated loops in the print path. As a consequence, the extra printed material needs to be cut out of the final structure, but also gives additional support to the overhangs in the model. Print Material and Technology. The patented twin-pipe pumping (TPP) system [7] was used to print the bridge segments. A cement-based mixture (without accelerator) and a limestone-based mixture (with a high dosage of accelerator) were delivered via two pumps and combined by using a helical static mixer right before extrusion. As these two streams move through the mixer, the mixing baffles divide the mixtures with an exponential increase in stratification. The accelerator in the limestone-based mixture now interacts with the cement in the cement-based mixture, leading to a fast stiffening rate and good shape stability of the printed segments. More details about the TPP system and the mix-proportions can be found in previous work of the authors [8].

40

T. Ooms et al.

Fig. 3. (a) Laterally supporting spacers and (b) bridging supports

Temporary Spacers and Other Supports. Based on numerical results and as an extra safety measure, additional lateral supports were desirable for the long straight walls of the prints to prevent early-age failure due to elastic buckling during printing. Therefore, 3D-printed spacers in PLA were inserted during the printing process at different levels to connect opposing wall segments to each other (see Fig. 3a). This was done manually and largely depends on the used print technology and material. In addition to these spacers, special bridging supports (see Fig. 3b) were purposely designed and 3D-printed in PLA to add underneath bridging sections in the print path (i.e. where the print path is not supported by an underlying layer of deposited material). This could have been avoided by subdividing the segments even further, however this would lead to more components and that requires more space to print the extra parts.

3.2 Transportation Once all segments were printed, they were transported carefully to the assembly location in the laboratory with a gantry crane. A lot of attention went into the proper way to clamp the printed parts with timber studs, as the highly irregular shape of each segment shifted the center of mass outwards. This process required meticulous movements in order to avoid cracking in the unreinforced and vulnerable segments. A numerical analysis was carried out to check whether the induced tensile stresses due to the handling did not exceed the expected tensile strength of the material. 3.3 Assembly Assembly and Reinforcement. The transportation was done in sequence with the installation of reinforcement. The reinforcement cage for the bridge deck as well as additional steel rebars inside of the struts of the girders (see Fig. 4) were added after each segment was placed on the assembly location. It was not feasible to construct the reinforcement cage as a whole, therefore overlaps were created between the consecutive reinforcement cages in the segments. Furthermore, the prestressing tendons were inserted and in order to ensure a parabolic tendon profile for the bottom strands, the

The Production of a Topology-Optimized 3D-Printed Concrete Bridge

41

Fig. 4. Reinforcement for the deck (blue) and struts (green), and prestressing tendons (red)

tendons were held in place with a steel wire in between the segments. A mortar adhesive was applied between the printed segments to avoid stress concentrations. Concrete Casting. According to the intended design, the hollow spaces of the printed elements were filled with concrete. The type of infill material and its properties were chosen to meet the requirements needed for this project. In this case, the printed concrete forms a closed space which does not allow many places to insert the cast concrete or apply a vibration rod. Furthermore, the narrow dimensions of the printed struts and ties in combination with the high surface area of the curved layers pose a risk of clogging during casting. Additionally, the curvature between the layers needs to be properly filled to obtain an optimal bond between the cast and printed material. Based on these concerns, only a self-compacting concrete (SCC) with a high flowability could meet these criteria. An SCC containing a maximum aggregate size of 8 mm was applied. The bridge was placed upside down in a horizontal position for casting. A vertical placement is not desirable as this would increase the concrete cast head and lead to potential cracking or blow out of the printed material. The voids between the formwork supports at the bottom of the bridge were filled with sand to fully support the bridge deck. Large openings of 68 mm were drilled out at the top of the girders to pump (under pressure) the SCC in the bridge at several inlets while smaller holes were drilled at specific locations to prevent air pockets. In addition to the gravitational flow, an increased pressure was carefully applied to ensure complete filling of the printed shell. During casting no visual cracking of the formwork is observed. After casting, the bridge is covered with a tarp to prevent external environmental influences. Although a high binder content is obtained (450 kg/m3 ) for the SCC, no specific measurements are taken to prevent thermal cracking. 24 h after casting, the printed shell is re-examined and no visual leakage at the interlayers is observed. Post-tensioning and Installation. After the casting procedure and sufficient hardening period of 14 days, post-tensioning was applied in a step-wise manner to the pre-installed 12.5 mm prestressing tendons (see Fig. 2 and 4). First, the straight tendons in the bridge deck were tensioned to a prestress of 40 kN. Afterwards, the curved tendons in the girders were also tensioned to 60 kN. In both cases the effective prestress values take into account

42

T. Ooms et al.

initial anchorage slip, however the losses due to sequential tensioning are not examined. For the final installation of the bridge structure, the bridge needed to be flipped around its longitudinal axis, as the bridge was assembled upside down for convenience. Finally, the bridge was installed in its final orientation on temporary supports outside the lab. More information and images are available on the project page [9].

4 Conclusion In this paper, the manufacturing process of a complex 3D-printed concrete bridge is briefly reported. This academic bridge project shows that it is possible to produce a topology-optimized 3D-printed bridge, involving prefabrication of 3D-printed hollow segments, assembly of segments and placement of reinforcement and prestressing tendons, filling the segments with SCC, and finally post-tensioning. Acknowledgements. We would like to thank the technical staff and other colleagues of the Magnel-Vandepitte laboratory at Ghent University for their help in the presented project. The authors are thankful for the valuable discussions with Oded Amir and grateful for the support of Freyssinet Belgium for providing the post-tensioning components.

References 1. Salet, T.A.M., Ahmed, Z.Y., Bos, F.P., Laagland, H.L.M.: Design of a 3D printed concrete bridge by testing. Virtual Phys. Prototyping 13(3), 222–236 (2018) 2. Langste 3D-geprinte brug ter wereld geopend. https://www.bam.com/nl/pers/persberichten/ 2021/9/langste-3d-geprinte-brug-ter-wereld-geopend. Accessed 25 Jan 2022 3. Vantyghem, G., De Corte, W., Shakour, E., Amir, O.: 3D printing of a post-tensioned concrete girder designed by topology optimization. Autom. Construct. 112, 103084 (2020) 4. Amir, O., Shakour, E.: Simultaneous shape and topology optimization of prestressed concrete beams. Struct. Multidiscip. Optim. 57(5), 1831–1843 (2017). https://doi.org/10.1007/s00158017-1855-5 5. Ooms, T., Vantyghem, G., Van Coile, R., De Corte, W.: A parametric modelling strategy for the numerical simulation of 3D concrete printing with complex geometries. Add. Manuf. 38, 101743 (2021) 6. Vantyghem, G., Ooms, T., De Corte, W.: VoxelPrint: A Grasshopper plug-in for voxel-based numerical simulation of concrete printing. Autom. Construct. 122, 103469 (2021) 7. Tao, Y., De Schutter, G., Van Tittelboom, K., Lesage, K.: Method for layer-by-layer deposition of concrete. WO 2021/214239 A1 (2021) 8. Tao, Y., Rahul, A.V., Lesage, K., Van Tittelboom, K., Yuan, Y., De Schutter, G.: Mechanical and microstructure properties of 3D printable concrete in the context of the twin-pipe pumping strategy. Cement Concrete Compos. 125, 104324 (2021) 9. OptiBridge: https://www.ugent.be/ea/structural-engineering/en/research/projects/allprojects/ optibridge2.htm. Accessed 25 Jan 2022

Injection 3D Concrete Printing (I3DCP) Combined with Vector-Based 3D Graphic Statics Yinan Xiao1(B) , Noor Khader1 , Aileen Vandenberg2 , Dirk Lowke2 , Harald Kloft1 , and Norman Hack1 1 Institute of Structural Design, Technische Universität Braunschweig, Pockelsstraße 4,

38106 Braunschweig, Germany [email protected] 2 Institute of Building Materials, Concrete Construction and Fire Safety, Technische Universität Braunschweig, Braunschweig, Germany

Abstract. This paper introduces the combination of Injection 3D Concrete Printing (I3DCP) with Vector-based 3D Graphic Statics (V3DGS). I3DCP is a technique that robotically injects concrete into a non-hardening carrier liquid which acts as a supporting structure for the printed strands. The printing path of I3DCP is exactly aligned with the spatial stress trajectories, which can be simply treated as strut-and-tie networks in the truss system. V3DGS matches the core of I3DCP, which becomes an ideal system for practically operating the strut-and-tie networks at the design phase, as well as intuitively improving the structural performance. Accordingly, grounded on V3DGS, the design framework called Combinatorial Equilibrium Modelling (CEM) lends itself particularly effective to be integrated with the design process. In terms of fabrication, a modularization strategy is described for fabricating objects larger than the suspension vessel. Finally, a segmentally printed and assembled demonstrator is presented. Keywords: Injection 3D Concrete Printing · Vector-based 3D graphic statics · Combinatorial Equilibrium Modelling · Fabrication-informed design

1 Introduction The emergence of various concrete 3D printing technologies has altered fundamentally our design thinking and building processes in the last few years. Despite the numerous advantages 3D concrete printing technologies have introduced, there exist certain fabrication constraints that need further exploration so that 3D printed concrete structures can be produced efficiently and utilized optimally [1]. To date, the most significant challenges facing 3D concrete printing technology include lengthy print time, monolithic volumes, and limited geometrical expression. To address these challenges, this paper proposes a novel technology, called Injection 3D Concrete Printing (I3DCP). In general, I3DCP is defined as the process of robotically injecting one fluid A into another fluid B, where one of the fluids is concrete. In our case, we injected concrete into a non-hardening carrier liquid, which acts as a supporting structure for the printed strands [2]. Compared © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 43–49, 2022. https://doi.org/10.1007/978-3-031-06116-5_7

44

Y. Xiao et al.

to layer-based extrusion techniques, I3DCP not only provides a high building rate but also enables spatial printing of complex geometries, in particular, spatial lattice concrete structures that are not feasible with conventional layered concrete printing [2]. As the I3DCP technology allows for creating material-efficient structures, its structural performance is essential for a successful implementation. Since the forces in the spatial truss system are, most of the time, simplified as axial forces with hinge joints among the elements, the stress-trajectories could be treated as strut-and-tie networks. Thus, a synthetic structural analysis and design method, known as vector-based 3D graphic statics [4] presents itself as an ideal system for providing an intuitive and practical methodology for architects and engineers to effectively create various spatial truss-like lattice structures with high structural performance. This paper explores the potential to combine Injection 3D Concrete Printing (I3DCP) and Vector-based 3D Graphic Statics (V3DGS) through the fabrication of a demonstrator. In addition, to address scale limitations imposed by the size of the suspension vessel, modularization techniques are developed, and jointing principles are investigated in this particular project.

2 State of the Art

Fig. 1. (a) Rapid liquid printing using urethane or rubber, MIT, 2017; (b) Prototypical application of I3DCP, TUBS, 2019.

The underlying concept of extrude constructing material into another liquid medium was first tested using a technique called “buoyant extrusion” by a research team from Princeton University in 2014 [5]. This concept was further developed into large-scale production by researchers from the Massachusetts Institute of Technology (Fig. 1a) [6]. Experiments with concrete printing in a gel were followed by the French startup company Soliquid. Simultaneously, experiments with concrete printing in ground limestone suspension were presented at Technische Universität Braunschweig (TUBS) (Fig. 1b) [2]. Recently, an in-depth investigation regarding the underlying physics, as well as the first large demonstrators, were presented in a joint publication of Soliquid, TU Braunschweig, and CY Cergy Paris University [3]. Current research at TUBS is focusing on improving the quality of the printing process driven by specific design methods and generating large-scale structures consisting of printed segments.

Injection 3D Concrete Printing (I3DCP)

45

3 Experimental Setup and Design Method In the next sections, the fabrication process of a demonstrator applying the combination of the design method grounded on vector-based 3D graphic statics and the I3DCP technique is described, as well as the investigation of the joint method among printed segments. The procedures were developed in the research studio “Digital Building Fabrication Studio” (DBFS) by researchers in the ITE institute with a group of students. 3.1 Robotic Setup

Fig. 2. The end-effector for I3DCP consists of a pinch valve connected to the concrete hose (not pictured), printing nozzle, and a pressurized air tube.

For fabrication, a UR10 type Universal Robot set-up is used. The software set-up involves the plugin Robots [7] that manages the communication between the custom Python program that orchestrates the specific commands for the end-effector and the motion of the robots, the graphical user interface (GUI) program Rhinoceros 3D’s Grasshopper which serves as the platform of the Robots plugin, and the UR robots, which are controlled by the URScript program. The control system runs on a laptop computer that is connected to the UR10 robot over TCP/IP socket. The printing tool of I3DCP is set as the end-effector on the UR10 (Fig. 2). It consists of a pinch valve that is connected to the concrete hose on the top, the printing nozzle on the bottom, and a pressurized air tube on the side. The condition of the pinch valve (open/closed) depends on the air pressure in the air tube, which is controlled by a magnetic valve. Both the magnetic valve and the concrete pump are integrated with the UR robots through the internal Inputs and Outputs (I/O) on the robot’s controller. 3.2 Materials The injection material in this research studio is a fine-grained ready-mix concrete (Nafufill KM 250, MC Bauchemie) with a maximum aggregate size of 2 mm. The dry mix compound is mixed with 4 L of water per 25 kg of dry mix using a pugmill mixer. The density of the ready-mix is 2,100 kg/m3 [3].

46

Y. Xiao et al.

As a carrier medium, a ground limestone suspension (Lhoist) is chosen because of the ease of its production with normal concrete mixing devices, its inert material properties, and its low cost. The density of the suspension is 1,850 kg/m3 [3]. 3.3 Design Method

a

b

Fig. 3. (a) Two options under the same force F1 . From the force diagram (F*), the upper option contains much less inner force which is more effective. Meanwhile, the e1 and e2 elements in the upper option have no structural function, thus they can be optimized; (b) CEM Samples.

The design and structural analysis method applied in this paper mainly relies on Vectorbased 3D Graphic Statics (V3DGS). The traditional graphic statics is a purely geometrical design and analysis method for planar trusses under static equilibrium. Based on vectors and projective geometry, it provides visual information of the reciprocal relationship between form and forces. [8] Thanks to the advancements in Computer-Aided Design (CAD), graphic statics has been recently extended to three-dimensional space following various approaches [9, 10]. Due to its intuitive advantages and seamless match with the I3DCP technology, V3DGS is adopted for this research. By observing the force diagram and operating reciprocally the form diagram, good control has been achieved for both the force and the load path in the strut elements (Fig. 3a). Hence, bringing in V3DGS makes it possible to considerably reduce the number of the redundant elements in the 3D lattice structures, and simultaneously ensures the equilibrium of the global system under certain load cases. Furthermore, V3DGS not only improves the structural utilization but also optimizes the structure based on compliance with certain constraints of the I3DCP technology. In terms of the design tool, a novel design framework called Combinatorial Equilibrium Modelling (CEM) [11] is chosen. Based on the V3DGS, the CEM tool brings in a new dimension - topological graph, to the traditional form and force diagrams combination. By doing that, series of designs with similar or even the same topological structure could be generated in a very short time (Fig. 3b). Different generations of designs are well organized in the CEM tool, remarkably shortening the design progress.

Injection 3D Concrete Printing (I3DCP)

47

4 Investigation of Joints and Fabrication Procedure 4.1 Investigation of Modularization Within the I3DCP approach, the available work envelope of the robot is regarded as a limiting constraint. Thus, a modular construction system should be considered to fabricate larger spatial structures by printing segments separately and subsequently assembling them in a specific method that ensures the whole system is still in global equilibrium. This concept was involved already in the design phase during the research studio. The students were requested to provide a spatial truss structure that could be split into segments in the size that is printable with the robotic set-ups in the lab. 4.2 Design and Fabrication Procedure

Fig. 4. (a) Form diagram of the table; (b) One leg of the table as a segment; (c) The process of assembling the printed segments into a whole table.

Fig. 5. The assembled lightweight concrete table demonstrator.

Driven by the constraints mentioned above, the final demonstrator was designed as a symmetric three-leg table (Fig. 4a). Each leg, as a segment, is connected with the other two, as well as the tabletop separately. The average printing duration of one leg is 2 min and 49 s. The average weight is just 2.95 kg. Finally, a lightweight concrete triangular table under only compression, with dimensions of 97 × 90 × 58 cm and a weight of 8.85 kg (main part except for the tabletop) is successfully achieved (Fig. 5).

48

Y. Xiao et al.

5 Conclusions and Outlook The investigated workflow and prototypical applications displayed the advantages of integrating Injection 3D Concrete Printing (I3DCP) with a structurally informed design process based on Vector-based 3D Graphic Statics (V3DGS). Utilizing the Combinatorial Equilibrium Modelling (CEM) tool for structural design increased the structural efficiency and precision of the structures in a geometrical approach. Considering the scale constraints of the robotic set-ups and the suspension vessels, a final demonstrator was printed and assembled in a modular system. To further improve the efficiency and precision of I3DCP the following challenges in terms of fabrication-constraint interaction have to be overcome. First of all, early experiments showed that not all lattice structures can be printed with satisfactory results. Fabrication and structural constraints need to be correlated, and investigated further, including e.g. the maximum length of printed strut elements, the overall number of struts connected at one node, the orientation and position of the struts in the suspension, and the minimum angles between the struts, to name a few. Considering the work envelope of the robot and the vessel size, methods to optimize the boundary of the printed modular segments and the connection strategy among them are becoming an important topic in future research. Furthermore, the assembling process also affects the original design at the boundaries. Thus, for complex, multi-component structures, assembly logics will require an integrated recalculation and hence redistribution of the force in the whole structure. Lastly, to avoid buckling issues in the struts, specific strategies have to be investigated concerning certain concrete properties. Here, the introduction of special fibers for reinforcement during printing would reduce the risk of the entire structure collapsing under shear or bending forces. These topics related to structural design will be combined again with V3DGS for optimizing the geometry to match the restrictions of the I3DCP in future research.

References 1. Buswell, R.A., De Silva, W.L., Jones, S.Z., Dirrenberger, J.: 3D printing using concrete extrusion: a roadmap for research. Cem. Concr. Res. 112, 37–49 (2018) 2. Hack, N., Dressler, I., Brohmann, L., Gantner, S., Lowke, D., Kloft, H.: Injection 3D concrete printing (i3dcp): basic principles and case studies. Materials 13(5), 1093 (2020) 3. Lowke, D., Vandenberg, A., Pierre, A., Thomas, A., Kloft, H., Hack, N.: Injection 3D concrete printing in a carrier liquid-Underlying physics and applications to lightweight space frame structures. Cement Concr. Comp. 124 (2021) 4. D’Acunto, P., Jasienski, J.-P., Ohlbrock, P.O., Fivet, C., Schwartz, J., Zastavni, D.: Vectorbased 3D graphic statics: a framework for the design of spatial structures based on the relation between form and forces. Int. J. Solids Struct. 167, 58 (2019) 5. Johns, R.L., Kilian, A., Foley, N.: Design approaches through augmented materiality and embodied computation. In: McGee, W., de Ponce Leon, M. (eds.) Robotic Fabrication in Architecture, Art and Design 2014, pp. 319–332. Springer, Cham (2014). https://doi.org/10. 1007/978-3-319-04663-1_22 6. Hajash, K., Sparrman, B., Guberan, C., Laucks, J., Tibbits, S.: Large-scale rapid liquid printing. 3D Print. Additive Manuf. 4(3), 123–132 (2017) 7. GitHub – visose/Robots. https://github.com/visose/Robots. Accessed 15 Jan 2022

Injection 3D Concrete Printing (I3DCP)

49

8. Maxwell, J.C.: On reciprocal figures, frames and diagrams of forces. Earth Environ. Sci. Trans. R. Soc. Edinburgh 26(1), 1–40 (1870) 9. D’Acunto, P., Ohlbrock, P.O., Jasienski, J.P., Fivet, C.: Vector-based 3D graphic statics (part I): evaluation of global equilibrium. In: Proceedings of the IASS Annual Symposia, pp. 1–10. International Association for Shell and Spatial Structures, Tokyo (2016) 10. Akbarzadeh, M., Van Mele, T., Block, P.: On the equilibrium of funicular polyhedral frames and convex polyhedral force diagrams. Comput. Aided Des. 63, 118–128 (2015) 11. Ohlbrock, P.O., D’Acunto, P.: A computer-aided approach to equilibrium design based on graphic statics and combinatorial variations. Comput. Aided Des. 121 (2020)

3DCP Structures: The Roadmap to Standardization Jolien Van Der Putten(B) , Maartje J. Hoogeveen, Marijn J. A. M. Bruurs, and Hans L. M. Laagland Witteveen+Bos NV, Blaak 16, Rotterdam, The Netherlands [email protected]

Abstract. During the last decades, 3D concrete printing (3DCP) is evolving rapidly. This newly developed technique allows the manufacturing of structures layer-by layer, based on a virtual model and without human intervention. The elimination of formwork, the reduction in labor time and cost and the increased architectural freedom are generally accepted as the major benefits compared to traditional concrete. At this moment, there are multiple examples of large-scale 3DCP projects worldwide, but the manufacturing process is still hindered from developing its full potential as developments are currently limited to trial and error related to large knowledge gaps in fundamental understanding, experimental methods and predictive models. Additionally, as concrete printing deviates on almost every aspect from mold-casted concrete, new standards with regards to the concrete mix design, the manufacturing process itself and the structural performance are required. All these categories should contain technical recommendations to ensure the structural stability, and should exclude the effect of the applied print parameters or cementitious material. To encounter the latter, researchers apply nowadays the ‘Design by Testing’ principle, with a project-specific testing program. This paper shows how structural components are designed based on the ‘Design by Testing’ principle and presents in addition an outlook for the legislation and standardization of future projects. Keywords: Standardization · Structural behavior · Full scale testing · Design by Testing

1 Problem Statement Worldwide, construction industry is facing major challenges. Urbanization, energy transition, a growing lack of natural resources, climate change, etc. are only a few factors that will affect the built environment. Investigations [1] recently showed that contractors have to realize more than 3600 houses per day by 2050 to meet the current needs. With these looming developments, the pressure on innovation is increasing in order to meet the demands of the future and new, more efficient and eco-friendly methods needs to be developed. Based on other industries, digital disruption could potentially become a game changer to meet these major challenges in the built environment and 3D concrete printing (3DCP) is one example of a promising manufacturing technique. The technique © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 50–55, 2022. https://doi.org/10.1007/978-3-031-06116-5_8

3DCP Structures: The Roadmap to Standardization

51

requires less material, no formwork, it reduces the amount of waste and construction time and allows more architectural freedom. Before a new construction technique can become a compelling alternative, some challenges have to be overcome and one important challenge is legislation. Current standards as for example Eurocode 2, do not consider the layered manufacturing process nor the anisotropic behavior of the printed specimens. Although there are at this moment numerous materials, equipment, and print processes available, clear requirements for materials, processes, calibration, tests, or general accepted standards are still missing. In the long run, this is not conducive to the quality assurance of 3DCP products. Therefore, it is necessary to create a full-scale series of design criteria, construction guidelines, and standard practices for 3DP in construction [2], which would reflect industry knowledge, help stimulate research, and promote implementation. This paper highlights a possible solution for this problem and shows how the ‘Design by Testing’ (DBT) principle can speed up the integration of 3DCP in construction industry and how the cooperation between research and industry can stimulate this process. In the future, the establishment of material standards, manufacturing processes, and structural designs will be essential before printed components can be used in buildings and the construction industry [3].

2 From Trial to Standards The first step in the quality assurance of new innovation techniques can be found in the ‘Design by Testing’ (DBT) principle, described in Eurocode 0 - NEN EN 1990 [4]. This method stipulates an elaborated determination of different material properties based on standardized tests, followed by various testing programs to map the structural parameters and a simulation of the real-life conditions and behavior of the construction. In general, the DBT-principle distinguishes seven different testing categories, indicated in Table 1. Table 1. Testing categories within the ‘Design by Testing’ principle [4] Category

Description

Testing category

a

Determination of characteristic and design values with their corresponding resistances

Structural design

b

Determination of material properties

Material properties and durability tests

c

Implementation of load models

Structural design

d

Implementation of resistance models

Structural design

e

Quality control

Construction test

f

Real scale testing

Construction test

g

Control testing of the structural element Construction test

Based on this principle, a flowchart for 3DCP has been provided and demonstrates a manner to cover all the different categories of the DBT-principle. The different testing

52

J. Van Der Putten et al.

categories are divided in a chronological order and every category has its corresponding test program to support the design phase. Important to note, in case of 3DCP there is a high interdependency between the different design phases. If little knowledge is available, more testing will be required as suggested by Fig. 1.

Fig. 1. ‘Design by Testing’ (DBT) flowchart for 3DCP

2.1 Material Design The development and characterization of a new printable material is time consuming and is mostly performed at research institutions. Compared to traditional concrete, 3D printed materials have to fulfill more specific requirements. Standard characterization methods like for example slump flow measurements cannot be used as they do not consider the behavior during and after extrusion nor the influence of different print process parameters. More advanced test methods related to the extrudability, buildability, hardening rate, structuration rate, etc. of the material should be taken into account in the fresh state characterization. During hardening, a distinction has to be made between the short and long-term behavior, or in other words the mechanical properties and the durability. Standardized test methods (i.e. compressive and flexural test methods, carbonation resistance tests, etc.) can be applied only if they are able to map the influence of the interlayers between subsequent printed layers. Intensive testing in different directions and sufficient sample size conform Eurocode 2 are additional requirements to map the orthotropic behavior (Fig. 2). After finishing this primary test flow chart, industry can implement these results and progress rapidly in future projects with the same material. 2.2 Structural Design After mapping the material properties, upscaled experiments are necessary to investigate the structural behavior of the construction. When little or nothing of the structure is known, a (scale) model, also known as Mock-up, has to be constructed. If only specific questions are unknown, e.g. the connection between different components, the model

3DCP Structures: The Roadmap to Standardization

53

Fig. 2. Flowchart of material test implementation [3]

can be downsized. Here, also the statistical reliability is important (Fig. 3A). An example of a Mock-up test executed during the design process of the bicycle bridge in Gemert is visualized in Fig. 3B. This test determines the behavior of the construction in ULS and SLS and registers the deformations during loading. When the deformations exceed the safety margin, the original design has to be improved. At the end of this testing phase, a duplicate of the final structure is often tested to determine the maximum load the structure can withstand.

Fig. 3. Flowchart of structural design implementation (A) and an example of a Mock-up test executed during the design process of a bicycle bridge in Gemert (B) [3]

2.3 Execution Design When the Mock-up test confirms the expected structural capacity and the reliability of the structure, the design can be printed, assembled and tested in-situ (Fig. 4A). These tests are non-destructive; the applied load is lower than the ULS load to avoid structural damage and are applied to check whether the quality of the final product corresponds with the predetermined assumptions and calculations. Similar tests are performed on a regular base to verify the capacity of the structure as a function of time. An example of the latter is represented in Fig. 4B, where the original structure on site is loaded by multiple containers filled with water. The holders were positioned in such a way that 100% of the ULS bending moment was achieved and the deformations were measured.

54

J. Van Der Putten et al.

Fig. 4. Execution test implementation (A) and final construction test of the footbridge in Gemert (B) [3]

2.4 Durability Design One of the most difficult aspects to guarantee is the long-term behavior. Elaborated testing procedures and results are currently missing, but can be counteracted by an intensive monitoring procedure of the construction on site. For example, considering the prestressed footbridge in Gemert, creep and shrinkage are considered as crucial parameters as they can result in a decrease of the prestressing force and the structural performance. Additionally, the deformations are extremely important and should be considered independently within each design or application. The flowchart represented in Fig. 5 provides a testing protocol in terms of durability design.

Fig. 5. Implementation of the durability tests [3]

2.5 Use Phase The start-up phase is the final step in the testing procedure and the structure will become asset management for the owner. When there is already enough confidence in the design and the structural behavior, one can decide to start directly with this phase and the manufacturing of the components on real scale. Unfortunately, irrespective of the design, intensive durability testing will remain necessary.

3DCP Structures: The Roadmap to Standardization

55

3 Implementation Further Ahead When the ‘Design by Testing’ protocol is sufficiently approved, the knowledge of the empirical data can be gathered into standardized documents, so-called technical recommendations. When developing these documents, it is crucial to make a valuable distinction between different categories (i.e. material, design and/or execution) and define for each category the level of abstraction (i.e. generic or specific). Generic standards for example only describe the essential broad outlines of the safety requirements that must be guaranteed. This implies a high degree of freedom and ensures a direct implementation of the innovation. However, a concrete handling manual is missing. Specific standards on the other hand are very clear and extensively written and results in a standard that offers solutions for a limited amount of 3DCP projects. In the most specific form, there can be for example a ‘Gemert-standard’, which serves as a standard for common like projects. The advantage is that nothing is unknown. The drawback is that it is only useful for duplicates of that bridge and offers no guidance for new deviating 3DCP projects. Independent organizations, like the Dutch standardization institute (NEN) or CROW normally guide the development of these standards and promote the upscaling towards international accepted documents.

4 Conclusions Standards are prescribed agreements and serve as a tool to assure safety and structural reliability in case of new innovative techniques. Standards are most of the time initiated by the market itself as they are necessary to create a consensus and give permission to apply new manufacturing techniques in reality. A first step in the standardization process is the Design by Testing principle, where elaborated testing procedures on different levels serve as a proof for the reliability of the materials and the construction technique. As described for the footbridge in Gemert, this complete process is very time consuming when executed for the first time, but the more knowledge is gathered, the faster the execution process will be in the future. This information is a first step in the development of national and international accepted standards.

References 1. Bertollini, V.: Here’s What Building the Future Looks Like for a 10-Billion-Person Planet (2018). https://redshift.autodesk.com/building-the-future/ 2. Ning, X., et al.: 3D printing in construction: Current status, implementation hindrances, and development agenda. Adv. Civ. Eng. 2021, 6665333 (2021) 3. Diks, T.: The roadmap to standards for 3D concrete printing: Research on the interplay between technological and legislative developments. In: Proceedings of the Construction Management and Engineering. University of Twente (2019) 4. ANB:2021, N.E.: Eurocode 0: Grondslag voor constructief ontwerp - Nationale bijlage. Belgisch Instituut Voor Normalisatie, Brussel (2021)

Binders and Aggregates 1: Aggregates

Mix Design for a 3D-Printable Concrete with Coarse Aggregates and Consideration of Standardisation Markus Taubert(B)

and Viktor Mechtcherine

Institute of Construction Materials, TU Dresden, 01062 Dresden, Germany [email protected]

Abstract. Driven by promising efficiency gains, 3D concrete printing is rapidly evolving. To facilitate the transfer of the findings into construction practice, it is recommended to develop concretes within the specifications of standards. It is shown, that limiting the fine particle content is a major challenge. A generalizable, numerically supported application of the particle size distribution according to andreasen and andersen is proposed as a solution. Its application in printing tests proves good extrudability and buildability of standard-compliant concrete with 16 mm largest grain size and a fine particle content of 500 kg/m3 . Keywords: Digital concrete · 3D concrete printing · Coarse aggregates · Particle size distribution · Standards · CONPrint3D

1 Introduction The additive manufacturing of concrete structures creates a high potential for efficiency increases with regard to labour and production times, the use of materials and the costs [1]. The majority of the digital fabrication technologies use extrusion of fresh concrete as deposition method [2], while nearly all mixtures only contain grain sizes < 4 mm and large proportions of fine particles (< 0.125 mm). A high paste content enables low inner friction reorganisation of the granular structure. This is measurable as low shear stresses [3]. On the other hand, it favours shrinkage, which leads to cracking. To ensure durability, the fine particle content for ordinary construction concretes is limited by standards [4]. Finally, fine-grained compositions require higher binder content, which also leads to poor environmental performance of the material. The aim of this research is to develop a concept for printable concrete, that meets the basic requirements of 3D printing (pumpability, extrudability, buildability) as well as the rules of European and German standards. Such concrete mixes would reduce considerably obstacles to the transfer of new technology into construction practice. The approach is developed by compiling the specifications of standards and presents a generalizable mix design. The article also presents first test results proving the validity of the suggested concept in the framework of CONPrint3D approach [1].

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 59–64, 2022. https://doi.org/10.1007/978-3-031-06116-5_9

60

M. Taubert and V. Mechtcherine

2 Standardisation The standards valid in Europe and Germany were taken as a base. These are in particular the European standard EN 206 as well as the national standard DIN 1045-2, which offers application rules for EN 206. EN 206 defines exposure classes and specifies classifications of fresh and hardened concrete properties. For cements, reference is made to EN 197-1, which describes 27 standard cement types, organised into five main types. These differ in the degree of substitution of the clinker phase and the types of substitutes. EN 197-5 added the sixth main type for composite cements with a clinker substitution of up to 65%. In addition, Portland composite cements were supplemented by the standard cement type CEM II/C-M, which allows a clinker substitution of up to 50% with any combination of substitutes. For aggregates, EN 206 provides for “natural normal aggregates, heavy aggregates as well as blast furnace slag according to EN 12620, light aggregates according to EN 13055 [as well as] recovered aggregates”. Since mainly natural round and crushed aggregates are used, the delimitation of these specifications is not discussed here. For 3D printable concretes limitations of the fine particle content are crucial. Fine particles are defined as the sum of solid mixture constituents with a particle size below 0.125 mm. They include cement, mineral powders and supplementary cementitious materials such as fly ash. The standard limits the absolute fines content depending on the cement content, the largest particle size, the compressive strength to be achieved and the exposure classes [4]. The specifications are summarized in Table 1. Table 1. Maximum permissible fine particle content according to DIN 1045–2 [4].

Compressive strength up to C50/60 from C55/67 class according to EN 206 Exposure class XF, XM other all ≥ 500 ≥ 400 any ≤ 400 Cement content [kg/m³] ≤ 300 ↔ ↔ linear interlinear interMax. permissible fine par400 500 500 600 polation polation ticle content [kg/m³] 550 if largest grain = 8 mm + 50 kg/m³ + 50 kg/m³

3 Mix Design The future oriented character of 3D concrete printing should be reflected in the material. To reduce environmental impact, a cement of the new standard cement type CEM II/C-M is chosen. In the binder, provided by Holcim (Deutschland) GmbH, half of the clinker is substituted in equal parts with granulated blast furnace slag and limestone powder. According to the manufacturer, the cement is approved for use in all exposure classes. The addition of fly ash has been generally ruled out, as the already limited availability will decrease due to Germany’s phase-out of coal-fired power generation. For the basic mixture, no other materials are added to the binder phase.

Mix Design for a 3D-Printable Concrete with Coarse Aggregates

61

To have as much fine particles in the mixture as allowed and cover almost all variations from Table 1, a target value of 500 kg/m3 is chosen for the fine particle content. Limited by the height of the CONPrint3D nozzle opening of 50 mm, the maximum grain size cannot be larger than 16 mm. The spaces between the coarse grains are to be filled with smaller fractions going down to a micrometre scale. If the largest possible grains are applied to fill the pores, the required fine particle fraction and the water demand are reduced. In order to achieve a continuous grading curve with a high packing density, a power function is used, which was derived by andreasen and andersen in 1929 [5]:  q d (1) A(d ) = 100% ∗ D

Volume fraction < 0.125 mm

D is the sieve opening of the coarsest grain fraction, A(d) is the cumulative sieve passage of any grain size d and q is the unitless distribution modulus. The smaller q, the finer the mixture is. A direct assignment of a distribution modulus to a chosen fines content requires a function of the largest grain size. Figure 1 shows the relationship as a nomogram. 1.0 0.8

q = 0.2 q = 0.25

0.6

q = 0.3 0.4

q = 0.35 q = 0.4

0.2

q = 0.45

0.0 1

2

4

8

16

32

q = 0.5

Maximum aggregate size [mm] Fig. 1. Fines’ volume fraction as a function of maximum aggregate size and exponent q.

Using Eq. (1), a mixture was composed of local gravel and river sand of fractions 0/2, 2/4, 4/8 and 8/16 as well as a quartz powder BCS 413 and the above-mentioned CEM II/C-M (S-LL) cement. The particle size distributions were measured in power steps from 2–11 mm to 24 mm by sieving and laser diffraction analysis. Figure 2 shows the results. Since the grading curves partly overlap, a trivial calculation of the optimal composition is not possible. So, for each grain size, an optimal content according to andreasen was calculated. Then the square deviation of the particle size distribution from the andreasen-optimum was calculated for each particle size and the mean value was formed over all intermediate results as a fitting quality indicator. A numerical optimisation tool is used to bring the mean value to a minimum by varying the component proportions and the exponent q. In doing that, constraints must be introduced. For example, the sum of all component proportions must be 100% and the fine particle

62

M. Taubert and V. Mechtcherine

content 500 kg/m3 . Additional specifications can be supplemented by further auxiliary conditions.

Fig. 2. Particle size distribution of all solid mixture constituents and the optimized mixture.

It must be noted that the selection of grain sizes included in the optimisation has an influence on the result. In the example proposed here, nine grain sizes ≤ 0.125 mm and seven grain sizes > 0.125 mm are taken into account. Consequently, the fine particle distribution has a greater weight in the optimisation than the rest. It has also to be mentioned that the andreasen particle size distribution indicates the composition in volume fractions. To complete the mixture, the water-to-cement ratio was chosen to 0.4. Herewith, the relevant requirements according to EN 206 are met for all exposure classes, while the consistency of fresh concrete can be well adjusted with a PCE superplasticiser.

4 Experimentation To investigate the consistency dependence on the fine particle content, grading curves with distribution moduli between 0.30 and 0.36 were determined. The finest-grained recipe served as the initial mix and was prepared with a w/c of 0.4 and a superplasticiser content of 1% by the weight of cement. The other mixtures were scaled to contain the same amount of cement. The next finer particle size distribution was subtracted from each of them. Thus, supplementary grading curves were calculated, which were added to the initial mixture successive after a flow table test according to EN 12350-5 respectively. This ensures a constant w/c and superplasticiser dosage. It must be noted, that there are influences of shear history and mixture age, which remained unquantified. The results are shown in Fig. 3. The slump flow decreases with increasing exponent q, i.e. the concrete becomes stiffer. The curve suggests a bilinear relationship. According to this, there is a critical range for the distribution modulus, below which fine particle oversaturation sets in. The grain contacts of coarse aggregates are reduced, making the concrete more

Mix Design for a 3D-Printable Concrete with Coarse Aggregates

63

Slump flow [cm]

flowable. It can be assumed, that this critical range is dependent on the grain shape and surface roughness. However, this needs to be confirmed with additional tests.

60 50

critical range of fine particle oversaturation

40 30 20 0.30

0.32 0.34 0.36 Distribution modulus q [ ]

Fig. 3. Bilinear relationship between distribution modulus q and slump flow of concrete with 16 mm maximum aggregate size, w/c 0.4 and 1% PCE superplasticiser by cement mass.

The mixture determined according to Sect. 3 has an exponent q = 0.339 and is composed as shown in Table 2. Table 2. Mix proportion of the optimized concrete. Component

Quantity [kg/m3 ]

CEM II/C-M (S-LL)

364.3

BCS 413

158.8

Sand 0/2

542.2

Gravel 2/4

330.2

Gravel 4/8

308.8

Gravel 8/16

481.0

Water

145.7

PCE-Superplasticizer

Adjusted

The composition yields a density of 2.33 t/m3 and a mass-related paste content of 27.7%. Trials on a lab-scale gantry printer with a screw conveyor and a rectangular nozzle opening (height of 50 mm, width of 150 mm) demonstrated the acceptable extrudability and good buildability; see Fig. 4. Slump-values between 8 cm and 12 cm are suitable for the given extrusion technology and were adjusted by superplasticiser dosage. No deformation of layers was observed. Ongoing tests indicated a good layer bond.

64

M. Taubert and V. Mechtcherine

Fig. 4. The developed concrete yields a good buildability and acceptable extrudability.

5 Conclusion The compliance with standards facilitates the transfer of 3DCP technology into the practice of construction. Therefore, the binder composition must be limited to the standard cement types. In addition, the content of fines must be limited. To do so, a generalizable approach with a particle size distribution that is extended to the micrometre scale was presented. A power equation according to andreasen and andersen was applied. By minimising the mean square deviation of the calculated particle size distribution from the andreasen optimum by varying the mixture component proportions and the distribution modulus q, an optimized grading curve can be found numerically. After introducing constraints, for example, a maximum fine particle content, the grading curve can be specified. Printing tests with a mixture determined in this way demonstrated the extrudability and buildability of a concrete having maximum aggregate size of 16 mm and a fines content of 500 kg/m3 . This 3D-printable concrete takes into account all standard specifications applicable in Germany. Funding. The authors thank the German Federation of Industrial Research Associations (AiF) for funding the project 21574 BR “Digitales Bauen – Großformatiger 3D-Druck mit Transportbeton”.

References 1. Mechtcherine, V., et al.: Large-scale digital concrete construction – CONPrint3D concept for on-site, monolithic 3D-printing. Autom. Constr. 107, 102933 (2019) 2. Buswell, R.A., et al.: A process classification framework for defining and describing digital fabrication with concrete. Cem. Concr. Res. 134, 106068 (2020) 3. Ivanova, I., et al.: Effects of volume fraction and surface area of aggregates on the static yield stress and structural build-up of fresh concrete. Materials 13(7), 1551 (2020) 4. DIN Deutsches Institut für Normung e. V.: Tragwerke aus Beton, Stahlbeton und Spannbeton: Teil 2: Beton – Festlegung, Eigenschaften, Herstellung und Konformität – Anwendungsregeln zu DIN EN 206–1; (DIN 1045–2) (2008) 5. Andreasen, A.H.M., et al.: Ueber die Beziehung Zwischen Kornabstufung und Zwischenraum in Produkten aus losen Körnern (mit einigen Experimenten). Kolloid-Zeitschrift 50, 217–228 (1930)

Fresh and Hardened Properties of 3D Printable Foam Concrete Containing Porous Aggregates Kirubajiny Pasupathy(B) , Sayanthan Ramakrishnan, and Jay Sanjayan Centre for Sustainable Infrastructure and Digital Construction, School of Engineering, Swinburne University of Technology, Hawthorn, VIC 3122, Australia [email protected]

Abstract. The foam concrete or aerated concrete is most widely used for construction applications where, lightweight, thermal insulation and fire resistance are crucial. While the application of foam concrete is successful in the construction industry, the emerging technology of concrete 3D printing possesses many challenges when using foam concrete. This is primarily due to the high flowability of foam concrete, since it contains a large proportion of air bubbles, resulting in poor shape retention and buildability characteristics. This study investigates the utilization of lightweight aggregate to reduce the foam content to achieve a lightweight 3D printable mix. The expanded perlite (EP) aggregate was used as a volumetric replacement to fine sand that substantially reduced the foam content in the mix. The effect of EP on the physical and mechanical properties was studied and compared with the control 3D printable foam concrete containing sand. The results showed that the replacement of sand with EP aggregate reduces the flow properties that is suitable for 3D printing. Besides, the compressive strength of 3D printed samples was also enhanced with the addition of EP aggregates. The compressive strength of EP based 3D printed specimens at 28 days was determined as 12.95 MPa, 15.5 MPa and 10.6 MPa in the perpendicular, longitudinal, and lateral directions respectively. On the other hand, control 3D printed samples displayed the compressive strength of 5.5 MPa, 8.4 MPa and 4.2 MPa at a slightly lower density range. Keywords: Lightweight concrete · 3D printing · Porous aggregate · Flow · Strength properties

1 Introduction 3D concrete printing (3DCP) technology becomes more popular in recent decades due to its inherent advantages for the construction industry. In comparison with the traditional construction method, the 3DCP method provides a significant reduction in construction wastes, labour cost, time and construction-related accidents as well as allowing architectural freedom to construct artistically intricate structures [1]. While the 3D printing technology could replace the traditional construction methods in most construction applications, one possible application that presents the major challenges for 3D printing is the lightweight concrete construction using 3D printing. The lightweight concrete has many © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 65–70, 2022. https://doi.org/10.1007/978-3-031-06116-5_10

66

K. Pasupathy et al.

applications including, firewalls, facade structures in buildings, theatres, auditoriums, hospitals, and highway sound barrier walls due to their inherent merits of lightweight, high thermal insulation, and good fire performance [2]. Therefore, 3DCP is a potential technology to construct complex architectural structures, particularly in façade and lightweight structures. In the past, the 3D printing of lightweight concrete is explored by a few researchers. For example, Weger et al. [3] introduced expanded glass beads aggregate as a replacement to normal sand in powder bed 3D printing to achieve the density of 1015 kg/m3 with a strength loss of up to 50%. Besides, the replacement of expanded clay aggregates up to 30% (by volume) of sand displayed a good extrudability [4]. Here, both studies have used the lightweight aggregate to reduce the density of the concrete. On the other hand, Ramakrishnan et al. [5] introduced a void (hollow) in the 3D printing structure to reduce the density of the concrete and they have determined that the 3D printed elements with 1369.6 kg/m3 have a compressive strength value of 19.3 MPa at 28 days. On the other hand, a few researchers have investigated the application of 3DCP for foam concrete production [6–9]. Foam concrete is classified as lightweight concrete, which is articulated by adding air voids/pores in the fresh cement matrix [10]. Markin et al. [11], developed the 3D printable foam concrete mixes in the density ranges of 800– 1200 kg/m3 and they have observed the compressive strength varied from 4.2–8.3 MPa after 35–38 days. In another study, it was stated that the 3D printed foam concrete has a compressive strength of 10 MPa for the density range of 1100 to 1580 kg/m3 [12]. The main drawback of using 3DCP for foam concrete is the high flowability and low yield stress at the fresh state, which influences the shape stability and buildability of foam concrete during the extrusion. To reduce this impact, the researchers have introduced some additives to enhance the buildability of foam concrete. For example, Liu et al. [9] proposed a method to add hydroxypropyl methylcellulose (HPMC) and silica fume as a viscosity modifier and thixotropic agents in foam concrete to enhance the foam stability and yield strength properties. In another study, the incorporation of 2% of Nanopowder and 10% of calcium sulfoaluminate cement has improved the fresh properties of foam concrete and increased the buildability. Therefore, based on the available literature, 3DCP has been used either with lightweight aggregate or foam concrete to achieve lightweight concrete. The combination of foam and lightweight aggregate in 3DCP has not been explored yet. The major advantage of combining foam and lightweight aggregate is the amount of foam required to achieve the lower density can be substantially reduced [13]. Therefore, this would be beneficial for enhancing the buildability and the hardened properties of 3D printing of foam concrete. In this study, expanded perlite (EP) was used as a lightweight aggregate to replace the normal sand in the foam concrete. The effect of EP aggregate on the flow properties and the hardened properties such as compressive strength and porosity was assessed.

Fresh and Hardened Properties of 3D Printable Foam Concrete

67

2 Methodology 2.1 Materials and Sample Preparation The foam concrete was prepared using General-purpose (Type GP) Portland cement complying with Australian standard AS 3972. The fine sand with a median diameter (D50) of 600 µm was used in the control mix and hydrophobic surface-modified expanded perlite (EP) was used as a lightweight aggregate. A polycarboxylate etherbased superplasticizer (SP) from BASF Australia was used in both types to control the workability. A diluted sodium dodecyl sulphate solution (SDS) was used to produce the premade foam. Table 1 shows the mix compositions of materials to prepare the 3D printable concrete. Compared to the control mix, the required water content used in the EP-3DCP type was high due to the hydrophobic nature and higher surface area of EP. Besides, the amount of foam content is varied in two types to attain approximate similar density due to the lightweight nature of EP in the EP-3DCP mix. The mixing process was as follows: the dry materials were placed in a Hobart mixer and the mixing was continued for 2 min at low speed. Then the required amount of water with SP was added into the dry mix and the mixing was conducted for another 3 min at the intermediate speed. Simultaneously, the foam was generated using a high shear mixer at 2000 RPM speed for 10 min. After preparing the foam, the required amount of foam was blended with cement slurry, and the mixing was performed another 2 min at a low speed. A gantry-type 3D printer with the working dimensions of 1800 mm (L) × 1600 mm (W) × 1800 mm (H) was used for the printing process. From each mix, 30 mm (W) × 20 mm (H) × 300 mm (L) size of layers were printed at a constant speed of 10 mm/s. Table 1. Mix compositions of 3D printable foam concrete. Mix No

Cement (g)

Sand (g)

EP (g)

Water (g)

Foam (g)

SP (g)

Control-3DCP

1000

856

-

320

50

1.6

EP-3DCP

1000

-

133.2

520

25

1.6

2.2 Testing Procedure The flowability of fresh foam concrete was measured in accordance with the ASTM C1437 using a flow table. The flow diameter was measured at two orthogonal directions and the average values were reported. The compressive strength of 3D printed foam concrete specimens was measured after 28 days of printing. It should be noted that the 3D printed concrete displayed an anisotropic behaviour and therefore the compressive strength test was determined in three different directions such as longitudinal, lateral and perpendicular direction. A 30 × 30 × 30 mm3 size of concrete specimens were extracted from the printed specimens and the strength test was performed using the Universal Mechanical Testing Systems (MTS) at the displacement control mode of 1 mm/min.

68

K. Pasupathy et al.

Three samples from each mix at each direction were tested, and the average values were reported for each mix and each direction. The porosity of the 3D printed filaments was evaluated on the 30 × 30 × 30 mm3 specimens extracted from printed filaments using the compaction method [14]. To conduct this test, three specimens with the dimensions of 30 × 30 × 30 mm3 were extracted from each type and the volume of the specimens was determined by measuring the linear dimensions using a Vernier calliper (Vo ). After that, the samples were powdered and compressed into a solid disc with a 25 mm diameter diecast set using a hydraulic jack operated at a constant pressure of 16 MPa. The volume of the solid disc was measured using the Vernier calliper (V1 ). Based on the volume measurements, the porosity (Ø) was calculated by the following formula: ∅ = (Vo − V1 )/Vo × 100%

(1)

3 Results and Discussion 3.1 Flow Measurement Figure 1(a) displayed the spread diameter of the fresh foam concrete mixes before and after 25 drops of the flow table. As can be seen in Fig. 1, before dropping the table, the spread diameter of both types of foam concrete mixes was almost similar, which is about 100 mm. Therefore, both mixes have zero slump values and this indicates both mixes have good shape retention during the extrusion process [15]. However, the flow diameter of the two types was significantly varied after dropping the table. The final flow diameter of control and EP base mixes was 213.5 mm and 167.5 mm, respectively. This suggests that the workability of the mix was reduced with the incorporation of EP, whereas sand mix is highly flowable. 3.2 Density and Compressive Strength Measurements The dry density of the Control-3DCP and EP-3DCP samples was determined as 1134 kg/m3 and 1245 kg/m3 , respectively. The 28 days compressive strength test results are shown in Fig. 1(b). As shown, the compressive strength values are varied with the testing direction, which is reflecting the anisotropic behaviour of 3D concrete printing. In both types, the strength values measured in the longitudinal (printing) direction are higher than the other two directions. This could be due to the higher compaction from the extrusion pressure and similar behaviour also observed by other researchers [5]. The comparison between the two groups reveals that the compressive strength of the EP-3DCP was higher than the Control-3DCP in all three directions. The 28 days compressive strength of EP-3DCP showed 135%, 152% and 84.5% enhancement in perpendicular, lateral and longitudinal directions, compared to the control mix. This could be due to the different amounts of foam content used in two types. Compared to the control mix, the amount of foam used in the EP-3DCP mix was low and the substantial reduction of foam content enhances the strength properties [13].

Fresh and Hardened Properties of 3D Printable Foam Concrete

69

3.3 Porosity Measurements The porosity values of Control-3DCP and EP-3DCP was determined as 42.7% and 37.8%, respectively. The variation in the porosity between the two types of mixes was due to the different amounts of air content (foam) in both types. Moreover, the porosity values are reflected in the compressive strength values. The compressive strength of foam concrete reduces with the increasing porosity as the air content increases, which reduces the matrix strength [16]. This is correlated well with the strength values observed for the 3D printed samples. 18

250

Spread diameter [mm]

200

150

100

50

16

Compressive strength [Nmm-2]

(a)

Before drop After drop

14

Perpendicular Longitudinal Lateral

(b)

12 10 8 6 4 2

0

0 Control-3DCP

EP-3DCP

Control-3DCP

EP-3DCP

Fig. 1. (a) Spread diameter of the fresh mix, (b) Compressive strength test results.

4 Conclusion This paper presented the results of a study on the 3D printable foam concrete with EP based lightweight aggregate. The flow measurements and hardened properties were compared with a control mix containing sand aggregates. Based on the presented study, the following conclusions can be drawn: • The spread diameter of EP based foam concrete is lower than the control mix, which is beneficial for shape retention and buildability in 3DCP. • The compressive strength of EP based 3D printed filaments were higher than the control specimens in all three testing directions. • Porosity analysis revealed that the porosity of EP based mix was lower than the control mix. This is due to the lower foam content in EP based mix than sand-based mix.

Acknowledgement. The authors acknowledge the Swinburne University of Technology and Australian Research Council (DE190100646) for supporting this work.

70

K. Pasupathy et al.

References 1. Wu, P., Wang, J., Wang, X.: A critical review of the use of 3-D printing in the construction industry. Autom. Constr. 68, 21–31 (2016) 2. Ramakrishnan, S., Pasupathy, K., Sanjayan, J.: Synthesis and properties of thermally enhanced aerated geopolymer concrete using form-stable phase change composite. J. Build. Eng. 40, 102756 (2021) 3. Weger, D., et al.: Lightweight concrete 3D printing by selective cement activation–investigation of thermal conductivity, strength and water distribution. In: Proceedings of the RILEM International Conference on Concrete and Digital Fabrication. Springer (2020)https://doi.org/ 10.1007/978-3-030-49916-7_17 4. Rahul, A., Santhanam, M.: Evaluating the printability of concretes containing lightweight coarse aggregates. Cement Concr. Compos. 109, 103570 (2020) 5. Ramakrishnan, S., et al.: Concrete 3D printing of lightweight elements using hollow-core extrusion of filaments. Cement Concr. Compos. 123, 104220 (2021) 6. Cho, S., et al.: Rheology and application of buoyant foam concrete for digital fabrication. Compos. B Eng. 215, 108800 (2021) 7. Lublasser, E., et al.: Robotic application of foam concrete onto bare wall elements-analysis, concept and robotic experiments. Autom. Constr. 89, 299–306 (2018) 8. Bedarf, P., et al.: Foam 3D printing for construction: A review of applications, materials, and processes. Autom. Constr. 130, 103861 (2021) 9. Liu, C., et al.: Influence of hydroxypropyl methylcellulose and silica fume on stability, rheological properties, and printability of 3D printing foam concrete. Cement Concr. Compos. 122, 104158 (2021) 10. Dhasindrakrishna, K., et al.: Collapse of fresh foam concrete: Mechanisms and influencing parameters. Cement Concr. Compos. 122, 104151 (2021) 11. Markin, V., et al.: 3D-printing with foam concrete: From material design and testing to application and sustainability. J. Build. Eng. 43, 102870 (2021) 12. Markin, V., et al.: Material design and performance evaluation of foam concrete for digital fabrication. Materials 12(15), 2433 (2019) 13. Pasupathy, K., Ramakrishnan, S., Sanjayan, J.: Enhancing the mechanical and thermal properties of aerated geopolymer concrete using porous lightweight aggregates. Constr. Build. Mater. 264, 120713 (2020) 14. Gao, H., et al.: A novel inorganic thermal insulation material utilizing perlite tailings. Energy and Buildings 190, 25–33 (2019) 15. Arunothayan, A.R., et al.: Development of 3D-printable ultra-high performance fiberreinforced concrete for digital construction. Constr. Build. Mater. 257, 119546 (2020) 16. Pasupathy, K., Ramakrishnan, S., Sanjayan, J.: Formulating eco-friendly geopolymer foam concrete by alkali-activation of ground brick waste. J. Clean. Prod. 325, 129180 (2021)

Sustainable 3D Concrete Printing with Large Aggregates Wilson Ricardo Leal da Silva(B) , Martin Kaasgaard, and Thomas J. Andersen Danish Technological Institute, Taastrup, Denmark [email protected]

Abstract. The number of large-scale projects featuring the use of 3D Concrete Printing (3DCP) had a steep increase in the past years. The most remarkable applications include apartment and residential buildings and promise to deliver a robust, cost-effective and sustainable solution compared to conventional construction. One of the paths to boost 3DCP’s sustainability lies in the material upscaling from mortars to concrete; thus, our work focuses on the concrete mixes for 3DCP. For that, we use a calcined-clay limestone-based cement (FutureCEM®) and large aggregates (up to 8.0 mm) to produce printable mixes. In this work, we present a comparative analysis of mixes produced with CEM I and FutureCEM. Specifically, mixes with strength classes of C25 and C45 produced with these cements are compared to a mortar produced with White Cement; a mortar produced with FutureCEM®; and a concrete produced with CEM I. Also, the CO2 footprint of the mixes is compared to that of Ready Mix Concrete produced in Denmark; such analysis validates the environmental benefits of material upscaling and points out that 3DCP mixes can be produced with a similar CO2 footprint to that of conventional concrete. Keywords: 3D Concrete Printing · Large aggregates · Sustainability

1 Introduction The number of large-scale projects featuring the use of extrusion-based 3D Concrete Printing (3DCP) as a construction technology had a steep increase over the past two years. A look into available project details shows that extrusion-based 3DCP is the most widespread technology, 3DCP is mainly used to print the contour (walls), and dry mortars remain as the main feedstock used in the process. While 3DCP is claimed to deliver a robust, cost effective and sustainable solution compared to conventional construction, the theme “sustainability” in 3DCP is still underexplored. There is a handful number of studies providing insights on 3DCP sustainability, e.g. [1–3], with little to none detailed data from full-scale projects. Besides architectural and structural design strategies that help reduce material consumption, the path to a more sustainable 3DCP lies in the material scaling up, since the material amount and type are pointed as critical factors on the environmental impact to produce an element such as a wall using 3DCP [3]. The most straightforward way to © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 71–77, 2022. https://doi.org/10.1007/978-3-031-06116-5_11

72

W. R. L. da Silva et al.

achieve a feedstock with low embodied energy is through the inclusion of large aggregates (as in conventional concrete) in the 3DCP mixes. This also enables producing a printing material that is cost competitive against conventional concrete produced locally, e.g. in a Ready-Mix concrete (RMC) plant. While the economic and environmental costs of 3DCP must be assessed in relation to the shape complexity of the final product [1], our focus is limited to the environmental costs and benefits of scaling up from mortar to concrete in extrusion-based 3DCP. The analysis presented in this article is part of the project “Next Generation 3Dprinted Concrete Structures” (N3XTCON) [4], which targets the development of sustainable 3DCP mix design approach. We opted to work with a calcined-clay limestone-based cement (FutureCEM) as the binder for 3DCP applications. In this study, we discuss the general aspects of N3XTCON’s 3DCP mix design approach and present a comparative analysis against 3DCP mortars produced with CEM I and FutureCEM. We also compare our 3DCP mixes to concrete from RMC plants. The obtained results validate the environmental benefits that yield from material scaling up from mortar to concrete in 3DCP as well as the competitiveness of the developed 3DCP mixes against conventional concrete from RMC plants.

2 N3XTCON Mix Design Strategy The N3XTCON mix design approach focuses on sustainability, rheology control and structural build up control. The first iteration of the approach is described as follows. First and foremost, sustainability is the backbone of the mix design. The material parameters considered in the design include targets for compressive strength, aggregate grading curve, particle packing density, aggregates’ shape, volumetric ratio of mortar, binder composition, and concrete slump. All of which to achieve a mix with an optimal cement content for a given strength class. The maximum aggregate particle size is set at 8.0 mm. It is evident that mix design parameters must be related to the pumping and extrusion setup. Relevant indicators are the ratio between the maximum particle size and the minimum extrusion nozzle size (α1 = Ømax /dmin-nozzle ), and the volume fraction of large aggregates (α2 = Vagg > 4mm /Vc ). The mixes tested so far have α1 < 23% and α2 < 26%. As for the specific target ranges for each mix design parameter listed above, we still depend on carrying out tests to come up with solid recommendations. Hence, the results presented in this article should be taken as a “work-in-progress”. So far, the developed procedure is as follows. Once a mix design is first suggested, we carry out small scale tests with concrete batches of 7 to 10 L to verify the concrete consistency via slump test. The target slump test is between 130 to 220 mm to achieve a pumpable and extrudable concrete – the mixes with greater consistency are recommended for 3DCP setups where accelerators are injected and mixed in the nozzle. Based on the slump test results, the initial yield stress (τy,s ) reads [5]:

Sustainable 3D Concrete Printing with Large Aggregates

τy,s = ρ ( 25.5 − Sh )/17.6, (for 5 < Sh < 25 cm)

73

(1)

where ρ is the density of the mix [kg/m3 ] and Sh the slump height [cm]. Thus, the target initial yield stress is between 0.5 and 1.6 kPa. After a base mix design is found and proved pumpable/extrudable, the design strategy shifts to rheology and structural build up control. The control of both rheology and structural build up, described as follows, is achieved by additives and admixtures. The rheology control secures a long open time (i.e. operational time of a fresh mix) to handle the concrete before extrusion takes place. This is achieved by using admixtures such as hydration retarders (e.g. sodium gluconate and tartaric acid-based admixtures) and plasticizers (mainly polycarboxylate-based high-range water reducing admixtures). The admixtures’ dosage depends upon 3DCP process parameters, production rate and conditions (temperature and humidity). After the material is extruded, it has to exhibit enough strength, so the printed element does not collapse. To enable that, the structural build up control strategy relies on adjusting the rate of cement hydration. A compelling review on the chemical reactions from different activation strategies used in digital fabrication processes can be found in [6]. To monitor the concrete structural build up, we rely on the use of penetration tests – with a Ø20 or Ø10 mm hemispherical tip. The concrete yield stress (τy,p ) evolution after activation based on the penetration load reads [7]: τy,p = F/3π R2 , (for hemispherical tips)

(2)

where F is the penetration load [N] and R the radius [mm] of the hemispherical tip. Right after activation, the mixes tested so far exhibit a τy,p within 1.1–2.5 kPa. The structural build-up is proportional to the dosage of “accelerators” added to the concrete mix. From a mix design perspective, the main point is to first determine the application where a new 3DCP mix is to be used. For example, when 3DCP is applied to print a house, the vertical build-up rate is in the range of 0.3 to 0.5m/h due to the usually large contour length; whereas the production of a column requires a structural build-up of at least 2.0 to 3.0 m/h. In other words, while the same base concrete mix (from a rheology control standpoint) can be suitable for both cases, it is the structural build up control that ensures that the printed element will not collapse. Based on the described approach, we developed and successfully tested different concrete mixes fit for 3DCP. Figure 1 depicts a cross section of one of the printed mixes in comparison to a 3DCP mortar produced with White Cement (CEM I 52.5 R SR5). The CO2 emissions associated with the produced mixes are discussed as follows.

74

W. R. L. da Silva et al.

Fig. 1. 3DCP samples produced with mortar (Ømax = 0.5 mm) and concrete (Ømax = 8.0 mm).

3 3DCP with Mortar and Concrete The idea in this analysis is to understand the effects of upscaling from mortars to concrete on the overall CO2 emission of 3DCP mixes. For the mortars, the maximum particle size was set at 0.5 mm, while the concrete contains aggregates up to 8.0 mm. In each analysis, the CO2 emissions of the mixes are normalised based on a reference mix, which is either a mortar or concrete. The mixes are produced with one of the cement types CEM I 52.5 R SR5 (White Cement), CEM I 52.5 N (MS) (LA) (Rapid Cement), and CEM II/B-M(Q-LL) 52.5 N (FutureCEM) – from Cementir Holding. For detailed information on the main performance parameters and composition requirements of these cements, please refer to EN 197-1 [8]. We designed printable concrete and mixes with strength class C25 and C45 using FutureCEM and Rapid Cement. These mixes are then evaluated against a Case 1) 3DCP mortar mix produced with White Cement and Case 2) FutureCEM. Also, we carried out a complementary analysis comparing our 3DCP concrete to RMC. The CO2 analysis for cases 1 and 2 are depicted in Fig. 2. The OPC content, type, and the mix label of each mix are also listed in Fig. 1. The strength class of R1 and F1 is C25, whereas R2 and F2 is C45. Note that the CO2 analysis does not account for the emissions from any of the components used to control rheology and structural build-up, namely plasticizers and accelerators. This is because their amounts in the concrete composition is rather low (0.1 to 2.5% bwoc.) and, thus, safely neglected. In Case 1 (Fig. 2a), the material upscaling contributes to a CO2 reduction from 63% to 78%, where the concrete mixes produced with FutureCEM feature a greater reduction. The upscaling effect on reducing CO2 emission is further supported by the replacement of the binder system, i.e. from White Cement to FutureCEM, which alone contributes to approximately 30% reduction in CO2 emissions. In Case 2 (Fig. 2b 4b), the material upscaling features the same trend, with an overall CO2 reduction ranging from 40 to 65% against the reference mix M2. The reduction is less dramatic than that from Case 1 because the reference mortar in Case 2 is produced with FutureCEM. Note that, the strength class of reference mortars in Cases 1 and 2 is greater than the concrete mixes R and F. Nonetheless, it is most likely that a 3DCP

Sustainable 3D Concrete Printing with Large Aggregates

75

Fig. 2. Normalised CO2 emission of 3DCP mortars and concrete: (a) Case 1 and (b) Case 2. The reference value in Case 1 and 2 are from our typical 3DCP mortar produced with 517 kg/m3 of Rapid Cement (M1) and 508 kg/m3 of FutureCEM (M2).

mortar composition with similar strength class to that of R and F would still feature a greater CO2 emission. For benchmarking purposes, the CO2 emissions from R1-2 and F1-2 were compared to that from concrete mixes from a local RMC company, see details in [9]. The obtained results are displayed in Fig. 3.

Fig. 3. Comparison of CO2 emission from 3DCP (concrete mixes) vs RMC at different strength classes. (Note: The RMC values account for conventional concrete mixes produced with aggregates up to 32 mm and slump within 40–120 mm, whereas the RMC-Green is an equivalent mix that is produced with the same aggregate type and slump range).

Figure 3 shows that both 3DCP mixes R and F have lower emissions than that from conventional RMC concrete. When compared to the so-called RMC-Green, which is a

76

W. R. L. da Silva et al.

sustainable RMC option, the emissions from 3DCP mixes R1-2 are in the same order of magnitude of RMC-Green. Whereas the emissions from F1-2 are even lower than RMC-Green. If we consider that the maximum particle size in the 3DCP and RMC-green mixes is different, it is safe to conclude that 3DCP concrete can be made sustainable and competitive from an environmental perspective against RMC concrete.

4 Conclusions The use of the N3XTCON mix design strategy enabled the production of a printable concrete mixes with a relatively low CO2 emission when compared to conventional RMC and even lower emission when compared to mortars. The relative drop in CO2 emissions should be in the same order of magnitude as presented in this work in case the presented mix design strategy is deployed in other countries. Notice that adjustments in terms of the target parameters are necessary to adapt to different printing processes. In any case, the N3XTCON mix design strategy provides companies with a sustainable alternative to pre-mixed materials and gradually paves the way towards a sustainable 3DCP. While the material upscaling from mortar to concrete is a necessary development track to leverage sustainability on large-scale 3DCP applications, the use of mortars should not be neglected. This is because there are characteristics intrinsic to design concepts that can only be achieved when using mortars with a high structural build up rate. Hence, locally produced concrete and dry mortars will gradually cement their markets as the number of 3DCP applications evolves. Acknowledgements. The support of the Innovation Fund Denmark (Grant no. 8055-00030B: N3XTCON) - Project partners: Danish Technological Institute; Technical University of Denmark; University of Southern Denmark; Henning Larsen Architects A/S; Bjarke Ingels Group A/S; CRH Concrete A/S; COBOD International A/S; NCC Danmark A/S; Aalborg Portland A/S; FB Gruppen A/S; and AP Pension.

References 1. De Schutter, G., Lesage, K., Mechtcherine, V., Nerella, V.N., Habert, G., Agusti-Juan, I.: Vision of 3D printing with concrete — Technical, economic and environmental potentials. Cement Concrete Res. 112, 25–36 (2018) 2. Han, Y., Yang, Z., Ding, T., Xiao, J.: Environmental and economic assessment on 3D printed buildings with recycled concrete. J. Clean. Prod. 278, 123884 (2021) 3. Agustí-Juan, I., Müller, F., Hack, N., Wangler, T., Habert, G.: Potential benefits of digital fabrication for complex structures: environmental assessment of a robotically fabricated concrete wall. J. Clean. Prod. 154, 330–340 (2017) 4. Leal da Silva, W.R., Garzon, S.F., Andersen, T.J., Ahrenkilde, I.: Towards sustainable 3D concrete printing. Construct. Print. Technol. CPT 2(-), 23–31 (2020) 5. Roussel, N.: Correlation between yield stress and slump: comparison between numerical simulations and concrete rheometers results. Mat. Struct. 39(4), 501 (2006) 6. Marchon, D., Kawashima, S., Bessaies-Bey, H., Mantellato, S., Ng, S.: Hydration and rheology control of concrete for digital fabrication: potential admixtures and cement chemistry. Cement Concrete Res. 112(S.I.), 96–110 (2018)

Sustainable 3D Concrete Printing with Large Aggregates

77

7. Lootens, D., Jousset, P., Martinie, L., Roussel, N., Flatt, R.J.: Yield stress during setting of cement pastes from penetration tests. Cem. Concr. Res. 39(5), 401–408 (2009) 8. EN 197-1:2012 - Cement - Part 1: Composition, specifications, and conformity criteria for common cements (2012) 9. Unicon - Cementir Holding: Link. https://www.unicon.dk/media/1660/prisliste-2021-web.pdf

Design and Fabrication of Spatially Graded Concrete Elements with Ice Aggregate Method Vasily Sitnikov1(B) , Lena Kitani1 , Artemis Maneka2 , Ena Lloret-Fritsch3 , Juney Lee2 , and Benjamin Dillenburger1 1 Digital Building Technologies, ETH Zurich, Zurich, Switzerland

[email protected]

2 Block Research Group, ETH Zurich, Zurich, Switzerland 3 Gramazio Kohler Research, ETH Zurich, Zurich, Switzerland

Abstract. An important research perspective for optimization of concrete structures today can be recognized in the concept of controlled spatial grading of concrete elements. Spatial grading aims to control the volumetric distribution of concrete within an element based on the internal stresses, which enables a more efficient structural performance with less material without changing the overall geometry of a structural part. Due to high geometric complexity, the fabrication of spatially graded structures are typically based on additive manufacturing techniques such as extrusion or voxel-based 3D printing. However, for large-scale production and fabrication, 3D printing is ultimately constrained by speed and cost. This article presents the potential of regular ice as aggregate for casting spatially graded concrete components. Prefabricated ice aggregates of varying size and geometry are carefully placed into a mould to form regions with lower and higher relative densities, into which self-compacting concrete is poured. We investigate various parameters of customized ice aggregates including geometry, scale and relative densities. The presented experiments demonstrate how ice aggregate can be used to efficiently fabricate spatially graded concrete elements with high geometric complexity, while improving the structural performance by concentrating concrete only where it is needed. In addition, the intricate patterns of voids created by ice aggregate result in unique visual and lighting qualities that would be difficult to achieve with other fabrication methods. This article addresses the fundamental research questions and provides a ground for further research in digitization and automation of the ice aggregate production method. Keywords: Ice formwork · Spatially graded concrete · High performance concrete · Compression-only structures · Zero waste formwork

1 Introduction The recent advancement in additive manufacturing has opened doors to cost-effective mass production of optimized and geometrically complex parts in many industries, from © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 78–83, 2022. https://doi.org/10.1007/978-3-031-06116-5_12

Design and Fabrication of Spatially Graded Concrete Elements

79

medicine to aeronautical industries. However, each sector is interested in specific materials and scale domains, thus the additive manufacturing principles have to be adapted to every specific application. In the construction industry, digital fabrication processes with concrete such as Digital Casting [1], Contour-Crafting [2] or extrusion-based 3D printing of concrete has made a great leap in recent years [3, 4]. 3D printing of concrete in particular, has excelled and can now be utilized effectively in large-scale applications. Nevertheless the geometry range of above mentioned methods comes with certain process-specific limitations. At the same time, binder jet printing provides almost full degree of geometrical design freedom by providing a constant vertical support for a fabricated part [5]. However the scalability and low mechanical strength of the binder jet prints is limiting their field of application. These limitations pose an obstacle in taking the full advantage of additive 3D manufacturing, dooming the whole area of lattice structures and functionally graded concrete elements to be inaccessible for practical application. The lattice structures, however, is a promising material strategy for lightweight structures and minimal material use in construction. Ice Formwork system, however, has shown tremendous potential for efficient fabrication of spatially grated concrete elements, combining both the geometrical complexity and the high performance concrete [6]. The system developed by Vasily Sitnikov in his PhD thesis provides a zero-waste and self-removing moulding for complex interlayered geometry in concrete. The thesis includes the scientific basis for use of ice as the moulding material for concrete, and demonstrates several processing methods compatible with this material configuration. Originated in the previous study, ice aggregate method represents a strategy for a fully automated production method of structurally graded concrete elements. However, to approach the questions of robotic automation the fundamental topics of this material system are to be assessed in this paper: (I) investigation of stochastic and non-stochastic spatial patterns for ice aggregates and their impact on the mechanical performance of the cast elements; and (II) implementation of graded structural concrete components using ice aggregate in a design workflow of a compression-only shell structure. The two topics were assessed in two parallel studies simultaneously. The first study analyzes the variation of spatial patterns with variable ice aggregates in a confined formwork. The goal was to determine the possible geometry and topology of the ice cells, in regards to their form and organizing patterns. In this respect, our interest was to set up a simulation framework, as well as to test the production of material samples and identify their mechanical properties. The second study presented explores how spatially graded elements can be implemented in an architectural design workflow. Since the scope of the research work was to investigate the applicability of ice aggregate for non-reinforced concrete structures, the focus was directed towards the design of a compression-only shell structure.

2 I - Spatial Patterns of Ice Aggregate and Their Properties The study of spatial patterns focuses on two types of aggregations; the autonomous (stochastic) and controlled (non-stochastic) packing. To orient ourselves within the field of possible spatial patterns, we studied the qualities of ice aggregate packing with the

80

V. Sitnikov et al.

help of a digital simulation framework. The computational setup used for the digital experiments was shared between the Rhinoceros, Grasshopper and Blender software platforms. Based on the packing patterns and the relative density, two specific aggregate geometries were selected as representative of the stochastic and non-stochastic systems, and produced physically for the material testing. In order to effectively transition from a digital model to a physical prototype, we developed a system of 3DP PLA moulds (Fig. 1). The moulds were used for compacting ice abrasion collected from a local ice rink, and physically similar to natural snow. Pressforming ice abrasion proved to be a rapid and precise production method for the repetitive cell units. Fabrication helped us assess this dynamic system, and laid the foundation for an ice aggregate-specific, automated setup (Fig. 2).

Fig. 1. PLA 3DP mould for fabrication of regular-shaped ice aggregates.

Fig. 2. Stochastic and nonstochastic spatial patterns; physical samples and digital models.

2.1 Uniaxial Compressive Strength Test To get an understanding of how geometry affects structural performance, uniaxial compression tests were performed on the two sets of samples demonstrating the main categories of ice aggregate packing. Two sets of three cubes of 150 mm and three beams of 40x40x160 mm were tested to assess the spatial pattern and the concrete batch respectively. Sample 1: was made with the stochastic self-packing of crushed-ice aggregates which have been sieved to the size of approx. 60 mm and fused together through cycles of melting and freezing, after which the test cubes were cast with concrete. This type of spatial pattern is considered to be the easiest and fastest to fabricate. Sample 2: Schwarz P was the aggregate form chosen to help us extract some understanding of the structural performance of non-stochastic patterns.

Design and Fabrication of Spatially Graded Concrete Elements

81

Table 1. Uniaxial compressive strength tests on stochastic and nonstochastic concrete lattices

The two concrete batches were of almost identical properties, namely the average compressive strength values were 59.2 MPa for Sample 1 and 56.6 MPa for Sample 2. The concrete lattice cubes performed much lower than their solid counterparts. However, it is important to notice that the Schwarz P samples performed 3.4 times better than the crushed ice aggregate cubes, where the compressive strength was equal to the bending strength of the control beams. Despite the two sample groups having similar relative density, the non-stochastic packing allowed the resulting concrete lattice to be more resistant to a compressive stress. The comparison between the two samples highlights the effect of geometry on performance (Table 1).

3 II - Design Application of Ice Aggregate The second study explores how ice aggregate can be used as an integral part of the design and fabrication process of discretized elements of a large-scale, compression-only shell structure. Compression-only structures eliminate the need for steel reinforcement, which provide an appropriate opportunity for the application of spatially graded concrete elements as load-bearing structural elements. Despite their structural efficiency, fabrication of formwork for doubly-curved concrete shell structures remains challenging. Additionally, while material efficiency could be improved further by reducing the thickness of the structure where the internal forces are lower, such procedure does not necessarily simplify the geometry of the formwork nor does it improve its fabricability. An alternative method for improving material efficiency and fabricability would be to maintain a constant thickness for all of the constituent elements, and remove the material from the interior region. In this regard, ice aggregate offers an effective method for controlling the gradation of concrete within each element while maximizing the structural depth, which results in a more robust structure that can accomodate a wider range of load cases. 3.1 Form Finding and Discretization The geometry of the compression-only shell structure was computed using RhinoVAULT 2 (Fig. 3a). This geometry is then post-processed using the COMPAS framework [7] and discretized into voussoirs that have planar top (extrados) and bottom (intrados) faces. Each voussoir consists of structural ribs and ornamental voids. The sizes of ribs are proportional to the magnitude of internal forces, while the level of gradation and porosity of the voids are determined by the relative z position of the blocks (Fig. 3b). The three levels of gradation are achieved by varying the sizes of the ice aggregate.

82

V. Sitnikov et al.

Fig. 3. A) shell structure; and B) the gradation of the voids based on its Z-axis position.

3.2 Fabrication of a Prototypical Voussoir The geometry of the prototypical element is inspired by an internal block of the shell structure. In order to demonstrate the various levels of gradation of concrete within this single element, the layout of the ribs is designed to partition the element into three concentric zones that each correspond to the three levels of void gradation.

Fig. 4. A) The layout of the ribs with temporary spacers; B) placement of ice aggregate with sizes that correspond to the zone; and C) demoulding of the boundary formwork and melting of the ice aggregate; D) spatially graded concrete element fabricated with ice aggregate.

The fabrication steps include the placing of the spacers (Fig. 4a), the placement of ice with sizes that correspond to each of the zones and the removal of the spacers once the aggregates are frozen together (Fig. 4b), and finally casting of self-compacting concrete (Fig. 4c). Once the concrete has been cured, the outer formwork can be removed and the ice aggregate melts away (Fig. 4d). Further research could look into improving the method with digital fabrication tools, and is to be addressed in future work.

4 Conclusions The Study I allowed us to see the possible performance variation in the selected concrete lattice samples and identify the role of geometry. Although the difference in performance was captured, this area requires more extensive study of the lattice geometries, and may

Design and Fabrication of Spatially Graded Concrete Elements

83

require further advancement of the simulation framework with FE tools for assessment of the mechanical properties. In turn, the Study II helped us to outline a possible area of application for spatially graded concrete elements in compression-only structures. The full-scale prototype was produced with analog production methods. The results of these studies allow further research into robotic automation of the ice aggregate fabrication and assembly. By automating this process the production scale and precision is expected to reach a decent level compatible with real-world applications.

References 1. Lloret-Fritschi, E., et al.: From smart dynamic casting to a growing family of digital casting systems. Cement Concrete Res. 134, 106071 (2020) 2. Zhang, J., Khoshnevis, B.: Optimal machine operation planning for construction by contour crafting. Autom. Constr. 29, 50–67 (2013) 3. Anton, A., et al.: Concrete choreography: prefabrication of 3D-printed columns. In: Burry, J., Sabin, J., Sheil, B., Skavara, M. (eds.) Fabricate 2020. UCL Press, pp. 286–293 (2020) 4. Bhooshan, S., et al.: Design, engineering and fabrication of a 3D-concrete-printed unreinforced masonry shell footbridge. In: Proceedings of 5th ICSA, Aalborg (2022) 5. Odaglia, P., Voney, V., Dillenburger, B., Habert, G.: Advances in binder-Jet 3D printing of non-cementitious materials. In: Bos, F.P., Lucas, S.S., Wolfs, R.J.M., Salet, T.A.M. (eds.) DC 2020. RB, vol. 28, pp. 103–112. Springer, Cham (2020). https://doi.org/10.1007/978-3-03049916-7_11 6. Sitnikov, V.: Ice Formwork: The Rationale and Potential of Ice-Based Moulding Systems for the Production of Complex-Geometry Precast Concrete, TRITA-ABE-DLT. Kungliga Tekniska högskolan, Stockholm (2020) 7. Van Mele, T., Liew, A., Méndez Echenagucia, T., Rippmann, M., et al.: COMPAS: A framework for computational research in architecture and structures (2017)

Binders and Aggregates 2: Alternative Binders

Accelerating Early Age Properties of Ultra-Low Clinker Cements for Extrusion-Based 3D Printing Rutendo Rusike1 , Michael Sataya2 , Alastair T. M. Marsh1 , Sergio Cavalaro3 , Chris Goodier3 , Susan A. Bernal1 , and Samuel Adu-Amankwah1(B) 1 School of Civil Engineering, University of Leeds, Woodhouse Lane, Leeds LS2 9JT, U.K.

[email protected]

2 Pavements and Materials, Arup, East West Building, Tollhouse Hill,

Nottingham NG1 5AT, U.K. 3 School of Architecture, Building and Civil Engineering, Loughborough University,

Loughborough L11 3TU, U.K.

Abstract. In this study, we investigated the influence of commercial sodium nitrate/thiocyanate accelerator compared to calcium sulfoaluminate cement addition on setting time, rheology and reaction kinetics of ultra-low clinker composite cement for extrusion-based 3D printing application. CEM I 52.5 N and a ternary composite cement with 70% clinker replaced by slag and limestone were evaluated. Results indicate that final setting time of 30 min and buildable yield stresses can be attained with less than 5% addition of calcium sulfoaluminate, with ettringite and C-(A)-S-H as main reaction products. This demonstrates the synergy between slag and calcium sulfoaluminate cements can be harnessed to control rheology and hardening. This is significance for evidencing suitability of ultralow clinker composite cements for extrusion-based 3DCP, thus helping to fulfil its wider potential as a low-carbon concrete technology. Keywords: 3D printing · Low-carbon cements · Admixtures · Setting time · Early age properties

1 Introduction The potential of 3DCP to revolutionize the construction industry is enormous. Despite this, application of the 3DCP by the construction industry is limited. Fabrication by extruding fast setting cementitious formulations is one of the most investigated 3DCP techniques. The technique successively deposits layers of fresh cementitious material, often mortars. These mixtures must be sufficiently workable for extrusion, and retain its shape upon deposition whilst undergoing rapid structuration and strength development to support subsequent layers without weakening the inter-layer bond strength [1, 2]. Such demanding characteristics of 3DCP cannot be achieved with conventional concrete mix design. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 87–92, 2022. https://doi.org/10.1007/978-3-031-06116-5_13

88

R. Rusike et al.

The concrete mix design controls the rheological properties which in turn determine extrudability, buildability and structuration [3]. These are characterized in terms of yield stresses, thixotropy, setting time and strength evolution. Several parametric studies exploring relationships between constituents, proportioning and rheology have been reported [3–5] and reviewed in [6]. For example, Rahul et al. [4] investigated the effect of plasticizer dosage and viscosity modifying agents (VMA) on Portland composite cements and found that printable mixes had yield stresses ranging between 1.5 – 2.5 kPa. Chaves et al. [7] evaluated similar parameters on limestone ternary GGBS or PFA cements. Yield stresses comparable to [4] were reported for printable mixes, but these increased proportionately as the VMA and inversely as plasticizer and SCM content. Yuan et al. [8] found significantly lower static yield stresses and structuration in mixes containing GGBS and PFA compared to VMAs such as nanoclay, carbonate or silica. These highlight fundamental challenges around mix design for 3DCP, i.e. overreliance on Portland cement and high cement content in mortars (typically > 800 kg/m3 ). One way to reduce the clinker factor is to use SCMs. However, SCMs hydrate slowly, requiring acceleration to be suitable for 3D printing. In this study, it is investigated the feasibility of using calcium sulfoaluminate cement as an accelerator in high SCM composite cement designed for digital fabrication applications. Results are compared to those of systems produced with commercial sodium nitrate/thiocyanate accelerators.

2 Materials and Methods 2.1 Materials Two cementitious systems, CEM I 52.5 N and ternary limestone-slag composite cement containing 70% OPC replacement, designated as C and CSL were used in this study. Commercial sodium nitrate/thiocyanate (designated Rp) or calcium sulfoaluminate cement was used as accelerator. Composition and particle size distribution (PSD) of the cementitious materials as determined by X-ray fluorescence (XRF) and laser difractometry are shown in Table 1 and Fig. 1 respectively. Table 1. Oxides composition of cement constituent materials as determined by XRF. Constituent

SiO2

Al2 O3

CaO

Fe2 O3

MgO

Na2 O

K2 O

CEM I (C) Slag (S)

19.4

5.1

64.9

3.1

1.1

0.1

34.9

11.6

41.8

0.5

5.8

0.1

Limestone (L)

2.0

0.8

53.1

0.3

0.6

Anhydrite

2.1

0.6

38.2

0.2

1.5

* Loss on ignition (LOI) obtained as mass loss up to 950 °C.

SO3

LOI*

0.6

3.1

1.9

0.5

3.1

1.5

-

0.1

0.1

42.3

-

0.2

52.3

3.7

Accelerating Early Age Properties of Ultra-Low Clinker Cements

89

2.2 Methods Flow table and rheology measurements were used to evaluate printability of the mortar samples whilst isothermal calorimetry and setting time provided an evaluation of structuration and the microstructure. Table 2 shows the investigated mixes. In preparing the ternary cement, the slag to limestone ratio was kept at 3.8:1 and the sulfate content adjusted to 3% using natural anhydrite. 100 90 80

% Passing

70 60 50

C S L Anhy

40 30 20 10 0 0.1

1

10

100

1000

Particle size,µm

Fig. 1. PSD of constituent materials as measured by laser diffraction

Pastes and mortar samples were made at 0.4 water to binder (w/b) ratio and 0.3% plasticizer content. A vortex mixer was used to prepare 9g paste samples for calorimetry. A planetary mixer was used to mix samples for setting, rheology and strength testing based on modified EN 196–1 procedure. The mortar samples were made with 1:2 binder:sand ratio. The plasticizer was added to the mixing water about 5 min before preparing the paste or mortar mix. Table 2. Composition of investigated mixes per 100 g of paste. Mix ID

CEM I

GGBS

Anhydrite

Accelerator

Plasticizer

w/b

C

100

-

-

-

0.3

0.4

C-Rp

100

-

-

3.5*

0.3

0.4

CSL

30

53.5

3.5

-

0.3

0.4

CSL-Rp

30

53.5

3.5

3.5*

0.3

0.4

3.5

3.5†

0.3

0.4

CSL-Es

30

53.5

Note: * is SikaRapid-1 accelerator and † is calcium sulfoaluminate cement

The liquid accelerator or calcium sulfoaluminate powder was added to the plain mortar and mixed at high shear for 30s before the measurements. Flow table and viscometry were measured on independently prepared samples; the measurement performed within 5 min after the high shear mixing. Flow table was measured using an automatic jolter conforming to EN1015–3. Photographs were taken after removing the mould and after

90

R. Rusike et al.

15 jolts. A HAAKE viscometer IQ apparatus was used for rheology measurement. After placing into the 60 mm diameter test container, the shear rate was linearly increased from 0.1 to 100 s−1 in 100 s and held for 5 s. Subsequently, the rate was decreased to 0.1 s−1 in 100 s. Static yield stress was taken as the peak stress in the shear stress versus shear rate plot whilst the Bingham model was fitted to the stabilized portion of the curve, the intercept taken as dynamic yield stress [8, 9]. Setting time was measured with an Automatic Vicat apparatus according to a modified EN 196–3 procedure but at 0.4 w/b ratio.

3 Results and Discussion Flowability and shape retention govern printability of cementitious materials. The flow table test indirectly evaluates these and have been used elsewhere to assess buildability [6]. Figure 2 shows representative photographs of the fresh mortars immediately after removing the conical moulds (i.e. 0 jolts) and after 15 jolts. With the commercial accelerator, the CEM I mortar (C-Rp) retained its shape upon removal of the cone and deformed minimally after 15 jolts. For the limestone ternary blend (CSL), addition of both accelerators improved the initial shape retention compared to the plain ternary cement mortar i.e. without the added accelerator. Despite improved shape retention, significant spread was noticed upon jolting when the commercial accelerator was added (i.e. CSL-Rp). Conversely, with the sulfoaluminate cement addition, shape retention before and after jolting was comparable to the reference C-Rp mortar. C-Rp

τS(Pa) τD(Pa)

1370 53.2

CSL

100 18.3

CSL-Rp

550 35.4

CSL-Es

2520 58.5

Fig. 2. Photograph of flow table test showing the influence of cement and accelerator types on flowability; and yield stresses of investigated mortars.

Also summarized in Fig. 2 are the rheological parameters, characterized in terms of the static (τS ) and dynamic (τD ) yield stresses. Lower dynamic yield stresses are requisite for extrudability, but higher static yield stresses facilitate buildability. Without the added accelerator, lower static and dynamic yield stresses were observed in the plain and

Accelerating Early Age Properties of Ultra-Low Clinker Cements

91

ternary mortars, both of which increased with the commercial accelerator but to a lesser extent in the composite cement than CEM I. For the ternary cement, flow resistance was greater in the calcium sulfoaluminate mix than the commercial accelerator. An inverse relationship between yield stresses and flowability is well established. The observed dynamic yield stresses are within the ranges reported in Yuan et al. [8] but an order of magnitude lower than [4], potentially due to lower w/b ratios and fibre inclusions in the latter [10]. Meanwhile, the sulfoaluminate accelerator improved dynamic yield stress to comparable level as the reference but the static yield stress was higher. Based on the static yield stress ranges for printability suggested in [4], mixes C-Rp and CSL-Es can be considered printable by extrusion. Isothermal calorimetry (Fig. 3) gives insight into kinetic factors controlling structuration. Compared to the ternary cement, more heat was generated in the CEM I pastes, implying faster structuration. The commercial accelerator shortened the dormant period in the CEM I (i.e. C-Rp), but the silicate reaction rate was comparable to the plain cement mix, C. This suggests the accelerator did not modify the silicate reaction rate significantly. However, the aluminate reaction shoulder was not seen, suggesting that its reaction plausibly preceded the silicate reaction, thus explaining the shortened dormant period. Meanwhile, the commercial accelerator did not alter onset of the silicate peak in CSL-Rp, but accelerated aluminate reactions. The sulfoaluminate cement however accelerated silicate and aluminate peaks in CSL-Es profoundly to a comparable extent to that of calcium aluminate cement on OPC, as reported in [11]. Irrespective of the cement, accelerated hydration led to lower cumulative heat at 48 h, suggesting that overall hydration would be lower in the accelerated mixes.

Cum. Heat, J/g paste

Heat, W/g of paste

0.002

200 150

Initial Final

600

100 50 0 0

6

12

18

24

30

36

42

48

0.001

Setting time, mins

250

0.003

700

C C-Rp CSL CSL-Rp CSL-Es

Si peak Al peaks

500 400 300 200 100

0.000 0

6

12

18

24

30

Time, hours

36

42

48

0 C

C-Rp

CSL

CSL-Rp CSL-Es

Fig. 3. Effect of cement and accelerator type on heat Fig. 4. Effect of cement and of reaction and cumulative heat (insert) accelerator type on setting time

Setting time provides further indication of microstructure development at early age and approximately defines the open-time during extrusion [1, 3]. The results in Fig. 4 confirm faster structure build-up in both cements when an accelerator (e.g. commercial of

92

R. Rusike et al.

sulfoaluminate cement) was added in the mix. The calcium sulfoaluminate-based accelerator enhanced structuration in the ternary cement more effectively than the commercial nitrate/thiocyanate accelerator. Concluding Remarks The rheology, kinetics of hydration and setting time results showed that the commercial sodium nitrate/thiocyanate accelerator was effective in CEM I mortar, but less effective in the ternary cement mortar. In contrast, a calcium sulfoaluminate cement is a highly effective accelerator for the ternary cement mortar. These results imply that small additions of calcium sulfoaluminate makes it possible to achieve printable mortars with ultra-low clinker composite cements. Further work on controlling setting time in the sulfoaluminate-ternary systems are ongoing. Acknowledgements. S. Adu-Amankwah would like to thank The Royal Society for funding this research through Grant number RGS\R1\211454. R. Rusike is grateful to the Laidlaw Foundation for sponsoring her internship on this project. Participation in this study of S.A. Bernal was sponsored by the Engineering and Physical Sciences Research Council (EPSRC) grants EP/S019650/1 and EP/R001642/.

References 1. Buswell, R.A., Leal de Silva, W.R., Jones, S.Z., Dirrenberger, J.: 3D printing using concrete extrusion: a roadmap for research. Cement and Concrete Res. 112, 37–49 (2018) 2. Wangler, T., Lloret, E., Reiter, L., Hack, N., Gramazio, F., Kohler, M., et al.: Digital concrete: opportunities and challenges. RILEM Tech. Lett. 1, 67–75 (2016) 3. Le, T.T., Austin, S.A., Lim, S., Buswell, R.A., Gibb, A.G.F., Thorpe, T.: Mix design and fresh properties for high-performance printing concrete. Mater. Struct. 45(8), 1221–1232 (2012) 4. Rahul, A.V., Santhanam, M., Meena, H., Ghani, Z.: 3D printable concrete: mixture design and test methods. Cement Concr. Compos. 97, 13–23 (2019) 5. Kolawole, J.T., Combrinck, R., Boshoff, W.P.: Measuring the thixotropy of conventional concrete: the influence of viscosity modifying agent, superplasticiser and water. Constr. Build. Mater. 225, 853–867 (2019) 6. Zhang, C., Nerella, V.N., Krishna, A., Wang, S., Zhang, Y., Mechtcherine, V., et al.: Mix design concepts for 3D printable concrete: a review. Cement Concr. Compos. 122, 104155 (2021) 7. Chaves Figueiredo, S., Romero Rodríguez, C., Ahmed, Z.Y., Bos, D.H., Xu, Y., Salet, T.M., et al.: An approach to develop printable strain hardening cementitious composites. Mater. Des. 169, 107651 (2019) 8. Yuan, Q., Zhou, D., Li, B., Huang, H., Shi, C.: Effect of mineral admixtures on the structural build-up of cement paste. Constr. Build. Mater. 160, 117–126 (2018) 9. Roussel, N.: Rheological requirements for printable concretes. Cem. Concr. Res. 112, 76–85 (2018) 10. Marchon, D., Kawashima, S., Bessaies-Bey, H., Mantellato, S., Ng, S.: Hydration and rheology control of concrete for digital fabrication: potential admixtures and cement chemistry. Cem. Concr. Res. 112, 96–110 (2018) 11. Reiter, L., Wangler, T., Anton, A., Flatt, R.J.: Setting on demand for digital concrete – principles, measurements, chemistry, validation. Cem. Concr. Res. 132, 106047 (2020)

Developing Printable Fly Ash–Slag Geopolymer Binders with Rheology Modification Tippabhotla A. Kamakshi(B) and Kolluru V. L. Subramaniam Indian Institute of Technology Hyderabad, Hyderabad, Telangana, India [email protected], [email protected]

Abstract. The rheology of mixtures of fly ash-slag geopolymers, optimized for strength is not favorable for printing. Rheology modification is required using additives, which provide specific improvements in yield stress and thixotropy. These binders typically exhibit a pseudo-yield type behavior with a continuously deformable response under applied stress. A printable (both extrudable and buildable) material requires a yield-type behavior and adequate thixotropy, which can be brought out by addition of clay and Carboxymethyl Cellulose (CMC). The modification in rheology is attempted using commonly available Kaolinite clay. Specific changes in rheology caused due to the rheology modifiers are evaluated and are related with the performance in printing. Addition of clay contributes to an increase in the stiffness of the paste and improves buildability of the mix. A synergy between clay and CMC is established for proper printability. Clay in combination with CMC increases the storage modulus and produces a yield type behavior. CMC improves flocculation of clay but delays buildup due to its negative influence on reaction kinetics. Excess CMC increases the resistance to flow and produces a continuously deformable Maxwell response, which is not suitable for buildability. Keywords: Geopolymer · Rheology · Clay · Thixotropy · Yield stress · Material formulation

1 Introduction Alkali-activated fly ash and slag (AAFS) binder pastes exhibit complex rheology depending on the blend of source materials and activating solution. In the case of AAFS, the activating solution is an alkaline solution with or without dissolved silica [1–5]. The paste compositions are often determined based on strength and are not suitable for extrusionbased 3D printing. Extrusion-based 3D printing requires the material to behave like a fluid under pressure (pumpability) and also requires a solid-like response in its rest state (structure build-up and shape retention). The paste composition that gives adequate strength does not exhibit adequate yield stress and thixotropy required for printing. Thus, the rheology of the basic AAFS paste needs to be altered to achieve high yield stress and thixotropy.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 93–98, 2022. https://doi.org/10.1007/978-3-031-06116-5_14

94

T. A. Kamakshi and K. V. L. Subramaniam

Commonly used rheology modifiers for changing a specific aspect of printability include clay, lime powder, silica fume, viscosity modifying agent (VMA), and Carboxymethyl Cellulose (CMC). Clay has been shown to significantly improve the yield stress and thixotropy [6–8]. Silica fume enhances the yield stress [8–10]. Lime improves the yield and thixotropy [11, 12]. Viscosity modifying Agents (VMS) and CMC are used for controlling the viscosity [9, 13]. The ideal requirements for printing have been established using cement-based binders. However, the same in the case of AAFS pastes are not yet determined. This paper attempts to relate the rheology with printing performance of AAFS binder pastes. Specific improvements in rheology achieved with the use of CMC and clay are evaluated and related with shape retention and buildability in extrusion-based 3D printing.

2 Materials and Methods The AAFS paste was made by mixing source materials with an activating solution made with sodium hydroxide. The fly ash and slag in the source materials were taken in equal mass proportions. The basicity of the activating solution was equivalent to 3M. Low calcium fly ash and ground granulated blast furnace slag (henceforth referred to as slag) were used as source materials. Class F Siliceous fly-ash found in the Indian sub-continent procured from local thermal power station confirming to IS 3812–2003 [14] and slag procured from JSW suppliers confirming to IS 12089–1987 [15] were used as the Raw materials. Kaolinite clay and CMC were used as the rheology modifiers in the present study. A commercially available CMC which was available as an aqueous solution with a viscosity of 60 cps was used. Laboratory grade NaOH was used and 3M NaOH has a viscosity of 0.05cps. The oxide compositions of the source materials and Kaolinite clay are presented in Tables 1 and 2, respectively. Table 1. Oxide composition (mass %) of source materials CaO Fly ash Slag

SiO2

Al2 O3

Fe2 O3

MgO

K2 O

SO3

TiO2

4.06

55.66

26.00

5.58

1.92

2.80

0.94

1.82

43.12

29.71

16.41

0.89

4.73

0.76

2.17

1.08

Table 2. Oxide composition (mass %) of kaolinite Na2 O

MgO

Al2 O3

SiO2

P2 O5

SO3

Cl

K2 O

CaO

TiO2

Fe2 O3

1.57

1.15

36.16

44.37

0.48

0.73

0.47

0.35

0.6

7.18

3.68

3 Results and Discussion The basic blend consisting of 50% slag and 50% fly ash with a solution to binder of 0.4 gives a 28-day compressive strength of around 26 MPa. However, the mix by itself is not

Developing Printable Fly Ash–Slag Geopolymer Binders

95

suitable for extrusion-based 3D Printing. To make it printable, rheology modification of yield stress and thixotropy is attempted with additions of commonly available Kaolinite clay and CMC as shown in Table 3. An extrusion-based 3D-Printer manufactured by 3D Potter has been used in the present study. In the print setup, addition of clay has been shown to arrest water migration and improve buildability. CMC was added for improving shape retention and buildability of the print. Clay and CMC contents were varied to arrive at the optimal consistency for the printing purpose. Table 3. Mix matrix Mix no

M0

M1

M2

M3

M4

FA

1

1

1

1

1

Slag

1

1

1

1

1

Kaolinite

0

0

0.16

0.16

0.16

CMC

0

0.005

0.005

0

0.007

Sol/Binder

0.33

0.33

0.33

0.33

0.33

Yield stress (Pa)

12

-

798

-

-

Apparent yield stress (Pa)

-

30

-

218

1514

Viscosity (Pa.s)

57

111

3100

515

8550

Initial set time (min)

80

100

165

90

160

Final set time (min)

130

150

215

135

205

The rheology of the AAFS binder pastes obtained from the constant strain rate test is shown in Fig. 1a. The shear resistance measured from the AAFS binder paste with CMC addition, M1, exhibits a very low resistance to flow. Under applied strain, the shear resistance of M1 increases and attains a constant value. The paste exhibits continuously deforming response, which is classified as a Maxwell-flow response [16]. The maximum stress in the response is identified as apparent yield stress [17]. The addition of clay to the AAFS paste increases the peak stress in the Maxwell flow response in M3. The Maxwell flow is produced by a continuous rearrangement of particles, which remain in a dispersed state to produce a continuously deforming response. The measured shear resistance in a Maxwell flow response to increasing shear strain increases continuously to attain a constant value at steady state flow. Addition of clay and CMC in M2 produces a further increase in the peak stress. More importantly, a yield type behavior is produced. The shear stress increases to a peak stress and with increasing strain, decreases to a steady state value. Yield stress is the peak resistance offered by an internal percolated network of solids particles within the paste. The disruption of the internal network of particles produces flow. There is an improvement in the rheology with a yield-type behavior produced by clay due to its dominant colloidal forces leading to rapid re-flocculation in the presence of CMC. The CMC enhances the flocculation of the clay leading to a fundamental change in

96

T. A. Kamakshi and K. V. L. Subramaniam

the behavior to a yield-type response. On increasing the CMC content, a Maxwell type response with a high apparent yield stress is produced in M4. The storage moduli of AAFS pastes with variation in clay and CMC are shown in Fig. 1b. The thixotropy is shown in the increase in the storage modulus with time. The increase in the storage modulus with time gives the structural build-up and is influenced by both flocculation effect and reactions within the paste, which produce a progressive increase in the elastic stiffness of the mixture. The initial elastic stiffness of the paste and its increase is influenced by the addition of CMC and clay. The addition of clay increases the initial stiffness in M3. The clay also showed a better structural build-up, which gets delayed with the addition of CMC. The use of CMC with clay in M2 produces a delay in structural buildup compared to M3. In mixtures with high CMC content, M4, there is a further delay in the buildup. The influence of CMC on rapid re-flocculation is evident in the high initial stiffness of pastes containing both clay and CMC. The CMC interferes with the reaction kinetics in the AAFS paste, which dominates the reaction procedure when added in higher quantities. The influence of CMC on the reaction kinetics is also seen in the increasing setting times given in Table 3. 1.6 K_M1 K_M3

Shear Stress (Pa)

1200

K_M5 K_M7

900 600

Storage Modulus (MPa)

1500

K_M1 K_M3 K_M5 K_M7

1.4 1.2 1 0.8 0.6 0.4

300

0.2 0

0.5

10.5 Step time (seconds)

(a)

0 10

100 Time (minutes)

(b)

Fig. 1. (a) Constant strain-rate response, (b) Increase in storage modulus with time.

Printing Performance The shape retention was established by printing a filament of length 12 mm and thickness 10 mm. The filament width was measured after 10 min and a spread of less than 2mm was considered to have a good shape retention [9]. Buildability was established by measuring the top width and the bottom width after the completion of the 10-layered print. The shape retention was achieved in pastes which exhibited less than 5 mm difference in the widths between the topmost and bottommost layers. The printing performance of the mixes is shown in Fig. 2.

Developing Printable Fly Ash–Slag Geopolymer Binders

97

Clay was found to help in arresting water migration and was responsible for buildability. Although M3 exhibits rapid structural buildup, the Maxwell type response does not result in shape retention. CMC was found to provide shape retention and buildability when used with clay. The yield type response with rapid structure buildup in M2 containing clay and CMC produced shape retention and buildability. CMC added in larger quantities (M4) affected the buildability due to its shift towards the Maxwell-type behavior and delayed structural buildup.

Mix Buildability Shape retention

M1 No No

M2 Yes Yes

M3 No No

M4 No Yes

Fig. 2. Printing performance of mixes

3.1 Summary and Findings In terms of rheology, clay was found to improve the yield stress and CMC was found to significantly increase the viscosity. CMC used with clay increases stiffness but delays the structural buildup. The findings of this study are summarized below. 1. The AAFS paste proportioned for strength does not possess adequate rheology suitable for printing. The yield stress and thixotropy are inadequate for shape retention and buildability. 2. It is essential to distinguish between Maxwell and yield type rheological responses of the AAFS paste. A Maxwell type continuously deformable response does not allow buildability even with a high resistance to applied shear strain. A yield-type response obtained from an internal structure within the paste is required for shape retention. 3. Stiffness of the paste required for shape retention is produced by the flocculation of clay. The internal structure within the AAFS paste is enhanced by the CMC, which produces a yield type response, which is required for shape retention in the printed paste. 4. CMC affects the kinetics, which influences the structure buildup within the paste. When added in lower quantities, the enhancement in the internal structure promoted by CMC offsets the delay in structure buildup to allow both shape retention and buildability.

98

T. A. Kamakshi and K. V. L. Subramaniam

References 1. Bhagath Singh, G.V.P., Subramaniam, K.V.L.: Evaluation of sodium content and sodium hydroxide molarity on compressive strength of alkali activated low-calcium fly ash. Cem. Concr. Compos. 81, 122–132 (2017). https://doi.org/10.1016/j.cemconcomp.2017.05.001 2. Reddy, K.C., Subramaniam, K.V.L.: Investigation on the roles of solution-based alkali and silica in activated low-calcium fly ash and slag blends. Cem Concr Compos 123, 104175 (2021). https://doi.org/10.1016/j.cemconcomp.2021.104175 3. Reddy, K.C., Gudur, C., Subramaniam, K.V.L.: Study on the influences of silica and sodium in the alkali-activation of ground granulated blast furnace slag. Constr. Build. Mater. 257, 119514 (2020). https://doi.org/10.1016/j.conbuildmat.2020.119514 4. Rafeet, A., Vinai, R., Soutsos, M., Sha, W.: Effects of slag substitution on physical and mechanical properties of fly ash-based alkali activated binders (AABs). Cem. Concr. Res. 122, 118–135 (2019). https://doi.org/10.1016/j.cemconres.2019.05.003 5. Phoo-Ngernkham, T., Maegawa, A., Mishima, N., et al.: Effects of sodium hydroxide and sodium silicate solutions on compressive and shear bond strengths of FA-GBFS geopolymer. Constr. Build. Mater. 91, 1–8 (2015). https://doi.org/10.1016/j.conbuildmat.2015.05.001 6. Kawashima, S., Chaouche, M., Corr, D.J., Shah, S.P.: Influence of purified attapulgite clays on the adhesive properties of cement pastes as measured by the tack test. Cem. Concr. Compos. 48, 35–41 (2014). https://doi.org/10.1016/j.cemconcomp.2014.01.005 7. Yuan, Q., Li, Z., Zhou, D., et al.: A feasible method for measuring the buildability of fresh 3D printing mortar. Constr. Build. Mater. 227, 116600 (2019). https://doi.org/10.1016/j.con buildmat.2019.07.326 8. Zhang, Y., Zhang, Y., She, W., et al.: Rheological and harden properties of the high-thixotropy 3D printing concrete. Constr. Build. Mater. 201, 278–285 (2019). https://doi.org/10.1016/j. conbuildmat.2018.12.061 9. Kondepudi, K., Subramaniam, K.V.L.: Formulation of alkali-activated fly ash-slag binders for 3D concrete printing. Cem. Concr. Compos. 119, 103983 (2021). https://doi.org/10.1016/ j.cemconcomp.2021.103983 10. Panda, B., Paul, S.C., Tan, M.J.: Anisotropic mechanical performance of 3D printed fiber reinforced sustainable construction material. Mater. Lett. 209, 146–149 (2017). https://doi. org/10.1016/j.matlet.2017.07.123 11. Alghamdi, H., Nair, S.A.O., Neithalath, N.: Insights into material design, extrusion rheology, and properties of 3D-printable alkali-activated fly ash-based binders. Mater. Des. 167, 107634 (2019). https://doi.org/10.1016/j.matdes.2019.107634 12. Baz, B., Remond, S., Aouad, G.: Influence of the mix composition on the thixotropy of 3D printable mortars. Mag. Concr. Res., 1–34 (2020). https://doi.org/10.1680/jmacr.20.00193 13. Figueiredo, S.C., Rodríguez, C.R., Ahmed, Z.Y., et al.: An approach to develop printable strain hardening cementitious composites. Mater. Des. 169, 107651 (2019). https://doi.org/ 10.1016/j.matdes.2019.107651 14. IS: 3812 (Part-1): Pulverized fuel ash — specification. Part 1: For use as Pozzolana in cement, Cement Mortar and Concrete (Second Revision). Bur Indian Stand, pp. 1–14 (2003) 15. Bureau of Indian Standard: IS:12089–1987: Specification for granulated slag for the manufacture of Portland slag cement. BIS, New Delhi, pp. 1–14 (1987) 16. Gadkar, A., Subramaniam, K.V.L.: An evaluation of yield and maxwell fluid behaviors of fly ash suspensions in alkali-silicate solutions. Mater. Struct. 52(6), 1–10 (2019). https://doi.org/ 10.1617/s11527-019-1429-7 17. Kondepudi, K., Subramaniam, K.V.L.: Extrusion-based three-dimensional printing performance of alkali-activated binders. ACI Mater. J. 118. https://doi.org/10.14359/51733107

Formulation and Characterization of a Low Carbon Impact Cementitious Ink for 3D Printing Estelle Hynek1,2(B) , David Bulteel1,2 , Antoine Urquizar3 , and Sébastien Remond4 1 ULR 4515 - LGCgE – Laboratoire de Génie Civil et géoEnvironnement, Univ. Lille, Institut

Mines-Télécom, Univ. Artois, Junia, 59000 Lille, France [email protected] 2 Centre for Materials and Processes, IMT Nord Europe, Institut Mines-Télécom, 59000 Lille, France 3 Constructions 3D, Bruay-sur-l’Escaut, France 4 INSA-CVL, LaMé – EA7494, Univ. Orleans, Univ. Tours, 8 Rue Léonard De Vinci, 45072 Orléans, France

Abstract. 3D printing is considered to be the most innovative construction technique. A challenge encountered with this technique is the ecological impact of the used mortar (ink). Indeed, due to the small particle size of the used sand, inks contain large amounts of Portland cement, which increases the carbon footprint of the material. This study is carried out in a partnership with Constructions 3D company, it concerns the design and characterization of a cementitious ink with a high cement substitution rate in order to reduce its environmental impact. This mortar is made from a reference ink composed of cement. We chose to work with constant workability based on the slump flow test and an equivalent Water/Binder ratio for the two mortars by adapting the admixture. A fresh state characterization allows evaluating the influence of the materials on the buildability of the mortars (fall cone test) and a hardened state characterization allows evaluating their influence on the mechanical aspect (compressive strength at 2, 7, 28 and 90 days, and shrinkage). Finally, full-scale prints are made in order to confirm the printability of this low carbon impact ink. Although 70% of the cement is replaced by metakaolin and ground granulated blast-furnace slag, the mortar produced is not less effective than the reference mixture and fully meets the specifications imposed by the 3D printing technique. The ternary ink formulated in this way will be tested on site to assess its properties in real conditions. Keywords: 3D printing · Low-carbon cementitious inks · Workability · Buildability · Mechanical characterization

1 Introduction This study is carried out in partnership with the Constructions 3D company whose goal is to sell large scale concrete printers. For this, the M-Tech Duomix 2000 printing system © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 99–104, 2022. https://doi.org/10.1007/978-3-031-06116-5_15

100

H. Estelle et al.

is used. This process requires the use of premix mortars and the incorporation of solid additives to the mixture. The objective of this study is to develop a low carbon printable mortar with reduced shrinkage compared to a totally cement-based ink. We chose to substitute the latter with a mixture of metakaolin and ground granulated blast furnace slag (GGBFS). These two materials were chosen for their ability to compensate for the weaknesses of each other and thus to reduce the amount of cement. Indeed, the GGBFS allows to maintain correct mechanical properties and a workability time adapted to 3D printing while the metakaolin helps to limit the shrinkage of mortars [1–4].

2 Materials and Methodology 2.1 Materials The two mixes were designed using an Ordinary Portland Cement Type 1 (CEM I 52.5 N), having a density of 3.16 g/cm3 and 8.2 μm median particle diameter “D50 ”, a GGBFS with a density of 2.90 g/cm3 and 11 μm D50 , a flash metakaolin with a density of 2.63 g/cm3 and 39 μm D50 , a crushed limestone sand having a density of 2.70 g/cm3 and a particle size distribution of 0 to 2 mm including 19% smaller than 63 μm. For admixtures, CHRYSO® high range water reducer (called HRWR), CHRYSO® viscosity modifying agent (named VMA-1) and CHRYSO® rheological stability agent (named VMA-2) were used. In order to adapt the cementitious inks to the printing process used by the partner, only solid admixtures were employed in this study to fit for the different mixing systems used. 2.2 Mix Design A reference ink which binder is composed of cement only is first designed in the laboratory. The Marsh cone test [5] is used to determine the superplasticizer demand of the mortar. In accordance with the supplier’s recommendations, the quantities of the combined VMAs are limited to 0.3% of the binder mass. The amount of water is adjusted in order to obtain a laboratory printable mortar, i.e. one which can be extruded without blocking and buildable [6]. A glue gun is used to simulate the extrudability and printability of inks. The reference ink, noted “Reference”, is then used to obtain a target spread value. The slump flow tests are carried out in accordance with the NF-EN 1015–3 standard directly at the end of mixing. Based on another study, the ternary ink, noted “Ternary”, is composed of 30% cement, 50% metakaolin and 20% GGBFS. Previous work has led to the conclusion that those percentages can limit the shrinkage of the mortar with metakaolin and compensate the mechanical strength reduction with GGBFS. Thus the formulated ink has a much lower carbon impact than an ink containing only cement. The volume proportions of binder and sand are kept equivalent to those of the reference mortar. The Water/Binder ratios are kept equal for the two mortars. The amount of superplasticizer was adjusted to give the same 15 shots spread value as the reference mortar. The ink is considered to be laboratory printable thanks to the gun simulation test. The compositions of the two mortars are presented in Table 1.

Formulation and Characterization of a Low Carbon Impact Cementitious Ink

101

Table 1. Mix composition of the mortars, in kg/m3 . Mortar

Cement

Reference

863

Ternary

259

Slag

Meta-kaolin

158

360

Sand

W/B

%HRWR

%VMA-1

%VMA-2

Spread (mm)

1076

0.38

0.40

0.1

0.2

146

1076

0.38

0.59

0.1

0.2

147

2.3 Experimental Procedures Fall cone test. Based on the ISO 17892–6 standard, the fall cone test is performed according to the method proposed by Baz et al. [7] in order to estimate the buildability of the studied printable mortars over time. After 30s of rest, measurements are taken every 150s. As shown by Estellé et al. [8], a yield stress value τ0 is obtained for each time period, allowing to determine a value of the structuration rate Athix . This yield stress value increases gradually when the material is left at rest. Roussel [9] proposes a linear relationship between the yield stress τ0, the structuration rate Athix and the resting time t [10, 11] as shown in Eq. (1), where τ0,0 is the initial yield stress. τO (t) = τ0,0 + Athix ∗ t

(1)

Mechanical performance. Flexural and compressive strength tests are carried out for both mortars at 2, 7, 28 and 90 days to assess the influence of the mineral additions combination on the ink mechanical performance. These tests were carried out in accordance with NF EN 196–1 standard. Shrinkage. The monitoring of the different types of shrinkage of the studied mortars is carried out according to the NF P15–433 standard. The specimens are kept in an environment at 20 °C and 50%RH to determine the total shrinkage. Autogenous shrinkage is also measured by insulating the mortar specimens with adhesive aluminum foil and using the same test conditions as for the total shrinkage monitoring. From these measurements, it is possible to deduce the drying shrinkage of the two mortars by assuming that the total shrinkage is equal to the sum of the autogenous and drying shrinkage [12]. 3D printing tests at full scale. A cartesian printer is used for testing the mortars at full scale. A cylinder of 150mm of radius is chosen to print the two mortars, using only one batch of 50L. The reference is printed first in order to set the speed of the printer leaving a pause time of 8 s between two superimposed layers. The pump flow rate is adapted to each mortar in order to obtain a layer width of about 5 ± 1cm.

3 Results and Discussion 3.1 Fresh State Properties Fall cone test. The fall cone tests were carried out three times for each mortar and the results, based on the mean values, are shown in Fig. 1. Tests were carried out over a

102

H. Estelle et al.

period of about 25 min but the Athix calculations are made on 930 s by neglecting the initial yield stress τ0,0 , period after which the relationship between yield stress and the structuring rate is no longer linear. Indeed despite triplicate testing, the dispersion of measurements is significant and is due to the simple experimental method used to estimate the buildability of the inks. The ternary ink appears to be slightly less thixotropic than the reference ink because some of the reactive clinker particles have been substituted by metakaolin and GGBFS particles. However, it shows that the stress increases over time and it would appear that this substitution of cement by metakaolin and GGBFS only slightly affects the fresh state global behavior of the mortar because even after 1000 s the two mortars have the same behavior. This means that the ternary ink is almost as buildable as the reference ink.

Fig. 1. Comparison of reference and ternary inks through fall cone test results.

3.2 Hardened State Properties Mechanical performance. As shown by the mean values in Table 2, the flexural and compressive strengths of the ternary ink are lower than those of the reference ink. This is due to the lower reactivity of the mineral additions compared to the cement which contains clinker particles. Table 2. Comparison of flexural and compressive strength of reference and ternary mortars at 2, 7, 28 and 90 days. Mean values are presented with standard deviation values. Mortar

Flexural strength (MPa) 2 days

7 days

Reference 9.1 ± 0.5 13 ± 2 Ternary

Compressive strength (MPa) 28 days 90 days 2 days

7 days

28 days 90 days

14 ± 2

15 ± 1

43 ± 3 50 ± 5 57 ± 1

58 ± 5

5.9 ± 0.5 10.3 ± 0.4 11 ± 1

10 ± 1

12 ± 1 29 ± 3 36 ± 4

45 ± 2

Shrinkage. As shown in Fig. 2, autogenous shrinkage is roughly the same for both inks. The total shrinkage, mainly due to drying shrinkage, is strongly reduced when

Formulation and Characterization of a Low Carbon Impact Cementitious Ink

103

the cement is substituted by metakaolin and GGBFS. All the results obtained are from triplicates.

Fig. 2. Total and autogenous shrinkage of ternary and reference inks.

3.3 3D Printing Tests 3D printing tests at full scale. The results are shown in Fig. 3. The complex shape of the mortar and the short inter-layer interval time cause unfavorable conditions for printing. In spite of these drastic printing conditions and as predicted by the laboratory test, both mortars are printable. For both prints, the cylinder diameter is reduced at the end of the test due to a lack of material. The printed reference mortar shown in Fig. 3a is the result of another study and has a W/B of 0.3725, i.e. and a slightly lower amount of water. The ternary ink, although containing a higher quantity of water which should penalize the buildability [6], is also printable under these particular printing conditions. To conclude, the substitution of 70% cement with GGBFS and metakaolin does not seem to prevent quick the ternary ink printing of a relatively narrow object with 50 layers of 1 cm in height. Furthermore, this is a more eco-friendly ink.

Fig. 3. Full scale printed mortars. Figure a: reference, b: ternary.

104

H. Estelle et al.

4 Conclusion The substitution of 70% cement by 50% metakaolin and 20% GGBFS leads to a less thixotropic ink, which nevertheless allows a narrow object to be printed quickly over a height of 50 cm, which is more than sufficient for the intended application. The mechanical strengths up to 28 days, in flexion and compression, are negatively affected by the high substitution rate. However, the strengths obtained remain quite correct for a printable mortar. Moreover, compared to the reference ink, the compressive strengths obtained at 90 days are much higher than at 28 days for the ternary ink. This is due to the pozzolanic effect of the mineral admixtures. Finally, the total shrinkage, mainly due to drying shrinkage, is improved when the cement is replaced by metakaolin and ground granulated blast-furnace slag from an early age. However, this ternary mixture does not seem to positively affect the autogenous shrinkage values.

References 1. Perlot, C., Rougeau, P.: Intérêt des métakaolins dans les bétons. In: Les éditions du CERIB (2007) 2. Gleize, P.J.P., Cyr, M., Escadeillas, G.: Effects of metakaolin on autogenous shrinkage of cement pastes. Cement Concr. Compos. 29, 80–87 (2007) 3. Wei, Y., Hansen, W., Biernacki, J.J., Schlangen, E.: Unified shrinkage model for concrete from autogenous shrinkage test on paste with and without groundgranulated blast-furnace slag. ACI Mater. J. 108, 13–20 (2011) 4. Lee, K.M., Lee, H.K., Lee, S.H., Kim, G.Y.: Autogenous shrinkage of concrete containing granulated blast-furnace slag. Cem. Concr. Res. 36, 1279–1285 (2006) 5. De Larrard, F., Bosc, F., Catherine, C., Deflorenne, F.: La nouvelle méthode des coulis de l’AFREM pour la formulation des bétons à hautes performances. Bulletin des laboratoires des Ponts et Chaussées 202, 61–69 (1996) 6. Khalil, N.: Formulation et caractérisation chimique et rhéologique des mortiers imprimables en 3D à base de mélanges de ciments Portland et sulfoalumineux. Génie civil. Ecole nationale supérieure Mines-Télécom Lille Douai (2018) 7. Baz, B., Remond, S., Aouad, G.: Influence of the mix composition on the thixotropy of 3D printable mortars. Mag. Concr. Res., 1–13 (2021) 8. Estellé, P., Michon, C., Lanos, C., Grossiord, J.L.: De l’intérêt d’une caractérisation rhéologique empirique et relative (2015) 9. Roussel, N.: A thixotropy model for fresh fluid concretes: Theory, validation and applications. Cem. Concr. Res. 36, 1797–1806 (2018) 10. Roussel, N., Cussigh, F.: Distinct-layer casting of SCC: The mechanical consequences of thixotropy. Cem. Concr. Res. 38, 624–632 (2008) 11. Ovarlez, G., Roussel, N.: A physical model for the prediction of lateral stress exerted by self-compacting concrete on formwork. Mater. Struct. 39(286), 269–279 (2006) 12. Aitcin, J.C., Neville, A., Acker, P.: Les différents types de retrait du béton. Bulletin des laboratoires des Ponts et Chaussées 215, 41–51 (1998)

Strategies for Reducing the Environmental Footprint of Additive Manufacturing via Sprayed Concrete Aurélie Favier(B)

and Agnès Petit

Mobbot SA, 1700 Fribourg, Switzerland [email protected]

Abstract. Lately, the construction industry has been trying to reduce its environmental footprint with the help of an additive manufacturing. This emerging technique allows for an automated construction of free-of-form structural elements. This, in turn, enables optimization of geometrical dimensions during design. However, the 3D concrete printing (3DCP) based on extrusion of material requires high amount of Portland cement, additives, and consumes an already scarce fine sand. On the other hand, the additive manufacturing using spraying (S-3DCP) offers larger flexibility in material composition, allowing for further reduction of the carbon footprint. It also facilitates implementation of larger aggregates, which improve material mechanical stability during manufacturing process and ameliorate its flexural strength in the hardened state. In this work, strategies related to the circular economy and the reduction of CO2 emissions through the concrete recipe optimization are assessed and demonstrated. The first strategy is based on the use of low-CO2 cement, which can reduce CO2 emissions by about 40% compared to using CEM II. The second strategy is based on the design of elements, which is to reduce the thickness while satisfying all mechanical constraints, that is, to reduce the concrete by approximately 25%. Finally, this paper will show the possibility of using demolition waste to replace 45% of aggregate (i.e., 35% of 0–4 mm sand and 10% of 48 mm aggregate) and the potential circular economy of S-3DCP. Due to the use of high-speed spraying (shotcrete) as the material placement method, continuous and isotropic elements with a low environmental footprint are produced. Keywords: Shotcrete · CO2 footprint · Circular economy · 3D concrete printing

1 Introduction In Switzerland, 68 million tons of building materials are consumed per year. 90% of it originates from primary sources [1, 2]. However, massive quantities of construction waste are generated by the demolition of buildings and the repair and maintenance of roads. In addition to excavated material, which accounts for more than 15 million tons per year, mineral construction waste is by far the largest waste stream generated in Switzerland [1, 2]. It consists of materials such as concrete, sand, asphalt and masonry. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 105–110, 2022. https://doi.org/10.1007/978-3-031-06116-5_16

106

A. Favier and A. Petit

For many years, this construction waste has been dumped in landfills, even though it could be useful to the resource-intensive construction industry. Closing material cycles and recycling therefore make sense in this sector. The construction of resource-friendly buildings within the recycling system has been made possible in Switzerland by standards in line with current practice and by the willingness to undertake extensive research and development work. One example is the MB SIA 2030 Technical Code of Practice for the safe use of concrete containing recycled aggregates in concrete construction. Another example is the development of the resource-saving Portland cement ZN/D 32.5 R. This is a new cement that closes the cycle of building materials by reusing local high quality mixed granular materials. Currently, the use of recycled materials and low-carbon cement is still marginal in additive manufacturing, but it is promising and essential to conserve natural resources, save landfill space and reduce emissions.

2 Concrete Design and Performances 2.1 Concrete Recipes Materials In this study, we use three type of cement and 1 supplementary cementitious material: • Holcim Fluvio 5 CEM II/A-LL 52.5 N • Holcim Susteno 3 R (ZN/D 32.5 R) contains Portland clinker, recycled mixed gravel, calcined shale and gypsum. • CEM I 52.5 N • Limestone Nekafill 15. Two fractions of aggregates are used in the S-3DCP process: 0–4 mm and 4–8 mm. Recycled aggregates are building demolition waste from SRREC, 4–8 mm are certified recycled aggregates then 0–4 mm are usually not valorized at all. Natural aggregates are certified round aggregates from Buhler Marin. Mixes We developed a concrete recipe to address a specific application for a delivery time after 24 h. The benchmarking of the present recipes was done and optimized to match the same application. Several recipes (Table 1) were assessed but two recipes: one for each cement were selected based on the pumpability and sprayable for S-3DCP. As the workability is a key point, the concrete must have a same consistency (same slump and spread) and can have different Water to Cement ratio (W/C). In the case of recipes with recycled aggregates, the water demand is still close to the one for recipes without recycled aggregates.

Strategies for Reducing the Environmental Footprint

107

Table 1. Concrete mixes Recipe 1 rec Recipe 1

Recipe2 rec Recipe 2

Recipe REF

Cement type

CEM II A/L: CEM II A/L: Susteno 3 R Susteno 3 R CEM I + Calcite Fluvio 5 Fluvio 5

Cement quantity (kg/m3)

500

500

500

500

420 + 140

% Natural sand 0-4mm

40%

75%

75%

75%

70%

% Natural sand 4-8mm

15%

25%

10%

25%

30%

% Recycled sand 35% 0-4mm

0%

0%

0%

0%

% Recycled sand 10% 4–8 mm

0%

15%

0%

0%

Water /Cement

0.45

0.36

0.42

0.42

0.4

Plasticizer%

0.9

0.9

0.9

0.9

0.7

Retarder %

0.1

0.1

0.1

0.1

0.1

4–5%

4–5%

4–5%

4–5%

Activator % 4–5% (based on cement content)

2.2 Mechanical Performances For recipes 1 and recipes 2, only compression tests were done following the norm SN-EN 12390–3. The tests specimens were drilled in a wall done by Mobbot S-3DCP technology to a core of 50x50 mm and lightly surfaced. For recipe 3, the tests specimens were drilled in a wall done by Mobbot S-3DCP technology to a core of 50x110 mm. Table 2 shows the average values on six specimens per age and per recipe. The target value to be delivered at 24 h is between 10 to 15 MPa (defined based on lifting with anchors), the target value at 28 days is strength class C40-C50. All recipes with or without recycled materials fulfill our needs for a 24 h delivery and a good compressive strength at 28 days. Table 2. Average values for compressive strength at different age Recipe 1 rec

Recipe 1

Recipe 2 rec

Recipe 2

Recipe REF

Compressive strength 1d (MPa)

25.6

33.7

10.5

12

20

Compressive strength 3d (MPa)

44

55

30.2

34.2

33

Compressive strength 28d (MPa)

59.1

73.7

48.4

56.9

58

108

A. Favier and A. Petit

2.3 Durability • Water permeability (including porosity) following the norm SIA 262-1A on 5 cores 50x50 mm at 27–29 days for each recipe. For all recipes, the water permeability q is lower than 10 g/m2h so the concrete is considered waterproof. The porosity is lower than 20% for the recipe without recycled aggregates and slightly higher than 20% for the ones with recycled aggregates. The porosity is highly dependent and sensitive to the water content so it can explain the difference between recipe with and without recycled aggregates. • Carbonation test following the norm SIA 262 -1 I on prism 360X120x120 at 27– 29 days. 4 sections were put under 4% CO2 and were taken at 0, 7, 28,63 days to measure the carbonation front (see Fig. 1). All specimens can be classified as XC4.

Fig. 1. Carbonation front at 7 days on recipe 1 rec.

• Freeze thaw with deicing agent – Method TfB [3] on 2 cores 50x50 mm drilled at the bottom and at the top of the wall at 27–29 days for each recipe. In all samples, the microstructure remains intact and homogeneous. The presence of recycled aggregates does not seem to create additional deterioration, on the contrary.

3 Environmental Assessment In this study, efforts to reduce the ecological impact of concrete were approached either by: – using low carbon cements – reducing the amount of concrete used for an element – using recycled materials. We will look at two environmental assessment parameters: – the embodied CO2 emissions in kgCO2 eq/m3 of concrete and in kgCO2 eq/m3 of concrete/MPa. – the circularity index (MCI) which measures the rate of recycled materials in relation to the total quantity of materials. Water is excluded from this calculation. The method used to calculate the material circularity rate is inspired by Eurostat [4].

Strategies for Reducing the Environmental Footprint

109

Table 3 shows the two assessment parameters, the amount of embodied CO2 is controlled by the quantity of cement. Table 3. Values of environmental parameters for different concretes. Concrete jobsite is CAN A with CEM II A/LL Recipe 1 rec

Recipe 1

Recipe2 rec

Recipe 2

Recipe REF

Concrete jobsite [5]

kgCO2 /m3

341

342

303

303

325

218

kgCO2 /m3 /MPa1day kgCO2 /m3 /MPa28days

13

10

28

25

16

68

6

5

6

5

6

11

MCI

35%

0%

15%

4%

0%

0%

The impact of recycled aggregates on CO2 emission is negligible. However, nonenergy mineral resources are a major economic issue in Switzerland and a source of conflict regarding their availability and accessibility. Ten million m3 of demolition waste are produced per year, and if converted into recycled aggregates, they become economically and ecologically interesting. Case-Study: Optimization of Standard Swisscom Type Manhole We provide a comparison for a standard product which is a Swisscom-type chamber realized by S-3DCP printed with different recipes and two thicknesses and a traditional method (onsite casting). This way, we can emphasize that the different recipes also allow a design optimization and maintain a 24–72 h delivery time. Figure 2 shows the comparison between a chamber made by S-3DCP and by in situ casting. The chamber cast onsite will be done with higher thickness (average 25 cm because it is cast directly against the pit wall). By S-3DCP, we can reduce the CO2 emissions by 15% minimum.

Fig. 2. CO2 saved for an element (manhole) done with different thickness and concrete recipe. 0% correspond to the manhole done at the job site.

110

A. Favier and A. Petit

With the S-3DCP system and the high-performance concretes, the design of chamber can be optimized to 15 cm then to 12 cm. Moreover, the concrete recipe can be adjusted to reduce CO2 emissions (replacement of Fluvio by Susteno) and we save from 11 to 23% of CO2 .

4 Conclusions In conclusion, the incorporation of recycled sand and aggregates allows the development of pumpable and sprayable recipes. Up to 45% of recycled sand and aggregates can be used without significant deterioration of the mechanical properties at early age. Moreover, durability is not affected by recycled aggregates. By using this recycled concrete, it is still possible to deliver chambers within 24 h. The factor controlling CO2 emissions is the quantity of cement. This quantity is often imposed by the need for mechanical properties and the quantity of fines to be pumpable. By playing with low carbon cements like Susteno, substituting part of the cement with other fines and reducing the quantity of concrete, we can reduce the CO2 emissions by more than 23% of the elements compared to a traditional method. As explained before, the use of recycled aggregates does not have a significant impact on CO2 emissions but plays a key role in the circular economy of construction. The S-3DCP process with recycled concrete recipes allows for the production and delivery of strong, durable, and more environmentally friendly elements within 24 h. Acknowledgement. We thank Holcim Switzerland and Kickstart project to support actively this study.

References 1. Federal Bureau for Statistics. https://www.bfs.admin.ch 2. Federal Bureau for Environment. https://www.bafu.admin.ch/ 3. Kurt, H., Maher B.: Béton résistant au gel et aux sels de déverglaçage, Bulletin du ciment (1997) 4. Eurostats.: Indicators - Circular economy - Eurostat (europa.eu) 5. Koordinationskonferenz der Bau- und Liegenschaftsorgane der öffentlichen Bauherren KBOB, Betonsortenrechner für Planende (treeze.ch)

Mechanical Performance of 3-D Printed Concrete Containing Fly Ash, Metakaolin and Nanoclay Ahmed Abdalqader1(B) , Mohammed Sonebi2 , Marie Dedenis3 , Sofiane Amziane4 , and Arnaud Perrot5 1 Tracey Concrete Ltd., Enniskillen, UK

[email protected]

2 Queen’s University Belfast, Belfast, UK 3 Eiffage Construction, Vélizy-Villacoublay, France 4 Université Clermont Auvergne, Clermont-Ferrand, France 5 University of South Brittany, Lorient, France

Abstract. Similar to conventional concrete, the use of supplementary cementitious materials (SCMs) in 3-D printing concrete (3DPC) can be technically and environmentally beneficial. 3DPC consumes larger amount of Portland cement compared to conventional concrete so replacing the cement with SCMs will reduce the carbon footprint of 3DPC. However, it is important that the level of replacement does not lead to a significance loss of the mechanical performance of 3DCP. Therefore, this study investigates the mechanical strengths of 3DPC with various mix compositions. Mixes containing fly ash, metakaolin, nanoclay, and combination of them were studied. The test program includes measuring the compressive strength and flexural strength on standard cubes/prisms as well as 3D printed cubes/prisms at 7 and 28 days. The findings showed that fly ash decreased the early age strengths but led to an increase in the late age strengths. On the other hand, metakaolin and nanoclay did not remarkedly affect the strengths. It was found that mixes containing both fly ash and metakaolin showed better performance. Finally, it was observed that samples from 3D printed concrete had lower strengths compared to the same mixes poured in standard cubes and prisms. Optimisation of the mix compositions using these SCMs is required to achieve acceptable mechanical performance. Keywords: 3D printing concrete · Fly ash · Metakaolin · Nanoclay · Supplementary cementitious materials

1 Introduction The interest in 3D concrete printing (3DCP) has exponentially increased in the last decade [1, 2]. Thanks to the competitive benefits this technology can offer to the construction industry such reducing construction time, cost and waste. Various mineral and chemical additives can be incorporated in 3DCP [3]. Conventional supplementary cementitious © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 111–116, 2022. https://doi.org/10.1007/978-3-031-06116-5_17

112

A. Abdalqader et al.

materials (SCMs) such as fly ash and metakaolin are usually incorporated in 3DCP mixes and their effects on the properties of 3DCP have been reported [4–6]. However, much less research was conducted on the use of nanoclay (NC) on 3DCP. Previous research showed that NC improved the compressive strength and the interfacial transition zone in cement composites [7, 8]. Therefore, this study investigates the effect of fly ash, metakaolin and nanoclay on the mechanical strengths of hardened 3D printed mortars. In addition, previous research concerned on mechanical strength of 3DCP was based on taking standard cubes/prisms to test them. However, the emphasis in this study is on considering testing cubes/prisms from the printed mortar.

2 Materials and Methods 2.1 Materials and Mix Proportions The cement used was Portland cement type CEM I 52.5 N, conforming to BS EN 1971:2011 [9]. Fly ash (FA), conforming to BS EN 450-1:2012 [10], was obtained from Scot Ash Ltd. This material had a surface gravity of 2.21 and a % passing 45 µm sieve of 85%. Metakaolin (MTK) is used in this experiment to replace some of the cement. A sand of a maximum particle diameter of 1.18 mm was used. The sand/binder and water/binder ratios were kept constant for all mixes at 2 and 0.5 (i.e. 305 kg/m3 ) respectively. The water temperature was maintained at around 16 ± 1 °C for all mixtures. The mortar temperature following the end of mixing was maintained at 20 ± 2 °C. Superplasticizer based on a polycarboxylate polymer solution, having a specific gravity of 1.07 g/ml, was used. The dosage of total SP mass was between 0.2 and 0.4 wt.% of the binder. Purified palygorskite nano-clay (PPNC) is used as a mineral VMA to control stability and flow of the fresh state, trademarked as Acti-Gel 208. The Acti-Gel 208 used is a highly purified magnesium aluminum silicate that is self-dispersing and was added at 5 and 7 kg/m3 in the mix. Natural fibres (NF, Sisal) were also used. These fibres were added to the blend at 3.6 kg/m3 (0.6% of binder and approximately 2.37 l/m3 of the total mix). The mix design of all mixes tested for 3D printing are presented in Table 1. More Table 1. Mixture proportions of mortars, kg/m3 Mix ID

CEMI

FA

REF NF3.6

607

0

FA30-SP0.2-NF3.6

435

172

MTK

Water

NC

SP

0

1230

0

0.136

0

1205

0

1.213

REF MTK15-SP0.44

516

0

91

1210

0

2.671

FA30-MTK10-SP0.44-NF3.6

364

182

61

1180

0

2.671

FA30-MTK15-SP0.44-NF3.6

334

182

91

1180

0

2.671

FA30-SP0.2-NF3.6-VMA5

425

182

0

1190

5.0

1.214

FA30-SP0.2-NF3.6-VMA7

425

182

0

1185

7.0

1.214

Mechanical Performance of 3-D Printed Concrete

113

information on the chemical and particle size distribution of the raw materials as well as mixing and printing procedure is detailed in [4, 11]. 2.2 Testing Procedure For each mixture, the compressive and flexural strength were tested. Compressive strengths were tested on laminated cubes and standard cubes (50 × 50 × 50 mm). The flexural strength was evaluated with 1- and 3-layer prisms and on standard prisms (50 × 50 × 200 mm). Compressive and flexural strength were tested at 7 and 28 days after casting. For each mixture, 6 standard cubes were prepared. The mortar was placed in the moulds and compacted for 5 s on a vibrating table to ensure compaction. The excess mortar is then removed. The cubes are then removed from the mould and the samples are placed in water at a constant temperature until the test days. Compressive strength is also important to calculate from printed cubes. This allows a comparison between uncompacted printed mortar cubes and standard compacted cubes. The moulds used in this experiment were cubic moulds of internal dimensions of 36 × 36 × 40 mm. To create these laminated cubes, 4-layer prisms are printed as shown in Fig. 1. The moulds are then applied vertically to the layers and the excess surrounding mortar is removed. These moulds are demoulded after about one hour. The only problem with this practice was that the upper and lower surface of the cubes were not perfectly smooth. Therefore, all laminated cubes were capped before placed in the compression machine.

Fig. 1. Process for developing layered cubes

The flexural test consists of a three-point bending test: the prism is positioned on two support points, 110 mm apart. The load is applied to the central point of the prism. The bending strength is evaluated at a constant load speed of 40 N/s.

114

A. Abdalqader et al.

Two types of prisms were tested: standard prisms according to the standard and n-layer prisms printed with the compressed air gun. After oiling the mould, the mortar is introduced into the mould and compacted on a vibrating table. The excess mortar on the surface is then removed and the samples are covered for 24 h. The specimens were immersed in water at a constant temperature until the appropriate test date. The same treatment was applied to printed samples that are extruded by the gun. These are composed of 1 and 3 layers.

3 Results and Discussions 3.1 Compressive Strength of Standard and Layered Cubes The compressive strength of all mixes presented in Fig. 2 shows that the compressive strength increased for all mixes with time. The addition of FA reduced the strength at both ages. The pozzolanic reaction of FA is known to start at later ages and contributes to improving the strength at ages beyond 28 days which was not tested in this research. The use of metakaolin slightly reduced the strength at 7 days but had no impact on the strength at 28 days when compared to the reference mix. When the FA mixtures incorporate MTK or NC (VMA) the strength did not change significantly at both ages. It is noted that the strength of the 3D printed concrete mixes remarkedly lower than their counterparts of standard cubes. The reduction in strength ranged between 40–52% at 7days and 12–57% at 28 days. This can be attributed to the preparation method of the 3D printed cubes which were neither vibrated nor compacted. Therefore, large number of voids and entrapped air were observed in these cubes, leading to reduction in strength.

Fig. 2. Compressive strength of standard and layered cubes at 7 and 28 days for all mixes

Mechanical Performance of 3-D Printed Concrete

115

The trend in the strength development and the effect of the FA, MTK and nanoclay were similar to that in standard cubes. However, a distinctive increase in 28-d strength was observed for mix FA30-MTK10-SP0.44-NF3.6. 3.2 Flexural Strength of Standard Prisms The flexural strength of standard prisms did not change significantly with time as shown in Fig. 3 with slight increase in strength. However, unexpectedly, mix FA30-SP0.2NF3.6-VMA7 exhibited lower flexural strength at 28 days than 7 days. This can be due to human error during the test. FA decreased the flexural strength after 7 days but had insignificant impact of flexural strength after 28 days. On the other hand, MTK and NC (VMA) had no noticeable effect on the flexural strengths at both ages. The flexural strength of 1-layer and 3-layer 3D printed prisms is shown in Fig. 3. The mixes gained comparable or better strength than those of standard prisms. However, it was clear that there was inconsistency of the results and a clear conclusion cannot be drawn from these results. The results indicate that the flexural strength of 1 layer is higher than those of 3 layers. This can be explained by the fact that 3-layer 3D printed prism had more entrapped air and cold joints, thus it is more heterogeneous and weaker than 1-layer 3D printed prism.

Fig. 3. Flexural strength of standard and layered cubes at 7 and 28 days for all mixes

4 Conclusion This mechanical strength, in terms of compressive strength and flexural strength, of 3D printed mortars was the main focus of this study. The effect of different additives was

116

A. Abdalqader et al.

explored. Understanding the influence of preparing the testing specimen in standard practice and 3D specimen was established. The main findings indicate that the replacement of cement by 30% FA lowered the strengths, particularly at early ages while the use of MTK and NC did not change the strengths. Standard cubes showed almost double higher compressive strength than 3-D printed cubes while the flexural strength of 3D printed prisms were comparable or higher than standard prism. However, there were discrepancies in the results of 1-layer and 3-layer 3D printed prism. Further investigation therefore is required to elucidate this behavior.

References 1. Mechtcherine, V., et al.: Extrusion-based additive manufacturing with cement-based materials – production steps, processes, and their underlying physics: a review. Cem. Concr. Res. 132, 106037 (2020) 2. Ma, G., Wang, L., Ju, Y.: State-of-the-art of 3D printing technology of cementitious material— an emerging technique for construction. Sci. China Technol. Sci. 61(4), 475–495 (2017). https://doi.org/10.1007/s11431-016-9077-7 3. Schmidt, W., Sonebi, M., Brouwers, H.J.H.J., Kühne, H.-C., Meng, B.: Rheology modifying admixtures: the key to innovation in concrete technology – a general overview and implications for Africa. Chem. Mater. Res. 5, 115–120 (2013). http://iiste.org/Journals/index.php/CMR/art icle/view/11354%5Cnpapers3://publication/uuid/3BCE5CBB-648B-46B1-8C67-21ABC2 25F6E9 4. Sonebi, M., Dedenis, M., Amziane, S., Abdalqader, A., Perrot, A.: Effect of red mud, nanoclay, and natural fiber on fresh and rheological properties of three-dimensional concrete printing. ACI Mater. J. 118, 97–110 (2021) 5. Panda, B., Ruan, S., Unluer, C., Tan, M.J.: Improving the 3D printability of high volume fly ash mixtures via the use of nano attapulgite clay. Compos. Part B Eng. 165, 75–83 (2019). https://doi.org/10.1016/j.compositesb.2018.11.109 6. Bohuchval, M., Sonebi, M., Amziane, S., Perrot, A.: Rheological properties of 3D printing concrete containing sisal fibres. In: 3rd International Conference on Bio-Based Building Materials, pp. 249–255, June 2019 7. Heikal, M., Ibrahim, N.S.: Hydration, microstructure and phase composition of composite cements containing nano-clay. Constr. Build. Mater. 112, 19–27 (2016). https://doi.org/10. 1016/j.conbuildmat.2016.02.177 8. Shebl, S.S., Seddeq, H.S., Aglan, H.: Effect of micro-silica loading on the mechanical and acoustic properties of cement pastes. Constr. Build. Mater. 25, 3903–3908 (2011) 9. British Standards Institution: BS EN 197-1 Cement Part 1: Composition, Specifications and Conformity Criteria for Common Cements, 50 (2011) 10. BSI: BS En 450-1 Fly ash for concrete. Definition, specifications and conformity criteria, British Standard Institution, UK, 34 (2012). http://shop.bsigroup.com/en/ProductDetail/?pid= 000000000030216589 11. Bohuchval, M., Sonebi, M., Amziane, S., Perrot, A.: Effect of metakaolin and natural fibres on three-dimensional printing mortar. Proc. Inst. Civ. Eng. Constr. Mater. 174, 115–128 (2021). https://doi.org/10.1680/jcoma.20.00009

Binders and Aggregates 3: Strain Hardening Materials

Incorporation and Characterization of Multi-walled Carbon Nanotube Concrete Composites for 3D Printing Applications Albanela Dulaj(B) , Monica P. M. Suijs, Theo A. M. Salet, and Sandra S. Lucas Eindhoven University of Technology, Eindhoven, The Netherlands [email protected]

Abstract. Being 3D concrete printing (3DCP) a relatively new technology, there is a constant need for development, not only in hardware but also in the materials we currently use. The use of nanomaterials, notably carbon nanotubes, in concrete and cement mixes is not new, but it is a novelty in 3DCP. This paper investigates the effect on the fresh and hardened state properties of adding multi-walled carbon nanotubes (MWCNTs) to a printable concrete mix. Cement composites reinforced with MWCNTs may exhibit an irreversible change in resistivity when subject to damage or microstructural changes caused by strain or stress. Research also points to an improvement in flexural and compressive strength and lower shrinkage for CNT-cement mortars. The combined electrical and mechanical properties of these mixtures are of interest in crack self-detection for structural health monitoring. In this paper, two commercial solutions of MWCNTs dispersed in water were used to determine their effect on a cement mortar suitable for 3D printing. Three different mixes were prepared: a reference mix, a mix with 0.05% of MWCNTs per binder content and a mix with 0.1% of MWCNTs per binder content. Higher percentages of carbon nanotubes resulted in a decrease in flowability and mechanical properties due to the difficulty in dispersing the nanotubes. The results show that one of the two batches of MWCNTs used performed better overall than the other and the mix with 0.05% of MWCNTs per binder content revealed a better performance both in the fresh and hardened state than the mix with 0.1% of MWCNTs. Keywords: 3D concrete printing · Multi-walled carbon nanotubes · Fresh concrete study · Hardened concrete

1 Introduction 3D Concrete Printing is a rapidly growing sector of the construction industry. There is a great deal of research underway into designing new concrete-based mixes suitable for 3D printing. The use of nanomaterials such as multi-walled carbon nanotubes (MWCNTs) to enhance certain properties has been studied deeply for conventional concrete but not yet for 3D printed concrete. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 119–125, 2022. https://doi.org/10.1007/978-3-031-06116-5_18

120

A. Dulaj et al.

Multi-walled carbon nanotubes have exceptional mechanical properties with an elastic modulus between 20 and 60 GPa and excellent thermal and electrical properties [1, 2]. Due to these properties, adding them to cementitious composites also improves the performance of the mortar by increasing its compressive and flexural strength, improving the bonding behavior of pulled-out fibers, and altering its conductivity [2]. Carbon nanotubes are indeed used in the concrete industry to create a smart material that changes its conductivity when subjected to stress. This effect can be used to monitor the health of a printed structure and increase its durability [3]. In this paper, multi-walled carbon nanotubes (MWCNTs) are added and mixed at different concentrations (0%, 0.05%, 0.1% by weight of MWCNTs per binder content) with a printable cement mortar and their effect on the fresh state and hardened state properties of the material are evaluated. An uniaxial unconfined compression test was performed at different ages of the mortar, up to 90 min after preparation, to evaluate how the strength and stiffness development was affected by the MWCNTs and how the built-up rate changed [4]. The material was then printed to evaluate its printability and aged 28 days to measure the compressive and flexural strength of the compositions. The strength of the printed samples was compared to the strength of cast samples of the same size and composition [5].

2 Experimental Procedure 2.1 Materials and Specimens The printable mix is a mortar developed to have a good thixotropic and mechanical behavior. Its composition is detailed in Table 1. Table 1. Composition of the printable mortar. Composition

Mix ratio

Type

Sand

Sand–binder ratio = 1.50

CEN-NORMSAND DIN EN 196–1

Binder

70% cement

CEM I 52.5R Heidelberg cement

20% fly ashes

Vliegasunie Class F

10% silica fumes

Elkem Microsilica 920D

Water

Water-binder ratio = 0.25

Additives (per binder content)

0.75% superplasticizer

Sika® ViscoCrete®-2640 con. 35% SPL

0.25% retarder

CUGLA Cretolent F con. 25% BT

1.0% accelerator

MasterSet AC 555 con 55%

In the mix, both a retarder and an accelerator are used because they react at different times: the retarder is used to maintain the mix flowable while it’s printed, while the accelerator starts to act when the mortar is deposited and helps the built-up of the material. The MWCNTs were supplied by NANOCYL in the form of a solution in water

Incorporation and Characterization of MWCNTs Concrete Composites

121

with a concentration of 3%. Two batches were used in this work: AQUACYL AQ0303 (batch 1) and AQUACYL AQ030X (batch 2). The two batches differ for the type of surfactants and additives used to spread the MWCNTs. To prepare the samples, the solution was mixed with part of the mortar’s water for 5 min using a mixer, and then the superplasticizer, accelerator and retarder were added once at a time mixing for 5 min after each addition for a total of 20 min of liquid mixing. The solution was then mixed with the solid components of the mortar for another 5 min. AQUACYL and water were weighed and mixed keeping into account that the final product should have a water binder ratio of 0.25 and the wanted MWCNTs concentration. 2.2 Uniaxial Unconfined Compression Test Uniaxial unconfined compression tests (UUCT) were performed on all the compositions to assess the strength σ and rigidity E (apparent Young’s modulus at 5% strain) development of the material over time up until 90 min from preparation. 5 cylindrical steel molds of 140 mm in height and 70 mm in diameter were prepared. The molds were covered with Teflon sheets inside to prevent fresh concrete from sticking to them. The fresh concrete was poured into the molds and compacted with a metallic rod. The tests were carried out after 0, 15, 30, 60 and 90 min from preparation. An Instron instrument with a 5 kN load cell was used to perform the tests (Fig. 1). A computer was connected to the instrument and recorded the vertical displacement and the applied force. The head speed was 30 mm/s. A camera Canon, type EOS 700D, with a resolution of 18 megapixels and a 2-s interval self-timer photographed the samples while they deformed (Fig. 1). The images were processed with the Vision Builder from National institute software to calculate the lateral deformation via optical analyses.

Fig. 1. UUCT test setup: 5 kN loadcell placed in the Instron to measure the deformation of fresh concrete cylindrical samples while a camera records the lateral deformation.

2.3 Hardened State Mechanical Tests The mechanical tests were performed following BS EN 1015–11:1999: a 3-point bending test and a compression test at the hardened state were performed on cast and printed samples aged 28 days. The cast samples were prepared by putting the fresh concrete on

122

A. Dulaj et al.

Styrofoam molds of 40 × 40 × 160 mm covered in plastic sheets for 24 h and then cured in water for 27 days. The printed samples were prepared with the same dimensions using an ABB IRB 1200 robot combined with a Makita DCG 140 Caulking gun, covered in plastic sheets for 24 h and then cured in water for 27 days. The tests were load controlled: the bending test had a load speed of 50 N/s, the compression test had a speed of 2400 N/s. The bending strength was determined by Eq. (1). fb = 1.5 · F · l / (b · d 2 )

(1)

The compression strength was determined by Eq. (2). fc = F · / (b · l)

(2)

where l = length, b = width and d = height of sample. F is the force that leads to failure (Fig. 2).

Fig. 2. The instrument is a simple bending and compression machine in which samples with standard dimension (following BS EN 1015-11:1999) are tested.

3 Experimental Results 3.1 Uniaxial Unconfined Compression Test Figure 3 shows the evolution of the compressive strength with time for the mixes with AQ0303 (Fig. 3.a) and the mixes with AQ030X (Fig. 3.b), and the results show a bi-linear trend that can be described by Eq. (3). σ =a·t+b

(3)

With σ = compressive strength in kPa, t = time in minutes, a and b the coefficients.

Incorporation and Characterization of MWCNTs Concrete Composites

(a)

123

(b)

Fig. 3. Development of compressive strength with time for a) compositions mixed with AQ0303 (batch 1) and b) compositions mixed with AQ030X (batch 2).

(a)

(b)

Fig. 4. Development of rigidity measured at 5%strain with time for a) compositions mixed with AQ0303 (lot 1) and b) compositions mixed with AQ030X (lot 2).

As shown in Fig. 3, the mixes with AQ030X have a better performance, with higher strength built up rate and, for the composition with 0.05% of MWCNTs from AQ030X, the strength built-up rate exceeds that of reference material, improving the fresh state mechanical properties. The rigidity is measured along with the strength, and its values are plotted in Fig. 4 for all the mixes. The results show again a bi-linear tendency described by Eq. (4). E =c·t+d

(4)

With E = rigidity measured at 5% strain in kPa, t = time in minutes, c and d the coefficients. Again, the tendency is the same as previously seen for the strength development, with a good improvement of the properties when AQ030X is used in the composition. The results indicate an improvement of the built-up rate and the strength and stiffness development when lot AQ030X is used. On the other hand, lot AQ0303 has detrimental effects on the built-up rate. This test confirms that the use of lot AQ030X is preferable for 3D printable concrete.

124

A. Dulaj et al.

3.2 Hardened State Mechanical Test on Printed Samples The mixes using AQ030X are printed with a rectangular nozzle of 1 × 4 cm and the layer height is about 9 mm (Fig. 5).

(a)

(b)

Fig. 5. Printing of the material: a) ABB-Makita system, b) printing of samples.

Cast samples are prepared alongside. The flexural strength (Fig. 6.a) and mechanical strength (Fig. 6.b) for cast and printed samples are measured at 28 days.

(a)

(b)

Fig. 6. Hardened state mechanical results of the cast and printed samples: a) flexural strength (MPa), b) compression strength (MPa).

The results show a slight decrease in flexural strength between cast and printed samples (between 1% and 5%) and a greater decrease in compressive strength (between 24% and 27%). This is in line with previous research [5]. The mechanical properties of the material decrease with the increase of the concentration of MWCNTs, especially in compression, for both printed and cast samples, probably due to the formation of MWCNT clusters within the material. The best mix appears to be the one with 0.05% MWCNTs.

Incorporation and Characterization of MWCNTs Concrete Composites

125

4 Conclusions Two different batches of MWCNTs suspended in a solution, AQ0303 and AQ030X, were mixed with a printable mortar at different concentrations (0%, 0.05% and 0.1% per binder content) to evaluate their effect on the fresh state properties such as the built-up rate and early age strength and stiffness development that assess the printability of the mortar. The compositions were then printed and cast samples with the same dimensions and composition were prepared alongside and their mechanical and flexural strength at 28 days of aging were measured and compared. The UUCT results show that mixes with AQ0303 have worse built-up rates and strength and stiffness development rates than the reference mortar, while mixes with AQ030X show an improvement of the properties at the fresh state, with higher strength and stiffness development rates than the reference. For this reason, AQ030X is more suitable for 3D printing and the printing trials were done using this lot. The compositions with different concentrations of AQ030X were all printable, and the samples printed show a slight decrease of the mechanical properties compared to cast samples, in line with previous research [5]. The increasing concentration of MWCNTs in the mix leads to a decrease in compressive strength due to the presence of clusters in the material. The mix with better mechanical performance appears to be the one with 0.05% MWCNTs. Further research is underway to improve MWCNTs dispersion (to improve mechanical properties) and to study the electrical properties, shrinkage and rheology of the compositions.

References 1. Suchorzewski, J., Prieto, M., Mueller, U.: An experimental study of self-sensing concrete enhanced with multi wall carbon nanotubes in wedge splitting test and DIC. Constr. Build. Mater. 262, 120871 (2020) 2. Ahmed, B.R., Hussein, A.J., Saleh, D., Rashid, R.S.M.: Influence of carbon nanotubes (CNTs) in the cement composites. In: IOP Conference Series: Earth and Environmental Science. Sustainable Civil and Construction Engineering Conference 2019, vol. 357. IOP Publishing (2019) 3. Coppola, L., Buoso, A., Corazza, F.: Electrical properties of carbon nanotubes cement composites for monitoring stress conditions in concrete structures. Appl. Mech. Mater. 82, 118–123 (2011) 4. Suiker, A.S.J., Wolfs, R.J.M., Lucas, S.M., Salet, T.A.M.: Elastic buckling and plastic collapse during 3D concrete printing. Cem. Concr. Res. 135, 106016 (2020) 5. Wolfs, R.J.M., Bos, F.P., Salet, T.A.M.: Hardened properties of 3D printed concrete: the influence of process parameters on interlayer adhesion. Cem. Concr. Res. 119, 132–140 (2019)

Properties of 3D-Printable Ductile Fiber-Reinforced Geopolymer Composite Shin Hau Bong1 , Behzad Nematollahi1,2(B) , Venkatesh Naidu Nerella3 , and Viktor Mechtcherine3 1 Centre for Smart Infrastructure and Digital Construction, Swinburne University of

Technology, Melbourne, Australia 2 Department of Civil and Structural Engineering, The University of Sheffield, Sheffield, UK

[email protected] 3 Institute of Construction Materials, TU Dresden, Dresden, Germany

Abstract. This paper presents the performances of a 3D printable ductile fiberreinforced geopolymer composite (3DP-DFRGC). An ambient temperature cured one-part geopolymer was utilized as the binder for manufacture of the developed 3DP-DFRGC, which eliminates the necessity for curing at elevated temperature and handling of user-hostile alkaline solutions. Herewith, it considerably enhances the possibility of in-situ applications and commercial viability of the 3DP-DFRGC. The rheological behavior and mechanical properties of the 3DPDFRGC were experimentally characterized. The mold-cast DFRGC was also prepared and tested for comparison. The 3DP-DFRGC exhibited pronounced deflection-hardening behavior under bending. The modulus of rupture and the corresponding deflection of the 3DP-DFRGC were 18% and 28% higher, respectively than those of the mold-cast DFRGC. This can be due to the preferential orientation of fibers in the 3D-printed specimens. Keywords: Ductile fiber-reinforced geopolymer composite · Geopolymer · Deflection-hardening · 3D concrete printing · Extrusion

1 Introduction Incorporating conventional reinforcement in the extrusion-based 3D concrete printing (3DCP) process is a challenge that still needs to be addressed [1]. A 3D-printable ductile fiber-reinforced cementitious composite (3DP-DFRCC) showing deflection-hardening behavior under bending could be a possible solution to lessen or remove the need for conventional steel reinforcement in 3DCP. However, high amount of ordinary Portland cement (OPC) typically used in DFRCC causes high heat release due to the OPC hydration, autogenous shrinkage, and cost. Moreover, the embodied energy and carbon emissions due to the manufacture of OPC can significantly impact the sustainability credentials of 3DCP with DFRCC. Developing a 3D-printable ductile fiber-reinforced geopolymer composite (3DP-DFRGC), in which the OPC is completely substituted by a geopolymer binder, can tackle the aforementioned problems. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 126–132, 2022. https://doi.org/10.1007/978-3-031-06116-5_19

Properties of 3DP-DFRGC

127

The authors have recently established the feasibility of developing a 3DP-DFRGC [2]. The printed specimens reinforced with 0.75 vol.% and 1.0 vol.% of polypropylene (PP) fibers showed deflection-hardening behavior with modulus of rupture of 6.1 to 7.5 MPa and mid-span deflection of 0.65 to 0.75 mm [2]. The previously developed 3DP-DFRGC has two main limitations. First, the deflection capacity of the printed specimens is low. Second, a conventional ‘two-part’ geopolymer binder, consisted of alkaline solutions and fly ash, was used to produce the 3DP-DFRGC, which required curing at elevated temperature and handling of corrosive alkaline liquids. To tackle the above limitations, the authors recently developed a new 3DP-DFRGC, the rheological behavior and mechanical properties of which are reported in this paper. A ‘one-part’ geopolymer binder, consisting of slag, fly ash and solid activator, was used for the manufacture of the new 3DP-DFRGC. This removes the need for handling corrosive alkaline liquids and elevated temperature curing, and thereby enhances its commercial viability. Instead of PP fibers, polyvinyl alcohol (PVA) fibers were used to improve the flexural performance of the composite. The rheological behavior and mechanical properties of the new 3DP-DFRGC were investigated. In addition, the conventionally mold-cast DFRGC was also tested for comparison.

2 Materials and Experimental Program 2.1 Materials and Mixture Design A combination of slag and fly ash (Class F) were used as the geopolymer precursor. The fine silica sand used had a D50 of 176 μm. GD Grade sodium silicate (SiO2 /Na2 O = 2.0) and anhydrous sodium metasilicate powders (SiO2 /Na2 O = 0.9) were used as the solid activators. Sucrose powder was used as a retarder. The density, elongation, length, diameter, elastic modulus and nominal strength of the PVA fiber were 1.3 g/cm3 , 6%, 8 mm, 39 μm, 41 GPa, and 1600 MPa, respectively. Table 1 presents the mixture proportions of the 3DP-DFRGC. The mass ratio of slag to fly ash ratio was 1.0. The mass ratio of GD Grade sodium silicate to anhydrous sodium metasilicate was also 1.0. Table 1. Mixture proportion of 3DP-DFRGC. Precursor

Activator

Sand

Retarder

Water

Fiber

1.00

0.10

0.10

0.005

0.367

0.014

Note: All numbers are mass ratios of the precursor weight (slag + fly ash) except fiber content (volume fraction). Note: The mixture proportions of the geopolymer matrix are adapted from [9].

2.2 Mixing and 3D Printing To make the fresh material, all dry ingredients were firstly blended in a Hobart mixer for 3 min. After that, water was slowly poured into the dry mix and the blending was continued at low speed for another 4 min. Subsequently, the fibers were gradually blended

128

S. H. Bong et al.

to the mixture and at low speed for another 3 min. Lastly, the blending was continued at high speed for another 3 min to achieve a homogeneous mixture. Using a gantry-type 3DCP machine with a 10 mm round printhead, a 10-layer slab (see Fig. 1a) and a two-layer slab (see Fig. 1b) were 3D-printed for the compression and flexural tests, respectively. The thickness of the layer was 5 mm. The print speed and extrusion speed were 2520 mm/min and 200 ml/min, respectively. A plastic sheet was used to cover the freshly printed slabs and left in the laboratory temperature (23 ± 3 °C) until the testing day, which was 28 days after the printing.

Fig. 1. (a) Top view of the 3DP-DFRGC slab (50 mm thick) for compression test and (b) top view of the 3DP-DFRGC slab (10 mm thick) for flexural test.

2.3 Characterization Methods The flowability of the fresh mix was measured by performing a flow table test in accordance with ASTM C1437-15. The rheological properties of the fresh DFRGC including static yield stress and structural recovery were determined via a Viskomat XL rheometer containing a six-blade vane probe (the diameter and height of the blades = 69 mm). The static yield stress evolution was measured by conducting constant shear rate test at the resting times of 1 min, 5 min, 10 min and 15 min. A constant rotational velocity of 0.6 rpm was applied to the fresh mixture for 4 s at each resting time. The highest torque value recorded in each measurement was then converted into static yield stress by using the Cauchy stress principle. The structural recovery behavior was evaluated by using a three-phase protocol to mimic the printing process. These phases included constant rotational velocities of (i) 0.1 rpm for 60 s, (ii) 30 rpm for 30 s and (iii) 0.1 rpm for 60 s. To determine the compressive strength of the 3DP-DFRGC, 50 mm cubes were sawn from the 10-layer 3D-printed slab and tested in three loading directions, namely X (longitudinally parallel to the layer interface), Y (laterally parallel to the layer interface) and Z (perpendicular to the layer interface) at the loading rate of 0.33 MPa/s. The flexural performance of the 3DP-DFRGC was evaluated by performing four-point bending tests (top span = 100 mm and bottom span = 300 mm). The 350 mm (length) × 60 mm (width) × 10 mm (height) panels were sawn from the two-layer 3D-printed slab and tested in Z-direction (perpendicular to the layer interface) at the rate of 0.5 mm/min.

Properties of 3DP-DFRGC

129

3 Results and Discussions 3.1 Rheological Proportions The mean spread diameter of the fresh mix was measured to be 170 mm after dropping of the flow table. Table 2 shows the static yield stress of the fresh material measured at four resting times. The static yield stress of the fresh material increased significantly with the resting time, which is desirable to ensure high shape retention and buildability. The continuous geopolymerization reactions in the material significantly increase its static yield stress over time. Moreover, physical interlocking and overlapping among fibers and their interactions with the solid particles in the matrix also increase the stress required to initiate the flow [3]. Table 2. Static yield stress of 3DP-DFRGC at different resting times. Resting time

1 min

5 min

10 min

15 min

Static yield stress

1472 Pa

3022 Pa

4227 Pa

5237 Pa

Figure 2 presents the structural recovery result of the fresh material. The developed fresh material exhibited recovery rate of 71.4% at the end of Phase III. This indicates that the developed material has good thixotropy behavior, which enables the material to retain its shape after being extruded and rapidly gain strength to sustain upper layers without experiencing considerable deformation at bottom layers.

Fig. 2. Structural recovery behavior of 3DP-DFRGC mixture.

3.2 Mechanical Properties The compressive strengths of 3DP-DFRGC in X-, Y-, and Z-directions were 50.3 ± 2.5 MPa, 44.9 ± 2.3 MPa, and 53.9 ± 0.9 MPa, respectively. The compressive strength

130

S. H. Bong et al.

of the mold-cast DFRGC was 64.7 ± 3.2 MPa, being 20% to 44% greater than the 3D-printed specimens, based on the loading orientation. The lower strength of the 3Dprinted fiber-reinforced cementitious composites was also reported by Zhu et al. [4]. The compressive strength of the 3DP-DFRGC showed anisotropic behavior depending on the testing direction. The compressive strength measured in Z-direction was the highest one, followed by that measured in X-direction and Y-direction. The strength measured in Z-direction was 7% and 20% higher as compared to that measured in X-direction and Y-direction, respectively. The maximum strength measured in Z-direction is most likely due to compaction of the fresh mixture in this direction by the self-weight of the layers [5]. The microstructure (void) morphology of 3D-printed concrete influences different orthogonal strengths [6]. Figure 3 presents the four-point bending test result of the mold-cast and 3Dprinted DFRGCs. Both types of DFRGCs showed deflection-hardening behavior. Table 3 presents the average first-crack strength, modulus of rupture (MOR) and their corresponding deflection values. The average MOR and deflection at peak load of the 3D-printed specimens were 18% and 28% higher as compared to those of the mold-cast specimens, respectively. In contrast to the random distribution of fibers in the mold-cast specimen, visual observation showed that in the 3DP-DFRGC most fibers were oriented parallel to the printing direction, leading to its superior deflection-hardening behavior [7]. The ductility of the specimens can be evaluated by the ductility index, which is the ratio of the deflection at peak load to the first-crack deflection (δMOR /δLOP ) [8]. Based on the deflection values shown in Table 3, the ductility index of the 3DP-DFRGC was calculated to be 13.1, which was 52% higher than the mold-cast DFRGC (8.6). The work-to-fracture values of both types of DFRGCs were determined from the area under the curves (see Fig. 3) up to the deflection corresponding to MOR; see Table 3. The work-to-fracture of 3DP-DFRGC was 47% higher than that of the mold-cast DFRGC, which is due to the superior flexural performance of the 3D-printed specimens.

Fig. 3. Flexural stress vs. mid-span deflection curves of mold-cast and 3D-printed DFRGCs.

Properties of 3DP-DFRGC

131

Table 3. Flexural test results of DFRGCs with different production processes. Production process

First-crack strength fLOP (MPa)

First-crack deflection δLOP (mm)

Modulus of Deflection at rupture peak load MOR (MPa) δMOR (mm)

Work-to-fracture (kJ/m2 )

3D-printing

3.0 ± 0.3

1.0 ± 0.2

5.2 ± 0.1

13.1 ± 0.3

57.1 ± 0.7

Mold-casting

2.9 ± 0.3

1.2 ± 0.4

4.4 ± 0.3

10.3 ± 2.3

38.9 ± 10.5

Note: The numbers indicate Average ± Standard Deviation.

4 Conclusion A 3DP-DFRGC was developed in this study showing deflection-hardening behavior and multiple fine cracks under flexure. Fresh properties and mechanical performances of the 3DP-DFRGC were measured and compared with the conventionally mold-cast DFRGC. The conclusions that can be made are as follows: • The developed 3DP-DFRGC showed superior flexural behavior to the mold-cast DFRGC. The modulus of rupture and deflection capacity of the 3D-printed specimens were 5.2 MPa and 13.1 mm, respectively. The corresponding values for the counterpart mold-cast specimens were 4.4 MPa and 10.3 mm. • The ductility index (δMOR /δLOP ) of the 3DP-DFRGC was 13.1, being 52% higher than the mold-cast DFRGC (8.6), confirming the superior ductility of the 3DP-DFRGC. • The work-to-fracture of the 3DP-DFRGC was 47% higher than the mold-cast DFRGC, owing to the superior flexural behavior of the 3D-printed specimens. • The compressive strength of the 3DP-DFRGC was in the range of 44.9–53.9 MPa, depending on the testing direction, which was 17–31% lower than that of the mold-cast DFRGC. • The static yield stress of the fresh 3DP-DFRGC increased exponentially as the resting time increased, which is beneficial to ensure high shape retention and buildability. The structural recovery of the fresh 3DP-DFRGC was determined to be 71.4%.

References 1. Mechtcherine, V., et al.: Integrating reinforcement in digital fabrication with concrete: a review and classification framework. Cem. Concr. Compos. 119, 103964 (2021) 2. Nematollahi, B., et al.: Effect of polypropylene fibre addition on properties of geopolymers made by 3D printing for digital construction. Materials 11(12), 2352 (2018) 3. Khayat, K.H., Meng, W., Vallurupalli, K., Teng, L.: Rheological properties of ultra-highperformance concrete—an overview. Cem. Concr. Res. 124, 105828 (2019) 4. Zhu, B., Pan, J., Nematollahi, B., Zhou, Z., Zhang, Y., Sanjayan, J.: Development of 3D printable engineered cementitious composites with ultra-high tensile ductility for digital construction. Mater. Des. 181, 108088 (2019) 5. Sanjayan, J.G., Nematollahi, B., Xia, M., Marchment, T.: Effect of surface moisture on interlayer strength of 3D printed concrete. Construct. Build. Mater. 172, 468–475 (2018)

132

S. H. Bong et al.

6. van den Heever, M., du Plessis, A., Kruger, J., van Zijl, G.: Evaluating the effects of porosity on the mechanical properties of extrusion-based 3D printed concrete. Cem. Concr. Res. 153, 106695 (2022) 7. Arunothayan, A.R., Nematollahi, B., Ranade, R., Bong, S.H., Sanjayan, J.: Development of 3Dprintable ultra-high performance fiber-reinforced concrete for digital construction. Construct. Build. Mater. 257, 119546 (2020) 8. Naaman, A., Reinhardt, H.: Characterization of high performance fiber reinforced cement composites’. In: High Performance Fiber Reinforced Cement Composites 2—HPFRCC 2, vol. 2, pp. 1–24. E & FN Spon (1996) 9. Bong, S.H., Nematollahi, B., Naidu Nerella, V., Mechtcherine, V.: Method of formulating 3Dprintable strain-hardening alkali-activated composites for additive construction. Cem. Concr. Compos. (2022, under review)

Feasibility of Using Ultra-High Ductile Concrete to Print Self-reinforced Hollow Structures Junhong Ye1 , Yiwei Weng1(B) , Hongjian Du2 , Mingyang Li3 , Jiangtao Yu4 , and Md Nasir Uddin4 1 Department of Building and Real Estate,

The Hong Kong Polytechnic University, Hong Kong, China [email protected] 2 Department of Civil and Environmental Engineering, National University of Singapore, Singapore, Singapore 3 Singapore Centre for 3D Printing, Nanyang Technological University, Singapore, Singapore 4 Department of Disaster Mitigation for Structures, Tongji University, Shanghai, China

Abstract. 3D concrete printing (3DCP) is facing the challenge of introducing steel reinforcement. Using ultra-high ductile concrete (UHDC) to print selfreinforced structures is a potential method to address the challenge. However, few researches have been conducted in this area. This work aims to demonstrate the feasibility of using UHDC to print self-reinforced structures with hollow sections. 6 different hollow beams without steel reinforcement were printed and tested under four-point loading. The results indicate that the printed beams with hollow sections have multiple cracking, flexural hardening, and ductile failure mode despite the absence of steel reinforcement. The flexural strength of hollow beams ranged from 6.06 MPa to 7.86 MPa. The deflection to span ratio of all printed beams was higher than 1/50 at the ultimate state. The combination of UHDC and hollow structures provides a new concept for 3DCP without steel reinforcement. Keywords: 3D concrete printing · Ultra-high ductile concrete · Hollow beam · Mechanical performance

1 Introduction 3D concrete printing (3DCP) has attracted significant attention due to its advantages, such as formwork-free construction, less wastage generation, enhanced productivity, etc. [1]. In 3DCP, the material is precisely deposited in a layer-atop-layer manner to build structures without formwork based on a predesigned 3D model [2]. This formwork-free construction approach allows engineers and architects to design unique buildings with functional performance [3, 4]. Nevertheless, 3DCP is still in the development stage, and many challenges exist in this area. One of the critical challenges is how to enhance the ductility of 3D printed structures. Various approaches were proposed to introduce reinforcement in 3DCP to improve its ductility [5]. However, these existing methods have certain limitations, such as increasing © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 133–138, 2022. https://doi.org/10.1007/978-3-031-06116-5_20

134

J. Ye et al.

the difficulties of system integrity due to adopting multi-printing systems to print concrete material and introduce the reinforcement concurrently. To avoid the abovementioned limitation, developing a self-reinforced material with favorable printability and superior mechanical properties, especially tensile strain capacity, is a potential method to tackle the challenges. Ultra-high ductile concrete (UHDC) is a promising material for constructing infrastructures without steel reinforcement due to its predominant tensile strain capacity of over 8.0% [6]. Yu et al. [7] developed a printable UHDC with a tensile strength of over 4.5 MPa and a strain capacity of 7.0% using short-cut fibers. This study aims to enhance the ductility of printed self-reinforced hollow concrete structures using UHDC. Four-point bending tests were conducted on UHDC solid beams and hollow beams. The experimental results of this study are expected to provide an innovative solution in 3DCP to construct the printed structure by using construction materials with high ductility.

2 Experimental Program Hollow beams with truss-like infill structures are adopted in the experimental part. Wang et al. [8] printed four types of hollow beams for bending tests, among which the hollow beams with truss-like sections outperformed others in flexural strain capacity and deformability. However, the truss-like beams in [8] still suffered from brittle failure due to the insufficient tensile strain capacity of printed materials. To further enhance the mechanical performance of printed beams from materials level and structural level, 3D printable UHDC with high superior tensile strain capacity was used to pinted hollow beams with truss-like sections. A total of 6 types of truss-like UHDC beams with hollow sections were designed and printed for four-point bending tests, considering the quantity of “trusses” (i.e., UHDC filaments) in pure-bending region and flexural shearing region. Two replicated specimens were printed for all types of beams. The 3D printable UHDC was developed in a previous study [7]. The compressive strength of printed UHDC ranged from 36.7 MPa to 44.6 MPa. The average tensile strength and strain capacity of printed UHDC were 5.75 MPa and 7.54%, respectively. The hollow beams were labelled by the quantity of UHDC filaments, as plotted in Fig. 1. For example, B-1-2-1 indicates that the hollow beam has two UHDC trusses in mid-span and one UHDC truss in each side-span. Two solid UHDC beams were printed as the comparison group, i.e., B-solid. The geometrical dimension of all beams is 150 × 150 × 825 mm. The span and loading spacing are 675 mm. A three-axis gantry-type concrete printer with a circle nozzle (30 mm in diameter) was adopted in this study to print UHDC components in the printing process. As shown in Fig. 2, the printed specimens were then placed in the curing room and naturally cured in the air for 28 days before testing. The cross-section of a single UHDC layer is 30 mm in width and 15 mm in height. Four-point bending tests were conducted on an MTS electro-servo machine with 500 kN capacity, as shown in Fig. 2. Three linear variable displacement transducers

Feasibility of Using Ultra-High Ductile Concrete

135

(LVDTs) were installed on the bottom of a specimen to measure the vertical displacement. The surface of the specimen was sprayed with white undercoat and black speckles for digital image correlation (DIC) measurement.

(a) B-1-2-1

(b) B-1-4-1

(c) B-2-2-2

(d) B-2-4-2

(d) B-3-2-3

(e) B-3-4-3

Fig. 1. Planar layout of truss-like UHDC beams with hollow sections.

Fig. 2. Printed UHDC beams and test setup.

3 Result and Discussion Figure 3 presents the crack patterns of different UHDC beams, using the horizontal strain contours obtained by DIC measurement. It is clear from Fig. 3a that B-solid exhibited ductile failure mode by triggering multiple fine cracks, despite the absence of steel reinforcement. Figures 3b–g show crack patterns of printed hollow columns. As can be seen from these figures, the fine crack of some hollow beams was extended over the whole span, such as the specimens of B-1-2-1 and B-1-4-1. The results indicate that less truss number in the flexural-shearing region has more benefits to triggering multiple fine cracks. The width of fine cracks in all beams ranged 0.1–0.2 mm, demonstrating the superior crack controllability of UHDC.

136

J. Ye et al.

The crack number of B-solid in the pure-bending and flexural-shearing regions was 92 and 72, respectively. The relations between crack number and truss number of hollow beams are plotted in Fig. 4. Regardless of the truss number in the pure-bending region, the crack number in the pure-bending region increases with the truss number. In contrast, the crack number in the flexural-shearing region decreases significantly with increasing the truss number. In addition, it is noted that the less truss number in the pure-bending region leads to more fine cracks regardless of the truss number in the flexural-shearing region. These results demonstrate that the appropriate design of the hollow section effectively triggers more fine cracks. The crack number of B-1-2-1 over the whole span is 262, which is nearly 1.6 times compared to that of B-solid.

(a) B-solid



(b) B-1-2-1

(c) B-2-2-2

(d) B-3-2-3

(e) B-1-4-1

(f) B-2-4-2

(g) B-3-4-3





Fig. 3. Crack patterns of all UHDC beams.

The load-deflection curves of tested beams are plotted in Fig. 5. Different levels of flexural-hardening behaviors were observed on the B-solid and different hollow beams. Due to the design of hollow sections, the hollow beams exhibited reduced load-bearing capacity and deformability compared to that of the B-solid. It is noted that B-1-2-1 with the fewest trusses outperformed other hollow beams in deformability, while B-3-4-3 with most trusses showed the highest load-bearing capacity. Table 1 summarizes the flexural properties of all tested beams, including flexural strength σ u , ultimate deflection δ u , deflection to span ratio δ u /L, strength to mass ratio σ u /m, and energy dissipation capacity T u . The highest flexural strength among the results of hollow beams is 7.86 MPa (B-3-4-3), which is 60% lower than that of B-solid. The strength to mass ratio of hollow beams ranges from 0.215 to 0.269, significantly smaller than that of B-solid. The ultimate deflection of B-1-2-1 and B-1-4-1 is 25.92 mm and 25.79 mm, respectively, slightly smaller than that of B-solid. The significant decrease in the load-bearing capacity and deformability of hollow beams may be due to the insufficient depth to width ratio. As a result, the contribution of trusses to the flexural properties is limited.

Feasibility of Using Ultra-High Ductile Concrete

(a) Crack number in the pure-bending region

137

 (b) Crack number in the flexural-shearing region

Fig. 4. The relations between crack number and truss number.

(a) All UHDC beams, including B-solid





(b) All UHDC beams with hollow sections

Fig. 5. Load-deflection curves of all UHDC beams.

Table 1. Flexural properties of all UHDC beams. Specimen ID

σu (MPa)

δu (mm)

δ u /L

σ u /m (MPa/kg)

Tu (kN·mm)

B-solid

13.12

29.04

1/23

0.383

1702.6

B-1-2-1

6.92

25.92

1/26

0.243

855.3

B-2-2-2

6.98

21.27

1/32

0.248

689.7

B-3-2-3

7.06

15.15

1/45

0.244

503.1

B-1-4-1

6.06

25.79

1/27

0.215

791.1

B-2-4-2

6.24

21.41

1/32

0.225

582.5

B-3-4-3

7.86

17.28

1/39

0.269

608.9

Figure 6 plots the relation between flexural property and truss number. It can be seen from Fig. 6a that the peak load of hollow beams is about 35 kN when the truss number in the pure-bending region keeps consistent in 2. When the truss number in the pure-bending region rises to 4, the load-bearing capacity of hollow beams increases considerably from

138

J. Ye et al.

30.3 kN to 39.3 kN with the increase of truss number in the flexural-shearing region. Regardless of the truss number in the pure-bending region, the ultimate deflection of hollow generally decreases with the increase of truss number in the flexural-shearing region.

(a) Peak load

(b) Ultimate deflection

Fig. 6. The relations between flexural properties with truss number.

4 Conclusion To sum up, the hollow beams have reduced load-bearing capacity and deformability compared to that of solid beams. However, the deflection to span ratio of hollow beams is higher than 1/50, satisfying the requirements on the deflection at serviceability limit states. Therefore, UHDC is a promising material for 3D printed self-reinforced hollow structures without steel reinforcement. However, further investigations are needed to be carried out to optimize the hollow structures and simultaneously enhance the mechanical properties.

References 1. Wangler, T., et al.: Digital concrete: a review. Cem. Concr. Res. 123, 105780 (2019) 2. Khoshnevis, B.: Automated construction by contour crafting—related robotics and information technologies. Autom. Constr. 13, 5–19 (2004) 3. Lim, S., et al.: Modelling curved-layered printing paths for fabricating large-scale construction components. Addit. Manuf. 12, 216–230 (2016) 4. Vantyghem, G., et al.: 3D printing of a post-tensioned concrete girder designed by topology optimization. Autom. Constr. 112, 103084 (2020) 5. Mechtcherine, V., et al.: Integrating reinforcement in digital fabrication with concrete: a review and classification framework. Cem. Concr. Compos. 119, 103964 (2021) 6. Yu, K., et al.: Feasibility of using ultra-high ductility cementitious composites for concrete structures without steel rebar. Eng. Struct. 170, 11–20 (2018) 7. Ye, J., et al.: Effect of polyethylene fiber content on workability and mechanical-anisotropic properties of 3D printed ultra-high ductile concrete. Constr. Build. Mater. 281, 122586 (2021) 8. Wang, L., et al.: Mechanical bahaviors of 3D printed lightweight concrete structure with hollow section. Arch. Civ. Mech. Eng. 20, 16 (2020)

Development of Cementitious Metamaterial with Compressive Strain Hardening Characteristics Keisuke Nishijo1(B) , Motohiro Ohno2 , and Tetsuya Ishida2 1 Maeda Corporation, Tokyo, Japan [email protected] 2 The University of Tokyo, Tokyo, Japan

Abstract. The aim of this study is to fabricate a cementitious metamaterial with compressive strain-hardening characteristics by directly 3D printing fiber reinforced cement mortar. The internal structure of the metamaterial was designed such that vertical and horizontal elliptical holes were hollowed out alternately from a rectangular prism. This structure was intended to “squeeze” inward when compressed even after cracking. To validate the conceptual design, specimens were prepared by using two methods: casting fiber reinforced cement mortar in 3D printed plastic molds and directly printing fiber reinforced mortar by an extrusion-based 3D mortar printer. The hollowed specimens, as well as a control cast specimen without holes, were subjected to compression testing, and the surface strain was measured by using Digital Image Correlation. The test result showed that the load first increased almost linearly with the increase of deformation, followed by 76% load drop at the onset of the first crack. After the load drop, the load increased again, instead of showing brittle failure, and could recover up to 76% of the peak. When compared with the filled specimen, the hollowed cast specimen showed significantly larger deformation capacity with 14 times larger energy absorption due to the post-peak strain hardening behavior. On the other hand, the hollowed 3D-printed specimen showed the strain-softening behavior after the peak load, which could be attributed to the lack of accuracy in printing the elliptical holes. Nevertheless, significantly high energy absorption could be achieved in the printed specimen as well. Keywords: Metamaterial · Compressive strain hardening · 3D printing · Digital Image Correlation · Energy absorption

1 Introduction 1.1 Metamaterial With the enhanced freedom of design, additive manufacturing of cementitious materials opens up a new possibility – cementitious metamaterial – which enables to acquire unprecedented properties that cannot be realized from inherent properties of the constituent material. The internal shape, size, and geometry of the material are architected © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 139–144, 2022. https://doi.org/10.1007/978-3-031-06116-5_21

140

K. Nishijo et al.

at mm-cm scales, to achieve unique mechanical properties such as auxetic behavior. This concept can be named as “mesoscale design” of cementitious materials, which is in between the material design and structural design in the concrete engineering. There have been a limited number of previous studies that dealt with such mesoscopically architected cementitious materials [1–4], but most of them fabricated the metamaterials by casting cementitious materials into 3D printed molds. The present study aims to directly 3D print cementitious metamaterials. Specifically, the mesoscale structure, rather than the constituent material, is tailored to achieve unique strain-hardening characteristics and higher energy absorption efficiency. The inherent brittle behavior of cement-based materials is to be overcome by the mesoscale design.

2 Materials and Methods 2.1 Metamaterial Design The conceptual design of the internal structure was derived from cementitious cellular composites with negative Poisson’s ratio [1]. It was designed such that vertical and horizontal elliptical holes were hollowed out alternately from a rectangular prism (Fig. 1a). This structure was intended to “squeeze” inward when compressed even after cracking (Fig. 1b). Figure 1c shows the Finite Element Analysis (FEA) simulation of the metamaterial under compression. Although a clear auxetic behavior cannot be seen in the simulation result, the metamaterial is relatively squeezed inward compared with a regular prismatic block showing a normal (i.e. positive) Poisson’s ratio. a)

b)

c) Compression

Max. Principal Strain (μ) 3000 2000

1000 0

Fig. 1. (a) Metamaterial design; (b) intended “squeezing” behavior leading to compression strainhardening; (c) FEA simulation for the compressive behavior

2.2 Specimen Preparation Three series of specimens were prepared in this study: a hollowed metamaterial made by casting (MM-C), a 3D-printed metamaterial (MM-P), and a prismatic filled specimen as a control (Fill-C). The specimen size for each series is shown in Fig. 2. The material used is fiber reinforced cement mortar with the water to cement ratio of 0.25, the sand to cement ratio of 0.20, and 1% fiber volume fraction. The cement is ordinary Portland

Development of Cementitious Metamaterial

141

cement, and the sand has the maximum size of 2 mm. The fiber is polyvinyl alcohol fiber with the length of 12 mm and the diameter of 39 microns (Kuraray REC15). A superplasticizer was added to adjust the rheology of the fresh mortar.

Fig. 2. Specimen sizes for three series

The Fill-C and MM-C specimens were made by casting mortar in molds. The molds were 3D printed by using a commercial resin 3D printer and a water-soluble BVOH (butenediol vinyl alcohol copolymer) filament (MELFIL, Mitsubishi Chemical). The fresh mortar was cast in the molds and the hardened specimens were then immersed into tap water. With this method, only the molds were dissolved and specimens with the complex internal structure could be easily prepared (Fig. 3a). A gantry-type 3D concrete printer was used to prepare the MM-P specimen. A circular nozzle with diameter of 10 mm was used and the height of a layer was set at 10 mm. Detailed printer settings and specifications can be found elsewhere [5]. The dimensions of the printed specimen are larger than those of other series in order to secure printing accuracy.

Fig. 3. Methods for specimen preparation: (a) casting fiber reinforced cement mortar into 3D printed molds, (b) directly printing the mortar by a 3D concrete printer

2.3 Compression Test Compression tests were conducted on the specimens at the age of 3 days. The surface strain/deformation of each specimen was measured by using two-dimensional Digital

142

K. Nishijo et al.

Image Correlation (DIC) technique. The surfaces of the specimens were sprayed with black and white paint to generate random speckle patterns. Images of the specimen were taken at a regular interval during the test by a digital single-lens reflex camera. The stress distribution in the metamaterial should be complex due to the existence of holes in the body. For simplicity, the overall stress was defined as the load divided by the cross-sectional area (a * b shown in Fig. 2) as in the filled prismatic specimen.

3 Results and Discussions 3.1 Cast Specimens Figure 4 shows the stress-strain curves of Fill-C and MM-C. The strain data were obtained by DIC. Brittle failure was observed on Fill-C when the stress reached the peak of 64 MPa at the compressive strain of 0.24%. On the other hand, MM-C exhibited a unique behavior instead of brittle failure; when the peak stress of 21 MPa at the strain of 0.60% was reached, the stress dropped to 5 MPa and then recovered as the strain increased. Then, the post-peak stress level eventually recovered up to 16 MPa at the stain of 5.9%, which was 76% stress recovery from the peak. Due to this post-peak strain hardening behavior, MM-C showed significantly larger deformation capacity with 14 times larger energy absorption compared with Fill-C.

Fig. 4. Stress-strain curves of Fill-C and MM-C (top) and the maximum principal strain maps of the MM-C at each point on the stress-strain curve (bottom)

Development of Cementitious Metamaterial

143

Figure 4 also shows distributions of maximum principal strain at the peak load and during the post-peak hardening stage. High-strain regions at the peak load (Fig. 4a) roughly match those in the FEA simulation (Fig. 1c). When the stress dropped after the peak, a diagonal crack was initiated and propagated through the high-strain regions, connecting the horizontally-oriented holes (Fig. 4b). Then, the stress increased again while the horizontal holes were being closed (Fig. 4c). After the second peak was reached, another diagonal crack crossing the previous one appeared, and the stress started decreasing again. Subsequently, the holes along the new crack were being closed while the stress increase was observed up to the third peak (Fig. 4d). The test results suggest that the intended stress distribution around the elliptical holes were roughly achieved in the cast hollowed specimen, although the expected “squeezing” behavior was not clearly observed. Instead, closure of the elliptical holes and the resultant interlocking of matrices around the holes resulted in the post-peak hardening behavior. 3.2 Printed Specimen Figure 5 shows the stress-strain curve of MM-P and its maximum principal strain maps. Unfortunately, some regions in the strain maps could not be analyzed by DIC, probably due to poor quality of random speckle patterns around the regions.

Fig. 5. The stress-strain curve and maximum principal strain maps of MM-P

Despite the relatively limited analyzed regions, it can be seen that distinct high-strain regions are not observed around the elliptical holes, unlike those in MM-C. This suggests that the intended stress distributions could not be achieved in MM-P. It should be noted that holes printed are not completely elliptical and that large voids were unintentionally

144

K. Nishijo et al.

created. As a result, this series did not clearly show strain-hardening characteristics, but the behavior seems rather strain-softening; after reaching the peak stress of 15 MPa at the stain of 0.14%, the stress level at the second peak was not surpassed by those at the subsequent peaks. Although distinct post-peak strain-hardening characteristics could not be achieved in the printed metamaterial due to the lack of printing accuracy, remarkably high energy absorption was demonstrated in MM-P as well, which cannot be achieved in the constituent material without architected internal structures.

4 Conclusions The present study has experimentally demonstrated the feasibility of fabricating cementitious metamaterials by direct printing of cement-based composites. Specifically, tailoring the mesoscale internal structure imparted post-peak compression-hardening characteristics to the constituent material that showed a brittle behavior under compression. However, the effect of the mesoscale design can be affected by the dimensional accuracy for 3D printing. Technological development in the field of 3D concrete printing is essential for this novel technology – cementitious metamaterial. To further advance the cementitious metamaterial technology, a more sophisticated design method for the internal structure should also be developed. Here, a computational generative design approach would be promising, where effective mesoscale designs are heuristically identified with the aid of Artificial Intelligence (AI). This will be the scope of the future study. Acknowledgments. This study was supported by JSPS KAKENHI Grant Number JP21H01403. The 3D printer used in this study was developed in a joint effort by the National Institute of Technology Ariake College, Taisei Corporation, and AKTIO Corporation.

References 1. Xu, Y., et al.: Cementitious cellular composites with auxetic behavior. Cem. Concr. Compos. 111, 103624 (2020) 2. Moini, M., et al.: Additive manufacturing and performance of architectured cement-based materials. Adv. Mater. 30(43), 1802123 (2018) 3. Rosewitz, J.A., et al.: Bioinspired design of architected cement-polymer composites. Cem. Concr. Compos. 96, 252–265 (2019) 4. Sajadi, S.M., et al.: Deformation resilient cement structures using 3D-printed molds. iScience 24(3), 102174 (2021) 5. Nishijo, K., Ohno, M., Ishida, T.: Quantitative evaluation of buildability in 3D concrete printing based on shear vane test. In: Wang, C.M., Dao, V., Kitipornchai, S. (eds.) EASEC16: Proceedings of the 16th East Asian-Pacific Conference on Structural Engineering and Construction, 2019, pp. 1891–1901. Springer, Singapore (2021). https://doi.org/10.1007/978-981-15-80796_174

Consistency of Mechanical Properties of 3D Printed Strain Hardening Cementitious Composites Within One Printing System Karsten Nefs1(B) , A. L. van Overmeir2 , Theo A. M. Salet1 , A. S. J. Suiker1 B. Šavija2 , E. Schlangen2 , and Freek Bos1

,

1 Eindhoven University of Technology, Eindhoven 5600 MB, The Netherlands

[email protected] 2 Delft University of Technology, Delft 2628 CN, The Netherlands

Abstract. Previous research has shown that the material properties of a threedimensional printed strain hardening cementitious composite (3DP-SHCC) can significantly vary, depending on the printing system with which it is produced. However, limited research has been performed on the reproducibility of hardened mechanical properties under identical printing conditions. In this study, the consistency of hardened properties, including compressive strength, flexural strength and deflection, and tensile strength and strain, was tested from materials printed during three separate but identical printing sessions. The research shows that with 3DPSHCC, significant variations in mechanical properties between printing sessions can be expected. Keywords: SHCC · 3DCP · Printing control · Variation

1 Introduction The limited ductility of 3D printable mortars is still one of the bigger challenges within the field of 3DCP. Nevertheless, a strain hardening cementitious composite (SHCC) is a material that displays a relatively ductile behaviour, and has been successfully applied in 3D concrete printing [1]. The effect of different printing systems on the mechanical properties of 3DP-SHCC was investigated by Figueiredo et al. [2]. Here it was found that both the mechanical and the physical properties of 3DP-SHCC specimens manufactured at Delft University of Technology (TUD) and Eindhoven University of Technology (TU/e) were significantly different, which was mainly attributed to the differences in equipment (geometries, brands, types, specifications). However, the consistency of mechanical properties of 3DP-SHCC elements printed on the same printing system, but over different printing sessions, was not investigated and is therefore the topic of the current study.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 145–151, 2022. https://doi.org/10.1007/978-3-031-06116-5_22

146

K. Nefs et al.

2 Methodology 2.1 Materials The material composition of the 3DP-SHCC mix used in the 3D printing experiments is presented in Table 1. A detailed description of the deployed mixing procedure can be found in Van Overmeir et al. [3]. The material is mixed in batches with a Hobart A200N mixers. Due to the maximum torque capacity of the mixer, the batches were limited to a volume of 3.5 L. Table 1. 3DP-SHCC composition, in [kg/m3 ] including viscosity modifying agent (VMA), super plasticizer (SP) and polyvinyl alcohol fibres (PVA). Blast furnace Cement Silica fume Limestone Sand Water VMA SP slag 42.5 N ( 0.05), (red/✘; probably not from the same distribution (p value < 0.05)). Printing session Print A Print B Print C

T-test (Compressive strength) Print A Print B Print C -

T-test T-test (Bending stress) (Uniaxial stress) Print A Print B Print C Print A Print B Print C -

4 Conclusions This aim of this study is to evaluate the consistency of mechanical properties of 3DPSHCC, obtained from separate but identical printing sessions. It may be concluded that: – Objects printed in separate printing sessions can display significant (but not dramatic) differences in their mechanical properties. – Properties related to deformation capacity (i.e., deformation in bending and strain in uniaxial tension) are quantitatively more strongly affected by the printing session than tensile strength properties (i.e., bending- and uniaxial strength). – The consistency between different printing sessions (i.e., whether results may have originated from the same statistical distribution) is not necessarily the same for compressive and (bending and uniaxial) tensile strength properties. From the above research results, it is concluded that several separate printing sessions are required to make unambiguous quantitative statements about mechanical properties of 3DP-SHCC (e.g. in the context of standardized strengths). Further research is necessary to determine whether this also applies to other types of printable mortars. Acknowledgements. This research was funded through the NWO Open Technology Program, project ‘High Performance 3D Concrete Printing’, grant number 17251.

References 1. Figueiredo, S.C., et al.: An approach to develop printable strain hardening cementitious composites. Mater. Des. 169, 107651 (2019). https://doi.org/10.1016/j.matdes.2019.107651 2. Figueiredo, S.C., et al.: Quality assessment of printable strain hardening cementitious composites manufactured in two different printing facilities. In: Bos, F.P., Lucas, S.S., Wolfs, R.J.M., Salet, T.A.M. (eds.) DC 2020. RB, vol. 28, pp. 824–838. Springer, Cham (2020). https://doi. org/10.1007/978-3-030-49916-7_81 3. Linde, A., van Overmeir, S.C., Figueiredo, B.Š, Bos, F.P., Schlangen, E.: Design and analyses of printable strain hardening cementitious composites with optimized particle size distribution. Constr. Build. Mater. 324, 126411 (2022). https://doi.org/10.1016/j.conbuildmat.2022.126411

Consistency of Mechanical Properties

151

4. Figueiredo, S.C., et al.: Mechanical behavior of printed strain hardening cementitious composites. Materials 13(10), 1–2 (2020). https://doi.org/10.3390/ma13102253 5. Ogura, H., et al.: Developing and testing of strain-hardening cement-based composites (SHCC) in the context of 3D-printing. Materials 11(8), 1–18 (2018) 6. Shapiro, S.S., Wilk, M.B.: An analysis of variance test for normality (complete samples). Biometrika 52(3/4), 591 (1965). https://doi.org/10.2307/2333709 7. Wolfs, R.: Experimental characterization and numerical modelling of 3D printed concrete: controlling structural behaviour in the fresh and hardened state. Ph.D. thesis, Eindhoven University of Technology (2019)

Design and Digital Workflow

Uncertainty Quantification of Concrete Properties at Fresh State and Stability Analysis of the 3D Printing Process by Stochastic Approach Zeinab Diab(B) , Duc Phi Do, Sébastien Rémond, and Dashnor Hoxha Univ Orléans, Univ Tours, INSA CVL, 7494 Lamé, EA, France [email protected]

Abstract. This study aims at developing an efficient method that allows quantifying the uncertainties of concrete properties and their effect on the stability of structure during the 3D printing process. Following that, the well-known Bayesian inference will be chosen to characterize the uncertainties of the elastic and plastic properties of the cementitious material at fresh state using the results of experiments available in the literature. These characterized mechanical properties and their associated uncertainty will be then taken as input parameters for the stochastic analysis through which the probability of instability of the printed structure due to plastic and buckling collapses can be estimated. Our numerical results highlight the significant effect of uncertainty on the stability during printing of concrete structure, that has been ignored in the literature. Keywords: 3D concrete printing · Fresh concrete · Uncertainty quantification · Bayesian inference · Stability analysis · Stochastic approach

1 Introduction The accuracy characterization of the buildability properties of fresh concrete at early-age as well as their evolution with time is a major challenge in the structure stability analysis of 3D concrete printing (3DCP). Although different attempts (e.g., using rotational rheometry, unconfined uniaxial compression tests UUCT, direct shear tests…) have been carried out; the difficulty is mainly related to the transient state of fresh concrete from viscous fluid in solids in the printing time range. A linear time evolution of the concrete properties at fresh state was highlighted in [1], it was also observed a non-negligible dispersion of the obtained results at each age represented by the high coefficient of variation COV which can reach up to more than 20%. Neglecting such uncertainty can affect the accuracy of failure prediction of concrete printed structure. The quantification of concrete properties’ uncertainty and its effect on the structure stability analysis of 3DCP is the main contribution of this work. To this end, the wellknown Bayesian inference (BI) is used for the uncertainty quantification whilst the stochastic analysis based on the Kriging metamodeling technique is chosen to estimate the failure probability during printing against plastic collapse and elastic buckling. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 155–160, 2022. https://doi.org/10.1007/978-3-031-06116-5_23

156

Z. Diab et al.

2 Chosen Methodology 2.1 Adopted Failure Criteria of the Printed Concrete Structure The stochastic analysis of the concrete structure during printing consists of estimating the failure probability of two failure modes (the plastic collapse and elastic buckling) by considering the uncertainty of the concrete properties at fresh state. The former failure mechanism depends strongly on the shear strength τy , which according to the Mohr coulomb criterion is a linear function of the normal stress σ n and characterized by the cohesion C and the friction angle φ (Eq. (2)). In the elastic buckling, the failure occurs when the axial loading S (i.e., the self-weight) applied on the structure exceeds the buckling strength capacity R which depends on the elastic modulus of fresh concrete and on the geometry of the printed structure. Based on experimental results, Wolfs et al. [1] proposed a constant friction angle φ while the cohesion C and young’s modulus E increase linearly as a function of time t: E = E0 + ξE .t

(1)

τy = c(t) + σn. tan(φ) = C0 + ξC t + σn. tan(φ)

(2)

The stability of the printed structure that verifies both failure modes can be finally presented by the following performance function g(X), (with g(X) < 0 if failure):   g(X ) = Min gp (X ), geb (X ) (3) where: gp (X ) = 1 −

σ1 [(1 − K0 ) − (1 + K0 )sin(φ)] 2Ccos(φ)

(4)

S R

(5)

geb (X ) = 1 −

gp (X) and geb (X) refers to the plastic collapse and elastic buckling failure criterion respectively, and K 0 = σ3 /σ1 is the ratio of minor and major stresses. The random vector X consists of five parameters (i.e., X = [E 0 , ξE , C 0 , ξC , Tan(φ)]) that present the mechanical properties of concrete at early age. 2.2 Uncertainty Quantification of Random Variables by Bayesian Inference The uncertainty quantification of input parameters (i.e., the vector of random variables X = [E 0 , ξE , C 0 , ξC , Tan (φ)]) is an essential step for the stochastic analysis of the printed structure. In this work, the well-known BI is chosen for this aim. The BI process consists of computing the probabilistic distribution p(X|y) of the random variables X conditional on the observed data y: p(X |y) =

p(y|X )π (X ) ∫ p(y|X )π (X )dX

(6)

Uncertainty Quantification of Concrete Properties

157

In Eq. (6), the prior distribution of the random variables π(X), which is usually expressed in form of a modified normal distribution, represents our beliefs about X a priori (i.e., before any data has been observed). Thus, the BI written in Eq. (6) updates the belief about X to the posterior function p(X|y) using the observed data that reduces the discrepancy between observation y and the model prediction f(X). Assuming Gaussian distribution with zero mean and an unknown variance σ2 of discrepancy, for a set of N experimental measurements we have (Eq. 7):  N 1 1 T  (7) p(y|X ) = exp − − f − f (X )) (X )) (y (y i i N i=1 2σ 2 2π σ 2 The posterior function p(X |y) is usually implicit and can only be explored numerically by using generally the sequential sampling technique Markov Chain Monte Carlo (MCMC) [2]. In this work, the Bayesian inference implanted in the Matlab toolbox UQLab [3] is chosen to quantify the uncertainty of the fresh concrete properties. 2.3 Kriging-Based Reliability Analysis In general, stability analysis of structure considering the uncertainty of the various input parameters, known as reliability analysis, is undertaken by the Monte Carlo Simulation (MCS). Despite its robustness and accuracy, this method requires a large number of structural evaluations which limits its effectiveness. To overcome the shortcoming of MCS in terms of computation time, especially when the numerical simulation of structure response must be performed, advanced techniques such as Kriging surrogate can be chosen [4]. The principal idea of this Kriging metamodeling technique is to approximate the performance function g(X ) by a Gaussian process g(X ) that can be constructed from a very reasonable number of evaluations of structure response. Using this Kriging metamodel, the MCS can straightforwardly be applied to estimate the failure probability when the structure response of each sample of random vector X (i) (i = 1, 2, N) can be easily interpolated from the constructed surrogate:  (i)    

(i) ) , I g(X (i) ) = 1 if g(X ) ≤ 0 (8) Pf = N1 N I g(X i=1 (i) 0 if g(X ) > 0

3 Numerical Applications The present study is conducted in the framework of the European Interreg project CIRMAP (CIrcular economy via customizable furniture with Recycled Materials for public Places) in which the development of a new Design Methodology for Customized Shapes (DMCS) is among the principal objectives. In this context, the stochastic analysis as presented in this work provides an important tool to predict the stability during printing of small concrete structures such as urban furniture. For the demonstration purpose, the considered furniture like a chair (see Fig. 2) will be chosen in this work. The chair

158

Z. Diab et al.

is printed with a velocity 6 m/min using the deposited layers of b = 3 cm of width and h = 1 cm of height. Following that, the BI is undertaken using the results of [1] who performed 25 experiments on 5 specimens at multiple ages t = 0, 15, 30, 60 and 90 min by both uniaxial compressive test and direct shear test. In this inversion process, whilst the mean values of the random variables (i.e., X = [E 0 , ξE , C 0 , ξC , Tan(φ)]) can be chosen arbitrary for the prior normal distribution, we use however the mean values obtained from the classical fitting method that minimizes the mean square error as shown in [1]. A large coefficient of variation COV = 50% is also taken for all parameters in this uncertainty quantification process.

Fig. 1. Posterior distribution, mean and standard deviation of random variables from BI.

The results obtained from the BI as highlighted in Fig. 1 show quite similar mean values as determined by Wolfs et al. [1]. But a significant uncertainty represented by a high standard deviation of the young’s modulus and the friction angle can be remarked from our statistical analysis. For the numerical investigation, the constant value of Poisson’s ratio ν = 0.3 and densityρ = 2070 kg/m3 are chosen similarly as in [1]. We note also that each deposited layer is considered as a homogeneous material whose time-dependent mechanical properties vary only in the vertical direction. In Fig. 2, we present firstly the deterministic results of the stability analysis of the concrete structure during printing which depends strongly on the chosen values of the mechanical properties of fresh concrete. The numerical simulations in 3D by Abaqus software of the printed structure show as expected the maximum compressive stress at the bottom layer at which plastic failure occurs (Fig. 2a) whilst the deformation pattern at the critical height of elastic buckling is shown in Fig. 2b. The variation of the critical height of each failure mode is significant when a lower value of elastic modulus (E 0 , ξE ) and shear strength (C 0 , ξC , Tan(φ)) induces the lower critical heights. For example, an increase of the plastic critical height is about 20 cm (i.e., 20 deposited layers) when one reduces twice the standard deviation in the mean values of mechanical properties of fresh concrete. The corresponding increase of the critical height of elastic buckling

Uncertainty Quantification of Concrete Properties

159

is about 6 cm (i.e., 6 deposited layers). Comparing the critical heights of the two failure modes, our results also highlight the dominant effect of elastic buckling on the stability of the chair during printing as discussed in [1].

Fig. 2. Deterministic results of the concrete structure behavior during printing: maximum compressive stress at the plastic critical height (a), deformation pattern of elastic buckling (b), critical height of both failure modes using with different values (μX + k · σX ) (with μX and σX are the mean and standard deviation) of the random vector X (c).

The Kriging-based reliability analysis is then conducted to study the stability of the printed structure which incorporates the uncertainties of these input parameters X = [E0 , ξE , C0 , ξ C , Tan(ϕ)]).

Fig. 3. Probability of plastic collapse and elastic buckling of the chosen concrete structured during printing as function of the target height H t (a); influence of the layer width b, layer thickness h and the printing velocity V n on the probability of elastic buckling (b).

As expected, the higher height of the printed structure increases the probability of failure of each failure mode (see Fig. 3a). Especially, the failure probability of the elastic buckling can reach 100% when the height of printed chair exceeds 28 cm (i.e., 28 deposited layers). Corresponding to this height, the failure probability of plastic collapse is largely smaller than 1% that confirms the dominant failure mechanism during printing of elastic buckling corresponding to the particular set of concrete properties and the chosen structure. We investigate in the next step the influence of the printing parameters

160

Z. Diab et al.

(e.g., the layer width b, the layer thickness h as well as the printing velocity V n ) on the elastic buckling failure probability. By fixing a target height of the printed structure H t = 0.3 (m), the parameters of the printing process are chosen: b ∈ [3, 6] cm, V n ∈ [4.2, 7.8] m/min, h ∈ [8, 12] mm which are in general dependent on the 3D printer machine limitation or preference. The results shown in Fig. 3b and 3c reveal that the elastic buckling depends strongly on the layer width b of the printed chair whilst the effect of the printing velocity V n , and the layer thickness h is moderate.

4 Conclusions In this study, a probabilistic methodology to analyze the stability of the concrete structure during the 3D printing process is presented. On the one hand, the Bayesian statistical method is chosen to quantify the uncertainties of the properties of cementitious materials in the fresh state. On the other hand, the reliability analysis using the Kriging metamodeling technique is carried out to quantify the propagation of the uncertainty on the probability of failure of the structure during printing which verifies two failure modes i.e., elastic buckling and plastic collapse. Corresponding to the chosen mechanical properties of fresh concrete and their quantified uncertainty, our results confirm the dominant effect of elastic buckling on the stability during printing. Our numerical investigations show that the stability is highly dependent on the width of the printed layer while the effect of the layer thickness and the printing speed is moderate. Acknowledgement. This research work has been carried out in the frame of the Interreg CIRMAP (No: NWE 1062), financed by the European Regional Development Fund (ERDF).

References 1. Wolfs, R.J.M., Bos, F.P., Salet, T.A.M.: Early age mechanical behaviour of 3D printed concrete: numerical modelling and experimental testing. Cem. Con. Res. 106, 103–116 (2018) 2. Rappel, H., Beex, L.A.A., Hale, J.S., Bordas, S.P.A.: Bayesian inference for the stochastic identification of elastoplastic material parameters: Introduction, misconceptions and insights. Mech. Time-Depend. Mater. 22, 221–258 (2018) 3. Marelli, S., Sudret, B.: UQLab: a framework for uncertainty quantification in Matlab. In: Proceedings of the 2nd International Conference on Vulnerability, Risk Analysis and Management (ICVRAM2014), Liverpool, United Kingdom, pp. 2554–2563 (2014) 4. Do, D.P., Vu, M.N., Tran, N.T., Armand, G.: Closed-form solution and reliability analysis of deep tunnel supported by a concrete liner and a covered compressible layer within the viscoelastic Burger rock. Rock Mech. Rock Eng. 54, 2311–2334 (2021)

Simulation of 3D Concrete Printing Using Discrete Element Method Knut Krenzer1(B) , Ulrich Palzer1 , Steffen Müller2 , and Viktor Mechtcherine2 1 IAB Weimar – Weimar Institute of Applied Construction Research, Weimar, Germany

[email protected] 2 Institute of Construction Materials, Technische Universität Dresden, Dresden, Germany

Abstract. The article at hand presents an approach for analyzing the 3D Concrete Printing (3DCP) by means of the Discrete Element Method (DEM). An advanced user-defined simulation material model for fresh printable concrete has been developed to simulate extrusion, discharge and deposition. In addition, a calibration procedure is shown to find a fitting parameter set for the material model parameters based on experimental data. The calibration of the latter is an iterative adaption process, leading to a realistic representation of real printable concrete. Finally, an extrusion-based 3DCP process is exemplary simulated to show the potential of the simulation method for process analyses and to verify the applicability of the model. The developed simulation tool enables a better understanding of the extrusion process during 3DCP and a profound analysis of the material flow within the extruder. Based on this information, improvements in the machine layout and the process parameter settings can be identified, allowing for further printing process optimizations. Keywords: 3D concrete printing · Digital concrete · Discrete element method · Extrusion · Calibration

1 Introduction The importance of 3D Concrete Printing (3DCP) has been continuously increasing over the last years. As the shortage of construction labor becomes more and more obvious, alternative ways of construction technologies must be developed based on automated processing of building materials. The concrete used for 3DCP usually differs considerably from conventional vibrated concrete. The lack of sound knowledge on the complex rheological behavior of printable concrete makes it difficult to predict its interaction with the processing equipment, such as pumping pipeline, extruder and nozzle. Thus, numerical simulation is a promising tool to predict the material flow depending on consistency, machine layout and processing parameters [1]. So far, simulations are predominantly limited to the prediction of the stability of printed layers and their failure behavior depending on the number of layers that are printed on top of it and the time at rest [2–5]. Mostly, such concrete layers are “artificial” © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 161–166, 2022. https://doi.org/10.1007/978-3-031-06116-5_24

162

K. Krenzer et al.

as they are created explicitly in their given shape, but are not “printed”, simulating the corresponding material flow completely. These simulation approaches neglect phenomena related to the process-dependent material deposition and flow. The main goal of the present research is to develop a material model that is capable to simulate 3D-printable concrete within the printing process, covering the extrusion process and the discharge of the concrete at the nozzle. This approach should enable the analysis and optimization of the flow inside the extruder as well as an estimation of the layer shape and stability after material deposition. To realistically simulate the 3DCP, the material model needs to be calibrated with respect to the concrete that is actually used. Obtaining the required rheological parameters from experiments and transfer them into the numerical simulation represent challenging aspects in the framework of DEM and, therefore, require a suitable calibration procedure. A calibration procedure must be developed to iteratively fit real material behavior and the parameters of the simulation material model.

2 Modelling Approach To enable valuable, realistic simulations of 3DCP, a numerical material model must correctly capture the rheological behavior of printable, fresh concrete. For that, the Bingham model with its determining parameters yield stress and plastic viscosity offers a suitable basis. In general, there are two established simulation techniques to model fresh concrete, namely, the Computational Fluid Dynamics (CFD) and the Discrete Element Method (DEM). Both exhibit specific advantages and shortcomings regarding the modeling accuracy of fresh concrete [6, 7]. For optimizing the printing head including the extruder for 3D concrete printing, the DEM seems to be the technique of choice because of its low computational costs with respect to the complexity of the machine geometry and movement compared with CFD. The IAB Weimar and the TU Dresden completed successfully several research projects by investigating approaches to cover the behavior of fresh concrete [7–10]. The therein-developed models with two-layer particles and shear rate estimation are the basis for the contact model at hand for printable concrete. First trials regarding their application to 3DCP revealed that the so-called stickiness as a strong cohesion factor should be additionally considered, besides the rheological parameters prescribed in the Bingham model. Noteworthy, the thixotropic behavior and the structural build-up upon material deposition or during process interruptions are not considered in the present study, but will be included in future analyses.

3 Material Model Calibration The material parameters required for the material model calibration were obtained in several experiments focusing on the rheological behavior of a printable concrete consisting of 392 kg CEM I 52.5 R, 214 kg Fly Ash (Steament H4), 107 kg Microsilicasuspension (Elkem 9/1), 253 kg fine Sand 0.06–0.2 (BCS 413), 253 kg local Sand 0–1, 759 kg local Sand 0–2, 246 kg H2 O and 9 kg superplasticizer (BASF Sky 593) for 1 m3 of concrete. All tests were repeated several times to capture the sensitivity of material and

Simulation of 3D Concrete Printing Using Discrete Element Method

163

measurement variations. In the following, the most relevant tests are described (also see Fig. 1). For the uniaxial compression tests, a cylinder (h = 80 mm, r = 25 mm) was filled with printed concrete. The subsequent steps were as follows: (1) remove the cylinder, (2) press a plate on the top of the concrete sample with a fixed speed of 0.5 mm/s and (3) measure the height dependent force on the plate. The test yields a force-displacement diagram (see Fig. 2) and the final shape (see Fig. 1). The inclined plane test captures the interaction between fresh concrete and wall surfaces of a machine, such as the extruder or nozzle. The printed concrete sample was put on an adjustable plane and the angle was continuously increased until the sample started to slide down the plane. This angle of the plane was documented for different plane materials and wetness states of the plane (see Fig. 1). The angles varied between 18° and 36°. The flow table test was performed regarding to DIN EN 12350-5 (see Fig. 1). The final diameter varied from 355 mm to 397 mm. The “break-off” behavior of the fresh concrete was captured during the extrusion process. A visual evaluation of the video recording enabled the capturing of the maximum overhang of the extruded concrete strand extruded before the material broke off and fell (see Fig. 1). The maximum overhang varied from 70 mm to 110 mm. Experiment

Picture / S ketch

Uniaxial compression test (Squeeze Test)

Inclined plane (slip behavior on wall material)

Flow table test DIN EN 12350-5

lifting

shocks

Break off test

Fig. 1. Experimental setup of the calibration experiments.

164

K. Krenzer et al.

The described experiments were simulated using the DEM by recreating the geometry with CAD, importing it into the simulation software and defining the geometry movement of the relevant parts. For the simulations representing the calibration experiments, a sensitivity analysis was performed to determine the most relevant material model parameters on the respective results. As expected, yield stress and cohesion exhibit a strong influence, but also the elastic modulus of the outer particle shell had a high impact. The effect of the viscosity indeed was lower. Further parameters did not have a significant effect on the results. Subsequently, a swarm optimization algorithm was initiated for fitting the relevant input parameters of the DEM material model. The fitted parameter sets showed a good agreement between the experimental and the simulation results of the individual experiments. In Fig. 2, the results of the uniaxial compression tests are shown. The same quality of agreement was reached for the inclined plane test and the break off test.

Fig. 2. Comparison of the force-displacement curves in the uniaxial compression test: the best-fit simulation result and the lower, average and upper values of the experimental results.

All three optimized parameter sets were similar, so that a single set of simulation material model parameters was determined by using a pareto optimization. This set could fit all three tests reasonably well. In contrast, the additionally performed simulation of the flow table test did not yield satisfactory results, i.e. the experimental data could not be achieved. The authors believe that this may be due to the missing consideration of the thixotropy effects in the current simulation model.

4 Simulating 3D Concrete Printing The fitted material parameter set and the above-described material model were used to simulate the 3DCP. For this purpose, the extruder-based printhead developed at the TU Dresden was reproduced within the simulation environment; see Fig. 3. The extruder conveys the material down to the nozzle where it is placed on the ground or the previously printed concrete layers. The printer nozzle creates a concrete layer with sharp edges and a dimension of about 50 × 150 mm. A good visual match between real and simulated concrete layers was found. This is illustrated by (i) the sharp shape of the simulated concrete layers and the low deformation

Simulation of 3D Concrete Printing Using Discrete Element Method

165

under self-weight observed, (ii) the clearly distinguishable layer boundaries, and (iii) an increased height of each beginning layer; see Fig. 3. Furthermore, the discharge speed at the nozzle was similar in the simulation and the experiment, indicating that the material behavior is correctly modeled.

Fig. 3. Digital geometry of the concrete printer (left), the simulated printed concrete layers (top right) and the real printed layers from the experiment (bottom right)

After printing several layers of concrete, the self-buckling failure scenario known from the experiments occurred within the simulation too, leading to an instability and a tilt of the printed wall; see Fig. 4.

Fig. 4. DEM simulation (left) and real application (right) of a multi-layer concrete print with insufficient material stability leading to a collapse of the layers, shown from different angles.

5 Conclusions and Outlook A novel simulation material model for 3D concrete printing was developed based on previous works at the IAB Weimar and TU Dresden. The newly adapted model yields

166

K. Krenzer et al.

the distinction between different wetness states of the material-wall interactions and an adapted cohesion property, which represents the increased stickiness of the printable concrete in comparison to ordinary concrete. A series of experiments was performed to determine the rheological behavior of a fresh printable concrete. A pareto optimization was used to find a fitting parameter set to achieve a good agreement for several simulation experiments at the same time. The 3D concrete printing was successfully simulated within the Discrete Element Method, showing a good qualitative agreement with the experimental findings regarding the shape of the printed layers and the speed of the material flow during extrusion. The integration of a thixotropy function into the simulation model is the target of further research to increase the predictability of layer stability over time. Acknowledgements. Authors express their sincere gratitude to the Federal Ministry for Economic Affairs and Climate Action for funding the ongoing research project “Betonextruder 2.0 – Entwicklung eines materialflussoptimierten Betonextruders zur additiven Fertigung mit zementgebundenen Werkstoffen unter Baustellenbedingungen”, grant number: AiF-ZIM, ZF4196105WO9.

References 1. Perrot, A., et al.: From analytical methods to numerical simulations: A process engineering toolbox for 3D concrete printing. Cem. Concr. Compos. 122, 104164 (2021) 2. Wolfs, R.J.M., Bos, F.P., Salet, T.A.M.: Early age mechanical behaviour of 3D printed concrete: numerical modelling and experimental testing. Cem. Concr. Res. 106, 103–116 (2018) 3. Reinold, J., Nerella, V.N., Mechtcherine, V., Meschke, G.: Particle finite element simulation of extrusion processes of fresh concrete during 3D-concrete-printing. In: II International Conference on Simulation for Additive Manufacturing, Sim-AM2019, pp. 428–439 (2019) 4. Spangenberg, J., da Silva, W.R.L., Comminal, R., Mollad, M.T., Andersen, T.J., Stang, H.: Numerical simulation of multi-layer 3D concrete printing. RILEM Tech. Lett. 6, 119–123 (2021) 5. Mengesha, M., Schmidt, A., Göbel, A., Lahmer, T.: Numerical modeling of an extrusionbased 3D concrete printing process considering a spatially varying pseudo-density approach. In: RILEM International Conference on Concrete and Digital Fabrication, pp. 323–332 (2020) 6. Roussel, N., et al.: Numerical simulations of concrete flow: a benchmark comparison. Cem. Concr. Res. 79, 265–271 (2016) 7. Mechtcherine, V., Gram, A., Krenzer, K., Schwabe, J.-H., Shyshko, S., Roussel, N.: Simulation of fresh concrete flow using discrete element method (DEM) – theory and applications. Mater. Struct. 47, 615–630 (2014) 8. Mechtcherine, V., Shyshko, S.: Simulating the behaviour of fresh concrete with the distinct element method – deriving model parameters related to the yield stress. Cem. Concr. Compos. 55, 81–90 (2015) 9. Krenzer, K., Palzer, S.: Neue Modellierungsansätze zur Simulation fließfähiger Betone. In: Ludwig, H.-M. (eds.) Proceedings of the 18th IBAUSIL, vol. 2, pp 115–122 (2012) 10. Krenzer, K., Mechtcherine, V., Palzer, U.: Simulating mixing processes of fresh concrete using the Discrete Element Method (DEM) under consideration of water addition and changes in moisture distribution. Cem. Concr. Res. 115, 274–282 (2019)

Influence of Infill Pattern on Reactive MgO Printed Structures AlaEddin Douba(B) , Palash Badjatya, and Shiho Kawashima Department of Civil Engineering and Engineering Mechanics, Columbia University, 500 West 120th Street, New York, NY 10027, USA [email protected]

Abstract. The construction industry’s increasing interest in additive manufacturing has generated a parallel interest in alternative and supplementary binders. Reactive magnesia (MgO) binders are an attractive and sustainable alternative to Portland cement; they are produced at lower temperatures and can potentially absorb the equivalent amount of CO2 emitted during production within their service life via carbonation curing. While excessive evaporation in 3D printed Portland cement is usually sought to be prevented, the resulting increase in porosity from the higher surface exposure is a desirable feature for 3D printed MgO that can lead to higher carbon intake. In this work, we utilize nanoclays in combination with methylcellulose as rheological modifiers to produce 25.4 mm MgO paste cylinders with different infill patterns: two open with continuous hollow channels and one solid/closed to mimic a conventionally cast one. Two additional configurations were introduced where the top and bottom layers of the open infills were replaced by a closed layer or a “lid” to conceal the infills’ internal structure. The results show that 3D printing of MgO improves compressive strength over cast ones by up to 380% at 28 days of carbonation curing, reaching 40–48 MPa at 1.1 w/b. The results also suggest that infill patterns play a more critical role in stress transfer than carbon intake as finite element analysis confirmed the introduction of localized stresses at the lid layers. Keywords: Magnesium Oxide · Magnesia · Nanoclays · Sustainability · Carbon sequestration

1 Introduction Concrete is the second most used material after water and consequently, the manufacture of its most carbon-intensive component, Portland cement, ends up accounting for 7–8% of global CO2 emissions [1]. Global cement consumption is increasing [2] and so will the associated emissions, unless steps are taken to modify the material to make it less of a carbon emitter or to replace it with alternatives. A viable solution is to use magnesiumbased binders that can sequester CO2 to form relatively stable carbonates that provide adequate strengths for construction under accelerated carbonation conditions [3, 4]. MgO as a material has been used as an additive for various applications but its use as a cement alternative is not as widespread [5, 6]. A non-trivial advantage of MgO is that it can be © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 167–172, 2022. https://doi.org/10.1007/978-3-031-06116-5_25

168

A. Douba et al.

mixed with water to form a paste that can be worked with just like conventional cement. More importantly, studies have shown that MgO can absorb CO2 to form carbonates without the need for elevated temperatures and pressures. Also, MgO can be obtained from non-carbonate sources (e.g. Mg-bearing silicates, industrial wastes, brine) thereby potentially having lower embodied carbon than conventional Portland cement. Better carbonation and consequently strength gain have been previously achieved through controlling mixture proportions, addition of hydrating agents, and adjusting curing conditions [7, 8]. However, unreacted material remains in the cured systems, potentially limiting further strength gain. 3D printing can help to facilitate carbonation degree as it allows control of geometry to reduce the distance atmospheric CO2 must travel to reach unreacted MgO, while limiting the amount of material required through topology optimization. Additionally, while excessive evaporation due to the absence of formwork and high surface exposure in 3D printed Portland cement systems is usually sought to be prevented, the resulting increase in porosity can lead to higher carbon intake in MgO systems. There has been relatively modest interest so far in printing using MgO [9] and though results have been promising, the study of the effect of various geometric configurations on strength is limited, which this study aims to address.

2 Materials and Methods 2.1 Materials The MgO used in this study was obtained from Martin Marietta Magnesia Specialties with the commercial name MagChem-30. The MgO was of light-burned type, available as a powder, and its chemical and physical properties are provided in Table 1. Table 1. Chemical and physical properties of magnesium oxide

MgO 98.2

CaO 0.8

Loose bulk density (g/cm3) 0.35

Chemical composition (%) Loss on ignition (LOI) SiO2 Fe2O3 Al2O3 Cl SO3 0.35 0.15 0.10 0.35 0.05 1.7% Physical properties Median Activity index % Passing 325 Surface area particle size (sec) mesh (m2/g) (micron) 3-8 20-30 18 99

Attapulgite nanoclay (NC) supplied by Active Minerals as Acti-Gel 208 were used as a thixotropic additive, which have a length of 1.5–2 µm and diameter of 30 nm. To enhance paste cohesion and bleeding resistance, a 14,000 molecular weight methyl cellulose (MC) from Millipore Sigma was also used. This combination of admixtures has been demonstrated to be effective in tailoring the rheology of Portland cement pastes [10]. From static yield stress measurements and printing performance tests, it was found that at 1.1 w/b ratio a dosage of 3 wt. % NC and 1.5 wt. % MC (by addition) was the

Influence of Infill Pattern on Reactive MgO Printed Structures

169

lowest satisfactory content, and was the dosage used in this study. The NC was stored as a solution using a magnetic stirrer while MC was added directly in its as-obtained dry powder form. 2.2 Experimental Methodology

Table 2. Various print configurations studied

Infill #2 (open)

Infill #3 (closed)

Capped Uncapped Cross-section (hidden infill) (exposed infill)

Infill #1 (open)

A syringe table-top printer was used to 3D print MgO cylinders 25.4 mm height and diameter with three infill patterns, as shown in Table 2. Additional cylinders were prepared for infills #1 and #2 where the top and bottom layers were replaced with closed ones to mimic capping, hiding away the internal open infill structure. In addition, cast specimens with conventional molds were prepared and tested for comparison. To minimize evaporation loss during the first 24 h, the printed specimens were loosely covered with a plastic sheet after printing while ensuring that the sheet did not touch and disturb the specimens. The control specimens were cast in cylindrical molds and covered with a plastic sheet. After 24 h in ambient lab conditions, the samples were transferred to an incubator maintained at 25 °C, 80 ± 5% RH and 20% CO2 to be carbonated. Compressive strength tests were carried out at the end of 28 days of incubation on an MTS Criterion C43 Electromechanical Testing Machine at a loading rate of 1.27 mm/min. The top and bottom surfaces of all specimens were covered with gypsum paste caps to ensure uniform loading. At least 3 specimens per batch were used to calculate the final average.

170

A. Douba et al.

2.3 Finite Element Modeling Finite element analysis (FEA) of all tested specimens was carried out using Autodesk Fusion 360 to investigate the effect of infill patterns and capping on stress transfer. Parabolic tetrahedral elements with 10 nodes were used for meshing. To avoid stress concentration from localized change in curvature, all edges were smoothed with circular parabolic with 0.25 mm radius. Up to 6 refinement steps were carried out per model to reach a convergence tolerance of 5%. The total number of elements used varied between 427,528–865,803. The elastic modulus of MgO was assumed to be 68 GPa with a Poisson’s ratio of 0.19 [11]. Boundary conditions were set to fix the bottom and apply a uniform 1 MPa stress at the top, and maximum Von Mises tensile stresses were extracted as an indicator of stress transfer efficiency.

3 Results and Discussions

Fig. 1. Aggregated compressive strength results for all batches cured for 28 days in accelerated CO2 conditions

The results of compressive strength after 28-days of carbonation curing are presented in Fig. 1. It is apparent that all of the 3D printed specimens show a higher strength than the cast control. Infill #3 specimens exceeded the strength of cast specimens by 380%, where the former had an average strength of 42.6 ± 4.7 MPa and the latter had a strength of 8.9 ± 2 MPa. The open infill printed specimens also exceeded the strength of the cast specimens by a significant margin. Between the two open infills, uncapped infill #2 had a higher average strength than that of uncapped infill #1: 47.2 ± 4.9 MPa vs 22.2 ± 1.7 MPa at 28 days, respectively. Uncapped infill #2 was within the margin of error of infill #3 at 28 days. Capping of infill #1 and infill #2 reduced strength by 24% and 42%, respectively. Results indicate that capping does not lead to any positive impact on strength and was in fact detrimental to the strength of specimens. These results provide preliminary insight into the effects of infill pattern on structural strength. 3D printing clearly has an advantage over conventional casting when using MgO cement. A possible reason for this behavior was speculated as higher porosity in the 3D printed material in the work of Khalil et al. [9]. A higher porosity, due to the lack of protective casing molds, can lead to higher diffusion of CO2 , more carbonation,

Influence of Infill Pattern on Reactive MgO Printed Structures

171

and thus higher strength. Another reason for the differences in strength could be the influence of specimen geometry on localized stresses, discussed below. Table 3. FEA results for compression under 1 MPa Infill #1 Uncapped

Infill #2 Capped

Uncapped

Infill #3

Cast

Capped

Volume (mm3 )

5729

6976

6267

7523

9407

10756

Number of elements

581064

658530

865803

735517

427528

680671

Max tension (MPa)

2.50

15.27

2.73

10.67

2.71

2.86

(a) Capped infill #1

(b) Capped infill #2

Fig. 2. FEA showing localized stress concentrations at the transfer layer between the cap and internal structure (sliced view).

To further understand the differences in compressive strength of tested configurations, FEA was carried out and the results are summarized in Table 3. Uncapped infills #1 and #2, infill #3, and cast show similar stress concentration levels corresponding to the 1 MPa load applied in the model. Therefore, the differences in strength observed in 3D printed specimens versus cast ones can be attributed to improved carbonation. The differences between infills #1, #2 and #3 can be correlated to differences in the amount of material present rather than stress transfer where infills #1 and #2 represent 39% and 33% reduction in volume, respectively. Capping introduced significant localized stress concentrations between the cap and the internal structure, as shown in Fig. 2, which partially explains the decrease in strength discussed earlier.

4 Conclusion In this work, we showed that 3D printed MgO, achieved by utilizing a mixture of nanoclays and methyl cellulose rheological modifiers, is a promising material that can reach

172

A. Douba et al.

compressive strengths of 50 MPa. The increase in strength of 3D printed parts compared to cast ones was mainly attributed to increased porosity resulting from higher water evaporation, which, in turn, facilitated CO2 penetration. Finally, FEA results support that the density of the infill pattern has greater influence on compressive strength if no abrupt changes occur in the structure, where stress concentrations resulted from the abrupt introduction of the capping layers. Acknowledgments. The authors would like to acknowledge the National Science Foundation (Award #1653419) and Columbia University’s School of Engineering and Applied Science (SEAS) Interdisciplinary Research Seed (SIRS) Program for financial support.

References 1. Andrew, R.M.: Global CO2 emissions from cement production. Earth Syst. Sci. Data 10(1), 195–217 (2018). https://doi.org/10.5194/essd-10-195-2018 2. IEA: Technology Roadmap - Low-Carbon Transition in the Cement Industry, Paris (2018). https://www.iea.org/reports/technology-roadmap-low-carbon-transition-in-the-cem ent-industry 3. Al-Tabbaa, A.: Reactive magnesia cement. In: Eco-Efficient Concrete, pp. 523–543. Elsevier (2013) 4. Harrison, A.J.W.: Reactive magnesium oxide cements. U.S. Patent No. 7,347,896 (2008) 5. Du, C.: A review of magnesium oxide in concrete. Concr. Int. 27(12), 45–50 (2005) 6. Walling, S.A., Provis, J.L.: Magnesia-based cements: a journey of 150 years, and cements for the future? Chem. Rev. 116(7), 4170–4204 (2016). https://doi.org/10.1021/acs.chemrev.5b0 0463 7. Ma, S., Akca, A.H., Esposito, D., Kawashima, S.: Influence of aqueous carbonate species on hydration and carbonation of reactive MgO cement. J. CO2 Utilization 41, 101260 (2020). https://doi.org/10.1016/j.jcou.2020.101260 8. Unluer, C., Al-Tabbaa, A.: Enhancing the carbonation of MgO cement porous blocks through improved curing conditions. Cem. Concr. Res. 59, 55–65 (2014). https://doi.org/10.1016/j. cemconres.2014.02.005 9. Khalil, A., Wang, X., Celik, K.: 3D printable magnesium oxide concrete: towards sustainable modern architecture. Addit. Manuf. 33, 101145 (2020). https://doi.org/10.1016/j.addma.2020. 101145 10. Douba, A., Kawashima, S.: Use of nanoclays and methylcellulose to tailor rheology for three-dimensional concrete printing. ACI Mater. J. 118(6), 275–289 (2021) 11. Pereira, A.H.A., Venet, M., Tonnesen, T., Rodrigues, J.A.: Development of equipment for non-destructive characterization of elastic moduli of ceramic materials. Ceramica 56(338), 118–122 (2010)

Durability

Evaluation of Durability of 3D-Printed Cementitious Materials for Potential Applications in Structures Exposed to Marine Environments Fabian B. Rodriguez1(B) , Cristian Garzon Lopez3 , Yu Wang1 , Jan Olek1 , Pablo D. Zavattieri1 , Jeffrey P. Youngblood2 , Gabriel Falzone4 , and Jason Cotrell4 1 Lyles School of Civil Engineering, Purdue University, West Lafayette, IN, USA

[email protected]

2 School of Materials Engineering, Purdue University, West Lafayette, IN, USA 3 Facultad de Ingeniería, Universidad Nacional de Colombia, Bogotá, Colombia 4 RCAM Technologies, Los Angeles, CA, USA

Abstract. The rising interest in 3D-printing of concrete structures for use in marine environments requires development of concrete mixtures with adequate mechanical and durability characteristics. The incorporation of alternative cementitious materials, combined with careful selection of printing parameters has emerged as an effective way of controlling not only the fresh properties and printability of mixtures, but also their mechanical and durability properties. This paper presents the results of various durability related tests performed on 3D-printed mortars, including density, porosity, rate of water absorption and resistance to chloride penetration. Results of these tests indicate that the performance of mortar elements 3D-printed using controlled overlap process was similar to the performance of conventionally cast mortar elements with the same composition. Moreover, the results of the chloride transport related tests obtained from all specimens evaluated during the course of the study indicate low chloride ion penetrability, thus re-affirming that combination of the proposed material and 3D-printing method of fabrication have a potential for producing structural elements for applications in marine environments. Keywords: 3D-printed concrete · Durability · Corrosion · Chlorides · Absorption

1 Introduction 3D-Concrete Printing (3DCP) process has been intensively explored for a variety of construction applications [1]. Early efforts focused on the development of 3D-printing systems and mixtures designs with suitable rheological and mechanical properties [2]. More recent efforts focus on 3D-printing of structures containing reinforcement and on evaluation of their mechanical and durability characteristics [3]. In addition, the increased global interest in renewable energy generation has accelerated the development © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 175–181, 2022. https://doi.org/10.1007/978-3-031-06116-5_26

176

F. B. Rodriguez et al.

of offshore wind energy projects. These projects represent an excellent opportunity for the use of 3DCP for manufacturing wind turbine towers, anchors, and foundations for offshore projects [4]. In these applications, the design and manufacture of 3D-printed structures must account for factors related to durability of concrete exposed to chlorides, sulfates, and magnesium ions. Appropriate testing methods are also required to estimate the durability performance of 3DCP materials [5]. The role of 3D-printing process-induces heterogeneities within the manufactured elements, such as cold joints, macropores and interfacial regions, along with parameters that control these heterogeneities, have already been extensively studied in the past [6, 7]. However, the study of their effects on the transport properties of chlorides and the susceptibility to corrosion has been limited. This paper presents the results of evaluation of selected durability characteristics of cementitious mixtures developed for potential applications in 3D-printed structures exposed to marine environment. The internal structure of elements produced a gantrystyle and robotic arm, was modified by adjusting the printing parameters that control the geometry of individual filaments to create an overlap of the adjoining filaments. Specifically, the influence of the degree of filament overlap on the absorption and transport characteristics of 3D-printed elements was investigated.

2 Materials and Methods The materials used in the fabrication of cast and 3D-printed specimens included the following: Type I ordinary portland cement (OPC) compliant with ASTM C150, limestone filler, silica fume, Class C fly ash compliant with ASTM C618, sand with a maximum particle size of 4.75 mm, polycarboxylate-based high range water-reducing admixture (HRWRA) meeting the requirements of ASTM C494 for type F admixtures, viscositymodifying admixture (VMA) meeting the requirements of ASTM C494 for type S admixtures and tap water. An additional mixture, M13+CC, used for evaluation of rapid chloride penetration and water absorption, included calcined clay as the supplementary cementitious material (SCM) and rheology modifying material. The proportions of the two types mixtures used in the study are given in Table 1. Table 1. Mixture proportioning of 3DCP. Name

Cementitious materials (kg/m3 ) Cement

Fly Ash class C

Silica fume

Admixtures (mL/100 kg) Limestone filler

Calcined Clay

HRWRA

VMA

M13

377.6

125.9

62.9

62.9

–

468

1192

M13+CC

377.6

125.9

62.9

62.9

31.5–

468

1192

The 3D-printed elements used in the study were produced using two types of printers: (a) the Hyrel Hydra 16A printer (0.6 m by 0.35 m by 0.2 m building size) with a 16 mm

Evaluation of Durability of 3D-Printed Cementitious Materials

177

nozzle opening (Fig. 1a), and (b) the ABB 6-axis robotic arm (2.65 m maximum reach) with a 25 mm nozzle opening (Fig. 1b). The filament height was set to 10 mm for the different printing configurations and the nominal filament width for the toolpath was set between 20 and 30 mm. However, the actual printed filament width (i.e., estimated extruded width based on the printing parameters) varied from 25 to 40 mm (Fig. 1c–1f). The printing parameters used for the different elements are presented in Table 2. The filament overlap percentage, defined as the difference between the printed width of the filament and the nominal filament width expressed as a percentage of the latter (Fig. 1f), was varied by altering the pump flow rate. The effect of overlap percentage on durability was subsequently evaluated. (a)

(c)

Nominal width Layer height

(f)

0% overlap – Clear interface

Printed width 25 mm

5-8% overlap 10 mm

(d) 26.9 mm 20 mm

(b)

26% overlap 10 mm

(e) 25 mm

60% overlap

25 mm 10 mm

40 mm

Fig. 1. a) Hyrel-16A printer; b) 6-axis robotic arm; c–e) 3DCP elements produced with different filament overlap values; f) Dimensions of filaments and overlap of the 3D-printed layers. Table 2. Printing parameters used in 3DCP process. Printer

Hyrel

Material

M13

Nozzle size

Printing speed

Nominal width

Estimated width

Overlap

(mm)

(mm/s)

(mm)

(mm)

(%)

25

26.9

8

16

50

Hyrel

M13

16

100

20

25.2

26

Robot Arm

M13

25

200

25

40.3

61

Robot Arm

M13+CC

25

150

30

31.4

5

Different types of durability-related tests were performed on specimens obtained from 100 mm diameter cast cylinders and cores extracted from 3D-printed prismatic elements. In general, the test procedures followed the relevant ASTM standards with small modifications as described in the next paragraph. All specimens were tested after

178

F. B. Rodriguez et al.

60 days of curing with the exception of samples from M13+CC mixture which were cored at 28 days. The tests performed are summarized as follows: The densities, absorptions and volumes of permeable voids were determined following the procedure given in ASTM C642 except that the vacuum saturation, rather than water boiling, was used for purposes of determination of internal porosity of the specimens. The water-absorption tests followed ASTM C1585, except that the specimens were dried at 60 ºC, rather than at 50 ºC following the procedure described by Zhutovsky and Hooton [8]. This process has been demonstrated to improve the uniformity of sorptivity measurements and provide data which correlate better with other transport properties. Indication of chloride ions penetration was evaluated by conducting electrical conductance tests following the ASTM- C1202 standard. The total charged passed was adjusted considering the nominal cross-sectional areas of the cylinder.

3 Experimental Results An initial assessment of the quality of the internal structure of 3D-printed elements was made through the determination of the bulk and apparent densities. Figure 2a shows that the measured densities values for different configurations of 3D-printed elements are comparable to the cast specimens. This suggests that the 3D-printing process which involves overlap of the filaments does not lead to the reduction in density commonly observed in 3DCP materials. Moreover, the results for the percentage of permeable pore space and absorption levels of 3D-printed specimens (Fig. 2b), indicate minor reduction with an increase in filament overlaps (2%–4% and 1%–2% respectively), possibly caused by the elimination of vertical interfaces. This hypothesis was evaluated further using the results of rate of absorption and chloride penetration tests. 3.0

b) 25.0

2.5

20.0

2.0

Cast elements 3D-printed 8% overlap 3D-printed 26% overlap 3D-printed 60% overlap

15.0

1.5

%

Mg/m³

a)

10.0

1.0

5.0

0.5 0.0

Bulk density, dry

Bulk density Bulk density after after immersion immersion and vacuum saturation

Apparent density

0.0 Volume of permeable pore space

Absorption after immersion

Absorp. after immersion and vacuum saturation

Fig. 2. a) Bulk densities and apparent density; b) Volume of permeable pore space an absorption, for cast elements and different types of 3D-printed elements.

The results of water absorption tests are presented on Fig. 3a–3c. Initial and secondary rates of absorption were determined based on a linear relationship between the absorption (I) and the square root of time (s1/2 ) with a correlation coefficient of at least 98%. Of the five types of specimens tested, the 3D-printed specimens from mixture M13+CC with 5% overlap and M13 with 8% overlap resulted, respectively, in the highest and second highest of both, the initial and the secondary rates of absorption. In contrast, mixture

Evaluation of Durability of 3D-Printed Cementitious Materials

179

M13 with 26% and 60% overlap exhibited rates of absorption similar to those of cast specimens. However, the variability of the results for specimens with 60% overlap was higher compared to that observed for cast and 3D-printed specimens with 26% overlap. This demonstrates that overlapping of adjacent filaments has a significant effect with respect to reducing the level of absorption of 3D-printed elements. However, it should be pointed out that extreme overlapping, such as 60%, did not result in significant additional improvement in the permeability values. The results of the rapid chloride penetration tests performed on the five sets of specimens are shown in Fig. 3d. These results also show a significant effect of filaments overlap on the total electrical charge passed. According to ASTM C1202, a very low potential for chloride ions penetrability is estimated for charge values between 100 and 1000 coulombs. It can be seen that three of the five types of specimens (i.e., cast and 3D-printed with 26% and 60% overlap) fall into this category. In contrast, 3D-printed elements with 5% and 8% overlap fall within the “low penetrability” category. a) 4.0

b) 5.0

M13 Cast elements

2.5 2.0 1.5

1.0

Cast elements 3DP 26% overlap 3DP 60% overlap 3DP 8% overlap 3DP 5% overlap M13+CC

0.5 0.0

0

200

Rate of absoprtion (mm/√s)

c)

400

600

800

M13 Cast elements M13+CC 3DP 5% overlap M13 3DP 8% overlap M13 3DP 26% overlap M13 3DP 60% overlap

1.4E-02

1.2E-02 1.0E-02 8.0E-03 6.0E-03 4.0E-03 2.0E-03 0.0E+00

Initial rate of absorption

Secondary rate of absorption

M13 3DP 8% overlap M13 3DP 26% overlap

3.0

M13 3DP 60% overlap

2.0 1.0

0.0

1000

Time [√s]

M13+CC 3DP 5% overlap

4.0

d) Charge Passed (Coulombs)

I [mm]

3.0

Absoprtion [I] (mm)

3.5

Initial absorption

Secondary absorption

2000 1500 1000 500

0

Charge passed

Fig. 3. (a) Results of a)water absorption showing the range of results and average absorption values (dotted lines); b) Initial and secondary absorption; c) Initial and secondary rates of absorption; d) Rapid chloride penetration test, for cast and 3D-printed specimens.

4 Discussion Based on the previous results, it is possible to identify the beneficial effect of adjusting the printing parameters to produce an overlap of the filaments with respect to the

180

F. B. Rodriguez et al.

designed toolpath. The data suggests that a significant improvement in the quality of the microstructure of 3D-printed elements was achieved when transitioning from an 8% to a 26% overlap. This was achieved by extruding material over a segment of previously deposited filament in order to “densify” the interfacial regions, which are typically more susceptible to development of higher porosity. At 5% and 8% overlap it is surmised that the material from freshly extruded filament does not merge with the existing filament creating a weak bond influenced by the speed of the nozzle and gravity acting on the filament. For intermediate overlap, such as 26%, results of absorption and chloride transport suggest that the internal structure of the elements, specially, the quality of the interfacial region is improved by fusing the extruded filament with the existing one, without affecting the stability of the layer. However, it is evident that an excessive overlap can have detrimental effects to the structure of the elements. Elements with 60% overlap show higher values of conductance, as well as higher variability in the capillary pore system. In this case, it is surmised that increasing flow of material extruded around the interfacial region of the filaments may cause the lateral displacement of the material in the previously deposited filament and subsequent creation of discontinuities and high-porosity regions, affecting the quality of the printed layer and the whole element. At present, only limited number of studies focused on quantitative evaluation of absorption and chloride penetration levels on 3D-printed elements [9]. Thus, comparisons are often made with studies conducted on cast concrete. However, significant differences in composition of mixtures used for 3D printing and those used for traditional, i.e., cast concretes pose a difficulty with respect to evaluating the effects of 3D-printing on durability of resulting elements.

5 Conclusions The current work demonstrated that the controlled overlap of the adjacent 3D-printed concrete filaments influenced the transport properties of resulting elements. 3D printed elements with a moderate level of filament overlap attained durability performance comparable to cast counterparts. This approach is useful for 3D-printing applications that require reduced permeability to avoid durability problems (e.g., reinforcement corrosion). Further exploration of the effects of the percentage of overlap, interlayer permeability and its interdependence on material’s rheology and printing time intervals is required to proof the efficacy of this approach in more general conditions.

References 1. Grasser, G., Pammer, L., Köll, H., Werner, E., Bos, F.P.: Complex architecture in printed concrete: the case of the Innsbruck University 350th Anniversary Pavilion COHESION. In: Bos, F.P., et al. (eds.): DC 2020, RILEM Bookseries, vol. 28, pp. 1116–1127 (2020). https:// doi.org/10.1007/978-3-030-49916-7_106 2. Blaakmeer, J., Lobo, B.: A robust mortar and printing system. In: Bos, F.P., et al. (eds.) DC 2020, RILEM Bookseries, vol. 28, pp. 1091–1103 (2020). https://doi.org/10.1007/978-3-03049916-7_104

Evaluation of Durability of 3D-Printed Cementitious Materials

181

3. Bos, F.P., Ahmed, Z.Y., Wolfs, R.J.M., Salet, T.A.M.: 3D printing concrete with reinforcement. In: Hordijk, D.A., Lukovi´c, M. (eds.) High Tech Concrete: Where Technology and Engineering Meet, pp. 2484–2493. Springer, Cham (2018). https://doi.org/10.1007/978-3-319-594712_283 4. Wu, Y.C., Cotrell, J., Li, M.: Interlayer effect on fracture behavior of 3D printing concrete. In: Bos, F.P., et al. (eds.) DC 2020, RILEM Bookseries, vol. 28, pp. 537–546 (2020). https://doi. org/10.1007/978-3-030-49916-7_55 5. Rehman, A.U., Kim, J.H.: 3D concrete printing: a systematic review of rheology, mix designs, mechanical, microstructural, and durability characteristics. Materials 14 (2021) 6. Moini, M., Olek, J., Magee, B., Zavattieri, P.D., Youngblood, J.P.: Additive manufacturing and characterization of architectured cement-based materials via X-ray micro-computed tomography. In: Wangler, T., Flatt, R.J. (eds.) DC 2018, RILEM Bookseries, vol. 19, pp. 176–189 (2019). https://doi.org/10.1007/978-3-319-99519-9_16 7. Wolfs, R.J.M., Bos, F.P., van Strien, E.C.F., Salet, T.A.M.: A real-time height measurement and feedback system for 3D concrete printing. In: Hordijk, D.A., Lukovi´c, M. (eds.) High Tech Concrete: Where Technology and Engineering Meet, pp. 2474–2483 (2017). https://doi.org/ 10.1007/978-3-319-59471-2_282 8. Zhutovsky, S., Douglas Hooton, R.: Role of sample conditioning in water absorption tests. Constr. Build. Mater. 215, 918–924 (2019) 9. Van Der Putten, J., De Volder, M., Van den Heede, P., De Schutter, G., Van Tittelboom, K.: 3D printing of concrete: the influence on chloride penetration. In: Bos, F.P., et al. (eds.) DC 2020, RILEM Bookseries, vol. 28, pp. 500–507 (2020). https://doi.org/10.1007/978-3-030-499167_51

Two Year Exposure of 3D Printed Cementitious Columns in a High Alpine Environment Timothy Wangler1(B) , Asel Maria Aguilar Sanchez1 , Ana Anton2 , Benjamin Dillenburger2 , and Robert J. Flatt1 1 Institute for Building Materials, ETH Zurich, Zurich, Switzerland

[email protected] 2 Institute for Technology in Architecture, ETH Zurich, Zurich, Switzerland

Abstract. While digital fabrication technologies with concrete have focused primarily on process development and compressive strength in terms of performance, more recently studies evaluating the durability performance of 3D printed cementitious materials have started to appear. To date, however, no field performance assessments have been published regarding 3D printed cementitious materials. In this study, we present the condition of bespoke 3D printed columns that have been exposed to the high alpine environment of Riom-Parsonz, Switzerland, for a period of two years. Damage levels are variable from column to column, often appearing as vertically oriented cracks that can also track along weak layer interfaces. The damage is conjectured to be linked primarily to freeze-thaw damage and/or thermal or differential shrinkage strains. The link between column design and damage is discussed in depth, in particular as it relates to the current predominant use case of 3D printed cementitious materials as a lost formwork. Keywords: 3D printing · Digital fabrication · Concrete · Durability · Field exposure

1 Introduction The recent interest in digital fabrication with concrete, on an academic and industrial level, has led to a growing number of practical applications and demonstrations of the technology in the real world [1]. Much research attention until now has been focused, understandably, on process development to ensure successful production of digitally fabricated structures. The link between processing, properties, and performance, however, is also of vital importance, and in this sense, most performance related research of digitally fabricated concrete has focused predominantly on the property of strength, and especially the anisotropy that can be induced by the process [2]. No less important, however, is the performance with respect to durability. Of prime importance in any reinforced concrete structure when it comes to durability is the protection of the steel reinforcement from agents that can lead to corrosion, thus the transport properties of the cover layer of concrete must be understood to predict when corrosion could be initiated. Until now, because of difficulties introducing traditional © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 182–187, 2022. https://doi.org/10.1007/978-3-031-06116-5_27

Two Year Exposure of 3D Printed Cementitious Columns

183

steel reinforcement bars into printing processes, the layered extrusion method has been used primarily to print lost formworks, in which hollow cores are printed where steel reinforcement is inserted and structural concrete is later cast [1]. This raises interesting questions of using the printed concrete not just structurally but also as a “durability” layer to help protect the steel reinforcement (for example, from carbonation). It is clear that the layer-by-layer fabrication method has the potential to introduce “channels” where transport is enhanced, as has been determined in some studies [3, 4]. Additionally, a recent study has demonstrated worse behavior of printed concrete with respect to cast concrete in terms of freeze-thaw performance [5]. Until now, field studies of printed structures are nonexistent, due to the paucity of their existence until just a couple years ago. We present here a first look at field performance of 9 structural scale printed components exposed in a high alpine environment after two years. These structures represent the typical use case of a lost formwork and can already give some insight into how the component construction design, material, and process can have critical impacts on the performance.

2 Concrete Choreography The design, production, and installation of 9 concrete columns for a dancing stage of the Origen Cultural Festival in Riom, Switzerland, has been described in detail elsewhere [6]. The 3D printed material is a cementitious mortar with crushed limestone aggregates (dmax = 2 mm), printed with a CEM I binder with 8% silica fume and 15% fine limestone substitution accelerated by a ca. 5% substitution of a calcium aluminate cement paste. Each column is approximately 2.7 m in height and each was printed in approximately 2.5 h on a hardened concrete slab.

Fig. 1. Column typologies in Concrete Choreography, with example cross section: (left) trigonometric function design engine, and (right) mesh subdivision design engine. The central dashed and circular section indicates the part into which structural concrete was cast.

The columns consisted of a central, circular void (approximately 25 cm in diameter) with an ornamented, aesthetic exterior design based on a computational generative design algorithm. Two design algorithms were used for the columns, one based on

184

T. Wangler et al.

trigonometric functions, and one based on mesh subdivisions, each producing a distinct design typology, noted in Fig. 1. The cross sectional profiles of the design typologies are also seen in Fig. 1, showing how the inner structural core is linked to the decorative exterior for the respective design typologies. Steel reinforcement cages were inserted and the inner core was cast with a standard C25/30 concrete between 7 and 30 days after printing. Casting took place in two sessions, with half of the core filled first, and the rest of the core filled two days later. The columns were then transported via truck from the production site to their installation site in Riom, Switzerland in early June 2019. The columns were initially left open on the top, accumulating rainfall during the summer of 2019, before having a bituminous layer applied on top to seal them. The bituminous layer has degraded over the two years on certain columns, likely due to wildlife and moisture (Fig. 2).

Fig. 2. Concrete Choreography columns in Riom, Switzerland, in June 2019. (Photo: Benjamin Hofer)

3 Exposure and Damage Observation 3.1 Exposure Riom is a small community in the Surses Valley of the canton of Graubünden, Switzerland, sitting on the east facing side at an elevation of about 1250 m, surrounded by the 3000 m peaks of the Swiss Alps. Most relevantly, the climate is classified as Dfb in the Köppen classification, typical of the Swiss Alps, with cold, snowy winters and warm summers with distributed precipitation: average highs and lows in January are 1 °C and −8 °C, respectively, and in July, 22 °C and 9 °C. Average snowfall amounts for Savognin, a ski resort town within 1 km of Riom, are approximately 250 cm annually, and total precipitation for the year is approximately 1400 mm. The columns sit on an exposed wooden platform, receiving direct sunlight with little protection from the wind. Snow removal is done manually, without the use of deicing salts.

Two Year Exposure of 3D Printed Cementitious Columns

185

3.2 Description of Damage Two trips to record observations of the columns were taken, one in October of 2020 and one in August of 2021. Observations during both trips showed three primary types of damage, with differing levels of severity from one year to the next. The three types of damage were: 1) macrocracks, often vertically oriented, 2) superficial granular disintegration, and 3) efflorescence, sometimes associated to the macrocracking and granular disintegration. The vertically oriented macrocracks have been observed in the first trip, with a small increase in number on the next trip, depending on the type. These cracks are predominantly associated to the columns that have been designed via trigonometric functions, tend to be closer to the bottoms of these columns, and tend to track within depressions. A particularly severe example is show in Fig. 3 (left). On the columns of the mesh subdivision typology, macrocracks are much less common, but tend to be associated to the top of the columns (Fig. 3, right), especially where moisture can accumulate. This type of damage is also observed on the tops of some of the trigonometric typologies, but only where the moisture can pool.

Fig. 3. Macrocracking damage observed on columns. (left) Vertically oriented crack near foot of a trigonometric typological column. (right) Cracks near the top of mesh subdivision typological column, in areas where moisture could pool.

Superficial granular disintegration is depicted in Fig. 4 (left). This type of damage is rather widespread on all columns and not generally associated to macrocracking, but is sometimes more severe in specific areas, such as specific layers, and of note is that this particular column suffered a production pause of 45 min where the damage is the highest. Similar to the granular disintegration, efflorescence was observed globally, especially in the first trip. However, strongly localized efflorescence could be observed in some macrocracks, as seen in Fig. 4 (right). A scraping and subsequent analysis of the efflorescence revealed predominantly alkali carbonates. These high concentrations of efflorescence also tended to be associated with darker spots on the cracks of the columns, indicating high local moisture transport.

186

T. Wangler et al.

Fig. 4. (left) Superficial granular disintegration, where filament protrusions can be seen to have eroded. The high damage zone on the bottom is associated with a cold joint. (right) Efflorescence in a vertical macrocrack, with dark zone indicating moisture.

4 Discussion The observed macrocracking of these columns is most likely due to one of two reasons: 1) thermal or shrinkage cracking, or 2) water pooling and freezing. Both of these mechanisms are most likely at play, depending on the design. For example, large, vertically oriented macrocracks tended to be associated only with the trigonometric typology, and one can see in Fig. 1a that the external shell is rather far from the core, with a very high effective diameter, and only connected by a few “spokes”. As this exterior shell shrinks, its only restraint comes from the concrete slab at the bottom and the “spoke” points. Stresses should tend to concentrate in the depressions and protrusions as well, which is what is observed. All of these observations point to shrinkage cracking as the main culprit in this typology. These cracks did not significantly worsen from the first trip to the second, as well, indicating that most of this type of damage was already completed within the first year, and corresponding to shrinkage processes. The macrocracking that appears in the mesh subdivision typology, however, is more likely due to freezing. As seen in Fig. 1, there is not an exaggerated exterior shell diameter, so shrinkage stresses are lower compared to the trigonometric typology, and overall these columns are in quite good condition from a macrocracking standpoint. The cracks in the top of the column appear only where water can pool at the top, are only evident in columns with poor coverage of the top bituminous layer, and worsened from one year to the next, indicating a repeated damage process such as freezing and thawing. As a final point on the macrocracking issue, it should be stated that water pooling and freezing at the bottom of the trigonometric typological columns could also be a contributor to the cracking on those columns. The granular disintegration can possibly be caused by either crystallization of salts, or freeze-thaw damage, but more likely from the latter. Another study has shown that a similar printed mixture to the one of these columns starts showing superficial granular disintegration after about 150 freeze-thaw cycles [5]. With five months where the average low temperature dips below zero in Riom, this amount of cycles can certainly be reached within a year or two. It is also worth noting that this damage was observed to have worsened from one year to the next, indicating a continual degradation from freeze-thaw cycles. Localization of this damage, especially to certain layers, could also indicate a certain dependence of the material properties to susceptibility of this type of damage.

Two Year Exposure of 3D Printed Cementitious Columns

187

It is possible that certain layers where the production process was not consistent, for example with either too much or too little fluidity, allow for locally susceptible points for damage, as seen in Fig. 4. The efflorescence indicative of salt crystallization did show some interesting aspects related to moisture transport. The near global presence of efflorescence on all columns indicated that the general movement of moisture out from the core, and localized efflorescence tended to come from areas where moisture accumulated as well, further highlighting the need to understand how design and moisture transport could be interlinked. Finally, while damage from freezing seems to be the source of the disintegration observed, salt crystallization cannot be completely ruled out as a reason.

5 Conclusion Overall, the main conclusions that can be drawn from this study are that for printed formwork columns, the design plays an enormous role in durability. Obvious design rules such as minimizing zones of moisture accumulation should be taken into account, but it should also be noted that if printable concrete is acting as formwork, the tendency to shrink on the cast concrete core should be taken into account by proper material matching, or adequate mitigating measures. The connection between processing and durability must also be investigated more thoroughly, especially as tight process requirements tend to limit certain fixes, such as adding an admixture. Acknowledgments. The authors acknowledge funding support from the Swiss National Science Foundation (NCCR Digital Fabrication, Agreement # 51NF40-141853). We acknowledge the essential contribution of Patrick Bedarf and Angela Yoo (Digital Building Technologies) in development of The Concrete Choreography Project, and the commitment and passion of our students from the MAS Dfab, ETH Zurich 2018/2019. The authors are grateful to Giovanni Netzer and the Origen Foundation for their support of the Concrete Choreography project and support during observation trips. The authors also acknowledge Heinz Richner, Arnesh Das, and Ylenia Praticó for their support during the observation trips.

References 1. Bos, F.P., Menna, C., Pradena, M., et al.: The realities of additively manufactured concrete structures in practice. Cem. Concr. Res. 156, 106746 (2022) 2. Wolfs, R.J.M., Bos, F.P., Salet, T.A.M.: Hardened properties of 3D printed concrete: the influence of process parameters on interlayer adhesion. Cem. Concr. Res. 119, 132–140 (2019) 3. Schröfl, C., Nerella, V.N., Mechtcherine, V.: Capillary water intake by 3D-printed concrete visualised and quantified by neutron radiography. In: Wangler, T., Flatt, R.J. (eds.) DC 2018. RB, vol. 19, pp. 217–224. Springer, Cham (2019). https://doi.org/10.1007/978-3-319-995199_20 4. Sanchez, A.M.A., Wangler, T., Stefanoni, M., Angst, U.: Microstructural examination of carbonated 3D-printed concrete. J. Microsc. (2022). https://doi.org/10.1111/jmi.13087 5. Das, A., Sanchez, A.M.A., Wangler, T., Flatt, R.J.: Freeze-thaw performance of 3D printed concrete: influence of interfaces. In: Proceedings of Digital Concrete (2022) 6. Anton, A., Bedarf, P., Yoo, A., Dillenburger, B.: Concrete choreography: prefabrication of 3D-printed columns. Fabricate (2020)

Salt Scaling Resistance of 3D Printed Concrete Manu K. Mohan(B)

, A. V. Rahul, Geert De Schutter , and Kim Van Tittelboom

Magnel-Vandepitte Laboratory, Department of Structural Engineering and Building Materials, Ghent University, Ghent, Belgium {Manu.KurungodMohan,Kim.VanTittelboom}@UGent.be

Abstract. Extrusion-based 3D concrete printing is an emerging technology in the construction field due to the many advantages associated with it as compared to conventional mould casting technology. However, many aspects like durability and long-term service performance are yet to be investigated in detail. The present study focuses on understanding the salt scaling resistance of 3D printed concrete samples. 3D printed concrete samples were prepared with a Portland cement mixture on the one hand and a mixture containing a blend of Portland cement and blast furnace slag on the other hand. The printed samples were subjected to freeze and thaw cycles with a 3% saltwater concentration. It was observed that the 3D printed samples exhibited better resistance against salt scaling compared to the mould cast samples made with the same mixture. The pore structure of the 3D printed samples was characterized by mercury intrusion porosimetry. It was observed that the presence of a higher amount of interconnected and coarser pores at the interlayer region of the 3D printed samples, acting like pockets of air voids, facilitates the release of ice crystallization pressure during the freezing phase. The study gives insights into the durability characteristics and feasibility of using 3D printed concrete elements exposed to aggressive environmental conditions. Keywords: Concrete · 3D printing · Salt scaling · Slag · Porosity

1 Introduction Extrusion-based 3D concrete printing (3DCP) is a novel method of constructing structures based on a pre-defined computer model [1]. The concrete is deposited in a layerby-layer manner without the need for formwork. 3DCP is getting wide popularity in the construction sector due to different advantages such as enhanced geometrical freedom, which enables topological optimization and thus increases material usage efficiency, cost effectiveness, sustainability, etc. [1, 2]. Although several aspects of the 3DCP technology are being studied, there exist seldom studies on the long-term durability performance. Salt scaling is one of the major durability issues of concrete. Salt scaling is defined as superficial damage caused by freezing a saline solution on the surface of a concrete body. The damage is progressive and consists of the removal of small chips or flakes of material. The damage induced by freeze–thaw cycles result into surface scaling attributable mainly to the crystallization pressure generated when the liquid water in the pore system changes to ice [3, 4]. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 188–193, 2022. https://doi.org/10.1007/978-3-031-06116-5_28

Salt Scaling Resistance of 3D Printed Concrete

189

De-icing salts and contact with sea water could exacerbate the salt scaling lowering the melting point of ice, which may induce thermal shock-mediated stress. Salt precipitation may also damage the pore system of the element. Other mechanisms involved in decay include internal crystallization and a physical development known as glue spalling [5]. The current study compares the salt scaling performance of 3D printed concrete elements with mould cast concrete samples with two different binder systems. Also, with the pore structure of the samples were characterized by mercury intrusion porosimetry.

2 Methodology 2.1 Materials and Mixtures The materials used in this study include a CEM 52.5 N Portland cement, CEM 52.5 R Portland cement and ground granulated blast furnace slag (GGBFS). Fine aggregates with a maximum particle size of 2 mm were also used to make the mixtures [6–9]. A polycarboxylate ether-based superplasticizer and a cellulose-based viscosity modifying agent were used as the chemical admixtures. The mixtures were prepared with a constant water-to-binder ratio of 0.35. The composition of the mixtures are given in Table 1. The detailed procedure of the mixture design is described in previous publications by the authors [6–8]. Table 1. Composition of the mixtures Material

Quantity (kg/m3 ) Mixture 1

Mixture 2

CEM I 52.5 N

376

–

CEM I 52.5 R

–

795.4

GGBFS

376

–

Sand

1279

1193.1

Water

263

302.3

Superplasticizer

5.27

3.12

Viscosity modifying agent

0.75

10.97

Total weight

2301.3

2285.6

w/b

0.35

0.38

Aggregate/binder

1.7

1.5

2.2 3D Print Experiments and Sample Preparation for Salt Scaling Tests 3D printing experiments were carried out using a six-axis industrial robot connected with a screw-based extrusion system. To assess the influence of interlayer on the salt scaling

190

M. K. Mohan et al.

performance, about 1 m long, eight layered wall elements (40 mm wide and 10 mm height layers) were printed with both the mixtures. After printing, the wall elements were cured until the age of 7 days at 25 °C and 95% relative humidity. 50 mm diameter and 50 mm height cylindrical samples were extracted from the wall elements as shown in Fig. 1. Salt scaling resistance was determined from the damage induced during freeze–thaw cycles based on the weight of the scaled material following the procedures mentioned in European standard CEN/TS 12390-9:2016 [10]. At the end of every 7 freeze thaw cycles, the mass of the scaled material was measured accurately and the average value was reported. The experiments were conducted on six 50 mm φ × 50 mm height samples per series. The samples were named as Mixture number-P for printed samples and Mixture number-MC for mould cast samples.

Fig. 1. Location of the samples for 3D printed samples

2.3 Mercury Intrusion Porosimetry About 1 cm × 2 cm size samples from the printed elements were cut for MIP studies. For comparison, same size samples are cut out from mould cast specimens. The experimental details are provided in previous publications of the authors [11]. At the age of 7 days, the samples were stored in isopropyl alcohol for up to 4 days to stop hydration by the solvent exchange method. After four days of immersion in isopropyl alcohol, the samples were stored inside a vacuum desiccator until testing. The MIP tests were carried out by using a Pascal 140–440 series porosimeter from Thermo Scientific. Mercury was intruded into each of the samples using a pressure range varying from vacuum to a maximum pressure of 200 MPa. MIP tests were repeated twice to check the repeatability and a representative curve from the two tests is presented.

3 Results and Discussions 3.1 Salt Scaling Shows the cumulative scaled of mass due to salt scaling for both the mixtures in 3D printed and mould cast conditions. The cumulative mass loss was normalized with the

Salt Scaling Resistance of 3D Printed Concrete

191

area of the exposed surface. It is interesting to note that the mass of scaled off material is significantly higher in the case of mould cast concrete compared to 3D printed samples. This drastic difference in salt scaling performance can be observed from Fig. 1 as well. In order to understand the possible mechanisms involved in such a different behaviour, a closer look at the pore structure characteristics and crystallization pressure due to the formation of ice is needed. It was reported that long and interconnected pores exist at the interface region of the 3D printed concrete elements [12]. This could result in the absorption and transport of water though the printed element [1]. The formation of ice crystals in interlayers may cause a similar effect to that of freezing in air voids. This in turn could cause a suction on the pore fluids of the bulk concrete which may compensate the glue spall stress [5] (Table 2 and Fig. 2). Table 2. Results of the salt scaling study Sample name

Cumulative mass of scaled material per unit area in kg/m2 (after 56 freeze-thaw cycles)

Mixture 1-P

0.11 ± 0.02

Mixture 1-MC

0.51 ± 0.03

Mixture 2-P

0.14 ± 0.02

Mixture 2-MC

0.57 ± 0.03

Fig. 2. Salt scaling damages observed on (a) printed and (b) mould cast samples.

3.2 Total Porosity from MIP Figure 3 shows the cumulative pore volume vs pore entry diameter plots for both the mixtures. The main difference between the 3D printed and mould cast samples lies on the pore volume at different pore size ranges. From Fig. 3, it can be observed that the volume of pores in the size range of 100–0.1 μm is lower in the case of 3D printed samples compared to the corresponding mould cast concrete samples for both the mixtures 1 and 2. Also, it must be noted that the volume of pore in the size e range of 50 μm is higher in the case of 3D printed samples and therefore, it can be considered as a system with higher

192

M. K. Mohan et al.

amount of air voids (similar to air entrained concrete). The formation of the ice crystal in an air void imposes suction in the pore fluid thereby contracting the porous skeleton. This compensates the glue spall stress generated from thermal expansion mismatch [5].

Fig. 3. Cumulative pore volume vs pore size curve of the 3D printbale mixtures (a) Mixture 1 (b) Mixture 2

4 Conclusions The current study attempts to characterize the salt scaling phenomenon in 3D printed concrete elements and compares this with mould cast concrete elements. The study was conducted with two different mixtures as among which a Portland cement-blast furnace slag system and a rapid hardening PC system. The salient conclusions from the study are listed below: • The printed concretes showed much higher resistance to salt scaling as compared to mould cast concrete. This could be due to the suction created from the ice formation in the interlayers of printed concrete, thereby compensating the glues spall stress from the ice formation on the surface concrete. • The volume of pores at a size range of 100–0.1 μm reduces nearby interfaces as observed from the MIP studies. This could indicate that the interfaces can act as air voids present in the system.

Acknowledgements. Authors would like to acknowledge the financial support provided by SIM (Strategic Initiative Materials in Flanders) and VLAIO (Flanders agency for innovation & entrepreneurship) towards the 3D2BGreen project. The authors also acknowledge the companies, BESIX, ResourceFull and Witteveen+Bos for being the partners of the 3D2BGreen project.

Salt Scaling Resistance of 3D Printed Concrete

193

References 1. Mohan, M.K., Rahul, A.V., De Schutter, G., Van Tittelboom, K.: Extrusion-based concrete 3D printing from a material perspective: a state-of-the-art review. Cem. Concr. Compos. 115, 103855 (2021). https://doi.org/10.1016/j.cemconcomp.2020.103855 2. Mohan, M.K., Rahul, A.V., De Schutter, G., Van Tittelboom, K.: Early age hydration, rheology and pumping characteristics of CSA cement-based 3D printable concrete. Constr. Build. Mater. 275, 122136 (2021). https://doi.org/10.1016/J.CONBUILDMAT.2020.122136 3. Valenza, J.J., Scherer, G.W.: Mechanism for salt scaling. J. Am. Ceram. Soc. 89, 1161–1179 (2006). https://doi.org/10.1111/j.1551-2916.2006.00913.x 4. Valenza, J.J., Scherer, G.W.: A review of salt scaling: I. Phenomenology. Cem. Concr. Res. 37, 1007–1021 (2007). https://doi.org/10.1016/j.cemconres.2007.03.005 5. Valenza, J.J., Scherer, G.W.: A review of salt scaling: II. Mechanisms. Cem. Concr. Res. 37, 1022–1034 (2007). https://doi.org/10.1016/j.cemconres.2007.03.003 6. Mohan, M.K., Rahul, A.V., Van Tittelboom, K., De Schutter, G.: Rheological and pumping behaviour of 3D printable cementitious materials with varying aggregate content. Cem. Concr. Res. 139, 106258 (2021). https://doi.org/10.1016/j.cemconres.2020.106258 7. Rahul, A.V., Mohan, M.K., De Schutter, G., Van Tittelboom, K.: 3D printable concrete with natural and recycled coarse aggregates: rheological, mechanical and shrinkage behaviour. Cem. Concr. Compos. 125, 104311 (2022). https://doi.org/10.1016/J.CEMCONCOMP.2021. 104311 8. Mohan, M.K., Rahul, A.V., Van Tittelboom, K., De Schutter, G.: Evaluating the influence of aggregate content on pumpability of 3D printable concrete. In: Bos, F.P., Lucas, S.S., Wolfs, R.J.M., Salet, T.A.M. (eds.) DC 2020. RB, vol. 28, pp. 333–341. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-49916-7_34 9. Mohan, M.K., Rahul, A.V., van Dam, B., Zeidan, T., De Schutter, G., Van Tittelboom, K.: Performance criteria, environmental impact and cost assessment for 3D printable concrete mixtures. Resour. Conserv. Recycl. 181, 106255 (2022). https://doi.org/10.1016/J.RESCON REC.2022.106255 10. CEN/TS 12390-9. Testing hardened concrete - Part 9 : Freeze-thaw resistance with de-icing salts - Scaling (2016) 11. Mohan, M.K., Rahul, A.V., De Schutter, G.: Interlayer bond and porosity of 3D printed concrete. In: RILEM Bookseries, pp. 1–10 (2021) 12. Van Der Putten, J., Deprez, M., Cnudde, V., De Schutter, G., Van Tittelboom, K.: Microstructural characterization of 3D printed cementitious materials. Materials (Basel) 12, 1–22 (2019). https://doi.org/10.3390/ma12182993

Influence of the Print Process on the Durability of Printed Cementitious Materials Jolien Van Der Putten(B)

, M. De Smet, P. Van den Heede , Geert De Schutter , and Kim Van Tittelboom

Magnel-Vandepitte Laboratory for Structural Engineering and Building Materials, Ghent University, Tech Lane Ghent Science Park, Campus A Building 60, 9052 Ghent, Belgium [email protected]

Abstract. 3D concrete printing (3DCP) is currently being explored both in academia and practice and ensures a fast, economic, safe and formwork-free construction process. Although the elimination of formwork is one of the biggest advantages, it also removes the protection between the curing concrete and the surrounding environment and consequently, cracking resulting from shrinkage can be more pronounced. Additionally, the effect of the layered fabrication process and the absence of compaction could increase the porosity and the occurrence of weakly bonded interfaces. The combination of these three phenomena might affect the durability of 3DCP elements, as these interfaces form ideal ingress paths for chemical substances. In order to improve the long-term behavior, this study aimed to comprehend the correlation between different print process parameters and the resistance against carbonation. Therefore, multi-layered elements were fabricated with two different print techniques (2D and 3D) and two interlayer time gaps (0 and 30 min). To enable a complete comparison between both fabrication processes also conventional cast elements were considered. In general, a more pronounced CO2 penetration could be observed for printed elements, related to the higher porosity. Additionally, enlarged time gaps tend to be detrimental for the durability, however, this effect could be counteracted the 3D-print technique. The higher pump pressure improves the bonding between subsequent layers and the general long-term behavior. Keywords: 3D printing · Carbonation resistance · Durability

1 Introduction With an annual production of six billion ton, concrete can be considered as the most widely used and most important construction material nowadays [1]. Even though the material science has been progressing rapidly, the improvements made in the manufacturing process are limited; the construction process is still labor intensive, the fabrication of molds and the integration of reinforcement is time consuming, especially when the complexity of the design increases. Therefore, additive manufacturing (AM), defined as ‘the process of joining materials to make objects from 3D model data, usually layer upon layer’ is gaining ground in the construction industry [2]. The most popular AM-technique © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 194–199, 2022. https://doi.org/10.1007/978-3-031-06116-5_29

Influence of the Print Process on the Durability

195

is 3D concrete printing (3DCP). This method allows a fast, economic, safe and formworkfree construction process with more architectural design freedom. Notwithstanding the many advantages of 3DCP, the layered end process could result in a higher porosity and the occurrence of weakly bonded interfaces [3, 4]. Additionally, the elimination of formwork goes along with the removal of the barrier between the curing concrete and the surrounding environment which could lead to more pronounced shrinkage related cracking [5]. These cracks, the higher porosity and the weakly bonded interfaces will not only weaken the structural properties of the elements, but might also affect the durability as they form ideal ingress paths for chemical substances [6]. At this moment, the addition of reinforcement is still an issue. However, once this will become daily practice, different corrosion-inducing mechanisms will become a threat and the transport properties and durability aspects will become of major importance. A typical deterioration mechanism associated with reinforced concrete structures is carbonation, decreasing the alkalinity of the concrete due to the reaction of CO2 with various hydrates. This study therefore aims to comprehend the correlation between the applied print parameters (i.e. print process and interlayer time gap) and the durability of the cementitious material. In addition, a comparative study between printed and cast elements is made.

2 Materials and Methods 2.1 Materials and Mix Composition The printable mixture was composed out of Portland Cement (CEM I 52.5 N) in combination with siliceous sea sand (0/2), water (W/C = 0.37) and a polycarboxylic ether-based superplasticizer (SP) (Glenium 51, conc. 35%, BASF, Germany). The mix composition can be found in Table 1. Table 1. Mix composition Component

CEM I 52.5

Sea sand

Water

SP

Amount [kg/m3 ]

703

1055

257

0.47% (woc)

2.2 Print Process and Sample Preparation A Quick-point mortar extruder, mounted vertically on a fixed steel frame, was used to simulate an extrusion-based print process (Fig. 1A). The mortar extruder could be manually altered in height to ensure a proper clearance between the nozzle and the building platform. The extrusion nozzle was elliptically shaped (28 × 18 mm2 ). The extruder is able to print layers up to 300 mm length, with different printing speeds. Within the scope of this research, a printing speed equal to 1.7 cm/s was selected. The width and height of the printed layers equaled 28 mm and 10 mm, respectively. Specimens printed with this technique will be denoted as 2D-printed samples.

196

J. Van Der Putten et al.

The upscaled 3D print experiments were executed with a six-axis industrial robot (Fig. 1B). The fresh cementitious material was transported to the print head with a progressive screw-based pump, connected to a 5 m long rubber hose. The layers were extruded with an elliptical nozzle (33 × 20 mm2 ) at a print velocity of 15 cm/s. The width and height of the printed layers equaled 33 mm and 10 mm, respectively. Specimens printed with this technique will be denoted as 3D-printed samples. Irrespective of the applied printing process, sample preparation starts by extruding the cementitious material through the nozzle with a predefined velocity. For each sample a single base layer was extruded. After a predefined time gap (zero or 30 min, denoted as T0 or T30, respectively), a second layer was deposited on top of the previous one, starting at the same initial X-position to ensure an equal time gap along the entire specimen. This process was repeated until four-layered samples were obtained. After printing, specimens were cured in standardized conditions (20 ± 2 °C, 60 ± 5% RH) until the day of testing. The resistance against carbonation of printed specimens was compared with moldcast elements, denoted as CAST. Therefore, prismatic molds (160 × 40 × 40 mm3 ) were filled with mortar in two steps and compacted on a jolting table by 60 jolts. These molds were then covered with plastic foil and stored in standardized conditions (20 ± 2 °C, 60 ± 5% RH). After 24 h, the prisms were demolded and stored in the same environment until the day of testing.

A

B

Fig. 1. Small-scale mortar extruder (A) and six-axis 3D-printing robot with pump (B).

2.3 Carbonation Resistance The carbonation resistance was investigated according to the standard CEN/TS 1239010. To start the procedure, printed and mold-cast specimens were sawn into smaller elements with dimensions as depicted in Fig. 2A and 2B. Thereafter, 5 of the 6 sample sides were coated with an epoxy resin (Episol Designtop SF – Resiplast, Fig. 2C) to ensure unidirectional CO2 transport through the non-coated front surface. At the age of 12 days, these coated elements were stored in a carbonation chamber at 20 °C and 60% RH containing 1 vol% CO2 . Due to the specimens age at the start of the CO2

Influence of the Print Process on the Durability

197

exposure, this procedure deviates from the one specified in the standard, but can be explained as follows. In general, the carbonation front proceeds when all the material is carbonated. Due to the higher binder content in printable mixtures and as Portland cement is the only binder component, the amount of carbonatable material is high. Assuming that the carbonation front would proceed slowly, specimens were therefore placed in the carbonation chamber after a hardening period of 12 days. After a CO2 exposure of 1, 3, 5 and 7 weeks, a minimum of 3 samples per test series was split perpendicular to their noncoated surface. The carbonation front was visualized by spraying a 1% phenolphthalein solution on these freshy split surfaces, where the non-carbonated region turns into a characteristic magenta color as this region is still highly alkaline. After photographing the split surfaces, ImageJ analysis was performed to measure the carbonation depth every millimeter as a function of time. Side effects were excluded as only the second and third layer were considered within the investigations.

Fig. 2. Schematic representation of a four-layered printed element (A) and a traditionally cast specimen (B), schematic representation of the sample preparation indicating the coated and noncoated surfaces (C) and the prepared samples (D)

3 Results and Discussion Based on the colorimetric visualization of the carbonation front and subsequent ImageJ analysis, the penetration of CO2 can be expressed as a function of time (Fig. 3A–E). The latter figures visualize the carbonation front over the specimen’s height, where the X-axis represents the position of the interlayer between the second and third layer. Comparing different manufacturing processes, one can conclude that printed specimens show a lower resistance against CO2 ingress. After one week of exposure, the carbonation front penetrated only 1 ± 0.01 mm into the cast samples, while in case of printed specimens approximately 5 ± 0.45 mm of the bulk material was carbonated, regardless of the print technique. This can be attributed to the higher porosity and the higher shrinkage related (micro-)cracks formed due to the lack of molding [6]. Both phenomena increase the amount of preferential ingress paths for chemical substances. In case of traditional mold-cast elements, the penetration front is straight and uniform over the entire height, highlighting the homogeneity of the material. When the applied time gap equals zero, a

198

J. Van Der Putten et al.

similar shape of the carbonation front can be observed for both print techniques. Although the penetration depth is higher (±200% higher in case of printed samples after one week of CO2 -exposure), the printable material can also be assumed as homogenous when the layers are printed directly upon each other. Enlarged time gaps induce preferential ingress paths at the interlayer, especially in case of 2D-printed elements (Fig. 3E). The latter phenomenon can be attributed to the decreased interlayer quality. The higher the time gap between the layers, the drier the substrate material and the lower the interlayer bonding [4, 6]. This effect seems to be counteracted by the higher pressure executed when making use of the 3D printing technique. As both the preferential ingress at the interlayer and the penetration depth in general are lower compared to 2D-printed elements, one can conclude that the application of the 3D printing technique enhances the long-term behavior of the specimens.

Fig. 3. CO2 ingress in case of traditional cast elements (A), 3D-elements fabricated with a zero (B) or 30-min time gap (C) and 2D-elements with a zero (D) or 30-min time gap (E) (n = 3, error bars are left out for the sake of clearance).

For each test series, the measured carbonation depths [mm] were plotted as a function of the square root of the √exposure time [weeks] to determine an experimental carbonation coefficient Acc [mm/ weeks]. The results of the latter are represented in Fig. 4 and confirms the conclusions mentioned above. Acc is the lowest in case of mold-cast samples due to the higher compaction degree and the lower porosity. Higher time intervals result in a higher carbonation coefficient, as observed for 2D-printed elements, and the detrimental effect of an enlarged time gap can be counteracted by the 3D print technique. The higher carbonation coefficient in case of 3D-printed elements can be related to the environmental conditions. Although the specimens were stored in climatized conditions, they were manufactured in nonstandardized lab conditions. The lower temperature and RH during the first hours after printing could enlarge the early-age drying process and result in a higher (micro)crack formation. Considering mold-cast specimens as the reference, one can conclude that

Influence of the Print Process on the Durability

199

Fig. 4. Carbonation coefficient in case of different manufacturing techniques (CAST, 2D or 3D) and different time gaps (T0 and T30)

the resistance against carbonation is inadequate and the material would result in corrosion of eventual embedded reinforcement. To improve the long-term behavior, print process parameters or the material composition can be adapted. Another possibility is the application of adequate curing methods. All this would limit the desiccation of the outer layers and lower their porosity, impeding the diffusion of CO2 and slowing down the carbonation process.

4 Conclusions Based on the current investigation, one can conclude that the manufacturing process and the individual print parameters (e.g. interlayer time gap) will not only affect the structural behavior, but also the long-term durability behavior. Due to the higher compaction degree and the lower porosity, traditional manufactured elements carbonate less compared with printed elements. When the interlayer time gap equals zero, the carbonation front is uniform, irrespective of the applied fabrication technique. Enlarged time gaps increase the preferential ingress paths due to the lower surface quality of the substrate layer. However, the higher print pressure in case of 3D printing is able to counteract the latter phenomenon.

References 1. De Schutter, G., et al.: Vision of 3D printing with concrete — technical, economic and environmental potentials. Cem. Concr. Res. 112, 25–36 (2018) 2. Bos, F., et al.: Additive manufacturing of concrete in construction: potentials and challenges of 3D concrete printing. Virt. Phys. Prototyp. 11(3), 209–225 (2016) 3. Nerella, V.N., Hempel, S., Mechtcherine, V.: Effects of layer-interface properties on mechanical performance of concrete elements produced by extrusion-based 3D-printing. Constr. Build. Mater. 205, 586–601 (2019) 4. Van Der Putten, J., Deprez, M., Cnudde, V., De Schutter, G., Van Tittelboom, K.: Microstructural characterization of 3D printed cementitious materials. Materials 12(18), 2993 (2019) 5. Buswell, R.A., et al.: 3D printing using concrete extrusion: a roadmap for research. Cem. Concr. Res. 112, 37–49 (2018) 6. Van Der Putten, J.: Mechanical Properties and Durability of 3D Printed Cementitious Materials, in Faculty of Engineering and Architecture, Ghent University, Ghent (2021)

Freeze-Thaw Performance of 3D Printed Concrete: Influence of Interfaces Arnesh Das(B) , Asel Maria Aguilar Sanchez, Timothy Wangler, and Robert J. Flatt Institute for Building Materials, ETH Zurich, Zurich, Switzerland [email protected]

Abstract. The long-term performance of 3D printed concrete structures is essential and among the various durability issues, frost damage is one of key importance, especially in cold locations such as Switzerland. For 3D printed materials, the presence of layer interfaces and cold joints is a potential issue in terms of frost resistance. Therefore, after extrusion, both cast and printed samples were prepared, and they were subjected to 300 cycles of freeze-thaw in accordance with ASTM C666. It was found that printed samples have lower resistance to freeze-thaw conditions compared to their cast counterparts. The lower resistance of the printed samples could be attributed to the heterogeneity in the microstructure, in particular to the higher capillary porosity in the interface region compared to that in the bulk. The higher capillary porosity could be confirmed based on the sorptivity test results. Keywords: 3D printing · Frost damage · Interface · Capillary porosity

1 Introduction 3D concrete printing offers several advantages over conventional construction methods, and this is the reason why interest in this field has been flourishing in the recent years [1–4]. The recent years have seen enough success with regard to fabrication of structures by this method, however, the proper long-term performance of the structures from the point of view of both strength and durability is equally crucial. The literature mentions about some studies whereby the mechanical properties of 3D printed structures were investigated [5, 6], but studies concerning the durability aspect are still quite limited. Among the various durability issues, frost damage is one of key importance, especially in cold locations such as Switzerland. For 3D printed materials, the presence of layer interfaces and cold joints is a potential issue in terms of frost durability. Deterioration of concrete saturated with water (or deicing salts) can take place through exposure to cyclic freezing temperatures. One of the major ways of mitigating frost damage is to use air-entraining admixtures (AEAs) whereby the entrained micro-sized air bubbles in the concrete matrix help by reducing the distance over which water is pushed, which limits the magnitude of the expansive stress on the cement paste [7, 8].

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 200–205, 2022. https://doi.org/10.1007/978-3-031-06116-5_30

Freeze-Thaw Performance of 3D Printed Concrete: Influence of Interfaces

201

Concerning 3D printing with concrete, there are a number of processing steps involved such as pumping, acceleration/mixing and extrusion. As a first step, the effect of the various processing conditions on the air void system was studied and their implications concerning freeze-thaw behavior was assessed [9, 10]. Hardened air void analysis was done on samples prepared after each processing step (including both cast and printed samples). It was found that extruded cast and extruded printed samples have very comparable void systems and spatial distribution of voids. However, the previously conducted analysis had a certain limitation in that capillary porosity, particularly in the interface region was not characterized. Therefore, the present work was motivated by the intention of determining whether this inhomogeneous air void analysis may be sufficient for predicting frost resistance of 3D printed samples. For that we report data on both cast and printed samples that are subjected to actual freeze-thaw conditions. Samples with AEA were also considered. Additionally, results about the water absorption behavior of the samples are further included.

2 Materials and Methods Experiments were done on both cast and printed samples. The role of AEA was also considered, whereby these admixtures were included in the accelerator paste. The AEA dosage was fixed at 0.2% by weight of cement. Concerning the mix used for printing, the maximum aggregate size was 2 mm, its volume fraction was 0.45 and the water to cement ratio was 0.38. For further details about mixture design, procedure followed for the printing process and groups of mixes investigated, readers can refer to our preceding publication on this work [9]. The freeze-thaw tests were conducted on extruded cast and extruded printed samples. Two samples of each mix group were used for the freeze-thaw test. The test protocol followed was in accordance with ASTM C666 (procedure B) [11] with little modifications, with respect to the operating temperatures of the cycles. The samples were fully saturated before starting the cycles. Each cycle lasted for 4 h meaning that 6 cycles were done in a day. The freezing of the samples was done in air at −20 °C while thawing was done at 20 °C while being immersed in water. Mass loss measurements was the basis for assessing damage and this was done once a week, so approximately after every 42 cycles. Damaged pieces were analyzed by SEM imaging, with polished sections produced via vacuum impregnation with an epoxy resin to preserve the crack structure. Analysis was done with an SEM FEI QUANTA 200 3D at high vacuum. For the water absorption tests, samples (sizes were similar to those used for freezethaw test) were first dried in the oven at 60 °C until constant mass. Sorptivity tests were then conducted by placing these samples on a tray with a water height of approximately 2–3 mm. Only the extruded cast and extruded printed samples (groups 3–6) were used for this test. For the printed samples, tests were done in two ways – one in which the interface layers were placed parallel to the direction of water movement and one for which it was perpendicular to the direction of water ingress. The gain in mass was noted after 1, 5, 10, 30, 60 and 90 min. The sorptivity values were calculated based on the

202

A. Das et al.

slope of gain in mass (per area of cross-section) and the square root of time in minutes, hence they are reported in units of g/cm2 /min1/2 . Additionally, all dried samples were completely immersed in water until saturation and % increase in mass was recorded.

3 Results and Discussion Based on the hardened air void analysis conducted previously, the spacing factor was computed for each mix group. They were found to be 232 µm, 273 µm, 208 µm, 220 µm, 161 µm and 171 µm respectively for groups 1–6 [9, 10]. None of the cast samples (groups 1, 2, 3 and 5) showed any signs of damage and the mass remained constant. However, the printed samples (both with and without AEA) showed signs of damage even though their spacing factor values were much similar to their corresponding cast samples. The non-AEA sample showed first signs of damage after 168 cycles while for the AEA one, it was after 264 cycles. Figure 1 shows the variation of mass versus number of freezethaw cycles, while Fig. 2 shows an example of one of the printed samples at 0 and 300 cycles of freeze-thaw. As can be seen, damage mostly occurred in the form of surface scaling. The damaged pieces fell off from the outer surface of the layers and in most cases, the thickness of the damaged piece was similar to the thickness of each layer of the print. Figure 3 shows SEM images of the damaged pieces, where cracks can be seen to originate in the cement paste, which is expected in frost damage. The cracks follow an orientation parallel to the exposure surface and go through aggregate, however, it is not possible to determine if the cracks originate in the interface region or the bulk of the printed layer.

Fig. 1. Variation of mass with cycles of freeze-thaw for printed samples

The fact that only the printed samples showed signs of damage even though their spacing factors/void systems were similar to their cast counterparts indicates some degree of heterogeneity in the former. It is true that in most cases, spacing factor values directly translate to performance under freeze-thaw conditions, but the concept of spacing factor assumes that voids are mono-sized spheres uniformly distributed in the cement paste [12], hence it does not account for any heterogeneity. However, the resistance to frost damage is not solely influenced by distribution of entrained air voids but also on the volume of capillary porosity and the tensile strength.

Freeze-Thaw Performance of 3D Printed Concrete: Influence of Interfaces

203

Fig. 2. Samples at 0 cycles (left) and 300 cycles (right) of freeze-thaw. During thawing, samples were immersed in water in the same way as shown in picture with the surface having maximum area facing downward.

Fig. 3. SEM images of damaged printed sample after 300 cycles of freeze-thaw.

It is likely that the interface region in the printed samples has higher capillary porosity and is weaker compared to the bulk, which can explain its lower resistance to frost conditions. Table 1 shows the water absorption test results for the extruded cast and extruded printed groups of mixes. (It is to be noted that based on the analysis conducted previously, the cement paste content for all these mixes was found to be similar). When the samples are held in a perpendicular direction, the sorptivity values are quite similar to their cast counterparts. On the other hand, the printed samples show higher sorptivity when the interface lines are aligned in the direction of water ingress. The high volume of capillary pores is consistent the sorptivity test results. Further microstructure investigations would be required to determine the actual volume and distribution of capillary porosity, which was however beyond the scope of this work. A similar observation about capillary porosity was made by Sanchez et al. [13], wherein they observed that the progression of carbonation front is much quicker in the interface region compared to that in the bulk. It could be argued that carbonation must have proceeded faster at the interface due to presence of entrapped voids. However, the rheology of the printed material used in their study was very similar to that in the present work and in either case, there were absolutely no signs of entrapped voids anywhere in the matrix. The rheology of the material after extrusion is fluid enough for large sized entrapped voids to easily escape.

204

A. Das et al.

Table 1. Water absorption test results for extruded cast and extruded printed groups of mixes. Second column shows condition of the sorptivity test for printed samples.

Concerning the water absorption test results, the general notion is that water movement in concrete takes place through the capillary pores and the entrained air voids do not take up any water [8, 14]. But entrained air voids may be filled with water for a material with open cell (interconnected air voids) network, which could be the case for lowdensity foamed concrete materials [15]. However, a contrasting observation is made for the cast samples wherein it is seen that mixes with higher AEA content have higher water absorption. This can be explained by the findings of Wong et al. [8] that air voids tend to affect packing of cement particles and thereby promote some degree of heterogeneity. The air void-paste interface region was found to locally have a much higher w/c ratio, similar to the interfacial transition zone (ITZ) found at the aggregate-paste interface. The higher proportion of entrained air voids may be thought of as inert inclusions (similar to increasing aggregate volume fraction) which do not influence transport properties of concrete but in fact, they can be connected through the much smaller sized capillary pores. The connectivity would essentially depend on the size of the voids entrained by the AEA. Additionally, it has also been reported in literature that air voids may also reach some degree of saturation if they have access to water for long enough period of time [8], which can explain the results of complete immersion in case of cast samples.

4 Conclusions The hardened air void analysis revealed that extruded cast and extruded printed samples have similar void systems and spatial distribution of voids. The PPV analysis also predicted similar performance under freezing conditions. However, in order to validate the air void analysis results and also develop a better understanding of the effect of presence of interfaces in printed sample, conducting actual freeze-thaw tests was necessary. It was found that printed samples (both with and without AEA) have lower resistance to freeze-thaw conditions compared to their cast counterparts. The lower resistance of the printed samples could be attributed to the heterogeneity in the microstructure, in particular to the higher capillary porosity in the interface region compared to that in the bulk. The higher capillary porosity could be confirmed based on the sorptivity test results.

Freeze-Thaw Performance of 3D Printed Concrete: Influence of Interfaces

205

Acknowledgements. This research was supported by the NCCR Digital Fabrication, funded by the Swiss National Science Foundation. (NCCR Digital Fabrication, Agreement # 51NF40141853). The authors would like to thank Concrete and Asphalt laboratory at EMPA, Switzerland for giving access to the freeze-thaw chamber and specially acknowledge the support provided by Janis Justs and Marcel Kappeli during the period of experiment at EMPA.

References 1. Khoshnevis, B.: Automated construction by contour crafting—related robotics and information technologies. Autom. Constr. 13(1), 5–19 (2004). https://doi.org/10.1016/j.autcon.2003. 08.012 2. Buswell, R.A., Leal de Silva, W.R., Jones, S.Z., Dirrenberger, J.: 3D printing using concrete extrusion: a roadmap for research. Cem. Concr. Res. 112, 37–49 (2018). https://doi.org/10. 1016/j.cemconres.2018.05.006 3. Roussel, N.: Rheological requirements for printable concretes. Cem. Concr. Res. 112, 76–85 (2018). https://doi.org/10.1016/j.cemconres.2018.04.005 4. Wangler, T., Lloret, E., Reiter, L., et al.: Digital concrete: opportunities and challenges. RILEM Tech. Lett. 1, 67–75 (2016). https://doi.org/10.21809/rilemtechlett.2016.16 5. Feng, P., Meng, X., Chen, J., Ye, L.: Mechanical properties of structures 3D printed with cementitious powders (2015). https://doi.org/10.1016/J.CONBUILDMAT.2015.05.132 6. Van Der Putten, J., Deprez, M., Cnudde, V., De Schutter, G., Van Tittelboom, K.: Microstructural characterization of 3D printed cementitious materials. Materials 12(18), 2993 (2019). https://doi.org/10.3390/ma12182993 7. Du, L., Folliard, K.J.: Mechanisms of air entrainment in concrete. Cem. Concr. Res. 35(8), 1463–1471 (2005). https://doi.org/10.1016/j.cemconres.2004.07.026 8. Wong, H.S., Pappas, A.M., Zimmerman, R.W., Buenfeld, N.R.: Effect of entrained air voids on the microstructure and mass transport properties of concrete. Cem. Concr. Res. 41(10), 1067–1077 (2011). https://doi.org/10.1016/j.cemconres.2011.06.013 9. Das, A., Song, Y., Mantellato, S., Wangler, T., Lange, D.A., Flatt, R.J.: Effect of processing on the air void system of 3D printed concrete. Cem. Concr. Res. 156, 106789 (2022). https:// doi.org/10.1016/j.cemconres.2022.106789 10. Das, A.: 3D concrete printing: early-age strength build-up and long-term durability, PhD thesis, (submitted). Published online (2022) 11. C09 Committee. Test method for resistance of concrete to rapid freezing and thawing. ASTM Int. https://doi.org/10.1520/C0666_C0666M-15 12. Powers, T.C., Helmuth, R.A.: Theory of volume changes in hardened portland-cement paste during freezing. Highw. Res. Board Proc. 32 (1953). https://trid.trb.org/view/102368. Accessed 30 Apr 2021 13. Microstructural examination of carbonated 3D-printed concrete – Sanchez. J. Microsc. Wiley Onl. Lib. https://doi.org/10.1111/jmi.13087. Accessed 4 Apr 2022 14. Nambiar, E.K., Ramamurthy, K.: Sorption characteristics of foam concrete. Cem. Concr. Res. 37(9), 1341–1347 (2007) 15. Das, A.: Microstructure characterization of foamed cement and concrete (MS Thesis). Published online (2018)

Heterogeneities and Defects

Mechanical Properties and Failure Pattern of 3D Printed Hollow Cylinders and Wall Segments Under Uniaxial Loading Shantanu Bhattacherjee1 , Smrati Jain1 , Manu Santhanam1(B) , and G. Thiruvenkatamani2 1 Department of Civil Engineering, IIT Madras, Chennai, India

[email protected] 2 Tvasta Manufacturing Solution Pvt. Ltd., Chennai, India

Abstract. Extrusion-based 3D printed structures are heterogeneous with a combination of solid layers and weak bonds. The weak bonds can be considered to be an amalgamation of hydrated products and air voids. The hardened state properties of 3D printed structures depend on several factors such as layer strength, bond strength, geometrical imperfection, anisotropy, and printing parameters. The geometrical imperfections may be due to compression of individual layers or localized buckling during printing. This research aims to study the mechanical properties of 3D printed hollow cylinders and wall segments. The hollow cylinders correspond to hollow printed columns, whereas wall segments are cut from printed walls. The wall segment used in this study had a fixed design. The study is divided into three phases: hollow cylinders were printed with different aspect ratios (L/D), and compressive strength was measured at different ages in the first phase. The second phase included displacement control tests on 150 mm diameter and 300 mm height hollow printed cylinders. The post-peak behaviour was evaluated. The cylinder fails much later than the initiation of the first crack, but the crack propagates diagonally through the bonds and layers at ultimate failure. The effect of curing (water and air curing) on the compressive strength of hollow cylinders was further evaluated. In the last phase, 1 m by 1 m walls were printed, and segments were cut. The compressive strength was evaluated on the cut segments. This study shows that the initiation of crack is majorly influenced by geometrical irregularity and bond strength between the layers for the hollow cylinders. Whereas, for the wall segment, cracks initiated and propagated through the wall leaves and ribs connection. Keywords: Hardened properties · Hollow cylinder · Aspect ratio · Uniaxial elastic modulus · Wall segment

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 209–215, 2022. https://doi.org/10.1007/978-3-031-06116-5_31

210

S. Bhattacherjee et al.

1 Introduction The mechanical testing of 3D printed structures is a critical challenge for researchers considering the heterogeneity and complexity of the elements. A few researchers determined the mechanical behaviour of the printed structure by extracting prisms and performing flexural and compressive strength tests. Nerella et al. [1] extracted prisms of 120 mm × 25 mm × 25 mm from a printed wall, whereas Paul et al. [2] extracted 50 mm cubes and 140 mm × 40 mm × 40 mm prisms from a larger printed structure. Alchar and Al-Tamimi [3] printed 50 mm cubes with 20 mm layer thickness. Most of these printed elements were 2 to 4 layers thick and the study was primarily concerned about the difference in behaviour between cast and printed specimens. The anisotropy in mechanical properties is noted by a number of authors and the defects between layers are reported to be a critical parameter for strength determination. Le et al. [4] concluded that the presence of faults between two printed layers leads to stress concentration, significantly influencing the behaviour. Ding et al. [5] reported the cracks to propagate diagonally or along the layer interfaces for compressive load with different loading directions. All the studies were performed on cut specimens. The objective of this study is to understand the behaviour of printed hollow cylinders under compressive load, to observe the crack propagation, and to determine the postpeak behaviour of the printed cylinders. The effect of size and curing condition is a primary concern in the paper. Further, the criticality of the joint between ribs and wall for a large-scale wall element with ribs as infill is studied.

2 Experimental Methodology 2.1 Materials and Mix Design The printable mix studied consists of 53 grade ordinary Portland cement (OPC) (confirming to IS12269-1987), processed class F fly ash (conforming to ASTM C618), quartz sand with maximum aggregate size of 2 mm, polypropylene fibers (12 mm in length and 40 microns in thickness), PCE based superplasticizers (SP), cellulose-based viscosity modifying agent (VMA), and aluminium sulphate based accelerating admixture (Acc.). The mix proportion is shown in Table 1. A pan-type mixer was used to mix the ingredients at 30 rpm. The mix with SP and VMA was pumped using a screw-based pump to the nozzle head where the accelerator was mixed before extrusion. Table 1. Mix proportion Material

OPC Quartz sand Fly-ash Water SP

Proportion (to 0.8 binder content)

1.5

0.2

0.32

VMA

Fibers Acc.

.05% (solid 0.22% 0.2% content)

a Note: SP, VMA, Fibers and Acc. are presented in terms of % of cementitious material

1.5–2%

Mechanical Properties and Failure Pattern of 3D Printed Hollow Cylinders

211

2.2 Test Procedures for Mechanical Properties of Printable Mix The mechanical properties of the printable mix with and without accelerator were evaluated in terms of cylinder compressive strength and uniaxial elastic modulus. The cylinders of 100 mm diameter and 200 mm height were cast using the extruded mix from the printer. The test was performed using a closed loop servo-hydraulic testing machine with a capacity of 1000 kN load. The movement of the actuator in the machine was captured using an inbuilt LVDT and the piston deformation was also measured. The stiffness of the configuration used was 740 kN/mm and required correction was performed to determine the exact compression of the specimen. Another strain gauge was connected at the center of cylinder (longitudinal side) with gauge length of 100 mm to determine the elastic modulus of the material following the procedure in [6]. 2.3 Test Procedures for Mechanical Properties of Printed Structures The mechanical properties of two types of printed structures were evaluated (Fig. 1) – a. hollow cylinder and b. wall segment. The cylinders were capped with mortar (cement to sand ratio of 1:4) and moist cured for 24 h before testing. The wall element was capped with plaster of paris and air cured for 24 h before testing. The capping was done to make the top and bottom surface parallel.

Fig. 1. Experimental setup: a) hollow cylinder and b) wall segment

Effect of Height to Diameter Ratio and Curing on Printed Cylinder The compressive strength test was performed on the printed cylinders of two different diameters - 150 mm and 200 mm - with a height of 300 mm, as per IS 516-1959 (loading rate at 140 kg/cm2 /min). Further, the post peak behaviour was determined using the same closed loop servo-hydraulic testing machine mentioned in previous section, with a loading rate of 0.005 mm/s. Failure Analysis of Printed Wall Segments Two wall segments were obtained from a 1 m by 1 m printed wall. The compressive strength test was performed, on specimens with a length of 300 mm, height of 195 mm, and width of 195 mm. The crack pattern and maximum load are reported.

212

S. Bhattacherjee et al.

3 Results and Discussions Mechanical Properties of Printable Mix The mechanical properties for the mixes with and without accelerator are highlighted in Table 2. The 28 days compressive strength and uniaxial elastic modulus is determined to be increased with the addition of accelerator. Further, the accelerated mix was used for studies with the printed samples. Hence, the mix used for printing is having a maximum compressive strength of 44.4 MPa and uniaxial elastic modulus of 27.6 MPa. Table 2. Mechanical properties of the mixes at 28 days Sample

Accelerator

Curing days

Cylinder compressive strength (MPa)

Elastic modulus (GPa)

Printable mix

✗

28

36.3 ± 0.2

24.9 ± 1.3

✓

28

44.4 ± 0.5

27.6 ± 1.3

Mechanical Properties of Printed Hollow Cylinders Cylinders with diameters of 150 mm and 200 mm were printed (height of 300 mm) using the accelerated mix. The stress for the printed cylinder is based on the area of loading surface, i.e., the compressive stress is the ratio of compressive load to initial loading surface contact area. The cylinders are grouped into three sets based on the extent of curing and age of testing. In set 1, the cylinders were moist cured for 7 days and tested at 7th day from printing. It is found from the results in Table 3 that the hollow cylinder with higher diameter has lower strength. Higher deviations could be seen in the compressive strength values. In set 2, the cylinders are moist cured for 7 days and tested at 28th day from printing. A marginal increase of about 20% and 30% is observed for cylinders of 150 mm and 200 mm diameter, respectively, as compared to the 7-day values. In set 3, the cylinders were moist cured for 28 days and tested at 28th day. The compressive strength increased significantly by 45% and 80% against the 7th day strength of Set 1 cylinders. The curing period is therefore a very significant factor for the strength evolution in 3D printed concrete structures. Additionally, the effect of height to diameter is significant as the compressive strength reduced from 23.5 MPa to 19.4 MPa on increase of L/D for the hollow printed cylinders. To further understand the failure of printed cylinders, a post peak analysis is performed as shown in Fig. 2. It is observed that even after the initiation of cracks, a few layers of the structure resisted a certain amount of load. This implies that a sudden failure may not be observed for a printed column, and a few uncracked layers can provide resistance against the load. Further studies are required to understand the post-peak behaviour of the printed compressive elements.

Mechanical Properties and Failure Pattern of 3D Printed Hollow Cylinders

213

Table 3. Compressive strength of printed hollow cylinders with different curing ages Sample

Height to diameter

Diameter (mm)

Printed cylinder with layer width of 30 mm and thickness of 15 mm

2

150

Curing days

7 28

1.5

200

7 28

Compressive strength (MPa) 7 days

28 days

16.1 ± 3.4

19.3 ± 0.8

–

23.5 ± 3.3

10.3 ± 4.4

13 ± 4.3

–

19.4 ± 1.6

Fig. 2. Post-peak behaviour of the printed compressive elements

The crack propagation in the cylinders is further evaluated by visual method. It is seen that a diagonal crack is propagated as shown in Fig. 3. Also, for a few cylinders multiple staggered cracks are observed to form near the loading end. Localised failure of individual layers and crack initiation due to layer joints may govern the failure.

Fig. 3. Crack propagation in printed hollow cylinders

214

S. Bhattacherjee et al.

Failure of Printed Wall Elements Wall segments were cut out from a printed wall as shown in Fig. 4. The failure load is found to be 863 kN (works out to approximately 15 MPa considering the loading area). The crack is observed to propagate through the joint between the wall leaves and the infill ribs. The failure is governed by the geometry of the structure, with joints playing a major role in the overall structural integrity. The role of the joints in governing the strength may be more critical than the strength of the mix. A more thorough study is required to understand the effect of geometry on the overall strength of the printed structures.

Fig. 4. Crack propagation in printed wall element

4 Conclusions An increase in diameter for a fixed height (reducing height to diameter ratio) leads to a reduction in overall compressive strength of a hollow printed cylinder. Both size and shape govern the strength of the printed elements. The failures at joints and localized layer failure also affect the overall strength. Curing increases the strength significantly and a prolonged curing may be prescribed for 3D printed elements with OPC and VMA. Further, a few uncracked layers in the elements resist the loads even when a part of the structure is cracked. It can be concluded from the studies that creation of a representative volumetric element is critical for understanding the printed element behaviour. The strength of the mix or cubes with two or three layers might not represent the behaviour of the whole printed structure.

References 1. Nerella, V.N., Hempel, S., Mechtcherine, V.: Effects of layer-interface properties on mechanical performance of concrete elements produced by extrusion-based 3D-printing. Constr. Build. Mater. 205, 586–601 (2019) 2. Paul, S.C., Tay, Y.W.D., Panda, B., Tan, M.J.: Fresh and hardened properties of 3D printable cementitious materials for building and construction. Arch. Civil Mech. Eng. 18(1), 311–319 (2017)

Mechanical Properties and Failure Pattern of 3D Printed Hollow Cylinders

215

3. Alchaar, A.S., Al-Tamimi, A.K.: Mechanical properties of 3D printed concrete in hot temperatures. Constr. Build. Mater. 266, 120991 (2021) 4. Le, T.T., et al.: Hardened properties of high-performance printing concrete. Cem. Concr. Res. 42, 558–566 (2012) 5. Ding, T., Xiao, J., Zou, S., Wang, Y.: Hardened properties of layered 3D printed concrete with recycled sand. Cem. Concr. Compos. 113, 103724 (2020). https://doi.org/10.1016/j.cemcon comp.2020.103724 6. Stephen, S.J., Júnior, E.Z., Gettu, R., Aguado, A., Vaishnav Kumar, S.: Determination of the complete stress-strain response of concrete under uniaxial compression. Ind. Concr. J. 1–25 (2021)

Impact of Drying of 3D Printed Cementitious Pastes on Their Degree of Hydration Rita M. Ghantous1(B) , Yvette Valadez-Carranza1 , Steven R. Reese2 , and W. Jason Weiss1 1 School of Civil and Construction Engineering, Oregon State University, Corvallis,

OR 97331, USA {ritamaria.ghantous,valadezy,Jason.Weiss}@oregonstate.edu 2 School of Nuclear Science and Engineering, Oregon State University, Corvallis, OR 97331, USA [email protected]

Abstract. The 3D printing of cementitious materials has been proposed as a method to reduce costs and waste associated with formwork. However, the 3D printing of concrete faces some challenges. One specific obstacle is related to the curing and shrinkage of the printed cementitious materials. These elements often contain a high paste content and poor curing conditions. Information is needed about the drying behavior of cementitious elements with very high surface to volume ratio that are exposed to drying nearly immediately upon placement. This study uses neutron radiography to evaluate the impact of the drying of printed cement paste samples on their degree of hydration. Exposing 3D printed samples to the atmosphere, (i.e., drying) severely limited the hydration. The impact of drying was highly dependent on the surface to volume ratio of the element. The degree of hydration of a material is linearly correlated to the square root of its volume to surface ratio. Keywords: 3D printed cementitious pastes · Drying behavior · Degree of hydration

1 Introduction The drying shrinkage of printed cementitious materials is a challenge facing 3D printing. The loss of water in fresh concrete can limit hydration and can lead to shrinkage that can result in crack development [1–7]. It is also known that drying shrinkage is heavily dependent on the surface to volume ratio of the element (S/V) [8]. 3D printed elements with 6 mm diameter have a S/V 200 times higher than the S/V of a conventional 300 mm thick concrete wall drying from one side. This high S/V of 3D printed elements indicates their sensitivity to drying and that there may be a need to re-evaluate their curing procedure. Researchers are working on addressing this point. Van Der Putten et al. [7] examined the usage of super absorbent polymers in 3D printed samples to reduce this shrinkage. Slavcheva [1] concluded that the addition of microsilica and superplasticizers © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 216–221, 2022. https://doi.org/10.1007/978-3-031-06116-5_32

Impact of Drying of 3D Printed Cementitious Pastes

217

will be beneficial in reducing the drying shrinkage of cement paste samples under different drying environment. However, information is still needed on the drying behavior of 3D printed cement paste elements exposed to drying nearly immediately after printing. This paper investigates the impact of drying on the degree of hydration of the printed materials. Second, the dependency between S/V of the element exposed to drying and their degree of hydration was studied.

2 Experimental Procedure 2.1 Preparation and Geometry of Cement Paste Samples Cement paste samples with a water to binder ratio of 0.26 were prepared for this study using type I ordinary Portland cement conforming to ASTM C150. A high range water reducing admixture (HRWR, MasterGlenium 7925) and a viscosity modifying admixture (VMA, MasterMatrix 450) were incorporated in the mixture design of the cement paste with a ratio of 0.5% and 1.2% by mass of cement [9]. The cement paste was mixed following a modified mixing procedure of ASTM C305-14. The VMA and HRWR were first mixed with the mixing water. Cement was then added and all the components were mixed for 90 s at 400 revolutions per minute in a twister evolution Venturi vacuum mixer at 80% vacuum level. The mixer was stopped for 15 s to scrap down into the mixing bowl any paste that may have collected on the sides. The cement paste was then vacuum mixed for 90 s at 400 revolutions per minute. For the 3D printed samples, after mixing, the cement paste was filled and consolidated in a 60 ml syringe. The syringe was mounted on a paste extruder that was connected to an Ultimaker© 2 + 3D printer using Polyethylene tubing. A stainless steel nozzle with an internal diameter (Di ) of 2.39 mm was used in this study (Fig. 1-a). The printing of the cement paste started 15 min after the beginning of the cement hydration. The printing speed was 540 mm/min. The nominal width of each printed filament was 2.87 mm, which is larger than Di due to the swelling of the extruded paste. The printing table was automated and programed to move with an accuracy of 12.5, 12.5 and 5 microns in the x, y and z directions, respectively. In addition, the amount of material was also programed to extrude at a rate that was relative to the rate at which the nozzle was moving [9]. The size of the prepared samples was 93 mm in length, 15 mm in height and 9 mm in depth (i.e., three filaments). One of the printed samples was exposed to drying environment (25 ± 1 °C temperature, and 50 ± 2% relative humidity) directly after printing in order to study the impact of drying on the hydration evolution of cement. This sample is labeled as sample 2 (Fig. 1-a). Another sample, labeled sample 1 in Fig. 1-a, was sealed directly after printing and was used as a reference for the degree of hydration value when the samples are not exposed to drying. Immediately after printing, the samples were transferred to an environmental chamber with a temperature of 25 ± 1 °C and a relative humidity (RH) of 50 ± 2% for a duration of 7 days. The degree of hydration of these samples was determined after the 7 days duration using neutron radiography as detailed later in this paper. The dependency of the extent of the hydration reaction on the S/V of a sample was studied by preparing samples with varying thicknesses. Figure 1-b illustrates the

218

R. M. Ghantous et al.

geometrical dimensions of the different cement paste samples prepared in this study. Four different thicknesses (t) were investigated: 4, 8, 16, 32 mm corresponding to cement paste samples with four different S/V: 500, 250, 125, 63 m2 /m3 respectively. Right after mixing, the cement paste was cast in the various molds and was consolidated to remove the entrapped air. After consolidation, the vertical sides of the molds were removed and the samples were exposed to 1D drying at 25 ± 1 °C temperature and 50 ± 2% RH for 8 days duration. At the end of this duration, the degree of hydration of these samples was determined using thermogravimetric analysis as detailed later in this paper.

(a)

(b)

Fig. 1. (a) A photo of the 3D printed samples preparation- sample 1: sealed 3D printed sample; sample 2: 3D printed sample exposed to drying; (b) Geometry of the cement paste samples with varying S/V obtained by varying the paste thicknesses (t): 4, 8, 16, 32 mm.

2.2 Degree of Hydration Measurements Using Neutron Radiography Neutron radiography was used to determining degree of hydration [10]. The approach consists on determining the volume of water in the hydration products after the free water is lost. In order to do so, at the end of the 7 days duration, all the free water remaining in the 3D printed samples was removed by exposing the samples to 105 °C until reaching less than 0.01% variation in mass within 24 h. The neutron radiography facility of the Oregon State TRIGA® Reactor (OSTR) was used for this study. A radiograph of the oven-dried sample was then captured using a D610 Nikon camera with a 50 mm f/1.2 lens. The radiographs resolution was 90 microns. The images were used to capture the transmitted neutrons intensities on a cesium iodide scintillation detector (CsI crystals with 5 µm in diameter doped with Gd). The radiographs of the oven-dried samples can be used to quantify the volume of non-evaporable water, i.e. water in hydration products, using Beer-Lambert law rewritten in Eq. (1).         IOD =− Vc + VHP + VGW + VCW + VCS TS ln c HP GW CW CS I0 (1) where, IODis the transmitted intensity through the oven dry sample, I0 is the beam intensity, c is the macroscopic cross section of the unhydrated cement (equal to

Impact of Drying of 3D Printed Cementitious Pastes

219

    0.01867 mm−1 for the cement used in this study), HP, GW, CW, CS are the macroscopic cross sections of the hydration products, gel water, capillary water and chemical shrinkage respectively, V is the volume fraction of cement (equal to 0.53 for the mixture design used in this study), hydration products (VHP ), gel water (VGW ), capillary water (VCW ), chemical shrinkage pores (VCS ), Ts is the sample thickness. During drying period, free water, i.e. capillary water and gel water, evaporates from the cement paste sample. Consequently, VGW , VCW are equal to zero in an oven-dried sample. In addition, the chemical shrinkage pores in an oven dried sample are empty (i.e. VCS = 0 in an oven dried cement paste sample). The volume of cement does not change in the sample. As a result, part of the hydration products can be considered to remain as the volume of the original cement [10]. The remaining volume fraction of the hydration products corresponds to the water that has reacted with the cement. Therefore, Eq. (2) can be used to determine the volume of non-evaporable water (V ) [10]. More details on this procedure can be found in [10]. ⎛ I  ⎞ OD  1 ⎝ ln I0 − Vc ⎠ (2) V = − c TS w  where, w is the macroscopic cross section of water (0.1208 mm−1 ). The degree of hydration can be then calculated using Eq. (3). DOH = C × V

(3)

where, DOH is the degree of hydration, C is found to be equal to 783 according to [10]. It should be noted that C changes with the mixture design. 2.3 Degree of Hydration Measurements Using Thermogravimetric Analysis The samples that were cast with varying S/V were tested in the thermogravimetric analysis (TGA, Q50, TA instrument) in order to determine their degree of hydration according to the procedure defined in [11]. A sample was taken from the middle section of each casted cement paste specimen after exposure to 8 days of drying at 25 ± 1 °C temperature and 50 ± 2% RH. It was then exposed to drying at 105 °C until reaching equilibrium. This sample was then ground and sieved through a 75 microns pore size sieve and tested in the TGA instrument according to the procedure detailed in [11]. The degree of hydration of each cement paste specimen was then determined using Eq. (4) [12, 13]. DOH (%) =

m1 −m2 ×(1+LOIc ) m2 ×(1−LOIc /0.23)

0.23

× 100

(4)

where, m1 and m2 are the sample weight recorded by the TGA at 105 °C and 1000 °C respectively. LOIc is the loss on ignition of the cement (measured to be equal to 1.3%). The degree of hydration of each specimen was determined by averaging the results obtained from four repeats in the TGA.

220

R. M. Ghantous et al.

3 Results and Discussion Figure 2-a illustrates the degree of hydration profiles for both sealed and unsealed 3D printed samples. It should be noted that degree of hydration profiles in this figure are an average across the cross section and variations due to the depth from the drying surface are not considered. It can be noted that the degree of hydration of samples exposed to drying ceased at an average value of 24 ± 4%, while sealed samples reached a degree of hydration of 64 ± 4% after a curing duration of 7 days. The water evaporation from the printed sample that was exposed to drying limited the extent of the hydration reaction. Consequently, curing of printed elements with a high S/V needs to be considered carefully in order to avoid drying and hydration limitation, which will then reduce the structure durability. Figure 2-b illustrates that the degree of hydration of a sample is linearly proportional to the square root of its volume to surface ratio. The slope of this linear fitting is equal to 319 (%DOH.m−1/2 ). The samples with the highest thickness (i.e., lowest S/V) showed the highest degree of hydration. Samples with 32 mm thickness showed a degree of hydration of 42 ± 4%. While the degree of hydration of samples with 16, 8, and 4 mm thickness were 29 ± 1, 19 ± 2, 13 ± 1% respectively. The severity of drying increased with the increase in the S/V leading to more limitation on the final degree of hydration.

20

40

60

80

100

Depth (mm)

0 3 6 9 12

Degree of hydration (%)

100

Degree of hydration (%) 0

80

Degree of hydration of sealed cured samples

60 40

32 mm 20 0 0.00

exposed to drying

(a)

sealed cured

4 mm

16 mm 8 mm

0.04

0.08

0.12

(Volume to surface ratio) 1/2 (m)1/2

(b)

Fig. 2. (a) Degree of hydration profiles in both sealed and unsealed 3D printed samples, (b) Correlation between the degree of hydration and the square root of the volume to surface ratio of the cement paste samples.

4 Conclusion 3D printed elements have a relatively high surface to volume ratio as compared to conventional structures. 3D printed elements are in general exposed to drying directly after printing due to the absence of a formwork. The direct exposure to drying leads to loss of water from the matrix more rapidly, which results in limitations on hydration. It was observed in this study that drying and the resulting degree of hydration are directly related to the S/V of the element. Specific curing procedures need to be studied to reduce the impact of the drying on the degree of hydration evolution in the printed elements.

Impact of Drying of 3D Printed Cementitious Pastes

221

References 1. Slavcheva, G.S.: Drying and shrinkage of cement paste for 3D printable concrete. IOP Conf. Ser. Mater. Sci. Eng. 481(1), 012043–012052 (2019) 2. Holt, E.: Contribution of mixture design to chemical and autogenous shrinkage of concrete at early ages. Cem. Concr. Res. 35(3), 464–472 (2005) 3. Siddika, A., Mamun, M.A.A., Ferdous, W., Saha, A.K., Alyousef, R.: 3D-printed concrete: applications, performance, and challenges. J. Sustain. Cement-Based Mater. 9(3), 127–164 (2020) 4. Radli´nska, A., Kaszy´nska, M., Zieli´nski, A., Ye, H.: Early-age cracking of self-consolidating concrete with lightweight and normal aggregates. J. Mater. Civ. Eng. 30(10), 04018242 (2018) 5. Hoffmann, M., Skibicki, S., Pankratow, P., Zieli´nski, A., Pajor, M., Techman, M.: Automation in the construction of a 3D-printed concrete wall with the use of a Lintel gripper. Materials (Basel) 13(8), 1800 (2020) 6. Van Der Putten, J., et al.: Neutron radiography to study the water ingress via the interlayer of 3D printed cementitious materials for continuous layering. Constr. Build. Mater. 258, 119587 (2020) 7. Van Der Putten, J., Deprez, M., Cnudde, V., De Schutter, G., Van Tittelboom, K.: Microstructural characterization of 3D printed cementitious materials. Materials (Basel, Switzerland) 12(18), 2993 (2019) 8. ACI Committee 209R-92-Creep and Volume Changes in Concrete: Prediction of creep, shrinkage, and temperature effects in concrete structures. American Concrete Institute, vol. 47, (2009) 9. Moini, M., Olek, J., Youngblood, J.P., Magee, B., Zavattieri, P.D.: Additive manufacturing and performance of architectured cement-based materials. Adv. Mater. 30(43), 1802123 (2018) 10. Moradllo, M.K., Montanari, L., Suraneni, P., Reese, S.R., Weiss, J.: Examining curing efficiency using neutron radiography. Transp. Res. Rec. 2672(27), 13–23 (2018) 11. Kim, T., Olek, J.: Effects of sample preparation and interpretation of thermogravimetric curves on calcium hydroxide in hydrated pastes and mortars. Transp. Res. Rec. 2290(1), 10–18 (2012) 12. Fagerlund, G.: Chemically bound water as measure of degree of hydration: method and potential errors. Division of Building Materials, Lund Institute of Technology Lund, Sweden (2009) 13. Copeland, L.E., Kantro, D.L.,Verbeck, G.J.: Chemistry of hydration of Portland cement. Portland Cement Association, Research and Development Laboratories (1960)

The Environment’s Effect on the Interlayer Bond Strength of 3D Printed Concrete Gerrit M. Moelich, J. J. Janse van Rensburg, Jacques Kruger, and Riaan Combrinck(B) Unit for Construction Materials, Department of Civil Engineering, Stellenbosch University, Stellenbosch, South Africa [email protected]

Abstract. Estimating the bond strength between extruded concrete filaments is of paramount importance for the structural design of 3D printed concrete elements. This study investigates the effect of the environmental printing conditions at various pass times on the interlayer bond strength (IBS), elaborates on the underlying mechanism and investigates a possible mitigation measure. Elements are printed with seven different pass times ranging from 1 min to 30 min while exposed to a benign (indoor-like) and severe (site-like) evaporation rates. The IBS is quantified through flexure tests, with the interlayer aligned vertically, at a concrete age of 28 days. The experimental results show that increasing the pass time from 1 min to 30 min decreases the IBS by 107% for the benign condition. When the specimens were exposed to the severe evaporation rate for 30 min the IBS decreased an additional 35% . Applying a curing compound to the exposed filament surface, to reduce the moisture evaporation, weakened the bond strength. Keywords: Concrete · 3D concrete printing · Mechanical strength · Interlayer bond strength · Hot weather concreting

1 Introduction The benefits of using concrete/mortar in additive manufacturing are the material’s mechanical strength, long-term durability, and scalability, in terms of raw material availability as well as element size. These attributes render 3D concrete printing (3DCP) suitable for creating structural elements for large-scale construction. Compared to the conventional construction method, 3DCP eliminates the need for formwork to reduce waste materials and manual labour while increasing design freedom and workplace safety [1]. One impediment to the structural integrity of printed concrete is the interlayer bond strength (IBS) between sequentially deposited filaments. The layer-by-layer deposition technique leaves the printed element significantly weaker in bending compared to the cast equivalent, especially in harsh on-site environments and extended pass times (time between sequential filament depositions) [2]. In a review paper, Kruger and Van Zijl [3] show an average maximum reduction in strength of 46% when the concrete is printed, and this reduction can range from approximately 10% to 90% depending on various © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 R. Buswell et al. (Eds.): DC 2022, RILEM Bookseries 37, pp. 222–227, 2022. https://doi.org/10.1007/978-3-031-06116-5_33

The Environment’s Effect on the Interlayer Bond Strength

223

parameters. This lack of IBS is concerning since the low w/b ratio, typically used in 3DCP mixes, is not fully utilised for structural performance and, secondly, the structural design becomes complex since many parameters can significantly affect the IBS, and consequently the degree of mechanical anisotropy. Tensile reinforcement would partly address this issue, but few practical reinforcement strategies are currently available. The mechanism responsible for the lack of IBS can be broadly categorised as mechanical, physical, and chemical effects [3, 4]. The mechanical effect includes the interlayer interlocking/tortuosity/roughness which is controlled by the mix’s rheological properties, printing parameters and the mix proportions and constituents [4, 5]. It is further a function of the pass time since the rheology properties of the substrate layer change as the material’s particles re-flocculation and structuration, and this affects the intermixing of filaments. The physical effect involves the air entrapment between consecutively placed filaments and depends on the material’s rheology as well as the printing parameters [6, 7]. The chemical effects include the bond produced by the hydration reaction of the binders. If surface moisture is lost to the environment, through evaporation, the chemical bond becomes pass time-dependent since there is less moisture available to hydrate the cementitious particles [5, 8]. Significant surface moisture evaporation also has a physical effect since the porosity of the interlayer zone increases [7, 9], and a mechanical effect since less pore water is present to facilitate the intermixing/interlocking between filaments during deposition. Significant pore water evaporation can also induce rapid plastic shrinkage [10], to potentially reduce the bond strength [6]. Literature agrees that surface moisture evaporation from the substrate layer is one of the dominant factors controlling the lack of IBS [3, 8, 9, 11]. A model to estimate the IBS of printed concrete has been proposed with pass time and the environment’s evaporation rate as main input parameters [2]. The model could accurately estimate the IBS reduction from 30% to 50%, but investigated only a relatively short pass time and low evaporation rate. This study investigates the effect of a long pass time and more extreme climate on the long-term IBS of 3D printed concrete and builds on reference [2]. The aim is to provide additional evidence, by using a different mix, w/c and test setup, and extend the limits for the proposed model.

2 Experimental Program Portland cement (PPC CEM II 52.5N) with 6 to 20% limestone replacement was used. Fly ash (DuraPozz Class F) and silica fume (SiliconSmelters Microfume) were added additionally as binders. Two natural, continuously graded, quarry sands were mixed to approximately fit the Fuller Thompson curve. A modified polycarboxylate polymer superplasticizer (Chryso Fluid Premia 310) was selected and added to obtain a minislump-flow result of 145 mm. One mix was used throughout the study and the proportions were as follow: CEM II 450 kg/m3 , fly ash 280 kg/m3 , silica fume 85 kg/m3 , fine aggregate (