V T 2 & F W U ' M Second Edition: EC OR ORM Orks SER S Anual

Download as pdf or txt
Download as pdf or txt
You are on page 1of 347

VECTOR2 & FORMWORKS USER’S MANUAL

SECOND EDITION

P. S. Wong

F.J. Vecchio

H. Trommels

August, 2013
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 xiv

List of Appendices xxi

Notation xxii

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 7

2.2.3 Equilibrium Relationships 8

2.2.4 Constitutive Relationships 9

2.2.5 Consideration of Local Crack Conditions 10

2.2.6 Consideration of Local Crack Conditions with FRC 12

2.3 Disturbed Stress Field Model (DSFM) 14

2.3.1 Compatibility Relationships 14

2.3.2 Equilibrium Relationships 16

2.3.3 Constitutive Relationships 17

2.3.4 Shear Slip Relationships 20

2.4 Finite Element Implementation 21

2.4.1 Composite Material Stiffness Matrix 23


iv
2.4.2 Element Stiffness Matrices 26

2.4.3 Pseudo Nodal Loads 26

2.4.4 Total Load Iterative Secant Stiffness Solution 28

2.5 Augmented Analysis Features 29

2.5.1 Modeling of Lateral Expansion Effect 29

2.5.2 Modeling of Triaxial Stresses 31

2.5.3 Analysis of Cyclically Loaded Structures 32

2.5.4 Analysis of Construction and Loading Chronology 34

2.5.5 Bond-Slip Mechanisms 35

2.5.6 Dynamic Analysis 36

2.5.6.1 Introduction 36

2.5.6.2 Newmark Direct Integration Method: Newmark Beta and Gamma Factors 37

2.5.6.3 Damping 37

2.5.6.4 Ground Acceleration 38

2.5.7 Accounting for Strain History 38

2.5.8 Accounting for Strain Rate Effects 38

2.5.9 Structural Damping 41

2.5.9.1 Rayleigh Damping 41

2.5.9.2 Alternative Damping 42

2.5.10 Geometric Nonlinearity 43

2.5.11 Crack Allocation 43

2.5.11.1 Uniform Crack Allocation (Deluce, Lee, and Vecchio, 2012) 43

2.5.11.2 Variable Crack Allocation (Sato 2002) 44

2.5.12 Additional Analysis Parameters 45

2.5.12.1 Concrete Aggregate Type 45

2.5.12.2 Concrete Conductivity 45

2.5.12.3 Concrete Fracture Energy 45


v
2.5.12.4 Prestressing Friction and Wobble Coefficient 46

2.5.12.5 Thermal Time-Stepping Factor 47

2.5.12.6 Material Resistance/Creep Factors 47

3 Element Library 48

3.1 Reinforced Concrete Elements 48

3.1.1 Constant Strain Triangle 48

3.1.2 Plane Stress Rectangle 50

3.1.3 Quadrilateral 51

3.1.3.1 Isoparametric Quadrilateral Elements 52

3.1.3.2 Degenerate Quadrilateral Elements 52

3.2 Reinforcement Elements 53

3.2.1 Truss Bar 53

3.3 Bond – Slip Elements 54

3.3.1 Link 55

3.3.2 Contact 56

4 Models for Concrete Materials 59

4.1 Compression Pre-Peak Response 59

4.1.1 Linear 60

4.1.2 Popovics 61

4.1.3 Hognestad (Parabola) 62

4.1.4 Popovics (High Strength) 62

4.1.5 Hoshikuma et al. 64

4.1.6 Smith-Young 65

4.1.7 Lee et al 2012 (FRC) (Lee et al, 2012) 65

4.2 Compression Post-Peak Response 66

4.2.1 Pre-Peak Base Curve 68

4.2.2 Modified Park-Kent 68


vi
4.2.3 Popovics / Mander 69

4.2.4 Hoshikuma et al. 70

4.2.5 Saenz / Spacone 71

4.2.6 Montoya 2003 73

4.2.7 Lee et al 2012 (FRC) 74

4.3 Compression Softening 74

4.3.1 No compression softening 76

4.3.2 Vecchio 1992-A (e1/e2-Form) 76

4.3.3 Vecchio 1992-B (e1/e0-Form) 77

4.3.4 Vecchio-Collins 1982 78

4.3.5 Vecchio-Collins 1986 79

4.4 Tension Stress-Strain Response 79

4.4.1 No Tension Stiffening 80

4.4.2 Modified Bentz 2003 81

4.4.3 Vecchio 1982 81

4.4.4 Collins-Mitchell 1987 82

4.4.5 Bentz 1999 82

4.4.6 Izumo, Maekawa Et Al. 83

4.4.7 Tension Chord Model (Kaufmann) 84

4.4.8 Lee 2011 (w/ Post Yield) (Lee et al. 2011) 85

4.5 Tension Softening 87

4.5.1 Not Considered 88

4.5.2 Linear 89

4.5.3 Bilinear 89

4.5.4 Nonlinear (Yamamoto 1999 – No residual) 91

4.5.5 Nonlinear (Hordijk) 92

4.5.6 Exponential 94
vii
4.5.7 Custom Input-Strain Based 94

4.5.8 Custom Input-Crack Based 95

4.5.9 FRC (fib Model Code 2010) 95

4.6 FRC Tension 96

4.6.1 Not Considered 96

4.6.2 SDEM-Monotonic (Lee et al 2012) 96

4.6.3 SDEM-Cyclic 98

4.6.4 Diverse Embedment Model (DEM, Lee et al 2011) 99

4.6.5 Variable Engagement Model (VEM: Voo & Foster 2003) 100

4.6.6 fib Model Code 2010 101

4.7 Lateral Expansion 101

4.7.1 Constant Poisson's ratio 101

4.7.2 Variable Poisson's Ratio – Kupfer 102

4.7.3 Variable Poisson’s Ratio – Montoya 102

4.7.4 Variable Montoya With Limit 104

4.8 Confinement Strength 104

4.8.1 Strength Enhancement Neglected 104

4.8.2 Kupfer / Richart Model 104

4.8.3 Selby Model 105

4.8.4 Montoya / Ottosen 106

4.9 Cracking Criterion 107

4.9.1 Uniaxial cracking stress 108

4.9.2 Mohr-Coulomb (Stress) 108

4.9.3 Mohr-Coulomb (Strain) 110

4.9.4 CEB-FIP Model 110

4.9.5 Gupta 1998 Model 110

4.10 Crack Stress Check 111


viii
4.10.1 Crack Stress Calc Omitted 111

4.10.2 Basic (DSFM/MCFT) 111

4.10.3 Advanced (Lee 2009) 112

4.11 Crack Width Check 112

4.12 Crack Slip Calculation 114

4.12.1 Not considered 115

4.12.2 Vecchio-Lai (Cyclic) [maybe don’t include] 115

4.12.3 Maekawa (Monotonic) 116

4.12.4 Stress Model (Walraven) 116

4.12.5 Hybrid-I, II, and III Models 116

4.13 Creep and Relaxation 117

4.14 Hysteretic Response 117

4.14.1 No Plastic Offsets 117

4.14.2 Plastic offsets; linear loading/unloading 118

4.14.3 Plastic offsets; nonlinear loading/unloading 120

4.14.4 Plastic offsets; nonlinear w/ cyclic decay (Palermo Model) 121

5 Models for other material types 126

5.1 Structural Steel 126

5.2 Masonry 127

5.2.1 Introduction 127

5.2.2 Compressive Behavior 128

5.2.3 Tensile Behavior 131

5.2.4 Slip Along the Joints 132

5.3 Wood (Fixed Orthotropic) 133

5.4 Concrete-Steel Laminate 134

5.5 Concrete-SFRC Laminate and Masonry-SFRC Laminate 135

5.6 Concrete-Ortho Laminate 136


ix
6 Models for Reinforcement Materials 137

6.1 Stress-Strain Response 137

6.1.1 Ductile Steel Reinforcement 137

6.1.2 Prestressing Steel 139

6.1.3 Tension Only Reinforcement 140

6.1.4 Compression Only Reinforcement 140

6.1.5 Externally Bonded FRP Reinforcement 140

6.2 Hysteretic Response 140

6.2.1 Linear 140

6.2.2 Seckin Model w/ Bauschinger Effect 141

6.2.3 Menegotto-Pinto Model w/ Bauschinger Effect 142

6.2.4 Elastic Hardening (Curvilinear) 144

6.2.5 Elastic-Hardening (Trilinear) 144

6.2.6 Elastic Plastic (Bilinear) 145

6.3 Dowel Action 145

6.3.1 Not considered 145

6.3.2 Tassios Model 146

6.4 Reinforcement Buckling 147

6.4.1 Dhakal-Maekawa Model 2002 (DM) Model 148

6.4.2 Refined Dhakal-Maekawa (RDM) Model 149

6.4.3 RDM and DM buckling models for hysteretic response 152

6.4.4 Determination of the Unsupported Length Ratio (b/t=L/D) 153

6.4.5 Asatsu Model 156

6.5 Steel Fibre Reinforced Concrete 156

6.5.1 Introduction 156

6.5.2 Stress-Strain Behavior of SFRC 157

6.6 Laminates 159


x
6.7 Shape Memory Alloy Type 1 and Type 2 160

6.7.1 Shape Memory Alloy 1 (SMA 1) 160

6.7.2 Shape Memory Alloy 2 (SMA 2) 161

7 Models for Bond 162

7.1 Bond Stress-Slip Models for Embedded Bars 162

7.1.1 Perfect bond 163

7.1.2 Eligehausen Model 163

7.1.3 Gan Model 166

7.1.4 Harajli Model 167

7.1.5 Fujii Model 169

7.1.6 Eligehausen and Gan Models with No Cyclic Damage 171

7.2 Bond Stress-Slip Models for Externally Bonded Plates or Sheets 172

7.2.1 Perfect bond 172

7.2.2 Other models 172

8 Part II: FormWorks 174

8.1 Introduction 174

8.2 Installing FormWorks 174

8.3 An Overview of the FormWorks Modeling Process 175

8.3.1 Before Using FormWorks 176

8.3.2 VecTor2 Input and Output Files 176

8.4 FormWorks Basics 178

8.4.1 The FormWorks Interface 178

8.4.2 Creating a New Workspace 179

8.4.3 Saving the Workspace 180

8.4.4 Opening a Saved Workspace File 181

8.5 Viewing and Printing the Workspace 181

8.5.1 Manipulating the View 181


xi
8.5.1.1 Horizontal and Vertical Scaling 181

8.5.1.2 Changing Display Limits 181

8.5.1.3 Zooming and Panning 182

8.5.1.4 Selecting Display Options 183

8.5.2 Printing the Workspace 186

9 The Job Data 187

9.1 The Job Control Page 187

9.1.1 Job Data Group 188

9.1.2 Structure Data Group 188

9.1.3 Load Data Group 188

9.1.3.1 Static Analysis 188

9.1.3.2 Dynamic Analysis 192

9.1.4 Analysis Parameters Group 193

9.2 The Model Page 194

9.2.1 Concrete Models 195

9.2.2 Reinforcement Models 197

9.2.3 Bond Models 197

9.2.4 Analysis Models 197

9.3 The Auxiliary Input Page 199

9.3.1 General 200

9.3.2 Dynamic Analysis Options 201

9.3.3 Tension Softening Options 202

9.3.4 Masonry Structures Data 203

9.3.5 Material Resistance/Creep Factors 205

10 The Structure Data 207

10.1 Preliminary considerations 207

10.2 Structure Limits 208


xii
10.3 Structure Information 209

10.4 Specifying Material Types 209

10.4.1 Reinforced Concrete Material Types 210

10.4.1.1 Reinforced Concrete 211

10.4.1.2 Structural Steel 216

10.4.1.3 Masonry 218

10.4.1.4 Wood-Orthotropic 220

10.4.1.5 Modeling Composite Materials 222

10.4.2 Reinforcement Material Types 229

10.4.2.1 Bond Types 230

10.4.3 Defining Nodes 232

10.4.4 Defining Elements 233

10.4.4.1 Viewing Elements Attributes 238

10.4.4.2 Viewing Multiple Element Layers 239

10.4.5 Assigning Material Types 239

10.4.6 Restraining the Structure 241

10.4.7 Automatic Method 243

10.4.7.1 Reinforced Concrete Regions 243

10.4.7.2 Reinforcement 252

10.4.7.3 Voids 257

10.4.7.4 Line Constraints 259

10.4.7.5 Point Constraints 260

10.4.7.6 Generating the Mesh 262

11 The Load Case Data 265

11.1 Load Limits 265

11.2 Selecting the Load Case 265

11.3 Load Information 265


xiii
11.4 Nodal Loads 266

11.5 Support Displacements 268

11.6 Gravity Loads 269

11.7 Temperature Loads 271

11.8 Concrete Prestrains 272

11.9 Ingress Pressures 273

11.10 Nodal Thermal Loads 273

11.11 Lumped Masses 276

11.12 Impulse Forces 278

11.13 Ground Acceleration Loads 280

12 Running VecTor2 283

12.1 Starting the Analysis 283

12.2 Example 1: Simply Supported Beam 285

12.3 Example 2: SW22 289

13 Summary & Recommendations 293

13.1 Summary 293

13.2 Recommendations 293

References 295

Appendix A – VecTor2 Input Files for Sample Problem 1 305

VecTor.JOB 306

Appendix B – VecTor2 Input Files for Sample Problem 2 311


xiv

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: Comparison of average and local stresses at a crack (a) average stresses between crack; (b) local

stresses at crack free surface 13

Figure 6: Deformation due to crack shear slip 15

Figure 7 VecTor2 nonlinear finite element analysis algorithm 22

Figure 8: Definition of secant moduli for a) concrete b) reinforcement 25

Figure 9: Rayleigh Damping (Chopra, 2007) 42

Figure 10: Prestressing Wobble and Friction Coefficients 46

Figure 11: Constant strain triangle element 49

Figure 12: Plane stress rectangle element 50

Figure 13: a) Quadrilateral element b) Decomposition of quadrilateral element into two constant strain

triangle elements 51

Figure 14: Degenerate Quadrilateral Element (Li, 2004) 52

Figure 15: Truss bar element 53

Figure 16 Link element 55

Figure 17: Contact element 57

Figure 18: Linear pre- and post-peak concrete compression response 60

Figure 19 Popovics pre- and post-peak concrete compression response 61

Figure 20: Hognestad parabolic pre- and post-peak concrete compression response 62

Figure 21: Popovics high strength pre- and post-peak concrete compression response 63

Figure 22: Hoshikuma et al. concrete compressive pre-peak response 64

Figure 23: Smith-Young concrete compressive pre-peak response 65

Figure 24: Smith-Young concrete compression post-peak response for fp <= f’c 68
xv
Figure 25: Modified Park-Kent post-peak concrete compression response 69

Figure 26: Hoshikuma et al. concrete compressive post-peak response 71

Figure 27: Saenz/Spacone concrete compressive post-peak response 72

Figure 28: Montoya concrete compressive post-peak response 73

Figure 29: Strength and strain-softened compression response 75

Figure 30: Strength-only softened compression response 76

Figure 31: Vecchio 1992-A compression softening model 77

Figure 32: Vecchio 1982 compression softening model 78

Figure 33: Vecchio 1982 and Collins-Mitchell tension stiffening response 82

Figure 34: Izumo, Maekawa et al. tensions stiffening response 83

Figure 35: Tension chord model 85

Figure 36: Lee 2010 stress-strain response 86

Figure 37: Linear tension softening response with and without residual 89

Figure 38: CEB-FIP tension softening curve (CEB-FIP, 1990) 90

Figure 39: Bilinear Tension softening model implemented in VecTor2 91

Figure 40: Yamamoto tension softening response with and without residual 92

Figure 41: Envelope curves for normal and light-weight concrete (Cornelissen et al, 1986) 92

Figure 42: Stress-crack opening relationships (Cornelissen et al., 1986) 93

Figure 43: Custom input-strain based tension softening in VecTor2 94

Figure 44 tension-crack opening relationship in fib model code 2010 (fib, 2010) 96

Figure 45: Stiffness for the re- and unloading response of FRC in cyclic tension 98

Figure 46: Kupfer variable Poisson’s ratio model 102

Figure 47: Lateral strain-axial strain curves for Imran-Pantazopoulou tests (Montoya, 2003) 103

Figure 48: Mohr-Coulomb (Stress) cracking criterion 109

Figure 49: Shear across with crack through aggregate interlock 111

Figure 50: Maximum crack width check including effect of element size 113

Figure 51: Palermo model of concrete hysteretic response in compression 124


xvi
Figure 52: Palermo model of concrete hysteretic response in tension 125

Figure 53: Stress-strain behavior of structural steel in VecTor2 126

Figure 54: Masonry joints 127

Figure 55: Ganz failure criterion implemented in VecTor2 (Ganz, 1985) 128

Figure 56: Masonry compressive stress-strain behavior 130

Figure 57: Variation in masonry elastic modulus 131

Figure 58: Masonry Joint Slip 132

Figure 59: Concrete-steel laminate element (Vecchio and McQuade, 2011) 134

Figure 60: Ductile steel reinforcement stress-strain response [left(a): Linear strain-hardening (Trilinear);

right (b): Nonlinear strain-hardening (HP4)] 138

Figure 61: Prestressing steel reinforcement stress-strain response 140

Figure 62: Seckin model of ductile steel reinforcement for hysteretic response [left (a): Seckin w/

Bauschinger-Trilinear; right (b): Seckin w/ Bauschinger-HP4] 141

Figure 63: Menegotto-Pinto model of ductile steel reinforcement for hysteretic response [left (a):

Menegotto-Pinto w/ Baushinger-Trilinear ; right (b): Menegotto-Pinto w/ Baushinger-HP4 143

Figure 64: curvilinear strain hardening 144

Figure 65: Trilinear stress-strain response 144

Figure 66: Elastic-plastic stress-strain response 145

Figure 67: Dowel resistance mechanism 147

Figure 68: Compressive stress-strain relationships (Dhakal and Maekawa, 2002a, b) 148

Figure 69: Compressive stress-strain relationships (Akkaya et al., 2013) 150

Figure 70: Hysteretic models for ductile reinforcement including buckling [left (a): Seckin; right (b):

Menegotto-Pinto 152

Figure 71: Values of nb and nl for common reinforcement 154

Figure 72: Tensile stress-strain response of FRC for low and high fibre volume contents (Susetyo, 2009)

158
xvii
Figure 73: Compressive stress-strain curve of SFRC, illustrating a) fibre content and b) fibre aspect ratio

(Susetyo, 2009) 159

Figure 74: Stress-strain response for SMA 1 160

Figure 75: Stress-strain response for SMA2 161

Figure 76: Eligehausen bond stress-slip response 165

Figure 77: Gan bond stress-slip response 167

Figure 78: Harajli bond stress-slip response 169

Figure 79: Eligehausen cyclic bond stress-slip relationship 171

Figure 80: Bond stress-slip response for externally bonded plates or sheets 173

Figure 81: Input and output files forVecTor2 analysis 176

Figure 82: FormWorks application window 179

Figure 83: Creating new Workspace windows 180

Figure 84: Display Limits dialog box 182

Figure 85: Display Options dialog box 184

Figure 86: Job Control property page 187

Figure 87: Monotonic type loading 189

Figure 88: Cyclic type loading 189

Figure 89: Reversed cyclic type loading 190

Figure 90: The Models page 195

Figure 91: The Auxiliary Input Page 199

Figure 92: Masonry joints 203

Figure 93: Masonry joint shear strength ratio 204

Figure 94: Structure Limits dialog box 208

Figure 95: Structure Information dialog box 209

Figure 96: Reinforced Concrete Materials Properties Dialog Box 211

Figure 97: Define Reinforced Concrete Properties Dialog Box: Structural Steel 216

Figure 98: Define Reinforced Concrete Properties: Masonry 218


xviii
Figure 99: Define Reinforced Concrete Properties: Wood-Orthotropic 220

Figure 100: Define Reinforced Concrete Properties: Concrete-Steel Laminate 222

Figure 101: Define Reinforced Concrete Properties: Concrete-SFRC Laminate 224

Figure 102: Define Reinforced Concrete Properties: Masonry-SFRC Laminate 226

Figure 103: Define Reinforced Concrete Properties: Concrete-Ortho Laminate 227

Figure 104: Reinforcement Materials Properties Dialog Box 229

Figure 105: Bond Properties Dialog Box 230

Figure 106: Create Nodes Dialog Box 232

Figure 107: Example of node creation entry 233

Figure 108: Create Rectangular Elements Dialog Box 234

Figure 109: Example of rectangular element creation entry 235

Figure 110: Example of link element creation entry 236

Figure 111: Example of contact element creation entry 237

Figure 112: Element attributes dialog 238

Figure 113: Assign Material types dialog box 240

Figure 114: Selecting elements for material type assignment with the mouse 241

Figure 115: Create Support Restraints Dialog Box 241

Figure 116: Selecting nodes for restraints with the mouse 242

Figure 117: Examples of legal and illegal region definitions and intersections 244

Figure 118: RC Regions Property Page 245

Figure 119: Example of reinforced concrete region creation 246

Figure 120: Hybrid Discretization 247

Figure 121: Grid Superposition Discretization 248

Figure 122: Division Point Insertion Discretization 248

Figure 123: Hybrid Discretization with Rectangle Option 249

Figure 124: Division Point Insertion Discretization with Quadrilateral Option 249

Figure 125: Division Point Insertion Discretization with Smoothed Triangles Option 250
xix
Figure 126: Reinforcement Page 253

Figure 127: Example of reinforcement path creation 255

Figure 128: Voids and Constraints 258

Figure 129: Example of point constraint creation 262

Figure 130: Mesh Creation Page 263

Figure 131: Load Limits Dialog Box 265

Figure 132: Load Information Dialog Box 266

Figure 133: Apply Nodal Loads Dialog Box 267

Figure 134: Apply Support Displacements Dialog Box 268

Figure 135: Apply Gravity Loads Dialog Box 269

Figure 136: Apply Temperature Loads Dialog 271

Figure 137: Apply Concrete Prestrains Dialog Box 272

Figure 138: Apply Ingress Pressures Dialog Box 273

Figure 139: Thermal profiles in VecTor2 (Zhou, 2004) 274

Figure 140: Apply Nodal Thermal Loads Dialog Box 275

Figure 141: Apply Lumped Masses Dialog Box 276

Figure 142: Impulse forces in VecTor2 278

Figure 143: Apply Impulse Forces Dialog Box 279

Figure 144: Apply Ground Acceleration Load Dialog 280

Figure 145: input file format of VECTOR.EQR file 281

Figure 146: Bandwidth Reduction Dialog Box 283

Figure 147: VecTor2 Analysis Proceeding 285

Figure 148: Reinforced Concrete Beam 286

Figure 150: Load versus midpsan deflection of reinforced concrete beam 287

Figure 149: FormWorks Model 287

Figure 151: Crack pattern of reinforced concrete beam at ultimate load 288

Figure 152: SW22 289


xx
Figure 153: FormWorks Model for SW22 290

Figure 154: Models used for SW22 291

Figure 155: Load-displacement response for SW22 291

Figure 156: Deflected Shape and Final Crack Pattern for SW22 292
xxi

List of Appendices

Appendix A – VecTor2 Input Files for Sample Problem 1 296

Appendix B – VecTor2 Input Files for Sample Problem 2 302


xxii

Notation

a = maximum aggregate size

𝐴𝑐𝑒 = effective cross-sectional area of concrete

c = convergence averaging factor

𝑐 = cohesion of the mortar bed joint interface (masonry)

ct = compression softening coefficient

[B] = element strain-displacement matrix

db = diameter of reinforcing bar

[D] = composite material stiffness matrix

[Dc] = concrete material stiffness matrix relative to x,y axes

[Dc]’ = concrete material stiffness matrix relative to principal axes

�𝐷𝑓 �′ = fibre stiffness matrix relative to the direction of the tensile stress attained by fibres

[Ds] = reinforcement material stiffness matrix relative to x,y axes

[Ds]’ = reinforcement material stiffness matrix relative to principal axes

Cd = compression softening, strain softening factor

Cs = compression softening, shear slip factor

𝐶𝑀𝑂𝐷𝑖 = crack mouth opening displacement

𝑑𝑓 = fibre diameter

Ec = initial tangent stiffness of concrete

𝐸𝑐,𝑑 = dynamic modulus of elasticity of concrete

E c1 = secant modulus of concrete in the principal tensile direction

E c2 = secant modulus of concrete in the principal compressive direction

𝐸𝑒ℎ = un- and re-loading stiffness due to mechanical anchorage effect of fibres

Ef1
= fibre stiffness matrix relative to the direction of the tensile stress attained by fibres

𝐸𝑚𝑥 = minimum modulus


xxiii
𝐸𝑚𝑦 = maximum modulus of elasticity

Es = initial tangent stiffness of reinforcement

Es = secant modulus of reinforcement

𝐸𝑠𝑡 = un- and re-loading stiffness due to frictional bond behavior of fibres

f’c = concrete cylinder uniaxial compressive strength

𝑓𝑐𝑑 = dynamic compressive strength of concrete

fcc = concrete cube strength

𝑓𝑐𝑠 = static compressive strength of concrete

fcx = average concrete axial stress in the x-direction

fcy = average concrete axial stress in the y-direction

𝑓𝑒ℎ = tensile stress attained by fibres due to the mechanical anchorage effect

𝑓𝑓 = tensile stress due to fibres at a crack

𝑓𝑓𝑡𝑠 = residual tensile stress at an initial crack opening

𝑓𝑖𝑡 = intermediate stress

𝑓𝑚𝑥 = compressive strength in the x- direction with bed joint aligned horizontally

𝑓𝑚𝑦 = compressive strength in the y-direction with bed joint aligned horizontally

fp = peak concrete compressive stress

𝑓𝑟1,3 = residual stresses input by user

𝑓𝑠𝑡 = tensile stress attained by fibres due to the frictional bond behavior

fsx = average stress of reinforcement parallel to the x-direction

fsy = average stress of reinforcement parallel to the y-direction

fscrx = local stress at a crack of reinforcement parallel to the x-direction

fscry = local stress at a crack of reinforcement parallel to the y-direction

fsyield = yield strength of reinforcement

f’t = concrete cracking strength

𝑓𝑡𝑑 = dynamic tensile strength of concrete


xxiv
𝑓𝑡𝑠 = static tensile strength of concrete

𝑓𝑡𝑦 = tensile strength of the mortar bed joint interface in the x-direction (perpendicular to the bed

joints)

𝑓𝑤,𝑐𝑟,𝑎𝑣𝑔 = tensile stress attained by fibres, which is evaluated for the average crack width

𝑓𝑤,𝑐𝑟,𝑚𝑎𝑥 = tensile stress attained by fibres, which is evaluated for the maximum crack width

[F*] = pseudo nodal load vector

[F] = applied external nodal load vector

[F’] = total nodal load vector

Gc = secant shear modulus of concrete

Gf = fracture energy parameter

ℎ = element thickness

k = post-peak decay parameter for stress-strain response of concrete in compression

𝑘 = stiffness matrix (for dynamic analyses)

𝑘𝑑 = shape factor based on post-peak compression stress-strain curve steepness

kr = spring stiffness of link element radial to reinforcement direction

kt = spring stiffness of link element tangential to reinforcement direction

[k] = element stiffness matrix

[kc] = element stiffness matrix contribution from concrete

[ks] = element stiffness matrix contribution from reinforcement

[K] = structure stiffness matrix

𝐾𝑑 = damage factor or fibre efficiency factor

𝐾𝑒ℎ = factors to represent the average pull-out stresses of a fiber due to the mechanical anchorage

effect

𝐾𝑓 = factor to account for fibre orientation and engagement

𝐾𝑠𝑡 = factors to represent the average pull-out stresses of a fiber due to frictional bond behaviour

𝑙𝑎,𝑐𝑟𝑖𝑡 = critical fibre length for fibre fracture


xxv
𝑙𝑓 = fibre length

𝐿/𝐷 = unsupported length to diameter ratio for steel reinforcing bars

Lr = characteristic length

𝑚 = mass matrix

n = curve fitting parameter for stress-stain response of concrete in compression

[r] = nodal displacement vector

𝑅 = ratio of principal stresses

s = average crack spacing

𝑠𝑒ℎ = slip at maximum tensile force due to the mechanical anchorage of a fiber with the inclination

angle of 0 degree to the normal of the crack surface

𝑠𝑓 = slip at the frictional bond strength for a fibre with the inclination angle of zero degree

smx = average crack spacing in the x-direction

smy = average crack spacing in the y-direction

𝑠𝑝𝑖 = joint spacing

sθ = average crack spacing perpendicular to the crack

𝑡ℎ𝑖 = thickness of joints

[T] = coordinate transformation matrix

v12,v21 = Poisson’s ratios of concrete in the principal directions

e
v12 = elastic component of Poisson’s ratio

*
v12 = residual component of Poisson’s ratio

vcxy = average concrete shear stress

vci = local concrete shear stress at a crack

𝑉𝑓 = volumetric fibre ratio

w = average crack width

𝑤𝑐𝑟,𝑎𝑣𝑔 = average crack width

𝑤𝑐𝑟,𝑒𝑥𝑝 = maximum experienced crack width


xxvi
𝑤𝑐𝑟,𝑚𝑎𝑥 = maximum crack width

α = orientation of reinforcement relative to positive x-axis

𝛼 = fibre engagement factor (for FRC Tension calculations)

𝛼 = direction of principal axes with respect to the direction of horizontal joints (masonry)

𝛼𝑓 = fibre orientation factor

𝛽 = Newmark beta factor (for dynamic analyses)

𝛽 = angle between principal axes and axes of orthotropy (masonry)

βcr = crack width check crack coefficient

𝛽𝑒ℎ = coefficients to consider the effect of a fiber slip on the longer embedment side on the

mechanical anchorage effect

βd = compression softening reduction factor

𝛽𝑓 = coefficients to consider the effect of a fiber slip on the longer embedment side on the frictional

behavior

βl = confinement strength enhancement factor

∆θε = post-cracking rotation of principal total strain field

∆θ σ = post-cracking rotation of the principal stress field

𝜀̇𝑐 = strain rate (compression)

𝜀̇𝑐𝑡 = tensile strain rate of concrete

εc1 = average net concrete axial strain, in the principal tensile 1-direction

εc2 = average net concrete axial strain, in the principal compressive 2-direction

𝜀𝑐80𝑖 = post-peak strain at 80% of peak stress

ε cf
= average net concrete tensile strain, in the direction of the tensile stress attained by fibres

𝜀𝑐ℎ = characteristic strain for concrete in tension

εcr = concrete cracking strain

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

εcy = average net concrete axial strain, in the y-direction


xxvii

ε oc = elastic strain offsets in concrete

ε cp = plastic strain offsets in concrete

ε cmx = maximum concrete compressive strain

ε tmx = maximum concrete tensile strain

𝜀𝑖 = intermediate strain

εo = concrete compressive strain corresponding to f’c

εp = concrete compressive strain corresponding to fp

εsh = reinforcement strain at onset of strain hardening

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

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

εsyield = yield strain of reinforcement

ε os = elastic strain offsets in reinforcement

ε sp = plastic strain offsets in reinforcement

εts = terminal strain in concrete tension softening response

εu = ultimate strain of reinforcement

εx = total axial strain in the x-direction

εy = total axial strain in the y-direction

ε sx = average axial strain in the x-direction, due to shear slip

ε sy = average axial strain in the y-direction, due to shear slip

𝜉 = damping ratio

𝛾 = Newmark gamma factor

γcxy = average net concrete shear strain

γxy = total shear strain

γ sxy = average shear strain due to shear slip


xxviii
𝜇 = the reduction coefficient for the strength 𝑓𝑚𝑥

𝜔 = natural frequency

𝜔𝑚 = ratio between the tensile strength in the x-direction (for 𝛼 = 0) and compressive strength 𝑓𝑚𝑦

Ψ𝑠 = bonded area per unit length

𝜙𝑛 = mode shape

𝜑 = friction angle of the mortar bed joint interface

θ = orientation of principal net concrete strain field, measured counter-clockwise from x-axis

𝜃𝑐𝑟𝑖𝑡 = critical fibre inclination angle which cause a fibre engagement for a given crack width

𝜃𝑓 = fibre inclination angle from axis perpendicular to the crack surface

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

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

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

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

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

ρ = reinforcement ratio

𝜌𝑒 = effective reinforcement ratio of steel bars

𝜌𝐹 = effective reinforcement ratio for FRP sheets

𝜎𝑓,𝑐𝑟,𝑎𝑣𝑔 = average fibre tensile stress at a crack

𝜎𝑓𝑢 = fibre tensile strength

σx = applied axial stress in the x-direction

σy = applied axial stress in the y-direction

[σo] = pseudo prestress vector

𝜏𝑏 = fibre bond strength

𝜏𝑏0 = max average bond stress in steel bars

𝜏𝑒ℎ,𝑚𝑎𝑥 = pull-out strength due to the mechanical anchorage of an end-hooked fiber

𝜏𝑓,𝑚𝑎𝑥 = frictional bond strength


xxix
τxy = applied shear stress

ψ = 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 7, 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, 6 and 7 describe constitutive and behavioral models pertaining to concrete materials, models for other

material types, reinforcement materials, and bond slip, respectively.

Part II, which includes chapters 8 through 12, serves as a user manual for the preprocessor program

FormWorks. These chapters describe the user-interface and the finite element modeling process.

Chapter 13 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. Dynamic analysis in

VecTor2 is also discussed, including an overview of damping and strain rate effects.

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.

σy
y
τxy
ρx, ρy,, fy, ES

σx
f’c, f’t, Ec, εo

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


6
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

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 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, θε, and orientations of principal stress, θσ, are the same

 perfect bond between reinforcement and concrete

 independent constitutive relationships for concrete and reinforcement

 negligible shear stresses in reinforcement.


7
2.2.2 Compatibility Relationships

The compatibility relationships pertain to the average strains in the concrete and reinforcement

components as shown in Figure 2.

y
εcy γcxy/2
1
εc1
θ
1
εc2

2
x
1 εcx

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, εc, and reinforcement, εs, 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 εsx and εsy,

respectively. The compatibility relationships are expressed by the following equations:

ε x = ε cx = ε sy (2.2.2.1)

ε y = ε cy = ε sy (2.2.2.2)

With a value for the shear strain, γxy, relationships from Mohr’s circle of strain determine the average

principal concrete tensile strain, εc1, and the average principal concrete compressive strain, εc2 as follows:

(ε ) [(ε ) ]
1
2 2
ε c1 , ε c 2 = 1
2 x + εy ± 1
2 x − εy + γ xy 2
(2.2.2.3)
8
Relationships from Mohr’s circle also determine the orientations of the average principal tensile strain

axis, θε, and stress axis, θσ, with respect to the x-axis:

 γ xy 
θ = θ ε = θ σ = 12 tan -1   (2.2.2.4)
 εx − εy 
 

2.2.3 Equilibrium Relationships

Consider the free body diagram of the membrane element, in Figure 3.

y σy
τxy

fc1
fcy

fc2
σx vcxy
1

θ fsy
fcx
2 x
fsx

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,

σx and σy, be balanced by the resultants of the average concrete stresses, fcx and fcy, and the reinforcement

stresses fsx and fsy. Equilibrium of moments requires that the applied shear stresses, τxy, are entirely resisted

by average shear stresses in the concrete, vcxy, (assuming the reinforcement does not exhibit dowel action).

These equilibrium relationships for average stresses may be summarized as follows:

σ x = f cx + ρ sx f sx (2.2.3.1)

σ y = f cy + ρ sy f sy (2.2.3.2)

τ xy = v cxy (2.2.3.3)

where ρsx and ρsy are the reinforcement ratios in the x and y directions, respectively.
9
As cracked concrete is orthotropic with respect to the principal stress directions, Mohr’s circle of

stress can be used to relate the average concrete stresses, fcx, and, fcy, to the average principal concrete

tensile stress, fc1:

f cx = f c1 − v cxy ⋅ cot (90 − θ σ ) (2.2.3.4)

f cy = f c1 − v cxy ⋅ tan (90 − θ σ ) (2.2.3.5)

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, fc2, to the principal compressive strain, εc2. Panel test results indicate that the

compressive strength and stiffness decrease as coexisting principal tensile strains, εc1, 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:

f c2 = c
[
f' 2(ε c 2 /ε o ) - (ε c 2 /ε o )
2
] (2.2.4.1)
0.8 − 0.34(ε c1 /ε o )

The term in the numerator is the Hognestad parabolic relationship for concrete in uniaxial compression,

often used for normal strength concretes. The value εo 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, fc1, to the principal tensile strain, εc1. It is first necessary to determine the uniaxial cracking
10
strength, f’t, and corresponding cracking strain, εcr. In the absence of information, they may be estimated

as follows:

f' t = 0.33 f' c (in MPa ) (2.2.4.2)

f' t
ε cr = (2.2.4.3)
Ec

where Ec is the initial tangent stiffness of the concrete, estimated as:

E c = 5000 f' c (in MPa ) (2.2.4.4)

Before cracking, the concrete behaves linear-elastically in tension as follows:

f c1 = E c ⋅ ε c1 for 0 < ε c1 < ε cr (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:

f' t
f c1 = (2.2.4.6)
1 + 200ε c1

As for the reinforcement in compression and tension, the MCFT uses a bilinear relationship

between the average stress, fs and average strain, εs. An initial ascending linear-elastic branch is followed

by a yield plateau, as described by the following equations:

f sx = E s ⋅ ε sx ≤ f sxyield (2.2.4.7)

f sy = E s ⋅ ε sy ≤ f syyield (2.2.4.8)

where Es is the elastic modulus of the reinforcement, and fsxyield and fsyyield 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
11
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

perpendicular to the principal tensile stress direction, and Figure 4b, which depicts the local stresses at the

free surface of the crack.

y y
fscry
fsy
a) θny b)

σx fc1 σx vci
1 1
θ θ
fsx fscrx
θnx
2 2
x x
τxy τxy
σy σy

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:

( )
f c1 = ρ x ( f s crx − f s x ) cos 2 θ nx + ρ y f s cry − f s y cos 2 θ ny (2.2.5.1)

where fscrx and fscry are the local reinforcement stresses at a crack, and θnx and θny 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:
12

( ) ( )
f c1 ≤ ρ x f sxyield − f s x cos 2 θ nx + ρ y f s yyield − f s y cos 2 θ ny (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, vci, are present on the crack

surface. Static equivalency of average and local stresses in the direction tangential to the crack determines

the local shear stresses as follows:

( )
v ci = ρ x ( f s crx − f s x ) cos θ nx ⋅ sin θ nx + ρ y f s cry − f s y cos θ ny ⋅ sin θ ny (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:

f' c
v ci ≤ (in mm, MPa ) (2.2.5.4)
0.31 + 24w/ (a + 26 )

The average crack width, w, is the product of the principle concrete tensile strain and the average crack

spacing perpendicular to the crack, sθ:

w = ε c1 sθ (2.2.5.5)

1
sθ = (2.2.5.6)
cos θ sin θ
+
s mx s my

The average crack spacing in the x-direction, smx, and y-direction, smy, 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.

2.2.6 Consideration of Local Crack Conditions with FRC

Where fibre reinforced concrete (FRC) is used, the fibres increase the maximum permissible tensile and

shear stress at a crack. The following figure depicts the local stresses at the free surface of a crack.
13

Figure 5: Comparison of average and local stresses at a crack


(a) average stresses between crack; (b) local stresses at crack free surface

For equilibrium, from Figure 5 above:

𝑓𝑐1 cos 𝜃 + 𝜌𝑠𝑥 𝑓𝑠𝑥 𝑐𝑜𝑠 𝜃 + 𝛼𝑎𝑣𝑔 𝑓𝑓 cos(𝜃 + 𝜃𝑓 ) = 𝜌𝑠𝑥 𝑓𝑠𝑐𝑟𝑥 cos 𝜃 − 𝑣𝑐𝑖,𝑐𝑟 sin 𝜃 + 𝑓𝑓 cos(𝜃1𝑛 + 𝜃𝑓 )

(2.2.6.1)

𝑓𝑐1 sin 𝜃 + 𝜌𝑠𝑦 𝑓𝑠𝑦 sin 𝜃 + 𝛼𝑎𝑣𝑔 𝑓𝑓 sin�𝜃 + 𝜃𝑓 � = 𝜌𝑠𝑦 𝑓𝑠𝑐𝑟𝑦 sin 𝜃 − 𝑣𝑐𝑖,𝑐𝑟 cos 𝜃 + 𝑓𝑓 sin�𝜃 + 𝜃𝑓 �

(2.2.6.2)

where 𝛼𝑎𝑣𝑔 is the average fibre orientation factor, 𝜃𝑓 is the angle of the tensile stress attained by fibres

from the axis perpendicular to the crack surface, 𝑓𝑓 is the tensile stress due to fibres for a given crack

width, and

𝛿
tan 𝜃𝑓 = 𝑤 𝑠 (2.2.6.3)
𝑐𝑟

Static equivalency of average and local stresses in the direction tangential to the crack determines the

local shear stresses as follows:

𝑣𝑐𝑖,𝑐𝑟 cos�𝜃+𝜃𝑓 � sin�𝜃+𝜃𝑓 �


= 𝜌𝑠𝑥 (𝑓𝑠𝑐𝑟𝑥 − 𝑓𝑠𝑥 ) − 𝜌𝑠𝑦 �𝑓𝑠𝑐𝑟𝑦 − 𝑓𝑠𝑦 � + �1 − 𝛼𝑎𝑣𝑔 �𝑓𝑓 � − �
cos 𝜃 sin 𝜃 cos 𝜃 sin 𝜃

(2.2.6.4)

𝑣𝑐𝑖,𝑐𝑟 = ∑𝑖 𝜌𝑠,𝑖 �𝑓𝑠𝑐𝑟,𝑖 − 𝑓𝑠,𝑖 � sin 𝜃𝑛,𝑖 cos 𝜃𝑛,𝑖 − �1 − 𝛼𝑎𝑣𝑔 �𝑓𝑓 sin 𝜃𝑓

(2.2.6.5)
14
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

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 6.
15
Meanwhile, concrete stresses are attributable only to the continuum straining in response to applied

stresses.

s 1
δs

2
x

Figure 6: Deformation due to crack shear slip

To reconcile this behavior, the DSFM expresses the total strains, εx, εy, and γxy, as the sum of net concrete

s s
strains, εcx, εcy, and γcxy, and strain due to shear slip, ε sx , ε y and γ xy :

ε x = ε cx + ε sx (2.3.1.1)

ε y = ε cy + ε sy (2.3.1.2)

γ xy = γ cxy + γ sxy (2.3.1.3)

Relationships from Mohr’s circle determine the principal net concrete tensile strain, εc1, and the

principal net concrete compressive strain, εc2, as follows:

(ε ) [(ε ) ]
1
2 2
ε c1 , ε c 2 = 1
2 cx + ε cy ± 1
2 cx − ε cy + γ cxy 2
(2.3.1.4)

s s
The crack slip shear strain components, ε sx , ε y and γ xy , are calculated from the average crack

slip shear strain γs. This quantity is defined as the crack slip, δs, divided by the average crack spacing, s,

as follows:

δs
γs = (2.3.1.5)
s

Relationships from Mohr’s circle resolve γs into the components εsx, εsy and γsxy:
16

ε sx = − 12 γ s sin (2θ ) (2.3.1.6)

ε sy = 12 γ s sin (2θ) (2.3.1.7)

γ sxy = γ s cos(2θ) (2.3.1.8)

The orientation of the principal net concrete strains, θ, and orientation of principal concrete stresses, θσ,

with respect to the x-axis are determined from Mohr’s circle relationships involving the net concrete

strain components as follows:

 γ cxy 
θ = θ σ = 12 tan -1   (2.3.1.9)
 ε cx − ε cy 
 

Likewise, the orientation of the principal total strain field, θε, is determined from the total strain

components:

 γ xy 
θ ε = 12 tan -1   (2.3.1.10)
 εx − εy 
 

The difference between the orientation of total strains and the orientation of the principal concrete

stresses defines the rotation lag, ∆θ:

∆θ = θ ε − θ σ (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:

ε sx = ε x (2.3.1.12)

ε sy = ε y (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:

σ x = f cx + ρ sx f sx (2.3.2.1)
17
σ y = f cy + ρ sy f sy (2.3.2.2)

τ xy = vcxy (2.3.2.3)

Additionally, the DSFM incorporates the equilibrium relationships for local stresses at the crack.

( )
f c1 = ρ x ( f s crx − f s x ) cos 2 θ nx + ρ y f s cry − f s y cos 2 θ ny (2.3.2.4)

( )
v ci = ρ x ( f s crx − f s x ) cos θ nx ⋅ sin θ nx + ρ y f s cry − f s y cos θ ny ⋅ sin θ ny (2.3.2.5)

Equilibrium of stress resultants normal to the crack surface results in the following equations:

( )
f c1 = ρ x ( f s crx − f s x ) cos 2 θ nx + ρ y f s cry − f s y cos 2 θ ny (2.3.2.6)

( )
v ci = ρ x ( f s crx − f s x ) cos θ nx ⋅ sin θ nx + ρ y f s cry − f s y cos θ ny ⋅ sin θ ny (2.3.2.7)

The average concrete tensile stress is subject to the limits of the yield strength of the reinforcement

traversing the crack:

( ) ( )
f c1 ≤ ρ x f sxyield − f s x cos 2 θ nx + ρ y f s yyield − f s y cos 2 θ ny (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, βd, to reflect the softening effect of the coexisting principal

tensile strains:

1
βd = ≤ 1.0 (2.3.3.1)
1 +C s C d

The factor, Cd, accounts for the softening effect of transverse tensile strains:
18

C d = 0.35(− ε c1 ε c 2 − 0.28)
0 .8
(2.3.3.2)

The factor Cs 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 Cs = 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 εc1/εc2 since the softening

effect is attributable only to the tensile strains. In this case, Cs = 0.55.

To soften the compressive strength and stiffness of the concrete, the concrete cylinder strength,

f’c, and corresponding peak strain, εo, are both reduced by the reduction factor to determine the peak

compressive stress, fp, and corresponding peak strain εp:

f p = −β d f' c (2.3.3.3)

ε p = −β d ε o (2.3.3.4)

Using the softened parameters, the following relationship determines the relationship between the

principal concrete compressive stress, fc2, and the principal net compressive strain, εc2:

f c2 = f p
(
n ε c2 ε p )
(2.3.3.5)
(n − 1) + (ε c 2 εp )nk

where

n = 0.80 - f p 17 (in MPa ) (2.3.3.6)

 1.0 for ε p < ε c 2 <0


k = (2.3.3.7)
0.67 - f p / 62 for ε c 2 < ε p <0

With regards to concrete in tension, the response before cracking is linear-elastic, as follows:

f c1 = E c ⋅ ε c1 for 0 ≤ ε c1 ≤ ε cr (2.3.3.8)

For cracked concrete, average concrete tensile stresses, f ca1 , 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):

f' t
f ca1 = for ε cr < ε c1 (2.3.3.9)
1 + ct ε c1
19
The coefficient, ct, proposed by Bentz (1999) incorporates the influence of reinforcement bond

characteristics and is computed as follows:

ct = 2.2m (2.3.3.10)

n
1 4 ρi
m
= ∑d
i =1 bi
| cos θ nii | (in mm ) (2.3.3.11)

where dbi is the bar diameter and ρi is the reinforcement ratio of each of the n reinforcement components.

In addition to tension stiffening, post-cracking tensile stresses, fc1b, 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:

b  (ε − ε cr ) 
f c1 = f' t 1 − c1  for ε cr < ε c1 < ε ts (2.3.3.12)
 (ε ts − ε cr ) 

The terminal strain, εts, (the strain at which tensile stresses in plain concrete reduce to zero) is determined

from the fracture energy parameter, Gf, (the area under the stress-strain curve of plain concrete, assumed

to be 75 N/m), and the characteristic length, Lr (assumed to be half the crack spacing):

Gf
ε ts = 2.0 (2.3.3.13)
f' t ⋅Lr

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:

( a
f c1 = max f c1 , f c1
b
) (2.3.3.14)

The DSFM constitutive model for reinforcement in tension or compression is trilinear to account

for strain-hardening phenomenon:

E s ⋅ ε s for 0 < ε s < ε syield




 f syield for ε syield < ε s < ε sh
fs =  (2.3.3.15)
 f syield + E sh (ε s − ε sh ) for ε sh < ε s < ε u


0 for ε u < ε s
20
where Es is the elastic modulus of the reinforcement, fsyield is the yield strength of the reinforcement, εsyield

is the yield strain of the reinforcement, εsh is the strain at the onset of strain hardening, εu 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, δs, to determine the crack slip shear strain, γs. One approach is to relate the crack

slip to the local shear stresses, vci, 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:

a v ci
δs = (2.3.4.1)
1.8w −0.8
( )
+ 0.234 w −0.707 − 0.20 ⋅ f cc

a
a δ
γs = s (2.3.4.2)
s

where fcc 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, θl, between the inclination of the principal total strain axis, θε and the inclination of

the principal stress axis, θσ. 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, ∆θε, of the principal total strain axis, relative to the orientation of the principal strains and

stresses at initial cracking, θic:


21
∆θ ε = θ ε − θ ic

(2.3.4.3)

The post-cracking rotation, ∆θσ, of the principal stress field is then related to ∆θε, by the constant rotation

lag is follows:

 ∆θ ε for | ∆θ ε |≤ θ l
∆θ σ = 
(
 ∆θ ε − θ
l
) for | ∆θ ε |> θ 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:

θ σ = θ ic + ∆θ σ (2.3.4.5)

Finally, relationships from Mohr’s circle determine the shear slip strain, γ bs :

( )
γ bs = γ xy ⋅ cos2θ σ + ε y − ε x ⋅ sin 2θ σ (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:

( a
γ s = max γ s , γ s
b
) (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 7. The following

discussion describes the details of some of these steps.


22

Input analysis control data Determine element Update strain and stress
stiffness matrix, [k] parameters
Input structure data
Determine strain offset
Input external load data displacements Proceed to next
[ ][, ][, ] [ ][, r ]
rco rcp rcs rso s
p load stage?

Update external loads for


load stage, [F] Determine pseudo nodal Analysis complete
load vector [F*]

Determine element strain


offset components Determine total load
[ ][ ][ ] [ ]
[εs] ε o ε p
c
ε os
c εp i s i vector [F’]=[F]+Σ[F*]

Determine average Determine structure


concrete and stiffness matrix
reinforcement stresses [K]=Σ[k]
fc1, fc2, fsi

Determine nodal
Determine local concrete displacements
and reinforcement stresses [r]=[K]-1[F’]
at cracks
vci, fscri
Determine element strains
[ε],[εc], [εs]
Determine crack slip
strain components, [εs]
Determine element
stresses, [σ]
Determine secant moduli

E c1 , E c 2 , G c , E si
Update secant moduli
E 'c1 , E 'c 2 , G 'c , E ' si
Determine material
component stiffness
matrices
[Dc], [Ds]
Secant moduli converged?

Determine composite
material stiffness matrix Steps to be performed for
[D]=[Dc]+Σ[Ds]i Output load stage results each element

Figure 7 VecTor2 nonlinear finite element analysis algorithm


23
2.4.1 Composite Material Stiffness Matrix

In the most general case, total strains, [ε] = [εx, εy, γxy] T, are comprised of net concrete strains

[ε c ] , elastic strain offsets [ε oc ] (due to thermal, prestrains, shrinkage and lateral expansion effects),

[ ]
plastic strain offsets in the concrete, ε cp , (due to cyclic loading or damage), and strains due to crack

shear slip, ε [ ] (as considered by the DSFM).


s

[ε] = [ε c ] + [ε oc ] + [ε cp ] + [ε s ] (2.4.1.1)

As well, compatibility relationships determine that the strain in the ith smeared reinforcement

component is the sum of the total strain, elastic strain offsets ε s [ ] (due to thermal and prestrain effects)
o
i

and plastic strain offsets ε s [ ] (due to cyclic loading or damage):


p
i

[ε s ]i = [ε] + [ε os ] i [ ]
+ ε sp i
(2.4.1.2)

At any point within the reinforced concrete continuum, the total strains are related to stresses [σ]

by the composite material stiffness matrix, [D], as follows:

[σ] = [D][ε] − [σ o ] (2.4.1.3)

The composite material stiffness matrix is the sum of the concrete material stiffness matrix, [Dc], and the

reinforcement component material stiffness matrices, [Ds]i, as follows

n
[D ] = [Dc ] + ∑ [Ds ]i (2.4.1.4)
i =1

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 [σo] calculated

as follows:

[σ ]= [D ]{[ε ] + [ε ] + [ε ]}+ ∑ [D ] {[ε ] + [ε ] }


n
o
c
o
c
p
c
s
s i
o
s i
p
s i (2.4.1.5)
i =1
24
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,[Dc]’, relative to these

directions. If it is assumed that the Poisson’s effect is negligible, then [Dc]’ is computed as follows:

 E c1 0 0 
 
[Dc ]' =  0 E c2 0  (2.4.1.6)
 
 0 0 G c 

The secant moduli E c1 , E c 2 , G c , as shown in Figure 8, are computed from the current values of the

principal stresses, fc1 and fc2, and the corresponding principal net concrete strains, εc1 and εc2, as follows:.

f c1 f E c1 ⋅ E c 2
E c1 = ; E c2 = c2 ; G c = (2.4.1.7)
ε c1 ε c2 E c1 + E c 2

Likewise, material stiffness matrices [Ds]i’ for each reinforcement component must first be

determined relative to their longitudinal axes. As the reinforcement is assumed only to resist uniaxial

stresses, [Ds]i’ is computed as follows:

ρ i E si 0 0
[Ds ]i ' =  0 0 0

(2.4.1.8)
 0 0 0

where ρi is the reinforcement ratio of the reinforcement component. The secant modulus E si , as shown in

Figure 8, is computed from its current value of stress, fsi and the corresponding strain, εsi as follows:

f si
E si = (2.4.1.9)
ε si
25

fc fs

fci fsi

Eci Es

1 1
ε ε
ε oc + ε cp +ε s
εci ε os + ε sp εsi
(a) (b)

Figure 8: Definition of secant moduli for a) concrete b) reinforcement

Likewise, in the case of FRC, the material stiffness matrix for fibres can be determined based on the net

strains in the direction of the crack. The fibre material stiffness matrix, �𝐷𝑓 � , is computed as follows:

����
𝐸𝑓1 0 0

�𝐷𝑓 � = � 0 0 0� (2.4.1.10)
0 0 0

The secant modulus, ����


𝐸𝑓1 , is computed from the tensile stress attained by fibres which is averaged

between cracks and from the strain:

𝛼𝑎𝑣𝑔 𝑓𝑓
����
𝐸𝑓1 = 𝜀 (2.4.1.11)
𝑐𝑓

𝜀𝑐1 +𝜀𝑐2 𝜀𝑐1 −𝜀𝑐2


𝜀𝑐𝑓 = + cos 2𝜃𝑓 (2.4.1.12)
2 2

𝛿
𝜃𝑓 = tan−1 𝑤 𝑠 (2.4.1.13)
𝑐𝑟

The material stiffness matrices, [Dc]’, [Df]’, and [Ds]’i, are transformed from their respective

principal axes to the x,y axes by means of the transformation matrix, [T], as follows:

𝑇 ′
[𝐷𝑐 ] = [𝑇𝑐 ]𝑇 [𝐷𝑐 ]′ [𝑇𝑐 ] + �𝑇𝑓 � �𝐷𝑓 � �𝑇𝑓 � (2.4.1.14)

where [𝑇𝑐 ] is calculated using 𝜃𝜎 , and �𝑇𝑓 � is calculated using 𝜃𝜎 + 𝜃𝑓 , and:

[Ds ]i = [Ts ]i T [Ds ]i ' [Ts ]i (2.4.1.15)

 cos 2 ψ sin 2 ψ cos ψ ⋅ sin ψ 


[T ] =  sin 2 ψ cos 2 ψ - cos ψ ⋅ sin ψ 

(2.4.1.16)
- 2 cos ψ ⋅ sin ψ 2 cos ψ ⋅ sin ψ
 (
cos 2 ψ − sin 2 ψ  )
26
For the concrete, the angle ψ is the inclination of the principal tensile stress axis, θσ, with respect

to the positive x-axis. For the fibres, the angle ψ is the inclination of the principal tensile stress axis,

θσ+θf with respect to the positive x-axis. For the reinforcement, the angle ψ is the orientation, αi, 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:

[k ] = ∫vol [B]T [D][B]dV (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

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, [kc], and stiffness of the reinforcement components, [ks]i, by substituting the respective

material stiffness matrix for the composite material stiffness matrix:

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

[k s ]i = ∫vol [B ]T [Ds ]i [B ]dV (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.
27

For the concrete:

ε oc   ε cxp   ε sx 
 
[ ]  
[ ] [ ]  
ε oc = ε oc  ; ε cp =  ε cyp  ; ε cs =  ε sy 
 
(2.4.3.1)
0  p 
γ s 
  γ  xy 
 cxy 

For the reinforcement components:

ε osi ⋅ (1 + cos 2α i )/ 2 ε sip ⋅ (1 + cos 2α i )/ 2


[ε ]
o
s i
 
[ ]
= ε osi ⋅ (1 − cos 2α i )/ 2 ; ε sp i
 
= ε sip ⋅ (1 − cos 2α i )/ 2 (2.4.3.2)
 ε osi ⋅ sin 2α i   ε sip ⋅ sin 2α i 
   

where αi is the orientation of the reinforcement component relative to positive x-axis.

[ ][ ][ ]
Integration of these strain components results in free nodal displacements rco , rcp , rcs , due to

[ ][ ]
concrete offset strains and shear slip, and displacements rso , rsp due to reinforcement strain offsets, as

follows:

[r ] = ∫ [ε ]dA ; [r ] = ∫ [ε ]dA ; [r ] = ∫ [ε ]dA


c
o o
c c
p p
c c
s s
(2.4.3.3)

[r ] = ∫ [ε ]dA ; [r ] = ∫ [ε ]dA
s
o o
s s
p p
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:

[F * ] = [k c ]{[rco ] + [rcp ] + [rcs ]}+ ∑ [k s ]i {[rso ] + [rsp ]}


n
(2.4.3.5)
i =1

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)
28
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
[K ] = ∑ [k ]i (2.4.4.1)
i =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]:

[F'] = [K ][r ] (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:

[r ] = [K ]−1 [F' ] (2.4.4.3)

The total element strains are determined from the nodal displacements, utilizing the strain-

displacement matrix.

[ε] = [B][r ] (2.4.4.4)

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:

[σ] = [Dc ][{ε] − [ε oc ] − [ε cp ] − [ε s ]}+ [Ds ][{ε] − [ε os ] − [ε op ]} (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:

E (i +1) = (1 − c ) ⋅ E i + c ⋅ E' i (2.4.4.6)


29
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.

For a linear-elastic material that is orthotropic in the principal directions, plane stresses

[σ1 σ2 τ12 ] T are related to plane strains [ε1 ε2 γ 12 ] T , by the material stiffness matrix, [D]:

 E1 v12 E1 0 
1 v E 
[D] = E2 0 (2.5.1.1)
1 − v12 v 21  
21 2

 0 0 (1 − v12 v 21 )G12 

where, v12 is the Poisson ratio relating ε1 to σ2, and v21 is the Poisson ratio relating ε2 to σ1. The number of

independent constants is reduced if the shear modulus, Gc, is approximated as follows:

E1 E 2
G12 ≈ (2.5.1.2)
E1 (1 + v12 ) + E 2 (1 + v 21 )

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:

v12 E1 = v 21 E 2 (2.5.1.3)
30
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 Poisson’s 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 fc2 < fc1 < 0, that

v12 > v21, Ec1 > Ec2 and v12 E c 1 ≠ v 21 E c 2 .

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, [ε ], as described in the preceding section.
o
c

For the purpose of illustration, consider an element subject to biaxial compression such that fc2 < fc1 <0.

The strains in the principal directions due to the Poisson’s effect are as follows:

f c2
ε c1 = - v12 (2.5.1.4)
E c2

f c1
ε c 2 = - v 21 (2.5.1.5)
E c1

For the reasons stated above, suppose that the lateral expansion, εc2, is smaller than the lateral expansion

εc1. The lateral expansion εc2 is modeled entirely by the elastic component, while the lateral expansion εc1

e
is modeled by both residual and elastic components. The elastic component of the Poisson’s ratio, v12 , is

computed by the reciprocity condition:

E c2
v12e = v 21 (2.5.1.6)
E c1
31
The elastic components of the lateral expansion effects are implemented in the symmetric concrete

material stiffness matrix, [Dc] as follows:

 E c1 e
v12 E c1 0 
 
[Dc ] = 1e v 21 E c 2 E c2 0  (2.5.1.7)
1 − v12 v 21  0
 0 (1 − e
v12 v 21 )
G c12 

where,

E c1 E c 2
G c12 (2.5.1.8)
(
E c1 1 + e
v12 )+ E c2 (1 + v 21 )
*
The residual component of the Poisson’s ratio, v12 , is the remaining portion of the expansion:

*
v12 = v12 − v12e (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 Poisson’s ratio:

f c2
ε oc1 = - v12
*
(2.5.1.10)
E c2

Models for determining the value of Poisson’s 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

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:

Ec  f f 
ε cz =  - v12 c 2 − v 21 c1  (2.5.2.1a)
E c + ρ z ⋅ E sz  E c2 E c1 

where ρz 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:

ρ z ⋅ f z, yield f c2 f c1
ε cz = - - v12 − v 21 (2.5.2.1b)
Ec E c2 E c1
32
The stress, fsz, in the out-of-plane reinforcement is determined as follows:

f sz = E s ε cz ≤ f z , yield (2.5.2.2)

where fz,yield is the yield strength of the out-of-plane reinforcement.

The resulting out-of-plane concrete compressive stress, fcz,, is determined from equilibrium as follows:

f cz = −ρ z ⋅ f sz

(2.5.2.3)

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,

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 Mohr’s circle.

If the inclination of the principal net concrete strains is θ, then the following relationships

p p
transform plastic strains, ε cxp , ε cy, γ cxy, with respect to the X,Y axes to plastic strains, ε c1p , ε c2
p
, with

respect to the principal 1,2 directions:

p p p p
ε cx + ε cy ε cx − ε cy
ε cp1 = + p
⋅ cos 2θ + γ cx ⋅ sin 2θ (2.5.3.1)
2 2
33
p p p p
ε cx + ε cy ε cx − ε cy
ε cp2 = − p
⋅ cos 2θ − γ cx ⋅ sin 2θ (2.5.3.2)
2 2

p p
In a given load step, the plastic strains change by the quantities ∆ε c1 and ∆ε c2 , 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:

∆ε cp1 ∆ε cp2
ε cxp ' = ε cxp + ⋅ (1 + cos 2θ) + ⋅ (1 − cos 2θ) (2.5.3.3)
2 2

p p
∆ε c2 ∆ε c2
ε cyp ' = ε cyp + ⋅ (1 − cos 2θ) + ⋅ (1 + cos 2θ) (2.5.3.4)
2 2

p
γ cxy p
' = γ cxy + ∆ε c1p ⋅ sin 2θ − ∆ε c2p ⋅ sin 2θ (2.5.3.5)

Maximum concrete strains are defined by maximum compressive strains, ε cmx , ε cmy, and γ cmxy ,

and maximum tensile strains ε tmx , ε tmy, and γ tmxy , 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 ε cmx , ε cmy, and γ cmxy describe a compressive strain envelope, rather

than a strain state, analogous relationships from Mohr’s circle transform them to and from the principal

and global axes.

The following relationships transform the maximum compressive strains, ε cmx , ε cmy, and γ cmxy ,

with respect to the x,y axes to maximum compressive strains, strains εcm1 and εcm2, with respect to the

principal 1,2 directions:

ε cmx + ε cmy ε cmx − ε cmy


ε cm1 = + ⋅ cos 2θ + γ cmx ⋅ sin 2θ (2.5.3.6)
2 2

ε cmx + ε cmy ε cmx − ε cmy


ε cm 2 = − ⋅ cos 2θ − γ cmx ⋅ sin 2θ (2.5.3.7)
2 2

In a given load step, the calculated principal net concrete strains are εc1 and εc2. If these principal

strains are more compressive than those described by εcm1 and εcm2, then the envelope of compressive
34
strains must be updated. The incremental maximum compressive strains in the principal directions are

defined as follows:

0 if ε c1 > ε cm1


∆ε cm1 =  (2.5.3.8)
ε c1 − ε cm1 if ε c1 < ε cm1

 0 if ε c 2 > ε cm 2
∆ε cm 2 =  (2.5.3.9)
ε c 2 − ε cm 2 if ε c 2 < ε cm 2

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:

∆ε cm1 ∆ε cm 2
ε cmx ' = ε cmx + ⋅ (1 + cos 2θ) + ⋅ (1 − cos 2θ) (2.5.3.10)
2 2

∆ε cm1 ∆ε cm 2
ε cmy ' = ε cmy + ⋅ (1 − cos 2θ) + ⋅ (1 + cos 2θ) (2.5.3.11)
2 2

γ cmxy ' = γ cmxy + ∆ε cm1 ⋅ sin 2θ − ∆ε cm 2 ⋅ sin 2θ (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

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
35
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, [ε], 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, ε cp and

[ε ]. Therefore, the newly engaged elements effectively behave as though they have been added in a zero
p
s

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.

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.
36
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.

2.5.6 Dynamic Analysis

2.5.6.1 Introduction

The structural modeling in FormWorks for dynamic analysis is much the same as for static analysis.

The mesh generation and material model selection methods are identical for dynamic and static analysis.

However, for certain dynamic analyses, parameters must be manually added to the analysis input files

created by FormWorks.
37
2.5.6.2 Newmark Direct Integration Method: Newmark Beta and Gamma Factors

VecTor2 employs Newmark’s method of direct integration to evaluate the dynamic response of

structures. The iterative Newmark integration method was introduced in 1959 (Wilson, 2002).

1
𝑢𝑡 = 𝑢𝑡−𝛥𝑡 + 𝛥𝑡 𝑢̇ 𝑡−𝛥𝑡 + � − 𝛽� 𝛥𝑡 2 𝑢̈ 𝑡−𝛥𝑡 + 𝛽𝛥𝑡 2 𝑢̈ 𝑡 (2.5.6.2.1)
2

𝑢̇ 𝑡 = 𝑢̇ 𝑡−𝛥𝑡 + (1 − 𝛾)𝛥𝑡𝑢̈ 𝑡−𝛥𝑡 + 𝛾𝛥𝑡𝑢̈ 𝑡 (2.5.6.2.2)

Updated in 1962 to eliminate the need for iteration and to introduce damping, the following equations

resulted:

𝑢̈ 𝑡 = 𝑏1 (𝑢𝑡 − 𝑢𝑡−𝛥𝑡 ) + 𝑏2 𝑢̇ 𝑡−𝛥𝑡 + 𝑏3 𝑢̈ 𝑡−𝛥𝑡 (2.5.6.2.3)

𝑢̇ 𝑡 = 𝑏4 (𝑢𝑡 − 𝑢𝑡−𝛥𝑡 ) + 𝑏5 𝑢̇ 𝑡−𝛥𝑡 + 𝑏6 𝑢̈ 𝑡−𝛥𝑡 (2.5.6.2.4)

where
1 1 1
𝑏1 = ; 𝑏2 = ; 𝑏3 = 𝛽 − ; 𝑏4 = 𝛾𝛥𝑡𝑏1 ; 𝑏5 = 1 + 𝛾𝛥𝑡𝑏2 ; 𝑏6 = 𝛥𝑡(1 + 𝛾𝑏3 − 𝛾)
𝛽𝛥𝑡 2 𝛽𝛥𝑡 2

The Newmark Beta Factor, β, and Newmark Gamma Factor, γ, are the constants used in the

method. The default value for γ in VecTor2 is 0.5; any other value will introduce artificial numerical

damping to the solution (Saatci, 2007). The β factor determines how the acceleration changes within a

time-step. A value of β = 0.25 is used in the integration when the acceleration is constant within a time-

step (constant acceleration method), while β = 1/6 is typically used when the acceleration changes

linearly through the time-step (linear acceleration method). The use of other values for β may greatly

affect the accuracy and stability of the solution, and are therefore not often used. The constant

acceleration method yielded the more stable results for the analyses found in Saatci (2007), and is the

recommended method. Thus, the default value for 𝛽 in VecTor2 is 0.25.

2.5.6.3 Damping

VecTor2 employs both Rayleigh and Alternative Damping. These two types of damping are

discussed in more detail in Section 2.5.9.


38
2.5.6.4 Ground Acceleration

Ground acceleration can be used in dynamic modeling in VecTor2, and can be input in two ways.

Ground acceleration may be specified in FormWorks or in a VECTOR.EQR file. The procedure for

inputting this data is discussed in more detail in the FormWorks section of this manual.

2.5.7 Accounting for Strain History

Accounting for strain history is one of the analysis options in the Model tab of the Define Job

dialog box. Accounting for strain history is the recommended option since it is important to consider

previous loads, particularly when performing cyclic or reverse cyclic analyses. If previous loading is

neglected, at the beginning of each new load stage, VecTor2 assumes that the structure is uncracked and

not damaged in any way.

2.5.8 Accounting for Strain Rate Effects

Accounting for strain rate effects becomes important when modeling structures under impact

loading or other high strain rate loading conditions. VecTor2 accounts for increased strain rates by

calculating Dynamic Increase Factors (DIFs) for different material types and loading conditions. For

concrete in tension and compression, the CEB-FIP 1990 Model Code DIFs are used. For both smeared

and truss steel reinforcement, the CEB-FIP 1988 bulletin 187 DIF formulations are used. For structural

steel, the DIFs are taken as 1 always.

In the following formulations, the DIF is equal to the ratio of the dynamic strength to the static

strength, dynamic modulus to static modulus, or dynamic strain to static strain. The CEB-FIP 1990

Model Code provides the following formulations strain rate effects for concrete.

For concrete in compression, the following CEB-FIP formulations are valid for strain rates in the

range of 30 × 10−6 s−1 < | εċ | < 3 × 102 s−1 (CEB-FIP, 1990):

𝜀̇ 1.026𝛼
� 𝑐̇ � 𝑓𝑜𝑟 𝜀𝑐̇ ≤ 30 𝑠 −1
𝑓𝑐𝑑 𝜀0
𝑓𝑐𝑠
=� 1 (2.5.8.1)
𝜀̇ 3
𝛾 �𝜀𝑐̇ � 𝑓𝑜𝑟 𝜀𝑐̇ > 30 𝑠 −1
0
39
where,
1
𝛼= 𝑓 (2.5.8.2)
5+9 𝑐𝑠
𝑓𝑐0

𝛾 = 106.156𝛼−2.0 (2.5.8.3)

𝜀0̇ = 30 × 10−6 𝑠 −1

𝑓𝑐0 = 10 𝑀𝑃𝑎

And 𝜀𝑐̇ is the strain rate, 𝑓𝑐𝑠 is the static compressive strength of concrete, 𝑓𝑐𝑑 is the dynamic

compressive strength of concrete.

In VecTor2, the equation for γ is slightly different, and is calculated as follows:

𝛾 = 106.156𝛼−0.492 (2.5.8.4)

For concrete in tension, for strain rates 3 × 10−6 s−1 < εcṫ < 3 × 102 s −1 , the following CEB-FIP

formulations apply:

𝜀̇ 1.016𝛿
� 𝜀𝑐𝑡̇ � 𝑓𝑜𝑟 𝜀𝑐𝑡̇ ≤ 30 𝑠 −1
𝑓𝑡𝑑 0
𝑓𝑡𝑠
=� 1 (2.5.8.5)
̇ 3
𝜀𝑐𝑡 −1
𝛽 �𝜀̇ � 𝑓𝑜𝑟 𝜀𝑐𝑡̇ > 30 𝑠
0

where,
1
𝛿= 𝑓 (2.5.8.6)
10+6 𝑐𝑠
𝑓𝑐0

𝛽 = 107.112𝛿−2.33 (2.5.8.7)

𝜀0̇ = 3 × 10−6 𝑠 −1

𝑓𝑐0 = 10 𝑀𝑃𝑎

And 𝑓𝑡𝑠 is the static tensile strength of concrete, and 𝑓𝑡𝑑 is the dynamic tensile strength of

concrete. In VecTor2, the formulation for β is:

𝛽 = 106.933𝛿−0.492 (2.5.8.8)

The CEB-FIP 1990 Model Code also provides a formulation for the effect of strain rate on the

modulus of elasticity of concrete.

In compression, for compressive strain rates within the range of 30 × 10−6 < |ε̇ | < 3 × 102 :

𝐸𝑐,𝑑 𝜀̇ 0.026
= �𝜀 𝑐̇ � (2.5.8.9)
𝐸𝑐𝑖 𝑐0
40
And in tension, for tensile strain rates within the range of 3 × 10−6 s −1 < εcṫ < 3 × 102 s −1 :

𝐸𝑐,𝑑 𝜀̇ 0.016
𝐸𝑐𝑖
= �𝜀 𝑐𝑡̇ � (2.5.8.10)
𝑐𝑡0

where,

𝜀𝑐0̇ = 30 × 10−6 𝑠 −1 for compression

̇ = 3 × 10−6 𝑠 −1 for tension


𝜀𝑐𝑡0

and 𝐸𝑐,𝑑 is the dynamic modulus of elasticity, 𝐸𝑐𝑖 is the modulus of elasticity of concrete at 28

days, and 𝜀̇𝑐 is the strain rate.

The formulation employed in VecTor2 for peak strain at maximum load, valid for compressive strain

rates of 30 × 10−6 < |ε̇ | < 3 × 102 and tensile strains rates of 3 × 10−6 s −1 < εcṫ < 3 × 102 s −1 :

𝜀𝑐1,𝑑 𝜀̇ 0.02
𝜀𝑐1
= �𝜀 𝑐̇ � (2.5.8.11)
𝑐0

where 𝜀𝑐1,𝑑 is the impact/dynamic strain at maximum load, 𝜀𝑐1 is the strain at maximum load for static

loading, and 𝜀̇𝑐0 and 𝜀̇𝑐𝑡0 have the same values as stated above.

For reinforcing steel, the CEB-FIP Bulletin 187 formulations are employed. The CEB-FIP

formulations are valid for steel strain rates of 5 × 10−5 s −1 ≤ εṫ ≤ 10 s −1 for both smeared and truss bar

reinforcement.

For the yield strength of reinforcement:

𝑓𝑦,𝑑 6 𝜀̇
𝑓𝑦
=1+ ∙ 𝑙𝑛 �𝜀 𝑡̇ � (2.5.8.12)
𝑓𝑦 𝑡0

And for ultimate strength of reinforcement steel:

𝑓𝑢,𝑑 7 𝜀̇
𝑓𝑢
=1+ ∙ 𝑙𝑛 �𝜀 𝑡̇ � (2.5.8.13)
𝑓𝑢 𝑡0

where

𝜀𝑡0̇ = 5 × 10−5 𝑠 −1
41
2.5.9 Structural Damping

Damping refers to the diminishing of the amplitude of the free vibration of a structure. The

damping matrix for a given structure is determined from its modal damping ratios. VecTor2 offers two

types of structural damping, Rayleigh Damping and Alternative Damping. In Rayleigh Damping,

specifying the damping ratio for two damping modes defines the damping frequency for all modes.

Alternative damping, consisting of the superposition of modal damping ratios, allows the user to specify

zero viscous damping for the final damping modes.

2.5.9.1 Rayleigh Damping

One of the most common formulations for proportional damping is Rayleigh Damping. In

Rayleigh Damping it is assumed that damping is proportional to both the mass and stiffness matrices

(Chopra, 2007).

For mass-proportional damping, the generalized damping equation for the nth mode is

𝐶𝑛 = 𝑎0 𝑀𝑛 (2.5.9.1.1)
𝑎0 1
where the modal damping ratio is 𝜁𝑛 = 2 𝜔𝑛
.

To calculate a specific value of damping ratio for any one mode, the equation 11.4.3 can be

modified as follows.

𝑎0 = 2𝜁𝑖 𝜔𝑖 (2.5.9.1.2)

Once 𝑎0 is determined, the damping matrix and damping ratio for every other mode is determined.

Likewise, for stiffness-proportional damping, the coefficient 𝑎1 is determined.

𝑎1 2𝜁𝑗
𝐶𝑛 = 𝑎1 𝜔𝑛2 𝑀𝑛 𝑎𝑛𝑑 𝜁𝑛 = 2
𝜔𝑛 ; 𝑎1 = 𝜔𝑗
(2.5.9.1.3)

Combining both mass-proportional and stiffness-proportional damping, Rayleigh damping is defined as

𝑐 = 𝑎0 𝑚 + 𝑎1 𝑘 (2.5.9.1.4)

Where the damping ratio is


𝑎0 1 𝑎1
𝜁𝑛 = 2 𝜔𝑛
+ 2
𝜔𝑛 (2.5.9.1.5)
42

Figure 9: Rayleigh Damping (Chopra, 2007)

The coefficients a0 and a1 can be determined from the specific damping ratios ζi and ζj for the ith and jth

modes.

1 1/𝜔𝑖 𝜔𝑖 𝑎0 𝜁𝑖
� � � � = � 𝜁𝑗 � (2.5.9.1.6)
2 1/𝜔𝑗 𝜔𝑗 𝑎1

If the same damping ratio is assumed for both modes, 𝑎1 and 𝑎0 can be expressed as follows:

2𝜔𝑖 𝜔𝑗 2
𝑎0 = 𝜁 𝜔 +𝜔 𝑎𝑛𝑑 𝑎1 = 𝜁 (2.5.9.1.7)
𝑖 𝑗 𝜔𝑖 +𝜔𝑗

2.5.9.2 Alternative Damping

Alternative Damping allows for damping behavior for the first j modes and no viscous damping

for the remaining modes.

The damping matrix for alternative damping is constructed as follows (Chopra, 2007):

2𝜁𝑛 𝜔𝑛
𝑐 = 𝑚 �∑𝑁
𝑛=1 𝑀𝑛
𝜙𝑛 𝜙𝑛𝑇 � 𝑚 (2.5.9.2.1)

where 𝑀𝑛 is the generalized mass matrix, 𝜙𝑛 is the mode shape, 𝑚 is the mass matrix, 𝑁 is the

total dynamic degrees of freedom.

The generalized mass for the nth node is:

𝑀𝑛 = 𝜙𝑛𝑇 × 𝑚 × 𝜙𝑛 (2.5.9.2.2)

The nth term in the big equation is the contribution of the nth mode to the damping matrix. Only the first

𝑗 modes in equation contribute to the damping response.


43
2.5.10 Geometric Nonlinearity

Geometric nonlinearity is used to consider secondary displacements and P-Delta effects and other

large displacements.

2.5.11 Crack Allocation

Two crack allocation models are available in VecTor2, uniform and variable. The uniform crack

allocation model is useful when determining crack widths and spacing for reinforced concrete with

regular deformed reinforcing steel or fibre reinforced concrete; the DSFM is used to determine an average

crack width and spacing. For concrete reinforced with FRP, the variable crack allocation model may be

more appropriate.

2.5.11.1 Uniform Crack Allocation (Deluce, Lee, and Vecchio, 2012)

In normal reinforced concrete, an average crack width and spacing are calculated using the DSFM

as follows (Vecchio, 2000):

𝑤 = 𝜀𝑐1 ∙ 𝑠 (2.5.11.1.1)

where 𝑤 is the average crack width, and 𝜀𝑐1 is the average tensile strain.

Under biaxial conditions, the following crack spacing parameters are calculated with respect to the

principal axis:

𝑠 𝑘1 𝑘2
𝑠 = 𝑠𝑐𝑟 = 2 �𝑐 + 10𝑏 � 𝑘3 + 𝑠𝑚
(2.5.11.1.2)

where,
1
𝑠𝑏 = 𝑐𝑜𝑠2 𝜃𝑖
(2.5.11.1.3)
∑𝑖
𝑠𝑏,𝑖

𝜋𝑑 2
𝑠𝑏𝑖 = 0.5� 𝜌 𝑏,𝑖 (2.5.11.1.4)
𝑠,𝑖

𝜌 𝛼𝑓 𝑉𝑓 𝑙 /𝑑
𝑠𝑚 = ∑𝑖 𝑑 𝑠,𝑖 𝑐𝑜𝑠 2 𝜃𝑖 + 𝑑𝑓
∙ 𝑚𝑎𝑥 � 𝑓50 𝑓 , 1.0� (2.5.11.1.5)
𝑏,𝑖

𝑚𝑖𝑛�𝑉𝑓 ,0.015� 50
𝑘1 = 0.4; 𝑘2 = 0.25; 𝑘3 = 1 − 0.015
∙ �1 − 𝑚𝑖𝑛 � , 1.0�� (2.5.11.1.6)
𝑙𝑓 /𝑑𝑓

𝑐 = 1.5𝑎𝑔𝑔 (2.5.11.1.7)
44
And agg is the maximum aggregate size, α is a fibre orientation factor (equal to 0.5 for an infinite

element).

2.5.11.2 Variable Crack Allocation (Sato 2002)

The variable crack allocation model is well-suited for the estimation of crack widths in reinforced

concrete members with fibre-reinforced polymer sheets.

To estimate the contribution of steel bars to crack formation, the stress at the final crack

formation is (Sato and Vecchio, 2003):

𝑠𝑟 𝛹𝑠,𝑖 𝜏𝑏0,𝑖 𝑐𝑜𝑠𝜃𝑠,𝑖


𝑓1′ = ∑𝑛𝑖=1 2𝐴𝑐𝑒
(2.5.11.2.1)

where 𝑠𝑟 is the crack spacing, Ψ𝑠 is the bonded area per unit length (mm2/mm), 𝜏𝑏0 is the maximum

average bond stress of the ith steel bar, 𝜃𝑠 is the angle between the reinforcement and principle tensile

concrete direction, and 𝐴𝑐𝑒 is the effective cross-sectional area of concrete.

The contribution of the FRP sheets to crack formation is as follows. Concrete stress reaches the

tensile strength ft′ at the midpoint between cracks, and a new crack develops at this section.

For equilibrium at a new crack formation:

𝑤𝐹,𝑗 𝑡𝐹,𝑗 𝐸𝐹,𝑗 𝛥𝜀𝐹0,𝑗 𝑐𝑜𝑠2 𝜃𝐹,𝑗


𝑓𝑡′ = ∑𝑛𝑗=1 𝐴𝑐𝑒
(2.5.11.2.2)

However, the only term required to model crack formation is 𝑡𝐹 𝐸𝐹 𝛥𝜀𝐹0 :

𝑡𝐹 𝐸𝐹 𝛥𝜀𝐹0𝑚𝑎𝑥 = 𝑐3 × 𝑚𝑖𝑛(1, 𝑠𝑟 /220) N/mm (2.5.11.2.3)

where

𝑐3 = �15.8 + 1.34�𝑡𝐹 𝐸𝐹 ��𝐺𝐹 N/mm (2.5.11.2.4)

When the average tensile stress in the cross-section of 𝑅𝑒𝐹 × 𝑤𝐹 at the midpoint between two cracks

approaches the tensile strength ft′ , a new crack develops.

Combining the two equations, equilibrium at the completion of new crack formation in

reinforced concrete members with both steel bars and FRP sheets is expressed in the following equation.

𝜌𝑒,𝑖 𝜏𝑠0,𝑖 𝑐𝑜𝑠𝜃𝑠,𝑖


𝑓𝑡′ = 2𝑠𝑟 ∑𝑚
𝑖=1 𝑑𝑏,𝑖
+ ∑𝑛𝑗=1�𝜌𝐹,𝑗 𝐸𝐹,𝑗 𝛥𝜀𝐹𝑚𝑎𝑥,𝑗 𝑐𝑜𝑠 2 𝜃𝐹,𝑗 � (2.5.11.2.5)
45
where 𝜌𝑒 is the effective reinforcement ratio for steel bars, 𝜌𝐹 is the effective reinforcement ratio

for FRP sheets (=𝑡𝐹 /𝑅𝑒𝐹 ), 𝑖 is the component (direction of steel bars), and j is the component of FRP

sheets.

The crack spacing can be calculated as:

𝑓𝑡′
⎧ 𝜌 𝜏 𝑐𝑜𝑠𝜃 𝜌𝐹,𝑗𝑐3,𝑗 𝑐𝑜𝑠2 𝜃𝐹,𝑗
𝑓𝑜𝑟 𝑠𝑟 ≤ 220𝑚𝑚
𝑐,𝑖 𝑏0,𝑖 𝑠,𝑖 1
⎪ 2 ∑𝑚
𝑖=1 + ∑𝑛
𝑑𝑏,𝑖 220 𝑗=1 𝑡𝑓,𝑗

𝑠 = 𝑠𝑟 = ′
2
𝑚 𝜌𝑓,𝑗 𝑐3,𝑗 𝑐𝑜𝑠 𝜃𝐹,𝑗
(2.5.11.2.6)
⎨𝑓𝑡 −∑𝑗=1 𝑡𝐹,𝑗
⎪ 𝑛 𝜌𝑒,𝑖 𝜏𝑏0,𝑖 𝑐𝑜𝑠𝜃𝑠,𝑖 𝑓𝑜𝑟 𝑠𝑟 > 220𝑚𝑚
⎩ 2 ∑𝑖=1 𝑑𝑏,𝑖

Details of the variable crack allocation model can be found in Sato (2003).

2.5.12 Additional Analysis Parameters

Different analysis options or material models available in VecTor2 require the user to specify

additional parameters. These are discussed in this section. Additionally, it is possible for users to specify

material resistance factors in FormWorks; this is discussed in this section and in the FormWorks section

of the manual.

2.5.12.1 Concrete Aggregate Type

The type of aggregate is important when considering fire loads, due to the fact that siliceous and

carbonate aggregates respond differently to elevated temperatures. Some characteristic properties, such

as 𝑓𝑐′ /𝑓𝑐20 are found in (Zhou, 2005). The thermal strains also differ between siliceous and carbonate

aggregate.

2.5.12.2 Concrete Conductivity

Concrete conductivity is important when determining thermal performance. In order to analyze

the response of concrete under extreme thermal loading (ie. fire), this input is required.

2.5.12.3 Concrete Fracture Energy

In order to predict crack formation in reinforced concrete, the fracture energy is required. The

fracture energy depends on the composite structure of the concrete and depends on the mechanical

interaction of the aggregates and cement matrix (Wittman, 2002). The concrete fracture energy is lowest
46
for pure hardened cement paste and increases as the maximum aggregate size increases. One equation for

determining the fracture energy is the following (Wittman, 2002):


𝑛
𝐺𝑓 = 𝑎𝛷𝑚𝑎𝑥 N/m (2.5.12.3.1)

Where 𝑎 = 80.6 and 𝑛 = 0.32.

Note that in VecTor2, the concrete fracture energy must be specified in units of kN/m. If a zero

is entered in FormWorks for fracture energy, a default value of 0.075 kN/m is assigned in VecTor2.

2.5.12.4 Prestressing Friction and Wobble Coefficient

The pre-stressing friction coefficient, specified in units of /m, is dependent on the type of tendon

and the radius of curvature of the tendon. Applicable to post-tensioned systems only, the range of friction

coefficients recommended by ACI and CEB is summarized in Figure 10 below (Collins and Mitchell,

1997). Similar to the friction coefficient, the prestressing wobble coefficient is applied only to post-

tensioned systems and is dependent on the type of tendon. Recommended values are also given in Figure

10.

Figure 10: Prestressing Wobble and Friction Coefficients


47
2.5.12.5 Thermal Time-Stepping Factor

The thermal time stepping factor allows transient thermal loading (ie. fire loading) to be

incorporated into VecTor2. The time step is used in the Crank-Nicolson Method, which is a finite

difference numerical integration method implemented in VecTor2 for the analysis of transient thermal

loading. Multiplying the thermal time stepping factor by the number of time steps gives the duration of

each time step.

2.5.12.6 Material Resistance/Creep Factors

Material Resistance Factors are applied to decrease the strength of specified materials. The

factored strength is equal to the input material strength multiplied by the material resistance factor.

Information on how to input the resistance factors for concrete, rebar steel, prestressing steel, structural

steel, masonry/mortar, and wood are provided in the FormWorks section of this manual. The concrete

creep coefficient and prestressing steel relaxation coefficients are discussed in the Concrete Models

section of this manual.


48

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:

[k ] = ∫vol [B]T [D][B]dV (3.1.1)

3.1.1 Constant Strain Triangle

The constant strain triangle, shown in Figure 11, 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
49
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.

y k
riy

i rix

thickness = t j
x

Figure 11: 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, [ε], are related to the nodal displacements, [r], by the strain-

displacement matrix [B] as follows:

[ε] = [Β][r ] (3.1.1.1)

where

[ε] = [ε x ε xy γ xy ]
T
(3.1.1.2)

[r ] = [rix riy r jx r jy rkx rky ]


T
(3.1.1.3)

 y jk 0 y ki 0 y ij 0
1  
[B] =  0 x kj 0 xik 0 x ji  (3.1.1.4)
2A
 x kj y jk xik y ki x ji y ij 

xij = xi - x j and y ij = y i - y j (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.


50
3.1.2 Plane Stress Rectangle

The plane stress rectangle, shown in Figure 12, 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.

2a

n m

2b
riy
rix
i j
thickness = t x

Figure 12: 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, [ε], are related to the nodal

displacements, [r], by the strain-displacement matrix, [B], as follows:

[ε] = [Β][r ] (3.1.2.1)

[ε] = [ε x ε xy γ xy ]
T
(3.1.2.2)

[r ] = [rix riy r jx r jy rmx rmy rnx rny ]


T
(3.1.2.3)

− (b − y ) 0 (b − y ) 0 (b + y ) 0 − (b + y ) 0 
 
[B] = 1  0 − (a − x ) 0 − (a + x ) 0 (a + x ) 0 (a − x )  (3.1.2.4)
4ab  
 − (a − x ) − (b − y ) − (a + x ) (b − y ) (a + x ) (b + y ) (a − x ) − (b + y )
51
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

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 13a, 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.

y y

n n
m m
TB
i i TA

j j
thickness = t x thickness = t x
(a) (b)

Figure 13: a) Quadrilateral element b) Decomposition of quadrilateral element into two constant strain
triangle elements

VecTor2 divides the quadrilateral element into two constant strain triangles TA and TB, sharing the

shortest diagonal as a common edge, as shown in Figure 13b. 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.
52
Two different types of quadrilateral elements are available in VecTor2, isoparametric and

degenerate. Both element types are discussed briefly below.

3.1.3.1 Isoparametric Quadrilateral Elements

For isoparametric elements, the same shape functions are used to represent the element geometry and

determine the displacements. In this case, the quadrilateral coordinates are the parameters that define the

shape functions, which are used to link the geometry and displacements.

3.1.3.2 Degenerate Quadrilateral Elements

Certain quadrilateral elements will degenerate into triangular elements. This is done by collapsing

one of the sides of the quadrilateral. For a 4-node quadrilateral, degeneration results in a 3-node

triangular element as illustrated in Figure 14..

Figure 14: Degenerate Quadrilateral Element (Li, 2004)

Degenerate elements allow quadrilateral and triangular elements to both be used, which is convenient

when dealing with complex structural geometry. Using degenerate elements, the shape functions of the

quadrilateral elements are defined in terms of local coordinates and the shape functions for the triangular

elements are defined in terms of global coordinates (Li, 2004).


53
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 15, 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.

y
j

α
i

thickness = t x

Figure 15: 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]:

[F ] = [k ][r ] (3.2.1.1)

where

[F ] = [Fix Fiy F jx F jy ]
T
(3.2.2.2)
54

[r ] = [rix riy r jx r jy ]
T
(3.2.2.3)

 cos 2 α cos α ⋅ sin α - cos 2 α - cos α ⋅ sin α 


 
 cos α ⋅ sin α 2
sin α − cos α ⋅ sin α - sin α 
2

[k ] = AE   (3.2.3.4)
L  cos 2 α - cos α ⋅ sin α cos 2 α cos α ⋅ sin α 
 
− cos α ⋅ sin α 2
− sin α cos α ⋅ sin α sin α 
2

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.
55
3.3.1 Link

The link element, (Ngo and Scordelis, 1967), shown in Figure 16, 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.

y concrete element
(e.g. quadrilateral) riy

link element θ
i
(nondimensional) rix

i,j kt
kr
j
truss bar
x

Figure 16 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]:

[∆] = [T ][r ] (3.3.1.1)

where,

[∆] = [∆ t ∆r ]
T
(3.3.1.2)
56

[r ] = [rix riy r jx r jy ] T
(3.3.1.3)

− cos θ − sin θ cos θ sin θ 


[T ] =   (3.3.1.4)
 sin θ − cos θ − sin θ cos θ

The force in the tangential spring, Ft, is found by multiplying the bond slip, ∆t, by the stiffness, kt,

and the bonded surface area tributary to the link, A. Likewise, the force in the radial spring, Fr, is found

by multiplying the deformation of the radial spring, ∆r, by the stiffness kr, and the bonded surface areas,

A, of embedded bars or external plates tributary to the link:

 Ft  k t 0 ∆t 
  = A   (3.3.1.5)
 Fr   0 k r  ∆ r 

The stiffness, kt, relates to the bond stress to the bond slip and is determined from the bond-slip curve. In

VecTor2, kr is assigned a large value of 100·kt, 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.

 Fix 
F 
 iy  = [T ]T  Ft  (3.3.1.6)
 F jx  F 
 r
 
F
 jy 

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]:

[F ] = [k ][r ] (3.3.1.7)

k t 0
[k ] = [T ]T A [T ] (3.3.1.8)
 0 k r 

3.3.2 Contact

The contact element, shown in Figure 17, 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
57
slippage, nodes, j and k must have the same coordinates. Either node j or node k must be attached to a

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.

y concrete element
t
(e.g. quadrilateral)
rmy
contact element r
θ
m
rmx
m,n n

j,k truss bar j


k
x

Figure 17: 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, τ,

is related to the tangential displacement, ∆t, by the stiffness, kt; the radial stress, σ, is related to the radial

displacement by the stiffness, kr, as follows:

 τ  k t 0 ∆ t 
 =   (3.3.2.1)
σ  0 k r  ∆ r 

The stiffness, kt, relates to the bond stress to the bond slip and is determined from the bond-slip curve. In

VecTor2, kr is assigned a large value of 100·kt, 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:
58
[F ] = [k ][r ] (3.3.2.2)

where

[F ] = [Fix Fiy F jx F jy Fmx Fmy Fnx Fny ] T


(3.3.2.3)

[r ] = [rix riy r jx r jy rmx rmy rnx rny ]


T
(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]:

[k ] = [T ]T [k ]' [T ] (3.3.2.5)

where

 2k t 0 − 2k t 0 kt 0 − kt 0 
 
 0 2k r 0 - 2k r 0 kr 0 - kr 
 
 − 2k t 0 2k t 0 - kt 0 kt 0 
 
 0 - 2k r 0 2k r 0 − kr 0 kr 
[k ]' = A  

(3.3.2.6)
6 kt 0 − kt 0 2k t 0 − 2k t 0
 
 0 kr 0 - 2k r 0 2k r 0 − 2k r 
 
 
 − kt 0 kt 0 - 2k t 0 2k t 0 
 
 0 − kr 0 kr 0 - 2k r 0 2k r 

cos θ sin θ 0 0 0 0 0 0 
 sin θ cos θ 0 0 0 0 0 0 

 0 0 cos θ sin θ 0 0 0 0 
 
 0 0 - sin θ cos θ 0 0 0 0 
[T ] =  (3.3.2.7)
0 0 0 0 cos θ sin θ 0 0 
 
 0 0 0 0 - sin θ cos θ 0 0 
 0 0 0 0 0 0 cos θ sin θ 
 
 0 0 0 0 0 0 - sin θ cos θ

and A is the bonded surface area tributary to the contact element.


59

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 FRC tension. Thirdly, confinement and lateral expansion

models are described. Fourthly, cracking criterion, crack stress calculations, crack width models, 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
60
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, fci, if the

compressive principal strain, εci, is less compressive than the strain, εp, corresponding to the peak

compressive stress, fp. The peak parameters are determined by adjusting the unconfined uniaxial concrete

cylinder strength, f’c, and the corresponding strain, εo, 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 18, is an elastic-plastic compression response curve, which is

not normally used.

 ε 
−  ci fp for ε p < ε ci < 0
 εp 
f ci =   
(4.1.1.1)
- f for ε ci < ε p < 0
 p

-f c
fp

εp -ε c

Figure 18: Linear pre- and post-peak concrete compression response


61
4.1.2 Popovics

Popovics (1973) presented stress-strain curves for a range of normal strength concretes. As shown

in Figure 19, 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:

ε  n
f ci = − ci fp for ε ci < 0 (4.1.2.1)
εp

 n -1+ ε ε
 ci p( )n

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 Ec, and secant stiffness, Esec. These values are computed as follows:

Ec
n= (4.1.2.2)
E c - E sec

fp
E sec = (4.1.2.3)
εp

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

-ε c

Figure 19 Popovics pre- and post-peak concrete compression response


62
4.1.3 Hognestad (Parabola)

The Hognestad parabola, as shown in Figure 20, is a simple compression response curve, suitable

for normal concrete strengths (<40 MPa).

-f c
fp

εp -ε c

Figure 20: Hognestad parabolic pre- and post-peak concrete compression response

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

 ε   ε ci 
2

 
f ci = − f p 2 ci −  <0 for ε ci < 0 (4.1.3.1)
  εp 
  ε p    

The stress-strain relationship is symmetric about εp, diminishing to zero stress at zero strain and 2εp. Note

that the Hognestad parabola predefines the initial tangent stiffness, Ec, as follows:

Ec = 2 f p ε p (4.1.3.2)

4.1.4 Popovics (High Strength)

Collins and Porasz modified the stress-strain curve proposed by Thorenfeldt, Tomaszewicz, and

Jensen (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
63
– High Strength response curve, as shown in Figure 21, 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:

ε  n
f ci = − ci fp for ε ci < 0
εp

 n -1+ ε ε
 ci p ( )
nk

(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 Ec, and secant stiffness Esec. It is given by the following equation:

fp
n = 0.80 +
17
(f p in MPa ) (4.1.4.2)

The parameter, k, increases the post-peak decay in stress and is calculated as follows:

1.0 for ε p < ε ci < 0



k= f' p (4.1.4.3)
0.67 +
62
≥ 1.0 (
for ε ci < ε p < 0 f p in MPa )

Note that the Popovics – High Strength relationship predefines the initial tangent stiffness, Ec, as follows:

fp n
Ec = ⋅ (4.1.4.4)
εp n -1

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

ε
- c

Figure 21: Popovics high strength pre- and post-peak concrete compression response
64
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 22. Experimental investigations

indicated that the peak compressive stress, fp, and corresponding strain, εp, depend upon the amount of

hoop reinforcement, whereas the initial stiffness, Ec, 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

 n −1

 1  ε ci  
f ci = E c ⋅ ε ci 1 − for ε p < ε ci < 0 (4.1.5.1)
 n  ε p  
 

The term in parentheses represents the deviation from linear-elastic response. The parameter, n, is

expressed in term of the initial tangent stiffness Ec, and secant stiffness Esec, as follows:

Ec
n= (4.1.5.2)
E c - E sec

fp
E sec = (4.1.5.3)
εp

-f c

fp = 70 MPa
fp = 60 MPa
fp = 50 MPa
fp = 40 MPa
fp = 30 MPa
fp = 20 MPa

-ε c

Figure 22: Hoshikuma et al. concrete compressive pre-peak response


65
The Hoshikuma et al. post-peak response is considered in the subsequent sections.

4.1.6 Smith-Young

Based on cylinder stress-strain curves, Smith and Young determined the following equation for

the stress-strain behaviour of concrete (Smith and Young, 1956).

𝜀 𝜀
𝑓𝑐 = 𝑓𝑐′ � 𝑐 � 𝑒𝑥𝑝 �1 − � 𝑐 �� (4.1.6.1)
𝜀𝑜 𝜀𝑜

The Smith-Young pre-peak curve is illustrated in Figure 23.

Figure 23: Smith-Young concrete compressive pre-peak response

4.1.7 Lee et al 2012 (FRC) (Lee et al, 2012)

For FRC in compression, the following stress-strain formulation is available in VecTor2:

𝐴(𝜀 /𝜀0 )
𝑓𝑐 = 𝑓𝑐′ �𝐴−1+(𝜀
𝑐
𝐵� (4.1.7.1)
𝑐 /𝜀0 )

where A and B are the parameters considering the effect of fibres, which are evaluated for pre-and

post-peak compressive behaviours, separately, as follows:


66
1
𝐴=𝐵= ′ 𝐵 𝑓𝑜𝑟 𝜀𝑐 /𝜀0 ≤ 1 (4.1.7.2)
𝑓
1−� 𝑐 �
𝜀0 𝐸𝑐

−0.957
𝑙𝑓
𝐴 = 1 + 0.723 �𝑉𝑓
𝑑𝑓
� 𝑓𝑜𝑟 𝜀𝑐 /𝜀0 > 1 (4.1.7.3)

0.064 −0.882
𝑓′ 𝑙
𝐵 = � 𝑐�
50
�1 + 0.882 �𝑉𝑓 𝑑𝑓 � �≥𝐴 𝑓𝑜𝑟 𝜀𝑐 /𝜀0 > 1 (4.1.7.4)
𝑓

The strain at the compressive strength and elastic modulus of FRC can be calculated as follows:

𝑙 0.12
𝜀0 = �0.0003𝑉𝑓 𝑑𝑓 + 0.0018� 𝑓𝑐′ (4.1.7.5)
𝑓

𝑙 0.41
𝐸𝑐 = �−367𝑉𝑓 𝑑 + 5520� 𝑓𝑐′
𝑓
(4.1.7.6)
𝑓

4.2 Compression Post-Peak Response

Beyond the peak compressive stress, fp, and corresponding strain, εp, 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

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, fci, if the

compressive principal compressive strain, εci, is more compressive than the strain, εp, corresponding to the

peak compressive stress, fp. The peak parameters are determined by adjusting the unconfined uniaxial
67
concrete cylinder strength, f’c, and the corresponding strain, εo, 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, fci,

is computed as follows:

f ci = (1 − c ) f cia + c ⋅ f cib (4.2.1)

where

 f p − f' c 
c = 4 , 0 ≤ c ≤ 1
 (4.2.2)
 f' c 

ε    
f cia = - f p  ci  ⋅ exp 1 -  ε ci  (4.2.3)
εp    ε p 
  

f cib = func(ε ci ) , as computed in the subsequent sections for confined concrete.

The stress computed as f cia is the Smith-Young post-peak branch for unconfined concrete, as shown in

Figure 24. The parameter, c, increases linearly from zero to one as fp 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.
68

-f c
fp

f cia

εp -ε c

Figure 24: Smith-Young concrete compression post-peak response for fp <= 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 0.2fp,

whenever the concrete is sufficiently confined such that fp exceeds f’c.

f ci = func (ε ci ) < - 0.2 f p if 0 < f' c < f p (4.2.4)

4.2.1 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.

4.2.2 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 25, was utilized to compute the flexural strength of reinforced concrete
69
columns confined by transverse hoop reinforcement. The linearly descending branch of the modified

stress-strain curve is liberally adapted for VecTor2 as follows:

[ ]
f cib = − f p + Z m f p (ε ci − ε p ) < 0 or - 0.2 f p for ε ci < ε p < 0 (4.2.2.1)

where

0.5
Zm = 0. 9
( f' c and f lat in MPa ) (4.2.2.2)
3 + 0.29 f ' c ε o   f lat 

⋅ + + εp
145 f ' c − 1000  − 0.002   170 

and flat, is summation of principal stresses, acting transversely to the direction under consideration:

f lat = f c1 + f c 2 + f c 3 - f ci ≤ 0 i = 1 or 2 (4.2.4.3)

-f c
fp

Zm

0.2f' c

εp -ε c

Figure 25: 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, Ec, is assigned a particular value as follows:

ε  n
f cib =  ci fp < 0.2 f p for ε ci < ε p < 0 (4.2.3.1)
εp

 n -1+ ε ε
 ci p ( )n
70
where

Ec
n= (4.2.3.3)
E c - E sec

fp
E sec = (4.2.3.2)
εp

E c = 5000 f' c (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 26, is idealized as a

straight line given as follows:

[ (
f cib = − f p + E des ε ci − ε p ) ] ≤ − 0.2 f p for ε ci < ε p < 0 (4.2.4.1)

where Edes 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 |0.5fp|, Hoshikuma et al.

proposed the following inverse relation between Edes and ρsfyh/f’c2, where ρs is the volumetric ratio of

hoop reinforcement, fyh is the yield strength of hoop reinforcement and ,f’c is the concrete cylinder

strength:

11.2
E des = 2
(4.2.4.2)
ρ s f yh f ' c

In VecTor2, the confining pressure, ρs fyh, provided by the yielding of the hoop reinforcement is replaced

by the mean lateral confining pressure, flat/2, and the deterioration rate is limited to half the initial tangent

modulus as follows:

11.2
E des = ≤ 0.5 E c (4.2.4.3)
( f lat 2) f 'c
2

where flat, is summation of principal stresses, acting transversely to the direction under consideration:
71
f lat = f c1 + f c2 + f c3 - f ci ≤ 0 i = 1 or 2 (4.2.4.4)

-f c
fp

E des

0.2f' c

εp -ε c

Figure 26: 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, εr , and corresponding stress, fr, 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 27, is given as follows:

ε 
K ⋅ f p  ci 
εp 
f ci =   (4.2.5.1)
2 3
ε     
1 + A ci  + B ε ci  + C  ε ci 
εp  εp  εp 
     

The parameters, A, B, C are expressed in terms of the stiffness ratio, strain ratio, and stress ratio,

Kσ, as follows:

A=C + K − 2 (4.2.5.2)

B = 1 − 2C (4.2.5.3)

(K σ − 1) 1
(4.2.5.4)
C=K −
(K ε − 1)2 Kε

where the stiffness ratio, K, compares the initial tangent stiffness, Ec, to the secant stiffness, Esec:
72
Ec
K= (4.2.5.5)
E sec

fp
E sec = (4.2.5.6)
εp

The strain ratio, Kε, compares the post-peak control point strain, εr, and the strain corresponding to the

peak stress, εp:

εr
Kε = (4.2.5.7)
εp

and the stress ratio, Kσ, compares the peak compressive stress, fp, to the post-peak control point stress, fr:

fp
Kσ = (4.2.5.8)
fr

It remains necessary to specify the post-peak control point strain, εr, and corresponding stress, fr.

Kwan and Spacone (2002) utilize the following empirical equation to determine the stress fr, which

increases as the confinement effect increases:

f p f 'c
fr = ≤ 1.4 f p (4.2.5.9)
5 f 'c − f p

The strain, εr, is computed as four times the strain corresponding to peak compressive stress:

ε r = 4ε p (4.2.5.10)

-f c
fp

0.2f' c

εp -ε c
Figure 27: Saenz/Spacone concrete compressive post-peak response
73
4.2.6 Montoya 2003

Montoya proposed a post-peak model for steel and FRC confined concrete. This model includes

a stress-strain curve which accounts for three-dimensional effects, concrete dilatation, strength

enhancement, and post-peak softening or increased strain hardening. In formulating this model, varying

concrete strengths, 20MPa to 100MPa, as well as confining pressure ratios, 0 to 100%, were studied.

Note that the dilatation component of the model is discussed in a subsequent section of this user manual.

For post-peak compression, the following model is proposed (Montoya, 2004).

𝑓𝑝𝑖
𝑓𝑐𝑖 = 2 (4.2.6.1)
𝐴�𝜀𝑐𝑖 /𝑓𝑝𝑖 � −𝐵�𝜀𝑐𝑖 /𝑓𝑝𝑖 �+𝐶+1.0

where

𝐴 = 𝑘𝑑 (4.2.6.2)

𝐴
𝐵=2 (4.2.6.3)
𝐸𝑠𝑒𝑐

𝐴
𝐶= 2 (4.2.6.4)
𝐸𝑠𝑒𝑐

𝑓𝑝𝑖
𝐸𝑠𝑒𝑐 = 𝜀𝑝𝑖
(4.2.6.5)

2
1 𝑓𝑝𝑖
𝑘𝑑 = � � (4.2.6.6)
4 𝜀𝑐80𝑖 −𝜀𝑝𝑖

𝜀𝑐80𝑖 𝑓
𝜀𝑐𝑜
= 1.5 + (89.5 − 0.60𝑓𝑐′ ) 𝑓𝑐1′ (4.2.6.7)
𝑐

And 𝑘𝑑 is the shape factor which is a function of the post-peak curve’s steepness, 𝜀𝑐80𝑖 is the

post-peak strain at 80% of the peak stress, 𝐸𝑠𝑒𝑐 is the secant Young’s modulus. The compressive stress-

strain curve is shown in Figure 28 below (Montoya, 2004).

Figure 28: Montoya concrete compressive post-peak response


74
4.2.7 Lee et al 2012 (FRC)

The formulations implemented in VecTor2 for compressive behavior of FRC are outlined in the Lee et al

2011 (FRC) section 4.1.7.

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, βd, 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 βd, 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 29, use βd to reduce both the uniaxial

compressive strength, f’c, and corresponding strain, ε0, to determine the peak compressive strength, fp, and

corresponding strain, εp, used in the compression response models. The value of βd is a function of εc1//εc2

– 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:

ε 
β d = func c1  (4.3.1)
 ε c2 

f p = β d β l f' c (4.3.2)

ε p = β d βl ε o (4.3.3a)

(The parameter, βl, accounts for strength enhancement from confinement effects.)
75
However, if the compression softening is significant enough that ε p = β d β l ε o is less compressive than

εo, then a modification must be made to the calculation of εp so that the compression response ascends for

strains up to εo and descends thereafter. When the compressive strain, εc2, is more compressive than

β d βl ε o , εp, is calculated as follows:

ε if ε o < ε c 2 < β d β l ε o < 0


ε p =  c2 (4.3.3b)
ε o if ε c 2 < ε o < β d β l ε o < 0

Further, for strains more compressive than εo, the softened response is computed by simply multiplying

the base response by βd.

-f c
f' c

βd f cbase

fp

εp εo -ε c

Figure 29: Strength and strain-softened compression response

Strength-only softened models, as shown in Figure 30, use βd to reduce only the uniaxial

compressive strength, f’c, to determine the peak compressive strength, fp, while the corresponding

strain, ε0,, is unmodified for compression softening. The value of βd is a function of εc1//ε0 – the ratio of

the principal tensile strain to the strain corresponding to the uniaxial compressive stress. As εc2 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.

ε 
β d = func c1  (4.3.4)
 εo 

f p = β d β l f' c (4.3.5)
76
ε p = βl ε o (4.3.6)

-f c

f' c

fp

εo -ε c

Figure 30: Strength-only softened compression response

It is assumed in the following discussion that the principal compressive strain εc2 and the strain

corresponding to peak stress, εo, are negative values while the principal tensile strain, εc1, is a positive

value.

4.3.1 No compression softening

The compression response is independent of the coexisting tensile strains:

βd = 1 (4.3.1.1)

4.3.2 Vecchio 1992-A (e1/e2-Form)

The strength-and-strained softened model, shown in Figure 31, 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 Cs 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.
77

1
βd = ≤1 (4.3.2.1)
1 + Cs ⋅ Cd

0 if r < 0.28
Cd =  (4.3.2.2)
0.35(r − 0.28)
0.80
if r > 0.28

− ε c1
r= ≤ 400 (4.3.2.3)
ε c2

0 if shear slip not considered


Cs =  (4.3.2.4)
0.55 if shear slip considered

f p = β d f' c (4.3.2.5)

ε p = βd εo (4.3.2.6)

1
βd

0.8

0.6

0.4

0.2 Cs=1.0
Cs=0.55
0 -ε c1/ε c2
0 5 10 15

Figure 31: 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 Cs

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
βd = ≤1 (4.3.3.1)
1 + Cs ⋅ Cd

 − ε c1 
C d = 0.27 − 0.37 , 0 ≤ C d ≤ 50 (4.3.3.2)
 εo 
78

0 if shear slip not considered


Cs =  (4.3.3.3)
0.55 if shear slip considered

f p = β d f' c (4.3.3.4)

ε p = εo (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.

1
βd = ≤1 (4.3.4.1)
ε 
0.85 − 0.27 c1 
 ε c2 

f p = β d f' c (4.3.4.2)

ε p = βd εo (4.3.4.3)

1
βd

0.8

0.6

0.4

0.2

0 -ε c1/ε o
0 5 10 15

Figure 32: Vecchio 1982 compression softening model


79
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
βd = ≤1 (4.3.5.1)
ε 
0.8 − 0.34 c1 
 εo 

f p = β d f' c (4.3.5.2)

ε p = εo (4.3.5.3)

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:

f c1 = E c ε c1 for 0 < ε c1 < ε cr (4.4.1)

where,

f cr
ε cr = (4.4.2)
Ec

εcr is the cracking strain , Ec is the initial tangent stiffness of concrete, εc1 is the principal tensile strain,

and fcr 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 fcr (or else additional cracking ensues), they act over a relatively large tributary
80
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

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 f ca1 , to distinguish it from the average concrete tensile stress due to

tension softening effects, denoted by f cb1 . The magnitude, f ca1 , 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
f c1 = max f c1 , f c1
b
) (4.4.3)

4.4.1 No Tension Stiffening

The tension stiffening effect is ignored and the post-cracking concrete tensile stress is zero.

f ca1 = 0 for 0 < ε cr < ε c1 (4.4.1.1)


81
4.4.2 Modified Bentz 2003

The Bentz 2003 formulation for tension stiffening incorporates the percentage of reinforcement as

well as bond characteristics; it is the default model in VecTor2. The formulation is as follows:

𝑓𝑡′
𝑓𝑐1 = 1+ 𝑓𝑜𝑟 𝜀𝑐1 > 𝜀𝑡′ (4.4.2.1)
�𝑐𝑡 𝜀𝑐1

where

𝑐𝑡 = 3.6 𝑡𝑑 ∙ 𝑚 (4.4.2.2)

𝑡𝑑 = 0.6 (4.4.2.3)

1/𝑚 = ∑𝑛𝑖=1 4𝜌𝑖 /𝑑𝑏𝑖 ∙ |cos(𝜃 − 𝛼𝑖 )| (4.4.2.4)

And 𝜌𝑖 is the reinforcement ratio, 𝑑𝑏𝑖 is the rebar diameter, 𝜃 is the inclination of the principle

direction, and 𝛼𝑖 is the inclination of reinforcement.

This formulation is modified when FRC is used (Lee et al. 2012). The following proposed

tension stiffening model is incorporated into VecTor2:

𝑓𝑡
𝑓𝑐1 = 1+ (4.4.2.5)
�𝑐𝑓 ∙3.6∙𝑀𝜀1

where
1.5
1 𝑙 �100𝑉 �
𝑐𝑓 = 0.6 + � 𝑓 � 𝑀0.8𝑓 𝑓𝑜𝑟 ℎ𝑜𝑜𝑘𝑒𝑑 − 𝑒𝑛𝑑 𝑓𝑖𝑏𝑟𝑒𝑠 (4.4.2.6)
0.034 𝑑𝑓

0.9
1 𝑙 𝑉𝑓
𝑐𝑓 = 0.6 + � 𝑓� 𝑀 0.8
𝑓𝑜𝑟 𝑠𝑡𝑟𝑎𝑖𝑔ℎ𝑡 𝑓𝑖𝑏𝑟𝑒𝑠 (4.4.2.7)
0.058 𝑑𝑓

4.4.3 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 33, is determined as:

f cr
f ca1 = for 0 < ε cr < ε c1 (4.4.3.1)
1 + 200ε c1
82

fc

f cr
ct = 200

ct = 500

ε cr εc

Figure 33: 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 33, is curve is determined as:

f cr
f ca1 = for 0 < ε cr < ε c1 (4.4.4.1)
1 + 500ε c1

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:

f cr
f ca1 = for 0 < ε cr < ε c1 (4.4.5.1)
1 + 3.6m ⋅ ε c1
83
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 34, is determined as:

 f cr for 0 < ε cr < ε c1 < 2ε cr


 c
f ca1 =   2ε cr  (4.4.6.1)
 f cr  ε  for 0 < 2ε cr < ε c1
  c1 

The parameter, c, reflects the bond characteristics. In VecTor2, the model is implemented with c = 0.4 ,

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.

fc
f cr

ε cr εc

Figure 34: Izumo, Maekawa et al. tensions stiffening response


84
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 (2003)

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 35. 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:

m 
 ρ e ,i cos θ s ,i  cos θ s ,i  n

f ca1 = s r τ bo ∑ 
i =1  d b ,i
⋅ min1,
 sr ε s
 ∑
+ ρ F , j E F , j ∆ε F cos 2 θ F , j
 j =1
(4.4.7.1)
  

The first term in the above equation represents the contribution of the steel reinforcement to tension

stiffening. The value sr is the crack spacing, τbo is the maximum average bond stress of the steel

reinforcement, and εs is the average reinforcement strain. For the ith steel reinforcement, ρe,i is the

effective reinforcement ratio in its tributary area, db,i is its diameter, and θs,i is the angle between principal
85
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 ∆εF is the

average difference between the average FRP strains and the local FRP strains at a crack. For the jth FRP

reinforcement, ρF,j is the effective reinforcement ratio in its tributary area, EF,j is the elastic modulus of

elasticity, ∆θF,j is the angle between principal concrete tensile stress direction and the axis of the FRP

reinforcement.

sr

fFcr fFcr

Figure 35: Tension chord model

4.4.8 Lee 2011 (w/ Post Yield) (Lee et al. 2011)

This model is based on the comparison of parametric studies with experimental results for

uniaxial members, and includes a proposed model for the behaviour of reinforced concrete after yielding

of reinforcement. The model allows VecTor2 to calculate the reinforcement stresses at crack locations,

and makes it possible to determine the average strain conditions that will result in the rupture of

reinforcement. Thus, the ductility of reinforced concrete members can be more accurately predicted.

Also, improved estimates of strength and deformation capacity for concrete members with nonductile

reinforcement can be made with this model.

Figure 36 shows the variation of the average tensile stress in concrete up to failure.
86

Figure 36: Lee 2010 stress-strain response

The proposed tension stiffening model for the stress-strain relationship of concrete after yielding

of the reinforcement is:


2
𝜀𝑡,𝑝𝑒𝑎𝑘 −𝜀𝑡,𝑎𝑣𝑔
𝑓𝑐𝑡,𝑝𝑒𝑎𝑘 − 𝑓𝑐𝑡,𝑝𝑒𝑎𝑘 � � 𝑓𝑜𝑟 𝜀𝑠𝑦 ≤ 𝜀𝑡,𝑎𝑣𝑔 ≤ 𝜀𝑡,𝑝𝑒𝑎𝑘
𝜀𝑡,𝑝𝑒𝑎𝑘 −𝜀𝑠𝑦
𝑓𝑐𝑡,𝑎𝑣𝑔 = � 𝑓𝑐𝑡,𝑝𝑒𝑎𝑘 −0.5𝑓𝑐𝑡,𝑝𝑒𝑎𝑘,𝜌
(4.4.8.1)
𝑓𝑐𝑡,𝑝𝑒𝑎𝑘 −
0.1−𝜀𝑡,𝑝𝑒𝑎𝑘
𝑚𝑖𝑛
�𝜀𝑡,𝑎𝑣𝑔 − 𝜀𝑡,𝑝𝑒𝑎𝑘 � ≥ 0.5𝑓𝑐𝑡,𝑝𝑒𝑎𝑘,𝜌 𝑓𝑜𝑟 𝜀𝑡,𝑎𝑣𝑔 ≥ 𝜀𝑡,𝑝𝑒𝑎𝑘
𝑚𝑖𝑛

where

𝜀𝑐𝑟 ∙𝐸𝑐
𝜌𝑚𝑖𝑛 = 𝑓 (4.4.8.2)
𝑠𝑦 −𝜀𝑐𝑟 ∙𝐸𝑠

𝜀𝑡,𝑝𝑒𝑎𝑘 = 0.01 + 0.001 ∙ 𝑚𝑎𝑥(15 − 𝑑𝑏 , 0) ≥ 𝜀𝑠ℎ (4.4.8.3)

𝑓𝑐𝑡,𝑝𝑒𝑎𝑘 = 𝑎�𝑓𝑐′ (4.4.8.4)

𝑎 = −0.0313𝜌𝑠0.57 𝑑𝑏 + 3.3881𝜌𝑠0.76 (4.4.8.5)

In this proposed model, the steel stress at the location of a crack is calculated from force

equilibrium as follows:

𝑓𝑠𝑐𝑟 = 𝑓𝑠,𝑎𝑣𝑔 + 𝑓𝑐𝑡,𝑎𝑣𝑔 /𝜌𝑠 (4.4.8.6)


87
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.

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 f cb1 , to distinguish it from the average concrete tensile stress due to

tension stiffening effects, denoted by f ca1 . The larger of the two tensile stresses is assumed to be the

average post-cracking concrete tensile stress:

( a
f c1 = max f c1 , f c1
b
) (4.5.1)

It is convenient to define common parameters in the tension-stiffening models. The fracture

energy, Gf, 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
88
fracture energy is independent of element size, and is assigned a value of 75N/m in VecTor2. The

representative length, Lr, 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, εch, of the tension softening curve is

determined as follows:

2G f
ε ch = , 1.1ε cr < ε ch < 10ε cr (4.5.2)
Lr ⋅ f cr

The corresponding tensile stress, fch, is the characteristic stress of the tension softening curve. The

terminal strain, εte, is the strain at which the tension softening stress diminishes to zero, determined as:

ε te = 5 ε ch (4.5.3)

In models, without residual tension, the tensile stress due to tension softening is equal to the

tensile stress, fts,base, computed from the tension softening base curve.
b
f c1 = f ts ,base (4.5.4)

In models with a sustained residual tension, fres, the residual stress is calculated as follows:

0.1 f cr for 0 < w < 5mm


f res =  (4.5.5)
0.1 f cr (3- w 2.5) ≥ 0 for 5mm < w

For these models, the tensile stress due to tension softening is taken as the maximum of the tensile stress,

and the residual stress, fres as follows:

f c1 = max( f ts ,base , f res ) (4.5.6)


b

Residual tension should not be utilized in structures predominantly subject to tension stresses.

4.5.1 Not Considered

Post-cracking tensile stresses due to tension stiffening are not considered:

b
f c1 = 0 (4.5.1.1)
89
4.5.2 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 37 with and without the residual branch, is given as follows:

 (ε − ε cr ) 
f ts, base = f cr 1 − c1  ≥ 0, for ε cr < ε c1 (4.5.2.1)
 (ε ch − ε cr ) 

fc
f cr

w/ residual

ε cr ε te =ε ch εc

Figure 37: Linear tension softening response with and without residual

4.5.3 Bilinear

There are numerous bilinear tension softening models that have been proposed by different

researchers. The CEB-FIP model is discussed in this manual.

The 1990 CEB-FIP model code provides a bilinear model for the stress-strain relationship of

concrete. For a cracked concrete section, Figure 38 summarizes the stress-strain relationship:
90

Figure 38: CEB-FIP tension softening curve (CEB-FIP, 1990)

The equations that summarize the CEB-FIP stress-strain relationship are (CEB-FIP, 1990):

𝑤
𝑓𝑐𝑡𝑚 �1 − 0.85 � 𝑓𝑜𝑟 0.15𝑓𝑐𝑡𝑚 ≤ 𝜎𝑐𝑡 ≤ 𝑓𝑐𝑡𝑚
𝑤1
𝜎𝑐𝑡 = � 0.15𝑓𝑐𝑡𝑚
(4.5.3.1)
(𝑤𝑐 − 𝑤) 𝑓𝑜𝑟 0 ≤ 𝜎𝑐𝑡 < 0.15𝑓𝑐𝑡𝑚
𝑤𝑐 −𝑤1

𝐺𝑓
𝑤1 = 2 𝑓 − 0.15𝑤𝑐 (4.5.3.2)
𝑐𝑡𝑚

𝐺𝑓
𝑤𝑐 = 𝛼𝐹 (4.5.3.3)
𝑓𝑐𝑡𝑚

where 𝑤 is the crack opening width (mm), 𝑤1 is the crack opening (mm) for 𝜎𝑐𝑘 = 0.15𝑓𝑐𝑡𝑚 , 𝑤𝑐

is the crack opening for 𝜎 = 0, 𝑓𝑐𝑡𝑚 is the tensile strength (MPa), and 𝛼𝐹 is a coefficient related to

maximum aggregate size.

Equations for fracture energy, 𝐺𝑓 , and tensile strength, 𝑓𝑐𝑡𝑚 , are also found in the CEB-FIP

design guide.

The bilinear tension softening model implemented in VecTor2, adapted from the above

formulations, is shown below in Figure 39:

𝜀 −𝜀
⎧ 𝑓𝑐𝑟 �1 − 0.80 �𝜀 𝑐1 −𝜀𝑐𝑟 �� 𝑓𝑜𝑟 𝜀𝑐1 < 𝜀𝑐ℎ3
⎪ 𝑐ℎ3 𝑐𝑟

𝑓𝑐1 = 𝜀 −𝜀 (4.5.3.4)
⎨0.2𝑓𝑐𝑟 �1 − � 𝑐1 𝑐ℎ3 �� 𝑓𝑜𝑟 𝜀𝑐ℎ3 ≤ 𝜀𝑐1 < 𝜀𝑐ℎ4
𝜀𝑐ℎ4 −𝜀𝑐ℎ3

⎩ 0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

where

𝜀𝑐ℎ3 = 0.64𝜀𝑐ℎ + 𝜀𝑐𝑟 (4.5.3.5)


91
εch4 = 6.8εch + εcr (4.5.3.6)

𝐺𝑓
𝜀𝑐ℎ = 𝐿 ; characteristic strain using Gf from (Bazant, 2002) (4.5.3.10)
𝑟𝑒𝑓 ∙𝑓𝑐𝑟

Figure 39: Bilinear Tension softening model implemented in VecTor2

4.5.4 Nonlinear (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 40 with and without the residual branch, is given as follows:

 f cr
 for ε cr < ε c1 < ε ch
1 + c(ε c1 − ε cr )
f ts, base = (4.5.3.1)
 f (ε te − ε c1 ) ≥ 0 for ε ch < ε c1


ch
(ε te − ε ch )
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:

Gf ∞

Lr
= ∫f
0
ts ,base dε c1 (4.5.3.2)
92

fc
f cr

w/ residual

ε cr ε ch ε te ε c

Figure 40: Yamamoto tension softening response with and without residual

4.5.5 Nonlinear (Hordijk)

Hordijk et al. carried out experiments on the crack softening behaviour of normal-weight and

light-weight concrete, resulting in nonlinear curves for the post-peak tensile stress-strain behaviour for

both concrete types. Both deformation controlled tensile tests and cyclic tests were performed. From

their experiments, Hordijk obtained stress-deformation envelope curves.

Figure 41: Envelope curves for normal and light-weight concrete (Cornelissen et al, 1986)
93
Stress-crack opening relationships were developed from the envelope curves shown in Figure 41

above, and a mathematical model was derived.

This mathematical model for both concrete types is shown below (Cornelissen et al, 1986):

𝜎/𝑓𝑡 = 𝑓(𝛿) − (𝛿/𝛿0 )𝑓(𝛿 = 𝛿0 ) (4.5.5.1)

𝑓(𝛿) = (1 + (𝐶1 𝛿/𝛿0 )3 ) 𝑒𝑥𝑝(−𝐶2 𝛿/𝛿0 ) (4.5.5.2)

Where δ0 is the crack opening at which stress cannot be transferred. The values for δ0 and the

two constants are shown in Figure 42.

Figure 42: Stress-crack opening relationships (Cornelissen et al., 1986)

The above model has been adapted for VecTor2 and is implemented using the following

formulation:

𝑤 3 𝑤 𝑤
𝑓𝑐1 = 𝑓𝑐𝑟 ��1 + �𝐶1 𝑤𝑐𝑟𝑥 � � 𝑒𝑥𝑝 �−𝐶2 𝑤𝑐𝑟𝑥 � − 𝑤𝑐𝑟𝑥 (1 + 𝐶13 ) 𝑒𝑥𝑝(−𝐶2 )� (4.5.5.3)
𝑢𝑙𝑡 𝑢𝑙𝑡 𝑢𝑙𝑡

where

𝐶1 = 3 (4.5.5.4)

𝐶2 = 6.93 (4.5.5.5)

5.136𝐺𝐹
𝑤𝑢𝑙𝑡 = 𝑓𝑐𝑟
(4.5.5.6)
94
4.5.6 Exponential

This tension softening formulation is recommended for FRC members, and is implemented in VecTor2 as

follows:

𝑓𝑐𝑟 𝑤𝑐𝑟
𝑓𝑐,𝑡 = 𝑓𝑐𝑟 exp �− 𝐺𝑓
� (4.5.6.1)

where Gf is the concrete fracture energy.

4.5.7 Custom Input-Strain Based

A custom tension softening curve can be input by entering four stress-strain values in the Auxiliary

Tab of the Model Page. This strain-based custom tension softening model is based on tension prism or

dog-bone responses (direct tension test results) and requires the manual input of fc1 − εc1 data. Based on

the input stress-strain values, VecTor2 constructs a tension softening response; for strain values between

the specified data points, the program calculates the stress through linear interpolation.

Figure 43 illustrates the custom input-strain based curve.

Figure 43: Custom input-strain based tension softening in VecTor2


95
4.5.8 Custom Input-Crack Based

This type of custom tension softening is also based on dog-bone responses and requires the

manual input of fc1 − w data.

For the crack based tension softening response, the user still inputs the tension softening data as

fc1 − εc1 data points, and VecTor2 converts them to fc1 − w data as follows:
𝜀𝑐1
𝑤 = 1000 ∙ (4.5.8.1)
𝑠𝑐𝑟1

Similar to the strain-based input, stress is calculated through linear interpolation between the

input data points.

4.5.9 FRC (fib Model Code 2010)

The fib Model Code formulation for FRC in tension is based on the bending test results of a

rectangular beam with a notch. The load-displacement response of the flexural member is used for an

evaluation of the tensile stress-crack mouth opening displacement (CMOD), which is then employed to

derive a simple linear post-crack constitutive law for the tensile stress-crack width response of FRC in

tension.

From the two residual flexural strengths, fR1 and fR3, at the CMODs of 0.5 and 2.5 mm, respectively, the

tensile stresses at a crack width, wcr, can be evaluated as follows:

𝑓𝐹𝑡𝑠 = 0.45 𝑓𝑅1 (4.5.9.1)


𝑤
𝑓𝐹𝑡𝑢 = 𝑓𝐹𝑡𝑠 − 𝑐𝑟 (𝑓𝐹𝑡𝑠 − 0.5𝑓𝑅3 + 0.2𝑓𝑅1 ) ≥ 0
𝐶𝑀𝑂𝐷
(4.5.9.2)
3

where, fFts is the residual tensile stress at an initial crack opening. The figures below illustrate the

relationship between the stress-CMOD response measured from the bending test and the stress-crack

width response of FRC in tension.


96

Figure 44 tension-crack opening relationship in fib model code 2010 (fib, 2010)

4.6 FRC Tension

A number of models are available in VecTor2 specifically for the consideration of the contribution of

steel fibre reinforcement to concrete tensile strength. This section describes the available formulations.

4.6.1 Not Considered

This option should be used for regular reinforced concrete, where no steel fibres are added.

4.6.2 SDEM-Monotonic (Lee et al 2012)

The Simplified Diverse Embedment Model (SDEM) for monotonic loading conditions expresses the

tensile stress in the fibres through the frictional bond behavior in straight fibres and the additional

mechanical anchorage in hooked-end fibres.

For both hooked and straight fibres the tensile stress attained by the frictional bond behavior of fibres is:
2
𝑙𝑓 2𝑤𝑐𝑟
𝑓𝑠𝑡 = 𝛼𝑓 𝑉𝑓 𝐾𝑠𝑡 𝜏𝑓,𝑚𝑎𝑥 �1 − � (4.6.2.1)
𝑑𝑓 𝑙𝑓
97
where

𝛽𝑓 𝑤𝑐𝑟
3 𝑠𝑓
𝑓𝑜𝑟 𝑤𝑐𝑟 ≤ 𝑠𝑓
𝐾𝑠𝑡 = � 𝑠𝑓 𝛽𝑓 𝑠𝑓
(4.6.2.2)
1 − �𝑤 + 3
�𝑤 𝑓𝑜𝑟 𝑤𝑐𝑟 > 𝑠𝑓
𝑐𝑟 𝑐𝑟

𝜏𝑓,𝑚𝑎𝑥 = 0.396�𝑓𝑐′ (4.6.2.3)

𝛽𝑓 = 0.6

and 𝛼𝑓 is the fibre orientation factor (=0.5 for a 3D infinite element), 𝑉𝑓 is the fibre volumetric

ratio, 𝐾𝑠𝑡 is the bond modulus, 𝑙𝑓 is the fibre length.

The value of 𝛽𝑓 = 0.6 is used in order to prevent the tensile stress due to fibres being overestimated,

which correlates well with data from the Diverse Embedment Model.

For end-hooked fibres, additional tensile stress is attained from the mechanical anchorage of the

hooked ends. The tensile stress attained through this anchorage is calculated in VecTor2 using the

following formulation:

2(𝑙𝑖 −2𝑤𝑐𝑟 )
𝑓𝑒ℎ = 𝛼𝑓 𝑉𝑓 𝐾𝑒ℎ 𝜏𝑒ℎ,𝑚𝑎𝑥 𝑑𝑓
(4.6.2.4)

where

2 𝑤𝑐𝑟 1 𝑤𝑐𝑟 2
⎧ 𝛽𝑒ℎ � − � � � 𝑓𝑜𝑟 𝑤𝑐𝑟 ≤ 𝑠𝑒ℎ
3 𝑠𝑒ℎ 5 𝑠𝑒ℎ
𝐾𝑒ℎ = 2 (4.6.2.5)
⎨1 + �7𝛽𝑒ℎ − 1� � 𝑠𝑒ℎ − 2��𝑤𝑐𝑟 −�𝑠𝑒ℎ � 𝑙𝑓 −𝑙𝑖
𝑓𝑜𝑟 𝑠𝑒ℎ < 𝑤𝑐𝑟 ≤
⎩ 15 𝑤𝑐𝑟 𝑙𝑓 −𝑙𝑖 2

4𝑃𝑒ℎ,𝑚𝑎𝑥
𝜏𝑒ℎ,𝑚𝑎𝑥 =
𝜋𝑑𝑓2
= 0.429�𝑓𝑐′ (4.6.2.6)

𝛽𝑒ℎ = 0.8 (4.6.2.7)

The value of 𝛽𝑒ℎ = 0.8 is used in order to prevent the tensile stress due to fibres being overestimated,

which correlates well with data from the Diverse Embedment Model.

Thus the tensile stress attained by steel fibres according to SDEM-Monotonic is:

𝑓𝑓 = 𝑓𝑠𝑡 𝑓𝑜𝑟 𝑠𝑡𝑟𝑎𝑖𝑔ℎ𝑡 𝑓𝑖𝑏𝑟𝑒𝑠 (4.6.2.8)

𝑓𝑓 = 𝑓𝑠𝑡 + 𝑓𝑒ℎ 𝑓𝑜𝑟 ℎ𝑜𝑜𝑘𝑒𝑑 𝑒𝑛𝑑 𝑓𝑖𝑏𝑟𝑒𝑠 (4.6.2.9)


98
4.6.3 SDEM-Cyclic

For the cyclic behavior of fibre-reinforced concrete, the SDEM-Cyclic model calculates the

unloading and reloading stiffness for the tensile stress attained by the fibres. For the envelope tensile

stress-crack width response of fibre-reinforced concrete, the formulations for the stress attained by the

fibres are the same as SDEM-Monotonic. The formulations for the stiffness of the unloading and

reloading paths are the following:

𝐸𝑢𝑛 = 𝐸𝑠𝑡 + 𝐸𝑒ℎ (4.6.3.1)

𝑙 𝛽 2𝑤𝑐𝑟,𝑒𝑥𝑝
𝐸𝑠𝑡 = 𝛼𝑓 𝑉𝑓 𝜏𝑓,𝑚𝑎𝑥 𝑓 𝑓 �1 − � (4.6.3.2)
𝑑 3𝑠 𝑓 𝑓 𝑙𝑓

𝑙 4𝛽𝑒ℎ 𝑙𝑖 −2𝑤𝑐𝑟,𝑒𝑥𝑝
𝐸𝑒ℎ = 𝛼𝑓 𝑉𝑓 𝜏𝑒ℎ,𝑚𝑎𝑥 𝑖
𝑑
(4.6.3.3)
𝑓 3𝑠𝑒ℎ 𝑙𝑓

In the above formulations, 𝐸𝑠𝑡 is the stiffness due to the frictional bond behavior of the steel

fibres, 𝐸𝑒ℎ is the stiffness due to mechanical anchorage, and 𝑤𝑐𝑟,𝑒𝑥𝑝 is the maximum experienced crack

width.

Figure 45: Stiffness for the re- and unloading response of FRC in cyclic tension
99
4.6.4 Diverse Embedment Model (DEM, Lee et al 2011)

The Diverse Embedment Model (DEM) considers various mechanisms involved in the tensile

behaviour of SFRC. The DEM considers the pullout behaviour of individual fibres, the random

distribution of fibres, and the effects of finite member dimensions. The fibre tensile stress at a crack can

be calculated from the analysis for the pullout behaviour of a fibre embedded on both sides; both the

frictional bond behaviour of a straight fibre and the mechanical anchorage effect of an end-hooked fibre

are accounted for.

The fibre tensile stress at a crack is affected by the fibre inclination angle and fibre embedment

length, thus the average fibre tensile stress at a crack, for a three-dimensional infinite element, can be

expressed by the following equation:

2 𝑙 /2 𝜋/2
𝜎𝑓,𝑐𝑟,𝑎𝑣𝑔 = ∫0 𝑓 ∫0 𝜎𝑓,𝑐𝑟 (𝑙𝑎 , 𝜃) sin 𝜃 𝑑𝜃𝑑𝑙𝑎 (4.6.4.1)
𝑙𝑓

where 𝑙𝑓 is the fibre length, 𝑙𝑎 is the fibre embedment length for the shorter embedded part across a crack,

𝜎𝑓,𝑐𝑟 is the fibre stress at a crack with a given fibre inclination angle and embedment length, and 𝜃 is the

fibre inclination angle from the axis perpendicular to the crack surface. In this formulation sin 𝜃 is the

probability function for the fibre inclination angle. From the average fibre stress at a crack, the tensile

stress attained by the fibres can be calculated by accounting for the fibre orientation factor and volumetric

ratio:

𝑓𝑓 = 𝛼𝑓 𝑉𝑓 𝜎𝑓,𝑐𝑟,𝑎𝑣𝑔 (4.6.4.2)

where 𝛼𝑓 is the fibre orientation factor and 𝑉𝑓 is the volumetric ratio. The fibre orientation factor varies

in regions where the distance to a boundary surface is less than the fibre length; moving inward away

from all boundaries, the factor converges to 0.5. In VecTor2, the fibre orientation factor is calculated

using the following expression, where ℎ is the member thickness:

2.8
ℎ ℎ
⎧−0.05 � � + 0.64 𝑓𝑜𝑟 ≤1
𝑙𝑓 𝑙𝑓
𝛼𝑓 = (4.6.4.3)
⎨ 0.087 �𝑙𝑓�1.12 + 0.5 𝑓𝑜𝑟

>1
⎩ ℎ 𝑙𝑓
100
4.6.5 Variable Engagement Model (VEM: Voo & Foster 2003)

The Variable Engagement Model (VEM) expresses the force in a single steel fibre based on the

calculation of the crack opening displacement at which the fibres become actively engaged; for

mechanically anchored fibres, once the adhesive between the concrete and fibre is broken, slip must occur

before anchorage is engaged.

In VEM, the tensile stress attained by fibres can be evaluated as:

𝑑𝑓
𝜎 = 𝐾𝑓 𝐾𝑑 𝑉𝑓 𝜏𝑏 (4.6.5.1)
𝑙𝑓

Where 𝜏𝑏 is the bond strength of a fibre, Kd is the damage factor or fibre efficiency factor, which

accounts for the effect that the pullout of adjacent fibres has on the bond efficiency in a particular fibre;

the damage factor decreases as the volume of fibres increases, but generally taken as one. Kf is the factor

to account for fibre orientation and engagement. Where fibre fracture is not possible, Kf is calculated

using the following equation:


2
tan−1 (𝑤/𝛼) 2𝑤
𝐾𝑓 = 𝜋
�1 − � (4.6.5.2)
𝑙𝑓

Where fibre fracture is possible, numerical integration is used to calculate Kf:

2 1 𝜃 𝑙 2𝑤
𝐾𝑓 = � ∙ ∫0 𝑐𝑟𝑖𝑡 ∫𝑤𝑎,𝑐𝑟𝑖𝑡 𝑘(𝑙𝑎 , 𝜃)𝑑𝑙𝑎 𝑑𝜃 � ∙ �1 − � ; (4.6.5.3)
𝜋 𝑙𝑓 /2−𝑤 𝑙𝑓

𝑑 𝜎𝑓𝑢
𝑙𝑎,𝑐𝑟𝑖𝑡 = min � 4𝑓 𝜏𝑏
+ 𝑤𝑒 , 𝑙𝑓 /2� (4.6.5.4)

𝜃𝑐𝑟𝑖𝑡 = tan−1 (𝑤/𝛼) (4.6.5.5)

where 𝜃 is a fibre inclination angle from the normal direction to crack surface, and la,crit is the critical fibre

embedment length for fibre fracture. When fibre fracture is not possible,the critical embedment length for

fracture is equal to lf/2 and equation (4.6.5.3) reduces to equation (4.6.5.2). In equation (4.6.5.2) and

equation (4.6.5.5), 𝛼 is an engagement factor which is usually chosen as 𝑑𝑓 /3.5.


101
4.6.6 fib Model Code 2010

FRC in tension, according to fib Model Code 2010, has been described in Section 4.5.9 of this manual.

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 Poisson’s ratios are computed in the following manner regardless of

the selected lateral expansion model. If the concrete is uncracked, then v12 and v21 are equal to the initial

Poisson’s ratio, vo. If the concrete is cracked, then the Poisson’s ratio decreases linearly from vo at the

cracking strain, εcr, to zero at two times the cracking strain.

v o for 0 < ε c1 < ε cr




v12 = v 21 =  ε  (4.7.1)
v o 1 − c1  ≥ 0 for ε cr < ε c1

  2ε cr 

The following lateral expansion models pertain to the determination of the Poisson’s ratio, vij,

relating the expansion of concrete in the principal i-direction due to compressive straining, εcj, in the

principal j-direction.

4.7.1 Constant Poisson's ratio

The Poisson’s ratios vij are always equal to the initial value, vo.

vij = v o for ε cj < 0 (4.7.1.1)


102
4.7.2 Variable Poisson's Ratio – Kupfer

Based on strain data of Kupfer et al. (1969), the Poisson’s ratio vij increases nonlinearly as

compressive strain increases. The Poisson’s ratio, shown in Figure 46, is computed as follows:

v o for - 0.5ε p < ε cj < 0



   
2
vij =    − 2ε cj (4.7.2.1)
v 1 + 1 . 5 − 1   ≤ 0.5 for ε cj < −0.5ε p
 o  εp  

    

where εp is the strain corresponding to the peak compressive stress.

v ij

0.5

vo

ε cji
0.5ε p

Figure 46: Kupfer variable Poisson’s ratio model

4.7.3 Variable Poisson’s Ratio – Montoya

This concrete dilation model from Montoya was proposed in conjunction with the concrete

compression post-peak stress-strain curve described in section 4.2.6. As mentioned previously, the model

was verified with and accounts for a wide range of confinement levels, with lateral pressures of up to

100% of the unconfined concrete strength. The formulation models concrete dilation as a function of the

lateral pressure ratio and the concrete compressive strength. The model proposed by Montoya was based

on cylinder tests carried out by Imran and Pantazopoulou, where the cylinders were subjected to different

confinement levels and loading paths. The resulting proposed model is (Montoya et al, 2006):

𝑓 𝜀 2
𝜀𝑐𝑙 = �1.9 + 24.2 𝑓𝑐𝑙′ � �𝜀 𝑐 � (4.7.3.1)
𝑐 𝑐𝑐
103
𝑓 𝜀
𝜈 = �1.9 + 24.2 𝑓𝑐𝑙′ � 𝜀2𝑐 ∙ 10−3 + 𝜈𝑜 (4.7.3.2)
𝑐 𝑐𝑐

𝜈𝑜 ≤ 𝜈 ≤ 1.1 (4.7.3.3)

�𝑓𝑐𝑗 +𝑓𝑐𝑘 �
𝑓𝑐𝑙𝑖 = − 2
; 𝑓𝑐𝑙𝑖 > 0, 𝑓𝑐𝑗 , 𝑓𝑐𝑘 < 0 (4.7.3.4)

where 𝑓𝑐1 is the lateral pressure, 𝑓𝑐′ is the unconfined concrete strength, 𝜀𝑐𝑐 is the strain at peak

stress, 𝜈 is the secant poisson ratio, and 𝑗, 𝑘, are the principle directions normal to i.

These relationships are based on cylinder tests with constant lateral pressure throughout loading.

However, they can also be used in incremental calculations with varying lateral pressure and strain at

peak stress. For constant lateral pressure and peak strain, the relationship becomes a straight line, where

the lateral strain increases proportional to axial strain. Comparisons to experimental results are shown in

Figure 47.

Figure 47: Lateral strain-axial strain curves for Imran-Pantazopoulou tests (Montoya, 2003)

There is no maximum value for the secant Poisson’s ratio. However, in finite element programs,

the deformation capacity of the elements used will provide a limit (Montoya, 2003).
104
4.7.4 Variable Montoya With Limit

In this model, the lateral pressure and poisson’s ratio is calculated in the same way as the Montoya

2003 model. The only modification in this model is that a limit of ν = 0.50 is imposed on the calculated

Poisson’s ratio.

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, βl, which is calculated by

the confinement strength models. The value of βl serves to modify the concrete compression response

curves by increasing both the uniaxial compressive strength, f’c, and corresponding strain, ε0, to determine

the peak compressive strength, fp, and corresponding strain, εp, as follows:

f p = β d β l f' c (4.8.1)

ε p = β d βl ε o (4.8.2)

(The parameter, βd, accounts for compression softening. Also, note that the calculation of εp is subject to

modification if certain compression softening models are selected.)

4.8.1 Strength Enhancement Neglected

The uniaxial compressive strength and corresponding strain are not enhanced for confinement.

βl = 1 (4.8.1.1)

4.8.2 Kupfer / Richart Model

In the case of triaxial compression, in which fc3< fc2< fc1<0, the strength enhancement factor for

the direction of the largest compressive stress, fc3, 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).


105

  f cn   f cn  
2
 f 
β l = 1 + 0.92  − 0.76   + 4.1 cl , f c 2 < f c1 < 0 (4.8.2.1)

  f 'c   f ' c    f 'c 

where f’c is the uniaxial concrete cylinder compressive strength, fcn is the difference in normal lateral

stresses acting on the concrete:

f cn = −( f c 2 − f c1 ) > 0 (4.8.2.2)

and fcl is the lateral confining stress on the concrete:

f cl = − f c1 > 0 (4.8.2.3)

The strength enhancement factor for the other compressive stress directions are similarly determined by

interchanging fc3, fc2, fc1 as necessary.

The above expressions can be extended to the case of biaxial compression, in which fc3< fc2<0 and

fc1 = 0, to determine the strength enhancement factor for the direction of the largest compressive stress, fc3.

Strength enhancement in the direction of fc2, is similarly determined by interchanging fc3 for fc2 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, fc3f, in the presence of compressive stresses, fc1 and fc2 is determined by numerically solving the

failure surface of Hsieh et al. (1979):

J2 J2 f c1 I
2.0108 + 0.9714 + 9.1412 + 0.2312 1 − 1 = 0 (4.8.3.1)
f 'c
2
f 'c f 'c f 'c

where I1 is the first stress invariant determined as:

I 1 = f c1 + f c 2 + f c 3 f (4.8.3.2)

and J2 is the second deviatoric stress invariant determined as:

J2 =
1
6
[ (
( f c1 − f c 2 )2 + f c 2 − f c3 f ) + (f
2
c3 f − f c1 )]
2
(4.8.3.3)

The stress enhancement factor βl is the ratio of the failure stress, fc3f, to the uniaxial concrete cylinder

strength, f’c:
106
f c3 f
βl = (4.8.3.4)
f 'c

4.8.4 Montoya / Ottosen

In this model, the stress required to cause failure the in the direction of the maximum compressive

stress, fc3f, in the presence of compressive stresses, fc1 and fc2 is determined by numerically solving the

four parameter failure surface of Ottosen (1979):

A⋅ J2 λ J2 B ⋅ I1
2
+ + −1 = 0 (4.8.4.1)
f 'c f 'c f 'c

I1 is the first stress invariant determined by Equation 4.8.3.2 and J2 is the second deviatoric stress

invariant determined by Equation 4.8.3.3.

The first dimensionless parameter A is computed as follows:

 2.731 for R < 0.2 and f' c ≤ 40 MPa


12.039 for R < 0.2 and f' c > 40 MPa
 (4.8.4.2)
A=
 2.327 for R > 0.2 and f' c ≤ 40 MPa

14.128 for R > 0.2 and f' c > 40 MPa

R is the ratio of the lateral compressive stress, fl to the uniaxial concrete strength, f’c, computed as

follows:

f cl
R= (4.8.4.3)
f 'c

( f c1 + f c2 )
f cl = − (4.8.4.4)
2

The second dimensionless parameter B is determined from the concrete tensile strength, ft, and the

concrete biaxial compressive strength, fbc, as follows:

A f bc − f t 1  f ' c f' 
B= ⋅ +  − c  (4.8.4.5)
9 f 'c 3  ft f bc 

f t = 0.65( f ' c )
0.33
(4.8.4.6)

f bc = 1.16 f ' c (4.8.4.7)

The value of λ is computed using the third and fourth dimensionless parameters, K1 and K2, as follows:
107
λ = K 1 + K 2 cos 3θ (4.8.4.8)

3  f ' c  A  f t  
K1 =  −   + 1 + 1 (4.8.4.9)
2  f t  3  f ' c  

3  f ' c  A  f t  
K2 =  −   − 1 − 2 B − 1 (4.8.4.10)
2  f t  3  f ' c  

The invariant cos3θ is computed as follows:

5.196 J 3
cos 3θ = 1.5
(4.8.4.11)
2J 2

where the third deviatoric stress invariant J3 is given by:

 I  I  I 
J 3 =  f c1 − 1  f c 2 − 1  f c 3 f − 1  (4.8.4.12)
 3  3  3

The stress enhancement factor βl is the ratio of the failure stress, fc3f, to the uniaxial concrete

cylinder strength, f’c:

f c3 f
βl = (4.8.4.13)
f 'c

4.9 Cracking Criterion

The cracking strength of concrete, fcr, 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 fcr, based on the coexisting compressive

stresses or strains. Therefore, fcr is generally different from the input value of concrete tensile strength, f’t.

Having determined fcr, the cracking strain, εcr, is computed by assuming a linear-elastic relationship

before cracking:

f cr
ε cr = (4.9.1)
Ec
108
where Ec is initial tangent stiffness of concrete. The values of fcr and εcr are then used in the tension

stress-strain response and tension stiffening models. The following discussion describes the available

cracking criterion models.

4.9.1 Uniaxial cracking stress

The cracking strength is taken as the specified uniaxial cracking strength:

f cr = f ' t (4.9.1.1)

4.9.2 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 Mohr’s circles defining combinations of shear stress, τ, and normal stresses, fc1

and fc3, resulting in shear failure.

As shown in Figure 48, the shear strength is comprised of a stress-independent component and a

stress-dependent component. The latter is the internal angle of friction, φ, 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, fc3, is equal to the concrete cylinder strength, f’c, and fc1 is zero. Therefore:

1 − sin φ
c = f 'c (4.9.2.1)
2 cos φ

Further, when fc3 = 0, the failure tensile stress fc1 is equal to fcru, which is computed from the envelope as:

2c ⋅ cos φ
f cru = f ' c (4.9.2.2)
2 cos φ

Given a set of principal concrete strains, εc3 < εc2 < εc1, the principal compressive stress, fc3, is

computed as follows:
109

  ε   ε c3 
2

− f ' c 2 c 3  −    for ε c 3 <ε o < 0
   ε o   εo  
  
f c3 = − f ' c for ε o < ε c 3 < 0 (4.9.2.3)
 0 for 0 < ε c 3


Having defined the principal compressive stress, the cracking strength is the principal tensile stress, fc1, of

the Mohr’s circle tangent to the failure envelope. The value of fcr is computed as follows:

 f 
f cr = f cru 1 + c 3 , 0.20 f t ' ≤ f cr ≤ f t ' (4.9.2.4)
 f 'c 

τ
φ=37°

fc

fcr

fc3 fcru

Figure 48: Mohr-Coulomb (Stress) cracking criterion


110
4.9.3 Mohr-Coulomb (Strain)

This model is similar to the Mohr-Coulomb (Stress) model. Given a set of principal concrete

strains, εc3 < εc2 < εc1, and the strain εo corresponding to the uniaxial concrete cylinder compressive

strength, the failure cracking strength is computed as follows:

 ε 
f cr = f cru 1 − c 3 , 0.20 f ' t ≤ f cr ≤ f ' t (4.9.3.1)
 εo 

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):

 f 
f cr = f cru 1 + 0.8 c 3 , 0.20 f t ' ≤ f cr ≤ f t ' (4.9.4.1)
 f 'c 

The principal compressive stress, fc3, is computed from equation 4.9.2.3. The unconfined cracking

strength, fcru, is computed from the compressive strength, f’c, as follows:

f cru = 0.6 f 'c (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, fcr. 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:

 ε 
f cr = f ' t 1 − c 2 , 0.25 f ' t ≤ f cr ≤ f ' t (4.9.5.1)
 εo 

where εc2 is the principal compressive strain, εo is the strain corresponding to the concrete cylinder

compressive strength, f’c.


111
4.10 Crack Stress Check

Calculation of the shear stress on a crack is included in VecTor2. The basic method for calculating

this stress is the DSFM/MCFT. An advanced option for calculation stress along a crack is also available

in VecTor2 and utilizes a different convergence criteria.

4.10.1 Crack Stress Calc Omitted

Calculation of shear stress on the crack is ignored.

4.10.2 Basic (DSFM/MCFT)

The basic option for crack stress calculation is based on the Modified Compression Field Theory

(MCFT) and Disturbed Stress Field Model (DSFM). The MCFT stress calculations are based on

Vecchio-Collins (1986). In the MCFT, stress is transferred across a crack through aggregate interlock.

Figure 49: Shear across with crack through aggregate interlock

The tensile stresses at a crack are zero, but the local shear stresses on the crack, 𝜈𝑐𝑖 are calculated

using the equation below, and are accompanied by small local compressive stresses across the crack, 𝑓𝑐𝑖 ,

based on work by Walraven (Vecchio-Collins, 1986).


2
𝑓𝑐𝑖
𝜈𝑐𝑖 = 0.18𝜈𝑐𝑖,𝑚𝑎𝑥 + 1.64𝑓𝑐𝑖 − 0.82 𝜈 (4.10.2.1)
𝑐𝑖,𝑚𝑎𝑥

where

�−𝑓𝑐′
𝜈𝑐𝑖,𝑚𝑎𝑥 = 0.31+24𝑤/(𝑎+16) (4.10.2.2)
112
𝑤 = 𝜖1 ∙ 𝑠𝑐𝑟 (4.10.2.3)

The formulation for scr is described in detail in section 2.5.11.

Shear stress on the crack can also be calculated by the Disturbed Stress Field Model (DSFM), which

builds on the MCFT. The calculation for the maximum shear stress on the crack is formulated in

VecTor2 as follows:

0.18�𝑓𝑐′
𝑣𝑐𝑖,𝑚𝑎𝑥 = 24𝑤𝑐𝑟 (4.10.2.4)
0.31+
𝑎𝑔 +16

4.10.3 Advanced (Lee 2009)

The advanced crack stress calculation differs in terms of the convergence criteria used; this model

employs a bi-section method of convergence. The crack stress formulation, however, is the same.

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

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, f c*2 , by a crack coefficient βcr as follows:


113

f c 2 = β cr f c*2 (4.11.1)

If the crack width check is omitted, βcr = 1.0. Otherwise βcr is computed as follows:

1 for w < wl
β cr =  (4.11.2)
1 − (w − wl ) 3 ≥ 0 for wl < w

where w is the crack width in the element, and wl 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.

When FRC is used, a maximum crack width check is automatically performed in VecTor2,

limiting the tensile stress attained by the fibres. The maximum crack width for an infinite element is

calculated from the average crack width, as follows (Deluce et al 2012):

𝑉𝑓 𝑙𝑓
𝑤𝑐𝑟,𝑚𝑎𝑥 = �1.7 + 3.4 𝑑𝑓
� 𝑤𝑐𝑟,𝑎𝑣𝑔 (4.11.1.1)

In order to consider the effect of element size on the difference between the maximum and average crack

widths, a linear interpolation between the average crack spacing and 1000 mm is employed, as illustrated

in Figure 50.

Figure 50: Maximum crack width check including effect of element size

As a result of this crack width check, the tensile stress attained by the fibres at an average crack width is

limited by the stress that can be attained by the fibres at the maximum crack width:

𝑓𝑓,𝑤𝑐𝑟,𝑎𝑣𝑔 ≤ 𝑓𝑓,𝑤𝑐𝑟,𝑚𝑎𝑥 (4.11.1.2)


114
4.12 Crack Slip Calculation

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, δs, along the crack: stress-based models,

constant rotation lag models, and hybrid models.

Stress-based models relate the shear slip, δs, along the crack to the local shear stress, vci, along

the crack. The shear slip strain, γs, is computed as the shear slip divided by the crack spacing, s:

a δs
γs = (4.12.1)
s

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, ∆θσ, to

the post-cracking rotation of the principal strain field, ∆θε, by a specified rotation lag, θl, as follows:

 ∆θ ε for | ∆θ ε |≤ θ l
∆θ σ = 
(
 ∆θ ε − θ
l
) for | ∆θ ε |> θ l
(4.12.2)

The inclination of the principal stress field is determined by adding ∆θσ to the inclination at cracking, θic:

θ σ = θ ic + ∆θ σ (4.12.3)

Shear strains can then be determined from Mohr’s circle transformation of the total strains as follows:

b
( )
γ s = γ xy ⋅ cos2θ σ + ε y − ε x ⋅ sin 2θ σ (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.

( a
γ s = max γ s , γ s
b
) (4.12.5)
115
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.

4.12.1 Not considered

As in the Modified Compression Field Theory, the analysis does not account for shear slip along

the crack:

γs = 0 (4.12.1.1)

4.12.2 Vecchio-Lai (Cyclic)

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, δs, is computed as follows:

ψ
δ s = δ *s ≤ 2w (4.12.2.1)
1− ψ

where

0.5v ci max + v co
δ *s = (4.12.2.2)
1.8w − 0.8
( )
+ 0.234 w −0.707 − 0.20 ⋅ f cc

v ci
ψ= (4.12.2.3)
v ci max

f 'c
v ci max = (4.12.2.4)
0.31 + 24 w (a + 16 )

f cc
v co = (in MPa ) (4.12.2.5)
30
116
Additionally, vci 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 fcc is the concrete cube strength, taken as

1.2f’c. The term vco, implements an initial offset in the crack shear-slip relationship.

4.12.3 Maekawa (Monotonic)

This model is proposed by Okamura and Maekawa (1991) and utilized in their fixed non-

orthogonal crack model. The slip along the crack, δs, is computed as follows:

ψ
δs = w ≤ 2w (4.12.3.1)
1− ψ

where,

v ci
ψ= (4.12.3.2)
v ci max

f 'c
v ci max = (4.12.3.3)
0.31 + 24 w (a + 16 )

Additionally, vci, 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, δs, is

computed as follows:

v ci + v co
δs = ≤ 2w (4.12.4.1)
1.8w − 0.8
( )
+ 0.234 w −0.707 − 0.20 ⋅ f cc

where vci, vco, w and fcc are as previously defined.

4.12.5 Hybrid-I, II, and III Models

The Hybrid-I Model combines the Walraven model with the constant rotation lag model.

The Hybrid-II Model combines the Vecchio-Lai model with the constant rotation lag model.

The Hybrid-III model combines the Maekawa model with the constant rotation lag model.
117
l
In these hybrid models, the constant rotation lag, θ , 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.

4.13 Creep and Relaxation

Currently, this option is not available in VecTor2.

4.14 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.

4.14.1 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 εc, the concrete

compressive stress, fc, is computed as follows:

0 for 0 < ε c


ε
f c =  c ⋅ f cm for ε cm < ε c < 0 (4.14.1.1)
 ε cm

 f bc (ε c )
 for ε c < ε cm < 0
118

where ε cm is the maximum previously attained compressive strain, f cm is the corresponding stress, and

f bc (ε c ) is the function defining the monotonic compression stress-strain base curve. In the latter case,

ε cm and f cm are updated as εc and fc respectively.

Unloading in compression results in concrete stresses as follows:


f c = E cm ⋅ εc (4.14.1.2)


where, E cm is the unloading modulus in compression, computed as:

− f cm
E cm = (4.14.1.3)
ε cm

When reloading in the tensile domain to a tensile strain of εc, the concrete tensile stress, fc, is

computed as follows:

 εc
 ⋅ f tm for 0 < ε c < ε tm

f c =  ε tm (4.14.1.4)

 f bt (ε c )
 for ε cm < ε c

where ε tm is the maximum previously attained tensile strain, f tm is the corresponding stress, and

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

f tm are updated as εc and fc respectively.

Unloading in tension results in concrete stresses as follows:


f c = E tm ⋅ εc (4.14.1.5)


where, E tm is the unloading modulus in tension, computed as:

− f tm
E tm = (4.14.1.6)
ε tm

4.14.2 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).


119
When reloading in the compression domain to a compressive strain of εc, the concrete

compressive stress, fc, is computed as follows:

0 for ε cp < ε c < 0 or 0 < ε c


 p
 ε − εc
fc =  c p
⋅ f cm for ε cm < ε c < ε cp < 0 (4.14.2.1)
 ε cm − ε c
 f bc (ε c )
 for ε c < ε cm < 0

where ε cp is the current plastic offset strain, ε cm is the maximum previously attained compressive strain,

f cm is the corresponding stress, and f bc (ε c ) is the function defining the monotonic compression stress-

strain base curve. In the latter case, ε cm and f cm are updated as εc and fc respectively.

At a given compressive strain, εc, the instantaneous plastic strain, ε cp ' , is computed as follows:

     
2

ε − ε 0.87 ε c  − 0.29 ε c   for 1.5ε p < ε c


c p
 εp  εp  
ε cp ' =        (4.14.2.2)
  εp 
ε c − 0.001305  for ε c < 1.5ε p


  0.002 

where εp is the strain corresponding to the peak stress on the compression base curve. If ε cp ' is more

compressive than ε cp , then ε cp is updated as ε cp ' .

Unloading in compression results in concrete stress, fc, is as follows:


f c = E cm (
ε c − ε cp ) (4.14.2.3)


where E cm is the unloading modulus in compression, computed as:

f cm

E cm = (4.14.2.4)
(ε cm − ε cp )
When reloading in the tensile domain to a tensile strain of εc, the concrete tensile stress, fc, is

computed as follows:

 ε c − ε cp

 p
⋅ f tm for ε cp < ε c < ε tm
f c =  ε tm − ε c (4.14.2.5)

 f bt (ε c )
 for ε cm < ε c
120

where ε cp is the current plastic offset strain , ε tm is the maximum previously attained tensile strain, f tm

is the corresponding stress, and f bt (ε c ) is the function defining the monotonic tensile stress-strain base

curve. In the latter case, ε tm and f tm are updated as εc and fc respectively.

For a given strain, εc, in the tension domain, the instantaneous plastic strain, ε cp ' , is initially the

strain at which the response first traverses from the compression to tension domain. Thereafter, the

monotonic stress-strain response is calculated with respect to that strain. Subsequently ε cp ' is updated

until non-negative and held at zero afterwards such that no plastic offset strains are considered in the

tension domain.

Unloading in tension results in concrete stress, fc, as follows:


f c = E tm (
ε c − ε cp ) (4.13.2.6)


where E tm is the unloading modulus in tension, computed as:

f tm

E tm = (4.14.2.7)
(ε tm − ε cp )

4.14.3 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 εc results in concrete stress, fc, is as follows:

E c (ε c − ε cm )
Nc
f c = f cm + E c (ε c − ε cm ) + for 1 ≤ N c ≤ 20 (4.14.3.1)
(
N c ε cp − ε cm )
N c −1

where ε cp is the current plastic offset strain, ε cm is the maximum previously attained compressive strain,

f cm is the corresponding stress. Nc 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, Ec, as follows:

Nc =
(
E c ⋅ ε cp − ε cm ) (4.14.3.2)
f cm + (
E c ε cp − ε cm )
121
In the case that Nc is less than one or greater than twenty, fc is computed by a linear unloading response

between from εc to ε cp :

(
f c = E c ε c − ε cp ) for N c ≤ 1 or 20 ≤ N c (4.14.3.3)

Unloading in tension results in concrete stress, fc, as follows:

E c (ε tm − ε c )
Nt
f c = f tm − E c (ε tm − ε c ) + for 1 ≤ N t ≤ 20 (4.14.3.4)
(
N t ε tm − ε cp )
N t −1

where ε cp is the current plastic offset strain, ε tm is the maximum previously attained tensile strain, f tm

is the corresponding stress. Nt is computed as follows such that the initial unloading modulus is equal to

the initial tangent stiffness of concrete, Ec:

Nt =
(
E c ⋅ ε tm − ε cp ) (4.14.3.5)
(
E c ε tm − ε cp )− f tm

In the case that Nt is less than one or greater than twenty, fc is computed by a linear unloading response

between from εc to ε cp :

(
f c = E c ε c − ε cp ) for N t ≤ 1 or 20 ≤ N t (4.14.3.6)

4.14.4 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 Figure 51 and Figure 52, respectively.

When reloading in the compression domain to a compressive strain of εc, the concrete compressive

stress, fc, is computed as follows:

+
f c = f ro + E cm (ε c − ε ro ) (4.14.4.1)

where ε ro is the strain at load reversal in the current hysteretic loop, f ro is the corresponding stress, ε cm

is the unloading strain in the current hysteretic loop are ε cm and f cm is the corresponding stress.
122
+
The reloading modulus in compression, E cm , is computed as follows

+ β d ⋅ f cm − f ro
E cm = (4.14.4.2)
ε cm − ε ro

+
The damage indicator, βd, serves to degrade E cm , 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,

εrec, and is computed as follows:

 1
for ε c < ε p

(
 1 + 0.10 ε rec ε p
0.5
)
βd =  (4.14.4.3)
 1
 for ε c > ε p
(
1 + 0.175 ε rec ε p
0.6
)
ε rec = ε cm − ε ro (4.14.4.4)

where εp is the strain corresponding to the peak stress in the base-curve.

The instantaneous plastic strain, ε cp ' , for the compression domain is computed as follows:

 ε 
2
 
ε cp ' = ε p 0.166 cm  + 0.132 ε cm  (4.14.4.5)
  εp   εp 
    

If ε cp ' is more compressive than ε cp , then ε cp is updated as ε cp ' .

Unloading in compression to a strain of εc results in concrete stress, fc, is as follows:

E c (0.071 − 1)(ε c − ε cm )
Nc
f c = f cm + E c (ε c − ε cm ) + (4.14.4.6)
(
N c ε cp − ε cm )
N c −1

where ε cp is the current plastic offset strain, ε cm is the maximum previously attained compressive strain,

f cm is the corresponding stress. Nc 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, Ec, at the beginning of the unloading branch and equal to 0.071 Ec at the end of the unloading

branch, as follows:

Nc =
(
E c (1 − 0.071) ε cp − ε cm ) (4.14.4.7)
f cm + (
E c ε cp − ε cm )
123
When reloading in the tensile domain to a tensile strain of εc, the concrete tensile stress, fc, is

computed as follows:

+
f c = β t ⋅ f tm − E tm (ε tm − ε c ) (4.14.4.8)

+
where ε tm is the unloading strain in the current hysteretic loop, f cm is the corresponding stress, and E cm

is the reloading modulus in tension, computed as follows

+ β t ⋅ f tm − f ro
E tm = (4.14.4.9)
ε tm − ε ro

+
The damage indicator, βt, serves to degrade E tm , 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,

εrec, and is computed as follows:

1
βd = (4.14.4.10)
1 + 1.15(ε rec )
0.25

ε rec = ε tm − ε ro (4.14.4.11)

The instantaneous plastic strain, ε cp ' , for the tension domain is computed as follows:

2
ε cp ' = 146ε tm + 0.523ε tm (4.14.4.12)

If ε cp ' is more tensile than ε cp , then ε cp is updated as ε cp ' .

Unloading in tension results in concrete stress, fc, as follows:

f c = f tm − E c (ε tm − ε c ) +
(E c )
− E cf (ε tm − ε c )
Nt
(4.14.4.13)
(
N t ε tm − ε cp )
N t −1

where ε cp is the current plastic offset strain, ε tm is the maximum previously attained tensile strain, f tm

is the corresponding stress. Nt is computed as follows such that the unloading modulus is equal to the

initial tangent stiffness of concrete, Ec at the beginning of the unloading branch, and equal to E cf at the

end of the unloading branch, and is computed as follows:

Nt =
(E c )(
− E cf ⋅ ε tm − ε cp ) (4.14.4.14)
(
E c ε tm − ε cp )− f tm
124

0.071 ⋅ E c (0.001 ε tm )
 for ε tm < 0.001
E cf =  (4.14.4.15)
0.053 ⋅ E c (0.001 ε tm )
 for ε tm < 0.001

Partial unloading and reloading is effected by similar equations with modifications to the reversal

and maximum strains under consideration.

-fc base curve for


compression softened
response
fp

ε cp εp −εc

Figure 51: Palermo model of concrete hysteretic response in compression


125
fc

fcr
base curve for tension
stiffened response

εcr εc
ε cp

Figure 52: Palermo model of concrete hysteretic response in tension


126

5 Models for other material types

The original material type modeled in VecTor2 is reinforced concrete; the majority of this manual

is dedicated to the discussion of material and behaviour models for concrete. The following materials can

also be modeled as the main material type in VecTor2: structural steel, masonry, and orthotropic wood.

In addition, using the main material types and smeared reinforcement types the following laminates can

be modeled in VecTor2: concrete-steel laminate, concrete-SFRC laminate, masonry-SFRC laminate,

concrete-wood laminate. This section contains a discussion of the above materials and how they are

analyzed in VecTor2. A discussion of how to model them in FormWorks is included in the Reinforced

Concrete Material Types part of the FormWorks section of this user manual.

5.1 Structural Steel

Structural steel is modeled in VecTor2 as a linear-elastic material up to the point of yielding, after

which plastic deformation and strain hardening occur. Strain hardening is modeled in VecTor2 using a

trilinear relationship, illustrated in Figure 53.

Figure 53: Stress-strain behavior of structural steel in VecTor2


127

Steel hysteresis is also considered, and is described in section 6.2. In VecTor2, the hysteresis

model for structural steel is the same as for reinforcing steel.

The main different between the modeling of structural steel and reinforcing steel in VecTor2 is the

consideration of biaxial loading. For structural steel, particularly steel plates, biaxial effects are

important; the Von Mises failure criterion is used for structural steel and is discussed further in the

concrete-steel laminate section of the manual, section 5.4. For reinforcing steel, biaxial effects are not

considered.

5.2 Masonry

5.2.1 Introduction

Masonry is an orthotropic material consisting of masonry units and mortar joints, shown in Figure

54. As with the smeared crack approach to the analysis of cracked concrete, for sufficiently large

masonry structures, the masonry can be modeled as a continuum with average properties where joint

failures are smeared across the single finite element.

Figure 54: Masonry joints


128

In Figure 54, sp2 is the spacing between head joints and sp1 is the spacing between bed joints.

These are entered in the Define Reinforced Concrete Materials tab of FormWorks as Sx and Sy ,

respectively.

5.2.2 Compressive Behavior

The uniaxial compressive pre-peak stress-strain behaviour of unreinforced masonry (URM) can

be modeled with a relationship similar to the one used to model the compressive response of reinforced

concrete. When modeling URM in VecTor2, because the initial tangent modulus is not necessarily

directly related to the peak stress-strain ratio, the Hoshikuma et al. model is recommended, as it allows for

the independent definition of peak stress, peak strain, and initial tangent modulus.

The material strength and elastic modulus of URM change with the direction of the axes of

orthotropy. Ganz’ failure criteria for URM is implemented in VecTor2, expressed in terms of the

principal stresses.

Figure 55: Ganz failure criterion implemented in VecTor2 (Ganz, 1985)


129

The following equations define the failure criteria above:

Equation Ib:

�(1 − ωm R)(ωm − R)�σ22 + t 2 �(R − 1)�fmy − fmx �(ωm + 1)2 − fmy R(1 + ω2m ) +

fmx (R − ω2m ) + 2ω2m fmy + fmx ωm (R − 1)� σ2 + �fmy − fmx �fmy ωm − fmx fmy ω2m ≤ 0 (5.2.2.1)

Equation II:

−Rσ22 − �t 2 �fmy − fmx �(1 − R) + fmx + Rfmy �σ2 − fmx fmy ≤ 0 (5.2.2.2)

Equation IIId:
2
�−(1 + ωm )2 (R − 1)2 t 4 + 2(R − 1)�−2ωm + R(1 + ω2m )� − �R(ωm − 1)� � σ22 +

�2ωm fmx (1 − ωm )�t 2 + R(1 − t 2 )�� σ2 − (ωm fmx )2 ≤ 0 (5.2.2.3)

Equation IIIa,b:

(t 2 + R2 − R2 t 2 )σ2 + fmx �t 2 + R(1 − t 2 )� ≤ 0 (5.2.2.4)

Equation IVa:
2
�(R − 1)2 t 2 (1 − t 2 ) − �tan φ �1 + t 2 (R − 1)�� � σ22 + 2c tan φ �1 + t 2 (R − 1)�σ2 − c 2 ≤ 0 (5.2.2.5)

Equation IVc:

�R2 �et 4 + t 2 �1 − 4m2 − 2λm(1 + 2m)� + (1 + 2λ)m2 � + R �−2et 4 + t 2 �4m(1 + λ) +

2m2 (1 + 2λ)� − 2m(1 + λ)� + et 4 − t 2 �1 + 2m(1 + λ)� + 1� σ22 + �2m�R + t 2 (1 − R)��fty


′ (1
+

′ (1
2λ) − ξ� + 2�1 + t 2 (R − 1)� �ξ − fty ′
+ λ)�� σ2 + fty ′
�fty ′
− 2�ξ − λfty �� ≤ 0 (5.2.2.6)

where

𝑡 = sin 𝛽

𝑅 = 𝜎1 /𝜎2

𝜆 = sin 𝜑 /(1 − sin 𝜑)


130
π φ
ξ = c ∙ tan � + �
4 2

f′ty
e = m�m + 2(1 + λ + λm)� where m = µf
mx

And 𝛽 is the angle between the principal axes and axes of orthotropy, 𝑅 is the ratio of principal

stresses, 𝜑 is the friction angle of the mortar bed joint interface, 𝑓𝑚𝑥 is the compressive strength in the x-

direction with bed joint aligned horizontally (at 𝛼 = 0), 𝑓𝑚𝑦 is the compressive strength in the y-direction


with bed joint aligned horizontally, 𝑐 is the cohesion of the mortar bed joint interface, 𝑓𝑡𝑦 is the tensile

strength of the mortar bed joint interface in the x-direction (perpendicular to the bed joints), 𝜔𝑚 is the

ratio between the tensile strength in the x-direction (for 𝛼 = 0) and compressive strength 𝑓𝑚𝑦 , and 𝜇 is the

reduction coefficient for the strength 𝑓𝑚𝑥 .

Within the implementation of the Ganz failure criteria in VecTor2, in order to properly account

for the orthotropic behaviour of masonry, the maximum compressive strength of the masonry is limited

by defining a factor βm , the ratio of the maximum principal compressive strength and the compressive

strength fmy .

𝑓𝑝 = 𝛽𝑑 𝛽1 𝛽𝑚 𝑓𝑐′ (5.2.2.7)

where 𝛽𝑑 is the compression softening parameter (coincident with the Vecchio 1992-A model for

RC) , 𝛽1 is the parameter accounting for confinement effect, 𝛽𝑚 (= 𝜎2,𝑚𝑎𝑥 /𝑓𝑚𝑦 ) is a parameter

accounting for orthotropic effect on compressive strength, 𝜎2,𝑚𝑎𝑥 is the principal compressive strength

defined according to the Ganz failure criteria, 𝑓𝑐′ = 𝑓𝑚𝑦 is the maximum compressive strength off

masonry, and fp is the actual compressive strength of masonry.

Figure 56: Masonry compressive stress-strain behavior


131

In VecTor2, when modeling masonry structures, the maximum compressive masonry strength,

fmy , is entered as the cylinder compressive strength, fc′ , in the Define Reinforced Concrete Properties tab

of FormWorks.

Given that masonry is considered orthotropic in VecTor2, the elastic moduli are different in each

direction of orthotropy. A smooth transition from the maximum modulus, 𝐸𝑚𝑦 , to the minimum

modulus, 𝐸𝑚𝑥 , is generated by evaluating the stresses and strains at various angles of β, and taking the

ratio of the two to determine the modulus at each angle. Once the 𝐸 − 𝛽 relationship is determined by

VecTor2, as shown in Figure 57, the initial tangent modulus can be evaluated directly as a function of the

direction of the principal axes with respect to the direction of the horizontal joints.

Figure 57: Variation in masonry elastic modulus

In the Define Reinforced Concrete Materials tab of FormWorks, the initial maximum elastic

modulus, 𝐸𝑚𝑦 , is entered in the Initial Tangent Elastic Modulus, 𝐸𝑐 , space. The ratio between the two

elastic moduli is entered in a different tab in FormWorks, as will be outlined in a subsequent section.

5.2.3 Tensile Behavior

The tensile behaviour of masonry in VecTor2 is modeled as isotopic linear elastic. The stress-

strain relationship is linear until the tensile strength, fty , is reached, after which tension softening occurs.

The suggested VecTor2 tension softening model is the Hordijk model. The tensile strength of masonry is
132
equal to the minimum tensile brick-mortar strength, perpendicular to the bed joints, and is entered as

tensile strength, ft′ , in the Define Reinforced Concrete Materials tab of FormWorks .

5.2.4 Slip Along the Joints

Within the DSFM adaptation for masonry materials, masonry is modeled as a continuum that may

slip along the head and bed joints, even when the material is uncracked.

From the total stress vector, the shear stress and shear strain parallel to the joints can be

determined. Based on the calculated shear strain along the joints, the shear slip, δs , can be determined:

𝛿𝑠 = 𝛾 ∙ 𝑡ℎ𝑖 (5.2.4.1)

where 𝑡ℎ𝑖 is the thickness of the joint.

The average shear slip strain can also be calculated as follows:

𝛿
𝛾𝑠 = 𝑠𝑝𝑠 (5.2.4.2)
𝑖

where 𝛿𝑠 is the slip along the joints and 𝑠𝑝𝑖 is the joint spacing.

Figure 58: Masonry Joint Slip

In VecTor2, by default, the crack slip check is not considered.


133
5.3 Wood (Fixed Orthotropic)

In VecTor2, wood is modeled as a fixed orthotropic material. The two directions of orthotropy

are parallel to the grain and perpendicular to the grain. The longitudinal direction is defined as parallel to

the grain; the transverse direction is perpendicular to the grain. When modeling wood in FormWorks, one

must define the compressive and tensile strength, as well as the elastic moduli and Poisson’s ratios, in

both the longitudinal and transverse directions. No default values will be applied by VecTor2 if an input

is left blank. This is discussed more in the FormWorks-Reinforced Concrete Material Types section of

the manual.

The stress-strain curve for wood is currently linear-elastic.

The following relations are used to model the stress-strain behaviour of wood in VecTor2.
1 𝜈𝑦𝑥
𝑒𝑥𝑥 = 𝜎 −𝐸 𝜎𝑦𝑦 (5.3.1)
𝐸𝑥𝑥 𝑥𝑥 𝑦𝑦

𝜈 1
𝑒𝑦𝑦 = − 𝐸𝑥𝑦 𝜎𝑥𝑥 + 𝜎 (5.3.2)
𝑥𝑥 𝐸𝑦𝑦 𝑦𝑦

1
𝛾𝑥𝑦 = 𝜎 (5.3.3)
𝐺𝑥𝑦 𝑥𝑦

In these equations, we can consider the x-axis to be the longitudinal direction. Thus, the stress-

strain behaviour of wood can be characterized by the following equations:


1 𝜈𝑡𝑙
𝑒𝑙𝑙 = 𝜎 − 𝜎 (5.3.4)
𝐸𝑙 𝑙𝑙 𝐸𝑡 𝑡𝑡

𝜈 1
𝑒𝑡𝑡 = − 𝐸𝑙𝑡 𝜎𝑙𝑙 + 𝜎 (5.3.5)
𝑙 𝐸𝑡 𝑡𝑡

1
𝛾𝑙𝑡 = 𝜎 (5.3.6)
𝐺𝑙𝑡 𝑙𝑡

where 𝑒𝑙𝑙 is the longitudinal strain, 𝑒𝑡𝑡 is the transverse strain, 𝐸𝑙 is the elastic modulus in the

longitudinal direction, 𝐸𝑡 is the elastic modulus in the transverse direction, 𝜈𝑙𝑡 is the Poisson’s ratio

(longitudinal stress-transverse strain), 𝜈𝑡𝑙 is the Poisson’s ratio (transverse stress-longitudinal strain), 𝜎𝑙𝑙

is the longitudinal stress, 𝜎𝑡𝑡 is the transverse stress.


134
5.4 Concrete-Steel Laminate

Steel-concrete composite wall elements typically consist of a thick concrete core integrated with

two thin steel faceplates, where forces are generally transferred between the concrete and the steel

through shear studs. In VecTor2, the DSFM is the basis for the analysis of concrete-steel laminates. As

VecTor2 is a 2D finite element program, the different materials must be modeled as a smeared/combined

element with properties representative of the two materials, as outlined below and described fully in

Vecchio and McQuade (2011). A typical concrete-steel laminate is illustrated in Figure 59.

Figure 59: Concrete-steel laminate element (Vecchio and McQuade, 2011)

The element is analyzed using the same principals of solid mechanics as a typical concrete

element, where:

[𝜎] = [𝐷][𝜀] − [𝜎 0 ] (5.4.1)

The composite material stiffness matrix, [D], and the prestress vector, [σ0 ], are formulated as

𝑡𝑐 𝑡𝑐 2𝑡𝑠
[𝐷] = [𝐷𝑐 ] + ∑𝑛𝑖=1 [𝐷𝑟 ]𝑖 + (𝑡 [𝐷𝑠 ] (5.4.2)
(𝑡𝑐 +2𝑡𝑠 ) (𝑡𝑐+2𝑡𝑠 ) 𝑐 +2𝑡𝑠 )
135

[𝜎 0 ] = [𝐷𝑐 ]�[𝜀𝑐𝑒 ] + �𝜀𝑐𝑝 � + [𝜀𝑐𝑠 ]� + ∑𝑛𝑖=1[𝐷𝑟 ]𝑖 �[𝜀𝑟𝑒 ]𝑖 + �𝜀𝑟𝑝 � � + [𝐷𝑠 ]�[𝜀𝑠𝑒 ] + �𝜀𝑠𝑝 �� (5.4.3)
𝑖

The formulation of [Dc ] and [Dr ] is described fully in (Vecchio, 2000b). The material stiffness matrix for

the steel faceplates is a diagonal matrix with respect to the principal stress directions.

For the yield criteria for the steel plates, the von Mises criterion is used, where:
2
(𝑓𝑠1 − 𝑓𝑠2 )2 + (𝑓𝑠2 − 𝑓𝑠3 )2 + (𝑓𝑠3 − 𝑓𝑠1 )2 = 2�𝑓𝑦 � (5.4.4)

In terms of the contribution of the steel plates to the tension stiffening response of concrete, no

experimental data is currently available. However, due to the fact that the steel plate is typically quite

thin compared to the concrete core and the bond between plates and the concrete is weaker than that of

fully embedded deformed rebar, it is assumed that the steel faceplates do not contribute to concrete

tension –stiffening (Vecchio and McQuade, 2011). In addition, in accordance with the Canadian Design

Code A23.3, for steel-concrete elements with no in-plane reinforcement, the maximum crack spacing is

considered equal to the element thickenss.

A detailed explanation of the modeling of steel-concrete composite structures, specifically wall

elements, can be found in Vecchio and McQuade (2011).

5.5 Concrete-SFRC Laminate and Masonry-SFRC Laminate

Similar to the concrete-steel laminate, a concrete-SFRC laminate combines a concrete core with

steel-fibre reinforced concrete faceplates. Using an SFRC laminate will result in a higher capacity and a

more ductile response; a common application of the SFRC laminate is in strengthening of RC slabs,

whereby a thin layer of SFRC is attached to an existing RC slab. Adding the SFRC laminate to the

compression zone (ie. the top of the slab) can significantly increase the flexural capacity of an existing

RC slab. There are also benefits in shear capacity. An overlayer of SFRC also allows for increased post-

cracking residual stress as the steel fibres are efficient at controlling large cracks in most circumstances.

For more details on the concrete or SFRC behaviour, see sections 4 or 6.5, respectively. For details

on how to model this material in FormWorks, see section 10.4.1.6.


136
The masonry-SFRC laminate is similar to the concrete-SFRC laminate, and is defined in FormWorks

in much the same way. For a detailed description of the behaviour of masonry, see section 5.2. Section

6.5 describes the stress-strain response of SFRC. For information on how to model masonry-SFRC

laminate in FormWorks, see section 10.4.1.7 of this manual.

5.6 Concrete-Ortho Laminate

Concrete-wood laminates are commonly used in both floor and beam construction. In new floor

construction, solid concrete is typically placed on timber floor beams or a solid layer of wood. The wood

layer functions to replace the cracked concrete-steel reinforcement section of a solid concrete slab, and

also reduces the need for formwork. Similarly, deep beams benefit from concrete-wood composite

construction, as wood can eliminate the high tensile stresses in the concrete. Hence, bridges can also

utilize composite concrete-timber beams. As with other laminates, forces must be transferred between the

concrete and the wood, most likely through shear studs.

Section 4 describes in detail the behaviour models for concrete. For the stress-strain behaviour of

orthotropic wood, see section 5.3. Details on how to model concrete-ortho laminates are provided in

section 10.4.1.8 of this manual.


137

6 Models for Reinforcement Materials

The following discussion describes the constitutive and behavioral models pertaining to the response

of reinforcement materials. The monotonic stress-strain response of reinforcement is discussed, followed

by the hysteretic response models. Subsequently, dowel action and reinforcement buckling models are

discussed.

After the discussion of conventional steel reinforcement, the modeling of Steel Fibre Reinforced

Concrete in VecTor2 is discussed. Subsequently, a brief discussion of laminates and shape memory

alloys in VecTor2 is included.

6.1 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.

6.1.1 Ductile Steel Reinforcement

The reinforcement stress-strain response is composed mainly of three parts, as shown in Figure 60,

including an initial linear-elastic response, a yield plateau, and either a linear or nonlinear strain-

hardening phase until rupture. Moreover, this monotonic stress-strain curve describes the back-bone curve

of the Seckin or Menegotto-Pinto models, as discussed below for the hysteretic response. The

reinforcement stress, fs, in tension and compression is determined as follows:


138

 Es ε s for ε s ≤ ε y
f for ε y < ε s ≤ ε sh
 y
 P
f s =    εu − ε s  (6.1.1.1)
 fu + ( f y − fu )  ε − ε  for ε sh < ε s ≤ ε u
  u sh 
0 for ε u < ε s

where εs is the reinforcement strain (εs=|εs|), εy is the yield strain, εsh is the strain at the onset of the strain

hardening, εu is the ultimate strain, Es is the elastic modulus, fy is the yield strength, fu is the ultimate

strength, and P is the strain-hardening parameter.

Figure 60: Ductile steel reinforcement stress-strain response


[left(a): Linear strain-hardening (Trilinear); right (b): Nonlinear strain-hardening (HP4)]

There are two options for the strain-hardening phases after the yield plateau; these are linear strain-

hardening (trilinear, P=1) and nonlinear strain-hardening (HP4, P=4), as shown in Figure 60. Elastic-

plastic or bilinear stress-strain curves are generated by the trilinear option given in Figure 60a. The default

option is HP4. (The default option can be changed from the hysteretic response section of reinforcement

models.) The strain hardening modulus, Esh is defined as follow:

( fu − f y )
Esh =    (6.1.1.2)
(ε u − ε sh )
139

6.1.2 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 61. The reinforcement stress, fs, in tension and compression is

determined by a Ramsberg-Osgood formulation as follows:


 1− A 

f s = Es ε s  A +  ≤ fu (6.1.2.1)

 [
1 + (Bε s )
C
]
1C

E sh
A= (6.1.2.2)
Es

E s (1 − A)
B= (6.1.2.2)
f s*

C = transition coefficient (6.1.2.3)

where εs is the reinforcement strain, Es is the initial elastic modulus, fu is the ultimate strength and f s* is

the value at which the second linear branch intercepts the stress axis at zero strain. A representative value

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.

fs
fu

Es

εs
140
Figure 61: Prestressing steel reinforcement stress-strain response

6.1.3 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.

6.1.4 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.

6.1.5 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.

6.2 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.

6.2.1 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.


141
6.2.2 Seckin Model w/ Bauschinger Effect

The following model was proposed by Seckin (1981), as shown in Figure 62. This model 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. There are two options when using this

model; these are Seckin w/ Bauschinger-Trilinear and Seckin w/ Bauschinger-HP4 as shown in Figure 62.

Figure 62: Seckin model of ductile steel reinforcement for hysteretic response
[left (a): Seckin w/ Bauschinger-Trilinear; right (b): Seckin w/ Bauschinger-HP4]

When reloading in a positive cycle to a strain of εj, the reinforcement stress is fs, reflects the Bauschinger

effect with a Ramberg-Osgood formulation as follows:

Em − Er
Er ( ε j − ε 0 ) + N −1 ( j
ε − ε0 )
N
f s=    (6.2.2.1)
N (ε m − ε 0 )

N=
( Em − Er )(ε m − ε 0 ) (6.2.2.2)
f m − Er ( ε m − ε 0 )
142

 Es for ( ε m − ε 0 ) ≤ ε y

  (ε − ε ) 
Er  Es 1.05 − 0.05 m 0 
= for ε y < ( ε m − ε 0 ) < 4ε y (6.2.2.3)
 εy 
  
E for 4ε y ≤ ( ε m − ε 0 )
 s

where, εo is the plastic offset strain in the current cycle, εy is the yield strain, εm is the maximum positive

strain attained in previous cycles, Em is the tangent stiffness at εm, Er is the unloading modulus, Es is the

elastic modulus of the monotonic stress-strain response, and fm is the stress corresponding to εm.

Unloading is linear and results in reinforcement stresses as follow:

fs =f sj −1 + Er ( ε j − ε j −1 )
   (6.2.2.4)

6.2.3 Menegotto-Pinto Model w/ Bauschinger Effect

The following model was proposed by Menegotto and Pinto (1973), as shown in Figure 63.

Similar to the Seckin model, this model includes the Bauschinger effect. There are two options when

using this model; these are Menegotto-Pinto w/ Bauschinger-Trilinear and Menegotto-Pinto w/

Bauschinger-HP4 as shown in Figure 63. The Menegotto-Pinto model defines an asymptotic curve

tangential to two asymptotic lines at the initial (origin) and end (target) points. The reloading part of the

stress-strain curve is defined in a normalized form as follows:

(1 − b ) ε eq
f eq  bε eq +
= (6.2.3.1)
(1 + ε )
1/ R
R
eq

fs − fr ε − εr
f eq = , ε eq = s , (εr=ε0 and fr=0) (6.2.3.2)
f p − fr ε p − εr

Em aξ
b=
Er
= R0 − 1 , ξε=
, R
a2 + ξ
y (ε m −εp ) (6.2.3.3)

 1 
=Er Es  0.8 +  (6.2.3.4)
 5 + 1000ε m 
143
where feq and εeq are normalized stress and strain, and fp and εp are stress and strain at the intersection

point between the initial tangent at the origin and the asymptote at the target, respectively. b is the strain-

hardening ratio of the intended slope (Em) at the target point to the unloading or initial reloading stiffness

(Er) at the origin. R is the independent parameter which defines the curvature according to experimental

parameters, R0, a1, and a2. In the VecTor2 implementation, R0, a1, and a2 are taken 15, 18.5 and 0.15,

respectively and the values of R are limited between 1 and 15. ξεy is the strain difference between the

strain at the intersection point, εp and the maximum (target) strain attained at the previous cycle, εm. In

addition, fm is the stress corresponding to εm, Em is the tangent stiffness at εm, Er is the unloading or initial

reloading modulus, proposed by Dodd and Restrepo-Posoda (1995), εo is the plastic offset strain in the

current cycle, fy is yield strength corresponding to the yield strain, εy, and Es is the elastic modulus of the

monotonic stress-strain response.

Figure 63: Menegotto-Pinto model of ductile steel reinforcement for hysteretic response
[left (a): Menegotto-Pinto w/ Baushinger-Trilinear ; right (b): Menegotto-Pinto w/
Baushinger-HP4

Unloading is linear and results in reinforcement stresses as follow:

fs =f sj −1 + Er ( ε j − ε j −1 )
   (5.2.3.5)
144
6.2.4 Elastic Hardening (Curvilinear)

Figure 64 illustrates the curvilinear stress-strain model.

Figure 64: curvilinear strain hardening

The equation for the curvilinear portion of the response is:

𝜀 −𝜀 4
𝑓𝑠 = 𝑓𝑠𝑢 − �𝑓𝑠𝑢 − 𝑓𝑠𝑦 � �𝜀 𝑠𝑢−𝜀 𝑠 � (6.2.4.1)
𝑠𝑢 𝑠ℎ

6.2.5 Elastic-Hardening (Trilinear)

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,

Es. As such, plastic strain offsets result.

Figure 65: Trilinear stress-strain response


145
6.2.6 Elastic Plastic (Bilinear)

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, Es. Figure 66 illustrates this stress-strain model.

Figure 66: Elastic-plastic stress-strain response

6.3 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, vdl, is computed

as a function of the shear slip, δs, at the crack. This shear resistance is subsequently subtracted from the

local shear stress on the crack, vci, thereby reducing the amount of computed shear slip.

6.3.1 Not considered

The shear resistance at a crack due to dowel action is not considered.


146
6.3.2 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 67. (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, Vd, due to the

relative displacement, δs, of the cracks is computed as follows:

Vd = E s I z λ3 δ s ≤ Vdu (6.3.2.1)

πd b4
Iz = (6.3.2.2)
64

kc db
λ=4 (6.3.2.3)
4E s I z

127 ⋅ c f ' c
kc = 23
(6.3.2.4)
db

c = 0.8 (6.3.2.5)

2
Vdu = 1.27 d b f 'c f y (6.3.2.6)

where δs is the shear slip along the crack, db is the diameter of the reinforcement, Es is the elastic modulus

of the reinforcement, fy is the yield strength of the reinforcement, f’c is the compressive strength of the

concrete. The parameter Iz is the area moment of inertia of the reinforcement. The parameter λ compares

the stiffness of the concrete to that to the reinforcing bar. The parameter kc is the stiffness of notional

concrete foundation, where c is an experimentally based coefficient to reflect the bar spacing. The

ultimate dowel force, Vdu, 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:

ρ sV d
vd = (6.3.2.7)
As

where ρs is the reinforcement ratio and As is the area of the reinforcement.


147

δs

Figure 67: Dowel resistance mechanism

The Tassios (Crack Slip) and Tassios (Crack Strength) options are available in VecTor2 for

calculating rebar dowel action. In the crack slip option, the dowel force 𝑉𝑑 is subtracted from the stress of

the concrete at a crack. For crack strength, the dowel force acts as a resisting force.

The default option is Tassios Crack Slip Model. In general, this option is used unless stability issues

related to reinforcement shearing occur during the analysis, in which case the Tassios Strength Model is

used.

6.4 Reinforcement Buckling

Three models are available for the consideration of reinforcing bar buckling. These are the Dhakal-

Maekawa 2002 model (DM), the Refined Dhakal-Maekawa model (RDM), and the Asatsu model. Both

the DM and RDM models are mainly based on the model of Dhakal and Maekawa (2002a, and b). By

using the formulations of Dhakal and Maekawa (2002b), the Seckin and Menegotto-Pinto hysteretic

models were modified by Akkaya et al. (2013) for ductile reinforcement including buckling effects. The

DM and the RDM models can be used only if the ductile steel reinforcement and one of the Seckin or the

Menegotto-Pinto models are selected. Both models assume that the reinforcement buckling begins to

occur when the unsupported length to diameter ratio (b/t=L/D) for the reinforcing bars exceeds 5.0 and

the compressive reinforcement strain exceeds its yield strain εy. Consequently, to take into account the

rebar buckling effect, both models require the input of L/D prior to an analysis for L/D ≥ 5.0. For the
148
default value of L/D and L/D<5, there is no rebar buckling effects considered for the reinforcing bars.

When using the Asatsu model, there is no requirement for the input of L/D.

6.4.1 Dhakal-Maekawa Model 2002 (DM) Model

The following model was proposed by Dhakal and Maekawa (2002a,b) as shown in Figure 68.

Figure 68: Compressive stress-strain relationships (Dhakal and Maekawa, 2002a, b)

The DM model defines an intermediate point (εi, fi) to obtain the average compressive stress-strain curves.

After this intermediate point, a constant negative stiffness of 0.02Es is assumed until the average

compressive stress becomes equal to 0.2fy. In this formulation,

fy L
rb =   (6.4.1.1)
100 D

 ε y [55 − 2.3rb ]        


εi = for ε i ≥ 7ε y  ,    
otherwise ε i =
7ε y (6.4.1.2)

2
 ε −ε 
fit =      fu + ( f y − fu )  u i  for (ε sh <    
f y for ε i ≤ ε sh ; fit =     εi ≤εu ) (6.4.1.3)
 ε u − ε sh 
2
 ε −ε 
f y for ε sc ≤ ε sh
f st =      fu + ( f y − fu )  u sc  for (ε sh <    
; f st =     ε sc ≤ ε u ) (6.4.1.4)
 ε u − ε sh 
149
where, rb is slenderness ratio, fy is the yield stress in MPa, fit and fst are the stress in the tension curve

corresponding to the intermediate strain εi and the current strain εsc, respectively. At the intermediate

point, the stress fi on the compression curve is determined as follows:

 fitα [1.1 − 0.016rb ]        


fi = for fi ≥ 0.2 f y  ,    
otherwise fi =
0.2 f y (6.4.1.5)

The following equations are used for the determination of α:

ε u − ε sh f
α= 0.75 + for α ≤ u and
      0.75 ≤ α ≤ 1.0 (6.4.1.6)
300ε y 1.5 f y

The compressive stress-strain (fsc,εsc) relationship are then determined as follows:

= Esε sc for ε sc ≤ ε y
f sc          (6.4.1.7)

  f   ε sc − ε y 
f sc=  f st 1 − 1 − i   for ε y < ε sc ≤ ε i
          (6.4.1.8)
  fit   ε i − ε y  

f sc  max ( f i − 0.02 Es ( ε sc − ε i ) ) ;0.2 f y          


= for ε i < ε sc ≤ ε u (5.4.1.9)

6.4.2 Refined Dhakal-Maekawa (RDM) Model

Although the DM model provides good approximation in general sense to the buckling behaviors

of reinforcing bars, to improve the performance of the DM model for reinforcing bars with any geometric

and material properties, and to eliminate some criticized points on the DM model, a refined model (RDM)

was proposed by Akkaya et al. (2013). As the RDM model covers the DM model, the performance of the

RDM model is similar to or better than the performance of the DM model. By comparison with the DM

model, the intermediate stress, fi is determined more effectively as per the location of the intermediate

strain, εi on the tension curve in the RDM model. In particular, RDM model gives better results than the
150
DM model when the intermediate strain, εi, is on the strain-hardening region of the tension curve for

reinforcing bars with a high strain-hardening parameter (P=4) and small slenderness ratios (rb<21), and

for reinforcing bars with tension stress-strain response where εi<εu and εimax>εu. A typical average

compressive response obtained from this model is presented in Figure 69.

Figure 69: Compressive stress-strain relationships (Akkaya et al., 2013)

The compression curves are developed through the definition of an intermediate point (εi , fi) on the

average compression stress-strain curves. The intermediate strain εi and the maximum intermediate strain

εimax are determined as follows:

fy L
rb =   (6.4.2.1)
100 D

 fy 
ε i max =
ε y 55 − 2.3 × 5 (6.4.2.2)
 100 

βε y [55 − 2.3rb ] for  7


εi = ε i ≥ ε y , otherwise ε i =
7ε y (6.4.2.3)

where rb is slenderness ratio and fy is the yield stress in MPa. Maximum intermediate strain εimax is used to

define different tension curves using β and P constants as follow:

=β ε u ε i max for  (ε iβ =1 < ε u and ε i max > ε u ); otherwise


= β 1 (6.4.2.4)
151

P = 1 for ( ε i max ≥ ε u and ε i = 7ε y ) ; otherwise P = 4 (6.4.2.5)

P
 (ε − ε ) 
  + ( f y − fu )  u i  for ε i > ε sh
fit = f y for ε i ≤ ε sh ; fit = fu  (6.4.2.6)
 (ε − ε ) 
 u sh 

P
 ( ε − ε sc ) 
  + ( f y − fu )  u
f st = f y for ε sc ≤ ε sh ; f st = fu   for ( ε sh < ε sc ≤ ε i ) (6.4.2.7)
 ( ε − ε ) 
 u sh 

where, fit and fst are the stress in the tension curve corresponding to the intermediate strain εi and current

strain εsc, respectively. At the intermediate point, the stress fi on the compression curves is determined as

follow:

=fi   α f y ; fi   
≤ fit (6.4.2.8)

The following equations are used for the determination of α based on the location of εi on the tension

curve:

 fu D 
[1.1 − 0.016rb ] 0.8 + 1.8
α= for ε i > ε sh
         (6.4.2.9)
 f y L 

 fu D 
0.75 [1.1 − 0.016rb ] 0.8 + 1.8
α= for ε i ≤ ε sh
          (6.4.2.10)
 f y L 

fit
α = 0.75 [1.1 − 0.016rb ]       (
for ε i max ≥ ε u and ε i = 7ε y )  (6.4.2.11)
fy

The compressive stress-strain (fsc,εsc) response including rebar buckling is then determined as follows:

= Esε sc for ε sc ≤ ε y
f sc          (6.4.2.12)
152

  f   ε sc − ε y 
f sc=  f st 1 − 1 − i   for ε y < ε sc ≤ ε i
          (6.4.2.13)
  fit   ε i − ε y 

f sc  max ( f i − 0.02 Es ( ε sc − ε i ) ) ;0.2 f y    (


= for ε i      
≤ ε sc < ε ii ) (6.4.2.14)

= max (
f sc  0.75 f i − 0.01Es ( ε sc − ε ii ) ) ;0.2 f y    
for (ε ii      
≤ ε sc < ε u ) (6.4.2.15)

6.4.3 RDM and DM buckling models for hysteretic response

When the unsupported length ratio is equal to or larger than 5, the backbone compressive stress-strain

curve are defined by the selected RDM or DM models as shown in Figure 70. The stiffness (Em) at the

target point and the unloading-reloading stiffness (Erb) on the compressive strain region are also changed

due to buckling.

Figure 70: Hysteretic models for ductile reinforcement including buckling


[left (a): Seckin; right (b): Menegotto-Pinto

In the Seckin model, Figure 70a, the stifness (Em) at the target point on the compressive strain zone is

assumed to be zero when the compressive stress (fm) is smaller than yield stress (fy). In Menegotto-Pinto
153
model, Figure 70b, the same stifness (Em) at the target stress-strain point (fm, εm) on the compressive strain

zone are limited with a negative value as follows:

=Em
(f m − fy )
≥ −0.02 Es (6.4.3.1)
(ε m −εy )

For both hysteretic models, in Figure 70, the unloading-reloading stifness (Erb) on the compressive strain

zone are reduced as follow:

2
 f 
=Erb Es  m  for ε m− ≤ ε i (6.4.3.2)
 f st 

 f  f 
=Er Es  m   i  for ε m− > ε i (6.4.3.3)
 
 f st   f y 

6.4.4 Determination of the Unsupported Length Ratio (b/t=L/D)

The accuracy of reinforcement buckling models depends on defining the unsupported length ratio (b/t)

correctly. Because RC members have the numerous reinforcement configurations, there are difficulties in

determining b/t automatically. Consequently, b/t should be input manually. One method is proposed by

Dhakal and Maekawa (2002c) for this purpose as follows:

This method requires determining the reduced flexural rigidity (Er I) of a rebar as follows:

=I π=
D 4 / 64, Er I 0.5 Es I / ( f y / 400) (6.4.4.1)
154
where, I, D, Es, and fy are the moment of inertia, diameter, modulus of elasticity, and yield strength (MPa)

of the reinforcing bar including buckling, respectively. Normalized stiffness k of the rebar, tie stiffness kt,

and equivalent stiffness ratio keq are calculated as follows:

π 4 Er I Et At nl kt
k= , kt = , keq = (6.4.4.2)
s 3
le nb k

where s is the tie spacing, At is the sectional area of the one tie leg, le is the length of the tie leg, Et is the

modulus of elasticity of the tie, nb is the number of longitudinal bars supported by these tie legs, and nl is

the number of tie legs parallel to the lateral load. Dhakal and Maekawa (2002c) propose the values for nb

and nl for common arrangements of longitudinal and lateral reinforcement as shown in Figure 71.

Figure 71: Values of nb and nl for common reinforcement

The buckling length is then determined as follow:

L= n ⋅ s (6.4.4.3)

where, n is the number of spaces between ties over the buckling length which may be chosen from

Table 1.
155
Table 1: Determination of n (Dhakal and Maekawa 2002c)

L=n s
keq n
keq>0.75 1
0.7500-0.5000 1 or 2
0.5000-0.1649 2
0.1649-0.0976 3
0.0976-0.0448 4
0.0448-0.0084 5
0.0084-0.0063 6
0.0063-0.0037 7
0.0037-0.0031 8

In Table 2, the slenderness ratio, rb indicates approximately the reduction of normalized stress (fi/fy) on the

modified compressive stress-strain curve due to buckling. The calculated unsupported length ratio can be

used together with Table 2 when determining the unsupported length ratio. If the calculated value of rb is

smaller than 8, there is no rebar buckling effect. For other values, Akkaya et al. (2013) recommend the

followings:

1) If the calculated value of rb is between 8 and 16, assume rb is 16.

2) If the calculated value of rb is between 16 and 21, assume rb is 21.

3) If the calculated value of rb is larger than 21, use the calculated value of rb.

Table 2: Relationships between the buckling length and normalized intermediate stress ratio

L fy Buckling affect
rb =
D 100 fi/fy level
<8 - No effect
(8-16) (1.0-0.8) Small
(16-34) (0.8-0.4) High
(34-50) (0.4-0.2) Very high
156
6.4.5 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 Gf.

6.5 Steel Fibre Reinforced Concrete

The following sections describe the response of Steel Fibre Reinforced Concrete (SFRC). The

two types of fibres that can be used to model SFRC in VecTor2 are straight fibres and hooked fibres. The

main difference between the two types of fibres is in the quality of bond between the fibres and the

concrete matrix. The deformed ends of the hooked fibres enable a stronger bond between the steel fibres

and the concrete compared to the straight fibres.

6.5.1 Introduction

The effect of fibres on the behaviour of concrete is dependent on fibre volume content, fibre length, fibre

aspect ratio, fibre tensile strength, concrete strength, and fibre orientation. The majority of these are

required inputs in FormWorks. The following properties are considered when SFRC is analyzed using

VecTor2:

- Fibre Volume Fraction, 𝑉𝑓

- Fibre Length, 𝐿𝑓

- Fibre Diameter, 𝐷𝑓

- Fibre Tensile Strength, 𝐹𝑢

- Fibre Bond Strength, 𝑇𝑢


157
Note that the fibre aspect ratio mentioned above is the ratio of fibre length to fibre diameter, and is a

measure of the stiffness of the fibre. A brief description of the effect of the factors listed above on the

behaviour of SFRC can be found in (Susetyo, 2009).

6.5.2 Stress-Strain Behavior of SFRC

This section outlines the tensile and compressive stress-strain behaviour of SFRC.

An important function of steel fibres in SFRC is to provide post-cracking tensile capacity. In

SFRC, while the concrete stress reduces after cracking, the strain may still increase substantially prior to a

reduction in tensile strength. At higher fibre volumes, strain hardening may also occur (Susetyo, 2009).

Figure 72 illustrates the stress-strain response of both high and low fibre volume content concrete.

There are a variety of options available with SFRC. Steel fibre reinforcement can be used in

normal SFRC or as part of a composite material. The two SFRC composite materials available in

VecTor2 are Concrete-SFRC Laminate and Wood-SFRC Laminate. For both regular SFRC and the

SFRC used in composite material models, either straight or hooked fibres can be used. The required input

properties for steel fibres are the same for types of fibres.

There are also a number of models available in VecTor2 for SFRC. For the compression stress-

strain response of SFRC, the Lee et al 2011 (FRC) model is available and is described in Section 4.2.7.

Two tension softening models for FRC are also available in VecTor2. An exponential tension softening

model and the fib FRC tension softening models are discussed in sections 4.5.6 and 4.5.9, respectively.
158
Numerous FRC tension models are also available and are described in section 4.6.

Figure 72: Tensile stress-strain response of FRC for low and high fibre volume contents
(Susetyo, 2009)

The addition of fibre to concrete makes the post-peak compressive response more gradual, less

steep compared to plain concrete. The higher the aspect ratio or volume content of the fibres, the more

gradual the post-peak compressive response. The compressive response of SFRC is illustrated in Figure

73.
159

Figure 73: Compressive stress-strain curve of SFRC, illustrating


a) fibre content and b) fibre aspect ratio (Susetyo, 2009)

6.6 Laminates

Four different types of laminate can be modeled in VecTor2. The most common is the concrete-

steel laminate, where a concrete core is located between two steel face plates.

The Steel Skin Plate option is used as the reference reinforcement type in the Define Reinforced

Concrete Properties tab when modeling concrete-steel composite elements. The steel plate is analyzed in

accordance with the DSFM, where steel stresses are determined from uniaxial stress-strain relationships.

For the monotonic response of steel skin plates, VecTor2 employs a tri-linear elastic-plastic strain

hardening model. Yielding, strain hardening, and Bauschinger’s effect for reverse cyclic loading are all

considered in VecTor2, and yielding of the plates under biaxial stress is determined using the Von Mises

criterion:

2
(𝑓𝑠1 − 𝑓𝑠2 )2 + (𝑓𝑠2 − 𝑓𝑠3 )2 + (𝑓𝑠3 − 𝑓𝑠1 )2 = 2�𝑓𝑦 � (6.6.1)

where fs1 , fs2 , and fs3 are the principal stresses in the plate, one of which is always zero.

The hysteretic response of the steel plate, which includes the Bauschinger effect as mentioned

above, is modeled using a modified Seckin formulation. After the steel plate buckles in compression, it

can sustain additional loading in tension but cannot sustain further compressive stress upon load reversal.

It is assumed in VecTor2 that the anchor studs prevent slip between the concrete and steel skin

plates. A more detailed response of the FEM formulation of steel-concrete elements is found in Vecchio

and McQuade (2011).

SFRC can also be used as a laminate; both SFRC—straight fibres and SFRC—hooked fibres are

possible laminate options. The stress-strain behaviour of SFRC was discussed in the preceding section.

Orthotropic laminate can also be used as smeared reinforcement. The stress-strain behaviour of

orthotropic wood was discussed in section 5.3.


160
6.7 Shape Memory Alloy Type 1 and Type 2

Shape Memory Alloy materials are used to replace conventional reinforcing steel under seismic

loading conditions, and are useful due to the material’s ability to dissipate large amounts of energy

without excessive permanent deformation. The hysteresis for conventional reinforcing steel includes

large strain offsets; after an earthquake, the structure may have large permanent displacements. Shape

memory alloys minimize or eliminate these large strain offsets such that after a seismic event, the

structure will retain its original shape or the deformations will be much smaller than if conventional

reinforcing steel was used.

6.7.1 Shape Memory Alloy 1 (SMA 1)

The idealized behaviour of the SMA 1, with no strain offsets is modeled with SMA 1. It has a

flag-shaped hysteresis, as shown in Figure 74.

Figure 74: Stress-strain response for SMA 1


161
6.7.2 Shape Memory Alloy 2 (SMA 2)

Developed at the University of Ottawa, the hysteresis for SMA 2 differs from SMA 1 in that it

incorporates strain hardening as well as small strain offsets. Figure 75 illustrates the stress-strain curve

and hysteresis for SMA 2.

Figure 75: Stress-strain response for SMA2

In the figure above, 𝑓𝑦 is the yield stress, 𝑓𝑢𝑛𝑙 is the unloading stress, 𝜀𝑝 is the strain offset, 𝜀𝑟1

and 𝜀𝑟2 are reference strains, and 𝜀𝑚 is the maximum strain.

Both the unloading stress and strain offset are functions of the maximum strain.
2)
𝑓𝑢𝑛𝑙 = 𝑓𝑦 ∙ (0.70188 − 0.003127𝜀𝑚 − 0.00003214𝜀𝑚 0.25𝑓𝑦 ≤ 𝑓𝑢𝑛𝑙 ≤ 0.75𝑓𝑦 (6.7.2.1)

2
𝜀𝑝 = 0.001𝜀𝑚 − 0.005𝜀𝑚 + 0.4323 (6.7.2.2)
162

7 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.

7.1 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, τ, and bond slips, ∆. Those pertaining to pullout failure are subscripted as τp, ∆p while those

pertaining to splitting failure are subscripted, τs, ∆s.

The actual bond stress-slip model is defined by a series of reference bond stresses and slips,

τsp, ∆sp. These are determined by linearly interpolating between the unconfined and confined reference

bond stresses and slips using the confinement pressure factor, β. 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, σ, the

confinement pressure factor may be computed as follows:

σ
β= , (in MPa) 0 ≤ β ≤ 1, (7.1.1)
7.5

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
163
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, db, as follows:

39.4 mm, 2.55 mm for 55 mm ≤ d b


30.6 mm, 2.20 mm for 43 mm ≤ d < 55 mm
 b
25.0 mm , 1.79 mm for 35 mm ≤ d b < 43 mm

20.9 mm , 1.48 mm for 29 mm ≤ d b < 35 mm
 (7.1.2)
S , H = 17.6 mm, 1.26 mm for 25 mm ≤ d b < 29 mm
13.6 mm, 0.98 mm for 19 mm ≤ d < 25 mm
 b
11.2 mm, 0.72 mm for 15 mm ≤ d b < 19 mm

 7.9 mm, 0.45 mm for 11 mm ≤ d b < 15 mm

 0.70d b , 0.04d b for 0 mm ≤ d b < 11 mm

For smooth bars, VecTor2 computes the lug spacing according to Equation 7.1.2, but computes the lug

height as follows:

S
H= (7.1.3)
75

The model for hooked bars is different is described subsequently.

7.1.1 Perfect bond

The bond material is assigned a numerically large stiffness and strength to prevent deformation of

the bond element.

7.1.2 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 76.

The confined stress-slip relationship is summarized as follows:

τ p1

(∆ ∆ ) p1
α
for ∆ ≤ ∆ p1
τ for ∆ p1 < ∆ ≤ ∆ p 2
 p2
τ= (
 ∆ − ∆ p2 )  (7.1.2.1)
τ p 2

-
(
 ∆ p 3 − ∆ p 2 )( )
τ p 2 − τ pf 

for ∆ p 2 < ∆ ≤ ∆ p 3
τ for ∆ p 3 < ∆
 pf
164
where

 d  f 'c
τ p1 =  20 − b  (7.1.2.2)
 4  30

τ p 2 = τ p1 (7.1.2.3)

 S  f 'c
τ pf =  5.5 − 0.07  (7.1.2.4)
 H  27.6

f 'c
∆ p1 = (7.1.2.5)
30

∆ p 2 = 3.0 mm (7.1.2.6)

∆ p3 = S (7.1.2.7)

α = 0.4 (7.1.2.8)

The unconfined stress-slip relationship is summarized as follows:

τ s1 (∆ ∆ s1 )α for ∆ ≤ ∆ s1

τ for ∆ s1 < ∆ ≤ ∆ s 2
 s2
τ=  (∆ − ∆ s 2 )  (7.1.2.9)
τ s 2 -  ( )
τ s 2 − τ sf  for ∆ s 2 < ∆ ≤ ∆ s 3
  (∆ s 3 − ∆ s 2 ) 
 τ
 sf for ∆ s 3 < ∆

where

f ' c ⋅c
τ s1 = 0.748 ≤ τ p1 (7.1.2.10)
db

τ s 2 = τ s1 (7.1.2.11)

f ' c ⋅c
τ sf = 0.234 ≤ τ pf (7.1.2.12)
db

 1  τ 
∆ s1 = ∆ p1 exp  ln s1  (7.1.2.13)
 α  τ p1 

∆ s2 = ∆ p2 (7.1.2.14)

∆ s3 = ∆ p3 (7.1.2.15)
165
Given a confinement pressure factor, β, the bond stress-slip relationship is defined as follows:

(
τ sp1 ∆ ∆ sp1 α

) for ∆ ≤ ∆ sp1

 (
 ∆ − ∆ sp12 ) 
τ sp1 - 
 (
 ∆ sp 3 − ∆ sp 2 )(
τ sp 2 − τ spf )

for ∆ sp1 < ∆ ≤ ∆ sp 2
τ= (7.1.2.16)
 (
 ∆ − ∆ sp 2 ) 
τ sp 2 - 
 (
 ∆ sp 3 − ∆ sp 2 )(
τ sp 2 − τ spf )

for ∆ sp 2 < ∆ ≤ ∆ sp 3


τ for ∆ sp 3 < ∆
 spf

(
τ sp1 = τ s1 + β τ p1 − τ s1 ) (7.1.2.17)

τ sp 2 = τ sp1 (7.1.2.18)

(
τ spf = τ s1 + β τ pf − τ sf ) (7.1.2.19)

(
∆ sp1 = ∆ s1 + β ∆ p1 − ∆ s1 ≥ ∆ s1 ) (7.1.2.20)

∆ sp 2 = ∆ p 2 (7.1.2.21)

∆ sp 3 = ∆ p 3 (7.1.2.22)

β=1

β=0

Figure 76: Eligehausen bond stress-slip response


166
7.1.3 Gan Model

The Gan confined bond stress-slip relationship, as shown in Figure 77, is the same as the

Eligehausen model for the confined stress-slip relationship as summarized by equations 7.1.2.1 to 7.1.2.8.

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:

τ s1 (∆ ∆ s1 )α for ∆ ≤ ∆ s1

 (∆ − ∆ s1 ) 

τ = τ s1 -  (
τ s1 − τ sf ) for ∆ s1 < ∆ ≤ ∆ s 2 (7.1.3.1)
  (∆ s 2 − ∆ s1 ) 
 τ for ∆ s 2 < ∆
 sf

where

f ' c ⋅c
τ s1 = 0.748 ≤ τ p1 (7.1.3.2)
db

τ s 2 = 0.15τ s1 (7.1.3.3)

τ sf = τ s 2 (7.1.3.4)

 1  τ 
∆ s1 = ∆ p1 exp  ln s1  (7.1.4.5)
 α  τ p1 

∆ s 2 = 2.0 mm (7.1.3.6)

∆ s3 = ∆ p3 (7.1.3.7)

Given a confinement pressure factor, β, the bond stress-slip relationship is defined by equation

7.1.2.16, where:

(
τ sp1 = τ s1 + β τ p1 − τ s1 ) (7.1.3.8)

(
τ sp 2 = τ s 2 + β τ p 2 − τ s 2 ) (7.1.3.9)

( )
τ spf = τ sf + β τ pf − τ sf < τ sf (7.1.3.10)

(
∆ sp1 = ∆ s1 + β ∆ p1 − ∆ s1 ≥ ∆ s1 ) (7.1.3.11)

2.0 mm if β = 0
∆ sp 2 =  (7.1.3.12)
∆ p 2 if 0 < β
167

2.0 mm if β = 0
∆ sp 3 =  (7.1.3.13)
∆ p 3 if 0 < β

β=1

β=0

Figure 77: Gan bond stress-slip response

7.1.4 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 78.

The confined stress-slip relationship is summarized as follows:

τ p1

(∆ ∆ ) p1
α
for ∆ ≤ ∆ p1
τ for ∆ p1 < ∆ ≤ ∆ p 2
 p2
τ= (
 ∆ − ∆ p2 )  (7.1.4.1)
τ p 2

-
(
 ∆ p 3 − ∆ p 2 )(
τ p 2 − τ pf )

for ∆ p 2 < ∆ ≤ ∆ p 3
τ for ∆ p 3 < ∆
 pf

where

τ p1 = 2.575 f ' c (7.1.4.2)

τ p 2 = τ p1 (7.1.4.3)
168
τ pf = 0.35τ p1 (7.1.4.4)

∆ p1 = 0.75(0.189 S + 0.18) (7.1.4.5)

∆ p 2 = 1.75(0.189 S + 0.18) (7.1.4.6)

∆ p3 = S (7.1.4.7)

α = 0.3 (7.1.4.8)

The unconfined stress-slip relationship is summarized as follows:

τ s1 (∆ ∆ s1 )
 α
for ∆ ≤ ∆ s1
τ= (7.1.4.9)

0 for ∆ s1 < ∆

where

 c
τ s1 =  0.249 + 0.291  f ' c ≤ τ p1 (7.1.4.10)
 d

τ s2 = 0 (7.1.4.11)

τ sf = 0 (7.1.4.12)

 1  τ 
∆ s1 = ∆ p1 exp  ln s1  (7.1.4.13)
 α  τ p1 

∆ s2 = ∆ p2 (7.1.4.14)

∆ s3 = ∆ p3 (7.1.4.15)

Given a confinement pressure factor, β, the bond stress-slip relationship is defined by equation

7.1.2.16 where:

(
τ sp1 = τ s1 + β τ p1 − τ s1 ) (7.1.4.16)

τ sp 2 = βτ p 2 (7.1.4.17)

τ spf = βτ pf (7.1.4.18)

(
∆ sp1 = ∆ s1 + β ∆ p1 − ∆ s1 ≥ ∆ s1) (7.1.4.19)

3.0 mm if β = 0
∆ sp 2 =  (7.1.4.20)
∆ p 2 if 0 < β
169

3.0 mm if β = 0
∆ sp 3 =  (7.1.4.21)
∆ p 3 if 0 < β

β=1
β=0

Figure 78: Harajli bond stress-slip response

7.1.5 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:
170

τ1 (∆ ∆ s1 ) for ∆ ≤ ∆ 1
τ for ∆ 1 < ∆ ≤ ∆ 2
 2
τ= (∆ − ∆ 2 )
(
τ 2 - (∆ − ∆ ) τ 2 - τ f
) for ∆ 2 < ∆ ≤ ∆ 3
(7.1.6.1)
3 2

τ
 f for ∆ 3 < ∆

In the case of β = 0 , the reference bond stress and slips are determined as follows:

τ1 = (Bi + 1.5) f 'c 8 (7.1.6.2)

τ 2 = τ1 (7.1.6.3)

τ f = 0.1 MPa (7.1.6.4)

τ1
∆1 = (7.1.6.5)
100

∆ 2 = 0.04 Bi − ∆ 1 (7.1.6.6)

∆ 3 = ∆1 2 0 + ∆ 2 (7.1.6.7)

Bi = min (Bis, Bic ) (7.1.6.8)

In the case of β ≠ 0 , the reference bond stress and slips are determined as follows:

0.313(0.307 Bi + 0.427 + 24.9Q ) f ' c for Q < 0.0035


τ1 =  (7.1.6.9)
0.313(0.307 Bi + 1.3) f ' c for Q > 0.0035

τ 2 = τ1 (7.1.6.10)

0.313(23.3Q + 0.3) f ' c for Q < 0.0035


τf =  (7.1.6.11)
0.35 f ' c for Q > 0.0035

τ1
∆1 = (7.1.6.12)
100

20.7Q + 0.53 − ∆ 1 2 for Q < 0.0035


∆2 =  (7.1.6.13)
1.25 − ∆ 1 2 for Q > 0.0035

∆ 3 = ∆1 2 + ∆ 2 (7.1.6.14)

Q = Q *β (Bis + 1) 88 (7.1.6.15)

1 for Bis < Bic


Q* =  (7.1.6.16)
 2 for Bic < Bis
171
7.1.6 Eligehausen and Gan Models with No Cyclic Damage

Figure 79 shows the analytical model developed by Eligehausen et al from cyclic loading tests.

Figure 79: Eligehausen cyclic bond stress-slip relationship

The initial loading curve is monotonic, where the bond element stiffness is calculated as follows

(Gan, 2000):

𝐺𝑛 = 𝐴𝑘 𝐺0 for linkage elements (7.1.6.1)

𝐺𝑛 = 𝐺0 for contact elements (7.1.6.2)

Where,

𝐺0 = 𝑑𝜏/𝑑𝛥 = 𝜏1 𝛼 ∙ 1001−𝛼 𝑎𝑡 𝛥/𝛥1 = 1/100 (7.1.6.3)

For all other loading increments, the stiffness is calculated as:

𝐺𝑛 = 𝐴𝑘 𝜏/𝛥 for linkage elements (7.1.6.4)

𝐺𝑛 = 𝜏/𝛥 for contact elements (7.1.6.5)

Unloading occurs with constant bond stiffness, which is the same bond stiffness used during

reloading. Unloading is shown by the curve ADE. The stress after unloading is related to the maximum

bond stress reached at the end of the previous loading, τpm , as follows:

𝜏𝑢𝑛 = −0.25𝜏𝑝𝑚 (7.1.6.6)


172
The damage index, D, is used to calculate the reduction in bond resistance as both the slip and

number of cycles increases. The damage index is equal to zero for no cyclic damage, and equal to one

when τ = 0 and bond has broken down entirely. Otherwise, D is calculated as follows.

𝐷𝑖 = 1 − 𝑒 𝑎 (7.1.6.7)

𝑎 = −1.2(𝐸𝑖 /𝐸0 )1.1 (7.1.6.8)

where 𝐸𝑖 is the hysteretic energy dissipation at the current unloading stage and 𝐸0 is the energy

equal to the area under the stress slip curve from the monotonic stage to Δ3 .

The application of the damage index, D, to different types of failure including pull-out, splitting

without confinement, and splitting with partial confinement is not outlined in this manual, but is discussed

fully in (Gan, 2000). The equations used to calculate the energy 𝐸𝑖 and 𝐸0 are also discussed there.

7.2 Bond Stress-Slip Models for Externally Bonded Plates or Sheets

7.2.1 Perfect bond

Bond materials are assigned a large stiffness and strength to prevent deformation bond elements.

7.2.2 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:

τ1 (∆ ∆ s1 ) for ∆ ≤ ∆ 1

 (∆ − ∆1 ) (
 τ1 + τ -τ ) for ∆ 1 < ∆ ≤ ∆ 2
 (∆ 2 − ∆1 ) 2 1
τ= (7.2.2.1)
τ + (∆ − ∆ 2 ) (τ - τ ) for ∆ 2 < ∆ ≤ ∆ 3
 2 (∆ 3 − ∆ 2 ) 3 2


τ 3 for ∆ 3 < ∆

The shape of the bond stress-slip relationship, as shown in Figure 80, may be modified by specifying the

values of the reference bond stresses, τ1, τ2, and τ3 and corresponding slips, ∆1<∆2< ∆3.
173

τ1
τ2

τ3


∆1 ∆2 ∆3

Figure 80: Bond stress-slip response for externally bonded plates or sheets
174

8 Part II: FormWorks

8.1 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.

8.2 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.

 JobOpt.fwd: the FormWorks job options data file.

 StrOpt.fwd: the FormWorks structure options data file.

 LoadOpt.fwd: the FormWorks load options data file.

 FWK.ico: the FormWorks file icon.

 Vt2.exe: the executable VecTor2 program.

 DOS4GW.exe: an auxiliary executable program required by VecTor2.


175
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 FWK 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 FWK.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.

8.3 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.
176
8.3.1 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.

8.3.2 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 81.

User input FormWorks

.JOB .S2R .L2R .L2R .L2R .L2R .L2R .A2R


~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~
~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~

VecTor2
Each load stage

.S2E .L2E .A2E .A2R


~~~~~ ~~~~~ ~~~~~ ~~~~~
~~~~~ ~~~~~ ~~~~~ ~~~~~

Figure 81: Input and output files forVecTor2 analysis


177
FormWorks receives user input and generates the following input files for VecTor2:

 Job Data File (*JOB)

VecTor2 reads one Job 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 Job 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.
178
 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.

8.4 FormWorks Basics

8.4.1 The FormWorks Interface

Upon starting FormWorks, the FormWorks application window appears as shown in Figure 82.

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
179
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 82: FormWorks application window

8.4.2 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 83. The FormWorks window is titled

FormWorks-Workspace2, indicating that the second Workspace is currently active.


180

Figure 83: Creating new Workspace windows

8.4.3 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 .FWK 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.
181
8.4.4 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 .FWK extension.

3. Click Open.

Alternatively, select the File menu and select from the four most recent Workspace files.

8.5 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.

8.5.1 Manipulating the View

8.5.1.1 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.

8.5.1.2 Changing Display Limits

To specify the portion of x,y plane that is visible in the Workspace, complete the following steps.
182
1. Select the View/Limits… menu item. The Display Limits dialog appears as shown in Figure 84.

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 84: Display Limits dialog box

8.5.1.3 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


183
 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.

8.5.1.4 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 85.

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


184
 Nodal Thermal Loads

Select to reveal nodal thermal loads for the current load case

 None

Select to hide the above load types for the current load case.

Figure 85: 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.


185
 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

Select to reveal the density in kg/m3 and G-forces applied to concrete elements in the load case.

 Element Temperature

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.

 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.
186
8.5.2 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.


187

9 The Job Data

The first step in creating the VecTor2 input is to define the Job Data. At the time of analysis,

FormWorks generates the *JOB file based on the defined Job Data.

9.1 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 86.

Figure 86: Job Control property page

2. Input the job data as described in the subsequent sections.


188
3. When done, select the Models page.

4. When done selecting appropriate models, select the Auxiliary page.

9.1.1 Job Data Group

These entry fields manage the creation of the *JOB 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 *JOB extension when saving the Job 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.

9.1.2 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.

9.1.3 Load Data Group

9.1.3.1 Static Analysis

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

structure, different load cases can have different load factors. For instance, lateral loads defined in one
189
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 Figure 87, Figure 88 and Figure 89, respectively.

Load Factor
30
25 Starting Load Stage No.: 1
No. of Load Stages: 26
20
Initial Factor: 0.
15
Final Factor: 25.
10
Inc. Factor: 1
5
Load Type: Monotonic
0
1 6 11 16 21 26
Load Stage No.

Figure 87: Monotonic type loading

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

Figure 88: Cyclic type loading


190

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

Figure 89: 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.
191
 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.


192
 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:

 LF f − LFi
 +1 for monotonic loading
 LS inc


  LF f − LFi   R ⋅ C inc 
No. Stages = 2(R ⋅ S ) + [S (S − 1)] + 1 for cyclic loading (9.1.3.1)
  LS
  LS inc   inc 

  LF f − LFi   2 R ⋅ C inc 
4(R ⋅ S ) + [S (S − 1)] + 1 for reversed cyclic loading
  LS

  LS inc   inc 

where LFi is the initial load factor, LFf is the final load factor, LSinc 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 Cinc is the cyclic

load factor increment.

9.1.3.2 Dynamic Analysis

To create a load case for a dynamic analysis:

The time-step size for the analysis is specified in the first load case. Therefore, the first load case should

contain the dynamic mass information.

1. Check the load Case box to activate the load case, as in a typical analysis. Note that the dynamic

information must be contained in the first load case.

2. Fill in the following fields:

 Load File Name

 Load Case Title


193
 Initial Factor: the initial factor coefficient is ignored for the first load case in dynamic analysis

(enter it as 1)

 Final Factor: the final factor is also ignored in the first load case (enter it as 1)

 Inc. Factor: this is the time increment between two recorded output files

 Load Type: The load type must be set to Cyclic in order to activate the Repetitions input

 Repetitions: Specify the number of time divisions between two output files. The time-step that

is actually used in the numerical integration for dynamic analysis is the Inc. Factor/Repetitions

As mentioned above, the initial and final load factors are ignored for the first load case in dynamic

analyses. However, if static loads are also present in the first load case file, the load factor for those static

loads is taken as 1.0.

In addition, selecting the proper time-step size for the analysis is very important, as it affects both

the accuracy and the stability of the solution. One way to determine if the time-step is sufficiently small

is to perform the analysis with a smaller time-step size and compare the two results. If the results are the

same, then the time-step size is appropriately small. If not, then a smaller time-step size should be used.

9.1.4 Analysis Parameters Group

This group controls the progress of the iterative solution procedure and the analysis output:

 Analysis Mode

There are five options for the analysis mode. The Linear Elastic option is the most basic

analysis mode. The Static Nonlinear-Load Step option is chosen for most static analyses where

the load is increasing in stages. The Static Nonlinear-Time Step option is used when modeling

time-varying loads such as surface thermal loads. The Dynamic Nonlinear-General option is

used for impact and impulse load analyses, including ground acceleration. Note that if this option

is used, the ground acceleration must be applied by selecting the “Considered” option in the

Apply Ground Acceleration field of the Auxiliary tab of the Define Job window. The Dynamic

Nonlinear-EQ Record option is used when the ground acceleration is specified by an earthquake

record, in the form of a VECTOR.EQR file.


194
 Seed File Name

Enter NULL if no seed file is used. Otherwise, enter the file name of the *A2R file.

 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.

9.2 The Model Page

1. Select the Models property page as shown in Figure 90.

2. Select the desired material and behavioral models. The following sections provide

abbreviated descriptions of the models.


195
3. When done, click Ok to store changes to the Job Data.

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 90: The Models page

9.2.1 Concrete Models

 Compression Pre-Peak Response

Select the ascending branch of the average concrete compression stress-strain response.

 Compression Post-Peak Response


196
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.

 FRC Tension

Select the FRC Tension model, if applicable.

 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 Poisson’s 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 Stress Calc

Select the model for calculating the shear stress at a crack.

 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.

 Crack Slip Calc

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
197
Theory. If any other option is selected, VecTor2 conducts the analysis based on the Disturbed

Stress Field Model.

 Creep and Relaxation (currently not available)

 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.

9.2.2 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.)

9.2.3 Bond Models

 Concrete Bond

Select the bond stress-slip relationship of between concrete and embedded reinforcing bars.

9.2.4 Analysis Models

 Strain History:

The previous loading should be considered for analyses involving cyclic and reversed cyclic

loading to include the hysteretic response models.


198
 Strain Rate Effects:

Strain rate effects may be considered when performing dynamic analyses. Strain rate effects

increase the dynamic strength of materials, and are discussed in the VecTor2 section of this

manual.

 Structural Damping

Damping is used for dynamic analyses; VecTor2 offers two types of damping, Rayleigh and

Alternative.

 Geometric Nonlinearity

 Crack Process

Two different crack allocation models are available. Uniform crack allocation is suitable for most

modeling purposes; the variable crack allocation model may be used where members are

reinforced with FRP sheets.


199
9.3 The Auxiliary Input Page

1. Select the Auxiliary input page as shown in Figure 91.

Figure 91: The Auxiliary Input Page

2. Enter applicable analysis parameters.

3. When done, click Ok to store changes to the Job Data.

For the majority of analyses, it is advisable to select the VecTor 2 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.


200
9.3.1 General

 Stiffness Matrix Solver

Select the solver to use, Solver 1 or Solver 2.

Default value: Solver 1

 Quadrilateral Element Type

Select which type of quadrilateral element to use in the mesh. The two options for quadrilateral

element type are isoparametric and degenerate. Degenerate elements allow for a mixture of both

quadrilateral and triangular elements and are convenient for structures with complex geometries.

Default value: Isoparametric

 Concrete Aggregate Type

Select the concrete aggregate type. The two options are carbonate and siliceous; the difference

between the two aggregate types is their response to thermal loading.

Default value: Carbonate

 Concrete Conductivity

Enter the concrete conductivity.

Default value: 2.19 W/mK.

 Concrete Fracture Energy

Enter the concrete fracture energy.

Default value: 0.75 kN/m.

 Pre-stressing Friction Coefficient

Enter the pre-stressing friction coefficient.

Default value: 0.30

 Pre-stressing Wobble Coefficient

Enter the pre-stressing wobble coefficient.

Default value: 0.0025/m.

 Thermal Time Stepping Factor


201
Enter the thermal time stepping factor. This factor determines the time step used in the Crank-

Nicolson Method. In general, the accuracy of the result will decrease with increasing time step

size.

Default value: 2/3 or 0.666667.

9.3.2 Dynamic Analysis Options

 Newmark Beta Factor

Enter the Newmark Beta Factor to be used in the analysis. As described before, the default value

in VecTor2 of β = 0.25 defines the constant acceleration method for Newmark’s method of

direct integration. For the linear acceleration method, β = 1/6 can be used, and caution is

advised when using any other value for β.

Default value: 0.25

 Newmark Gamma Factor

Enter the Newmark Gamma Factor.

Default value: γ = 0.5.

 Reference Mode #1

Enter reference mode #1, to be used in Rayleigh damping.

Default value: 1

 Reference Mode #2

Enter reference mode #2, for Rayleigh damping.

Default value: 2

 Damping Factor #1

Enter the damping factor assigned to Reference Mode #1.

Default value: 0 (no damping)

 Damping Factor #2

Enter the damping factor assigned to Reference Mode#2.

Default value: 0 (no damping)


202
 Ground Acceleration in x-direction

Choose to Consider or Not Consider the ground acceleration in the x-direction. The Considered

option is to be chosen when the Analysis Mode on the Job Control page is set to Dynamic

Nonlinear—General.

 Ground Acceleration in y-direction

Choose to Consider or Not Consider the ground acceleration in the y-direction. The Considered

option is to be chosen when the Analysis Mode on the Job Control page is set to Dynamic

Nonlinear—General.

9.3.3 Tension Softening Options

By specifying tension softening points, a custom tension softening curve can be developed.

 Tension Softening Pt 1: Strain

Enter the strain, expressed in units of millistrain, for the first tension softening point

Default value: 0

 Tension Softening Pt 1: Stress

Enter the stress, expressed in MPa, for the first tension softening point.

Default value: 0

 Tension Softening Pt 2: Strain

Enter the strain for the second point on the tension softening curve.

Default value: 0.5

 Tension Softening Pt2: Stress

Enter the stress for the second point on the tension softening curve.

Default value: 2 MPa.

 Tension Softening Pt 3: Strain

Enter the strain for the third point on the tension softening curve.

Default value: 1

 Tension Softening Pt 3: Stress


203
Enter the stress for the third point on the tension softening curve.

Default value: 1 MPa

 Tension Softening Pt 4: Strain

Enter the strain for the fourth point on the tension softening curve.

Default value: 2

 Tension Softening Pt 4: Stress

Enter the stress for the fourth point on the tension softening curve.

Default value: 0.1 MPa

9.3.4 Masonry Structures Data

 Principal direction wrt x-axis

In degrees, enter the direction of Joint 1, as shown on the figure below. Once the direction of

Joint 1 is set, Joint 2 is considered to be perpendicular to it. The default value of the principal

direction in VecTor2 is 0 degrees, meaning that it is assumed to be aligned with the x-axis.

Default value: 0 degrees

Figure 92: Masonry joints


204
 Masonry Joint 1: Thickness

Enter the thickness of Joint 1, th1, in millimeters.

Default value: 10 mm.

 Masonry Joint 2: Thickness

Enter the thickness of Joint 2, th2, in millimeters.

Default value: 10 mm.

 Joint Shear Strength Ratio

Enter the Joint Shear Strength Ratio, c/fmy, for the masonry. This is the ratio between the shear

strength of the joints,c, and the maximum compressive strength, fmy .

Default value: 0.01

Figure 93: Masonry joint shear strength ratio

 Masonry Strength Ratio

Enter the masonry strength ratio, fmx /fmy. This defines the ratio between the minimum

compressive strength of masonry, fmx , and the maximum compressive strength of masonry, fmy .

Default value: 0.5

 Elastic Modulus Ratio


205
Enter the elastic modulus ratio, Emy /Emx. This defines the ratio between maximum initial elastic

modulus, Emy , and the minimum initial elastic modulus, Emx in each direction of orthotropy. As

Default value: 0.5

 Friction Angle

Enter the friction angle, φ, of the joints in degrees. The friction angle is illustrated in the Joint

Shear Strength Ratio subsection.

Default value: 37 degrees

 Brick Strength Ratio

Enter the brick strength ratio. This defines the ratio between the tensile strength of masonry in

the x-direction, evaluated when α = 0, and the compressive strength, fmy .

Default value: 0.1

 Strength Reduction Factor

Enter the shear strength reduction factor, µ. This is the reduction factor for the fmx strength,

applicable in cases of uniaxial compression. The reduction factor, µ, accounts for the decrease in

stress parallel to the bed joints under uniaxial compression.

Default value: 1

9.3.5 Material Resistance/Creep Factors

 Concrete Resistance Factor

Enter the concrete resistance factor.

Default value: 1

 Rebar Steel Resistance Factor

Enter the rebar steel resistance factor.

Default value: 1

 P/S Steel Resistance Factor

Enter the pre-stressing steel resistance factor.

Default value: 1
206
 Structural Steel Resistance Factor

Enter the structural steel resistance factor.

Default value: 1

 Masonry/Mortar Resistance Factor

Enter the masonry/mortar resistance factor.

Default value: 1

 Wood/Ortho Resistance Factor

Enter the wood resistance factor.

Default value: 1

 Concrete Creep Coefficient

Enter the creep coefficient of concrete.

Default value: 0 (no creep)

 P/S Relaxation Coefficient

Enter the relaxation coefficient for the pre-stressing steel.

Default value: 0 (no relaxation)


207

10 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.

10.1 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.
208
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:

Bandwidth ≈ max{2[2(i − j ) − R ] − 1} (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.

10.2 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 94.

Figure 94: Structure Limits dialog box


209

10.3 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 95. These

values are updated as the model is constructed.

Figure 95: Structure Information dialog box

10.4 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.
210
10.4.1 Reinforced Concrete Material Types

In VecTor2 and FormWorks, reinforced concrete materials encompass reinforced concrete, as

well as other materials that are modeled with or without smeared reinforcement components or laminate

components. Reinforced concrete materials are applied to rectangular, quadrilateral, or triangular

elements.

Although referred to as Reinforced Concrete, a wide variety of materials can be analyzed in

VecTor2 and modeled in FormWorks. The Reinforced Concrete materials are:

- Reinforced concrete

- Structural steel

- Masonry

- Wood (fixed orthotropic)

With those four reinforced concrete materials, by choosing the appropriate type of smeared reinforcement,

the user is able to model the following composite materials:

- Concrete-Steel Laminate

- Concrete-SFRC Laminate

- Masonry-SFRC Laminate

- Concrete-Ortho Laminate

The reinforcement components that are available to be used with the reinforced concrete materials are:

- Ductile steel reinforcement

- Prestressing steel

- Tension only reinforcement

- Compression only reinforcement

- External bonded FRP Fabric

- Steel-fibre—hooked

- Steel-fibre—straight

- Steel skin plate

- SFRC laminate—hooked fibre


211
- SFRC laminate—straight fibre

- Orthotropic laminate

- Shape Memory Alloy Type 1 and Type 2

Brief descriptions of how to model the above concrete materials and reinforcement components in

FormWorks are included in the sections 10.4.1.1-10.4.1.5.

10.4.1.1 Reinforced Concrete

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 96.

Figure 96: Reinforced Concrete Materials Properties Dialog Box


212
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.

 Cylinder Compressive Strength, f’c:

Enter the peak compressive stress of a standard concrete test cylinder, in MPa.

 Tensile Strength, f’t:

Enter the uniaxial cracking strength of concrete in MPa.

Default value: f' t = 0.33 f ' c MPa.

 Initial Tangent Modulus, Ec:

Enter the tangent stiffness of the concrete stress-strain response at zero-strain, in MPa.

Default value: E c = 5500 f ' c MPa.

 Cylinder Strain at f’c, eo:

Enter the compressive strain, εo, corresponding to f’c, as a positive value in millistrain.

Default value: ε o = 1.8 + 0.0075 f ' c millistrain.

 Poisson’s Ratio, Mu:

Enter the initial Poisson’s ratio, vo, of the concrete as a positive value.

Default value: 0.15.

 Thermal Expansion Coefficient, Cc:

Enter the concrete strain increase per temperature increase of 1°C.

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/m3.


213
3
Default value: 2400 kg/m .

 Thermal Diffusivity, Kc:

Default value: 1.20 s/mm2.

 Average Crack Spacing:

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. If “0” is specified, the maximum crack

spacing is taken as 1000 mm. If the evaluated crack spacing is larger than 1000 mm, 1000

mm is used to calculate the crack width. This value can be reduced by specifying a negative

number for crack spacing “-NUM”; in this case, the maximum crack spacing is NUM.

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.
214
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.

 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


215
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, Fy/Es.

 Thermal Expansion Coefficient, Cs

Enter the reinforcement strain increase per temperature increase of 1°C.

Default value: 10x10-6 /°C.

 Prestrain, ∆εp

Enter the elastic strain offset of the reinforcement relative to the unstrained concrete, in

millistrain.

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.


216
3. Click Delete in the Reinforcement Components group. The reinforcement component is deleted

from the list box and the remaining reinforcement components are renumbered.

When modeling SFRC in FormWorks the selections that options that need to be selected in the Define

Reinforced Concrete Properties tab are Reinforced Concrete as the reference type in the Concrete

Properties section, and either Steel Fibre-Hooked or Steel Fibre-Straight as the reference type in the

Reinforcement Component Properties section. In FormWorks, the required inputs in the Reinforcement

Component Properties section of the Define Reinforced Concrete Properties tab are the same for both the

straight and hooked fibres.

10.4.1.2 Structural Steel

The Define Reinforced Concrete Properties dialog box appears as shown in Figure 97 when structural

steel is selected as the main material type.

Figure 97: Define Reinforced Concrete Properties Dialog Box: Structural Steel
217
Enter the following properties in the Concrete Properties group. Any default values that are assigned

when a “0” is entered in FormWorks are noted.

 Thickness, T

Enter the out of plane thickness of the steel in millimetres.

 Yield Strength, 𝐅𝐲

Enter the stress of the yield plateau in MPa.

 Ultimate Strength, 𝐅𝐮

Enter the maximum stress the steel can attain before rupturing, in MPa. The ultimate strength

must be greater than or equal to the yield strength.

Default value: Fu = 1.5Fy MPa

 Elastic Modulus, 𝐄𝐬

Enter the stiffness of the initial linear-elastic branch of the stress-strain response, in MPa.

Default value: Es = 200 000 MPa

 Strain Hardening Strain, 𝛆𝐬𝐡

Enter the strain at which the steel 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, Fy/Es.

Default value: esh = 5 me

 Ultimate Strain, 𝛆𝐮

Enter the strain at which the steel ruptures in millistrain.

Default value: eu = 150 me

 Thermal Expansion Coefficient, 𝐂𝐜

Enter the steel strain increase per temperature increase of 1 °C.

Default value: Cc = 10x10−6 /°C

 Poisson’s Ratio, 𝛎

Enter the initial Poisson’s ratio, νo , of the steel as a positive value.

Default value: ν = 0.30


218
 Density

Enter the mass density of the steel in kg/m3.

Default value: ρ = 7850 kg/m3

 Thermal Diffusivity, 𝐊 𝐬

Enter the thermal diffusivity of the steel in mm2/s.

 Unsupported Length Ratio, 𝐛/𝐭

Enter the unsupported length ratio of the steel.

10.4.1.3 Masonry

The Define Reinforced Concrete Properties dialog box appears as shown in Figure 98 when masonry is

selected as the main material type.

Figure 98: Define Reinforced Concrete Properties: Masonry

Enter the following properties in the Concrete Properties group. Properties marked by * are assigned

default values when ‘0’ are entered in their fields.


219
 Thickness, T

Enter the out of plane thickness of the masonry in millimetres.

 Cylinder Compressive Strength, 𝐟𝐜′

Enter the peak compressive stress in MPa. As explained in section 5.2.2, the maximum

compressive masonry strength, fmy , is entered as the cylinder compressive strength, fc′ .

 Tensile Strength, 𝐟𝐭′

Enter the uniaxial cracking strength of the masonry in MPa.

Default value: ft = 0.33�fc′ MPa

 Initial Tangent Elastic Modulus, 𝐄𝐜

Enter the tangent stiffness of the masonry stress-strain response at zero-strain in MPa.

2000f′c
Default value: Ec = εo
MPa (if εo > 0); Ec = 3320�fc′ + 6900 MPa

 Cylinder Strain at 𝐟𝐜′ , 𝛆𝐨

Enter the compressive strain, εo , corresponding to fc′ , as a positive value in millistrain.

2000f′c
Default value: ε0 = if Ec > 0; ε0 = 1.8 + 0.0075fc′ me
Ec

 Poisson’s Ratio, 𝛎

Enter the initial Poisson’s ratio, νo , of the masonry as a positive value.

Default value: ν = 0.15

 Thermal Expansion Coefficient, 𝐂𝐜

Enter the masonry strain increase per temperature increase of 1 °C.

Default value: Cc = 10 × 10−6 /°C

 Maximum Aggregate Size, a

Enter the maximum aggregate size in millimetres.

Default value: a = 20 mm

 Density

Enter the mass density of the masonry in kg/m3.

Default value: ρ = 2400 kg/m3


220
 Thermal Diffusivity, 𝐊 𝐜

Enter the thermal diffusivity of the masonry in mm2/s.

Default value: K c = 1.20 mm2/s

 Joint Spacing

Enter the joint spacing parallel to the y-axis for Sx and parallel to the x-axis for Sy . Sx and Sy

are defined as the spacings between the head joints and bed joints. See section 5.2.1 for a

description and diagram.

10.4.1.4 Wood-Orthotropic

The Define Reinforced Concrete Properties dialog box appears as shown in Figure 99 when wood is

selected as the main material type.

Figure 99: Define Reinforced Concrete Properties: Wood-Orthotropic


221
Enter the following properties in the Concrete Properties group. There are no VecTor2 default

properties for this material type.

 Thickness, T

Enter the thickness of the wood in the out of plane direction in millimetres.

 Longitudinal Direction from x-axis

Enter the longitudinal direction of the wood from the x-axis in degrees. In wood material

models, the longitudinal direction is the direction of the grain of the wood.

 Compressive Strength—Longitudinal, 𝐟𝐜−𝐥

Enter the compressive strength in the longitudinal direction in MPa.



 Compressive Strength—Transverse, 𝐟𝐜−𝐭

Enter the compressive strength in the transverse direction, perpendicular to the direction of

the grain, in MPa.



 Tensile Strength—Longitudinal, 𝐟𝐭−𝐥

Enter the tensile strength in the longitudinal direction in MPa.



 Tensile Strength—Transverse, 𝐟𝐭−𝐭

Enter the tensile strength in the transverse direction, perpendicular to the direction of the

grain, in MPa.

 Shear Strength

Enter the shear strength, νlt , in MPa. This is the wood shear strength parallel to the grain, the

longitudinal shear strength of the wood. Shear strength, νlt , varies depending on the type

of wood, and the water content of the wood.

 Density

Enter the density of the wood in kg/m3.

 Elastic Modulus—Longitudinal, 𝐄𝐥

Enter the elastic modulus for the longitudinal direction in MPa.

 Elastic Modulus—Transverse, 𝐄𝐭

Enter the elastic modulus for the transverse direction in MPa.


222
 Poisson’s Ratio—Longitudinal stress-transverse strain, 𝛎𝐥𝐭

Specify the initial Poisson’s ratio for longitudinal stress-transverse strain.

 Poisson’s Ratio—Transverse stress-longitudinal strain, 𝛎𝐭𝐥

Specify the initial Poisson’s ratio for transverse stress-longitudinal strain.

10.4.1.5 Modeling Composite Materials

As mentioned previously, composite materials can be modeled in VecTor2 using the four main reinforced

concrete materials and the appropriate type of smeared reinforcement. This section outlines how to model

each type of composite. The four composites that can be modeled are: concrete-steel laminate, concrete-

SFRC laminate, masonry-SFRC laminate, and concrete-ortho laminate.

10.4.1.5.1 Concrete-Steel Laminate

To model concrete-steel laminate the material type selected must be reinforced concrete. The Define

Reinforced Concrete Properties dialog box appears as shown in Figure 100.

Figure 100: Define Reinforced Concrete Properties: Concrete-Steel Laminate


223
After the two materials, the concrete and reinforcement, are defined in FormWorks, VecTor2 creates a

combined material stiffness which is used for the analysis, as discussed previously in section 5.4. Regular

ductile steel reinforcement can also be specified as a reinforcement component for the concrete in the

concrete-steel laminate, although this is not required. If specified, the ductile steel reinforcing bars will

also be incorporated into the combined material stiffness matrix.

Concrete Properties

See section 10.4.1.1 for the required concrete inputs. Note that that the thickness entered in the Concrete

Properties section is the thickness of the concrete core only and does not include the steel skin plate, the

thickness of which is specified in the Reinforcement Component Properties section.

Reinforcement (Steel Skin Plate) Properties

For the most part, the input properties required for the steel skin plate are similar to the input requirements

for the other steel reinforcement types. The input parameters unique to the steel skin plate reinforcement

are:

 Laminate Thickness, 𝐓𝐒

Enter the sum of the thicknesses of the steel plates in millimetres.

 Poisson’s Ratio, 𝛎

Enter the initial Poisson’s ratio, νo , for the steel plate.

The other parameters that must be entered are:

- Yield Strength, 𝐹𝑦

- Ultimate Strengh, 𝐹𝑢

- Elastic Modulus, 𝐸𝑠

- Strain Hardening Strain, 𝜀𝑠ℎ

- Ultimate Strain, 𝜀𝑢

- Thermal Expansion Coefficient, 𝐶𝑠 (default value assigned if left as “0”)


224
- Prestrain (if applicable)

- Unsupported Length Ratio, 𝑏/𝑡

For further instruction on how to enter these parameters, see section 10.4.1.1.

10.4.1.5.2 Concrete-SFRC Laminate

Defining the concrete-SFRC laminate properly in FormWorks is important. In the Define Reinforced

Concrete Properties tab, in the Concrete Properties subgroup, the reference type must be set to

Reinforced Concrete. In the Reinforcement Component Properties subgroup, the reference type must be

set to either SFRC Laminate-Hooked Fibre or SFRC Laminate-Straight Fibre. The difference

between the two fibre types is discussed briefly in the reinforcement materials section of this user manual.

The user can also choose to include normal ductile reinforcing steel in the concrete, although this is not

required.

The Define Reinforced Concrete Properties dialog box appears as shown in Figure 101.

Figure 101: Define Reinforced Concrete Properties: Concrete-SFRC Laminate


225
Concrete Properties

See section 10.4.1.1 for the required concrete inputs.

SFRC Laminate Reinforcement Properties

Enter the following properties in the Reinforcement Component Properties group. Properties marked

by * are assigned default values when ‘0’ are entered in their fields. The same properties must be entered

regardless of whether the hooked fibre or straight fibre laminate is used.

 Fibre Volume Fraction, 𝐕𝐟

Enter the fibre volume fraction in %.

 Fibre Length, 𝐋𝐟

Enter the fibre length used in the SFRC in millimetres.

 Fibre Diameter, 𝐃𝐟

Enter the fibre diameter in millimetres.

 Fibre Tensile Strength, 𝐅𝐮

Enter the fibre tensile strength in MPa.

Default value: Fu = 1100 MPa

 Fibre Bond Strength, 𝐓𝐮

Enter the fibre bond strength in MPa.

 Cylinder Compressive Strength, 𝐟𝐜′

Enter the peak compressive stress in MPa.

 Tensile Strength, 𝐟𝐭′

Enter the uniaxial cracking strength of the SFRC in MPa.

Default value: ft′ = 0.33�fc′ MPa

 Initial Tangent Elastic Modulus, 𝐄𝐜

Enter the tangent stiffness of the SFRC stress-strain response at zero-strain in MPa.

Default value: Ec = 3320�fc′ + 6900 MPa

 Cylinder Strain at 𝐟𝐜′ , 𝛆𝐨


226
Enter the compressive strain, εo , corresponding to fc′ , as a positive value in millistrain.

Default value: ε0 = 1.8 + 0.0075fc′ me

 Maximum Aggregate Size, a

Enter the maximum aggregate size in millimetres.

Default value: a = 20 mm

 Laminate Thickness, T

Enter the SFRC laminate total thickness in millimetres.

10.4.1.5.3 Masonry-SFRC Laminate

The masonry-SFRC laminate is defined in FormWorks in basically the same way as the concrete-SFRC

laminate. In the Define Reinforced Concrete Properties tab, in the Concrete Properties subgroup, the

reference type must be set to Reinforced Concrete. In the Reinforcement Component Properties

subgroup, the reference type must be set to either SFRC Laminate-Hooked Fibre or SFRC Laminate-

Straight Fibre.

The Define Reinforced Concrete Properties dialog box appears as shown in Figure 102.

Figure 102: Define Reinforced Concrete Properties: Masonry-SFRC Laminate


227
Masonry Properties

The same masonry properties must be entered as are required to define regular reinforced or unreinforced

masonry. See section 10.4.1.3 for the required concrete inputs.

SFRC Laminate Reinforcement Properties

The SFRC laminate properties are entered as they are entered for concrete-SFRC laminate. See section

10.4.1.5.2 for details on the required inputs.

10.4.1.5.4 Concrete-Ortho Laminate

The Concrete-Ortho Laminate is a combination of a concrete core and wood (orthotropic) faceplates. In

the Define Reinforced Concrete Properties tab, in the Concrete Properties subgroup, the reference type

must be set to Concrete-Ortho Laminate. In the Reinforcement Component Properties subgroup, the

reference type must be set to Orthotropic Laminate.

The Define Reinforced Concrete Properties dialog box appears as shown in Figure 103.

Figure 103: Define Reinforced Concrete Properties: Concrete-Ortho Laminate


228
Concrete Properties

The same concrete properties must be entered as are required to define regular reinforced concrete. See

section 10.4.1.1 for the required concrete inputs.

Orthotropic Laminate Properties

The inputs required are the same as for the wood (fixed orthotropic) material type, with the exception of

density. The wood density is not a required input for orthotropic laminate. The required inputs are

specified in section 10.4.1.4.


229
10.4.2 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 104.

Figure 104: 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 mm2.

 Color
230
Enter the display color of the reinforcement material type in the FormWorks Workspace.

10.4.2.1 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

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 105.

Figure 105: 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.


231
 Reference Type

As described in Chapter 7, 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.

 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, β, as follows:

σ
β= , (in MPa ) 0 ≤ β ≤ 1, (10.4.2.1.1)
7.5

where σ 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 mm2, per

element.
232
 Bond Stress-Slip Curve Reference Points

Enter three reference bond stresses, Ui, and corresponding slips, Si, to define a multilinear

bond stress-slip relationship. Enter bond stresses, Ui, in MPa and corresponding slips, Si, in

millimeters.

10.4.3 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 106.

Figure 106: Create Nodes Dialog Box

2. Complete the fields to create a lattice of nodes separated by constant spacing and node number.

Refer to Figure 107 as an example corresponding to the values shown in Figure 106.
233

37 38 39 40 41 42

5 nodes spaced from node 13 31 32 33 34 35 36


@ 0 mm in the x-direction
@ 100 mm in the y-direction 25 26 27 28 29 30
increasing in node number by 6

19 20 21 22 23 24

13 14 15 16 17 18

6 nodes spaced from node 13


Node 13 located @ @ 125 mm in the x-direction
(100,200) mm. @ 0 mm in the y-direction
Increasing in node number by 1
Enter: Node X Y #nodes dnode dx dy #nodes dnode dx dy
13 100. 200. 6 1 125. 0. 5 6 0. 100.

Figure 107: 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.

10.4.4 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
234
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 108.

Figure 108: Create Rectangular Elements Dialog Box


235
2. Complete the fields to create a lattice of elements with constant change in element numbering and

defining nodes. Refer to Figure 109 as an example for creating rectangular elements. The rules for

different element types are subsequently discussed.

37 38 39 40 41 42
16 17 18 19 20

31 32 33 34 35 36
4 elements from element 1 11 12 13 14 15
Increasing in element number by 5
Increasing in defining node number by 6 25 26 27 28 29 30
6 7 8 9 10

19 20 21 22 23 24
1 2 3 4 5

13 14 15 16 17 18

Element 1 5 elements from element 1


Defined by nodes 13, 14 20, 19 Increasing in element number by 1
Increasing in defining node number by 1

Enter: Elmt node 1 2 3 4 #elmts delmt dnode #elmts delmt dnode


1 13 14 20 19 5 1 1 4 5 6

Figure 109: 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.


236
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 110, 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.

19 20 21 22 23 24
11 12 13 14 15

13 14 15 16 17 18
90 91 92 93 94 95

33 46 34 47 35 48 36 49 37 50 38

Node Pairs with same coordinates:


Elements: 13 & 33
Rectangular: 12 to 15 14 & 34
Truss: 46 to 40 15 & 35
Link: 90 to 95 16 & 36
17 & 37
18 & 38
Enter: Elmt node 1 2 #elmts delmt dnode #elmts delmt dnode
90 13 33 6 1 1 1 1 1

Figure 110: Example of link element creation entry


237
For Contact elements, enter the four node numbers that define the nodes of the source element.

As shown in the example of Figure 111, 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.

19 20 21 22 23 24
11 12 13 14 15

13 14 15 16 17 18
90 91 92 93 94

33 46 34 47 35 48 36 49 37 50 38

Node Pairs with same coordinates:


Elements: 13 & 33
Rectangular: 12 to 15 14 & 34
Truss: 46 to 40 15 & 35
Contact: 90 to 94 16 & 36
17 & 37
18 & 38
Enter: Elmt node 1 2 3 4 #elmts delmt dnode #elmts delmt dnode
90 13 33 14 34 5 1 1 1 1 1

Figure 111: 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
238
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.

10.4.4.1 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 112. The element

number, element type, incident nodes and their coordinates, material types and assigned loads are shown.

Figure 112: Element attributes dialog


239
10.4.4.2 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.

10.4.5 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.
240
 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 113.

Figure 113: 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 114. Only elements

compatible with material type may be selected.


241

Figure 114: 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.

10.4.6 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 115.

Figure 115: Create Support Restraints Dialog Box


242
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 116. Check the X and/or Y box to restrain the selected nodes against

displacements in the x and Y directions, respectively.

Figure 116: 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.


243
10.4.7 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.

10.4.7.1 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.


244
Legal and illegal polygon regions are shown in Figure 117.

Legal Legal Perfect Edge Intersection of Regions

Illegal Imperfect Edge


Illegal Clockwise Definition Illegal Self-Intersection Intersection of Regions

Figure 117: 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 115, with the RC regions property page as the active page.
245

Figure 118: 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 119 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
246
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) (1000, 1000) (2000, 1000) (3000, 1000)

Nodes to be
Enforced intermediate created by
nodal locations FormWorks
restrained in the
x direction
(0,0) (3000,0)

Entry: X Y Reps Dx Dy Restrain Edge


0. 0. 1 0. 0. X Y +
3000. 0. 1 0. 0. X Y +
3000. 1000. 3 -1000. 0. X Y +
0. 1000. 1 0. 0. X Y Close

Figure 119: Example of reinforced concrete region creation


247
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 120. Like the Grid Superposition

algorithm, FormWorks attempts to discretize the

Figure 120: Hybrid Discretization


248
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 121. 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 121: 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 122. 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 122: Division Point Insertion Discretization


249

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

123 shows the Hybrid discretization of Figure 120

with the rectangle option

Figure 123: 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 124 shows the

Division point insertion discretization of Figure

122 with the quadrilateral option.

Figure 124: Division Point Insertion Discretization with Quadrilateral Option


250
 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 125

shows the Division point insertion discretization of

Figure 122 with the smoothed triangles option.

Figure 125: 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
251
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.
252
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.

10.4.7.2 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.
253
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 126.

Figure 126: 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


254
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 127 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.
255
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.  +
450. 65. 1 0. 0. +
3000. 65. 1 0. 0. +

Figure 127: 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.
256
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.


257
10.4.7.3 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 128.

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.
258

Figure 128: 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
259
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.

10.4.7.4 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.
260
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 128.

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.

10.4.7.5 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
261
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 129 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.
262
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  Y  +
2750. 800. 5 50. 0. 4 0. 50. X Y +

Figure 129: 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.

10.4.7.6 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 130.


263

Figure 130: 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
264
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.


265

11 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 Job Data.

11.1 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 131.

Figure 131: Load Limits Dialog Box

11.2 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 Job Data may be selected.

11.3 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
266
Information dialog appears as shown in Figure 132.

Figure 132: Load Information Dialog Box

11.4 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 133.


267

Figure 133: Apply Nodal Loads Dialog Box

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.


268

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.

11.5 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 134.

Figure 134: 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.


269

11.6 Gravity Loads

Gravity loads are applied to rectangular, quadrilateral and triangular elements to include the

weight, W, of the element, as follows.

W x = +GX ⋅ 9.81 ⋅ ρ ⋅ V (11.6.1a)

W y = −GY ⋅ 9.81 ⋅ ρ ⋅ V (11.6.1b)

where, Wx and Wy 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, ρ 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 135.

Figure 135: 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/m3.


270
 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.)

 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.


271

11.7 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:

[ε o
cx ε ocy γ ocxy ] + = [α
T
c ⋅ ∆T α c ⋅ ∆T 0] T (11.7.1)

ε os = α s ⋅ ∆T (11.7.2)

where ε oc and ε os are elastic strain offsets for the concrete and reinforcement, respectively, αc and αs 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 Element Temperature toolbar

button. The Apply Temperature Loads dialog box appears as shown in Figure 136.

Figure 136: 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
272
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.

11.8 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:

[ε o
cx ε ocy γ ocxy ] + = [ε
T
ps ε ps 0 ] T
(11.8.1)

where ε oc is elastic strain offsets of the concrete, and εps 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 137.

Figure 137: 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.
273
 d strain

Enter the increment in prestrain between successive elements in a row/column.

11.9 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 138.

Figure 138: 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.

11.10 Nodal Thermal Loads

Nodal thermal loads are time varying temperature loads applied to the surface of rectangular,
274
quadrilateral, and triangular elements. These loads are to be used with the time step analysis mode.

Types of Thermal Loads

In VecTor2, there are 5 types of thermal loads, specified in FormWorks by entering type number 1-5 in

the Apply Nodal Thermal Loads dialog box. The different thermal profiles are outlined in Figure 139.

Figure 139: Thermal profiles in VecTor2 (Zhou, 2004)


275
To apply Nodal Thermal Loads, complete the following steps.

1. Click the Nodal Thermal Loads toolbar button. The Apply Nodal Thermal Loads dialog box

appears as shown in Figure 140.

2. Specify the nodal thermal loads by one of the following methods.

Figure 140: Apply Nodal Thermal Loads Dialog Box

To Select Surfaces by Node Number

Complete the Node, #Nodes, and dNode inputs to select the desired concrete elements. Specify the Type

of thermal load, and the applicable Tmi and Tpi.

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 Type of thermal load, and required Tmi and Tpi.

Note that for thermal load types 4 and 5, the standard fire models, require no user input of time-

temperature data.

3. Click Apply or type Enter. The applied nodal thermal loads are drawn in the Workspace and

added to the drop list in the dialog box.


276
Nodal thermal loads can be modified by repeating the above procedure. Previously defined nodal thermal

loads will be overwritten. Nodal thermal loads are deleted by selecting them either from the drop list or

by using the mouse, then clicking Delete.

11.11 Lumped Masses

Impact loads can be modeled with lumped masses when the structural mass is known; this is

typically the case where dropped weights are used. When modeling lumped masses, the structural mass

is lumped at the nodes, and is specified using the Apply Lumped Masses option in FormWorks. The most

general way to calculate the nodal masses is to calculate the mass of an element and divide it by the

number of nodes; more than one element can contribute to an individual nodal mass. Currently in

VecTor2, the density entered in the Concrete Material Properties section is not converted to nodal mass

automatically, but is included as static self-weight loads. If a non-zero GF-X or GF-Y is specified for the

lumped mass, and the default concrete density is assumed, the self-weight will have been double counted.

It is recommended that when using non-zero static load multipliers with lumped masses, that a very small

positive density be used to avoid double counting.

To Apply Lumped Masses, complete the following procedure.

1. Select the Load/Apply Lumped Masses menu item or click the Lumped Masses toolbar

button. The Apply Lumped Masses dialog box appears as shown in Figure 141.

Figure 141: Apply Lumped Masses Dialog Box

2. Specify the nodes and lumped masses by one of the following methods:

To select nodes by node number


277
Complete Node, #Nodes, and dNodes entry fields to select the desired nodes. Specify the

loads.

 DOF-X

If the mass is activated in the x-direction, check this box. Otherwise, leave it unchecked.

 DOF-Y

If the mass is activated in the y-direction, check this box. Otherwise, leave it unchecked.

 Mass (kg)

Enter the nodal mass in kg.

 GF-X

Enter the multiplier for the static load in the x-direction. This converts the nodal mass to a

force and applies it as a static load in the x-direction. Typically, values of +1 or -1 are used to

consider self-weight. For the conversion into a force, g=9.81 m/s2 is used.

 GF-Y

Enter the multiplier for the static load in the y-direction. In order to specify a gravity load

downward, the direction must be specified as -1.

 Vo-X

Enter the contact velocity (the initial velocity) of the lumped mass in the x-direction.

 Vo-Y

Enter the contact velocity of the lumped mass in the y-direction. A downward velocity is

specified with a negative.

Alternatively, to Select Nodes 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 DOF-X, DOF-Y, Mass (kg), GF-X, GF-Y, Vo-X, and

Vo-Y as described above.


278
Lumped masses can be modified by repeating the above procedure. Previously defined

lumped masses will be overwritten. Lumped masses can be deleted by selecting tem from the

drop list or by using the mouse, then clicking Delete.

11.12 Impulse Forces

Impulse loads are time varying nodal forces. In VecTor2, up to 20 impulse forces can be input

for a given node, and VecTor2 does not make any assumptions about the force returning to zero.

VecTor2 does assume that the impulse starts at t=0 at F=0 kN, and will connect the origin to the first

point with a straight line. After defining the impulse in FormWorks, there is the option to display the

impulse force for a given node, as shown below.

Figure 142: Impulse forces in VecTor2

When modeling the impulse from a blast, if the impact load-time history is not previously known,

VecTorBlast can be used to determine the impulse forces at different locations of the structure.

Note that for impulse loads, nodal/lumped masses must be specified in order for a dynamic

analysis to occur. The lumped masses do not need to be assigned any initial speed or acceleration,

however, masses only need to be assigned to the required nodes.


279
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 143. The figure shows the points that

must be input to produce the graph shown in Figure 142.

Figure 143: 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.

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.

Note: A total of 20 points can be defined for each node.


280
11.13 Ground Acceleration Loads

Ground acceleration loads may be specified to subject the structure to seismic-type loads. As

mentioned previously, there are two methods for specifying the ground acceleration in VecTor2. Ground

acceleration in terms of time and x- and y-accelerations can be input through FormWorks using the Apply

Ground Acceleration Load window. Alternatively, ground acceleration can be input using a

VECTOR.EQR file.

To specify user-defined 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 144.

Figure 144: 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/s2, 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.

Ground accelerations can also be specified using the VECTOR.EQR file.

The earthquake record should be available online as a text file. Online earthquake record databases

include:
281

Pacific Earthquake Engineering Research Center-Ground Motion Database

http://peer.berkeley.edu/smcat/

Pacific Earthquake Engineering Research Center: NGA Database

http://peer.berkeley.edu/nga/

Incorporated Research Institutions for Seismology

http://www.iris.edu/

Cosmos Virtual Data Center

http://db.cosmos-eq.org

After downloading the earthquake record, it must be renamed as VECTOR.EQR and put in the same

folder as the other input files. The VECTOR.EQR file must be of the following format (Saatci and

Vecchio, 2007):

Figure 145: input file format of VECTOR.EQR file

1. Number of lines to be skipped containing supplementary information.

2. Lines containing supplementary information.

3. Information: total number of data points, time increment, number of data points/row, data

unit
282
4. Three lines to be skipped. (contain supplementary info)

5. Acceleration data for earthquake event in row-wise order.

Note that when the ground acceleration is entered using the VECTOR.EQR file, the Ground Acceleration

option in the Auxiliary tab of the Define Job window must be set to “Considered”. Further, no Ground

Acceleration Record should be specified in FormWorks when using a VECTOR.EQR file.


283

12 Running VecTor2

Once the Job, 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.

12.1 Starting the Analysis

1. As described in Section 8.4.3, save the FormWorks Workspace file by selecting the File/Save

menu item or clicking the Save toolbar item. It is recommended that .FWK 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 146. Complete the following procedure.

Figure 146: Bandwidth Reduction Dialog Box


284
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 Job

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 Job Data. Click Save.


285
The VecTor2 analysis proceeds as shown in Figure 147. 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 147: VecTor2 Analysis Proceeding

12.2 Example 1: Simply Supported Beam

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 148 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 Hybrid II

Vecchio-Lai model.
286

220 1830 1830 220

2- No. 10, As = 200 mm2, db = 11.3 mm, fy = 315 MPa, fu = 460 MPa
63
D5 @ 210, ρ=0.101%, α = 90°, fy = 600MPa, fu = 649 MPa
361
2- No. 25, As = 1000 mm2, db = 25..2 mm, fy = 445,MPa, fu = 680 MPa
2 -No. 30, As = 1500 mm2, db = 29.9 mm, fy =436 MPa, fu =700 MPa
63
63
Concrete: f’c = 32.6 MPa, f’t = 1.819 MPa, eo = 1.902 me
305
All dimensions in mm

Figure 148: 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 149. 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.


287

Figure 149: FormWorks Model

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 150.

Figure 150: Load versus midpsan deflection of reinforced concrete beam


288
At the ultimate load corresponding deflection of 450 kN and 17 mm, respectively, VecTor2

determines the cracking pattern as shown in Figure 151.

Figure 151: Crack pattern of reinforced concrete beam at ultimate load


289
12.3 Example 2: SW22

This second example shows the FormWorks modeling and VecTor2 calculation of load-deflection

response for a shear wall. The shear wall, SW22, in this example was one of 13 shear walls tested under

axial load and monotonically increasing lateral load by Lefas et al in 1990 and modeled in Vecchio

(1992). The shear wall dimensions and reinforcement properties are illustrated in Figure 152.

Figure 152: SW22

The vertical reinforcement consists of two layers of 8 mm-diameter bars at 62 mm spacing; horizontal

reinforcement consists of 6.25 mm-diameter bars at 115 mm spacing. A 140 mm wide internal column

was formed at the end of each wall. The yield strengths for the 8 mm and 6.25 mm bars were taken as

470 MPa and 520 MPa, respectively. The modulus of elasticity for the reinforcement was 210,000 MPa,
290
with a strain hardening modulus of approximately 10,000 MPa, The concrete cylinder compressive

strength used in the model was 36.6 MPa; the tensile strength used was 2.16 MPa, with an initial tangent

elastic modulus of 32800 MPa. Poisson’s ratio was assumed to be 0.15. An axial load of 182 kN was

applied to the wall.

All reinforcement for SW22 was modeled as smeared reinforcement in VecTor2; there are no discrete

reinforcing bars in the model. The mesh for SW22 is shown in Figure 153.

Figure 153: FormWorks Model for SW22

As shown in the mesh, there are three concrete types. Type 1 is used in the centre of the shear wall, type

2 is used for the exterior columns, Type 3 is used for the base and top. The specific properties used for

each concrete material type can be found in the excerpt from the SW22 structure file.

The models used in the analysis of SW22 are summarized in Figure 154:
291

Figure 154: Models used for SW22

The load-deflection response for SW22 as calculated by VecTor2 is shown in Figure 155.

Figure 155: Load-displacement response for SW22


292

At the ultimate load condition of 20mm and 154kN, the crack pattern and deflected shape of SW22 are

shown in Figure 156.

Figure 156: Deflected Shape and Final Crack Pattern for SW22
293

13 Summary & Recommendations

13.1 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.

13.2 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.
294
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.
295

References

Akkaya, Y., Guner, S. and Vecchio, F. J., 2013, “Modeling inelastic buckling behavior of reinforcing bars

in compression”

Akkaya, Y., Guner, S. and Vecchio, F. J., 2013, “Modeling hysteretic response of reinforcing bars

including buckling in reinforced concrete members”

Akkaya, Y., Guner, S. and Vecchio, F. J., 2012, “Modeling cyclic behavior of reinforcing steel bars

including buckling,”

Bazant, Z.P., 2002. “Concrete fracture models: testing and practice”. Engineering Fracture Mechanics.

Vol. 69, pp. 165-205

Bentz, E.C., 2000. "Sectional Analysis of Reinforced Concrete Members", Ph.D. Thesis, Department of

Civil Engineering, University of Toronto, 310 pp.

Chopra, A., 2007. “Dynamics of Structures”. Prentice Hall. New Jersey.

Collins, M.P., and Mitchell, D., 1997. “Prestressed Concrete Structures”, Response Publications, Canada,

766 pp.

Comite Euro-International du Beton. 1990. “CEB-FIP Model Code 1990”. Thomas Telford. London.

Cornelissen, H.A., Hordijk, D.A., Reinhardt, H.W. 1986. “Experimental determination of crack softening

characteristics of normal weight and lightweight concrete”. HERON Vol. 31, No. 2
296
Deluce, J.D., Lee, S.-C., and Vecchio, F.J. 2012. “Crack Model for SFRC Members Containing

Conventional Reinforcement,”ACI Structural Journal, under review.

Dhakal, R.P. and Maekawa, K., 2002a, “Modeling for post-yield buckling of reinforcement,” Journal of

Structural Engineering, ASCE, Vol. 128, No. 9, pp. 1139-1147.

Dhakal, R.P., and Maekawa, K., 2002b, “Reinforcement Stability and Fracture of Cover Concrete in

Reinforced Concrete Members,” Journal of Structural Engineering, ASCE, Vol. 128, No.10, pp. 1253-

1262.

Dhakal, R.P., and Maekawa, K., 2002c, “Path-dependent Cyclic Stress-Strain Relationship of Reinforcing

bar including buckling,” Engineering Structures, Vol. 24, No. 11, pp. 1383-1396.

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.

fib. 2010. “First complete draft: bulletin 55”. Vol 1

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.

Ganz, H.R., 1985. “Mauerwerksscheiben unter Normalkraft und Schub”. Dissertation 7849, Institute of

Structural Engineering (IBK), Swiss Federal Institute of Technology (ETH). Zurich, Switzerland

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.


297

Harajli, M.H. and Mukaddam, M.A., 1988. “Slip of Steel Bars in Concrete Joints under Cyclic Loading”,

ASCE Journal 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 Journal of Structural Engineering, Vol. 123, No. 5, pp.

624-633.

Hsieh, S.S., Ting, E.C., and Chen, W.F., 1979. “An Elastic-Fracture Model for Concrete.” Proc. 3rd 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 Journal of

Structural Engineering, Vol. 124, No. 12, pp. 1467-1475.

Kent, D.C., and Park, R., 1971. “Flexural Members with Confined Concrete”, ASCE Journal 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 Journal of

Structural Engineering, Vol. 113, No. 10, pp. 2160-2173.


298
Kupfer, H.B., and Gerstle, K.H., 1973. “Behaviour of Concrete under Biaxial Stresses”, ASCE Journal 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

Journal, 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.

Lee, S.-C., Cho, J.-Y., and Vecchio, F.J. 2011a. “Diverse Embedment Model for Steel Fiber-Reinforced

Concrete in Tension: Model Development”. ACI Materials Journal, Vol. 108, No. 5, pp.516-525

Lee, S.-C., Cho, J.-Y., and Vecchio, F.J. 2011b. “Diverse Embedment Model for Steel Fiber-Reinforced

Concrete in Tension: Model Verification”. ACI Materials Journal, Vol. 108, No. 5, pp.526-535

Lee, S.-C., Cho, J.-Y., and Vecchio, F.J. 2011c. “Model for Post-Yield Tension Stiffening and Rebar

Rupture in Concrete Members”. Engineering Structures, Vol. 33, No. 5, pp.1723-1733

Lee, S.-C., Cho, J.-Y., and Vecchio, F.J. 2012a. “Simplified Diverse Embedment Model for SFRC

Elements in Tension”. ACI Materials Journal, under review

Lee, S.-C., Cho, J.-Y., and Vecchio, F.J. 2012b. “Tension Stiffening Model for Steel Fiber Reinforced

Concrete Containing Conventional Reinforcement”. ACI Structural Journal, under review


299
Lee, S.-C., Oh, J.-H., and Cho, J.-Y. 2012. “Compressive Behavior of Normal and High Strength

Concrete with End-Hooked Steel Fibers”. in progress

Li, L., Han, X., and Xu, S. 2004. “Study on the degeneration of quadrilateral element to triangular

element”. Communications in Numerical methods in Engineering. pp 671-679.

Mander, J.B., Priestley, M.J.N., and Park. R., 1988. "Theoretical Stress-Strain Model for Confined

Concrete", ASCE Journal of Structural Engineering, Vol. 114, No. 8, pp. 1804-1826.

Minelli, F., and Vecchio, F.J. 2006. “Compression Field Modeling of Fiber Reinforced Concrete

Members Under Shear Loading”. ACI Structural Journal, Vol. 103, No. 2, pp. 244-252

Montoya, E. 2003. “Behaviour and Analysis of Confined Concrete”. PhD Thesis. Department of Civil

Engineering. University of Toronto. 321 pp.

Montoya, E., Vecchio, F.J., and Sheikh, S.A. 2004. “Numerical evaluation of the behavior of steel- and

FRP-confined concrete columns using compression field modeling”. Engineering Structures, Vol. 26,

pp. 1535-1545

Montoya, E., Vecchio, F.J., and Sheikh S.A. 2006. “Compression Field Modeling of Confined Concrete:

Constitutive Models”. ASCE Journal of Materials in Civil Engineering, Vol. 18, No. 4, pp. 1-8

Ngo, D. and Scordelis, A.C., 1967. “Finite Element Analysis of Reinforced Concrete Beams”, ACI

Journal, 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.


300

Ottosen, N.S. 1977. “A Failure Criterion for Concrete”, ASCE Journal 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.

Saatci, S. 2007. "Behaviour and Modeling of Reinforced Concrete Structures Subjected to Impact Loads",

PhD Thesis, Department of Civil Engineering, University of Toronto, 316 pp.

Saatci, S. and Vecchio, F.J. 2007. “Dynamic Analysis with VecTor2”. FormWorks Bulletin

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.


301
Sato, Y. and Vecchio, F.J., 2003. “Tension Stiffening and Crack Formation in RC Members with FRP

Sheets”, ASCE Journal of Structural Engineering, Vol. 129, No 6, pp. 717-724

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 Journal, 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.

Smith, G.M., and Young, L.E. 1956. “Ultimate Flexural Analysis Based on Stress-Strain Curve of

Cylinder”. ACI Journal. Vol. 53, No. 6, pp. 597-609

Susetyo, J. 2009. “Fibre Reinforcement for Shrinkage Crack Control in Prestressed, Precast Segmental

Bridges”. PhD Thesis. Department of Civil Engineering. University of Toronto. 532 pp.

Thorenfeldt, E., Tomaszewicz, A., and Jensen, J.J., 1987. "Mechanical Properties of High-Strength

Concrete and Application in Design", Proceedings of the Symposium "Utilization of High Strength

Concrete," Stavanger, Norway, June 1987, Tapir, Trondheim, pp. 149-159.

Vecchio, F.J., 1992. "Finite Element Modeling of Concrete Expansion and Confinement", ASCE Journal

of Structural Engineering, Vol. 118, No. 9, pp. 2390-2406.


302

Vecchio, F.J. 1999. "Towards Cyclic Load Modeling of Reinforced Concrete", ACI Structural Journal,

Vol. 96, No. 2, pp.193-202.

Vecchio, F.J., 2000. "Analysis of Shear-Critical Reinforced Concrete Beams", ACI Structural Journal,

Vol. 97, No.1 pp. 102-110.

Vecchio, F.J., 2000. “Disturbed Stress Field Model for Reinforced Concrete: Formulation”, ASCE

Journal 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 Journal Vol. 83, No. 2, pp. 219-231.

Vecchio, F.J. and Collins, M.P., 1993. "Compression Response of Cracked Reinforced Concrete", ASCE

Journal 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

Journal of Engineering Mechanics, Vol. 121, No. 6 pp. 702-708.

Vecchio, F.J. and Lai, D., 2004. “Crack Shear-Slip in Reinforced Concrete Elements”, Journal of

Advanced Concrete Technology, Vol. 2, No. 3, pp. 289-300

Vecchio, F.J. and McQuade, I., 2011. “Towards improved modeling of steel-concrete composite wall

elements”, Journal of Nuclear Engineering and Design, Vol. 241, No. 8, pp. 2629-2642
303

Walraven, J.C., 1981. “Fundamental Analysis of Aggregate Interlock”, ASCE Journal of the Structural

Division, Vol. 107, No. 11, pp. 2245-2270.

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.

Wilson, E., 2002. “Three-Dimensional Static and Dynamic Analysis of Structures”. Computers and

Structures, Berkeley.

Wittmann, F. H. 2002. “Crack formation and fracture energy of normal and high strength concrete”.

Qingdao: Centre for Durability, Maintenance, and Repair

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.

Zhou, C.E. 2004. “Nonlinear Finite Element Analysis of Reinforced Concrete Structures Subjected to

Transient Thermal Loads”. M.A.Sc. Thesis. Department of Civil Engineering. University of Toronto.

153 pp.
304
Zhou, C. E. and Vecchio, F. J., 2005. "Nonlinear Finite Element analysis of Reinforced Concrete

Structures Subjected to Transient Thermal Loads", Computers and Structures, Vol. 2, No. 6, pp. 455-

480.
305

Appendix A – VecTor2 Input Files for Sample Problem 1


306
VecTor.JOB

VER 3.5
* * * * * * * * * * * *
* V E C T O R *
* J O B D A T A *
* * * * * * * * * * * *

Job Title (30 char. max.) : Enter Job Title


Job File Name ( 8 char. max.) : VecTor
Date (30 char. max.) : Enter Date

STRUCTURE DATA
--------------
Structure Type : 2
File Name ( 8 char. max.) : Struct

LOADING DATA
------------
No. of Load Stages : 201
Starting Load Stage No. : 1
Load Series ID ( 5 char. max.) : ID

Load File Name Factors


Case (8 char. max.) Initial Final LS-Inc Type Reps C-Inc
1 Case1 0.0000 50.0000 0.2500 1 1 0.0000
2 NULL 0.0000 0.0000 0.0000 1 1 0.0000
3 NULL 0.0000 0.0000 0.0000 1 1 0.0000
4 NULL 0.0000 0.0000 0.0000 1 1 0.0000
5 NULL 0.0000 0.0000 0.0000 1 1 0.0000

ANALYSIS PARAMETERS
-------------------
Analysis Mode (1-2) : 1
Seed File Name (8 char. max.) : NULL
Convergence Limit (>1.0) : 1.000010
Averaging Factor (<1.0) : 0.600
Maximum No. of Iterations : 60
Convergence Criteria (1-5) : 2
Results Files (1-4) : 2
Output Format (1-3) : 1

MATERIAL BEHAVIOUR MODELS


-------------------------
Concrete Compression Base Curve (0-3) : 1
Concrete Compression Post-Peak (0-3) : 1
Concrete Compression Softening (0-8) : 1
Concrete Tension Stiffening (0-6) : 1
Concrete Tension Softening (0-3) : 1
Concrete Tension Splitting (0-1) : 0
Concrete Confined Strength (0-2) : 1
Concrete Dilation (0-1) : 1
Concrete Cracking Criterion (0-4) : 1
Concrete Crack Stress Calculation (0-2) : 1
Concrete Crack Width Check (0-2) : 1
Concrete Bond or Adhesion (0-3) : 1
Concrete Creep and Relaxation (0-1) : 1
Concrete Hysteresis (0-2) : 1
Reinforcement Hysteresis (0-2) : 1
Reinforcement Dowel Action (0-1) : 1
Reinforcement Buckling (0-1) : 1
Element Strain Histories (0-1) : 1
Element Slip Distortions (0-4) : 1
Strain Rate Effects (0-1) : 1
Structural Damping (0-1) : 1
Geometric Nonlinearity (0-1) : 1
Crack Allocation Process (0-1) : 1

<<< JOB FILE NOTES>>>


307
Struct.S2R
* * * * * * * * * * * * * * * * * * *
* V e c T o r 2 *
* S T R U C T U R E D A T A *
* * * * * * * * * * * * * * * * * * *

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

Structure Title (30 char. max.) : Enter Structure Title


Structure File Name ( 8 char. max.) : Struct
No. of R.C. Material Types : 2
No. of Steel Material Types : 3
No. of Bond Material Types : 0
No. of Rectangular Elements : 418
No. of Quadrilateral Elements : 0
No. of Triangular Elements : 0
No. of Truss Bar Elements : 114
No. of Linkage Elements : 0
No. of Contact Elements : 0
No. of Joints : 468
No. of Restraints : 13

MATERIAL SPECIFICATIONS
***********************

(A) REINFORCED CONCRETE


-----------------------
<NOTE:> TO BE USED IN RECTANGULAR AND TRIANGULAR ELEMENTS ONLY

CONCRETE
--------
MAT REF Ns T f'c [ f't Ec e0 Mu Cc Agg Dens Kc ] [Sx Sy]
TYP TYP # mm MPa MPa MPa me /C mm kg/m3 mm2/s mm mm
1 1 0 305.000 22.600 1.819 23770.000 1.902 0.150 0.000 20.000 2400.000
0.000 0.000 0.000
2 1 1 305.000 22.600 1.819 23770.000 1.902 0.150 0.000 20.000 2400.000
0.000 0.000 0.000
/
REINFORCEMENT COMPONENTS
------------------------
MAT REF DIR As Db Fy Fu Es esh eu Cs Dep b/t
TYP TYP deg % mm MPa MPa MPa me me /C me
2 1 90.000 0.101 6.400 600.000 649.000 200000.000 4900.000 0.000 0.000
0.000 0.000
/
(B) STEEL
---------
<NOTE:> TO BE USED FOR TRUSS ELEMENTS ONLY
MAT REF AREA Db Fy Fu Es esh eu Cs Dep b/t
TYP TYP mm2 mm MPa MPa MPa me me /C me
1 1 200.600 11.300 315.000 460.000 200000.000 10.000 0.000 0.000 0.000
0.000
2 1 997.600 25.200 445.000 680.000 200000.000 10.000 0.000 0.000 0.000
0.000
3 1 1404.400 29.900 436.000 700.000 200000.000 10.000 0.000 0.000 0.000
0.000
/
(C) BOND
--------
<NOTE:> TO BE USED FOR EXTERIOR/INTERIOR BONDED ELEMENTS
MAT REF { Ao U1 U2 U3 S1 S2 S3 }/{ CPF Cmin No. HOOK }
TYP TYP mm^2 MPa MPa MPa mm mm mm 0-1 mm LYR 0/1
/
ELEMENT INCIDENCES
******************

(A) RECTANGULAR ELEMENTS


------------------------
<<<<< FORMAT >>>>>
ELMT INC1 INC2 INC3 INC4 [#ELMT d(ELMT) d(INC)] [#ELMT d(ELMT) d(INC)]/
1 38 39 27 26 1 1 1 1 1 1 /
2 14 26 27 15 1 1 1 1 1 1 /
308

(B) QUADRILATERAL ELEMENTS


--------------------------
<<<<< FORMAT >>>>>
ELMT INC1 INC2 INC3 INC4 [#ELMT d(ELMT) d(INC)] [#ELMT d(ELMT) d(INC)]/
/
(C) TRIANGULAR ELEMENTS
-----------------------
<<<<< FORMAT >>>>>
ELMT INC1 INC2 INC3 [ #ELMT d(ELMT) d(INC) ] [ #ELMT d(ELMT) d(INC)] /
/
(D) TRUSS ELEMENTS
------------------
<<<<< FORMAT >>>>>
ELMT INC1 INC2 [ #ELMT d(ELMT) d(INC)] [#ELMT d(ELMT) d(INC) ]/
419 2 14 1 1 1 1 1 1 /
420 14 26 1 1 1 1 1 1 /
421 26 38 1 1 1 1 1 1 /

(E) LINKAGE ELEMENTS


---------------------
<<<<< FORMAT >>>>
ELMT INC1 INC2 [ #ELMT d(ELMT) d(INC) ] [ #ELMT d(ELMT) d(INC) ]
/
(F) CONTACT ELEMENTS
---------------------
<<<<< FORMAT >>>>
ELMT INC1 INC2 INC3 INC4 [ #ELMT d(ELMT) d(INC) ] [ #ELMT d(ELMT) d(INC) ]
/
MATERIAL TYPE ASSIGNMENT
************************
<<<<< FORMAT >>>>>
ELMT MAT ACT [ #ELMT d(ELMT)] [ #ELMT d(ELMT) ] /
1 2 1 1 1 1 1 /
2 2 1 1 1 1 1 /
3 2 1 1 1 1 1 /
/
COORDINATES
***********
<NOTE:> UNITS: in 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 1 1 0.000 0.000 /
2 0.000 63.000 1 1 0.000 0.000 1 1 0.000 0.000 /
3 0.000 126.000 1 1 0.000 0.000 1 1 0.000 0.000 /
/
SUPPORT RESTRAINTS
******************
<NOTE:> CODE: '0' FOR NOT RESTRAINED NODES AND '1' FOR RESTRAINED ONES
<<<<< FORMAT >>>>>
NODE X-RST Y-RST [ #NODE d(NODE) ] /
49 0 1 1 1 /
457 1 0 1 1 /
458 1 0 1 1 /
/

<<< STRUCTURE FILE NOTES >>>


309
Case1.L2R
* * * * * * * * * * * * *
* V e c T o r 2 *
* L O A D D A T A *
* * * * * * * * * * * * *

LOAD CASE PARAMETERS


********************

Structure Title (30 char. max.) : Enter Structure Title


Load Case Title (30 char. max.) : Enter load case title
Load Case File Name (8 char. max.) : Case1
No. of Loaded Joints : 0
No. of Prescribed Support Displacements : 1
No. of Elements with Gravity Loads : 0
No. of Elements with Temperature Loads : 0
No. of Elements with Concrete Prestrain : 0
No. of Elements with Ingress Pressure : 0
No. of Nodes with Thermal Load : 0
No. of Nodes with Lumped Masses : 0
No. of Nodes with Impulse Forces : 0
Ground Acceleration Record (0-1) : 0

JOINT LOADS
***********
<NOTE:> UNITS: KIPS OR KN
<<<<< FORMAT >>>>>
NODE Fx Fy [ #NODE d(NODE) d(Fx) d(Fy) ] /
/
SUPPORT DISPLACEMENTS
*********************
<NOTE:> UNITS: MM OR IN
<<<<< FORMAT >>>>>
JNT DOF DISPL [ #JNT d(JNT) ] /
468 2 -1.000 1 1/
/
GRAVITY LOADS
*************
<NOTE:> UNITS: KG/M3
<<<<< FORMAT >>>>>
ELMT DENS GX GY [#ELMT d(ELMT)] [ #ELMT d(ELMT)] /
/
ELEMENT TEMPERATURE
*******************
<NOTE:> UNITS: F OR C
<<<<< FORMAT >>>>>
ELMT TEMP [ #ELMT d(ELMT) d(TEMP) ] [ #ELMT d(ELMT) d(TEMP) ] /
/
CONCRETE PRESTRAINS
*******************
<NOTE:> UNITS: me
<<<<< FORMAT >>>>>
ELMT STRAIN [ #ELMT d(ELMT) d(STRAIN) ] [ #ELMT d(ELMT) d(STRAIN) ] /
/
INGRESS PRESSURES
*****************
<NOTE:> UNITS: MPa
<<<<< FORMAT >>>>>
ELMT PRESSURE [ #ELMT d(ELMT) d(PRS) ] [ #ELMT d(ELMT) d(PRS) ] /
/
NODAL THERMAL LOADS
*******************
<NOTE:>
<<<<< FORMAT >>>>>
NODE TYPE Tm1 Tp1 Tm2 Tp2 Tm3 Tp3 [ #NODE d(NODE) ] [ #NODE d(NODE) ] /
/
LUMPED MASSES
*************
<NOTE:> UNITS: kg, m/s
<<<<< FORMAT >>>>>
NODE DOF-X DOF-Y MASS GF-X GF-Y Vo-X Vo-Y [ #NODE d(NODE) ] /
/
IMPULSE FORCES
310
**************
<NOTE:> UNITS: Sec, kN
<<<<< FORMAT >>>>>
NODE DOF T1 F1 T2 F2 T3 F3 T4 F4 [ #NODE d(NODE) ] /
/
GROUND ACCELERATION
*******************
<NOTE:> UNITS: Sec, G
<<<<< FORMAT >>>>>
TIME ACC-X ACC-Y
/

<<< LOAD FILE NOTES >>>


311

Appendix B – VecTor2 Input Files for Sample Problem 2


312
VecTor.JOB
VER 3.5
* * * * * * * * * * * *
* V E C T O R *
* J O B D A T A *
* * * * * * * * * * * *

Job Title (30 char. max.) : Shear Wall SW22


Job File Name ( 8 char. max.) : SW22
Date (30 char. max.) : May 2011

STRUCTURE DATA
--------------
Structure Type : 2
File Name ( 8 char. max.) : SW22

LOADING DATA
------------
No. of Load Stages : 201
Starting Load Stage No. : 1
Load Series ID ( 5 char. max.) : SW22

Load File Name Factors


Case (8 char. max.) Initial Final LS-Inc Type Reps C-Inc
1 SW22DL 0.0000 100.0000 0.5000 1 1 0.0000
2 SW22VL 1.8200 1.8200 0.0000 1 1 0.0000
3 NULL 0.0000 0.0000 0.0000 1 1 0.0000
4 NULL 0.0000 0.0000 0.0000 1 1 0.0000
5 NULL 0.0000 0.0000 0.0000 1 1 0.0000

ANALYSIS PARAMETERS
-------------------
Analysis Mode (1-2) : 1
Seed File Name (8 char. max.) : NULL
Convergence Limit (>1.0) : 1.000010
Averaging Factor (<1.0) : 0.500
Maximum No. of Iterations : 100
Convergence Criteria (1-5) : 2
Results Files (1-4) : 2
Output Format (1-3) : 1

MATERIAL BEHAVIOUR MODELS


-------------------------
Concrete Compression Base Curve (0-3) : 1
Concrete Compression Post-Peak (0-3) : 1
Concrete Compression Softening (0-8) : 1
Concrete Tension Stiffening (0-6) : 1
Concrete Tension Softening (0-3) : 1
Concrete Tension Splitting (0-1) : 1
Concrete Confined Strength (0-2) : 1
Concrete Dilation (0-1) : 1
Concrete Cracking Criterion (0-4) : 1
Concrete Crack Stress Calculation (0-2) : 1
Concrete Crack Width Check (0-2) : 2
Concrete Bond or Adhesion (0-3) : 1
Concrete Creep and Relaxation (0-1) : 1
Concrete Hysteresis (0-2) : 1
Reinforcement Hysteresis (0-2) : 1
Reinforcement Dowel Action (0-1) : 1
Reinforcement Buckling (0-1) : 1
Element Strain Histories (0-1) : 1
Element Slip Distortions (0-4) : 1
Strain Rate Effects (0-1) : 1
Structural Damping (0-1) : 1
Geometric Nonlinearity (0-1) : 1
Crack Allocation Process (0-1) : 1

<<< JOB FILE NOTES>>>


313
SW22.S2R
* * * * * * * * * * * * * * * * * * *
* V e c T o r 2 *
* S T R U C T U R E D A T A *
* * * * * * * * * * * * * * * * * * *

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

Structure Title (30 char. max.) : Shear Wall SW22


Structure File Name ( 8 char. max.) : SW22
No. of R.C. Material Types : 3
No. of Steel Material Types : 0
No. of Bond Material Types : 0
No. of Rectangular Elements : 496
No. of Quadrilateral Elements : 0
No. of Triangular Elements : 0
No. of Truss Bar Elements : 0
No. of Linkage Elements : 0
No. of Contact Elements : 0
No. of Joints : 559
No. of Restraints : 46

MATERIAL SPECIFICATIONS
***********************

(A) REINFORCED CONCRETE


-----------------------
<NOTE:> TO BE USED IN RECTANGULAR AND TRIANGULAR ELEMENTS ONLY

CONCRETE
--------
MAT REF Ns T f'c [ f't Ec e0 Mu Cc Agg Dens Kc ] [Sx Sy]
TYP TYP # mm MPa MPa MPa me /C mm kg/m3 mm2/s mm mm
1 1 2 65.000 36.600 2.160 32800.000 2.000 0.150 0.000 10.000 2400.000
0.000 0.000 0.000
2 1 4 65.000 36.600 2.160 32800.000 2.000 0.150 0.000 10.000 2400.000
0.000 0.000 0.000
3 1 3 200.000 36.600 2.160 32800.000 2.000 0.150 0.000 10.000 2400.000
0.000 0.000 0.000
/
REINFORCEMENT COMPONENTS
------------------------
MAT REF DIR As Db Fy Fu Es esh eu Cs Dep b/t
TYP TYP deg % mm MPa MPa MPa me me /C me
1 1 0.000 0.820 6.000 520.000 650.000 200000.000 10.000 150.000 0.000
0.000 0.000
1 1 90.000 2.090 6.000 470.000 650.000 200000.000 10.000 150.000 0.000
0.000 0.000
2 1 0.000 0.820 6.000 520.000 650.000 200000.000 10.000 150.000 0.000
0.000 0.000
2 1 0.000 0.336 6.000 420.000 650.000 200000.000 10.000 150.000 0.000
0.000 0.000
2 1 90.000 3.312 8.000 470.000 650.000 200000.000 10.000 150.000 0.000
0.000 0.000
2 1 361.000 0.900 6.000 420.000 650.000 200000.000 10.000 150.000 0.000
0.000 0.000
3 1 0.000 2.500 6.000 420.000 650.000 200000.000 10.000 150.000 0.000
0.000 0.000
3 1 90.000 2.500 6.000 420.000 650.000 200000.000 10.000 150.000 0.000
0.000 0.000
3 1 361.000 2.500 6.000 420.000 650.000 200000.000 10.000 150.000 0.000
0.000 0.000
/
(B) STEEL
---------
<NOTE:> TO BE USED FOR TRUSS ELEMENTS ONLY
MAT REF AREA Db Fy Fu Es esh eu Cs Dep b/t
TYP TYP mm2 mm MPa MPa MPa me me /C me
/
(C) BOND
--------
<NOTE:> TO BE USED FOR EXTERIOR/INTERIOR BONDED ELEMENTS
314
MAT REF { Ao U1 U2 U3 S1 S2 S3 }/{ CPF Cmin No. HOOK }
TYP TYP mm^2 MPa MPa MPa mm mm mm 0-1 mm LYR 0/1
/
ELEMENT INCIDENCES
******************

(A) RECTANGULAR ELEMENTS


------------------------
<<<<< FORMAT >>>>>
ELMT INC1 INC2 INC3 INC4 [#ELMT d(ELMT) d(INC)] [#ELMT d(ELMT) d(INC)]/
1 9 10 33 32 1 1 1 1 1 1 /
2 11 12 35 34 1 1 1 1 1 1 /
3 10 11 34 33 1 1 1 1 1 1 /
4 35 12 13 36 1 1 1 1 1 1 /
/
(B) QUADRILATERAL ELEMENTS
--------------------------
<<<<< FORMAT >>>>>
ELMT INC1 INC2 INC3 INC4 [#ELMT d(ELMT) d(INC)] [#ELMT d(ELMT) d(INC)]/
/
(C) TRIANGULAR ELEMENTS
-----------------------
<<<<< FORMAT >>>>>
ELMT INC1 INC2 INC3 [ #ELMT d(ELMT) d(INC) ] [ #ELMT d(ELMT) d(INC)] /
/
(D) TRUSS ELEMENTS
------------------
<<<<< FORMAT >>>>>
ELMT INC1 INC2 [ #ELMT d(ELMT) d(INC)] [#ELMT d(ELMT) d(INC) ]/
/
(E) LINKAGE ELEMENTS
---------------------
<<<<< FORMAT >>>>
ELMT INC1 INC2 [ #ELMT d(ELMT) d(INC) ] [ #ELMT d(ELMT) d(INC) ]
/
(F) CONTACT ELEMENTS
---------------------
<<<<< FORMAT >>>>
ELMT INC1 INC2 INC3 INC4 [ #ELMT d(ELMT) d(INC) ] [ #ELMT d(ELMT) d(INC) ]
/
MATERIAL TYPE ASSIGNMENT
************************
<<<<< FORMAT >>>>>
ELMT MAT ACT [ #ELMT d(ELMT)] [ #ELMT d(ELMT) ] /
1 3 1 1 1 1 1 /
2 3 1 1 1 1 1 /
3 3 1 1 1 1 1 /
/
COORDINATES
***********
<NOTE:> UNITS: in OR mm
<<<<< FORMAT >>>>>
NODE X Y [ #NODES d(NODES) d(X) d(Y) ] [ #NODES d(NODES) d(X) d(Y) ] /
1 -200.000 0.000 1 1 0.000 0.000 1 1 0.000 0.000 /
2 -150.000 0.000 1 1 0.000 0.000 1 1 0.000 0.000 /
3 -100.000 0.000 1 1 0.000 0.000 1 1 0.000 0.000 /
4 -50.000 0.000 1 1 0.000 0.000 1 1 0.000 0.000 /
/
SUPPORT RESTRAINTS
******************
<NOTE:> CODE: '0' FOR NOT RESTRAINED NODES AND '1' FOR RESTRAINED ONES
<<<<< FORMAT >>>>>
NODE X-RST Y-RST [ #NODE d(NODE) ] /
1 1 1 1 1 /
2 1 1 1 1 /
3 1 1 1 1 /
4 1 1 1 1 /

<<< STRUCTURE FILE NOTES >>>


315
SW22VL.L2R
* * * * * * * * * * * * *
* V e c T o r 2 *
* L O A D D A T A *
* * * * * * * * * * * * *

LOAD CASE PARAMETERS


********************

Structure Title (30 char. max.) : Shear Wall SW22


Load Case Title (30 char. max.) : Axial Ld (100 kN)
Load Case File Name (8 char. max.) : SW22VL
No. of Loaded Joints : 13
No. of Prescribed Support Displacements : 0
No. of Elements with Gravity Loads : 0
No. of Elements with Temperature Loads : 0
No. of Elements with Concrete Prestrain : 0
No. of Elements with Ingress Pressure : 0
No. of Nodes with Thermal Load : 0
No. of Nodes with Lumped Masses : 0
No. of Nodes with Impulse Forces : 0
Ground Acceleration Record (0-1) : 0

JOINT LOADS
***********
<NOTE:> UNITS: KIPS OR KN
<<<<< FORMAT >>>>>
NODE Fx Fy [ #NODE d(NODE) d(Fx) d(Fy) ] /
519 0.000 -7.692 1 1 0.000 0.000/
520 0.000 -7.692 1 1 0.000 0.000/
521 0.000 -7.692 1 1 0.000 0.000/
522 0.000 -7.692 1 1 0.000 0.000/
523 0.000 -7.692 1 1 0.000 0.000/
524 0.000 -7.692 1 1 0.000 0.000/
525 0.000 -7.692 1 1 0.000 0.000/
526 0.000 -7.692 1 1 0.000 0.000/
527 0.000 -7.692 1 1 0.000 0.000/
528 0.000 -7.692 1 1 0.000 0.000/
529 0.000 -7.692 1 1 0.000 0.000/
530 0.000 -7.692 1 1 0.000 0.000/
531 0.000 -7.692 1 1 0.000 0.000/
/
SUPPORT DISPLACEMENTS
*********************
<NOTE:> UNITS: MM OR IN
<<<<< FORMAT >>>>>
JNT DOF DISPL [ #JNT d(JNT) ] /
/
GRAVITY LOADS
*************
<NOTE:> UNITS: KG/M3
<<<<< FORMAT >>>>>
ELMT DENS GX GY [#ELMT d(ELMT)] [ #ELMT d(ELMT)] /
/
ELEMENT TEMPERATURE
*******************
<NOTE:> UNITS: F OR C
<<<<< FORMAT >>>>>
ELMT TEMP [ #ELMT d(ELMT) d(TEMP) ] [ #ELMT d(ELMT) d(TEMP) ] /
/
CONCRETE PRESTRAINS
*******************
<NOTE:> UNITS: me
<<<<< FORMAT >>>>>
ELMT STRAIN [ #ELMT d(ELMT) d(STRAIN) ] [ #ELMT d(ELMT) d(STRAIN) ] /
/
INGRESS PRESSURES
*****************
<NOTE:> UNITS: MPa
<<<<< FORMAT >>>>>
ELMT PRESSURE [ #ELMT d(ELMT) d(PRS) ] [ #ELMT d(ELMT) d(PRS) ] /
/
NODAL THERMAL LOADS
*******************
316
<NOTE:>
<<<<< FORMAT >>>>>
NODE TYPE Tm1 Tp1 Tm2 Tp2 Tm3 Tp3 [ #NODE d(NODE) ] [ #NODE d(NODE) ] /
/
LUMPED MASSES
*************
<NOTE:> UNITS: kg, m/s
<<<<< FORMAT >>>>>
NODE DOF-X DOF-Y MASS GF-X GF-Y Vo-X Vo-Y [ #NODE d(NODE) ] /
/
IMPULSE FORCES
**************
<NOTE:> UNITS: Sec, kN
<<<<< FORMAT >>>>>
NODE DOF T1 F1 T2 F2 T3 F3 T4 F4 [ #NODE d(NODE) ] /
/
GROUND ACCELERATION
*******************
<NOTE:> UNITS: Sec, G
<<<<< FORMAT >>>>>
TIME ACC-X ACC-Y
/

<<< LOAD FILE NOTES >>>


317
SW22DL.L2R
* * * * * * * * * * * * *
* V e c T o r 2 *
* L O A D D A T A *
* * * * * * * * * * * * *

LOAD CASE PARAMETERS


********************

Structure Title (30 char. max.) : Shear Wall SW22


Load Case Title (30 char. max.) : Horiz Disp (1 mm)
Load Case File Name (8 char. max.) : SW22DL
No. of Loaded Joints : 0
No. of Prescribed Support Displacements : 1
No. of Elements with Gravity Loads : 0
No. of Elements with Temperature Loads : 0
No. of Elements with Concrete Prestrain : 0
No. of Elements with Ingress Pressure : 0
No. of Nodes with Thermal Load : 0
No. of Nodes with Lumped Masses : 0
No. of Nodes with Impulse Forces : 0
Ground Acceleration Record (0-1) : 0

JOINT LOADS
***********
<NOTE:> UNITS: KIPS OR KN
<<<<< FORMAT >>>>>
NODE Fx Fy [ #NODE d(NODE) d(Fx) d(Fy) ] /
/
SUPPORT DISPLACEMENTS
*********************
<NOTE:> UNITS: MM OR IN
<<<<< FORMAT >>>>>
JNT DOF DISPL [ #JNT d(JNT) ] /
525 1 1.000 1 1/
/
GRAVITY LOADS
*************
<NOTE:> UNITS: KG/M3
<<<<< FORMAT >>>>>
ELMT DENS GX GY [#ELMT d(ELMT)] [ #ELMT d(ELMT)] /
/
ELEMENT TEMPERATURE
*******************
<NOTE:> UNITS: F OR C
<<<<< FORMAT >>>>>
ELMT TEMP [ #ELMT d(ELMT) d(TEMP) ] [ #ELMT d(ELMT) d(TEMP) ] /
/
CONCRETE PRESTRAINS
*******************
<NOTE:> UNITS: me
<<<<< FORMAT >>>>>
ELMT STRAIN [ #ELMT d(ELMT) d(STRAIN) ] [ #ELMT d(ELMT) d(STRAIN) ] /
/
INGRESS PRESSURES
*****************
<NOTE:> UNITS: MPa
<<<<< FORMAT >>>>>
ELMT PRESSURE [ #ELMT d(ELMT) d(PRS) ] [ #ELMT d(ELMT) d(PRS) ] /
/
NODAL THERMAL LOADS
*******************
<NOTE:>
<<<<< FORMAT >>>>>
NODE TYPE Tm1 Tp1 Tm2 Tp2 Tm3 Tp3 [ #NODE d(NODE) ] [ #NODE d(NODE) ] /
/
LUMPED MASSES
*************
<NOTE:> UNITS: kg, m/s
<<<<< FORMAT >>>>>
NODE DOF-X DOF-Y MASS GF-X GF-Y Vo-X Vo-Y [ #NODE d(NODE) ] /
/
IMPULSE FORCES
318
**************
<NOTE:> UNITS: Sec, kN
<<<<< FORMAT >>>>>
NODE DOF T1 F1 T2 F2 T3 F3 T4 F4 [ #NODE d(NODE) ] /
/
GROUND ACCELERATION
*******************
<NOTE:> UNITS: Sec, G
<<<<< FORMAT >>>>>
TIME ACC-X ACC-Y
/

<<< LOAD FILE NOTES >>>

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy