VECTOR2 & FORMWORKS USERS MANUAL

P. S. Wong
F.J. Vecchio




August, 2002


ii
Abstract
VecTor2 is a program based on the Modified Compression Field Theory for nonlinear finite
element analysis of reinforced concrete membrane structures. Considering the inherent intricacies of
nonlinear finite element analysis and VecTor2, user facilities are imperative to their rational and
convenient application.
Program documentation for VecTor2 and a new preprocessor software entitled FormWorks are
presented in this report. The program documentation describes the theoretical basis of VecTor2, the finite
element library and the constitutive models for concrete, reinforcement and bond materials. This
documentation and the FormWorks manual provide guidance for modeling and data input.
FormWorks is a graphics-based preprocessor program for the Windows environment.
FormWorks includes facilities for data visualization and input, bandwidth reduction and automatic mesh
generation. The latter requires only economical user input, permits a high degree of user control over
mesh topology and generates mixed element type meshes for reinforced concrete structures.




iii
Table of Contents
Abstract ii
List of Figures xi
List of Appendices xv
Notation xvi
1 Introduction 1
1.1 Background 1
1.2 The VecTor2 Program 2
1.3 Research Objectives 3
1.4 Organization 4
2 Part I: VecTor2 5
2.1 Introduction 5
2.2 Modified Compression Field Theory (MCFT) 5
2.2.1 Assumptions 6
2.2.2 Compatibility Relationships 6
2.2.3 Equilibrium Relationships 7
2.2.4 Constitutive Relationships 9
2.2.5 Consideration of Local Crack Conditions 10
2.3 Disturbed Stress Field Model (DSFM) 12
2.3.1 Compatibility Relationships 13
2.3.2 Equilibrium Relationships 15
2.3.3 Constitutive Relationships 16
2.3.4 Shear Slip Relationships 19
2.4 Finite Element Implementation 20
2.4.1 Composite Material Stiffness Matrix 22

iv
2.4.2 Element Stiffness Matrices 24
2.4.3 Pseudo Nodal Loads 25
2.4.4 Total Load Iterative Secant Stiffness Solution 26
2.5 Augmented Analysis Features 27
2.5.1 Modeling of Lateral Expansion Effect 27
2.5.2 Modeling of Triaxial Stresses 29
2.5.3 Analysis of Cyclically Loaded Structures 30
2.5.4 Analysis of Construction and Loading Chronology 33
2.5.5 Bond-Slip Mechanisms 34
3 Element Library 35
3.1 Reinforced Concrete Elements 35
3.1.1 Constant Strain Triangle 35
3.1.2 Plane Stress Rectangle 37
3.1.3 Quadrilateral 38
3.2 Reinforcement Elements 39
3.2.1 Truss Bar 39
3.3 Bond Slip Elements 40
3.3.1 Link 41
3.3.2 Contact 42
4 Models for Concrete Materials 45
4.1 Compression Pre-Peak Response 45
4.1.1 Linear 46
4.1.2 Popovics 47
4.1.3 Hognestad (Parabola) 48
4.1.4 Popovics (High Strength) 48
4.1.5 Hoshikuma et al. 50
4.2 Compression Post-Peak Response 51
v
4.2.1 Pre-Peak Base Curve 53
4.2.2 Modified Park-Kent 53
4.2.3 Popovics / Mander 54
4.2.4 Hoshikuma et al. 55
4.2.5 Saenz / Spacone 56
4.3 Compression Softening 58
4.3.1 No compression softening 60
4.3.2 Vecchio 1992-A (e1/e2-Form) 61
4.3.3 Vecchio 1992-B (e1/e0-Form) 62
4.3.4 Vecchio-Collins 1982 62
4.3.5 Vecchio-Collins 1986 63
4.4 Tension Stress-Strain Response 64
4.4.1 No Tension Stiffening 65
4.4.2 Bentz 2003 65
4.4.3 Vecchio 1982 65
4.4.4 Collins-Mitchell 1987 66
4.4.5 Bentz 1999 66
4.4.6 Izumo, Maekawa Et Al. 67
4.4.7 Tension Chord Model (Kaufmann) 68
4.5 Tension Softening 69
4.5.1 Not Considered 71
4.5.2 Linear 71
4.5.3 Yamamoto 1999 No residual 72
4.6 Tension Splitting 72
4.6.1 Not Considered 73
4.6.2 DeRoo 1995 73
4.7 Lateral Expansion 74
vi
4.7.1 Constant Poisson's ratio 74
4.7.2 Variable Poisson's Ratio Kupfer 74
4.7.3 Variable Poissons Ratio Montoya 75
4.8 Confinement Strength 76
4.8.1 Strength Enhancement Neglected 76
4.8.2 Kupfer / Richart Model 76
4.8.3 Selby Model 77
4.8.4 Montoya / Ottosen 78
4.9 Cracking Criterion 79
4.9.1 Uniaxial cracking stress 80
4.9.2 Mohr-Coulomb (Stress) 80
4.9.3 Mohr-Coulomb (Strain) 81
4.9.4 CEB-FIP Model 81
4.9.5 Gupta 1998 Model 82
4.10 Crack Slip Check 82
4.10.1 Crack shear check omitted 82
4.10.2 Vecchio-Collins 1986 83
4.10.3 Gupta 1998 Model 83
4.11 Crack Width Check 83
4.12 Element Slip Distortion 84
4.12.1 Not considered 86
4.12.2 Stress Model (Lai-Vecchio) 86
4.12.3 Stress Model (Maekawa) 86
4.12.4 Stress Model (Walraven) 87
4.12.5 Hybrid-I, II, and III Models 87
4.12.6 Rotation lag of n degrees 87
4.13 Hysteretic Response 88
vii
4.13.1 No Plastic Offsets 88
4.13.2 Plastic offsets; linear loading/unloading 89
4.13.3 Plastic offsets; nonlinear loading/unloading 91
4.13.4 Plastic offsets; nonlinear w/ cyclic decay (Palermo Model) 92
5 Models for Reinforcement Materials 96
5.1 Stress-Strain Response 96
5.1.1 Ductile Steel Reinforcement 96
5.1.2 Prestressing Steel 97
5.1.3 Tension Only Reinforcement 98
5.1.4 Compression Only Reinforcement 98
5.1.5 Externally Bonded FRP Reinforcement 99
5.2 Hysteretic Response 99
5.2.1 Linear 99
5.2.2 Seckin Model w/ Bauschinger Effect 99
5.2.3 Elastic-Plastic 100
5.2.4 Elastic-Plastic w/ Hardening 100
5.3 Dowel Action 101
5.3.1 Not considered 101
5.3.2 Tassios Model 101
5.4 Reinforcement Buckling 102
5.4.1 Not Considered 102
5.4.2 Asatsu Model 103
6 Models for Bond 104
6.1 Bond Stress-Slip Models for Embedded Bars 104
6.1.1 Perfect bond 105
6.1.2 Eligehausen Model 105
6.1.3 Gan Model 108
viii
6.1.4 Harajli Model 109
6.1.5 Hooked Bars 111
6.1.6 Fujii Model 112
6.2 Bond Stress-Slip Models for Externally Bonded Plates or Sheets 114
6.2.1 Perfect bond 114
6.2.2 Other models 114
7 Part II: FormWorks 115
7.1 Introduction 115
7.2 Installing FormWorks 115
7.3 An Overview of the FormWorks Modeling Process 116
7.3.1 Before Using FormWorks 117
7.3.2 VecTor2 Input and Output Files 117
7.4 FormWorks Basics 119
7.4.1 The FormWorks Interface 119
7.4.2 Creating a New Workspace 120
7.4.3 Saving the Workspace 121
7.4.4 Opening a Saved Workspace File 122
7.5 Viewing and Printing the Workspace 122
7.5.1 Manipulating the View 122
7.5.1.1 Horizontal and Vertical Scaling 122
7.5.1.2 Changing Display Limits 122
7.5.1.3 Zooming and Panning 123
7.5.1.4 Selecting Display Options 124
7.5.2 Printing the Workspace 127
8 The Job Data 128
8.1.1 The J ob Control Page 128
8.1.1.1 J ob Data Group 129
ix
8.1.1.2 Structure Data Group 129
8.1.1.3 Load Data Group 129
8.1.1.4 Analysis Parameters Group 133
8.1.2 The Model Page 134
8.1.2.1 Concrete Models 136
8.1.2.2 Reinforcement Models 137
8.1.2.3 Bond Models 137
8.1.2.4 Analysis Models 138
9 The Structure Data 139
9.1.1 Preliminary considerations 139
9.1.2 Structure Limits 140
9.1.3 Structure Information 141
9.1.4 Specifying Material Types 141
9.1.4.1 Reinforced Concrete Material Types 142
9.1.4.2 Reinforcement Material Types 146
9.1.4.3 Bond Types 147
9.1.5 Defining Nodes 150
9.1.6 Defining Elements 151
9.1.6.1 Viewing Elements Attributes 156
9.1.6.2 Viewing Multiple Element Layers 157
9.1.7 Assigning Material Types 157
9.1.8 Restraining the Structure 159
9.1.9 Automatic Method 161
9.1.9.1 Reinforced Concrete Regions 161
9.1.9.2 Reinforcement 170
9.1.9.3 Voids 175
9.1.9.4 Line Constraints 177
x
9.1.9.5 Point Constraints 178
9.1.9.6 Generating the Mesh 180
10 The Load Case Data 183
10.1.1 Load Limits 183
10.1.2 Selecting the Load Case 183
10.1.3 Load Information 183
10.1.4 Nodal Loads 184
10.1.5 Support Displacements 186
10.1.6 Gravity Loads 186
10.1.7 Temperature Loads 188
10.1.8 Concrete Prestrains 190
10.1.9 Ingress Pressures 191
10.1.10 Surface Thermal Loads 191
10.1.11 Impulse Loads 193
10.1.12 Ground Acceleration Loads 194
11 Running VecTor2 195
11.1 Starting the Analysis 195
11.2 A Simple Example 197
12 Summary & Recommendations 201
12.1 Summary 201
12.2 Recommendations 201
References 203
Appendix A VecTor2 Input Files for Sample Problem 209

xi
List of Figures
Figure 1: Reinforced concrete membrane element subject to in-plane stresses 5
Figure 2: Average concrete strains due to average stress-strain response of concrete 7
Figure 3 Free body diagram of reinforced concrete element showing average stresses 8
Figure 4 Comparison of average and local stresses at a crack 11
Figure 5: Deformation due to crack shear slip 14
Figure 6 VecTor2 nonlinear finite element analysis algorithm 21
Figure 7: Definition of secant moduli for a) concrete b) reinforcement 24
Figure 8: Constant strain triangle element 36
Figure 9: Plane stress rectangle element 37
Figure 10: a) Quadrilateral element
b) Decomposition of quadrilateral element into two constant strain triangle elements 38
Figure 11: Truss bar element 39
Figure 12 Link element 41
Figure 13: Contact element 43
Figure 14: Linear pre- and post-peak concrete compression response 46
Figure 15 Popovics pre- and post-peak concrete compression response 47
Figure 16: Hognestad parabolic pre- and post-peak concrete compression response 48
Figure 17: Popovics high strength pre- and post-peak concrete compression response 50
Figure 18: Hoshikuma et al. concrete compressive pre-peak response 51
Figure 19: Smith-Young concrete compression post-peak response for f
p
<=f
c
53
Figure 20: Modified Park-Kent post-peak concrete compression response 54
Figure 21: Hoshikuma et al. concrete compressive post-peak response 56
Figure 22: Saenz/Spacone concrete compressive post-peak response 58
Figure 23: Strength and strain-softened compression response 59
Figure 24: Strength-only softened compression response 60
Figure 25: Vecchio 1992-A compression softening model 62
xii
Figure 26: Vecchio 1982 compression softening model 63
Figure 27: Vecchio 1982 and Collins-Mitchell tension stiffening response 66
Figure 28: Izumo, Maekawa et al. tensions stiffening response 67
Figure 29: Tension chord model 69
Figure 30: Linear tension softening response with and without residual 71
Figure 31: Yamamoto tension softening response with and without residual 72
Figure 32: DeRoo tension splitting response 73
Figure 33: Kupfer variable Poissons ratio model 75
Figure 34: Mohr-Coulomb (Stress) cracking criterion 81
Figure 35: Palermo model of concrete hysteretic response in compression 95
Figure 36: Palermo model of concrete hysteretic response in tension 95
Figure 37: Ductile steel reinforcement stress-strain response 97
Figure 38: Prestressing steel reinforcement stress-strain response 98
Figure 39: Seckin model of reinforcement hysteretic response 100
Figure 40: Dowel resistance mechanism 102
Figure 41: Eligehausen bond stress-slip response 107
Figure 42: Gan bond stress-slip response 109
Figure 43: Harajli bond stress-slip response 111
Figure 44: Hooked bar bond stress-slip response 112
Figure 45: Bond stress-slip response for externally bonded plates or sheets 114
Figure 46: Input and output files forVecTor2 analysis 117
Figure 47: FormWorks application window 120
Figure 48: Creating new Workspace windows 121
Figure 49: Display Limits dialog box 123
Figure 50: Display Options dialog box 125
Figure 51: J ob Control property page 128
Figure 52: Monotonic type loading 130
xiii
Figure 53: Cyclic type loading 130
Figure 54: Reversed cyclic type loading 131
Figure 55: The Models page 135
Figure 56: Structure Limits dialog box 140
Figure 57: Structure Information dialog box 141
Figure 58: Reinforced Concrete Materials Properties Dialog Box 142
Figure 59: Reinforcement Materials Properties Dialog Box 147
Figure 60: Bond Properties Dialog Box 148
Figure 61: Create Nodes Dialog Box 150
Figure 62: Example of node creation entry 151
Figure 63: Create Rectangular Elements Dialog Box 152
Figure 64: Example of rectangular element creation entry 153
Figure 65: Example of link element creation entry 154
Figure 66: Example of contact element creation entry 155
Figure 67: Element attributes dialog 156
Figure 68: Assign Material types dialog box 158
Figure 69: Selecting elements for material type assignment with the mouse 159
Figure 70: Create Support Restraints Dialog Box 159
Figure 71: Selecting nodes for restraints with the mouse 160
Figure 72: Examples of legal and illegal region definitions and intersections 162
Figure 73: RC Regions Property Page 163
Figure 74: Example of reinforced concrete region creation 164
Figure 75: Hybrid Discretization 165
Figure 76: Grid Superposition Discretization 166
Figure 77: Division Point Insertion Discretization 166
Figure 78: Hybrid Discretization with Rectangle Option 167
Figure 79: Division Point Insertion Discretization with Quadrilateral Option 167
xiv
Figure 80: Division Point Insertion Discretization with Smoothed Triangles Option 168
Figure 81: Reinforcement Page 171
Figure 82: Example of reinforcement path creation 173
Figure 83: Voids and Constraints 176
Figure 84: Example of point constraint creation 180
Figure 85: Mesh Creation Page 181
Figure 86: Load Limits Dialog Box 183
Figure 87: Load Information Dialog Box 184
Figure 88: Apply Nodal Loads Dialog Box 184
Figure 89: Apply Support Displacements Dialog Box 186
Figure 90: Apply Gravity Loads Dialog Box 187
Figure 91: Apply Temperature Loads Dialog 189
Figure 92: Apply Concrete Prestrains Dialog Box 190
Figure 93: Apply Ingress Pressures Dialog Box 191
Figure 94: Apply Surface Thermal Loads Dialog 192
Figure 95: Apply Impulse Forces Dialog Box 193
Figure 96: Apply Ground Acceleration Load Dialog 194
Figure 97: Bandwidth Reduction Dialog Box 195
Figure 98: VecTor2 Analysis Proceeding 197
Figure 99: Reinforced Concrete Beam 198
Figure 100: Finite element mesh for half of beam 199
Figure 101: Load versus midpsan deflection of reinforced concrete beam 199
Figure 102: Crack pattern of reinforced concrete beam at ultimate load 200

xv
List of Appendices
Appendix A VecTor2 Input Files for Sample Problem 209


xvi
Notation
a =maximum aggregate size
c =convergence averaging factor
c
t
=compression softening coefficient
[B] =element strain-displacement matrix
d
b
=diameter of reinforcing bar
[D] =composite material stiffness matrix
[D
c
] =concrete material stiffness matrix relative to x,y axes
[D
c
] =concrete material stiffness matrix relative to principal axes
[D
s
] =reinforcement material stiffness matrix relative to x,y axes
[D
s
] =reinforcement material stiffness matrix relative to principal axes
C
d
=compression softening, strain softening factor
C
s
=compression softening, shear slip factor
E
c
=initial tangent stiffness of concrete
1 c E =secant modulus of concrete in the principal tensile direction
2 c E =secant modulus of concrete in the principal compressive direction
E
s
=initial tangent stiffness of reinforcement
s E =secant modulus of reinforcement
f
c
=concrete cylinder uniaxial compressive strength
f
cc
=concrete cube strength
f
cx
=average concrete axial stress in the x-direction
f
cy
=average concrete axial stress in the y-direction
f
p
=peak concrete compressive stress
f
sx
=average stress of reinforcement parallel to the x-direction
f
sy
=average stress of reinforcement parallel to the y-direction
f
scrx
=local stress at a crack of reinforcement parallel to the x-direction
xvii
f
scry
=local stress at a crack of reinforcement parallel to the y-direction
f
syield
=yield strength of reinforcement
f
t
=concrete cracking strength
[F*] =pseudo nodal load vector
[F] =applied external nodal load vector
[F] =total nodal load vector
c G =secant shear modulus of concrete
G
f
=fracture energy parameter
k =post-peak decay parameter for stress-strain response of concrete in compression
k
r
=spring stiffness of link element radial to reinforcement direction
k
t
=spring stiffness of link element tangential to reinforcement direction
[k] =element stiffness matrix
[k
c
] =element stiffness matrix contribution from concrete
[k
s
] =element stiffness matrix contribution from reinforcement
[K] =structure stiffness matrix
L
r
=characteristic length
n =curve fitting parameter for stress-stain response of concrete in compression
[r] =nodal displacement vector
s =average crack spacing
s
mx
=average crack spacing in the x-direction
s
my
=average crack spacing in the y-direction
s

=average crack spacing perpendicular to the crack

[T] =coordinate transformation matrix
v
12
,v
21
=Poissons ratios of concrete in the principal directions
e
v
12
=elastic component of Poissons ratio
*
v
12
=residual component of Poissons ratio
xviii
v
cxy
=average concrete shear stress
v
ci
=local concrete shear stress at a crack
w =average crack spacing
=orientation of reinforcement relative to positive x-axis

cr
=crack width check crack coefficient

d
=compression softening reduction factor

l
=confinement strength enhancement factor

=post-cracking rotation of principal total strain field

=post-cracking rotation of the principal stress field

c1
=average net concrete axial strain, in the principal tensile 1-direction

c2
=average net concrete axial strain, in the principal compressive 2-direction

cr
=concrete cracking strain

cx
=average net concrete axial strain, in the x-direction

cy
=average net concrete axial strain, in the y-direction
o
c
=elastic strain offsets in concrete
p
c
=plastic strain offsets in concrete
cmx
=maximum concrete compressive strain
tmx
=maximum concrete tensile strain

o
=concrete compressive strain corresponding to f
c

p
=concrete compressive strain corresponding to f
p

sh
=reinforcement strain at onset of strain hardening

sx
=average strain of reinforcement parallel to the x-direction

sy
=average strain of reinforcement parallel to the y-direction

syield
=yield strain of reinforcement
o
s
=elastic strain offsets in reinforcement
xix
p
s
=plastic strain offsets in reinforcement

ts
=terminal strain in concrete tension softening response

u
=ultimate strain of reinforcement

x
=total axial strain in the x-direction

y
=total axial strain in the y-direction
s
x
=average axial strain in the x-direction, due to shear slip
s
y
=average axial strain in the y-direction, due to shear slip

cxy
=average net concrete shear strain

xy
=total shear strain
s
xy
=average shear strain due to shear slip
=orientation of principal net concrete strain field, measured counter-clockwise from x-axis

ic
=orientation of principal strain and stress field at initial cracking

l
=rotation lag between orientation of principal strain and principal stress fields in concrete

n
=angle between normal to crack surface and longitudinal axis of reinforcement

=orientation of principal total strain field, measured counter-clockwise from x-axis

=orientation of principal stress field, measured counter-clockwise from x-axis

=reinforcement ratio

x
=applied axial stress in the x-direction

y
=applied axial stress in the y-direction
[
o
] =pseudo prestress vector

xy
=applied shear stress
=transformation angle between principal direction and positive x-axis



1
1 Introduction
1.1 Background
VecTor is a suite of computer programs dedicated to nonlinear finite element analysis of
reinforced concrete (NLFEARC). These programs have been developed at the University of Toronto by
researchers studying reinforced concrete behavior and applications of the finite element method over the
last two decades. By combining a variety of realistic nonlinear models for reinforced concrete with the
powerful analytical capabilities of finite element analysis, these programs permit more accurate
assessments of structural performance (strength, post-peak behavior, failure mode, deflections and
cracking) than can be achieved by linear-elastic methods. At the same time, the finite element method
allows analysts to address the composite nature of reinforced concrete material, changing material
properties due to progressive cracking, challenging geometries and loadings complexities which might
thwart conventional analysis techniques.
As the results of NLFEARC have become increasingly reliable and modern digital computers
advance in performance, NLFEARC has evolved from the research domain to a practical design tool.
However, the capabilities that make these NLFEARC programs so remarkable, pose potential difficulties
to the non-specialist user. While the reinforced concrete models are critical to the efficacy of the analysis,
they may also be obscure to its users. Without an understanding of the theoretical basis of programs and
their operation, the impressive output of the programs cannot be regarded with confidence. As such, the
rational and effective use of NLFEARC programs requires guidance and user experience.
Moreover, generating input for finite element analysis requires knowledge of the finite element
method itself. In order to idealize a structure to a valid mathematical model, users must exercise their
discretion regarding the mesh topology, material selection, boundary conditions, load representation and
computational efficiency. Lacking preprocessor facilities, the time-consuming and error-prone nature of
data input may discourage the practical application of NLFEARC.
2
The availability of user facilities must be regarded as necessary for progress in the practical
application of NLFEARC. It is the objective of this report to address this need by developing program
documentation and preprocessor software and for the nonlinear finite element program VecTor2.
1.2 The VecTor2 Program
VecTor2 is a nonlinear finite element program for the analysis of two-dimensional reinforced
concrete membrane structures. The program has been developed at University of Toronto since 1990,
when its original version was known as TRIX . This development has coincided with experimental tests
to corroborate the ability of VecTor2 to predict the load-deformation response of a variety of reinforced
concrete structures exhibiting well-distributed cracking when subject to short-term static monotonic,
cyclic and reverse cyclic loading.
The theoretical bases of VecTor2 are the Modified Compression Field Theory (Vecchio and
Collins, 1986) and the Disturbed Stress Field Model (Vecchio, 2000) analytical models for predicting
the response of reinforced concrete elements subject to in-plane normal and shear stresses. VecTor2
models cracked concrete as an orthotropic material with smeared, rotating cracks. The program utilizes an
incremental total load, iterative secant stiffness algorithm to produce an efficient and robust nonlinear
solution.
Originally, VecTor2 employed the constitutive relationships of the MCFT. Subsequent
developments have incorporated alternative constitutive models for a variety of second-order effects
including compression softening, tension stiffening, tension softening, and tension splitting. Also, the
capabilities of the VecTor2 have been augmented to model concrete expansion and confinement, cyclic
loading and hysteretic response, construction and loading chronology for repair applications, bond slip,
crack shear slip deformations, reinforcement dowel action, reinforcement buckling, and crack allocation
processes.
Finite element models constructed for VecTor2 use a fine mesh of low-powered elements. This
methodology has advantages of computational efficiency and numerical stability. It is also well suited to
reinforced concrete structures, which require a relatively fine mesh to model reinforcement detailing and
3
local crack patterns. The element library includes a three-node constant strain triangle, a four-node plane
stress rectangular element and a four-node quadrilateral element for modeling concrete with smeared
reinforcement; a two-node truss-bar for modeling discrete reinforcement; and a two-node link and a four-
node contact element for modeling bond-slip mechanisms.
VecTor2 reads ASCII text files for input and outputs binary and ASCII text files for analysis
results. The software Augustus provides graphical post-processing capabilities for the analysis results
of VecTor2.
1.3 Research Objectives
The objective of this research is to make NLFEARC with VecTor2 more amenable to practical
application for the non-specialist user. The strategy to meet these objectives is to develop documentation
for VecTor2 and a new preprocessor software program, called FormWorks .
The purpose of the program documentation is to provide guidance for the rational usage of VecTor2.
In a descriptive treatment, the documentation encompasses the following topics in relation to VecTor2:
the Modified Compression Field Theory and Disturbed Stress Field Model
the finite element formulation
the element library
the material models.
The FormWorks preprocessor software and its accompanying manual address the ease of use of
VecTor2 as an analytical tool, by decreasing the potential for input error and expediting the modeling
process. To fulfill this function, the FormWorks preprocessor includes the following amenities:
an instructive manual for using FormWorks and VecTor2
a graphical interface for data input and visualization in the Windows environment
input data checking and editing capabilities
a bandwidth reduction algorithm for improving computational efficiency
an automatic mesh-generator.
Specific criteria for the automatic mesh-generator are as follows:
4
accommodation of multiply connected regions and arbitrary material delineations
generation of mixed element-type meshes using the VecTor2 element library
preferential generation of isothetic rectangular elements for computational accuracy
good computational characteristics of the mesh
efficient and robust mesh generation algorithm
high degree of user control over mesh topology
aesthetically pleasing mesh topology for ease of result interpretation
economical demands on human input effort.
1.4 Organization
The report is subdivided into the following parts and chapters.
Part I, which includes Chapters 2 through 6, pertains to the program VecTor2. Chapter 2 presents
the theory and implementation of VecTor2. The Modified Compression Field Theory and Disturbed
Stress Field Model are described as the theoretical bases of the VecTor2 analysis. The chapter
subsequently discusses the implementation of these theories in the finite element algorithm and
incorporation of advanced analysis features. Chapter 3 presents the VecTor2 finite element library and the
role of elements in modeling reinforced concrete and reinforcement materials and bond-slip. Chapters 4, 5
and 6 describe constitutive and behavioral models pertaining to concrete materials, reinforcement
materials and bond slip, respectively.
Part II, which includes chapters 7 through 11, serve as a user manual for the preprocessor program
FormWorks. These chapters describe the user-interface and the finite element modeling process.
Chapter 12 summarizes this research and provides recommendations for further developments.
5
2 Part I: VecTor2
2.1 Introduction
This chapter begins with a review of the Modified Compression Field Theory and Disturbed Stress
Field Model the theoretical bases of VecTor2 analyses. The discussion subsequently focuses on the
finite element implementation of these theories for monotonically loaded structures. Finally, consideration
is given to the implementation of augmented analysis features for modeling lateral expansion, triaxial
stresses, cyclic loading, construction and loading chronology, and bond-slip.
2.2 Modified Compression Field Theory (MCFT)
The MCFT (Vecchio and Collins 1986) is an analytical model for predicting the load-deformation
response of reinforced concrete membrane elements subjected to shear and normal stresses, as shown in
Figure 1. The MCFT determines the average and local strains and stresses of the concrete and
reinforcement, and the widths and orientation of cracks throughout the load-deformation response of the
element. Based on this information, the failure mode of the element can also be determined.

Figure 1: Reinforced concrete membrane element subject to in-plane stresses

The MCFT models cracked concrete as an orthotropic material using a smeared rotating crack
approach. Cracked concrete is treated as a solid continuum with cracks distributed over the membrane
y
x

xy

x
,
y,
, f
y
,E
S
f
c
, f
t
, E
c
,
o

x
6
element, as opposed to a solid interrupted by discrete physical discontinuities. The smeared cracks freely
reorient, remaining coaxial with the changing direction of the principal concrete compressive stress field.
As well as being computationally convenient, the smeared rotating crack approach is consistent with the
distributed and meandering crack patterns observed in many reinforced concrete structures.
The theory is comprised of three sets of relationships: compatibility relationships for concrete and
reinforcement average strains; equilibrium relationships involving average stresses in the concrete and
reinforcement; and constitutive relationships for cracked concrete and reinforcement. The constitutive
relationships for cracked concrete result from tests of reinforced concrete panels using a purpose-built
Panel Element Tester at the University of Toronto. As such, the formulation of the MCFT incorporates
realistic constitutive models for concrete based on experimentally observed phenomena. While cracks are
smeared and the relationships are formulated in terms of average stresses and strains, a critical aspect of
the MCFT is the consideration of local strain and stress conditions at cracks.
2.2.1
2.2.2
Assumptions
The MCFT utilizes the following assumptions:
uniformly distributed reinforcement
uniformly distributed and rotating cracks
uniformly applied shear and normal stresses
unique stress state for each strain state, without consideration of strain history
strains and stresses are average over a distances including several cracks
orientations of principal strain,

, and orientations of principal stress,

, are the same

perfect bond between reinforcement and concrete
independent constitutive relationships for concrete and reinforcement
negligible shear stresses in reinforcement.
Compatibility Relationships
The compatibility relationships pertain to the average strains in the concrete and reinforcement
components as shown in Figure 2.
7
y

Figure 2: Average concrete strains due to average stress-strain response of concrete

With the perfect bond assumption, it follows that average strains experienced by the concrete are equally
experienced by the reinforcement. Therefore, for non-prestressed reinforcement, the average strains in the
concrete,
c
, and reinforcement,
s
, will be the same. Although the MCFT can address any number of
reinforcement components and orientations, consider the orthogonally reinforced membrane element of
Figure 1. The strains of the reinforcement parallel to the x and y directions are denoted by
sx
and
sy
,
respectively. The compatibility relationships are expressed by the following equations:
sy cx x
= = (2.2.2.1)
sy cy y
= = (2.2.2.2).
With a value for the shear strain,
xy
, relationships from Mohrs circle of strain determine the average
principal concrete tensile strain,
c1
, and the average principal concrete compressive strain,
c2
as follows:
( ) ( ) [ ]
2
1
2 2
2
1
2
1
2 1
,
xy y x y x c c
+ + = (2.2.2.3).
Relationships from Mohrs circle also determine the orientations of the average principal tensile strain
axis,

, and stress axis,

, with respect to the x-axis:

= = =

y x
xy
-1
2
1
tan (2.2.2.4).
2.2.3 Equilibrium Relationships
Consider the free body diagram of the membrane element, in Figure 3.
x

cx

cy
1
1

cxy
/2
1

c2

c1
8

Figure 3 Free body diagram of reinforced concrete element showing average stresses

Equilibrium of forces in the x and y-directions requires that the resultants of the applied normal stresses,

x
and
y
, be balanced by the resultants of the average concrete stresses, f
cx
and f
cy
, and the reinforcement
stresses f
sx
and f
sy
. Equilibrium of moments requires that the applied shear stresses,
xy,
are entirely resisted
by average shear stresses in the concrete, v
cxy,
(assuming the reinforcement does not exhibit dowel action).
These equilibrium relationships for average stresses may be summarized as follows:
sx sx cx x
f f + = (2.2.3.1)
sy sy cy y
f f + = (2.2.3.2)
cxy xy
v = (2.2.3.3)
where
sx
and
sy
are the reinforcement ratios in the x and y directions, respectively.
As cracked concrete is orthotropic with respect to the principal stress directions, Mohrs circle of
stress can be used to relate the average concrete stresses, f
cx
, and, f
cy
, to the average principal concrete
tensile stress, f
c1
:
( )

= 90 cot
1 cxy c cx
v f f (2.2.3.4)
( )

= 90 tan
1 cxy c cy
v f f (2.2.3.5).
y
y

xy
f
c1
f
cy
f
c2
x

x
f
sx
f
sy
f
cx
v
cxy
1

2
9
2.2.4 Constitutive Relationships
Constitutive models are necessary to relate the strains in the compatibility relationships with the
stresses in the equilibrium relationships. As described by Vecchio and Collins (1986), thirty panels
measuring 890x890x70mm were subject to in-plane stress conditions by the Panel Element Tester at the
University of Toronto. The test results were analyzed to develop constitutive models for cracked concrete
in compression and tension. It should be noted that other relationships have since been implemented in
VecTor2.
With regards to concrete in compression, the constitutive relationship relates the principal
compressive stress, f
c2
, to the principal compressive strain,
c2
. Panel test results indicate that the
compressive strength and stiffness decrease as coexisting principal tensile strains,
c1
, increase. This
phenomenon, known as compression softening, is incorporated by softening the stress-strain response of
concrete in uniaxial compression. The proposed relationship is as follows:
( ) ( ) [ ]
( )
o c
o c o c c
c
/ . .
/ - / f'
f


=
1
2
2 2
2
34 0 8 0
2
(2.2.4.1)
The term in the numerator is the Hognestad parabolic relationship for concrete in uniaxial compression,
often used for normal strength concretes. The value
o
is the concrete cylinder strain (a negative value)
corresponding to the peak compressive stress, f
c
, as determined from uniaxial compression tests of
concrete cylinders. The term in the denominator reflects the softening effect of principal tensile strains.
With regards to concrete in tension, the constitutive relationship relates the principal tensile
stress, f
c1
, to the principal tensile strain,
c1
. It is first necessary to determine the uniaxial cracking
strength, f
t
, and corresponding cracking strain,
cr
. In the absence of information, they may be estimated
as follows:
( ) MPa in f' . f'
c t
33 0 = (2.2.4.2)
c
t
cr
E
f'
= (2.2.4.3)
where E
c
is the initial tangent stiffness of the concrete, estimated as:
10
( MPa in f' E
c c
5000 = ) (2.2.4.4).
Before cracking, the concrete behaves linear-elastically in tension as follows:

cr c c c c
for E f < < =
1 1 1
0 (2.2.4.5).
After cracking, tensile stresses may continue to exist in the concrete between cracks in reinforced
concrete due to bond interactions between the concrete and reinforcement. To model this phenomenon,
known as tension stiffening, the concrete tensile stress decays from the tensile strength as the principal
concrete tensile strain increases. The MCFT proposed relationship is as follows:
1
1
200
c
t
c
1
f'
f
+
= (2.2.4.6).
As for the reinforcement in compression and tension, the MCFT uses a bilinear relationship
between the average stress, f
s
and average strain,
s
. An initial ascending linear-elastic branch is followed
by a yield plateau, as described by the following equations:
sxyield sx s sx
f E f = (2.2.4.7)
syyield sy s sy
f E f = (2.2.4.8)
where E
s
is the elastic modulus of the reinforcement, and f
sxyield
and f
syyield
are the yield stress of the
reinforcement in the x and y-directions, respectively.
2.2.5 Consideration of Local Crack Conditions
Given a compatible average strain condition, the preceding relationships can determine the
average stresses in the concrete and reinforcement and the applied shear and normal stresses that they
equilibrate. However, it would be unconservative to disregard the possibility that the element response is
governed by local yielding of the reinforcement at the crack or sliding shear failure along a crack. To
address these possibilities, the MCFT limits the local stresses at the crack and the average concrete tensile
stress.
Stresses fields in reinforced concrete vary from the average condition between cracks to the local
condition at the crack. Consider Figure 4a, which depicts the average stresses at a section between cracks
11
perpendicular to the principal tensile stress direction, and Figure 4b, which depicts the local stresses at the
free surface of the crack.
y

Figure 4 Comparison of average and local stresses at a crack
a) average stresses between crack b) local stresses at crack free surface

At a free surface of a crack, the average concrete tensile stresses diminish to virtually zero. To transmit
the average tensile stress across the crack, the reinforcement stress and strain must increase locally at the
crack. Static equivalency of the average and local stresses in the direction normal to the crack surface
results in the following equation:
( ) ( )
ny y s cry s y nx x s crx s x c
f f f f f
2 2
1
cos cos + = (2.2.5.1)
where f
scrx
and f
scry
are the local reinforcement stresses at a crack, and
nx
and
ny
are the angles between
the normal to the crack and the reinforcement. Considering the above equation, it is apparent that the
average tensile concrete stress is limited by the yielding of the reinforcement at the crack. If the
reinforcement yield strength is substituted for the local reinforcement stresses, the terms in parentheses
define the reserve capacity of the reinforcement, which limits the post-cracking concrete tensile stress as
follows:
( ) ( )
ny
2
y s yyield s y nx x s sxyield x c
f f f f f + cos cos
2
1
(2.2.5.2).
As a principal plane, shear stresses are absent from the section in Figure 4a. However, as the
reinforcement generally crosses the crack at a skew angle, local shear stresses, v
ci
, are present on the crack
y
x

xy
f
sx
f
c1

1
2
f
sy
a)

xy
f
scrx
v
ci

1
2
f
scry

ny

nx
b)
x
12
surface. Static equivalency of average and local stresses in the direction tangential to the crack determines
the local shear stresses as follows:
( ) ( )
ny ny y s cry s y nx nx x s crx s x ci
f f f f v + = sin cos sin cos (2.2.5.3).
Independently of the above equation, local shear stresses can only become so large before sliding shear
failure occurs. The shear stress is limited by aggregate interlock mechanisms, which decrease in efficacy
as the crack width, w, increases and the maximum aggregate size, a, decreases. Based on the analysis of
aggregate interlock by Walraven (1981), the MCFT limits the shear stress on the crack as follows:
( )
( ) MPa mm, in
a w/ .
f'
v
c
ci
26 24 31 0 + +
(2.2.5.4).
The average crack width, w, is the product of the principle concrete tensile strain and the average crack
spacing perpendicular to the crack, s

= s w
c1
(2.2.5.5)
my mx
s s
s

sin cos
1
(2.2.5.6).
The average crack spacing in the x-direction, s
mx
, and y-direction, s
my
, may be estimated from the bond
properties and layout of the reinforcement. Refer to Collins and Mitchell, (1997).
If either the maximum permitted average concrete tensile stress or local shear stress at a crack is
exceeded, then the strain state of the element is modified to result in a lower average concrete tensile
stress.
2.3 Disturbed Stress Field Model (DSFM)
The DSFM (Vecchio, 2000) addresses systematic deficiencies of the MCFT in predicting the
response of certain structures and loading scenarios. In lightly reinforced elements, where crack shear slip
is significant, the rotation of the principal stress field tends to lag the greater rotation of the principal
strain field. For such elements, the shear stiffness and strength is generally overestimated by the MCFT,
which assumes the rotations are equal. Conversely, in elements that exhibit limited rotation of the
13
principal stress and strain fields, the MCFT generally underestimates the shear stiffness and strength,
partly because the concrete compression response calibrated for the MCFT is overly softened for the
effect of principal tensile strains.
The DSFM is conceptually similar to the MCFT, but extends the MCFT in several respects. Most
importantly, the DSFM augments the compatibility relationships of the MCFT to include crack shear slip
deformations. The strains due to these deformations are distinguished from the strains of the concrete
continuum due to stress. As such, the DSFM decouples the orientation of the principal stress field from
that of the principal strain field, resulting in a smeared delayed rotating-crack model. Moreover, by
explicitly calculating crack slip deformations, the DSFM eliminates the crack shear check as required by
the MCFT. Constitutive relationships for concrete and reinforcement are also refined. The following
discussion presents the compatibility, equilibrium and constitutive relationships of the DSFM, with an
emphasis on differences from the MCFT.
2.3.1 Compatibility Relationships
While the MCFT assumes that principal strain and principal stress axes remain coaxial, panel
tests results indicate that this assumption is not always true after the cracking. The evidence demonstrates
the principal strain field generally changes inclination at a larger rate than the principal stress field,
resulting in a differential lag between the principal strain and principal stress axes.
The phenomenon is attributable to the manner in which the strain and stress fields are determined.
The measured strains are total strains, which are attributable to straining of the concrete continuum in
response to applied stresses as shown in Figure 2 and discontinuous shear slip as shown in Figure 5.
Meanwhile, concrete stresses are attributable only to the continuum straining in response to applied
stresses.
14
y

Figure 5: Deformation due to crack shear slip

To reconcile this behavior, the DSFM expresses the total strains,
x
,
y
, and
xy
, as the sum of net concrete
strains,
cx
,
cy
, and
cxy
, and strain due to shear slip, , and :
s
x

s
y

s
xy

s
x cx x
+ = (2.3.1.1)
s
y cy y
+ = (2.3.1.2)
s
xy cxy xy
+ = (2.3.1.3).
Relationships from Mohrs circle determine the principal net concrete tensile strain,
c1,
and the
principal net concrete compressive strain,
c2
, as follows:
( ) ( ) [ ]
2
1
2 2
2
1
2
1
2 1
,
cxy cy cx cy cx c c
+ + = (2.3.1.4).
The crack slip shear strain components, , and , are calculated from the average crack
slip shear strain
s
x

s
y

s
xy

s
. This quantity is defined as the crack slip,
s
, divided by the average crack spacing, s,
as follows:
s
s
s

= (2.3.1.5).
Relationships from Mohrs circle resolve
s
into the components
s
x
,
s
y
and
s
xy
:
( ) = 2 sin
2
1
s
s
x
(2.3.1.6)
( ) = 2 sin
2
1
s
s
y
(2.3.1.7)
x

1
2
s

s
w
15
( ) = 2 cos
s
s
xy
(2.3.1.8).
The orientation of the principal net concrete strains, , and orientation of principal concrete stresses,

,
with respect to the x-axis are determined from Mohrs circle relationships involving the net concrete
strain components as follows:

= =

cy cx
cxy
-1
2
1
tan (2.3.1.9).
Likewise, the orientation of the principal total strain field,

, is determined from the total strain

components:

y x
xy
-1
2
1
tan (2.3.1.10).
The difference between the orientation of total strains and the orientation of the principal concrete
stresses defines the rotation lag, :

= (2.3.1.11).
Although the DSFM can address any number of reinforcement components and orientations,
consider the orthogonally reinforced membrane element of Figure 1. Assuming perfect bond, the average
strains of the reinforcement components in the x- and y-directions are equal to the total strains:
x sx
= (2.3.1.12)
y sy
= (2.3.1.13).
2.3.2 Equilibrium Relationships
Again, consider the orthogonally reinforced membrane element of Figure 1. The average stress
equilibrium relationships of the DSFM are the same as those of the MCFT. They are summarized as
follows:
sx sx cx x
f f + = (2.3.2.1)
sy sy cy y
f f + = (2.3.2.2)
cxy xy
v = (2.3.2.3).
16
Additionally, the DSFM incorporates the equilibrium relationships for local stresses at the crack.
( ) ( )
ny y s cry s y nx x s crx s x c
f f f f f + =
2 2
1
cos cos (2.3.2.4)
( ) ( )
ny ny y s cry s y nx nx x s crx s x ci
f f f f v + = sin cos sin cos (2.3.2.5).
Equilibrium of stress resultants normal to the crack surface results in the following equations:
( ) ( )
ny y s cry s y nx x s crx s x c1
f f f f f + =
2 2
cos cos (2.3.2.6)
( ) ( )
ny ny y s cry s y nx nx x s crx s x ci
f f f f v + = sin cos sin cos (2.3.2.7).
The average concrete tensile stress is subject to the limits of the yield strength of the reinforcement
traversing the crack:
( ) ( )
ny y s yyield s y nx x s sxyield x c
f f f f f +
2 2
1
cos cos (2.3.2.8).
Unlike the MCFT, however, the tensile stress is not subject to the limitation of shear stresses at a crack,
since the DSFM explicitly incorporates deformations due to shear slip rather than ascribing a limiting
stress corresponding to shear slip failure.
2.3.3 Constitutive Relationships
Between the development of the MCFT and the DSFM, the constitutive models for cracked
concrete were revised and refined. Although a variety of alternative models are available, the following
discussion presents the models of the DSFM.
As previously discussed, the response of concrete in compression depends on both the principal
compressive strain and coexisting principal tensile strain. Analyses of additional test panels (Vecchio and
Collins, 1993) led to the reduction factor,
d
, to reflect the softening effect of the coexisting principal
tensile strains:
0 1
1
1
.
C C
d s
d

+
= (2.3.3.1).
The factor, C
d
, accounts for the softening effect of transverse tensile strains:
(
8 0
2 1
28 0 35 0
.
c c d
. . C = ) (2.3.3.2).
17
The factor C
s
recognizes whether or not the analysis accounts for element slip deformations. If the
analysis couples the inherent softening effect of shear slippage with the softening effect due to tensile
strains (as in the MCFT), then C
s
=1.0. Conversely, if the analysis considers elements slip distortion, as
in the DSFM, then the softening effect appears to be less for the same value of
c1
/
c2
since the softening
effect is attributable only to the tensile strains. In this case, C
s
=0.55.
To soften the compressive strength and stiffness of the concrete, the concrete cylinder strength,
f
c
, and corresponding peak strain,
o
, are both reduced by the reduction factor to determine the peak
compressive stress, f
p
, and corresponding peak strain
p
:
c d p
f' f = (2.3.3.3)
o d p
= (2.3.3.4).
Using the softened parameters, the following relationship determines the relationship between the
principal concrete compressive stress, f
c2
, and the principal net compressive strain,
c2
:
( )
( ) ( )
nk
p c
p c
p c
n
n
f f
+

=
2
2
2
1
(2.3.3.5)
where
( ) MPa in f - . n
p
17 80 0 = (2.3.3.6)

< <
< <
=
0 62 67 0
0 0 1
2
2
p c p
c p
for / f - .
for .
k (2.3.3.7).
With regards to concrete in tension, the response before cracking is linear-elastic, as follows:
cr c c c c
for E f =
1 1 1
0 (2.3.3.8).
For cracked concrete, average concrete tensile stresses, , due to tension stiffening, can be generally be
modeled by the nonlinearly decaying relationship (but must not exceed the bound imposed by equation
2.3.2.6):
a
c
f
1
1
1
1
1
c cr
c t
t a
c
for
c
f'
f <
+
= (2.3.3.9).
18
The coefficient, c
t
, proposed by Bentz (1999) incorporates the influence of reinforcement bond
characteristics and is computed as follows:
m . c
t
2 2 = (2.3.3.10)
( mm in |
d

m
i i
i
n
n
1 i b
i
| cos
4 1
=

=
) (2.3.3.11).
where d
bi
is the bar diameter and
i
is the reinforcement ratio of each of the n reinforcement components.
In addition to tension stiffening, post-cracking tensile stresses, f
c1
b
, arise in plain concrete due to
fracture mechanisms a phenomenon known as tension softening. Tension softening effects may be
significant in lightly reinforced concrete structures. The cracking tensile stress due to tension softening
may be calculated by the following linear relationship:
( )
( )
ts c cr
cr ts
cr c
t
b
c
for f' f < <

=
1
1
1
1 (2.3.3.12).
The terminal strain,
ts
, (the strain at which tensile stresses in plain concrete reduce to zero) is determined
from the fracture energy parameter, G
f
, (the area under the stress-strain curve of plain concrete, assumed
to be 75 N/m), and the characteristic length, L
r
(assumed to be half the crack spacing):
r t
f
ts
L f'
G
.

= 0 2 (2.3.3.13).
Finally, the post-cracking principal tensile stress in the concrete is taken as the larger of the
values predicted by the tension stiffening and tension softening phenomena:
( )
b
c
a
c c
f , f max f
1 1 1
= (2.3.3.14).
The DSFM constitutive model for reinforcement in tension or compression is trilinear to account
for strain-hardening phenomenon:
( )

<
< < +
< <
< <
=
s u
u s sh sh s sh syield
sh s syield syield
syield s s s
s
for
for E f
for f
for E
f
0
0
(2.3.3.15).
19
where E
s
is the elastic modulus of the reinforcement, f
syield
is the yield strength of the reinforcement,
syield

is the yield strain of the reinforcement,
sh
is the strain at the onset of strain hardening,
u
is the ultimate
strain of the reinforcement.
2.3.4
)
Shear Slip Relationships
Having defined the compatibility, equilibrium and constitutive relationships, it remains necessary to
model the crack slip,
s
, to determine the crack slip shear strain,
s
. One approach is to relate the crack slip
to the local shear stresses, v
ci
, at the crack. Based on the analysis of aggregate interlock by Walraven
(1981), the following relationship may be used to determine the crack-slip and shear slip strain:
(
cc
. 0.8
ci a
s
f . w . w .
v
+
=

20 0 234 0 8 1
707 0
(2.3.4.1)
s
a
s a
s

= (2.3.4.2).
where f
cc
is the concrete cube strength, w is the average crack spacing and s is the average crack spacing.
However, the above approach is problematic in two respects. First, the equilibrium equation
2.3.2.5 predicts that the shear stress at a crack, and hence the crack slip, are always zero for unreinforced
elements. This is equivalent to ignoring crack shear stresses arising from aggregate interlock the sliding
friction between the exposed aggregate and cement paste at the crack and the plastic deformation of the
cement paste due to contact stresses. Secondly, the slip relationship does not account for the initial crack
slip that occurs before contact areas develop between the rough crack surfaces.
To address these deficiencies, a second approach for modeling the shear slip is to specify a
constant rotation lag,
l
, between the inclination of the principal total strain axis,

and the inclination of

the principal stress axis,

. Indeed, results of panel tests indicate that the lag is established soon after
cracking, and generally falls in the range of 5 to 10 until the yielding of a reinforcement component,
whereupon the lag increases. To implement this approach, it is necessary to define the post-cracking
rotation,

, of the principal total strain axis, relative to the orientation of the principal strains and
stresses at initial cracking,
ic
:
20
ic
=

(2.3.4.3).
The post-cracking rotation,

, of the principal stress field is then related to

, by the constant rotation

lag is follows:
( )

>

=

for
for
l l
l
| |
| |
(2.3.4.4).
The orientation of the principal stress field is determined as the sum its orientation at initial cracking, and
its post-cracking rotation:

+ =
ic
(2.3.4.5).
Finally, relationships from Mohrs circle determine the shear slip strain, :
b
s

( )

+ = 2 sin
x y xy
b
s
cos2 (2.3.4.6).
Further, it is possible to combine the two approaches in a hybrid model for shear slip. The shear
slip strain is computed by both approaches and the larger of values is utilized:
( )
b
s
a
s s
max = , (2.3.4.7).
When the concrete element is unreinforced or when the local shear stress on the crack is small, the
constant rotation lag governs the shear slip, reflecting the initial slip occurring prior to development of
shear stresses at a crack. Conversely, when the shear stresses on the crack are large, the shear slip is
predicted by the stress-based formulation. As such, using the both approaches collectively is consistent
with the actual determinants of the shear slip.
2.4 Finite Element Implementation
Displacement-based finite element methods for structural analysis result in a system of equations
relating unknown nodal displacements to specified forces by the structure stiffness matrix. The VecTor2
algorithm for nonlinear finite element analysis is summarized by the flow chart in Figure 6. The following
discussion describes the details of some of these steps.
21
Input analysis control data
Update strain and stress
parameters
Determine element
stiffness matrix, [k]

Figure 6 VecTor2 nonlinear finite element analysis algorithm

Determine average
concrete and
reinforcement stresses
f
c1
, f
c2
, f
si
Determine local concrete
and reinforcement stresses
at cracks
v
ci
, f
scri
Determine crack slip
strain components, [
s
]
Determine material
component stiffness
matrices
[D
c
], [D
s
]
Determine composite
material stiffness matrix
[D]=[D
c
]+[D
s
]
i
Determine pseudo nodal
load vector [F*]
Input structure data
Input external load data
Update external loads for
load stage, [F]
[ ]
o
c
[ ]
p
c
[ ]
i
o
s
[ ]
i
p
s

Determine element strain

offset components
[
s
]
Determine structure
stiffness matrix
[K]=[k]
Determine total load
vector [F]=[F]+[F*]
Determine nodal
displacements
[r]=[K]
-1
[F]
Determine element strains
[],[
c
], [
s
]
Determine element
stresses, []
Update secant moduli

si c c c
E G E ' , ' , ' , '
2 1
E
[ ] [ ] [ ]
s
c
p
c
o
c
r r , , r [ ] [ ]
p
s
o
s
r r ,
Determine strain offset
displacements

Determine secant moduli

i s c c c E G E E , , , 2 1
Proceed to next
load stage?
Analysis complete
Secant moduli converged?
Output load stage results
Steps to be performed for
each element
22
2.4.1 Composite Material Stiffness Matrix
In the most general case, total strains, [] =[
x
,
y
,
xy
]
T
, are comprised of net concrete strains
, elastic strain offsets [ ]
c
[ ]
o
c
(due to thermal, prestrains, shrinkage and lateral expansion effects),
plastic strain offsets in the concrete, [ ]
p
c
, (due to cyclic loading or damage), and strains due to crack
shear slip, [ ]
s
(as considered by the DSFM).
[ ] [ ] [ ] [ ] [ ]
s p
c
o
c c
+ + + = (2.4.1.1).
As well, compatibility relationships determine that the strain in the i
th
smeared reinforcement
component is the sum of the total strain, elastic strain offsets [ ]
i
o
s
(due to thermal and prestrain effects)
and plastic strain offsets [ ]
i
p
s

(due to cyclic loading or damage):
[ ] [ ] [ ] [ ]
i
p
s i
o
s i s
+ + = (2.4.1.2).
At any point within the reinforced concrete continuum, the total strains are related to stresses []
by the composite material stiffness matrix, [D], as follows:
[ ] [ ][ ] [ ] D
o
= (2.4.1.3).
The composite material stiffness matrix is the sum of the concrete material stiffness matrix, [D
c
], and the
reinforcement component material stiffness matrices, [D
s
]
i
, as follows
[ ] [ ] [ ]

=
+ =
n
1 i
i s c
D D D (2.4.1.4).
While the composite material stiffness matrix operates on total strains, element stresses can be directly
related only to net strains of the concrete and reinforcement. Therefore, it is necessary to subtract the
stress contribution of strain offsets and shear slip strains by use of the pseudo stress vector [
o
] calculated
as follows:
[ ] [ ] [ ] [ ] [ ] { } [ ] [ ] [ ] {
i
p
s i
o
s
n
1 i
i s
s p
c
o
c c
o
D D + + + + =

=
} (2.4.1.5).


23
As the MCFT and DSFM model the reinforced concrete as an orthotropic material in the principal
stress directions, it is necessary to formulate the concrete material stiffness matrix,[D
c
], relative to these
directions. If it is assumed that the Poissons effect is negligible, then [D
c
] is computed as follows:
[ ]

=
c
c
c
c
G
E
E
' D
0 0
0 0
0 0
2
1
(2.4.1.6).
The secant moduli , , , 2 1 c c c G E E as shown in Figure 7, are computed from the current values of the
principal stresses, f
c1
and f
c2,
and the corresponding principal net concrete strains,
c1
and
c2
, as follows:.
1
1
1
c
c
c
f
E

= ;
2
2
2
c
c
c
f
E

= ;
2 1
2 1
c c
c c
c
E E
E E
G
+

= (2.4.1.7).
Likewise, material stiffness matrices [D
s
]
i
for each reinforcement component must first be
determined relative to their longitudinal axes. As the reinforcement is assumed only to resist uniaxial
stresses, [D
s
]
i
is computed as follows:
[ ]

=
0 0 0
0 0 0
0 0 si
i
i s
E
' D (2.4.1.8)
where
i
is the reinforcement ratio of the reinforcement component. The secant modulus si E , as shown in
Figure 7, is computed from its current value of stress, f
si
and the corresponding strain,
si
as follows:
si
si
si
f
E

= (2.4.1.9).

24
f
c f
s

Figure 7: Definition of secant moduli for a) concrete b) reinforcement

The material stiffness matrices, [D
c
] and [D
s
]
i,
are transformed from their respective principal
axes to the x,y axes by means of the transformation matrix, [T], as follows:
[ ] [ ] [ ] [ ]
i c i c
T
i c i c
T D T D ' = (2.4.1.10)
[ ] [ ] [ ] [ ]
i s i s
T
i s i s
T D T D ' = (2.4.1.11)
[ ]
( )

=
2 2
2 2
2 2
sin cos sin cos sin cos
sin cos cos sin
sin cos sin cos
2 2 -
- T (2.4.1.12).
For the concrete, the angle is the inclination of the principal tensile stress axis,

, with respect
to the positive x-axis. For the reinforcement, the angle is the orientation,
i
, of each reinforcement
component, with respect to the positive x-axis.
2.4.2 Element Stiffness Matrices
The element stiffness matrix, [k] relates nodal forces to nodal displacements of the element. It is
determined from the composite material stiffness matrix as follows:
[ ] [ ] [ ][ ]

=
vol
T
dV B D B k (2.4.2.1).
The strain-displacement matrix [B], interpolates strains throughout the element continuum by operating
on nodal displacements of the element. The form of the strain-displacement matrix depends on the type of

ci
E
ci
1
f
ci f
si

si
s p
c
o
c
+ +
1
p
s
o
s
+
E
s
(a) (b)
25
the element, and the resulting value of the above integration will depend on the composite material
stiffness matrix, the element geometry and the exactness of the integration method.
It is also possible to separate the element stiffness matrix into contributions from the stiffness of
the concrete, [k
c
], and stiffness of the reinforcement components, [k
s
]
i
, by substituting the respective
material stiffness matrix for the composite material stiffness matrix:
[ ] [ ] [ ][ ]

=
vol
c
T
c
dV B D B k (2.4.2.2)
[ ] [ ] [ ] [ ]

=
vol
i s
T
i s
dV B D B k (2.4.2.3).
2.4.3 Pseudo Nodal Loads
The element stiffness matrix operates on total nodal displacements in a manner that is analogous
to the material stiffness matrix operating on total strains. Like the use of the pseudo-stress vector, it is
necessary to subtract the nodal force contributions of the nodal displacements due to strain offsets and
crack shear slip by the use of a pseudo nodal load vector.
First, the offset and crack shear slip strains are resolved into components relative to the x,y axes.
For the concrete:
[ ]

=
0
o
c
o
c
o
c
; ; (2.4.3.1) [ ]

=
p
cxy
p
cy
p
cx
p
c
[ ]

=
s
xy
s
y
s
x
s
c
For the reinforcement components:
[ ]
( )
( )

+
=
i
o
si
i
o
si
i
o
si
i
o
s
/
/
2 sin
2 2 cos 1
2 2 cos 1
; [ ]
( )
(

+
=
i
p
si
i
p
si
i
p
si
i
p
s

/
/
2 sin
2 2 cos 1
2 2 cos 1
) (2.4.3.2)
where
i
is the orientation of the reinforcement component relative to positive x-axis.
Integration of these strain components results in free nodal displacements[ ] [ ] [ ]
s
c
p
c
o
c
r r r , , , due to
concrete offset strains and shear slip, and displacements [ ] [ ]
p
s
o
s
r r , due to reinforcement strain offsets, as
follows:
26
[ ] [ ]dA r
o
c
o
c

= ; ; [ ] [ ]dA r
p
c
p
c

= [ ] [ ]dA r
s s
c

= (2.4.3.3)
[ ] [ ]dA r
o
s
o
s

= ; (2.4.3.4).
The pseudo nodal loads, [F*], required to produce equivalent nodal displacements for each
element are determined by multiplying the free nodal displacements by the contributions of the concrete
and reinforcement to the element stiffness matrix:
[ ] [ ]dA r
p
s
p
s

=
[ ] [ ] [ ] [ ] [ ] { } [ ] [ ] [ ] { }

=
+ + + + =
n
1 i
p
s
o
s i s
s
c
p
c
o
c c
r r k r r r k * F (2.4.3.5).
The pseudo nodal loads are added to the applied external loads, [F], to determine the total nodal
force vector, [F]:
[ ] [ ] [ ] * F F F' + = (2.4.3.6).
2.4.4 Total Load Iterative Secant Stiffness Solution
The global stiffness matrix, [K], of the entire structure is assembled by the summation of all the
element stiffness matrices:
[ ] [ ]

=
=
m
1 i
i
k K (2.4.4.1).
Having determined the global stiffness matrix and the total nodal load vector, the following
system of equations results involving the unknown nodal displacements, [r]:
[ ] [ ][ ] r K F' = (2.4.4.2).
Providing the structure is adequately restrained, the unknown nodal displacements are determined
by pre-multiplying total nodal load vector by the inverse of the global stiffness matrix:
[ ] [ ] [ ] F' K r
1
= (2.4.4.3).
The total element strains are determined from the nodal displacements, utilizing the strain-
displacement matrix.
[ ] [ ][ ] r B = (2.4.4.4).
27
The shear and normal stresses acting upon the element are determined by multiplying the element
strains by the material stiffness matrices and subtracting the pseudo stress vector:
[ ] [ ] [ ] [ ] [ ] [ ] { } [ ] [ ] [ ] [ ] { } D D
o
p
o
s s
s p
c
o
c c
+ = (2.4.4.5).
However, the secant moduli, E , used to determine the structure stiffness matrix are only
estimates that are initially based on the stress-strain state of the previous load step. Owing to the finite
size of each load step, the stress-strain state corresponding to the current load step differs from that of the
previous load step by a finite amount, resulting in an error in the computed nodal displacements. The error
can be made arbitrarily small by an iterative refining the secant moduli. The secant moduli are
recomputed as ' E for the calculated stress-strain state, and then averaged with the assumed values E as
follows:
( )
i
1) (i ' E c E c E
i
+ = + 1 (2.4.4.6).
where c is a specified convergence averaging factor and i is the iteration number. When the secant moduli
have acceptably converged or a specified number of iterations have been performed, VecTor2 concludes
the analysis for the load step, stores the analysis results and proceeds to the next load step, if any.
2.5 Augmented Analysis Features
While the MCFT and DSFM form the basis for VecTor2, the program has been augmented to
model the effects of lateral expansion, triaxial stresses, cyclic and reverse cyclic loading, construction and
loading chronology, and bond slip. Each of these features is subsequently described.
2.5.1 Modeling of Lateral Expansion Effect
When reinforced concrete is uncracked or experiences relatively small tensile strains, lateral
expansion may account for a significant portion of the total strains in the principal maximum strain
direction. If these strains are incorrectly attributed to tensile straining due to stress, the compression
softening effect may be overestimated. To address this potential inaccuracy, Vecchio (1992) implements
concrete lateral expansion effects by modifying the formulation of the concrete material stiffness matrix.
28
For a linear-elastic material that is orthotropic in the principal directions, plane stresses
are related to plane strains [
T

12 2 1
] [ ]
T

12 2 1
, by the material stiffness matrix, [D]:
[ ]
( )

=
12 21 12
2 2 21
1 12 1
21 12
1 0 0
0
0
1
1
G v v
E E v
E v E
v v
D (2.5.1.1)
where, v
12
is the Poisson ratio relating
1
to
2
, and v
21
is the Poisson ratio relating
2
to
1
. The number of
independent constants is reduced if the shear modulus, G
c
, is approximated as follows:
( ) ( )
21 2 12 1
2 1
12
1 1 v E v E
E E
G
+ + +
(2.5.1.2).
Further, the reciprocity requires that the strain energy of the element be independent of the sequence of
loading. It follows that the remaining four constants are related as follows:
2 21 1 12
E v E v = (2.5.1.3).
As the above terms are the off-diagonal elements of the [D], the above condition is necessary for the
symmetry of [D]. If it is assumed that the Poissons ratio is zero, and the elastic moduli are replaced by
secant moduli for concrete, the concrete material stiffness matrix of equation 2.4.1.5 results.
The response of concrete violates the above condition, such that the material stiffness matrix
becomes unsymmetrical. As concrete is subjected to increasing compressive stress, progressive internal
micro-cracking accelerates the rate of lateral expansion. The same mechanism is responsible for the
softening of the compression stress-strain response. Therefore, it is generally true for f
c2
<f
c1
<0, that
v
12
>v
21
, E
c1
>E
c2
and .
2 21 1 12 c c
E v E v
This difficulty is circumvented by separating the lateral expansion effects into elastic and residual
components. The elastic component is modeled by a symmetric concrete material stiffness matrix, while
the residual component is modeled by elastic strain offsets, [ ]
o
c
, as described in the preceding section.
For the purpose of illustration, consider an element subject to biaxial compression such that f
c2
<f
c1
<0.
The strains in the principal directions due to the Poissons effect are as follows:
2
2
12 1
c
c
c
E
f
v - = (2.5.1.4)
29
1
1
21 2
c
c
c
E
f
v - = (2.5.1.5).
For the reasons stated above, suppose that the lateral expansion,
c2
, is smaller than the lateral expansion

c1
. The lateral expansion
c2
is modeled entirely by the elastic component, while the lateral expansion
c1

is modeled by both residual and elastic components. The elastic component of the Poissons ratio, , is
computed by the reciprocity condition:
e
v
12
1
2
21 12
c
c
e
E
E
v v = (2.5.1.6).
The elastic components of the lateral expansion effects are implemented in the symmetric concrete
material stiffness matrix, [D
c
] as follows:
[ ]
( )

=
12
21 12
2 2
21
1
12
1
21 12
1 0 0
0
0
1
1
c
e
c c
c
e
c
e
c
G v v
E E v
E v E
v v
D (2.5.1.7)
where,
( ) ( )
21
2
12
1
2 1
12
1 1 v E v E
E E
G
c
e
c
c c
c
+ + +
(2.5.1.8).
The residual component of the Poissons ratio, , is the remaining portion of the expansion:
*
v
12
e *
v v v
12 12 12
= (2.5.1.9).
The residual component of the lateral expansion in the 1-direction is computed as an elastic strain offset
using the residual component of the Poissons ratio:
2
2 *
12 1
c
c o
c
E
f
v - = (2.5.1.10).
Models for determining the value of Poissons ratio are described in subsequent sections.
2.5.2 Modeling of Triaxial Stresses
Although the MCFT is formulated for the plane stress state, VecTor2 accounts for out-of-plane
stresses in the z-direction due to confinement of lateral expansion by out-of-plane reinforcement. The
30
triaxial stress state is then utilized in computing the strength enhancement effects due to confinement. The
out-of-plane concrete strain is computed as follows:

+
=
1
1
21
2
2
12
c
c
c
c
sz z c
c
cz
E
f
v
E
f
v -
E E
E
(2.5.2.1a)
where
z
is the reinforcement ratio of the out-of plane reinforcement. If however, the out-of-plane
reinforcement has yielded, the out-of-plane concrete strain is computed as follows:

1
1
21
2
2
12
c
c
c
c
c
yield z, z
cz
E
f
v
E
f
v -
E
f
-

= (2.5.2.1b).
The stress, f
sz
, in the out-of-plane reinforcement is determined as follows:
(2.5.2.2)
yield z cz s sz
f E f
,
=
where f
z,yield
is the yield strength of the out-of-plane reinforcement.
The resulting out-of-plane concrete compressive stress, f
cz
,, is determined from equilibrium as follows:
(2.5.2.3).
sz z cz
f f =
2.5.3 Analysis of Cyclically Loaded Structures
A simplifying assumption of the MCFT is ignoring the effect of strain history in determining the
stress-strain response of the concrete and reinforcement components. Therefore, the constitutive
relationships described so far correlate a unique stress value for any strain value. While this is adequate
for monotonic loading, it is not adequate for cyclic loading due to the hysteretic response of concrete and
reinforcement. When subjected to load reversals, these materials exhibit non-coincident stress-strain
responses, which give rise to plastic strain offsets. Moreover, the hysteretic responses depend upon the
plastic strain offset and maximum strains and stresses previously experienced by the materials. The
hysteretic response of concrete and reinforcement are described as material models in subsequent
sections. The following discussion describes how VecTor2 records the plastic strains and strain envelopes
for the concrete.
When considering constitutive relationships for concrete, the pertinent strain directions are those
parallel to the principal net concrete strains. However, these axes rotate as the load changes. As such,
31
VecTor2 records concrete plastic strains and strain envelopes with respect to the x,y axes and transforms
them to and from arbitrary orientations of the principal axes using relationships from Mohrs circle.
If the inclination of the principal net concrete strains is , then the following relationships
transform plastic strains, , with respect to the X,Y axes to plastic strains, , , with
respect to the principal 1,2 directions:
p
cx

p
cy,

p
cxy,

p
c1

p
c2

+

+
+
= 2 sin 2 cos
2 2
1
p
cx
p
cy
p
cx
p
cy
p
cx
p
c
(2.5.3.1)

+
= 2 sin 2 cos
2 2
2
p
cx
p
cy
p
cx
p
cy
p
cx
p
c
(2.5.3.2).
In a given load step, the plastic strains change by the quantities and , with respect to
the principal 1 and 2 directions, respectively. The following relationships transform the plastic strain
increments and add them to the previous plastic strains to update the plastic strains with respect to the x,y
axes:
p
c1

p
c2

( ) (

+ +

+ = 2 cos 1
2
2 cos 1
2
'
2 1
p
c
p
c p
cx
p
cx
) (2.5.3.3)
( ) ( ) +

+

+ = 2 cos 1
2
2 cos 1
2
'
p
c2
p
c2 p
cy
p
cy
(2.5.3.4)
+ = 2 sin 2 sin '
p
c2
p
c1
p
cxy
p
cxy
(2.5.3.5).
Maximum concrete strains are defined by maximum compressive strains, , and ,
and maximum tensile strains , and , with respect to the X, Y directions. For brevity, the
following discussion describes the treatment of the compressive strains only, but similar relationships
result for the tensile strains. Although , and describe a compressive strain envelope, rather
than a strain state, analogous relationships from Mohrs circle transform them to and from the principal
and global axes.
cmx

cmy,

cmxy

tmx

tmy,

tmxy

cmx

cmy,

cmxy

32
The following relationships transform the maximum compressive strains, , and ,
with respect to the x,y axes to maximum compressive strains, strains
cmx

cmy,

cmxy

cm1
and
cm2
, with respect to the
principal 1,2 directions:
+

+
+
= 2 sin 2 cos
2 2
1 cmx
cmy cmx cmy cmx
cm
(2.5.3.6)

+
= 2 sin 2 cos
2 2
2 cmx
cmy cmx cmy cmx
cm
(2.5.3.7).
In a given load step, the calculated principal net concrete strains are
c1
and
c2.
If these principal
strains are more compressive than those described by
cm1
and
cm2
, then the envelope of compressive
strains must be updated. The incremental maximum compressive strains in the principal directions are
defined as follows:

<
>
=
1 1 1 1
1 1
1
0
cm c cm c
cm c
cm
if
if
(2.5.3.8)

<
>
=
2 2 2 2
2 2
2
0
cm c cm c
cm c
cm
if
if
(2.5.3.9).
The following relationships transform the maximum compressive strain increments and add them to the
previous maximum compressive strains to update the maximum compressive strains with respect to the
x,y axes:
( ) ( )

+ +

+ = 2 cos 1
2
2 cos 1
2
'
2 1 cm cm
cmx cmx
(2.5.3.10)
( ) ( ) +

+

+ = 2 cos 1
2
2 cos 1
2
'
2 1 cm cm
cmy cmy
(2.5.3.11)
+ = 2 sin 2 sin '
2 1 cm cm cmxy cmxy
(2.5.3.12).
Collectively, the plastic strains and strain envelopes and other stress envelope quantities allow
VecTor2 to implement the hysteretic response models for the concrete. Likewise, VecTor2 records the
strain and stress history of the reinforcement (without the additional complexity introduced by the
transformations), to implement hysteretic response models for the reinforcement. The plastic strains
33
resulting from the hysteretic response of both material components are treated by the pseudo nodal load
approach described in the preceding section.
2.5.4 Analysis of Construction and Loading Chronology
Repaired reinforced concrete structures present an example where construction and loading
chronologies significantly influence the load-deformation behavior. Initial loading of the original
structure results in distress to the concrete and reinforcement materials. At some later time, possibly while
still in service, the structure may be strengthened by casting of additional concrete, addition of reinforcing
bars or the lay up of FRP composite materials. After repair, entirely new loads may be applied to the
repaired portions of the structure. The effectiveness of the repair procedure depends upon the load sharing
between the original and newly added portions of the structure. In turn, this depends not only upon the
final configuration of the structure and loads, but also the extent of damage prior to repair and the strain
differentials between the original material and repair materials at the time of repair.
VecTor2 models construction and loading chronology by allowing elements to be engaged and
disengaged during loading. Engaged elements represent portions of the structure that are currently
present. They contribute to the strength and stiffness of the structure. Conversely, disengaged elements
represent portions the structure that are currently absent. They do not contribute to the strength and
stiffness of the structure. In cases where regions of a structure will be replaced by repair materials,
engaged and disengaged elements occupy the same space in the mesh, resulting in a double meshed
region.
At a given load stage, the total strains of disengaged elements, [], are compatible with strains of
the adjoining engaged elements. However, when disengaged elements are activated at a later load stage,
the previously experienced total strains are recorded entirely as plastic strain offset strains, [ ]
p
c
and [ ]
p
s
.
Therefore, the newly engaged elements effectively behave as though they have been added in a zero
elastic strain condition to previously engaged elements, which may be already distressed to some degree.
At each load stage, the strain history of the elements are updated and recorded as necessary.
34
2.5.5 Bond-Slip Mechanisms
Bond action is the means by which stress transfers between the concrete and reinforcement and
allows the two materials to behave compositely. As the concrete adjacent to reinforcement deforms,
resisting bond stresses act tangentially to the reinforcement. In the case of embedded bar reinforcement,
the bond stresses arise from chemical adhesion, friction and mechanical interlock with the concrete. In the
case of externally bonded reinforcements, the bond stresses arise from the adhesive properties and shear
resistance of materials adjoining the concrete and reinforcement. Bond action is critical in anchorage
zones and near cracks, where high stress gradients in the reinforcement necessitate large bond stresses.
Perfect bond may exist in regions where bond stresses are sufficiently low. In this case, the
change in reinforcement strain in response to loading is equal to the change in the total concrete strain. At
higher levels of bond stress, damage to the bond results in imperfect bond action. Consequently,
differential straining of the concrete and reinforcement occurs. This manifests as relative displacements or
bond slip between the concrete and reinforcement. Among other factors, the bond stress is related to the
extent of bond slip.
For many reinforced concrete structures subject to monotonic loading, proper reinforcement
detailing ensures that the strength capacity is not governed by failure of bond action. The limited localized
bond slip has minimal influence on the load-deformation response. For these structures, the perfect bond
assumption of the MCFT is acceptable for modeling their behavior. In contrast, structures that are
susceptible to bond failures may exhibit substantial bond slip that modifies the internal stress distribution
and load-deformation response. As well, cyclically loaded structures experience a gradual degradation of
bond that may significantly influence the hysteretic response.
VecTor2 models bond-slip mechanisms by use of the bond elements and bond stress-slip curves
for embedded reinforcement and externally bonded reinforcement. The bond elements serve as
deformable interfaces between concrete and discrete reinforcement elements, thereby removing the
perfect bond relationship between the concrete and reinforcement materials. The stress-slip response of
the bond elements is governed by the bond stress-slip curve. The bond stress-slip curves depend upon
material properties, reinforcement layout and confining pressures.
35
3 Element Library
The VecTor2 element library is subdivided into three element categories depending on the type of
material the element models. The first category consists of planar triangular, rectangular and quadrilateral
elements which model concrete with or without smeared reinforcement. The second category consists of
the linear truss bar element which models discrete reinforcement. The third category consists of the non-
dimensional link and contact elements, which model bond-slip.
All elements are simple and low powered, with minimal nodes, straight conforming boundaries,
and linear displacement functions. Compared to higher powered elements, these lower powered elements
are advantageous in two respects. First, VecTor2 explicitly calculates their stiffness coefficients without
resorting to numerical integration. Second, low powered elements are not susceptible to spurious
behaviors such as zero-energy modes. The following discussion describes the element type.
3.1 Reinforced Concrete Elements
Reinforced concrete elements are used to model plain concrete or concrete with smeared
reinforcement, representing regions of a structure having well-distributed reinforcement. Owing to
compatibility between the concrete and smeared reinforcement comprising the element, the concrete and
reinforcement are perfectly bonded. The element stiffness matrices, [k], are formulated with the
composite material stiffness matrices by closed form evaluation of the following integral:
[ ] [ ] [ ][ ]

=
vol
T
dV B D B k (3.1.1).
3.1.1 Constant Strain Triangle
The constant strain triangle, shown in Figure 8, is a three-noded element with uniform thickness,
t. The element is defined by the three node numbers, in counterclockwise sequence i,j,k. As each node
displaces in the x and y directions, the element has a total of six degrees of freedom. Although the triangle
36
may assume any orientation and shape, its accuracy degrades as the shape deviates from an equilateral
triangle. As such, distorted and elongated triangles should be avoided.

Figure 8: Constant strain triangle element

The displacement of any point in the element is expressed as a linear combination of the x and y
coordinates. The element strains, [], are related to the nodal displacements, [r], by the strain-
displacement matrix [B] as follows:
[ ] [ ][ ] r = (3.1.1.1)
where
y
k
(3.1.1.2) [ ] [
T
xy xy x
= ]
[ ] [ ]
T
ky kx jy jx iy ix
r r r r r r r = (3.1.1.3)
[ ]

=
ij ji ki ik jk kj
ji ik kj
ij ki jk
y x y x y x
x x x
y y y
2A
1
B 0 0 0
0 0 0
(3.1.1.4)
j i ij j i ij
y - y y and x - x x = = (3.1.1.5).
Considering the terms of [B], it is apparent that the strains and stresses are constant throughout the
element. Consequently, the constant strain triangle behaves poorly in bending and a finer mesh is
generally is required to achieve the same accuracy as a mesh constructed with plane stress rectangles. For
these reasons, use of the constant strain triangle should be limited to accommodating edges that are not
parallel to the X or Y axes, and to make transitions in element size.
i
j
x
thickness =t
r
iy
r
ix
37
3.1.2
]
]
Plane Stress Rectangle
The plane stress rectangle, shown in Figure 9, is a four-noded element with uniform thickness, t.
The element is defined by four node numbers in counterclockwise sequence, i,,j,m,n. As each node
translates in the x and y directions, the element has a total of eight degrees of freedom. The rectangle must
be oriented isothetically, with its edges parallel to the x and y axes. Although the rectangle may assume
any width and height, its accuracy degrades as the shape deviates from a square. As such, rectangles with
aspect ratios exceeding 3:2 should be avoided.

Figure 9: Plane stress rectangle element

The displacement of any point in the element is expressed as a bilinear combination of the x and y
coordinates, including the mixed xy term. The element strains, [], are related to the nodal
displacements, [r], by the strain-displacement matrix, [B], as follows:
[ ] [ ][ ] r = (3.1.2.1)
[ ] [
T
xy xy x
= (3.1.2.2)
[ ] [
T
ny nx my mx jy jx iy ix
r r r r r r r r r = (3.1.2.3)
[ ]
( ) ( ) ( ) ( )
( ) ( ) ( ) ( )
( ) ( ) ( ) ( ) ( ) ( ) ( ) (

+ + + +
+ +
+ +
=
y b x a y b x a y b x a y b x a
x a x a x a x a
y b y b y b y b
ab
B 0 0 0 0
0 0 0 0
4
1
)
(3.1.2.4).
Considering the terms of [B], it is apparent that the plane stress rectangle can represent linear gradients of
strains and stress across its width and height. Evaluation of strains at the center of the rectangular element
i
j
m
y
x
thickness =t
2a
n
2b
r
iy
r
ix
38
avoids parasitic shear. Given that the plane stress rectangle is less overly stiff than the constant strain
triangle, the plane stress rectangle is preferable for modeling reinforced concrete regions.
3.1.3 Quadrilateral
The quadrilateral, shown in Figure 10a, is a four-noded element with uniform thickness, t. The
element is defined by four node numbers in counterclockwise sequence, i, j, m, n. As each node translates
in the X and Y directions, the element has a total of eight degrees of freedom. The quadrilateral may
assume any orientation and shape in the X, Y coordinate system. For accurate results, the quadrilaterals
should be compact with approximately equal interior angles, and lacking in excessive skew, taper or
warp.

Figure 10: a) Quadrilateral element b) Decomposition of quadrilateral element into two constant strain
triangle elements

VecTor2 divides the quadrilateral element into two constant strain triangles T
A
and T
B
, sharing the
shortest diagonal as a common edge, as shown in Figure 10b. VecTor2 analyzes the two triangular
elements separately, each with their own stiffness matrix. Having solved for the nodal displacements,
VecTor2 computes the strains in the quadrilateral element as the area average of the strains in each
triangular element. For the same reasons that apply to the constant strain triangles, the plane stress
rectangle is preferable to the quadrilateral element. Use of the quadrilateral element should be limited to
accommodating edges that are not parallel to the x or y axes, and to make transitions in element size.
i
j
m
x
y
thickness =t
n
i
j
m
y
n
T
B
T
A
thickness =t x
(a)
(b)
39
3.2 Reinforcement Elements
Reinforcement elements are used to model reinforcement bars or FRP layers, or a lumped
collection of them in close proximity relative to the surrounding element sizes. The stiffness matrix, [k],
of these elements are formulated with the reinforcement material stiffness matrix only. Use of the discrete
reinforcement elements is required when either the behavior of individual bars or bond-slip mechanisms
are of interest. Generally, as a matter of accuracy and computational economy, primary reinforcement is
modeled by discrete reinforcement elements whereas well-distributed reinforcement is modeled by
smeared reinforcement in the reinforced concrete elements.
3.2.1 Truss Bar
The truss bar, shown in Figure 11, is a two-noded element with uniform cross-sectional area, A.
The element is defined by two node numbers, i, j. As each node displaces in the x and y directions, the
element has a total of four degrees of freedom. The truss element may assume any orientation in the x, y
coordinate system.

Figure 11: Truss bar element
The truss bar exhibits resistance only to elongation along its axis. The following equations relate
the nodal forces, [F], and nodal displacements, [r], with the truss bar element stiffness matrix [k]:
[ ] [ ][ ] r k F = (3.2.1.1)
where
[ ] [ ]
T
jy jx iy ix
F F F F F = (3.2.2.2)
i
j
y

thickness =t x
40
[ ] [
T
jy jx iy ix
r r r r r = ] (3.2.2.3)
[ ]

=
2 2
2 2
2 2
2 2
sin sin cos sin sin cos
sin cos cos sin cos cos
sin sin cos sin sin cos
sin cos cos sin cos cos
-
-
- -
L
AE
k (3.2.3.4).
3.3 Bond Slip Elements
Without bond-slip elements, the reinforced concrete elements and discrete reinforcement elements
are defined by common nodes. As such, the displacement of these elements is compatible, representing a
condition of perfect bond between the concrete and discrete reinforcement elements.
Bond-slip elements serve as deformable interfaces between reinforced concrete elements and
discrete reinforcement elements. Prior to slippage, bond-slip elements are defined by paired nodes having
the same coordinates. A reinforced concrete element is attached to one of the nodes, and a discrete
reinforcement element is attached to the other node. Increases of bond stresses are accompanied by bond
slip, manifested by relative displacement of the paired nodes and the attached reinforced concrete and
discrete reinforcement elements. VecTor2 includes two bond-slip elements: the link element and the
contact element.
The accuracies of the link and contact element depend upon the variation in slip between
successive nodes of the same discrete reinforcement elements. One measure of accuracy is the
discrepancy between the theoretical strain energy and strain energy of bond elements to model given slip
variations. As discussed by Keuser and Mehlhorn (1987), link elements model constant bond slips
exactly, but provide poor accuracy for non-constant slip variations. Contact elements with a linear
displacement function model constant and linear slip variations exactly, and provide good approximations
of nonlinear slip variations. Nevertheless, it has been found that in practice (Gan, 2000, Wong, 2001) that
both elements types provide similar and good accuracy when the mesh is sufficiently fine.
41
3.3.1 Link
The link element, (Ngo and Scordelis, 1967), shown in Figure 12, is a two-noded, non-
dimensional element. The element is defined by two different node numbers, i and j. Prior to slippage, the
paired nodes must have the same coordinate. One node must be incident to a reinforced concrete element,
and the other node must be incident to a discrete reinforcement element. As each node displaces in the x
and y directions, the element has a total of four degrees of freedom.


Figure 12 Link element

The link element may be conceptualized as two orthogonal springs, linking the reinforced
concrete element and the discrete reinforcement element. One spring deforms tangentially to the discrete
reinforcement element, representing bond slip and bond stresses. The other spring deforms radially to the
discrete reinforcement element, representing radial displacements and stresses.
The nodal displacements of the elements in the x,y coordinate system, [r], are transformed to
deformations of the tangential spring (the bond slip),
t
, and the deformation of the radial spring,
r
, by
the coordinate transformation matrix [T]:
[ ] [ ][ ] r T = (3.3.1.1)
where,
[ ] [ ]
T
r t
= (3.3.1.2)
i,j
y
concrete element
(e.g. quadrilateral)
truss bar
x
link element
(nondimensional)
j
i
k
t
r
iy

r
ix
k
r
42
[ ] [
T
jy jx iy ix
r r r r r = ]
3.3.2
(3.3.1.3)
[ ]

=
cos sin cos sin
sin cos sin cos
T (3.3.1.4).
The force in the tangential spring, F
t
, is found by multiplying the bond slip,
t
, by the stiffness, k
t
,
and the bonded surface area tributary to the link, A. Likewise, the force in the radial spring, F
r
, is found
by multiplying the deformation of the radial spring,
r
, by the stiffness k
r
, and the bonded surface areas,
A, of embedded bars or external plates tributary to the link:
(3.3.1.5).

r
t
r
t
r
t
k
k
A
F
F
0
0
The stiffness, k
t
, relates to the bond stress to the bond slip and is determined from the bond-slip curve. In
VecTor2, k
r
is assigned a large value of 100k
t
, to suppress the radial displacement of reinforcement
element.
The nodal forces in the x,y directions, [F], are determined by transforming the spring forces with
transpose of the transformation matrix, [T]
T
.
[ ]

r
t T
jy
jx
iy
ix
F
F
T
F
F
F
F
(3.3.1.6).
Substituting equations 3.3.1.1 and 3.3.1.2 in equation 3.3.1.3, provides the stiffness relationship in the x,y
coordinate system with the link element stiffness matrix [k]:
[ ] [ ][ ] r k F = (3.3.1.7)
[ ] [ ] [ ] T
k
k
A T k
r
t
T

=
0
0
(3.3.1.8).
Contact
The contact element, shown in Figure 13, is a four-noded element, having only a linear
dimension. The element is defined by four different node numbers in the sequence, j, k, m, n. Prior to
slippage, nodes, j and k must have the same coordinates. Either node j or node k must be attached to a
43
reinforced concrete element. The other node must be attached to a discrete reinforcement element. The
same condition applies to the nodes m and n. As each node displaces in the x and y directions, the element
has a total of eight degrees of freedom. The contact element may assume any orientation in the x,y plane.

Figure 13: Contact element

The contact element models a continuous interface along the length of the discrete reinforcement
element. With two node pairs defining the contact interface, the displacement of any point along the
contact element is linearly interpolated from the nodal displacements to ensure conformal deformations
with the reinforced concrete element and discrete reinforcement element. At any point, the bond stress, ,
is related to the tangential displacement,
t
, by the stiffness, k
t
; the radial stress, , is related to the radial
displacement by the stiffness, k
r
, as follows:

r
t
r
t
k
k
0
0
(3.3.2.1).
The stiffness, k
t
, relates to the bond stress to the bond slip and is determined from the bond-slip curve. In
VecTor2, k
r
is assigned a large value of 100k
t
, to suppress the radial displacement of reinforcement
element.
By minimizing of the potential energy of a loaded element, the following equilibrium relationship
is found relating the nodal forces [F] and displacements [r] in the x,y coordinate system:
[ ] [ ][ ] r k F = (3.3.2.2)
j,k
y
concrete element
(e.g. quadrilateral)
truss bar
x
contact element
j

r
mx
r
my
m,n
k
m
t
r
n
44
where
[ ] [
T
ny nx my mx jy jx iy ix
F F F F F F F F F = ]
]
(3.3.2.3)
[ ] [
T
ny nx my mx jy jx iy ix
r r r r r r r r r = (3.3.2.4).
The contact element stiffness matrix, [k], in the x,y coordinate system is found by transforming the
contact element stiffness matrix, [k], in the local coordinate system with the transformation matrix [T]:
[ ] [ ] [ ] [ ] T k T k
T
' = (3.3.2.5)
where
[ ]

=
r r r r
t t t t
r r r r
t t t t
r r r r
t t t t
r r r r
t t t t
k k - k k
k k - k k
k k k - k
k 0 k k k
k k k k -
k k - k k
k - k k - k
k k k k
6
A
k
2 0 2 0 0 0
0 2 0 2 0 0
2 0 2 0 2 0 0
0 2 2 0 0
0 0 2 0 2 0
0 0 0 2 0 2
0 0 2 0 2 0
0 0 0 2 0 2
' (3.3.2.6)
[ ]

=
cos sin 0 0 0 0 0 0
sin cos 0 0 0 0 0 0
0 0 cos sin 0 0 0 0
0 0 sin cos 0 0 0 0
0 0 0 0 cos sin 0 0
0 0 0 0 sin cos 0 0
0 0 0 0 0 0 cos sin
0 0 0 0 0 0 sin cos
-
-
-
T (3.3.2.7)
and A is the bonded surface area tributary to the contact element.
45
4 Models for Concrete Materials
The concrete constitutive and behavioral models are paramount to the accuracy of the VecTor2
results. At each load step, the structure stiffness is determined from the stresses and strains calculated
from the constitutive models. Further, the inclusion or omission of a model determines whether the effects
of pertinent behaviors are included in the analysis, and the suitability of the results to the purpose of the
analysis. Many of the models in VecTor2 include multiple options differing in degree and type, which
may produce a divergence of results.
Regarding the constitutive relationships, VecTor2 utilizes Cauchy-type models, which describe the
concrete response via nonlinear functions of stress and strain. This approach is amenable to concrete
given that the combined behavior of aggregates, cement and reinforcement which that can often only be
described by empirical relationships. These relationships typically involve mechanical properties
determined from standard specimens under specific stress and strain conditions, rather than being inherent
material properties. Bearing this in mind, it is necessary to select models judiciously for each analysis.
The following discussion describes the constitutive and behavioral models pertaining primarily to
the response of the concrete material, although many models must be discussed in the context of
reinforced concrete. First, the compressive stress-strain response is discussed followed by the associated
models of compression softening. Second, the tensile stress-strain response is discussed followed by the
associated models of tension softening, and tension splitting. Thirdly, confinement and lateral expansion
models are described. Fourthly, the crack slip check, crack width and slip distortion models are
considered. Finally, hysteretic response models are presented.
4.1 Compression Pre-Peak Response
The stress-strain response of concrete in uniaxial compression is nonlinear beyond low
compressive stresses, despite the fact that the constituent cement paste and aggregates exhibit linear-
elastic behavior in compression. The apparent contradiction is explained by the softening effect of internal
microcracks that form as a result of stress concentrations at the interface of the cement paste and
46
aggregates. The ascending branch of the nonlinear stress-strain response is the subject of the subsequent
discussion.
Compression pre-peak response models compute the principal compressive stress, f
ci
, if the
compressive principal strain,
ci
, is less compressive than the strain,
p
, corresponding to the peak
compressive stress, f
p
. The peak parameters are determined by adjusting the unconfined uniaxial concrete
cylinder strength, f
c
, and the corresponding strain,
o
, for compression softening due to transverse tensile
strains, and strength enhancement due to confinement. Some of the following compression pre-peak
response models are described by functions that extended into the post-peak range. In these cases, both
the ascending and descending branches are presented.
4.1.1 Linear
The linear model, shown in Figure 14, is an elastic-plastic compression response curve, which is
not normally used.

< <
< <

=
0
0
p ci p
ci p p
p
ci
ci
for f -
for f
f (4.1.1.1).
-
c
-f
c
f
p

p

Figure 14: Linear pre- and post-peak concrete compression response
47
4.1.2 Popovics
Popovics (1973) presented stress-strain curves for a range of normal strength concretes. As shown
in Figure 15, these curves reflect the greater stiffness and linearity of the ascending branch and the
reduced ductility of concretes as the peak compressive stress increases.
The stress-strain curve is given by the following equation:
( )
0
1
<
+

=
ci
n
p ci
p
p
ci
ci
for
- n
n
f f (4.1.2.1).
The long fraction represents the deviation from linear-elastic response. The curve fitting parameter, n,
captures the greater linearity of higher strength concrete through the diminishing difference between the
initial tangent stiffness E
c
, and secant stiffness, E
sec
. These values are computed as follows:
sec c
c
E - E
E
n = (4.1.2.2).
p
p
sec
f
E

= (4.1.2.3)

-
c
-f
c
fp =50 MPa
fp =40 MPa
fp =30 MPa
fp =20 MPa

Figure 15 Popovics pre- and post-peak concrete compression response

48
4.1.3 Hognestad (Parabola)
The Hognestad parabola, as shown in Figure 16, is a simple compression response curve, suitable
for normal concrete strengths (<40 MPa).
-
c
-f
c
f
p

p

Figure 16: Hognestad parabolic pre- and post-peak concrete compression response

The stress-strain curve is described by the following relationship:
0 0 2
2
< <

=
ci
p
ci
p
ci
p ci
for f f (4.1.3.1).
The stress-strain relationship is symmetric about
p
, diminishing to zero stress at zero strain and 2
p
. Note
that the Hognestad parabola predefines the initial tangent stiffness, E
c
, as follows:
p p c
f E = 2 (4.1.3.2).
4.1.4 Popovics (High Strength)
Collins and Porasz modified the stress-strain curve proposed by Thorenfeldt, Tomaszewicz, and
J ensen (1987) and Popovics (1973) to accommodate the behavior of high strength concrete in
compression. Experimental studies demonstrate that as the concrete strength increases, the response is
linear to a greater percentage of the maximum compressive stress, the strain corresponding to the peak
compressive stress increases, and the descending branch of the stress-strain curve declines more steeply.
Also, intermediate high strength concretes exhibit a decreased ultimate compressive strain. The Popovics
49
High Strength response curve, as shown in Figure 17, primarily differs from the Popovics response
curve in the more rapid post-peak stress decay for higher strength concretes.
The stress-strain curve is given by the following equation:
( )
0
1
<
+

=
ci
nk
p ci
p
p
ci
ci
for
- n
n
f f (4.1.4.1).
The long fraction represents the deviation from linear-elastic response. The curve fitting parameter, n,
captures the greater linearity of higher strength concrete through the diminishing difference between the
initial tangent stiffness E
c
, and secant stiffness E
sec
. It is given by the following equation:
( ) in MPa f
f
. n
p
p
17
80 0 + = (4.1.4.2).
The parameter, k, increases the post-peak decay in stress and is calculated as follows:
( )

MPa in f for .
f'
.
for .
k
p p ci
p
ci p

< < +
< <
=
0 0 1
62
67 0
0 0 1
(4.1.4.3).
Note that the Popovics High Strength relationship predefines the initial tangent stiffness, E
c
, as follows:
1 - n
n
f
E
p
p
c

= (4.1.4.4).

50
-
c
-f
c
fp = 70 MPa
fp =60 MPa
fp =50 MPa
fp =40 MPa
fp =30 MPa
fp =20 MPa

Figure 17: Popovics high strength pre- and post-peak concrete compression response
4.1.5 Hoshikuma et al.
In a study of confinement effects of reinforced concrete bridge piers with a variety of cross
sectional shapes, hoop reinforcement spacing and volumetric ratio, Hoshikuma, et al. (1997) proposed a
stress-strain curve for concrete in compression, as shown in Figure 18. Experimental investigations
indicated that the peak compressive stress, f
p
, and corresponding strain,
p
, depend upon the amount of
hoop reinforcement, whereas the initial stiffness, E
c
, does not. In this context, it is noted that Hognestad
parabolic is deficient in that the initial stiffness is predetermined by equation. 4.1.3.2 in terms of the peak
parameters, and is therefore an implicit function of the amount hoop reinforcement. To reconcile this
inconsistency, the following relationship was proposed for the ascending branch of the compression
stress-strain curve
0
1
1
1
< <

=

ci p
n
p
ci
ci c ci
for
n
E f (4.1.5.1).
The term in parentheses represents the deviation from linear-elastic response. The parameter, n, is
expressed in term of the initial tangent stiffness E
c
, and secant stiffness E
sec
, as follows:
51
sec c
c
E - E
E
n = (4.1.5.2)
p
p
sec
f
E

= (4.1.5.3).

-
c
-f
c
fp = 70 MPa
fp =60 MPa
fp =50 MPa
fp =40 MPa
fp =30 MPa
fp =20 MPa

Figure 18: Hoshikuma et al. concrete compressive pre-peak response
The Hoshikuma et al. post-peak response is considered in the subsequent sections.
4.2 Compression Post-Peak Response
Beyond the peak compressive stress, f
p
, and corresponding strain,
p
, concrete resists substantial
compressive stress under continued compressive straining. Moreover, confinement of the concrete by
transverse stresses enhances the strength and ductility of the concrete, transforming the failure in
compression from brittle to ductile. This residual compressive strength and ductility may allow localized
regions of a reinforced concrete structure to fail, but gradually unload so as to redistribute internal stresses
and forestall total failure of the structure until additional deformation occurs. The effect of such post-peak
behavior may be beneficial by allowing greater economy in design, or detrimental, should overstrength of
the desired failure mode result in an undesirable failure mode. Post-peak behavior may also be significant
52
in over-reinforced structures. As such, the selection of the compression post-peak response is significant
to a realistic analysis of the load-deformation response.
Compression post-peak response models compute the principal compressive stress, f
ci
, if the
compressive principal compressive strain,
ci
, is more compressive than the strain,
p
, corresponding to the
peak compressive stress, f
p
. The peak parameters are determined by adjusting the unconfined uniaxial
concrete cylinder strength, f
c
, and the corresponding strain,
o
, for compression softening due to
transverse tensile strains, and strength enhancement due to confinement.
Of the following models presented, the Modified Park-Kent, Popovics/Mander, Hoshikuma et al.
and Saenz/Spacone models are formulated in the context of confined concrete. If the Modified Park-Kent,
Popovics/Mander, or Hoshikuma et al. models are selected and the concrete is not sufficiently confined,
then an alternative formulation is necessary to compute the compression post-peak response. (The
Saenz/Spacone is exempt from the following computation, owing to its inclusion of a control point on the
post-peak stress-strain curve.) For these post-peak response models, the compression post-peak stress, f
ci
,
is computed as follows:
( )
b
ci
a
ci ci
f c f c f + = 1 (4.2.1)
where
1 0 , 4

= c
f'
f' f
c
c
c p
(4.2.2)
- exp f - f
p
ci
p
ci
p
a
ci

= 1 (4.2.3)
) (
ci
b
ci
func f = , as computed in the subsequent sections for confined concrete.
The stress computed as is the Smith-Young post-peak branch for unconfined concrete, as shown in
Figure 19. The parameter, c, increases linearly from zero to one as f
a
ci
f
p
increases from f
c
to 1.25f
c
. When
used in equation 4.2.1, c effects a linear transition from the unconfined to confined compression post-peak
response.
53
-
c
-f
c

p
f
p
a
ci
f

Figure 19: Smith-Young concrete compression post-peak response for f
p
<= f
c


In addition, when the Modified Park-Kent, Popovics/Mander, Hoshikuma et al. and Saenz/Spacone
models are selected, the post-peak response is assumed to have a sustaining branch equal to 0.2f
p
,
whenever the concrete is sufficiently confined such that f
p
exceeds f
c
.
( )
p c p ci ci
f f' 0 if f . - func f < < < = 2 0 (4.2.4).
4.2.1
4.2.2
Pre-Peak Base Curve
The Pre-Peak Base Curve option is a valid selection for the compression post-peak response if the
Linear, Popovics, Hognestad (Parabola) or Popovics (High Strength) compression pre-peak response
model is selected. The compression post-peak stress is computed using the equations presented in the
previous section for the descending branch of the selected stress-strain curve.
Modified Park-Kent
Park, Priestly and Gill (1982) modified a stress-strain curve proposed by Kent and Park to
account for the enhancement of concrete strength and ductility due to confinement. The stress-strain
curve, as shown in Figure 20, was utilized to compute the flexural strength of reinforced concrete
54
columns confined by transverse hoop reinforcement. The linearly descending branch of the modified
stress-strain curve is liberally adapted for VecTor2 as follows:
[ ] 0 2 0 0 ) ( < < < + =
p ci p p ci p m p
b
ci
for f . - or f Z f f (4.2.2.1)
where
( MPa in f and f'
f
. f
f
Z
lat c
p
lat
o
c
c
m
+

+
=
9 . 0
170 002 0 1000 ' 145
' 29 . 0 3
5 . 0
) (4.2.2.2)
and f
lat
, is summation of principal stresses, acting transversely to the direction under consideration:
2 1 0
3 2 1
or i f - f f f f
ci c c c lat
= + + = (4.2.4.3).
-
c
-f
c
f
p

p
0.2f'
c
Z
m
1

Figure 20: Modified Park-Kent post-peak concrete compression response

4.2.3 Popovics / Mander
Mander, Priestley and Park (1988), proposed a compression stress-strain curve for confined
concrete with transverse hoop reinforcement. The form of stress-strain curve is the same as that proposed
by Popovics (1973), except that the initial tangent stiffness, E
c
, is assigned a particular value as follows:
( )
0 2 0
1
< < <
+

=
p ci p
n
p ci
p
p
ci b
ci
for f .
- n
n
f f (4.2.3.1)
55
where
sec c
c
E - E
E
n = (4.2.3.3)
p
p
sec
f
E

= (4.2.3.2)
c c
f' E 5000 = (4.2.3.4).
4.2.4 Hoshikuma et al.
In a study of confinement effects of reinforced concrete bridge piers with a variety of cross
sectional shapes, and transverse hoop reinforcement spacing and volumetric ratios, Hoshikuma et al.
(1997) proposed a stress-strain curve for concrete in compression. Based on experimental investigations,
the descending branch of the compression stress-strain curve, as shown in Figure 21, is idealized as a
straight line given as follows:
( ) [ ] 0 for f E f f
p ci p p ci des p
b
ci
< < + = 2 . 0 (4.2.4.1)
where E
des
is the deterioration rate of the descending branch. According to a regression analysis of
experimental post-peak stress-strain curves for compressive stresses less than |0.5f
p
|, Hoshikuma et al.
proposed the following inverse relation between E
des
and
s
f
yh
/f
c
2
, where
s
is the volumetric ratio of
hoop reinforcement, f
yh
is the yield strength of hoop reinforcement and ,f
c
is the concrete cylinder
strength:
2
'
2 . 11
c yh s
des
f f
E

= (4.2.4.2).
In VecTor2, the confining pressure,
s
f
yh
, provided by the yielding of the hoop reinforcement is replaced
by the mean lateral confining pressure, f
lat
/2, and the deterioration rate is limited to half the initial tangent
modulus as follows:
( )
c
c lat
des
E
f f
E 5 . 0
' 2
2 . 11
2
= (4.2.4.3)
where f
lat
, is summation of principal stresses, acting transversely to the direction under consideration:
56
2 1 0 or i f - f f f f
ci c3 c2 c1 lat
= + + = (4.2.4.4).
-
c
-f
c
f
p

p
0.2f'
c
E
des
1

Figure 21: Hoshikuma et al. concrete compressive post-peak response
4.2.5 Saenz / Spacone
Saenz (1964) noted that higher strength concretes exhibit more rapidly descending compression
post-peak responses. To reflect the shape of the descending branch, Saenz proposed that the compression
stress-strain curve pass through a post-peak control point strain,
r
, and corresponding stress, f
r
, as well as
satisfy stress and stiffness boundary conditions at zero stress and the peak compressive stress. The general
form of such a curve, shown in Figure 22, is given as follows:

C B A
f K
f
p
ci
p
ci
p
ci
p
ci
p
ci
3 2
1

= (4.2.5.1).
The parameters, A, B, C are expressed in terms of the stiffness ratio, strain ratio, and stress ratio,
K

, as follows:
2 + = K C A (4.2.5.2)
C B 2 1 = (4.2.5.3)
57
( )
( )

=
K
K
K
K C
1
1
1
2
(4.2.5.4)
where the stiffness ratio, K, compares the initial tangent stiffness, E
c
, to the secant stiffness, E
sec
:

E
E
K
c
sec
= (4.2.5.5)

f
E
p
p

=
sec
(4.2.5.6).
The strain ratio, K

, compares the post-peak control point strain,

r
, and the strain corresponding to the
peak stress,
p
:
K
p
r

(4.2.5.7)
and the stress ratio, K

, compares the peak compressive stress, f

p
, to the post-peak control point stress, f
r
:

f
f
K
r
p
=

(4.2.5.8).
It remains necessary to specify the post-peak control point strain,
r
, and corresponding stress, f
r
.
Kwan and Spacone (2002) utilize the following empirical equation to determine the stress f
r
, which
increases as the confinement effect increases:
p
p c
c p
r
f .
f f
f f
f 4 1
' 5
'

= (4.2.5.9).
The strain,
r
, is computed as four times the strain corresponding to peak compressive stress:
p r
= 4 (4.2.5.10).
58
-
c
-f
c
f
p

p
0.2f'
c

Figure 22: Saenz/Spacone concrete compressive post-peak response
4.3 Compression Softening
Compression softening in cracked concrete is the reduction of compressive strength and stiffness,
relative to the uniaxial compressive strength, due to coexisting transverse cracking and tensile straining.
This reduction can be substantial and have considerable effects on the load-deformation response of
reinforced concrete structures, in terms of stiffness, ultimate strength capacity and ductility.
In VecTor2, the compression softening is effected by a softening parameter,
d
, with a value
between zero and one, which is calculated by the compression softening models. These models are
determined by statistical analysis of the stress-strain states of selected panel elements (890x890x70mm)
and shell elements (1450x1450x350mm) tested at the University of Toronto (Vecchio and Collins, 1992).
Depending on how the models calculate and apply
d
, the following compression softening models may
be classified into two types: strength-and-strained softened and strength-only softened models.
Strength-and-strain softened models, as shown in Figure 23, use
d
to reduce both the uniaxial
compressive strength, f
c
, and corresponding strain,
0
, to determine the peak compressive strength, f
p
, and
corresponding strain,
p
, used in the compression response models. The value of
d
is a function of
c1/
/
c2

59
the ratio of the principal tensile strain to the principal compressive strain. In general, the calculation and
application of strength-and-strained softened models is as follows:

=
2
1
c
c
d
func (4.3.1)
c l d p
f' f = (4.3.2)
o l d p
= (4.3.3a).
(The parameter,
l
, accounts for strength enhancement from confinement effects.)
However, if the compression softening is significant enough that
o l d p
= is less compressive than

o
, then a modification must be made to the calculation of
p
so that the compression response ascends for
strains up to
o
and descends thereafter. When the compressive strain,
c2
, is more compressive than
,
o l d

p
, is calculated as follows:

< < <

< < <
=
0
0
2
2 2
o l d o c o
o l d c o c
p
if
if
(4.3.3b).
Further, for strains more compressive than
o
, the softened response is computed by simply multiplying
the base response by
d
.
-
c
-f
c

f'
c
f
p

p
o

d
f
cbase

Figure 23: Strength and strain-softened compression response

60
Strength-only softened models, as shown in Figure 24, use
d
to reduce only the uniaxial
compressive strength, f
c
, to determine the peak compressive strength, f
p
, while the corresponding
strain,
0,
, is unmodified for compression softening. The value of
d
is a function of
c1/
/
0
the ratio of
the principal tensile strain to the strain corresponding to the uniaxial compressive stress. As
c2
is not
considered as a variable, strength-only softened models are more amenable for manual calculation, but
marginally weaker in correlation with experimental results and in accuracy.

=
o
c
d
func
1
(4.3.4)
c l d p
f' f = (4.3.5)
o l p
= (4.3.6).
-
c
-f
c

o
f'
c
f
p

Figure 24: Strength-only softened compression response

It is assumed in the following discussion that the principal compressive strain
c2
and the strain
corresponding to peak stress,
o
, are negative values while the principal tensile strain,

c1
, is a positive
value.
4.3.1 No compression softening
The compression response is independent of the coexisting tensile strains:
61
1 =
d
(4.3.1.1).
4.3.2 Vecchio 1992-A (e1/e2-Form)
The strength-and-strained softened model, shown in Figure 25, is based on the results of 116
panel and shell element tests. The ratio of the principal tensile strain to principal compressive stress is
limited to 400 to avoid overestimation of the softening effect when the principal tensile strains are very
large (e.g. when the reinforcement has yielded). The factor C
s
recognizes whether shear slip deformations
are considered (see discussion on DSFT). Both the uniaxial compressive strength and corresponding
strain are softened. The model was originally developed for the Popovics (High Strength) compression
stress-strain curve.
1
1
1

+
=
C C
d s
d
(4.3.2.1)
( )

>
<
=
r if r
r if
C
d
28 . 0 28 . 0 35 . 0
28 . 0 0
80 . 0
(4.3.2.2)
400
2
1

= r
c
c
(4.3.2.3)

=
considered slip shear if
considered not slip shear if
C
s
55 . 0
0
(4.3.2.4)
c d p
f' f = (4.3.2.5)
o d p
= (4.3.2.6).
62

0
0.2
0.4
0.6
0.8
1
0 5 10 15
-
c1/

c2

d
Cs=1.0
Cs=0.55

Figure 25: Vecchio 1992-A compression softening model
4.3.3 Vecchio 1992-B (e1/e0-Form)
The model is the strength-only softened version of the Vecchio 1992-A model. The factor C
s

recognizes whether shear slip deformations are considered (see discussion on DSFT). The model was
originally developed for the Popovics (High Strength) compression stress-strain curve.
1
1
1

+
=
C C
d s
d
(4.3.3.1)
C C
d
o
c
d
50 0 , 37 . 0 27 . 0
1

= (4.3.3.2)

=
considered slip shear if
considered not slip shear if
C
s
55 . 0
0
(4.3.3.3)
c d p
f' f = (4.3.3.4)
o p
= (4.3.3.5).
4.3.4 Vecchio-Collins 1982
The strength-and-strain softened model is based on the results of thirty panel tests. The model
was originally developed for the Hognestad Parabola compression stress-strain curve. Note that this
model does not recognize the softening effects of shear slip when they are included in the analysis.
63
1
27 . 0 85 . 0
1
2
1

=
c
c
d
(4.3.4.1)
c d p
f' f = (4.3.4.2)
o d p
= (4.3.4.3).

0
0.2
0.4
0.6
0.8
1
0 5 10 15
-
c1/

d

Figure 26: Vecchio 1982 compression softening model
4.3.5 Vecchio-Collins 1986
The model is strength-only softened version of the Vecchio-Collins 1982 model. The model was
originally developed for the Hognestad Parabola compression stress-strain curve. Note that this model
does not recognize the softening effects of shear slip when they are included in the analysis.
1
34 . 0 8 . 0
1
1

=
o
c
d
(4.3.5.1)
c d p
f' f = (4.3.5.2)
o p
= (4.3.5.3).
64
4.4 Tension Stress-Strain Response
In tension, concrete is predominantly brittle and its response can be differentiated into uncracked
and cracked response.
Prior to cracking the response is assumed to be linear-elastic, as follows:
(4.4.1)
cr c c c c
for E f < < =
1 1 1
0
where,
c
cr
cr
E
f
= (4.4.2).

cr
is the cracking strain , E
c
is the initial tangent stiffness of concrete,
c1
is the principal tensile strain,
and f
cr
is the cracking stress of the concrete determined by the cracking criterion model.
After cracking in reinforced concrete structures, the concrete tensile stresses diminish virtually to
zero at the free surface of cracks. However, owing to bond action with the reinforcement, average
concrete tensile stresses continue to exist in the concrete between the cracks in the vicinity of the
reinforcement. With additional tensile straining, cracks widen, the bond action degrades near the cracks
and the average concrete tensile stresses gradually diminish to zero. While these average concrete tensile
stresses must be less than f
cr
(or else additional cracking ensues), they act over a relatively large tributary
area of the reinforcement. As such the stiffness of the reinforced concrete in tension is greater than that of
the reinforcement alone. For this reason, the phenomenon is known as tension stiffening.
Tension stiffening is important to modeling the load-deformation behavior, particularly in the
finite element context. If tension stiffening is neglected, the concrete tensile stress reduces immediately to
zero upon cracking and the tensile stress must be redistributed entirely to the reinforcement. The
discontinuous change in the stiffness may manifest as an unrealistic abrupt deviation in the load-
deformation response and pose difficulties to the solution convergence for lightly reinforced structures.
The cracked concrete exhibiting tension stiffening must be within the tributary area of the
reinforcement. In VecTor2, the tributary area of discrete reinforcement elements is delineated by a
distance of 7.5 bar diameters from the discrete reinforcement element. (Note that in this manner, the
tension stiffening effect depends somewhat on the coarseness of the finite element mesh). When the
65
Tension Chord Model is selected, the tributary area of externally bonded FRP plates and sheets is based
upon crack formation and bond considerations.
In VecTor2, the tension stiffening is effected by a gradually decreasing average stress-strain
response of concrete in tension, as presented in the following discussion. The average concrete tensile
stress determined is denoted by , to distinguish it from the average concrete tensile stress due to
tension softening effects, denoted by . The magnitude, , is always limited by the yielding of the
reinforcement of the crack, and additionally by the maximum shear stress at the crack when slip
deformations are not included. The larger of the two tensile stresses is assumed to be the average post-
cracking concrete tensile stress:
a
c
f
1
b
c
f
1
a
c
f
1
( )
b
c
a
c c
f , f max f
1 1 1
= (4.4.3).
4.4.1
4.4.2
4.4.3
No Tension Stiffening
The tension stiffening effect is ignored and the post-cracking concrete tensile stress is zero.
1 1
0 0
c cr
a
c
for f < < = (4.4.1.1).
Bentz 2003
The development of this model is in progress and is currently the same as the Bentz 1999 model.
When complete, it will be a more rigorous adaptation of the Bentz 1999 model to two-dimensional stress
conditions.
Vecchio 1982
This model, originally used in the Modified Compression Field Theory, is based upon tests
conducted at the University of Toronto of thirty panel elements measuring 890x890x70mm with welded
wire mesh reinforcement. Compared to the Collins-Mitchell 1987 model, this model is more appropriate
for smaller scale elements and structures. The average concrete tensile stress-strain response is curve,
shown in Figure 27, is determined as:
66
1
1
1
0
200 1
c cr
c
cr a
c
for
f
f < <
+
= (4.4.3.1).

c
f
c
f
cr

cr
ct =200
ct =500

Figure 27: Vecchio 1982 and Collins-Mitchell tension stiffening response
4.4.4 Collins-Mitchell 1987
The model, a modification of the Vecchio 1982 model, is based upon the results of shell elements,
measuring 1450x1450x350mm with reinforcing bars, tested at the University of Toronto. This model
results in a more rapidly diminishing tension stiffening effect compared to the Vecchio 1982 model and is
more appropriate for larger scale elements and structures. The average concrete tensile stress-strain
response, shown in Figure 27, is curve is determined as:
1
1
1
0
500 1
c cr
c
cr a
c
for
f
f < <
+
= (4.4.4.1).
4.4.5 Bentz 1999
Bentz (1999) proposes a tension stiffening formulation that incorporates the bond characteristics
of the reinforcement, given that the tension stiffening effect relies upon bond action. The model,
originally formulated for sectional analysis of reinforced concrete members, has been adapted by Vecchio
for VecTor2 to account for two dimensional stress conditions and the placement of smeared and discrete
reinforcement. The average concrete tensile stress-strain response is curve is determined as:
67
1
1
1
0
6 . 3 1
c cr
c
cr a
c
for
m
f
f < <
+
= (4.4.5.1).
In brief, the bond parameter, m reflects the ratio of the area of concrete to the bonded surface area of the
reinforcement that is tributary to the concrete. For a constant area of reinforcement, the bond
characteristics of numerous bars of smaller diameter are superior to those of fewer large diameter bars. As
such, the former will have a smaller value of m and result in a larger tension stiffening effect.
4.4.6 Izumo, Maekawa Et Al.
Izumo et al. (1992) propose an analytical model for reinforced concrete panels subjected to in-
plane stresses using a smeared crack approach. The model exhibits good accordance with the
experimental results for seventeen of the Vecchio and Collins panels. The average concrete tensile stress-
strain response curve, shown in Figure 28, is determined as:

< <

< < <

=
1
1
1
1
2 0
2
2 0
c cr
c
c
cr
cr
cr c cr cr
a
c
for f
for f
f (4.4.6.1)
The parameter, c, reflects the bond characteristics. In VecTor2, the model is implemented with 4 . 0 = c ,
the recommended value for deformed reinforcing bars. Unlike the preceding models, the tensile stress
remains at the cracking stress up to two times the cracking strain before descending nonlinearly.

c
f
c
f
cr

cr

Figure 28: Izumo, Maekawa et al. tensions stiffening response
68
4.4.7 Tension Chord Model (Kaufmann)
The preceding tension stiffening models are empirical relationships based on investigations of
concrete reinforced with embedded steel bars or wire, rather than externally bonded FRP plates or sheets.
Considering the differences in dimensional and mechanical properties between the steel and FRP, and the
more brittle bond between FRP and concrete, it follows that the preceding tension stiffening models are
not ideal for determining tension stiffening effects due to FRP plate or sheet reinforcement.
To address crack spacing and tension stiffening effects in concrete reinforced with externally
bonded FRP plates and sheets, as well as conventional reinforcement, Sato and Vecchio (2002)
implemented the tension chord model (Kauffmann and Marti, 1998). This tension chord model considers
a truss-like segment of reinforced concrete, as shown in Figure 29. The two ends of the chord represent
two consecutive cracks, and the length of the chord is equal to the crack spacing. The chord is subject to
tensile stresses along the longitudinal axis, which are resisted entirely by the reinforcement at the ends,
and by a distribution of tensile reinforcement and concrete stresses between the two cracks. For a given
value of tensile reinforcement stress at the ends of the chords, the tension chord model considers the bond
stresses and slips between the concrete and reinforcement to determine the distribution of tensile stresses
for the two materials between the cracks.
With its dependence on the crack spacing and bond stress-slip relationships, the tension chord
model cannot be simply summarized as stress-strain relationship for concrete. While the complete
description is beyond the scope of this discussion, the average concrete tensile stress is determined as
follows:

= =
+

=
n
j
j F F j F j F
m
i s r
i s
i b
i s i e
bo r
a
c
E
s d
s f
1
,
2
, ,
1
,
,
, ,
1
cos
cos
, 1 min
cos
(4.4.7.1).
The first term in the above equation represents the contribution of the steel reinforcement to tension
stiffening. The value s
r
is the crack spacing,
bo
is the maximum average bond stress of the steel
reinforcement, and
s
is the average reinforcement strain. For the ith steel reinforcement,
e,i
is the
effective reinforcement ratio in its tributary area, d
b,i
is its diameter, and
s,i
is the angle between principal
69
concrete tensile stress direction and the axis of the steel reinforcement. The second term in the above
equation represents the contribution of the FRP reinforcement to tension stiffening. The value
F
is the
average difference between the average FRP strains and the local FRP strains at a crack. For the jth FRP
reinforcement,
F,j
is the effective reinforcement ratio in its tributary area, E
F,j


is the elastic modulus of
elasticity,
F,j
is the angle between principal concrete tensile stress direction and the axis of the FRP
reinforcement.


s
r
Figure 29: Tension chord model
4.5 Tension Softening
Tension softening refers to the presence of post-cracking tensile stresses in plain concrete. Under
increased tensile straining, the tensile stresses diminish to zero. This phenomenon is attributable to the
fact that concrete is not perfectly brittle. Rather, as described by fracture mechanics approaches, the
formation of a localized crack requires energy. As the fracture process progresses and the crack widens,
concrete in the vicinity of the crack is gradually relieved of stress, and the dissipated energy propagates
the crack tip.
Tension softening is significant in several ways to the analysis of reinforced concrete structures,
particularly those having lightly reinforced regions. The tension softening response may be important to
modeling the stress redistribution and localized damage of lightly reinforced structures exhibiting brittle
failure modes. By including a descending post-cracking stress-strain branch for plane concrete, it is
possible to more accurately determine the load-deformation response and ductility of the member.
Further, tension softening may mitigate inaccuracies associated with the coarseness of the finite element
mesh. Due to their finite size, the elements invariably include both cracked and uncracked concrete.
f
Fcr f
Fcr
70
Accounting for the post-cracking tensile stress in cracked elements represents to some extent the stiffness
contribution of uncracked concrete, and prevents undue stress concentration in adjacent uncracked
elements.
In VecTor2, tension softening is effected by descending post-cracking average tensile stress-
strain curves for concrete described in the subsequent discussion. The average concrete tensile stress due
to tension softening is denoted by , to distinguish it from the average concrete tensile stress due to
tension stiffening effects, denoted by . The larger of the two tensile stresses is assumed to be the
average post-cracking concrete tensile stress:
b
c
f
1
a
c
f
1
( )
b
c
a
c c
f , f max f
1 1 1
= (4.5.1).
It is convenient to define common parameters in the tension-stiffening models. The fracture
energy, G
f
, is the energy required to form a complete crack of unit area. It describes the resistance of the
concrete to cracking and is equivalent to the area beneath a plot of tensile stress versus crack width. The
fracture energy is independent of element size, and is assigned a value of 75N/m in VecTor2. The
representative length, L
r
, is the distance over which the crack is assumed to be uniformly distributed, and
is assigned a value of half the crack spacing. The characteristic strain,
ch
, of the tension softening curve is
determined as follows:
cr ch cr
cr r
f
ch
. ,
f L
G
< <

= 10 1 1
2
(4.5.2).
The corresponding tensile stress, f
ch
, is the characteristic stress of the tension softening curve. The
terminal strain,
te
, is the strain at which the tension softening stress diminishes to zero, determined as:
ch te
= 5 (4.5.3).
In models, without residual tension, the tensile stress due to tension softening is equal to the
tensile stress, f
ts,base
, computed from the tension softening base curve.
base ts
b
c1
f f
,
= (4.5.4).
In models with a sustained residual tension, f
res
, the residual stress is calculated as follows:
71
( )

w mm for w - f .
mm w for f .
f
cr
cr
res

<
< <
=
5 0 5 . 2 3 1 0
5 0 1 0
(4.5.5).
For these models, the tensile stress due to tension softening is taken as the maximum of the tensile stress,
and the residual stress, f
res
as follows:
( )
res base ts
b
c
f , f max f
, 1
= (4.5.6).
Residual tension should not be utilized in structures predominantly subject to tension stresses.
4.5.1
4.5.2
Not Considered
Post-cracking tensile stresses due to tension stiffening are not considered:
0
1
f
b
c
= (4.5.1.1).
Linear
The tension softening base curve descends linearly from the cracking stress and strain to zero
stress at the characteristic strain (i.e. the terminal strain is taken as the characteristic strain). The curve, as
shown in Figure 30 with and without the residual branch, is given as follows:
( )
( )
c1 cr
cr ch
cr c
cr base ts,
for f f <

= , 0 1
1
(4.5.2.1).

c
f
c
f
cr

cr

te
=
ch
w/ residual

Figure 30: Linear tension softening response with and without residual
72
4.5.3 Yamamoto 1999 No residual
As proposed by Yamamoto (1999), the tension softening base curve descends non-linearly from
the cracking stress to the characteristic stress and strain, then linearly to zero stress at the terminal strain.
The curve, shown in Figure 31 with and without the residual branch, is given as follows:
( )
( )
( )

<


< <
+
=
c1 ch
ch te
c te
ch
ch c1 cr
cr c
cr
base ts,
for f
for
c
f
f
0
1
1
1
(4.5.3.1)
The tension softening coefficient, c, is determined such that area beneath the tension softening curve is
equal to the ratio of the fracture energy to the characteristic length as follows:

=
0
1 , c base ts
r
f
d f
L
G
(4.5.3.2).

c
f
c
f
cr

cr

ch
w/ residual

te

Figure 31: Yamamoto tension softening response with and without residual
4.6 Tension Splitting
Tension splitting refers to the formation of splitting cracks parallel to reinforcing bars in tension,
due to the prying action of bar deformations. As described by Vecchio and DeRoo (1995) this
phenomenon results in anomalous post-cracking dilation behavior in panels with small amounts of
73
transverse reinforcement that are loaded in uniaxial tension or combined tension and shear. Contrary to
the usual Poissons effect, the concrete dilates in the direction transverse to the tension, such that the
principal compressive strains become tensile soon after cracking. Such dilation may significantly affect
the load-deformation response of confined concrete.
In VecTor2, the effects of tension splitting are effected by modifying the post-cracking Poissons
ratio, v
21
, which relates strains in the principal compressive strain direction due to actions in the principal
tensile strain direction.
4.6.1
4.6.2
Not Considered
The computation post-cracking Poissons ratio follows the equations in Section 4.7.
DeRoo 1995
The post-cracking Poissons ratio, v
21
, as shown in Figure 32, is computed as follows:
( )
1
1
1
21
06 . 0
c cr
c
cr c cr o
for
v
v <

= (4.6.2.1)
where v
o
is the pre-cracking Poissons ratio (typically between 0.10 and 0.25),
cr
is the cracking strain,
and
c1
is the principal concrete tensile strain. The value of -0.06 is the experimentally observed post-
cracking incremental Poissons ratio in panels with some transverse reinforcement. Therefore, v
21,
may
become negative to reflect the dilation effect of tension splitting.

-0.10
-0.05
0.00
0.05
0.10
0.15
0.20
0.000 0.002 0.004 0.006 0.008 0.010 0.012

c 1
v
21

Figure 32: DeRoo tension splitting response
74
4.7 Lateral Expansion
Lateral expansion effects are modeled in VecTor2 according to the formulation discussed in
Section 2.5.1. Due to internal microcracking, the rate of concrete lateral expansion increases as the
compressive stress increases. Near the peak compressive stress, the volume of concrete expands as
cracking becomes extensive. When confined by reinforcement, the lateral expansion gives rise to passive
confining pressures, which may significantly enhance the strength and ductility of concrete in
compression.
For concrete in tension, the Poissons ratios are computed in the following manner regardless of
the selected lateral expansion model. If the concrete is uncracked, then v
12
and v
21
are equal to the initial
Poissons ratio, v
o
. If the concrete is cracked, then the Poissons ratio decreases linearly from v
o
at the
cracking strain,
cr
, to zero at two times the cracking strain.

<

< <
= =
1
1
1
21 12
0
2
1
0
c cr
cr
c
o
cr c o
for v
for v
v v (4.7.1).
The following lateral expansion models pertain to the determination of the Poissons ratio, v
ij
,
relating the expansion of concrete in the principal i-direction due to compressive straining,
cj
, in the
principal j-direction.
4.7.1
4.7.2
Constant Poisson's ratio
The Poissons ratios v
ij
are always equal to the initial value, v
o
.
0 < =
cj o ij
for v v (4.7.1.1).
Variable Poisson's Ratio Kupfer
Based on strain data of Kupfer et al. (1969), the Poissons ratio v
ij
increases nonlinearly as
compressive strain increases. The Poissons ratio, shown in Figure 33, is computed follows:
75

<

+
< <
=
p cj
p
cj
o
cj p o
ij
. for v
. - for v
v
5 0 5 . 0 1
2
5 . 1 1
0 5 0
2
(4.7.2.1)
where
p
is the strain corresponding to the peak compressive stress.

cji
v
ij
v
o
0.5
0.5
p

Figure 33: Kupfer variable Poissons ratio model
4.7.3 Variable Poissons Ratio Montoya
This preliminary version of the Montoya model determines the Poissons ratio according to the
triaxial compression state, defined by principal concrete stresses, f
c3
<f
c2
<f
c1
<0. For expansion in the i-
direction, the Poissons ratio is computed as follows:
2
max
p
c
o ijk
v v

+ = (4.7.3.1).
where v
o
is the initial Poissons ratio,
cmax
is the most compressive of principal strains
cj
and
ck
,
orthogonal to the i-direction, and
p
is the strain corresponding to the peak compressive stress. Both
cmax

and
p
are expressed as positive values in millistrain. The parameter is determined by the triaxial
compressive stress state as follows:
3
'
18 +

=
c
cl
f
f
(4.7.3.2).
76
( )
0
2
>
+
=
f f
f
ck cj
cl
(4.7.3.3).
4.8 Confinement Strength
Confined concrete exhibits enhanced strength and ductility in compression. In VecTor2, strength
enhancement due to confinement is effected by a strength enhancement factor,
l
, which is calculated by
the confinement strength models. The value of
l
serves to modify the concrete compression response
curves by increasing both the uniaxial compressive strength, f
c
, and corresponding strain,
0
, to determine
the peak compressive strength, f
p
, and corresponding strain,
p
, as follows:
c l d p
f' f = (4.8.1)
o l d p
= (4.8.2).
(The parameter,
d
, accounts for compression softening. Also, note that the calculation of
p
is subject to
modification if certain compression softening models are selected.)
4.8.1
4.8.2
Strength Enhancement Neglected
The uniaxial compressive strength and corresponding strain are not enhanced for confinement.
1 =
l
(4.8.1.1).
Kupfer / Richart Model
In the case of triaxial compression, in which f
c3
<f
c2
<f
c1
<0, the strength enhancement factor for
the direction of the largest compressive stress, f
c3
, is determine by the equation below. The first term is an
adaptation of the relationship proposed by Kupfer et al. (1969) to determine the strength of concrete
subject to biaxial compression. The second term is the stress enhancement in columns with spiral
reinforcement as noted by Richart et al. (1928).
0 ,
'
1 . 4
'
76 . 0
'
92 . 0 1
1 2
2
f f
f
f
f
f
f
f

c c
c
cl
c
cn
c
cn
l
< <

+ = (4.8.2.1)
77
wheref
c
is the uniaxial concrete cylinder compressive strength, f
cn
is the difference in normal lateral
stresses acting on the concrete:
( ) 0
1 2
> =
c c cn
f f f (4.8.2.2)
and f
cl
is the lateral confining stress on the concrete:
0
1
> =
c cl
f f (4.8.2.3).
The strength enhancement factor for the other compressive stress directions are similarly determined by
interchanging f
c3
, f
c2
, f
c1
as necessary.
The above expressions can be extended to the case of biaxial compression, in which f
c3
<f
c2
<0 and
f
c1
=0, to determine the strength enhancement factor for the direction of the largest compressive stress, f
c3
.
Strength enhancement in the direction of f
c2
, is similarly determined by interchanging f
c3
for f
c2
in the
above equations. Note that in either case, the second term of equation 4.8.2.1 is zero.
4.8.3 Selby Model
In this model, the stress required to cause failure the in the direction of the maximum compressive
stress, f
c3f
, in the presence of compressive stresses, f
c1
and f
c2
is determined by numerically solving the
failure surface of Hsieh et al. (1979):
0 1
'
2312 . 0
'
1412 . 9
'
9714 . 0
'
0108 . 2
1 1 2
2
2
= + + +
c c
c
c
c
f
I
f
f
f
J
f
J
(4.8.3.1)
where I
1
is the first stress invariant determined as:
f c c c
f f f I
3 2 1 1
+ + = (4.8.3.2)
and J
2
is the second deviatoric stress invariant determined as:
( ) ( ) ( [ )
2
1 3
2
3 2
2
2 1 2
6
1
c f c f c c c c
f f f f f f J + + = ] (4.8.3.3).
The stress enhancement factor
l
is the ratio of the failure stress, f
c3f
, to the uniaxial concrete cylinder
strength, f
c
:
c
f c
l
f
f
'
3
= (4.8.3.4).
78
4.8.4 Montoya / Ottosen
In this model, the stress required to cause failure the in the direction of the maximum compressive
stress, f
c3f
, in the presence of compressive stresses, f
c1
and f
c2
is determined by numerically solving the
four parameter failure surface of Ottosen (1979):
0 1
' '
'
1
2
2
2
=

c c
c
f
I B
f
J
f
J A
(4.8.4.1).
I
1
is the first stress invariant determined by Equation 4.8.3.2 and J
2
is the second deviatoric stress
invariant determined by Equation 4.8.3.3.
The first dimensionless parameter A is computed as follows:

> >
>
> <
<
=
MPa f' and . R for .
MPa f' and . R for
MPa f' and . R for
MPa f' and . R for
A
c
c
c
c
40 2 0 128 14
40 2 0 327 . 2
40 2 0 039 . 12
40 2 0 731 . 2
(4.8.4.2).
R is the ratio of the lateral compressive stress, f
l
to the uniaxial concrete strength, f
c
, computed as
follows:
c
cl
f
f
R
'
= (4.8.4.3)
( )
2
2 1 c c
cl
f f
f
+
= (4.8.4.4).
The second dimensionless parameter B is determined from the concrete tensile strength, f
t
, and the
concrete biaxial compressive strength, f
bc
, as follows:

=
bc
c
t
c
c
t bc
f
f
f
f
f
f f A
B
' '
3
1
' 9
(4.8.4.5)
( )
33 . 0
' 65 . 0
c t
f f = (4.8.4.6)
c bc
f f ' 16 . 1 = (4.8.4.7).
The value of is computed using the third and fourth dimensionless parameters, K
1
and K
2
, as follows:
+ = 3 cos
2 1
K K (4.8.4.8)
79

= 1 1
' 3
'
2
3
1
c
t
t
c
f
f A

f
f
K (4.8.4.9)

= 1 2 1
' 3
'
2
3
2
B
f
f A
f
f
K
c
t
t
c
(4.8.4.10).
The invariant cos3 is computed as follows:
5 . 1
2
3
2
196 . 5
3 cos
J
J
= (4.8.4.11)
where the third deviatoric stress invariant J
3
is given by:

=
3 3 3
1
3
1
2
1
1 3
I
f
I
f
I
f J
f c c c
(4.8.4.12).
The stress enhancement factor
l
is the ratio of the failure stress, f
c3f
, to the uniaxial concrete
cylinder strength, f
c
:
c
f c
l
f
f
'
3
= (4.8.4.13).
4.9 Cracking Criterion
The cracking strength of concrete, f
cr
, is not an inherent material property. In addition to factors
such as specimen size, and compressive strength, the cracking strength varies with the stress states. In
particular, the cracking strength generally decreases as transversely acting compressive stresses increase.
The cracking criterion accounts for this effect by computing f
cr
, based on the coexisting compressive
stresses or strains. Therefore, f
cr
is generally different from the input value of concrete tensile strength, f
t
.
Having determined f
cr
, the cracking strain,
cr
, is computed by assuming a linear-elastic relationship
before cracking:
c
cr
cr
E
f
= (4.9.1)
where E
c
is initial tangent stiffness of concrete. The values of f
cr
and
cr
are then used in the tension
stress-strain response and tension stiffening models. The following discussion describes the available
cracking criterion models.
80
4.9.1
4.9.2
Uniaxial cracking stress
The cracking strength is taken as the specified uniaxial cracking strength:
t cr
f f ' = (4.9.1.1).
Mohr-Coulomb (Stress)
The Mohr-Coulomb criterion is often used to determine the failure shear stress and failure plane
for a given set of coexisting normal stresses in a frictional material, concrete in this case. It is implicitly
assumed that the concrete is sufficiently ductile to redistribute stresses along the failure plane. The failure
envelope is tangent to the Mohrs circles defining combinations of shear stress, , and normal stresses, f
c1

and f
c3
, resulting in shear failure.
As shown in Figure 34, the shear strength is comprised of a stress-independent component and a
stress-dependent component. The latter is the internal angle of friction, , which VecTor2 assumes is 37.
The former is the cohesion, c, which is determined by noting that at failure in uniaxial compression, the
maximum compressive stress, f
c3
, is equal to the concrete cylinder strength, f
c
, and f
c1
is zero. Therefore:

=
cos 2
sin 1
'
c
f c (4.9.2.1).
Further, when f
c3
=0, the failure tensile stress f
c1
is equal to f
cru
, which is computed from the envelope as:

=
cos 2
cos 2
'
c
f f
c cru
(4.9.2.2).
Given a set of principal concrete strains,
c3
<
c2
<
c1
, the principal compressive stress, f
c3
, is
computed as follows:

<
< <
< <

=
3
3
3
2
3 3
3
0 0
0 '
0 2 '
c
c o c
o c
o
c
o
c
c
c
for
for f
for f
f (4.9.2.3).
Having defined the principal compressive stress, the cracking strength is the principal tensile stress, f
c1
, of
the Mohrs circle tangent to the failure envelope. The value of f
cr
is computed as follows:
81
' ' 3
20 . 0 ,
'
1
t cr t
c
c
cru cr
f f f
f
f
f f

+ = (4.9.2.4).

Figure 34: Mohr-Coulomb (Stress) cracking criterion
4.9.3 Mohr-Coulomb (Strain)
This model is similar to the Mohr-Coulomb (Stress) model. Given a set of principal concrete
strains,
c3
<
c2
<
c1
, and the strain
o
corresponding to the uniaxial concrete cylinder compressive
strength, the failure cracking strength is computed as follows:
t cr t
o
c
cru cr
f f f f f ' ' 20 . 0 , 1
3

= (4.9.3.1).
4.9.4 CEB-FIP Model
The CEB-FIP model reduces the cracking strength for increasing biaxial compression, based on
the linear relationship proposed by Kupfer et al. (1973):
' ' 3
20 . 0 ,
'
8 . 0 1
t cr t
c
c
cru cr
f f f
f
f
f f

+ = (4.9.4.1).
=37
c
f
cru
f
cr

f
c
f
c3
82
The principal compressive stress, f
c3
, is computed from equation 4.9.2.3. The unconfined cracking
strength, f
cru
, is computed from the compressive strength, f
c
, as follows:
c cru
f f ' 6 . 0 = (4.9.4.2)
4.9.5 Gupta 1998 Model
In an investigation of reinforced concrete members subject to high levels of axial compression and
shear, Gupta (1998) observes that the predicted shear capacity is sensitive to the assumed cracking
strength, f
cr
. In examining data from Kupfer et al. (1969), Gupta notes that the cracking strength of
concrete reduces rapidly at high levels of compression due to internal microcracking. Gupta proposes a
linear reduction in cracking strength, with increasing compressive strain as follows:
t cr t
o
c
t cr
f f f f f ' ' 25 . 0 , 1 '
2

= (4.9.5.1)
where
c2
is the principal compressive strain,
o
is the strain corresponding to the concrete cylinder
compressive strength, f
c
.
4.10 Crack Slip Check
When element crack slip distortions are not considered, as in the MCFT, it is necessary to check
that the local shear stresses, , at a crack do not exceed a maximum shear stress, , corresponding to
sliding shear failure. If this value is found to be exceeded, then the average concrete tensile stresses, f
ci
v
max
ci
v
c1
,
must by reduced by the factor
ci ci
v v
max
and the stress and strain state of element is reconsidered. The
following discussion presents models for the determining the maximum value of local shear stress, . v
ci
max
4.10.1 Crack shear check omitted
No limit is imposed on the value of v
ci
.
83
4.10.2 Vecchio-Collins 1986
Based on the work of Walraven (1981), Vecchio and Collins (1986) propose that the local shear
stress be limited as follows:
( )
( ) MPa mm,
a w/ .
f'
v
c
ci
26 24 3 0
max
+ +
= (4.10.2.1)
where f
c
is the uniaxial concrete cylinder compressive strength, w is the crack spacing and a is the
maximum aggregate size. The limiting shear stress increases as the uniaxial concrete cylinder
compressive strength increases, the crack width decreases, or the maximum aggregate size increases.
4.10.3 Gupta 1998 Model
Gupta (1998) conducted a series of shear transfer tests and developed a shear friction model.
Based on the test results, Gupta modified the Vecchio-Collins 1986 model for concretes in which the
crack surface passes through the aggregates, as is typical in higher strength concretes. The maximum
shear stress at the crack interface is computed as follows:
( )
( MPa mm, f' .
w
-
a w/ .
R
v
c ci
001 0
10
1
26 24 3 0
max

+ +
= ) (4.10.3.1)
The value, R, is a roughness factor computed as follows:
( ) MPa f' U v
c ci
5 . 0 2
max
+ = (4.10.3.2).
The 2 MPa accounts for the natural roughness of the crack surface. The parameter U is the proportion of
unfractured aggregates on the crack surface, expressed as a value between zero and one, and therefore
accounts for the roughness due to aggregates protruding from the cement at the crack surface. Typically,
U decreases as the concrete strength increases.
4.11 Crack Width Check
The crack width check serves to reduce average compressive stresses when crack widths exceed a
specified limit. This check was implemented for the analysis shear-critical reinforced concrete beams
having little or no shear reinforcement (Vecchio, 2000), in which element shear-slip distortions were not
84
considered. Such beams exhibit a dominant shear crack of considerable width at failure. It is necessary to
limit the compressive stress in elements with excessive crack widths for two reasons. First, concrete near
the crack exhibits tensile strains, which may exceed the calibration range of compression softening
models and warrant additional softening. Secondly, if shear-slip distortions are not considered, the
reorientation of the principal stresses tends to be overestimated, thereby implying a transmission of local
compressive stresses across the crack. Yet, such transmission is unlikely when cracks exceed the specified
limit. It was found that rapidly reducing the average compressive stress when the crack limit is exceeded
provides more accurate predictions of the load-deformation response. However, it may be preferable to
include element slip distortions in the analysis instead of including the crack width check.
The crack width check is implemented by reducing the average compressive stress computed
from the stress-strain response, , by a crack coefficient
*
2 c
f
cr
as follows:
*
2 2 c cr c
f f = (4.11.1).
If the crack width check is omitted,
cr
=1.0. Otherwise
cr
is computed as follows:
( )

<
<
=
w w for w w
w w for
l l
l
cr
0 3 1
1
(4.11.2)
where w is the crack width in the element, and w
l
is the limiting crack width. The limiting crack width
may be selected as a quarter or half the aggregate size, or widths of 1mm, 2mm, or 5mm.
4.12 Element Slip Distortion
The element slip distortion models allow the VecTor2 analysis to explicitly account for strains due
to shear slip along the crack, in the manner of the Disturbed Stress Field Model. As described in Section
2.3.4, three methods are available for determining the shear slip,
s
, along the crack: stress-based models,
constant rotation lag models, and hybrid models.
Stress-based models relate the shear slip,
s
, along the crack to the local shear stress, v
ci
, along
the crack. The shear slip strain,
s
, is computed as the shear slip divided by the crack spacing, s:
85
s
s a
s

= (4.12.1).
Note that the implemented versions of these models do not account for shear slip before the crack surfaces
develop traction. As well, these models compute zero shear slip in unreinforced elements, as the local
shear stress for such elements is always computed as zero.
Constant rotation lag models relate the post-cracking rotation of the principal stress field,

, to
the post-cracking rotation of the principal strain field,

, by a specified rotation lag,

l
, as follows:
( )

>

=

for
for
l l
l
| |
| |
(4.12.2).
The inclination of the principal stress field is determined by adding

to the inclination at cracking,
ic
:

+ =
ic
(4.12.3).
Shear strains can then be determined from Mohrs circle transformation of the total strains as follows:
( )

+ = 2 sin
x y xy
b
s
cos2 (4.12.4)
Note that constant rotation lag models may be unsuitable at higher load intensities when the amount of lag
increases and shear slip depends increasingly on the shear stress on the crack.
Hybrid models compute the shear slip strains according to both a stress-based model and a
constant rotation lag model, and utilize the greater of the two values.
( )
b
s
a
s s
max = , (4.12.5).
At low load intensities, the shear slip is generally governed by the constant rotation lag model, which
captures the initial slip as the crack surfaces develop traction. At higher load intensities, the shear slip is
generally governed by the stress-based model, which captures the increasing slip.
Lai (2001) made the following conclusions regarding the merits of the approaches. In monotonic
loading, the three stress based models (Walraven, Maekawa, and Lai/Vecchio) provided comparable
levels of accuracy. In cyclic loading the Walraven model appears to be more accurate and stable. In
constant rotation lag models, the lag should be taken as 10 for unreinforced elements, 7.5 for uniaxially
reinforced elements and 5 for biaxially reinforced elements.
86
4.12.1 Not considered
As in the Modified Compression Field Theory, the analysis does not account for shear slip along
the crack:
0 =
s
(4.12.1.1).
4.12.2 Stress Model (Lai-Vecchio)
This model, proposed by Vecchio and Lai (2002), is a combination of the Walraven and
Maekawa models and is found to provide levels of accuracy comparable to these models. The slip along
the crack,
s
, is computed as follows:
w
s s
2
1
*

= (4.12.2.1)
where
( )
cc
co ci
s
f w w
v v
+
+
=

20 . 0 234 . 0 8 . 1
5 . 0
707 . 0 8 . 0
max *
(4.12.2.2)
max ci
ci
v
v
= (4.12.2.3)
( ) 16 24 31 . 0
'
max
+ +
=
a w
f
v
c
ci
(4.12.2.4)
( ) MPa in
f
v
cc
co
30
= (4.12.2.5).
Additionally, v
ci
is the local shear stress on the crack, f
c
is the concrete cylinder compressive strength, w
is the average crack width, a is the maximum aggregate size, and f
cc
is the concrete cube strength, taken as
1.2f
c
. The term v
co
, implements an initial offset in the crack shear-slip relationship.
4.12.3 Stress Model (Maekawa)
This model is proposed by Okamura and Maekawa (1991) and utilized in their fixed non-
orthogonal crack model. The slip along the crack,
s
, is computed as follows:
87
w w
s
2
1

= (4.12.3.1)
where,
max ci
ci
v
v
= (4.12.3.2)

( ) 16 24 31 . 0
'
max
+ +
=
a w
f
v
c
ci
(4.12.3.3).
Additionally, v
ci
, f
c
, w and a are as previously defined.
4.12.4 Stress Model (Walraven)
This model is an adaptation of the formulations of proposed by Walraven and Reinhardt (1981),
based on an analysis of crack structure and contact area of crack faces. The slip along the crack,
s
, is
computed as follows:
( )
w
f w w
v v
cc
co ci
s
2
20 . 0 234 . 0 8 . 1
707 . 0 8 . 0

+
+
=

(4.12.4.1)
where v
ci
, v
co
, w and f
cc
are as previously defined.
4.12.5
4.12.6
Hybrid-I, II, and III Models
The Hybrid-I Model combines the Lai/Vecchio model with the constant rotation lag model.
The Hybrid-II Model combines the Maekawa model with the constant rotation lag model.
The Hybrid-III model combines the Walraven model with the constant rotation lag model.
In these hybrid models, the constant rotation lag,
l
, is taken as 10 for unreinforced elements,
7.5 for elements with one smeared reinforcement component and 5 for elements with two or more
smeared reinforcement components.
Rotation lag of n degrees
Constant rotation lag models are available with
l
values of 5, 7.5, 10, and 15.
88
4.13 Hysteretic Response
As a consequence of internal damage, the stress-strain response curves of concrete under loading,
unloading and reloading are non-coincident. The resulting plastic offset strains along with the area
delineated by the hysteretic loops, are indicative of the internal damage and energy dissipation under
cyclic loading. As such, the hysteretic response is a critical influence on the strength and ductility of
reinforced concrete structures subjected to cyclic and reverse cyclic loading.
The following hysteretic response models describe how concrete reloads to and unloads from the
monotonic concrete stress-strain curve. As such, the monotonic stress-strain curve is likened to a
backbone, to which unloading and reloading curves attach. The hysteretic response also incorporates the
plastic strain offsets that define the unloading path.
4.13.1 No Plastic Offsets
The concrete reloads linearly from and unloads linearly to the point of zero strain and zero stress
of the monotonic stress-strain curve.
When reloading in the compression domain to a compressive strain of
c
, the concrete
compressive stress, f
c
, is computed as follows:
( )

< <
< <

<
=
0
0
0 0
cm c c bc
c cm cm
cm
c
c
c
for f
for f
for
f (4.13.1.1).
where is the maximum previously attained compressive strain, is the corresponding stress, and
is the function defining the monotonic compression stress-strain base curve. In the latter case,
and are updated as
cm

cm
f
( )
c bc
f
cm

cm
f
c
and f
c
respectively.
Unloading in compression results in concrete stresses as follows:
c cm c
E f =

(4.13.1.2)
where, is the unloading modulus in compression, computed as:

cm
E
89
cm
cm
cm
f
E

(4.13.1.3).
When reloading in the tensile domain to a tensile strain of
c
, the concrete tensile stress, f
c
, is
computed as follows:
( )

<
< <

=
c cm c bt
tm c tm
tm
c
c
for f
for f
f
0
(4.13.1.4).
where is the maximum previously attained tensile strain, is the corresponding stress, and
is the function defining the monotonic tensile stress-strain base curve. In the latter case, and
are updated as
tm

tm
f
( )
c bt
f
tm

tm
f
c
and f
c
respectively.
Unloading in tension results in concrete stresses as follows:
c tm c
E f =

(4.13.1.5)
where, is the unloading modulus in tension, computed as:

tm
E
tm
tm
tm
f
E

(4.13.1.6).
4.13.2 Plastic offsets; linear loading/unloading
This model is similar to the preceding model, except that it includes plastic offset strains as
proposed by Vecchio (1999).
When reloading in the compression domain to a compressive strain of
c
, the concrete
compressive stress, f
c
, is computed as follows:
( )

< <
< < <


< < <
=
0
0
0 0 0
cm c c bc
p
c c cm cm
p
c cm
p
c c
c c
p
c
c
for f
for f
or for
f (4.13.2.1).
90
where is the current plastic offset strain, is the maximum previously attained compressive strain,
is the corresponding stress, and
p
c

cm

cm
f ( )
c bc
f is the function defining the monotonic compression stress-
strain base curve. In the latter case, and are updated as
cm

cm
f
c
and f
c
respectively.
At a given compressive strain,
c
, the instantaneous plastic strain, , is computed as follows: '
p
c

<

<

=
p c
p
c
c p
p
c
p
c
p c
p
c
. for
. for
5 1
002 . 0
001305 . 0
5 1 29 . 0 87 . 0
'
2
(4.13.2.2)
where
p
is the strain corresponding to the peak stress on the compression base curve. If is more
compressive than , then is updated as .
'
p
c

p
c

p
c
'
p
c

Unloading in compression results in concrete stress, f

c
, is as follows:
( )
p
c c cm c
E f =

(4.13.2.3)
where is the unloading modulus in compression, computed as:

cm
E
( )
p
c cm
cm
cm
f
E

=

(4.13.2.4).
When reloading in the tensile domain to a tensile strain of
c
, the concrete tensile stress, f
c
, is
computed as follows:
( )

<
< <


=
c cm c bt
tm c
p
c tm
p
c tm
p
c c
c
for f
for f
f (4.13.2.5).
where is the current plastic offset strain , is the maximum previously attained tensile strain,
is the corresponding stress, and
p
c

tm

tm
f
( )
c bt
f is the function defining the monotonic tensile stress-strain base
curve. In the latter case, and are updated as
tm

tm
f
c
and f
c
respectively.
For a given strain,
c
, in the tension domain, the instantaneous plastic strain, , is initially the
strain at which the response first traverses from the compression to tension domain. Thereafter, the
'
p
c

91
monotonic stress-strain response is calculated with respect to that strain. Subsequently is updated
until non-negative and held at zero afterwards such that no plastic offset strains are considered in the
tension domain.
'
p
c

Unloading in tension results in concrete stress, f

c
, as follows:
( )
p
c c tm c
E f =

(4.13.2.6)
where is the unloading modulus in tension, computed as:

tm
E
( )
p
c tm
tm
tm
f
E

=

(4.13.2.7).
4.13.3 Plastic offsets; nonlinear loading/unloading
This model proposed by Vecchio is similar to the preceding model except that unloading in the
compression and tension domains follows nonlinear Ramsberg-Osgood formulations.
Unloading in compression to a strain of
c
results in concrete stress, f
c
, is as follows:
( )
( )
( )
20 1
1



+ + =

c
N
cm
p
c c
N
cm c c
cm c c cm c
N for
N
E
E f f
c
c
(4.13.3.1)
where is the current plastic offset strain, is the maximum previously attained compressive strain,
is the corresponding stress. N
p
c

cm

cm
f
c
is the Ramsberg-Osgood power term representing the deviation from
linear elasticity. It is computed such that the initial unloading modulus is equal to the initial tangent
stiffness of concrete, E
c
, as follows:
( )
( )
cm
p
c c cm
cm
p
c c
c
E f
E
N
+

= (4.13.3.2).
In the case that N
c
is less than one or greater than twenty, f
c
is computed by a linear unloading response
between from
c
to :
p
c

( )
c c
p
c c c c
N or N for E f = 20 1 (4.13.3.3).
Unloading in tension results in concrete stress, f
c
, as follows:
92
( )
( )
( )
20 1
1



+ =

t
N
p
c tm t
N
c tm c
c tm c tm c
N for
N
E
E f f
t
t
(4.13.3.4)
where is the current plastic offset strain, is the maximum previously attained tensile strain,
is the corresponding stress. N
p
c

tm

tm
f
t
is computed as follows such that the initial unloading modulus is equal to
the initial tangent stiffness of concrete, E
c
:
( )
( )
tm
p
c tm c
p
c tm c
t
f E
E
N


= (4.13.3.5).
In the case that N
t
is less than one or greater than twenty, f
c
is computed by a linear unloading response
between from
c
to :
p
c

( )
t t
p
c c c c
N or N for E f = 20 1 (4.13.3.6).
4.13.4
)
Plastic offsets; nonlinear w/ cyclic decay (Palermo Model)
The model proposed by Palermo and Vecchio (2002) is similar to the preceding model.
Modifications include modeling of damage in the reloading curves, consideration for partial unloading
and reloading, the shape of the unloading of curves, and calculations of the instantaneous plastic offset
strains, in both the compression and tension domains. The shape of the hysteretic responses in
compression and tension are shown in Figures 35 and 36, respectively.
When reloading in the compression domain to a compressive strain of
c
, the concrete compressive
stress, f
c
, is computed as follows:
(
ro c cm ro c
E f f + =
+
(4.13.4.1)
where is the strain at load reversal in the current hysteretic loop, is the corresponding stress,
ro

ro
f
cm

is the unloading strain in the current hysteretic loop are
cm
and is the corresponding stress.
cm
f
The reloading modulus in compression, , is computed as follows
+
cm
E
ro cm
ro cm d
cm
f f
E


=
+
(4.13.4.2).
93
The damage indicator,
d
, serves to degrade , such that additional straining is required to intersect the
base-curve. The damage is a function of the strain recovered in unloading of the current hysteretic loop,

+
cm
E
rec
, and is computed as follows:
( )
( )

>
+
<
+
=
p c
p rec
p c
p rec
d
for
for
6 . 0
5 . 0
175 . 0 1
1
10 . 0 1
1
(4.13.4.3).
ro cm rec
= (4.13.4.4).
where
p
is the strain corresponding to the peak stress in the base-curve.
The instantaneous plastic strain, , for the compression domain is computed as follows: '
p
c

=
p
cm
p
cm
p
p
c
132 . 0 166 . 0 '
2
(4.13.4.5).
If is more compressive than , then is updated as . '
p
c

p
c

p
c
'
p
c

Unloading in compression to a strain of

c
results in concrete stress, f
c
, is as follows:
( )
( )( )
( )

N
E
E f f
c
c
N
cm
p
c c
N
cm c c
cm c c cm c
1
1 071 . 0

+ + = (4.13.4.6).
where is the current plastic offset strain, is the maximum previously attained compressive strain,
is the corresponding stress. N
p
c

cm

cm
f
c
is the Ramsberg-Osgood power term representing the deviation from
linear elasticity. It is computed such that the unloading modulus is equal to the initial tangent stiffness of
concrete, E
c,
at the beginning of the unloading branch and equal to 0.071 E
c
at the end of the unloading
branch, as follows:
( )( )
( )
cm
p
c c cm
cm
p
c c
c
E f
E
N
+

=
071 . 0 1
(4.13.4.7).
When reloading in the tensile domain to a tensile strain of
c
, the concrete tensile stress, f
c
, is
computed as follows:
94
(
c tm tm tm t c
E f f =
+
) (4.13.4.8)
where is the unloading strain in the current hysteretic loop, is the corresponding stress, and
is the reloading modulus in tension, computed as follows
tm

cm
f
+
cm
E
ro tm
ro tm t
tm
f f
E


=
+
(4.13.4.9).
The damage indicator,
t
, serves to degrade , such that additional straining is required to intersect the
base-curve. The damage is a function of the strain recovered in unloading of the current hysteretic loop,

+
tm
E
rec
, and is computed as follows:
( )

rec
d
25 . 0
15 . 1 1
1
+
= (4.13.4.10)
ro tm rec
= (4.13.4.11).
The instantaneous plastic strain, , for the tension domain is computed as follows: '
p
c

tm tm
p
c
+ = 523 . 0 146 '
2
(4.13.4.12).
If is more tensile than , then is updated as . '
p
c

p
c

p
c
'
p
c

Unloading in tension results in concrete stress, f

c
, as follows:
( )
( )( )
( )

N
E E
E f f
t
t
N
p
c tm t
N
c tm
f
c c
c tm c tm c
1


+ = (4.13.4.13).
where is the current plastic offset strain, is the maximum previously attained tensile strain,
is the corresponding stress. N
p
c

tm

tm
f
t
is computed as follows such that the unloading modulus is equal to the
initial tangent stiffness of concrete, E
c
at the beginning of the unloading branch, and equal to at the
end of the unloading branch, and is computed as follows:
f
c
E
( ) ( )
( )
tm
p
c tm c
p
c tm
f
c c
t
f E
E E
N


= (4.13.4.14)
( )
( )

<
<
=
001 . 0 001 . 0 053 . 0
001 . 0 001 . 0 071 . 0
tm tm c
tm tm c
f
c
for E
for E
E (4.13.4.15).
95
Partial unloading and reloading is effected by similar equations with modifications to the reversal
and maximum strains under consideration.

-f
c
base curve for
compression softened
response

Figure 35: Palermo model of concrete hysteretic response in compression


Figure 36: Palermo model of concrete hysteretic response in tension
f
p

p
p
c

f
c
f
cr
base curve for tension
stiffenedresponse

cr

c
p
c

96
5 Models for Reinforcement Materials
The following discussion describes the constitutive and behavioral models pertaining to the response
of reinforcement materials, which includes both steel and fiber reinforced polymers. The monotonic
stress-strain response of reinforcement is discussed, followed by the hysteretic response models.
Subsequently, dowel action and reinforcement buckling models are discussed.
5.1 Stress-Strain Response
The following models describe the monotonic stress-strain response of reinforcement materials.
These models are categorized by reference types of reinforcement: ductile steel reinforcement,
prestressing steel, tension only reinforcement, compression only reinforcement, and externally bonded
FRP reinforcement.
5.1.1 Ductile Steel Reinforcement
The reinforcement stress-strain response is trilinear, as shown in Figure 37. It consists of an initial
linear-elastic response, a yield plateau, and a linear strain-hardening phase until rupture. The
reinforcement stress, f
s
, in tension and compression is determined as follows:
( )

<
< +
<

=
s u
u s sh sh s sh y
sh s y y
y s s s
s
for
for - E f
for f
for E
f
0
(5.1.1.1)
( )
sh
y u
sh u
E
f f
+ = (5.1.1.2)
where
s
is the reinforcement strain,
y
is the yield strain,
sh
is the strain at the onset of strain hardening,

u
is the ultimate strain, E
s
is the elastic modulus, E
sh
is the strain hardening modulus, f
y
is the yield
strength, and f
u
is the ultimate strength.
97

s
f
s
- f
y
E
sh
1

y

u

sh

sh
y
u
f
y
f
u
-f
u
E
s

Figure 37: Ductile steel reinforcement stress-strain response
5.1.2 Prestressing Steel
This model is appropriate for cold-worked steel reinforcement that does not exhibit a distinct
yield plateau, but rather an initial linear-elastic branch, followed by a transition curve to a second
hardening linear branch, as shown in Figure 38. The reinforcement stress, f
s
, in tension and compression is
determined by a Ramsberg-Osgood formulation as follows:
( ) [ ]
u
C
C
s
s s s
f
B
A
A E f

+

+ =
1
1
1
(5.1.2.1)
s
sh
E
E
A = (5.1.2.2)
( )
*
1
s
s
f
A E
B

= (5.1.2.2)
t coefficien transition C = (5.1.2.3)
where
s
is the reinforcement strain, E
s
is the initial elastic modulus, f
u
is the ultimate strength and is
the value at which the second linear branch intercepts the stress axis at zero strain. A representative value
*
s
f
98
of the transition coefficient, C, for prestressing strands with an ultimate strength of 1860 MPa is 10 for
low-relaxation steel, or 6 for stress-relieved steel.
s
f
s
f
u
E
s
1

Figure 38: Prestressing steel reinforcement stress-strain response
5.1.3
5.1.4
Tension Only Reinforcement
In tension, the stress-strain response is the same as the trilinear relationship of the ductile steel
reinforcement model. In compression, the reinforcement stress is always zero. This model may be utilized
to represent fabric-type FRP reinforcement that offers no resistance in compression.
Compression Only Reinforcement
In compression, the stress-strain response is the same as the trilinear relationship of the ductile
steel reinforcement model. In tension, the reinforcement stress is always zero. This response may be
utilized to model contact phenomenon between two disjoint structures, by specifying an initial prestrain
representing the relative movement prior to contact.
99
5.1.5
5.2.1
5.2.2
Externally Bonded FRP Reinforcement
This model is similar to the tension only reinforcement model, in that the reinforcement does not
exhibit compressive stress. Additional modifications allow the local crack stresses in externally bonded
FRP reinforcement to be more accurately calculated.
5.2 Hysteretic Response
The following hysteretic response models describe how the reinforcement reloads to and unloads
from the monotonic stress-strain curve. As such, the monotonic stress-strain curve is likened to a
backbone, to which unloading and reloading curve attach.
Linear
The reinforcement reloads linearly from and unloads linearly to the point of zero strain and zero
stress of the monotonic stress-strain curve. No plastic offset strains result.
Seckin Model w/ Bauschinger Effect
The following model is proposed by Seckin (1981), as shown in Figure 39. It includes the
Bauschinger effect, in which the reinforcement exhibits premature yielding upon load reversal after
plastic prestraining due to stress changes at the microscopic level.
When reloading in a positive cycle to a strain of
i
, the reinforcement stress is f
s
, reflects the
Bauschinger effect with a Ramsberg-Osgood formulation as follows:
( )
( )
(
N
o i
N
o m
r m
o i r s
N
E E
E f

)

+ =
1
(5.2.2.1)

( )(
(
)
)
o m r m
o m r m
E f
E E
N


= (5.2.2.2)

( )
( )
( )

<
< <

<
=
o m y s
y o m y
y
o m
s
y o m s
r
for E
for E
for E
E
4
4 05 . 0 05 . 1 (5.2.2.3)
100
where,
o
is the plastic offset strain in the current cycle,
y
is the yield strain,
m
is the maximum positive
strain attained in previous cycles, E
m
is the tangent stiffness at
m
, E
r
is the unloading modulus, E
s
is the
elastic modulus of the monotonic stress-strain response, and f
m
is the stress corresponding to
m
.
Unloading is linear and results in reinforcement stresses as follows:
(5.2.2.4). (
1
1

+ =

i i r s s
E f f
i
)
5.2.3
5.2.4

f
s
E
sh
1
f
y
E
s
1

s
Figure 39: Seckin model of reinforcement hysteretic response
Elastic-Plastic
The monotonic stress-strain curve is bilinear, comprised of an initial linear-elastic branch,
followed by a yield plateau. The reinforcement unloads and reloads linearly with a modulus equal to
elastic modulus of the reinforcement, E
s
. As such, plastic strain offsets result.
Elastic-Plastic w/ Hardening
The monotonic stress-strain curve is trilinear, as described for the ductile steel reinforcement. The
reinforcement unloads and reloads linearly with a modulus equal to elastic modulus of the reinforcement,
E
s
. As such, plastic strain offsets result.
101
5.3 Dowel Action
Dowel action refers to shear resistance offered by reinforcing bars crossing a crack as the crack
slips transversely to the axis of the reinforcement. In some circumstances, such as beams with small
amounts of transverse reinforcement, dowel action may contribute significantly to the shear strength and
post-peak ductility of reinforced concrete members.
The model is used in conjunction with the Vecchio/Lai and Walraven stress-based element slip
distortion models and their hybrid formulations. The shear resistance due to dowel action, v
dl
, is computed
as a function of the shear slip,
s
, at the crack. This shear resistance is subsequently subtracted from the
local shear stress on the crack, v
ci
, thereby reducing the amount of computed shear slip.
5.3.1
5.3.2
Not considered
The shear resistance at a crack due to dowel action is not considered.
Tassios Model
The dowel force-displacement relationship is modeled as elastic-plastic. For simplicity, consider a
reinforcing bar perpendicularly crossing a crack as shown in Figure 40. (Reinforcing bars crossing cracks
at oblique angles are also accommodated.) As described by He and Kwan (2001), the dowel action of the
reinforcement may modeled as a beam on an elastic concrete foundation. The dowel force, V
d
, due to the
relative displacement,
s
, of the cracks is computed as follows:
(5.3.2.1)
du s z s d
V I E V =
3

64
4
b
z
d
I

= (5.3.2.2)
4
4
z s
b c
I E
d k
= (5.3.2.3)

3 2
' 127
b
c
c
d
f c
k

= (5.3.2.4)
8 (5.3.2.5) . 0 = c
102

y c b du
f f d V ' 27 . 1
2
= (5.3.2.6)
where
s
is the shear slip along the crack, d
b
is the diameter of the reinforcement, E
s
is the elastic modulus
of the reinforcement, f
y
is the yield strength of the reinforcement, f
c
is the compressive strength of the
concrete. The parameter I
z
is the area moment of inertia of the reinforcement. The parameter compares
the stiffness of the concrete to that to the reinforcing bar. The parameter k
c
is the stiffness of notional
concrete foundation, where c is an experimentally based coefficient to reflect the bar spacing. The
ultimate dowel force, V
du
, corresponds to plastic hinging of the reinforcement and crushing of the
surrounding concrete in multiaxial compression.
The shear resistance due to dowel action is computed as a smeared contribution as follows:

s
d s
d
A
V
v

= (5.3.2.7)
where
s
is the reinforcement ratio and A
s
is the area of the reinforcement.

s

Figure 40: Dowel resistance mechanism
5.4 Reinforcement Buckling
Reinforcement bars may be susceptible to buckling failures when located near the surface of a
concrete member and subjected to high levels of compressive stress. When bars buckle, the associated
concrete cover and reinforcement cease to contribute to the flexure and shear resistance of the member.
5.4.1 Not Considered
If not considered, the capacity of reinforcement bars in compression is dictated only by the
compressive stress-strain response of the reinforcement material.
103
5.4.2 Asatsu Model
In order to use this model, the reinforcement bars subject to buckling must be modeled discretely by
truss bar elements and connected to the concrete elements with bond (link or contact) elements. A full
description of the determination of the buckling length and buckling criteria are beyond the scope of this
discussion. In brief, the criteria for reinforcement buckling are that reinforcement is plasticized, is subject
to compressive stresses exceeding eighty percent of the yield strength and that the bond deterioration is
severe, such that the cumulative energy consumption exceeds the fracture energy G
f
.

104
6 Models for Bond
The following discussion describes the monotonic bond stress-slip models for bond elements. In
VecTor2, the bond materials are divided into two categories: models for embedded bars (smooth and
embedded) and models for externally bonded plates or sheets. For the former, the bond stress-slip
relationship is internally calculated by VecTor2 according to the selected model, while for the latter, the
bond stress-slip relationship is specified by the user by a series of reference bond stress and slips.
Hysteretic response is included in the models, but not discussed here.
6.1 Bond Stress-Slip Models for Embedded Bars
For embedded bars, the bond model first determines the stress-slip relationship for two distinct
cases: confined bars and unconfined bars. The stronger relationship for confined bars corresponds to pull-
out type bond failure, whereas the relationship for unconfined bars corresponds to splitting failure. Both
the confined and unconfined bond stress-slip relationships are defined by a series of reference bond
stresses, , and bond slips, . Those pertaining to pullout failure are subscripted as
p
,
p
while those
pertaining to splitting failure are subscripted,
s
,
s
.
The actual bond stress-slip model is defined by a series of reference bond stresses and slips,

sp
,
sp.
These are determined by linearly interpolating between the unconfined and confined reference
bond stresses and slips using the confinement pressure factor, . A confinement pressure of zero
corresponds to the unconfined case of splitting failure, while a confinement pressure of 7.5 MPa
corresponds to the confined case of pullout failure. Based on the anticipated confining pressure, , the
confinement pressure factor may be computed as follows:
, 1 0 ,
5 . 7

= MPa) (in (6.1.1).

The following models utilize several properties pertaining to the reinforcing bars and their
placement to compute the bond stress-slip relationship. The minimum of the concrete cover and half of
105
the bar spacing is denoted, c. For deformed reinforcement bars, VecTor2 computes the lug spacing, S, and
lug height, H, depending on the reinforcement bar diameter, d
b
, as follows:
(6.1.2).

<
<
<
<
<
<
<
<

=
mm d mm 0 for d d
mm d mm for mm mm
mm d mm for mm mm
mm d mm for mm mm
mm d mm for mm mm
mm d mm for mm , mm
mm d mm for mm , mm
mm d mm for mm mm
d mm for mm mm
H S
b b b
b
b
b
b
b
b
b
b
11 04 . 0 , 70 . 0
15 11 45 . 0 , 9 . 7
19 15 72 . 0 , 2 . 11
25 19 98 . 0 , 6 . 13
29 25 26 . 1 , 6 . 17
35 29 48 . 1 9 . 20
43 35 79 . 1 0 . 25
55 43 20 . 2 , 6 . 30
55 55 . 2 , 4 . 39
,
For smooth bars, VecTor2 computes the lug spacing according to Equation 6.1.2, but computes the lug
height as follows:

75
S
H = (6.1.3)
The model for hooked bars is different is described subsequently.
6.1.1
6.1.2
Perfect bond
The bond material is assigned a numerically large stiffness and strength to prevent deformation of
the bond element.
Eligehausen Model
As proposed by Eligehausen et al. (1983), the confined and unconfined bond stress-slip
relationships are described by an ascending non-linear branch, a constant bond stress plateau, a linearly
declining branch, and a sustaining residual stress branch, as shown in Figure 41.
The confined stress-slip relationship is summarized as follows:

( )
( )
( )
( )

<
<

<

=

3
3 2 2
2 3
2
2
2 1 2
1 1 1
p pf
p p pf p
p p
p
p
p p p
p p p
for
for -
for
for
(6.1.2.1)
106
where

30
'
4
20
1
c b
p
f d

= (6.1.2.2)
(6.1.2.3)
1 2 p p
=

6 . 27
'
07 . 0 5 . 5
c
pf
f
H
S

= (6.1.2.4)

30
'
1
c
p
f
= (6.1.2.5)
(6.1.2.6) mm
p
0 . 3
2
=
(6.1.2.7) S
p
=
3
(6.1.2.8). 4 . 0 =
The unconfined stress-slip relationship is summarized as follows:

( )
( )
( )
( )

<
<

<

=

3
3 2 2
2 3
2
2
2 1 2
1 1 1
s sf
s s sf s
s s
s
s
s s s
s s s
for
for -
for
for
(6.1.2.9)
where

1 1
'
748 . 0
p
b
c
s
d
c f

= (6.1.2.10)
(6.1.2.11)
1 2 s s
=

pf
b
c
sf
d
c f

=
'
234 . 0 (6.1.2.12)

=
1
1
1 1
ln
1
exp
p
s
p s
(6.1.2.13)
(6.1.2.14)
2 2 p s
=
(6.1.2.15).
3 3 p s
=
107
Given a confinement pressure factor, , the bond stress-slip relationship is defined as follows:

( )
( )
( )
( )
( )
( )
( )

<
<

<

3
3 2 2
2 3
2
2
2 1 2
2 3
12
1
1 1 1
sp spf
sp sp spf sp
sp sp
sp
sp
sp sp spf sp
sp sp
sp
sp
sp sp sp
for
for -
for -
for
(6.1.2.16)
( )
1 1 1 1 s p s sp
+ = (6.1.2.17)
(6.1.2.18)
1 2 sp sp
=
( )
sf pf s spf
+ =
1
(6.1.2.19)
( )
1 1 1 1 1 s s p s sp
+ = (6.1.2.20)
(6.1.2.21)
2 2 p sp
=
(6.1.2.22).
3 3 p sp
=

=1
=0

Figure 41: Eligehausen bond stress-slip response
108
6.1.3 Gan Model
The Gan confined bond stress-slip relationship, as shown in Figure 42, is the same as the
Eligehausen model for the confined stress-slip relationship as summarized by equations 6.1.2.1 to 6.1.2.8.
The unconfined bond stress-slip relationship is described by an ascending non-linear branch, a
descending linear branch, and a sustaining residual stress branch, summarized as follows:

( )
( )
( )
( )

<
<

2
2 1 1
1 2
1
1
1 1 1
s sf
s s sf s
s s
s
s
s s s
for
for -
for
(6.1.3.1)
where

1 1
'
748 . 0
p
b
c
s
d
c f

= (6.1.3.2)
(6.1.3.3)
1 2
15 . 0
s s
=
(6.1.3.4)
2 s sf
=

=
1
1
1 1
ln
1
exp
p
s
p s
(6.1.4.5)
(6.1.3.6) mm
s
0 . 2
2
=
(6.1.3.7).
3 3 p s
=
Given a confinement pressure factor, , the bond stress-slip relationship is defined by equation
6.1.2.16, where:
( )
1 1 1 1 s p s sp
+ = (6.1.3.8)
( )
2 2 2 2 s p s sp
+ = (6.1.3.9)
( )
sf sf pf sf spf
< + = (6.1.3.10)
( )
1 1 1 1 1 s s p s sp
+ = (6.1.3.11)
(6.1.3.12)
if
if mm
p
sp

<
=
=
0
0 0 . 2
2
2
109
(6.1.3.13).
if
if mm
p
sp

<
=
=
0
0 0 . 2
3
3

=1
=0

Figure 42: Gan bond stress-slip response
6.1.4 Harajli Model
Both the confined and unconfined bond stress-slip relationships are described by an ascending
non-linear branch, a constant bond stress plateau, and linearly declining branch, and a sustaining residual
stress branch, as shown in Figure 43.
The confined stress-slip relationship is summarized as follows:

( )
( )
( )
( )

<
<

<

=

3
3 2 2
2 3
2
2
2 1 2
1 1 1
p pf
p p pf p
p p
p
p
p p p
p p p
for
for -
for
for
(6.1.4.1)
where

c p
f ' 575 . 2
1
= (6.1.4.2)
(6.1.4.3)
1 2 p p
=
110
(6.1.4.4)
1
35 . 0
p pf
=
(6.1.4.5) ( 18 . 0 189 . 0 75 . 0
1
+ = S
p
)
)
3
(6.1.4.6) ( 18 . 0 189 . 0 75 . 1
2
+ = S
p
(6.1.4.7) S
p
=
3
(6.1.4.8). . 0 =
The unconfined stress-slip relationship is summarized as follows:

( )

<

=

1
1 1 1
0
s
s s s
for
for
(6.1.4.9)
where

1 1
' 291 . 0 249 . 0
p c s
f
d
c

+ = (6.1.4.10)
(6.1.4.11) 0
0
2
=
s
(6.1.4.12) =
sf

=
1
1
1 1
ln
1
exp
p
s
p s
(6.1.4.13)
(6.1.4.14)
2 2 p s
=
(6.1.4.15).
3 3 p s
=
Given a confinement pressure factor, , the bond stress-slip relationship is defined by equation
6.2.1.16 where:
( )
1 1 1 1 s p s sp
+ = (6.1.4.16)
(6.1.4.17)
2 2 p sp
=
(6.1.4.18)
pf spf
=
( )
1 1 1 1 1 s s p s sp
+ = (6.1.4.19)
(6.1.4.20)
if
if mm
p
sp

<
=
=
0
0 0 . 3
2
2
111
(6.1.4.21).
if
if mm
p
sp

<
=
=
0
0 0 . 3
3
3

=1
=0

Figure 43: Harajli bond stress-slip response
6.1.5 Hooked Bars
Regardless of the model selected, the bond stress-slip relationship for hooked bars, shown in Figure
44, is summarized as follows:

( )
( )

<

=


for MPa
for MPa
p
1
1
1
22
22
(6.1.5.1)
where
p1
and are determined by the selected model.
112

Figure 44: Hooked bar bond stress-slip response
6.1.6 Fujii Model
The Fujii model provides the lowest bond strength and ductility among the available models.
Note that this is recommended for analyses where cover splitting is anticipated, but does not consider the
effect of hooked bars. The model determines whether one of two cover splitting modes governs: side
splitting, in which the entire volume of concrete cover spalls, and corner splitting, in which a fraction of
the concrete cover spalls. Side splitting occurs when the side splitting length ratio, Bis, is less than the
corner splitting length ratio, Bic, and vice versa. The value Bis compares the thickness of the concrete to
the cumulative diameter of reinforcing bars, while Bic compares the cover to the diameter of a single bar.
The exact computation is beyond the scope of this discussion.
The bond stress-slip relationships is described by an ascending linear branch, a constant stress
plateau, a descending linear branch, and a sustaining residual stress branch, summarized as follows:

( )
( )
( )
( )

<
<

<

=
3
3 2 2
2 3
2
2
2 1 2
1 1 1
for
for - -
for
for
f
f
s
(6.1.6.1).
113
In the case of , the reference bond stress and slips are determined as follows: 0 =
( ) f Bi
c
8 ' 5 . 1
1
+ = (6.1.6.2)
(6.1.6.3)
1 2
=
(6.1.6.4) MPa
f
1 . 0 =

100
1
1

= (6.1.6.5)
(6.1.6.6)
1 2
04 . 0 = Bi

2 1 3
0 2 + = (6.1.6.7)
(6.1.6.8). ( Bic Bis Bi , min = )
In the case of , the reference bond stress and slips are determined as follows: 0

( )
( )

> +
< + +
=
0.0035 Q for f Bi
0.0035 Q for f Q Bi
c
c
' 3 . 1 307 . 0 313 . 0
' 9 . 24 427 . 0 307 . 0 313 . 0
1
(6.1.6.9)
(6.1.6.10)
1 2
=

( )

>
< +
=
0035 0 ' 35 . 0
0035 0 ' 3 . 0 3 . 23 313 . 0
. Q for f
. Q for f Q
c
c
f
(6.1.6.11)

100
1
1

= (6.1.6.12)

. Q for
. Q for Q

>
< +
=
0035 0 2 25 . 1
0035 0 2 53 . 0 7 . 20
1
1
2
(6.1.6.13)

2 1 3
2 + = (6.1.6.14)
( ) 88 1
*
+ = Bis Q Q (6.1.6.15)

<
<
=
Bis Bic for
Bic Bis for
Q
2
1
*
(6.1.6.16).
114
6.2 Bond Stress-Slip Models for Externally Bonded Plates or Sheets
6.2.1
6.2.2
Perfect bond
Bond materials are assigned a large stiffness and strength to prevent deformation bond elements.
Other models
Regardless of which imperfect bond model is selected, the bond stress-slip relationship for
externally bonded plates or sheets is described by a multilinear relationship as follows:

( )
( )
( )
( )
( )
( )
( )

<
<


+
<


+

=
3 3
3 2 2 3
2 3
2
2
2 1 1 2
1 2
1
1
1 1 1
for
for -
for -
for
s
(6.2.2.1).
The shape of the bond stress-slip relationship, as shown in Figure 45, may be modified by specifying the
values of the reference bond stresses,
1
,
2
, and
3
and corresponding slips,
1
<
2
<
3
.

1
2
3

Figure 45: Bond stress-slip response for externally bonded plates or sheets

115
7 Part II: FormWorks
7.1 Introduction
FormWorks is a preprocessor software that generates input files for VecTor2, the nonlinear finite
element analysis program for membrane structures. The role of FormWorks is to provide a user interface
for generating, visualizing and checking the finite element model. This chapter provides assistance for
using FormWorks and constructing the finite element model.
7.2 Installing FormWorks
The FormWorks program is written in the object-oriented C++programming language and Microsoft
Foundation Classes, and compiled with Microsoft Visual C++Version 6.0. Formworks is designed to run
in the Microsoft Windows operating system. The recommended minimum system requirements are:
PC with 233 MHz or higher processor and 16 MB of RAM
5MB of free hard-disk space
Microsoft Windows 95 or later operating system
16-bit color monitor
mouse or equivalent pointing device.
At a minimum, the FormWorks package consists of the following files:
FormWorks.exe: the executable FormWorks program.
J obOpt.fwd: the FormWorks job options data file.
StrOpt.fwd: the FormWorks structure options data file.
LoadOpt.fwd: the FormWorks load options data file.
FWS.ico: the FormWorks file icon.
Vt2.exe: the executable VecTor2 program.
DOS4GW.exe: an auxiliary executable program required by VecTor2.
116
To install and run FormWorks, complete the following procedure:
1. Create a new folder entitled VecTor in the C:\Program Files folder.
2. Copy all of the above files to the newly created folder.
3. Locate the Control Panel folder in the hard drive.
4. Open the Folder Options folder.
5. Select the File Types tab.
6. Click the New button.
7. In the Create New Extension dialog box, enter FWS in file extension field.
8. Click OK.
9. In the File Types tab, click the Opens with: Change button.
10. Click the Browse button.
11. Browse for and select the FormWorks.exe file in the newly created VecTor folder.
12. Click the Open button, then the Ok button.
13. In the File Types tab, click the Advanced button.
14. In the Edit File Type dialog box, click the Change Icon... button.
15. Browse for and select the FWS.ico file in the newly created VecTor folder.
16. Click the Ok button.
17. Double-click the FormWorks.exe icon in the newly created VecTor folder to run FormWorks.
7.3 An Overview of the FormWorks Modeling Process
The goal of the VecTor2 analysis is to approximate the response of an actual reinforced concrete
structure to a given loading scenario and thereby solve a specific engineering question. The approach is to
partition the structure into finite elements, generate solutions for each element, assemble the solutions and
thereby determine the response of the entire structure. A requisite task for an accurate and relevant
solution is a carefully considered finite element model. The subsequent discussion attempts to bring forth
initial considerations in this process and describe the role of FormWorks in generating input files.
117
7.3.1
7.3.2
Before Using FormWorks
It is advisable to carefully define the structural analysis problem to be addressed by VecTor2
before embarking on the finite element modeling process. With some forethought and experience, it is
possible to simplify input, create an efficient finite element model, and avoid the need for extraneous
analyses. As VecTor2 provides copious amounts of analysis results, it is advisable whenever possible to
construct the finite element model with the intent of determining a specific aspect of structural response.
A numerical objective such as the maximum load factor, displacement or stress levels can suggest the
necessary mesh features and appropriate material models. Assumptions should be considered regarding
the stress-state, material properties, boundary conditions, structural and loading symmetry, and the nature
of the loads. Further, it is useful to hypothesize which mechanisms are critical to the structural response
so that they are adequately represented and suitable material models may be chosen.
VecTor2 Input and Output Files
To run a VecTor2 analysis, several input files are required. As the analysis proceeds, VecTor2
generates several output files. This process is summarized in Figure 46.

Figure 46: Input and output files forVecTor2 analysis

.JOB
~~~~~
~~~~~
.S2R
~~~~~
~~~~~
.L2R
~~~~~
~~~~~
.L2R
~~~~~
~~~~~
.L2R
~~~~~
~~~~~
.L2R
~~~~~
~~~~~
.L2R
~~~~~
~~~~~
.A2R
~~~~~
~~~~~
.S2E
~~~~~
~~~~~
.L2E
~~~~~
~~~~~
.A2E
~~~~~
~~~~~
.A2R
~~~~~
~~~~~
User input FormWorks
VecTor2
Each load stage
118
FormWorks receives user input and generates the following input files for VecTor2:
Job Data File (*JOB)
VecTor2 reads one J ob Data File. This file manages the execution of the VecTor2 analysis. It
specifies the names of the Structure Data file and Load Case file, the load factors, the iteration
parameters, and the selected material and analysis models.
Reduced Structure Data File (*S2R)
VecTor2 reads one Reduced Structure Data File. This file describes the material properties,
elements, material assignments, nodes and restraints in an abbreviated format.
Reduced Load Case Data File (*L2R)
VecTor2 reads one to five Load Case Data Files. Each file describes nodal loads, support
displacements, concrete prestrains, gravity loads, temperature loads, and ingress pressures for
each load case in an abbreviated format. Individual load factors for each load case are specified in
the J ob Data file.
For some analyses, VecTor2 can also read an input seed file. Seed files are Reduced Analysis Data
files with the extension *A2R, generated as output from a previous load stage. Seed files store the strain
and stress history of the structure. These files may be used for analyzing repaired structures by running
the analysis to an intermediate load stage, modifying the structure by disengaging or engaging elements,
and resuming the analysis with the seed file.
During the analysis process, VecTor2 generates the following output files:
Expanded Structure Data File (*S2E)
VecTor2 generates one expanded structure data file. This file prints out the structure data in
greater detail than the reduced structure file by printing the attributes of every material type, node
and element.
Expanded Load Case Data File (*L2E)
VecTor2 generates one expanded load case data file for each reduced load case data file. This file
prints the load case data in greater detail than the reduced load case data file.

119
Expanded Analysis Data File (*A2E)
If desired, VecTor2 generates one expanded analysis file for each load stage, in ASCII format that
may be read with a text editor. This file prints out convergence parameters, reactions,
displacements, crack widths, stresses and strains for concrete and reinforcement, bond stresses
and slips, and stiffness matrix coefficients.
Reduced Analysis Data File (*A2R)
If desired, VecTor2 generates one reduced analysis file for each load stage, in binary format.
These files are also used as seed files.
7.4 FormWorks Basics
7.4.1 The FormWorks Interface
Upon starting FormWorks, the FormWorks application window appears as shown in Figure 47.
The exact configuration of the FormWorks screen elements may vary with the operating system and
display hardware.
FormWorks is a multiple document interface. The FormWorks application window encloses one
or more child Workspace windows. Each Workspace is a unique document that can be saved and opened
as a FormWorks file and contains all the information required to generate the input files for one VecTor2
finite element mode. The application title bar indicates the name of the active workspace in brackets. In
this case, the active workspace is Workspace1, which is created by default when the FormWorks
application opens. The finite element model appears in the Workspace window as it is created.
A menu bar appears near the top of the FormWorks window. When a Workspace is open, the
File, Edit, View, Structure, Load, Analysis, Window and Help menu items are visible. Some menu
items appear greyed-out and become enabled as the finite element model proceeds.
Several toolbars are docked below the menu bar. These buttons provide easier access to the menu
items having the same icon. Similarly, some buttons appear greyed-out and become enabled as the finite
120
element model proceeds. As the mouse lingers over the toolbar buttons, names of the buttons appear
beneath the mouse pointer.
A status bar appears at the bottom of the FormWorks window. A prompt on the left side of the
status bar describes the function of menu items and toolbar buttons as the mouse pointer lingers over
them. In the right side, the Coordinate Pane displays the coordinates of the mouse crosshairs in the active
Workspace window.


Figure 47: FormWorks application window

7.4.2 Creating a New Workspace
To start a new finite element model:
1. Select the File/New menu item. Or, click the New toolbar button.
By default, the new Workspace name reflects the order of their creation. For instance, two Workspace
child windows, Workspace1 and Workspace2, appear in Figure 48. The FormWorks window is titled
FormWorks-Workspace2, indicating that the second Workspace is currently active.
121

Figure 48: Creating new Workspace windows
7.4.3 Saving the Workspace
It is advisable to regularly save the Workspace for backup and later retrieval as the finite element
model progresses. To save the active Workspace:
1. Select the File/Save menu item. Or, click the Save toolbar button. The Save As dialog
box appears.
2. Select a directory in which to save the Workspace. It is recommended that the .FWS file be saved
in the directory containing the FormWorks application.
3. Enter a name for the Workspace in the File Name field.
4. Click Save.
A new Workspace file with the .FWS extension is created in the specified directory. Select the File/Save
menu item or click the Save toolbar button to update an existing Workspace file. To save a version of the
same Workspace in a different file, select the File/Save As menu item and follow the preceding
procedure.
122
7.4.4
7.5.1
Opening a Saved Workspace File
To open a saved Workspace file:
1. Select the File/Open menu item. Or, click the Open toolbar button. The Open
dialog box appears.
2. Browse for and select the desired FormWorks file with the .FWS extension.
3. Click Open.
Alternatively, select the File menu and select from the four most recent Workspace files.
7.5 Viewing and Printing the Workspace
The active Workspace window displays a limited region of an infinite (x,y) plane, corresponding to
the plane of stresses. As the modeling process proceeds, it is necessary to manipulate this view to display
different parts of the finite element model. Furthermore, it may be desirable to control which model
attributes are displayed. The following sections describe how to manipulate and print the Workspace.
Manipulating the View
7.5.1.1 Horizontal and Vertical Scaling
On the screen, the Workspace may be displayed with either equal or unequal scaling of the x and
y distances. The former is generally desirable and is the default option. The latter option may be
convenient for viewing structures in which the width is much greater than the height or vice versa.
Select the View/Maintain Aspect Ratio menu item to toggle between the two modes. When
checked, the scaling is equal in the x and y directions. When unchecked, the scaling is unequal in the x
and y directions, and depends on the current display limits.

7.5.1.2 Changing Display Limits
To specify the portion of x,y plane that is visible in the Workspace, complete the following steps.
123
1. Select the View/Limits menu item. The Display Limits dialog appears as shown in Figure 49.
The current limits of the Workspace window are shown in the dialog box.
2. In the Min X entry field, enter the desired bottom coordinate of the Workspace window.
3. In the Min Y entry field, enter the desired left coordinate of the Workspace window.
4. In the Max X entry field, enter the desired right coordinate of the Workspace window.
5. In the Max Y entry field, enter the desired top coordinate of the Workspace window.
6. Click Ok.
The Workspace window is redrawn with (Min X, Min Y) as the bottom-left coordinates. If the Maintain
Aspect Ratio item is checked the top-right coordinates of the Workspace window are adjusted to maintain
equal horizontal and vertical scaling on the screen. If the Maintain Aspect Ratio item is not checked, the
Workspace window is redrawn with (Max X, Max Y) as the top-right coordinates of the Workspace
window.

Figure 49: Display Limits dialog box

7.5.1.3 Zooming and Panning
Five options exist for zooming and panning the Workspace view. Select the desired option from
the View/Zoom menu item or click the corresponding toolbar button.
Select the Zoom All menu item or click the Zoom All toolbar button to display the
entire finite element model in the Workspace View
124
Select the Zoom In menu item or click the Zoom In toolbar button to increase the scale of the
Workspace View by 10%.
Select the Zoom Out menu item or click the Zoom Out toolbar button to decrease the
scale of the Workspace View by 10%.
Select the Zoom Window menu item or click the Zoom Window toolbar button. The
mouse pointer appears as a magnifying glass. Left-click and drag the mouse pointer to
specify the view window.
Select the Pan Menu item or click the Pan toolbar button. The mouse pointer appears
as a hand. Left -click and drag the mouse pointer to translate the Workspace view window.

7.5.1.4 Selecting Display Options
The display options hides or reveal attributes of the finite element model in the Workspace view.
Select the View/Display Options menu item or click the Display Options toolbar button to
display the Display Options dialog box shown in Figure 50.
Node Options
Node Numbers
Check to reveal the node number beside each node of the finite element model.
Restraints
Check to reveal the support restraints on each node.
Nodal Loads
Select to reveal applied nodal forces for the current load case.
Support Displacements
Select to reveal imposed displacements for the current load case.
Impulse Forces
Select to reveal time-varying forces for the current load case
None
Select to hide the above load types for the current load case.
125

Figure 50: Display Options dialog box
Element Options
Element Numbers
Check to reveal the element number in the center of each element of the finite element model.
Material Color
Check to display elements with the color of the assigned material type or default color. Uncheck
to view elements drawn in black and white.
Material Type Number
Select to reveal the material types labels in the center of each element. Reinforced concrete
materials are displayed as Cn, reinforcement materials are displayed as Rn and bond materials
are displayed as Bn, where n is the ordinal of the material type.
Gravity Loads
126
Select to reveal the density in kg/m
3
and G-forces applied to concrete elements in the load case.
Temperature Loads
Select to reveal the temperature gradient in degrees Celsius for concrete and reinforcement
elements in the current load case.
Concrete Prestrains
Select to reveal the elastic strain offset in millistrain applied to concrete elements in the load case.
Ingress Pressures
Select to reveal ingress pressures in mega Pascal, applied to concrete elements in the load case.
Surface Thermal Loads
Select to reveal time varying thermal loads in degrees Celsius applied to the free surfaces of
concrete elements in the current load case
None
Select to hide all of the above load types for the current load case.

Element Filters
Elements are displayed in layers according to their element type. From bottom to top, the drawing
order is as follows: rectangular elements, quadrilateral elements, triangular elements, truss elements, link
elements, and contact elements. Occasionally, elements conceal elements or their attributes drawn beneath
them. The Element Filters reveals or hides element types or makes them ineligible for mouse selection.
Rectangular, Quadrilateral and Triangular
Check to hide the element attributes, (but not the element itself) and make the elements ineligible
for mouse selection.
Truss, Link and Contact
Check to hide the elements and their attributes, and make the elements ineligible for mouse
selection.
127
7.5.2 Printing the Workspace
FormWorks allows the finite element model to be printed with a standard printer. The entire finite
element model is scaled to fit the selected page format and printed with the same attributes that are shown
in the Workspace view.

To print the finite element model, complete the following procedure:
1. Select the File/Print Setup menu item. Select the desired paper properties and click Ok
2. Select the File/Print Preview menu item to preview the finite element model.
3. Select the File/Print menu item or click the Print toolbar button.
4. Click Ok to print the Workspace.
128
8 The Job Data
The first step in creating the VecTor2 input is to define the J ob Data. At the time of analysis,
FormWorks generates the *J OB file based on the defined J ob Data.
8.1.1 The Job Control Page
1. Select the Job/Define Job menu item or click the Define Job toolbar button. The
Define Job property sheet appears with the Job Control page displayed as shown in Figure 51.

Figure 51: Job Control property page

129
2. Input the job data as described in the subsequent sections.
3. When done, select the Models page.

8.1.1.1 J ob Data Group
These entry fields manage the creation of the *J OB file.
Job File Name
Enter an alpha-numeric file name up to 8 characters long without spaces. This defines the file
name to which FormWorks appends the *J OB extension when saving the J ob Data file.
Job Title
Enter a descriptive identifier up to 30 characters long to differentiate this analysis from others.
Date
Enter the date in a string up to 30 characters long.

8.1.1.2 Structure Data Group
These entry fields manage the creation of the *S2R file.
Structure File Name
Enter an alpha-numeric file name up to 8 characters long without spaces. This defines the file
name to which FormWorks appends the *S2R extension when saving the Structure Data file.
Structure Title
Enter a descriptive identifier up to 30 characters long for the structure being analyzed.
Structure Type
Select Plane Membrane 2-D for the VecTor2 analysis.

8.1.1.3 Load Data Group
Each load case consists of a set of loads that are proportionally increased or decreased by a
common load factor from one load stage to the next. While all load cases act simultaneously on the
130
structure, different load cases can have different load factors. For instance, lateral loads defined in one
load case may be monotonically increased, while gravity loads defined in another load case remain
constant.
Each load case is assigned one of three loading types, Monotonic, Cyclic and Reversed Cyclic.
Examples of each are illustrated in Figures 52, 53 and 54, respectively.
0
5
10
15
20
25
30
1 6 11 16 21 26
Load Stage No.
Load Factor
Starting Load Stage No.: 1
No. of Load Stages: 26
Initial Factor: 0.
Final Factor: 25.
Inc. Factor: 1
Load Type: Monotonic

Figure 52: Monotonic type loading

0
5
10
15
20
25
30
35
40
1 51 101 151 201 251 301 351
Load Stage No.
Load Factor
Starting Load Stage No.: 1
No. of Load Stages: 361
Initial Factor: 0.
Final Factor: 25.
Inc. Factor: 1
Load Type: Cyclic
Repetitions: 3
Cyclic Inc.Factor: 10


Figure 53: Cyclic type loading
131
-40
-30
-20
-10
0
10
20
30
40
1 51 101 151 201 251 301 351 401 451 501 551 601 651 701 751
Load Stage No.
Load Factor
Starting Load Stage No.: 1
No. of Load Stages: 721
Initial Factor: 0.
Final Factor: 25.
Inc. Factor: 1
Load Type: Reversed Cyclic
Repetitions: 3
Cyclic Inc.Factor: 10

Figure 54: Reversed cyclic type loading

The size of the load steps, which is be controlled by the size of the load factor increment, can
appreciably impact the efficiency of the solution convergence. Many small load increments may be
preferable to fewer large load increments, especially when the structure is at an advanced state of distress.
Smaller load increments allow the solution to properly converge in a fewer number of iterations before
the analysis proceeds to the next load step. Excessively large load increments may result in improper
convergence. Given the overall softening response of concrete, improper converge may overestimate the
strength for an imposed displacement, and underestimate the displacement for an imposed load.
The following entry fields are common to all load cases. They define the name of the *A2E and
*A2R output files generated by VecTor2 and the number of loads stages to be analyzed. As the analysis
proceeds, VecTor2 generates output files having the name LoadCaseID_N.A2E and/or
LoadCaseID_N.A2R, where N is the current load stage number.
Load Case ID
Enter an alpha-numeric file name up to 5 characters long without spaces. This defines the file
name to which VecTor2 appends the *A2E and/or *A2R extension when storing analysis results.



132
Starting Load Stage Number
Enter an integer greater than or equal to 1. This defines the number of the first *A2E or *A2R file
that is stored by VecTor2. When resuming an analysis, enter an integer greater than the last
completed load stage to avoid overwriting previously generated output files.
No. of Load Stages
Enter an integer greater than or equal to 1. This defines the number of load stages analyzed by
VecTor2. The total number of required stages is defined by equation 8.1.1.3.1.

To create a load case, complete the following steps:
1. Check the load Case box to activate the load case. Only active load cases can be modified.
2. Complete the following entry fields for the activated load case.
Load File Name
Enter an alpha-numeric file name up to 8 characters long without spaces. This defines the file
name to which FormWorks appends the *L2R extension when generating the Load Case Data
files.
Load Case Title
Enter a descriptive identifier up to 30 characters long to differentiate the load case.
Initial Factor
Enter the load factor of the first load stage.
Final Factor
For monotonic loading, enter the load factor of the last load stage. For cyclic and reversed
loading, enter the maximum load factor of the first set of repetitions.
Inc. Factor
Enter the change in load factor from one load stage to the next.
Load Type
Select the desired loading type from the drop-list.

133
Repetitions
Enter the number of cycles per set. (For cyclic and reversed cyclic loading only.)
Cyclic Inc.Factor
Enter the change in final load factor from one set of repetitions to the next. For uniformity in load
stage increments, the value should be a multiple of the load stage load factor increment. (For
cyclic and reversed cyclic loading only.)

Having specified the load factors and load factor increments, the number of load stages required to
analyze all load stages can be computed as follows:
( ) ( ) [ ]
( ) ( ) [ ]

=
loading cyclic reversed for S S
LS
C R
LS
LF LF
S R
loading cyclic for S S
LS
C R
LS
LF LF
S R
loading monotonic for
LS
LF LF
Stages No.
inc
inc
inc
i f
inc
inc
inc
i f
inc
i f
1 1
2
4
1 1 2
1
(8.1.1.3.1)
where LF
i
is the initial load factor, LF
f
is the final load factor, LS
inc
is the load factor increment for each
load stage, R is the number of repetitions, S is the number of sets of full repetitions and C
inc
is the cyclic
load factor increment.

8.1.1.4 Analysis Parameters Group
This group controls the progress of the iterative solution procedure and the analysis output:
Analysis Mode
Select Load Step from the drop-list. Alternatively, select Time Step from the drop-list if utilizing
time varying loads such as surface thermal loads, impulse forces, or ground acceleration loads.
Seed File Name
Enter NULL if no seed file is used. Otherwise, enter the file name of the *A2R file.

134
Max. No. of Iterations
Enter the maximum number of iterations VecTor2 performs for each load stage. Regardless of the
convergence quality, VecTor2 proceeds to the next load stage when this limit is reached.
Averaging Factor
Enter the weighting factor between 0 and 1 used to update the value of the material stiffness
coefficients between iterations. Structures exhibiting less stability such as lightly reinforced
structures require values closer to zero. Alternatively, check the dynamic averaging factor to
allow VecTor2 to automatically choose a value based on response of the structure.
Convergence Limit
Enter a value greater than 1 for the maximum ratio of the convergence criteria that must be
satisfied before the VecTor2 proceeds from on load stage to the next. As the value approaches 1,
the more stringent the convergence criterion becomes.
Convergence Criteria
Select the parameter by which the solution convergence is judged against the convergence limit
before VecTor2 proceeds to the next load step.
Results Files
Select the file type of the analysis output. ASCII files are extended analysis files that can be read
by text editors. Binary files are reduced analysis files that may be used as seed files.
Output Format
Select To Computer.
8.1.2 The Model Page
1. Select the Models property page as shown in Figure 55.
2. Select the desired material and behavioral models. The following sections provide
abbreviated descriptions of the models.
3. When done, click Ok to store changes to the J ob Data.
135
For the majority of analyses, it is advisable to select the VecTor2 default models. Ultimately,
however, it remains the responsibility of the analyst to exercise his or her discretion regarding the
appropriateness of the models and the reliability of the results.


Figure 55: The Models page






136
8.1.2.1 Concrete Models
Compression Pre-Peak Response
Select the ascending branch of the average concrete compression stress-strain response.
Compression Post-Peak Response
Select the descending branch of the average concrete compression stress-strain response.
Compression Softening
Select the model for reducing the concrete compressive strength and stiffness relative due to the
presence of coexisting transverse tensile strains.
Tension Stiffening
Select the post-cracking average tensile stress-strain response of reinforced concrete.
Tension Softening
Select the post-cracking average tensile stress-strain response of plain concrete.
Tension Splitting
Select the post-cracking dilation response of concrete in the direction that is transverse to the
direction of the principal tensile strains.
Confined Strength
Select the model for strength and ductility enhancement of concrete subject to biaxial or triaxial
compressive stress states.
Dilation
Select the model for computing the post-cracking Poissons effect for expansion in the direction
transverse to compressive stresses..
Cracking Criterion
Select the model for determining the concrete cracking strength based on the stress or strain state
of an assumed failure condition.
Crack Slip Check
Select the model for determining the maximum local shear stress permitted at a crack when
element slip distortions are not considered.
137
Crack Width Check
Select the crack width beyond width the average concrete compressive stress is reduced to effect
the inability of the concrete to transmit compressive stresses across large crack widths.
Slip Distortions
Select the model for determining the crack slip strains as a component of the total strains. If Not
Considered is selected, VecTor2 conducts the analysis based on Modified Compression Field
Theory. If any other option is selected, VecTor2 conducts the analysis based on the Disturbed
Stress Field Model.
Hysteretic Response
Select the model for the average stress-strain response of concrete when subjected to unloading
and unloading, and the resulting plastic strain offsets.

8.1.2.2 Reinforcement Models
Hysteretic Response
Select the model for the average stress-strain response of reinforcement when subjected to
unloading and unloading, and the resulting plastic strain offsets.
Dowel Action
Select the model for determining the contribution to shear resistance of reinforcing bars crossing
cracks.
Buckling
Select the model for determining the failure of the failure of truss bar elements in compression
due to buckling and associated splitting of the concrete cover. (Note that the truss bar elements
must be connected to the concrete with bond elements to use the Asatsu model.)

8.1.2.3 Bond Models
Concrete Bond or Adhesion
Select the bond stress-slip relationship of between concrete and embedded reinforcing bars.
138

8.1.2.4 Analysis Models
Strain History:
The previous loading should be considered for analyses involving cyclic and reversed cyclic
loading to include the hysteretic response models.
139
9 The Structure Data
The second step in creating the input is to define the Structure Data, which describes the finite
element model itself. The following sections describe how to define material properties, create nodes and
elements, restraint the structure and assign material types.
9.1.1 Preliminary considerations
At this stage, a sketch is useful to map out the proposed finite element mesh. Often, a practical
mesh requires some simplification of the actual structure. Nonetheless, the sketch should delineate
geometric features, changes in concrete thickness and mechanical properties, changes in distributed
reinforcement amounts and properties, the locations of concentrated reinforcement, and essential
boundary conditions.
An appropriate mesh topology depends on several factors including the required degree of
accuracy, anticipated stress gradients, selected element types, changes in material types, placement of
loads, and computational limits. Considering the low-power of elements and the formulation of the
Modified Compression Field Theory, elements should be sized so that assumptions of uniform stress and
crack distributions are passable within elements. Analysis results indicating abrupt stress variations in
adjacent elements may suggest the need for mesh refinement.
FormWorks provides two methods for defining the nodes and elements of the finite element
mesh: the Manual Method and the Automatic Method. In the Manual Method, the recommended
sequence of steps for creating the mesh is to specify material properties, specify the numbering and
location of nodes, specify elements, assign nodal restraints, and assign material types to elements. This
method offers complete control over the mesh topology and its computational characteristics, but may be
more time consuming for complicated geometries. In the Automatic Method, the structure is described by
a series of mesh boundaries and reinforcement locations. A mesh generation facility automatically creates
and numbers the nodes, elements, and restraints and assigns material types. This method allows easier
redefinition or refinement of the mesh, but sacrifices some control over the mesh topology.
140
When utilizing the Manual Method, the nodes should be systematically numbered to minimize the
bandwidth of the structure stiffness matrix [K]. The computation time required by VecTor2 is
approximately proportional to the square of the bandwidth. The bandwidth may be approximated as:
( ) [ ] { } 1 2 2 max R j i andwidth B (9.1.1.1)
where, i is largest node number of an element, j is the smallest node number of the same element, R is the
number restrained degrees of freedom for nodes with numbers between i and j inclusive. Therefore,
smaller bandwidths result when the nodes of the elements have numbers that are as similar as possible.
For simply connected meshes, this can typically be achieved by consecutive node numbering along the
shortest dimension of the structure. FormWorks includes a bandwidth reduction algorithm. While it is no
substitute for well planned node numbering, this feature is useful for geometrically complicated meshes
where an efficient node numbering is either non-obvious or impractical to input. It is also useful if the
structure is subsequently altered.
9.1.2 Structure Limits
VecTor2 limits the number of material types, nodes,
elements and a maximum bandwidth permitted for any finite
element mesh. These limits vary with the version of VecTor2. To
view these limits, select the Structure/Structure Limits menu
item. The Structure Limits dialog box appears as shown in
Figure 56.









Figure 56: Structure Limits dialog box
141
9.1.3 Structure Information
Select the Structure/Structure Information menu item or click the Structure
Information toolbar button to determine the number of defined materials types, nodes and elements
currently defined in the model. The Structure Information dialog appears as shown in Figure 57. These
values are updated as the model is constructed.

Figure 57: Structure Information dialog box
9.1.4 Specifying Material Types
VecTor2 includes three types of materials: reinforced concrete, reinforcement, and bond
materials. Their uses and description are described in the following sections.

142
9.1.4.1 Reinforced Concrete Material Types
Reinforced concrete materials types described concrete with or without one or more smeared
reinforcement components. These material types are applied to rectangular, quadrilateral, or triangular
elements.

To add, modify or delete concrete material types, click the Structure/Define
Reinforced Concrete Materials menu item or click the Define Reinforced Concrete
Materials toolbar button. The Define Reinforced Concrete Materials dialog box appears as shown in
Figure 58.

Figure 58: Reinforced Concrete Materials Properties Dialog Box

To Add Concrete Material Types, complete the following steps:
1. Enter the following properties in the Concrete Properties group. Properties marked by * are
assigned default values when 0 are entered in their fields.
Thickness, T
Enter the out of plane thickness of the concrete in millimeters.
143
Cylinder Compressive Strength, fc:
Enter the peak compressive stress of a standard concrete test cylinder, in MPa.
Tensile Strength, ft:
Enter the uniaxial cracking strength of concrete in MPa.
Default value:
c t
f f' ' 33 . 0 = MPa.
Initial Tangent Modulus, Ec:
Enter the tangent stiffness of the concrete stress-strain response at zero-strain, in MPa.
Default value:
c c
f E ' 5500 = MPa.
Cylinder Strain at fc, eo:
Enter the compressive strain,
o
, corresponding to f
c
, as a positive value in millistrain.
Default value:
c o
f ' 0075 . 0 8 . 1 + = millistrain.
Poissons Ratio, Mu:
Enter the initial Poissons ratio, v
o
, of the concrete as a positive value.
Default value: 0.15.
Thermal Expansion Coefficient, Cc:
Enter the concrete strain increase per temperature increase of 1C.
Default value: 10x10
-6
/C.
Maximum Aggregate Size, a:
Enter the maximum aggregate size, in millimeters.
Default value: 10 mm.
Density:
Enter the mass density of the concrete in kg/m
3
.
Default value: 2400 kg/m
3
.
Thermal Diffusivity, Kc:
Default value: 1.20 s/mm
2
.
Average Crack Spacing:
144
Enter the crack control parameter indicating the spacing of cracks parallel to the y-axis for
Sx, and parallel to the x-axis for Sy, in millimeters.
Default value: computed by the CEB-FIP model, unless the Tension Chord (Kauffmann)
model is selected for tension stiffening.
Color:
Select the display color of the concrete material type in the FormWorks Workspace.
2. Click Add in the Concrete Types group. The newly added concrete type appears in the Concrete
Types list box and the reinforcement component properties fields are enabled.

To Modify Concrete Material Types, complete the following steps:
1. In the Concrete Types list box, select the concrete type to be modified.
2. Re-enter the properties in the Concrete Properties group as desired.
3. Click Update in the Concrete Types group to store the modified concrete properties.

To Delete Concrete Material Types, complete the following steps:
1. In the Concrete Types list box, select the concrete type to be deleted.
2. Click Delete in the Concrete Types group. The concrete material type is deleted from the list box
and the remaining concrete types are renumbered.

To Add Reinforcement Components to a concrete material type, complete the following steps.
1. In the Concrete Types list box, select the concrete type to which the reinforcement component
will belong. The selected concrete type appears highlighted in the list box.
2. Enter the following properties in the Reinforcement Component Properties group.
Reference Type
Refer to Section 5.1 for the stress-strain response of different reinforcement types.
Out of Plane Reinforcement
Check this box to orient the reinforcement perpendicularly to the x,y plane.
145
Direction from X-Axis
For in plane reinforcement, enter the inclination of the reinforcement axis, measured
counterclockwise from the positive x-axis in degrees. Enter a value between 0 and 360.
Reinforcement Ratio, As
Enter the ratio of cross-sectional area of the reinforcement to the area of concrete over which
it is smeared, expressed as a percentage.
Reinforcement Diameter, Db
Enter the size of the reinforcing bar, in millimeters.
Yield Strength, Fy
Enter the stress of the yield plateau.
Ultimate Strength, Fu
Enter the maximum stress the reinforcement can attain before rupturing, in MPa. The ultimate
strength must be greater than or equal to the yield strength.
Elastic Modulus, Es
Enter the stiffness of the initial linear-elastic branch of the stress-strain response, in MPa.
Strain Hardening Modulus, Esh
Enter the stiffness of the hardening branch of the stress-strain response, in MPa.
Strain Hardening Strain, esh
Enter the strain at which the reinforcement stress-strain response begins to ascend from the
yield plateau to the ultimate strength, in millistrain. The value must be greater than or equal to
the yield strain, F
y
/E
s
.
Thermal Expansion Coefficient, Cs
Enter the reinforcement strain increase per temperature increase of 1C.
Default value: 10x10
-6
/C.
Prestrain,
p

Enter the elastic strain offset of the reinforcement relative to the unstrained concrete, in
millistrain.
146
3. Click Add in the Reinforcement Components group. The newly added reinforcement
component type appears in the Reinforcement Component Types list box.

To Modify Reinforcement Components of a concrete material type, complete the following steps:
1. In the Concrete Types list box, select the concrete type to which the reinforcement component
belongs. The selected concrete type appears highlighted.
2. In the Reinforcement Components list box, select the component to be modified. The selected
reinforcement component appears highlighted.
3. Re-enter the properties in the Reinforcement Component Properties group as desired.
4. Click Update in the Reinforcement Components group to store the modified reinforcement
component properties.

To Delete Reinforcement Components of a concrete material type, complete the following steps:
1. In the Concrete Types list box, select the concrete type to which the reinforcement component
belongs. The selected concrete type appears highlighted.
2. In the Reinforcement Components list box, select the component to be deleted. The selected
reinforcement component appears highlighted.
3. Click Delete in the Reinforcement Components group. The reinforcement component is deleted
from the list box and the remaining reinforcement components are renumbered.

9.1.4.2 Reinforcement Material Types
Reinforcement materials types describe steel or FRP reinforcement materials for truss bar
elements. To add, modify or delete reinforcement material types, click the Structure/Define
Reinforcement Materials menu item or click the Define Reinforcement Materials toolbar
button. The Define Reinforcement Materials dialog box appears as shown in Figure 59.
147

Figure 59: Reinforcement Materials Properties Dialog Box

The procedures for adding, modifying and deleting reinforcement material types are similar to those for
reinforced concrete materials. The properties for reinforcement materials are similar to those for
reinforcement components. Only the unique values are described below.
Cross-Sectional Area
Enter the cross-sectional area of reinforcement assigned to truss elements, in mm
2
.
Color
Enter the display color of the reinforcement material type in the FormWorks Workspace.

9.1.4.3 Bond Types
Bond types describe bond stress-slip relationships between concrete and discrete
reinforcement, and are applied to link and contact elements. To add, modify or delete bond
148
types, click the Structure/Define Bond Properties menu item or click the Define Bond Properties
toolbar button. The Define Bond Properties dialog box appears as shown in Figure 60.

Figure 60: Bond Properties Dialog Box

The procedures for adding, modifying and deleting bond types are similar to those for reinforced concrete
materials. The properties for bond materials are described below.

Reference Type
As described in Chapter 6, VecTor2 determines bond stress-slip relationships for Embedded
Deformed Bars and Embedded Smooth Bars based on the selected concrete bond model.
For Externally Bonded Plates and Sheets, the bond stress-slip relationship is explicitly
specified by a series of reference bond stress and corresponding slips.
149
Color
Select the display color of the bond type in the FormWorks Workspace.

If Embedded Deformed Bars or Embedded Deformed Bars is selected:
Confinement Pressure Factor
The confinement pressure factor is used to interpolate the bond stress-slip relationship
between unconfined splitting failure and confined pull-out failure. Compute the confinement
pressure factor, , as follows:
( ) , 1 0 ,
5 . 7

= MPa in (9.1.4.3.1)
where is the anticipated confining pressure on the embedded bar in MPa.
Min (Bar Clear Cover, Spacing), Cmin
Enter the lesser of the clear cover of the embedded bar and half of the spacing between the
embedded bars, in millimeters.
No. of Reinforcement Layers thru Depth
Enter the number of layers of reinforcing bars represented by truss elements to be attached to
the bond element.

If Externally Bonded Plates or Sheets is selected:
Bonded Surface Area, Ao
Enter the area of the external plate or sheet that is adherent to the concrete, in mm
2
, per
element.
Bond Stress-Slip Curve Reference Points
Enter three reference bond stresses, U
i
, and corresponding slips, S
i
, to define a multilinear
bond stress-slip relationship. Enter bond stresses, U
i
, in MPa and corresponding slips, S
i
, in
millimeters.
150
9.1.5 Defining Nodes
The following section describes how to add and delete nodes by the Manual Method.
The rules for defining nodes in VecTor2 are as follows:
A finite element mesh having n nodes must contain all nodes numbered consecutively from
1 to n.
All nodes must be attached to at least one element.
FormWorks, circumvents the first rule and allows nodes to be arbitrarily numbered. If the rule is
violated or if the bandwidth reduction algorithm is used, FormWorks renumbers the nodes when
generating the input files. This feature may be useful when it is more convenient to edit a mesh in
progress or modify an existing mesh than to define a new one. Still, it is advisable to number nodes
according to the rule as the renumbering may appear disorderly.

To Create Nodes, complete the following steps.
1. Select the Structure/Create Nodes menu item or click the Create Nodes toolbar
button. The Create Nodes dialog box appears as shown in Figure 61.

Figure 61: Create Nodes Dialog Box

2. Complete the fields to create a lattice of nodes separated by constant spacing and node number.
Refer to Figure 62 as an example corresponding to the values shown in Figure 61.

151

Figure 62: Example of node creation entry

3. Click Add or type Enter. The generated nodes are drawn in the Workspace, the total number of
nodes is updated, and the last node entry is added to the drop list in the dialog box.

To Delete Nodes complete the following steps.
1. In the drop list, select the nodes to be deleted. The selected node entry appears highlighted. Note
that only nodes that are unattached to elements may be deleted.
2. Click Delete. The deleted nodes disappear from the Workspace, the total number of nodes is
updated, and the node entry is deleted from the drop list in the dialog box. Also any nodal loads
or support displacements assigned to these nodes are also deleted.
9.1.6 Defining Elements
The following section describes how to create and delete elements by the Manual Method.
Element types may be categorized by their function in the finite element model. Rectangular,
Quadrilateral and Triangular elements are used in conjunction with Reinforced Concrete material types
to model reinforced concrete regions. Truss Bar elements are used in conjunction with Reinforcement
Enter: Node X Y #nodes dnode dx dy #nodes dnode dx dy
13 100. 200. 6 1 125. 0. 5 6 0. 100.
6 nodes spaced from node 13
@ 125 mm in the x-direction
@ 0 mm in the y-direction
Increasing in node number by 1
Node 13 located @
(100,200) mm.
13 14 15 16 17 18
19 20 21 22 23 24
25 26 27 28 29 30
31 32 33 34 35 36
37 38 39 40 41 42
5 nodes spaced from node 13
@ 0 mm in the x-direction
@ 100 mm in the y-direction
increasing in node number by 6
152
material types to model reinforcement in a discrete manner. Link and Contact elements are used to
model the interface between rectangular, quadrilateral or triangular elements and truss bar elements.
In VecTor2, the following rules exist for defining elements:
A finite element mesh having n elements must contain all elements numbered from 1 to n.
Elements numbers must be ordered by element type from lowest to highest: rectangular,
quadrilateral, triangular, truss, link and contact.
FormWorks circumvents these two rules and allows elements to be arbitrarily numbered. If either
rule is violated, FormWorks automatically renumbers the elements when generating input files. This
feature may be useful when it is more convenient to edit a mesh in progress or modify an existing mesh
than to define a new one. Nevertheless, it is advisable to obey the rules, as renumbering may be disorderly
and make result interpretation difficult.

The procedure for creating and deleting elements is similar for all six element types. A generic
procedure follows, with necessary rules for each element type.
To Create Elements, complete the following procedure.
1. Select the Structure/Create [Type] Elements
menu item or click the Create [Type] Elements
toolbar button for the desired element type. As an example the Create Rectangular Elements
dialog box is shown in Figure 63.


Figure 63: Create Rectangular Elements Dialog Box

153
2. Complete the fields to create a lattice of elements with constant change in element numbering and
defining nodes. Refer to Figure 64 as an example for creating rectangular elements. The rules for
different element types are subsequently discussed.


Figure 64: Example of rectangular element creation entry

For Rectangular elements, enter four node numbers that define the corner nodes of the source
element. The four nodes must be predefined and form the corners of an isothetic rectangle in
cyclic counter-clockwise order.

For Quadrilateral elements, enter the four node numbers that define the vertex nodes of the
source element. The four nodes must be predefined and form the corners of a quadrilateral in
cyclic counter-clockwise order. Note that quadrilateral elements that are rectangular in shape are
not analytically equivalent to rectangular elements.
Enter: Elmt node 1 2 3 4 #elmts delmt dnode #elmts delmt dnode
1 13 14 20 19 5 1 1 4 5 6
5 elements from element 1
Increasing in element number by 1
Increasing in defining node number by 1
Element 1
Defined by nodes 13, 14 20, 19
13 14 15 16 17 18
19 20 21 22 23 24
25 26 27 28 29 30
31 32 33 34 35 36
37 38 39 40 41 42
4 elements from element 1
Increasing in element number by 5
Increasing in defining node number by 6
16 17 18 19 20
11 12 13 14 15
6 7 8 9 10
1 2 3 4 5
154
For Triangular elements, enter the three node numbers that define the vertex nodes of the source
element. The three nodes must be predefined and may be in either cyclic clockwise or cyclic
counterclockwise order.

For Truss elements, enter the two node numbers that define the end nodes of the source element.
The two nodes must be predefined.

For Link elements, enter the two node numbers that define the nodes of the source element. The
two nodes must be predefined with the same coordinates so that the link is non-dimensional. As
shown in the example of Figure 65, defining link elements requires creating layers of nodes. One
node must be an incident node of a rectangular, quadrilateral or triangular element. The other
node must be an incident node of a truss bar element. Further, if two or more truss bars intersect
at a common node and each truss bar is connected to the concrete by a link element, it is
necessary to define separate nodes and link elements for each truss bar at the intersection
location.

90 13 33 6 1 1 1 1 1
Figure 65: Example of link element creation entry

Enter: Elmt node 1 2 #elmts delmt dnode #elmts delmt dnode
Elements:
Rectangular: 12 to 15
Truss: 46 to 40
Link: 90 to 95

11 12 13 14 15
13 14 15 16 17 18
33 34 35 36 37 38
46 47 48 49 50
90 91 92 93 94 95
Node Pairs with same coordinates:
13 & 33
14 & 34
15 & 35
16 & 36
17 & 37
18 & 38
19 20 21 22 23 24
155
For Contact elements, enter the four node numbers that define the nodes of the source element.
As shown in the example of Figure 66, nodes 1 and 2 must have the same coordinate, and nodes 3
and 4 must have the same coordinate. Of each node pair, one node must be an incident node of a
rectangular, quadrilateral or triangular elements, while the other node must be an incident node of
a truss bar element. Further, if two or more truss bars intersect at a common node and each truss
bar is connected to the concrete by a contact element, it is necessary to define separate nodes and
contact elements for each truss bar at the intersection location.

Figure 66: Example of contact element creation entry

3. Click Add or type Enter. The generated elements are drawn in the Workspace, the total number
of elements is updated, and the last element entry is added to the drop list in the dialog box. At
this stage, the elements have no material types assigned to them, and are assigned the following
default colors:
Rectangular, Quadrilateral and Triangular: White
Truss Bar: Magenta
Link: Blue
Contact: Cyan

Enter: Elmt node 1 2 3 4 #elmts delmt dnode #elmts delmt dnode
90 13 33 14 34 5 1 1 1 1 1
Elements:
Rectangular: 12 to 15
Truss: 46 to 40
Contact: 90 to 94

11 12 13 14 15
13 14 15 16 17 18
33 34 35 36 37 38
46 47 48 49 50
90 91 92 93 94
Node Pairs with same coordinates:
13 & 33
14 & 34
15 & 35
16 & 36
17 & 37
18 & 38
19 20 21 22 23 24
156
To Delete Elements, complete the following steps.
1. In the drop list, select the elements to be deleted. The selected element entry appears highlighted.
2. Click Delete. The deleted elements disappear from the Workspace, the total number of elements
is updated and the element entry is deleted from the drop list in the dialog box. Any associated
material type assignments and loads for these elements are also deleted.

9.1.6.1 Viewing Elements Attributes
To view a summary of the attributes of an element, position the mouse cross-hairs within the
boundaries of the desired element and click the left mouse button. The selected element appears
highlighted in green and the Element Attributes dialog box appears as shown in Figure 67. The element
number, element type, incident nodes and their coordinates, material types and assigned loads are shown.

Figure 67: Element attributes dialog

157
9.1.6.2 Viewing Multiple Element Layers
Occasionally, it is necessary to create layers of elements that occupy the same space. In the
Workspace, only the top layer is visible and conceals the elements beneath it. To view concealed element
layers, it is necessary to send the topmost layer backwards.

To send one element backwards:
1. Position the mouse cross-hairs within the boundary of the element.
2. Click the right mouse button. The context menu appears below the mouse cross-hairs.
3. Select Send This Element Back. The element beneath the mouse cross-hairs is sent to the back,
revealing the element beneath it.

To send multiple elements backwards:
1. Position the mouse cross-hairs within the boundary of any element.
2. Click the right mouse button. The context menu appears below the mouse cross-hairs.
3. Select Send Multiple Elements Back. The Send Elements Back dialog box appears.
4. Click the Send button. The mouse cross-hair becomes a pick arrow.
5. Click the left mouse button and drag pick arrow over the elements to be sent backwards. The
selected elements appear highlighted in bright green.
6. Click Send. The selected elements are sent to the back, revealing the elements beneath them.
7. Click Done.
9.1.7 Assigning Material Types
Having defined elements and material types, the following section describes how to assign the
material types to the elements.
In VecTor2, the following rules exist for assigning material types:
All Rectangular, Quadrilateral and Triangular elements must be assigned a Reinforced Concrete
material type.
158
All Truss Bar elements must be assigned a Reinforcement material type.
All Link and Contact elements must be assigned a Bond type.

To Assign Material Types, complete the following procedure.
1. Select the Structure/Assign Material Type menu item or click the Assign Material
Types toolbar button. The Assign Material Types dialog box appears as shown in
Figure 68.

Figure 68: Assign Material types dialog box

2. Specify the material types and select elements by one of the following two methods.
To Select Elements by Element Number
Complete the entry fields to select the desired elements. The element selection method is similar
to the specification of element numbers. Select a compatible material type from the drop list.
Check the Active box if the selected elements are engaged, so that they contribute to the stiffness
and strength of the structure. Uncheck the box if the selected elements are disengaged, so that
they do not contribute to the strength and stiffness of the structure. In the latter case, all element
strains are computed as plastic strain offsets. Use this feature to disengage and engage elements
for repair applications.

Alternatively, to Select Elements with the Mouse:
Select the material type from the drop list. Click Select. The mouse cross-hairs become a pick
arrow. Click the left mouse button and the drag pick arrow over the elements to be selected. The
selected elements appear highlighted in bright green, as shown in Figure 69. Only elements
compatible with material type may be selected.
159

Figure 69: Selecting elements for material type assignment with the mouse

3. Click Assign or type Enter. The assigned material types are drawn in the Workspace, and the
material type assignments are added to the drop list in the dialog box. Active concrete elements
are filled with solid colors, while inactive concrete elements are filled with hatched colors. Active
truss, link and contact elements are drawn as thicker lines than inactive elements.

Material type assignments can be modified by repeating the above procedure. Previous material
assignments will be overwritten.
9.1.8 Restraining the Structure
Having defined the nodes, the following section describes how to add support restraints.

To Add Support Restraints, complete the following procedure.
1. Select the Structure/Create Support Restraints menu item or click the Create
Support Restraints toolbar button. The Create Support Restraints dialog box appears, as
shown in Figure 70.

Figure 70: Create Support Restraints Dialog Box
160
2. Specify the restraints and select nodes by one of the following two methods.
To Select Nodes by Node Number
Complete the entry fields to select the desired nodes. The node selection method is similar to the
specification of node numbers. Check the X and/or Y box to restrain the selected nodes against
displacements in the x and Y directions, respectively.

Alternatively, to Select Nodes with the Mouse:
Click Select. The mouse cross-hairs become a pick arrow. Position the pick arrow over desired
nodes to restrain and click the left mouse button. The selected nodes appear highlighted in bright
green, as shown in Figure 71. Check the X and/or Y box to restrain the selected nodes against
displacements in the x and Y directions, respectively.

Figure 71: Selecting nodes for restraints with the mouse
3. Click Add or type Enter. The restraints are drawn in the Workspace and the restraint entries are
added to the drop list in the dialog box.

Restraints may be removed by selecting the restraint entry in the drop list and clicking Remove, or
selecting nodes with the mouse and clicking Remove.
161
9.1.9 Automatic Method
The Automatic Method provides a method to define nodes, elements, restraints, and assign
material types without having to manually create and number the nodes and elements. Instead, concrete
regions are defined by one or more polygonal regions. These regions may contain voids, and linear and
point constraints for the mesh. Truss bars elements are defined by linear paths. Link and contact elements
can be created for segments of the discrete reinforcement. Several automatic discretization and meshing
options are provided to accommodate different structure geometries, meshing needs, element types and
the extent of manual control. Material types should be defined before proceeding to the subsequent
sections.

9.1.9.1 Reinforced Concrete Regions
The first step is to define one or more Reinforced Concrete Regions. Each region is defined as a
polygon that will be meshed with rectangular, quadrilateral, and/or triangular elements. Each region can
have different meshing parameters, concrete material type assignments, and numbers of layers. The
guidelines for defining reinforced concrete regions are as follows:
The region is defined by a nontrivial closed polygon.
The vertices defining the region are entered in cyclic counter-clockwise order.
The edges of a region must be simple and not self-intersecting.
Regions can be attached along common edges, but cannot overlap. Care should be taken to define
the common vertices of regions with exactly the same coordinates.
162
Illegal and illegal polygon regions are shown in Figure 72.

Legal Perfect Edge Intersection of Regions
Legal
Illegal Imperfect Edge
Intersection of Regions Illegal Self-Intersection Illegal Clockwise Definition
Figure 72: Examples of legal and illegal region definitions and intersections

To Define a Concrete Region, complete the following steps.
1. Select the Structure / Define and Mesh Structure menu item or select the Define
and Mesh Structure toolbar button. The Define and Mesh Structure property sheet
appears as shown in Figure 73, with the RC regions property page as the active page.
163

Figure 73: RC Regions Property Page
2. Click Create New Region to start a new region. To abort the region at any time, click the same
button, which is now labeled as Cancel New Region.
3. In the Vertices and Restraints group, complete the following procedure to define the region. An
example is shown in Figure 74 to create a rectangle with one restrained edge.
i) Enter the following fields to define a vertice of the region. It is necessary to define only the
vertices at the ends of edges of the region. If desired, one can also define intermediate
vertices to enforce the location of boundary nodes. By doing so, and selecting the Do Not
164
Discretize Boundary manual override option, it is possible to designate the location of all
boundary nodes.
X
Enter the x-coordinate of a vertice, in millimeters.
Y
Enter the y-coordinate of a vertice, in millimeters.
Reps
Enter the number of vertices to create along a line.
Dx
Enter the spacing of vertices in the x-direction from X, in millimeters.
Dy
Enter the spacing of vertices in the y-direction from Y, in millimeters.
Restrain Edge X,Y
Check the appropriate degree of freedom to restrain all nodes on the edge defined by the
current vertice and the next vertice against displacements in the x and/or y directions.
ii) Click +after each vertice entry, except the last vertice. After the last vertice, click Close. The
region is drawn in the Workspace as vertices are added.

(0, 1000) (3000, 1000) (2000, 1000) (1000, 1000)
Nodes to be
created by
FormWorks
restrained in the
x direction
Enforced intermediate
nodal locations
(0,0) (3000,0)
Entry: X Y Reps Dx Dy Restrain Edge
0. 0. 1 0. 0. X Y +
3000. 0. 1 0. 0. Xa Y +
3000. 1000. 3 -1000. 0. X Y +
0. 1000. 1 0. 0. X Y Close
Figure 74: Example of reinforced concrete region creation

165
4. In the Material Layers group, complete the following procedure to add a layer of concrete
elements. In general, only one layer of elements is necessary. More than one layer may be added
to double-meshed a region. Each layer of the region may be assigned different reinforced concrete
material types. If None is selected, FormWorks creates a layer of elements without any material
type assignments. If no layers are added, FormWorks creates no elements, whatsoever.
i) In the drop list, select the reinforced concrete material type to assign the elements of the
region.
ii) Check or uncheck Active if the layer of elements is to be engaged or disengaged,
respectively.
iii) Click + to add the layer of elements with the selected material type. (Click to delete defined
layers from the drop list).
iv) Repeat steps i) to iii) to add additional layers of elements, if desired.
5. In the Discretization and Mesh Type group, select from the following options.
(Note that each region can have different discretization and mesh types.)

The following options determine the manner in which
the FormWorks discretizes the region.
Hybrid
This option is recommended for structures
comprised primarily by one or more rectangular
regions, as is encountered in the elevation of
common reinforced concrete beams, columns and
walls. The discretization is shown for a simple
region in Figure 75. Like the Grid Superposition
algorithm, FormWorks attempts to discretize the
Figure 75: Hybrid Discretization

166
region with a highly regular grid of nodes. Like the Point Insertion Algorithm, it
accommodates nonlinearity of the region boundary and interfering features with local
variations in node spacing. A constrained Delaunay algorithm triangulates the inserted nodes.

Grid Superposition
This option is recommended for polygons of arbitrary shape, having relatively coarse
boundary features. The discretization for a simple region is shown in Figure 76. FormWorks
superimposes the region with a regular grid of rectangular elements, and then removes
elements that are too close to the boundary or other
interfering features within the polygon. The
boundary nodes of the rectangular grid are then
triangulated with a constrained Delaunay
algorithm. This meshing procedure was
liberally adapted from the concepts described by
Petersen, Rodrigues and Martins (2000).
Figure 76: Grid Superposition Discretization

Division Point Insertion
This option is recommended for polygons of arbitrary shape, having several nonlinear edges.
The discretization for a simple region is shown in Figure 77. FormWorks discretizes the
region by inserting nodes at approximately equal
intervals in the x and y direction. Nonlinearity of the
region boundary and interfering features within the
polygon are accommodated by local variations the
node spacing. A constrained Delaunay algorithm
triangulates the inserted nodes.

Figure 77: Division Point Insertion Discretization
167

The following options determine how FormWorks treats the triangular elements created by
the Delaunay triangulation.
Rectangles
If checked, FormWorks merges adjacent pairs of
triangles to create as many rectangular elements of
suitable quality. If the Quadrilaterals option is also
selected, rectangular elements are formed
preferentially over quadrilateral elements. Figure 78
shows the Hybrid discretization of Figure 75 with
the rectangle option
Figure 78: Hybrid Discretization with Rectangle Option

Quadrilaterals
If checked, FormWorks merges adjacent pairs of triangular elements to form quadrilateral
elements. FormWorks selects the best quality quadrilaterals based upon the resulting aspect
ratio and internal angles. As many quadrilateral
elements as possible are created to minimize the
number of residual triangular elements. If the
Rectangles option is also selected, rectangular
elements are created preferentially over
quadrilateral elements. Figure 79 shows the
Division point insertion discretization of Figure 77
with the quadrilateral option.
Figure 79: Division Point Insertion Discretization with Quadrilateral Option



168
Smoothed Triangles
If checked, the shape of the triangular elements are
optimized by adjusting the position of free nodes
by iterative Lagrangian averaging, such that final
position of each node is approximately an average
of the position of its contiguous nodes. Figure 80
shows the Division point insertion discretization of
Figure 77 with the smoothed triangles option.
Figure 80: Division Point Insertion Discretization with Smoothed Triangles Option

6. In the Mesh Parameters group, specify the parameters.
Elmt Size X and Elmt Size Y
For grid superposition and hybrid discretization types, enter the average length of elements in
the x and y directions, respectively, in millimeters.
Tri Elmt. Size
For the division point insertion discretization type, enter the average side length of triangular
elements, in millimeters.
Max. Aspect Ratio
For grid superposition and hybrid discretization types, enter the maximum permitted ratio of
the rectangular element height to width, or vice versa. If the maximum aspect ratio is
exceeded, Elmt Size X and Elmt Size Y are automatically adjusted to the largest possible
dimensions that satisfy the maximum aspect ratio.
Complexity Factor
Enter the decimal fraction of the element size used for the minimum permissible distance
between nodes created by the FormWorks discretization. As the complexity factor decreases,
nodes may be created close the boundary region and other interfering features within the
region. A smaller value may result in higher mesh quality for regions having fine boundary
169
features, such as tight re-entrant corners or small internal angles, or for regions having a
layout of reinforcement lines that is congested relative to the element size.
7. In the Manual Overrides group, specify the following overrides.
Do Not Discretize Boundary
Check this option to prevent FormWorks from creating additional nodes on the region
boundary, other than those specified as intermediate vertices. Use this option to control the
boundary discretization.
Do Not Discretize Interior
Check this option to prevent FormWorks from discretizing the interior of the region with
additional nodes, other than those specified by reinforcement lines, voids, and line and point
constraints. Use this option to control the interior discretization.
Do Not Adjust Elmt Size
Check this option to prevent FormWorks from adjusting the Elmt Size X and Y, regardless of
the maximum aspect ratio specified. When using the grid superposition discretization type,
utilize this option with the Superposition Corner Origin override to control the placement of
the grid.
Do Not Refine Near Truss
For the grid superposition discretization type, check this option to prevent FormWorks from
creating nodes offset from vertical and horizontal segments of reinforcement lines. The
purpose of the offsetting is to create more regularly spaced rectangular elements near truss
elements.
Superposition Corner Origin
For the grid superposition discretization type, specify the corner of the bounding rectangle of
the region from which the superposition grid originates. This option has no effect unless the
Do Not Adjust Elmt Size option is checked.
8. Click Add Region to store the region. The region is added to the list.

170
To Modify a Region, complete the following steps.
1. Select the region in the list box. The selected region appears highlighted and its current attributes
are displayed in the entry fields.
2. Modify the attributes of the region. Note that the vertices of the region cannot be changed.
3. Click Update Region to store the changes.

To Delete a Region, complete the following steps.
1. Select the region in the list box. The selected region appears highlighted and its current attributes
are displayed in the entry fields.
2. Click Delete Region. The selected region is removed from the list box and disappears from the
Workspace. The remaining regions are renumbered. Alternatively, click Delete All to remove all
defined regions.

9.1.9.2 Reinforcement
The second step is to define Reinforcement Paths, if any. Each reinforcement path is defined by
a series of line segments, which will be meshed with truss bar elements. If desired, segments of the path
can be attached with link or bond elements to the concrete elements. Each reinforcement path can have
several layers of truss bar and bond elements, with different material type assignments. The reinforcement
path assumes the meshing parameters of the surrounding concrete and the mesh will be constrained to the
reinforcement path. The guidelines for defining reinforcement paths are as follows:
The reinforcement path cannot intersect itself.
The reinforcement path may traverse boundaries of regions and voids. Reinforcement paths
may intersect each other. It is not necessary to define the points of intersection.
Only segments of paths that are contained inside or on a reinforced concrete region are meshed.
To create a bare bar, it is necessary to surround the bar with a region having no material layers.


171
To Define a Reinforcement Path, complete the following steps.
1. Select the Reinforcement property page of the Define and Mesh Structure property sheet, as
shown in Figure 81.

Figure 81: Reinforcement Page
2. Click Create New Truss to start a new reinforcement path. To abort the path at any time, click
the same button, which is now labeled as Cancel New Truss.
3. In the Bond Information group, specify the following options.
Truss Reinforcement Is Perfectly Bonded Over Entire Length
172
Check this option if the truss bar elements created over the whole reinforcement path are to
be connected directly to the nodes of the concrete elements. Uncheck this option if any
segment of the path is to be connected with bond elements.
Attach Imperfectly Bonded Truss Segments to Concrete with
If the above selection is unchecked, select either Link or Contact elements as the type of
bond element for this reinforcement path.
4. In the Vertices group, complete the following procedure to define the reinforcement path. An
example is shown in Figure 82 to create a rectangle with one restrained edge.
i) Enter the following fields to define a vertice of the reinforcement path. It is necessary to
define only the vertices at the ends of the path segments. If desired, one can also define
intermediate vertices to enforce the location of nodes defining the truss bar elements. Indeed
by doing so, and selecting the Do Not Discretize Truss manual override option, it is possible
to designate the discretization of all truss bar elements on the path.
The X, Y, Reps, Dx and Dy fields have the same meaning as for defining regions.
Imperfect Bond
If checked, all truss elements between the current vertice and the next vertice will be
connected with bond elements. Only segments between two successive perfectly bonded
vertices are attached directly to the concrete elements.
ii) Click + after each entry. The reinforcement path is drawn in the Workspace.

173

Figure 82: Example of reinforcement path creation
5. In the Material Layers group complete the following procedure to add layers of truss bars
elements and contact elements (if imperfectly bonded). More than one layer may be added to
double mesh the path. Each layer of the reinforcement path may be assigned different
reinforcement and bond material types. If None is selected, FormWorks creates a layer of
elements without any material type assignments. If no layers are added, FormWorks creates no
elements, whatsoever.
i) In the Truss Material drop list, select the reinforcement material type to assign to the truss
elements.
ii) In the Bond Material drop list, select the bond material type to assign to the link or contact
elements, if the reinforcement path is not perfectly bonded along its entire length.
iii) Click + to add a layer of elements to the reinforcement path with the currently selected
material type. (Click to delete the defined layers from the drop list.)
iv) Repeat steps i) to iii) to add additional layers of elements (e.g. double-meshing).
6. In the Manual Overrides group, specify the following overrides.
Do Not Discretize Truss
Check this option to prevent FormWorks from creating additional nodes on the reinforcement
path, other than those specified as vertices. Use this option to control the path discretization.
Truss elements created in this path
segment to be connected with
bond elements
Existing reinforced
concrete region
(50,65) (450,65) (3000,65)
Entry: X Y Reps Dx Dy Imperfect Bond
50. 65. 1 0. 0. a +
450. 65. 1 0. 0. +
3000. 65. 1 0. 0. +
174
7. Near the top of the page, complete the following entry fields to create multiple reinforcement
paths, offset by constant displacements. This is a convenient way to define repetitive
reinforcement details, such as beam stirrups and column ties, without having to separately enter
the vertices of each reinforcement path.
Reps
Enter the number of repetitions of the reinforcement path.
Dx
Enter the translation in the x direction between successive paths, from the defined vertices.
Dy
Enter the translation in the y direction between successive paths, from the defined vertices.
8. Click Add Truss to store the reinforcement path(s). The reinforcement path(s) are added to the
list and appear in the Workspace as red lines.

To Modify a Reinforcement Path, complete the following steps.
1. Select the reinforcement path in the list box. The selected path appears highlighted and its current
attributes are displayed in the entry fields.
2. Modify the attributes of the region. With regard to the path itself, vertices can only be added.
3. Click Update Truss to store the changes.

To Delete a Reinforcement Path, complete the following steps.
1. Select the reinforcement path in the list box. The selected reinforcement path appears highlighted
and its current attributes are displayed in the entry fields.
2. Click Delete Truss. The selected reinforcement path is removed from the list box and disappears
from the Workspace. The remaining reinforcement paths are renumbered. Alternatively, click
Delete All Truss to remove all defined reinforcement paths.

175
9.1.9.3 Voids
The third step is to define Voids, if any, in the reinforced concrete regions, if any. As examples,
voids may represent an opening such as a window, the space between framing members, or the cavities of
cellular structures. Like Concrete Regions, each void is defined as a polygon. No elements of any type are
created within the void. The void boundary assumes the meshing parameters of the surrounding concrete.
The guidelines for defining voids are as follows:
The void is defined by a nontrivial closed polygon.
The vertices defining the voids are defined in cyclic clockwise order.
The edges of the void must be simple and not self-intersecting.
Voids can transverse reinforced concrete regions and reinforcement paths. It is not necessary to
define intersection points.

To Define a Void, complete the following steps.
1. Select the Voids & Constraints property page of the Define and Mesh Structure property sheet,
as shown in Figure 83.
2. In the Voids group, click Create New Void to start a new void. To abort the void at any time,
click the same button, which is now labeled as Cancel New Void.
3. Complete the following procedure to define the void.
i) Enter the fields defining the location of the void vertices, in clockwise order. It is necessary to
define only the vertices at the ends of the void segments. If desired, one can also define
intermediate vertices to enforce the location of nodes on the void boundary. Indeed, by doing
so, and selecting the Do Not Discretize manual override option, it is possible to designate the
location of all nodes on the void boundary.
The X, Y, Reps, Dx and Dy fields have the same meaning as for defining regions.
ii) Click + after each vertice entry, except the last. After the last vertice, click Close.


176

Figure 83: Voids and Constraints

4. Click the Do Not Discretize option to prevent FormWorks from creating additional nodes on the
void boundary other that those specified as vertices. Use this option to control the void boundary
discretization.
5. Near the top of the Void group, complete the following entry fields to create multiple voids,
offset by constant displacements. This is a convenient way to define repetitive voids, such as
177
evenly spaced equal-sized openings in frame structures, without having to separately enter the
vertices of each void.
Reps
Enter the number of repetitions of the void.
Dx
Enter the translation in the x direction between successive voids, from the defined vertices.
Dy
Enter the translation in the x direction between successive voids, from the defined vertices.
6. Click Add Void to store the void(s). The void(s) are added to the list and appear in the
Workspace as polygons composed of dashed blue lines.
To Delete a Void, complete the following steps.
1. Select the void in the list box. The selected void appears highlighted and its current attributes are
displayed in the entry fields.
2. Click Delete Void. The selected void is removed from the list and disappears from the
Workspace. The remaining voids are renumbered. Alternatively, click Delete All Voids to
remove all defined voids.

9.1.9.4 Line Constraints
The fourth step is to define Line Constraints, if any, in the reinforced concrete regions. Line
constraints are used to create material type boundaries or any other type of conformal edge inside within
the reinforced concrete region. FormWorks constrains element edges to line-up with line constraints. As
such, it is not necessary to define reinforced concrete regions for every localized change in material type.
Instead, larger regions containing several regions can be defined and different material types can be
assigned within the region after the mesh is created. Line constraints assume the meshing parameters of
the surrounding concrete. The guidelines for creating line constraints are the same as those for
reinforcement paths.

178
To Define a Line Constraint, complete the following steps.
1. Select the Voids & Constraints property page of the Define and Mesh Structure property sheet,
as shown in Figure 83.
2. In the Line Constraints group, click Create New Line to start a new line constraint. To abort the
constraint at any time, click the same button, which is now labeled as Cancel New Line.
3. Define a line constraint in the same manner that reinforcement paths are created, by entering the
location of vertices and clicking +.
4. Click the Do Not Discretize box to prevent FormWorks from creating nodes along the line
constraint, other than those specified as vertices. Use this option to control the line constraint
discretization.
5. Near the top of the Line Constraints group, complete the Reps, Dx, and Dy entry fields to create
multiple line constraints, offset by constant displacements. This is a convenient way to define
repetitive line constraints, without having to separately enter the vertice of each line constraint.
6. Click Add Line to store the line constraint(s). The line constraint(s) are added to the list and
appear in the Workspace as dashed green lines.

To Delete a Line Constraint, complete the following steps.
1. Select the line in the list box. The selected line appears highlighted and its current attributes are
displayed in the entry fields.
2. Click Delete Line. The selected line is removed from the list and disappears from the Workspace.
The remaining lines are renumbered. Alternatively, click Delete All Lines to remove all defined
line constraints.

9.1.9.5 Point Constraints
The fifth step is to define Point Constraints, if any, in the reinforced concrete regions. A node is
always created at the location of a point constraint. Point constraints have three possible uses. First, point
constraints can be used to specify the exact locations of restraints, nodal loads or support displacements
179
acting on the structure. Second, point constraints can be used to locally refine a mesh by placing nodes at
a finer resolution than the specified element sizes for enclosing reinforced concrete region. Finally, it is
possible to fully control the mesh discretization by specifying the location of all nodes with point
constraints and selecting the Do Not Discretize override options in the preceding sections. This technique
would resemble the Manual Method for creating nodes and elements, except that neither the nodes nor
elements need to be manually numbered. The guidelines for creating point constraints are as follows:
Point Constraints may be placed anywhere, including a reinforced concrete region, on region
boundaries, on reinforcement paths, on void boundaries, or on line constraints.
Point constraints that are not inside a reinforced concrete region are ignored.

To Define Point Constraints, complete the following steps.
1. In the Point Constraints group, complete the following fields. Figure 84 shows an example.
X
Enter the x coordinate of the point constraint, in millimeters.
Y
Enter the y coordinate of the point constraint, in millimeters.
Reps
Enter the number of points in the row/column.
Dx
Enter the spacing of successive points of a row/column in the x direction, in millimeters.
Dy
Enter the spacing of successive points of a row/column, in the y direction, in millimeters.
Restrain X
Check this option to create a support restraint in the x direction for the point constraint(s).
Restrain Y
Check this option to create a support restraint in the y direction for the point constraint(s).
2. Click + to store the point constraint(s). Point constraints appear in the Workspace as green dots.
180
Reinforced concrete region
Local mesh refinement
Line constraint
Reinforcement Path
Pin support location
Entry: X Y Reps Dx Dy Reps Dx Dy Restrain
100. 0. 1 0. 0. 1 0. 0. X a Y a +

2750. 800. 5 50. 0. 4 0. 50. X Y +
Figure 84: Example of point constraint creation

To Delete a Point Constraint, complete the following steps.
1. Select the point constraint in the drop list. The selected point constraint appears highlighted.
2. Click . The selected point is removed from the list and disappears from the Workspace. Clicking
will remove point constraints one at a time. Alternatively, click All, to remove all the point
constraints at once.

9.1.9.6 Generating the Mesh
Once the concrete regions, reinforcement paths, voids, line constraints and point constraints have
been defined, the mesh can be generated.

To Generate the Mesh, complete the following steps.
1. Select the Create Mesh property page of the Define and Mesh Structure property sheet, as
shown in Figure 85.
181

Figure 85: Mesh Creation Page
2. Check the Bypass Edge Constraint Check to speed-up the meshing process by not explicitly
checking the conformity of element edges with the various boundaries and constraints. Generally,
this option should not be checked, especially for regions that are not strictly convex.
3. Click Create Mesh. The meshing generation proceeds until complete as indicated by the status
bar. In the Mesh Information group, the Mesh column indicates the number of nodes, elements
and restraints required by the meshed region. The Total column indicates these values plus the
182
number of nodes, elements and restraints already defined in the Structure Data. The Limit Check
indicates whether the Structure Limits have been exceeded.
4. Inspect the resulting mesh in the Workspace view. If the Structure Limits have been exceeded or
elements are misshapen, modify the meshing parameters or redefine the regions, reinforcement
paths, voids, and constraints as necessary. Then, repeat Steps 1 to 3. When using the Grid
Superposition method or Division Point Insertion discretization types, judiciously adding point
constraints often provides a simple correction.
5. If the mesh is acceptable, click Add Mesh to Structure and close the Define and Mesh
Structure property sheet. The nodes, elements and restraints are then stored in the Structure Data.
The regions, reinforcement paths, voids and constraints are stored for future revision whenever
the Define and Mesh Structure property sheet is reopened.
6. If necessary, refine material type assignments and add support restraints to the new mesh using
the Manual Method to complete the Structure Data.
183
10 The Load Case Data
The third step in creating the input is to define the data for one or more load cases. As previously
described, all the loads belonging to a load case are multiplied by the load factors defined for that load
case in the J ob Data.
10.1.1 Load Limits
VecTor2 limits the number of loads acting on nodes and elements in each load case. These limits
vary with the version of VecTor2. To view these limits, select the Load/Load Limits menu item. The
Load Limits dialog box appears as shown in Figure 86.

Figure 86: Load Limits Dialog Box
10.1.2
10.1.3
Selecting the Load Case
Before assigning any loads, select the Load/Select
Load Case [n] menu item or click the Load Case [n] tool bar button to choose the load case to which the
loads will be added. Only the load cases that are active in the J ob Data may be selected.
Load Information
Select the Load/Load Information menu item or click the Load Information toolbar
button to view the number of defined loads of each type for the current load case. The Load
Information dialog appears as shown in Figure 87.
184

Figure 87: Load Information Dialog Box
10.1.4 Nodal Loads
These loads are concentrated forces acting in either the x or y directions, on individual nodes.

To Add Nodal Loads, complete the following procedure.
1. Select the Load/Apply Nodal Loads menu item or click the Nodal Loads toolbar button. The
Apply Nodal Loads dialog box appears as shown in Figure 88.

Figure 88: Apply Nodal Loads Dialog Box
185

2. Specify the nodes and nodal loads by one of the following two methods.
To Select Nodes by Node Number
Complete the Node, #nodes and dnode entry fields to select the desired nodes. The node
selection method is similar to specifying node numbers when creating nodes. Specify the loads.
Fx
Enter the force acting on Node in the x direction, in kN. Positive values act in the direction of the
positive x axis, while negative values act towards the negative x axis.
Fy
Enter the force acting on Node in the y direction, in kN. Positive values act in the direction of the
positive y axis, while negative values act towards the negative y axis.
dFx
Enter the increment of the force Fx, acting on successively selected nodes, in kN.
dFy
Enter the increment of the force Fy, acting on successively selected nodes, in kN.

Alternatively, to Select Nodes with the Mouse:
Click Select. The mouse cross-hairs become a pick arrow. Position the pick arrow over nodes to
be selected and click the left mouse button. The selected nodes are highlighted in bright green.
Enter the values for Fx and Fy, as described above. The same force is applied to all selected
nodes.
3. Click Apply or type Enter. The applied nodal loads are drawn in the Workspace and added to the
drop list in the dialog box.

Nodal loads can be modified by repeating the above procedure. Previously applied nodal loads
will be overwritten. Nodal loads can be deleted by selecting them either from the drop list or by using
the mouse, then clicking Delete.
186
10.1.5 Support Displacements
These loads are imposed displacements of nodes of the structure.
Select the Load/Apply Support Displacements menu item or click the Support Displacements
toolbar button. The Apply Support Displacements dialog box appears as shown in Figure 89.

Figure 89: Apply Support Displacements Dialog Box

The procedure for adding support displacements is similar to that for adding nodal loads. The unique
entry fields are described as follows.
D.O.F. X,Y
Select either the X or Y direction for the nodal displacement. Separate entries must be made to
displace the node is both degrees of freedom.
Displacement
Enter the imposed displacement of the nodes, in millimeters. Enter positive values for
displacements in the direction of the positive X or Y axis, and negative values for displacements
in the direction of the negative X or Y axis.
10.1.6 Gravity Loads
Gravity loads are applied to rectangular, quadrilateral and triangular elements to include the
weight, W, of the element, as follows.
V GX W
x
+ = 81 . 9 (10.1.6.1a)
V GY W
y
= 81 . 9 (10.1.6.1b)
187
where, W
x
and W
y
are the components of the weight acting in the X and Y directions, respectively, GX
and GY are the G-forces acting in the X and Y directions, respectively, is the mass density of the
reinforced concrete, and V is the volume of the element. The weight components are then evenly
distributed as nodal loads among the nodes of the element.

To Apply Gravity Loads, complete the following procedure.
1. Select the Load/Apply Gravity Loads menu item or click the Gravity Loads toolbar button.
The Apply Gravity Loads dialog box appears as shown in Figure 90.

Figure 90: Apply Gravity Loads Dialog Box

2. Specify the elements and gravity loads by one of the following methods:
To Select Elements by Element Number
Complete Elmt, #elmts and delmt entry fields to select the desired concrete elements. The
element selection method is similar to specifying element numbers when creating elements.
Specify the loads.
Dens
Enter the mass density of the reinforced concrete element, in kg/m
3
.
Gx
Enter the G-force acting in the x direction. Enter positive values for gravity forces acting in
the direction of the positive x-axis. (If the x-axis of the Workspace is parallel to the ground,
this value is typically zero. Otherwise, if the coordinate axis has been rotated so that the x-
axis is not parallel to the ground, a nonzero value can be used to specify a component of the
gravitational force.)
188
Gy
Enter the G-force acting in the y direction. Enter positive values for gravity forces in the
direction of the negative Y-axis. (If the y-axis is perpendicular to the ground, this value is
typically positive one. Otherwise, if the coordinate axis has been rotated so that the y-axis is
not perpendicular to the ground, a nonzero value can be used to specify a component of the
gravitational force.)

Alternatively, to Select Elements with the Mouse:
Click Select. The mouse cross-hairs become a pick arrow. Click the left mouse button and drag
the pick arrow over the elements to be selected. The selected elements are highlighted in bright
green. Enter the values for Gx, Gy and Dens as described above.
3. Click Apply or type Enter. The applied gravity loads are drawn in the Workspace and added to
the drop list in the dialog box.

Gravity loads can be modified by repeating the above procedure. Previously defined gravity loads
will be overwritten. Gravity loads are deleted by selecting them either from the drop list or by using the
mouse, then clicking Delete.
10.1.7 Temperature Loads
Temperature loads are applied to rectangular, quadrilateral, and triangular elements and truss bar
elements. The applied temperature loads model only expansion and contraction effects of temperature
gradients. VecTor2 incorporates temperature loads by converting them to elastic strain offsets, as follows:
[ ] [ ]
T
c c
T
o
cxy
o
cy
o
cx
T T 0 = + (10.1.7.1)
T
s
o
s
= (10.1.7.2).
189
where and are elastic strain offsets for the concrete and reinforcement, respectively,
o
c

o
s

c
and
s
are
the coefficients of thermal expansion for concrete and reinforcement (defined as material properties) and
T is the temperature load acting on the element.

Select the Load/Apply Temperature Loads menu item or click the Temperature Loads toolbar
button. The Apply Temperature Loads dialog box appears as shown in Figure 91.

Figure 91: Apply Temperature Loads Dialog

The procedure for adding temperature loads is similar to that for adding gravity loads. The unique
entry fields are described as follows.
Temp
Enter the temperature, T, of the source element with respect to the temperature of unloaded
elements, in C.
dtemp
Enter the increment in temperature, in C, between successive elements in the row/column.

Temperature loads can be modified by repeating the above procedure. Previously defined
temperature loads will be overwritten. Temperature loads are deleted by selecting them either from the
drop list or by using the mouse, then clicking Delete.

190
10.1.8 Concrete Prestrains
Concrete prestrain loads are applied to rectangular, quadrilateral and triangular
elements. Example applications of concrete prestrains include negative shrinkage strains or positive
expansive strains due to alkali-aggregate reactivity. VecTor2 directly incorporates the prestrain load as an
elastic strain offset as follows:
[ ] [ ]
T
ps ps
T
o
cxy
o
cy
o
cx
0 = + (10.1.8.1).
where is elastic strain offsets of the concrete, and
o
c

ps
is the applied prestrain load. As such, the load
translates the concrete stress-strain response by a strain amount equal to the specified prestrain.

Select the Load/Apply Concrete Prestrains menu item or click the apply Concrete Prestrains
toolbar button. The Apply Concrete Prestrains dialog box appears as shown in Figure 92.

Figure 92: Apply Concrete Prestrains Dialog Box

The procedure for applying, modifying and deleting concrete prestrain loads is similar to that for
applying temperature loads. The unique entry fields are described as follows:
Strain
Enter the prestrain value, in millistrain. For instance, enter a negative value for shrinkage strains.
d strain
Enter the increment in prestrain between successive elements in a row/column.

191
10.1.9 Ingress Pressures
Ingress pressure loads are applied to rectangular, quadrilateral and triangular elements.
When the element is cracked, the pressure acts on the out-of-plane dimension in a hydrostatic state.
VecTor2 multiplies the specified pressure by the out-of-plane area of each edge of the membrane element
to determine equivalent nodal loads in the x and y directions.

Select the Load/Apply Ingress Pressures menu item or click the Ingress Pressures toolbar
button. The Apply Ingress Pressures dialog box appears as shown in Figure 93.

Figure 93: Apply Ingress Pressures Dialog Box

The procedure for applying, modifying and deleting ingress pressures is similar to that for
applying temperature loads. The unique entry fields are described as follows:
Pressure
Enter the ingress pressure acting on the element, in MPa.
dPressure
Enter the increment in ingress pressure between successive elements selected in a row/column.
10.1.10 Surface Thermal Loads
Surface thermal loads are time varying temperature loads applied to the free surface of
rectangular, quadrilateral and triangular elements. These loads are to be used with the time step analysis
mode.
To Apply Surface Thermal Loads, complete the following steps.
192
1. Select the Load/Apply Surface Thermal Loads menu item or click the Surface Thermal Loads
toolbar button. The Apply Surface Thermal Loads dialog box appears as shown in Figure 94.

Figure 94: Apply Surface Thermal Loads Dialog

2. Specify the surface and temperature loads by one of the following methods.
To Select Surfaces by Node Number
Complete the following entry fields.
Node 1 and Node 2
Enter the two node numbers describing a free surface of a rectangular, quadrilateral or
triangular element.
NumSurf
Enter the number of surfaces to which loads are applied.
Dnode
Enter the increment in node number between successive surfaces.
Tmi
Enter the time, in seconds, at which the temperature Tpi occurs.
Tpi
Enter the temperature, in C, corresponding Tmi.

Alternatively, to Select Surfaces with the Mouse:
Click Select. The mouse cross-hairs become a pick arrow. Position the pick arrow over the free
surfaces of the structure to be selected and click the left mouse button. The selected free-surfaces
are highlighted in bright green. Enter the values of Tmi and Tpi as described above.
193
3. Click Apply or type Enter. The applied surface thermal loads are drawn in the Workspace and
added to the drop list in the dialog box.

Surface thermal loads can be modified by repeating the above procedure. Previously defined
surface thermal loads will be overwritten. Surface thermal loads are deleted by selecting them either from
the drop list or by using the mouse, then clicking Delete.
10.1.11 Impulse Loads
Impulse loads are time varying nodal forces. These loads are to be used with the time step
analysis mode.

Select the Load/Apply Impulse Loads menu item or click the Impulse Loads toolbar button.
The Apply Impulse Loads dialog box appears as shown in Figure 95.

Figure 95: Apply Impulse Forces Dialog Box

The procedure for applying impulse loads is similar to that for applying nodal loads and support
displacements. The unique entry fields are described as follows.
dof
Select either the x or y direction in which the impulse force acts.
Ti
Enter the time, in seconds, corresponding Fi.
Fi
Enter the force, in kN, corresponding Ti. Positive forces act in the direction of the positive
axes.
194

Impulse loads can be modified by repeating the above procedure. Previously defined impulse
loads will be overwritten. Impulse loads are deleted by selecting them either from the drop list or by using
the mouse, then clicking Delete.
10.1.12 Ground Acceleration Loads
Ground acceleration loads may be specified to subject the structure to seismic-type loads. These
loads are to be used with the time step analysis mode.

To Ground Acceleration Loads, complete the following steps.
1. Select the Load/Apply Ground Acceleration menu item or click the Ground Acceleration
toolbar button. The Apply Ground Acceleration dialog box appears as shown in Figure 96.

Figure 96: Apply Ground Acceleration Load Dialog
2. Specify the ground acceleration record by completing the following fields.
Time
Enter the time, in seconds.
Acc X, Y
Enter the ground acceleration in the x and y directions, in m/s
2
, occurring at the above time.
3. Click Add or type Enter.
4. Repeat Steps 2 to 4 to enter the remainder of the ground acceleration record. The acceleration
record should conclude with a final entry with a time of 99 999 seconds.
195
11 Running VecTor2
Once the J ob, Structure and Load Case Data have been properly defined, it is possible to start the
VecTor2 analysis. Complete the following steps to proceed with the VecTor2 analysis.
11.1 Starting the Analysis
1. As described in Section 7.4.3, save the FormWorks Workspacefile by selecting the File/Save
menu item or clicking the Save toolbar item. It is recommended that .FWS file be saved in the
directory containing the FormWorks application.
2. Select the Analysis/Run VecTor2 Processor menu item or click the Run VecTor2
Processor toolbar button.
3. FormWorks presents the option to attempt to reduce the bandwidth. A reduced bandwidth
decreases the computation time by renumbering the nodes in a more computationally efficient
manner. If No is selected, proceed to Step 4. If Yes is selected the Bandwidth Reduction dialog
appears as shown in Figure 97. Complete the following procedure.

Figure 97: Bandwidth Reduction Dialog Box
196
i) Click Start. FormWorks determines the bandwidth of the mesh and displays it as Original
Bandwidth.
ii) Enter a value for the Target Maximum Bandwidth. Choose a trial value less than the
original bandwidth.
iii) Enter a value for the Maximum No. of Iterations. This controls how many iterations
FormWorks attempts before pausing the bandwidth reduction algorithm. Begin with a value
approximately equal to one tenth of the number of nodes in the structure.
iv) Click Resume. The bandwidth reduction proceeds and stops when either the bandwidth is
less than the Target Maximum Bandwidth, or the Maximum Number of Iterations is reached.
v) Inspect the Current Bandwidth of the structure. The iterative bandwidth reduction algorithm
used by FormWorks may converge slowly for some numbering schemes, resulting in an
initial increase of bandwidth followed by substantial reductions. In this case, additional
iteration is required. For some numbering schemes, it may actually increase the bandwidth. In
this case, click No and proceed to Step 4.
vi) Repeat steps ii to v, and until an acceptable bandwidth is achieved or further iterations do not
reduce the bandwidth.
vii) If the bandwidth reduction is acceptable, click Yes. Otherwise, click No to proceed with the
original node numbering.
4. The Save Job File dialog appears. The job file must be saved as Vector.job for analysis. Select
the directory corresponding to that containing the FormWorks and VecTor2 program and click
Save.
5. The Save Structure File dialog appears. Accept the structure file name, as assigned in the J ob
Data. Click Save.
6. For each load case, the Save Load Case File dialog appears. Accept the load case file name, as
assigned in the J ob Data. Click Save.

197
The VecTor2 analysis proceeds as shown in Figure 98. Providing there are no errors in the input, the
analysis proceeds until all specified load steps are performed, or until the stiffness matrix is no longer
invertible.

Figure 98: VecTor2 Analysis Proceeding
11.2 A Simple Example
A simple example is presented in this section for the FormWorks modeling procedure. This example
is not intended to rigorously analyze and interpret the results, but rather to illustrate a possible means of
discretization and model selection.
It is proposed to determine the ultimate load and corresponding deflection of the reinforced concrete
beam shown in Figure 99 subject to center point loading, as tested by Shim (2002). VecTor2 input files,
which are equivalent to those used in this example, are provided in Appendix A.
The Hognestad model was selected for both the pre and post-peak compression response, with the
Vecchio 1992-A compression softening model. The Bentz 2003 model was selected for tension stiffening.
A crack width limit of 2 mm was imposed and element slip-distortions were included with the Vecchio-
Lai model.

198
220

Figure 99: Reinforced Concrete Beam

It is decided to model the beam with rectangular elements for the concrete, and truss bar elements for
the longitudinal reinforcing bars. Two reinforced concrete material types are utilized. One type represents
the plain concrete cover. The other type models the web region of the beam with one smeared
reinforcement component, which represents the stirrup reinforcement. Three ductile steel reinforcement
material types are utilized; one each for the pair of No. 10 bars, the pair of No. 25 bars, and the pair of
No. 30 bars.
As both the beam and the loading conditions are symmetrical about the midspan, only half of the
beam needs to be modeled. The automatic mesh generation facility with the hybrid discretization type was
used to create the mesh shown in Figure 100. Each pair of longitudinal reinforcing bars was entered as a
separate reinforcement path with its corresponding material type. Nodes at the midspan of the beam are
restrained from displacements in the longitudinal direction. The node at the support is restrained from
displacements in the transverse direction.
63
63
63
361
2- No. 10, As =200 mm
2
, db =11.3 mm, fy =315 MPa, fu =460 MPa
2- No. 25, As =1000 mm
2
, db =25..2 mm, fy =445,MPa, fu =680 MPa
2 -No. 30, As =1500 mm
2
, db =29.9 mm, fy =436 MPa, fu =700 MPa
D5 @ 210, =0.101%, =90, fy =600MPa, fu =649 MPa
305
Concrete: fc =32.6 MPa, ft =1.819 MPa, eo =1.902 me
All dimensions in mm
220
1830 1830
199

Figure 100: Finite element mesh for half of beam

One load case was utilized to impose a support displacement of 1 mm at the midspan. The load
factor was increased monotonically from zero to failure in increments of 0.25 mm. The self-weight of the
beam is not included.
VecTor2 determines the load vs. midspan deflection response as shown in Figure 101.
0
100
200
300
400
500
0 5 10 15 20
Midspan deflection (mm)
V
e
r
t
i
c
a
l

L
o
a
d

(
k
N
)

Figure 101: Load versus midpsan deflection of reinforced concrete beam

At the ultimate load corresponding deflection of 450 kN and 17 mm, respectively, VecTor2
determines the cracking pattern as shown in Figure 102.

200

Figure 102: Crack pattern of reinforced concrete beam at ultimate load
201
12 Summary & Recommendations
12.1 Summary
User facilities have been presented for the two-dimensional nonlinear finite element analysis
program, VecTor2. These user facilities consist of program documentation for VecTor2 and a newly
developed preprocessor program called FormWorks. With the purpose of providing a basis for rational
usage, the program documentation has addressed the formulation and implementation of the VecTor2
program, its element library and its numerous constitutive models. The FormWorks program has been
developed to expedite the modeling process by providing a graphical interface for data visualization and
input, data checking procedures, a two-dimensional automatic mesh generator, and a bandwidth reduction
algorithm. In order to provide a high degree of user control, the automatic mesh generator is capable of
producing mixed element meshes and accommodating arbitrary material delineations, constraints, and
local variations in element size.
12.2 Recommendations
It is apparent that one of the fundamental challenges facing a user of a nonlinear finite element
program such as VecTor2 is the modeling process itself. Unless a generalized NLFEARC program exists,
it remains necessary to extract a portion of the structure to analyze, determine appropriate boundary and
loading conditions, and select appropriate analysis models. In this regard, the automatic mesh generator
provides convenience utility, but does not provide context to the analysis. This may be achieved by
implementing a catalog of common structural members in a preprocessor program with familiar boundary
conditions and recommended analysis models. Alternatively, it may be beneficial to implement
procedures that harmonize or automatically prescribe material models. Possibly, different material models
could be applied to different mesh regions depending on the predicted stress state.

202
With regards to NLFEARC as a design tool, it would be beneficial to integrate NLFEARC into the
iterative design process. In this regard, the importance of graphical preprocessor facilities to efficiency
should not be discounted, particularly for NLFEARC programs that conduct three-dimensional analyses
or utilize more geometrically complex elements. Conceivably, the input interface could be CAD-based
and include mesh generators dedicated for finite element analysis. Other analysis procedures could also be
integrated to provide a means of verifying results against more conventional analysis methods or accepted
design standards.

203
References
Bentz, E.C., 2000. "Sectional Analysis of Reinforced Concrete Members", Ph.D. Thesis, Department of
Civil Engineering, University of Toronto, 310 pp.

Collins, M.P., and Mitchell, D., 1997. Prestressed Concrete Structures, Response Publications, Canada,
766 pp.

Eligehausen, R., Popov, E., and Bertero, V. 1983. Local Bond Stress-Slip relationship of Deformed Bars
under Generalized Excitations, Report No. UCB/EERC-83/23, Earthquake Engineering Center,
University of California, Berkeley.

Gan, Y., 2000. "Bond Stress and Slip Modeling in Nonlinear Finite Element Analysis of Reinforced
Concrete Structures", M.A.Sc Thesis, University of Toronto, 251 pp.

Gupta, P.R.., 1998. "Shear Design of Reinforced Concrete Members under Axial Compression", Ph.D.
Thesis, Department of Civil Engineering, University of Toronto, 189 pp.

Harajli, M.H. and Mukaddam, M.A., 1988. Slip of Steel Bars in Concrete J oints under Cyclic Loading,
ASCE J ournal of Structural Engineering, Vol. 114, No. 9, pp. 2017-2035.

He, X.G., and Kwan, A.K.H., 2001. Modeling Dowel Action of Reinforcement Bars for Finite Element
Analysis of Concrete Structures, Computers and Structures, Vol. 79, No.6, pp. 595-604.

Hoshikuma, J ., Kawashima, K., Nagaya, K. and Taylor, A.W., 1997. "Stress-Strain Model for Confined
Reinforced Concrete in Bridge Piers", ASCE J ournal of Structural Engineering, Vol. 123, No. 5, pp.
624-633.

204
Hsieh, S.S., Ting, E.C., and Chen, W.F., 1979. An Elastic-Fracture Model for Concrete. Proc. 3
rd
Eng.
Mech. Div. Spec. Conf., ASCE, Austin, Texas, pp. 437-440.

Izumo, J ., Shin. H., Maeakawa, K., Okamura, H., 1992. "An Analytical Model for RC Panels Subjected to
In-Plane Stresses", Concrete Shear in Earthquake, Elsevier Applied Science, London and New York,
pp. 206-215.

Kaufmann, W. and Marti, P., 1998. Structural Concrete: Cracked Membrane Model, ASCE J ournal of
Structural Engineering, Vol. 124, No. 12, pp. 1467-1475.

Kent, D.C., and Park, R., 1971. Flexural Members with Confined Concrete, ASCE J ournal of the
Structural Division, Vol. 97, No. ST7, Proc. Paper 8243, pp. 1341-1360.

Keuser, M. and Mehlhorn, G., 1987. Finite Element Models for Bond Problems, ASCE J ournal of
Structural Engineering, Vol. 113, No. 10, pp. 2160-2173.

Kupfer, H.B., and Gerstle, K.H., 1973. Behaviour of Concrete under Biaxial Stresses, ASCE J ournal of
Engineering Mechanics, Vol. 99, EM4, pp. 853-866.

Kupfer, H., Hilsdorf, H.K.and Rusch, H., 1969. "Behavior of Concrete under Biaxial Stress", ACI
J ournal, Vol. 87, No. 2, pp. 656-666.

Kwon, M. and Spacone, E., 2002. "Three-Dimensional Finite Element Analyses of Reinforced Concrete
Columns", Computers and Structures. Vol. 80, No. 2, pp.199-212.

Lai, D. 2001. "Crack Shear-Slip in Reinforced Concrete Elements", M.A.Sc. Thesis, Department of Civil
Engineering, University of Toronto, 154 pp.
205

Mander, J .B., Priestley, M.J .N., and Park. R., 1988. "Theoretical Stress-Strain Model for Confined
Concrete", ASCE J ournal of Structural Engineering, Vol. 114, No. 8, pp. 1804-1826.

Ngo, D. and Scordelis, A.C., 1967. Finite Element Analysis of Reinforced Concrete Beams, ACI
J ournal, Vol. 64, No. 3 , pp. 152-163.

Okamura, H., and Maekawa, K. 1991. Nonlinear Analysis and Constitutive Models of Reinforced
Concrete, University of Tokyo, ISBN 7655-1506-0, 182 pp.

Ottosen, N.S. 1977. A Failure Criterion for Concrete, ASCE J ournal of Engineering Mechanics
Division, Vol. 103, No. 4, pp. 527-535.

Palermo, D., and Vecchio, F.J ., 2002, Behaviour and Analysis of Reinforced Concrete Walls Subjected
to Reversed Cyclic Loading, Publication No. 2002-01, Department of Civil Engineering, University
of Toronto, 351 pp.

Petersen, S.B., Rodrigues, J .M.C, and Martins, P.A.F., 1993. Automatic Generation of Quadrilateral
Meshes for the Finite Element Analysis of Metal Forming Processes, Finite Elements in Analysis and
Design, Vol. 35, pp. 157-168.

Popovics, S., 1973. "A Numerical Approach to the Complete Stress-Strain Curve of Concrete", Cement
and Concrete Research, Vol. 3, No.5, pp. 583-599.

Richart, F.E., Brandtzaeg, A., and Brown, R.L., 1928. A Study of the Failure of Concrete under
Combined Compressive Stresses, Bulletin No.185, University of Illinois Engineering Experimental
Station, Urbana, Illinois, 104 pp.
206
Saenz, L.P., 1973. Discussion of Equation for the Stress-Strain Relation for Concrete, Cement and
Concrete Research. Vol. 3, No. 5, pp. 583-599.

Sato, Y. and Vecchio, F.J ., 2002. Tension Stiffening and Crack Formation in RC Members with FRP
Sheets, ASCE J ournal of Structural Engineering, (submitted).

Scott, B.D., Park, R., and Priestley, M.J .N. 1982. "Stress-Strain Behavior of Concrete Confined by
Overlapping Hoops at Low and High Strain Rates", ACI J ournal, Vol. 79, No. 1, pp.13-27.

Seckin, M., 1981. "Hysteretic Behaviour of Cast-in-Place Exterior Beam-Column-Slab Subassemblies",
Ph.D. Thesis, Department of Civil Engineering, University of Toronto, 266 pp.

Selby, R.G. 1993. "Three-Dimensional Constitutive Relations for Reinforced Concrete", Publication No.
93-02, Department of Civil Engineering, University of Toronto, 147 pp.

Shim, W., 2002. Analysis of Beams with Low Shear Reinforcement, M.A.Sc. Thesis, Department of
Civil Engineering, University of Toronto, 358 pp.

Thorenfeldt, E., Tomaszewicz, A., and J ensen, J .J ., 1987. "Mechanical Properties of High-Strength
Concrete and Application in Design", Proceedings of the Symposium "Utilization of High Strength
Concrete," Stavanger, Norway, J une 1987, Tapir, Trondheim, pp. 149-159.

Vecchio, F.J ., 1992. "Finite Element Modeling of Concrete Expansion and Confinement", ASCE J ournal
of Structural Engineering, Vol. 118, No. 9, pp. 2390-2406.

Vecchio, F.J . 1999. "Towards Cyclic Load Modeling of Reinforced Concrete", ACI Structural J ournal,
Vol. 96, No. 2, pp.193-202.
207

Vecchio, F.J ., 2000. "Analysis of Shear-Critical Reinforced Concrete Beams", ACI Structural J ournal,
Vol. 97, No.1 pp. 102-110.

Vecchio, F.J ., 2000. Disturbed Stress Field Model for Reinforced Concrete: Formulation, ASCE
J ournal of Structural Engineering, Vol. 126, No. 9, pp. 1070-1077.

Vecchio, F.J . and Collins, M.P., 1982. "Response of Reinforced Concrete to In-Plane Shear and Normal
Stresses", Publication No. 82-03, Department of Civil Engineering, University of Toronto, 332 pp.

Vecchio, F.J . and Collins, M.P., 1986. "The Modified Compression Field Theory for Reinforced Concrete
Elements Subject to Shear", ACI J ournal Vol. 83, No. 2, pp. 219-231.

Vecchio, F.J . and Collins, M.P., 1993. "Compression Response of Cracked Reinforced Concrete", ASCE
J ournal of Structural Engineering, Vol. 119, No. 12, pp. 3590-3610.

Vecchio, F.J . and DeRoo, A., 1995. "Smeared-Crack Modeling of Concrete Tension Splitting", ASCE
J ournal of Engineering Mechanics, Vol. 121, No. 6 pp. 702-708.

Vecchio, F.J . and Lai, D., 2002. Crack Shear-Slip in Reinforced Concrete Elements, ACI Structural
J ournal, (submitted).

Walraven, J .C., 1981. Fundamental Analysis of Aggregate Interlock, ASCE J ournal of the Structural
Division, Vol. 107, No. 11, pp. 2245-2270.

208
Walraven, J .C., and Reinhardt, H.W., 1981. Theory and Experiments on the Mechanical Behaviour of
Cracks in Plain and Reinforced Concrete Subjected to Shear Loading, Concrete Mechanics Part A,
Heron, Vol. 26, No. 4, 65 pp.

Wong, R., 2001. Towards Modelling of Reinforced Concrete Members with Externally-Bonded Fibre
Reinforced Polymer (FRP) Composites, M.A.Sc. Thesis, Department of Civil Engineering,
University of Toronto, 298 pp.

Yamamoto, T., 1999. "Nonlinear Finite Element Analysis of Transverse Shear and Torsional Problems in
Reinforced Concrete Shells," M.A.Sc. Thesis, Department of Civil Engineering, University of
Toronto, 112 pp.

209






Appendix A VecTor2 Input Files for Sample Problem

















210
WSA1.JOB
* * * * * * * * * * * *
* V E C T O R *
* J O B D A T A *
* * * * * * * * * * * *

J ob Ti t l e ( 30 char . max. ) : WSA1
J ob Fi l e Name ( 8 char . max. ) : WSA1
Dat e ( 30 char . max. ) : August 11, 2002

STRUCTURE DATA
- - - - - - - - - - - - - -
St r uct ur e Type : 2
Fi l e Name ( 8 char . max. ) : WSA1

LOADI NG DATA
- - - - - - - - - - - -
No. of Load St ages : 201
St ar t i ng Load St age No. : 1
Load Ser i es I D ( 5 char . max. ) : WSA

Load Fi l e Name Fact or s
Case ( 8 char . max. ) I ni t i al Fi nal LS- I nc Type Reps C- I nc
1 WSA1 0. 000 50. 000 0. 250 1 1 0. 000
2 NULL 0. 000 0. 000 0. 000 1 1 0. 000
3 NULL 0. 000 0. 000 0. 000 1 1 0. 000
4 NULL 0. 000 0. 000 0. 000 1 1 0. 000
5 NULL 0. 000 0. 000 0. 000 1 1 0. 000

ANALYSI S PARAMETERS
- - - - - - - - - - - - - - - - - - -
Anal ysi s Mode ( 1- 2) : 1
Seed Fi l e Name ( 8 char . max. ) : NULL
Conver gence Li mi t ( >1. 0) : 1. 000010
Aver agi ng Fact or ( <1. 0) : 0. 500
Maxi mumNo. of I t er at i ons : 60
Conver gence Cr i t er i a ( 1- 5) : 1
Resul t s Fi l es ( 1- 4) : 2
Out put For mat ( 1- 3) : 1

MATERI AL BEHAVI OUR MODELS
- - - - - - - - - - - - - - - - - - - - - - - - -
Concr et e Compr essi on Base Cur ve ( 0- 3) : 2
Concr et e Compr essi on Post - Peak ( 0- 3) : 0
Concr et e Compr essi on Sof t eni ng ( 0- 8) : 1
Concr et e Tensi on St i f f eni ng ( 0- 6) : 1
Concr et e Tensi on Sof t eni ng ( 0- 3) : 1
Concr et e Tensi on Spl i t t i ng ( 0- 1) : 1
Concr et e Conf i ned St r engt h ( 0- 2) : 1
Concr et e Di l at i on ( 0- 1) : 1
Concr et e Cr acki ng Cr i t er i on ( 0- 4) : 1
Concr et e Cr ack Sl i p Check ( 0- 2) : 1
Concr et e Cr ack Wi dt h Check ( 0- 2) : 4
Concr et e Bond or Adhesi on ( 0- 3) : 1
Concr et e Cr eep and Rel axat i on ( 0- 1) : 1
Concr et e Hyst er esi s ( 0- 2) : 1
Rei nf or cement Hyst er esi s ( 0- 2) : 1
Rei nf or cement Dowel Act i on ( 0- 1) : 1
Rei nf or cement Buckl i ng ( 0- 1) : 1
El ement St r ai n Hi st or i es ( 0- 1) : 1
El ement Sl i p Di st or t i ons ( 0- 4) : 1
St r ai n Rat e Ef f ect s ( 0- 1) : 1
St r uct ur al Dampi ng ( 0- 1) : 1
Geomet r i c Nonl i near i t y ( 0- 1) : 1
Cr ack Al l ocat i on Pr ocess ( 0- 1) : 1

211
WSA1.S2R
* * * * * * * * * * * * * * * * * * *
* V e c T o r 2 *
* S T R U C T U R E D A T A *
* * * * * * * * * * * * * * * * * * *

STRUCTURAL PARAMETERS
*********************

St r uct ur e Ti t l e ( 30 char . max. ) : WSA1
St r uct ur e Fi l e Name ( 8 char . max. ) : WSA1
No. of R. C. Mat er i al Types : 2
No. of St eel Mat er i al Types : 3
No. of Bond Mat er i al Types : 0
No. of Rect angul ar El ement s : 410
No. of Quadr i l at er al El ement s : 0
No. of Tr i angul ar El ement s : 0
No. of Tr uss Bar El ement s : 123
No. of Li nkage El ement s : 0
No. of Cont act El ement s : 0
No. of J oi nt s : 462
No. of Rest r ai nt s : 12

MATERI AL SPECI FI CATI ONS
***********************

( A) REI NFORCED CONCRETE
- - - - - - - - - - - - - - - - - - - - - - -
<NOTE: > TO BE USED I N RECTANGULAR AND TRI ANGULAR ELEMENTS ONLY

CONCRETE
- - - - - - - -
MAT Ns T f ' c [ f ' t Ec e0 Mu Cc Agg Dens Kc ] [ Sx Sy]
TYP # mm MPa MPa MPa me / C mm kg/ m3 mm2/ s mm mm
1 0 305. 22. 6 1. 819 23770. 1. 902 0. 150 0. 20. 2400. 0. 0. 0.
2 1 305. 22. 6 1. 819 23770. 1. 902 0. 150 0. 20. 2400. 0. 0. 0.
/
REI NFORCEMENT COMPONENTS
- - - - - - - - - - - - - - - - - - - - - - - -
MAT REF DI R As Db Fy Fu Es Esh esh Cs Dep
TYP TYP deg % mm MPa MPa MPa MPa me / C me
2 1 90. 0. 101 6. 400 600. 649. 200000. 10. 4900. 0. 0.
/
( B) STEEL
- - - - - - - - -
<NOTE: > TO BE USED FOR TRUSS ELEMENTS ONLY
MAT REF AREA Db Fy Fu Es Esh esh Cs Dep
TYP TYP mm2 mm MPa MPa MPa MPa me / C me
1 1 200. 6 11. 3 315. 000 460. 000 200000. 000 14500. 000 10. 000 0. 000 0. 000
2 1 997. 6 25. 2 445. 000 680. 000 200000. 000 23500. 000 10. 000 0. 000 0. 000
3 1 1404. 4 29. 9 436. 000 700. 000 200000. 000 26400. 000 10. 000 0. 000 0. 000
/
( C) BOND
- - - - - - - -
<NOTE: > TO BE USED FOR EXTERI OR/ I NTERI OR BONDED ELEMENTS
MAT REF { Ao U1 U2 U3 S1 S2 S3 }/ { CPF Cmi n No. HOOK }
TYP TYP mm^2 MPa MPa MPa mm mm mm 0- 1 mm LYR 0/ 1
/
ELEMENT I NCI DENCES
******************

( A) RECTANGULAR ELEMENTS
- - - - - - - - - - - - - - - - - - - - - - - -
<<<<< FORMAT >>>>>
ELMT I NC1 I NC2 I NC3 I NC4 [ #ELMT d( ELMT) d( I NC) ] [ #ELMT d( ELMT) d( I NC) ] /
1 1 12 13 2 41 1 11 10 41 1 /
/
( B) QUADRI LATERAL ELEMENTS
- - - - - - - - - - - - - - - - - - - - - - - - - -
<<<<< FORMAT >>>>>
ELMT I NC1 I NC2 I NC3 I NC4 [ #ELMT d( ELMT) d( I NC) ] [ #ELMT d( ELMT) d( I NC) ] /
/



212
( C) TRI ANGULAR ELEMENTS
- - - - - - - - - - - - - - - - - - - - - - -
<<<<< FORMAT >>>>>
ELMT I NC1 I NC2 I NC3 [ #ELMT d( ELMT) d( I NC) ] [ #ELMT d( ELMT) d( I NC) ] /
/
( D) TRUSS ELEMENTS
- - - - - - - - - - - - - - - - - -
<<<<< FORMAT >>>>>
ELMT I NC1 I NC2 [ #ELMT d( ELMT) d( I NC) ] [ #ELMT d( ELMT) d( I NC) ] /
411 2 13 41 1 11 1 1 1 /
452 3 14 41 1 11 1 1 1 /
493 10 21 41 1 11 1 1 1 /
/
( E) LI NKAGE ELEMENTS
- - - - - - - - - - - - - - - - - - - - -
<<<<< FORMAT >>>>
ELMT I NC1 I NC2 [ #ELMT d( ELMT) d( I NC) ] [ #ELMT d( ELMT) d( I NC) ]
/
( F) CONTACT ELEMENTS
- - - - - - - - - - - - - - - - - - - - -
<<<<< FORMAT >>>>
ELMT I NC1 I NC2 I NC3 I NC4 [ #ELMT d( ELMT) d( I NC) ] [ #ELMT d( ELMT) d( I NC) ]
/
MATERI AL TYPE ASSI GNMENT
************************
<<<<< FORMAT >>>>>
ELMT MAT ACT [ #ELMT d( ELMT) ] [ #ELMT d( ELMT) ] /
1 1 1 41 1 1 1 /
42 2 1 41 1 8 41 /
370 1 1 41 1 1 1 /
411 3 1 41 1 1 1 /
452 2 1 41 1 1 1 /
493 1 1 41 1 1 1 /
/
COORDI NATES
***********
<NOTE: > UNI TS: i n OR mm
<<<<< FORMAT >>>>>
NODE X Y [ #NODES d( NODES) d( X) d( Y) ] [ #NODES d( NODES) d( X) d( Y) ] /
1 0. 000 0. 000 1 1 0. 000 0. 000 4 11 55. 000 0. 000 /
2 0. 000 63. 000 1 1 0. 000 0. 000 4 11 55. 000 0. 000 /
3 0. 000 126. 000 7 1 0. 000 51. 572 4 11 55. 000 0. 000 /
10 0. 000 487. 000 2 1 0. 000 63. 000 4 11 55. 000 0. 000 /
45 220. 000 0. 000 1 1 0. 000 0. 000 38 11 49. 459 0. 000 /
46 220. 000 63. 000 1 1 0. 000 0. 000 38 11 49. 459 0. 000 /
47 220. 000 126. 000 7 1 0. 000 51. 572 38 11 49. 459 0. 000 /
54 220. 000 487. 000 2 1 0. 000 63. 000 38 11 49. 459 0. 000 /
/
SUPPORT RESTRAI NTS
******************
<NOTE: > CODE: ' 0' FOR NOT RESTRAI NED NODES AND ' 1' FOR RESTRAI NED ONES
<<<<< FORMAT >>>>>
NODE X- RST Y- RST [ #NODE d( NODE) ] /
452 1 0 11 1 /
45 0 1 1 1 /
/
213
WSA1.L2R * * * * * * * * * * * * *
* V e c T o r 2 *
* L O A D D A T A *
* * * * * * * * * * * * *

LOAD CASE PARAMETERS
********************

St r uct ur e Ti t l e ( 30 char . max. ) : WSA1
Load Case Ti t l e ( 30 char . max. ) : 1mm
Load Case Fi l e Name ( 8 char . max. ) : WSA1
No. of Loaded J oi nt s : 0
No. of Pr escr i bed Suppor t Di spl acement s : 1
No. of El ement s wi t h Gr avi t y Loads : 0
No. of El ement s wi t h Temper at ur e Loads : 0
No. of El ement s wi t h Concr et e Pr est r ai n : 0
No. of El ement s wi t h I ngr ess Pr essur e : 0
No. of El ement Sur f aces w/ Ther mal Load : 0
No. of Nodes wi t h I mpul se For ces : 0
Gr ound Accel er at i on Recor d ( 0- 1) : 0

J OI NT LOADS
***********
<NOTE: > UNI TS: KI PS OR KN
<<<<< FORMAT >>>>>
NODE Fx Fy [ #NODE d( NODE) d( Fx) d( Fy) ] /
/
SUPPORT DI SPLACEMENTS
*********************
<NOTE: > UNI TS: MM OR I N
<<<<< FORMAT >>>>>
J NT DOF DI SPL [ #J NT d( J NT) ] /
462 2 - 1. 000 1 1/
/
GRAVI TY LOADS
*************
<NOTE: > UNI TS: KG/ M3
<<<<< FORMAT >>>>>
ELMT DENS GX GY [ #ELMT d( ELMT) ] [ #ELMT d( ELMT) ] /
/
TEMPERATURE LOADS
*****************
<NOTE: > UNI TS: F OR C
<<<<< FORMAT >>>>>
ELMT TEMP [ #ELMT d( ELMT) d( TEMP) ] [ #ELMT d( ELMT) d( TEMP) ] /
/
CONCRETE PRESTRAI NS
*******************
<NOTE: > UNI TS: me
<<<<< FORMAT >>>>>
ELMT STRAI N [ #ELMT d( ELMT) d( STRAI N) ] [ #ELMT d( ELMT) d( STRAI N) ] /
/
I NGRESS PRESSURES
*****************
<NOTE: > UNI TS: MPa
<<<<< FORMAT >>>>>
ELMT PRESSURE [ #ELMT d( ELMT) d( PRS) ] [ #ELMT d( ELMT) d( PRS) ] /
/
SURFACE THERMAL LOADS
*********************
<NOTE: > UNI TS: Sec, Degr ees C
<<<<< FORMAT >>>>>
NODE1 NODE2 Tm1 Tp1 Tm2 Tp2 Tm3 Tp3 [ #SURF d( NODE) ] [ #SURF d( NODE) ] /
/
I MPULSE FORCES
**************
<NOTE: > UNI TS: Sec, kN
<<<<< FORMAT >>>>>
NODE DOF T1 F1 T2 F2 T3 F3 T4 F4 [ #NODE d( NODE) ] /
/
GROUND ACCELERATI ON
*******************
<NOTE: > UNI TS: Sec, m/ s2
<<<<< FORMAT >>>>>
TI ME ACC- X ACC- Y