Seismic Performance Analysis of Concrete Gravity Dams by Gaohui Wang, Wenbo Lu, Sherong Zhang

Advanced Topics in Science
and Technology in China 57
Gaohui Wang
Wenbo Lu
Sherong Zhang
Seismic Performance
Analysis of Concrete
Gravity Dams
Advanced Topics in Science and Technology
in China
Volume 57
Zhejiang University is one of the leading universities in China. In Advanced Topics
in Science and Technology in China, Zhejiang University Press and Springer jointly
publish monographs by Chinese scholars and professors, as well as invited authors
and editors from abroad who are outstanding experts and scholars in their fields.
This series will be of interest to researchers, lecturers, and graduate students alike.
Advanced Topics in Science and Technology in China aims to present the latest
and most cutting-edge theories, techniques, and methodologies in various research
areas in China. It covers all disciplines in the fields of natural science and
technology, including but not limited to, computer science, materials science, the life
sciences, engineering, environmental sciences, mathematics, and physics.
If you are interested in publishing your book in the series, please contact:
Dr. Mengchu Huang, E-mail: mengchu.huang@springer.com
This book series is index by the SCOPUS database.
More information about this series at http://www.springer.com/series/7887

Gaohui Wang Wenbo Lu
• •
Sherong Zhang
Seismic Performance
Analysis of Concrete Gravity
Dams
123
Gaohui Wang Wenbo Lu
State Key Laboratory of Water Resources State Key Laboratory of Water Resources
and Hydropower Engineering Science and Hydropower Engineering Science
Wuhan University Wuhan University
Wuhan, China Wuhan, China
Sherong Zhang
State Key Laboratory of Hydraulic
Engineering Simulation and Safety
Tianjin University
Tianjin, China
ISSN 1995-6819 ISSN 1995-6827 (electronic)

Advanced Topics in Science and Technology in China
ISBN 978-981-15-6193-1 ISBN 978-981-15-6194-8 (eBook)
https://doi.org/10.1007/978-981-15-6194-8
Jointly published with Zhejiang University Press
The print edition is not for sale in China Mainland. Customers from China Mainland please order the
print book from: Zhejiang University Press.
© Zhejiang University Press and Springer Nature Singapore Pte Ltd. 2021
This work is subject to copyright. All rights are reserved by the Publishers, 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 publishers, 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 publishers nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publishers remain 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
In order to meet the ever-increasing demand for power, irrigation, drinking water,
etc., numerous high concrete dams are being built or to be built and the majority
of them are located in active seismic regions. Due to the low tensile resistance of
concrete material, concrete dams will inevitably occur cracking damage under
strong ground motions. Considering that the possible dam failure due to seismic
activities could result in heavy loss of human life and substantial property damages,
seismic safety evaluation of high dams remains a crucial problem in dam con-
struction. To achieve a reasonable assessment of the seismic safety of dams, study
on the cracking damage processes and potential failure modes of concrete dams
subjected to strong earthquakes is indispensable and deserves more investigation
and attention. This book addresses the challenges in approach and application of
seismic performance evaluation of concrete gravity dams.
Earthquake ground motion is a complex natural phenomenon associated with the
abrupt energy release caused by the fault rupture, and it is influenced by many
factors, such as earthquake source mechanism, propagation path of waves, and soil
condition at the site. The seismic performance assessment of concrete dams under
strong ground motions depends primarily on how accurate the constitutive models
adopted in describing the structural behavior and in predicting possible future
earthquake events at the precise site. The objectives of this book are aimed to
evaluate the seismic performance of concrete gravity dams with consideration of
the effects of strong motion duration, mainshock–aftershock seismic sequence, and
near-fault ground motion. Both extended finite element method (XFEM) and
concrete damaged plasticity (CDP) model are employed to characterize the
mechanical behavior of concrete gravity dams under strong ground motions
including the dam–reservoir–foundation interaction. In addition, the effects of the
initial crack, earthquake direction, and cross-stream seismic excitation on nonlinear
dynamic response and damage-cracking risk of concrete gravity dams to strong
ground motions are discussed.
This book is comprised of ten chapters.
Chapter 1 provides an overview on the effects of strong earthquake ground
motions on concrete dams.
v
vi Preface
Chapter 2 provides two fracture modeling approaches, i.e. XFEM and CDP
model, to describe nonlinear dynamic response and seismic failure process of
concrete gravity dams under strong ground motions.
Chapter 3 deals with the numerical prediction of crack propagation in concrete
gravity dams with single and multiple initial cracks. The effects of the initial crack
position and length on the crack propagation and seismic response of dam–reser-
voir–foundation systems are studied.
Chapter 4 generalizes five potential failure modes of concrete gravity dams by
applying the incremental dynamic analysis method based on the XFEM.
Chapter 5 proposes local and global damage indices to assess quantitatively the
effects of strong motion duration on the accumulated damage of concrete gravity
dams. The definition of single-component durations that exhibit the strongest
influence on structural damage is determined.
Chapter 6 proposes a general integrated duration definition for multi-component
seismic excitations based on the existing concept of strong motion duration.
Chapter 7 assesses the effects of aftershocks on concrete gravity dams and
provides a quantitative description of the damage demands prior to and following
the aftershocks.
Chapter 8 examines the influence of earthquake direction on seismic perfor-
mance of concrete gravity dams subjected to seismic sequences.
Chapter 9 proposes a systematic approach for seismic performance evaluation of
concrete gravity dams subjected to near-fault and far-fault ground motions based on
the presented performance criteria.
Chapter 10 establishes a three-dimensional model to obtain the more realistic
seismic response of concrete gravity dams under strong ground motions. This
chapter conducts a systematic study on the 3D seismic damage-cracking behavior
of concrete gravity dams with all the following factors considered: contraction joint
nonlinearity, cross-stream earthquake excitation, and dam–foundation–reservoir
interaction.
The authors gratefully appreciate the supports from the National Natural Science
Foundation of China (No. 51939008) and the Technology and Industry for National
Defense of China (No. JCKY2018110C162).
Wuhan, China Gaohui Wang

Wuhan, China Wenbo Lu
Tianjin, China Sherong Zhang
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Objective and Scope of This Research . . . . . . . . . . . . . . . . . . . 4
1.3 Organization of the Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2 Comparative Analysis of Nonlinear Seismic Response
of Concrete Gravity Dams Using XFEM and CDP Model . . . .... 11
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 11
2.2 Method for Dynamic Failure Analysis of Concrete
Dams Under Strong Earthquake . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2.1 Prototype Observation . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2.2 Model Test Method . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2.3 Numerical Simulation Method . . . . . . . . . . . . . . . . . . . 16
2.3 eXtended Finite Element Method (XFEM) . . . . . . . . . . . . . . . . 21
2.3.1 XFEM Approximation . . . . . . . . . . . . . . . . . . . . . . . . 21
2.3.2 Enrichment Functions . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.3.3 Discrete Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.4 Concrete Damaged Plasticity (CDP) Model . . . . . . . . . . . . . . . 27
2.4.1 Damage Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.4.2 Yield Criterion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.4.3 Flow Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.5 Lagrangian Formulation for Dynamic Interaction
of Dam-Reservoir-Foundation Systems . . . . . . . . . . . . . . . .... 32
2.6 Application of the Two Models in Concrete Gravity Dams .... 35
2.6.1 Description of Koyna Gravity Dam-Reservoir-
Foundation System . . . . . . . . . . . . . . . . . . . . . . . .... 35
2.6.2 A Comparative Study on Seismic Nonlinear
Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 37
vii
viii Contents
2.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3 Seismic Cracking Analysis of Concrete Gravity Dams
with Initial Cracks Using XFEM . . . . . . . . . . . . . . . . . . . . . . . . .. 53
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 53
3.2 Validation Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 55
3.3 Seismic Crack Propagation Analysis of Koyna Gravity
Dam with Initial Crack . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 57
3.3.1 Initial Cracking Models . . . . . . . . . . . . . . . . . . . . . .. 57
3.3.2 Seismic Crack Propagation Process with no Initial
Crack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 59
3.3.3 Seismic Crack Propagation Process for Single Initial
Cracking Models . . . . . . . . . . . . . . . . . . . . . . . . . . .. 60
3.3.4 Seismic Crack Propagation Process for Multiple
Initial Cracking Model . . . . . . . . . . . . . . . . . . . . . . .. 62
3.4 Seismic Crack Propagation Analysis of Guandi Gravity Dams
with Initial Cracks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 66
3.4.1 FEM Model and Material Properties . . . . . . . . . . . . .. 66
3.4.2 Initial Crack Position . . . . . . . . . . . . . . . . . . . . . . . .. 67
3.4.3 Crack Propagation Process of the Dam with no Initial
Crack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.4.4 Influence of Initial Crack Position . . . . . . . . . . . . . . . . 69
3.4.5 Influence of Initial Crack Length . . . . . . . . . . . . . . . . . 73
3.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4 Seismic Potential Failure Mode Analysis of Concrete Gravity
Dam–Water–Foundation Systems Through Incremental
Dynamic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 79
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 79
4.2 Nonlinear Dynamic Response of Guandi Dam Under Design
Peak Ground Acceleration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
4.2.1 FEM Model and Material Properties . . . . . . . . . . . . . . 81
4.2.2 Seismic Response and Crack Propagation Analysis . . . 82
4.3 Seismic Potential Failure Mode Analysis . . . . . . . . . . . . . . . . . 85
4.3.1 Database of as-Recorded Acceleration Records . . . . . . 85
4.3.2 Incremental Dynamic Analysis . . . . . . . . . . . . . . . . . . 89
4.3.3 Typical Failure Modes of the Guandi Gravity
Dam for Each Record . . . . . . . . . . . . . . . . . . . . . . . .. 91
4.3.4 Generalization of Potential Failure Modes for
Concrete Gravity Dams . . . . . . . . . . . . . . . . . . . . . .. 93
4.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 95
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 95
Contents ix
5 Correlation Between Single Component Durations and Damage

Measures of Concrete Gravity Dams . . . . . . . . . . . . . . . . . . . . . .. 99
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 99
5.2 Strong Motion Duration-Related Measure Used
in this Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
5.2.1 Definitions of Strong Motion Duration
for Single Component Ground Motion . . . . . . . . . . . . . 103
5.2.2 Accelerogram Selection and Correction . . . . . . . . . . . . 107
5.2.3 Strong Motion Duration Prediction . . . . . . . . . . . . . . . 107
5.2.4 Correlation Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 110
5.3 Seismic Accumulated Damage Indices . . . . . . . . . . . . . . . . . . . 113
5.4 Seismic Damage Analysis of Koyna Dam . . . . . . . . . . . . . . . . 115
5.4.1 Description of Koyna Gravity Dam Model
Used for Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . 115
5.4.2 Strong Motion Duration Effects on Accumulated
Damage of Concrete Gravity Dams . . . . . . . . . . . . . . . 116
5.5 Correlation Study Between Strong Motion Durations
and Damage Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
5.5.1 Damage Measures of Local and Global Damage
Indices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
5.5.2 Damage Measures of Peak Displacement
and Damage Energy Dissipation . . . . . . . . . . . . . . . . . 123
5.5.3 Identifying the Influence of Single Component
Duration on Damage Measures . . . . . . . . . . . . . . . . . . 125
5.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
6 Integrated Duration Effects on Seismic Performance
of Concrete Gravity Dams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
6.2 A General Definition of Integrated Duration . . . . . . . . . . . . . . . 135
6.2.1 Single Component Duration . . . . . . . . . . . . . . . . . . . . 135
6.2.2 Integrated Duration of Multi-component Ground
Motions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
6.3 Database of as-Recorded Acceleration Records . . . . . . . . . . . . . 137
6.3.1 Accelerogram Selection and Integrated Duration
Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
6.3.2 Relationship Between Integrated and Single
Component Durations . . . . . . . . . . . . . . . . . . . . . . . . . 137
6.4 Influence of Two-Component Ground Motions
on Nonlinear Dynamic Response . . . . . . . . . . . . . . . . . . . . . . . 141
6.4.1 Displacement Response . . . . . . . . . . . . . . . . . . . . . . . 143
6.4.2 Stress Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
x Contents
6.4.3 Damage Dissipation Energy Response . . . . . . . . . . . . . 145

6.4.4 Damage Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
6.5 Correlation Between Integrated Durations and Damage
Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
6.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
7 Damage Demand Assessment of Concrete Gravity Dams
Subjected to Mainshock-Aftershock Seismic Sequences . . . . . . . . . . 155
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
7.2 Mainshock–Aftershock Seismic Sequences . . . . . . . . . . . . . . . . 160
7.2.1 Construction Method of Mainshock–Aftershock
Seismic Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
7.2.2 As-Recorded Mainshock–Aftershock Seismic
Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
7.2.3 Correlation Between Ground Motion Characteristics
of Mainshocks and Major Aftershocks . . . . . . . . . . . . . 165
7.3 Finite Element Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
7.3.1 Koyna Dam–Reservoir–Foundation System . . . . . . . . . 167
7.3.2 Seismic Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
7.4 Nonlinear Behavior of the Koyna Dam–Reservoir–Foundation
System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
7.4.1 Structural Damage . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
7.4.2 Displacement Response . . . . . . . . . . . . . . . . . . . . . . . 170
7.4.3 Damage Dissipated Energy . . . . . . . . . . . . . . . . . . . . . 171
7.5 Estimation of Damage Demands for Mainshock–Aftershock
Seismic Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
7.5.1 Effects of as-Recorded Mainshock–Aftershock
Seismic Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
7.5.2 Comparative Analysis of Damage Demands Between
as-Recorded Seismic Sequences and Repeated
Earthquakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
7.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
8 Earthquake Direction Effects on Nonlinear Dynamic
Response of Concrete Gravity Dams to Seismic Sequences . . . . . . . 185
8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
8.2 Earthquake Incident Direction for Two-Dimensional
Seismic Performance Analysis . . . . . . . . . . . . . . . . . . . . . . . . . 187
8.3 Finite Element Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
8.3.1 Description of Finite Element Model
of Dam-Foundation-Reservoir Systems . . . . . . . . . . . . 188
8.3.2 Input Ground Motions . . . . . . . . . . . . . . . . . . . . . . . . 188
Contents xi
8.4 Effects of Earthquake Direction . . . . . . . . . . . . . . . . . . . . . . . . 190

8.4.1 Direction Effects of Single Earthquake Events . . . . . . . 190
8.4.2 Direction Effects of Mainshock–Aftershock Seismic
Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
8.4.3 Effects of Aftershock Polarity . . . . . . . . . . . . . . . . . . . 196
8.5 Influence of the Ground Motion Intensity on the Seismic
Sequence Direction Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
8.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
9 Seismic Performance Evaluation of Dam-Reservoir-Foundation
Systems to Near-Fault Ground Motions . . . . . . . . . . . . . . . . . . . . . 207
9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
9.2 Characteristics of Near-Fault Ground Motions . . . . . . . . . . . . . 210
9.2.1 Forward Directivity Effect . . . . . . . . . . . . . . . . . . . . . . 211
9.2.2 Fling Step Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
9.2.3 Hanging Wall Effect . . . . . . . . . . . . . . . . . . . . . . . . . . 213
9.3 Near-Fault and Far-Fault Ground Motion Records Considered
in This Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
9.4 Seismic Performance Evaluation Methods . . . . . . . . . . . . . . . . 215
9.5 Near-Fault Ground Motion Effects on Seismic Performance
of Concrete Gravity Dams Using Linear and Nonlinear
Evaluation Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
9.5.1 Seismic Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
9.5.2 Seismic Performance Evaluation Using Linear
Dynamic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
9.5.3 Seismic Performance Evaluation Using Nonlinear
Dynamic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
9.6 Nonlinear Dynamic Response of Concrete Gravity Dams
Subjected to Near-Fault and Far-Fault Ground Motions . . . . . . . 231
9.6.1 Nonlinear Displacement Response . . . . . . . . . . . . . . . . 232
9.6.2 Seismic Damage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
9.6.3 Damage Energy Dissipation . . . . . . . . . . . . . . . . . . . . 236
9.6.4 Identifying the Influence of Near-Fault Ground
Motions on Seismic Damage of Dams . . . . . . . . . . . . . 237
9.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
xii Contents
10 Deterministic 3D Seismic Damage Analysis of Guandi Concrete

Gravity Dam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
10.2 Material Properties of Mass Concrete and Contraction
Joint Nonlinearity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
10.2.1 Material Properties of Mass Concrete . . . . . . . . . . . . . 246
10.2.2 Contraction Joint Nonlinearity . . . . . . . . . . . . . . . . . . . 246
10.3 3D Lagrangian Finite Element Formulation . . . . . . . . . . . . . . . 248
10.4 Validation Test for 3D Model . . . . . . . . . . . . . . . . . . . . . . . . . 249
10.5 3D Guandi Gravity Dam-Reservoir-Foundation System . . . . . . . 250
10.5.1 Introduction to Guandi Gravity Dam . . . . . . . . . . . . . . 250
10.5.2 3D Finite Element Model . . . . . . . . . . . . . . . . . . . . . . 253
10.6 Nonlinear Seismic Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . 255
10.6.1 Seismic Damage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
10.6.2 Maximum Stream Displacement . . . . . . . . . . . . . . . . . 256
10.6.3 Contact Behavior of Contraction Joints . . . . . . . . . . . . 257
10.7 Evaluation of the Key Factors Affecting Damage-Cracking
Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
10.7.1 Discussion of Results from 3D
Dam-Reservoir-Foundation Model . . . . . . . . . . . . . . . . 260
10.7.2 Comparisons of 3D and Quasi-3D Analysis
Results for Representative Monoliths . . . . . . . . . . . . . . 261
10.8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
Principal Symbols and Abbreviations
Symbols
B ¼ rw Matrix of derivatives of extended shape functions wri

Bef Strain–displacement matrix of the fluid element
b Body force vector per unit mass
bj Vector of corresponding additional degrees of freedom for
modeling crack faces (not crack tips)
C Damping matrix
Cc Damping matrix of the coupled system
Cf Elasticity matrix consisting of diagonal terms
E Elastic tensor
f External load vector
H Matrix consisting of interpolation functions of the fluid
element
hs Vector consisting of interpolation functions of the free surface
fluid element
K Stiffness matrix
Kc Stiffness matrix of the coupled system
Kf Stiffness matrix of the fluid system
Kf System stiffness matrix including the free surface stiffness
M Mass matrix
Mf Mass matrix of the fluid system
Mc Mass matrix of the coupled system
n External unit vector to C
Rf Time-varying nodal force vector for the fluid system
Rc Time-varying nodal force vector of ground acceleration

S
Deviatoric part of the effective stress tensor r
Sf Stiffness matrix of the free surface of the fluid system
t Applied traction force vector on the Neumann boundary Ct
ui Classical degrees of freedom for node i.
xiii
xiv Principal Symbols and Abbreviations
u_ Velocity
u Applied displacement vector on the Dirichlet boundary Cu
[u] Jump in the displacement
Uf Nodal displacement vector
Usf Vertical nodal displacement vector
Uc Vector of the displacement of the coupled system
U_ Nodal velocity vector
U_ c Vector of the velocity of the coupled system
€f
U Nodal acceleration
€c
U Vector of the acceleration of the coupled system
u€ Acceleration
x Sample gauss point
x* Closest point to x
r Cauchy stress tensor
C11 Bulk modulus of fluid
C22, C33 and C44 Rotation constraint parameter
DP&A Park and Ang index
DILi Local damage index for crack path i
DIG Global damage index
E Elasticity modulus
E0 Initial (undamaged) elastic stiffness of the material
EH Hysteretic energy
Ei Damage dissipation energy at the crack path i
Fy Elastic strength
Ed Damage dissipated energy
G Scalar plastic potential function
Fl1 ðxÞ and Fl2 ðxÞ Asymptotic crack-tip enrichment functions
Fl (r, h) Crack-tip enrichment functions
Gf Fracture energy
H(x) Heaviside function
H(t) Husid diagram as a function of time t
I Set of all nodes in the mesh
I0 Arias intensity
I0H Arias intensities of the horizontal component
I0V Arias intensities of the vertical component
I0H1 and I0H2 Arias intensities of the two horizontal ground motions
I1 First effective stress invariant
J2 Second effective deviatoric stress invariant
JI Set of nodes whose shape function support is cut by a crack
K1 I Set of nodes whose shape function support contains the first
crack tip in their influence domain
K2 I Set of nodes whose shape function support contains the second
crack tip in their influence domain
Principal Symbols and Abbreviations xv
Kc Strength ratio of concrete under equal biaxial compression to

triaxial compression
Li Total length to which crack path i is expected to grow
M Mainshock magnitude
M0 Aftershock magnitude
Mw Moment magnitude
N Number of seismic events
Ni(x) Shape function associated with node i
P Pressure
Px, Py and Pz Corresponding rotational stresses
P0 Pressure value at zero opening
Pz Rotational stress parameter
R Correlation coefficient
R′ Source distance
Sij Effective deviatoric stress tensor
Tp Predominant period
T0 Total duration of the record
TB Bracketed duration
TU Uniform duration
TE Effective duration
TS Significant duration
TS(70%) Time interval between 15 and 85% of the Husid diagram
TS(90%) Time interval between 5 and 95% of the Husid diagram
TH Strong motion durations of the horizontal component
TV Strong motion durations of the vertical component
TI Integrated duration
T H1 and T H2 Strong motion durations of two perpendicular horizontal
components
TBI Integrated bracketed duration
TUI Integrated uniform duration
TSI Integrated significant duration
TBH Bracketed duration of the horizontal component
TBV Bracketed duration of the vertical component
Un Normal component of the interface displacement
Wx Rotation about the axis x
Wy Rotation about the axis y
Wz Rotation about the axis z
a Ground acceleration
b Constant
bi1 and bi2 Length of the edges of the quadrilateral constant strain
elements beside node i on the upstream surface of the dam
c0 Initial contact distance
cl1 l2
k and ck
Vector of corresponding additional degrees of freedom which
are related to the modeling of crack tips
xvi Principal Symbols and Abbreviations
d Scalar stiffness degradation variable

dt Tension damage
dc Compression damage
g Acceleration due to gravity
fc0 Compressive strength of the concrete
ft Tensile strength of the concrete
h Depth of water
jt, Tensile damage variable
jc Compressive damage variable
lDi Length of the damage path in crack path i
m0i Westergaard virtual mass
n Unit outward normal to the crack at x*
nA and nB Index sets of the nodes of superposed element A and element B
p Normal contact pressure
p Effective hydrostatic pressure
q Mises equivalent effective stress
l Coefficient of friction
ufn Normal component of the free surface displacement
v Volume
yi Distance from node i to the water surface
(r, h) Local polar coordinate system with its origin at the crack tip
pffiffi
r sin h=2 Discontinuous across the crack plane, whereas the last three
functions are continuous
a and b Dimensionless material constants
m Poisson’s ratio
w Dilation angle measured in the p–q plane at high confining
pressure
wri Extended shape functions
3 Parameter that defines the rate at which the function
approaches the asymptote defined by w
C Boundary
Cu Dirichlet boundary
Ct Neumann boundary
q Initial mass density
r Effective stress
^max
r Algebraically maximum eigenvalue of r
rb0 Concrete strength under equal biaxial compression
rc0 Initial compressive yield stress
rc Effective compressive cohesion stresses
rt Effective tensile cohesion stresses
rt0 Uniaxial tensile stress at failure
e Total strain tensor
ee Elastic strain
ep Plastic strain
Principal Symbols and Abbreviations xvii
~epc Equivalent compressive plastic strain

~ept Equivalent tensile plastic strain
eel Elastic strain
ev Volumetric strain
k_ Nonnegative function referred to as the plastic consistency
parameter
dn Crack normal opening
dt Crack tangential sliding
cw Weight density of fluid
qf Mass density of the fluid
qw Mass density of water
XA Subdomain element A
XB Subdomain element B
f Viscous damping ratio
sc Cohesive traction applied on the discontinuity surface
Abbreviations
CDP Concrete damaged plasticity

CMOD Crack mouth opening displacement
COD Crack opening displacement
COSMOS Consortium of Organizations for Strong Motion Observation
Systems
CRCM Coaxial rotating crack model
DCR Demand–capacity ratio
DDA Discontinuous deformation analysis
DE-BE Distinct element–boundary element
DEM Distinct element method
DI Damage index
DP Drucker−Prager
FBG Fiber Bragg grating
FCM-VSRF Fixed crack model with a variable shear resistance factor
FEM Finite element method
IDA Incremental dynamic analysis
LEFM Linear elastic fracture mechanics
MDOF Multi-degree of freedom
MFCM Multidirectional fixed crack model
NLFM Nonlinear fracture mechanics
NPSBFEM3D 3D nonlinear polyhedron scaled boundary finite element
OMFCM Orthogonal multi-fixed crack model
PEER Pacific Earthquake Engineering Research Center
PGA Peak ground acceleration
xviii Principal Symbols and Abbreviations
PGD Peak ground displacement

PGV Peak ground velocity
PUM Partition of unity method
RC Reinforced concrete
RCC Roller-compacted concrete
RFPA Rock failure process analysis
SDOF Single degree of freedom
SMD Strong motion duration
SPH Smoothed particle hydrodynamics
XFEM Extended finite element method
Chapter 1
Introduction
1.1 Background
Dams are important lifeline engineering, which has contributed to the development
of civilization for a long time. In order to meet the ever-increasing demand for flood
control, hydropower, irrigation and drinking water etc., the majority of high dams
have been built, are being built or to be built in the southwest region of China with
active seismic activities. Table 1.1 shows some typical high dam projects, in which
the height of the dam is from 100 to 250 m. The design peak ground acceleration
(PGA) is also given in Table 1.1. It can be found that the design PGA for some dams
is very high. The seismic safety problem of high dams is very outstanding.
The earthquake is a devastating natural disaster. Due to the outburst of strong
earthquakes like Wenchuan earthquake in 2008 (8.0 M w ), Haiti earthquake in 2010
(7.3 M w ), Tohoku earthquake in 2011(9.0 M w ), and Chile earthquake in 2015 (8.3
M w ), it seems that the earth has come into an era with more and stronger earth-
quakes. When subjected to strong ground motions, mass concrete dams are likely
to experience cracking due to the low tensile resistance of concrete. Meanwhile,
the potential crack initiation and propagation would adversely affect the static and
dynamic performance of dams. As cracks penetrate deep inside a dam, its structural
resistance may be considerably weakened, thereby increasing the risk of the dam
failure and endangering the safety of the dam. However, the possible failure of dams
retaining large quantities of water can cause a considerable amount of devastation
in the downstream populated area during strong earthquakes. Hence, dams must be
completely safe and stable. The seismic safety evaluation of high dams remains a
crucial problem in dam construction.
The prototype observation is a very useful method to acknowledge the dynamic
and earthquake behavior of high dams to actual earthquakes. While there are many
high concrete dams throughout the world, only some of them have experienced
ground shaking induced structural damage. To name just a few, the Hsingfengkiang
dam, China, 1962; Koyna dam, India, 1967; and Sefid-Rud dam, Iran, 1990 are the
© Zhejiang University Press and Springer Nature Singapore Pte Ltd. 2021 1
G. Wang et al., Seismic Performance Analysis of Concrete Gravity Dams,
Advanced Topics in Science and Technology in China 57,
https://doi.org/10.1007/978-981-15-6194-8_1
2 1 Introduction
Table 1.1 Concrete gravity

Concrete gravity dam Maximum height (m) PGA (g)
dams with the height from
100 to 250 m in the southwest Longtan 216.5 0.20
region of China Huangdeng 203 0.251
Guandi 168 0.34
Xiangjiaba 162 0.222
Jinanqiao 160 0.399
Ahai 138 0.344
Kalasuke 121.5 0.21
Longkaikou 119 0.394
ones that have suffered from damage in earthquakes (Nuss et al. 2012), as shown in
Fig. 1.1.
The Hsinfengkiang dam is a concrete buttress dam with a crest length of 440 m, a
crest width of 5 m, and a maximum structural height of 105 m, which was completed
in 1962 located Guangdong Province, China. The magnitude M 6.1 Hsinfengkiang
earthquake occurred on March 19, 1962 with the focal depth of 5 km. This earthquake
is considered to be a reservoir triggered (Tsung-ho et al. 1976). A horizontal crack
82 m long on the monoliths 13# to 17# developed on the right side of the downstream
face around the elevation of 108 m at which the slope changes abruptly. A few smaller
cracks developed on the left side of the dam at the same elevation as the crack on the
right side.
The Koyna dam is a concrete gravity dam, which was built from 1954 to 1963
located in the southwestern region of India. The dam has a crest length of 853 m,
a crest width of 14.8 m, a maximum structural height of 103 m, and a maximum
base width of 70.2 m. A significant change of slope on the downstream face is 37 m
below the dam crest. The Koyna earthquake occurred on December 11, 1967 with the
epicenter 13 km from the dam with a focal depth of 8–13 km. The magnitude is M 6.5.
The maximum accelerations were recorded at the foundation gallery in the stream
direction of 0.49 g, cross-stream direction of 0.63 g, and cross-stream direction of
0.34 g. The Koyna earthquake has caused very serious structural damage to the dam,
including horizontal cracks on the upstream and downstream faces of a number of
non-overflow monoliths around the elevation at which the slope of the downstream
face changes abruptly. Leakage was found in some of these monoliths near the
changes in the slope of the downstream face, implying the complete penetration
from the upstream face to the downstream face (Chopra and Chakrabarti 1973).
The Sefid Rud Dam is a concrete buttress dam, which was built from 1956 until
1962. The project is located in Manjil in Gilan Province, northern Iran. The crest
length of the dam is 425 m, the highest monolith is 106 m high. There are 23 buttresses,
14 m wide, with a web thickness of 5 m. The dam was originally designed to with-
stand a 0.25 g using the quasi-static method. An exceptionally strong ground motion
with a magnitude of M 7.6 hit the area of the dam on June 20, 1990. Two strong
aftershocks having magnitude in the 6.2–6.5 magnitude range occurred several hours
1.1 Background 3
(a)
(b)
(c)
Fig. 1.1 Location of earthquake induced crack. a Hsinfengkiang dam; b Koyna dam; c Sefid Rud
dam
4 1 Introduction
after the main shock. Damages to the dam structure consisted mainly of cracks along
the horizontal construction joints and of spalling of concrete along the vertical joints
between buttress heads (Ghaemmaghami and Ghaemian 2010). Following the earth-
quake, the most serious observed damage to the dam was horizontal cracks that
appeared in the upper parts of the monoliths, especially in the highest monolith. A
major crack ran almost the whole length of the dam at about 14 m below the crest.
Leakage was reported through some of the cracks. Other damage included minor
damage and displacement of all the gates, varying types of damage at the dam crest.
Nonlinear dynamic response and damage propagation process of high concrete
dams to strong ground motions are very complicated. Seismic safety evaluation of
high dams remains a crucial problem in dam construction. The investigation and
survey on the actual earthquake damage of high dams is an important way to analyze
the seismic performance and failure pattern of high dams. However, there is just a
little actual seismic damage of high dams to strong ground motions. The nonlinear
dynamic response, damage mechanism, cracking process, potential failure mode,
and seismic performance of high concrete dams under strong ground motions have
not been further understanding. In the absence of monitoring data of actual dam
damage process, dynamic model test and finite element method are the main means
to understand the nonlinear dynamic response behavior and earthquake damage
modes for high concrete dams, which are very important to dam seismic performance
evaluation.
1.2 Objective and Scope of This Research
This book addresses the challenges in approach and application of seismic perfor-
mance evaluation of concrete gravity dams. The seismic performance assessment of
concrete dams under strong ground motions depends primarily on how accurate the
constitutive models adopted in describing the structural behavior and in predicting
possible future earthquake events at the precise site. In fact, many nonlinear consti-
tutive models have also been developed in order to forecast the nonlinear dynamic
response of concrete dams under strong ground motions. There are based either on
the discrete crack approach (Ingraffea and Saouma 1985; Ayari and Saouma 1990;
Ahmadi et al. 2001; Omidi et al. 2013) or the smeared crack approach (Bhattacharjee
and Léger 1994; Léger and Leclerc 1996; Ghaemian and Ghobarah 1999; Wang et al.
2000). In addition, other models, such as plastic-damage model (Lee and Fenves
1998; Omidi et al. 2013), rock failure process analysis (Zhong et al., 2011), distinct
element method (Pekau and Cui 2004), discontinuous deformation analysis (Wang
et al. 2006), mesh-free particle method called smoothed particle hydrodynamics
(Das and Cleary 2013) and hybrid distinct element-boundary element (Mirzayee
et al. 2011), are also used in some cases to analyze the seismic failure behavior of
concrete dams.
It is well known that the earthquake ground motion is a complex natural
phenomenon associated with the abrupt energy release caused by fault rupture, and
1.2 Objective and Scope of This Research 5
it is influenced by many factors, such as earthquake source mechanism, propagation

path of waves, soil condition at the site, and so on. Earthquake strong ground motion
can be characterized by many different parameters including amplitude, frequency
content and strong motion duration, each of which reflects some particular feature
of the shaking. The most frequently used intensity parameters are the peak ground
acceleration (PGA), peak ground velocity (PGV), strong motion duration (SMD) and
spectral amplitude for different characteristic periods of the strong motion records.
The importance of the amplitude and frequency content has been universally recog-
nized. But strong motion duration is also one of the key intensity parameters that
contribute to the seismic performance of structures. However, current approaches for
earthquake-resistant design and structural analysis based on the response spectrum
have not yet considered the influence of the ground motion duration.
Earthquakes are usually part of a sequence of ground motions which can be defined
as the foreshock, mainshock, and aftershock. Seismic sequences characterized by a
mainshock followed by strong aftershocks in a short time have been observed in
many areas. Most aftershocks are located over the full area of fault rupture and either
occur along the fault plane itself or along other faults within the volume affected by
the strain associated with the mainshock. Strong aftershocks are dangerous because
they are usually unpredictable, and have the potential to cause additional structural
damage (Li and Ellingwood 2007; Ruiz-García and Negrete-Manriquez 2011; Goda
2012; Salami et al. 2019). The structure already damaged from the mainshock and
not yet repaired, which may be incapable of resisting the excitation of the strong
aftershocks, may be collapsed or become completely unusable under mainshock-
aftershock seismic sequences (Li et al. 2014; Omranian et al. 2018; Shokrabadi et al.
2018; Shokrabadi and Burton 2018). This characteristic is very important and the
influence of seismic sequences can’t be ignored. However, most structures designed
according to the modern seismic codes only apply a single earthquake on structure
modeling and analysis. The seismic performance of concrete gravity dams has been
examined by the isolated ‘design earthquake’ without taking the influence of multiple
earthquake phenomena into account. In this case, the structure may sustain damage in
the event of the ‘design earthquake’, and this single seismic design philosophy does
not take the effect of strong aftershocks on the accumulated damage of structures into
account. However, in the real earthquake event, bigger earthquakes may trigger more
and larger aftershocks and the sequences can last for years or even longer, the tremors
always occurred repeatedly. So, this structure located in seismic regions is not only
exposed to a single seismic event, but also to a seismic sequence. Therefore, it is of
great importance and urgency to investigate the influence of as-recorded mainshock-
aftershock seismic sequences on the dynamic response and accumulated damage of
concrete gravity dams.
Recordings obtained in recent earthquakes revealed that seismic ground motions
recorded within the near-fault region are quite different from the usual far-fault
ground motions observed at large distance in many respects such as the period of
earthquake continuity, peak ground acceleration, velocity and displacement, rupture
directivity, fling step and pulse properties (Chopra and Chintanapakdee 2001).
Forward directivity and fling effects have been identified by seismologists as the
6 1 Introduction
primary characteristics of near-fault ground motions (Mavroeidis and Papageorgiou

2003). Fault-normal components of ground motions often contain a large displace-
ment and velocity pulses, which expose the structure to high input energy in the begin-
ning of the earthquake. The pulses are strongly influenced by the rupture mechanism,
the slip direction relative to the site, and the location of the recording station relative
to the fault which is termed as ‘directivity effect’ due to the propagation of the rupture
toward the recording site (Bray and Rodriguez-Marek 2004). Because of the unique
characteristics of near-fault ground motion, the ground motions recorded in the near-
fault region have the potential to cause considerable damage to structures. Therefore,
structural response to near-fault ground motions has received much attention in recent
years. The effects of near-fault ground motions on many civil engineering structures,
such as buildings (Bhagat et al. 2018; Yaghmaei-Sabegh et al. 2019; Yang et al. 2019)
and bridges (Ardakani and Saiidi 2018; Xiang and Alam 2019; Xin et al. 2019), have
been investigated in many recent studies. It can be clearly seen from these studies that
the near-fault ground motion has a significant influence on the nonlinear dynamic
response of structures. Structures located within the near-fault region suffered more
severe damage than those located in the far-fault zone. The related study is therefore
a very important topic for the structural engineering. However, there is no sufficient
research about the near-fault ground motion effects on the seismic performance and
accumulated damage of concrete gravity dams.
For these problems above, the objectives of this book are aimed to evaluate the
seismic performance of concrete gravity dams with consideration of the effects
of strong motion duration, mainshock-aftershock seismic sequence, and near-fault
ground motion. Both extended finite element method (XFEM) and concrete damaged
plasticity (CDP) model are employed to characterize the mechanical behavior of
concrete gravity dams under strong ground motions including the dam-reservoir-
foundation interaction. In addition, the effects of the initial crack, earthquake
direction, and cross-stream seismic excitation on nonlinear dynamic response and
damage-cracking risk of concrete gravity dams to strong ground motions are
discussed.
1.3 Organization of the Book
This book is comprised of ten chapters. In this chapter, the research background
and objectives have been presented. An overview on the effects of strong earthquake
ground motions on concrete dams is provided.
Chapter 2 provides two fracture modelling approaches, i.e. XFEM and CDP
model, to describe the nonlinear dynamic response and seismic failure process of
concrete gravity dams under strong ground motions. The Lagrangian approach is
used for the finite element modeling of the dam-reservoir-foundation interaction
problem. The effects of different nonlinear approaches on the seismic response of
the dam are compared and discussed.
1.3 Organization of the Book 7
The seismic crack propagation of concrete gravity dams with initial cracks at the
upstream and downstream faces has rarely been studied during strong earthquakes.
Chapter 3 deals with the numerical prediction of crack propagation in concrete gravity
dams with single and multiple initial cracks. The crack progress of a scaled-down
1:40 model of a gravity dam with the initial notch in the upstream wall is simulated to
verify the validity of the calculation model. The effects of the initial crack position and
length on the crack propagation and seismic response of dam-reservoir-foundation
systems are studied.
Chapter 4 generalizes five potential failure modes of concrete gravity dams by
applying the incremental dynamic analysis method based on the XFEM. Considering
the uncertainty of the ground motion input, 40 as-recorded accelerograms with each
scaled to 8 increasing intensity levels are selected as seismic excitations. Based on
320 numerical simulation results, typical failure processes are presented.
Chapter 5 proposes local and global damage indices to quantitatively assess
the effects of strong motion duration on the accumulated damage of concrete
gravity dams. This chapter describes numerically the interdependency between single
component durations and structural accumulated damage indices. The definition of
single component durations that exhibit the strongest influence on structural damage
is determined. Then, Chap. 6 proposes a general integrated duration definition for
multi-component seismic excitations based on the existing concept of strong motion
duration. The relationship between integrated and single horizontal or vertical dura-
tions is investigated. A series of nonlinear dynamic analyses is performed to quantify
the effects of integrated duration and vertical seismic excitations on the nonlinear
response of concrete gravity dam-reservoir-foundation systems.
Chapter 7 assesses the effects of aftershocks on concrete gravity dam–reservoir–
foundation systems and provides a quantitative description of the damage demands
prior to and following the aftershocks. The correlation between ground motion char-
acteristics (i.e. frequency content, strong motion duration, and amplitude) of the
mainshocks and major aftershocks is discussed. Then, Chap. 8 examines the influence
of earthquake direction on seismic performance of concrete gravity dams subjected
to seismic sequences. Nonlinear dynamic analyses of the dam-reservoir-foundation
system subjected to as-recorded mainshock-aftershock sequences are conducted to
investigate the effects: (1) direction of single earthquake events; (2) direction of
seismic sequences; (3) aftershock polarity; and (4) earthquake intensity.
Chapter 9 deals with the characterization and modeling of near-fault and far-
fault ground motions. A systematic approach for seismic performance evaluation
and assessment of the probable level of damage based on the proposed performance
criteria is presented by using linear analysis results. To validate the presented seismic
performance evaluation method, nonlinear dynamic damage analyses of the concrete
dam under near-fault ground motions are conducted.
Owing to the presence of contraction joints, the above-mentioned analyses that
deal with the nonlinear dynamic response and seismic safety evaluation of concrete
gravity dams are based on the two-dimensional model, with the underlying assump-
tion that the monoliths behave independently during earthquake activities. Chapter 10
establishes a three-dimensional model to obtain the more realistic seismic response
8 1 Introduction
of concrete gravity dams. This chapter conducts a systematic study on the 3D seismic
damage-cracking behavior of concrete gravity dams with all the following factors
considered: contraction joint nonlinearity, cross-stream earthquake excitation, and
dam-foundation-reservoir interaction.
References
Ahmadi, M. T., Izadinia, M., & Bachmann, H. (2001). A discrete crack joint model for nonlinear
dynamic analysis of concrete arch dam. Computers & Structures, 79(4), 403–420.
Ardakani, S. M. S., & Saiidi, M. S. (2018). Simple method to estimate residual displacement
in concrete bridge columns under near-fault earthquake motions. Engineering Structures, 176,
208–219.
Ayari, M. L., & Saouma, V. E. (1990). A fracture mechanics based seismic analysis of concrete
gravity dams using discrete cracks. Engineering Fracture Mechanics, 35(1–3), 587–598.
Bhagat, S., Wijeyewickrema, A. C., & Subedi, N. (2018). Influence of near-fault ground motions
with fling-step and forward-directivity characteristics on seismic response of base-isolated
buildings. Journal of Earthquake Engineering 1–20.
Bhattacharjee, S. S., & Léger, P. (1994). Application of NLFM models to predict cracking in concrete
gravity dams. Journal of Structural Engineering, 120(4), 1255–1271.
Bray, J. D., & Rodriguez-Marek, A. (2004). Characterization of forward-directivity ground motions
in the near-fault region. Soil Dynamics and Earthquake Engineering, 24(11), 815–828.
Chopra, A. K., & Chakrabarti, P. (1973). The Koyna earthquake and the damage to Koyna Dam.
Bulletin of the Seismological Society of America, 63(2), 381.
Chopra, A. K., & Chintanapakdee, C. (2001). Comparing response of SDF systems to near-fault
and far-fault earthquake motions in the context of spectral regions. Earthquake Engineering and
Structural Dynamics, 30(12), 1769–1789.
Das, R., & Cleary, P. W. (2013). A mesh-free approach for fracture modelling of gravity dams under
earthquake. International Journal of Fracture, 179(1–2), 9–33.
Ghaemian, M., & Ghobarah, A. (1999). Nonlinear seismic response of concrete gravity dams with
dam–reservoir interaction. Engineering Structures, 21(4), 306–315.
Ghaemmaghami, A. R., & Ghaemian, M. (2010). Shaking table test on small-scale retrofitted
model of Sefid-rud concrete buttress dam. Earthquake Engineering and Structural Dynamics,
39, 109–118.
Goda, K. (2012). Nonlinear response potential of mainshock–aftershock sequences from Japanese
earthquakes. Bulletin of the Seismological Society of America, 102(5), 2139–2156.
Ingraffea, A. R., & Saouma, V. (1985). Numerical modeling of discrete crack propagation in
reinforced and plain concrete. In Fracture mechanics of concrete: Structural application and
numerical calculation (pp. 171–225). Netherlands: Springer.
Lee, J., & Fenves, G. L. (1998). A plastic-damage concrete model for earthquake analysis of dams.
Earthquake Engineering and Structural Dynamics, 27(9), 937–956.
Léger, P., & Leclerc, M. (1996). Evaluation of earthquake ground motions to predict cracking
response of gravity dams. Engineering Structures, 18(3), 227–239.
Li, Q., & Ellingwood, B. R. (2007). Performance evaluation and damage assessment of steel
frame buildings under main shock-aftershock earthquake sequences. Earthquake Engineering
and Structural Dynamics, 36(3), 405–427.
Li, Y., Song, R., & Van de Lindt, J. W. (2014). Collapse fragility of steel structures subjected
to earthquake mainshock-aftershock sequences. Journal of Structural Engineering, 140(12),
4014095.
Mavroeidis, G. P., & Papageorgiou, A. S. (2003). A mathematical representation of near-fault ground
motions. Bulletin of the Seismological Society of America, 93(3), 1099–1131.
References 9
Mirzayee, M., Khaji, N., & Ahmadi, M. T. (2011). A hybrid distinct element-boundary element
approach for seismic analysis of cracked concrete gravity dam-reservoir systems. Soil Dynamics
and Earthquake Engineering, 31(10), 1347–1356.
Nuss, L. K., Matsumoto, N., & Hansen, K. D. (2012). Shaken but not stirred-earthquake performance
of concrete dams (pp. 1511–1530). In USSD Proceedings.
Omidi, O., Valliappan, S., & Lotfi, V. (2013). Seismic cracking of concrete gravity dams by plastic–
damage model using different damping mechanisms. Finite Elements in Analysis and Design,
63, 80–97.
Omranian, E., Abdelnaby, A. E., & Abdollahzadeh, G. (2018). Seismic vulnerability assessment of
RC skew bridges subjected to mainshock-aftershock sequences. Soil Dynamics and Earthquake
Engineering, 114, 186–197.
Pekau, O. A., & Cui, Y. (2004). Failure analysis of fractured dams during earthquakes by DEM.
Engineering Structures, 26(10), 1483–1502.
Ruiz-García, J., & Negrete-Manriquez, J. C. (2011). Evaluation of drift demands in existing
steel frames under as-recorded far-field and near-fault mainshock-aftershock seismic sequences.
Salami, M. R., Kashani, M. M., & Goda, K. (2019). Influence of advanced structural modeling
technique, mainshock-aftershock sequences, and ground-motion types on seismic fragility of
low-rise RC structures. Soil Dynamics and Earthquake Engineering, 117, 263–279.
Shokrabadi, M., Burton, H. V., & Stewart, J. P. (2018). Impact of sequential ground motion pairing
on mainshock-aftershock structural response and collapse performance assessment. Journal of
Structural Engineering, 144(10), 4018177.
Shokrabadi, M., & Burton, H. V. (2018). Risk-based assessment of aftershock and mainshock-
aftershock seismic performance of reinforced concrete frames. Structural Safety, 73, 64–74.
Tsung-ho, H., Hsueh-hai, L., Tu-hsin, H., & Cheng-jung, Y. (1976). Strong-motion observation
of water-induced earthquakes at Hsinfengkiang reservoir in China. Engineering Geology, 10(2),
315–330.
Wang, G., Pekau, O. A., Zhang, C., & Wang, S. (2000). Seismic fracture analysis of concrete gravity
dams based on nonlinear fracture mechanics. Engineering Fracture Mechanics, 65(1), 67–87.
Wang, J., Lin, G., & Liu, J. (2006). Static and dynamic stability analysis using 3D-DDA with
incision body scheme. Earthquake Engineering and Engineering Vibration, 5(2), 273–283.
Xiang, N., & Alam, M. S. (2019). Displacement-based seismic design of bridge bents retrofitted
with various bracing devices and their seismic fragility assessment under near-fault and far-field
ground motions. Soil Dynamics and Earthquake Engineering, 119, 75–90.
Xin, L., Li, X., Zhang, Z., & Zhao, L. (2019). Seismic behavior of long-span concrete-filled steel
tubular arch bridge subjected to near-fault fling-step motions. Engineering Structures, 180, 148–
159.
Yaghmaei-Sabegh, S., Neekmanesh, S., & Ruiz-García, J. (2019). Evaluation of the coefficient
method for estimation of maximum roof displacement demand of existing buildings subjected to
near-fault ground motions. Soil Dynamics and Earthquake Engineering, 121, 276–280.
Yang, D., Guo, G., Liu, Y., & Zhang, J. (2019). Dimensional response analysis of bilinear
SDOF systems under near-fault ground motions with intrinsic length scale. Soil Dynamics and
Earthquake Engineering, 116, 397–408.
Zhong, H., Lin, G., Li, X., & Li, J. (2011). Seismic failure modeling of concrete dams considering
heterogeneity of concrete. Soil Dynamics and Earthquake Engineering, 31(12), 1678–1689.
Chapter 2
Comparative Analysis of Nonlinear
Seismic Response of Concrete Gravity
Dams Using XFEM and CDP Model
2.1 Introduction
The majority of high concrete dams are being built or to be built in countries with
active seismic activities. However, concrete dams will inevitably occur cracking
damage due to the low tensile resistance of concrete material under strong ground
motions. The seismic safety evaluation of concrete dams has received consider-
able attention since the outburst of strong earthquakes. Seismic failure analysis of
concrete dams is a complex problem associated with the highly nonlinear mechan-
ical behavior of concrete material under strong ground motions. The linear elastic
model in conjunction with criteria based on an allowable maximum tensile stress
is commonly utilized in the seismic design and safety evaluation of dams (Chopra
1988). Complex nonlinear constitutive models are required in order to capture the
concrete material’s behavior and assess the seismic safety of concrete dams in
earthquake-prone areas.
An investigation of the cracking mechanism and nonlinear dynamic response of
concrete dams under strong earthquakes is critically important for a rigorous seismic
safety evaluation, and nonlinear models simulating crack propagation within the dam
body need to be employed. However, there is no unified fracture modelling approach
to capture damage and failure in concrete dams. Different nonlinear models may
lead to some differences in cracking behavior of concrete dams. Several material
models have been introduced in order to model the nonlinear behavior of the concrete
material. The available literature includes models based on the theories of hypoelas-
ticity, hyperelasticity, plasticity, fracture mechanics, plastic-fracture, and continuum
damage (Cervera et al. 1995).
Different fracture methods with various material constitutive models will obtain
different nonlinear behaviors, which may significantly influence on the dam safety
evaluation. The present work will make use of both XFEM and CDP model to char-
acterize the mechanical behavior of concrete gravity dams under strong ground
motions. The seismic fracture response of concrete gravity dams is investigated
https://doi.org/10.1007/978-981-15-6194-8_2
12 2 Comparative Analysis of Nonlinear Seismic Response …
with considering the effects of dam-reservoir-foundation interaction. The Lagrangian

approach is used for the finite element modeling of dam-reservoir-foundation interac-
tion problem. For numerical application, seismic analyses of the Koyna gravity dam
are performed by using the 1967 Koyna earthquake records. How the two models
tackle the initiation and propagation of cracks in the dam body is investigated. The
crack propagation processes and failure modes of the dam structures are compared.
2.2 Method for Dynamic Failure Analysis of Concrete

Dams Under Strong Earthquake
Due to the low tensile strength of concrete materials, concrete dams are likely to
experience cracking. Hence, nonlinear dynamic analysis of concrete dams to earth-
quake loading will be inevitable. However, nonlinear dynamic response and failure
processes of concrete dams subjected to strong ground motions are very complex.
In the absence of monitoring data of the actual seismic crack processes of dams, the
dynamic model test and finite element method are important means to understand
the nonlinear dynamic behavior and damage pattern of high concrete dams under
strong ground motions. This section mainly reviewed the dynamic analysis methods
for concrete dams to strong earthquake loading. The model test, fracture mechanics,
and damage mechanics methods are introduced emphatically. The advantages and
disadvantages of these methods are discussed. At the same time, the latest dynamic
methods for simulating the damage processes of concrete dams to strong earthquakes
are reviewed.
2.2.1 Prototype Observation
The prototype observation is a very direct method to understand the seismic perfor-
mance of high dams to actual earthquakes. However, concrete dams have exhibited
extremely well seismic performance, even when subjected to strong ground motions
far in excess of their design. Hence, there are few concrete dams that have suffered
earthquake failure or major damages. This section will review some concrete dams
that have been reported earthquake-induced failure or major damages (Nuss et al.
2012), as summarized in Table 2.1.
2.2.2 Model Test Method
The dynamic model test is an important method to understand the nonlinear dynamic
response and seismic failure modes of the dam. The model test mainly includes the
Table 2.1 Damaged concrete dams under strong ground motions
Dam Country Height (m) Crest feet (m) Event Mag. Results
Koyna gravity dam India 103 853 Koyna 6.5 Cracks mainly appeared in the non-overflow
Dec 11, 1967 monoliths. 18 horizontal cracks developed on
the upstream face and 7 cracks developed on
the downstream face around the elevation at
which the slope of the downstream face
changes abruptly. Leakage was found in
some of these monoliths near the changes in
the slope of the downstream face, implying
the complete penetration from the upstream
face to the downstream face
Shih Kang gravity dam Taiwan, China 21.4 357 Chi-Chi 7.6 The fault rupture extended both upstream and
Sep 21, 1999 downstream of the dam and caused extensive
damage to bays 16–18. The ground
movement led to a vertical differential
movement of about 9 m in these bays (the left
part of these bays raised about 11, and the
right side raised about 2 m). There was also a
diagonal horizontal offset through the dam of
about 7 m, and the dam collapsed with
2.2 Method for Dynamic Failure Analysis of Concrete Dams …
uncontrolled release of water. This is the first

concrete dam, which has failed due to an
earthquake
Uh gravity dam Japan 14 34 Western Tottori 7.3 Although the Uh gravity dam was subjected
Oct 6, 2000 to strong ground motion, the only damage to
the dam was cracking 10–30 mm wide on the
spillway channel near the base of the
downstream face
(continued)
13
Table 2.1 (continued)
14
Dam Country Height (m) Crest feet (m) Event Mag. Results
Pacoima arch dam USA 113 180 Northridge 6.8 The main damage is that the contraction joint
Jan 17, 1994 between the arch dam and thrust block at the
left abutment opened 50 mm. The thrust
block and underlying rock mass may have
moved away from the dam about 13 mm
Hsinfengkiang buttress dam China 105 440 Reservoir 6.1 A horizontal crack 82 m long on the
Mar 19,1962 monoliths 13# to 17# developed on the right
side of the downstream face around the
elevation of 108 m at which the slope
changes abruptly. A few smaller cracks
developed on the left side of the dam at the
same elevation as the crack on the right side
Sefid Rud buttress dam Iran 106 417 Manjil 7.7 Horizontal cracks appeared in the upper parts
Jun 21, 1990 of the monoliths, especially in the highest
monolith. A major crack ran almost the
whole length of the dam at about 14 m below
the crest. Leakage was reported through
some of the cracks
2 Comparative Analysis of Nonlinear Seismic Response …
2.2 Method for Dynamic Failure Analysis of Concrete Dams … 15
centrifuge test and shaking table test. With the extensive construction of high DAMS
in the world, the model test technology for the dam has been developed rapidly. Many
researchers have investigated the nonlinear dynamic response behavior and cracking
failure processes of dams under strong earthquakes by using model tests.
Chen et al. National Research Council (Dams, 1990a) employed the shaking table
test to investigate the seismic failure process of the Koyna gravity dam under strong
ground motions. Pekau et al. (1995) investigated the cracking process of the Koyna
dam under simplified loading conditions on a shaking table. Tinawi et al. (2000)
presented shake table tests to study the dynamic cracking and sliding response of
concrete gravity dams, the experimental results have been compared with a simula-
tion. Zhou et al. (2000) performed a series of dynamic experiments of high arch dams
to investigate the seismic response and failure modes. Morin et al. (2002) conducted
shake table tests to study the seismic behavior and the applicability of the joint model
on a 3.4 m high post-tensioned gravity dam. Li et al. (2005) investigated the seismic
response of the powerhouse monolith of the Three Gorges dam through the model
test on a shaking table. Wang and Li (2006) carried out a seismic overloading model
test to study the dynamic behavior of an arch dam with the height of 278 m in proto-
type. Ghaemmaghami and Ghaemian (2008, 2010) conducted a shaking table test
to study the seismic failure process of the Sefid-Rud concrete dam under the strong
earthquake of M 7.6, where the scale of the model is 1:30. Rochon-Cyr and Léger
(2009) performed a series of shaking table sliding tests to investigate the dynamic
sliding response of a 1.5 m high concrete gravity dam model including water uplift
pressure. Zhong et al. (2011) investigated the dynamic response and failure modes
of the Dagangshan arch dam based on the model test. Chen et al. (2013) studied
the dynamic failure process of the Ahai concrete gravity dam used the shaking table
model test, where the interaction between the reservoir water and dam is consid-
ered, and the foundation is assumed to be rigid. Resatalab et al. (2013) conducted
an experimental study on seismic response behavior of concrete gravity dams on
the shaking table with considering the interaction between the dam-foundation and
reservoir. Aldemir et al. (2015) conducted a pseudo-dynamic testing to investigate
the seismic cracking of a concrete gravity dam with a 1/75 scaled. Roşca (2008)
used the physical model to study the dynamic behavior and failure mechanism of
concrete dams to strong ground motions. He thought that the simulation of boundary
conditions is one major problem, and the simulation of material properties seems
to be the most difficult and important for the small-scale modeling. Phansri et al.
(2010) conducted two small-scale model tests on the shaking table to examine the
seismic-induced damage of a concrete gravity dam. They simulated the two shaking
table tests to study the crack/behavior of the two small-scale dams using the plastic-
damage model. Wang et al. (2014) conducted a comparative model experiment test
on the shaking table to study the seismic failure process of a concrete gravity dam
with and without the reinforcement measure, and evaluate the effectiveness of the
strengthening measure. Wang et al. (2019) conducted a dynamic fracture test for
a small-scale model on the shaking table to investigate the dynamic characteristics
and failure modes of a concrete gravity dam with the prototype height of 278 m.
In their test, the Fiber Bragg Grating (FBG) strain sensor is employed to obtain the
dynamic strain and residual strain. Rodríguez et al. (2020) compared the seismic
performance of the Koyna dam using both the numerical method and shake table
testing. They concluded that the scale model testing can provide invaluable insights
into the nonlinear dynamic response and failure mechanisms of concrete gravity
dams, which is even more important in terms of climate change to support capital
planning and improve dam safety throughout the world.
Although the dynamic model test can capture the dynamic response and failure
process of concrete dams, the investment of the model test is large and the period
is very long. Many issues need to be clarified. It is difficult to satisfy the similar
rate requirements of the mechanical properties of materials, loads and boundary
conditions at the same time. And there are still some technical problems, such as the
method of overloading, similar criterion of failure test, and the change of material
properties in the strength reserve method.
2.2.3 Numerical Simulation Method
With the development of computer technology and numerical analysis, the applica-
tion of nonlinear finite element methods provides a new way to study the dynamic
response and failure processes of concrete dams under earthquake loads. The crack
plays an important role in the dynamic response behavior of concrete dams. How
to simulate crack propagation becomes an important problem. Over the past several
decades, extensive research has been carried out and many crack models have been
developed to simulate the nonlinear seismic response of concrete gravity dams.
There are two traditional fracture models, namely, discrete crack approach
(Hillerborg et al. 1976) and smeared crack approach (Rashid 1968). In the discrete
crack model, fracture mechanics concepts are used to model the discrete cracking.
The discontinuous displacement field at a crack is accounted for by introducing a
discontinuity interface into the solid and describing its behavior by a discrete traction-
separation law (Lagier et al. 2011). In a finite element mesh, discontinuity interfaces
are placed at element boundaries. Hence, they need remeshing algorithms to accom-
modate crack propagation (Ingraffea and Saouma 1985). This approach would be
efficiently applied to problems where only a few well-defined fractures are encoun-
tered. There are two methods that can be used in the discrete crack model, i.e., the
linear elastic fracture mechanics (LEFM) and nonlinear fracture mechanics (NLFM).
When the applied loads are very slow and are also the impulsive loads, the frac-
ture behavior of concrete structures seems to be adequately predicted by the LEFM
model. However, in the intermediate range, from short-term static loading to seismic-
induced strain rates, the NLFEM model appears to be more appropriate to describe
the fracture process (Bhattacharjee and Leger 1993). Ingraffea (1990) employed the
mixed-mode LEFM implemented within a discrete crack method to elucidate the
crack initiation and trajectory of the Fontana dam. Ayari and Saouma (1990) devel-
oped an LEFM criterion with the discrete crack model for seismic analysis of concrete
gravity dams. Bhattacharjee and Léger (1994) investigated the fracture propagation
process of concrete gravity dams with initial cracks by using the discrete crack model
based on the nonlinear fracture mechanics (NLFM). Ahmadi et al. (2001) presented a
discrete crack joint model to investigate the nonlinear dynamic response of concrete
arch dams. Shi et al. (2003) developed an extended fictitious crack model to analyze
the multiple discrete cracking behavior in concrete dams. The original single-crack
and multiple-crack problems are involved in the numerical modeling. Lohrasbi and
Attarnejad (2008) investigated the crack propagation process of concrete gravity
dams using the discrete crack method, and concluded that the discrete crack method
can demonstrate the real crack and its opening, but it required an expense of time and
money because of frequent meshing. Shi et al. (2014) presented a two-step approach
for discrete crack analysis of concrete gravity dams to earthquake force.
In the smeared crack model which is based on changes in the constitutive laws
governing the behavior of concrete, cracking is modeled by modifying the strength
and stiffness of concrete and by distributing or “smearing” the dissipated energy
along the finite width of the localization band (Theiner and Hofstetter 2009; Markovič
et al. 2013). The structural integrity is evaluated on a local basis. El-Aidi and Hall
(1989a, b) discussed the nonlinear dynamic response and seismic performance of
concrete dams based the smeared crack approach. Bhattacharjee and Leger (1993)
employed a nonlinear smeared fracture model to predict the failure process and
energy response of Koyna gravity dams under strong ground motions. Ghaemian and
Ghobarah (1999) used the smeared crack model to investigate the seismic cracking
and response of concrete gravity dams. In their model, the staggered solution method
is presented to consider the dam-reservoir interaction, and the crack propagation
criterion is based on the nonlinear fracture mechanics. Lotfi and Espandar (2004)
combined the discrete crack and non-orthogonal smeared crack technique to inves-
tigate the nonlinear seismic behavior of concrete arch dams. They found that the
developed discrete crack and non-orthogonal smeared crack is a more rigorous,
consistent and realistic approach. Calayir and Karaton (2005a) presented a co-axial
rotating crack model (CRCM) to discuss the nonlinear fracture process of concrete
gravity dams subjected to strong earthquakes with considering the effects of dam-
reservoir interaction. Mirzabozorg and Ghaemian (2005) proposed a smeared crack
approach to model the three-dimensional dynamic behavior of the Koyna gravity
dams and Morrow Point arch dams. Hariri-Ardebili et al. (2013) proposed a coaxial
rotating smeared crack model to study the seismic failure behavior of the Koyna
gravity dam considering the fracture energy effects and the dam-reservoir and dam-
foundation interactions. Zhang et al. (2014) employed the smeared crack model to
present the cracking characteristics of concrete gravity dams with longitudinal joints.
Hariri-Ardebili and Seyed-Kolbadi (2015) assessed the seismic cracking of the Koyna
gravity dam, Sefidrud buttress dam, and Dez arch dam using an improved 3D co-axial
rotating smeared crack model. Their simulation results are reasonably matched with
experimental tests. Hariri-Ardebili et al. (2016) investigated FEM-based parametric
analysis of a typical concrete gravity dam. In their finite element model, a devel-
oped rotating smeared crack model is used to obtain the nonlinear dynamic response
of the dam. Kong et al. (2017) used the co-axial rotating smeared crack model to
study the seismic cracking behavior of the conventional reinforced concrete face
slab and ductile fiber-reinforced cement-based composite face slab for concrete-
faced rockfill dams. Moradloo et al. (2018) employed a modified three-dimensional
rotating smeared crack model to obtain the nonlinear behavior of concrete arch dams
and evaluate the seismic fragility of dams through nonlinear incremental analysis.
Alijani-Ardeshir et al. (2019) discussed the effects of three smeared crack models
on seismic response of the Pine Flat dam, the three models are the multidirectional
fixed crack model (MFCM), coaxial rotating crack model (CRCM), and orthogonal
multi fixed crack model (OMFCM). Pirooznia (2019) investigated the isolation layer
effects on the seismic improvement of concrete gravity dams considering the dam-
reservoir interaction based on the smeared crack method. Their results revealed that
the isolation layer can reduce the earthquake response and crack propagation process.
Under the traditional finite element framework, the description of discontinuous
displacement field is usually realized by coordinating the element boundary with
discontinuous interface and setting double nodes in corresponding positions. The
simulation of evolving discontinuous interface requires continuous remeshing. This
will result in a large amount of pretreatment workload, which is not conducive to
engineering applications. The extended finite element method (XFEM) (Belytschko
and Black 1999; Moës et al. 1999; Stazi et al. 2003; Belytschko et al. 2009), which
is based on the cohesive segments method (Remmers et al. 2008) in conjunction
with the phantom node technique (Hansbo and Hansbo 2004; Song et al. 2006),
can be used to simulate crack initiation and propagation along an arbitrary path,
since the crack propagation is not tied to the element boundaries in a mesh. With this
approach, it is not necessary to define the crack tip position, but it is possible to simply
define a reference region in which the crack will propagate. The near-tip asymptotic
singularity is not needed, and only the displacement jump across a cracked element
is considered. Hence, the crack must propagate across an entire element at a time to
avoid the need to model the stress singularity. The phantom node method describes
discontinuity by superposing the phantom elements instead of introducing additional
degrees of freedom, and it is easy to incorporate into conventional finite element
codes. The XFEM is applied to dynamic problems. Pan et al. (2014) compared the
seismic failure process of concrete gravity dams using different fracture approaches
including the XFEM with a cohesive law, the crack band finite element method with
a plastic-damage relation, and the Drucker-Prager elasto-plastic model. Wang et al.
(2015) investigated the potential failure modes of the Guandi gravity dam based
on the XFEM. They obtained the typical failure processes and five potential failure
modes of concrete gravity dams through incremental dynamic analysis. Zhang et al.
(2013) presented the XFEM to predict the seismic crack propagation process of
concrete gravity dams with single and multiple initial cracks. Pirboudaghi et al.
(2018) proposed an extended finite element-wavelet transform coupled procedure
for seismic damage detection based system identification of concrete dams. The
advantage of the structural health monitoring used the XFEM is that the whole dam
structure is potentially under damage risk without predefined damage. Huang (2018)
simulated the seismic crack propagation of concrete gravity dams using the XFEM. In
his numerical framework, an incrementation-adaptive multi-transmitting boundary
is employed to model the seismic wave propagation in the reservoir and foundation.
Ma et al.(2019) investigated the seismic dynamic response and crack propagation

process of the Koyna gravity dam with different initial crack lengths based on the
XFEM.
Concrete is a kind of quasi-brittle material, which behaves as linear elastic
behavior under the action of the small ground motion. However, with the increase of
ground motion load, the micro-cracks in concrete will gradually develop into macro-
scopic cracks. The nonlinear dynamic behavior analysis of concrete dams based
on the plastic damage mechanics has a sufficient theoretical basis. The damage
mechanics method has become an important means to reflect the nonlinearity of
concrete materials. In order to describe the complex mechanical behavior of concrete
material under earthquake conditions, a number of damage constitutive models have
been developed, including the isotropic damage models (Lubliner et al. 1989; Mazars
and Pijaudier-Cabot 1989; Lee and Fenves 1998a, b), anisotropic damage models
(Dragon and Mroz 1979; Murakami and Ohno 1981), and the damage models
(Cervera et al. 1995, 1996; Yazdchi et al. 1999) for concrete gravity dams under
seismic loads. The concrete damaged plasticity (CDP) model which is developed by
Lubliner et al. (1989) and modified by Lee and Fenves (1998a, b) offers a particu-
larly interesting context where damage evolution can be simulated. This approach
describing the nonlinear behavior of each compounding substance of a multiphase
composite material is used for seismic cracking analysis of concrete dams. In this
model, the uniaxial strength functions are factorized into two parts to represent the
permanent (plastic) deformation and degradation of stiffness (degradation damage).
It assumes that there are two main failure mechanisms of the concrete material, one
for tensile cracking, and the other for compressive crushing. Lee and Fenves (1998a,
b) used the modified plastic-damage constitutive model to evaluate the nonlinear
seismic response of the Koyna dam under the 1967 Koyna earthquake. The rate-
independent and strain softening effects have been considered. Sarkar et al. (2007)
investigated the influence of the reservoir height and foundation modulus on the
nonlinear seismic response of the Koyna gravity dam through the concrete damaged
plasticity model. Omidi et al. (2013) employed the plastic-damage model to study
the damping mechanism effects on the seismic cracking response of concrete gravity
dams including the dam-reservoir interaction. Zhang and Wang (2013) discussed
the effects of near-fault ground motions on the nonlinear dynamic response and
accumulated damage of concrete gravity dams using the concrete damage plasticity
model. Ghaedi et al. (2015) studied the size and shape effects of the gallery on
seismic nonlinear dynamic response of roller compacted concrete gravity dams based
on the concrete damaged plasticity model. Chen et al. (2016) presented a failure
analysis approach for concrete arch dams based on the elastic strain energy crite-
rion using the orthotropic damage mode. Yazdani and Alembagheri (2017) inves-
tigated the nonlinear seismic response of gravity dams under near-fault ground
motions based on the plastic-damage method. Omidi and Lotfi (2017) presented
a combined discrete crack and plastic-damage (DC-PD) method to investigate the
seismic nonlinear response of concrete arch dams. In their technique, the discrete
crack approach is used to model the joints, and the plastic-damage model is employed
to study the damage process of arch dams. Khazaei Poul and Zerva (2018) conducted
nonlinear time-domain dynamic analyses to evaluate the input motion mechanism

effect on concrete gravity dams based on the concrete damage plasticity model.
Chen et al. (2019) used the concrete damage plasticity model to investigate the
seismic performance and failure modes of the Jin’anqiao concrete gravity dam.
In their model, the viscous-spring boundary is employed to model the interaction
between the dam and foundation. Zhao et al. (2019) investigated the seismic damage
of concrete gravity dams using the concrete damaged plasticity constitutive model.
The earthquake damage has been quantitatively assessed using electro-mechanical
impedance measurements. Daneshyar and Ghaemian (2019) developed a new rate-
dependent anisotropic damage-plastic model to model the seismic nonlinear response
of concrete arch dams with the coupled adhesive-frictional joint response. Pang et al.
(2020) used the plastic-damage model to describe the seismic nonlinearity of concrete
face slabs, and investigate the seismic fragility of concrete face rockfill dam. Lu et al.
(2019) discussed the spatial variability effects of strength parameters on the seismic
response characteristics of concrete gravity dams based on the concrete damage
plasticity model.
In addition, other models, such as the rock failure process analysis (RFPA),
distinct element method (DEM), discontinuous deformation analysis (DDA), mesh-
free particle method called smoothed particle hydrodynamics (SPH) and hybrid
distinct element-boundary element (DE-BE) et al., are also methods for simula-
tion of concrete dams and their behavior. Zhong et al. (2011)employed the rock
failure process analysis (RFPA) proposed by Tang and Kou (1998) to investigate
the typical failure process and failure modes of concrete gravity dams considering
the uncertainty in ground motion input and concrete material. Pekau and Cui (2004)
conducted a comprehensive study on the nonlinear dynamic response behavior of the
fractured Koyna dam subjected to strong ground motions using the distinct element
method (DEM). Bretas et al. (2014) also used the DEM to evaluate the seismic
safety of masonry gravity dams. Wang et al. (2006) used the discontinuous defor-
mation analysis (DDA) to investigate the dynamic stability of a fractured concrete
gravity dam under seismic ground motions. Das and Cleary (2013) explored a mesh-
free particle method called smoothed particle hydrodynamics (SPH) to model the
seismic failure process of concrete gravity dams subjected strong ground motions.
Mirzayee et al. (2011) proposed a new algorithm called the hybrid distinct element-
boundary element (DE-BE) to study the nonlinear dynamic behavior of cracked
concrete gravity dam-reservoir systems. Pan et al (2011) employed the Drucker–
Prager (DP) elasto-plastic model to discuss the seismic crack process of concrete
dam and compare with the extended finite element method (XFEM) and crack band
finite element method. Chen et al. (2017) proposed a 3D nonlinear polyhedron scaled
boundary finite element (NPSBFEM3D) for elasto-plastic analysis with the advan-
tages of FEM, BEM and polyhedrons. The authors have successfully carried out
seismic cracking analysis of concrete gravity dams with this method.
2.3 eXtended Finite Element Method (XFEM) 21
2.3 eXtended Finite Element Method (XFEM)
In the standard FEM, cracks are required to follow element edges for the discontinuity
modeling. In contrast, the crack geometry in the extended finite element method
(XFEM) need not be aligned with the element edges and thus is independent of
the background mesh. Such an independence is made by enriching the standard
displacement-based finite element approximation with some pre-knowledge of the
physics of crack. It is based on the partition of unity method (PUM) introduced by
Melenk and Babuška (1996), which allows local enrichment functions to be easily
incorporated into a finite element approximation. This enrichment function typically
consists of some near-tip asymptotic functions that capture the singularity around
the crack tip and a discontinuous function that represents the jump in displacement
across the crack surface. By implementing the generalized Heaviside function (Moës
et al. 1999), the method was further enhanced, avoiding taking into account the
complicated mapping for arbitrary curved cracks.
2.3.1 XFEM Approximation
The XFEM enriches a standard displacement based finite element approximation

with discontinuous functions. The approximation for a displacement vector function
u with the partition of unity enrichment (Fig. 2.1) in the XFEM takes the following
form (Moës et al. 1999):
Fig. 2.1 Enriched nodes in the XFEM


uxfem (x) = ui Ni (x)
i∈I

+ b j N j (x)H (x)
j∈J

only Heaviside nodes
4 4

+ Nk (x) ckl1 Fl1 (x) + Nk (x) ckl2 Fl2 (x) (2.1)
k∈K 1 l=1 k∈K 2 l=1

only crack - tip nodes
where x = {x, y} is the two-dimensional coordinate system, I is the set of all nodes
in the mesh, N i (x) is the shape function associated with node i, ui are the classical
degrees of freedom for node i. J ⊂ I is the set of nodes whose shape function
support is cut by a crack, bj is the vector of corresponding additional degrees of
freedom for modeling crack faces (not crack-tips). If the crack is aligned with the
mesh, bj represent the opening of the crack, H(x) is the Heaviside function. K 1
⊂ I and K 2 ⊂ I are the set of nodes whose shape function support contains the
first and second crack tips in their influence domain, respectively. ckl1 and ckl2 are
the vector of corresponding additional degrees of freedom which are related to the
modeling of crack-tips, as the near-tip regions are enriched with four different crack
functions. Fl1 (x) and Fl2 (x) are the asymptotic crack-tip enrichment functions. If
there is no enrichment, then the above equation reduces to the classical finite element
approximation ufem (x) = i ui Ni (x).
The first term on the right-hand side of the above equation (Eq. (2.1)) is applied
to all the nodes while the second term is only valid for nodes whose shape function
support is cut by the crack interior, and the third (fourth) term is used only for nodes
whose shape function support is cut by the crack tip.
2.3.2 Enrichment Functions
To model the discontinuity in displacement field, the enrichment function H(x) which
we refer to as a generalized Heaviside enrichment function is implemented in the
simulation of powder-die contact surface. The function H(x) takes the value of +1
above the crack, and −1 below the crack. The function H(x) is given by

1 if (x − x∗ ) · n ≥ 0
H (x) = (2.2)
−1 otherwise
where x is a sample gauss point, x* (lies on the crack) is the closest point to x
(Fig. 2.2), and n is the unit outward normal to the crack at x*. Figure 2.2 illustrates
the discontinuous jump function across the crack surface.
Fig. 2.2 Representation of normal and tangential coordinates for a smooth crack
In order to model the crack-tip and also to improve the representation of crack-
tip fields, crack-tip enrichment functions are used in the element which contains the
crack tip. For an isotropic material, the crack-tip enrichment functions F l (r, θ ) which
are also shown in Fig. 2.2 are given as

√ θ √ θ √ θ √ θ
{Fl (r, θ )}l=1
4
= r sin , r cos , r sin θ sin , r sin θ cos (2.3)
2 2 2 2
where (r, θ ) is the local polar coordinate system with its origin at the crack tip,
√ θ = 0 is tangent to the crack at the tip. Note that the first function in Eq. (2.3)
and
r sin θ 2, is discontinuous across the crack plane, whereas the last three functions
are continuous. It bears emphasis that the near-tip discontinuity can be represented
with other sets of functions, or even a single function which is discontinuous across
the crack tip geometry. Multiple cracks can be treated in the above framework, by
incorporating additional discontinuous and near-tip enrichment.
The cohesive segments method (Remmers et al. 2008) in conjunction with
phantom nodes proposed by Song et al. (2006) has been used in the framework of
the XFEM to simulate the crack initiation and propagation along arbitrary, solution-
dependent paths in the bulk materials for brittle or ductile fracture. The cohesive
segments are not restricted to being located along element boundaries, but can be
located at arbitrary locations and in arbitrary directions, allowing for the resolu-
tion of complex crack patterns. In this case the near-tip asymptotic singularity is not
needed, and only the displacement jump across a cracked element, which is described
using the phantom node method, is considered. Therefore, the crack has to propagate
across an entire element at a time to avoid the need to model the stress singularity.
Phantom nodes, which are superposed on the original real nodes, are utilized to
represent the discontinuity of the cracked elements, as illustrated in Fig. 2.3 (Song
et al. 2006). Propagation of a crack along an arbitrary path is made possible by the
use of phantom nodes that initially have exactly the same coordinates as the real
nodes and that are completely constrained to the real nodes up to damage initiation.
In an uncrack element, each phantom node is completely constrained to their corre-
sponding real nodes (n1 to n4 ). But when crossed by a crack at c , the element is
partitioned into two subdomains, Ω A and Ω B . The discontinuity in the displacement
Fig. 2.3 Representation of cracked elements by implementation of the phantom node method
is made possible by adding phantom nodes (n 1 to n 4 ) superimposed to the original

nodes. The existing element is replaced by two sub-elements, referred to as element
A and element B. Each sub-element is formed by a combination of some real nodes
(the ones corresponding to the cracked part) and phantom nodes (the ones corre-
sponding to the respective part of the original element). The two sub-elements are
constituted by the nodes n 1 , n 2 , n3 and n4 (nA ) and n1 , n2 , n 3 , and n 4 (nB ). Each
phantom node and its corresponding real node are no longer tied together and can
move apart. Both elements are only partially active, the active part of element A is
Ω A and the active part of element B is Ω B . This is represented numerically in the
definition of the displacement field: the displacement of a point with coordinates x
is computed by

uA (x, t) = uAj (t)N j (x), x ∈ ΩA
(2.4)
uB (x, t) = uBj (t)N j (x), x ∈ ΩB
The approximation of the displacement field is then given by:

u(x, t) = uAj (t)N j (x) H (− f (x)) + uBj (t)N j (x) H ( f (x)) (2.5)
j∈n A
j∈n B

uA (x,t) uB (x,t)
where nA and nB are the index sets of the nodes of superposed element A and element
B, respectively; f (x) is the signed distance measured from the crack.
The crack normal opening δ n and the tangential sliding δ t are shown in the
following equation:

δn = n[u]
(2.6)
δt = [u] − nδn
where [u] is the jump in the displacement given as

[u] = N j uAj − N j uBj (2.7)
j∈n A j∈n B
In order to control the magnitude of separation, the cohesive law is defined. A

separation occurs when the cohesive strength of the cracked element is zero, after
which the phantom and the real nodes move independently. This method which
provides an effective and attractive engineering approach, has been used to simulate
the initiation and growth of multiple cracks in solids by Remmers et al. (2008) and
Song et al. (2006). The detail of the flowchart for crack propagation simulation can
be found in references (Ye et al. 2012).
In the XFEM, the mesh is not required to conform to the geometric discontinuities.
Two signed distance functions per node are generally required to describe the crack
location, including the location of crack tips, in a cracked geometry. The first signed
distance function describes the crack surface, while the second is used to construct an
orthogonal surface so that the intersection of the two surfaces gives the crack front.
The first signed distance function is assigned only to nodes of elements intersected
by the crack, while the second is assigned only to nodes of elements containing the
crack tips.
2.3.3 Discrete Equations
The proposed approach is based on an elastodynamics behavior for XFEM analysis

of the two-dimensional dam section subjected to an earthquake. We write the strong
form of the momentum conservation law in terms of the Cauchy stress tensor, for
the current configuration described in Fig. 2.4, as follows:
Fig. 2.4 Notations used for

a two-dimensional domain
∂σi j
+ ρbi − ρ üi = 0 ∈ Ω (2.8)
∂x j
where σ is the Cauchy stress tensor, ρ is the initial mass density, b is the body force
vector per unit mass, ü is the acceleration.
The boundary conditions are
σi j n j = t̄i ∈ Γt (2.9)
ui = ūi ∈ Γu (2.10)
where n is the external unit vector to Γ , t̄ is the applied traction force vector on
the Neumann boundary Γt , ū is the applied displacement vector on the Dirichlet
boundary Γu . It can be noted that Γu ∩ Γt = Γ and u ∪ Γt = ∅.
The weak form of the momentum equation for dynamic problems in the current
configuration is given by

ρ üδudΩ + Cu̇δudΩ + σδεdΩ + c
δ[u]dΓc = ρbδudΩ + t̄δudΓt
Ω Ω Ω Γc Ω Γt
(2.11)
where t̄ is the normalized traction prescribed on Γt ; τc is the cohesive traction applied

on the discontinuity surface; u̇ is the velocity; C is the damping matrix. In the XFEM,
approximation (2.1) is utilized to calculate the displacement uh (x) for a typical point x
in the dam section. Using the standard Bubnov-Galerkin procedure, the equilibrium
discrete system of equations for a two-dimensional dam section subjected to an
earthquake with XFEM obtained from Eq. (2.11) can be written as
Müh + Cu̇h + Kuh = f (2.12)
where M is the mass matrix; C is the damping matrix; K is the stiffness matrix;
f is the external load vector; uh = {u, a, c}T is the vector of nodal parameters,
including displacements u, Heaviside and crack tip enrichment degrees of freedom
a and c, respectively. Rayleigh damping assumption is used for material damping;
thus viscous damping matrix is defined as
C = αM + βK (2.13)
where α and β are the constant factors determined by the given free-vibration
frequencies of the dam structure to damping ratio.
These matrices for an element e are defined as
⎡ ⎤
Miuuj Miuaj Miucj
⎢ ⎥
Miej = ⎣ Miauj Miaaj Miacj ⎦ (2.14)
Micuj Micaj Miccj
⎡ ⎤
Kiuuj Kiuaj Kiucj
⎢ au ⎥
Ki j = ⎣ Ki j
e
Kiaaj Kiacj ⎦ (2.15)
Kicuj Kicaj Kiccj
T
fie = fiu , fia , fic (2.16)
The components of consistent mass matrix Mirsj (r, s = u, a, c), stiffness matrix
Ki j (r, s = u, a, c) and force vector fir (r = u, a, c) include all parts of the classical
rs
FEM(uu ), Heaviside enrichment(aa ), orthotropic crack tip enrichment(cc ) and the

coupled parts of XFEM approximation(ua, uc, ac ) are given by:

Mirsj = ρζir ζ js dΩ (r, s = u, a, c) (2.17)
Ω

Kirsj = (Bir )T D(Bsj )dΩ (r, s = u, a, c) (2.18)
Ωe

fir = ψir tdΓ + ψir bdΩ (r = u, a, c) (2.19)
Γt Ωe
Where
⎧
⎨ Ni r = u
ζir = N H r = a (α = 1, 2, 3and4) (2.20)
⎩ i
Ni Fαi r = c
⎧
⎨ Ni r = u
ψir = Ni H r = a (α = 1, 2, 3 and 4) (2.21)
⎩
Ni Fα r = c
and B = ∇ψ is the matrix of derivatives of extended shape functions ψir .
2.4 Concrete Damaged Plasticity (CDP) Model
In order to describe the complex mechanical behavior of concrete material under

earthquake conditions, a number of constitutive models have been developed,
including the isotropic damage models (Lubliner et al. 1989; Mazars and Pijaudier-
Cabot 1989; Lee and Fenves 1998a, b), the anisotropic damage models (Dragon
and Mroz 1979; Murakami and Ohno 1981), and the damage models (Cervera et al.
1995, 1996; Yazdchi et al. 1999) for concrete gravity dams under seismic loads. In
this section, a basic constitutive model developed by Lubliner et al. (1989) and modi-
fied by Lee and Fenves (1998a, b) is presented. The model describing the nonlinear
behavior of each compounding substance of a multiphase composite material is
commonly used for seismic cracking analysis of concrete dams. In this model, the
uniaxial strength functions are factorized into two parts to represent the permanent
(plastic) deformation and degradation of stiffness (degradation damage). It assumes
that there are two main failure mechanisms of the concrete material, one for tensile
cracking, and the other for compressive crushing.
2.4.1 Damage Evolution
In the incremental theory of plasticity, the total strain tensor, ε, is decomposed into
the elastic part, εe , and the plastic part, εp , which for linear elasticity is given by:
ε = εe + ε p (2.22)
The state of the nonlinear local problem with variables {εe , εp , κ} is assumed to
be known at time t. With this information, the stress tensor is given by
σ = (1 − d)σ̄ = (1 − d)E 0 (ε − ε p ) and d = d(κ) (2.23)
where E 0 is the initial (undamaged) elastic stiffness of the material, d is the scalar
stiffness degradation variable, which can take values in the range from 0 (undamaged
material) to 1 (fully damaged material). The damage associated with the failure
mechanisms of the concrete (cracking and crushing) therefore results in a reduction
in the elastic stiffness, which is assumed to be a function of a set of the internal
variable κ consisting of tensile and compressive damage variables, i.e. κ = {κ t ,
κ c }. Damage functions in tension d t and in compression d c are nonlinear functions
calculated by comparison of the uniaxial response with experimental data. Following
the usual notions of continuum damage mechanics, the effective stress σ̄ is defined
as
σ
σ̄ = = E 0 (ε − ε p ) (2.24)
1−d
Similarly, the first effective stress invariant I¯1 and the second effective deviatoric
stress invariant J¯2 are defined in terms of the effective stress tensor.
I¯1 = σ̄ii (2.25)

2.4 Concrete Damaged Plasticity (CDP) Model 29
1
J¯2 = S̄i j S̄i j (2.26)
2
where S̄i j is the effective deviatoric stress tensor.

The stress-cracking strain curves for uniaxial tension and the stress-crushing strain
in uniaxial compression are needed to define elastic, plastic and damage behaviors,
as shown Fig. 2.5. The stress-strain relations under uniaxial tension and compression
loading are
p
σt = (1 − dt )E 0 (εt − εt ) (2.27)
σc = (1 − dc )E 0 (εc − εcp ) (2.28)
Fig. 2.5 Response of concrete under a uniaxial tension and b uniaxial compression
2.4.2 Yield Criterion
The yield function proposed by Lubliner et al. (1989) and modified by Lee and
Fenves (1998) is adopted. The Concrete Damaged Plasticity (CDP) model uses
the yield condition to account for different evolution of strength under tension and
compression. In terms of effective stresses, the yield function takes the following
form
1
F= q̄ − 3α p̄ + β ε̃ p σ̄ˆ max − γ −σ̄ˆ max − σ̄c ε̃cp ≤ 0 (2.29)
1−α
with

σbo σc0 − 1
α= 0 ≤ α ≤ 0.5, (2.30)
2σbo σc0 − 1
p
σ̄c ε̃c
β = p (1 − α) − (1 + α), (2.31)
σ̄t ε̃t
3(1 − K c )
γ = , (2.32)
2K c − 1
σ̄
p̄ = − : I, (2.33)
3
!
3
q̄ = S̄ : S̄ (2.34)
2
where α and β are dimensionless material constants, σ̄ˆ max is the algebraically
maximum eigenvalue of σ̄ , σb0 is the concrete strength under equal biaxial compres-
sion, σc0 is the initial compressive yield stress, σ̄c and σ̄t are the effective compressive
p p
and tensile cohesion stresses respectively, ε̃c and ε̃t are the equivalent compressive
and tensile plastic strains respectively, K c is the strength ratio of concrete under equal
biaxial compression to triaxial compression, p̄ is the effective hydrostatic pressure,
q̄ is the Mises equivalent effective stress, and S̄ is the deviatoric part of the effective
stress tensor σ̄ .
Typical yield surfaces in the deviatoric plane are shown in Figs. 2.6, and 2.7 shows
the initial shape of the yield surface in the principal plane stress space.
2.4.3 Flow Rule
The plastic strain rate is evaluated by the flow rule, which is defined by a scalar plastic
potential function, G. During plasticity, the normality plastic flow rule is applied as
2.4 Concrete Damaged Plasticity (CDP) Model 31
Fig. 2.6 Yield surfaces in the deviatoric plane
Fig. 2.7 Initial yield function in plane stress space
∂G(σ̄ )
ε̇ p = λ̇ (2.35)
∂ σ̄
where λ̇ is a non-negative function referred to as the plastic consistency parameter.

A Drucker-Prager hyperbolic function is used as the plastic potential function
"
G= ( σt0 tan ψ)2 + q̄ 2 − p̄ tan ψ (2.36)
where ψ is the dilation angle measured in the p–q plane at high confining pressure;
σt0 is the uniaxial tensile stress at failure; and is a parameter, referred to as the
eccentricity, that defines the rate at which the function approaches the asymptote
defined by ψ (the flow potential tends to a straight line as the eccentricity tends to
zero). This flow potential, which is continuous and smooth, ensures that the flow
direction is defined uniquely.
2.5 Lagrangian Formulation for Dynamic Interaction

of Dam-Reservoir-Foundation Systems
In order to consider the reservoir effect on the behavior of the dam under strong
ground motions, three approaches are generally used in the analyses of fluid-structure
interaction problems. The simplest one is the added mass approach initially proposed
by Westergaard (1933) (with added masses on the dam). Another approach is the
Eulerian approach (Maity and Bhattacharyya 2003), in which the displacements are
the variables in the structure and the pressure or velocity potential is the variables in
the fluid. Since these variables in the structure and fluid are different in the Eulerian
approach, a special-purpose computer program is required for the solution of coupled
systems. The third way to represent the fluid-structure interaction is the Lagrangian
approach (Wilson and Khalvati 1983; Calayir and Dumanoǧlu 1993), where the
displacements are the variables for both the fluid and the structure) approaches. For
that reason, Lagrangian displacement-based fluid elements can be easily incorporated
into a general-purpose computer program for structural analysis, because special
interface equations are not required. Dynamic response of fluid-structure systems
using the Lagrangian approach has been investigated by many researchers (Calayir
and Karaton 2005b; Bilici et al. 2009).
The formulation of the fluid-solid system based on Lagrangian approach is given
according to Refs (Wilson and Khalvati 1983; Calayir and Dumanoǧlu 1993). In
this approach, displacements are selected as the variables in both fluid and structure
domains. Fluid is assumed to be linearly elastic, inviscid, and irrotational. For a
general two-dimensional fluid element, the stress-strain relationships can be written
as:

#
#
P C11 0 εv
= (2.37)
Pz 0 C22 WZ
where P is the pressure, C 11 is the bulk modulus of fluid and εv is the volumetric
strain; W z is the rotation about the axis z, Pz and C 22 are the rotational stress and
constraint parameter related with W z , respectively. Note that the rotation constraint
parameter C 22 in the above stress–strain relationship of fluid is introduced to enforce
2.5 Lagrangian Formulation for Dynamic Interaction … 33
the irrotationality of fluid by penalty method. It should be as high as necessary to

prevent fluid rotation but small enough to avoid causing numerical ill-conditioning
in the assembled stiffness matrix.
In the analysis, the effects of the small amplitude free surface waves, which are
commonly referred to as the sloshing effects, are taken into account. The sloshing
effects cause a pressure at the free surface of fluid, which is given by
P = −γw u fn (2.38)
where γ w is the weight density of fluid and ufn is the normal component of the free
surface displacement. The free surface stiffness for fluid is obtained from the discrete
form of Eq. (2.38).
In this study, the equations of motion of the fluid system are obtained using energy
principles. Using the finite element approximation, the total strain energy of the fluid
system can be written as
1 T
πe = U KfUf (2.39)
2 f
where Uf is the nodal displacement vector, Kf is the stiffness matrix of the fluid
system. Kf is obtained by the sum of the stiffness matrices of the fluid elements as
follows
⎫
K f = Kef ⎬
$ eT (2.40)
K f = B f C f B f dV ⎭
e e e
V
where Cf is the elasticity matrix consisting of diagonal terms in Eq. (2.37). Bef is the
strain-displacement matrix of the fluid element.
An important behavior of fluid systems is the ability to displace without a change
in volume. For reservoir and storage tanks, this movement is known as sloshing waves
in which the displacement is in the vertical direction. The increase in the potential
energy of the system because of the free surface motion can be written as
1 T
πs = U S f Us f (2.41)
2 sf
where Usf and Sf are the vertical nodal displacement vector and the stiffness matrix
of the free surface of the fluid system, respectively. Sf is obtained by the sum of the
stiffness matrices of the free surface fluid elements as follows
⎫
S f = Sef ⎬
$ T (2.42)
S f = ρ f g hs hs dA ⎭
e e
A
where hs is the vector consisting of interpolation functions of the free surface fluid
element. ρ f and g are the mass density of the fluid and the acceleration due to gravity,
respectively. In addition, the kinetic energy of the system can be written as
1 T
T = U̇ M f U̇ f (2.43)
2 f
where U̇ and Mf are the nodal velocity vector and the mass matrix of the fluid system,
respectively. Mf is also obtained by the sum of the mass matrices of the fluid elements
as follows
⎫
M f = Mef ⎬
$ T (2.44)
M f = ρ f H HdV ⎭
e e
V
where H is the matrix consisting of interpolation functions of the fluid element.

If Eqs. (2.39), (2.41) and (2.43) are combined using the Lagrange’s equations, the
following set of equations is obtained
M f Ü f + K∗f U f = R f (2.45)
where Ü f , U f , K∗f and Rf are the nodal acceleration, the nodal displacement, the
system stiffness matrix including the free surface stiffness and the time-varying nodal
force vector for the fluid system, respectively. In the formation of the fluid element
matrices, reduced integration orders are utilized.
The equations of motion of the fluid system (Eq. (2.45)), have a similar form
to those of the structure system. For obtaining the coupled equations of the fluid-
structure system, it is required to determine the interface condition. Because the
fluid is assumed to be inviscid, only the displacement in the normal direction to
the interface is continuous at the interface of the system. Assuming that positive
face is the structure and negative face is the fluid, the boundary condition at the
fluid-structure interface is
Un− = Un+ (2.46)
where U n is the normal component of the interface displacement (Akkas et al. 1979).
Using the interface condition, the equations of motion of the coupled system to ground
motions including damping effects are given by
Mc Üc + Cc U̇c + Kc Uc = Rc (2.47)
where Mc , Cc and Kc are the mass, damping and stiffness matrices for the coupled
system, respectively. Uc , U̇c and Üc are the vectors of the displacement, velocity,
acceleration of the coupled system, respectively. Rc is the time-varying nodal force
vector of ground acceleration.
2.6 Application of the Two Models in Concrete Gravity Dams 35
2.6 Application of the Two Models in Concrete Gravity

Dams
2.6.1 Description of Koyna Gravity

Dam-Reservoir-Foundation System
Koyna concrete gravity dam in India, 103 m high and 70.2 m wide at its base, is
selected for numerical application. The Koyna earthquake of magnitude 6.5 on the
Richter scale on December 11, 1967, with maximum acceleration measured at the
foundation gallery of 0.49 g and 0.34 g in horizontal and vertical direction, has
caused very serious structural damage to the dam, including horizontal cracks on
the upstream and downstream faces of a number of non-overflow monoliths around
the elevation at which the slope of the downstream face changes abruptly (Chopra
and Chakrabarti 1973). Leakage was observed in some of these monoliths near the
changes in the slope of the downstream face, implying a complete penetration from
the upstream face to the downstream face.
This problem has been extensively analyzed by a number of investigators. In
this section, dynamic response analyses of the Koyna dam are performed by using
the XFEM and CDP model. The time history of the Koyna earthquake is shown
in Fig. 2.8. Two-dimensional 4-node plane strain elements (CPE4R) with reduced
integration and hourglass control are used to discretize the concrete dam. The mesh
of the concrete dam is adequately refined at the base and near the change in the slope
of the downstream face where crack growth is expected. The results of the mesh
sensitivity analysis suggest that the mesh sizes along the base and near the change
in the downstream slope are 1.0 m and 0.5 m, respectively. The element size for
other parts is about 1.5 m. The finite element model of the dam-reservoir-foundation
interaction system is given in Fig. 2.9a, and the dimensions of the dam are given in
Fig. 2.9b.
For the initial time step, nodal displacements on the left and right truncated
boundary of the dam-reservoir-foundation system are assumed to be zero in the
normal direction. In addition, the foundation base is fully constrained. In the sequent
0.4 Horizontal Component 0.4 Vertical Component

Acceleration (g)
Acceleration (g)
0.2 0.2
0.0 0.0
-0.2 -0.2
-0.4 -0.4
0 2 4 6 8 10 0 2 4 6 8 10
Time (s) Time (s)
(a) (b)
Fig. 2.8 Koyna earthquake on December 11, 1967. a Horizontal component; b vertical component
Fig. 2.9 Finite element discretization for the dam- reservoir-foundation system of Koyna dam:
a dam-reservoir-foundation system and b concrete gravity dam
dynamic analysis, all displacement constraints are released and the selected earth-
quake accelerations are applied to the foundation base as the input load. At the fluid-
solid interface, the displacement in the normal direction to the interface is assumed
to be continuous during the entire simulation. The foundation rock is assumed to be
linearly elastic and massless. The massless foundation is used to avoid reflection of
the outgoing waves. The reservoir water is assumed to be linearly elastic, irrotational,
and inviscid.
The material parameters for concrete, foundation rock, and water are listed in
Table 2.2. The material parameters used in the CDP model for concrete is shown
in Table 2.3. In order to consider strain rate effects of concrete material, a dynamic
magnification factor of 1.2 is implemented to the tensile strength of concrete. To
account for the energy dissipation in the dam-reservoir-foundation system, a Rayleigh
damping ratio of 5% is applied to the natural frequencies of the first and second
vibration modes of the entire system, which are 2.55 Hz and 6.13 Hz, respectively.
Table 2.2 Material

Material name Property name Value Unit
parameters for concrete,
foundation rock, and water Concrete Young’s modulus 31.0 GPa
Mass density 2643 kg/m3
Tensile strength 2.9 MPa
Compressive strength 24.1 MPa
Poisson’s ratio 0.2 –
Foundation rock Young’s modulus 21.6 GPa
Poisson’s ratio 0.2 –
Water Bulk modulus 2.07 GPa
Mass density 1000 kg/m3
Table 2.3 Material parameters used in the CDP model for concrete
Concrete compression hardening and damage Concrete tension stiffening and damage
Stress (MPa) Crushing strain Damage Stress (MPa) Cracking strain Damage
13 0 0 2.9 0 0
24.1 0.00016 0 2.76 0.000012 0.068
20.64 0.00054 0.244 2.30 0.000036 0.219
12.66 0.00132 0.541 1.57 0.000085 0.435
6.45 0.00288 0.756 0.94 0.000181 0.640
3.15 0.00599 0.876 0.54 0.000375 0.787
1.54 0.01221 0.938 0.31 0.000761 0.878
0.76 0.02464 0.969 0.19 0.001535 0.961
The values of the mass and stiffness proportional damping coefficients used for both
fluid and structure are then calculated to be α = 1.1317 and β = 0.001834.
2.6.2 A Comparative Study on Seismic Nonlinear Response
Dynamic crack propagation processes of Koyna gravity dam under the 1967 earth-
quake including dam-reservoir-foundation interaction are conducted employing both
the XFEM based on the cohesive segments method and the CDP model made in
ABAQUS program. The integration time step used in the analysis is 0.01 s. For
dynamic input, the transverse and vertical acceleration components of the 1967
Koyna earthquake are selected. The linear and nonlinear solutions obtained from
the dynamic analysis of the dam-reservoir-foundation system are compared with
each other. The cracking effects of the concrete material on the seismic response of
the dam-reservoir-foundation system are examined.
The static solutions of the dam due to its gravity loads and hydrostatic loads are
taken as initial conditions in the dynamic analyses of the system. For previously
evaluating the seismic performance and predicting the overstressed regions where
cracking may occur in the dam body, the maximum and minimum principal stresses
for the linear dynamic analysis case are shown in Fig. 2.10. The numeric values
marked on these envelopes are given in terms of Pa. From the response results, it
can be found that a larger tensile stress is observed at the dam heel and the upper
part of the dam, especially near the change in the slope of the downstream face.
The maximum tensile stress is 9.30 MPa on the downstream face at the elevation of
the downstream slope change, which is approximately 3.2 times the tensile strength
selected for the concrete material. These maximum tensile stress on the upstream
face exceed 4.5 MPa, and exceed 7.0 MPa around at the dam heel. Compressive
stresses also take larger values in the upper part of the upstream and downstream
faces, especially around the elevation of the downstream slope change. The maximum
Fig. 2.10 Envelopes of maximum (a) and minimum (b) principal stresses of Koyna dam for the
linear dynamic analysis case
compressive stress is 11.55 MPa, which is smaller than the compressive strength of
concrete material. In these overstressed regions, cracks in the dam body are expected
to occur. Therefore, it can be stated that the linear time history analysis of the dam-
reservoir-foundation system is insufficient, and the nonlinear time history analysis
is required to further assess the dam damage and estimate the performance more
accurately.
The results corresponding to the linear dynamic response clearly indicate that
significant nonlinear deformation of Koyna gravity dam should be expected for the
1967 Koyna earthquake. The seismic failure processes of Koyna dam during the
Koyna earthquake are predicted using both the XFEM and CDP model. The crack
propagation processes and final crack profile of Koyna dam obtained from the XFEM
are shown in Fig. 2.11. As shown, smooth curvature discrete cracks penetrating the
elements are obtained. The accumulated damage processes of Koyna dam during
the Koyna earthquake as predicted using the CDP model are given in Fig. 2.12. The
damage of a crack at an element integration point is indicated by shading the related
area with red color. As shown, the crack propagation processes and failure modes
are obtained.
As observed by comparing Figs. 2.11 and 2.12, the crack propagation processes
are more or less similar, with the time for two initial cracks beginning to propagate
at the dam heel and around the elevation at which the slope of the downstream face
changes abruptly. Subsequently, crack trajectories obtained from the XFEM and CDP
model have some differences. For both the nonlinear constitutive models, the initial
crack propagation in dam firstly occurs at the dam heel at the time t = 2.71 s, and
Fig. 2.11 Crack propagation processes of Koyna gravity dam obtained from the XFEM at six
selected times. a t = 2.71 s; b t = 3.87 s; c t = 3.92 s; d t = 4.25 s; e t = 4.27 s; f t = 4.48 s
other initial crack in the dam is observed at 3.87 s near the change in the slope of the
downstream face. At these locations, stresses are concentrated and the tensile stresses
take large values. After t = 3.92 s, there is some difference between the cracking
propagation process from the XFEM and CDP model. For the XFEM technique, the
initial crack near the change in the slope of the downstream face extends deeper inside
of the dam, and propagates about two-third through the width of the dam section at t
= 4.25 s. The crack trajectory curves down due to the compressive stresses resulting
from rocking of the top block (Fig. 2.11c, d. After the time instant t = 4.25 s, the
crack propagates horizontally toward to upstream face. At t = 4.48 s, the downstream
crack extends completely to the upstream face at a height of 60.0 m above the base,
penetrating the whole section of the dam. It can be noted that the crack propagation
processes are mostly concentrated at the time from 3.87 to 4.48 s when the dynamic
displacement response of the dam crest is larger, and the dam remains its overall
safety. While, for the CDP model, initial crack begins to initialize near the middle
Fig. 2.12 Crack propagation processes of Koyna gravity dam obtained from the CDP model at six
selected times. a t = 2.71 s; b t = 3.87 s; c t = 3.92 s; d t = 4.21 s; e t = 4.67 s; f t = 5.56 s
of the upstream face (Fig. 2.12d) due to the vibration characteristics at t = 4.21.
As the dam oscillates during the earthquake, these previously formed cracks at the
upstream and downstream faces extend into the dam, penetrating the whole section
of the dam. Based on the XFEM and CDP model, it can be found that the Koyna
earthquake causes serious structural damage to the Koyna dam.
The final crack profiles of Koyna gravity dam under Koyna ground motion
predicted by the XFEM and CDP model are compared with the previous research
results and model test. Figure 2.13 summarizes the final crack profiles obtained by
different approaches. By comparing the current results with the Koyna dam proto-
type observation (Chopra and Chakrabarti 1973), the model test (Dams 1990b) and
previous research results, it can be noted that the crack trajectories in the dam are
different from the various fracture procedures. While, the finial crack profiles are
Fig. 2.13 Comparison of the final crack profile of Koyna gravity dam during 1967 Koyna earthquake
obtained by different approaches. a Extended finite element method (XFEM); b Concrete damaged
plasticity (CDP) model; c Experimental result (Dams 1990b); d Smeared crack model (Mirzabozorg
and Ghaemian 2005); (e) Non-linear smeared fracture (Bhattacharjee and Leger 1993); f Smeared
carck model considering fracture energy effects (Hariri-Ardebili et al. 2013); g Nonlinear fracture
mechanics (Wang et al. 2000); h DruckerPrager elasto-plastic model(Pan et al. 2011); i Crack
band finite element method (Pan et al. 2011); j Plastic-damage model (Lee and Fenves 1998a,
b); k Plastic-damage model employing the damage-dependent damping mechanism (Omidi et al.
2013); l Continuum damage concrete model (Calayir and Karaton 2005b); m co-axial rotating crack
model (Calayir and Karaton 2005a); n Damage rupture process (Chen et al. 2014); o Rock failure
process analysis (Zhong et al. 2013); p Mesh-free approach (SPH) (Das and Cleary 2013)
more or less similar. The failure mechanism is formed by two main damage zones,
one at the base, and one in the upper part of the dam around the elevation at which the
slope of the downstream face changes abruptly. The top cracking profiles are almost
either nearly horizontal or sloping downward from the downstream face toward
the upstream face. But in some analyses, cracks are predicted to initialize near the
middle of the upstream face or the downstream face, and extend into the dam. From
the Fig. 2.13, it may be concluded that both the XFEM and CDP model can predict
effectively the crack propagation processes in concrete gravity dams under seismic
conditions.
The seismic response of the dam-reservoir-foundation system is investigated by
comparison of the linear and nonlinear dynamic procedures. Figure 2.14 shows the
time history graphs of the horizontal and vertical displacements of the dam crest
(point P1) obtained from the linear elastic, XFEM, and CDP model. The positive
directions of the horizontal and vertical displacement are in the downstream and
upward directions, respectively. There is no cracking during the relatively small
amplitude motion. Although cracking in the dam is firstly observed at 2.71 s around
the dam heel, the depth of the crack propagation is smaller. Subsequently, other initial
crack in the dam is observed at 3.87 s near the change in the slope of the downstream
face, which extends deeper inside of the dam. The nonlinear displacement responses
1.5
4 XFEM XFEM
Displacement (cm)
Displacement (cm)
CDP 1.0 CDP

Linear Linear
2
0.5
0 0.0
-2 -0.5
-1.0
-4
-1.5
0 2 4 6 8 10 0 2 4 6 8 10
Time (s) Time (s)
(a) (b)
Fig. 2.14 Comparison of the horizontal and vertical displacement histories of the point P1 at the
dam crest from the different procedures. a Horizontal displacement; b vertical displacement
based on the XFEM and CDP before 3.87 s are almost coincident with those calcu-
lated from the linear elastic analysis. The subsequent displacements obtained from
the linear and nonlinear solutions separately with each other as the cracks form and
propagate in the dam. It can be seen from Fig. 2.14 that displacement responses are
significantly different from the various procedures as the cracks propagate in the
dam. The differences in the displacement amplitude, which are initially small, reach
important levels as the cracks propagate deeper inside of the dam. The vibration
periods of displacement response are also changed by the crack propagation, and a
lengthening vibration period from the XFEM and CDP model is found, implying that
the rigidity of the dam is gradually decreased due to concrete softening. It has been
also found that after the strong earthquake, more significant residual displacements
using the CDP model remain in the dam crest due to plastic strain during the cyclic
loading. At all times, nonlinear response analyses from the two fracture procedures
indicate that the dam-reservoir-foundation system remains stable. However, the final
conclusions about the dynamic stability should be based on a separate rigid-body
analysis of the block considering large displacements and water-fracture interaction
mechanisms.
Figure 2.15 shows the time history graphs of the maximum and minimum principal
stresses occurred in the center of the element E1 located at the changes in the slope of
the downstream face. It is seen from Fig. 2.15 that the stress responses obtained from
different procedures are also significantly different. The maximum peak value of the
maximum principal stresses for the linear case is larger than that in the nonlinear
case. The maximum peak value of the maximum principal stresses for the linear
case is 8.53 MPa. While, the maximum peak values of those for the two nonlinear
cases are about the tensile strength of the concrete material (2.9 MPa). Figure 2.15a
confirms that the tensile strength obtained from the XFEM is completely removed
after cracking. In CDP model, the damage associated with the failure mechanisms
of the concrete will result in a reduction of the elastic stiffness. Therefore, the tensile
strength using the CDP model does not completely reduce to zero after damage
Minimum principal stress (MPa)
Maximum principal stress (MPa)
XFEM 0
8
CDP -2
Linear
6 -4
-6
4
-8
2 -10
XFEM
-12
CDP
0 -14 Linear
0 2 4 6 8 10 0 2 4 6 8 10
Time (s) Time (s)
(a) (b)
Fig. 2.15 Comparison of the maximum (a) and minimum (b) principal stress histories from the
different procedures occurred in the center of the element E1 (near the changes in the slope of the
downstream face)
Minimum principal stress (MPa)

8 XFEM
CDP
0
Linear
6
-2
4
-4
2
XFEM
-6 CDP
0 Linear
0 2 4 6 8 10 0 2 4 6 8 10
Time (s) Time (s)
(a) (b)
Fig. 2.16 Comparison of the maximum (a) and minimum (b) principal stress histories from the
different procedures occurred in the center of the element E2 (at the dam heel)
occurring. The minimum principal stress for the nonlinear cases generally takes
larger peak values. The maximum peak values of the minimum principal stresses for
the XFEM and the CDP model are 12.28 MPa and 14.55 MPa, respectively, which
is smaller than the compressive strength of the concrete material. The time history
graphs of the maximum and minimum principal stresses occurred in the center of
element E2 located at the dam heel are given in Fig. 2.16. The variations in these
stresses are similar to those of the element E1. After 2.71 s, the stress responses
obtained from the linear and nonlinear solutions separately with each other as the
cracks form and propagate in the dam.
The XFEM technique based on the cohesive segments method in conjunction
with the phantom node technique is used to model crack initiation and propagation
along an arbitrary path. Therefore, the crack response behavior can be obtained
based on the XFEM. The time history graphs of the crack opening displacement
at the downstream and upstream faces near the elevation at which the slope of the
downstream face changes abruptly are presented in Fig. 2.17. The peak value of crack
opening displacement at the downstream end happens at 4.32 s, reaching 14.76 mm,
and at the upstream end happens at 4.55 s, reaching 10.91 mm. As shown in Fig. 2.17,
it is known that the crack opening displacement at the downstream end is dominant
Fig. 2.17 Time history 16

14.76mm Downstream
graphs of the crack opening 14 Upstream
displacement (COD) 12 10.91mm
10
COD (mm)
8
6
4
Initial crack
2
0
Penetrated crack
-2
0 2 4 6 8 10
Time (s)
Fig. 2.18 Sliding 2
Sliding along the crack (mm)

1.58mm
displacement histories of the
upper block of the dam after 1
the formation of the 0
penetrated crack
-1 Penetrated crack
-2
-3
-3.73mm
-4
0 2 4 6 8 10
Time (sec)
compared with that at the upstream end. The reason for this is mainly because the
upstream-sloped crack, which increases the resistance against downstream sliding
of the upper block, makes the upper block easier to rotate toward the upstream
direction rather than downstream direction under seismic conditions. Figure 2.18
shows the histories of the sliding of the upper block of the dam after the formation
of the penetrated crack obtained by the XFEM. As may be seen from Fig. 2.18,
the maximum sliding displacement toward the downstream direction is 1.58 mm
at 4.52 s, and toward the upstream direction is −3.73 mm at 4.75 s. It may be
concluded that when the Koyna earthquake causes the penetrated crack damage to
Koyna dam, the upper block remains stable against overturning, as the maximum
sliding displacement of 3.73 mm is almost negligible, and the residual displacement
is very small. It should be noted that these crack response behaviors can’t be obtained
from the CDP model.
2.7 Conclusions
In this study, nonlinear dynamic response and seismic failure processes of

concrete gravity dams are investigated with considering the effects of dam-
reservoir-foundation interaction. The Lagrangian approach is used for the finite
element modeling of the dam-reservoir-foundation interaction problem. Two fracture
modelling approaches, XFEM and CDP model, are compared for failure analyses
of concrete gravity dams. The XFEM based on the cohesive segments method in
conjunction with the phantom node technique can be used to model crack initiation
and propagation along an arbitrary path. While, the CDP model including the strain
hardening or softening behavior is presented to describe the accumulated damage
processes of structures under strong ground motions. The Koyna gravity dam is
selected as a numerical application for analyzing the failure processes using the two
approaches. The effects of different nonlinear approaches on the seismic response
of the dam are compared and discussed.
Comparison of the results of the two different nonlinear approaches, the Koyna
dam prototype observation, model test and those of available methods in the literature
indicates that both the XFEM and the CDP model selected for concrete gravity dams
can be used in the seismic failure analysis. The crack initiation obtained from the two
procedures is more or less similar, with the same time for two initial cracks beginning
to propagate at the dam heel and downstream face. While, the crack trajectories in
the dam are different from the various fracture procedures. The failure mechanism
is formed by two main damage zones, one at the base, and one in the upper part of
the dam around the elevation at which the slope of the downstream face changes
abruptly. For the XFEM technique, the initial crack near the change in downstream
slope extends completely to the upstream face, penetrating the whole section of the
dam. While, crack obtained from the CDP model is also predicted to initialize near
the middle of the upstream face, these previously formed cracks at the upstream and
downstream faces extend into the dam, which penetrates the whole monolith of the
dam.
Nonlinear dynamic response of the dam-reservoir-foundation system is inves-
tigated by comparison of the linear and nonlinear dynamic procedures. During
the relatively small amplitude motion, nonlinear dynamic responses based on the
XFEM and CDP model are almost coincident with those calculated from the linear
elastic analysis. While, displacement and stress responses are significantly different
from the various procedures as the cracks propagate in the dam. Significant residual
displacements are predicted by the CDP model due to plastic strain during the cyclic
loading, which haven’t been found in the XFEM. However, the XFEM can be used
to predict opening and sliding displacements of the cracks. In order to capture the
cracking mechanism and nonlinear dynamic response of concrete gravity dams under
strong ground motions, appropriate nonlinear models simulating crack propagation
are critically important for a rigorous seismic safety evaluation.
References
Akkas, N., Akay, H. U., & Yilmaz, C. (1979). Applicability of general-purpose finite element
programs in solid-fluid interaction problems. Computers & Structures, 10(5), 773–783.
Aldemir, A., Binici, B., Arici, Y., Kurc, O., & Canbay, E. (2015). Pseudo-dynamic testing of a
concrete gravity dam. Earthquake Engineering and Structural Dynamics, 44(11), 1747–1763.
Alijani-Ardeshir, M., Neya, B. N., & Ahmadi, M. (2019). Comparative study of various smeared
crack models for concrete dams. Journal of the Croatian Association of Civil Engineers, 71(4),
305–318.
Belytschko, T., Gracie, R., & Ventura, G. (2009). A review of extended/generalized finite element
methods for material modeling. Modelling and Simulation in Materials Science and Engineering,
17(4), 43001.
References 47
Belytschko, T., & Black, T. (1999). Elastic crack growth in finite elements with minimal remeshing.
International Journal for Numerical Methods in Engineering, 45(5), 601–620.
Bhattacharjee, S. S., & Leger, P. (1993). Seismic cracking and energy dissipation in concrete gravity
dams. Earthquake Engineering & Structural Dynamics, 22(11), 991–1007.
Bilici, Y., Bayraktar, A., Soyluk, K., Haciefendioğlu, K., Ateş, Ş., & Adanur, S. (2009). Stochastic
dynamic response of dam–reservoir–foundation systems to spatially varying earthquake ground
motions. Soil Dynamics and Earthquake Engineering, 29(3), 444–458.
Bretas, E. M., Lemos, J. V., & Lourenço, P. B. (2014). A DEM based tool for the safety analysis of
masonry gravity dams. Engineering Structures, 59, 248–260.
Calayir, Y., & Dumanoǧlu, A. A. (1993). Static and dynamic analysis of fluid and fluid-structure
systems by the Lagrangian method. Computers & Structures, 49(4), 625–632.
Calayir, Y., & Karaton, M. (2005a). Seismic fracture analysis of concrete gravity dams including
dam-reservoir interaction. Computers & Structures, 83(19), 1595–1606.
Calayir, Y., & Karaton, M. (2005b). A continuum damage concrete model for earthquake analysis
of concrete gravity dam-reservoir systems. Soil Dynamics and Earthquake Engineering, 25(11),
857–869.
Cervera, M., Oliver, J., & Faria, R. (1995). Seismic evaluation of concrete dams via continuum
damage models. Earthquake Engineering and Structural Dynamics, 24(9), 1225–1245.
Cervera, M., Oliver, J., & Manzoli, O. (1996). A rate-dependent isotropic damage model for the
seismic analysis of concrete dams. Earthquake Engineering and Structural Dynamics, 25(9),
987–1010.
Chen, B., Gu, C., Bao, T., Wu, B., & Su, H. (2016). Failure analysis method of concrete arch dam
based on elastic strain energy criterion. Engineering Failure Analysis, 60, 363–373.
Chen, D., Yang, Z., Wang, M., & Xie, J. (2019). Seismic performance and failure modes of the
Jin’anqiao concrete gravity dam based on incremental dynamic analysis. Engineering Failure
Analysis, 100, 227–244.
Chen, H., Li, D., & Guo, S. (2014). Damage-rupture process of concrete dams under strong
earthquakes. International Journal of Structural Stability and Dynamics, 14(7), 1450021.
Chen, J., Wang, M., & Fan, S. (2013). Experimental investigation of small-scaled model for
powerhouse dam section on shaking table. Structural Control and Health Monitoring, 20(5),
740–752.
Chen, K., Zou, D., & Kong, X. (2017). A nonlinear approach for the three-dimensional polyhe-
dron scaled boundary finite element method and its verification using Koyna gravity dam. Soil
Dynamics and Earthquake Engineering, 96, 1–12.
Chopra, A. K. (1988). Earthquake response analysis of concrete dams. In Advanced dam engineering
for design, construction and rehabilitation (pp. 416–465). Boston, MA: Springer.
Dams, N.R.C.U. (1990a). Earthquake engineering for concrete dams: Design, performance, and
research needs. National Academies Press.
Dams, N.R.C.U. (1990b). Earthquake engineering for concrete dams: Design, performance, and
research needs. National Academies Press.
Daneshyar, A., & Ghaemian, M. (2019). Seismic analysis of arch dams using anisotropic damage-
plastic model for concrete with coupled adhesive-frictional joints response. Soil Dynamics and
Earthquake Engineering, 125, 105735.
Dragon, A., & Mroz, Z. (1979). A continuum model for plastic-brittle behaviour of rock and
concrete. International Journal of Engineering Science, 17(2), 121–137.
El-Aidi, B., & Hall, J. F. (1989a). Non-linear earthquake response of concrete gravity dams part 1:
Modelling. Earthquake Engineering & Structural Dynamics, 18(6), 837–851.
El-Aidi, B., & Hall, J. F. (1989b). Non-linear earthquake response of concrete gravity dams part 2:
Behaviour. Earthquake Engineering and Structural Dynamics, 18(6), 853–865.
Ghaedi, K., Jameel, M., Ibrahim, Z., & Khanzaei, P. (2015). Seismic analysis of roller compacted
concrete (RCC) dams considering effect of sizes and shapes of galleries. KSCE Journal of Civil
Engineering.
Ghaemmaghami, A. R., & Ghaemian, M. (2008). Experimental seismic investigation of Sefid-rud
concrete buttress dam model on shaking table. Earthquake Engineering and Structural Dynamics,
37(5), 809–823.
Ghaemmaghami, A. R., & Ghaemian, M. (2010). Shaking table test on small-scale retrofitted
model of Sefid-rud concrete buttress dam. Earthquake Engineering and Structural Dynamics,
39, 109–118.
Hansbo, A., & Hansbo, P. (2004). A finite element method for the simulation of strong and weak
discontinuities in solid mechanics. Computer Methods in Applied Mechanics and Engineering,
193(33), 3523–3540.
Hariri-Ardebili, M. A., Seyed-Kolbadi, S. M., & Kianoush, M. R. (2016). FEM-based parametric
analysis of a typical gravity dam considering input excitation mechanism. Soil Dynamics and
Hariri-Ardebili, M. A., Seyed-Kolbadi, S. M., & Mirzabozorg, H. (2013). A smeared crack model for
seismic failure analysis of concrete gravity dams considering fracture energy effects. Structural
Engineering and Mechanics, 48(1), 17–39.
Hariri-Ardebili, M. A., & Seyed-Kolbadi, S. M. (2015). Seismic cracking and instability of concrete
dams: Smeared crack approach. Engineering Failure Analysis, 52, 45–60.
Hillerborg, A., Mod E Er, M., & Petersson, P. (1976). Analysis of crack formation and crack growth
in concrete by means of fracture mechanics and finite elements. Cement and Concrete Research,
6(6), 773–781.
Huang, J. (2018). An incrementation-adaptive multi-transmitting boundary for seismic fracture
analysis of concrete gravity dams. Soil Dynamics and Earthquake Engineering, 110, 145–158.
Ingraffea, A. R. (1990). Case studies of simulation of fracture in concrete dams. Engineering
Fracture Mechanics, 35(1–3), 553–564.
reinforced and plain concrete. In Fracture Mechanics of Concrete: Structural application and
Khazaei Poul, M., & Zerva, A. (2018). Nonlinear dynamic response of concrete gravity dams
considering the deconvolution process. Soil Dynamics and Earthquake Engineering, 109, 324–
338.
Kong, X., Zhang, Y., Zou, D., Qu, Y., & Yu, X. (2017). Seismic cracking analyses of two types of face
slab for concrete-faced rockfill dams. Science China Technological Sciences, 60(4), 510–522.
Lagier, F., Jourdain, X., De Sa, C., Benboudjema, F., & Colliat, J. B. (2011). Numerical strategies for
prediction of drying cracks in heterogeneous materials: Comparison upon experimental results.
Lee, J., & Fenves, G. L. (1998a). A plastic-damage concrete model for earthquake analysis of dams.
Lee, J., & Fenves, G. L. (1998b). Plastic-damage model for cyclic loading of concrete structures.
Journal of Engineering Mechanics, 124(8), 892–900.
Li, Q. S., Li, Z. N., Li, G. Q., Meng, J. F., & Tang, J. (2005). Experimental and numerical seismic
investigations of the Three Gorges dam. Engineering Structures, 27(4), 501–513.
Lohrasbi, A., & Attarnejad, R. (2008). Crack growth in concrete gravity dams based on discrete
crack method. American Journal of Engineering and Applied Sciences, 1(4), 318–323.
Lotfi, V., & Espandar, R. (2004). Seismic analysis of concrete arch dams by combined discrete
crack and non-orthogonal smeared crack technique. Engineering Structures, 26(1), 27–37.
References 49
Lu, X., Wu, Z., Pei, L., He, K., Chen, J., Li, Z., et al. (2019). Effect of the spatial variability of strength
parameters on the dynamic damage characteristics of gravity dams. Engineering Structures, 183,
281–289.
Lubliner, J., Oliver, J., Oller, S., & Onate, E. (1989). A plastic-damage model for concrete.
International Journal of Solids and Structures, 25(3), 299–326.
Ma, Y., Qin, Y., Chai, J., & Zhang, X. (2019). Analysis of the effect of initial crack length on concrete
members using extended finite element method. International Journal of Civil Engineering,
17(10), 1503–1512.
Maity, D., & Bhattacharyya, S. K. (2003). A parametric study on fluid–structure interaction
problems. Journal of Sound and Vibration, 263(4), 917–935.
Markovič, M., Krauberger, N., Saje, M., Planinc, I., & Bratina, S. (2013). Non-linear analysis of
pre-tensioned concrete planar beams. Engineering Structures, 46, 279–293.
Mazars, J., & Pijaudier-Cabot, G. (1989). Continuum damage theory-application to concrete.
Melenk, J. M., & Babuška, I. (1996). The partition of unity finite element method: Basic theory and
applications. Computer Methods in Applied Mechanics and Engineering, 139(1), 289–314.
Mirzabozorg, H., & Ghaemian, M. (2005). Non-linear behavior of mass concrete in three-
dimensional problems using a smeared crack approach. Earthquake Engineering and Structural
Dynamics, 34(3), 247–269.
Moës, N., Dolbow, J., & Belytschko, T. (1999). A finite element method for crack growth without
remeshing. International Journal for Numerical Methods in Engineering, 46(1), 131–150.
Moradloo, J., Naserasadi, K., & Zamani, H. (2018). Seismic fragility evaluation of arch concrete
dams through nonlinear incremental analysis using smeared crack model. Structural Engineering
& Mechanics, 68(6), 747–760.
Morin, P. B., Léger, P., & Tinawi, R. (2002). Seismic behavior of post-tensioned gravity dams:
shake table experiments and numerical simulations. Journal of Structural Engineering, 128(2),
140–152.
Murakami, S., & Ohno, N. (1981). A continuum theory of creep and creep damage. Creep in
Structures, 28, 422–444.
Nuss, L. K., Matsumoto, N., & Hansen, K. D. (2012). Shaken but not stirred-earthquake performance
of concrete dams. In USSD Proceedings (pp. 1511–1530).
63, 80–97.
Omidi, O., & Lotfi, V. (2017). Seismic plastic–damage analysis of mass concrete blocks in arch
dams including contraction and peripheral joints. Soil Dynamics and Earthquake Engineering,
95, 118–137.
Pan, J., Feng, Y., Jin, F., Zhang, C., & Owen, D. R. J. (2014). Comparison of different fracture
modelling approaches to gravity dam failure. Engineering Computations, 31(1), 18–32.
Pan, J., Zhang, C., Xu, Y., & Jin, F. (2011). A comparative study of the different procedures for
seismic cracking analysis of concrete dams. Soil Dynamics and Earthquake Engineering, 31(11),
1594–1606.
Pang, R., Xu, B., Zhou, Y., Zhang, X., & Wang, X. (2020). Fragility analysis of high CFRDs
subjected to mainshock-aftershock sequences based on plastic failure. Engineering Structures,
206, 110152.
Pekau, O. A., Lingmin, F., & Chuhan, Z. (1995). Seismic fracture of Koyna Dam: Case study.
Phansri, B., Charoenwongmit, S., Warnitchai, P., Shin, D. H., & Park, K. H. (2010). Numerical
simulation of shaking table test on concrete gravity dam using plastic damage model.
Pirboudaghi, S., Tarinejad, R., & Alami, M. T. (2018). Damage detection based on system identifi-
cation of concrete dams using an extended finite element-wavelet transform coupled procedure.
Journal of Vibration and Control, 24(18), 4226–4246.
Pirooznia, A. (2019). Seismic improvement of gravity dams using isolation layer in contact
area of dam-reservoir in smeared crack approach. Iranian Journal of Science and Technology,
Transactions of Civil Engineering, 43(2), 137–155.
Rashid, Y. R. (1968). Ultimate strength analysis of prestressed concrete pressure vessels. Nuclear
Engineering and Design, 7(4), 334–344.
Remmers, J. J. C., de Borst, R., & Needleman, A. (2008). The simulation of dynamic crack prop-
agation using the cohesive segments method. Journal of the Mechanics and Physics of Solids,
56(1), 70–92.
Resatalab, S., Attarnejad, R., & Ghalandarzadeh, A. (2013). Experimental investigation on inter-
action of concrete gravity dam-reservoir-foundation on shaking table. Technical Journal of
Engineering and Applied Sciences, 3(16), 1756–1762.
Rochon-Cyr, M., & Léger, P. (2009). Shake table sliding response of a gravity dam model including
water uplift pressure. Engineering Structures, 31(8), 1625–1633.
Rodríguez, J., Goyal, S., Tiwari, A., Issa, A., Alam, M. S., Seethaler, R., Alam, T., et al. (2020).
Comparative seismic performance of Koyna and Corra Linn dams using numerical analysis and
shake table testing. In ICOLD 2020 Annual Meeting/Symposium, New Delhi, India.
Roşca, B. (2008). Physical model method for seismic study of concrete dams. Bulletin of the
Polytechnic Institute of Jassy: Constructions, Architechture Section LIV (LVIII), 3, 57–76.
Sarkar, R., Paul, D. K., and Stempniewski, L. (2007). Influence of reservoir and foundation on the
nonlinear dynamic reservoir of concrete gravity dams. ISET Journal of Earthquake Technoloy,
44, 377–389.
Shi, Z., Nakano, M., Nakamura, Y., & Liu, C. (2014). Discrete crack analysis of concrete gravity
dams based on the known inertia force field of linear response analysis. Engineering Fracture
Mechanics, 115, 122–136.
Shi, Z., Suzuki, M., & Nakano, M. (2003). Numerical analysis of multiple discrete cracks in concrete
dams using extended fictitious crack model. Journal of Structural Engineering, 129(3), 324–336.
Song, J., Areias, P. M. A., & Belytschko, T. (2006). A method for dynamic crack and shear band
propagation with phantom nodes. International Journal for Numerical Methods in Engineering,
67(6), 868–893.
Stazi, F. L., Budyn, E., Chessa, J., & Belytschko, T. (2003). An extended finite element method
with higher-order elements for curved cracks. Computational Mechanics, 31(1), 38–48.
Tang, C. A., & Kou, S. Q. (1998). Crack propagation and coalescence in brittle materials under
compression. Engineering Fracture Mechanics, 61(3), 311–324.
Theiner, Y., & Hofstetter, G. (2009). Numerical prediction of crack propagation and crack widths
in concrete structures. Engineering Structures, 31(8), 1832–1840.
Tinawi, R., Léger, P., Leclerc, M., & Cipolla, G. (2000). Seismic safety of gravity dams: From snake
table experiments to numerical analyses. Journal of Structural Engineering, 126(4), 518–529.
Wang, G., Wang, Y., Lu, W., Zhou, C., Chen, M., & Yan, P. (2015). XFEM based seismic potential
failure mode analysis of concrete gravity dam–water–foundation systems through incremental
dynamic analysis. Engineering Structures, 98, 81–94.
Wang, H., & Li, D. (2006). Experimental study of seismic overloading of large arch dam. Earthquake
Engineering and Structural Dynamics, 35(2), 199–216.
References 51
Wang, M., Chen, J., Fan, S., & Lv, S. (2014). Experimental study on high gravity dam strength-
ened with reinforcement for seismic resistance on shaking table. Structural Engineering and
Mechanics, 51(4), 663–683.
Wang, M., Chen, J., Wei, H., Song, B., & Xiao, W. (2019). Investigation on seismic damage model
test of a high concrete gravity dam based on application of FBG strain sensor. Complexity, 2019,
1–12.
Westergaard, H. M. (1933). Water pressures on dams during earthquakes. Transactions ASCE, 95(2),
418–433.
Wilson, E. L., & Khalvati, M. (1983). Finite elements for the dynamic analysis of fluid-solid systems.
Yazdani, Y., & Alembagheri, M. (2017). Nonlinear seismic response of a gravity dam under near-
fault ground motions and equivalent pulses. Soil Dynamics and Earthquake Engineering, 92,
621–632.
Yazdchi, M., Khalili, N., & Valliappan, S. (1999). Non-linear seismic behaviour of concrete
gravity dams using coupled finite element-boundary element technique. International Journal
for Numerical Methods in Engineering, 44(1), 101–130.
Ye, C., Shi, J., & Cheng, G. J. (2012). An eXtended finite element method (XFEM) study on the
effect of reinforcing particles on the crack propagation behavior in a metal–matrix composite.
International Journal of Fatigue, 44, 151–156.
Zhang, L., Zhang, H., & Hu, S. (2014). Failure patterns of shear keys and seismic resistance of a
gravity dam with longitudinal joints. Journal of Earthquake and Tsunami, 08(01), 1450004.
Zhang, S., Wang, G., & Yu, X. (2013). Seismic cracking analysis of concrete gravity dams with
initial cracks using the extended finite element method. Engineering Structures, 56, 528–543.
Zhang, S., & Wang, G. (2013). Effects of near-fault and far-fault ground motions on nonlinear
dynamic response and seismic damage of concrete gravity dams. Soil Dynamics and Earthquake
Zhao, S., Fan, S., & Chen, J. (2019). Quantitative assessment of the concrete gravity dam damage
under earthquake excitation using electro-mechanical impedance measurements. Engineering
Zhong, H., Wang, N., & Lin, G. (2013). Seismic response of concrete gravity dam reinforced with
FRP sheets on dam surface. Water Science and Engineering, 6(4), 409–422.
Zhou, J., Lin, G., Zhu, T., Jefferson, A. D., & Williams, F. W. (2000). Experimental investigations
into seismic failure of high arch dams. Journal of Structural Engineering, 126(8), 926–935.
Chapter 3
Seismic Cracking Analysis of Concrete
Gravity Dams with Initial Cracks Using
XFEM
3.1 Introduction
Concrete gravity dams are distinguished from other concrete structures because
of their size and their interactions with the reservoir and foundation. In practical
service, concrete gravity dams normally might have well cracked at their base or
at the upstream and downstream faces caused by internal and external temperature
variations, shrinkage of the concrete, differential foundation settlement, previous
earthquakes, or other reasons (Ingraffea 1990; Feng et al. 1996). These cracks with
limited depth will possibly develop in the dam body or, through the monoliths under
static or dynamic conditions. As a result, these existing cracks weaken the seismic
capacity of concrete gravity dams mainly due to the nonlinear behavior, during strong
ground motions. Hence, the real analytical challenge for the crack analysis of concrete
gravity dams is to predict their propagation paths under seismic loading conditions.
Then, countermeasures can be taken at an early stage to stem their further growth.
Nonlinear response of cracked concrete gravity dams has been of great interest in
engineering. Many studies (Ayari and Saouma 1990; Bhattacharjee and Léger 1994;
Léger and Leclerc 1996; Wang et al. 2000; Ahmadi et al. 2001) focused mainly on
the propagation of cracks in the dam, which is accompanied by opening and closing
of the cracks, including shake table experiments (Hall 1988; Li et al. 2005; Rochon-
Cyr and Léger 2009). However, little efforts have been done for the case in which
cracks are expected to appear at the upstream and downstream faces in non-overflow
monoliths of the dam. In this case, the seismic response becomes more complex.
Shi et al. (2003) used a well-quoted scale model of a concrete gravity dam to
analyze multiple discrete cracks in concrete and obtain various kinds of cracking
behaviors. Barpi and Valente (2000) employed the cohesive crack model to investi-
gate the behavior of a concrete gravity dam of 103 m high with an initial crack in
the upstream face. Their results showed that the initial notch in the upstream face
served as the starting point of a crack that propagated toward the foundation during
the loading process. Bolzon (2004) compared the merits of Linear Elastic Fracture
https://doi.org/10.1007/978-981-15-6194-8_3
54 3 Seismic Cracking Analysis of Concrete Gravity …
Mechanics (LEFM) and cohesive crack approach on the evaluation of safety against
ultimate failure of large concrete gravity dams with an initial notch at the base. Batta
and Pekau (1996) extended the two-dimensional boundary element procedure for
analyzing the propagation of a single discrete crack to simultaneous multiple cracking
in concrete gravity dams. Bhattacharjee and Léger (1994) considered both the coaxial
rotating crack model (CRCM) and the fixed crack model with a variable shear resis-
tance factor (FCM-VSRF) to study the nonlinear response of a model concrete gravity
dam with initial notch on the upstream surface and a full-scale concrete gravity dam
with the pre-assigned imperfection located on the upstream side at the elevation of
the downstream slope change. Their results showed that the ultimate response of the
full-scale dam was not sensitive to the depth of initial imperfection placed on the
upstream side. Oliver et al. (2002) presented the strong discontinuity approach to
observe the fracture process of a reduced model of a concrete gravity dam with an
initial notch on the upstream surface. Tinawi et al. (2000) conducted the shake table
experiment and numerical simulation on four 3.4-m-high plain concrete gravity dam
models with initial notch to study their dynamic cracking and evaluate the seismic
safety. Zhang et al. (2013) presented the extended finite element method (XFEM) to
analyze the seismic crack propagation of concrete gravity dams with initial cracks at
the upstream and downstream faces. Pekau and Cui (2004) used the distinct element
method (DEM) to study the seismic behavior of the fractured Koyna dam during
earthquakes. Their results showed that the safety of the dam was ensured if the crack
shape was horizontal or upstream-sloped, and it was very dangerous if the crack
slopes downstream. Javanmardi et al. (2005) combined the discrete crack model
with a theoretical model for uplift pressure variations along a cracked dam to study
the seismic stability of concrete gravity dams. Wang et al. (2006) studied the stability
of a gravity dam on jointed rock foundation and the seismic stability of the fractured
Konya gravity dam using 3D-discontinuous deformation analysis (3D-DDA). Pekau
and Zhu (2006), Zhu and Pekau (2007) proposed a rigid model and a flexible FE
model to study the seismic behavior of cracked concrete gravity dams. Mirzayee et al.
(2011) proposed a hybrid distinct element-boundary element (DE-BE) approach for
modeling the nonlinear seismic behavior of fractured concrete gravity dams consid-
ering dam-reservoir interaction effects. Shi et al. (2013) used the scaled boundary
polygons coupled with the interface element to investigate the crack propagation in
concrete gravity dams with a preset notch at the upstream face. Jiang et al. (2013)
develop a dynamic contact model and a simplified reinforcing steel constitutive
model to analyze the failure process of the cracked gravity dam with and without
reinforcement.
Cracking plays an important role in the concrete structural behavior, and the
modeling of crack growth is a problem of great importance in the simulation of
failure. In this chapter, the extended finite element method (XFEM), which is based
on the cohesive segments method in conjunction with the phantom node technique,
is used to study the crack propagation and nonlinear fracture behavior of concrete
gravity dams with initial cracks under earthquake conditions. The XFEM can be
used to simulate crack initiation and propagation along an arbitrary path, since the
crack propagation is not tied to the element boundaries in a mesh. A scaled-down
3.1 Introduction 55
1:40 model of a gravity dam with an initial notch on the upstream wall is analyzed
for accuracy verification. The Lagrangian approach is used for the finite element
modeling of dam-reservoir-foundation interaction problem. Subsequently, seismic
cracking analyses of Koyna gravity dam and Guandi gravity dam with initial cracks
are performed. The Koyna dam with single and multiple initial cracks is investigated
to evaluate the seismic crack propagation of concrete gravity dams with initial cracks.
The influence of the initial crack position and length on the crack propagation process
of Guandi gravity dams subjected to design earthquake is also discussed. The crack
propagation process and failure modes of the dam with cracks are obtained.
3.2 Validation Test
A scaled model of a concrete gravity dam tested by Carpinteri et al. (1992) is analyzed
using the XFEM. The model, which contains a horizontal notch of 15 cm on the
upstream face located at a quarter of the dam height, is loaded with an equivalent
hydraulic load in order to induce a curved crack that propagates from the tip of the
notch towards the downstream face, as shown in Fig. 3.1. Numerical simulation of
this test was reported in several public actions using the FEM in combination with
CCMs (Barpi and Valente 2000; Shi et al. 2003).
Finite element model for the dam with the setup is shown in Fig. 3.2. The mesh
of the dam is adequately refined at the lower portion of the dam, in which crack
propagation is expected. The material properties of the model dam are: elasticity
modulus E = 35,700 MPa, Poisson’s ratio v = 0.1, tensile strength σ u = 3.6 MPa,
and fracture energy Gf = 184 N/m. The density of the material is assumed to be
2400 kg/m3 . Following an unsuccessful experimental attempt to simulate the self-
weight condition, in which an unstable failure occurred along the base of the model,
Fig. 3.1 Model test of the

concrete gravity dam with
the initial crack (Carpinteri
et al. 1992)
Fig. 3.2 Dam model, Scale 1:40, notch depth of 15 cm (dimensions in cm)
the repaired model as the second model has been tested without any adjustment of
the self-weight condition. In the present study, only the second model is analyzed,
and the predicted response of the single crack is compared with the documented
experimental and discrete crack analysis results of Carpinteri et al. (1992). The
crack mouth opening displacement (CMOD) at a rate of 1.2 µm/s is applied as the
control parameter in the present nonlinear analyses performed using the XFEM.
The hydraulic thrust was generated by means of a servo-controlled actuator with a
2000 kN capacity and applied to the upstream side. This force was distributed in four
concentrated loads whose intensity is indicated in Fig. 3.3. This force is gradually
increased until the failure of the dam.
Fig. 3.3 Total upstream face 1000

load versus CMOD Experimental result
(Carpinteri, 1992)
displacement 800 Cohesive crack model
Applied force (kN)
XFEM
600
400
200
0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
CMOD (mm)
3.2 Validation Test 57
Numerical results from the XFEM procedure Previous research results
Fig. 3.4 Numerical and experimental crack trajectories for the concrete gravity dam. a Numerical
results from the XFEM procedure; b Previous research results (Carpinteri et al. 1992; Barpi and
Valente 2000)
The relations between the total upstream face load and CMOD obtained from
the XFEM, cohesive crack model (Barpi and Valente 2000)and experimental results
(Carpinteri et al. 1992) are shown in Fig. 3.3. As is apparent from Fig. 3.3 that these
curves are close to each other. The predicted peak load of the numerical simulation
based on the XFEM agrees well with the experiment results.
Figure 3.4 shows the computed and experimental crack trajectories. It can be noted
that the occurring crack profile shows a good agreement with the experimental results
of the model test. By comparing the current results with previous research results
(Carpinteri et al. 1992; Barpi and Valente 2000), notice that the curved character of
the crack trajectory is correctly captured by the XFEM, indicating that the XFEM
procedure can predict effectively the crack propagation process in concrete gravity
dams with initial cracks.
3.3 Seismic Crack Propagation Analysis of Koyna Gravity

Dam with Initial Crack
3.3.1 Initial Cracking Models
In this section, the dynamic cracking analysis of Koyna dam is performed by using
the extended finite element method for the concrete material. The time history of the
Koyna earthquake is shown in Fig. 2.8 (Chap. 2). Finite element model for the tallest
section of the dam is shown in Fig. 3.5. The mesh of the dam is adequately refined
near the changes in the slope of the downstream faces, in which crack propagation is
expected. The reason for this is because damage due to tensile stresses is expected to
initiate near stress concentrations in those zones. The foundation of the dam is taken
as being rigid. Westergaard virtual mass (Westergaard 1933) is employed to include
the hydrodynamic effect. The value of the Westergaard virtual mass m i at node i on
the upstream surface of the dam is
Fig. 3.5 Finite element model of Koyna dam
7
m i = hyi ρw (bi1 + bi2 ) 2 (3.1)
8
where h is the depth of water; yi is the distance from node i to the water surface,
ρ w is the mass density of water; and bi1 and bi2 are the length of the edges of the
quadrilateral constant-strain elements beside node i on the upstream surface of the
dam.
It should be noted that the seismic water pressure effects inside the cracks are not
considered in the analysis. Further studies of the effect of seismic water pressure on
the crack propagation and dynamic response of the dam are deemed necessary.
In order to reveal the seismic behavior and the crack propagation process of the
cracked concrete gravity dams, the Koyna concrete gravity dam with three different
sets of initial cracks from the preassigned imperfection, located on the upstream and
downstream faces at the elevation of the downstream slope change, is assumed as
shown in Fig. 3.6. The initial crack at the base is also considered, but an initial crack
modeled at this location did not propagate during seismic analysis.
The depths of the initial cracks are all assumed to be 0.4 m. Their corresponding
XFEM discretizations are depicted in Fig. 3.6, and titled as Case I for the profile
with a horizontal crack at the upstream face (Crack C1), Case II for a horizontal
crack at the downstream face (Crack C2), Case III for an upstream-sloped crack at
the downstream face (Crack C3), Case IV for both the Crack C1 and Crack C2, Case
V for both the Crack C1 and Crack C3. A coefficient of friction of 0.7 is assumed
for all the cases to consider the effective interlock in the cracks, and the cohesion
coefficient of the crack is set at zero. The material parameters and loadings are
assumed as aforementioned.
It is evident that the model for the actual cracking of the dam is more likely to be
associated with crack C2 and crack C3. This is because the change in the slope of the
downstream face provides a singular point for the first crack formation. However,
3.3 Seismic Crack Propagation Analysis of Koyna Gravity Dam with Initial Crack 59
Fig. 3.6 Details of Koyna dam with different initial cracks
the initial cracks are not solely stress-induced but may also arise due to a variety of
other causes including shrinkage, temperature effects, etc. (Batta and Pekau 1996),
the additional three fracture models are included in the following examination of the
fracture process of Koyna dam.
3.3.2 Seismic Crack Propagation Process with no Initial

Crack
The crack propagation processes of the Koyna dam during the Koyna earthquake as
predicted using the XFEM procedures is shown in Fig. 3.7. As shown, smooth curva-
ture discrete cracks penetrating the elements are obtained. The fracture propagation
processes can be described as follows:
The initial crack in the dam is firstly observed at 3.87 s near the changes in the slope
of the downstream face (Fig. 3.7a). At this location, stresses are concentrated and the
tensile stresses take large values. As the vibration characteristics, the crack extends
deeper inside of the dam. After a period of 0.4 s, i.e. at t = 4.27 s, the downstream
crack propagates about three-fourth through the width of the dam section. The crack
trajectory curves down due to the compressive stresses resulting from rocking of
the top block (Fig. 3.7b–d). After the time instant t = 4.27 s, the crack propagates
horizontally toward to upstream face. At t = 4.48 s, the downstream crack extends
completely to the upstream face at a height of 56.7 m above the base, penetrating
the whole section of the dam. It can be noted that the crack propagation process is
mostly concentrated at the time from 3.87 to 4.48 s when the dynamic displacement
response of the dam crest is larger, and the dam remains its overall safety.
(a) t=3.87s (b) t=4.00s (c) t=4.27s (d) t=4.48s
Fig. 3.7 Processes of Koyna dam crack propagation at four selected times. a t = 3.87 s; b t =
4.00 s; c t = 4.27 s; d t = 4.48 s
3.3.3 Seismic Crack Propagation Process for Single Initial

Cracking Models
Dynamic crack propagation and response analysis of Koyna gravity dam with signal
initial crack under the 1967 earthquake are conducted employing the XFEM-based
cohesive segments method. Results for seismic crack propagation with initial cracks
C1, C2 and C3 considered separately are presented in Fig. 3.8, Figs. 3.9 and 3.10,
respectively.
Figure 3.8 shows the crack propagation processes of the Koyna dam with initial
crack C1. As shown, it is known that the final cracking profile with initial crack
C1 is quite different from those with the initial cracks C2 and C3. As the vibration
characteristics, the initial crack propagation in the dam is firstly observed at 3.20 s
near the initial crack C1. After a stretch of horizontal propagation, the crack profile
gradually curves downward due to the increasing compressive stresses on the down-
stream side. At t = 4.49 s, the crack extends about five-sixth through the width of
(a) t=3.20s (b) t=3.80s (c) t=4.15s (d) t=4.49s
Fig. 3.8 Crack propagation processes of Koyna dam for single initial crack C1(Case I) at four
selected times. a t = 3.20 s; b t = 3.80 s; c t = 4.15 s; d t = 4.49 s
(a) t=3.87s (b) t=3.90s (c) t=4.25s (d) t=4.49s
Fig. 3.9 Crack propagation processes of Koyna dam for single initial crack C2(Case II) at four
(a) t=3.87s (b) t=4.00s (c) t=4.27s (d) t=4.49s
Fig. 3.10 Crack propagation processes of Koyna dam for single initial crack C3(Case III) at four
the dam section, and then stops propagation. The final crack profile is presented in
Fig. 3.8d, and no penetrating crack appears in this case.
Dynamic crack propagation processes of Koyna dam with initial cracks C2 and
C3 are shown in Figs. 3.9 and 3.10, respectively. The final cracking profiles for the
initial cracks C2 and C3 models are more or less similar, with the initial cracks
breaking through to the opposite face of the dam to cause complete rupture, but the
crack trajectories are very different. The crack trajectory for the Case III is similar to
the Case that the dam with no initial crack (Fig. 3.7), the reason for this is because the
initial crack directions are similar. As the vibration characteristics, the initial crack
propagation in the dam with initial cracks C2 and C3 also first occurs at 3.87 s near
the discontinuity in the slope of the downstream face. After time t = 4.49 s, variation
in this damage is insignificant level. In the absence of other cracking, the initial
horizontal crack C2 of Fig. 3.9 propagates horizontally and emerges close to the
upstream face at the elevation near the changes in the downstream slope. Similarly,
Displacement (cm) 4
0
No initial crack
-2 Initial crack C1(Case I)
Initial crack C2(Case II)
-4 Initial crack C3(Case III)
-6
0 2 4 6 8 10
Time(sec)
Fig. 3.11 Time history graphs of the horizontal displacements of the point p1 at the dam crest
when only initial crack C3 is considered (Fig. 3.10), the crack extends toward the
upstream face along the initial crack direction.
The time history graphs of the horizontal displacement at the point P1 on the dam
crest for different crack models are given in Fig. 3.11, with positive displacement
in the downstream direction. In addition to Case I, the curves of the horizontal
displacement are basically similar in other cases. Although initial cracks near the
changes in the slope of the downstream face cause different cracking profiles, there
is only a slight difference in the horizontal displacement obtained with and without
initial crack. It is also seen from this figure that crack propagation is not necessarily
in phase with the displacement of the crest of the dam.
Figures 3.12 and 3.13 show the time history graphs of the crack opening displace-
ments, and the history of the sliding of the upper block of the dam after the formation
of the penetrated crack. The results obtained from different crack models have some
differences. These differences for the COD reach to an important level in point of
amplitude and frequency content (especially between t = 3.20 and 4.50 s) as the
cracks propagate in the dam. There is only crack opening/closing behavior in case I,
because no penetrated crack occurred in the dam. From the time history graphs of the
COD of Fig. 3.12, it is evident that the cracks remain more or less closed for almost
the entire time. As may be seen from Fig. 3.13, the maximum sliding displacement
toward the upstream direction is −3.88 mm at 4.58 s under Case III. The residual
displacement for the crack C2 is 0.35 mm toward the downstream direction due to the
water pressure, and the residual displacement for the crack C3 is −0.24 mm toward
the upstream direction due to the upstream-sloped crack. It may be concluded that
the upper block remains stable against overturning, as the residual displacement is
almost negligible.
3.3.4 Seismic Crack Propagation Process for Multiple Initial

Cracking Model
The results of the seismic crack propagation processes with multiple cracking models
are given in Figs. 3.14 and 3.15. As is to be expected, the above more complicated
18
10.08mm 17.10mm
10
15 Downstream
Upstream
Upstream
8
12
COD (mm)
COD (mm)
6 9
4 6
3.54mm
2 3
Initial crack Initial crack
0 0
Penetrated crack
0 2 4 6 8 10 0 2 4 6 8 10
Time (sec) Time (sec)
(a) (b)
14
13.70mm
12 Downstream
Upstream
10
COD (mm)
8.49mm
8
6
4
2 Initial crack
0
Penetrated crack
-2
0 2 4 6 8 10
Time (sec)
(c)
Fig. 3.12 Time history graphs of the crack opening displacement (COD) for single initial cracking
models. a Case I; b Case II; c Case III
2
1.45mm
1
0.35mm
0 -0.01mm
-0.24mm
-1
No initial crack
-2 Initial crack C2 (Case II)
Initial crack C3 (Case III)
-3 -3.28mm
-3.89mm
-4
0 2 4 6 8 10
Time (sec)
Fig. 3.13 Sliding displacement histories of the upper block of the dam with single initial crack
after the formation of the penetrated crack
cracking pattern associated with the multiple cracking model increases the compu-
tational effort. As shown, the crack propagation processes are more or less similar,
with time for both initial cracks beginning to propagate. In both the multiple cracking
models, the initial crack propagation in dam firstly occurs on the upstream face at
the time t = 3.20 s. After a stretch of horizontal propagation, the initial downstream
(a) t=3.20s (b) t=3.87s (c) t=4.00s (d) t=4.49s
Fig. 3.14 Crack propagation processes of Koyna dam for multiple initial crack C1 and C2(Case
IV) at four selected times. a t = 3.20 s; b t = 3.87 s; c t = 4.00 s; d t = 4.49 s
(a) t=3.20s (b) t=3.87s (c) t=4.15s (d) t=4.46s
Fig. 3.15 Crack propagation processes of Koyna dam for multiple initial crack C1 and C3(Case
V) at four selected times. a t = 3.20 s; b t = 3.87 s; c t = 4.15 s; d t = 4.46 s
crack begins to propagate toward the upstream face at the time t = 3.87 s. But crack
trajectories obtained from single and multiple initial cracking models have some
differences. These differences reach to an important level in point of processes as
the cracks propagate in the dam. For single initial cracking models, initial cracks
propagate through to the opposite face of the dam to cause serious damage. But the
initial cracks at the downstream and upstream faces extend into the dam for multiple
initial cracking models, nearly penetrating the whole section of the dam.
The time history graphs of the COD cracked at both the upstream and downstream
faces are given in Fig. 3.16, and show much more complexity due to the existence
of multi-cracks. It is seen from Fig. 3.16 that the cracks also remain more or less
closed for almost the entire time. Although the initial cracks have some effects on
the final cracking profile of the dam, there is only a slight difference in the time
history of the COD obtained with the multiple cracking models. Upstream crack C1
is first to propagate and experiences an increase in the magnitude of COD compared
to the corresponding single crack model. This crack propagation is associated with
12
11.47mm 12 12.27mm
10
8.75mm 10 Downstream
Downstream 8.38mm
8 Upstream
COD (mm)
Upstream
COD (mm)
8
6
6
4
4
2 Initial crack 2 Initial crack
0 0
Initial crack Initial crack
-2 -2
0 2 4 6 8 10 0 2 4 6 8 10
Time (sec) Time (sec)
(a) (b)
Fig. 3.16 Time history graphs of the crack opening displacement (COD) for multiple initial
cracking models. a Case IV; b Case V
relatively large crack opening displacement at approximately 4.17 s. After the time
t = 4.49 or 4.46 s, the cracks do not propagate, but the crack opening displacement
continues to increase. As shown in Fig. 3.16a, the peak of crack opening displacement
at the downstream end happens at 4.54 s, reaching 11.47 mm, and at the upstream
end happens at 4.29 s, reaching 8.75 mm. Figure 3.16b shows that the peak of
crack opening displacement at the downstream end also happens at 4.54 s, reaching
12.27 mm, and at the upstream end happens at 4.29 s, reaching 8.75 mm. Figure 3.17
shows the sliding displacement history of the upper block of the dam. As may be seen
from Fig. 3.17, the maximum sliding displacement toward the upstream direction is
−2.28 mm at 4.54 s under Case V.
2
1.16mm
1
0.19mm
0
-0.18mm
Initial crack C1 and C2 (Case (IV)
-1 Initial crack C1 and C3 (Case V)
-2 -2.28mm
0 2 4 6 8 10
Time (sec)
Fig. 3.17 Sliding displacement histories of the upper block of the dam with multiple initial crack
3.4 Seismic Crack Propagation Analysis of Guandi Gravity

Dams with Initial Cracks
3.4.1 FEM Model and Material Properties
The Guandi gravity dam is currently under construction in Southwest China. The
dam is located in a strong earthquake region with the design peak ground accelera-
tion (PGA) of 0.34 g. A typical non-overflow monolith of the dam, which is 142 m
high with a 138 m deep reservoir, is employed to model its seismic damage process.
Four-node plane strain quadrilateral isoparametric elements with 2 * 2 Gauss inte-
gration are utilized to discretize the rock foundation and dam structure as shown
in Fig. 3.18. Foundation corresponding to twice the dam height from the dam heel
to upstream, from the dam toe to downstream and from the dam bottom downward
is modeled with massless foundation. Since the presented research is aimed at the
seismic damage process of concrete gravity dams, the nonlinearity of the foundation
rock is not considered. In addition, Lagrangian approach is used for the finite element
modeling of dam-reservoir-foundation interaction problem. The material parameters
of the fluid are assumed as aforementioned. The energy dissipation of the monolith is
considered by the Rayleigh damping method with 5% damping ratio. The traditional
massless foundation approach is utilized herein to account for the dam-foundation
interaction. Applied loads include self-weight of the dam, hydrostatic, uplift, hydro-
dynamic, and earthquake forces. Static solutions of the dam due to its gravity loads
and hydrostatic loads are taken as initial conditions in the dynamic analyses of the
system. Three indices of concrete are employed, i.e. C15, C20, and C25. The material
properties are listed in Table 3.1. The tensile strength is taken as 10% of its compres-
sive counterpart. The elasticity modulus and Poisson’s ratio of the foundation rock
are taken as 21.6 GPa and 0.20, respectively. To account for the effect of strain rate,
modulus and strength of the dam and modulus of the foundation rock are increased
Dam-reservoir-foundation system Dam
Fig. 3.18 Finite element meshes of the dam-reservoir-foundation system

3.4 Seismic Crack Propagation Analysis of Guandi Gravity Dams … 67
Table 3.1 Material properties of the dam

Concrete Modulus Poisson’s Density Compressive Tensile Fracture
(GPa) ratio (kg/m3 ) strength (MPa) strength energy
(MPa) (N/m)
C15 56.0 0.167 2552 14.53 1.45 205
C20 57.6 0.167 2552 19.38 1.94 257
C25 58.8 0.167 2552 21.45 2.15 300
by 30% according to the Code for Seismic Design of Hydraulic Structures in China.
Both stream and vertical directions are subjected to earthquake excitation.
3.4.2 Initial Crack Position
In order to predict the locations of potential cracking, stress analysis of the intact
Guandi gravity dam subjected to the recorded Koyna earthquake with the design peak
ground acceleration (PGA) of 0.34 g as well as the static loads (concrete dead weight
plus reservoir hydrostatic pressure) is first performed. Envelope of the resulting
principal tensile stresses for the dam is presented in Fig. 3.19. It can be seen from
that the zones of high tensile stress exceed the tensile strength of concrete are mainly
observed on the downstream surface (near the change in downstream slope) and the
upstream surface (near the slope change at two heights of 8 and 48 m above the base,
the dam heel and the upper part of the upstream surface). On the slope change and
the dam heel, due to the stress concentration, the stress singularity results in a high
computed tensile stress almost three times the tensile strength of concrete.
Fig. 3.19 Envelope of

principal tensile stresses for
seismic analysis without
cracks
Based on the above, the Guandi concrete gravity dam with six different sets of
small initial cracks from the pre-assigned imperfection, located on the upstream and
downstream faces, is assumed as shown in Fig. 3.20. The initial crack at the dam
heel is also considered, but an initial crack modeled at this location did not propagate
during seismic analysis. The seismic performance and crack propagation process of
the cracked concrete gravity dams are investigated for each of these six models of
cracking. In order to examine the influence of the initial crack position on the crack
propagation trajectory, the depths of the initial cracks are all assumed to be 0.4 m,
which is introduced separately on the downstream and the upstream faces of the dam.
Their corresponding XFEM discretizations are depicted in Fig. 3.18, and titled as
Case I for an intact profile with no initial crack, Case II an upstream-sloped crack at
the downstream face (Crack C1), Case III for the profile with a horizontal crack at
the upstream face near the change in the slope of the face at a height of 8 m above
the base (Crack C2), Case IV for a horizontal crack at the upstream face near the
slope change at a height of 48 m above the base (Crack C3), Case V, VI and VII
for a horizontal crack on the upstream face at the a height of 123.35 m, 110 m and
95 m above the base (Crack C4, Crack C5 and Crack 6), respectively. A coefficient
of friction of 0.7 is assumed for all the cases to consider the effective interlock in the
cracks, and the cohesion coefficient of the crack is set at zero.
It is evident that the model for the actual cracking of the dam is more likely to
be associated with crack C1, C2 and crack C3. This is because the change in the
slope of the downstream and upstream faces provides a singular point for the first
crack formation. However, the initial cracks are not solely stress-induced but may
also arise due to a variety of other causes including shrinkage, temperature effects,
etc., the additional three fracture models are included in the following examination
of the fracture process of the dam.
Fig. 3.20 Details of Koyna

dam with different initial
cracks
(a) t=4.18s (b) t=4.24s (c) t=4.55s (d) t=5.56s
Fig. 3.21 Crack propagation processes of the dam with no initial crack (Case I) at four selected
times. a t = 4.18 s; b t = 4.24 s; c t = 4.55 s; d t = 5.56 s
3.4.3 Crack Propagation Process of the Dam with no Initial

Crack
In order to investigate the influence of initial cracks on the seismic performance and
crack propagation of concrete gravity dams, dynamic crack process and response
analysis of the intact Guandi gravity dam under the 1967 earthquake with the design
peak ground acceleration (PGA) of 0.34 g are firstly conducted employing the XFEM-
based cohesive segments method.
Figure 3.21 shows the crack propagation processes of the intact dam (Case I) as
predicted using the XFEM procedures. As shown, no penetrating crack appears in
this case, and the failure mechanism is formed of two main damage zones, one near
the dam heel and one at the change in the downstream slope. An initial crack in the
dam is predicted to initially occur near the changes in the slope of the downstream
face at 4.18 s. At this location, stresses are concentrated and the tensile stresses take
large values. As the vibration characteristics, the crack extends deeper inside of the
dam at approximately a 45° angle to the vertical. The crack trajectory curves down
due to the compressive stresses resulting from rocking of the top block. Furthermore,
an initial crack in the dam heel is observed at 4.24 s, which is probably due to stress
concentration. It can be noted that the crack at the slope change stops expanding after
5.56 s, and the crack extends into the dam about 10 m.
3.4.4 Influence of Initial Crack Position
Seismic crack propagation analyses of the dam with initial cracks C1–C6 considered
separately are presented in Figs. 3.22, 3.23, 3.24, 3.25, 3.26 and 3.27, respectively.
It can be seen from that the initial crack positions have significant influence on
the seismic performance and crack propagation processes of concrete gravity dams,
which will cause more severe damage to the dam body than the intact dam profile.
(a) t=3.42s (b) t=4.24s (c) t=4.60s (d) t=4.63s
Fig. 3.22 Crack propagation processes of the dam for single initial crack C1 (Case II) at four
(a) t=4.26s (b) t=4.61s (c) t=4.67s (d) t=5.00s
Fig. 3.23 Crack propagation processes of the dam for single initial crack C2 (Case III) at four
(a) t=4.21s (b) t=4.62s (c) t=4.66s (d) t=4.87s
Fig. 3.24 Crack propagation processes of the dam for single initial crack C3 (Case IV) at four
In some cases, cracks extend completely to the opposite face, penetrating the whole
section of the dam and separating the crest from the upper part of the dam.
Figure 3.22 shows the crack propagation processes of the Guandi gravity dam
with initial crack C1. As shown, the crack propagation processes for the Case II
are more or less similar to the Case I that the dam with no initial crack as shown in
Fig. 3.21. But smooth curvature discrete cracks penetrating the elements are obtained
in the Case II. At t = 3.42 s, the initial crack C1 near the changes in the slope of the
(a) t=4.24s (b) t=4.26s (c) t=4.40s (d) t=4.83s
Fig. 3.25 Crack propagation processes of the dam for single initial crack C4 (Case V) at four
(a) t=4.24s (b) t=4.26s (c) t=4.30s (d) t=4.81s
Fig. 3.26 Crack propagation processes of the dam for single initial crack C5 (Case VI) at four
selected times. (a) t = 4.24 s; (b) t = 4.26 s; (c) t = 4.30 s; (d) t = 4.81 s
(a) t=4.24s (b) t=4.26s (c) t=4.30s (d) t=4.68s
Fig. 3.27 Crack propagation processes of the dam for single initial crack C6 (Case VII) at four
downstream face is beginning to propagate almost perpendicular to the downstream

surface. With the on-going acceleration excitation, the crack extends deeper inside of
the dam. The crack trajectory curves down due to the compressive stresses resulting
from rocking of the top block (Fig. 3.22a–c). After a period of 1.18 s, i.e. at t =
4.60 s, the downstream crack propagates about four-fifth through the width of the
dam section (Fig. 3.22c). After the time instant t = 4.60 s, the crack propagates
horizontally toward to upstream face. At t = 4.48 s, the downstream crack extends

completely to the upstream face at a height of 116.2 m above the base, penetrating
the whole section of the dam (Fig. 3.22d).
Dynamic crack propagation processes of the dam with initial cracks C2 and C3 are
shown in Figs. 3.23 and 3.24, respectively. The crack propagation processes and the
final cracking profiles for the initial cracks C2 and C3 models are more or less similar,
in which a crack is observed to initially occur near the change in the slope of the
downstream face and then extends into the dam, and three damage zones (the change
in downstream and upstream slopes and near the dam heel) are clearly identified. For
the initial crack C2 model (Case III), the initial crack C2 breaks through to the base of
the dam to cause complete rupture (Fig. 3.23c). In addition, one crack develops near
the change in downstream slope and another crack extends near the change in the
upstream slope (Fig. 3.23). For the initial crack C3 model (Case IV), the initial crack
C3 at the change in the slope of the upstream face continues to grow downwards
toward downstream in a slightly inclined direction (about 10° to the horizontal) with
the continuous vibration of the dam. The crack propagates up to about 27.8 m and
then stops propagation (Fig. 3.24d). The crack propagation depth at this location for
Case IV is longer than the Case III.
The results of the seismic crack propagation processes of the dam with initial
cracks C4, C5, and C6 are given in Figs. 3.25, 3.26 and 3.27. As shown, it is known
that the final cracking profiles with initial cracks C4, C5, and C6 are quite different
from those with the initial cracks C1, C2, and C3. The final cracking trajectories
for the initial cracks C4, C5 and C6 models are more or less similar, with the initial
cracks breaking through to the opposite face of the dam to cause serious damage. For
the initial crack C4 model (Case V), a crack in the upper of the dam upstream face
propagates downwards toward downstream in a slightly inclined direction and finally
reaches the downstream face, penetrating the whole section of the dam (Fig. 3.25). In
Case VI (the initial crack C5 model), the initial crack C5 approximately horizontal
toward the downstream face, separates the crest from the upper part of the dam
(Fig. 3.26). For the initial crack C5 model (Case VII), the initial crack propagation
in the dam is firstly observed at 4.26 s near the initial crack C5 due to the vibration
characteristics. After a stretch of horizontal propagation, the crack profile gradually
curves downward due to the increasing compressive stresses on the downstream side.
At t = 4.68 s, the crack extends about five-sixth through the width of the dam section,
and then stops propagation. The final crack profile is presented in Fig. 3.27d, and no
penetrating crack appears in this case.
3.4.5 Influence of Initial Crack Length
The finite element model (Fig. 3.18) with initial cracks at the downstream face (Crack
C1) and the upstream face near the slope change at a height of 48 m above the base
(Crack C3) is analyzed with four values of initial crack lengths, 0.2 m, 0.4 m, 1.0 m
and 2.0 m, to study the sensitivity of the predicted response to this initial parameter.
Finial failure patterns of the dam with different initial crack lengths are shown in
Figs. 3.28 and 3.29. A comparison of horizontal displacement time history of the dam
crest with that of no initial crack is shown in Fig. 3.30, with positive displacement
in the downstream direction.
The influence of initial crack lengths on the crack propagation trajectory can be
addressed by comparing Figs. 3.21d, 3.28 and 3.29. It can be found that the initial
crack length has some influence on the crack propagation depths. When the initial
crack length at the downstream face is 0.2 m, there is no penetrating crack. With the
increase of the initial crack length to 0.4 m, the downstream crack extends completely
to the upstream face, penetrating the whole section of the dam. The ultimate failure
profiles are very similar for the initial crack length of 0.4 m, 1.0 m and 2.0. However,
(a) Length=0.2m (b) Length=0.4m (c) Length=1.0m (d) Length=2.0m
Fig. 3.28 Final crack profiles of the dam with different initial crack lengths at the downstream
face: a length = 0.2 m; b length = 0.4 m; c length = 1.0 m; b length = 2.0 m
(a) Length=0.2m (b) Length=0.4m (c) Length=1.0m (d) Length=2.0m
Fig. 3.29 Final crack profiles of the dam with different initial crack lengths at the upstream face
near the slope change at a height of 48 m above the base: a length = 0.2 m; b length = 0.4 m;
c length = 1.0 m; d length = 2.0 m
Fig. 3.30 Time history 6

No initial carck
graphs of the horizontal Length=0.4m
Displacement (cm)
4
displacements of the dam Length=2.0m
crest with initial cracks at 2
a the downstream face; b the 0

upstream face near the slope
-2
change at a height of 48 m
above the base -4
0 2 4 6 8 10
Time (s)
(a)
5
No initial crack
4 Length=0.4m
Displacement (cm)
Length=2.0m
3
2
1
0
-1
-2
0 2 4 6 8 10
Time (s)
(b)
there is a significant impact on the horizontal displacement response of the dam with
different initial crack lengths at the downstream face as shown in Fig. 3.30a.
As observed by comparing Figs. 3.21d and 3.29, it can be found that the position
of the initial crack has significant influence on the crack propagation process of
concrete gravity dams. While the crack trajectories are very similar for different
initial crack lengths located in the upstream face near the slope change at a height
of 48 m above the base, the reason for this is because the initial crack directions
are similar. The crack propagation depths at the upstream face for the initial crack
length of 0.2 m, 0.4 m, 1.0 m and 2.0 m are 22.75 m, 27.80 m, 28.63 m and 31.02 m,
respectively. Although the initial cracks at the upstream face near the slope change
cause different cracking depths, there is only a slight difference in the horizontal
displacement obtained with different initial crack lengths.
3.5 Conclusions
The objective of this study is to evaluate the influence of the initial crack position and
length on the seismic performance and crack propagation of concrete gravity dams
with considering the effects of dam-reservoir-foundation interaction. The reservoir
water is modeled using two-dimensional fluid finite elements by the Lagrangian
3.5 Conclusions 75
approach. The extended finite element method (XFEM), which is based on the cohe-
sive segments method in conjunction with the phantom node technique, is used to
model the cracked concrete gravity dam and predict the crack propagation process.
The performance of the XFEM procedure for the analysis of crack propagation in
concrete gravity dams with initial cracks has been demonstrated in this work. For
this purpose, cracking processes of a scaled-down 1:40 model of a gravity dam with
an initial notch on the upstream wall loaded with an equivalent hydraulic load are
analyzed for accuracy verification. The results show that the crack profile match well
with the experimental results from the model test, indicating that the XFEM proce-
dure can effectively capture the crack propagation processes and the crack trajectory
in concrete gravity dams with initial cracks.
Seismic cracking analysis of the cracked concrete gravity dam indicates that the
initial cracks are important in the estimation of the seismic response and crack prop-
agation process. The crack propagation processes for the multiple cracking models
are considerably different those for the single initial crack models. In addition to
single initial crack in the upstream face of the dam, other possible patterns of initial
cracks, including multiple initial on both upstream and downstream face of the dam,
result in nearly penetrating cracks in the dam.
Seismic cracking analyses of the Guandi gravity dam with six different sets of
small initial cracks at different locations along the upstream and downstream faces are
performed. The results with and without initial cracks are compared, and significant
differences in terms of the crack profile are observed, which indicate that the initial
crack position has significant influence on the crack propagation process of concrete
gravity dams. The cracked dam will cause more severe damage to the dam body than
the intact dam profile. The critical cracking is found to be associated with fracture
originating at the point of downstream slope change and penetrating the dam almost
instantaneously to separate the crest from the upper part of the dam. Other possible
patterns of initial cracks on the upstream face of the upper of the dam also result in
a complete rupture. The influence of the initial cracks with different lengths at the
downstream and upstream faces is also discussed. The results show that the initial
crack length has some impact on the crack propagation depths and displacement
response. The crack trajectories are very similar for cases with different initial crack
lengths.
References
Barpi, F., & Valente, S. (2000). Numerical simulation of prenotched gravity dam models. Journal
of Engineering Mechanics, 126(6), 611–619.
Batta, V., & Pekau, O. A. (1996). Application of boundary element analysis for multiple seismic
cracking in concrete gravity dams. International Journal of Rock Mechanics and Mining Sciences
and Geomechanics Abstracts, 33(6), A277.
Bolzon, G. (2004). LEFM and cohesive-crack approaches to safety evaluation of concrete gravity
dams. In Computational Mechanics WCCM VI in conjunction with APCOM’04, Beijing (pp. 1–5).
Carpinteri, A., Valente, S., Ferrara, G., & Imperato, L. (1992). Experimental and numerical fracture
modelling of a gravity dams. In Fracture Mechanics of Concrete Structures, Proceedings of the
First International Conference on Fracture Mechanics of Concrete Structures, Elsevier Applied
Science, Breckenridge, CO (pp. 351–360).
Feng, L. M., Pekau, O. A., & Zhang, C. H. (1996). Cracking analysis of Arch Dams by 3D boundary
element method. Journal of Structural Engineering, 122(6), 691–699.
Hall, J. F. (1988). The dynamic and earthquake behaviour of concrete dams: review of experimental
behaviour and observational evidence. Soil Dynamics and Earthquake Engineering, 7(2), 58–121.
Ingraffea, A. R. (1990). Case studies of simulation of fracture in concrete dams. Engineering
Fracture Mechanics, 35(1–3), 553–564.
Javanmardi, F., Léger, P., & Tinawi, R. (2005). Seismic structural stability of concrete gravity dams
considering transient uplift pressures in cracks. Engineering Structures, 27(4), 616–628.
Jiang, S., Du, C., & Hong, Y. (2013). Failure analysis of a cracked concrete gravity dam under
earthquake. Engineering Failure Analysis, 33, 265–280.
Li, Q. S., Li, Z. N., Li, G. Q., Meng, J. F., & Tang, J. (2005). Experimental and numerical seismic
investigations of the Three Gorges dam. Engineering Structures, 27(4), 501–513.
Oliver, J., Huespe, A. E., Pulido, M. D. G., & Chaves, E. (2002). From continuum mechanics to
fracture mechanics: the strong discontinuity approach. Engineering Fracture Mechanics, 69(2),
113–136.
Pekau, O. A., & Zhu, X. (2006). Three-degree-of-freedom rigid model for seismic analysis of
cracked concrete gravity dams. Journal of Engineering Mechanics, 132(9), 979–989.
Rochon-Cyr, M., & Léger, P. (2009). Shake table sliding response of a gravity dam model including
water uplift pressure. Engineering Structures, 31(8), 1625–1633.
Shi, M., Zhong, H., Ooi, E. T., Zhang, C., & Song, C. (2013). Modelling of crack propagation of
gravity dams by scaled boundary polygons and cohesive crack model. International Journal of
Fracture, 183(1), 29–48.
Shi, Z., Suzuki, M., & Nakano, M. (2003). Numerical analysis of multiple discrete cracks in concrete
dams using extended fictitious crack model. Journal of Structural Engineering, 129(3), 324–336.
Tinawi, R., Léger, P., Leclerc, M., & Cipolla, G. (2000). Seismic safety of gravity dams: From snake
table experiments to numerical analyses. Journal of Structural Engineering, 126(4), 518–529.
Westergaard, H. M. (1933). Water pressures on dams during earthquakes. Transactions ASCE, 95(2),
418–433.
References 77
Zhang, S., Wang, G., & Yu, X. (2013). Seismic cracking analysis of concrete gravity dams with
Zhu, X., & Pekau, O. A. (2007). Seismic behavior of concrete gravity dams with penetrated cracks
and equivalent impact damping. Engineering Structures, 29(3), 336–345.
Chapter 4
Seismic Potential Failure Mode Analysis
of Concrete Gravity
Dam–Water–Foundation Systems
Through Incremental Dynamic Analysis
4.1 Introduction
When subjected to strong ground motions, mass concrete dams are likely to experi-
ence cracking due to the low tensile resistance of concrete. Meanwhile, the potential
crack initiation and propagation would adversely affect the static and dynamic perfor-
mance of dams. As cracks penetrate deep inside a dam, its structural resistance may be
considerably weakened, thereby endangering the safety of the dam. While there are
many high concrete dams throughout the world, only some of them have experienced
ground shaking induced structural damage. To name just a few, the Hsingfengkiang
dam, China, 1962; Koyna dam, India, 1967; and Sefid-Rud dam, Iran, 1990 are the
ones that have suffered from damage in earthquakes.
The shaking table test, which can reproduce the seismic response of structures
to seismic loadings, is one method to investigate the seismic failure process of high
concrete dams. However, many issues need to be clarified for the model test. Among
them, the problem of similarity relation is the most difficult, particularly for the
seismic failure process of high concrete dams. Apart from the shaking table test,
many nonlinear models have also been developed in order to assess the seismic safety
of concrete dams. There are based either on the discrete crack approach (Ingraffea
and Saouma 1985; Ayari and Saouma 1990; Ahmadi et al. 2001) or the smeared
crack approach (Bhattacharjee and Léger 1994; Léger and Leclerc 1996; Ghaemian
and Ghobarah 1999; Wang et al. 2000). In addition, other models, such as the plastic-
damage model, extended finite element method (XFEM), discrete element method
(DEM), discontinuous deformation analysis (DDA), are also used in some cases to
analyze the seismic failure behavior of concrete dams.
Without being exhaustive, some contributions to these methods and their applica-
tions in the seismic crack analysis of concrete dams are worth mentioning. Lee and
Fenves (1998) developed a new plastic-damage constitutive model for earthquake
analysis of concrete dams, and Long et al. (2009) used this model to study the seismic
damage mechanics of the gravity dam and evaluate the effect of the reinforcement.
https://doi.org/10.1007/978-981-15-6194-8_4
80 4 Seismic Potential Failure Mode Analysis of Concrete …
Calayir and Karaton (2005a, b) presented a continuum damage model and a co-axial
rotating crack model for earthquake damage response analysis of concrete gravity
dams with the effects of dam-reservoir interaction considered. Zhang et al. (2013a,
b) investigated the influences of strong motion duration and mainshock-aftershock
seismic sequences on the accumulated damage of concrete gravity dams using a
concrete damage plasticity (CDP) model. Omidi et al. (2013) examined the seismic
fracture responses of concrete gravity dams due to constant and damage-dependent
damping mechanisms. They used a plastic-damage model to simulate the irreversible
damage occurring in the fracturing process of concrete. Pan et al. (2011) compared
the cracking process and profiles of concrete dams using the XFEM with cohesive
constitutive relations, crack band finite element method with plastic-damage rela-
tions, and finite element Drucker-Prager Elasto-plastic model. Zhang et al. (2013c)
presented the XFEM for analyzing the seismic crack propagation of concrete gravity
dams with initial cracks at the upstream and downstream faces. Zhong et al. (2011)
extended the Rock Failure Process Analysis (RFPA) to study the failure process
of high concrete dams subjected to strong earthquakes. They presented the typical
failure process and failure modes of concrete gravity dams with considering the
uncertainty in the ground motion input and concrete material. Pekau and Cui (2004)
used the DEM to study the seismic behavior of the fractured Koyna dam during
earthquakes. Their results showed that the safety of the dam is ensured if the crack
shape is horizontal or upstream-sloped, and it is very dangerous if the crack slopes
downstream. Bretas et al. (2014) developed the DEM to investigate failure mecha-
nisms in masonry gravity dams. Mirzayee et al. (2011) proposed a hybrid distinct
element-boundary element approach to model the nonlinear seismic behavior of
cracked concrete gravity dams considering dam-reservoir interaction effects. Wang
et al. (2006) employed the DDA to study the seismic stability of the upper part of the
fractured Koyna dam. Das and Cleary (2013) explored a mesh-free particle method
called smoothed particle hydrodynamics (SPH) for the modeling of gravity dam
failure subjected to fluctuating dynamic earthquake loads.
Although all the aforementioned research investigated the failure process of dams
under some specific seismic ground excitations, only a quite limited number of them
have been performed to generalize the seismic failure modes of concrete dams, espe-
cially from a large and reliable database of ground motions. In this contribution,
we analyze the complicated seismic failure process of the Guandi concrete dam
located in China, which involves lots of mechanisms such as elastic deformation and
initiation and propagation of cracks leading to ultimate failure. The XFEM, which
was originally proposed by Belytschko and Black (1999), is employed to investigate
the nonlinear dynamic behavior of concrete gravity dams. Moreover, the cohesive
segments method (Remmers et al. 2008) in conjunction with the phantom node
technique (Hansbo and Hansbo 2004; Song et al. 2006) is introduced to the XFEM
framework in order to simulate crack initiation and propagation along arbitrary paths,
This strategy provides great flexibility and versatility in the numerical modeling of
discontinuities, since the finite element mesh conforming to cracks is not required
and the remeshing during crack growth is avoided in this strategy. These advan-
tages are attributed to the enrichment functions added to the standard finite element
4.1 Introduction 81
approximation, which is based on the partition of unity concept. Furthermore, the

interaction between the impounded water and the dam-foundation system is also
explicitly taken into account by modeling the reservoir water with two-dimensional
fluid finite elements in the Lagrangian formulation. That is to say, the presented
XFEM framework is applied in the fluid-solid coupled systems for solid (concrete
dam) modeling.
More importantly, the presented study also attempts to generalize the potential
failure modes of concrete gravity dams from a large database of ground motions.
Considering the uncertainty in the ground motion input, 40 as-recorded accelero-
grams with each scaled to 8 increasing intensity levels are selected as seismic
excitations. Seismic cracking damage process and dynamic response of concrete
gravity dams including the effect of dam-reservoir-foundation interaction are simu-
lated through the incremental dynamic analysis (IDA) (Alembagheri and Ghaemian
2013a, b; Billah and Alam 2014; Nik Azizan et al. 2018; Sotoudeh et al. 2018; Chen
et al. 2019) based XFEM. The typical failure process and potential failure modes of
concrete gravity dams under strong earthquake ground motions are presented.
4.2 Nonlinear Dynamic Response of Guandi Dam Under

Design Peak Ground Acceleration
4.2.1 FEM Model and Material Properties
In this section, the seismic potential failure modes of the Guandi concrete gravity
dam-reservoir-foundation system are studied. The Guandi gravity dam with a design
PGA of 0.34 g is located on the Yalong River in the Sichuan Province of China.
A typical non-overflow monolith with a height of 142 m is employed to model its
seismic damage process. The finite element model of the dam-reservoir-foundation
interaction system is shown in Fig. 4.1. It should be noted that the upstream face
near the dam heel and the downstream face near the dam toe have been simplified as
compared with Sect. 3.4.1.
The Lagrangian approach is used for the finite element modeling of dam-reservoir-
foundation interaction problem. For the initial time step, displacements of nodes on
the left and right truncated boundary of the dam-reservoir-foundation system are
assumed to be zero in the normal direction. In addition, the base of the foundation
is fully constrained. For the sequent dynamic analysis, all displacement constraints
are released, and the horizontal and vertical components of the selected earthquake
accelerations are applied to the foundation base as the input loading. The material
parameters of the fluid are assumed as aforementioned. The traditional massless
foundation approach is utilized herein to account for the dam-foundation interaction.
Since the presented research is aimed at the seismic damage processes of concrete
gravity dams, nonlinearity of the foundation rock is not considered. The energy
(a) Dam-reservoir-foundation system (b) Guandi gravity dam
Fig. 4.1 Finite element meshes of the Guandi graivty dam-reservoir-foundation system. a Dam-
reservoir-foundation system; b Guandi gravity dam
dissipation of the monolith is considered by the Rayleigh damping method with 5%

damping ratio.
Three indices of concrete are employed, i.e. C15, C20 and C25. The material
properties of the dam concrete are listed in Table 3.1. The tensile strength is taken
as 10% of its compressive counterpart. The elasticity modulus and Poisson’s ratio of
the foundation rock are taken as 21.6 GPa and 0.20, respectively. To account for the
effects of strain rate, modulus and strength of the dam and modulus of the foundation
rock are increased by 30% according to the Code for Seismic Design of Hydraulic
Structures in China. Both horizontal direction and vertical direction are subjected to
earthquake excitation.
4.2.2 Seismic Response and Crack Propagation Analysis
In this section, the effects of the concrete cracking damage on the seismic response of
the dam-reservoir-foundation system under design peak ground acceleration (PGA)
of 0.34 g are investigated by comparison of the linear and nonlinear dynamic solutions
of the system. The linear solution is obtained by assuming the concrete to be linear
elastic while the nonlinear dynamic solution of the dam is obtained from the XFEM-
based cohesive segments method. Namely, the concrete cracking damage resulting
from the seismic ground motion is considered in the nonlinear solution. Besides
the effects of the concrete cracking damage, we also analyze and discuss the full
process of dynamic cracking of the Guandi dam. The Koyna earthquake, which is
scaled to have 0.34 g spectral acceleration in the stream direction, is used as seismic
excitation. The implicit Newmark-β method is used in the incremental dynamic
analysis for time stepping, which is unconditionally stable. That is to say, any time
4.2 Nonlinear Dynamic Response of Guandi Dam Under Design Peak … 83
step will lead to stable results. However, in order to ensure accurate time integration,
sufficiently small time increment is needed. On the other hand, excessive time steps
result in high burden and poor computational efficiency. In practice, an adequate
balance between accuracy and efficiency can be obtained by choosing the following
combination: the initial increment size and maximum increment size are 0.001 s
and 0.01 s, respectively. Meanwhile, automatic time incrementation is used for the
dynamic analysis to further improve the efficiency.
Figure 4.2 compares the history of horizontal and vertical displacements at the dam
crest obtained from the linear elastic method and the XFEM, respectively, when the
Guandi dam is under the combined static and earthquake loads. The solid lines are the
response of the dam using the XFEM, whereas the dashed lines are the dam response
assuming the concrete is linear elastic. The positive directions of the horizontal and
vertical displacement are in the downstream and upward directions, respectively. As
can be seen in Fig. 4.2, the nonlinear dynamic response before cracking initiation
(1.90 s) is coincident with the response calculated from the linear elastic analysis.
This means that the maximum principal stress in the dam does not exceed the tensile
strength of the concrete during the relatively small amplitude motion, and there is no
cracking before t = 1.90 s. The subsequent displacements obtained from the linear
and nonlinear solutions separately with each other as the cracks form and propagate
in the dam. It is clear from Fig. 4.2 that the nonlinear response has a substantially
different displacement history than the linear model after t = 3.67 s. The vibration
periods of displacement response are also changed by the crack propagation, and a
lengthened vibration period from the XFEM procedure is found. It implies that the
crest displacements are dominated by rigid–body rocking of the upper portion of the
dam after the formation of cracks near the change in the slope of the downstream
face. At all times, the nonlinear response analysis shows that the dam remains stable.
Figures 4.3 and 4.4 present the time history graphs of the crack opening displace-
ments, and the history of the sliding of the upper block of the dam after the formation
of the penetrated crack. As shown in Fig. 4.3, the peak of crack opening displace-
ment at the downstream end happens at 4.32 s, reaching 1.79 cm, and at the upstream
end happens at 3.92 s, reaching 0.77 cm. It can also be found from Fig. 4.3 that
the crack opening displacement at the downstream end is dominant compared with
1.0
Linear
2 1.91cm 2.00cm 0.92cm Linear
Displacement (cm)
Displacement (cm)
Non-linear Non-linear
0.5
0.28cm
1
0.0
0
-0.5
-0.19cm
-1 -1.0
-0.99cm
-2 -1.97cm -1.5
-1.60cm
0 2 4 6 8 10 0 2 4 6 8 10
Time(s) Time(s)
(a) (b)
Fig. 4.2 Time history graphs of the horizontal and vertical displacements of the point P1 at the
dam crest (a) Horizontal displacement and (b) Vertical displacement
2.0
1.79cm Downstream
Upstream
1.5
COD (cm)
1.0 0.77cm
0.5
0.0
Initial crack Penetrated crack
-0.5
0 2 4 6 8 10
Time (s)
Fig. 4.3 Time history graphs of the crack opening displacement (COD)
0.5
0.26cm
Sliding along the crack (cm)
0.0
-0.5
Penetrated crack
-1.0
-1.5
-2.0
-2.08cm
-2.5
0 2 4 6 8 10
Time (s)
Fig. 4.4 Sliding displacement histories of the upper block of the dam after the formation of
the penetrated crack
that at the upstream end. The reason for this is mainly because the upstream-sloped
crack, which increases the resistance against downstream sliding of the upper block,
makes the upper block easier to rotate toward the upstream direction rather than
downstream direction under seismic conditions. As may be seen from Fig. 4.4, the
maximum sliding displacement toward the downstream direction is 0.26 cm at 3.69 s,
and toward the upstream direction is −2.08 cm at 3.92 s. It may be concluded that
when the Koyna earthquake scaled to have 0.34 g spectral acceleration in the stream
direction causes the penetrated crack damage to the dam, the upper block remains
stable against overturning, as the maximum sliding displacement of 2.08 cm is almost
negligible, and the residual displacement is very small.
Crack propagation processes in the Guandi gravity dam at four selected times as
predicted using the XFEM procedure are shown in Fig. 4.5. As shown, the smooth
curvature discrete crack penetrating the elements is obtained. Damage initiates from
4.2 Nonlinear Dynamic Response of Guandi Dam Under Design Peak … 85
(a) t=1.90s (b) t=2.59s (c) t=3.67s (d) t=3.98s
Fig. 4.5 Processes of the crack propagation in the Guandi gravity dam at four selected times. a t
= 3.87 s; b t = 4.00 s; c t = 4.27 s; d t = 4.48 s (Amplification factor: 100)
regions with very high stress concentrations. An upstream-sloped in the dam is initi-
ated at the discontinuity in the slope of the downstream face at about t = 1.90 s
(Fig. 4.5a). At this location, stresses are concentrated and the tensile stresses take
large values. As the vibration characteristics, the crack extends deeper inside of the
dam. The initial crack propagates almost perpendicular to the downstream surface.
The crack trajectory curves down due to the compressive stresses resulting from
rocking of the top block (Fig. 4.5a, b). At t = 2.59 s, the upper crack propagates
to about three-fifth through the thickness of the dam section (Fig. 4.5b). After the
time instant t = 2.59 s, the crack propagates approximately horizontally toward
the upstream face. At t = 3.67 s, the downstream crack extends completely to the
upstream face at a height of 118.4 m above the base, penetrating the whole section
of the dam (Fig. 4.5c). Because of the thrust of impounded water which is opposing
the tendency of the top section to slide along the crack in the upstream direction, the
computed crack profiles in the upper part of the dam can be considered neutral or
favorable conditions to maintain stability.
4.3 Seismic Potential Failure Mode Analysis
4.3.1 Database of as-Recorded Acceleration Records
In order to investigate the potential failure modes of concrete gravity dam-reservoir-

foundation systems subjected to strong ground motions, 40 real earthquake records,
which have been selected from the databases of the Consortium of Organizations
for Strong Motion Observation Systems (COSMOS) (COSMOS 2019) and Pacific
Earthquake Engineering Research Center (PEER) (PEER 2019) with a broad distribu-
tion of durations, are applied to the structure. The seismic sequences with different
earthquake magnitudes (i.e. M w 6.0–7.8) that have occurred in the United States,
Taiwan and India, are shown in Table 4.1. The data sample includes well-known
earthquakes with strong seismic activity, such as Loma Prieta (1989), Northridge
Table 4.1 Earthquake records used in the failure mode analysis
86
No. Earthquake Date Distance to fault (km) Station location Mw Comp. PGA (cm/s2 )
1 Koyna 11-12-1967 13.0 Koyna Dam 6.5 0 464.78
2 Northridge 17-01-1994 12.5 Coldwater Canyon School #5309 6.7 180 313.30
3 Northridge 17-01-1994 35.8 Los Angeles Terrace #24592 6.7 180 310.10
4 Northridge 17-01-1994 15.7 Epiphany Litheran Church #5353 6.7 196 404.20
5 Northridge 17-01-1994 18.4 Los Angeles Mulholland Dr #5314 6.7 35 641.90
6 Northridge 17-01-1994 8.6 Los Angeles Dam #2141 6.7 334 419.10
7 Northridge 17-01-1994 12.9 CA—White Oak Covenant Church #5303 6.7 180 467.90
8 Northridge 17-01-1994 7.1 Newhall, CA—Los Angeles County Fire 6.7 360 578.20
#24279
9 Northridge 17-01-1994 8.6 Sylmar, CA—Jensen Filtration Plant #655 6.7 22 560.30
10 Northridge 17-01-1994 29.4 Warm Springs #24272 6.7 90 221.20
11 Northridge 17-01-1994 17.9 Tarzana, CA—Cedar Hill #24436 5.3 90 365.30
12 Chi-Chi 20-09-1999 8.3 Taichung, Taiwan #TCU078 7.6 0 301.80
15 Chi-Chi 20-09-1999 32.0 Ilan, Taiwan #ILA067 7.6 90 195.70
(continued)
4 Seismic Potential Failure Mode Analysis of Concrete …
16 Chi-Chi 20-09-1999 37.2 Chiayi, Taiwan #CHY014 7.6 0 254.90
18 Chi-Chi 20-09-1999 31.1 Chiayi, Taiwan #CHY086 7.6 0 201.60
19 Loma Prieta 18-10-1989 17.2 South St and Pine Dr #47524 7.0 90 174.50
20 Loma Prieta 18-10-1989 15.9 Capitola, CA #47125 7.0 90 390.80
21 Loma Prieta 18-10-1989 11.9 Santa Teresa Hills #57563 7.0 315 223.40
22 Loma Prieta 18-10-1989 6.3 Gilroy Array Sta 3 # 47381 7.0 90 362.00
23 Loma Prieta 18-10-1989 22.7 Gilroy Array Sta 7 #57425 7.0 90 314.30
24 Loma Prieta 18-10-1989 2.8 Corralitos, CA #57007 7.0 90 469.40
4.3 Seismic Potential Failure Mode Analysis
25 Loma Prieta 18-10-1989 16.9 Coyote Lake Dam, CA #57217 7.0 285 471.00
26 Loma Prieta 18-10-1989 2.8 Gilroy Array Sta 1, CA #47379 7.0 90 433.60
27 Imperial Valley 15-10-1979 12.9 El Centro, CA—Array Sta 7 #5028 6.5 52 205.76
28 Imperial Valley 15-10-1979 5.6 El Centro, CA—Differential Array #5165 6.5 360 473.64
29 Imperial Valley 15-10-1979 9.8 Casa Flores Mexicali #6619 6.5 0 238.10
(continued)
87
88
32 Imperial Valley 15-10-1979 8.8 Holtville, CA—Post Office #5055 6.5 225 242.96
33 Imperial Valley 15-10-1979 21.8 Superstition Mtn, CA—Camera Site #0286 6.5 135 182.19
34 Petrolia 17-08-1991 11.7 General Store #89156 6.0 90 488.70
35 Parkfield 28-09-2004 14.0 USGS Parkfield Dense #02 6.0 360 170.00
36 Superstition Hills 24-11-1987 13.0 Westmorland, CA #11369 6.6 180 203.56
37 Kern County 21-07-1952 36.2 Lincoln School #1095 7.5 111 175.95
38 San Simeon 22-12-2003 14.8 Caltrans Bridge Grnds #37737 6.5 90 175.00
39 Landers 28-06-1992 28.6 Whitewater Canyon #5072 7.3 270 124.92
40 Landers 28-06-1992 10.0 Joshua Tree #22170 7.3 90 278.40
4 Seismic Potential Failure Mode Analysis of Concrete …
4.3 Seismic Potential Failure Mode Analysis 89
(1994), and Chi-Chi (1999) events. Owing to the uncertainty of earthquake activ-
ities, the amplitude of earthquakes may vary over a wide range according to the
seismic hazard analysis with a different occurrence probability. Eight levels of the
acceleration amplitude are employed in the analysis, i.e. 0.2, 0.3, 0.34, 0.4, 0.50,
0.6, 0.7 and 0.8 g. By increasing the earthquake intensity, the structure is shifted
from its initial elastic state into a series of successive inelastic states and finally to its
collapse. In each seismic analysis first the static loads including the self-weight of
the dam and the hydrostatic load on upstream face of the dam are applied, and then
the two components of the selected earthquakes are uniformly applied at the base of
the foundation.
4.3.2 Incremental Dynamic Analysis
Failure mode analysis of concrete gravity dams-reservoir-foundation systems is

carried out through nonlinear IDA based on XFEM. Final cracking profiles of the
Guandi gravity dam under the 1967 Koyna record at different shaking intensity levels
are shown in Fig. 4.6. The crack initiation occurs at different times and locations
depending on the amplitude. When the dam is loaded with only static loads, stress
level in the dam is relatively low, no visible crack is detected. For the seismic motions
(a) PGA=0.2g (b) PGA=0.3g (c) PGA=0.34g (d) PGA=0.4g
(e) PGA=0.5g (f) PGA=0.6g (g) PGA=0.7g (h) PGA=0.8g
Fig. 4.6 Final cracking profiles of the Guandi gravity dam subjected to Koyna record at various
intensity levels. a PGA = 0.2 g, b PGA = 0.3 g, c PGA = 0.34 g, d PGA = 0.4 g, e PGA = 0.5 g,
f PGA = 0.6 g, g PGA = 0.7 g, and h PGA = 0.8 g
associated with a PGA of 0.2 g, very minor damage is identified near the change in
the downstream slope. With the increasing acceleration excitation, the downstream
surface and upstream surface as well as the dam heel of the dam are characterized
by high tensile stress and compressive stress. At the 0.30 g level, some moderate
damage is identified, but it does not seem to reach a level that could compromise the
integrity of the section. The upper crack propagates to about three-fifth through the
thickness of the dam section. The results corresponding to the 0.34–0.80 g PGA level
clearly indicate that significant damage is detected. At the 0.34 g level, the upper
crack propagates completely to the upstream face, causing the dam head detached
from the dam body. At the 0.6 g level, cracks appear not only at the dam heel and the
change in the downstream slope, but also near the change in the slope of the upstream
face. As the dam oscillates during an earthquake, these cracks extend into the dam.
The case associated with a PGA of 0.70 g shows clear indications of significant
strength degradation in the dam, with two cracking patterns that completely extend
across the upper and middle sections.
Typical damage processes of the Guandi gravity dam-reservoir-foundation system
subjected to the Koyna acceleration time history with an amplitude of 0.7 g, about 2.0
times the design PGA, are plotted in Fig. 4.7. As shown, three damage zones (i.e. the
change in the slope of the down face, the change in upstream slope, and the dam heel)
are clearly identified and they correspond to the areas associated with the maximum
tensile demands. At 1.84 s, the initial crack in the dam is firstly observed near the
(a) t=1.84s (b) t=1.89s (c) t=1.95s (d) t=2.23s
(e) t=2.32s (f) t=2.77s (g) t=3.69s (h) t=4.42s
Fig. 4.7 Typical damage processes of the Guandi gravity dam under Koyna acceleration time
histories with an amplitude of 0.7 g. a t = 1.84 s, b t = 1.89 s, c t = 1.95 s, d t = 2.23 s, e t =
2.32 s, f t = 2.77 s, g t = 3.69 s, and h t = 4.42 s (Amplification factor: 100)
change in the slope of the down face (Fig. 4.7a). With the on-going acceleration
excitation, the crack extends deeper inside of the dam. During the next seconds, this
crack propagates quickly and finally to the upstream face at 1.95 s, penetrating the
whole section of the dam (Fig. 4.7c). In addition, a crack develops in the dam heel at
2.23 s, which is probably due to stress concentration (Fig. 4.7d). After the time instant
t = 2.32 s, a crack in the middle of the dam is initiated at the change in the slope of
the upstream face (Fig. 4.7e) and the tensile stress is remarkably concentrated here.
With the continuous vibration of the dam, this crack continues to grow downward
toward downstream in a slightly inclined direction (about 10° to the horizontal). The
crack propagates up to 27 m and then grows upward toward downstream at an angle
of about 50° with respect to the dam bottom (Fig. 4.7f-g). At t = 4.42 s, the upstream
crack extends completely to the downstream face, penetrating the whole section of
the dam (Fig. 4.7h). This finally separates a major portion of the structure and causes
catastrophic failure.
4.3.3 Typical Failure Modes of the Guandi Gravity Dam

for Each Record
In order to obtain the potential failure modes of concrete gravity dams, 40 real
earthquake records are selected as earthquake excitation. Each record is scaled to 8
increasing intensity levels in the stream direction to generate eight groups of acceler-
ation excitations. More than 320 cases were calculated in the present study. But only
some typical results are shown in this section. Typical cracking profiles of the Guandi
gravity dam for each record at a certain shaking intensity level are shown in Fig. 4.8.
The deformations of the dam at the end of the earthquakes can also be observed in
Fig. 4.8. It should be noted that the deformation has been magnified 100 times. These
figures depict the damage predicted for different real ground motions considered in
this study. From the numerical results, it can be found that the selected real strong
motion records have significant influence on the crack propagation process and the
finial cracking profile of concrete gravity dams, which may be mainly because they
have a different intensity parameter, such as the peak ground acceleration (PGA),
the strong motion duration (SMD) and the spectral amplitudes for different char-
acteristic periods. In addition to the previously mentioned three damage zones (the
dam neck, dam heel, and change in upstream slope), other two damage zones (the
downstream face and concrete partition junction) are also detected. Top cracking
profiles are almost either nearly horizontal or sloping downward from the down-
stream face toward the upstream face. It can also be found from Fig. 4.8 that for
concrete gravity dams, the dam neck is vulnerable to damage due to stress concen-
tration. At this location, penetrated cracks are detected even under some smaller
ground motions. In some records, there are uncompleted cracks at the neck of the
PGA=0.7g PGA=0.4g PGA=0.6g PGA=0.5g PGA=0.6g
(a) No. 1 (b) No. 2 (c) No. 3 (d) No. 4 (e) No. 5
(f) No. 6 (g) No. 7 (h) No. 8 (i) No. 9 (j) No. 10
(k) No. 11 (l) No. 12 (m) No. 13 (n) No. 14 (o) No. 15
(p) No. 16 (q) No. 17 (r) No. 18 (s) No. 19 (t) No. 20
(u) No. 21 (v) No. 22 (w) No. 23 (x) No. 24 (y) No. 25
Fig. 4.8 Typical cracking profiles for the Guandi gravity dam under real earthquake records at
different shaking intensity levels (amplification factor: 100)
(z) No. 26 (aa) No. 27 (bb) No. 28 (cc) No. 29 (dd) No. 30
(ee) No. 31 (ff) No. 32 (gg) No. 33 (hh) No. 34 (ii) No. 35
(jj) No. 36 (kk) No. 37 (ll) No. 38 (mm) No. 39 (nn) No. 40
Fig. 4.8 (continued)
dam. But in some earthquake analyses, there is no cracking at the neck of the dam.
With the increasing acceleration excitation, cracks are predicted to initialize near
the middle of the upstream face or the downstream face, and multiple failure modes
occur in the dam, which may be because that the downstream and upstream surfaces
as well as the dam heel of the dam are characterized by high tensile stress under
strong ground motions.
4.3.4 Generalization of Potential Failure Modes for Concrete

Gravity Dams
From the 320 numerical simulation results, five typical failure modes are obtained,
as shown in Fig. 4.9. The first failure mode (Fig. 4.9a) is cracking that initiates at
the change in downstream slope and propagates almost perpendicular to the down-
stream surface for a certain distance. Then the cracking propagates approximately
horizontally toward the upstream face and separates the crest from the upper part of
I
v
iv
II
II
III
(a) Failure mode I (b) Failure mode II (c) Failure mode III (d) Failure mode iv (e) Failure mode v
Fig. 4.9 Typical failure modes of concrete gravity dams under strong ground motions. a Failure
mode I, b Failure mode II, c Failure mode III, d Failure mode iv, and e Failure mode v
the dam. The second failure mode (Fig. 4.9b) is cracking that initiates at the dam heel
and progresses a certain distance from the upstream face to the downstream face. In
the third failure mode (Fig. 4.9c), cracking in the middle of the dam are initiated at
the changes in the slope of the upstream face and propagate toward downstream in a
slightly inclined direction. Finally the cracking reaches the downstream face. These
cracks finally separate a major portion of the structure and cause catastrophic failure.
The fourth failure mode (Fig. 4.9d) is cracking that initiates from the downstream
face and travels almost normal to the initiation surface. Then the cracking propagates
across another crack initiated from the change in upstream slope and finally pene-
trates the whole section of the dam. In the fifth failure mode (Fig. 4.9e), crack begins
to initialize near the upper of the upstream face, which is especially easy for the
concrete partition junction. Then it propagates approximately horizontally toward
the opposite face and finally reaches the downstream face, with a cracking pattern
that extends completely across the upper section.
The first two failure modes can be caused by relatively minor earthquakes. Since
the water pounding function of the reservoir is retained to a great extent, no disastrous
results would occur toward the downstream areas. These latter three failure modes
require much stronger seismic excitations (maybe 1.5 times the design PGA or even
greater than that), as compared to the former two failure modes. Although earthquake
shock much stronger than the design value is of low probability, the experiences from
the Hsinfengkiang gravity dam in China, Koyna gravity dam in India, and Sefid Rud
dam in Iran show that these could happen. The possible failure of these modes of
dams retaining large quantities of water will cause the most undesirable impact on
the downstream populated area along with a considerable amount of devastation.
This enlightens the importance for the dam designers and related researchers to
obtain insights into fracture mechanisms of the dam under various dynamic load
cases. Understanding the failure pattern can also be used to assist in improving dam
design and provide more realistic failure scenarios for subsequent dam break flood
modeling.
4.4 Conclusions 95
4.4 Conclusions
Dam failure will lead to catastrophic consequences including loss of human life and
property damage. The objective of this study is to predict the possible failure modes of
concrete gravity dams with considering the effects of dam-reservoir-foundation inter-
action. The reservoir water is modeled using two-dimensional fluid finite elements
by the Lagrangian approach. The extended finite element method (XFEM) based
on the cohesive segments method in conjunction with the phantom node technique
is presented to deal with the numerical prediction of crack propagation in concrete
gravity dams.
40 as-recorded strong ground motion records considered in this study are used
as seismic excitations. The crack propagation processes and potential failure modes
of the Guandi gravity dam with considering the effects of dam-reservoir-foundation
interaction are investigated by applying the IDA method based on the XFEM. Damage
and cracking of concrete gravity dams in earthquakes are caused mainly by excessive
stress. The selected real strong motion records have significant influence on the crack
propagation processes and the finial cracking profile of concrete gravity dams. The
dam neck is vulnerable to damage, in which penetrated cracks are detected even
under some smaller ground motions. With the increasing acceleration excitations,
multiple failure modes occur in the dam. Cracks at the dam neck, dam heel and
abrupt change in slope of upstream face are most often observed. Based on the
320 numerical simulation results, five typical failure modes are obtained. Cracks
may mainly occur in the following regions: the discontinuity in the slope of the
downstream face, the change in upstream slope, the dam heel, the downstream face
and the concrete partition junction. The obtained potential failure modes of the dam
can be used to assist in improving dam design and provide more realistic failure
scenarios for subsequent dam break flood modeling.
References
Alembagheri, M., & Ghaemian, M. (2013a). Damage assessment of a concrete arch dam through
nonlinear incremental dynamic analysis. Soil Dynamics and Earthquake Engineering, 44, 127–
137.
Alembagheri, M., & Ghaemian, M. (2013b). Incremental dynamic analysis of concrete gravity
dams including base and lift joints. Earthquake Engineering and Engineering Vibration, 12(1),
119–134.
Belytschko, T., & Black, T. (1999). Elastic crack growth in finite elements with minimal remeshing.
Billah, A. H. M. M., & Alam, M. S. (2014). Seismic performance evaluation of multi-column bridge
bents retrofitted with different alternatives using incremental dynamic analysis. Engineering
Structures, 62–63, 105–117.
Bretas, E. M., Lemos, J. V., & Lourenço, P. B. (2014). A DEM based tool for the safety analysis of
masonry gravity dams. Engineering Structures, 59, 248–260.
Calayir, Y., & Karaton, M. (2005a). A continuum damage concrete model for earthquake analysis
857–869.
Calayir, Y., & Karaton, M. (2005b). Seismic fracture analysis of concrete gravity dams including
dam-reservoir interaction. Computers & Structures, 83(19), 1595–1606.
Chen, D., Yang, Z., Wang, M., & Xie, J. (2019). Seismic performance and failure modes of the
Jin’anqiao concrete gravity dam based on incremental dynamic analysis. Engineering Failure
Analysis, 100, 227–244.
COSMOS. (2019). Strong motion virtual data center. http://strongmotioncenter.org/vdc/scripts/def
ault.plx.
Hansbo, A., & Hansbo, P. (2004). A finite element method for the simulation of strong and weak
discontinuities in solid mechanics. Computer Methods in Applied Mechanics and Engineering,
193(33), 3523–3540.
reinforced and plain concrete. In Fracture mechanics of concrete: Structural application and
Lee, J., & Fenves, G. L. (1998). A plastic-damage concrete model for earthquake analysis of dams.
Long, Y., Zhang, C., & Xu, Y. (2009). Nonlinear seismic analyses of a high gravity dam with and
without the presence of reinforcement. Engineering Structures, 31(10), 2486–2494.
Nik Azizan, N. Z., Majid, T. A., Nazri, F. M., Maity, D., & Abdullah, J. (2018). Incremental dynamic
analysis of koyna dam under repeated ground motions. IOP Conference Series: Materials Science
and Engineering, 318, 12021.
63, 80–97.
1594–1606.
PEER. (2019). Pacific Earthquake Engineering Research Center. http://peer.berkeley.edu/peer_g
round_motion_database/.
Remmers, J. J. C., de Borst, R., & Needleman, A. (2008). The simulation of dynamic crack prop-
agation using the cohesive segments method. Journal of the Mechanics and Physics of Solids,
56(1), 70–92.
Song, J., Areias, P. M. A., & Belytschko, T. (2006). A method for dynamic crack and shear band
propagation with phantom nodes. International Journal for Numerical Methods in Engineering,
67(6), 868–893.
References 97
Sotoudeh, M. A., Ghaemian, M., & Sarvghad Moghadam, A. (2018). Determination of limit-states
for near-fault seismic fragility assessment of concrete gravity dams. Scientia Iranica.
Zhang, S., Wang, G., Pang, B., & Du, C. (2013a). The effects of strong motion duration on
the dynamic response and accumulated damage of concrete gravity dams. Soil Dynamics and
Zhang, S., Wang, G., & Sa, W. (2013b). Damage evaluation of concrete gravity dams under
mainshock–aftershock seismic sequences. Soil Dynamics and Earthquake Engineering, 50,
16–27.
Zhang, S., Wang, G., & Yu, X. (2013c). Seismic cracking analysis of concrete gravity dams with
Chapter 5
Correlation Between Single Component
Durations and Damage Measures
of Concrete Gravity Dams
5.1 Introduction
As is well known, earthquake ground motion can be characterized by amplitude,

frequency content and strong motion duration, each of which reflects some partic-
ular feature of the shaking. Amplitude is generally characterized by the peak ground
acceleration (PGA), the peak ground velocity (PGV) and the peak ground displace-
ment (PGD). The frequency content is generally described by the Fourier spectrum
of the ground motion. Nonetheless, both amplitude and frequency distribution can
be described by the widely accepted response spectrum (in terms of acceleration,
velocity, or displacement). The importance of the amplitude and frequency content
has been universally recognized. However, the conclusions with regard to the rele-
vance of strong motion duration to structural response differ widely, ranging from
null to significant, which remains a topic of considerable debate. This is mainly
because the influence of strong motion duration on structural response and damage
depend on many factors including the type of structure examined, the construction
model, the other parameters used to characterize the ground motion, the measure of
structural damage employed, and the large number of widely differing duration defi-
nitions that have been proposed (Bommer et al. 2004, 2006; Hancock and Bommer
2006).
There are more than 30 different definitions of strong motion duration (Bommer
and Martínez-Pereira 1999). While there is no unanimous view regarding which of
the definitions of strong motion duration is to be preferred, which probably reflects the
fact that different definitions may be more or less suitable for different applications.
Although a large number of definitions of strong motion duration have been presented
in the literature, the available definitions can be grouped into four different categories:
(a) bracketed duration (Ambraseys and Sarma 1967; Bolt 1973); (b) uniform duration
(Bommer et al. 2009); (c) effective duration (Bommer and Martínez-Pereira 1999);
(d) significant duration (Trifunac and Brady 1975). Subsequently, several new defi-
nitions and prediction models of strong motion duration have been put forward. For
https://doi.org/10.1007/978-981-15-6194-8_5
100 5 Correlation Between Single Component Durations and Damage …
example, Taflampas et al. (2009) proposed a new definition of strong motion dura-
tion combining the alternative bracketed and significant duration definitions based
on the time integral of the absolute ground velocity, and their presented bracketed-
significant duration was found to be well correlated with the strong motion part of
the records. Montejo and Kowalsky (2008) proposed a procedure for estimation of
frequency dependent strong motion duration based on the continuous wavelet trans-
form and the decomposition of the earthquake record. Arjun and Kumar (2011) devel-
oped a neural network approach for estimation of strong motion duration based on
earthquake records and site characteristics. Yaghmaei-Sabegh et al. (2014) presented
a simple and effective empirical model for predicting the significant duration of
ground motions based on recorded earthquake events in Iran.
Experiences from a number of earthquakes show that a ground motion with
moderate peak ground acceleration and a long duration may cause greater strength
and stiffness degradation than a ground motion with a large acceleration and a small
duration (Bommer and Martínez-Pereira 1999). The duration of strong motion may
significantly affect the damage of structures and plays an important role in assessing
the damage potential of earthquake ground motions. However, current approaches
for the earthquake-resistant design and structural analysis based on the response
spectrum have not yet considered the influence of the ground motion duration. There
are many studies reporting that link structural damage to parameters related either
directly or indirectly to strong motion duration. However, the relevance of strong
motion duration to structural response remains an open question, with some research
indicating no effect (Cornell 1997; Shome et al. 1998) and other research indicating a
possible correlation (Reinoso et al. 2000; Hancock and Bommer 2004). At least part
of the reason that researches have differing conclusions on the importance of strong
motion duration is the use of different duration definitions, structural models, and
damage metrics. For example, Hancock and Bommer (2006) presented a summary
and critical review of the literature with regard to the influence of strong motion
duration on structural demand, and concluded that those studies employing damage
measures related to cumulative energy usually found a positive correlation between
strong motion duration and structural damage, while those using damage measures,
such as maximum response parameters, generally found little or no correlations
between duration and damage.
In order to investigate the influence of the strong motion duration on structural
response and damage, a substantial amount of research has been carried out over the
past decades. Mahin (1980) found that strong motion duration might play an impor-
tant role in the inelastic deformation and energy dissipation demands of short period
structures. Léger and Leclerc (1996) suggested that short duration analytic records
should not be used as a substitute for other types of more appropriate records in the
earthquake safety evaluation of concrete dams. Youd et al. (2001) clearly recognized
that the strong motion duration has profound effects on the behavior of saturated soil.
Kunnath and Chai (2004) found that the long duration will increase inelastic design
base shear. Bommer et al (2004) showed that the duration of strong motion can make
a significant influence on the strength degradation of masonry structures. Iervolino
5.1 Introduction 101
et al. (2006) addressed the question of which nonlinear demand measures are sensi-
tive to ground motion duration by statistical analyses of several case studies. The
results led to the conclusion that duration of ground motion does not have a signif-
icant influence on displacement ductility and cyclic ductility demand. Hancock and
Bommer (2007) revealed that duration of strong motion has no influence on damage
measures employing the peak response such as inter-storey drift, but if cumulative
parameters are used to measure the damage, the duration of strong motion is found to
have a significant influence on the inelastic structural response. Ruiz-García (2010)
conducted a comprehensive analytical study to evaluate the influence of strong motion
duration on the residual displacement demands of SDOF and MDOF systems. They
found that long duration ground motions may lead to larger median residual displace-
ment ratios for SDOF systems, and tend to increase residual drift demands in the
upper stories of MDOF systems. Sarieddine and Lin (2013) discussed the correla-
tions between the strong motion duration and structural damage. They emphasized
that there is no correlation between the structural response and ground motion dura-
tion. Raghunandan and Liel (2013) suggested that the ground motion duration should
be considered in the structural design and assessment of seismic risk. Because the
ground motion duration is found to have a significant influence on the collapse risk
of the analyzed RC buildings. Ou et al. (2014) examined the seismic behavior of RC
bridge columns under long duration ground motions. Their result indicated that the
column tested under long duration ground motions shows a similar peak strength but
a lower ductility capacity. Chandramohan et al. (2015) isolated and quantified the
influence of ground motion duration on the structural collapse capacity of a modern
steel moment frame and an RC bridge pier. They thought the significant duration defi-
nition is the most suitable metric to describe the strong motion duration for structural
analysis. Hou and Qu (2015) concluded that duration of spectrally matched ground
motions does not have the effect on the central tendency of ductility demand, but
it has a significant influence on the hysteretic energy dissipation demands. Khaloo
et al. (2016) investigated the influence of reducing earthquake record duration on
fragility curves of RC frames. They concluded that the seismic performance of struc-
tures is dependent on ground motion duration of records when the energy-based or
combination indices are employed as the damage measures. Chandramohan et al.
(2016) developed a procedure to compute the source-specific probability distribu-
tion of the strong motion duration, and evaluated the influence of hazard-consistent
ground motion duration on the collapse risk of an eight-story RC moment frame
building. Barbosa et al. (2016) evaluated the effects of ground motion duration on
structural damage of steel moment resisting frame buildings. Their results indicated
that strong motions duration has a significant influence on the deformations and
measures when the spectral accelerations are large. Belejo et al. (2017) found that
the strong motion duration has little effect on the displacement and drift responses of
the plan-asymmetric RC building, but it considerably affects the damage predicted
using the Park-Ang and Reinhorn and Valles damage indices. Tombari et al. (2017)
employed a beam nonlinear Winkler foundation model to investigate the effects of
ground motion duration and soil nonlinearity on the seismic performance of single
piles based on the incremental dynamic analysis method. Their results showed that
the long ground motion duration will exacerbate the effects of the cyclic degrada-
tion/hardening and the gapping. Bravo-Haro and Elghazouli (2018) examined the
influence of ground motion duration on the nonlinear dynamic response of steel
moment frames considering the cyclic degradation effects. They found that long
duration ground motions will increase the probability of collapse of steel moment
frames, and considerably reduce the structural collapse capacity up to 40%. Pan et al.
(2018) investigated the effects of strong motion duration on the seismic performance
and collapse rate of three low-rise light-frame wood houses. Kiani et al. (2018) exam-
ined the role of the conditioning intensity measure in the influence of ground motion
duration on the dynamic response of building structures. They found that ground
motion duration considerably affects the structural response in terms of cumulative
absolute velocity and spectral acceleration. Other conditioning intensity measures,
such as the peak ground acceleration and velocity, spectrum intensity, and spectral
acceleration, are not substantially affected by ground motion duration. Samanta and
Pandey (2018) studied the effects of strong motion duration on the seismic perfor-
mance of a fifteen storied building structure. They concluded that ground motion
duration decreases the peak floor acceleration and average floor spectral acceleration
only at higher hazard. Zhang et al. (2018) concluded that the ground motion duration
has a significant influence on the seismic performance of high-rise intake towers in
terms of the seismic displacement along the main stream. Molazadeh and Saffari
(2018) investigated the destructive effects of strong motion duration on ductility and
hysteretic energy dissipation demands of SDOF systems. Their results showed that
the strong motion duration has a significant influence on the pinching-degrading
behavior of SDOF systems with short periods. Kabir et al. (2019) evaluated the
seismic fragility of a multi-span RC bridge subjected to near-fault, far-field and long
duration ground motions. Their results showed that failure probabilities of bridge
piers and isolation bearings are dominated by long duration ground motions than
those of near-fault and far-fault ground motions. Kiani et al. (2019) examined the
importance of ground motion duration on the risk-based structural responses, and
thought the ground motion duration based on the definition of significant duration
even in case of shallow events is an important parameter which should be considered
in ground motion selection. Greco et al. (2019) discussed the correlation between
the ground motion duration and damping reduction factor using the random vibra-
tion theory. Their results indicated that the damping reduction factor is sensitive to
the ground motion duration, where the value of the damping reduction factor will
diminish with the increase of effective duration values. Pan et al. (2019) studied the
effects of long duration ground motions on the seismic performance and collapse
capacity of a mid-rise wood frame building through incremental dynamic analysis
method. Their results indicated that long duration ground motions will reduce the
median collapse capacity by 18%, and increase the estimated median Park and Ang
damage index by 36%. Pan et al. (2020) explored the seismic damage of four repre-
sentative light-frame wood houses under long duration earthquakes, and developed
the fragility curves for conventional low-rise light-frame wood buildings.
It should be noted that few studies have focused their attention on the nonlinear
dynamic response and seismic damage of concrete gravity dams subjected to earth-
quake motions with different strong motion durations. For example, Zhang et al.
(2013a) investigated the effects of strong motion duration on the dynamic response
and accumulated damage of concrete gravity dams based on the definition of signif-
icant duration. Their result showed that strong motion duration is insignificant to
peak displacement response assessment. While studies employing damage measures
using local and global damage indices showed that strong motion duration is posi-
tively correlated to the accumulated damage for events with similar response spec-
trum. Wang et al. (2018) investigated the strong duration effects on the nonlinear
dynamic response of a dam-reservoir-foundation system. Their results indicated that
the larger is the ground motion intensity, the more pronounced is the ground motion
duration effects on dam responses. Xu et al. (2018) investigated the influence of
strong motion duration on the seismic performance of high concrete-faced rockfill
dams based on elastoplastic analysis. They found that the strong motion duration
was positively correlated with the displacement and plastic shear strain, but weakly
correlated with face-slab damage index.
The objective of this chapter is to provide a method for quantifying the interrela-
tionship between strong motion durations and damage measures of concrete gravity
dams. 20 as-recorded accelerograms with a wide range of durations, which are scaled
and matched to match a 5% damped target spectrum, are selected in this study. Three
different definitions of significant duration, bracketed duration and uniform duration
are presented for measuring strong motion durations. Local damage index, global
damage index, peak displacement, and damage energy dissipation are employed as
the measures of structural damage. A Concrete Damaged Plasticity (CDP) model
including the strain hardening or softening behavior is selected for the concrete
material. Nonlinear dynamic response and seismic damage analyses of Koyna gravity
dam under different strong motion durations are conducted to furnish the structural
damage status. The interrelationships between the different strong motion durations
and the damage measures are given.
5.2 Strong Motion Duration-Related Measure Used in this

Study
5.2.1 Definitions of Strong Motion Duration for Single

Component Ground Motion
Any attempt to study on the correlation between strong motion durations and struc-
tural damage levels immediately faces the problem that there is currently no univer-
sally accepted definition of strong motion duration. Several researches in the past
have been conducted for the quantification of strong motion duration (Shoji et al.
2005; Bommer et al. 2009; Taflampas et al. 2009; Arjun and Kumar 2011), and
there are more than 30 different definitions of strong motion duration (Bommer and
Martínez-Pereira 1999) based on one component of ground motions. There is no clear
consensus as to which of the multiple definitions of duration is to be preferred, which
probably reflects the fact that different definitions may be more or less suitable for
different applications. In the past, a number of researchers had proposed procedures
to compute strong motion duration of an earthquake record. In general, the available
definitions can be classified into four different groups:
(a) Bracketed duration (T B ) (Ambraseys and Sarma 1967; Bolt 1973) in which the
duration is defined as the time interval between the first and the last exceedances
of a particular threshold of acceleration (usually 0.05 g);
(b) Uniform duration (T U ) (Bommer et al. 2009), which rather than a continuous
time window is defined as the sum of time intervals during which the record
exceeds a particular acceleration threshold;
(c) Effective duration (T E ) (Bommer and Martínez-Pereira 1999) that defines the
duration of strong motion as the time interval between two particular thresholds
of the Arias intensity, and;
(d) Significant duration (T S ) (Trifunac and Brady 1975), it is defined to be the
interval between the time at which a given percentage of Arias intensity of the
record is reached.
Most of the proposed definitions are applied directly to recordings of the ground
motion, but a few apply one of these generic definitions to dynamic response of
structures. Within the four generic definitions, distinction, which is related to whether
the thresholds of acceleration or energy are absolute values or defined relative to the
maximum value attained in the accelerogram, can also be found. Absolute definition
includes bracketed duration (T B ), uniform duration (T U ) and effective duration (T E ),
and relative definition contains significant duration.
Among different seismic parameters, attention is focused on those which have a
high level of correlation with the examined damage indices. The definition of effective
duration is not considered in this study because it is very difficult to ascertain the
particular thresholds of the Arias intensity for the effective duration. Therefore, the
bracketed duration, the uniform duration, and the significant duration according to
Trifunac/Brandy are chosen in this study.
The uniform duration only considers the intervals for which the ground accel-
eration is above the threshold, which differs from the bracketed duration, therefore
uniform duration is always shorter than bracketed duration for a given acceleration
record, as shown in Fig. 5.1. For these two definitions, the absolute threshold such
as 0.05 g or 0.10 g can be used.
It is well known that the destructiveness of a seismic excitation can be described
using several intensity parameters. The Arias intensity (Arias 1970) is an energy-
related seismic index which is designed to reflect the total energy content of a seismic
excitation, and is defined by the following relation:
5.2 Strong Motion Duration-Related Measure Used in this Study 105
Acceleration (g) 0.3

0.2 Acceleration record
0.1
0.0
-0.1
-0.2
-0.3
0 2 4 6 8 10
Time (s)
(a) A natural accelerogram
Absolute acceleration (g)
0.3
Accelaration record
0.2 Bracketed duration (TB )
Absolute threshold 0.05g
0.1 TB=3.76s
0.0
0 2 4 6 8 10
Time (s)
(b) The diagram of bracketed duration
Absolute acceleration (g)
0.3
Accelaration record
Uniform duration (TU )
0.2
Absolute threshold 0.05g
0.1 TU=2.36s
0.0
0 2 4 6 8 10
Time (s)
(c) The diagram of uniform duration
Fig. 5.1 Diagram of bracketed and uniform durations of a natural accelerogram
T0
π
I0 = a 2 (t)dt (5.1)
2g
0
where, I 0 is the Arias intensity, T 0 is the total duration of the record and a is the
ground acceleration.
Trifunac and Brady (1975) have defined the significant duration of strong ground
motion as the time interval between 5 and 95% of the Arias intensity. For this defini-
tion, the Husid diagram (Husid 1969) is used. The Husid diagram is the time history
of the seismic energy content scaled to the total energy content and it is given by
π
t
2g 0 a 2 (t)dt
H (t) = (5.2)
I0
where, a is the ground acceleration and H(t) is the Husid diagram as a function of
time t. The Husid diagram of a natural accelerogram is shown in Fig. 5.2. In this
figure strong motion duration after Tricunac/Brandy can be computed easily.
gravity dams
0.3
Acceleration (g)
0.0
-0.3
0 10 20 30 40 50 60
Time (s)
100
85% 95%
Arias intensity (%)
80
T70%
60
T90% H(t)
40
T70%=18.73s
20 T90%=26.04s
5% 15%
0
0 10 20 30 40 50 60
Time (s)
Fig. 5.2 Husid diagram of a natural accelerogram
Because the thresholds of Arias intensity used in this study are relative to the
total Arias intensity of the record, the symbol T S (significant duration) is used, with
the subscript indicating the relative threshold. Two definitions of T S(70%) (15–85%)
(Takizawa and Jennings 1980) and T S(90%) (5–95%) (Trifunac and Brady 1975) are
used, as shown in Fig. 5.2. The first strong motion duration is defined as the time
interval between 15 and 85% of the Husid diagram and the second strong motion
duration is defined as the time interval between 5 and 95% of the Husid diagram.
The former definition is intended to capture the energy from the body waves whereas
the latter includes the full wave train. The T S is defined by the following relation
TS = T2 − T1 (5.3)

H (T1 ) = 15% H (T1 ) = 5%
or (5.4)
H (T2 ) = 85% H (T2 ) = 95%
where, T S is the significant duration of ground motion, T 2 is the time at the 85% or
95% of the Husid diagram and T 1 is the time at the 15% or 5% of the Husid diagram.
5.2.2 Accelerogram Selection and Correction
In order to identify the relationship between the accumulated damage and strong
ground motion duration, 20 real earthquake records used in this study have been
selected from the databases of the COSMOS (COSMOS 2019) and PEER (PEER
2019) with a broad distribution of durations. Table 5.1 shows the data of the acceler-
ation time history with different earthquake magnitudes (i.e. M w 6.0–7.8) that have
occurred in the United States, Taiwan and India. The data sample includes world-
wide well-known earthquakes with strong seismic activity, such as the Loma Prieta
(1989), Northridge (1994), and Chi-Chi (1999) events.
For the purpose of analyzing strong motion duration effects on nonlinear demand,
it is necessary to minimize the influence of the spectral amplitude and other factors.
The ground motion records are normalized to have the same peak ground acceleration
(PGA) equal to 0.3 g. Besides, the accelerograms are scaled and matched to the target
spectrum at a given damping level using the SeismoSignal. The SeismoSignal uses
wavelets to adjust the accelerograms to match the response spectra, while minimizing
changes to the other ground motion characteristics. The adjusted accelerograms have
a good match on average to the target spectrum, which ensures that the main differ-
ence between the records is the strong motion duration. The modified records with
different range durations are illustrated in Fig. 5.3, and the corresponding response
spectrum is shown in Fig. 5.4.
5.2.3 Strong Motion Duration Prediction
In order to develop vector predictions of duration in conjunction with other ground-

motion parameters related to amplitude and energy content. Three different levels of
peak ground acceleration (PGA) are considered for the input motions: 0.25, 0.30 and
0.35 g. Among different strong motion durations, attention is focused on those which
have a high level of correlation with the examined damage measures. Therefore, the
definitions of the significant, bracketed, and uniform duration are used in this study.
According to these definitions, strong motion durations for all seismic excitations
described in Table 5.1 are calculated, as shown in Table 5.2. The duration parameters
of the non-modified real accelerograms are also given in Table 5.2. It can be seen from
Table 5.2 that these earthquakes with larger PGA will increase absolute measures like
bracketed duration (T B ) and uniform duration (T U ) whereas the relative measures
[e.g. significant duration: T S(70%) (15–85%) and T S(90%) (5–95%)] are not influenced
by the increased amplitude.
Table 5.1 Strong motion database (non-modified records)
108
1 Koyna 11-12-1967 13.0 Koyna Dam 6.5 0 480.20
2 Northridge 17-01-1994 35.8 Los Angeles Terrace #24592 6.7 180 310.10
3 Northridge 17-01-1994 15.7 Epiphany Litheran Church #5353 6.7 196 404.20
4 Loma Prieta 18-10-1989 17.2 South St and Pine Dr #47524 7.0 90 174.50
6 Petrolia 17-08-1991 11.7 General Store #89156 6.0 90 488.70
7 Northridge 17-01-1994 18.4 Los Angeles Mulholland Dr #5314 6.7 35 641.90
8 Loma Prieta 18-10-1989 15.9 Capitola, CA #47125 7.0 90 390.80
10 Landers 28-06-1992 28.6 Whitewater Canyon #5072 7.3 270 124.92
11 Northridge 17-01-1994 8.6 Los Angeles Dam #2141 6.7 334 419.10
12 Loma Prieta 18-10-1989 11.9 Santa Teresa Hills #57563 7.0 315 223.40
13 Superstition Hills 24-11-1987 13.0 Westmorland, CA #11369 6.6 180 203.56
14 Kern County 21-07-1952 36.2 Lincoln School #1095 7.5 111 175.95
15 Landers 28-06-1992 10.0 Joshua Tree #22170 7.3 90 278.40
16 San Simeon 22-12-2003 14.8 Caltrans Bridge Grnds #37737 6.5 90 175.00
17 Parkfield 28-09-2004 14 USGS Parkfield Dense #02 6.0 360 170.00
(continued)
5 Correlation Between Single Component Durations and Damage …
18 Imperial Valley 15-10-1979 9.8 Casa Flores Mexicali #6619 6.5 0 238.10
19 Northridge 17-01-1994 12.5 Coldwater Canyon School #5309 6.7 180 313.30
5.2 Strong Motion Duration-Related Measure Used in this Study
109
1 2 3 4 5 6
7 9 10 11
8
0.6 g
12 13 14 15
16 17 18 19 20
20 s
Fig. 5.3 Acceleration time histories for the modified strong motion records used in this study. For
ground motion information see Table 5.1
gravity dams
3.0
2.5
Amplification factor
2.0
1.5
1.0
0.5
0.0
0 1 2 3 4
Period (s)
Fig. 5.4 Acceleration spectrum for scaled earthquake records
5.2.4 Correlation Analysis
Figure 5.5 shows the correlation between total duration and different duration defi-
nitions of the modified records with a PGA of 0.30 g, and Fig. 5.6 illustrates a
case of correlation between significant duration (15–85% Arias intensity) and other
different duration definitions of the modified records with a PGA of 0.30 g. As shown,
it is known that although durations calculated using different definitions are generally
poorly correlated with the total duration, there is an approximate correlation between
Table 5.2 Strong motion durations obtained from modified records
No. Earthquake T 0 (s) Significant Bracketed duration T B (s) Uniform duration T U (s)
duration T S (s)
T S(70%) T S(90%) PGA = 0.30 g PGA = 0.25 g PGA = 0.35 g PGA = 0.30 g PGA = 0.25 g PGA = 0.35 g
1 Koyna 10.00 3.08 5.02 6.92 6.92 8.98 2.62 2.16 3.11
2 Northridge 20.00 5.49 10.27 14.31 12.03 16.21 4.82 3.72 5.94
3 Northridge 20.00 7.14 10.12 13.74 11.12 16.58 6.22 5.12 7.20
4 Loma Prieta 40.00 13.00 23.70 33.18 25.98 33.22 10.60 8.08 12.74
5 Chi-Chi 60.00 18.73 26.04 31.31 31.28 31.33 12.24 10.05 14.27
6 Petrolia 10.00 2.00 3.26 3.76 3.70 5.14 2.36 1.90 2.74
7 Northridge 25.00 4.46 7.62 9.28 8.56 9.28 3.76 3.12 4.44
8 Loma Prieta 39.82 8.48 13.16 14.62 14.46 14.62 5.74 4.88 6.62
9 Chi-Chi 40.00 15.24 23.55 30.52 25.02 30.56 12.38 10.53 14.40
10 Landers 56.55 23.59 31.36 43.04 40.18 49.16 20.00 16.84 22.58
11 Northridge 26.55 4.06 6.67 7.02 6.66 9.01 3.08 2.43 3.62
12 Loma Prieta 25.00 6.60 9.32 16.28 15.18 19.46 6.62 5.74 7.60
13 Superstition 59.82 8.82 17.04 23.24 23.20 23.24 7.44 5.76 8.78
5.2 Strong Motion Duration-Related Measure Used in this Study
Hills
14 Kern County 54.22 13.86 28.86 33.84 32.98 35.34 8.50 6.44 10.74
15 Landers 40.00 21.48 26.08 31.44 31.10 36.68 14.14 11.88 16.48
(continued)
111
112
No. Earthquake T 0 (s) Significant Bracketed duration T B (s) Uniform duration T U (s)
duration T S (s)
T S(70%) T S(90%) PGA = 0.30 g PGA = 0.25 g PGA = 0.35 g PGA = 0.30 g PGA = 0.25 g PGA = 0.35 g
16 San Simeon 25.00 5.86 10.19 10.97 9.14 11.66 3.28 2.39 4.09
17 Parkfield 20.00 6.41 11.63 16.58 14.86 18.68 5.93 4.81 6.99
18 Imperial 16.95 6.65 10.56 16.27 15.69 16.69 7.02 5.87 8.03
Valley
19 Northridge 21.86 8.20 15.86 18.82 18.80 19.04 6.30 5.06 7.66
20 Chi-Chi 50.00 17.62 25.77 33.52 32.55 34.02 18.61 16.05 20.51
40
TS(70%)(15~85%)
Strong motion duration (s)

TS(90%)(5~95%)
TB (0.05g)
30
TU (0.05g)
20
10
0
0 10 20 30 40 50 60
Total duration (s)
Fig. 5.5 Correlation between total duration and different definitions of duration (PGA = 0.3 g)
TS(90%)(5~95%)
Strong motion duration (s)
40
TB (0.05g)
TU (0.05g)
30
20
10
0
0 5 10 15 20 25
Significant duration TS(70%) (15~85%) (s)
Fig. 5.6 Correlation between significant duration [T S(70%) (15–85%)] and different definitions of
duration (PGA = 0.3 g)
significant duration [T S(70%) (15–85%)] and other different definitions of duration

for these modified records used in this study. This is because the records have been
scaled to have the same PGA, and have been matched to the target spectrum at a
given damping level.
5.3 Seismic Accumulated Damage Indices
We evaluate the effects of mainshock-aftershock seismic sequences on accumulated

damage based on the ultimate state of the dam. As a new indicator that means the
soundness of a dam against crack penetration failure, both local and global damage
indices are proposed to assess the accumulated damage of structures quantitatively. In
this model, the global damage is obtained as a weighted average of the local damage
at the ends of each element, with the dissipated energy as the weighting function.
The local damage index (Fig. 5.7) is given by the following relation:
The length of the residual path in crack

path 1 where cracking did not occur
lD1
L1
The length of the residual path in crack

path 2 where cracking did not occur
lD2
L2
Fig. 5.7 Establishment of the crack paths

l Di
D I Li = (5.5)
Li
where, DI Li is the local damage index for crack path i, L i is the total length to which
crack path i is expected to grow, and lDi is the length of the damage path in crack
path i. The damage of a crack at an element integration point is indicated by shading
the related area with red color. As shown in Fig. 5.7, crack paths along the damaged
elements are obtained. The residual crack paths are established a priori, and assumed
to propagate approximately horizontally toward the opposite face, with a cracking
pattern that extends completely across the section.
In the present case, the global damage index is a weighted average of the local
damage indices and the damage dissipation is chosen as the weighting function. The
global damage index is given by
n
i=1 D I Li E i
D IG = n (5.6)
i=1 E i
where, DI G is the global damage index, DI Li is the local damage index, E i is the
damage dissipation energy at the crack path i (the energy dissipated in the whole (or
partial) model by damage), n is the number of crack paths at which the local damage
is computed.
5.4 Seismic Damage Analysis of Koyna Dam 115
5.4 Seismic Damage Analysis of Koyna Dam
5.4.1 Description of Koyna Gravity Dam Model Used

for Evaluation
The Koyna concrete gravity dam in India, 103 m high, and 70.2 m wide at its base,
which is one of a few concrete dams that have experienced a destructive earthquake,
is selected as an application. Finite element model for the tallest section of the dam
is shown in Fig. 5.8. The mesh of the dam is adequately refined at the base and near
the changes in the slope of the downstream face, in which the crack propagation is
expected. The reason is that damage due to tensile stresses is expected to initiate near
stress concentrations in those zones. The foundation of the dam is taken as being
rigid.
The material parameters of the Koyna dam concrete are as follows: the elasticity
modulus E = 3.1 × 104 MPa, the Poisson’s ratio ν = 0.2, the mass density ρ =
2643 kg/m3 , the fracture energy is 250 N/m. The tensile and compressive strength
of the dam are 2.9 and 24.1 MPa, respectively. A dynamic magnification factor of
1.2 is considered for the tensile strength to account for strain rate effects. The energy
dissipation of the monolith is considered by the Rayleigh damping method with 5%
damping ratio. The maximum reservoir water level of 96.5 m is considered. Applied
Fig. 5.8 Finite element model of Koyna dam

loads include self-weight of the dam, hydrostatic, uplift, hydrodynamic, and earth-
quake forces. The static solution of the dam due to its gravity loads and hydrostatic
loads is taken as initial conditions in the dynamic analyses of the system. Westergaard
virtual mass (Westergaard 1933) is employed to include the hydrodynamic effect.
5.4.2 Strong Motion Duration Effects on Accumulated

Damage of Concrete Gravity Dams
Earthquakes often lead to stiffness and strength deterioration of structures. For two
earthquake ground motions with similar spectral amplitude but of different duration,
the motion of longer duration would be expected to more damaging. Hence, the
duration of earthquake ground motion should be considered an important parameter
in addition to the maximum amplitude and frequency content for adequately charac-
terizing the effect of earthquake ground motion on seismic damage of structures. In
order to investigate the effects of strong motion duration on the accumulated damage
of concrete gravity dams, the dynamic damage analyses of Koyna gravity dam under
selected real earthquakes are conducted employing the Concrete Damaged Plas-
ticity (CDP) model developed by Lubliner et al. (1989) and modified by Lee and
Fenves (1998) (see Zhang et al. 2013b) for a detailed description of the constitu-
tive model). The integration time step used in the analysis is 0.01 s. Only the one
component (horizontal) of the seismic input is considered in these analyses, which
are conducted considering 20 different earthquake records, with records selected to
represent a wide range of intensities from relatively weak motions to very strong
shaking. These records are modified to match a 5% damped target spectrum.
The accumulated damage profiles of Koyna dam during the modified real accelero-
grams with a PGA of 0.3 g are shown in Fig. 5.9. These figures depict the damage
predicted for different range durations of real ground motions considered in this study.
The shaded area related to red color indicates those elements that experienced some
level of tensile damage over the duration of the analysis. As shown, the crack prop-
agation process and failure modes are obtained. From the cracking profiles shown
in Fig. 5.9a–t, it can be observed that the failure mechanism is formed of two main
damage zones, one at the base and one in the upper part of the dam. In almost all the
analyses, the cracking is always initiated at the dam heel which may be due to stress
concentration, and then progresses a long way from the upstream face to the down-
stream face. These cracking profiles in the upper part of the dam are always initiated
near the discontinuity in the slope of the downstream face. Top cracking profiles
are almost either nearly horizontal or sloping downward from the downstream face
toward the upstream face at approximately a 45° angle to the vertical. The crack
trajectory curves down due to the compressive stresses resulting from rocking of the
top block. But in some analyses, cracks are predicted to initialize near the middle
of the upstream face or the downstream face, and extend into the dam. Because of
the thrust of impounded water which is opposing the tendency of the top section to
5.4 Seismic Damage Analysis of Koyna Dam 117
TS (70%)=3.08s TS (70%)=5.49s TS (70%)=7.14s TS (70%)=13.00s TS (70%)=18.73s

TB=6.92s TB=14.31s TB=13.74s TB=33.18s TB=31.31s
TU=2.62s TU=4.82s TU=6.22s TU=10.60s TU=12.24s
(a) No. 1 (b) No. 2 (c) No. 3 (d) No. 4 (e) No. 5

(f) No. 6 (g) No. 7 (h) No. 8 (i) No. 9 (j) No.10
(k) No. 11 (l) No. 12 (m) No. 13 (n) No. 14 (o) No. 15
TS (70%)=6.41s TS (70%)=6.65s TS (70%)=8.20s TS (70%)=17.62s
TS (70%)=5.86s TS (90%)=15.86s
TS (90%)=11.63s TS (90%)=10.56s TS (90%)=25.77s
TS (90%)=10.19s TB=16.58s TB=18.82s
TB=10.97s TB=16.27s TB=33.52s
TU=5.93s TU=7.02s TU=6.30s TU=18.61s
TU=3.28s
(p) No. 16 (q) No. 17 (r) No. 18 (s) No. 19 (t) No. 20
Fig. 5.9 Cracking profiles for Koyna dam under modified real accelerations with a PGA of 0.30 g
(for ground motion information see Table 5.1)
slide along the crack in the upstream direction, the computed crack profiles in the
upper part of the dam can be considered neutral or favorable conditions to maintain
stability.
For the seismic motions associated with a short duration, the actual response of
the dam will exhibit some tensile cracking, and some small damage is identified.
But it will not drastically affect the results of the dynamic behavior of the dam. On
the other hand, the cases associated with a moderate duration, the moderate damage
is found, but it does not seem to reach a level that could compromise the integrity
of the section. However, the results corresponding to the input motions with a long
duration clearly show indications of significant strength degradation in the dam, with
a cracking pattern that extends completely across the upper section. Longer duration
will lead to greater accumulate damage to which aseismic design of the dam should
be given attention.
5.5 Correlation Study Between Strong Motion Durations

and Damage Measures
The influence of strong motion duration on nonlinear structural response, however,

remains a topic of much debate and universal conclusions are unlikely to reached
since the resolution of the issue is complicated by the variety of definitions of duration
and the variety of structural behavior, as well as the difficulty of decoupling the
specific effect of duration from other features of the ground motion. In order to
analysis the interrelationships between the different strong motion durations and the
damage measures, both relative (15–85% significant duration and 5–95% significant
duration) and absolute (0.05 g bracketed duration and 0.05 g uniform duration)
definitions are used to determine the influence of ground motion duration on different
damage measures such as peak displacement, local damage index, global damage
index and damage energy dissipation. The seismic response analyses are carried
out by scaling each input record, to progressively increase the intensity of ground
shaking by increments of about 0.05 g. Real earthquake records are scaled to different
intensity levels in order to produce responses ranging from elastic to large nonlinear
responses of the dam. A further aspect which has been taken into consideration is
the expected damage potential of strong motion duration on concrete gravity dams
to be analyzed.
5.5.1 Damage Measures of Local and Global Damage Indices
After the nonlinear dynamic response analyses of the examined concrete gravity dam
for all the selected accelerograms at different intensity levels, the peak displacement
and damage energy dissipation are obtained, and the local and global damage indices
5.5 Correlation Study Between Strong Motion Durations … 119
are computed according to Eqs. (5.5) and (5.6) for each seismic excitation. The
damage index values for each used accelerogram are given in Table 5.3. The influence
of strong motion duration on the damage of concrete gravity dams is illustrated by
using different symbols for different levels of duration on plots of average spectral
accelerations versus damage indices. Correlations are also made between strong
motion duration and peak displacement and damage energy dissipation.
As mentioned above, the failure mechanism is formed by two main damage zones,
one at the base and one in the upper part of the dam. Hence, two local damage indices
are calculated. To analyze the effects of strong motion duration on the accumulated
damage of concrete gravity dams, Figs. 5.10, 5.11 and 5.12 are generated by plotting
the accumulated damage of the dam imparted by the 20 records for a given level of
intensity in terms of local and global damage indices. Trend lines (straight lines) for
each level of intensity are displayed with the aim of identifying a general tendency (if
any). This also applies to the other figures of this type in the article. However, S-from
trend lines are fitted to the plots to show the general trends for the correlation between
strong motion duration and local damage index of the dam heel. In some cases, the
S-form trend lines are also used to identify the correlation between the uniform
duration and damage measures. The trend lines show that, as would be expected,
the local and global damage indices of the examined dam during the modified real
accelerograms with longer duration are generally greater than that under shorter
events for the same level of spectral acceleration. It can also be found that the ground
motions with greater peak ground acceleration (PGA) would cause greater damage
to concrete gravity dams than smaller events. However, the correlation is relatively
weak with significant overlap between uniform duration and damage measures with
different PGA (Fig. 5.10d, 5.11d, and 5.12d). The damage demand examined by the
significant duration (15–85% significant duration and 5–95% significant duration)
is proportional to the peak ground acceleration of ground motions. Results from
the bracketed duration and uniform duration measures are not very similar. This
is because the relative measures (significant duration) are not influenced by the
increased amplitude.
As can be seen from Figs. 5.10, 5.11 and 5.12, studies employing damage measures
using local and global damage indices show that strong motion durations calculated
from different definitions (bracketed duration, uniform duration and significant dura-
tion) are all positively correlated to the accumulated damage for events with similar
response spectrum. Damage measures such as the local damage index for the dam
upper zone and global index for the dam are consistently greater for ground motions
with longer duration. While, the accumulated damage on the dam heel is not very
sensitive to strong motion duration. The accumulated damage for the upper zone of
the dam is more sensitive to ground motion duration, which gives more importance
to the dissipated energy during the hysteretic behavior of the structure since the high
seismic response zone is mainly located in the upper zone of the dam. After the long
duration earthquake, more significant accumulated damage remains in the upper zone
of the dam due to plastic strain during cyclic loadings.
Table 5.3 Damage index values (for ground motion information see Tables 5.1 and 5.2)
120
No. Earthquake Local damage index Global damage index

The dam heel The upper part of the dam
PGA (0.25 g) PGA (0.30 g) PGA (0.35 g) PGA (0.25 g) PGA (0.30 g) PGA (0.35 g) PGA (0.25 g) PGA (0.30 g) PGA (0.35 g)
1 Koyna 0.08 0.09 0.14 0.08 0.21 0.28 0.08 0.12 0.18
2 Northridge 0.14 0.19 0.22 0.08 0.26 0.32 0.12 0.21 0.25
3 Northridge 0.19 0.25 0.28 0.26 0.48 0.65 0.21 0.31 0.38
4 Loma Prieta 0.20 0.27 0.31 0.40 0.61 0.79 0.25 0.36 0.44
5 Chi-Chi 0.20 0.24 0.28 0.58 0.90 1.00 0.29 0.40 0.46
6 Petrolia 0.09 0.14 0.17 0.05 0.10 0.13 0.08 0.13 0.16
7 Northridge 0.10 0.14 0.17 0.18 0.30 0.38 0.12 0.18 0.23
8 Loma Prieta 0.17 0.22 0.25 0.25 0.42 0.64 0.19 0.27 0.35
9 Chi-Chi 0.17 0.22 0.25 0.38 0.63 0.86 0.22 0.32 0.40
10 Landers 0.23 0.26 0.31 0.69 1.00 1.00 0.35 0.44 0.48
11 Northridge 0.05 0.08 0.11 0.02 0.15 0.21 0.04 0.10 0.14
12 Loma Prieta 0.13 0.17 0.22 0.17 0.28 0.41 0.14 0.20 0.27
13 Superstition 0.16 0.22 0.27 0.23 0.44 0.55 0.18 0.28 0.34
Hills
14 Kern 0.20 0.23 0.27 0.36 0.64 0.90 0.25 0.34 0.43
County
(continued)
No. Earthquake Local damage index Global damage index
The dam heel The upper part of the dam
PGA (0.25 g) PGA (0.30 g) PGA (0.35 g) PGA (0.25 g) PGA (0.30 g) PGA (0.35 g) PGA (0.25 g) PGA (0.30 g) PGA (0.35 g)
15 Landers 0.22 0.25 0.30 0.59 0.90 1.00 0.31 0.42 0.48
16 San Simeon 0.20 0.25 0.28 0.30 0.32 0.54 0.23 0.27 0.35
17 Parkfield 0.14 0.20 0.23 0.24 0.31 0.55 0.17 0.23 0.32
18 Imperial 0.14 0.17 0.22 0.21 0.43 0.55 0.16 0.24 0.31
Valley
19 Northridge 0.16 0.19 0.22 0.34 0.44 0.67 0.20 0.25 0.34
20 Chi-Chi 0.22 0.26 0.31 0.40 0.75 1.00 0.27 0.39 0.50
5.5 Correlation Study Between Strong Motion Durations …
121
0.3 0.3
Local damage index

Local damage index
0.2 0.2
0.1 0.1
PGA=0.30g PGA=0.30g
PGA=0.25g PGA=0.25g
PGA=0.35g PGA=0.35g
0.0 0.0
0 5 10 15 20 25 0 5 10 15 20 25 30 35
Duration (s) Duration (s)
(a) Significant duration, TS (70%) (15-85%) (b) Significant, TS (90%) (5-95%)
0.3 0.3
Local damage index
Local damage index

0.2 0.2
0.1 0.1
PGA=0.30g PGA=0.30g
PGA=0.25g PGA=0.25g
PGA=0.35g PGA=0.35g
0.0 0.0
0 10 20 30 40 50 0 5 10 15 20 25
(c) Bracketed duration, TB (0.05g) (d) Uniform duration, TU (0.05g)
Fig. 5.10 Influence of strong motion duration on the local damage index measure for the dam heel.
a T S(70%) (15–85%); b T S(90%) (5–95%); c T B (0.05 g); d T U (0.05 g)
1.0 1.0
Local damage index
Local damage index
0.8 0.8
0.6 0.6
0.4 0.4
PGA=0.30g PGA=0.30g
0.2 0.2
PGA=0.25g PGA=0.25g
PGA=0.35g PGA=0.35g
0.0 0.0
0 5 10 15 20 25 0 5 10 15 20 25 30 35
1.0 1.0
Local damage index
Local damage index
0.8 0.8
0.6 0.6
0.4 0.4
PGA=0.30g PGA=0.30g
0.2 0.2
PGA=0.25g PGA=0.25g
PGA=0.35g PGA=0.35g
0.0 0.0
0 10 20 30 40 50 0 5 10 15 20 25
Fig. 5.11 Influence of strong motion duration on the local damage index measure for the upper
part of the dam. a T S(70%) (15–85%); b T S(90%) (5–95%); c T B (0.05 g); d T U (0.05 g)
Global damage index 0.5 0.5
Global damage index

0.4 0.4
0.3 0.3
0.2 0.2
0.1 PGA=0.30g 0.1 PGA=0.30g

PGA=0.25g PGA=0.25g
PGA=0.35g PGA=0.35g
0.0 0.0
0 5 10 15 20 25 0 5 10 15 20 25 30 35
0.5 0.5
Global damage index
Global damage index

0.4 0.4
0.3 0.3
0.2 0.2
PGA=0.30g PGA=0.30g
0.1 0.1
PGA=0.25g PGA=0.25g
PGA=0.35g PGA=0.35g
0.0 0.0
0 10 20 30 40 50 0 5 10 15 20 25
Fig. 5.12 Influence of strong motion duration on the global damage index measure for the dam.
a T S(70%) (15–85%); b T S(90%) (5–95%); c T B (0.05 g); d T U (0.05 g)
5.5.2 Damage Measures of Peak Displacement and Damage

Energy Dissipation
To analyze the effects of strong motion duration on the damage of concrete gravity
dams, Fig. 5.13 is generated to identify whether the peak displacement response of
the dam is related to strong motion duration, and Fig. 5.14 are generated by plotting
the damage energy dissipation of the dam imparted by the 20 records for a given level
of intensity. From Fig. 5.13, it can be observed that there is no significant relationship
between any of the three duration definitions (bracketed duration, uniform duration,
and significant duration) under consideration and the damage measure based on the
peak displacement of the structure. However, the results from the damage energy
dissipation measure are similar to that from the local and global index measures.
Stronger duration dependence might be found for higher loading levels, which would
cause greater deformation and damage energy dissipation in the dam. From Fig. 5.14,
it can also be noticed that the bracketed and uniform durations significantly increase
the scatter of damage measures around the trend lines.
Peak displacement (cm) 6 6
Peak displacement (cm)

5 5
4 4
3 3
2 2
PGA=0.30g PGA=0.30g
1 PGA=0.25g 1 PGA=0.25g
PGA=0.35g PGA=0.35g
0 0
0 5 10 15 20 25 0 5 10 15 20 25 30 35
(a) Significant duration, TS (70%) (15- 85%) (b) Significant, TS (90%) (5-95%)
6 6

5 5
4 4
3 3
2 2
PGA=0.30g PGA=0.30g
1 PGA=0.25g 1 PGA=0.25g
PGA=0.35g PGA=0.35g
0 0
0 10 20 30 40 50 0 5 10 15 20 25
Fig. 5.13 Influence of strong motion duration on peak displacement of the dam. a T S(70%) (15–
85%); b T S(90%) (5–95%); c T B (0.05 g); d T U (0.05 g)
Damage energy dissipation (kN.m)
80 80
PGA=0.30g PGA=0.30g
PGA=0.25g PGA=0.25g
60 PGA=0.35g 60 PGA=0.35g
40 40
20 20
0 0
0 5 10 15 20 25 0 5 10 15 20 25 30 35
(a) Significant duration, TS (70%)(15-85%) (b) Significant, TS (90%) (5-95%)
80 80
PGA=0.30g PGA=0.30g
PGA=0.25g PGA=0.25g
60 PGA=0.35g 60 PGA=0.35g
40 40
20 20
0 0
0 10 20 30 40 50 0 5 10 15 20 25
Fig. 5.14 Influence of strong motion duration on the damage energy dissipation measure for the
dam. a T S(70%) (15–85%); b T S(90%) (5–95%); c T B (0.05 g); d T U (0.05 g)
5.5.3 Identifying the Influence of Single Component

Duration on Damage Measures
The above analyses show that strong motion duration has significant influence on the
local damage of the dam upper part, the global damage, and the damage energy dissi-
pation. Hence, Fig. 5.15 mainly shows the comparison of the correlations between
strong motion duration and these damage measures during the modified real accelero-
grams with a PGA of 0.3 g. The trend lines in Fig. 5.15 show that the events with
longer duration generally cause greater damage on the dam than the shorter events
for the same level of spectral acceleration. The findings do indicate that the seismic
performance of concrete gravity dams would be improved by taking account of the
expected duration of earthquake ground motion. The comparison of the scatter of the
observations around the trend lines illustrated in Fig. 5.15 indicates that the signifi-
cant duration obtained based on the time between 15 and 85% of the Arias intensity,
is the duration measure which slightly more correlates with accumulated damage.
This is to be expected, as this measure is the best indication to assess the potential
damage energy of earthquake ground motion, as noted earlier. However, duration
measures with 5–95% significant duration, 0.05 g bracketed duration and 0.05 g
uniform duration generally increase the scatter of the damage measures around the
trend lines. In order to justify this assertion and emphasize the grade of interrelation
0.5
1.0
Global damame index
0.4
Local damame index
0.8
0.3
0.6
0.2 TS(70%) (15~85%)

0.4 TS(70%) (15~85%)
TS(90%) (5~95%) TS(90%) (5~95%)
0.2 TB (0.05g) 0.1 TB (0.05g)
TU (0.05g) TU (0.05g)
0.0 0.0
0 10 20 30 40 0 10 20 30 40
(a) The local damage index measure for the dam upper part (b) The global damage index measure
30
20
TS(70%) (15~85%)
10 TS(90%) (5~95%)
TB (0.05g)
TU (0.05g)
0
0 10 20 30 40
Duration (s)
(c) Damage energy dissipation measure
Fig. 5.15 Comparison of the correlations between strong motion duration with different definitions
and a the local damage index measure for the dam upper part; b the global damage index measure;
c damage energy dissipation measure
between strong motion duration and damage measures, the correlation coefficient
(R) has been used, which is defined by the following relation:
N
(xi − x̄)(yi − ȳ)
R = i=1
(5.7)
N 2 N
i=1 (x i − x̄) i=1 (yi − ȳ)
2
where x i and x̄ are the actual and its average values; yi and ȳ are the predicted and its
average values; and N denotes the number of data in the analysis. The values (R) of
correlation coefficients between strong motion durations with different definitions
and damage measures are tabulated in Table 5.4. Correlation coefficients between
absolute durations with different thresholds (0.05 and 0.10 g bracketed durations,
0.05 and 0.10 g uniform durations) and damage measures are also compared in
Table 5.4.
Through the correlation coefficients presented in Table 5.4, it can be seen that the
strong motion duration as defined by Trifunac and Brady has a very high correlation
with accumulated damage and damage energy dissipation. This is due to the fact
that their definition does take into account the seismic energy content. The two kinds
Table 5.4 Values (R) of correlation coefficient between strong motion duration and damage
measures
Damage measures Duration measures The value (R) of correlation coefficient
PGA(0.25 g) PGA(0.30 g) PGA(0.35 g)
The local damage index T S(70%) (15–85%) 0.942 0.980 0.940
measure for the dam T S(90%) (5–95%) 0.894 0.938 0.939
upper part
T B (0.05 g) 0.884 0.926 0.896
T U (0.05 g) 0.831 0.916 0.886
T B (0.10 g) 0.851 0.926 0.887
T U (0.10 g) 0.766 0.843 0.815
The global damage index T S(70%) (15–85%) 0.908 0.929 0.888
measure for the dam T S(90%) (5–95%) 0.892 0.917 0.886
T B (0.05 g) 0.872 0.909 0.877
T U (0.05 g) 0.814 0.872 0.860
T B (0.10 g) 0.522 0.371 0.307
T U (0.10 g) 0.476 0.345 0.285
The damage energy T S(70%) (15–85%) 0.920 0.930 0.913
dissipation measure for T S(90%) (5–95%) 0.880 0.913 0.898
the dam
T B (0.05 g) 0.865 0.881 0.884
T U (0.05 g) 0.823 0.813 0.828
T B (0.10 g) 0.486 0.429 0.437
T U (0.10 g) 0.447 0.374 0.382
of significant durations (T S(70%) (15–85%) and T S(90%) (5–95%)) provide almost the
same interrelation grade between the examined damage measures. It must be pointed
out that the duration measure predicted by the 15–85% significant duration has shown
a slightly higher correlation with accumulated damage when compared with the 5–
95% significant duration. However, both of them have a higher correlation when
compared with the absolute duration. Compare the correlations between damage
measures and absolute durations with different thresholds (0.05 and 0.10 g), it can
be seen that there is a very high correlation between damage measures and absolute
duration with 0.05 g threshold. On the other hand, a very poor correlation can be
observed between damage measures and absolute duration with 0.10 g threshold.
Among the examined definitions of strong motion duration, the 15–85% significant
duration has the strongest correlation with the local damage indices.
The analyses presented in this chapter show that for ground motions with compa-
rable modified real records, those with longer duration or greater magnitude generally
cause greater damage, as measured by local damage index and global damage index,
in concrete gravity dams. This is simply an empirical verification of expected struc-
tural behavior, which suggests that shorter strong motion duration records should not
be used as the seismic motion input in the earthquake safety evaluation of concrete
gravity dams. It should be noted that the correlation grade between the examined
strong motion durations and the damage indices presented in this study is valid for
the specific case of spectrum compatible accelerograms as presented in Sect. 5.2.
Further investigations will be carried out using the abundant strong motion records.
5.6 Conclusions
Researchers have shown increasing interest in relating the structural damage caused
by earthquakes to parameters associated with ground motion duration. In this chapter,
a methodology for quantifying the interrelationship between strong motion durations
and damage measures of concrete gravity dams is presented. The questions which
definition of the strong motion duration is the most useful indicator of earthquake
damage potential and which damage measure is the most useful index of the structural
performance are evaluated. The nonlinear dynamic response and seismic damage
process of concrete gravity dams during ground motions with different durations are
conducted according to the Concrete Damaged Plasticity (CDP) model. The local
damage index, global damage index, peak displacement response and damage energy
dissipation are considered as the damage measures.
20 real earthquake records with a wide range of durations, which are scaled to
match the target spectrum at a given damping level, are considered in this study. The
definitions of significant duration, bracketed duration and uniform duration are used
for characterizing the strong motion duration. From the calculated durations with
different definitions, it can be found that durations are generally poorly correlated
with the total duration. However, there is an approximate correlation between signif-
icant duration [T S(70%) (15–85%)] and other different definitions of duration for these
modified records.
The results show that, for concrete gravity dams, the influence of strong motion
duration (measured by different definitions) depends on both the peak ground accel-
eration and the damage measure employed. Although measure using peak response is
predominantly used in design and assessment applications because of their concep-
tual simplicity, the damage measure based on the peak displacement response exhibits
a poor correlation with the duration of the ground motion. However, strong motion
duration has been found to be significant to the cumulative damage of the dam.
Damage measures such as local damage index, global damage index and damage
energy dissipation are consistently greater for ground motions with longer duration.
Among the examined definitions of strong motion duration, the 15–85% signifi-
cant duration has the strongest correlation with the accumulated damage, as this
duration definition is intended to assess the potential damage energy of earthquake
ground motions. T S(70%) (15–85%) displays the highest correlation coefficient with
the local damage index for the dam upper part. Good correlation can be observed
between damage measures and absolute duration with 0.05 g threshold, whereas
absolute duration with 0.10 g threshold provides poor or fair correlation with damage
measures.
The influence of duration depends on a number of different factors including
the definition of strong motion duration, damage measure, strong motion record,
and structural model. These promising results are the starting point for further
exploration, of more seismic records and of other types of dams.
References
Ambraseys, N. N., & Sarma, S. K. (1967). The response of earth dams to strong earthquakes.
Geotechnique, 17(3), 181–213.
Arias, A. (1970). A measure of earthquake intensity. Seismic Design for Nuclear Power Plants RJ
Hanson, 438–483.
Arjun, C. R., & Kumar, A. (2011). Neural network estimation of duration of strong ground motion
using Japanese earthquake records. Soil Dynamics and Earthquake Engineering, 31(7), 866–872.
Barbosa, A. R., Ribeiro, F. L. A., & Neves, L. A. C. (2016). Influence of earthquake ground-
motion duration on damage estimation: application to steel moment resisting frames. Earthquake
Engineering & Structural Dynamics.
Belejo, A., Barbosa, A. R., & Bento, R. (2017). Influence of ground motion duration on damage
index-based fragility assessment of a plan-asymmetric non-ductile reinforced concrete building.
Engineering Structures, 151, 682–703.
Bolt, B. A. (1973). Duration of strong ground motion. In Proceedings 5th World Conference on
Earthquake Engineering (pp. 1304–1313), Rome, Italy.
Bommer, J. J., Hancock, J., & Alarcón, J. E. (2006). Correlations between duration and number
of effective cycles of earthquake ground motion. Soil Dynamics and Earthquake Engineering,
26(1), 1–13.
References 129
Bommer, J. J., Magenes, G., Hancock, J., & Penazzo, P. (2004). The influence of strong-motion
duration on the seismic response of masonry structures. Bulletin of Earthquake Engineering, 2(1),
1–26.
Bommer, J. J., & Martínez-Pereira, A. (1999). The effective duration of earthquake strong motion.
Journal of Earthquake Engineering, 3(2), 127–172.
Bommer, J. J., Stafford, P. J., & Alarcon, J. E. (2009). Empirical equations for the prediction of
the significant, bracketed, and uniform duration of earthquake ground motion. Bulletin of the
Seismological Society of America, 99(6), 3217.
Bravo-Haro, M. A., & Elghazouli, A. Y. (2018). Influence of earthquake duration on the response
of steel moment frames. Soil Dynamics and Earthquake Engineering, 115, 634–651.
Chandramohan, R., Baker, J. W., and Deierlein, G. G. (2015). Quantifying the influence of ground
motion duration on structural collapse capacity using spectrally equivalent records. Earthquake
Spectra.
Chandramohan, R., Baker, J. W., & Deierlein, G. G. (2016). Impact of hazard-consistent ground
motion duration in structural collapse risk assessment. Earthquake Engineering & Structural
Dynamics.
Cornell, C. A. (1997). Does duration really matter? Proceedings of the FHWA/NCEER Workshop on
the National Representation of Seismic Ground Motion for New and Existing Highway Facilities
(pp. 125–133).
COSMOS. (2019). Strong Motion Virtual Data Center, http://strongmotioncenter.org/vdc/scripts/
default.plx.
Greco, R., Vanzi, I., Lavorato, D., & Briseghella, B. (2019). Seismic duration effect on damping
reduction factor using random vibration theory. Engineering Structures, 179, 296–309.
Hancock, J., & Bommer, J. J. (2004). The influence of phase and duration of earthquake damage in
degrading structures. Canada 1990.
Hancock, J., & Bommer, J. J. (2006). A state-of-knowledge review of the influence of strong-motion
duration on structural damage. Earthquake Spectra, 22(3), 827–845.
Hancock, J., & Bommer, J. (2007). Using spectral matched records to explore the influence of strong-
motion duration on inelastic structural response. Soil Dynamics and Earthquake Engineering,
27(4), 291–299.
Hou, H., & Qu, B. (2015). Duration effect of spectrally matched ground motions on seismic demands
of elastic perfectly plastic SDOFS. Engineering Structures, 90, 48–60.
Husid, R. L. (1969). Analisis de terremotos: analisis general. Revista del IDIEM, 8(1), 21–42.
Iervolino, I., Manfredi, G., & Cosenza, E. (2006). Ground motion duration effects on nonlinear
seismic response. Earthquake Engineering and Structural Dynamics, 35(1), 21–38.
Kabir, M. R., Billah, A. H. M. M., & Alam, M. S. (2019). Seismic fragility assessment of a multi-
span RC bridge in Bangladesh considering near-fault, far-field and long duration ground motions.
Khaloo, A., Nozhati, S., Masoomi, H., & Faghihmaleki, H. (2016). Influence of earthquake record
truncation on fragility curves of RC frames with different damage indices. Journal of Building
Kiani, J., Camp, C., & Pezeshk, S. (2018). Role of conditioning intensity measure in the influence of
ground motion duration on the structural response. Soil Dynamics and Earthquake Engineering,
104, 408–417.
Kiani, J., Camp, C., & Pezeshk, S. (2019). The importance of non-spectral intensity measures on
the risk-based structural responses. Soil Dynamics and Earthquake Engineering, 120, 97–112.
Kunnath, S. K., & Chai, Y. H. (2004). Cumulative damage-based inelastic cyclic demand spectrum.
Lee, J., & Fenves, G. L. (1998). Plastic-damage model for cyclic loading of concrete structures.
Mahin, S. A. (1980). Effects of duration and aftershocks on inelastic design earthquakes. Istanbul,
5, 677–680.
Molazadeh, M., & Saffari, H. (2018). The effects of ground motion duration and pinching-degrading
behavior on seismic response of SDOF systems. Soil Dynamics and Earthquake Engineering,
114, 333–347.
Montejo, L. A., & Kowalsky, M. J. (2008). Estimation of frequency-dependent strong motion
duration via wavelets and its influence on nonlinear seismic response. Computer-Aided Civil and
Infrastructure Engineering, 23(4), 253–264.
Ou, Y., Song, J., Wang, P., Adidharma, L., Chang, K., & Lee, G. (2014). Ground motion dura-
tion effects on hysteretic behavior of reinforced concrete bridge columns. Journal of Structural
Engineering, 140(3), 4013065.
Pan, Y., Ventura, C. E., & Finn, W. D. L. (2018). Effects of ground motion duration on the seismic
performance and collapse rate of light-frame wood houses. Journal of Structural Engineering,
144(8), 4018112.
Pan, Y., Ventura, C. E., Finn, W. D. L., & Xiong, H. (2019). Effects of ground motion duration
on the seismic damage to and collapse capacity of a mid-rise woodframe building. Engineering
Structures, 197, 109451.
Pan, Y., Ventura, C. E., & Tannert, T. (2020). Damage index fragility assessment of low-rise light-
frame wood buildings under long duration subduction earthquakes. Structural Safety, 84, 101940.
PEER. (2019). Pacific Earthquake Engineering Research Center, http://peer.berkeley.edu/peer_g
Raghunandan, M., & Liel, A. B. (2013). Effect of ground motion duration on earthquake-induced
structural collapse. Structural Safety, 41, 119–133.
Reinoso, E., Ordaz, M., & Guerrero, R. U. L. (2000). Influence of strong ground-motion duration
in seismic design of structures (p. 1151).
Ruiz-García, J. (2010). On the influence of strong-ground motion duration on residual displacement
demands. Earthquakes and Structures, 1(4), 327–344.
Samanta, A., & Pandey, P. (2018). Effects of ground motion modification methods and ground
motion duration on seismic performance of a 15-storied building. Journal of Building Engi-
neering, 15, 14–25.
Sarieddine, M., & Lin, L. (2013). Investigation correlations between strong-motion duration and
structural damage. Structures Congress, 10, 2926–2936.
Shoji, Y., Tanii, K., & Kamiyama, M. (2005). A study on the duration and amplitude characteristics
of earthquake ground motions. Soil Dynamics and Earthquake Engineering, 25(7), 505–512.
Shome, N., Cornell, C. A., Bazzurro, P., & Carballo, J. E. (1998). Earthquakes, records, and nonlinear
responses. Earthquake Spectra, 14(3), 469–500.
Taflampas, I. M., Spyrakos, C. C., & Koutromanos, I. A. (2009). A new definition of strong motion
duration and related parameters affecting the response of medium-long period structures. Soil
Dynamics and Earthquake Engineering, 29(4), 752–763.
Takizawa, H., & Jennings, P. C. (1980). Collapse of a model for ductile reinforced concrete frames
under extreme earthquake motions. Earthquake Engineering & Structural Dynamics, 8(2), 117–
144.
Tombari, A., Naggar, M. H. E., & Dezi, F. (2017). Impact of ground motion duration and soil non-
linearity on the seismic performance of single piles. Soil Dynamics and Earthquake Engineering,
100, 72–87.
Trifunac, M. D., & Brady, A. G. (1975). A study on the duration of strong earthquake ground
motion. Bulletin of the Seismological Society of America, 65(3), 581–626.
Wang, C., Hao, H., Zhang, S., & Wang, G. (2018). Influence of ground motion duration on responses
of concrete gravity dams. Journal of Earthquake Engineering, 1–25.
Westergaard, H. M. (1933). Water pressures on dams during earthquakes. Trans. ASCE, 98, 418–432.
References 131
Xu, B., Wang, X., Pang, R., & Zhou, Y. (2018). Influence of strong motion duration on the seismic
performance of high CFRDs based on elastoplastic analysis. Soil Dynamics and Earthquake
Yaghmaei-Sabegh, S., Shoghian, Z., & Neaz Sheikh, M. (2014). A new model for the prediction of
earthquake ground-motion duration in Iran. Natural Hazards, 70(1), 69–92.
Youd, T. L., Idriss, I. M., Andrus, R. D., Arango, I., Castro, G., Christian, J. T., et al. (2001). Liquefac-
tion resistance of soils: summary report from the 1996 NCEER and 1998 NCEER/NSF workshops
on evaluation of liquefaction resistance of soils. Journal of Geotechnical and Geoenvironmental
Engineering, 127(10), 817–833.
16–27.
Zhang, H. Y., Zhang, L. J., Wang, H. J., & Guan, C. N. (2018). Influences of the duration and
frequency content of ground motions on the seismic performance of high-rise intake towers.
Engineering Failure Analysis, 91, 481–495.
Chapter 6
Integrated Duration Effects on Seismic
Performance of Concrete Gravity Dams
6.1 Introduction
Ground motions induced by earthquake events contain three translational compo-

nents, which can be directly recorded by strong-motion seismometers. When time
history analyses are performed to evaluate the seismic performance of important
infrastructures (e.g. high dams, bridges and tall buildings), the effects of ground
motion components should be considered, and as-recorded accelerograms including
multiple components are usually selected as the seismic excitation. For concrete
gravity dams, both the vertical and stream components of earthquake ground motions
are usually considered in two-dimensional models (Calayir and Karaton Calayir and
Karaton 2005; Omidi et al. 2013; Zhang et al. 2013a, b; Hariri-Ardebili and Saouma
2015; Wang et al. ; Hariri-Ardebili et al. 2016a, b), whereas stream, cross-stream
and vertical motions are usually selected as seismic excitations for three-dimensional
analyses (when required) (Wang et al. 2012; Arici et al. 2014). On the other hand,
the seismic response analysis of arch dams is, in general, conducted based on three-
dimensional models, in which three components of earthquake ground motions are
applied simultaneously (Engineering Manual 1110-2-6051 2003; Hariri-Ardebili and
Kianoush 2014; Hariri-Ardebili et al. 2016a, b).
Earthquake ground motions can be characterized by three main parameters (i.e.
amplitude, frequency content, and duration). As one of the major characteristics of
ground motions, the effects of strong motion duration on seismic demands imposed
on structures have been widely studied (Bommer et al. 2004; Raghunandan and Liel
2013; Zhang et al. 2013a, b; Ou et al. 2014; Song et al. 2014; Chandramohan et al.
2015; Hou and Qu 2015; Wang et al. 2015a, b, c; Kiani et al. 2018; Molazadeh
and Saffari 2018; Pan et al. 2018; Samanta and Pandey 2018; Zhang et al. 2018;
Kabir et al. 2019). A common argument is that the degree of influence of duration
on structural damage depends on several factors including the definition of strong
https://doi.org/10.1007/978-981-15-6194-8_6
134 6 Integrated Duration Effects on Seismic Performance …
motion duration, the type of structure examined, the strong motion record selected as
seismic excitations, and the demand measures used to quantify the damage (Hancock
and Bommer 2006).
An important issue encountered in the duration effect analysis is to determine
the strong motion duration of ground motions. As summarized in Ref (Bommer
and Martínez-Pereira 1999), more than 30 different strong motion duration defi-
nitions have been presented in the literature and without being exhaustive, some
contributions are worth mentioning. There are four typical and widely used defini-
tions of strong motion duration, i.e. (a) bracketed duration (Ambraseys and Sarma
1967; Bolt 1973); (b) uniform duration (Bommer et al. 2009); (c) effective duration
(Bommer and Martínez-Pereira 1999), and; (d) significant duration (Trifunac and
Brady 1975), and then classified by whether the amplitude or energy thresholds used
for their measurement are absolute or relative to the peak value in the recording.
Subsequently, several new definitions and prediction models of strong motion dura-
tion have been put forward. Subsequently, several new definitions and prediction
models of strong motion duration have been put forward (Kempton and Stewart
2006; Bommer et al. 2009; Taflampas et al. 2009; Yaghmaei-Sabegh et al. 2014).
It should be emphasized that all the aforementioned duration definitions can only
determine the strong motion duration based on one component of ground motions,
either horizontal seismic excitation or vertical one. However, multiple components
of earthquake ground motions are usually selected as seismic excitations in practice.
To account for the duration contributions of all ground motion components, Wang
et al. (2015a, b, c) proposed a concept of integrated duration based on the integrated
Husid diagram. Nonetheless, this definition of integrated duration can only calculate
the significant duration of multi-component ground motions.
This study aims to answer the question how to predict the integrated duration
when selecting multiple components of ground motions for structural response anal-
ysis. For this purpose, a general concept of integrated duration is proposed based
on the existing definitions of strong motion duration. Integrated Bracketed dura-
tion, integrated uniform duration, and integrated significant duration are chosen as
the duration measures. The peak displacement, maximum principal stress, damage
dissipation energy, and local damage index are adopted as the seismic demands.
The relationship between integrated and single horizontal or vertical durations is
investigated based on 20 as-recorded accelerograms with a wide range of duration.
A series of nonlinear dynamic analyses is performed to quantify the effects of inte-
grated duration and vertical seismic excitations on the nonlinear response of concrete
gravity dam-reservoir-foundation systems. The results reveal the correlation between
different integrated durations and damage measures.
6.2 A General Definition of Integrated Duration 135
6.2 A General Definition of Integrated Duration
6.2.1 Single Component Duration
As mentioned above, the most widely used measures for the strong motion duration
based on one component of ground motions can be classified into four categories: (a)
bracketed duration (T B ) (Ambraseys and Sarma 1967; Bolt 1973); (b) uniform dura-
tion (T U ) (Bommer et al. 2009); (c) effective duration (T E ) (Bommer and Martínez-
Pereira 1999); (d) significant duration (T S ) (Trifunac and Brady 1975). In this study,
the definition of effective duration is not considered since it is very difficult to obtain
the particular thresholds of the Arias intensity for each record. The bracketed duration
(T B ) and uniform duration (T U ) belong to the class of absolute duration definitions.
One specific acceleration threshold (i.e. 0.05 g) is considered for both the absolute
duration definitions in this study. In contrast, the significant duration (T S ) is a relative
duration concept in the sense that the significant duration is unaffected by the scaling
of ground motion records.
6.2.2 Integrated Duration of Multi-component Ground

Motions
In seismic performance analysis, multiple components of ground motions are usually

chosen as seismic excitations for structures. In order to bridge the gap between strong
motion duration and multi-component excitations, a general definition of integrated
duration is proposed based on the existing concepts of strong motion duration.
As well known, the destructiveness of ground motions can be characterized
by several intensity parameters. The Arias intensity I 0 is a popular energy-related
seismic intensity metric reflecting the total energy content of a ground motion. With
the Arias intensity being selected as the weighting function, the integrated duration is
proposed to be a weighted average of the corresponding single component duration.
Take two-component excitations for example, the proposed integrated duration (T I )
can be defined as
T H × I0H + T V × I0V
TI = (6.1)
I0H + I0V
where I0H and I0V are the Arias intensities of horizontal and vertical components,
respectively. T H and T V are the strong motion durations of horizontal and vertical
components, respectively. Note that T H and T V can be calculated based on any
aforementioned duration definitions.
For three-component excitations, the integrated duration (T I ) can also be easily
obtained as follows
T H1 × I0H1 + T H2 × I0H2 + T V × I0V

TI = (6.2)
I0H1 + I0H2 + I0V
where T H1 and T H2 are the strong motion durations of two perpendicular horizontal
components; I0H1 and I0H2 are the Arias intensities of the two horizontal ground
motions.
In this chapter, we focus on two-component ground motions (with horizontal
and vertical components) since the seismic response of concrete gravity dams is
investigated based on a 2D model. According to Eq. (6.1) and single component
duration definitions, integrated bracketed duration (TBI ), integrated uniform duration
(TUI ), and integrated significant duration (TSI ) can be obtained, serving as a unified
duration measure of two-component ground motions. For example, the integrated
bracketed duration (TBI ) can be calculated by the following relation
TBH × I0H + TBV × I0V

TBI = (6.3)
I0H + I0V
where TBH and TBV are the bracketed duration of horizontal and vertical components,
respectively. For illustrative purpose, the entire calculation process of the integrated
bracketed duration is shown in Fig. 6.1 for an as-recorded accelerogram.
Fig. 6.1 Illustration of the integrated bracketed duration

6.3 Database of as-Recorded Acceleration Records 137
6.3 Database of as-Recorded Acceleration Records
6.3.1 Accelerogram Selection and Integrated Duration

Prediction
This study focuses on how to determine the integrated durations and their effects
on the nonlinear dynamic demands imposed to concrete gravity dams. For this
purpose, as-recorded acceleration time histories with two components are needed for
performing nonlinear dynamic analyses. In this contribution, a total of 20 real earth-
quake records from around the world, covering a broad range of duration, are used
with each of which containing two accelerograms (horizontal and vertical compo-
nents). The relevant information of these earthquakes, obtained from the database
provided by the Consortium of Organizations for Strong Motion Observation Systems
(COSMOS) (COSMOS 2019), is summarized in Table 6.1. This information contains
the magnitude, the epicenter distance, the total duration T 0 , and the peak ground
acceleration (PGA).
In order to isolate and quantify the influence of duration on the nonlinear dynamic
response of concrete gravity dams, it is necessary to minimize the effects of PGA,
spectral amplitude and other ground motion characteristics. All horizontal accelero-
grams are normalized to have a PGA of 0.3 g, and the PGA of vertical accelerograms
is scaled to 0.2 g (two thirds of the horizontal PGA). The spectral acceleration of all
horizontal ground motions is scaled and adjusted to have a good match on average
to the target spectrum (as shown in Fig. 6.2) by using wavelets. Vertical accelero-
grams are also adjusted along with horizontal components. By doing so, the major
difference between these records is ensured to be the strong motion duration. Based
on the aforementioned definitions, both the single component and integrated strong
motion durations are calculated for all seismic excitations listed in Table 6.1, which
are summarized in Table 6.2.
6.3.2 Relationship Between Integrated and Single

Component Durations
In order to study whether the integrated durations have similar duration distribution as
their corresponding single component durations, the relationship between integrated
and single component durations is investigated in this section. Furthermore, the
correlation between horizontal and vertical component durations is also analyzed.
To measure the linear correlation between strong motion durations, the Pearson
correlation coefficient is calculated.
Based on the single component durations listed in Table 6.2, relationship between
horizontal and vertical component durations is shown in Fig. 6.3. The correlation
coefficients are also provided in this figure for each duration definition. As can
Table 6.1 List of as-recorded earthquake ground motions with two components (non-modified
records)
No. Earthquake Date Station M w Distance T 0 (s) Original record
location to fault PGA(cm/s2 )
(km) Horizontal Vertical
1 Koyna 11-12-1967 Koyna Dam 6.5 13.0 10.00 480.20 333.20
2 Northridge 17-01-1994 Los Angeles 6.7 35.8 20.00 310.10 131.90
Terrace
#24592
3 Northridge 17-01-1994 Epiphany 6.7 15.7 20.00 404.20 409.52
Litheran
Church #5353
4 Loma Prieta 18-10-1989 South St and 7.0 17.2 40.00 174.50 193.21
Pine Dr
#47524
5 Chi-Chi 20-09-1999 Taichung, 7.6 8.3 60.00 301.80 170.37
Taiwan
#TCU078
6 Petrolia 17-08-1991 General Store 6.0 11.7 10.00 488.70 164.30
#89156
Mulholland
Dr #5314
8 Loma Prieta 18-10-1989 Capitola, CA 7.0 15.9 39.82 390.80 500.10
#47125
9 Chi-Chi 20-09-1999 Taichung, 7.6 7.9 40.00 360.10 276.00
Taiwan
#TCU072
10 Landers 28-06-1992 Whitewater 7.3 28.6 56.55 124.92 110.97
Canyon
#5072
Dam #2141
12 Loma Prieta 18-10-1989 Santa Teresa 7.0 11.9 25.00 223.40 205.02
Hills #57563
13 Superstition 24-11-1987 Westmorland, 6.6 13.0 59.82 203.56 187.63
Hills CA #11369
14 Kern 21-07-1952 Lincoln 7.5 36.2 54.22 175.95 102.85
County School #1095
15 Landers 28-06-1992 Joshua Tree 7.3 10.0 40.00 278.40 177.73
#22170
16 San Simeon 22-12-2003 Caltrans 6.5 14.8 25.00 175.00 86.65
Bridge Grnds
#37737
(continued)

No. Earthquake Date Station M w Distance T 0 (s) Original record
location to fault PGA(cm/s2 )
(km) Horizontal Vertical
17 Parkfield 28-09-2004 USGS 6.0 14 20.00 170.00 86.00
Parkfield
Dense #02
18 Imperial 15-10-1979 Casa Flores 6.5 9.8 16.95 238.10 327.70
Valley Mexicali
#6619
19 Northridge 17-01-1994 Coldwater 6.7 12.5 21.86 313.30 256.25
Canyon
School #5309
20 Chi-Chi 20-09-1999 Taichung, 7.6 8.9 50.00 127.50 85.95
Taiwan
#TCU050
3.0
Target spectrum
2.5 Adjusted spectrum
Amplification factor
2.0
1.5
1.0
0.5
0.0
0 1 2 3 4
Period (s)
Fig. 6.2 Acceleration spectrum for scaled earthquake records
be seen, the horizontal component duration calculated using five definitions under
consideration is positively related to the corresponding vertical duration. In general,
this means that the longer the horizontal duration is, the larger the vertical duration
is. It can also be seen that the significant duration measure (TS(5−95%) ) gives much
higher correlation coefficient as compared with the uniform duration definition.
Figure 6.4 depicts the correlation between integrated durations calculated using
different definitions and their corresponding single component durations for hori-
zontal and vertical accelerograms. The values (R) of correlation coefficient are
summarized in Table 6.3. In order to investigate the influence of the PGA of vertical
140
Table 6.2 Strong motion durations obtained from modified records

No. Single component duration (s) Integrated duration (s)
Horizontal component Vertical component
H
Ts(15−85%) H
Ts(5−75%) H
Ts(5−95%) H
TB(0.05g) H
TU(0.05g) V
Ts(15−85%) V
Ts(5−75%) V
Ts(5−95%) V
TB(0.05 v I I I I I
g) TU(0.05 g) Ts(15−85%) Ts(5−75%) Ts(5−95%) TB(0.05 g) TU(0.05 g)
1 3.08 2.42 5.02 6.92 2.62 3.84 3.37 5.58 7.94 2.66 3.59 2.95 5.29 7.39 2.64
2 5.49 5.69 10.27 14.31 4.82 7.05 7.42 11.98 14.62 3.00 7.00 8.31 11.43 14.45 4.01
3 7.14 6.42 10.12 13.74 6.22 4.84 5.54 10.70 7.84 2.22 6.74 6.36 10.26 12.54 5.41
4 13.00 9.42 23.7 33.18 10.60 13.12 8.90 23.10 13.48 2.80 13.36 11.02 24.38 29.68 9.21
5 18.73 15.83 26.04 31.31 12.24 18.46 17.11 25.63 32.77 7.92 18.53 15.99 25.52 31.78 10.86
6 2.00 2.24 3.26 3.76 2.36 3.34 3.78 6.96 5.06 1.64 2.60 3.6 4.15 4.21 2.11
7 4.46 5.04 7.62 9.28 3.76 5.12 5.56 8.76 10.26 3.28 5.08 5.26 7.79 9.70 3.55
8 8.48 5.56 13.16 14.62 5.74 6.02 6.30 10.64 11.54 4.3 6.70 5.72 12.84 13.52 5.22
9 15.24 15.34 23.55 30.52 12.38 11.05 13.42 24.59 20.19 3.66 14.73 15.24 23.23 28.57 10.73
10 23.59 20.33 31.36 43.04 20.00 21.20 20.14 31.36 24.01 7.66 21.69 20.66 29.45 37.83 16.62
11 4.06 3.75 6.67 7.02 3.08 4.21 4.09 7.59 8.04 2.06 3.96 4.63 7.38 7.36 2.74
12 6.60 6.24 9.32 16.28 6.62 6.38 7.30 10.74 11.52 3.12 6.48 6.48 9.75 15.17 5.80
13 8.82 9.98 17.04 23.24 7.44 11.38 11.02 15.92 17.34 5.62 10.10 12.04 17.86 21.17 6.80
14 13.86 10.50 28.86 33.84 8.50 20.26 15.42 33.58 39.24 6.44 16.30 11.62 30.20 36.00 7.68
15 21.48 21.58 26.08 31.44 14.14 22.28 22.50 26.34 29.86 9.66 21.51 22.72 26.38 30.91 12.62
16 5.86 5.05 10.19 10.97 3.28 7.79 6.75 13.17 14.23 3.74 6.73 6.69 12.02 12.44 3.49
17 6.41 5.83 11.63 16.58 5.93 8.58 7.20 14.76 19.72 6.52 7.71 7.36 13.65 18.18 6.23
18 6.65 5.71 10.56; 16.27 7.02 8.26 8.09 12.96 15.38 5.62 7.31 6.16 11.33 15.99 6.58
19 8.20 7.00 15.86 18.82 6.30 6.62 5.86 11.20 11.52 3.56 8.28 6.84 16.26 16.83 5.55
20 17.62 18.40 25.77 33.52 18.61 9.51 9.33 18.17 20.87 5.00 17.63 20.37 26.62 31.17 16.08
6 Integrated Duration Effects on Seismic Performance …
40 TS(15-85%)
TS(5-75%)
Vertical component duration (s)

35 TS(5-95%)
TB(0.05g) R=0.946
30
TU( 0.05g)
25
R=0.774
20
15 R=0.887
10 R=0.897
5
R=0.637
0
0 10 20 30 40 50
Horizontal component duration (s)
Fig. 6.3 Correlation between horizontal and vertical component durations
accelerograms on the correlation between integrated and single component dura-

tions, the vertical accelerograms are also scaled to have the same PGA with that of
horizontal components. The corresponding correlation coefficients are also tabulated
in Table 6.3. As shown, the integrated durations predicted by the proposed general
definition are positively correlated with their corresponding single component dura-
tions (both horizontal and vertical components). Overall, the correlation coefficients
for the horizontal component are higher than those of the vertical component. This
is because that the Arias intensity is selected as the weighting function. Due to the
seismic characteristics of the selected records, the average value of the Arias intensity
for horizontal accelerations is greater than that of vertical accelerations. In addition,
it can be seen that the difference of the correlation coefficient between horizontal
and vertical components relies on the adopted duration definition. More specifically,
the correlation difference is small for significant and bracketed durations, whereas
uniform duration gives much larger correlation difference. When the PGA of vertical
accelerograms is changed from two thirds of the horizontal PGA to full horizontal
PGA, a certain decline in the correlation coefficient between integrated and hori-
zontal durations is observed, whereas the correlation coefficient between integrated
and vertical durations slightly increases.
6.4 Influence of Two-Component Ground Motions

on Nonlinear Dynamic Response
In order to identify the effects of integrated duration and two-component seismic

excitations on the structural response, the Koyna concrete gravity dam is selected
as the representative numerical example. In order to model the dam-reservoir-
foundation interaction system, the Lagrangian finite element formulation is utilized
Single component duration (s)

25 Horizontal component Horizontal component
25
Vertical component Vertical component
R=0.990
20 20
R=0.988
15 15
10 10
R=0.924 R=0.885
5 5
0 0
5 10 15 20 5 10 15 20
Integrated duration (s) Integrated duration (s)
(a) Significant duration ( TS(15 -85%) ) (b) Significant duration ( TS(5 - 75%) )
35 40

Horizontal component Horizontal component
30 Vertical component Vertical component
25 30
R=0.989
20
20
15 R=0.948
10 10 R=0.853
R=0.995
5
0 0
5 10 15 20 25 30 10 20 30 40
(c) Significant duration ( TS(5 -95%) ) (d) Bracketed duration ( TB(0.05g) )
Horizontal component
25
Vertical component
20
15 R=0.997
10
5
R=0.672
0
0 5 10 15 20
Integrated duration (s)
(e) Uniform duration ( TU(0.05g) )
Fig. 6.4 Correlation between integrated durations and single component durations for five duration
definitions: a significant duration (TS(15−85%) ); b significant duration (TS(5−75%) ); c significant
duration (TS(5−95%) ); d bracketed duration (TB(0.05 g) ); e uniform duration (TU(0.05 g) )
to take into account the fluid-solid interaction. In this formulation, the displacements
are supposed to be the unknown variables for both the water and dam/foundation
domains. The Concrete Damaged Plasticity (CDP) model including the strain hard-
ening/softening behavior is employed to capture crack initiation and propagation in
the dam body. The geometry and finite element (FE) discretization of the Koyna
dam–reservoir–foundation system are illustrated in Sect. 2.5 (Chap. 2). The material
parameters for concrete, foundation rock, and water are the same as Table 2.1.
Firstly, the influence of vertical component ground motions on nonlinear dynamic
response of the Koyna gravity dam-reservoir-foundation system under typical earth-
quakes is investigated. The response history of the Koyna dam subjected to two repre-
sentative earthquakes, i.e. Loma Prieta #47524 (No. 4) and Kern County #1095 (No.
6.4 Influence of Two-Component Ground Motions on Nonlinear Dynamic Response 143
Table 6.3 Values (R) of correlation coefficients between integrated durations and single component
durations
Correlation The PGA of vertical accelerograms The PGA of vertical accelerograms
coefficients is scaled to two thirds of the PGA of is the same with that of horizontal
horizontal components components
Horizontal Vertical Horizontal Vertical
component component component component
I
Ts(15−85%) 0.988 0.924 0.980 0.959
I
Ts(5−75%) 0.990 0.885 0.985 0.958
I
Ts(5−95%) 0.995 0.948 0.989 0.982
I
TB(0.05 0.989 0.853 0.973 0.896
g)
I
TU(0.05 0.997 0.672 0.991 0.723
g)
14), are selected for examination, since the vertical component of Loma Prieta #47524
earthquake increases the structural demands, while that of Kern County #1095 earth-
quake reduces the demands. The examined responses include displacement, stress,
dissipation energy and crack patterns of the dam body. Subsequently, the influence
of vertical seismic excitations on the seismic performance of Koyna concrete gravity
dam-reservoir-foundation system is quantified by using 20 as-recorded seismic
ground motions.
6.4.1 Displacement Response
The horizontal displacement histories at the crest of the dam subjected to two repre-
sentative earthquakes are shown in Fig. 6.5. Both the results for single and two-
component earthquake excitations are presented. Here, the positive and negative
signs of the displacement value represent the downstream and upstream directions,
respectively. It is evident that although the maximum absolute displacement is slightly
increased (Fig. 6.5a) or reduced (Fig. 6.5b) when both the horizontal and vertical
4 4
Dam crest Dam crest
Displacement (cm)
Displacement (cm)
2 2
0 0
-2 -2
Horizontal Horizontal
-4 Horizontal+Vertical -4 Horizontal+Vertical
0 10 20 30 40 0 5 10 15 20 25
Time (s) Time (s)
(a) (b)
Fig. 6.5 Horizontal displacement histories at the crest of the dam under two representative ground
motions. a Loma Prieta #47524; b Kern County #1095
excitations are considered, the overall displacement histories for two-component

excitations almost coincide with that when only horizontal component is considered.
The displacement response processes are less likely to be affected by the vertical exci-
tations when the PGA of vertical accelerations is scaled to be two thirds of horizontal
PGA.
6.4.2 Stress Response
Figure 6.6 shows the maximum principal stress history at the dam heel and near
the change in the slope of the downstream face of the dam under two representa-
tive ground motions. It can be seen that the vertical ground motion has significant
influence on the response history of maximum principal stress. At the beginning, the
dynamic stress responses obtained from two-component excitations are close to those
under horizontal ground motions alone. The subsequent stress responses obtained
from the single and two-component seismic excitations deviate from each other.
Overall, the stress response under two-component excitations is slightly smaller
than that subjected to single-component excitations.
3 3
2 Horizontal+Vertical 2 Horizontal+Vertical
1 1
Near the change
0 0 in downstream slope
Dam heel
-1 -1
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)
(a)
3 3
Near the change

in downstream slope
2 2
Dam heel
1 1
0 0
Horizontal+Vertical Horizontal+Vertical
-1 -1
0 5 10 15 20 25 0 5 10 15 20 25
Time (s) Time (s)
(b)
Fig. 6.6 Maximum principal stresses at the dam heel and near the change in downstream slope of
the dam under two representative ground motions. a Loma Prieta #47524; b Kern County #1095
Damage dissipation energy (kN.m)

25 25
20 20
Horizontal
15 Horizontal+Vertical 15 Horizontal
Horizontal+Vertical
10 10
5 5
0 0
0 10 20 30 40 0 5 10 15 20 25
Time (s) Time (s)
(a) (b)
Fig. 6.7 Damage dissipation energy of the dam under two representative ground motions. a Loma
Prieta #47524; b Kern County #1095
6.4.3 Damage Dissipation Energy Response
Figure 6.7 shows the damage-induced energy dissipation curves for the entire dam
body subjected to single and two-component ground motions. The dash lines repre-
sent the dissipated energy of the dam subjected to both horizontal and vertical accel-
erations, whereas the solid lines denote that under the horizontal excitations alone.
As can be seen from Fig. 6.7, the vertical seismic excitation has a certain influence on
the damage dissipation energy. The final damage dissipation energy is 23.56 kNm for
the two-component Loma Prieta (#47524) earthquake, whereas a value of 21.15 kNm
is obtained when only the horizontal component is considered. On the other hand,
the final damage dissipation energy obtained from the Kern Count record (#1095) is
23.55 and 18.99 kNm for single and two-component excitations, respectively.
6.4.4 Damage Analysis
Seismic damage profiles of Koyna dam resulted from the two representative ground
motions are depicted in Fig. 6.8. Both the damage profiles for single and two-
component earthquake excitations are presented for comparison purpose. The
contour value between 0 and 1 indicates the concrete damage status ranging from
intact to fully damaged material. One can see that although the vertical excitation
does not significantly affect the dynamic displacement response, the finial cracking
profile may be changed a lot by the vertical excitation. It should be noted that the
inclination of the cracking profile could affect the stability of the separated upper
block of the dam body when penetrated cracks are formed under strong ground
motions.
Horizontal Horizontal and vertical
(a)
Horizontal Horizontal and vertical
(b)
Fig. 6.8 Cracking profiles of dam body under two representative ground motions. a Loma Prieta
#47524; b Kern County #1095
6.4.4.1 Identifying the Influence of Vertical Seismic Excitations
To quantify the influence of vertical seismic excitations on the nonlinear response

of Koyna concrete gravity dam-reservoir-foundation system, a series of nonlinear
dynamic analyses are carried out, with the dam-reservoir-foundation system
subjected to the 20 as-recorded seismic ground motions. Note that each ground
motion includes both the horizontal and vertical components. For comparison
purpose, the nonlinear response of the dam-reservoir-foundation system subjected
to horizontal excitations alone is also examined. Three engineering demand parame-
ters are considered in this study, these being peak displacement, damage dissipation
energy, and local damage index. The local damage index is defined as the ratio of
the length of a cracking path to the total cross-sectional length along the path.
The results of maximum horizontal displacements at the crest, damage dissipa-
tion energy, and local damage index are compared in Fig. 6.9 for single (horizontal
7
Single component (H)
Two components (H+V)
6
Displacement (cm)
3.73 cm
4
3 3.63 cm
0 5 10 15 20
No.
(a)
40

30
20 15.34 kN.m
10
14.40 kN.m
0
0 5 10 15 20
No.
(b)
1.0
0.8
Local damage index
0.6 0.49
0.4
0.48
0.2
0.0
0 5 10 15 20
No.
(c)
Fig. 6.9 Effects of vertical seismic excitations on the nonlinear dynamic response of Koyna concrete
gravity dam: a maximum absolute horizontal displacements at dam crest; b damage dissipation
energy; c local damage index for the upper part
excitation alone) and two-component (horizontal and vertical excitations) ground

motions. The black solid and magenta dashed lines represent the average response of
the dam under single and two-component seismic excitations, respectively. It is very
interesting to see that the vertical ground motions do not always increase the seismic
demands. As can be further seen, the nonlinear response of the dam is mainly domi-
nated by the horizontal excitations. The average values of the maximum horizontal
displacement (Fig. 6.9a), damage dissipation energy (Fig. 6.9b), and local damage
index measure (Fig. 6.9c) predicted by using two-component seismic excitations are
higher than those with horizontal excitations. For example, the average local damage
index for the upper part of the dam under two-component seismic excitations is 0.49,
while the corresponding value for horizontal accelerograms is 0.48. In this case, the
vertical excitations only increase this structure demand very slightly. It should be
noted that the observation on the average demand increase due to vertical excitations
is obtained from the response of the Koyna gravity dam subjected to a specific set
of earthquake ground motion. Thus, it might not apply to other gravity dams and
ground motions.
6.5 Correlation Between Integrated Durations and Damage

Measures
In the literature, many studies have been conducted concerning the effects of strong
motion duration on the nonlinear dynamic response of concrete gravity dams (Ou
et al. 2014; Song et al. 2014). However, most of the past studies are based on hori-
zontal seismic excitations and the corresponding single component durations. Now
we have a general definition of integrated duration that accounts for the duration
contributions from all components of ground motions. This section aims to quan-
tify the correlation between the general integrated duration and seismic demands
of Koyna concrete gravity dam-reservoir-foundation system, considering both the
horizontal and vertical ground motions. To the best knowledge of the authors, this
study presents the first reference in this respect.
In order to present a comprehensive study on the effects of integrated duration,
we consider five integrated durations including both relative (15–85, 5–75 and 5–
95% significant durations) and absolute (0.05 g bracketed and uniform durations)
definitions and three damage measures in terms of peak displacement, damage dissi-
pation energy and local damage index. The correlation between employed integrated
durations and damage measures is shown as data pairs in Figs. 6.10, 6.11 and 6.12.
Straight trend line, least-squares fitted to the data points, is displayed for each pair of
integrated duration definition and damage measure, aiming at identifying the possible
tendency.
As can be seen from Figs. 6.10, 6.11 and 6.12, integrated durations based on
different definitions are all positively correlated with the adopted damage measures.
To wit, the nonlinear dynamic response of the Koyna dam under two-component
6.5 Correlation Between Integrated Durations and Damage Measures 149
7 7
TsI(15-85%) I
TB(0.05g)
I
6 Ts(5 -75%) 6 I
TU(0.05g)
Displacement (cm)
Displacement (cm)
I
Ts(5 -95%)
R=0.783 R=0.778
R=0.701
5 R=0.689 5
R=0.790
4 4
3 3
2 2
0 10 20 30 40 0 10 20 30 40
(a) (b)
Fig. 6.10 Correlation between integrated duration and peak displacement at the dam crest: a relative
durations; b absolute durations
40 Damage dissipation energy (kN.m) 40

35 R=0.882 35
30 R=0.840 30
25 R=0.809 25 R=0.803 R=0.836
20 20
15 15
10 TsI(15- 85 %) 10
I I
Ts(5
5 - 75%)
5 TB(0.05g)
I I
Ts(5 -95%) TU(0.05g)
0 0
0 10 20 30 40 0 10 20 30 40
(a) (b)
Fig. 6.11 Correlation between integrated duration and damage dissipation energy: a relative
durations; b absolute durations
1.0 R=0.885 1.0 R=0.864
0.8 R=0.812 0.8

Damage index
Damage index
R=0.867
R=0.858
0.6 0.6
0.4 0.4
0.2 TsI(15-85%) 0.2

I I
Ts(5 -75%) TB(0.05g)
0.0 I
Ts(5
0.0 I
TU(0.05g)
-95%)
0 10 20 30 40 0 10 20 30 40
(a) (b)
Fig. 6.12 Correlation between integrated duration and local damage index: a relative durations;
b absolute durations
seismic excitations with longer integrated duration is in general larger than that
under shorter earthquake events for the same level of PGA.
The correlation coefficients R between integrated durations and damage measures
are also given in Figs. 6.10, 6.11 and 6.12. It is found that 15–85% integrated signif-
I
icant duration (Ts(15−85%) ) has the strongest correlation with the adopted damage
measures among the examined five integrated duration definitions. But there is
no homogeneous trend for other four integrated duration definitions. For peak
displacement demand (as shown in Fig. 6.10), 5–75% integrated significant duration
I
(Ts(5−75%) ) exhibits a higher correlation compared with 5–95% integrated significant
I
duration (Ts(5−95%) ). Furthermore, the correlation coefficient obtained from 0.05 g
I
integrated uniform duration (TU(0.05g) ) is higher than that of 0.05 g integrated brack-
I
eted duration (TB(0.05g) ). However, an opposite tendency is observed for damage
dissipation energy (Fig. 6.11) and local damage index (Fig. 6.12) demands: 5–95%
I
integrated significant duration (Ts(5−95%) ) shows slightly higher correlation than 5–
I
75% integrated significant duration (Ts(5−75%) ) and 0.05 g integrated bracketed dura-
I I
tion (TB(0.05g) ) is better than 0.05 g integrated uniform duration (TU(0.05g) ) in terms
of the degree of correlation. Another interesting observation is that the correlation
coefficients for peak displacement measure are lower than those for damage dissi-
pation energy and local damage index. It means that the integrated duration could
affect the accumulative damage more than the peak displacement response.
For comparison purpose, the correlation between horizontal durations and damage
measures under horizontal seismic excitations alone is provided in Fig. 6.13. It can
be seen from Fig. 6.13a that the peak displacement is mildly-to-weakly correlated
with horizontal durations from a statistical point of view. The correlation coeffi-
cients obtained from horizontal seismic excitations are lower than those from two-
component ground motions. In addition, the slope of trend lines between peak
displacements and horizontal durations is very gentle. The maximum slopes of the
trend lines, which represent the degree of influence, are 0.049 and 0.165 for single
and two-component seismic excitations, respectively. This observation indicates a
weaker influence of horizontal duration on peak displacement demand, as compared
with that of integrated duration under two-component seismic excitations.
On the other hand, the other damage measures such as damage dissipation energy
and local damage index are considerably greater for ground motions with longer
duration. Correlation coefficients between horizontal duration and these two damage
measures (damage dissipation energy and local damage index) are slightly higher
than those of integrated duration (Fig. 6.13b–c) under both the horizontal and vertical
excitations. This is because that the vertical excitations may increase or reduce the
damage measures, which could affect the correlation between integrated durations
and damage measures.
It is also interesting to find that among the examined horizontal duration defini-
tions, the significant horizontal duration associated with 15–85% of the Arias inten-
sity exhibits the strongest correlation with the employed damage measures. This
6.5 Correlation Between Integrated Durations and Damage Measures 151
7
H
Ts(15 -85%)
H
Ts(5 - 75%)
6
Displacement (cm) H
Ts(5 - 95%)
H
TB(0.05g)
5 H
TU(0.05g)
R=0.629 R=0.669 R=0.650
R=0.540
4
R=0.626
3
2
0 10 20 30 40
Horizontal duration (s)
(a)
40
35 R=0.930
R=0.913
30
25
R=0.861
20 R=0.813
H
Ts(15 -85%)
15 R=0.881 H
Ts(5 - 75%)
10 H
Ts(5 - 95%)
5 H
TB(0.05g)
H
TU(0.05g)
0
0 10 20 30 40
(b)
R=0.980
1.0
R=0.938
0.8 R=0.916
Damage index
R=0.926
R=0.928
0.6
H
0.4 Ts(15 -85%)
H
Ts(5 - 75%)
0.2 H
Ts(5 - 95%)
H
TB(0.05g)
0.0 H
TU(0.05g)
0 10 20 30 40
(c)
Fig. 6.13 Correlation between horizontal duration and damage measures under single horizontal
seismic excitations: a peak displacement; b damage dissipation energy; c Local damage index
observation is consistent with what we found for the integrated duration. On the
other hand, the 5–75% significant horizontal duration has the weakest correlation
with damage measures among the examined three significant duration definitions.
6.6 Conclusions
Most previous studies with respect to the influence of strong motion duration on the
nonlinear dynamic response of structures are based on horizontal seismic excitations
and the corresponding single component durations. In contrast, this study investigates
the integrated duration effects on concrete gravity dam-reservoir-foundation systems
subjected to both horizontal and vertical ground motions. The novelty of this work
lies in the proposal of a general duration definition to take into account the duration
contributions of all ground motion components. The unified integrated duration can
be calculated based on any existing concept of strong motion duration. In this study,
we consider integrated bracketed, uniform, and significant durations to measure dura-
tions for two-component ground motions. 20 as-recorded earthquake records with a
wide range of durations are applied to Koyna dam-reservoir-foundation system, in
order to quantify the influence of vertical excitation and integrated durations. The
following conclusions can be drawn from the study:
(1) The horizontal duration computed by the selected duration definitions is posi-
tively correlated to the corresponding vertical duration. Moreover, integrated
durations are also positively correlated with both the horizontal and vertical
durations. The correlation coefficients between integrated and horizontal
durations are larger than those between integrated and vertical durations.
(2) The displacement response is very slightly affected by vertical excitations,
whereas vertical excitations have a certain influence on the history of maximum
principal stress and damage dissipation energy. The cracking profile may also be
changed by the vertical motions. It is also found that the vertical motions do not
always increase the seismic demands such as peak displacement, accumulated
local damage, and dissipation energy.
(3) Nonlinear dynamic response of the Koyna dam under two-component seismic
excitations with longer integrated duration is in general greater than that under
ground motions with shorted integrated duration. 15-85% integrated signifi-
cant duration has the strongest correlation with the examined damage measures
among the five integrated duration definitions.
(4) Although horizontal duration is mildly-to-weakly correlated with peak displace-
ment under single seismic excitations, the slope of trend lines between horizontal
durations and peak displacements is very gentle. Compared with the horizontal
duration, the integrated duration exhibits higher correlation and more significant
influence on the peak displacement demand of the Koyna concrete gravity dam
subjected to two-component ground motions.
6.6 Conclusions 153
Based on the results of this investigation, it is suggested that the effects of inte-
grated duration should be taken into account for seismic performance assessment of
concrete gravity dams.
References
Ambraseys, N. N., & Sarma, S. K. (1967). The response of earth dams to strong earthquakes.
Geotechnique, 17(3), 181–213.
Arici, Y. N., Binici, B., & Aldemir, A. (2014). Comparison of the expected damage patterns from
two- and three-dimensional nonlinear dynamic analyses of a roller compacted concrete dam.
Structure and Infrastructure Engineering, 10(3), 305–315.
Bolt, B. A. (1973). Duration of strong ground motion. In Proceedings 5th World Conference on
Earthquake Engineering, Rome, Italy (pp. 1304–1313).
Bommer, J. J., Magenes, G., Hancock, J., & Penazzo, P. (2004). The influence of strong-motion
duration on the seismic response of masonry structures. Bulletin of Earthquake Engineering, 2(1),
1–26.
Bommer, J. J., Stafford, P. J., & Alarcon, J. E. (2009). Empirical equations for the prediction of
the significant, bracketed, and uniform duration of earthquake ground motion. Bulletin of the
Seismological Society of America, 99(6), 3217.
Bommer, J. J., & Martínez-Pereira, A. (1999). The effective duration of earthquake strong motion.
Journal of Earthquake Engineering, 3(2), 127–172.
Calayir, Y., & Karaton, M. (2005). A continuum damage concrete model for earthquake analysis
857–869.
Chandramohan, R., Baker, J. W., & Deierlein, G. G. (2015). Quantifying the influence of ground
motion duration on structural collapse capacity using spectrally equivalent records. Earthquake
Spectra.
COSMOS. (2019). Strong Motion Virtual Data Center. http://strongmotioncenter.org/vdc/scripts/
default.plx.
Engineering Manual 1110-2-6051. (2003). Time-history dynamic analysis of concrete hydraulic
structures. In Structural performance and damage criteria. U.S. Army Corps of Engineers
(USACE).
Hancock, J., & Bommer, J. J. (2006). A state-of-knowledge review of the influence of strong-motion
duration on structural damage. Earthquake Spectra, 22(3), 827–845.
Hariri-Ardebili, M. A., Furgani, L., Meghella, M., & Saouma, V. E. (2016a). A new class of seismic
damage and performance indices for arch dams via ETA method. Engineering Structures, 110,
145–160.
Hariri-Ardebili, M. A., Seyed-Kolbadi, S. M., & Kianoush, M. R. (2016b). FEM-based parametric
Hariri-Ardebili, M. A., & Kianoush, M. R. (2014). Integrative seismic safety evaluation of a high
concrete arch dam. Soil Dynamics and Earthquake Engineering, 67, 85–101.
Hariri-Ardebili, M. A., & Saouma, V. (2015). Quantitative failure metric for gravity dams.
Hou, H., & Qu, B. (2015). Duration effect of spectrally matched ground motions on seismic demands
of elastic perfectly plastic SDOFS. Engineering Structures, 90, 48–60.
Kempton, J. J., & Stewart, J. P. (2006). Prediction equations for significant duration of earthquake
ground motions considering site and near-source effects. Earthquake Spectra, 22(4), 985–1013.
Kiani, J., Camp, C., & Pezeshk, S. (2018). Role of conditioning intensity measure in the influence of
ground motion duration on the structural response. Soil Dynamics and Earthquake Engineering,
104, 408–417.
Molazadeh, M., & Saffari, H. (2018). The effects of ground motion duration and pinching-degrading
behavior on seismic response of SDOF systems. Soil Dynamics and Earthquake Engineering,
114, 333–347.
63, 80–97.
Ou, Y., Song, J., Wang, P., Adidharma, L., Chang, K., & Lee, G. (2014). Ground motion dura-
tion effects on hysteretic behavior of reinforced concrete bridge columns. Journal of Structural
Engineering, 140(3), 4013065.
Pan, Y., Ventura, C. E., & Finn, W. D. L. (2018). Effects of ground motion duration on the seismic
performance and collapse rate of light-frame wood houses. Journal of Structural Engineering,
144(8), 4018112.
Raghunandan, M., & Liel, A. B. (2013). Effect of ground motion duration on earthquake-induced
structural collapse. Structural Safety, 41, 119–133.
Samanta, A., & Pandey, P. (2018). Effects of ground motion modification methods and ground
motion duration on seismic performance of a 15-storied building. Journal of Building Engi-
neering, 15, 14–25.
Song, R., Li, Y., & van de Lindt, J. W. (2014). Impact of earthquake ground motion characteristics
on collapse risk of post-mainshock buildings considering aftershocks. Engineering Structures,
81, 349–361.
Taflampas, I. M., Spyrakos, C. C., & Koutromanos, I. A. (2009). A new definition of strong motion
duration and related parameters affecting the response of medium-long period structures. Soil
Dynamics and Earthquake Engineering, 29(4), 752–763.
Trifunac, M. D., & Brady, A. G. (1975). A study on the duration of strong earthquake ground
motion. Bulletin of the Seismological Society of America, 65(3), 581–626.
Wang, G., Wang, Y., Lu, W., Zhou, C., Chen, M., & Yan, P. (2015a). XFEM based seismic potential
Wang, G., Wang, Y., Lu, W., Zhou, W., & Zhou, C. (2015b). Integrated duration effects on
seismic performance of concrete gravity dams using linear and nonlinear evaluation methods.
Soil Dynamics and Earthquake Engineering, 79, 223–236.
Wang, G., Zhang, S., Zhou, C., & Lu, W. (2015c). Correlation between strong motion durations
and damage measures of concrete gravity dams. Soil Dynamics and Earthquake Engineering, 69,
148–162.
Wang, H., Feng, M., & Yang, H. (2012). Seismic nonlinear analyses of a concrete gravity dam with
3D full dam model. Bulletin of Earthquake Engineering, 10(6), 1959–1977.
Yaghmaei-Sabegh, S., Shoghian, Z., & Neaz Sheikh, M. (2014). A new model for the prediction of
earthquake ground-motion duration in Iran. Natural Hazards, 70(1), 69–92.
Zhang, H. Y., Zhang, L. J., Wang, H. J., & Guan, C. N. (2018). Influences of the duration and
frequency content of ground motions on the seismic performance of high-rise intake towers.
Engineering Failure Analysis, 91, 481–495.
Zhang, S., Wang, G., & Yu, X. (2013b). Seismic cracking analysis of concrete gravity dams with
Chapter 7
Damage Demand Assessment of Concrete
Gravity Dams Subjected
to Mainshock-Aftershock Seismic
Sequences
7.1 Introduction
In the performance-based design of concrete dams, the dams are designed to meet the
normal operating requirement when subjected to frequent earthquakes, to undergo
repairable damage during a design basis earthquake, and to prevent uncontrollable
failure during a calibrated earthquake (Chinese 2011). In China, these earthquake
loadings, without exception, are specified in the current seismic codes as single
events (Chinese 2001). However, earthquakes, in general, do not occur as a single
event but as a series of shocks. A large mainshock usually triggers numerous after-
shocks in a short period. For example, the 2008 Wenchuan earthquake had a main-
shock measuring 8.0 M w followed by approximately five aftershocks of magnitude
greater than 6.0 (Zhai et al. 2013). Mainshock–aftershock seismic sequences have
also been observed in many dam projects, as shown in Table 7.1 (Alliard 2006).
When a dam has been damaged by a mainshock, strong aftershocks may cause addi-
tional cumulative damage to the dam before it is repaired, increasing the risk of a
major damage or collapse. Thus, mainshock–aftershock seismic sequences represent
a realistic situation that requires special treatment in the seismic design.
The seismic performance of structures that are subjected to multiple earthquakes
has gained a great deal of attention from researchers. Several investigations have
been conducted to study the damage potential of aftershocks. Those studies vary
in several aspects: the type of structures (reinforced concrete buildings (Liolios
et al. 2015; Zafar and Andrawes 2015), reinforced concrete containments (Zhai et al.
2015, 2017), bridges (Jeon et al. 2016), etc.), model representation (single-degree-of-
freedom (Hatzigeorgiou and Beskos 2009; Goda 2012)to multi-degree-of-freedom
(Liolios et al. 2015; Zafar and Andrawes 2015)), generation of mainshock–after-
shock seismic sequences (as-recorded sequences (Zafar and Andrawes 2015), arti-
ficial records (Goda and Taylor 2012; Han et al. 2015), actual records with scaling
(Hatzigeorgiou 2010a; Hatzivassiliou and Hatzigeorgiou 2015), etc.), and objective
(seismic demand (Hatzigeorgiou 2010b; Goda and Salami, 2014), life-cycle cost
https://doi.org/10.1007/978-981-15-6194-8_7
156 7 Damage Demand Assessment of Concrete Gravity …
Table 7.1 Mainshock-aftershock sequences recorded in the dam projects (Alliard 2006)
Name Country Year of Year Mainshock Largest Largest Type
impounding of magnitude aftershock aftershock
main magnitude delay
shock
Koyna India 1962 1967 6.3 5.1 10 months Concrete
gravity
Kremasta Greece 1965 1966 6.2 5.5 Several Rockfilled
months
Hsinfengkiang China 1959 1962 6.1 5.3 1 year 1/2 Buttress
Kariba Zimbabwe 1958 1963 6.0 5.8 2 Days Arch
Oroville USA, Ca. 1967 1975 5.7 5.2 1 day Earth-filled
Aswan Egypt 1964 1981 5.6 4.6 9 months Earth and
rock filled
Lake Mead USA, Co. 1935 1938 5.0 4.0 2h Arch
(Hoover dam)
Benmore New 1964 1966 5.0 4.4 1 year Earth-filled
Zealand
Marathon Greece 1929 1938 5.0 – – Concrete
gravity
Monteynard France 1962 1963 4.9 4.5 – Arch
Bhatsa India 1977 1983 4.9 3.9 4 months Masonry
gravity
Nurek Tadjikistan 1972 1972 4.6 4.3 Several Earth-filled
hours
(Liek Yeo and Allin Cornell 2009), risk assessment (Song et al., 2014), vulnerability
assessment (Li et al. 2014; Raghunandan et al. 2015), etc.).
Without being exhaustive, some studies on the aftershock effects on the seismic
performance of structures are worth mentioning. Salami et al. (2019) quantified the
influence of real mainshock-aftershock sequences on the seismic fragility of a 2-
strory RC structure. They found that considering aftershocks will increase the prob-
ability of exceedance of extensive and complete damage up to 10% for the crustal
ground motions and 5% for the inslab and interface records. Shokrabadi and Burton
(2018) quantified the impact of aftershock and mainshock-aftershock on the seismic
risk of three reinforced concrete moment frame structures. Their results showed that
both the increased post-mainshock seismic hazard as well as the reduction in the struc-
tural capacity have a significant influence on the seismic hazard. Shokrabadi et al.
(2018) evaluated the influence of mainshock-aftershock ground motion pairing on
the structural response and collapsed performance of five ductile reinforced concrete
frames with varying heights. Their results showed that the mainshock-aftershock
may overestimate or underestimate the seismic demand and risk relative to the use of
more-appropriate mainshock-aftershock record pairs. Hosseinpour and Abdelnaby
(2017) concluded that the earthquake direction, aftershock polarity, and vertical
component can have a significant influence on the nonlinear dynamic behavior of RC
building structures subjected to multiple earthquakes. Zhai et al. (2015) explored the
seismic response of RC containment building subjected to as-records mainshock-
aftershock seismic sequences, their results indicated that aftershocks have a signif-
icant influence on the dynamic response of RC buildings in terms of maximum
top response and accumulated damage. Ruiz-García et al. (2014) evaluated the
effects of mainshock-aftershock seismic sequences on the dynamic response of RC
framed-buildings located in soft-soil sites. Their results showed that the relationship
between the period of the building after the mainshock and the predominant period
of the aftershock has a significant influence on the seismic response of RC build-
ings to aftershock ground motions. Silwal and Ozbulut (2018) examined the after-
shock collapse performance of steel frame buildings with and without superelastic
viscous dampers subjected to mainshock-aftershock seismic sequences. Ruiz-García
et al. (2018) explored the effects of as-recorded mainshock-aftershock sequences
on the seismic response of three-dimensional steel moment-resisting buildings in
terms of lateral interstory drift demands. Song et al. (2016) proposed a framework
to examine the effects of aftershock ground motions on the seismic loss of steel
buildings. The uncertainty in earthquake ground motions, structural model, damage
and loss was considered by using the Monte Carlo Simulation with Latin Hyper-
cube Sampling. Amiri and Bojórquez (2019) quantified the residual displacement
ratios of structures as bi-linear SDOF systems under mainshock-aftershock seismic
sequences. The effects of ground motion characteristics of seismic sequences such
as site condition, magnitude, epicentral distance, and duration have been consid-
ered. Rinaldin et al. (2017) investigated the effects of repeated seismic sequences
on SDOF systems with hysteretic or damped dissipative behavior. They concluded
that the seismic design codes should consider the influence of mainshock-aftershock
seismic sequences on structural response. Yaghmaei-Sabegh and Ruiz-García (2016)
investigated the nonlinear dynamic response of SDOF systems subjected to as-
recorded Varzaghan-Ahar seismic sequences in terms of constant-ductility strength
reduction factor and inelastic displacement ratios, and thought that it is not reason-
able to repeat the mainshock as the aftershock. Zhai et al. (2014) investigated the
effects of mainshock-aftershock ground motions on the ductility demand, normal-
ized hysteretic energy and damage index of inelastic SDOF structures, and found
that the aftershock has a more significant influence on the normalized hysteretic
energy and damage index than on ductility demand. Furtado et al. (2018) assessed
the seismic vulnerability of undamaged and damaged structures with and without
infill masonry walls under mainshock-aftershock sequences. They found that after-
shocks will cause higher seismic vulnerability to the damaged RC structures than the
undamaged structures. Morfidis and Kostinakis (2017) explored the role of masonry
infills on the damage response of 3D RC buildings with different heights, structural
systems and distribution of masonry infills subjected to seismic sequences, and found
that seismic sequences has larger influence on the structural damage for the infilled
buildings than that for bare structures. Tesfamariam et al. (2015) studied the seismic
vulnerability of RC buildings with unreinforced masonry infill walls subjected to
mainshock-aftershock seismic sequences. Their results showed that the mainshock-
aftershock seismic sequences will increase the seismic demands when compared
with single mainshock ground motions. Omranian et al. (2018) performed a cloud
analysis method to study the seismic vulnerability of RC skew bridges subjected
to as-recorded mainshock-aftershock sequences. The effects of the skew angle of
the deck and direction of the seismic excitation on the fragility curves have been
discussed. They thought it was unconservative if only the mainshocks are consid-
ered. Pang and Wu (2018) investigated the seismic fragility of multispan RC bridges
subjected to mainshock-aftershock sequences, and found that aftershocks can be
harmful to both bridge components and system. Shin and Kim (2017) explored the
effects of frequency contents of aftershock ground motions on nonlinear dynamic
response of RC bridge columns, and demonstrated that the frequency contents of
aftershocks have great influence on the seismic performance of RC columns. Dong
and Frangopol (2015) performed a framework for risk and resilience assessment
of highway bridges subjected to mainshock and aftershock sequences. The effects
of the uncertainties associated with seismic scenarios and consequence evaluation
were incorporated within their framework. Their results showed that the strong
aftershocks have a significant influence on the repair loss and residual function-
ality of bridges. Fakharifar et al. (2015) evaluated the efficacy of three types of
repair jackets on mainshock-damaged RC bridge columns subjected to aftershock
attacks with various intensities using the incremental dynamic analyses method,
and found the three repair jackets can effectively improve the seismic performance
and collapse capacity of bridges by approximately 20% under severe aftershocks.
Konstandakopoulou and Hatzigeorgiou (2017) evaluated the seismic performance
of water and wastewater steel tanks subjected to as-recorded and artificial seismic
sequences. They thought the single earthquake records would lead to underestimate
the seismic demands of the steel tanks in terms of bearing capacity and deformation.
Singh et al. (2018) employed a unidirectional shake table to investigate the seismic
behavior of the damaged tunnel subjected to aftershock ground motions, and found
that the damaged tunnel is more vulnerable to the low frequency aftershocks. Sun
et al. (2020) evaluated the effects of mainshock-aftershock seismic sequences on
the nonlinear dynamic response of hydraulic arched tunnels using the lining local
damage index and lining global damage index. Their results showed that seismic
sequences have a negative influence on the seismic performance of the tunnel, and
will cause relatively severe cumulative damage in comparison with single mainshock
ground motions. Yu et al. (2019) concluded that strong aftershocks have a signifi-
cant influence on the seismic performance and floor acceleration response spectrum
of the AP1000 nuclear island building and will aggravate the potential cumulative
damage. Zhao et al. (2020) thought the aftershocks have only a minor impact on the
seismic behavior of isolated AP1000 nuclear island building because of the isolation
systems.
Concrete gravity dams located in earthquake-prone regions are not only exposed
to a single seismic event but also to a sequence of seismic shocks. Strong after-
shocks have the potential to cause additional cumulative damage to the mainshock-
damaged dams. Despite the fact that this problem has been qualitatively acknowl-
edged, nonlinear demands of concrete gravity dams (Oudni and Bouafia 2015; Wang
et al. 2015a; Alembagheri, 2016; Hariri-Ardebili et al. 2016a, b) are mainly based
on a single and rare design earthquake. To the best of our knowledge, only a few
researchers have evaluated the effect of mainshock–aftershock seismic sequences
on the nonlinear behavior of dams. Alliard and Léger (2008) studied the earthquake
safety evaluation of concrete gravity dams considering the aftershocks, they found
that strong aftershocks can cause additional damage and sliding displacements to
concrete gravity dams. Xia et al. (2010) employed a fully coupled method to analyze
the response of the earth embankment subjected to repeated earthquakes. Zhang et al.
(2013) examined the effects of as-recorded mainshock–aftershock seismic sequences
on the accumulated damage of concrete gravity dams. Their results indicated that
mainshock-aftershock seismic sequences have a significant influence on the accu-
mulated damage of concrete gravity dams. Pang et al. (2020) discussed the effects of
mainshock-aftershock seismic sequences on the probability of the high concrete face
rockfill dam based on the vertical deformation and damage index, and concluded that
strong aftershocks will significantly increase the fragility of the mainshock-damaged
concrete face rockfill dam.
It should be noted that most of the aforementioned studies were focused on the
effects of seismic sequences (as-recorded sequences, artificial sequences, repeated
sequences, etc.) on the seismic demands of reinforced concrete buildings. The limited
works (Alliard and Léger 2008; Xia et al. 2010; Zhang et al. 2013a) mentioned above
considered only the influence of the mainshock–aftershock seismic sequences on
concrete gravity dams and the effect of the damage due to mainshock on the damage
potential of aftershocks is not well understood. Besides, they did not provide statis-
tical information about the seismic demands of concrete gravity dams subjected to
single seismic events (both mainshocks and aftershocks) and mainshock–aftershock
seismic sequences. Furthermore, it is interesting to know whether the repeated earth-
quake approach can be used to estimate the seismic performance of concrete gravity
dams that are subjected to mainshock–aftershock seismic sequences.
The objective of this study is to understand the effects of the as-recorded main-
shock–aftershock seismic sequences on the damage demands of concrete gravity
dam–reservoir–foundation systems. To achieve this, 20 as-recorded mainshock–
aftershock seismic sequences are considered in this investigation. The objectives
of the current research are as follows: (a) to investigate the correlation between
the ground motion characteristics (i.e. frequency content, strong motion duration,
and amplitude) of the mainshocks and the major aftershocks, (b) to study the
differences in the nonlinear response of concrete gravity dam–reservoir–foundation
systems subjected to single mainshocks, single aftershocks, and mainshock–after-
shock seismic sequences, (c) to quantify the effects of aftershocks on the damage
demands of concrete gravity dams, and (d) to compare the damage demands in terms
of local and global damage indices between the as-recorded seismic sequences and
the repeated earthquakes.
7.2 Mainshock–Aftershock Seismic Sequences
7.2.1 Construction Method of Mainshock–Aftershock

Seismic Sequences
At present, the most widely used construction methods of mainshock–aftershock

seismic sequences for the structural seismic response analysis can be classified into
three categories: (a) as-recorded sequences (Ruiz-Garc I A and Negrete-Manriquez
2011; Efraimiadou et al. 2013); (b) repeated sequences (Hatzigeorgiou and Beskos
2009; Hatzigeorgiou and Liolios 2010a); (c) artificial sequences(Hatzigeorgiou and
Liolios 2010b; Moustafa and Takewaki 2011).
The as-recorded mainshock-aftershock seismic sequence is the first choice to
investigate the effects of the seismic sequence on nonlinear dynamic response of
structures. However, there are few records of as-recorded mainshock-aftershock
seismic sequences which are exactly in line with the actual research need. Due to the
large difference in the spectral characteristics between the mainshock and the after-
shock, it is uncertain whether the artificial seismic sequence can reflect the seismic
characteristics of the real seismic sequence. When an as-recorded single earthquake
is combined to form a mainshock-aftershock earthquake sequence (i.e. repeated
sequences), it is often necessary to determine the relationship of the magnitude and
peak acceleration between the mainshock and aftershock.
Hatzigeorgiou and Beskos (2009) employed the well-known Gutenberg-Richter
Law (Gutenberg and Richter 1965) to express the relationship of magnitudes between
mainshocks and aftershocks. The Gutenberg-Richter Law is often used to describe
the relationship between the magnitude and total number of earthquakes in any given
region and time period, which can be expressed as
N = 10b(M−M0 ) or log(N ) = b(M − M0 ) (7.1)
where N is the number of seismic events with a given magnitude range; M is the
magnitude of the mainshock; M 0 is the magnitude of the aftershock; b is the constant,
which is typically equal to 1.0. It can be seen from Eq. (7.1) that when the mainshock
magnitude of a given earthquake event is M, there will be 2 earthquakes with the
aftershock magnitude M 0 = M − 0.3010(log(2) ≈ 0.3010), and 3 earthquakes with
the aftershock magnitude M 0 = M − 0.4771(log(3) ≈ 0.4771). For example, for
the given mainshock with M = 7.0, there will be 2 earthquakes with the aftershock
magnitude M 0 = 7.0 − 0.3010 ≈ 6.7, and 3 earthquakes with the aftershock magni-
tude M 0 = 7.0 − 0.4771 ≈ 6.5. When the correlation of the magnitude between the
mainshock and aftershock is obtained, the relationship of the peak ground acceler-
ation (PGA) between the mainshock and aftershock can be calculated by using the
well-known Joyner–Boore (Joyner and Boore 1982) attenuation relation, which is
given by
7.2 Mainshock–Aftershock Seismic Sequences 161

log(PGA) = 0.49 + 0.23(M − 6) − log R 2 + 82 − 0.0027 R 2 + 82 (7.2)
Where PGA is expressed in gravitational acceleration, g; R is the source distance,

in km; M is the magnitude. This equation can be combined with Eq. (7.1)
to calculate the peak ground accelerations for 1, 2, and 3 earthquake events.
According to Eqs. (7.1) and (7.2), irrespectively of source distance and magnitude
of seismic events, the ratios between the peak ground accelerations (PGA) for the
above-mentioned cases can be expressed as
PGA(2−E V E N T S) PGA(M−0.3010)
=
PGA(1−E V E N T S) PGA(M)
√ √
R 2 +82 −0.0027 R 2 +82 .
100.49+0.23(M−0.3010−6)−log (7.3)
= √ √
100.49+0.23(M−6)−log R 2 +82 −0.0027 R 2 +82
= 0.8526
PGA(3−E V E N T S) PGA(M−0.4771)
=
PGA(1−E V E N T S) PGA(M)
√ √
R 2 +82 −0.0027 R 2 +82 .
100.49+0.23(M−0.4771−6)−log (7.4)
= √ √
100.49+0.23(M−6)−log R 2 +82 −0.0027 R 2 +82
= 0.7767
According to Eqs. (7.3) and (7.4), when the PGA of the mainshock is Ag,max ,
there will be 2 earthquakes with the aftershock PGA equal to 0.8526 · Ag,max , and 3
earthquakes with the aftershock PGA equal to 0.7767 · Ag,max .
7.2.2 As-Recorded Mainshock–Aftershock Seismic

Sequences
In order to investigate the nonlinear behavior of the mainshock-damaged concrete

gravity dams subjected to strong aftershocks, seismic sequences including the main-
shock and one major aftershock are needed for performing the nonlinear dynamic
analyses. In this study, as-recorded mainshock–aftershock seismic sequences are
selected according to the following criteria: (a) the magnitude of the mainshocks is
equal to or greater than 5.5 and that of the major aftershocks is equal to or greater than
4.0, (b) accelerograms are recorded at stations, which are located in free fields or low-
rise buildings, and (c) the predominant period (T p ) of the mainshocks is less than 0.4 s,
which corresponds to the typical dam sites. With these criteria, 20 as-recorded main-
shock–aftershock seismic sequences shown in Table 7.2, which were recorded at the
same station and in the same direction, are selected from the strong motion database
of the COSMOS (COSMOS 2019). The data sample includes well-known earth-
quakes with strong seismic activities, such as the Mammoth Lakes (1980), Kozani
Table 7.2 List of mainshock-aftershock seismic sequences considered in this investigation
162
No. Seismic Station Comp. Date Mag R (km) Recorded PGA Predominant Duration
sequences location (cm/s2 ) period (s) (5–95%) (s)
1 Mammoth 54099 90 1980-05-25(16:33:44) 6.1(ML ) 9.1 402.19 0.22 8.96
Lakes 1980-05-25(20:35:48) 5.7(ML ) 3.4 338.87 0.16 3.54
2 Mammoth 54099 180 1980-05-25(16:33:44) 6.1(ML ) 9.1 392.09 0.14 8.94
Lakes 1980-05-25(20:35:48) 5.7(ML ) 3.4 490.54 0.14 3.54
3 Chalfant 54428 270 1986-07-21(14:42:26) 6.4(ML ) 18.5 435.80 0.20 7.26
Valley 1986-07-21(14:51:09) 5.6(ML ) 27.9 156.56 0.12 7.66
4 Chalfant 54428 360 1986-07-21(14:42:26) 6.4(ML ) 18.5 394.60 0.24 8.16
Valley 1986-07-21(14:51:09) 5.6(ML ) 27.9 101.54 0.12 11.16
5 Chalfant 54171 180 1986-07-21(14:42:26) 6.4(ML ) 23.7 249.10 0.24 12.76
Valley 1986-07-31(07:22:40) 5.8(ML ) 15.5 179.66 0.34 14.16
6 Chalfant 54171 270 1986-07-21(14:42:26) 6.4(ML ) 23.7 169.00 0.36 17.04
Valley 1986-07-31(07:22:40) 5.8(ML ) 15.5 121.44 0.24 17.18
7 Coalinga CHP 0 1983-07-22(02:39:54) 6.0(Mw ) 13.4 350.10 0.14 9.01
1983-07-25(22:31:39) 5.3(Mw ) 12.7 521.20 0.22 2.77
8 Imperial 5055 315 1979-10-15(23:16:53) 6.5(Mw ) 8.8 209.40 0.22 13.22
Valley 1979-10-15(23:19:35) 5.0(ML ) 12.3 246.06 0.22 12.84
9 Kozani ITSAK 0 1995-05-13(08:47:00) 6.1(ML ) 19.5 211.03 0.20 5.84
1995-05-17(04:14:00) 5.3(ML ) 12.1 126.19 0.22 5.41
10 Kozani ITSAK 90 1995-05-13(08:47:00) 6.1(ML ) 19.5 136.00 0.20 7.72
1995-05-17(04:14:00) 5.3(ML ) 12.1 119.26 0.14 5.63
11 Whittier 24400 270 1987-10-01(14:42:20) 6.1(Mw ) 14.2 399.10 0.20 7.94
Narrows (continued)
7 Damage Demand Assessment of Concrete Gravity …
No. Seismic Station Comp. Date Mag R (km) Recorded PGA Predominant Duration
sequences location (cm/s2 ) period (s) (5–95%) (s)
1987-10-04(10:59:38) 5.3(Mw ) 16.3 333.30 0.14 5.18
12 Whittier 24400 360 1987-10-01(14:42:20) 6.1(Mw ) 14.2 420.10 0.18 7.14
Narrows 1987-10-04(10:59:38) 5.3(Mw ) 16.3 308.60 0.16 6.2
13 Whittier 24401 270 1987-10-01(14:42:20) 6.1(Mw ) 15.5 136.50 0.12 11.44
Narrows 1987-10-04(10:59:38) 5.3(Mw ) 15.1 137.20 0.30 3.74
14 Whittier 24401 360 1987-10-01(14:42:20) 6.1(Mw ) 15.5 183.80 0.12 7.34
Narrows 1987-10-04(10:59:38) 5.3(Mw ) 15.1 193.40 0.28 2.76
15 Whittier 24402 0 1987-10-01(14:42:20) 6.1(Mw ) 19.4 299.30 0.10 4.54
Narrows 1987-10-04(10:59:38) 5.3(Mw ) 18.2 266.00 0.14 2.14
7.2 Mainshock–Aftershock Seismic Sequences
16 Whittier 24402 90 1987-10-01(14:42:20) 6.1(Mw ) 19.4 157.90 0.22 8.1

Narrows 1987-10-04(10:59:38) 5.3(Mw ) 18.2 187.90 0.14 3.04
17 Cape 1585 270 1992-04-25(18:06:04) 7.0(Mw ) 26.5 317.60 0.16 10.65
Mendocino 1992-04-26(07:41:00) 6.6(Ms ) 33.3 344.20 0.30 6.76
/Petrolia
18 Cape 1585 360 1992-04-25(18:06:04) 7.0(Mw ) 26.5 471.00 0.24 10.42
Mendocino 1992-04-26(07:41:00) 6.6(Ms ) 33.3 431.40 0.34 6.83
/Petrolia
19 Cape 1583 360 1992-04-25(18:06:04) 7.0(Mw ) 32.7 280.10 0.4 15.06
Mendocino 1992-04-26(07:41:00) 6.6(Ms ) 47.1 271.00 0.38 9.75
/Petroli
20 Cape 1586 270 1992-04-25(18:06:04) 7.0(Mw ) 35.1 261.50 0.2 16.19
Mendocino 1992-04-26(07:41:00) 6.6(Ms ) 45.3 228.40 0.36 10.78
/Petroli
163
400 Mammoth Lakes seismic sequence 150

Kozani seismic sequence
Acceleration (cm/s2)
Station: 54099 Convict Creek-Comp.:90
100
200 Station: ITSAK-Comp.:90
50 Time gap (10 s)
Time gap (10 s)
0 0
1983/07/25
-50
1980/05/25
1980/05/25
1983/07/22
22:31:39
20:35:48
02:39:54
16:33:44
-200
-100
-400 -150
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 40
Time (s) Time (s)
(a) (b)
200 400 Cape Mendocino/Petrolia seismic sequence

Whittier Narrows seismic sequence
Station: 24402 -Comp.:90 Station: 1585-Comp.:270
100 200
Time gap (10 s) Time gap (10 s)
0 0
1987/10/01
1987/10/04
14:42:20
10:59:38
1992/04/26
-100
1992/04/25
-200
07:41:00
18:06:04
-200
-400
0 5 10 15 20 25 30 35 40 45 0 5 10 15 20 25 30 35 40 45 50 55
Time (s) Time (s)
(c) (d)
Fig. 7.1 Four representative as-recorded mainshock–aftershock seismic sequences: a Mammoth

Lakes, Station: 54099-Comp. 90, b Kozani, Station: ITSAK-Comp. 90, c Whittier Narrows, Station:
24402-Comp. 90, and d Cape Mendocino/Petrolia, Station: 1585-Comp. 270
(1995), Whittier Narrows (1987), and Cape Mendocino/Petrolia (1992) events. For
each of the 20 sequential ground motion records, the time gap between the main-
shock and the aftershock seismic events is set to be 10 s as illustrated in Fig. 7.1,
which shows the aforementioned four representative mainshock–aftershock seismic
sequences. This gap has zero acceleration ordinates and is adequate to allow the
structure to cease moving after the previous seismic event. It should be noted that
the selected aftershock records might be greater than the typical aftershock records
that are actually observed in the field. This is because the ground motion database,
such as COSMOS, lists only the stronger aftershock records. Therefore, the results
presented in this study should be interpreted cautiously.
The most important characteristics that describe an earthquake ground motion are
the maximum absolute amplitude, frequency content, and strong motion duration.
These characteristics of the seismic sequences used in this study are listed in Table 7.2.
The peak ground acceleration (PGA) is selected as the amplitude of the accelerogram.
The predominant period (T p ), which is a measure of the frequency content of the
ground motion, is defined as the period at which the maximum spectral acceleration
occurs in an acceleration response spectrum calculated at 5% damping. Various
definitions (Wang et al. 2015b, 2016) of strong motion duration have been proposed;
however, the use of the significant duration seems more appropriate and justified
because the ground motion records are required to be scaled for analysis, and the
significant duration is not affected by scaling (Zhang et al. 2013b; Wang et al. 2015c).
In this study, significant duration T 90 % (5–95%), which is defined as the time interval
between 5% and 95% of the Arias intensity, is selected to measure the duration of
ground motions.
7.2.3 Correlation Between Ground Motion Characteristics

of Mainshocks and Major Aftershocks
In order to investigate the similarity between the ground motion characteristics of the
selected aftershock accelerations and their corresponding mainshock time-histories,
the following parameters are calculated for each ground motion: (a) the predominant
period (T p ), (b) the strong motion duration (T 90 % ), and (c) the peak ground accel-
eration (PGA). For the purpose of illustration, the elastic spectra of the four typical
seismic sequences at a viscous damping ratio of ζ = 5% are presented in Fig. 7.2, and
the corresponding power amplitude spectra are shown in Fig. 7.3. The predominant
periods (T p ) of the mainshocks and major aftershocks are also indicated in Fig. 7.2.
It can be clearly observed that the predominant periods and power amplitudes of the
mainshocks and their corresponding aftershocks are significantly different, which
means that the mainshock and the major aftershock have different frequency contents.
Thus, it may not be appropriate to assume that the mainshock–aftershock seismic
sequences have identical frequency content (Zhai et al. 2013).
1.8 0.6
Spectral acceleration (g)
1.6
Mainshock 0.5 Mainshock
1.4
Aftershock Aftershock
1.2 0.4
1.0
0.3
0.8
0.6 0.2
0.4
0.1
0.2
0.0 0.0
0 2 4 0 2 4
Period (s) Period (s)
(a) (b)
1.0 1.4
Mainshock 1.2 Mainshock

0.8
Aftershock 1.0 Aftershock
0.6 0.8
0.4 0.6
0.4
0.2
0.2
0.0 0.0
0 2 4 0 2 4
Period (s) Period (s)
(c) (d)
Fig. 7.2 Elastic pseudo-acceleration response spectra of four typical seismic sequences (5%).
a Mammoth Lakes, Station: 54099-Comp. 90, b Kozani, Station: ITSAK-Comp. 90, c Whittier
Narrows, Station: 24402-Comp. 90, and d Cape Mendocino/Petrolia, Station: 2585-Comp. 270
0.18 0.12
Mainshock Mainshock
0.15 Aftershock Aftershock
Power amplitude
Power amplitude
0.12 0.08
0.09
0.06 0.04
0.03
0.00
0.00
0.1 1 10 0.1 1 10
Frequency (Hz) Frequency (Hz)
(a) (b)
0.12 0.28
Mainshock Mainshock
Aftershock 0.24 Aftershock
Power amplitude
Power amplitude
0.08 0.20
0.16
0.12
0.04
0.08
0.04
0.00
0.00
0.1 1 10 0.1 1 10
Frequency (Hz) Frequency (Hz)
(c) (d)
Fig. 7.3 Power amplitude spectra of four typical mainshock–aftershock sequences: a Mammoth
The correlation between the three major characteristics of the mainshocks and their
corresponding aftershocks is quantified in Fig. 7.4a (predominant period), Fig. 7.4b
(strong motion duration), and Fig. 7.4c (peak ground acceleration, PGA). The corre-
lation coefficient for the predominant period is 0.30, indicating that the predominant
periods of the mainshock ground motions and that of the aftershock ground motions
are weakly correlated from a statistical point of view. Based on this observation,
it can be concluded that the simulation approach of repeating the mainshock as an
aftershock may not be appropriate. A much higher correlation coefficient (0.74) is
obtained for the significant duration than the predominant period, meaning that the
strong motion duration of the mainshocks is mildly correlated with that of the corre-
sponding aftershocks. Furthermore, it is observed that, on an average, the strong
motion duration of the mainshocks is approximately 28.6% longer than that of the
corresponding aftershocks.
In addition, it can be clearly observed from Fig. 7.4c that the PGA of the main-
shocks follows an approximately linear trend with respect to that of the major after-
shocks (i.e. the PGA of the major aftershocks increases with an increase in the PGA
of the mainshocks). For the selected seismic sequences, it was observed that, on an
0.4 20
18
T90% of aftershock (s)

Tp of aftershock (s)
16
0.3 14
12
0.2 0.30 10
8
6 0.74
0.1 4
2
0.0 0
0.0 0.1 0.2 0.3 0.4 0 2 4 6 8 10 12 14 16 18 20
Tp of mainshock (s) T90% of mainshock (s)
(a) (b)
600
PGA of aftershock (cm/s 2)
500
400
300
200 0.64
100
0
0 100 200 300 400 500 600
PGA of mainshock (cm/s2)
(c)
Fig. 7.4 Relationship between characteristic parameters of mainshocks and their corresponding
aftershocks. a Predominant period, b strong motion duration, and c peak ground acceleration
average, the PGA of the mainshock ground motions is 13.2% larger than that of
their major aftershocks. It should be noted that the correlation phenomena quantified
in this study are specific to the selected 20 as-recorded seismic sequences. Further
investigations will be carried out by using abundant as-recorded seismic sequences.
7.3 Finite Element Model
7.3.1 Koyna Dam–Reservoir–Foundation System
In order to quantify the influence of the major aftershocks on the nonlinear dynamic
response of the mainshock-damaged concrete gravity dams, the Koyna concrete
gravity dam is selected as a representative numerical example. The geometry and
finite element (FE) discretization of the Koyna dam–reservoir–foundation system are
illustrated in Sect. 2.5 (Chap. 2). The material parameters for concrete, foundation
rock, and water are the same as Table 2.1.
7.3.2 Seismic Input
A set of 20 as-recorded seismic sequences with one major aftershock is consid-

ered as the seismic input, as summarized in Table 7.2. The mainshocks are scaled
to have a PGA of 0.40 g. The aftershock accelerograms are adjusted along with
their corresponding mainshock accelerogram. Thus, all the seismic sequences are
multiplied by appropriate factors. For example, the aforementioned four typical
sequential ground motions are multiplied by 0.976 (Mammoth Lakes, Station:
54099-Comp. 90), 2.885 (Kozani, Station: ITSAK-Comp. 90), 2.485 (Whittier
Narrows, Station: 24402-Comp. 90), and 1.236 (Cape Mendocino/Petrolia, Station:
2585-Comp. 270).
In order to investigate the influence of the aftershocks on the damage demands
of the mainshock-damaged dam, repeated earthquakes (i.e. back-to-back approach)
are also considered as the seismic input. The back-to-back approach assumes that
the features of the ground motion, such as the frequency content and strong motion
duration of the mainshocks and aftershocks, are the same. The amplitude (i.e. PGA)
of the mainshock is multiplied by a scaling factor (0.8526 (Hatzigeorgiou and Beskos
2009) or 1.0) to obtain the PGA of the aftershock.
7.4 Nonlinear Behavior of the Koyna

Dam–Reservoir–Foundation System
In this section, the nonlinear dynamic behavior of the concrete gravity dam–reser-
voir–foundation system subjected to the four representative seismic sequences is
discussed. This study focuses on the following design parameters: structural damage,
displacement response, and damage dissipated energy.
7.4.1 Structural Damage
The damage profiles of the Koyna dam–reservoir–foundation system subjected to

both single seismic events (mainshock or aftershock) and mainshock–aftershock
seismic sequences are shown in Fig. 7.5. The contour values shown in Fig. 7.5
are calculated using the CDP model, and they indicate the damage variable of the
concrete. In the CDP model, the damage variable of the concrete is determined
by using the corresponding tensile and compressive strains, and it ranges from 0
7.4 Nonlinear Behavior of the Koyna Dam–Reservoir–Foundation System 169
Mainshock Aftershock Mainshock–aftershock

1980-05-25 1980-05-25 seismic sequence
(16:33:44) (16:33:44)
(a)
Mainshock Aftershock Mainshock–aftershock

1995-05-13 1995-05-17 seismic sequence
(08:47:00) (04:14:00)
(b)
Aftershock
Mainshock Mainshock–aftershock
1987-10-01 1987-10-04
seismic sequence
(14:42:20) (10:59:38)
(c)
Mainshock Aftershock
Mainshock–aftershock
1992-04-25 1992-04-26
seismic sequence
(18:06:04) (07:41:00)
(d)
Fig. 7.5 Damage profiles of the Koyna dam under single seismic events and seismic sequences.
(intact material) to 1 (fully damaged material). This figure depicts the damage of
the dam–reservoir–foundation system due to the four representative ground motions
considered in this study. It is evident that the aftershocks influence the accumulated
damage of the post-mainshock concrete gravity dams significantly. In some cases,
the results corresponding to the mainshock–aftershock ground motions clearly show
a significant degradation in the strength of the dam, which is indicated by a crack
that extends completely across the upper section. In general, it is observed that the
seismic sequences do not lead to changes in the damage pattern of the dam during
the aftershocks. In other words, if the dam suffers damage during the mainshock, it
is likely that the damage accumulates sequentially during the aftershock.
It is also observed that a significant difference exists in the damage profile of the
upper part between the single mainshock and aftershock events. Such a difference
may be contributed to the distinctive ground motion characteristics (i.e. amplitude,
frequency content, and strong motion duration) between the mainshock and after-
shock motions. Not only a single mainshock event can cause serious damage to the
dam, but also a single aftershock event has the potential to cause extensive structural
damage. In some cases, a single aftershock event can cause more damage to the dam
than a single mainshock event.
It is concluded that seismic sequences can lead to a higher accumulated damage
to concrete gravity dams; hence, the seismic design of concrete gravity dams should
be given more attention.
7.4.2 Displacement Response
The time history of the horizontal displacements at the crest of the dam, which is
subjected to the four representative sequential ground motions, is shown in Fig. 7.6
with the positive displacement toward the downstream direction. The displacement
response of the dam subjected to single aftershock events is also illustrated in Fig. 7.6.
For comparison, the starting time of the single aftershock event is adjusted to match
the starting time of the aftershock in the seismic sequence.
It is observed from Fig. 7.6 that the dam is excited in a different manner for each
individual seismic event, and the nonlinear response that is obtained from the seismic
sequences has a substantially different displacement history than those obtained
from the single aftershocks. The maximum upstream horizontal displacements (indi-
cated by negative values in Fig. 7.6) corresponding to the seismic sequences are
greater than that corresponding to the single mainshocks. This is because that the
examined dam exhibits residual displacements, which are accumulated during earth-
quakes, and therefore, the maximum displacements appear to increase during seismic
sequences when compared to single mainshocks. This means that mainshock–after-
shock seismic sequences require increased displacement demands when compared
to single mainshocks.
Seismic sequences have a significant influence on the residual displacements
in addition to the maximum displacements. In some cases, single aftershocks
7.4 Nonlinear Behavior of the Koyna Dam–Reservoir–Foundation System 171
4
4 Seismic sequence Seismic sequence
Single aftershock Single aftershock
2
Displacement (cm)
Displacement (cm)
Static loading Static loading
2
0
0
-2
-2
Mainshock Gap (10 s) Aftershock -4 Mainshock Gap(10s) Aftershock
-4
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 40
Time (s) Time (s)
(a) (b)
6 4
Seismic sequence Seismic sequence
4 Single aftershock 2 Single aftershock
Displacement (cm)
Static loading
Displacement (cm)
Static loading
2
0
0
-2
-2
-4
-4
Mainshock Gap(10s) Aftershock Mainshock Gap (10 s) Aftershock
-6 -6
0 5 10 15 20 25 30 35 40 45 0 5 10 15 20 25 30 35 40 45 50 55
Time (s) Time (s)
(c) (d)
Fig. 7.6 Time histories of horizontal displacement at crest of dam for typical seismic sequences.
may induce residual displacements toward the downstream direction, unlike the
corresponding mainshocks, which induce displacements in the upstream direction.
However, in most of the cases, seismic sequences generally lead to a higher and more
intense response when compared to the worst single seismic event.
7.4.3 Damage Dissipated Energy
Another critical parameter is the damage dissipated energy (E d ), which can be

computed by numerical quadrature according to the following equation.
dt
1 el
Ed = ε : E: εel dddv (7.5)
2
v 0
where d t is the current damage level, E is the elastic tensor, εel is the elastic strain,
and v is the volume. This parameter is useful to evaluate the potential damage that
the structure has sustained. Figure 7.7 shows the damage-induced dissipated energy
curves of the Koyna dam subjected to sequential and single seismic events. The
solid lines represent the energy dissipated during the mainshock–aftershock seismic
sequences, whereas the dashed lines denote the energy dissipated during the single
aftershock ground motions. For comparison, the starting time of the single aftershock
Damage dissipated energy (KN.m)

30 30
Seismic sequence 27.8 KN.m
Seismic sequence
25 Single aftershock 25 Single aftershock
6.2 KN.m
20 20
12.9 KN.m
15 15
9.2 KN.m
10 10
5 5
0 0 Aftershock
Mainshock Gap (10 s) Aftershock Mainshock Gap (10 s)
-5 -5
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 40
Time (s) Time (s)
(a) (b)

80 35
69.1 KN.m Seismic sequence
70 Seismic sequence 30 Single aftershock
60 Single aftershock
25 12.26 KN.m 19.8 KN.m
50
37.7 KN.m 20
40
30 15
20 10
10 5
0 Gap (10 s) Aftershock 0 Mainshock Gap (10 s)
Mainshock Aftershock
-10 -5
0 5 10 15 20 25 30 35 40 45 0 5 10 15 20 25 30 35 40 45 50 55
Time (s) Time (s)
(c) (d)
Fig. 7.7 Damage dissipated energy curves of the dam for typical seismic sequences. a Mammoth
is adjusted to be the same as that of the aftershock in the seismic sequence. It is

evident that more energy is dissipated during the seismic sequences than during the
corresponding mainshocks. In some cases, the energy dissipation induced by the
single aftershock is larger than that induced by the mainshock–aftershock seismic
sequence. It is also interesting to observe that the energy dissipated during the single
aftershock ground motions is higher than the additional energy dissipated during the
seismic sequences. This means that the mainshock-damaged condition will strongly
influence the subsequent nonlinear dynamic behavior.
7.5 Estimation of Damage Demands

for Mainshock–Aftershock Seismic Sequences
As described in Sect. 7.4, the typical seismic sequences require the increased capacity
to withstand nonlinear demands (i.e. damage accumulation, residual displacement,
7.5 Estimation of Damage Demands for Mainshock–Aftershock Seismic Sequences 173
and damage dissipated energy) when compared to the corresponding single main-
shocks. This section will examine the estimation of damage demands for both the as-
recorded seismic sequences and repeated earthquakes. In order to quantify the struc-
tural damage demand of the concrete gravity dam under investigation, the damage
index (DI) and damage dissipated energy are computed.
7.5.1 Effects of as-Recorded Mainshock–Aftershock Seismic

Sequences
The damage demands of the concrete gravity dam subjected to mainshock–aftershock

seismic sequences can be characterized by the additional damage due to the after-
shock ground motions with respect to the mainshock ground motions. Figure 7.8
shows the local and global damage indices corresponding to the selected 20 as-
recorded mainshock–aftershock seismic sequences. The damage level for the single
mainshocks and single aftershocks is indicated in this figure. The mean values of
damage indices for the single and sequential seismic events are also indicated in
Fig. 7.8. It can be observed that in any as-recorded seismic sequences, strong after-
shocks cause increased damage both in the local and global levels. This implies that
strong aftershocks influence the accumulated damage significantly during multiple
seismic events. However, it should be emphasized that aftershock risks are not
always high for all situations. The ground motion database COSMOS lists only
the major aftershock records, and thus the selected mainshock–aftershock records
are demanding or more critical cases with respect to the other typical cases.
As shown in Fig. 7.8, the seismic sequences increase the damage indices (both the
local and global indices) when compared to the single mainshock ground motions.
On an average, the global index of the dam corresponding to the as-recorded seismic
sequences is 1.41 times of that corresponding to the single mainshocks. The local
damage indices of the upper part and heel of the dam subjected to the mainshock–
aftershock seismic sequences appear to increase by 45% and 25%, respectively, with
respect to that obtained for the corresponding single mainshocks. The mean values of
the additional damage indices of the upper part of the dam, heel of the dam, and entire
dam corresponding to the as-recorded mainshock–aftershock seismic sequences are
0.24, 0.04, and 0.09, respectively. This indicates that the additional damage caused by
the aftershocks is more pronounced for the upper part of the dam, and the accumulated
damage of the heel of the dam is not very sensitive to the seismic sequences.
From Fig. 7.8, it can be observed that in some cases, the damage index corre-
sponding to the single strong aftershocks may be larger than that corresponding to
the single mainshocks, which implies that a single aftershock event may cause more
damage to the dam than a single mainshock event. This is because the mainshock and
the corresponding aftershock have significantly different ground motion characteris-
tics. In some seismic sequences, although the magnitude of the mainshock is greater
Fig. 7.8 Damage indices of the dam for typical seismic sequences. a Local damage index for the
upper part of the dam, b local damage index for the dam heel, and c global damage index for the
whole dam
than that of the major aftershock, the intensity of the major aftershock (measured by
the PGA) is greater than that of the corresponding mainshock.
Figure 7.9 shows the ratios of the additional damage demands induced by the
aftershock of as-recorded mainshock–aftershock seismic sequences (DIas ) to those
of the undamaged dam subjected to the single aftershocks (DIas). The additional
damage index depends not only on the level of damage following the mainshock but
also on the ground motion characteristics of the aftershock. It is interesting to observe
that for all the ratios lower than 1.0, strong aftershocks will cause greater damage
1.0
Upper part; Mean=0.39
Dam heel; Mean=0.22
0.8 Global; Mean=0.31
ΔDIas/DIas
0.6
0.4
0.2
0.0
0 2 4 6 8 10 12 14 16 18 20
No.
Fig. 7.9 Ratios of additional damage indices of the mainshock-damaged dam subjected to
aftershocks (DIas ) to damage indices of the undamaged dam subjected to single aftershocks
(DIas )
to the undamaged dam than to the mainshock-damaged dam, as shown in Fig. 7.9.
Especially, the local damage index ratio and the global index ratio shown in Fig. 7.9
are lower than 0.5 for most seismic sequence ground motions. This may be because
the downstream and upstream surfaces of the upper part of the dam as well as the
heel of the dam are characterized by high tensile stress under strong ground motions.
The initial damage always appears in these high-stress areas and quickly extends to
the inside of the dam. After the crack propagates to a certain depth, the development
of the damage slows down even under the same loading condition. Hence, strong
aftershocks will cause a smaller additional damage to the mainshock-damaged dam
than when they are applied to the undamaged dam. It should also be noted that this
observation might be specific to the selected strong aftershock records, dam types,
and modeling assumptions adopted in the structural model as well as the damage
measures. When moderate or small aftershocks, which may not cause damage to the
undamaged dam, are selected as the seismic input, the relevant conclusion may not
be applicable.
The mean ratios of the local damage indices of the upper part and heel of the dam
are 0.39 and 0.22, respectively, whereas the mean ratio of the global damage index
is 0.31. This means that the structural damage induced by the mainshock–aftershock
seismic sequences is not a linear superposition of the single mainshock and single
aftershock. On an average, the additional damage indices of the mainshock-damaged
dam subjected to aftershocks decreased by approximately 60% with respect to the
undamaged dam subjected to the corresponding single aftershocks.
The ratios of the additional damage dissipated energy of the mainshock-damaged
dam subjected to aftershocks (DDEas ) to that of the undamaged dam subjected
to single aftershocks (DDEas ) are also computed for each as-recorded mainshock–
aftershock ground motion, as shown in Fig. 7.10. The mainshock–aftershock seismic
sequences strongly affect damage dissipated energy, in addition to the aforemen-
tioned damage indices. It is obvious that all the ratios in Fig. 7.10 are lower than
1.0
Mean=0.59
0.8
ΔDDEas/DDEas
0.6
0.4
0.2
0.0
0 2 4 6 8 10 12 14 16 18 20
No.
Fig. 7.10 Ratios of additional damage dissipated energy of the mainshock-damaged dam subjected
to aftershocks (DDEas ) to that of the undamaged dam subjected to single aftershocks (DDEas )
1.0, which also implies that strong aftershocks will cause greater damage to the
undamaged dam than to the mainshock-damaged dam.
7.5.2 Comparative Analysis of Damage Demands Between

as-Recorded Seismic Sequences and Repeated
Earthquakes
In order to identify the influence of ground motion characteristics of aftershocks

on the damage demands of the mainshock-damaged dam, the earthquake accelero-
grams of aftershocks are constructed by using different methods. The as-recorded
mainshock–aftershock seismic sequences presented in Sect. 7.2.1 are considered
as the reference Case 1. For comparison, the PGA of the aftershock of the as-
recorded seismic sequences is also scaled to have the same value of 0.34 g (with
a scaling factor of 0.8526 (Hatzigeorgiou and Beskos 2009)), namely, Case 2. Addi-
tionally, earthquake accelerograms of the mainshocks are simply repeated to form
the earthquake accelerograms of the aftershock. In this way, the mainshock–after-
shock seismic sequences are modeled by using back-to-back identical accelerograms,
namely, repeated earthquakes. Two different levels of PGA are considered for the
aftershock ground motion: 0.4 g (without scaling the mainshock to obtain the after-
shock) and 0.34 g for Case 3 and Case 4, respectively. Table 7.3 summarizes the cases
for each seismic sequence considered in this study. The PGA ratio for the seismic
sequences is defined as the ratio of the PGA of the aftershock (PGAas ) to the PGA
of the mainshock (PGAms ). The analyses described in Sects. 7.4 and 7.5.1 are with
respect to Case 1.
Figure 7.11 shows the additional damage index for the upper part of the dam,
heel of the dam, and entire dam that are subjected to both the as-recorded seismic
sequences and the repeated earthquakes. Table 7.4 summarizes the mean values of the
Table 7.3 List of the computed cases and information of mainshock-aftershock seismic sequences
considered in this investigation
Description Case Number of PGA Ratio
records Mainshock (g) Aftershock (g) (PGAas /PGAms )
As-recorded Case 1 20 0.40 Adjusting along The same with

seismic with the the non-modified
sequences corresponding records
mainshock
Case 2 20 0.40 0.34 0.8526
Repeated Case 3 20 0.40 0.40 1.0
seismic Case 4 20 0.40 0.34 0.8526
sequences
damage indices corresponding to the mainshocks and the additional damage indices
corresponding to the aftershocks for the upper part of the dam, heel of the dam, and
entire dam. The increasing damage index ratio is defined as the ratio of the additional
damage indices of the mainshock-damaged dam subjected to aftershocks (DIas ) to
the damage indices of the undamaged dam subjected to single mainshocks (DIms ).
It can be observed that the selection of the aftershock ground motions influences the
damage demands of the concrete gravity dam significantly. The damage demands
triggered by the mainshocks are larger than those triggered by the aftershocks are.
It can be observed from the results of Case 1 and Case 2 that the mean values
of the additional damage indices corresponding to the as-recorded aftershocks are
approximately the same, especially for the local damage index of the upper part
of the dam. This means that when the as-recorded mainshock–aftershock seismic
sequences are selected as the seismic input, the aftershocks can also be scaled to
have the same PGA based on an empirical PGA ratio, which is derived using the
ratio of the empirical attenuation functions (Hatzigeorgiou and Beskos 2009) (e.g.,
a PGA ratio of 0.8526 for the aftershocks).
It can be observed from the results of Case 3 and Case 4 that the repetition of
the identical earthquake ground motions (Case 3) triggers slightly more additional
damage demands of the dam at the end of the same earthquake ground motion (Case
4). It can be observed that the smaller the PGA of the aftershocks in the repeated
approach, the lower the additional damage indices.
Comparing the damage demands of the as-recorded seismic sequences (Case 1 and
Case 2) and repeated earthquakes (Case 3 and Case 4), the additional damage indices
corresponding to the aftershocks are more pronounced when the aftershocks are as
recorded with respect to the mainshocks. An important finding for the performance
assessment of the concrete gravity dam subjected to a mainshock–aftershock scenario
is that the damage demands computed from the repeated seismic sequences are, in
general, smaller than that computed from the as-recorded seismic sequences. This
means that the use of repeated earthquakes tends to underestimate the amplitude of
the damage demands, especially when the PGA of the aftershocks of the repeated
seismic sequences is scaled with a PGA ratio of 0.8526 (Case 4).
0.6 Case 1 Mean=0.240

Case 2 Mean=0.244
Additional damage index
0.5 Case 3 Mean=0.198

Case 4 Mean=0.111
0.4
0.3
0.2
0.1
0.0
0 2 4 6 8 10 12 14 16 18 20
No.
(a)
0.18
Case 1 Mean=0.040
0.15 Case 2 Mean=0.046
Case 3 Mean=0.035
0.12 Case 4 Mean=0.015
0.09
0.06
0.03
0.00
-0.03
0 2 4 6 8 10 12 14 16 18 20
No.
(b)
0.25
Case 1 Mean=0.088
Case 2 Mean=0.092
0.20 Case 3 Mean=0.074

Case 4 Mean=0.038
0.15
0.10
0.05
0.00
0 2 4 6 8 10 12 14 16 18 20
No.
(c)
Fig. 7.11 Additional damage index caused by different seismic sequence cases. a Upper part of
the dam, b dam heel, and c entire dam
It should be noted that the approach of using the mainshock as an aftershock,

has been found in (Li and Ellingwood 2007; Ruiz-García and Negrete-Manriquez
2011; Goda 2015) to overestimate the drift demands of the frame structures when
compared to the as-recorded mainshock–aftershock seismic sequences. From these
studies, one may naturally expect the same outcome for concrete dams and conclude
Table 7.4 Average values of damage indices under mainshocks and additional damage indices under aftershocks
Case Local damage index for the upper part of the Local damage index for the dam heel Global damage index for the dam
dam
DIms DIas DIas /DIms (%) DIms DIas DIas /DIms (%) DIms DIas DIas /DIms (%)
Case 1 0.539 0.240 44.5 0.198 0.040 20.2 0.280 0.088 31.4
Case 2 0.539 0.244 45.3 0.198 0.046 23.2 0.280 0.092 32.9
Case 3 0.539 0.198 36.7 0.198 0.035 17.7 0.280 0.074 26.4
Case 4 0.539 0.111 20.6 0.198 0.015 7.6 0.280 0.038 13.6
7.5 Estimation of Damage Demands for Mainshock–Aftershock Seismic Sequences
179
that it is conservative (or safe) to use repeated earthquakes of this kind. However,
the presented study, which is specific to the concrete gravity dams, suggests that
repeated earthquakes of this kind should be used with more caution. The underlying
reason for this observation may be attributed to the fact that concrete gravity dams
are massive concrete structures. The structural characteristics, seismic failure modes,
and damage measures of concrete gravity dams are quite different from those of the
frame structures. According to the results of the study, once a crack propagates to a
certain depth in the dam body (mainshock-damaged dam), strong ground motions,
which have the same frequency content as that of the mainshock, tend to slow down
subsequent damage development when compared to the as-recorded aftershocks. In
addition, it should be noted that the observation of this chapter might be specific to
the selected strong aftershock records, dam types, modeling assumptions as well the
damage measures.
Comparing Case 2 and Case 4, the additional damage indices in the upper
part of the dam, heel of the dam, and entire dam computed from the as-recorded
seismic sequences are larger by approximately 54.5%, 67.4%, and 58.7%, respec-
tively, than the corresponding additional damage indices calculated from the repeated
earthquakes.
7.6 Conclusions
This study presents the damage demand assessment for the concrete gravity dam–
reservoir–foundation systems subjected to as-recorded and artificial mainshock–
aftershock seismic sequences. The main innovation of this work is the quantification
of the effects of aftershocks directly into the damage demands of the mainshock-
damaged dam. For this purpose, a set of 20 as-recorded mainshock–aftershock earth-
quake ground motions is considered in this study, and four approaches are employed
to form the mainshock–aftershock seismic sequences. The following conclusions are
drawn from this investigation:
(1) Although the strong motion duration and PGA amplitude of the mainshocks
and the major aftershocks considered in this investigation are correlated in an
approximately linear fashion, the frequency content of the mainshocks, which
is measured by using the predominant period of the ground motion, is weakly
correlated with their corresponding frequency content of the aftershocks. Thus,
it can be concluded that the simulation approach of repeating the mainshock as
an aftershock is not appropriate.
(2) There exists a significant contrast in the damage profile between a single
mainshock event and a single aftershock event because of their different
ground motion characteristics. However, the damage induced by the mainshock
propagates sequentially during the aftershock.
(3) Mainshock–aftershock seismic sequences lead to an increase in the accumu-
lated damage of the concrete gravity dam. Strong aftershocks will cause greater
7.6 Conclusions 181
damage demands to the undamaged dam than to the mainshock-damaged dam.

The additional damage indices of the mainshock-damaged dam that is subjected
to aftershocks decrease by approximately 60% with respect to the undamaged
dam that is subjected to only the aftershocks.
(4) The aftershock ground motions selected by using different approaches influ-
ence the damage demands of the concrete gravity dam–reservoir–foundation
system significantly. The as-recorded aftershocks cause a greater additional
damage to the mainshock-damaged dam than the artificial aftershocks of the
repeated approach. Repeated earthquakes tend to underestimate the level of
damage demands, especially for a PGA ratio of 0.8526.
When the as-recorded mainshock–aftershock seismic sequences are selected as
the seismic input, the as-recorded aftershock accelerograms can be adjusted along
with their corresponding mainshocks or scaled to have the same PGA by using an
empirical PGA ratio.
References
Alembagheri, M. (2016). Earthquake damage estimation of concrete gravity dams using linear
analysis and empirical failure criteria. Soil Dynamics and Earthquake Engineering, 90, 327–339.
Alliard, P., & Léger, P. (2008). Earthquake safety evaluation of gravity dams considering aftershocks
and reduced drainage efficiency. Journal of Engineering Mechanics, 134(1), 12–22.
Alliard, P. M. (2006). Mainshocks and aftershocks sequences database. http://www.polymtl.ca/str
uctures/en/telecharg/index.php.
Amiri, S., & Bojórquez, E. (2019). Residual displacement ratios of structures under mainshock-
aftershock sequences. Soil Dynamics and Earthquake Engineering, 121, 179–193.
Chinese, S. (2001). Specifications for seismic design of hydraulic structures. Beijing: Chinese
Electric Power Press.
Chinese, S. 2011. Code for Design of Seismic of Hydropower Projects., Beijing: Chinese Electric
Power Press.
COSMOS. (2019). Strong Motion Virtual Data Center. http://strongmotioncenter.org/vdc/scripts/
default.plx.
Dong, Y., & Frangopol, D. M. (2015). Risk and resilience assessment of bridges under mainshock
and aftershocks incorporating uncertainties. Engineering Structures, 83, 198–208.
Efraimiadou, S., Hatzigeorgiou, G. D., & Beskos, D. E. (2013). Structural pounding between adja-
cent buildings subjected to strong ground motions. Part II: the effect of multiple earthquakes.
Earthquake Engineering & Structural Dynamics, 42(10), 1529–1545.
Fakharifar, M., Chen, G., Sneed, L., & Dalvand, A. (2015). Seismic performance of post-mainshock
FRP/steel repaired RC bridge columns subjected to aftershocks. Composites Part B Engineering,
72, 183–198.
Furtado, A., Rodrigues, H., Varum, H., & Arêde, A. (2018). Mainshock-aftershock damage
assessment of infilled RC structures. Engineering Structures, 175, 645–660.
Goda, K. (2012). Nonlinear response potential of mainshock–aftershock sequences from Japanese
earthquakes. Bulletin of the Seismological Society of America, 102(5), 2139–2156.
Goda, K. (2015). Record selection for aftershock incremental dynamic analysis. Earthquake
Goda, K., & Salami, M. R. (2014). Inelastic seismic demand estimation of wood-frame houses
subjected to mainshock-aftershock sequences. Bulletin of Earthquake Engineering, 12(2), 855–
874.
Goda, K., & Taylor, C. A. (2012). Effects of aftershocks on peak ductility demand due to strong
ground motion records from shallow crustal earthquakes. Earthquake Engineering and Structural
Dynamics, 41(15), 2311–2330.
Gutenberg, B., & Richter, C. F. (1965). Seismicity of the earth and associated phenomena. New
York: Hafner.
Han, R., Li, Y., & Lindt, J. (2015). Assessment of seismic performance of buildings with
incorporation of aftershocks. Journal of Performance of Constructed Facilities, 29(3), 1–17.
Hariri-Ardebili, M. A., Saouma, V. E., & Porter, K. A. (2016a). Quantification of seismic potential
failure modes in concrete dams. Earthquake Engineering and Structural Dynamics, 45(6), 979–
997.
Hariri-Ardebili, M. A., Seyed-Kolbadi, S. M., & Kianoush, M. R. (2016b). FEM-based parametric
Hatzigeorgiou, G. D. (2010a). Behavior factors for nonlinear structures subjected to multiple near-
fault earthquakes. Computers & Structures, 88(5–6), 309–321.
Hatzigeorgiou, G. D. (2010b). Ductility demand spectra for multiple near- and far-fault earthquakes.
Soil Dynamics and Earthquake Engineering, 30(4), 170–183.
Hatzigeorgiou, G. D., & Beskos, D. E. (2009). Inelastic displacement ratios for SDOF structures
subjected to repeated earthquakes. Engineering Structures, 31(11), 2744–2755.
Hatzigeorgiou, G. D., & Liolios, A. A. (2010). Nonlinear behaviour of RC frames under repeated
strong ground motions. Soil Dynamics and Earthquake Engineering, 30(10), 1010–1025.
Hatzivassiliou, M., & Hatzigeorgiou, G. D. (2015). Seismic sequence effects on three-dimensional
reinforced concrete buildings. Soil Dynamics and Earthquake Engineering, 72, 77–88.
Hosseinpour, F., & Abdelnaby, A. E. (2017). Effect of different aspects of multiple earthquakes
on the nonlinear behavior of RC structures. Soil Dynamics and Earthquake Engineering, 92,
706–725.
Jeon, J., DesRoches, R., & Lee, D. H. (2016). Post-repair effect of column jackets on aftershock
fragilities of damaged RC bridges subjected to successive earthquakes. Earthquake Engineering
Joyner, W. B., & Boore, D. M. (1982). Prediction of earthquake response spectra (pp. 34–45). US
Geological Survey.
Konstandakopoulou, F. D., & Hatzigeorgiou, G. D. (2017). Water and wastewater steel tanks under
multiple earthquakes. Soil Dynamics and Earthquake Engineering, 100, 445–453.
Li, Q., & Ellingwood, B. R. (2007). Performance evaluation and damage assessment of steel
frame buildings under main shock-aftershock earthquake sequences. Earthquake Engineering
Li, Y., Song, R., & Van de Lindt, J. W. (2014). Collapse fragility of steel structures subjected
to earthquake mainshock-aftershock sequences. Journal of Structural Engineering, 140(12),
4014095.
Liek Yeo, G., & Allin Cornell, C. (2009). Building life-cycle cost analysis due to mainshock and
aftershock occurrences. Structural Safety, 31(5), 396–408.
Liolios, A., Karabinis, A., Liolios, A., Radev, S., Georgiev, K., & Georgiev, I. (2015). A computa-
tional approach for the seismic damage response under multiple earthquakes excitations of adja-
cent RC structures strengthened by ties. Computers & Mathematics with Applications, 70(11),
2742–2751.
Morfidis, K., & Kostinakis, K. (2017). The role of masonry infills on the damage response of R/C
buildings subjected to seismic sequences. Engineering Structures, 131, 459–476.
Moustafa, A., & Takewaki, I. (2011). Response of nonlinear single-degree-of-freedom structures
to random acceleration sequences. Engineering Structures, 33(4), 1251–1258.
References 183
Omranian, E., Abdelnaby, A. E., & Abdollahzadeh, G. (2018). Seismic vulnerability assessment of
RC skew bridges subjected to mainshock-aftershock sequences. Soil Dynamics and Earthquake
Oudni, N., & Bouafia, Y. (2015). Response of concrete gravity dam by damage model under seismic
excitation. Engineering Failure Analysis, 58, 417–428.
Pang, R., Xu, B., Zhou, Y., Zhang, X., & Wang, X. (2020). Fragility analysis of high CFRDs
subjected to mainshock-aftershock sequences based on plastic failure. Engineering Structures,
206, 110152.
Pang, Y., & Wu, L. (2018). Seismic fragility analysis of multispan reinforced concrete bridges using
mainshock-aftershock sequences. Mathematical Problems in Engineering, 2018, 1–12.
Raghunandan, M., Liel, A. B., & Luco, N. (2015). Aftershock collapse vulnerability assessment of
reinforced concrete frame structures. Earthquake Engineering and Structural Dynamics, 44(3),
419–439.
Rinaldin, G., Amadio, C., & Fragiacomo, M. (2017). Effects of seismic sequences on structures
with hysteretic or damped dissipative behaviour. Soil Dynamics and Earthquake Engineering,
97, 205–215.
Ruiz-Garc, I. A. J., & Negrete-Manriquez, J. C. (2011). Evaluation of drift demands in existing
steel frames under as-recorded far-field and near-fault mainshock–aftershock seismic sequences.
Ruiz-García, J., Marín, M. V., & Terán-Gilmore, A. (2014). Effect of seismic sequences in reinforced
concrete frame buildings located in soft-soil sites. Soil Dynamics and Earthquake Engineering,
63, 56–68.
Ruiz-García, J., Yaghmaei-Sabegh, S., & Bojórquez, E. (2018). Three-dimensional response of steel
moment-resisting buildings under seismic sequences. Engineering Structures, 175, 399–414.
Salami, M. R., Kashani, M. M., & Goda, K. (2019). Influence of advanced structural modeling
technique, mainshock-aftershock sequences, and ground-motion types on seismic fragility of
low-rise RC structures. Soil Dynamics and Earthquake Engineering, 117, 263–279.
Shin, M., & Kim, B. (2017). Effects of frequency contents of aftershock ground motions on
reinforced concrete (RC) bridge columns. Soil Dynamics and Earthquake Engineering, 97, 48–59.
Shokrabadi, M., Burton, H. V., & Stewart, J. P. (2018). Impact of sequential ground motion pairing
on mainshock-aftershock structural response and collapse performance assessment. Journal of
Structural Engineering, 144(10), 4018177.
Shokrabadi, M., & Burton, H. V. (2018). Risk-based assessment of aftershock and mainshock-
aftershock seismic performance of reinforced concrete frames. Structural Safety, 73, 64–74.
Silwal, B., & Ozbulut, O. E. (2018). Aftershock fragility assessment of steel moment frames with
self-centering dampers. Engineering Structures, 168, 12–22.
Singh, D. K., Mandal, A., Karumanchi, S. R., Murmu, A., & Sivakumar, N. (2018). Seismic
behaviour of damaged tunnel during aftershock. Engineering Failure Analysis, 93, 44–54.
Song, R., Li, Y., & van de Lindt, J. W. (2014). Impact of earthquake ground motion characteristics
on collapse risk of post-mainshock buildings considering aftershocks. Engineering Structures,
81, 349–361.
Song, R., Li, Y., & Van de Lindt, J. W. (2016). Loss estimation of steel buildings to earthquake
mainshock-aftershock sequences. Structural Safety, 61, 1–11.
Sun, B., Zhang, S., Deng, M., & Wang, C. (2020). Nonlinear dynamic analysis and damage
evaluation of hydraulic arched tunnels under mainshock–aftershock ground motion sequences.
Tunnelling and Underground Space Technology, 98, 103321.
Tesfamariam, S., Goda, K., & Mondal, G. (2015). Seismic vulnerability of reinforced concrete frame
with unreinforced masonry infill due to main shock-aftershock earthquake sequences. Earthquake
Spectra, 31(3), 1427–1449.
Wang, G., Wang, Y., Lu, W., Yan, P., Zhou, W., & Chen, M. (2016). A general definition of integrated
strong motion duration and its effect on seismic demands of concrete gravity dams. Engineering
Wang, G., Zhang, S., Zhou, C., & Lu, W. (2015c). Correlation between strong motion durations
and damage measures of concrete gravity dams. Soil Dynamics and Earthquake Engineering, 69,
148–162.
Xia, Z., Ye, G., Wang, J., Ye, B., & Zhang, F. (2010). Fully coupled numerical analysis of repeated
shake-consolidation process of earth embankment on liquefiable foundation. Soil Dynamics and
Earthquake Engineering, 30(11), 1309–1318.
Yaghmaei-Sabegh, S., & Ruiz-García, J. (2016). Nonlinear response analysis of SDOF systems
subjected to doublet earthquake ground motions: A case study on 2012 Varzaghan-Ahar events.
Engineering Structures, 110, 281–292.
Yu, N., Zhao, C., Peng, T., Huang, H., Roy, S. S., & Mo, Y. L. (2019). Numerical investigation of
AP1000 NIB under mainshock-aftershock earthquakes. Progress in Nuclear Energy, 117, 103096.
Zafar, A., & Andrawes, B. (2015). Seismic behavior of SMA–FRP reinforced concrete frames under
sequential seismic hazard. Engineering Structures, 98, 163–173.
Zhai, C., Wen, W., Chen, Z., Li, S., & Xie, L. (2013). Damage spectra for the mainshock–aftershock
sequence-type ground motions. Soil Dynamics and Earthquake Engineering, 45, 1–12.
Zhai, C., Wen, W., Li, S., Chen, Z., Chang, Z., & Xie, L. (2014). The damage investigation of
inelastic SDOF structure under the mainshock-aftershock sequence-type ground motions. Soil
Dynamics and Earthquake Engineering, 59, 30–41.
Zhai, C., Zheng, Z., Li, S., & Pan, X. (2017). Damage accumulation of a base-isolated RCC building
under mainshock-aftershock seismic sequences. KSCE Journal of Civil Engineering, 21(1), 364–
377.
Zhai, C., Zheng, Z., Li, S., & Xie, L. (2015). Seismic analyses of a RCC building under mainshock–
aftershock seismic sequences. Soil Dynamics and Earthquake Engineering, 74, 46–55.
16–27.
Zhao, C., Yu, N., Peng, T., Lippolis, V., Corona, A., & Mo, Y. L. (2020). Study on the dynamic
behavior of isolated AP1000 NIB under mainshock-aftershock sequences. Progress in Nuclear
Energy, 119, 103144.
Chapter 8
Earthquake Direction Effects
on Nonlinear Dynamic Response
of Concrete Gravity Dams to Seismic
Sequences
8.1 Introduction
Historical earthquakes have shown that a large mainshock can usually trigger
numerous aftershocks within a short period of time. For example, the 2010 Chile
earthquake had a mainshock measuring 8.8 M w followed by 306 aftershocks with
the magnitude greater than 5.0, and 21 aftershocks with the magnitude greater than 6.0
(Ruiz-García and Negrete-Manriquez 2011). This means that man-made civil engi-
neering structures (e.g., concrete gravity dams) in the seismically active region are
not exposed to a single seismic event, but to a sequence of seismic shocks consisting
of foreshock, mainshock, and aftershock. However, the modern performance-based
assessment methodology for evaluation of concrete gravity dams is only based on a
single earthquake event in current seismic codes [e.g. in China (Chinese 2011)]. The
phenomenon of multiple earthquakes has not been considered adequately.
It has been recognized that strong aftershocks have the potential to cause addi-
tional cumulative damage to the mainshock-damaged structures. Despite the fact that
the influence of mainshock-aftershock seismic sequences on the nonlinear dynamic
response of structures has been qualitatively acknowledged (Fakharifar et al. 2015;
Jeon et al. 2016; Song et al. 2016; Konstandakopoulou and Hatzigeorgiou 2017;
Moshref et al. 2017; Rinaldin et al. 2017; Shin and Kim 2017; Zhai et al. 2017),
nonlinear dynamic response analysis and seismic performance evaluation of concrete
gravity dams are mainly based on a single earthquake event (Zhang et al. 2013b;
Hariri-Ardebili and Saouma 2015; Wang et al. 2015; Hariri-Ardebili et al. 2016;
Soysal et al. 2016; Bybordiani and Arıcı 2017; Yazdani and Alembagheri 2017). To
the best of our knowledge, only a few researchers have investigated the influence of
strong aftershocks on the seismic demands of dams (Alliard and Léger 2008; Xia
et al. 2010; Zhang et al. 2013a; Wang et al. 2017a). In addition, these studies all
revealed that strong aftershocks have a significant influence on the nonlinear seismic
demands of dams. Thus, ground accelerations of mainshock-aftershock sequences
represent a real situation that requires special treatment in the dam seismic design.
https://doi.org/10.1007/978-981-15-6194-8_8
186 8 Earthquake Direction Effects on Nonlinear Dynamic Response …
The time history analysis method has been widely used to evaluate the seismic
performance of dams. When as-recorded ground motion accelerograms are selected
as the seismic input for two-dimensional analyses of concrete gravity dams, one
horizontal seismic component is generally applied to the foundation base in the
upstream-downstream direction. The downstream-upstream direction has not been
considered. However, previous studies indicate that the orientation of the seismic
input with regard to the principal structural axes has a significant influence on the
structural response. Changing the orientation of the seismic input may lead to a
quite different structural response, which can radically alter the analysis results in
terms of the damage level and damage characteristics. Rigato and Medina (2007)
examined the influence of the earthquake direction on seismic demands for inelastic
single-story structures subjected to bi-directional ground motions. They concluded
that applying bi-directional ground motions only along the principal axes of struc-
tures would underestimate the inelastic peak deformation demand. Hosseinpour and
Abdelnaby (2017) investigated the performance of the irregular structures subjected
to seismic sequences in different directions. Their results indicated that changing the
earthquake direction will significantly affect the residual drifts in most cases. Hatzi-
vassiliou and Hatzigeorgiou (2015) examined the effect of seismic sequence direction
incident on the nonlinear response of building structures for different siting config-
urations. Ruiz-García and Aguilar (2015) studied the effect of aftershock direction
(aftershock polarity) on the aftershock collapse potential. Their observation showed
that the aftershock direction has a significant influence not only on the aftershock
collapse capacity but also in the aftershock capacity against demolition. Morfidis
and Kostinakis (2017) employed 12 different incident angles to account for the influ-
ence of earthquake direction on the dynamic response of RC buildings with masonry
infills to seismic sequences. They found that it is of great importance to account for
the direction of the seismic motion. Raghunandan et al. (2015) mentioned that when
the residual drift resulting from a mainshock is high, the direction of the aftershock
with respect to the mainshock is important. Kostinakis et al. (2015) investigated the
impact of the seismic excitation angle on the damage level of 3D multi-story RC
buildings. They proposed that the common practice of applying the seismic ground
motions along the structural axes can lead to significant underestimation of structural
damage. However, the aforementioned studies are aimed at the reinforced concrete
building structures. Despite the significance of the incidence directions of the ground
motion records, few studies have focused their attention on the nonlinear response of
concrete gravity dam-reservoir-foundation systems subjected to seismic sequences
in different incidence directions.
The seismic sequence direction may influence the cumulative damage processes
of the dams, which will lead to different failure mechanisms. The objective of this
chapter is to better understand differences in structural response and damage on
concrete gravity dams when subjected to mainshock-aftershock seismic sequences
with different incidence directions. To accomplish this purpose, two seismic incident
directions (i.e. the upstream-downstream and downstream-upstream directions) are
applied along horizontal axes of the two-dimensional model. Nonlinear dynamic
response of the concrete gravity dam-reservoir-foundation system subjected to as-

recorded mainshock-aftershock sequences is conducted. The effects of direction of
single earthquake events, direction of seismic sequences, aftershock polarity are
investigated. The influence of the intensity of ground motions on the seismic sequence
direction effects is also discussed.
8.2 Earthquake Incident Direction for Two-Dimensional

Seismic Performance Analysis
The length of the dam axis of concrete gravity dams is generally very long. In
order to adapt the deformation requirements of the foundation and minimize the
cracking caused by the shrinkage, concrete gravity dams are constructed as indi-
vidual monoliths that are separated by contraction joints. Owing to the presence
of contraction joints, most of the nonlinear dynamic response analyses of concrete
gravity dams are through two-dimensional (2D) model (Wang et al. 2017b). When
the time history analysis method is selected to assess the seismic performance of
concrete gravity dams based on the two-dimensional model, earthquake acceler-
ations (artificial or as-recorded) are applied to the foundation base as the input
load. However, due to the uncertainty of the location of the epicenter of the next
earthquake, different incidence directions of the selected strong ground motion
records should be considered in order to assess the maximum structural demands.
Hence, for two-dimensional seismic performance analysis of concrete gravity dam-
reservoir-foundation systems, two earthquake incidence directions should be consid-
ered. Figure 8.1 shows the configurations of two incident directions for the two-
dimensional model of concrete gravity dams. One is the upstream-downstream
direction. Another one is the downstream-upstream direction).
(a) (b)
Fig. 8.1 Two siting configurations of earthquake direction for the concrete gravity dam-reservoir-
foundation systems based on two-dimensional model: a the upstream-downstream direction, and
b the downstream-upstream direction
8.3.1 Description of Finite Element Model

of Dam-Foundation-Reservoir Systems
In order to evaluate the influence of the earthquake direction on the nonlinear dynamic
response of concrete gravity dams to mainshock-aftershock seismic sequences, the
Koyna concrete gravity dam is selected as a representative numerical example. The
finite element (FE) discretization of the Koyna dam–reservoir–foundation system is
illustrated in Sect. 2.5. The foundation rock is assumed to be massless to avoid wave
propagation effects when applying free field earthquake records at the foundation
base. The reservoir water is assumed to be linearly elastic, irrotational, and inviscid.
The energy dissipation of the dam-reservoir-foundation system is considered by the
Rayleigh damping method with 5% damping ratio. Applied loads include the self-
weight of the dam, hydrostatic, uplift, hydrodynamic, and earthquake forces. The
static solutions of the dam due to its gravity loads and hydrostatic loads are taken as
initial conditions in the dynamic analyses of the system.
For the initial time step, the nodal displacements on the left and right truncated
boundary of the dam–reservoir–foundation system are assumed as zero in the normal
direction. In addition, the foundation base is fully constrained. In the sequent dynamic
analysis, all the displacement constraints are released, and the selected earthquake
accelerations are applied to the foundation base as the input load. At the fluid–
solid interface, the displacement in the direction normal to the interface is assumed
continuously during the entire simulation.
8.3.2 Input Ground Motions
In order to obtain the nonlinear behavior of concrete gravity dams subjected to

seismic sequences by time history analysis method, four as-recorded mainshock-
aftershock earthquake records, namely, Mammoth Lakes (1980), Kozani (1995),
Whittier Narrows (1987), and Cape Mendocino/Petrolia (1992) events, are selected
from the database provided by the COSMOS (COSMOS 2019). The details of these
ground motion records are listed in Table 8.1. Each selected mainshock–aftershock
seismic sequence is recorded at the same station and in the same direction.
For the comparison purpose, mainshock accelerograms of aforementioned seismic
sequences are scaled to have the same PGA of 0.40 g. The aftershock accelero-
grams are adjusted along with their corresponding mainshock accelerograms. Thus,
aforementioned four typical sequential ground motions are multiplied by 0.976
(Mammoth Lakes seismic sequences), 2.885 (Kozani seismic sequences), 2.485
(Whittier Narrows seismic sequences), and 1.236 (Cape Mendocino/Petrolia seismic
sequences). The scaled mainshock-aftershock seismic records with different dura-
tions are illustrated in Fig. 8.2. Every sequential ground motion is constructed as a
Table 8.1 List of mainshock-aftershock seismic sequences considered in this investigation
No. Seismic sequences Station Comp. Date Mag R (km) Recorded PGA Predominant Duration
location (cm/s2 ) period (s) (5–95%) (s)
1 Mammoth Lakes 54099 90 1980-05-25 6.1(ML ) 9.1 402.19 0.22 8.96

(16:33:44)
1980-05-25 5.7(ML ) 3.4 338.87 0.16 3.54
(20:35:48)
2 Kozani ITSAK 90 1995-05-13 6.1(ML ) 19.5 136.00 0.20 7.72
(08:47:00)
1995-05-17 5.3(ML ) 12.1 119.26 0.14 5.63
(04:14:00)
3 Whittier Narrows 24402 90 1987-10-01 6.1(Mw ) 19.4 157.90 0.22 8.1
(14:42:20)
1987-10-04 5.3(Mw ) 18.2 187.90 0.14 3.04
(10:59:38)
4 Cape 1585 270 1992-04-25 7.0(Mw ) 26.5 317.60 0.16 10.65
Mendocino/Petrolia (18:06:04)
1992-04-26 6.6(Ms ) 33.3 344.20 0.30 6.76
(07:41:00)
189
400 Mammoth Lakes seismic sequence 400

Kozani seismic sequence
Acceleration (cm/s )
2
2
200 200
Time gap (10 s)
Time gap (10 s)
0 0
1983/07/25
1980/05/25
1980/05/25
1983/07/22
22:31:39
-200
20:35:48
02:39:54
16:33:44
-200
-400 -400
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 40
Time (s)
Time (s)
(a) (b)
Whittier Narrows seismic sequence Cape Mendocino/Petrolia seismic sequence
400 400
2
2
200 200
Time gap (10 s) Time gap (10 s)
0 0
1987/10/01
1987/10/04
14:42:20
-200 10:59:38
1992/04/26
-200
1992/04/25
07:41:00
18:06:04
-400 -400
0 5 10 15 20 25 30 35 40 45 0 5 10 15 20 25 30 35 40 45 50 55
Time (s) Time (s)
(c) (d)
Fig. 8.2 Four as-recorded mainshock–aftershock seismic sequences: a Mammoth Lakes, b Kozani,
c Whittier Narrows, and d Cape Mendocino/Petrolia
unique seismic record where between mainshock-aftershock seismic events a time

gap of ten seconds is applied. This gap has zero acceleration ordinates and is adequate
to allow the structure to cease moving after the previous seismic event.
8.4 Effects of Earthquake Direction
In order to see the seismic performance of concrete gravity dams subjected to ground
motions in different directions, different sequences are employed and applied in both
upstream-downstream and downstream-upstream directions. The effects of direction
of single earthquake events, direction of seismic sequences, and aftershock polarity
on the nonlinear dynamic response of concrete gravity dams to strong ground motions
are evaluated. It should be noted that applying aftershocks without their mainshock
to the dam is aimed to observe the influence of the aftershock on the accumulated
damage of the mainshock-damaged dams.
8.4.1 Direction Effects of Single Earthquake Events
Figure 8.3 shows the final damage profile of the Koyna gravity dam subjected to single
mainshock and aftershock events with different incidence directions. Figure 8.3
clearly shows the differences between the damage characteristics of the Koyna dam
8.4 Effects of Earthquake Direction 191
Mainshock Mainshock Aftershock Aftershock

Upstream- Downstream- Upstream- Downstream-
downstream upstream downstream upstream
direction direction direction direction
(a)

(b)

(c)
(d)
Fig. 8.3 Damage profiles of the Koyna dam under single seismic events in different incidence
directions. a Mammoth Lakes, b Kozani, c Whittier Narrows, and d Cape Mendocino/Petrolia
when it was subjected to the selected as-record records applied along the upstream-
downstream and downstream-upstream directions. Changing the earthquake direc-
tion significantly affects the damage propagation processes of the upper part of the
dam. The accumulated damage on the heel is not very sensitive to the earthquake
incidence direction. Although the cracking profiles in the upper part of the dam are
initiated at the change in the slope of the downstream face in most analyses, the crack
propagation direction and crack length in the subsequent analysis are significantly
different when subjected to different earthquake directions. In some stations, cracks
are predicted to initialize near the middle of the upstream face and extend into the
dam toward the downstream face when subjected to ground motions in the upstream-
downstream direction. However, this cracking path of the dam will be completely
changed when subjected to the same ground motions in the downstream-upstream
direction.
It is also observed from Fig. 8.3 that there is a significant difference in the damage
profile of the upper part of the dam under the single seismic events. This may be
because that the selected single ground motions have different ground motion char-
acteristics (i.e. frequency content, peak ground acceleration, and strong motion dura-
tion). In some cases, the single aftershock ground motions can also cause serious
damage to the dam with the formation of the penetrated crack. This means that the
strong aftershock ground motions have the potential to cause additional damage to
the dam and cannot be simply repeated by the mainshocks.
The time history of the horizontal displacements at the crest of the dam subjected
to the single mainshock ground motions with two incidence directions is shown in
Fig. 8.4. The horizontal displacement response of the dam subjected to single after-
shock events with two incidence directions is also illustrated in Fig. 8.5. The positive
displacement is assumed toward the downstream direction. The initial displacement
is 0.29 cm, which is caused by the static loads.
It can be observed from Figs. 8.4 and 8.5 that the earthquake direction has a
significant influence on the nonlinear displacement response of the dam. In most
cases, changing the earthquake direction will affect the direction of the maximum
horizontal displacements (indicated by the positive and negative values in Figs. 8.4
and 8.5). In some stations, the nonlinear displacement response in two directions
is not significantly different at the beginning, but the difference increases under
the subsequent seismic input. In some cases, the upstream-downstream incidence
direction may induce residual displacements toward the downstream direction, and
the downstream-upstream incidence direction may induce residual displacements
toward the upstream direction. The differences between the total residual displace-
ments are more than 2.35 cm for two earthquake directions (see Fig. 8.5c). However,
in some cases, the residual displacements are toward the same direction under the
two earthquake directions.
Upstream-downstream direction
3 3 Downstream-upstream direction
Displacement (cm)
Displacement (cm)
2 2
1 1
0 0
-1 -1
-2 -2
-3 Upstream-downstream direction -3
Downstream-upstream direction
-4 -4
0 3 6 9 12 15 0 5 10 15
Time(s) Time(s)
(a) (b)
4
4 Upstream-downstream direction Upstream-downstream direction
Downstream-upstream direction 3 Downstream-upstream direction
3
Displacement (cm)
Displacement (cm)
2 2
1 1
0 0
-1 -1
-2 -2
-3
-3
-4
-4
0 5 10 15 20 0 5 10 15 20 25
Time(s) Time(s)
(c) (d)
Fig. 8.4 Horizontal displacements of the dam crest under single mainshock events in different inci-
dence directions. a Mammoth Lakes, b Kozani, c Whittier Narrows, and d Cape Mendocino/Petrolia
4
Displacement (cm)
Downstream-upstream direction 3 Downstream-upstream direction

3
Displacement (cm)
2
2
1
1
0 0
-1 -1
-2 -2
-3 -3
-4 -4
0 3 6 9 0 5 10 15
Time(s) Time(s)
(a) (b)
5
4
Displacement (cm)
4 Downstream-upstream direction Downstream-upstream direction

3 3
Displacement (cm)
2 2
1 1
0 0
-1
-1
-2
-3 -2
-4 -3
-5 -4
0 5 10 15 0 5 10 15 20
Time(s) Time(s)
(c) (d)
Fig. 8.5 Horizontal displacements of the dam crest under single aftershock events in different inci-
dence directions. a Mammoth Lakes, b Kozani, c Whittier Narrows, and d Cape Mendocino/Petrolia
8.4.2 Direction Effects of Mainshock–Aftershock Seismic

Sequences
Figures 8.6 and 8.7 show the final damage profiles of the Koyna dam subjected to the
as-recorded mainshock-aftershock seismic sequences with two earthquake incidence
directions. As shown, it can be found that the earthquake direction has a significant
influence on the damage propagation path and failure mechanism of the upper part of
the dam to seismic sequence. The more serious damage the single aftershock events
with different directions cause to the dam, the greater impact on the final damage the
mainshock-aftershock seismic sequence direction has, such as the Mammoth Lakes
and Whittier Narrows seismic sequences. The same seismic sequence with different
incidence directions will cause more or less similar damage profile and damage level
to the dam.
By comparing Fig. 8.3 (single seismic events) with Figs. 8.6 and 8.7 (mainshock-
aftershock seismic events), it can be noted that the damage level of the upper part of
(a) (b) (c) (d)
Fig. 8.6 Damage profiles of the Koyna dam under the selected mainshock-aftershock seismic
sequences in the upstream-downstream direction. a Mammoth Lakes, b Kozani, c Whittier Narrows,
and d Cape Mendocino/Petrolia
(a) (b) (c) (d)
sequences in the downstream-upstream direction. a Mammoth Lakes, b Kozani, c Whittier Narrows,
and d Cape Mendocino/Petrolia
the dam subjected to the single mainshock ground motions is smaller in comparison to
that subjected to mainshock-aftershock seismic sequences. Strong aftershock ground
motions can cause significant additional damage to the mainshock-damaged dam.
Although there are different failure modes of the dam when subjected to single
mainshock and single aftershock, the seismic sequences do not lead to changes
in the damage pattern of the dam during the aftershocks in general. Results of this
study indicate that considering just a mainshock to evaluate the structural response in
different directions may cause an underestimation in the difference between structural
demands in different directions.
Figure 8.8 shows the horizontal displacement time history of the dam crest to the
mainshock-aftershock seismic sequences with two incidence directions. It is obvious
that the mainshock-aftershock seismic sequence direction has a significant influence
on the peak displacement demands. For example, the maximum positive and nega-
tive horizontal displacements of the crest dam under the Mammoth Lakes seismic
sequences in the upstream-downstream direction are 3.70 cm and −2.73 cm, respec-
tively. However, the maximum positive and negative displacements of the crest dams
to the Mammoth Lakes seismic sequences in the downstream-upstream directions
are 3.00 cm and −3.73 cm, respectively. Residual displacements are accumulated
during the mainshock-aftershock seismic sequences. However, due to the influence of
the mainshock-damaged, the residual displacements are always toward the upstream
direction under the two earthquake directions.
4 4
Displacement (cm)
Upstream-downstream direction Upstream-downstream direction

Displacement (cm)
Downstream-upstream direction Downstream-upstream direction

2 Static loading 2 Static loading
0 0
-2 -2
-4
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 40
Time (s) Time (s)
(a) (b)
6 4
Displacement (cm)
Displacement (cm)
Upstream-downstream direction Downstream-upstream direction

4 Downstream-upstream direction 2 Static loading
Static loading
2 0
0
-2
-2
-4
-4
-6 -6
0 5 10 15 20 25 30 35 40 45 0 5 10 15 20 25 30 35 40 45 50 55
Time (s) Time (s)
(c) (d)
Fig. 8.8 Time histories of horizontal displacement at the dam crest under mainshock-aftershock
seismic sequences with different incidence directions. a Mammoth Lakes, b Kozani, c Whittier
Narrows, and d Cape Mendocino/Petrolia
8.4.3 Effects of Aftershock Polarity
Previous studies have shown that the mainshock or the aftershock accelerograms in a
sequence could have the opposite as-recorded value. This means that the mainshock-
aftershock could have different polarity (Raghunandan et al. 2012; Ruiz-García and
Aguilar 2015). To investigate the effects of aftershock polarity on the nonlinear
dynamic response in this study, the aftershock ground motions followed by the
mainshocks are assumed to have the same incidence direction (i.e. positive polarity)
and the opposite incidence direction (i.e. negative polarity) with the corresponding
mainshock ground motions.
Figure 8.9 shows the damage profiles of the dam under the mainshock-aftershock
seismic sequences with considering the negative polarity of aftershocks. These results
are compared with the cases when the mainshock and aftershock are in the same
direction (Fig. 8.6). Findings indicate that when the single aftershock ground motions
cause small damage to the undamaged dam, the aftershock polarity has a slight effect
on the damage propagation process and damage demands of the dam. Unlike the
direction effects of mainshock-aftershock seismic sequences, the aftershock polarity
will not significantly change the damage pattern of the dam in general. This is because
that the dam suffers the mainshocks with the same direction, and it is likely that the
mainshock-damaged accumulates sequentially during the aftershock with different
polarity.
Figure 8.10 shows the horizontal displacement response of the crest dam under
the selected mainshock-aftershock seismic sequences with different polarity for the
aftershocks. In general, there is no significant difference between the displacement
responses of the dam. However, maximum displacement demands are affected by
aftershock polarity. In some stations (Fig. 8.10a), the aftershock earthquake ground
motion with the negative polarity tends to move the dam to the opposite direction of
the original post-mainshock residual displacement. This means that the aftershock
polarity may lead to a recentering behavior.
(a) (b) (c) (d)
sequences with considering the negative polarity of aftershocks. a Mammoth Lakes, b Kozani,
8.5 Influence of the Ground Motion Intensity on the Seismic … 197
4
4
Displacement (cm)
Displacement (cm)
Static loading 2 Static loading

2
0
0
-2
-2
-4
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 40
Time (s) Time (s)

(a) (b)
6 4
Displacement (cm)
4 Downstream-upstream direction
Displacement (cm)
Static loading
2 Static loading
2
0
0
-2
-2
-4 -4
-6 -6
0 5 10 15 20 25 30 35 40 45 0 5 10 15 20 25 30 35 40 45 50 55
Time (s) Time (s)
(c) (d)
Fig. 8.10 Time histories of horizontal displacement at the dam crest with considering the aftershock
polarity effects. a Mammoth Lakes, b Kozani, c Whittier Narrows, and d Cape Mendocino/Petrolia
8.5 Influence of the Ground Motion Intensity

on the Seismic Sequence Direction Effects
In order to investigate the influence of the intensity of ground motions on the damage
characteristics of the dam under seismic sequences in different incidence directions,
as-recorded mainshock-aftershock seismic sequences are scaled to different inten-
sity levels to produce response ranging from small to large nonlinear response of the
dam. For this purpose, the nonlinear dynamic response analyses are carried out by
scaling each real mainshock-aftershock record, to progressively increase the intensity
of ground motions by increments of about 0.1 g. Hence, the mainshock accelero-
grams of the aforementioned seismic sequence are scaled to have the PGA of 0.3 g,
0.4 g, and 0.5 g, respectively. The aftershock accelerograms are adjusted along with
their corresponding mainshock accelerograms. It should be noted that the nonlinear
dynamic responses of the dam under seismic sequences with the mainshock PGA of
0.4 g are shown in Sect. 8.4.2.
Figure 8.11 and Fig. 8.12 show the damage profiles of the Koyna dam subjected
to seismic sequences in two directions with the mainshock PGAs of 0.3 g and 0.5 g,
respectively. Comparing Figs. 8.6 and 8.7 with Figs. 8.11 and 8.12, it can be seen that
the intensity of ground motions has a significant influence on the seismic sequence
direction effects. When the mainshock PGA of seismic sequences is not high (i.e.
0.3 g), there is a slight difference between the damage demands and damage profiles
for the dam subjected to seismic sequences in two incidence directions. In most cases,
cracks in the upper part of the dam are initiated at the downstream slope change
discontinuity and extend into the dam toward the upstream face when subjected to
Downstream- Upstream- Downstream-

Upstream-
upstream downstream upstream
downstream
direction direction direction
direction
(a) (b)
(c) (d)
Fig. 8.11 Damage profiles of the Koyna dam subjected to mainshock-aftershock seismic sequences
in different incidence directions with the mainshock PGA of 0.3 g. a Mammoth Lakes, b Kozani,

(a) (b)

(c) (d)
Fig. 8.12 Damage profiles of the Koyna dam subjected to mainshock-aftershock seismic sequences
in different incidence directions with the mainshock PGA of 0.5 g. a Mammoth Lakes, b Kozani,
two earthquake directions with the mainshock PGA of 0.3 g. However, as the PGA of
the seismic sequences increases, the difference between the damage characteristics
increases, and the structural demands also increase. This indicates that the damage
demand and failure mechanism of the concrete gravity dam are more sensitive to the
earthquake directions when subjected to seismic sequences with a higher PGA.
In order to quantify the structural damage demands of the concrete gravity dams
subjected to different incidence directions, the local damage index (DI), which is
defined as the ratio of the length of a cracking path to the total cross-sectional length
along the path, is computed. Figure 8.13 shows the local damage index of the upper
part of the dam subjected to single mainshock events and mainshock-aftershock
seismic sequences in different incidence directions. As shown, local damage indices
of the examined dam during the selected as-recorded mainshock-aftershock accelero-
grams with large PGA are generally greater than those under small events. It can also
be found that earthquake direction causes different damage demands to the dam. The
difference between structural damage demands in different directions increases with
the increase of the PGA. However, it should be noted that when strong ground motions
(i.e. PGA=0.5 g) cause very serious damage to the dam, such as the penetrated crack,
Mammoth Lakes Kozani Whittier Narrows Cape Mendocino/Petrolia
1.0
Damage index
0.5
0.0
0.3g 0.4g 0.5g 0.3g 0.4g 0.5g 0.3g 0.4g 0.5g 0.3g 0.4g 0.5g
(a)
1.0
Damage index
0.5
0.0
0.3g 0.4g 0.5g 0.3g 0.4g 0.5g 0.3g 0.4g 0.5g 0.3g 0.4g 0.5g
(b)
Fig. 8.13 Local damage index for the upper part of the dam subjected to single mainshock events
and mainshock-aftershock seismic sequences in different incidence directions. a Single mainshock
ground motions; b mainshock-aftershock seismic sequences
0.5
Additional damage indices
0.4
0.3
0.2
0.1
0.0
0.3g 0.4g 0.5g 0.3g 0.4g 0.5g 0.3g 0.4g 0.5g 0.3g 0.4g 0.5g
Fig. 8.14 Additional local damage index for the upper part of the dam caused by mainshock-
aftershocks seismic sequences
the local damage index cannot be used to evaluate the earthquake direction effects
on the structural damage demands for the dam.
Figure 8.14 shows the additional local damage index for the upper part of the dam
caused by the aftershock ground motions. It is obvious that mainshock-aftershock
seismic sequences cause an increment of the local damage index in comparison with
single mainshock events. However, the influence degree of the mainshock-aftershock
seismic sequences on the structural damage demands depends on the intensity of
the selected ground motions. Weak aftershock ground motions have less effect on
structural damage demands of the dam than strong aftershocks, which will lead
to little additional damage to concrete gravity dams. In general, strong aftershock
ground motions have a significant influence on the seismic performance of the dam.
However, if strong mainshock ground motions have led to serious damage to the
dam, the strong aftershocks may also cause little accumulated damage to the dam.
Figures 8.15 and 8.16 show the time history of the horizontal displacements at
the crest dam to different PGAs. It is obvious that seismic sequences, earthquake
directions, and PGA strongly affect the nonlinear dynamic response behavior of
the dam. As it can be seen in Fig. 8.15, the nonlinear dynamic responses in two
directions are not significantly different when subjected to seismic sequences with
the PGA of 0.3 g. However, the difference, especially for the residual displacements,
significantly increases under the high PGA of 0.5 g (Fig. 8.16). It can also be found
that the residual displacements are accumulated during the seismic sequences.
Figure 8.17 shows the maximum positive and negative horizontal displacements
of the dam crest subjected to mainshock-aftershock seismic sequences with different
PGAs. Figure 8.18 illustrates the residual displacements for sequence ground motions
with different PGAs. It can be seen from Figs. 8.17 and 8.18 that the earthquake
direction has a significant influence on the maximum horizontal displacement and
residual displacements. The calculated maximum horizontal displacements and
residual displacements are all becoming more sensitive to the earthquake direction
with the increase of the PGA. For example, maximum horizontal displacements of
the dam subjected to the Kozani seismic sequence in the upstream-downstream and
downstream-upstream directions with the mainshock PGA of 0.5 g are −6.12 cm and
4 4
Displacement (cm)
Displacement (cm)
Static loading Static loading
2 2
0 0
-2 -2
Mainshock Gap (10 s) Aftershock Mainshock Gap(10s) Aftershock
-4 -4
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 40
Time (s) Time (s)
(a) (b)
4
Displacement (cm)
Displacement (cm)
2 Static loading 2 Static loading
0
0
-2
-4 -2
-6 Mainshock Gap(10s) Aftershock Mainshock Gap (10 s) Aftershock

-4
0 5 10 15 20 25 30 35 40 45 0 5 10 15 20 25 30 35 40 45 50 55
Time (s) Time (s)
(c) (d)
Fig. 8.15 Horizontal displacement of the crest dam subjected to mainshock-aftershock seismic
sequences in different incidence directions with the mainshock PGA of 0.3 g. a Mammoth Lakes,
b Kozani, c Whittier Narrows, and d Cape Mendocino/Petrolia
4 6
4
Displacement (cm)
Static loading
Displacement (cm)
2 Static loading
2
0 0
-2 -2
-4
-4
Mainshock Gap (10 s) Aftershock Mainshock Gap(10s) Aftershock
-6
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 40
Time (s) Time (s)
(a) (b)
4 Upstream-downstream direction 4 Upstream-downstream direction
Displacement (cm)
2
Displacement (cm)
Static loading
2 Static loading
0
0
-2
-2
-4
-4
-6
-6 Mainshock Gap(10s) Aftershock Mainshock Gap (10 s) Aftershock
-8
0 5 10 15 20 25 30 35 40 45 0 5 10 15 20 25 30 35 40 45 50 55
Time (s) Time (s)
(c) (d)
Fig. 8.16 Horizontal displacement of the crest dam subjected to mainshock-aftershock seismic
sequences in different incidence directions with the mainshock PGA of 0.5 g. a Mammoth Lakes,
b Kozani, c Whittier Narrows, and d Cape Mendocino/Petrolia
6
0.5g 0.3g 0.4g 0.5g 0.3g 0.4g 0.5g
Maximum displacement (cm)
0.3g 0.4g 0.5g 0.3g 0.4g
4
2
0
A
-2
-4
Kozani
-6 Mammoth Lakes
Upstream-downstream direction Whittier Narrows
Downstream-upstream direction Cape Mendocino/Petrolia
-8
Fig. 8.17 Maximum horizontal displacement of the crest dam under different intensity levels of
ground motions
0.5
0.3g 0.4g 0.5g 0.3g 0.4g 0.5g 0.3g 0.4g 0.5g 0.3g 0.4g 0.5g
Residual displacement (cm)
0.0
A
-0.5
-1.0
-1.5 Mammoth Lakes
Kozani
-2.0
-2.5 Whittier Narrows
Upstream-downstream direction Cape Mendocino/Petrolia

-3.0
-3.5
Fig. 8.18 Residual displacement of the crest dam under different intensity levels of ground motions
4.26 cm, respectively. And the residual displacements are −3.12 cm and −0.34 cm,
respectively. It can also be found that the ground motions with greater PGA would
cause greater residual displacement to the dam than smaller events.
Changing in the damage, maximum displacement, and residual displacement
demands with the earthquake direction shows different performances in different
directions. This might be mainly due to the distinctive variations and degradation
of the structural properties (i.e. strength, stiffness, and damping) in different direc-
tions. Therefore, it is necessary to include the effects of earthquake direction on the
nonlinear dynamic behavior when evaluating the seismic performance of concrete
gravity dams subjected to strong mainshock-aftershock seismic sequences.
8.6 Conclusions
The present study investigated the nonlinear dynamic behavior of the two-
dimensional concrete gravity dam–reservoir–foundation systems subjected to single
seismic events and mainshock-aftershock seismic sequences in different directions.
The main innovation of this work is to evaluate the earthquake direction effects on
8.6 Conclusions 203
the structural demands of the dam based on the two-dimensional model. For this
purpose, two seismic incident directions (i.e. as-recorded acceleration time history
and opposite direction of the as-recorded acceleration time history) are applied along
horizontal axes of the two-dimensional model. Nonlinear dynamic response history
analyses are conducted to investigate the effects of direction of single earthquake
events, direction of seismic sequences, and aftershock polarity. The influence of the
intensity of ground motions on the direction effects of seismic sequences is also
discussed. Based on the results, the following conclusions may be drawn:
1. Changing the earthquake direction of single seismic events or mainshock-
aftershock seismic sequences can significantly affect the damage, maximum
displacement and residual displacement demands of the dam. The more different
damage the single seismic events with different directions cause to the dam, the
greater impact on the final damage the direction of mainshock-aftershock seismic
sequences has. The application of earthquake records along the horizontal axes
in the upstream-downstream direction may lead to significant underestimation
of seismic damage.
2. In general, the aftershock polarity will not significantly change the damage
pattern of the dam. In some cases, the aftershock earthquake ground motion
with the negative polarity tends to move the dam to the opposite direction of the
original post-mainshock residual displacement, which means that the aftershock
polarity may lead to a recentering behavior.
3. The intensity of ground motions has a significant influence on the seismic
sequence direction effects. When the mainshock PGA of seismic sequences is
not high, there is a slight difference between the damage demands and damage
profiles for the dam subjected to two incidence directions. However, as the PGA
of the seismic sequences increases, the damage propagation processes, damage
demands, maximum displacement, and residual displacements are becoming
more sensitive to the earthquake direction.
4. The influence degree of the mainshock-aftershock seismic sequences on the addi-
tional damage demands depends on the intensity of the selected ground motions.
Weak aftershock ground motions have less effect on structural damage demands
of the dam than strong aftershocks. In general, strong aftershock ground motions
will significantly increase the accumulated damage of the dam. However, if strong
mainshock ground motions have led to serious damage to the dam, the strong
aftershocks may also cause little accumulated damage to the dam.
The aforementioned conclusions are obtained from the nonlinear dynamic
response of the Koyna gravity dam subjected to four specific mainshock-aftershock
seismic sequences. Thus, these conclusions may not apply to all gravity dams and
ground motions. To better assess the earthquake direction effects, further study using
seismic fragility analysis, abundant strong motion records, and more incident angles
will be undertaken in the future.
References
Alliard, P., & Léger, P. (2008). Earthquake safety evaluation of gravity dams considering aftershocks
and reduced drainage efficiency. Journal of Engineering Mechanics, 134(1), 12–22.
Bybordiani, M., & Arıcı, Y. (2017). The use of 3D modeling for the prediction of the seismic
demands on the gravity dams. Earthquake Engineering & Structural Dynamics.
Chinese, S. (2011). Code for design of seismic of hydropower projects. Beijing: Chinese Electric
Power Press.
COSMOS. (2019). Strong Motion Virtual Data Center, http://strongmotioncenter.org/vdc/scripts/
default.plx.
Fakharifar, M., Chen, G., Sneed, L., & Dalvand, A. (2015). Seismic performance of post-mainshock
FRP/steel repaired RC bridge columns subjected to aftershocks. Composites Part B Engineering,
72, 183–198.
Hariri-Ardebili, M. A., & Saouma, V. (2015). Quantitative failure metric for gravity dams.
Hatzivassiliou, M., & Hatzigeorgiou, G. D. (2015). Seismic sequence effects on three-dimensional
reinforced concrete buildings. Soil Dynamics and Earthquake Engineering, 72, 77–88.
Hosseinpour, F., & Abdelnaby, A. E. (2017). Effect of different aspects of multiple earthquakes
on the nonlinear behavior of RC structures. Soil Dynamics and Earthquake Engineering, 92,
706–725.
Jeon, J., DesRoches, R., & Lee, D. H. (2016). Post-repair effect of column jackets on aftershock
fragilities of damaged RC bridges subjected to successive earthquakes. Earthquake Engineering
Konstandakopoulou, F. D., & Hatzigeorgiou, G. D. (2017). Water and wastewater steel tanks under
multiple earthquakes. Soil Dynamics and Earthquake Engineering, 100, 445–453.
Kostinakis, K., Morfidis, K., & Xenidis, H. (2015). Damage response of multistorey r/c buildings
with different structural systems subjected to seismic motion of arbitrary orientation. Earthquake
Morfidis, K., & Kostinakis, K. (2017). The role of masonry infills on the damage response of R/C
buildings subjected to seismic sequences. Engineering Structures, 131, 459–476.
Moshref, A., Khanmohammadi, M., & Tehranizadeh, M. (2017). Assessment of the seismic capacity
of mainshock-damaged reinforced concrete columns. Bulletin of Earthquake Engineering, 15(1),
291–311.
Raghunandan, M., Liel, A. B., & Luco, N. (2015). Aftershock collapse vulnerability assessment of
reinforced concrete frame structures. Earthquake Engineering and Structural Dynamics, 44(3),
419–439.
Raghunandan, M., Liel, A. B., Ryu, H., Luco, N., & Uma, S. R. (2012). Aftershock fragility curves
and tagging assessments for a mainshock-damaged building. In Proceedings of the 15th World
Conference on Earthquake Engineering, Lisbon.
Rigato, A. B., & Medina, R. A. (2007). Influence of angle of incidence on seismic demands
for inelastic single-storey structures subjected to bi-directional ground motions. Engineering
Structures, 29(10), 2593–2601.
Rinaldin, G., Amadio, C., & Fragiacomo, M. (2017). Effects of seismic sequences on structures
with hysteretic or damped dissipative behaviour. Soil Dynamics and Earthquake Engineering,
97, 205–215.
Ruiz-García, J., & Aguilar, J. D. (2015). Aftershock seismic assessment taking into account post-
mainshock residual drifts. Earthquake Engineering and Structural Dynamics, 44(9), 1391–1407.
References 205
Shin, M., & Kim, B. (2017). Effects of frequency contents of aftershock ground motions on
reinforced concrete (RC) bridge columns. Soil Dynamics and Earthquake Engineering, 97, 48–59.
Song, R., Li, Y., & Van de Lindt, J. W. (2016). Loss estimation of steel buildings to earthquake
mainshock-aftershock sequences. Structural Safety, 61, 1–11.
Soysal, B. F., Binici, B., & Arici, Y. (2016). Investigation of the relationship of seismic intensity
measures and the accumulation of damage on concrete gravity dams using incremental dynamic
analysis. Earthquake Engineering and Structural Dynamics, 45(5), 719–737.
Wang, G., Wang, Y., Lu, W., Yan, P., Zhou, W., & Chen, M. (2017a). Damage demand assess-
ment of mainshock-damaged concrete gravity dams subjected to aftershocks. Soil Dynamics and
Wang, G., Wang, Y., Lu, W., Yu, M., & Wang, C. (2017b). Deterministic 3D seismic damage analysis
of Guandi concrete gravity dam: A case study. Engineering Structures, 148, 263–276.
Wang, G., Wang, Y., Lu, W., Zhou, C., Chen, M., & Yan, P. (2015). XFEM based seismic potential
Xia, Z., Ye, G., Wang, J., Ye, B., & Zhang, F. (2010). Fully coupled numerical analysis of repeated
shake-consolidation process of earth embankment on liquefiable foundation. Soil Dynamics and
Yazdani, Y., & Alembagheri, M. (2017). Nonlinear seismic response of a gravity dam under near-
621–632.
Zhai, C., Zheng, Z., Li, S., & Pan, X. (2017). Damage accumulation of a base-isolated RCC building
under mainshock-aftershock seismic sequences. KSCE Journal of Civil Engineering, 21(1), 364–
377.
Zhang, S., Wang, G., & Sa, W. (2013a). Damage evaluation of concrete gravity dams under
16–27.
Chapter 9
Seismic Performance Evaluation
of Dam-Reservoir-Foundation Systems
to Near-Fault Ground Motions
9.1 Introduction
Ground motion records obtained in recent major strong earthquakes, such as Loma
Prieta (1989), Northridge (1994), Kobe (1995), Kocaeli (1999), and Chi-Chi (1999),
revealed unique characteristics of ground motions in a near-fault area. The seismic
ground motions recorded within the near-fault region of an earthquake at stations
located toward the direction of the fault rupture are significantly different from the
usual far-fault ground motions observed at a large distance (Chopra and Chintana-
pakdee 2001). Forward directivity and fling effects have been identified by seismol-
ogists as the primary characteristics of near-fault ground motions (Mavroeidis and
Papageorgiou 2003). The fault-normal components of ground motions often contain
large displacements and velocity pulses. Such a pronounced pulse does not exist in
far–fault ground motions. The pulses are strongly influenced by the rupture mecha-
nism, the slip direction relative to the site, and the location of the recording station
relative to the fault which is termed as ‘directivity effect’ due to the propagation of
the rupture toward the recording site (Wang et al. 2002; Somerville 2003; Bray and
Rodriguez-Marek 2004).
Because of the unique characteristics of near-fault ground motion, the ground
motions recorded in the near-fault region, which expose the structure to high input
energy in the beginning of the earthquake (Liao et al. 2004), have the potential
to cause a large response and considerable damage to structures. Therefore, struc-
tural response to near-fault ground motions has received much attention in recent
years. The effects of near-fault ground motions on many civil engineering structures,
such as buildings and bridges, have been investigated in many recent studies. Xiang
and Alam (2019) evaluated the seismic vulnerability of double-column bridge bents
retrofitted with different braces under near-fault and far-field ground motions. Xin
et al. (2019) investigated the seismic behaviors of long-span concrete-filled steel
tubular arch bridges under near-fault fling-step motions, and the effects of different
components in fling-step motions on the seismic response of the bridge are conducted
https://doi.org/10.1007/978-981-15-6194-8_9
208 9 Seismic Performance Evaluation of Dam-Reservoir-Foundation …
by the parametric analysis. Ma et al. (2019) developed a probabilistic seismic demand

model to evaluate the seismic performance of regular continuous bridges subjected to
pulse-like near-fault ground motions. Their results showed that the near-fault ground
motions can cause more serious damage to the bridge. Kabir et al. (2019) assessed
the seismic vulnerability of a multi-span bridge subjected to near-fault, far-fault, and
long duration ground motions. They thought the failure probabilities of the bridge
system were dominated by the long duration ground motions than those of near-
fault and far-fault ground motions. Ardakani and Saiidi (2018) developed a simple
empirical method to evaluate the residual displacements of concrete bridge columns
subjected to near-fault ground motions. Hedayati Dezfuli et al. (2017) concluded
that near-fault ground motions will cause more damage to the bridge structure than
far-fault ground motions in terms of maximum lateral displacement, generated shear
force, residual deformation, and energy dissipation capacity of the bridge bearings,
duck, and piers. Yang et al. (2017) demonstrated that the seismic demands of the seis-
mically isolated bridge located in the vicinity of a surface fault rupture or crossing
a fault rupture zone will be significantly underestimated when the high-pass filtered
ground motion is utilized. Li et al. (2017) evaluated the seismic response of super-
span cable-stayed bridges built in the forward region, middle region, and backward
region relative to fault rupture. Their results showed that the bridge located in the
middle region will experience a larger response than that in the forward region and
backward region. Yang et al (2019) proposed an intrinsic length scale significance
to analyze the dimensional response of bilinear SDOF systems subjected to near-
fault ground motions. Manafpour and Kamrani Moghaddam (2019) investigated the
effects of seismic sequences on the dynamic response of RC SDOF systems subjected
to near-fault and far-fault ground motions. Both the pulse-like and no-pulse motions
were considered in the near-fault ground motions. Du et al. (2020) investigated the
seismic performance of bucking-restrained braced RC frame structures subjected
to near-fault ground motions with forward directivity and fling-step effects. Their
results indicated that near-fault ground motions with forward-directivity and fling-
step effects will produce larger maximum interstory drift ratio and story drift than
non-pulse-like near-fault ground motions. Bilgin and Hysenlliu (2020) predicted the
nonlinear displacement demands of low and mid-rise masonry buildings subjected
to near-fault and far-fault ground motions. They concluded that the near-fault ground
motions have significantly more damage potential on masonry structures than far-
fault ground motions. Bhagat et al. (2018) investigated the nonlinear seismic response
of base-isolated buildings subjected to near-fault ground motions with fling-step and
forward-directivity characteristics. Behesthi Aval et al. (2018) discussed the seismic
performance of tunnel form buildings subjected to near-fault and far-fault ground
motions through incremental dynamic analysis method. They found that the proba-
bility of reaching structural elements to preliminary damage levels under near-fault
ground motions was increased up to 20% when compared with far-fault ground
motions. Cao et al. (2016) investigated the effects of wave passage on torsional
response of elastic buildings under near-fault pulse-like ground motions. A simpli-
fied pulse model is used to represent the main features of near-fault ground motions
which will cause large torsional response to the buildings. Tajammolian et al. (2016)
investigated the effects of mass eccentricity on the seismic response of isolated

structures under near-fault ground motions. Eslami et al. (2016) conducted a numer-
ical investigation on the seismic retrofitting of RC buildings subjected to near-fault
ground motions using the carbon fiber-reinforced polymer sheets. They concluded
that the retrofitting scheme can substantially improve the seismic performance of
RC buildings under impulsive ground motions. Beiraghi et al. (2016a) examined
the seismic behavior of RC wall tall buildings subjected to forward directivity near-
fault and far-fault ground motions, and concluded that near-fault ground motions
will produce considerably larger seismic demands (i.e. curvature ductility, inter-
story drift and displacement) as compared with the far-fault ground motions. They
also found that the near-fault motion pulse will transfer more energy quantity to
the RC core-wall buildings (Beiraghi et al. 2016b). Sun et al. (2020) compared the
far-fault and near-fault ground motion effects on the probabilistic seismic response
of arched hydraulic tunnels, and found that the near-fault ground motions with fling-
step effects can cause higher probability of exceedance to the arched hydraulic tunnel
than other types of ground motions. Konstandakopoulou et al. (2019) investigated
the nonlinear dynamic behavior of 3D offshore platforms including the pile-soil-
structure interaction effects under near-fault ground motions. They found that near-
fault ground motions from reverse or oblique reverse faults will cause more intense
structural response as compared with ground motions that correspond to strike-slip
faults. Chen et al. (2020) examined the elastoplastic dynamic response of intake
tower structures subjected to near-fault ground motions with forward-directivity and
fling-step effects. Their results showed that the near-fault ground motions can lead
to great nodal displacement of intake tower structures. It can be clearly seen from
these studies that the importance of near-fault ground motion effects on the nonlinear
dynamic response of structures has been highlighted. Aseismic design of structures
should be given attention to the characteristics of near-fault ground motions due to
their impulsive effects on structures.
It should be noted that few studies have focused their attention on the nonlinear
dynamic response and seismic damage of concrete gravity dams subjected to near-
fault ground motions. For example, Akköse and Şimşek (2010) studied the seismic
response of a concrete gravity dam subjected to near-fault and far-fault ground
motions including dam-water-sediment-foundation rock interaction. They found that
plastic deformations in the dam subjected to near-fault ground motion are greater than
those subjected to far-fault ground motion. Bayraktar et al. (2009; 2010) examined the
effects of near-fault and far-fault ground motions on the nonlinear response of gravity
dams. The results revealed that there are more seismic demands on displacements and
stresses when the dam is subjected to near-fault ground motion. Yazdani and Alem-
bagheri (2017a, b) investigated nonlinear seismic response analyses and seismic
vulnerability assessment of gravity dams located in near-fault areas by proposing a
probabilistic seismic demand model for gravity dams considering near-field earth-
quake ground motions. Although previous studies provided some information on the
effects of near-fault ground motions on the response of dams, there is no sufficient
research about the near-fault ground motion effects on the seismic damage of concrete
gravity dams. Zou et al (2019) explored the seismic slope stability of a high rockfill
dam considering strain softening subjected to pulse-like ground motions. The results
showed that the pulse-like ground motion has a significant impact on sliding displace-
ment of the rockfill dam. Gorai and Maity (2019) compared the effects of near-fault
and far-fault ground motions on the seismic performance of concrete gravity dams
in terms of demand capacity ratio, cumulative inelastic duration, and spatial extent
of overstress region.
The main objective of this chapter is to investigate and compare the dynamic
behavior of concrete gravity dams subjected to near-fault and far-fault ground motion
excitations with considering the effects of dam-reservoir-foundation interaction. The
Concrete Damage Plasticity (CDP) model which includes the strain hardening or
softening behavior is adopted to study the seismic response of concrete gravity dams
under earthquake conditions. The 1979 Imperial Valley, 1989 Loma Prieta, 1994
Northridge, and 1999 Chi-Chi earthquake records which display ground motions
with an apparent velocity pulse are selected to represent the near-fault ground motion
characteristics. In this study, the term “near-fault ground motion” is referred to the
ground motion record obtained in the vicinity of a fault with the apparent velocity
pulse (pulse duration larger than 1.0 s), and the peak ground velocity/peak ground
acceleration (PGV/PGA) value which is larger than 0.1 s. The earthquake ground
motions recorded at the same site from other events that the epicenter far away from
the site are employed as the far-fault ground motions, which are used to compare with
near-fault ground motions. The Lagrangian approach is used for the finite element
modeling of dam-reservoir-foundation interaction problem. The Koyna gravity dam
is employed as a numerical application. Nonlinear seismic damage analyses of the
selected concrete dam subjected to both near-fault and far-fault ground motions are
performed. The influence of near-fault ground motions on the dynamic response and
seismic damage of concrete gravity dams is examined.
9.2 Characteristics of Near-Fault Ground Motions
It is well known that the earthquake is a complex natural phenomenon associated

with the abrupt energy release caused by the fault rupture. There are many factors that
affect the types of ground motions at a site, such as the underlying soil condition at
the site, earthquake source mechanism, the magnitude of the earthquake, the distance
from the earthquake, propagation path of waves, and topographic features. A near-
fault ground motion is one that is experienced near the rupturing fault line. When a
site lies within the near-fault region (usually less than 10 km), a unique set of factors
controls the motion that is recorded, which will be different from those observed
further away from the seismic source.
The characteristics of near-fault ground motions are linked to the fault geometry
and the orientation of the traveling seismic waves. The most primary characteristics
of near-fault ground motions concerning the field of structural and civil engineering
are the impulsive characteristic of the velocity (i.e. the forward directivity effect) and
displacement ground motions affecting at long periods (i.e. fling step effect), which
9.2 Characteristics of Near-Fault Ground Motions 211
have caused severe structural damage in recent major earthquakes. Another feature
is the hanging wall/footwall effect. Although the near-fault ground motion has the
typical characteristic of a short duration, high amplitude pulse, these phenomena
are not always observed, which are based on the focal mechanism and the direction
of the fault rupture. This means that not all sites within the near-fault region will
experience the high amplitude pulse-like behavior during the earthquake.
9.2.1 Forward Directivity Effect
One of the critical factors affecting ground motion in the near field area is the direction
in which rupture progresses from the hypocenter along the zone of rupture. An earth-
quake is a shear dislocation that begins at a point on a fault and spreads throughout
the fault plane at a velocity that is almost equal to the shear wave velocity. The rupture
velocity is generally slightly less than the shear wave velocity. If a site is located at
one end of a fault and the rupture propagates from the other end. The rupture velocity
is on the average 80% of the shear wave velocity. (Near-fault ground motions and
structural design issues.)
Figure 9.1 shows the forward directivity phenomenon. When the fault ruptures
toward the site A, the shear wave energy will arrive at this site within a short time
frame (almost the same time). Because of this, the seismic energy from each fault
segment arrives together at site A, and accumulates in front of the propagating
rupture, which is expressed in the forward directivity region as a relatively short
duration record containing large velocity amplitude. However, the seismic energy
will be distributed when arrives at Site B over a longer period of time, resulting in
a longer duration record with lower velocity amplitudes. This can be explained by
other scientific phenomena such as the sonic boom or the Doppler effect, which is
essentially the same principle and may be thought of accordingly.
Forward directivity effects can be classified as forward, reverse, or neutral, as
shown in Fig. 9.2. When the rupture propagates toward the site and the angle between
the fault and the direction from the hypocenter to the site is reasonably small, this
site is likely to generate the forward directivity effects. This is because that the
rupture often propagates at a velocity close to the velocity of shear wave radiation,
energy is accumulated in front of the propagating rupture and is expressed in the
forward directivity region as a large velocity pulse (Somerville et al. 1997). Forward
directivity effects can be presented both for strike-slip and dip-slip events. In strike-
slip events, forward directivity conditions are typically largest for sites near the end
of the fault when the rupture front is moving towards the site and slip vector points
toward the site as well. In dip-slip events, forward directivity conditions occur at sites
located in the up-dip projection of the fault plane (Bray and Rodriguez-Marek 2004).
This means that the rupture direction is aligned up on the fault plane and the slip
vector points upwards as well. The conditions that lead to high forward directivity
can be summarized as follows (Somerville 2005):
Fig. 9.1 Effects of accumulation of seismic energy

n
io
t
ul
ct
Fa
ire
Neutral
D
Forward
re
tu
up
R
Reverse Neutral
Fig. 9.2 Defining a site as forward, reverse, or neutral directivity

9.2 Characteristics of Near-Fault Ground Motions 213
• The smaller the angle between the direction of the rupture propagation and the
direction of the seismic waves travelling from the fault to the site, the higher the
forward directivity effects.
• The larger the fraction of the fault rupture surface that lies between the hypocenter
and the site, the higher the forward directivity effects.
• Forward directivity does not exist if the slip is concentrated near one end of the
fault where the site is located, even if these geometrical conditions are satisfied.
If the rupture propagates away from the site, this site is likely to demonstrate
reverse directivity, and the ground motion is characterized by long duration and low
amplitude at long periods. When the rupture propagation is neither predominantly
towards nor away from the site, this means that the site is located close to the epicenter
of the strike-slip faulting and located off the end of the up-dip projection of reverse
faults. In this condition, the site experiences the neutral directivity. The phrase “direc-
tivity effects” usually refers to “forward directivity effects”, as this case is expected
to result in ground motions that are more critical to engineering structures.
9.2.2 Fling Step Effect
The fling step effects are the consequences of a permanent ground displacement
due to the static deformation field of the earthquake, occurring over a discrete time
interval of several seconds as the fault slip is developed (Stewart et al., 2002). This
effect is observed close to the fault and can be significant when excessive tectonic
deformations occur due to large slip on the fault plane (as shown Fig. 9.3). For a
strike-slip fault, the offset is partitioned primarily on the fault parallel component.
For a dip-slip fault, the offset is partitioned chiefly on the fault normal and vertical
components. Pulses from the fling step effects have different characteristics than
those from forward directivity effects. The forward directivity effect is a dynamic
phenomenon that produces no permanent ground displacement and hence two-sided
velocity pulses. Whereas the fling step effect is characterized by a unidirectional
large amplitude velocity pulse and a discrete step in the displacement time history.
The effects of both fling step and forward directivity can be superimposed on one
another in the fault normal component of a dip-slip earthquake, whereas the parallel
component of the fault will show neither effects. In general, the fault normal and fault
parallel components show similar patterns in the far-field ground motion. However,
the near-fault ground motion can show a substantial difference.
9.2.3 Hanging Wall Effect
In the near-fault earthquake, there are not only the forward directivity effect and
the fling step effect, but also the hanging wall effect. Since the sites located on the
200
150
100
Velocity (cm/s)
50
0
-50
-100
-150
-200
0 5 10 15 20 25 30 35 40 45 50 55
Time (s)
200
Displacement (cm)
150
100
50
0
-50
-100
-150
-200
0 5 10 15 20 25 30 35 40 45 50 55
Time (s)
Fig. 9.3 Typical velocity and displacement time history curves of 1999 Kocaeli earthquake with
fling step effect
hanging wall of a dip-slip fault are generally closer to the fault than those located
on the foot wall at the same closest distance, the ground motions of the hanging
wall are larger than those located on the foot wall at the same closest distance. This
is the hanging wall effect of the near-fault ground motion, as shown in Fig. 9.4.
Abrahamson and Somerville (1996) observed that the peak horizontal accelerations
on the hanging wall sites were about 50% greater than those located on the foot wall
over a distance range of 10–20 km during the 1994 Northridge earthquake. Hanging-
wall effects have also been observed in some reverse dip-slip faulting earthquakes
of the past and can be expected in future reverse events.
Fig. 9.4 Hanging wall effect of near-fault ground motions

9.3 Near-Fault and Far-Fault Ground Motion Records Considered in This Study 215
9.3 Near-Fault and Far-Fault Ground Motion Records

Considered in This Study
One of the most important characteristics of near-fault events concerning the field
of structural and civil engineering is the impulsive character of the velocity and
displacement ground motions affecting at long periods and thereby having a severe
influence on the structures. The selection of ground motion is one of the important
parts of seismic evaluation of gravity dams. In this study, 10 as-recorded near-fault
ground records are selected as the input ground motion from the 1979 Imperial
Valley, 1989 Loma Prieta, 1994 Northridge, and 1999 Chi-Chi earthquakes. These
records, which display ground motions with an apparent velocity pulse, are selected
to represent the near-fault ground motion characteristics. The velocity pulse dura-
tion in the near-fault ground motions is larger than 1.0 s. In addition, the ratio of the
peak ground velocity (PGV) to the peak ground acceleration (PGA) of the near-fault
ground motions is larger than 0.1 s. On the contrary, another set of ground motion
records, recorded at the same site condition from the same earthquake events with
epicenter far away from the site, is employed to represent the characteristics of far-
fault ground motion. The properties of these records are depicted in Table 9.1. The
ground motion records are obtained from the databases of the COSMOS (COSMOS
2019) and PEER (PEER 2019). As a typical example, Fig. 9.5 shows the accelera-
tion, velocity and displacement time histories of the fault normal component of the
near-fault and far-fault ground motions. Figure 9.6 shows the acceleration response
spectrum with damping ratio 5% for selected near-fault and far-fault ground motion
records in the Northridge earthquake.
As shown in Fig. 9.5, it can be seen from that there is a distinct difference between
the velocity pulse of ground motions recorded at near-fault and far-fault regions. The
near-fault ground motion possesses a significantly long period pulse in the accelera-
tion time history that is consistent with velocity and displacement histories. The long
period response of the near-fault ground motion is more excessive than the far-fault
ground motion.
The near-fault and far-fault ground motion records are normalized to have the peak
ground acceleration (PGA) equal to 0.3 g. In these analyses, only the first 15 s of the
Northridge, Imperial Valley, and Loma Prieta earthquake records and the first 40 s of
the Chi-Chi earthquake records are considered. Figure 9.7 presents acceleration time
history for scaled near-fault and far-fault ground motion records. Figure 9.8 presents
velocity time history for scaled near-fault and far-fault ground motion records.
9.4 Seismic Performance Evaluation Methods
A systematic and rational methodology for seismic performance assessment and

qualitative damage estimation using standard results from linear time-history anal-
ysis is presented. The performance evaluation and assessment of the probable level of
Table 9.1 Properties of selected near-fault and far-fault ground motions considered in this investigation
216
No. Ground motion Earthquake Distance to fault (km) Station location Mw Comp. PGA PGV (cm/s) PGV/PGA (s)
(cm/s2 )
1 Near-fault Northridge 1994 7.1 Newhall, CA–Los 6.7 360 578.20 94.70 0.164
Angeles County Fire
#24279
2 Far-fault Northridge 1994 18.4 Los Angeles, CA - 6.7 35 576.93 29.75 0.052
Fire Station 108
#5314
3 Near-fault Northridge 1994 8.6 Sylmar, CA - Jensen 6.7 22 560.30 77.23 0.138
Filtration Plant #655
4 Far-fault Northridge 1994 29.4 Warm Springs 6.7 90 221.20 13.40 0.061
#24272
5 Near-fault Northridge 1994 8.6 Los Angeles 6.7 64 317.6 44.54 0.140
Reservoir #2141
6 Far-fault Northridge 1994 17.9 Tarzana, CA - Cedar 5.3 90 365.30 11.80 0.032
Hill #24436
7 Near-fault Imperial Valley 1979 5.2 El Centro, CA - Array 6.5 230 360.37 95.89 0.266
Sta 5 #0952
8 Far-fault Imperial Valley 1979 21.7 El Centro, CA - Array 6.5 230 131.07 12.70 0.096
Sta 13 #5059
9 Near-fault Imperial Valley 1979 8.8 Holtville, CA - Post 6.5 225 242.96 51.90 0.214
Office #5055
10 Far-fault Imperial Valley 1979 21.8 Superstition Mtn, CA 6.5 135 182.19 8.65 0.047
- Camera Site #0286
11 Near-fault Loma Prieta 1989 6.3 Gilroy Array Sta 3 # 6.5 90 362.00 43.80 0.121
47381
(continued)
9 Seismic Performance Evaluation of Dam-Reservoir-Foundation …
No. Ground motion Earthquake Distance to fault (km) Station location Mw Comp. PGA PGV (cm/s) PGV/PGA (s)
(cm/s2 )
12 Far-fault Loma Prieta 1989 22.7 Gilroy Array Sta 7 6.5 90 314.30 16.30 0.052
#57425
13 Near-fault Loma Prieta 1989 2.8 Corralitos, CA 6.5 90 469.40 47.50 0.101
#57007
14 Far-fault Loma Prieta 1989 16.9 Coyote Lake Dam, 6.5 285 471.00 37.50 0.079
CA #57217
15 Near-fault Chi-Chi 1999 8.9 Taichung, Taiwan 7.6 90 142.70 32.40 0.227
#TCU050
16 Far-fault Chi-Chi 1999 32.0 Ilan, Taiwan #ILA067 7.6 90 195.70 11.40 0.058
9.4 Seismic Performance Evaluation Methods
#TCU072
18 Far-fault Chi-Chi 1999 37.2 Chiayi, Taiwan 7.6 0 254.90 22.90 0.090
#CHY014
#TCU076
20 Far-fault Chi-Chi 1999 31.1 Chiayi, Taiwan 7.6 0 201.60 17.9 0.089
#CHY086
217
0.6 Near-fault 0.6 Far-fault
Acceleration (g)
Acceleration (g)
0.3 0.3
0.0 0.0
-0.3 -0.3
-0.6 -0.6
0 3 6 9 12 15 0 3 6 9 12 15
Time (s) Time (s)
100 100
Velocity (cm/s)
Velocity (cm/s)
50 50
0 0
-50 -50
-100 -100
0 3 6 9 12 15 0 3 6 9 12 15
Time (s) Time (s)
50 50
Displacement (cm)
Displacement (cm)
25 25
0 0
-25 -25
-50 -50
0 3 6 9 12 15 0 3 6 9 12 15
Time (s) Time (s)
(a) Near-fault ground motion (b) Far-fault ground motion
Fig. 9.5 Sample of acceleration, velocity, and displacement time histories for a near-fault ground
motion recorded at Newhall station, and b far-fault ground motion recorded at Los Angeles station
in Northridge earthquake
Fig. 9.6 Acceleration 2.5

response spectrum with
damping ratio 5% for
Response Acceleration (g)
2.0
selected near-fault and
far-fault ground motions Near-fault
1.5 Far-fault
1.0
0.5
0.0
0 1 2 3 4
Period (s)
damage are formulated based on the demand-capacity ratios for stresses, the cumu-
lative duration of stress excursions beyond the tensile strength of the concrete and
the spatial extent of overstressed regions. This evaluation is applied to the damage
control range of strains shown in Fig. 9.9.
9.4 Seismic Performance Evaluation Methods 219
1 2 3 4
5 6 7 8
9 10 11 12
0.6 g
13 14 15
16
17 19 20
18
20 s
Fig. 9.7 Acceleration time histories for the scaled near-fault and far-fault ground motions used in
this study. For ground motion information see Table 9.1
1 2
3 5
4 6
7 8
9 10 11 12
13 14
120 cm/s
15
16
17 18
20
19
20 s
Fig. 9.8 Velocity time histories for the scaled near-fault and far-fault ground motions used in this
study. For ground motion information see Table 9.1
The demand-capacity ratio (DCR) for concrete gravity dams is defined as the
ratio of the calculated maximum principal stress to tensile strength of the concrete.
The tensile strength of the plain concrete used in this definition is the static tensile
strength characterized by the uni-axial splitting tension tests or from
f t = 1.7 f c2/3 (9.1)

Demand-Capacity Ratio 2
1 T=0.24s
-1
-2
0.0 0.24 0.48 0.72 0.96 1.20 1.44 1.68 1.92

(a) Time (s)
Duration=0.4s
0.5 Stress Damage Control Region

Cumulative Inelastic Duration (s)
3.4fc
/ 2/3
Significant damage
0.4 Assess using nonlinear analysis B Apparent Tensile
Strength
0.3
0.2
1.7fc
/ 2/3
A
0.1 Low to moderate damage
Assess using linear analysis
0.0
0.0 2.0
Demand-Capacity Ration (DCR) Strain
(b) (c)
Fig. 9.9 Illustration of seismic performance and damage criteria (Ghanaat 2004)
proposed by Raphael (1984), where f c is the compressive strength of the concrete.

The maximum permitted DCR for the linear analysis of dams is 2. This corresponds
to a stress demand twice the static tensile strength of the concrete. As illustrated in
the stress-strain curve in Fig. 9.9c, the stress demand associated with a DCR of 2
corresponds to the so called “apparent dynamic” tensile strength of the concrete, a
quantity proposed by Raphael for evaluation of the results of linear dynamic analysis.
The cumulative inelastic duration of stress excursions is defined as the total dura-
tion of the stress excursions above a stress level associated with a DCR ≥ 1. The
higher the cumulative duration, the higher is the possibilities for more damage. The
sinusoidal stress history with five stress cycles is shown in Fig. 9.9a. The sinusoidal
cycle duration of the stress excursion above the tensile strength is equal to T/3, where
T is the period of the sinusoid. The total inelastic duration for all five stress excursions
(shaded area) amounts to 5T/3. Considering that the period of the signal is 0.24 s, the
cumulative inelastic duration of stress excursions is 0.4 s. The cumulative duration
for a DCR of 2 is assumed zero. For gravity dams a lower cumulative duration of
0.3 s (in Fig. 9.10) is assumed, mainly because gravity dams resist loads by cantilever
mechanism only, as opposed to arch dams that rely on both the arch and cantilever
actions.
9.4 Seismic Performance Evaluation Methods 221
Fig. 9.10 Seismic 0.5

performance threshold
Cumulative Inelastic Duration (s)

curves for concrete gravity
dams 0.4
Significant damage
Assess using nonlinear analysis
0.3
0.2
0.1
Low to moderate damage
Assess using linear analysis
0.0
1.0 2.0
Demand-Capacity Ration (DCR)
The seismic performance and probable level of damage of concrete gravity dams
are evaluated on the basis of the demand-capacity ratios, the cumulative inelastic
duration and the spatial extent of overstressed regions described above. Three
performance levels are considered (Ghanaat 2004):
(1) Minor or no damage. The dam response is assumed to be elastic if the demand
capacity ratio (DCR) ≤ 1. The dam is considered to behave in the elastic range
with little or no possibility of damage.
(2) Acceptable level of damage. If the estimated DCR > 1, the dam will exhibit a
nonlinear response in the form of cracking and joint opening. If the estimated
DCR < 2, the overstressed regions are limited to 15 percent of the dam cross-
section surface area, and the cumulative duration of stress excursion also falls
below the performance curve given in Fig. 9.10, the level of nonlinear response
or cracking can be considered acceptable with no possibility of failure.
(3) Severe damage. The damage is considered as severe if the DCR > 2, or the
cumulative overstress duration for all DCR values between 1 and 2 falls above
the performance curves as shown in Fig. 9.10. In these situations, nonlinear time
history analysis may be required to further investigate the performance of the
dam.
9.5 Near-Fault Ground Motion Effects on Seismic

Performance of Concrete Gravity Dams Using Linear
and Nonlinear Evaluation Methods
Due to the unique characteristics of the near-fault ground motions and their potential
to cause severe damage to structures, the interest on the near-fault effect on struc-
tural response has increased. Earthquake engineers have been considering methods
to incorporate near-fault effects in engineering design (Mavroeidis and Papageor-

giou 2003). The seismic safety evaluation of high dams subjected to near-fault
ground motions remains a crucial problem in dam construction. In this study, a
set of selected near-fault and far-fault ground motion records is used to examine the
near-fault ground motion effects on seismic performance of concrete gravity dams
with considering the effects of dam-reservoir-foundation interaction. Only the hori-
zontal component of the seismic input is considered in this analysis. A 5% damping
ratio has been assumed in all dynamic analyses. The integration time step used in the
time history analysis is 0.005 s. The geometry and finite element (FE) discretization
of the Koyna dam–reservoir–foundation system are illustrated in Sect. 2.5 (Chap. 2).
The material parameters for concrete, foundation rock, and water are the same as
Table 2.1.
It is known that seismic performance assessment of concrete gravity dams has
become significantly important in view of the observed lack of seismic resistance of
dams during strong earthquakes. Due to the severe damage effects of impulsive type
motions on structures, satisfactory seismic performance of concrete gravity dams
located in the near-fault region is necessary because the release of the impounded
large quantity of water can cause a considerable amount of devastation in the
downstream populated areas. The state-of-the-art for evaluating seismic response of
concrete gravity dams has progressively moved from elastic static analysis to elastic
dynamic, nonlinear static and finally nonlinear dynamic analysis. The traditional
seismic performance evaluation method is based on linear theory. It is assumed that
the tensile stresses should be less than the dynamic tensile strength of the concrete
material. However, in practice up to five stress excursions above the tensile strength of
the concrete material has been considered acceptable based on engineering judgment
and other considerations (USACE, 2003). To overcome the above shortcomings, a
new performance evaluation approach mainly based on the demand-capacity ratios
(DCR) and cumulative overstress duration, was proposed by Ghanaat (2002, 2004).
Yamaguchi et al. (2004) used this method to evaluate the seismic performance of
concrete gravity dams. Sevim (2011) investigated the effect of material properties on
the seismic performance of arch dam-reservoir-foundation interaction systems based
on the Lagrangian approach using demand-capacity ratios.
9.5.1 Seismic Input
A type of 8 records from Table 9.1 is selected as the input ground motion for seismic
performance of concrete gravity dams. The assembled database can be investigated
in two sub-data sets. The first set which displays ground motions with an apparent
velocity pulse is selected to represent the near-fault ground motion characteristics.
On the contrary, the second set of four ordinary ground motion records is employed
to represent the characteristics of far-fault ground motion. Informations pertinent to
the ground motion data sets, including closest fault distance, station, component of
earthquake and peak ground acceleration (PGA), peak ground velocity (PGV), and
9.5 Near-Fault Ground Motion Effects on Seismic Performance of Concrete … 223
the ratio of PGV to PGA, are presented in Tables 9.2 and 9.3. The ground motion
records are obtained from the databases of the COSMOS (COSMOS 2019) and PEER
(PEER 2019).
The near-fault and far-fault ground motion records are normalized to have the peak
ground acceleration (PGA) equal to 0.30 g. In the analyses, only the first 15 s of the
earthquake records are considered. Two different levels of peak ground acceleration
(PGA) are considered for the input motions: 0.20 and 0.30 g.
9.5.2 Seismic Performance Evaluation Using Linear

Dynamic Analysis
The seismic performance of Koyna dam is firstly evaluated by using the linear time
history analysis results for both near-fault and far-fault ground motions scaled to 0.20
and 0.30 g. The response of the dam has been determined in terms of the response
parameters such as the demand-capacity ratios (DCR), cumulative overstress duration
and overstressed areas.
The maximum principle tensile stress histories for both near-fault and far-fault
earthquake records with a PGA level of 0.20 g are plotted by displaying the demand-
capacity ratios (DCR) in Figs. 9.11 and 9.12. Figure 9.13 presents the corresponding
performance evaluation curves for both near-fault and far-fault earthquakes at the
0.20 g level. Figure 9.14 shows the overstressed regions of the dam in terms of the
DCR values for near-fault ground motions scaled to 0.20 g. The shaded areas indi-
cated that the concrete material regions where the computed DCR values exceeded
1.0. It is clear from Fig. 9.11 and Fig. 9.13a that the dam subjected to far-fault ground
motions with a PGA level of 0.20 g is considered to behave in the elastic range with
little or no possibility of damage, because the maximum principle tensile stresses are
almost less than the tensile strength of the concrete material (2.9 MPa).
It can be seen from Fig. 9.12 that the maximum principal stress peaks in excess
of the tensile strength of the concrete material vary from 4 to 7 cycles for different
near-fault earthquake records with a PGA level of 0.20 g. Figure 9.13b shows that
the cumulative overstress durations for DCR = 1 are over than 0.3 s for all cases,
but most of the resulting cumulative duration curves are located within the zone
of acceptable performance, and the areas of overstressed regions associated with
DCR > 1 represent only 2.5, 2.0, 5.5, and 1.2% of the total section area (Fig. 9.14).
It is therefore concluded that the actual seismic performance of the dam subjected
to near-fault ground motions at the 0.20 g level is likely to exhibit some tensile
cracking but the global consequences of the resulting damage are expected to be
minor. The damage is considered as an acceptable level. Based on these results, it
can be concluded that the results from the linear time history analysis still provide
sufficient information to characterize the response of the dam for both near-fault and
far-fault ground motions with a PGA level of 0.20 g.
224
Table 9.2 Properties of selected near-fault ground motions considered in this investigation
No. Earthquake Year Distance to fault (km) Station location Mw Comp. PGA (cm/s2 ) PGV (cm/s) PGV/PGA (s)
1 Northridge 1994 8.6 Sylmar, CA - Jensen 6.7 22 560.30 77.23 0.138
Filtration Plant #655
2 Imperial Valley 1979 5.2 El Centro, CA - Array 6.5 230 360.37 95.89 0.266
Sta 5 #0952
3 Loma Prieta 1989 2.8 Corralitos, CA #57007 6.5 90 469.40 47.50 0.101
4 Kobe 1995 1.0 JMA station 6.9 0 805.45 81.3 0.101
9 Seismic Performance Evaluation of Dam-Reservoir-Foundation …
Table 9.3 Properties of selected far-fault ground motions considered in this investigation
No. Earthquake Year Distance to fault (km) Station location Mw Comp. PGA (cm/s2 ) PGV (cm/s) PGV/PGA (s)
1 Northridge 1994 29.4 Warm Springs #24272 6.7 90 221.20 13.40 0.061
2 Imperial Valley 1979 21.8 Superstition Mtn, 6.5 135 182.19 8.65 0.047
CA-Camera Site #0286
3 Loma Prieta 1989 16.9 Coyote Lake Dam, CA 6.5 285 471.00 37.50 0.079
#57217
4 Kobe 1995 22.5 Kakogawa 6.9 0 246.58 18.74 0.076
9.5 Near-Fault Ground Motion Effects on Seismic Performance of Concrete …
225
Max. Principal Stress (MPa)

(a) Northridge #24272 (b) Imperial Valley #0286
6 DCR=2 6 DCR=2
4 4
DCR=1 DCR=1
2 2
0 0
0 3 6 9 12 15 0 3 6 9 12 15
Time (s) Time (s)
(a) (b)

DCR=2 (c) Loma Prieta #57217 DCR=2 (d) Kobe #Kakogawa
6 6
4 4
DCR=1 DCR=1
2 2
0 0
0 3 6 9 12 15 0 3 6 9 12 15
Time (s) Time (s)
(c) (d)
Fig. 9.11 Time histories of maximum principal tensile stresses for far-fault ground motions with
a PGA level of 0.20 g. a Northridge #24272, b Imperial Valley #0286, c Loma Prieta #57217, and
d Kobe # Kakogawa
DCR=2 (a) Northridge #655 (b) Imperial Valley #0952

6 6 DCR=2
4 4
DCR=1 DCR=1
2 2
0 0
0 3 6 9 12 15 0 3 6 9 12 15
Time (s) Time (s)
(a) (b)
DCR=2 (c) Loma Prieta #57007 DCR=2 (d) Kobe #JMA

6 6
4 4
DCR=1 DCR=1
2 2
0 0
0 3 6 9 12 15 0 3 6 9 12 15
Time (s) Time (s)
(c) (d)
Fig. 9.12 Time histories of maximum principal tensile stresses for near-fault ground motions with
d Kobe # JMA
Figure 9.15 and Fig. 9.16 display the maximum principle tensile stress histories
for both near-fault and far-fault earthquake records with a PGA level of 0.30 g.
The corresponding performance evaluation curves are shown in Fig. 9.17. The over-
stressed regions computed using the linear time history analyses in terms of the
DCR values for near-fault ground motions scaled to 0.30 g are shown in Figs. 9.18
Cumulative Overstress Duration (s) 0.4 0.4
Cumulative Overstress Duration (s)

Northridge Northridge
Imperail Valley Imperail Valley
0.3 Loma Prieta 0.3 Loma Prieta
Kobe Kobe
Target Target
0.2 0.2
0.1 0.1
0.0 0.0
1.0 1.2 1.4 1.6 1.8 2.0 1.0 1.2 1.4 1.6 1.8 2.0
Demand-Capacity Ration (DCR) Demand-Capacity Ration (DCR)
(a) Far-fault earth quakes (b) Near-fault earth quakes
Fig. 9.13 Performance assessment curves for Koyna dam subjected to near-fault and far-fault
earthquakes with a PGA level of 0.20 g
Fig. 9.14 Overstressed regions for Koyna dam under near-fault ground motions with a PGA level
of 0.20 g. a Northridge #655, (b) Imperial Valley #0952, c Loma Prieta #57007, and d Kobe # JMA
and 9.19. The shaded areas indicated that the concrete material regions where the
computed DCR values exceeded 1.0.
It can be seen from Figs. 9.15, 9.16a and 9.18 that the dam subjected to far-
fault earthquake records with a PGA level of 0.30 g suffers an acceptable level
damage with no possibility of failure as the estimated DCR < 2, most of the resulting
cumulative duration curves are located within the zone of acceptable performance,
and the overstressed regions associated with DCR > 1 are less than 15% of the total
section area. Therefore, some tensile cracking damage around the neck of the dam
is expected but it will not drastically affect the resulting dynamic behavior.
It is clear from Fig. 9.16 that the maximum principal stress peaks in excess of the
tensile strength of the concrete material vary from 7 to 13 cycles for different near-
fault earthquake records with a PGA level of 0.30 g. Some values of the maximum
principal stress are over than DCR = 2 (3–5 cycles) for all near-fault earthquakes.
The results in Fig. 9.17b show that the stress demand-capacity ratios exceed 2 and
the cumulative inelastic duration is substantially greater than the acceptance level.
It means that the near-fault ground motions with a PGA level of 0.30 g will cause

10 10
8 8
DCR=2 DCR=2
6 6
4 4
DCR=1 DCR=1
2 2
0 0
0 3 6 9 12 15 0 3 6 9 12 15
Time (s) Time (s)
(a) (b)

10 10
(c) Loma Prieta #57217 (d) Kobe #Kakogawa
8 8
DCR=2 DCR=2
6 6
4 4
DCR=1 DCR=1
2 2
0 0
0 3 6 9 12 15 0 3 6 9 12 15
Time (s) Time (s)
(c) (d)
Fig. 9.15 Time histories of maximum principal tensile stresses for far-fault ground motions with
d Kobe # Kakogawa
10 10
8 8
DCR=2 DCR=2
6 6
4 4
DCR=1 DCR=1
2 2
0 0
0 3 6 9 12 15 0 3 6 9 12 15
Time (s) Time (s)
(a) (b)
10 10
(c) Loma Prieta #57007 (d) Kobe #JMA
8 8
DCR=2 DCR=2
6 6
4 4
DCR=1 DCR=1
2 2
0 0
0 3 6 9 12 15 0 3 6 9 12 15
Time (s) Time (s)
(c) (d)
Fig. 9.16 Time histories of maximum principal tensile stresses for near-fault ground motions with
d Kobe # JMA
a considerable damage on the dam body as the performance curve is above the
acceptance curve. Therefore, it can be stated that the linear time history analysis of
the dam is insufficient, and the nonlinear time history analysis is required to estimate
the performance more accurately.

0.4 1.2
Northridge Northridge
Imperail Valley 1.0 Imperail Valley
0.3 Loma Prieta Loma Prieta
Kobe Kobe
0.8
Target Target
0.2 0.6
0.4
0.1
0.2
0.0 0.0
1.0 1.2 1.4 1.6 1.8 2.0 1.0 1.2 1.4 1.6 1.8 2.0
Demand-Capacity Ration (DCR) Demand-Capacity Ration (DCR)
(a) Far-fault earth quakes (b) Near-fault earth quakes
Fig. 9.17 Performance assessment curves for Koyna dam subjected to near-fault and far-fault
earthquakes with a PGA level of 0.30 g
Fig. 9.18 Overstressed regions for Koyna dam under far-fault ground motions with a PGA level
of 0.30 g. a Northridge #24272, b Imperial Valley #0286, c Loma Prieta #57217, and d Kobe #
Kakogawa
Fig. 9.19 Overstressed regions for Koyna dam under near-fault ground motions with a PGA level
of 0.30 g. a Northridge #655, b Imperial Valley #0952, c Loma Prieta #57007, and d Kobe # JMA
Based on these results, it can be concluded that there are more seismic performance
demands for the dam subjected to near-fault ground motions. The nonlinear analysis
is required for near-fault ground motions at the 0.30 g level to further assess the dam
damage.
9.5.3 Seismic Performance Evaluation Using Nonlinear

Dynamic Analysis
According to the proposed performance criteria, the results corresponding to the

0.30 g PGA level clearly indicated significant nonlinear deformation should be
expected for near-fault ground motions. The nonlinear analysis must be carried out for
seismic performance assessment of the dam. To validate this aspect of the proposed
damage criteria, nonlinear dynamic damage analyses of the selected concrete dam
under near-fault ground motions are conducted employing the Concrete Damaged
Plasticity (CDP) model with the strain hardening or softening behavior. In order
to study the influence of near-fault ground motions on the seismic performance of
concrete gravity dams, nonlinear seismic analyses of concrete gravity dams subjected
to far-fault ground motions are also performed.
Seismic damage profiles of the Koyna gravity dam during the real near-fault and
far-fault ground motions with a PGA of 0.3 g are shown in Figs. 9.20 and 9.21,
respectively. The shaded area related with red color indicates those elements that
experienced some level of tensile damage. The figures depict the damage predicted
for real ground motions considered in this study. From the cracking profiles shown
in Figs. 9.20 and 9.21, it can be observed that the failure mechanism is formed of
two main damage zones, one at the base and one in the upper parts of the dam
(downstream face or upstream face). Two damage zones are clearly identified and
they correspond to the areas associated with the maximum tensile demands predicted
Fig. 9.20 Cracking profiles of Koyna dam under near-fault ground motions with a PGA level of
0.30 g. a Northridge #655, b Imperial Valley #0952, c Loma Prieta #57007, and d Kobe # JMA
Fig. 9.21 Cracking profiles of Koyna dam under far-fault ground motions with a PGA level of
0.30 g. a Northridge #24272, b Imperial Valley #0286, c Loma Prieta #57217, and d Kobe #
Kakogawa
by the previous linear analyses. It can be also seen from Figs. 9.20 and 9.21 that as-
recorded near-fault ground motions have significant influence on the seismic damage
of concrete gravity dams. For the far-fault ground motions associated with a PGA
level of 0.30 g, some moderate damage is identified (Fig. 9.21) but it does not seem
to reach a level that could compromise the integrity of the section. On the other hand,
the results corresponding to the input near-fault ground motions associated with a
PGA level of 0.30 g clearly show indications of significant strength degradation in
the dam, with a cracking pattern that extends completely across the upper section.
Comparison of the dam crest horizontal displacement histories from the nonlinear
analyses of the concrete gravity dam subjected to near-fault and far-fault ground
motion indicates that a remarkable upstream deviation appears in the dam response
under some near-fault ground motion cases, and there is a little, if any, residual
deformation of the dam under far-fault ground motions compared with the initial
displacement when imposing the seismic load. The maximum residual displacement
of about 1.98 cm occurs under the near-fault ground motion recorded at Corralitos,
CA #57007 station in the Loma Prieta earthquake, with respect to the equivalent
deformation caused by the static loads.
9.6 Nonlinear Dynamic Response of Concrete Gravity

Dams Subjected to Near-Fault and Far-Fault Ground
Motions
Due to the unique characteristics of the near-fault ground motions and their potential
to cause severe damage to structures, the interest on the near-fault effect on structural
response has increased. Earthquake engineers have been considering methods to
incorporate near-fault effects in engineering design (Mavroeidis and Papageorgiou
2003). In this thesis, a set of selected near-fault and far-fault ground motion records (as
shown Table 9.1) is used to examine the near-fault ground motion effects. Nonlinear
seismic analyses of dam-reservoir-foundation systems subjected to both near-fault
and far-fault ground motions are performed. The influence of near-fault and far-fault
ground motions on the dynamic response and seismic damage of concrete gravity
dams is studied and discussed. Dynamic damage analysis of the selected concrete
dam is conducted employing the Concrete Damaged Plasticity (CDP) model with
the strain hardening or softening behavior. The integration time step used in the
nonlinear analysis is 0.005 s.
9.6.1 Nonlinear Displacement Response
Two sets of analyses, namely the near-fault case and far-fault case, are conducted
to study the influence of near-fault and far-fault ground motions on the seismic
response of dam-reservoir-foundation systems. Each set contains 10 as-recorded
ground motion records. Figure 9.22 shows the time histories of horizontal displace-
ments at the dam crest obtained from nonlinear analyses for near-fault and far-
fault ground motions, with positive displacement in the downstream direction. It is
clear from Fig. 9.22 that the non-linear response obtained from near-fault ground
motions has a substantially different displacement history than those obtained from
far-fault ground motions. The crest horizontal displacement values for near-fault
ground motions are greater than those for far-fault ground motions although the
peak ground acceleration of near-fault and far-fault records are the same.
Comparison of the dam crest horizontal displacement histories from the nonlinear
analyses of the dam-reservoir-foundation system subjected to near-fault and far-
fault ground motion indicates that a remarkable upstream deviation appears in the
dam response under some near-fault ground motion cases, and there is a little, if
any, residual deformation of the dam under far-fault ground motions compared with
the initial displacement when imposing the seismic load. The maximum residual
displacement of about 3.75 cm occurs under the near-fault ground motion recorded
at Taichung, Taiwan #TCU076 station in Chi-Chi earthquake, with respect to the
equivalent deformation caused by the static loads. It is also seen from Fig. 9.22 that
plastic deformations in the dam subjected to near-fault ground motions are larger
than those subjected to far-fault ground motions.
9.6.2 Seismic Damage
Seismic damage profiles of the Koyna gravity dam during the real near-fault and
far-fault ground motions (Northridge earthquake, Imperial Valley earthquake, Loma
Prieta earthquake, Chi-Chi earthquake) with a PGA of 0.3 g including dam-reservoir-
foundation interaction are shown in Figs. 9.23, 9.24, 9.25 and 9.26, respectively.
9.6 Nonlinear Dynamic Response of Concrete Gravity Dams Subjected … 233
Displacement (cm) 6 6 6
Displacement (cm)
Displacement (cm)
No.1ñNear-fault No.3ñNear-fault No. 5ñNear-fault
No.2ñFar-fault No.4ñFar-fault No. 6ñFar-fault
3 3 3
0 0 0
-3 -3 -3
-6 -6 -6
0 3 6 9 12 15 0 3 6 9 12 15 0 3 6 9 12 15
Time (s) Time (s) Time (s)
(a)
6 6
Displacement (cm)
Displacement (cm)
No. 7ñNear-fault No. 9ñNear-fault
No. 8ñFar-fault No. 10ñFar-fault
3 3
0 0
-3 -3
-6 -6
0 3 6 9 12 15 0 3 6 9 12 15
Time (s) Time (s)
(b)
6 6
Displacement (cm)
No. 11ñNear-fault Displacement (cm) No. 13ñNear-fault

No. 12ñFar-fault No. 14ñFar-fault
3 3
0 0
-3 -3
-6 -6
0 3 6 9 12 15 0 3 6 9 12 15
Time (s) Time (s)
(c)
6 6
Displacement (cm)
6
Displacement (cm)
Displacement (cm)
No. 15ñNear-fault No. 17ñNear-fault No. 19ñNear-fault

No. 16ñFar-fault No. 18ñFar-fault No. 20ñFar-fault
3 3 3
0
0 0
-3
-3 -3
-6
-6 -6
0 8 16 24 32 40 0 8 16 24 32 40 0 8 16 24 32 40
Time (s) Time (s) Time (s)
(d)
Fig. 9.22 Time histories of horizontal displacements at the crest of the dam subjected to near-fault
and far-fault ground motions from the a Northridge, b Imperial Valley, c Loma Prieta, and d Chi-Chi
earthquakes. For ground motion information see Table 9.1
The shaded area related with red color indicates those elements that experienced
some level of tensile damage over the duration of the analysis. The figures depict
the damage predicted for real ground motions considered in this study. From the
cracking profiles shown in Figs. 9.23, 9.24, 9.25 and 9.26, it can be observed that
as-recorded near-fault ground motions have significant influence on the seismic
damage of concrete gravity dams. In some cases, the results corresponding to the
input motions with near-fault ground motions clearly show indications of significant
strength degradation in the dam, with a cracking pattern that extends completely
across the upper section.
It can also be seen from Figs. 9.23, 9.24, 9.25 and 9.26 that the failure mechanism
is formed by two main damage zones, one at the base and one in the upper parts of
the dam. In almost all the analyses, the cracking is always first initiated at the dam
heel, and then progresses a long way from the upstream face to the downstream face.
No. 1 No. 3 No. 5
(a)
No. 2 No. 4 No. 6
(b)
Fig. 9.23 Cracking profiles of Koyna dam under the Northridge earthquake: a near-fault ground
motions b far-fault ground motions
No. 7 No. 9 No. 8 No. 10
(a) (b)
Fig. 9.24 Cracking profiles of Koyna dam under the Imperial Valley earthquake: a near-fault ground
No. 11 No. 13 No. 12 No. 14
(a) (b)
Fig. 9.25 Cracking profiles for Koyna dam under the Loma Prieta earthquake: a near-fault ground
Fig. 9.26 Cracking profiles of Koyna dam under the Chi-Chi earthquake: a near-fault ground
Cracking profiles in the upper part of the dam are always initiated at the point of slope
discontinuity on the downstream face. Top cracking profiles are almost either nearly
horizontal or sloping downward from the downstream faces towards the upstream
faces. But in some analyses, cracks are predicted to initialize near the middle of the
upstream face or the downstream face, and extend into the dam. Because of the thrust
of impounded water which is opposing the tendency of the top section to slide along
the crack in the upstream direction, the computed crack profiles in the upper part of
the dam can be considered neutral or favourable conditions to maintain stability.
9.6.3 Damage Energy Dissipation
Figure 9.27 shows the energy dissipation curves of the dam under near-fault and far-
fault ground motions. The damage energy dissipation capacity of the dam increases
when subjected to near-fault ground motions. The cumulative damage energy dissipa-
tion of the dam for near-fault ground motions is 4.68 times that for far-fault ground
motions, on average. This demonstrates that near-fault ground motion will cause
Damgage Energy Dissipation (KN.m)
60
50 No. 1
50 No. 7
No. 2 No. 8
40 No. 3 No. 9
No. 4 40 No. 10
No. 5
30 No. 6
30
20 20
10 10
0 0
0 3 6 9 12 15 0 3 6 9 12 15
Time (s) Time (s)
(a) (b)
40 100 No. 15
No. 16
No. 11
No. 17
No. 12 80 No. 18
30 No. 13
No. 19
No. 14
No. 20
60
20
40
10
20
0 0
0 3 6 9 12 15 0 10 20 30 40
Time (s) Time (s)
(c) (d)
Fig. 9.27 Damage energy dissipation curves of the dam for near-fault and far-fault ground motions.
a Northridge earthquakes, b Imperial Valley earthquakes, c Loma Prieta earthquakes, and d Chi-Chi
earthquakes
more severe damage to the dam body than far-fault ground motion. As can be seen
in Fig. 9.27, the amount of the damage can also be assessed by the damage energy
dissipation.
9.6.4 Identifying the Influence of Near-Fault Ground

Motions on Seismic Damage of Dams
The absolute maximum horizontal displacements obtained from the nonlinear anal-
yses for both near-fault and far-fault ground motions are given in Fig. 9.28. It can
be seen from Fig. 9.28 that the values of maximum horizontal displacements for the
dam under near-fault ground motions are generally greater than those subjected to
far-fault ground motions.
In order to analyze the effects of near-fault ground motion on seismic damage
of concrete gravity dams, Figs. 9.29, 9.30 and 9.31 are generated by plotting the
accumulated damage of the dam imparted by the 20 records for a given level of
intensity in terms of the local and global damage indices. Trend lines are displayed
with the aim of identifying an average value for each case.
From Figs. 9.29, 9.30 and 9.31, it can be observed that studies employing damage
measures using local and global damage indices show that near-fault ground motion
may have a significant influence on the accumulated damage. The global index of the
dam for near-fault ground motions is 1.89 times that for far-fault ground motions,
on average. It can also be seen from Figs. 9.29-9.31 that the upper zone of concrete
gravity dams is more vulnerable to near-fault earthquakes in comparison to a corre-
sponding far-fault ground motion. Damage measures such as the local damage index
Near-fault 19
7 Far-fault
5
6
Displacement (cm)
3 Average of 5.06cm
15
5
11 17
1 7 13
1.74cm 9
4 8
2 18 20
3 6 16
10 12 14
4
Average of 3.32cm
2
0 2 4 6 8 10 12 14 16 18 20
No.
Fig. 9.28 Absolute maximum horizontal displacements at crest of the dam under near-fault and
far-fault ground motions. For ground motion information see Table 9.1
19
1.0 Average of 0.84
9
1 13
11
Local damage index
0.8 15
3 7 17
5
Near-fault
0.6 0.44
6 Far-fault
2 8 16
12 14
0.4
4 18 20
Average of 0.40 10
0.2
0 2 4 6 8 10 12 14 16 18 20
No.
Fig. 9.29 Influence of near-fault and far-fault ground motions on the local damage index measure
for the upper part of the dam (near the changes in the slope of the downstream face). For ground
motion information see Table 9.1
0.25 5
Average of 0.18 15
7
Local damage index
19
0.20 1 3 9 12
0.15 0.07 16 17 20
8 13
14
0.10 18
6 11
0.05 2 4 Average of 0.11
10 Near-fault
Far-fault
0.00
0 2 4 6 8 10 12 14 16 18 20
No.
Fig. 9.30 Influence of near-fault and far-fault ground motions on the local damage index measure
for the dam heel. For ground motion information see Table 9.1
for the upper zone of the dam are consistently greater for near-fault ground motions.
While, the accumulated damage on the dam heel is not very sensitive to near-fault
ground motions. The local damage index under near-fault ground motion for the
upper part of the dam is 0.84, and for the dam heel is 0.18, on average. The accumu-
lated damage for the upper zone of the dam is more sensitive to near-fault ground
motion, which gives more importance to the dissipated energy during the hysteretic
behavior of the structure since the high seismic response zone is mainly located in
the upper zone of the dam.
9.7 Conclusions 239
0.4
1 5 15 19
7 9
Global damage index
3 17
0.3 13
Average of 0.34 0.16 11
12 16
8
0.2 14
2 6 20
18
4 Average of 0.18
0.1 10 Near-fault
Far-fault
0 2 4 6 8 10 12 14 16 18 20
No.
Fig. 9.31 Influence of near-fault and far-fault ground motions on the global damage index measure
for the dam. For ground motion information see Table 9.1
9.7 Conclusions
The purpose of this study was to analytically evaluate the near-fault and far-fault
ground motion effects on nonlinear dynamic response and seismic damage of
concrete gravity dams with considering the effects of dam-reservoir-foundation inter-
action. The reservoir water is modeled using two-dimensional fluid finite elements
by the Lagrangian approach. The Koyna gravity dam is chosen for analysis. 20 as-
recorded near-fault and far-fault strong ground motion records considered in this
study are used as seismic excitations. Nonlinear seismic analyses of concrete gravity
dams under earthquake conditions are performed according to the Concrete Damaged
Plasticity (CDP) model including the strain hardening or softening behavior.
It is concluded from the study that the nonlinear response obtained from near-
fault ground motions has a substantially different displacement history than those
obtained from far-fault ground motions. The crest horizontal displacement values
for near-fault ground motions are greater than those for far-fault ground motions.
A comparison of the dam crest horizontal displacement history from the nonlinear
analyses of the dam-reservoir-foundation system subjected to near-fault and far-fault
ground motion indicates that a remarkable upstream deviation appears in the dam
response under some near-fault ground motion cases.
The performed seismic damage analyses show that accumulated damage of dams
under consideration is found to be significantly affected by near-fault ground motion.
The upper zone of concrete gravity dams is more vulnerable to near-fault earthquakes
in comparison to a corresponding far-fault ground motion. In some cases, the results
corresponding to the input motions with near-fault ground motions clearly show
indications of significant strength degradation in the dam, with a cracking pattern
that extends completely across the upper section.
According to this study, near-fault earthquake ground motions have remarkable

effects on the nonlinear dynamic response and accumulated damage of concrete
gravity dams due to their impulsive effects on structures. Near-fault ground motions
have the potential to cause more severe damage to the dam body than far-fault ground
motions. The effects of near-fault ground motions on concrete gravity dams should
be taken into account to obtain more realistic results.
References
Abrahamson, N. A., & Somerville, P. G. (1996). Effects of the hanging wall and footwall on ground
motions recorded during the Northridge earthquake. Bulletin of the Seismological Society of
America, 86S(1B), S93–S99.
Akköse, M., & Şimşek, E. (2010). Non-linear seismic response of concrete gravity dams to near-fault
ground motions including dam-water-sediment-foundation interaction. Applied Mathematical
Modelling, 34(11), 3685–3700.
Ardakani, S. M. S., & Saiidi, M. S. (2018). Simple method to estimate residual displacement
in concrete bridge columns under near-fault earthquake motions. Engineering Structures, 176,
208–219.
Bayraktar, A., AltuniAik, A. C., Sevim, B. A., Kartal, M. E., Turker, T., & Bilici, Y. (2009). Compar-
ison of near-and far-fault ground motion effect on the nonlinear response of dam-reservoir-
foundation systems. Nonlinear Dynamics, 58(4), 655.
Bayraktar, A., Türker, T., Akk Se, M., & Ate, E. (2010). The effect of reservoir length on seismic
performance of gravity dams to near-and far-fault ground motions. Natural Hazards, 52(2), 257–
275.
Behesthi Aval, S. B., Mohsenian, V., & Sadegh Kouhestani, H. (2018). Seismic performance-based
assessment of tunnel form building subjected to near- and far-fault ground motions. Asian Journal
of Civil Engineering, 19(1), 79–92.
Beiraghi, H., Kheyroddin, A., & Kafi, M. A. (2016a). Forward directivity near-fault and far-fault
ground motion effects on the behavior of reinforced concrete wall tall buildings with one and
more plastic hinges. Structural Design of Tall and Special Buildings, 25(11), 519–539.
Beiraghi, H., Kheyroddin, A., & Kafi, M. A. (2016b). Energy dissipation of tall core-wall structures
with multi-plastic hinges subjected to forward directivity near-fault and far-fault earthquakes.
Structural Design of Tall and Special Buildings, 25(15), 801–820.
Bhagat, S., Wijeyewickrema, A. C., & Subedi, N. (2018). Influence of near-fault ground motions
with fling-step and forward-directivity characteristics on seismic response of base-isolated
buildings. Journal of Earthquake Engineering 1–20.
Bilgin, H., & Hysenlliu, M. (2020). Comparison of near and far-fault ground motion effects on low
and mid-rise masonry buildings. Journal of Building Engineering, 30, 101248.
Bray, J. D., & Rodriguez-Marek, A. (2004). Characterization of forward-directivity ground motions
in the near-fault region. Soil Dynamics and Earthquake Engineering, 24(11), 815–828.
Cao, Y., Meza-Fajardo, K. C., Mavroeidis, G. P., & Papageorgiou, A. S. (2016). Effects of wave
passage on torsional response of symmetric buildings subjected to near-fault pulse-like ground
motions. Soil Dynamics and Earthquake Engineering, 88, 109–123.
Chen, X., Liu, Y., Zhou, B., & Yang, D. (2020). Seismic response analysis of intake tower structure
under near-fault ground motions with forward-directivity and fling-step effects. Soil Dynamics
and Earthquake Engineering, 132, 106098.
Chopra, A. K., & Chintanapakdee, C. (2001). Comparing response of SDF systems to near-fault
and far-fault earthquake motions in the context of spectral regions. Earthquake Engineering and
Structural Dynamics, 30(12), 1769–1789.
References 241
COSMOS. 2019. Strong Motion Virtual Data Center. http://strongmotioncenter.org/vdc/scripts/def

ault.plx.
Du, K., Cheng, F., Bai, J., & Jin, S. (2020). Seismic performance quantification of buckling-
restrained braced RC frame structures under near-fault ground motions. Engineering Structures
211.
Eslami, A., Ronagh, H. R., & Mostofinejad, D. (2016). Analytical assessment of CFRP retrofitted
reinforced-concrete buildings subjected to near-fault ground motions. Journal of Performance of
Constructed Facilities, 30(6), 4016044.
Ghanaat, Y. (2002). Seismic performance and damage criteria for concrete dams. In Proceedings of
the 3rd US-Japan Workshop on Advanced Research on Earthquake Engineering for Dams, San
Diego.
Ghanaat, Y. (2004). Failure modes approach to safety evaluation of dams. In 13th World Conference
on Earthquake Engineering, Vancouver.
Gorai, S., & Maity, D. (2019). Seismic response of concrete gravity dams under near field and far
field ground motions. Engineering Structures, 196, 109292.
Hedayati Dezfuli, F., Li, S., Alam, M. S., & Wang, J. (2017). Effect of constitutive models on
the seismic response of an SMA-LRB isolated highway bridge. Engineering Structures, 148,
113–125.
Konstandakopoulou, F. D., Evangelinos, K. I., Nikolaou, I. E., Papagiannopoulos, G. A., & Pnev-
matikos, N. G. (2019). Seismic analysis of offshore platforms subjected to pulse-type ground
motions compatible with European Standards. Soil Dynamics and Earthquake Engineering
105713.
Li, S., Zhang, F., Wang, J., Alam, M. S., & Zhang, J. (2017). Seismic responses of super-span cable-
stayed bridges induced by ground motions in different sites relative to fault rupture considering
soil-structure interaction. Soil Dynamics and Earthquake Engineering, 101, 295–310.
Liao, W., Loh, C., & Lee, B. (2004). Comparison of dynamic response of isolated and non-isolated
continuous girder bridges subjected to near-fault ground motions. Engineering Structures, 26(14),
2173–2183.
Ma, H., Zhuo, W., Lavorato, D., Nuti, C., Fiorentino, G., Gu, Y., et al. (2019). Probabilistic seismic
response analysis on continuous bridges under near-fault ground motions. Iranian Journal of
Science and Technology, Transactions of Civil Engineering, 43(3), 491–500.
Manafpour, A. R., & Kamrani Moghaddam, P. (2019). Performance capacity of damaged RC
SDOF systems under multiple far- and near-field earthquakes. Soil Dynamics and Earthquake
Mavroeidis, G. P., & Papageorgiou, A. S. (2003). A mathematical representation of near-fault ground
motions. Bulletin of the Seismological Society of America, 93(3), 1099–1131.
PEER. (2019). Pacific Earthquake Engineering Research Center. http://peer.berkeley.edu/peer_g
Raphael, J. M. (1984). Tensile strength of concrete. ACI Journal Proceedings 158–165.
Sevim, B. (2011). The effect of material properties on the seismic performance of Arch Dams.
Natural Hazards and Earth System Science, 11(8), 2253–2261.
Somerville, P. (2005). Engineering characterization of near fault ground motions. In New Zealand
Society for Earthquake Engineering Conference.
Somerville, P. G. (2003). Magnitude scaling of the near fault rupture directivity pulse. Physics of
the Earth and Planetary Interiors, 137(1–4), 201–212.
Somerville, P. G., Smith, N. F., Graves, R. W., & Abrahamson, N. A. (1997). Modification of
empirical strong ground motion attenuation relations to include the amplitude and duration effects
of rupture directivity. Seismological Research Letters, 68(1), 199–222.
Stewart, J. P., Chiou, S., Bray, J. D., Graves, R. W., Somerville, P. G., & Abrahamson, N. A.
(2002). Ground motion evaluation procedures for performance-based design. Soil Dynamics and
Sun, B., Zhang, S., Deng, M., & Wang, C. (2020). Inelastic dynamic response and fragility analysis of
arched hydraulic tunnels under as-recorded far-fault and near-fault ground motions. Soil Dynamics
and Earthquake Engineering, 132, 106070.
Tajammolian, H., Khoshnoudian, F., & Mehr, N. P. (2016). Seismic responses of isolated structures
with mass asymmetry mounted on TCFP subjected to near-fault ground motions. International
Journal of Civil Engineering, 14(8A), 573–584.
USACE. (2003). Time history dynamic analysis of concrete hydraulic structures. Engineer Manual
Em 1110-2-6051.
Wang, G. Q., Zhou, X. Y., Zhang, P. Z., & Igel, H. (2002). Characteristics of amplitude and duration
for near fault strong ground motion from the 1999 Chi-Chi, Taiwan Earthquake. Soil Dynamics
Xiang, N., & Alam, M. S. (2019). Displacement-based seismic design of bridge bents retrofitted
with various bracing devices and their seismic fragility assessment under near-fault and far-field
ground motions. Soil Dynamics and Earthquake Engineering, 119, 75–90.
Xin, L., Li, X., Zhang, Z., & Zhao, L. (2019). Seismic behavior of long-span concrete-filled steel
tubular arch bridge subjected to near-fault fling-step motions. Engineering Structures, 180, 148–
159.
Yamaguchi, Y., Hall, R., Sasaki, T., Matheu, E., Kanenawa, K., Chudgar, A., & Yule, D.
(2004). Seismic performance evaluation of concrete gravity dams. In 13th World Conference
on Earthquake Engineering, Vancouver.
Yang, D., Guo, G., Liu, Y., & Zhang, J. (2019). Dimensional response analysis of bilinear
SDOF systems under near-fault ground motions with intrinsic length scale. Soil Dynamics and
Yang, S., Mavroeidis, G. P., Ucak, A., & Tsopelas, P. (2017). Effect of ground motion filtering on the
dynamic response of a seismically isolated bridge with and without fault crossing considerations.
Yazdani, Y., & Alembagheri, M. (2017a). Seismic vulnerability of gravity dams in near-fault areas.
Yazdani, Y., & Alembagheri, M. (2017b). Nonlinear seismic response of a gravity dam under near-
621–632.
Zou, D., Han, H., Ling, H. I., Zhou, Y., & Liu, J. (2019). An approach for the real-time slip
deformation coupled with strain softening of a high rockfill dam subjected to pulse-like ground
motions. Soil Dynamics and Earthquake Engineering, 117, 30–46.
Chapter 10
Deterministic 3D Seismic Damage
Analysis of Guandi Concrete Gravity
Dam
10.1 Introduction
During the recent decades, the growing demand for green energy has boosted the
construction of high concrete dams, many of which are located in seismically active
regions. For instance, tens of concrete dams over 150 m in height are either being
built or planned in Southwest China, with the design peak ground acceleration value
up to 0.56 g. The seismic safety of concrete dams is one of the most important issues
in the engineering field since earthquake activities may impair the proper functioning
of these critical infrastructures and trigger catastrophic failure that results in heavy
loss of human life and property.
One of the major reasons for seismic damage and failure of such structures is the
initiation, growth and coalescence of cracks in concrete (Mridha and Maity 2014).
Due to the growing concern on the seismic safety of concrete dams, nonlinear numer-
ical models, which can predict crack initiation and propagation through the dam body,
have attracted great research interest in the community of dam engineering and been
widely used for evaluating the seismic performance of concrete dams. These models
can be grouped into two families: continuum crack approaches with no intention to
capture individual cracks such as smeared crack approach (Bhattacharjee and Léger
1994; Léger and Leclerc 1996) and plastic-damage model (Lee and Fenves 1998a,
b; Omidi et al. 2013), and discrete crack approaches aimed at resolving individual
cracks, including the fracture mechanics approach (Ayari and Saouma 1990) and
extended finite element method (XFEM) (Zhang et al. 2013a, b; Wang et al. 2015a,
b). While discrete crack approaches provide an explicit representation of discon-
tinuous displacement fields, they need to keep track of the evolution of individual
cracks, therefore imposing significant challenges to finite element (FE) implemen-
tations, especially in three dimensional cases. On the other hand, the continuum
crack models treat cracks as the ultimate consequence of damage accumulation in
https://doi.org/10.1007/978-981-15-6194-8_10
244 10 Deterministic 3D Seismic Damage Analysis …
the constitutive relations with strain softening. While discrete crack models can be
regarded as more accurate modeling techniques than continuum methods, continuum
crack approaches are computationally very attractive and better suited for complex
engineering problems (Wang and Waisman 2016).
In order to meet the deformation requirements of the foundation and to minimize
the cracking caused by shrinkage, concrete gravity dams are constructed as indi-
vidual monoliths that are separated by contraction joints. Owing to the presence of
contraction joints, most of the analyses that deal with the nonlinear dynamic response
and seismic safety evaluation of concrete gravity dams are through either two dimen-
sional (2D) (Huang and Zerva 2013; Omidi et al. 2013; Alembagheri and Ghaemian
2013a; Alembagheri and Seyedkazemi 2015) or quasi three dimensional (3D) (Pan
et al. 2011; Mridha and Maity 2014) models, with the underlying assumption that
the monoliths behave independently during earthquake activities. The tallest mono-
lith of concrete gravity dams, usually located on the valley bottom, is often selected
for dynamic analysis since this monolith is the most critical part of dam structures.
Based on 2D or quasi-3D simulations, many factors which may influence the dynamic
behavior of concrete gravity dams have been studied, including the input excitation
mechanism, strong motion duration, near-fault ground motions, aftershocks, reser-
voir length, to name just a few(Zhang and Wang 2013; Wang et al. 2015a, b, 2016,
2017).
However, three dimensional analyses are required in order to obtain a more real-
istic seismic response of concrete gravity dams. The reasons are threefold. First,
monoliths of a concrete gravity dam have different heights and cross-sections. Such
difference leads to distinct periods of vibration and damage characteristics during
seismic events. For instance, the Koyna earthquake caused varying degrees of damage
to different monoliths of the Koyna dam (Chopra and Chakrabarti 1973), including
horizontal cracks on the downstream face of monoliths 13–18 and 25–30 and on
the upstream face of monoliths 10–18 and 24–30. Second, the contraction joints
are expected to open/close and slide cyclically when subjected to severe earthquake
ground motions. The nonlinear interaction between concrete monoliths would result
in the redistribution of stresses in dam bodies and thus may influence the pattern
and degree of the damage-cracking of concrete gravity dams. When high concrete
gravity dams are built in narrow valleys, vast height variations exist between adja-
cent monoliths and dam foundations are usually tilted in the cross-stream direc-
tion. Under cross-stream seismic excitations, these factors would further increase
the nonlinear response of contraction joints of dams, which may in turn affect the
overall seismic performance of concrete gravity dams significantly. Last, gravity
dam-foundation systems generally undergo significant dynamic forces (from reser-
voirs) when subjected to earthquakes, which may lead to crack initiation and prop-
agation in dam bodies. In addition, the dynamic interaction between the dam and
foundation can also affect the system response through the reflection and radiation
of seismic waves. However, the significance of the dam-foundation-reservoir inter-
action can only be partially considered by the widely used 2D or quasi-3D models,
in which the cross-stream interaction is completely neglected.
Up to now, 3D finite element analyses accounting for the effects of contraction

joints (Feng et al. 2011; Hariri-Ardebili et al. 2013; Omidi and Lotfi 2013; Wang et al.
2013; Alembagheri and Ghaemian 2013b, 2015; Hariri-Ardebili and Kianoush 2014)
are mainly carried out in reference to arch dams and there are only a few studies on the
3D dynamic response of concrete gravity dams. Azmi and Paultre (2002) studied the
influence of contraction joints opening-closing as well as shear sliding on the seismic
behavior of Outardes 3 gravity dam comprised of 19 monoliths. Wang et al. (2012a,
b) employed a 3D model to analyze the earthquake response of a concrete gravity dam
based on the linear material model of mass concrete. In their study, dynamic contact
between monoliths was taken into account. The results showed that increasing the
connection between the monoliths can enhance the aseismic capacities of the dam.
In order to investigate the nonlinear seismic response of a typical gravity dam as
accurate as possible, Hariri-Ardebili et al. (2016) modeled one of the non-overflow
monoliths in 3D geometry with three times of the actual thickness of the monolith
and only the results at the middle section were extracted. Arici et al. (2014) and
Yilmazturk et al. (2015) conducted seismic response analyses of a roller compacted
concrete (RCC) gravity dam including 3D dam-reservoir-foundation interaction. By
comparing the results with those from 2D analyses, they found that the 3D behavior of
the dam is substantially different from that obtained from the idealized 2D analyses.
They emphasized the importance and necessity of full 3D analyses for gravity dams,
especially when they are built in relatively narrow canyons.
In this contribution, a systematic study has been conducted on the 3D seismic
damage-cracking behavior of concrete gravity dams with all the following factors
considered: contraction joint nonlinearity, cross-stream earthquake excitation, and
dam-foundation-reservoir interaction. The effects of the cross-stream seismic excita-
tion and contraction joints on the cracking characteristics of a practical gravity dam
have been investigated. To this end, Guandi concrete gravity dam, built in the highly
seismic zone of China, is considered. The Concrete Damaged Plasticity (CDP) model
(Lubliner et al. 1989; Lee and Fenves 1998a, b) including strain hardening/softening
behavior is used to capture crack initiation and propagation in the dam body. The joint
nonlinearity associated with opening and slipping of contraction joints is modeled
using two different surface-to-surface contact models (soft and hard formulations)
so that the effects of joint infills can be quantified. The Lagrangian approach is used
for the modeling of the dam-reservoir-foundation interaction. The proposed model is
first validated against the experimental cracking results of the Koyna concrete gravity
dam. Then a comprehensive investigation of the seismic cracking characteristics of
the Guandi concrete gravity dam-foundation-reservoir system is carried out based on
3D nonlinear dynamic analyses. Several case studies are conducted to investigate the
influence of the cross-stream seismic excitation and contraction joints on nonlinear
dynamic responses of the Guandi gravity dam.
10.2 Material Properties of Mass Concrete

and Contraction Joint Nonlinearity
Two sources of nonlinearity are involved in the dynamic behavior of concrete gravity
dams: material nonlinearity due to concrete cracking and contact nonlinearity owing
to the opening/closure and sliding of contraction joints. Numerical models for both
sources of nonlinearity are detailed in this section. FE analyses presented in this work
are performed using the general-purpose commercial software ABAQUS/Standard
V 6.11(Version 2011).
10.2.1 Material Properties of Mass Concrete
In general, both the degradation of stiffness (damage) due to micro-cracking and

unrecoverable deformation (plasticity) are exhibited by concrete materials. There-
fore, a proper constitutive model for concrete should address the two physically
distinct behaviors. The Concrete Damaged Plasticity (CDP) model, proposed by
Lubliner et al. (1989) and modified by Lee and Fenves (1998a, b), offers a particularly
efficient context where damage evolution and plastic deformation can be modeled
simultaneously. In this model, uniaxial strength functions are factorized into two
parts to represent the permanent (plastic) deformation and degradation of stiffness
(damage). Two main failure mechanisms of the concrete material are assumed in the
CDP model: one is tensile cracking and the other is compressive crushing. Accord-
ingly, different strength and post-peak behaviors of concrete can be taken into account
for tension and compression. This approach, which can efficiently describe complex
mechanical behavior of concrete under cyclic loadings, is selected in this study for
the seismic cracking analysis of concrete dams. The interested reader is referred to
(Zhang et al. 2013a, b) for more details concerning the constitutive model, including
the damage evolution, the yield criterion, and the flow rule.
10.2.2 Contraction Joint Nonlinearity
Concrete gravity dams are constructed as individual monoliths which are separated
by contraction joints. The contraction joints are expected to open/close and slide
cyclically during severe earthquake ground motions. The presence of these weak
planes between monoliths may significantly affect the dynamic behavior of concrete
gravity dam-reservoir-foundation systems.
In order to model the joint behavior, a surface-to-surface contact model is used
in the present study. The interaction between contacting surfaces consists of two
components: normal interaction due to loads perpendicular to surfaces in contact, and
tangential one, referring to possible relative sliding between two adjacent surfaces.
10.2 Material Properties of Mass Concrete and Contraction Joint Nonlinearity 247
Shear keys are not considered in contraction joints. To quantify the effect of joint
infill, two different contact formulations are adopted in this study to describe the
normal behavior, i.e. hard and soft contact models as shown in Fig. 10.1. We define the
clearance as the normal separation between two surfaces. In hard contact formulation,
the contact constraint is applied when the clearance becomes zero. There is no limit
on the magnitude of contact pressure that can be transferred by the contact surfaces.
The surfaces start to separate when the contact pressure becomes zero or negative,
in which case the contact constraint is removed. Thus, a dramatic change in contact
pressure may occur when the contact condition changes from “open” (a positive
clearance) to “close” (zero clearance). On the other hand, the normal stress is an
exponential function of the clearance in the soft contact formulation, which is suited
to describe the contact behavior of soft, thin layer of grouting materials between the
contraction joint surfaces. The adopted exponential function is given as follows:

p=
0, c ≥ c0 (10.1)
P0
exp(1)−1
1− c
c0
exp 1 − c
c0
− 1 , c < c0
where c0 is the initial contact distance, p is the normal contact pressure, P0 is the
pressure value at zero opening. In this model, the surfaces begin to transmit normal
pressure once the distance between them, measured in the normal direction, reduces
to the initial contact distance c0 . For each joint, the initial contact distance c0 = 1 mm
and the pressure value P0 = 50 MPa at zero opening are selected (Chen et al. 2012;
Wang et al. 2012a, b).
In addition to determining whether contact occurs at a particular point, the relative
sliding of the contact surfaces is also calculated. When surfaces are in contact, they
usually transfer not only normal pressure but also shear forces. Thus, the shear forces
that resist the relative sliding of the surfaces should be taken into consideration. In
this study, the Coulomb friction model is employed to characterize the frictional
Fig. 10.1 Contact pressure-clearance relationships for hard and soft contact models
behavior between the surfaces. The tangential motion is zero until the shear traction
reaches a critical value, which depends on the normal contact pressure as follows
τcrit = μp (10.2)
where μ is the coefficient of friction and p is the contact pressure between the two
surfaces. This equation gives the limiting frictional shear stress for the contacting
surfaces.
10.3 3D Lagrangian Finite Element Formulation
The adopted finite element formulation (Wilson and Khalvati 1983; Calayir and
Dumanoǧlu 1993) for a three dimensional fluid-solid coupled system is based on
the Lagrangian approach. In this approach, displacements are selected to be the
unknown variables for both the fluid and structure domains. Fluid is assumed to be
linearly elastic, inviscid, and irrotational. The stress-strain relationships for a general
three-dimensional fluid element read:
⎧ ⎫ ⎧ ⎫⎧ ⎫
⎪
⎪ P ⎪ ⎪ C11 0 0 0 ⎪ ⎪ εv ⎪
⎨ ⎪ ⎬ ⎪ ⎨ ⎬⎪
⎪ ⎨ ⎪
⎬
Px 0 C22 0 0 Wx
= (10.3)
⎪
⎪ P ⎪ ⎪ 0 0 C33 0 ⎪ ⎪ Wy ⎪
⎩ y⎪ ⎭ ⎪ ⎩ ⎭⎪
⎪ ⎩ ⎪
⎭
Pz 0 0 0 C44 Wz
where P is the pressure, C 11 is the bulk modulus of fluid, εv is the volumetric strain;
W x , W y and W z are the rotations with respect to the Cartesian axis x, y, z, respec-
tively, Px , Py and Pz are the corresponding rotational stresses, C 22 , C 33 and C 44 are
the constraint parameters. Note that the rotation constraint parameters in the above
stress–strain relationships are introduced to enforce the irrotationality of fluid by
means of a penalty method. It should be as high as necessary to prevent fluid rota-
tion but small enough to avoid causing numerical ill-conditioning in the assembled
stiffness matrix.
In the analysis, the effect of the small amplitude free surface waves, which is
commonly referred to as the sloshing effect, is taken into account. The sloshing
effect causes a pressure at the free surface of fluid, which is given by
P = −γw u fn (10.4)
where γ w is the weight density of fluid and ufn is the normal component of the
free surface displacement. The free surface stiffness for fluid is obtained from the
discrete form of Eq. (10.4). Detailed formulation on the Lagrangian finite element
formulation for the simulation of the dam-reservoir-foundation system can be found
in Ref. Calayir and Dumanoǧlu (1993).
10.4 Validation Test for 3D Model 249
10.4 Validation Test for 3D Model
In order to validate the effectiveness of the numerical framework, a quasi 3D model of

a single monolith is established to analyze the cracking behavior of Koyna concrete
gravity dam under the 1967 Koyna earthquake. Three components of the Koyna
earthquake are shown in Fig. 10.2. Only the stream and vertical components of the
seismic ground motions are considered in the validation test, which is consistent
with the seismic input for the model test reported in (Mridha and Maity 2014). The
geometric characteristics and FE mesh of the quasi 3D model are shown in Fig. 10.3.
The material parameters for the Koyna dam-reservoir-foundation system are listed
in Table 2.2 (Chap. 2). The tensile and compressive strength of concrete are 2.9
and 24.1 MPa, respectively. The fracture energy is 250 N/m. In order to avoid wave
propagation effects when applying free field earthquake records at the foundation
base, the foundation rock is assumed to be massless. The reservoir water is assumed
to be linearly elastic, inviscid, and irrotational. The rotation constraint parameter of
the fluid is assumed to be 1000 times of its bulk modulus (Akkose et al. 2008). A
dynamic magnification factor of 1.2 is considered for the tensile strength to account
for strain rate effects. The energy dissipation of the dam-reservoir-foundation system
is considered by the Rayleigh damping method with 5% damping ratio. Transmitting
boundaries have been used at the truncated boundary of the reservoir. It should be
noted that the wave absorption at the reservoir bottom has not been considered.
0.4 Component L 0.4 Component T

Acceleration(g)
Acceleration(g)
0.2 0.2
0.0 0.0
-0.2 -0.2
-0.4 -0.4
0 2 4 6 8 10 0 2 4 6 8 10
Time (s) Time (s)
(a) (b)
0.4 Component V
Acceleration (g)
0.2
0.0
-0.2
-0.4
0 2 4 6 8 10
Time (s)
(c)
Fig. 10.2 Koyna earthquake on December 11, 1967. a Component L, b Component T, and
c Component V. (L: stream direction, T: cross-stream direction, and V: vertical direction)
Fig. 10.3 Geometry and finite element discretization for the Koyna dam-reservoir-foundation
system. a Quasi 3D model of the dam-reservoir-foundation system, and b enlarged plot of the
dam body
For the initial time step, the nodal displacements on the truncated boundary of the
reservoir and foundation are assumed to be zero in the normal direction. In addition,
the bottom boundary of the foundation is fully constrained. After the static analysis,
all the displacement constraints are released, and the stream and vertical components
of the Koyna earthquake accelerations are applied to the base of the foundation as
the input loading.
Figure 10.4 compares the final cracking profile of Koyna dam obtained from the
quasi-3D numerical simulation with that from the model test (Mridha and Maity
2014). The damage zone, which indicates the cracking profiles, is shaded in red
color. As can be clearly seen, the numerical cracking profile matches reasonably
well with the experimental results. It can be concluded that the presented numerical
framework can predict effectively the crack propagation process in concrete gravity
dams under seismic loadings. It should be noted that the presented test does not
consider the effect of contraction joints and a better validation may be pursued if the
experimental data for the full 3D model with multiple monoliths become available.
10.5 3D Guandi Gravity Dam-Reservoir-Foundation

System
10.5.1 Introduction to Guandi Gravity Dam
Guandi hydropower station (see Fig. 10.5), which is a roller compacted concrete
(RCC) dam, is located on the Yalong River in the Sichuan Province of China. The
crest length of the dam is 467 m, the highest monolith is 168 m high, and the normal
depth of reservoir water is 164 m. The maximum thickness of the dam base is 153.2 m
10.5 3D Guandi Gravity Dam-Reservoir-Foundation System 251
Fig. 10.4 Final failure mode of Koyna dam under the 1967 earthquake given by a three dimensional
finite element meshes; b experimental results from the model test with full reservoir. The elasticity
modulus, density, tensile, and compressive strength of the model test are 203.68 MPa, 2578 kg/m3 ,
0.19023 MPa and 0.01869 MPa, respectively (Mridha and Maity 2014)
Fig. 10.5 Site photos of the Guandi hydropower project: a upstream side; b downstream side
while the thickness at the crest of the dam is 20 m. The dam was built with 22
contraction joints, with the space between each other being approximately 20 m.
These contraction joints form a set of 23 monoliths.
The basic material parameters for the Guandi gravity dam-reservoir-foundation
system are listed in Table 10.1. It should be noted that three indices of concrete are
employed, i.e. C15, C20 and C25. However, only C20 concrete is employed for the
Table 10.1 Dynamic

Material Modulus Poisson’s Density (kg/m3 )
material parameters for the
(GPa) ratio
Guandi gravity
dam-reservoir-foundation Concrete 57.6 0.167 2552
system Foundation 21.6 0.2 Massless
rock (assumption)
Reservoir 2.07 (Bulk – 1000
water modulus)
whole dam in this Chapter due to the complexity of the 3D dam-reservoir-foundation

system. The dynamic tensile and compressive strength of concrete are 1.94 MPa and
19.38 MPa, respectively (obtained by applying a dynamic magnification factor of
1.3 to the static strength). The fracture energy is 257 N/m. The uniaxial tension
and compression stress-strain behavior used in the CDP model were obtained from
the concrete experiments conducted for the Guandi gravity dam. These stress-strain
curves, which define the nonlinear behavior of concrete, are tabulated in Table 10.2.
The material parameters for the reservoir water are assumed to be the same with those
used in the validation test. The traditional massless foundation approach is utilized
here to avoid wave propagation effects when applying free field earthquake records
at the foundation base. In this model, only the stiffness of the foundation medium is
considered. It should be noted that the massless foundation model may overestimate
the nonlinear dynamic response of concrete dams in comparison with massed foun-
dation model (Hariri-Ardebili and Mirzabozorg 2013; Hariri-Ardebili et al. 2016).
The material damping for the dam-reservoir-foundation system is considered via the
Rayleigh damping assumption with 5% damping ratio in the analysis. The value of
the damping ratio is selected according to the current seismic codes in China. We
should mention that this value may be overestimated as the nonlinear processes in
the dam will produce additional damping. While transmitting boundaries are used at
Table 10.2 Uniaxial tension and compression behavior used in the CDP model
Compression hardening and damage Tension stiffening and damage
Stress (MPa) Crushing strain Damage Stress (MPa) Cracking strain Damage
15.50 0 0 1.94 0 0
19.38 0.00007 0 1.90 0.000003 0.057
17.35 0.00024 0.228 1.74 0.000010 0.170
11.87 0.00057 0.505 1.41 0.000024 0.346
6.65 0.00124 0.724 1.00 0.000051 0.545
3.43 0.00259 0.856 0.66 0.000104 0.713
1.72 0.00528 0.927 0.41 0.000212 0.829
0.86 0.01067 0.963 0.26 0.000428 0.901
0.43 0.02143 0.981 0.16 0.000859 0.944
0.22 0.04297 0.991 0.10 0.001721 0.968
10.5 3D Guandi Gravity Dam-Reservoir-Foundation System 253
the truncated boundary of the reservoir, we do not consider the wave absorption at
the reservoir bottom.
The Guandi gravity dam is located in a highly seismic region with a design peak
ground acceleration (PGA) of 0.34 g, and found on very dense soil and soft rock with
shear wave velocity more than 500 m/s. The static solution of the dam-reservoir-
foundation system is taken as the initial condition in the nonlinear dynamic analysis.
The static loads considered here are the body self-weight, the hydrostatic pressure of
the impounded water, and the uplift pressure. The dynamic loads include earthquake
excitations in three directions and the hydrodynamic pressure. Due to the lack of
real earthquake records from a location close to the dam, three seismic records from
similar sites (i.e. 1967 Koyna #Koyna dam, 1994 Northridge #5353 and 1989 Loma
Prieta #57563) are selected as the earthquake excitation for the 3D seismic analysis
of the Guandi gravity dam. All acceleration components are scaled by the same factor
such that the PGA along the stream direction is 0.34 g. After analyzing the results,
we found that although the selected three seismic records lead to differences in the
damage-cracking pattern and response level, the observations about the effects of the
cross-stream seismic excitation and contraction joints on the cracking characteristics
of the Guandi gravity dam are more or less the same for the three earthquakes. For
clarity, we only discuss the results from the Koyna earthquake in the following.
10.5.2 3D Finite Element Model
The finite element discretization of the Guandi gravity dam-reservoir-foundation

system, together with the layout of 22 contraction joints, is shown in Fig. 10.6.
The interaction between the impounded water and the dam-foundation system is
explicitly taken into account by modeling the reservoir water with three dimensional
fluid elements in the Lagrangian formulation. The foundation extends to a distance
of 168 m (the height of the tallest monolith) in downward, upstream, downstream,
leftward, and rightward directions. The solid domain (dam body and foundation rock)
is discretized with eight-node solid elements whereas eight-node Lagrangian fluid
elements are adopted for the modeling of reservoir water. A relatively fine mesh is
used for the entire dam. The model consists of 189,411 elements and 223,017 nodes,
among them 59,465 elements are used for the dam, and 36,175 elements are used for
the reservoir domain.
For the initial time step, the displacements of nodes on the truncated boundary
of the three-dimension dam-reservoir-foundation system are assumed as zero in the
normal direction. In addition, the base of the foundation is fully constrained. In the
subsequent nonlinear dynamic analysis, all displacement constraints are released, and
three components of the selected earthquake accelerations are applied to the foun-
dation base as the input loading. The implicit Newmark-β time integration method
is applied to conduct the seismic response analysis. The initial increment size and
maximum increment size are 0.001 s and 0.01 s, respectively.
Fig. 10.6 Finite element model of the Guandi dam-foundation-reservoir system: a geometric
configuration of the coupled system; b mesh for the dam body and foundation; c layout of contraction
joints
10.6 Nonlinear Seismic Behavior 255
10.6 Nonlinear Seismic Behavior
Nonlinear dynamic analysis is performed based on the established three-dimensional

FE model, in which the effects of concrete cracking, contraction joint nonlinearity and
dam-reservoir-foundation interaction are taken into account. Furthermore, we also
investigate the potential influence of joint grouting materials on the dynamic response
of the Guandi dam by using both hard and soft pressure-clearance relationships. The
following performance criterion has been used to evaluate the seismic performance:
(1) No damage: from the initial intact state to some minor cracks (in numerical
simulation this corresponds to a small damage area with a low damage value).
The dam is considered to behave in the elastic range with little or no possibility
of damage;
(2) Slight damage: from a few minor cracks to cracks penetrating the one-third
width of the dam section. The level of nonlinear response or cracking can be
considered acceptable with no possibility of failure;
(3) Moderate damage: the crack depth is from one-third to two-third of the dam
section width. The dam can be easily repaired without affecting the dam
operation;
(4) Severe damage: crack depth is from two-third of the dam section width to almost
perforation. This damage state is associated with extensive damages requiring
immediate retrofit and rehabilitation actions.
10.6.1 Seismic Damage
The final damage profiles of the Guandi gravity dam are compared in Fig. 10.7 for
hard and soft contact formulations, where the serial number for each monolith (1–23)
is also given for better understanding. It can be seen in Fig. 10.7 that the final damage
profiles from both contact models are quite similar to each other. As can also be
seen in Fig. 10.7, cracks mainly develop in the following regions: the discontinuity
in the slope of the downstream face (monoliths 5–10, 16–20), the junction of the
pier and spillway sill (monoliths 11–15), the dam heel (monoliths 12–15), and the
upstream face (monoliths 8–10). In the simulation, the gates of the crest spillway
have not been modeled, and the piers are modeled as freestanding cantilevers, using
the constitutive model for mass concrete. Due to the cross-stream movement of the
dam, severe damage is localized in the piers. The actual damage level is expected to
be smaller since there are gates as well as traffic beams on the top of the piers that
can enhance the stiffness. In addition, piers are reinforced concrete structures and the
adopted material model could overestimate the damage of piers. Based on the above
mentioned performance criteria, the Koyna earthquake causes moderate damage to
the middle non-overflow and spillway monoliths.
Fig. 10.7 Final damage profiles of Guandi concrete dam: a upstream side, hard contact; b upstream
side, soft contact; c downstream side, hard contact; d downstream side, soft contact
10.6.2 Maximum Stream Displacement
The distribution of the maximum stream displacements along the crest is shown in
Fig. 10.8 for hard and soft contact formulations, where the positive displacement is in
the downstream direction. As can be seen, while the maximum positive displacements
of each monolith from the hard contact formulation are slightly larger than those from
the soft one, the hard contact model leads to smaller maximum displacements in the
upstream direction. Overall, the difference in the maximum stream displacements
along the crest is minor between hard and soft contact models. For non-overflow
60 Positive-Hard contact
Negative-Hard contact
Displacement (mm)
Positive-Soft contact
40 Negative-Soft contact
20
-20
-40
0 2 4 6 8 10 12 14 16 18 20 22 24
Monolith number
Fig. 10.8 Maximum positive and negative stream displacements along the dam crest
10.6 Nonlinear Seismic Behavior 257
monoliths (1–9, 17–23), the maximum stream displacements on the dam crest are
closely related to the height of monoliths: the maximum positive displacement for
non-overflow monoliths from the hard contact formulation is 39.3 mm at monolith 9,
with a height of 142 m. For overflow monoliths, the maximum positive displacement
from the hard contact formulation is 54.4 mm at monolith 11, whereas the maximum
displacement in the upstream direction is relatively small. This is because that the
structural feature of piers limits the upstream movement.
10.6.3 Contact Behavior of Contraction Joints
Envelopes of the resulting contact pressure on 22 contraction joints based on hard and
soft contact formulations are presented in Fig. 10.9. As can be seen, the envelopes of
the contact pressure resulted from the two contact models are quite close to each other.
Fig. 10.9 Envelopes of contact pressure of 22 contraction joints using a hard contact formulation
and b soft contact formulation (unit: Pa)
Fig. 10.10 Envelopes of contraction joint opening displacement for representative joints: a J3;
b J9; c J10 (unit: m)
Great contact pressure on contraction joints is generated by the dynamic interaction

between monoliths. The zones of high contact pressure are mainly located on the dam
crest and the spillway sill (near the junction of the pier and spillway). The maximum
contact pressure from the hard contact formulation is 6.40 MPa, arising on J15.
The above analyses indicate that the material properties of the joint infill have
limited influence on the nonlinear dynamic response of the Guandi dam. It is also
found that the major conclusions drawn from the following case studies are not
affected by the adopted contact model. Therefore, for clarity, the following results that
consider contraction joint nonlinearity are all based on the hard contact formulation.
Envelopes of opening displacement for three representative contraction joints J3,
J9, and J10 are depicted in Fig. 10.10. It can be found that the maximum opening
displacement appears on the dam crest. Figure 10.11 shows the time history of the
opening and sliding displacements of the three contraction joints (J3, J9 and J10) on
the dam crest. As can be seen in Fig. 10.11, the opening and sliding displacements
of J10 is significantly larger than those of J3 and J9. This can be explained by the
presence of piers in the overflow monoliths which reduce the cross-stream stiffness.
10.7 Evaluation of the Key Factors Affecting

Damage-Cracking Characteristics
In this section, four different cases are investigated in order to obtain insight into
the effects of contraction joints and cross-stream excitation on the seismic damage-
cracking behavior of the concrete gravity dam. The Guandi gravity dam-reservoir-
foundation system presented in Sect. 10.5 is taken as the reference Case A. The basic
features of the four simulation cases are summarized as follows:
(1) Case A: 3D dam-reservoir-foundation model which considers material nonlin-
earity in terms of concrete damage-cracking, joint nonlinearity in terms of
contraction joint opening and sliding, and three-component ground motion
including stream, cross-stream and vertical components.
10.7 Evaluation of the Key Factors Affecting Damage-Cracking Characteristics 259
70
J3
60 J9
J10
Contraction joint
50
openting (mm)
40
30
20
10
0
0 2 4 6 8 10
Time (s)
(a)
35
J3
30 J9
Contraction joint
25 J10
sliding (mm)
20
15
10
5
0
-5
0 2 4 6 8 10
Time (s)
(b)
Fig. 10.11 Displacement histories of three representative contraction joints (J3, J9, and J10) on
the dam crest: a opening displacement; b sliding displacement
(2) Case B: same with Case A except that the dam body is modeled as a whole
without contraction joints.
(3) Case C: same with Case A except that only stream and vertical seismic
excitations are considered.
(4) Case D: quasi-3D model is used for representative monoliths (monoliths 9,
11, and 16) where material nonlinearity and two-component seismic excitation
including stream and vertical components are considered.
It should be noted that in Case D the quasi-3D meshes for monoliths 9, 11 and
16 are the same with the meshes used in Case A for those monoliths. All analyses
are performed on a PC with an Intel(R) Core (TM) i7-4770 CPU at 3.40 GHz and
16 GB RAM. The CPU times for Case A, Case B, Case C, and Case D (monoliths
9) are 321,552 s, 5,292 s, 176,736 s, and 711 s, respectively.
10.7.1 Discussion of Results from 3D

Dam-Reservoir-Foundation Model
Figure 10.12 compares the final damage-cracking patterns of the Guandi gravity dam
for Cases A, B, and C. Note that the damage profile shown in Fig. 10.12a is the same
as that in Fig. 10.7a. As can be seen from Fig. 10.12, both the contraction joints and
cross-stream seismic excitation have a significant influence on the cracking behavior
of the dam body. The damage area for Case B in which the dam body is modeled as
a whole is smaller in comparison to the damage zone for reference Case A. This is
because that the existence of contraction joints could reduce the integral rigidity of
the dam body and thus leads to larger nonlinear demands. Severe damage region for
Case B appears primarily in the piers on the spillway sections due to the cross-stream
movement of the dam.
By comparing Fig. 10.12a, c, it can also be found that the damage level for Case C
(without cross-stream seismic excitation) is significantly lower than that of reference
Case A. There is almost no damage in the piers of the overflow monoliths in Case C,
Fig. 10.12 Comparison of seismic damage-cracking behaviors for: a Case A; b Case B (no contrac-
tion joint); c Case C (no cross-stream seismic excitation). The figures in the left column are views
from the upstream side whereas those in the right column are views from the downstream side
which implies that the damage of piers in Case A is mainly resulted from the cross-
stream seismic excitation. This is expected since the piers have larger stiffness in the
stream direction than that in the cross-stream direction. The damage on the upstream
face of monoliths 8–10 occurring in reference Case A is not observed in Case C.
This means that dynamic interaction between monoliths caused by the cross-stream
ground motion could cause significant damage to the dam.
10.7.2 Comparisons of 3D and Quasi-3D Analysis Results

for Representative Monoliths
The results from the 3D dam-reservoir-foundation FE model (Cases A, B, and C) are

compared with the results from quasi-3D models for three representative monoliths
(Case D). Figures 10.13, 10.14 and 10.15 depict the damage-cracking behaviors of
three representative monoliths (non-overflow monolith 9, overflow monolith 11, and
middle spillway monolith 16) for all four cases. It can be seen that the damage pattern
in reference Case A is significantly different from that of other cases. Compared with
quasi-3D modeling (Case D), more severe damage is observed when considering
the contraction joints and cross-stream seismic excitation (Case A), especially in
overflow monolith 11. This observation indicates the importance of the full three
dimensional modeling for the seismic safety assessment of concrete gravity dams.
In addition, it can also be observed that the damage-cracking behavior for Case C
is quite close to that from quasi-3D modeling (Case D), implying that the dynamic
interaction between monoliths is mainly caused by the cross-stream excitation.
Figure 10.16 compares the time history of the stream displacement on the dam
crest for four simulation cases. Due to hydrostatic pressure, initial stream displace-
ment exists with the maximum value occurring in Case D and the minimum value
occurring in Case B for each representative monolith. As shown in Fig. 10.16,
Fig. 10.13 Seismic damage characteristics of non-overflow monolith 9: a reference Case A; b Case
B (no contraction joint); c Case C (no cross-stream seismic excitation); d Case D (quasi-3D model
for single monolith)
Fig. 10.14 Seismic damage characteristics of overflow monolith 11: a reference Case A; b case B
(no contraction joint); c Case C (no cross-stream seismic excitation); d Case D (quasi-3D model
for single monolith)
Fig. 10.15 Seismic damage characteristics of middle spillway monolith 16: a reference Case A;
b Case B (no contraction joint); c Case C (no cross-stream seismic excitation); d Case D (quasi-3D
model for single monolith)
the dynamic responses in the stream direction almost coincide with each other
before 4.0 s and then start to deviate from each other. However, Case D
shows a similar behavior to Case C during the entire dynamic response process.
In Case A, the residual plastic displacements on the dam crest are −7.14 mm,
22.7 mm, and −11.2 mm for representative monoliths 9, 11 and 16, respectively,
which are higher than those in other cases. The nonlinear displacement response
analysis shows that the Guandi dam remains stable during the selected seismic
event. However, the final conclusions about the dynamic stability of the Guandi
dam should be based on a continuous analysis of the post-cracking dam under new
ground motions.
The cross-stream displacement time history of the three representative monoliths
is also compared in Fig. 10.17 for all cases, where the positive displacement is
in the leftward direction. Although the three-component ground motion is applied
as the input loading for Cases A and B, the cross-stream displacement in Case
Displacement (mm) 45
30
15
0
Case A
-15 Case B
Case C
-30 Case D
0 2 4 6 8 10
Time (s)
(a)
60
Case A
Displacement (mm)
50
Case B
40
Case C
30 Case D
20
10
0
-10
-20
-30
0 2 4 6 8 10
Time (s)
(b)
45
Displacement (mm)
30
15
0
Case A
-15 Case B
-30 Case C
Case D
-45
0 2 4 6 8 10
Time (s)
(c)
Fig. 10.16 Comparison of stream displacements on the dam crest: a non-overflow monolith 9;
b overflow monolith 11; c middle spillway monolith 16
15
Displacement (mm)
-15 Case A
Case B
-30 Case C
Case D
0 2 4 6 8 10
Time (s)
(a)
45
Displacement (mm)
30
15
0
-15 Case A
Case B
-30 Case C
-45 Case D
0 2 4 6 8 10
Time (s)
(b)
30
Displacement (mm)
15
-15 Case A
Case B
-30 Case C
Case D
0 2 4 6 8 10
Time (s)
(c)
Fig. 10.17 Comparison of cross-stream displacements on the dam crest: a non-overflow monolith
9; b overflow monolith 11; c middle spillway monolith 16
A is substantially higher than that in Case B. This means that the contract joint
has significant influence on the nonlinear dynamic response along the cross-stream
direction. Although the cross-stream seismic excitation is not considered in Case
C, small cross-stream displacement occurs due to the dynamic interaction between
monoliths.
What can be concluded with the comparisons presented in the subsection is that
contraction joints and cross-stream seismic excitation play a role of great importance
in the seismic performance of the concrete gravity dam. Performing quasi-3D or 2D
analyses, in which the aforementioned two factors cannot be appropriately captured,
may significantly underestimate the damage-cracking risk of concrete gravity dams
when subjected to strong ground motions, which is non-conservative for the seismic
design of concrete gravity dams. It is worth noting that the obtained conclusions
are based on the deterministic analyses using three real earthquake records. The
influence of contraction joints and cross-stream seismic excitation on the dynamic
response of concrete gravity dams depends on a variety of factors including the
ground motion input and site conditions. To better assess such influence and in the
spirit of performance based earthquake engineering (Hariri-Ardebili and Saouma
2016a, b), seismic fragility analysis using abundant strong motion records is planned
as future work.
10.8 Conclusions
In this contribution, a comprehensive study of the seismic damage-cracking behavior

of the Guandi concrete gravity dam is presented based on the three dimensional
dynamic nonlinear finite element model. All of the following important factors are
taken into account: the interaction of dam-reservoir-foundation, the opening/closing
and sliding of contraction joints, and the cross-stream seismic excitation. Compar-
isons are made between the results from both the hard contact formulation and expo-
nential contact model. The effectiveness of the numerical framework which considers
the dam-reservoir-foundation interaction is first validated against the experimental
results of the Koyna concrete gravity dam. Then case studies are conducted in order to
characterize the effects of contraction joints and cross-stream ground motion on the
seismic performance of the Guandi concrete gravity dam. The following conclusions
can be drawn based on the analysis results:
(1) The type of selected contact model has limited influence on the nonlinear
dynamic response of Guandi dam.
(2) Nonlinear dynamic responses are closely related to the height and structural
feature of monoliths. Large damage appears primarily in the middle monoliths of
the river bed. Higher strength concrete can be used in these regions to minimize
the expected seismic damage and improve the safety of Guandi dam.
(3) Contraction joints are expected to open/close and slide repeatedly during strong
ground motions. The existence of contraction joints could reduce the integral
rigidity of the monolithic model, and thus lead to larger nonlinear demands
in the three dimensional dam-reservoir-foundation system. Dynamic interac-
tion between monoliths caused by the cross-stream ground motion could cause
significant damage to the dam.
(4) Damage-cracking level from the three-dimensional model with the contrac-
tion joints and cross-stream seismic excitation considered is much severer
than the quasi-3D results. Hence, the quasi-3D or 2D analyses may signifi-
cantly underestimate the damage-cracking risk of concrete gravity dams when
subjected to strong ground motions, which is non-conservative for the seismic
design of concrete gravity dams. The observation indicates the importance of
three-dimensional modeling for the seismic safety evaluation.
The aforementioned conclusions are obtained from the response of the Guandi
gravity dam subjected to three specific earthquake ground motions. Thus, these
conclusions may not apply to all gravity dams and ground motions. Further
study using incremental dynamic analysis, richer ground motion database and
reinforcement concrete material model for the piers will be undertaken in the future.
References
Akkose, M., Baraktar, A., & Dumanoglu, A. A. (2008). Reservoir water level effects on nonlinear
dynamic response of arch dams. Journal of Fluids and Structures, 24(3), 418–435.
Alembagheri, M., & Ghaemian, M. (2013a). Seismic assessment of concrete gravity dams using
capacity estimation and damage indexes. Earthquake Engineering and Structural Dynamics,
42(1), 123–144.
Alembagheri, M., & Ghaemian, M. (2013b). Damage assessment of a concrete arch dam through
nonlinear incremental dynamic analysis. Soil Dynamics and Earthquake Engineering, 44, 127–
137.
Alembagheri, M., & Ghaemian, M. (2015). Seismic performance evaluation of a jointed arch dam.
Alembagheri, M., & Seyedkazemi, M. (2015). Seismic performance sensitivity and uncertainty
analysis of gravity dams. Earthquake Engineering and Structural Dynamics, 44(1), 41–58.
Arici, Y. N., Binici, B., & Aldemir, A. (2014). Comparison of the expected damage patterns from
two- and three-dimensional nonlinear dynamic analyses of a roller compacted concrete dam.
Azmi, M., & Paultre, P. (2002). Three-dimensional analysis of concrete dams including contraction
joint non-linearity. Engineering Structures, 24(6), 757–771.
Calayir, Y., & Dumanoǧlu, A. A. (1993). Static and dynamic analysis of fluid and fluid-structure
systems by the Lagrangian method. Computers & structures, 49(4), 625–632.
Chen, D., Du, C., Yuan, J., & Hong, Y. (2012). An investigation into the influence of damping on
the earthquake response analysis of a high arch dam. Journal of Earthquake Engineering, 16(3),
329–349.
References 267
Feng, J., Wei, H., Pan, J., Jian, Y., Wang, J., & Zhang, C. (2011). Comparative study procedure for
the safety evaluation of high arch dams. Computers and Geotechnics, 38(3), 306–317.
Hariri-Ardebili, M. A., Mirzabozorg, H., & Kianoush, M. R. (2013). Seismic analysis of high
arch dams considering contraction-peripheral joints coupled effects. Central European Journal
of Engineering, 3(3), 549–564.
Hariri-Ardebili, M. A., & Kianoush, M. R. (2014). Integrative seismic safety evaluation of a high
concrete arch dam. Soil Dynamics and Earthquake Engineering, 67, 85–101.
Hariri-Ardebili, M. A., & Mirzabozorg, H. (2013). A comparative study of seismic stability of
coupled arch dam-foundation-reservoir systems using infinite elements and viscous boundary
models. International Journal of Structural Stability and Dynamics, 13(06), 1350032.
Hariri-Ardebili, M. A., & Saouma, V. E. (2016a). Sensitivity and uncertainty quantification of the
cohesive crack model. Engineering Fracture Mechanics, 155, 18–35.
Hariri-Ardebili, M. A., & Saouma, V. E. (2016b). Collapse fragility curves for concrete dams:
Comprehensive study. Journal of Structural Engineering, 142(10), 4016075.
Huang, J., & Zerva, A. (2013). Earthquake performance assessment of concrete gravity dams
subjected to spatially varying seismic ground motions. Structure and Infrastructure Engineering,
10(8), 1011–1026.
Lee, J., & Fenves, G. L. (1998a). A plastic-damage concrete model for earthquake analysis of dams.
Lee, J., & Fenves, G. L. (1998b). Plastic-damage model for cyclic loading of concrete structures.
Mridha, S., & Maity, D. (2014). Experimental investigation on nonlinear dynamic response of
concrete gravity dam-reservoir system. Engineering Structures, 80, 289–297.
63, 80–97.
Omidi, O., & Lotfi, V. (2013). Earthquake response of concrete arch dams: A plastic-damage
approach. Earthquake Engineering and Structural Dynamics, 42(14), 2129–2149.
1594–1606.
Version, A. (2011). ABAQUS user’s manual. Dassault Systèmes, SIMULIA.
Wang, G., Wang, Y., Lu, W., Yan, P., Zhou, W., & Chen, M. (2016). A general definition of integrated
strong motion duration and its effect on seismic demands of concrete gravity dams. Engineering
Wang, G., Wang, Y., Lu, W., Yan, P., Zhou, W., & Chen, M. (2017). Damage demand assess-
ment of mainshock-damaged concrete gravity dams subjected to aftershocks. Soil Dynamics and
Wang, H., Feng, M., & Yang, H. (2012a). Seismic nonlinear analyses of a concrete gravity dam
with 3D full dam model. Bulletin of Earthquake Engineering, 10(6), 1959–1977.
Wang, J., Lv, D., Jin, F., & Zhang, C. (2013). Earthquake damage analysis of arch dams considering
dam-water-foundation interaction. Soil Dynamics and Earthquake Engineering, 49, 64–74.
Wang, J., Zhang, C., & Jin, F. (2012b). Nonlinear earthquake analysis of high arch dam-water-
foundation rock systems. Earthquake Engineering and Structural Dynamics, 41(7), 1157–1176.
Wang, Y., & Waisman, H. (2016). From diffuse damage to sharp cohesive cracks: A coupled XFEM
framework for failure analysis of quasi-brittle materials. Computer Methods in Applied Mechanics
and Engineering, 299, 57–89.
Wilson, E. L., & Khalvati, M. (1983). Finite elements for the dynamic analysis of fluid-solid systems.
Yilmazturk, S. M., Arici, Y., & Binici, B. (2015). Seismic assessment of a monolithic RCC gravity
dam including three dimensional dam-foundation-reservoir interaction. Engineering Structures,
100, 137–148.
Zhang, S., Wang, G., & Sa, W. (2013a). Damage evaluation of concrete gravity dams under
16–27.
Zhang, S., & Wang, G. (2013). Effects of near-fault and far-fault ground motions on nonlinear
dynamic response and seismic damage of concrete gravity dams. Soil Dynamics and Earthquake

Seismic Performance Analysis of Concrete Gravity Dams by Gaohui Wang, Wenbo Lu, Sherong Zhang

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Seismic Performance Analysis of Concrete Gravity Dams by Gaohui Wang, Wenbo Lu, Sherong Zhang

Uploaded by

Copyright:

Available Formats

Advanced Topics in Science

and Technology in China 57

More information about this series at http://www.springer.com/series/7887

ISSN 1995-6819 ISSN 1995-6827 (electronic)

Wuhan, China Gaohui Wang

5 Correlation Between Single Component Durations and Damage

6.4.3 Damage Dissipation Energy Response . . . . . . . . . . . . . 145

8.4 Effects of Earthquake Direction . . . . . . . . . . . . . . . . . . . . . . . . 190

10 Deterministic 3D Seismic Damage Analysis of Guandi Concrete

B ¼ rw Matrix of derivatives of extended shape functions wri

Kc Strength ratio of concrete under equal biaxial compression to

d Scalar stiffness degradation variable

~epc Equivalent compressive plastic strain

CDP Concrete damaged plasticity

PGD Peak ground displacement

Table 1.1 Concrete gravity

1.2 Objective and Scope of This Research

it is influenced by many factors, such as earthquake source mechanism, propagation

primary characteristics of near-fault ground motions (Mavroeidis and Papageorgiou

1.3 Organization of the Book

with considering the effects of dam-reservoir-foundation interaction. The Lagrangian

2.2 Method for Dynamic Failure Analysis of Concrete

2.2.1 Prototype Observation

2.2.2 Model Test Method

uncontrolled release of water. This is the first

2.2.3 Numerical Simulation Method

Ma et al.(2019) investigated the seismic dynamic response and crack propagation

nonlinear time-domain dynamic analyses to evaluate the input motion mechanism

2.3 eXtended Finite Element Method (XFEM)

2.3.1 XFEM Approximation

The XFEM enriches a standard displacement based finite element approximation

Fig. 2.1 Enriched nodes in the XFEM

2.3.2 Enrichment Functions

is made possible by adding phantom nodes (n 1 to n 4 ) superimposed to the original

The approximation of the displacement field is then given by:

where [u] is the jump in the displacement given as

In order to control the magnitude of separation, the cohesive law is defined. A

2.3.3 Discrete Equations

The proposed approach is based on an elastodynamics behavior for XFEM analysis

Fig. 2.4 Notations used for

where t̄ is the normalized traction prescribed on Γt ; τc is the cohesive traction applied

Müh + Cu̇h + Kuh = f (2.12)

FEM(uu ), Heaviside enrichment(aa ), orthotropic crack tip enrichment(cc ) and the

and B = ∇ψ is the matrix of derivatives of extended shape functions ψir .

2.4 Concrete Damaged Plasticity (CDP) Model

In order to describe the complex mechanical behavior of concrete material under

2.4.1 Damage Evolution

σ = (1 − d)σ̄ = (1 − d)E 0 (ε − ε p ) and d = d(κ) (2.23)

I¯1 = σ̄ii (2.25)

where S̄i j is the effective deviatoric stress tensor.

σc = (1 − dc )E 0 (εc − εcp ) (2.28)

2.4.2 Yield Criterion

2.4.3 Flow Rule

Fig. 2.6 Yield surfaces in the deviatoric plane

Fig. 2.7 Initial yield function in plane stress space

where λ̇ is a non-negative function referred to as the plastic consistency parameter.

2.5 Lagrangian Formulation for Dynamic Interaction

the irrotationality of fluid by penalty method. It should be as high as necessary to

where H is the matrix consisting of interpolation functions of the fluid element.

Un− = Un+ (2.46)

Mc Üc + Cc U̇c + Kc Uc = Rc (2.47)

2.6 Application of the Two Models in Concrete Gravity

2.6.1 Description of Koyna Gravity

0.4 Horizontal Component 0.4 Vertical Component

is made possible by adding phantom nodes (n 1 to n 4 ) superimposed to the original