Internal Swelling Reactions

in Concrete
As critical concrete infrastructure deteriorates, engineers need efficient and
reliable techniques to appraise the causes and the extent of deterioration,
to evaluate the structural consequences and to select effective management
protocols and rehabilitation strategies. This book looks at deterioration
caused by internal swelling reaction (ISR) mechanisms in concrete, such
as alkali-­aggregate reaction, delayed ettringite formation and freeze-­thaw
cycles.
The book provides accessible and comprehensive coverage of recent work
and developments on the most common ISR mechanisms leading to induced
expansion and deterioration. It addresses the implications of ISR on differ-
ent scales (micro, meso and macro), outlines qualitative and quantitative
techniques to assess the condition of affected concrete and introduces the
multi-­level assessment protocol, using advanced microscopic and mechani-
cal techniques, particularly the stiffness damage test and damage rating
index, as a reliable approach to appraise ISR-­affected infrastructure. Also
included is a detailed case study of the Robert-­Bourassa Charest Overpass
in Quebec.
Internal Swelling Reactions in Concrete: Mechanisms and Condition
Assessment is primarily intended for undergraduate and graduate students,
as well as academics interested in the field of concrete durability and condi-
tion assessment of concrete. It will also be of interest to engineers and infra-
structure owners dealing with ISR-­related problems.
 Internal Swelling
Reactions in Concrete
Mechanisms and Condition
Assessment

Leandro F. M. Sanchez
Cover image: Leandro F. M. Sanchez

First edition published 2024

by CRC Press
2385 NW Executive Center Drive, Suite 320, Boca Raton FL 33431

and by CRC Press

4 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN

CRC Press is an imprint of Taylor & Francis Group, LLC

© 2024 Leandro F. M. Sanchez

Reasonable efforts have been made to publish reliable data and information, but the
author and publisher cannot assume responsibility for the validity of all materials or
the consequences of their use. The authors and publishers have attempted to trace the
copyright holders of all material reproduced in this publication and apologize to copy-
right holders if permission to publish in this form has not been obtained. If any copy-
right material has not been acknowledged please write and let us know so we may
rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted,
reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other
means, now known or hereafter invented, including photocopying, microfilming, and
recording, or in any information storage or retrieval system, without written permis-
sion from the publishers.

For permission to photocopy or use material electronically from this work, access
www.copyright.com or contact the Copyright Clearance Center, Inc. (CCC), 222
Rosewood Drive, Danvers, MA 01923, 978-750-8400. For works that are not available
on CCC please contact [email protected]

Trademark notice: Product or corporate names may be trademarks or registered trade-

marks and are used only for identification and explanation without intent to infringe.

ISBN: 978-1-032-03598-7 (hbk)

ISBN: 978-1-032-03599-4 (pbk)
ISBN: 978-1-003-18815-5 (ebk)

DOI: 10.1201/9781003188155

Typeset in Sabon
by SPi Technologies India Pvt Ltd (Straive)
Life is a challenging and yet different journey for everyone. Nobody knows
exactly the main objective(s) of such a journey, and all of us have very dis-
tinct interests; however, I believe most of us agree that as long as we do
things with passion, we are on the right track.
Born and raised in a country with extreme challenges, I have been taught
from my early years that education should be amongst my main priorities.
My parents have done everything they could so that my sister and I could
have the best education possible, both informal (at home) and formal (at
school). They made me see how important professors were for society and
how these outstanding human beings could positively impact the lives of
people. These professionals give much more than they take, teach much
more than they learn and share much more than they withhold. Yes, profes-
sors change lives! They simply do! From teaching important and specific
topics to making your dreams possible and helping create the means, profes-
sors make the impossible come true. From my best professors, I learnt that
if I kept books as my best friends, I should be ok. And my life went on this
way…always with a book as my best friend.
Life was super kind to me on this purpose. I have had the opportunity to
have amazing professors in my life, much more than I could ever list herein;
however, some of them were not only important to my technical knowledge
and background but were crucial to shaping the person I am. My first pro-
fessors were my parents. My mom taught me Portuguese, English, history
and geography, while my dad taught me math and physics. Besides being my
superheroes, I had two role models directly at home to follow from the very
beginning, and I am so glad for it. Growing up, I met numerous great profes-
sors at distinct levels of my formal education, and each of them taught me
incredible lessons I still remember nowadays.
When I was 16, I was unsure which path to follow and was losing motiva-
tion at school. Yet I met a very important physics professor, Ms. Virginia
Simoes Cortez, who was the first person who told me I had talent and should
pursue a STEM degree. This dictated the choices I would have made in the
years to come. Already at the university, when the future was still unclear to
me, I registered myself on a course called Applied Math to Life; in this
course, I met the incredible professor Aguinaldo Prandici Ricieri, who taught
me, among integrals, derivatives and others math and physics tricks, the love
for science and the possibility to achieve my dreams in a research-oriented
career, as long as I kept my passion and determination. Upon graduation,
uncertain of how the best manner would be to formally initiate a graduate
programme, I had the opportunity to meet Professor Cardoso, dean of the
Polytechnic School of Sao Paulo at the time. Professor Cardoso was one of
the best mentors I have ever had in my life. He explained to me, little by
little, the career of a researcher and pointed me to the graduate programme
at the Polytechnic School of Sao Paulo, where everything started.
I then initiated a master of applied sciences (MASc) programme under the
supervision of Professors Paulo Helene and Selmo Kuperman; after a pre-
liminary proposal to conduct a project on steel corrosion, for some reason
beyond my control, they changed completely my project to work on an
international interlab led by a Canadian Institution (CANMET) to study
internal swelling reactions (ISRs) in concrete, particularly alkali-aggregate
reaction (AAR); I did not know at the time, but I was about to fall in love
with ISR and this new project would change my entire career, for the better.
Moreover, Professors Paulo Helene and Selmo Kuperman became much
more than merely supervisors to me but my career inspirations, mentors and
very close friends. Yet during my MASc, I had the opportunity to meet and
discuss with so many great professors, and amongst them professor Maria
Alba Cincotto, who not only taught me Cement Chemistry but also unfor-
gettable lessons on determination, science and ethics, lessons that I carry
with me until today. From time to time, I remember her great lessons (inside
and outside the classroom) on the Polytechnic School campus in Sao Paulo,
and I still miss them.
When my MASc degree was successfully completed, I knew I wanted
more, but I could not imagine that the best was yet to come. It was then
when life played a very important role, and I received a formal contact from
Dr. Benoit Fournier, who had recently quit CANMET and became a profes-
sor at Laval University, to pursue a PhD under him on AAR. I promptly
accepted the invitation and took it not only as a likely great experience but
as a dream coming true and the opportunity of my life. Over the years as a
PhD student, I realized very quickly that research, especially on ISR, was
exactly what I wanted for my life. Dr. Benoit Fournier, definitely amongst
the best researchers in the field, was not only a great supervisor for me but
also a wonderful mentor, and why not a father? I am so glad and thankful
for all the lessons I learnt from him and will never forget all the help and
support I received. I remember thinking when I started my position at the
University of Ottawa that if I could be half of what Benoit was for me to my
students, I would be extremely satisfied!
Towards the end of my PhD, I was introduced to professors Denis Mitchell
and Jose Bastien, who finally supervised and guided me in a postdoctoral
programme. They also were so important to me, not only on the research
programme itself but on everything else that an early career researcher
needs, such as support, guidance and motivation. I will never forget some of
the discussions we had at McGill. I also miss this time.
Finally, I would like to dedicate this book to my mentor Ramon, along
with my beloved Diana (wife), Ian and Noah (kids), who are my daily
professors in the art of love…before knowing you, I could never imagine
love could be so huge.

This book is dedicated to the professionals who change lives, the ones
who take care more of others than themselves, to all professors, particu-
larly to the professors of my life.
Leandro F. M. Sanchez
July 23, 2023
Ottawa, ON, Canada
“In the end, I don’t really know whether this work belongs to me,
although I’m pretty sure that I completely belong to it”.
Leandro F. M. Sanchez
Contents

About the Author xiii

Foreword xiv
Preface xvi
Acknowledgments xviii

1 Introduction 1
1.1 Introduction 1
1.2 Chapter descriptions 4
References 4

2 Internal swelling reactions (ISRs) mechanisms 6

2.1 Introduction: internal swelling reactions (ISRs) 6
2.2 Alkali-aggregate reaction (AAR) 6
2.2.1 Induced expansion mechanism 7
2.2.2 AAR microscopic damage features in concrete 9
2.2.3 AAR influence on mechanical properties of
affected concrete 11
2.3 Delayed ettringite formation (DEF) 12
2.3.1 Induced expansion mechanism 13
2.3.2 DEF microscopic damage features in concrete 14
2.3.3 DEF influence on mechanical properties
of affected concrete 16
2.4 Other mechanisms 17
2.4.1 Freezing and thawing cycles 17
2.4.2 Internal sulphate attack (ISA) 21
2.4.3 Thaumasite formation (TF) 23
2.4.4 Physical sulphate attack (PSA) 24
2.4.5 Hydration of crystalline MgO and CaO 24
References 26

ix
x Contents

3 Assessing the condition of ISR-affected concrete 30

3.1 Service life of concrete infrastructure 30
3.2 Condition assessment of concrete infrastructure 31
3.2.1 Examination of current documentation 32
3.2.2 Visual inspection (VI) 34
3.2.3 Field testing 35
3.2.4 Laboratory analyses 37
3.2.5 Structural analyses 37
3.3 ISRs as ongoing deterioration processes 39
References 41

4 Visual inspection and non-destructive testing (NDT) 44

4.1 Introduction 44
4.2 Visual inspection (VI) 44
4.2.1 ISR-induced surface damage signs on concrete 45
4.2.2 Quantifying ISR-induced deterioration
in concrete 50
4.2.3 Other VI techniques 52
4.3 Non-destructive testing (NDT) 53
4.3.1 Electrical resistivity (ER) 54
4.3.2 Surface thermography (ST) 56
4.3.3 Ground penetrating radar (GPR) 57
4.3.4 Stress waves 58
4.3.5 Resonant frequency (RF) 61
References 63

5 Microscopic analyses 68
5.1 Introduction 68
5.2 Petrographic analysis: assessing the damage cause(s) 68
5.2.1 Introduction 68
5.2.2 Sample preparation for petrography 69
5.2.3 Evaluation of ISR in concrete through petrography 70
5.3 Petrographic analysis: assessing the damage degree 73
5.3.1 Introduction 73
5.3.2 Damage rating index (DRI) 74
5.3.3 Image analysis 82
5.4 Emerging technologies to quantify damage in concrete 84
References 86
 Contents xi

6 Mechanical tools 90
6.1 Introduction 90
6.2 Test procedures in compression 91
6.2.1 Compressive strength test 91
6.2.2 Modulus of elasticity (ME) 96
6.2.3 Stiffness damage test (SDT) 99
6.3 Test procedures in tension 106
6.3.1 Tensile strength test 106
6.4 Test procedures in shear 114
6.4.1 Direct shear test 114
References 118

7 Multi-level assessment of ISR-affected concrete 121

7.1 Introduction 121
7.2 Multi-level assessment protocol 122
7.3 Practical use of the multi-level assessment protocol 123
7.3.1 Using the multi-level assessment protocol
to understand the impact of ISR under
unrestrained conditions 124
7.3.2 Using the multi-level assessment protocol
to understand the impact of ISR under
restrained conditions 134
References 136

8 Forecasting future performance and managing critical

infrastructure 139
8.1 Introduction 139
8.2 Prognosis and the potential for further deterioration 139
8.2.1 Laboratory techniques 140
8.2.2 Prognosis estimation based on laboratory test
procedures 144
8.2.3 Modelling 149
8.3 Management and rehabilitation of ISR-affected concrete
infrastructure 155
8.3.1 Management protocols 155
8.3.2 Rehabilitation techniques 159
References 163
xii Contents

9 Case study: condition assessment of the

Robert-Bourassa/Charest (RBC) overpass 167
9.1 Introduction 167
9.2 Condition assessment of RBC overpass 169
9.2.1 Definition of exposure conditions (microclimate) 169
9.2.2 Visual inspection and crack measurements 169
9.2.3 Coring 174
9.2.4 Laboratory test methods 176
9.3 Results 178
9.3.1 Stiffness damage test (SDT) 178
9.3.2 Damage rating index (DRI) 184
9.3.3 Compressive strength results 186
9.4 Discussion 188
9.4.1 Multi-level assessment application 188
9.4.2 Visual vs multi-level assessment 192
9.4.3 Potential structural implications on ISR-affected
columns 193
References 194

10 Conclusions and future works 196

10.1 Conclusions, current challenges and future opportunities 196

Index 198
About the Author

Dr. Leandro F. M. Sanchez obtained a bachelor’s

and MASc in civil engineering from Maua School
of Engineering (Sao Caetano, Brazil, 2004) and
Polytechnic of Sao Paulo (Sao Paulo, Brazil,
2008), respectively; a PhD in earth sciences from
Université Laval (Quebec City, Canada, 2014);
and a postdoctoral fellowship in civil engineer-
ing from McGill University (Montreal, Canada,
2015). He joined the Department of Civil
Engineering at the University of Ottawa in 2015,
where he currently is an associate professor. His
expertise is related to the development of sustain-
able concrete materials and concrete durability, especially the diagnosis
and prognosis of concrete affected by internal swelling reactions (ISRs). He
has co-­authored over 150 refereed publications on concrete sustainability
and durability and contributed to the development of standard test proto-
cols, along with descriptive, empirical and numerical models for the design
and assessment of concrete. Dr. Sanchez is an active participant on several
national and international committees and is currently the deputy chair of
the RILEM TC 300 -ARM: Alkali-reaction mitigation in concrete and the
chair of ACI Committee 221 – Aggregates. He is also the current chair of the
17th International Conference on Alkali-­Aggregate Reaction in Concrete
(ICAAR), to be held in Ottawa in 2024. Dr. Sanchez is a reviewer for a
number of important journals in the field and is a recipient of highly pres-
tigious awards such as the VANIER PhD scholarship (2010), NSERC Early
Career Researcher Award-Discovery Launch Supplement (2016), NFRF-­
Exploration grant (2019), the distinguished Ontario Ministry of Colleges
and Universities Early Career Researcher Award (ERA, 2021) and the
Faculty of Engineering Early Career Researcher Award (ECRA, 2023).

xiii
Foreword

Professor Leandro F. M. Sanchez presents to the scientific and academic

community an extremely important publication on internal swelling reac-
tions (ISR), their mechanisms and how to assess the condition of the ISR-­
affected concrete structures.
There are currently, in the world, thousands of concrete structures that
display deterioration signs at different levels caused by ISR, such as alkali-­
aggregate reaction (AAR) and internal sulphate attack (ISA); the latter may
be caused by either delayed ettringite formation (DEF) or by use of sulphide-­
bearing aggregates in concrete.
This book covers in detail the most common ISR mechanisms affecting
concrete infrastructure worldwide, such as AAR, DEF and freeze-­thaw (FT)
cycles, which can be particularly a problem in cold climate countries, such
as Canada (Chapters 1 and 2). With a very didactic scheme, the author then
approaches conventional and advanced techniques used to assess the condi-
tion of ISR-­affected structures, helping the reader to better diagnose and
make predictions (i.e., prognose) of affected structures (Chapter 3).
Techniques such as visual inspection (VI) and non-­destructive testing (NDT),
which are very useful for anyone performing infrastructure evaluations, are
comprehensively covered in Chapter 4. Furthermore, two chapters of inter-
est for students and engineers – namely, microscopic analyses and mechani-
cal tests – are thoroughly described in Chapters 5 and 6, emphasizing the
various microscopic techniques currently available for diagnosing ISR in
concrete along with assisting the physical integrity appraisal of affected con-
crete, respectively. The multi-­level assessment protocol is then presented in
Chapter 7, encompassing selected microscopic and mechanical tools earlier
discussed and displaying a novel and promising approach to detect the cause
and extent of ISR-­induced deterioration in concrete.
The sequence of the book brings a very important chapter (Chapter 8 –
forecasting future behaviour and managing critical infrastructure affected
by ISR), which by far is the least understood topic covered in the book by
the ISR community and the one requiring further research and develop-
ments; this chapter will greatly help engineers and infrastructure owners.
Yet, perhaps the most important chapter in this book is Chapter 9, where a

xiv
 Foreword xv

study case is presented by Professor Sanchez, in which a clear understanding

of the impact of ISR, particularly AAR and FT, on the durability and long-­
term performance of a deteriorated overpass in Quebec City, Canada, is
achieved via the use of the multi-­ level assessment. This chapter clearly
emphasizes the suitability of the proposed approach to recognize the cause(s)
and extent of deterioration of affected structures and support infrastructure
owners to make informed decisions on their assets.
Finally, Professor Sanchez, with many years of experience in dealing with
cases and studies on internal swelling reactions, points to the future, indicat-
ing recommendations and further studies yet to be performed for a better
understanding and management of ISR in concrete.
Selmo Chapira Kuperman
Sao Paulo, SP, Brazil
Preface

Critical infrastructure is crucial to society, driving the economy, connecting

businesses, communities and people and improving the quality of life of
human beings. Nowadays, it is recognized that ISRs are amongst the most
harmful deterioration mechanisms affecting the durability, serviceability and
long-­term performance of critical concrete infrastructure around the globe.
There is currently a large amount of research conducted over the last
decades on the most common ISR mechanisms, such as AAR and DEF; how-
ever, the available documents bear two important drawbacks: (a) they nor-
mally focus on a single ISR mechanism, and thus the information on the
impact of various ISR mechanisms on concrete is quite spread out, difficult
to find and compare to one another; furthermore, the assessment of com-
bined mechanisms, which is very likely to occur in practice, is generally not
addressed. And (b) most of these works were developed through a single
“scale” approach; this means that either microscale (e.g., microscopy),
mesoscale (e.g., mechanical tests) or even macroscale (i.e., structural behav-
iour) is addressed in most of the available documentation. This makes the
knowledge developments in this field super deep and specific, which is scien-
tifically outstanding on the one hand but makes the overall understanding
of the state of the art in the area quite complicated. The idea of this book
was then to develop a single, accessible and comprehensive document in
which the recent works and developments on the most common ISR mecha-
nisms leading to induced expansion and deterioration could be discussed;
moreover, the implications of ISRs on different scales (i.e., micro, meso and
macro) are also addressed, along with discussions on the various stages of
the condition assessment of ISR-­affected structures, from preliminary (i.e.,
diagnosis), to middle (i.e., prognosis) or even final stages (i.e., rehabilitation
strategies), which can help infrastructure owners and engineers to better
cope with ISR-­induced deterioration in practice. This book is primarily
intended for undergraduate and graduate students and academics interested
in the field of concrete durability and condition assessment of concrete as

xvi
 Preface xvii

well as engineers and infrastructure owners dealing with problems related to

ISRs. Hopefully, this book will be valuable and will support engineers and
owners to make better and more informed decisions on the management of
ISR-­deteriorated infrastructure assets.
Leandro F. M. Sanchez
Ottawa, ON, Canada
Acknowledgments

I was hired in July 2015 to work in the Department of Civil Engineering

at the University of Ottawa. Over the last eight years, my life has never
been so busy and hectic, and yet I have never had so much fun! We created
a new research group (i.e., μStructure) that works on innovative and chal-
lenging theoretical and applied research related to concrete technology, sus-
tainability and durability. This group currently comprises over 25 students
among volunteers, undergrads, MAScs, PhDs and postdoctoral fellows. It is
composed of people with distinct backgrounds and interests, bearing high
technical skills, ethics and values.
As time went by, our relationship became above and beyond research and
academia, and we actually developed a “concrete” family to which we all
feel we belong. As I always tell them, it is definitely not about “me” but
rather about “us”; it is always because of the group and for the group. This
book would not be possible without the extremely important help and sup-
port I have had from several of my super bright students. From drafting and
reading texts, finding issues and correcting typos, to working on figures,
tables and references, these people were crucial for this project to take place.
My special thanks and gratitude to my very special, passionate and talented
students: Cassandra Trottier (PhD student), Ana Bergmann (PhD student),
Rennan Medeiros (PhD student), Diego Jesus de Souza (former PhD stu-
dent), Andisheh Zahedi (former PhD student) and Thuc Nguyen (former
PhD student). You all rocked it and were side by side with me from the
beginning to the end of this project. I am so glad and lucky to have all of you
in our family and will never forget what you have done!
Leandro F. M. Sanchez
Ottawa, ON, Canada

xviii
Chapter 1

Introduction

1.1 INTRODUCTION

Our lives critically depend on concrete infrastructure on an everyday basis.

Such infrastructure is designed with a lifespan of 50–75 years according
to the code used and the structure type (i.e., dams, bridges, tunnels, build-
ings, etc.). A large number of critical structures built in the 1960s to 1980s
worldwide are now reaching the end of their service lives, besides present-
ing major distress signs caused by numerous damage mechanisms. Action is
then required to ensure adequate performance over their last few years of
service or even to extend their lifespan.
Amongst the main processes affecting critical concrete infrastructure
around the globe, internal swelling reactions (ISRs) are likely the most
harmful mechanisms, leading to induced expansion, mechanical properties
and physical integrity reductions of affected structures and structural mem-
bers, and decreasing their durability and serviceability performance. ISR is
conventionally associated with alkali-­aggregate reaction (AAR) and delayed
ettringite formation (DEF), although other mechanisms, such as freeze and
thaw and internal sulphate attack derived from sulphide-­bearing aggregates,
may also induce internal swelling processes leading to induced expansion
and deterioration (Sanchez et al., 2018; Noël et al., 2018).
AAR is normally divided into alkali-­silica reaction (ASR) and alkali-­
carbonate reaction (ACR); ASR is by far the most common process affecting
concrete structures around the world (Fournier & Bérubé, 2000). ASR is a
chemical reaction between the alkali hydroxides (i.e., Na+, K+, OH−) from
the concrete pore solution and certain unstable mineral phases found in the
aggregates used to make concrete; ASR produces a secondary reaction prod-
uct (i.e., ASR gel) that swells upon moisture uptake, leading to induced
expansion and cracking of the affected material (Fournier & Bérubé, 2000).
Otherwise, ACR is a much less common reaction whose distress mechanism
is still mostly unknown, being considered as a form of ASR by some authors
(Katayama, 2010; Katayama & Grattan-­Bellew, 2012) while other research-
ers believe that ACR follows a “different” and unique mechanism (Canada
Standards Association [CSA Group], 2019 – Appendix B). Nevertheless,

DOI: 10.1201/9781003188155-1 1
2 Internal Swelling Reactions in Concrete

ACR is assumed to take place with the use of carbonate rocks in concrete
through a process called “dedolomitization” with the formation of brucite
and calcite (Fournier & Bérubé, 2000).
DEF is defined as the formation of ettringite in concrete after setting (or
at least a substantial portion of the whole setting process) without the pen-
etration or diffusion of any supplementary external sulphate source from
the surroundings (Taylor et al., 2001; Collepardi, 2003; Martin, 2011).
Usually, DEF takes place whether the concrete undergoes temperatures
above 65°C in the very early hours after pouring, which is very common in
either steam-­cured members at high temperatures or in mass concrete due to
the temperature rise during cement hydration (Taylor et al., 2001; Collepardi,
2003; Martin, 2011). Upon high relative humidity over continuous time
periods, DEF leads to induced swelling and deterioration of affected con-
crete (Taylor et al., 2001; Collepardi, 2003; Martin, 2011).
Over the last decades, several approaches and recommendations, includ-
ing a variety of laboratory test procedures, have been developed worldwide
to assess the “potential” of ISR-­induced expansion and deterioration in con-
crete, along with the efficiency of preventive measures before concrete pour-
ing. Despite some issues with some of these test procedures, the majority of
experts agree that, in general, it is now possible to build new concrete infra-
structure with minimum or calculated risk of ISR. However, there is cur-
rently no consensus about the most efficient method(s) that should be
implemented and when for the rehabilitation of existing concrete structures
and structural members affected by ISR. In this context, numerous engineers
and researchers have been proposing appraisal techniques that are able to
determine both the damage’s cause and extent (i.e., diagnosis) and the
potential of further distress (i.e., prognosis) of ISR-­affected concrete, which
are essential steps in selecting efficient rehabilitation methods and optimum
application periods for affected concrete infrastructure.
The Institut Français des Sciences et Technologies des Transports, de
l’Aménagement et des Réseaux (Ifsttar 2003; now Gustave Eiffel University)
Bérubé et al. (2005), Fournier et al. (2010) and Godart et al., 2013 have
developed comprehensive management protocols for the diagnosis and
prognosis of ISR-­affected concrete structures. The proposed protocols are
based upon a series of comparative field and laboratory investigations to
confirm that ISR is the main cause (or a significant contributor to the overall
deterioration observed), thus aiming to select appropriate rehabilitation
strategies. Such investigations include one or several of the following steps
illustrated in Figure 1.1 and described next (Fasseu & Mahut, 2003; Bérubé
et al., 2005; Fournier et al., 2010; Godart et al., 2013):

• Routine field inspection of the structure under study to identify the

presence, distribution and severity of the defects affecting the various
structural members (especially the ISR-­related features), as well as the
exposure conditions to which the structure is subjected.
 Introduction 3

Figure 1.1 G eneral flowchart for management protocols of ISR-­a ffected structures.

• Preliminary in situ monitoring programme of deterioration (especially

signs of induced expansion and deformation) to quantify the progress
(i.e., rate) of deterioration on selected structural elements.
• Whenever appropriate (depending on the nature of deterioration and
importance of the structure), the implementation of a detailed investi-
gation programme including extensive in situ activities and laboratory
test procedures (i.e., petrographic characterization, chemical, physical
and mechanical tests) on samples collected from one or several com-
ponents of the ISR-­affected concrete structure.

Although the above protocols and respective steps are very detailed and
have been published for quite some time (i.e., 8–16 years), there is still a
lack of general understanding of the use of qualitative and quantitative in
situ and/or laboratory procedures to appraise the current and future con-
dition of ISR-­affected concrete. The goal of the present book is thus to
4 Internal Swelling Reactions in Concrete

thoroughly describe the most common ISR mechanisms leading to induced

expansion and deterioration along with presenting the state of the art of
the distinct approaches (i.e., visual inspection, non-­destructive tests, chemi-
cal procedures, microscopic analyses and mechanical techniques) commonly
used or with the potential to be implemented to appraise current and future
conditions of ISR-­affected structures; the ultimate aim of this book is to
provide engineers and infrastructure owners with tools for making better
decisions and selecting more appropriate strategies to cope with ISR in criti-
cal infrastructure/assets.

1.2 CHAPTER DESCRIPTIONS

This book is divided into ten chapters. Chapter 1 describes the importance
of the book, the scope and content of the work, along with the descrip-
tion of the chapters. Chapter 2 discusses in detail the most common ISR
mechanisms inducing expansion and deterioration in concrete. Chapter 3
presents the common practices to assess the condition in ageing concrete
infrastructure, introduces ISR as an ongoing damage mechanism in con-
crete, establishes the definition of damage within the context of this work
and discusses diagnosing ISR-­affected concrete. Chapter 4 presents visual
and non-­destructive tests (NDT) as promising tools to assess the condition
of ISR-­affected concrete, while Chapters 5 and 6 display in detail established
and promising microscopic and mechanical techniques, respectively.
Chapter 7 introduces the multi-­level assessment, a combination of micro-
scopic and mechanical techniques, to quantitatively assess the cause and
extent of damage in concrete affected by ISR, as well as discusses the impact
of different ISR mechanisms on the engineering properties and structural
responses of affected concrete members.
Chapter 8 presents the most common prognosis techniques and manage-
ment protocols used worldwide and discusses their positive aspects and
limitations. Chapter 9 displays a detailed case study from a Canadian over-
pass affected by ISR after nearly 50 years of service where the proposed
multi-­level assessment has been successfully implemented. Finally, Chapter
10 presents the conclusions of the book and discusses the need for future
work in the field.

REFERENCES

Bérubé, M.-A., Smaoui, N., Bissonnette, B., & Fournier, B. (2005). Outil d’évaluation
et de Gestion Des Ouvrages d’art Affectés de Réactions Alcalis-­ Silice (RAS).
http://www.mtq.gouv.qc.ca/portal/page/portal/Librairie/Publications/fr/ministere/
recherche/etudes/rtq0608.pdf
Canada Standards Association (CSA Group). (2019). CSA A23.1:19/CSA A23.2:19.
 Introduction 5

Collepardi, M. (2003). A state-­of-­the-­art review on delayed ettringite attack on

concrete. Cement and Concrete Composites, 25(4-­5 SPEC), 401–407. https://doi.
org/10.1016/S0958-­9465(02)00080-­X
Fasseu, P., & Mahut, B. (2003). Aide à La Gestion Des Ouvrages Atteints de Réactions
de Gonflement Interne. Guide Technique Des LPC (LCPC, 2003).
Fournier, B., & Bérubé, M. A. (2000). Alkali-­ aggregate reaction in concrete: A
review of basic concepts and engineering implications. Canadian Journal of Civil
Engineering, 27(2), 167–191. https://doi.org/10.1139/l99-­072
Fournier, B., Bérubé, M.-A., Folliard, K. J., & Thomas, M. (2010). Report on
Diagnosis, Prognosis and Mitigation of ASR in Transportation Structures. Federal
Highway Administration Publications FHWA-­HRT-­04-­113 (2004) and Techbrief
FHWA-­HRT-­06-­071, issued 2010.
Godart, B., De Rooij, M., & Wood, J. G. M. (2013). Guide to Diagnosis and
Appraisal of AAR Damage to Concrete in Structures – Part 1-­Diagnosis. RILEM
State-­of the-­Art Report (STAR), Springer, 2013.
Katayama, T., & Grattan-­Bellew, P. E. (2012). Petrography of Kingston Experimental
Sidewalk at Age 22 Years. ASR as the Cause of Deleteriously Expansive, so-­Called
Alkali-­Carbonate Reaction. In 14th ICAAR - International Conference on Alkali-­
Aggregate Reaction in Concrete.
Katayama, T. (2010). The So-­Called Alkali-­Carbonate Reaction (ACR) - Its mineral-
ogical and geochemical details, with special reference to ASR. Cement and Concrete
Research, 40(4), 643–675. https://doi.org/10.1016/j.cemconres.2009.09.020
Martin, R.-P. (2011). Analyse Sur Structures Modèles Des Effets Mécaniques de La
Réaction Sulfatique Interne Du Béton.
Noël, M., Sanchez, L., & Tawil, D. (2018). Structural implications of internal swell-
ing reactions in concrete: Review and research needs. Magazine of Concrete
Research, 70(20), 1052–1063. https://doi.org/10.1680/jmacr.17.00383
Sanchez, L.F.M., Drimalas, T., Fournier, B., Mitchell, D., & Bastien, J. (2018).
Comprehensive damage assessment in concrete affected by different Internal
Swelling Reaction (ISR) mechanisms. Cement and Concrete Research, 107. https://
doi.org/10.1016/j.cemconres.2018.02.017
Taylor, H. F.W., Famy, C., & Scrivener, K. L. (2001). Delayed Ettringite Formation.
Cement and Concrete Research 31(5), 683–693. https://doi.org/10.1016/S0008-
­8846(01)00466-­5
Chapter 2

Internal swelling reactions

(ISRs) mechanisms

2.1 INTRODUCTION: INTERNAL SWELLING

REACTIONS (ISRS)

ISRs are deterioration processes leading to induced expansion and damage

of affected concrete, usually in the presence of moisture. ISRs are generally
associated with a reduction in the mechanical properties, durability and long-­
term performance of affected concrete (Fournier & Bérubé, 2000; Martin,
2010; Sanchez, 2014). Among existing ISRs, alkali-­silica reaction (ASR) and
delayed ettringite formation (DEF) are likely the most observed worldwide
(Sanchez et al., 2018). However, other mechanisms inducing expansion and
damage, such as freeze and thaw cycles (FT), internal sulphate attack (ISA),
thaumasite formation (TF), physical sulphate attack (PSA) and hydration of
crystalline MgO and CaO, might also be classified as ISR.

2.2 ALKALI-­AGGREGATE REACTION (AAR)

AAR is a term used to describe chemical reactions between certain min-

eral phases from the aggregates used in concrete (fine and coarse) and the
alkali hydroxides (i.e., Na+, K+ and OH−) from the concrete pore solution.
The first recorded cases of AAR date from the early 1940s in California
(United States) by Thomas Stanton (1941); afterwards, numerous cases have
been identified worldwide. Nowadays, AAR is recognized as one of the most
harmful processes affecting the durability and long-­term performance of
concrete infrastructure in the field.
AAR-­induced expansion and damage in concrete are quite heterogeneous.
The reaction kinetics (i.e., induction period, expansion rate and ultimate
expansion) depends upon several parameters, such as temperature, alkali-­
loading of the concrete, type and nature (i.e., particle size and mineralogy)
of aggregates, presence of moisture and stress-­state. Nevertheless, three con-
ditions are simultaneously necessary to trigger AAR in concrete (Figure 2.1):
(a) the presence of reactive aggregates in the mixture, (b) high concentration

6 DOI: 10.1201/9781003188155-2
 Internal swelling reactions (ISRs) mechanisms 7

Figure 2.1 (a)

 Three conditions simultaneously required to trigger AAR. (b)
Concrete member from Robert-­B ourassa/Charest overpass built in
1966 in Quebec City, Canada, using an alkali-­silica reactive coarse
limestone aggregate.

of alkali hydroxides (i.e., NaOH, KOH) in the concrete pore solution and
(c) high humidity conditions. AAR often leads to extensive cracking, loss of
material’s physical integrity and, in some cases, the functionality of affected
structures or structural members.

2.2.1 Induced expansion mechanism

AAR can be divided into two main reaction types: ASR and alkali-­carbonate
reaction (ACR). ASR is by far the most common reaction type found around
the world, and its distress mechanism is already well understood, at least in
its major steps. It involves a chemical reaction between the alkali hydroxides
from the concrete pore solution and poorly crystalline or metastable silica
mineral forms in natural or synthetic aggregates. The reaction generates a
secondary product (the so-­called ASR-­gel) that induces expansive pressures
within the reacting aggregate material(s) and the adjacent cement paste upon
moisture uptake (Fournier & Bérubé, 2000). ASR-­induced development can
be divided into three major steps (Rajabipour et al., 2015): (a) interaction of
the alkaline environment and dissolution of the metastable siliceous mate-
rial, (b) formation and gelation of colloidal silica and (c) osmotic moisture
absorption and secondary products (i.e., ASR-­gel) expansion, as presented
in Equation 2.1.


a 
b 
c

( 2 )solid ( 2 )aqueous ( 2 )gel swelling of gel

SiO → SiO → SiO →
(2.1)

When metastable silica encounters the alkaline solution from the concrete
pore, a layer of cations (i.e., calcium, potassium and/or sodium) develops on
8 Internal Swelling Reactions in Concrete

its surface; such a layer tends to have a higher concentration of calcium,

which contributes to delaying the release of amorphous silica, while the dif-
fusion of alkali ions is not affected in the same magnitude. Nevertheless, the
metastable silica gradually breaks down and starts bonding with the soluble
alkali ions. The resulting product within the aggregate particles bears a
somewhat uniform composition, often characterized by a high Si/Ca ratio
and high concentrations of Na+ and K+ (Rodrigue et al., 2020). However,
once cracks extend from the reactive particles towards the cement paste, the
reaction products uptake calcium ions encountered at locations close to the
interface between the aggregates and the cement paste, which makes the
composition of reaction products change with an increase in calcium con-
tent, stiffness, viscosity, and swelling potential (Leemann & Lura, 2013).
On the other hand, ACR is a much less common deterioration process
whose mechanism is still mostly unknown. ACR is viewed by many
researchers as a reaction that occurs between the alkali hydroxides and
certain types of dolomitic limestones. However, the threshold between
ACR and ASR is unclear in most of the studies. It is generally agreed that
ACR is accompanied by the process of dedolomitization and formation of
calcite and brucite, displayed in Equation 2.2. This reaction reduces the
volume of solids by approximately 13% (Thomas & Folliard, 2007);
therefore, induced expansion is attributed to a parallel mechanism.
Theories have been proposed to explain the process of ACR-­ induced
expansion, such as (a) hydraulic pressure caused by the migration of water
molecules and alkaline ions, (b) absorption of alkaline ions and water
molecules on the surfaces of active clay materials scattered by dolomite
grains and (c) developments and rearrangement of dedolomitization prod-
ucts and formation and growth of crystalline products in confined spaces.
However, it has been claimed that ASR might also play a role in ACR-­
induced expansion and deterioration in concrete; petrographic examina-
tion made on ACR-­affected concrete specimens demonstrated the presence
of deleterious ASR-­gel from cryptocrystalline quartz along with harmless
dedolomitization products not associated with induced expansion and
crack formation (Katayama & Grattan-­Bellew, 2012; Katayama, 2010).
Although the aforementioned studies suggest a critical role of ASR in the
so-­called ACR, the induced expansion and deterioration caused by ACR
are significantly different from ASR in affected concrete (i.e., higher expan-
sion rate, mechanical property losses and different crack pattern) as veri-
fied by Sanchez et al. (2015).

CaMg ( CO3 )2 + 2 [ Na or K ] OH → CaCO


3

Dolomite Calcite

+ Mg ( OH )2 + [ Na or K ]2 CO3

Brucite (2.2)
 Internal swelling reactions (ISRs) mechanisms 9

2.2.2 AAR microscopic damage features in concrete

AAR damage development is often directly correlated to the level of induced
expansion triggered by the physicochemical mechanism. Yet, important
progress has been made over the last decades to better understand and
explain the impact of AAR-­induced microscopic damage features on the
mechanical properties of affected concrete.
ASR-­secondary products’ location and morphology depend on the miner-
alogical nature of the aggregates (Dunant & Scrivener, 2010). According to
the authors, two large classes of aggregate types can be distinguished: (a)
slowly reactive and (b) rapidly reactive aggregates. The distress caused by
slowly reactive aggregates, which are often used in conventional concrete, is
characterized by the formation of secondary products within the aggregates.
This phenomenon induces cracks inside the aggregate particles, which reach
the bulk cement paste with the increase in expansion. On the other hand,
rapidly reactive aggregates are more homogeneous in composition than
slowly reactive aggregates where ASR is mainly produced on the surface of
their particles; this results in cracks forming in the outer part of the aggre-
gate particles, which leads to cracking (and thus further damage) in the bulk
cement paste at early reaction levels (Giaccio et al., 2008). Bérard and Roux
(1986) suggested the existence of three types of ASR-­induced damage mech-
anisms as per distinct reactive rock types encountered in Quebec, Canada.
These types are illustrated in Figure 2.2:

• Peripheral reactions of non-­porous aggregates are shown in Figure

2.2a.
• Diffuse reactions cause the swelling of the bulk reactive aggregate par-
ticles, as shown in Figure 2.2b.
• Internal reactions lead to the formation of veins of ASR-­secondary
products, as displayed in Figure 2.2c.

According to Reinhardt and Mielich (2011), two different distress mech-

anisms are proposed for ASR-­induced development. The first mechanism
suggests that the dissolution process takes place at the aggregate particles’

Figure 2.2 A
 AR-­induced damage mechanisms as described by (Bérard & Roux,
1986): (a) peripheral reactions, (b) diffuse reactions and (c) internal
reactions.
10 Internal Swelling Reactions in Concrete

surfaces; thus, ASR-­secondary products and associated cracks are formed

at the interfacial transition zone (ITZ) and easily reach the bulk cement
paste upon swelling. The second approach states that cracks are generated
within the aggregate particles, reaching the cement paste with the increase
in induced expansion. This approach assumes that the critical aggregate
expansion must be achieved before new cracks are generated in the system;
moreover, a critical “crack length” should also be reached before cracks
are propagated within the aggregate particles, reaching the cement paste at
later stages. Furthermore, the ITZ is observed to remain intact in most cases
during ASR development, except in locations close to radial cracks formed
within the particles per Golterman (1995). Figure 2.3 illustrates common
ASR-­induced damage features.
Further investigations made by Sanchez et al. (2015) with a wide range of
reactive aggregate types (i.e., fine and coarse aggregates) and natures (i.e.,
lithotypes), as well as concrete mixtures (i.e., 25, 35 and 45 MPa), demon-
strated that at the beginning of the physicochemical reaction (e.g., low
expansion levels up to 0.05%), ASR-­induced cracks are mainly generated
within the aggregate particles. As the expansion progresses and for moder-
ate expansion levels (e.g., ± 0.12%), new cracks are still developed within
the aggregate particles; yet, the pre-­existing cracks keep growing in length
and width, with some cracks reaching the cement paste. At high expansion
levels (e.g., ± 0.20%), most of the previously generated cracks may already
be found in the cement paste. Furthermore, from this stage and onwards,
due to the “minimum energy law”, it is easier for ASR-­induced expansion to
keep increasing pre-­existing cracks than generating new cracks in the sys-
tem. Finally, at very high expansion levels (e.g., ≥0.30%), the cracks previ-
ously formed throughout the whole physicochemical process connect to one
another in the cement paste, forming an important cracking network that
directly impacts the mechanical properties of the affected concrete, espe-
cially its compressive strength. Figure 2.4 illustrates the qualitative and

Figure 2.3 A
 SR-­induced damage features: (a) scheme with a single aggregate par-
ticle and (b) affected concrete specimen (Golterman, 1995).
 Internal swelling reactions (ISRs) mechanisms 11

Figure 2.4 Q
 ualitative damage model based on distinct levels of expansion
(Sanchez et al., 2015).

descriptive damage model proposed by Sanchez et al. (2015) and described

earlier.
Conversely, ACR-­ induced expansion and damage development were
found to be quite different from ASR, at least in its microscopic damage
features (Sanchez et al., 2017). Furthermore, although there is no model
proposed to explain ACR-­induced crack development (i.e., cracks genera-
tion and propagation), it has been verified that very few cracks are observed
within the aggregate particles at low and moderate expansion levels. In con-
trast, important cracks are already verified in the cement paste from the
beginning of the physicochemical process. Moreover, minimal amounts of
reaction products are observed, even at high expansion levels. Therefore,
from an engineering and performance point of view, ACR should be consid-
ered as a different mechanism when compared to ASR, causing a distinct
impact on the mechanical response of affected concrete.

2.2.3 AAR influence on mechanical properties of

affected concrete
Evaluating AAR-­induced expansion and microscopic damage features dis-
cussed in Section 2.2.2 helps us to understand their associated impact on
the mechanical properties of the affected concrete. Although conventional
mechanical test procedures (i.e., compressive and tensile strengths, modulus
of elasticity (ME), etc.) are often used to assess the condition of concrete
structures affected by AAR, it is important to discuss the diagnostic charac-
ter of each of them prior to their use; hence, a proper understanding of the
impact of AAR on the distinct mechanical properties of affected concrete
as a function of its induced development is required (Sanchez et al., 2017).
Literature shows that the stiffness (i.e., ME) and tensile strength are the
most influenced mechanical properties of AAR-­ deteriorated concrete
(Sanchez et al., 2017). The ME is primarily governed by the mechanical
properties of the aggregates, especially the coarse aggregate (Mehta &
Monteiro, 2017). Therefore, the deterioration caused by AAR in the aggre-
gate particles is responsible for the significant decrease in ME of affected
12 Internal Swelling Reactions in Concrete

Figure 2.5 O
 verall mechanical properties (compressive strength, tensile strength
and modulus of elasticity) losses of ASR-­a ffected concrete specimens.

concrete. Results demonstrate that losses in ME may vary from 30% at low
expansion levels to 65% at high levels of expansion. Likewise, the tensile
strength of the concrete is much more affected by AAR than its compressive
strength (Giaccio et al., 2008). It has been verified that tensile strength
reductions of AAR-­affected concrete may drop by about 70%, even at low
or moderate levels of expansion (Sanchez et al., 2017). On the other hand,
compressive and direct shear strengths seem to be less affected, although
some contradictory results are often found, especially in shear (Souza et al.,
2019). Numerous experimental campaigns verified that compressive strength
is a mechanical property that is not considerably affected by AAR, at least
in its early stages. It has been reported that AAR-­affected structures display-
ing expansion levels lower than 0.10% may, in general, still efficiently with-
stand service loads. Moreover, significant reductions in compressive strength
are only observed for high and very high expansion levels (i.e., > 0.20%) as
per (Sanchez et al., 2017, 2018; Kubo & Nakata, 2012). Figure 2.5 illus-
trates the average values obtained by Sanchez et al. (2017) of compressive
strength, tensile strength and ME reductions as a function of AAR-­induced
expansion.

2.3 DELAYED ETTRINGITE FORMATION (DEF)

DEF is a form of ISA, and it is defined as the formation of ettringite in con-

crete after its setting (or at least a substantial portion of the whole setting)
without the contribution of any supplementary external sulphate source
from the surrounding environment (Taylor et al., 2001). Besides being con-
sidered as a type of ISA (described in Section 2.4.2), DEF can be distin-
guished due to the “nature” of ettringite formation. Usually, DEF takes place
when the concrete undergoes temperatures > 65°C in the very early hours
 Internal swelling reactions (ISRs) mechanisms 13

after pouring. Cement hydration and calcium silicate hydrate (C-­S-­H) for-
mation are greatly accelerated with the increase in the curing temperature.
With sustained temperatures above 65°C, ettringite becomes thermodynam-
ically unstable; therefore, previously formed ettringite decomposes, and the
sulphate ions return to the solution (Taylor et al., 2001), being absorbed by
C-­S-­H. Later, when SO42− is desorbed, the reformation of ettringite leads to
induced expansion and cracking. Figure 2.6 illustrates the appearance of
DEF-­affected columns after 15 years of service.
Although ettringite is commonly seen as homogeneously distributed into
the cement paste, DEF-­induced expansion and damage in concrete are quite
heterogeneous. The reaction kinetics (i.e., induction period, expansion rate
and ultimate expansion) depends upon several parameters, such as tempera-
ture, sulphate concentration of the cement, environmental conditions (i.e.,
temperature and humidity) and alkali-­loading of the concrete.

2.3.1 Induced expansion mechanism

To better understand the unique damage mechanism of DEF, a brief dis-
cussion on the hydration of the aluminate phases of the clinker (C3A and
C4AF), especially C3A, is required. The reaction of C3A with water is imme-
diate; crystalline hydrates, such as C3AH6, C4AH19 and C2AH8, are quickly
formed, releasing a large amount of heat. Gypsum is commonly added to
the anhydrate clinker to control such rapid hydration. In general, gypsum
and alkalis are quickly released into the concrete pore solution moments
after the contact between anhydrate cement and water. Hence, once calcium
and aluminum ions become available, there is the precipitation of calcium

Figure 2.6 T
 ypical appearance of columns affected by DEF after approximately 15
years (Thomas et al., 2008).
14 Internal Swelling Reactions in Concrete

aluminate trisulphate hydrate (AFt – ettringite). Ettringite is usually the first

hydrate to crystallize due to the high sulphate/aluminum ratio in the aque-
ous phase during the first hour of hydration. Later, after most of the sulphate
ions are consumed in the Aft’s formation, the aluminate concentration rises
again, and ettringite becomes unstable, being converted into calcium alumi-
nate monosulphate hydrate (AFm), the final product of Portland cement’s
hydration bearing over 5% of C3A.
DEF driven by curing temperatures above 65°C changes the concentra-
tion of sulphate and aluminum ions in the concrete pore solution, thus mod-
ifying the overall hydration process of the aluminate phases of the clinker.
Such high temperatures are normally achieved in steam-­ cured concrete
members or massive structures due to their significant temperature rise dur-
ing cement hydration. In these conditions, sulphate and aluminate ions tend
to be entrapped into C-­S-­H. Thus, once concrete temperature drops to ambi-
ent temperature at later stages, most of the aluminum ions remain firmly
bound into C-­S-­H, whereas SO42- is released in the concrete pore solution
and continues the overall ettringite formation. Since the concentration of
the aluminum ions in the solution is low, most of the ettringite is, therefore,
likely to be formed close to sources Al, such as AFm particles (Taylor et al.,
2001). This process develops important crystallization pressure in confined
spaces. Furthermore, since AFm is homogeneously distributed within the
cement paste, the formation of further ettringite leads to fairly homoge-
neous swelling of the cement paste, developing an extensive crack formation
in the affected concrete.

2.3.2 DEF microscopic damage features in concrete

The most acceptable DEF mechanism is characterized by a fairly homog-
enous expansion of the cement paste (in the presence of moisture), forming
circular or peripheral cracks (i.e., gaps, bands) around the aggregate par-
ticles, “detaching” them from the cement paste. Thus, “gaps” at the ITZ are
progressively generated and filled with large amounts of ettringite (Martin,
2010). Moreover, DEF-­induced swelling provides hydraulic tension inside
the aggregate particles and radial (in tension), and tangential (in compres-
sion) stresses in the cement paste. The radial tension presents its peak at the
ITZ; thus, cracks are often developed at these locations. Figure 2.7 presents
the common damage features generated by DEF in concrete. If the tensile
strength of the aggregate particle is lower than the tensile strength of the
ITZ, the fracture surface occurs at locations just inside the aggregate par-
ticle (outlining the particle’s shape; Goltermann, 1995).
According to Sanchez et al. (2020) DEF-­affected concrete displays already
at low and moderate expansion levels (i.e., ≤ 0.12%) an important number
of open cracks in the cement paste. At this level, although most of the cracks
are observed in the paste, some cracks can also be verified within the aggre-
gate particles, which might be due to the important stress concentration
 Internal swelling reactions (ISRs) mechanisms 15

Figure 2.7 C
 hemical/physical induced damage features (or aggregate shrinkage):
(a) scheme with a single aggregate particle and (b) affected concrete
specimen (Golterman, 1995).

taking place at the ITZ during DEF-­induced development. As the expansion

increases to high levels (i.e., up to 0.30%), the cracks previously generated
keep increasing (in length and width), but new cracks are also formed in the
system. Furthermore, some cracks are observed within the aggregate parti-
cles. At very high expansion levels (i.e., ≥ 0.30%), a major crack network is
observed in the cement paste, leading to an important physical integrity loss
of the affected material. For expansion levels greater than 0.50%, two new
phenomena are observed: debonding and disaggregation of the aggregate
particles, which increase, even more, the overall damage of the material.
Recent studies suggest the presence of different damage mechanisms in
concrete, such as FT, AAR (as displayed in Figure 2.8), wetting/drying cycles

Figure 2.8 C
 -­b eam specimen showing signs of DEF deterioration at Texas A&M
Riverside campus (Karthik et al., 2016).
16 Internal Swelling Reactions in Concrete

or even the introduction of mechanical loads, may increase the likelihood of

DEF-­induced development. Moreover, the combination of DEF and ASR is
chemically self-­supported, and thus, this combined mechanism is often
observed in the field. It has been found that ASR reaction products enable
the mobility of the ions and thus facilitate DEF-­ induced development.
Furthermore, numerous authors believe that DEF and ASR coupling occurs
because the consumption of alkalis due to ASR development to form ASR-­
secondary products contributes towards DEF development (Thomas et al.,
2008; Martin, 2010). Finally, although quantifying the actual contribution
of each mechanism towards the overall deterioration process of the concrete
is not an easy task, it is commonly agreed that DEF plays a major role in the
combined deterioration process (Thomas et al., 2008).

2.3.3 DEF influence on mechanical properties of

affected concrete
Evaluating DEF-­ induced expansion and microscopic damage features
discussed in Section 2.3.2 helps the understanding of its impact on the
mechanical properties of the affected concrete. According to Sanchez
et al. (2018), at expansion levels up to 0.12%, DEF-­affected concrete
may display a ME 50% lower than sound concrete. Likewise, the tensile
strength of the concrete may drop about 70%, even at low or moderate
levels of expansion, while the compressive strength reduction may lessen
by 10%. The explanation for the low reduction in compressive strength is
that (1) the cracks are highly localized at the ITZ and (2) upon loading,
existing cracks are partially stopped as per the so-­called arrest mechanism
provided by the aggregates in the system (Mindess et al., 2003). As the
expansion level rises (i.e., up to 0.30%), the ME decreases even more, but
at a lower rate (down to about 60%). On the other hand, the compressive
strength reduction significantly increases, reaching values of about 40%.
At this stage, it is possible to note a trend where some of the ITZ cracks
formed at distinct locations start linking to one another, which creates
an important net cracking at the cement paste and decreases the physical
integrity of the affected material. For expansion levels higher than 0.50%,
two new phenomena start taking place: debonding and disaggregation
of the aggregate particles, which accelerate once more the drop in stiff-
ness (i.e., ME) of the affected concrete down to values of about 85%.
Likewise, the compressive strength drop reaches values of about 50% at
this stage. Yet, the cracking extension process at this level of expansion
does not seem to further affect the tensile strength of the concrete since
these cracks have already reached their “critical length” to cause tensile
failure at low to moderate levels of expansion. Figure 2.9 illustrates the
average compressive strength, tensile strength and ME losses as a func-
tion of DEF-­induced expansion of affected concrete (Sanchez et al., 2018;
Yammine et al., 2020; Karthik et al., 2016).
 Internal swelling reactions (ISRs) mechanisms 17

Figure 2.9 O
 verall mechanical properties (compressive strength, tensile strength
and modulus of elasticity) losses of DEF-­a ffected concrete specimens.

2.4 OTHER MECHANISMS

2.4.1 Freezing and thawing cycles

It is well established that porous materials containing moisture are suscep-
tible to deterioration under repeated FT. Freezing water has proven to be
quite deleterious to concrete, yet several factors can influence the FT mecha-
nism. Amongst them, the main factors are (a) aggregate features (i.e., par-
ticle size, porosity and permeability), (b) the presence of entrained air, (c)
the water-­to-­cement ratio and (d) the degree of saturation of the concrete.
Hydraulic pressure is generated when water freezes in the concrete capillar-
ity pores, causing induced expansion since frozen water swells at about 9%
of its original volume. Thus, repeatable FT cycles may cause severe physical
damage in concrete, such as internal cracking and surface scaling. Figure
2.10 illustrates a concrete sidewalk deteriorated by FT damage.

2.4.1.1 Induced expansion mechanism

When ice forms in the capillary pore structure of concrete, it can lead to
induced expansion and microcracking mainly caused by (a) hydraulic pres-
sure generation due to ice formation, (b) osmotic pressure generation due to
solute concentration increase in the pore solution adjacent to freezing sites,
(c) water desorption from C-­S-­H and (d) ice segregation. The most common
processes presented in the literature are the first two, hydraulic (caused by
the increase in the specific volume of water through freezing in large cavi-
ties) and osmotic pressure (due to salt concentration differences in the pore
water/solution; Mindess et al., 2003). It is generally believed that osmotic
pressure predominantly damages concrete exposed to saltwater (e.g., from
18 Internal Swelling Reactions in Concrete

Figure 2.10 (a) Picture of a severely scaled sidewalk. (b) Sidewalk that was
severely scaled, which led to a complete loss of mechanical integrity,
probably by internal frost damage (Valenza & Scherer, 2006).

de-­icing salts, seawater), while hydraulic pressure governs frost deteriora-

tion whether no external saltwater is present (Fagerlund, 1994).
Water in the capillary pores of concrete is not pure; it contains several
soluble substances, such as alkalis, chlorides and calcium hydroxide.
Moreover, it is quite common to find different salt concentrations in the bulk
concrete, causing gradient osmotic pressures (Mehta & Monteiro, 2017).
Likewise, it is observed that the exposed surface of concrete members cools
faster and freezes first. As freezing occurs in the first surface layer of the con-
crete, water can no longer move through the frozen region. Therefore, con-
tinued freezing accompanied by expansion forces water away from the
freezing front into the concrete. The result is the pressure exceeding the ten-
sile strength of the cement paste, causing microcracking. This hypothesis also
explains why FT-­induced deterioration in concrete often occurs with the
presence of cracks parallel to the exposed surface (Powers, 1945; Deschenes
Jr., 2017).
As previously mentioned, several factors may influence FT-­induced mech-
anisms. For instance, the aggregates’ features (i.e., size, porosity and perme-
ability) used in concrete are essential in the damage process. According to
Verbeck and Landgren (1996), the aggregates used in concrete can be
divided into three categories: (a) first-­class aggregates (i.e., low permeabil-
ity), which are normally not harmed by FT cycles; (b) second-­class aggre-
gates (i.e., average permeability), where their particle size is critical since the
larger the particle size, the greater the risk of FT damage; and (c) third-­class
aggregates (i.e., with high permeability), which enable easy water penetra-
tion, inducing FT damage in the ITZ rather than within their particles.
Air-­entraining admixtures are widely known as FT inhibitors; normally,
air-­entraining admixtures are selected as a volume percentage (%) of the con-
crete. However, the most important concept behind the use of air entrainment
 Internal swelling reactions (ISRs) mechanisms 19

is not the total volume (%) of entrained air but rather the incorporation of a
given amount of round empty voids (which act as valves) spaced between
0.1 and 0.2 mm from one another in the hardened cement paste. However,
for practical reasons, the entrained air in (%) is the parameter typically
required while mix-­proportioning concrete mixtures. Values ranging from
3% to 7% of air, depending on the maximum aggregate size and the sur-
rounding environment, are often selected for concrete mixtures in cold
regions. These values are considered very effective in proportion to risk-­free
FT mixtures, especially if aggregates which are not FT susceptible are used.
In general, when the concrete water-­to-­cement ratio increases, the mechan-
ical properties of the material decrease due to the rise in overall porosity
(i.e., mainly capillary pores). Since water freezes primarily in the concrete
pores, the higher the concrete porosity, the higher its likely susceptibility to
FT deterioration. However, higher mechanical properties do not necessarily
lead to proper FT performance since the presence of air entrainment agents
is crucial. Therefore, air-­entrained conventional concrete mixtures are
expected to behave better against FT-­induced deterioration than non-­air-­
entrained, high-­performance concrete (Mehta & Monteiro, 2017).

2.4.1.2 FT microscopic damage features in concrete

Following Sanchez et al. (2020) findings, it has been observed that at the
beginning of FT-­induced development and for low expansion levels (i.e.,
0.05%), cracks are mainly formed in the ITZ or the bulk cement paste/pores.
Moreover, at this expansion level, a few cracks may also be observed within
the aggregate particles. As the expansion level increases and for moderate
expansion levels (i.e., 0.12%), the cracks previously formed at the begin-
ning of the deterioration process increase in length and width, and some
new cracks are also generated in the cement paste (i.e., ITZ or bulk-­paste/
pores) and aggregate particles. For high expansion levels (i.e., 0.20%), the
distress mechanism keeps progressing, either through the increase of pre-­
existent cracks or by forming new cracks in the cement paste. Otherwise, the
generation of cracks within the aggregate particles seems to stop completely.
Finally, when the expansion level reaches very high degrees (i.e., 0.30% and
beyond), most of the cracks present in the cement paste link to one another,
resulting in an important crack network formation.

2.4.1.3 FT influence on mechanical properties of affected

concrete
Evaluating FT-­induced expansion and microscopic damage features helps
to understand their associated impact on the mechanical properties of the
affected concrete. According to Sanchez et al. (2018), for low expansion lev-
els, a significant amount of cracks are found within the aggregate particles,
20 Internal Swelling Reactions in Concrete

yet with much less intensity than for ASR-­affected specimens, for instance.
Therefore, the FT-­affected specimens may experience a decrease of 35%
in ME. Comparably to ASR and DEF, FT-­affected concrete displays tensile
strength reductions down to 70% at low expansion amplitudes. The fail-
ure mechanism in tension is indeed a direct and brittle mechanism dictated
by “fracture mechanics”, where “stress concentrations” are generated in
the presence of micro defects/pores of concrete due to FT-­induced devel-
opment. Likewise, since FT cracks are mainly formed within the cement
paste, a much more significant drop in compressive strength, reaching values
between 15% and 30% of loss, is observed at low expansions. At moder-
ate expansion levels (i.e., 0.12%), the ME keeps dropping down to values
of about 40%, while compressive strength reductions reach values between
20% and 35%, and the tensile strength keeps dropping, yet at a lower rate.
As mentioned in Section 2.4.1.2, from expansion amplitudes of 0.12% up
to 0.20%, the damage generation in the cement paste keeps progressing,
but the cracks within the aggregate particles stop being generated. This dis-
tress feature continues at about 0.30% of expansion; thus, the drop in the
modulus reaches values of about 50%. The tensile strength loss at this phase
seems similar to that obtained at moderate expansion levels, which means
that since the tensile strength is extremely dependent on the crack length
(i.e., maximum length needed to cause the crack’s propagation leading to
failure), its threshold was likely already attained at lower expansion lev-
els. In addition, the compressive strength drop reaches about 40%, being
more significant than ASR and almost the same as DEF for the same expan-
sion levels. Figure 2.11 illustrates the average compressive strength, tensile
strength and ME losses in the function of the expansion amplitudes of FT-­
affected concrete (Sanchez et al., 2018).

Figure 2.11 O
 verall mechanical properties (compressive strength, tensile strength
and modulus of elasticity) losses of FT-­a ffected concrete specimens.
 Internal swelling reactions (ISRs) mechanisms 21

2.4.2 Internal sulphate attack (ISA)

The term sulphate attack comprises complex processes of chemical reac-
tions between sulphate ions (SO42−) and hydrated products from the cement
paste in the presence of high moisture. Overall, these processes include salt
crystallization (i.e., physical attack) and chemical attack by sulphate ions
from internal (i.e., contaminated materials, sulphate-­bearing aggregates or
binder materials) or external (i.e., soil, groundwater, seawater, etc.) sources
in the concrete. Nevertheless, besides a few particularities of the differ-
ent named mechanisms of sulphate attack, the global chemistry involving
SO42− anions in these processes are quite similar; its swelling behaviour is
commonly attributed to the formation of gypsum and secondary ettringite.
Equation 2.3 illustrates the formation of gypsum from the reaction of sul-
phate ions and portlandite and the formation of secondary ettringite from
its further reaction with calcium monosulfoaluminate hydrate (AFm):

SO24− + Ca ( OH )2 + 2H 2O → CaSO4 .2H 2O+ 2OH −

Gypsum
Portlandite (2.3)

2CaSO4 .4H 2O+ 4CaO.Al2O3.SO3.12H 2O

Gypsum AFm
+ 16H 2O → 6CaO.Al2O3.3SO3.32H 2O
Ettringite
(2.4)

It is worth noting that ettringite is a common product that originates from

cement hydration; its formation only becomes harmful when it occurs after
the concrete’s setting. The formation of gypsum and secondary ettringite
commonly yields changes in volume by about 1.2 to 2.2 times higher than
the reactant products, leading to an increase in the internal pressure on its
surroundings and causing volumetric deformation, cracking, spalling and
mechanical properties reductions. Moreover, the magnitude of the induced
damage development depends upon several parameters, such as sulphate
concentration, humidity, exposure and curing temperature, along with the
cation associated with SO42−.
Most standards worldwide suggest restrictive limits on both sulphur tri-
oxide and C3A contents of cement; the purpose of this double restriction is
to prevent internal/external sulphate attack. Such restriction was recognized
when it was observed that extensive induced expansion might occur in the
absence of an external source of sulphate. By definition, ISA takes place
when there is a delayed release of sulphates in the hardened cement paste or
in the presence of iron sulphide-­bearing aggregates (e.g., aggregates bearing
pyrrhotite, pyrite). In general, pyrrhotite and pyrite minerals are oxidized
(as shown in Equations 2.5 and 2.6, respectively) in the presence of water
22 Internal Swelling Reactions in Concrete

and oxygen, producing ferrous ions (Fe2+) and sulphate (SO42−), and releas-
ing hydrogen ions (H+) into the cement paste. Besides the further reaction of
the SO42− to form gypsum and secondary ettringite, the oxidation of pyr-
rhotite/pyrite leads to a decrease in the pH of the concrete pore solution,
accelerating the oxidation of the minerals (Dobrovolski et al., 2021).
Moreover, the oxidation rate of sulphide minerals increases directly with
increasing relative humidity to values up to 80% (Steger, 1982).

Fe1− xS + ( 2 − x / 2 ) O2 + xH 2O → (1 − x ) Fe2 + + SO24 − + 2 xH +

(2.5)

FeS2 + 7 / 2O2 + H 2O → Fe2+ + 2SO42− + 2H + (2.6)

Concrete deterioration due to the combined effects of the oxidation of

iron sulphides (e.g., in sulphide-­bearing aggregates as seen in Figure 2.12a)
and ISA in the cement paste (i.e., due to the release of sulphide ions) gener-
ates secondary minerals with high swelling properties, causing expansion
development in both concrete microstructure phases, aggregates and cement
paste. The expansion in the aggregate particles is commonly attributed to
the oxidation of iron, similar to steel corrosion in reinforced concrete.
Petrographic evaluations, illustrated in Figure 2.12b, demonstrate that
important cracks are observed around and through the aggregate particles
towards the cement paste of affected concrete (Rodrigues et al., 2012).
Moreover, the aggregate particles are completely covered with iron oxy-­
hydroxide and are surrounded by a whitish halo; the latter was verified as
being a region rich in ettringite.
Furthermore, changing humidity and temperature can either intensify/
accelerate the damage development or even change the mechanism of ISA.

Figure 2.12 F eatures of concrete deterioration caused by iron sulphide-­b earing

aggregates: (a) Cracking in a house's foundations. Open cracks are
typically more pronounced at the corners of the foundation blocks,
often next to rain gutters. Yellowish surface coloration is often seen
on the exposed foundation walls. (b) Stereomicroscopic views of a
deteriorated concrete foundation block (Rodrigues et al., 2012).
 Internal swelling reactions (ISRs) mechanisms 23

For instance, high-­temperature conditions during the hardening process of

the concrete can lead to further development of DEF (described in Section
2.3), whereas low temperature conditions can lead to TF, whose mechanism
will be further discussed in the following section.

2.4.3 Thaumasite formation (TF)

In recent years, sulphate attack caused by TF has received considerable
attention. TF involves sulphate and carbonate ions in the presence of high
humidity and “preferentially” low temperature (< 15°C). However, TF has
already been reported at higher temperatures (Bassuoni & Nehdi, 2009).
Differently from the pyrrhotite/pyrite oxidation or DEF, TF mainly “targets”
the SO42− ions that are present in the C-­S-­Hs (the main binding phase in
concrete) rather than the ones present in the calcium hydroxide and cal-
cium aluminate phases. Equation 2.7 demonstrates thaumasite chemical
formation. In severe cases of TF, besides the intense expansion development
proportioned by the sulphate attack and ettringite formation, the hardened
cement paste is expressively replaced by thaumasite, transforming the con-
crete into a white and incohesive mush (Skalny et al., 2002).

3CaO  2SiO2 3H 2O+ 2Ca ( OH )2 + 2SO24 − + 2CaCO3

C −S − H Portlandite
+ 24H 2O → 2 ( CaSiO3  CaCO3  CaSO4 15H 2O )
Thaumasite
(2.7)

It is worth mentioning that TF may occur in any type of sulphate attack

as long as carbonate ions and preferably low temperatures are included in
the process. Moreover, it is unlikely that thaumasite occurs without the for-
mation of ettringite. Most case studies of TF-­related deterioration show that
thaumasite and ettringite coexist in the microstructure of concrete (Rahman
& Bassuoni, 2014). According to the literature, there are two possible
“routes” for TF: direct and indirect. In the former, sulphate ions react with
carbonate (either CO32− ions or atmospheric CO2) and C-­S-­H to form thau-
masite, as demonstrated in Equation 2.7, whereas in the indirect “route,”
ettringite is a precursor for TF (Bensted, 2003). In other words, the damage
development taking place in TF-­affected concrete tends first to follow the
primary source of SO42− release; however, with the worsening factor of
expressive strength loss due to the deterioration of C-­S-­H. Finally, it is worth
mentioning that the mechanism of TF is still under debate in the scientific
community; therefore, an in-­depth investigation of TF conditions is still
required and could shed some light on preventive measures against TF-­
induced development.
24 Internal Swelling Reactions in Concrete

2.4.4 Physical sulphate attack (PSA)

The so-­called PSA, also called sulphate salt crystallization or salt hydration
distress, usually refers to repeated cycles of recrystallization of mirabilite
(Na2SO4·10H2O) into thenardite (Na2SO4), which lays on the constant vari-
ation of both humidity and temperature. This temperature-­dependent pro-
cess leads to repeated volume variation, illustrated by Figure 2.13, leading
to fatigue of the cement paste and its subsequent loss of cohesion (Skalny
et al., 2002). First, high moisture conditions solubilize sodium sulphate salts
(internally or externally to concrete), which facilitates its transport and dif-
fusion through concrete. Afterwards, once the exposure conditions change
and the water starts evaporating, the sodium and sulphate ions “reattach”
to one another, forming sodium sulphate decahydrate (Na2SO4∙10H2O).
Moreover, the continuous changes in temperature or humidity over time are
followed by the sodium sulphate anhydrite (Na2SO4) and vice versa.
In a PSA, the involvement of the hydrated Portland cement products is not
“mandatory”. However, the damage mechanism of PSA and whether it is
independent of chemical sulphate attack is still controversial. The reaction
and crystal formation commonly take place in any empty flaws within the
concrete, such as pre-­existent cracks and further developed cracks, air voids,
pores in the cement paste or aggregates, at the ITZ. Scaling and flaking of
the concrete’s surface are typical signs of PSA because salts tend to crystal-
lize in pores around the evaporation surfaces. Figure 2.14 illustrates deterio-
ration due to PSA.

2.4.5 Hydration of crystalline MgO and CaO

The expansive effect of crystalline magnesium oxide hydration in concrete
was recognized more than a hundred years ago (i.e., 1884) when the first
cases were reported in France and Germany, while it was quite common
to find cement containing about 25% of MgO in its chemical composition

Figure 2.13 T
 emperature-­d ependent phase diagram for sodium sulphate (Tsui
et al., 2003).
 Internal swelling reactions (ISRs) mechanisms 25

Figure 2.14 M
 echanism of PSA damage with (a) specimen placed in partial immer-
sion, (b) salts wick up through the concrete and deposit on its sur-
face and (c) resulting surface damage (Alyami et al., 2019).

(Mehta & Monteiro, 2017). Nowadays, different standards (e.g., ASTM

C150, 2018) require that the MgO content in cement shall not exceed
6%. However, MgO expansive agents are commonly used to compensate
for temperature shrinkage of mass concrete (Mo et al., 2014). Although
approximately 118% of solid volume increase is caused when MgO trans-
forms into Mg(OH)2, the corresponding induced expansion is not precisely
equal to the increase of solid volume. Overall, MgO hydrates slowly, and
it is associated with delayed induced expansion in concrete through the
formation of Mg(OH)2 (i.e., brucite). There are several theories developed
to explain the expansion mechanism of MgO (Mo et al., 2014); the two
main common theories are (a) crystal growth pressure due to the growth of
Mg(OH)2 and (b) swelling pressure due to water absorption by very small
crystals of Mg(OH)2. During the hydration of MgO, a supersaturated solu-
tion of Mg2+ and OH− are firstly formed; then Mg(OH)2 crystals precipitate
and grow in a confined region, resulting in crystal growth pressure leading
to induced expansion (Chatterji, 1995). Nevertheless, the most important
factors affecting the expansion of MgO concrete are the dosage and reac-
tivity of MgO and temperature (i.e., the higher the curing temperature, the
higher the hydration kinetics of MgO).
The harmful effect due to the hydration of partially crystalline CaO
(poorly burnt from CaCO3) was reported in the 1930s in the United States.
Although the presence of a high volume of partially crystalline CaO in
cement can cause a severe induced expansion in concrete, this phenomenon
is virtually unseen in modern concrete construction due to the better manu-
facturing protocols of a Portland cement clinker. However, the driving force
leading to induced expansion while the hydration of partially crystalline
CaO is the crystal growth pressure generated after later Ca(OH)2 formation
26 Internal Swelling Reactions in Concrete

under restricted conditions near the surfaces of poorly burnt lime grains
(Chatterji, 1995). Therefore, the solid volume of Ca(OH)2 formed (from the
later hydration of CaO) doubles the initial volume of CaO and thus, expan-
sive forces are generated, causing the swelling of the cement paste and fur-
ther cracking. Yet, the severity of the induced expansion may be decreased
in low alkali systems, likely due to the release of calcium ions in the concrete
pore solution (Min et al., 1996).

REFERENCES

Alyami, M. H., Alrashidi, R. S., Mosavi, H., Almarshoud, M. A., & Riding, K. A.
(2019). Potential accelerated test methods for physical sulphate attack on con-
crete. Construction and Building Materials 229, 116920. https://doi.org/10.1016/
j.conbuildmat.2019.116920
ASTM C150. 2018). Standard Specification for Portland Cement, issued 2018.
Bassuoni, M. T., & Nehdi, M. L. (2009). Durability of self-­ consolidating con-
crete to different exposure regimes of sodium sulphate attack. Materials and
Structures/Materiaux et Constructions 42, 1039–1057. https://doi.org/10.1617/
s11527-­008-­9442-­2
Bensted, J. (2003). Thaumasite – direct, woodfordite and other possible formation
routes. Cement and Concrete Composites 25(8), 873–77. https://doi.org/10.1016/
S0958-­9465(03)00115-­X
Bérard, J., & Roux, R. (1986). La Viabilité Des Bétons Du Québec: Le Rôle Des
Granulats. Canadian Journal of Civil Engineering 13(1), 12–24. https://doi.org/
10.1139/l86-­003
Chatterji, S. (1995). Mechanism of expansion of concrete due to the presence of
dead-­burnt CaO and MgO. Cement and Concrete Research 25(1), 51–56. https://
doi.org/10.1016/0008-­8846(94)00111-­B
Deschenes Jr. R. A. (2017). Mitigation and Evaluation of Alkali-­Silica Reaction
(ASR) and Freezing and Thawing in Concrete Transportation Structures.
ProQuest Dissertations and Theses. Ann Arbor: University of Arkansas. https://
login.proxy.bib.uottawa.ca/login?url=https://www.proquest.com/dissertations-­
theses/mitigation-­e valuation-­a lkali-­s ilica-­r eaction-­a sr/docview/1949666139/
se-­2?accountid=14701
Dobrovolski, M. E. G., Munhoz, G. S., Pereira, E., & Medeiros-­Junior, R. A. (2021).
Effect of crystalline admixture and polypropylene microfiber on the internal sul-
phate attack in portland cement composites due to pyrite oxidation. Construc­
tion and Building Materials 308(September), 125018. https://doi.org/10.1016/
j.conbuildmat.2021.125018
Dunant, C. F., & Scrivener, K. L. (2010). Micro-­mechanical modelling of alkali-­silica-­
reaction-­induced degradation using the AMIE framework. Cement and Concrete
Research 40(4), 517–25. https://doi.org/10.1016/j.cemconres.2009.07.024
Fagerlund, G. (1994. The Water Absorption Process in the Air-­Pore System WATER
ABSORPTION PROCESS IN THE AIR-­PORE SYSTEM Report TVBM-­7085.
Fournier, B., & Bérubé, M.-A. (2000). Alkali-­ aggregate reaction in concrete: A
review of basic concepts and engineering implications. Canadian Journal of Civil
Engineering 27(2), 167–91. https://doi.org/10.1139/cjce-­27-­2-­167
 Internal swelling reactions (ISRs) mechanisms 27

Giaccio, G., Zerbino, R., Ponce, J. M., & Batic, O. R. (2008). Mechanical behavior of
concretes damaged by alkali-­silica reaction. Cement and Concrete Research 38(7),
993–1004. https://doi.org/10.1016/j.cemconres.2008.02.009
Goltermann, P. (1995). Mechanical predictions of concrete deterioration; Part 2:
Classification of crack patterns. ACI Materials Journal 92(1), 1–6. https://doi.
org/10.14359/1177
Karthik, M. M., Mander, J. B., & Hurlebaus, S. (2016). Deterioration data of a
large-­scale reinforced concrete specimen with severe ASR/DEF deterioration.
Construction and Building Materials 124, 20–30. https://doi.org/10.1016/
j.conbuildmat.2016.07.072
Katayama, T. (2010). The So-­Called Alkali-­Carbonate Reaction (ACR) - Its mineral-
ogical and geochemical details, with special reference to ASR. Cement and Concrete
Research 40(4), 643–75. https://doi.org/10.1016/j.cemconres.2009.09.020
Katayama, T., & Grattan-­ Bellew, P. E. (2012). Petrography of the Kingston
Experimental Sidewalk at Age 22 Years–ASR as the Cause of Deleteriously
Expansive, So-­ Called Alkali-­Carbonate Reaction. In Proceedings of the 14th
International Conference on Alkali-­ Aggregate Reaction in Concrete. Austin,
Texas, 10.
Kubo, Y., & Nakata, M. (2012). Effect of reactive aggregate on mechanical proper-
ties of concrete affected by alkali-­silica reaction. In 14th International Conference
on Alkali-­Aggregate Reaction in Concrete, electronic. Austin (Texas).
Leemann, A., & Lura, P. (2013). E-­Modulus of the alkali-­silica-­reaction product
determined by micro-­indentation. Construction and Building Materials 44, 221–
27. https://doi.org/10.1016/j.conbuildmat.2013.03.018
Martin, R. P. (2010). Analyse Sur Structures Modèles Des Effets Mécaniques de La
Réaction Sulfatique Interne Du Béton. Université Paris-­Est.
Mehta, P. K., & Monteiro, P. J. M. (2017. “CONCRETE Microstructure, Properties
and Materials.”
Min, D., Dongwen, H., Xianghui, L., & Mingshu, T. (1996). Mechanism of expan-
sion in hardened cement pastes with hard-­ burnt lime. Cement and Concrete
Research 26(4), 647–48. https://doi.org/10.1016/0008-­8846(96)00021-­x
Mindess, S., Young, J. F., & Darwin, D. (2003). Concrete. Prentice-­ Hall Civil
Engineering and Engineering Mechanics Series. Prentice-­ Hall. https://books.
google.ca/books?id=38VoQgAACAAJ
Mo, L., Deng, M., Tang, M., & Al-­Tabbaa, A. (2014). MgO expansive cement and
concrete in China: Past, present and future. Cement and Concrete Research 57,
1–12. https://doi.org/10.1016/j.cemconres.2013.12.007
Powers, T. C. (1945). A working hypothesis for further studies of frost resistance of
concrete. ACI Journal Proceedings 41(1). https://doi.org/10.14359/8684
Rahman, M. M., & Bassuoni, M. T. (2014). Thaumasite sulphate attack on con-
crete: Mechanisms, influential factors and mitigation. Construction and Building
Materials 73(12), 652–62. https://doi.org/10.1016/j.conbuildmat.2014.09.034
Rajabipour, F., Giannini, E., Dunant, C., Ideker, J. H., & Thomas, M. D. A. (2015).
Alkali-­Silica reaction: Current understanding of the reaction mechanisms and
the knowledge gaps. Cement and Concrete Research 76(6), 130–46. https://doi.
org/10.1016/j.cemconres.2015.05.024
Reinhardt, H. W., & Mielich, O. (2011). A fracture mechanics approach to the crack
formation in alkali-­sensitive grains. Cement and Concrete Research. https://doi.
org/10.1016/j.cemconres.2010.11.008
28 Internal Swelling Reactions in Concrete

Rodrigue, A., Duchesne, J., Fournier, B., Champagne, M., & Bissonnette, B. (2020).
Alkali-­Silica reaction in Alkali-­activated combined slag and fly ash concretes: The
tempering effect of fly ash on expansion and cracking. Construction and Building
Materials 251. https://doi.org/10.1016/j.conbuildmat.2020.118968
Rodrigues, A., Duchesne, J., Fournier, B., Durand, B., Rivard, P., & Shehata, M.
(2012). Mineralogical and chemical assessment of concrete damaged by the oxida-
tion of sulfide-­bearing aggregates: Importance of thaumasite formation on reac-
tion mechanisms. Cement and Concrete Research 42(10), 1336–1347. https://doi.
org/10.1016/j.cemconres.2012.06.008
Sanchez, L. F.M., Drimalas, T., Fournier, B., Mitchell, D., & Bastien, J. (2018).
Comprehensive damage assessment in concrete affected by different Internal
Swelling Reaction (ISR) mechanisms. Cement and Concrete Research 107
(February), 284–303. https://doi.org/10.1016/j.cemconres.2018.02.017
Sanchez, L. F. M., Drimalas, T., & Fournier, B. (2020). Assessing condition of con-
crete affected by Internal Swelling Reactions (ISR) through the Damage Rating
Index (DRI). Cement 1–2(September), 100001. https://doi.org/10.1016/j.cement.
2020.100001
Sanchez, L. F. M., Fournier, B., Jolin, M., & Duchesne, J. (2015). Reliable quan-
tification of AAR damage through assessment of the Damage Rating Index
(DRI). Cement and Concrete Research 67(1), 74–92. https://doi.org/10.1016/j.
cemconres.2014.08.002
Sanchez, L. F. M., Fournier, B., Jolin, M., Mitchell, D., & Bastien, J. (2017). Overall
assessment of Alkali-­Aggregate Reaction (AAR) in concretes presenting differ-
ent strengths and incorporating a wide range of reactive aggregate types and
natures. Cement and Concrete Research 93, 17–31. https://doi.org/10.1016/j.
cemconres.2016.12.001
Sanchez, L. F. M. (2014). Contribution to the assessment of damage in aging con-
crete infrastructures affected by alkali-­aggregate reaction. 341.
Skalny, J., Marchand, J., & Odler, I. (2002). Sulphate attack on concrete (1st ed.).
London and New York: Spon Press.
Souza, D. J. De, Sanchez, L. F. M., & De Grazia, M. T. (2019). Evaluation of a direct
shear test setup to quantify AAR-­induced expansion and damage in concrete.
Construction and Building Materials 229, 116806. https://doi.org/10.1016/j.
conbuildmat.2019.116806
Steger, H. F. (1982). Oxidation of sulfide minerals. VII. Effect of temperature and
relative humidity on the oxidation of pyrrhotite. Chemical Geology 35(3–4), 281–
295. https://doi.org/10.1016/0009-­2541(82)90006-­7
Taylor, H. F. W., Famy, C., & Scrivener, K. L. (2001). Delayed ettringite forma-
tion. Cement and Concrete Research 31(5), 683–693. https://doi.org/10.1016/
S0008-­8846(01)00466-­5
Thomas, M. D. A., & Folliard, K. J. (2007). Concrete aggregates and the durability
of concrete. Durability of Concrete and Cement Composites, 247–281. https://doi.
org/10.1533/9781845693398.247
Thomas, M., Folliard, K., Drimalas, T., & Ramlochan, T. (2008). Diagnosing delayed
ettringite formation in concrete structures. Cement and Concrete Research 38(6),
841–847. https://doi.org/10.1016/j.cemconres.2008.01.003
Tsui, N., Flatt, R. J., & Scherer, G. W. (2003). Crystallization damage by sodium
sulphate. Journal of Cultural Heritage 4(2), 109–115. https://doi.org/10.1016/
S1296-­2074(03)00022-­0
 Internal swelling reactions (ISRs) mechanisms 29

Valenza, J. J., & Scherer, G. W. (2006). Mechanism for salt scaling. Journal of the
American Ceramic Society 89(4), 1161–1179. https://doi.org/10.1111/j.1551-
­2916.2006.00913.x
Verbeck, G., & Landgren, R. (1996). Influence of physical characteristics of aggre-
gates on frost resistance of concrete. In Proceedings of the American Society for
Testing and Materials (ASTM ’96), 1063–1079. Philadelphia, USA.
Yammine, A., Leklou, N., Choinska, M., Bignonnet, F., & Mechling, J. M. (2020).
DEF damage in heat cured mortars made of recycled concrete sand aggregate.
Construction and Building Materials 252, 119059. https://doi.org/10.1016/
j.conbuildmat.2020.119059
Chapter 3

Assessing the condition of

ISR-­affected concrete

3.1 SERVICE LIFE OF CONCRETE INFRASTRUCTURE

Critical concrete infrastructure, including bridges, overpasses, dams, tun-

nels, industrial buildings and stadiums, is designed with a specified lifes-
pan known as service life. The concept of service life dates back to early
observations made by builders who noticed that certain materials and
approaches had longer lifespans than others (ACI 365.1R-­00 2000; Davey
1961). Historically, the prediction and evaluation of service life have relied
on qualitative and empirical methods. However, advancements in under-
standing the common deterioration mechanisms of concrete have paved the
way for more quantitative predictions of concrete infrastructure lifespan
(ACI 365.1R-­00 2000).
Typically, critical infrastructure is designed with a service life ranging
from 50 to 75 years, depending on the structure type and design standards.
The service life of such infrastructure is not solely determined by the risk of
failure but also considers overall functionality. Excessive operating costs, for
example, may render a structure economically unviable, which in turn may
require its replacement and, thus, the end of service life. It is important to
note that the terms “durability” and “service life” are distinct and are often
mistakenly used interchangeably. Durability refers to the ability of a prod-
uct, component, assembly or construction to maintain its intended function(s)
over a specified time, while service life is defined as the period after place-
ment during which all properties consistently exceed the minimum accept-
able values with routine maintenance (ACI 365.1R-­00 2000).
According to the ACI 365 committee, three types of service life can be
defined: technical service life, functional service life and economic service
life (Sommerville 1986; Sommeville 1992):

• Technical service life: elapsed time until a defined unacceptable state

is reached, such as significant concrete spalling, safety levels falling
below acceptable thresholds or member’s failure.

30 DOI: 10.1201/9781003188155-3
 Assessing the condition of ISR-affected concrete 31

• Functional service life: refers to the duration until a structure no longer

meets functional requirements or becomes obsolete due to changing
demands, such as the need for increased clearance or road widening.
• Economic service life: the duration at which replacing the structure or
a portion of it becomes more financially advantageous than maintain-
ing it in service.

Service life concepts and methodologies are applicable in both the design and
operation phases of structures. Parameters such as water-­to-­cementitious
materials ratio, concrete cover and inspection and maintenance strategies
are considered in the design phase to ensure the desired service life. Over
the years, various methodologies have been developed to predict the service
life of critical concrete infrastructure. These methodologies rely on informa-
tion about the current condition of the structure (or structural member)
under analysis, deterioration rates, past and future loading conditions and
definition of the end of service life (Clifton 1991). Based on predictions of
remaining life, economic decisions can be made regarding whether to repair,
rehabilitate or replace the structure. The end of service life is typically deter-
mined by criteria such as (ACI 365.1R-­00 2000)

• severe material deterioration or exceeding the design load-­carrying

capacity,
• maintenance requirements surpass available resources,
• aesthetics becoming unacceptable and
• functional capacity of the structure no longer meeting demand.

Ultimately, decisions regarding the end of service life are influenced by con-
siderations of human safety and economics. In many cases, the condition,
appearance or capacity of concrete structures can be improved to acceptable
levels; however, the associated costs may be prohibitive. Therefore, manage-
ment strategies for critical concrete infrastructure should be based upon a
thorough assessment of the structure’s condition (ACI 365.1R-­00 2000).

3.2 CONDITION ASSESSMENT OF CONCRETE

INFRASTRUCTURE

Condition assessment refers to the comprehensive evaluation of the over-

all state of a structure or its components. Traditional protocols for condi-
tion assessment begin with a detailed examination of relevant information
pertaining to the structure, including soil profiles, foundation type, as-­built
drawings, construction records and a history of deterioration. This initial
phase is followed by field surveys and testing that encompass visual inspec-
tions, the application of non-­destructive testing (NDT), structural health
monitoring (SHM) and long-­term monitoring techniques. In certain cases,
32 Internal Swelling Reactions in Concrete

laboratory and structural analyses may also be employed as necessary

(Blight & Alexander 2011; Omar & Nehdi 2018; Saviotti 2014). Once the
aforementioned tasks have been completed, structural evaluations are con-
ducted to assess the current load-­bearing capacity of the structure (or struc-
tural member(s)) under analysis and ascertain its ability to withstand service
loads. These evaluations involve the use of structural testing and/or model-
ling techniques (ACI 365.1R-­00 2000; Blight & Alexander 2011; Kennedy
1958). Figure 3.1 illustrates the common sequence of activities while the
condition assessment of concrete infrastructure.
A comprehensive and systematic condition assessment is essential for
accurately predicting the current and future performance of infrastructure,
as well as for optimizing maintenance, rehabilitation and replacement strat-
egies. However, it is worth noting that many ageing structures are still evalu-
ated using simplistic, qualitative and subjective methods such as visual
inspection, which can be influenced by the personal judgement of the inspec-
tor. Nonetheless, significant advancements have been made in recent decades
in the field of advanced techniques for condition assessment, including NDT,
SHM, laboratory analyses and more. In the following sections, a brief over-
view of these techniques will be provided, discussing their fundamental con-
cepts and specific applications, both individually and in combination, for
achieving a reliable assessment of critical infrastructure.

3.2.1 Examination of current documentation

Prior to conducting any on-­site inspection, it is essential to thoroughly
review all relevant documents and information pertaining to the structure.
This comprehensive review serves as the foundation for the subsequent
field inspection. The following information should be carefully sought and
examined:

• Identification of the structure type and purpose

• Assessment of environmental conditions and the degree of exposure
• Date of completion and any historical records
• Documentation of prior rehabilitation or maintenance work, whether
minor or major
• Details regarding the design loading and construction specifics
• Examination of foundation characteristics and soil profiles
• Evaluation of the drainage system and joint arrangements
• Identification of the reinforcement, including its positioning, amount
and detailing
• Documentation of concrete properties, such as mix-­proportions and
the sources of raw materials employed (e.g., cement and aggregates)
• Review of previous inspection reports and any past field and labora-
tory analyses conducted
 Assessing the condition of ISR-affected concrete
Figure 3.1 C ondition assessment flowchart.
(Adapted from ACI 365.1R-­0 0 2000; Omar & Nehdi 2018.)

33
34 Internal Swelling Reactions in Concrete

• Identification of the dates when signs of deterioration were first

observed
• Consideration of any additional information that may be relevant to
the field inspection, such as the performance of neighbouring or simi-
lar structures, age of the structure and any design modifications that
have been made

Once all the necessary information has been gathered, field inspections can
be scheduled. It is important to note that the validity and reliability of the
available documents, particularly test results, should be carefully evaluated
and, if possible, cross-­verified through on-­site observations.
During the site inspection, visible indications of deterioration must be
meticulously documented. This includes recording crack widths, directions,
spacing and precise locations. These observations should be captured photo-
graphically and annotated on scale drawings or, alternatively, recorded
through freehand sketches with measured dimensions and directional indica-
tors, such as a north point (Blight & Alexander 2011; Fournier et al. 2010).

3.2.2 Visual inspection (VI)

VI serves as the primary task in conducting a comprehensive condition
assessment of concrete infrastructure. In various regions, including North
America, routine VI is typically performed at intervals of approximately
two years, with the specific frequency dependent on the type and condition
of the structure being analysed. For ageing infrastructure (i.e., over 30 years
old), enhanced VI is conducted at six-­year intervals. Additionally, detailed
emergency inspections are performed when there is an imminent risk of
failure or following catastrophic failure to ensure public safety (Akula et al.
2014; Omar & Nehdi 2018). VI plays a crucial role in detecting and iden-
tifying signs and patterns of deterioration, including flaws, defects, crack
networks, discolorations and the presence of secondary products associated
with various deterioration mechanisms. However, it is important to note
that VI is a qualitative and subjective procedure influenced by the equip-
ment used and operator interpretation. In many cases, it may not provide
an accurate assessment of the cause and extent of deterioration or facilitate
the selection of appropriate rehabilitation strategies for affected structures
(Blight & Alexander 2011; Moore et al. 2000; Zahedi et al. 2022). Various
equipment can be utilized during VI to aid in the inspection process, such as
(Blight & Alexander 2011; Fournier et al. 2010)

• hand lens,
• binocular,
• crack gauge,
• measuring tape and
• camera.
 Assessing the condition of ISR-affected concrete 35

However, there is a pressing need for the development of improved and

standardized VI guidelines tailored to different types of structures. This
would ensure that VI outcomes can be reliably utilized in the overall assess-
ment of critical infrastructure, particularly those affected by ISR. Chapter
4 explores both conventional and advanced VI approaches employed for
assessing critical concrete infrastructure.

3.2.3 Field testing
Several techniques can be employed in the field to assess the condition of
critical infrastructure, with three of the most commonly used being NDT,
SHM and load testing (LT).

3.2.3.1 Non-­destructive testing (NDT)

The early detection of deterioration in concrete, prior to major damage, is
facilitated by the use of NDT. NDT methods can be integrated into VI pro-
tocols for concrete infrastructure to evaluate parameters such as stiffness,
strength, moisture content and the presence of flaws, defects and cracks.
Over the past decades, several NDT methods have been developed, each
exploring distinct phenomena, such as acoustic, seismic, electric, thermal
and electromagnetic, based on the predominant deterioration mechanisms
leading to damage (Gucunski et al. 2013). For instance, the probability of
active corrosion can be assessed using methods including half-­cell poten-
tial (Pradhan & Bhattacharjee 2009), electrical resistivity (Browne 1980)
and ground penetrating (Varnavina et al. 2015) techniques, while corro-
sion rate can be identified by the polarization resistance method (Cady &
Gannon 1992). Furthermore, the presence of vertical cracks, which results
in a reduced modulus of elasticity of concrete, can be captured using the
ultrasonic surface wave method (Nazarian et al. 1993). Concrete delamina-
tion can be detected using techniques such as impact echo (Kee et al. 2012;
Shokouhi et al. 2014), pulse echo (Krause et al. 2011) and infrared thermog-
raphy tests (Kee et al. 2012; Washer et al. 2009). Additional information on
the theoretical foundations, instrumentations, applications and data analy-
sis of NDT can be found in Chapter 4.

3.2.3.2 Structural health monitoring (SHM)

SHM is a non-­destructive technique employed for in situ evaluation. It uti-
lizes multiple sensors embedded in the structural member(s) of interest to
monitor their structural response and identify any abnormal behaviour.
The primary objective of SHM techniques is to estimate deterioration and
assess its impact on the structural response and capacity. In recent years, sev-
eral SHM systems have been developed and implemented to provide valu-
able information on concrete infrastructure. These systems typically share
36 Internal Swelling Reactions in Concrete

fundamental elements, including sensor measurements and instrumentation,

structural assessment through peak strains or modal analysis and support
for maintenance, rehabilitation and replacement strategies (Alampalli 2012).
The effectiveness of an SHM system depends on the type and quantity of
sensors employed. Monitoring systems can utilize single or multiple sensor
types, which can be customized to capture various physical measurements
associated with loads and environmental conditions (Wong 2007). In addi-
tion to structural evaluation, SHM systems, equipped with diverse sensor
types, can also identify material properties such as concrete creep, shrinkage
and corrosion. Furthermore, they can assess environmental effects, such as
temperature gradients and dynamic responses (e.g., traffic-­induced vibra-
tions; Feng & Feng 2018). While SHM methods hold great promise, their
widespread implementation is still limited. The absence of technical stan-
dards for SHM systems has historically hindered their extensive adoption in
the condition assessment of critical infrastructure (Bulletin fib Task Group
3.3. 2022; TG3.3 fib Bulletin 2023).

3.2.3.3 Load testing (LT)

LT is a valuable technique for assessing critical infrastructure, especially
bridges (AASHTO 2011). It enables the determination of safe loading
levels for structures or structural members. Through forced static and
dynamic load testing in varied load conditions, the maximum response
can be detected using strain transducers placed at critical locations of the
structure or structural member of interest (Omar & Nehdi 2018). Load
tests can be broadly classified into two categories: proving load tests,
which serve as self-­ supporting alternatives to theoretical assessments,
and supplementary load tests, which complement theoretical calculations
(Zhang et al. 2016). Load ratings are then established using methods such
as allowable stress, load factor or load and resistance factor. Additionally,
load testing can be combined with preliminary or advanced structural
analysis, including finite element models (FEM), as well as NDT and SHM
techniques. This integration, within a structural reliability framework,
facilitates the determination of practical safety factors for ageing concrete
infrastructure (Wang et al. 2011).

3.2.3.4 Field testing combination for assessing the condition

of concrete infrastructure
Various condition assessment protocols and frameworks have been devel-
oped over the years based on the field methodologies discussed earlier. One
of the most recent and comprehensive approaches was introduced by the
International Federation for Structural Concrete (TG3.3 fib Bulletin 2023).
This framework revolves around the acquisition and utilization of a set of
indicators that pertain to the serviceability and safety of the structure (or
 Assessing the condition of ISR-affected concrete 37

structural member) being analysed. These indicators are derived from data
obtained through different yet complementary methods, including VI, NDT
and SHM. By utilizing these methodologies, the framework enables the cal-
culation of indicators that provide insight into the current state and load-
ing conditions of the structure. The indicators are classified based on the
method used to gather the information and the level of assessment, whether
it is local or global. It is important to highlight that within this proposed
approach, NDT is particularly suited for localized damage detection evalu-
ation, while SHM is typically utilized to obtain global performance data
for the structure or its components (Figure 3.2) (Bulletin fib Task Group
3.3., 2022).

3.2.4 Laboratory analyses
The condition of the structure (or structural member) under analysis may
require additional laboratory tests. These tests encompass various tech-
niques, such as chemical, microscopic and mechanical assessments, each
serving its own purpose. Chemical assessment focuses on investigating the
presence of secondary products that can indicate or confirm the occurrence
of chemical deterioration in concrete. This deterioration can lead to material
degradation and, consequently, structural implications. Microscopic analy-
ses are indicated to evaluate concrete microstructure, which could suggest
the presence of secondary products and or crack patterns related to physical
or chemical deterioration. Mechanical assessment searches on the impact
of physical or chemical deterioration on the engineering properties of con-
crete, such as tensile, compressive and shear strengths, as well as stiffness.
These methods are typically selected when there is a significant degree of
visible damage to the structure under investigation or when ISR mechanisms
are suspected to be involved in the overall damage process. More detailed
discussions on these procedures can be found in Chapters 5 (microscopic
techniques) and 6 (mechanical testing).

3.2.5 Structural analyses
A thorough structural analysis should be conducted using either traditional
or numerical methods to assess the structural performance of concrete
infrastructure after field and/or laboratory testing. Traditional methods
rely on conventional design standards and guidelines, offering effective
approaches but often characterized by conservative assumptions. On the
other hand, numerical assessment, such as FEM, enables a more reliable,
accurate and realistic representation of the structure’s condition (Omar &
Nehdi 2018). In FEM, various variables are typically considered, including
the properties of the materials, the geometry of the structure and/or struc-
tural components, the design and detailing of reinforcement, the current
state of damage, the construction process and the environmental conditions
 38
Internal Swelling Reactions in Concrete
Figure 3.2 C ondition assessment framework.
(Adapted from Bulletin fib Task Group 3.3., 2022.)
 Assessing the condition of ISR-affected concrete 39

(Sousa et al. 2014). Therefore, the use of FEM is extremely popular to evalu-
ate the structural capacity of ISR-­affected infrastructure.

3.3 ISRS AS ONGOING DETERIORATION PROCESSES

Significant progress has been made over the years in the field of condition
assessment of concrete infrastructure. Traditional assessment methods,
including some of the methodologies previously discussed in this chapter,
have proven effective for evaluating ageing and deterioration processes,
such as load-­ induced cracks or damage caused by aggressive external
agents (e.g., carbonation, chloride penetration leading to steel corrosion,
sulphate attack), where the deterioration rates caused by these mecha-
nisms are fairly well understood via transport mechanisms, besides being
easily detected (or measured) in the field. However, when internal damage
mechanisms, such as ISRs, occur within the concrete material, the appli-
cation of these tools and protocols becomes more complex and requires
further examination.
ISR mechanisms are harmful, ongoing processes leading to induced
expansion and deterioration which impact the durability, serviceability and
structural performance of affected structures. The rate of induced expansion
and deterioration varies depending on factors such as the type of ISR (e.g.,
alkali-­aggregate reaction, delayed ettringite formation, freeze and thaw (FT)
damage), environmental conditions (global and local factors like tempera-
ture, relative humidity, exposure degree and solar radiation), concrete mix-­
proportions and the geometry and confinement of the affected structures (or
structural members). It is generally accepted that ISR-­induced expansion
follows an S-­ shaped curve over time, although linear trends are often
observed in practice, especially in large structures such as dams. Figure 3.3
provides an illustration of the theoretical evolution of ISR-­induced expan-
sion as a function of time.
The assessment of structures affected by ISR is a complex undertaking since
besides detecting the main cause(s) generating damage, along with their cur-
rent deterioration extent, the so-­called diagnosis, it is also important to fore-
cast their future behaviour over time (i.e., prognosis). The techniques
mentioned earlier in this chapter, although useful, offer only a limited assess-
ment of ISR-­affected structures. They may provide insights into either the
cause(s) or the extent of the damage but generally yield short-­term informa-
tion that is insufficient for predicting the long-­term performance and service
life of the affected structures (Kennedy 1958). This means that the develop-
ment of a comprehensive and reliable protocol enabling the evaluation of the
current cause(s) and degree of damage, along with the deterioration rate over
time (e.g., coupling laboratory test data with in situ monitoring and model-
ling) of ISR-­affected concrete, is crucial and yet a quite complex task, requir-
ing engineering analysis, judgement and interpretation (ACI 365.1R-­00 2000).
40 Internal Swelling Reactions in Concrete

Figure 3.3 T heoretical evolution of ISR-­induced expansion over time.

The proposed protocol should integrate service history, materials and

structural members’ features, current damage cause(s) and extent, structural
analyses and a comprehensive degradation model to enable proper predic-
tion of future behaviour. Furthermore, such methodology should be able to
appraise the role of maintenance in extending service life or structural reli-
ability of damaged structures (Naus & Oland 1994; Sanchez et al. 2020).
Moreover, it is essential to recognize that different ISR mechanisms, such as
AAR, FT and DEF, exhibit unique deterioration processes that have varying
impacts on the engineering properties of the affected concrete. Therefore, it
becomes necessary to define the concept of “damage” in a broader and more
comprehensive manner that encompasses these distinctions. While the term
“damage” has been interpreted differently in various studies and by differ-
ent authors, it can be broadly defined as the “harmful and measurable con-
sequences of various mechanisms (e.g., loadings, shrinkage, creep, ASR,
DEF, freezing and thawing) on the mechanical properties, physical integrity,
and durability of concrete” (Sanchez et al. 2020). In practical terms, damage
can be defined as (a) engineering properties reductions (i.e., compressive,
tensile and direct shear strengths) of concrete, (b) stiffness reduction of con-
crete and (c) physical integrity and/or durability loss, which is directly
related to the existing cracking network. Figure 3.4 illustrates the broad
damage definition for ISR-­affected concrete (Sanchez et al. 2020).
It is important to note that it is outside of the scope of the book to discuss
available techniques, models and calculations to be used to perform struc-
tural assessment of ISR-­affected structures. However, the primary objective
 Assessing the condition of ISR-affected concrete 41

Figure 3.4 G lobal classification of damage in concrete.

(Adapted from Sanchez et al. 2014.)

of this book is to comprehensively examine the efficiency of various tech-

niques and protocols that can be employed to assess the current and future
condition of ISR-­affected concrete. This is crucial to provide input and data
for subsequent structural performance analyses. In this context, various
visual, non-­destructive, microscopic and mechanical techniques have been
utilized to assess the condition of ISR-­affected concrete. Some of these tech-
niques have shown considerable promise, while others have been deemed
obsolete. In the forthcoming chapters, the implementation, adaptation and
advancement of field and laboratory techniques capable of efficiently quan-
tifying ISR-­induced deterioration will be extensively discussed. The advan-
tages and disadvantages of each technique will be presented, and protocols
enabling the development of frameworks to support infrastructure owners
in making informed decisions on their assets will be introduced.

REFERENCES

AASHTO. (2011). Manual for Bridge Evaluation (2nd ed.). American Association of
State Highway and Transportation Officials (AASHTO). ISBN: 978-­1-­56051-­519-­7
ACI 365.1R-­00. (2000). Service-­Life Prediction-­State-­of-­the-­Art Report.
Akula, M., Zhang, Y., Kamat, V. R., & Lynch, J. P. (2014). Leveraging structural
health monitoring for bridge condition assessment. Construction Research
Congress 2014, 1159–1168. https://doi.org/10.1061/9780784413517.119
Alampalli, S. (2012). Special Issue on Nondestructive evaluation and testing for
bridge inspection and evaluation. Journal of Bridge Engineering, 17(6), 827–828.
https://doi.org/10.1061/(ASCE)BE.1943-­5592.0000430
Blight, G. E., & Alexander, M. G. (2011). Alkali-­aggregate reaction and structural
damage to concrete. CRC Press. https://doi.org/10.1201/b10773
Browne, R. D. (1980). Mechanisms of corrosion of steel in concrete in relation to
design, inspection, and repair of offshore and coastal structures. ACI Special
Publication, 65, 169–204.
Bulletin fib Task Group 3.3. (2022). Existing concrete structures: Life management,
testing and structural health monitoring (in preparation).
Cady, P. D., & Gannon, E. J. (1992). Condition evaluation of concrete bridges rela-
tive to reinforcement corrosion. Volume 8: Procedure manual. National Research
Council.
Clifton, J. R. (1991). Predicting the remaining service life of concrete and concrete
technology. National Institute of Standards and Technology.
42 Internal Swelling Reactions in Concrete

Davey, N. (1961). A history of building materials. Phoenix House.

Feng, D., & Feng, M. Q. (2018). Computer vision for SHM of civil infrastructure:
From dynamic response measurement to damage detection – A review. Engineering
Structures, 156, 105–117. https://doi.org/10.1016/j.engstruct.2017.11.018
Fournier, B., Bérubé, M. A., Folliard, K., & Thomas, M. (2010). Report on the diag-
nosis, prognosis, and mitigation of Alkali-­Silica Reaction (ASR) in transportation
structures.
Gucunski, N., Imani, A., Romero, F., Nazarian, S., Yuan, D., Wiggenhauser, H.,
Shokouhi, P., Taffe, A., & Kutrubes, D. (2013, January 13). Nondestructive test-
ing to identify concrete bridge deck deterioration. Proceedings of the 92nd Meet.
https://doi.org/10.17226/22771
Kee, S.-H., Oh, T., Popovics, J. S., Arndt, R. W., & Zhu, J. (2012). Nondestructive
bridge deck testing with air-­ coupled impact-­ echo and infrared thermography.
Journal of Bridge Engineering, 17(6), 928–939. https://doi.org/10.1061/(ASCE)
BE.1943-­5592.0000350
Kennedy, T. B. (1958). Laboratory testing and the durability of concrete. In ASTM
STP 236 (Ed.), Symposium on Approaches to Durability in Structures. ASTM.
Krause, M., Mayer, K., Friese, M., Milmann, B., Mielentz, F., & Ballier, G. (2011).
Progress in ultrasonic tendon duct imaging. European Journal of Environmental
and Civil Engineering, 15(4), 461–485. https://doi.org/10.1080/19648189.2011.
9693341
Moore, M., Phares, B., Graybeal, B., Rolander, D., & Washer, G. (2000). Reliability of
visual inspection for highway bridges. Federal Highway Administration (FHWA).
Naus, D. J., & Oland, C. B. (1994). Structural aging program technical progress
report for period.
Nazarian, S., Baker, R., & Crain, K. (1993). Development and testing of a seismic
pavement analyzer; report SHRP-­H-­375.
Omar, T., & Nehdi, M. (2018). Condition assessment of reinforced concrete bridges:
Current practice and research challenges. Infrastructures, 3(3), 36. https://doi.
org/10.3390/infrastructures3030036
Pradhan, B., & Bhattacharjee, B. (2009). Half-­cell potential as an indicator of chloride-­
induced rebar corrosion initiation in RC. Journal of Materials in Civil Engineering,
21(10), 543–552. https://doi.org/10.1061/(ASCE)0899-­1561(2009)21:10(543)
Sanchez, L. F. M., Fournier, B., Jolin, M., & Bastien, J. (2014). Evaluation of the
stiffness damage test (SDT) as a tool for assessing damage in concrete due to
ASR: Test loading and output responses for concretes incorporating fine or coarse
reactive aggregates. Cement and Concrete Research, 56, 213–229. https://doi.
org/10.1016/j.cemconres.2013.11.003
Sanchez, L. F. M., Fournier, B., Mitchell, D., & Bastien, J. (2020). Condition
assessment of an ASR-­ affected overpass after nearly 50 years in service.
Construction and Building Materials, 236, 117554. https://doi.org/10.1016/j.
conbuildmat.2019.117554
Saviotti, A. (2014). Bridge assessment, management and life cycle analysis. Modern
Applied Science, 8(3). https://doi.org/10.5539/mas.v8n3p167
Shokouhi, P., Wolf, J., & Wiggenhauser, H. (2014). Detection of delamination in con-
crete bridge decks by joint amplitude and phase analysis of ultrasonic array mea-
surements. Journal of Bridge Engineering, 19(3). https://doi.org/10.1061/(ASCE)
BE.1943-­5592.0000513
 Assessing the condition of ISR-affected concrete 43

Sommerville, G. (1986). Design life of structures. The Structural Engineer, 64A(2),

60–71.
Sommeville, G. (1992). Service life prediction – an overview. Concrete International,
14(11), 45–49.
Sousa, H., Bento, J., & Figueiras, J. (2014). Assessment and management of con-
crete bridges supported by monitoring data-­ based finite-­
element modeling.
Journal of Bridge Engineering, 19(6). https://doi.org/10.1061/(ASCE)BE.1943-­
5592.0000604
TG3.3 fib Bulletin. (2023). Condition assessment of reinforced concrete structures:
State of the art: knowledge and case studies in the TG3.3 fib Bulletin.
Varnavina, A. V., Khamzin, A. K., Torgashov, E. V., Sneed, L. H., Goodwin, B. T., &
Anderson, N. L. (2015). Data acquisition and processing parameters for con-
crete bridge deck condition assessment using ground-­coupled ground penetrating
radar: Some considerations. Journal of Applied Geophysics, 114, 123–133. https://
doi.org/10.1016/j.jappgeo.2015.01.011
Wang, N., Ellingwood, B. R., & Zureick, A.-H. (2011). Bridge rating using sys-
tem reliability assessment. II: Improvements to bridge rating practices. Journal
of Bridge Engineering, 16(6), 863–871. https://doi.org/10.1061/(ASCE)BE.1943-­
5592.0000171
Washer, G., Fenwick, R., Bolleni, N., & Harper, J. (2009). Effects of environmen-
tal variables on infrared imaging of subsurface features of concrete bridges.
Transportation Research Record: Journal of the Transportation Research Board,
2108(1), 107–114. https://doi.org/10.3141/2108-­12
Wong, K.-Y. (2007). Design of a structural health monitoring system for long-­span
bridges. Structure and Infrastructure Engineering, 3(2), 169–185. https://doi.
org/10.1080/15732470600591117
Zahedi, A., L. F. Sanchez, & Noël, M. (2022). Appraisal of visual inspection tech-
niques to understand and describe ASR-­induced development under distinct con-
finement conditions. Construction and Building Materials, 323, 126549. https://
doi.org/10.1016/j.conbuildmat.2022.126549
Zhang, Q., Alam, M. S., Khan, S., & Jiang, J. (2016). Seismic performance compari-
son between force-­based and performance-­based design as per Canadian Highway
Bridge Design Code (CHBDC) 2014. Canadian Journal of Civil Engineering,
43(8), 741–748. https://doi.org/10.1139/cjce-­2015-­0419
Chapter 4

Visual inspection and

non-­destructive testing (NDT)

4.1 INTRODUCTION

Visual inspection (VI) and non-­destructive testing (NDT) are frequently

the first procedures conducted while assessing the condition of concrete
infrastructure, regardless of the damage type and degree observed. These
techniques allow for gathering preliminary information on the presence of
deterioration while deciding if further assessment is required (Kim et al.
2019). This chapter presents a thorough description of visual inspection
procedures along with commonly used NDT to assess the condition of ISR-­
affected concrete.

4.2 VISUAL INSPECTION (VI)

VI is normally the very first step towards assessing the presence of dete-
rioration, irregularities and displacements of concrete members, ensuring
they still meet structural safety regulations and expected service require-
ments (Henrickson et al. 2016; Koch et al. 2015). VI processes generally
include a descriptive survey of the type(s) and degree of existing deteriora-
tion observed on the surface of concrete members (David & Gregory 2017;
Gattulli & Chiaramonte 2005; Kabir 2010a). Although VI is qualitative
by nature, it can be combined with quantitative measurements, such as the
measurement of crack widths and computation of crack density, to provide
more detailed information and classification on the overall deterioration
process. For instance, Table 4.1 illustrates the accepted crack widths in rein-
forced concrete members under service loads as per the American Concrete
Institute (ACI) 224R-­01 – Control of cracking in concrete structures (ACI
224R-­19 2019).
Regardless of the mechanism inducing damage, performing routine VI is
considered crucial to identify early deterioration and thus prevent damage
escalation and important structural implications leading to catastrophic
failure (Golden et al. 2018; Koch et al. 2015; Wood 2008). Moreover,

44 DOI: 10.1201/9781003188155-4
 Visual inspection and non-destructive testing (NDT) 45

Table 4.1 A
 cceptable crack widths in reinforced concrete members under service
loads (ACI 224R-­19 2019)

Crack width

Exposure condition mm inches

Dry air or protective membrane 0.41 0.016

Humidity, moist air, soil 0.30 0.012
De-­icing chemicals 0.18 0.007
Seawater and seawater spray, wetting and drying 0.15 0.006
Water-­retaining structures (excluding nonpressure pipes) 0.10 0.004

although the full recognition of the main cause(s) leading to deterioration

cannot be made only through VI, the knowledge of common damage pat-
terns and features associated with distinct deterioration processes may help
in their identification in the field. The next section presents common “mac-
roscopic” damage features related to ISR in affected concrete.

4.2.1 ISR-­induced surface damage signs on concrete

4.2.1.1 Alkali-­aggregate reaction (AAR)
The main damage indicators and features associated with AAR-­induced
deterioration are as follows (ACI 201.1R 2008):

• Deposits: upon water penetration in concrete, deposits may be gener-

ated on the concrete member’s surface due to the dissolved or leached
chemicals. Figure 4.1 illustrates distinct deposits that may be presented
as (a) efflorescence, where whitish and powdery salt is observed; (b)
exudation, where a liquid or gel-­like deposit is observed in voids or
cracks; (c) incrustation, where a hard crust is formed and; (d) stalac-
tite, where a downward pointing formation suspended from the con-
crete surface is formed, usually shaped like an icicle.
• Map cracking: A typical indication of AAR-­induced development
where cracks form a repetitive polygonal pattern that partially or
locally covers the surface of affected concrete. These randomly dis-
tributed cracks may become more oriented in locations bearing
higher confinement and/or restraint degrees. Figure 4.2 illustrates a
y-­shaped column seated in a foundation block. The block displays a
random map cracking since it bears minor reinforcement, whereas
the cracks present in the y-­shaped columns follow a more oriented
path (i.e., parallel to the main reinforcement) due to the reinforce-
ment restraint.
46 Internal Swelling Reactions in Concrete

Figure 4.1 D
 eposit signs on the surface of AAR-­d eteriorated concrete: (a) efflo-
rescence along major cracks and joints, (b) exudation of gel-­l ike prod-
uct staining the concrete adjacent to the cracks, (c) encrustation on
the surface of the concrete and (d) stalactites.
([a] photograph courtesy of Leandro Sanchez; [b–d] photograph cour-
tesy of Cassandra Trottier.)

Figure 4.2 A
 lkali-­s ilica reaction (ASR)-affected concrete member in Quebec City,
Canada.
(Photograph courtesy of Leandro Sanchez.)
 Visual inspection and non-destructive testing (NDT) 47

Figure 4.3 T
 ypical DEF damage features in massive concrete members from (a)
(Thomas et al. 2008) and (b) (Godart & Divet 2013).

4.2.1.2 Delayed ettringite formation (DEF)

The main damage indicators and features associated with DEF are similar to
AAR (Martin et al. 2015). The main difference is that DEF-­induced expan-
sion and deterioration typically occur in massive concrete structures such as
piers, pier caps, bridge girders and foundation blocks, as illustrated in Figure
4.3; DEF is also observed in steam-­cured concrete, normally used in precast
members (Godart & Divet 2013; Karthik et al. 2016b; Karthik et al. 2016a).

4.2.1.3 Other mechanisms: freezing and thawing (FT)

Deterioration caused by fluctuating temperatures is largely observed under
harsh climates where the mean temperature varies above and below the
freezing point of water (i.e., 0°C); this variation leads to the occurrence of
FT cycles, which may cause a wide range of deterioration signs, including
the following (ACI 201.1R 2008):

• Scaling: Local flaking or peeling of finished surfaces of hardened con-

crete. Scaling can be categorized as marginal, moderate and severe
depending on the deterioration depth (d) and area, i.e., marginal scal-
ing (loss of surface mortar, no coarse aggregates’ exposure); moderate
scaling (loss of surface mortar from 5 < d < 10 mm and some aggre-
gate’s exposure) as displayed in Figure 4.4a; severe scaling (loss of
surface mortar from 5 < d < 10 mm surrounding 10 or 20 mm coarse
aggregate particles; very severe (loss of surface mortar and coarse
aggregate particles, with d > 20 mm), as per Figure 4.4b.
• Delamination: A type of scaling affecting a larger area, where two lay-
ers of concrete are separated from one another, as illustrated by Figure
4.4c; delamination may not appear at the surface of affected concrete
but is detectable through NDT.
48 Internal Swelling Reactions in Concrete

Figure 4.4 C
 oncrete surface scaling at two levels of damage: (a) small scaling
showing small dispersed aggregate particles, (b) medium scaling show-
ing dispersed aggregate particles and some pits of mortar and (c)
severe scaling showing a large number of exposed aggregate particles.

Figure 4.5 ( a) Typical D-­c racking (Auberg & Setzer 2014) and (b) pop-­o ut in
concrete.
(Photograph courtesy of Cassandra Trottier.)

• Cracking: FT-­induced cracking is generally propagated in the form of

the so-­called D-­cracking, which is near and roughly parallel to joints
and edges. Such damage can be observed in concrete pavements after
a few years of service, as illustrated in Figure 4.5a.
• Pop-­outs: damage feature defined by the spalling of small portions of
concrete incorporating coarse aggregate particles, as shown in Figure
4.5b; normally, pop-­outs take place during the use of highly permeable
coarse aggregate particles that break under pressure caused by water
freezing.

4.2.1.4 Other mechanisms: sulphide-­bearing aggregates

The deterioration caused by sulphide-­bearing aggregates has gained signifi-
cant attention in the last couple of years in regions such as North America
and northern Europe (Casanova et al. 1996; Chinchón et al. 1995; Lugg &
Probert 1996), especially after the occurrence of important deterioration in
Trois-­Rivière (Quebec, Canada), where concrete foundations in a number
 Visual inspection and non-destructive testing (NDT) 49

of houses exhibited quite impressive damage signs within a relatively short

period of time (i.e., three to five years; Duchesne et al. 2021; Rodrigues et al.
2012). The most common deterioration signs of sulphide-­bearing aggregates
deterioration include (Rodrigues et al. 2012):

• Cracking: As for AAR and DEF, sulphide-­bearing aggregate induces

randomly distributed cracks in concrete, the so-­called map cracking
(Figure 4.6a), with greater prominence at the edges of affected struc-
tural members or areas where water accumulates, such as locations
close to rain gutters (Figure 4.6b). The crack openings can reach up to
10 mm, and in regions of more severe damage, these values can esca-
late to as high as 40 mm (Figure 4.6c).
• Deposits and discolouration: The surface of concrete affected by
sulphide-­bearing aggregates often displays a yellowish coloration.
Iron hydroxides and rust can be observed within the cracks, indicating

Figure 4.6 S ulphide-­b earing aggregates induced deterioration: (a) map-­c racking
pattern on interior home foundation painted wall in Connecticut,
United States (Geiss & Gourley 2019); (b) cracking observed at a
house foundation in Trois-­R ivière, Canada; the top green bars of the
crack comparator card represents 1 cm and the bottom represents
1″ (Rodrigues et al. 2016); (c) a 12 mm (1/2″) crack opening in a con-
crete foundation in Connecticut, United States (Zhong & Wille 2018);
and (d) a slab-­o n-­g rade with yellowish and rust stains in Trois-­R ivière,
Canada (Rodrigues et al. 2012).
50 Internal Swelling Reactions in Concrete

the presence of chemical reactions and oxidation processes. Reddish-­

brown discolouration (i.e., rust spots or stains) may also be observed
ay the concrete’s surface, as illustrated in Figure 4.6d (Jana 2022).
• Pop-­out: Pop-­outs are quite common on the interior side of foundation
walls. These pop-­outs are characterized by the presence of oxidized
aggregate particles surrounded by a whitish or yellowish powdery
deposit. They commonly occur in areas where the spalling of small
portions of concrete, incorporating coarse aggregate particles, has
taken place.
• Excessive deformations: Due to the extensive concrete cracking and
displacement, deformations of concrete members even showing sec-
tion losses can be observed (Figure 4.7).

4.2.2 Quantifying ISR-­induced deterioration in

concrete
Semi-­quantitative approaches have been developed to complement the out-
comes of qualitative and descriptive visual inspections; besides quantifying
the surface damage degree of affected concrete, they can be used to esti-
mate the rate of deterioration over time. Amongst the proposed methods,
the cracking index (CI) is probably the most conventional technique used to
appraise ISR-­induced deterioration, providing an indication of the damage
degree of the deteriorated concrete member under evaluation.
CI was proposed in France in the 1990s and requires drawing a region of
interest (ROI) for analysis, which is normally 0.5 m by 0.5 m square divided
into even intervals (Figure 4.8); the cracks intercepting the drawn segments
are counted, and their widths measured. Sometimes, a cross is drawn in the
interior of the square to increase precision in capturing cracks from distinct
directions (Figure 4.8a). The size of the ROI may be adjusted (i.e., increased

Figure 4.7 A
 map-­c racked concrete house foundation presenting very large crack
openings in Connecticut, United States (Rodrigues et al. 2016).
 Visual inspection and non-destructive testing (NDT) 51

Figure 4.8 C
 I method: (a) schematic of the original CI as per (Fasseu & Michel
1997) and (b) adapted CI grid drawn on the surface of the deterio-
rated concrete.
(Photograph courtesy of Cassandra Trottier.)

or decreased) based on the specific structural member being investigated.

For larger members such as dams and retaining walls, a square with dimen-
sions of at least 1 m by 1 m is recommended. Otherwise, for slender mem-
bers, such as precast beams, squares smaller than 0.5 m by 0.5 m are suitable
(Figure 4.8b). It is crucial to adapt the ROI to ensure accurate measurements
for different types of structural members. Furthermore, CI becomes not only
a tool enabling assessment of the current deterioration but also of the dete-
rioration rate once the same ROI is assessed over time (Fasseu & Michel
1997; Fournier et al. 2010; Leemann et al. 2021).
The calculation of the CI involves dividing the total summation of crack
openings by the size of the square segments selected (i.e., base length) as
illustrated by Equation 4.1:

∑ Crack openings
CI = (4.1)
Base length

Concrete members presenting CI values greater than 0.5 mm/m and/or

cracks with widths exceeding 0.15 mm require additional investigations for
better identifying the cause(s) leading to deterioration, along with the extent
of internal damage (Fasseu & Michel 1997; Fournier et al. 2010; Leemann
et al. 2021). Building on the previous equation, Zahedi et al. (2022) pro-
posed a modified version of the CI calculation, allowing the estimation of
the attained expansion in reinforced concrete displaying distinct levels of
reinforcement, as described in Equation 4.2. Figure 4.9 illustrates the cor-
relation between attained expansion and the proposed approach, allowing
a much better estimation when compared solely with the CI value, yet the
52 Internal Swelling Reactions in Concrete

Figure 4.9 U
 sage of CI outcomes to estimate induced expansion of ISR-­a ffected
concrete: (a) correlation between CI and actual expansion (measured
via demec points) and (b) correlation between the proposed approach
by Zahedi et al. and actual expansion (Zahedi et al. 2022).

application of the proposed approach for concrete members bearing distinct

geometry, confinement/restraint and environmental conditions still requires
further investigation (Zahedi et al. 2022).

∑ Crack openings
Base length
ε (%) = (4.2)
n

The value of “n” represents the count of cracks observed within the base-
line length over which the CI measurement is performed.

4.2.3 Other VI techniques
The rise of new and sophisticated technologies has enabled the development
of tools used to perform VI in a more efficient and timely manner, increasing
the accessibility of the members under evaluation and, thus, more thorough
and frequent inspections. Such works have been applied to a wide range of
structures and are normally conducted via image analysis (or digital image
 Visual inspection and non-destructive testing (NDT) 53

processing) techniques. Image analysis is a tool used to quantify objects

and patterns in images and can, therefore, complement VI protocols since,
amongst others, VI bears the purpose of recognizing distinct damage fea-
tures and patterns associated with the various types of ISRs. The automatic
detection of these numerous damage features has been thus the focus of
several studies, from mapping deposits and discoloration (Valença et al.
2013) to identifying potential areas of interest showing signs of corrosion,
delamination and efflorescence (Ogawa et al. 2022; Savino & Tondolo
2021; Valença et al. 2013) and cracking in general (Billah et al. 2019;
Carrasco et al. 2021; Dias et al. 2021; Jang et al. 2019; Kim et al. 2020;
Liu et al. 2020; Ogawa et al. 2022; Savino & Tondolo 2021; Zhu et al.
2011). In the aforementioned works, cameras from smartphones, drones
or vehicles were used (Miyamoto 2013; Radopoulou & Brilakis 2017;
Zhang et al. 2017, 2018), all of which with the goal to decrease the time
and increase the accessibility of visual inspections.
However, only a few works have used image analysis and corresponding
automated techniques to visually assess the surface of concrete affected by
ISR. Kabir (Kabir 2010b) evaluated the level of damage induced by AAR
using a thermographic image coupled with artificial intelligence to charac-
terize and quantify cracking. Moliard (Moliard et al. 2016) digitized images
of concrete surfaces affected by ISR and applied an automated version of the
French CI while updating the index to detect anisotropic behaviour.
It is expected that automated techniques may significantly enhance the VI
and condition assessment of concrete affected by ISR, especially when per-
formed over time (Moliard et al. 2016; Vlahović et al. 2012). Moreover, the
progression of the damage and differential movements on the surface of
ISR-­affected concrete could be obtained by acquiring images at successive
time intervals. The analysis of the differential movements, named digital
image correlation (DIC), is also amongst the emerging technologies used to
monitor concrete structures (Hansen & Hoang 2021; Mahadevan et al.
2017; Thériault et al. 2022). These types of visual inspection techniques are
non-­destructive in nature and are promising tools to enhance the quality of
current VI approaches.

4.3 NON-­DESTRUCTIVE TESTING (NDT)

NDT can complement VI since visual damage signs may not appear until
reaching moderate-­to high-­damage levels (Gunn et al. 2017). The concept
of NDT is generally based on physical principles such as energy, electricity
and motion. NDTs enable the estimation of mechanical properties and the
detection of defects, flaws and imperfections, thus facilitating damage assess-
ment with more or less limited accuracy. Despite notable advancements in
data processing capabilities and the integration of sophisticated techniques
(e.g., reflection attenuation, diffraction, scattering, complex theories of
54 Internal Swelling Reactions in Concrete

heterogeneous media and digital imaging processing), NDT still possesses

limitations in quantitatively evaluating ISR-­affected concrete infrastructure.
However, these techniques may also provide interesting qualitative or rela-
tive appraisal, comparing sound versus damaged locations, for instance.
Various NDTs have been explored for evaluating concrete distressed by
ISR in the field. These techniques include electrical resistivity, surface ther-
mography, ground penetrating radar (GPR), stress waves and resonant fre-
quency analysis. The following sections present the most used NDT for
assessing critical concrete infrastructure affected by ISR, highlighting their
advantages and disadvantages, as well as discussing the future challenges
associated with their field implementation.

4.3.1 Electrical resistivity (ER)

ER is an intrinsic property of materials that hinders the flow of electrical
current through their bodies, regardless of their geometry. While dry con-
crete does not exhibit an electrical response, the presence of a pore solu-
tion allows for the passage of current. This pore solution contains various
ions, including Ca++, K+, Na+, OH− and SO− (as discussed in Chapter 2),
which undergo changes in concentration over time and affect electrical flow.
Consequently, ER becomes a sensitive parameter for detecting the progres-
sion of ISR.
A range of set-­ups can be employed to measure ER, with the Wenner
probe and the two-­point uniaxial methods being the most widely used for
surface and bulk ER measurements, respectively.
The Wenner probe (i.e., surface ER) presents a configuration consisting of
four electrodes; these electrodes are evenly spaced and placed on the con-
crete surface. The outer electrodes apply an alternating current to the con-
crete, while the two inner electrodes measure the potential (Figure 4.10b).

Figure 4.10 C oncrete resistivity testing set-­u ps: (a) surface ER and (b) bulk ER.
 Visual inspection and non-destructive testing (NDT) 55

By analysing the difference in potential relative to the applied current, the

resistivity of the concrete surface is determined. The two-­point method (i.e.,
bulk ER) involves placing the concrete specimen between two parallel metal
plates acting as electrodes. It is important to ensure a reliable connection
between the specimen and the electrodes, and this can be enhanced by incor-
porating moist sponges between the metal plates and the specimen. Once
the specimen is appropriately positioned and connected to the electrodes, an
alternating current is applied, and the resulting voltage or potential reduc-
tion, is measured (Figure 4.10a). Detailed information about the testing set-
­up can be found in (Morris et al. 1996).
While ER initially showed promise for assessing ISR-­induced damage,
certain challenges have emerged in practice, including the significant influ-
ence of moisture conditions on the concrete’s ER. Figure 4.11a illustrates the
typical trend of ER versus induced expansion, while Figure 4.11b shows the
relationship between ER and time. Notably, surface ER demonstrates a sen-
sitive correlation with expansion up to a damage level of 0.05%, beyond
which no correlation is observed.

Figure 4.11 E R of non-­d amaged and damaged concrete as a function of (a) expan-
sion (%) and (b) time. (in weeks).
(Adapted from Rivard & Saint-­P ierre 2009.)
56 Internal Swelling Reactions in Concrete

Recently, interesting improvements in the sensitivity and accuracy of sur-

face ER for assessing ISR damage have been demonstrated. However, due to
variations among results, further clarification is still needed regarding
parameters that may directly affect surface ER outcomes. Additionally,
although limited tests have been performed, bulk ER has shown interesting
correlations with microscopic test procedures used to characterize damage
in ISR-­affected concrete, such as the damage rating index (DRI) (Grazia
2023; Strow et al. 2022); the DRI will be discussed in detail in Chapter 5.
Besides its application in assessing ISR-­damaged concrete, ER has proven
useful for evaluating the effectiveness of ISR-­preventive measures, such as
the use of supplementary cementitious materials (SCMs) in concrete.
However, it is important to note that the curing temperature significantly
influences the estimation of ISR-­ induced expansion using surface ER
(Chopperla & Ideker 2022; Wang et al. 2022). Finally, given the variety of
results obtained, further research is still necessary to better understand and
clarify the parameters that directly affect surface ER outcomes when assess-
ing ISR.

4.3.2 Surface thermography (ST)

Surface thermography is a non-­ intrusive and contactless method that
monitors temperature changes on the surface of an object over time using
thermographic technologies. By employing an infrared detector, infrared
thermography captures thermal patterns on the object’s surface, providing
valuable insights into both surface and subsurface irregularities (Rogalski
2011). Concrete surfaces can be accurately evaluated via thermal energy
using emissivity, allowing for the identification of warmer and cooler
regions, as illustrated in Figure 4.12. In the assessment of large concrete
areas, the sun serves as a suitable and cost-­effective heat energy source,
ensuring even heat distribution across the entire surface. It is important to
note that surfaces with varying degrees of roughness may absorb radiation
at different levels.
Defects in concrete, such as delamination, poor consolidation, water infil-
tration or ISR deterioration, introduce heterogeneities that result in distinct
heat flow patterns when compared to sound regions. ISR-­induced cracking
in concrete leads to variations in temperature absorption, providing an
opportunity to identify damaged regions through the analysis of heat flow
across the concrete surface.
Statistical texture analysis is commonly applied, followed by supervised
neural network classification approaches to enhance the accuracy of ther-
mography imaging for assessing ISR-­affected structures. These techniques
enable the extraction of cracking patterns, including measurements of width,
length and area (Kabir 2010b).
While infrared thermography shows promise in aiding routine inspec-
tion for assessing concrete infrastructure affected by ISR, further research
 Visual inspection and non-destructive testing (NDT) 57

Figure 4.12 E ffect of internal anomaly on surface temperature during heat flow.
(a) Hot zone above flow due to inwards heat flow. (b) Cold zone
above flaw due to outwards heat flow.
(Adapted from Malhotra & Carino 2003.)

investigations are necessary to improve the accuracy of outcomes and fully

explore the potential of this technique when combined with appropriate
data treatments.

4.3.3 Ground penetrating radar (GPR)

GPR is a geophysical method utilized to image sub-­surfaces of structures
by employing high-­frequency electromagnetic waves. The antenna used in
geophysical surveys serves the purpose of transmitting short pulses of elec-
tromagnetic energy within a specific frequency range (50–150 MHz) into
the surveyed material. This antenna can be employed in two ways: it can
either be dragged across the surface or attached to a survey vehicle. This
technique capitalizes on the disparities in dielectric constants at material
boundaries, where a portion of the energy is reflected while the remainder
penetrates deeper into the structure. Figure 4.13 demonstrates a scheme of
the GPR test.
GPR has gained widespread adoption as an NDT for assessing various
aspects of concrete structures, including delamination, voids, reinforcing
rebars and concrete thickness (ACI 201.1R 2008).
In the context of ISR-­damaged infrastructures, GPR serves to complement
VI data by assisting in the selection of coring locations and aiding in the
classification of low-­and high-­damage regions within the structure. ISR
58 Internal Swelling Reactions in Concrete

Figure 4.13 R
 eflections of electromagnetic radiation pulse at interfaces between
materials with different relative dielectric constants.
(Adapted from Malhotra & Carino 2003.)

products exhibit distinct dielectric constants compared to concrete hydrated

phases, which are discernible through variations in signal amplitude.
Additionally, the time taken for the wave to bounce back to the receiver is
utilized to determine the depth and location of the boundary, offering a
somewhat accurate estimation of the ISR damage location, as observed in
Figure 4.14.
GPR has proven to be an interesting auxiliary tool for assessing damaged
concrete structures, enabling the estimation of regions impacted by ISR
prior to the manifestation of visual signs. Nonetheless, further improve-
ments in data treatment techniques and additional investigations are neces-
sary to enhance the condition assessment of concrete infrastructure affected
by ISR.

4.3.4 Stress waves
In isotropic and elastic media, stress waves propagate through two principal
modes: dilatational waves, also known as compression waves or P-­waves,
 Visual inspection and non-destructive testing (NDT) 59

Figure 4.14 G
 PR treated outcomes (a) amplitude of dielectric constant and (b)
dielectric constant.
(Adapted from Klysz et al. 2006.)

and distortional waves, known as shear waves or S-­waves. These modes are
distinguished by the direction of particle motion in relation to the propaga-
tion direction of the wavefront. Stress waves can be generated using various
sources to evaluate damaged concrete, such as pulse-­echo, impact-­echo and
impulse-­response techniques (Malhotra & Carino 2003).
The most used method is the pulse echo, also known as ultrasonic pulse
velocity (UPV). This test is performed by placing two transducers on the
concrete surface in specific positions. These positions can be direct, semi-­
direct and indirect, as per Figure 4.15. One transducer will transmit a fre-
quency, usually 54 kHz, and the other transducer will receive it; the more
continuous the media, the higher the wave velocity. The wave mode depends
on the transducer type applied.
60 Internal Swelling Reactions in Concrete

Figure 4.15 S chematic approaches to assess concrete through UPV: (a) direct, (b)
semi-­d irect and (c) indirect.

P-­wave analysis has been widely employed in the assessment of ISR-­

damaged concrete, particularly to appraise damage caused by ASR; its sen-
sitivity is, however, limited, as demonstrated by Figure 4.16. The relationship
between P-­wave velocity and expansion varies depending on factors such as
aggregate type (i.e., fine vs coarse), nature (i.e., lithotype) and amount of
ASR-­secondary products formed. Nonetheless, despite its limitations, the
P-­wave analysis is recommended in various protocols as a complementary
tool for routine inspections (Godart et al. 2013).
Further research is, therefore, to explore the correlation between P-­wave
over ASR damage severity. This will contribute to refining the interpretation
of P-­wave data and enhancing the accuracy of ASR assessment using stress
wave analysis. Additionally, investigating the effect of other parameters,
such as specimen size and testing conditions, on the P-­wave behaviour in
ASR-­damaged concrete would provide valuable insights for the develop-
ment of more reliable assessment protocols.

Figure 4.16 U
 PV (P-­w ave) in assessing ASR-­a ffected concrete laboratory-­m ade
samples over expansion.
(Adapted from Rivard & Saint-­P ierre 2009; Sanchez 2014; Sargolzahi
et al. 2010.)
 Visual inspection and non-destructive testing (NDT) 61

4.3.5 Resonant frequency (RF)

The natural frequency of vibration in an elastic system plays a crucial role
in characterizing its dynamic behaviour since it is closely related to the
dynamic modulus of elasticity and density of the system. By measuring the
natural frequency of vibration, the dynamic modulus of elasticity of materi-
als can be determined. In the assessment of concrete, the RF can be obtained
through forced resonance tests and impact resonance tests (ASTM C 215
2019). RF analysis can be employed to evaluate ISR-­affected concrete using
both laboratory specimens or cores extracted from structures and real-­scale
members. The impact approach has been widely used to assess ISR, which
can be obtained in three modes, as depicted in Figure 4.17.
The sensibility to detect ISR progression via RF, particularly for ASR-­
induced deterioration, has been observed to be influenced by several factors,
such as aggregate’s type (i.e., fine vs coarse), nature (i.e., lithotype), and
amount of ASR-­secondary products formed. Depending on the aggregate’s
lithotype, the reduction trend of the frequency can be observed in low (≈
0.05%) and moderate (≈ 0.12) expansion levels. It is worth noting that no
significant reduction in frequency is observed for ASR-­affected concrete
incorporating fine reactive aggregates, as illustrated in Figure 4.18. However,
it is important to acknowledge that there are limitations to using RF as a
sole indicator for assessing concrete distressed by ASR.
The assessment of real-­scale concrete members affected by ASR using RF
analysis has already been evaluated and yields valuable insights into the
deterioration process. Specifically, the dynamic modulus of elasticity, as
determined through the longitudinal frequency, exhibits a reduction in low-­
age specimens, indicating the sensitivity of RF to ASR progression at the
real-­scale level. However, according to (Siegert et al. 2005), it is important
to note the presence of a rigidity restoration phenomenon attributed to a
healing process which takes place over ASR development as per Figure 4.19.
This restoration in rigidity is associated with the sealing of microcracks and
pores filling by the ASR products, resulting in an overall improvement in the
concrete stiffness. Furthermore, it should be acknowledged that structural
effects, such as load redistribution and stress relaxation, along with environ-
mental conditions (i.e., temperature, moisture, etc.), may also contribute
significantly to the healing process observed in real-­scale members.
Further research is still necessary to enhance the reliability and accuracy
of RF analysis for ISR assessment. The investigation of the impact of various
parameters, such as aggregate features, specimen size and geometry, and
testing conditions on the RF responses is deemed crucial to enable the use of
RF in a more diagnostic manner. Additionally, exploring the correlation
between RF and other mechanical properties of ISR-­affected concrete can
contribute to a more comprehensive understanding of the deterioration pro-
cess and aid in the development of effective evaluation methodologies.
 62
Internal Swelling Reactions in Concrete
Figure 4.17 R F testing set-­u p for impact excitation: (a) transverse mode, (b) longitudinal mode and (c) torsional mode.
 Visual inspection and non-destructive testing (NDT) 63

Figure 4.18 R F as a function of ASR damage degree.

(Adapted from Malone et al. 2021; Rivard & Saint-­
P ierre 2009;
Sargolzahi et al. 2010.)

Figure 4.19 D
 ynamic modulus of elasticity over time of beams made of ASR-­
reactive and non-­r eactive aggregates.
(Adapted from Siegert et al. 2005.)

REFERENCES

ACI 201.1R. (2008). Guide for conducting a visual inspection of concrete in service.
American Concrete Institute.
ACI 224R-­19. (2019). Control of cracking in concrete structures. American Concrete
Institute.
ASTM C 215. (2019). Standard test method for fundamental transverse, longitudinal,
and torsional resonant frequencies of concrete specimens. ASTM International.
Auberg, R., & Setzer, M. (2014). Frost resistance of concrete RILEM workshop on
resistance of concrete to freezing and thawing with and without de-­icing chemi-
cals. CRC Press.
64 Internal Swelling Reactions in Concrete

Billah, U. H., Tavakkoli, A., & La, H. M. (2019). Concrete crack pixel classification
using an encoder decoder based deep learning architecture. In Advances in visual
computing (pp. 593–604). Springer International Publishing.
Carrasco, M., Araya-­Letelier, G., Velázquez, R., & Visconti, P. (2021). Image-­based
automated width measurement of surface cracking. Sensors, 21(22). https://doi.
org/10.3390/s21227534
Casanova, I., Agulló, L., & Aguado, A. (1996). Aggregate expansivity due to sulfide
oxidation — I. Reaction system and rate model. Cement and Concrete Research,
26(7), 993–998. https://doi.org/10.1016/0008-­8846(96)00085-­3
Chinchón, J. S., Ayora, C., Aguado, A., & Guirado, F. (1995). Influence of weathering
of iron sulfides contained in aggregates on concrete durability. Cement and Concrete
Research, 25(6), 1264–1272. https://doi.org/10.1016/0008-­8846(95)00119-­W
Chopperla, K. S. T., & Ideker, J. H. (2022). Using electrical resistivity to determine
the efficiency of supplementary cementitious materials to prevent alkali-­ silica
reaction in concrete. Cement and Concrete Composites, 125, 104282. https://doi.
org/10.1016/j.cemconcomp.2021.104282
David, L., & Gregory, M. (2017). Review of robotic infrastructure inspection sys-
tems. The Journal of Infrastructure Systems, 23(3), 1–12.
Dias, I. S., Flores-­Colen, I., & Silva, A. (2021). Critical analysis about emerging
technologies for Building’s façade inspection. Buildings, 11(2), 1–19. https://doi.
org/10.3390/buildings11020053
Duchesne, J., Rodrigues, A., & Fournier, B. (2021). Concrete damage due to oxi-
dation of pyrrhotite-­bearing aggregate: A review. RILEM Technical Letters, 6,
82–92. https://doi.org/10.21809/rilemtechlett.2021.138
Fasseu, P., & Michel, M. (1997). Détermination de l’indice de fissuration d’un pare-
ment de béton; Méthode d’essai LCPC N0. 47.
Fournier, B., Bérubé, M. A., Folliard, K., & Thomas, M. (2010). Report on
the Diagnosis, Prognosis, and Mitigation of Alkali-­ Silica Reaction (ASR) in
Transportation Structures.
Gattulli, V., & Chiaramonte, L. (2005). Condition assessment by visual inspec-
tion for a bridge management system. Computer-­Aided Civil and Infrastructure
Engineering, 20(2), 95–107.
Geiss, C. E., & Gourley, J. R. (2019). A thermomagnetic technique to quantify the
risk of internal sulfur attack due to pyrrhotite. Cement and Concrete Research,
115, 1–7. https://doi.org/10.1016/j.cemconres.2018.09.010
Godart, B., & Divet, L. (2013). Lessons learned from structures damaged by delayed
ettringite formation and the french prevention strategy. Fifth International
Conference on Forensic Engineering, Institution of Civil Engineers, 389–400.
https://hal.science/hal-­00945667/document
Godart, B., Rooij, M., & Wood, J. G. M. (2013). Guide to diagnosis and appraisal of
AAR damage to concrete in structures: Part 1 diagnosis (AAR 6.1) (RILEM). Springer.
Golden, J., Gomes, E., Anurag, R., & Su, Y. T. (2018). Analysis of a historical failure,
report no. 1: University of California, Berkeley.
Grazia, M. T. (2023). Short and long-­term performance of eco-­efficient concrete mix-
tures [PhD]. University of Ottawa.
Gunn, R.M., Scrivener, K.L., & Leemann, A. (2017). The identification, extent and
prognosis of alkali-­aggregate reaction related to existing dams in Switzerland. In
A. Sellier, É. Grimal, S. Multon, & É. Bourdarot (Eds.), Swelling concrete in dams
and hydraulic structures (pp. 117–143).
 Visual inspection and non-destructive testing (NDT) 65

Hansen, S. G., & Hoang, L. C. (2021). Anisotropic compressive behaviour of con-

crete from slabs damaged by alkali-­silica reaction. Construction and Building
Materials, 267. https://doi.org/10.1016/j.conbuildmat.2020.120377
Henrickson, J. V., Rogers. C., Lu, H. H., Valasek, J., & Shi, Y. (2016). Infrastructure
assessment with small unmanned aircraft systems. 2016 International Conference
on Unmanned Aircraft Systems (ICUAS), 933–942.
Jana, D. (2022). Cracking of residential concrete foundations in eastern Connecticut,
USA from oxidation of pyrrhotite. Case Studies in Construction Materials, 16,
e00909. https://doi.org/10.1016/j.cscm.2022.e00909
Jang, K., Kim, N., & An, Y. K. (2019). Deep learning–based autonomous concrete
crack evaluation through hybrid image scanning. Structural Health Monitoring,
18(5–6), 1722–1737. https://doi.org/10.1177/1475921718821719
Kabir, S. (2010a). Imaging-­based detection of AAR induced map-­crack damage in
concrete structure. NDT and E International, 43(6), 461–469. https://doi.org/
10.1016/j.ndteint.2010.04.007
Kabir, S. (2010b). Imaging-­ based detection of AAR induced map-­ crack dam-
age in concrete structure. NDT & E International, 43(6), 461–469. https://doi.
org/10.1016/j.ndteint.2010.04.007
Karthik, M. M., Mander, J. B., & Hurlebaus, S. (2016a). ASR/DEF related expansion in
structural concrete: Model development and validation. Construction and Building
Materials, 128, 238–247. https://doi.org/10.1016/j.conbuildmat.2016.10.084
Karthik, M. M., Mander, J. B., & Hurlebaus, S. (2016b). Deterioration data of a
large-­scale reinforced concrete specimen with severe ASR/DEF deterioration.
Construction and Building Materials, 124, 20–30. https://doi.org/10.1016/j.
conbuildmat.2016.07.072
Kim, H., Ahn, E., Shin, M., & Sim, S. (2019). Crack and Noncrack classification from
concrete surface images using machine learning. Structural Health Monitoring,
18(3), 725–738. https://doi.org/10.1177/1475921718768747
Kim, J. J., Kim, A. R., & Lee, S. W. (2020). Artificial neural network-­based automated
crack detection and analysis for the inspection of concrete structures. Applied
Sciences (Switzerland), 10(22), 1–13. https://doi.org/10.3390/app10228105
Klysz, G., Lataste, J.-F., Fnine, A., Dérobert, X., Piwakowski, B., & Buyle-­Bodin,
F. (2006). Auscultation non destructive du chevêtre du pont de la Marque (59).
Revue Européenne de Génie Civil, 10(1), 7–24. https://doi.org/10.1080/1774712
0.2006.9692813
Koch, C., Georgieva, K., Kasireddy, V., Akinci, B., & Fieguth, P. (2015). A review on
computer vision based defect detection and condition assessment of concrete and
asphalt civil infrastructure. Advanced Engineering Informatics, 29(2), 196–210.
https://doi.org/10.1016/j.aei.2015.01.008
Leemann, A., Menénendez, E., & Sanchez, L. (2021). Assessment of damage and expan-
sion. In V. E. Saouma (Ed.), Diagnosis & Prognosis of AAR Affected Structures
State-­of-­the-­Art Report of the RILEM Technical Committee 259-­ISR. Springer.
Liu, Y., Yeoh, J. K. W., & Chua, D. K. H. (2020). Deep learning–based enhance-
ment of motion blurred UAV concrete crack images. Journal of Computing in
Civil Engineering, 34(5), 04020028. https://doi.org/10.1061/(asce)cp.1943-­5487.
0000907
Lugg, A., & Probert, D. (1996). ‘Mundic’-type problems: A building material
catastrophe. Construction and Building Materials, 10(6), 467–474. https://doi.
org/10.1016/0950-­0618(95)00095-­X
66 Internal Swelling Reactions in Concrete

Mahadevan, S., Neal, K., Nath, P., Bao, Y., Cai, G., Orme, P., Adams, D., & Agarwal,
V. (2017). Quantitative diagnosis and prognosis framework for concrete degrada-
tion due to alkali-­silica reaction. American Institute of Physics (AIP) Conference
Proceedings, 1806(080006). https://doi.org/10.1063/1.4974631
Malhotra, V. M., & Carino, N. J. (2003). Handbook on nondestructive testing of
concrete (2nd ed.). CRC Press (Taylor & Francis Group).
Malone, C., Zhu, J., Hu, J., Snyder, A., & Giannini, E. (2021). Evaluation of alkali-­
silica reaction damage in concrete using linear and nonlinear resonance techniques.
Construction and Building Materials, 303, 124538. https://doi.org/10.1016/j.
conbuildmat.2021.124538
Martin, R. P., Céline, B., & Toutlemonde, F. (2015). Alkali aggregate reaction and
delayed ettringite formation: common features and differences. May 2012. https://
hal.archives-­ouvertes.fr/hal-­00852367
Miyamoto, A. (2013). Development of a remote collaborative visual inspection sys-
tem for road condition assessment. Key Engineering Materials, 569–570, 135–
142. https://doi.org/10.4028/www.scientific.net/KEM.569-­570.135
Moliard, J.-M., Baltazart, V., Bérenger, B., Perrin, T., & Tessier, C. (2016). Digitized
measurement of the cracking index on the facings of concrete structures. 8th
RILEM International Conference on Mechanisms of Cracking and Debonding in
Pavements, 13, 731–737. https://doi.org/10.1007/978-­94-­024-­0867-­6
Morris, W., Moreno, E. I., & Sagüés, A. A. (1996). Practical evaluation of resistivity
of concrete in test cylinders using a Wenner array probe. Cement and Concrete
Research, 26(12), 1779–1787. https://doi.org/10.1016/S0008-­8846(96)00175-­5
Ogawa, N., Maeda, K., Ogawa, T., & Haseyama, M. (2022). Deterioration level esti-
mation based on convolutional neural network using confidence-­aware attention
mechanism for infrastructure inspection. Sensors, 22(1). https://doi.org/10.3390/
s22010382
Radopoulou, S. C., & Brilakis, I. (2017). Automated detection of multiple pave-
ment defects. Journal of Computing in Civil Engineering, 31(2), 1–14. https://doi.
org/10.1061/(asce)cp.1943-­5487.0000623
Rivard, P., & Saint-­Pierre, F. (2009). Assessing alkali-­silica reaction damage to concrete
with non-­destructive methods: From the lab to the field. Construction and Building
Materials, 23(2), 902–909. https://doi.org/10.1016/j.conbuildmat.2008.04.013
Rodrigues, A., Duchesne, J., Fournier, B., Durand, B., Rivard, P., & Shehata, M.
(2012). Mineralogical and chemical assessment of concrete damaged by the oxida-
tion of sulfide-­bearing aggregates: Importance of thaumasite formation on reac-
tion mechanisms. Cement and Concrete Research, 42(10), 1336–1347. https://doi.
org/10.1016/j.cemconres.2012.06.008
Rodrigues, A., Duchesne, J., Fournier, B., Durand, B., Shehata, M. H., & Rivard, P.
(2016). Evaluation protocol for concrete aggregates containing iron sulfide miner-
als. ACI Materials Journal, 113(3). https://doi.org/10.14359/51688828
Rogalski, A. (2011). Recent progress in infrared detector technologies. Infrared Physics
& Technology, 54(3), 136–154. https://doi.org/10.1016/j.infrared.2010.12.003
Sanchez, L. F. M. (2014). Contribution to the assessment of damage in aging con-
crete infrastructures affected by alkali-­aggregate reaction, PhD. 377. https://api.
semanticscholar.org/CorpusID:139714980
Sargolzahi, M., Kodjo, S. A., Rivard, P., & Rhazi, J. (2010). Effectiveness of non-
destructive testing for the evaluation of alkali-­ silica reaction in concrete.
Construction and Building Materials, 24(8), 1398–1403. https://doi.org/10.1016/j.
conbuildmat.2010.01.018
 Visual inspection and non-destructive testing (NDT) 67

Savino, P., & Tondolo, F. (2021). Automated classification of civil structure defects
based on convolutional neural network. Frontiers of Structural and Civil
Engineering, 15(2), 305–317. https://doi.org/10.1007/s11709-­021-­0725-­9
Siegert, D., Multon, S., & Toutlemonde, F. (2005). Resonant frequencies monitor-
ing of alkali aggregate reaction (AAR) damaged concrete beams. Experimental
Techniques, 29(6), 37–40. https://doi.org/10.1111/j.1747-­1567.2005.tb00245.x
Strow, M., Bevington, P., Bentivegna, A., Bakhtiari, S., Aranson, I., Ozevin, D., &
Heifetz, A. (2022). Monitoring accelerated alkali-­ silica reaction in concrete
prisms with petrography and electrical conductivity measurements. Materials and
Structures, 55(4), 119. https://doi.org/10.1617/s11527-­022-­01942-­8
Thériault, F., Noël, M., & Sanchez, L. (2022). Simplified approach for quantita-
tive inspections of concrete structures using digital image correlation. Engineer­
ing Structures, 252(June 2021), 1–12. https://doi.org/10.1016/j.engstruct.2021.
113725
Thomas, M., Folliard, K., Drimalas, T., & Ramlochan, T. (2008). Diagnosing delayed
ettringite formation in concrete structures. Cement and Concrete Research, 38(6),
841–847. https://doi.org/10.1016/j.cemconres.2008.01.003
Valença, J., Gonçalves, L. M. S., & Júlio, E. (2013). Damage assessment on concrete
surfaces using multi-­spectral image analysis. Construction and Building Materials,
40, 971–981. https://doi.org/10.1016/j.conbuildmat.2012.11.061
Vlahović, M. M., Savić, M. M., Martinović, S. P., Boljanac, T. D., & Volkov-­Husović,
T. D. (2012). Use of image analysis for durability testing of sulfur concrete
and Portland cement concrete. Materials and Design, 34, 346–354. https://doi.
org/10.1016/j.matdes.2011.08.026
Wang, Y., Ramanathan, S., Chopperla, K. S. T., Ideker, J. H., & Suraneni, P. (2022).
Estimation of non-­ traditional supplementary cementitious materials poten-
tial to prevent alkali-­silica reaction using pozzolanic reactivity and bulk resis­
tivity. Cement and Concrete Composites, 133, 104723. https://doi.org/10.1016/j.
cemconcomp.2022.104723
Wood, J. G. M. (2008). Implications of the collapse of the de La Concorde overpass.
4th International Conference on ‘Forensic Engineering’, 16–18.
Zahedi, A., Sanchez, L.F.M., & Noël, M. (2022). Appraisal of visual inspection tech-
niques to understand and describe ASR-­induced development under distinct con-
finement conditions. Construction and Building Materials, 323, 126549. https://
doi.org/10.1016/j.conbuildmat.2022.126549
Zhang, A., Wang, K. C. P., Fei, Y., Liu, Y., Tao, S., Chen, C., Li, J. Q., & Li, B. (2018).
Deep learning–based fully automated pavement crack detection on 3D asphalt
surfaces with an improved CrackNet. Journal of Computing in Civil Engineering,
32(5), 1–14. https://doi.org/10.1061/(asce)cp.1943-­5487.0000775
Zhang, A., Wang, K. C. P., Li, B., Yang, E., Dai, X., Peng, Y., Fei, Y., Liu, Y., Li,
J. Q., & Chen, C. (2017). Automated pixel-­level pavement crack detection on
3D asphalt surfaces using a deep-­learning network. Computer-­Aided Civil and
Infrastructure Engineering, 32(10), 805–819. https://doi.org/10.1111/mice.12297
Zhong, R., & Wille, K. (2018). Deterioration of residential concrete foundations:
The role of pyrrhotite-­bearing aggregate. Cement and Concrete Composites, 94,
53–61. https://doi.org/10.1016/j.cemconcomp.2018.08.012
Zhu, Z., German, S., & Brilakis, I. (2011). Visual retrieval of concrete crack proper-
ties for automated post-­earthquake structural safety evaluation. Automation in
Construction, 20(7), 874–883. https://doi.org/10.1016/j.autcon.2011.03.004
Chapter 5

Microscopic analyses

5.1 INTRODUCTION

As previously discussed in Chapter 4, damage at the surface of concrete

detected through visual inspection (VI) methods (i.e., qualitative descrip-
tions and cracking index – CI) and non-­destructive testing (NDT) can give
a first glance at its condition. Mechanical testing subsequently corresponds
to the material’s macro performance and ability to continue withstanding
loads (see Chapter 6). Although the exact cause of damage may not neces-
sarily be captured by VI, NDT and certain mechanical tests, it is widely
accepted that a combination of microscopy tools (i.e., qualitative and quan-
titative) can offer insight into the cause(s) leading to concrete deterioration
(e.g., internal swelling reaction (ISR) type) along with the damage degree of
the affected material. This chapter intends to display the common qualita-
tive and quantitative microscopic procedures used to assess the cause and
extent of ISR damage in concrete in both micro and mesoscales.

5.2 PETROGRAPHIC ANALYSIS: ASSESSING THE

DAMAGE CAUSE(S)

5.2.1 Introduction
Petrographic analysis refers to the study of concrete (or rock) thin or pol-
ished concrete sections, allowing a petrographer to reveal certain character-
istics of the concrete (or rock) microstructure, such as aggregates mineralogy,
various phases of cement hydration and secondary reaction products. For
many years, the science of petrography has played an important role in con-
crete research, especially in the condition assessment of damaged concrete
(Jana, 2005). Over time, petrography became widely used in the concrete
construction industry with the aim of

a. characterizing concrete’s composition, mineralogical and textural proper-

ties, including aggregate lithotypes, size and shape, cement paste features,
presence of supplementary cementing materials and chemical admixtures

68 DOI: 10.1201/9781003188155-5
 Microscopic analyses 69

(Erlin & Stark, 1990) as per the Standard Guide for Petrographic
Examination of Aggregates for Concrete (ASTM C295, 2019);
b. providing a quality control tool for concrete construction (i.e., incor-
rect proportioning, mixing, placement, finishing or curing opera-
tions) as per the Standard Practice for Petrographic Examination of
Hardened Concrete (ASTM C856-­20, 2020);
c. diagnosing the cause(s) of concrete damage due to deleterious chemi-
cal (e.g., acid, alkali, sulphate, chloride, seawater attack) and physical
(e.g., frost, fire) mechanisms (Erlin & Stark, 1990; A. B. Poole & Sims,
2016; Walker et al., 2006); and
d. appraising the efficiency of rehabilitation strategies (e.g., repair materials).

Petrographic evaluations are generally conducted using optical microscopes

(OM), scanning electron microscopes (SEM) or both (ASTM C295, 2012;
ASTM C856-­20, 2020; ASTM C1723-­16, 2016; A. B. Poole & Sims, 2016;
Walker et al., 2006). Chemical compositions of distinct concrete microstruc-
ture phases can also be determined using energy-­dispersive X-­ray spectros-
copy (EDS/EDX). Petrography should be conducted by highly qualified,
skilled and experienced petrographers as per American standards to be reli-
able (ASTM C295, 2019; ASTM C856-­20, 2020). Moreover, it is suggested
that a petrographer be familiar with the local geology of the concrete struc-
ture under evaluation (Fernandes et al., 2016).

5.2.2 Sample preparation for petrography

Concrete specimens, either cores extracted from concrete structures or
structural members or even manufactured in the laboratory, are sized to
adequately represent the region of interest for analysis. Initially, cores are
cleaned after their extraction and then wrapped in self-­adhesive film and/or
sealed bags to prevent dehydration and further deterioration (Broekmans,
2012; CUR-­Recommendation-­102, 2008). The specimens are then cut to
size (i.e., thin or polished sections), cleaned and dried. Afterwards, the sec-
tions are impregnated with a clear or fluorescent dye epoxy (ASTM C856-­
20, 2020; Broekmans, 2012; CUR-­Recommendation-­102, 2008; Poole &
Sims, 2016) to create a contrast using ultraviolet (UV) lights where the resin
is deposited, such as in voids, porous media and cracks. Finally, the samples
are further ground/polished to achieve minimal topographical variations
and light scattering, thus enhancing the distinction between concrete com-
ponents under SEM and X-­ray microanalyses (Goldstein et al., 2003).
Figure 5.1 illustrates conventional (i.e., 50 mm by 28 mm) and large-­size
(i.e., 50 mm by 100 mm) thin sections (A and B, respectively) along with
mounted and impregnated polished concrete sections (C and D) prepared
for petrographic analysis. Several guides are readily available for further
reference on aggregate and concrete petrography (Fernandes et al., 2016;
Ingham, 2013; Poole & Sims, 2016).
70 Internal Swelling Reactions in Concrete

Figure 5.1 Images taken with a flatbed scanner of (a) conventional size thin sec-
tion of 50 mm by 28 mm, (b) large thin section of 50 mm by 100 mm,
(c) 4 cm diameter concrete section and (d) 2.5 cm diameter mounted
concrete section.
(Photographs courtesy of Andreas Leemann, Leandro Sanchez,
Cassandra Trottier.)

5.2.3 Evaluation of ISR in concrete through

petrography
The first step in the evaluation of ISR-­induced deterioration in concrete is
the selection of regions of interest (ROI) in the sample under analysis (ASTM
C856-­20, 2020); normally, this selection is made via a thorough appraisal
using the stereomicroscope. Figure 5.2 illustrates a stereomicroscope and a
prepared specimen with highlighted ROI.
Once the ROI is defined, an optical microscope is used to detect the pres-
ence of reactive minerals within the aggregates, reaction products, mixture
proportions (estimated range of water-­to-­cement ratio, air-­void characteris-
tics), etc. This appraisal is normally conducted through cross-­ polarized
transmitted light or reflected light at magnifications up to 600x (Jana, 2005).
Complimentary evaluations under the SEM with EDS/EDX might be per-
formed; the latter allows for microscopic examinations at higher magnifica-
tions, along with providing chemical composition analysis. Figure 5.4a and
 Microscopic analyses 71

Figure 5.2 S tereomicroscope equipped with an automated stage for automatic

image acquisition and linear traverse stepwise operations used for
preliminary analysis and selection of ROI on polished concrete
sections.
(Photograph courtesy of Francisco Locati and Cassandra Trottier.)

b illustrate secondary reaction products of ettringite and alkali-­silica reac-

tion (ASR) with their corresponding EDS/EDX, respectively. For example,
ettringite may have the appearance of needle-­like crystals filling or lining
empty spaces in the concrete microstructure, presenting silica, alumina, and
calcium as the main oxides (Melo et al., 2023; Thomas et al., 2008).
Otherwise, ASR secondary products (i.e., the so-­called ASR “gel”) may dis-
play various morphologies (i.e., amorphous or crystalline) and varied chem-
ical composition according to its age and location in the concrete
microstructure (higher amounts of calcium are verified in regions closer to
the cement paste); nevertheless, silica, sodium, potassium, and calcium are
the main oxides present in ASR secondary products (Leemann et al., 2016).
Ultimately, to confirm whether the products encountered via EDS/EDX are
the ones suspected, their atomic mass percentages may be evaluated through
ratio plots, as illustrated in Figure 5.3a and b, in the case of secondary ettr-
ingite and ASR, respectively. If the ratios measured via EDS/EDX are close
to (or move towards) the ettringite (Ett.) dot in Figure 5.4a or show total
alkalis to silica ratio (i.e., Na+K+Cs/Si) values between 0.20 and 0.35 as per
Leemann et al. and Ahmed et al., 2022 in Figure 5.4b, then these products
are confirmed to be secondary ettringite or ASR products, respectively. The
previous discussion highlights the great importance and reliability of petro-
graphic analysis to assess the cause(s) of deterioration of concrete infra-
structure; nevertheless, the following limitations still remain during the use
of this technique: (a) high expertise is required to properly prepare
72 Internal Swelling Reactions in Concrete

Figure 5.3 S EM images of (a) ettringite-­filled crack in the aggregate (shown with
solid arrows) extending into the cement paste (shown with dashed
arrow), (b) ettringite lining a void at the top left corner (shown with
dashed arrow) and surrounding an aggregate particle (shown with
solid arrows), (c) EDS/EDX analysis showing the presence of ettrin-
gite (Loser & Leemann, 2016) and (d) EDS/EDX analysis showing the
presence of ASR secondary products.
(Figure courtesy of Leandro Sanchez).

specimens, conduct the analysis and evaluate outcomes to reduce subjectiv-

ity; (b) the specimen size assessed may not be representative to explain the
whole deterioration process observed in the concrete structure or structural
member under analysis; (c) although reliable to assess the cause(s) leading to
deterioration, conventional petrography provides limited quantitative infor-
mation on the extent of deterioration experienced by the affected concrete.
Such information is crucial to help engineers and infrastructure owners
make better decisions in terms of management and rehabilitation protocols
(Hollis et al., 2006; Poole & Sims, 2016). In this context, a number of
microscopic techniques have been proposed over the last years to address
the aforementioned drawbacks, and they will be discussed in the following
section.
 Microscopic analyses 73

Figure 5.4 C
 rack pattern in concrete affected by (a) ASR reactive coarse aggre-
gate, (b) ASR reactive sand, (c) FT cycles and (d) DEF. The concrete's
surface is shown on the left-­h and side.
(Adapted from British Cement Association (BCA), 1992.)

5.3 PETROGRAPHIC ANALYSIS: ASSESSING THE

DAMAGE DEGREE

5.3.1 Introduction
Understanding the petrographic distress features and degree of concrete is
a matter of scale; depending on the evaluation size, different microstructure
features (i.e., cracks, defects, flaws, etc.) are observed, yet the analysis scale
should be selected as per the examiner’s purpose. In Chapter 3, damage
was globally defined as (1) engineering properties reductions (i.e., compres-
sive, tensile and direct shear strengths) of concrete, (2) stiffness reduc-
tion of concrete and (3) physical integrity and/or durability loss, which is
directly related to the existing crack network. Therefore, whether or not
74 Internal Swelling Reactions in Concrete

petrographic analysis is to be used to evaluate damage, it should be able

to correlate the observed petrographic distress features with the previously
described damage parameters. In this context, the literature demonstrates
that petrographic distress features observed at the mesoscale (i.e., polished
concrete sections at about 15x–20x magnification) may properly correlate
with macro performance when compared to standardized mechanical and
durability testing protocols (Sanchez et al., 2017, 2018). Furthermore, at
the mesoscale, different damage patterns may be observed as per the dete-
rioration mechanism(s) leading the overall damage process, which may not
only facilitate the detection of the may cause(s) of damage but actually help
understand the potential impact on the engineering properties and durabil-
ity of the material. For instance, Figure 5.4 illustrates common petrographic
damage features and patterns observed for distinct ISR mechanisms. For
ASR-­induced deterioration, cracks originate and are mainly found within
the aggregate particles, either coarse (Figure 5.4a) or fine (Figure 5.4b).
Otherwise, both freeze and thaw (FT) and DEF induce cracks in the cement
paste, where FT-­induced cracks are present in the bulk cement paste and
pores, being fairly parallel to the concrete’s surface in locations exposed
to the environment, while DEF-­induced cracks are more located in the
ITZ and connect to one another in the cement paste at high deterioration
levels. A number of petrographic protocols at the mesoscale have been
developed over the past decades to semi or fully quantitatively appraise
deterioration in concrete. Amongst them, the Damage Rating Index (DRI),
a semi-­quantitative petrographic procedure and the image analysis, a fully
quantitative approach, seem to be the most effective. The next section will
present these methods along with some emerging technologies to quantita-
tively evaluate ISR-­induced deterioration in concrete.

5.3.2 Damage rating index (DRI)

The DRI method consists of analysing 1 cm2 grids (field of view at 15x–16x
magnification) defined at the surface of polishing concrete sections using a
stereomicroscope (reflected light). In each square, various types of petro-
graphic damage features are counted and multiplied by weighting factors
whose main purpose is to balance their relative importance towards the
overall deterioration of the concrete. The previous calculation (i.e., num-
ber of cracks multiplied by the respective weighting factors) corresponds to
the DRI number; the higher the DRI number, the higher the deterioration
of the concrete. Ideally, a surface of at least 200 cm2 should be used for
analysis, and it may be greater in the case of mass concrete incorporating
coarse aggregates with larger size fractions. Nonetheless, for comparative
purposes, the final DRI number is normalized to a 100 cm2 area (Figure 5.5).
The DRI was initially idealized by (Grattan-­Bellew & Danay, 1992) for
assessing the damage caused by ASR; therefore, the first weighting factors
were proposed to indicate the level of development, or progress, of ASR in
 Microscopic analyses 75

Figure 5.5 A prepared specimen for DRI analysis.

(Photograph courtesy of Cassandra Trottier.)

concrete (Grattan-­Bellew & Danay, 1992). Over time, the weighting factors
were modified so that the DRI number could better correlate with induced
expansion, physical integrity and engineering properties reductions of ISR-­
affected concrete (Sanchez et al., 2015, 2017, 2018, 2020; Villeneuve, 2011).
The types of cracks appraised in the DRI method are described in Table 5.1,
along with their respective weighting factors. It is worth noting that higher
weighting factors are attributed to features deemed more important. For
instance, a closed crack in the aggregate may not represent ISR damage, as
it is most likely produced by rock weathering and aggregate processing;
hence, a factor below unity (i.e., 0.25) is attributed to this feature. Otherwise,
an opened crack in the aggregate (with or without secondary products) can
indicate that a reactive aggregate is present or even damage in the aggregate
has occurred due to pressure from its surroundings; a factor of 2 is attrib-
uted to such cracks. One should note that secondary products can be washed
away during polishing, besides being difficult to observe at 15x–16x magni-
fication. Thus, the same factor has been attributed to open cracks in the
aggregates with and without products. Moreover, a disaggregated/corroded
particle has the same weight as cracks within aggregate particles since the
damage remains within the particle and is of rare occurrence yet may indi-
cate the presence of ASR.
Cracks in the cement paste represent the highest level of damage with a
weighting factor of 3. Such cracks indicate either the presence of mecha-
nisms generating and propagating damage in the cement paste, such as DEF
and FT, or the presence of a well-­developed ASR, where its cracks have
elongated from the aggregate particles into the cement paste. Like the reac-
tion products within the aggregate particles and for the same reasons, the
cement paste cracks with and without reaction products are weighted the
same. Finally, the debonded aggregate feature, which also bears a weighting
factor of 3, is quite common for mechanisms generating cracks in the cement
paste, especially DEF, while remaining a rare occurrence in ASR-­affected
concrete.
The DRI has been shown over the last years to be a reliable tool to
appraise damage caused by ISR in concrete (Sanchez et al., 2018, 2020).
76 Internal Swelling Reactions in Concrete

Table 5.1 The DRI weighting factors associated with ISR damage

Distress feature Description Acronym Weighting factor

Closed crack in A crack in which no opening is

CCA 0.25
aggregate observed
Opened crack in A crack with an opening
OCA 2
aggregate between its edges
Opened crack in A crack with an opening
aggregate with between its edges and filled OCA-­RP 2
reaction product with a substance
Disaggregated/ A reacted particle in which its
corroded centre has been eroded and DAP 2
aggregate particleb removed
Crack in the cement An open crack propagating
CCP 3
paste through the cement paste
Crack in the cement An open crack filled with
paste with a substance propagating CCP-­RP 3
reaction product through the cement paste
Coarse aggregate A crack highlighting the
CAD 3
debondinga aggregate-­paste interface
a A debonded aggregate is considered when a crack in the ITZ surrounds 50% of an ASR reactive
particle. When DEF is suspected, a debonded aggregate is considered when the crack in the ITZ
surrounds 75% of the aggregate.
b A disaggregated/corroded particle must not contain a countable crack network.

However, its non-­negligible subjective character, which may cause variabil-

ity in the test outcomes, is still criticized by engineers and practitioners.
Amongst the aspects that may potentially bring important subjectivity and
variability to the method, the distinct crack features recognition and count
are the most important. Otherwise, like conventional petrography, the abil-
ity to recognize different crack features comes with practice and training.
Figure 5.6 shows a few of the distress features observed through a stereomi-
croscope at 16x magnification. It is important to note that the count of
cracks may be quite complicated, especially in systems containing important
crack networks, such as illustrated in Figure 5.7. In these systems, a node
and segment method may be used to count the cracks. However, other tech-
niques may also be used, such as defining a starting point (i.e., reference)
and following the cracks from this point onwards until their ends; both
methods result in the same counts. In Figure 5.7, one verifies the node and
segment calculation where the colour coded lines reveal distinct cracks (i.e.,
segments: 1-­2-­3, 9-­6, 4-­5, 7-­8, 10-­11-­12, 13, 15 and 14-­16; nodes: 1 to 8).
Once all cracks observed on the sample are counted, they are summed
and normalized to 100 cm2, as per Equation 5.1. It is suggested to electroni-
cally tabulate the counts using software to calculate the final DRI number
and produce the corresponding bar charts, illustrated in Figure 5.8.
 Microscopic analyses 77

Figure 5.6 M
 icrographs of 1 cm 2 presenting distress features counted through
the DRI of ISR in concrete: (a) concrete affected by ASR with open
cracks in the aggregate with gel (OCAG), closed crack in the aggregate
(CCA) and crack in the cement paste (CCP), (b) concrete affected by
alkali-­c arbonate reaction with an open crack in the aggregate (OCA)
and a crack in the cement paste, (c) concrete affected by FT with
cement paste crack (CCP) parallel to the exposed surface and (d)
DEF affected concrete with open cracks in the aggregate (OCA) and
cracks in the cement paste with and without reaction product in the
ITZ.
(Photographs courtesy of Andisheh Zahedi and Cassandra Trottier.)

DRI 
(
 ∑ 0.25 (CCA ) + 2 (OCA + OCARP + DAP ) + 3 (CCP + CCPRP + CAD) )  × 100 cm 2
 Number of analysed 1 cm by 1 cm squares 
 
(5.1)

The bar chart in Figure 5.8 illustrates not only the DRI number (i.e.,
x-­axis) but also the distinct crack features (i.e., distinct colours) observed,
78 Internal Swelling Reactions in Concrete

Figure 5.7 Point count method for crack networks.

(Adapted from Sims et al., 1992.)

Figure 5.8 D RI bar chart comparing ASR and DEF.

which gives a better understanding of the mechanism(s) leading to the dete-

rioration of the affected concrete. For example, DEF generates an important
number of cracks in the cement paste (CCP – checker chart), several of them
causing debonding of the aggregates from the cement paste (CAD – dotted
chart). Otherwise, ASR generates cracks within the aggregate particles
(OCA – diagonal striped chart), some of them presenting secondary prod-
ucts (OCARP – vertical stripe chart); upon development, some of these cracks
elongate to the cement paste, generating cracks with (CCPRP – diamond
chart) and without (CCP – checker chart) reaction products.
Other than the DRI bar charts, the DRI can be performed in its extended
version (Sanchez et al., 2015). In this extended method, a thorough analysis of
the distinct petrographic damage features is performed in both counts/100
cm2 and percentages without the use of the weighting factors. Figure 5.9 illus-
trates plots in counts (Figure 5.9a) and percentages (Figure 5.9b). These visual
representations help to distinguish the leading crack types, which may be
composed of (i) CCA, (ii) OCA+OCARP, (iii) CCP+CCPRP and (iv) CAD. In
addition, the crack density as the summation of OCA+OCARP and CCP+CCPRP
counts per 1 cm2 can be plotted as a function of expansion (Figure 5.9c).
 Microscopic analyses 79

Figure 5.9 E xtended version of the DRI with (a) crack counts per 100 cm 2 , (b)
proportions and (c) crack density.
(Sanchez et al., 2020).

The annotation, description and record of other petrographic features

that are observed while gathering the crack counts for the DRI calculation
is highly recommended to supplement the investigation into the cause of the
damage. These features can include but are not limited to the deposit of
product in voids, reaction rims where a dark rim is observed outlining the
aggregate particles, the crack pattern, the identification of extensions such
as cracks extending from aggregate to cement paste and so on. Some of the
features may not necessarily represent damage to concrete but can indicate
the presence of ISR.
80 Internal Swelling Reactions in Concrete

5.3.2.1 Sample preparation for DRI

Sample preparation for the DRI procedure is fairly simple and follows steps
that are common for microscopy. Initially, a concrete core or specimen is
cut in half longitudinally using a masonry saw equipped with a diamond
blade, either unnotched or notched (Figure 5.10a). The masonry saw can
include a water kit or other coolants, such as paraffin oil or kerosene; care
must be taken when using coolants other than water and all safety protocols
followed. It is suggested to cut the concrete core or specimen in one single
motion to reduce deep saw marks on the finished surface (Figure 5.10b).
The cut is then followed by a series of subsequent grinding/polishing stages
to achieve a flat and highly reflective surface. A suggested sequence of lap-
ping disks from coarsest to finest is as follows: 30 grit (to remove saw marks
if necessary), 60 grit, 140 grit, 280 grit (80–100 μm), 600 grit (20–40 μm),
1,200 grit (10–20 μm) and 3,000 grit (4–8 μm). However, this series is based
on the supplier and type of abrasive material used. Figures 5.11a and b

Figure 5.10 D
 RI sample cutting: (a) longitudinal cutting of concrete specimen
using a wet masonry saw equipped with a 14-­i nch diamond blade and
(b) saw marks visible on the surface after cutting.
(Photographs courtesy of Cassandra Trottier.)

Figure 5.11 (a) 18-­

i nch magnetic diamond-­ i mpregnated laps, (b) mechanically
rotating lapping machine and (c) handheld polisher.
(Photographs courtesy of Cassandra Trottier.)
 Microscopic analyses 81

show 18-­inch lapping disks and a mechanically rotating wheel, respectively,

while Figure 5.11c shows a handheld mechanical polisher. A glass plate can
also be used onto which abrasive powders are scattered (Figure 5.12a) and
mixed with a lubricant such as water to produce a grinding/polishing paste
(Figure 5.12b). Manual random circular motions are then used while apply-
ing some pressure and ensuring that the entire surface is ground/polished
(Figure 5.12c). The time required for each step is dependent on the equip-
ment/method used and the types of materials comprised in the concrete. In
any case, the surface is cleaned (using a soft bristle brush or compressed air)
after each step to remove the slurry (Figure 5.12d), any loose abrasives or
aggregate/cement paste fragments that may have become dislodged during
sample preparation. A quick examination of the surface quality after each
step can be performed by reflecting sunlight or indoor ceiling lights onto
the surface to observe if any portion of the surface lacks reflectiveness, as
shown in Figure 5.15a, where the centre portion of the surface is duller and
whiter than the sides. Moreover, deep scratches or marks from previous
steps visible on the aggregates with the naked eye should be removed prior
to the polishing stages. Marks from previous steps should be progressively
less visible on the aggregates as the surface becomes ready for the next step.
If the aggregates have the appearance of being scuffed when using impreg-
nated disks, it is an indication that more lubricant, such as water, is required
during the polishing step.
When the surface of the section is considered suitable for analysis, having
a highly reflective surface (Figure 5.13b), a grid of 1×1 cm squares is drawn
on the surface using a thin tip permanent marker; an alternative to drawing
and permanently marking the surface of the polished concrete section under
analysis is the use of a 3D printed grid that is simply placed on the speci-
men’s surface, as shown in Figure 5.13c. Stiff gardening mesh with a 1 cm
by 1 cm grid can also be used. Generally, the first row is used to number the

Figure 5.12 A
 brasive powder (a) scattered onto a wet glass plate, (b) mixed by
hand to form a slurry, (c) placing of the concrete specimen with its
working surface against the glass plate and (d) the excess of slurry
onto the working surface prior to cleaning.
(Photographs courtesy of Cassandra Trottier and Francisco Locati.)
82 Internal Swelling Reactions in Concrete

Figure 5.13 A
 concrete surface (a) not well finished where a non-­reflective area
is observed at the centre of the surface shown as a whiter colour
compared to both sides, (b) a well-­finished, highly reflective surface
with a grid drawn where light reflection is visible in the bottom right
corner and numbering convention shown on the first row and col-
umn and (c) with a 3D-­p rinted grid placed over the finished surface.
(Photographs courtesy of Cassandra Trottier.)

columns (for example, 1 to 20, as Figure 5.13b illustrates), while the first
column is used to label the rows (for example, A to I, as Figure 5.13b
illustrates).

5.3.3 Image analysis
Image analysis (or digital image processing – DIP) can also be applied to
quantify damage in concrete. Once captured, images are analysed either
manually or using software where quantitative information pertaining to the
deterioration characteristics is extracted, such as the crack length, width and
pattern (i.e., isotropic vs anisotropic), amongst many others often observed
in 2D images (a discussion on 3D images is presented in Section 5.4). Cracks
characteristics such as location, orientation and pattern with respect to (a)
the loading direction of the structural member and (b) the exposed surface
and the presence of reaction products are to be considered when identifying
the cause(s) and extent of damage.
Image analysis is normally used, as the DRI, after conventional petrogra-
phy, where the main cause(s) leading to deterioration is found. Impregnation
is recommended using a fluorescent dye to highlight the cracks encountered
and thus facilitate precise quantification of the petrographic damage features
(providing greater contrast between cracks, voids and flaws and the back-
ground). Figure 5.14 illustrates concrete polished samples impregnated with
a fluorescent dye, where Figure 5.14a shows an ASR specimen presenting a
 Microscopic analyses 83

Figure 5.14 F luorescent epoxy impregnated 5 cm by 10 cm concrete section illu-

minated by UV lights: (a) low ASR-­induced expansion specimen (i.e.,
0.05%) and (b) high ASR-­induced expansion specimen (i.e., 0.20%).
(Photograph courtesy of Cassandra Trottier and Leandro Sanchez.)

low deterioration, and Figure 5.14b illustrates an ASR sample displaying a

high deterioration.
The outcomes of image analysis, such as crack length, width, pattern,
orientation and density, can be associated with (a) the level of damage and
(b) the impact of the observed cracks on the mechanical properties, stiffness
and durability of the ISR-­affected concrete. Crack lengths and orientation
were found to be the best computable indicators to estimate induced expan-
sion caused by ASR when evaluating laboratory and field specimens (Rivard
et al., 2000). Meanwhile, the “Norwegian crack index” applied to ASR-­
affected concrete where a plane polished surface is impregnated with fluo-
rescent material and illuminated with UV light without using a microscope,
in which the proportions of the cracks in the aggregate particles and cracks
extending into the cement paste from the aggregates are counted and
summed over 100 cm2, was used to estimate ASR-­ induced expansion
(Jensen & Sujjavanich, 2016; Lindgård et al., 2004, 2012).

5.3.3.1 Sample preparation for image analysis

The sample preparation for image analysis consists of applying a fluorescent
dye to enhance the appearance of cracks under UV light. Specimens are cut
axially and ground flat. The fluorescent epoxy resin is then applied to the
flat dry surface (Figure 5.15a), placed under a vacuum to ensure that cracks
are filled (Figure 5.15b) and left to cure as per the manufacturer’s details.
Figure 5.15c shows concrete sections with hardened/cured fluorescent
epoxy prior to it being removed to expose the cracked surface. Once the
coat has hardened, it is carefully removed by following a similar grinding/
polishing sequence as for the DRI while ensuring that the resin-­filled cracks
remain at the surface without removing the impregnated cracked material.
If cracks are previously filled with reaction products, it may not be feasible
to fill such cracks with resin, and therefore, this should be considered dur-
ing the analysis. An important note when using fluorescence is to ensure
84 Internal Swelling Reactions in Concrete

Figure 5.15 F luorescent epoxy resin (a) being applied to a concrete section, (b)
vacuum chamber and (c) hardened/cured surfaces ready for removal.
(Photographs courtesy of Cassandra Trottier and Francisco Locati.)

that the proportions of fluorescent powder and resin are compatible with
the UV light source. Trials may be required to determine the proportions
unless a pre-­mixed fluorescent resin is used. A flatbed scanner can further
be used for image analysis as an effective approach to acquire an intact
image (i.e., without stitching several images) and has been previously used
to evaluate the amount of entrained air in concrete (Fonseca & Scherer,
2015; Song et al., 2017). An example of an image taken with a flatbed
scanner is shown in Figure 5.5.

5.4 EMERGING TECHNOLOGIES TO QUANTIFY

DAMAGE IN CONCRETE

Evidently, the digitization of images used for image analysis leads to auto-
mating the manual classification of petrographic damage features and the
mathematical interpretation of extracted data. Although many of the proce-
dures of image analysis are automated to filter the images and remove noise
and unwanted features, such as voids, stitching and thresholding, amongst
many other tedious tasks, very few automated protocols have been devel-
oped to quantitatively assess ISR-­induced deterioration in concrete (Rivard,
1998). Dehghan et al. (2016) used the DRI on digital images taken with two
methods: (a) a flatbed scanner and (b) a stereomicroscope equipped with a
digital camera, which is a form of image analysis in a point count or segmen-
tation sense. However, with the rise in new technologies, more sophisticated
image acquisition equipment and availability of previously written scripts,
it becomes less of a challenge to automate image analysis procedures. Some
works have used artificial intelligence to identify cracks and classify them
based on the DRI’s classification system (Bezerra, 2020), which shows a clear
contrast more easily detectable when compared to fluorescence, as shown in
Figure 5.16. Moreover, since the arrival of artificial intelligence techniques,
 Microscopic analyses 85

Figure 5.16 M
 icrographs of 1 cm 2 polished section from a Robert Bourassa-­
Charest overpass extracted core on the left-­h and side and traced
cracks using an automated procedure (Bezerra, 2020) on the right-­
hand side: (a) shows CCP and ITZ, while (b) shows cracks mainly in
the aggregate with the reaction product.
(Photos courtesy of Agnes Bezerra, Haoye Lu and Cassandra Trottier.)

object recognition in images has been continuously improving, which could

potentially replace the time-­consuming sample preparation used to enhance
the contrast between the cracks and other components of the concrete (i.e.,
cement paste, voids and aggregate) while reducing the potential for induced
shrinkage cracks often observed after drying the concrete specimens before
application of fluorescent dyes.
Three-­dimensional image analysis using micro-­ computed tomography
(micro-­CT) has begun to gain more interest in the diagnosis of ISR-­affected
concrete, particularly to measure cracks generation and propagation, along
with reaction product formation. As such, a number of researchers have used
micro-­CT to evaluate ASR in mortar specimens (Hernández-­Cruz et al., 2016;
Marinoni et al., 2009; Voltolini et al., 2011), whereas others have evaluated
the progression of ASR in concrete over time (Shakoorioskooie et al., 2021).
Moreover, Suzuki et al. (2017) CT scan appraised cores extracted from con-
crete columns affected by FT to estimate the induced deterioration in correla-
tion with acoustic emission. Likewise (Joshi et al., 2022), studied the influence
86 Internal Swelling Reactions in Concrete

of DEF on concrete under compressive loads and quantified its crack propa-
gation via CT scan. These works evidently highlight the ability of micro-­CT
to detect damage caused by various ISRs. Therefore, tomography techniques
have immense potential to provide 3D images of ISR-­affected concrete with-
out the application of techniques to estimate the 3D representation from 2D
petrographic damage features. Yet, the limitations of tomography remain
with the ability to resolve small objects, such as the width of cracks below its
spatial resolution and the low contrast between the cracks with or without
reaction products and the bulk cement paste. Leemann and Münch (2019)
observed that the addition of caesium and barite to fresh concrete could help
enhance contrast and thus facilitate recognition of ASR secondary products
within aggregate particles or cement paste, respectively. Otherwise, contrast
remains a challenge for other ISR mechanisms. Nonetheless, further research
is still required to improve the current limitations of micro-­CT scan, along
with enhancing its potential to become a reliable technique to quantitatively
appraise induced expansion and damage in concrete caused by ISR.

REFERENCES

Ahmed, H., Zahedi, A., Sanchez, L. F. M., & Fecteau, P.-L. (2022). Condition assess-
ment of ASR-­affected reinforced concrete columns after nearly 20 years in ser-
vice. Construction & Building Materials, 347, 128570. https://doi.org/10.1016/
j.conbuildmat.2022.128570
ASTM C295. (2012). Standard guide for petrographic examination of aggregates for
concrete, ASTM International, West Conshohocken (USA).
ASTM C295. (2019). Standard guide for petrographic examination of aggregates
for concrete.
ASTM C856-­20. (2020). Standard practice for petrographic examination of hard-
ened concrete. In Annual book of ASTM standards. https://doi.org/10.1520/C0856
ASTM C1723-­16. (2016). Standard Guide for Examination of Hardened Concrete
Using Scanning Electron Microscopy. In Annual book of ASTM standards. https://
doi.org/10.1520/C1723-­16.can
Bezerra, A. (2020). The Use of Artificial Intelligence for Assessing Damage in
Concrete Affected by Alkali-­Silica Reaction (ASR) [Master’s Thesis.]. University
of Ottawa.
British Cement Association (BCA). (1992). The diagnosis of alkali-­silica reaction
(p. 44) [Report of a Working Party]. British Cement Association (BCA).
Broekmans, M. (2012). Chapter 7: Deleterious reactions of aggregate with alkalis in
concrete. Reviews in Mineralogy & Geochemistry, 74, 279–364.
CUR-­Recommendation-­102. (2008). Inspection and assessment of concrete struc-
tures in which ASR is suspected or has been confirmed, official English translation
(p. 31). Gouda, Netherlands: Centre for Civil Engineering Research and Codes.
Dehghan, A., Zhang, P., Ossetchkina, E., Sloan, D., & Peterson, K. (2016). Digital
microscopy applied to damage rating index for alkali-­silica reaction in concrete. In
D. Cong & D. Broton (Eds.), Advances in cement analysis and concrete petrography
(pp. 105–125). ASTM International. https://doi.org/10.1520/STP161320180003
 Microscopic analyses 87

Erlin, B., & Stark, D. (1990). Petrography applied to concrete and concrete aggre-
gates. American Society for Testing and Materials, STP 1061.
Fernandes, I., Ribeiro, M., Broekmans, M. A. T. M., & Sims, I. (2016). Petrographic
Atlas: Characterisation of Aggregates Regarding Potential Reactivity to Alkalis:
RILEM TC 219-­ACS Recommended Guidance AAR-­1.2, for Use with the RILEM
AAR-­ 1.1 Petrographic Examination Method (1st ed.). Springer Netherlands.
https://doi.org/10.1007/978-­94-­017-­7383-­6
Fonseca, P. C., & Scherer, G. W. (2015). An image analysis procedure to quantify the
air void system of mortar and concrete. Materials and Structures, 48(10), 3087–
3098. https://doi.org/10.1617/s11527-­014-­0381-­9
Goldstein, J., Newbury, D., Joy, D., Lyman, C., Echlin, P., Lifshin, E., Sawyer, L., &
Michael, J. (2003). Scanning electron microscopy and X-­ray microanalysis. A text
for biologists, materials scientists, and geologists (3rd ed.). Berlin: Springer Verlag.
Grattan-­Bellew, P. E., & Danay, A. (1992). Comparison of Laboratory and Field
Evaluation of Alkali-­Silica Reaction in Large Dams. Proc. International Conference
on Concrete Alkali-­Aggregate Reactions in Hydroelectric Plants and Dams, Sept 28th
to October 2, 1992. Canadian Electrical Association in Association with Canadian
National Committee of the International Commission of Large Dams, pp. 23.
Hernández-­Cruz, D., Hargis, C. W., Dominowski, J., Radler, M. J., & Monteiro, P. J.
M. (2016). Fiber reinforced mortar affected by alkali-­silica reaction: A study by
synchrotron microtomography. Cement and Concrete Composites, 68, 123–130.
https://doi.org/10.1016/j.cemconcomp.2016.02.003
Hollis, N., Walker, D., Lane, S., & Stutzman, P. E. (2006). Petrographic methods of
examining hardened concrete: A petrographic manual, FHWA-­HRT-­04-­150.
Ingham, J. P. (2013). Geomaterials under the microscope. Academic Press.
Jana, D. (2005). Concrete Petrography – Past, Present, and Future. 10th Euroseminar
on Microscopy Applied to Building Materials, Scotland.
Jensen, V., & Sujjavanich, S. (2016). Alkali silica reaction in concrete foundations
in Thailand. 16th International Conference on Alkali-­ Aggregate Reactions in
Concrete (ICAAR), Sao Paulo, Brazil.
Joshi, N. R., Matsumoto, A., Asamoto, S., Miura, T., & Kawabata, Y. (2022).
Investigation of the mechanical behaviour of concrete with severe delayed ettrin-
gite formation expansion focusing on internal damage propagation under various
compressive loading patterns. Cement and Concrete Composites, 128, 104433.
https://doi.org/10.1016/j.cemconcomp.2022.104433
Leemann, A., Katayama, T., Fernandes, I., & Broekmans, M. A. T. M. (2016).
Types of alkali-­aggregate reactions and the products formed. Proceedings of the
Institution of Civil Engineers - Construction Materials, 169, 128–135. https://doi.
org/10.1680/jcoma.15.00059
Leemann, A., & Münch, B. (2019). The addition of caesium to concrete with
alkali-­silica reaction: Implications on product identification and recognition of
the reaction sequence. Cement and Concrete Research, 120, 27–35. https://doi.
org/10.1016/j.cemconres.2019.03.016
Lindgård, J., Haugen, M., Castro, N., & Thomas, M. D. A. (2012). Advantages of
using plane polished section analysis as part of microstructural analyses to describe
internal cracking due to alkali-­silica reactions. 14th International Conference on
Alkali-­Aggregate Reactions in Concrete (ICAAR), Austin, Texas.
Lindgård, J., Skjølsvold, O., Haugen, M., Hagelia, P., & Wigum, B. J. (2004).
Experience from evaluation of degree of damage in fluorescent impregnated
88 Internal Swelling Reactions in Concrete

plan polished sections of half cores based on the cracking index method. 12th
International Conference on Alkali-­Aggregate Reactions in Concrete (ICAAR),
Beijing, China.
Loser, R., & Leemann, A. (2016). An accelerated sulfate resistance test for con-
crete. Materials and Structures, 49(8), 3445–3457. https://doi.org/10.1617/
s11527-­015-­0731-­2
Marinoni, N., Voltolini, M., Mancini, L., Vignola, P., Pagani, A., & Pavese, A. (2009).
An investigation of mortars affected by alkali-­silica reaction by X-­ray synchro-
tron microtomography: A preliminary study. Journal of Materials Science, 44(21),
5815–5823. https://doi.org/10.1007/s10853-­009-­3817-­9
Melo, R. H. R. Q., Hasparyk, N. P., & Tiecher, F. (2023). Assessment of concrete
impairments over time triggered by DEF. Journal of Materials in Civil Engineering,
35(8). https://doi.org/10.1061/JMCEE7.MTENG-­15041
Poole, A. B., & Sims, I. (2016). Concrete petrography, a handbook of investigative
techniques. (2nd ed.). London: CRC Press (Taylor & Francis Group).
Rivard, P. (1998). Quantification de l’endommagement Du Béton Atteint de Réaction
Alcalis-­Silice Par Analyse d’images [Master’s Thesis.]. Université de Sherbrooke.
Rivard, P., Fournier, B., & Ballivy, G. (2000). Quantitative petrographic technique
for concrete damage due to ASR: Experimental and application. Cement, Concrete
and Aggregates, 22(1), 63–72. https://doi.org/10.1520/CCA10465J
Sanchez, L. F. M., Drimalas, T., & Fournier, B. (2020). Assessing condition of concrete
affected by internal swelling reactions (ISR) through the Damage Rating Index
(DRI). Cement, 1–2, 100001. https://doi.org/10.1016/j.cement.2020.100001
Sanchez, L. F. M., Drimalas, T., Fournier, B., Mitchell, D., & Bastien, J. (2018).
Comprehensive damage assessment in concrete affected by different internal
swelling reaction (ISR) mechanisms. Cement and Concrete Research, 107, 284–
303. https://doi.org/10.1016/j.cemconres.2018.02.017
Sanchez, L. F. M., Fournier, B., Jolin, M., & Duchesne, J. (2015). Reliable quan-
tification of AAR damage through assessment of the Damage Rating Index
(DRI). Cement and Concrete Research, 67, 74–92. https://doi.org/10.1016/j.
cemconres.2014.08.002
Sanchez, L. F. M., Fournier, B., Jolin, M., Mitchell, D., & Bastien, J. (2017). Overall
assessment of Alkali-­Aggregate Reaction (AAR) in concretes presenting differ-
ent strengths and incorporating a wide range of reactive aggregate types and
natures. Cement and Concrete Research, 93, 17–31. https://doi.org/10.1016/j.
cemconres.2016.12.001
Shakoorioskooie, M., Griffa, M., Leemann, A., Zboray, R., & Lura, P. (2021). Alkali-­
silica reaction products and cracks: X-­ray micro-­tomography-­based analysis of
their spatial-­temporal evolution at a mesoscale. Cement and Concrete Research,
150, 106593. https://doi.org/10.1016/j.cemconres.2021.106593
Sims, I., Hunt, B., & Miglio, B. (1992). Quantifying microscopical examinations of
concrete for Alkali Aggregate Reactions (AAR) and other durability aspects.
Song, Y., Zou, R., Castaneda, D. I., Riding, K. A., & Lange, D. A. (2017). Advances
in measuring air-­void parameters in hardened concrete using a flatbed scanner.
Journal of Testing and Evaluation, 45(5), 20150424. https://doi.org/10.1520/
JTE20150424
Suzuki, T., Shiotani, T., & Ohtsu, M. (2017). Evaluation of cracking dam-
age in freeze-­ thawed concrete using acoustic emission and X-­ ray CT image.
Construction & Building Materials, 136, 619–626. https://doi.org/10.1016/j.
conbuildmat.2016.09.013
 Microscopic analyses 89

Thomas, M., Folliard, K., Drimalas, T., & Ramlochan, T. (2008). Diagnosing delayed
ettringite formation in concrete structures. Cement and Concrete Research, 38(6),
841–847. https://doi.org/10.1016/j.cemconres.2008.01.003
Villeneuve, V. (2011). Détermination de l’endommagement du béton par méthode
pétrographique quantitative [M.Sc.]. Université Laval.
Voltolini, M., Marinoni, N., & Mancini, L. (2011). Synchrotron X-­ray computed
microtomography investigation of a mortar affected by alkali-­silica reaction: A
quantitative characterization of its microstructural features. Journal of Materials
Science, 46(20), 6633–6641. https://doi.org/10.1007/s10853-­011-­5614-­5
Walker, H. N., Lane, D. S., & Stutzman, Paul E. (2006). Petrographic methods of
examining hardened concrete: A petrographic manual (FHWA-­HRT-­04-­150;
p. 351). Federal Highway Administration (FHWA).
Chapter 6

Mechanical tools

6.1 INTRODUCTION

Well-­designed and “sound” concrete generally presents good strength and

modulus of elasticity (ME), reasonable tensile strength, a brittle response
under uniaxial loading (i.e., compression or tension) and an increase in
both ductility and strength (in compression and tension) in a confined envi-
ronment (Crouch & Wood 1990). This sensitivity to confinement is related
to the presence of small flaws or microcracks in the material. Thus, even
for “sound” concrete under a triaxial compression load, there will always
be locations in tension within the bulk material due to its heterogeneous
and “defective” nature (Crouch & Wood 1990). Moreover, the complex-
ity of concrete behaviour under stress is even greater when the material
is deteriorated (Crouch & Wood 1990). Hence, the understanding of the
mechanical properties changes (i.e., compressive and tensile strengths, ME,
aggregate interlock through the direct shear, and also the stress/strain rela-
tionship) as a function of ISR-­induced development is a very important step
for assessing condition (i.e., damage extent), along with designing rehabili-
tation strategies for affected structures and structural members (Kubo &
Nakata 2012).
Different ISR mechanisms may influence differently the mechanical
properties of affected concrete (Sanchez et al., 2018). This dissimilar
impact is related to the unique microscopic distress features and general
“pattern” associated with each of these deterioration mechanisms (see
Chapter 5; Sanchez et al. 2020). To understand the impact of the various
ISRs on the mechanical response of affected concrete, it is crucial to first
discuss the response of conventional and “sound” concrete with respect to
standardized and advanced mechanical test procedures conducted in com-
pression, tension and shear. Depending on the “sound” concrete response
and failure mode, a given mechanical test procedure may or may not be
considered suitable or “diagnostic” to appraise the damage extent of ISR-­
affected concrete.

90 DOI: 10.1201/9781003188155-6
 Mechanical tools 91

6.2 TEST PROCEDURES IN COMPRESSION

6.2.1 Compressive strength test

Compressive strength is the most common mechanical test procedure per-
formed in concrete due to four main reasons: (1) it is assumed that most
of the mechanical properties of concrete are directly related to compres-
sive strength (although not always true); (2) concrete is primarily used in
compression due to its low tensile strength; (3) structural design codes are
mainly based on the compressive strength; (4) the compressive strength test
is a fast, easy and relatively inexpensive procedure (Mindess et al. 2003).
This method is commonly implemented as per national/international stan-
dards such as ASTM C39 or similar (ASTM C39 2003). The compressive
strength test outcomes may vary as a function of a number of parameters
such as temperature and humidity, sample size and geometry, loading rate,
etc. Therefore, it is important that the aforementioned parameters be prop-
erly controlled so that the test outcomes gathered from different labora-
tories can be compared to one another. In this section, some of the most
important parameters of the test will be discussed; further information can
be found in (Mindess et al. 2003; ASTM C39 2003).
Compressive strength tests as per ASTM C 39 are conducted using cylin-
drical specimens or cores with a length-­to-­diameter ratio (l/d) of 2:1 (i.e.,
usually 100 mm by 200 mm or 150 mm by 300 mm specimens). If the speci-
mens are manufactured in the laboratory, they should be moulded in layers,
depending on the method of consolidation; if the specimens are rodded,
three equal layers are required. However, if they are vibrated, only two lay-
ers are to be used. Once manufactured, the specimens should be moist cured
in a standard moist room or in saturated lime water at 23 ± 2°C until testing
(normally over 28 days or as per the project’s rationale). Conversely, if the
test is conducted on cores extracted from concrete structures, variations in
the moisture condition of the specimens from coring to testing may occur,
which can significantly influence the test results (ASTM C42 2008). As per
ASTM C 42, cores should be tested “in the same moisture condition than
that they were in the field”. However, since moisture gradients are often
observed within concrete elements (in contact or not with external moisture
sources), it is recommended that upon coring, the specimens should be
wrapped in an impermeable plastic film and stored for at least five days
before testing (ASTM C42 2008). Moreover, according to Canadian
Standards, cores extracted from structures should be rewetted for 48h in a
standard moist curing room prior to testing in order to reset their moisture
conditions (CSA.A23.2-­14C 2009).
After removal from the moist room and before testing, the specimen’s
ends should be treated to become smooth, plane and perpendicular to the
longitudinal axis. Normally, planeness is achieved by mechanical grinding
or by capping through the use of stiff Portland cement pastes, high-­strength
92 Internal Swelling Reactions in Concrete

gypsum plasters or even sulphur mortars. Once the ends are smooth and
flat, the specimen is ready for testing. Loading is then applied to the speci-
men at a rate ranging from 0.15 to 0.35 Mpa/s for hydraulic machines or at
a deformation rate of 1 mm/min for mechanical machines until failure. The
maximum load and type of failure are finally reported. It is worth noting
that different geometries (such as cubes) and test conditions are used in
other standards and countries, particularly in some European countries
(Mindess et al. 2003). Correction factors should be then implemented to
compare the test outcomes bearing distinct parameters. Table 6.1 illustrates
the correction factors normally adopted to correlate compressive strength
responses of cylinders and cubes.
Regardless of the test conditions, the response of concrete under non-­
confined, uniaxial compression is considered a progressive phenomenon
related to the generation and propagation of microcracks in the system
(Mindess et al. 2003). At the macroscale, cracks are generated parallel to the
loading direction (i.e., vertically), forming a conic shape at the top and bot-
tom ends of the specimen due to the existing friction between the ends and
the loading plate. In a condition of non-­friction, completely parallel and
vertically aligned cracks would be expected before failure. Figure 6.1 illus-
trates the macroscale cracking process. Disregarding the presence of fric-
tion, an important number of vertical cracks is nevertheless generated; these
vertical cracks create various “slender little columns” in the specimen, which
leads to local buckling. Ultimately, failure in unconfined uniaxial compres-
sion is caused by tension failure of multiple and simultaneous planes.
On the other hand, at the micro-­mesoscale, the phenomenon of failure
under a non-­confined uniaxial compression is slightly more progressive.
First, cracks are initially generated at the interface between the aggregate

Table 6.1 C
 orrection factors used to correlate compressive strength
results from cylinders and cubes

Cube strength (Mpa) Cylinder strength (Mpa) Correction factor

9 7 1.29
16 12 1.33
20 15.5 1.29
24.5 20 1.23
27.5 23 1.20
35.5 32 1.11
42 36 1.17
44 37.5 1.17
48 44 1.09
52 50.5 1.03

Source: Adapted from (Mindess et al. 2003)

 Mechanical tools 93

Figure 6.1 M
 acroscale cracking formation in an unconfined uniaxial compressive
strength test: (a) confinement at both ends, (b) confinement at the
bottom and splitting failure at the top and (c) splitting failure.
(Adapted from Mindess et al. 2003; Swamy, 1979.)

particles and the cement paste, the so-­ called interfacial transition zone
(ITZ), when the load reaches about 30%–40% of the ultimate capacity of
the material being tested; then, a process of “slow/stable crack propagation”
takes place where the tensile strength of the material is locally reached due
to stress intensity factors and thus, the load is transferred elsewhere (Mindess
et al. 2003). The propagation of cracks generated at the ITZ slowly contin-
ues with loading, and at 50%–70% of the ultimate capacity, multiple cracks
reach the cement paste. Some of those cracks are “arrested” by aggregate
particles present in the granular system and need to outline them to keep
propagating, while others remain free in the cement paste. At about 75% of
the ultimate capacity, these cracks increase significantly in length and width,
and some of them start linking to one another. If the load is sustained, fail-
ure may take place from this stage and onwards. At 95%–100% of the
ultimate capacity, most of the cracks are linked, and the phenomenon
becomes “unstable”, with fast crack propagation leading to failure (Mindess
et al. 2003). Figure 6.2 illustrates the failure process previously described.
Understanding the failure process of “sound” concrete in unconfined uni-
axial compression is imperative to evaluate whether this test procedure may
be considered a diagnostic tool to appraise ISR-­deteriorated concrete. The
literature suggests that the compressive strength of ISR-­affected concrete
94 Internal Swelling Reactions in Concrete

Figure 6.2 C oncrete failure in unconfined uniaxial compression.

(Adapted from Mindess et al. 2003.)

varies as a function of the damage mechanism. Figure 6.3 illustrates the

impact of different ISRs on the compressive strength of affected concrete.
For alkali-­ aggregate reaction (AAR)–affected concrete, the compressive
strength decrease is only significant (i.e., 15% or more) at high and very
high expansion levels (i.e., > 0.20% of expansion or strain; Nixon &
Bollinghaus 1985; Pleau et al. 1989; Smaoui, Bérubé, et al., 2004b; Sanchez
et al., 2018, 2017). This happens particularly because at low and moderate
expansion levels (i.e., 0.05%–0.12% of expansion or strain), the vast major-
ity of AAR-­ induced cracks are generated within the aggregate particles
(Sanchez et al., 2015a; Sanchez, Fournier, Jolin, Bedoya, et al., 2016); in
other words, the AAR-­ induced cracks do not directly interact with the
cracks formed over the loading process in compression, as previously
described in this section. Otherwise, the response of concrete affected by
other ISR mechanisms, for example, delayed ettringite formation (DEF) and
freeze and thaw (FT), is slightly different, especially at the early stages of the
reaction. DEF-­and FT-­ affected concrete display a higher compressive
strength reduction (i.e., 10% to 20%) for low and moderate expansion lev-
els (i.e., 0.05%–0.12% of expansion or strain) than AAR-­affected concrete;
this happens since DEF and FT generate microcracks in the cement paste,
particularly at the ITZ and bulk cement paste/pores, respectively. Therefore,
these cracks interact more with the new cracks developed over the loading
process in compression or even increase (i.e., in length and width) while
loading, which explains the more pronounced impact. This impact could be
 Mechanical tools 95

Figure 6.3 Impact of distinct ISR on the compressive strength of concrete. The
mixtures presented in the labels are 35 MPa concrete mixtures incor-
porating either New Mexico or Texas sand as coarse or fine reactive
aggregates, respectively. These mixtures developed DEF, FT and ASR
deterioration (Sanchez et al., 2018).

even higher in the compressive strength reduction of DEF-­and FT-­affected

concrete; however, two mechanisms are verified to inhibit further losses,
such as (1) the cracks being extremely localized, especially in DEF cases, and
(2) the “arrest mechanism” provided by the aggregates, which stops the
cracks’ propagation once they get in contact with the aggregate particles, as
described by Mindess et al. (2003)). Conversely, at high and very high
expansion levels (i.e., 0.20%–0.30%), the compressive strength reductions
are quite similar whatever the ISR mechanism (i.e., 20%–25%) since the
cracks generated by the three mechanisms are observed in the cement paste
at later stages, which makes a similar interaction with the compression load.
Interestingly, combined mechanisms display a more important impact on
the compressive strength of affected concrete, particularly at the early stages
of DEF + ASR-­and FT + ASR-­affected concrete, as a few examples; this
influence is more pronounced when ASR is triggered from a reactive coarse
aggregate. This behaviour indicates a higher interaction of the cracks in the
cement paste and aggregates over the compression loading, besides suggest-
ing the decrease of the so-­called arrest mechanism previously described since
the cracks go through and split the coarse aggregate particles.
The results previously discussed demonstrate that the compressive
strength test is not always effective in diagnosing the damage degree (i.e.,
expansion level) of concrete affected by ISR. On the one hand, it is very
96 Internal Swelling Reactions in Concrete

important to acknowledge the actual and residual strength of ISR-­affected

concrete to understand the potential structural implications of the current
deterioration in affected structures or structural components. On the other
hand, having a test procedure suitable to estimate the current induced
expansion of distinct ISR mechanisms is crucial to recognize the potential
for further deterioration of affected concrete members. All of the aforemen-
tioned should be carefully evaluated prior to the use of compressive strength
in practice.

6.2.2 Modulus of elasticity (ME)

The stiffness of concrete is conventionally represented by the ME. However,
it is important to notice that concrete is a nonlinear inelastic material under
both tension and compression loads. Therefore, the concept of ME needs to
be carefully applied in this context since it does not represent a single value
as for linear elastic materials (Mindess et al. 2003).
It is widely known and accepted that the stiffness response of various
materials, such as aggregates, cement paste, mortar and concrete, is dissimi-
lar (Kosmatka et al. 2003). Figure 6.4 illustrates the stress-­strain relation-
ship of the previously mentioned materials. Normally, aggregates display
higher stiffness than cementitious composites (i.e., cement paste, mortar and
concrete), although the stiffness response of aggregates depends on their
nature (i.e., lithotype). Moreover, an almost elastic linear behaviour is
observed in the stress-­strain curve of aggregates’ specimens. Conversely, one

Figure 6.4 S tress vs strain relationship for various materials.

(Adapted from Kosmatka et al. 2003.)
 Mechanical tools 97

might expect that the cement paste (which does not bear aggregates) is the
material displaying the lowest stiffness, along with nonlinear inelastic
behaviour. Concrete and mortar mixtures are within these two extremes
(Kosmatka et al. 2003) due to the inclusion of aggregates, which display a
higher stiffness.
There are several possibilities for adopting the ME of concrete. Figure 6.5
illustrates some of the moduli that can be gathered from the stress-­strain
curve of concrete in compression. The initial tangent modulus is likely the
closest approximation of the ME derived from a linear elastic material. This
parameter is not conventionally used in the design of concrete structures
since it is considered to overestimate the stiffness of the material, besides
being captured with small values of stresses and strains that do not match
the ones applied in practice. Otherwise, the secant ME (i.e., the slope of the
secant between the origin and a selected point in the stress-­strain curve)
bears an element of nonlinearity, being a more conservative parameter than
the initial tangent modulus and better represents the behaviour of the mate-
rial in the field (Mindess et al. 2003). The secant ME (normally the slope
between the origin and 40% of the ultimate stress) is often adopted to design
concrete structures. The initial tangent or the secant moduli are not always
easily determined due to errors in measurement caused by the specimen’s
seating at the beginning of the test or by the presence of cracks in the speci-
men’s evaluated that close under load (Mindess et al. 2003). As such, the
chord modulus, as per ASTM C 469, is used, which is the slope of a line
drawn between two points on the stress-­strain curve. The initial tangent
modulus typically corresponds to the slope of the curve at a strain of

Figure 6.5 S tress-­s train curve of concrete in compression displaying the various
elastic moduli.
(Adapted from Mindess et al. 2003.)
98 Internal Swelling Reactions in Concrete

0.00005 (lower limit) and 40% of the compressive strength value (upper
limit) of the material tested (ASTM C 469 2006). The chord modulus is a
more conservative measure than the initial tangent modulus and is easily
gathered experimentally (Mindess et al. 2003).
The specimens (i.e., laboratory cylinders or cores) should be manufac-
tured or cored, moisture conditioned and their ends treated similarly to the
compressive strength test to perform the ME test as per ASTM C 469. Then,
a load of 0.25 MPa/s ± 0.05 MPa/s at a constant rate should be applied until
it reaches 40% of the compressive strength of the concrete under analysis.
Upon reaching this value, the load is then reduced to zero at the same rate.
Three loading-­unloading cycles are normally performed, where the first
cycle is discarded. The average of the second and third cycles is often selected
to represent the stiffness or ME of the concrete (ASTM C 469 2006).
Understanding the stiffness response of concrete, normally adopted as the
secant or chord ME, is crucial to appraise its efficiency as a diagnostic pro-
tocol to assess ISR-­affected concrete. As for the compressive strength, the
literature suggests that the ME of ISR-­affected concrete varies as a function
of the damage mechanism. Figure 6.6 illustrates the impact of different ISRs
on the ME of affected concrete. For AAR-­affected concrete, the ME is
already significantly affected at low and moderate expansion levels (i.e.,
0.05%–0.12% of expansion or strain), where reductions of until 30% are

Figure 6.6 Impact of distinct ISR on the ME of affected concrete. The mixtures
presented in the labels are 35 Mpa concrete mixtures incorporating
either NM or TX as coarse or fine reactive aggregates, respectively.
These mixtures developed DEF, FT and ASR deterioration (Sanchez
et al., 2018).
 Mechanical tools 99

frequently observed (Sanchez et al., 2017, 2018). This happens due to the
generation of an important number of cracks within the reactive aggregate
particles (i.e., coarse and or fine; Sanchez, Fournier, Jolin, Bedoya, et al.,
2016; Sanchez et al. 2015a), thus weakening the stiffness of the aggregate.
As the induced expansion keeps progressing, the cracks previously formed
in the aggregates expand to the cement paste. Therefore, the ME of the
affected concrete keeps decreasing but at a lower degree. For very high
expansion levels (i.e., 0.30% of expansion or strain), the ME reduction of
AAR-­ affected concrete can reach 50% (Sanchez et al., 2018, 2017).
Otherwise, the response of concrete affected by other ISR mechanisms, for
example, DEF and FT, is different, especially at the early stages of the dam-
age process. DEF-­and FT-­affected concrete display a higher decrease in ME
at low and moderate expansion levels (i.e., 0.05%–0.12%) than AAR-­
affected concrete; this happens because DEF and FT generate an important
number of microcracks at the interface between aggregates and cement paste
(i.e., ITZ). Therefore, these cracks may “break” the bond between the aggre-
gates and cement paste and thus significantly drop the stiffness of the mate-
rial as the stress transfer “bridge” created by that bond is eliminated. The
decrease in ME of DEF-­and FT-­affected concrete may vary from either 35%
to 55% for DEF cases or 25% to 40% for FT cases at low and moderate
expansion levels (Sanchez et al., 2018). Conversely, at high and very high
expansion levels (i.e., 0.20%–0.30% of expansion or strain), the ME reduc-
tions are quite similar, whatever the ISR mechanism (i.e., 50%–60%), since
the cracks generated by the three mechanisms are very spread in the affected
concrete at later stages (i.e., cement paste, ITZ and aggregates), which makes
a roughly similar interaction with a controlled compression load. Interestingly,
combined mechanisms display a dissimilar impact on the ME of affected
concrete. DEF + ASR coupling seems to further reduce the ME of the affected
concrete, whereas FT + ASR coupling seems to lessen the impact on the stiff-
ness of the material; the more pronounced influence of DEF + ASR coupling
is likely due to the higher chemical affinity and microscopic distress features
interaction (i.e., sharper and wider cracks at the ITZ and aggregate particles)
present in this combined mechanism. Nevertheless, the results discussed in
this section demonstrate that the ME is a suitable mechanical procedure to
appraise the current stage of ISR-­affected concrete.

6.2.3 Stiffness damage test (SDT)

Concrete under cyclic loading behaves quite differently than under sus-
tained loading. During repeated cycles in compression, concrete displays
a change in the stress-­strain curve from a concave downward to a concave
upward shape (Mindess et al. 2003). This change is also characterized by
an increase in secant modulus (i.e., stiffness) over the initial cycles followed
by a reduction of stiffness or secant modulus at later stages. Furthermore,
the accumulation of strain is rapid during the initial cycles, slows for the
100 Internal Swelling Reactions in Concrete

intermediate cycles and increases again for the final cycles prior to failure.
Figure 6.7 illustrates the stress-­strain curves of concrete under cyclic load in
compression.
The shape of the stress-­strain curves at early cycles suggests that pre-­
existing defects and or microcracks are closed over the first compression
cycles and may be reopened and propagated at later cyclic stages. The mech-
anism of damage development and failure under cyclic loading is explained
through fracture mechanics concepts (i.e., stress concentration), where
important stresses are generated at the tip of the flaws, defects and micro-
cracks generated during the hydration process and often observed at the
ITZ of concrete (Mindess et al. 2003). At the microscale, when concrete is
tested under repeated cycles of loading, these pre-­existing flaws and cracks
tend to grow. This is demonstrated in Figure 6.7 by the formation of hyster-
esis loops. The hysteresis area loops represent the irreversible energy of
deformation, some of which represent the sliding across surfaces of cracks,
while others demonstrate the energy to extend the pre-­ existing cracks
(Mindess et al. 2003). The hysteresis area loops decrease at first with succes-
sive load cycles but then begin to increase again prior to failure. In the
beginning, the crack growth is slow, yet due to stress concentrations, pre-­
existing microcracks eventually increase in size (i.e., length and width) with
the rise in the number of cycles and interactions with the aggregates in the
cementitious system, which leads to the stabilization of larger cracks growth;
however, the energy supplied by the repeated loadings eventually increases
the deterioration at the cracks tips, leading to crack propagation until fail-
ure (Mindess et al. 2003).
If one evaluates the mechanism of failure under cyclic loading, along with
the changes in the distinct variables that comprise the test outcomes, such as

Figure 6.7 S tress-­s train curves of concrete under cyclic loading.

(Adapted from Mindess et al. 2003.)
 Mechanical tools 101

the stress-­strain curve shapes, the hysteresis area loops and plastic deforma-
tion (or strain), one verifies an opportunity of using cyclic tests for assessing
pre-­existing deterioration in concrete. In this context, the SDT has been pro-
posed, and, differently from the compressive strength and ME, it is a fairly
new procedure; its use has recently been adopted to assess the condition of
concrete affected by ISR; hence, a detailed description of the test develop-
ment is deemed appropriate herein.
The SDT is a mechanical and cyclic test procedure used to assess the
condition (i.e., damage extent) of concrete affected by ISR mechanisms
(Sanchez, Fournier, et al., 2014). The SDT was initially developed by Walsh,
who observed a good correlation between the crack density and the cycles
of loading/unloading (i.e., stress/strain relationship) of rock specimens
(Walsh 1965); Crouch then adapted this procedure for concrete specimens
in 1987 (Crouch 1987). Following those developments, Chrisp et al. pro-
posed the use of the SDT for assessing concrete affected by ASR through
the application of a fixed stress of 5.5 MPa at a loading rate of 0.10 MPa/s
(Chrisp et al. 1993; Chrisp et al. 1989); the authors initially wanted the
procedure to remain “non-­destructive”, thus enabling the reuse of the test
specimens for further analyses. Chrisp and coworkers conducted over 1,000
tests on cores extracted from ASR-­affected concrete structures; after apprais-
ing the stress-­strain curves of the deteriorated specimens, the authors pro-
posed the following test outcomes to assess the damage extent of affected
concrete (Chrisp et al. 1993; Chrisp et al. 1989):

• ME (E): average secant ME of the last four cycles since affected

concrete often demonstrates lower ME than undamaged or sound
specimens.
• Hysteresis area (H, in J/m3): area of the hysteresis loops (i.e., area
under the stress-­strain curves) averaged over the last four cycles since
affected concrete displays higher energy loss (or hysteresis area loops)
than undamaged or sound concrete.
• Nonlinearity index (NLI): it represents the ratio between the slope of
the stress-­strain curve at half of the maximum load and the secant ME
(E). This parameter provides information on the damage extent and
pattern (i.e., cracks orientation).

Chrisp et al. observed that the hysteresis area of the first cycle was much
greater than that of the following four consecutive cycles; this behaviour
has been attributed to the sliding effect across open crack surfaces and their
closure upon loading. Therefore, the results from the first cycle were sug-
gested to be disregarded to only obtain a response from the material itself.
Moreover, the ME was found to be the most sensitive test outcome for
slightly damaged concrete. However, for higher degrees of deterioration, the
hysteresis area has been verified as the critical test outcome for assessing the
extent of damage. Finally, it has been observed that the crack pattern could
102 Internal Swelling Reactions in Concrete

influence the results of the test since deteriorated concrete presenting cracks
mostly oriented perpendicular to the load direction were found to display a
low ME, a larger hysteresis area and an NLI greater than unity, while those
with cracks mainly oriented parallel to the load direction were seen to yield
a high ME, a smaller hysteresis area and an NLI lower than unity (Chrisp
et al. 1993; Chrisp et al. 1989).
Later, Smaoui et al. (Smaoui, Fournier et al., 2004a) further evaluated the
efficiency of the SDT on laboratory-­ made concrete specimens mix-­
proportioned as per the concrete prism test (CPT) according to ASTM C
1293, incorporating a variety of reactive aggregate types (i.e., fine versus
coarse) and natures (i.e., lithotypes), and presenting various expansion lev-
els following storage at 38°C at 100% R.H. (ASTM C1293 2015). After
performing numerous tests, it has been verified that the most efficient SDT
outcome was the hysteresis area of the first cycle; moreover, they observed
that the SDT should be conducted with a fixed load of 10 MPa instead of
5.5 MPa to increase the diagnostic character of the test since lower stress
levels were found to be unsuitable to allow ASR-­induced microcracks to suf-
ficiently reclose, which decreased the test efficiency and reliability. Smaoui et
al. also verified that the correlation between the expansion and the plastic
deformation obtained after the five loading/unloading cycles was fairly sat-
isfactory. However, they noted a high variability for both the hysteresis area
and plastic deformation of ASR-­affected concrete incorporating different
reactive aggregates. These variabilities were found to be possibly associated
with the type (fine or coarse) and nature (lithotypes) of reactive aggregates
used in concrete, which could lead to distinct deterioration patterns (i.e.,
cracks density, orientation, locations, etc.; Smaoui, Fournier et al., 2004a).
Sanchez et al. (Sanchez, Fournier, et al., 2014, Sanchez et al. 2015b,
Sanchez, Fournier, Jolin, Bastien, et al. 2016) performed a comprehensive
experimental campaign, evaluating numerous concrete mixtures presenting
distinct mechanical properties and reactive aggregates, on the use of SDT for
appraising concrete affected by AAR but also deteriorated by other ISR
mechanisms such as DEF and FT, single or combined. Further details on the
procedure developments and specific considerations may be found in
Sanchez et al. (Sanchez, Fournier, et al. 2014, Sanchez et al. 2015b, Sanchez,
Fournier, Jolin, Bastien, et al. 2016). Briefly, it has been found that the SDT
should be conducted at 40% of the material’s mechanical capacity at the
loading rate proposed by Chrisp et al. of 0.10 MPa/s. Furthermore, the
method was considered statistically efficient and reliable for assessing the
damage extent of ISR-­affected concrete, especially through the use of the
Stiffness Damage Index (SDI) and Plastic Deformation Index (PDI) out-
comes; the SDI and PDI represent, respectively, the ratio of dissipated energy/
plastic deformation to the total energy/deformation implemented in the sys-
tem over five cycles of loading-­unloading (i.e., SI/(SI + SII) and DI/(DI +
DII)). Finally, as per Chrisp et al. (Chrisp et al. 1993; Chrisp et al. 1989), the
Nonlinearity Index (NLI), defined as the secant modulus at half the
 Mechanical tools 103

maximum load (Secant 2) divided by the secant modulus at the maximum

load (Secant 1) has been confirmed as an interesting outcome to assess the
damage extent and pattern in concrete affected by ISR. Figure 6.8 illustrates
the SDT set-­up and the calculation of the SDI, PDI and NLI parameters.
Besides optimizing the testing load and outcomes, Sanchez et al. (2015b;
Sanchez, Fournier, Jolin, Bastien, et al. 2016) also studied some practical
parameters that might interfere with the SDT results, such as the specimen’s
size (i.e., 100 vs 150 mm diameter, distinct length-­to-­diameter ratios), ends
rectification type (i.e., grinding versus capping), moisture condition, drying
and rewetting effects, and location of the core in the affected structure (i.e.,
surface vs core of the members under analysis). It has been verified that the
sample’s diameter and the end rectification type do not significantly influ-
ence the SDT outcomes. However, the sample length-­to-­diameter ratio, the
moisture condition and the specimen’s location were indeed found to
impact the test results. Therefore, for practical purposes and following the
Canadian Standards (CSA.A23.2-­14C 2009), it has been proposed that
cores extracted from structures should be cut and rectified to have a length-­
to-­diameter ratio close to 2:1; moreover, the specimens should be rewetted
for 48 h in a standard moist curing room prior to stiffness damage testing
to decrease the test variability. Figure 6.9 illustrates a flowchart developed
by Sanchez et al. to describe the practical tasks required before conducting
the SDT in the laboratory (Sanchez et al. 2015b; Sanchez, Fournier, Jolin,
Bastien, et al., 2016).
Most of the works developed so far using the SDT to assess the damage
extent of affected concrete were conducted evaluating AAR-­affected con-
crete. However, the test procedure has been shown to also be effective and
reliable in appraising other ISR mechanisms such as DEF and FT. Moreover,

Figure 6.8 S DT set-­u p and stress-­s train curve of ISR-­a ffected concrete: calcula-
tion of the indices SDI, PDI and NLI.
(Adapted from Kongshaug et al., 2002.)
104 Internal Swelling Reactions in Concrete

Figure 6.9 S DT flowchart displaying practical tasks required to conduct the test
in the laboratory.
(Adapted from Sanchez et al. 2015b; Sanchez, Fournier, Jolin, Bastien,
et al. 2016.)

as previously discussed in this section, the SDT outcomes (i.e., particularly

the SDI and PDI) measure the number of inner cracks (or physical integrity)
of affected concrete. Therefore, these parameters should not largely vary
from one ISR mechanism to another as a function of induced development.
Figure 6.10 demonstrates the SDI and PDI values of concrete mixtures
affected by ASR, DEF and FT (Sanchez et al. 2018). Analysing the data, it is
clear that both SDI and PDI are very similar for ASR and FT mechanisms,
varying from 0.10 to 0.45 and from 0.08 to 0.45, respectively, for low (i.e.,
0.05%) and very high (i.e., 0.30%) expansion levels. Conversely, the results
obtained for DEF, although following the same trend, are slightly lower than
the ones for ASR and FT, ranging from 0.10 to 0.25 and from 0.08 to 0.20,
respectively. These findings might be related to (1) the thinner and more
spread crack pattern found in ASR and FT deterioration when compared to
the sharper and more localized pattern observed in DEF-­affected concrete
and (2) the constantly observed cracks outlining the aggregate particles (i.e.,
ITZ) in DEF-­affected concrete, which could limit their full closure under
compressive loads (Sanchez et al. 2018). It should be noted that SDI and PDI
values under 0.10 generally represent sound conventional concrete.
 Mechanical tools 105

Figure 6.10 Impact of distinct ISR on the (a) SDI and (b) PDI of affected con-
crete. The mixtures presented in the labels are 35 MPa concrete
mixtures incorporating either NM or TX as coarse or fine reactive
aggregates, respectively. These mixtures developed DEF, FT and ASR
deterioration (Sanchez et al., 2018).

The previous results indicate that the SDT is an efficient tool for assessing
the induced expansion and deterioration in concrete caused by distinct ISR
mechanisms. Further studies should be conducted to understand the param-
eters influencing the SDI and PDI fluctuations for distinct ISR mechanisms,
yet, in general, a consistent deterioration trend is observed for all mecha-
nisms evaluated so far.
106 Internal Swelling Reactions in Concrete

6.3 TEST PROCEDURES IN TENSION

6.3.1 Tensile strength test

Tensile strength is a parameter that is not commonly used for designing rein-
forced concrete structures (other than pavements), and it is usually adopted
(for sound and conventional concrete) as being about 10% of the compres-
sive strength of the material at a given age (Mindess et al. 2003; Swamy
1979). Otherwise, the literature suggests that this parameter is much more
influenced by ISR deterioration and cracking in general than the compres-
sive strength, and thus its evaluation could likely add interesting informa-
tion on ISR-­induced development (Sanchez, Multon et al., 2014; Sanchez
et al., 2017, 2018).
Differently from the compressive strength, the response of “sound” con-
crete under tension is very brittle and is governed by fracture mechanics
principles (Broek 1982; Mindess et al. 2003; Swamy 1979). According to
fracture mechanics, the mechanical response of materials containing flaws,
defects and microcracks, such as concrete, is very diminished when com-
pared to its theoretical potential. Results very often tend to be between 10
to 100 times lower than theoretically expected (Broek 1982; Mindess et al.
2003; Swamy 1979). Failure takes place as per the formation of macro-
cracks generated from pre-­existing defects and flaws in the material; these
macrocracks are formed due to the presence of stress intensity factors at the
tips of pre-­existing microdefects, increasing their size (i.e., length and width)
until they reach the so-­called critical length; defects displaying sizes above
and beyond the critical length lead to brittle failure of concrete.
At the macroscale, the response of “sound” concrete under tension may
be divided into two stages: before and after the formation of the lead mac-
rocrack bearing the critical length. Before such a macrocrack is formed,
the response of the material depends on its microstructure quality (i.e.,
cement paste, ITZ and aggregates; Swamy 1979). Conversely, once the
lead macrocrack is formed, the mechanical response is solely dependent on
the friction (i.e., slide across surfaces) of the lead macrocrack. On the
other hand, at the micro-­mesoscale, the phenomenon of failure in tension
of “sound” concrete is normally divided into five steps. First, microcracks
are developed (normally in the cement paste) from pre-­existing flaws or
defects in the material at low loading levels (step 1), followed by a process
of slow crack propagation (step 2). Then, some of the formed cracks are
“arrested” by the aggregate particles present in the granular system and
need to outline them to keep propagating, while others remain free in the
cement paste (step 3). Finally, the critical crack length is reached (step 4),
and thus the process changes from a stable (i.e., slow crack propagation) to
an unstable (i.e., fast crack propagation) process, leading to a sudden and
brittle failure (step 5) (Swamy 1979). Figure 6.11 illustrates the previously
described failure process.
 Mechanical tools 107

Figure 6.11 C oncrete failure in tension.

(Adapted from Swamy 1979.)

Several test set-­ups have been developed to assess the tensile strength of
concrete; the most common approaches are the direct tension test, the bend-
ing tensile test and the splitting tensile test (Mindess et al. 2003). The direct
tension test is theoretically a quite suitable technique to evaluate the “pure”
or “true” tensile strength of concrete. However, premature failure at the
ends of the specimens caused by secondary stresses induced by the grips is
quite often verified (Mier & Van Vliet 2002). Therefore, the test has not
been adopted by ASTM as a standardized test procedure. Figure 6.12 illus-
trates the direct tension test set-­up normally used. Nevertheless, RILEM
displays a recommendation for a direct tensile strength test in concrete that
was primarily developed for research (Zheng et al. 2001; Mindess et al.
2003). This method involves applying direct tension to cylindrical or pris-
matic specimens through end plates glued to the specimens. The ends of the
specimens should be rectified and carefully cleaned so that they become flat
and smooth, enabling proper adhesion of the glue (normally a polyepoxy
resin); they must be perpendicular to the axis of the specimen within 1/4°
(Zheng et al. 2001; Mindess et al. 2003). The load is applied at a rate of
0.05MPa/s until failure takes place. The US Bureau of Reclamation also
specifies a direct tension test that uses bonded end plates (Zheng et al. 2001;
Mindess et al. 2003). Otherwise, the bending tensile and splitting tensile
tests are both standardized techniques that measure “indirectly” the tensile
strength of concrete; therefore, they are perceived as “less diagnostic” tools
to assess ISR-­induced damage development since failure happens in desig-
nated areas (i.e., failure planes) in both procedures.
108 Internal Swelling Reactions in Concrete

Figure 6.12 D irect tension test set-­u p.

The splitting tensile test as per ASTM C 496 is performed on standard

cylinders, tested on their side in diametral compression, as illustrated by
Figure 6.13 (ASTM C496 2008). The load is normally applied through a
narrow bearing strip of relatively soft material. The tensile stress distribu-
tion along the vertical diameter of the specimen is not uniform; there are

Figure 6.13 Splitting tensile strength test set-­

u p and corresponding stress
distribution.
(Adapted from Mindess et al. 2003.)
 Mechanical tools 109

high compressive stresses near the ends of the vertical diameter while a
nearly uniform tensile stress area acting over the middle-­two thirds of the
specimen takes place (ASTM C496 2008; Mindess et al. 2003). Since con-
crete is much weaker in tension than in compression, it will split in tension
at a much lower load than is required to fail the specimen in compression.
There is no simple relationship between the values of tensile strength
found from the splitting test and those measured in direct tension. However,
it is commonly assumed that the splitting tensile strength values are
5%–12% higher than the direct strength values, although some works have
shown close results from these two test procedures (Zheng et al. 2001;
Mindess et al. 2003). Detailed analyses from the splitting tensile strength
test have shown that the test as per ASTM C 496 is not a “true” property of
the material since the results are dependent on the size, width and type of
the load-­bearing strips (Zheng et al. 2001; Mindess et al. 2003).
The flexural strength of concrete may be determined as per ASTM C78.
Prismatic specimens (i.e., 150 mm by 150 mm by 500 mm) are manufac-
tured in two consecutive and equal layers; each layer is compacted by either
rodding (i.e., 60 times per layer) or vibration (as per compression tests). The
specimens are cured in a standard fashion and then tested in flexure in four
or third-­point loading, as illustrated in Figure 6.14 (ASTM C78 2010). The
flexural strength of concrete, as per ASTM C78, can also be conducted with
specimens sawn from concrete components.
The specimens manufactured in the laboratory should be tested and
turned on their sides with respect to their position as moulded. This should
provide smooth, plane and parallel faces for loading (ASTM C78 2010;
Mindess et al. 2003). Otherwise, the specimens should be ground or capped.
The specimens are loaded at a rate of 860–1,200 KPa/min, and the theoreti-
cal maximum tensile strength, also called modulus of rupture (R), is calcu-
lated according to Equation 6.1.

R = Pl / ( bd ^ 2 ) , (6.1)

Figure 6.14 F lexural test as per ASTM C 78.

(Adapted from Mindess et al. 2003.)
110 Internal Swelling Reactions in Concrete

where P is the peak load, l is the span length, b is the specimen width and d is
the specimen depth. Equation 6.1 holds if the specimen breaks between the
two interior loading points (i.e., the middle third of the beam). If the beam
breaks outside of these points, but not more than 5% of the span length,
Equation 6.2 should be used (ASTM C78 2010; Mindess et al. 2003).

R = 3Pa / ( bd ^ 2 ) , (6.2)

where a is the average distance between the point of fracture and the nearest
support.
The flexural test tends to overestimate the “pure” or “true” tensile strength
of concrete by about 50%, especially due to the fact that the flexure formula
assumes that the stress varies linearly across the cross-­section of the speci-
men, which is not true due to the nonlinear stress-­strain behaviour of con-
crete. Nevertheless, the flexure test remains quite useful, especially to
measure the response of concrete mixtures used in components under bend-
ing in the field, such as pavements (Mindess et al. 2003).
The gas pressure tension test, also known as the indirect tension test, is a
non-­ standardized technique that has been developed by the Building
Research Council of Waterford (United Kingdom) as a means of investigat-
ing anisotropic behaviour of materials along with overcoming the chal-
lenges of premature failure of the direct tensile strength. The test procedure
uses compressed gas to apply a uniformly distributed pressure to the curved
surface of standard (i.e., 100 mm by 200 mm) concrete cylinders or cores
(Komar et al. 2013; Sanchez, Multon et al., 2014). The apparatus consists
of a hollow cylindrical test chamber that envelops the curved surface of the
test cylinder. At either end of the testing chamber, rubber “O-­rings” are used
to seal the compressed gas so that it only acts upon the curved surface of the
specimen. Figure 6.15 illustrates the cross-­section of a gas pressure cham-
ber. Both ends are left open to atmospheric pressure, resulting in a biaxial
loading configuration. Gas pressure is monotonically increased until the test
cylinder fails in a plane transverse to the axis of the testing chamber (Komar
et al. 2013; Sanchez, Multon et al., 2014).
The gas pressure applied to the curved surface is a biaxial loading condi-
tion, but the reaction forces within the diphase model differ. In particular,
the pore water reacts hydrostatically, whereas the solid phase reacts biaxi-
ally, resulting in a net internal tensile force driven by the pore fluid. The
resultant internal tension force is the primary reason why the pressure ten-
sion method is thought to measure the “pure” or “true” tensile strength of
concrete, along with being well suited for detecting durability issues that
affect the concrete microstructure and physical integrity (Komar et al.,
2013; Sanchez, Multon et al., 2014).
Understanding the failure process in tension, along with the influence of
the various tension test set-­ups on the mechanical responses of “sound,” but
especially ISR-­affected concrete, is crucial to evaluate and compare their
 Mechanical tools 111

Figure 6.15 C ross-­s ection of the pressure chamber (Sanchez, Multon et al., 2014).

diagnostic character. For instance, Figure 6.16 illustrates the comparison

amongst the direct tensile, splitting tensile and gas pressure tension test
results measured on AAR-­affected concrete as a function of induced expan-
sion. Analysing the results, it is clear to see that there is a decrease in tensile
strength as a function of AAR-­induced development from the beginning of
the expansion development for all the set-­ups presented. Moreover, the test
responses vary according to the set-­up used; the splitting test yields lower
mechanical reductions (i.e., higher total values) as a function of induced
expansion than both the direct tensile and the gas pressure tension tests,
which are very close to one another. This behaviour is observed because the
splitting tensile strength is an indirect test where the failure plane is previ-
ously “established” or “selected” prior to testing (i.e., vertical plane).
Therefore, whether the most prominent cracks due to ISR (i.e., AAR in this
case) are not located in the selected failure plane, the method is less effective
in assessing the deterioration present in the specimen. Conversely, the direct
tensile strength and the gas pressure tension test display a quite sharp drop
in tensile strength from the beginning of the expansion development, pre-
senting very similar results to one another as a function of the induced
expansion. This indicates these two set-­ups are able to measure the “pure”
or “true” tensile strength of the affected concrete, creating a fracture process
that is less established from the beginning of the test and thus being more
effective to appraise the induced development of ISR mechanisms. Since the
literature shows a number of practical issues related to the direct tensile
strength test, such as failure at the ends of the specimens due to secondary
stresses induced by the grips, the gas pressure tension emerges as a promis-
ing technique.
The gas pressure tension has been used to assess ISR-­affected specimens
and demonstrated to be suitable to capture the induced expansion and
112 Internal Swelling Reactions in Concrete

Figure 6.16 C
 omparison of tensile strength test results as a function of AAR-­
induced expansion for direct tensile and splitting tensile for (a)
non-­r eactive aggregates, (b) highly reactive Texas sand and (c) highly
reactive Québec City limestone.
(Adapted from Smaoui et al. 2006.)
 Mechanical tools 113

damage development in its early stages (i.e., low and moderate expansion
levels), regardless of the ISR mechanism (Sanchez et al., 2017, 2018). Figure
6.17 illustrates the gas pressure tension results obtained for ASR and FT-­
affected specimens displaying from low (i.e., 0.05%) to very high (i.e.,
0.30%) expansion levels. Observing the data, one notices that the drop in

Figure 6.17 G
 as pressure tension results for (a) ASR and (b) combined with FT-­
affected concrete. The mixtures presented in the labels are 35 MPa
concrete mixtures incorporating either NM or TX as coarse or fine
reactive aggregates, respectively. These mixtures developed DEF, FT
and ASR deterioration.
(Adapted from Sanchez et al., 2018.)
114 Internal Swelling Reactions in Concrete

tensile strength of FT-­affected specimens (i.e., from 45% to 70%) at low

expansion levels is higher than ASR-­affected concrete (i.e., from 30% to
60%). Yet, for high and very high expansion levels, the difference between
mechanisms seems to decrease. These results are caused by the more impor-
tant presence of cracks in the cement paste (i.e., ITZ, bulk cement paste and
pores) taking place in FT deterioration when compared to ASR, where the
cracks are mainly within the aggregate particles in the early stages of the
mechanism. However, very likely due to the test conditions (i.e., pure ten-
sion governed by fracture mechanics), the gas pressure set-­up seems to not
be able to distinguish different damage scenarios beyond “moderate” dam-
age degrees since an important levelling off trend is observed from 0.12%
expansion and onwards (Sanchez et al., 2017, 2018). Further analysis should
still be performed to better understand the limitation of the gas pressure
tension to appraise ISR-­affected concrete at high and very high expansion
levels since the use of this test procedure in this perspective is fairly new.

6.4 TEST PROCEDURES IN SHEAR

6.4.1 Direct shear test

Shear strength in concrete is a property governed by tension and compression
forces. Once cracked (and concrete will always present a certain number of
inner cracks, flaws, etc.), concrete may transfer shear forces across the cracks
through two distinct mechanisms: (a) dowel effect and (b) shear friction
(Haskett et al. 2011). The dowel effect is related to the reinforcement used in
concrete members, whereas the shear friction is associated with the concrete
features themselves. Shear friction is defined as the “frictional resistance of
cracks to sliding” (Haskett et al. 2011). Under initially cracked conditions,
the sliding plane surfaces of concrete may be assumed as “rough and irregu-
lar” due to the presence of aggregates (normally rough and irregular), along
with the rough texture of the cracks’ surfaces themselves. These rough and
irregular aggregate particles “force the sliding planes apart, which induces
normal stresses in the reinforcement crossing the sliding planes, restricting
their opening” (Haskett et al. 2011). Confinement to the sliding planes pro-
vides frictional resistance to sliding and allows the shear forces to transfer
across the cracked planes (Saouma et al. 2016; Haskett et al. 2011). Under
high confinement levels, significant shear stresses are transferred across the
crack surfaces through shear friction. The shear friction phenomenon, often
called “aggregate interlock”, is an important component while designing
reinforced concrete members in North America, particularly in Canada
(CSA A23.3-­14 2017).
A number of test methods have been developed over the years to evaluate
the direct shear capacity and shear friction of reinforced and unreinforced
concrete (Banks-­Sills & Arcan 1983; Adams & Walrath 1987; Richard
1981). Figure 6.18 illustrates the proposed set-­ups.
 Mechanical tools 115

Figure 6.18 S hear set-­u ps to test concrete specimens: (a) double-­n otched cylin-
drical shear specimen (Barr & Hasso 1986), (b) reinforced concrete
shear set-­u p (Bazant & Pfeiffer 1985) and (c) small-­s cale direct shear
set-­u p (Barr & Hasso 1986).

Barr et al. (Barr & Hasso 1986) performed a comparative study on differ-
ent set-­ups to evaluate the direct shear of plain concrete. One of the most
promising set-­ups used was the modified cylinder specimen (i.e., double-­
notched cylindrical specimen), illustrated in Figure 6.18a. The proposed
geometry enabled the use and comparison of distinct types of materials and
specimens such as concrete (samples or cores made of plain concrete, fibre
reinforced concrete, etc.), rock, and similar materials (Barr & Hasso 1986).
However, some concerns were raised with the use of the double-­notched
specimen, such as the difficulty of making similar and aligned notches along
with the application of a direct longitudinal load on the specimen.
Gao et al. (1979) developed a new shear set-­up to study the brittle fracture
of reinforced concrete composites, which is illustrated in Figure 6.18b; later
on, this set-­up was adopted for other studies with promising results. Recent
studies adapted the same geometry proposed by Gao et al., yet an upper
notch has been added to the superior part of the specimen (Gao et al. 1979).
116 Internal Swelling Reactions in Concrete

Bazant and Pfeiffer (1985) have used the geometry proposed in Figure
6.18b to study the shear strength of concrete and mortar beams. Barr and
Thomas (Barr & Derradj 1990) have used the same geometry to determine
the fracture characteristics of glass-­reinforced cement. However, a great deal
of care was necessary during the manufacture of the test specimens to ensure
a proper alignment and the same depth of the two notches. Nonetheless, the
experimental programme presented interesting outcomes with low variabil-
ity and high reproducibility (Barr & Derradj 1990; Barr and Hasso, 1986).
Barr and Hasso (1986) adapted the previous set-­ups and proposed a test
apparatus to assess small-­scale concrete samples (i.e., 100 mm by 200 mm),
which is illustrated in Figure 6.18c. The most important change the authors
made was the use of a circumferential notch instead of two aligned ones.
Studies demonstrated that circumferential notches of about 20 to 25 mm
might be able to ensure the shear-­type failure of the specimens without leav-
ing a too-­small area of the sample to be tested, which might cause an increase
in the discrepancy of the results (Barr & Hasso 1986).
Sanchez et al. (2015a; Sanchez, Fournier, Jolin, Bedoya, et al., 2016) veri-
fied through the use of advanced microscopic analyses that AAR-­induced
cracks development follows a two-­step process: (1) the cracks are generated
within the reactive aggregate particles (fine or coarse) in the bulk concrete
volume and (2) the cracks increase in length and width and eventually run
out to the cement paste. The deterioration mechanism described earlier
raised concerns about the shear response of concrete members affected by
ISR mechanisms since shear friction (or aggregate interlock) may be drasti-
cally decreased due to AAR-­induced cracks formation within the aggregate
particles. Otherwise, if the aggregate interlock of concrete specimens is
directly influenced by AAR, a shear test set-­up might be able to capture AAR
development and thus become a potentially promising tool to evaluate
AAR-­induced damage.
De Souza et al. (2019) evaluated a wide range of concrete mixtures dete-
riorated by AAR (i.e., ASR and ACR) presenting distinct mechanical proper-
ties (i.e., 25, 35 and 45 MPa) and incorporating numerous reactive aggregate
types (i.e., fine and coarse) and natures (i.e., lithotypes) using the direct
shear test set-­up as per Barr and Hasso (Barr & Hasso 1986). Figure 6.19a
illustrates the test set-­up adopted, while Figure 6.19b gives a plot of the
results obtained in this work. Analysing the results obtained, it is verified
that there is a consistent decrease in the shear strength as a function of AAR-­
induced development. This decrease is not linear but very sharp at low to
moderate expansion levels (i.e., 0.05% to 0.12%), where reductions from
10% to 25% are often observed, followed by a period of slight decrease at
high and very high expansion levels (i.e., 0.20% to 0.30%), where reduc-
tions of 20% to 35% are verified.
Besides the promising results demonstrating the potential of the direct
shear test of being a diagnostic procedure to capture AAR-­induced expan-
sion and damage development, analyses on the failure mode of tested
 Mechanical tools 117

Figure 6.19 D
 irect shear test: (a) set-­u p and (b) shear strength decrease as a
function of AAR-­induced expansion (Souza et al., 2019).

Figure 6.20 D
 irect failure plane of AAR-­a ffected concrete generated by (a) reac-
tive fine aggregates and (b) reactive coarse aggregates. The width the
image represents is 6 cm. The dashed arrows point to failure through
the aggregate particles, and the solid arrows point to failure at the
cement paste, causing the debonding of the aggregate particles.
(Photos courtesy of Rouzbeh Ziapour.)

specimens showed that the procedure is able to “recognize” the distinct

microscopic damage features developed when AAR is triggered by reactive
coarse or fine aggregates, as illustrated in Figure 6.20. For instance, the fail-
ure plane takes place through the cement paste when AAR is generated by
reactive fine aggregates (Figure 6.20a) while through the coarse aggregates
when AAR is triggered by reactive coarse aggregate particles (Figure 6.20b).
The latter means that, first, the direct shear test set-­up is able to capture
AAR-­induced damage, detecting distinct AAR cracking patterns and thus
generating failure at distinct planes. With the promising results obtained for
AAR-­affected concrete, the direct shear test demonstrates a great potential
to evaluate other ISR mechanisms. Further analysis should be conducted in
this regard.
118 Internal Swelling Reactions in Concrete

REFERENCES

Adams, D. F., & Walrath, D. E. (1987). Current status of the Iosipescu shear test
method. Journal of Composite Materials, 21(6): 494–507. https://doi.org/10.
1177/002199838702100601
ASTM C 469. (2006). Standard Test Method for Static Modulus of Elasticity and
Poisson’s Ratio of Concrete in Compression. https://doi.org/10.1520/C0469
ASTM C1293. (2015). Standard Test Method for Determination of Length Change
of Concrete Due to Alkali-­Silica Reaction. ASTM Standard Book. www.astm.org
ASTM C39. (2003). Standard test method for compressive strength of cylindri-
cal concrete specimens. ASTM Standard Book, 4(March), 1–5. https://doi.org/
10.1520/C0039
ASTM C42. (2008). Standard Test Method for Obtaining and Testing Drilled Cores
and Sawed Beams of Concrete. ASTM Standard Book. https://doi.org/10.1520/
mnl10913m
ASTM C496. (2008). Standard Test Method for Splitting Tensile Strength of
Cylindrical Concrete Specimens. ASTM Standard Book. https://doi.org/10.1520/
mnl10881m
ASTM C78. (2010). Standard Test Method for Flexural Strength of Concrete (Using
Simple Beam with Third-­ Point Loading). ASTM Standard Book. https://doi.
org/10.1520/C0078
Banks-­Sills, L., & Arcan, M. (1983). An Edge-­Cracked Mode II Fracture Specimen.
Experimental Mechanics, 23(3), 257–261. https://doi.org/10.1007/bf02319251
Barr, B., & Derradj, M. (1990). Numerical study of a shear (Mode II) type test speci-
men geometry. Engineering Fracture Mechanics, 35(1–3), 171–180. https://doi.
org/10.1016/0013-­7944(90)90194-­L
Barr, B., & Hasso, E. B. D. (1986). Development of a Compact Cylindrical Shear
Test Specimen. Journal of Materials Science Letters, 5(12), 1305–1308. https://doi.
org/10.1007/BF01729401
Bazant, Z. P., & Pfeiffer, P. A. (1985). Tests on Shear Fracture and Strain-­Softening
in Concrete. In 2nd Symposium on Interaction of Non-­Nuclear Munitions with
Structures. Held in Pana beach, Florida, pp. 254–264.
Broek, D. (1982). Elementary engineering fracture mechanics (3rd ed.). Martinus
Nijhoff Publishers. ISBN: 978-­94-­010-­8425-­3
Chrisp, T. M., Waldron, P., & Wood, J. G. M. (1993) Development of a non-­destructive
test to quantify damage in deteriorated concrete. Magazine of Concrete Research,
45(165), 247–256. https://doi.org/10.1680/macr.1993.45.165.247
Chrisp, T. M., Wood, J. G. M., & Norris, P. (1989). Towards quantification of micro-
structural damage in AAR deteriorated concrete. In Fracture of concrete and rock:
Recent developments (p. 755).
Crouch, R. S. (1987). Specification for the determination of stiffness damage param-
eters from the low cyclic uniaxial compression of plain concrete cores, revision A.
Mott. Hay & Anderson. Special Services Division. Internal Technical Note.
Crouch, R. S., & Wood, J. G. M. (1990). Damage evolution in AAR affected
concretes. Engineering Fracture Mechanics, 35(1–3), 211–218. https://doi.
org/10.1016/0013-­7944(90)90199-­Q
CSA A23.3-­14. (2017). Design of Concrete Structures.
CSA.A23.2-­ 14C. (2009). Obtaining and Testing Drilled Cores for Compressive
Strength Testing.
 Mechanical tools 119

Gao, H., Wang, Z., Yang, C., & Zhou, A. (1979). An investigation on the brittle
fracture of KrKn composite mode cracks. Acta Metallurgica Sinica, 15, 380–391.
Haskett, M., Oehlers, D. J., Mohamed Ali, M. S., & Sharma, S. K. (2011). Evaluating
the shear-­ friction resistance across sliding planes in concrete. Engineering
Structures, 33(4), 1357–1364. https://doi.org/10.1016/j.engstruct.2011.01.013
Kongshaug, S. S., Oseland, O., Kanstad, T., Hendriks, M.A.N., Rodum, E., &
Markeset, G. (2020). Experimental investigation of ASR-affected concrete – The
influence of uniaxial loading on the evolution of mechanical properties, expansion
and damage indices. Construction and Building Materials, 245, 118384. https://
doi.org/10.1016/j.conbuildmat.2020.118384
Komar, A., Hartell, J., & Boyd, A. J. (2013). Pressure tension test: Reliability for
assessing concrete deterioration. In Proceedings of the Seventh International
Conference on Concrete under Severe Conditions, China, pp. 340–347.
Kosmatka, S. H., Kerkhoff, B., Panarese, W. C., MacLeod, N. F., & McGrath, R. J.
2003. Design and control of concrete mixtures. Canadian Cement Association
(CCA).
Kubo, Y., & Nakata, M. (2012). Effect of types of reactive aggregate on mechani-
cal properties of concrete affected by Alkali-­Silica reaction. In 14th International
Conference on Alkali-­Aggregate Reaction, 021711, p. 10.
Mier, J. G. M. Van, & Van Vliet, M. R. A. (2002). Uniaxial tension test for the deter-
mination of fracture parameters of concrete: State of the Art. Engineering Fracture
Mechanics, 69, 235–247.
Mindess, S., Young, J. F., & Darwin, D. (2003). Concrete (2nd ed.). Prentice Hall –
Person Education Inc.
Nixon, P. J., & Bollinghaus, R. 1985. The effect of Alkali aggregate reaction on the
tensile strength of concrete. Durability of Building Materials, 2, 243–248.
Pleau, R., Bérubé, M. A., Pigeon, M., Fournier, B., & Raphael, S. (1989). Mechanical
behavior of concrete Aaffected by AAR. In 8th ICAAR – International Conference
on Alkali-­Aggregate Reaction in Concrete (pp. 721–726).
Richard, H. A. (1981). A new shear compact test specimen. International Journal of
Fracture, 17, 5–7.
Sanchez, L.F.M., Drimalas, T., & Fournier, B. (2020). Assessing condition of concrete
affected by Internal Swelling Reactions (ISR) through the Damage Rating Index
(DRI). Cement, 1–2(September), 100001. https://doi.org/10.1016/j.cement.2020.
100001
Sanchez, L. F. M., Drimalas, T., Fournier, B., Mitchell, D., & Bastien, J. (2018).
Comprehensive damage assessment in concrete affected by different internal swell-
ing reaction (ISR) mechanisms. Cement and Concrete Research, 107(February),
284–303. https://doi.org/10.1016/j.cemconres.2018.02.017
Sanchez, L.F.M., Fournier, B., Jolin, M., & Bastien, J. (2014). Evaluation of the
Stiffness Damage Test (SDT) as a Tool for Assessing Damage in Concrete Due
to ASR: Test Loading and Output Responses for Concretes Incorporating Fine
or Coarse Reactive Aggregates. Cement and Concrete Research, 56. https://doi.
org/10.1016/j.cemconres.2013.11.003
Sanchez, L.F.M., Fournier, B., Jolin, M., & Bastien, J. (2015b). Evaluation of the
Stiffness Damage Test (SDT) as a tool for assessing damage in concrete due
to Alkali-­ Silica Reaction (ASR): Input parameters and variability of the test
responses. Construction and Building Materials, 77. https://doi.org/10.1016/j.
conbuildmat.2014.11.071
120 Internal Swelling Reactions in Concrete

Sanchez, L.F.M., Fournier, B., Jolin, M., Bastien, J., & Mitchell, D. 2016. Practical
use of the Stiffness Damage Test (SDT) for assessing damage in concrete infra-
structure affected by alkali-­silica reaction. Construction and Building Materials,
125. https://doi.org/10.1016/j.conbuildmat.2016.08.101
Sanchez, L.F.M., Fournier, B., Jolin, M., Bedoya, M.B.A., Bastien, J., & Duchesne,
J. (2016). Use of damage rating index to quantify alkali-­silica reaction damage in
concrete: Fine versus coarse aggregate. ACI Materials Journal, 113(4). https://doi.
org/10.14359/51688983
Sanchez, L. F.M., Fournier, B., Jolin, M., & Duchesne, J. (2015a). Reliable quan-
tification of AAR damage through assessment of the Damage Rating Index
(DRI). Cement and Concrete Research, 67, 74–92. https://doi.org/10.1016/j.
cemconres.2014.08.002
Sanchez, L.F.M., Fournier, B., Jolin, M., Mitchell, D., & Bastien, J. (2017). Overall
Assessment of Alkali-­Aggregate Reaction (AAR) in concretes presenting differ-
ent strengths and incorporating a wide range of reactive aggregate types and
natures. Cement and Concrete Research, 93. https://doi.org/10.1016/j.cemconres.
2016.12.001
Sanchez, L.F.M., Multon, S., Sellier, A., Cyr, M., Fournier, B., & Jolin, M. (2014).
Comparative study of a chemo-­mechanical modeling for Alkali Silica Reaction
(ASR) with experimental evidences. Construction and Building Materials, 72.
https://doi.org/10.1016/j.conbuildmat.2014.09.007
Saouma, V. E.., Hariri-­Ardebili, M.A., Le Pape, Y., & Balaji, R. 2016. Effect of alkali-­
silica reaction on the shear strength of reinforced concrete structural members. A
numerical and statistical study. Nuclear Engineering and Design, 310, 295–310.
https://doi.org/10.1016/j.nucengdes.2016.10.012
Smaoui, N., Bérubé, M. A., Fournier, B., & Bissonnette, B. (2004b). Influence of
specimen geometry, orientation of casting plane, and mode of concrete consolida-
tion on expansion due to ASR. Cement, Concrete and Aggregates, 26(2), 58–70.
https://doi.org/10.1520/cca11927
Smaoui, N., Bissonnette, B., Bérubé, M.A., Fournier, B., & Durand, B. (2006).
Mechanical properties of ASR-­affected concrete containing fine or coarse reactive
aggregates. Journal of ASTM International, 3(3), 1–16. https://doi.org/10.1520/
jai12010
Smaoui, N., Fournier, B., Bérubé, M. A., Bissonnette, B., & Durand, B. 2004a.
Evaluation of the expansion attained to date by concrete affected by alkali-­silica
reaction. Part II: Application to nonreinforced concrete specimens exposed outside.
Canadian Journal of Civil Engineering, 31(6): 997–1011. https://doi.org/10.1139/
L04-­074
Souza, D. J. De, Sanchez, L. F.M., & De Grazia, M. T. (2019). Evaluation of a direct
shear test setup to quantify AAR-­induced expansion and damage in concrete.
Construction and Building Materials, 229, 116806. https://doi.org/10.1016/
j.conbuildmat.2019.116806
Swamy, R.N. (1979). Fracture mechanisms applied to concrete. In F.D. Lydon (Eds.),
Fracture measurements of cementitious composites (pp. 221–281). London:
Applied Science Publishers.
Walsh, J. B. (1965). The effect of cracks on the uniaxial elastic compression of
rocks. Journal of Geophysical Research, 70(2), 399–411. https://doi.org/10.1029/
jz070i002p00399
Zheng, W., Kwan, A. K.H., & Lee, P.K.K. (2001). Direct tension test of concrete. ACI
Materials Journal, 98 (1): 63–71. https://doi.org/10.14359/10162
Chapter 7

Multi-­level assessment of
ISR-­affected concrete

7.1 INTRODUCTION

In the preceding chapters, an extensive discussion unfolded on the distinct

internal swelling reaction (ISR) mechanisms inducing deterioration in con-
crete, the importance of assessing the condition of ISR-­affected concrete and
the different test procedures (i.e., visual, non-­destructive, chemical, micro-
scopic and mechanical) enabling the appraisal of structures and structural
members affected by ISR. The analysis of these chapters revealed the pres-
ence of advantages and disadvantages associated with each recommended
technique, necessitating a case-­by-­case evaluation. Depending on factors
such as the type of structure, accessibility, extent of damage and environ-
mental conditions, certain techniques demonstrate greater efficiency than
others. However, it became evident from the preceding chapters that the
most effective approach lies in combining these techniques to create compre-
hensive evaluation protocols. The selection of techniques should go beyond
mere “data gathering and correlation” and instead focus on their synergistic
nature, providing supplementary information necessary for a thorough diag-
nosis (i.e., cause(s) and extent) of the damage affecting concrete structures
or structural members under analysis. Chapter 5 extensively discussed the
fact that different deterioration mechanisms associated with ISR can induce
distinct microscopic damage features, such as cracks, in affected concrete;
Figure 7.1 illustrates the various damage features associated with different
ISR mechanisms.
Analysis of Figure 7.1 reveals that ASR-­induced deterioration predomi-
nantly originates within the reactive aggregate particles (i.e., either fine or
coarse) in the concrete microstructure. In contrast, DEF-­induced deteriora-
tion often manifests as cracks outlining the aggregate particles at the inter-
facial transition zone (ITZ) and subsequently extending into the cement
paste, ultimately connecting with one another. In the case of freeze and thaw
(FT) deterioration, cracks tend to develop primarily in the cement paste and
macro pores (assuming the use of FT-­resistant aggregates in the mixture)
rather than exclusively at the ITZ. Moreover, it is worth noting that the
proximity of FT cracks to the surface of affected concrete members

DOI: 10.1201/9781003188155-7 121

122 Internal Swelling Reactions in Concrete

Figure 7.1 C
 rack pattern in concrete affected by (a) alkali-­s ilica reaction (ASR)
reactive coarse aggregate, (b) ASR reactive sand, (c) freezing and
thawing cycles and (d) delayed ettringite formation (DEF). The con-
crete's surface is shown on the left-­h and side.
(Adapted from British Cement Association (BCA) 1992.)

correlates with a higher likelihood of parallel alignment. These observations

emphasize the fact that the overall impact of deterioration on the engineer-
ing properties of concrete varies depending on the underlying mechanism(s)
causing damage and, significantly, the degree of its development. Therefore,
an efficient and comprehensive assessment protocol should encompass the
ability to determine these variations.

7.2 MULTI-­LEVEL ASSESSMENT PROTOCOL

Literature reports a wide range of test procedures (e.g., visual, non-­

destructive, microscopic and mechanical) employed to quantify the impact
of ISR-­induced deterioration in affected concrete over time. However, most
of the available data is presented through a “decoupled” pathway. This means
that either microscopic and descriptive data are provided to confirm the
presence of ISR through the identification of secondary products or visual,
non-­destructive and mechanical data shed light on the influence of current
ISR-­induced deterioration on the durability and engineering properties of the
affected material. The limitation of conducting “decoupled” analyses is the
inability to determine the cause and extent (i.e., complete diagnosis) of the
 Multi-level assessment of ISR-affected concrete 123

deterioration. Furthermore, without establishing a quantitative correlation

between the “inner quality” of the affected concrete microstructure and
its macro performance, the potential progression of the induced damage
and its associated impact remains elusive. To overcome this challenge, a
multi-­level assessment protocol, recognized as a multi-­scale approach, has
been developed to facilitate a more comprehensive understanding, correla-
tion and quantitative description of how the “inner quality” of the con-
crete microstructure (i.e., porosity, number, type and location of cracks;
presence of secondary products; discoloration; and other relevant factors)
influences its macro performance (Martin et al. 2017; Sanchez et al. 2017,
2018). Essentially, the macro performance of concrete, comprehending engi-
neering properties, such as stiffness, compressive, tensile and direct shear
strength, is explained over time via a thorough and quantitative examina-
tion of its microstructure. The multi-­level protocol recommends the inte-
gration of quantitative microscopic techniques, such as the Damage Rating
Index (DRI) and image analysis, combined with mechanical tests, such as
the stiffness damage test (SDT) and direct shear test, to establish a cohesive
relationship between the microstructure and macro performance. Over the
years, this protocol has exhibited a high efficiency and reliability in assessing
the condition of concrete damaged by ISR, thereby enabling a deeper com-
prehension of its present state and the progressive advancement of induced
deterioration in ISR-­affected critical concrete infrastructure.

7.3 PRACTICAL USE OF THE MULTI-­LEVEL

ASSESSMENT PROTOCOL

In the pursuit of a comprehensive understanding of ISR-­affected concrete,

Sanchez et al. and Martin et al. (Martin et al. 2017; Sanchez et al. 2017,
2018) conducted an extensive experimental campaign using numerous con-
crete specimens/cores with distinct raw materials and compositions. As a
result, a pragmatic multi-­level approach was proposed, which integrates
quantitative microscopic and mechanical data into a unified four-­quadrant
chart (Figure 7.2).
Within this chart, the expansion experienced by ISR-­affected concrete at
a specific age (represented on the positive x-­axis) can be easily correlated
with mechanical data (expressed as Stiffness Damage Index (SDI) on the
positive y-­axis) and microscopic data (DRI number divided by 1,000 on the
negative y-­axis). Moreover, additional parameters of interest, such as modu-
lus of elasticity (ME), compressive strength, direct shear strength and more,
can be incorporated on the negative x-­axis, thereby enabling a comprehen-
sive evaluation of the overall degree of deterioration.
The compilation and analysis of data through the four-­quadrant chart for
ISR-­affected concrete, incorporating a wide variety of raw materials and
compositions, enabled the development of benchmark tables that outline
124 Internal Swelling Reactions in Concrete

Figure 7.2 Four-­

q uadrant chart for the assessment of ISR-­
a ffected concrete
(Sanchez et al. 2017, 2018).

the impact of different ISR mechanisms on the induced expansion and dete-
rioration of affected concrete. Table 7.1 presents examples of such bench-
mark tables for ASR, DEF and FT mechanisms. Notably, this experimental
campaign also investigated induced deterioration caused by alkali-­carbonate
reaction (ACR) (i.e., Kings + Lav 35), which allowed a deeper understanding
of its distinct impact on concrete integrity and engineering properties.
However, it is important to note that ACR was not included in the bench-
mark tables due to the utilization of only one aggregate source for this par-
ticular mechanism.
The utilization of benchmark tables holds considerable advantages for
engineers in practice, as it enhances decision-­making by providing a more
accurate understanding of the entire deterioration process with a reduced
reliance on extensive data collection. However, it is important to note that
these benchmark tables have primarily been established based on laboratory
test samples subjected to free-­induced expansion conditions, while it is evi-
dent that structures and structural members experience a diverse range of
stress-­state conditions in the field. Nevertheless, this approach should be
regarded as a “reference” and can be considered the “worst-­case scenario”
for ISR-­affected concrete.

7.3.1 Using the multi-­level assessment protocol to

understand the impact of ISR under unrestrained
conditions
The evaluation of ISR-­affected concrete commonly relies on quantifying
the degree of expansion, which serves as a measure of its deterioration.
Table 7.1 Benchmark tables of ISR-­induced development on affected concrete

Damage results

Distress Classification of ASR Reference expansion Compressive Tensile strength

mechanism damage degree (%) level (%)a Stiffness loss (%) strength loss (%) loss (%) SDI DRI
ASR Negligible 0.00–0.03 – – – 0.06–0.16 100–155
Marginal 0.04 ± 0.01 5–37 (−)10–15 15–60 0.11–0.25 210–400
Moderate 0.11 ± 0.01 20–50 0–20 40–65 0.15–0.31 310–500
High 0.20 ± 0.01 35–60 13–25 45–80 0.19–0.32 500–765
Very high 0.30 to 0.50 ± 0.01 40–67 20–35 0.22–0.36 600–925
Ultra high 0.50 to 1.00 ± 0.01 – – – – –

Multi-level assessment of ISR-affected concrete

≥ 1.00 ± 0.01 – – – – –
FT and FT Negligible 0.00–0.03 – – – 0.11 147–154
+ ASR Marginal 0.04 ± 0.01 0.23–0.35 0.12–0.32 0.44–0.67 0.16–0.23 496–684
Moderate 0.11 ± 0.01 0.28–0.36 0.21–0.32 0.62–0.67 0.25–0.28 590–950
High 0.20 ± 0.01 0.33–0.46 0.22–0.37 0.65–0.67 0.27–0.41 677–963
Very high 0.30 to 0.50 ± 0.01 0.37–0.52 0.24–0.40 0.65–0.73 0.34–0.45 800–1300
Ultra high 0.50 to 1.00 ± 0.01 – – – – –
≥ 1.00 ± 0.01 – – – – –
DEF and Negligible 0.00–0.03 – – – 0.11 110–147
DEF + Marginal 0.04 ± 0.01 – – – – –
ASR Moderate 0.11 ± 0.01 0.35–0.56 0.09–0.34 – 0.17–0.20 355–599
High 0.20 ± 0.01 – – – – –
Very high 0.30 to 0.50 ± 0.01 0.55–0.62 0.29–0.43 – 0.19–0.28 623–710
Ultra high 0.50 to 1.00 ± 0.01 0.56–0.77 0.40–0.47 – 0.27–0.43 828–1022
≥ 1.00 ± 0.01 0.60–0.86 0.40–0.50 0.30–0.54 841–1363

125
a These levels of expansion should not be considered as strict limits between the various of damage degree but rather indicators/reference levels of the three different
mechanisms studied in this work.
Source: Adapted from Sanchez et al. (2017, 2018)
126 Internal Swelling Reactions in Concrete

Extensive investigations have revealed a consistent and progressive cor-

relation between expansion and the overall deterioration of the concrete
(Mohammadi et al. 2020; Sanchez et al. 2017). Specifically, as the expan-
sion increases, there is a corresponding escalation in the loss of mechanical
properties, physical integrity and stiffness of the affected material. This cor-
relation emphasizes the importance of accurately assessing and quantifying
the extent of expansion to effectively gauge the severity of deterioration
in ISR-­affected concrete. In this context, extensive investigations have been
conducted to understand the particularities associated with different ISR
mechanisms, including ASR, ACR, DEF and FT cycles.

7.3.1.1 Alkali-­silica reaction (ASR)

In an extensive investigation conducted by (Sanchez et al. 2017), the dete-
rioration caused by ASR was thoroughly examined. The study involved the
assessment of 20 different concrete mixtures, each characterized by distinct
mechanical properties (e.g., 25, 35 and 45 MPa) and incorporating a total
of 13 different lithotypes of reactive aggregates. The integrity of the concrete
was evaluated using the DRI, while the mechanical properties were assessed
through the SDT, along with measurements of ME, tensile and compressive
strength.
Remarkably, the investigation revealed a pronounced correlation between
the development of distress features and changes in the mechanical proper-
ties of the concrete specimens in relation to the extent of ASR-­induced
expansion. These significant findings are depicted in Figure 7.3.
For negligible (0.00%) and marginal levels (0.05%) of expansion, a pro-
nounced decrease in ME (up to 30%) and tensile strength (30% to 70%) is
clearly observed. It is worth noting that the majority of ASR-­induced cracks
are localized within the aggregate particles at these expansions (Goltermann
1995; Sanchez et al. 2017), as extensively discussed in Chapters 2 and 5. As
elaborated in Chapter 6, this significant decline in concrete’s ME can be
attributed to the reduced stiffness of the aggregates caused by the generation
of cracks at the submicroscopic level within the aggregate particles. Similarly,
the notable reduction in tensile strength can be attributed to the principles
of fracture mechanics (Sanchez et al. 2017; Zahedi et al. 2021). Moreover,
Chapter 6 thoroughly examines the failure mechanism of concrete under
compression, which elucidates the fact that ASR-­induced deterioration only
slightly impacts the compressive strength of affected concrete (i.e., about
5% on average) at this initial stage of expansion.
As ASR progresses and attains a moderate level of expansion (0.12%),
new cracks begin to form within the aggregate particles, while pre-­existing
ones extend in length and width, gradually infiltrating the surrounding
cement paste (Sanchez et al. 2017; Zahedi et al. 2021). Consequently, the
decline in both the ME and tensile strength of the affected concrete contin-
ues, albeit at a more gradual pace (up to 35% and 55%–75%, respectively).
 Multi-level assessment of ISR-affected concrete 127

Notably, the compressive strength experiences a minor decrease of approxi-

mately 10% at this stage.
At higher levels of expansion (0.20%), the ASR-­induced deterioration pri-
marily propagates existing cracks within the cement paste rather than gen-
erating new ones. Chapter 6 provides detailed insights into this phenomenon.
Although the ME of the concrete continues to decline at a slower rate, the

Figure 7.3 L osses in mechanical properties of ASR-­a ffected concrete incorpo-

rating distinct aggregate types versus induced expansion: (a) SDI, (b)
modulus of elasticity reduction.
128 Internal Swelling Reactions in Concrete

Figure 7.3 (Continued) L osses in mechanical properties of ASR-­affected con-

crete incorporating distinct aggregate types versus
induced expansion: (c) compressive strength reduction
and (d) tensile strength reduction.
(Adapted from Sanchez et al. 2017, 2018.)

loss in tensile strength appears to be similar to that observed at the moderate

expansion levels (0.12%). In contrast, the compressive strength of the
affected concrete begins to decline more significantly, experiencing a reduc-
tion of around 25%. Ultimately, when ASR reaches very high levels of
expansion (≥ 0.30%), a complex network of cracks permeates the entire
 Multi-level assessment of ISR-affected concrete 129

cement paste, leading to a substantial drop in the compressive strength of

the affected concrete (35%–40%). However, the loss in both ME and tensile
strength appears to level off (Sanchez et al. 2017).

7.3.1.2 Alkali-­carbonate reaction (ACR)

In the same experimental investigation conducted as part of the ASR sec-
tion, an assessment of ACR-­induced development was performed on 35 MPa
concrete mixtures incorporating a highly reactive coarse aggregate sourced
from Kingston, Ontario, Canada (Sanchez et al. 2017). ACR-­induced devel-
opment was monitored over time, and the concrete microstructure’s anal-
ysis through the DRI, along with the evaluation of ACR’s impact on the
mechanical properties of affected concrete via SDT, ME, tensile and com-
pressive strength, were conducted. While no descriptive deterioration model
has been proposed to fully elucidate the development of ACR-­induced dete-
rioration, it is evident that the microscopic damage features resulting from
ACR-­induced expansion differ significantly from those observed in the case
of ASR crack development (see Figure 7.4).
During the initial stages of the chemical reaction and at lower levels of
expansion (0.05%), the observations revealed the presence of a few closed
and open cracks within the aggregate particles. However, it is noteworthy
that significant cracking without secondary products (i.e., gel) formation
was already detected in the cement paste, primarily concentrated at the ITZ.
The mechanism behind this type of cracking at the ITZ can be attributed to
two distinct phenomena – namely, expansion of the cement paste (which is
unlikely in this case) or shrinkage of the aggregate particles, as extensively
discussed in Chapter 2.
For moderate levels of expansion (0.12%), the network of cracking within
the cement paste continues to intensify, ultimately resulting in a substantial
density of cracks in the affected concrete. Although a very different cracking
pattern is observed when compared to ASR-­ induced deterioration, the
reduction in mechanical properties seems to be quite similar; however, the
reduction in compressive strength of ACR-­affected concrete is more pro-
nounced since the primary cracks generated in the mechanism are located
within the ITZ and bulk cement paste.

7.3.1.3 Delayed ettringite formation (DEF)

Experimental campaigns have been conducted to investigate the implica-
tions of DEF on the integrity of concrete and its subsequent reduction in
mechanical properties (E. Giannini et al. 2018; Melo et al. 2023; Sanchez
et al. 2018). The findings obtained from these experimental programmes
shed some light on the crack generation and progression within the ITZ
during the initial stages of expansion (up to 0.12%). Notably, the presence
of cracks in the ITZ adversely affects the bond between the aggregate and
130 Internal Swelling Reactions in Concrete

Figure 7.4 C
 omparison of DRI between ACR and ASR: (a) plot and (b) bar charts
(Sanchez et al. 2017).

cement paste. Consequently, a significant decrease in the ME, reaching up to

50%, is observed. Moreover, the reduction in compressive strength, averag-
ing around 10%, is relatively higher when compared to ASR-­affected con-
crete. The relatively limited decrease in compressive strength at this stage of
the reaction, despite the majority of cracks being present in the ITZ, can be
 Multi-level assessment of ISR-affected concrete 131

attributed to either the highly localized nature of DEF-­induced cracking or

the potential inhibitory effect caused by the “stop/arrest mechanism” pro-
vided by aggregate particles, as previously proposed by (Mehta & Monteiro
2014; Sanchez et al. 2018).
As the expansion level increases to a higher range (0.30–0.50%), in accor-
dance with the minimum energy law, the pre-­existing cracks in the system
become more significant in terms of length and width, while the formation of
new cracks becomes limited. Cracks within the ITZ start to interconnect, giv-
ing rise to a network of cracks within the cement paste. Consequently, the ME
continues to decrease, reaching a reduction of up to 60%, while the compres-
sive strength is notably diminished by up to 40%. At very high levels of
expansion (> 0.50%), the aggregate particles undergo debonding and disag-
gregation, leading to a substantial decrease in the ME of the affected concrete,
with reductions of up to 85%. Likewise, the compressive strength experiences
a significant decline of up to 50% due to the formation of a cracking network
within the bulk cement paste (Sanchez et al. 2018) (Figure 7.5).

7.3.1.4 Freeze and thawing (FT)

Research has been conducted to understand FT-­ induced expansion and
deterioration and its impact on the mechanical properties of affected con-
crete. This understanding is crucial to understand the behaviour of concrete
under harsh climates (Sanchez et al. 2018, 2020; Zahedi et al. 2022a). The
findings obtained allowed the description of FT-­induced deterioration (i.e.,
generation and propagation) as a function of expansion, as follows.
At the initial stage of damage (0.05%), FT-­induced cracks are observed in
either the ITZ or the bulk cement paste. While the ME does not appear to
be significantly affected when compared to ASR, the tensile strength of the
concrete at this stage matches that of ASR-­affected concrete, with a reduc-
tion of approximately 60% (Komar & Boyd 2017). However, it is worth
noting that FT-­induced cracks are narrower and more widespread when
compared to DEF-­induced cracks. Consequently, a slightly higher loss of
compressive strength is observed at this stage for FT deterioration, ranging
between 15% and 30% (Sanchez et al. 2018).
As the expansion progresses to 0.12%, there is no significant change in
the ME and tensile strength when compared to the previous stage of dam-
age. However, the compressive strength continues to decrease, reaching a
reduction of approximately 20% to 35%. This decrease can be attributed to
the propagation of most FT-­induced cracks into the cement paste, either
within the ITZ or the bulk paste/pores.
At higher levels of expansion (0.20%), the FT cracks continue to propa-
gate, resulting in the interconnection of various cracks in the cement paste,
as well as some cracks within the aggregate particles. This leads to the for-
mation of an important cracking network, which significantly diminishes
the physical integrity of the affected material.
132 Internal Swelling Reactions in Concrete

Figure 7.5 M
 echanical properties losses DEF-­a ffected concrete specimens incor-
porating distinct aggregate types as a function of concrete expansions:
(a) modulus of elasticity and (b) compressive strength.

Finally, at very high expansion levels (0.30%), the loss of ME (approxi-

mately 50%) becomes almost identical to that of ASR-­affected specimens,
although still lower than DEF-­induced deterioration (Sanchez et al. 2018).
This behaviour may be attributed to the observation of the aggregate parti-
cles’ debonding in DEF deterioration, as reported by (BCA 1992; Poole &
Sims 2016; Thomas et al. 2008). Furthermore, at this expansion level, a
drop in compressive strength of around 40% to 45% is experienced, which
is nearly identical to DEF and slightly higher than ASR-­affected concrete
under similar levels of damage (Sanchez et al. 2018) (Figure 7.6).
 Multi-level assessment of ISR-affected concrete 133

Figure 7.6 M
 echanical properties losses FT-­a ffected concrete specimens incor-
porating distinct aggregate types as a function of concrete expansions:
(a) modulus of elasticity, (b) tensile strength and (c) compressive
strength.
134 Internal Swelling Reactions in Concrete

7.3.2 Using the multi-­level assessment protocol to

understand the impact of ISR under restrained
conditions
Extensive research has been conducted to evaluate the behaviour of vari-
ous ISR mechanisms under unrestrained (free expansion) conditions.
However, only a limited number of studies have investigated the influ-
ence of confinement on ISR-­induced expansion and deterioration (Zahedi
et al., 2022c, 2022d). The majority of these studies have concurred that
restraint helps mitigate ISR-­induced expansion (Multon & Toutlemonde
2006). It has also been observed the anisotropic character of ISR under
restraint since its induced expansion and deterioration are diminished in
the direction of confinement while being transferred (and thus increas-
ing) to unrestrained directions (Morenon et al. 2017; Multon et al. 2005;
Multon & Toutlemonde 2006; Zahedi et al., 2022c, 2022d). For instance,
previous observations have verified that unrestrained directions of rein-
forced concrete blocks reinforced in a single dimension (i.e., 1D) with
a reinforcement ratio of 2% exhibited approximately 40% higher ISR-­
induced expansion when compared to the confined direction (Zahedi
et al. 2021; Zahedi et al. 2022b, 2022c, 2022d). This suppression of
induced expansion in the main reinforcement direction is attributed to
the elastic restraint of the reinforcement, which induces tensile stresses in
the steel rebars.
On the one hand, anisotropy plays an important role in the deformation
of ISR-­affected members; an interesting example was observed in the study
of thick slabs bearing rectangular cross-­sections and asymmetrically distrib-
uted longitudinal reinforcement around their centroidal axis; it has been
verified that the original rectangular cross-­sections deformed into trapezoi-
dal sections after approximately 0.15%–0.20% of induced expansion
(Allard et al. 2018). Schematics illustrating such deformations can be found
in Figure 7.7. On the other hand, anisotropy can significantly influence the
orientation of induced cracks based on the boundary conditions and rein-
forcement configuration of affected concrete members. While unconfined
members typically uncover a random crack pattern (i.e., map cracking),
most ISR-­induced cracks in restrained concrete tend to be generated parallel
to the main reinforcement (Allard et al. 2018; Barbosa et al. 2018; Zahedi
et al., 2022c, 2022d). Schematics depicting the development of ISR-­induced
cracks in unrestrained, 1D and 2D reinforced concrete blocks can be found
in Figure 7.8 (Zahedi et al. 2021).
In addition to influencing crack orientation, anisotropy can also impact
the generation and propagation of cracks in reinforced concrete. Compared
to concrete under free expansion conditions, ISR-­affected reinforced con-
crete exhibits a higher number of open cracks in the cement paste, along
with evidencing a higher amount of secondary products/deposits for the
same level of development, particularly in ASR deterioration. Moreover, the
 Multi-level assessment of ISR-affected concrete 135

Figure 7.7 S chematics depicting (a) damaged ASR-­a ffected restrained concrete
block and (b) undamaged concrete block (Zahedi et al. 2022b).

Figure 7.8 T
 ypical ASR-­induced crack orientation from (a) unconfined block (top
view), (b) 1D confined block (top view), (c) 1D confined block (side/
longitudinal view), (d) 2D confined block (top view) and (e) 2D con-
fined block (side/longitudinal view) (Zahedi et al. 2021).
136 Internal Swelling Reactions in Concrete

generation of an important number of cracks in non-­reactive aggregate par-

ticles is often observed (Zahedi et al. 2022b).
The observation of such singular petrographic damage features in rein-
forced concrete can lead to the identification of different mechanical prop-
erty losses in restrained vs unrestrained members. The degradation of the
ME and compressive strength in ISR-­affected reinforced concrete follows
similar anisotropic trends (Abd-­Elssamd et al. 2020; Giannini & Folliard
2012; Hayes et al. 2020), and higher reductions in compressive strength and
ME are observed in unrestrained directions when compared to the restrained
directions (Barbosa et al. 2018; Zahedi et al., 2022c, 2022d). For instance,
the compressive strength of cores extracted perpendicular to the main
restrained direction may experience higher reductions ranging from 7% to
35% when compared to those retrieved parallel to the main rebars (Bach et
al. 1993; Barbosa et al. 2018; Jones et al. 1994; Kongshaug et al. 2020;
Morenon et al. 2017). This behaviour matches the number of cracks
observed in the cores, where cores parallel to the steel bars tend to display a
slightly lower number of cracks when compared to those extracted perpen-
dicular to the rebars (Zahedi et al., 2022c, 2022d).
Ultimately, when comparing the impact of various confinement configura-
tions (i.e., none, uniaxial, biaxial and triaxial) on ISR-­ induced damage
development, it is observed that concrete members under triaxial conditions
exhibit lower expansion, microscopic damage and mechanical degradation
than the other restraint configurations over time.

REFERENCES

Abd-­Elssamd, A., Ma, Z. J., Le Pape, Y., Hayes, N. W., & Guimaraes, M. (2020).
Effect of alkali-­silica reaction expansion rate and confinement on concrete degra-
dation. ACI Materials Journal, 117(1). https://doi.org/10.14359/51720294
Allard, A., Bilodeau, S., Pissot, F., Fournier, B., & Bissonnette, B. (2018). Expansive
behavior of thick concrete slabs affected by alkali-­silica reaction (ASR). Construction
and Building Materials, 171, 421. https://doi.org/10.1016/j.conbuildmat.2018.
03.111
Bach, F., Thorsen, T. S., & Nielsen, M. P. (1993). Load-­carrying capacity of structural
members subjected to alkali-­silica reactions. Construction and Building Materials,
7(2), 109–115. https://doi.org/10.1016/0950-­0618(93)90040-­J
Barbosa, R. A., Hansen, S. G., Hansen, K. K., Hoang, L. C., & Grelk, B. (2018).
Influence of alkali-­silica reaction and crack orientation on the uniaxial compres-
sive strength of concrete cores from slab bridges. Construction and Building
Materials, 176, 440–451. https://doi.org/10.1016/j.conbuildmat.2018.03.096
British Cement Association (BCA). (1992). The diagnosis of alkali-­silica reaction
(p. 44) [Report of a Working Party]. British Cement Association (BCA).
Giannini, E., & Folliard, K. (2012). Stiffness damage and mechanical testing of core
specimens for the evaluation of structures affected by ASR. 14th International
Conference on Alkali-­Aggregate Reaction. International Conference on Alkali-­
Aggregate Reaction, Texas, USA.
 Multi-level assessment of ISR-affected concrete 137

Giannini, E., Sanchez, L. F. M., Tuinukuafe, A., & Folliard, K. J. (2018).

Characterization of concrete affected by delayed ettringite formation using the
stiffness damage test. Construction and Building Materials, 162, 253–264. https://
doi.org/10.1016/j.conbuildmat.2017.12.012
Goltermann, P. (1995). Mechanical predictions of concrete deterioration; Part 2:
Classification of crack patterns. ACI Materials Journal, 92(1). https://doi.org/
10.14359/1177
Hayes, N. W., Giorla, A. B., Trent, W., Cong, D., Le Pape, Y., & Ma, Z. J. (2020).
Effect of alkali-­silica reaction on the fracture properties of confined concrete.
Construction and Building Materials, 238, 117641. https://doi.org/10.1016/j.
conbuildmat.2019.117641
Jones, A. E. K., Clark, L. A., & Amasaki, S. (1994). The suitability of cores in predict-
ing the behaviour of structural members suffering from ASR. Magazine of Concrete
Research, 46(167), 145–150. https://doi.org/10.1680/macr.1994.46.167.145
Komar, A. J. K., & Boyd, A. J. (2017). Evaluating freeze-­thaw deterioration with
tensile strength. IOP Conference Series: Materials Science and Engineering, 216,
012024. https://doi.org/10.1088/1757-­899X/216/1/012024
Kongshaug, S. S., Oseland, O., Kanstad, T., Hendriks, M. A. N., Rodum, E., &
Markeset, G. (2020). Experimental investigation of ASR-­affected concrete – The
influence of uniaxial loading on the evolution of mechanical properties, expansion
and damage indices. Construction and Building Materials, 245, 118384. https://
doi.org/10.1016/j.conbuildmat.2020.118384
Martin, R.-P., Sanchez, L., Fournier, B., & Toutlemonde, F. (2017). Evaluation of
different techniques for the diagnosis & prognosis of Internal Swelling Reaction
(ISR) mechanisms in concrete. Construction and Building Materials, 156, 956–
964. https://doi.org/10.1016/j.conbuildmat.2017.09.047
Mehta, P. K., & Monteiro, P. J. M. (2014). Concrete: Microstructure, properties, and
materials (4th ed.). New York: McGraw-­Hill Education.
Melo, R. H. R. Q., Hasparyk, N. P., & Tiecher, F. (2023). Assessment of concrete
impairments over time triggered by DEF. Journal of Materials in Civil Engineering,
35(8). https://doi.org/10.1061/JMCEE7.MTENG-­15041
Mohammadi, A., Ghiasvand, E., & Nili, M. (2020). Relation between mechani-
cal properties of concrete and alkali-­silica reaction (ASR); a review. Construc­
tion and Building Materials, 258, 119567. https://doi.org/10.1016/j.conbuildmat.
2020.119567
Morenon, P., Multon, S., Sellier, A., Grimal, E., Hamon, F., & Bourdarot, E. (2017).
Impact of stresses and restraints on ASR expansion. Construction and Building
Materials, 140, 58–74
Multon, S., Seignol, J.-F., & Toutlemonde, F. (2005). Structural behavior of concrete
beams affected by alkali-­silica reaction. ACI Materials Journal, 102(2), 67–76.
Multon, S., & Toutlemonde, F. (2006). Effect of applied stresses on alkali-­silica
reaction-­induced expansions. Cement and Concrete Research, 36(5), 912–920.
https://doi.org/10.1016/j.cemconres.2005.11.012
Poole, A., & Sims, I. (2016). Concrete petrography: A handbook of investigative
techniques (2nd ed.). CRC Press, Taylor & Francis Group, a Balkema book; Gale
Academic OneFile. https://doi.org/10.1201/b18688
Sanchez, L. F. M., Drimalas, T., & Fournier, B. (2020). Assessing condition of concrete
affected by internal swelling reactions (ISR) through the Damage Rating Index
(DRI). Cement, 1–2, 100001. https://doi.org/10.1016/j.cement.2020.100001
138 Internal Swelling Reactions in Concrete

Sanchez, L. F. M., Drimalas, T., Fournier, B., Mitchell, D., & Bastien, J. (2018).
Comprehensive damage assessment in concrete affected by different internal
swelling reaction (ISR) mechanisms. Cement and Concrete Research, 107, 284–
303. https://doi.org/10.1016/j.cemconres.2018.02.017
Sanchez, L. F. M., Fournier, B., Jolin, M., Mitchell, D., & Bastien, J. (2017). Overall
assessment of Alkali-­Aggregate Reaction (AAR) in concretes presenting differ-
ent strengths and incorporating a wide range of reactive aggregate types and
natures. Cement and Concrete Research, 93, 17–31. http://dx.doi.org/10.1016/
j.cemconres.2016.12.001
Thomas, M., Folliard, K., Drimalas, T., & Ramlochan, T. (2008). Diagnosing delayed
ettringite formation in concrete structures. Cement and Concrete Research, 38(6),
841–847. https://doi.org/10.1016/j.cemconres.2008.01.003
Zahedi, A., Komar, A., Sanchez, L. F. M., & Boyd, A. J. (2022a). Global assessment
of concrete specimens subjected to freeze-­thaw damage. Cement and Concrete
Composites, 133, 104716. https://doi.org/10.1016/j.cemconcomp.2022.104716
Zahedi, A., Sanchez, L. F. M., & Noël, M. (2022b). Appraisal of visual inspection
techniques to understand and describe ASR-­induced development under distinct
confinement conditions. Construction and Building Materials, 323, 126549.
https://doi.org/10.1016/j.conbuildmat.2022.126549
Zahedi, A., Trottier, C., Sanchez, L. F. M., & Noël, M. (2021). Microscopic assessment
of ASR-­affected concrete under confinement conditions. Cement and Concrete
Research, 145, 106456. https://doi.org/10.1016/j.cemconres.2021.106456
Zahedi, A., Trottier, C., Sanchez, L. F. M., & Noël, M. (2022c). Condition assessment
of alkali-­silica reaction affected concrete under various confinement conditions
incorporating fine and coarse reactive aggregates. Cement and Concrete Research,
153, 106694. https://doi.org/10.1016/j.cemconres.2021.106694
Zahedi, A., Trottier, C., Sanchez, L., & Noël, M. (2022d). Evaluation of the induced
mechanical deterioration of alkali-­silica reaction affected concrete under distinct
confinement conditions through the stiffness damage test. Cement and Concrete
Composites, 126, 104343. https://doi.org/10.1016/j.cemconcomp.2021.104343
Chapter 8

Forecasting future performance and

managing critical infrastructure

8.1 INTRODUCTION

The management of internal swelling reaction (ISR)–affected concrete infra-

structure comprises three main aspects: (a) diagnosis, (b) prognosis and (c)
maintenance strategies. The diagnosis task aims to find the main cause(s)
leading to deterioration and appraises the current condition (i.e., dam-
age degree) of affected structures or structural members under evaluation.
Otherwise, the prognosis consists of assessing the potential for further dete-
rioration and the associated structural implications over time. Finally, the
outcomes gathered from diagnosis and prognosis are critical for determin-
ing appropriate maintenance strategies.
Chapters 1 and 2 presented in detail the most common ISR mechanisms
inducing expansion and deterioration in concrete and the need for a proper
diagnosis, whereas Chapters 3 to 7 displayed numerous diagnosis tech-
niques and approaches (i.e., visual, non-destructive, microscopic and
mechanical tests), enabling a reliable appraisal of the main cause(s) and
extent of deterioration of affected concrete. This chapter aims to discuss the
most common prognosis procedures, frameworks and models adopted in
the literature to evaluate the potential of further development (i.e., expan-
sion and deterioration) of ISR in concrete. It also highlights possible reha-
bilitation strategies and available management protocols.

8.2 PROGNOSIS AND THE POTENTIAL FOR FURTHER

DETERIORATION

The prognosis of ISR-affected infrastructure should consider the likelihood

of future deterioration and its associated structural implications. Therefore,
laboratory test methods coupled with mathematical models are normally
used for this purpose. Likewise, field monitoring (i.e., measurements of dis-
placement, temperature, relative humidity, pressure, etc.) may also be effec-
tively implemented as input or fitting parameters in mathematical models,
supporting infrastructure owners in decision-making. The following sections

DOI: 10.1201/9781003188155-8 139

140 Internal Swelling Reactions in Concrete

will present the most common laboratory techniques and engineering-based

models to forecast the behaviour of ISR-affected concrete.

8.2.1 Laboratory techniques
Laboratory test procedures are normally the first approach used to evalu-
ate the potential for future deterioration of ISR-affected concrete. Various
test methods were developed over the last decades for this purpose, and
although far from being perfect, they may provide valuable information.
These tests are normally based on expansion or chemical procedures.

Expansion procedures
• Objective: Expansion procedures aim to estimate the potential of
future expansion and deterioration of ISR-affected concrete. These
procedures are generally divided into two categories: (a) assessing
residual expansion or (b) absolute expansion. Although these tests do
not represent the behaviour of affected structures or structural mem-
bers in the field, they may provide valuable insights into ISR-induced
development, such as the expected expansion kinetics (or rate) and its
potential of levelling off over time (Merz & Leemann, 2013).
• Sample preparation and methodology: Expansion procedures are
conducted following the coring of regions of interest of affected struc-
tures or structural members. Some common procedures should be per-
formed to avoid variability in the test outcomes; first, the extracted
cores from the distinct regions of interest should be carefully labelled
and wrapped in plastic film to avoid moisture loss. Then, the wrapped
cores should be kept in a controlled environment with a tempera-
ture of 23°C ± 2°C for five days. This allows for achieving consistent
and homogeneous moisture content within the samples. Afterwards,
the cores’ ends are thoroughly prepared through cutting and grind-
ing (or capping) processes until the desired length-to-diameter ratio
of 1:2 (or close to) is achieved. Finally, studs are installed at the core
ends to enable expansion monitoring over time. Figure 8.1 illustrates
examples of cores’ extraction (Figure 8.1a) labelling (Figure 8.1b),
end studs’ installation and storage (Figure 8.1c), and measurement
(Figure 8.1d).

A number of test protocols have been developed worldwide for evaluating

the expansion potential involving different test set-ups, environmental con-
ditions, and measuring parameters (Saouma, 2021). Table 8.1 presents a
comprehensive overview of current testing protocols, including set-up varia-
tions and primary outcomes for distinct ISR mechanisms. Among these, the
most common procedure used in Canada, North America and various coun-
tries in Europe involves subjecting core samples to 38°C and 100% relative
humidity for a duration of one year (or until the expansion levels off) to
 Forecasting future performance and managing critical infrastructure 141

Figure 8.1 (a) Cores’ extraction, (b) Core’s labelling, (c) End studs’ installation
and storage and (d) Core’s measurement. Core sample preparation
and measurement.
(Photos courtesy of Leandro Sanchez.)

determine the residual expansion potential (EXP) of the concrete. This pro-
cedure is normally used to assess AAR and AAR + DEF potential for further
deterioration. Whether DEF is primarily intended to be appraised, the core
specimens should be soaked in water at 38°C and monitored for one year
(or until the expansion levels off) instead of stored under 100% RH
(Kawabata et al., 2016; Martin, 2010; Martin et al., 2017).
Another approach commonly used during the appraisal of future deterio-
ration caused by AAR entails the soaking of core specimens in 1M NaOH
solutions at 38°C for one year; this procedure aims to measure the degree of
absolute reactivity of aggregates (ABR). Figure 8.2 illustrates the test set-ups
commonly used to obtain (EXP) and (ABR) outcomes.
Residual expansion tests are quite easy and relatively non-expensive pro-
cedures enabling the evaluation of the future expansion behaviour of ISR-
affected concrete. However, important discrepancies are often observed
when comparing the results obtained in the laboratory with the field perfor-
mance of affected structures. Issues such as alkalis leaching, alkali release
from aggregates, restraint/confinement, load and environmental conditions,
which are different in the lab and field, have been pointed out as the main
142 Internal Swelling Reactions in Concrete

Table 8.1 ISR residual expansion testing protocols

ISR mechanisms Test set-up Outcomes Reference

AAR 95% RH and Phases of residual Canadian Standards

38°C expansion Association (CSA)
(2000); ISE (1992);
LCPC (2003);
Multon et al. (2008)
AAR 1M NaOH and Residual aggregate Bérubé et al. (2002)
38°C reactivity potential
AAR 1M NaOH and Residual expansion Gao et al. (2011)
60°C testing aggregates
from the structure
AAR 0.7M NaOH Residual aggregate Zubaida (2020)
and 38°C, reactivity potential
wrapped (improvement of
samples 1M NaOH and
38°C method)
AAR 4% NaCl and Structures exposed Swamy (1991, 1997)
38°C to salt
AAR 1M NaOH and Faster in reaching the Bérubé et al. (2002)
80°C ultimate residual
expansion plateau
DEF Soaked in Remaining potential Martin et al. (2017)
water and of DEF-affected
38°C concrete
DEF, DEF + ASR Lime water Remaining potential Folliard et al. (2006);
and 23°C of DEF-affected Ramlochan et al.
concrete (2003)

Figure 8.2 R
 esidual test set-ups performed to measure (a) the residual EXP and
(b) the ABR. EXP can be used in both AAR and DEF cases, whereas
ABR is used in AAR appraisal.
 Forecasting future performance and managing critical infrastructure 143

causes leading to the discrepancies verified (Bérubé et al., 2002; Sellier et al.,
2009). Therefore, research is still required to improve the current test set-
ups to better represent field performance. Nevertheless, more important
than discussions on the reliability and efficiency of distinct residual expan-
sion protocols is the interpretation and usage of the test outcomes obtained
from them, which is a challenging task and should be conducted carefully
and in a conservative fashion, accounting for the potential differences
between lab and field.

Soluble alkalis
• Objective: Besides expansion tests, some chemical procedures may also
estimate the potential of further expansion and deterioration of ISR-
affected concrete; amongst them, the soluble alkalis method stands
out as a promising protocol for appraising ISR future deterioration,
particularly for AAR-affected concrete.
It is widely known that the higher the amount of alkalis in the con-
crete pore solution (i.e., Na+, K+ and OH-), the higher the likelihood of
AAR-induced expansion and damage; therefore, assessing the remain-
ing alkali content at a given age may provide insight into the remain-
ing expansion yet to take place.
• Sample preparation and methodology: A few methods have been
proposed over the past few decades to quantify the soluble alkalis in
concrete; some of them include the extraction of concrete pore solu-
tion under pressure, while the vast majority consist of filtering under
pressure of crushed and finely-ground concrete samples, followed by
analysis of the alkali concentration in the obtained solution.
Overall, the sample preparation process prior to filtering under
pressure begins with the longitudinal splitting of concrete cores. The
split core is then broken into large particles measuring approximately
25 mm. Subsequently, the sample undergoes multiple crushing stages
until particle sizes of about 5 mm are reached. From this point, a rep-
resentative subsample weighing 1 kg is selected and pulverized until
the material reaches particle sizes of 150 μm. By splitting the material
once more, two or three subsamples of 10 g are obtained and stored in
airtight bags to prevent carbonation until testing.

The following techniques are considered the most reliable for extracting the
pore solution and determining the pH and free alkali content: pore-water
expression (PWE), in situ leaching (ISL), ex situ leaching (ESL), cold water
extraction (CWE), hot water extraction (HWE) and espresso method (EM).
Brief descriptions of these techniques are presented hereafter:

• PWE: This technique involves subjecting concrete samples to high

pressure, resulting in pore solution extraction. PWE is a simple, effec-
tive and quite reliable test procedure widely implemented to evaluate
144 Internal Swelling Reactions in Concrete

the alkalis composition of concrete pore solution. PWE can be used

for cement paste, mortar and concrete samples, yet attention should
be paid to the interference of aggregates on the test outcomes, particu-
larly for concrete specimens.
• ISL: This method is developed by the creation of a small hole in the
concrete member under analysis, followed by the injection of a leach-
ing solution; the concrete pore solution is then dissolved into the
leaching solution, which is collected for further chemical analysis.
• ESL: ESL corresponds to the extraction and soaking of concrete cores
into a leaching solution, allowing the alkalis to dissolve; the solution
is then chemically evaluated and the alkalis amount quantified.
• CWE: CWE is a quite used technique due to its simplicity and minimal
equipment requirements. Concrete powders (i.e., particles < 80 μm)
are immersed in cold water (liquid-to-solid ratio equal to 1) for a 5
min leaching time. During this process, alkalis from the pore solution
diffuse into water, which is then filtered and evaluated.
• HWE: This method follows the same general procedure of the CWE,
yet concrete powders (i.e., particles < 160 μm) are boiled in 100 ml of
deionized water over 10 min and left standing overnight, enhancing
the diffusion of alkalis into the water.
• EM: This method involves the extraction of soluble alkalis by passing
300 ml of deionized boiling water through the concrete sample (i.e.,
10 g of particles < 150 μm). Before being analysed, the solution is
topped up to 500 mL with distilled water.

The literature shows that although all of the aforementioned methods dis-
play advantages and limitations, the EM is the technique that provides
more accurate and less variable outcomes to assess the potential of further
AAR-induced development (Plusquellec et al., 2017). However, there is still
a debate in the research community on how to use those test outcomes to
quantitatively estimate the potential of further AAR expansion in the field.
On the other hand, the prognosis of other ISR mechanisms, such as DEF
via pore solution analysis, is quite complex, and there is currently a lack
of research, methods and quantitative data in this regard. However, it is
accepted that the determination of pH, alkalis and sulphate contents of pore
solution may help in understanding the potential of future deterioration
caused by any ISR mechanism.

8.2.2 Prognosis estimation based on laboratory test

procedures
Besides conducting tests that provide physical (i.e., expansion) or chemi-
cal (i.e., pore solution analysis) outcomes “individually”, understanding
how to combine and implement them into a prognosis protocol is prob-
ably amongst the most current challenges in the area. In this context,
 Forecasting future performance and managing critical infrastructure 145

Bérubé et al. proposed a framework for combining the information from

laboratory tests into a protocol to estimate the potential expansion rate
of AAR-affected structures in the field. Figure 8.3 illustrates the proposed
framework (Bérubé et al., 2002), which is based upon research projects
dealing with the assessment of AAR-affected dams.
In this framework, the residual and absolute expansions (EXP and ABR,
respectively) are combined with soluble alkalis results (ALK) along with ser-
vice conditions (i.e., temperature, relative humidity, confinement/restraint) to
estimate the potential of further expansion (PFE) caused by AAR.
The coefficient EXP is normally the first coefficient gathered through
residual expansion tests conducted at 38°C and 100% RH (see Section
8.2.1); this test is considered the most realistic laboratory test procedure
to assess the PFE of AAR-affected members in service since the cores
retrieved from affected structures are tested with their actual alkali

Figure 8.3 F ramework for estimating AAR-induced expansion rate in the field.
146 Internal Swelling Reactions in Concrete

Table 8.2 Classification of expansion potential, from Bérubé et al. (2002b)

Coefficient EXP – Residual concrete expansivity in the laboratory (core testing in

air at > 95 RH and 38°C)

Exp./year % Residual exp EXP Exp./year % Residual exp EXP

< 0.003 Negligible 0 0.015 to 0.02 Moderate 6

0.003 to 0.005 Very low 1 0.02 to 0.025 High 9
0.005 to 0.01 Low 2 0.025 to 0.03 High 12
0.01 to 0.015 Moderate 4 > 0.03 Very high 16

content (Bérubé et al., 1994). Annual residual expansion results ranging

from negligible (i.e., less than 0.003% per year) to very high expansion (i.e.,
over 0.03% per year) may be obtained (Bérubé et al., 2002; Fournier et al.,
2010) as per Table 8.2. However, it should be mentioned that EXP can often
be underestimated due to alkalis leaching taking place over the test.
Therefore, if the coefficient EXP is considered relatively low, it should be
replaced by the product of the two coefficients ABR and ALK. The ABR
coefficient is obtained following a one-year test at 38°C and 1M NaOH%
(see Section 8.2.1), and the results can be classified from negligible (i.e., less
than 0.04%) to very high (i.e., higher than 0.20%), as per Table 8.3.
Otherwise, the soluble alkalis, and thus the coefficient ALK, is measured as
per one of the procedures described in Section 8.2.1. Bérubé et al. proposed
the use of the HWE method for such purpose (Bérubé et al., 1994). This
index, which varies from very low (i.e., < 1.0 kg/m3) to very high (i.e., > 2.5
kg/m3), as illustrated in Table 8.4, indicates the remaining alkalis available
to keep inducing further AAR expansion in the field. It is worth noting that

Table 8.3 C
 lassification of the absolute degree of reactivity of the aggregates
present in the concrete under study, based on expansion tests on cores
in 1 N NaOH solution at 38°C or on CSA A23.2-14A-94 concrete
prism tests on aggregates extracted from cores (ABR) (CSA, 2014)

1-year expansion (%) (After Absolute degree of Coefficient

preconditioning in the case of cores) reactivity of aggregatesa ABRa

< 0.04 Negligible 0 (or ≥ 0)

0.04 to 0.08 Low 1 (or ≥ 1)
0.08 to 0.12 Medium 2 (or ≥ 2)
0.12 to 0.20 High 3 (or ≥ 3)
> 0.20 Very high 4
a When testing cores, the qualification of the absolute degree of reactivity and the correspond-
ing coefficient ABR are considered as minimum if the concrete specimens were abnormally
fissured or porous compared to the overall concrete member under study or if the concrete
is quite impermeable to the alkaline solution.
 Forecasting future performance and managing critical infrastructure 147

Table 8.4 C
 lassification of the water-soluble alkali content of the
concrete under study, as measured by the HWE method,
after correction for aggregate contribution (ALK)

Corrected soluble alkali

content (kg/m3 Na2Oeq) Classification Coefficient ALK

< 1.0 Very low 0

1.0–1.5 Low 1
1.5–2.0 Medium 2
2.0–2.5 High 3
> 2.5 Very high 4

if, on the one hand, significant expansion obtained over the aforementioned
expansion tests does not necessarily mean that the concrete under study will
swell in service; for instance, if the humidity conditions in the field are too
low to sustain alkali-silica reaction (ASR) and or the restraints a confine-
ment effects are sufficiently high to suppress induced expansion. On the
other hand, low expansions indicate that the concrete under study should
not swell in service unless the concrete specimens tested are very deterio-
rated and or are much more porous than the concrete member under analy-
sis. In other words, the samples used for testing do not represent, on average,
the member appraised in the field. Once EXP, ABR and ALK are determined,
parameters related to field conditions (i.e., humidity, temperature and stress
state) should be obtained.
The humidity coefficient (HUM) can be gathered via commercial probes
that can measure internal humidity along small drill holes (Bérubé et al.,
1994); the range of values that can be obtained and their associated AAR
risk are displayed in Table 8.5. If internal humidity values are not available,
external humidity could be used to estimate the risk for AAR continuation
as per Table 8.6.
The TEM is proposed to account for the effect of temperature on AAR-
induced development; for example, the expansion tests conducted in the

Table 8.5 C
 lassification of internal humidity conditions
as regards the risk of ASR (HUM)

Relative humidity inside

concrete (% RH) “Humidity risk” for ASR Coefficient HUM

< 80 Very low 0

80–85 Low 0.25
85–90 Medium 0.5
90–95 High 0.75
95–100 Very high 1
148 Internal Swelling Reactions in Concrete

Table 8.6 C
 lassification of external humidity conditions as regards
the risk of ASR (HUM)

Non-massive concretea Mass concretea

Humidity conditions “Humidity Coefficient “Humidity Coefficient

(%RH) risk” for ASR HUM risk” for ASR HUM

Interior < 70% Very low 0 Low 0.25

Interior 70%–80% Low 0.25 Medium 0.5
Interior 80%–90% Medium 0.5 High 0.75
Interior 90%–95% High 0.75 Very high 1
Interior 95%–100% or ’ 1
Very high 1
immersed
Exterior not exposed High 0.75
Medium 0.5
to rainb
Exterior exposed to Very high 1
High 0.75
rainb
Exterior immersed or ’ 1
Very high 1
buried

a  If the concrete component is at the same time exposed to humid and dry conditions, there is a
risk for anisotropic expansion.
b In the case of a tempered climate as the one prevailing in North America.

laboratory are normally performed at 38°C, whereas most concrete struc-

tures exposed outdoors (e.g., North America) are submitted to yearly aver-
age lower temperatures. The idea of TEM is thus to take into account this
effect, as per Table 8.7.
The coefficient of confinement/restraint (STR) is the most difficult param-
eter to estimate due to the often-limited amount of information (i.e., draw-
ings, members detailing, etc.) of ageing infrastructure. The values proposed
in Table 8.1 are based on “average results” encountered in the Institution of
Structural Engineers report (ISE, 1992). It is important to note that the STR
coefficient is valid in the direction parallel to the main rebars since literature

Table 8.7 P
 roposed values for the coefficient of thermal correction temperature
coefficient (TEM)

Annual average temperature (°C)a Coeff. of thermal correction TEM

<0 0.4
0–10 0.55
10–20 0.7
20–30 0.85
> 30 1.0

a In most urban areas in Canada, the annual average temperature is between 0°C and 10°C.
 Forecasting future performance and managing critical infrastructure 149

Table 8.8 P
 roposed values for the coefficient of correction for reinforcement
and stresses applied to concrete in service (STR)

Amount of steel Coeff. of correction Internal or external Coeff. of correction

reinforcement (%)a for stresses STR restraints (MPa)b for stresses STR

0 1.0 0 1.0
0.25 0.75 0.25 0.85
0.5 0.55 0.5 0.7
0.75 0.4 0.75 0.55
1 0.3 1 0.4
2 0.25 1.5 0.3
≥3 0.2 2 0.2
≥3 0.1

a  When rebars are installed on a single plane (1D or 2D reinforcement) or on many parallel planes
(2D or 3D reinforcement) but without any anchorage between the different planes, the coefficient
STR applies in the direction(s) of the rebars only.
b  In the cases of uniaxial (1D) or biaxial (2D) compressive stresses, the coefficient STR applies in the

direction of the stresses only.

shows that due to anisotropy, induced expansion is redistributed and

becomes higher in directions perpendicular to the main reinforcement bars.
Finally, upon obtaining all parameters, the potential of further expansion
(PFE) of the structure or structural member under evaluation can be computed
and correlated with a potential expansion rate(% per year) as per Figure 8.3.

8.2.3 Modelling
Mathematical models play a crucial role in the representation and under-
standing of physicochemical phenomena. A wide range of models have been
developed in the past to describe ISR-induced expansion and deterioration,
from describing the chemical phenomena to evaluating the performance and
capacity of critical infrastructure in the field. The goal of this chapter is not
to debate the pros and cons of each of those methods. Moreover, it is out
of the scope of the current book to discuss “structural” models, normally
run by finite elements, and to appraise service and ultimate limit states of
affected structures. This is considered the following step of our proposed
condition assessment protocol, normally conducted by structural engineers
after diagnosis and prognosis tasks. Yet, the idea of this section is to discuss
models that aim to evaluate the potential of further continuation of ISR-
induced development and ultimately estimate the expansion as a function of
time of the affected concrete in the “material’s scale”.
ISR models can be divided into four categories: micromodels (based on
ion diffusion/reaction products), micro-mesomodels (based on secondary
products generation), mesomodels (based on internal pressure) and macro-
models (based on expansion). Micro-, micro-meso- and mesomodels are
150 Internal Swelling Reactions in Concrete

often considered for materials’ scale appraisal, whereas macromodels can be

used for material and structural evaluations (Esposito, 2015).
One example of a very promising micro-mesomodel to describe AAR
effects in concrete is a chemo-mechanical elastic model developed by
researchers from the University Paul Sabatier in Toulouse, France (Multon
et al., 2009). This model considers the chemical interactions between the
alkalis in the concrete pore solution and the non-stable siliceous phases
from the aggregates within a representative elemental volume (REV) of con-
crete; parameters such as alkali diffusion into aggregates, secondary prod-
ucts formation and deposition in the concrete porosity are considered. As a
result, a volumetric amount of AAR secondary products is formed and
induces expansion, as illustrated in Figure 8.4; moreover, these products,
after filling the initial porosity of the concrete, produce pressure which
induces cracks in the cement paste once the tensile stresses generated exceed
the tensile strength of the concrete.
The initial calibration of this chemo-mechanical elastic model was based
on experimental data obtained in mortars, which yielded satisfactory results.
The model was further validated using extensive data on crack generation

Figure 8.4 M
 icro-mesomodel to describe AAR-induced product formation (a)
and stresses and cracking generated (b).
 Forecasting future performance and managing critical infrastructure 151

and propagation and mechanical properties reductions from Sanchez et al.

(Sanchez et al., 2014), demonstrating its promising character to predict
induced expansion and damage. Another example of an interesting AAR
mesoscopic model is the one proposed by Dunant and Scrivener at EPFL in
Lausanne, Switzerland (Dunant & Scrivener, 2010). This approach uses a
finite element framework, where the aggregates are assumed to be com-
pletely spherical, both the aggregates and cement paste are considered
quasi-brittle, and the secondary products generated behave in a linear-elas-
tic fashion. The model considers that AAR randomly generates secondary
products that fill voids, flaws or previous cracks within the reactive aggre-
gates, tending to form “pockets” within the particles. An enrichment func-
tion (Moës et al., 2003) is used to simulate the exact contact between the
secondary products and the aggregates. Induced expansion is then devel-
oped whenever the amount of products overcomes the initial porosity of
the aggregates. Furthermore, cracks and reduced stiffness are computed
once expansion is induced. As for the chemo-mechanical elastic model,
encouraging correlations were observed between simulations and experi-
mental data (Haha et al., 2007) concerning crack propagation and stiffness
reduction (Dunant & Scrivener, 2010).
Macromodels, primarily aiming to assess induced expansion, which
account indirectly for the distinct aspects of ISR physicochemical processes,
are probably the most used type of models adopted for engineering and
decision-making purposes. Most of these models are based on Larive’s
model (Larive, 1997), which is a semi-empirical approach that was devel-
oped to appraise ASR-induced expansion after a large experimental cam-
paign in the laboratory. The model describes induced expansion and ASR
kinetics using three parameters: latency (τl), characteristic (τc) and ultimate
expansion (ε∞), as shown in Equation 8.1.
t
−
1 − e c( )
τ θ
ε ( t, θ )
=
( t −τ l (θ ))
∗ε∞
1 + e c( )
τ θ
(8.1)
The induced expansion described by Larive’s model often presents an S-shape
(i.e., sigmodal) curve, which may be broken down into four phases, as illus-
trated in Figure 8.5. Initially, in phase A, the curve represents the formation
and accommodation of ASR secondary products within the reactive aggre-
gate particles and surrounding flaws in the cement paste (i.e., ITZ), result-
ing in little to no expansion. The ascending period (phase B) of the curve
corresponds then to the swelling of ASR secondary products upon mois-
ture uptake from the surroundings; during this phase, significant expansion
and crack formation are expected, explaining the fast kinetics and convex
shape of the curve. The subsequent phase (phase C), distinguished from the
previous by a concave behaviour, suggests the decrease in expansion rate
due to the increased number of cracks generated in the system, creating
152 Internal Swelling Reactions in Concrete

Figure 8.5 S -shape curve described by Larive’s model for ASR-induced development.

a supplementary space to accommodate secondary products. In the final

phase, or phase D, the reactants (i.e., alkalis and/or silica) are expected to be
depleted from the system, which makes the curve level off.
Larive’s model has shown to be a quite promising approach to describe
ASR-induced development and has been successfully used in numerous
applications; however, since latency (τl) and characteristic (τc) parameters
are primarily mathematical terms rather than parameters representing phys-
ical or chemical variables; they may be hardly estimated and are often “fit”
in a case-by-case approach to match the behaviour observed by the affected
concrete under evaluation. This aspect limits its ability to predict behaviour
of ASR-affected concrete in the field whether no monitoring or tests are
conducted.
Trying to better understand and represent the influence of key compo-
nents of ASR on latency (τl) and characteristic (τc) parameters, De Grazia
et al. (2021) proposed that measurable parameters, such as temperature,
aggregate type (i.e., fine or coarse) and reactivity (i.e., marginal, moderate,
high and very high), humidity and alkali content be used to “calibrate” and
or modify latency (τl) and characteristic (τc) parameters. The authors sug-
gested that new coefficients should be added in the initial Larive’s equation
for this purpose; these coefficients were established in the laboratory using
a wide variety of test procedures, mix proportions and aggregate reactivity
levels. The modified form of Larive’s model considering the new coefficients
is presented in Equation 8.2.

t
−
τ ckc ,T kc ,RH kc ,% Akc ,E
1− e
( t, θ ) ε=
ε= (t ) ε
∞
( t −τ l kL,T kL,RH kL,%A kL,E )
∗ ( kInf ,T kInf , RH kInf , E kInf ,%A ) ε ∞ ,
−
τ ckc ,T kc ,RH kc ,% Akc ,E
1+ e (8.2)
 Forecasting future performance and managing critical infrastructure 153

where t is the elapsed time; ε(t) is the expansion at a given elapsed time; ε∞ is
the maximum expansion at infinity (or ultimate expansion); τc is the charac-
teristic time (as a function of the aggregate type and nature/reactivity); kc, T,
kc, RH, kc, % A, kc, E is the temperature, humidity, alkali content and exposure
coefficients impacting the characteristic time; kc, Tkc, RHkc, % Akc, E is the tem-
perature, humidity, alkali content and exposure coefficients impacting the
latency time; kInf, T kInf, RH kInf, E kInf, % A is the temperature, humidity, exposure
and alkali content coefficients influencing the maximum expansion.
The model proposed by De Grazia et al. (2021) showed promising results
in describing the behaviour of ASR and ACR-induced expansion in the labo-
ratory, and it has even been adopted and slightly modified to appraise an
ASR-affected reinforced concrete overpass after nearly 50 years of service in
Quebec City, Canada (Gorga et al., 2018). Yet, it is worth noting that De
Grazia’s model has been primarily developed to estimate the behaviour of
AAR-induced development in the laboratory, and thus care should be taken
for its use in field applications.
Another contribution to the use of semi-empirical macromodels to
describe AAR-induced expansion has been made by Nguyen et al. (2022). In
this model, the conventional Larive’s model is adopted, and its latency (τl)
and characteristic (τc) parameters are calibrated by laboratory tests (i.e.,
concrete prism test – CPT) and then used to estimate the behaviour of con-
crete exposed to field conditions but bearing the same raw materials and
mix proportions to the mixture tested in the laboratory. To improve the cor-
relation between laboratory and field, this model accounts for (a) the amount
of alkalis leached over the accelerated test in the laboratory through the
computation of a “virtual expansion” showing no-leaching and (b) the
potential alkalis contribution from the aggregates. Figure 8.6 illustrates the
flowchart proposed by Nguyen et al. (Nguyen et al., 2022) with the concepts

Figure 8.6 F lowchart proposed for the semi-empirical approach as per (Nguyen
et al., 2022).
 154 Internal Swelling Reactions in Concrete

of the proposed approach. Strong correlations to estimate field behaviour

from laboratory testing were obtained, showing a promising character of
the proposed model (Nguyen et al., 2022).
Although the number of models developed for AAR applications is
probably much higher than for other ISR mechanisms, numerical works
have also been conducted to describe induced expansion caused by inter-
nal sulphate attack, particularly DEF. As for AAR, most of the approaches
use Larive’s model to describe the expansion development over time. In
this context, an interesting approach has been proposed by Brunetaud
(Brunetaud, 2005), where a new time-dependent parameter was pro-
posed to Larive’s model to better describe the long-term behaviour of
DEF in concrete. The proposed model is presented in Equation 8.3, and
it has shown great results in describing DEF-induced expansion in the
laboratory.

 t 
− 
 τc 
1− e  Φ 
=ε (t ) ε ∞ ∗ 1 − ,
 t τl 
− +   δ + 1 
 τc τc 
1+ e (8.3)

where Φ and δ are corrective parameters, and t is elapsed time (Kawabata

et al., 2016).
As presented in Chapter 2, AAR and DEF are often combined in the field
due to their synergetic character. Therefore, developing a model that couples
both phenomena is crucial. Martin (Martin, 2010) proposed a novel semi-
empirical approach for coupled AAR and DEF mechanisms, which is based
upon Larive’s (1997) and Brunetaud’s (2005) approaches for AAR and DEF,
respectively. The proposed model (presented in Equation 8.4) accounts for
the contribution of both AAR and DEF separately, yielding a total expan-
sion over time for each mechanism.

 t 
 − τ C 
1− e  DEF 
ε (t ) =
ε DEF + ε ASR = ε ∞ _ DEF ∗  t −τ LDEF 
 
τ CDEF 
1 + e
 t 
 − τ 
 Φ  1− e  C _ ASR 
× ∗1 −
δ + 1  + ε ∞ _ ASR ∗ ×  t −τ L _ ASR 
   τ 
1 + e C _ ASR  (8.4)

Martin et al. (2010) successfully applied this combined model to repre-

sent experimental data obtained from specimens distressed by the combined
effects of ASR and DEF at 100% RH, effectively capturing the expansion
over time caused by each mechanism, as per Figure 8.7.
 Forecasting future performance and managing critical infrastructure 155

Figure 8.7 E xperimental versus numerical results.

(Adapted from R. P. Martin et al., 2010 approach.)

8.3 MANAGEMENT AND REHABILITATION OF

ISR-AFFECTED CONCRETE INFRASTRUCTURE

Upon identifying the current cause(s) and extent of damage, as well as esti-
mating future deterioration and likely structural implications of ISR-affected
infrastructure, efficient management protocols should be implemented to
better cope with the deterioration as a function of time. Normally, man-
agement protocols comprehend two major steps: (a) a series of field and
laboratory test procedures along with modelling, to appraise and ensure
serviceability and safety as a function of time, and; (b) rehabilitation strate-
gies aiming to stop or at least mitigate the rate of deterioration over time.
The next two sections will cover the most common management protocols
and rehabilitation procedures implemented in ISR-affected structures.

8.3.1 Management protocols
From previous chapters’ discussions, it is quite clear that management pro-
tocols for ISR-affected infrastructure should bear at least four distinct yet
complementary steps: (a) field assessment, where visual and non-destructive
techniques are used to gather preliminary information; (b) laboratory tests,
where chemical, microscopic and mechanical tests are conducted to under-
stand the cause(s) and extent of deterioration (i.e., diagnosis); (c) data analy-
sis, where the prognosis appraisal along with the current and future structural
implications are assessed; and (d) decision-making, where maintenance and
156 Internal Swelling Reactions in Concrete

rehabilitation strategies are decided. Figure 8.8 illustrates such a hypotheti-

cal four-phase management protocol.
Over time, a few specific management protocols have been developed to
help infrastructure owners better cope with ISR-affected infrastructure.
Amongst them, four protocols stand out as promising frameworks, starting
from literature survey and condition assessment to rehabilitation strategies:
(a) Bérubé et al. (Bérubé & Fournier, 2005), (b) RILEM (Godart et al.,
2013), (c) (Fournier et al., 2010) and (d) IFSTTAR (LCPC, 2003). Figure 8.9
illustrates the content of the proposed flowcharts.
From the aforementioned flowcharts, one observes that the proposed
management protocols bear similarities and differences, yet the four
phases illustrated in the hypothetical ISR-management protocol appear
(with more or less detail) in each of them. Moreover, these protocols
incorporate several of the field and laboratory test procedures (i.e., visual
inspection, cracking monitoring, non-destructive testing, microscopy and
mechanical tests) discussed in detail in Chapters 3, 4, 5 and 6. Prognosis
is also discussed on them, based on laboratory tests and modelling (Section
8.2 of this chapter). However, they are rather descriptive and thus subjec-
tive by nature, heavily relying on the expertise of the engineer(s) dealing
with the affected structure to decide whether further assessment is
required following visual inspection. Nevertheless, some descriptive and

Figure 8.8 C omprehensive management protocol for ISR-affected structures.

Figure 8.9 C
 omparison among the proposed management protocols: Bérubé
et al., RILEM, FHWA and IFSTTAR.
 Forecasting future performance and managing critical infrastructure 157

semi-quantitative guidelines are provided to better guide infrastructure

owners in decision-making, particularly regarding surface deterioration
signs, such as the classification of the probability of AAR (based on expan-
sion/displacement of members and surface deterioration), the classifica-
tion of the AAR development degree and the classification of the
importance of crack widths and cracking index values, as illustrated in
Tables 8.9 to 8.11, respectively.
Regardless of the protocol used, if the engineer(s) in charge of the evalua-
tion considers that further assessment is required, coring is then conducted
and laboratory (i.e., chemical, microscopic, mechanical, etc.) tests are per-
formed with the aim of diagnosing the current deterioration (i.e., cause and
extent of damage). For this step, various approaches are available in the lit-
erature, yet a broad discussion, along with the implementation of the multi-
level assessment, considered a quite reliable protocol, is presented in Chapter
7 of this book. Upon diagnosis completion, the prognosis is required and
then the evaluation of the potential of further deterioration may be achieved
via laboratory tests (e.g., residual expansion, soluble alkalis), modelling or a

Table 8.9 Classification of probability of AAR occurrence

Probability of AAR occurrence

Feature Low Medium High

Expansion and/or None Some evidence (e.g., Fair to extensive signs

displacement of closure of joints in of volume increase
elements pavements, Jersey leading to spalling at
barriers, spalls, joints, displacement
misalignments between and/or misalignment
structural members) of structural members
Cracking and None Some crack patterns Extensive map cracking
crack pattern typical of ASR (e.g., or cracking aligned
map cracking or cracks with major stress or
aligned with major reinforcement
reinforcement or
stress)
Surface None Slight surface Many cracks with dark
discoloration discoloration associated discoloration and
with some cracks adjacent zones
of light-coloured
concrete
Exudations None White exudation around Colourless, jelly-like
some cracks, possibility exudations readily
of colourless, jelly-like identifiable as ASR
exudations gel associated with
several cracks

Source: Modified from (CSA A864-00 (R2005), 2005)

158 Internal Swelling Reactions in Concrete

Table 8.10 Classification of AAR development degree (CSA A864-00 (R2005), 2005)

AAR development degree Nature extent of features

Low No gel present, no sites of expansive reaction, presence

of other indicative features were rarely found. Small
amount of isolated ASR gel was sporadically present in
the cement paste, but very few reactive particles were
identified. Absence of other expansive-related products
and features caused by some other mechanisms. Quality
of concrete microscopically is reasonably good.
Medium to High Presence of some or all features is generally consistent with
AAR, such as:
• cracking and microcracking, especially when associated
with known reactive aggregates.
• presence of potentially reactive aggregates.
• internal fracturing of known reactive aggregates.
• darkening of cement paste around aggregate particles,
cracks or voids.
• presence of reaction rims around the internal periphery
of reactive aggregate particles.
• presence of damp patches on core surfaces.
Very High Presence of features such as evidence of sites of expansive
reaction, that is, locations within the concrete where
evidence of reaction and emanation of swelling pressure
can be positively identified, for example, streaming of
ASR gel from a reacted and cracked aggregate particle
into the adjoining cement paste with development of
cracks both in the cement paste and along the paste-
aggregate interface.
Presence of ASR gel in cracks and voids associated with
reactive particles and readily visible to normal or
corrected-to-normal vision or under low magnification.

Table 8.11 C
 lassification of the importance of crack widths (CSA, 2014) and
cracking index values (Fasseu & Michel, 1997)

CSA IFSTTAR

Crack width Cracking

(mm) Description CI importance

< 0.1 Fine Usually present 0–0.5 Negligible

0.1–0.3 Normal Normal limit for RC 0.5–1 Marginal
0.3–0.5 Large Over limit 1–2 Moderate
0.5–1.0 Moderately wide Record all 2–5 High
2.0–1.0 Wide Refer to Engineer 5–10 Very high
>5.0–10.0 Very wide Refer to Engineer > 10 Ultra-high
 Forecasting future performance and managing critical infrastructure 159

combination of both (Chapter 8). Structural evaluation should also be con-

ducted by experts at this phase, taking into consideration both current and
future deterioration based on diagnosis and prognosis, respectively. The
acceptable threshold of induced expansion and associated deterioration
depends on the type and importance of the affected structure; however, it is
quite common to define the “end of service life” of structures affected by
ISR, when its members reach, expansion levels of 0.20%, since such a value
indicates yielding of the rebars and thus potential structural implications. It
is worth noting that this threshold value may be selected on a case-by-case
basis and could be much lower whether serviceability aspects are consid-
ered. Finally, depending on the current damage degree and the potential for
further deterioration, infrastructure owners or engineers in charge of critical
infrastructure may select rehabilitation strategies, which may largely vary
and depend upon the structure type, importance and deterioration degree.
The next section discusses some of the techniques used to mitigate/reduce
deterioration caused by ISR in concrete infrastructure.

8.3.2 Rehabilitation techniques
The efficiency of rehabilitation strategies depends on the ISR type, along with
the current damage and the potential for further deterioration. Moreover, it
is quite clear that the sooner maintenance and rehabilitation strategies are
implemented, the more efficient they are and the less structural implications
they might have. Figure 8.10 associates tasks implemented in ISR-affected

Figure 8.10 C
 lassification of residual expansion potential and its correspondent
intervention strategy.
160 Internal Swelling Reactions in Concrete

concrete structures as a function of their degree of damage (i.e., low, moder-

ate, severe and very severe). The following subsections discuss the strategies
adopted in each of these deterioration levels.

8.3.2.1 Mitigate or inhibit deterioration progress

The strategies employed to mitigate or inhibit the progress of deterioration
in concrete infrastructure usually involve limiting the sustaining conditions
of the main damage mechanism(s) inducing deterioration. As discussed in
Chapter 2, water has been identified as a key factor in ISR, especially in AAR
and DEF cases. Therefore, reducing water in the system can effectively miti-
gate (i.e., decrease the deterioration rate) of affected concrete (De Souza and
Sanchez, 2023). Common strategies in this regard include drainage improve-
ment, coatings and crack injections. However, when inhibiting the deterio-
ration progress is the goal, chemical aspects should also be considered, and
thus specific compounds may be used, such as lithium-based products.

8.3.2.1.1 Drainage system
Structural members that are constantly exposed to the environment and
water, such as highway structures (i.e., abutments, piers, girders, pile caps,
decks, curbs and sidewalks), foundation blocks and dams, require an effi-
cient drainage system. The implementation of a well-designed drainage sys-
tem allows for the reduction of internal humidity, aiming to decrease the
expansion rate of ISR-affected concrete. Therefore, it is crucial to conduct
regular inspections and cleaning of drainage systems along with perform-
ing necessary maintenance to ensure their optimal operation. In the case of
ISR-affected structures, drainage is often combined with cladding, which
provides an additional layer of protection.

8.3.2.1.2 Moisture control by coatings

The implementation of methods that efficiently reduce the internal relative
humidity of concrete is essential to control ISR-induced development. While
surface coatings that prevent moisture ingress are advantageous, it is also
important to consider the moisture within the concrete to mitigate AAR.
Thus, when considering surface coating, the goal is not to limit external
water infiltration but rather to facilitate internal humidity reduction. This
promotes the gradual drying of the concrete and leads to an overall decrease
in internal relative humidity. Coatings, sealers and waterproofs used for
moisture control of ISR-affected concrete should have specific properties,
including resistance to water absorption, the ability to penetrate into mea-
surable depths, durability against ultraviolet (UV) degradation and long-
term stability in alkaline environments.
 Forecasting future performance and managing critical infrastructure 161

The efficiency of products such as silanes, siloxanes, and polyurethanes in

controlling moisture content and minimizing AAR-induced expansion has
been thoroughly appraised and attested in numerous studies (Bérubé et al.,
2002a; Champagne et al., 2019). These products are also recognized for
their ability to enhance frost resistance, reduce chloride ingress and protect
steel reinforcement from corrosion (Deschenes, 2017).

8.3.2.1.3 Crack injection
Crack filling (or injection) is considered in cases where ISR-induced crack-
ing is excessively present on the surface of affected concrete. Thresholds for
crack filling have been established at 0.15 mm and 0.30 mm for reinforced
and non-reinforced concrete members, respectively, while for prestressed
concrete, the limit is lower (i.e., 0.10 mm) (Fournier et al., 2010). Crack
filling involves the use of flexible materials, such as flexible grouts, caulk-
ings and polymers, preventing water and other substances from entering
through the cracks. While rigid polymers and cement grouts can be used to
temporarily stabilize cracks, cracks adjacent to the filled ones may eventu-
ally appear due to the rigid nature of these materials and their strong bond
to the concrete substrate. Crack injection is a symptomatic treatment and
should be used to restore concrete monolithic conditions rather than miti-
gating ISR; it is therefore recommended that crack filling be used as part of
a comprehensive approach to inhibit induced deterioration, such as combin-
ing it with surface coatings or sealers (LCPC, 2003).

8.3.2.1.4 Lithium-based admixtures
Lithium-based admixtures (whether applied topically or introduced into the
concrete through vacuum impregnation or electrochemical chloride removal
process) have been widely used in both new and existing concrete structures
to mitigate ASR-induced expansion and slow down its damage develop-
ment (Folliard et al., 2003; Stokes et al., 1997; E.R. Giannini, et al. 2011).
Among these, lithium nitrate solutions, being pH neutral, easy to handle and
possessing higher penetration rates than lithium hydroxide solutions, have
gained popularity in recent years.

Topical treatment: Lithium may be topically applied to ASR-affected

concrete members; however, previous works of topical lithium ap-
plication have raised concerns about the extent of penetration and
its ability to reduce ASR deterioration. It is important to note that
the extent of concrete distress at the time of the treatment signifi-
cantly influences the potential for lithium ingress. This means that
the sooner the lithium is applied, the better and more efficient it may
be. Results indicate that for concrete cracks with average widths of
0.2 mm or more, lithium penetration can reach depths of about 50
162 Internal Swelling Reactions in Concrete

mm (Ekolu et al., 2017; Kawamura & Kodera, 2005; Kobayashi &

Takagi, 2020; Souza et al., 2017).
Electrochemical impregnation method: Electrochemical methods can
be used to impregnate lithium into concrete. This technique has been
widely employed with various lithium compounds, such as lithium
borate, -nitrate and -hydroxide. While only a limited number of tests
have been conducted to determine the extent of lithium penetration
through electrochemical impregnation, it has been observed that a pen-
etration depth of at least 30 mm is often achieved (De Souza, 2016).
Vacuum impregnation: Vacuum impregnation creates a negative pressure
that enables repair products, including lithium-based admixtures, to
penetrate into materials, filling and interconnecting cracks, voids and
even microcracks. Lithium has been applied via vacuum to treat ASR-
induced deterioration in a wide range of concrete members, such as
abutment walls, sidewalks, parapets and decks (Fournier et al., 2010).

8.3.2.2 Displacement monitoring and loading carry capacity

The ongoing swelling deterioration process in ISR-affected structures can
induce movement, potentially leading to structural instability. In such
a sense, monitoring the structure displacement through devices and sen-
sors such as inclinometers, extensometers, accelerometers and fibre optics
may indicate the intervention time and structure member to be intervened.
Furthermore, monitoring the structure loading carrying capacity with active
solutions (e.g., distributed load balancing, active dumping systems or active
vibration control) can prevent further damage to the structure.

8.3.2.3 Mechanical intervention
When the damage extent reaches a point where it may cause serious struc-
tural implications or when other mitigating systems cannot be applied,
mechanical intervention becomes necessary.

Strengthening: Post-tensioning (using tendons or cables) is currently the pri-

mary structural-level intervention employed to restore the integrity of
ISR-affected infrastructure; it is commonly used for non-massive struc-
tural reinforced concrete members using 1D or 2D encapsulation (CSA,
2000). Besides post-tensioning, the introduction of reinforcement in the
form of straps, steel plates and bolts tensioning has proven to be efficient
in containing ASR-induced expansion (ISE, 1992). Moreover, rock an-
chors and encapsulation methods have been utilized to restrain expan-
sion and movement in mass concrete foundations, including tower blocks
(Villemure et al., 2019).
Stress relief via slot cutting: Stress relief via slot cutting is one of the com-
mon strategies to relieve stresses at selected locations of ISR-affected
 Forecasting future performance and managing critical infrastructure 163

structures and thus isolate further swelling (ISE, 1992). This technique
is widely used in massive structures, particularly in dams (Charlwood &
Solymar, 1995). However, it is important to note that slot cutting may
only provide a “temporary solution” since ISR (e.g., AAR or DEF) are
ongoing deterioration processes that keep evolving; therefore, further
cuts may be necessary over time, increasing the cost of the rehabilitation
process. Slot cutting often causes the “local” increase in expansion rate
before the relief gap is closed (Charlwood & Solymar, 1995). Moreover,
since the cut alters the overall stress state of surrounding areas, additional
reinforcement may be necessary to ensure the stability of concrete ele-
ments during and after slot cutting.

8.3.2.4 Replacement
While the replacement of ISR-affected concrete may be the most efficient
remedial measure, it is rarely economically feasible. A common approach
used in practice is to replace only a small portion of the affected structure
while modifying or reinforcing the vast majority of deteriorated structural
components so that the structure meets acceptable service conditions (Blight
& Ballim, 2000).

REFERENCES

Bérubé, M.-A., Chouinard, D., Pigeon, M., Frenette, J., Rivest, M., & Vézina, D.
(2002a). Effectiveness of sealers in counteracting alkali-silica reaction in high-
way median barriers exposed to wetting and drying, freezing and thawing, and
deicing salt. Canadian Journal of Civil Engineering, 29(2), 329–337. https://doi.
org/10.1139/l02-010
Bérubé, M.-A., Duchesne, J., Dorion, J. F., & Rivest, M. (2002b). Laboratory assess-
ment of alkali contribution by aggregates to concrete and application to con-
crete structures affected by alkali-silica reactivity. Cement and Concrete Research,
32(8), 1215–1227. https://doi.org/10.1016/S0008-8846(02)00766-4
Bérubé, M. A., & Fournier, B. (2005). Outil d’évaluation et de gestion des ouvrages
d’art affectés de réactions alcalis-silice (RAS).
Bérubé, M.-A., Frenette, J., Pedneault, A., & Rivest, M. (2002). Laboratory assess-
ment of the potential rate of ASR expansion of field concrete. Cement, Concrete
and Aggregates, 24(1), 13. https://doi.org/10.1520/CCA10486J
Bérubé, M. A., Pedneault, J. F., & Rivest, M. (1994). Laboratory assessment of
potential for future expansion and deterioration of concrete affected by ASR.
Blight, G. E., & Ballim, Y. (2000). Properties of AAR-affected concrete studied over
20 years. In M. A. Bérubé, B. Fournier, & B. Durand (Eds.), 11th international
conference on Alkali-aggregate reaction (pp. 1109–1118). Centre de recherche
interuniversitaire sur le béton (CRIB).
Brunetaud, X. (2005). Étude de l’influence de différents paramètres et de leurs inter-
actions sur la cinétique de l’amplitude de la réaction sulfatique interne au béton
[PhD]. Châtenay-Malabry, Ecole centrale de Paris.
Canadian Standards Association (CSA). (2000). Guide to the Evaluation and
Management of Concrete Structures Affected by Alkali-Aggregate Reaction.
164 Internal Swelling Reactions in Concrete

Champagne, M., Roy-Tremblay, M., Fournier, F., Duchesne, F., & Bissonnette, B.
(2019). Long-term effectiveness of sealers in counteracting alkali-silica reaction in
highway median barriers exposed to wetting and drying, freezing and thawing,
and de-icing salts. 17th Euroseminar on Microscopy Applied to Building Materials.
Charlwood, R. G., & Solymar, Z. V. (1995). Long-term Management of AAR-
Affected Structures – An International Perspective. AAR in Hydroelectric Plants
and Dams: Proceedings of the 2nd International Conference, 19–55.
CSA. (2000). CSA A864-00 - Guide to the Evaluation and Management of
Concrete Structures Affected by Alkali-Aggregate Reaction. Canadian Standards
Association.
CSA. (2014). A23.1-14/A23.2-14 Concrete materials and methods of concrete con-
struction/Test methods and standard practices for concrete. Canadian Standards
Association (CSA).
CSA A864-00 (R2005). (2005). Guide to the Evaluation and Management of
Concrete Structures Affected by Alkali-Aggregate Reaction. In CSA A864-00
R2005 Constr. Mater. Build. Struct. Build. Concr. Struct (p. 108).
De Grazia, M. T., Goshayeshi, N., Gorga, R., Sanchez, L. F. M., Santos, A. C., &
Souza, D. J. (2021). Comprehensive semi-empirical approach to describe alkali
aggregate reaction (AAR) induced expansion in the laboratory. Journal of Building
Engineering, 40(January). https://doi.org/10.1016/j.jobe.2021.102298
De Souza, D. J., & Sanchez, L. F. M. (2023). Evaluating the efficiency of SCMs
to avoid or mitigate ASR-induced expansion and deterioration through a multi-
level assessment. Cement and Concrete Research, 173, 107262. https://doi.
org/10.1016/j.cemconres.2023.107262
De Souza, L. M. S. (2016). Electrochemical lithium migration to mitigate alkali-silica
reaction in existing concrete structures. Delft University of Technology.
Deschenes, R. A. (2017). Mitigation and Evaluation of Alkali-Silica Reaction
(ASR) and Freezing and Thawing in Concrete Transportation Structures [PhD].
University of Arkansas.
Dunant, C. F., & Scrivener, K. L. (2010). Micro-mechanical modelling of alkali-silica-
reaction-induced degradation using the AMIE framework. Cement and Concrete
Research, 40(4), 517–525. https://doi.org/10.1016/j.cemconres.2009.07.024
Ekolu, S., Rakgosi, G., & Hooton, D. (2017). Long-term mitigating effect of lith-
ium nitrate on delayed ettringite formation and ASR in concrete – Microscopic
analysis. Materials Characterization, 133, 165–175. https://doi.org/10.1016/
j.matchar.2017.09.025
Esposito, R. (2015). The deteriorating impact of Alkali-Silica reaction on concrete:
Expansion and mechanical properties. Delft University of Technology.
Fasseu, P., & Michel, M. (1997). Détermination de l’indice de fissuration d’un pare-
ment de béton; Méthode d’essai LCPC N0. 47.
Folliard, K. J., Barborak, R., Drimalas, T., Du, L., Garber, S., Ideker, J., Ley, T.,
Williams, S., Juenger, M., Fournier, B., & Thomas, M. D. A. (2006). Preventing
ASR/DEF in new concrete: Final report (0-4085-5). Texas Department of
Transportation and the Federal Highway Administration.
Folliard, K.J., Thomas, M.D.A., Ideker, J.H., East, B. & Fournier, B. (2009). Case
Studies Treating ASR-Affected Structures with Lithium Nitrate. Transportation
Research Board Annual Meeting 2009 Paper #09-2685.
Fournier, B., Bérubé, M. A., Folliard, K., & Thomas, M. (2010). Report on the diag-
nosis, prognosis, and mitigation of Alkali-Silica Reaction (ASR) in transportation
structures.
 Forecasting future performance and managing critical infrastructure 165

Gao, X. X., Multon, S., Cyr, M., & Sellier, A. (2011). Optimising an expansion test
for the assessment of alkali-silica reaction in concrete structures. Materials and
Structures, 44(9), 1641–1653. https://doi.org/10.1617/s11527-011-9724-y
Giannini, E.R., Bentivegna, A.F., & Folliard, K.J. (2011). Coatings and overlays for
concrete affected by Alkali-Silica reaction. In V. Mechtcherine, U. Schneck, & M.
Grantham (Eds.), 4th international conference on concrete repair. Dresden, Germany.
Godart, B., Rooij, M., & Wood, J. G. M. (2013). Guide to diagnosis and appraisal
of AAR damage to concrete in structures: Part 1 diagnosis (AAR 6.1) (RILEM).
Springer.
Gorga, R. V., Sanchez, L. F. M., & Martín-Pérez, B. (2018). FE approach to per-
form the condition assessment of a concrete overpass damaged by ASR after 50
years in service. Engineering Structures, 177, 133–146. https://doi.org/10.1016/j.
engstruct.2018.09.043
Haha, M. Ben, Gallucci, E., Guidoum, A., & Scrivener, K. L. (2007). Relation of
expansion due to alkali silica reaction to the degree of reaction measured by SEM
image analysis. Cement and Concrete Research, 37(8), 1206–1214. https://doi.
org/10.1016/j.cemconres.2007.04.016
ISE. (1992). Structural effects of alkali-aggregate reaction: technical guidance on the
appraisal of existing structures. The Institution of Structural Engineers (ISE).
Kawabata, Y., Martin, R.-P., Seignol, J.-F., & Toutlemonde, F. (2016). Modelling of
evolution of transfer properties due to expansion of concrete induced by internal
swelling reaction. 5th International Conference on Alkali Aggregate Reaction.
Kawamura, M., & Kodera, T. (2005). Effects of externally supplied lithium on the
suppression of ASR expansion in mortars. Cement and Concrete Research, 35(3),
494–498. https://doi.org/10.1016/j.cemconres.2004.04.032
Kobayashi, K., & Takagi, Y. (2020). Penetration of pressure-injected lithium nitrite
in concrete and ASR mitigating effect. Cement and Concrete Composites, 114,
103709. https://doi.org/10.1016/j.cemconcomp.2020.103709
Larive, C. (1997). Apports combinés de l’expérimentation et de la modélisation à
la compréhension de l’alcali-réaction et de ses effets mécaniques [PhD]. Ecole
Nationale des Ponts et Chaussées.
LCPC. (2003). Aide à la gestion des ouvrages atteints de réactions de gonflement
interne. Techniques et méthodes des laboratoires des ponts et chaussées, Guide
méthodologique (pp. 1–66). Laboratoire central des ponts et chaussées.
Martin, R. P. (2010). Analyse sur structures modèles des effets mécaniques de la réac-
tion sulfatique interne du béton. Laboratoire central de ponts et chausses (LCPC).
Martin, R. P., Renaud, J. C., & Toutlemonde, F. (2010). Experimental investigations
concerning combined delayed ettringite formation and alkali aggregate reaction.
6th International Conference on Concrete under Severe Conditions CONSEC10.
Martin, R.-P., Sanchez, L., Fournier, B., & Toutlemonde, F. (2017). Evaluation of
different techniques for the diagnosis & prognosis of Internal Swelling Reaction
(ISR) mechanisms in concrete. Construction and Building Materials, 156, 956–
964. https://doi.org/10.1016/j.conbuildmat.2017.09.047
Merz, C., & Leemann, A. (2013). Assessment of the residual expansion potential of
concrete from structures damaged by AAR. Cement and Concrete Research, 52,
182–189. https://doi.org/10.1016/j.cemconres.2013.07.001
Moës, N., Cloirec, M., Cartraud, P., & Remacle, J.-F. (2003). A computational approach
to handle complex microstructure geometries. Computer Methods in Applied
Mechanics and Engineering, 192(28–30), 3163–3177. https://doi.org/10.1016/
S0045-7825(03)00346-3
166 Internal Swelling Reactions in Concrete

Multon, S., Barin, F.-X., Godart, B., & Toutlemonde, F. (2008). Estimation of
the residual expansion of concrete affected by Alkali Silica reaction. Journal of
Materials in Civil Engineering, 20(1), 54–62. https://doi.org/10.1061/(ASCE)
0899-1561(2008)20:1(54)
Multon, S., Sellier, A., & Cyr, M. (2009). Chemo-mechanical modeling for prediction
of alkali silica reaction (ASR) expansion. Cement and Concrete Research, 39(6),
490–500. https://doi.org/10.1016/j.cemconres.2009.03.007
Nguyen, T. N., Sanchez, L. F. M., Li, J., Fournier, B., & Sirivivatnanon, V. (2022).
Correlating alkali-silica reaction (ASR) induced expansion from short-term laboratory
testings to long-term field performance: A semi-empirical model. Cement and Concrete
Composites, 134, 104817. https://doi.org/10.1016/j.cemconcomp.2022.104817
Plusquellec, G., Geiker, M. R., Lindgård, J., Duchesne, J., Fournier, B., & De Weerdt,
K. (2017). Determination of the pH and the free alkali metal content in the pore
solution of concrete: Review and experimental comparison. Cement and Concrete
Research, 96, 13–26. https://doi.org/10.1016/j.cemconres.2017.03.002
Ramlochan, T., Zacarias, P., Thomas, M. D. A., & Hooton, R. D. (2003). The effect
of pozzolans and slag on the expansion of mortars cured at elevated temperature:
Part I: Expansive behaviour. Cement and Concrete Research, 33(6), 807–814.
https://doi.org/10.1016/S0008-8846(02)01066-9
Sanchez, L. F. M., Multon, S., Sellier, A., Cyr, M., Fournier, B., & Jolin, M. (2014).
Comparative study of a chemo-mechanical modeling for alkali silica reaction
(ASR) with experimental evidences. Construction and Building Materials, 72,
301–315. https://doi.org/10.1016/j.conbuildmat.2014.09.007
Saouma, E. V. (2021). Diagnosis & prognosis of AAR-affected structures state-of-
the-art report of the RILEM technical committee 259-ISR. In RILEM State-of-
the-Art Reports. Springer.
Sellier, A., Bourdarot, E., Multon, S., Cyr, M., & Grimal, E. (2009). Combination
of structural monitoring and laboratory tests for assessment of Alkali-aggregate
reaction swelling: Application to gate structure dam. Combination of Structural
Monitoring and Laboratory Tests for Assessment of Alkali-Aggregate Reaction
Swelling: Application to Gate Structure Dam, 281–290.
Souza, L. M. S., Polder, R. B., & Çopuroğlu, O. (2017). Lithium migration in a
two-chamber set-up as treatment against expansion due to alkali-silica reaction.
Construction and Building Materials, 134, 324–335. https://doi.org/10.1016/
j.conbuildmat.2016.12.052
Stokes, D., Wang, H., & Diamond, S. (1997). A lithium-based admixture for
ASR control that does not increase the pore solution pH. Proceedings of the
Fifth CANMET/ACI Inter-national Conference on Superplasticizers and Other
Chemical Admixtures in Concrete: ACI SP-173 (pp. 855–867). Skokie, IL,
American Concrete Institute.
Swamy, R. N. (1991). The Alkali-Silica Reaction in Concrete (R. N. Swamy, Ed.).
CRC Press. https://doi.org/10.4324/9780203036631
Swamy, R. N. (1997). Assessment and rehabilitation of AAR-affected structures.
Cement and Concrete Composites, 19(5–6), 427–440. https://doi.org/10.1016/
S0958-9465(97)00035-8
Villemure, F. A., Fiset, M., Bastien, J., Mitchell, D., & Fournier, B. (2019). Behavior
of epoxy bonded bars in concrete affected by alkali-silica reaction. ACI Material
Journal, 116(6).
Zubaida, N. (2020). Evaluation of the potential of residual expansion of concrete
affected by alkali aggregate reaction [Master]. University of Ottawa.
Chapter 9

Case study
Condition assessment of the Robert-Bourassa/
Charest (RBC) overpass

9.1 INTRODUCTION

The Robert-Bourassa/Charest (RBC) overpass, located in Quebec City,

Canada, was a highway bridge structure constructed in 1966 using an alkali-
silica reactive coarse limestone aggregate. The overpass consisted of a deck rest-
ing on Y-shaped reinforced concrete columns, which in turn were supported by
massive concrete foundation blocks (Figure 9.1) (Fournier et al. 2015). While
there is no specific information available regarding the concrete mix designs,
technical reports indicate that the 28-day design strengths were 24 MPa
for the foundation blocks and 28 MPa for the columns and decks (ICAAR
Visit Report 2000). Over the course of the past three decades, the structural
members of the RBC overpass have exhibited various signs of deterioration.
These signs include extensive steel corrosion and concrete delamination and
spalling at the deck level, along with map cracking, scaling, disaggregation
and pop-outs affecting the foundation blocks due to alkali-silica reaction
(ASR) and freeze and thaw (FT) cycles. Additionally, steel corrosion and
concrete spalling have been observed on the columns and foundation blocks
exposed to saltwater spray from traffic on the Robert-Bourassa highway
(Bérubé et al. 2005). To address the concerns regarding the long-term per-
formance of the Y-shaped columns of the RBC overpass, a series of reha-
bilitation techniques were implemented in the year 2000; these techniques
aimed to mitigate further expansion and damage caused by ASR. The fol-
lowing is a detailed description of the selected products used for this pur-
pose (Sanchez et al. 2020):

• Waterproof polymer-modified sealant (WPMS): This hand-spray-

applied material was used to prevent water leakage and moisture
uptake into the structural member.
• Cementitious coating: A protective coating applied to the concrete sur-
face to prevent moisture uptake and improve aesthetics.
• Silane/siloxane sealants: These water-repellent materials do not cre-
ate a barrier but are effective in minimizing moisture uptake from the
environment.

DOI: 10.1201/9781003188155-9 167

168 Internal Swelling Reactions in Concrete

Figure 9.1 R BC overpass after nearly 50 years in service (Sanchez et al. 2020).

• Glass fibre reinforced polymer (GFRP) sheet: High-performance car-

bon sheets utilized to reinforce the distressed concrete members. They
also serve to prevent water leakage and moisture uptake.
• Epoxy coating: A protective polymeric coating applied to concrete
members to prevent moisture uptake.
• Copolymer-modified coating: This high-build, copolymer-modified
coating is specially formulated to prevent water leakage and moisture
uptake.

Through a series of site inspection surveys, field monitoring and labora-

tory tests conducted on the RBC overpass, a comprehensive assessment of
the structure’s condition was performed. The results of these investigations
served as the basis for the development of a guide for the diagnosis, prog-
nosis and management of ASR-affected structures proposed by Bérubé et
al. and presented in Chapter 8 (Bérubé et al. 2005). The guide included
a wide range of laboratory tests, such as the stiffness damage test (SDT),
Damage Rating Index (DRI), residual expansion and water-soluble alkalis,
among others. While the guide showed promise, it was found that certain
test procedures, particularly the SDT and DRI, were found not to be fully
diagnostic, limiting the efficiency of the proposed protocol. Additionally,
 Case study 169

the visual inspection analyses comparing the performance of the different

rehabilitation products used on the columns did not yield conclusive results.
Therefore, questions were raised regarding the efficiency of these products
in mitigating further expansion and deterioration caused by ASR and FT
cycles (Sanchez et al. 2020).
Before the demolition of the RBC overpass in 2010/2011, several cores
were extracted from the various structural members, including the foun-
dation blocks (FB), treated and untreated columns (C), and the bridge
deck (BD). These cores allowed for a comprehensive multi-level assess-
ment of the overall condition of the structure, following the protocols
outlined in (Sanchez et al. 2014, 2015, 2016a, 2016b, 2017, 2018) and
described in Chapter 7. In addition to the cores’ extraction, a thorough
visual inspection was conducted on all columns, both treated and untreated,
prior to the extraction of the cores.

9.2 CONDITION ASSESSMENT OF RBC OVERPASS

9.2.1 Definition of exposure conditions (microclimate)

It should be noted that the RBC overpass consisted of two parallel bridge
structures: the South and North Bridges. While the overall macroclimate
conditions were considered similar for all members of the overpass, varia-
tions in microclimate conditions were observed. Considering this, the expo-
sure conditions for the different members were defined as follows:

• FB: Distinct locations of the FB were categorized as either exposed (E)

or non-exposed (NE). The E condition referred to locations directly
exposed to weather elements such as wind, rain and splash zones. On
the other hand, the NE condition denoted locations protected by the
BD.
• Columns: The first column at the edge of the deck was designated as a
highly exposed (HE) condition. The second and third columns towards
the centre of the BD were classified as moderately exposed (ME), while
the remaining columns were considered NE. Refer to Figure 9.2 for
further details.

No specific exposure conditions were assigned to the BD itself since it was

completely covered by an asphalt concrete layer.

9.2.2 Visual inspection and crack measurements

A visual inspection was conducted on all Y-shaped columns of the RBC
overpass, including treated and untreated columns in both E and NE con-
ditions. The objective of this inspection was twofold: (a) to compare the
170 Internal Swelling Reactions in Concrete

Figure 9.2 Identification of exposure conditions for columns: (a) north-west

group, (b) south-west group and (c) south-east group (Sanchez et al.
2020).

different treatments applied to the columns and (b) to assess any potential
disparity between surface damage (evaluated through visual inspection) and
internal damage (assessed through microscopic and/or mechanical tests on
extracted cores). Due to significant deterioration, a detailed visual inspec-
tion was deemed unnecessary for the other members (i.e., BD and FB), and
coring was directly proceeded with.
During the visual inspection of RBC columns, a qualitative and semi-
quantitative evaluation of the observed deterioration degree was performed.
 Case study 171

Each column received a qualitative rating based on its overall visual condi-
tion using the following scale (Sanchez et al. 2020):

• 0: Undamaged
• 1: Very minor signs of damage
• 2: Noticeable signs of damage
• 3: Moderate damage
• 4: High damage
• 5: Very high damage

Furthermore, semi-quantitative measurements were conducted to assess the

importance of cracks, specifically crack openings, in each column. Four very
fine cracks and four wide-opened cracks were selected from each column
arm to capture the “damage range” for each column. Table 9.1 provides a
summary of the main results from the visual inspection of the RBC columns.
Figure 9.3 depicts the average maximum crack opening values (average of
the four wide-opened cracks) measured on the columns, while the complete
range of crack openings is presented in Table 9.1. Figure 9.4 showcases the
general condition of selected evaluated columns, such as columns 17, 25,
28 and 29.

9.2.2.1 Untreated columns
The evaluation of untreated columns on the RBC overpass revealed notable
findings. HE columns from both the South and North structures, namely
Y25, Y32, Y41, Y48 and Y17, displayed visual distress ratings ranging from
high to very high, with values between 4 and 5, as shown in Table 9.1.
Similarly, ME columns, including Y18, Y26 and Y42, exhibited significant
visual distress ratings ranging from 3 to 4. Moreover, both HE and ME
columns showed elevated average maximum crack openings, some reaching
1.5 mm–2.0 mm. In contrast, the untreated and NE columns, specifically
Y19, Y20 and Y21, exhibited much less visual distress when compared to
their ME and HE counterparts. The protection provided by their location
shielded them from the detrimental effects of the environment, resulting in
better visual conditions.

9.2.2.2 Treated columns
The visual assessment of the RBC overpass columns revealed interesting
findings. The HE-Y25 column, treated with a waterproof sealant, exhib-
ited a level of visual deterioration comparable to the untreated Y17 col-
umn. Both columns received a distress rating of 4 and displayed an average
maximum crack opening of 1.8 mm. However, Y17 appeared wetter in
appearance when compared to Y25, a characteristic previously reported
by Bérubé et al. (Bérubé et al. 2002) after applying waterproof sealants,
172 Internal Swelling Reactions in Concrete

Table 9.1 Visual inspection results from distinct RBC columns (Sanchez et al. 2020)

Exposure Qualitative Crack opening

Group Column Treatment class damage ratinga (mm)b

South- Y25 Waterproof HE 4 0.60–3.0

West sealant
Y26 – ME 3–4 1.00–4.0
Y27 – ME 3–4 0.60–1.0
Y28 Waterproof NE 2 0.20–1.3
sealant
Y29 Epoxy coating NE 2 0.60–1.5
Y30 Waterproof ME 1 0.10–0.6
sealant +
copolymer-
modified
coating
Y32 – HE 5 0.60–1.0

South- Y41 – HE 5 0.40–2.0

East Y42 – ME 4 0.50–5.0
Y43 – ME 3 0.25–1.0
Y46 – ME 3 0.33–3.0
Y48 – HE 4–5 0.60–2.0
North- Y17 – HE 4 0.40–2.0
West Y18 – ME 4 1.0
Y19 – ME 2 0.80–2.0
Y20 – NE 2 0.40–0.8
Y21 – NE 2 0.40–0.6
Y22 Waterproof ME 1–2 0.00–0.5
sealant +
Cementitious
coating
Y23 Waterproof ME 1 0.00–0.1
sealant +
Silane/siloxane
sealants
Y24 GFRP sheet HE 0–1 0.00
a Definitions according to 4.1.
b Cracks range observed in the distinct columns.

specifically silane-based products, to ASR-affected members. In contrast, NE

columns – namely, Y28 and Y29 – treated with a waterproof sealant and
epoxy coating, respectively, showed some degree of deterioration but to a
lesser extent than the HE-Y25 column. They received a qualitative rating of
2 and exhibited average maximum crack openings of 0.5 mm and 0.9 mm,
 Case study 173

Figure 9.3 M
 aximum crack opening of distinct treated and untreated RBC col-
umns: (a) south-west (left) and south-east (right) groups, (b) north-
west group (Sanchez et al. 2020).

respectively. The observed damage on these NE columns closely resembled

that of the untreated columns (i.e., Y19 and Y21). Moving on to the ME col-
umns, including Y30, Y22 and Y23, they were initially treated with various
products: waterproof sealant, copolymer-modified coating, cementitious
coating and silane/siloxane-based products, respectively. All these columns
displayed a low to very low qualitative visual damage level. Cracks with
very small openings, measuring 0.3 mm, 0.2 mm and 0.1 mm, respectively,
were observed on each of them. Lastly, the HE-Y32 column, which under-
went GFRP wrapping, showed no noticeable surface cracking.
174 Internal Swelling Reactions in Concrete

Figure 9.4 V
 isual aspect of RBC columns: (a) HE untreated Y17 column, (b) HE
waterproof treated Y25 column, (c) NE waterproof treated Y28 col-
umn and (d) NE epoxy coating Y29 column (Sanchez et al. 2020).

9.2.3 Coring
Core samples were carefully extracted from the various members of the
RBC overpass, including the FB, columns (C), and BD of both the South
and North structures. In the case of the BD, the coring process involved the
extraction of three large samples with a diameter of 800 mm (Figure 9.5a).
These large cores were then stored at a controlled temperature of 23 ± 2°C
for 25 weeks. Subsequently, a second coring step was conducted, resulting
in smaller cores with a diameter of 100 mm and a length of 200 mm. These
smaller cores were obtained from three directions: vertical (V), transversal
(T) and longitudinal (L) within each large sample (Figure 9.5b). From each
core, three specimens measuring 100 by 200 mm were obtained per direc-
tion, as shown in Figure 9.6a (dashed lines). It should be noted that the first
inch of each core from both ends was discarded due to surface flaws and
defects resulting from the coring process.
 Case study 175

Figure 9.5 ( a–f) Coring from BD, columns and FB (Sanchez et al. 2020).

Furthermore, ten distinct columns from the South and North structures,
numbered 17, 22, 23, 24, 25, 28, 29, 30, 32 and 42, were evaluated. These
columns underwent different treatments and were exposed to varying condi-
tions (Table 9.1). After the demolition of the RBC overpass, some columns
were saw-cut and stored for approximately two weeks prior to coring (Figure
9.5c and d). Eight core specimens were extracted from each column, and the
samples’ cover was subsequently removed. Therefore, a single specimen per
core was obtained, as depicted in Figure 9.6b; only the “internal” specimens
located within the reinforcement cage of the columns were evaluated.
In the case of the FB, coring was conducted at exposed and NE locations.
Eighteen samples were extracted per location in a semi-circular manner, as
shown in Figure 9.5e and f. Similar to the BD, these samples were saw-cut,
discarding the first inch at each side (Figure 9.6c). Consequently, two speci-
mens measuring 100 by 200 mm were obtained per core per location.
Upon extraction, all 100 mm diameter cores were carefully wrapped in
plastic film to prevent moisture loss and stored at a temperature of 12°C.
This storage condition was implemented to halt any further deterioration of
the cores, as recommended by (Sanchez et al. 2017). Prior to conducting
mechanical testing, the specimens were saw-cut to a length of 200 mm and
then placed in a moist curing room at a temperature of 23 ± 2°C and 100%
relative humidity for 48 hours. The ends of all specimens were ground using
a mechanical grinder to ensure uniformity. The samples for petrography via
the DRI were axially cut into two halves, and one of the resulting flat sur-
faces was polished. A portable hand-polishing device equipped with dia-
mond-impregnated rubber disks was utilized to achieve the desired surface
quality for microscopic assessment. The disks used ranged in coarseness
from no. 50 (coarse) to 1,500, 3,000 (very fine). This approach proved effec-
tive in obtaining surfaces suitable for detailed microscopic examination.
176 Internal Swelling Reactions in Concrete

Figure 9.6 ( a–c) Schematics of cores preparation (cutting and grinding) for test-
ing (Sanchez et al. 2020).

Eight 100 mm by 200 mm samples per member per condition were selected
for assessment of RBC members, as described in Table 9.2.

9.2.4 Laboratory test methods

Following coring, the multi-level assessment as per Sanchez et al. (2017, 2018)
was conducted; it is important to mention that conventional petrography to
 Case study 177

Table 9.2 N
 umber of cores extracted from the different RBC members
(Sanchez et al. 2020)

Condition

Concrete members E (surface) E (core) NE (surface) NE (core)

FB 8 8 8 8

Longitudinal Vertical Transversal

Ta Bb Ta Bb Ta Bb BD
8 8 8 8 8 8

Exposure Class Treatment Samples

Column 17 HE UN 8
T; Waterproof sealant + Cementitious
Column 22 ME 8
coating
T; Waterproof sealant + Silane/siloxane
Column 23 ME 8
sealants
Column 24 HE T; GFRP sheet 8
Column 25 HE T; Waterproof sealant 8
Column 28 NE T; Waterproof sealant 8
Column 29 NE T; Epoxy coating 8
T; Waterproof sealant + copolymer-
Column 30 ME 8
modified coating
Column 32 HE UN 8
Column 42 NE UN 8

a T: Top of the BD;

b B: Bottom of the BD.

detect the cause(s) of deterioration was not performed because ASR and FT
had already been detected in RBC members in previous evaluations as per
Bérubé et al. 2005. Therefore, it has been decided to proceed directly with
the multi-level protocol according to the tests presented hereafter:

• SDT: The SDT was employed to quantify the damage degree of the
distinct cores extracted from the various RBC members. The SDT
involves subjecting concrete specimens to five compressive cycles,
using 40% of the compressive strength of a companion sound con-
crete with similar features and maturity. Chapter 6 provides a com-
plete description of the test procedure. Concrete specimens from each
set of cores, as outlined in Table 9.2, underwent five loading-unloading
cycles at a controlled loading rate of 0.10 MPa/s. Each set consisted of
three concrete cores obtained from the same structural member, with
similar environmental conditions, depth from the surface, direction
and treatment type.
178 Internal Swelling Reactions in Concrete

• DRI: The DRI was also used to assess the damage degree of the
distinct cores extracted from the various RBC members. The DRI
involves microscopically analysing polished concrete sections under
a stereomicroscope with a magnification of 15x–16x. Petrographic
damage features observed are then counted within 1 cm2 grids drawn
on the polished concrete sections. Weighting factors are multiplied by
the feature counts to balance their relative importance towards the
overall deterioration of the material. Finally, the DRI number is com-
puted; the higher the DRI number, the higher the deterioration of the
concrete. Ideally, a surface of at least 200 cm2 should be considered per
concrete member, yet for comparative purposes, the final DRI num-
ber is normalized to 100 cm2. Chapter 5 provides a comprehensive
description of the DRI procedure. The DRI analysis was conducted on
two companion cores (i.e., 100 cm2 each – total of 200 cm2) from each
set of cores extracted from the same structural member, with similar
environmental conditions, depth from the surface, direction and treat-
ment type, as specified in Table 9.2.
• Compressive strength tests: Although the compressive strength of con-
crete is less affected by internal swelling reaction (ISR), particularly
ASR, when compared to properties such as tensile strength and modu-
lus of elasticity (Bérubé et al. 2002; ISE 1992; Nixon & Bollinghaus
1985; Smaoui et al. 2004; Wood & Johnson 1993), compressive
strength tests were performed on the cores extracted from the differ-
ent RBC members. The purpose of these tests was twofold: (a) first,
to determine the ultimate strength capacity of non-damaged or less-
damaged samples from the distinct members, which served as the basis
for the 40% value used in the SDT, and (b) second, to evaluate any
structural implications of ASR-induced damage over time on the RBC
members. Therefore, two initial concrete cores were extracted from
undamaged or less-damaged members/locations from the columns, BD
and FB and compressive strength tests were conducted. Furthermore,
compressive strength tests were also performed on all concrete cores
obtained from the distinct RBC-affected members after stiffness dam-
age testing. The latter has been shown to be valid by Sanchez et al.
(2014).

9.3 RESULTS

9.3.1 Stiffness damage test (SDT)

9.3.1.1 Untreated members
The results of the SDT, including the Stiffness Damage Index (SDI), Plastic
Deformation Index (PDI), modulus of elasticity (E) reductions in percentage
and E values in GPa, are depicted in Figure 9.7. Upon analysing the data,
 Case study 179

Figure 9.7 ( a) SDI, (b) PDI.

180 Internal Swelling Reactions in Concrete

Figure 9.7 (Continued) (c–d) E results (E reduction, % – left; E values, GPa –

right) from RBC untreated members (Sanchez et al.
2020).
 Case study 181

it is evident that the FB specimens exhibited higher SDI and PDI values, as
well as greater E reductions and lower E values when compared to the BD
and C cores. The results also varied depending on the specimen’s condition,
such as the type of environment and depth from the surface. Notably, the
exposed FB cores demonstrated higher SDI and PDI values when compared
to the NE specimens. This observation aligns with the fact that the presence
of high relative humidity or moisture is a prerequisite for ISR development.
Additionally, surface specimens displayed more significant damage than
internal samples. For the FB samples, the SDI values ranged from 0.20 to
0.35, while the PDI results varied from 0.16 to 0.25.
Different levels of damage were observed in the BD cores, depending on
their orientation. Cores oriented vertically (V) and transversally (T) with
respect to traffic exhibited higher damage, as indicated by higher SDI, PDI,
greater E reductions and lower E values. In contrast, longitudinal (L) speci-
mens showed lower levels of deterioration. Additionally, cores taken from
the top of the deck (LT, TT, VT) displayed less damage when compared to
those taken from the bottom (LB, TB, VB). In the case of the columns (C),
the damage degree was lower than that of the FB and BD cores, with SDI
ranging from 0.12 to 0.15 and PDI from 0.08 to 0.12. This can be attributed
to the amount of restraint due to reinforcement, as well as the compression
service loading experienced by these members.
Overall, the FB cores exhibited a very high degree of damage, as classified
by Sanchez et al. (2017, 2018)), while the BD cores displayed moderate dam-
age and C cores a marginal deterioration. It is worth noting that several FB
and BD cores displayed modulus of elasticity values below 20 GPa, which is
significantly low compared to expected values (34–37 GPa, depending on the
member) for sound concrete with similar strength and aggregate type
(Sanchez et al. 2017).

9.3.1.2 Treated and untreated columns

The SDT results obtained from the treated and untreated columns of the
RBC overpass are presented in Figure 9.8. Upon analysis, it can be observed
that the North-West specimens (i.e., 14, 22, 23 and 24) exhibited higher
SDI and PDI values when compared to the South-West (i.e., 25, 28, 29, 30
and 32) and South-East (i.e., 42) cores. Surprisingly, the untreated (UN) and
HE columns (i.e., 17, 32 and 42) did not necessarily display higher levels of
damage, despite the expectation that the presence of high relative humidity
(or moisture) would contribute to ASR-induced development. The SDI val-
ues for the HE untreated columns were 0.15, 0.14 and 0.12, while the PDI
results were 0.12, 0.11 and 0.08, respectively.
Among all the members, columns 22, 23 and 24 exhibited the highest SDI
and PDI values, with values of 0.16, 0.18 and 0.17 for SDI and 0.10, 0.15,
and 0.10 for PDI, respectively. These columns were treated (T) and classified
as ME/HE, according to Table 9.2. Columns 25 and 30 also showed
182 Internal Swelling Reactions in Concrete

Figure 9.8 ( a) SDI, b) PDI.

 Case study 183

Figure 9.8 (Continued) (c-d) E results (E reduction, % – left; E values, GPa – right)

from RBC treated and untreated columns (Sanchez et al.
2020).
184 Internal Swelling Reactions in Concrete

relatively high SDI and PDI values, with values of 0.14, 0.15 and 0.12 for
SDI and 0.13 for PDI. These columns were treated and classified as HE and
ME, respectively. Finally, members 28 and 29 displayed the lowest SDI and
PDI results (except for the SDI value of column 42), with values of 0.13 and
0.12 (SDI) and 0.07 (PDI), respectively. These columns were treated and
categorized as NE.
In general, the damage levels observed in most columns were classified as
marginal to moderate, according to Sanchez et al. (2017, 2018)). Additionally,
the damage levels in the untreated (UN) or treated (T) columns were much
lower compared to the damage observed in the FB and BD cores. The modu-
lus of elasticity values obtained exhibited a strong correlation with the SDI
and PDI values, indicating that higher SDI and PDI values were associated
with lower modulus of elasticity values. The modulus of elasticity (E) values
ranged from 30 GPa (in the case of column 42, which was untreated and
HE) to 17 GPa (in the case of column 23, which was treated and ME). This
represents a stiffness decrease ranging from 20% to 50%, with an average
decrease of approximately 30%. From a structural engineering perspective,
such a decrease in stiffness can be considered significant.

9.3.2 Damage rating index (DRI)

9.3.2.1 Untreated members
Figure 9.9 depicts the microscopic damage features and DRI numbers
observed on cores extracted from the distinct untreated members of RBC
(i.e., FB, BD and C). Upon examination of the plots, it becomes evident that
all specimens exhibited closed cracks within the aggregate particles (CCA).

Figure 9.9 D RI results from RBC untreated members (Sanchez et al. 2020).
 Case study 185

These CCA cracks are likely a result of aggregate production processes

(e.g., crushing, sieving) and weathering and may not be directly associated
with ISR-induced deterioration. However, the presence of open cracks with
and without gel within the aggregate particles (OCA and OCARP, respec-
tively) indicates ISR deterioration, particularly caused by ASR. Similarly,
cracks in the cement paste with and without gel (CCP and CCPRP) are also
attributed to ASR. The DRI numbers obtained align with the SDT results,
as higher values are observed in the FB specimens, followed by BD and C
cores. Once again, the extent of damage varied depending on the condition
of the cores. Exposed FB cores exhibited more damage when compared
to NE specimens, while surface specimens displayed greater damage than
internal samples.
In the case of BD cores, the damage levels varied according to the orienta-
tion of the cores. Vertical (V) cores demonstrated greater damage when
compared to longitudinal (L) and transverse (T) specimens. These findings
slightly differ from the mechanical analyses, where T specimens showed
higher damage than L cores. Additionally, specimens from the top of the BD
displayed lower damage when compared to specimens from the bottom. The
C cores exhibited lower damage degrees than FB and BD cores, which can
be attributed to the amount of restraint due to reinforcement present in
these members.

9.3.2.2 Treated and untreated columns

Figure 9.10 provides a visual representation of the microscopic damage
features and DRI numbers observed on cores extracted from the distinct
treated and untreated columns of RBC. It is important to note that the anal-
ysis focused on samples located within the reinforcement cage, disregard-
ing the columns’ cover, as shown in Figure 9.3b. Upon examination, it is
evident that all specimens exhibited CCA, which, as previously mentioned,
is not necessarily indicative of ISR-induced damage. However, the presence
of opened cracks with and without gel within the aggregate particles (OCA
and OCARP, respectively) and the cement paste (CCP and CCPRP) are indeed
features associated with ISR (i.e., ASR more specifically) and were observed
in all the specimens extracted from various columns. The DRI numbers
obtained for both the treated and untreated columns ranged from 200
to 400, indicating a marginal to moderate level of damage in accordance
with (Sanchez et al. 2017, 2018), which aligns with the mechanical findings
obtained from the SDT analysis.
The DRI numbers obtained from the cores extracted from the North-
West group exhibited a range of 200 to 300. Minimal variations were
observed between individual cores within this group. On the other hand, the
South-West/East columns displayed slightly higher average DRI values, rang-
ing from 200 to 400. Notably, the highly and ME columns (e.g., 25 and 30)
generally exhibited greater damage when compared to the NE columns
186 Internal Swelling Reactions in Concrete

Figure 9.10 D
 RI results from the distinct RBC treated and untreated columns
(Sanchez et al. 2020).

(e.g., 28 and 29). Furthermore, a notable presence of cracks within the

cement paste was observed, particularly in comparison to the North-West
group, suggesting a more pronounced manifestation of ISR-induced devel-
opment in this group of columns. It is plausible that the North-West col-
umns experienced a combination of mechanisms such as ASR and FT,
resulting in cracks within the aggregates and cement paste. Overall, the val-
ues obtained from the different untreated (UN) or treated (T) columns were
lower than the damage levels observed in the FB and BD cores.

9.3.3 Compressive strength results

Figure 9.11 illustrates the obtained compressive strength results of the cores
extracted from the various members of the RBC structure. The compressive
strengths ranged from 20 to 33 MPa for the BD, 20 to 28 MPa for the FB
and 33 to 42 MPa for the different columns. Additionally, it was observed
that the compressive strengths of the BD cores varied depending on their ori-
entation. Specifically, the vertical cores exhibited slightly lower compressive
strength values in comparison to the transverse and longitudinal samples.
Half of the evaluated cores from the FB, specifically four out of eight
cores (50%), exhibited compressive strength values lower than the 24 MPa
design values. Similarly, most of the samples from the BD, with 10 out of 17
samples (58%), displayed compressive strength values below the 28 MPa
design value. On the other hand, all the cores from the C demonstrated
compressive strengths exceeding the 28 MPa design value.
 Case study 187

Figure 9.11 C
 ompressive strength results for the distinct RBC members: (a)
Bridge deck, (b) Foundation block and (c) Columns (Sanchez et al.
2020).
188 Internal Swelling Reactions in Concrete

9.4 DISCUSSION

9.4.1 Multi-level assessment application

A comprehensive approach to appraise ISR-induced damage in concrete was
developed by (Sanchez et al. 2017, 2018), particularly for alkali-aggregate
reaction (AAR), DEF and FT deterioration, single or combined. Through
an extensive analysis of various concrete strengths (25, 35 and 45 MPa),
aggregate types (fine vs coarse) and natures (distinct lithotypes), a chart of
data envelopes was formulated (Table 9.3) with a 95% confidence level.
It is important to note that the chart shows some overlap of data, both
microscopic and mechanical, which can be attributed to the inherent hetero-
geneity of concrete and its constituents, particularly the type and nature of
reactive aggregates (Figure 9.12a – envelopes for 35 MPa concrete illustrat-
ing 12 distinct ASR reactive aggregates). To address this variability, Sanchez
et al. (2017, 2018) suggested two approaches for utilizing the chart. Firstly,
practitioners may select data points from similar reactive aggregates present
in the concrete under examination (e.g., reactive limestone aggregate in this
case). Secondly, if the reactive aggregate is unknown, averaging the enve-
lope values can be considered. It should be noted that Table 9.3 was estab-
lished based on laboratory test samples under free-expansion conditions,
while structures and structural members experience various stress states and
confinement conditions. Nonetheless, this chart serves as a valuable refer-
ence, providing an assessment of the “worst-case scenario” for ISR-affected
concrete.
The strength of employing the multi-level approach lies in its correlation
with ISR-induced expansion, offering engineers responsible for ageing
infrastructure insights into the current state of the chemical reaction and
its potential for further development. Such information is crucial for
selecting appropriate rehabilitation techniques for damaged structural
components. Additionally, it is important to highlight that Table 9.3 spe-
cifically presents values obtained from AAR-affected concrete. Data from

Table 9.3 Multi-level assessment results of AAR-affected concrete (Sanchez et al. 2017)

Assessment of ASR
Classification
of ASR Reference Compressive Tensile
damage expansion Stiffness strength strength
degree (%) level (%) loss (%) loss (%) loss (%) SDI DRI

Negligible 0.00–0.03 - - - 0.06–0.16 100–155

Marginal 0.04 ± 0.01 5–37 (-)10–15 15–60 0.11–0.25 210–400
Moderate 0.11 ± 0.01 20–50 0–20 40–65 0.15–0.31 330–500
High 0.20 ± 0.01 35–60 13–25 0.19–0.32 500–765
45–80
Very high 0.30 ± 0.01 40–67 20–35 0.22–0.36 600–925
 Case study 189

Figure 9.12 M
 ulti-level assessment plots proposed by Sanchez et al. for (a) AAR-
affected 35 MPa mixtures containing 12 distinct reactive aggregates
and (b) AAR-affected 25 and 35 MPa mixtures incorporating a reac-
tive limestone aggregate similar to the aggregate used in RBC mem-
bers (Sanchez et al. 2020).

other ISR mechanisms, such as DEF and FT, either individually or in combi-
nation with AAR, have also been collected and documented in the literature,
as discussed by Sanchez et al. (2018)).

9.4.1.1 Untreated members
Upon examining the microscopic and mechanical results obtained in this
study (Figures 9.7–9.11) and comparing them with the data presented in
190 Internal Swelling Reactions in Concrete

Table 9.3 and Figure 9.12b, which represent an AAR-affected concrete data-
base utilizing a reactive limestone aggregate, it becomes evident that the FB,
BD and C members exhibit varying degrees of damage (Sanchez et al. 2017,
2018, 2020). Specifically, FB is characterized by a very high level of damage,
while BD and C display high and moderate levels of deterioration, respec-
tively. These observations align with the “free-expansion” values, which are
approximately 0.30%, 0.20% and 0.08% for the FB, BD and C, respectively.
Analysis of the DRI results reveals the significant presence of cracks
within the aggregate particles, both with and without gel, indicating ASR as
the primary cause of the ongoing damage process, as anticipated. Moreover,
substantial cracks in the cement paste, partially filled with gel, are observed,
particularly in the FB specimens. This suggests not only high levels of ASR
development but also the likelihood of combined mechanisms such as freez-
ing and thawing.
From an engineering standpoint, the compressive strength values obtained
from the untreated members raise concerns. A notable portion of the FB and
BD specimens exhibit compressive strength results lower than the expected
design values, which brings doubts on the overpass’s ability to withstand
service loads as intended and in accordance with Canadian standards and
safety protocols. Additionally, most of the untreated specimens display very
low stiffness, as indicated by their modulus of elasticity results. This low
stiffness may pose potential issues regarding the structure’s serviceability,
including deflection, displacement or even local deformation of members.
Previous visual inspections have already identified significant deflections in
certain bridge members, which can be partially attributed to the substantial
decrease in the material’s stiffness.
Furthermore, the values obtained from untreated members suggest that
the unrestrained concrete expansion in some elements surpasses the yielding
threshold of the steel reinforcement (0.20%), raising structural concerns.
Additionally, as emphasized by Sanchez et al. (2017)), ASR cracks may
diminish the aggregate interlock effects in damaged concrete due to their
microscopic distress features, such as cracks splitting aggregate particles.
Consequently, the overall shear capacity of affected structural members may
be reduced. Therefore, the proper management of ASR-affected infrastruc-
ture necessitates in-depth evaluations of its potential structural implications.
It is important to note that there is currently a lack of literature data and
research on this subject matter.

9.4.1.2 Treated columns
The expansion levels observed in the distinct RBC columns, which underwent
various rehabilitation procedures, ranged from 0.05% to 0.12%, as depicted
in Figure 9.13 based on Table 9.3 and Figure 9.12b. These expansion values
indicate a marginal to moderate degree of damage, on average. The North-
West columns group exhibited higher levels of expansion compared to the
 Case study 191

Figure 9.13 P
 otential expansion attained for the different RBC columns (Sanchez
et al. 2020).

South-West/East group. The efficiency of the distinct treatment procedures

could not be clearly determined in this study due to the low damage state
of the columns and the lack of knowledge regarding the “0” damage values
prior to treatment. However, it can be observed that, for the majority of the
columns, a more exposed condition resulted in higher expansion levels. For
instance, columns 22, 23 and 24 exhibited greater expansion when com-
pared to columns 28 and 29. Furthermore, when comparing these results
with the previously obtained data from untreated FB and BD members, it is
evident that the deterioration found in the treated and untreated columns is
significantly lower.
Although most of the treated and untreated columns in the North-West
and South-West/East groups can be classified as having marginal to moder-
ate damage according to Table 9.3, the specific damage features observed in
each group of columns provide valuable insights into their distinct distress
processes. This can be readily observed by analysing the DRI bar charts
presented in Figures 9.9 and 9.10. The relatively low number of cracks in
the cement paste (indicated by the checker an initial stage of ASR). In con-
trast, the higher occurrence of cement paste cracks in the South-West/East
group indicates a greater degree of deterioration. However, upon closer
examination of the charts for the South-West/East group, it becomes evident
that although the cement paste cracks are more pronounced on average, the
cracks in the aggregate particles are less prominent. This suggests the pres-
ence of another mechanism, such as FT, contributing to the overall deterio-
ration process of the concrete in combination with ASR.
The evaluation of compressive strength results obtained from the treated
columns reveals that most of these members exhibit compressive values
higher than anticipated. This suggests that their structural capacity is less of
a concern when compared to the FB and BD members. Nevertheless, the
previous evaluations and discussions indicate that the bridge did not comply
192 Internal Swelling Reactions in Concrete

with the current Canadian standards and safety regulations at the time of its
demolition in 2010/2011.

9.4.2 Visual vs multi-level assessment

Visual Inspection is typically employed as the initial non-destructive tech-
nique for inspecting concrete infrastructure. While it is often considered
descriptive, qualitative and somewhat subjective, it plays a crucial role in
decision-making and determining areas that require further assessment. In
this study, both untreated and treated columns of RBC were subjected to
visual inspection and multi-level appraisal, enabling a comparison between
these two techniques. The outcomes obtained from the non-destructive
visual inspection and the destructive microscopic and mechanical assess-
ments are presented in Table 9.4.
Upon analysing the data presented in Table 9.4, it is observed that out of
the ten columns assessed, only two consistent results were found between
visual inspection and multi-level assessments (columns 28 and 29).
Additionally, two types of inconsistencies were identified: visual inspection
yielded higher damage levels than the multi-level assessment in four col-
umns (25, 32, 42, 17), while in four other columns (30, 22, 23, 24), visual
inspection indicated a lower damage degree when compared to the multi-
level appraisal. These results suggest a weak correlation between the surface

Table 9.4 Comparison between visual and multi-level assessments (Sanchez et al. 2020)

Visual inspection Multi-level assessment

Qualita­ Crack
tive opening Damage Expansion Damage Classifi­
Group Column rating (mm) degree (%) degree cation

South- Y25 4 0.60–3.0 High 0.07 Marginal ✕

West Y28 2 0.20–1.3 Marginal 0.06 Marginal ✓
Y29 2 0.60–1.5 Marginal 0.05 Marginal ✓
Y30 1 0.10–0.6 Undamaged 0.09 Moderate ✕
Y32 5 0.60–1.0 Very high 0.08 Marginal/ ✕
Moderate
South- Y42 4 0.50–5.0 High 0.05 Marginal ✕
East
North- Y17 4 0.40–2.0 High 0.08 Moderate ✕
West Y22 1 to 2 0.00–0.5 Minor/ 0.10 Moderate ✕
Marginal
Y23 1 0.00–0.1 Minor 0.11 Moderate ✕
Y24 0 to 1 0 Undamaged/ 0.09 Moderate ✕
minor
 Case study 193

condition and the internal damage degree of ISR-affected concrete. Therefore,

the extent of surface cracking should not be solely relied upon to evaluate
the inner condition and performance of ISR-affected concrete members.
Nonetheless, this does not mean that surface cracking (and thus visual
inspection) should not be considered whenever rehabilitation procedures
are requested, especially when aiming to inhibit further durability-related
issues such as FT and steel corrosion.

9.4.3 Potential structural implications on ISR-

affected columns

9.4.3.1 Stress state of the stirrups

Three-metre sections were taken from two reinforced columns – one from
an exposed site (E) and the other from a NE location to assess the potential
structural implications of ISR on the affected columns, which posed sig-
nificant concerns for the structural engineers overseeing the RBC project,
as depicted in Figure 9.14. It is important to note that the E column had
undergone FRP sheet wrapping in 2000 due to the severity of its deteriora-
tion, as illustrated in Figure 9.14b. In the laboratory, the concrete cover of
the columns was removed at selected locations and the stirrups sanded and
cleaned; then, strain gauges were glued on the stirrups to measure their
stress relief upon cutting. A total of six stirrups were selected for testing at
each column.
The stress relief in terms of micro strains (με) observed after cutting the
instrumented stirrups from both columns is illustrated in Figure 9.15. It is
important to note that the columns had been stored in the laboratory for
several weeks prior to testing, during which no live loads were applied to the
members. Therefore, it can be reasonably assumed that the strains resulting
from stress relief of the stirrups are primarily attributed to ISR-induced
expansion (ASR and FT in this case). The stress-state assessment of the stir-
rups’ cutting was conducted in two ways. Initially, the stirrups were cut at
their edges following the removal of the concrete cover. However, as the stir-
rups were often bent at the edges and sometimes tied to the columns’ ends,

Figure 9.14 R BC columns: (a) NE column and (b) E column.

194 Internal Swelling Reactions in Concrete

Figure 9.15 S tress relief measured on the two columns after stirrups cutting: (a)
E column and (b) NE column. Blue bars represent the stirrups’ edges,
and red bars represent the centre of the cross-section.

analyses were also performed at the centre of the column’s cross-section to

determine if higher and potentially more reliable values could be obtained.
The strain values observed in the stirrups of the NE column were lower
when compared to those of the E column, particularly when the analyses
were conducted at the centre of the cross-section (indicated by the orange
bars). When comparing data from the edge cuttings (blue bars) and centre
cuttings (orange bars), higher values were found at the centre for the E
column, while the edge values were relatively similar for both columns.
This indicates that more accurate and reliable results are obtained from
the centre cuttings of the stirrups. The average strain values obtained from
the centre cuttings of the NE and exposed columns were 1290 and 570
μstrain, respectively. These results, especially for the E column, are consid-
ered quite high, particularly considering that the columns had been stored
in the laboratory before testing, with no live loads applied that could con-
tribute to the stress state of the stirrups at the time of cutting. Therefore,
these values can be directly attributed to the development of ISR and could
have been even higher if the cutting had been evaluated under actual ser-
vice conditions. This emphasizes not only the well-known durability and
serviceability concerns associated with ISR but also the potential struc-
tural implications caused by ISR-induced expansion and damage in con-
crete structures.

REFERENCES

Bérubé, M.-A., Chouinard, D., Pigeon, M., Frenette, J., Rivest, M., & Vézina, D.
(2002). Effectiveness of sealers in counteracting alkali-silica reaction in high-
way median barriers exposed to wetting and drying, freezing and thawing, and
deicing salt. Canadian Journal of Civil Engineering, 29(2), 329–337. https://doi.
org/10.1139/l02-010
 Case study 195

Bérubé, M.-A., Smaoui, N., Fournier, B., Bissonnette, B., & Durand, B. (2005).
Evaluation of the expansion attained to date by concrete affected by alkali-silica
reaction. Part III: Application to existing structures. Canadian Journal of Civil
Engineering, 32(3), 463–479. https://doi.org/10.1139/l04-104
Fournier, B., Sanchez, L., Beauchemin, S., & Au doctorat, candidat. (2015). Outils
d’investigation de la réactivité alcalis-granulats dans les infrastructures en béton
Rapport Final par.
ICAAR Visit Report. (2000). Report of the visit of structures affected by AAR in the
Quebec City area. In 11th International conference on alkali-aggregate reaction,
Québec, Canada.
ISE. (1992). Structural effects of alkali-aggregate reaction: technical guidance on the
appraisal of existing structures. The Institution of Structural Engineers (ISE).
Nixon, P. J., & Bollinghaus, R. (1985). The effect of alkali-aggregate reaction on the
tensile strength of concrete. Durable Construction Material, 2(3), 243–248.
Sanchez, L. F. M., Drimalas, T., Fournier, B., Mitchell, D., & Bastien, J. (2018).
Comprehensive damage assessment in concrete affected by different internal swell-
ing reaction (ISR) mechanisms. Cement and Concrete Research, 107(February),
284–303. https://doi.org/10.1016/j.cemconres.2018.02.017
Sanchez, L. F. M., Fournier, B., Jolin, M., & Bastien, J. (2014). Evaluation of the
stiffness damage test (SDT) as a tool for assessing damage in concrete due to
ASR: Test loading and output responses for concretes incorporating fine or coarse
reactive aggregates. Cement and Concrete Research, 56, 213–229. https://doi.org/
10.1016/j.cemconres.2013.11.003
Sanchez, L. F. M., Fournier, B., Jolin, M., Bastien, J., & Mitchell, D. (2016a). Practical
use of the Stiffness Damage Test (SDT) for assessing damage in concrete infra-
structure affected by alkali-silica reaction. Construction and Building Materials,
125, 1178–1188. https://doi.org/10.1016/j.conbuildmat.2016.08.101
Sanchez, L. F. M., Fournier, B., Jolin, M., & Duchesne, J. (2015). Reliable quan-
tification of AAR damage through assessment of the Damage Rating Index
(DRI). Cement and Concrete Research, 67, 74–92. https://doi.org/10.1016/j.
cemconres.2014.08.002
Sanchez, L. F. M., Fournier, B., Jolin, M., Mitchell, D., & Bastien, J. (2017). Overall
assessment of Alkali-Aggregate Reaction (AAR) in concretes presenting differ-
ent strengths and incorporating a wide range of reactive aggregate types and
natures. Cement and Concrete Research, 93, 17–31. https://doi.org/10.1016/j.
cemconres.2016.12.001
Sanchez, L. F. M., Fournier, B., Mitchell, D., & Bastien, J. (2020). Condition
assessment of an ASR-affected overpass after nearly 50 years in service.
Construction and Building Materials, 236, 117554. https://doi.org/10.1016/j.
conbuildmat.2019.117554
Sanchez, L., Fournier, B., Jolin, M., Bedoya, M. A. B., Bastien, J., & Duchesne, J.
(2016b). Use of damage rating index to quantify alkali-silica reaction damage in
concrete: Fine versus coarse aggregate. ACI Materials Journal, 113(4). https://doi.
org/10.14359/51688983
Smaoui, N., Bérubé, M. A., Fournier, B., & Bissonnette, B. (2004). Influence of speci-
men geometry, orientation of casting plane and mode of concrete consolidation on
expansion due to ASR. Cement, Concrete and Aggregates, 26(2), 58–70.
Wood, G. M., & Johnson, R. A. (1993). The appraisal and maintenance of structures
with alkali-silica reaction. The Institution of Structural Engineers (ISE).
Chapter 10

Conclusions and future works

10.1 CONCLUSIONS, CURRENT CHALLENGES AND

FUTURE OPPORTUNITIES

Internal swelling reactions (ISRs) are extremely complex damage mecha-

nisms leading to induced expansion and deterioration of affected concrete.
If, on the one hand, Chapters 1 and 2 demonstrate the complicated scientific
nuances (yet to be fully understood) of the most common ISR mechanisms
in concrete, they also display, on the other hand, the basic understanding
(at least in it major steps) that engineers and infrastructure owners need to
have while dealing with these unique and “ongoing” mechanisms in con-
crete infrastructure.
Over time and with the increase in technology, a wide range of techniques
and devices aiming to assess the condition of concrete from both “materi-
als” and “structures” scales have been developed; however, most of these
“traditional methods”, although quite promising for evaluating distress
mechanisms coming from the outside to the inside of concrete (e.g., steel
corrosion triggered by carbonation and or chloride penetration, external
sulphate attack, or even static or dynamic loads), do not seem to be suitable
to appraise the extent of internal deterioration caused by ISRs. In this con-
text, Chapter 3 clearly points out the differences between external and inter-
nal damage mechanisms and highlights the need for a more comprehensive
evaluation of the current (i.e., diagnosis: understanding the cause(s) and
extent of deterioration) and future (i.e., prognosis: the possibility of further
deterioration over time) condition whenever concrete structures are affected
by ISRs.
Chapters 4 through 6 present various visual, non-­ destructive testing,
microscopic and mechanical test protocols aiming to contribute towards
ISRs diagnosis in concrete along with the comprehension of their impact on
the engineering properties of the affected material. Although research and
further developments/improvements are never-­ending processes, it is widely
accepted by the ISR scientific community that the current tools to diagnose
and assess the condition of ISR-­affected concrete are suitable for this pur-
pose. On this basis, Chapter 7 demonstrates the use of a novel approach, the

196 DOI: 10.1201/9781003188155-10

 Conclusions and future works 197

so-­called multi-­level assessment protocol, that showed to be quite promising

in assessing the cause and extent of damage of ISR-­ affected concrete.
Chapter 9 displays a successful condition assessment campaign performed
with the use of the multi-­level assessment protocol to appraise a concrete
overpass located in Quebec City, Canada, after nearly 50 years of service.
Otherwise, the same cannot be said for the “prognosis” and “management”
aspects of ISR-­related research, where very little has been developed over the
recent years. Chapter 8 demonstrates quite clearly that the physical and
chemical test procedures developed in the laboratory to assess the potential
for further development of ISRs in concrete are still preliminary and incom-
plete, being either qualitative or partially quantitative, where the use of the
test outcomes to reassess the behaviour of affected structures is yet to be
understood or implemented. Likewise, management protocols are very
descriptive and heavily rely on the experience and expertise of the engineer(s)
conducting the evaluation. Finally, mathematical models, although quite
interesting and promising, require laboratory or field data to be adjusted.
Since the laboratory tools are yet to be developed and most of the structures
do not have consistent monitoring, the suitability of using modelling to fore-
cast behaviour of ISR-­affected concrete is limited. The lack of knowledge on
“prognosis” aspects of ISRs makes the rehabilitation strategies and mainte-
nance protocols adopted in practice purely experience-­related, based upon
past trial-­and-­error experiences.
The previous scenario clearly states the progress made over the last
decades on the understanding of ISRs and their implication in concrete,
along with tools to assess the condition of deteriorated infrastructure.
However, it also emphasizes quite important challenges that engineers and
researchers currently face to cope with ISRs in practice, particularly related
to the potential of further deterioration of the ongoing established mecha-
nisms. Recent developments in new technology, such as the use of high-­
quality sensors and imaging, coupled with artificial intelligence, etc., could
improve the state-­of-­the-­art of “prognosis” and “management” aspects of
ISR research and are nowadays hot topics in the area, bearing new opportu-
nities for further development.
Index

Pages in italics refer to figures and pages in bold refer to tables.

Affected Alkali-­aggregate reaction (AAR), 6, 7,

Affected concrete, 1, 9, 10, 11, 12, 9, 14–16, 19, 23–25, 27, 29,
13, 14, 15, 17, 20, 21, 22, 24, 30, 33, 46, 58, 63–65, 71,
26–30, 32–35, 37, 38, 40, 41, 82–85, 105, 106, 112, 116,
46, 48, 49, 51, 52, 53, 55, 117, 120, 121, 129, 134–138,
57–59, 60, 61–65, 68, 71, 72, 142, 145, 150, 156, 157, 159,
74, 78, 79, 90, 93, 95, 96, 98, 160–165, 168, 169, 171,
101, 103, 104, 108, 111–114, 172, 175, 176, 178–184,
116, 117, 119–122, 124, 125, 188, 189
127, 128, 129, 131, 132, Alkali-­carbonate reaction (ACR), 19,
135, 138, 139–147, 149–153, 20, 23, 25, 26, 29, 45, 77, 130,
155–159, 161, 167, 170, 173, 134, 142, 144, 147, 148, 171
178, 179, 181 Alkali-­silica reaction (ASR), 10, 12, 19,
Affected structure(s), 14–16, 19, 21, 23–30, 34, 38, 44, 45, 46, 58,
22, 30, 52, 57, 58, 74, 83, 108, 60, 61, 64, 72, 73, 77, 78, 79,
114, 121, 157–159, 163, 167, 81, 82, 83, 85, 89, 91–94, 95,
173, 174, 177, 178, 181, 182, 96, 98, 100, 101, 103, 104,
184, 186 105, 106, 113, 117, 119, 120,
Aggregate, 1, 6, 7, 9, 13, 14, 19, 23–29, 122, 123, 127, 128, 130, 131,
32–35, 37, 38, 40, 44–47, 132, 134, 135, 137–140, 142,
48, 57, 59, 63, 65–68, 78, 79, 143–148, 150, 152–156, 160,
82, 84–87, 90, 92–99, 101, 165, 166, 169, 170–172, 175,
103–106, 108, 110–113, 176, 179, 180–187, 188,
117, 120, 122, 124, 127, 195
128, 132, 134, 135, 137–140, ASR gel, 19, 176
142, 144–147, 149–151, 154, Assessment, 1, 3, 10–16, 22, 23, 46,
156, 160, 165, 169–171, 176, 49, 50–52, 54–62, 69, 71, 74,
181–185 76, 78, 79, 82–86, 104, 106,
Aggregate particle(s), 10, 15, 26–29, 113, 124, 137–145, 147, 149,
32–34, 37, 38, 40, 65, 66, 68, 151–153, 155, 156, 163, 167,
90, 92, 93, 96, 97, 101, 104, 173–175, 181–186, 189, 193,
112, 113, 117, 122, 124, 132, 194
134, 135, 139, 144, 147, 149, Multi-­level assessment, 11, 12, 141,
169, 176 143, 145, 147, 149, 151, 153,
Aggregate reactivity, 170 155, 175, 194

198
 Index 199

Cement, 6, 20, 23, 25–29, 31–46, 50, 134, 135, 136–141, 143, 147,
60, 82, 84–86, 88–90, 92–97, 149, 150, 154–156, 161, 169,
99, 101, 103–107, 109, 173, 175, 177–181, 183, 185,
111–115, 117, 124, 132, 134, 186, 188–191, 195
135, 137–139, 144, 145, Damage cause, 10, 58, 86
147–149, 152, 154–156, 162, Damage extent, 108, 119, 120, 121,
168, 169, 176, 179, 181–184 180
Cement paste, 25–29, 31–34, 36–42, Damage feature(s), 9, 10, 15, 27–29,
44, 86, 89, 90, 92–97, 99, 101, 32–34, 37, 63, 65, 66, 71, 92,
103, 104, 111–115, 117, 124, 96, 100, 102, 104, 135, 139,
132, 134, 135, 139, 144, 145, 147, 154, 195
147–149, 152, 162, 168, 169, Damage level(s), 71, 73, 191
176 Damage mechanism(s), 9, 19, 22, 27,
Compressive strength, 11, 12, 17, 20, 31, 33, 42, 112, 116, 178
28, 30, 34, 35, 38, 92, 93–95, Damage rating index (DRI), 10, 12,
109–114, 116, 119, 124, 125, 46, 74, 75, 76, 77, 78, 79, 80,
128, 132, 133, 136, 141, 92–98, 100–102, 104, 106,
144–150, 154, 187, 188, 130, 137, 138, 141, 143, 144,
195 147, 148, 155, 184, 186, 193,
Concrete, 1, 3, 6, 9, 10, 11, 13–16, 195
18–162, 164–168, 170–190, Microscopic damage, 9, 27, 29, 32,
192, 194–195, 197 34, 37, 135, 139, 147, 154
Concrete core(s), 85, 98, 136, 141, Surface damage, 10, 25, 43, 63, 68,
154, 161, 162, 175, 176, 177, 188
195 Delayed ettringite formation (DEF), 1,
Condition, 1, 3, 10, 12, 14, 16, 22, 9, 13, 14, 15, 16, 19, 20, 23,
29, 46, 48, 49, 50–62, 71, 24, 30–35, 38, 41, 45, 47, 57,
76, 82–84, 86, 104, 106, 58, 65, 67, 73, 77–79, 83, 84,
108–110, 119, 121, 137, 139, 91–94, 95, 96, 98, 104, 105,
141, 155–157, 167, 174, 183, 106, 112, 113, 116, 117, 120,
185–187, 189, 193, 194 121, 122, 123, 125, 131, 132,
Condition assessment, 10, 12, 16, 139, 140, 142, 143, 144, 147,
33, 38, 49, 50, 51, 54, 56, 59, 149, 150, 155, 159, 160, 162,
60, 61, 71, 76, 82–84, 86, 156, 172, 178, 180, 182, 183
167, 174, 183, 185, 187 Detection, 53, 55, 60, 71, 83, 84, 85, 92
Crack(s), 12, 22, 26, 29, 32, 33, 37, Deterioration, 1, 10, 11, 14–16, 19–21,
38, 45, 48–50, 52, 55, 62, 63, 22, 24, 26, 29, 33–37, 40, 41,
67–69, 72, 73, 76, 77–79, 82, 45, 48, 49, 52, 53, 55, 57–60,
83, 85, 90, 91, 93–97, 100, 62, 63, 65–69, 74, 79, 83–90,
101, 104, 111, 118, 119, 122, 92, 95, 100–103, 108, 113,
124, 132, 135, 140, 147, 150, 114, 116, 118–120, 122–124,
152–155, 157, 158, 168, 169, 129, 131, 132, 134, 137,
173, 175, 176, 178, 179, 187, 139–142, 144, 145, 147, 149,
189, 190, 191 150, 152, 155–159, 161, 162,
167, 173, 175, 177–182, 185,
Damage, 1, 9, 10, 11, 12, 18, 19, 20, 187–190, 193, 195
22–24, 27–29, 31–40, 41, 42, Deterioration mechanism(s), 16, 48,
43, 46, 47, 48, 53, 55, 57–60, 52, 53, 108, 134, 139
62, 63, 65–69, 71, 73–76, Diagnosis, 13, 16, 20, 23, 57, 82–84,
78, 81–87, 91–94, 96, 97, 103, 104, 140, 154, 155, 157,
100–102, 104–106, 108, 112, 167, 173, 175, 177, 183, 184,
113, 116–121, 125, 131, 132, 186
200 Index

Expansion, 1, 9, 16, 19, 20–22, 24–40, 145, 146, 147, 149, 152, 161,
41, 43, 44, 46, 57, 58, 69, 70, 163, 165, 167, 169, 171, 172,
73, 74, 78, 79, 83, 93, 96, 101, 177, 179, 180, 182, 184
104, 105, 112–114, 116, 117, Internal swelling reactions (ISR), 1, 6,
120, 122, 123, 129, 130–132, 9–17, 19–39, 40, 41–51, 52,
134, 135, 138, 141, 142, 53–64, 66, 68, 70–75, 76,
143–147, 149, 150, 152, 154, 78–80, 82–84, 86, 88, 90,
155, 157, 158, 159, 160–169, 92–94, 95, 96, 97, 98,
171, 172, 175, 177–182, 100–104, 105, 106, 108,
183–187, 191 110–123, 124, 125, 126,
Expansion level(s), 11, 28, 29, 30, 128–151, 152, 153–162,
32–34, 37, 38, 79, 112, 113, 164, 166–170, 172–174,
116, 117, 122, 125, 131, 132, 176–184, 186, 188, 190,
134, 146, 149, 150, 158, 159, 192, 194, 195, 197
177, 188 Internal sulphate attack (ISA), 9, 14, 19,
Expansion over time, 58 24, 30, 39, 40
Induced expansion, 9, 16, 19, 20, 21,
24–26, 28, 29, 31, 34, 35, 37, Management, 1, 11, 15, 17, 20, 21, 59,
39, 43, 44, 46, 57, 58, 70, 73, 60, 61, 82, 90, 157, 173, 174,
74, 101, 104, 117, 123, 129, 182, 186
130, 135, 138, 142, 145–147, Management protocol(s), 11, 20, 21,
149, 152, 161, 163, 165, 167, 156, 157, 173, 174
169, 171, 172, 177, 179, 180, Mechanical properties, 9, 12, 17, 19,
182, 184 20, 24, 26–30, 34, 37–39, 58,
71, 79, 101, 108, 109, 120,
Freeze and thaw (FT), 14, 15, 20, 24, 138, 144, 145, 147, 149, 150,
33, 35–38, 57, 58, 65, 66, 73, 155, 169, 182
77, 91–93, 95, 98, 103, 105, Microscopy, 16, 86, 98, 104, 105, 174,
112, 113, 116, 117, 120, 121, 182
122, 123, 131, 132, 133, 139, Microscopic, 1, 9, 10, 14, 22, 27, 29,
142, 143, 144, 149, 151, 185, 32, 34, 37, 55, 59, 74, 86–91,
187, 195 93, 95, 97, 99, 101, 103, 105,
107, 108, 117, 134, 135, 139,
Induced, 1, 9, 10, 14, 16, 19, 20–22, 140, 141, 147, 154, 156, 157,
24–39, 41, 43, 44, 46, 54, 173, 175, 182, 188, 193
57–61, 63, 65–68, 70, 71, Microscopic damage, 9, 27, 29, 32,
73, 74, 79, 83, 85, 88, 92, 34, 37, 135, 139, 147, 154
93, 101–104, 108, 112, 114, Modelling, 11, 44, 57, 150, 152, 167,
117, 120, 122–125, 129, 130, 173–175, 182, 183
134, 135, 138–147, 149, 150, Multi-­level assessment, 11, 12, 141,
152–156, 158, 161–163, 165, 143, 145, 147, 149, 151, 153,
167–172, 177–180, 182–184, 155, 175, 188, 189, 192, 194
195
Induced damage, 9, 10, 15, 27, 28,
33, 73, 125, 134, 135, 141, Non-­destructive testing (NDT), 10, 14,
154, 195 22, 49, 50, 53–55, 62, 63, 65,
Induced expansion, 9, 16, 19–21, 67, 69, 71–73, 75, 77, 79, 81,
24–26, 28, 29, 31, 34, 35, 37, 83, 85, 86, 174
39, 40, 43, 44, 46, 52, 57, 58,
70, 73, 74, 83, 101, 104, 117, Petrography, 10, 23, 45, 85–90, 94,
123, 129, 130, 135, 138, 142, 104–106, 155, 193, 194
 Index 201

Petrographic analysis, 10, 86, 87, Structural

91, 92 Structural members, 20, 25, 54, 57,
Physical sulphate attack (PSA), 9, 24, 58, 69, 87, 108, 138, 139, 142,
25, 42–44 155, 157, 158, 175, 178, 187
Prognosis, 11, 13, 16, 20, 22, 23, 57, Structural analysis, 55
60, 82–84, 155, 157, 162, 167, Structures
173–175, 177, 183, 184 Affected structures, 14, 15, 16, 19,
Protocol, 1, 11, 14, 57, 58, 84, 140, 21, 22, 30, 52, 57, 58, 74, 83,
141, 142, 152, 161–163, 167, 108, 114, 157–159, 163, 167,
174, 175, 186, 195 173, 174, 178, 181, 182, 184,
186
Reaction, 1, 6, 9, 13, 14, 19, 23–29, Structures, 14, 15, 16, 19, 20–23,
31, 34, 39, 40, 42, 44–46, 57, 25, 29, 30, 32, 44–46, 49,
59, 60, 63, 64, 82–86, 88, 89, 50, 52–54, 57–63, 65, 70,
93–97, 100, 101, 103–107, 71, 74–76, 79, 81–85, 87,
112, 128, 136–140, 142, 144, 105–109, 114, 115, 119,
147, 148, 154–157, 165, 167, 121, 124, 136, 137, 139,
176, 181–185, 195 142, 154, 156–160, 163,
Reaction product(s), 26, 29, 34, 45, 167, 173, 174, 178–184,
86, 88, 89, 93–96, 100, 101, 186, 187, 189, 192, 193
103, 104, 106, 167
Rehabilitation, 1, 11, 16, 20, 50, 52, 54, Tensile strength
87, 90, 173, 174, 177, 180, Tensile strength, 11, 12, 17, 20,
184, 187 29, 30, 32, 34, 35, 38,
108, 109, 111, 112, 124,
Stiffness damage test (SDT), 1, 11, 125, 126, 127, 128, 129,
12, 60, 103, 104, 117, 119, 130, 132, 133, 136, 137,
120–123, 137, 141, 144, 147, 143, 144, 146, 149, 151,
155, 156, 186, 195 155, 168, 188, 195
Specimens, 12, 26, 30, 35, 38, 79, 81, Tensile strength test, 11, 108, 125,
87, 90, 101, 103, 109, 114, 126, 130
116, 119, 120, 121, 125, 127, Thaumasite formation (TF), 9, 24, 41,
129, 131–136, 138, 141, 144, 46, 84
150, 151, 154, 156, 159, 162,
164, 165, 172, 192, 193, 195
Strength, 11, 12, 28, 29, 30, 32, 34–36, Visual inspection, 10, 12, 14, 22, 52,
38, 41, 53, 108–114, 116, 119, 60–63, 65, 67, 69, 71, 73, 75,
124–130, 132, 134–138, 141, 77, 79, 81, 83–86, 156, 172,
143–151, 154, 155, 168, 195 174, 187–190, 192