Proceedings of the 6th International Conference on Civil Engineering, ICOCE 2022, Singapore: Innovations in Civil Engineering 9811939829, 9789811939822

Citation preview

Lecture Notes in Civil Engineering

Eric Strauss   Editor

Proceedings of the 6th International Conference on Civil Engineering, ICOCE 2022, Singapore Innovations in Civil Engineering

Lecture Notes in Civil Engineering Volume 276

Series Editors Marco di Prisco, Politecnico di Milano, Milano, Italy Sheng-Hong Chen, School of Water Resources and Hydropower Engineering, Wuhan University, Wuhan, China Ioannis Vayas, Institute of Steel Structures, National Technical University of Athens, Athens, Greece Sanjay Kumar Shukla, School of Engineering, Edith Cowan University, Joondalup, WA, Australia Anuj Sharma, Iowa State University, Ames, IA, USA Nagesh Kumar, Department of Civil Engineering, Indian Institute of Science Bangalore, Bengaluru, Karnataka, India Chien Ming Wang, School of Civil Engineering, The University of Queensland, Brisbane, QLD, Australia

Lecture Notes in Civil Engineering (LNCE) publishes the latest developments in Civil Engineering - quickly, informally and in top quality. Though original research reported in proceedings and post-proceedings represents the core of LNCE, edited volumes of exceptionally high quality and interest may also be considered for publication. Volumes published in LNCE embrace all aspects and subfields of, as well as new challenges in, Civil Engineering. Topics in the series include: • • • • • • • • • • • • • • •

Construction and Structural Mechanics Building Materials Concrete, Steel and Timber Structures Geotechnical Engineering Earthquake Engineering Coastal Engineering Ocean and Offshore Engineering; Ships and Floating Structures Hydraulics, Hydrology and Water Resources Engineering Environmental Engineering and Sustainability Structural Health and Monitoring Surveying and Geographical Information Systems Indoor Environments Transportation and Traffic Risk Analysis Safety and Security

To submit a proposal or request further information, please contact the appropriate Springer Editor: - Pierpaolo Riva at [email protected] (Europe and Americas); - Swati Meherishi at [email protected] (Asia - except China, and Australia, New Zealand); - Wayne Hu at [email protected] (China). All books in the series now indexed by Scopus and EI Compendex database!

More information about this series at https://link.springer.com/bookseries/15087

Eric Strauss Editor

Proceedings of the 6th International Conference on Civil Engineering, ICOCE 2022, Singapore Innovations in Civil Engineering

123

Editor Eric Strauss Michigan State University Dimondale, MI, USA

ISSN 2366-2557 ISSN 2366-2565 (electronic) Lecture Notes in Civil Engineering ISBN 978-981-19-3982-2 ISBN 978-981-19-3983-9 (eBook) https://doi.org/10.1007/978-981-19-3983-9 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. 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 Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Preface

2022 6th International Conference on Civil Engineering (ICOCE 2022) was supposed to be held in Singapore during March 27–29, 2022. However, given the unstable situation of the epidemic, the conference was held as a fully virtual one in Zoom. We scheduled the equipment test for each presenter in advance to ensure successful delivery and avoid any unwanted technical glitches. For those who had Internet issues, a pre-recorded video was made as an alternative to a live presentation. The entire conference was archived for future references and streaming purpose. This conference is intended as a vehicle for the dissemination of research results on the latest advances made in the area of civil engineering, which gave an excellent opportunity for the scientists and researchers around the world to have a common platform to exchange their findings and to discuss the developments and make a base for starting collaborations in the national and international levels. It was also aimed at promoting contacts among researchers and research groups for the creation of multinational thematic and research networks, as well as promoting contacts for future collaborative joint projects within some of the international funding programs. The technical program committee put together a program, which included presentations from four keynote speakers: Prof. Benjamin Sovacool (University of Sussex, UK), Prof. Prashant Kumar (University of Surrey, UK), Prof. Pen-Chi Chiang (National Taiwan University, Taiwan), and Dr. Christopher H. T. Lee (Nanyang Technological University, Singapore) and 28 papers that were accepted for publication in the ICOCE’22 conference proceedings after a rigorous review, which covers the following four specific areas: concrete structure design and analysis, structural mechanics and structural engineering, geological exploration and earthquake engineering, and urban planning and environmental assessment. On behalf of the organizing committee, we extend our heartfelt gratitude to the keynote speakers and the authors of the papers selected for presentation and publication. We also extend our thanks to the international reviewers and the members of the program committee for their contribution and commitment to putting together

v

vi

Preface

a program of presentations that conference participants should find beneficial in their professional endeavor. ICOCE 2022 Conference Organizing Committees

Organization

Committees Conference Chairs Zongjin Li Shane Snyder

University of Macau, China Nanyang Technological University, Singapore

Conference Co-chairs Pen-Chi Chiang Prashant Kumar

National Taiwan University, Taiwan University of Surrey, UK

Program Chairs Roger Ruan Benjamin K. Sovacool

University of Minnesota, USA University of Sussex, UK

Publication Chair Eric Strauss

Michigan State University, USA

Technical Program Committees Akmal Abdelfatah Mubarak Al Alawi Godwin Akpeimeh J. Amudhavel Irina Benedyk Klaas van Breugel Ta-Peng Chang

American University of Sharjah, UAE Sultan Qaboos University, Oman University of Leeds, UK Bhopal–Indore Highway, India University at Buffalo, USA Delft University of Technology, The Netherlands Civil and Construction Engineering, Taiwan Tech, Taiwan

vii

viii

Fadi Hage Chehade Chow Ming Fai Yasmin Fathy Hassan Hemida Yuner Huang Farhad Jazaei Fei Jin Hue Thi Nguyen Kedsarin Pimraksa Zhao Qin Yongmin Kim Siti Fatin Mohd. Razali Korb Srinavin Sudharshan N. Raman Kong Fah Tee Linh Truong-hong Chien Ming Wang Yuandong Wang Yan Xiao Jinhui Yan Mijia Yang Xiong Yu

Organization

Lebanese University, Beirut, Lebanon Monash University Malaysia, Malaysia University of Cambridge, UK University of Birmingham, UK The University of Edinburgh, UK The University of Memphis, USA Cardiff University, Wales, UK University of Wisconsin-Madison, Vietnam Chiang Mai University, Thailand Syracuse University, USA University of Glasgow Singapore, Singapore Universiti Kebangsaan Malaysia (UKM), Malaysia Khon Kaen University, Thailand Monash University Malaysia, Malaysia University of Greenwich, UK Delft University of Technology, The Netherlands The University of Queensland, Australia University of Utah, USA Dalian University of Technology, China University of Illinois, USA North Dakota State University, USA Case Western Reserve University, Cleveland, USA

Contents

Concrete Structure Design and Analysis Linear Time Invariant Property of Modelling the Water Absorption Process within Cement-Based Materials . . . . . . . . . . . . . . . . . . . . . . . . . Jinliang Liu, Li Song, Chenxing Cui, and Ran Liu

3

Study of Failure Patterns in a Reinforced Concrete Beam Subjected to a Varying Range of Blast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anita Bhatt and Sajad Ahmed Bhat

15

Corroded Stirrups Effects on the Shear Behavior of Reinforced Concrete Slender Beams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pier Paolo Rossi and Nino Spinella

30

Response Control on Seismic Retrofit of Low-Rise RC Frame Using Viscous Damper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Panumas Saingam

38

Strength and Durability Properties Empty Fruit Bunch Ash (EFBA) as Partial Replacement of Cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gunalaan Vasudevan, Kribadharan Uthayasirppi, and Mohd Mawardi bin Hassim Experimental Study on the Performance of FRP Grid Reinforced ECC Composite Layer-Concrete Interface . . . . . . . . . . . . . . . . . . . . . . . Weiwen Li, Jie Liu, Jiahao Zhang, Shiying Tang, Meizhong Wu, and Xu Yang

49

58

Structural Mechanics and Structural Engineering Dynamic Strain Estimations of Beam Ends in Steel Moment-Resisting Frames Using Acceleration Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Xiyang Yu and Xiaohua Li

73

ix

x

Contents

Studies on the Relationship Between Anchor Force of Prestressed Anchor Cable and Nonlinear Vibration of Anchor Head . . . . . . . . . . . . Hao Li and Hui Cao

87

Interrelationships of Load and Displacement of Barrette Piles for Various Interpretation Criteria Subjected to Uplift Loading . . . . . . . . . Yit-Jin Chen, Anjerick Topacio, and Suneelkumar Laveti

98

Research on the Influence of Water Horse on the Vortex Induced Resonance Response of Bridges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 Bifeng Liu and Changzhao Qian The Degradation of Avalanche Anchorage Systems . . . . . . . . . . . . . . . . 120 C. Paglia and C. Mosca Modeling and Analysis of Shanghai Central Tower . . . . . . . . . . . . . . . . 128 Yuchi Liu, Yiwen Chen, and Peitong Li From Digitized Systemization to New Era of Autonomous Materials Facilitation in Architectural Design and Actuation . . . . . . . . . . . . . . . . . 138 Mohammed Abbood and Ghada Al Slik Geological Exploration and Earthquake Engineering Analytical Investigation of the Influence of Seismicity on the Construction Cost of the R/C Load-Bearing Structure of a 15-Storey Building . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Georgios Papadimitriou and Theodoros Chrysanidis The Influence of Buried Depth on the Stress Mode of Independent Foundation in Rock Foundation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 Yunpeng Dong and Yin Ke Study on the Strength Characteristics of Metamorphic Slate Based on Point Load Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Wei Xingcan, Qi Shaoli, Yang Baowen, Wang Wenpo, and Zhong Zhibin Soıl Structure Interactıon Effects on the Seısmıc Response of Rc Structure wıth Rıgıd and Flexıble Foundatıons . . . . . . . . . . . . . . . 190 Gokaran Prasad Awadhya, Aayush Jha, Mrinal Thakur, and Sanidhya Sharma Study on Treatment of Ground Subsidence in Goaf . . . . . . . . . . . . . . . . 212 Hong Qiang Zhang, Ling Gao, and Ming Yue Feng Forward Modeling and Detection of the Potential Slip in Artificial Slope by GPR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 Zhenjun Zhang, Luo Ye, and Gao Lv

Contents

xi

Application of High Density Electrical Method in Karst Area . . . . . . . . 229 Hong Qiang Zhang, Zheng Bo Cao, and Wei Li Green Building Construction Implementation Barriers in the Philippines a Hierarchical Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234 Cris Edward Monjardin, Lovely Jasmin Dela Cruz, Ezekiel Esguerra, and Ondrea Nhika Dangzalan Urban Planning and Environmental Assessment Wind-Induced Dynamic Response Analysis of the Frame Lightning Rod . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245 Ran Liu, Li Song, Chenxing Cui, and Jinliang Liu Research on Green Renewal Strategy Based on LID Adaptation Toolkit Selection–The Example of the North Bay Section of Macau Inner Harbour Terminal Neighborhood . . . . . . . . . . . . . . . . . . . . . . . . . 258 Linsheng Huang The Impact of Urbanisation on Catchment Discharge in Malaysia: A Case Study on Sungai Selangor Catchment . . . . . . . . . . . . . . . . . . . . 268 Mayuran Jayatharan, Andreas Aditya Hermawan, Amin Talei, and Izni Zahidi Can an Environmentally Sustainable Construction be Affordable? . . . . 281 Paulo Mendonça and Clara Vieira Climate Change and Sustainable Campus Planning: A Review of Michigan Universities’ Climate-Related Plans . . . . . . . . . . 289 AyseOzcan Buckley and Eric J. Strauss Synthesis of a Novel pH-Responsive Emulsifier Based on Dynamic Covalent Bond and Its Application in Reversible Oil-Based Drilling Fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313 Guoshuai Wang, Guancheng Jiang, Jun Yang, Yinbo He, and Yue Fu Removal of Aqueous Cu2+ by NaCl Modified 5A Zeolite . . . . . . . . . . . . 324 Jiantao Yang and Jianwen Wei Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331

Concrete Structure Design and Analysis

Linear Time Invariant Property of Modelling the Water Absorption Process within Cement-Based Materials Jinliang Liu1(B) , Li Song1,2 , Chenxing Cui1 , and Ran Liu1 1 School of Civil Engineering, Central South University, Shaoshan South Road 68,

Changsha 410082, China [email protected] 2 National Engineering Laboratory for High Speed Railway Construction, Changsha 410082, China

Abstract. As a transmission medium of harmful ions, water has an important impact on the durability of concrete. The purpose of this study is to investigate the characteristics of water transfer in sound and cracked cement-based materials. Water absorption tests are performed considering initial saturation, damage crack, and water-cement ratio. Numerical models based on assumption of diffusion mechanism are proposed and solved by finite element method, and different constitutive models are employed and verified. The results show that water absorptivity is intensely affected by initial saturation and damage cracks. Besides, on the basis of diffusion equation, the linear relationship between cumulative water absorption and square root of time is found to be rigidly satisfied whatever the formula of diffusion coefficient. Keywords: Water absorption · Cement-based material · Modelling · Diffusion coefficient · Long term

1 Introduction As is known that ingression of water and harmful ions are two crucial issues that cause durability problem of cement-based materials. Freeze-thaw cycle damage caused by water will degrade the concrete strength, resulting in cracks and spalling [1–3]. On the other hand, water acts as the transport medium of aggressive ions such as chloride and sulfate ions that transfer through mechanism of diffusion or convection [4, 5]. Water itself and issues caused by it, account for most degradation problems concerning both durability and function. For concrete surface of nearly 0–20 mm, water capillary absorption in each wettingdrying cycle is regarded as the main mechanism of chloride ion transferring into a porous concrete because diffusion flux is quite small compared with capillary absorption. Hence, examining the capillary behavior in concrete is important to predict ion transfer coupling wetting-drying process. Weighing method is commonly employed to estimate water absorption characteristic of materials, and the water absorptivity can be deduced © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 3–14, 2023. https://doi.org/10.1007/978-981-19-3983-9_1

4

J. Liu et al.

through water capillary test of a dried specimen [6–8]. However, the disadvantage of water absorption test is that we cannot obtain the time-varying water content distribution and wetting front within the specimen. To predict the concentration profile of chloride ion, water distribution in the process of capillary absorption is necessary. Methods of nuclear magnetic resonance, gamma rays and neutron radiography have been used to monitor the process of water transfer in real time [9–11]. The geometric quantitative information of pores such as pore size distribution was acquired from mercury intrusion porosimetry (micropore of smaller than 10 µm). In addition, numerical model based on diffusion mechanism is applied to simulate the water content profile with an assumed hydraulic diffusion coefficient [12, 13]. Mesoscale models considering aggregate, cement paste and interface zone were proposed to get a more detailed comprehension instead of homogeneous medium assumption [14, 15]. Cracks and damage caused by loading provide fast paths for water absorption, and the influence on water absorptivity has been examined by tests [16–18]. In this paper, theoretical models based on continuum medium are reviewed including governing equation, initial and boundary conditions, and the saturation of concrete is chosen as unknown field variable. Then water absorption tests of dried mortar specimens, as well as cement paste cracked by temperature of 105 °C, were performed by weighing method. On the basis of test results, the values of hydraulic diffusion coefficients were estimated, and then employed in the numerical model to simulate saturation distribution within specimens. Finally, the disadvantages of known diffusion coefficient models, and of numerical model applied in long term absorption, were discussed, as well as the influence of initial saturation.

2 Models 2.1 Governing Equation As an extension of Darcy’s law, the unsaturated flow theory can be formulated as: u = −k(S)∇ψ

(1)

where u is the flow velocity (m/s); k(S) is the unsaturated permeability coefficient (m/s); S is the saturation of media (0–1); ψ is the height of water head (m). According to the law of mass conservation, that water flowing into media per unit time is equal to the mass increase, we can obtain ∂S + ∇ · u = 0. ∂t

(2)

Substituting Eq. (1) into Eq. (2), we can deduce [19] ∂S + ∇ · (−k(S)∇ψ) = 0. ∂t

(3)

In terms of the chain rule, Eq. (3) can be transformed into dψ ∂S + ∇ · (−k(S) ∇S) = 0 ∂t dS

(4)

Linear Time Invariant Property of Modelling the Water Absorption Process

5

from which we can extract the definition of hydraulic diffusion coefficient in the diffusion system as: D(S) = k(S)

dψ . dS

(5)

Finally, the governing equation of water transferring in concrete can be inferred as: ∂S + ∇ · (−D(S)∇S) = 0. ∂t

(6)

2.2 Hydraulic Diffusion Coefficient Known studies show that, in the initial time (roughly 20 h), water absorption mass is proportional to the square root of time, that is i = A · t 0.5

(7)

where i is the equivalent height of water absorption (m); A is the water absorptivity (m/s0.5 ); t is the time (s). Through experimental results, we can obtain i=

Δm aρw

(8)

where Δm is the cumulative absorption mass (kg); a is the area of face immersed into water (m2 ); ρw is the density of water (kg/m3 ). The hydraulic diffusion coefficient formulated in Eq. (5) still lacks of theoretical calculation method, and the empirical formula is often employed to estimate, that is [7] D(S) = Dd0 exp(nS)

(9)

where Dd0 is the diffusion coefficient of dry concrete (m2 /s); n is the material coefficient that is taken as 6 for concrete [20]. In addition, supposing that the initial saturation of concrete is zero and Eq. (9) is used in the water diffusion models, [21] deduces the theoretical relationship among A, n and Dd0 , that is 1 1 A 2 1 ( )2 = Dd0 [exp(n)( − 2 ) − ( − 2 )] φ n n n n

(10)

where φ is the porosity of concrete. 2.3 Boundary and Initial Conditions Water contact surface of concrete can be regarded saturated because water is sufficient in the semi-infinite space. Hence, at the interface, we can define a Dirichlet boundary type as: S = 1.0

(11)

6

J. Liu et al.

Concrete in service is hardly dried to the saturation of zero. However, in water absorption test, concrete is always completely dried at 105 °C so that the saturation becomes zero. Therefore, initial saturation is taken as: S = S0

(12)

where S0 is equal to zero for the dry concrete, or a distribution function for the non-dry concrete.

3 Experiments 3.1 Materials and Specimens Cement type used in this test was P · I 42.5 (China standard GB175-2007), a reference cement with 100% cement clinker. Water was general tap water. The fine aggregate was natural river sand with the size of 0.08 mm to 20 mm. A series of mortar and cement paste cubic specimens with length of 100 mm, were cast with the water-cement ratio of 0.4 and 0.5, and the mix proportion was listed in Table 1. All specimens were demoulded after 24 h, and cured for 28 days in the condition of 20 °C and 90% RH. Table 1. Mix proportion and saturated water content of concrete (kg/m3 ). Specimens

w/c

Cement

Water

Dry density

Water loss

Cement

0.4

495

198

Sand

1612

321

Cement

0.5

495

248

1503

346

Mortar

0.4

495

198

1114

2103

136

Mortar

0.5

495

248

1114

1993

172

3.2 Water Capillary Absorption Tests Three specimens of every type were used to measure the maximum water absorption. Before the capillary test, all specimens were immersed into water for seven days to reach saturated (msat ), and then dried at 105 °C for seven days (mdry ). Hence, the absorption saturation can be defined as: S=

m − mdry msat − mdry

(13)

where m is the current mass of concrete (kg). The cement paste specimens were significantly damaged by drying process, and many scaly cracks occurred on all surfaces because of dry shrinkage effect. However, for the mortar specimens, no significant damage was observed on all surfaces though certain damage exists.

Linear Time Invariant Property of Modelling the Water Absorption Process

7

Accuracy of the weighing balance is 0.01 g, and indoor temperature is 26 °C during the tests. A demoulding side face was employed as the absorption surface that was immersed in water roughly 2 mm. Then the specimens were weighted after 10 min, 30 min, 2 h, 3 h, 4 h, 6 h, 1 days, 2 days, 3 days, and 4 days. The absorption surfaces should be wiped by a damp rag before weighting. After the capillary test, all specimens were again immersed into water to obtain the maximum absorption mass. For the test of non-dry specimens whose five faces were sealed by epoxy, they were place in a sealed container for 30 days after the water absorption reached the predetermined degree of saturation. The redistribution of water will contribute to a uniform initial state.

4 Results and Numerical Verification 4.1 Process of Cumulative Water Absorption Figure 1a shows the absorption process of initially dry mortar specimens. The relationship between cumulative absorption mass and square root of time presents two significant linear stages, and the time intersection is roughly between 6 h and 24 h. It can be seen that in the first stage of absorption, the absorption process is stable and basically the same for the specimens with same w/c. For the dry mortar specimens, the absorption mass within four hours is approximately 12–20 g; however, the cumulative absorption difference increased to approximately 20 g after four days. The absorptivity of the specimens with w/c of 0.5 is 1.65 times that of the specimens with w/c of 0.4, and the ratios of second slope to first are 0.2 and 0.24, respectively (see Table 2).

(a)

(b)

Fig. 1. Water capillary absorption process of mortar specimens, (a) different water-cement ratio and (b) different saturation.

8

J. Liu et al. Table 2. Water absorptivity calculated through the test results.

Specimen Mortar

w/c

A(m/s0.5 )

Ratio*

0.4

1.01 × 10−5

0.240

0.5

1.66 × 10−5

0.200

0.65

0.4

2.00 × 10−6

0.450

0.30

0.4

4.40 × 10−6

0.580

0.4

1.19 × 10−4

0.019

Dry

0.5

1.58 × 10−4

0.005

0.65

0.4

2.27 × 10−5

0.060

0.4

1.18 × 10−5

0.060

0.4

6.08 × 10−5

0.110

0.4

3.08 × 10−5

1.000

Saturation Dry Dry

Cement

Dry

0.65 0.40 0.40

*Ratio of water absorptivity in the second stage to that in the first stage.

Figure 1b shows the absorption processes of non-initially dry mortar specimens (w/c of 0.4) whose saturation were 0.65 and 0.30, respectively. We can see that the water absorptivity decreases with an increase in saturation, and the two-stage law of absorption still exists. The absorption mass within four hours drops to 2–6 g due to the initial water content. The water absorptivities decrease by 80% and 56%, respectively, for saturation of 0.65 and 0.30, and the ratios of second slope to first increase in comparison with that of dry specimens (see Table 2). On the basis of the measured results, if the water absorptivity and saturation are linearly fitted, the goodness of fit (R2 ) is 0.91. It suggests that this linear model can be applied to predict the water absorptivity under any saturation. Figure 2a shows the absorption process of cement paste specimens damaged by high temperature of 105 °C. It can be seen that the absorptivity is much greater than that of mortar specimens, reaching one order of magnitude (see Table 2). The absorption mass within four hours is 130–195 g, nearly half of the maximum water contents of 321 g and 346 g. The test results indicate that the development of micro and macro cracks will result in a significant variation in water absorptivity. Though the water absorption rate increases largely, the linear law of the first stage still remains undisturbed. Therefore, we suppose that in the first stage of water absorption the linear law should always exist regardless of the degree and method of damage.

Linear Time Invariant Property of Modelling the Water Absorption Process

(a)

9

(b)

Fig. 2. Water capillary absorption process of cement specimens with crack damage, (a) different water-cement ratio and (b) different saturation.

Figure 2b shows the water absorption of cement paste specimens (w/c of 0.4) whose initial saturation were 0.65 and 0.40, respectively. The absorptivities of the specimens is listed in Table 2, which are the same order of magnitude as that of dry mortar specimens. It can be seen that the linear law in the first stage is still valid in different saturation; however, the absorption characteristics show a great variation, for example, nearly two times difference between the parallel results, and an absorption curve with a single slope. Therefore, as both damage and saturation factors included, the water absorptivity and long-term absorption characteristic will grow more complex. For the cement paste tests, if the mean value of absorptivities is used to linearly fit the relationship between absorptivity and saturation, the goodness of fit (R2 ) is 0.92. 4.2 Verification of Numerical Results Through the numerical solution of water distribution, we can calculate the total water absorption as:  Δm = wsat SdV − S0 (14) V

where wsat is the saturated water content (kg/m3 ). The numerical model is valid in the first linear stage of water absorption because the hydraulic diffusion coefficient is defined according to the water absorptivity. Table 3 lists the simulated test results and the parameters employed in the numerical models. The partial differential governing equation Eq. (6) was solved by finite element method that will not be detailed in this paper. In addition, the constitutive equations and boundary conditions are presented in Sect. 2 as well, resulting in a highly nonlinear problem.

10

J. Liu et al. Table 3. Parameters used in the numerical models of different saturation. A(1e−5 m/s0.5 )

Dd0 (1e−11 m2 /s)

Specimen

Porosity

Saturation

Mortar

0.136

0.00

1.01

4.5

0.136

0.30

0.44

0.9

0.136

0.65

0.20

0.2

0.321

0.00

11.90

111.6

0.321

0.40

6.08

29.1

0.321

0.65

2.27

4.1

Cement

Figure 3 shows the verification of cumulative absorption mass, and numerical water content profiles of mortar specimens (w/c of 0.4) with initial saturation of 0, 0.3 and 0.65. It can be seen that all simulated absorption curves are completely linear; however, only the case of initially dry shows great agreement with the test results. The deviation of simulation results grows with the increase of initial saturation, which indicates that the suitability of Eq. (9) encounters challenge for the non-initially dry condition. Nevertheless, it is interesting that the calculated absorption curve maintains a linear increase regardless of the initial saturation and the nonlinear hydraulic diffusion coefficient. Technique of nuclear magnetic resonance (NMR) was reported to measure the water content profile in the process of water capillary [20]. However, precise water profile is not the focus of this paper; thus, only numerical results of water profiles are shown according to the mechanism models. After absorbing for an hour, the response depths of saturation are 6 mm, 4 mm and 3 mm, respectively, for initial saturation of 0, 0.3, and 0.65. Apparently, the maximum influence depth is nearly 12 mm even if a certainly imprecise calculation is taken into consideration for the non-initially dry conditions. In addition, we can see that the profiles will grow linear with the increase of initial saturation.

Linear Time Invariant Property of Modelling the Water Absorption Process

11

Fig. 3. Verification and numerical results of mortar: (a) and (b) initial saturation of zero; (c) and (d) initial saturation of 0.3; (e) and (f) initial saturation of 0.65.

5 Discussion 5.1 Assumption of the Hydraulic Diffusion Coefficient As shown in Fig. 3, the numerical results of cumulative absorption always grow linearly with the square root of time though distribution curves are greatly different from each other. Therefore, a series of attempts were made to validate the linear law, and a constant diffusion coefficient was used instead of Eq. (9). As expected, we obtain a similar linear law, and the constant value of D is determined to be 4.3 × 10−9 m2 /s as the absorption curve coincides with the test results, shown in Fig. 4a. Consequently, for the proposed

12

J. Liu et al.

numerical model including the Dirichlet boundary Eq. (11) and the governing equation Eq. (6), the relationship between total absorption mass and time must be Δm ∝ t 0.5 , regardless of the form of hydraulic diffusion coefficient. Figure 4b shows the concomitant water content profiles as the constant value of diffusion coefficient was employed. We can obtain a totally different saturation profile under the same volume integral of saturation; hence, a constitutive equation reflecting the actual mechanism is required to predict saturation profile.

(a)

(b)

Fig. 4. Application of constant diffusion coefficient, (a) verification of cumulative absorption mass, and (b) numerical water content profiles.

5.2 Modelling the Long-Term Water Absorption Exponential and constant constitutive equations were respectively used to modelling the long-term water transfer in nearly 0–4 days, and the numerical results are illustrated in Fig. 5. It can be seen that the absorption curves of the both methods always coincide the whole time, and well agree with the test result in first stage, as shown in Fig. 5a. However, in the second stage of longer time, the numerical results show a growing deviation (roughly twofold) that cannot be neglected. Figure 5b shows the water content profiles of the second stage as a result of different assumptions of hydraulic diffusion coefficient. Because the cumulative absorption of simulation is greater than the test value, we speculate that the actual water distribution curve is slightly gentle (like the Constant ones) and the intrusion depth is smaller (like the Exponent ones). Apparently, a new theory is needed to eliminate the significant modelling deviation in the second stage if the long-term water absorption has to be taken into consideration.

Linear Time Invariant Property of Modelling the Water Absorption Process

(a)

13

(b)

Fig. 5. Prediction of long-term water transfer, (a) cumulative water absorption, and (b) numerical water content profiles solved by exponential and constant diffusion coefficients.

6 Conclusions Considering water-cement ratio, initial saturation and damage cracks, water absorption tests of mortar and cement paste were carried out, and different numerical models were compared with the test results. The following conclusions were drawn. • Due to the initial saturation of not zero, such as 0.65 and 0.3, the water absorptivities nearly decrease by an order of magnitude for both mortar and cement paste. • The increased porosity in terms of water-cement ratio can improve water absorptivities, and the impact caused by large amounts of damage cracks is more significant. • The linear relationship between cumulative water absorption and square root of time is universal and firm in the first stage, regardless of initial saturation and damage degree. • For the diffusion model used in water transfer, linear relationship between cumulative water absorption and square root of time is rigidly satisfied whatever the formula of diffusion coefficient that rules the solution of saturation distribution.

Acknowledgements. This research was supported by grants from the National Natural Science Foundation of China (Grant Nos. 52078492, 51778631 and U1934217), which are gratefully acknowledged.

References 1. Sun, W., Zhang, Y., Yan, H., Mu, R.: Damage and damage resistance of high strength concrete under the action of load and freeze-thaw cycles. Cem. Concr. Res. 29(9), 1519–1523 (1999)

14

J. Liu et al.

2. Zhang, P., Wittmann, F.H., Vogel, M., Müller, H.S., Zhao, T.: Influence of freeze-thaw cycles on capillary absorption and chloride penetration into concrete. Cem. Concr. Res. 100, 60–67 (2017) 3. Bao, J., Xue, S., Zhang, P., Dai, Z., Cui, Y.: Coupled effects of sustained compressive loading and freeze–thaw cycles on water penetration into concrete. Struct. Concr. 22, E944–E954 (2021) 4. Polder, R.B., Peelen, W.H.: Characterisation of chloride transport and reinforcement corrosion in concrete under cyclic wetting and drying by electrical resistivity. Cem. Concr. Compos. 24(5), 427–435 (2002) 5. Chen, Y., Gao, J., Tang, L., Li, X.: Resistance of concrete against combined attack of chloride and sulfate under drying–wetting cycles. Constr. Build. Mater. 106, 650–658 (2016) 6. Hall, C., Yau, M.R.: Water movement in porous building materials—IX. The water absorption and sorptivity of concretes. Build. Environ. 22(1), 77–82 (1987) 7. Hall, C.: Water sorptivity of mortars and concretes: a review. Mag. Concr. Res. 41(147), 51–61 (1989) 8. Yang, L., Liu, G., Gao, D., Zhang, C.: Experimental study on water absorption of unsaturated concrete: w/c ratio, coarse aggregate and saturation degree. Constr. Build. Mater. 272, 121945 (2021) 9. Cid, J., Alquier, J., Crausse, P.: Study of moisture transfer in a deformable porous medium through attenuation of two different energy gamma rays. Rev. Sci. Instrum. 63(3), 2057–2064 (1992) 10. Maier, M., et al.: Mass transport in polymer electrolyte membrane water electrolyser liquidgas diffusion layers: a combined neutron imaging and x-ray computed tomography study. J. Power Sources 455, 227968 (2020) 11. Luki´c, B., Tengattini, A., Dufour, F., Briffaut, M.: Visualising water vapour condensation in cracked concrete with dynamic neutron radiography. Mater. Lett. 283, 128755 (2021) 12. Bažant, Z., Najjar, L.: Nonlinear water diffusion in nonsaturated concrete. Matériaux et Constr. 5(1), 3–20 (1972) 13. Yoo, J.-H., Lee, H.-S., Ismail, M.A.: An analytical study on the water penetration and diffusion into concrete under water pressure. Constr. Build. Mater. 25(1), 99–108 (2011) 14. Abyaneh, S.D., Wong, H., Buenfeld, N.: Computational investigation of capillary absorption in concrete using a three-dimensional mesoscale approach. Comput. Mater. Sci. 87, 54–64 (2014) 15. Sun, H.-M., Yang, W., Chen, R.-P., Kang, X.: A coarse-grained water model for mesoscale simulation of clay-water interaction. J. Mol. Liq. 318, 114085 (2020) 16. Smyl, D., Ghasemzadeh, F., Pour-Ghaz, M.: Modeling water absorption in concrete and mortar with distributed damage. Constr. Build. Mater. 125, 438–449 (2016) 17. Eghtesadi, S., Nokken, M.: Effect of cracking and improper consolidation as important concrete defects on water absorption and electrical conductivity. J. Mater. Civ. Eng. 29(11), 04017201 (2017) 18. Li, X., Chen, S., Xu, Q., Xu, Y.: Modeling capillary water absorption in concrete with discrete crack network. J. Mater. Civ. Eng. 30(1), 04017263 (2018) 19. Richards, L.A.: Capillary conduction of liquids through porous mediums. J. Appl. Phys. 1(5), 318–333 (1931) 20. Leech, C., Lockington, D., Dux, P.: Unsaturated diffusivity functions for concrete derived from NMR images. Mater. Struct. 36(6), 413–418 (2003) 21. Lockington, D., Parlange, J.-Y., Dux, P.: Sorptivity and the estimation of water penetration into unsaturated concrete. Mater. Struct. 32(5), 342 (1999)

Study of Failure Patterns in a Reinforced Concrete Beam Subjected to a Varying Range of Blast Anita Bhatt(B)

and Sajad Ahmed Bhat

Department of Civil Engineering, IIT Roorkee, Roorkee 247667, India [email protected]

Abstract. Blast load is assumed to cause local failure, which causes progressive collapse of structure globally. A reinforced concrete beam may experience failure in various forms, i.e., shear, flexure, and combined failure. P-I (Pressure-impulse) curves based on a single degree of freedom analysis are used to describe structural response against blast load. The design codes also suggest preparing P-I curves based on SDOF to obtain structural response subjected to blast. The structural dynamic response of R.C. flexure members for local analysis and complete structural frame for global analysis under a varying range of explosion scenarios was predicted by investigations. For distant blast, the model-based on flexure failure mode predicts approximately good results, while for near blast shear failure modebased model understandably gives better results. A model based on one failure mode will not yield relatively accurate results for various blast ranges. The present study suggests that the model formulation that can capture both flexure and shear response yields an accurate response in the case of the near blast as well as distant blast, irrespective of the structure’s flexibility. Keywords: Distant blast · Near blast · Local failure · Global failure · Flexibility

1 Introduction The fundamentals of blast load and seismic design depend on dynamic behaviour and energy dissipation mechanisms [1]. Many kinds of researches were conducted in the past to assess a structure’s vulnerability subjected to blast loading. Investigations were conducted for R.C. structures to study their vulnerability when subjected to blast loads [2]. The natural period of vibration of structure and the higher modes play a vital role when the response of member/building against blast load is considered [3, 4]. Mindlin and Deresiewicz correctly articulated the fact that the shear stress distribution on a section depends on the mode shape of vibration [5]. They further stated that the maximum shear stress occurs at the neutral axis for the lower vibration modes, while for significantly higher modes of vibration, the shear stress is minimum at the neutral axis. Dynamic response of a structure subjected to blast loading involves non-linear inelastic material behaviour and high strain rate, thus increasing the complexity in analyzing the problem © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 15–29, 2023. https://doi.org/10.1007/978-981-19-3983-9_2

16

A. Bhatt and S. A. Bhat

[2, 6]. To simplify the problem, sometimes blast loading can be represented by using a triangular pulse for a single degree of freedom system [7]. The traditional one-dimensional Euler-Bernoulli elastic beam theory for flexural elastic beam vibrations becomes an inadequate model for analysis if higher modes are adopted [3]. The Euler-Bernoulli beam theory is not suitable for transverse impulsive-type loadings because the propagation velocity of the disturbance approaches infinity as its wave-length approaches zero [8]. The shearing deformations and rotary inertia become increasingly vital if higher modes are considered, thus indicating the Timoshenko beam theory formulation most suitable for analyzing structural elements subjected to blast loading. Since the structural elements experience considerable deformations under blast loading, the non-linear dynamic analysis must be performed, considering the high strain behaviour of materials and thus observing various modes of failure experienced by structural elements under blast loading. Another approach for analyzing the given problem can be blast wave-structure interaction [9]. There are various technical manuals available for blast-resistant design, which can be adopted for guidelines [10, 11]. We can also simulate the problem via computer programming in software like ABAQUS, MATLAB, ANSYS, etc. [2, 12]. There are various mechanisms [13–15] in which R.C. structural elements can fail, namely flexural failure with formations of flexural plastic hinges [1, 7], flexure shear failure characterized by the formation of inclined tension cracks and flexural cracks [9, 16] and direct shear failure [8, 17, 18] near supports or joints of the elements that comprise the structure [19]. The present study analyzes an isolated reinforced concrete column and the reinforced concrete frame using SDOF modelling based on Euler-Bernoulli theory and continuum FEM modelling using ABAQUS software to observe local and global failure, respectively. Enhancing values of shear capacity and bending capacity can take care of high strain rate effects that arise during impulse loading and treating the Timoshenko beam as a viscoelastic beam. The research aims to study various numerical and analytical approaches involving highly dynamic events towards civil structure response subjected to blast loading.

2 Methodology 2.1 Analysis of Reinforced Concrete Column/Beam Subjected to Blast Loading External beams and columns are severely affected by blast loading as they are in close vicinity of the blast, thus making them primary load resisting members. Internal members are only affected by the load transmitted from the primary members. The analysis of individual reinforced concrete fixed columns subjected to blast loads is performed to observe the behaviour of individual members. The data regarding geometric/material properties of the column is summarized in Table 1 below: In the present study, a reinforced concrete column with fixed boundary conditions is adopted to observe local failure. ABAQUS was used for the modelling and analysis. The three-dimensional models of the reinforced concrete column were developed, and the reinforcement was modelled as solid circular bars, as shown in Fig. 1(a) and 1(b).

Study of Failure Patterns in a Reinforced Concrete Beam Subjected

17

Table 1. Geometric/material properties of the column. Column section

Material parameters

Reinforcement

L (mm)

B (mm)

H (mm)

fck (MPa)

Ec (GPa)

fy (MPa)

Es (GPa)

No. of bars

Dia (mm)

4100

450

400

48.5

34.8

415

200

10

25

(a)

(b)

Fig. 1. (a) Three-dimensional model of the column in ABAQUS. (b) ABAQUS modelling of reinforcement details.

The Concrete Damage Plasticity (CDP) model was used as the material model for concrete, whereas reinforcing steel was modelled as elastic-perfectly plastic material. The grade of concrete was taken to be M48.5, and that of steel was Fe 415. The materials models are represented graphically below as Fig. 2(a) and 2(b).

18

A. Bhatt and S. A. Bhat

(a)

(b)

Fig. 2. (a) Elastic-perfectly plastic stress-strain model for reinforcing steel of grade Fe 415. (b) Concrete Damage Plasticity (CDP) stress-strain model for M 48.5 concrete

2.2 Analysis of Reinforced Concrete Frame Subjected to Blast Loading Reinforced concrete frames may suffer significant damage when subjected to blast loadings. A complete frame model should be analyzed under these loads to have a general idea about the damage caused by the blast at a global level. Analysis of a complete R.C. frame subjected to blast loading is considered. The frame chosen for this purpose was a three-storey plane frame with two bays, as shown in Fig. 3. Frame Details Number of storeys 3 (G +2) Number of bays

2

Clear storey height 4 m Clear bay width

5.4 m

Beam

300 mm* 450 mm Three no. bars 20 mm dia. top

Reinforcement Column

Three no. bars 20 mm dia. bottom 450 mm*400 mm

Reinforcement

Ten no. bars 25 mm dia.

Concrete

M 48.5

Steel

Fe 415

Fig. 3. 3D model of frame in ABAQUS.

The three-dimensional model of the frame was developed in ABAQUS. The concrete elements of the frame were modelled as three-dimensional solid elements, while the

Study of Failure Patterns in a Reinforced Concrete Beam Subjected

19

reinforcement steel was modelled as a truss element. The model was meshed with solid brick elements of size 100 mm. The element type used for analysis was C3D8R, which uses three degrees of freedom per node with a total of 8 nodes in the element. This element uses reduced integration for analysis with stiffness hourglass control. The section details of elements of the frame are described in Fig. 4.

Fig. 4. Section details of beam and column used in the frame.

The behaviour of steel reinforcement was modelled as a simple elastic-perfectly plastic stress-strain curve. As the grade of steel used is Fe 415, the yield stress is 415 MPa. The yield strain was obtained by using the elastic modulus (E) of 200 GPa. The stressstrain curve is presented in Fig. 2(a). The model used for concrete is the Concrete Damage Plasticity (CDP) model, which ensures the quasi-brittle behaviour of materials and is available in ABAQUS. The grade of the concrete used was M 48.5 (fck = 48.5 MPa). 2.3 Blast Load Calculation Numerous mathematical models were developed to estimate the blast loading approaching the structure. Brode introduced a formula based on the reduced distance to calculate the peak pressure) [2]. Later Newmark and Hensen established a relationship for calculating maximum blast overpressure [20]. Krauthammer et al. [15] simplified the loading function to a simple triangular pulse. Kingery and Bulmash (1984) [21] developed complete charts based on the reduced distance for calculating all blast wave parameters. In the present study, blast pressure exerted on the column was calculated by the MATLAB program for three blast cases, as shown in Table 2 subsequently for comparing the FEM-based analysis with the SDOF-based analysis.

20

A. Bhatt and S. A. Bhat Table 2. Details of Blast load exerted on the column.

Case

Mass (kg)

Standoff distance (m)

Peak Reflected pressure (kPa)

Peak load for SDOF (kN)

Duration (ms)

1

100

10

830

980.1

2

300

10

2400

2833.04

6.04

3

600

10

4800

5667.8

10.35

1

The fixed boundary condition was simulated by arresting the degree of freedom for a particular set of nodes. The different nodes with arrested degrees of freedom were carefully selected to closely simulate the behaviour of the support. The loading on the column was applied as uniform pressure. The models were analyzed for loading mentioned in the aforementioned Table 2. The results corresponding to load case 2 (Peak pressure of 2400 kPa with a duration of 6.04 ms) are discussed in the subsequent section. The face of the column exposed to blast was selected, and a uniform pressure varying with time was applied to that face. The temporal loading variation is shown in Fig. 5(a) and 5(b).

(a)

(b)

Fig. 5. (a) Blast load variation with time. (b) Blast loading as applied to the column.

The load to the frame and column was applied using the CONWEP extension of ABAQUS. The load was applied to only the exposed surface of the frame. The CONWEP simulates the hemispherical blast wave and calculates the pressure function for each element separately. The program automatically calculates the arrival time for the pressure wave based on the geometry of the structure and the blast. The charge assumed for analysis was 300 kg of TNT at a distance of 10 m from the frame. The peak pressure at the middle of the ground floor column as generated by CONWEP is given in Fig. 6.

Study of Failure Patterns in a Reinforced Concrete Beam Subjected

21

Pressure -Time History Blast Pressure (Pa)

1400000 1200000 1000000 800000 600000 400000 200000 0 0

0.01

0.02

0.03

0.04

0.05

0.06

Time (s) Fig. 6. Blast pressure variation at the centre of the first column.

3 Results and Discussion Non-linear time history analysis was conducted under given loading (Table 2). The maximum displacement was recorded for the three cases. The comparison of the maximum displacement from elastic-plastic SDOF idealization and continuum finite element model was done for all three load cases. The displacement time history plots based on SDOF based analysis and FEM-based analysis for case 2 are shown subsequently in Fig. 7. 0.03

FEM

Displacement (m)

0.025

SDOF

0.02 0.015 0.01 0.005 0 0

1

2

3

4

5

6

7

8

t/td (Time as rao of blast duraon) Fig. 7. Displacement response of midpoint of a fixed column from elastic-plastic SDOF and FEM model.

The behaviour of blast load on an isolated reinforced column is observed in detail using ABAQUS software based on FEM modelling. As the blast load approaches the structural element, it observes tension and compression damage. The column stiffness decreases as the damage in the concrete increases. The damage in the column is higher

22

A. Bhatt and S. A. Bhat

at the maximum displacement. Figure 8(a) and 8(e) display the blast pressure at a time of 10 ms and 16 ms, respectively. Figure 8(b) and 8(f) depict the displacement corresponding to 10 ms and 16 ms, respectively. Figure 8(c) and 8(d) display the compression and tension damage plots for a time of 10 ms. Figure 8(g) and 8(h) display the compression and tension damage plots for a time of 16 ms.

(a)

(b)

(c)

(d)

Fig. 8. (a) Blast pressure at time 10 ms. (b) Displacement at time 10 ms. (c) Damage compression at time 10 ms. (d) Damage tension at time 10 ms. (e) Blast pressure at time 16 ms. (f) Displacement at time 16 ms. (g) Damage compression at time 16 ms. (h) Damage tension at time 16 ms.

Study of Failure Patterns in a Reinforced Concrete Beam Subjected

(e)

(f)

(g)

(h)

Fig. 8. continued

23

24

A. Bhatt and S. A. Bhat

As the blast load approaches the structural element, both tension and compression damages were observed. Significant tension damages were observed at the outer face in the middle of the column and at the inner face of the support. 3.1 Progressive Damage The damage in the frame increases as the blast wave moves forward with time because more area is exposed to the blast pressure as time passes. The results are presented in the form of damage plots at different times obtained using FEM-based analysis. The compression damage response, tension damage response, and displacement time history were obtained corresponding to the blast pressure variation. The damage plots help identify or locate the zones where the damage is maximum for the structure. The located damage zones can be optimally reinforced to prevent collapse. The damage results can also prompt the designer to use different materials or structural components. The results of the analysis are described below. Figure 9(a) and 9(e) display the blast pressure at a time of 10 ms and 16 ms, respectively. Figure 9(b) and 9(f) depict the displacement corresponding to a time of 10 ms and 16 ms, respectively. Figure 9(c) and 9(d) display the compression and tension damage plots for a time of 10 ms. Figure 9(g) and 9(h) depict the compression and tension damage plots for a time of 16 ms. The analysis of the frame demonstrated the propagation of the blast pressure as well as the damage. Elements directly exposed to the blast were severely affected and suffered more damages. The damage started from the column exposed at the ground floor and propagated in the direction away from the blast. This analysis further emphasized the need for local analysis of exposed elements. Figure 10(a) and 10(b) display the displacement time histories for isolated fixed column and reinforced concrete frames at different floor levels, respectively. The maximum displacement of 9.34 mm at time 0.013 s was observed for the isolated column. The maximum displacement of 9.12 mm at 0.013 s for the ground floor column, 15 mm at 0.015 s for the middle floor column, and 15.2 mm at 0.052 s for the top floor column were observed for framed structure. Figure 11(a) and 11(b) display the tension damage time history observed for an isolated fixed reinforced concrete column and reinforced concrete frame, respectively. The tension damage started at time 0.009 s, and correspondingly tensile stress 3.26e06 N/m2 was observed for the isolated column. The tension damage started at time 0.009 s, and correspondingly tensile stress 3.26e06 N/m2 was observed for the combined isolated column and frame. The analysis of the frame demonstrated the propagation of the blast pressure as well as the damage. Elements directly exposed to the blast were severely affected and suffered more damages. The damage started from the column exposed on the ground floor and propagated in the direction away from the blast. This analysis further emphasized the need for local analysis of exposed elements.

Study of Failure Patterns in a Reinforced Concrete Beam Subjected

(a)

(c)

25

(b)

(d)

Fig. 9. (a) Blast pressure at time 10 ms. (b) Displacement at time 10 ms. (c) Damage compression at time 10 ms. (d) Damage tension at time 10 ms. (e) Blast pressure at time 16 ms. (f) Displacement at time 16 ms. (g) Damage compression at time 16 ms. (h) Damage tension at time 16 ms.

26

A. Bhatt and S. A. Bhat

(e)

(f)

(g)

(h)

Fig. 9. continued

Study of Failure Patterns in a Reinforced Concrete Beam Subjected 0.01

0.016

0.009

0.014

Displacement (m)

Displacement (m)

0.008 0.007 0.006 0.005 0.004 0.003 0.002

27

Ground Storey Middle storey Top storey

0.012 0.01 0.008 0.006 0.004 0.002

0.001

0

0

0

0.02

0.04

0.06

0

Time (s)

0.02

0.04

0.06

Time (s)

(a)

(b)

1

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

0.9 0.8 0.7

Damage

Damage

Fig. 10. (a) Displacement time history for a fixed column. (b) Displacement time history for the frame on different floors.

0.6 0.5 0.4

Ground Storey

0.3 0.2

Middle Storey

0.1

Top Storey

0 0

0.02

0.04

Time (s) (a)

0.06

0

0.02

0.04

0.06

Time (s) (b)

Fig. 11. (a) Tension damage for a fixed column. (b) Tension damage for the frame on different floors.

Figure 12(a) and 12(b) display the compression damage time history observed for an isolated fixed reinforced concrete column and reinforced concrete frame, respectively. The tension damage started at time 0.009 s, and correspondingly tensile stress 3.26e06 N/m2 was observed for the isolated column. The tension damage started at time 0.01 s, and correspondingly tensile stress 1.62e07 N/m2 was observed for the combined isolated column and frame. The compression damage in the column of the top storey is negligible.

28

A. Bhatt and S. A. Bhat 0.3

0.16

0.25

0.14

Ground Storey Middle Storey

Damage

Damage

0.12 0.2 0.15 0.1

0.1 0.08 0.06 0.04

0.05

0.02

0 0

0.02

0.04

Time (s) (a)

0.06

0 0

0.02

0.04

Time (s)

0.06

(b)

Fig. 12. (a) Compression damage for a fixed column. (b) Compression damage for the frame on different floors.

4 Conclusion The dynamic response of reinforced concrete isolated column and frame subjected to blast loading was studied to observe the structure’s local and global behaviour. The response was observed using SDOF analysis (based on Euler-Bernoulli theory) and continuum finite element modelling software ABAQUS (based on Timoshenko theory). The propagation of damage for an isolated column and different elements of the frame was studied using ABAQUS. The following outcome can be derived from the research. The response of the reinforced concrete frame deviates from the traditional dynamic (seismic) response. The local element failure is predominant, and failure propagates from the elemental level to the global level. In a reinforced concrete frame, regions closest to the blast loading are most vulnerable to damage, and the damage propagation starts from this region. Additional care should be taken while designing such elements. The reinforced concrete frame and isolated reinforced concrete column observed both tension and compression damage, indicating both flexure and shear failure probability of structure. SDOF response based on a single mode of failure may not provide information regarding the actual behaviour of a structural element. Thus analysis models which can cover various modes of failure for a given instance should be preferred. This aspect still requires thorough validation. The behaviour and vulnerability of reinforced concrete frames have to be studied in detail to provide required strengthening design aspects for resisting blast loads, and will be undertaken by the authors.

References 1. Smith, P.D., Hetherington, J.G.: Blast and Ballistic Loading of Structures (1994) 2. Jayasooriya, R.: Vulnerability and Damage Analysis of Reinforced Concrete Framed Buildings (2010) 3. Van Der Meer, L.J., Kerstens, J.G.M., Bakker, M.C.M.: P-I diagrams for linear-elastic cantilevered Timoshenko beams including higher modes of vibration. Heron 55, 51–83 (2010)

Study of Failure Patterns in a Reinforced Concrete Beam Subjected

29

4. Bhatt, A., Bhargava, P., Maheshwari, P.: Contribution of higher modes in the dynamic response of reinforced concrete member subjected to blast. In: IOP Conference on Series Earth Environmental Science, vol. 871, p. 012003 (2021). https://doi.org/10.1088/1755-1315/871/1/ 012003 5. Mindlin, R.D., Deresiewicz, H.: Thickness-shear and flexural vibrations of a circular disk. J. Appl. Phys. 25, 1329–1332 (1954). https://doi.org/10.1063/1.1721554 6. Nassr, A.A., Razaqpur, A.G., Tait, M.J., Campidelli, M., Foo, S.: Strength and stability of steel beam columns under blast load. Int. J. Impact Eng. 55, 34–48 (2013). https://doi.org/10. 1016/j.ijimpeng.2012.11.010 7. Syrmakezis, C.A.: Introduction to Structural Dynamics (1996) 8. Ross, B.T.J.: Impulsive direct shear failure in RC slabs. J. Struct. Eng. 111, 1661–1677 (1985) 9. Ekström, J.: Concrete structures subjected to blast loading, p. 20 (2015) 10. Department of Defense of UFC 3-340-02. Structures to resist the effects of the accidental explosions. Unified Facil. Criteria. 1943 (2008) 11. IS:4991-1968: Criteria for Blast Resistant Design of Structures for Explosion Above Ground. Bur. Indian Stand, pp. 1–43 (1968) 12. Ibrahim, Y.E., Ismail, M.A., Nabil, M.: Response of reinforced concrete frame structures under blast loading. Procedia Eng. 171, 890–898 (2017). https://doi.org/10.1016/j.proeng. 2017.01.384 13. French, R., Maher, E., Smith, M., Stone, M., Kim, J., Krauthammer, T.: Direct shear behavior in concrete materials. Int. J. Impact Eng. 108, 89–100 (2017). https://doi.org/10.1016/j.iji mpeng.2017.03.027 14. Nassr, A.A., Ghani Razaqpur, A., Campidelli, M.: Effect of initial blast response on RC beams failure modes. Nucl. Eng. Des. 320, 437–451 (2017). https://doi.org/10.1016/j.nuceng des.2017.06.019 15. Krauthammer, T.: Analysis of impulsively loaded reinforced concrete structural elements-I. Theory 48, 851–860 (1993) 16. Yu, R., Zhang, D., Chen, L., Yan, H.: Non-dimensional pressure–impulse diagrams for blastloaded reinforced concrete beam columns referred to different failure modes. Adv. Struct. Eng. 21, 2114–2129 (2018). https://doi.org/10.1177/1369433218768085 17. Wang, W., Zhang, D., Lu, F.Y., Tang, F.J., Wang, S.C.: Pressure-impulse diagram with multiple failure modes of one-way reinforced concrete slab under blast loading using SDOF method. J. Cent. South Univ. 20, 510–519 (2013). https://doi.org/10.1007/s11771-013-1513-z 18. Krauthammer, T., Astarlioglu, S.: Direct shear resistance models for simulating buried RC roof slabs under airblast-induced ground shock. Eng. Struct. 140, 308–316 (2017). https:// doi.org/10.1016/j.engstruct.2017.02.056 19. Ross, T.J.: Direct Shear Failure in Reinforced (1983) 20. Ngo, T., Mendis, P., Gupta, A., Ramsay, J.: Blast loading and blast effects on structures - an overview. Electron. J. Struct. Eng. 7, 76–91 (2007) 21. Kingery, C.N., Bulmash, G., Bulmash, G.: U.S. Army Ballistic Research Laboratory; U.S. Army Ballistic Research Laboratory. Airblast Parameters from Tnt Spherical Air Burst and Hemispherical Surface Burst; Technical Report Arbrl-Tr, 02555; U.S. Army Armament and Development Center, Ballistic Research Laboratory: Aberdeen Proving Ground, Maryland (1984)

Corroded Stirrups Effects on the Shear Behavior of Reinforced Concrete Slender Beams Pier Paolo Rossi(B)

and Nino Spinella

Department of Civil Engineering and Architecture, University of Catania, 95123 Catania, Italy {pierpaolo.rossi,nino.spinella}@unict.it

Abstract. Stirrups corrosion causes damage effects on the shear response of reinforced concrete elements. Such damage mainly includes cracking and spalling of concrete in the cover, and reduction of the cross-sectional area of the reinforcing steel. The need for tools able to predict the reduction of the load-bearing capacity of reinforced concrete element has led to several formulations based on the numerical regression of available experimental results. This paper proposes a mechanical model for the prediction of the shear response of reinforced concrete slender beams with corroded reinforcement. The proposed procedure, based on the Modified Compression Field Theory, can reproduce the entire load-displacement curve. The corrosion effects on the effective beam width and rebar cross-section are taken into account. The accuracy of the proposed model is assessed using a database of sixty-two specimens. The proposed model achieves accurate results. Keywords: Shear · Corrosion · Stirrups · Beams

1 Introduction The lack of maintenance of Reinforced Concrete (RC) beams have led to deterioration of the mechanical properties of these structural elements, especially in existing bridges [1–3] and buildings [4]. Corrosion causes cracking and spalling of the concrete cover, and reduction of the cross-sectional area of the reinforcing steel [5]. These effects also depend on the type of corrosion (uniform or concentrated) of the reinforcement. Therefore, significant reduction of the ultimate strength and maximum deflection may occur. Moreover, concentrated corrosion (i.e. “pitting”) may also change the failure mode of the beam, from ductile for bending to brittle for shear [6–8]. Fernandez et al. [5] observed that pitting is significant for structures where the moment redistributions is achievable. Corrosion of the longitudinal reinforcing steel has largely been studied, and a large literature is available. By contrast, a few research studies are focused on corrosion of stirrups, and consequently on the reduction of the shear capacity of RC beams. Stirrups confine concrete, and then they reduce the critical shear crack width. Therefore, corrosion of stirrups significantly reduces the plasticity branch of the steel rebars and the shear strength capacity. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 30–37, 2023. https://doi.org/10.1007/978-981-19-3983-9_3

Corroded Stirrups Effects on the Shear Behavior

31

It follows that stirrups corrosion is a critical issue. It affects not only the shear strength but also the ductility of the beam. Therefore, the deformation capacity of RC beams needs to be assessed. However, most of the models proposed in the literature are able to predict only the load-bearing capacity at the Ultimate Limit State (ULS) of corroded RC beams, and, in many cases, the basic equations are empirical [9]. Recently, Cladera et al. [10] modified the Compression Chord Capacity Model (CCCM) to predict the shear strength of corrosion-damaged RC beams. Predictions were compared to the experimental results of 146 slender and deep beams failing in shear, where stirrups and/or longitudinal reinforcement were subjected to corrosion, and achieved satisfactory results. This study aims to propose a theoretical model to predict the entire load-deformation curve of corroded RC beams by extending a previously established procedure based on the Modified Compression Field Theory (MCFT) [11]. A crack element containing longitudinal and transverse smeared steel rebars is considered. Moreover, both reduced rebar cross-section of and concrete cover because of corrosion are taken into account. The proposed model allows the calculation of the strain and stress fields by respecting all the equilibrium and compatibility equations. A validation of the proposed model, against an experimental test database reported in the literature, is presented.

2 Stirrups Corrosion Structural Effects Steel rebars with uniform and/or concentrated corrosion show a reduction in the cross-sectional area (especially due to pitting) and deterioration of the geometric and mechanical properties of the concrete cover due to cracking and spalling. 2.1 Cross-Section Reduction of Steel Rebar The cross-sectional loss ratio, ηa , can be evaluated as: ηa =

φ2 − φ2 As0 − As × 100 = 0 2 × 100 As0 φ0

(1)

where As0 and φ 0 are the cross-sectional area and diameter, respectively, of the steel rebar before corrosion; As and φ are the cross-sectional area and diameter after corrosion. The maximum value (ηam ) along a length of the steel rebar is an estimate of pitting. It is safe to use ηam for predicting the shear strength of RC members. When only the weight loss ratio is reported, ηw = 100 × (m0 – m)/m0 (m0 = mass of a length of the steel rebar before corrosion; and m = mass after corrosion), it should been converted into cross-section loss ratios, ηam . With this aim, Cladera et al. [10] recently simplified a discontinued equation between ηam and ηw , originally introduced by Lu et al. [9], in the following form: ηam = 1.36 ηw

(2)

32

P. P. Rossi and N. Spinella

2.2 Effective Beam Width Higgins et al. [12] observed that the volumetric expansion of corrosion products of steel stirrups causes spalling of the concrete cover. Thus, the external part of the beam is cracked and not able to carry stress. This corrosion-structural effect depends on stirrup spacing (sv ) and concrete cover (c). Higgins and co-authors proposed two equations, based on empirical data and theoretical computations, to estimate the effective beam width (bw,eff ). In the following, these equations are reported in the form proposed by Cladera et al. [10], presenting continuity for the case sv = 5.5(c + φ v ), where φ v is the stirrup diameter: bw,eff = bw − 2(c + φv ) + bw,eff = bw −

sv 5.5

5.5 (c + φv )2 sv

(3) (4)

Equations (3) and (4) are for sv ≤ 5.5(c + φ v ) and sv > 5.5(c + φ v ), respectively. Moreover, the same authors suggested to reduce the beam width (bw ) to the effective value (bw,eff ) only when stirrups exhibit at least 10% of cross-section loss.

3 Proposed Model for the Shear Response of Corroded RC Beams The proposed model predicts the shear-displacement response of slender RC beams with corroded rebars. It is based on the MCFT equations, and, as assumed by other researchers in their formulations [13, 14], the flexural model is solved separately by the shear one. 3.1 Flexural Model A top axial strain (εct ) value is assumed, then the flexural cross-section strain is estimated by adjusting the longitudinal strain of bottom reinforcement (εsb ) until the internal axial load equals the external one [15]. Flexural analysis provides the bending moment (M), the curvature (χ ), and the average axial strain from flexure (εxf ) at the centroid level of the cross-section (see Fig. 1).

Corroded Stirrups Effects on the Shear Behavior

33

Fig. 1. a) Lateral view and b) cross-section of the RC beam. c) Flexural and d) Shear model.

3.2 Shear Model The equilibrium conditions of MCFT require that: σcv = σc1 − τ/cotθ

(5)

σc2 = σc1 − τ (1/cotθ + cotθ )

(6)

with σ cv = concrete stress along the vertical direction; σ c1 and σ c2 = concrete principal tensile and compressive stresses, respectively; τ = shear stress; and θ = strut angle (see Fig. 1). Clamping stress (σ v ) can be obtained as σ v = σ cv + ρ sv σ sv ; where ρ sv = Asv /(bw sv ) = geometric ratio of the transverse reinforcement; Asv is the total area and sv spacing of stirrups, respectively; and σ sv is the stirrup stress. In slender beams, the clamping stress along a cross-section enough far from the point load can be neglected [16], and then Eq. (6) allows calculation of the shear stress as follows:   τ = (σc1 + ρsv σsv )cotθ = (σc1 − σc2 )cotθ/ 1 + cot 2 θ (7) Hence, the strut angle inclination θ is obtained: cot 2 θ = (−σc2 − ρsv σsv )/(σc1 + ρsv σsv )

(8)

At the same time, the compatibility equations of the MCFT require that: εx + εv = ε1 + ε2

(9)

cot 2 θ = (εv − ε2 )/(εx − ε2 )

(10)

34

P. P. Rossi and N. Spinella

with εx and εv = horizontal and vertical strains, respectively; and ε1 and ε2 = principal tensile and compressive strains, respectively. Equations (8) and (10) return the same value. However, the two equations of the strut angle give the vertical strain (εv ), which is independent of the crack angle (θ ). In any case, elastic or plastic behavior of transverse reinforcement must be distinguished. If stirrups are elastic, the steel stress is σ sv = E s εv , and a 2nd-order equation in εv is obtained. The close-solution for εv is the follow:    2 εv = −Be + Be − 4Ae Ce /(2Ae ) (11) where coefficients Ae , Be , and C e are: Ae = ρsv Es

(12)

Be = σc1 +Aεv (εx − 2ε2 )

(13)

Ce = σc2 (εx − ε2 ) − σc1 ε2

(14)

If stirrups yield, εv is the solution of a linear equation: εv = −Cy /By

(15)

By = σc1 + ρsv fyv

(16)

Cy = σc2 (εx − ε2 ) − σc1 ε2 + ρsv fyv (εx − 2ε2 )

(17)

where coefficients By and C y are:

The compression strength of concrete in a diagonally cracked web is reduced as a function of the principal strain. Therefore, the following equations are used [6, 11]:       ε2 2 ε2 − σc2 = fce 2 (18) εc0 εc0 η=

1 fce ≤1 = fc 0.8 + 0.34(ε1 /εc0 )

(19)

where f ce = effective compressive strength of concrete; and εc0 = strain at peak stress in concrete. Cracked concrete is assumed to carry tensile stress by tension stiffening [11]:   (20) σc1 = fct / 1 + 500 ε1 The tensile stress (σ c1 ) contribution to shear strength [Eq. (7)] has an upper limit (σ c1,max ) due to the capacity of cracked concrete to bridge forces across the crack sides. The maximum admissible value of the tensile stress is the minimum of the two

Corroded Stirrups Effects on the Shear Behavior

35

expressions derived from equilibrium of forces across the crack in the longitudinal and transverse directions:  

σc1,max = min τi cotθ + ρsl fyl − σsl ; τi /cotθ + ρsv fyv − σsv (21) τi =

0.18 fc 0.31 +

24w dg +16

    ≤ ρsv fyv − σsv − ρsl fyl − σsl sinθ cosθ

(22)

where f yl and f yv = yielding strength of the longitudinal and vertical rebars, respectively; τ i = local shear stress; d g = maximum coarse aggregate size (in mm); w = ε1 S mθ is the average crack width (in mm); and f c is in MPa. In Eqs. (21) and (22), the residual reinforcement stresses at the crack are taken as zero when the yield stress is reached. To estimate the average diagonal crack spacing, S mθ = 1/(sinθ /S ml + cosθ /S mv ), the average crack spacing along the two orthogonal directions S ml and S mv are assumed equal to the stirrups spacing (sv ) and the effective depth (d) of the cross-section, respectively. 3.3 Analytical Procedure For a generic top-strain value (εct ), a flexural analysis is performed, and values of bending moment (M), curvature (χ ) and flexural axial strain (εxf ) are obtained. For a simply supported beam, the shear force is constant along the shear span (a); thus, the applied shear stress at the critical section - away av from the support - is taken as τ flex = (M/av )/(bw d). The shear stress capacity [Eq. (7)] must be ≥ τ flex . If this is not the case, shear failure is achieved. Summarizing, the following steps are involved in the shear model: Set a value of ε2 . Set a value of ε1 , and then calculate η from Eq. (19) and σ c2 from Eq. (18). Use the stored values of θ and εv to calculate w and σ c1 . Calculate εv from Eq. (11) or (15). Calculate σ sv = min{E s εv ; f yv }. Calculate the new θ by Eq. (10) and σ c1,max from Eqs. (21) and (22). If σ c1 > σ c1,max , then set σ c1 = σ c1,max and return to Step 3. 7. Check if the value of ε2 obtained by Eq. (9) is equal to the assumed one. In case not, return to Step 2 and adjust ε1 . 8. Calculate τ from Eq. (7). If τ = τ flex return to Step 1 and adjust ε2 .

1. 2. 3. 4. 5. 6.

4 Validation of the Proposed Model A database recently collected by Cladera et al. [10] is used here to validate the proposed model. It includes 62 slender beams with corroded stirrups, which failed in shear. All the specimens have rectangular cross-sections, and the height of cross-section is less than 350 mm. Moreover, the RC beams are over-reinforced in bending (ρ l = 0.99 ÷ 3.01%). The concrete compressive strength (f cm ) is ranging between 22.5 and

36

P. P. Rossi and N. Spinella

50 MPa. The geometric percentage of stirrups (ρ v ) is between 0.14 and 0.52%. The corrosion degree of stirrups reaches the maximum value of 97.2%. The experimental versus numerical results at failure are reported in Fig. 2. They highlight that the proposed model accurately predicts the shear strength of RC beams with corroded stirrups. A mean value equal to 1.01 and a Coefficient of Variation (CoV) equal to 0.29 have been achieved for the ratio of the experimental to numerical shear strength (See Fig. 2).

Fig. 2. Comparison between the experimental and numerical shear strength.

5 Conclusions This research study presents an analytical model to predict the shear behavior of RC beams with corroded stirrups. The proposed procedure take into account the structural effects of corrosion: rebar cross-section loss and effective beam width reduction due to spalling of cover. The numerical results reported in the paper show that the proposed model provides good performance in terms of shear strength prediction (mean = 1.01 and CoV = 0.29). Further investigations, also with the help of sophisticated tools like the Non-Linear Finite Elements Methods, are needed to better figure out the relationship between the inclination angle of strut and the rate of corrosion of stirrups.

References 1. Colajanni, P., Recupero, A., Ricciardi, G., Spinella, N.: Failure by corrosion in PC bridges: a case history of a viaduct in Italy. Int. J. Struct. Integrity 7(2), 181–193 (2016). https://doi. org/10.1108/IJSI-09-2014-0046 2. Recupero, A., Spinella, N.: Experimental tests on corroded prestressed concrete beams subjected to transverse load. Struct. Concr. 20, 2220–2229 (2019). https://doi.org/10.1002/suco. 201900242 3. Recupero, A., Spinella, N.: Preliminary results of flexural tests on corroded prestressed concrete beams. Proc. Fib. Symp. (2019). Krakow, Poland

Corroded Stirrups Effects on the Shear Behavior

37

4. Lo Presti, A., Recupero, A., Spinella, N.: Influence of rebar corrosion on RC frame push-over response. In: Hordijk, D.A., Lukovi´c, M. (eds.) High Tech Concrete: Where Technology and Engineering Meet, pp. 2118–2126. Springer, Cham (2018). https://doi.org/10.1007/978-3319-59471-2_242 5. Fernandez, I., Herrador, M.F., Marí, A.R., Bairán, J.M.: Structural effects of steel reinforcement corrosion on statically indeterminate reinforced concrete members. Mater. Struct. 49(12), 4959–4973 (2016). https://doi.org/10.1617/s11527-016-0836-2 6. Spinella, N., Colajanni, P., Recupero, A., Tondolo, F.: Ultimate shear of RC beams with corroded stirrups and strengthened with FRP. Buildings 9(2), 34 (2019). https://doi.org/10. 3390/buildings9020034 7. Campione, G., Cannella, F.: Engineering failure analysis of corroded R.C. beams in flexure and shear. Eng. Fail. Anal. 86, 100–114 (2018). https://doi.org/10.1016/J.ENGFAILANAL. 2017.12.015 8. Recupero, A., Spinella, N., Tondolo, F.: Failure analysis of corroded RC beams subjected to shear-flexural actions. Eng. Failure Anal. 93, 26–37 (2018). https://doi.org/10.1016/j.engfai lanal.2018.06.025 9. Lu, Z.H., Li, H., Li, W., Zhao, Y.G., Dong, W.: An empirical model for the shear strength of corroded reinforced concrete beam. Constr. Build. Mater. 188, 1234–1248 (2018). https:// doi.org/10.1016/J.CONBUILDMAT.2018.08.123 10. Cladera, A., Marí, A., Ribas, C.: Mechanical model for the shear strength prediction of corrosion-damaged reinforced concrete slender and non slender beams. Eng. Struct. 247, 113163 (2021). https://doi.org/10.1016/J.ENGSTRUCT.2021.113163 11. Vecchio, F.J., Collins, M.P.: The modified compression-field theory for reinforced concrete elements subjected to shear. ACI J. Proc. 83(2), 219–231 (1986). https://doi.org/10.14359/ 10416 12. Higgins, C., et al.: Shear capacity assessment of corrosion-damaged reinforced concrete beams. Corvallis, Oregon, USA (2003) 13. Colalillo, M.A., Sheikh, S.A.: Behavior of shear-critical reinforced concrete beams strengthened with fiber-reinforced polymer—analytical method. ACI Struct. J. 111(6), 1385–1396 (2014). https://doi.org/10.14359/51687036 14. Mostafaei, H., Vecchio, F.J.: Uniaxial shear-flexure model for reinforced concrete elements. J. Struct. Eng. 134(9), 1538–1547 (2008). https://doi.org/10.1061/(ASCE)0733-9445(200 8)134:9(1538) 15. Spinella, N.: N-M-χ interaction for arbitrary cross section under biaxial bending and axial load. Pollack Periodica 8(3), 87–100 (2013). https://doi.org/10.1556/Pollack.8.2013.3.9 16. Acevedo, A.B., Bentz, E.C., Collins, M.P.: Influence of clamping stresses in the shear strength of concrete slabs under uniform loads. J. Earthq. Eng. 13(1), 1–17 (2009). https://doi.org/10. 1080/13632460902813190

Response Control on Seismic Retrofit of Low-Rise RC Frame Using Viscous Damper Panumas Saingam(B) Department of Civil Engineering, School of Engineering, King Mongkut’s Institute of Technology Ladkrabang, Bangkok 10520, Thailand [email protected]

Abstract. In recent years, seismic design and detailing requirements for buildings have considerably improved worldwide. For example, in Thailand, a new seismic design code was published in 2021, but many existing buildings do not satisfy the new code and require retrofit. The seismic retrofit is required to improve the seismic performance of the existing building. However, the response control method to control the target story drift ratio of the retrofitted RC buildings using the viscous damper is lack introduction. This study proposes a response control retrofit strategy using viscous dampers, designed using an equivalent linearization approach. A constant stiffness method is introduced to efficiently distribute the dampers along with the building height. The stiffness of the damper is equally distributed for all stories. A design example is introduced of a low-rise reinforced concrete school building in Thailand, which was damaged in the 2014 Mae Lao earthquake. Nonlinear response history analysis is used to validate the introduced method. The results indicate that the average peak story drifts ratios can be controlled within the target story drift ratio of 0.67% rad. Keywords: Response control · Seismic retrofit · Low-rise RC building · Viscous dampers

1 Introduction Thailand has historically been considered to have a low seismic hazard, and the most current existing buildings were designed to resist only gravity load. However, the seismic resistance was not considered in the designs. In recent years, the earthquake has damaged several buildings around the world. Therefore, in 2009, the Department of Public Works and Town & Country Planning of Thailand published a seismic design specification for new buildings [1], followed by a specification for seismic retrofit [2]. Shortly after the May 15, 2014 Mae Lao earthquake struck, causing extensive damage to older buildings that were constructed before the seismic specifications were implemented. Much of the damage was observed in reinforced concrete (RC) structures, as reported in [3, 4] including some school buildings. Figure 1a shows a typical 2-story RC school building in Thailand, which was constructed in many places in the country. The buildings are non-ductile RC moment frames with vertical © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 38–48, 2023. https://doi.org/10.1007/978-981-19-3983-9_4

Response Control on Seismic Retrofit of Low-Rise RC Frame

39

irregularities due to infill masonry walls. The 2-story building, as shown in Fig. 1b, received significant structural damage during the Mae Lao earthquake, as indicated by the severe damage to the beam-column joint at the top of the ground story columns. According to the severe damage from the Mae Lao in 2014, the Ministry of Interior of Thailand published a new seismic design code, which is a regulation in 2021 [5]. This affects that all buildings in the seismic region in Thailand are required to improve their seismic performance by strengthening or retrofitting. A conventional retrofit solution for seismically deficient reinforce concrete (RC) frames is to install a stiff shear wall [6, 7], which limits drift and ensures that the mainframe remains elastic but imposes large floor accelerations. This implies extensive nonstructural damage, as building contents and nonstructural components are unlikely to be detailed for seismic resistance in Thailand. An alternative retrofit solution is to employ energy dissipation devices to control both drift and accelerations while protecting the existing structure. Energy dissipation devices have been reported to be an effective seismic retrofit solution for RC frame buildings and have been applied in practice. A retrofit design method for RC frame structures, where the buckling-restrained braces (BRBs), are installed in parallel with a supplementary elastic steel frame was introduced in [8, 9]. In addition, a retrofit RC building was proposed in the study [10]. The results from the study [8–12] indicated that energy-dissipating devices can improve efficiently the seismic performance of the retrofitted RC buildings. The challenge in Thailand is that the seismic hazard is relatively small, with response control retrofits requiring smaller dampers installed at fewer stories than in a typical Japanese application. While still a potentially effective retrofit solution, the low demands introduce unique challenges in determining an efficient number, size and distribution of dampers, as the optimal damper type, distribution and design approach may be different from those countries. Although the study [9] proposed installing viscous dampers into the target building, the controlled response on seismic retrofit RC building method with a viscous damper is rarely proposed. This study introduces a response control method to retrofit the RC buildings, which is named a constant stiffness method. The method is used to design and retrofit the example 2-story building as shown in Fig. 1. The seismic performance of the existing RC building and the retrofitted RC building is investigated and compared. The design is verified through nonlinear response history analysis.

Fig. 1. School building: (a) Elevation of 2-story, (b) Observed damage.

40

P. Saingam

2 Constant Stiffness Method The inelastic story force-displacement response of the bare RC frame is first obtained through pushover analysis. While the example building as shown in Fig. 1 was subjected to large drift and strength degradation due to column bending failure in Mae Lao 2014 earthquake, only the response up to the target story drift is needed for this analysis, which is set as θ tar = 1/150 (0.67% rad) to validate the proposed design method. A tri-linear degrading Takeda model is adopted to represent the existing RC frame [13] and is calibrated to match the area under the pushover curve at each story ith story. The post-yield response is assumed perfectly plastic (α 2 = 0), the yield story drift θ fy is limited to 1/100–1/300 rad, and the crack (δ fci ) to yield (δ fyi ) displacement ratio is initially set as μc = 0.1 [13], but permitted up to μc = 0.2. The yield shear force Qfyi and displacement δ fyi are then estimated, and the cracking shear force Qfci and displacement δ fci are adjusted to produce the same shear force ratio N = Qfyi /Qfci and cracked stiffness ratio α 1 = [(Qfyi − Qfci )/(δ fyi − δ fci )]/K fi at all stories, where the initial story stiffness K f0i = Qfci /δ fci . This treatment reduces the multi-story frame to a simplified representation, with the same pre-yield stiffness ratio α 1 , crack-to-yield drift ratio μc , ductility μf = δ tar /δ fyi , and secant stiffness K fμ = pK f 0 at each story. The multi-degree of freedom (MDOF) model is then reduced to an equivalent single-degreeof-freedom (SDOF) system using the equivalent height (H eq ), mass (M eq ), and stiffness (K f ) [8]. The cyclic hysteretic response of the SDOF system is shown in Fig. 2 for the cracked and yielding stages. pKf0=Kfμ

RC frame (μc 1, μf < 1) and (b) Yielding (μf > 1).

The hysteretic energy dissipated by the RC frame (E f ) depends on the unloading stiffness (K ul ), with the unloading stiffness degradation parameter λ assumed as 0.4 [13]. The equivalent hysteretic damping for a constant cyclic displacement (hf μ = hf 0 + Ef /4πEfe ) is then determined from the hysteretic energy E f , strain energy E fe and intrinsic damping hf 0 is assumed to be 0.03 for RC structures.

Response Control on Seismic Retrofit of Low-Rise RC Frame

41

As displacement ductility in each cycle varies when subjected to earthquake excitation, the study [14] introduced the average damping concept (Eq. 1) is employed. hf μ

1 = hf 0 + μtar

μtar  hf μ − hf 0 d μ

(1)

1

However, for simplicity, the average equivalent damping (hfμ ) (Eq. 2) may be estimated from the equivalent damping of the maximum cycle (hf μ ) and a calibrated damping reduction factor (Rfμ ). The average hfμ and peak hf μ equivalent damping are shown in Fig. 3a and the corresponding reduction factors Rfμ is shown in Fig. 3b.   hf μ = hf 0 + Rf μ hf μ − hf 0 (2)

h`fμ hfμ Rfμ

h`fμ ,hfμ

0.15 0.12 0.09 0.06 0.03 0

0

0.4

0.8 μf 1.2

1.6

2

1 0.8 0.6 0.4 0.2 0

Rfμ=0.4

0

0.4

0.8 μf 1.2

1.6

2

Fig. 3. Equivalent damping reduction factor: (a) Equivalent damping and (b) Rfμ .

The spectral displacement S d (T fμ ,hfμ ) of the bare RC frame is estimated from the design elastic displacement response spectrum at the secant period (T fμ ), reduced from the 5% damped spectrum using equivalent damping (hfμ ) and reduction factor proposed which was introduced as a = 25 in the study [15]. The secant period is given by Eq. 3 and uses the secant stiffness (K fu ) of the bare RC frame at the target drift. The roof drift of the bare RC frame (θ fμ ) is estimated from Eq. 4 and dampers are required if θ fμ exceeds the target story drift θ tar .  Meq Tf μ = 2π (3) Kf μ θf μ =

Sd (Tf μ , hf μ ) Heq

(4)

42

P. Saingam

The viscous (VS) dampers are velocity-dependent devices, which are effective in controlling drifts and enhancing the system energy dissipation. The VS are typically installed in series with an elastic brace element, with the assembly acting in parallel to the RC and supplemental steel frames, as indicated by Fig. 4(a). The component force-displacement relationships are shown in Fig. 4(b), where E dVS is the equivalent damping of the viscous damper, Ka the loss stiffness Eq. 8, C d the damping coefficient, K b the brace stiffness, ω the circular frequency, E e the equivalent potential energy of a total system, Ka the storage stiffness Eq. 5, and ηa the brace-damper subassembly’s loss factor Eq. 7. Cd2 Kb ω2

Ka = Ka = Kf

(5)

Kb2 + Cd2 ω2  2 θf μ p θtar Dh2 − p γs +

ηa =

(6)

1 ηa

Ka Kb =  Ka Cd ω

(7)

(a) Viscous damper EdVS

Q

Steel Frame

Qdy

Q

Ka ” +

μfyδfy

δ

QSFμ

KSF μfyδfy δ (b)

Fig. 4. (a) Viscous damper with the brace, RC frame and elastic steel frame model. (b) Viscous damper and elastic steel frame force-displacement model.

Response Control on Seismic Retrofit of Low-Rise RC Frame

Ka =

Cd Kb2 ω Kb2

43

(8)

+ Cd2 ω2

The required loss stiffness ratio (rdVS ) = Ka /Kf of the brace-damper subassembly is referred to as the added component and is given by for the RC frame cracking and yielding stages as shown in Eqs. 9(a) and 9(b), respectively. 

rdVS =

Ka = Kf



 μ (1−p) 1 + 25 hf 0 + π1 · pμc +μc · Rf μ f   (μc μf > 1, μf < 1) 1 + 25hf 0 (γs + 1/ηa ) + (0.5x25ReqVS ) ⎛ ⎞⎞  λ 

⎛ pμ +μc −p μf (1+μc ) θf μ 2 ⎝1 + 25⎝hf 0 + 1 · f ⎠⎠ p − 1 · R f μ π pμ +μc θ

K  rdVS = a = Kf

p

θf μ θ

2

−1



(9a)

f



 1 + 25hf 0 (γs + 1/ηa ) + (0.5x25ReqVS )

(μf > 1)

(9b)

ReqVS

Though the supplemental damping provided by the viscous dampers is velocity, rather than displacement dependent, the hysteretic damping of the RC frame still contributes to equivalent damping of the system, which consequently varies cycle by cycle. The reduction factor ReqVS relating the average (heq ) and peak cycle (heq ) equivalent damping is shown in Fig. 5.

1 0.8 0.6 0.4 0.2 0

ReqVS =0.50

0

0.4

3.0 2.0 1.0 0.5 Kd /K f= 0.1

0.8 μf 1.2

1.6

2

Fig. 5. Damping reduction factor for the system with VS dampers (ReqVS ).

3 Design Example and Validation 3.1 Design Example This section applies the constant drift method procedure to the 2-story RC school buildings depicted in Fig. 1, which requires seismic retrofit. Thailand Seismic Design Code [5]. The newest seismic Thai code has been written based on ASCE 7-05 [16], and the design level spectral response acceleration parameters for these structures are S DS = 0.56(g) and S D1 = 0.24(g) (site class D, Phan, Chiang Rai), approximately half of the seismic demands in Japan.

44

P. Saingam

The story masses of the 2-story building are 266 and 172 tons at the first and roof stories, respectively, and the fundamental period of the bare RC frame is 0.59 s in both the longitudinal and transverse directions. Pushover curves and calibrated tri-linear Takeda models for the 2-story building are shown in Fig. 6 for the first story and second story, respectively. Figure 6a and 6b illustrate the story shear to story displacement of the 1st story for longitudinal and transverse directions, respectively. Figure 6c and 6d show the story shear to story displacement of the 2nd story for longitudinal and transverse directions, respectively. Structural properties of the bare SDOFRC structures are summarized in Table 1. The ratios of the area under the pushover curves and tri-linear model (Apushover /Atri ) are close to 1.0 at each story, indicating a good fit. Table 2 shows damper distributions for 2-story building model.

1F Pushover curve 1F Takeda Model

300

θtar=1/150 rad.

200

α1Kf1 = 0.6Kf1 θfy=1/275 rad. (0.36%)

Qfc=130 100 0

500 Qfy=459 400

H1 = 3450 mm

Story shear (kN)

Story shear (kN)

500 Qfy=442 400

δfc=2.51

δ =12.55 δtar=23

10 fy

20 30 40 50 Stroy displacement (mm)

60

0

2F Pushover cur 2F Takeda Mode θtar=1/150 rad. (0.67%) α1Kf2 = 0.60Kf2 θfy=1/275 rad. (0.36%) Kf2 =40.67 kN/mm δfc=2.58

0

δfy=12.91 δtar=23.67

10

0

500

Story shear (kN)

Story shear (kN)

0

Kf1 = 53.76 kN/mm δfc=2.51

δfy=12.55 δtar=23

10

20 30 40 50 Story displacement (mm)

20 30 40 50 Story displacement (mm)

60

70

H2 = 3550 mm

Qfy=442 400

400 Qfy= 357 300

Qfc =105 100

α1Kf1 = 0.60Kf1 θfy =1/275 rad. (0.36%)

(b) Transverse direction for 1st story

H2 = 3550 mm

200

θtar=1/150 rad. (0.67%)

200

70

(a) Longitudinal direction for 1st story 500

1F Pushover Curve 1F Takeda Modal

300

Qfc=135 100

Kf1 =51.90 kN/mm

0

H1 = 3450 mm

300

θtar =1/150 rad. (0.67%)

200 Qfc=130 100 0

60

(c) Longitudinal direction for 2nd story

2F Pushover Curv 2F Takeda Modal

α1Kf2 = 0.60Kf2 θfy =1/275 rad (0.36%) Kf2 =50.41 kN/mm

0

δfc=2.58

δfy=12.91 δtar=23.67

10

20 30 40 50 Story displacement (mm)

60

7

(d) Transverse direction for 2nd story

Fig. 6. Pushover curve and tri-linear model of the 2-story building (a) Longitudinal direction for the first story, (b) Transverse direction for the first story, (c) Longitudinal direction for the second story, (d) Transverse direction for the second story. Table 1. Characteristic of bare RC frame Direction

μf

K f0 kN/mm

H eq mm

M eq ton

μc

α1

T fμ sec

K fμ kN/mm

Rfμ = 0.6 and θ tar = 1/150 rad Longitudinal

1.83

46.6

4510

412

0.20

0.60

0.97

17.3

Transverse

1.83

49.4

4469

415

0.20

0.60

0.95

18.3

Response Control on Seismic Retrofit of Low-Rise RC Frame

45

Table 2. Damper distributions for 2-story building model Direction

Story

Ka /Kf

K fi kN/mm

 Kai kN

Kb kN/mm

heq

Cd kN · s/mm

ni

Longitudinal

2

0.11

40.7

4.4

68.2

0.104

0.36

2

51.9

5.7

68.7

0.46

2

50.4

4.5

68.2

0.36

1

53.8

4.8

68.7

0.39

1

1 Transverse

2 1

0.09

0.097

3.2 Validation of Effectiveness of Constant Stiffness Method To validate the retrofit designs, nonlinear response history analyses were performed, targeting the design acceleration response spectrum described earlier (S DS = 0.56(g) and S D1 = 0.24(g)). Two suites of ground motions were used, reflecting common practice in Japan and the US, which the Thai code is based upon. First, a suite of four earthquake ground motions were spectrally matched, consisting of El Centro NS (1940), JMA Kobe NS (1995), TAFT EW (1925), and Hachinohe NS (1968). The duration of four observed waves was 30 s for each wave and compared to the design spectrum in Fig. 7(a). Additionally, a suite of 11 scaled single component records were selected from the PEER NGA2 ground motion database 2 (Fig. 7(b)). Scaling was conducted over a target period range of 0.2T 1,min and 1.5T 1,max following ASCE 7-16 [17], where T 1,min and T 1,max are the minimum and maximum fundamental periods from the two models, resulting in a target period range of 0.1 to 2 s. Records were limited to strike-slip events with magnitudes of 6 ≤ M w ≤ 7.5 within 20 km and on soil class D (180 ≤ V s,30 ≤ 360 m/s), consistent with the dominant seismic hazard in the Chiang Rai province and local site conditions. Scale factors varied from 0.5 to 2.0, and the average spectrum matches or exceeds the target spectrum over the range of interest. While the average acceleration response spectra are similar for both suites, the average displacement spectra exceed the design spectra by a relatively large margin for the scaled suite at periods greater than 1 s, as shown in Fig. 7 while the displacement spectra are shown in Fig. 7(c). The peak story drift ratios of the existing RC frame and the retrofitted models using viscous dampers are shown in Fig. 8 for the 2-story building. Only the longitudinal direction is shown here as the response is similar in the two orthogonal directions. Drift is concentrated at the first story, exceeding the target story drift angle and matching the observed damage experienced during the Mae Lao earthquake. Figure 8(a) and Fig. 8(b) show the peak story drift ratios of the existing RC frame and the retrofitted with viscous dampers, respectively. Using the spectrally matched suite, adding dampers in proportion to the RC frame stiffness using the constant stiffness method improves the seismic performance of the retrofitted building. The second story drift under all ground motions is 0.2% for the design using the constant stiffness method (Fig. 8(b)) but increases to 0.67% to 0.78% at the first story. The scaled ground motions produce a similar average drift distribution for the 2-story building (Fig. 8) but exhibit greater record-to-record variability.

46

P. Saingam

800

800

Acceleration (cm/s 2 )

1000

Acceleration (cm/s2 )

1000

600 400 200

600 400 200 0

0 0

0.5

1 Period (s)

1.5

0

2

1.5

2

(b) Scaled suite

25

Displacement spectrum ,Sd (cm)

1 Period (s)

(a) Matched suite

Average(Spectral Average matched suite) (Scaled suite)

h=5

20 15

Pseudo disp design spectra

10 5 0

0.5

0

0.5

1 1.5 Period (s)

2

(c) Displacement spectra

Fig. 7. 5% damped response spectra: (a) Matched suite, (b) Scaled suite and (c) Displacement spectra.

2

2

θtar = 1/150 (0.67%)

θtar = 1/150 (0.67%)

Average (Matched suite)

Average (Matched suite)

Story

Story

Average (Scaled suite)

Average (Scaled suite) 1

0.0

0.5 1.0 Peak story drift (%)

(a) Existing RC frame

1.5

1

0.0

0.5 1.0 Peak story drift (%)

1.5

(b) Retrofitted building

Fig. 8. Peak story drifts of the 2-story building in longitudinal direction: (a) Bare RC frame, (b) Retrofitted building.

Response Control on Seismic Retrofit of Low-Rise RC Frame

47

4 Conclusions A response control retrofit based on equivalent linearization, which is called a constant stiffness method, was introduced to assign an efficient damper distribution. The peak story drift ratios of the introduced method were compared to the existing RC frame on the example RC building. The results indicated that assigning viscous damper in proportion to the bare RC frame stiffness using the constant stiffness method improved the seismic performance of the retrofitted building. In addition, the average peak story drift ratios from both matched and scaled suites can be controlled within the target story drift ratio of 0.67% rad. Further study should apply and investigate the performance of the proposed retrofit method to taller RC buildings.

References 1. Department of Public Works and Town & Country Planning (DPT). Thailand Seismic Design Specification (2009) 2. Department of Public Works and Town & Country Planning (DPT). Strengthening Buildings Recommendation Specification (2014) 3. Lukkunaprasit, P., et al.: Performance of structures in the Mw 6.1 Mae Lao earthquake in Thailand on May 5, 2014 and implications for future construction. J. Earthq. Eng. 20, 219–242 (2015) 4. Ornthammarath, T., Warnitchai, P.: 5 May 2014 MW 6.1 Mae Lao (Northern Thailand) earthquake: interpretations of recorded ground motion and structural damage. Earthq. Spectra 32, 1209–1238 (2016) 5. Ministry of Interior. Thailand Seismic Design Code (2021) 6. Canbay, E., Ersoy, U., Ozcebe, G.: Contribution of reinforced concrete infills to seismic behavior of structural systems. ACI Struct. J. 100, 637–643 (2003) 7. Foutch, D.A., Hjelmstad, K.D., Calderon, E.D.V., Gutierrez, E.F., Downs, R.E.: The Mexico earthquake of September 19, 1985: case studies of seismic strengthening for two buildings in Mexico City. Earthq. Spectra 5, 153–174 (1989) 8. Sutcu, F., Takeuchi, T., Matsui, R.: Seismic retrofit design method for RC buildings using buckling-restrained braces and steel frames. J. Constr. Steel Res. 101, 304–313 (2014) 9. Lee, D., Taylor, D.P.: Viscous damper development and future trends. Struct. Des. Tall Build. 10, 311–322 (2001) 10. Saingam, P., et al.: Composite behavior in RC buildings retrofitted using buckling-restrained braces with elastic steel frames. Eng. Struct. 219, 110896 (2020) 11. Di Sarno, L., Manfredi, G.: Experimental tests on full-scale RC unretrofitted frame and retrofitted with buckling-restrained braces. Earthq. Eng. Struct. Dyn. 41, 315–333 (2012) 12. Saingam, P., Matsuzaki, R., Nishikawa, K., Sitler, B., Terazawa, Y., Takeuchi, T.: Experimental dynamic characterization of friction brace dampers and application to the seismic retrofit of RC buildings. Eng. Struct. 242, 112545 (2021) 13. Takeda, T., Sozen, M.A., Norby Nielsen, N.: Reinforced concrete response to simulated earthquakes. J. Struct. Div. 96(12), 2557–2573 (1970). https://doi.org/10.1061/JSDEAG.000 2765 14. Newmark, N.M., Rousenblueth, E.: Fundamentals of Earthquake Engineering. Prentice-Hall Inc. (1971) 15. Kasai, K., Ito, H.: Passive control design method based on tuning of stiffness, yield strength, and ductility of elasto-plastic damper. J. Struct. Constr. Eng. AIJ 595, 45–55 (2005)

48

P. Saingam

16. American Society of Civil Engineers (ASCE). Minimum Design Loads for Buildings and Other Structures 2005 (ASCE/SEI 7-05) 17. American Society of Civil Engineers (ASCE). Minimum Design Loads for Buildings and Other Structures 2016 (ASCE/SEI 7-16)

Strength and Durability Properties Empty Fruit Bunch Ash (EFBA) as Partial Replacement of Cement Gunalaan Vasudevan1(B) , Kribadharan Uthayasirppi1 , and Mohd Mawardi bin Hassim2 1 Department of Construction Management, Faculty of Built Environment, Tunku Abdul

Rahman University College Malaysia, Kuala Lumpur, Malaysia [email protected] 2 Jurutera Awam, Cawangan Jalan, Ibu Pejabat JKR Malaysia, Kuala Lumpur, Malaysia

Abstract. This research showed the results of experiments evaluating the use of empty fruit bunch ash (EFBA) from the oil palm industry as partial replacement for ordinary Portland cement. Research on the reuse of waste materials in the concrete industry has been quite intensive in the past decade. The objective of this research is to identify the performance of EFBA as a partial cement replacement in the production of concrete. EFBA of various amounts, namely 5%, 10%, 15% and 20% by volume, was added as a replacement for ordinary Portland cement. The results showed that EFBA concrete greatly improved the compressive and flexural strength of concrete. The rate of water absorption of EFBA concrete was reduced as EFBA filled up the existing voids, making it more impermeable. However, the compressive strength of the EFBA concrete decreases gradually when the amount of EFBA increased. It can be concluded that the optimum percentage of EFBA as a partial cement replacement is 15%. In this direction, an experimental investigation of ultrasonic pulse velocity, carbonation test, compressive strength, flexural strength and water absorption was undertaken to test the performance of EFBA and admixtures as partial replacement for cement in concrete. Keywords: Empty fruit bunch ash · Admixture · Strength and durability concerete properties

1 Introduction Oil palm trunks and oil palm fronds represent about 75% of those wastes and that they are left rotten within the plantations for mulching and nutrient exercise. The remaining 25% consisting of palm kernel shells, monocarp fibre and empty fruit bunches are dried and utilized palm oil mills or power plants to generate electricity through combustion at 800–1000 °C [1]. Burning of oil palm by-products results in the formation of oil palm ash (or palm oil fuel ash) that is disposed to open fields because it is useless as fertilizer

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 49–57, 2023. https://doi.org/10.1007/978-981-19-3983-9_5

50

G. Vasudevan et al.

because of its less nutrient worth. The ash will simply get carried by wind and end in air pollution on a moist day inflicting traffic hazards conjointly potential of health hazards resulting in bronchi and respiratory organ diseases. Therefore, a way of using Empty Fruit Bunch Ash (EFBA) as a possible replacement for cement should be projected to cut back the environmental result of cement and therefore the value to provide cement [2]. Concrete is one of the highly used constructional material in the world. Annually almost 2.9 billion tons of cement are produced and at this pace, it is expected to reach 4 billion tons by 2020. Clinker is the main ingredient in cement which is obtained by heating up limestone which involves burning of fossil fuel that accounts for 40–50% of emission. Additionally, decomposing of limestone in this process accounting for the remaining 50– 60% which in combined accounts for 5–7% of global CO2 . The combustion process in producing cement mainly contributes to the emission of CO2 which is a major greenhouse gas (GHG). Build-up of large amount of GHG is the primary cause of global warming. Consequently, results in severe climate changes such as sea-level rise from thermal expansion and ocean acidification from increased absorption of CO2 which then pours down as acid rain affecting the integrity of modern structures [3]. Such issues have raised concerns on environment where it has become a necessity to come up with alternative cementitious material for concrete construction.

2 Methodology The Fig. 1 show conceptual framework to determine the performance of the concrete that contain the Empty Fruit Bunch Ash (EFBA) by preparing concrete cube sample and tested to obtain some of the basic engineering properties. The concrete mix design is done by systematic analysis and chooses the proportion of the ingredient to use the concrete mix to produce an economical concrete and also with strength that desired when the cube is hardened. The variables which can be controlled are water cement ratio, maximum aggregate size, aggregate grading and use of admixture.

Strength and Durability Properties Empty Fruit Bunch Ash (EFBA)

51

Fig. 1. Conceptual framework.

3 Result and Discussion 3.1 Sieve Analysis Based on Fig. 2, show that as for Empty Fruit Bunch Ash (EFBA), the dried EFBA was sieved through a 150 um sieve to remove coarse and undesired particles which were incompletely burnt during combustion. The EFBA used had a specific gravity and Blaine’s surface area of 1.88 and 506 m2 /kg, respectively. Normal mining sand with sizes between 0.3 and 5 mm was used as fine aggregate in this investigation whereas the water absorption, specific gravity and fineness modulus of the sand were found as 0.81%, 2.79 and 2.88, respectively [4]. As for the coarse aggregate, an aggregate size of between 4.7 mm–10 mm were used which had a water absorption, specific gravity and bulk density of 0.74%, 2.65 and 1455 kg/m3 , respectively.

52

G. Vasudevan et al.

Fig. 2. Sieve analysis.

3.2 Compressive Strength Test Based on Fig. 3, the graph indicates a linier increase in compressive strength from 0% of EFBA to 15% of EFBA for all 3 curing days with 15% EFBA cured for 28 days sample obtaining the highest reading of 34.85 MPa.Upon curing for 28 days, the compressive increased even more even for the controlled sample which obtained 26.22 MPa. As for 5%, 10% and 15%, the reading also increased gradually to 28.88 MPa, 32.98 MPa and 34.85 MPa, respectively. The increase of compressive strength to that of the controlled sample for 5%, 10% and 15% is about 10.1%, 25.8% and 32.9%, respectively. In short, the increasing trend of the compressive strength linier to the increasing percentage of EFBA obtaining up to twice the compressive strength compared to controlled sample

Fig. 3. Compressive strength test.

Strength and Durability Properties Empty Fruit Bunch Ash (EFBA)

53

Fig. 4. Results of UPV

indicates that EFBA is equivalent to cement in acting as a binder. EFBA forms a stronger bond when added with cement up to a certain percentage. 3.3 Ultrasonic Pulse Velocity (UPV) Based on Fig. 4 the UPV values are increasing gradually with the addition of EFBA up to 15% for all 7, 14 and 28 days of curing. Beyond 15% which is at 20% of EFBA replacement, the UPV value reduced but not lower than the controlled sample which indicates that the pozzolanic material still makes a significant difference. At 28 days of curing, the UPV value recorded 30.52 m/s for the controlled sample which was the highest amongst the 3 curing days sample. The UPV value for 5%, 10% and 15% gradually increased to 36.95 m/s, 40.52 m/s and 45.6 m/s or about 21.1%, 32.8% and 49.4% of that of the controlled sample, respective but then reduces to 41.44 m/s for 20% EFBA yet still higher than the controlled sample value. Based on the result, it indicates that the test run for 7 days of cured sample shows the highest average UPV value amongst the other curing days. This could be due to the ongoing pozzolanic reaction which forms pores within the sample which affects the velocity of pulse travelling through the sample. 3.4 Flexural Strength Test Based on the Fig. 5, the flexural strength of the samples increased with curing time but decreases with an increase in EFBA replacement beyond 15%. When EFBA is ground to a reasonably high fineness, the rate of flexural strength gain of the samples were significantly improved due to the hydration reaction, nucleation effect, packing effect and pozzolanic reaction [4]. These results conclude that EFBA with high fineness can be used as a good pozzolan in cement-based materials and can be used to replace OPC up to 15%. At 28 days, the controlled sample obtained 2.8 MPa flexural strength or about 21.7% increase compared to the controlled sample of 7 days. The flexural strength for 5%, 10% and 15% of EFBA increased to 2.9 MPa, 3.1 MPa and 3.2 MPa or about 3.6%, 10.7% and 14.3%, respectively. At 20% EFBA, the pattern is still the same as other curing days.

54

G. Vasudevan et al.

Fig. 5. Flexural strength test.

3.5 Rebound Hammer Concrete Test The Fig. 6, show at the 28 days of curing, again the trend is the same as 7 days and 14 days of curing but record the best or highest average result compared to other samples. The controlled sample obtained a total of 21 MPa of force which is the highest amount recorded amongst the controlled samples. As the EFBA replacement are increased to 5%, 10% and 15%, the value obtained increased to 25 MPa, 26 MPa and 28 MPa or about 19%, 23.8% and 33.3% of that of the controlled sample.

Fig. 6. Rebound hammer test.

3.6 Microstructures of Empty Fruit Bunch Ash (EFBA) The apparent morphologies of the EFBA examined by SEM are shown in Fig. 7. Scanning Electron Microscope (SEM) test was conducted on the control mix sample and on the

Strength and Durability Properties Empty Fruit Bunch Ash (EFBA)

55

15% EFBA, 28 days cured sample [5]. Adding EFBA to the concrete had provided additional calcium oxide which is an important component in producing secondary calcium-silicate-hydrate (C-S-H) gel. It is found that by adding EFBA into the sample, small pores are formed in the structure and C-S-H covers almost the entire fractured surface. The majority large spaces were filled with C-S-H gel because of the increased pozzolanic reactivity and hydration rate thus creating a denser structure compared to the controlled sample [6]. With increasing curing time, production of Ca(OH)2 decreases leading to an increased formation of C-S-H. Hence, EFBA can still be used as a partial replacement for cement without affecting the water absorption rate.

Fig. 7. (a) SEM images of control mix, (b) SEM images of EFBA.

3.7 Carbonation Test Carbonation is a consequence of the transformation of calcium hydro-oxide [Ca(OH)2 ] to calcium carbonate [CaCO3 ] altering the microstructure of the cement paste by decreasing porosity. Figure 8 shows that about 50% relative humidity (RH) and about 5% CO2 for 28 days can cause the pore voids being filled and also likely for C-S-H attack which is what defects reinforcements.Oxygen and moisture are the components required for corrosion of reinforcement steel. Thru this test, the depth of carbonation is determined. The test is carried out by spraying Phenolphthalein indicator on freshly exposed surface of concrete sample. Carbonation depth is accessed when the indicator turns pink in contact with alkaline in sample with pH value greater than 9. At lower lever pH, the

56

G. Vasudevan et al.

indicator is colourless as the sample is carbonated and the protective layer gets destroyed exposing the reinforcement to corrosion. With enhanced dispersion and filler effect, the sample shows a fairly low carbonation which indicates that EFBA can serve as a favourable pozzolanic material to replace OPC in producing a low carbonation depth and high strength mortar.

Fig. 8. Carbonation test.

4 Conclusion and Recommendation The overall objective of the work was to investigate the feasibility of EFBA as a partial replacement of cement to concrete mix. This study included the preparation of concrete mixes containing and the evaluation of EFBA concrete properties in fresh and hardened states. The studying properties involved mix workability, compressive strength and flexural strength. According to experimental results, the usage of in EFBA concrete mixes as an alternative of disposal for EFBA is possible. The current type of EFBA accumulated in dumping sites and the expected future type were used in making concrete mixes. The influence of both types on concrete properties was studied. In all cases there was an optimum quantity of EFBA which can be used without introducing any change in mix preparation and acceptable properties were still be produced [7]. The study showed that the EFBA can be used in production of cement without changing the normal industrial process. Recommendation for further research The following recommendations are proposed for further research and study in order to from a complete picture of using empty fruit bunch ash in concrete mixes: 1) Investigation is needed on the different superplasticizer used together with empty fruit bunch ash in the concrete mix design. 2) Study is required for palm kernel shell powder mix with other by-products in the concrete mix design. 3) Further investigation of the higher percentage of replacement of the cement by the empty fruit bunch ash in concrete mix design.

Strength and Durability Properties Empty Fruit Bunch Ash (EFBA)

57

4) Longer curing time is required in order to understand the behaviour of the concrete, and it might be giving the different result for empty fruit bunch ash. 5) Durability test such as sulphate attack test must be further study in order to understand percentage loss in weight of the empty fruit bunch ash concrete by sulphate attack.

Acknowledgment. This study was supported by the Faculty of Built Environment, Tunku Abdul Rahman University College under the Centre for Construction Research.

References 1. Vasudevan, G.: Performance of Alum Sludge as partial replacement for cement adding superplasticizer. In: IOP Conference Series: Materials Science and Engineering (2019) 2. Thomas, S., Kumar, S., Arel, H.S.: Sustainable concrete containing palm oil fuel ash as a supplementary cementitious material – a review. Renewable and Sustainable Energy Reviews 80, 550–561 (2017) 3. Kroehong, W.: Effect of palm oil fuel ash fineness on the microstructure of blended cement paste. Constr. Build. Mater. 25, 4095–4104 (2011) 4. Islam, M.M.U., Mo, K.H., Alengaram, U.J., Jumaat, M.Z.: Mechanical and fresh properties of sustainable oil palm shell lightweight concrete incorporating palm oil fuel ash. J. Clean. Prod. 115, 307–314 (2016) 5. Hamada, H.M., Yahaya, F.M., Muthusamy, K., Jokhio, G.A., Humada, A.M.: Fresh and hardened properties of palm oil clinker lightweight aggregate concrete incorporating Nano-palm oil fuel ash. Constr. Build. Mater. 214, 344–354 (2019) 6. Khankhaje, E., Hussin, M.W., Mirza, J., Rafieizonooz, M., et al.: On blended cement and geopolymer concretes containing palm oil fuel ash. Mater. Design 89, 385–398 (2016) 7. Vasudevan, G.: Performance of steel slag as a partial replacement fine aggregate incorporating superplasticizer. In: Awang, M., Meor M Fared, M. (eds.) ICACE 2019. Lecture Notes in Civil Engineering, vol. 59, pp. 151–160. Springer, Singapore (2020). https://doi.org/10.1007/978981-15-1193-6_17

Experimental Study on the Performance of FRP Grid Reinforced ECC Composite Layer-Concrete Interface Weiwen Li1 , Jie Liu1 , Jiahao Zhang1 , Shiying Tang1 , Meizhong Wu1 , and Xu Yang2(B) 1 Guangdong Provincial Key Laboratory of Durability for Marine Civil Engineering, College of

Civil and Transportation Engineering, Shenzhen University, Shenzhen 518060, China 2 College of Civil and Environmental Engineering, Harbin Institute of Technology, Shenzhen 518055, China [email protected]

Abstract. In this paper, the bond slip between the composite reinforcement layer and concrete is studied by single shear test. The composite reinforcement layer contains the CFRP (Carbon Fiber Reinforced Polymer) grid embedded within ECC (Engineering cementitious composites) matrix. Considering the influence of bond length, matrix type and ECC strength, the failure mode of the interface and the effect of various factors are clarified through the discussion of the experimental phenomenon, load slip relationship and bond slip relationship. The result indicated that the effective bond length increases with the decrease of ECC matrix strength, when the effective bond length exceeds 550 mm or the ECC matrix strength is greater than E40, the failure load can be increased. In comparison with mortar composite layer, ECC composite layer tends to be relatively ductile. Finally, two of existing FRP-concrete interface models were selected to fit the experimental curve, and the fitting effect of existing FRP-concrete interface models is acceptable. Keywords: Interface · ECC · FRP · Concrete · Bond behavior · Single shear test

1 Introduction As a kind of material for reinforcement, carbon fiber reinforced polymer (CFRP) has a lots of advantages, such as light weight, high strength, and corrosion resistance. Nevertheless, what can’t be ignored is the disadvantage of its poor bonding performance with concrete. So many researchers use inorganic cement-based materials to replace epoxy resin, which is commonly used as a binder for FRP. Therefore, many studies on the FRP-concrete interface using inorganic cementitious materials as binders have begun to emerge, among which fabric-reinforced cementitious matrix (FRCM), textile-reinforced mortar (TRM) [1–3] and other reinforcements have emerged. This research direction produced a lots of achievement, including the interface of FRP strip-concrete [4], the interface of FRP woven mesh-concrete [5], etc. However, these materials tend to have © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 58–70, 2023. https://doi.org/10.1007/978-981-19-3983-9_6

Experimental Study on the Performance of FRP Grid

59

low stiffness and poor bonding performance, which leads to unsatisfactory reinforcement effects on the structure. A lot of researches have shown that CFRP grid has the characteristics of high rigidity and good stability, and has obvious advantages in the reinforcement of concrete structures [6]. At the same time, engineering cementitious composites (ECC) have the advantages of multiple cracks and strong adhesion to concrete [7–9]. According to the above summary, we can know that the composite layer formed by the CFRP grid embedded with ECC matrix (ECC composite layer) can provide a new idea for the reinforcement of structures in engineering. Therefore, this paper researched the bonding performance of the interface between ECC composite layer and concrete. The failure form and failure load are discussed firstly, and then the load-slip curve and bond-slip curve are analyzed, the experimental parameters including the bond length and the types of matrix as well as the strength of ECC matrix. Finally, the constitutive model of the interface between FRP and concrete was used to fit the experimental bond-slip curve. The results show that the model used in this paper is great, and provides a good empirical theory and research direction for the subsequent study of the bond performance of the ECC composite layer-concrete interface.

2 Experimental Study In this single shear test, the composite layer, bonded on the concrete block, as shown in Fig. 1. A total of 20 specimens were tested, and the research variables are shown in Table 1. 2.1 Materials The average cube strength and elastic modulus of the concrete is 25.5 MPa, 26.8 GPa. The average cube strength of mortar is 34.5 MPa, which is close to ECC within 60%

Fig. 1. Specimen dimension (units: mm).

60

W. Li et al.

fly ash. And configure ECC with different strengths by adjusting the content of fly ash [10], the details of ECC (unit in g/L) are listed in Table 2 as follows [11].

Fig. 2. CFRP grid.

The size of the CFRP grid is 50 mm × 50 mm, and the cross-sectional area of a single grid is 26.4 mm2 (See in Fig. 2). The test results show that the tensile strength of the CFRP grid is 1515 MPa and the elastic modulus is 114.8 GPa. 2.2 Specimen Preparation The dimensions of each part of the test specimens are shown in Fig. 1. The layout of the strain gauge can be seen from Fig. 3. It should be note that the pouring sequence is to place a layer of 10 mm thick matrix firstly, and them the grid is placed on the surface of the matrix, finally a layer of 10 mm thick matrix is covered.

Fig. 3. Details of strain gauge (units: mm).

2.3 Test Methods This single shear test adopts displacement-controlled loading, and the loading rate is 0.2 mm/min. As shown in Fig. 4. The machine applied load to the composite layer by clamping the end reinforcement sheet of grid. Paste three metal sheets (MS) and 5 displacement meters (DM) are used to measure the displacement, and several strain gauges (SG) can collect data of strain of grid.

Experimental Study on the Performance of FRP Grid

Fig. 4. Schematic of single shear test.

Table 1. Details of characterizations of specimens. Specimens

Matrix types

Bond length (mm)

Numbers of specimens

M-250

Mortar

250

5

M-350

350

M-450

450

M-550

550

M-650

650

E40-250

ECC containing 40% FA

250

E40-350

350

E40-450

450

E40-550

550

E40-650 E60-250

650 ECC containing 60% FA

250

E60-350

350

E60-450

450

E60-550

550

E60-650 E80-250

5

5

650 ECC containing 80% FA

250

E80-350

350

E80-450

450

E80-550

550

E60-650

650

5

61

62

W. Li et al. Table 2. Mixture proportions and properties of ECC.

Types

Cement

Quartz sand

Fly ash

Water

Water reducer

PE fiber

Ultimate compressive strength/MPa

Ultimate tensile strength/MPa

E40

800

E60

533.3

266.6

533.3

400

5

15

51.5

6.489

266.6

800

400

6

15

34.9

5.821

E80

266.6

266.6

1066.6

400

7.5

15

22.1

2.878

3 Result 3.1 Failure Mode The test results show that, regardless of ECC composite layer or the mortar composite layer, most of the failure modes of the specimens show the interfacial debonding (DB), and only the failure mode of specimen E80-650 shows grid fracture and interface is not debonded (FGB). By comparison, the mortar composite layer is brittle failure and irregular to follow, while the ECC composite layer is relatively ductile and its debonding process shows two different forms of sudden peeling and gradual debonding as the bond length increases. The load-slip curve is shown in Fig. 6. In a word, sudden peeling mainly occurs in the

(a)

(b)

(c)

(d) Fig. 5. Details of cracks (a, b and c are for ECC, d is for mortar).

Experimental Study on the Performance of FRP Grid 35 30

Load(kN)

25 20 15

E40-250 E40-350 E40-450 E40-550 E40-650

10 5 0 0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

S (mm)

(a)E40

35 30

Load(kN)

25 20 15

E60-250 E60-350 E60-450 E60-550 E60-650

10 5 0 0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

S (mm)

(b) E60

45 40 35

Load(kN)

30 25 20

E80-250 E80-350 E80-450 E80-550 E80-650

15 10 5

0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0

S (mm)

(c) E80

35 30

Load(kN)

25 20 15

M-250 M-350 M-450 M-550 M-650

10 5 0 0.0

0.5

1.0

1.5

2.0

2.5

S (mm)

3.0

3.5

4.0

4.5

(d) M

Fig. 6. Load-slip curve of specimens.

63

64

W. Li et al.

specimens with bond lengths of 250 mm and 350 mm. There are only a few cracks before the specimens are destroyed, and finally the interface suddenly deboned. Nevertheless, gradually debonding occurs in the specimens with bond lengths of 450 mm, 550 mm, and 650 mm. Its failure characteristics are mainly termed (1) the specimen is in the elastic stage and there is no obvious damage when the load is small in the early stage of the test; (2) As the load increases to larger, the upper surface of the composite layer loading end began to bring fine cracks (weft cracks) into being, which are perpendicular to the loading direction; (3) When the load continued to increase in the middle of the test, the longitudinal cracks will appear, the number of cracks gradually increases and develops toward the free end, and the composite layer peels off at the loading end partially so that the amount of interface slip increases; (4) In the latest stage, The width and number of cracks in the composite layer further expand and become stable. The range of local debonding extends to the free end, and the amount of interface slip is about to reach its peak. Finally, the entire composite layer is completely peeled and destroyed. Cracks can be seen in Fig. 5. 3.2 Failure Load The failure load of each specimen is summarized in Table 3, and the influence of factors such as the bond length, the strength of ECC matrix and the type of matrix are discussed. In view of the effect of bonding length, the failure load increases with the bond length of the composite layer increases. When the bond length of the composite layer exceeds the effective bond length, the failure load will not continue increase. Therefore, the effective bond length range of different matrix composite layers can be roughly determined. The effective bond length ranges of E40 and E60 specimens is 550 mm– 650 mm, and the effective bonding length of E80 specimens is greater than 650 mm, as is shown in Table 3. Among them, the composite layer of the matrix of E40, E60 and Mortar showed many decreases after the bond length exceeded 550 mm. The author believes that this is mainly due to the inevitable shrinkage cracks formed during curing [12]. When the strength of ECC matrix is different, it is found that the interface failure load does not change linearly with the decrease in the strength of the ECC matrix when the bond length exceeds 450 mm. In general, the greater strength of the ECC matrix results in the greater ultimate bond stress of the interface, so that the interface failure load will also increase. However, the test results show that the failure load of E60 group of specimens has decreased, while the failure load of E80 group of specimens has increased significantly. The reason is that the effective bond length of E80 specimens is affected by the strength of the ECC matrix, and it also can be said that the smaller the strength of ECC matrix, the longer effective bond length. So under the coupling effect of the maximum bond stress of interface and the effective bond length, the failure load of E80 specimens increased. The type of matrix is different, and the bonding performance is also different. As shown in Table 3, the failure load and the slip of ECC composite layer are significantly greater than that of the mortar composite layer. It can be seen that ECC composite layer matrix has significant advantages of high ductility and large bearing capacity of interface.

Experimental Study on the Performance of FRP Grid

65

2.5

Stress(MPa)

2.0

1.5

1.0

E40 E60 E80

0.5

0.0 0.00

0.25

0.50

0.75

1.00

1.25

S(mm)

(a) Bond-slip curve of E40, E60 and E80

(b) Each stage of bond -slip curve Fig. 7. Bond-slip curve.

Table 3. Test result. Specimens

Global slip/mm

Failure load/kN

Failure mode

Prediction of effective bond length

E40-250

1.24

17.4

DB

550 mm–650 mm

E40-350

1.78

24.8

DB

E40-450

2.46

17.0

DB

E40-550

3.37

33.0

DB

E40-650

3.99

30.5

DB

E60-250

1.79

19.1

DB

550 mm–650 mm (continued)

66

W. Li et al. Table 3. (continued)

Specimens

Global slip/mm

Failure load/kN

Failure mode

E60-350

1.27

18.1

DB

E60-450

2.43

18.4

DB

E60-550

3.09

21.19

DB

E60-650

4.03

22.5

DB

E80-250

1.59

22.8

DB

E80-350

1.36

17.4

DB

E80-450

2.7

28.1

DB

E80-550

4.5

40.7

DB

E80-650

5.5

43.3

FGB

M-250

0.5

11.4

DB

M-350

1.7

14.6

DB

M-450

2.2

23.3

DB

M-550

3.0

20.5

DB

M-650

3.5

17.5

DB

Prediction of effective bond length

≥650 mm

450 mm–550 mm

3.3 Bond-Slip Curve According to the results, the displacement value of the composite reinforcement layer at the free end is almost zero, so it can be ignored. The value of the strain gauge is calculated by differential interpolation to obtain the bond-slip curve. According to the shape of the curves, the bond-slip curves of E40, E60, and E80 specimens can be divided into three stages: elastic rising stage, plastic rise stage, and plastic fall stage, as shown in Fig. 7. Among them, the elastic rise stage curve is roughly linear, the bond stress of interfacial increases rapidly with the increase of slip, and the bonding stiffness of the interface is larger; The interface shear stress of the plastic rise stage is greater than the chemical bonding force between concrete and ECC, the growth rate of interface shear stress decreases gradually with the increase of slip, and the bonding stiffness of the interface gradually becomes smaller; When the shear stress peak is reached, the curve enters the plastic fall stage. At this time, the interface is damaged, it is also means that the aggregate on the interface between ECC and concrete wears, the shear stress of the interface decreases and the amount of slippage increases rapidly. The overall shape of the bond slip curve, meanwhile, is not affected by the strength of the ECC matrix, and before the composite layer is completely peeled off, a residual stress section similar to the FRCM bond slip curve appears at the end of the bond slip curve [4] This is mainly due to the mechanical occlusion of the ECC and concrete interface and the bridging effect of the fibers at the ECC matrix interface, and the bond-slip curve shows a certain degree of ductility.

Experimental Study on the Performance of FRP Grid

67

3.4 Comparision of Models Considering that there are few studies on prediction models related to the interface between ECC and concrete, but the prediction models related to the interface between FRP and concrete have a wide range of applications. Among them, the models proposed by He [13] and Wu [14] have great prediction effects, as seen Eq. 1 and Eq. 2. These two models are fitted to the experimental results, and the fitted bond-slip curve is consistent with the experimental results, as shown in Fig. 8. In the rising section of the curve, both the He model and the Wu model are in good agreement with the experimental values; In the falling section, the experimental data are basically distributed on both sides of the fitting curve of Wu model, while the fitting curve of He model has relatively poor accuracy. Table 4 lists the relevant fitting parameters and the fitting coefficients of each model. The average values of the fitting coefficients R2 of He model and Wu model are 0.818 and 0.965, respectively. It can be seen that using Wu model as the constitutive

Fig. 8. Fitting curve of bond-slip relationship.

68

W. Li et al.

model of bond-slip relationship between the ECC composite layer and concrete will be more accurate. He model: s n ) (1) τ (s) = τmax ( s0 (n − 1) + (s/s0 )n Wu model:

   −ωs −ωs  τmax (ω + eω − 1) C + 2e−ω − 1 ωs s0 + C − τ (s) =  · e × 1 − e s0 2 ω ω+e −1 s0 1 − e−ω (ω + Ceω + 1) (2) where n is a parameter that mainly controls the softening branch; ω and C are infinite constants; τmax is the average maximum bond stress; s0 are the slip at the maximum bond stress. Table 4. Fitting result. Specimens

τ max /MPa

s0 /mm

n

R2 (He model)

C

ω

R2 (Zhou model)

E40-250

2.1

0.33

1.701

0.945

1.152

0.441

0.983

E40-450

2.39

0.21

1.412

0.976

1.135

0.384

0.993

E40-550

2.78

0.27

1.575

0.767

1.103

0.362

0.954

E60-250

1.26

0.13

1.453

0.836

1.155

0.650

0.923

E60-350

1.21

0.10

1.386

0.917

1.198

0.681

0.995

E60-450

1.74

0.11

1.539

0.680

1.155

0.639

0.945

E80-350

0.80

0.10

2.955

0.933

1.112

0.944

0.995

E80-450

0.93

0.31

1.219

0.377

1.115

0.962

0.906

E80-550

1.10

0.12

2.166

0.938

1.303

0.874

0.994

Average value

0.818

0.965

4 Summary In this study, the bonding behavior of the interface between the FRP grid-reinforced ECC composite layer and concrete was studied by single shear test. The conclusions are as follows: (1) The debonding process of the interface between ECC composite layer and concrete is mainly divided into sudden peeling and gradual debonding as the bond length increases. Furthermore, the surface crack morphology of the ECC composite layer at the time of failure is not affected by factors such as the strength of ECC matrix, the types of ECC matrix, and the bond length.

Experimental Study on the Performance of FRP Grid

69

(2) The failure load of the interface between ECC composite layer and concrete is affected by the coupling of the effective bond length and the strength of ECC matrix. Under the condition that the effective bond length increases with the decrease of ECC matrix strength, when the effective bond length exceeds 550 mm or the ECC matrix strength is greater than E40, the failure load can be increased; (3) The test results show that the bond-slip curve of the interface between the FRP grid-ECC composite layer and the concrete is basically the same. It can be divided into three stages: elastic rising stage, plastic rising stage, and plastic falling stage. And fitting it with the bond-slip model of the interface between FRP and concrete can get a good fitting work.

Acknowledgment. The authors would like to appreciate the financial support provided by the National Natural Science Foundation of China (NSFC Grant Numbers: 51878415, 51908373). Technical support is acknowledged from Guangdong Provincial Key Laboratory of Durability for Marine Civil Engineering (SZU), No. 2020B1212060074.

References 1. Bournas, D.A., Lontou, P.V., Papanicolaou, C.G., Triantafillou, T.C.: Textile-reinforced mortar versus fiber-reinforced polymer confinement in reinforced concrete columns. ACI Struct. J. 104(6), 740 (2007) 2. ACI Committee 549: Guide to design and construction of externally bonded fabric-reinforced cementitious matrix (FRCM) systems for repair and strengthening concrete and masonry structures (2013). ACI 549.4R-13 3. Focacci, F., D’Antino, T., Carloni, C.: The role of the fiber–matrix interfacial properties on the tensile behavior of FRCM coupons. Constr. Build. Mater. 265, 120263 (2020) 4. Nelson, L.A., Al-Allaf, M., Weekes, L.: Analytical modelling of bond-slip failure between epoxy bonded FRP and concrete substrate. Compos. Struct. 251, 112596 (2020) 5. Jiang, J., Jiang, C., Li, B., Feng, P.: Bond behavior of basalt textile meshes in ultra-high ductility cementitious composites. Compos. B Eng. 174, 107022 (2019) 6. Liu, Z., Yue, Q., Li, R., Ding, Y.: FRP grid in civil engineering applications. In: The 9th National Symposium on FRP application in Construction Engineering (2015) 7. Li, V.C., Leung, C.K.Y.: Steady-state and multiple cracking of short random fiber composites. J. Eng. Mech. 118(11), 2246–2264 (1992) 8. Yang, X., Gao, W.Y., Dai, J.G., Lu, Z.D., Yu, K.Q.: Flexural strengthening of RC beams with CFRP grid-reinforced ECC matrix. Compos. Struct. 189, 9–26 (2018) 9. Zheng, Y.: Experiment and calculation method research on reinforced concrete (RC) beams strengthened with the composite of FRP grid and ECC. A dissertation submitted to Southeast University (2018) 10. Liu, C., Bi, Y., Hua, Y.: Mechanical properties of PVA-ECC with high volume fly ash and mechanism analysis of fly ash. Bull. Chin. Ceram. Soc. 36(11), 3739–3744 (2017) 11. Fiber reinforced polymer composite grids for civil engineering. GB/T 36262—2018 12. Li, M., Li, V.C.: Behavior of ECC/concrete layered repair system under drying shrinkage conditions/Das Verhalten eines geschichteten Instandsetzungssystems aus ECC und Beton unter der Einwirkung von Trocknungsschwinden. Restorat. Build. Monuments 12(2), 143– 160 (2006)

70

W. Li et al.

13. He, W., Wang, X., Ding, L., Wu, Z.: Experimental study on bond behavior of interface between fiber-reinforced polymer grids and concrete substrate. Compos. Struct. 257, 113164 (2021) 14. Zhou, Y.W., Wu, Y.F.: General model for constitutive relationships of concrete and its composite structures. Compos. Struct. 94(2), 580–592 (2012)

Structural Mechanics and Structural Engineering

Dynamic Strain Estimations of Beam Ends in Steel Moment-Resisting Frames Using Acceleration Data Xiyang Yu and Xiaohua Li(B) School of Civil Engineering, Chongqing University, Chongqing 400045, China [email protected]

Abstract. Dynamic-strain-based damage detection methods for fractured beams in steel moment-resisting frames have been proposed previously. However, it is impractical to install a dense-array strain sensing system in a real steel frame, while using acceleration measurements to estimate strain responses of the steel frame can make the collection of strain data more convenient. In this study, a linear-modelbased state observer (LMBO) was used to estimate dynamic strain responses of beam ends in a steel moment-resisting frame subjected to base excitations, and required measurements were only story accelerations. In the state observer, a steel frame experiencing base motions is equivalent to the frame with corrective forces and added grounded dampers. The effectiveness of this method was numerically investigated through a finite element model of a three-story steel moment-resisting frame in SAP 2000 software. The results show that when the input excitation was white noise or an earthquake motion recorded in the 1989 Loma Prieta earthquake, estimated strain responses were close to the corresponding simulated results, and the values of the Time Response Assurance Criterion (TRAC) between the both strain responses exceeded 0.8. Keywords: Dynamic strain estimation · State observer · Steel moment-resisting frame

1 Introduction After a major earthquake, it is essential to evaluate the post-earthquake safety of buildings, which will be helpful to decision making and re-occupancy [1]. For the steel moment-resisting frame, the earthquake-induced damages are more likely to occur at welded beam ends due to the strong-column and weak-beam design philosophy. For example, reconnaissance reports of the 1994 Northridge earthquake and the 1995 Kobe earthquake pointed out that a large number of steel moment-resisting frames suffered damages at welded beam ends [2–4]. Furthermore, this kind of structural damages will threaten the safety of steel buildings [5, 6]. Thus, accurate information of local damages (i.e., fractures near beam ends) of the steel moment-resisting frame will improve the reliability of the post-earthquake performance evaluation. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 73–86, 2023. https://doi.org/10.1007/978-981-19-3983-9_7

74

X. Yu and X. Li

For reasons such as limitations of manual inspections, many structural damage detection theories have been proposed in the past decades [7], such as model parameter-based methods [8, 9] and signal processing-based methods [10, 11]. For example, Li et al. proposed an approach based on time-frequency analysis and fractal dimension, which was successfully used to determinate locations of seismic damages in a ten-story shear-type building model [11]. Many of these methods are capable of determining variations in structural global vibration characteristics or further identifying damage locations, but it is challenging to quantify the damage severity. In recent years, Li (the author) et al. have conducted a series of studies to deal with the problem [12–15]. By comparing strain responses on steel beams under small-amplitude dynamic excitations before and after a major earthquake, Li et al. first proposed a dynamic-strain-based damage index [12] and developed simplified derivation of a damage curve for the seismically induced beam fractures [13], and then Li et al. presented a decoupling algorithm for evaluating multiple beam damages in a steel frame [14]. In the experimental investigations, the above methods determined the locations of damages at beam ends in a five-story steel moment-resisting frame and quantified the corresponding damage extents. It was worth noting that the strain data used in the methods proposed by Li et al. were measured under small-amplitude loads (e.g., ambient vibrations), and the steel frame was in the elastic stage. However, a dense-array strain sensing system is necessary for the dynamic-strainbased methods, which is difficult to meet in reality. With the progress of strong-motion instrumentation programs (SMIP) in seismically active regions, using acceleration data to estimate strain responses of the steel moment-resisting frame can make the collection of strain data more convenient, which is the main motivation of this study. In the field of strain estimation, Hernandez et al. proposed a linear-model-based state observer (LMBO) [16, 17], where a finite element model was utilized to estimate local strain responses of a structure in the elastic stage by using limited acceleration measurements. This method has been successfully applied in a seven-story full-scale reinforced concrete cantilever shear wall tested by a shaking table [18], and compared with the measured strain data, the accuracy of the estimated results of the method is significantly higher than that of the results from directly applying the base excitation to the finite element model of the shear wall. However, the application of this method in strain estimations of the steel moment-resisting frame has not been studied before. In this paper, the effectiveness of the method was numerically studied, thus laying the foundation for the subsequent experimental verification and dynamic-strain-based damage detection. Firstly, Sect. 2 presents the simplified derivation of the LMBO which is a feedback system combining a mathematical model with measured data [19]. Then in Sect. 3.1 and 3.2, a finite element model was established in SAP 2000 software [20] according to a three-story steel moment-resisting frame in the structural laboratory of the Disaster Prevention Research Institute, Kyoto University, Japan. In Sect. 3.3 and 3.4, the estimated strain responses at beam ends from the LMBO were compared with the simulated results from the original finite element model, when the input excitation was a stationary random process (banded white noise) or a nonstationary random process (an earthquake motion recorded in the 1989 Loma Prieta earthquake).

Dynamic Strain Estimations of Beam Ends in Steel Moment-Resisting Frames

75

2 Linear-Model-Based State Observer The detailed derivation of the LMBO method proposed in [16–18] is briefly summarized here for the sake of completeness. The building structure in the elastic stage is linear and time-invariant. In the state observer (i.e., LMBO method), a structure subjected to external excitations is equivalent to a structure with corrective forces and added grounded dampers. The motion equation of a building structure with n degrees of freedom is defined as: M x¨ (t) + C D x˙ (t) + Kx(t) = b1 u(t)

(1)

where M  Rn×n is a diagonal mass matrix; x(t)  Rn , x˙ (t)  Rn and x¨ (t)  Rn are the displacement, velocity and acceleration relative to the ground respectively; K  Rn×n is a stiffness matrix of the structure; u(t)  Rz is a matrix of excitations and z is the number of external excitations. The excitations are the equivalent loads related to seismic accelerations when the structure is subjected to earthquakes; b1  Rn×z is the spatial distribution of the excitations. In the LMBO method, assuming that the velocity measurements are available (in this paper, the velocity measurements are obtained from the acceleration measurements), the  estimated displacement x(t)  Rn can be given by the following equation (state observer): ¨ ˙    M x(t) + (C D + λT βλ) x(t) + K x(t) = λT β q˙ k (t)

(2)

˙ ¨   where x(t)  Rn and x(t)  Rn are the estimated velocity and acceleration relative to the r×n is a matrix which is decided by the locations of measurement points, ground; λ R and r is the number of measurement points; β Rr×r is a symmetric matrix which is free to select but affects the accuracy of the estimated results; q˙ k (t)  Rr is a matrix consisting of the velocity measurements. After comparing Eq. (1) with Eq. (2), it can be found that a structure subjected to a seismic load can be equivalent to the structure with corrective forces and added grounded dampers by using the LMBO. Figure 1 shows a shear model with corrective forces F = λT β q˙ k (t) and added grounded dampers C = λT βλ. Such a physical interpretation allows the state observer to be implemented in a finite element software and solved quickly, which is one of its significant advantages [16]. In order to obtain accurate estimation results, it is necessary to find the appropriate β. The trace of the covariance matrix of the displacement estimation error is set as the objective function L:      L = tr E (x(t) − x(t))(x(t) − x(t))T  +∞ = tr( 0 b1 Suu (w)bT1  ∗0 + Γ 0 λT βSvv (w)β T λΓ ∗0 )dw (3) −∞

and Γ 0 = (−Mw2 + C D iw + λT βλiw + K)−1

(4)

76

X. Yu and X. Li

where Suu (w) and Svv (w) are the power spectral density (PSD) matrices of the external excitation and measurement noise, respectively. Then the matrix β corresponding to the minimum value of L is the optimal solution. To simplify the process of solving for the optimal solution of β, let β = θ Ir , where I is an identity matrix. Thus, the selection of β can be expressed concisely as: min(L) = L(θoptimal ∈ R+ )

(5)

Fig. 1. Demonstration of the implementation of the LMBO method: (a) a shear model subjected to a seismic load, (b) a shear model with the LMBO

3 Numerical Example 3.1 Basic Information A finite element model of a three-story steel frame was established in SAP 2000 according to a three-story steel moment-resisting frame tested on a shaking table in the structural laboratory of the Disaster Prevention Research Institute, Kyoto University, Japan. Then the simulated acceleration measurements from the finite element model were used to estimate the strain responses of the bottom flanges at the beam ends, so as to numerically verify the effectiveness of the LMBO method in strain prediction of the steel frame in the elastic phase. The three-story frame specimen is shown in Fig. 2a. The steel material of the frame was SS400. It had two bays in X direction, 4 m in total, and one bay in Y direction, 1 m in total. The heights of the first to the third stories were 915 mm, 880 mm and 880 mm, with mass of 1700 kg, 1600 kg and 1600 kg respectively. Special design of the damage link at the beam end is shown in Fig. 2b. More design details about this steel frame can be found in [21, 22]. Figure 3 shows sensor deployment. In the numerical example, the locations of simulated acceleration measurements were the same as the

Dynamic Strain Estimations of Beam Ends in Steel Moment-Resisting Frames

77

locations of the real accelerometers, and the locations of simulated and estimated strain responses were at the beam ends (as shown in Fig. 3). Moreover, sampling frequencies of all measurements were 100 Hz. Due to the symmetry of the steel frame, the strain responses at the 1st to 12th measurement points (as shown in Fig. 3) were estimated in this paper. It was worth noting that the input excitations were all in the X direction and only the measurement data in the X direction were considered because the frame was rigid (moment-resisting) in the X direction.

Fig. 2. Shaking table test of the three-story steel moment-resisting frame: (a) overview, (b) beam– column joint

Fig. 3. Deployment of sensors and measurement points

3.2 Model Verification Figure 4 shows the finite element model of the three-story steel frame simulated by using SAP 2000. The beams and columns consist of linear elastic frame elements. The beams

78

X. Yu and X. Li

and columns are rigidly connected. The material properties were selected as follows: Young modulus E = 200.5 GPa and Poisson ratio v = 0.3. Due to the special design of the damage link, it was replaced by I-shaped cross section with the same section modulus in bending, and the section modulus of an undamaged link is 8667 mm3 [21]. In addition, all structural damping ratios were classically taken to be 5%. The simulated results of the finite element model and the measured results of the shaking table test were compared from the aspects of frequency domain and time domain to verify the feasibility of finite element model. In the frequency domain, the first three structural frequencies in the X direction identified from three types of data: (1) Measured acceleration responses. They were measured from the real steel frame subjected to a white noise base excitation (with vibration frequency band of 0–50 Hz and effective value of 7 gal), and then structural frequencies in the X direction were obtained by using a frequency domain decomposition method. The frequencies from the first to third orders were 4.364 Hz, 10.240 Hz, and 11.550 Hz respectively; (2) Results of modal analysis. In SAP 2000, the modal analysis of the finite element model was carried out to obtain the three natural frequency in the X direction, and the corresponding results were 4.509 Hz, 9.915 Hz, 11.636 Hz; (3) Simulated acceleration responses. The same white noise excitation as the vibration test was applied to the finite element model in SAP 2000, and then the simulated acceleration of the third floor was decomposed in frequency domain to obtain the structural frequency. The corresponding results were 4.382 Hz, 9.192 Hz, 12.570 Hz. Based on the frequencies obtained from the measured acceleration (i.e., 4.364 Hz, 10.240 Hz, and 11.550 Hz), the differences of the same order frequencies in the three types of data were less than 3.5%, 10.3% and 8.9% respectively. In the time domain, Fig. 5 compares the measured and simulated absolute accelerations of the third story when the Kobe-30% seismic wave (i.e., the amplitude of the input wave is 30% of the original wave recorded in the Kobe earthquake, and the peak acceleration is 2.53 m/s2 ) was input to the real specimen and the finite element model. It could be found that the both time histories were relatively consistent. Thus, the finite element model is verified in linear range with measurement results.

Fig. 4. Finite element model of three-story steel frame

Dynamic Strain Estimations of Beam Ends in Steel Moment-Resisting Frames

79

Fig. 5. Absolute acceleration response of 3rd story under Kobe-30%

3.3 White Noise Base Excitation A banded white noise base excitation (with vibration frequency band of 0.2–30 Hz and a sampling frequency of 100 Hz) was input to the finite element model. This kind of load could be regarded as a stationary random process. A white noise process with a 10% root mean square (RMS) noise-to-signal value was superimposed on the three simulated acceleration responses of the third story (three accelerometers at the third story), and then the processed acceleration data were regarded as measurement data. In order to obtain a suitable feedback matrix β, the three-story frame was simplified as a time-invariant shear model with lumped mass. And structural parameters of this shear model were the same as these of the finite element model. The inter-story stiffnesses of the first to third stories were 15900 kN/m, 22300 kN/m and 10700 kN/m respectively. Mass matrix M = diag (1700 kg, 1600 kg,1600 kg). Meanwhile, it was assumed that structural damping matrix CD was formulated using the Rayleigh damping model: CD =

2ξ ω1 ω2 2ξ M+ K ω1 + ω2 ω1 + ω2

(6)

where ω1 and ω2 = first two natural frequencies of the structure; ξ = 5%. In this study, the velocity signals were integrated from the accelerations and then applied to a highpass filter with a 0.2 Hz cutoff frequency. The power spectral density (PSD) matrices of the external excitation and measurement noise were given by Suu (w) = diag (35.08, 31.05, 31.05) and Svv (w) = 8.02 × 10−7 I1 . In this case, an optimal value was θ optimal = 2550 Ns/m with L = 2.11 × 10–9 . Figure 6 shows the relationships between L and θ when the PSD of measurement noise was Svv , 2Svv or 3Svv . A suboptimal value of θ was 1350 Ns/m when the PSD of measurement noise was 2Svv , and it yielded a value of L = 2.21 × 10–9 which was closed to the optimal. More importantly, as can be seen from Fig. 6, this suboptimal θ was still valid under stronger measurement noise. Thus, θ = 1350 Ns/m was selected to implement the LMBO method. In addition, the damper value (i.e., θ ) was divided by three and assigned to the added grounded dampers at the locations of accelerometers of the third story in the X direction. The dynamic responses at the beam ends, including the bending moment and axial force, could be obtained from the finite element model through SAP 2000 software. The strain responses of the bottom flange at the beam ends were calculated by using material mechanics and then applied to a band-pass filter (0.4–15 Hz). Due to limited space, Figs. 7, 8, 9 and 10 compare the simulated and estimated strain responses at the first to

80

X. Yu and X. Li

Fig. 6. Relationships between L and θ

Fig. 7. Simulated and estimated strains at 1st measurement point: (a) time history, (b) detail time history, (c) frequency domain

fourth measurement points, and it could be found that they were relatively consistent in both time and frequency domains. Figure 11 compares the values of the Time Response Assurance Criterion (TRAC) [23, 24] between the simulated and estimated strains at all twelve measurement points. TRAC =

[{Strains (t)}T {Straine (t)}]2 [{Strains (t)} {Strains (t)}][{Straine (t)}T {Straine (t)}] T

(7)

where {Strains (t)} = time history of the simulated strain signal; {Straine (t)} = time history of the estimated strain signal. The value of TRAC is between zero and unity, and the closer it is to unity, the higher the agreement between the two signals. It could be found from Fig. 11 that the values of TRAC at all twelve measurement points exceeded 0.8, which indicted a high agreement between the simulated and estimated strains. Thus, in linear elastic range, the LMBO method was capable of estimating strain responses at beam ends in the steel moment-resisting frame when the input excitation was a stationary random process. 3.4 Earthquake Excitation An earthquake motion recorded in the 1989 Loma Prieta earthquake was input to the finite element model. This kind of load could be regarded as a nonstationary random

Dynamic Strain Estimations of Beam Ends in Steel Moment-Resisting Frames

81

Fig. 8. Simulated and estimated strains at 2nd measurement point: (a) time history, (b) detail time history, (c) frequency domain

Fig. 9. Simulated and estimated strains at 3rd measurement point: (a) time history, (b) detail time history, (c) frequency domain

82

X. Yu and X. Li

Fig. 10. Simulated and estimated strains at 4th measurement point: (a) time history, (b) detail time history, (c) frequency domain

Fig. 11. Values of TRAC at all twelve measurement points

process. A white noise process with a 10% root mean square (RMS) noise-to-signal value was superimposed on the three simulated acceleration responses of the third story, and then the processed acceleration data were regarded as the measurement data. The velocity signals were obtained by using the same data processing method as Sect. 3.3. Through the time-invariant shear model in Sect. 3.3, a value of θ = 1100 Ns/m was selected to implement the LMBO method. It was worth noting that the strain responses were calculated by using material mechanics and then applied to a band-pass filter (2.5–15 Hz). Figures 12, 13, 14 and 15 show the simulated and estimated strain responses at the first to fourth measurement points, and two kinds of signals were also relatively consistent in both time and frequency domains. Figure 16 shows the values of TRAC, and they all exceeded 0.83, which indicted a high agreement between the simulated and estimated strains. Thus, in linear elastic range, the LMBO method was also capable of estimating strains at beam ends in the steel moment-resisting frame when the input excitation was a nonstationary random process.

Dynamic Strain Estimations of Beam Ends in Steel Moment-Resisting Frames

83

Fig. 12. Simulated and estimated strains at 1st measurement point: (a) time history, (b) detail time history, (c) frequency domain

Fig. 13. Simulated and estimated strains at 2nd measurement point: (a) time history, (b) detail time history, (c) frequency domain

84

X. Yu and X. Li

Fig. 14. Simulated and estimated strains at 3rd measurement point: (a) time history, (b) detail time history, (c) frequency domain

Fig. 15. Simulated and estimated strains at 4th measurement point: (a) time history, (b) detail time history, (c) frequency domain

Fig. 16. Values of TRAC at all twelve measurement points

Dynamic Strain Estimations of Beam Ends in Steel Moment-Resisting Frames

85

4 Conclusion This paper presents the use of a linear-model-based state observer (LMBO) to estimate the strain responses at beam ends in a steel moment-resisting frame. Although this method has been proposed before, it was the first time to use it in estimating strain responses of the steel moment-resisting frame. The effectiveness of the LMBO method was numerically verified through a finite element model of a steel frame in the SAP 2000 software. The results show that when the input excitation was a stationary random process (banded white noise) or a nonstationary random process (an earthquake motion recorded in the 1989 Loma Prieta earthquake), estimated strain responses were close to the corresponding simulated results, and the values of Time Response Assurance Criterion (TRAC) between the both strain responses exceeded 0.8.

5 Future Work Further research is required for the strain estimation and damage detection of beam ends in steel moment-resisting frames. The future work is as follows: (1) It is necessary to find a better method to obtain the suitable value of θ; (2) The effectiveness of the LMBO method in strain estimation will be experimentally verified; (3) The LMBO method will be combined with the dynamic-strain-based damage detection method proposed by the author Li in order to identify and quantify seismic damages at the beam ends of steel frames. Acknowledgments. This work was supported by the National Natural Science Foundation of China (Grant No. 52178454) and Natural Science Foundation of Chongqing (Grant No. cstc2019jcyj-msxmX0254).

References 1. Celebi, M., et al.: Real-time seismic monitoring needs of a building owner—and the solution: a cooperative effort. Earthq. Spectra 20(2), 333–346 (2004) 2. Nakashima, M.: Reconnaissance report on damage to steel buildings structures observed from the 1995 Hyogoken-Nanbu (Hanshin/Awaji) earthquake, Abridged English edition. Steel Committee of Kinki Branch, the Architectural Institute of Japan (AIJ) (1995) 3. Youssef, N.F.G., Bonowitz, D., Gross, J.L.: A survey of steel moment-resisting frame buildings affected by the 1994 Northridge earthquake. US National Institute of Standards and Technology (1995) 4. Mahin, S.A.: Lessons from damage to steel buildings during the Northridge earthquake. Eng. Struct. 20(4–6), 261–270 (1998) 5. Rodgers, J.E., Mahin, S.A.: Effects of connection fractures on global behavior of steel moment frames subjected to earthquakes. J. Struct. Eng. 132(1), 78–88 (2006) 6. Lignos, D.G., et al.: Numerical and experimental evaluation of seismic capacity of high-rise steel buildings subjected to long duration earthquakes. Comput. Struct. 89(11–12), 959–967 (2011) 7. Hou, R., Xia, Y.: Review on the new development of vibration-based damage identification for civil engineering structures: 2010–2019. J. Sound Vib. 491, 115741 (2021)

86

X. Yu and X. Li

8. Yoon, M.K., Heider, D., Gillespie, J.W., et al.: Local damage detection with the global fitting method using operating deflection shape data. J. Nondestr. Eval. 29(1), 25–37 (2010) 9. Caddemi, S., Caliò, I.: The exact explicit dynamic stiffness matrix of multi-cracked EulerBernoulli beam and applications to damaged frame structures. J. Sound Vib. 332(12), 3049– 3063 (2013) 10. Huang, N.E.: The Hilbert-Huang Transform in Engineering. Taylor and Francis Group, New York (2005) 11. Li, H., Tao, D., Huang, Y., et al.: A data-driven approach for seismic damage detection of shear-type building structures using the fractal dimension of time–frequency features. Struct. Control. Health Monit. 20(9), 1191–1210 (2013) 12. Li, X., Kurata, M., Nakashima, M.: Evaluating damage extent of fractured beams in steel moment-resisting frames using dynamic strain responses. Earthq. Eng. Struct. Dyn. 44(4), 563–581 (2015) 13. Li, X., Kurata, M., Nakashima, M.: Simplified derivation of a damage curve for seismically induced beam fractures in steel moment-resisting frames. J. Struct. Eng. 142(6), 04016019 (2016) 14. Li, X., Kurata, M., Suzuki, A.: Decoupling algorithm for evaluating multiple beam damages in steel moment-resisting frames. Earthq. Eng. Struct. Dyn. 46(7), 1045–1064 (2017) 15. Suzuki, A., Kurata, M., Li, X., et al.: Residual structural capacity evaluation of steel momentresisting frames with dynamic-strain-based model updating method. Earthq. Eng. Struct. Dyn. 46(11), 1791–1810 (2017) 16. Hernandez, E.M.: A natural observer for optimal state estimation in second order linear structural systems. Mech. Syst. Signal Process. 25(8), 2938–2947 (2011) 17. Erazo, K., Hernandez, E.M.: A model-based observer for state and stress estimation in structural and mechanical systems: experimental validation. Mech. Syst. Signal Process. 43(1–2), 141–152 (2014) 18. Erazo, K., Hernandez, E.M.: High-resolution seismic monitoring of instrumented buildings using a model-based state observer. Earthq. Eng. Struct. Dyn. 45(15), 2513–2531 (2016) 19. Roohi, M., Hernandez, E.M., Rosowsky, D.: Nonlinear seismic response reconstruction and performance assessment of instrumented wood-frame buildings—validation using NEESWood Capstone full-scale tests. Struct. Control. Health Monit. 26(9), e2373 (2019) 20. Wilson, E.L., Habibullah, A.: SAP 2000 Software, Version 21. Computer and Structures, Inc. (CSI), Berkeley (2019) 21. Nishino, H.: Evaluation of residual seismic performance of steel buildings using local damage identification sensing and model updating. Kyoto University, Kyoto (2014) 22. Matarazzo, T.J., Kurata, M., Nishino, H., et al.: Postearthquake strength assessment of steel moment-resisting frame with multiple beam-column fractures using local monitoring data. J. Struct. Eng. 144(2), 04017217 (2018) 23. Avitabile, P., Pingle, P.: Prediction of full field dynamic strain from limited sets of measured data. Shock. Vib. 19(5), 765–785 (2012) 24. Maes, K., Iliopoulos, A., Weijtjens, W., et al.: Dynamic strain estimation for fatigue assessment of an offshore monopile wind turbine using filtering and modal expansion algorithms. Mech. Syst. Signal Process. 76, 592–611 (2016)

Studies on the Relationship Between Anchor Force of Prestressed Anchor Cable and Nonlinear Vibration of Anchor Head Hao Li1(B) and Hui Cao1,2 1 School of Civil Engineering, Chongqing University, Chongqing 400045, China

[email protected] 2 MOE Key Lab of New Technology for Construction of Cities in Mountain Area, Chongqing

University, Chongqing 400045, China

Abstract. In order to ensure the normal operation of prestressed anchor cable, the detection of anchor force is a very important link in the health detection of prestressed anchor cable. By analyzing the relationship between nonlinear vibration characteristics and anchoring force, this paper explores the nondestructive testing method of anchoring force. The concrete workbench was used to simulate the soil anchored by the anchor cable. Under the condition of maintaining the tension at all levels by the pull-out apparatus, the lateral excitation was carried out at the anchor head of the anchor cable, and the free vibration acceleration signal of this part was recorded. The first-order frequency signal was decomposed by AMD, and Hilbert transform was performed on it to obtain the relationship between the frequency, amplitude and time of the signal, and then the frequencyamplitude curve reflecting the nonlinear vibration characteristics was obtained. The same amplitude segment is selected on the frequency-amplitude curve, and the corresponding frequency difference is used as the nonlinear index to analyze the relationship between the tension (anchoring force) and the nonlinear vibration degree of each level. The test and analysis results show that the degree of nonlinear vibration decreases with the decrease of anchorage force. Therefore, the nonlinear index can be used to judge the anchoring force of anchor cable. Keywords: Prestressed anchor cable · Nonlinear vibration characteristics · Anchoring force · Nondestructive testing · AMD decomposition

1 Introduction Prestressed anchor cable is widely used in practical engineering. In the long-term work of the anchor cable, the cable relaxation, soil creep and other unfavorable factors are easy to reduce the anchorage force of the anchor cable, which leads to anchorage failure. Anchoring force is an important index to evaluate the supporting quality of anchorage system, so the anchorage force detection of prestressed anchor cable is a very important subject.

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 87–97, 2023. https://doi.org/10.1007/978-981-19-3983-9_8

88

H. Li and H. Cao

At present, the detection of anchorage force of prestressed anchor cable mainly adopts mechanical or drawing method. Nondestructive testing methods, such as vibration signal analysis of anchorage force research is relatively less. Dang et al. [1] proposed and verified that the PCA-based algorithm can filter the influence of ambient temperature on the impedance monitoring of anchorage zone of anchor cable, and found that the impedance sensor arranged near the anchor head had better effect on the damage monitoring of anchorage zone. Huang et al. [2] verified the reliability of the nondestructive testing instrument based on the stress wave method by testing the anchorage quality of the bolt and anchor cable in the project. Based on the understanding that the stiffness of the structure changes with the amplitude before and after the damage, many scholars use nonlinear vibration characteristics to study the nondestructive testing of the structure. Spalvier [3] proposed torsionalvibration testing technology to characterize the nonlinear of concrete structure to determine the stress state of the structure. Alnuaimi [4] compared the linear and nonlinear ultrasonic methods to analyze the internal damage of concrete structures, and proved that the nonlinear ultrasonic parameters were more sensitive to the changes in the internal structure of materials. Wan et al. [5] used nonlinear characteristics to detect gas on nano-beam surface. Cao et al. [6, 7] studied the identification of debonding by analyzing the nonlinear characteristics of CFST. At the anchor head of the anchor cable, the anchoring force makes the anchor plate closely contact with the concrete anchor pier. Under transverse excitation, the interface effect similar to the crack of concrete beam exists on the interface between the anchor plate and the concrete pier, which makes the vibration of the anchor head present nonlinear vibration characteristics. In the case of a certain roughness of the interface, the interface effect is mainly affected by the size of the anchorage force. The larger the anchorage force, the closer the interface contact, the more obvious the interface effect and the stronger the nonlinear vibration characteristics. Therefore, it is worth studying to analyze the nonlinear vibration characteristics of the anchor head so as to realize the anchorage force detection. In this paper, the dynamic detection of prestressed anchor cable specimens under different working conditions is carried out, and the relationship between the nonlinear vibration characteristics of prestressed anchor cable and the anchoring force is analyzed to explore the feasibility of non-destructive testing of anchor cable anchoring force based on nonlinear vibration characteristics.

2 Principle and Method 2.1 Relationship Between Nonlinear Vibration Characteristics and Anchoring Force The so-called interface effect refers to the micro-viscosity, micro-collision and micro-slip on the interface during vibration. Heller et al. [8] analyzed the interface effect between the metal sheets tightened by two bolts. With the increase of bolt loosening, the interface effect between the metal sheets decreased, and the nonlinear vibration characteristics caused by the interface effect also weakened.

Studies on the Relationship Between Anchor Force

89

Similar to the relationship between the bolt tightness and the interface effect of the metal sheet, the interface effect between the anchor plate and the concrete anchor pier of the anchor cable will also be affected by the anchoring force. It is expected that with the decrease of the anchoring force, the interface effect decreases and the nonlinear characteristics weaken. 2.2 AMD and Hilbert Transform The free vibration signal of the anchor head of prestressed anchor cable under external excitation has multiple frequencies. Firstly, improved AMD [9] (analytical mode decomposition) is used to decompose the first-order frequency component from the free vibration signal, and then Hilbert transform is performed on the decomposed component to obtain the frequency-amplitude curve, and then the nonlinear degree of the signal is analyzed. AMD constructs the orthogonal function of specific frequency by setting the binary cutoff frequency parameter, and conducts Hilbert transform on the product of it and signal, so as to filter out the components less than specific frequency [10]. AMD sets different binary cutoff frequency parameters to decompose each single-component signal in the original signal. For the mixed signal x(t), it is composed of n single-component signals n  d xid (t). xi (t)(i = 1, 2, · · · , n), whose expression is: x(t) = i=1

The frequency of each single component ωi (t)(i = 1, 2, · · · , n) satisfies the following: |ω1 (t)| < ωb1 (t) < |ω2 (t)| < · · · < ωb(n−2) (t) < |ωn−1 (t)| < ωb(n−1) (t) < |ωn (t)| (1) where ωbi (t) ∈ (ωi (t), ωi+1 (t))(i = 1, 2, · · · , n − 1) is the binary cut-off frequency parameter set. By setting this parameter, each signal component of the original signal can be decomposed, as shown in Formula (2):  d xi (t) = si (t), i = 1 (2) xid (t) = si (t) − si−1 (t), i = 2, · · · , n For non-stationary signals with time-varying frequency, si (t) can be calculated by Formula (3):           si (t) = sin θbi (t) H x(t) cos θbi (t) − cos θbi (t) H x(t) sin θbi (t) (i = 1, 2, · · · , n − 1)

where θ bi (t) is the phase angle; H {·} is Hilbert transform.

(3)

90

H. Li and H. Cao

For real signal x(t), the Hilbert transformation is as follows:  1 x(τ ) 1 = dτ H [x(t)] = x(t) ∗ πt π R t−τ

(4)

Hilbert transform can get the amplitude and frequency of x(t), which can be calculated by the following formula: ⎧

⎪ ⎨ A(t) = x2 (t) + H 2 [x(t)]   (5) 1 d θ (t) 1 d H [x(t)] ⎪ = arctan ⎩ f (t) = 2π dt 2π dt x(t) The relationship between the amplitude, frequency and time of the signal can be obtained by Formula (5). 2.3 Calculation of Nonlinear Vibration Characteristics Compared with the high-order frequency component, the first-order frequency component is easily excited under external excitation, and is less affected by environmental noise. Therefore, according to the measured acceleration signal, the spectrum map is used to identify the binary cut-off frequency for AMD decomposition, and the first-order frequency component is separated. The first-order frequency component is integrated twice to obtain the displacement signal. The relationship between the amplitude, frequency and time of the signal is obtained by Hilbert transform. Using the frequency variation corresponding to the same amplitude segment to quantify the nonlinear vibration characteristics of the specimen, the nonlinear index of the specimen can be obtained.

3 Dynamic Test of Anchor Cable 3.1 Design and Fabrication of Specimens A 700 mm × 700 mm × 300 mm concrete block (hereinafter referred to as the workbench) is used to simulate the soil working with the anchor cable, and the PVC pipe is used in the center to retain a 30 mm diameter through the hole along the length direction. The steel strand with a nominal diameter of 15.2 mm is used as the anchor cable through the channel of the center of the workbench. The section of the specimen is shown in Fig. 1, the specific parameters are shown in Table 1, and the pouring and working state of the workbench are shown in Fig. 2. Considering two different anchor cable conditions, two specimens are designed: specimen 1 does not deal with, simulate the anchor head in good condition of prestressed anchor cable; the anchor plate of specimen 2 was corroded, and the prestressed anchor cable with serious corrosion of anchor head was simulated. The other parameters of the two specimens were the same.

Studies on the Relationship Between Anchor Force

91

The workbench is firmly connected to the laboratory floor through four anchor bolts, as shown in Fig. 2(b). The concrete workbench has large quality and is closely connected with the ground anchorage. By knocking on the specimen, the measured vibration signal is analyzed, and it is found that the vibration amplitude of the workbench can be ignored.

Fig. 1. The concrete workbench design

Table 1. Specimen parameters Specimen

Parameters

Size of workbench (Length × width × height /mm)

700 × 700 × 300

Concrete grade

C40

Nominal diameter of steel strand (mm)

15.2

Penetration channel diameter (mm)

30

Length of steel strand (mm)

200 + 700 + 600

Standard value of tensile strength of steel strand (Mpa)

1860

Design value of tensile strength of steel strand (MPa)

1320

Size of anchor plate (Length × width × thickness /mm)

120 × 120 × 20

Measuring range of acceleration sensor (g)

±5

Frequency response range of acceleration sensor (Hz)

0.2–1600

Axial sensitivity of acceleration sensor (mV/g)

1006

(a)Acceleration sensor under anchor plate

(b)Pull-out apparatus in work

Fig. 2. Drawing of specimen making and working state

92

H. Li and H. Cao

3.2 Experimental Process The maximum working tension of the steel strand in this test was calculated according to the domestic and foreign specifications, and the maximum tension was set to 100 kN considering the limitation of the tonnage of the drawing instrument for the test. The specific test steps are as follows: (1) The pull-out apparatus is used to load the load step by step, 10 kN is the first level, until 100 kN, each load is maintained for 5 min; (2) Under each level of load, the hammer imposes transverse impact load on the upper part of the anchor plate at the anchorage end. The free vibration acceleration signal is collected by the acceleration sensor attached to the bottom of the anchor plate, and the sampling frequency is 5000 Hz. (3) When the load reaches 100 kN, the load is unloaded step by step, and each step is kept for 5 min. The force hammer is used to excite the anchor plate, and the acceleration signal is collected. The sampling frequency is 5000 Hz. The test steps of the two specimens are the same.

4 Analysis of Relationship Between Nonlinear Vibration Characteristics and Tension 4.1 Nonlinear Index Taking the signal of a specimen1 under 50 KN tension as an example, the calculation of signal nonlinear index is introduced. For the measured signal, according to the method described in Sect. 2.3, the firstorder frequency component is separated from the signal and the displacement signal is obtained by integration. The frequency-time curve and amplitude-time curve are obtained by Hilbert transform. The least square method is used to fit the frequencyamplitude curve according to the time correspondence. The curves were normalized, that is, the values of each ordinate of the curve were subtracted from their maximum values, and the frequency-amplitude curves after normalization at all levels of tension in the loading stage of specimen 1 were plotted as shown in Fig. 3. It can be seen in the figure that the first-order frequency decreases with the increase of amplitude at all levels of tension, showing nonlinear vibration characteristics. But the amplitude of the firstorder frequency reduction given by these curves is different. In order to quantitatively compare the nonlinear vibration characteristics of specimens under different tension levels, the absolute value of frequency variation corresponding to the same amplitude section on the frequency-amplitude curve can be taken as a nonlinear index. The larger the nonlinear index is, the stronger the nonlinear characteristics are.

Studies on the Relationship Between Anchor Force

93

It can be seen from Fig. 3 that when the tension increases from 30 kN to 40 kN, the amplitude-frequency curve decreases significantly. The analysis shows that when the tension is less than a certain value (30 kN), the asperity between the contact surface has a certain extrusion, showing interface effect, resulting in nonlinear vibration. When the tensile force exceeds this value, the contact surface is compacted, and the untouched asperities also turn to contact state. The interface effect is significantly enhanced, and the nonlinearity is significantly increased.

Fig. 3. The specimen 1 frequency-amplitude curve

4.2 Relationship Between Nonlinear Index and Tension The frequency-amplitude curve of the first-order vibration component of the two specimens under different tensile forces is taken as the same amplitude segment (0.1–4) × 10–7 m, and the corresponding nonlinear index is calculated. The relationship between the nonlinear indexes of specimen 1 and specimen 2 and the tensile force is shown in Fig. 4 and Fig. 5. It can be seen that the tensile force of the two specimens is positively correlated with the nonlinear index, and the nonlinear index of specimen 2 is generally greater than that of specimen 1.

94

H. Li and H. Cao

Fig. 4. The specimen 1 trend graph of nonlinear index

Fig. 5. The specimen 2 trend graph of nonlinear index

In order to investigate the variation of nonlinear index with tension in detail, the state of tension at 10kN is taken as the initial state, and the nonlinear index f 1 is the reference value. Define the percentage of nonlinear index corresponding to different tension: δ i = f i /f 1 , reflect the change degree of nonlinear index and initial state of different tension; define nonlinear index percentage difference: i = δ i − δ i-1 , reflecting the speed of nonlinear index change; similarly, the percentage of non-linear indicators under unloading condition is defined: δ u = f u /f 1 , and the percentage difference of nonlinear indicators is defined: u = δ u − δ u-1 . The specific calculation results are shown in Table 2 and Table 3.

Studies on the Relationship Between Anchor Force

95

Table 2. The specimen 1 comparison table of nonlinear index Tensioning force P/kN

Loading stage

Unloading stage

Nonlinear index f i

Nonlinear index percentage δi

Percentage Nonlinear difference index f u of nonlinear index i

Nonlinear index percentage δu

Nonlinear index percentage difference u

10

0.408

100.0%

0.328

100.0%

20

0.482

118.1%

18.1%

0.513

156.4%

56.4%

30

0.658

161.3%

43.1%

0.587

179.0%

22.6%

40

1.132

277.5%

116.2%

0.764

232.9%

54.0%

50

1.249

306.1%

28.7%

0.936

285.4%

52.4%

60

1.475

361.5%

55.4%

1.371

418.0%

132.6%

70

1.513

370.8%

9.3%

1.626

495.7%

77.7%

80

1.626

398.5%

27.7%

1.768

539.0%

43.3%

90

1.783

436.5%

38.0%

1.781

543.6%

4.6%

100

1.855

454.7%

18.1%

1.855

565.5%

22.0%

Table 3. The specimen 2 comparison table of nonlinear index P/kN

Loading stage fi

δi

10

0.541

100.0%

20

0.788

145.6%

30

1.195

40

1.371

50 60

Unloading stage i

fu

δu

u

0.513

100.0%

45.6%

0.853

166.3%

66.3%

220.9%

75.3%

0.980

191.1%

24.8%

253.4%

32.5%

1.175

229.2%

38.0%

1.508

278.6%

25.3%

1.356

264.5%

35.4%

1.582

292.3%

13.7%

1.456

283.9%

19.4%

70

1.731

319.9%

27.5%

1.711

333.7%

49.8%

80

1.755

324.4%

4.5%

1.700

331.4%

−2.2%

90

1.793

331.3%

7.0%

1.814

353.7%

22.3%

100

1.828

337.9%

6.5%

1.828

356.5%

2.8%

The following conclusions can be drawn from the above chart: (1) The change trends of nonlinear indexes of specimens 1 and 2 are similar, and the tension is positively correlated with the nonlinear indexes. This is because the

96

H. Li and H. Cao

greater the tensile force, the closer the contact between the anchor plate and the concrete workbench, the stronger the interface effect; (2) With the increase of tensile force, the growth rate of nonlinear index slows down. This is because when the tension is large, the increase of contact tightness slows down; (3) The nonlinear index of specimen 2 is generally greater than that of specimen 1. This is because the corrosion surface roughness of the specimen 2 anchor plate is greater than that of the anchor plate without corrosion, which enhances the interface effect and increases the degree of nonlinear vibration. In summary, the nonlinear index of steel strand (prestressed anchor cable) always keeps the same direction as the tensile force (anchorage force), that is, the nonlinear index increases with the increase of tensile force (loading stage), and decreases with the decrease of tensile force(unloading stage). Therefore, the nonlinear index can be used as an index to determine the anchoring force of anchor cable. The corrosion of the anchor head has an increasing effect on the nonlinear index, which can be appropriately considered when estimating the anchorage force.

5 Conclusion By analyzing the relationship between anchorage force and nonlinear vibration characteristics, this paper explores the nondestructive testing method of anchorage force. The prestressed anchor cable is simulated by the steel strand through the concrete worktable. The free vibration acceleration signal of the anchor head of the anchor cable is collected under different tensile forces. The first-order frequency component is obtained by AMD decomposition, and the nonlinear vibration index of the anchor cable is obtained by Hilbert transform. The relationship between the nonlinear index of the anchor cable and the tensile force is studied. The following conclusions are obtained. (1) On the interface between the anchor plate of the prestressed anchor cable and the concrete workbench, the interface effect is obvious, and the lateral free vibration signal of the anchor plate has obvious nonlinear characteristics. Using AMD and Hilbert transform, the index reflecting the degree of nonlinearity can be successfully obtained. (2) The nonlinear index of prestressed anchor cable is positively correlated with the tensile force, and the nonlinear index is sensitive to the contact surface state. When the anchor head corrosion is large, the nonlinear index is larger. (3) If only considering the contribution of the interface effect of the anchor head to the nonlinear vibration, the size of the anchorage force of the anchor cable can be judged by using the nonlinear index as the basis and considering the corrosion condition of the anchor head. Nondestructive testing of anchorage force by nonlinear index is a feasible idea. Considering the corrosion degree of anchor head and the influence of field measured noise, further work is needed to study the non-destructive testing method of anchor cable anchorage force based on the practical application of this paper.

Studies on the Relationship Between Anchor Force

97

References 1. Thanh, C.H., Dang, N.L., Kim, J.T.: PCA-based filtering of temperature effect on impedance monitoring in prestressed tendon anchorage. Smart Struct. Syst. 22(1), 57–70 (2018) 2. Kejun, H., Guanghui, W., Xueming, W., et al.: Testing technology and application of bolting and quality inspection of roadway based on principle of stress wave. Coal Technol. 36(04), 76–77 (2017) 3. Spalvier, A., et al.: Torsional vibration technique for the acoustoelastic characterization of concrete. Mater. Struct. 53(1), 67–78 (2020) 4. Alnuaimi, H.N., et al.: Monitoring concrete curing by linear and nonlinear ultrasonic methods. ACI Mater. J. 118(3), 61–69 (2021) 5. Lei, W., Canchang, L., Weixu, K., et al.: Nonlinear vibration analysis of a nano-beam considering gas diffusion surface stresses. J. Vibr. Shock 40(10), 50–56 (2021) 6. Jun, W., Hui, C.: Discriminant analysis of concrete debonding of CFST based on nonlinear vibration characteristics. J. Civ. Archit. Environ. Eng. 40(01), 48–54 (2018) 7. Hui, C., Yaxiang, L.: Nonlinear vibration characteristics and de-bonding recognition of concrete-filled steel tube columns. J. Vibr. Shock 39(01), 260–265 (2020) 8. Heller, L., Foltête, E., Piranda, J.: Experimental identification of nonlinear dynamic properties of built-up structures. J. Sound Vib. 327(1), 183–196 (2009) 9. Zhixiang, H., Zuocai, W., Yixin, R., et al.: On the analytical mode decomposition theory and algorithm for discrete vibration signal processing. J. Vibr. Eng. 29(02), 348–355 (2016) 10. Genda, C., Zuocai, W.: A signal decomposition theorem with Hilbert transform and its application to narrowband time series with closely spaced frequency components. Mech. Syst. Signal Process. 28, 258–279 (2012)

Interrelationships of Load and Displacement of Barrette Piles for Various Interpretation Criteria Subjected to Uplift Loading Yit-Jin Chen(B) , Anjerick Topacio, and Suneelkumar Laveti Chung Yuan Christian University, Taoyuan, Taiwan [email protected]

Abstract. This paper evaluates various interpretation criteria for barrette piles subjected to uplift loading conditions. Eight load test results were gathered and employed for the analysis in order to determine the application of these interpretation criteria to barrette piles. The database was divided into drained and undrained soil conditions. Analysis of each of the interpretation criteria was performed in relation to the displacement ranges of each of the interpreted capacities. It was found out that the interpretation criteria L1 provided the initial linear elastic stage or the serviceability design at mean displacements of 4.1 mm and 7.3 mm, respectively, for drained and undrained soil conditions. On the other hand, the interpretation criteria of DeBeer, van der Veen, Terzaghi and Peck, Davisson, L2 , and slope tangent fell on the same ranges of interpreted capacities with mean displacements ranging from 15 to 25 mm for drained and from 21 to 34 mm for undrained soil conditions. Finally, the interpretation criteria of DIN4026, Fuller and Hoy, and Chin all over-estimate the capacity with mean displacement exceeding 40 mm for drained and 53 mm for undrained soil conditions. In addition, the interrelationships of the load and the displacement for each of the interpretation criteria were further analyzed. A normalized load-displacement curve was determined in order to assess the corresponding mean displacements at which each of these interpretation criterion’s loads are mobilizing along the curve. Statistical analysis was also applied to determine the consistency and reliability of each of the interpretation criteria. Normalized load-displacement equations for barrette piles subjected to uplift loading condition were also calculated for both drained and undrained soil conditions to be utilized and recommended for engineering practice and design of barrette piles for uplift loading. Keywords: Uplift loading · Barrette piles · Displacement · Interpretation criteria · Load test

1 Introduction Various conditions of a pile foundations can lead into various load (Q) – displacement (ρ) curve types that is gathered from axial load tests on such foundations. These varieties may exhibit any one of three shapes, A, B, or C, as shown in Fig. 1. But due to the © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 98–107, 2023. https://doi.org/10.1007/978-981-19-3983-9_9

Interrelationships of Load and Displacement of Barrette

99

requirements of structures that can only withstand a range of displacements, most of the load-displacement curves that are gathered from load test results resemble that of curve C. This may pose a dilemma as the capacity of the pile is not clearly visible on such condition of the load-displacement curve. Therefore, the capacity almost always needs to be interpreted from the load test results. Interpretation criteria (e.g., [1–10]) have been proposed over the years for interpreting such failure load. Table 1 defines nine representative criteria for the interpreted failure load based on a variety of assumptions, individual judgments, extrapolations, and others from the measured load–displacement curve. As found in practice, these interpretation criteria will give different results that can vary substantially.

Fig. 1. Typical load–displacement curves for pile foundations

With these uncertainties in the interpretation of the capacity of a foundation, it is of utmost importance to analyze the application of these interpretation on various conditions and pile types. These load test data may provide vital information in determining the effects of different loading conditions to various soil and pile properties. Various researchers have also compiled relational databases of axial load test on different types of piles [11–16]. And since the 1980s, Kulhawy and co-workers have examined this issue in detail for drilled foundations. Their research [9, 10] and [17–20] mainly focused on the L1 (elastic limit) – L2 (failure threshold) method. Later, Chen and co-authors ([13, 16] and [21–24]) performed a more extensive evaluation to cover the existing representative uplift and compression interpretation criteria for various soil and pile types. What lacked in these analyses is a detailed comparison of various interpretation criteria when they are applied to barrette piles under uplift loading conditions. Therefore, in this paper, nine representative uplift interpretation methods are examined in detail to assess their relative merits and their interrelationships. A database consisting of axial uplift load tests for barrette piles under drained and undrained soil conditions is used for this purpose. The results are compared statistically and graphically, and conclusions are reached for consistent use in practice.

100

Y.-J. Chen et al. Table 1. Definitions of representative uplift interpretation criteria for pile foundations Method

Classification

van der Veen (1953)

Mathematical modeling

Chin (1970)

Mathematical modeling

Terzaghi and Peck (1967)

Settlement limitation

Settlement limitation Settlement Fuller and Hoy (1970) limitation Settlement DIN4026 (1975) limitation DeBeer (1970)

Davisson (1972)

Graphical construction

slope tangent (1985)

Graphical construction

L1 - L2 (1989, 2002)

Graphical construction

Definition of interpreted capacity, Q Value of Qult which is the ultimate load that gives a straight line when log (1-Q/Qult) is plotted versus total settlement. Load is equal to inverse slope, , of line with Q = load and s = total settlement. Load occurs at 1.0 in (25.4 mm) total settlement. Load occurs at which change in slope on log-log total settlement curve. Minimum load occurs at a rate of plastic settlement of 0.05 in per ton (0.14 mm/kN). Load corresponds to displacement at 2.5% B. Load occurs at a displacement equal to the pile elastic compression line, , plus 0.15 in (3.8 mm) + B (in inch or mm)/120, in which Q = load, D = depth, A = area, E = Young’s modulus, B = pile diameter. Load occurs at a displacement equal to the initial slope of the load-displacement curve plus 0.15 in (3.8 mm) + B (in inch or mm)/120. L1 and L2 designate the elastic limit and failure threshold, respectively. Failure is defined qualitatively as the load beyond which a small increase in load produces a significant increase in displacement.

2 Database The database that was utilized in this study consisted of eight (8) load test results of barrette piles under uplift loading conditions. These load tests were done both in drained and undrained soil conditions, thus, the database was further divided into the said soil conditions, respectively. Division of the database into drained and undrained groups is governed by the prominent soil type along the pile length of each load test. Table 2 shows the soil and pile parameters that have been utilized in the study for its analysis. It can also be calculated in the table that the average equivalent diameter of the database is at 1.98 m while average pile length is at 41.7 m ranging from 3.5 to 57.5 m.

Interrelationships of Load and Displacement of Barrette

101

3 Interpreted Axial Uplift Capacity As discussed, nine different criteria were used to analyze the interpreted capacity, as given in Table 1. These criteria were selected because they represent various displacement ranges and may represent the distribution of the interpreted results from the lower, middle and higher ranges as seen in past researches. Table 3 shows the results of each of the interpreted capacities (Q) based on each of the methods and represents different ranges of the capacities. However, during extrapolation, some load tests were terminated before achieving the available interpolated values. Following the conclusions of Phoon and Tang [25], bias is deemed inside a reasonable range for extrapolation from a load test terminated at 75% or higher of the actual Davisson capacity which is around 133% or lower of the final terminated load from any load test. Thus, these interpreted results were denoted as greater than (>) the value of 133% of the terminated load. In addition to the results of the interpreted capacities, the relative displacements (ρ) are also determined in order to assess the location of each of the interpreted capacities Table 2. Soil and pile information for barrette piles Shaft No.

Test location Soil layer description

Soil condition

Width, W (m)

Side, S (m)

Equivalent Pile diameter B length (m) L (m)

L/B

TPU1

Taipei, Taiwan

Clay, sand and rock

Undrained

1.20

2.70

2.03

34.50

17.0

TPU2

Taipei, Taiwan

Clay, sand and rock

Undrained

1.20

2.70

2.03

50.30

24.8

TPU3

Taipei, Taiwan

Clay and silty gravels

Drained

1.30

2.70

2.11

52.00

24.6

TPU4

Taipei, Taiwan

Silt sand, silt clay

Drained

1.20

2.70

2.03

51.40

25.3

TPU5

Bangkok, Thailand

Soft and hard clay and dense to very dense sand

Undrained

1.50

3.00

2.40

57.50

24.0

TPU6

Taipei, Taiwan

Silty clay, sandy silt, and gravel

Drained

0.8

2.7

1.66

45.3

27.3

TPU7

Taipei, Taiwan

Silty clay, sandy silt, and gravel

Drained

0.80

2.60

1.63

39.50

24.2

TPU8

Zurich, Switzerland

Loose to dense moraine

Drained

1.00

3.00

1.95

3.50

1.8

102

Y.-J. Chen et al. Table 3. Interpreted uplift capacities utilizing various interpretation criteria

Shaft no.

Soil condition

Interpreted capacity, Q(kN) L1

L2

DAV*

ST*

T&P*

DeBeer

DIN*

F&H*

VDV*

TPU3

Drained

12500

20000

22700

22400

22482

20400

24297

24600

21000

26259

TPU4

8000

15000

>15656

15400

>15656

13200

>15656

>15656

13000

>15656

TPU6

12500

20800

21000

20300

20549

20000

22300

23000

17000

25771

TPU7

8000

12000

12200

12100

12473

12500

13304

13600

11000

>14352

9800

14000

15700

15600

15798

12700

>15960

13800

12000

>15656

20000

30000

32400

31200

31093

25000

33699

34000

26000

36785

TPU2

15000

29700

26400

26400

25302

25100

30549

32000

28000

38550

TPU5

35250

61500

62100

66700

57773

60500

69525

69000

60000

>71820

TPU8 TPU1

Undrained

Chin

Note: *: DAV – Davisson, ST – Slope-tangent, T&P – Terzaghi and Peck, DIN – DIN4026, F&H – Fuller and Hoy, VDV – van der Veen

and their distribution along the load-displacement curve as seen in Table 4. Furthermore, comparison between each of the interpreted capacities was done in order to assess where each of the interpretation methods are distributed along the normalized load displacement curve. In order to assess this, a normalizing interpretation method must be determined in order to check each of the other methods’ location in the curve in relation to the normalizing method. the L2 method was used as the normalizing criterion. This graphical method interprets the capacity as the start of the load–displacement curve’s final linear region. This method is effective for load–displacement curves resembling that of curves B and C with the application of a hyperbolic extension. After normalizing the interpreted capacities and calculating the mean values for both drained and undrained soil conditions, it can be found that the interpreted load of the L1 provided the initial linear elastic stage of the developed normalized load–displacement curve. It may be used to predict the serviceability load that can be resisted by barrette piles or designs that require displacements that do not exceed mean displacements of 4.1 mm and 7.3 mm, respectively for drained and undrained soil conditions.

Interrelationships of Load and Displacement of Barrette

103

Table 4. Relative displacements utilizing various interpretation criteria Shaft no.

Soil condition

Relative displacement, ρ (mm) L1 L2

DAV* ST* T&P* DeBeer DIN* F&H* VDV* Chin

TPU3 Drained

3.9 13.6 27.2

24.8 25.4

14.9

52.9

63.3

17.0

>63.3

TPU4

2.1 16.5 24.6

20.7 25.4

8.0

24.6

24.6

7.5

>24.6

TPU6

6.1 27.0 28.4

23.9 25.4

22.4

41.5

53.6

12.5

>53.6

TPU7

5.8 20.5 22.4

21.4 25.4

25.7

40.7

50.6

14.1

>50.6

TPU8

2.4

8.5 23.1

21.2 25.4

5.3

30.1

7.8

4.3

>30.1

TPU1 Undrained 5.5 20.6 34.4

26.0 25.4

9.9

50.8

56.8

11.2

>56.8

TPU2

8.5 44.6 28.9

28.9 25.4

24.8

50.8

65.0

35.3

>65.0

TPU5

8.0 32.0 33.3

46.7 25.4

30.0

59.8

57.0

29.1

>76.0

Note: *: DAV – Davisson, ST – slope-tangent, T&P – Terzaghi and Peck, DIN – DIN4026, F&H – Fuller and Hoy, VDV – van der Veen

Most of the interpretation criteria fell on the transition region of the normalized load– displacement curve with the methods of L2 , Davisson, slope-tangent, Terzaghi and Peck, van der Veen, and DeBeer. These interpretation criteria provided good estimates of the capacity of barrette piles and are effective for designs that require mean displacements that do not exceed a range from 15 to 25 mm for drained and from 21 to 34 mm for undrained soil conditions. Lastly, the methods of Fuller and Hoy, DIN4026, and Chin had overestimated capacities and thus were un-conservative in interpreting the capacity of barrette piles for both drained and undrained soil conditions with mean displacements exceeding 40 mm for drained and 53 mm for undrained soil conditions. Graphical representation of the location for these methods on the normalized loaddisplacement curve can be seen in Figs. 2 and 3 for drained and undrained soil conditions, respectively.

104

Y.-J. Chen et al.

Fig. 2. Normalized load displacement curve for barrette piles in drained soils

Fig. 3. Normalized load displacement curve for barrette piles in undrained soils

It can also be seen in the calculated results that the drained soils mobilize capacity at a lower displacement values in comparison to the undrained soil conditions. This means that lower capacities can be expected from sandy soils at lower displacements in comparison to clayey soils at slightly higher displacements. Normalized load-displacement curve equations are also computed based on the data that were interpreted for both drained and undrained soil conditions. These equations may help in simplifying the analysis of each interpreted capacities in relation to that of the normalizing method that is L2 . The equations are listed below for drained and undrained soil conditions, respectively. Q ρ = QL2 3.21 + 0.82ρ

  for drained soils r2 = 0.99

(1)

Interrelationships of Load and Displacement of Barrette

Q ρ = QL2 7.09 + 0.78ρ

  for undrained soils r2 = 0.99

105

(2)

Furthermore, the results of this preliminary analysis may be able to shed light on the behaviour of each of the interpretation criteria when applied to barrette piles under uplift loading conditions. In order to increase the reliability and decrease the uncertainty of the results of the analysis, additional load tests should be employed to the database. It is therefore recommended for future expansion of the study the addition of load test data, in order to increase the range of pile and soil properties included in the analysis. Also, analysis of the behaviour of the interpretation criteria to the side and tip resistances is advised in order to present a more robust comparison between the interpretation methods that are being studied.

4 Summary and Conclusions Axial uplift load test data were used to evaluate the capacity of barrette piles in various soil conditions. The database included 8 field uplift load tests, including 5 drained and 3 undrained soil conditions. Nine representative interpretation criteria were utilized to evaluate the available data. From these analyses, the following results were drawn: 1. L1 method provided the initial linear elastic stage or the serviceability region of the developed normalized load–displacement curve with mean displacements that do not exceed 4.1 mm and 7.3 mm, respectively, for drained and undrained soil conditions. 2. The methods of L2 , Davisson, slope-tangent, Terzaghi and Peck, van der Veen, and DeBeer are located at the transition region to the initial stage of the final linear region of the curve. These interpretations yielded at mean displacements that do not exceed a range from 15 to 25 mm for drained soil conditions. 3. For undrained soil conditions, the methods of L2 , Davisson, slope-tangent, Terzaghi and Peck, van der Veen, and DeBeer yielded at mean displacements ranging from 21 to 34 mm. 4. The methods of Fuller and Hoy, DIN4026, and Chin have overestimated capacities and thus were un-conservative in interpreting the capacity of barrette piles for both drained and undrained soil conditions. These methods have mean displacements exceeding 40 mm and 53 mm for drained and undrained soil conditions, respectively. 5. Normalized load-displacement curves and their respective equations have been presented to be utilized for future designs of barrette piles in different soil conditions. ρ with an r 2 = 0.99; Drained soil conditions yielded an equation of QQL2 = 3.21+0.82ρ ρ while undrained soil conditions yielded an equation of QQL2 = 7.09+0.78ρ with an r 2 = 0.99. 6. In order to increase the reliability of the analysis, additional load test data is necessary for the analysis.

Acknowledgements. This study was supported by the Ministry of Science and Technology of Taiwan under the contract number MOST 110–2221-E-033–010-MY2 and the John Su Foundation.

106

Y.-J. Chen et al.

References 1. van der Veen, C.: Bearing capacity of a pile. In: Proceedings of the 3rd International Conference on Soil Mechanics and Foundation Engineering, Zurich, Switzerland, pp. 16–27 August 1953. International Society for Soil Mechanics and Geotechnical Engineering, London, vol. 2, pp. 85–90 (1953) 2. Terzaghi, K., Peck, R.B.: Soil Mechanics in Engineering Practice. Wiley, New York, 2 (1967) 3. Chin, F.K.: Estimation of the ultimate load of piles not carried to failure. In: Proceedings of the 2nd Southeast Asian Conference on Soil Engineering, Singapore, pp. 81–90 (1970) 4. DeBeer, E.E.: Experimental determination of the shape factors and bearing capacity factors of sand. Géotechnique 20(4), 387–411 (1970). https://doi.org/10.1680/geot.1970.20.4.387 5. Fuller, F.M., Hoy, H.E.: Pile load tests including quick load test method, conventional methods, and interpretations. Highw. Res. Board 333, 74–86 (1970) 6. Davisson, M.T.: High capacity piles. In: Proceedings of the Lecture Series on Innovations in Foundation Construction, American Society of Civil Engineers, Illinois Section, Chicago, Ill. p. 52 (1972) 7. DIN 4026: Beiblatt, Rammpfahle, Herstellun Bemessung und Zulassige Belastun Eraulterungen (1975). https://www.umwelt-online.de/recht/bau/din/402664B._C 8. O’Rourke, T.D., Kulhawy, F.H.: Observations on load tests on drilled shafts. In: Proceedings of Drilled Piers and Caissons II. Edited by C.N. Baker. American Society of Civil Engineers, New York, pp. 113–128 (1985) 9. Hirany, A., Kulhawy, F.H.: Interpretation of load tests on drilled shafts. II: axial uplift. In: Kulhawy, F.H. (ed.) Proceedings of the Foundation Engineering: Current Principles and Practices. GSP 22, pp. 1150–1159. American Society of Civil Engineers, New York (1989) 10. Hirany, A., Kulhawy, F.H.: On the interpretation of drilled foundation load test results. In: O’Neill, M.W., Townsend, F.C. (eds.) Proceedings of Deep Foundations. GSP 116, pp. 1018– 1028. American Society of Civil Engineers, Reston, Va (2002) 11. Long, J.H., Shimel, S.: Drilled shafts – A database approach. In: Proceedings of the Foundation Engineering Congress, pp. 1091–1108 (1989) 12. Wysockey, M.H., Long, J.H.: Utility of drilled shaft load test results. In: Proceedings of the International Conference on Design and Construction of Deep Foundations, pp. 1789–1803 (1994) 13. Marcos, M.C., Chen, Y.-J., Chang, K.C.: Evaluation of interpretation criteria for piles under compression loading. J. Adv. Eng. 9, 177–182 (2014) 14. Chen, Y.-J., Liao, M.-R., Lin, S.-S., Huang, J.-K., Marcos, M.C.: Development of an integrated web-based system with a pile load test database and pre-analyzed data. Geomech. Eng. 7(1), 37–53 (2014). https://doi.org/10.12989/gae.2014.7.1.037 15. Kumari, A., Thakare, S.W., Dhatrak, A.I.: Lateral and uplift capacities of barrette pile in sandy soil. In: Latha Gali, M., Raghuveer Rao, P. (eds.) Construction in Geotechnical Engineering. LNCE, vol. 84, pp. 215–235. Springer, Singapore (2020). https://doi.org/10.1007/978-98115-6090-3_15 16. Chen, Y.-J., Chu, T.-C., Topacio, A., Marcos, M.C.: Interrelationships of load and displacement of barrette piles for various interpretation criteria under drained loading. In: Proceedings of the 2021 International Conference on Civil, Materials, and Environmental Engineering, Malaysia, pp. 1–7 (2021) 17. Jeon, S.S., Kulhawy, F.H.: Evaluation of axial compression behavior of micropiles. In: Brandon, T.L. (ed.) Proceedings of the Foundation and Ground Improvement. GSP 113. American Society of Civil Engineers, Reston, Va (2001) 18. Cushing, A.G., Kulhawy, F.H.: Drained elastic behavior of drilled shafts in cohesionless soils. In: O’Neill, M.W., Townsend, F.C. (eds.) Proceedings of DEEP Foundations. GSP 116, pp. 22–36. American Society of Civil Engineers, Reston, Va (2002)

Interrelationships of Load and Displacement of Barrette

107

19. Kulhawy, F.H.: On the axial behavior of drilled foundations. In: Turner, J.P., Mayne, P.W. (eds.) Proceedings of Geosupport 2004: Drilled Shafts, Micropiling, Deep Mixing, Remedial Methods, and Specialty Foundation Systems. GSP 124, pp. 34–51. American Society of Civil Engineers, Reston, Va (2004) 20. Chen, J.-R.: Axial behavior of drilled shafts in gravelly soils. Ph.D. thesis, Department of Civil and Environmental Engineering, Cornell University, N.Y. (2004) 21. Chen, Y.-J., Chang, H.-W., Kulhawy, F.H.: Evaluation of uplift interpretation criteria for drilled shaft capacity. J. Geotech. Geoenvironmental Eng. 134(10), 1459–1468 (2008). https://doi. org/10.1061/(ASCE)1090-0241(2008)134:10(1459) 22. Chen, Y.-J., Fang, Y.-C.: Critical evaluation of compression interpretation criteria for drilled shafts. J. Geotech. Geoenvironmental Eng. 135(8), 1056–1069 (2009). https://doi.org/10. 1061/(ASCE)GT.1943-5606.0000027 23. Chen, Y.-J., Chu, T.-H.: Evaluation of uplift interpretation criteria for drilled shafts in gravelly soils. Can. Geotech. J. 49, 70–77 (2012) 24. Chen, Y.-J., Lin, W.-Y., Topacio, A., Phoon, K.-K.: Evaluation of interpretation criteria for drilled shafts with tip post grouting. Soils Found. J. 1–16 (2021) 25. Phoon, K.K., Tang, C.: Effect of extrapolation on interpreted capacity and model statistics of steel H-piles. Georisk: Assess. Manag. Risk Eng. Syst. Geohazards 13(4) 291–302 (2019). https://doi.org/10.1080/17499518.2019.1652920

Research on the Influence of Water Horse on the Vortex Induced Resonance Response of Bridges Bifeng Liu1,2 and Changzhao Qian1,2(B) 1 School of Civil Engineering and Architecture,

Xiamen University of Technology, Xiamen 361024, Fujian, China [email protected] 2 Fujian Provincial Key Laboratory of Wind Disaster and Wind Engineering, Xiamen 361024, Fujian, China

Abstract. Vortex-induced resonance is the most common wind-induced vibration phenomenon in bridge structures. On May 5, 2020, the vortex-induced resonance phenomenon occurred on Humen Bridge. The reason is that the water horses continuously installed along the side guardrails of the bridge changed the steel box. The aerodynamic shape of the beam causes the wind-induced vortex-induced resonance phenomenon in the bridge due to the incoming wind. Therefore, it is necessary to study the influence of the water horse on the vortex-induced resonance performance of the bridge structure. In this paper, with the Xiamen Haicang Bridge as the background, a vortex-vibration wind tunnel test of a segment model with a scale ratio of 1:25 is carried out. The main girder segment model is used to arrange water at the windward side railings of the bridge when the wind attack angle is 0°. The effect of water horses on the railings on the windward and leeward side and the water horses on the railings on both sides on the bridge vortex induced resonance. Keywords: Wind tunnel test · Segment model · Bridge structure · Water horse · Vortex induced resonance

1 Introduction Vortex-induced resonance is a wind-induced vibration phenomenon that is very likely to occur in long-span bridges at low wind speeds [1]. It is a self-excited limiting vibration. It is caused by airflow passing through the surface of the main beam. Caused by vortices that fall off regularly at a certain time interval. In a certain range of wind speed, when the frequency of the vortex shedding is close to a certain natural frequency of the bridge, the bridge will produce vortex vibration. At the same time, the bridge vibration will give feedback to the shedding vortex, making a certain wind speed.The frequency of vortex shedding in the interval is locked around the natural frequency of the bridge and does not change with the change of wind speed [1]. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 108–119, 2023. https://doi.org/10.1007/978-981-19-3983-9_10

Research on the Influence of Water Horse on the Vortex

109

Vortex-induced resonance has occurred in many bridges at home and abroad, such as the Xihoumen Bridge in China, the Humen Bridge in China, the Tokyo Bay Passage Bridge in Japan, and the Kossock Cable-stayed Bridge in the United Kingdom etc. Therefore, in order to ensure the aerodynamic stability of the bridge structure, the study of vortex vibration is essential when the bridge is designed for wind resistance. At present, the research on vortex vibration phenomenon is difficult to use analytical methods. Therefore, the use of wind tunnel test technology is the most widely used method in the study of bridge vortex vibration phenomenon, and the segment model wind tunnel test is one of the most conventional test methods [2]. In order to study the vortex vibration performance of long-span bridges more accurately, the wind tunnel test of the segment model with large scale ratio is more and more adopted. When Larose et al. conducted a wind tunnel test of a segment model of the Stonecutter Bridge with a large scale ratio, they found that the Reynolds number has a direct effect on the vortex detachment characteristics when the bridge reaches the vortex lock interval [3]. Sun Yanguo, Liao Haili, etc. conducted wind tunnel tests on a segment model of a long-span suspension bridge with a scale ratio of 1:20 to study the vortex vibration experiment of the segment model, and analyzed the damping and wind attack. The different influences of other factors such as angle on the vortex vibration response of the bridge [4]; Li Yongle, Hou Guangyang etc.took a long-span highway suspension bridge as an example, produced a section model of the main girder with a scale ratio of 1:45 and performed the vortex vibration wind The tunnel test is used to study and analyze the influence of bridge railings, deflectors, different inspection lane positions and wind attack angles on the vortex-induced resonance response of the main girder. At the same time, aerodynamic control measures to improve the vortex vibration of the main girder are also proposed [5]. On May 5, 2020, the deck of the suspension bridge of the Humen Bridge in Guangdong experienced large vibrations, which aroused strong attention from the academic community. According to the discussion results of the experts, the main reason for the vibration of the Humen Bridge this time was that the water horses were continuously installed along the side guardrails of the bridge, which changed the aerodynamic shape of the steel box girder. The phenomenon of vortex induced resonance. Therefore, this paper uses the Xiamen Haicang Bridge as the research background, and uses a 1:25 scaled segment model to conduct wind tunnel tests to study the response of the water horse on the bridge deck to the vortex induced resonance of a long-span bridge.

2 Engineering Background The research in this paper is based on the Xiamen Haicang Bridge. The Xiamen Haicang Bridge is a three-span continuous full-floating steel box girder suspension bridge connecting Xiamen’s main island and Xiamen’s Haicang District. The span is 230 m + 648 m + 230 m, which is a large-span flexible bridge structure. Due to the geographical environment and wind load, the wind-induced vibration response of the bridge is very obvious, and it causes structural fatigue damage. In order to ensure the safety of the structure of the bridge against wind, the wind tunnel test of the Haixifeng Engineering Research Center of Xiamen University of Technology was carried out to study the

110

B. Liu and C. Qian

vortex induced resonance of the bridge, and the effect of water horses on the vortex vibration response of the bridge was extended to different positions on the bridge deck. The standard cross section of the bridge stiffener is shown in Fig. 1.

Fig. 1. Standard cross section view of stiffening beam

3 Overview of Wind Tunnel Test The vortex-induced resonance wind tunnel test of the 1:25 main girder segment model of the Haicang Bridge was carried out in the Wind Tunnel Laboratory of the Haixifeng Engineering Research Center of Xiamen University of Technology, as shown in Fig. 2. This segment model A total of 4 test conditions were studied in wind tunnel tests, including bare bridge status, water horses at the windward side railings, water horses at the leeward side railings, and water horses at both side rails. The water horses are also in accordance with actual engineering. The size used is made according to the scale ratio of 1:25, and its size is: 4 m × 0.032 m, and the test conditions involved are all completed in a uniform flow field at a wind attack angle of 0°.

Fig. 2. Segment model wind tunnel test

Research on the Influence of Water Horse on the Vortex

111

3.1 Segment Model Design The geometric scale ratio of the main girder segment model used for the vortex wind tunnel test is λL = 1:25. In order to reduce the influence of the three-dimensional flow at the end of the segment model, the length of the main girder model is taken as L = 4 m and the main girder width B = 1.28 m, the model height is H = 0.12 m, and the model aspect ratio is about 3.125. Use ANSYS software to establish the structural model of the Haicang Bridge, and perform dynamic response analysis to obtain the structural dynamic characteristic parameters, so that the correspondence between the main parameters of the real bridge and the main parameters of the segment model can be determined (Table 1). In order to make the segment model’s own structure have good rigidity performance, the main girder segment model is composed of high-strength steel frame and light wood cladding to fully ensure the similarity of geometric shapes. The railings on the main girder are made of acrylic panels in proportion to the fineness. The shape and air permeability of the railing are simulated. Table 1. Design parameter table of segment model Parameter

Symbol

Unit

Design value

Length of main beam

L

m

4

Width

B

m

1.28

Heigth

H

m

0.12

Equivalent mass

meq

kg/m

229

Mass moment of inertia

Imeq

kg·m2 /m

49.33

Verticalfundamental frequency

fh

Hz

1.625

Torsion fundamental frequency

ft

Hz

3.629

Torsion frequency ratio

ε

/

2.233

Vertical bending wind speed ratio

mh

/

2.5

Torsional wind speed ratio

mt

/

3.3

3.2 Allowable Value of Vortex Induced Resonance Amplitude According to the “Code for Wind Resistance Design of Highway Bridges” [6], the allowable amplitude values of vertical vortex-induced resonance and torsional vortexinduced resonance when the existing bridge of Xiamen Haicang Bridge is completed are: [ha ] = 0.04/fv = 0.2384 m [θa ] = 4.56/Bfv = 0.2890◦

112

B. Liu and C. Qian

4 Analysis of Test Results 1. Figures 3 and 4 are the vertical displacement amplitude-wind speed curve and the torsional displacement amplitude-wind speed curve of the main beam segment model at 0° wind attack angle.

Fig. 3. Vertical displacement amplitude-wind speed

Fig. 4. Torsional displacement amplitude-wind speed

Figure 5 is the vertical vibration time history diagram corresponding to the four working conditions when the wind speed is 5 m/s. Figures 6 to 8 are the bare bridge and the windward side railing at the wind speed 5 m/s with the water horses and the two side railings. Vertical vortex frequency corresponding to water horse. It can be seen from Figs. 3 and 5 that when the wind speed is 5 m/s, the bare bridge, the water horses at the windward side railings, and the water horses at the railings on both sides all

Research on the Influence of Water Horse on the Vortex

113

have vertical vortex-induced resonance, and the amplitude is 0.01153 m respectively, 0.109 m, 0.1176 m, the frequency when the vibration occurs and the vertical fundamental frequency of this wind speed are both 0.1625 Hz (as shown in Fig. 7–9), and the amplitude of the vertical vortex-induced resonance displacement of the main beam is less than “The allowable value of the “Code for Design of Wind Resistance of Highway Bridges”, while the bridge structure does not have vertical vortex-induced resonance when the water horse is arranged on the leeward side.

Fig. 5. 5 m/s vertical vibration time history chart

Fig. 6. 5 m/s vertical vortex frequency of bare bridge

114

B. Liu and C. Qian

Fig. 7. Vertical vortex frequency of windward side

Fig. 8. Vertical vortex frequency of leeward

2. When the water horse is arranged at the railing on the windward side of the main beam segment model, Fig. 9 and Fig. 10 are the torsional vibration time-history diagram and torsional vortex vibration frequency corresponding to the main beam when the wind speed is 8.75 m/s, as shown in Fig. 4, It can be seen that the torsional vortex induced resonance occurs in the wind speed range of 7.5 m/s–10 m/s, with an amplitude of 0.1066°, and its vibration frequency is consistent with the torsional fundamental frequency at this wind speed, which is 0.4799 Hz (Fig. 10) Shown). Figure 11 and Fig. 12 are the torsional vibration time-history diagram and torsional vortex frequency corresponding to the main beam when the wind speed is 15 m/s. It can be seen from Fig. 4 that torsion occurs in the wind speed range of 11.25 m/s– 16.25 m/s For vortex-induced resonance, the wind speed with the largest amplitude of 6 m/s is selected for analysis. Its amplitude is 0.3362°, which is greater than the allowable value of the specification. The vibration frequency and the torsional fundamental frequency at this wind speed are both 0.4799 Hz (as shown in Fig. 12).

Research on the Influence of Water Horse on the Vortex

Fig. 9. 8.75 m/s time history diagram of torsional vibration

Fig. 10. 8.75 m/s torsional vortex frequency

Fig. 11. 15 m/s time history diagram of torsional vibration

115

116

B. Liu and C. Qian

Fig. 12. 15 m/s torsional vortex frequency

3. When water horses are arranged at the railings on both sides of the main girder segment model, Fig. 13 and Fig. 14 are the torsional vibration time-history diagram and torsional vortex vibration frequency corresponding to the main girder when the wind speed is 9.9 m/s. It can be seen that the torsional vortex induced resonance occurs in the wind speed range of 8.25 m/s–11.55 m/s, with an amplitude of 0.0492°, and its vibration frequency is consistent with the vertical fundamental frequency at this wind speed, which is 0.4799 Hz (As shown in Fig. 9). Figure 15 and Fig. 16 are the torsional vibration time-history diagram and torsional vortex vibration frequency corresponding to the main beam when the wind speed is 23.1 m/s. It can be seen from Fig. 4 that it occurs in the wind speed range of 16.5 m/s–24.75 m/s For torsional vortex induced resonance, the wind speed with the largest amplitude of 23.1 m/s is selected for analysis. Its amplitude is 0.5806°, which is much larger than the allowable value of the specification. The vibration frequency and the vertical fundamental frequency at this wind speed are both 0.4799 Hz (as shown in the Fig. 16).

Fig. 13. 9.9 m/s time history diagram of torsional vibration

Research on the Influence of Water Horse on the Vortex

Fig. 14. 9.9 m/s torsional vortex frequency

Fig. 15. 23.1 m/s time history diagram of torsional vibration

Fig. 16. 23.1 m/s torsional vortex frequency

117

118

B. Liu and C. Qian

5 Conclusion The vortex induced resonance phenomenon of the main girder structure generally occurs in the lower wind speed range, and it is very sensitive to the changes in the aerodynamic shape of the bridge. The water horses arranged on the bridge will change the aerodynamic shape of the bridge structure, making the main girder structure It is no longer the original streamlined section. When subjected to wind load, due to being blocked by the water horse, a larger vortex will be formed. The original bare bridge is very different. When the frequency of the force is close to the natural frequency of the bridge structure, the bridge will have a vortex-induced resonance phenomenon, and the amplitude is very large. Comparing the wind tunnel test on the influence of the vortex-induced resonance response of the bridge when the water horses are arranged at different positions at a wind attack angle of 0°, the following conclusions can be drawn. 1) When water horses are arranged at the windward side railings of the bridge deck, water horses are arranged at the railings on both sides, and the bare bridge is in the state of the main girder when the wind speed is 5 m/s, vertical vortex induced resonance occurs, and the amplitudes are 0.109 m and 0.1176 m respectively, 0.01153 m are within the allowable value of the specification, and the corresponding vertical vortex frequency and vertical fundamental frequency are consistent with 0.1625 Hz, 2) When the water horse is arranged at the railing on the windward side of the bridge, the main girder has torsional vortex induced resonance, and the vortex vibration lock interval is 7.5 m/s–10 m/s and 11.25 m/s–16.25 m/s, the largest The torsion amplitude is 0.3362°, which is greater than the allowable value in the specification. When the water horses are arranged at the railings on both sides of the bridge deck, the main beam also has torsional vortex induced resonance, and the vortex vibration lock interval is 8.25 m/s–11.55 m/s and 16.5 m/s–24.75 m/s, the maximum torsion The amplitude of 0.5806° is much larger than the allowable value of the specification. It can be concluded that: arranging water horses on the windward side railings and the railings on both sides of the bridge will cause the main beam to produce vertical vortex induced resonance and torsional vortex induced resonance, and the torsional vortex induced resonance is the largest in the two states The amplitude exceeds the allowable value of the specification. At the same time, the vertical vortex-induced resonance amplitude and the torsional vortex-induced resonance amplitude of the water horses arranged at the railings on both sides are larger than the vortex vibration amplitudes of the water horses arranged at the windward side railings. 3) When the water horse is arranged on the leeward side railing of the bridge deck, there is no vortex-induced resonance phenomenon in the main beam. It may be that when the water horse is arranged on the leeward side, the bridge deck does not easily cause aerodynamic flow, resulting in no vortex-induced resonance of the main beam.

Acknowledgement. The authors would like to acknowledge with great gratitude for the supports of the Science and Technology Project in Xiamen (Grant No: 3502Z20183050), National Science Foundation of Fujian Province (Grant No: 2019J01866), National Science Foundation of China (Grant No: 52178510), Science and Technology Project of Construction in Xiamen (Grant No: XJK-2021–12).

Research on the Influence of Water Horse on the Vortex

119

References 1. Huayang, L.: Analysis of Vortex-Induced Vibration Response of Long-Span Suspension Bridge Suspenders. Zhengzhou University, Zhengzhou (2018) 2. Li, Z., Yaojun, G.: Research on vortex induced resonance test of segment model of Shanghai bridge vehicle-bridge system. Chin. Civil Eng. J. 40(8), 61–66 (2007) 3. Larose, G.L., Larsen, S.V., Larsen, A., et al.: Sectional model experiments at high reynolds number for the deck of a 1018 m span cable-stayed bridge. In: Proceeding of 11th International Conference on Wind Engineering. TTU Press, Lubbock, pp. 373–380 (2003) 4. Yanguo, S., Haili, L., Mingshui, L.: Suspension bridge vortex vibration suppression measures based on segment model test. J. Southwest Jiaotong Univ. 47(2), 218–223 (2012) 5. Li, Y., Hou, G., Xiang, Q., et al.: Wind tunnel test study on optimization of vortex vibration performance of steel box girder of long-span suspension bridge. Acta Aerodyn. 29(6), 702–708 (2011) 6. Ministry of communications of the People’s Republic of China. JTG/T D60–01–2004 Code for Wind Resistance Design of Highway Bridges. People’s Communications Publishing House, Beijing (2004)

The Degradation of Avalanche Anchorage Systems C. Paglia(B) and C. Mosca DACD, University of Applied Sciences of Southern Switzerland, Trevano, CP 12, 6952 Canobbio, Switzerland [email protected]

Abstract. Avalanche protection systems are decisive to protect urban zones in mountain locations during winter. The anchor rods are firmly inserted in the ground down to the base of the rock and are mainly made of steel incorporated in injection cementitious grouts. In the present work, 30 to 40 years old anchorages from the Alps were extracted until 4 m in depth and the degradation was investigated. They generally exhibited a good conservation state. This was also due to the relatively low aggressiveness of ground-infiltrating waters. The carbonation of the injection mortars was very limited and corrosion was present only at an initial stage, along zones no longer than 30 cm and not around the entire steel bar perimeter. Exfoliation and localized corrosion with small craters were rarely present. This latter intensified corrosion was identified in zones where an incomplete wrapping of the anchor bar with injection grout was observed. The same behavior was present where the polymeric nets did not expand during the grout injection process and remained in contact with the steel. This was particularly seen in the lower parts of the anchorages down to 3.0 m, where the cementitious material has more difficulties to be injected and to embed the steel. In this concern, Oxygen-rich or depleted zones did not create corrosion macro-elements. Degradation susceptible zones, such as threaded heads portions or steel protection sheaths exhibited a slight uniform corrosion, but no reduction of the anchorage diameters. The bolts did not shifted during the tensile extraction tests and loads around 300 kN were registered. Therefore, uniform corrosion on threaded anchorage heads did not influence the bolt resistance. Keywords: Avalanche · Steel · Anchorage · Degradation

1 Introduction Pre-stressed cables and anchorages are widely used in the construction sector. They are placed in bridges, within cooling towers, in water tanks, off-shore tank, water plants, wall anchorages and foundations. They are mainly composed of an anchorage head with a fixed plate, a twisted structure with an injection/deareation channel an enlargement and a steel bar protection sheath. A cement-based injection grout is injected around the metal bars to protect them from corrosion. The durability of anchorage systems is influenced by the development of high strength cementitious injection materials, the © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 120–127, 2023. https://doi.org/10.1007/978-981-19-3983-9_11

The Degradation of Avalanche Anchorage Systems

121

content of corrosion inhibitors or silica fume and steel alloying elements [1]. Several alternative protection systems to steel are used: PVC, HDPE and PP sheaths. Additionally non-destructive diagnostic methods such as corrosion potential, polarization resistance, galvanostatic pulse, impedance spectroscopy, ultrasonic measurements, impact echo, radiography, radar and others contribute to control the degradation [2, 3]. Lately, anchorage or pre-stressed cables are put in place with an electrical separation from other metal parts of the structure, in particular rebars or foundation. Steel anchorage are often used in walls or in slope static issues. In this case, the anchorage-mortar system are in contact with the ground or the rocks. The steel anchorage are important parts of avalanche defense structures. These are widely widespread in mountain regions, in particular along the Alps. Under these extreme atmospheric conditions, the durability assumes a decisive role in the safety and functionality of the structures. In order to gain experience in the degradation of the steel anchorage and sheaths up to 40 years old, systems have been extracted from the soil and destructive investigations were carried out.

2 Experimental Procedure The anchorage were extracted from the soil. Twelve samples up to 4 m long, including the embedded cementitious material were brought to the laboratory. Visual and optical (binocular lens) investigations were carried out. The carbonation depth in the mortar was detected with Phenolphtalein [4], while the microstructure was observed with an optical microscope. The anchorage bolt extraction tests were carried out with a conventional tensile testing equipment. The Brinell hardness was measured according to the norm EN ISO 6506 [5] on the surface and on the anchorage section. For the comparison between hardness and tensile strength were used the tables displayed in the norm EN 18265 [6]. The main parameters of the chemistry of groundwater and a geology overview were also analyzed.

3 Results and Discussion The old anchorages exhibited a nominal diameter of 36 mm and were still partially wrapped in the mortar. The mortar diameter reached values up to 12 cm. In some cases, a decentralized position of the anchorage with respect to the mortar was observed (Fig. 1). This may be accounted for a difficult embedment and penetration of the cementitious material along the entire length of the system, especially with depth. The mortar cover thickness varied from 0 to 120 mm, depending on the position. Locally, steel parts were directly exposed to the surface, i.e. to the soil and groundwater. The steel parts embedded in mortar did not exhibit a significant corrosion. This was also due to the general low carbonation rate of the injection grout into the ground, with 3–4 mm over 40 years. Mortar micro-cracks longitudinal, transverse and radial to the anchorage were also observed. These may be caused in situ during slope slowly movements or more

122

C. Paglia and C. Mosca

Fig. 1. Cross section of the anchorage. Note the decentralized position of the steel bar.

likely during extraction tests or transportation, since no particular enhanced ingress of aggressive agents or degradation is observed on steel surface in the crack regions. The mortar leaching phenomena were not significant, despite the sporadic presence of water. The water exhibited a pH around 8, corrosive for steel. The water carbonate hardness was 27.3 °F, typical for bicarbonate-rich water and not particularly aggressive for mortars (leaching). The water sulphate content of 46.5 mg/l (200–600 mg/l) as well as the Ammonium content 6.0, the converted σucs will increase significantly. When I s(50) > 10.0, σucs has been greater than 500 MPa. In a word, the exponential formula would overestimate the σucs of rocks with high strength. (2) According to Fig. 8, the σucs obtained by different empirical formulas are quite different. The σucs of axial loading σca = 59.4 – 151.1 MPa, and the range reaches 91.7 MPa, while that of radial loading is σcd = 23.9 – 53.6 MPa, and the range is 29.7 MPa. The maximum value is obtained by logarithmic function of the 8th , and the maximum discreteness is obtained by power function of the 5th . Therefore, the greater the rock strength is, the greater the difference of σucs obtained by different empirical formulas, the lower the reliability of the conversion results. (3) Among the 10 selected models, the σucs obtained by power empirical formula recommended by Standard for engineering classification of rock mass is the smallest, which means the result is more conservative. the σucs and the discrete coefficient obtained by linear empirical formula recommended by ISRM are more reasonable when the lower limit is taken. (4) The anisotropic coefficients of sandy slate obtained by different methods are all greater than 2.0, and the anisotropic coefficients obtained by 5th and 9th empirical formular are even greater than 4.0, indicating that the strength of sandy slate is obviously anisotropic, and the σucs perpendicular to the plane is much larger. Point load test has been widely used in practical engineering due to its flexibility, convenience and low cost. Point load test is a test method that can quickly and effectively obtain the uniaxial compressive strength of rock. However, the empirical relationship

184

W. Xingcan et al.

between point load strength index and uniaxial compressive strength is complex. There is no recognized and reliable unified empirical conversion formula so far. Most of them prefer to carry out both point load test of rock blocks and uniaxial compression test of standard rock samples at the same time, and then established the empirical formular. Finally, the uniaxial compressive strength of similar rock is calculated by the empirical formular. It is inevitable to carry out the uniaxial compressive tests of standard rock samples, which weaken the advantages of point load test. This is one of the biggest obstacles to limiting the application of point load test in rock engineering. In this paper, based on a large number of point load test results of sandy slate cores, 10 different empirical formulars are used to predict the uniaxial compressive strength. The comparative analysis show that the prediction effect of exponential function and logarithmic function is poor for sandy slate with obviously anisotropic. The effectiveness of linear function depends on the selected proportional coefficient. The lower limit recommended by ISRM is better than those others. The power function recommended by Standard for engineering classification of rock mass (SECR) is conservative. Table 2. Some empirical equations correlating the UCS to point load Strength Index of rock. Serial number

Type

Empirical relationship

Applicable conditions

1

Linearity

Source

σ c = 20I s(50)

Schist

ISRM [7]

2

σc = 23.62I s(50) −2.69

Coal rock

Kahraman [18]

3

σ c = 21.65I s(50)

Red sandstone

Dai Ling et al. [8]

4

σ c = 13I s(50)

Sandstone, I s(50) < Tsiambaos [24] 2 MPa

σ c = 7.3I s(50) 1.71

Soft rock

σc = 22.82I s(50) 0.75

Not applicable to conglomerate and σ c < 5 MPa very soft rock

Standard for engineering classification of rock mass [29]

5 6

Power function

7

Exponential function

σ c = 2.27exp(1.04 I s(50) )

Dry and saturated pyroclastic rocks

Kahraman [22]

8

Logarithmic function

σ c = 100ln(I s(50) ) + 13.9

Sandstone and other 17 kinds of rocks

Teymen [28]

9

Parabola

σ c = 3.86I s(50) 2 + Pyroclastic rock 5.65 I s(50)

Quane [26]

σ c = –0.66I s(50) 2 + 21.15I s(50)

Zhang Jianming et al. [13]

10

0 < I s(50) < 15 MPa

Study on the Strength Characteristics of Metamorphic Slate Based on Point Load Tests

185

Table 3. Experimental results of Point Load Tests. Category

Load Mean Standard Confidence Coefficient Number Coefficient direction strength deviation interval of of of /MPa dispersion effective anisotropy tests

Carbonaceous Vertical slate Parallel

3.5

1.9

[2.8,4.2]

0.81

32

2.4

1.7

[1.4,3.3]

0.26

15

Sandy slate

Vertical

4.6

2.4

[4.1,5.0]

0.65

112

Parallel

1.6

1.1

[1.4,1.8]

0.89

96

1.46 2.88

Table 4. Experimental results of uniaxial compressive strength for sandy slate. Serial number

Load direction

Mean σucs /MPa

Standard deviation

95% confidence interval

1 ISRM

Axial

91.4

47.9

[32.3,41.0]

Coefficient of dispersion 0.68

Radial

36.7

21.8

[32.3,41.0]

0.69

2 3 4 5 6 SECR 7 8 9 10

Axial

105.2

56.5

[94.6,115.8]

0.67

Radial

40.6

25.7

[35.4,45.9]

0.74

Axial

98.9

51.8

[89.2,108.6]

0.65

Radial

39.7

23.6

[34.9,44.5]

0.69

Axial

59.4

31.1

[53.6,65.2]

0.65

Radial

23.9

14.2

[21.0,26.7]

0.69

Axial

114

97

[95.9,132.2]

0.96

Radial

24.9

23.5

[20.2,29.7]

1.10

Axial

69.5

27.8

[64.3,74.7]

0.50

Radial

34.8

16.0

[31.5,38.0]

0.53

Axial

5191.6

17628.9

[1890.8,8492.5]

1.30

Radial

31.5

48.3

[21.7,41.2]

0.99

Axial

151.1

56.5

[140.1,161.7]

0.52

Radial

53.6

70.3

[39.3,67.8]

1.00

Axial

128.3

113.1

[107.1,149.4]

0.95

Radial

27.9

24.9

[22.8,32.9]

0.98

Axial

79.1

34.2

[72.7,85.5]

0.55

Radial

35.8

19.9

[31.8,39.8]

0.65

Coefficient of anisotropy 2.49 2.59 2.49 2.49 4.58 2.00 164.81 2.82 4.60 2.21

186

W. Xingcan et al. Table 5. Statistical results of converted uniaxial compressive strength of slate.

Category

Load direction

Carbonaceous slate

Vertical

Parallel

Sandy slate

Vertical

Parallel

σc mean value /MPa

Mean difference/%

ISRM

69.9

18.88

SECR

56.7

Empirical relation

ISRM

47.0

SECR

41.8

ISRM

91.4

SECR

69.5

ISRM

36.7

SECR

34.8

11.06

23.96

5.18

Standard deviation

95% confidence interval

Coefficient of dispersion

37.6

[55.3,84.4]

0.81

23.8

[47.5,65.9]

0.59

34.2

[28.0,65.9]

0.26

21.4

[29.9,53.7]

0.19

47.9

[32.3,41.0]

0.68

27.8

[64.3,74.7]

0.50

21.8

[32.3,41.0]

0.69

16.0

[31.5,38.0]

0.53

Number of effective tests 32

15

112

96

4.3 Comparative Analysis Based on the above-mentioned, the linear and power empirical function recommended by ISRM and Standard for engineering classification of rock mass, respectively, have a better prediction effect on the uniaxial compressive strength of sandy slate. Therefore, these two methods are used to analyze the results of point load test of carbonaceous and sandy slate (Table 5), It shows that: (1) The average difference of σucs obtained by the two methods is small. The minimum difference is 5.18% under radial loading and the maximum difference is 23.96% under axial loading. (2) For vertical loading, the strength of sandy slate is significantly higher than that of carbonaceous slate, indicating that the strength of sandy slate block itself is larger. However, for the condition of parallel loading, the strength of carbonaceous slate is slightly higher than that of sandy slate, indicating that the cementation between carbonaceous slate strata is better.

5 Conclusion Based on a diversion tunnel project in western Sichuan, this paper takes the thin carbonaceous slate and sandy slate of the Lower Ordovician Wachang Formation (O1 w) exposed during excavation as the objects, and comprehensively studies the strength characteristics of slate by XRD, thin section projection microscopy and rock block point load test. The following conclusions are obtained: (1) Due to the complex geological structures, the Ordovician Wachang group (O1 w) slate presents poor integrity, thin layer structure and joints development. It is mainly contained quartz, some albite, muscovite, pyrite, etc. (2) The point load strength of carbonaceous slate and sandy slate meet well of the normal distribution. The anisotropy coefficients are 1.46 and 2.88 respectively, which have obvious anisotropy characteristics. The overall strength of sandy slate is larger.

Study on the Strength Characteristics of Metamorphic Slate Based on Point Load Tests

187

(3) Among the selected 10 kinds of empirical formulas between uniaxial compressive strength and point load index, the exponential and logarithmic function have the worst prediction effect, and the prediction effect of linear function depends on the selected proportional coefficient. The lower limit proportional coefficient recommended by ISRM is better, while the prediction result of power function recommended by Standard for engineering classification of rock mass is conservative. (4) The difference of uniaxial compressive strength obtained by different empirical formulas would be greater for the rock with greater strength, and the reliability of conversion results is lower. (5) The method recommended by ISRM and Standard for engineering classification of rock mass are more suitable for the prediction of uniaxial compressive strength of slate. The results show that the strength of sandy slate is larger than that of carbonaceous slate, while carbonaceous slate has the better cementation. (6) In consideration of the microscopic structure and strength characteristics of the lower Ordovician Wachang Formation (O1 w) slate, in practical engineering, it should be paid more attention to the relationship between the rock layer and loading direction, since the stability of surrounding rock with the same lithology varies greatly with the combination of angles. Meanwhile, the water rationality of the slate is poor which the softening effect of water will significantly reduce the strength of rock and interlayer cementation. The results can provide reference for the stability evaluation of surrounding rock for a large number of deep underground engineering in slate area of Western Sichuan Plateau. Due to the extreme bedding structure and strength anisotropy of slate, the angle between the attitude of rocks and tunnel axis is an important factor affecting the engineering characteristics of slate in practical engineering. However, the complex tectonic action in western Sichuan has caused drastic changes in the attitude of rock mass. Therefore, it is suggested that dynamic construction and support design should be carried out under the premise of fully studying the strength and deformation anisotropy characteristics of slate in the area, combined with the measurement of exposed surrounding rock, so as to effectively reduce the construction risk.

References 1. Yunpeng, X., Qiunan, C., Xiaocheng, H., et al.: An experimental study of microstructure and uniaxial compression test of carbonaceous slate in a deep buried tunnel. Hydrogeol. Eng. Geol. 47(1), 96–102 (2020). (in Chinese) 2. Du, Y.: Study on the mechanical characteristics of support structure and deformation control technology of carbonaceous slate surrounding tunnel. Chang’an University (2017). (in Chinese) 3. Xiaohui, X., Yongjian, B., Yongbo, T., et al.: Geological structure and rock deformation characteristics and their geohazard-controlling mechanism in Yajiang area, west Sichuan. Geoscience 35(1), 145–152 (2021). (in Chinese) 4. Gang, C., Gong Lun, Y., Yu, J., et al.: Study on large deformation characteristics and construction control technology in high altitude slate tunnel. Chin. J. Undergr. Space Eng. 16(S2), 744–751 (2020). (in Chinese)

188

W. Xingcan et al.

5. Cai, J., Du, G., Ye, H., et al.: A slate tunnel stability analysis considering the influence of anisotropic bedding properties. Adv. Mater. Sci. Eng. 2019, 1–17 (2019) 6. Garcia-Fernandez, C.C., Alvarez-Fernandez, M.I., Cardoso, R., Gonzalez-Nicieza, C.: Effect of environmental relative humidity in the tensile strength of layering in slate stone. Bull. Eng. Geol. Environ. 79(3), 1399–1411 (2020). https://doi.org/10.1007/s10064-019-01619-7 7. Franklin, J.A.: Suggested method for determining point load strength. Int. J. Rock Mech. Min. Sci. Geomech. Abstr. 22(2), 51–61 (1985) 8. Ling, D., Huayan, Y., Feiyang, Z., et al.: Experimental study on point load strength of red sandstone with different shapes. Sci. Technol. Eng. 19(07), 209–214 (2019). (in Chinese) 9. Jun, D., Chenchen, W., Siqi, W., et al.: Correlation of compressive strength of irregular coal samples by point load tests. Sci. Technol. Eng. 18(05), 291–296 (2018). (in Chinese) 10. Hu, Q.: Correlation between point load test and standard compressive strength of rock. Water Resour. Hydropower Eng. 24–26 (1997). (in Chinese) 11. Li Hongpeng, W., Yiyang, G.C., et al.: The relation between point load strength and compressive and tensile strength of marble. Sci. Technol. Eng. 19(32), 294–299 (2019). (in Chinese) 12. Lin, C.: Study on large deformation mechanism and deformation control technology of broken carbonaceous slate surrounding tunnel. Southwest Jiaotong University (2019). (in Chinese) 13. Zhang, J.M., Tang, Z.C., Liu, Q.S.: Relation between point load index and uniaxial compressive strength for igneous rock. Rock Soil Mech. 36(S2), 595–602 (2015). (in Chinese) 14. Yu, Z., Quan, G., Fei, Y., et al.: Study of weathering grouping and strength characteristics of wudang group schist based on point load testing. Rock Soil Mech. 33(S1), 229–232 (2012). (in Chinese) 15. Guo, M.L.: Discussion on adaptability of point load test of rocks. Rock Soil Mech. 24(3), 488–489 (2003). (in Chinese) 16. Wang, R., Tang, C., Wang, S.: Study on several problems about point load test of rock. J. Northeast. Univ. Nat. Sci. 29(01), 130–132 (2008). (in Chinese) 17. Azimian, A., Ajalloeian, R., Fatehi, L.: An empirical correlation of uniaxial compressive strength with P-wave velocity and point load strength index on marly rocks using statistical method. Geotech. Geol. Eng. 32(1), 205–214 (2014). https://doi.org/10.1007/s10706-0139703-x 18. Kahraman, S.: Evaluation of simple methods for assessing the uniaxial compressive strength of rock. Int. J. Rock Mech. Min. Sci. 38(7), 981–994 (2001) 19. Guo, Y., Chen, Y., Jiang, Y., et al.: Study on stability control measures for large deformation of carbonaceous slate at diversion tunnel of Guzeng hydropower station. Pearl River 41(12), 58–64 (2020). (in Chinese) 20. Ping, S.U.N., Yueping, Y.I.N., Shuren, W.U., et al.: Experimental study of microstructure and mechanical properties of rocks from Donghekou landslide. Chin. J. Rock Mech. Eng. 29(S1), 2872–2878 (2010). (in Chinese) 21. Ng, I., Yuen, K., Lau, C.: Predictive model for uniaxial compressive strength for Grade III granitic rocks from Macao. Eng. Geol. 199, 28–37 (2015) 22. Kahraman, S.: The determination of uniaxial compressive strength from point load strength for pyroclastic rocks. Eng. Geol. 170, 33–42 (2014) 23. Broch, E.M., Franklin, J.A.: The point load strength test. Int. J. Rock Mech. Min. Sci. Geomech. Abstr. 9, 669–697 (1972) 24. Tsiambaos, G., Sabatakakis, N.: Considerations on strength of intact sedimentary rocks. Eng. Geol. 72(3–4), 261–273 (2004) 25. Forster, I.R.: The influence of core sample geometry on the axial point-load test. Int. J. Rock Mech. Min. Sci. Geomech. Abstr. 20(6), 291–295 (1983) 26. Quane, S.L., Russell, J.K.: Rock strength as a metric of welding intensity in pyroclastic deposits. Eur. J. Mineral. 15, 855–864 (2003)

Study on the Strength Characteristics of Metamorphic Slate Based on Point Load Tests

189

27. Diamantis, K., Gartzos, E., Migiros, G.: Study on uniaxial compressive strength, point load strength index, dynamic and physical properties of serpentinites from Central Greece: test results and empirical relations. Eng. Geol. 108(3–4), 199–207 (2009) 28. Kılıç, A., Teymen, A.: Determination of mechanical properties of rocks using simple methods. Bull. Eng. Geol. Env. 67(2), 237–244 (2008). https://doi.org/10.1007/s10064-008-0128-3 29. National standard of the People’s Republic of China. Standard for engineering classification of rock mass. Beijing: Ministry of Water Resources of the People’s Republic of China (2014). (in Chinese)

Soıl Structure Interactıon Effects on the Seısmıc Response of Rc Structure wıth Rıgıd and Flexıble Foundatıons Gokaran Prasad Awadhya, Aayush Jha(B) , Mrinal Thakur, and Sanidhya Sharma Department of Civil Engineering, Delhi Technological University, Bawana Road, Delhi, India [email protected], [email protected]

Abstract. During earthquakes, soil conditions are extremely important. The study of the energy transmission mechanism from soil to buildings during seismic waves is important for multi-story building seismic design and complements the notion of soil structure for building design. In this paper, G + 10 symmetrical model with fixed (NSSI) and flexible (SSI) bases has been added at the base to investigate the effects of soil structure interaction during seismic waves. Using the Bowles technique area springs and line springs of required stiffness were assigned to the flexible base model. We have proceeded using the response spectrum method in our research because it complies to the linear dynamic features of seismic analysis. Short, non-deterministic, and transient dynamic events such as earthquakes are estimated using the response Spectrum approach. The numerical findings and the graphs obtained using etabs with the SSI model are compared to those obtained using fixed-base support modelling assumptions. The study revealed that soil structure interaction needs to be considered especially while designing buildings in high seismic zones. The SSI responses were studied using the peak responses of story shear, story moment, story displacement, story drift, column moments, base reactions, axial forces, and soil pressure. Keywords: Spectrum · Soil structure interaction · Story drift · Column moments · Seismic prone area

1 Introduction The interaction between the structural elements determines the reaction of the structural system composed of numerous elements. In the case of seismic waves, the soil-structure interaction effect plays a crucial role in influencing the structure’s response. Seismic waves cause motion in structures, which is determined by the structure’s vibrational properties and layout. The magnitude at which earthquake motions alter the features observable at the foundation is entirely determined by the soil’s relative mass and stiffness attributes. In this paper we have proceeded using the response spectrum method because it demonstrates the linear dynamic property of seismic analysis. In this paper, the study and comparison of the soil structure interaction under the effect of earthquake according to the Indian Standard Code. A 10-storey building situated in Kathmandu © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 190–211, 2023. https://doi.org/10.1007/978-981-19-3983-9_17

Soıl Structure Interactıon Effects on the Seısmıc Response

191

(Nepal) has been taken for this study. Earthquake load is applied on buildings as per IS Code 1893 part (1): 2006. Analysis is done using ETABS software. Building is analysed using Response spectrum method under-ground and random motion.Movement of the ground surface in the non-existence of the structure and its foundation is termed as free-field motion. This movement is greater than the Foundation Input Motion i.e., when the motion is applied to structure in presence of line springs and area springs attached to the foundation. Because of the passive response of the structure and the deflections that response creates in the well-defined foundation springs, foundation input motion differs from true foundation base motion. Inertial interaction of the structure and foundation is the name given to this response. Because of the three-dimensional movement of the ground, kinematic interaction transforms the free-field ground action into foundation input motion. Because of the foundation system’s stiffness and resilience, the three-dimensional movement is rounded closer within the building shroud and over the foundation depth within the foundation mark.In some circumstances, Soil Structure Interaction can put up considerable variation in exactly how the structure will perform during the seismic waves and the forces employed for their design. SSI effect shows the difference between the actual response and theoretical response. In this paper, an effort is made to find the detrimental consequence of inertial interaction and are compared for both fixed and flexible base foundations.The effect of SSI subjected to gravity loads and seismic stresses is investigated using ETABS analytical software [1]. The building is modelled with assumed material properties and structural details. The building is presumed to be located in a seismic prone area. The model is simulated under two different situations namely: fixed base (NSSI) and flexible base (SSI). Due to complexities associated with the modelling of structure using soil structure interaction effect, this practice is often neglected, and certain assumptions are made for simplifying the model [2].

2 Literature Review In last half century understanding of earthquake has taken a considerable progress, which has given rise to many model’s concepts to improve the seismic performance of building under dynamic condition. The major earthquakes around the world showed the important role of soil-structure interaction (SSI) in determining dynamic behaviour of structure. Many examples may be found, such as the 1985 earthquake in Mexico City, when soft soil enhanced the ground motion and frequency of ground shaking, causing havoc. Similarly, soil motion amplification caused the cypress freeway to collapse during the 1989 Loma Prieta earthquake. [Pallavi Badri et al. 2016] looked at the SSI interaction analysis of an asymmetric building supported on piled raft in the 2015 Nepal earthquake and discovered that the more complex the building, the higher the risk during an earthquake event, and that the building’s response is governed by the peak ground acceleration rather than the magnitude of the earthquake. [Julio et al. 2008] analysed G + 6 storey rcc frame building under the influence of SSI 4. Under the influence of SSI horizontal spectral acceleration values of structure were decreased. Also, the paper stated vibration period and damping were increased when SSI model was introduced to the dynamic behaviour of the building in comparison to the fixed base model [3]. [Anand et al. (2010)] used Etabs to investigate the g + 15 space frame with and without shear

192

G. P. Awadhya et al.

wall for various soil conditions [4]. Between two frames, the dynamic quantities were compared. When the soil shifted from hard to medium to soft, the base shear, axial force, and lateral displacement in the column all rose at the same time, according to the report. The effect of SSI on different foundations of the construction was compared by [Priyanka et al. (2012). Response spectrum Method was used to examine the structures. [Gopal 0. Dhoot. et al. (2016)] studied performance-based design of RCC with pushover analysis under the effect of SSI. According to the findings, the seismic design provides a viable methodology for retrofitting the structure to meet the needed goal.[Behzad Fatahi et al. (2011)] investigated 15-story concrete-resisting buildings in Australia’s Ce, De, and Ee soil classifications. According to the article, the influence of SSI on inelastic seismic design based on soil type Ce is minimal, however the model’s performance level significantly raised in soil types De and Ee. [Ebrahim Rayat Roknabadi et al. (2017)] analysed the dynamic response of building using experimental tests and finite element analysis. Shaking table results were compared to SSI system finite element analysis results, which included shear velocity, frequency, damping ratio, and PSA [5]. The paper concluded SSI effects reduces with the increase of the foundation embedment. [Shylaja N et al. (2017)] analysed SSI behaviour of asymmetric 3d building frames with isolated footing under modified wrinkler model. The fundamental natural period, base shear, lateral displacement, axial force, and bending moment were all examined in this article [6]. The paper stated bending moment increased with the lessening in modulus of shear in the soil.According to [Nikhil N. Chopade et al. (2018)], the traditional inelastic design technique without SSI is insufficient to ensure the safety of RC buildings resting on medium and soft soil. SSI impacts on seismic response needs of multi-story MRF building on raft foundation were investigated by [Tarek M. A. Alazrak et al. (2014)]. The equivalent static load, response spectrum approaches, and nonlinear time history analysis were all used to analyze the model. According to the findings, SSI has a considerable impact on the building’s base forces and roof displacement when compared to a model that ignores interaction.

3 Soil-Structure Interaction The responsive effect of soil and sub-grade with structure is referred to as soil structure interaction (SSI). The SSI is the difference between the structure’s actual response and the response when the structure’s base is believed to be rigid. Inertial effects, kinematic interaction effects, and soil-foundation flexibility effects are the three types of SSI effects.These effects can be describedin terms of engineering analysis as foundation stiffness and damping, foundation deformation and difference between free-field ground motion and foundation input motion [7]. At the interface, vibrations, displacements, and rotations cause shear, moment, and torsion, which dissipate energy and affect the system’s response. These are caused by the foundation’s elasticity.The flexural, axial, and shear deformations of the structural foundation elements result in forces and displacements applied by the building and the soil medium. The kinematic interaction and relative deformation at the soil foundation interface account for foundation deformation and difference between free-field ground motion and foundation input motion.

Soıl Structure Interactıon Effects on the Seısmıc Response

193

Two methods are used for evaluating the above effects. In Direct method the soil and structure are assumed to be the same model and the system is analyzed. Because this method takes a long time to compute, it is often avoided. InSubstructure methodthe first step in this method is to compute free field soil motions, followed by the computation of transfer functions, which are utilized to convert free field motions into FIM. Finally, springs are assigned to represent stiffness and damping at the soil-structure contact, and the entire model system is analysed [8]. SSI is determined by structure stiffness in relation to soil stiffness, structure height or slenderness in relation to footing width and the mass of the structure in relation to the mass of the earth that supports the structure. When seismic waves reach the foundation’s bottom, they are separated into two types, transmission waves that enter the structure and reflection waves that bounce back into the ground. When a transmission wave enters a structure, creating vibration, it reflects and travels back to the foundation, causing the soil-structure phenomena. The radiation wave is a portion of the wave that is sent back to the earth. Seismic waves are trapped when radiated waves are insignificant, and the building begins to vibrate continuously for a long time [9]. In comparison to the structure itself, radiation damping generates an increase in the total damping of the soil-structure system. The natural frequency of the soil structure system is also lower than the natural frequency of the soil due to the SSI effect. This contact increases the structure’s overall displacement, allowing the foundation to translate and rotate. There are two types of interactions in the soil-structure interaction. Kinematic Interaction which explains how translation excitation causes kinematic interaction and Inertial Interaction which accounts for the fact that the structure’s mass and subsequent hammering can produce soil compliance. Inertial Interaction (SSI) is induced by inertial forces and is associated with the mass of the structure, resulting in overturning moment and transverse shear in the vibrating structure. 3.1 Respone Spectrum Response-Spectrum Analysis (RSA) is a linear dynamic analysis approach for estimating a structure’s peak response from a subset of the system’s odes recovered using the Eigen frequency extraction procedure. It investigates dynamic behavior by determining pseudospectral acceleration, velocity, displacement, and damping over a particular time. A reaction spectrum is a graph of the greatest linear SDOF response for a certain component of earthquake ground motion. In this plot, natural period is on the x axis whereas response quantity (max. displacement, max. velocity or max. acceleration) is on the y-axis as shown in Fig. 1 [10]. Maximum response of SDOF system depends only on the natural frequency and damping.

194

G. P. Awadhya et al.

Fig. 1. Response spectrum graph.

Dynamic analysis is basically performed for regular building in zone IV and V for height greater than 40 m and in zone II and III for height greater than 90 m and for irregular buildings located in zone IV and V for height greater than 12 m and zone II and III for height greater than 40 m as per IS code. Due to lateral load from earthquake forces the building displaces. The response of the structure, due to application of lateral load induces displacement, velocity, and acceleration in the structure. The assessment of these factors is studied in response spectrum method.

4 High Structure-to-Soil Stiffness Ratios Lengthens Period and Change Design Forces If the structure is comparatively stiff in contrast to the soil, foundation rotation and translation comparative to the free-field motion can arise adding up to the structural displacements and expanding or lengthening the structural period of the structure. These increased periods can influence the related spectral accelerations employed for the design purpose. This impact normally arises in structures with a concentrated lateral forceresisting structure. On the other hand, the effect of soil flexibility on the building reaction is generally quite minimal in structures with large, stiff foundations on very stiff soils and flexible structures.The fundamental period of vibration is lengthened by soil flexibility. The line springs and area springs attached to the bottom demonstrate the soil and structure’s vertical flexibility. In the horizontal displacement of the model, there might be vertical displacement of the springs positioned at the bottom of the foundation and the rotation of foundation might occur. This increase in displacement corresponds to an increase in the structure’s time period [11].

Soıl Structure Interactıon Effects on the Seısmıc Response

195

The time-period refers to the building’s fundamental translation or natural period. It signifies the natural period of the oscillation of the building when it is subjected to ground shaking during seismic waves. The height of the building taken for analysis is 30.48 m which gives the time-period to be 0.9729 s according to the formula given below as per IS 1893:2016(part 1). Time-period for bare moment resisting frame buildings. Ta = 0.0075 h0.75 ............eqn

(1)

where, h = height of the building.

5 Site Characterization The site considered for this project lies in the severe most seismic zone i.e., Seismic Zone V with highest seismic zone factor of 0.36. Soil type II is considered for this project as per IS 1893: 2006 part (1). The soil structure interaction model took into account medium and stiff soil with poorly graded sands and gravel (SP) with few particles. The safe bearing capacity of the soil is set at 135 KN/m2 , and the stiffness (ks) of the line and area springs is calculated using the Bowles Method. The foundation of the 10-storey building is positioned on the upper most soil layer with unit weight of soil as 20 KN/m3 as per the statistical records obtained through official website. The site condition consists of a thick deposit of black silty beds, the “Kalimati Formation” for 3 m and Greyish Moderately Weathered Rock beneath. 5.1 Modelling and Formulation The G + 10 Storeys system geometry is located in Kathmandu and has a plan dimension of 14.7486 m × 10.261 m. The structure will be utilized as a dwelling. Reinforced concrete frames support the lateral and vertical load resisting systems. Columns, primary beams, and secondary beams make up the frames. The material and geometric properties are shown in Table 1. (Table 2 and 3). Table 1. Material and geometric properties of frame, raft and concrete S. no

Structure

Component

Details

1

Frame

a. Storey height b. Beam size c. Column size

3.048 m 300 mm × 400 mm 450 mm × 450 mm

2

Raft

a. Size b. Thickness c. Beam size

243.16 m2 600 mm 600 mm × 600 mm

3

Concrete

a. For columns b. For beams and slabs c. For raft

M25 M25 M25

4

Shell

a. Slab

125 mm

196

G. P. Awadhya et al. Table 2. List of seismic parameters

Category

Parameter

Zone

V

Zone factor

0.36

Importance factor

1.2

Response reduction factor

5

Vertical irregularity in geometry

NO

Soil type

II i.e., Medium soil or Stiff soil

Table 3. Load combinations S. no

Load combination

1

1.5DL

2

1.5(DL+ LL)

3 4

S. no

Load combination

S. no

Load combination

8

1.5(DL-EQx)

15

1.2(DL+LL+RSx)

9

1.5(DL+EQy)

16

1.2(DL+LL+RSy)

1.2(DL+LL+EQx)

10

1.5(DL-EQy)

17

1.5(DL+RSx)

1.2(DL+LL-EQx)

11

0.9DL+1.5EQx

18

1.5(DL+RSy)

5

1.2(DL+LL+EQy)

12

0.9DL+1.5EQy

19

0.9DL+1.5RSx

6

1.2(DL+LL-EQy)

13

0.9DL-1.5EQx

20

0.9DL+1.5RSy

7

1.5(DL+EQx)

14

0.9DL-1.5EQy

5.2 Models 5.2.1 NSSI (Non-soil Structure Interaction) In this G + 10 model, the structure’s base is rigid, i.e., a fixed support is allocated to the structure’s base as shown in Fig. 2. This model is assigned different load scenarios, such as dead load, live load, and seismic loads, employing 20 distinct load combinations, as previously indicated [12]. The response spectrum method was used to model this construction, and it passed the concrete frame design examination [13].

Soıl Structure Interactıon Effects on the Seısmıc Response

197

Fig. 2. NSSI model.

5.2.2 SSI (Soil Structure Interaction) In this G + 10 model at the base of the structure raft foundation is designed using Bowles method. Then using the empirical formulae foe stiffness of the spring, area springs were assigned underneath the raft foundation and the line springs were also assigned at the base of the raft foundation along the beams connecting the columns. The load combinations were same as in the fixed support model. The model shown in Fig. 3 personifies the soil structure interaction study [14].

198

G. P. Awadhya et al.

Fig. 3. SSI model.

6 Results and Discussion 6.1 Story Drift The largest comparable displacement of each level divided by the height of the floor in question is the story drift ratio. For a 10-story building, the maximum story drift response plots of fixed and flexible bases are evaluated and scrutinized. The response spectrum approach is used to calculate the seismic response plots in accordance with IS 1893:2006 part 1. For fixed and flexible base structures, the maximum story drift response values are 0.001448 and 0.001735, as shown in Fig. 4 and Fig. 5 rspectively [15].The drift ratio increases gradually and reaches its maximum value at the second storey level, as seen in the graph. The intensity of the story drift ratio increases as soil stiffness decreases. Because the underneath soil state varies from stiff to soft, the story drift ratio is higher for flexible mat foundation. The upward trend is more pronounced in the upper and lower storeys. As the number of stories rises, the Soil Structure Interaction effect on the tale drift ratio becomes increasingly significant [16].

Soıl Structure Interactıon Effects on the Seısmıc Response

Fig. 4. Story drift (SSI)

Fig. 5. Story drift (NSSI).

199

200

G. P. Awadhya et al.

6.2 Story Shear The lateral load imposed by seismic waves is depicted per storey by story shear [17]. The maximum storey shear value for fixed and flexible foundation buildings can be calculated using the response spectrum approach. The narrative shear values rise as soil stiffness falls. The flexible base foundation’s minimum story shear is -623.423703 KN (See Fig. 6), which appears in the first floor of a 10-story structure, while the rigid base foundation’s minimum story shear is -637.05047 KN (See Fig. 7), which appear in the ground floor of the building [18].

Fig. 6. Story shears (SSI)

Fig. 7. Story shears (NSSI)

Soıl Structure Interactıon Effects on the Seısmıc Response

201

6.3 Maximum Story Displacement The SSI model amplifies lateral displacements and inter-story drifts, especially for MRF buildings resting on relatively soft soil. This increase in lateral deformations may affect the building frames’ performance. For a G + 10 storey building, the maximum story displacement response plots of fixed and flexible foundation are explored. The displacement increases in SSI models; the displacement grows more in foundations located on soft soil, and this value reduces as soil stiffness increases. For fixed and flexible foundation constructions, the maximum story displacement response values are 18.68255 and 56996658655 respectively (See Fig. 8 and Fig. 9). Because of the flexible basis, the story displacement value for the flexible base construction exceeds the IS code limit. In accordance with IS code 1893:2016 (part 1), maximum story displacement should be less than 0.004 times the height of the building [19].

Fig. 8. Maximum story displacement (SSI)

Fig. 9. Maximum story displacement (NSSI)

202

G. P. Awadhya et al.

6.4 Story Overturning Moment Moments, uplift forces, and shears applied to structures cause the footing to become unstable, resulting in the footing turning over and the structure failing during highintensity seismic waves. In the case of rigid foundations, however, these overturning moments have no influence on the structure. Resisting moments at the soil-foundation contact help to sustain the footing and prevent it from overturning. Fixed and flexible bases have maximum overturning moments of 0 KNm and 24.846326 KNm, respectively (See Fig. 10 and Fig. 11). These fixed and flexible base values demonstrate the need of addressing soil structure interaction when constructing a structure in order to achieve a stable and efficient structure. The overturning moment increases as the stiffness of the soil increases, as seen by these data and respective graphs. At the edge of the footing, these overturning tests are presented [20].

Fig. 10. Story Overturning Moment (SSI)

Soıl Structure Interactıon Effects on the Seısmıc Response

203

Fig. 11. Story overturning moment (NSSI)

6.5 Soil Pressure The precise estimate of contact pressure for flexible foundations is a difficult task. As a result, Terzaghi has assumed that the subgrade response modulus (Ks) for the entire footing remains constant. This means that the pressure ratio, as well as the factor of safety and settlement, will remain constant at all positions of the footing. The maximum soil pressure intensity, 140 KN/m2 , is found near the center of the footing (See Fig. 12). As a result, the highest displacement will occur at this spot, which is known as the dished effect. As we travel away from the center, both the soil pressure and the displacement decrease. Only 25% of the mat foundation surface area is influenced by each node of a perfect square or rectangular foundation baseThe soil pressure at each site can alternatively be determined by dividing the support base response by the tributary area at that point. Because the rigid foundation settles flat, the distribution of the subgrade reaction across the foundation base must be planar for permanent foundations [22]. Flexible foundations, on the other hand, do not have a linear subgrade reaction since their base is dependent on the foundation’s compressibility and structural stiffness. For studying the effect of Soil and Structure Interaction, it is critical to design a flexible foundation that takes into account the effect of modulus of subgrade reaction. The relative displacement of the structure decreases as the structural rigidity of the foundation increases, and the soil pressure beneath the structure decreases. Even with higher modulus of subgrade reaction the pressure distribution also reduces. Which means the Ks will take up more stress at the point of load application [21].

204

G. P. Awadhya et al.

Fig. 12. Soil pressure of SSI model

6.6 Column Moments The main sources of moments in the columns under gravity action are lateral stresses from wind or seismic waves, as well as eccentricity of the supported beams. The strength and bending qualities of the columns are monitored by the concrete’s compressive strength, column shape, and even the longitudinal strength of the reinforcements [23]. Moment 3-3 is the bending of a beam or frame around the 3-3 axis (See Fig. 14 and Fig. 16), while Moment 2-2 denotes the bending of a column on the positive and negative 2 faces around the 1 axis of consideration (See Fig. 13 and Fig. 15). The largest peak moment occurs along the 2-2 axis in the local direction. Because of larger displacement, overturning moments, and other factors, these moments are higher in flexible base structures.

Soıl Structure Interactıon Effects on the Seısmıc Response

Fig. 13. M 2-2 (SSI)

205

206

G. P. Awadhya et al.

Fig. 14. M 3-3 SSI

Soıl Structure Interactıon Effects on the Seısmıc Response

Fig. 15. M 2-2 NSSI

207

208

G. P. Awadhya et al.

Fig. 16. M 3-3 NSSI

Soıl Structure Interactıon Effects on the Seısmıc Response

209

7 Conclusion Seismic forces were applied to a G + 10 storeyed building frame with a fixed and flexible base, which were analyzed and developed using ETABS software. Using the plots indicated above, the seismic responses of the building frames such as story shear, story moment, story displacement, story drift, column moments, base reactions, axial forces, and soil pressure were analyzed. When soil structure interaction is taken into account, the study of the soil-structure model yields a wide range of responses. The fixed support model’s values were within limitations, i.e., maximum storey displacement was less than 0.004 * building height, and storey drifts were less than 0.004. Fixed base buildings had lower lateral deflection, storey drifts, base shear, and moment values than flexible foundation buildings. When the soil condition at the site is not adequate, i.e., when the structure rests on loose soil, the soil structure interaction effect must be considered.This research could be useful in developing safe design guidelines for seismic design of structures that take into account soil flexibility. The Response Spectrum approach is used to investigate the effects of slab, column, mat foundation, structure, and soil model components (IS 1893 part1:2006). The findings obtained using a fixed base model and a flexible base model assumption are compared to analyze soil structure interaction. The stiffness of the foundation base reduces as the soil springs are added, but the fundamental period of the structure increases, implying that the fundamental period of the structure is not only a function of structure height, but also of soil condition. As a result, it improves the concept of modeling the structure while considering the soil-structure interaction concept. SSI models will thus have a longer period than NSSI models. The story displacement response of SSI models with flexible base foundations is higher than that of fixed base models (NSSI). The consequences of soil structure interaction become exacerbated as the structure’s storey increases. Soil structure interaction has a greater impact on the bottom storey displacements than on the rest of the storeys. As a result, if the SSI principle is not considered while designing multistorey buildings, they are far more likely to fail. With a decrease in soil stiffness and even an increase in the structure’s storey, the story drift ratio rises. The tale drift ratio of the SSI model with flexible base was higher than the fixed base model (NSSI). The effects of SSI on overturning moments, base reactions, and column moments are then evaluated using this model. In comparison to the conventional hypothesis, which ignored soil structure interaction, the investigation demonstrates that the SSI has a significant impact on every structural response. The structure’s dynamic responses differ from the fixed base condition. The seismic behavior of mid-rise structures is heavily influenced by these dynamic responses. Thus, SSI effects are critical for seismic design of mid-rise structures, especially when they are built on soft soil bases and in high seismic zones like Seismic Zone IV and V. If SSI is not taken into account during analysis and design, the precision in determining structural safety when confronted with seismic waves may not be consistent. As a result, while designing building frames for seismic stresses, a proper foundation system that takes into account the influence of soil stiffness ”must be used.

210

G. P. Awadhya et al.

References 1. Wong, F.S.: Uncertainties in dynamic soil-structure interaction. J. Eng. Mech. 110(2), 308– 324 (1984). https://doi.org/10.1061/(asce)0733-9399(1984)110:2(308) 2. Wong, B.H.L.: “In the absence of soil-structure interaction effects have been presented by Mita and Luco (1990a, 1990 ) and Luco et al. (1991). The form and effectiveness of active control in the presence of soil-structure interaction effects are examined by considering structures with different heights supported on soils with different rigidities. The effects on the response of the structures and on the amplitudes of the required control Structural Response with Absorbing Boundary To formulate and solve the complete problem, we first consider the motion of the superstructure for a given translation and rotation of the base. Let u ( x ) e’ °  equal the harmonic translation of the structure with respect to an inertial frame of reference . For small displacements, the total displacement u ( x ) can be decomposed into u ( x ) = u B + xd B + u R ( x ),” vol. 117, no. 10, pp. 2237–2250 (1992) 3. Pilecki, T.J.: Interface model for dynamic soil-structure interaction. J. Geotech. Eng. 113(5), 537–538 (1987). https://doi.org/10.1061/(ASCE)0733-9410(1987)113:5(537) 4. Dicleli, M., Albhaisi, S., Mansour, M. Y.: Static soil–structure interaction effects in seismicisolated bridges, 10(1), 22–33 (2005) https://doi.org/10.1061/(ASCE)1084-0680(2005)10 5. S. S. I. Analysis and P. For, . “i: a,” vol. 125, no. January, pp. 26–37, (1999) 6. Jonathan, P., Gregory, L., Raymond, B., Stewart, B.J.P., Fenves, G. L., Seed, R.B.: S Eismic S Oil -S Tructure I Nteraction in B Uildings, vol. 125, pp. 38–48 (1999) 7. Renzi, S., Madiai, C., Vannucchi, G.: A simplified empirical method for assessing seismic soil-structure interaction effects on ordinary shear-type buildings. Soil Dyn. Earthq. Eng. 55, 100–107 (2013). https://doi.org/10.1016/j.soildyn.2013.09.012 8. Ghosh, B., Madabhushi, S.P.G.: Centrifuge modelling of seismic soil structure interaction effects. Nucl. Eng. Des. 237(8), 887–896 (2007). https://doi.org/10.1016/j.nucengdes.2006. 09.027 9. Wolf, J.P., Song, C.: Some cornerstones of dynamic soil-structure interaction. Eng. Struct. 24(1), 13–28 (2002). https://doi.org/10.1016/S0141-0296(01)00082-7 10. Spyrakos, C.C., Maniatakis, C.A., Koutromanos, I.A.: Soil-structure interaction effects on base-isolated buildings founded on soil stratum. Eng. Struct. 31(3), 729–737 (2009). https:// doi.org/10.1016/j.engstruct.2008.10.012 11. Tang, Y., Zhang, J.: Probabilistic seismic demand analysis of a slender RC shear wall considering soil-structure interaction effects. Eng. Struct. 33(1), 218–229 (2011). https://doi.org/ 10.1016/j.engstruct.2010.10.011 12. Ganjavi, B., Hao, H.: A parametric study on the evaluation of ductility demand distribution in multi-degree-of-freedom systems considering soil-structure interaction effects. Eng. Struct. 43, 88–104 (2012). https://doi.org/10.1016/j.engstruct.2012.05.006 13. Stefanidou, S.P., Sextos, A.G., Kotsoglou, A.N., Lesgidis, N., Kappos, A.J.: Soil-structure interaction effects in analysis of seismic fragility of bridges using an intensity-based ground motion selection procedure. Eng. Struct. 151, 366–380 (2017). https://doi.org/10.1016/j.eng struct.2017.08.033 14. Sucasaca, J., Sáez, E.: Topographical and structure-soil-structure interaction effects on dynamic behavior of shear-wall buildings on coastal scarp. Eng. Struct. 247, 113113 (2021). https://doi.org/10.1016/j.engstruct.2021.113113 15. Kausel, E.: Early history of soil-structure interaction. Soil Dyn. Earthq. Eng. 30(9), 822–832 (2010). https://doi.org/10.1016/j.soildyn.2009.11.001 16. Zogh, P., Motamed, R., Ryan, K.: Empirical evaluation of kinematic soil-structure interaction effects in structures with large footprints and embedment depths. Soil Dyn. Earthq. Eng. 149, 106893 (2021). https://doi.org/10.1016/j.soildyn.2021.106893

Soıl Structure Interactıon Effects on the Seısmıc Response

211

17. Matinmanesh, H., Asheghabadi, M.S.: Seismic analysis on soil-structure interaction of buildings over sandy soil. Procedia Eng. 14, 1737–1743 (2011). https://doi.org/10.1016/j.proeng. 2011.07.218 18. Luco, J.E., Trifunac, M.D., Wong, H.L.: Isolation of soil-structure interaction effects by fullscale forced vibration tests. Earthq. Eng. Struct. Dyn. 16(1), 1–21 (1988). https://doi.org/10. 1002/eqe.4290160102 19. Wong, H.L.: Wong-SFSIThesisCaltech.pdf 20. Stewart, J.P., Fenves, G.L.: System identification for evaluating soil-structure interaction effects in buildings from strong motion recordings. Earthq. Eng. Struct. Dyn. 27(8), 869–885 (1998). https://doi.org/10.1002/(SICI)1096-9845(199808)27:8%3c869::AID-EQE762%3e3. 0.CO;2-9 21. Baker, J.W.: Measuring bias in structural response caused by ground motion scaling. In: Pacific Conference on Earthquake Engineering, no. 056, pp. 1–6 (2007). https://doi.org/10.1002/eqe 22. Stewart, J.P.: Soıl-Structure Interactıon Effects by Raymond, B., Seed and Gregory, L. Fenves Report No . PEER-98/07 Pacific Earthquake Engineering Research Center University of California Berkeley, California November 1998 ii, Environ. Eng., no. November (1998) 23. Isbiliroglu, Y., Taborda, R., Bielak, J.: Coupled soil-structure interaction effects of building clusters during earthquakes. Earthq. Spectra 31(1), 463–500 (2015). https://doi.org/10.1193/ 102412EQS315M

Study on Treatment of Ground Subsidence in Goaf Hong Qiang Zhang(B) , Ling Gao, and Ming Yue Feng Hebei Provincial Communications Planning, Design and Research Institute Co., Ltd, Shijiazhuang 050011, China [email protected]

Abstract. In the construction of expressway, the ground collapse caused by underground mining restricts the route selection of expressway. Therefore, before the construction of expressway, it is necessary to find out the impact of Goaf Collapse on expressway construction. In order to reduce the risk of geological disasters, safeguard social and public interests and alleviate social contradictions, it is very necessary to carry out gob survey and treatment research. Find out the distribution of goaf through data collection, investigation and visit, measurement, engineering geological mapping, geophysical exploration, drilling, in-situ test, sampling test and other methods, and obtain the optimal treatment scheme in combination with the actual situation of the site and relevant engineering experience for peer reference. Keywords: Goaf · Ground Subsidence · Expressway · Treatment

1 Introduction Due to the failure to fill in time after the underground goaf is formed, the original stress balance state of the overburden is damaged. When highway construction is carried out above the goaf, when the influence depth of subgrade load overlaps with the caving zone and fracture zone of the goaf, the goaf may be “activated”. The “activation” process of goaf is difficult to observe directly. Its “activation” process is affected by a variety of natural and human factors. The impact on the surface is quite complex and a longterm process, which may occur a few years or decades or even hundreds of years after mining. As long as there is residual deformation in the “activated” goaf, it is still possible to reactivate, The main threat objects are the life and property safety of passing vehicles and pedestrians and the stability of expressway subgrade. Therefore, it is necessary to control the goaf [1–3]. This paper investigates and studies the treatment scheme for the ground collapse of goaf passing by an expressway.

2 Engineering Geological The overlying loess soil layer in the exploration area is 0.5–12 m thick, with slight to medium collapsibility and a small amount of gravel in some parts; The overlying © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 212–220, 2023. https://doi.org/10.1007/978-981-19-3983-9_18

Study on Treatment of Ground Subsidence in Goaf

213

rock mass in the goaf is mostly gray and gray white strongly ~ moderately weathered dolomitic limestone with brecciated dolomitic limestone and limestone. Affected by the intrusion of tectonic magma, the rock mass is relatively broken ~ extremely broken, mostly in fragment shape, a small amount in column shape, and mixed with multi-layer iron bearing amphibolite and diorite vein, which is irregularly distributed. 2.1 Meteorology and Hydrology The exploration area has a temperate monsoon climate, with large temperature difference and four distinct seasons. It is dry and windy in spring, hot and rainy in summer, cool and cool in autumn, cold and dry in winter with little rain and snow. The annual average temperature is 13.5 °C, and the average temperature in the coldest month is − 2.3 °C. The average temperature in the hottest July is 26.9 °C, the extreme minimum temperature is −19.9 °C, and the extreme maximum temperature is 42.5 °C; The annual average precipitation is about 549 mm, with uneven distribution within the year and large interannual variation. The annual precipitation is concentrated from June to September, accounting for about 75% of the total annual precipitation, and mostly in the form of rainstorm. The annual sunshine is 2200−2700 h, and the frost free period is 191−215 d. Affected by the high-pressure air flow in Inner Mongolia in winter, this area is mostly northwest wind in winter and southwest wind in spring and autumn. Strong winds of force 7−8 often occur, with an annual average wind speed of 2.6 m/s. The maximum frozen soil depth is 0.53 m. Nanming River is about 2 km to the southwest of the exploration area, which originates from the south foot of Motianling in the territory, turns from northwest to southeast to northeast, and intersects Beiming River to the east of Yonghe village, with a total length of 95 km. This river is a seasonal river. Except for a small amount of spring water in the upstream guantaochuan, the rest are dried up for a long time, and the flood is discharged only in flood season, with an annual runoff of 71.8 million m3 . Surface water flows in the river in rainy season and fluctuates with rainfall. 2.2 Topographic Features The project area is located at the eastern foot of Taihang Mountain, which is surrounded by xiaomotianling and Shibapan mountains, with an Intermountain basin in the middle. The exploration area is located in the gentle slope zone of low mountains, heavy hills and belongs to the low mountain area (II 2) with erosion structure dominated by carbonate rocks. The terrain is high in the West and low in the East. The elevation is 433.2−448.3 m. The surface vegetation is not developed, and more crops are planted. 2.3 Formation Lithology The bedrock at the top of the exploration area is exposed, and most of it is covered with quaternary Middle Pleistocene slope diluvium. The strata from new to old are described as follows: Quaternary Middle Pleistocene proluvial and Deluvial loess like silty clay and silty clay (Q2 pl + dl ): distributed in the slope zone of the hillside, with maroon silty clay at the

214

H. Q. Zhang et al.

bottom and gravel layer. The soil layer has vertical joints, the bedding is not obvious, has the general characteristics of Malan loess, and has slight to medium collapsibility. The thickness varies, mostly 0.5−12 m. Middle Ordovician Majiagou Formation (O2): gray and gray white dolomitic limestone with breccia dolomitic limestone. Affected by structural intrusion, the rock mass is relatively broken ~ extremely broken, mostly in fragment shape, a small amount in column shape, and mixed with multi-layer iron bearing amphibolite and dioritevein. Yanshanian magmatic rocks (δ52 ): magma intrudes along the NNE trending fault beam, and the surface is irregularly distributed locally. The lithology is complex, mainly diorite and amphibolite, intercalated with multi-layer iron bearing amphibolite. Ore bodies occur in veins in amphibolite metamorphic strata and are irregularly distributed. 2.4 Neotectonic Movement and Earthquake There is a small north-east transverse concealed normal fault 500 m south of the bridge site area. The overall strike of the fault is 23−36° NNE and the dip angle is 55−75°. The hanging wall of the fault is Sinian chert dolomite and dolomitic limestone, and the footwall is Ordovician Majiagou Formation limestone. In the route area, the southwest fault trend gradually changes from NNE to NE, and is concealed under the overburden. Through this survey, the fault is not a Holocene Active fault. Affected by the fault, structural breccia is developed in the Middle Ordovician (O2 ) limestone layer in the bridge site area, structural fractures are developed along the rock contact zone, water erosion marks on the contact surface are obvious, and dissolution is generally developed. Geological Structure. The exploration area is located in the North Korean paraplatform I2 of the first-class tectonic unit, the Xishan fault uplift II 23 of the second-class tectonic unit, the Taihang arch short beam III 211 of the third-class tectonic unit, and the Wu’an concave fault beam IV 233 of the fourth-class tectonic unit. Most faults strike NNE and NW, and the development degree is weak in the West and strong in the East. Most magmatic rocks intrude along the faults in the East. The intrusive rocks in the exploration area intrude along the NNE trending fault bundle, and the surface presents irregular local distribution. Ore bodies occur in veins in amphibolite metamorphic strata and are irregularly distributed. Earthquake. According to the county chronicles, there has been no earthquake above magnitude 6 in and around the exploration area, which belongs to a relatively stable area. According to the seismic ground motion parameter zoning map of China (GB 18306– 2015), the basic seismic fortification intensity in the exploration area is VII and the basic seismic acceleration is 0.15 g [4].

2.5 Hydrogeological Conditions Surface Water. Groundwater in the exploration area is mainly bedrock fissure water and pore phreatic water in Quaternary loose overburden. Phreatic water mainly occurs in the contact part between slope diluvium and bedrock. Due to the high terrain, there is little

Study on Treatment of Ground Subsidence in Goaf

215

occurrence; Bedrock fissure water mainly occurs in shallow moderately weathered rock fissures and goaf. The recharge source is atmospheric precipitation, which is discharged to low-lying areas, and the dynamic change of groundwater is greatly affected by climate factors. There was no surface water during the survey. After the rainstorm in June 2017, the deep goaf was filled with groundwater. According to the water quality analysis results of water samples taken from nearby sites, groundwater is slightly corrosive to steel bars in concrete structures and reinforced concrete structures. To sum up, the conditions of groundwater infiltration, runoff and discharge in the exploration area are good, and its circulation and discharge are fast. The rapid flow of groundwater makes the shallow rock and soil in the goaf soften, disintegrate and sink. Especially under the condition of rainstorm, the rainwater pours back, and the messy pillars in the original goaf are washed away and collapsed. In addition, the surface scouring induces the further development of ground cracks and collapses.

3 Ground Subsidence in Goaf 3.1 Overview of Mineral Exploitation According to the surface mapping and survey visit, the iron ore mining is composed of four main mines in the periphery, one main mine in the middle and three to four small mines, of which the main mine 1 is located near 70 m to the right of K25 + 042; The main mine 2 is located near 30 m to the left of K25 + 054; Main mine 3 is located near 60 m to the left of K25 + 228; The main mine 4 is located near 153 m to the right of K25 + 184; The middle main mine 5 is located near 26 m to the right of K25 + 172. The small mines are basically buried, and the four main mines in the periphery are mined in the middle. The mining methods are shaft and ramp, and the mining roadway and main mining area are basically without support or temporary support. The main mining area of the iron mine is banded, located in the west of K25 + 100−K25 + 200, with a width of 60−100 m, a buried depth of 60−100 m and a maximum mining thickness of 20 m. All mines are mined at this banded vein, and a small amount of chicken nest mining is carried out in other places, which is irregular. 3.2 Distribution Characteristics of Goaf Surface Collapse Collapse area: it is banded, and it basically intersects vertically at K25 + 160−k25 + 170. The intersection is narrow and 10 m wide. The main collapse part is 60−100 m west of K25 + 100−K25 + 200, 50–60 m wide. Two collapse pits are formed in the middle of ore body mining. Collapse pit 1 is 60−90 m right of K25 + 120−K25 + 160, about 20−30 m wide. The edge surface is mostly messy, and a collapse pit is formed at the bottom, with a depth of 15−20 m, It is now a garbage dump, which has not been filled; The collapse pit 2 is located at 120−130 m on the right side of K25 + 150−K25 + 170, 10 m wide and 5−10 m deep. It has been filled with garbage and there is no obvious trace on the surface.

216

H. Q. Zhang et al.

Crack area: located at the edge of the collapse area and the extension part of the depression, 5 cracks are mainly distributed. Crack 1 is located 40m to the left of K25 + 133, intersects the route K25 + 125, runs 40° to the northwest, nearly parallel to the collapse edge, extends 140m to 118 m to the right of K25 + 125, with a width of 10−50 cm and a visible depth of 1−3 M; Crack 2 is located 56.8 m to the left of K25 + 150, intersects the route K25 + 148, extends 52° to the northwest, extends 55 m to the collapse edge, with a width of 3−10 cm and a visible depth of 0.5−2 m; Crack 3 is located 58 m to the left of K25 + 164, intersects the route K25 + 174, extends 45−50° to the northwest, extends 120 m along the collapse edge to the tail end of crack 1, with a width of 5−20 cm and a visible depth of 0.5–3M; Crack 4 is located 41 m to the left of K25 + 225, intersects K25 + 210, extends 72 m to the northwest 45° to the collapse edge. Generally, the main mine 3 extends to 1−2 small mines, with a width of 10−50 cm, the widest part is nearly 1m, the visible depth is 2−4 m, and the surrounding loess has been covered by collapse; Crack 5 is located 12 m to the left of K25 + 232, intersects the route K25 + 230, extends 22° to the northwest, extends intermittently from main mine 3 to main mine 4, and is mostly covered by loess collapse. The local visible width is 3−6 cm and the visible depth is 0.5−1.5 m. In addition, some fine cracks are distributed at the collapse edge, which is caused by shallow collapse and tension crack. 3.3 Cause Analysis of Goaf Ground Collapse Most of the mining roadway and goaf have no support or temporary support, generally no filling, and the roof of the goaf is free to collapse. The overlying rock mass of the goaf is strongly ~ moderately weathered dolomitic limestone with breccia limestone. Affected by the intrusion of tectonic magma, the rock mass is relatively broken ~ extremely broken, mixed with multi-layer iron bearing amphibolite and diorite vein. The engineering geological properties are poor, and the soil is above it. Therefore, the goaf surface will produce strong displacement, settlement and deformation. When the collapse zone and fracture zone develop into the soil layer, collapse will occur under the action of strong permeability of drainage. Since the mining has been terminated for nearly 15 years, and after the rainstorm in June 2017, the rainwater is poured along the cracks or collapse of the goaf, and there is no obvious change in the collapse area, indicating that the two collapse areas have completely collapsed and have no further development. Therefore, the hidden danger body in the current goaf has been basically stable, the impact of continuous deformation is light, and the impact on the geological environment is small. However, when engineering construction is carried out above the goaf, when the influence depth of building load overlaps with the caving zone and fracture zone of the goaf, it may cause “activation” of the goaf, cause large land subsidence and deformation, and cause building damage.

4 Ground Subsidence in Goaf According to the survey visit and survey results, combined with the collapse and fracture development form, it is determined that the goaf belongs to small kiln goaf. According

Study on Treatment of Ground Subsidence in Goaf

217

to technical rules for highway design and construction in goaf (JTG/T d31–03-2011), the limit equilibrium analysis method is adopted for the roadway in goaf, and the critical depth Hcr to maintain the stability of roadway roof is calculated according to appendix d.0.5, so as to obtain the stability coefficient of roof in goaf [5]. After exploration and analysis, the cavities under the subgrade are divided into three types: residual shallow branch roadway, deep branch roadway, deep main roadway and goaf. The width of branch roadway is 2−5m, and the width of main roadway and goaf is 5−10 m. According to the technical rules for design and construction of goaf highway (JTG/T d31–03-2011), the limit equilibrium analysis method is adopted for goaf roadway. When h reaches a certain depth, the roof rock stratum just maintains natural balance (i.e. q = 0), at this time, h is called the critical depth Hcr. Calculate the critical depth Hcr to maintain the stability of roadway roof according to appendix d.0.5, as shown in formula 1.    μ tan ϕ Bγ + B2 γ 2 + 4Bp0 γ 1−μ   (1) Hcr = μ tan ϕ 2γ 1−μ where: B - roadway width in goaf (m); P0 - subgrade base pressure of highway, including driving load, subgrade and pavement load (kPa); γ —Weighted average weight of overburden (kN/m3); ϕ—Weighted average internal friction angle of overlying strata (°); μ—Weighted average Poisson’s ratio of overlying strata. Among them, the roadway width adopts the upper limit of the range known in the investigation, and the highway base pressure is the sum of the pressure generated by the Quaternary overburden (10 kPa) and the driving load (20 kPa). According to the geotechnical physical and mechanical indexes, the weighted average gravity, average internal friction angle and average Poisson’s ratio of the overburden are taken as 25 kN/m3, 55° and 0.20 respectively. The average internal friction angle of overlying strata is given by referring to the similar friction angle of limestone in the engineering geology manual. The similar internal friction angle takes into account the cohesion of rock. It is concluded that the roof stability evaluation result of shallow support roadway is unstable, which needs to be treated by engineering measures, and the roof of deep main roadway is basically stable, which cannot be treated; There is no need to deal with the stability of goaf roof in deep branch roadway.

5 Study on Treatment Scheme of Ground Collapse in Goaf The small pile abutment is flat, with pebbles on the surface and bedrock under it. Good abutment stability. The slope of abutment with large chainage is about 37°, the surface is exposed as syenite porphyry, the rock stratum is stably distributed, and no adverse geology such as landslide and collapse is found. Good abutment stability. Pile foundation is adopted in abutment design, which can effectively increase abutment stability and reduce abutment slope sliding risk. If the gravity abutment is adopted, the sliding and

218

H. Q. Zhang et al.

overturning stability shall be checked. Slope stability shall be considered in abutment subgrade excavation and bridge head cone slope design. 5.1 Governance Principles According to the surface mapping and survey visit, the iron ore mining is composed of four main mines in the periphery, one main mine in the middle and three to four small mines, of which the main mine 1 is located near 70 m to the right of K25 + 042; The main mine 2 is located near 30 m to the left of K25 + 054; Main mine 3 is located near 60 m to the left of K25 + 228; The main mine 4 is located near 153 m to the right of K25 + 184; The middle main mine 5 is located near 26 m to the right of K25 + 172. The small mines are basically buried, and the four main mines in the periphery are mined in the middle. The mining methods are shaft and ramp, and the mining roadway and main mining area are basically without support or temporary support. Ensure the scientificity, feasibility and rationality of the prevention and control project. The gob collapse prevention and control project shall be planned uniformly, highlight the key points according to the geological environment background conditions, adjust measures to local conditions, and take comprehensive treatment measures. The prevention and control scheme shall have the characteristics of reliable technology, reasonable economy, simple structure and strong operability. Comprehensive prevention and control should implement the principle of combining engineering measures with administrative measures. 5.2 Common Treatment Schemes Grouting Method. Grout materials with filling and cementation properties are injected into cavities, collapse zones, fracture zones and geotechnical voids in goaf manually, so as to increase their strength after hardening, support overlying strata, and control and reduce the damage of ground deformation to ground buildings. It is applicable to the treatment of goaf and roadway without underground construction conditions. Dry Masonry Method. The abandoned but not fully collapsed roadway with underground construction conditions shall be backfilled and built manually with rubble. Mortar Method. It is mainly used for the roadway with defective supporting structure. Excavation and Backfilling Method. Excavate the shallow buried goaf or roadway, and then backfill with dry masonry, mortar masonry or other methods. Bridge Span Method. Span the influence range of goaf or roadway in the form of bridge. It is applicable to goaf or roadway with shallow buried depth (several meters to tens of meters).

Study on Treatment of Ground Subsidence in Goaf

219

5.3 Governance Scheme Adopted for the Project In recent years, with the development of China’s expressway construction, Shanxi and Henan have many successful examples in the management of Expressway goaf, such as Zheng Shao line in Henan, Chang Jin line, Yang Hou line, Fen off-line line, Lijun line and Dayun line in Shanxi. The buried depth of the managed goaf ranges from tens of meters to more than 500 m, which have obtained successful experience. The buried depth of the goaf of the project is large, the width of the influence zone of the goaf is large, the roof has collapsed after mining, and obvious ground collapse has been formed. Obviously, mortar masonry, dry masonry, excavation and backfilling are not available. If the bridge span method adopts pile foundation, the stable stratum of the goaf floor needs to be used as the pile end bearing stratum. The pile length is too large, and it is difficult to cast into pile due to the existence of cavity, which is not technically feasible. If long-span bridges are used, the construction is difficult and the cost is high due to the large span; The goaf does not have the conditions for underground construction. Combined with the successful experience of previous projects, the grouting method full filling pressure grouting method should be selected for the project.

6 Abutment Stability Evaluation The small pile abutment is flat, with pebbles on the surface and bedrock under it. Good abutment stability. The slope of abutment with large chainage is about 37°, the surface is exposed as syenite porphyry, the rock stratum is stably distributed, and no adverse geology such as landslide and collapse is found. Good abutment stability. Pile foundation is adopted in abutment design, which can effectively increase abutment stability and reduce abutment slope sliding risk. If the gravity abutment is adopted, the sliding and overturning stability shall be checked. Slope stability shall be considered in abutment subgrade excavation and bridge head cone slope design.

7 Conclusion The exploration of the project has collected the preliminary work results, basic geological data and hydrogeological data around. Through investigation and visit, geological mapping, geophysical exploration, drilling and other exploration means, the geological environment background conditions of the exploration area have been basically found out. Through the analysis of the ground collapse stability of goaf, the treatment schemes of ground collapse in different goaf areas have been studied, Finally, the grouting method - full filling pressure grouting method is selected for treatment.

References 1. Gou, D., Tian, J., Li, J., et al.: Grouting treatment technology in goaf of small coal mine under expressway tunnel. J. For. Eng. 3(5), 142–149 (2018) 2. Yan, X., Ke, B., Luo, l., et al.: Construction and quantitative analysis of stability evaluation index system of a goaf. Mod. Min. (8), 194–196 (2016)

220

H. Q. Zhang et al.

3. Yin, L., Li, W., Yang, X., et al.: Application of grouting method in foundation treatment engineering in goaf 4. Seismic ground motion parameter zonation map of China (GB 18306–2015) 5. Technical rules for design and construction of goaf highway (JTG/T d31–03–2011)

Forward Modeling and Detection of the Potential Slip in Artificial Slope by GPR Zhenjun Zhang1 , Luo Ye1 , and Gao Lv1,2(B) 1 Mechanical Engineering College, Xi’an Shiyou University, Xi’an, China

[email protected] 2 Shaanxi Key Laboratory of Geotechnical and Underground Space Engineering, Xi’an

University of Architecture and Technology, Xi’an, China

Abstract. Human activities have a huge impact on natural slopes and can easily cause natural disasters such as landslides. This leads to obvious environmental damage. Based on the method of Ground Penetrating Radar (GPR), the space position and filling properties of artificial high embankment slope slip planes were studied. With the numerical simulation method, the transient reflection-transmission rule and characteristics of geological radar waves were studied. The imaging differences of geological radar were caused by the inhomogeneous distribution of relative dielectric constant and conductivity of artificial layers. GPR data was closely related to the relative dielectric constant of each artificial layer. The echo data had less affected by the relative dielectric constant of the deep artificial layer. The spatial position of the slip plane can be precisely deduced by the hierarchical numerical model. These conclusions of the artificial slope slip surface should be an important guiding significance for detection. Keywords: GPR · Artificial slope · Finite-different time-domain method

1 Introduction Artificial slope was a new artificial slope in construction. It’s influenced by human factors, also limited by the bearing capacity of the artificial foundation. And it is probably slipped during the process of construction or project operations [1, 2]. The surface detection technique of the non -disturbance has great importance to monitor potential slips of the fill slope if it was prior considered [3–5]. In this paper, the Ground Penetrating Radar (GPR) of non-destructively detecting can confirm the slips’ space position and fill shape timely and effectively was introduced. The GPR of detecting in slope’s engineering mainly has the following results: Sass et al. used the GPR to collect large date to study the Alps slope slips [6], which are caused by unconformity layer and stratified cuttings of persistent layers, they also analyzed the slip plane’s thickness and internal structure of the slope of the Swabian Alb Jurassic sedimentary rock field [7], the slope’s marl surface are soil clay and limestone. Zieli´nski A et al. used the GPR of detecting to analyze a certain Poland highway slope [8], and they proposed warning measures for potential dangers by analyzing the slope’s internal © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 221–228, 2023. https://doi.org/10.1007/978-981-19-3983-9_19

222

Z. Zhang et al.

structural characteristics. Fankhauser K et al. used the method of the ground and airborne geological radar to analyze and determine the position of the bedrock of Switzerland’s steep slope [9], also the time and space distribution of water content of slope are analyzed with the method of conductivity in their analysis. CHEN Jigang et al. adopted a water injection system drilling TV system and the GPR to detect the depth of deep - mining and heavy - slanted seam floor, and digitally analyzed the number of floor cracks and borehole depth crack width and quantities before and the floor is mined [10]. Wang Yong-Qiang et al. used the GPR to detect the open mine slope and revealed the potential geological defects of the interior of the steep slope [11]. Li Daxin et al. speculated the causes of the slope’s collapsed by using the results of detecting Chongqing steel company material farm bank slope [12], which in virtue of high positioning accuracy and high resolution of the GPR. Recently, the research on the problem of sliding off the artificial slope of artificial fill slope by the positive inverse method is less studied. The influence laws of different compactness and water content of soil on the deep harmful geologic body was studied based on theoretical derivation numerical calculation and layered construction. The article also studied the spatial distribution of slope slip under laminar filling by FiniteDifferent Time-Domain and positive inverse method.

2 Impact of Human Engineering on the Environment Since the reform and opening up, China’s economic construction has made great achievements, but the development model of high input, high consumption and high pollution has directly led to the severe situation of excessive resource and environmental costs of economic growth. Human beings are increasingly aware of the negative impact of engineering construction on the environment, thus investing a lot of human and financial resources in design, materials, construction, recycling, supervision and other aspects of continuous efforts to achieve green ecology, and even give up economic benefits for the environment in order to achieve sustainable development. Engineering construction should be synchronized with ecological environment construction planning, synchronized implementation, synchronized development, realize the unity of economic benefits, social benefits, ecological benefits, environmental benefits, in the protection of development, in the development of protection, tend to benefit and avoid harm, to avoid shortcomings, to promote the coordinated development of engineering construction and ecological environment, to achieve the harmony of man and nature.

3 The GPR Echo Model of the Layered Electromagnetic wave reflection can be generated in different dielectric layers, the GPR wave will generate reflection in each interface, while the GPR wave in the refractive part will also be discounted gradually, according to this principle, the multilayer geologic model of artificial high-fill is established, and the optimized 2d multilayer geological model as Fig. 1.

Forward Modeling and Detection of the Potential Slip

223

Fig. 1. Geoelectric model of the multilayer fill.

Supposing the layer of the filling is evenly distributed, the relative permittivity is ε1–εn, the thickness is h1–hn and the one-way trip time of each layer of GPR is t1–in. The thickness of each filling is defined as follows: h = h1 + h2 + h3 + · · · + hn

(1)

And the total detection time is defined as follows: t = 2 × (t1 + t2 + t3 + · · · + tn ) h1 hn v1 = √cε = t ; vn = √cεn = t n 1

(2)

1

Finally, two-way travel time equation of GPR wave is as follows h c

=

t1 √ ε1

+

t2 √ ε2

+

t3 √ ε3

+···+

tn−1 √ εn−1

+

tn √ εn

h = h1 + h2 + h3 + · · · + hn−1 + hn t = 2 × (t1 + t2 + t3 + · · · + tn−1 + tn )

(3)

4 Reflection Mechanism of Special Medium Interface The difference of reflection is generated by different dielectric constants can be further analyzed. Based on GPRMAX2D, this section adopts the Finite-Different Time-Domain to carry out a more specific image analysis of the reflection characteristics of differences of dielectric constant. In this section, the three kinds of material air water, and metals, which have obvious dielectric characteristics in the artificial filling layer are analyzed, and these three substances can form reflection features in the artificial layer. Based on its characteristics, the geological model is established as Fig. 2.

224

Z. Zhang et al.

Fig. 2. Test model of the special medium interface reflection.

The basic setting of this model is as follows: the horizontal distance is 2.0 m, the detecting depth is 0.7 m, the unit grid is 0.0025 m × 0.0025 m, the boundary conditions are the perfect matching interface, the time depth is 40 ns, wavelet frequency is 900 MHz, the excitation source is the Ricker wavelet, the thickness of artificial is 0.35 m, reflector, and the special medium reflector is 0.35 m. The transmitting of GPR and receiving of the antenna in the surface of artificial scans from left to right, and then the reflection images can be drawn about the air-water and metal. At the same time, setting up a special medium reflector of pure soil as a contrast model, as shown in Fig. 3.

Fig. 3. (a)–(d) Echo characteristics of (a) loess layer, (b) Air layer, (c) Water layer and (d) Metal layer, respectively.

The images of different reflectors obtained by GPR are shown in Fig. 3, in which diagram (a) is a set of pure artificial without reflection, the diagram shows that, in the 1 ns–5 ns range, the waveform has very obvious change, this is the characteristics of the amplitude of Ricker wavelet. The whole waveform of 5 ns–40 ns is of uniformity, it indicates that there is no obvious reflection characteristic of geological radar wave inhomogeneous artificial. Diagram (b) shows that the GPR wave contacts with the air interface at a depth of 0.35 m, and then cause the waveform changes, the waveform and energy will invert when the geologic radar wave meets the air interface in the artificial. Diagram (c) shows the interface reflection of the GPR wave when it spreads from the artificial to the water at 0.35 m. Similarly, the water makes an obvious reflection of the GPR wave, about 99% of the waves generate reflection, the rest electromagnetic signals transmit and continue to spread, the deep image shows multiple reflected waves.

Forward Modeling and Detection of the Potential Slip

225

Diagram (d) shows the reflection of the metal interface encountered by GPR waves. The reflection of metal on electromagnetic waves is close to 1, so the electromagnetic wave will generate complete reflection, and form the most obvious reflection images. As a kind of new idea and new method of geotechnical engineering, groundpenetrating radar is convenient and flexible in specific engineering, this paper combines the feature of relative permittivity properties of artificial and the feature image of analysis and back analysis, also the typical applications in the project of the GPR is the high fill slope slip detection.

5 Application in Zhen’an Slope 5.1 Overview of the Landfill Project of Artificial Slope The Zhen’an Pumped Storage Project is located in Pusatian Village, Yuehe Town, Zhen’an County, Shangluo City. The slope of the slag site is mainly Platform 1. The #1 construction platform is mainly located near the upper and lower storage link road pile number 4 + 800. The site covers an area of about 49,000 m2 in the middle and lower sections of the Red Cliff Gulch. The site is centered on the Red Cliff Gulch, with lush vegetation on both sides of the slope. The highest elevation of the site is 1110 m and the lowest elevation is 990 m, with a maximum height difference of 120 m. The slope of the site is about 40°–60°, with a gentle slope up and down. The bottom of the trench is about 3 m–5 m wide. Longitudinal slope drops of 28%. The ditch has perennial flowing water. The trench has lush vegetation, bedrock outcrops. No accumulation of loose material. 1# Construction platform affected by terrain. The site can only be spread along the Red Cliff Gulch and both banks. The site mainly outcrops marble. The depth of weak weathering is generally greater than 15 m–20 m, the rock body is more complete, and the development of fissures is common. Within 1 m of the surface layer, the rock body is fragmented and locally weathered into gravelly soil with plant roots interspersed, as shown in Fig. 4. 5.2 Detection and Analysis of Slope Slip Surface of Artificial Landfill According to the field situation, given the sliding position of the artificial high fill slope, the internal slip plane should be east-west, therefore the lines from east to west are respectively set in the edge and inner edge of the filling platform, and lines connecting the edge and inside are also stetted, to comprehensively understand the space position of the deep sliding surface of the filled foundation. The primary detection of the 10–20 m underground was detected by 100 MHz antennas. The main data are as follows: Diagram (a), (b) in Fig. 5 respectively correspond to line 1 and line 2. According to line 1 in diagram (a), 0 ns–40 ns indicates the surface of the filled soil, the working range is covered with Artificial, the GPR wave reflected even signals in this scope. The plus or minus the amplitude approximately, which means that the filling layer is compacted and average. Within the scope of 40 ns–120 ns, more obvious reflectors are appeared, according to the proposed rule, the relative dielectric constant of the filled soil at that point is significantly greater than the surface, and therefore, there is higher water content

226

Z. Zhang et al.

Fig. 4. Construction platforms and on-site testing

(a)Scanning image in a succession of GPR - line 1

(b) Scanning image in a succession of GPR - line 2 Fig. 5. Scanning image in a succession of GPR

Forward Modeling and Detection of the Potential Slip

227

in the soil layer. The positive and negative wave reflections are more uniform in 120 ns– 240 ns, there is no obvious structural damage to the filling layer, and the water content is higher than the surface layer. At 240 ns–500 ns, the reflection signal is weak, and the interference signal is significantly increased. In this range, it entered into the bedrock body, and the influence of bedrock on electromagnetic wave attenuation is significant, it suggests that the water content in bedrock is high. According to line 2 in the diagram (b), 0 ns–40 ns indicates the surface of the filled soil, the working range is covered with artificial, the GPR wave reflected even signals in this scope, plus or minus the amplitude approximately, it shows that the filling layer is compacted and average. The reflectors are more obvious in the cope of 40 ns–120 ns, according to the proposed rule, the relative dielectric constant of the filled soil at that point is significantly greater than the surface relative dielectric constant, therefore, there is higher water content in the soil layer. The positive and negative wave reflections are more uniform in 120 ns–320 ns, there is no obvious structural damage to the filling layer, and the water content is higher than the surface layer. At 320 ns–500 ns, the reflection signal is weak, and the interference signal is significantly increased. In this range, it entered into the bedrock body, and the influence of bedrock on electromagnetic wave attenuation is significant, it suggests that the water content in bedrock is high.

6 Conclusions Based on the study of reflection and transmission mechanism of the GPR wave in the artificial layered soil, and the practical application of detection of the slope sliding surface, the conclusions are as follows: 1. The signal energy of the ground-penetrating radar is proportional to the square of the wave amplitude, so the energy size of the echo signal can also reflect the numerical of the wave amplitude. And the echo energy and wave amplitude are related to the relative permittivity of each layer medium. 2. With the increase of water content of fill layers, the multiple waves of the groundpenetrating radar gradually became obvious, and the position of multiple waves gradually extended to the time, the reason is that the speed of an electromagnetic wave in the media of higher relative dielectric constant increases gradually. When the filling layer has high conductivity, it has a strong attenuation effect on the electromagnetic wave, thus affecting the detection of the GPR wave on the deep filling. 3. The propagation wave amplitude of radar waves changes a little in a simple artificial sample. When radar waves reflect in the air interface, the amplitude of the wave is gradually shifted from positive to negative, this change is due to the shift of characteristics of the wave amplitude when it spreads from the artificial to the air. The artificial are kind of medium with high relative dielectric constant while the air is a lower kind of relative dielectric constant. The reflection of the water interface has the characteristics when the wave spreads from the medium of the low relative dielectric constant to the medium of high relative dielectric constant. The reflection on the metal interface is more obvious, and the reflection is almost twice as much as the reflection of the water interface.

228

Z. Zhang et al.

Acknowledgment. Thanks the financial: Provincial College Students’ innovation and Entrepreneurship Project (s201910705073). Shaanxi Key Laboratory of Geotechnical and Underground Space Engineering (JT201901), Shaanxi Province Science Foundation for Youths (Grant No. 2018JQ5203). Shaanxi province housing and construction research and development projects (2020-K41).

References 1. Hu, C.M., Mei, Y., Liu, Z.R.: Deformation mode and stability analysis of high sticking slope of collapsible losse. Chin. J. Rock Mech. Eng. 12, 2585–2592 (2012) 2. Ma, Y., Wang, J.D., Peng, S.J.: Deformation and failure mechanism of high sticking artificial slope. Chin. J. Geotech. Eng. 03, 518–528 (2016) 3. Zhu, C.H., Li, N.: Mesoscopic deformation mechanism of loses high-fill foundation based on soil electrical resistivity. Chin. J. Rock Mech. Eng. 03, 640–648 (2013) 4. Zhu, C.H., Li, N., Liu, M.Z.: Spatiotemporal laws of the post-construction settlement of the artificial-filled foundation of Lüliang airport. Chin. J. Geotech. Eng. 02, 293–301 (2013) 5. Su, L.H., Li, N., Lv, G.: Artificial soil water content fast nondestructive testing of airport surface. J. Xi’an Univ. Technol. 01, 40–44 (2015) 6. Sass, O., Krautblatter, M.: Debris flow-dominated and rockfall - dominated talus slopes: genetic models derived from GPR measurements. Geomorphology 86(1), 176–192 (2007) 7. Sass, O., Bell, R., Glade, T.: Comparison of GPR, 2D-resistivity, and traditional techniques for the subsurface exploration of the Öschingen landslide, Swabian Alb (Germany). Geomorphology 93(1), 89–103 (2008) 8. Zieli´nski, A., Mazurkiewicz, E., Łyskowski, M., et al.: Use of GPR method for investigation of the mass movements development on the basis of the landslide in Kałków. Roads and Bridges-Drogi i Mosty 15(1), 61–70 (2016) 9. Fankhauser, K., Guzman, D.R.L., Oggier, N., et al.: Seasonal response and characterization of a scree slope and active debris flow catchment using multiple geophysical techniques: the case of the meretschibach catchment Switzerland. EGU Gen. Assembly Conf. Abs. 17, 11833 (2015) 10. Chen, J.G., Xiong, Z.Q., Li, H.: Failure characteristics of the floor under pressure inclined and extra-thick coal seam in full-mechanized top coal caving faces. Chin. J. Rock Mech. Eng. S1, 3018–3023 (2016) 11. Wang, Y.Q., Cao, Z., Tan, Q.W.: GPR detection of open-pit landslide. J. China Coal Soc. 36(7), 1093–1097 (2011) 12. Li, D.X., Qi, M.S., Wang, C.L.: Georadar investigation for hidden danger in river dykes. Chin. J. Geol. Hazard Control 01, 21–25 (1996)

Application of High Density Electrical Method in Karst Area Hong Qiang Zhang(B) , Zheng Bo Cao, and Wei Li Hebei Provincial Communications Planning, Design and Research Institute Co., Ltd., Shijiazhuang 050011, China [email protected]

Abstract. When building bridges in limestone areas, karst development often affects the quality of piles and bridge safety. Karst development presents diversity, complexity and irregularity. If the development degree characteristics of karst cannot be found out, it will increase great difficulties and potential safety hazards for later construction. In the past, the development degree of karst cannot be completely found out by drilling alone. This requires a combination of geophysical exploration and drilling. Using high-density electrical method, the karst development can be determined according to the inversion abnormal low resistance area. According to the geophysical results and combined with the drilling data, the characteristics of karst development degree can be fully displayed in space, and the design and construction work can be carried out on the basis of accurate and detailed geological data, so as to save the project cost, shorten the construction period and ensure the safety of bridge construction and later operation. Keywords: High density electrical method · Bridge survey · Karst

1 Introduction With the construction of more and more bridges in mountainous areas, a single drilling method cannot meet the requirements of builders. Especially in limestone area, karst development often affects the pile quality and bridge safety. If we can’t find out the development degree of karst, it will increase great difficulties and potential safety hazards for later construction. In this paper, the bridge survey in limestone area of a project is studied. The high-density electrical method combined with drilling can well find out the degree of karst development and provide accurate geological data for designers [1–3].

2 Project Overview 2.1 Bridge Structure The upper part of a bridge is designed to adopt 16–30 m continuous T-beam, the lower part adopts column pier and column/ribbed slab abutment, and the foundation adopts pile foundation [3, 4]. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 229–233, 2023. https://doi.org/10.1007/978-981-19-3983-9_20

230

H. Q. Zhang et al.

2.2 Topography and Geomorphology The bridge site belongs to river terrace landform with slight topographic fluctuation. The surface elevation of the bridge site is 464.0−507.0 m. 2.3 Meteorological The project area has a warm temperate humid and semi humid continental monsoon climate with four distinct seasons, The general characteristics are: sudden cold and warm in spring (March−May), dry and lack of rain, mild climate and abundant rainfall in summer (June−August), sunny weather in autumn (September−November), large temperature difference, cold, dry and little snow in winter (December−February). The annual average temperature is 6.5 °C−10.3 °C. The data show that the minimum temperature in this area is minus 28.6 °C (January 15, 1981) and the maximum temperature is 36.6 °C (July 16, 1972). The precipitation in the year is unevenly distributed, mainly from June to September, accounting for about 80%−90% of the annual precipitation; the annual average rainfall is 727 mm, the maximum annual precipitation is 1122.6 mm, the maximum frozen soil depth is 1.2 m, and the annual average frost free period is about 135 days. The northwest monsoon prevails in winter and the southeast monsoon in summer. 2.4 Formation Lithology The surface layer of the bridge site area is covered with quaternary Holocene alluvial pebble layer and partially covered with quaternary upper Pleistocene proluvial and Deluvial gravel; The underlying bedrock is mainly middle Ordovician limestone and calcareous sandstone. The large chainage abutment is exposed as Yanshanian circumgyrate porphyry syenite, and the Carboniferous mudstone and structural breccia zone are partially exposed. The engineering geological conditions are complex. 2.5 Geological Structure There is a small north-east transverse concealed normal fault 500 m south of the bridge site area. The overall strike of the fault is 23−36° NNE and the dip angle is 55−75°. The hanging wall of the fault is Sinian chert dolomite and dolomitic limestone, and the footwall is Ordovician Majiagou Formation limestone. In the route area, the southwest fault trend gradually changes from NNE to NE, and is concealed under the overburden. Through this survey, the fault is not a Holocene Active fault. Affected by the fault, structural breccia is developed in the Middle Ordovician (O2 ) limestone layer in the bridge site area, structural fractures are developed along the rock contact zone, water erosion marks on the contact surface are obvious, and dissolution is generally developed. 2.6 Hydrogeological Conditions Surface water. The surface water in the bridge site is xidaogou River, a tributary of Liuhe river. It is a seasonal river. The water volume changes with rainfall and has the characteristics of sudden rise and fall. It is a typical mountain river.

Application of High Density Electrical Method in Karst Area

231

Groundwater. During the survey, the groundwater level in the bridge site area is 7.30−13.50 m, the groundwater is mainly Quaternary pore water, the atmospheric precipitation infiltration and river lateral supply are the main supply modes of groundwater, and the discharge modes are mainly groundwater runoff, river discharge and manual mining. According to the water quality analysis results of the detailed engineering geological survey report of Chengping expressway, the chemical type of shallow groundwater in this area is mainly SO4-CaMg, which is slightly corrosive to the steel bars in concrete structures and reinforced concrete structures. It is recommended to take corresponding anti-corrosion measures.

3 High Density Resistivity Method 3.1 Working Methods Duk-4 high density resistivity measurement system produced by Chongqing Geological Instrument Factory is used for the high density resistivity method instrument. The system includes host dzd-8 multi-functional DC method instrument, mis-10c cascade electrode converter and distributed large line cable. The system has the characteristics of large storage capacity, accurate and fast measurement and convenient operation. Its main functions include automatic signal acquisition and automatic storage, The system is equipped with domestic advanced high-density electrical method forward and inverse interpretation software, which greatly improves the accuracy and efficiency of geophysical interpretation. High density resistivity method is to infer the formation change by measuring the resistivity change of the measured section and analyzing the geological drilling data. Different formation lithology has certain electrical differences, which can be inferred and interpreted according to electrical differences. The measurement and observation system adopts Wenner device, and the interpretation section is inverted trapezoid. The distance between the measuring points is 5 m. The field recorded data is replayed to the computer. After inspection and calibration, it is processed by professional processing software. The processing mainly includes section

Fig. 1. Schematic diagram of high density resistivity method device.

232

H. Q. Zhang et al.

resistivity contour drawing, forward and reverse interpretation, and color section drawing of forward and reverse interpretation. According to the resistivity contour map, forward and reverse interpretation resistivity section map, geological interpretation is carried out in combination with drilling data, and the geophysical interpretation geological profile is drawn. As shown in Fig. 1. 3.2 Technical Principle High density resistivity method is to infer the formation change by measuring the resistivity change of the measured section and analyzing the geological drilling data. Different formation lithology has certain electrical differences, which can be inferred and interpreted according to electrical differences.

4 Workload Layout According to the 2016 edition of code for seismic design of buildings (gb50011–2010) and code for seismic design of Highway Engineering (jtgb02–2013), combined with the survey data, the geological structure of the project site is relatively stable and belongs to the general seismic section. The equivalent shear wave velocity of the site soil layer is 500 ≥ VSE > 250 m/s, the site soil type is medium hard soil, and the project site type is class II [4, 5].

5 Abutment Stability Evaluation The small pile abutment is flat, with pebbles on the surface and bedrock under it. Good abutment stability. The slope of abutment with large chainage is about 37°, the surface is exposed as syenite porphyry, the rock stratum is stably distributed, and no adverse geology such as landslide and collapse is found. Good abutment stability. Pile foundation is adopted in abutment design, which can effectively increase abutment stability and reduce abutment slope sliding risk. If the gravity abutment is adopted, the sliding and overturning stability shall be checked. Slope stability shall be considered in abutment subgrade excavation and bridge head cone slope design.

6 Conclusion There is a small north-east transverse concealed normal fault 500 m south of the bridge site area. The overall strike of the fault is 23−36° NNE and the dip angle is 55−75°. The hanging wall of the fault is Sinian chert dolomite and dolomitic limestone, and the footwall is Ordovician Majiagou Formation limestone. In the route area, the southwest fault trend gradually changes from NNE to NE, and is concealed under the overburden. Through this survey, the fault is not a Holocene Active fault. Affected by the fault, structural breccia is developed in the Middle Ordovician (O2 ) limestone layer in the bridge site area, structural fractures are developed along the rock contact zone, water erosion marks on the contact surface are obvious, and dissolution is generally developed.

Application of High Density Electrical Method in Karst Area

233

The results of high-density electrical method reveal that there is a closed low resistance anomaly within 37 m to the right of K7 + 285 ~ k7 + 285 and 31−54 m in depth. It is speculated that this is a karst cave development area or stratum rich in water. There is closed low resistivity anomaly within the range of 38 m to the right of K7 + 495 ~ k7 + 495 and the depth of 25−38 m. It is speculated that there is karst cave or stratum rich water. There is closed low resistivity anomaly within 25 m from the left of K7 + 555 to 30 m from the right of K7 + 555, and the depth is 25−65 m. It is speculated that this is a karst cave development area or stratum rich in water.

References 1. Zhao, G., Chen, X., Tang, J.: New progress and development trend of earth electromagnetic method in China. Adv. Geophys. (4) (2007) 2. Yin, J., Ling, W., Lin, Y.: Application of high frequency magnetotelluric in deep tunnel detection. Water Conservancy Plan. Design, (9), 12–13 (2015) 3. Liu, H., Wang, H.: Application of comprehensive exploration technology in expressway along the river. Eng. Constr. (6), 803–805 (2009) 4. Code for seismic design of buildings (gb50011–2010), Edition 2016 5. Code for seismic design of Highway Engineering (jtgb02–2013)

Green Building Construction Implementation Barriers in the Philippines a Hierarchical Model Cris Edward Monjardin(B) , Lovely Jasmin Dela Cruz, Ezekiel Esguerra, and Ondrea Nhika Dangzalan Mapua University Intramuros, Manila, Philippines [email protected], {ljdelacruz,elesguerra, oncdangzalan}@mymail.mapua.edu.ph

Abstract. The Philippines is one of the developing countries that attempt to practice green building. Although with 300 + registered and certified green building projects acknowledged by LEED, green building is less practiced in the Philippine Construction Industry due to several barriers. Limited studies were conducted that discuss green building in the Philippines, especially the barriers to its implementation. This research created hierarchical models of barriers to green building implementation in the Philippines by identifying, classifying, and ranking the existing barriers. The RRL looked into the technicalities of the green building certification, compiled the barriers identified by previous studies, and provided this research with its theoretical framework as the basis for the proposed methodology. The objective was achieved using two (2) types of survey questionnaires, answered by the two (2) chosen respondents: multi-unit residential occupants and qualified construction experts with knowledge or experience with green building. These questionnaires are put through statistical analysis which resulted in the research to identify and also rank the green building barriers under Social and Economic aspects of sustainability, significant in the Philippines. Finally, this research is geared towards assisting professionals in the construction industry, decision-makers, and policymakers in determining the best strategy for dealing with these challenges. Keywords: Green building · Building · Barriers · Hierarchical · Construction · Engineering

1 Introduction The concept of sustainable green building has been pioneered in the late 19th century and continues to be adopted throughout the 21st century by more nations. Green building’s global trend or status in each country may be distinguished by Green Building Rating Tools; also known as certification, utilized to assess and recognize buildings that meet certain green requirements or standards. From the Green Building Rating Tools, the number of green buildings which are qualified and have met such requirements and standards in each country is totaled and reflected. Globally, there is a variety of Green © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 234–242, 2023. https://doi.org/10.1007/978-981-19-3983-9_21

Green Building Construction Implementation Barriers

235

Building Rating Tools, although the major certifications such as LEED, BREEAM, Passive house, DGNB, and EDGE continue to be widely used [1]. Specifically, the LEED or Leadership in Energy and Environmental Design is the most recognized and utilized green building certification system among them. The countries: Mainland China, Canada, India, Brazil, and the Republic of Korea are the consistent countries and regions, outside the United States, ranked by LEED based on gross square meters of the green building projects [2]. In summary, the emergence of the green building movement brings benefits specifically in the three pillars of sustainability: namely, environment, economic and social. Its adaptation and implementation are increasing in different countries and are currently a primary solution for mitigating negative impacts of the building and construction industry. However, wide-scale implementation of green buildings has not yet emerged globally, especially for developing countries, because of varying hindrances in each country, such as the degree of economic growth, geographical location, per capita capital, and many more. The research delved into the identification of the barriers of green building construction implementation in the Philippines based on the perception of possible building occupants and engineers and presented the results through a hierarchical model. Green Building in the Philippines. In the Philippine setting, green building practices were pioneered and have been slowly adopted throughout the years. Specifically, an organization named Philippine Green Building Council (PHILGBC) was established in 2007 to promote the building industry’s exchange of green building expertise to ensure a long-term sustainable environment. Parallel to LEED, a rating tool: Building for Ecologically Responsive Design Excellence (BERDE), was also established by the organization alongside the Philippine government, to assess, measure, monitor, and certify the performance of local and national green building projects following environmental laws, regulations and mandatory standards [3]. However, BERDE holds very limited data on the projects they have certified, thus, it is better suggested to look up the green building projects which are LEED certified. According to LEED’s database, three hundred twenty-six (326) projects in the Philippines are registered for certification or certified. Two hundred thirty-two (232) of which are in the National Capital Region (NCR), which is the chosen scope of this research due to a large number of registered green building projects in the area. Out of the two hundred thirty-two (232) registered in LEED, only one hundred one (101) are certified. The remaining one hundred thirty (130) are not yet certified, some of which are still under construction or had been reviewed for certification, but their points did not reach the threshold to be certified. In addition, there are very few studies that investigate the causes for the slow development of green building in the Philippines. Regardless of the notable benefits for the environment and inhabitants, some aspects of green building construction make it less prevalent as the main method which hinders its steady progress. These so-called aspects may be referred to as barriers; serve as the hindrances preventing the implementation and acknowledgement of its potential as a better alternative to the conventional method of construction. Green Building Barriers. The review of related literature revealed the green building barriers that exist in the international setting. They are summarized and categorized in Table 1.

236

C. E. Monjardin et al. Table 1. Green building barriers and codes

Code

Barriers

Social: awareness and education AE01

inadequate knowledge and information about green building concept [4]

AE02

not enough research and case studies [4]

AE03

Lack of building management role in supporting green building movement, [4]

AE04

Lack of public awareness [4–6]

AE05

Lack of database and information in GB [5, 6]

Social: people and culture PC01

Burdensome implementation [4]

PC02

lack of supportive atmospheres [4]

PC03

Resistance to change [4]

PC04

Negligence [4]

PC05

Inconsistencies between formal regulations (e.g., existing federal, state and local legislation) and LEED [7, 8]

PC06

Lack of government support [6, 9]

PC07

Lack of interest from clients and market demand [5, 6]

PC08

Lack of government incentives [5, 6]

PC09

Lack of green building codes and regulation [5, 6]

PC10

Lack of GB promotion by government [5]

PC11

High degree of distrust of green building technology [5]

Economic: financial F01

Higher costs for green building options than conventional ones. [4]

F02

Risk of investment. [4, 6]

F03

high cost of certification [4, 7–9]

F04

Deficient financial support from the government and banks [4]

F05

Scarcity of insurance solutions [7, 8]

F06

Rental or resale value loss due to delay related to green construction procedures and conditions [7, 8]

F07

Failure to use of financial incentives (tax/loan discounts, low financing rates) because of delays or lower certification levels than expected [7, 8]

F08

High initial/investment cost [6, 9]

F09

Long payback period [5, 9]

F10

Higher cost of GBT products [5]

F11

Lack of financing schemes [5, 6] (continued)

Green Building Construction Implementation Barriers

237

Table 1. (continued) Code

Barriers

F12

High market prices and rental charges of green buildings resulting from green building technologies application [5, 6]

Economic: management M01

Absence of further evaluation, [4]

M02

Insufficient supervision from responsible parties, [4, 9]

M03

Inadequate definition of project parties’ contractual roles and responsibilities, [7, 8]

M04

Lack of importance attached to GBs by senior management, [5]

M05

Lack of strategy to promote GB, [6]

M06

Lack of design and construction team, [6]

M07

Risks and uncertainties involved in adopting new technologies, [5]

M08

Limited experience with the use of nontraditional procurement methods, [5]

MO9

Resistance to change from the use of traditional construction methods, [5]

M10

Conflict of interests among various stakeholders in adopting green building technologies, [5]

M11

Adoption of green building technologies is time consuming and causes project delays, [5]

Economic: technical T01

Lack of green construction experience and qualification, [5–8]

T02

Contractors and subcontractors agreeing to standards that are not within their expertise and competence, [7, 8]

T03

Lack of expertise in new products/technologies, [5–9]

T04

Complex and rigid requirements involved in adopting green building technologies, [5]

T05

Lack of local instituteUnavailability of local suppliers of GBTs and facilities for research and development (R&D) of GB, [5, 9]

Economic: resource R01

Doubts about long-term viability and performance of new and untested products, materials and technologies, [7, 8]

R02

Faulty performance of HVAC/electrical/plumbing systems and alternative water systems/alternative power generating equipment, [7, 8]

R03

Failure to receive materials/products in a timely fashion causing delays, [7, 8]

R04

Lack/unavailability of local suppliers of GBT products, [4, 5, 9]

R05

Lack of green technologies, [6, 9]

238

C. E. Monjardin et al.

2 Methodology The methodology of this study is a quantitative research design based on the combined methodologies of three related studies by Wimala et al. [4], Chan et al. [5], and Dehdasht [10]. In Phase 1, the researchers collected information regarding green building and its rating system to have an understanding of the benefits green building construction offers to the society, environment, and the builder’s reputation. The information gathered is a huge component of the scope of the introduction of this paper and are greatly sourced from the websites of green building councils including WGBC, PHILGBC, BERDE, and LEED. Moving on to Phase 2, the researchers then identified the barriers of green building implementation existing in other countries such as Ghana, Malaysia, and Indonesia. The barriers identified are listed and categorized in Table 1 alongside their respective codes, which were used throughout the proceeding parts of this paper. In Phase 3, the preparation for the data gathering started in order to contextualize the found barriers in the Philippine setting. The barriers were summarized for the construction of the questionnaires. An initial questionnaire was made for the construction experts which contains 4-point Likert Scales to let them rate how strongly they agree or disagree if such a barrier exists or occurs in the country (4 = Strongly Agree, 3 = Agree, 2 = Disagree, 1 = Strongly Disagree). In Phase 4, the pilot testing of the questionnaire was conducted through Google Forms to 30 construction experts, which are engineers, architects, project managers, consultants, quantity surveyors and other professionals in the construction industry. The results of the pilot testing underwent a statistical analysis: Cronbach’s alpha, which will determine the reliability of the questionnaire. Once the questionnaire was considered reliable, the researchers proceeded to Phase 5 of this study after determining the sample size for both types of respondents. In Phase 5, the actual data gathering commenced with the qualified construction experts as respondents. The snowball sampling method was utilized via social media for the 50 participants. After which, the data gathered underwent the ANOVA Test to determine whether the categories have significant differences among the categories. The normalized values of the data gathered were also identified to determine which barriers are significant in the country and were included in the final questionnaire for the multiunit residential occupants. In Phase 6, the questionnaire was then distributed in social media to reach 400 multi-unit residential occupants in NCR who are familiar about green building.

3 Results and Discussion Pilot Testing. The data from the pilot testing, which underwent Cronbach’s alpha, is equal to 0.869851. It obtained a value greater than 0.6 which interprets that the survey questions are internally reliable (or consistent). The last column is the value of Cronbach’s alpha IF we delete that question. For each question, the Cronbach’s alpha is still greater than 0.6 thus, deleting any question to make the survey questionnaire reliable is not needed.

Green Building Construction Implementation Barriers Table 2. P-value of the categories according to qualified experts Category

P-value

Awareness and education

0.716485

People and culture

239

Table 3. P-value of the categories according to multi-unit residential building occupants Category

P-value

0

Awareness and education

0

Financial

0.000282

People and culture

0

Management

0.000007

Financial

0

Technical

0.426928

Management

0

Resource

0.070613

Technical

0.013755

Resource

0.000104

ANOVA Test. The ANOVA Test was done to determine the P-Values of the categories which dictates whether the barriers within that category have significant differences or none. An obtained P-value less than 0.05 signifies the difference is significant (Tables 2 and 3). Normalized Values. The critical barriers to the adoption of green building implementation in the Philippines are defined as those with normalized values not less than 0.50. After normalizing the values, the results are as follows. Table 4. Significant barriers of each categories according to qualified experts Significant barriers

AE

PC

F

M

T

R

AE03

PC05

F01

M01

T01

R01

AE05

PC06

F03

M02

T05

R04

PC08

F04

M03

PC10

F05

M04

F08

M05

F10

M07

F11

M08

F12

M09 M10

Table 4 shows what barriers are retained from the statistically analyzed data accumulated from the answers of the qualified experts in the Philippines. Table 5 shows what barriers are retained from the statistically analyzed data accumulated from the answers of the Multi-unit Residential Building Occupants in the Philippines.

240

C. E. Monjardin et al.

Table 5. Significant barriers of each categories according to multi-unit residential building occupants Significant barriers

AE

PC

F

M

T

R

AE04

PC02

F04

M02

T05

R04

PC03

F08

M03

PC04

F11

M04

PC05

M05

PC06

M08

PC08

M09

PC010

M10

R05

4 Conclusion The research concluded that even with the positive route of green buildings, barriers are still present to its implementation which are summarized and categorized in Table 1. Contextualizing the barriers and after implementing the ANOVA test, normalization, and Tukey HSD resulted in two (2) hierarchical models. The first hierarchical model shows the barriers according to construction experts. All of the barriers in the model are statistically identified as significant in the country. First, under the Social aspect; the barriers under the Awareness and Education category have no statistically significant difference, however, the ranking is as follows; (1) AE03, and (2) AE05. Under the People and Culture category, the barriers have a significant difference between them, and the ranking is as follows; (1) PC10, (2) PC08, (3) PC06, and (4) PC05. Next, under the Economic aspect; the barriers under the Financial category have a statistically significant difference, and the ranking is as follows; (1) F03, (2) F01 and F10, (3) F04 and F08, (4) F12, (5) F05, and (6) F11. Under the Management category, the barriers have a significant difference between them, and the ranking is as follows; (1) M05, (2) M09, (3) M08, (4) M04, (5) M10, (6) M01, (7) M07, (8) M02, and (9) M03. Under the Technical category, the barriers have no statistically significant difference, however, the ranking is as follows; (1) T05, and (2) T01. Lastly, under the Resource category, the barriers have no statistically significant difference, however, the ranking is as follows; (1) R04, and (2) R01. The second hierarchical model shows the barriers according to multi-unit residential occupants. Similar to Fig. 1, the barriers in the model are statistically identified as significant in the country. First, under the Social aspect; the barriers under the Awareness and Education category only one is significant, which is (1) AE04. Under the People and Culture category, the barriers have a significant difference between them, and the ranking is as follows; (1) PC10, (2) PC06, (3) PC08, (4) PC04, (5) PC03, (6) PC05 and (7) PC02. Next, under the Economic aspect; the barriers under the Financial category have a statistical significant difference, and the ranking is as follows; (1) F11, (2) F04, and (3) F08. Under the Management category, the barriers have a significant difference between them, and the ranking is follows; (1) M05, (2) M04, (3) M03, (4) M10, (5) M09, (6)

Green Building Construction Implementation Barriers

241

M08, and (7) M02. Under the Technical category, there is only one significant barrier, which is (1) T05. Lastly, under the Resource category, the barriers have a significant difference, and their ranking is as follows; (1) R04, and (2) R05. In conclusion, there are barriers to green building construction implementation in the Philippines according to the perspectives of qualified construction experts and multiunit residential occupants in NCR. The hierarchical models of the barriers provided an organized presentation, allowing future readers to easily identify according to the aspects of sustainability (social and economic), categories, and what is more significant among all barriers.

Fig. 1. Hierarchical model of green building construction implementation barrier in the Philippines according to construction experts

References 1. United Nations Environment Programme. Global Status Report for Buildings and Construction: Towards a Zero-emission, Efficient and Resilient Buildings and Construction Sector. Nairobi (2020) 2. Stanley, S. U.S. green building council announces Top 10 countries and regions for LEED green building | U.S. green building council. Usgbc.org. https://www.usgbc.org/articles/ us-green-building-council-announces-top-10-countries-and-regions-leed-green-building. Accessed 04 Nov 2021 3. Culiao, R., Tae, S., Kim, R.: A review of the Philippine green building rating system, BERDE in comparison with G-SEED and LEED. Sbt-Durabi.org, 87–94 (2018) 4. Wimala, M., Akmalah, E., Sururi, M.R.: Breaking through the barriers to green building movement in Indonesia: Insights from building occupants. Energy Procedia 100, 469–474 (2016) 5. Chan, A.P., Darko, A., Olanipekun, A.O., Ameyaw, E.E.: Critical barriers to green building technologies adoption in developing countries: the case of Ghana. J. Clean. Prod. 172, 1067– 1079 (2018)

242

C. E. Monjardin et al.

6. Samari, M., Ghodrati, N., Esmaeilifar, R., Olfat, P., Shafiei, M.W.M.: The investigation of the barriers in developing green building in Malaysia. Mod. Appl. Sci. 7(2) (2013) 7. Gurgun, A.P., Arditi, D., Vilar, P.C.: Impacts of construction risks on costs in Leed-certified projects. J. Green Build. 11(4), 163–181 (2016) 8. Chiveralls, K., Zillante, G., Zuo, J., Wilson, L., Pullen, S.: Constructing corporate social responsibility: encouraging CSR through legislation and regulation. Paper presented at the COBRA 2011 - Proceedings of RICS Construction and Property Conference, pp. 1272–1284 (2011) 9. Shen, W., Tang, W., Siripanan, A., Lei, Z., Duffield, C., Hui, F.: Understanding the green technical capabilities and barriers to green buildings in developing countries: a case study of Thailand. Sustainability 10(10), 3585 (2018) 10. Dehdasht, G.: Barriers of sustainable lean construction implementation: hierarchical model (2018). https://doi.org/10.1007/978-3-030-02101-6_4

Urban Planning and Environmental Assessment

Wind-Induced Dynamic Response Analysis of the Frame Lightning Rod Ran Liu1(B) , Li Song1,2 , Chenxing Cui1 , and Jinliang Liu1 1 School of Civil Engineering, Central South University, Changsha 410075, China

[email protected] 2 National Engineering Laboratory for High Speed Railway Construction, Changsha 410075,

China

Abstract. This paper simulated and analyzed the wind-induced dynamic response of a 500 kV substation’s typical frame lightning rod structure using the fluid-solid coupling numerical method. The analysis shows that the high-frequency vibration response of the upper members is significant, and it is greatly affected by multiorder vibration mode, the crosswind vortex-induced resonance is prone, which is unfavorable to the whole stress of the structure. The upper two sections of the lightning rod were optimized by applying spiral side plates and carried out the whole analysis. It is found that the maximum stress of the frame lightning rod occurred on the beam of the intersecting joint within the basic wind speed. With the increase of wind speed, the maximum stress position moves upward along the lightning rod and occurs at the bottom of the lightning rod sections. The structural optimization measures can effectively suppress the vortex-induced resonance of the lightning rod. The vortex-induced amplitude of the structure in the crosswind can be significantly reduced in the basic wind speed. So their joint action should be considered in the structural design and stress analysis. Keywords: Frame lighting rod · Scale model design · Structural optimization design · Wind-induced vibration · Fluid-solid coupling

1 Introduction The lightning rods are mostly installed on the frame to form a frame lightning rod [1]. The frame lightning rod structure is slender and long, determining that the wind load is its control load. With the improvement of social demand and technological progress, the lightning rod structure is gradually developing in a higher and lighter direction, which further enhances its sensitivity to wind load [2]. The frame lightning rod structure will sustain vibration under wind load and experience a large number of stress cycles during service, and its fatigue damage will be sustained accumulate and structural failure may occur [3]. In China, the collapse accidents of the frame lightning rod are not uncommon. As early as 1991, a 220 kV substation in Yuanjiang had a lightning rod breaking and collapse accident. Then, there were five collapse accidents of the 750 kV substation

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 245–257, 2023. https://doi.org/10.1007/978-981-19-3983-9_22

246

R. Liu et al.

frame lightning rod from 2008 to 2015 [4–6]. The formal operation time of these broken substation frame lightning rods did not reach the structure’s design life and caused significant economic losses and social impacts. A 220 kV frame lightning rod of a 500 kV substation broke and collapsed in winter under a strong wind climate. The fracture position of the lightning rod occurred at the intersecting node between the bottom of the lightning rod and the beam [7], the fracture mode is torn from the beam’s cross-section. Therefore, it is essential to analyze the fracture cause, structural optimization design, and the wind-vibration response of the substation frame with lightning rods. After the accident, the department concerned conducted experimental research on the material chemical composition and mechanical properties of the broken part of the lightning rod, and the results met the requirements [6]. Chen Yiwen et al. [7] conducted modeling and simulation analysis on the structure and summarized the main reasons for the bottom fracture of the lightning rod: the code ignored the vortex-induced vibration of the frame lightning rod under wind load and its fatigue effect. The stress concentration of the steel tubular intersecting joint is prominent. The intersecting joint at the bottom of the lightning rod under the alternating wind load has an excellent fatigue impact, which causes the crack to expand continuously and leads to the lightning rod fracture accident. The plasticity and toughness of steel in the heat-affected zone decrease with repeated stress, and the fracture form changes from plastic fracture to brittle fracture. The research about the fracture causes of the frame lightning rod showed that wind load significantly impacts the fatigue of the lightning rod intersecting joint. Still, there are few studies on the quantitative analysis of its impact degree. In this paper, the fluidsolid coupling calculation method is used to simulate and analyze the wind-induced dynamic characteristics of the lightning rod and optimize the lightning rod structure.

2 Scale Model Design of the Frame Lightning Rod 2.1 The Whole Frame Lightning Rod Structure The research object of this paper is a 220 kV frame lightning rod of a 500 kV substation. The whole structure diagram is shown in Fig. 1. The material parameters are shown in Table 1. Studies had found that [7] there was apparent stress concentration at the intersecting joint of the middle-span (A-span) frame lightning rod under wind load. When the A-span lightning rod is in the whole structure, the variation law of stress and displacement at the connection points of each lightning rod section under different wind load conditions are the same as that in the single lightning rod so that the A-span lightning rod can be analyzed separately. The A-span single frame lightning rod is scaled according to the ratio of 1:10 to reduce the amount of calculation, and similar parameters are shown in Table 2.

Wind-Induced Dynamic Response Analysis of the Frame Lightning Rod

247

Fig. 1. The schematic diagram of the whole frame lightning rod.

Table 1. Material parameters [7] Steel model

Elastic modulus

Poisson’s ratio

Design strength

Yield strength

Mass density

Q235B

2.06 × 105

0.3

215 MPa

235 MPa

7850 kg/m3

Q345B

2.06 × 105

0.3

310 MPa

345 MPa

7850 kg/m3

Table 2. Similar parameters for the aeroelastic model design Similar parameters Similarity ratio

λL

λM

λU

λEI

λf

n

n3

n0.5

n5

n−0.5

1:10

1:103

1: 3.16

1:105

3.16:1

2.2 Aeroelastic Model Design Due to the different bending stiffness of the bottom support of the A-span frame structure in each direction, the aeroelastic model adopts a rectangular steel tube structure. The calculation principle of the steel tube size is to keep the longitudinal and transverse flexibility ratio of the aeroelastic model consistent with the original structure, and the modal frequency meets the similarity ratio. The size is 20 mm × 40 mm × 1 mm, as shown in Fig. 2. There are five sections of the A-span lightning rod, the sizes from top to bottom are 48 × 3.5 × 1500, 89 × 5 × 2500, 140 × 4 × 2500, (325/220) × 6 × 4500, 325 × 10 × 4000, the diameter and length of the aeroelastic model circular tube can be calculated by the size similarity ratio λL , the thickness can be calculated according to the bending stiffness ratio λEI , the sizes of the aeroelastic model from top to bottom are 4.8 × 0.0858 × 150, 8.9 × 0.1284 × 250, 14 × 0.111 × 250, (32.5/22) × 0.1704 × 450, 32.5 × 0.2757 × 400. Because the thickness of each section of the

248

R. Liu et al.

lightning rod is too small, it is not easy to converge in numerical calculation. So the thickness of each section is increased by three times. The material’s elastic modulus is reduced to 1/3 of the original to meet the bending stiffness scale ratio. After the size of the lightning rod aeroelastic model is determined, the additional counterweight required for each section is calculated according to the mass scale ratio, and the material density is adjusted to meet the mass scale ratio requirement. The size of the cross beam and lightning rod section of the prototype intersecting joint at the bottom of the lightning rod are 480 mm × 6 mm and 325 mm × 8 mm. To ensure the aeroelastic model’s intersecting joint and the prototype intersecting joint meet the scale ratio, the size of the cross beam is 48 mm × 0.207 mm.

Fig. 2. The schematic diagram of the S-axis and W-axis direction of the aeroelastic model.

3 Fluid-Solid Coupling Analysis of the Aeroelastic Model The finite element analysis process of fluid-solid coupling of the frame lightning rod structure is shown in Fig. 3. The solution steps are as follows: (1) The finite element solid model and fluid domain model of the frame lightning rod are established by rhinoceros. Input the material parameters of each section of the frame lightning rod in the Engineering data module of the workbench. (2) Import the solid structure model into the Transient Structural module in the workbench, then define the material parameters and steel pipe thickness of each section of the lightning rod, set the structural boundary conditions (fixed bottom surface of the support) and the solid coupling surface, and finally divide the structure mesh (see Fig. 4). (3) Import the fluid domain model into the HyperMesh software to divide the fluid domain 3D mesh, then import the mesh model into the fluid module in the workbench, and set the boundary conditions and coupling surface of the fluid domain (see Fig. 5). (4) Set the fluid-solid coupling solution parameters through the SYTEM COUPLING solver, establish the fluid-solid coupling interface for data transmission of the fluid domain and solid structure, and use the spring smoothing method to update the data. In this paper, the weak coupling method is used to solve the fluid-solid coupling problem of the substation frame lightning rod, so the solution order is to calculate the fluid domain first and then the solid structure.

Wind-Induced Dynamic Response Analysis of the Frame Lightning Rod

249

Start

Establish the finite element model of the frame lightning rod (rhino)

Establish fluid domain model and 3D mesh generation (HyperMesh)

Import into the Transient Structural module in Workbench

Import the mesh model into the Fluent module in Workbench

Input the material parameters of each section of the frame lightning rod Engineering data

Set the fluid-solid coupling surfaces of solid structure and fluid domain respectively

Define the boundary conditions of the fluid domain (see Fig. 3)

The data transmission of fluid-solid coupling interface is carried out through SYTEM COUPLING platform, and the data is updated by spring smoothing method In the weak coupling method, calculate the fluid domain first, and then calculate the solid structure End

Fig. 3. The analysis process of the frame lightning rod structure

3.1 Finite Element Model of the Frame Lightning Rod and Fluid Domain In this paper, the finite element model of the frame lightning rod is established using the shell element, and the fluid domain model is established in the HyperMesh, as shown in Fig. 4–5. There are five boundary conditions in the fluid domain, including the inlet and outlet boundaries, the upper and lower boundaries, and the outer surface boundary of the lightning rod boundary. The fluid domain is defined based on the fifth section size of the lightning rod, the distances from its centroid to the inlet boundary and the upper boundary and lower boundary are the same, which are set to 10D, the distance from its centroid to the outlet boundary is set to 25D, and the height of the fluid domain is set to 2 m. The variation law of wind speed and turbulence characteristics (including turbulent kinetic energy and turbulent dissipation rate) at the inlet boundary with the height above the ground is programmed and input through UDF [8].

250

R. Liu et al.

Fig. 4. The aeroelastic model of the frame lightning rod.

(a) Two-dimensional fluid domain

(b) Fluid domain model

Fig. 5. Fluid domain analysis model.

The working environment of the frame lightning rod: basic wind pressure is ω0 = 0.35 kN/m2 , average altitude is z = 136.4 m, basic wind speed calculation formula:   2w0 2 × 0.35 v0 = = = 23.83 m/s (1) ρ 0.00125e−0.0001×136.4 where ρ = 0.00125e−0.0001z . The exponential law formula proposed by A.G. Davenport [9] for the average wind profile is widely used in civil engineering design. According to the geomorphic features of the substation, it belongs to the class B geomorphic area in China’s “Building Structure Load Code” [10]. The parameter value can refer to China’s load code [11]. The formula is as follows: U (z) = 23.83(z/10)0.15 , z < 350 m

(2)

3.2 Model Validation The fluid-solid coupling calculation of the aeroelastic model is carried out. The position of the measurement point at the bottom of each section of the lightning rod is consistent with the test. The first two sections are 20 mm above the bottom joint, and the third section is 25 mm above the bottom joint. The last two sections are 40 mm above the bottom joint. When the wind speed is 5 m/s, and the wind angle is 0° (perpendicular to the S axis),

Wind-Induced Dynamic Response Analysis of the Frame Lightning Rod

251

comparing the simulation calculation results with the test results, as shown in Table 3. It is found that the errors of W4 and W5 measuring points are 23.439% and 20.844%, and the other measuring points are less than 20%. Because of the test equipment errors and differences in the specifications of aeroelastic models, it is difficult to completely reproduce the test results in the numerical simulation, so the error is acceptable. Table 3. Mean square value of measuring point strain under 0° wind angle and 5 m/s wind speed (με). Test value [12]

Analog value

Error (%)

Test value [12]

Analog value

Error (%)

S1

3.394

2.993

19.401

S2

3.133

2.857

14.81

W1

6.141

5.752

6.763

W2

8.445

7.291

15.828

S3

3.232

3.613

S4

5.602

4.703

2.278

W3

10.15

9.68

4.855

15.67

W4

10.717

8.682

23.439

S5

11.321

9.274

17.525

W5

8.273

6.846

20.844

4 Wind-Induced Dynamic Response Analysis of the Frame Lightning Rod 4.1 Dynamic Response Analysis of the Aeroelastic Model At the local height of 10 m, the basic wind speed with a recurrence period of 50 years is 23.83 m/s, which corresponds to the wind speed of the aeroelastic model is 7.5 m/s. The strain time history of each measuring point of the prototype frame lightning rod under the basic wind speed is calculated using the fluid-solid coupling simulation method. And the strain power spectrum can be obtained by Fourier transform, as shown in Fig. 6. It can be seen that the vibration form of the first two sections of the lightning rod is the most complicated, and their frequency response includes the first four frequencies. The third section consists of the first, second, and fourth frequency components, and the last two sections are mainly first-order frequencies. Therefore, the high-frequency response of the lightning rod structure is primarily reflected in the upper members of the structure, and their vibration is jointly controlled by multi-order vibration modes, which is unfavorable to the structure.

R. Liu et al.

40

30

32

25

30

20

28

15 10

50

26 24

10

15

0

20

5

10

Time (s) 23.83 m/s, W-1 (0°)

1.0

X:61.04 Y:1.154

Strain power spectral density

Strain power spectral density

1.2 1.0 0.8 0.6 0.4

X:7.56 Y:0.20

0.2 0.0

X:17.94 Y:0.08 0

20

X:39.06 Y:0.19 40

Frequency (Hz)

15

25

20

Time (s)

60

80

23.83 m/s, W-2 (0°)

X:7.56 Y:0.95

0.8

0.6

0.4 X:17.94 Y:0.22

0.2

0.0

0

20

X:39.06 Y:0.11 40

Frequency (Hz)

5

10

Time (s)

15

1.5

1.0

0.5

0.0

80

20

23.83 m/s, W-3 (0°)

X:61.04 Y:1.154

X:61.04 Y:1.154

X:61.16 Y:0.05 60

0

2.0

Strain power spectral density

5

40

30

18 0

45

35

20

0

23.83m/s,W-3 (0°)

55

22

5

-5

60

23.83m/s,W-2 (0°)

34

Strain (με)

Strain(με)

36

23.83m/s,W-1 (0°)

35

Strain (με)

252

0

20

X:61.04 Y:1.154 40

60

80

Frequency (Hz)

Fig. 6. The fixed-point strain time history and the strain power spectrum of the lightning rod section before optimization (v = 23.83 m/s).

4.2 The Frame Lightning Rod Aeroelastic Model Optimization Design Through the wind vibration response analysis of the prototype frame lightning rod aeroelastic model, the vibration forms of the first and second lightning rod sections are more complicated, and vortex-excited resonance in the crosswind is prone to occur, which is unfavorable to the structure response. Therefore, the structural optimization measure adopted in this paper is to attach spiral side plates on the surface of the lightning rod to suppress the occurrence of vortex-excited resonance. The factors that affect the suppression efficiency of the spiral side plate include the shape and the coverage rate of the side plate [13]. In this paper, the size of the spiral side plate is selected: the number of screw heads is 3, the screw pitch is 5D, the screw height is 0.25D, and the coverage rate is 50% (D is the outer diameter of the lightning rod tube). The optimized frame lightning rod finite element model is shown in Fig. 7.

Fig. 7. Finite element scale model of the optimized frame lightning rod.

Wind-Induced Dynamic Response Analysis of the Frame Lightning Rod

253

4.3 The Whole Analysis of the Optimized Frame Lightning Rod Aeroelastic Model Figure 8 shows the stress cloud picture of the aeroelastic model of the frame lightning rod before and after optimization under the wind speed of 5 m/s at the wind angle of 0°. It can be seen from the figure that the maximum stress of the aeroelastic model of the frame lightning rod before and after optimization occurs at the bottom of the second section of the lightning rod and on the beam respectively, the maximum stress is 94.547 Mpa and 28.043 Mpa. According to the similarity ratio between the simulated wind speed, and the actual wind speed and Eq. (3), the maximum stress of the frame lightning rod before and after optimization under 15.812 m/s actual wind speed can be calculated to be 283.641 Mpa and 84.129 Mpa, which does not exceed the yield strength of the material. The stress conversion relationship between the aeroelastic model and prototype structure: σp =

Ep εp Ep σm = σm = 3σm Em εm Em

(3)

σp is the actual stress of the middle-span lightning rod in the whole prototype structure, and σm is the stress of the aeroelastic model.

(a) The aeroelastic model before optimization (b) The optimized aeroelastic model Fig. 8. The whole stress cloud diagram of the aeroelastic model at the 20 s under 5 m/s wind speed.

Analyze the influence of wind speed on the overall stress of the optimized frame lightning rod, the maximum stress, and its location under 3.162 m/s−31.623 m/s wind speeds and 0° wind angle are listed in table 4. Through comparison, it is found that the maximum structural stress increases with the increase of wind speed. When the wind speed is 31.623 m/s, the maximum stress value is 550.860 Mpa, which has exceeded the yield strength of the material. Within the basic wind speed, the maximum stress occurs on the beam. With the increase of wind speed, the location of the maximum stress moves upward along the frame lightning rod. For each lightning rod section, the most unfavorable position of stress occurs at the bottom of the rod. It shows that within the basic wind speed, the continuous cyclic vibration of the structure has the most significant

254

R. Liu et al.

influence on the cumulative fatigue damage of the intersecting joint. The upper sections of the lightning rod are more sensitive to the increase in wind speed. Table 4. Maximum structural stress under each wind speed and 0° wind angle. Wind speed (m/s)

Maximum Mises stress (Mpa)

Maximum stress position

3.162

3.810

Beam

6.325

20.807

Beam

9.487

40.857

Beam

12.649

44.211

Beam

15.812

84.129

Beam

18.974

98.010

Beam

22.136

132.162

Beam

25.298

226.377

The bottom of the third section of the lightning rod

28.461

227.688

The bottom of the third section of the lightning rod

31.623

550.860

The bottom of the first section of the lightning rod

4.4 Analysis of the Frame Lightning Rod Aeroelastic Model Joints Since each lightning rod section’s most unfavorable stress position occurs at the bottom, the following mainly carried out joints analysis for the bottom of each lightning rod section. In this paper, the measuring points of the first four sections of the lightning rod aeroelastic model are arranged at 20 mm above the bottom of the lightning rod section, the measurement point of the fifth section of the lightning rod aeroelastic model is set at 40 mm above the centroid of the beam. At the height of the measuring point of each lightning rod section, two measuring points are set along the strong axis (S axis) and the weak axis (W axis). Figure 9 shows the comparison results of fixed-point strain time history and the strain power spectrum of each lightning rod section before and after optimization under 0° wind angle and 23.83 m/s basic wind speed. It is found that under the basic wind speed, the spiral side plate can significantly reduce the vibration amplitude of the fixedpoint strain of the lightning rod. The fixed-point strain response of the second section is increased, and the other lightning rod sections are slightly reduced, indicating that the spiral side plate increases the downwind resistance of the second section of the lightning rod. Each lightning rod section is mainly dominated by the first-order frequency, it can be found that the spiral side plates reduce the frequency components in the frequency response of the upper sections of the frame lightning rod, which can effectively avoid high-frequency response.

Wind-Induced Dynamic Response Analysis of the Frame Lightning Rod 50 0°,W1 (original structure) 0°,W1 (optimized structure)

45 40

0°,W3 (original structure) 0°,W3 (optimized structure)

60

50

30

50

20 15 10 5

Strain (με )

25

40

Strain (με )

Strain (με)

0°,W2 (original structure) 0°,W2 (optimized structure)

60

35

255

30

40

0

20

-5

30

-10 0

5

10

15

10

20

Time (s)

10

0.8 X:12.70 Y:2.24

Strain power spectral density

Strain power spectral density

2.0

5

15

0

20

1.5

1.0

0.5

X:12.55 Y:0.75

0.7 0.6 0.5 0.4 0.3 0.2

0.0 0

5

10

10

15

20

25

15

20

Time (s) 0.40

23.83 m/s, W-2 (0°)

0.1 0.0

5

Time (s)

2.5 23.83 m/s, W-1 (0°)

0

Strain power spectral density

-15

23.83 m/s, W-3 (0°)

X:25.10 Y:0.38

0.35 0.30 0.25 0.20 0.15 0.10 0.05

0

5

10

15

20

25

Frequency (Hz)

Frequency (Hz)

0.00

0

10

20

30

Frequency (Hz)

Fig. 9. The fixed-point strain time history and the strain power spectrum of the lightning rod section before and after optimization (v = 23.83 m/s)

Figure 10 shows the comparison results of each fixed-point strain response of crosswind at 0° wind angle before and after structural optimization. Before the structural optimization, the second and fourth sections of the lightning rod structure have vortexexcited resonance at about 21 m/s wind speed. The third and fifth sections of the lightning rod structure have vortex-excited resonance at about 24 m/s wind speed. And the first section of the lightning rod has obvious vortex-excited resonance around the two wind speeds. There is no obvious vortex-excited resonance in each lightning rod section of the optimized structure, indicating that the structural optimization measure effectively suppressed the vortex-excited resonance of the structure.

0°,S1 0°,S2 0°,S3 0°,S4 0°,S5

100 80

Maximum absolute value of strain(με)

Maximum absolute value of strain(με)

60

120

60 40 20 0

0

5

10

15

20

Wind speed(m/s)

25

30

0°,S1 0°,S2 0°,S3 0°,S4 0°,S5

40

20

0

0

5

10

15

20

25

30

35

Wind speed(m/s)

(a) The aeroelastic model before optimization (b) The optimized aeroelastic model Fig. 10. The fixed-point crosswind strain responses of the aeroelastic model at 0° wind angle

256

R. Liu et al.

5 Conclusion In this paper, taking a 500 kV substation 220 kV frame lightning rod intersecting joint fracture accident as the engineering background, the optimization design is carried out of the frame lightning rod and analyzes its wind-induced vibration response based on the fluid-structure coupling numerical method. The main conclusions are as follows: (1) The whole analysis of the optimized aeroelastic model found that the greater the wind speed, the greater the structure’s stress. Within the basic wind speed, the maximum stress of the frame lightning rod occurs on the beam of the intersecting joint, indicating that the continuous cyclic vibration of the structure has the most significant impact on the cumulative fatigue damage of the intersecting joint. With the increase of wind speed, the maximum stress position moves upward along the lightning rod. When the wind speed reaches 28.461m/s, the maximum stress of the first section of the lightning rod at a 0° wind angle exceeds the yield strength of the material, and the stress of the structure is unsafe. (2) The vortex-excited resonance of the structure can be effectively suppressed by adding spiral side plates to the upper sections of the lightning rod. It can significantly reduce the vortex-excited amplitude of the structure of crosswind within the basic wind speed. Still, it will hurt harm the crosswind vibration of the structure under higher wind speeds.

References 1. Sun, Z., Yu, C.: On installing the overhead lightning rod on the top of the terminal tower of 35 kV power transmission lines toward transformer substations. J. Hefei Univ. Technol. 25(1), 127–130 (2002) 2. Wang, S.: Studies of wind-induced vibration and wind-induced fatigue on high-rise structures. Zhejiang University (2005) 3. Repetto, M.P., Solari, G.: Wind-induced fatigue collapse of real slender structures. Eng. Struct. 32(12), 3888–3898 (2010) 4. Huang, B., Zou, B.: Analysis of lightning rod falling accident in 220 kV Yuanjiang substation. East Chin. Electric Power 30(8), 127–130 (1992) 5. Xu, X., Wu, G., Li, Y.: Cause analysis and Countermeasures of Substation lightning rod fracture. Ningxia Electric Power, (c00), 94–99 (2010) 6. Sun, T., Li, Y., Wen, D.: The effect on karman vortex street to the mechanical strength of the lightning rod of steel tube. Insulators Surge Arresters, 01(003), 142–146 (2017) 7. Chen, Y.: Wind-induced dynamic response analysis of substation framework with lightning rod. Zhengzhou University (2017) 8. Sun, X.: Simulation of data exchange of multi-physics coupling surfaces. Comput. Simul. 32(1), 23–28 (2015) 9. Zhang, H.: Non-oscillating and non-free-parameter dissipative finite difference scheme. Acta Aerodynamica Sin. 6(2), 143–165 (1988) 10. Ministry of Construction of the People’s Republic of China.: GB50009–2012. Building Structure Load Code. China Construction Industry Press, Beijing (2012) 11. Zdravkovich, M.M.: Review and classification of various aerodynamic and hydrodynamic means for suppressing vortex shedding. J. Wind Eng. Ind. Aerodyn. 7(2), 145–189 (1981)

Wind-Induced Dynamic Response Analysis of the Frame Lightning Rod

257

12. Yi, Y.: Bearing Capacity Test Study and Fatigue Life Analysis of Planar Joints of Lightning Rod in Substation. Zhengzhou University (2018) 13. Zheng, T.: Study on the mechanism of spiral side plate for suppressing vortex-Induced vibration of ocean riser. Jiangsu University of Science and Technology (2013)

Research on Green Renewal Strategy Based on LID Adaptation Toolkit Selection--The Example of the North Bay Section of Macau Inner Harbour Terminal Neighborhood Linsheng Huang1,2(B) 1 Faculty of Humanities and Arts, Macau University of Science and Technology,

Avenida Wai Long, Taipa, Macau, China [email protected] 2 Faculty of Civil Engineering, Putian University, Xueyuan Middle Street, Fujian, China

Abstract. This paper focuses on the application of low impact development (LID) strategy in the adaptation of urban built environment and nature, pointing out that the sustainable regeneration of marginal historical areas should be highlighted in the practice of urban development from incremental large-scale development to stock quality improvement. This paper focuses on the specific problems faced by the gradually depolarized and fragmented historic space of the North Bay section of Macau’s Inner Harbour Terminal, with the operational model and technical approach of low-impact development, combining the summary and cross-examination of existing engineering cases, to construct an adaptive strategic toolkit for regional regeneration. The green regeneration for the North Bay section of the inner harbor pier is used as a framework to analyze its role in solving flooding problems, optimizing the spatial interface, and bringing social benefits, as well as its contribution to the sound development of the historic space of the Macao city port. Keywords: Low impact development · Inner harbor pier · Technical strategy · Toolkit · Practical significance

1 Background 1.1 Low Impact Development (LID) Model LID, or Low Impact Development, was first proposed by the United States in the 1990s from the perspective of ecology [1], mainly focusing on the design and bioretention of stormwater management sites, and then gradually involving planning and design, technical measures and maintenance management. Low Impact Development (LID) can be considered as a sustainable model of urban regeneration based on decentralized, micro-renewal and small-scale control measures under cost control to enhance urban adaptation to nature. The LID adaptation toolkit is one of the tools to select the best matching regeneration strategy for different areas. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 258–267, 2023. https://doi.org/10.1007/978-981-19-3983-9_23

Research on Green Renewal Strategy Based on LID Adaptation Toolkit

Fig. 1. Shanghai Sponge city practice. Reference: https://baijia hao.baidu.com/s?id=160066997 8082459030)

259

Fig. 2. U.S. Green street diagram. Reference: http:// www.yuanlin8.com/landscape/7430.html

Fig. 3. North Bay section of inner harbor pier

The focus of LID research is different in here and abroad. Domestic research focuses on urban road landscape planning, road stormwater landscape design, urban stormwater flood control and drainage management, and the selection and layout of road stormwater facilities, such as the urban environmental management and mitigation of internal disasters in the sponge city renovation in China guided by the LID concept (e.g. Fig. 1); foreign research focuses on both stormwater landscape design and management. In this way, according to the actual situation of each region, the management manual of stormwater runoff has been introduced [2], such as the U.S. “Michigan Low Impact Development Manual”, and the Australian Water Sensitive Urban Design (WSUD) (e.g. Fig. 2), which are used to guide the specific practice. 1.2 The North Bay Section of Macau Inner Harbour Terminal The Macau Inner Harbor Pier area is an early site of human activities in Macau, which has accumulated deep urban memory and witnessed the historical transformation of Macau changing from a small fishing village before the opening of the port to a modern city. The 16 existing pier buildings from the 1940s to 1980s in the North Bay section of the inner harbor retain the early reclaimed land pattern in terms of bayside space, road system, and ramp space (e.g. Fig. 3), with strong cultural presence and deep memory of the place. And it is a rare historical and cultural landscape of the fishing port that integrates port history, neighborhood environment and architectural culture. From the perspective of the current problems the neighborhood environment, the lack of detailed planning has led to the disorderly increase and decrease in the construction and renovation of urban infrastructure in the North Bay section of the Inner Harbor Pier in recent years, and lack of control and implementation standards for the elements, in the pier architecture and landscape a greater threat. With the gradual implementation of

260

L. Huang

the dual-carbon goal, the lack of green facilities has further made the North Bay section of the inner harbor pier gradually become a marginal space of the city of Macau. At the same time, the development of a strategic toolkit based on the specific mode of LID application for realistic needs has become an urgent issue for the urban historic space of Macao.

2 Analysis of Current Problems and Case Studies 2.1 Problem ANALYSIS Flooding Problem. In recent years, global warming, sea level rise, and extreme weather conditions have been occurring frequently, testing Macau’s urban drainage, water storage, and flood prevention capabilities. In terms of relative elevation, the elevation of the ground level of the inner harbor pier is (+1.3 m−2.3 m) MSL, which is lower than the storm surge height (+1.95–3.78 m) MSL during typhoons, and the relatively low terrain increases the probability of flooding (e.g. Fig. 4). According to statistics, the North Bay section of the Inner Harbor Pier suffers from flooding to varying degrees throughout the year (e.g. Fig. 5), with flooding lasting up to 8 h, which seriously affects the use of the North Bay section of the Pier and the production and living of the residents in the city.

Fig. 4. List of historical flood heights and tide heights (storm) in the North Bay section of Inner harbour pier.

Hard Impact of Waterfront Barges. Since 1831, the inner harbor has undergone many large-scale artificial reclamation. Due to the lack of technology and concept, the early waterfront areas of the inner harbor terminals were mainly built with hard barges (e.g. Fig. 6). The non-ecological shoreline formed by the hard barge broke the original natural shoreline pattern, destroyed the river ecosystem in the pier area [3], and completely changed the ecological appearance of the original mangrove wetlands along the inner harbor. This anti-ecological reclamation is one of the main reasons for the extensive flooding and ecosystem problems in the inner harbor.

Research on Green Renewal Strategy Based on LID Adaptation Toolkit

Fig. 5. Flooding

261

Fig. 6. Current state of hard barges

Fig. 7. Street bottom interface and mid-dimensional interface status.

Single Multi-level Street Interface. The streetscape of the North Bay section of the Inner Harbor Pier is similar to that of other traditional streets, showing more of a standardized street interface pattern, i.e., the ground floor interface of the street is mainly oriented by two two-way traffic lanes, with a high percentage of impervious material surfaces. The North Bay section of the street is narrow due to historical reasons, and the space of the ramp is occupied by hawkers and stores, which makes it even more constrained. Motorized vehicles, non-motorized vehicles, and pedestrians are not effectively separated, resulting in a low sense of safety. There is a serious lack of greenery in the streets, and the surface is overly hardened, resulting in stagnation and low infiltration and storage of rainwater, making it impossible for surface water to infiltrate into the ground or drain out in time during heavy rains and astronomical tides, causing street flooding. At the same time, the chaotic flower cages and monotonous columns are not conducive to the control and enhancement of the overall character of the street community in the middle dimension of the building façade (e.g. Fig. 7). Fading Memory of the Place. After the 1980s, most of the pier buildings in the area lacked effective maintenance and repair, and make with the real problems of aging of the building body, disorganized internal space, deteriorated external form, and loss

262

L. Huang

of pier functions. The whole area gradually lost the support of the main industry and degenerated into a passive space of the city, face severing the daily connection with the citizens’ life [4], leisure, and work, and losing the sense of co-frequent development and public historical identity with the city.

2.2 Case Study The existing problems of flooding, barging, street interface and cultural recognition of the site in the North Bay section of the Inner Harbor Pier are one of the objective and prominent problems faced by the Port Terminal area. In order to respect the history and culture of the port terminal and to improve the environment of the area, the LID strategy is a better way to be chosen through the perspective analysis of several cases. Shanghai Yangpu Riverfront Demonstration Site Renovation. The Yangpu riverfront demonstration site in Shanghai was previously a low-lying, waterlogged area, which was the first foreign yarn factory in Shanghai since the construction of Yangpu Road by the public concession in 1869. The LID approach is used to preserve the geomorphic state, using rain gardens, plant configurations, bioretention and other means to store precipitation and stagnate rainwater, taking into account both technical and aesthetic interests [5], and the urban industrial culture is continued (e.g. Fig. 8a). Transformation of Singapore Barge Terminal. The Singapore Barge Terminal is part of the historic district of Singapore, which was the center of the city’s early commercial and transportation port. By preserving, repairing, and faithfully restoring the historic buildings in the barge area [6], upgrading and transforming the space of the site, and designing the barge treatment and intimate platform, the barge terminal site is preserved and strengthened as a whole, and the transformation of its function to accommodate commercial tourism is promoted. It creates a historical urban open space against the background of urban skyscrapers, preserving the memory of the place and deepening cultural identity (e.g. Fig. 8b). Green Streets in Vermont, USA. Based on the LID strategy, different cities in the U.S. established green street manuals for local conditions, Vermont published the Vermont Green Street Guide on the basis of the manuals, proposing five levels of street green renewal, i.e., Level 1 increases the amount of vegetation on both sides of the street and reduces impervious surfaces, and Level 2 increases the amount of vegetation on both sides of the street and reduces impervious surfaces. Level 2 is a large number of landscaped areas and a wide tree canopy to retain rainfall, Level 3 is the installation of stormwater landscaping facilities, Level 4 is a combination of Level 2 and 3 features proposed to be integrated with multiple access modes, and Level 5 is the management and renovation of the overall scope of the street. The effectiveness, means, cost, and complexity of the project increase step by step, and in view of the built environment and actual urban traffic conditions, the first, second, and third level of street renovation is generally the focus [7] (e.g. Fig. 8c).

Research on Green Renewal Strategy Based on LID Adaptation Toolkit

a Shanghai Yangpu Riverfront

b Singapore Barge Terminal

Update

263

c U.S. Navy Pier Philadelphia

Fig. 8. Three case diagrams.

The above cases all take into account the scale, status, capital and other constraints of the region, combine cultural, landscape, and facility elements with parallel characteristics, and respect regional attributes to promote regeneration practices. To summarize the specific application of different strategies of LID presented in the cases (e.g. Fig. 9), it is clear that they are functionally effective but also have limitations in use. For the North Bay section of the Inner Harbor Pier, the above case studies will serve as a useful toolkit to guide the selection of green regeneration strategies for the North Bay section in terms of development intensity, capital investment, regional space, and effectiveness.

3 Strategy Selection Basis and Toolkit Construction 3.1 Basis for Strategy Selection The choice of specific strategies for LID varies from region to region due to different urban problems, policy context, and ideology. To summarize the above case studies and actual situations, the main approaches and practices of LID include bioretention areas, cisterns and rain barrels, grass pond filter strips, green roofs (photovoltaic panels), site design, permeable paving, road and parking lot design, vegetated pockets, and infiltration wells. However, not every strategy is suitable for the regeneration of the North Bay section of the Macau Ferry Terminal, facing the main considerations being: First, The built-up area is relatively stable and has a traffic function. The area has been built for many years and can be spatially divided into a riverbank area, a road traffic area, and a pedestrian area of the rides, etc. The overall environment of the area is relatively stable and is not suitable for large-scale planning and renovation; Second The area has many historical elements and is not suitable for high-intensity development from the perspective of heritage conservation and utilization [8] (e.g. Fig. 9); Third, The space in the area is compact and there are no patches of unused vacant land. The area is spatially compact and there are no patches of idle vacant land, which makes it unsuitable for regeneration with high land use requirements; Fourth, Some of the fishery functions are still present in the area, so the needs of industrial development and the appropriateness of the strategy should be taken into consideration. In addition, given that the North Bay section is located in the marginal spatial area of the city of Macau, the current social concern is low and the inclination of capital is limited, so the high cost of development investment is not appropriate.

264

L. Huang

3.2 Construction of the LID Strategy Toolkit Based on the above case study, we further analyze the implementation effects of LID strategy in terms of functional role, economic cost and landscape appreciation, and based on the realistic needs and implementation limitations of the North Bay section of the Inner Harbor Pier, we establish a matching analysis table for the green regeneration strategy of the North Bay section under the principles of minimum intervention, minimum cost and maximum effect. The regeneration effect on the street area is strong and prominent, and the vegetation paved area is stronger and has average function ((e.g. Fig. 10), and this is used to construct the LID adaptation strategy toolkit for the North Bay section of the Inner Harbour Terminal.

Fig. 9. List of historic buildings

Fig. 10. LID strategy matching analysis

Fig. 11. Green renewal strategy for the North Bay section of inner harbour terminal based on LID adaptation toolkit.

Bioretention. The North Bay section of the Inner Harbor Terminal can be constructed as a bioretention area by combining the underutilized space formed by the spacing between

Research on Green Renewal Strategy Based on LID Adaptation Toolkit

265

the terminal buildings and the stepped greenery to soften the hard embankment. Soil, plants and microorganisms can be used to treat stormwater, so as to alleviate surface runoff [9], and at the same time enhance the ornamental landscape of the waterfront. Low-Impact Site Design. Low-impact site design is a green regeneration strategy that has minimal impact on the historic streetscape and features of the North Bay section. From the bottom interface, as in the Yangpu River demonstration section and Vermont green streets, the integrated street site consideration, the unused riverfront walkway is simply transformed into a landscape grass pond planting surface or modular grass pond, the elevation is lowered to below the road surface, while both sides of the carriageway and the curb hard interface can be considered in combination with gravel paved drainage ditches to increase the drainage surface, thus using sunken grass pond, permeable curb interface, etc. Through scientific planning, the rainwater infiltrated by grass ponds and ditches will form a network through underground water collection structures or drainage systems to effectively mitigate surface runoff and reduce flooding caused by heavy rainfall. From the middle dimension, the vertical greening transformation can be carried out by using the skin of the new building façade and the vertical surface of the building such as the columns of the rider building, which are incompatible with the pier building, to intercept rainwater in multiple directions, reduce the probability of flooding and optimize the overall landscape interface of the street, weakening the huge contrast between the old and new buildings along the street surface. Permeable Paving Blocks. At present, the North Bay section of the inner harbor pier is paved with cement, whether it is for the riding building or the pedestrian walkway along the river. For example, in the old port of Marseille, France6 , from the city landscape beautification and maintenance of the historical environment, the selection of local characteristics of permeable paving blocks to update the hardened walkway, forming a linear permeable area of the street, slowing down the speed of runoff into the underground drainage network and less likely to cause long-term accumulation of water on the ground [10], but also play a role in enhancing the historical landscape of the port. Vegetated Pockets. Given the spatial constraints of the North Bay section, for the existing site, long and shallow vegetated pockets or vegetated ditches can be created along the riverwalk to mitigate surface runoff and facilitate stormwater infiltration. From these analysis, it can be seen that different strategic choices of LID adaptation toolkits will inevitably have different effects on the application targets. In light of the reality of the North Bay neighborhood, a flexible approach to building a strategic toolkit is undoubtedly the best guarantee for the neighborhood’s vitality (e.g. Fig. 11).

4 Practical Implications of LID Adaptation Toolkit 4.1 Enhance the social value of the neighborhood through green renewal The revitalization of the North Bay section of the Inner Harbour Pier must combine the protection of the historical environment with the protection of the landscape environment,

266

L. Huang

through the selection and intervention of the LID adaptation toolkit strategy, the aquatic ecological system of the North Bay section of the neighborhood is complemented, from softening the hard barges, optimizing the spatial interface of roads and building facades, and reorganizing the ecological system and the industrial environment. By implanting the cultural elements of the pier, the enhancement of the street environment will definitely increase the interactive relationship with public life, reconstruct the ecological system and industrial landscape of the street, and strengthen the cultural memory of the street. 4.2 Green Renewal of Infrastructures to Promote Diversified Economic Models Based on the LID adaptive toolkit, the original monotonous pedestrian space is broken up and a sustainable cultural slow walking system such as a riverfront walkway and a pedestrian walkway with a riding building is constructed to maintain the diversity of landscape corridors. In order to restore the fishing port style of the North Bay section of the Inner Harbor Pier, upgrade and adjust the industrial attitude along the shoreline, and gradually form a sustainable urban waterfront leisure district with tourism and service industries as the core under the premise of environmental improvement, and integrate the culture of fishery and pier to create a waterfront leisure place for citizens and enhance the attractiveness of the district. Summary The LID adaptive toolkit strategy is not a unidirectional choice, but a multidimensional possibility. Based on the above analysis, it can be seen that the LID adaptation toolkit for the North Bay section of the Inner Harbour Pier, which is based on bioretention, low-impact site design, permeable paving blocks and vegetation pockets, can meet the current needs and desires of the neighborhood enhancement and achieve the goals of regional environmental, social and economic development, as well as provide a useful reference for the green regeneration practice of marginal urban historic spaces.

References 1. Sun, Z.P.: Xiamen,s infrastructure based on the concept of “low impact concept” of retrofit sutdy, Anhui University of Construction (2021) 2. Zhi, X.f.: Research on public facilities design of urban waterfront space under the interaction design concept, Harbin Institute of Technology, pp. 8–11 (2016) 3. Deng, Y.: Renovation and adaptive-reuse for the waterfront historical industrial estate: analysis on the renewal along Singapore River. Mod. Urban Res. 8, 25–32 (2008) 4. Sun, M., Li, Z.Y.: Research on regeneration design strategy of waterfront industrial heritage in urban inner fringe case study on Philadelphia navy yard. Urban Architect. (8), 26–29 (2017) 5. Zhang, M., Zhang, J., Qin, S.: Media of landscape-four stepped narrations from the construction of rainwater garden at Yangpu waterfront public space, Chin. Landscape Architect. (7), 49–54 (2021) 6. Liu, S.Y., Yang, F.: Adaptive rouse of historic buildings/neighborhoods in the mind frame of sustainable development. Hist. Building Conserv. 2, 4–11 (2009) 7. Fan, W.: Study on rainwater landscape facility design of green street—a case of Vermont in the USA. Urbanism Architect. 18(17), 148–150 (2021)

Research on Green Renewal Strategy Based on LID Adaptation Toolkit

267

8. Lou, J.K., Lin, A.L.: Research on street green space landscape preference based on narrative preference method proceedings of the annual china urban planning conference, pp. 26–29 (2021) 9. Wang, Y.C.: Research on the prevention and control of waterlogging in soutuern china based on low impact development model. Water Conservancy Sci. Technol. Econ. 08(4), 17–20 (2021) 10. Wang, X.H.: Study on the Renewal Planning of Dalian Port Area Under the Perspective of Historical Heritag, Harbin Institute of Technology (2019)

The Impact of Urbanisation on Catchment Discharge in Malaysia: A Case Study on Sungai Selangor Catchment Mayuran Jayatharan(B)

, Andreas Aditya Hermawan , Amin Talei , and Izni Zahidi

Discipline of Civil Engineering, School of Engineering, Monash University Malaysia, 47500 Bandar Sunway, Selangor Darul Ehsan, Malaysia [email protected]

Abstract. Malaysia is one of the countries that experience a tropical climate with an abundant amount of rainfall. However, this large amount of rainfall paired up with rampant development has increased the severity and frequency of floods occurring in the nation. This research investigates the utilisation of a hydrological model to assess the effects of urbanisation on the river flow regime. The software used to develop the hydrological model for this research is the MIKE 11 NedbørAfstrømnings-Model or better known as MIKE 11 NAM. Urban development is captured by developing runoff coefficients with land use and land cover (LULC) data. The applicability of Landsat-8 satellite images in determining LULC is also explored within this research. In addition, future development scenarios are developed based on the rate of urbanisation within the Sungai Selangor catchment. The results show that the MIKE 11 NAM software captured the river flow regime of the study site relatively well, with overall performance criteria score for the coefficient of determination, r2 = 0.7 and Nash- Sutcliffe efficiency, CE = 0.6. Landsat-8 satellite images can estimate the rate of urbanisation within the study area. The runoff coefficients developed from the LULC data improved the overall performance criteria score. Lastly, the increase of developed areas in the expanse of forest area caused a sharp increase in the value of the runoff coefficient. Keywords: Urbanisation · Hydrological modelling · Tropical flood · MIKE 11 NAM; Land use · Landsat-8

1 Introduction The rapid rate of urbanisation has caused an increase in the severity and frequency of river floods occurring globally. In addition, the increase in population density and rampant development in flood-prone areas has intensified the flood tolls. Thus, it is essential to assess the effects of urban development on flood severity. Studies in the past suggest that the type of land use and land cover (LULC) significantly influences flood peaks and flood volume [1, 2]. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 E. Strauss (Ed.): ICOCE 2022, LNCE 276, pp. 268–280, 2023. https://doi.org/10.1007/978-981-19-3983-9_24

The Impact of Urbanisation on Catchment Discharge

269

The increase in developed areas tends to lead towards higher impervious surfaces. These impervious surfaces contribute to larger discharge peaks and volume due to rainfall being unable to infiltrate the soil. Thus, the quantification of impervious land becomes vital in determining the influences of LULC on urban development. Based on the paper done by Chabaeva, Civco [3], there are three main approaches for quantifying the amount of impervious cover. These methods consist of (1) interpretive detection, (2) spectral pattern recognition, and (3) statistical and mathematical modelling. Among these three methods, interpretive detection produces the most accurate results, although at the same time also the most time-consuming. In recent years, remote sensing techniques have been popular for researchers to quantify and classify land use within an area of interest (AOI). Landsat images have been at the forefront as the primary data source for reliable remote-sensing images for LULC analysis [4–7]. Landsat images are moderate resolutions satellite images provided by the United States Geological Survey (USGS), freely accessible to the public domain. Landsat-8 is the latest satellite launched into space in 2013 and takes approximately 725 images per day. Users can access these images as soon as 8 h after acquisition [8]. Therefore, Landsat images are an ideal choice for researchers to determine long-term land-use changes within a particular AOI. The current or past status of LULC is important for hydrologists as it allows the determination of runoff coefficient within an AOI. Runoff coefficient, in simple terms, can be defined as the ratio of runoff to rainfall [9]. It is predominantly used as a parameter in hydrological modelling, where it represents the amount of runoff generation within a catchment area. In the past, the runoff coefficient has been used to assess the influence of land-use types on runoff generation. It was found that land use type had a significant influence on runoff coefficient compared to other factors such as topography, vegetation cover, soil type, and moisture content of the soil [10–12]. Runoff coefficient is an essential parameter within hydrological models [13]. In the past, researchers have tried to establish a proper methodology in assessing the effects of LULC on river flood regimes. The paper done by Salvadore, Bronders [14] states that there is currently no universal methodology for accurately simulating the urban water system. However, the article suggests that remote sensing data and measurable parameters within hydrological models might be the step to reduce uncertainties. Thus, this research aims to assess the relationship between urbanisation and river flood discharge by (1) investigating the applicability and reliability of utilising Landsat Images and a hydrological model in determining runoff coefficient, (2) assessing the effects of urbanisation towards river flood discharge in future development plans.

2 Study Site and Data Used The study area chosen for this research is the Sungai Selangor River basin situated within the west coast of Peninsular Malaysia in the state of Selangor. This area has a relatively humid climate with slight variation in temperature throughout the year [15]. The catchment area of the river basin is 1955 km2 and length of the main river is 110 km with several tributaries. The main river experiences an average discharge of 57 m3 /s

270

M. Jayatharan et al.

throughout the year with seasonal rainfall variations causing the flow to vary between 23 m3 /s and 122 m3 /s [16]. Historical records suggest that occasional overspills from the Northeast monsoon cause large-scale heavy rainstorms in December with a mean monthly rainfall ranging between 200 mm to 260 mm. The dry months from January to February alongside June to August have a mean monthly rainfall range from 120 mm to 150 mm [16]. The LULC within the study area is obtained from the Department of Town and Country Planning Malaysia (JPBD). The upstream region of the basin consists of a mountainous area with forested foothills. The middle region comprises agricultural plantations comprising rubber trees and oil palm. The downstream region consists mainly of lowlying peat swamp forests. Furthermore, developed areas make up 47% of the study area. The developed areas consist of industrial, commercial, residential, and general utilities infrastructures. The remaining 53% consists of water bodies, reserved forest area, and bare land. In addition, a total of 19 telemetric rainfall stations within the catchment area were considered for this research. Hydrological data of 15-min resolution is provided by each station and was obtained from the Hydrology Unit of the Department of Irrigation and Drainage Malaysia (DID). The data range considered is from the year 2013–2018.

3 Methodology 3.1 Development of the Hydrological Model The hydrological model was developed for the downstream streamflow station of the study area by using the MIKE 11 NAM software. The MIKE 11 NAM model is a conceptual model developed by the Danish Hydraulic Institute [17]. An event-based rainfall-runoff (R-R) modelling approach was adopted to tackle the several data gaps within the dataset used. Flood events were used for calibrating and validating the hydrological model.. The total catchment area for the downstream flow station is 1186 km2 , shown in Fig. 1. A total of 55 rainfall-runoff events were considered to develop the hydrological model, where 45 events were considered for calibration and 10 for validation. This research adopts several performance criteria to assess the model performance, including the coefficient of determination (r2 ), Nash – Sutcliffe coefficient of efficiency (CE) and relative peak error (RPE). Table 1 below summarises the criteria scores used to evaluate the flood events based on past research in assessing hydrological models [18–20].

The Impact of Urbanisation on Catchment Discharge

271

3.2 LULC Classification Landsat-8 images were used to determine and classify the land use within the catchment area. Landsat-8 images were obtained from the USGS earth explorer website (https:// earthexplorer.usgs.gov/), with a total of 12 images spanning from 2015 to 2021. Landsat images with a cloud cover of below 30% were considered before further inspection was conducted for land use classification. The classification was done using ArcGIS software, and the Supervised classification method was adopted. Land use types were classified into four main types: developed area, water bodies, forest area, and agricultural land. An accuracy assessment was done to each Landsat-8 image considered for the classification. This assessment was done by reusing the chosen Landsat-8 image as ground truth survey points were not available for this research. The results obtained from the land use classification done with the Landsat-8 images were then compared with official land use data obtained from JPBD. 3.3 Development and Validation of Runoff Coefficient The areas obtained from the official land use data were used to develop the runoff coefficient for the catchment area. The runoff coefficient was generated using the impervious coefficients within the local stormwater management guideline, MSMA [21]. Flood events were used to validate the runoff coefficient derived from the local guideline. The model parameter that represents imperviousness was tweaked to assess its performance. The flood hydrograph generated using the runoff coefficient derived from the official land use data were compared with the simulated hydrograph obtained from the validated hydrological model. The performance criteria values obtained from the simulated hydrographs determine the capabilities of utilising land use data to derive runoff coefficients as input for hydrological models. Table 1. Performance rating table [18–20] Model performance criteria

Performance rating Poor

Fair

Fairly good Good

Excellent

Coefficient of determination (r 2 )

0.5

>0.6

>0.75

>0.9

Nash-sutcliffe coefficient of efficiency (CE)

0.5

>0.65

>0.7

>0.8

Relative peak error (RPE)

>0.7