Anda di halaman 1dari 64

Predicting shear type crack initiation and

growth in concrete with non-linear finite


element method
Richard Malm

April 2009
TRITA-BKN. Bulletin 97, 2009
ISSN 1103-4270,
ISRN KTH/BKN/B97SE

c
Richard
Malm 2009
Royal Institute of Technology (KTH)
Department of Civil and Architectural Engineering
Division of Structural Design and Bridges
Stockholm, Sweden, 2009

The men of experiment are like the ant, they only collect and use; the
reasoners resemble spiders, who make cobwebs out of their own substance. But the bee takes the middle course: it gathers its material from
the flowers of the garden and field, but transforms and digests it by a
power of its own.
Francis Bacon 1561 -1626

Abstract
In this thesis, the possibility to numerically describing the behaviour that signifies
shear type cracking in concrete is studied. Different means for describing cracking
are evaluated where both methods proposed in design codes based on experiments
and advanced finite element analyses with a non-linear material description are
evaluated. It is shown that there is a large difference in the estimation of the crack
width based on the calculation methods in design codes. The large difference occurs
due to several of these methods do not account for shear friction in the crack face.
The finite element method is an important tool for analysing the non-linear behaviour caused by cracking. It is especially of importance when combined with
experimental investigations for evaluating load bearing capacity or establishing the
structural health. It is shown that non-linear continuum material models can successfully be used to accurately describe the shear type cracking in concrete. A
method based on plasticity and damage theory was shown to provide accurate estimations of the behaviour. The methods based on fracture mechanics with or without
inclusion of damage theory, overestimated the stiffness after crack initiation considerably. The rotated crack approach of these methods gave less accurate descriptions
of the crack pattern and underestimated the crack widths.
After verification of the material model, realistic finite element models based on
plasticity and damage theory are developed to analyse the cause for cracking in two
large concrete structures. The Storfinnforsen hydropower buttress dam is evaluated
where the seasonal temperature variation in combination with the water pressure
have resulted in cracking. With the numerical model the cause for cracking can be
explained and the crack pattern found in-situ is accurately simulated. The model
is verified against measurements of variation in crest displacement and crack width
with close agreement. The construction process of a balanced cantilever bridge,
Grndal Bridge, is numerically simulated and a rational explanation of the cause
for cracking is presented. It is shown that large stresses and micro-cracks develop
in the webs during construction, especially after tensioning the continuing tendons
in the bottom flange. Further loads from temperature variation cause cracking in
the webs that is in close agreement with the cracking found in-situ. The effect of
strengthening performed on this bridge is also evaluated where the vertical Dywidag
tendons so far seem to have been successful in stopping further crack propagation.
Key words: Non-linear finite element analysis, concrete, crack width, crack propagation, shear, time-dependant properties, thermal loads, serviceability state,

Sammanfattning
I freliggande avhandling studeras mjligheten att numeriskt beskriva uppsprickning
i betong, orsakat av en kombination av normal- och skjuvspnningar. Olika metoder
fr att beakta uppsprickning utvrderas, dr bde metoder i dimensioneringsnormer
som r baserade p experiment och avancerade finita element analyser med en ickelinjr beskrivning av materialet studeras. Studien visar att det r en stor variation p
den berknade sprickbredden bland de metoder som finns i dimensioneringsnormer.
Denna stora skillnad mellan normerna beror frmst p att flera av dessa metoder
inte beaktar skjuvfriktion mellan sprickytorna.
Finita element metoden r ett viktigt verktyg fr att analysera det icke-linjra beteendet hos betong som orsakas av uppsprickning. Den r srskilt viktig vid kombinering med experimentella underskningar, fr utvrdering av brfrmga eller
fr tillstndsbedmning av en konstruktion. I denna avhandling illustreras att ickelinjra kontinuumbaserade materialmodeller med stor framgng kan anvndas fr
att noggrant beskriva initiering och propagering av skjuvsprickor orsakade av kombination av normal- och skjuvspnningar. Srskilt en metod, baserad p plasticitetsoch skadeteori, gav en mycket god uppskattning av beteendet. De metoder som r
baserade p brottmekanik, med eller utan skadeteori, verskattade styvheten hos materialet efter uppsprickning. Den roterande sprickmetoden i dessa materialmodeller
gav dessutom mindre noggranna uppskattningar av sprickmnster och underskattade
sprickbredder.
Efter att materialmodellen verifierats, har realistiska finita element modeller baserade p plasticitets- och skadeteori utvecklats fr att analysera orsaker till uppsprickning i tv stora betongkonstruktioner. Storfinnforsens kraftverksdamm r en lamelldamm av betong, dr den rliga temperaturvariationen i kombination med vattentryck har orsakat uppsprickning. Med den numeriska berkningsmodellen gr det att
frklara orsaken till sprickorna samt att sprickmnster, variation in krnfrskjutning
och sprickbredd stmmer vl verens med observationer och fltmtningar. Konstruktionsprocessen hos en konsolutbyggd bro, Grndalsbron, har simulerats och
dr en rationell frklaring till den observerade uppsprickningen presenteras. Hga
spnningar och ven microsprickor uppkommer i balkliven till fljd av konstruktionsprocessen, srskilt i samband med efterspnningen av kontinuitetskablarna i
underflnsen. Ytterligare belastning av temperaturvariation orsakar uppsprickning
i balkliven som stmmer vl verens med det observerade. Frstrkningstgrden som genomfrts p bron har ocks utvrderats dr de vertikala Dywidagstagen
verkar, n s lnge, begrnsa fortsatt sprickpropagering.

vii

Preface
The research work presented in this thesis was carried out from January 2004 to
February 2009 at the Department of Civil and Architectural Engineering, Royal
Institute of Technology (KTH). The research have been carried out under supervision
of Prof. Hkan Sundquist and Assoc. Prof. Anders Ansell, and was financed by the
Swedish research council Formas.
First of all, I express my gratitude to my supervisor Hkan Sundquist and my assistant supervisor Anders Ansell for their guidance, support and valuable discussions.
I especially thank Hkan Sundquist for giving me the opportunity to work in this
research area and for providing inspiring research environment. I am very grateful
to my assistant supervisor Anders Ansell for his encouraging support and our long
and fruitful discussions.
A special thanks to the people who have been co-authors or in some other way been
of great assistance regarding the papers in this thesis. Adj. Prof. Bo Westerberg
and Prof. Jonas Holmgren for their valuable advice and discussions regarding design
in the serviceability state and behaviour of concrete. In addition, I thank Adj. Prof.
Costin Pacoste for reviewing the thesis and his valuable advice.
I would also like to thank my colleagues at the Division of Structural Design and
Bridges and the Division of Concrete Structures that have assisted me in the work
with this thesis. I especially thank the PhD students of these divisions, for providing
a friendly atmosphere. I am forever grateful to Andreas Andersson for our interesting
and sometimes endless discussions, regarding our research projects.
Last but not least, my warmest gratitude to my family and friends and especially
to my beloved fiance Grel Lindstrm.

Stockholm, February 2009


Richard Malm

ix

List of Publications
This thesis is based on the work presented in the following publications:
I Malm, R. and Holmgren, J., 2008 Cracking in deep beams owing to shear
loading. Part 1: Experimental study and assessment. Magazine of Concrete
Research, Vol. 60, No. 5, 371379.
II Malm, R. and Holmgren, J., 2008 Cracking in deep beams owing to shear
loading. Part 2: Non-linear analysis. Magazine of Concrete Research, Vol. 60,
No. 5, 381388.
III Malm, R. and Ansell, A., Nonlinear analysis of thermally induced cracking of
a concrete dam, Submitted to ACI Structural Journal, July 2008
IV Malm, R., James, G. and Sundquist, H., 2006. Monitoring and evaluation of
shear crack initiation and propagation in webs of concrete box-girder sections.
In: International Conference on Bridge Engineering - Challenges in the 21st
Century. Hong Kong.
V Malm, R. and Sundquist, H., Time-dependent analyses of segmentally constructed balanced cantilever bridges. Submitted to Engineering Structures,
January 2009
VI Malm, R. and Ansell, A., Non-linear analyses of cracking in segmentally constructed concrete box-girder bridges. Submitted to Engineering Structures,
January 2009
The planing, numerical implementations and analyses as well as the main part of the
writing have been performed by Malm in all the papers. The co-authors have guided
the work and contributed to the papers with comments and revisions. In paper III
the planing and writing of the paper have been done in collaboration with the coauthor. Malm has not been performed the experimental tests presented in paper I,
IV and in section 3.2, but is responsible for the evaluation of all experimental data
presented.
In addition to the publications mentioned above, some additional publications have
been made during the project. These publications are:
Malm, R. 2008. FE Analysis of Cracking in Concrete due to Shear Loading.
In: XXth Symposium on Nordic Concrete Research & Development, Blsta,
June 2008.

xi

Ansell, A. and Malm, R. 2008. Modelling of Thermally Induced Cracking of a


Concrete Buttress Dam. In: XXth Symposium on Nordic Concrete Research
& Development, Blsta, June 2008.
Ansell, A. and Malm, R. 2008. Modelling of Thermally Induced Cracking of
a Concrete Buttress Dam. Nordic Concrete Research, Vol. 38, No. 2, 69-88.
Ansell, A., Ekstrm T., Hassanzadeh M. and Malm R. 2009. Verification of
the cause of the cracks in a Buttress Dam. Accepted for the 23rd ICOLD
CONGRESS, Brazil, May 2009
Malm, R., 2005. Analys av sprickbildning och sprickbredd vid plant spnningstillstnd i balkliv av armerad betong (In Swedish). Tech. report 88,
Royal Institute of Technology (KTH), Stockholm.
Malm, R., 2006. Shear cracks in concrete structures subjected to in-plane
stresses. Lic. Thesis, Royal Institute of Technology (KTH), Stockholm.
Ansell, A., Ekstrm, T., Hassanzadeh, M., Malm, R. 2009. Cracked concrete
buttress dams Part 2. Elforsk Report, Stockholm (in press)

xii

Contents
Abstract

Sammanfattning

vii

Preface

ix

List of Publications

xi

1 Introduction

1.1

Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.2

Outline of the thesis . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.3

Aims and scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.4

Research contribution . . . . . . . . . . . . . . . . . . . . . . . . . . .

2 Numerical models of concrete


2.1

Constitutive models for concrete cracking . . . . . . . . . . . . . . . .

2.1.1

Non-linear fracture mechanics . . . . . . . . . . . . . . . . . .

2.1.2

Non-linear fracture mechanics based on crack bands . . . . . .

2.1.3

Plasticity theory . . . . . . . . . . . . . . . . . . . . . . . . . 10

2.1.4

Damage theory . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2.1.5

Damage-coupled plasticity theory . . . . . . . . . . . . . . . . 14

2.2

Dilatancy in the damaged plasticity model . . . . . . . . . . . . . . . 15

2.3

Numerical instability . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

3 Numerical studies
3.1

23

Deep beams - shear type cracking . . . . . . . . . . . . . . . . . . . . 23

xiii

3.2

Buttress dam - thermal cracking . . . . . . . . . . . . . . . . . . . . . 25

3.3

Bridges - shear and thermal cracking . . . . . . . . . . . . . . . . . . 31

4 Conclusions

35

4.1

General conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

4.2

Further research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

Bibliography

43

Appended Papers

45

Cracking in deep beams owing to shear loading. Part 1: Experimental study


45
and assessment

II Cracking in deep beams owing to shear loading. Part 2: Non-linear analysis 57


III Nonlinear analysis of thermally induced cracking of a concrete dam

67

IV Monitoring and evaluation of shear crack initiation and propagation in


webs of concrete box-girder sections

83

V Time-dependent analyses of segmentally constructed balanced cantilever


bridges

97

VI Non-linear analyses of cracking in segmentally constructed concrete boxgirder bridges


109

xiv

Chapter 1
Introduction
1.1

Background

Limiting the crack width or the reinforcement stress to a predefined value in the
serviceability state is a requirement in most design codes. The proposed methods
in most of these codes are based on experimental data where the cracks are results
from bending or tensile loads. These methods do not thereby include the effect of
internal friction and aggregate interlocking. Inclined cracks, initiating from tensile
stresses and propagating due to a combination of shear and tension, can be significantly affected by aggregate interlocking. Cracking in concrete is different from that
in materials such as metal and glass, characterised by continuous formation and
bridging of micro-cracks that develop into a macro-crack. The formation of microcracks in concrete is macroscopically represented as a softening behaviour, which
causes localisation and redistribution of strain. The softening behaviour also causes
a stiffness degradation, which is evident in cyclic tests. These tests also show that
concrete under tensile loading is characterised by plastic deformations, i.e. permanent deformations. A detailed description of the constitutive behaviour of concrete
should account for stiffness degradation, permanent residual strains and aggregate
interlocking. The latter can be implemented in several different ways, through shear
retention factor, dilation, frictional stresses in the crack plane, etc.
The finite element method (FEM) is an important tool for analysing the non-linear
behaviour caused by cracking. This is of great importance when combined with
experimental investigations to evaluate load bearing capacity or establishing the
structural health. Large concrete structures seldom collapse due to exceeding the
ultimate load capacity. Instead, it is more common that these are not able to satisfy
the requirements in the serviceability state with deflections or crack widths that
may compromise the durability. In this thesis it will be shown that non-linear finite
element analyses successfully can be used to predict these cracks and deflections
with sufficient accuracy.
Cracks were found on two newly built balanced cantilever bridges, Grndal and
Alvik bridges, after two years of service. A photo of the Alvik Bridge is shown
in Figure 1.1. Subsequent inspections showed that the number of cracks increased

CHAPTER 1. INTRODUCTION

as well as the crack widths. This initiated discussion regarding available methods
for estimating crack widths of inclined shear type cracks and their reliability. The
idea of this PhD project was to use the Grndal and Alvik bridges as a reference to
study shear type cracking which hopefully could lead to recommendations of suitable
methods.

Figure 1.1: The balanced cantilever constructed Alvik Bridge. Photo S. Trillkott

1.2

Outline of the thesis

The thesis consists of the introductory part and six appended papers. Most of the
results are presented in the papers. The introductory part summarise the main
aspects of the papers and also contribute with additional information. The theories
that constitute the background to the numerical methods used in the papers are
also presented in the introductory part. Because the introductory part includes
additional information, references not included in the appended papers are given.
The main purpose of the introductory part of this thesis is to describe the role each
paper has for the PhD project. A reader may start with the introductory part for
an overview and after this continue with the papers.
In all papers, the possibility of numerically describing the behaviour that signifies
concrete cracking is studied. In Paper I and II, finite element material models and
simplified calculation methods of deep beams subjected to shear type loading are
compared to experiments. Paper III presents finite element analyses of temperature
induced cracking in a concrete buttress dam. In Paper IV, numerical methods in
several design codes for the crack width estimation are studied. In Paper V and VI,
finite element analyses of the construction process of a balanced cantilever bridge
are presented.

1.3. AIMS AND SCOPE

1.3

Aims and scope

The objective of this PhD project was to increase the understanding of cracking in
concrete structures, where different means for estimating this is analysed. Describing
cracks that propagate in combination of shear and tension is the main focus in the
thesis. These are primarily studied in the serviceability state, but when evaluating
different non-linear numerical methods the analyses are performed up to the load
carrying capacity.
The aim of this work is to study shear type crack initiation and propagation in
concrete structures. In the first part of this thesis the possibility of using finite
element analysis as a tool for predicting shear type cracking is studied. The fundamental theories that form the basis for the constitutive concrete material models are
described. Recommendations regarding some specific material parameters and procedures to avoid convergence difficulties are presented. A selection has been made
to the theories implemented in the commercial finite element programs Atena and
Abaqus, two of the programs used for analyses in this thesis. The purpose of the
numerical studies is to analyse concrete beams from crack initiation and follow the
crack propagation towards shear type failure. The models are used to compare their
numerical response with the experimentally obtained. The reason for this is that
the capability of the finite element programs to describe the shear crack initiation
and propagation must be studied and compared. It is also of importance to study
how shear type cracks are treated in the concrete design codes. Numerical studies
are performed on the Grndal Bridge based on design loads in the serviceability
state, as well as an estimation of the actual crack widths for load conditions that is
probable during the first years of the bridge.
When a reliable modelling method is chosen, the next phase is to develop realistic
finite element models to analyse the cause for cracking in large concrete structures.
The knowledge from performed measurements may act as input but is especially used
for verification of the calculated results. The studied objects are verified against insitu observations of crack patterns and crack widths. One of the structures is also
verified against measured displacements and variations in crack width.
One of the limitations in this thesis is that only shear type cracks are studied. These
are except for flexural or tensile cracks, i.e. cracks that both initiate and propagate
in tension, probably the most common type of cracks. Cracks that initiate and
propagate from direct shear are hence not studied here. Only continuum based
material models are studied in this thesis, as distinct from discrete or cohesive
interface material models. The continuum based models are more computational
effective and cracks can initiate anywhere within the structure, not only at predefined
interface elements. Already available methods for describing the cracking behaviour
are analysed. The thesis focus on the evaluation and application of these methods
and how drawbacks in the methods can be avoided.

CHAPTER 1. INTRODUCTION

1.4

Research contribution

From a scientific point of view, the original feature of this thesis is that non-linear
finite element analysis is shown to estimate crack widths of shear type cracks more
accurately than the commonly used methods in the design codes. The finite element
material model used for evaluation is compared to other methods and verified against
laboratory test. From a practical point of view, the original feature is that the
numerical modelling method is used to analyse to two large concrete structures
where the cause for cracking is evaluated and verified against observations made
in-situ and experimental data.
Different methods for describing cracking in concrete are critically evaluated. When
comparing the methods, it was important that all were created with as similar
material description as possible. In other studies found in the literature only one
modelling approach is usually presented. In their studies, it is not always evident
whether the material properties have been altered to give as good fit to an experiment as possible. The purpose of the numerical results presented in this thesis is
not to receive an exact fit to measured data. Instead, it is of importance that the
material models can predict the behaviour and give a good result even if the material properties are not entirely accurate. The material parameters in all presented
analyses have been kept constant throughout each of the separate projects, even
though a better fit could be received if any of the parameters were changed. For
instance, in Paper II a perfect fit would have been possible to achieve with all the
different programs, but then each program would have used a different description
of the material, i.e. different values of tensile strength, fracture energy, etc.
The same procedure was utilised in the comparison of the calculation methods in
the design codes for crack width estimation. All preconditions have been assumed
equal in the calculations when comparing the methods. These calculations also have
been made with material properties based on each of the studied codes to determine
whether differences in the material definition could result in a different conclusion
regarding which method is most suitable.
When a suitable modelling method was chosen and it was verified that it could describe shear type cracking, cyclic thermal cracking was studied. The Storfinnforsen
hydropower dam, see Figure 1.2, is evaluated where the seasonal temperature variation in combination with the water pressure have led to cracking in the buttress
dam. It was shown that the installation of an insulation wall has resulted in formation of new cracks. It was also shown that an alternative placement of the wall
would have been more suitable and would not have caused these new cracks.
As the modelling method was verified against cracking due to shear and thermal
actions, it was used to analyse the construction of the Grndal Bridge. In this
thesis a rational explanation of the cause for cracking is presented. It is shown that
large stresses and micro-cracks develop in the webs during construction, especially
after tensioning the continuing tendons in the bottom flange. Temperature variation
causes cracking in the webs that is in close agreement with the cracking found insitu. The effect of the strengthening on this bridge is also presented where the

1.4. RESEARCH CONTRIBUTION

vertical Dywidag tendons so far seem to have been successful to stop further crack
propagation.

Figure 1.2: Storfinnforsen hydropower plant in the Faxlven River. Photo from
P. Stenstrm, WSP.

Chapter 2
Numerical models of concrete
2.1

Constitutive models for concrete cracking

The material models used in the finite element analyses are described in the appended papers. This section is intended to be a compliment to the descriptions
presented in the papers, where a wider description of the theories behind the constitutive material models is presented.
The behaviour at failure of concrete differs, depending on if it subjected to low or
high pressures. At low pressures the failure is typically brittle in nature, which is
the case for tensile stresses and compressive stresses at low hydrostatic pressures. If
concrete is subjected to higher hydrostatic pressures, on the other hand, the material
can deform plastically on the failure surface like a ductile material before failure
strains are obtained, Chen (1982). This section is intended for readers familiar with
the non-linear behaviour of concrete and describes the implementations and the
theories used to simulate the non-linear behaviour of concrete. There are several
textbooks and journal papers on this matter and an interested reader is referred
to the following literature regarding the numerical implementations studied in this
thesis for describing concrete mechanics; Malm (2006); Chen (1982); Chen and Han
(1995); Elfgren (1989); Bangash (2001); Jirsek and Baant (2002); de Borst (1986);
Rots (2002); Mang et al. (2003); Karihaloo (2003) and Weihe et al. (1998).
The three main theories for continuum mechanics describing concrete cracking are
fracture mechanics, plasticity theory and damage theory. Combinations of these theories are implemented in the finite element programs used for the analyses presented
in the appended papers.

2.1.1

Non-linear fracture mechanics

Attempts to describe concrete with linear elastic fracture mechanics have been made
since the 1950s. In the early applications of fracture as a constructive phenomenon,
the material was considered to be totally brittle, i.e. a complete loss of stiffness
at failure, Rashid (1968); Cervenka (1970). These models was shown to have nu-

CHAPTER 2. NUMERICAL MODELS OF CONCRETE

merical difficulties and also problems with stress singularity at the crack tip and
was sensitive to the size and direction of the mesh. The first non-linear theory of
fracture mechanics was proposed by Hillerborg et al. (1976). It includes the tension
softening fracture process zone through a fictitious crack in front of the pre-existing
crack whose faces are acted upon by certain closing stresses such that there is no
stress concentration at the tip of this extended crack, as shown in Figure 2.1. The
stress increases from zero at the tip of the pre-existing traction-free macro-crack to
the full strength of the material ft . The proposed model assumes that the fracture
process zone is of negligible thickness and thereof the alternative name discrete crack
model. (Karihaloo, 2003)
Studies of the behaviour of the fracture process zone subjected to combined normal and shear displacements is given by Hassanzadeh (1991a,b); van Mier et al.
(1991). From the experiments it can be seen that essentially the same mechanism
of aggregate interlock occurs between surfaces of micro-cracks and macro-cracks.
s(w)

wc

ft

w
KI = 0

a0

lp
crack length

Figure 2.1: A traction-free crack of length a0 terminating in a fictious crack with


residual stress transfer capacity (w) whose faces close smoothly near
its tip (the critical stress intensity factor KI = 0). Reproduction from
Hillerborg et al. (1976); Karihaloo (2003).

The parameters shown in Figure 2.1 are:


lp
the length of the fracture process zone,
w
inelastic crack opening displacement, and
wc
the crack opening displacement when the crack can be considered as
traction-free.
Two material parameters are essential to describe the material behaviour in the
discrete crack model; the stress-displacement relation (w) in the softening zone
and the fracture energy Gf , which is defined as the area under the tension softening
curve, Karihaloo (2003). The fracture energy defines the amount of energy that is
needed to open a unit area of a crack to obtain a stress free crack. The fracture
energy can be determined according to the test method designed by RILEM 50FMC (1985) for a beam subjected to three-point bending. For ordinary concrete
the fracture energy is normally between 50 and 200 N/m depending on concrete
strength and the size of the ballast, Bangash (2001). The three most common
numerical descriptions of the softening are a linear, a bilinear or an exponential
function. The linear softening behaviour can in most cases provide a solution that

2.1. CONSTITUTIVE MODELS FOR CONCRETE CRACKING

gives accurate results, even though the material response tends to be slightly too stiff.
The stiffness problem with the linear softening reduces with decreased element size,
as shown by Malm (2006). The bilinear description was derived by Hillerborg (1985)
and have been shown for several cases to be sufficiently accurate, e.g. by Gylltoft
(1991). The exponential function was experimentally derived by Cornelissen et al.
(1986) and is considered by to be the best and most accurate description, according
to Karihaloo (2003).

2.1.2

Non-linear fracture mechanics based on crack bands

Micro-cracking and the bridging in the fracture process zone is not continuous and
it does not necessary develop in narrow discrete regions in line with the continuous
traction-free crack. Because of this, it has been argued that the tension softening
relation (w) can equally well be approximated by a strain softening relation (),
i.e. decreasing stress with increasing inelastic strain. However, this strain is related
to the inelastic crack opening displacement w and the fracture energy. This means
that the ultimate strain at rupture, c is related to wc and in other words the strain
is now defined by a fracture criterion. Characterisation of the tension softening
behaviour of concrete in this smeared manner through a strain softening constitutive
relation, was introduced by Baant (1979) and was further developed by Baant and
Oh (1983). To relate the inelastic strain to the crack opening displacement w and
the fracture energy Gf , a width h had to be defined where = w/h. It is assumed
that the micro-cracks in the fracture process zone are distributed over a band of
width h, hence the name crack band model. As the micro-cracks are assumed to
be smeared over an element, the whole element fractures as the tensile strength is
reached. (Karihaloo, 2003)
The total strain in an element is defined as the sum of the elastic strain and the
inelastic strain. There are two approaches for determining the crack direction in the
smeared crack methods. The first method, called the fixed crack model, is obtained if
the smeared cracks in the crack band are normal to the principal tensile stress at the
moment of crack initiation and their direction do not change during the subsequent
growth of the crack band even when the direction of the principal stress changes. In
the second method the crack rotates as the direction of the principal stress changes.
With this approach, the crack is therefore always normal to the principal tensile
stress direction, hence the name rotated crack model. One main difference between
these two approaches is how the shear stresses and shear strains are treated at the
smeared crack bands. In the fixed crack model the need for a shear retention factor
which reduces the shear modulus is needed to avoid convergence difficulties and
to avoid physically unrealistic and distorted crack patterns. In concrete the shear
retention factor allows for the roughness of crack faces due to aggregate interlocking,
etc. The shear modulus is normally reduced with increasing strain, which represents
a reduction of the shear stiffness due to crack opening. In the rotated crack model
no shear stresses can occur in the crack plane since the crack follows the direction of
the principal stress. In this case an implicit shear modulus is calculated to provide
co-axiality between the rotating principal stress and strain. (Rots, 2002)

CHAPTER 2. NUMERICAL MODELS OF CONCRETE

2.1.3

Plasticity theory

Plasticity is a tool for describing ductile material behaviour. It is, however, also
used for finite element analysis of brittle behaviour. The classical theory of plasticity, like any mathematical representation of the mechanical behaviour of solids, may
be viewed in two ways; as a translation of the physical reality or as a model that
approximates the behaviour under certain circumstances. Regarding concrete it has
generally been acknowledged that such prominent features of plasticity theory as
a well-defined yield criterion and a strictly elastic unloading are approximations at
best. Nevertheless, many problems involving brittle materials have been quite successfully treated by means of plasticity theory, according to Lubliner et al. (1989). In
problems in which tension, with the attendant crack development, plays a significant
role - such as shear failure in reinforced concrete structures - the usual procedure is
to apply plasticity theory in the compression zone, and treat the zones in which at
least one principal stress is tensile by one of several versions of fracture mechanics.
(Lubliner et al., 1989)
Yield and failure functions
In biaxial tests on concrete and geomaterials it is usually found that the various
critical surfaces in stress space are similar. Concrete can exhibit a significant volume change when subjected to severe inelastic states. In Figure 2.2 (a) it can be
seen that the increase in volume can be more than twice as large for the hydrostatic compressive stress state, illustrated as 1 /2 = 1/ 1 in Figure 2.2, as
for uniaxial compression 1 /2 = 1/0. In Figure 2.2 (a) the points marked on
the stress-volumetric strain diagrams indicate the limit of elasticity, the point of
inflection in the volumetric strain, the bendover point corresponding to the onset
of instability or localisation of deformation and the ultimate load. The critical surfaces corresponding to these material states are illustrated in Figure 2.2 (b). The
surfaces give an indication of the expansion of the failure surface, i.e. the reserves
of strength that concrete has from the moment its elastic limit is reached until it
completely ruptures. (Karihaloo, 2003)
The same result is not found in triaxial compression tests, at least not for sufficiently
high hydrostatic pressures. This means that, while the yield surface (however defined) is closed, the failure surface is open in the direction of hydrostatic pressure,
Lubliner et al. (1989). Since the critical surfaces are similar in the biaxial behaviour
of concrete, a yield function is used in the plasticity based models. This yield function is initially defined to correspond to the elastic limit and the surface expands
during subsequent loading up to failure. The size of the yield function is based on
the material properties defined for the uniaxial behaviour of concrete. The yield
surface is based on the strengths defined as the material no longer acts elastic and
the failure surface is based on the ultimate strengths. In uniaxial tension, the material is normally defined to be elastic up to the tensile strength. This means that
in the biaxial tensile meridian the yield surface is equal to the failure surface. In
compression the material is usually assumed to be initially elastic up to 30 60 %
of the compressive strength, Chen (1982).

10

2.1. CONSTITUTIVE MODELS FOR CONCRETE CRACKING

(a)

(b)

Figure 2.2: (a) Volumetric strain in biaxial compression, (b) typical loading curves
under biaxial stresses. From Karihaloo (2003).

Several different failure (or yield) criteria have been developed for concrete materials. In Figure 2.3 some common failure surfaces are presented together with
experimental data from Kupfer and Gerstle (1969). For steel usually the von Mises
failure criterion is used, see Figure 2.3 (a). The yield criteria most often used for
concrete materials are the Drucker-Prager and Mohr-Coulomb criteria, see Figure
2.3 (b) and (c), respectively. According to Lubliner et al. (1989), these criteria do
not represent experimental results for concrete very well unless suitably modified.
One modification is for instance to use a combination of the Mohr-Coulomb and
the Drucker-Prager yield functions, where the Drucker-Prager is used for biaxial
compression and the Mohr-Coulomb is used otherwise.
s/fc

s/fc

s/fc

-1

-1

-1

-0.5

-2

-2

-2

-1

-3
-3 -2 -1

(a) von Mises

1
s/fc

-3
-3 -2

1
s/fc
(b) Drucker-Prager
-1

-3
-3 -2

s/fc
0.5
0

1
s/fc
(c) Mohr-Coulomb
-1

Figure 2.3: Failure criteria for biaxial stress state illustrated for plane stress state.
Reproduction from Jirsek and Baant (2002).

11

-1.5
-1.5 -1 -0.5

CHAPTER 2. NUMERICAL MODELS OF CONCRETE

Hardening
The strain tensor in plasticity based models is decomposed into an elastic part and
a plastic part. Plasticity theory permits a description of the dependence of strain in
the material on its history through the introduction of an internal scalar variable,
here denoted as . The internal variable usually describes irreversible material behaviour and its evolution is expressed by means of rate equations which are functions
of the plastic strain rate, = f (p ). In both the work-hardening hypothesis and
the strain-hardening hypothesis the internal parameter, called hardening parameter
is integrated along the loading path to give the scalar hardening variable, . The
hardening rule defines the motion of the subsequent yield surfaces during plastic
loading. The yield condition generalises the concept of yield stress to multi-axial
stress states and includes the history dependence through the scalar hardening variable. Since the yield function is dependent on the loading history through it can
only expand or shrink in the stress space, not translate or rotate. Such hardening
is called isotropic hardening, whether the work-hardening or the strain-hardening
approach is used. The direction of the plastic strain tensor p is determined from
the derivative of the plastic potential function which is illustrated in Figure 2.4.
(Karihaloo, 2003)
Flow rule
As previously mentioned the shape of the yield surface at any given loading condition
can be determined by the hardening rule. The connection between the yield surface
and the stress-strain relationship is determined with a flow rule.
Concrete can exhibit a significant volume change when subjected to severe inelastic
states. This change in volume, usually referred to as dilation, is caused by plastic
distortion and can be reproduced by using an adequate plastic potential function G,
Lubliner et al. (1989). The evolution of the inelastic displacements in the fracture
process zone is defined through the flow rule. The flow rule is defined as
p =

(2.1)

where 0 is a scalar hardening parameter which can vary throughout the straining
process. The gradient of the potential surface G/ defines the direction of the
plastic strain increment vector p , and the hardening parameter determines its
length. (Chen and Han, 1995)
The most simple case is when the plastic potential function coincides with the yield
surface, i.e. has the same shape. The plastic flow develops along the normal to
the yield surface. This is called an associated flow, because the plastic flow is
connected or associated with the yield criterion. The other approach where two
separate functions are used for the plastic flow rule and the yield surface, and where
they do not coincide is called a non-associated flow. In this case the plastic flow
develops along the normal to the plastic flow potential and not to the yield surface.
(Glvez et al., 2002)

12

2.1. CONSTITUTIVE MODELS FOR CONCRETE CRACKING

Hardening
.
ep

_
q

Hyperbolic Drucker-Prager
flow potential
_
p

Figure 2.4: The Drucker-Prager hyperbolic plastic potential function in the meridional plane. From Hibbitt et al. (2007).

An example of a plastic potential function is illustrated in Figure 2.4. This function


is the Drucker-Prager hyperbolic plastic potential function used in the material
model concrete damaged plasticity in the finite element program Abaqus, (Hibbitt
et al., 2007). In the figure hardening through the scalar hardening parameter is
illustrated. This function will be described in more detail in Section 2.2. The
parameters in Figure 2.4 are:
p

2.1.4

the hydrostatic pressure stress, which is a function of the first stress invariant I1 , defined as
p = I1 /3 = (11 + 22 + 33 )/3
the Mises
equivalent effective stress, described as
q

q = 32 S : S = 3J2 where J2 is the second deviatoric stress invariant


2
2
J2 = 11
+ 22
11 22 for biaxial loading
and S is the effective deviatoric stress tensor S = pI +

the dilation angle, measured in the p q plane at high confining pressure.

Damage theory

The progressive evolution of micro-cracks and nucleation and growth of voids are
represented in concrete damage models by a set of variables which alter the elastic
and/or plastic behaviour of concrete at the macroscopic level. In practical implementation, the damage models are very similar to the plasticity theory described in
Section 2.1.3. In all concrete damage models, unloading leaves no residual damage
regardless of the degree of damage induced up to the instant of unloading. This
means that there is no plastic deformation, which is rather unrealistic. For this
reason there are theories that couple the stiffness reduction with the plastic deformation, later described in Section 2.1.5. (Karihaloo, 2003)
In damage models the total stress-strain relation have the following form
= Ds :

13

(2.2)

CHAPTER 2. NUMERICAL MODELS OF CONCRETE

where is the stress tensor, is the strain tensor. The secant stiffness tensor Ds of
the damaged material depends on a number of internal variables which can be tensorial, vectorial or scalar. The expression in Eq. (2.2) differs from classical non-linear
elasticity by a history dependence, which is introduced through a loading-unloading
function F . This function vanishes upon loading and is negative otherwise. It is the
counterpart of the yield function in plasticity theory. For damage growth, F must
remain zero for an infinitesimal period, so that its derivative dF/dt = 0. The theory
is completed by specifying the appropriate material dependent evolution equations
for the internal variables. (de Borst, 2002)
Isotropic damage models
For the case with isotropic damage evolution, the total stress-strain in Eq. (2.2) specialises so that the initial shear modulus and the initial bulk modulus are degraded
with separate scalar damage variables d1 and d2 . A simplification of the isotropic
model can be made by assuming that the degradation of the secant shear stiffness
(1 d1 )G and the secant bulk moduli (1 d2 )K degrade in the same manner during damage growth, i.e. d d1 = d2 . This means that the Poissons ratio of the
material remains unchanged during damage growth and leads to
= (1 d)D0 :

(2.3)

where the damage variable d grows from zero at an undamaged state to one at
complete loss of integrity. The stiffness tensor D0 represent the stiffness of the
undamaged material. (de Borst, 2002)
In case of proportional loading of concrete structures, where cracking often results
from a practically uniaxial tensile stress, isotropic models are generally considered as
sufficiently accurate. It should be noted that the stiffness of the compressive struts
in reinforced concrete members may be considerably underestimated, according to
Mang et al. (2003). According to Grassl and Rempling (2007), isotropic damage
models often fail to describe realistic crack patterns in mixed mode fracture tests.
In these cases, an anisotropic damage model or damage-coupled plasticity is needed.

2.1.5

Damage-coupled plasticity theory

The damage strain described in Section 2.1.4 is not a permanent strain. It is fully
recovered on unloading unlike the equivalent plastic strain. It is known that for
concrete after unloading some permanent strains remain due to sliding and friction
in the micro-cracks. Since failure mechanics of concrete in tension as well as in
compression for low levels of confinement is associated with stiffness degradation
as well as with inelastic deformations, models characterised by a coupling between
damage and plasticity have been developed.
In continuum damage mechanics, the stiffness degradation can be modeled by defining the relationship between stresses and effective stresses. The simplest mode of
coupling between damage and plasticity is a scalar damage elasto-plastic model

14

2.2. DILATANCY IN THE DAMAGED PLASTICITY MODEL

based on the effective stress concept developed by Ju (1989). The stress-strain


equation for the model is
= (1 d)D0 : ( p )

(2.4)

According to the effective stress concept, the plastic yield function is formulated in
terms of effective stress. The effective stress is calculated according to

2.2

1d

(2.5)

Dilatancy in the damaged plasticity model

The constitutive concrete model used in most analyses presented in this thesis, is the
concrete damaged plasticity model in Abaqus. A detailed description of this method
is presented in Paper VI. In this section only some additional information will be
given, aiming at describing a suitable choice for the dilation angle. In this section,
parameter studies for dilation angle in shear type cracking problems and description
of measurements and observations regarding dilatancy are presented. The uniaxial
behaviour of this model is presented in Papers II, III and VI. The biaxial yield
function used in concrete damaged plasticity was developed by Lubliner et al. (1989)
and includes the modifications that later were proposed by Lee and Fenves (1998).
The yield function is a combined geometric shape of two different Drucker-Prager
type functions as illustrated in Figure 2.5. The equation for the yield function is
presented in Eq. (3) in Paper VI.

>

uniaxial tension
s2
1
(q -3 a p + bs2 ) = fc0
1-a

ft0
s1

uniaxial compression
biaxial tension

>

1
(q -3 a p + bs1 )=fc0
1- a
sc0

(f b0 ,f b0)
biaxial compression

1
1-a (q -3 a p ) = fc0

Figure 2.5: Biaxial yield surface in the constitutive model concrete damaged plasticity. From Hibbitt et al. (2007).

The yield criterion is based on an invariant function of the state of stress, i.e. independent of the choice of the coordinate systems. The parameters in Figure 2.5 are:

15

CHAPTER 2. NUMERICAL MODELS OF CONCRETE

a dimensionless coefficient, described as


fb0 fc0
where 0 0.5
=
2fb0 fc0
fc0
the initial uniaxial compressive yield stress.
fb0
the initial equibiaxial compressive yield stress
ft0
the uniaxial tensile stress at failure

a dimensionless coefficient, described as

c (
pc )
=
( 1) ( + 1)

t (
pt )
c (

pc ) is the effective compressive cohesion stress, and


t (

pt ) is the effective tensile cohesion stress.


Dilation
The Drucker-Prager hyperbolic plastic potential function used in concrete damaged
plasticity in Abaqus is illustrated in Figure 2.4 and defined as
p
(2.6)
G = (ft0 tan )2 + q2 p tan
where,


is the eccentricity, which defines the rate at which the plastic potential
function approaches the asymptote. Increasing value of  provides more
curvature to the low potential, and
is the dilation angle, measured in the p q plane at high confining pressure.

The flow in the Drucker-Prager function is non-associative (not identical with the
yield surface) in the meridional plane if the dilation angle and the material friction
angle are different.
The dilation angle is used as a material parameter in Abaqus. It measures the
inclination the plastic potential reaches for high confining pressures. Low values of
the dilation angle will produce brittle behaviour while higher values will produce
more ductile behaviour. A parameter study on a reinforced beam subjected to fourpoint bending presented in Malm et al. (2006) is shown here in Figure 2.6. In the
figure it can be seen that the difference in behaviour is rather small between 20
and 40 and especially between 30 and 40 where the difference is only marginal.
The best agreement with the experimental response was reached for a dilation angle
between 30 and 40 . In the verification of the material model developed by Lee
and Fenves (1998), the dilation is defined as 31 to describe both uniaxial tensile
and compressive failure, while a slightly lower dilation of about 25 for verifying
the biaxial compressive failure with the experimental curves from Kupfer et al.
(1969). Their material properties was assumed equal to ft = 3.5 MPa and fc =
27.6 MPa for the tensile and compressive strength respectively. The initial elastic
modulus was 31.7 MPa in their analyses. Jankowiak and Lodygowski (2005) presents
a study with parameter identification of the dilation angle used in Abaqus. In
their study, it was determined to 38 based on a minimisation of the error of the
biaxial failure envelope of Kupfer et al. (1969) and the yield surface in Abaqus.
The properties of the concrete was experimentally determined to ft = 2.8 MPa

16

2.2. DILATANCY IN THE DAMAGED PLASTICITY MODEL

and fc = 50.0 MPa as tensile and compressive strength, respectively, according to


Jankowiak and Lodygowski (2005). A rather low elastic modulus of E = 19.7 GPa
was determined in their experiments. These studies shows that the dilation angle
in concrete damaged plasticity should be defined between approximately 25 and
40 to describe normal grade concrete to describe both tension and compression in
biaxial stress states. Choosing a dilation in the higher range will result in a better
description for the behaviour for a low degree of confining pressure. In the analyses
presented in the appended papers the dilation angle is chosen within 35 and 38 .
In the next part of this section, measurements and observation of the dilation angle
found in the literature are presented.
80
70

Load [kN]

60
50
40
30
20
10
0
0

3
4
5
Mid deflection [mm]

Experiment
y = b = 56.3
y = 50
y = 40
y = 30
y = 20
y = 10
6
7

Figure 2.6: Parameter study of the dilation angle, from Malm et al. (2006).

In rock mechanics, roughness is often introduced where frictional forces are transmitted along the inclined sliding surfaces of the asperities, Patton (1966). In the
case of a smooth joint, i.e. with no relevant asperity effect, the response is characterised by no dilatancy coupled with constant shear effect. In rough joints the
response depends on the slip direction, forward or backward. In forward slip, the
joint dilates and its shear strength is also higher. On the contrary, in backward slip
the joint contracts and slides at a lower shear force. Puntel and Saouma (2008)
measured the shear response of a sandblasted specimen under confining pressure,
the roughness was laser-scanned and varied only about 4 mm along the crack. In
this case the confining pressure might besides this have smoothen the contact points
and hence resulting in a smooth joint where nearly no dilation was measured. They
measured a dilation angle of about 10 that reduced during subsequent load cycles.
Alejano and Alonso (2005) states that the peak dilation angle at very low stress
levels can be equal to the material friction angle. Increasing degree of confinement
and plastic strain reduces the dilation angle. It is shown for several tests that the
dilation angle is above 50 for a low degree of confinement and axial plastic strain.
The dilation angle reduces to only a few degrees for high plastic axial strain and
confining compressive stress.

17

CHAPTER 2. NUMERICAL MODELS OF CONCRETE

Figure 2.7: Illustration of a crack.

In the analyses presented in the appended papers, the size of ballast will produce
a roughness that is much larger than in the case presented by Puntel and Saouma
(2008). Furthermore, according to Johansson and Stille (2009), a small variation in
the matedness of the crack will have a significant impact on the dilation angle. This
means the better two crack faces fit together, the higher dilation angle. The cracks
studied in this thesis, will have a high degree of matedness, or even be perfectly
mated, i.e. a perfect fit between the two crack faces. This is since the cracks
develop from tensile stresses and propagate from tensile and shear stresses. This
means that the normal stress, i.e. confinement, is small which is also a factor that
indicates that the dilation angle should be much higher than presented by Puntel and
Saouma (2008). According to Johansson and Stille (2009) a perfectly mated crack
in hard rock under low normal stresses could have a dilation angle of approximately
35 , which is similar to the values of the dilation angle presented by Jankowiak and
Lodygowski (2005); Malm et al. (2006). In shear test results made for asperities of
nine granite samples presented by Grasseli (2001) a dilation angle between 29 and
39 was found. These experiments were conducted with a rather low normal stress
of about 1.1 MPa to 2.3 MPa.

2.3

Numerical instability

Some sources of instability include local yield, snap-through, surface wrinkling or


localised material failure. When local instabilities develop, it may not be possible
to obtain a converging solution. At a brittle failure, the diagonal shear type crack
usually initiates rapidly, from nearly invisible just before the peak load and after the
peak load it dominates the appearance. This may result in convergence difficulties
if an implicit solution scheme is used.
For models where there is no reinforcement in significant regions of the model, the
tension softening approach based on a stress-strain relationship will introduce unreasonable mesh sensitivity into the results, according to Hibbitt et al. (2007) and
Cervenka et al. (2005). For these cases it is better to define the fracture energy
or defining the stress and crack opening displacement curve. According to Elfgren
(1989) the element size dependency can be reduced if associated with a regularisa-

18

2.3. NUMERICAL INSTABILITY

tion technique, thereby distributing the dissipated energy over the elements. The
localised crack is distributed over a continuum finite element, provided that the crack
opening w is equal to the element strain times the crack band width h, which for
low order elements often is equal to the element length, according to Rots and Ivernizzi (2004). The size of the crack band is associated with the size, orientation
and integration scheme of the finite element.
Avoiding convergence difficulties
Avoiding convergence difficulties can be done in several ways. The most obvious
method is to define the mesh, boundary conditions and loads to minimise potential
numerical singularities. Another technique is to change the convergence criteria and
allowing a larger number of iterations before convergence is checked. If possible, the
problem might be converted into a quasi-static analysis where an explicit solution
scheme can be used. Point loads and distributed loads can be converted to velocities
of a small magnitude, hence ensuring small kinematic energy content compared to
the strain energy. Regularisation techniques mentioned above, may also help to
improve the convergence. Another way to overcome convergence difficulties is to
introduce artificial damping. This reduces the rate of energy dissipation from one
cracked element to its surroundings.
Most analyses presented in the appended papers are performed with an implicit
solution scheme. The simulations performed in Abaqus had a tendency to interrupt
due to convergence difficulties. The most effective method to continue beyond this,
is in the authors opinion, reducing the tolerances and increasing the number or
iterations performed before the rate of convergence is checked. Typically, the tolerances are reduced to 10 % of the default values and limiting the maximum number of
iterations between two and four times the default values. This results in small time
increments when reaching a point where convergence might be compromised and
thereby increasing the cpu-time. In combination with these measures, a small fraction of visco-plastic regularisation of the constitutive equations was used. A small
fraction of stabilising energy also was included. It is shown in Paper III, Malm and
Ansell (2008), that even with a high content of stabilising energy, the crack pattern
could still be accurately described.
Mesh dependency
The results showed that the order of the element was found to be of no importance.
In the analyses, only finite-strain elements were used and all element types had
passed both force and displacement patch tests, for both membrane and plate bending. Patch test in the finite element method is a simple indicator of the quality of
an element, developed by Irons and Razzaque (1972). The patch test uses a partial
differential equation on a domain consisting from several element set-ups so that the
exact solution is known. Elements pass the patch test, if the finite element solution
is the same as the exact solution. The shell elements presented in the appended
papers are all modeled with general purpose elements that allow transverse shear

19

CHAPTER 2. NUMERICAL MODELS OF CONCRETE

deformation. They use Mindlin thick shell theory as the shell thickness increases and
becomes discrete Kirchhoff thin shell elements as the thickness decreases, i.e. the
transverse shear deformation becomes very small as the shell thickness decreases.
The presented results in Paper III and Papers V - VI are mainly shown for triangular
three node elements (S3) or hexahedral four node elements with reduced integration
(S4R), respectively.
The analyses also showed that the results in Abaqus was not especially dependent on
the element size or element direction. This may be a result of the tension softening
being defined in displacements or fracture energy and that the damage is isotropic
which means that it is damaged in all directions. The difference in results based on
a fine and a coarse mesh presented by Malm (2006) is shown in Figure 2.8.
Experiment
ABAQUS fine mesh, L = 5 mm
ABAQUS coarse mesh, L = 50 mm

60

Load [kN]

50
ABAQUS fine mesh, L = 5 mm

40
30
20
10
0

0.5

1.5

2.5

3.5

ABAQUS coarse mesh, L = 50 mm

Mid deflection [mm]

Figure 2.8: Calculations with a fine and a coarse mesh made with the concrete damaged plasticity material model, from Malm (2006).

In Atena the following failure band approach is utilized to reduce the effect of mesh
sensitivity, both for element size and element orientation for skew meshes. This
approach was also used in Abaqus when calculating the crack widths. The element
length for skew meshes was calculated as
L

= L

= 1 + (max 1)
where

,
45

0 45

(2.7)

is the minimum of the angles 1 and 2 between the direction normal to


the failure plane and the element sides, see Figure 2.9.
0
is the length of the crack band with skew mesh. The length is denoted Lt
0
for tension and Lc for compression.
is a factor accounting for skew mesh, the recommended value of max = 1.5.

The crack width is calculated as the total crack opening displacement within the
crack band
0
(2.8)
w = cr Lt

20

2.3. NUMERICAL INSTABILITY

where
cr

is the crack opening strain, which is equal to the strain normal to the
crack direction in the cracked state after the complete stress release.
4 noded element
y

crack
direction

q2
q1

Lc
Lt

Figure 2.9: Definition of localization (crack) band. From Cervenka et al. (2005).

21

Chapter 3
Numerical studies
Analysing crack initiation and crack propagation is not an easy task even though
commercial finite element codes with highly sophisticated material models for describing the behaviour exist. This PhD project was initiated as a result of analysing
the means for estimating crack initiation and propagation in concrete structures.
After two years of service cracks were found on two newly built balanced cantilever
bridges, called the Grndal and Alvik bridges. These cracks are a result of several
combined effects such as time-dependent material effects, thermal and shear loads.
The problem has been divided into steps where the different causes for cracking
can be studied separately. This is done by starting with analyses of deep beams
where cracking is governed by shear loading. After this, a case study of the inclined cracking due to thermal effects in the Storfinnforsen hydropower dam was
analysed. The last case study is the Grndal Bridge where the experience from the
previous case studies constitutes the starting point, where the combined effects from
time-dependent material effects, thermal loads and shear loads are studied.

3.1

Deep beams - shear type cracking

The first two numerical studies of deep beams are presented in Papers I and II,
(Malm and Holmgren, 2008a,b). The experimentally obtained crack pattern is illustrated in Figure 3.1 for one of the studied beams. In Paper I the laboratory
experiments are presented along with calculations based on truss and strut-and-tie
models. The most accurate method, developed by Zhang and Tan (2007), gave an
average calculated ultimate load for the ten beams equal to 0.9 of the experimental load. The accuracy of the different methods and for the different beams varied
considerably. Almost all estimations were conservative and sometimes very conservative. It was interesting to note that the EC 2 (2004) model gave the best result for
the beams with higher reinforcement ratio and therefore resulted in a shear compressive failure. The Zhang and Tan (2007) method, however, gave the best estimations
of the beams with the lower amount of web reinforcement which therefore resulted
in the diagonal tensile failure.
Paper II presents non-linear numerical analyses of the experiments. Two different

23

CHAPTER 3. NUMERICAL STUDIES

finite element programs Abaqus, (Hibbitt et al., 2007), and Atena, (Cervenka et al.,
2005), and one program based on sectional forces called Response, (Bentz, 2000),
are studied. These programs are chosen both due to accessibility and since these
are some of the most respected and widely used for concrete analyses. The Abaqus
model is primarily based on plasticity and damage theory, but it also includes the
fracture process zone from fracture mechanics. The theories behind the concrete
damaged plasticity material model are presented in the following sections 2.1.3, 2.1.4
and the combination of these theories in 2.1.5. The Atena models are based on
fracture mechanics and damage theory for the tensile behaviour of concrete where
the direction of a crack can be either fixed or rotating in respect to a changing
stress state. The theories behind the material model SBETA in Atena are presented
in the following sections 2.1.2 and 2.1.4. The Response model is based on the
modified compression field theory, Vecchio and Collins (1986), and is similar to the
rotating crack model in Atena. The purpose with these calculations was to study
the difference in calculation methods. Therefore, all material data was described as
equal as possible in all analyses.

300

B5
375

300
225
(190)

225

300
(260)

300

300

375

300
225

300

300

300

300 300

225

225

225

300

300

300

Figure 3.1: Illustration of a studied deep beam and the experimentally obtained crack
pattern.

The analyses showed that one of the strut-and-tie models estimated the ultimate
load, almost as accurately as the more complicated non-linear numerical analyses.
These deep beams may be considered as rather straightforward, hence it is not surprising that a well-defined strut-and-tie model could be sufficient to receive accurate
estimations of the ultimate load. The analyses with the different numerical programs
all showed good agreement with the measured ultimate load. One exception was
the beams without web reinforcement that could not be analysed accurately with
Response, where the analyses ended prematurely. Cracking was shown to be more
complicated to accurately describe numerically with some methods and the deflections too, as shown in Paper II. The material models based on fracture mechanics
gave too stiff behaviour while the model based on plasticity instead gave slightly too
soft behaviour. Crack pattern and crack widths could not be successfully captured
with the rotating crack models. The fixed crack approach gave a much better representation of crack pattern and crack widths. The overall behaviour with ultimate
load, deflection, crack pattern and crack widths was best described with the damaged plasticity model in Abaqus. This is the reason all following analyses are made
with this material model.

24

3.2. BUTTRESS DAM - THERMAL CRACKING

These two papers contribute to the thesis through the verification of the calculation
method used in future papers compared to experiments, other non-linear numerical
methods and simplified design calculation methods. The results in these papers
show that the studied smeared crack models, i.e. based on fracture mechanics, with
either rotating or fixed crack models are not to prefer, due to the over stiff behaviour.
The rotated models also gave a less accurate description of the crack pattern and
especially crack widths. The rotating crack models, missed the shear type failure in
some cases, because the crack direction rotated. With the rotating model in Atena,
the estimation of crack width could be corrected, if the user manually added the
calculated crack width from all adjacent cracked elements that described the same
physical crack.

3.2

Buttress dam - thermal cracking

The presented analyses in Paper III, Malm and Ansell (2008), were made as part
of an Elforsk1 project to explain and numerically describe the crack pattern found
in the Storfinnforsen buttress dam. A photo of the storfinnforsen hydropower dam
is shown in Figure 3.2. The results should be able to describe the development
during 60 years of the service life of the dam to describe the crack pattern found
in-situ. This has numerically been done covering five years of thermal steady state
calculations through changing from summer to winter conditions respectively.

Figure 3.2: The monoliths of Storfinnforsen hydropower plant in the Faxlven River.
Photo from T. Ekstrm.

It is shown that seasonal thermal variation initiates cracks in the same region that
was found in the late 1980s cracks on the Storfinnforsen dam. A stabilised crack
pattern, i.e. no new cracks are initiated or propagates, was reached after two years in
the finite element (FE) analyses. At this point, the front-plate and the buttress wall
1

A Swedish energy research and development organisation

25

CHAPTER 3. NUMERICAL STUDIES

of the dam had extensive cracking. When these cracks were found on the real dam,
an insulation wall was installed to reduce the temperature gradient over the frontplate. This was also included in the FE analysis and it is shown that this insulation
wall is the cause of several new cracks. The final crack pattern that occurs in the
FE analysis after years of cycling the temperature corresponds with the observed
in-situ. The use of three-dimensional shell models proved to very accurate although
rather computationally demanding with a CPU time of approximately 7.5 hours for
the seasonal analysis.
The contribution of this paper to the thesis, is a verification that the modelling
method accurately can describe crack initiation and propagation due to thermal
cyclic loads. The results presented in this paper is only verifying the crack pattern.
Hence, in the next section calculated behaviour is compared to measurements of
variation in crest movement and crack width.
Verification of the FE model
This analysis starts with the fully cracked dam presented in Paper III where a
detailed simulation of a year with transient heat transfer analysis is simulated. This
is done with a restart analysis of the previous model presented in the paper where
a detailed transient temperature variation is introduced as predefined fields in the
cracking model. The purpose is to verify the numerical model with the in-situ
measurements of the variation in crest movement and the crack widths.
The instrumentation and measurements are performed by the company EnergoRetea
AB and included as a separate part of the Elforsk project. To verify the finite element model the measured data have been given to the author. The measurements
are presented in mainly internal reports, made to the dam owners E.ON Vattenkraft
Sverige AB. The measurements performed on the dam consists of temperature measurements, pendulum measurements of the crest displacement and variation in crack
width measured with linear variable differential transformer (LVDT) sensors. The
two largest monoliths, monolith 42 and 43, are monitored where the FE-model is
most similar to monolith 42. A description of the measurement setup and results is
given by Ansell et al. (2007).
The measured temperature signal is shown in Figure 3.3 together with a smoothen
curve with averaged weekly variation. Measurements of the outside air temperatures
at Storfinnforsen hydropower dam are unfortunately not complete, since for some
periods of time the thermal gauges failed to record the temperature. The time
period between 2008-01-26 and 2008-04-08 (ten weeks) is the longest period without
available data. In addition to this period, the measurements are incomplete between
2007-12-13 and 2008-01-20 (5 weeks). The temperature measurements are sampled
six times each day. The dam seems to be mainly dependent on the seasonal variation,
hence the measured temperatures are smoothened to weekly average values.

26

3.2. BUTTRESS DAM - THERMAL CRACKING

30
25

Measured temperature
Smoothen temperature

Temperature (C)

20
15
10
5
0
-5
-10
-15
11/01 12/01 01/01 02/01 03/01 04/01 05/01 06/01 07/01 08/01 09/01 10/01 11/01

2007

Date

2008

Figure 3.3: Temperature measured at Storfinnforsen and the smoothen weekly average temperature.

To receive complete temperature signals, the measured temperatures were compared with the temperatures for the same period at stersund, obtained from an
online temperature database (http://www.temperatur.nu, 2009). The dam is located close to stersund, within a distance of approximately 100 km. As shown in
Figure 3.4 the temperatures are very similar for these two regions, hence including
the temperatures measured in stersund for the parts where the measured signal
at Storfinnforsen is incomplete will give sufficient accuracy. The temperature signal measured at Storfinnforsen contains less scatter and is slightly lower than the
measured temperature at stersund, mainly an effect due to averaging.

Figure 3.4: Temperature from stersund compared with temperature measured at


Storfinnforsen.

27

CHAPTER 3. NUMERICAL STUDIES

Local peaks where sorted out from the smoothen data, in order for the FE input to
include as much of the variation as possible, when decimating the signal one value
per week. The finite element calculation would take an unreasonable amount of
cpu-time if more than one calculation per week was made. The temperature input
data in the FE analysis is shown in the upper graph of Figure 3.7. From studying
the measured signals, it looks like a sample frequency of once per week seems to be
sufficient to capture the main behaviour. Daily effects are not visible in the variation
of the crack width. Besides this, the crest movement was measured less than twice
per month.
Along with the outside temperature, the temperature within the thermally isolated
area is also measured. This signal is also incomplete for the time periods mentioned
earlier. Hence, some procedure has to be done to create a complete signal. In
this case, the temperature within the isolated area seems rather similar from one
year to another. Therefore, the indoor temperature for the missing parts have been
taken from the previous years of measurements. The indoor temperature from the
previous year shows that the temperature varies about 2 C during December to
April. The temperature changes slowly and hence a possible error in the input data
to the FE analysis may not affect the results. The measured indoor temperature for
the specific time period with incomplete measurements is shown in Figure 3.5.
14
13

Temperature (C)

12

2006 - 2007
2007 - 2008
Combination of the two years

11
10
9
8
7
6
5
12/01

01/01

02/01

03/01

04/01

05/01

06/01

Date
14

Temperature ( C)

2006 - measured
2007
Figure 3.5: Temperature
in the isolated area 2006-2007 and 2007-2008 re2007In- 2008
13 spectively.
addition to this, the combined temperature curve used in
Combination of the two years
the
FE
analysis
is shown.
12
11

Temperatures that also affect the behaviour of the dam are the water temperature
10
and the temperature of the bedrock. These temperatures have been described in
Bjrnstrm et al. (2006); Ansell et al. (2007) where both can be
accordance with
9
assumed to follow a sine curve varying from +2.5 C to +12.5 C and 2.5 C to
8
+12.5 C for the surface water and the bedrock respectively. The surface water is
7 water in the top 10 metres from the crest. The water below this is
defined as the
assumed have a constant temperature of +4 C.
6

5
12/01

01/01

02/01

03/01

28
Date

04/01

05/01

06/01

3.2. BUTTRESS DAM - THERMAL CRACKING

In Figure 3.6 the fully cracked dam presented in Paper III, is the initial state for
the analysis presented in this section. The variation in crack width is illustrated for
the inclined crack that originates from the inspection gangway, the crack referred to
as crack 3 in Paper III. The measured variation in crack width is measured with a
LVDT that is 0.4 m long and hence indirect also measuring the variation in strain in
the un-cracked concrete near the crack. The LVDT is placed at the distance of 1.1 m
and 1.3 m from the inspection gangway on monolith 42 and 43, respectively. In the
FE model this has been taken at the widest representation of the crack, 1.2 m from
the inspection gangway. The crack width is calculated as the maximum principal
strain multiplied with the characteristic length of the element.
Crest displacement

Crack width

Figure 3.6: Illustration of the cracked buttress wall, showing the definition of positive
crest displacement and point where the crack width is calculated.

The calculated and the measured crest displacement is illustrated in the centre graph
in Figure 3.7. The displacement when the measurements started is initially defined
as the zero level in the measurements. It is unknown at exactly which position the
dam crest was at this time. Therefore, the measured displacement has been added
a constant value to define the zero level in the same way as the FE model, i.e.
according to the design drawings. The measured total displacement during a year
is 9.6 mm while the peak to peak value from the finite element analysis (FEA) is
10.4 mm. A positive displacement is defined in the downstream direction, i.e. when
the dam is moving away from the water.

29

CHAPTER 3. NUMERICAL STUDIES

Temperature ( C)

30

Measured
FEA

20
10
0
-10

Crack width variation (mm)

Crest displacement (mm)

11/01 12/01 01/01 02/01 03/01 04/01 05/01 06/01 07/01 08/01 09/01 10/01 11/01
0

Measured (monolith 43)


FEA

-2
-4
-6
-8
-10
-12

11/01 12/01 01/01 02/01 03/01 04/01 05/01 06/01 07/01 08/01 09/01 10/01 11/01

0.15

Measured (monolith 43)


Measured (monolith 42)
FEA

0.10
0.05
0
-0.05
-0.10

11/01 12/01 01/01 02/01 03/01 04/01 05/01 06/01 07/01 08/01 09/01 10/01 11/01

2007

Date

2008

Figure 3.7: Comparison of the measured and calculated behaviour of the dam. The
upper graph shows the outdoor temperature, the centre shows the crest
displacement and the lower graph shows the variation in crack width.

Crack width variation (mm)

The variation in crack width is shown in the bottom graph of Figure 3.7. The
initial value of all curves is adjusted so their average value is equal to zero. It is
only the variation in crack width that is measured and hence only the variation is
compared. The measured
variation in crack width, especially in monolith 42, and the
Measured (monolith 43)
0.15
calculated
variation
shows
good
Measured (monolith
42) agreement. During the summer, the measured crack
0.10
FEA
width on
0.05 monolith 42 is slightly higher and the calculated crack width variation is
in closer 0agreement with the measured on monolith 43. The peak to peak variation
of the -0.05
measured crack width on monolith 42 is 0.34 mm while it is 0.26 mm for the
-0.10
calculated variation. Due to the incomplete measured signal of monolith 43 within
-0.15
this period,
the minimum crack width variation that occurred in the winter was not
11/01 12/01 01/01 02/01 03/01 04/01 05/01 06/01 07/01 08/01 09/01 10/01 11/01
measured.
measured during the previous winter,
a
2007Considering the minimum value
Date
2008
peak to peak variation of 0.24 mm for monolith 43 is obtained.
Good agreement is found between measured and calculated displacement as well as
crack width. The largest deviation between the measured crest displacement and
the calculated can be found from approximately 2008-04-01 to 2008-06-01. During
this period the measured crest displacement increases much faster than the calculated displacement. There could be several reasons for this deviance between the

30

3.3. BRIDGES - SHEAR AND THERMAL CRACKING

measured and the calculated behaviour. Especially one measured value deviate the
most, the measurement performed 2008-04-24. In fact, at this date the measurement
technicians attached a LVDT to the pendulum, to receive continuous measurements
of the crest displacement. It was noticed that the LVDT sensor caused significant
disturbance to the measurement, resulting in almost twice as large crest displacements as expected. When they noticed this, the LVDT was removed. It is possible
that there still is some disturbance due to this which may have increased the displacement, even though the measured value from 2008-04-24 shown in Figure 3.7 is
made 15 minutes after the LVDT was removed. Comparing the measured behaviour
in crest displacement and crack width, it can be seen that the crack width does
not increase until the outdoor temperature increases. This also implies that the
measured value of the crest displacement in the end of April could be wrong. Other
reasons for the difference may be that the crest displacement is always measured
during daytime when the temperature is as highest. During this period there is
a large difference in daily temperature variation and since the calculation is made
with averaged weekly values some thermal effects may have been missed. Another
probable reason for this difference is that the temperature sensors are isolated to not
include the additional temperature from the sun radiation. This means that in the
finite element analysis, the effect of solar radiation is not included, only the actual
air temperature.

3.3

Bridges - shear and thermal cracking

The Grndal Bridge and the Alvik Bridge are studied in Papers IV, V and VI,
Malm et al. (2006); Malm and Sundquist (2009); Malm and Ansell (2009). The
Grndal Bridge is shown in Figure 3.8. In Paper IV, measurements performed on
the bridges and methods in design codes for crack width estimations are evaluated.
The monitoring of the bridges shows that the strengthening with post-tensioned
Dywidag tendons seems to have been successful, but the carbon fibre laminates
have a questionable effect regarding limiting the crack widths in the serviceability
state. The results show that the temperature is the main cause for crack width
variation. Two sensors on the Alvik Bridge, sensors MP1 and MP5, showed large
increasing trends in crack width variation. The sensor MP1 is the only sensor located
in a section not strengthened with Dywidag tendons. The sensor MP5 is measuring
the horizontal displacement between the top flange and the web. In Paper IV this
sensor is referred to as having a decreasing crack sliding displacement which is not
the case. The crack sliding is increasing but in the opposite direction of sensors
positive direction. Both the Grndal Bridge and Alvik Bridge show large horizontal
sliding movement in the cracks at the conjunction between the top flange and the
webs. High stresses in this region are shown in both Paper V and VI. Analyses
performed on the Kllsundsbron, a bridge of similar type and dimensions but with
a hinge in midspan, shows also high stresses in the conjunction between web and
top flange, Plos and Gylltoft (2004, 2006); Broo (2008).

31

CHAPTER 3. NUMERICAL STUDIES

Figure 3.8: The Grndal Bridge. Photo taken by S. Trillkott.

It is also shown in Paper IV that the crack width estimations in the national codes,
BBK 04 (2004); NS 3473 (2004); EC 2 (2004), are often sensitive to how the reinforcement stresses/strains are calculated. This can result in large differences in
estimation of crack width, depending on which method is used. An extensive description and comparison of the methods in design codes presented in this paper is
given by Malm (2005, 2006). The calculation methods in the national codes where
the direction of the crack is equal to the principal stress direction, i.e. a rotating
crack approach, that does not account for the shear stresses in the crack plane therefore give results that are conservative. In this case, these methods gave crack widths
and reinforcement stresses several times larger than the other, more realistic models.
The calculation methods that in some way include the effect of shear forces in the
crack plane, give much more realistic estimations of crack widths and reinforcement
stresses. For a load combination that describe the state at Grndal Bridge after two
years, gives crack widths of 0.13 and 0.23 mm with one method in BBK 04 (2004)
and the general method in NS 3473 (2004), respectively.
Paper V presents results from simulation of the construction process of the Grndal Bridge with time-dependent material properties. The calculated displacement
during construction is compared with measured displacements with good agreement. Creep and non-uniform drying shrinkage will affect the behaviour significantly. Creep has a large impact, even during cantilevering and the non-uniform
shrinkage will result in increasing displacements and stresses when the superstructure is complete. Creep resulted in a underestimation of the deflections of the cantilever of approximately 30 % if it neglected and the effect of creep resulted in three
times higher displacements in the mid span after two years of service, if creep was
included. The combined effect of creep and shrinkage is larger than their separate
effects, hence superposition will underestimate the result. The non-uniform drying shrinkage must be accounted for if detailed stresses and deflections are wanted,
when there is a large difference in thickness between the bottom and top flange.

32

3.3. BRIDGES - SHEAR AND THERMAL CRACKING

Significant effect of creep appearing only a short time after casting, can be found
in experiments presented in the literature, e.g. Scordelis et al. (1979). In their experiments, prototypes of two-span box-girder bridges were studied. The measured
displacements increased by 50 % due to creep after only 27 days in their experiment.
In the last paper of the thesis, Paper VI, the segmental construction is analysed
where areas subjected to high tensile stresses are defined with material models for
describing cracking. It is shown that cracks initiate during construction, especially
when the continuity cables in the bottom flange are post-tensioned. During thermal
loads those cracks propagate and new cracks initiate. During the summer, crack
widths of similar size as observed in-situ are found in the web on the side facing
inwards. The cracks calculated from the finite element model are up to 0.24 mm in
the most severely cracked section. Calculating the cracks based on the reinforcement
stresses in this section gives crack widths of 0.15 0.18 mm with the methods in
the design codes, BBK 04 (2004); NS 3473 (2004); EC 2 (2004). The crack widths
presented in this paper agree with the estimations in Paper IV and with the crack
widths found in this section in-situ, which mainly are 0.1 0.2 mm. The sensor
MP1 on the Alvik Bridge is located near the centre of gravity of the cross-section
at the distance 0.1L from the pier, where L is the length of the span as shown in
Paper IV. In Figure 11 of Paper V and in Figure 6 of Paper VI it can be seen that
rather large tensile stresses in the web at this position are found on the Grndal
Bridge. This confirms that there are large stresses within this region which could
be the reason for the increasing crack width measured at the sensor MP1.

33

Chapter 4
Conclusions
4.1

General conclusions

From a scientific point of view, the original feature of this thesis is that non-linear
finite element analysis was shown to estimate crack widths of shear type cracks
more accurately than the commonly used methods in the design codes. The finite
element material model used for evaluation is compared to other non-linear methods
and verified against laboratory test. From a practical point of view, the original
feature is that the numerical modelling method is used to analyse to two large
concrete structures where the cause for cracking is evaluated and verified against
observations made in-situ and experimental data.
There is today not one universal numerical model that in every aspect perfectly can
describe cracking in concrete. Some material models have much better means to
give good estimations than others. In this thesis solutions and recommendations to
avoid convergence difficulties are presented. If non-exclusively monotonic loading is
to be studied, the numerical methods should be able to describe both the stiffness
degradation and the plastic deformations that real concrete exhibit in experiments.
One big disadvantage with most numerical models is that they can not be combined
with the visco-elastic material description of creep. After comparing material models
based on different theories and verifying these against laboratory tests, one modelling approach seems to be especially superior. Even though most of the material
models could predict the ultimate load accurately, the predictions in crack pattern
and especially the crack widths and deflections differed considerably. The studied
methods primarily based on fracture mechanics gave too stiff behaviour after cracks
had initiated. It is commonly known that fixed crack models, can overestimate the
stiffness as an effect of stress-locking. This is not the case here since both the fixed
and the rotated model experienced about the same overestimation of the stiffness. It
is likely to assume that these models overestimate the aggregate interlocking effect
in the cracks, which in the fixed crack model is govern by the shear retention factor
and in the rotated model through an implicit reduction of the shear modulus based
on the principal stresses.
The rotated crack models also distributed physical cracks over several elements, re-

35

CHAPTER 4. CONCLUSIONS

gardless of element size. This resulted in that all these parallel cracks had to be
summarised to receive a reasonable crack width. The most severe problem with
rotating crack models is that shear type failure can be missed and the analysis continues past the real failure load. This is because the crack direction rotates and a
false equilibrium state can be found making it possible to continue past the point of
failure. The material model based on plasticity in combination with isotropic damage proved to be accurate in describing crack initiation and propagation. With the
constitutive model the non-linear behaviour of concrete, including crack pattern,
crack widths, load and displacements could be accurately predicted. The drawback of this model was the risk of convergence difficulties which might interrupt a
simulation.
There is a large difference in the estimation of the crack width based on the calculation methods in various design codes. The large difference appears since several of
these methods do not account for shear friction in the crack face. The calculation
methods that in some way include the effect of shear forces in the crack plane, give
much more reasonable estimations of the crack width and reinforcement stresses.
Implicit inclusion of the effect of aggregate interlocking could be made by changing
the angle of the cracks by a few degrees. Changing the direction so that equal reinforcement stresses in the two directions appears, as showed for the case presented in
Paper IV, can be sufficient to receive realistic estimation of the crack widths. This
method was only studied for this case and it is not certain that it works properly
for all cases. The more advanced method based on the modified compression field
theory, but assumed as initially cracked, is superior since it include the aggregate
interlock effects in a more precise matter.
When a suitable modelling method was chosen and it was verified that it could
describe shear type cracking, cyclic thermal cracking was studied. The analyses of
Storfinnforsen hydropower dam show that the calculated response is in good agreement with the measured variation in crack width and displacement and the observed
crack pattern. It was shown that the dam cracks due to seasonal temperature variation and the modification with an insulating wall have resulted in formation of
new cracks. Some of the new cracks seems to be of more serious nature regarding
the dam safety. It is also shown that an alternative placement of the insulation
would have been more suitable and would not have caused these new cracks. The
calculated crack pattern after temperature variation is compared to the observed
in-situ. In addition to this the crest movement and crack width variation for a year
are compared with measured data. All the calculated results are in close agreement
with the measured in-situ. The dam owners E.ON are now planing to strengthen
the Storfinnforsen dam and a similar dam in Ramsele both crossing the Faxlven
River for approximately 1500 million SEK.
As the modelling method was verified against cracking due to shear and thermal
actions, it was used to analyse the construction of the Grndal Bridge. It was
shown that creep is important already during cantilevering. For the Grndal Bridge
it resulted in a underestimation of the deflections of the cantilever of approximately
30 % if it was neglected. Creep was of even greater importance on the long-term
displacements. After two years of service, which corresponds to the age of the
Grndal Bridge when the cracks were first found, the displacement was three times

36

4.2. FURTHER RESEARCH

higher in the mid span if creep is included. The effect of non-uniform shrinkage rate
due to the time it takes for the drying shrinkage to occur is significant especially after
two years of service. The results showed that to create a finite element model of the
Grndal Bridge that accurately could describe the cracking in the webs, the timedependent effects have to be included. The analyses of the Grndal Bridge show that
cracks initiate during construction, especially when the continuing tendons in the
bottom flange are tensioned. Further thermal loads propagate the cracks and initiate
new cracks. The calculated behaviour of the bridge is similar to the observations
made on both the Grndal Bridge and the Alvik Bridge. Both crack pattern and
crack widths are in agreement with those found in-situ. Calculated displacements
during construction is in close agreement with the camber measured in-situ. The
monitoring of the bridges shows that the strengthening with post-tensioned Dywidag
tendons seems, so far, to have been successful. The carbon fibre laminates have a
small effect regarding limiting the crack widths in the serviceability state.

4.2

Further research

Continuum material models can successfully be used to accurately describe the shear
type cracking in concrete. However, there is still a need for large amount of research
regarding the development of numerical material models for non-linear finite element
methods. There are drawbacks regarding all present models whether it is computational efficiency, convergence difficulties, accuracy or the possibility of combining
cracking with other inelastic effects such as creep. The discrete crack models will
surely at some point be enough computational effective for analyses of large structures. Their drawback today is, in addition to long cpu-times required, that the
user manually have to define interface elements where cracks can initiate. Discrete
models would be very promising to use for large structures, if this procedure could
be simplified and if it is used in combination with adaptive meshing. Of especial
interest would be numerical material models that could describe the long-term behaviour of concrete structures by capturing the deterioration process with chloride
penetration that results in reinforcement corrosion. It is not evident that discrete
material models will be better in describing the crack initiation and propagation
than the continuum approaches. Concrete is an in-homogenous material where local
material variations can initiate cracks. This is one reason why continuum approaches
probably always will be most suitable for describing the cracking in concrete.
According to the design codes, crack widths should be limited to a predefined value
depending on the environmental conditions. In reality it is the crack width at the
level of reinforcement that should be limited and not the crack width at the face
of the concrete member. The corrosion of reinforcement is governed by the amount
of chloride ions reaching the reinforcement. It would be interesting with statistical
studies of the variation in crack width along the crack direction. In this way it
should be possible to determine better criteria for maximum allowed crack widths
in the serviceability state.
The methods available in design codes are based on laboratory experiments where
initially un-cracked concrete beams are loaded, usually in tension. Several experi-

37

CHAPTER 4. CONCLUSIONS

ments of shear type cracking can be found in the literature where calculation methods such as the modified compression field theory, Vecchio and Collins (1986), and
the softened-truss models, Hsu (1993), are developed from these experiments. It
is seldom that experiments of initially cracked concrete structures are presented in
the literature. Experiments of this type usually are intended to analyse the effect
of aggregate interlock such as the work presented by Walraven (1980). Puntel and
Saouma (2008) presents results from cyclic shear tests of concrete members with
a predefined crack subjected to compressive confinement. The results in this paper intend to show the frictional behaviour of a predefined crack. Pimanmas and
Maekawa (2001) presents tests where the shear capacity of reinforced concrete beams
that have been pre-cracked is studied. These beams have been pre-cracked in directions opposite to the direction for failure loading. Experiments of pre-cracked
concrete can be made to study the dilation of concrete during shear type loading.
These tests should be conducted for varying concrete quality and size of ballast. The
cracks should be scanned with an optic scanner to determine the roughness of the
crack faces. These experiments should also be performed with different specimen
sizes to determine the size effect.
Probabilistic models are used to account for uncertainties regarding material properties and loads. The probabilistic models are used to develop standards and construction codes in the area for dam safety. To apply these probabilistic models,
predefined failure modes are necessary. Combining the non-linear finite element
method with probabilistic models has great advantages. The failure mode and load
can be determined with the non-linear finite element model and with the means of
probabilistic method variations of material properties and loads can be investigated.
Combining these two methods should result in more reliable determination of the
safety index, as long as a proper finite element model and reasonable variations are
considered.

38

Bibliography
Alejano, L., Alonso, E., 2005. Considerations of the dilatancy angle in rocks and
rock masses. International Journal of Rock Mechanics and Mining Sciences, Vol.
42, No. 4, 481507.
Ansell, A., Bjrnstrm, J., Ekstrm, T., Hassanzadeh, M., Unosson, M., 2007.
Spricktillvxt i lamelldamm - Tillmpning av icke-linjra modeller Del I (In
Swedish). Report 08:21, Elforsk.
Bangash, M., 2001. Manuals of numerical methods in concrete - Modelling and
applications validated by experimental and site-monitoring data. Thomas Telford,
ASCE Press, Reston.
Baant, Z., 1979. Instability, ductility and size effect in strain-softening concrete.
Journal of Engineering Mechanics Division, Vol. 102, No. 2, 331344.
Baant, Z., Oh, B., 1983. Crack band theory for fracture of concrete. Materials and
Structures, Vol 16, No. 93, 155177.
BBK 04, 2004. Boverkets handbok om betongkonstruktioner, 3rd Edition. Boverket,
Byggavdelningen, Karlskrona.
Bentz, E., 2000. RESPONSE 2000: 2D Sectional analysis of beams and columns,
(www.ecf.utoronto.ca/bentz/r2k.htm).
Bjrnstrm, J., Ekstrm, T., Hassanzadeh, M., 2006. Spruckna betongdammar versikt och berkningsmetoder (In Swedish). Report 06:29, Elforsk.
Broo, H., 2008. Shear and torsion in concrete structures - Non-linear finite element
analysis in design and assessment. Ph.D. thesis, Chalmers University of Technology, Gothenburg.
Cervenka, V., 1970. Inelastic finite element analysis of reinforced concrete panels
under in-plane loads. Ph.D. thesis, University of Colorado, Boulder.
Cervenka, V., Jendele, L., Cervenka, J., 2005. ATENA program documentation Part 1 theory. Prague.
Chen, W., 1982. Plasticity in reinforced concrete. McGraw-Hill Book Company Inc.,
New York.
Chen, W., Han, D., 1995. Plasticity for structural engineers. Gau Lih Book Co.,
Ltd, Taiwan.

39

BIBLIOGRAPHY

Cornelissen, H., Hordijk, D., Reinhardt, H., 1986. Experimental determination of


crack softening characteristics of normal weight and lightweight concrete. Heron,
Vol. 31, No. 2, Delft University of Technology, Delft.
de Borst, R., 1986. Non-linear analysis of frictional materials. Ph.D. thesis, Delft
University of Technology, Delft.
de Borst, R., 2002. Fracture in quasi-brittle materials: A review of continuum
damage-based approaches. Engineering Fracture Mechanics, Vol. 69, No. 2, 95
112.
EC 2, 2004. Eurocode 2: Design of concrete structures - Part 1-1: General rules and
rules for buildings. EN 1992-1-1. Brussels.
Elfgren, L., 1989. Fracture mechanics of concrete structures - From theory to applications, Report of the Technical Committe 90-FMA Fracture Mechanics to
Concrete - Applications, RILEM Edition. Chapman and Hall Ltd, London.
Glvez, J., Cervenka, J., Cendn, D., Saouma, V., 2002. A discrete crack approach
to normal/shear cracking of concrete. Cement and Concrete Research, Vol. 32,
No. 10, 15671585.
Grasseli, G., 2001. Shear strength of rock joints based on quantified surface description. Ph.D. thesis, Ecole Polytechnique Fdrale de Lausanne (EPFL), Lausanne.
Grassl, P., Rempling, R., 2007. Influence of volumetric-deviatoric coupling on crack
prediction in concrete fracture tests. Engineering Fracture Mechanics, Vol. 74, No.
10, 16831693.
Gylltoft, K., 1991. Fracture mechanics analyses using ABAQUS. In: Analysis of Concrete Structures by Fracture Mechanics: Proceedings of the International RILEM
Workshop Dedicated to Professor Arne Hillerborg. Abisko, Sweden, June 28-30,
1989. Editors L. Elfgren and S. Shah. Chapman and Hall, London, pp. 152159.
Hassanzadeh, M., 1991a. Behaviour of fracture process zones in concrete influenced
by simultaneously applied normal and shear displacements. Ph.D. thesis, Lund
Institute of Technology, Lund.
Hassanzadeh, M., 1991b. Considerations regarding fracture zone response to simultaneous normal and shear displacement. In: Analysis of Concrete Structures by
Fracture Mechanics: Proceedings of the International RILEM Workshop Dedicated to Professor Arne Hillerborg. Abisko, Sweden, June 28-30, 1989. Editors
L. Elfgren and S. Shah. Chapman and Hall, London, pp. 4454.
Hibbitt, H., Karlsson, B., Sorensen, E., 2007. ABAQUS version 6.7 finite element
program - standard users manual. Pawtucket.
Hillerborg, A., 1985. The theoretical basis of a method to determine the fracture
energy Gf of concrete. Materials and Structures, Vol. 18, No. 4, 291296.
Hillerborg, A., Moder, M., Petersson, P., 1976. Analysis of crack formation and
growth in concrete by means of fracture mechanics and finite elements. Cement
and Concrete Research, Vol. 6, No. 6, 773782.

40

BIBLIOGRAPHY

Hsu, T., 1993. Unified theory of reinforced concrete. CRC Press, Boca Raton,
Florida.
http://www.temperatur.nu, 2009. Online temperature database.
Irons, B., Razzaque, A., 1972. Experience with the patch test for convergence of
finite elements. In: The Mathematical Foundations of the Finite Element Method
with Applications to Partial Differential Equations, Editor A.K. Aziz. Academic
Press, New York, pp. 557587.
Jankowiak, T., Lodygowski, T., 2005. Identification of parameters of concrete damage plasticity constitutive model. Foundations of Civil and Environmental Engineering, No. 6, 5369.
Jirsek, M., Baant, Z., 2002. Inelastic analysis of structures. John Wiley & Sons,
New York.
Johansson, F., Stille, H., 2009. En konceptuell modell fr uppskattning av maximal skjuvhllfasthet fr ra ofyllda sprickor av olika storlek (In Swedish). In:
Bergmekanikdagen. SveBeFo, Stockholm.
Ju, J., 1989. On energy-based coupled elastoplastic damage theories: constitutive
modeling and computational aspects. International journal of solids and structures
Vol. 25, No. 7, 803833.
Karihaloo, B., 2003. Failure of concrete. Comprehensive Structural Integrity Vol.
2.10, 475546.
Kupfer, H., Gerstle, K., 1969. Behavior of concrete under biaxial stresses. Journal
of Engineering Mechanics Division, Vol. 99, No. 4, 853866.
Kupfer, H., Hilsdorf, H., Raush, H., 1969. Behavior of concrete under biaxial stresses.
ACI Journal, Vol. 66, No. 8, 656666.
Lee, J., Fenves, G., 1998. Plastic-damage model for cyclic loading of concrete structures. Journal of Engineering Mechanics, Vol. 124, No. 8, 892900.
Lubliner, J., Oliver, J., Oller, S., Onate, E., 1989. A plastic-damage model for
concrete. International Journal of Solids and Structures Vol. 25, No. 3, 299326.
Malm, R., 2005. Analys av sprickbildning och sprickbredd vid plant spnningstillstnd i balkliv av armerad betong (In Swedish). Tech. report 88, Royal Institute
of Technology (KTH), Stockholm.
Malm, R., 2006. Shear cracks in concrete structures subjected to in-plane stresses.
Lic. Thesis, Royal Institute of Technology (KTH), Stockholm.
Malm, R., Ansell, A., 2008. Nonlinear analysis of thermally induced cracking of a
concrete dam. Submitted to ACI Structural Journal.
Malm, R., Ansell, A., 2009. Non-linear analyses of cracking in segmentally constructed concrete box-girder bridges. Submitted to Engineering Structures.

41

BIBLIOGRAPHY

Malm, R., Holmgren, J., 2008a. Cracking in deep beams owing to shear loading.
Part 1: Experimental study and assessment. Magazine of Concrete Research, Vol.
60, No. 5, 371379.
Malm, R., Holmgren, J., 2008b. Cracking in deep beams owing to shear loading. Part
2: Non-linear analysis. Magazine of Concrete Research, Vol. 60, No. 5, 381388.
Malm, R., James, G., Sundquist, H., 2006. Monitoring and evaluation of shear crack
initiation and propagation in webs of concrete box-girder sections. In: International Conference on Bridge Engineering - Challenges in the 21st Century. Hong
Kong.
Malm, R., Sundquist, H., 2009. Time-dependent analyses of segmentally constructed
balanced cantilever bridges. Submitted to Engineering Structures.
Mang, H., Lackner, R., Meschke, G., Mosler, J., 2003. Computional modelling of
concrete structures. Comprehensive Structural Integrity, Vol. 3.10, 536601.
NS 3473, 2004. Concrete structures - Design and detailing rules, 6th Edition. Standards Norway.
Patton, F., 1966. Multiple modes of shear failure in rock and related materials.
Ph.D. thesis, University of Illinois at Urbana-Champaign, Urbana.
Pimanmas, A., Maekawa, K., 2001. Finite element analysis and behaviour of precracked reinforced concrete members in shear. Magazine of Concrete Research,
Vol. 53, No. 5, 263282.
Plos, M., Gylltoft, K., 2004. Structural assessment of a prestressed box girder concrete bridge, using non-linear finite element analysis. In: Second International
Conference on Bridge Maintenance Safety and Management, IABMAS. Kyoto.
Plos, M., Gylltoft, K., 2006. Evaluation of shear capacity of a prestressed concrete
box girder bridge using non-linear FEM. Structural Engineering International,
Vol. 16, No. 3, 213221.
Puntel, E., Saouma, V., 2008. Experimental behavior of concrete joint interfaces
under reversed cyclic loading. Journal of Structural Engineering, Vol. 134, No. 9,
15581568.
Rashid, Y., 1968. Ultimate strength analysis of prestressed concrete pressure vessels.
Nuclear Engineering and Design, Vol. 7, No. 4, 334344.
RILEM 50-FMC, 1985. Determination of the fracture energy of mortar and concrete
by means of three-point bend tests on notched beams. Materials and Structures
Vol. 18, No. 4, 285290.
Rots, J., 2002. Comparative study of crack models. In: DIANA Conference, October,
ISBN NR: 90 5809 530 4.
Rots, J., Ivernizzi, S., 2004. Regularized sequentially linear saw-tooth softening
model. International Journal for Numerical and Analytical Methods in Geomechanics, Vol. 28, No. 7/8, 821856.

42

BIBLIOGRAPHY

Scordelis, A., Elfgren, L., Larsen, P., 1979. Time-dependent behavior of concrete
box girder bridges. ACI Journal, Vol. 76, No. 9, 159177.
van Mier, J., Nooru-Mohamed, M., Schlangen, E., 1991. Experimental analysis of
mixed mode I and II behaviour of concrete. In: Analysis of Concrete Structures
by Fracture Mechanics: Proceedings of the International RILEM Workshop Dedicated to Professor Arne Hillerborg. Abisko, Sweden, June 28-30, 1989. Editors
L. Elfgren and S. Shah. Chapman and Hall, London, pp. 3243.
Vecchio, F., Collins, M., 1986. The modified compression-field theory for reinforced
concrete elements subjected to shear. ACI Journal, Vol. 83, No. 2, 219231.
Walraven, J., 1980. Discontinuous modeling of strain localisation and failure. Ph.D.
thesis, Delft University of Technology, Delft.
Weihe, S., Krplin, B., de Borst, R., 1998. Classification of smeared crack models
based on material and structural properties. International Journal of Solids and
Structures, Vol. 35, No. 12, 12891308.
Zhang, N., Tan, K.-H., 2007. Direct strut-and-tie model for single span and continuous deep beams. Engineering Structures, Vol. 29, No. 11, 29873001.

43

List of Bulletins from the Department of Structural Engineering,


The Royal Institute of Technology, Stockholm
TRITA-BKN. Bulletin
Pacoste, C., On the Application of Catastrophe Theory to Stability Analyses of Elastic Structures.
Doctoral Thesis, 1993. Bulletin 1.
Stenmark, A-K., Dmpning av 13 m lng stlbalk "Ullevibalken". Utprovning av dmpmassor och
faststtning av motbalk samt experimentell bestmning av modformer och frlustfaktorer. Vibration tests
of full-scale steel girder to determine optimum passive control.
Licentiatavhandling, 1993. Bulletin 2.
Silfwerbrand, J., Renovering av asfaltgolv med cementbundna plastmodifierade avjmningsmassor.
1993. Bulletin 3.
Norlin, B., Two-Layered Composite Beams with Nonlinear Connectors and Geometry Tests and
Theory.
Doctoral Thesis, 1993. Bulletin 4.
Habtezion, T., On the Behaviour of Equilibrium Near Critical States.
Licentiate Thesis, 1993. Bulletin 5.
Krus, J., Hllfasthet hos frostnedbruten betong.
Licentiatavhandling, 1993. Bulletin 6.
Wiberg, U., Material Characterization and Defect Detection by Quantitative Ultrasonics.
Doctoral Thesis, 1993. Bulletin 7.
Lidstrm, T., Finite Element Modelling Supported by Object Oriented Methods.
Licentiate Thesis, 1993. Bulletin 8.
Hallgren, M., Flexural and Shear Capacity of Reinforced High Strength Concrete Beams without
Stirrups. Licentiate Thesis, 1994. Bulletin 9.
Krus, J., Betongbalkars lastkapacitet efter miljbelastning.
1994. Bulletin 10.
Sandahl, P., Analysis Sensitivity for Wind-related Fatigue in Lattice Structures.
Licentiate Thesis, 1994. Bulletin 11.
Sanne, L., Information Transfer Analysis and Modelling of the Structural Steel Construction Process.
Licentiate Thesis, 1994. Bulletin 12.
Zhitao, H., Influence of Web Buckling on Fatigue Life of Thin-Walled Columns.
Doctoral Thesis, 1994. Bulletin 13.
Kjrling, M., Dynamic response of railway track components. Measurements during train passage and
dynamic laboratory loading.
Licentiate Thesis, 1995. Bulletin 14.
Yang, L., On Analysis Methods for Reinforced Concrete Structures.
Doctoral Thesis, 1995. Bulletin 15.
Petersson, ., Svensk metod fr dimensionering av betongvgar.
Licentiatavhandling, 1996. Bulletin 16.
Lidstrm, T., Computational Methods for Finite Element Instability Analyses.
Doctoral Thesis, 1996. Bulletin 17.
Krus, J., Environment- and Function-induced Degradation of Concrete Structures.

119

Doctoral Thesis, 1996. Bulletin 18.


Editor, Silfwerbrand, J., Structural Loadings in the 21st Century.
Sven Sahlin Workshop, June 1996. Proceedings. Bulletin 19.
Ansell, A., Frequency Dependent Matrices for Dynamic Analysis of Frame Type Structures.
Licentiate Thesis, 1996. Bulletin 20.
Troive, S., Optimering av tgrder fr kad livslngd hos infrastrukturkonstruktioner.
Licentiatavhandling, 1996. Bulletin 21.
Karoumi, R., Dynamic Response of Cable-Stayed Bridges Subjected to Moving Vehicles.
Licentiate Thesis, 1996. Bulletin 22.
Hallgren, M., Punching Shear Capacity of Reinforced High Strength Concrete Slabs.
Doctoral Thesis, 1996. Bulletin 23.
Hellgren, M., Strength of Bolt-Channel and Screw-Groove Joints in Aluminium Extrusions.
Licentiate Thesis, 1996. Bulletin 24.
Yagi, T., Wind-induced Instabilities of Structures.
Doctoral Thesis, 1997. Bulletin 25.
Eriksson, A., and Sandberg, G., (editors), Engineering Structures and Extreme Events proceedings
from a symposium, May 1997. Bulletin 26.
Paulsson, J., Effects of Repairs on the Remaining Life of Concrete Bridge Decks.
Licentiate Thesis, 1997. Bulletin 27.
Olsson, A., Object-oriented finite element algorithms.
Licentiate Thesis, 1997. Bulletin 28.
Yunhua, L., On Shear Locking in Finite Elements.
Licentiate Thesis, 1997. Bulletin 29.
Ekman, M., Sprickor i betongkonstruktioner och dess inverkan p bestndigheten.
Licentiate Thesis, 1997. Bulletin 30.
Karawajczyk, E., Finite Element Approach to the Mechanics of Track-Deck Systems.
Licentiate Thesis, 1997. Bulletin 31.
Fransson, H., Rotation Capacity of Reinforced High Strength Concrete Beams.
Licentiate Thesis, 1997. Bulletin 32.
Edlund, S., Arbitrary Thin-Walled Cross Sections. Theory and Computer Implementation.
Licentiate Thesis, 1997. Bulletin 33.
Forsell, K., Dynamic analyses of static instability phenomena.
Licentiate Thesis, 1997. Bulletin 34.
Ikheimonen, J., Construction Loads on Shores and Stability of Horizontal Formworks.
Doctoral Thesis, 1997. Bulletin 35.
Racutanu, G., Konstbyggnaders reella livslngd.
Licentiatavhandling, 1997. Bulletin 36.
Appelqvist, I., Sammanbyggnad. Datastrukturer och utveckling av ett IT-std fr byggprocessen.
Licentiatavhandling, 1997. Bulletin 37.
Alavizadeh-Farhang, A., Plain and Steel Fibre Reinforced Concrete Beams Subjected to Combined
Mechanical and Thermal Loading.
Licentiate Thesis, 1998. Bulletin 38.

120

Eriksson, A. and Pacoste, C., (editors), Proceedings of the NSCM-11: Nordic Seminar on Computational
Mechanics, October 1998. Bulletin 39.
Luo, Y., On some Finite Element Formulations in Structural Mechanics.
Doctoral Thesis, 1998. Bulletin 40.
Troive, S., Structural LCC Design of Concrete Bridges.
Doctoral Thesis, 1998. Bulletin 41.
Trno, I., Effects of Contour Ellipticity upon Structural Behaviour of Hyparform Suspended Roofs.
Licentiate Thesis, 1998. Bulletin 42.
Hassanzadeh, G., Betongplattor p pelare. Frstrkningsmetoder och dimensioneringsmetoder fr plattor
med icke vidhftande spnnarmering.
Licentiatavhandling, 1998. Bulletin 43.
Karoumi, R., Response of Cable-Stayed and Suspension Bridges to Moving Vehicles. Analysis methods
and practical modeling techniques.
Doctoral Thesis, 1998. Bulletin 44.
Johnson, R., Progression of the Dynamic Properties of Large Suspension Bridges during Construction
A Case Study of the Hga Kusten Bridge.
Licentiate Thesis, 1999. Bulletin 45.
Tibert, G., Numerical Analyses of Cable Roof Structures.
Licentiate Thesis, 1999. Bulletin 46.
Ahlenius, E., Explosionslaster och infrastrukturkonstruktioner - Risker, vrderingar och kostnader.
Licentiatavhandling, 1999. Bulletin 47.
Battini, J-M., Plastic instability of plane frames using a co-rotational approach.
Licentiate Thesis, 1999. Bulletin 48.
Ay, L., Using Steel Fiber Reinforced High Performance Concrete in the Industrialization of Bridge
Structures.
Licentiate Thesis, 1999. Bulletin 49.
Paulsson-Tralla, J., Service Life of Repaired Concrete Bridge Decks.
Doctoral Thesis, 1999. Bulletin 50.
Billberg, P., Some rheology aspects on fine mortar part of concrete.
Licentiate Thesis, 1999. Bulletin 51.
Ansell, A., Dynamically Loaded Rock Reinforcement.
Doctoral Thesis, 1999. Bulletin 52.
Forsell, K., Instability analyses of structures under dynamic loads.
Doctoral Thesis, 2000. Bulletin 53.
Edlund, S., Buckling of T-Section Beam-Columns in Aluminium with or without Transverse Welds.
Doctoral Thesis, 2000. Bulletin 54.
Lfsjgrd, M., Functional Properties of Concrete Roads General Interrelationships and Studies on
Pavement Brightness and Sawcutting Times for Joints.
Licentiate Thesis, 2000. Bulletin 55.
Nilsson, U., Load bearing capacity of steel fibree reinforced shotcrete linings.
Licentiate Thesis, 2000. Bulletin 56.
Silfwerbrand, J. and Hassanzadeh, G., (editors), International Workshop on Punching Shear Capacity of
RC Slabs Proceedings. Dedicated to Professor Sven Kinnunen. Stockholm June 7-9, 2000. Bulletin 57.

121

Wiberg, A., Strengthening and repair of structural concrete with advanced, cementitious composites.
Licentiate Thesis, 2000. Bulletin 58.
Racutanu, G., The Real Service Life of Swedish Road Bridges - A case study.
Doctoral Thesis, 2000. Bulletin 59.
Alavizadeh-Farhang, A., Concrete Structures Subjected to Combined Mechanical and Thermal Loading.
Doctoral Thesis, 2000. Bulletin 60.
Wppling, M., Behaviour of Concrete Block Pavements - Field Tests and Surveys.
Licentiate Thesis, 2000. Bulletin 61.
Getachew, A., Trafiklaster p broar. Analys av insamlade och Monte Carlo genererade fordonsdata.
Licentiatavhandling, 2000. Bulletin 62.
James, G., Raising Allowable Axle Loads on Railway Bridges using Simulation and Field Data.
Licentiate Thesis, 2001. Bulletin 63.
Karawajczyk, E., Finite Elements Simulations of Integral Bridge Behaviour.
Doctoral Thesis, 2001. Bulletin 64.
Thyr, T., Strength of Slotted Steel Studs.
Licentiate Thesis, 2001. Bulletin 65.
Tranvik, P., Dynamic Behaviour under Wind Loading of a 90 m Steel Chimney.
Licentiate Thesis, 2001. Bulletin 66.
Ullman, R., Buckling of Aluminium Girders with Corrugated Webs.
Licentiate Thesis, 2002. Bulletin 67.
Getachew, A., Traffic Load Effects on Bridges. Statistical Analysis of Collected and Monte Carlo
Simulated Vehicle Data.
Doctoral Thesis, 2003. Bulletin 68.
Quilligan, M., Bridge Weigh-in-Motion. Development of a 2-D Multi-Vehicle Algorithm.
Licentiate Thesis, 2003. Bulletin 69.
James, G., Analysis of Traffic Load Effects on Railway Bridges.
Doctoral Thesis 2003. Bulletin 70.
Nilsson, U., Structural behaviour of fibre reinforced sprayed concrete anchored in rock.
Doctoral Thesis 2003. Bulletin 71.
Wiberg, A., Strengthening of Concrete Beams Using Cementitious Carbon Fibre Composites.
Doctoral Thesis 2003. Bulletin 72.
Lfsjgrd, M., Functional Properties of Concrete Roads Development of an Optimisation Model and
Studies on Road Lighting Design and Joint Performance.
Doctoral Thesis 2003. Bulletin 73.
Bayoglu-Flener, E., Soil-Structure Interaction for Integral Bridges and Culverts.
Licentiate Thesis 2004. Bulletin 74.
Lutfi, A., Steel Fibrous Cement Based Composites. Part one: Material and mechanical properties.
Part two: Behaviour in the anchorage zones of prestressed bridges.
Doctoral Thesis 2004. Bulletin 75.
Johansson, U., Fatigue Tests and Analysis of Reinforced Concrete Bridge Deck Models.
Licentiate Thesis 2004. Bulletin 76.

122

Roth, T., Langzeitverhalten von Spannsthlen in Betonkonstruktionen.


Licentitate Thesis 2004. Bulletin 77.
Hedebratt, J., Integrerad projektering och produktion av industrigolv Metoder fr att frbttra
kvaliteten.
Licentiatavhandling, 2004. Bulletin 78.
sterberg, E., Revealing of age-related deterioration of prestressed reinforced concrete containments in
nuclear power plants Requirements and NDT methods.
Licentiate Thesis 2004. Bulletin 79.
Broms, C.E., Concrete flat slabs and footings New design method for punching and detailing for
ductility.
Doctoral Thesis 2005. Bulletin 80.
Wiberg, J., Bridge Monitoring to Allow for Reliable Dynamic FE Modelling - A Case Study of the New
rsta Railway Bridge.
Licentiate Thesis 2006. Bulletin 81.
Mattsson, H-., Funktionsentreprenad Brounderhll En pilotstudie i Uppsala ln.
Licentiate Thesis 2006. Bulletin 82.
Masanja, D. P, Foam concrete as a structural material.
Doctoral Thesis 2006. Bulletin 83.
Johansson, A., Impregnation of Concrete Structures Transportation and Fixation of Moisture in Water
Repellent Treated Concrete.
Licentiate Thesis 2006. Bulletin 84.
Billberg, P., Form Pressure Generated by Self-Compacting Concrete Influence of Thixotropy and
Structural Behaviour at Rest.
Doctoral Thesis 2006. Bulletin 85.
Enckell, M., Structural Health Monitoring using Modern Sensor Technology Long-term Monitoring of
the New rsta Railway Bridge.
Licentiate Thesis 2006. Bulletin 86.
Sderqvist, J., Design of Concrete Pavements Design Criteria for Plain and Lean Concrete.
Licentiate Thesis 2006. Bulletin 87.
Malm, R., Shear cracks in concrete structures subjected to in-plane stresses.
Licentiate Thesis 2006. Bulletin 88.
Skoglund, P., Chloride Transport and Reinforcement Corrosion in the Vicinity of the Transition Zone
between Substrate and Repair Concrete.
Licentiate Thesis 2006. Bulletin 89.
Liljencrantz, A., Monitoring railway traffic loads using Bridge Weight-in-Motion.
Licentiate Thesis 2007. Bulletin 90.
Stenbeck, T., Promoting Innovation in Transportation Infrastructure Maintenance Incentives,
Contracting and Performance-Based Specifications.
Doctoral Thesis 2007. Bulletin 91
Magnusson, J., Structural Concrete Elements Subjected to Air Blast Loading.
Licentiate Thesis 2007. Bulletin 92.

123

Pettersson, L., G., Full Scale Tests and Structural Evaluation of Soil Steel Flexible Culverts with low
Height of Cover.
Doctoral Thesis 2007. Bulletin 93
Westerberg, B., Time-dependent effects in the analysis and design of slender concrete compression
members.
Doctoral Thesis 2008. Bulletin 94
Mattsson, H-, Integrated Bridge Maintenance. Evaluation of a pilot project and future perspectives.
Doctoral Thesis 2008. Bulletin 95
Andersson, A., Utmattningsanalys av jrnvgsbroar. En fallstudie av stlbroarna mellan Stockholm
Central och Sder Mlarstrand, baserat p teoretiska analyser och tjningsmtningar.
Licentiate Thesis 2009. Bulletin 96

The bulletins enumerated above, with the exception for those which are out of print, may be purchased
from the Department of Civil and Architectural Engineering, The Royal Institute of Technology, SE-100
44 Stockholm, Sweden.
The department also publishes other series. For full information see our homepage http://www.byv.kth.se

124