Červenka Consulting s.r.o.

Na Hrebenkach 55
150 00 Prague
Czech Republic
Phone: +420 220 610 018
E-mail: cervenka@cervenka.cz
Web: http://www.cervenka.cz

ATENA Program Documentation

Part 1

Theory

Written by
Vladimír Červenka, Libor Jendele,
and Jan Červenka

Prague, March 31st, 2021

Ackowledgements:
The software was developed with partial support of Eurostars funding program.
The following models were developed with the financial support of TA ČR:
CC3DNonLinCementitious2HPRFC, CC3DNonLinCementitious2FRC,
CC3DNonLinCementitious2SHCC, CCModelGeneral, CCTransportMaterial
Durability Anslysis module has been developed during the project TA04031458, „Software for
prediction and modelling of durability and safety of transportation structures“ funded by TA ČR
ATENA3DPrint module has been developed during the project TF04000051, “digiCON2 –
Simulation Service for Digital Concrete Production” funded by TA ČR.
ATENA modul CeSTaR – for modelling of prefabricated reinforced concrete segments with multi-
spiral reinforcement was developed under the project TF05000040 „CeSTaR - Computer simulation
and experimental validation - complex service for flexible and efficient design of pre-cast concrete“
funded by TA ČR.

Trademarks:
ATENA is registered trademark of Vladimir Cervenka.
Other names may be trademarks of their respective owners.

Copyright © 2000-2021 Červenka Consulting s.r.o.

Contents
1 CONTINUUM GOVERNING EQUATIONS 1
1.1 Introduction 1

1.2 General Problem Formulation 2

1.3 Stress Tensors 4

1.3.1 Cauchy Stress Tensor 4
1.3.2 2nd Piola-Kirchhoff Stress Tensor 4

1.4 Strain Tensors 5

1.4.1 Engineering Strain 5
1.4.2 Green-Lagrange Strain 5

1.5 Constitutive Tensor 6

1.6 The Principle of Virtual Displacements 6

1.7 The Work Done by the External Forces 8

1.8 Problem Discretisation Using Finite Element Method 9

1.9 Stress and Strain Smoothing 10

1.9.1 Extrapolation of Stress and Strain to Element Nodes 11

1.10 Simple, Complex Supports and Master-Slave Boundary Conditions. 12

1.11 References 13
2 CONSTITUTIVE MODELS 15
2.1 Constitutive Model SBETA (CCSbetaMaterial) 15
2.1.1 Basic Assumptions 15
2.1.2 Stress-Strain Relations for Concrete 18
2.1.3 Localization Limiters 24
2.1.4 Fracture Process, Crack Width 25
2.1.5 Biaxial Stress Failure Criterion of Concrete 25
2.1.6 Two Models of Smeared Cracks 27
2.1.7 Shear Stress and Stiffness in Cracked Concrete 29
2.1.8 Compressive Strength of Cracked Concrete 29
2.1.9 Tension Stiffening in Cracked Concrete 30
2.1.10 Summary of Stresses in SBETA Constitutive Model 30

ATENA Theory i
 2.1.11 Material Stiffness Matrices 31
2.1.12 Analysis of Stresses 32
2.1.13 Parameters of Constitutive Model 33

2.2 Fracture–Plastic Constitutive Model (CC3DCementitious,

CC3DNonLinCementitious, CC3DNonLinCementitious2,
CC3DNonLinCementitious2User, CC3DNonLinCementitious2Variable,
CC3DNonLinCementitious2FRC, CC3DNonLinCementitious2SHCC,
CC3DNonLinCementitious3) 34
2.2.1 Introduction 34
2.2.2 Material Model Formulation 35
2.2.3 Rankine-Fracturing Model for Concrete Cracking 35
2.2.4 Plasticity Model for Concrete Crushing 38
2.2.5 Combination of Plasticity and Fracture model 41
2.2.6 Variants of the Fracture Plastic Model 44
2.2.7 Tension Stiffening 47
2.2.8 Crack Spacing 47
2.2.9 Fixed or Rotated Cracks 48
2.2.10 Fatigue 48
2.2.11 Fiber Reinforced Concrete (FRC) Material 52
2.2.12 Strain Hardening Cementitious Composite (SHCC, HPFRCC) Material 52
2.2.13 Confinement-Sensitive Constitutive Model 55

2.3 Von Mises Plasticity Model 59

2.4 Drucker-Prager Plasticity Model 62

2.5 User Material Model 63

2.6 Interface Material Model 63

2.7 Reinforcement Stress-Strain Laws 67

2.7.1 Introduction 67
2.7.2 Bilinear Law 67
2.7.3 Multi-line Law 68
2.7.4 No Compression Reinforcement 69
2.7.5 Cyclic Reinforcement Model 69
2.7.6 Cyclic Reinforcement Model – Steel DRC 70

2.8 Reinforcement Bond Models 72

2.8.1 CEB-FIP 1990 Model Code 72

ii
 2.8.2 Bond Model by Bigaj 74
2.8.3 Memory Bond Material 75

2.9 Microplane Material Model (CCMicroplane4) 76

2.9.1 Equivalent Localization Element 77

2.10 References 80
3 FINITE ELEMENTS 85
3.1 Introduction 85

3.2 Truss 2D and 3D Element 87

3.3 Plane Quadrilateral Elements 91

3.4 Plane Triangular Elements 97

3.5 3D Solid Elements 99

3.6 Spring Element 110

3.7 Quadrilateral Element Q10 112

3.7.1 Element Stiffness Matrix 112
3.7.2 Evaluation of Stresses and Resisting Forces 115

3.8 External Cable 117

3.9 Reinforcement Bars with Prescribed Bond 118

3.10 Interface Element 124

3.11 Truss Axi-Symmetric Elements. 128

3.12 Ahmad Shell Element 130

3.12.1 Coordinate Systems. 132
3.12.2 Geometry Approximation 137
3.12.3 Displacement Field Approximation. 138
3.12.4 Strain and Stresses Definition. 139
3.12.5 Serendipity, Lagrangian and Heterosis Variant of Degenerated Shell Element. 140
3.12.6 Smeared Reinforcement 145
3.12.7 Transformation of the Original DOFs to Displacements at the Top and Bottom of the
Element Nodal Coordinate System 145
3.12.8 Shell Ahmad Elements Implemented in ATENA 149

3.13 Curvilinear Nonlinear 2D Isoparametric Layered Shell Quadrilateral Elements 149

ATENA Theory iii

 3.13.1 Geometry and displacements 150
3.13.2 Connection of the shell2D to an ambient solid element 153
3.13.3 Green-Lagrange strains 156

3.14 Curvilinear Nonlinear 2D Isoparametric Layered Shell Triangular Elements 161

3.15 Curvilinear Nonlinear 3D Isoparametric Layered Shell Hexahedral Elements 162

3.15.1 Geometry and displacements 165
3.15.2 Green-Lagrange strains 166

3.16 Curvilinear Nonlinear 3D Isoparametric Layered Shell Wedge Elements 171

3.17 Curvilinear Nonlinear 3D Beam Element 173

3.17.1 Geometry and Displacements and Rotations Fields 173
3.17.2 Strain and Stress Definition 176
3.17.3 Matrices Used in the Beam Element Formulation 176
3.17.4 The Element Integration 183

3.18 Curvilinear Nonlinear 3D Isoparametric Beam Element 185

3.19 Curvilinear Nonlinear 1D element 187

3.19.1 Connection of the beam1D to an ambient solid element 187

3.20 Integrated forces and moments for shells 189

3.21 Integrated forces and moments for beams 191

3.22 Global and Local Coordinate Systems for Element Load 191

3.23 Digital printing of concrete structures 194

3.23.1 Simplified strength and stability assessments of extruded structures 195
3.23.2 Steps to carry on analyses of extruded structures 209

3.24 References 214

4 SOLUTION OF NONLINEAR EQUATIONS 215
4.1 Linear Solvers 215
4.1.1 Direct Solver 216
4.1.2 Direct Sparse Solver 217
4.1.3 Iterative Solver 217
4.1.4 Parallel Direct Sparse Solver PARDISO 221

4.2 Full Newton-Raphson Method 223

iv
4.3 Modified Newton-Raphson Method 225

4.4 Arc-Length Method 226

4.4.1 Normal Update Method 229
4.4.2 Consistently Linearized Method 229
4.4.3 Explicit Orthogonal Method 230
4.4.4 The Crisfield Method. 231
4.4.5 Arc Length Step 232

4.5 Line Search Method 232

4.6 Parameter  233

4.7 Band Width Optimization 235

4.8 References 238

5 CREEP AND SHRINKAGE ANALYSIS 241
5.1 Implementation of Creep and Shrinkage Analysis in ATENA 241
5.1.1 Basic Theoretical Assumptions 241

5.2 Approximation of Compliance Functions  (t , t ') by Dirichlet Series. 243

5.3 Step by Step Method 244

5.4 Integration and Retardation Times 245

5.5 Creep and Shrinkage Constitutive Model 247

5.6 References 256

6 DURABILITY ANALYSIS 259
6.1 Carbonation 260
6.1.1 Example of Carbonation 261

6.2 Chlorides 261

6.3 Diffusion coefficient for chlorides 263

6.4 MODELS for PROPAGATION PHASE 266

6.4.1 Carbonation during propagation phase 266
6.4.2 Chloride ingress during propagation phase 267
6.4.3 Cracking of concrete cover 267
6.4.4 Spalling of concrete cover 268

ATENA Theory v
6.5 Alkali-Aggregate Reaction 268
6.5.1 Introduction of alkali-aggregatea model for concrete 268
6.5.2 Model for ASR kinetics 270
6.5.3 Prediction of ASR swelling  cal

273
6.5.4 Influence of moisture FM 274
6.5.5 ASR for 3D conditiions 275
6.5.6 Validation on free expansion 277
6.5.7 Implementation in Atena 279
6.5.8 Comments 281

6.6 References 282

7 TRANSPORT ANALYSIS 285
7.1 Numerical Solution of the Transport Problem – Spatial Discretisation 288

7.2 Numerical Solution of the Transport Problem – Temporal Discretisation 294

7.2.1  -parameter Crank Nicholson Scheme 295
7.2.2 Adams-Bashforth Integration Scheme 295
7.2.3 Reduction of Oscillations and Convergence Improvement 296

7.3 Material Constitutive Model 296

7.4 Fire Element Boundary Load 312

7.4.1 Hydrocarbon Fire 312
7.4.2 Fire Exposed Boundary 313
7.4.3 Implementation of Fire Exposed Boundary in ATENA 313

7.5 Moisture-Heat Element Boundary Load 314

7.6 References 316

8 DYNAMIC ANALYSIS 319
8.1 Structural Damping 323

8.2 Spectral analysis 325

9 EIGENVALUES AND EIGENVECTORS ANALYSIS 329
9.1 Inverse Subspace Iteration 329
9.1.1 Rayleigh-Ritz Method 330
9.1.2 Jacobi Method 330
9.1.3 Inverse Iteration Method 331

vi
 9.1.4 Algorithm of Inverse Subspace Iteration 332
9.1.5 Sturm Sequence Property Check 334

9.2 References 334

10 GENERAL FORM OF DIRICHLET BOUNDARY CONDITIONS 335
10.1 Theory Behind the Implementation 335
10.1.1 Single CBC 336
10.1.2 Multiple CBCs 338

10.2 Application of Complex Boundary Conditions 341

10.2.1 Finite Element Mesh Refinement 341
10.2.2 Mesh Generation Using Sub-Regions 342
10.2.3 Discrete Reinforcement Embedded in Solid Elements 343
10.2.4 Curvilinear Nonlinear Beam and Shell Elements 344

10.3 References 345

INDEX 347

ATENA Theory vii

1 CONTINUUM GOVERNING EQUATIONS

1.1 Introduction
This chapter presents the general governing continuum equations for nonlinear analysis. In
general, there exist many variants of nonlinear analysis depending on how many nonlinear
effects are accounted for. Hence, this chapter first introduces some basic terms and entities
commonly used for nonlinear structural analyses, and then it concentrates on formulations that
are implemented in ATENA.
It is important to realize that the whole structure does not have to be analyzed using a full
nonlinear formulation. However, a simplified (or even linear) formulation can be used in many
cases. It is a matter of engineering knowledge and practice to assess, whether the inaccuracies
due to a simplified formulation are acceptable or not.
The simplest formulation, i.e., linear formulation, is characterized by the following assumptions:
The constitutive equation is linear, i.e., the generalized form of Hook's law is used.
The geometric equation is linear; that is, the quadratic terms are neglected. It means that during
analysis, we neglect the change of shape and position of the structure.
Both loading and boundary conditions are conservative, i.e., they are constant throughout the
whole analysis irrespective of the structural deformation, time etc.
Generally, linear constitutive equations can be employed for a material, which is far from its
failure point, usually up to 50% of its maximum strength. Of course, this depends on the type of
material, e.g., rubber needs to be considered as a nonlinear material earlier. But for usual civil
engineering materials, the previous assumption is satisfactory.
Geometric equations can be considered linear if the deflections of a structure are much smaller
than its dimensions. This must be satisfied not only for the whole structure but also for its parts.
Then the geometric equations for the loaded structure can then be written using the original
(unloaded) geometry.
It is also important to realize that a linear solution is permissible only in the case of small strains.
This is closely related to the material property because if strains are high, the stresses are usually,
although not necessarily, high as well.
Despite the fact that for the vast majority of structures linear simplifications are quite acceptable,
there are structures when it is necessary to take into account some nonlinear behavior. The
resulting governing equations are then much more complicated, and normally they do not have a
closed-form solution. Consequently, some nonlinear iterative solution schemes must be used (see
Chapter Solution of Nonlinear Equations further in this document).
Nonlinear analysis can be classified according to a type of nonlinear behavior:
Nonlinear material behavior only needs to be accounted for. This is the most common case for
ordinary reinforced concrete structures. Because of serviceability limitations, deformations
are relatively small. However, the very low tensile strength of concrete needs to be accounted
for.
Deformations (either displacements only or both displacements and rotations) are large enough
such that the equilibrium equations must use the deformed shape of the structure. However,
the relative deformations (strains) are still small. The complete form of the geometric

ATENA Theory 1
 equations, including quadratic terms, has to be employed, but constitutive equations are
linear. This group of nonlinear analyses includes most stability problems.
The last group uses nonlinear both material and geometric equations. In addition, it is usually not
possible to suddenly apply the total value of load, but it is necessary to integrate in time
increments (or loading increments). This is the most accurate and general approach but
unfortunately, is also the most complicated.
There are two basic possibilities for formulating the general structural behavior based on its
deformed shape:
Lagrange formulation:
In this case, we are interested in the behavior of infinitesimal particles of volume dV . Their
volume will vary dependent on the loading level applied and, consequently, on the extent of
current deformations. This method is usually used to calculate civil engineering structures.
Euler formulation:
The essential idea of Euler's formulation is to study the "flow" of the structural material through
infinitesimal and fixed volumes of the structure. This is the favorite formulation for fluid
analysis, analysis of gas flow, tribulation etc., where large material flows exist.
For structural analysis, however, the Lagrangian formulation is better, and therefore the attention
will be restricted to this. Two forms of the Lagrangian formulation are possible. The governing
equations can either be written with respect to the original undeformed configuration at time t =
0 or with respect to the most recent deformed configuration at time t. The former case is called
Total Lagrangian formulation (TL), while the latter one is called the Updated Lagrangian
formulation (UL).
It is difficult to say which formulation is better because both have their advantages and
drawbacks. Usually, it depends on a particular structure being analyzed and which one to use is a
matter of engineering judgment. Generally, provided the constitutive equations are adequate, the
results for both methods are identical.
ATENA currently uses the Updated Lagrangian formulation (which is described later in this
chapter) and supports the highest, i.e., 3rd level of nonlinear behavior. Soon, it should also
support Total Lagrangian formulation.

1.2 General Problem Formulation

A general analysis of a structure usually consists of the application of many small load
increments. At each of those increments, an iterative solution procedure has to be executed to
obtain a structural response at the end of the increment. Hence, denoting the start and end of the
load increment by t and t  t , at each step, we know the structural state at the time t (from the
previous steps) and solve for the state at the time t  t . This procedure is repeated as many
times as necessary to reach the final (total) level of loading.
This process is depicted in Fig. 1-1. At the time t  0 the volume of the structure is 0V , the
surface area is 0 S , and any arbitrary point M has coordinates 0 X 1 , 0 X 2 , 0 X 3 . Similarly, at the
time t the same structure has a volume tV , surface area t S , and coordinates of the point M are
t
X 1 , t X 2 , t X 3 . A similar definition applies for the time t  t by replacing index t by t  t .

2
 [0X3, tX3, t+tX3]

M0 M1 M2
Configuration t+t
Configuration 0 Configuration t

0 t t+t
[0X2, tX2, t+tX2]
[ X1, X1, X1]

Fig. 1-1 The movement of body of structure in Cartesian coordinate system.

For the derivations of nonlinear equations, it is important to use clear and simple notations. The
same system of notation will be used throughout this document:
Displacements u are defined in a similar manner to that adopted for coordinates, hence t ui is the
i -th element of the displacement vector at the time t ,
ui  t t X i  t X i is i -th element of the vector of displacement increments at the time t ,
The left superscript denotes the time corresponding to the value of the entity, the left subscript
denotes the configuration with respect to which the value is measured, and subscripts on the
right identify the relationships to the coordinate axis. Thus, for example t 0t ij denotes
element i , j of stress tensor  at the time t  t with respect to the original (undeformed)
configuration.
For derivatives, the abbreviated notation will be used, i.e., all right subscripts that appear after a
comma declare derivatives. For example:
t t  t t
u 
0 i, j ui (1.2)
X j
The general governing equations can be derived in the form of a set of partial differential
equations (for example, using the displacement method), or an energy approach can be used. The
final results are the same.
One of the most general methods of establishing the governing equations is to apply the principle
of virtual work. There are three basic variants of this:
The principle of virtual displacements,
The principle of virtual forces,
The Clapeyron divergent theorem.
Using the virtual work theorems, it is possible to derive several different variation principles
(Lagrange principle, Clapeyron principle, Hellinger-Reissner principle, Hu-Washizu principle
etc.). There are popular especially in linear analysis. They can be used to establish equilibrium
equations, to study possible deformation modes in finite element discretization etc.
Unfortunately, in the nonlinear analysis, they do not always work.

ATENA Theory 3
In this document, all the following derivations will be presented in their displacement forms, and
consequently, the principle of virtual displacements will be used throughout.
The following section deals with the definition of the stress and strain tensors, which are usually
used in nonlinear analysis. All of them are symmetric.

1.3 Stress Tensors

1.3.1 Cauchy Stress Tensor

This tensor is well known from linear mechanics. It expresses the forces that act on
infinitesimally small areas of the deformed body at time t. Sometimes, these are also called an
"engineering" stress. The Cauchy stress tensor is the main entity for checking ultimate stress
values in materials. In the following text, it will be denoted by  . It is energetically conjugated
with an Engineering strain tensor described later.

1.3.2 2nd Piola-Kirchhoff Stress Tensor

The 2nd Piola-Kirchhoff tensor is a fictitious entity, having no physical representation of it as in
the case of the Cauchy tensor. It expresses the forces, which act on infinitesimal areas of the
body in the undeformed configuration. Hence it relates forces to the shape of the structure, which
no longer exists.
The mathematical definition is given by:
0
 0
Sij  t X i , m  mn t X j ,n
t t 0
(1.3)

0 t

where
0

is the ratio of density of the material at time 0 and t ,
t

t
 mn is the Cauchy stress tensor at time t ,
0
t X i , m is the derivative of coordinates, ref. (1.5).

Using inverse transformation, we can express Cauchy stress tensor in terms of the 2nd Piola-
Kirchhoff stress tensor, i.e.:
t
 t
 mn 
t t t
0 X m ,i 0 S mn 0 X n , j (1.4)
0

The elements 0t X i ,m are usually collected in the so-called Deformation gradient matrix:

X   0 t X T 
t T
0 (1.5)

where:

T
    
0   0
T
, 0 , 0 
  X1  X 2  X 3 
t
X T   t X 1 , t X 2 , t X 3 

4
 0

The ratio can be computed using:
t

0
  t  det( 0t X ) (1.6)
Expression (1.6) is based on the assumption that the weight of an infinitesimal particle is
constant during the loading process.
Some important properties can be deduced from the definition of 2nd Piola-Kirchhoff tensor (1.3)
:
at time 0, i.e., the undeformed configuration, there is no distinction between 2nd Piola-Kirchhoff
0

and Cauchy stress tensors because 00 X  E , i.e., unity matrix and the density ratio = 1.,
t

2nd Piola-Kirchhoff tensor is an objective entity in the sense that it is independent of any
movement of the body provided the loading conditions are frozen. This is a very important
property. The Cauchy stress tensor does not satisfy this because it is sensitive to the rotation
of the body. It is energetically conjugated with the Green-Lagrange tensor described later.
They're some other stress tensor commonly used for nonlinear structural analysis, e.g., Jaumann
stress rate tensor (describes stress rate rather than its final values) etc.; however, they are not
used in ATENA and therefore not described in this document.

1.4 Strain Tensors

1.4.1 Engineering Strain

It is the most commonly used strain tensor, comprising strains that are called Engineering strains.
Its main importance is that it is used in linear mechanics as a counterpart to the Cauchy stress
tensor.
1  u u 
e
t mn   t m  t n  (1.7)
2   Xn  Xm 

1.4.2 Green-Lagrange Strain

This is the energy conjugate of the 2nd Piola-Kirchhoff tensor and its properties are similar (i.e.,
objective etc.). It is defined as:
1 t
 
t
0 ij
2
 0 ui , j  0 u j ,i  0 uk ,i 0 uk , j 
t t t
(1.8)

If we calculate the length of an infinitesimal fibber prior to and after deformation in the original
coordinates, we get exactly the terms of the Green-Lagrange tensor.
The following equation gives a relation between variation of Green-Lagrange and Engineering
strain tensors:
t X m t X n
  0t  ij     t emn  (1.9)
0 X i 0 X j

ATENA Theory 5
These are the strain tensors used in ATENA. From the other strain tensors commonly used in the
nonlinear analysis we can mention Almansi strain tensor, co-rotated logarithmic strain, strain
rate tensor etc.

1.5 Constitutive Tensor

Although the whole chapter later in this document is dedicated to the problem of constitutive
equations and to material failure criteria, assume for the moment that stress-strain relation can
be written in the following form:
t
0 Sij 0t Cijrs 0t  rs (1.10)

where t0 Cijrs is the constitutive tensor.

This form is applicable for linear materials, or in its incremental form, it can also be used for
nonlinear materials. The following important relations apply for transformation from coordinates
to time 0 to coordinates at the time t :
t
t
t
Cmnpq  t t t t
0 xm ,i 0 xn , j 0Cijrs 0 x p , r 0 xq , s (1.11)
t 0

or in the other direction
0
0
t
Cijrs  x 0 x tC 0
x 0x (1.12)
0 t
 t i ,m t j ,n t mnpq t r , p t s ,q
Using constitutive tensor (1.11) and Almansi strains tt  , we can write for Cauchy stresses (with
respect to coordinates at time t ):
t
 ij tt Cijrs tt rs (1.13)

Almansi strains are defined (related to Green-Lagrange strains t0  ij by

t
t  mn t0 xi ,m t0 x j ,n 0t  ij (1.14)

or can be calculated directly:

1 t
t
 
t ij
2
 t ui , j  t u j ,i  t uk ,i t uk , j 
t t t
(1.15)

The equation (1.13) is equivalent to the equation (1.10) that was written for the original
configuration of the structure. It is very important to know, with respect to which coordinate
system the stress, strain, and constitutive tensors are defined, as the actual value can significantly
differ. ATENA currently assumes that all these tensors are defined at coordinates at time t .

1.6 The Principle of Virtual Displacements

This section presents how the principle of virtual displacement can be applied to the analysis of
a structure. For completeness, both the Lagrangian Total and Updated formulations will be
discussed. In all derivations, it is assumed that the response of the structure up to time t is
known. Now, at the time t  t we apply load increment and using the principle of virtual
displacement will solve for the state of the structure at t  t .
Virtual work of the structure yields the following. For Total formulation:

6
  Sij   
 ij  dV  0t  dt R
t  dt t  dt
0 0 (1.16)
0
V

for Updated formulation:

 Sij   
 ij  dV  tt  dt R
t  dt t  dt
t t (1.17)
t
V

where 0V , t V denotes the structure volume corresponding to time 0 and t t  dt R is the total
virtual work of the external forces. The symbol  denotes variation of the entity. Since energy
must be invariant with respect to the reference coordinate system (1.16) and (1.17) must lead to
identical results.
Substituting expressions for strain and stress tensors, the final governing equation for structure
can be derived. They are summarized in (1.18) through (1.29). Note that the relationships are
expressed with respect to configurations at an arbitrary time t and an iteration ( i ) . Typically, the
time t may by 0 , in which case we have Total Lagrangian formulation or t  t (i  1) , in which
case, we have Updated Lagrangian formulation, where some terms can be omitted. ATENA also
supports "semi" Updated Lagrangian formulation when t conforms to time at the beginning of
time increment, i.e., the beginning of load step. The following table compares the above-
mentioned formulations:
Table 1.6-1 Comparison of different Lagrangian formulation.

Transform each Transform each load Transform Calculate

Lagrangian iteration increment stress and
u for t eij
t t ( i 1)
t i, j
strain for
formulation IP state Material IP state Material
output
variables properties variables properties
Total No No No No Yes Yes
Updated Yes Yes Yes Yes No No
"Semi"- No No Yes Yes No Yes
Updated
Governing equations:


t t
t Sij( i )   t tt ij( i )  t dV  t t R (1.18)
t
V

nd
where 2 Piola-Kirchhoff stress and Green Lagrange strain tensor are:
t

t t
Sij( i )  t ( i ) t t ( i )
x
t t i ,m t
t mn t t x (ji,n) (1.19)

t t t

1 t t ( i )
 t tt ij( i )  
2
 t ui, j  t t ( i )
u 
t i, j
t t ( i ) t t ( i )
u
t k ,i u
t k, j  (1.20)

The stress and strain increments:

t t ( i )
S  t tt Sij(i 1)  t Sij( i )
t ij (1.21)

ATENA Theory 7
 t t ( i )
t ij   
t t ( i 1)
t ij  t  ij( i )
(1.22)
t  (i )
ij  e  t
(i )
t ij
(i )
ij

where linear part of the strain increment is calculated by:

1
e(i )
t ij 
2
 t ui , j  t ui , j 
(i ) (i ) t t ( i 1)
t uk ,i t uk , j 
(i )
t u k , j t u k ,i 
t t ( i 1) (i )
(1.23)

and nonlinear part by:

1
t ij( i ) 
2
 t u k ,i t u k , j 
(i ) (i )
(1.24)

Using constitutive equations in form:

t t ( i )
S
t ij  t Cijrs t  rs( i ) (1.25)
(i )
where t Cijrs is tangent material tensors and noting that   t  tt ij( i )     t  ij( i )  , an incremental form of
(1.18) can be derived:

e  tij( i )    t eij( i )  tij( i )  dV  Sij( i )   t eij( i )  tij( i )  dV t t R

t t
 
(i ) t t
t Cijrs t ij t (1.26)
t
t t
V V

After linearisation, i.e., neglecting 2nd order terms in (1.26):

e  tij( i )    t eij( i )  tij( i )  dV  Cijrs t t eij( i )   t eij( i )  dV

t t
 
(i )
t Cijrs t ij t (1.27)
t
t t
V V

we arrive to the final form of the governing equations:

Cijrs t ers( i )   t eij( i )  dV  Sij( i 1)   tij( i )  dV 

t t
 
t t
t t
t t
V V
(1.28)
 e  dV
t

t t t t ( i 1)
R t S ij
(i )
t ij
t
V

Note that the term   t eij( i )     t eij  is constant, i.e., independent of t ui( i ) , hence it is on RHS of
(1.28).

1.7 The Work Done by the External Forces

So far only the incremental virtual internal work has been considered. This work has to be
balanced by the work done by the external forces. It is calculated as follows:
 2 t tt ui( i 1)
t t
R 
t t
fbi   t u  dV  
i t t
fsi   t u  dS   i t
 dV (1.29)
t
V t
S t
V
t 2

where fbi and fsi are body and surface forces, t S and t V denotes integration with respect to the
surface with the prescribed boundary forces and volume of the structure (at the time and t ).
The 1st integral in (1.29) accounts for external work on a surface (e.g., external forces), the
second one for work done by body forces (e.g., weight), and the last one accounts for work done
by inertia forces, which are applicable only for dynamic analysis problems).
8
At this point, all the relationships for incremental analysis have been presented. In order to
proceed further, the problem must be discretized and solved by iterations (described in Chapter
Solution of Nonlinear Equations).

1.8 Problem Discretisation Using Finite Element Method

Spatial discretization consists of discretizing the primary variable, (i.e., deformation in case of
ATENA) over the domain of the structure. It is done in ATENA by the Finite Element Method.
The domain is decomposed into many finite elements, and at each of these elements, the
deformation field is approximated by
t
ui  h j t uij (1.30)

where
j is the index for finite element node, j  1...n ,
n is the number of element nodes,
h j are interpolation function usually grouped in matrix H j   h1 ( r , s, t ), h2 ( r , s, t ).....hn ( r , s, t )  ,
r , s, t are the local element coordinates.
The interpolation functions h j are usually created in the way that h j  1 at the node j and
h j  0 at any other element nodes.

Combining (1.30) and equation for strain definition (1.8) it can be derived:
t t
t  ( i )   t BL 0  t tt BL( i11)  t tt BNL  U
( i 1) t t ( i )
(1.31)

where
t
t t ( i )
is the vector of Green-Lagrange strains,
t t
U ( i ) is the vector of displacements,
t t ( i 1) t t ( i 1) st
t BL 0 , t BL1 , t BNL are linear strain-displacements transformation matrices (the 1 two of
them) and nonlinear strain-displacements transformation matrix (the last one).
A similar equation can also be written for stress tensor.
t t
t S (i )  t t
t C (i ) t t
t  (i ) (1.32)
where:
t t ( i )
tS is vector of 2nd Piola-Kirchhoff stress tensor and
t t
t C ( i ) is incremental stress-strain material properties matrix.
Applying the above discretization for each finite element of the structure and assembling the
results, the continuum based governing equations in (1.28) can be re-written in the following
form:
 t t
t
M U ( i )  ( t K L  t tt K NL
( i 1)
)  t tU ( i )  t t R  t t F ( i 1) (1.33)
t 2
where

t K L is the linear strain incremental stiffness matrix,

t t ( i 1)
t K NL is the nonlinear strain incremental stiffness matrix,
t
M is the structural mass matrix,

ATENA Theory 9
 t tU ( i ) is the vector of nodal point displacements increments at the time t  t , iteration i ;
 t t ( i )
t+ t

t 2
 U  is the vector of nodal accelerations,
t t
R , t t F ( i 1) is the vector of applied external forces and internal forces,
(i )
, ( i1) superscripts indicate iteration numbers.
Note that (1.33) also contains inertial term needed only for dynamic analysis. Finite element
matrices in (1.33) and corresponding analytical expressions are summarized:
 
Cijrs t ers( i )   t eij( i )  dV
t
K L U ( i )    t BLT t C t BL dV  U ( i )  
t t  t
V  t
V

 
Sij( i 1)   tij( i )  dV
t
t t ( i 1)
K NL U ( i )    t t ( i 1) T
BNL t t
Sij( i 1) t t ( i 1)
BNL dV  U ( i )   t t
t t t t t  t
V  t
V

Sij( i 1)   t eij( i )  dV

t
t t
F ( i 1)   t t
t Sij( i 1) dV   t t
t (1.34)
t t
V V

H H
t t T t t T t t
R t f A dV dA  t f B dV  t t R
t t
A V

     t t ui( i ) t   t t ui( i ) 

t
M 2 t t
U (i )
   H T t  H dV  2 U 
t t (i )
   dV
t t  t t 2  t
2

V  t
V

1.9 Stress and Strain Smoothing

All derivations and solution procedures in ATENA software are based on the deformational form
of the finite element method. Any structure is solved using the weak (or integral) form of
equilibrium equations. The whole structure is divided into many finite elements, and
displacement u at each particular element (at any location) is approximated by approximation
functions hi and element displacements u i as follows: u   hi u i , ( i is index of an element
i
node). It is important to note that in order not to lose any internal energy of the structure, the
displacements over the whole structure must be continuous. The continuity within finite
elements is trivial. The use of continuous approximation functions h j ensures this requirement.
A bit more complicated situation is on boundaries between adjacent elements; however, if the
adjacent elements are of the same type, their displacements are also continuous. Note that there
exist are some techniques that alleviate the continuity requirement, but in ATENA they are not
used.
Unlike displacements, stress and strain fields are typically discontinuous. Moreover, a structure
is investigated within so-called material (or integral) points, which are points located somewhere
within each element. Their position is derived from the requirement to minimize the
approximation error. In other words, the standard finite element method provides stress and
strain values only at those material points, and these values must be later somehow extrapolated

10
into element nodal points. Often, some sort of smoothing is required in order to remove the
mentioned stress and strain discontinuity. This section describes how this goal is done in
ATENA.
There are two steps in the process of stress and strain smoothing: 1/ extrapolation of stress and
strain from material points to element nodes and 2/ averaging of stress in global node. The whole
technique is described briefly. All details and derivations can be found e.g. (ZIENKIEWICZ,
TAYLOR 1989) and ČERVENKA et. al. 1993.

1.9.1 Extrapolation of Stress and Strain to Element Nodes

The extrapolation is done as follows (for each component of structural stress  and strain  ).

Let us define a vector of stresses xx at element nodes i such as xx  xx ,1 , xx ,2 ,....xx ,n  ,
T

where the 2nd index indicates element node number. Let us also define a vector
Pxx   Pxx ,1 , Pxx ,2 ,....Pxx ,n  , whose component are calculated
T

Pxx ,i   hi  xx d  e (1.35)
e

The nodal value xx (with values of  xx at nodes i =1..n ) is then calculated as follows:

xx   M  Pxx
inv
(1.36)

where:

M ij   hi h j d  e (1.37)
e

In the above  xx is an extrapolated field of the stress  xx calculated by FEM. It is typically

discontinuous. n is the number of element nods, e is the volume of the investigated finite
element. The same strategy is also used for the remaining stress and strain components.
This smoothing technique is called variational as it is base on averaging energy over the element.
In addition, ATENA also supports another way of extrapolating vales from integration points to
element nodes. In this case, (1.37) is assumed to be a "lumped" diagonal matrix in order to
eliminate the need for solving a system of linear equations. The process of lumping is
characterized as follows:
M ij   hi  hk  ij d e (1.38)
e
k 1, n

As most element space approximations satisfy h

k 1, n
k  1 , the above equation is simplified to:

M ij   hi  ij d  (1.39)
e

where  ij is Kronecker delta. This "lumped" formulation ATENA uses by default.

The above values are output as nodal element stress/strain values. It follows to calculate
   
averaged stress/stain value  i   xx ,  yy ,..... xz i in a global node i that is participated by all
elements k with an incidence at the global node i .

ATENA Theory 11
 
   i ek
i  k
(1.40)
 k
ek

where is the vector of stresses i  xx , yy ,.....xz i at a node i ,  ek is the volume of the
element k that has the incidence of global node i . It should be noted that in ATENA, the same
extrapolation techniques are used for other integration point quantities as well such as: fracturing
strains, plastic strains and others.

1.10 Simple, Complex Supports and Master-Slave Boundary

Conditions.
Simple support and complex support boundary conditions represent boundary conditions of
Dirichlet types, i.e., boundary conditions that prescribe displacements. On the other hand, Simple
load boundary conditions are an example of von Neumann type boundary conditions when
forces are prescribed.
Let K is structural stiffness matrix, u is the vector of nodal displacements, and R is a vector of
nodal forces. Further, let u is subdivided into the vector of free degrees of freedom uN (with
von Neumann boundary conditions) and constrained degrees of freedom uD (with Dirichlet
boundary conditions):

u 
u  N (1.41)
 uD 
The problem governing equations can then be written:

 K NN K ND   uN   RN 
K  (1.42)
 DN K DD   uD   RD 
ATENA software supports that any constrained degree of freedom can be a linear combination
of other degrees of freedom plus some constant term:
uDi  uDi ,0    k uNk (1.43)
k

where uDi ,0 is the constant term and  k are coefficients of the linear combination. Of course, the
equation (1.43) can also include the term  u
l
l
l
D ; however, it is transformed into the constant

term.
The free degrees of freedom are then solved by
uN   K NN 
1
R N  K ND RD  (1.44)

and the dependent RD is solved by

RD  K DN uN  K DD uD (1.45)

12
The ATENA simple support boundary conditions mean that the boundary conditions use only
constant terms are uDi ,0 , (i.e.  k  0 ). The complex support boundary conditions use the full
form of (1.43).
The boundary conditions as described above allow to specify for one degree of freedom either
Dirichlet, or von Neumann boundary condition, but not both of them at the same time. It comes
from the nature of the finite element method. However, ATENA can also deal with this case of
more complex boundary conditions by introducing Lagrange multipliers. The derivation of the
theory behind this kind of boundary conditions is beyond the scope of this manual. Details can
be found elsewhere, e.g., in (Bathe 1982). To apply this type of boundary conditions in ATENA,
specify for those degrees of freedom both simple load and complex support boundary condition,
the latter one with the keyword "RELAX" keyword in its definition.
A useful feature of ATENA is that at any time, it stores in RAM only K NN and all the
elimination with the remaining blocks of K is done at element level at the process of assembling
the structural stiffness matrix.
A special type of complex boundary conditions of the Dirichlet type are so-called master-slave
boundary conditions. Such a boundary condition specifies that all (available) degrees of one
finite node (i.e., slave node) are equal to degrees of freedom of another node (i.e., master node).
If more master nodes are specified, then these master nodes are assumed to form a finite element
and degrees of freedom of the slave node are assumed to be a node within that element. Its
(slave) degrees of freedom are approximated by element nodal (i.e., master) degrees of freedom
in the same way as displacements approximation within a finite element. The coefficients  k in
(1.43) are thus calculated automatically. This type of boundary condition is used for example, for
fixing discrete reinforcement bars to the surrounding solid element.

1.11 References
BATHE, K.J. (1982), Finite Element Procedures In Engineering Analysis, Prentice-Hall, Inc.,
Englewood Cliffs, New Jersey 07632, ISBN 0-13-317305-4.
ČERVENKA, J., KEATING, S.C., AND FELIPPA, C.A. (1993), "Comparison of strain
recovery techniques for the mixed iterative method", Communications in Numerical Methods in
Engineering, Vol. 9, 925-932.
ZIENKIEWICZ, O.C., TAYLOR, R.L., (1989), The Finite Element Method, Volume 1,
McGraw-Hill Book Company, ISBN 0-07-084174-8.

ATENA Theory 13
14
2 CONSTITUTIVE MODELS

2.1 Constitutive Model SBETA (CCSbetaMaterial)

2.1.1 Basic Assumptions

2.1.1.1 Stress, Strain, Material Stiffness

The formulation of constitutive relations is considered in the plane stress state. A smeared
approach is used to model the material properties, such as cracks or distributed reinforcement.
This means that material properties defined for a material point are valid within a certain
material volume, which is in this case associated with the entire finite element. The constitutive
model is based on the stiffness and is described by the equation of equilibrium in a material
point:

s  De, s   x ,  y ,  xy  , e   x ,  y ,  xy 
T T
(2.1)

where s, D and e are a stress vector, a material stiffness matrix and a strain vector, respectively.
The stress and strain vectors are composed of the stress components of the plane stress state
 x ,  y ,  xy , Fig. 2-1, and the strain components  x ,  y ,  xy , Fig. 2-2, where  xy is the engineering
shear strain. The strains are common for all materials. The stress vector s and the material matrix
D can be decomposed into the material components due to concrete and reinforcement as:
s  s c  s s , D  Dc  D s (2.2)
The stress vector s and both component stress vectors sc , s s are related to the total cross section
area. The concrete stress s c is acting on the material area of concrete A c , which is approximately
set equal to the cross section of the composite material A c  A (the area of concrete occupied by
reinforcement is not subtracted).
The matrix D has a form of the Hooke's law for either isotropic or orthotropic material, as will be
shown in Section 2.1.11.

Fig. 2-1 Components of plane stress state.

ATENA Theory 15
 Fig. 2-2 Components of strain state.

The reinforcement stress vector ss is the sum of stresses of all the smeared reinforcement
components:
n
s s   s si (2.3)
i 1

where n is the number of the smeared reinforcement components. For the ith reinforcement, the
global component reinforcement stress ssi is related to the local reinforcement stress  si, by the
transformation:
s si  T pi  si, (2.4)
Asi
where pi is the reinforcing ratio pi  , Asi is the reinforcement cross section area. The local
Ac
reinforcement stress  si, is acting on the reinforcement area Asi
The stress and strain vectors are transformed according to the equations bellow in the plane
stress state. New axes u, v are rotated from the global x, y axes by the angle The angle is
positive in the counterclockwise direction, as shown in Fig. 2-3.

Fig. 2-3 Rotation of reference coordinate axes.

The transformation of the stresses:
s (u )  T s ( x ) (2.5)

16
  cos( ) 2 sin( ) 2 2 cos( ) sin( ) 
 
T   sin( ) 2
cos( ) 2
2 cos( ) sin( )  (2.6)
  cos( ) sin( ) cos( ) sin( ) cos( )2  sin( ) 2 
 

s ( u )   u ,  v , uv  , s ( x )   x ,  y ,  xy 
T T

The transformation of the strains:

e( u )  T e ( x ) (2.7)

 cos( ) 2 sin( ) 2 cos( ) sin( ) 

 
T   sin( ) 2
cos( ) 2
 cos( ) sin( )  (2.8)
 2 cos( ) sin( ) 2 cos( ) sin( ) cos( ) 2  sin( ) 2 
 

e(u )   u ,  v ,  uv  , e( x )   x ,  y ,  xy  .
T T

The angles of principal axes of the stresses and strains, Fig. 2-1, Fig. 2-2, are found from the
equations:
2 xy  xy
tan(2 )  , tan(2 )  (2.9)
 x  y x y
where  is the angle of the first principal stress axis and  is the angle of the first principal
strain axis.
In case of isotropic material (un-cracked concrete) the principal directions of the stress and
strains are identical; in case of anisotropic material (cracked concrete) they can be different. The
sign convention for the normal stresses, employed within this program, uses the positive values
for the tensile stress (strain) and negative values for the compressive stress (strain). The shear
stress (strain) is positive if acting upwards on the right face of a unit element.

2.1.1.2 Concept of Material Model SBETA

The material model SBETA includes the following effects of concrete behavior:
non-linear behavior in compression including hardening and softening,
fracture of concrete in tension based on the nonlinear fracture mechanics,
biaxial strength failure criterion,
reduction of compressive strength after cracking,
tension stiffening effect,
reduction of the shear stiffness after cracking (variable shear retention),
two crack models: fixed crack direction and rotated crack direction.
Perfect bond between concrete and reinforcement is assumed within the smeared concept. No
bond slip can be directly modeled except for the one included inherently in the tension stiffening.
However, on a macro-level a relative slip displacement of reinforcement with respect to concrete
over a certain distance can arise if concrete is cracked or crushed. This corresponds to a real
mechanism of bond failure in case of the bars with ribs.
The reinforcement in both forms, smeared and discrete, is in the uniaxial stress state and its
constitutive law is a multi-linear stress-strain diagram.

ATENA Theory 17
The material matrix is derived using the nonlinear elastic approach. In this approach the elastic
constants are derived from a stress-strain function called here the equivalent uniaxial law. This
approach is like the nonlinear hypo-elastic constitutive model, except that different laws are used
here for loading and unloading, causing the dissipation of energy exhausted for the damage of
material. The detailed treatment of the theoretical background of this subject can be found, for
example, in the book CHEN (1982). This approach can be also regarded as an isotropic damage
model, with the unloading modulus (see next section) representing the damage modulus.
The name SBETA comes from the former program, in which this material model was first used.
It means the abbreviation for the analysis of reinforced concrete in German language -
StahlBETonAnalyse.

2.1.2 Stress-Strain Relations for Concrete

2.1.2.1 Equivalent Uniaxial Law

The nonlinear behavior of concrete in the biaxial stress state is described by means of the so-
called effective stress  c , and the equivalent uniaxial strain  eq . The effective stress is in most
ef

cases a principal stress.

The equivalent uniaxial strain is introduced to eliminate the Poisson’s effect in the plane stress
state.
 ci
 eq  (2.10)
Eci
The equivalent uniaxial strain can be considered as the strain, that would be produced by the
stress  ci in a uniaxial test with modulus Eci associated with the direction i. Within this
assumption, the nonlinearity representing a damage is caused only by the governing stress  ci .
The details can be found in CHEN (1982).
The complete equivalent uniaxial stress-strain diagram for concrete is shown in Fig. 2-4.

Fig. 2-4 Uniaxial stress-strain law for concrete.

The numbers of the diagram parts in Fig. 2-4 (material state numbers) are used in the results of
the analysis to indicate the state of damage of concrete.

18
Unloading is a linear function to the origin. An example of the unloading point U is shown in
Fig. 2-4. Thus, the relation between stress  c and strain  eq is not unique and depends on a load
ef

history. A change from loading to unloading occurs when the increment of the effective strain
changes the sign. If subsequent reloading occurs the linear unloading path is followed until the
last loading point U is reached again. Then, the loading function is resumed.
The peak values of stress in compression f’cef and in tension f’tef are calculated according to the
biaxial stress state as will be shown in Sec.2.1.5. Thus, the equivalent uniaxial stress-strain law
reflects the biaxial stress state.
The above defined stress-strain relation is used to calculate the elastic modulus for the material
stiffness matrices, Sect. 2.1.11. The secant modulus is calculated as
c
s
Ec  (2.11)
 eq
It is used in the constitutive equation to calculate stresses for the given strain state, Sect. 2.1.12.
The tangent modulus Ect is used in the material matrix Dc for construction of an element stiffness
matrix for the iterative solution. The tangent modulus is the slope of the stress-strain curve at a
given strain. It is always positive. In cases when the slope of the curve is less then the minimum
value Emint the value of the tangent modulus is set Ect = Emint. This occurs in the softening ranges
and near the compressive peak.
Detail description of the stress-strain law is given in the following subsections.

2.1.2.2 Tension before Cracking

The behavior of concrete in tension without cracks is assumed linear elastic. Ec is the initial
elastic modulus of concrete, f t 'ef is the effective tensile strength derived from the biaxial failure
function, Section 2.1.5.2.
 c ef  Ec  eq , 0   c  ft ' ef (2.12)

2.1.2.3 Tension after Cracking

Two types of formulations are used for the crack opening:
A fictitious crack model based on a crack-opening law and fracture energy. This formulation
is suitable for modeling of crack propagation in concrete. It is used in combination with the
crack band, see Sect.2.1.3.
A stress-strain relation in a material point. This formulation is not suitable for normal cases of
crack propagation in concrete and should be used only in some special cases.
In following subsections are described five softening models included in SBETA material
model.

ATENA Theory 19
 (1) Exponential Crack Opening Law

Fig. 2-5 Exponential crack opening law.

This function of crack opening was derived experimentally by HORDIJK (1991).

   w 3   w w
 1   c1   exp  c2   1  c13  exp  c2  , (2.13)
f t ' ef   wc    wc  wc

Gf
wc  5.14
f t ' ef
where w is the crack opening, wc is the crack opening at the complete release of stress, is the
normal stress in the crack (crack cohesion). Values of the constants are, c1 =3, c2 =6.93. Gf is the
fracture energy needed to create a unit area of stress-free crack, f t ' ef is the effective tensile
strength derived from a failure function, Eq.(2.22). The crack opening displacement w is derived
from strains according to the crack band theory in Eq.(2.18).
(2) Linear Crack Opening Law

Fig. 2-6 Linear crack opening law.

 c ef ft ' 2G
  wc  w , wc  ' f (2.14)
ft ' ef wc ft

20
 (3) Linear Softening Based on Local Strain

Fig. 2-7 Linear softening based on strain.

The descending branch of the stress-strain diagram is defined by the strain c3 corresponding to
zero stress (complete release of stress).
(4) SFRC Based on Fracture Energy

Fig. 2-8 Steel fiber reinforced concrete based on fracture energy.

f1 f2 2G f
Parameters: c1  , c2  , wc 
ft ' ef
ft ' ef
f1  f 2
(5) SFRC Based on Strain

Fig. 2-9 Steel fiber reinforced concrete based on strain.

ATENA Theory 21
 f1 f2
Parameters: c1  ' ef
, c2 
ft f t ' ef
Parameters c1 and c2 are relative positions of stress levels, and c3 is the end strain.

2.1.2.4 Compression before Peak Stress

The formula recommended by CEB-FIP Model Code 90 has been adopted for the ascending
branch of the concrete stress-strain law in compression, Fig. 2-10. This formula enables wide
range of curve forms, from linear to curved, and is appropriate for normal as well as high
strength concrete.
kx  x 2  E
 c ef  f c ' ef ,x  ,k  o (2.15)
1  (k  2) x c Ec

Fig. 2-10 Compressive stress-strain diagram.

Meaning of the symbols in the above formula in:
cef -concrete compressive stress,
f c'ef - concrete effective compressive strength (See Section 2.1.5.1)
x - normalized strain,
 - strain,
c - strain at the peak stress f’cef ,
k - shape parameter,
Eo - initial elastic modulus,
f ' ef
Ec - secant elastic modulus at the peak stress, Ec  c .
c
Parameter k may have any positive value greater than or equal 1. Examples: k=1. linear, k=2. -
parabola.
As a consequence of the above assumption, distributed damage is considered before the peak
stress is reached. Contrary to the localized damage, which is considered after the peak.

22
2.1.2.5 Compression after Peak Stress
The softening law in compression is linearly descending. There are two models of strain
softening in compression, one based on dissipated energy, and other based on local strain
softening.

2.1.2.5.1 Fictitious Compression Plane Model

The fictitious compression plane model assumes, that compression failure is localized in a plane
normal to the direction of compressive principal stress. All post-peak compressive displacements
and energy dissipation are localized in this plane. It is assumed that this displacement is
independent on the size of the structure. This hypothesis is supported by experiments conducted
by Van MIER (1986).
This assumption is analogous to the Fictitious Crack Theory for tension, where the shape of the
crack-opening law and the fracture energy are defined and are considered as material properties.

Fig. 2-11 Softening displacement law in compression.

In case of compression, the end point of the softening curve is defined by means of the plastic
displacement wd. In this way, the energy needed for generation of a unit area of the failure plane
is indirectly defined. From the experiments of Van MIER (1986), the value of wd =0.5mm for
normal concrete. This value is used as default for the definition of the softening in compression.
The softening law is transformed from a fictitious failure plane, Fig. 2-11, to the stress-strain
relation valid for the corresponding volume of continuous material, Fig. 2-10. The slope of the
softening part of the stress-strain diagram is defined by two points: a peak of the diagram at the
maximal stress and a limit compressive strain d at the zero stress. This strain is calculated from
a plastic displacement wd and a band size L'd (see Section 2.1.3) according to the following
expression:
wd
d  c  (2.16)
L'd
The advantage of this formulation is reduced dependency on finite element mesh.

2.1.2.5.2 Compression Strain Softening Law Based on Strain.

A slope of the softening law is defined by means of the softening modulus Ed . This formulation
is dependent on the size of the finite element mesh.

ATENA Theory 23
2.1.3 Localization Limiters
So-called localization limiter controls localization of deformations in the failure state. It is a
region (band) of material, which represents a discrete failure plane in the finite element analysis.
In tension it is a crack, in compression it is a plane of crushing. These failure regions have some
dimension. However, since according to the experiments, the dimensions of the failure regions
are independent on the structural size, they are assumed as fictitious planes. In case of tensile
cracks, this approach is known as rack the “crack band theory“, BAZANT, OH (1983). Here is
the same concept used also for the compression failure. The purpose of the failure band is to
eliminate two deficiencies, which occur in connection with the application of the finite element
model: element size effect and element orientation effect.
4 noded element
y
crack
direction
2
1

Lc
Lt x

Fig. 2-12 Definition of localization bands.

2.1.3.1 Element Size Effect.

The direction of the failure planes is assumed to be normal to the principal stresses in tension
and compression, respectively. The failure bands (for tension Lt and for compression Ld) are
defined as projections of the finite element dimensions on the failure planes as shown in Fig.
2-12.

2.1.3.2 Element Orientation Effect.

The element orientation effect is reduced, by further increasing of the failure band for skew
meshes, by the following formula (proposed by CERVENKA et al. 1995).
Lt   Lt , Ld   Ld
' '


  1  ( max  1) ,    0; 45 (2.17)
45

24
An angle  is the minimal angle ( min 1 , 2  ) between the direction of the normal to the failure
plane and element sides. In case of a general quadrilateral element the element sides directions
are calculated as average side directions for the two opposite edges. The above formula is a
linear interpolation between the factor 1.0 for the direction parallel with element sides, and
  max , for the direction inclined at 45o. The recommended (and default) value of  max =1.5.

2.1.4 Fracture Process, Crack Width

The process of crack formation can be divided into three stages, Fig. 2-13. The uncracked stage
is before a tensile strength is reached. The crack formation takes place in the process zone of a
potential crack with decreasing tensile stress on a crack face due to a bridging effect. Finally,
after a complete release of the stress, the crack opening continues without the stress.
The crack width w is calculated as a total crack opening displacement within the crack band.
w   cr L't (2.18)
where  cr is the crack opening strain, which is equal to the strain normal to the crack direction in
the cracked state after the complete stress release.

Fig. 2-13 Stages of crack opening.

It has been shown that the smeared model based on the refined crack band theory can
successfully describe the discrete crack propagation in plain, as well as reinforced concrete
(CERVENKA et al. 1991, 1992, and 1995).
It is also possible, that the second stress, parallel to the crack direction, exceeds the tensile
strength. Then the second crack, in the direction orthogonal to the first one, is formed using the
same softening model as the first crack. (Note: The second crack may not be shown in a
graphical post-processing. It can be identified by the concrete state number in the second
direction at the numerical output.)

2.1.5 Biaxial Stress Failure Criterion of Concrete

2.1.5.1 Compressive Failure

A biaxial stress failure criterion according to KUPFER et al. (1969) is used as shown in Fig.
2-14. In the compression-compression stress state the failure function is

ATENA Theory 25
 Fig. 2-14 Biaxial failure function for concrete.
1  3.65a ' 
f c ' ef  f c , a  c1 (2.19)
(1  a ) 2
 c2
where  c1 ,  c 2 are the principal stresses in concrete and f’c is the uniaxial cylinder strength. In
the biaxial stress state, the strength of concrete is predicted under the assumption of a
proportional stress path.
In the tension-compression state, the failure function continues linearly from the point
 c1  0 ,  c 2  f c' into the tension-compression region with the linearly decreasing strength:
 c1
f c ' ef  f c' rec , rec  (1  5.3278 ), 1.0  rec  0.9 (2.20)
f c'
where rec is the reduction factor of the compressive strength in the principal direction 2 due to
the tensile stress in the principal direction 1.

2.1.5.2 Tensile Failure

In the tension-tension state, the tensile strength is constant and equal to the uniaxial tensile
strength f’t. In the tension-compression state, the tensile strength is reduced by the relation:
f t ' ef  f t ' ret (2.21)
where ret is the reduction factor of the tensile strength in the direction 1 due to the compressive
stress in the direction 2. The reduction function has one of the following forms, Fig. 2-15.
 c2
ret  1  0.95 (2.22)
f c'
A  ( A  1) B 
ret  , B  Kx  A, x  c '2 (2.23)
AB fc
The relation in Eq.(2.22) is the linear decrease of the tensile strength and (2.23) is the hyperbolic
decrease.

26
Two predefined shapes of the hyperbola are given by the position of an intermediate point r, x.
Constants K and A define the shape of the hyperbola. The values of the constants for the two
positions of the intermediate point are given in the following table.

type point parameters

r x A K
a 0.5 0.4 0.75 1.125
b 0.5 0.2 1.0625 6.0208

Fig. 2-15 Tension-compression failure function for concrete.

2.1.6 Two Models of Smeared Cracks

The smeared crack approach for modeling of the cracks is adopted in the model SBETA. Within
the smeared concept two options are available for crack models: the fixed crack model and the
rotated crack model. In both models the crack is formed when the principal stress exceeds the
tensile strength. It is assumed that the cracks are uniformly distributed within the material
volume. This is reflected in the constitutive model by an introduction of orthotropy.

2.1.6.1 Fixed Crack Model

In the fixed crack model (CERVENKA 1985, DARWIN 1974) the crack direction is given by
the principal stress direction at the moment of the crack initiation. During further loading this
direction is fixed and represents the material axis of the orthotropy.

ATENA Theory 27
 Fig. 2-16 Fixed crack model. Stress and strain state.
The principal stress and strain directions coincide in the uncracked concrete, because of the
assumption of isotropy in the concrete component. After cracking the orthotropy is introduced.
The weak material axis m1 is normal to the crack direction, the strong axis m2 is parallel with the
cracks.
In a general case the principal strain axes and rotate and need not to coincide with the axes
of the orthotropy m1 and m2. This produces a shear stress on the crack face as shown in Fig.
2-16. The stress components c1andc2 denote, respectively, the stresses normal and parallel to
the crack plane and, due to shear stress, they are not the principal stresses. The shear stress and
stiffness in the cracked concrete is described in Section 2.1.7.

2.1.6.2 Rotated Crack Model

In the rotated crack model (VECCHIO 1986, CRISFIELD 1989), the direction of the principal
stress coincides with the direction of the principal strain. Thus, no shear strain occurs on the
crack plane and only two normal stress components must be defined, as shown in Fig. 2-17.

Fig. 2-17 Rotated crack model. Stress and strain state.

If the principal strain axes rotate during the loading the direction of the cracks rotate, too. In
order to ensure the co-axiality of the principal strain axes with the material axes the tangent shear
modulus Gt is calculated according to CRISFIELD 1989 as
 c1   c 2
Gt  (2.24)
2(1   2 )

28
2.1.7 Shear Stress and Stiffness in Cracked Concrete
In case of the fixed crack model, the shear modulus is reduced according to the law derived by
KOLMAR (1986) after cracking. The shear modulus is reduced with growing strain normal to
the crack, Fig. 2-18 and this represents a reduction of the shear stiffness due to the crack
opening.

Fig. 2-18 Shear retention factor.

 1000 u 
 ln  
G  rg Gc , rg  c3  c1  (2.25)
c2
c1  7  333( p  0.005),c2  10  167( p  0.005),0  p  0.02
where rg is the shear retention factor, G is the reduced shear modulus and Gc is the initial
concrete shear modulus:
Ec
Gc  (2.26)
2(1   )
where Ec is the initial elastic modulus and  is the Poisson's ratio. The strain  is normal to the
crack direction (the crack opening strain), c1 and c2 are parameters depending on the reinforcing
crossing the crack direction, p is the transformed reinforcing ratio (all reinforcement is
transformed on the crack plane) and c3 is the user’s scaling factor. By default, c3=1. In ATENA
the effect of reinforcement ratio is not considered, and p is assumed to be 0.0.
There is an additional constraint imposed on the shear modulus. The shear stress on the crack
plane  uv  G is limited by the tensile strength f’t. The secant and tangent shear moduli of
cracked concrete are equal.

2.1.8 Compressive Strength of Cracked Concrete

A reduction of the compressive strength after cracking in the direction parallel to the cracks is
done by a similar way as found from experiments of VECCHIO and COLLINS 1982 and
formulated in the Compression Field Theory. However, a different function is used for the
reduction of concrete strength here, to allow for user's adjustment of this effect. This function
has the form of the Gauss's function, Fig. 2-19. The parameters of the function were derived
from the experimental data published by KOLLEGER et al. 1988, which also included data of
Collins and Vecchio (VECCHIO at al.1982)
ATENA Theory 29
 2
f c ' ef  rc f c' , rc  c  (1  c) e  (128u ) (2.27)

For the zero normal strain, there is no strength reduction, and for the large strains, the
strength is asymptotically approaching to the minimum value f c ' ef  cf c' .

Fig. 2-19 Compressive strength reduction of cracked concrete.

The constant c represents the maximal strength reduction under the large transverse strain. From
the experiments by KOLLEGGER et all. 1988, the value c = 0.45 was derived for the concrete
reinforced with the fine mesh. The other researchers (DYNGELAND 1989) found the reductions
not less than c=0.8. The value of c can be adjusted by input data according to the actual type of
reinforcing.
However, the reduction of compressive strength of the cracked concrete does not have to be
affected only by the reinforcing. In the plain concrete, when the strain localizes in one main
crack, the compressive concrete struts can cross this crack, causing so-called "bridging effect".
The compressive strength reduction of these bridges is also captured by the above model.

2.1.9 Tension Stiffening in Cracked Concrete

The tension stiffening effect can be described as a contribution of cracked concrete to the tensile
stiffness of reinforcing bars. This stiffness is provided by the uncracked concrete or not fully
opened cracks and is generated by the strain localization process. It was verified by simulation
experiments of HARTL, G., 1977 and published in the paper (MARGOLDOVA et.al. 1998).
Including an explicit tension stiffening factor would result in an overestimation of this effect.
Therefore, in the ATENA versions up to1.2.0 no explicit tension stiffening factor is possible in
the input.

2.1.10 Summary of Stresses in SBETA Constitutive Model

In the case of uncracked concrete, the stress symbols have the following meaning:
 c1 - maximal principal stress
 c 2 - minimal principal stress
(tension positive, compression negative)
In the case of cracked concrete, Fig. 2-16 stresses are defined on the crack plane:
 c1 - normal stress normal to the cracks
 c 2 - normal stress parallel to the cracks

30
  c - shear stress on the crack plane

2.1.11 Material Stiffness Matrices

2.1.11.1 Uncracked Concrete

The material stiffness matrix for the uncracked concrete has the form of an elastic matrix of the
isotropic material. It is written in the global coordinate system x and y.
 
1  0 
E  
Dc  2 
 1 0  (2.28)
1  
1  
0 0 
 2 
In the above E is the concrete elastic modulus derived from the equivalent uniaxial law. The
Poisson's ratio  is constant.

2.1.11.2 Cracked Concrete

For the cracked concrete, the matrix has the form of the elastic matrix for the orthotropic
material. The matrix is formulated in a coordinate system m1, m2, Fig. 2-16 and Fig. 2-17, which
is coincident with the crack direction. This local coordinate system is referred to the superscript
L later. The direction 1 is normal to the crack and the direction 2 is parallel with the crack. The
definition of the elastic constants for the orthotropic material in the plane stress state follows
from the flexibility relation:
 1  
 E  21 0 
E2
 1   1   1 
    12 1  
 2     0   2  (2.29)
   1 E E
 
2    
 1
 0 0
G 

First, we eliminate the orthotropic Poisson’s ratios for the cracked concrete, because they are
commonly not known. For this we use the symmetry relation 12 E2   21E1 . Therefore, in (2.29)
there are only three independent elastic constants E1 , E2 , 21 . Assuming that  21   is the
Poisson's ratio of the uncracked concrete and using the symmetry relation, we obtain
E1
 12   (2.30)
E2
L
The stiffness matrix Dc is found as the inverse of the flexibility matrix in (2.30):

   0 
D c  H  1 0 ,
L

 0 0 G  (2.31)
E
  1 ,H  E1 (1   2 )
E2

ATENA Theory 31
In the above relation E2 must be nonzero. If E2 is zero and E1 is nonzero, then an alternative
1 E
formulation is used with the inverse parameter  2 . In case that both elastic modules are
 E1
L
zero, the matrix Dc is set equal to the null matrix.
L
The matrix Dc is transformed into the global coordinate system using the transformation matrix
T from (2.8).
D c  TT D cL T (2.32)
The angle  is between the global axis x and the 1st material axis m1, which is normal to the
crack, Fig. 2-16.

2.1.11.3 Smeared Reinforcement

The material stiffness matrix of the ith smeared reinforcement is
 cos(  i )4 cos(  i )2 sin(  i ) 2 cos(  i )3 sin(  i ) 
 
D si  pi Esi cos(  i ) 2 sin(  i ) 2 sin(  i ) 4 cos(  i ) sin(  i )3  (2.33)
 cos(  i )3 sin(  i ) cos(  i ) sin(  i )3 cos(  i ) 2 sin(  i ) 2 

The angle  is between the global axis x and the ith reinforcement direction, and Esi is the elastic
modulus of reinforcement. The reinforcing ratio pi =As/Ac.

2.1.11.4 Material Stiffness of Composite Material

The total material stiffness of the reinforced concrete is the sum of material stiffness of concrete
and smeared reinforcement:
n
D  Dc   D si (2.34)
i 1

The summation is over n smeared reinforcing components. In ATENA the smeared

reinforcement is not added on the constitutive level, but it is modeled by separate layers of
elements whose nodes are connected to those of the concrete elements. This corresponds to the
assumption of perfect bond between the smeared reinforcement and concrete.

2.1.11.5 Secant and Tangent Material Stiffness

The material stiffness matrices in the above Subsections 2.1.11.1, 2.1.11.2, 2.1.11.3, 2.1.11.4 are
either secant or tangent, depending on the type of elastic modulus used.
The secant material stiffness matrix is used to calculate the stresses for the given strains, as
shown in Section 2.1.12.
The tangent material stiffness matrix is used to construct the element stiffness matrix.

2.1.12 Analysis of Stresses

The stresses in concrete are obtained using the actual secant component material stiffness matrix
s
s c  Dc e (2.35)

32
 s
where Dc is the secant material stiffness matrix from Section 2.1.11 for the uncracked or
cracked concrete depending on the material state. The stress components are calculated in the
global as well as in the local material coordinates (the principal stresses in the uncracked
concrete and the stresses on the crack planes).
The stress in reinforcement and the associated tension stiffening stress is calculated directly from
the strain in the reinforcement direction.

2.1.13 Parameters of Constitutive Model

Default formulas of material parameters:

Parameter: Formula:
Cylinder strength f c'  0.85 f cu'

Tensile strength 2
ft  0.24 f
' ' 3
cu

Initial elastic modulus Ec  (6000  15.5 f cu' ) f cu'

Poisson's ratio   0.2

Softening compression wd  0.0005mm
Type of tension softening 1 – exponential, based on GF
Compressive strength in cracked concrete c = 0.8
Tension stiffening stress  st  0.
Shear retention factor variable (Sect.2.1.7)
Tension-compression function type linear
Fracture energy Gf according to VOS 1983 GF  0.000025 f t ' ef [MN/m]

Orientation factor for strain localization  max  1.5 (Sect.2.1.3)

The SBETA constitutive model of concrete includes 20 material parameters. These parameters
are specified for the problem under consideration by user. In case of the parameters are not
known automatic generation can be done using the default formulas given in the table above. In
such a case, only the cube strength of concrete f’cu (nominal strength) is specified and the
remaining parameters are calculated as functions of the cube strength. The formulas for these
functions are taken from the CEB-FIP Model Code 90 and other research sources.
Used units are MPa.
The parameters not listed in the table have zero default value.
The values of the material parameters can be also influenced by safety considerations. This is
particularly important in cases of a design, where a proper safety margin should be met. For that
reason, the choice of material properties depends on the purpose of analysis and the filed of an
application. The typical examples of the application are the design, the simulation of failure and
the research.

ATENA Theory 33
In case of the design application, according to most current standards, the material properties for
calculation of structural resistance (failure load) are considered by minimal values with applied
partial safety factors. The resulting maximum load can be directly compared with the design
loads.
According to some researchers, more appropriate approach would be to consider the average
material properties in nonlinear analysis and to apply a safety factor on the resulting integral
response variable (force, moment). However, this safety format is not yet fully established.
In cases of the simulation of real behavior, the parameters should be chosen as close as possible
to the properties of real materials. The best way is to determine these properties from mechanical
tests on material sample specimens.

2.2 Fracture–Plastic Constitutive Model (CC3DCementitious,

CC3DNonLinCementitious, CC3DNonLinCementitious2,
CC3DNonLinCementitious2User,
CC3DNonLinCementitious2Variable,
CC3DNonLinCementitious2FRC,
CC3DNonLinCementitious2SHCC,
CC3DNonLinCementitious3)

2.2.1 Introduction
Fracture-plastic model combines constitutive models for tensile (fracturing) and compressive
(plastic) behavior. The fracture model is based on the classical orthotropic smeared crack
formulation and crack band model. It employs Rankine failure criterion, exponential softening,
and it can be used as rotated or fixed crack model. The hardening/softening plasticity model is
based on Menétrey-Willam failure surface. The model uses return mapping algorithm for the
integration of constitutive equations. Special attention is given to the development of an
algorithm for the combination of the two models. The combined algorithm is based on a
recursive substitution, and it allows for the two models to be developed and formulated
separately. The algorithm can handle cases when failure surfaces of both models are active, but
also when physical changes such as crack closure occur. The model can be used to simulate
concrete cracking, crushing under high confinement, and crack closure due to crushing in other
material directions.
Although many papers have been published on plasticity models for concrete (for instance,
PRAMONO, WILLAM 1989, MENETREY et al 1997, FEENSTRA 1993, 1998 ETSE 1992) or
smeared crack models (RASHID 1968, CERVENKA and GERSTLE 1971, BAZANT and OH
1983, DE BORST 1986, ROTS 1989), there are not many descriptions of their successful
combination in the literature. OWEN et al. (1983) presented a combination of cracking and
visco-plasticity. Comprehensive treatise of the problem was provided also by de BORST (1986),
and recently several works have been published on the combination of damage and plasticity
(SIMO and JU 1987, MESCHKE et al. (1998). The presented model differs from the above
formulations by ability to handle also physical changes like for instance crack closure, and it is
not restricted to any shape of hardening/softening laws. Also, within the proposed approach it is
possible to formulate the two models (i.e. plastic and fracture) entirely separately, and their
combination can be provided in a different algorithm or model. From programming point of
view such approach is well suited for object-oriented programming.

34
The method of strain decomposition, as introduced by DE BORST (1986), is used to combine
fracture and plasticity models together. Both models are developed within the framework of
return mapping algorithm by WILKINS (1964). This approach guarantees the solution for all
magnitudes of strain increment. From an algorithmic point of view the problem is then
transformed into finding an optimal return point on the failure surface.
The combined algorithm must determine the separation of strains into plastic and fracturing
components, while it must preserve the stress equivalence in both models. The proposed
algorithm is based on a recursive iterative scheme. It can be shown that such a recursive
algorithm cannot reach convergence in certain cases such as, for instance, softening and dilating
materials. For this reason, the recursive algorithm is extended by a variation of the relaxation
method to stabilize convergence.

2.2.2 Material Model Formulation

The material model formulation is based on the strain decomposition into elastic  ije , plastic  ijp
and fracturing  ijf components (DE BORST 1986).

 ij   ije   ijp   ijf (2.36)

The new stress state is then computed by the formula:

 ijn   ijn 1  Eijkl ( kl   klp   klf ) (2.37)

where the increments of plastic strain  ijp and fracturing strain  ijf must be evaluated based on
the used material models.

2.2.3 Rankine-Fracturing Model for Concrete Cracking

Rankine criterion is used for concrete cracking

Fi f   ii t  f ti  0 (2.38)

It is assumed that strains and stresses are converted into the material directions, which in case of
rotated crack model correspond to the principal directions, and in case of fixed crack model, are
given by the principal directions at the onset of cracking. Therefore,  ii t identifies the trial
stress and f ti tensile strength in the material direction i . Prime symbol denotes quantities in the
material directions. The trial stress state is computed by the elastic predictor.
 ij t   ij n 1  Eijkl  kl (2.39)

If the trial stress does not satisfy (2.38), the increment of fracturing strain in direction i can be
computed using the assumption that the final stress state must satisfy (2.40).
Fi f   ii n  f ti   ii t  E iikl  kl f  f ti  0 (2.40)
This equation can be further simplified under the assumption that the increment of fracturing
strain is normal to the failure surface, and that always only one failure surface is being checked.
For failure surface k , the fracturing strain increment has the following form.
 Fk f
 ij  
f
   ik (2.41)
  ij

ATENA Theory 35
After substitution into (2.40) a formula for the increment of the fracturing multiplier  is
recovered.
 kk t  f tk  kk t  f t( wkmax )
   and wkmax  Lt (ˆ kk f   ) (2.42)
E kkkk E kkkk
This equation must be solved by iterations since for softening materials the value of current
tensile strength f t( wkmax ) is a function of the crack opening w , and is based on Hordijk’s formula
(defined in SBETA model).
The crack opening w is computed from the total value of fracturing strain ˆ kk f in direction k ,
plus the current increment of fracturing strain  , and this sum is multiplied by the
characteristic length Lt . The characteristic length as a crack band size was introduced by
BAZANT and OH. Various methods were proposed for the crack band size calculation in the
framework of finite element method. FEENSTRA (1993) suggested a method based on
integration point volume, which is not well suited for distorted elements. A consistent and rather
complex approach was proposed by OLIVIER. In the presented work the crack band size Lt is
calculated as a size of the element projected into the crack direction, Fig. 2-20. CERVENKA V.
et al. (1995) showed that this approach is satisfactory for low order linear elements, which are
used throughout this study. They also proposed a modification, which accounts for cracks that
are not aligned with element edges.

Fig. 2-20 Tensile softening and characteristic length

Equation (2.42) can be solved by recursive substitutions. It is possible to show by expanding

f t( wkmax ) into a Taylor series that this iteration scheme converges if:

f t( wkmax ) E kkkk

  (2.43)
w Lt
Equation (2.43) is violated for softening materials only when snap back is observed in the stress-
strain relationship, which can occur if large finite elements are used. In the standard
displacement based finite element method, the strain increment is given, therefore, a snap back
on the constitutive level cannot be captured. This means that the critical region, with snap back
on the softening curve, will be skipped in a real calculation, which physically means, that the
energy dissipated by the system will be over estimated. This is of course undesirable, and finite

36
 E kkkk f t(0) f (0)
elements smaller then L  should be used, where t denotes the initial slope
w w
of the crack softening curve.

It is important to distinguish between total fracturing strain ˆij f , which corresponds to the
maximal fracturing strain reached during the loading process, and current fracturing strain  ij f ,
which can be smaller due to crack closure, and is computed using (2.44) derived by ROTS and
BLAUWENDRAAD.

 kl f  ( Eijkl  Eijkl

 cr ) 1 E klmn  mn
 , and Eijlk
cr is defined by  ij  Eijkl
 cr  kl f (2.44)

cr represents the cracking stiffness in the local material directions.

The fourth order crack tensor Eijkl
In the current formulation, it is assumed, that there is no interaction between normal and shear
components. Thus, the crack tensor is given by the following formulas.
cr  0 for i  k and j  l
Eijkl (2.45)

Mode I crack stiffness equals

f t( wimax )
E cr
iiii , (no summation of indices) (2.46)
ˆii f
and mode II and III crack stiffness is assumed as:
cr  sF min  Eiiii
Eijij ´cr
, E ´jjjj
cr
 , (no summation of indices) (2.47)

where i  j , and sF is a shear factor coefficient that defines a relationship between the normal
and shear crack stiffness. The default value of sF is 20.
Shear strength of a cracked concrete is calculated using the Modified Compression Field Theory
of VECHIO and COLLINS (1986).
0.18 f c
 ij  , i j (2.48)
24 w
0.31 
ag  16

Where f c is the compressive strength in MPa, a g is the maximum aggregate size in mm and w
is the maximum crack width in mm at the given location. This model is activated by specifying
the maximum aggregate size a g otherwise the default behavior is used where the shear stress on
a crack surface cannot exceed the tensile strength.
The secant constitutive matrix in the material direction was formulated by ROTS and
BLAUWENDRAAD in the matrix format.
E s  E - E(E cr  E) -1 E (2.49)

Strain vector transformation matrix T (i.e. global to local strain transformation matrix) can be
used to transform the local secant stiffness matrix to the global coordinate system.
T
E s  T Es T (2.50)
It is necessary to handle the special cases before the onset of cracking, when the crack stiffness
approaches infinity. Large penalty numbers are used for crack stiffness in these cases.

ATENA Theory 37
2.2.3.1 Unloading Direction
Crack closure stiffness is controlled by the unloading factor (material parameter) 0 ≤ fU < 1.
The value of 0 corresponds to unloading to origin (default value for backward compatibility),
fU =1 means unloading direction parallel to the initial elastic stiffness.

2.2.4 Plasticity Model for Concrete Crushing

New stress state in the plastic model is computed using the predictor-corrector formula.
 ij  ( n1) ij  Eijkl ( kl   klp )   ijt  Eijkl  klp   ijt   ijp
(n)
(2.51)

The plastic corrector  ijp is computed directly from the yield function by return mapping
algorithm.
F p ( ijt   ijp )  F p ( ijt  lij )  0 (2.52)

The crucial aspect is the definition of the return direction lij , which can be defined as

 G p ( klt )  G p ( ijt )
lij  E ijkl then  ijp   (2.53)
  kl  ij
where G ( ij ) is the plastic potential function, whose derivative is evaluated at the predictor stress
state  ijt to determine the return direction.

The failure surface of MENETREY, WILLAM is used in the current version of the material
model.

 
2
   
3P . '   m
F   15
p
' r ( , e )  c  0 (2.54)
 fc   6 f c 3 f c' 

where
f c'2  f t '2 e 4(1  e 2 ) cos2   (2e  1) 2
m3 , r ( , e)  1
f c' ft ' e  1

2(1  e ) cos   (2e  1) 4(1  e ) cos   5e  4e
2 2 2 2
 2

In the above equations ( ,  , ) are Heigh-Vestergaard coordinates, f c and f t is compressive

strength and tensile strength respectively. Parameter e  0.5,1.0 defines the roundness of the
failure surface. The failure surface has sharp corners if e  0.5 , and is fully circular around the
hydrostatic axis if e  10
. .
The position of failure surfaces is not fixed but it can move depending on the value of strain
hardening/softening parameter. The strain hardening is based on the equivalent plastic strain,
which is calculated according to the following formula.
 eqp  min( ijp ) (2.55)

For Menétrey-Willam surface the hardening/softening is controlled by the parameter c  0,1 ,

which evolves during the yielding/crushing process by the following relationship:

38
 2
 f c( eqp ) 
c  (2.56)
 f 
 c 
In the above two formulas the expression f c( eqp ) indicates the hardening/softening law, which is
based on the uniaxial compressive test. The law is shown in Fig. 2-21, where the softening curve
is linear, and the elliptical ascending part is given by the following formula:
2
  c   eqp 
  f co   f c  f co  1 
   (2.57)
 c 

f'c

f’c0= 2f’t
eq
p

c =f'c/E
p

Fig. 2-21. Compressive hardening/softening and compressive characteristic length. Based on

experimental observations by VAN MIER.

The law on the ascending branch is based on strains, while the descending branch is based on
displacements to introduce mesh objectivity into the finite element solution, and its shape is
based on the work of VAN MIER. The onset of nonlinear behavior f c'0 is an input parameter as
well as the value of plastic strain at compressive strength  cp . The Fig. 2-21 shows typical values
of these parameters. In general case, however,  cp should be calculated from the total strain at the
f c'
peak by subtracting the elastic part  cp  1  , where 1 is the compressive strain when the
E
compressive strength f c' is reached. Especially the choice of the parameter f c'0 should be
selected with care, since it is important to ensure that the fracture and plastic surfaces intersect
each other in all material stages. On the descending curve the equivalent plastic strain is
transformed into displacements through the length scale parameter Lc . This parameter is defined
by analogy to the crack band parameter in the fracture model in Sec. 2.2.3, and it corresponds to
the projection of element size into the direction of minimal principal stresses. The square in
(2.56) is due to the quadratic nature of the Menétry-Willam surface.
Return direction is given by the following plastic potential
1
G p ( ij )   I1  2 J 2 (2.58)
3
where  determines the return direction. If   0 material is being compacted during crushing,
if   0 material volume is preserved, and if   0 material is dilating. In general, the plastic
model is non-associated, since the plastic flow is not perpendicular to the failure surface

ATENA Theory 39
The return mapping algorithm for the plastic model is based on predictor-corrector approach as
is shown in Fig. 2-22. During the corrector phase of the algorithm the failure surface moves
along the hydrostatic axis to simulate hardening and softening. The final failure surface has the
apex located at the origin of the Haigh-Vestergaard coordinate system. Secant method-based
Algorithm 1 is used to determine the stress on the surface, which satisfies the yield condition and
also the hardening/softening law.

Fig. 2-22 Plastic predictor-corrector algorithm.

Fig. 2-23. Schematic description of the iterative process (2.73). For clarity shown in two dimensions.

40
Algorithm 1: (Input is  ij , ( n 1) ijp ,  ( n ) ij )
( n 1)

Elastic predictor:  ijt  ( n1) ij  Eijkl  ( n ) kl (2.59)

Evaluate failure criterion: f Ap  F p ( ijt , ( n 1) ijp ) ,  A  0 (2.60)

If failure criterion is violated i.e. f Ap  0

 G p ( ijt )
Evaluate return direction: mij  (2.61)
 ij

Return mapping: F p ( ijt  B E mij ,  ijp )  0  B

( n 1)
(2.62)

Evaluate failure criterion: f Bp  F p ( ijt  B E mij , ( n 1)

 ijp  B mij ) (2.63)

Secant iterations (i ) as long as  A    B  e (2.64)

B   A
New plastic multiplier increment:    A  f Ap (2.65)
f Bp  f Ap

 G p ( ijt   E ( i 1)
mij )
New return direction: (i )
mij  (2.66)
 ij

Evaluate failure criterion: f p  F p ( ijt   E (i )

mij ,  ijp  
( n 1) (i )
mij ) (2.67)

New initial values for secant iterations:

f Bp  0  f Bp  f p , B   (2.68)

f Bp  0  f Ap  f Bp ,  A  B , f Bp  f p , B   (2.69)

End of secant iteration loop
End of algorithm update stress and plastic strains.
(n)
 ijp  ( n 1) ijp  B (i ) mij ,  ij   ijt  B E (i ) mij
(n)
(2.70)

2.2.5 Combination of Plasticity and Fracture model

The objective is to combine the above models into a single model such that plasticity is used for
concrete crushing and the Rankine fracture model for cracking. This problem can be generally
stated as a simultaneous solution of the two following inequalities.
F p ( ( n 1) ij  Eijkl ( kl   klf   klp ))  0 solve for  klp (2.71)

F f ( ( n 1) ij  Eijkl ( kl   klp   klf ))  0 solve for  klf (2.72)

Each inequality depends on the output from the other one, therefore the following iterative
scheme is developed.

Algorithm 2:

ATENA Theory 41
Step 1: F p ( ( n 1) ij  Eijkl ( kl   (i 1) klf  b ( i 1) klcor   ( i ) klp ))  0 solve for  (i ) klp

Step 2: F f ( ( n 1) ij  Eijkl ( kl   (i ) klp   (i ) klf ))  0 solve for  (i ) klf

Step 3:  (i ) ijcor   (i ) ijf   (i ) ijf (2.73)

Iterative correction of the strain norm between two subsequent iterations can be expressed as
 (i ) ijcor  (1  b)  f  p  ( i ) ijcor (2.74)

 (i ) ijf   (i 1) ijf  ( i ) ijp   (i ) ijp

where   f
, 
p

 (i ) ijp   (i 1) ijp  ijcor

and b is an iteration correction or relaxation factor, which is introduced to guarantee

convergence. It is to be determined based on the run-time analysis of  f and  p , such that the
convergence of the iterative scheme can be assured. The parameters  f and  p characterize the
mapping properties of each model (i.e. plastic and fracture). It is possible to consider each model
as an operator, which maps strain increment on the input into a fracture or plastic strain
increment on the output. The product of the two mappings must be contractive to obtain a
convergence. The necessary condition for the convergence is:
(1  b) f  p  1 (2.75)

If b equals 0 , an iterative algorithm based on recursive substitution is obtained. The

convergence can be guaranteed only in two cases:
One of the models is not activated (i.e. implies  f or  p  0 ),
There is no softening in either of the two models and dilating material is not used in the plastic
part, which for the plastic potential in this work means   0 , (2.58). This is a sufficient but
not necessary condition to ensure that  f and  p  1 .
It can be shown that the values of  f and  p are directly proportional to the softening rate in
each model. Since the softening model remains usually constant for a material model and finite
element, their values do not change significantly between iterations. It is possible to select the
scalar b such that the inequality (2.75) is satisfied always at the end of each iteration based on
the current values of  f and  p . There are three possible scenarios, which must be handled, for
the appropriate calculation of b :
 f p   , where  is related to the requested convergence rate. For linear rate it can be
set to   1 / 2 . In this case the convergence is satisfactory and b  0 .

   f p  1 , then the convergence would be too slow. In this case b can be estimated
 f p
as b  1  , in order to increase the convergence rate.

1   f  p , then the algorithm is diverging. In this case b should be calculated as

b  1 to stabilize the iterations.
 f p

42
This approach guarantees convergence as long as the parameters  p , f do not change
drastically between the iterations, which should be satisfied for smooth and correctly formulated
models. The rate of convergence depends on material brittleness, dilating parameter  and finite
element size. It is advantageous to further stabilize the algorithm by smoothing the parameter b
during the iterative process:
b  ( ( i )b  (i 1)b) / 2 (2.76)
where the superscript i denotes values from two subsequent iterations. This will eliminate
problems due to the oscillation of the correction parameter b . Important condition for the
convergence of the above Algorithm 2 is that the failure surfaces of the two models are
intersecting each other in all possible positions even during the hardening or softening.
Additional constraints are used in the iterative algorithm. If the stress state at the end of the first
step violates the Rankine criterion, the order of the first two steps in Algorithm 2 is reversed.
Also, concrete crushing in one direction influences the cracking in other directions. It is assumed
that after the plasticity yield criterion is violated, the tensile strength in all material directions is
set to zero.
On the structural level secant matrix is used to achieve a robust convergence during the strain
localization process.
The proposed algorithm for the combination of plastic and fracture models is graphically shown
in Fig. 2-23. When both surfaces are activated, the behavior is quite like the multi-surface
plasticity (SIMO et al. 1988). Contrary to the multi-surface plasticity algorithm the proposed
method is more general in the sense that it covers all loading regimes including physical changes
such as for instance crack closure. Currently, it is developed only for two interacting models, and
its extension to multiple models is not straightforward.
There are additional interactions between the two models that need to be considered to properly
describe the behavior of a concrete material:
(a) After concrete crushing the tensile strength should decrease as well
(b) According to the research work of Collins (VECHIO and COLLINS (1986)) and
coworkers it was established the also compressive strength should decrease when
cracking occurs in the perpendicular direction. This theory is called compression field
theory and it is used to explain the shear failure of concrete beams and walls.
The interaction (a) is resolved by adding the equivalent plastic strain to the maximal fracturing
strain in the fracture model to automatically increase the tensile damage based on the
compressive damage such that the fracturing strains satisfies the following condition:

ft p
ˆkk f   eq (2.77)
f c
The compressive strength reduction (b) is based on the following formula based proposed by
Collins:

 c  rc f c
1
rc  , rclim  rc  1.0 (2.78)
0.8  170 1

ATENA Theory 43
Where 1 is the tensile strain in the crack. In ATENA the largest maximal fracturing strain is
used for 1 and the compressive strength reduction is limited by rclim . If rclim is not specified, then
no compression reduction is considered.

2.2.6 Variants of the Fracture Plastic Model

The several ATENA material models are based on the above theories:
CC3DCementitious,
CC3DNonLinCementitious,
CC3DNonLinCementitious2,
CC3DNonLinCementitious2Variable,
CC3DNonLinCementitious2Fatigue (described in section 2.2.10),
CC3DNonLinCementitious2User,
CC3DNonLinCementitious2FRC (described in section 2.2.11),
CC3DNONLINCEMENTITIOUS2SHCC,
CC3DNONLINCEMENTITIOUS2HPFRC (described in section 2.2.12),
and CC3DNonLinCementitious3 (described in section 2.2.13),
with the following differences: CC3DCementitious assumes linear response up to the point when
the failure envelope is reached both in tension and compression. This means that there is no
hardening regime in Fig. 2-21. The material CC3DNonLinCementitious on the contrary assumes
a hardening regime before the compressive strength is reached. The material
CC3DNonLinCementitious2 is equivalent to CC3DNonLinCementitious but purely incremental
formulation is used (in CC3DNonLinCementitious a total formulation is used for the fracturing
part of the model), therefore this material can be used in creep calculations or when it is
necessary to change material properties during the analysis. The material
CC3DNonLinCementitious2Variable is based on the material CC3DNonLinCementitious2 and it
allows to define history evolution laws for selected material parameters. The following material
parameters can be defined using an arbitrary evolution laws: young modulus E , tensile strength
f t ' , compressive strength f c' and f c'0 . It is the responsibility of the user to define the parameters
in a meaningful way. It means that at any time (please note compressive strength parameters f c'
and f c'0 are defined as negative values in ATENA):

ft '  1
2 f c'0 (2.79)

f c'0  f c'0 , f c'0  0 (2.80)

The material CC3DNonLinCementitious2User allows for user defined laws for selected material
laws such as: diagrams for tensile and softening behavior (see Fig. 2-24 and Fig. 2-25), shear
retention factor (Fig. 2-26) and the effect of lateral compression on tensile strength (Fig. 2-27).

44
 t/f’t

1.0
t t
(f1 ) Lt/Lch
loc

t loc f1
~
Fig. 2-24. An example of a user defined tensile behavior for CC3DNonLinCementitious2User
material.
c/f’c

1.0
p c
(eqc ) Lc/Lch
loc

cloc
~ peq
Fig. 2-25. An example of a user defined compressive behavior for CC3DNonLinCementitious2User
material.
G/Gc

1.0
f sh t
(1 ) Lt/Lch
loc

sh
loc
~f1
Fig. 2-26. An example of a user defined shear retention factor for shear stiffness degradation after
cracking.

ATENA Theory 45
In the user defined material mode II and III crack stiffness are evaluated with the help of the
shear retention factor rg as:

rg G
cr 
Eijij (2.81)
1  rg

where i  j , rg  min(rgi , rgj ) is the minimum of shear retention factors on cracks in directions
i , j , and G is the elastic shear modulus. Shear retention factor on a crack in direction i is
evaluated from the user specified diagram as shown in Fig. 2-26.
In the above diagrams Lt and Lc represents the crack band size and crush band size respectively
as it is defined Section 2.1.3. Ltch and Lcch represents a size for which the tensile and compression
diagram respectively is valid. For instance, it represents the measuring base that was used in an
experiment to determine the strain values in the diagrams above.  loc represents the strain value,
after which strain localization can be expected. Usually, this is the strain after which the diagram
is entering into the softening regime. For instance, the strain value that is used to determine the
tensile strength is calculated based on the following assumptions:
if 1f   loc
f

1f  1f
else

1f   locf  (1f   locf ) Lt (2.82)

Ltch
The calculation of the strain value for graphs in Fig. 2-25 and Fig. 2-26 is analogical to Eq.
(2.82) but the appropriate values of  loc , L and Lch should be used. It should be noted that the
strain 1f is the strain that is calculated from the strain tensor at the finite element integration
points, while the strain 1f is used to determine the current tensile strength from the provided
stress-strain diagram (see Fig. 2-24). The equation (2.82) then represents a scaling that considers
the difference between the experimental size and the size of the integration point. This approach
guarantees that the same amount of energy is dissipated when using large and small finite
elements.
It is also possible to define a material law for the shear strength of a cracked concrete and for the
compressive strength reduction after cracking.

Compressive strength of cracked concrete  c  rc (1f ) f c (2.83)

Shear strength of cracked concrete  ij  f sh (1f ) ft (2.84)

It should be realized that the compressive strength of the cracked concrete i.e. (2.83) is a
function of the maximal fracturing strain, i.e. maximal tensile damage at the given point. The
shear strength should be a function of the crack opening. Because of that the shear strength is
specified as a function of the fracturing strain 1f after the localization transformation (2.82).
The shear strength law is specified as a value relative to ft . The compressive strength reduction
is specified as a function relative to f c .

46
 t/f’t

1.0

1.0 3/f’c
Fig. 2-27. An example of a user defined tensile strength degradation law due to lateral compressive
stress.

2.2.7 Tension Stiffening

In heavily reinforced concrete structures, the cracks cannot fully developed and concrete
contributes to the steel stiffness. This effect is called tension stiffening and in
CC3DNonLinCementitious2 material it can be simulated by specifying a tension stiffening
factor cts . This factor represents the relative limiting value of tensile strength in the tension
softening diagram. The tensile stress cannot drop below the value given by the product of cts f t
(see Fig. 2-28). The recommended default value for cts is 0.4 as recommended by CEB-FIP
Model Code 1990.


ft

c ts ft


Fig. 2-28: Tension stiffening.

2.2.8 Crack Spacing

In heavily reinforced concrete structures, or structures with large finite elements, when many
reinforcement bars are crossing each finite element, the crack band approach described in
Section 2.1.3 will provide too conservative results, and the calculated crack widths may be

ATENA Theory 47
overestimated. This is the consequence of the fact that the crack band approach assumes that the
crack spacing is larger than a finite element size. In heavily reinforced structures, or if large
finite elements are used, it may occur that the crack spacing will be smaller than finite element
size. This is especially true if shell/plate elements are used. In this case, typically large finite
elements can be used, and they usually contain significant reinforcement. In these cases, it is
useful to provide the crack spacing manually, since otherwise the program will overestimate the
cracking and due to that also larger deflections may be calculated. The program ATENA allows
the user to manually define the crack spacing. This user defined spacing is used as crack band
size Lt in cases when the user defined crack spacing is smaller than the Lt that would be
calculated by formulas presented in Section 2.1.3.

2.2.9 Fixed or Rotated Cracks

Similarly, to the SBETA material, the Cementitious material family offers the choice of fixed
and rotated crack models (see section 2.1.6). The fixed crack material parameter determines at
which maximum residual tensile stress level the crack direction gets fixed. In other words, 0.0
means fully rotated crack model (as 0 in SBETA), 1.0 means fixed crack model (as 1 in
SBETA), values between 0.0 and 1.0 determine the crack direction locking level, e.g., 0.7 fixes
the crack direction as soon it opens so far that the softening law drops to 0.7 times the [initial]
tensile strength.

2.2.10 Fatigue
For modelling fatigue behavior of concrete (CEB 1988 and SAE AE-4) under tensile load, a new
material has been implemented in ATENA. The new material
(CC3DNonLinCementitious2Fatigue) is based on the existing three-dimensional fracture plastic
material (CC3DNonLinCementitious2) and uses a stress-based model (2.2.10.1). It has an
additional parameter,  fatigue , and additional data attributes for  base , N , and  fatigue , used in the
damage calculation as described in section 2.2.10.2. For details and validation against tests
conducted by KESSLER-KRAMER (2002) see ČERVENKA, PRYL (2007) or PRYL,
CERVENKA, PUKL (2010). Modelling 3-point bending tests with this material is presented in
PRYL, PUKL, CERVENKA (2013) and PRYL, D., MIKOLÁŠKOVÁ, J., PUKL, R. (2014).

2.2.10.1 Stress Based Models

In this approach the fatigue is represented by the so-called S-N curves relating the applied stress,
S, and the number of cycles, N, to failure. Such curves must be determined by tests, see Fig.
2-29.
For steel reinforcement bars the performance can be normally expressed as a simple power law
by BASQUIN (1910).
 rm N  C (2.85)
where  r is the stress range, N is the number of cycles to failure and m and C are constants.
This means a linear relationship between  and N in a full logarithmic diagram. The equation
(2.85) is generally valid for the high-cycle range.
For plain concrete, the performance can normally be expressed as a straight line in a semi-
logarithmic diagram of the form:
 max
 1   1  R  log N (2.86)
f

48
  min
where  max is the maximum stress, f is static concrete strength, R  ,  min is minimum
 max
stress and  is a material constant. The equation (2.86) holds for both compressive and tensile
stresses, however, the value of  is not necessarily the same for tensile and compressive
behavior of a material. The value should be determined from experiments. For example,
 =0.052 was used based on the experimental results for load levels 0.7 and 0.9 Fstat when
modelling the test on a probe sealed during curing with a notch from section 3.5.2.4 of
KESSLER-KRAMER (2002) for validation.

Fig. 2-29: Typical S-N line for concrete in compression (KLAUSEN (1978))
The S-N relations mentioned above are mainly obtained by constant amplitude tests. However,
in real structures the stresses are varying. One method which can be of help in this context is the
well-known Palmgren-Miner hypothesis PALMGREN (1924), MINER (1945).
k
ni
N
i 1
1 (2.87)
i

where ni is the number of constant amplitude cycles at stress level i , N i is the number of cycles
to failure at stress level i , and k is the number of stress levels. As a rough tool this hypothesis is
useful, especially concerning steel. It can also be used for concrete although some investigations
have suggested that a value lower than 1 should be used.

ATENA Theory 49
2.2.10.2 Fatigue Damage Calculation
In the implemented model, fatigue damage consists of a contribution based on cyclic stress
(2.2.10.2.1), and an additional contribution from crack opening and closing in each cycle
(2.2.10.2.3). The former is dominant before cracking occurs, the latter in already cracked
regions.

2.2.10.2.1 Stress Based Contribution

The number of cycles to failure N is determined from a simple stress based model, so called S-
N or Wöhler curve as described in the previous section 2.2.10.1.
  upper 
 1 
 f 
  fatigue 1 R  
 upper  
 1   fatigue 1  R  log N , i.e., N  10  
, where  upper stands for the maximum
f
 base
tensile or compressive stress and f for the corresponding strength, ft or f c , R  .
 upper

Then, the damage due to fatigue after n cycles is calculated as an increase of the maximum
fracturing strain ˆij f (see section 2.2.3). The maximum fracturing strain in each principal
direction is adjusted by adding
w fatigue n
 fatigue  , where w fatigue  w fail and the failing displacement for the given stress
ElemSize N
w fail  invert _ soft _ law( upper ) (see Fig. 2-30).

Fig. 2-30: Softening law and fatigue damage.

50
In ATENA 4.0, a single value of  fatigue is used to calculate fatigue damage caused by both
tensile and compressive stresses. So far, there is also no special provision implemented for loads
crossing zero, i.e., changing from tension to compression and back in each cycle, which lead to
faster damage according to experimental results presented in CEB 1988 and SAE AE-4. In that
situation, the damage is calculated separately for cyclic loading from 0 to max. compression and
from 0 to max. tension, and then the worse of both damage values is considered. It should be
also noted that the damage is only introduced in form of maximum fracturing strain, which has
no direct impact on compressive material properties, i.e., the fatigue damage effectively only has
influence on tensile behaviour of the material.

2.2.10.2.2 Stress Based Contribution with Trilinear Damage 1

Hardcoded definition of damage evolution during the fatigue process, with the breakpoints
wf1 = wfr_1 * wfail and wf2 = wfr_2 * wfail.
w fatigue = n * wf1 / N1 for n_tot < N1

wf1 + ((n_tot - N1) * (wf2 - wf1) / (N2 - N1)) - wf_curr for N1 <= n_tot < N2
wf2 + ((n_tot - N2) * (wfail - wf2) / (N - N2)) - wf_curr for N2 <= n_tot < N
wfail * n_tot / N - wf_curr for N <= n_tot
where
n_tot = n + N_beg, N_curr = N - N_beg,
N_beg = wf_curr * N1 / wf1 for wf_curr < wf1
N1 + (wf_curr - wf1) * (N2 - N1)/(wf2 - wf1) for wf1 <= wf_curr < wf2
N2 + (wf_curr - wf2) * (N - N2)/( wfail - wf2) for wf2 <= wf_curr
0 < N_beg < N
and
wfr_1 = 0.1, Nr_1 = 0.1, wfr_2 = 0.5, Nr_2 = 0.9, N1 = Nr_1 * N, N2 = Nr_2 * N.

2.2.10.2.3 Crack Opening Based Contribution

The damage due to cracks that open and close during the cyclic loading is determined as
w
 fatigueCOD  fatigueCOD , where w fatigueCOD  n  fatigue / RCOD c fatigueCODload COD 2, RCOD is the crack
ElemSize
opening ratio (similar to the cycle asymmetry ratio R used in the stress based contribution; with
a bottom limit of 0.01), and COD denotes the difference between the maximum and minimum
crack opening during a cycle. The resulting  fatigueCOD is added to  fatigue before the fatigue
damage is introduced into the material.

1
Available since version 5.3.0
2
In ATENA versions prior to 5.1.3 and 5.3.4: w fatigueCOD  n  fatigue c fatigueCODload COD

ATENA Theory 51
2.2.10.3 Bringing in Fatigue Damage
It is recommended to introduce the fatigue induced damage into the unloaded structure (i.e., at
the lower stress level). Several other approaches of introducing the damage into the model were
also tested, i.e., introducing the damage at the upper load level or during reloading, but they
usually bring more convergence problems, especially during unloading.

2.2.11 Fiber Reinforced Concrete (FRC) Material

The CC3DNonLinCementitious2FRC material model is based on CC3DNonLinCementitious2
as described above in Sections 2.2.1 - 2.2.6. In case of FRC, the fibers added to the concrete
mixture increase the residual strength and ductility of the material, which is reflected by the
tension softening law. In the FRC material model, the added fractural energy approach proposed
by Juhász (2013) is implemented in the stress-crack width diagram. The total fractural energy of
the fiber reinforced concrete reads:

GFFRC  GF  GFf ,
where GFFRC and GF, are the fractural energies of the fiber reinforced concrete and the plain
concrete matrix, respectively, and GFf is the additional fractural energy, which corresponds to the
pull-out energy of the fibers.

Fig. 2-31: Crack opening law for FRC using the added fractural energy approach.
The fracture energy added by the fibers is assumed as:
GFf  w f  f t ,
where wf is the maximum crack opening width of the FRC, which depends on the type and
length of the fibers, and ft is the post-cracking residual tension strength. It should be noted that
the value of ft defined as fFtu in the fib model code 2010 (Taerwe and Matthys, 2013).

2.2.12 Strain Hardening Cementitious Composite (SHCC, HPFRCC) Material

The CC3DNONLINCEMENTITIOUS2SHCC is suitable for fibre reinforced concrete, such as
SHCC (Strain Hardening Cementitious Composites) and HPFRCC or UHPFRC (high and ultra-
high performance fiber reinforced concrete) materials. The theory of this material model is
identical to those described in Sections 2.2.1 - 2.2.6. The tensile softening regime (Fig. 2-33) and
the shear retention factor (Eq. (2.94)) are modified based on the model, proposed in KABELE, P.
(2002). This model is based on a notion of a representative volume element (RVE), which
contains distributed multiple cracks (hardening) as well as localized cracks (softening) – see Fig.
2-32.

52
 a) Multiple cracking regime b) Localized cracking regime
 
seco
nda

 
ry c
rack

t
y crack se
set

r
prima

Fig. 2-32: Representative volume element with cracks.

2.2.12.1 Basic Assumptions

a) multiple cracking regime (hardening)
 A set of parallel planar multiple cracks forms when maximum principal stress max = fc (first
crack strength).
 Crack planes are perpendicular to the direction of max (-axis).
 The direction of a crack set is fixed.
 Secondary crack set may form in direction perpendicular to primary set if the maximum
normal stress in the corresponding direction (-axis) exceeds fc.
 Cracks may slide if the direction of principal stress changes.
 Crack opening and sliding are resisted by fiber bridging.
 Crack opening and sliding displacements are averaged over the RVE as cracking strains
mc , mc ,
ij , ij (notation: lower indices – components of tensor or vector, upper indices –
multiple or localized crack mc, lc and association with primary or secondary crack direction ,
)
b) localized cracking regime (softening)
 A localized crack forms within a set of multiple cracks if the corresponding normal cracking
strain exceeds the level of mcmb (cracking strain capacity, a material constant).

 Opening and sliding displacements of the i , i localized cracks are treated by the crack
lc, lc,
band model (i.e. they are transformed into cracking strains ij , ij by dividing them with
corresponding band width wc or wc).
The overall strain of the RVE is then obtained as a sum of strain of material between cracks
(which may possibly contain nonlinear plastic strain due to compressive yielding), cracking
strains due to multiple cracks, and cracking strains due to localized cracks:
s mc , mc , lc, lc ,
ij ij ij ij ij ij (2.88)
where  ijs represents the strain of the continuous material between cracks.

ATENA Theory 53
2.2.12.2 Crack Opening Model
The crack-normal stress components are related to cracking strains corresponding to opening of
multiple and localized cracks by piecewise linear relations depicted in Fig. 2-33 [although linear
hardening and softening are shown, a user should be allowed to input piecewise linear curves].
Note that for multiple cracks, it is assumed that they do not close unless exposed to crack-normal
compression (plasticity-like unloading) while a localized crack is assumed to close so that
normal stress decreases linearly to reach zero at zero COD [these assumptions may need to be
revised in the future to some combination of plasticity and damage-like closure]. See also section
2.2.3.

multiple cracking regime localized cracking regime

 
crack
loading
opening

unloading/ unloading/
reloading reloading
mcmb 

cracking strain mc cracking strain mc COD 

Fig. 2-33: Stress vs. cracking strain relations in crack-normal direction.

2.2.12.3 Crack Sliding Model

The model for crack sliding phenomena is implemented by means of a variable shear retention
factor  The shear retention factor is defined as a ratio of the material post-cracking shear
stiffness Gc to its elastic shear stiffness G,

Gc
 . (2.89)
G
Let us determine stiffness Gc, while considering the most general 2-D case of an element, which
contains two perpendicular sets of multiple cracks and two perpendicular localized cracks. If the
problem is defined in plane , then the total engineering shear strain has only one non-zero
component, which is obtained as:
s mc , mc , lc , lc ,
2 2 2 2 2 , (2.90)
which can be rewritten with use of the shear bridging model (Kabele, 2000) as:
1 1 1 1 1 1
G M mc ,
M mc ,
wc L wc L Gc (2.91)

Functions M and L are defined by

Vf k Gf
M ( )  (2.92)
2

54
 1   Vf k Gf
L()     , for    0
 2 2 0   4 k G f   
2

 1    
 3 E f  d f  
 
L (  )  0 , for    0 (2.93)
Here Vf is the fiber volume fraction, Gf is the fiber shear modulus, Ef is the fiber Young’s
modulus, df is the fiber diameter, and k is the fiber cross-section shape correction factor. The
quantity  and  indicates the crack opening in direction  and  respectively. The parameter
 0 represents the limiting value of the crack opening displacement, when no tensile stress can be
transferred across the crack, i.e. the point when the stress-displacement diagram in Fig. 2-33
drops to zero. These parameters are to be supplied by the user except for the parameter  0 ,
which is automatically extracted from the provided stress-strain law for tension. The shear
retention factor is then expressed as
1
 (2.94)
 1 1 1 1 
1 G        
 M (  ) M ( ) wc L(  ) wc L(  ) 
mc , mc , 

Note that for an element containing only multiple cracks (before localization) 0 and
mc , mc ,
1/L terms approach zero. For an uncracked element, 0 and 1/M and 1/L
approach zero, giving =1.

2.2.13 Confinement-Sensitive Constitutive Model

The CC3DNonLinCementitious3 fracture-plastic constitutive model is an advanced version of
the CC3DNonLinCementitious2 material that can handle the increased deformation capacity of
concrete under triaxial compression. It is suitable for problems including confinement effects
such as confined reinforced concrete members (columns, bridge piers), nuclear vessels and
triaxial compression tests of plain concrete. A detailed description of the model formulation is
presented in PAPANIKOLAOU and KAPPOS (2007). In this section, only the main differences
between the CC3DNonLinCementitious3 and the CC3DNonLinCementitious2 model are
described, which are mainly focused on the plasticity part of the model (section 2.2.4).

2.2.13.1 Hardening and Softening Function

The position of failure surface can expand and move along the hydrostatic axis (simulating the
hardening and softening stages), based on the value of the hardening/softening parameter (κ). In
the present model, this parameter identifies with the volumetric plastic strain (GRASSL et al.,
2002) :
dκ  dε pv  dε1p  dε 2p  dε 3p (2.95)
The instantaneous shape and location of the loading surface during hardening is defined by a
hardening function (k), which depends on the hardening/softening parameter (κ). This function is
directly incorporated in the Menétrey-Willam failure surface equations (2.54), operating as a

ATENA Theory 55
scaling factor on the compressive concrete strength (fc). It has the same elliptic form with
CC3DNonLinCementitious2 (2.57), but herein in terms of the plastic volumetric strain:
2
 ε pv,t  ε pv 
k(κ)  k(ε )  k o  1  k o   1  
p
 (2.96)
v  εp
 v,t 
where ε pv,t is the plastic volumetric strain at uniaxial concrete strength (onset of softening) and ko
is the value that defines the initial yield surface that bounds the initial elastic regime (onset of
plasticity). At the end of the hardening process, the hardening function retains a constant value
of unity and the material enters the softening regime, which is controlled by the softening
function (c). This function simulates the material decohesion by shifting the loading surface
along the negative hydrostatic axis. It is assumed that it follows the softening function originally
proposed by VAN GYSEL and TAERWE (1996) for uniaxial compression:
2
 
 1 
 2 
c(κ)  c(ε v )    n1  1  
p
(2.97)
1
  n 2  1  
 
 
where:
n1  ε pv / ε pv,t (2.98)

n 2  (ε pv,t  t) / ε pv,t (2.99)

Parameter t in equation (2.99) controls the slope of the softening function and the outmost square
is necessary due to the quadratic nature of the loading surface. The softening function value
starts from unity and complete material decohesion is attained at c = 0. The evolution of both
hardening and softening functions with respect to the hardening/softening parameter is
schematically shown in Fig. 2-34.

56
 k(κ) / c(κ)
c k
1.0

k
0.8

0.6
c

0.4

0.2
ko

0.0
p
ε v,t κ = εpv

Fig. 2-34: Evolution of hardening (k) and softening (c) functions with respect to the plastic
volumetric strain.

2.2.13.2 Plastic Potential Function

The present plasticity model incorporates a non-associated flow rule using a polynomial plastic
potential function (g), with Lode angle (θ) dependency and adjustable order (n):
n
 ρ   1  ρ ξ
g  A    C  (B  C)(1  cos3θ)    a (2.100)
 k  c  fc   2  k  c  fc k  c  fc
Parameters A, B and C define the shape of the plastic potential function in stress space and their
calibration is based on the assumption that the inclination (ψ) of the incremental plastic strain
vector identifies with the inclination of the total plastic strain vector at three distinct stress states,
namely the uniaxial, biaxial and triaxial compressive concrete strength (Fig. 2-35). The attraction
constant (a) is included for mathematical clarity and is not a user parameter, due to plastic
potential function differentiation in the flow rule.

ATENA Theory 57
 Fig. 2-35: Direction (ψ) of the incremental (a) and total (b) plastic strain vectors.

2.2.13.3 Suggested Model Parameters

A detailed calibration scheme for the plasticity model parameters, based on and extensive
experimental database can be found in PAPANIKOLAOU and KAPPOS (2007) and suggested
values (including the fracture model parameters) for various uniaxial compressive concrete
strengths (fc) are shown in the following table (see Atena Input File Format document for the
material definition details):
Table 2.2-1 Suggested parameters for the fracture and plasticity models

fc (ΜPa) 20 30 40 50 60 70
Εc (MPa) 24377 27530 30011 32089 33893 35497
ν 0.2 0.2 0.2 0.2 0.2 0.2
ft (MPa) 1.917 2.446 2.906 3.323 3.707 4.066
λt 1.043 1.227 1.376 1.505 1.619 1.722
e 0.5281 0.5232 0.5198 0.5172 0.5151 0.5133
fco (MPa) -4.32 -9.16 -15.62 -23.63 -33.14 -44.11

ε pv,t 4.92·10-4 6.54·10-4 8.00·10-4 9.35·10-4 1.06·10-3 1.18·10-3

t 1.33·10-3 2.00·10-3 2.67·10-3 3.33·10-3 4.00·10-3 4.67·10-3

A 7.342177 5.436344 4.371435 3.971437 3.674375 3.43856
B -8.032485 -6.563421 -5.73549 -5.430334 -5.202794 -5.021407
C -3.726514 -3.25626 -3.055953 -2.903173 -2.797059 -2.719067
n 3 3 3 3 3 3
Gf (MN/m) 4.87·10-5 6.47·10-5 7.92·10-5 9.26·10-5 1.05·10-4 1.17·10-4

58
 fc (ΜPa) 80 90 100 110 120
Εc (MPa) 36948 38277 39506 40652 41727
ν 0.2 0.2 0.2 0.2 0.2
ft (MPa) 4.405 4.728 5.036 5.333 5.618
λt 1.816 1.904 1.986 2.063 2.136
e 0.5117 0.5104 0.5092 0.5081 0.5071
fco (MPa) -56.50 -70.30 -85.48 -102.01 -114.00

ε pv,t 1.30·10-3 1.41·10-3 1.52·10-3 1.62·10-3 1.73·10-3

t 5.33·10-3 6.00·10-3 6.67·10-3 7.33·10-3 8.00·10-3

A 3.245006 3.082129 2.942391 2.820644 2.713227
B -4.871993 -4.745867 -4.637358 -4.542587 -4.458782
C -2.659098 -2.611426 -2.572571 -2.540158 -2.512681
n 3 3 3 3 3
Gf (MN/m) 1.29·10-4 1.40·10-4 1.50·10-4 1.61·10-4 1.71·10-4

2.3 Von Mises Plasticity Model

Von Mises plasticity model called also as J2 plasticity is based only on one parameter k. The
yield function is defined as:
F p ( ij )  J 2  k   eqp   0 (2.101)

where J2 denotes the second invariant of stress deviator tensor. The parameter
k  p
eq  1
3  y   eqp  is the maximal shear stress and  y is the uniaxial yield stress. This
parameter controls the isotropic hardening of the yield criterion.
N inc
 y   eqp    y  H  eqp ,  eqp   2
3  ε p
: ε p  (2.102)
i 1

 y is the yield stress, H the hardening modulus and  eqp is the equivalent plastic strain
calculated as a summation of equivalent plastic strains during the loading history.
In case of von Mises plasticity the plastic potential function is identical with the yield function:
G p ( ij )  F P ( ij ) (2.103)
The associated flow rule is assumed. The background information can be found in (CHEN,
SALEEB 1982, Sec.5.4.2).
The Von Mises model could be used to model cyclic steel behavior including Bauschinger
effect. In this case the yield function is modified as:

ATENA Theory 59
 1
2  σ  X  :  σ  X   k  eqp   (r  1)k0  0 (2.104)

where σ is the deviatoric stress, k0 is an initial value of k ( eqp ) according to (2.102), X is the so
called back stress controlling the kinematic hardening:
X  2 3 k1 ε p  k2 X  eqp (2.105)

In equations (2.104) and (2.105) quantities r , k1 , k2 are material parameters for the cyclic
response. If r is non-zero, the cyclic model is activated, and it controls the radius of the Von
Mises surface. If r  1 the yielding will start exactly when  y is reached. For lower values, the
non-linear behavior starts earlier, and the slope of the response is mainly affected by parameter
k1 (larger value – higher slope). Parameter k2 on the other hand affects the memory of the cyclic
response. Some examples of various parameter combinations are shown at Fig. 2-36.

60
Fig. 2-36: Effect of material parameter choice on cyclic response for E=210 GPa and  y = 200 MPa.

ATENA Theory 61
2.4 Drucker-Prager Plasticity Model
Drucker-Prager plasticity model is based on a general plasticity formulation that is described in
Section 2.2.4. The yield function is defined as:
p
FDP ( ij )   I 1  J 2  k  0 (2.106)

Where  and k are parameters defining the shape of the failure surface. They can be estimated
by matching with the Mohr-Coulomb surface. If the two surfaces are to agree along the
compressive meridian, i.e.   00 , the formulas are:
2sin  6 c cos 
 , k (2.107)
3  3  sin   3  3  sin  
This corresponds to a outer cone to the Mohr-Coulomb surface. The inner cone, which passes
through the tensile meridian where   600 has the constants given by the following expressions:
2sin  6 c cos 
 , k (2.108)
3  3  sin   3  3  sin  
The position of failure surfaces is not fixed but it can move depending on the value of strain
hardening/softening parameter. The strain hardening is based on the equivalent plastic strain,
which is calculated according to the following formula.
 eqp  min( ijp ) (2.109)

Hardening/softening in the Drucker-Prager model is controlled by the parameter k . This

parameter is selected such that the surface at the peak passes through the uniaxial compressive
strength, and it changes according to the following expression.
f c( eqp )
k' k (2.110)
f c
The symbol k’ in the above formula replaces k in (2.106). In the above two formulas the
expression f c( eqp ) indicates the hardening/softening law, which is based on the uniaxial
compressive test. The law is shown in Fig. 2-37.

Fig. 2-37. Linear softening in the Drucker-Prager material model

Return direction is given by the following plastic potential:

1
G p ( ij )   I1  2 J 2 (2.111)
3

62
where  determines the return direction. If   0 material is being compacted during crushing,
if   0 material volume is preserved, and if   0 material is dilating. In general, the plastic
model is non-associated, since the plastic flow is not perpendicular to the failure surface
The return mapping algorithm for the plastic model is based on predictor-corrector approach as
is shown in Fig. 2-22. During the corrector phase of the algorithm the failure surface moves
along the hydrostatic axis to simulate hardening and softening. The final failure surface has the
apex located at the origin of the Haigh-Vestergaard coordinate system. Secant method-based
Algorithm 1 is used to determine the stress on the surface, which satisfies the yield condition and
also the hardening/softening law.

2.5 User Material Model

In some situations, none of the standard material models available in ATENA can describe the
behavior sufficiently. Many such cases can be handled by defining user laws in the fracture-
plastic material model (see CC3DNonLinCementitious2User described in section 2.2.6), in the
others the user can provide a dynamic link library implementing his own material model. The
user material is based on the elastic isotropic material, adding new material parameters and state
variables (both limited to floating point values). See the User Material DLL Manual for
description and reference, and the CCUserMaterialExampleDLL directory in Atena Science
Examples for an example project including the source code in C and a window help file version
of the manual, AtenaV4_UserMaterialDLL.chm. Please note that the behavior of the user model
may have influence on convergence of the analysis.

2.6 Interface Material Model

The interface material model can be used to simulate contact between two materials such as for
instance a construction joint between two concrete segments or a contact between foundation
and concrete structure. The interface material is based on Mohr-Coulomb criterion with tension
cut off. The constitutive relation for a general three-dimensional case is given in terms of
tractions on interface planes and relative sliding and opening displacements.
 1   Ktt 0 0   v1 
    
 2    0 Ktt 0  v2  (2.112)
  K nn   u 
   0 0
For two-dimensional problems second row and column are omitted.
The initial failure surface corresponds to Mohr-Coulomb condition (2.113) with ellipsoid in
tension regime. After stresses violate this condition, this surface collapses to a residual surface
which corresponds to dry friction.
  c    ,   0 (2.113)

   c  ,  
2
c ft 2 
  0 1 , c   , 0    ft
 ft   c  c  2 ft 
2 0
 c2
1
 ft   c 
2

  0 ,   ft

ATENA Theory 63
In tension the failure criterion is replaced by an ellipsoid, which intersect the normal stress axis
at the value of ft with the vertical tangent and the shear axis is intersected at the value of c (i.e.
cohesion) with the tangent equivalent to  .
The parameters for the interface model cannot be defined arbitrarily; there is certain dependence
of some parameters on the others. When defining the interface parameters, the following rules
should be observed:
c
ft  , ft  c
 (2.114)
c  0, f t  0,   0
It is recommended that parameters c, f t ,  are always greater than zero. In cases when no
cohesion or no tensile strength is required, some very small values should be prescribed.

Trial stress


1 c Initial surface
Final stress
Residual surface

ft 
Fig. 2-38: Failure surface for interface elements.
In general three-dimensional case  in Fig. 2-38 and equation (2.113) is calculated as:

   12   22 (2.115)

64
 
c 
min
Ktt
1

Ktt

1
(a) v


ft

min
Knn Knn
1
1

(b) u
Fig. 2-39: Typical interface model behavior in shear (a) and tension (b)
The K nn , K tt denote the initial elastic normal and shear stiffness, respectively. Typically for
zero thickness interfaces, the value of these stiffnesses correspond to a high penalty number. It is
recommended not to use extremely high values as this may result in numerical instabilities. It is
recommended to estimate the stiffness value using the following formulas
E G
K nn  , K tt  (2.116)
t t
where E and G is minimal elastic modulus and shear modulus respectively of the surrounding
material. t is the width of the interface zone. Its value can be selected either based on the reality.
For instance, for mortar between masonry bricks the value is typically 10-20 mm. Alternatively,
it can be estimated as a dimension, which can be considered negligible with respect to the
structural size. For instance, in case of a dam analysis, where the dam dimensions are typically in
the order of 100 meters, the width of the interface zone can be estimated to be 0.5 meters. It is

ATENA Theory 65
suitable due to numerical reasons if stiffness is about 10 times of the stiffness of adjacent finite
elements.
There are two additional stiffness values that need to be specified in the ATENA input. They are
denoted in Fig. 2-39 as K nnmin and K ttmin . They are used only for numerical purposes after the
failure of the element to preserve the positive definiteness of the global system of equations.
Theoretically, after the interface failure the interface stiffness should be zero, which would mean
that the global stiffness will become indefinite. These minimal stiffnesses should be about 0.001
times of the initial ones.
It is possible to define evolution laws for tensile as well as shear softening by arbitrary
multilinear laws. Examples of such laws are shown in Fig. 2-40. The figure describes bi-linear
softening laws. The break point of this law can be determined for instance by the formula
proposed by Bruehwiler and Wittman (1990).
ft GF
s1  , v1  0.75 (2.117)
4 ft

ft cc 0

I II
GF GF
s1 s1c

u1 u v 1c v
u eqf
u eqf
Fig. 2-40: Example of a softening law for tension and cohesion.
The evolution law depends on the equivalent nonlinear interface relative displacement

ueqf  u 2f  v 2f 1  v 2f 2 in 3D and ueqf  u 2f  v 2f in 2D (2.118)

Where u f and v fi are the inelastic components of the relative interface displacement on the
basis of their decomposition into elastic and nonlinear, i.e. fracturing part.
u  ue  u f
(2.119)
vi  vi  v fi
This approach ensures that the degradation in shear affects also tensile strength and vice versa.
For instance, when the interface is damaged in shear, the tensile strength is reduced as well. The
typical behavior of the interface model with the softening evolution laws is shown in Fig. 2-39
by the dotted lines. The default behavior when no softening law is given is brittle with
immediate drop to zero in tension and to the residual dry friction in shear. The behavior is shown
in Fig. 2-39 by the solid black line.
When user softening laws are defined for the interface material, it is recommended that the
softening law for cohesion is always more ductile then the one for tensile strength, i.e. the
cohesion should be higher than the tensile strength at any time during the softening process.

66
Fig. 2-41: Example of a cyclic response of the model in shear under constant normal pre-stress.

2.7 Reinforcement Stress-Strain Laws

2.7.1 Introduction
Reinforcement can be modeled in two distinct forms: discrete and smeared. Discrete
reinforcement is in form of reinforcing bars and is modeled by truss elements. The smeared
reinforcement is a component of composite material and can be considered either as a single
(only one-constituent) material in the element under consideration or as one of the more such
constituents. The former case can be a special mesh element (layer), while the later can be an
element with concrete containing one or more reinforcements. In both cases the state of uniaxial
stress is assumed, and the same formulation of stress-strain law is used in all types of
reinforcement. More info about discrete reinforcement is available in Section 10.2.3 Discrete
Reinforcement Embedded in Solid Elements, located near the end of this manual.

2.7.2 Bilinear Law

The bilinear law, elastic-perfectly plastic, is assumed as shown in Fig. 2-42.

ATENA Theory 67
 Fig. 2-42 The bilinear stress-strain law for reinforcement.
The initial elastic part has the elastic modulus of steel Es. The second line represents the
plasticity of the steel with hardening and its slope is the hardening modulus Esh. In case of
perfect plasticity Esh =0. Limit strain L represents limited ductility of steel.

2.7.3 Multi-line Law

The multi-linear law consists of four lines as shown in Fig. 2-43. This law allows to model all
four stages of steel behavior: elastic state, yield plateau, hardening and fracture. The multi-line is
defined by four points, which can be specified by input.

Fig. 2-43 The multi-linear stress-strain law for reinforcement.

The above-described stress-strain laws can be used for the discrete as well as the smeared
reinforcement. The smeared reinforcement requires two additional parameters: the reinforcing
ratio p (see Section 2.1.1.1) and the direction angle  as shown in Fig. 2-44.

68
 Fig. 2-44 Smeared reinforcement.
The spacing s of the smeared reinforcement is assumed infinitely small. The stress in the
smeared reinforcement is evaluated in the cracks, therefore it should also include a part of stress
due to tension stiffening (which is acting in concrete between the cracks, section 2.1.9).
 scr '   s '   ts (2.120)
where  s is the steel stress between the cracks (the steel stress in smeared reinforcement),  scr
' '

is the steel stress in a crack. If no tension stiffening is specified  ts =0 and  scr   s . In case of
' '

the discrete reinforcement the steel stress is always  s .

'

2.7.4 No Compression Reinforcement

Normally all reinforcement material models in ATENA exhibit the same behavior in tension as
well as in compression. The material types CCReinforcement and
CCSmearedReinforcement include the capability to deactivate the compressive response of
the reinforcement. This is sometimes useful, if this material model is used to simulate the
behavior of reinforcement elements that have a very low bending stiffness, so it can be assumed
that when the reinforcement is loaded by compressive forces, buckling occurs and the strength of
the elements in compression is negligible. This is controlled by the command COMPRESSION
0 or 1, which deactivates and activates the compressive response respectively (for more details
see ATENA Input File Format).

2.7.5 Cyclic Reinforcement Model

The reinforcing steel stress-strain behavior can be described by the nonlinear model of
Menegotto and Pinto (1973). In ATENA this model is extended to account of the isotropic
hardening due to an arbitrary hardening law that can be specified for reinforcement (see Sections
2.7.2, 2.7.3). The stress in the cyclic model is calculated according to the following expression.
   0   r   *   r (2.121)
where

  b 
* * 1  b   * , * 
  r
R  R0 
c1
(2.122)
,
1    0  r c2  
1
*R
R

ATENA Theory 69
where R0 , c1 and c2 are experimentally determined parameters, and b the current hardening
modulus. The Fig. 2-45 shows the meaning of strain values  r ,  0 ,  and stress values  r and
 0 . These values changes for each cycle. The values with the subscript r indicate the point
where the cycle started, and the subscript 0 indicates the theoretical yield point that would be
reached during the unloading if the response would not have been modified by the hysteretic
behavior. During the calculation of this point the material stress-strain law is considered (see
Sections 2.7.2, 2.7.3)
Nincr .
  f R   eq  ,  eq 
*
  i
eq (2.123)
i 1

Fig. 2-45: Cyclic reinforcement model based on Menegotto and Pinto (1973).

2.7.6 Cyclic Reinforcement Model – Steel DRC

Another nonlinear constitutive model for reinforcement which captures cyclic behavior and is
implemented in ATENA is described by Dodd and Restrepo (1995) and further improved by Se-
Hyung Kim (2015).

70
 Fig. 2-46: Cyclic reinforcement model based on Dodd and Restrepo (1995) – backbone curve
definition points.

Fig. 2-47: Cyclic reinforcement model based on Dodd and Restrepo (1995) – effect of parameter .

ATENA Theory 71
2.8 Reinforcement Bond Models
The basic property of the reinforcement bond model is the bond-slip relationship. This
relationship defines the bond strength (cohesion)  b depending on the value of current slip
between reinforcement and surrounding concrete. ATENA contains three bond-slip models:
according to the CEB-FIB model code 1990, slip law by Bigaj and the user defined law. In the
first two models, the laws are generated based on the concrete compressive strength,
reinforcement diameter and reinforcement type. The important parameters are also the
confinement conditions and the quality of concrete casting.

2.8.1 CEB-FIP 1990 Model Code

b

Fig. 2-48: Bond-slip law by CEB-FIP model code 1990.


s
b   
max  s 
, 0  s  s1 (2.124)

 1

b   max
, s1  s  s 2 (2.125)

72
  ss 
b     max   f  , s 2  s  s3
2
(2.126)
s s 
max

3 2

b   f
, s3  s (2.127)

Table 2.8-1: Parameters for defining the mean bond strength-slip relationship for ribbed bars.

2 3 4 5
Value Unconfined concrete* Confined concrete**
Bond conditions Bond conditions
Good All other cases Good All other cases
S1 0.6 mm 0.6 mm 1.0 mm
S2 0.6 mm 0.6 mm 3.0 mm
S3 1.0 mm 2.5 mm clear rib spacing
 0.4 0.4

max
2.0 f 1.0 f 2.5 f 1.25 f
C C C C

f
0.15  max 0.40  max
* Failure by splitting of the concrete
**Failure by shearing of the concrete between the ribs
Table 2.8-2: Parameters for defining the bond strength-slip relationship for smooth bars.

Values Cold drawn wire Hot rolled bars

Bond conditions Bond conditions
Good All other Good All other cases
cases
0.01 mm 0.1 mm
s1  s 2  s3
 0.5 0.5

ATENA Theory 73
 max   f 0.1 f 0.05 f 0.3 f 0.15 f
C C C C

2.8.2 Bond Model by Bigaj

The second pre-defined bond model available in ATENA is based on the work by BIGAJ 1999.
This model depends on the bond quality, concrete cubic compressive strength f cu' and
reinforcement bar radius D . The slip law for this model is shown in Fig. 2-49.

3 4

Fig. 2-49: Bond law by BIGAJ 1999

The ascending part of the stress-slip law i.e. part a is modeled by a bi-linear curve. The
coordinates of the four points defining this stress-slip relationship are listed in the table below.

Table 2.8-3: Parameters for defining the bond strength-slip relationship for ribbed bars.

Concrete Bond Point 1 Point 2 Point 3 Point 4

Type quality
Excelent s/D 0.000 0.020 0.044 0.480

 b / 0.8 f cu' 0.500 3.000 0.700 0.000

f c' < 60
Good s/D 0.000 0.030 0.047 0.480

 b / 0.8 f cu' 0.500 2.000 0.700 0.000

Bad s/D 0.000 0.040 0.047 0.480

 b / 0.8 f cu' 0.500 1.000 0.700 0.000

Excelent s/D 0.000 0.012 0.030 0.340

74
  b / 0.88 f cu' 0.600 2.500 0.900 0.000
f c' > 60
Good s/D 0.000 0.020 0.030 0.340

 b / 0.88 f cu' 0.600 1.900 0.900 0.000

Bad s/D 0.000 0.025 0.030 0.340

 b / 0.88 f cu' 0.600 1.100 0.900 0.000

2.8.3 Memory Bond Material

The Memory Bond material is an improvement to better capture the response during cyclic
loading and unloading in general. It can be used with any of the above-mentioned bond strength
– bond slip envelope functions. The response only differs after the bond stress sign changes.
Instead of following the same envelope as during loading, the maximum bond stress is
determined by the additional parameter  1 , see Fig. 2-50. Admissible values are  res ≤  1 ≤  max ,
where  res is the residual bond stress (last value from the bond strength – bond slip function) and
 max the maximum bond stress (max. value from the bond strength – bond slip function).
In the figure, s is the current slip value, smax the maximum of the absolute slip value ever reached
(damage variable),   f ( s ) is the bond strength function.

Fig. 2-50: Memory Bond working diagram

The response for a slip change si  si 1  si is defined separately for 2 cases:

(1) Loading range s  smax

  f (s)
(2) Unloading range -smax < s < smax

ATENA Theory 75
 s  0    1

s  0    1

2.9 Microplane Material Model (CCMicroplane4)

The basic idea of the microplane model is to abandon constitutive modelling in terms of tensors
and their invariants and formulate the stress-strain relation in terms of stress and strain vectors
on planes of various orientations in the material, now generally called the microplanes. This
idea arose in G.I. Taylor’s (TAYLOR 1938) pioneering study of hardening plasticity of
polycrystalline metals. Proposing the first version of the microplane model, BAZANT 1984, in
order to model strain softening, extended or modified Taylor’s model in several ways (in detail
see BAZANT et al. 2000), among which the main one was the kinematic constraint between the
strain tensor and the microplane strain vectors. Since 1984, there have been numerous
improvements and variations of the microplane approach. A detailed overview of the history of
the microplane model is included in BAZANT et al 2000 and CANER and BAZANT 2000. In
what follows, we briefly review the derivation of the microplane model that is used in this work.
In the microplane model, the constitutive equations are formulated on a plane, called
microplane, having an arbitrary orientation characterized by its unit normal ni . The kinematic
constraint means that the normal strain  N and shear strains  M ,  L on the microplane are
calculated as the projections of the macroscopic strain tensor  ij :

1 1
 N  ni n j  ij ,  M 
2
 mi n j  m j ni   ij ,  L   li n j  l j ni   ij
2
(2.128)

where mi and li are chosen orthogonal vectors lying in the microplane and defining the shear
strain components. The constitutive relations for the microplane strains and stresses can be
generally stated as:
 N (t )  Ft0  N ( ),  L ( ),  M ( ) 
 M (t )  Gt 0  N ( ),  L ( ),  M ( ) (2.129)
 L (t )  Gt 0  N ( ),  L ( ),  M ( )
where F and G are functionals of the history of the microplane strains in time t. For a detailed
derivation of these functionals a reader is referred to BAZANT et al 2000 and CANER and
BAZANT 2000. The macroscopic stress tensor is obtained by the principle of virtual work that is
applied to a unit hemisphere . After the integration, the following expression for the
macroscopic stress tensor is recovered (BAZANT 1984):

3 Nm
 
 ij   s d   6  w sij(  ) , where sij   N ni n j  M  mi n j  m j ni   L  li n j  l j ni  (2.130)
2 
ij
 1 2 2
where the integral is approximated by an optimal Gaussian integration formula for a spherical
surface; numbers  label the points of the integration formula and w are the corresponding
optimal weights.

76
2.9.1 Equivalent Localization Element
The objective of the equivalent localization element is to achieve equivalence with the crack
band model. This basic idea is that the material properties and parameters of the softening
material model are not modified to account for the differences in the finite element size, but
rather the softening crack band is coupled in series with an elastically behaving layer, to obtain
equivalence. For brevity, this layer will henceforth be called the `spring’. For large finite
elements, the effective length of this added elastic spring, representing the thickness of the added
elastic layer having the elastic properties of the material, will be much larger than the size (or
thickness) of the localization zone (crack band). Thus, after the crack initiation, the energy stored
in the elastic spring can be readily transferred to the localization zone and dissipated in the
softening (i.e., fracturing) process.
Inside each finite element at each integration point, an equivalent localization element is
assumed. The localization element is a serial arrangement of the localization zone, which is
loading, and an elastic zone (spring), which is unloading. The total length of the element is
equivalent to the crack band size L (width), and can be determined using the same methods as
described in Section 2.1.3 (see Fig. 2-12). The width of the localization zone is given either by
the characteristic length of the material or by the size of the test specimen for which the adopted
material model has been calibrated.
The three-dimensional equivalent element is constructed by three serial arrangements of the
elastic zone (spring) and localization band. The spring-band systems are perpendicular to each
other, and they are arranged parallel to the principal strain directions (Fig. 2-51). The simplified
two-dimensional version is shown in Fig. 2-52. In this arrangement of spring-band systems it is
possible to identify the following unknown stresses and strains:
 ijb , 1 iju , 2 iju , 3 iju and  ijb , 1 iju , 2 iju , 3 iju
m
where superscript b denotes the quantities in the localization band and the symbol x u with
superscripts u and m defines the quantities in the elastic spring in the direction m .

ATENA Theory 77
 
ij ,ij
u u

ij ,ij
b b


ij ,ij
u u

3
L

ij ,ij
3 u u
h

1
h 2
h
1
L 2
L
Fig. 2-51: The arrangement of the three-dimensional equivalent localization element.

Finite element Localization

element
1

Elastic
springs
1
ij
u

2
2
1
h ijb iju


Localization h
2 band
L 2
L
Fig. 2-52: The simplified two-dimensional view of the spring-band arrangement.
Ideally, the chosen directions should be perpendicular to the planes of failure propagation. In
ATENA, it is assumed for them to be aligned with the principal axes of the total macroscopic
strain tensor, which in most cases should approximately correspond to the above requirement.

78
Altogether there are 48 unknown variables. In the subsequent derivations, it is assumed that
these stresses and strains are defined in the principal frame of the total macroscopic strain tensor.
The set of equations available for determining these variables starts with the constitutive
formulae for the band and the elastic springs:
 ijb  F ( ijb ) (2.131)
m
 iju  Dijlk m klu for m  1...3 (2.132)

The first formula (2.131) represents the evaluation of the non-linear material model, which in our
case is the microplane model for concrete. The second equation (2.132) is a set of three elastic
constitutive formulations for the three linear zones (springs) that are involved in the arrangement
at Fig. 2-51. This provides the first 24 equations, which can be used for the calculation of
unknown strains and stresses.
The second set of equations is provided by the kinematic constrains on the strain tensors.
1  b1 1 u 1
11  1
L  
11 h  11 L  1h  
1

 22  2  22b 2h  2 22u 2 L  2h 
L

1  b 3

 33  3  33 h  3 33u 3 L  3h 
L

1 1 (2.133)
1 
2 L
 L

12   1 12b 1h  112u 1L  1h   2 12b 2h  212u  2

L  2h  

1 1 1 
 23   2  23b 2h  2 23u  2 L  2h    3  23b 3h  3 23u  3 L  3h   
2 L L 
1 1 1 
13   1 13b 1h  113u  1L  1h    3 13b 3h  313u  3 L  3h   
2 L L 
These 6 additional equations can be written symbolically as:
11 1 
 ij   i  ijb i h  i iju  i L  i h    j  ijb j h  j iju  j L  j h    (2.134)
2 L L 
The next set of equations is obtained by enforcing equilibrium in each direction between the
corresponding stress components in the elastic zone and in the localization band. For each
direction m , the following condition must be satisfied:
 ijb m e j  m iju m e j for m  1...3 (2.135)

where m e j denotes coordinates of a unit direction vector for principal strain direction m . Since
the principal frame of the total macroscopic strain tensor is used the unit vectors have the
following coordinates:
1
e j  1, 0, 0  , 2 e j   0,1, 0  , 3e j   0, 0,1 (2.136)
The remaining equations are obtained by enforcing equilibrium between tractions on the other
surfaces of the band and the elastic zone (layer) imagined as a spring:
 ijb m e j  n iju m e j where m  1..3, n  1...3, m  n (2.137)

ATENA Theory 79
The equation (2.137) is equivalent to a static constraint on the remaining stress and strain
components of the elastic springs. Formulas (2.135) and (2.137) together with the assumption of
stress tensor symmetry represent the remaining 18 equations that are needed for the solution of
the three-dimensional equivalent localization element. These 18 equations can be written as:
 ijb  m iju for m  1...3 (2.138)
This means that the macroscopic stress must be equal to  ijb , i.e., the stress in the localization
element, and that the stresses in all the three elastic zones must be equal and to the microplane
stress  ijb . This also implies the equivalence of all the three elastic strain tensors.
Based on the foregoing derivations, it is possible to formulate an algorithm for the calculation of
unknown quantities in the three-dimensional equivalent localization element.
Input:  ij ,  ij ,  ijb ,  iju (2.139)

Initialization:  ijb   iju   ij (2.140)

(i )
i
L jh  j L ih
Step 1: d  iju  i j
Cijkl rkl(i 1) (2.141)
2L L
(i ) ( i 1) (i )
Step 2:  iju   iju  d  iju (2.142)

(i ) 2iL jL 2 i L j L  i L jh  j L ih u
Step 3:  ijb   ij   ij (2.143)
i
L jh  j L ih i
L j h  j L ih
(i ) (i )
Step 4: rij(i )   ijb   iju (2.144)

where Cijlk is the compliance tensor. The above iterative process is controlled by the following
convergence criteria;
d  iju (i ) rij( i ) rij( i ) T d  iju ( i )
e , e , e (2.145)
 ij  ijb  ijb  ij
The macroscopic stress is then equal to the stress in the localization band  ijb . More details about
the derivations of the above algorithm as well as various examples of application can be obtained
from the original reference CERVENKA et al. 2004. It should be noted that the described
equivalent localization element is used only if the calculated crack band size L (see Section
2.1.3) in each principal strain direction is larger than the prescribed localization band size h . For
smaller element sizes the equivalent localization approach is not used and mesh-dependent
results may be obtained.

2.10 References
BASQUIN, H.O. (1910), The exponential law of endurance tests, Proc. ASTM, 10 (II).
BAZANT, Z.P, OH, B.H (1983) - Crack Band Theory for Fracture of Concrete, Materials and
Structures, RILEM, Vol. 16, 155-177.
BAŽANT, Z.P., (1984), ‘Microplane model for strain controlled inelastic behavior’, Chapter 3 in
Mechanics of Engineering Materials (Proc., Conf. held at U. of Arizona, Tucson, Jan. 1984),
C.S. Desai and R.H. Gallagher, eds., J. Willey, London, 45-59.

80
BAŽANT, Z.P., CANER, F.C., CAROL, I., ADLEY, M.D., AND AKERS, S.A., (2000),
‘Microplane Model M4 for Concrete: I. Formulation with Work-Conjugate Deviatoric Stress’, J.
of Engrg. Mechanics ASCE, 126 (9), 944-961.
BIGAJ, A.J (1999) - Structural Dependence of Rotation Capacity of Plastic Hinges in RC Beams
and Slabs, PhD Thesis, Delft University of Technogy, ISBN 90-407-1926-8.
BRUEHWILER, E., and WITTMAN, F.H. (1990), “The Wedge Splitting Test, A New Method
of Performing Stable Fracture-Mechanics Tests”, Engineering Fracture Mechanics, Vol. 35, No.
1-3, pp. 117-125.
CANER, F.C., AND BAŽANT, Z.P., (2000) ‘Microplane Model M4 for Concrete: II. algorithm
and calibration.", J. of Engrg. Mechanics ASCE, 129 (9), 954-961.
CEB-FIP Model Code 1990, First Draft, Comittee Euro-International du Beton, Bulletin
d'information No. 195,196, Mars.
CEB 1988, Bulletin D’Information No 188, Fatigue of concrete structures, State of the art report.
CERVENKA, V., GERSTLE, K. (1972) - Inelastic Analysis of Reinforced Concrete Panels: (1)
Theory, (2) Experimental Verification and application, Publications IABSE, Zürich, V.31-00,
1971, pp.32-45, and V.32-II,1972, pp.26-39.
CERVENKA, V. (1985) - Constitutive Model for Cracked Reinforced Concrete, Journal ACI,
Proc. V.82, Nov-Dec., No.6,pp.877-882.
CERVENKA, V., PUKL, R., ELIGEHAUSEN, R. (1991) - Fracture Analysis of Concrete Plane
Stress Pull-out Tests, Proceedings, Fracture process in Brittle Disordered Materials, Noordwijk,
Holland, June 19-21.
CERVENKA, V., PUKL, R., OZBOLT, J., ELIGEHAUSEN, R. (1995), Mesh Sensitivity
Effects in Smeared Finite Element Analysis of Concrete Structures, Proc. FRAMCOS 2, 1995,
pp 1387-1396.
CERVENKA, V., PUKL, R. (1992) - Computer Models of Concrete Structures, Structural
Engineering International, Vol.2, No.2, May 1992. IABSE Zürich, Switzerland, ISSN 1016-
8664, pp.103-107.
CERVENKA, V., PUKL, R., OZBOLT, J., ELIGEHAUSEN, R. (1995) - Mesh Sensitivity
Effects in Smeared Finite Element Analysis of Concrete Fracture, Proceedings of FRAMCOS2,
Zurich, Aedificatio.
CERVENKA, V., CERVENKA, J. (1996) - Computer Simulation as a Design Tool for Concrete
Structures, ICCE-96, proceedings of The second International Conference in Civil Engineering
on Computer Applications Research and Practice, 6-8 April, Bahrain.
CERVENKA, J, CERVENKA, V., ELIGEHAUSEN, R. (1998), Fracture-Plastic Material Model
for Concrete, Application to Analysis of Powder Actuated Anchors, Proc. FRAMCOS 3, 1998,
pp 1107-1116.
ČERVENKA, J., BAŽANT Z.P., WIERER, M., (2004), `Equivalent Localization Element for
Crack Band Approach to Mesh Sensitivity in Microplane Model’, submitted for publication, Int.
J. for Num. Methods in Engineering.
ČERVENKA, J., PRYL, D., (2007), `Fatigue Modelling of Crack Growth by Finite Element
Method and Smeared Crack Approach’, Internal Report 2007-08-03-2002-DP, Cervenka
Consulting.
CRISFIELD, M.A., WILLS, J. (1989)- The Analysis of Reinforced Concrete Panels Using
Different Concrete Models, Jour. of Engng. Mech., ASCE, Vol 115, No 3, March, pp.578-597.

ATENA Theory 81
CRISFIELD, M.A. (1983) - An Arc-Length Method Including Line Search and Accelerations,
International Journal for Numerical Methods in Engineering, Vol.19,pp.1269-1289.
CHEN, W.F, SALEEB, A.F. (1982) - Constitutive Equations For Engineering Materials, John
Willey \& Sons, ISBN 0-471-09149-9.
DARWIN, D., PECKNOLD, D.A.W. (1974) - Inelastic Model for Cyclic Biaxial Loading of
Reinforced Concrete, Civil Engineering Studies, University of Illinois, July.
DE BORST, R. (1986), Non-linear analysis of frictional materials, Ph.D. Thesis, Delft
University of Technology, 1986.
DRUCKER, D.C., PRAGER, W., Soil Mechanics and Plastic Analysis or Limit Design, Q.
Appl. Math., 1952, 10(2), pp 157-165.
DYNGELAND, T. (1989) - Behavior of Reinforced Concrete Panels, Dissertation, Trondheim
University, Norway, BK-report 1989:1
FEENSTRA, P.H., Computational Aspects of Bi-axial Stress in Plain and Reinforced Concrete.
Ph.D. Thesis, Delft University of Technology, 1993.
FEENSTRA, P.H., ROTS, J.G., AMESEN, A., TEIGEN, J.G., HOISETH, K.V., A 3D
Constitutive Model for Concrete Based on Co-rotational concept. Proc. EURO-C 1998, 1, pp.
13-22.
Taerwe, L., and Matthys, S. (2013), Fib model code for concrete structures 2010, Wilhelm Ernst
& Sohn, Berlin, Germany, ISBN 978-3-433-03061-5.
fib Model Code for Concrete Structures 2010, (2013), Wilhelm Ernst & Sohn, Berlin, Germany,
ISBN 978-3-433-03061-5.
ETSE, G., Theoretische und numerische Untersuchung zum diffusen und lokalisierten Versagen
in Beton, Ph.D. Thesis, University of Karlsruhe 1992.
FELIPPA, C. (1966) - Refined Finite Element Analysis of Linear and Nonlinear Two-
Dimensional Structures, Ph.D. Dissertation, University of California, Engineering, pp.41-50.
GRASSL, P., LUNDGREN, K., and GYLLTOFT, K. (2002) “Concrete in compression: A
plasticity theory with a novel hardening law”, International Journal of Solids and Structures,
39(20), 5205-5223.
VAN GYSEL, A., and TAERWE, L. (1996) “Analytical formulation of the complete stress-
strain curve for high strength concrete”, Materials and Structures, RILEM, 29(193), 529-533.
HARTL, G. (1977) “Die Arbeitlinie Eingebetete Staehle bei erst und kurz=Belastung”,
Dissrtation, Univbersitaet Innsbruck
HORDIJK, D.A. (1991) - Local Approach to Fatigue of Concrete, Doctor dissertation, Delft
University of Technology, The Netherlands, ISBN 90/9004519-8.
Juhász, K. P. (2013) “Modified fracture energy method for fibre reinforced concrete”, Fibre
Concrete 2013, Prague, Czech Republic, pp. 89-90, ISBN 978-80-01-05238-9.
KABELE, P. (2002) - Equivalent Continuum Model of Multiple Cracking, Engineering
Mechanics 2002, 9 (1/2), pp.75-90, Assoc.for Engineering Mechanics, Czech Republic
KESSLER-KRAMER, CH., (2002) “Zugverhalten von Beton unter Ermüdungsbeanspruchung”,
Schriftenreihe des Instituts für Massivbau und Baustofftechnologie, Heft 49, Karlsruhe.
KLAUSEN, D. (1978), Festigkeit und Schadigung von Beton bei haufig wiederholter
Beanschpruchung, PhD Thesis, University of Technology Darmstadt, 85 pp.

82
KOLLEGGER, J. - MEHLHORN, G. (1988) - Experimentelle und Analytische Untersuchungen
zur Aufstellung eines Materialmodels für Gerissene Stahbetonscheiben, Nr.6 Forschungsbericht,
Massivbau, Gesamthochschule Kassel.
KOLMAR, W. (1986) - Beschreibung der Kraftuebertragung über Risse in nichtlinearen Finite-
Element-Berechnungen von Stahlbetontragwerken", Dissertation, T.H. Darmstadt, p. 94.
KUPFER, H., HILSDORF, H.K., RÜSCH, H. (1969) - Behavior of Concrete under Biaxial
Stress, Journal ACI, Proc. V.66, No.8, Aug., pp.656-666.
MARGOLDOVA, J., CERVENKA V., PUKL R. (1998), Applied Brittle Analysis, Concrete
Eng. International, November/December 1998.
MENETREY, P., WILLAM, K.J. (1995), Triaxial failure criterion for concrete and its
generalization. ACI, Structural Journal, 1995, 92(3), pp 311-318.
MENETREY, Ph., WALTHER, R., ZIMMERMAN, Th., WILLAM, K.J., REGAN, P.E.
Simulation of punching failure in reinforced concrete structures. Journal of Structural
Engineering, 1997, 123(5), pp 652-659.
MIER J.G.M van (1986) - Multiaxial Strain-softening of Concrete, Part I: fracture, Materials
and Structures, RILEM, Vol. 19, No.111.
MINER M.A. (1945), Cumulative damage in fatigue. Transactions of the American Society of
Mechanical Engineering, 67:A159-A164.
OLIVIER, J., A Consistent Characteristic Length for Smeared Cracking Models, Int. J. Num.
Meth. Eng., 1989, 28, pp 461-474.
OWEN, J.M., FIGUEIRAS, J.A., DAMJANIC, F., Finite Element Analysis of Reinforced and
Pre-stressed concrete structures including thermal loading, Comp. Meth. Appl. Mech. Eng., 1983,
41, pp 323-366.
PALMGREN, A. (1924), Die Lebensdauer von Kugellagern. Zeitschrift Verein Deutscher
Ingenieure, 68(14):339-341.
PAPANIKOLAOU, V.K., and KAPPOS, A.J. (2007) “Confinement-sensitive plasticity
constitutive model for concrete in triaxial compression”, International Journal of Solids and
Structures, 44(21), 7021-7048.
PRAMONO, E, WILLAM, K.J., Fracture Energy-Based Plasticity Formulation of Plain
Concrete, ASCE-JEM, 1989, 115, pp 1183-1204.
PRYL, D., CERVENKA, J., and PUKL, R. (2010) “Material model for finite element modelling
of fatigue crack growth in concrete”, Procedia Engineering, 2 (2010) 203–212.
PRYL, D., PUKL, R., and CERVENKA, J. (2013) “Modelling high-cycle fatigue of concrete
specimens in three-point bending”, Life-Cycle and Sustainability of Civil Infrastructure Systems
(Eds. Strauss, Frangopol & Bergmeister) 1303–1306.
PRYL, D., MIKOLÁŠKOVÁ, J., PUKL, R. (2014) “Modeling Fatigue Damage of Concrete”,
Key Engineering Materials, 2014 Vols. 577-578, pp. 385-388, ISSN: 1662-9795
RAMM, E. (1981) - Strategies for Tracing Non- linear Responses Near Limit Points, Non- linear
Finite Element Analysis in Structural Mechanics, (Eds. W.Wunderlich, E.Stein, K.J.Bathe)
RASHID, Y.R. (1968), Ultimate Strength Analysis of Pre-stressed Concrete Pressure Vessels,
Nuclear Engineering and Design,1968, 7, pp 334-344.
ROTS, J.G., BLAAUWENDRAAD, J., Crack models for concrete: discrete or smeared? Fixed,
multi-directional or rotating? HERON 1989, 34(1).

ATENA Theory 83
SAE, AE-4, Fatigue Design Handbook
SIMO, J.C., JU, J.W., Strain and Stress-based Continuum Damage Models-I. Formulations, II-
Computational Aspects, Int. J. Solids Structures, 1987, 23(7), pp 821-869.
SIMO, J.C., KENNEDY, J.G., GOVINDJEE, S., (1988), Non-smooth Multi-surface Plasticity
and Visco-plasticity. Loading/unloading Conditions and Numerical Algorithms, Int. J. Num.
Meth. Eng., 26, pp 2161–2185.
TAYLOR, G.I., (1938), ‘Plastic strain in metal’, J. Inst. Metals, 62, 307-324.
UIJL, J. den, BIGAJ, A. J. (1996): A Bond Model for Ribbed Bars Based on Concrete
Confinement. Heron, Vol.41, No.3.
VECCHIO, F.J., COLLINS, M.P (1986)- Modified Compression-Field Theory for Reinforced
Concrete Beams Subjected to Shear, ACI Journal, Proc. V.83, No.2, Mar-Apr., pp 219-231.
VOS, E. (1983) - Influence of Loading Rate and Radial Pressure on Bond in Reinforced
Concrete, Dissertation, Delft University, pp.219-220.
WILKINS, M.L., Calculation of Elastic-Plastic Flow, Methods of Computational Physics, 3,
Academic Press, New York, 1964.
DODD, L. L., and J. I. RESTREPO-POSADA. "Model for predicting cyclic behavior of
reinforcing steel." Journal of structural engineering 121.3 (1995): 433-445.
KIM, SE-HYUNG. Cyclic uniaxial constitutive model for steel reinforcement. Diss. Virginia
Tech, 2015.

84
3 FINITE ELEMENTS

3.1 Introduction
The preceding chapters dealt with the general formulation of the problem, geometric and
constitutive equations. All expressions were derived independently of the structural shape, the
finite elements used etc. Here, an information about finite elements currently implemented in
ATENA is given.

t
h2
3
7 6 h1

4 9 2

5 s
8
r 1

t t

3 3 h6
h8 7 6
7 6
2 4 9 2
4 9
5 5 s
8 s 8

1 r 1
r

t
h9
3
7 6

4 9 2

5 s
8
r 1

Fig. 3-1 Examples of interpolation function for plane quadrilateral elements.

The available elements can be divided into three groups: plane elements for 2D, 3D and axi-
symmetric analysis, solid 3D elements and special elements, which comprises elements for
modeling external cable, springs, gaps etc.
With few exceptions all elements implemented in ATENA are constructed using isoparametric
formulation with linear and/or quadratic interpolation functions. The isoparametric formulation
of one-, two- and three-dimensional elements belong to the "classic" element formulations. This

ATENA Theory 85
is not because of its superior properties, but since it is a versatile and general approach with no
hidden difficulties and, also very important, these elements are easy to understand. This is very
important particularly in nonlinear analysis. For example, it is highly undesirable to add element-
related problems to problems related to e.g. material modeling.
Big advantage of ATENA isoparametric elements is that their interpolation functions hi (r , s, t )
are constructed in hierarchical manner. Take an example of plane quadrilateral element. Some of
its interpolation functions are depicted in Fig. 3-1. The 1st four functions, i.e. functions h1 (r , s, t )
to h4 (r , s, t ) has to be always present in the interpolation set, (to ensure bilinear approximation).
Then, any additional function h6 (r , s, t ) through h9 (r , s, t ) can be added independently. This
would involve adding the new function itself and amendments to the already present
interpolation functions. This approach (and use of C++ templates) makes possible that one
element formulation generates quadrilateral elements with nodes (1,2,3,4), (1,2,3,4,5),
(1,2,3,4,6), ... (1,2,3,4,8), (1,2,3,4,9), (1,2,3,4,5,6), (1,2,3,4,5, 7), ... (1,2,3,8,9), ...
(1,2,3,4,5,6,7,8,9). Additional mid-side points are particularly useful for changing mesh density,
(i.e. element size), see Fig. 3-2, as they allow change of mesh density without need triangular
elements.
Although the concept of hierarchical elements was described for plane quadrilateral elements, in
ATENA it applies for plane triangular elements, 3D bricks, tetrahedral and wedge elements, too.
Always there is a set of basic interpolation function that can be extended by any “higher”
interpolation function.
Apart of interpolation functions finite element properties depend strongly on numerical
integration scheme used to integrate element stiffness matrix, element nodal forces etc. In Atena,
majority of elements are integrated by Gauss integration scheme that ensure n(n  1) order
accuracy, where n is degree of the polynomial used to approximate the integrated function.

9 10 11 9 10 11

6 7 8 7
6 8

3 4 5 3 4 5

1 2 1 2

Solution with Standard

hierarchical solution
elements

Fig. 3-2 Change of finite element mesh density.

86
3.2 Truss 2D and 3D Element
2D and 3D truss elements in ATENA are coded in group of elements CCIsoTruss<xx> ...
CCIsoTruss<xxx>. The string in < > describes present element nodes, (see Atena Input File
Format document for more information). These are isoparametric elements integrated by Gauss
integration at 1 or 2 integration points for the case of linear or quadratic interpolation, i.e. for
elements with 2 or 3 element nodes, respectively. They are suitable for plane 2D as well as 3D
analysis problems. Geometry, interpolation functions and integration points of the elements are
given in Fig. 3-3, Table 3.2-1 to Table 3.2-3.

s
y
2 r
3 CCIsoTruss<xx>
1
CCIsoTruxx<xxx>
x

Fig. 3-3 Geometry of CCIsoTruss<...> elements.

Table 3.2-1 Interpolation functions of CCIsoTruss<...> elements.

Node Function Include only if node 3 is

i defined
hi

1 1 1
(1  r )  h3
2 2
2 1 1
(1  r )  h3
2 2
3 (1  r 2 )

Table 3.2-2 Sample points for Gauss integration of 1 node CCIsoTruss<xx> element.

Integration point Coordinate r Weight

1 0. 2.

ATENA Theory 87
 Table 3.2-3 Sample points for Gauss integration of 2 and 3 nodes CCIsotruss<xxx> elements.

Integrati Coordinate r Weight

on point

1 0.577350269189626 1.
2 -0.577350269189626 1.
The element vectors and matrices for Total Lagrangian formulation, configuration at time t and
iteration (i) are as follows. Note that they are equally applicable for Updated Lagrangian
formulation upon applying changes related to the element reference coordinate system
(undeformed vs. deformed element axis.). The formulation is present for 3-nodes element option.
The 2-nodes variant is obtained by simply neglecting the terms for the element mid-point.
An arbitrary point on the truss element has at reference time t coordinates t X  [t x1 , t x1 , t x1 ] :
t
x1 t x11 h1  t x12 h2  t x13 h3

t
x2 t x12 h1  t x22 h2  t x23 h3 (3.1)

t
x3 t x31 h1  t x32 h2  t x33 h3

At time t  t ( i 1) the same point has coordinates t  t

X ( i 1) :

t  t
x1( i 1)  (t x11  t u11(i 1) ) h1  (t x12  t u12(i 1) )h2  (t x13  t u13(i 1) )h3

t  t
x2( i 1)  (t x12  t u1(2 i 1) )h1  (t x22  t u22(i 1) ) h2  (t x23  t u23(i 1) )h3 (3.2)

t  t
x3( i 1)  (t x31  t u31(i 1) )h1  (t x32  t u32(i 1) ) h2  (t x33  t u33(i 1) )h3
t  t
and at time t  t (i ) coordinates X (i )
t  t
x1( i )  (t x11  t u11( i ) ) h1  (t x12  t u12( i ) )h2  (t x13  t u13(i ) )h3

t  t
x2( i )  (t x12  t u1(2 i ) )h1  (t x22  t u22(i ) ) h2  (t x23  t u23( i ) )h3 (3.3)

t  t
x3( i )  (t x31  t u31( i ) )h1  (t x32  t u32( i ) ) h2  (t x33  t u33(i ) )h3

Increment of Green Lagrange strain t 11( i )  

t  t ( i ) t  t
t 11  t  tt11( i 1) (at time t  t , iteration (i )
with
to configuration at time t ) is calculated:

88
   t  t (i )  2  t  t (i 1) 2 
 l   l  
   r  r
(i ) 1
    
 (3.4)
t 11 2
2  tl  
   

  r  

where truss length differentials are
2 2 2 2
  t l    t x1    t x2    t x3 
       
 r   r   r   r 

2 2 2 2
  t  t l (i 1)    t  t x1(i 1)    t  t x2(i 1)    t  t x3(i 1) 
        (3.5)
 r   r   r   r 

2 2 2 2
  t  t l (i )    t  t x1(i )    t  t x2(i )    t  t x3( i ) 
       
 r   r   r   r 

Substituting (3.5), (3.3) into (3.4) after some math manipulation it can be derived:
 h1 h1 t 1 h1 h2 t 2 h1 h3 t 3 
 r r x1  r r x1  r r x1 
 
 h2 h1 t x1  h2 h2 t x 2  h2 h3 t x 3 
 r r 1 r r 1 r r 1 
 
 h3 h1 t x11  h3 h2 t x12  h3 h3 t x13 
 r r r r r r 
 h h h h h h 
 1 1 t x12  1 2 t x22  1 3 t x23 
 r r r r r r 
t  t 1  h2 h1 t 1 h2 h2 t 2 h2 h3 t 3 
t BL 0  2  x2  x2  x2 
  t l   r r r r r r 
   h3 h1 t 1 h3 h2 t 2 h3 h3 t 3 
 r   x2  x2  x2 
 r r r r r r 
 h1 h1 t 1 h1 h2 t 2 h1 h3 t 3 
 r r x3  r r x3  r r x3 
 
 h2 h1 t x1  h2 h2 t x 2  h2 h3 t x 3 
 r r 3 r r 3 r r 3 
 
 h3 h1 t x31  h3 h2 t x32  h3 h3 t x33  (3.6)
 r r r r r r 

ATENA Theory 89
  h1 h1 t  t 1(i 1) h1 h2 t  t h1 h3 t  t 3(i 1) 
 r r u1  u12(i 1)  u1 
r r r r
 
 h2 h1 t  t u1(i 1)  h2 h2 t  t 2( i 1)
u1 
h2 h3 t  t 3(i 1) 
u1
 r r 1
r r r r 
 
 h3 h1 t  t u11( i 1)  h3 h2 t  t 2( i 1)
u1 
h3 h3 t  t 3(i 1) 
u1
 r r r r r r 
 h h h h h h 
 1 1 t  t u1(2 i 1)  1 2 t  t 2( i 1)
u2  1 3 t  t u23(i 1) 
 r r r r r r 
t  t 1  h2 h1 t  t 1(i 1) h2 h2 h h 
t BL( i11)  2  u2  t  t 2( i 1)
u2  2 3 t  t u23(i 1) 
  t l   r r r r r r 
   h3 h1 t  t 1( i 1) h3 h2  h h 
 r   u2  t  t 2( i 1)
u2  3 3 t  t u23(i 1) 
 r  r r r r r 
 h1 h1 t  t 1(i 1) h1 h2 t  t 2( i 1) h1 h3 t  t 3(i 1) 
 r r u3  u3  u3 
r r r r
 
 2 1 t  t u1(i 1)  h2 h2
 h h  t  t 2( i 1)
u3 
h2 h3 t  t 3(i 1) 
u3
 r r 3
r r r r 
 
 h3 h1 t  t u31( i 1)  h3 h2 h h
t  t 2( i 1)
u3  3 3 t  t u33(i 1)  (3.7)
 r r r r r r 
and
 h1 h2 h3 
 r 0 0 0 0 0 0 
r r
 
1  h1 h2 h3
t  t ( n 1)
BNL  t 0 0 0 0 0 0  (3.8)
l 
t
r r r
  h1 h2

h3 
 r   0 0 0 0 0 0
 r r r 

The 2nd Piola-Kirchhoff stress matrix and tensor are:

 t  tt S11(i 1) 0 0 
t  t  
t S ( i 1)  0 t  t ( i 1)
t S11 0 ,
t  t
t S (i 1)  [ t  tt S11( i 1) ] (3.9)
 0 0 t  t ( i 1) 
t S11
 
t  t (i )
The formulation is completed by relationship for element deformation gradient t X 1,1 , which
yields:
  t  t l (i ) 
 
t  t  r 
t
(i )
X 1,1  (3.10)
 tl 
 
 r 
Note that 2-nodes truss element has constant strains along its length and thus the increment of
Green Lagrange strain can be calculated directly, (i.e. not using differentials truss length as it
was the case of (3.4) ):

90
 1 l   l  
 t  t ( i ) 2 t  t (i 1) 2 
 (i )
 (3.11)
2 
t 11 t 2
l
 
This yields a bit simpler element formulation (with the same results). However, for the sake of
preserving unified approach to all truss elements, ATENA uses even in this case the equation
(3.4).

3.3 Plane Quadrilateral Elements

Plane quadrilateral elements in ATENA are coded in group of elements CCIsoQuad<xxxx> ...
CCIsoQuad<xxxxxxxxx>. The string in < > describes present element nodes (see Atena Input
File Format document for more information). These are isoparametric elements integrated by
Gauss integration at 4 or 9 integration points for the case of bilinear or bi-quadratic interpolation,
i.e. for elements with 4 or 5 and more element nodes, respectively. They are suitable for plane
2D, axisymmetric and 3D problems.
CCIsoQuad2_5<...> elements present a simplified 3D formulation of the CCIsoQuad<...>
elements. Their higher execution performance is achieved at cost of omitting some nonlinear
terms, see below.
Geometry, interpolation functions and integration points of the elements are given in Fig. 3-4
and in the subsequent tables.

y 2 s
5
1 CCIsoQuad<xxxx>
6
CCIsoQuad<xxxxx>
9 CCIsoQuad<xxxx_x>
8 r ....
3 CCIsoQuad<xxxx_x_x_>
7 ....
4 x CCIsoQuad<xxxxxxxxx>

Fig. 3-4 Geometry of CCIsoQuad<...> elements.

ATENA Theory 91
 Table 3.3-1: Interpolation functions of CCIsoQuad<...> elements.

Node Function hi Include only if node i is defined

i

i=5 I=6 i=7 i=8 i=9

1 1 1 1 1
(1  r )(1  s )  h5  h8 h9
4 2 2 4
2 1 1 1 1
(1  r )(1  s )  h5  h6 h9
4 2 2 4
3 1 1 1 1
(1  r )(1  s)  h6  h7 h9
4 2 2 4
4 1 1 1 1
(1  r )(1  s)  h7  h8 h9
4 2 2 4
5 1 1
(1  r 2 )(1  s )  h9
2 2
6 1 1
(1  s 2 )(1  r )  h9
2 2
7 1 1
(1  r 2 )(1  s )  h9
2 2
8 1 1
(1  r )(1  s 2 )  h9
2 2
9 1
(1  r 2 )(1  s 2 )
2

Table 3.3-2: Sample points for Gauss integration of 4 nodes CCIsoQuad<...> element.

Integration Coordinate r Coordinate s Weight

point

1 0.577350269189626 0.577350269189626 1.
2 0.577350269189626 -0.577350269189626 1.
3 -0.577350269189626 0.577350269189626 1.
4 -0.577350269189626 -0.577350269189626 1.

92
 Table 3.3-3: Sample points for Gauss integration 5 to 9 nodes CCIsoQuad<...> elements.

Integrati Coordinate r Coordinate s Weight

on point

1 0.774596669241483 0.774596669241483 0.3086419753

2 0.774596669241483 0. 0.4938271605
3 0.774596669241483 -0.774596669241483 0.3086419753
4 0. 0.774596669241483 0.4938271605
5 0. 0. 0.7901234568
6 0. -0.774596669241483 0.4938271605
7 -0.774596669241483 0.774596669241483 0.3086419753
8 -0.774596669241483 0. 0.4938271605
9 -0.774596669241483 -0.774596669241483 0.3086419753

Equations (3.12) through (3.21) present CCIsoQuad<...> axisymmetric element formulation. 2D

element formulation is simply obtained by removing terms associated with circumferential
strains and stresses t  tt 33(i ) , t  tt S33( i ) .
Incremental strains:


(i )
t 11  t u1,1
(i )
 t  tt u1,1
( i 1)
t u1,1 
(i ) t  t ( i 1)
t u2,1 t u2,1 
(i ) 1
2
 u    u  
(i ) 2
t 1,1 t
(i ) 2
2,1

t  22(i )  t u2,2
(i )
 t  tt u1,2
( i 1)
t u1,2 
(i ) t  t ( i 1)
t u2,2 t u2,2 
(i ) 1
2
 ut 1,2   u  
(i ) 2
t
(i ) 2
2,2

1 (i )

(i )
t 12 
2
 t u1,2  t u2,1(i )  
1 t  t (i 1) (i ) t  t ( i 1) (i ) t  t (i 1) (i ) t  t (i 1) (i )
2
 t u1,1 t u1,2  t u2,1 t u2,2  t u1,2 t u1,1  t u2,2 t u2,1  
1 (i ) (i )
2
 t u1,1 t u1,2  t u2,1(i ) t u2,2(i ) 
2
u (i ) t  t
u1( i ) u1( i ) 1  u (i ) 
 (i )
 t1    t1 
 x
t 33
x1 t 2
2  x1  (3.12)
1

ATENA Theory 93
Displacement derivatives:
  t  t ui(i ) t  t ui(i 1) 
t
(i )
u
i, j 
t x j
(3.13)
t  t ( i 1)
t  t ( i 1)  u
t i, j u  i

 xjt

Strains and matrices to calculate them:

t  ( i )  t  tt BL(i 1) U (i )

t  ( i )   t 11(i ) , t  22(i ) , 2 t 12(i ) , t  33(i )  (3.14)

U (i ) t  t U ( i ) t  t U (i 1)  u11(i ) , u1(2 i ) , u12(i ) , u22(i ) ,.....u1n (i ) , u2n (i ) 

Linear strain-displacement matrix:

t  t
t BL(i 1)  t  t
t BL 0  t  tt BL(i11) (3.15)
Linear strain-displacement matrix – constant part:
 t h1,1 0 t h2,1 0 ... 0 0 
 0 
 t h1,2 0 t h2,2 ... 0 t hn ,2 
t  t
B
t L0   t h1,2 t h1,1 t h2,2 t h2,1 ... t hn ,2 t hn ,1
 (3.16)
 
 h1 0
h2
0
h
... t n 0 
 t x1 t
x1 x1 

where

hi
t hi , j 
t x j

(3.17)
ui(i ) t  t ui(i ) t  t ui(i 1)

n
t
x1   hk t x1k
k 1

94
Linear strain-displacement matrix – non-constant part:
 t t ( i 1)
l11 t h1,1 t t ( i 1)
l
21 t 1,1 h t t ( i 1)
l
11 t 2,1 h t t ( i 1)
l h
21 t 2,1
 t t ( i 1) t t ( i 1) t t ( i 1) t t ( i 1)
 l11 t h1,2 l
21 t 1,2 h l
11 t 2,2 h l h
21 t 2,2
t t
t
( i 1)
B
L1   t t ( i 1)
l11 t h1,2  t t l11( i 1) t h1,1 t t ( i 1)
l21 t 1,2h  t t ( i 1)
l
21 t 1,1 h t t ( i 1)
l
11 t 2,2 h t t ( i 1)
 l
11 t 2,1 h t t ( i 1)
l h 
21 t 2,2
t t ( i 1)
l h
21 t 2,1

 t t ( i 1) h1 t t ( i 1) h
l33 t 0 l 2
0
 x1
33 t
x
 1

... t t ( i 1)

l h
11 t n ,1
t t ( i 1)
l
21 t n ,1 h 
t t ( i 1) t t ( i 1) 
... l h
11 t n ,2 l
21 t n ,2 h 
(3.18)
... t t ( i 1)
l h
11 t n ,2  t t ( i 1)
l11 t n ,1 h t t ( i 1)
l
21 t n ,2 h t t ( i 1)
 l
21 t n ,1 h 

t t ( i 1) hn 
... l 0
33 t
x 
1 
where
n
t  t ( i 1)
l
11   t hk ,1 t  tt u1k ( i 1)
k 1

n
t  t ( i 1)
l
12   t hk ,2 t  tt u1k (i 1)
k 1

n
t  t ( i 1)
l
21   t hk ,1 t  tt u2k (i 1) (3.19)
k 1

n
t  t ( i 1)
l
22   t hk ,2 t  tt u2k (i 1)
k 1

1 n
t  t ( i 1)
l  t  hk t  t k ( i 1)
u
x1 k 1
33 t 1

Nonlinear strain-displacement matrix

 t h1,1 0 t h2,1 0 ... hn ,1
t 0 
 h 0 0 
 t 1,2 t h2,2 0 ... t hn ,2

t  t ( i 1)
 0 t h1,1 0 t h2,1 ... 0 t hn ,1

t BNL   (3.20)
 0 t h1,2 0 t h2,2 ... 0 t hn ,2 
 h1 h2 h 
 t 0 t
0 ... t n 0 
 x1 x1 x1 

ATENA Theory 95
2nd Piola-Kirchhoff stress tensor and vector
 t  tt S11( i 1) t  t ( i 1)
t 12S 0 0 0 
 t  t ( i 1) t  t ( i 1) 
 t S 21 t 22S 0 0 0 
t S ( i 1)
 0 0 t  t ( i 1)
t 11S t  t ( i 1)
t 12S 0 
 t  t ( i 1) t  t ( i 1) 
 0 0 t 21S t 22S 0 
 0 0 0 0 t  t ( i 1) 
 t S33 
(3.21)
t S (i )   t  tt S11( i 1) t  t
t S 22(i 1) t  t
t S 21(i 1) t  t
t S33(i 1) 

In case of the simplified 3D analysis, i.e. elements CCIsoQuad2_5<...>, the equations are further
extended as follows:
All element matrices and vectors are computed with respect to element local coordinate
system xlocal ,1 , xlocal ,2 using equations in (3.12) through (3.21). They are transformed into
3D global coordinate system by means of simple transformation:
M global  T M local T T , vglobal  T vlocal (3.22)

where
M global , M local , vglobal , vlocal are global and local finite element matrices and vectors,

T is transformation matrix from local to global coordinate system:

 cos( xlocal ,1 , xglobal ,1 ), cos( xlocal ,2 , xglobal ,1 ) 
 
T   cos( xlocal ,1 , xglobal ,2 ), cos( xlocal ,2 , xglobal ,2 )  (3.23)
 cos( xlocal .1 , xglobal ,3 ), cos( xlocal ,2 , xglobal ,3 ) 
 
where:
xlocal ,i , xglobal , i are local and global coordinates (in 2D and 3D space).

The local element coordinate system (see Fig. 3-5) is defined by local xlocal ,1 , xlocal ,2 , xlocal ,3
coordinates. All of them pass through origin of the global (reference) coordinate system. The
axes xlocal ,1 and xlocal ,2 constitute a local coordinates element plane that is parallel to the element.
The axis xlocal ,3 is perpendicular to the element and the axis xlocal ,1 is defined as a projection of
global x1 axis to the local coordinate element plane. An exception to that is, when the element is
normal to the global x1 . In this case the local xlocal ,1 coincides with the global x2 axis.

The present definition of local element coordinate system depends on plane of the finite element,
but it does not depend on its shape itself. This is very important property, as ATENA supports
use of local (instead of global) nodal degrees of freedom and, (of course) these degrees of
freedom must refer to a coordinate system common to all elements of the plane, in which they
lie.

96
 x3
3

xl o c a l , 3

4 O

X
X’ 2
xl o c a l , 2

1 x2
x1 xl o c a l , 1

Fig. 3-5 Local plane element coordinate system.

Full 3D formulation of the CCIsoQuad<...> elements is much the same as that for simplified 3D
elements CCIsoQuad2_5<...>. The only difference is that the matrix t0 BNL will include also terms
related to the „out-of-element-plane“ direction:
 t h1,1 0 0 t h2,1 0 0 t h3,1 0 0 t hN ,1 0 0 
 0 0 0 0 
 t h1,2 0 t h2,2 0 t h3,2 0 t hN ,2 0
t  t ( i 1)
 0 h 0 0 h2,1 0 0 h3,1 0 0 hN ,1 0 
t BNL  t 1,1 t t
...... t
 (3.24)
 0 t h1,2
0 0 t h2,2
0 0 t h3,2
0 0 t hN ,2
0 
 0 0 t h1,1 0 0 t h2,1 0 0 t h3,1 0 0 t hN ,1

 
 0 0 t h1,2 0 0 t h2,2 0 0 t h3,2 0 0 t hN ,2 


3.4 Plane Triangular Elements

Plane triangular elements in ATENA are coded in group of elements CCIsoTriangle<xxx> ...
CCIsoTriangle<xxxxxx>. The string in < > describes present element nodes (see Atena Input
File Format document for more information). These are isoparametric elements integrated by
Gauss integration at 1 or 3 integration points for the case of bilinear or bi-quadratic interpolation,
i.e. for elements with 3 or 4 and more element nodes, respectively. They are suitable for plane

ATENA Theory 97
2D, axisymmetric and 3D problems. Geometry, interpolation functions and integration points of

3 s
CCIsoTriangle<xxx>
...
6 CCIsoTriangle<xxxxxx>
y
5

1
4 r
2

the elements are given in x

Fig. 3-6, Table 3-1, Table 3-2, and Table 3-3.

3 s
CCIsoTriangle<xxx>
...
6 CCIsoTriangle<xxxxxx>
y
5

1
4 r
2

x
Fig. 3-6: Geometry of CCIsoTriangle<...> elements.

Table 3-1: Interpolation functions of CCIsoTriangle<...> elements.

Node Function hi Include only if node i is

i defined
i=4 i=5 i=6
1 1 r  s 1 1
 h4  h6
2 2
2 r 1 1
 h4  h5
2 2
3 s 1 1
 h5  h6
2 2

98
 4 4r (1  r  s)

5 4rs
6 4s (1  r  s )

Table 3-2: Sample point for Gauss integration of 3 nodes CCIsoTriangle<...> elements.

Integration point Coordinate r Coordinate s Weight

1 1/3 1/3 1/2
Table 3-3: Sample points for Gauss integration of 3 to 6 nodes CCIsoTriangle<...> elements.

Integration Coordinate r Coordinate s Weight

point

1 1/6 1/6 1/6

2 2/3 1/6 1/6
3 1/6 2/3 1/6
All the above expressions for the formulation for plane quadrilateral elements remain valid also
for the triangular elements, including the extension from 2D to simplified and full 3D analysis.
The expressions only use different approximation functions hi (r , s, t ) and different integration
points [r , s, t ] , see Table 3-1, Table 3-2, and Table 3-3.

3.5 3D Solid Elements

ATENA finite element library includes the following group of 3D solid elements:
tetrahedral elements CCIsoTetra<xxxx> ... CCIsoTetra<xxxxxxxxxx> with 4 to 10 nodes,
see Fig. 3-7,
brick elements CCIsoBrick<xxxxxxxx> ... CCIsoBrick<xxxxxxxxxxxxxxxxxxxx> with 8 up
to 20 nodes see Fig. 3-8 and
wedge elements CCIsoWedge<xxxxxx> ... CCIsoWedge<xxxxxxxxxxxxxxx> with 6 to 15
nodes, see Fig. 3-9.
The string in < > describes present element nodes (see Atena Input File Format document for
more information). These are isoparametric elements integrated by Gauss integration at
integration points given in the following tables. Interpolation functions for all variants of the
elements are also given in the tables below.

ATENA Theory 99
 t
4
z CCIsoTetra<xxxx>
9 .............
10 s
3 CCIsoTetra<xxxxxxxxxx>
8

7 6
1
5 r
2 y
x
t
Fig. 3-7 Geometry of CCIsoTetra<...> elements.
t
3
11 10
4 2
12 9 CCIsoBrick<xxxxxxxx>
1
..................
z20 19
CCIsoBrick<xxxxxxxxx...x>
7 18
15 s

8 17 14
6
r 16
13
5 y
x

Fig. 3-8 Geometry of CCIsoBrick<...> elements.

t
9 3
1
8
7
z 2 15 CCIsoWedge<xxxxxx>
.............
13 14 s
6 CCIsoWedge<xxxxxxxxxx xxxxx>

12 11
4
10 r
5 y
x

Fig. 3-9 Geometry of CCIsoWedge<...> elements.

100
 Table 3.5-1 Interpolation functions of CCIsoTetra<...> elements.

Node Function hi Include only if node i is defined

i
i=5 i=6 i=7 i=8 i=9 i = 10
1 1 r  s  t 1 1 1
 h5  h7  h10
2 2 2
2 r 1 1 1
 h5  h6  h8
2 2 2
3 s 1 1 1
 h6  h7  h9
2 2 2
4 t 1 1 1
 h8  h9  h10
2 2 2
5 4r (1  r  s  t )

6 4rs (1  t )

7 4 s (1  r  s  t )

8 4rt (1  s)

9 4 s t (1- r )

10 4t (1- r - s - t )

Table 3.5-2 Sample point for Gauss integration of 4 nodes CCIsoTetra<...> element.

Integration point Coordinate r Coordinate s Coordinate t Weight

1 1/4 1/4 1/4 1/6

ATENA Theory 101

 Table 3.5-3 Sample points for Gauss integration of 5 to 10 nodes CCIsoTetra<...> elements.

Integration point Coordinate r Coordinate s Coordinate t Weight

1 0.13819660 0.13819660 0.13819660 1/24

2 0.13819660 0.13819660 0.58541020 1/24
3 0.58541020 0.13819660 0.13819660 1/24
4 0.13819660 0.58541020 0.13819660 1/24

Table 3-4 Interpolation functions of CCIsoBrick<...> elements.

Function hi Include only if node i is defined

Node i

i=9 i = 10 i = 11 i = 12 i = 13 i = 14 i = 15 i = 16 i = 17 i = 18 i = 19 i = 20

1 1 1
(1  r )(1  s)(1  t )  h9
1
 h12
1
 h17
8 2 2 2

2 1 (1  r )(1  s)(1  t )  1 h 1 1
 h18
9  h10
8 2 2 2

3 1 (1  r )(1  s)(1  t ) 1 1
 h10  h11
1
 h19
8 2 2 2

4 1 (1  r )(1  s)(1  t ) 1 1
 h11  h12
1
 h20
8 2 2 2

5 1 (1  r )(1  s)(1  t ) 1
 h13
1 1
 h16  h17
8 2 2 2

6 1 (1  r )(1  s)(1  t ) 1 1
 h13  h14
1
 h18
8 2 2 2

7 1 (1  r )(1  s)(1  t ) 1 1
 h14  h15
1
 h19
8 2 2 2

102
 8 1
(1  r )(1  s)(1  t )
1 1
 h15  h16
1
 h20
8 2 2 2

9 1
(1  r 2 )(1  s)(1  t )
4

10 1
(1  r )(1  s 2 )(1  t )
4

11 1
(1  r 2 )(1  s)(1  t )
4

12 1
(1  r )(1  s 2 )(1  t )
4

13 1
(1  r 2 )(1  s)(1  t )
4

14 1
(1  r )(1  s 2 )(1  t )
4

15 1
(1  r 2 )(1  s)(1  t )
4

16 1
(1  r )(1  s 2 )(1  t )
4

17 1
(1  r )(1  s)(1  t 2 )
4

18 1 (1  r )(1  s)(1  t 2 )
8

19 1
(1  r )(1  s)(1  t 2 )
4

20 1
(1  r )(1  s)(1  t 2 )
4

Table 3.5-4 Sample points for Gauss integration of 8 nodes CCIsoBrick<...> element.

Inte- Coordinate r Coordinate s Coordinate t Weight

gration
point
1 0.5773502691896 0.5773502691896 0.577350269189626 1.
26 26
2 0.5773502691896 0.5773502691896 - 1.
26 26 0.577350269189626
3 0.5773502691896 - 0.577350269189626 1.
26 0.5773502691896
26

ATENA Theory 103

 4 0.5773502691896 - - 1.
26 0.5773502691896 0.577350269189626
26
5 - 0.5773502691896 0.577350269189626 1.
0.5773502691896 26
26
6 - 0.5773502691896 - 1.
0.5773502691896 26 0.577350269189626
26
7 - - 0.577350269189626 1.
0.5773502691896 0.5773502691896
26 26
8 - - - 1.
0.5773502691896 0.5773502691896 0.577350269189626
26 26

Table 3.5-5 Sample points for Gauss integration of 9 to 20 nodes CCIsoBrick<...> element.

Inte- Coordinate r Coordinate s Coordinate t Weight

gration
point
1 0.7745966692414 0.7745966692414 0.774596669241483 0.1714677641
83 83
2 0.7745966692414 0.7745966692414 0. 0.2743484225
83 83
3 0.7745966692414 0.7745966692414 - 0.1714677641
83 83 0.774596669241483
4 0.7745966692414 0. 0.774596669241483 0.2743484225
83
5 0.7745966692414 0. 0. 0.4389574760
83
6 0.7745966692414 0. - 0.2743484225
83 0.774596669241483
7 0.7745966692414 - 0.774596669241483 0.1714677641
83 0.7745966692414
83

104
 8 0.7745966692414 - 0. 0.2743484225
83 0.7745966692414
83
10 0. 0.7745966692414 0.774596669241483 0.2743484225
83
11 0. 0.7745966692414 0. 0.4389574760
83
12 0. 0.7745966692414 - 0.2743484225
83 0.774596669241483
13 0. 0. 0.774596669241483 0.4389574760
14 0. 0. 0. 0.7023319616
15 0. 0. - 0.4389574760
0.774596669241483
16 0. - 0.774596669241483 0.2743484225
0.7745966692414
83
17 0. - 0. 0.4389574760
0.7745966692414
83
18 0. - - 0.2743484225
0.7745966692414 0.774596669241483
83
19 - 0.7745966692414 0.774596669241483 0.1714677641
0.7745966692414 83
83
20 - 0.7745966692414 0. 0.2743484225
0.7745966692414 83
83
21 - 0.7745966692414 - 0.1714677641
0.7745966692414 83 0.774596669241483
83
22 - 0. 0.774596669241483 0.2743484225
0.7745966692414
83
23 - 0. 0. 0.4389574760
0.7745966692414
83

ATENA Theory 105

 24 - 0. - 0.2743484225
0.7745966692414 0.774596669241483
83
25 - - 0.774596669241483 0.1714677641
0.7745966692414 0.7745966692414
83 83
26 - - 0. 0.2743484225
0.7745966692414 0.7745966692414
83 83
27 - - - 0.1714677641
0.7745966692414 0.7745966692414 0.774596669241483
83 83

Table 3.5-6 Interpolation functions of CCIsoWedge<...> elements.

hh1  (1  r  s )
hh2  r
hh3  s
hh4  4r (1  r  s )
hh5  4rs
hh6  4 s (1  r  s )
1 t
hv1 
2
1 t
hv2 
2
hv3  (1  t 2 )

106
 Function Include only if node i is defined
Node I

hi i=7 i=8 i=9 i = 10 i = 11 i = 12 i = 13 i = 14 i = 15

1 hh1 hv1 1 1 1
 h7  h9  h13
2 2 2
2 hh2 hv1 1 1 1
 h7  h8  h14
2 2 2
3 hh3 hv1 1 1 1
 h8  h9  h15
2 2 2
4 hh1 hv2 1 1 1
 h10  h12  h13
2 2 2
5 hh2 hv2 1 1 1
 h10  h11  h14
2 2 2
6 hh3 hv2 1 1 1
 h11  h12  h15
2 2 2
7 hh4 hv1

8 hh5 hv1

9 hh6 hv1

10 hh4 hv2

11 hh5 hv2

12 hh6 hv2

13 hh1 hv3

14 hh2 hv3

15 hh3 hv3

Table 3.5-7 Sample points for Gauss integration of 6 nodes CCIsoWedge<...> element.

Integration point Coordinate r Coordinate s Coordinate t Weight

1 1/6 1/6 0.577350269189626 1/6

2 2/3 1/6 0.577350269189626 1/6
3 1/6 2/3 0.577350269189626 1/6

ATENA Theory 107

 4 1/6 1/6 -0.577350269189626 1/6
5 2/3 1/6 -0.577350269189626 1/6
6 1/6 2/3 -0.577350269189626 1/6
Table 3.5-8 Sample points for Gauss integration of 7 to 15 nodes CCIsoWedge<...> element.

Integration point Coordinate r Coordinate s Coordinate t Weight

1 1/6 1/6 0.774596669241483 0.0925925926
2 2/3 1/6 0.774596669241483 0.0925925926
3 1/6 2/3 0.774596669241483 0.0925925926
4 1/6 1/6 0. 0.1481448148
5 2/3 1/6 0. 0.1481448148
6 1/6 2/3 0. 0.1481448148
7 1/6 1/6 -0.774596669241483 0.0925925926
8 2/3 1/6 -0.774596669241483 0.0925925926
9 1/6 2/3 -0.774596669241483 0.0925925926
Formulation of 3D solid elements is given in the following equations:
Incremental strains:
1 (i ) 1 t  t (i 1) ( i ) t  t ( i 1) ( i ) 1 (i ) (i )
t  ij( i ) 
2
 t ui , j  t u j , i  
(i )

2
 t uk , i t uk , j  t u k , j t uk , i  
2
 t uk , i t uk , j  (3.25)

where indices i, j, k  1...3 

Displacement derivatives:

  t  t ui(i ) t  t ui(i 1) 

t
(i )
u
i, j 
t x j
(3.26)
t  t ( i 1)  t  t ui(i 1)
u
t i, j 
t x j

Strains and matrices to calculate them:

108
 t  ( i )  t  tt BL(i 1) U (i )

t  ( i )   t 11(i ) t  22( i ) t  33(i ) 2 t 12(i ) 2 t  23( i ) 2 t 13(i ) 

(3.27)

U (i ) t  t U ( i ) t  t U (i 1) 
u11(i ) u1(2 i ) u31(i ) u12(i ) u22(i ) u32( i ) ... u1n (i ) u2n (i ) u3n ( i ) 

Linear strain-displacement matrix:

t  t
t BL(i 1)  t  t
t BL 0  t  tt BL(i11) (3.28)
Linear strain-displacement matrix – constant part:
 t h1,1 0 0 t h2,1 0 0 ... t hn ,1 0 0 
 0 0 
 t h1,2 0 0 t h2,2 0 ... 0 t hn ,2

t  t
 0 0 t h1,3 0 0 t h2,3 ... 0 0 t hn ,3

t BL 0    (3.29)
 t h1,2 t h1,1 0 t h2,2 t h2,1 0 ... t hn ,2 t hn ,1 0 
 0 t 1,3h t h1,2 0 t h2,3 t h2,2 ... 0 t hn ,3 t hn ,2

 
 t h1,3 0 h
t 1,1 t h2,3 0 t h2,1 ... t hn ,3 0 t hn ,1 

where

hi
t hi , j 
t x j (3.30)

ui(i ) t  t ui(i ) t  t ui(i 1)

Linear strain-displacement matrix – non-constant part:

 t  t ( i 1)
l11 t h1,1 t  t ( i 1)
l
21 t 1,1 h t  t ( i 1)
l 31 t 1,1h t  t ( i 1)
l
11 t 2,1 h
 t  t ( i 1) t  t ( i 1) t  t ( i 1) t  t ( i 1)
 l12 t h1,2 l
22 t 1,2 h l h
32 t 1,2 l
12 t 2,2 h
t  t
 t  t ( i 1)
l h t  t ( i 1)
l h t  t ( i 1)
l h t  t ( i 1)
l h
t BL(i11)   t  t (i 1) 13 t t t1,3 (i 1) t  t ( i 1)
23 t 1,3
t  t ( i 1) t  t ( i 1)
33 t 1, 3
t  t ( i 1) t  t ( i 1)
13 t 2,3
t  t ( i 1)
 l11 t h1,2  l11 t h1,1 l 21 t 1,2 h  l
21 t 1,1 h l h 
31 t 1,2 l h
31 t 1,1 l
11 t 2,2 h  l h
12 t 2,1
 t  t l (i 1) h  t  t l (i 1) h t  t ( i 1)
l h  t  t ( i 1)
l h t  t ( i 1)
l h  t  t ( i 1)
l h t  t ( i 1)
l h  t  t ( i 1)
l h
 t  t 12(i 1)t 1,3 t  t 13(i 1)t 1,2 22 t 1,3
t  t ( i 1)
23 t 1,2
t  t ( i 1)
32 t 1,3
t  t ( i 1)
33 t 1,2
t  t ( i 1)
12 t 2,3
t  t ( i 1)
13 t 2,2
t  t ( i 1)
 l11 t h1,3  l13 t h1,1 l 21 t 1,3 h  l
23 t 1,1 h l h 
31 t 1,3 l h
33 t 1,1 l
11 t 2,3 h  l h
13 t 2,1

... t  t ( i 1)
l
31 t n ,1 h 
t  t ( i 1) 
... l
31 t n ,2 h 
... t  t ( i 1)
l
33 t n ,3 h  (3.31)
t t ( i 1) t t ( i 1) 
... l
31 t n ,2 h  l
32 t n ,1 h 
... t t ( i 1)
l h  t t ( i 1)
l h 
32 t n ,3
t t ( i 1)
33 t n ,2
t t ( i 1)

... l
31 t n ,3 h  l
33 t n ,1 h 

where
n
t  t ( i 1)
ij l   t hk , j t  t k ( i 1)
u
t i (3.32)
k 1

ATENA Theory 109

Nonlinear strain-displacement matrix
 t h1,1 0 0 t h2,1 ... t hn ,1 0 0 
 h 0 0 h2,2 ... hn ,2 0 0 
 t 1,2 t t

 t h1,3 0 0 t h2,3 ... t hn ,3 0 0 

 
 0 t h1,1 0 0 ... 0 t hn ,1 0 
t  t
B ( i 1)
 0 t h1,2 0 0 ... 0 t hn ,2 0  (3.33)
t NL
 
 0 t h1,3 0 0 ... 0 t hn ,3 0 
 0 0 
t h1,1 0 ... 0 0 t hn ,1
 
 0 0 t h1,2 0 ... 0 0 h
t n ,2 
 0 
 0 t h1,3 0 ... 0 0 t hn ,3 

2nd Piola-Kirchhoff stress tensor and vector

 t  tt S11(i 1) t  t ( i 1)
S
t 12
t  t ( i 1)
t 13S 0 0 0 0 0 0 
 t  t (i 1) t  t ( i 1) t  t ( i 1) 
 t S 21 S
t 22 t 23S 0 0 0 0 0 0 
 t  tt S31(i 1) t  t ( i 1)
S
t 32
t  t ( i 1)
t 33S 0 0 0 0 0 0 
 t  t ( i 1) t  t ( i 1) t  t ( i 1) 
 0 0 0 S
t 11 S
t 12 S
t 13 0 0 0 
S ( i 1)
 0 0 0 t  t ( i 1)
S t  t ( i 1)
S t  t ( i 1)
S 0 0 0 
t
 t 21
t  t ( i 1)
t 22
t  t ( i 1)
t 23
t  t ( i 1)

 0 0 0 S
t 31 S
t 32 S
t 33 0 0 0 
 0 0 0 0 0 0 t  t ( i 1)
S t  t ( i 1)
S t  t ( i 1) 
 t 11 t 12 t S13 
t  t ( i 1) t  t ( i 1) t  t ( i 1)
 0 0 0 0 0 0 S
t 21 S
t 22 S
t 23 
 t  t ( i 1) t  t ( i 1) t  t ( i 1) 
 0 0 0 0 0 0 S
t 31 S
t 32 t S33 

S ( i )   t  tt S11( i 1) t  t

t
( i 1)
S 22 t  t
t S33( i 1) t  t ( i 1)
S
t 12
t  t
t
( i 1)
S 23 t  t ( i 1)
S
t 13

(3.34)

3.6 Spring Element

Spring elements in ATENA are used to model spring-like boundary conditions, i.e. situation
where external forces acting on boundary of the structure are linearly proportional to the
associated displacements. Three elements of this type are available, see also Fig. 3-10, Fig. 3-11:
CCSpring – 2D and 3D element to model spring-like boundary conditions at a point,
CCLineSpring – 2D element to model spring-like boundary conditions along a line
CCPlaneSpring – 3D element to model spring-like boundary conditions along a triangular area.
All these elements are derived from 2D or 3D formulation of the CCIsoTruss<xx> element
described earlier in this chapter. For example, CCSpring element consists of one
CCIsoTruss<xx> element. The 1st node of each CCIsoTruss<xx> coincides with one node of the
CCSpring element, whereas the 2nd node of the CCIsoTruss<xx> is set by direction vector, see
Fig. 4-4. Note that as the analysis is nonlinear, length of the direction does matter. This vector is
specified in ATENA &SPRING_GEOMETRY_SPEC command and is common for all spring
elements that use this geometry.

110
CCLineSpring and CCPlaneSpring elements were created to enable convenient definition of
„uniform“ spring-like conditions along the boundaries. The boundary force at a node i of the
spring element is calculated:
ui kA
Ri  (3.35)
n direction
where
k is spring material stiffness parameter set by &MATERIAL SPRING command,
(parameter k has character of multi-linear Young modulus),
ui is displacement at spring element node i ,
A is the area of CCPlaneSpring element or length of CCLineSpring multiplied by
thickness (which defaults to 1 if not specified in element geometry) or the area defined in
element geometry for CCSpring (similarly, with a default of 1 if not specified) for the
respective element,
n is number element nodes, i.e. 1, 2 or 3 for CCSpring, CCLineSpring or CCPlaneSpring
element respectively,
direction is Euclidean norm (i.e., length) of the direction vector, see above.

CCSpring CCLineSpring

Fig. 3-10 Geometry of 2D CCSpring and CCIsoLineSpring.

ATENA Theory 111

 z

CCPlaneSpring

CCSpring

area A

Fig. 3-11 Geometry of 3D CCSpring and CCPlaneSpring.

3.7 Quadrilateral Element Q10

3.7.1 Element Stiffness Matrix

The quadrilateral finite element Q-10 is derived from a six-node triangle (CCQ10<xxxx>,
CCQ10Sbeta<xxxx>). The derivation of the stiffness matrix is taken from FELIPPA 1966. The
position of any internal point P in the element is defined by the triangular coordinates (called
also natural coordinates). These coordinates are expressed by means of areas within the triangle
as shown in Fig. 3-12. Sub-areas Ai are subtended by the point P and two corners. A is the area
of triangular element.

Fig. 3-12 Coordinate systems of the six-node triangular element.

A1 A A
1  , 2  2 , 3  3 (3.36)
A A A
1   2   3  1

112
Using the quadratic interpolation function, the displacement components u(  i ), v(  i ) is written
in the terms of triangular coordinates  i and nodal displacement vectors :

u ( i )  F( i )T u, v( i )  F( i )T v (3.37)

The displacement vectors u, v contain six components of the nodal displacements and the vector
F ( i ) contains the quadratic interpolation functions in triangular coordinates:

u   u1  , v   v1 
T T
u2 u3 u4 u5 u6 v2 v3 v4 v5 v6 (3.38)

F( i )   1 (2 1 1)  2 (2 2  1)  3 (2 3  1) 4 1 2 4 3 1

T
4 2 3 (3.39)
A general procedure to construct the element stiffness matrix is described by the set of following
equations:
(a) The constitutive equation:
sD e (3.40)
(b) The strain-displacement equations in the Cartesian coordinates:
u  x, y  v  x, y  u  x, y  v  x, y 
x  , y  ,    (3.41)
x y y x
which is written in terms of the natural coordinates  i and the nodal displacements vectors u, v:

u 
 ( i )  F T   (3.42)
v
The stiffness matrix:
K   F T D F dV (3.43)
V

The matrix F contains partial derivatives of the interpolation function F and the integral in the
last equation is made over the element volume V. The details of the derivation can be found in
FELIPPA 1966 and here only the final matrix equations are presented.

Fig. 3-13 Quadrilateral element (b) composed from two triangular elements (a).
The quadrilateral finite element is composed from two 4-node triangular elements, as shown in
Fig. 3-13. Two degrees of freedom in a node are the horizontal and vertical displacements. The
triangular element is derived from the 6-node triangle by imposing kinematic constraints on two
mid-side nodes. The resulting strain-displacement matrix relation for the 4-node triangle is:
ATENA Theory 113
 e x   U O 
e  Bd
    u 
e y   O V    (3.44)
 g  V U   v 
   
where ex, ey are the normal strain vectors, g is the shear strain vector (engineering type) and O is
the null matrix. The strain and displacement vectors contain nodal components:

e x   x1  x 2  x 3 , e y   y1  y 2  y 3 , g   x1  x 2  x 3
T T T
(3.45)

u   u1  , v   v1 
T T
u2 u3 u4 v2 v3 v4 (3.46)

The strain interpolation function in the element is linear and is uniquely specified by three nodal
values in the corners of the triangular element, while the displacement interpolation function is
quadratic and is specified by three corners and one mid-side nodal displacement. The
components ui, vi are the horizontal and vertical displacements, respectively, in the node i. The
indexes 1, 2 and 3 denote the corner nodes of a sub-triangle and the index 4 is for the mid-side
node, see Fig. 3-13 (a). The strain-displacement sub-matrices in (3.44) are
3b1  2b3 b2 b3 4b2 
1 
U b1 3b2  b3 b3 4b1 
2S 
 b1 b2 b3 . 

3a1  2a3 a2 a3 4a2 

1 
V  a1 3a2  a3 a3 4a1  (3.47)
2S 
 a1 a2 a3 . 

a1  x3  x2 b1  y2  y3
a2  x1  x3 b2  y3  y1
a3  x2  x1 b3  y1  y2
2S  a3 b2  a2 b3
where xi, yi are the global Cartesian coordinates of the node i in a sub-triangle, S is the area of the
sub-triangle.
The element stiffness matrix for the 4-node sub-triangle is
K K uv 
K   uu (3.48)
 K vu K vv 

The stiffness matrix K has an order 8 and is so partitioned that the upper four rows correspond to
the horizontal displacement components (index u) and the lower four rows correspond to the
vertical displacement components (index v). The integration of the stiffness coefficients is made
exactly, and the resulting sub-matrices are:
K uu  St  d11 A  d13 (H  HT )  d33 C 

K vv  St  d 22 C  d 23 (H  HT )  d33 A 

K uv  St  d12 H  d13 A  d 23 C  d33 HT  (3.49)

114
where t is the thickness of the element, dij are the coefficients of the material stiffness matrix D,
(3.40). The integration in (3.43) is done explicitly by the following matrix multiplication:
A  UT QU, H  UT QV, C  VT QV (3.50)
Where the area integration matrix Q is:
2 1 1
1 
Q  1 2 1  (3.51)
12
1 1 2 

The element stiffness matrix of the 5-node quadrilateral, Fig. 3-13(b), is composed of the two 4-
node sub-triangles by summing the stiffness coefficients of the appropriate nodes. The resulting
matrix of the 5-node quadrilateral K10 has the order 10. The coefficients of the matrix can be
rearranged according to the external (index e) and internal (index i) degrees of freedom:
K K ei 
K10   ee (3.52)
 K ie K ii 

The sub-matrices corresponding to two internal degrees of freedom are eliminated by

condensation procedure and the final element stiffness matrix K of the order 8 is obtained:
K  K ee  K ei K ii 1 K ie (3.53)

Fig. 3-14 Subdivision of quadrilateral element.

The subdivision of the quadrilateral element into the triangular elements must be done in an
optimal way and it is preformed automatically by the program. The examples of the subdivisions
are illustrated by Fig. 3-14. Due to this method of the subdivision, a concave form of the
quadrilateral element is acceptable. This element form could not be achieved by an isoparametric
element.

3.7.2 Evaluation of Stresses and Resisting Forces

For the given displacement field, the strains and stresses are evaluated in the center of the
quadrilateral element. The stresses at this point are obtained from material laws as functions of
strains according to Section 2.1.12. Also, the constitutive law for the element and the matrix D
are calculated from the stresses and strains at the center of the element. These stresses and strains
are written in the output file as a part of the results.
The calculation of resisting nodal forces of the sub-triangle for a current displacement field and a
constitutive law is done by the following equation:
R  t BT Q 9 s 9 (3.54)
where R is the vector of nodal forces (same arrangement and numbering as in the vector d in
(3.44)). The matrix Q9 contains three integration matrices Q in the diagonal. The stress vector s9

ATENA Theory 115

(same numbering as the vector e, (3.40), is calculated from the current strains and secant material
matrix, Section 2.1.12.
There are two variations of this element in program ATENA: CCQ10<xxxx> and
CCQ10Sbeta<xxxx>. The main difference between these two elements lies in the way how the
resisting forces are calculated. In case CCQ10<xxxx>, they are computed as described by
Equation (3.54). In the second case, however, the material law is evaluated only at the element
centroid. Based on the current state of damage a secant constitutive matrix is calculated and it is
used to determine the integration point stresses and resulting resisting forces. This element type
is almost identical to the element that was implemented in the program SBETA, i.e. the former
version of this program. Due to this approach, there are some limitations for usage of this
element with respect to some material models. It can be only used with material models that are
able to calculate and exact secant constitutive matrix. This means that only the following
material models can be used with the element CCQ10Sbeta<xxxx>: CCElastIsotropic and
CCSbetaMaterial.

116
3.8 External Cable
External pre-stressing cables are reinforcing bars, which are not connected with the most of the
concrete body, except of limited number of points, so called deviators, as shown in Fig. 3-15.
This element type is denoted in ATENA as CCExternalCable.

Fig. 3-15 External cable model.

Each cable has two ends provided with anchors. The anchor, where the pre-stressing force is
applied is denoted as the active anchor, the anchor on the other side is the passive anchor. The
points between the anchors are called deviators (or links). After applying pre-stressing the cable
is fixed at anchors. In the deviators, cable can slide while its movements and the forces are
governed by the law of dry friction. The slips of the cable in the deviators (the relative
displacement of the cable ends with respect to the deviators) are denoted as 1 ,  2 ... They are
introduced as variables to be determined by the analysis.

Fig. 3-16 Forces at the deviator.

The forces, F1 and F2 acting on a deviator i are the cable forces at the adjacent cable sections,
Fig. 3-16. Their difference Pi = F1 -F2, (F1> F2) is the loss of the pre-stressing force due to
friction in the deviator i. The relation between these forces according to the law of friction is
expressed as:

 
F2  F1 e i p  Q f ( ) f r (r ) (3.55)

ATENA Theory 117

In this expression  i is the angular change of the cable direction at the deviator i, R is the radius
of the deviator, (i.e. the product R i is the length the contact between the cable and the
deviator.).  is the friction coefficient. The constant part of the friction is Qi  p c f R  i , where
c f is the cohesion (a constant part of the friction) of the cable per unit length and unit perimeter.
p stands for reinforcement bar perimeter. If the constant part of friction is neglected, the term Q
is zero. f ( ) , f r (r ) are user defined function that enable change of deviator's properties
depending on value of slip  and deviator position coordinate r (measured from its starting
point). By default, these functions are set to one.
Introducing dia  e i p f ( ) f r (r ) and dib  p c f R  i f ( ) f r (r ) we can simplify (3.55) to

F2  F1 d1a  d1b (3.56)

Fig. 3-17 Forces and displacements in the cable element (cable section).
A section of the cable between the deviators is considered as the uniaxial bar element, Fig. 3-17.
The force F in the cable element depends on the pre-stressing force P, the displacements of ends
u1, u2 due to structural deformation and the cable slips 1 ,  2 in the deviators. The slips  are
introduced as an additional variable for the external cables. The equilibrium equation of the
cable section is:
F  P  K (u2  u1   2  1 ) (3.57)
The element stiffness K = Es A/L, where A, L are the cable’s cross section and length,
respectively, and Es is the actual secant or tangent modulus derived in the same way as in case of
other reinforcement using bilinear or multi-linear law.
The cable forces F1, F2, … are determined by applying the above equations for all cable
deviators, i.e. an iterative solution is executed for displacements u, (outer iterations loop), and on
slips  i , (inner iteration loop).
Introduction of pre-stressing is accomplished by applying an initial slip (cable pull-out) at the
anchor end until a prescribed pre-stressing force is reached. This procedure reflects a real
process of pre-stressing and considers the loss of pre-stressing due to friction deviators and
deformation of the structure.

3.9 Reinforcement Bars with Prescribed Bond

Reinforcement bars with prescribed bonds are an extension of the external cables described in
the previous section. The main difference is that they can also account for a bond between the

118
bar and the surrounding concrete body. This connection need not be perfect, because the
cohesion strength has a limited value. It is inputted in form of a “bond” cohesion stress.
This type of element is denoted as CCBarWithBond in ATENA. A typical reinforcement bar of
this type is depicted in the figure below. The detail shows undeformed and deformed shape of a
segment of the bar. The original length l0 will change to l due to displacement u of the
surrounding body and bar slips  .

c c
1 2 i-1 i m

1 2 i-1 i i+1 m+1

undeformed truss i

deformed truss i

u  
i i i+1
u
i+1
lo
l

Fig. 3-18 Reinforcement bar with slips.

Normal stress at element i is calculated by:

(ui  i   i 1  ui   i )
i  E (3.58)
li

ATENA Theory 119

  x
Its derivative is compared with the total cohesion stress  c , i.e.
  c . If the cohesion stress
x
between the bar and the surrounding concrete is to be exceeded, the bar will slip to reduce this
stress. Otherwise, the slips  will remain unchanged (or initially equal to zero), which
correspond to the case of perfect bond.
The total cohesion stress consists of two parts: base cohesion stress and so-called wobble
cohesion stress, i.e. an extra cohesion dependent on axial stress in the bar, (see the term  x f w
below). The wobble cohesion is derived as follows: Prestress losses are calculated by:
 r   p (1  e    r )
 r   p     p e   r
 r (3.59)
    p e    r     r
r
 r
    p e    r  f w  r
r

The wobble related cohesion stress is thus     r  f w r .

Realizing that the cohesion stress can be constant, or it can be defined as a function of  and r ,
we can calculate the total cohesion stress  c as follows:

 c  f r (r ) fT (T ) f c (ccorr )  c 0 f ( )   x f w 
corr
(3.60)

f ( ) , f r (r ) are the same as those described for external cables near (3.55),  c 0 is reference
base cohesion stress due to slipping (to be inputted),  c is total cohesion stress due to slipping
and wobble cohesion, p is perimeter of the reinforcement bar, r is location at the bar.  x is
normal stress in the bar and f w states for wobble coefficient. The remaining parameters are:  r
is axial normal stress in the bar in direction of local coordinate axis r (in direction of the bar) and
p, A means perimeter and cross-sectional area of the bar, (again similar to r in the case of
external cable). Function f ccorr (ccorr ) and fT (T ) expresses, how the cable’s cohesion depends on
current temperature and corrosion ratio Acurr / Aorig at a point of the cable. Acurr , Aorig is current
and original (i.e. before corrosion started) area of cross section of the cable.

Fig. 3-19 Forces at node i.

120
The discretized solution equation for node i, (considering elements i  1, i ), reads (the bars are of
constant strain type):

BAR WITH PRESCRIBED BOND:

Fi L  Ai 1  i 1
Fi R  Ai  i
(3.61)
l l   
for ( Fi  Fi ) :
R L
Fi  Fi  i i 1 p   c  c  i 
R L

2   i 
l l    
for ( Fi R  Fi L ) : Fi R  Fi L    i i 1 p   c  c  i  
 2  i 
   

If this element acts as the external cable, see the previous section, then

EXTERNAL CABLE WITH DEVIATORS:

for ( Fi R  Fi L ) : Fi L  Fi R dia  dib

Fi R  Fi L  Fi R  Fi R dia  dib

  
Fi R  Fi L  Fi R 1  dia  dib )  (3.62)

for ( Fi R  Fi L ) : Fi R  Fi L dia  dib

Fi R  Fi L   Fi L  Fi L dia  dib

  
Fi R  Fi L   Fi L 1  dia  dib 

Assembling (3.61) and (3.62) yields final (in)equations for force difference at node i:

EXTERNAL CABLE WITH DEVIATORS AND PRESCRIBED BOND:

li  li 1 l  l  c
for ( Fi R  Fi L ) : 
Fi R  Fi L  Fi R 1  dia  dib  2
p  c  p i i 1
2  i
 ik (3.63)

 l l l  l  c 
for ( Fi R  Fi L ) :  
Fi R  Fi L    Fi L 1  dia  dib  i i 1 p  c  p i i 1
2 2  i
 ik 
 

ATENA Theory 121

Note that at this stage we solve for slips  , (while keeping constant cable displacements u ). As
the reference cohesion stress is a function of  , i.e.  c 0   co ( ...) , in the above equations we

use its Taylor approximation  c  c  i
 i
The above set of (in)equations is calculated in iterative manner. Assume we know the forces at
iteration (k  1) , then the forces at iteration k are:
EAi
Fi R , k 1  (u i1  u i   ik11   ik 1 )
li
EAi 1
Fi L , k 1  (u i  u i1   ik 1   ik11 )
li 1

EAi
Fi R , k  Fi R , k 1  ( ik1   ik )
li
EAi 1
Fi L , k  Fi L , k 1  ( ik   ik1 ) (3.64)
li 1

 k   k 1   k
i 1 i 1 i 1

 
k
i i
k 1
  i k

 k   k 1   k
i 1 i 1 i 1

and

122
 for ( Fi R  Fi L ) :
EAi EA l l 
Fi R  ( ik1   ik )  Fi L  i 1 ( ik   ik1 )  i i 1 p c  ik 
li li 1 2  i
li  li 1
 
Fi R 1  dia  dib 
2
p c

 EAi 1  k  EAi 1 EAi  k  EAi  k li  li 1  c

   i1      i      i1  p  ik 
 li 1   li 1 li   li  2  i
 l l 
   
Fi R  Fi L   Fi R 1  dia  dia  i i 1 p  c 
 2 

for ( Fi R  Fi L ) :
EAi EA l l 
Fi R  ( ik1   ik )  Fi L  i 1 ( ik   ik1 )  i i 1 p c  ik 
li li 1 2  i
 l l 

 
  Fi L 1  dia  dib  i i 1 p  c 
2 
 EAi 1   EAi EAi 1  k  EAi  l l 
   i1      i     ik1   i i 1 p c  ik 
k

 li 1   li li 1   li  2  i
 l l 
   
Fi R  Fi L   Fi L 1  dia  dib  i i 1 p  c 
 2 
(3.65)

If the above equation is written for all nodes on the bar, we obtain a set of inequalities. It has to
be solved in iterative manner (within each iteration of the main solution loop).

Atena also support so called CCBarWithMemoryBond 2D and 3D elements. They differ from
their original formulation, (i.e. elements CCBarWithBond), in that they have different function
f ( ) for "loading“ and f ,unload ( ) for "unloading" regime. This means   ( min ,  max ) in the
former and   ( min ,  max ) in the latter case.

ATENA Theory 123

 Fig. 3-20 Bond function for CCBarWithMemoryBond element

To obtain more realistic shape, the resulting cohesion stresses are prior their output smoothed.
The smoothing operation for node i is expressed as follows:
 i 1li 1   ili
 right 
li 1  li

 ili   i 1li 1
 left  (3.66)
li  li 1

( right   left ) A

c 
pli
The equation (3.58) together with (3.61) completes the element description. The element can be
used to realistically model cohesion between reinforcement bar and concrete. Such a model is
needed for analysis of pullout tests etc. Although the adopted solution is simple, it provides
reasonable results accuracy at low computation cost. A more elaborate model of cohesion
between reinforcement bar and surrounding concrete can be achieved by using special interface
elements that is described in the next section.

3.10 Interface Element

The interface elements are used to model a contact between two surfaces. Currently, the
following element types are available: CCIsoCCIsoGap<xxxx> and CCIsoGap<xxxxxx>,
CCIsoGap<xxxxxxxx> for 2D and 3D analysis, respectively. These elements use linear
approximation of geometry. For the case of nonlinear geometry, use element type

124
CCIsoGap<xxxxxx> for 2D and CCIsoGap<xxxxxxxxxxxx> or
CCIsoGap<xxxxxxxxxxxxxxxx> for 3D. The string in < > describes present element nodes, (see
Atena Input File Format document for more information). The elements are derived from the
corresponding isoparametric elements (described in sections 3.3 and 3.4), i.e. they use the same
geometry and nodal ids etc. Geometry of the supported gap elements is depicted in Fig. 3-21.

Linear geometry Nonlinear geometry

v 1
2D v 1
r,u(r) u(r)
r,
2 2 5
4 4
6
3 3
CCIsoGap<xxxx> CCIsoGap<xxxxx_x>

w 3D 10 w

1 1
2 7
2 9
r,u(r,s) r,u(r,s) 8
4 3 4
3
5 5 12
11
6 6

s,v(r,s) s,v(r,s)
CCIsoGap<xxxxxx> CCIsoGap<xxxxxxxxxxxx>

13 w
w 1
1 9
2 2
12 16
5 5
4 r,u(r,s) 6 10 4
r,u(r,s) 6 11
3 3
14
8 8
7 15
7
s,v(r,s) s,v(r,s)
CCIsoGap<xxxxxxxx>
CCIsoGap<xxxxxxxxxxxxxxxx>

Fig. 3-21 CCIsoGap elements

The interface is defined by a pair of lines, (or surfaces in 3D) each located on the opposite side
of interface. In the original (i.e. undeformed) geometry, the interface lines/surfaces can share the
same position, or they can be separated by a small distance. In this case we speak about the
interface with nonzero thickness.
In the following, the interface behavior is explained on a simple 2-dimensional case, see section
2.6 for a full description of the interface material.
The interface element has two states:
 Open state: There is no interaction of the contact sides.
 Closed state: There is full interaction of the contact sides. In addition, friction sliding of
the interface is possible in case of interface element with a friction model.

ATENA Theory 125

Penalty method is employed to model the above behavior of the interface. For this purpose, we
define a constitutive matrix of the interface in the form:
 F  K 0   u 
F       tt    Du (3.67)
 F   0 K nn   v 

in which u, v are the relative displacements of the interface sides (sliding and opening
displacements of the interface) in the local coordinate system r , s and K tt , K nn are the shear and
normal stiffness, respectively. This coefficient can be regarded as stiffness of one material layer
(real, or fictious) having a finite thickness. The layer is only a numerical tool to handle the gap
opening and closing. F , F are forces at the interface, (again at the local coordinate system).
The actual derivation of gap elements is now demonstrated for the case of linear 2D gap element
CCIsoGap<xxxx>, see Fig. 3-21. The other elements are constructed in a similar way.
The element has two degrees of freedom defined in the local coordinate system, which is aligned
with the gap direction. They are relative displacements v, u and are defined as follows:
1 1
h1  (1  r ), h2  (1  r )
2 2

 u   h1u1,4  h2 u2,3 
u      
 v   h1v1,4  h2 v2,3 
 u1 
v 
 1
u2 
 
 h 0 h2 0 h2 0  h1 0   v2 
u   1  Bu
 0 h1 0 h2 0  h2 0 h1  u3 
 
 v3  (3.68)
u 
 4
 v4 

The rest of the element derivation is the same as in case of any other elements, i.e. the stiffness
matrix K   ΒT DBdV , vector of internal forces Q   ΒT FdV etc. A numerical integration in
two Gauss points is used to integrate the interface element stiffness matrix. The matrix K and the
vector Q are in local coordinate system and therefore before they are assembled in the problem
governing equations, they must be transformer in global coordinates.
The stiffness coefficients depend on the gap state. The interface is considered open, if the normal
force F >Rti (Rti is the interface tensile strength force) and the corresponding constitutive law is
(stress free interface):
 F  0 
   (3.69)
 F  0 

126
The stiffness coefficients are set to small, but nonzero values K ttop , K nn
op
.
The interface element is considered closed if F  Rti. The stiffness coefficients are set to large
values K ttcl , K nn
cl
. It should be noted that the stiffness coefficients are defined only for the purpose
of the numerical iterative solution. (Hint: The values of coefficients in the closed state (the large
values) are based on thickness comparable to the size of neighbor quadrilateral elements. The
minimum values in the open state can be about 1000 times smaller. )
The interface thickness in the out-of plane direction is normally provided as an input parameter.
In the case of axi-symmetric analysis it is however calculated using the formula:
t  2 x (3.70)
where x is the distance from the axis of symmetry.
There are two special options for processing the gap elements:
Initial gap opening
It is possible to "open" gap at a particular load step, typically the first step of the analysis, i.e. we
can introduce to the gaps something like initial element strains in case of ordinary finite
elements. This is achieved by LOAD INITIAL GAP ... INIT_STEP_ID step_id command. Upon
that, during calculation of the (gap) element at the step step_id an artificial opening of the
interface is introduced. Its value is the distance between upper and lower element surfaces/lines
(with reference to undeformed structural shape).
The GAP element load is typically used as follows: we have a structure with a base and upper
block. The upper block falls towards the base block that is typically fixed. The structure is solved
by introducing a layer of gap elements between the base and upper blocks and applying the GAP
element load (for these gaps elements) in the 1st step. As a result, in the first steps the gaps will
open to the distance between the blocks. It involves some tensional forces, but as the interface
material usually sustains only compression forces, they can be neglected. In next steps the upper
block gradually is falling to the base block until it hits it. At this moment interface gaps get fully
closed, they change their regime form tension to compression and the upper block gets fully
supported by the base block.
Moving gaps 3
Suppose we have a structure has a base block and an upper block sitting on the base block. The
base block is fixed, the upper block is dragged on the upper surface of the base block. The blocks
are not mutually interconnected, only some friction and cohesion forces exist between them.
Such problems can be modelled by the RESET_DISPLS n flag for the CC2DInterface /
CC3DInterface. If this flag is input, then the upper and bottom surface/lines for all corresponding
elements are realigned at the end of each step as shown for 2D elements in the following picture.
The 3D gaps element is realigned in the same way.
Of course, the boundary surface/lines projection of the gap interface (and thus its "moving" can
be used in more complex situation, but the essence of the described technique remains the same.
The layer of interface elements is typically connected to the bottom/ upper block of structure by
MASTER SLAVE NODAL LISTS boundary conditions, where we must not forget to use the
flag PROCESS_FLAG USE_CURRENT_COORDS. It will assure that after realigning the
interface gets properly connected to the rest of the (deformed) structure.

3
Available starting from ATENA version 4.3.1.

ATENA Theory 127

 Fig. 3-22 Moving gap 2D element
Note that the option of the gap's initial opening and the reset displacements flag can be
combined. Both these special processing options are possible, because the ATENA software uses
incremental approach to solve the structure. Thus, changing shape of the gap (at the end of the
steps) will not harm governing equilibrium equations.

3.11 Truss Axi-Symmetric Elements.

In the following a circumferential truss element for axisymmetric analysis are described. The
elements call CCCircumferentialTruss and CCCircumferentialTruss2 and they are aimed mainly
for modeling structural circumferential reinforcement. For radial reinforcement refer to
CCIsoTruss<xx> and CCIsoTruss<xxx> elements.
The CCCircumferentialTruss has one node only, whereas the CCCircumferentialTruss2 has
nodes two. They behave much the same, the difference being only in calculation of their “cross-
sectional area”. In case of the CCCircumferentialTruss element the area is entered directly from
input data. The CCCircumferentialTruss2 element calculate the area as its thickness (defined in
its geometry data) multiplied by its length. Unlike isoparametric elements thses elements are
derived and computed analytically.
Geometry, interpolation functions and integration points of the elements are given in
Fig. 3-23.

y
y
2
1
1

CCCircumferentialTruss CCCircumferentialTruss2

x x

Fig. 3-23 Geometry of CCCircumferentialTruss and CCCircumferentialTruss2 elements.

128
In the following structural vectors and matrices for the CCIsoTruss element are derived.
Development of the CCIsoTruss2 is much the same. In fact, it is CCIsoTruss acting at the centre-
point of the CCIsoTruss2 element with its cross-sectional area calculated as explained above.
The element vectors and matrices for Total Lagrangian formulation (TL), configuration at time t
and iteration (i) are as follows. Note that they are equally applicable for Updated Lagrangian
formulation (UL) upon applying changes related to the element reference co-ordinate system
(undeformed vs. deformed element axis.).
The truss element center has at reference time t and t  t ( i 1) co-ordinates t X  [t x1 , t x1 ] and
t  t ( i 1) ( i 1) ( i 1)
X  [t  t x1 , t  t x1 ] , respectively. The element length (at respective time) is its length is
t
l  2 t x1 and t  t ( i 1)
l  2 t (t x11  t u11( i 1) ) .

Increment of Green Lagrange strain t 11( i )  

t  t ( i ) t  t
t 11  t  tt11(i 1) (at time t  t , iteration (i )
with
to configuration at time t ) is calculated:

1 l   l  
 t  t (i ) 2 t  t ( i 1) 2 

(i )
 (3.71)
2 tl 
t 11 2

 
where truss length t  t ( i )
l  2 t (t x11  t u11(i 1)  t u11( i ) ) . Note that t u11(i ) is co-ordinate increment
(i ) ( i 1)
(t  t x1 t  t x1 ) . Substituting expressions for element length into (3.71) yields:

(i )

4 2  x  u
t 1
1
t 1( i 1)
1  t u11(i )    t x11  t u11(i 1) 
2 2

 x
t 11
t 1 2
1
(3.72)
2
u1( i )
u u t 1( i 1) 1( i )
1 u  1( i )
t 1
 1 t 1
  
t 1

 x1  t 1 2
t 1 t 1
x 1 2 x  1

Separating t u11(i ) from (3.72) and rearranging in matrix form we obtain:

t  t 1
t BL 0  t
(3.73)
x11

t  t
t
u11(i 1)
BL(i11)  (3.74)
 x
t
t 1 2
1

and
t  t ( n 1) 1
BNL  (3.75)
 x11 
t 2
t

The 2nd Piola-Kirchhoff stress matrix and tensor are:

t  t
t S (i 1)  t  t
t S (i 1)  [ t  tt S11( i 1) ] (3.76)

ATENA Theory 129

 t  t (i )
The formulation is completed by relationship for element deformation gradient t X 1,1 , which
yields:

 x11  t u11( i ) 
t 2
t  t
t X (i )
1,1 1 e 
(i )
t 11 t
(3.77)
x11

where engineering strain t e11(i ) is calculated by

 2 t 1 2   
 4  x1  u1   4  x1     x11  t u11(i )  t x11 
2 t 1 t 1( i ) 2 t 2
t  t ( i )
l
e 
(i )
t 11    (3.78)
4 2  t x11 
t t 1
l 2 x1

3.12 Ahmad Shell Element

This section describes Ahmad shell element implemented in ATENA, see Fig. 3-27. It can be
used to model thin as well as thick shell or plate structures. It accounts for both plane and
bending structural stiffness. The element features quadratic geometry and displacement
approximation and therefore, the element’s shape can be non-planar. It is possible to account for
structural curvatures. Big advantage of this element is that it is seamlessly connectible to true 3D
ATENA elements.
Three modifications of this element are supported, and these are characterized by Lagrangian,
Serendipity and Heterosis variant of geometry and displacement field approximation. To avoid
or minimize membrane and shear locking of shell element it is further possible to use full
integration scheme, as well as reduced and/or selective integration. The problems concerned with
combination of displacement approximation and integration scheme with respect to locking
phenomena are discussed.
The element is derived in a way similar as the other finite elements, which are described in this
manual. Hence, in the present description will concentrate mainly on features that are specific for
this element. Following Total Lagrangian formulation of the problem, the principle of virtual
displacement is used to assemble incremental form of governing equations of structure.
The present Ahmad element belongs to group of shell element formulation that is based on 3D
elements’ concept. Nevertheless, it uses some assumptions and restrictions, so that the originally
3D element is transformed into 2D space only. It saves computational time and it also avoids
some formulation difficulties pertaining to 3D elements.
The element’s in-plane integration is carried out in usual way by Gauss integration scheme,
whilst in the 3rd dimension (i.e. perpendicular to mid surface of element) the integration can be
done in closed (analytical) form. However, in order to enable accounting for nonlinearity of
constitutive equations, the so-called layer concept is used instead. Hence, in the 3rd dimension
simple quadrilateral integration is employed.
The present degenerate continuum element was originally proposed by Ahmad et al. (Ahmad,
Irons et al. 1970). Following general shell element theory concept, every node of element has
five degree of freedom, e.g. three displacements and two rotations in planes normal to mid-
surface of element. In order to facilitate a simple connection of this element with other true 3D
elements, the (original) five degrees of freedom are transformed into x,y,z displacement of a top
node and x,y displacement of a bottom node degrees of freedom. The two nodes are located on
the normal to mid-surface passing thru the original mid-surface element’s node, see Fig. 3-28.

130
The essential point in the element’s derivation is that displacements and rotations fields are
approximated "independently", (see e.g. (Jendele 1981), where similar approach is used for
plates). This means that they are handled separately. Unlike in true Mindlin theory our
formulation matches geometric equations automatically. However, a special technique is used to
improve the element’s shear behavior (Hinton and Owen 1984).
The first formulation of this element proposed by Ahmad was linear but since that time many
improvements have been achieved. The most important is the application of reduced or selective
integration scheme that reduces or totally removes locking of the element. Also, many authors
extended the original formulation to geometrically and later also materially nonlinear analysis.
One such an advanced form of the element is the formulation implemented in ATENA.
On input, the Ahmad element uses the same geometry as 20 nodes isoparametric brick element,
i.e. CCIsoBrick<xxxxxxxxxxxxxxxxxxxx>, see Fig. 3-27. This is needed, in order to be able to
use the same pre- and postprocessors’ support for the shell and native 3D brick (i.e. hexahedron)
elements. After the 1st step of the analysis, the input geometry will automatically change to the
external geometry from Fig. 3-27. As nodes 17 and 18 contain only so-called bubble function,
the element is post-processed in the same way is it would be the element
CCIsoBrick<xxxxxxxxxxxxxxxx>. Internally, all element’s vectors and matrices are derived
based on the internal geometry as depicted also in Fig. 3-27.
With shell elements, the best connection at edges is to cut both at 45 degrees, or a different
corresponding angle if the thicknesses are not the same, or if connected at other than right angle,
see Fig. 3-24 (a). Another option is to use a volume brick element at the corner, which is the only
feasible way when more than two shells are connected, see Fig. 3-24 (b). The nodes on the
surface connected to the volume element have to be listed in the INTERFACE subcommand in
the shell geometry definition for correct behavior. Connecting like in Fig. 3-25 is not
recommended, as the master-slave relations induced by the fixed thickness of the shell may
cause numerical problems.

Shell1

Brick Shell2

Shell3
(a) (b)
Fig. 3-24: Ahmad Shell - recommended connection (a) 2 shells (b) 3 shells

ATENA Theory 131

 Fig. 3-25: Ahmad Shell - not recommended connection

3.12.1 Coordinate Systems.

The essential point in the element’s derivation is to understand coordinate systems that are used
within the derivation. These are as follows. Note that all vectors indicating coordinate systems’
axes are normalized. Thus, any directional cosines are simply computed as scalar products that
need not be divided by the vectors’ norm.
Global coordinate system.
It is used to define the whole FE model. Global coordinates are denoted by t x1 , t x2 , t x3 , where
the index t referrers to time. Note that we are using Modified Lagrangian formulation, in which
model configuration is updated after each time step, while within one step (for iterating) the
configuration from the step beginning is employed. Thus, 0 x1 , 0 x2 , 0 x3 are a point global
coordinates prior any load has been applied.
Nodal coordinate system
This coordinate system is defined at each point of element mid-plane surface, i.e. mid-nodes 1-9.
t t t
At a node k it is specified by vectors V1k , V2k , V3k , see Fig. 3-26.

132
 Fig. 3-26 Ahmad element coordinate systems
t t
The vectors V1k , V2k , V3k are defined as follows: Firstly, two auxiliary vectors V1 , V3 are
t t t

t
calculated. Vector V3 at a point is defined as a line joining bottom and top coordinates at the
t
node k (prior any deformations, i.e. at reference configuration). The second vector V1 is normal
t
to V3 and is parallel to plane of global X 1G and
0 0
X 3G . Hence:

V3   V3 , V3 , V3 

t t t t

 1 2 3
(3.79)
V1   V1 , V1 , V1    V3 ,0,  V3 
t t t t t t

 1 2 3  3 1

t t
If V3 is parallel to X 2G (i.e. V3  V3  0 ), V1 is defined by
0 t t

1 3

V1    V3 ,0,0 

t t
(3.80)
 2 
t t t t
After that, the coordinate system V1k , V2k , V3k itself is defined. The vector V3k is constructed in
t
the same way as was the vector V3 , however, current, i.e. deformed configuration is used. The
remaining two vectors are defined as vector product:
t t
V2k  V3k  V1
t
(3.81)
t t t
V1k  V2k t V3k (3.82)

ATENA Theory 133

 t t t
The vectors V1k , V2k , V3k define local nodal shell coordinate system in which the shell rotations
are specified. As already mention, the original formulation of the element has 5 DOFs per nodes.
These are 3 displacements, expressed in the global coordinate systems and two rotations  ,  .
t t t
They are rotations along the vectors V1k , V2k . It comes from definition that V3 need not be
normal to the element surface. It must only connect the top and bottom nodes of the shell.
t
Sometimes, it is advantageous to modify the nodal coordinate system so that V3k remains
t t t
unchanged but V1k and V2k are rotated (along V3k ) to a certain direction. Note however, that
t t t
mutual orthogonality of V1k , V2k , V3k must not be damaged.

Local coordinate system

t
Local coordinates are denoted by x1L , t x2L , t x3L . The system refers to coordinate axes
t t t
X 1L , X 2L , X 3L . It is used mainly at sampling (integration) points to calculate strains and
t t t
stresses. The vector axes X 1L , X 2L , X 3L are defined by:

  t x1    t x1 
   
 r   s 
t  t x   t x 
X 3L   2    2  (3.83)
 t r   t s 
  x3    x3 
 r   s 
   
t t t
X 2L  X 3L  V1k
t t t
X 1L  X 2L  X 3L (3.84)

t t t t
As the nodal coordinate system V1k , V2k , V3k can rotate along V3k , the local coordinate system
t t t t
would X 1L , X 2L , X 3L rotate simultaneously along X 3L . This definition allows for user defined
shell local coordinate system that is common for all shell elements, irrespective of their
t t
incidences. Note that unlike V3k the vector X 3L is always normal to the element mid-plane
surface.
Curvilinear coordinate system
This system is used to calculate derivatives and integration in element integration points. Its
coordinates are r , s for in-plane direction and t in direction of element thickness, see Fig. 3-26.
The in-plane displacements are approximated by Lagrange, Hetherosis or Serendipity
approximation similar 2D isoparametric elements. For the 3rd direction, i.e. through the depth of
the element. linear approximation is used within the frame of the shell layer concept.

134
 t3 Input geometry
11 10
4 2
12 9
1
20 19
7 18
15 s

8 17 14
6
r 16
13
5

t3
11 10 External geometry
4 2
12 9
117
7
15 s

8 14
18 6
r 16
13
5

t
Internal geometry
1
2 8
3 9 7
s
4 6
5
r

Fig. 3-27 Geometry and the element’s nodes

ATENA Theory 135

 Fig. 3-28 Degenerate shell Ahmad element – coordinate systems and degree of freedoms
(DOFs)

136
3.12.2 Geometry Approximation
The coordinates of the top and bottom element surface are used to define the element geometry:
 t x1    t x1k ,top   t x1k , bot  
t  t N
 1  t  t k ,top  1  t  t k , bot  
 x2   x   hk  2  x2   2  x2   (3.85)
 t x3  k 1   t x3k ,top   t x3k , bot  
      
where N=8 is number of nodes per element, (geometry is always interpolated by 8-nodes
Serendipity interpolation, irrespective of displacement interpolation), h(r,s) is k-th interpolation
 t x1k , top   t x1k , bot 
   
function, r,s,t are isoparametric coordinates (see Fig. 3-27),  t x2k , top  and  t x2k , bot  are vector of
 t x3k , top   t x3k , bot 
   
top and bottom coordinates of point k, see Fig. 3-29.

t
X k ,top  [ t x1k ,top , t x2k ,top , t x3k ,top ]

node k t
X k ,mid  [t x1k , mid , t x2k ,mid , t x3k ,mid ]

t
X k ,bot  [t x1k ,bot , t x2k ,bot , t x3k ,bot ]

Fig. 3-29 Approximation of the element geometry

Using the above the equation (3.85) can be rewritten in the following form:
 t k , mid  tV k  
 x1 
t
  x1   3 1
 
t  t N
  t k , mid  t  t k  
 x2   x   hk   x2   2  V3 2  thick k  (3.86)
 t x3  k 1
  t x3k , mid  t k  
    V3  
 
 3 
where thick k is element thickness in node k (i.e. distance between top and bottom points) and

 t x1mid    t x1top   t x1bot  

 t mid  1   t top   t bot  
 x 2     x2    x2   (3.87)
2   t xtop   t x bot  
 t x3mid 
 k  3 k  3 k 
are coordinates of mid surface.

ATENA Theory 137

3.12.3 Displacement Field Approximation.
The general concept of displacement approximation is very similar, (although not identical) to
geometry approximation. As already mentioned, the original version of Ahmad element uses 5
T
degrees of freedom per node, see Fig. 3-28. These are  t u1mid ,t u2mid ,t u3mid ,t  ,t   , where
u1 , u2 , u3 are displacements of the element’s node at the mid-surface and t  ,t  are
t mid t mid t mid

rotations with respect to vectors v1k and v 2k respectively. These degrees of freedoms (DOFs)
are used throughout the whole element’s development. However, in order to improve
compatibility of the present shell element with other 3D elements implemented in ATENA,
T
externally the element uses  t u1top ,t u2top ,t u3top ,t u1bot ,t u2bot  DOFs, i.e. displacements at the top and
bottom of the element. The 6th displacement, i.e. u3bot is eliminated due to application of shell
theory that assumes  33  0 .
Approximation of the original three "displacement" and two rotation degrees of freedom is
independent. Nevertheless, the curvatures used in governing element equations use all of them in
the sense dictated by geometric equations. This approach enables to satisfy not only equilibrium
equations for membrane stresses and in-plane shear (in mid-surface) as it is the case of popular
Kirchhoff hypothesis, but also to satisfy equilibrium condition for transversal shears (normal to
mid-surface).
Note that in the following derivation of the element we will deal with the original set of
element’s DOFs , see (10). Every point thus has five degree of freedom,
T
 t u1mid ,t u2mid ,t u3mid ,t  ,t   . Displacement vector is calculated by:

 t k , mid   tV k t
V1k  
u 
t k
1   u1   2 1 1 
t k   
t k
  t N

 t k , mid  t 
 u   u   hk   u2   2 thick k   V2 2
t k t k
V1  t k   (3.88)
2 2
u 
t k k 1
  t u3k , mid   t k t k 
  
 3
   V2 V1  
  3 3 
T
The original displacement vector at point k has the form  t u1mid ,t u2mid ,t u3mid ,t  ,t   . Unlike in the
case of geometry approximation, were N=8, displacements approximation accounts also for
displacement in the element mid-point, i.e. N=9. The ninth function h is so called bubble
function.
u k ,mid  [ t u1k ,mid , t u2k ,mid , t u3k ,mid ]
t
V3k

node k

k
t t
V2k
t
 k

t
V1k

Fig. 3-30 Displacement approximation

138
3.12.4 Strain and Stresses Definition.
The 2nd Piolla Kirchhoff tensor and Green Lagrange strain tensor is used. They are calculated
and printed in the local coordinate system t x '1 , t x '2 and t x '3 .
Green - Lagrange tensor.
The general definition for Green-Lagrange strain tensor has the form (see eq. (1.8)):
1 t
 
t
0 ij
2
 0 ui , j  0 u j , i  0 uk , i 0 uk , j 
t t t
(3.89)

Using the above equation and applying the Von-Karman assumption, Eqn. (3.89) can be written
as:
  t u1 
   2

  t x1   1   u3  
t

 tu   2   t x1  
 0 11  
t 2
  2 
   x2
t
 t   1   t u3  
 0  22    t u  t u    t  
 2 0t12    t 1  t 2    2   x2    0t  L  0t  NL (3.90)
 t    x2  x1    t u  t u 
 2 0 13   t   3 3

 2 t    u1   u3    t x2  t x1 
t

 0 23   t
 x3  t x1   0 
 t   
  u2  u3  
t
0 
 t x  t x 
 3 2

The Von-Karman assumptions simplify the calculation of strain by accepting that:

All strains are relatively small,

The deflection normal to mid surface of shell is of order of thickness,

The both curvatures are much smaller than 1.,

The in-plane displacements are much smaller than transverse displacement and thus their
derivatives in 2nd order terms can be neglected.

t
0 L and 0t  NL represents linear and nonlinear part of strain vector, respectively. More
information about their calculation is beyond the scope of this publication. It is available e.g. in
(Jendele 1992).
2nd Piolla Kirchhoff tensor.
Energetically conjugated with Green - Lagrange tensor is 2nd Piolla Kirchhoff tensor, and this
tensor is used by the present shell element. Remind that we account for all stresses with
exclusion of normal stress which is perpendicular to shell mid surface (as it is usual practice in
shell analysis). This is the reason, why we introduced local coordinate system and all expression
are derived with respect to it.
Obviously, the local coordinate system varies depending on element deformation and thus it is
necessary to re-compute (each iteration) the transformation matrix T (that relates local and
global coordinate systems).

ATENA Theory 139

To compute internal forces, we will use 2nd Piolla Kirchhoff tensor in vector form (in a node k):
 0t S    0t S11 0
t
S 22 t
S
0 12
t
S
0 13 0
t
S 23  (3.91)
k k

Note that that it is possible to abbreviate full 3 by 3 element tensor to the above vector, because
of adopting Von Karmann simplifying assumption.

3.12.5 Serendipity, Lagrangian and Heterosis Variant of Degenerated Shell

Element.
Until now no information about interpolation function h and number of integration points were
given. The present shell element analysis uses Serendipity interpolation functions. Note that
bubble function h9 (used in displacement approximation only) represents relative departure of
approximated function with respect to the function value calculated by previous eight
approximation functions.
The interpolation functions hi read:
1
h1 (r , s )  (1- r )(1- s)(-r - s -1)
4
1
h2 (r , s )  (1- s )(1- r 2 )
2
1
h3 (r , s )  (1  r )(1- s )(r - s -1)
4
1
h4 (r , s )  (1  r )(1- s 2 )
2
1
h5 (r , s )  (1  r )(1  s )(r  s -1)
4
1
h6 (r , s )  (1  s )(1- r 2 )
2
1
h7 (r , s )  (1  r )(1- s )(r - s -1)
4
1
h8 (r , s )  (1- r )(1- s 2 )
2
1
h9 (r , s )  (1- r 2 )(1- s 2 ) (3.92)
2
The actual values in center point can be calculated by:
8
a9   ai hi (r  0, s  0)  a9 (3.93)
i 1

where hi are values of interpolation function at point (0,0), ai are corresponding node values,
a9 is departure in the center (i.e. computed value corresponding to degree of freedom at center)
and a9 is total value in center.
Depending on combination how many nodes and integration points are used, we distinguish the
Serendipity, Lagrange and Heterosis degenerated element variants, see Fig. 3-31.

140
Serendipity element.
This element was used in the original Ahmad work. It comprises eight nodal points (center point
corresponding to bubble function is omitted).
Gauss integration scheme is used for integration. It can be integrated by full, reduced, or
selective integration procedure. Using full integration, i.e. at three by three sample points,
element exhibits shear locking for thin and even moderately thick element. If reduced integration
is used. the problem of locking is significantly improved without creating spurious energy modes
on structure level. However, in case of thin element there are two non communicable spurious
energy mode on element level.
It should be noted that there were reported some difficulties if some unfavorable constraints are
applied. Nevertheless, the element is popular. If reduced integration is used the provided results
are relatively good.

Fig. 3-31 Node notation for element variants of the Ahmad shell element
Nine node Lagrangian element.
The nine-point Lagrangian element is still considered to be the best variant of the degenerated
element. This is especially because of its versatility. For full integration scheme there is no
problem with membrane and shear locking in very thin plate and shell application. If element is
moderately thick, shear locking can be improved by reduced integration scheme. However, in
that case the element exhibits rank deficiency.

ATENA Theory 141

Heterosis element.
The Heterosis element is very similar to Lagrangian element. The difference is that it assumes
first three DOFs at the element centre to be constrained, (i.e. only the rotations are retained)
The element exhibits better behavior compared with previous quadratic elements and especially
in combination with selective integration scheme no locking is produced. With reduced
integration the element provides results better than Lagrangian element. In that case there are
some spurious mechanisms, but for practical solution there are not probable.

Fig. 3-32 Summary of locking and spurious energy modes

Problem of membrane and shear locking in linear analysis are summarized in Fig. 3-32. In the
case of nonlinearity, the situation is much more complicated and depends primarily on the
material state at the sampling points. For more information refer to (Jendele, Chan et al. 1992)
Element’s integration
In previous paragraphs we mentioned many time full, reduced, and selective integration scheme.
The sense of these procedures is best to demonstrate in Fig. 3-33.

142
 Fig. 3-33 Integration schemes and sampling point notation
The steps during selective integration of shear can been explained on example of integration
arbitrary function f (r , s ) :
1/ First we calculate the value of f at sampling points that corresponds to two-by-two integration
rule, i.e
f (-0.5773,-0.5773), f (-0.5773,0.5773), f (0.5773,-0.5773), f (0.5773,0.5773)
2/ Using bilinear approximation we calculate the values of f at points that correspond to three-
by-three integration rule. There are two possibility to that.
The first one is based on original approximation area and the main idea is that we calculate the
value of function f at "corners" of isoparametric element (i.e. r  1., s  1. ):

ATENA Theory 143

 4
f (0.5773, 0.5773)   f i hi' (0.5773, 0.5773)
k 1
4
f (0.5773,0.5773)   fi hi' (0.5773,0.5773)
k 1
4
(3.94)
f (0.5773, 0.5773)   f h (0.5773, 0.5773)
i i
'

k 1
4
f (0.5773,0.5773)   f i hi' (0.5773,0.5773)
k 1

where fi are element nodal values of function f and hi' are interpolation functions corresponding
to two-by-two interpolation and a node i.

Fig. 3-34 Extension of bilinear approximation function for arbitrary rectangular

The set of equation (3.94) can be solved for fi . Having these value, we can bi-linearly
approximate function f and compute function value at any point, i.e. also at sampling points
corresponding to three by three integration rule.
The second and more elegant solution is direct approximation. The interpolation function hi are
presented for a square area of the size 2x2 units, but they can be extended to a rectangular of any
size, as shown in Fig. 3-34.
Since the functional values for the 2x2 sampling points in the corner of the square
lr  ls  2 x0.5775 are available, the approximation functions hi' can be used directly to calculate
the values of the function f at sampling points corresponding to the 3x3 integration rule.
For integration in direction perpendicular to r - s plane, that is in t coordinate it is also possible to
use Gauss integration, but due to material nonlinearity there is more advantages to use direct
rectangular integration. This concept is called the Layered model, see Fig. 3-35.
The main idea of it is to divide the element along the thickness to layers whereby in particular
layer the values of strain and stresses are expected to be constant and equal to their value at
weight point of layer. Hence the integration in t direction is computed as a sum of integrated
expressions multiplicated by adequate area of layer for all layers from bottom to top of element.
It was found that to achieve good accuracy it is necessary to about six to ten layers.
This concept. i.e. layer model is advantageous because it enables us to create for example
reinforcing layers in element and also we can use finer division near top and bottom of shell,
where higher stress level can be expected.

144
 Fig. 3-35 The layer model

3.12.6 Smeared Reinforcement

The ATENA implementation of the Ahmad shell element supports embedding of smeared
reinforcement layers. In this concept, reinforcement bars with the same coordinate z, (see Fig.
3-35), material and the same directions are replaced by a layer of smeared reinforcement. Such a
layer is placed at the same elevation z as the original reinforcement bars and its thickness is
calculated so that sum of cross-sectional area of the bars and the replacing smeared
reinforcement layer is the same. The layer is usually superimposed over existing concrete layers
and it employs CCSmeardReinforcement material law, which makes possible to account for the
original reinforcement bars’ direction.
Because each layer of the Ahmad shell can use a distinct material model, concrete and smeared
reinforcement layers are treated in similar way. (Constitutive equations, i.e. material law are
placed outside of ATENA finite elements’ code). Description of syntax of related input
commands is beyond scope of this document, but it can be found in the “ATENA Input File
Format” document.
Note that the support for smeared reinforcement does not exclude use of structural discrete
reinforcement. Both the type of reinforcement can be combined in one model to achieve the best
likeness of the the real structure with its numerical model.

3.12.7 Transformation of the Original DOFs to Displacements at the Top and

Bottom of the Element Nodal Coordinate System
This section describes in detail the whole procedure of transforming Ahmad elements from its
original formulation to the new one used by ATENA SW. Just to remind you: The original
formulation (described in the previous sections) differs from the new one in selection of element
degree of freedom, see Section 3.12.3.
Let us start to work in nodal coordinate system first. The following equation states
transformation rules for transforming three global displacements and two nodal rotations at the
element mid-plane, (i.e. the original DOFs at a node k), to 6 displacements at nodal coordinate
system, three at the top and three at the bottom surface of the shell. Note the right superscripts
“N” that indicate nodal coordinate system.

ATENA Theory 145

 t k t t 1 
 V1 1 V1k V1k thick 0 
2 3 2
 
  tV k 1
 u1  thick   t u k , mid 
t k , N , top t t
V2k V2k 0  t u1k , mid 
 t k , N , top   2 1 2 3 2  1
 u2  t k   t u k , mid   t k , mid 
 u2 
t t
 t u3k , N , top   V3 1 V3k V3k 0 0  2 
  t u3k , mid   T1  t u3k , mid 
2 3
 t k , N , bot    (3.95)
 u1   tV k t
V k t
V k 1
 thick 0  t k   t k 
u   11 1 2 1 3
      
t k , N , bot
2
 t k , N , bot   t
2
  t  k   tk 
 
 u3   V2k t t 1 
V2k V2k 0 thick
 1 2 3 2 
t k t t 
 V3 1 V3k V3k 0 0 
2 3

Transformation from nodal to global coordinated system

The next step in the element’s derivation is to write transformation of the left-hand side vector of
(3.95) from nodal to global coordinate system. It reads:

 tV k t
V2k
t
V3k 0 0 0 
 11 1 1  t k , N ,top
 u1   t k
t k , top
 u   t u1k , N ,top 
0   t 1k , N ,top 
t t
 t k , top   V1 2 V2k V3k 0 0  t k , N ,top 
2 2  u
 u2   t k t  2   u2 
 t u3k , top   V1 3 V2k V3k 0 0 0   t u k , N ,top   t k , N , top 
u
3 3
 t k , bot      T2  t k , N , bot 
3 3

t k  t k , N , bot
(3.96)
 u1   0 V3   u1  u1
t t
0 0 V k
1 1 V k
2 1  
 t u k , bot   1
  t k , N , bot 
u  t u k , N , bot 
 t 2k , bot   0 0 0
t
V1k
t
V2k V3   t k , N , bot 
t k 2
 t 2k , N , bot 
 u3   2 2 2  u
 3   u3 
 0 0 0
t
V1k V2k V3 
t k

 3 3 3

Complete transformation of the original DOFs to the new element formulation DOFs
The final transformation from the original to the new element DOFs at a node k is obtain by
substituting (3.95) into (3.96). Thus, we can write
 t u1k , top 
 t k , top   t u1k , mid   t u1k , mid 
u
 2   t k , mid   t k , mid 
 t u3k , top   u2   u2 
 t k , bot   T2 T1  u3   T  t u3k , mid 
t k , mid
(3.97)
 u1   t k   t k 
 t u k , bot       
 t k , bot   k   tk 
  
2 t
 
 3 
u
where T

In a very similar way, we can define inverse transformation, i.e. from the new DOFs to original
one. Without any derivation the matrix reads:

146
  t u1k , top 
 t k , mid
1 u   t k , top 
 t k , mid
u
  u2 
 2   t u k , top 
 t k , mid
3 u   T'  t 3k , bot  (3.98)
   u1 

t k
   t u k , bot 
 t
k   t 2k , bot 
 
 u3 

 T11 T12 T13 T11 T12 T13 

 2 2 2 2 2 2 
 
 T21 T22 T23 T21 T22 T23 
 2 2 2 2 2 2 
 
 T31 T32 T33 T31 T32 T33 
where T'   2 2 2 2 2 2 
 t k t t t t t k 
 V1 1 V1k V1k  V1k  V1k  V1 
2 3 1 2 3
 thick k thick k
thick k
thick k
thick k
thick k 
 t t t t t t k

  V2k V k
2 2 V k
2 3 V k
2 1 V2k V2 
 1 2 3

 thick k thick k
thick k thick k thick k thick k 
Constraining the redundant DOF to comply with shell theory
As noted earlier, the original set of DOFs at a node comprises 5 DOFs, whilst the new one has
T
six DOFs. Consequently, one DOF from  t u1k , top ,t u2k , top ,t u3k , top ,t u1k , bot ,t u2k , bot ,t u3k ,bot  must be
fixed. The presented work prefers to constrain t u3k , bot but u1k ,bot or u2k , bot are also good
candidates, if t u3k ,bot can not be fixed due some numerical problems, usually due to a special
position of the element with respect to global coordinate system.
Derivation of the constrain is now demonstrated on the case of t u3k ,bot . Using (3.97)

t
u3k , bot t u3k , top  (t u3k , bot t u3k , top ) 
t

u3k ,top  T16  T13  u1k , mid  T26  T23  u2k , mid  ....T56  T53   k 
t t t
 (3.99)
t t t t
t
u3k ,top  thick k V1k  k  thick k V2k  k
3 3

Now in (3.99) eliminate  k and t  k using (3.98). Thus, we obtain one equation relating
T
 t u1k , top ,t u2k ,top ,t u3k , top ,t u1k , bot ,t u2k , bot ,t u3k , bot  , which is then used to constrain t u3k ,bot as a linear
combination of t u1k , top ,t u2k , top ,t u3k , top ,t u1k , bot ,t u2k , bot :

u3k , bot  c1top t u1k , top  c2top t u2k , top  c3top t u3k , top  c1bot t u1k , bot  c2bot t u2k , bot (3.100)
where:

ATENA Theory 147

 t t t t
V1k V1k + V2k V2k
ctop
 3 1 3 1

 V    V  1
1 2 2
t k t k
1 3 2 3

t t t t
V1k V1k + V2k V2k
c2top  3 2 3 2

 V    V  1
t 2 t 2
k k
1 3 2 3

1   V    V  yy
t 2 t 2
k k 2
1 3 2 3 z
ctop

 V    V  1
3 2 2
t k t k
1 3 2 3

t t t t
V1k V1k + V2k V2k
cbot
 3 1 2 1

 V    V  1
1 2 2
t k t k
1 3 2 3

t t t t
V1k V1k + V2k V2k
c2bot   3 2 3 2

 V    V  1
t 2 t 2
k k
1 3 2 3
(3.101)
The DOFs u1k , bot or u2k , bot can be eliminated in the same way. During the execution of the
t
element, it is recommended to constrain one of u1k , bot ,t u2k , bot ,t u3k , bot based on which solution is
the most stable, (i.e. maximum denominator in (3.101)).
Constraining DOFs at the centre of Hetherosis element
A special attention needs to be paid to the 9th mid-plane node of Hetherosis element when we
have to additionally constrain t u1k , mid , t u2k , mid , t u3k , mid . Thus, of the 6 DOFs we need to constrain 4
of them.
t
For example, suppose we want to keep free u2k , top and t
u3k , top and we need to
fix t u1k ,top ,t u1k ,bot ,t u2k ,bot ,t u3k ,bot . Equation (3.99) from the previous paragraph needs to be added by
three more equations. These are:

 t u1k , top 
 t k , top 
u
 t u1k , mid  T11' T12' T13' T14' T15' T16   t 2k , top  0 
'

 t k , mid   '  u 
 u2   T21 T22 T23 T24 T25
' ' ' '
T26'   t 3k , bot   0  (3.102)
u
 t u3k , mid  T31' T32' T33' T34' T35'
   T36'   t 1k , bot  0 
u 
 t 2k , bot 
 u3 
t
Equations (3.99) and (3.102) are then solved for u1k , top ,t u1k , bot ,t u2k , bot ,t u3k , bot as a linear
combination of t u2k ,top and t u3k ,top .

148
 1
 t u1k , top   T11' T14' T15' T16'   T12' t u2k , top  T13' t u3k , top 
 k , bot   '   ' t k , top ' t k , top 
 T22 u2  T23 u3
' '
 u1     T21 T24 T25 T26'   (3.103)
 u2k ,bot   T31' T34' T35' T35'   T32' t u2k ,top  T33' t u3k , top 
 t k , bot   top   top t k ,top top t k , top 
 u3  c1 1 c1bot c2bot  c2 u2  c3 u3 
Again, there are several alternatives regarding of which of the 6 DOFs to keep and which to
eliminate. The best option is chosen the same way as described in Section 0.

3.12.8 Shell Ahmad Elements Implemented in ATENA

Several modifications the Ahmad shell elements are implemented in ATENA. They are listed
in the following table:
Table 3-5 Ahmad shell elements.

Element name Type of Number of Number of Comment

approximation in-plane in-plane
integration integration
points per points per
axis direction axis direction
for bending for shear
CCAhmadElement33L9 Lagrange 3 3 No spurious
modes, locking
in this shells
CCAhmadElement32L9 Lagrange 3 2
CCAhmadElement33H9 Heterosis 3 3
CCAhmadElement32H9 Heterosis 3 2 Good
compromise
between locking
and spurious
energy modes
CCAhmadElement22S8 Serendipity 2 2 Fast, but
spurious modes

3.13 Curvilinear Nonlinear 2D Isoparametric Layered Shell

Quadrilateral Elements
This section describes shell elements that model a structure by a curvilinear 2D surface. The
element uses hierarchical geometry and displacement interpolation. It can have from 4 to 9
nodes, each of them having 5 DOFs: 3 displacements in direction of global X,Y,Z axis and 2
rotations along user defined vectors V1 ,V2 . If the shell is in the XY plane, then typically
V1  X ,V2  Y .

ATENA Theory 149

The element uses linear geometry and displacement interpolation in the direction of its thickness
and quadratic or linear approximation in the element's plane. If quadratic approximation is used,
behavior of the element resembles behavior of Ahmad shell element described in the previous
section. 4 nodes version of this element, i.e. the element with linear approximation, does not
perform well, (the element is too stiff), and thus it is recommended only for some local links etc.
On the other hand, both bending and membrane behavior of 8-9 nodes version of the elements is
great.
The elements are derived based on the Shell theory, (similarly to Ahmad element). As a result, it
is assumed  t  0,  t is negligible and the element cannot change its thickness. (t indicates local
axis in the shell's thickness).
Depending on number of element nodes these finite elements call CCIsoShellQuad<xxxx> ...
CCIsoShellQuad<xxxxxxxxx>

Fig. 3-36 CCIsoShell2D elements

3.13.1 Geometry and displacements

The shell’s geometry at the configuration t and t  dt is defined by:

150
  t 
t
xi  hk  t X ik  ak tVi nk 
 2 
 t 
xi  hk  t  t X ik ,(i 1)  ak t  tVi nk ,( i 1) 
t  t ( i 1)
(3.104)
 2 
t  t ( i )  t 
xi  hk  t  t X ik ,(i )  ak t  tVi nk ,(i ) 
 2 
where i=1,2,3 is index relating to global axes x1 , x2 , x3 , (i.e. x,y,z), k  1...nG , nG = number of
the element's nodes used to approximate geometry, typically 8 or 9. Note that due to Shell
theory the shell thickness at node k 0 ak t ak t  t ak(i 1) t  t ak(i )  ak . The symbol t  t Vi nk ,( i ) is
ith coordinate, ( i  1, 2,3 for coordinate x, y, z ), of the vector V n at node k at time t  t ,
iteration (i ) . The vector V n is normal to the shell. Later we will also use vectors V 1 ,V 2 ,
(V 1 V 2 V n ) . They will constitute base vectors for shell's bending rotations  ,  .

Similarly, displacements at time t  t , iteration (i-1):

t  t
ui(i 1) t  t xi(i 1) t xi (3.105)
Substituting (3.104) into (3.105)

 t   t 
t  t
ui(i 1)  hk   t  t X ik ,(i 1)  ak Vi nk ,(i 1)    t X ik  ak tVi nk
t  t

 2   2 
(3.106)
 t 
 hk   t  t X ik ,(i 1) t X ik   ak
 2
 t  t
Vi nk ,( i 1)
 Vi
t nk
 

Note that in this case k  1...n , n is number of nodes to approximate displacements. Current
implementation of the shell elements assumes ng  n , (which differs for Ahmads elements).

Displacement increments within an iteration (at time t  t ) are:

 t 
ui  hk  

t  t
X ik ,(i )  t  t X ik ,(i 1)   ak
2

Vi nk ,(i ) t  t Vi nk ,(i 1) 
t  t



(3.107)
 t 
 2

 hk  U ik  ak t  t Vi nk ,( i ) t  t Vi nk ,(i 1) 



At each node, the element has 5 DOFs: 3 displacements U ik and two rotations  k ,  k described
below:
Let us define at each node of the shell a local coordinate system specified by three vectors
t  t 1k ,( i 1) t  t 2k ,( i 1) t  t nk ,( i 1)
Vi , Vi , Vi , see Fig. 3-36. The last vector is vector normal to surface of
the shell at node k and the first and second vectors are calculated as follows:
t  t
 
Vi1k ,(i 1)  e2  t  tVi nk ,(i 1) / e2  t  tVi nk ,(i 1)
(3.108)
2k ,( i 1) nk ,( i 1)
t  t
Vi t  t Vi t  t Vi1k ,(i 1)

ATENA Theory 151

 T
For the next derivation let us assume a general vector vL  v1L , v2 L , v3 L  with unit length that is
subject to rotations
[ L ,  L ,  L ]T , (where the subscript L indicates that both the vector and the rotations are defined
with respect to the local coordinate system (defined by t  t Vi1k ,( i 1) , t  tVi 2k ,(i 1) , Vi nk ,( i 1) ). The
t  t

rotations of the vector will produce displacements, (all in the local CS)
 u1L   0 v3 L v2 L   L   v3 L  L  v2 L L 
 u    v 0
  
v1L    L    v3 L L  v1L L  (3.109)
 2 L   3L
 u3 L   v2 L v1L 0    L   v2 L L  v1L  L 

Transforming the displacements from local to global coordinate system

 u1L   V1 k V1nk ,(i 1)   u1L 

t  t 1 ,( i 1)
 u1 
t  t
V12k ,(i 1) t  t

u   T u    t  t V 1k ,( i 1) t  t 2k ,( i 1) t  t nk ,( i 1)   

 2 L 2G  2 L   2 V2 V2  u 2 L  
 u3   u3 L   t  t V31k ,( i 1) t  t 2k ,( i 1)
V3 t  t nk ,( i 1) 
V3  u3 L 
 t  t V11k ,( i 1) t  t
V12k ,(i 1) t  t
V1nk ,( i 1)   v3 L  L  v2 L L 
 t  t 1k ,( i 1) t  t nk ,( i 1)   
  v3 L L  v1L L  
t  t 2k ,( i 1)
 V2 V2 V2
 t  t V31k ,( i 1) t  t nk ,( i 1)    (3.110)
  v2 L L  v1L  L 
t  t 2k ,( i 1)
 V3 V3
 t  t V11k ,( i 1)  v3 L  L  v2 L L   t  t V12k ,(i 1)  v3 L L  v1L L   t  t V1nk ,(i 1)  v2 L L  v1L  L  
 
 t  t V21k ,( i 1)  v3 L  L  v2 L L   t  t V22k ,(i 1)  v3 L L  v1L L   t  t V2nk ,(i 1)  v2 L L  v1L  L  
 
 t  t V31k ,( i 1)  v3 L  L  v2 L L   t  t V32k ,(i 1)  v3 L L  v1L L   t  t V3nk ,(i 1)  v2 L L  v1L  L  
 

Now assume the same behavior for a vector normal to the shell's surface (again in the local CS
and unit length), i.e. vL t  t ViLnk ,( i 1)   0,0,1 . When this vector gets rotated, it produces
displacements, (see (3.110):

 u1   V1 k
t  t 1 ,( i 1)
  L  t  t V12k ,(i 1)   L  
u    t  t V 1k ,( i 1)   t  t V 2k ,(i 1)  
 2  2  L 2  L  (3.111)
 u3   V3
t  t 1k ,( i 1)
  L   V3
t  t 2 k ,( i 1)
  L  

Substituting now  L t  t  k ,( i 1) ,  L t  t  k ,( i 1) and uk t  t Vi nk ,(i 1) , k  1..3 we can write
final equations for displacements due to rotations, (for iteration (i-1) and (i) and the difference):

152
t  t
Vi nk ,(i 1)   t  t
 k ,(i 1) Vi 2k ,(i 1)  t  t  k ,( i 1) t  tVi1k ,( i 1)
t  t

t  t
Vi nk ,(i )   t  t
 k ,(i ) t  t
Vi 2k ,(i 1)  t  t  k ,( i ) t  tVi1k ,(i 1)

Vi nk ,(i ) t  t Vi nk ,(i 1)    t  t  k ,( i ) t  t  k ,(i 1)  Vi 2k ,( i 1)   t  t  k ,(i ) t  t  k ,( i 1) 

t  t t  t
t  t
Vi1k ,( i 1)
 k Vi 2k ,( i 1)   k
t  t
Vi1k ,(i 1)
t  t

(3.112)

Hence, they represent rotation along two user defined vectors t  t

Vi nk ,(i 1) . It is important to note
that the vectors t  t
Vi nk ,(i 1) moves as the structure deforms.

Using (3.112) in (3.107) yields (and assuming shell thickness at a node k )

 t 
ui  hk  U ik  ak   k
 2
 t  t
Vi 2k ,( i 1)   k t  t

Vi1k ,(i 1) 

(3.113)

Note that the vectors t  t

Vi1k ,( i 1) , Vi 2k ,(i 1) ,
t  t t  t
Vi nk ,( i 1) must be normalized. Also note that
(3.111) should be used to connect dofs of shell and solid elements.

3.13.2 Connection of the shell2D to an ambient solid element

Connection of the shell2D element to an ambient structure consists of two part:

1. fix a FE node with [u, v, w] displacement within the shell2D element,

2. fix two rotation dofs of the shell2D element within ambient elements.

3.13.2.1 Fixing a FE node with [u, v, w] displacement within the shell2D element

Using the shell2D approximation the shell's displacement at the bottom uibot and at the top uitop
are:

 a 
uibot  hk  U ik  k   k
 2
 t  t
Vi 2k ,(i 1)   k Vi1k ,(i 1) 
t  t



 a 
uitop  hk  U ik  k   k
 2
 t  t 2k ,( i 1)
Vi   k t  tVi1k ,( i 1) 

 (3.114)

uitop  bot  hk ak   k  t  t
Vi 2k ,(i 1)   k Vi1k ,(i 1)
t  t


The index i is 1..3 for x..z displacements. Using the shell3D approximation displacement at the
same locations can be calculated by:
ATENA Theory 153
 uuibot  hhlbot UU ibot ,l
uuitop  hhltop UU itop ,l (3.115)
uuitop  bot  uuitop  uuibot

where hhlbot and hhltop are the solid's shell3D interpolation functions at location top and bottom
of the shells at node i, UU ibot ,l ,UU itop ,l are corresponding nodal displacements of the solid
element. Comparing (3.115) and (3.114) it can be shown that

hk  hhktop  hhkbot
(3.116)
t hk  hhktop  hhkbot

Thus, to fix [u, v, w] doffs of a node with shell2D elements we first calculate hhi values for the
case of shell3D approximation. Then, these are used to get hi , (see (3.115), comprised in the
shell2D approximation. It remains to compute shell2D rotation  i ,  i and this is (again) done by
comparing 2D and 3D approximation in (3.114) and (3.115). After some mathematical
manipulation we will arrive to the final expressions:
 top ak 2k ak 1k 
 hhk  hhk
bot
0 0  Vx  hhktop  hhkbot  Vx  hhktop  hhkbot    uk 
2 2  
 u1     vk 
u    a ak 1k
 2  0 hhktop  hhkbot 0  k Vy2k  hhktop  hhkbot  Vy  hhktop  hhkbot    wk 
k
 2 2  
 u3     k 
ak 2k ak 1k
 0 0 hhktop  hhkbot  Vz  hhktop  hhkbot  Vz  hhktop  hhkbot     
 2 2   k 
(3.117)

3.13.2.2 Fixing two rotation dofs of the shell2D element within ambient elements

Derivation of expressions to fix shell2D rotations in ambient elements is based (similarly to the
previous section) on comparing the shell2D and shell3D approximation of top and bottom nodes.
What we do is we first we fix the top and bottom in the ambient element using (solid) 3D
approximation. It yields expression something like:
uuitop  hhltop UU itop ,l  ......
(3.118)
uuibot  hhltop UU ibot , l  ......
Note that rhs of (3.118) may also include rotations. The resulting equations for shell2D rotation
 ,  are:

154
D   ak (V 1k , x 2 V 2k , x V 2k , y  V 1k , x 2 V 2k , y 2  V 1k , x 2 V 2k , z 2  V 1k , x V 1k , y V 2k , x 2 
V 1k , x V 1k , y V 2k , x V 2k , y  V 1k , x V 1k , y V 2 k , z 2  2V 1k , x V 1k , z V 2k , x V 2k , z 
V 1k , x V 1k , z V 2k , y V 2k , z  V 1k , y V 1k , z V 2k , y V 2k , z  V 1k , z 2 V 2k , x 2  V 1k , z 2 V 2k , y 2 )

 cf ktop
,1  ( V 1k , x V 2 k , y  V 1k , x V 1k , y V 2 k , x  V 1k , x V 1k , z V 2 k , z  V 1k , z V 2 k , x ) / D
2 2

 cf ktop
,2  (V 1k , x V 2 k , y  V 1k , x V 1k , y V 2 k , x  V 1k , y V 1k , z V 2 k , z  V 1k , z V 2 k , y ) / D
2 2

 cf ktop
,3  (V 1k , x V 2 k , z  V 1k , x V 1k , y V 2 k , z  V 1k , x V 1k , z V 2 k , x  V 1k , x V 1k , z V 2 k , y ) / D
2

 cf kbot
,1  (V 1k , x V 2 k , y  V 1k , x V 1k , y V 2 k , x  V 1k , x V 1k , z V 2 k , z  V 1k , z V 2 k , x ) / D
2 2

 cf kbot
,2  ( V 1k , x V 2 k , y  V 1k , x V 1k , y V 2 k , x  V 1k , y V 1k , z V 2 k , z  V 1k , z V 2 k , y ) / D
2 2

 cf kbot
,3  ( V 1k , x V 2 k , z  V 1k , x V 1k , y V 2 k , z  V 1k , x V 1k , z V 2 k , x  V 1k , x V 1k , z V 2 k , y ) / D
2

 cf ktop
,1  ( V 1k , x V 2 k , y  V 1k , x V 2 k , z  V 1k , y V 2 k , x V 2 k , y  V 1k , z V 2 k , x V 2 k , z / D
2 2

 cf ktop
,2  (V 1k , x V 2 k , x V 2 k , y  V 1k , y V 2 k , x  V 1k , y V 2 k , z  V 1k , z V 2 k , y V 2 k , z / D
2 2

 cf ktop
,3  (V 1k , x V 2 k , x V 2 k , z  V 1k , y V 2 k , y V 2 k , z  V 1k , z V 2 k , x  V 1k , z V 2 k , y / D
2 2

 cf kbot
,1  (V 1k , x V 2 k , y  V 1k , x V 2 k , z  V 1k , y V 2 k , x V 2 k , y  V 1k , z V 2 k , x V 2 k , z / D
2 2
(3.119)
 cf bot
k ,2  (V 1k , x V 2k , x V 2k , y  V 1k , y V 2k , x  V 1k , y V 2k , z  V 1k , z V 2k , y V 2k , z / D
2 2

 cf kbot
,3  ( V 1k , x V 2 k , x V 2 k , z  V 1k , y V 2 k , y V 2 k , z  V 1k , z V 2 k , x  V 1k , z V 2 k , y / D
2 2

 k   cfl top
,i ( hhl
top
UU itop ,l  ...)   cfl bot bot
,i ( hhk UU ibot ,l  ...)
 k   cf l top
,i ( hhl
top
UU itop ,l  ...)   cf l bot bot
,i ( hhk UU ibot ,l  ...)

(3.120)
k  1..number of approximation shell 2 D nodes
i  1..3, ( x..z )
l  1..number of approximation solid 3 D nodes

where V 1k , i  Vi1k V 2k , i  Vi 2k .

Note that displacement dofs are fixed by (3.117).

If either bottom or top node gets outside the ambient element, the middle point is used instead.
 1
Equation (3.120) is still valid but it is necessary to use D  D that replaces D to calculate the
2
 cf k ,1 ... cf k ,3 coefficients.
top bot

ATENA Theory 155

3.13.3 Green-Lagrange strains

The elements are derived using Green-Lagrange strains and 2nd Piola Kirchhoff stresses. Green-
Lagrange strains at (i-th iteration), i,j-axis x,y,z are calculated as follows :

t  t ( i )
t
2

1 t  t ( i ) ( i ) t  t ( i ) ( i ) t  t ( i ) t  t (i )
 ij  t ui , j  t u j ,i  t um , i t um , j 
1
  ( t  tt ui(,i j1)  t ui , j )  ( t  tt u (ji, i 1)  t u j , i )  ( t  tt um(i,i1)  t um , i )( t  tt um(i,j1)  t um , j )  (3.121)
2
t  t ( i 1)
 t  ij  t eij  tij
where:
1 t  t (i 1) t  t (i 1) t  t ( i 1) t  t (i 1)

t  t ( i 1)
t ij 
2
 t ui, j  t u j ,i  t uk ,i t uk , j 
1
e 
t ij
2
 t ui , j  t u j , i 
t  t ( i 1)
t u m , i t um , j  t u m , j t um , i 
t  t ( i 1)

 t eij0  t eij1
1 (3.122)
t eij 
0

2
 t ui, j  t u j ,i 
1 t  t ( i 1)
t eij 
1

2
 t um,i t um, j  t  tt um(i,j1) t um,i 
1
t ij 
2
 t um , i t um , j 

Element's displacements u are approximated by isoparametric interpolation. Hence, it is simple

to calculate their derivatives with respect to local coordinate r,s,t. Using an arbitrary function
f(x,y,z) Eqn. (3.123) to (3.125) show, how to compute its derivatives with respect to global x,y,z
axis.
Calculation of derivatives:

 f   x y z   f   f 
 r   r r 
r   x   x 
     
 f    x y z   f   f 
 J  (3.123)
 s   s s 
s y   y
      
 f   x y z   f   f 
 t   t t  
t   z   z 

 f   f 
   r 
 x   
 f  1  f 
 y   J  s  (3.124)
   
 f   f 
 z   t 

156
Derivatives of coordinates at t with respect to r,s,t to calculate J:
 t xi hk  t k t t t k 
  X i  ai Vni 
r r  2 
 xi hi  t k t t t k 
t
  X i  ai Vni  (3.125)
s s  2 
 t xi 1 
 hi  t ai tVnki 
t 2 

Derivatives of displacement increments at time t ...t  t (i 1) with respect to r,s,t:

 t  t ui(i 1) hk  t  t k ,(i 1) t k t 

r
 
r 
Xi  X i   ak
2
 t  t
Vi nk ,( i 1) t Vi nk 


t  t ( i 1)
 ui h  t 
s
 k   t  t X ik ,(i 1) t X ik   ak
s  2
 t  t
Vi nk ,( i 1) t Vi nk 

 (3.126)

 t  t ui(i 1) 1 
t
 hk  ak
2
 t  t
Vi nk ,(i 1) t Vi nk 


 t  t ui(i 1) hk  t  t k ,(i 1) t 
  Ui  ak t  t
dVi nk ,(i 1) 
r r  2 
t  t ( i 1)
 ui h  t 
 k  t  t U ik ,(i 1)  ak t  t
dVi nk ,(i 1) 
s s  2 
 t  t ui(i 1) 1  (3.127)
 hk  ak t  t
dVi nk ,(i 1) 
t 2 

t  t
U ik ,( i 1)  t  t X ik ,( i 1) t X ik
t  t
dVi nk ,(i 1)  t  tVi nk ,(i 1) t Vi nk

Derivatives of displacement increments at time t  t within iteration i with respect to r,s,t:

ui hk  k t
r

r
 U i  ak  
 2
Vi 
k t  t 2k ,( i 1)
  k t  tVi1k ,(i 1)  
ui h 
s
t

 k  U ik  ak   k t  tVi 2k ,(i 1)   k t  tVi1k ,(i 1)
s  2
 
ui 1 
t 2

 hk  ak   k t  tVi 2k ,( i 1)   k t  tVi1k ,( i 1) 

 (3.128)

U ik t  t U ik ,( i )  t  tU ik ,(i 1)

ATENA Theory 157

To proceed further in the derivation of the 3D isoparametric element, we need to calculate
derivatives of the displacement increments with respect to t x   t x1 , t x2 , t x3  . This is achieved
using (3.124) thru (3.128).
Derivatives of displacement increments at time t ...t  t ( i 1) with respect x1 , x2 , x3 :

 t  t ui(i 1) t inv , k  t  t ui(i 1) t inv , k  t  t ui(i 1) t inv , k  t  t ui(i 1)
 J j1  J j2  J j3 (3.129)
 t xj r s t

 t  t ui(i 1) t inv , k hk  t  t k ,( i 1) t 

 J j1  Ui  ak t  t
dVi nk ,(i 1) 
 txj r  2 
hk  t  t k ,(i 1) t 
 t J inv
j2
,k
 Ui  ak t  t dVi nk ,(i 1) 
s  2 
 1 t  t nk ,(i 1) 
 t J inv ,k
j 3 hk  ak dVi 
2 
 h hk t inv 
 t  tU ik ,(i 1)  k t J inv
j1  J j2 
 r s 
ak   hk t inv hk t inv  t inv 
 t  t dVi nk ,(i 1) t J j1  J j2   J j 3 hk  (3.130)
2   r s  
t
t  t k ,( i 1) t t  t nk ,( i 1) ak k
 U i h 
k
j dVi Gj
2

hk t inv , k hk t inv , k

t
h kj  J j1  J j2
r s
G j   t t h kj  t J inv
t k
j3
,k
hk 

Derivatives of displacement increments at time t  t within iteration i with respect to x1 , x2 , x3 :

ui , k ui , k ui , k ui

 t J inv
j1  t J inv
j2  t J inv
j3 (3.131)
 xj
t
r s t

ui , k hk  t 
 xj
t
 t J inv
j1
r 
 U i  ak  
k

2
k
 t  t
Vi 2k ,(i 1)   k 
Vi1k ,(i 1) 
t  t


hk  k t 
 t J inv
j2
,k

s 
 U i  ak  
2

k t  t 2k ,( i 1)
Vi   k t  tVi1k ,(i 1) 

 (3.132)

1 
 t J inv
j3
,k

2

hk  ak   k t  tVi 2k ,(i 1)   k t  tVi1k ,(i 1) 


After some rearrangement Eqn. (3.155) yields:

158
 ui  h h ,k 
 U ik  k t J inv
j1
,k
 k t J inv
j2 
 xj
t
 r s 
ak   h h ,k  t 
k Vi 2k ,(i 1)  t  k t J inv
t  t
j1
,k
 k t J inv
j2   J j 3 hk 
inv , k

2   r s  
a   h h ,k  t 
  k k t  tVi1k ,(i 1)  t  k t J inv
j1
,k
 k t J inv
j2   J j 3 hk 
inv , k

2   r s  
U ik t h kj   k t  t
gi1k ,( i 1) t G kj   k t  t
g i2k ,(i 1) t G kj
(3.133)

ak
t  t
gi1k ,(i 1)   t  t
Vi 2k ,(i 1)
2
ak
t  t
gi2k ,(i 1)  t  t
Vi1k ,(i 1)
2

At this place, we can derive final expression to compute linear and nonlinear strains increments.
Linear strains t eij(i ) are calculated as follows:

 t e11(i )  
 (i )   u1(i )
 t e22   
 e(i )   ... 

t e (i )   t 33(i )    B1L0  B1L1 ... B kL0  B kL1 ... B n  B n  u (ki ) 
L0 L1
(3.134)
 2 t e12   
 2 e(i )   ... 
 t 23  u (i ) 
(i )  n 
 2 t e13 

 t h1k 0 0 t  t
g11k ,( i 1) t G1k t  t
g12k ,( i 1) t G1k 
 
 0
t
h2k 0 t  t
g 12k ,( i 1) t G2k t  t 2k ,( i 1) t k
g2 G2 
 0 0 t
h3k g 31k ,( i 1) t G3k
t  t t  t 2k ,( i 1) t k
g3 G3 
B kL0   t k t  t 1k ,( i 1) t k  (3.135)
 h2
t k
h1 0 g1 G2  t  t g 12k ,(i 1) t G1k t  t 2k ,( i 1) t k
g1 G2  t  t g 22k ,(i 1) t G1k 
 0 t
h3k t
h2k t  t 1k ,( i 1) t k
g2 G3  t  t g31k ,(i 1) t G2k g2 G3  t  t g 32k ,(i 1) t G2k 
t  t 2k ,( i 1) t k
t k t  t 1k ,( i 1) t k

 h3 0 t
h1k g1 G3  t  t g31k ,(i 1) t G1k t  t 2k ,( i 1) t k
g1 G3  t  t g32k ,(i 1) t G1k 

where
 T
u (ki )  U1k ,(i ) ,U 2k ,( i ) ,U 3k ,(i ) , k ,(i ) ,  k ,( i )  at node k.

The second part of B kL , i.e. B kL1 , is derived from 2 t eij1  B kL1 u (ki ) , at node k:

ATENA Theory 159

 1 t  t (i 1)
e 
1
t ij
2
 t um,i t um, j  t  tt um(i,j1) t um,i 

1 t  t ( i 1) k t k

2
t um , i 
U m h j   k t  t g 1mk ,(i 1) t G kj   k t  t g m2k ,(i 1) t G kj 
1

 t  tt um(i,j1) U mk t hik   k t  t g 1mk ,(i 1) t Gik   k t  t g m2k ,(i 1) t Gik
2


1 k t  t ( i 1) t k

2

U m t um , i h j   k t  tt um(i,i1) t  t g 1mk ,( i 1) t G kj   k t  tt um(i,i1) t  t g m2k ,( i 1) t G kj 
1

 U mk t  tt um( i,j1) t hik   k t  tt um(i,j1) t  t g 1mk ,( i 1) t Gik   k t  tt um( i,j1) t  t g m2k ,( i 1) t Gik
2


Introducing

( i 1)  t  t um(i 1) t  t ( i 1)

l mj   u
t m, j
t x j
1k ,(j i 1)   t  t g 1mk ,(i 1) lmj( i 1) (3.136)
m

 k2 ,(j i 1)   t  t g m2k ,(i 1) lmj(i 1)

m

we can write
1 k ( i 1) t k
e 
1
t ij
2
U m lmi h j   k 1ki,(i 1) t G kj   k  2k i,( i 1) t G kj 
(3.137)
1
 U mk lmj(i 1) t hik   k 1k ,(j i 1) t Gik   k  k2 ,(j i 1) t Gik 
2

 l11( i 1) t h1k l21(i 1) t h1k l31( i 1) t h1k


 l12( i 1) t h2k l22(i 1) t h2k l32( i 1) t h2k
 l13( i 1) t h3k l23(i 1) t h3k l33( i 1) t h3k
B kL1   (i 1)
l11 h2  l12(i 1) t h1k
t k
l21(i 1) h2  l22( i 1) t h1k
t k
l31(i 1) h2  l32( i 1) t h1k
t k

l (i 1) h3  l13(i 1) t h2k

t k
l22(i 1) h3  l23
t k ( i 1) t k
h2 l32( i 1) h3  l33( i 1) t h2k
t k
 12(i 1)
l11 h3  l13(i 1) t h1k
t k
l21(i 1) h3  l23 h1
t k ( i 1) t k
l31(i 1) h3  l33( i 1) t h1k
t k

... k ,( i 1) t k
11 G1  k21,(i 1) t G1k 
k ,( i 1) t k k ,( i 1) t k 
...  12 2 G 22 G 2 
...  k ,( i 1) t k
G k ,( i 1) t k
G 
13
k ,( i 1) t k
3
k ,( i 1) t
23
k ,( i 1) t k
3
k ,( i 1) t k
(3.138)
...  11 G 
2 12 G1k  21 G 
2 22 G1 
...  k ,( i 1) t k
G  k ,( i 1) t
G2k  k ,( i 1) t k
G  k ,( i 1) t
G2k 
12
k ,( i 1) t k
3 13
k ,( i 1) t
22
k ,( i 1) t k
3 23
k ,( i 1) t

...  11 G 
3 13 G1k  21 G 
3 23 G1k 

160
The energy of nonlinear strains:
t  t
Let t S ( i 1) is a matrix storing stresses sij at time t  t , iteration (i-1):

 S11 
0 S11 
 
0 0 S11 SYM 
 
 S12 0 0 S 22 
t  t
t S ( i 1)
 0 S12 0 0 S 22  (3.139)
 
0 0 S12 0 0 S 22 
S 0 0 S 23 0 0 S33 
 13 
0 S13 0 0 S 23 0 0 S33 
0 0 S13 0 0 S 23 0 0 S33 


Then matrix B NL  B1NL ... B kNL ... B nNL  is composed so that (at a node k)

 
ij sij ij   u (ki )   B kNL 
T T t  t
t S (i 1) B kNL u (ki ) (3.140)

where  states for variation of the following entity. It can be shown that the matrix B NL can be
set in the following shape:
 t h1k 0 0 t  t
g11k ,(i 1) t G1k t  t
g12k ,(i 1) t G1k 
 t  t 2k ,( i 1) t k 
 0
t
h1k 0 t  t
g 12k ,(i 1) t G1k g2 G1 
 0 t k
0 t
h1k t  t
g31k ,(i 1) t G1k t  t 2
g3 k ,( i 1)
G1
t k t  t 2k ,( i 1) t k 
 h2 0 0 t  t
g11k ,(i 1) t G2k g1 G2 
t k
B kNL  0 t
h2k 0 t  t
g 12k ,(i 1) t G2k t  t 2
g2 k ,( i 1)
G2 (3.141)
 t k t  t 1k ,( i 1) t k t  t 2k ,( i 1) t k

 0 0 h2 g3 G2 g3 G2 
 t hk t  t 2k ,( i 1) t k 
0 0 t  t
g11k ,(i 1) t G3k g1 G3
 3 
 0 t
h3k 0 t  t
g 12k ,(i 1) t G3k t  t 2k ,( i 1) t k
g2 G3 
 t  t 2k ,( i 1) t k 
 0 0 t
h3k t  t
g31k ,(i 1) t G3k g3 G3 

Having the matrices (3.135), (3.138), (3.139) and (3.141) these are used to compute the element's
stiffness matrix, mass matrix, element loads etc. in exactly the same way as it is done for other
ATENA's element.

3.14 Curvilinear Nonlinear 2D Isoparametric Layered Shell

Triangular Elements
This section describes triangular shell finite elements. Their properties and their derivation are
much the same as that for quadrilateral shell finite elements CCIsoShellQuad<xxxx> ...
CCIsoShellQuad<xxxxxxxxx> described in the previous chapter. The only difference in that
they feature triangular shape. Their geometry is depicted in the figure below

ATENA Theory 161

Depending on number of element nodes these finite elements call CCIsoShellTriangle<xxx> ...
CCIsoShellTriangle<xxxxxx>.

Fig. 3-37 CCIsoShell2D triangular elements

3.15 Curvilinear Nonlinear 3D Isoparametric Layered Shell

Hexahedral Elements

A family of 3D isoparametric shell elements is presented, see the figure below. Their properties
lie between degenerated Ahmad shell elements from Section 3.12 and full 3D brick elements
from Section 3.5.
Shape and kinematic behaviour resembles that of the shell's element. All points through the
shell's thickness remain located on a line passing thru the corresponding top and bottom nodes of
the shell, however unlike in the classical shell theory, their distance can change. As for degrees
of freedom, (DOFS), a typical 3D isoparametric shell element has 9 nodes at the top and nine
nodes at the botom surface, each of them having 3 DOFS, (i.e. 3 displacements). A similar 2D

162
shell element would feature 9 nodes located at the shell's midplane, each of them having 5
DOFS, (3 displacements plus 2 rotations).
The new elements use full 3D static equations. i.e. the elements consider all 6 components of 3D
stress and strain vector. Geometrical and material nonlinearity is supported. The governing
equations are calculated and integrated in material points. Gauss integration is used in shell's
plane direction, whilst layered concept is employed throughout the thickness of the shells, (i.e.
rectangur quadrature). As each layer can use different material model, some layers can be
employed for modelling of embedded reinforcement. The elements typically use 3 x 3 x
number_of_layers integration (i.e. material) points.
The elements are suitable for both shallow and deep shells and are extremely simple for use,
because they can be input and output as usual 3D solid hexahedral elements with 8, 20 or 27
nodes. Hence, these shells can be hadled with most 3D pre- and post-processors. They also use
standard 3D material models, element loads and other boundary conditions designed for
hexahedral elements.
The presented shell elements are particularly useful for structures that combine solid 3D
elements and shell elements, because they do not imply any additional shell kinematic constraint
that would harm an anjancent 3D solid elements. (Typical shell elements assume  t  0 that
enforces the same displacements of the corresponding top and bottom nodes in direction of their
connecting line). They are designed for bent shells and to analyze these structures (with the same
accuracy) they require far less finite elements compared to a similar analysis using standard
hexahedral elements. On the other hand, the 3D behaviour of these elements involves a small
overhead, so that standard 2D shell elements (with only 5 stress/strain components per material
point) can perform in some cases slightly better. Nevertheless, the overhead is well paid off by
easy of use of the presented elements, their nice 3D visualization, simple connection to adjacent
3D solid parts of the structure etc. In addition, the hiearchical isoparametric space interpolation
(used for the presented 3D shell elements) ensures that finer and coarser meshes are easy to
connect. Of coarse, this feature must be supported by pre- and postprocessor being used.

ATENA Theory 163

 Fig. 3-38 Isoparametric 3D shell element - coordinate systems

Geometry and displacements are approximated by hiearchical isoparametric spatial interpolation,

(similar to other 2D and 3D elements defined in previous sections). The elements have at
minimum 4 points at its top and 4 points at its bottom surface. It corresponds to linear
approximation and the element's name CCIsoShellBrick<xxxxxxxx>. The most accurate version
of the elements uses nodes 1 to 16 and 21,22, see the figure above. Its name is
CCIsoShellBrick<xxxxxxxxxxxxxxxxxx>. Such element can have curvilinear shape and features
quadratic displacement approximation. Hierarchical concept the shell element is employed.
Hence, the 3D shell element can have from 8 to 18 nodes. The nodes 1-8 are compulsory. Nodes
9-16 and 21,22 are optional. Nodes 17 to 22 are automatically removed from the element's
incidences. They are considered only for the sake of compatibility with input data preprocessor.
The <xxxxx..> string in the element name (following CCIsoShellBrick) specifies, which of the
element's node is (or is not) included. An included node is market as "x", a node not included is
marked as "_", (underscore). The shell's nodes are maped into the string as follows:
<1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,21,22>. For example,
CCIsoShellBrick<xxxxxxxxx_x_x_x_xx> uses nodes 1-8,9,11,13,15,21,22. Note that the bottom
and top surface must use the same number and location of the optional nodes. Hence, if node 9 is
included, node 13 must be included, too.

164
3.15.1 Geometry and displacements

The shell’s geometry at the configuration time t and t  dt , (iteration (i-1) and (i)), is defined by:
 1  t t k , top 1  t t k , bot 
t
xi  hk  Xi  Xi 
 2 2 
t  t ( i 1)  1  t t  t k , top ( i 1) 1  t t  t k ,bot ( i 1) 
xi  hk  Xi  Xi  (3.142)
 2 2 
t  t ( i )  1  t t  t k , top (i ) 1  t t  t k , bot ( i ) 
xi  hk  Xi  Xi 
 2 2 

where i=1,2,3 is index relating to global axes x1 , x2 , x3 , (i.e. x,y,z), hk  hk (r , s ) is k-th

interpolation function, (see Table 3-4), k  1...nG is number of the shell's nodes, nG = number of
t
the element's nodes used to approximate geometry, typically 8 or 9. xi represents i-th
coordinate of a node of the element (at the specified time).
Displacements at time t  t ( i 1) , i=1,2,3 for global axes x,y,z, at iteration (i  1) reads :
t  t
ui(i 1)  t  t xi( i 1)  t xi (3.143)

Substituting (3.142) into (3.143), i=1,2,3 for global axes x,y,z, we can derive

t  t  1  t t  t k , top (i 1) 1  t t  t k , bot (i 1)   1  t t k , top 1  t t k , bot 

ui(i 1)  hk  Xi  Xi   hk  Xi  Xi  
 2 2   2 2 
 1  t t  t k , top (i 1) t k , top 1  t  t  t k , bot ( i 1) 1  t t k , bot  
 hk   Xi  Xi    Xi  Xi 
 2 2  2 
1 t t  t 1 t 
 hk  U ik , top (i 1)  t  t
U ik ,bot ( i 1) 
 2 2 
(3.144)
where

t  t
U ik , top (i 1)  t  t X ik , top (i 1)  t X ik , t  t
U ik , bot ( i 1)  t  t X ik , bot ( i 1)  t X ik

t  t
Displacement increments within i-th iteration are calculated as ui(i )  t  t xi( i )  t xi(i 1) :

1 t t  t 1 t t  t 
ui(i )  hk  U ik , top ( i )  U ik , bot (i )  (3.145)
 2 2 

ATENA Theory 165

 t  t
where U ik ,top (i )  t  t X ik ,top ( i )  t X ik (i 1) , t  t
U ik , bot (i )  t  t X ik , bot (i )  t X ik ( i 1) . In the above X ik ,top
t  t
and X ik , bot is top and bottom nodal coordinate of node i. Similarly, U ik , top ( i 1) , t  t
U ik , bot ( i 1)
denotes displacements at the same node.

3.15.2 Green-Lagrange strains

The elements are derived using Green-Lagrange strains and 2nd Piola Kirchhoff stresses. Total
Lagrangian formulation is employed, but after each load step we transform the analyzed model
(and its stress and other tensors) to the coordinate system defined by the current shape of the
model. (The standard Total Lagrangian formulation calculates all with respect to the original
coordinate system without any transformation; Updated Lagrangian formulation carries all the
transformation each transformation, BATHE(1982). )
The shell's total strains at time t  t , i-th iteration, are calculated: ( i, j=1..3 for axis x,y,z)
1 t  t (i ) t  t ( i ) t  t (i ) t  t ( i )

t  t ( i )
t ij 
2
 t ui, j  t u j ,i  t uk ,i t uk , j 
1

  t  tt ui(,i j1)  t ui(,i )j    t  tt u (ji, i1)  t u (ji,)i    t  tt uk( i,i 1)  t uk(i,)i  t  tt uk(i,j1)  t uk(i,)j 
2
 (3.146)

 t  ij  t eij(i )  tij(i )
t  t ( i 1)

t  t ( i ) t  t
where u
t i, j is derivative of displacement ui( i ) with respect to axis t x j at time t, i.e. at the
beginning of time step. (i) refers to iteration number. Similarly, t ui(,i )j denotes displacement
increment at the current iteration.
1 t  t ( i 1) t  t (i 1) t  t (i 1) t  t ( i 1)
Subtracting
2
 t ui, j  t u j ,i  t uk ,i t uk , j  from (3.146) we can calculate

t  t ( i 1)
t ij 

linear and nonlinear strain increments t eij(i ) and tij(i ) :

1
e 
(i )
t ij
2
 t ui , j  t u j , i 
(i ) (i ) t  t ( i 1)
t uk , i t uk , j 
(i )
t uk , j t uk , i 
t  t ( i 1) (i )

(3.147)
1
t  (i )
ij   t uk( i,)i t uk(i,)j 
2

Derivatives with respect to global x   x1 , x2 , x3  are calculated in standard way from derivatives
with respect to curvilinear isoparametric coordinates r   r , s, t    r1 , r2 , r3  . For example,
derivatives of a function f ( x1 , x2 , x3 ) is:

166
  f   x1 x2 x3   f   f 
 r   r    
r r   x1   x1 
   
 f    x1 x2 x3   f   f  f x j f f
   J  , i.e.   J ij (3.148)
 s   s s 
s  x2 
     x2  ri ri x j x j
 f   x1 x2 x3   f   f 
 t   t t  
t   x   
 3  x3 

 f   f 
   r 

 1x
 
 f  1  f  f f
   J   , i.e.  J inv
ji (3.149)
 x2   
s xi rj
 f   f 
   t 
 x3 

The presented shell elements employs isoparametric hierarchical interpolation. Hence,

coordinates t x of a point are calculated by:
 1  t t k , top 1  t t k , bot 
t
xi  hk  Xi  Xi  (3.150)
 2 2 
 t xi
where the interpolation functions hk (r , s ) are enlisted in Table 1-3-1 and their derivatives
ri
with respect to r,s,t (to calculate J) are:

 t xi hk
 1  t t k , top 1  t t k , bot 
  Xi  Xi 
r r
 2 2 
 xi hk  1  t t k , top 1  t t k , bot 
t
  Xi  Xi  (3.151)
s s  2 2 
 t xi hk
t

2
 t
X ik , top  t X ik , bot 

t  t
The above expressions are employed to obtain derivatives of (total) displacements ui(i 1) with
respect to r,s,t. They are needed to calculate strains (3.147).

 t  t ui(i 1) hk  1  t t  t 1 t 

  U ik , top (i 1)  t  t
U ik , bot (i 1) 
r r  2 2 
 t  t ui(i 1) hk  1  t t  t 1 t 
-   U ik , top (i 1)  t  t
U ik , bot (i 1)  (3.152)
s s  2 2 
 t  t ui(i 1) hk
t

2
 t  t
U ik , top (i 1)  t  t
U ik , bot (i 1) 
ATENA Theory 167
Derivatives of displacement increments with respect to r,s,t:
 t  t ui(i ) hk 1 t t  t 1 t t  t 
  U ik , top ( i )  U ik , bot (i ) 
r r  2 2 
 t  t ui(i ) h  1  t t  t 1 t t  t 
 k U ik , top ( i )  U ik , bot (i )  (3.153)
s s  2 2 
 t  t ui(i ) hk
t

2
 t  t
U ik , top ( i )  t  t
U ik , bot (i ) 

To proceed further in the derivation of the 3D isoparametric element, we need to calculate

derivatives of the displacement increments with respect to t x   t x1 , t x2 , t x3  . This is achieved
using (3.149) and (3.153):
ui(i ) t inv ui(i ) t inv ui(i ) t inv ui(i ) t inv ui(i )
 J jl  J j1  J j2  J j3 (3.154)
 t xj rl r s t

ui(i ) t inv hk  1  t t  t 1 t t  t 

 J j1  U ik , top (i )  U ik , bot (i ) 
 txj r  2 2 
hk  1  t t  t k , top (i ) 1  t t  t k , bot (i ) 
 t J inv
j2  Ui  Ui  (3.155)
s  2 2 
hk t  t k , top (i ) t  t k , bot (i )
 t J inv
j3
2
 Ui  Ui 
After some rearrangement Eqn. (3.155) yields:

ui(i ) t  t k , top (i )  t inv hk 1  t t inv hk 1  t t inv hk 

 Ui  J j1  J j2  J j3 
 txj  r 2 s 2 2 
t  t  hk 1  t t inv hk 1  t t inv hk 
U ik , bot (i )  t J invj1  J j2  J j3  (3.156)
 r 2 s 2 2 
hktop, j t  tU ik , top ( i )  hkbot, j t  tU ik , bot (i )

where
 hk 1  t t inv hk 1  t t inv hk  bot  t inv hk 1  t t inv hk 1  t t inv hk 
hktop, j   t J inv
j1  J j2  J j3  , hk , j   J j1  J j2  J j3 
 r 2 s 2 2   r 2 s 2 2 
At this place, we can derive final expression to compute linear and nonlinear strains increments.
Linear strains t eij(i ) are calculated as follows, see (3.147):

168
  t e11(i )  
 (i )   u1(i )
 t e22   
 t e33 (i )   ... 

t e (i )   (i )    B1L0  B1L1 ... B kL0  B kL1 ... B n  B n  u (ki ) 
L0 L1
(3.157)
 2 t e12   
 2 e(i )   ... 
 t 23  u (i ) 
(i )  n 
 2 t e13 

 t hktop,1 0 0 t hkbot,1 0 0 
 top top 
 0 t hk ,2 0 0 t hk ,2 0 
 0 0 hktop,3 0 0 top 
t hk ,3
B kL0   top top
t
top top  (3.158)
 t hk ,2 t hk ,1 0 t hk ,2 t hk ,1 0 
 0 top top top top 
t hk ,3 t hk ,2 0 t hk ,3 t hk ,2
 top top top top

 t hk ,3 0 t hk ,1 t hk ,3 0 t hk ,1 


 T
where u (ki )  U1k , top ( i ) ,U 2k , top ( i ) ,U 3k , top ( i ) ,U1k , bot ( i ) ,U 2k , bot ( i ) ,U 3k , bot ( i )  at node k.

Introducing
 t  t ui(i 1)
lij(i 1)   t  t ( i 1)
u
t i, j (3.159)
t x j

we can write
t  t ( i 1)
u
t m,i t um(i,) j  t  tt um(i,j1) t um(i,)i 
lmi(i 1)  hktop, j t  tU mk ,top (i )  hkbot, j t  t
U mk , bot (i )   lmj( i 1)  hktop,i t  tU mk , top ( i )  hkbot, i t  t
U mk , bot ( i )   (3.160)
U mk , top (i )  lmi( i 1) hktop, j  lmj( i 1) hktop, i   t  t U mk , bot (i )  lmi(i 1) hkbot, j  lmj(i 1) hkbot, i 
t  t

Finally, matrix B L1 yields

 l11(i 1)t hktop,1 l21( i 1)t hktop,1 l31(i 1)t hktop,1 l11(i 1)t hkbot,1

 l12(i 1) hktop,2 l22(i 1) hktop,2 l32(i 1) hktop,2 l12(i 1) hkbot,2
 l13(i 1) hktop,3 l23(i 1) hktop,3 l33(i 1) hktop,3 l13(i 1) hkbot,3
B kL1   (i 1) top ( i 1) top
l11 hk , 2  l12 hk ,1 l21( i 1) hktop,2  l22(i 1) hktop,1 l31(i 1) hktop,2  l32( i 1) hktop,1 l11( i 1) hkbot,2  l12( i 1) hkbot,1
l (i 1) htop  l (i 1) htop l22(i 1) hktop,3  l23(i 1) hktop,2 l32(i 1) hktop,3  l33( i 1) hktop,2 l12(i 1) hkbot,3  l13( i 1) hkbot,2
 12(i 1) ktop,3 13(i 1) ktop,2
l11 hk ,3  l13 hk ,1 l21(i 1) hktop,3  l23(i 1) hktop,1 l31(i 1) hktop,3  l33( i 1) hktop,1 l11(i 1) hkbot,3  l13( i 1) hkbot,1

ATENA Theory 169

 ... ( i 1) bot
l21 t hk ,1 l31(i 1)t hkbot,1 

... l22(i 1) hkbot,2 l32(i 1) hkbot,2 
... l23(i 1) hkbot,3 ( i 1) bot
l33 hk ,3 
 (3.161)
... l21(i 1) hkbot,2  l22( i 1) hkbot,1 l31( i 1) hkbot,2  l32( i 1) hkbot,1 
... l22(i 1) hkbot,3  l23( i 1) hkbot,2 l32(i 1) hkbot,3  l33(i 1) hkbot,2 

... l21(i 1) hkbot,3  l23(i 1) hkbot,1 l31(i 1) hkbot,3  l33(i 1) hkbot,1 

t  t

Assembling stresses at time t  t , iteration (i-1) into matrix t S (i 1) , participation of nonlinear
strains tij(i ) is, see (3.147)

1 
t  t
t Sij( i 1)   tij(i )   t  t
t Sij(i 1)   tij(i )   t  t
t Sij(i 1)    t uk(i,)i t uk(i,)j  
2 
1 
 t  tt Sij(i 1)   t uk( i,)i t uk(i,)j  t uk(i,)i t  uk( i,)j  
2 
 t Sij  t uk ,i t uk , j
t  t ( i 1) (i ) (i )

 T  (3.162)
  u1( i )  u1(i )
   
 ...    ... 
  u k   B 
 NL T t  t ( i 1) NL  ( i )
( i )
tS B u k 
   
 ...   ... 
 u   ( i ) u (i ) 
 n   n 

B NL B1NL ... B kNL ... B nNL 

 t hktop,1 0 0 t hkbot,1 0 0 
 
 0 t hktop,1 0 0 t hkbot,1 0 
 0 0 t hktop,1 0 0 bot 
t hk ,1
 top 
 t hk ,2 0 0 t hkbot,2 0 0 
B kNL  0 hktop,2 0 0 hkbot,2 0 
 t t

 0 0 t hktop,2 0 0 h bot
t k ,2 
 htop 0 0 hkbot,3 0 0  (3.163)
 t k ,3 t 
 0 t hktop,3 0 0 t hkbot,3 0 
 bot 
 0 0 t hktop,3 0 0 t hk ,3 

170
  t  tt s11(i 1) 
 t  t ( i 1) 
 0 s
t 11 symmetric 
 0 0 t  t ( i 1)
s
t 11

 t  t (i 1) t  t ( i 1) 
  t s12 0 0 s
t 22 
t  t ( i 1)
tS  0 t  t ( i 1)
s 0 0 t  t ( i 1)
s  (3.164)
 t 12
t  t ( i 1)
t 22
t t ( i 1)

 0 0 s
t 12 0 0 s
t 22 
 t  t s ( i 1) 0 0 t  t ( i 1)
s 0 0 t  t ( i 1)
s 
 t 13 t 23 t 33 
t  t ( i 1) t  t ( i 1) t  t ( i 1)
 0 s
t 13 0 0 s
t 23 0 0 s
t 33 
 t  t ( i 1) t t ( i 1) t  t ( i 1) 
 0 0 s
t 13 0 0 s
t 23 0 0 t s33 

Using (3.148) and (3.149) it follows to present final expression for computation of space
derivatives of f ( x1 , x2 , x3 ) :

f  h k  
 J inv
ji  Fk  ; h k  h k (r , s, t )  hˆ k (r , s ) h k (t )
xi  rj 
 
   hˆ k  k 
f   ˆk k  
xi
 J ji  Fk
inv
 
h (r , s ) h (t )   J ji   Fk 

inv

   rj
h  hˆ
k k  h
  
rj   
 rj    
  k
   hˆ k   h k     hˆ k  h k     hˆ k   h 
J1i   Fk 
inv
h  hˆ
k k
    J 2i   Fk 
inv
h  hˆ
k k
    J 3i   Fk 
inv
h  hˆ
k k
 
   r r       s s       t t   
        
k
ˆ
 h k k 
 h k ˆ k 
 h 
J1inv
i  Fk h   J 2invi k  F h   J 3invi  Fk hˆ k 
 r   
   s   t 
(3.165)

Having all the matrices and relationships above, the rest of derivation of the presented
isoparametric shell elements is straightforward. Simply use the matrices B L0 , B L1 , B NL and

t  t ( i 1)
tS to calculate structural stiffness matrices t K L , t  tt K NL
( i 1)
, vectors of nodal forces
t  t
F ( i 1) , and loads t  t R as described in the Section Problem Discretisation Using Finite
Element Method earlier in this document.

3.16 Curvilinear Nonlinear 3D Isoparametric Layered Shell Wedge

Elements
This section describes wedge shell finite elements. Their properties and their derivation are much
the same as that for hexahedral shell finite elements CCIsoShellBrick<xxxxxxxx> ...
CCIsoShellBrick<xxxxxxxxxxxxxxxxxx> described in the previous chapter. The only diference
in that they feature wedge shape. Their geometry is depicted in the figure below

ATENA Theory 171

Depending on number of element nodes these finite elements call CCIsoShellWedge<xxxxxx>
... CCIsoShellWedge<xxxxxxxxxxxx>.

Fig. 3-39 CCIsoShell3D wedge elements

172
3.17 Curvilinear Nonlinear 3D Beam Element
A curvilinear 3D beam finite element CCBeamNL is described here. The element is based on a
similar beam element from BATHE (1982). It is fully nonlinear, in terms of its geometry and
material response. It uses quadratic approximation of its shape, so it can be curvilinear, twisted,
with variable dimensions of the cross-sections. Moreover, beam’s cross-sections can be of any
shape, optionally even with holes.
The element belongs to the group of isoparametric elements with Gauss integration along its axis
and trapezoidal (Newton-Cotes) quadrature within the cross-section. The integration (or
material) points are placed in a way similar to the layered concept applied to shell elements,
however, the “layers” are located in both “s,t” directions.

3.17.1 Geometry and Displacements and Rotations Fields

Geometry of the element is depicted in Fig. 3-40. The depicted brick nodes specification is
employed to ensure compatibility of the element with ATENA preprocessor. The beam 3D nodes
definition is used by ATENA postprocessor. The element response is computed within the 1D
beam geometry. Thus, on input the element has 20 nodes, while during the calculation it has only
15 nodes, i.e. 12 nodes for 3D beam shape definition and 3 nodes for the 1D beam geometry.
Any of the 15 nodes can be subject to a kinematic or static constraint. The 1D beam nodes have
6 degrees of freedom (dofs) – three displacements and three rotations with respect to global
coordinate axes. The 3D beam nodes allocate only the three displacement dofs per node. The
redundant brick nodes are ignored, and they allocate no dofs.
The element uses three configurations. The reference configuration corresponds to shape of the
beam at the beginning of the step, i.e. prior any load in the current step is applied was employed.
It is used as a reference coordinate system for all calculation within a loading step t, with respect
to which all derivatives are computed. This configuration is denoted by a t superscript left to a
referred symbol, e.g. t x . The element shape after all previous iterations within the current step
and prior the current iteration is denoted by t  dt superscript, t  dt x . Increments within the
current iteration do not use any superscript, e.g. x.

ATENA Theory 173

 t
a2 Geom etry
t
b2

t 1
Vt

t
t
b3 a3

6
13
t
a1 1
Vs 14 5 Bric k nodes
7 16
t
s
b1 18 15 8

19
r 17 6
20 5 Beam 3D nodes
7
10 8
2
9
1 11
10
12 9
3 11 Beam 1D nodes
4 14
12

2
1 15 t

3
4
w
z
13 z 2

2 s
2
y r
x
x
y v
u Isoparam etric shape
Global c oord. system and elem ent dofs
Fig. 3-40 CCBeamNL element

174
The beam’s geometry at the configuration t and t  dt is defined by:
 t s 
t
x  hi  t X i  t ai tVi tx  t bi tVi sx 
 2 2 
 t t s s 
t
y  hi  t Yi  t ai tVi y  t bi tVi y  (3.166)
 2 2 
 t s 
t
z  hi  t Z i  t ai tVi tz  t bi tVi sz 
 2 2 
In the above i refers to axial nodes, i.e. i  1..3 for the nodes 13,14,15, see the 1D beam nodes.
T
hi  hi (r ) is i-th nodal interpolation function i described in Section 3.2.  t X i ,t Yi ,t Z i  are
T T
global coordinates of a node i at time t. The vectors  t Vi tx ,t Vi y ,t Vi tz  ,  t Vi sx ,t Vi y ,t Vi sz  are the
t s

vectors t Vt , tVs depicted in Fig. 3-40, in a cross section i, at time t, which define local coordinate
axis s,t. The symbols t ai , t bi refers to dimensions of the cross section i, time t; see the figure,
too.
Geometry of the beam at time t  dt is defined in a similar way:
t  dt  t s 
x  hi  t  dt X i  t ai t  dtVi tx  t bi t  dtVi sx 
 2 2 
t  dt  t t s s 
y  hi  t  dt Yi  t ai t  dtVi y  t bi t  dtVi y  (3.167)
 2 2 
t  dt  t s 
z  hi  t  dt Z i  t ai t  dtVi tz  t bi t  dtVi sz 
 2 2 
The element’s displacements at time t  dt is calculated as follows:
t  dt
u  t  dt x t x
t  dt
v  t  dt y t y (3.168)
t  dt
w  t  dt z t z
and displacement increments within a iteration:
 t s 
u  hi  U i  t ai Vi tx  t bi Vi sx 
 2 2 
 t t s s 
v  hi  Vi  t ai Vi y  t bi Vi y  (3.169)
 2 2 
 t s 
w  hi  Wi  t ai Vi tz  t bi Vi sz 
 2 2 
In the above equation the vectors Vi t , Vi s are Vi t t  dt Vi t t Vi t and Vi s t  dt Vi s t Vi s are
approximated by

ATENA Theory 175

 Vi tx   Vi tz iy 
t  dt t
Vi y iz
t  dt

t
Vi y   t  dt
Vi tx iz  Vi tz ix
t  dt

Vi tz   t  dt t
Vi y ix  Vi tx iy
t  dt

(3.170)
Vi sx   t  dt
Vi sz iy  t  dt s
Vi y iz 
s
Vi y   t  dt
Vi sx iz  t  dt
Vi sz ix 
Vi sz   t  dt
Vi y ix 
s t  dt
Vi sx iy 
The parameters ix ,iy ,iz are rotations around the global axis, with respect to beginning of the
current load step. Note that (3.170) is valid only approximately.

3.17.2 Strain and Stress Definition

The element uses Green-Lagrange strain and Piola-Kirchhof stresses, see Section 1.4.2 and
Section 1.3.2. transformed to the local isoparametric r,s,t coordinate system. As the beam theory
implies, only normal strain component  r and shear components  rs ,  rt are considered. The
stress vector includes the corresponding  rr , rs , rt entries, whereby the remaining strains have
to remain zero. The procedure of calculation stress-strain response is as follows:
1. Calculate all 6 components of Green-Lagrange strains (1.8) and their increments within
global coordinate systems. The increments are computed with respect to the beginning of
the current load step.
2. Transform the strains increments into local r,s,t coordinate system.
3. Zeroise components  ss ,  tt ,  st .
4. Execute material law to compute corresponding stresses.
5. Transform the stresses to the global coordinate system.
The following expressions are used to calculate displacement derivatives needed for calculation
of the strains:
1
 df   dx dy dz   df   df 
 dx   dr dr dr   dr   dr 
       
 df   dx dy dz   df   J 1  df 
 dy    ds ds ds   ds   ds 
(3.171)
       
 df   dx dy dz   df   df 
 dz   dt dt dt   dt   dt 
where f is a displacement function to be derived.

3.17.3 Matrices Used in the Beam Element Formulation

Substituting equations (3.166) to (3.171) into the expressions for calculating element matrices
(1.31) to (1.34) all important matrices and vectors of the beam element can be calculated. Their

176
explicit presentation is beyond the scope of this document. Nevertheless, the most important
ones are now given:
The Jacobian matrix:
 t x hi  t t t t t x s t t sx 
J11    X i  ai Vi  bi Vi 
r r  2 2 
 y hi  t
t
t t t ty s t t sy 
J12    Yi  ai Vi  bi Vi 
r r  2 2 
 t z hi  t t t t t z s t t sz 
J13    Z i  ai Vi  bi Vi 
r r  2 2 
t x 1 
J 21   hi  t bi tVi sx 
s 2 
t y 1 s 
J 22   hi  t bi tVi y 
s 2 
y
t
1 
J 23   hi  t bi tVi sz 
t 2 
t x 1 
J 31   hi  t ai tVi tx 
t 2 
t y 1 t 
J 32   hi  t ai tVi y 
t 2 
(3.172)
z
t
1 
J 33   hi  t ai tVi tz 
t 2 

ATENA Theory 177

 t  dt
The matrix t BNL :
It is constructed in the way that

U 1 
 u  V 
 x   1
  W1 
 u   x
 y  1 
  1y 
 u   z
 z  1 
  U 
 v   2
 x   V2 
 v  W 
   t  dt
B  2  t  dt
BNL U
 y 
t NL
 2x  t

 v   y
   2 
 z   2z 
 w   
  U 3 
 x  V 
 w   3
  W3 
 y   x 
 w   3y  (3.173)
  3 
 z   z
 3 
The detailed expressions for calculating t  dtt BNL are given in (3.176) and (3.177). The equations
are important because they present the way, how spatial derivatives of all the displacements are
calculated. The entries in t  dtt BNL are thus used to setup also the matrix t  dtt BL 0 and t  dtt BL1 .
These matrices are computed as follows:

t  dt t  dt
t BL 0(1, i )  t BNL (1, i )
t  dt t  dt
t BL 0(2, i )  t BNL (5,i )
t  dt t  dt
t BL 0(3,i )  t BNL (9,i )
t  dt t  dt
(3.174)
t BL 0(4, i )  t BNL (4, i )  t  dtt BNL (2, i )
t  dt t  dt
t BL 0(5, i )  t BNL (6, i )  t  dtt BNL (8,i )
t  dt t  dt
t BL 0(6, i )  t BNL (7, i )  t  dtt BNL (3, i )

178
 t  dt  t  dt u t  dt  t  dt v t  dt  t  dt w t  dt
t BL1(1, i )  B
t NL (1, i ) + B
t NL (4, i ) + t BNL (7, i )
x x x
t  dt  t  dt u t  dt  t  dt v t  dt  t  dt w t  dt
t BL1(2,i )  B
t NL (2, i ) + B
t NL (5, i ) + t BNL (8, i )
y y y
t  dt  t  dt u t  dt  t  dt v t  dt  t  dt w t  dt
t BL1(3, i )  B
t NL (3, i ) + B
t NL (6, i ) + t BNL (9 i )
z z z
t  dt  t  dt u t  dt  t  dt u t  dt  t  dt v t  dt
t BL1(4,i )  t BNL (2, i )  t BNL (1, i ) + t BNL (5, i ) +
x y x
 t  dt v t  dt  t  dt w t  dt  t  dt w t  dt
t BNL (4, i ) + t BNL (8, i ) + t BNL (7, i )
y x y
t  dt  t  dt u t  dt  t  dt u t  dt  t  dt v t  dt
t BL1(5,i )  t BNL (3, i )  t BNL (2, i ) + t BNL (6, i ) +
y z y
 t  dt v t  dt  t  dt w t  dt  t  dt w t  dt
B
t NL (5, i ) + B
t NL (9, i ) + z t BNL (8,i )
z y y
t  dt  t  dt u t  dt  t  dt u t  dt  t  dt v t  dt
t BL1(6,i )  B
t NL (1, i )  B
t NL (3, i ) + t BNL (4, i ) +
z x z
(3.175)
 t  dt v t  dt  t  dt w t  dt  t  dt w t  dt
t BNL (6, i ) + t BNL (7, i ) + t BNL (9, i )
x z x

ATENA Theory 179

 t  dt 1 hi
t BNL (1,1)  J1,1
r
t  dt
t BNL (1,2)  0
t  dt
t BNL (1,3)  0
t  dt
t BNL (1,4)  0
t  dt 1hi  t t t  dt tz s t t  dt sz  1 1 t t  dt sz 1 1 hi t t  dt tz
t BNL (1,5)  J1,1  ai Vi  bi Vi   J1,2 hi bi Vi  J1,3 ai Vi
r  2 2  2 2 r
t  dt 1 hi  t t t  dt t y s t t  dt s y  1 1 t t  dt s y 1 1 hi t t  dt t y
t BNL (1,6)  J1,1   ai Vi  bi Vi   J1,2 hi bi Vi  J1,3 ai Vi
r  2 2  2 2 r
t  dt 1 hi
t BNL (2,1)  J 2,1
r
t  dt
B
t NL (2,2)  0
t  dt
t BNL (2,3)  0
t  dt
t BNL (2,4)  0
t  dt 1hi  t t t  dt t z s t t  dt sz  1 1 t t  dt sz 1 1 hi t t  dt tz
t BNL (2,5)  J 2,1  ai Vi  bi Vi   J 2,2 hi bi Vi  J 2,3 ai Vi
r  2 2  2 2 r
t  dt 1 hi  t t t  dt t y s t t  dt s y  1 1 t t  dt s y 1 1 hi t t  dt t y
t BNL (2,6)  J 2,1   ai Vi  bi Vi   J 2,2 hi bi Vi  J 2,3 ai Vi
r  2 2  2 2 r
t  dt 1 hi
t BNL (3,1)  J 3,1
r
t  dt
t BNL (3,2)  0
t  dt
t BNL (3,3)  0
t  dt
t BNL (3,4)  0
t  dt 1 hi  t t t  dt tz s t t  dt sz  1 1 t t  dt sz 1 1 hi t t  dt tz
t BNL (3,5)  J 3,1  ai Vi  bi Vi   J 3,2 hi bi Vi  J 3,3 ai Vi
r  2 2  2 2 r
t  dt 1 hi  t t t  dt t y s t t  dt s y  1 1 t t  dt s y 1 1 hi t t  dt t y
t BNL (3,6)  J 3,1   ai Vi  bi Vi   J 3,2 hi bi Vi  J 3,3 ai Vi
r  2 2  2 2 r
t  dt
t BNL (4,1)  0
t  dt 1 hi
t BNL (4,2)  J1,1
r
t  dt
t BNL (4,3)  0
t  dt 1 hi  t t t  dt tz s t t  dt sx  1 1 t t  dt sz 1 1 hi t t  dt tz
t BNL (4,4)  J1,1   ai Vi  bi Vi   J1,2 hi bi Vi  J1,3 ai Vi
r  2 2  2 2 r
t  dt
t BNL (4,5)  0
t  dt 1 hi  t t t  dt tx s t t  dt sx  1 1 t t  dt sx 1 1 hi t t  dt tx
t BNL (4,6)  J1,1  ai Vi  bi Vi   J1,2 hi bi Vi  J1,3 ai Vi
r 2 2  2 2 r
t  dt
t BNL (5,1)  0
t  dt 1 hi
t BNL (5,2)  J 2,1
r
t  dt
t BNL (5,3)  0
t  dt 1 hi  t t t  dt tz s t t  dt sx  1 1 t t  dt sz 1 1 hi t t  dt tz
t BNL (5,4)  J 2,1   ai Vi  bi Vi   J 2,2 hi bi Vi  J 2,3 ai Vi
r  2 2  2 2 r
t  dt
t BNL (5,5)  0 (3.176)
t  dt 1 hi  t t t  dt tx s t t  dt sx  1 1 t t  dt sx 1 1 hi t t  dt t x
t BNL (5,6)  J1,1  ai Vi  bi Vi   J1,2 hi bi Vi  J1,3 ai Vi
r 2 2  2 2 r

180
 t  dt
t BNL (6,1)  0
t  dt 1 hi
t BNL (6,2)  J 3,1
r
t  dt
t BNL (6,3)  0
t  dt 1 hi  t t t  dt tz s t t  dt sx  1 1 t t  dt sz 1 1 hi t t  dt tz
t BNL (6,4)  J 3,1   ai Vi  bi Vi   J 3,2 hi bi Vi  J 3,3 ai Vi
r  2 2  2 2 r
t  dt
t BNL (6,5)  0
t  dt 1 hi  t t t  dt tx s t t  dt sx  1 1 t t  dt sx 1 1 hi t t  dt tx
t BNL (6,6)  J 3,1  ai Vi  bi Vi   J 3,2 hi bi Vi  J 3,3 ai Vi
r 2 2  2 2 r
t  dt
t BNL (7,1)  0
t  dt
t BNL (7,2)  0
t  dt 1 hi
t BNL (7,3)  J1,1
r
t  dt 1 hi  t t t  dt t y s t t  dt s y  1 1 t t  dt s y 1 1 hi t t  dt t y
t BNL (7,4)  J1,1  ai Vi  bi Vi   J1,2 hi bi Vi  J1,3 ai Vi
r  2 2  2 2 r
t  dt 1 hi  t t t  dt tx s t t  dt sx  1 1 t t  dt sx 1 1 hi t t  dt tx
t BNL (7,5)  J1,1   ai Vi  bi Vi   J1,2 hi bi Vi  J1,3 ai Vi
r  2 2  2 2 r
t  dt
t BNL (7,6)  0
t  dt
t BNL (8,1)  0
t  dt
t BNL (8,2)  0
t  dt 1 hi
t BNL (8,3)  J1,1
r
t  dt 1 hi  t t t  dt t y s t t  dt s y  1 1 t t  dt s y 1 1 hi t t  dt t y
t BNL (8,4)  J 2,1  ai Vi  bi Vi   J 2,2 hi bi Vi  J 2,3 ai Vi
r 2 2  2 2 r
t  dt 1 hi  t t t  dt tx s t t  dt sx  1 1 t t  dt sx 1 1 hi t t  dt tx
t BNL (8,5)  J 2,1   ai Vi  bi Vi   J 2,2 hi bi Vi  J 2,3 ai Vi
r  2 2  2 2 r
t  dt
t BNL (8,6)  0
t  dt
t BNL (9,1)  0
t  dt
t BNL (9,2)  0
t  dt 1 hi
t BNL (9,3)  J 3,1
r
t  dt 1 hi  t t t  dt t y s t t  dt s y  1 1 t t  dt s y 1 1 hi t t  dt t y (3.177)
t BNL (9,4)  J 3,1  ai Vi  bi Vi   J 3,2 hi bi Vi  J 3,3 ai Vi
r  2 2  2 2 r
t  dt 1 hi  t t t  dt tx s t t  dt sx  1 1 t t  dt sx 1 1 hi t t  dt tx
t BNL (9,5)  J 3,1   ai Vi  bi Vi   J 3,2 hi bi Vi  J 3,3 ai Vi
r  2 2  2 2 r
t  dt
t BNL (9,6)  0

ATENA Theory 181

 t  t
The stress matrix t Sij from (1.34) has he form:

 t  tt xx t  t

t xy
t  t

t xz

 

t  t
t yy t  t

t yz 
 t  t

t zz

 

t  t
t xx t  t

t xy 
t  t
t xz 
t  t
t Sij 
 t  t
 
t  t  (3.178)
 t yy t yz

 
t  t
t zz 
  xx  
t  xz 
t  t t  t t  t
sym.
 t t xy

 t  t

t yy
t  t
t  yz 
 
t  zz 
t  t


As already mentioned, stress-strain relations are calculated in r,s,t coordinate system, hence we
need equations for their transformations from global x,y,z coordinate system to the isoparametric
system with r,s,t coordinates and vice versa.
Let us denote t  dt T , t  dtT transformation matrices for strain and stress transformation from
global to isoparametric coordinate system, so that:
 t  tt xx 
 t  t 
 t  rr 
t  t  t  yy 
 t  t  t  dt  t  zz 
t  t

 t rs    T  
 t  rt 
t  t 
t  t
t  xy 
   t  t 
 t  tt yz 
 t  xz 
(3.179)
 t  tt xx 
 t  t 
 t  rr 
t  t  t  yy 
 t  t  t  dt  t  zz 
t  t

 t  rs   T  t  t 
 t  tt rt   t xy 
   t  t 
 t  tt yz 
 t  xz 
Then the transformation matrices are calculated by:


 t  dt V rx 2
    V rz 
ry 2 2 ry
t  dt t  dt
V 2t  dt V rx t  dt
V

t  dt  ry t  dt sy s ry t  dt
T   t  dt V rx t  dtV sx t  dt
V V t  dt
V rz t  dt
V sz t  dt
V rx t  dt
V y  t  dt V V sx
 t  dt V rx t  dtV tx t  dt
V
ry t  dt
V
ty t  dt
V rz t  dt
V sz t  dt
V rx t  dt t
V y  t  dt V
ry t  dt
V tx



182
 ... 2t  dt V
ry t  dt
V rz 2t  dt V rx  t  dt
V rz
t  dt ry t  dt sy

... V V sz  t  dt V rz t  dt
V t  dt rx t  dt s z
V V  t  dt V rz t  dtV sx  (3.180)
t  dt ry t  dt t  dt rz t  dt ty t  dt rx t  dt t z 
... V V tz
V V V V  t  dt V rz t  dtV tx 


 t  dt V rx 2
    V rz 
ry 2 2
t  dt t  dt
V

t  dt  ry t  dt sy
T   2t  dt V rx t  dtV sx 2t  dt V V 2t  dt V rz t  dt
V sz
 2t  dt V rx t  dtV tx 2t  dt V
ry t  dt
V
ty
2t  dt V rz t  dt
V sz



... t  dt
V
ry t  dt
V rz t  dt
V rx  t  dt
V rz
t  dt ry t  dt sy

... V V sz  t  dt V rz t  dt
V t  dt rx t  dt sz
V V  t  dt V rz t  dtV sx  (3.181)
t  dt ry t  dt t  dt rz t  dt ty t  dt rx t  dt t z 
... V V tz
V V V V  t  dt V rz t  dtV tx 

V s   t  dt V sx t  dt V y t  dt V sz  , t  dt V t   t  dt V tx t  dt V y t  dt V tz  are
t  dt s t
where vectors
vectors of unity length from Fig. 3-40. The remaining vector is calculated as a vector product of
the previous two vectors:
V r   t  dt V rx V rz  t  dt V s  t  dtV t
t  dt t  dt ry t  dt
V (3.182)

Inverse transformation matrices are calculated as:

t  dt
T1 t  dt TT
(3.183)
t  dt
T1 t  dt TT

3.17.4 The Element Integration

The element is integrated numerically. Along its longitudinal axis the element is integrated by
standard two to six nodes Gaussian integration. The table below lists r coordinates and
associated weights for utilized integration points:
Table 3-6: Gaussian integration of the beam element along the longitudinal axis

Integrat Coordinate r Weight

Number of
ion
integ. points
point

1 0.577350269189626 1.
2
2 -0.577350269189626 1.

ATENA Theory 183

 1 0.774596669241483 0.555555555555556
3 2 0. 0.888888888888889
3 -0.774596669241483 0.555555555555556
1 0.861136311594053 0.347854845137454
2 0.339981043584856 0.652145154862546
4
3 -0.339981043584856 0.652145154862546
4 0.861136311594053 0.347854845137454
1 0.906179845938664 0.236926885056189
2 0.538469310105683 0.478628670499366
5 3 0. 0.568888888888889
4 -0.538469310105683 0.478628670499366
5 -0.906179845938664 0.236926885056189
1 0.932469514203152 0.171324492379170
2 0.661209386466265 0.360761573048139
3 0.238619186083197 0.467913934572691
6
4 -0.238619186083197 0.467913934572691
5 -0.661209386466265 0.360761573048139
6 -0.932469514203152 0.171324492379170
In most cases the 2-nodes integration should be sufficient, for a higher order integration schemes
oscillatory shear stresses and forces may be observed along the length of the beam.
As for integration within the cross-section, i.e. in s,t coordinates, trapezoidal quadrature is used.
The element cross-section is subdivided into ns , nt “strips” as depicted in the following figure.
t

dtnt
2 s

dt2
dt1
ds1 dsns ind ividual “weight“
ds2 2 and m aterial

Fig. 3-41 The beam cross section integration

184
The integration is then carried out by summing functional values in center of all quadrilaterals
multiplied by their area.
Note that the element is integrated within the isoparametric coordinate system, hence we have to
use dx dy dz  det( J ) dr ds dt , see (3.171).
Nice feature of the ATENA’s implementation of the beam is that each of the quadrilaterals in a
cross section adopts an artificial input weight factor. By default, such a “weight” is equal to one,
however, if we set its value to zero, essentially a hole is introduced. This mechanism, together
with possibility of defining a customized material law in each of the quadrilaterals facilitates to
analyze beams that have a arbitrary shape of cross-sections.
The present beam implementation supports also smeared reinforcement. This is done in the same
way as it was for the Ahmad elements described in the previous section.

3.18 Curvilinear Nonlinear 3D Isoparametric Beam Element

CCIsoBeamBrick20_3, CCIsoBeamBrick12_3D and CCIsoBeamBrick8_3D are beam curved
isoparametric elements similar to the previous CCBeamNL_3D element. They use similar
geometry, node numbering etc., but differ from CCBeamNL_3D in that they account for all 6
components of 3D strains and stress vectors. They comply with all 3D static equations and no
additional static or kinematic constrains are imposed. The comparison of CCBeamNL_3D vs.
CCIsoBeamBrick12_3D resembles that of CCAhmad vs. CCIsoShell elements described above.
The CCIsoBeamBrick20_3, CCIsoBeamBrick12_3D and CCIsoBeamBrick8_3D are easy to use,
they preserve their 3D volume and they are nicely visualized during pre and post processing.
They can be input, loaded, and output in the same way as CCIsoBrick hexahedral elements.
CCIsoBeam8_3D features linear geometry and displacement approximation, (i.e. it has nodes
1...8, see the figure below), whilst CCIsoBeam12_3D has reduced quadratic approximation, (i.e.
it has nodes 1...12). CCIsoBeam20_3D comprises 20 nodes as shown in the sub-figure “Brick
nodes” below and it has full serendipity displacements approximation.

ATENA Theory 185

 Geometry

t
Vt
1

6
13
Vs
1
14 5 Brick nodes
7 16
s
18 15 8

19
r 17 6
20 5 Beam 3D nodes
7
10 8
2
9
1 11
10
12 9
3 11 4
12

2
1 t

3
4
w
z
z 2

2 s
2
y r
x
x
y v
u Isoparametric shape
Global coord. system and element dofs
Fig. 3-42 CCIsoBeamBrick12_3D and CCIsoBeamBrick8_3D elements

Shape of cross section can be any quadrilateral, i.e. it need not be only a rectangle as depicted
above. The elements are particularly useful for analyses of structures, where beam elements must
be combined with 3D solid and/or shell elements.
Derivation of the element is much the same as that for CCIsoShell element, i.e. Equations
(3.143) and (3.145) thru (3.165) remain valid. Geometry and displacement approximation
(3.144) is replaced by:
 1  s t k , front 1  s t k ,back  1  t t k ,top 1  t t k ,bot 
t
xi  hk  Xi  Xi  Xi  Xi 
 2 2  2 2 
 1  s t  t k , front (i 1) 1  s t t k ,back (i 1)  1  t t t k ,top (i 1) 1  t 
t t ( i 1)
xi  hk  Xi  Xi  Xi  t  t
X ik , bot (i 1)  (3.184)
 2 2  2 2 
t t (i )  1  ds t t 1  s t t k ,back (i )  1  t t t 1 t t  t 
xi  hk  X ik , front (i )  Xi  X ik ,top (i )  X ik ,bot (i ) 
 2 2  2 2 
hk  hk (r ) are 1D interpolation functions, see the interpolation function for CCIsoTruss
elements. The same notation is used for CCIsoShell Elements.

186
The element is calculated in integration points, (i.e. material points) that are located similar to
CCBeamNL_3D elements, refer to Fig. 3-41. The element can use any 3D material model.
Different materials can be specified for each material points, (or points in cross section). Some
of them can be used for modelling of embedded reinforcement. (Btw. discrete reinforcement can
be employed, too). The elements support both material and geometric nonlinearity.

3.19 Curvilinear Nonlinear 1D element

The elements CCIsoBeamBar<xx> and CCIsoBeamBar<xxx> are from the point of view of
mechanics nearly identical to the element described in Section 3.13, the difference being only in
that that these elements are specified by their axis as 1D beams. The first element has 2 nodes
(and uses linear interpolation of its geometry and displacements). The latter element has 3 nodes
(and uses quadratic interpolation of its geometry and displacements, which is identical to
CCBeamNL element referred above). The elements can be curved and can have variable height,
width and orientation of their cross section. All these parameters are input in CCBeam1D
geometry in form of algebraic expressions. The expression are functions of beam's coordinates
x,y,z. Similar to CCBeamNL element, these elements are also integrated by Gauss integration
along the beam's axis while grid quadrature is used for the remaining 2 directions (within cross
sections). The elements support embedded reinforcements, holes different materials in different
integration points etc. in the same way as it is the case of CCBeamNL element. They are suitable
for modeling of both shallow and deep beams. Note that CCIsoBeamBar<xx> has far worse
properties compared to CCIsoBeamBar<xxx>. Hence, the linear element should be used only to
model some links and connections within the structures.
2
Beam 1D nodes

3 t

w
z
1 z 2

2 s
2
y r
x
x
y v
u Isoparametric shape

Global coord. system and element dofs

Fig. 3-43 CCIsoBeamNLBar<xxx> element

3.19.1 Connection of the beam1D to an ambient solid element

The procedure to connect beam1D's dofs to an ambient element is like that for shell2D elements,
see 3.13.2. Again, it consists of two parts:
1. fix a FE node with [u, v, w] displacement within the beam1D element,

ATENA Theory 187

 2. fix three rotation dofs of the beam1D element within ambient elements

3.19.1.1 Fixing a FE node with [u, v, w] displacement within the beam1D element

Using (3.169) and (3.170) write expression for beam1D displacements at the top uitop and bottom
uibot , i.e. s  0, t  1 of a cross section. Do the same for right uiright and left uileft point, i.e.
s  1, t  0 .

Write 3D solid approximation for the same 4 nodes. Then, if we compare the 1D and 3D
approximation, after some mathematical manipulation we derive
hhk14 p 58  hhk (r , s  1, t  1)  hhk (r , s  1, t  1)  hhk ( r , s  1, t  1)  hhk (r , s  1, t  1)
hhk14 m 58  hhk (r , s  1, t  1)  hhk (r , s  1, t  1)  hhk (r , s  1, t  1)  hhk (r , s  1, t  1)
hhk15 m 48  hhk (r , s  1, t  1)  hhk (r , s  1, t  1)  hhk (r , s  1, t  1)  hhk (r , s  1, t  1)

 14 p 58 ak 14 m 58 bk 15 m 48   uk  (3.185)
 hhk 0 0 0
2
hhk 
2
hhk   vk 
 u1   
u    0 ak 14 m58   wk 
 2  
hhk14 p 58 0 
2
hhk 0 0
  x,k
  
k
 u3  
 0 0 hhk14 p 58 hhk15 m 48 0 0  y,k 

   
 z,k 

3.19.1.2 Fixing three rotation dofs of the beam3D element within ambient elements
Similarly, to the expressions for shell2D the resulting equations for beam1D rotation  x , y , z
are

 x 
   MM UU top UU top UU top UU bot UU bot UU bot UU right UU right UU right UU left UU left UU left T
 y  1 2 3 1 2 3 1 2 3 1 2 3 

 z 
(3.186)

188
  Vr Vs VsxVry VrxVs y VsxVrz VrxVsz 
 1/ 2 x x 1 / 2  1 / 2  
 a a a a a 
 VrxVs y VsxVry Vr Vs Vs yVrz VryVsz 
 1 / 2 a  a 1/ 2 y y 1 / 2 
 a a a 
 VrxVsz VsxVrz VryVsz Vs yVrz Vr Vs 
 1 / 2 a  a 1 / 2
a

a
1/ 2 z z
a 
 
 Vr Vs Vs Vr Vr Vs VsxVrz VrxVsz 
 1 / 2 x x 1/ 2 x y  x y 1/ 2 
a a a a a 
 
 1 / 2VrxVs y  VsxVry 1 / 2
VryVs y
1/ 2
Vs yVrz VryVsz 

 a a a a a 
 
 1 / 2VrxVsz  VsxVrz Vr Vs Vs Vr VrzVsz 
1/ 2 y z  y z 1 / 2
 a a a a a 
MM T   
 Vr Vt Vt Vr Vr Vt Vt xVrz VrxVt z 
1 / 2 x x 1/ 2 x y  x y 1/ 2 
 b b b b b 
 
 1 / 2VrxVt y  Vt xVry 1 / 2
VryVt y
1/ 2
Vt yVrz VryVt z 

 b b b b b 
 
 1 / 2VrxVt z  Vt xVrz Vr Vt Vt Vr VrzVt z 
1/ 2 y z  y z 1 / 2
 b b b b b 
 Vt xVry VrxVt y 
 Vr Vt Vt xVrz VrxVt z 
1/ 2 x x 1 / 2  1 / 2 
 b b b b b  (3.187)
 Vr Vt Vt Vr Vr Vt Vt yVrz VryVt z 
 1 / 2 x y  x y 1/ 2 y y 1 / 2  
 b b b b b 
 VryVt z Vt yVrz 
 1 / 2VrxVt z  Vt xVrz 1 / 2 
Vr Vt
1/ 2 z z 
 b b b b b 

where Vrk , i  Vi rk Vsk , i  Vi sk , a  ak , b  bk

Note that displacement dofs are fixed by(3.185).

If either bottom or top node gets outside the ambient element, the middle point is used instead.
Equations (3.186) and (3.187) are still valid but it is necessary to use
MM  (i, j )  1 MM (i, j ), j  1..6, MM
  (i, j )  MM (i, j ), j  7..12 to calculate    T .
2  x y z

Similarly, if either right or bottom node gets outside the ambient element, the middle point is
used instead. Then, it is necessary to use
  (i, j )  MM (i, j ), j  1..6, MM 1
  (i, j )  MM (i, j ), j  7..12 to calculate     . T
MM  x y z
2

3.20 Integrated forces and moments for shells

Integrated forces for shells are computed as follows:

ATENA Theory 189

 t/2
Nx'    x ' x ' dz '
t / 2
t/2
Ny'    y ' y ' dz '
t / 2
t/2
Nz '    z ' z ' dz '
t / 2
t/2
Qx ' z '    x ' z ' dz '
t / 2
t/2
Qx ' y '    dz ' (3.188)
t / 2 x ' y '
t/2
Qy ' z '    y ' z ' dz '
t / 2
t/2
M y'    x ' x ' z dz '
t / 2
t/2
Mx'    y ' y ' ( z ) dz '
t / 2
t/2
Kx' y '    ( z ) dz '
t / 2 x ' y '

The above forces and moments act on planes indicated below:

y ' z ' :  N x ' Qx ' y ' Qx ' z ' Kx' y ' M y' K x ' z '  0 
x ' z ' : Qy ' x '  Qx ' y ' N y ' Qy ' z ' M x' K y ' x '  Kx ' y ' K y ' z '  0 
x ' y ' : Qz ' x '  Qx ' z ' Qz ' y '  Qy ' z ' Nz ' K z ' x '  0 K z ' y '  0 M z '  0 

The actual values of the forces and moments are calculated by extrapolation of stresses from IPs
into finite element nodes, (please refer to Section "Extrapolation of Stress and Strain to Element
Nodes" in Chapter CONTINUUM GOVERNING EQUATIONS. The process is as follows:

Let us take an example of N x ' that is calculated by integration of  x ' x ' thru element's thickness.

The stress  x ' x ' at element nodes is extrapolated from stresses in IPs ˆ x ' x ' by

 x ' x '   M  Px ' x '

inv

Pxx , i   hi ˆ x ' x ' dVe (3.189)

Ve

M ij   hi h j dVe
Ve

where Ve stands for element volume. Using (3.188) and writing (3.189) for extrapolation within
shell mid-plane  e , (i.e. integration over  e instead of Ve ) we can write

N x ' x '   MM  PPx ' x '

inv

PPxx , i  
e  t/2

t / 2 
hi ˆ x ' x 'dz d e   hi ˆ x ' x ' dVe
Ve
(3.190)

1
MM ij   hi hj d e   hi hj dVe
e Ve t

190
 where t  t (r , s ) is element thickness at r,s. The integration for extrapolation is carried out over
 e , because the forces and moments are the same through shell thickness. Note that
h  h (r , s ) is interpolation function in the shell mid-plane and it is independent of t coordinate,
k k

(unlike hi  hi (r , s, t ) in (3.189)). Therefore, we can write, (see the last equation in (3.190):

Ve
hi hj dVe  
e  t/2

t / 2 i 
h (r , s ) hj (r , s )dt d e

h (r , s ) h (r , s )   dt  d 
t/2
 i j e
e t / 2
(3.191)
 hi (r , s ) hj (r , s ) t (r , s) d  e   hi hj t d  e
e e

1 
 h h dVe   hi hj d e  MM ij
Ve t i j e

3.21 Integrated forces and moments for beams

Integrated forces for beams are computed as follows:
t/2
Nx'    x ' x ' dy ' dz '
t / 2
t/2
Qx ' y '    x ' y ' dy ' dz '
t / 2
t/2
Qx ' z '    dy ' dz '
t / 2 x ' z '
t/2
(3.192)
Ky'z'   ( x ' y ' ( z ')   x ' z ' y ') dy ' dz '
t / 2
t/2
M y'   ( x ' x ' z ') dy ' dz '
t / 2
t/2
Mz'    x ' x ' ( y ') dy ' dz '
t / 2

The forces and moments act on the plane (x'y'). They are calculated similar way to (3.190),
1
however, MM ij   hi h j dr   hi h j dVe , where b h is area of the beam's cross section and
le Ve b h

le is element length.

3.22 Global and Local Coordinate Systems for Element Load

Most element loads can be defined in global or local coordinate system. Global coordinate
system is always available, hence using it is usually the safest way to input a desired element
load. Nevertheless, some elements are internally defined in a local coordinate system and it can
be employed for an element load definition, too. Location of such a local system, (if it exists) has
been described together with description of the associated finite element. For example, local
coordinate systems are defined for plane 3D isoparametric elements, shell, and beam elements
etc. On the other hand, elements such as tetrahedrons, bricks and others are defined in directly in

ATENA Theory 191

global coordinate system and therefore a local element load is treated as if it were input as a
global element load.
An exception to the above are truss elements. Although they are defined in global coordinate
system, they do support local element load. Their local coordinate system (for element loading
only) is defined as follows:
 local X axis points in direction of the truss element,
 local Y axis is normal to local X axis and lies in the global XY plane,
 its positive orientation is chosen so that the local X and local Y forms a right-hand (2D)
coordinate system in the plane defined by these local axes,
 local Z axis is vector product of the local X and local Y axes, (for 3D case only).
 if the truss is parallel to global z, then local X points in direction of global Z, local Y
coincides with global Y and local Z has opposite direction of the global X, (for 3D case
only).

2D 3D

ZG
YG N2
YG
YL
N1

N1 XL YL
YG
XG

XL N2
XG

Fig. 3-44 Local and global coordinate systems for truss element N1-N2, (e.g. loaded element edge)
Specification of a boundary load deserves slightly more attention. Firstly, it is applied only to an
element’s edge or an element’s surface, (see also the note below), as opposed to e.g. an element
body load that is for the whole element. Local coordinate system is thus defined by location of
the loaded edge or surface. Secondly, a boundary load definition must include a reference to a
selection, which contains nodes to be loaded. Their order in the list is irrelevant, as what really
matters is the order in which they appear in the element incidences. When processing a boundary
load, ATENA loops thru all element’s surfaces and edges, (in the order specified in the table
below) and checks appropriate incidental nodes. If the tested node is present in the list of loaded
boundary nodes, it is picked up and put into incidences of a new planar or line element. This
element is later used to process the boundary load. It is its local coordinate system, that is
(possibly) used to deal with local/global load transformations.
The table below defines the orders, in which element surfaces and edges are tested for a surface
or edge element load. (It is assumed that element incidences are (n1 , n2 , ... nnum _ elem _ nodes ) ). It
describes linear elements, but surfaces and edges of nonlinear elements are treated in the same
order.

192
Table 3-7: Order of element surface and nodes as they are tested within a boundary load definition.

Element shape Type Surface/node incidences

Truss Edge (n1 , n2 )
Triangle Surface (n1 , n2 , n3 )
Edge (n1 , n2 ); (n2 , n3 ); (n3 , n1 )
Quad Surface (n1 , n2 , n3 , n4 )
Edge (n1 , n2 ); (n2 , n3 ); (n3 , n1 ); (n4 , n1 )
Hexahedron, Surface (n1 , n2 , n3 , n4 ); (n5 , n6 , n7 , n8 ); (n1 , n4 , n8 , n5 ); (n2 , n3 , n7 , n6 );
(brick) (n1 , n2 , n6 , n5 ); ( n4 , n3 , n7 , n8 );
Edge (n1 , n2 ); (n2 , n3 ); ( n3 , n4 ); ( n4 , n1 );
(n5 , n6 ); (n6 , n7 ); (n7 , n8 ); (n8 , n5 );
(n1 , n5 ); (n2 , n6 ); ( n3 , n7 ); (n4 , n8 )
Tetrahedron Surface (n1 , n2 , n3 ); (n1 , n2 , n4 ); (n1 , n3 , n4 ); (n2 , n3 , n4 )
Edge (n1 , n2 ); (n2 , n3 ); (n3 , n1 );
(n4 , n1 ); (n4 , n2 ); (n4 , n3 )
Wedge Surface (n1 , n2 , n3 ); (n4 , n5 , n6 );
(n1 , n2 , n5 , n4 ); (n6 , n5 , n2 , n3 ); (n4 , n6 , n3 , n1 )
Edge (n1 , n2 ); (n2 , n3 ); (n3 , n1 );
(n4 , n5 ); (n5 , n6 ); (n6 , n4 );
(n1 , n4 ); (n2 , n5 ); (n3 , n6 );

Note that only one surface or one edge of each element can be loaded in a single boundary load
specification. If more element’s surfaces or edges are to be loaded, use more boundary load
definitions. Violation of this rule causes an error report and skipping of the offending boundary
load.

ATENA Theory 193

 3D edge load
2D edge load planar element
planar element

n1 ZG n1 XL ZL
YG XL

YL
YL n2 n3
n2
n3
YG
XG
XG

3D edge load
solid element 3D surface load
solid element
n1 X L Z L n1
ZG ZL
ZG
YL
YL
n3 XL
n2 n2 n3

n6 n6
n5 YG YG
XG n5
XG

Fig. 3-45 Examples of positioning local coordinate system used by surface and element load for 2D
and 3D elements
Transport analysis does not distinguish between local and global element loads. Hence, a local
element “load” is treated as being a global load. The actual load value is always scalar, (unlike
vectors in statics) and it is assumed positive for flow out of the element.

3.23 Digital printing of concrete structures4

Digital 3D printing of concrete and reinforced concrete structures seems to be an innovated,
progressive, and economically effective method for building civil engineering structures in
future. It has several advantages in comparison to the traditional methods in building industry.

4
Not available in ATENA version 5.7.0 and older

194
For example, it allows for miscellaneous shapes of the structures, so that they can be designed
more favorably for their static and functional behavior, architectonic design etc. It enables better
optimization of the structures resulting in reduced cost, less labor-intensity, less waste produced,
greater integration of function and increased speed of the whole construction process. Although
most printing methods have not yet showed their full potentials, most engineers agree that they
are the right way for civil engineering in near future, because they contribute to better design of
the structures and their higher industrialization.
There exists a variety of 3D printing methods used at construction scale, name e.g. 3D extrusion,
powder/particle bed printing, 3D block assembling, spraying etc. This Section presents ATENA
support for analyses of printed structures using 3D extrusion and describes, how such a
construction process can be modelled by this software. It is characterized by printing the
structure by layers, i.e. pressing concrete mix thru the nuzzle moving alongside a stepwise linear
polygon line that corresponds to individual walls of the structure. Often, some walls are too wide
to be printed by one pass of the nuzzle and two or more (parallel) printing passes are needed.
Once the current layer has been completed, the printing head returns to its origin, moves one
layer upwards and starts printing next layer until full height walls of the structure is produced.

3.23.1 Simplified strength and stability assessments of extruded structures

This section brings preliminary considerations and requirements that should be addressed in
design and fabrication of extruded concrete structures. Some derivations below are inspired by
papers (Roussel 2018) and (Wolfs at.al. 2018).

3.23.1.1 Material model for stability assessments

Material behavior used for digital fabrication of concrete structures can be modelled by
viscoplastic and elastoplastic materials. The former model is suitable for times when the material
is being pumped and is flowing to a place of its final position. This time period is not addressed
here. We will rather concentrate on the later times, when the material is still fresh, but it is
already in rest. At that time, the material features approximately elastoplastic behavior.
There exist several kinds of yield surfaces that define threshold between elastic and fully plastic
behavior. Using a few material parameters that are typically obtained from laboratory tests they
define general 3D stress-strain conditions when the material start to yield. Uniaxial tensile
strength, shear tensile strength etc. are examples of such parameters.
Stress-strain conditions in printed walls are close to 1D conditions, (with self-weight body load
only) and thus, throughout all the derivation here we assume 1D elastic behavior up to the
material compression f c . Nevertheless, as some people prefer to measure and use the material
shear strength f sh , we will show how to convert 1D material strength f1D to f sh and vice versa.
Using e.g. Mises yield surface, https://en.wikipedia.org/wiki/Von_Mises_yield_criterion

1 
 v  3 J 2    11   22    22   33    33   11   6  232   312   122   
2 2 2
(3.193)
2 

calculate equivalent von Mises stress  v ( J 2 is the second invariant of stress deviator tensor) for
uniaxial test conditions f1D   11  0,  ij  0 for (i  1)  ( j  1) and pure shear test conditions

ATENA Theory 195

 f sh   12   21  0, otherwise  ij  0 . By comparing the corresponding equivalent von Mises
stresses, we get the required strength conversion formula:
 v   11  f1D
 v  3  12  3 f sh (3.194)
f1D  3 f sh
Another option is to use maximum shear stress theory, see
http://thegateacademy.com/files/wppdf/Theories-of-failure.pdf. It defines yield surface by
constraining maximum shear
 1   2 2  3  3  1 
 sh,max  max abs( ), abs( ), abs( ) (3.195)
 2 2 2 
where  1 , 2 , 3 are principal stresses. Substituting the above two stress test conditions in
(3.195) we get
 11f1D
 1   11 , 2   3  0   sh,max  
2 2
  ( 12 )
 1  0, 2   12 , 3   12   sh ,max  12   12  f sh (3.196)
2
f1D  2 f sh

Total strain theory postulates, see also the above reference:

 tst   12   22   32  2  1 2   2 3   1 3  (3.197)

Then
 1   11 , 2   3  0   tst   11  f1D
 1  0, 2   12 , 3   12   tst  2 1     12  2 1    f sh (3.198)
f1D  2 1    f sh

Of course, a more elaborate and precise yield surface can be employed but we believe that for
the preliminary assessment the above simple expressions serve enough accuracy. After all, in
ATENA computer analyses one can use any material model suitable for cementitious material. It
is more accurate but at the same time also computationally expensive.

3.23.1.2 Strength-based stability of an individual layer

Let as assume a simplified time development of material yield stress f c (t )

f c (t )  min( f c ,0  fc t , f c ,max ) (3.199)

196
where f c ,0 is yield stress at time t  0 , (i.e. initial value just after material depositing), f c ,max is
maximum f and f is structuration rate. The layer is loaded primarily by its gravity self-
c c

weight and therefore

f c ,0  h  g (3.200)
Maximum height of one printed layer is

f c ,0
h (3.201)
g

If surface tension  is considered, it produces stresses of order  st  . Comparing with
h
(3.201) we get

h g 
h
(3.202)

h
g

Pa 0.1
For example, for   0.1 , (=water) we calculate h   0.002m . Therefore, for
m 2300 10
printed structures stability contribution of f c ,0 is more important than contribution of surface
tension.

3.23.1.3 Collective strength-based stability of more layers

If we consider the case of several printed layers, the lowest layer must resist vertical load H  g ,
where H is total height of the structure.
 11  H  g  f c
f c  f c ,0  fc t  H  g  vv t  g (3.203)
f
fc  vv  g  c ,0
t
where vv is vertical printing speed. The last expression in (3.203) states minimum structuration
rate f for being able to print the top layer at time t.
The total time ttot for printing height H of the structure is

ATENA Theory 197

 l
H
H H H tl vh H l
ttot      (3.204)
vv h h h vh h
tl
In the above tl is time to print a single layer, i.e. time necessary for printing head’s move along
the printing polygon that has total length l.

3.23.1.4 Collective plastic collapse criterion implemented in ATENA

This section describes steps that are executed to estimate plasticity-based criterion in ATENA.
The procedure is inspired by (Suiker 2020) presentation at DC2020 conference in Eindhoven in
2020.

3.23.1.4.1 Linear material curing function

The stability criterion is similar to that presented in the previous section; however, it is expressed
in slightly different form. It assumes linear material curing function, i.e.

 p (t )   p ,0 (1   t ) (3.205)

where  p (t ) is material yield strength at time t,  p ,0 is its initial value at t  0 and  represents
material linear curing rate of the yield stress.
Vertical stress  v at the bottom of the wall is, (H is the wall height,  is concrete density and g
states for gravity acceleration)
v  H  g (3.206)
and we require
 p  v (3.207)

For the following derivation, lets introduce dimensionless parameter

 p ,0
   (3.208)
 gvv

Note that vertical printing speed vv is in the (Suiker 2020) paper (and Atena) denoted as l .
Substituting H  vv t into (3.206) we get

 p   p ,0 * (1   t )  vv  g (3.209)

198
After some mathematical manipulation it yields

1
lp 
1  
(3.210)
 p ,0
lp  lp
g

where l p is the maximum wall height before the collapse.

 p
If  p ,0  0 , then the wall is stable for  vv  g , i.e. the case, when time rate of increase of
t
material strength is higher than the rate of increase of vertical stress during printing of the wall.
This condition also indicates unlimited wall height.
The paper (Suiker 2020) also discusses, how to calculate  p . For the case of pressure-
dependent shear failure, they recommend Mohr-Coulomb theory

2c cos( )
p  (3.211)
1  K  (1  K )sin( )
In the above  is material frictional angle, c states for material cohesion and K  min( K y , K z )
is minimum of coefficient of lateral stresses K y   y /  x , K z   z /  x , (axis x is vertical, axes
y,z are lateral, i.e. horizontal.
Substituting (3.211) into (3.205) yields

  p   p c   1   p   p c  
 p   p ,0 (1   pt )   p ,0     t   
p ,0  1     t  (3.212)
  t c t    p ,0   t c t  

From the above

1   p   p c 
p     (3.213)
 p ,0   t c t 

where
 p 2c sin( ) 2c cos( ) 2 (1  K )
 
 1  K  (1  K )sin( ) (1  K  (1  K )sin( )) 2
(3.214)
 p 2cos( )

c 1  K  (1  K )sin( )

3.23.1.4.2 Exponential material curing function

ATENA Theory 199

This section describes a similar stability assessment; however, exponential decaying curing
process is assumed now. This means that Eqn. (3.205) changes to
 p (t )   p ,0 ( p  (1   p ) e t ) (3.215)

 p ( )
where  p  and  is now coefficient of compression strength exponential curing rate.
 p ,0

Substituting (3.215) into (3.206) and (3.207) yields

1     p 0   p  1    p  p 0  
tcollapse   gvv  W   e gvv      p  p 0  (3.216)
 vv g   gvv   
   

W(z) states for Lambert W(z) function. The maximal wall height at collapse is
l p  tcollapsevv (3.217)

3.23.1.5 Buckling stability

Buckling stability of the printed structures may limit the structure even more that strength-base
stability. It is computed using Euler Buckling Theory, see
http://www.continuummechanics.org/columnbuckling.html. Let us start our derivation with
classic beam bending equation that reads
E I u  M (3.218)
where E, I state for Young modulus and quadratic moment of inertia, x is longitudinal coordinate
d 2u
of the beam with its origin at the bottom, u  u ( x) is deformation and u  2 is its second
dx
derivation with respect to x. M is loading moment. Let us assume 1m long section of the wall. It
can be modelled by a vertical beam supported at the bottom and loaded by a vertical force P at
its top, i.e. M   Pu . Solving differential equation (3.218) yields
 P   P 
u  A sin  x   B cos  x  (3.219)
 EI   EI 
A, B are two constants to be solved from the beam’s boundary conditions u (0)  u ( H )  0 and
u (0)  u ( H )  0 . It yields B  0 and when looking for a nontrivial solution, we get
P
H   , from which we derive the well-known final expression for critical force
EI

 2E I
P (3.220)
H2

200
The same applies for boundary conditions u (0)  u (0)  u ( H )  0 and u ( H )  0 . For a general
case
 2 EI
P (3.221)
(kH ) 2

Substituting P  H  g A , (A is cross section of the 1m long wall section), we get

1 3
 EI
2  2E
1w
 2 E w2
H2   12 
Pk 2 k 2 H  g 1 w k 212 H  g (3.222)
1  Ew 2 2
H 3
k 12  g
Equation (3.222) states critical height of a printed wall to prevent its collapse due to losing
stability. W states for the wall width.

Finally, using (3.222) and (3.203) calculate a threshold H, below which the strength-based
stability criterion (3.203) is dominant whilst above it the buckling limit is more restrictive.

1  2 E w2 1  2 E w2
H 3 
k 12  g k 3 12 f c
H (3.223)
w E
H
k 12 f c

3.23.1.6 Elastic buckling collapse criterion implemented in ATENA

The paper by (Suiker 2020), (Suiker 2018) also presents an estimation of elastic buckling
stability of the printed walls. It is more accurate than the criterion from the previous section
because it allows for clamp or simple support boundary conditions along the wall vertical edges.

3.23.1.6.1 Linear material curing function

Like 3.23.1.4.1 the material linear curing rate is assumed

E (t )  E0 (1   E )t (3.224)

where E (t ) is material Young modulus at time t, E0 is its initial value at t  0 and

 E represents material linear curing rate of elasticity modules.

ATENA Theory 201

The employed method is in detail derived in (Suiker 2018). It presents a semi numerical-
analytical solution expressed in forms of easily useable plots. The recommended procedure is
implemented in ATENA.

The solution uses three dimensionless parameters

 gh
lcr  3 lcr
D0
 gh
bcr  3 b (3.225)
D0
D0  E
E  3
 gh vv
In the above equations lcr , b, h is critical buckling height, horizontal length, (i.e. width), and
thickness of the wall, respectively. Vertical printing speed is:
q t
vv  l   l (3.226)
vn hTl Tl
with q  vn h tl being the material volume discharged from the printing nozzle per unit time, Tl is
the period required for printing an individual layer and tl is height, (i.e. thickness) of the printed
layer, see the figure below

Fig. 3-46 The buckling wall

D0 states for initial wall bending stiffness defined by

202
 E0 h3
D0  (3.227)
12(1   2 )

where E0 is initial Young modulus and  is Poisson ratio of the material.

The procedure to calculate critical wall height lcr is as follows
1. Calculate D0 , (3.227).

2. Calculate  E , bcr , (3.225).

3. For the particular support conditions along vertical edges of the wall use
Fig. 3-47 and find lcr that corresponds to the above  E , bcr .

4. Using inverse of the expression for lcr calculate lcr , (3.225).

If the printed wall is not supported along its vertical edges, use the dash line for free wall in Fig.
3-47, (i.e. for bcr   ). The dash lines for the case of clamped and simply supported wall yield
the same the same lcr .

Alternatively, (Suiker 2020) recommends lcr  1.98635  0.996  E 0.793 .

ATENA Theory 203

Fig. 3-47 Critical dimensionless buckling length versus dimensionless linear curing rate for the case
of fully clamped and simply supported wall.

3.23.1.6.2 Exponential material curing function

This section provides solution for buckling stability subject to exponential material curing rate
E (t )  E0 ( E  (1   E )e  E t ) (3.228)
E ( )
Notation used is similar to the above, i.e.  E  and  E is now coefficient of exponential
E0
Young modulus curing rate.

204
The overall solution is the same as it was for the case of linear curing, only instead of Fig. 3-47
the solution with exponential curing rate requires to use plots Fig. 48 thru Fig. 50. These plots
also comes from (Suiker 2018).

Fig. 48 Critical dimensionless buckling length versus dimensionless exponential curing rate for the
case of free wall.

Note that (Suiker 2018) provides the above plots only for  E  {2..10} . It is sufficient for
modelling some laboratory experiments, but practical analyses typically require values of  P
much higher.

ATENA Theory 205

Fig. 49 Critical dimensionless buckling length versus dimensionless exponential curing rate for the
case of simply supported wall.

206
Fig. 50 Critical dimensionless buckling length versus dimensionless exponential curing rate for the
case of fully clamped wall.

ATENA Theory 207

3.23.1.7 Comparison of buckling stability results provided in Section3.23.1.5 and
Section 3.23.1.6 – linear material curing rate

If we assume f c   p , then the previously presented strength stability criteria in Section 3.23.1.3
and Section 3.23.1.4 yield the same results. However, the buckling stability criterion in Section
3.23.1.6 is more sophisticated than that from Section 3.23.1.5. It is mainly improved in that it
can account for additional boundary conditions along the printed wall’s vertical edges.
Nevertheless, for the case of unsupported, (i.e. free) vertical edges the two models should yield
similar results. This is checked here.

Using Young modulus from (3.224) and vertical printing speed vv from (3.226) we can write,
(see Section 3.23.1.5)
1  2 E0 (1   E )t w2
H 3  vv t (3.229)
k 12  g
where the wall width w  h , see Fig. 3-46. Solving the above equation for t yields critical wall
height lcr

E0 Tl 3 h5vn 3 E 3
  3 4.5 1014 h 2 E0 q 3 g 2 k 3  2  h 2 E0 q 2 g 2 k 2  2 2.4649 1028  2.025 1029 q 2 k 2
qgk 
vn hTl Tl 2 h 4 vn 2 E0  E 1
t  0.000009700895963 2 2   28232.21472
Q gk  Qk 
tQ
lcr  t vv 
vn hTl
(3.230)

Example: substituting wall parameters

208
 b
Tl 
vn
E0  48500 Pa
1
 E  0.000895
sec
m3
q  54427  0.0013
sec
kg
  2100
m3
6.25 m
vn 
60 sec
m
g  10. 2
sec
k 1
h  0.055 m
b 1 m (3.231)

the expression (3.230) and (3.225) calculates critical wall height 0.188m an 0.1825m
respectively. For the case of h  0.1m the expression (3.230) and (3.225) results in 0.305m and
0.292 m.

3.23.2 Steps to carry on analyses of extruded structures

A typical analysis of a structure built by 3D extrusion slightly differs from usual analyses. All
the required steps are now described:

Step 1. Prepare a FE model of the structure neglecting the printing process:

The analysis starts by creating a full FE models whereby the process of the printing is ignored. It
means that we model the final geometry, properties, and conditions of the structure. Any
available FE preprocessor can be used to achieve the goal. Use appropriate (time independent)
material model and supply parameters that correspond to the final (long age) material properties.

Step 2. Calculate time of construction ticonstr of each part of the structure, i.e. for each individual
element:
Use ATENA UPDATE_ELEMENT_CONSTRUCT_TIME command to accomplish this step. It
requires the following data:
 List of element groups that are printed. It is assumed that all elements of the groups are
constructed in this way. Actual group’s ids are entered via an ATENA selection list.
 Horizontal velocity of the printing head vh , about 1-10 cm/s.

ATENA Theory 209

  Thickness of one printed layer h, usually 1-10 cm.
 Width of the printed layer w, typically 5-25 cm
 Vector of vertical move from one layer to the next layer n .
 Track polygon of the printing head’s motion. It is specified as an ATENA selection
containing ids of FE nodes thru which the printing head passes. The track consists of any
number of linear segments. If some segments are not mutually connected, i.e. the track is
broken, separate the corresponding segments by inserting id=0 between their adjacent
end nodes.
 Set start time tstart of the track polygon. Typically, tstart  0 , however if the structure is
printed using several track polygons (with e.g. different width), then tstart of the current
polygon equals to time corresponding to the last point of the previous polygon.

Having all the above it follows to calculate time ticonstr of each element. Let PEC   x y z  are
T

coordinates of center of the element. The element is printed when the head is at the closest
position. The track polygon of the moving head is input by setting location of its bottom right
edge. Hence, in the following derivations we work with a point P, (instead of PEC ):

 x   h nx  wvx  / 2 
 
P   y   h n y  wv y  / 2  (3.232)
 
 z   h nz  wvz  / 2 

210
 PEC

w
h P

R
B


n Q
 
v AB

Fig. 3-51 Calculation of age of a particular printed element.

Element construction time is calculated as follows:

 
 AB  AP
PQ  
AB

 AP  PQ 
  2  2
AQ 

   AB
Q  A  AQ  (3.233)
AB
 
    n PQ 
QR  cos( PQ, n) PQ    PQ
n  PQ

   n
R  Q  QR 
n

ATENA Theory 211

   
The element is printed as a part of a segment AB , if point Q  AB and its distance
 
 
PR  h / 2 . It is printed in a layer id l  int QR / t  1 and has construction time

AR
ticonstr   l  1 tlayer  t prev _ segs  tcur _ seg  (3.234)
AB

where tlayer is total time to print one layer, i.e. its length divided by vh , t prev _ segs is time to print

element in the current layer up to point A and tcur _ seg is time to print the current segment AB .
The symbol  and  stand for cross and dot product, respectively. The remaining symbols in
the equations are depicted in Fig. 3-104.

3. Account for construction time ticonstr during the analysis:

ATENA calculates structures step by step. Each step has its time t and it stepwise increases.
When executing an analysis step, its time is compared with ticonstr of each printed element. If
t  ticonstr , then the element’s contribution is assembled as usually, i.e. at its full values. For
elements with t  ticonstr ATENA offers two options:

 The element is calculated as usually, i.e. neglecting its ticonstr . It yields unreduced stresses
(corresponding to deformation), vector of element forces and matrix of element stiffness.
However, before their assembly into global data structures, the vector and matrix is
multiplied by a reduction coefficient   1 . This simulates that the element does not yet
exist. The coefficient is defined by ATENA command
NEGLIGIBLE_ELEMENT_CONTRIBUTION_COEFF  . If   0 , the element does
not contribute at all.
Although this approach is simple, it has several disadvantages: it is computationally
inefficient because it calculates at each time step all elements despite their contribution to
the whole structure is possibly later minimized by the coefficient  . The next
disadvantage is that it involves some element forces’ redistribution, (i.e. some additional
iterations), when the element transfers from t  ticonstr to t  ticonstr status. Note that it
happens in spite of ATENA uses incremental solution technique.
As discussed previously, the stresses are computed always in full value, i.e. neglecting
ticonstr . Now at t  ticonstr we calculate element forces by something like
T T
Fi    i B dV    i 1  Ei  i  B dV . If the structures does not exhibit any
i i
T
deformation increment at the current step, then  i  0  Fi    i 1B dV , which is
i
T
differs from what we used in the previous step, (= Fi 1     i 1B dV )!
i
On the other hand, this solution approach simulates better the case, when we require print
layers having a constant height, (although not quite exactly).
 The second method is to mark all elements active only on condition t  ticonstr . Use an
ATENA command something like
SELECTION "SOLID_BOX_ELEMENTS"
CONSTRUCT_TIME_DEPENDENT_ACTIVE GROUP 1

212
 It ensures that elements with t  ticonstr are skipped. They are not computed, not assembled,
they don’t contribute the structure. They also do not deform, unless dictated by their
adjacent elements. This solution is more effective because it calculates only “printed”
parts of the structure. Also, no additional iterations are needed. It corresponds to the case
when we keep constant top position of each layer, (while its height slightly increases).
This method is preferable over the previous one.

4. Account for time dependent material behavior

For this kind of analysis, it is essential to use a material model whose properties vary in time.
Mechanical properties of a fresh concrete are certainly significantly different from those for the
mature material. For this purpose, ATENA offers CCMaterialWithVariableProperties material
model. It builds up on any ATENA material model, but it updates its parameters using explicitly
given time functions. Of course, CCMaterialWithVariableProperties accounts for ticonstr , i.e. the
time functions receive (t  ticonstr ) argument. If creep and shrinkage analysis is required, one
should use ATENA MATERIAL id MAT_CONSTR_TIME t command. The material model
then calculates behavior of the material being by t younger, i.e. current and load time t , t ' is
replaced by t  t , t ' t .

5. Loading
A structure produced by digital 3D extrusion requires typically three kinds of boundary
conditions:
 Kinematic boundary condition, i.e. definitions of supports etc. They are much the same as
for traditionally built structure.
 Self-weight loading: This is modelled by element BODY LOAD option. Use its new
“INSIDE_T_TDT_ONLY” flag to add the element’s weight only once and at the proper
time. For example, use the command something like
LOAD BODY group 1 INSIDE_T_TDT_ONLY VALUE Z -0.023 ;
 Material shrinkage: This loading is input as element INITIAL STRAIN load, whereby we
must consider element construction time ticonstr . It is achieved by using a new element
load’s flag CONSIDER_CONSTR_TIME VALUE. At a particular time, younger
elements will exhibit a smaller shrinkage than the older ones. For example, use the
command something like LOAD TOTAL FUNCTION 100 INITIAL STRAIN group 1
CONSIDER_CONSTR_TIME VALUE X 1. Y 1.000 Z 1.000 ;
Note that for the sake of convenience it is recommended to input the load as total load.
Therefore, the loading function is defined as TOTAL. (By default, ATENA assumes
incremental load, i.e. LOAD INCREMENTAL FUNCTION….).

6. Visualization of printing process

By default, ATENA draws only elements that are active and/or elements active on condition
provided t  ticonstr . However, it can be overridden by checking a special switch, in which case
ATENA draws active element only if t  ticonstr and/or it draws conditionally active elements

ATENA Theory 213

despite their t  ticonstr status. As such, it is always possible to view full or only printed part of the
structure.

3.24 References
AHMAD, S., B. M. IRONS, ET AL. (1970). "Analysis of Thick and Thin Shell Structures by
Curved Finite Elements." International Journal of Numerical Methods in Engineering 2:
419-451.
BATHE, K.J.(1982), Finite Element Procedures In Engineering Analysis, Prentice-Hall, Inc.,
Englewood Cliffs, New Jersey 07632, ISBN 0-13-317305-4.
CRISFIELD, M.A. (1983) - An Arc-Length Method Including Line Search and Accelerations,
International Journal for Numerical Methods in Engineering, Vol.19,pp.1269-1289.
FELIPPA, C. (1966) - Refined Finite Element Analysis of Linear and Nonlinear Two-
Dimensional Structures, Ph.D. Dissertation, University of California, Engineering, pp.41-50.
HINTON, E. AND D. R. J. OWEN (1984). Finite Element Software for Plates and Shells,
Peridge Press.
JENDELE, L. (1981). Thick Plate Finite Element based on Mindlin's Theory. Prague, student
research work.
JENDELE, L. (1992). Nonlinear Analysis of 2D and Shell Reinforced Concrete Structures
Including Creep and Shrinkage. Civil Engineering Department. Glasgow, University of
Glasgow: 393.
JENDELE, L., A. H. C. CHAN, ET AL. (1992). "On the Rank Deficiency of Ahmad's Shell
Element." Engineering Computations 9(6): 635-648.
RAMM, E. (1981) - Strategies for Tracing Non- linear Responses Near Limit Points, Non- linear
Finite Element Analysis in Structural Mechanics, (Eds. W.Wunderlich, E.Stein, K.J.Bathe)
ROUSSEL, N. (2018). - Rheological Requirements for Printable Concretes, Cement and
Concrete Research 112: 76-85.
SUIKER, A. S. (2018). "Mechanical Performance of Wall Structures in 3D Printing
Processes: Theory, Design Tools and Experiments." International Journal of Mechanical
Sciences, 137: 145-170.
SUIKER, A. S., R. J. M. WOLFS, ET AL. (2020). "Elastic Buckling and Plastic Collapse During
3D Concrete Printing." Cement and Concrete Research 135: 1-16.
WOLFS, R. J. M., F. P. BOS, ET AL. (2018). Early Age Mechanical Behaviour of 3D Printed
Concrete: Numerical Modelling and Experimental Testing, Cement and Concrete Research 106:
103-116.
Suiker, A. S. (2018). "Mechanical Performance of Wall Structures in 3D Printing
Processes: Theory, Design Tools and Experiments." International Journal of Mechanical
Sciences, 137: 145-170.

214
4 SOLUTION OF NONLINEAR EQUATIONS
The main objective of this chapter is to review methods for the solution of a set of nonlinear
equations. Several methods, which are implemented in ATENA are described later in this
Chapter. However, all of them need to solve a set of linear algebraic equations in the form
Ax  b (4.1)
where A, x , b stands for a global structural matrix and vectors of unknown variables and rhs of
the problem, respectively. Hence, this problem is discussed first.

4.1 Linear Solvers

Two types of solvers are supported: direct and iterative, each of them having some pros and
cons. Without going into details, a direct solver is recommended for smaller problems or
problems. It is more robust and manages better ill-posed equations systems. On the other hand,
iterative solvers are typically more efficient to solve large (well-posed) 3D analyses. In addition,
two sparse direct solvers are provided. They intend to borrow advantages from both direct and
iterative solvers.
The two approaches (i.e., direct and iterative) differ in the way they store the structural matrix
A . It comes from the nature of FEM that the structural matrices have sparse character, with most
of nonzero elements located near the diagonal. The matrix has banded pattern and ATENA
works with band of variable width.
If a direct solver is used, then each column of matrix A stores all entries between the diagonal
element and the last nonzero element in the column. This structure is sometimes called sky-line
profile structure. The matrix A
 a11 a12 a13 a15 
a a22 a23 a24 a25 
 21 
 a31 a32 a33 a34 a35 
 
A a42 a43 a44 a45 a46  (4.2)
 a51 a52 a53 a54 a55 a56 
 
 a64 a65 a66 a67 
 a76 a77 

is thus stored in three vectors d , u , l with actual data and one vector p with information about
matrix’s profile:

d   a11 a77 
T
a22 a33 a44 a55 a66
u   a12 a67 
T
a13 a23 a13 a24 a34 a15 a25 a35 a45 a46 a56
l   a21 a31 a32 a76 
T
a13 a43 a43 a51 a52 a53 a54 a64 a65 (4.3)

p   0 1 3 5 9 11 12
T

For each column i of the matrix A the vector p stores location of ai (i 1) within the array u ,
resp. l . If A is symmetric, then u  l and only l is stored. Note the a direct solver we have to

ATENA Theory 215

store all elements within the bandwidth, even though some of them may be equal to zero,
because that they can become nonzero in the process of solution, (i.e., matrix factorization).
Iterative solver can store only true nonzero elements, irrespective of whether they are located
above or below the skyline. Suppose the matrix A from (4.2) that stores some zero elements
below the skyline

 a11 0 a13 a15 

0 a22 a23 a24 0 
 
 a31 a32 a33 a34 0 
 
A a42 a43 a44 0 a46  (4.4)
 a51 0 0 0 a55 0 
 
 a64 a65 a66 a67 
 a76 a77 

All iterative solvers would store the matrix A in three vectors. All the data are stored in a vector
a and location of the stored element is maintained in vectors r , c . The above matrix is stored as
follows:
a   a11 a31 a51 a22 a32 a42 a33 a13 a23 a43 ... a77 a67 
(4.5)
c  1 3 5 2 3 4 3 1 2 3 ... 7 6 
r   1 4 7 21 23 

The vector a stores for each column of A first diagonal element, followed by all nonzero
elements, from the top to the bottom of the column. The vector c stores row index of each entry
in the vector a . r stores location of all diagonal elements aii within a appended by an artificial
pointer to an 1n 1 , where n  dim( A) .

4.1.1 Direct Solver

The well-known Cholesky decomposition is used to solve the problem. The matrix A is
decomposed into
A  LDU (4.6)
where L, U is lower and upper matrix and D is diagonal matrix. The method to compute the
decomposition is described elsewhere, e.g. (Bathe 1982). Equation (4.1) is then solved in two
steps:
v  L1b
(4.7)
x   DU  v
1

Both of the above equations are computed easily, because the involved matrices have triangular
pattern. Hence, the solution of (4.7) represents back substitution only. If A is symmetric, (which
is usually the case), then

216
 U  LT (4.8)

4.1.2 Direct Sparse Solver

Direct sparse solvers are similar to the above Direct solvers; however, they should work more
economically both in terms of RAM and CPU requirements. They belong to a group of direct
(i.e., non-iterative) solution methods. They are based on matrix decomposition similar to (4.6).
The decomposition can be LU or LDU for non-symmetric matrices and/or LLT or LDLT
decomposition for symmetric matrices.
The main difference between these solvers and those from Section 4.1.3 is that they run the so-
called pre-factorization procedure before the actual factorization is executed. Such a pre-
factorization has two jobs:
1. Find out, what initially zero aij entries of the matrix A (that are stored below the skyline)
become nonzero due to factorization of A. Such entries are called fill-in.
2. Per mutate lines and columns of A so that the filling gets minimum.
Once a map of fill-in is known, it is added to the originally nonzero data of A and only these data
are to be stored and maintained in the next operations. Hence, as it is not necessary to store and
work upon all data below the skyline of A (as it is he case of solvers in Section 4.1.1); we can
use here a sparse matrix storage scheme. The incurred savings in both RAM and CPU resources
is significant and it pays off well for a computation overhead caused by the pre-factorization
phase and a bit more complicated storage scheme in use.
It is beyond the scope of this document to describe all details about the implementation of this
solver. It is based on (Vondracek, 2006) and (Davis et. al, 1995). A number of optimization
techniques are used to speed up the solution procedure, such as the problem (4.6) can be solved
using a block structure. This applies to pre-factorization, factorization as well as for
backward/forward substitution phases. The typical size of such a block is 2x2 .. 6x6. The bigger
block size, the smaller overhead for pre-factorization and mapping of the matrix and the faster
the operation to actually factorize and solve the problem (4.6). Use of a bigger block, however,
results also in a higher waste of RAM because all nonzero data and fill-in are rounded into a
storage with block pattern.
Direct sparse solvers are a compromise between Direct Solvers and Sparse Solvers. They
typically need more RAM and CPU than Sparse solvers do (and less than Direct Solvers),
however, they never diverge and bring uncertainties as what precoditioner to use, etc. Therefore,
they are recommended for middle size (may-be ill-conditioned) problems, the solution of which
would not fit into RAM subject a Direct Solver is used, and for which Sparse solvers are not
sufficiently robust.

4.1.3 Iterative Solver

The table below lists all solvers in ATENA that can solve the problem (4.1) iteratively. Although
the list is long, from the practical point of view only a few of them are recommended, see the
column “Description”. In addition, only the methods DCG and ICCG are designed to take full
advantage of symmetry of A (if present). The remaining solvers would store only the symmetric
part of A, however, they will operate on it in the same way as it is not symmetric. Therefore, for
symmetric problems, the solvers DCG and ICCG are preferable.
Each of the iterative solvers typically consists of two routines, one for “preparation” of the
solution and the other for the solution itself, i.e., “execution” phase. The former routine is

ATENA Theory 217

particularly important for the case of preconditioned iterative solvers. This is where a
preconditioning matrix is created.
The most efficient preconditioning routine are based on incomplete Cholesky decomposition
(Rektorys 1995). The preconditioning matrix A' is decomposed in the same way as (4.6), i.e.
A'  L'D'U' (4.9)
Comparing A and A' , it can be written
for aij  0 a 'ij  aij
(4.10)
for aij  0 a 'ij  aij
The incomplete Cholesky decomposition is carried out in the same way as complete Cholesky
decomposition (4.6), however, entries in A , which were originally zero and became nonzero
during the factorization are ignored, i.e., they stay zero even after the factorization. The incurred
inaccuracy is the penalty for memory savings due to usage of the iterative solvers’ storage
scheme. For symmetric problem, use ssics routine, for non-symmetric problems the ssilus is
available to construct A'  L'D'(L')T or A'  L'D'U' .
Last but not least, note that each solver needs some temporary memory. Such requirements are
included in the table below. Typically, the more advanced the iterative solver, the more extra
memory it needs and the fewer the number of iterations needed to achieve the same accuracy.

Table 4.1-1 SOLVER TYPES.

Type D/I Prep. Exec. Sym/N Temporary memory Description

phase phase on- required
sym
LU D --- --- S,NS ----- For smaller or ill-
posed probems
JAC I ssds sir S,NS 4*(11)+8*(1+4*n) Simple, not
recommended
GS I --- sir S,NS 4*(11+nel+n+1)+8*(1+3
*n+nel)
ILUR I ssilus sir S,NS 4*(13+4*n+nu+nl)+8*(1
+4*n+nu+nl)
DCG I ssds scg S 4*(11)+8*(1+5*n) For large symmetric
well-posed problems
ICCG I ssics scg S 4*(12+nel+n)+8*(1+5*n For large symmetric
+nel) problems,
recommended
DCGN I ssd2s scgn S,NS 4*(11)+8*(1+8*n) For large non-
symmetric well-
posed problems
LUCN I ssilus scgn S,NS 4*(13+4*n+nl+nl)+8*(1 For large non-
+8*n+nl+nu) symmetric problems,
recommended

218
DBCG I ssds sbcg S,NS 4*(11)+8*(1+8*n)
LUBC I ssilus sbcg S,NS 4*(13+4*n+nl+nu)+8*(1
+8*n+nu+nl)
DCGS I ssds scgs S,NS 4*(11)+8*(1+8*n)
LUCS I ssilus scgs S,NS 4*(13+4*n+nl+nu)+8*(1
+8*n+nu+nl)
DOMN I ssds somn S,NS 4*(11)+8*(1+4*n+nsave
+3*n*(nsave+1))
LUOM I ssilus somn S,NS 4*(13+4*n+nu+nl)+8*(1
+nl+nu+4*n+nsave+3*n
*(nsave+1))
DGMR I ssds sgmres S,NS 4*(31)+8*(2+n+n*(nsav
e+6)+nsave*(nsave+3))
LUGM I ssilus sgmres S,NS 4*(33+4*n+nl+nu)+8*(2
+n+nu+nl+n*(nsave+6)+
nsave*(nsave+3))
In the above:
n is the number of degree of freedom of the problem. nel is the number of nonzeros in the lower
triangle of the problem matrix (including the diagonal). nl and nu is the number of nonzeros in
the lower resp. upper triangle of the matrix (excluding the diagonal).

Table 4.1-2: EXECUTION PHASES.

Phase name Description

sir Preconditioned Iterative Refinement sparse Ax = b solver. Routine to solve a

general linear system Ax = b using iterative refinement with a matrix
splitting.
scg Preconditioned Conjugate Gradient iterative Ax=b solver. Routine to solve a
symmetric positive definite linear system Ax = b using the Preconditioned
Conjugate Gradient method.
scgn Preconditioned CG Sparse Ax=b Solver for Normal Equations. Routine to
solve a general linear system Ax = b using the Preconditioned Conjugate
Gradient method applied to the normal equations AA'y = b, x=A'y.
sbcg Solve a Non-Symmetric system using Preconditioned BiConjugate Gradient.
scgs Preconditioned BiConjugate Gradient Sparse Ax=b solver. Routine to solve a
Non-Symmetric linear system Ax = b using the Preconditioned BiConjugate
Gradient method.
somn Preconditioned Orthomin Sparse Iterative Ax=b Solver. Routine to solve a
general linear system Ax = b using the Preconditioned Orthomin method.

ATENA Theory 219

sgmres Preconditioned GMRES iterative sparse Ax=b solver. This routine uses the
generalized minimum residual (GMRES) method with preconditioning to
solve non-symmetric linear systems of the form: A*x = b.

Table 4.1-3: PREPARATION PHASES.

Phase name Description

ssds Diagonal Scaling Preconditioner SLAP Set Up. Routine to compute the
inverse of the diagonal of a matrix stored in the SLAP Column format.
ssilus Incomplete LU Decomposition Preconditioner SLAP Set Up.Routine to
generate the incomplete LDU decomposition of a matrix. The unit lower
triangular factor L is stored by rows and the unit upper triangular factor U is
stored by columns. The inverse of the diagonal matrix D is stored. No fill in
is allowed.
ssics Incompl Cholesky Decomposition Preconditioner SLAP Set Up. Routine to
generate the Incomplete Cholesky decomposition, L*D*L-trans, of a
symmetric positive definite matrix, A, which is stored in SLAP Column
format. The unit lower triangular matrix L is stored by rows, and the inverse
of the diagonal matrix D is stored.
ssd2s Diagonal Scaling Preconditioner SLAP Normal Eqns Set Up. Routine to
compute the inverse of the diagonal of the matrix A*A'. Where A is stored in
SLAP-Column format.

As for the solution procedure, i.e., the latter of the two solution phases, the most commonly used
method is the Conjugate gradient method (with incomplete Cholesky preconditioner) (Rektorys
1995). The flow of execution is as follows:

220
 r 1  b  A x1
z 1  M 1 r 1

r i zi
  i 1 i 1
i

r z
p i  z i   i p i 1
(4.11)
r i zi
  i
i

p Ap i  
x i 1  x i   i p i


r i 1  r i   i Ap i 
z i 1  M 1 r i 1
i  i 1
This solution procedure is implemented in scg routine.
The iterative solvers in ATENA are based on SLAP package (Seager and Greenbaum 1988) that
were modified to fit into ATENA framework. The authors of the package refer to (Hageman and
Young 1981), where all of the implemented solution techniques are fully described.

4.1.4 Parallel Direct Sparse Solver PARDISO 5

This solver uses PARDISO parallel direct sparse solver from the Math Kernel Library (MKL)
provided by Intel together with Intel Composer XE 2011. The solver has been developed within
the PARDISO Project, (see for example http://www.pardiso-project.org/). It is aimed for large
sparse symmetric and un-symmetric linear systems with shared memory. It offers direct or
iterative solver algorithms. The solver is well established and used by many software packages.
A lot of literature is related to the PARDISO project. For more information, refer to
http://fgb.informatik.unibas.ch/people/oschenk/index.html. Also, basic information is given in
the Intel Composer XE 2011 manuals.
A simplified version of this solver is also included in Atena. For the sake of simplicity, most
solution parameters are kept with their default value. The exception to that is the parameter
"PARDISO_REQUIRED_ACCURACY". It is input via the Atena "SET" input command. It
specifies, whether use of direct method with LU decomposition or iterative method with CGS
preconditioning is preferred. In the latter case, it also set a required solution accuracy. (For more
information refer to the Atena Input File Manual).
The following solver description is taken from the MKL manual provided by with Intel
Composer XE 2011, (also at http://software.intel.com/sites/products/documentation/hpc/
mkl/mklman/GUID-7E829836-0FEF-46B2-8943-86A022193462.htm.
Symmetric Matrices:
The solver first computes a symmetric fill-in reducing permutation P based on either the
minimum degree algorithm (Liu, 1985) or the nested dissection algorithm from the METIS
package (Karypis, 1998) (both included with Intel MKL), followed by the parallel left-right
looking numerical Cholesky factorization (Schenk, 2000) of PAPT = LLT for symmetric

5
Available starting from ATENA version 5.

ATENA Theory 221

positive-definite matrices, or PAPT = LDLT for symmetric indefinite matrices. The solver uses
diagonal pivoting, or 1x1 and 2x2 Bunch and Kaufman pivoting for symmetric indefinite
matrices, and an approximation of X is found by forward and backward substitution and iterative
refinements.
Whenever numerically acceptable 1x1 and 2x2 pivots cannot be found within the diagonal super-
node block, the coefficient matrix is perturbed. One or two passes of iterative refinements may
be required to correct the effect of the perturbations. This restricting notion of pivoting with
iterative refinements is effective for highly indefinite symmetric systems. Furthermore, for a
large set of matrices from different application areas, this method is as accurate as a direct
factorization method that uses complete sparse pivoting techniques(Schenk, 2004).
Another method of improving the pivoting accuracy is to use symmetric weighted matching
algorithms. These algorithms identify large entries in the coefficient matrix A that, if permuted
close to the diagonal, permit the factorization process to identify more acceptable pivots and
proceed with fewer pivot perturbations. These algorithms are based on maximum weighted
matchings and improve the quality of the factor in a complementary way to the alternative idea
of using more complete pivoting techniques.
The inertia is also computed for real symmetric indefinite matrices.
Unsymmetric Matrices:
The solver first computes a non-symmetric permutation PMPS and scaling matrices Dr and Dc
with the aim of placing large entries on the diagonal to enhance reliability of the numerical
factorization process (Duff and Koster 1999). In the next step the solver computes a fill-in
reducing permutation P based on the matrix PMPSA + (PMPSA)T followed by the parallel
numerical factorization
QLUR = PPMPSDrADcP
with super-node pivoting matrices Q and R. When the factorization algorithm reaches a point
where it cannot factor the super-nodes with this pivoting strategy, it uses a pivoting perturbation
strategy similar to (Li and Demmel 1999). The magnitude of the potential pivot is tested against
a constant threshold of alpha = eps*||A2||inf , where eps is the machine precision, A2 =
P*PMPS*Dr*A*Dc*P, and ||A2||inf is the infinity norm of the scaled and permuted matrix A.
Any tiny pivots encountered during elimination are set to the sign (lII)*eps*||A2||inf, which
trades off some numerical stability for the ability to keep pivots from getting too small. Although
many failures could render the factorization well-defined but essentially useless, in practice the
diagonal elements are rarely modified for a large class of matrices. The result of this pivoting
approach is that the factorization is, in general, not exact and iterative refinement may be needed.
Direct-Iterative Preconditioning.
The solver enables to use a combination of direct and iterative methods (Sonneveld 1989) to
accelerate the linear solution process for transient simulation. Most of the applications of sparse
solvers require solutions of systems with gradually changing values of the nonzero coefficient
matrix, but the same identical sparsity pattern. In these applications, the analysis phase of the
solvers has to be performed only once and the numerical factorizations are the important time-
consuming steps during the simulation. PARDISO uses a numerical factorization A = LU for the
first system and applies the factors L and U for the next steps in a preconditioned Krylow-
Subspace iteration. If the iteration does not converge, the solver automatically switches back to
the numerical factorization. This method can be applied to un-symmetric and structurally
symmetric matrices in PARDISO. For symmetric matrices, Conjugate-Gradients method is
applied. You can select the method using only one input parameter.

222
Separate Forward and Backward Substitution.
The solver execution step can be divided into two or three separate substitutions: forward,
backward, and possible diagonal. This separation can be explained by the examples of solving
systems with different matrix types.
A real symmetric positive definite matrix A is factored by PARDISO as A = L*LT . In this case
the solution of the system A*x=b can be found as a sequence of substitutions: L*y=b (forward
substitution) andLT*x=y (backward substitution).
A real unsymmetric matrix A is factored by PARDISO as A = L*U . In this case the solution of
the system A*x=b can be found by the following sequence: L*y=b (forward substitution) and
U*x=y (backward substitution).
Note that different pivoting (1x1, 2x2...) produces different LDLT factorization. Therefore results
of forward, diagonal and backward substitutions with diagonal pivoting can differ from results of
the same steps with Bunch and Kaufman pivoting. Of course, the final results of sequential
execution of forward, diagonal and backward substitution are equal to the results of the full
solving step regardless of the pivoting used.
Sparse Data Storage.
Sparse data storage in PARDISO follows the scheme described above.

4.2 Full Newton-Raphson Method

Using the concept of incremental step by step analysis, we obtain the following set of nonlinear
equations:
K ( p ) p  q  f ( p ) (4.12)

where:
q is the vector of total applied joint loads,

f ( p ) is the vector of internal joint forces,

p is the deformation increment due to loading increment,

p are the deformations of the structure prior to load increment,

K ( p ) is the stiffness matrix, relating loading increments to deformation increments.

The R.H.S. of (4.12) represents out-of-balance forces during a load increment, i.e., the total load
level after applying the loading increment minus internal forces at the end of the previous load
step. Generally, the stiffness matrix is deformation dependent, i.e., a function of p , but this is
usually neglected within a load increment in order to preserve linearity. In this case, the stiffness
matrix is calculated based on the value of p pertaining to the level prior to the load increment.

The set of equations (4.12) is nonlinear because of the nonlinear properties of the internal forces:
f (kp )  kf ( p ) (4.13)

and nonlinearity in the stiffness matrix

K ( p )  K ( p  p ) (4.14)

where k is an arbitrary constant.

ATENA Theory 223

The set of equations represents the mathematical description of structural behavior during one
step of the solution. Re-writing equations (4.12) for the i-th iteration within a distinct loading
increment we obtain:
K ( pi 1 ) pi  q  f ( pi 1 ) (4.15)

All the quantities for the (i-1)-th iteration have already been calculated during previous solution
steps. Now we solve for p i at load level q using:

pi  pi 1  pi (4.16)

As pointed out earlier, equation (4.15) is nonlinear, and therefore it is necessary to iterate until
some convergence criterion is satisfied. The following possibilities are supported in ATENA
( k marks k -th component of the specified vector):

piT pi
  rel . disp
piT pi

( q  f ( pi 1 ))T ( q  f ( pi 1 ))
  rel . force
f ( pi )T f ( pi )
(4.17)
piT ( q  f ( pi 1 ))
  rel .energy
piT f ( pi )

max(( q k  f k ( pi 1 ))) max( ( q k  f k ( pi 1 )))

  abs . force
max( f k ( pi )) max( f k ( pi ))

The first one checks the norm of deformation changes during the last iteration whereas the
second one checks the norm of the out-of-balance forces. The third one checks out-of-balance
energy, and the fourth condition checks out-of-balanced forces in terms of maximum
components (rather than Euclid norms). The values of the convergence limits  are set by
default to 0.01 or can be changed by the input command SET.
The concept of solving nonlinear equation set by Full Newton-Raphson method is depicted in
Fig. 4-1:

224
 Loading
q

Loading increment

p p p Deformation
0 1 2

Fig. 4-1 Full Newton-Raphson method.

4.3 Modified Newton-Raphson Method

The most time-consuming part of solution (4.15) is the re-calculation of the stiffness matrix
K ( pi 1 ) at each iteration. In many cases this is not necessary and we can use matrix K ( p0 ) from
the first iteration of the step. This is the basic idea of the so-called Modified Newton-Raphson
method. It produces very significant time saving, but on the other hand, it also exhibits worse
convergence of the solution procedure.
The simplification adopted in the Modified Newton-Raphson method can be mathematically
expressed by:
K ( pi 1 )  K ( p0 ) (4.18)

The modified Newton-Raphson method is shown in Fig. 4-2. Comparing Fig. 4-1 and Fig. 4-2 it
is apparent that the Modified Newton-Raphson method converges more slowly than the original
Full Newton-Raphson method. On the other hand, a single iteration costs less computing time,
because it is necessary to assemble and eliminate the stiffness matrix only once. In practice, a
careful balance of the two methods is usually adopted in order to produce the best performance
for a particular case. Usually, it is recommended to start a solution with the original Newton-
Raphson method and later, i.e., near extreme points, switch to the modified procedure to avoid
divergence.

Loading
q

Loading increment

p0 p1 p2 Deformation
p3 p4

Fig. 4-2 Modified Newton-Raphson method

ATENA Theory 225

4.4 Arc-Length Method
Next to the Modified Newton-Raphson method, the most widely used method is the Arc-length
method. This method was first employed about fifteen years ago to solve geometrically nonlinear
structures. Because of its excellent performance, it is now quite well established for geometric
nonlinearity and for material nonlinearity as well. Many workers have been interested in using
and improving Arc-length procedures. In Atena, it can be used within CCStructures module, i.e.
for static analysis.
The main reason for the popularity of this method is its robustness and computational efficiency
which assures good results even in cases where traditional Newton-Raphson methods fail. Using
an Arc-length method stability problems such as snap back and snap through phenomena can be
studied as well as materially nonlinear problems with non-smooth or discontinuous stress-strain
diagrams. This is possible due to the changing load conditions during iterations within an
increment.
The main idea of this method is well explained by its name, arc-length. The primary task is to
observe complete load-displacement relationship rather than applying a constant loading
increment as it is in the Newton-Raphson method. Hence this method fixes not only the loading
but also the displacement conditions at the end of a step. There are many ways of fixing these,
but one of the most common is to establish the length of the loading vector and displacement
changes within the step.
From the mathematical point of view, it means that we must introduce an additional degree of
freedom associated with the loading level (i.e., a problem has n displacement degrees of freedom
and one for loading) and in addition, a constraint for the new unknown variable must be
introduced. The new degree of freedom is usually named . There are many possibilities for
defining constraints on  and those implemented in ATENA are briefly reviewed in the
following sections.
To derive the Arc-length method, we re-write the set of equations (4.12) in the form of (4.19),
where  defines the new loading factor:
K ( p ) p   q  f ( p ) (4.19)

Now re-writing (4.19) in a form suitable for iterative solution:

K ( pi 1 ) pi   q  f ( pi 1 )   q  f i 1 (4.20)

pi  pi 1  pi  pi 1  i 1 i 1 (4.21)

pi  pi 1  i 1 i 1 (4.22)

i  i 1  i 1 (4.23)
The notation is explained in Fig. 4-3. The matrix K can be recomputed for every iteration
(similar to the Full Newton-Raphson method) or it can be fixed based on the 1st iteration for all
subsequent iterations (Modified Newton Raphson method). The vector q does not mean in this
case the total loading at the end of the step but only a reference loading "type". The actual
loading level is a multiple of this.

226
The scalar  is an additional variable introduced by the Line-search method, which will be
discussed later. The scalar  is used to accelerate solutions in cases of well-behaved load-
deformation relationships or to damp possible oscillations if some convergence problems arose,
e.g., near bifurcation and extreme points.

q 0

0 q

q 1
1 q
q 2 g
q 3 1
R2
g0
Load increment

Loading R1

T 0

 0
0
q  start
p 1

p 2
p 3
00 11 2 2

p0 p1 p2 Deformation

Fig. 4-3 The Arc-length method

Additional notation is defined as follows:
Out-of-balance forces in i-th iteration:
g ( pi )  g i  f i  i q  f  ( i 1  i 1 ) qi (4.24)

R.H.S vector in i-th iteration:

RHS i  i q  f i 1  i 1q  gi 1 (4.25)

Substituting (4.21) through (4.25) into (4.20), the deformation increment  i 1 can be calculated
from:
K i 1  RHS i 1  i 1q  gi 1 (4.26)

Hence:

 i 1   i 1  i 1 T (4.27)

where
 i 1   K 1 gi 1
(4.28)
T  K q 1

It remains only to set the additional constraint for i 1 and i 1 and the whole algorithm is
defined. Thus compared to the Newton-Raphson methods in which we solve n dimensional
nonlinear problem, the Arc-length method need to solve a (n + 2) dimensional problem, where
the first n unknowns correspond to deformations and the last two are i 1 and i 1 .

ATENA Theory 227

If we set i 1  1 , then we deal with an (n + 1) dimensional problem that corresponds to the pure
Arc-length method, otherwise, a combination of Arc-length and Line search must be employed.
The Line search method is discussed later in this chapter. Note that all vectors including  i 1 ,
 T are of order (n + 1). Their (n + 1)-th coordinate corresponds to the loading dimension  and
it is set to zero.
Now, introduce two new vectors ti 1 and ni 1 as shown in Fig. 4-4. There are defined by:
ti 1  pi 1   ( i 1  start ) (4.29)

ni 1   i 1  i 1 (4.30)
where:
 is scalar that relates dimensions of  to size of deformation space,

i 1 is a (n + 1) dimensional vector with its firth n coordinates set to zero (deformation

space) and its (n + 1)-th coordinate equal to i 1 .

start is a (n+1) dimensional vector similar to i 1 , however its (n + 1)-th coordinate equal
to start .

n1
 n2
n3

t1 t3
t2

Fig. 4-4 The vectors ti and ni and scalar  .

It is then obvious that
ti  ti 1  nn 1 (4.31)
Defining the residual R :
Ri 1  ti 1 ni 1 (4.32)
equations (4.20) through (4.32) lead to the final expression for the unknown i 1 (noting that
piT1 i 1  piT1 i 1  0 ):

228
 Ri 1  piT1  i 1
i 1  (4.33)
piT1 T   2 (i 1  start )
To obtain i 1 by (4.33) the residual Ri 1 must be defined. In fact, it also defines the type of
Arc-length constrain being used. The types supported in ATENA are described below.

4.4.1 Normal Update Method

Vector ti 1 and ni 1 are normals in this case, hence residual Ri 1  0 , see Fig. 4.4-3.

n1
 n2
n3

t1 t3
t2

Fig. 4.4-3 Normal update method.

The main advantage of this method is its simplicity. The Normal update plane is relatively
reliable, but it can fail if the l-p diagram suddenly changes its slope or turns back or down (snap
back and snap through). Nevertheless, if these special conditions are treated by this method, then
a very significant reduction in step length is unavoidable.

4.4.2 Consistently Linearized Method

The residual Ri 1 is defined in this case by

Ri 1  ti T1 ni 1  ti 1 ni 1 cos( )   ti 1 ( ti 1  s ) (4.34)

The step length s and angle  are depicted in Fig. 4.3-4. The norm of the vector ti 1 is
calculated using (4.29):
2
ti 1  piT1pi 1   2 ( i 1  start ) 2 (4.35)

ATENA Theory 229

 
n
i-1

ti-1
current s
required
s
ti s = step length

Fig. 4.4-4 Consistently linearized method.

Substituting (4.34) and (4.35) in (4.33) we obtain the final expression for i 1 . It should be
noted that the scalar s is set 'a priori' and governs the actual step length. Of course, the proper
choice of this parameter is essential for the solution and therefore it will be discussed later in
more detail.
This method is especially suitable for solutions that embrace   p diagrams with sudden breaks
and discontinuities, e.g. for materially nonlinear problems.

4.4.3 Explicit Orthogonal Method

The basic constraint for i 1 in this case is that ti 1  ti  s , where s is some distinct 'a priori'
set step length. Similar to the previous method, we also have to evaluate the residual Ri 1 :

Ri 1  ti T1 ni 1  ti 1 ni 1 cos( )   ti 1 ri 1 (4.36)

Based on the similar triangles (see Fig. 4.4-), the following can be derived:
rl 1 tl 1
 (4.37)
t 'l 1  s t 'l

 s 2 ( ti '  s )
Ri 1  (4.38)
ti '

ti '  ti 1  ni'1 (4.39)

2 2 2
ti '  ti ' 1   2 i21   2  i 1 (4.40)

The vector ti ' 1 is calculated using (4.35). By substituting the above equations into (4.33) the
final expression for i 1 is obtained.
From the above derivation, it is clear that in practice we at first employ Normal Update Method
(Chapter 4.4.1) to solve for ti ' and ni'1 and thereafter, we correct the i 1 in order to satisfy
the constraint ti 1  ti  s .

230
 ||ri-1|| n’i-1


|| r’i-1|| = ||t’i - s||
||ti-1||
ti-1 ni-1

ti
s
||t’i||

s = step length

p
Fig. 4.4-5 Explicit orthogonal method.
This method is usually utilized to analyze geometrically nonlinear structures, particularly
stability problems. Its main feature is robustness and compared with the "classical" Crisfield
cylinder method (see below) it avoids the problem of the choice of the proper i 1 root (the
condition ti 1  ti  s while expressing vector length analytically). As for convergence, the
method is comparable to the method 4.4.3, but has the advantage that it preserves the step length.

4.4.4 The Crisfield Method.

The Crisfield method is derived directly from the constraint of constant step length
ti 1  ti  s The residual Ri 1 is not used in this case and we substitute equations (4.20)
through (4.31) straight into the above constraint. It leads to the following equation for i 1 :

a1i21  a2 i 1  a3  0 (4.41)

where:
a1   2 TT  T   2

a2  2  2 ( i 1  start )  2 TT  i 1 2 (4.42)

a3   2 ( i 1  start ) 2   2 i T1 i 1  s 2
Equation (4.41) has generally two roots i 1 and hence we must decide which of them to use.
There exist several strategies but ATENA chooses that root i 1 , for which cos( ti 1 , ti )  0 (or
higher of them), i.e., direction of new increment as close as possible to direction of the previous
increment (within the same step).

ATENA Theory 231

4.4.5 Arc Length Step
The proper step length is of essential importance for good execution performance. It directly
influences the convergence radius on the one hand and the number of required steps on the other.
ATENA uses the following procedure to set (or optimize) s :
(1) Set loading vector q and thus define a reference loading level (within one load
increment).
(2) Structural response to this load in the 1st execution step, the 1st iteration defines step
length s1 in the 1st step. In the subsequent steps, the step length is kept fixed or optimized
(based on SET ATENA input command, subcommand
&ARC_LENGTH_OPTIMISATION:
ni 1
si  si 1 (4.43)
n

ni 1
si  4 si 1 (4.44)
n
n
si  si 1 (4.45)
ni 1
where
si and si 1 is Arc length step length in the current and the previous load increment,
respectively.
n and ni 1 is desired number of iterations and number of iterations in the previous step.
n is typically 5-6.

4.5 Line Search Method

The objective of this method is to calculate the parameter  that was already introduced in the
Chapter 4.4 Arc-Length Method The method can be used either independently or in combination
with Arc length method. The primary reason for introducing a new parameter (i.e. a new degree
of freedom to the set of equations) is to accelerate or to damp the speed of analysis of the load-
displacement relationship.
The basic idea behind  is to minimize work of current out-of-balance forces on displacement
increment.
Let us assume that we have already solved already two points p0 and p0   '  p and thus we
have also calculated out-of-balance forces g ( p0 ) and g ( p0   '  ) at these points. The aim of
this method is to set the parameter  so that the work being done by out-of-balance forces at
point p0   is minimum.

The work of out-of-balance forces is:

p
 ( p )   ( p0 )   g ( p )T dp  minimum (4.46)
po

Hence:

232
  ( p )   p T 
p p
 0  p g ( p)   g ( p )T 0 (4.47)
d p  0   

Interpolating linearly out-of-balance forces between points p0 and p0   ' 

 g ( p0   '  )  g ( p0 )  g ( p0   '  )  g ( p0 )
g ( p0   )  g ( p0 )    || p0    p0 || g ( p0 )  
 || p0   '   p0 ||   '
 
(4.48)
and using :
p  p0  
p (4.49)


The final expression for  ' can be derived:
g ( p0 )T 
  ' (4.50)
g ( p0 )T   g ( p0   '  )T 

Thus, the Line search method can be summarized:

Use any method to calculate displacement increment  , (see Fig. 4-3 and (4.28)). The
parameter  ' can be set from the last load increment or simply to unity.
Calculate out-of-balance forces for both g ( p0 ) and g ( p0   '  ) .

Use (4.50) to calculate new value for  .

As all the above equations are nonlinear, the parameter  must be solved by iterations until
g ( p0  
 a specified energy drop, typically < 0.6 – 0.8 >.
g ( p0 )

Practical experience suggests that the value of parameter  should be kept in interval < 0.1 – 5>.

4.6 Parameter 
The parameter  scales the deformation space p to the loading dimension  . If   0 , the
solution for i 1 is searched on an area of a cylindrical shape of radius equal to step length
s (Crisfield method) and the axis normal to the p (deformation) space. The solution is the point
of intersection of this area and the line, defined by the energy gradients of structure and by the
applied load at point p . If   0 , the solution is carried out in the same way on ellipsoidal or
spherical space.
The higher value of  , the higher "weight factor" for changes in loading space compared to
displacement increments.
ATENA currently supports the following formulae for setting and optimization of  (for current
step j ). They are reviewed below.

ATENA Theory 233

The first strategy requires the “load to displacement” increment ratio (4.51) is constant
throughout all steps, (e.g., input value  req )

 
  req (4.51)
( p )

Then, at the end step j-1 we can calculate

j 1   j 1
j 1  (4.52)
( j 1 p )

This value (due to nonlinearities) will not match req . Therefore, for step j we will modify j 
as follows:

j   req  j 1 
req
 (4.53)
j 1 
req req ( j 1 p )
      req
j 1   j 1
j 1 j 1 j 1
 j 1
j
j 1 
( j 1 p )

The above optimization process is initialized in the first step by assuming that
0   1, 0   1,  ( j 1 p )   T , where  T is displacement corresponding to master Arc-length
load increment defined earlier in this chapter. Hence
req req req
  0      req  T (4.54)
j 1   j 1
1
j 1  1
( j 1 p) T

The parameters j  in all subsequent steps are calculated using (4.53). If the ratio of
displacements changes ( j p ) to load changes  ( j  ) in the last load step increase, then the
equation (4.54)(4.55) increases  in the current step, thereby puts higher „weight factor“ on
loads compared to displacements. Hence, the equation (4.54) tends to keep constant importance
of loading space irrespective of displacements. Note that the equation (4.54) corresponds to
BETA_FORCES_DISPLS_RATIO_CONSTANT.
The second supported strategy is different. In ATENA, it is referred to as
BETA_RATIO_CONSTANT method and it tries to keep constant  coefficients, whilst
managing the coefficients  . Thus, it works in the opposite way as compared to the first strategy
described above.
From (4.52) we can write for steps (j-1) and j

234
 j 1  ( j 1 p )
j 1 
 j 1

j  ( j p)
j 
 j

Now requiring j 1   j  we have

j  ( j p) j 1  ( j 1 p)

 j  j 1
 ( j 1 p)
(4.56)
j  j 1

j 1  ( j p)
 j

 j 1  j j  j
and if we assume  , then  and the above equation yields
 ( j 1 p ) ( j p) j 1  j 1 

( j 1 p)
 j 1
j   j 1  (4.57)
( j p)
 j

 ( j 1 p )
 j 1
If in subsequent steps changes, the procedure is trying to compensate for that by re-
( j p)
 j
( p)
adjusting the coefficients  . In other words, this strategy is trying to keep constant,

(i.e., the relative importance of load vs. displacement spaces).

4.7 Band Width Optimization

The way in which individual structural degrees of freedom (dofs) are mapped into the global
structural matrices has a significant impact on their size and cost of the solution in terms of
required CPU and RAM resources.
Let us assume the 2D example of the 3 bars element from Fig. 4-5. The structure consists of
three beam elements 1,2,3. It has four global nodes with three degrees of freedom in each of
them, i.e., two displacements and one rotation. Suppose the structure is solved by a direct solver,
i.e., we use half-band skyline storage scheme (4.4).
By default, i.e., without any optimization, the structural degrees of freedom are allocated
sequentially starting from the node 1 up to the last node n, i.e., 4. Hence, the jth degree of
freedom at the node i has number ndof (i  1)  j , where ndof is number of dofs per node.

ATENA Theory 235

If the structural nodes are numbered as indicated, then the beam 1,2 and 3 have nodal incidences
1-3, 3-4 and 4-2, respectively and the final stiffness matrix K has the pattern from the left-
bottom part of Fig. 4-5. Note that the matrix K must also store the entries depicted as circles
without filling. Although they are initially zero, they may turn nonzero during the matrix
decomposition needed to solve the problem, i.e., we must store the matrix with 69 entries and
maximum half-band width 9.
On the other hand, if nodal degrees of freedom are numbered as shown in the right-bottom part
of Fig. 4-5, then the matrix K must store only 51 entries and has maximum half bandwidth only
6.
The two examples document, how important efficient numbering of the degrees of freedom of
the structure is. If the structure (to be solved) is simple, then a suitable dofs' numbering can be
done manually by appropriate numbering of the structural nodes. However, in the more complex
cases (and in particular if a model of the structure is generated automatically), an optimal dofs
mapping must be calculated.
There are number of algorithms that deliver more or less efficient dofs mapping. Probably the
best established algorithm of that kind is Cuthill-McKee algorithm (Cuthill, McKee 1969). This
is not due to its superior property, but due it has been developed as first. The algorithm produces
an ordered n-tuple R of vertices which is the new order of the structural vertices. It numbers the
vertices according to a particular breadth-first traversal, where neighboring vertices are visited in
order from lowest to highest vertex order.
The reverse Cuthill–McKee algorithm (RCM) is the alternative of the Cuthill-McKee algorithm,
in which the vertices are visited in reverse order, i.e. form the highest to the lowest vertex.
ATENA implements Gibbs and Sloan dofs optimization algorithms:

236
 Fig. 4-5 Optimization of dofs numbering

The Sloan algorithm (Sloan, Randolf (1983)

In an effort to obtain an optimum elimination order, the algorithm first renumbers the nodes, and
then uses this result to resequence the elements. This intermediate step is necessary because of
the nature of the frontal solution procedure, which assembles variables on an element-by-
element basis but eliminates them node by node. To renumber the nodes, a modified version of
the King’ algorithm is used. In order to minimize the number of nodal numbering schemes that
need to be considered, the starting nodes are selected automatically by using some concepts from
graph theory. Once the optimum numbering sequence has been ascertained, the elements are
then reordered in an ascending sequence of their lowest-numbered nodes. This ensures that the
new elimination order is preserved as closely as possible. For meshes that are composed of a
single type of high-order element, it is only necessary to consider the vertex nodes in the
renumbering process. This follows from the fact that mesh numberings which are optimal for
low-order elements are also optimal for high-order elements. Significant economies in the
reordering strategy may thus be achieved.

ATENA Theory 237

The Gibbs et. al. algorithm (Gibbs et. al. 1976)
This algorithm typically produces bandwidth and profile, which are comparable to those of the
commonly-used reverse Cuthill–McKee algorithm, yet it requires significantly less computation
time. Nevertheless, it delivers dofs mapping that is usually slightly less efficient than that by the
Sloan algorithm and therefore, it is less preferred option the optimization.
Note that the above algorithms optimize dofs numbering by reordering the structural nodes. They
do not account for possible different number of dofs within a particular node. Note also that in
order to minimize cost of the dofs remapping, the optimization is carried out before assembling
the structural global matrices and vectors. Thus, they are assembled directly into their final,
optimized location.
Iterative solvers use data storage scheme (4.3). As the storage scheme stores only nonzero
elements, the solution is less sensitive to a bad dofs mapping. For huge analyses it is nevertheless
suggested to carry out a dofs mapping optimization, as it typically yields individual elements
entries stored closer to each other with positive effect on solution convergence and RAM data
management.
A detailed description of the above algorithms is above scope of the publication. For more
information the reader is suggested to study the given references.

4.8 References
BATHE, K.J.(1982), Finite Element Procedures In Engineering Analysis, Prentice-Hall, Inc.,
Englewood Cliffs, New Jersey 07632, ISBN 0-13-317305-4.
CRISFIELD, M.A. (1983) - An Arc-Length Method Including Line Search and Accelerations,
International Journal for Numerical Methods in Engineering, Vol.19, pp.1269-1289.
CUTHILL, E. and J. MCKEE (1969). Reducing the Bandwidth of Sparse Symmetric Matrices.
Proc. 24th Nat. Conf. ACM.
DAVIS, T., AMESTOY, P., DUFF, I.S (1995) - An Aproximate Minimum Degree Ordering
Algorithm, Comp. and Information Science Dept., University of Florida, Tech. Report
TR-94-039.
DUFF, I. S. and KOSTER, J. (1999) - The Design and Use of Algorithms for Permuting Large
Entries to the Diagonal of Sparse Matrices. SIAM J. Matrix Analysis and Applications,
20(4):889-901.
FELIPPA, C. (1966) - Refined Finite Element Analysis of Linear and Nonlinear
Two-Dimensional Structures, Ph.D. Dissertation, University of California, Engineering,
pp.41-50.
GIBBS, N. E., W. G. POOLE, et al. (1976). "An Algorithm for Reducing the Bandwidth and
Prole of a Sparse Matrix." SIAM Journal of Numerical Analysis 13(2).
KARYPIS, G. and KUMAR, V. (1998) - A Fast and High Quality Multilevel Scheme for
Partitioning Irregular Graphs. SIAM Journal on Scientific Computing, 20(1):359-392.
LIU, J.W.H. (1985) - Modification of the Minimum-Degree Algorithm by Multiple Elimination.
ACM Transactions on Mathematical Software, 11(2):141-153.
LI, X.S., DEMMEL, J, W. (1999) - A Scalable Sparse Direct Solver Using Static Pivoting. In
Proceeding of the 9th SIAM conference on Parallel Processing for Scientific Computing, San
Antonio, Texas, March 22-34.

238
RAMM, E. (1981) - Strategies for Tracing Non- linear Responses Near Limit Points, Non- linear
Finite Element Analysis in Structural Mechanics, (Eds. W.Wunderlich,E.Stein, K.J.Bathe)
REKTORYS, K. (1995). Přehled užité matematiky. Prague, Prometheus.
SLOAN, S. W. and M. F. RANDOLF (1983). "Automatic Element Reordening for Finite
Element Analysis with Frontal Solution Schemes." Int. Journal for Numerical Methods in Eng.
19: 1153-1181.
SEAGER, M. K. and A. GREENBAUM (1988). A SLAP for the Masses, Lawrence Livermore
National Laboratory
SCHENK, O., GARTNER, K. and FICHTNER, W. (2000) - Efficient Sparse LU Factorization
with Left-right Looking Strategy on Shared Memory Multiprocessors. BIT, 40(1):158-176.
SCHENK, O. and GARTNER, K. (2004) - On Fast Factorization Pivoting Methods for Sparse
Symmetric Indefinite Systems. Technical Report, Department of Computer Science, University
of Basel, submitted.
SONNEVELD, P. (1989) - CGS, a Fast Lanczos-Type Solver for Nonsymmetric Linear
Systems. SIAM Journal on Scientific and Statistical Computing, 10:36-52.
VONDRACEK, R. (2006) - The Use Of The Sparse Direct Solver In The Egineering
Applications Of The Finite Element Method. Theses for Ph.D. Czech Technical University,
Prague.

ATENA Theory 239

5 CREEP AND SHRINKAGE ANALYSIS
Creep and shrinkage are undoubtedly features that have a significant influence on concrete
behaviour. Although creep and shrinkage analysis can be neglected in the design of most civil
structures, there exist cases when these phenomena have to be accounted for. The Ref. (Bazant
and Baweja 1999) provides a five levels classification of structures that can serve as simple
guidelines for making a decision, when creep and shrinkage analysis is needed and when it is not
needed. The recognized levels of structures are as follows:
Level 1: Reinforced concrete beams, frames, and slabs with span under 20m and heights of up to
30m, plain concrete footings, retaining walls.
Level 2. Prestressed beams or slabs of spans up to 20m, high-rise building frames up to 100m
high.
Level 3. Medium-span box girder, cable-stayed or arch bridges with spans of up to 80m,
ordinary tanks, silos, pavements.
Level 4. Long-span prestressed box-girder, cable-stayed or arched bridges; large bridges built
sequentially in stages by joining parts, large gravity, arch or buttress dams, cooling towers, large
roof shells, very tall buildings.
Level 5. Record span bridges, nuclear containments and vessels, large offshore structures, large
cooling towers, record-span thin roof shells, record-span slender arch bridges.
Full creep and shrinkage analysis is mandatory for the design of structures level 4 and 5 and it is
recommended also for the level 3 structures.

5.1 Implementation of Creep and Shrinkage Analysis in ATENA

ATENA software provides a powerful method for creep and shrinkage analysis for most
problems from engineering practice. It is based on the so-called cross-sectional approach,
meaning that the analysis builds upon creep and shrinkage behavior of the whole cross-section
rather than the behavior of individual material points only. The reason for choosing this method
is that at this moment, there are available numerous models for predicting creep and shrinkage
behavior of a concrete cross-section, whereas there is very low evidence about the same behavior
at the material point level. The second reason is that its accuracy suffices for most analyses from
engineering practice, and it is much less expansive in terms of computational cost.

5.1.1 Basic Theoretical Assumptions

The implemented creep and shrinkage analysis is based on the assumption of linear creep, which
in other words means that the material compliance function (t , t ') and accompanying function
for shrinkage  0 (t ) depends only on material composition, temperature, shape, and time at
observation t and at loading t ' . It does not depend on stress-strain conditions. In spite of the
simplifications, the provided analysis is sufficiently accurate in most practical cases and it is fast
and efficient. On the other hand, it is applicable only for structures, where the stress value does
not exceed about 60% of the ultimate strength of concrete. For higher load levels, the material

ATENA Theory 241

nonlinearity becomes significant and a more elaborate solution has to be employed. The above
simplification applies to time-dependent (i.e., long-term) material behavior only. For short-term
behavior of the material, model retains its nonlinearity, i.e., it accounts for phenomena such as
cracks, plasticity.
The creep and shrinkage analysis is based on the assumption of Stieltjes integral, which is
written for the case of 1D analysis in the following form:
t 
 (t )   (t , t ') d   0 (t ) (4.58)
t' 
where:
t = observation time,
t ' = loading time,
 (t ) =stress at the time t ,

 0 (t ) = initial stress-independent strain such as concrete shrinkage,

(t , t ') = compliance function of concrete.

Fig. 5-1 Decomposition of stress history into stress steps (left) or impulses (right).
The sense of Stieltjes integral is given in the above figure.
Equation (4.58) has to be modified for the case of 2 and 3D analyses for practical analyses. This
is done below. It is important to note that (4.58) applies to any stress and strain history, and it is
defined in incremental form. It means that at a particular time t , stress at t  t depends only

on the current material state at time t and stress increment at a time t  t , i.e.   d .

The final form of the above equations reads:
t   B( ( )) 
 (t )   (t , )  B( ( ))   ( )  d   0 (t ) (4.59)
t'
   
where:
 (t ) = is stress vector at a time t , (note the bar atop of a symbol indicates vector),

 0 (t ) = vector of initial strains, such as shrinkage,

B( ( )) = matrix accounting for multiaxial stress-strain conditions, including all material
short-term nonlinearities.

242
Notice the way the equation (4.59) is written. Long-term and short-term material behavior is
separated. The former is encapsulated in the compliance function (t , t ') , whereas the short-
term behavior is comprised in the matrix B( ( )) . This assumption brings significant
simplification of the creep and shrinkage analysis, and it is believed that for most practical
analysis, the induced inaccuracy is acceptable.
Substituting t  t '  t ,  t  0 into (4.59) and applying load increment  (t ')   (t ') (i.e.,
loading from the zero level) at a time t ' , it can be derived
 (t '  t )   (t '  t , t ')B( (t ')) (t ')   0 (t '  t ) (4.60)
Comparison of (4.60) with similar equations for constitutive relations for short-term loading
conditions, i.e. t '  t  t ' , yields instantaneous secant material rigidity matrix:

D(t ') =  B( (t ')) (t ', t ') 

1
(4.61)

The matrix D(t ') corresponds to the reciprocal value of the well-known secant Young modulus
E (t ') in the case of 1D stress-strain conditions. In the case of plane stress conditions, the matrix
B( ( )) reads (4.62), etc.

 1  0 

B= 1 0  (4.62)
 sym. 2(1   ) 

5.2 Approximation of Compliance Functions  (t , t ') by Dirichlet

Series.
Ref. (Bazant and Spencer 1973) and others show that significant improvement of computational
efficiency can be obtained if the original material compliance function (t , t ') is during the
creep solution approximated by Dirichlet series  '(t , t ') as follows:

n  
 t t ' 

1 1    

 '(t , t ')   1 e 
(4.63)
E (t ')  1 E (t ')  
 
where :
  = are so-called retardation times,

n = number of approximation functions, i.e., this parameter is related to the input parameter
number of retardation times.
E (t ') = instant Young modulus at the time t ' ,

E (t ') =coefficients for the approximation functions.

ATENA Theory 243

Fig. 5-2 Approximation of compliance (or retardation) function curve at age t' at loading by a sum of
exponentials used as shape functions of Dirichlet series
The effect of the use of Dirichlet series approximation is depicted in the above figure. A single
approximation exponential is drawn in sub-figure (a), while the whole process of decomposition
of compliance and retardation curves is depicted in the sub-figures (b), (c), respectively.
The incorporation of the Dirichlet series  '(t , t ') brings the following benefits:

- Creep analysis is independent of the material creep prediction model.

- Time integration is exact; hence, fewer temporal increments are necessary.
- Less demand of computer storage needed for storing data from the previous temporal
steps of the analysis. It suffices to store data from the previous analysis step only, rather
than the complete stresses-strain history of the analyzed structure.

5.3 Step by Step Method

Equation (4.59) (upon substitution (4.63) is solved numerically. The structure is discretized in
space by the finite element method (described elsewhere in this document). As for time, the
solution is carried out by the Step-by-step method (SBS) (Bazant 1988). The structural behavior
is analyzed in several time steps, i.e. in time increments, as it corresponds to (4.59). After some
mathematical manipulations (Jendele and Phillips 1992), the final solution equations read:
 r 1/ 2  B 1 ( -  r )
 r  E (4.64)
r -1/ 2 r

 (tr )   r   r 1   r (4.65)
 (tr )   r   r 1   r (4.66)

1 1 n
1
   1   , r  (4.67)
 r 1/ 2
E Er 1/ 2  1 E , r 1/ 2

 t
 r

 
r  1  e   (4.68)
  tr
 

244
 n  i 
t

r   1  e    *r 1   r0

(4.69)
 1  

 (tr )   r   r 1   r (4.70)

  e
*

tr

 * 
 E
r

 r 1/ 2   
r  (4.71)
r r 1
Er 1/2

In the above the following notation is used:

r = identification of temporal increments, r  1..N , where N is number of time
increments for the analysis,
tr  tr  tr 1 = time increment,

 r   r   r 1 = stress increment in time tr ,

 r*   * (tr ) =internal variables at time tr ,

0 0
 r   (tr ) = shrinkage at time tr ,
1
Er 1/ 2   E (tr )  E (tr 1 )  = constant average secant Young modulus at time incremenent
2
tr ,

1 1
E r 1/ 2 
2
 2

E (tr )  E (tr 1 )   Er  Er 1  = constant average value of Dirichlet

coefficient E at tr ,

1
B r 1/ 2   B(tr )  B(tr 1 )  = average value of the matrix B at tr .
2
Equation (4.64) thru (4.71) defines all necessary relations to complete the creep and shrinkage
analysis in ATENA. Of course, they are supplemented by relations used by the short-term
material constitutive model, i.e., equations for calculating the matrix B.
At each time increment, a typical short-term alike analysis is carried. The difference between the
short-term analysis and the described analysis of one step of the creep and shrinkage is that the
latter one uses especially adjusted Young modulus E  r 1/ 2 and initial strain increments  r to
account for creep and shrinkage. After each step, these have to be updated. It involves mainly
update of  and  r . With these values, a new E
r
 r 1/ 2 is calculated and the next temporal
analysis step is carried out.

5.4 Integration and Retardation Times

Appropriate selection of retardation and integration times is of crucial importance for accurate
and efficient creep and shrinkage analysis. The choice of retardation times has a direct impact on
the accuracy of approximation of an original compliance function by Dirichlet series, see

ATENA Theory 245

Equation (4.63) and Fig. 5-2, whilst the choice of integration times affects the accuracy of the
approximation of loading function of the structure, see Equation (4.58) and Fig. 5-1. If the
number of times is too low, some important features of concrete behavior can be disregarded.
The opposite extreme, i.e., using too many retardation or integration times results in worthless
lengthy solution of the problem.
The ATENA software respects recommendation in (Bazant and Whittman 1982). Retardation
times are spread uniformly in log(t ) space and they are automatically calculated as follows:
  i 1
    i  10 m
( i 1   i ),   1, 2..n, i  0,  i  0
 1
(4.72)
   10 m
( 1 ),   1, 2..n

In the above m is the number of retardation times per log(t ) unit, m  1 . By default, this
constant is in ATENA set to 1. If required, a more detailed approximation is possible, i.e., any
value m  1 can be used. In the program, this parameter is input as a number of retardation times
per time unit in logarithmic scale. For a typical concrete creep law, a certain optimal value can
be determined, and it is independent of a structure being analyzed. Note, however, that the value
depends on the choice of time units.
Example: If the retardation times parameter is set to 2, the creep law will be approximated by
two approximation points for the time interval between 0 - 1 day, two points for the interval 1 -
10 days, then two points for 10 - 100 days, etc.
Therefore, the proper values will depend on the choice of time units. If the time unit is a day, the
recommended value is 1 - 2.
Start time  1 must be chosen sufficiently low, so that Dirichlet series can account for processes
in very young concrete right after its loading has been applied. As a default, ATENA uses the
expression  1  0.1 t ' .

As for the upper limit for   , it is required:

t
n  (4.73)
2
The above limits are applicable for the case when the coefficients E (t ') of the Dirichlet series
in (4.63) are calculated by the Least-square method (Jendele and Phillips 1992).
ATENA also supports an alternative way of calculation of the coefficients E (t ') of the Dirichlet
series in (4.63). In this case, Inverse Laplace transformation (Bazant and Xi 1995) is used
instead. This method requires  1  0 , typically 1E-3 and

n  t (4.74)
Comparing the above two approaches, it can be said that the Least-square method yields
approximation of the compliance function at discrete times, whereby Inverse transformation is
based on continuous approach. In some cases, the Least-square method results in better
convergence behavior; however it sometimes suffers from numerical problems during

246
calculation due to an ill-posed problem for solution of E (t ') . It is left to experience and
engineering judgment to decide, which of the method is more appropriate for a particular
solution.
Integration times or sample times tr are calculated in a similar way. In this case, the times are
uniformly spread in log(t  t ') time scale. They are generated starting from the 1st loading time
t ' . Hence, we can write
r  i 1
tr  ti  10 l
 ti 1  ti  , t '  ti (4.75)

where l  2 is the number of time increments per unit of log(t  t ') and ti 1  t ' 0.1  ti  0.1
days. Each new major load increment or decrement causes the generation procedure (4.75) must
start again from small time increments. This parameter defines the number of time steps that the
program will use to integrate the structural behavior. Creep or other nonlinear effects will cause
a redistribution of stresses inside the structure. In order to properly capture such processes, a
sufficiently small time steps are needed. Its definition depends on the type of the analyzed
structure as well as on the choice of time units. For typical reinforced concrete structures and for
the time unit being a day, it is recommended to set this parameter to 2. This will mean that for
each load interval longer then 1 day, two sub-steps will be added. For a load that is interval
longer than 10 days, 4 sub-steps will be added. For an interval longer than 100 days, it will be 6
sub-steps, etc.
The creep and shrinkage analysis in ATENA requires that the user set number of retardation
times m and the number of time increments l per unit of log time, (unless the default values are
OK). He/she also specifies time span, i.e.,  1 and  n . Then, retardation times are generated, i.e.,
an appropriate command is issued. It follows to set stop time of the analysis. Usual input data
describing structural shape, material etc. are given thereafter; however, there are three important
differences from the time-independent analysis:
1. Material model for concrete contains data for long‐term as well as for short‐term material
model.
2. Step data must include information about the time at which the step is applied.
3. It is recommended to input data for all intended load time steps prior to the steps are executed.
It helps the generation of integration (intermediate) times

Intermediate time steps, i.e., times tr as well retardation times are generated automatically. The
analysis proceeds until the stop time is reached. If no stop time is specified, it is assumed to be
the time of the last load step. If the time span for retardation times does not cover step load
times, the solution is aborted, giving an appropriate error message.

5.5 Creep and Shrinkage Constitutive Model

In the above sections, it was silently assumed that the long-term part of the material model, i.e.,
compliance function (t , t ') and shrinkage function  r0 for concrete, is known and it was shown
how it is utilized within creep and shrinkage analysis. It is the primary intention of this section to

ATENA Theory 247

describe what long-term creep and shrinkage prediction models are implemented in ATENA and
how they should be used.
Generally speaking, ATENA applies no restriction on the kind and shape of both  (t , t ') and
 r0 , as it adopts the SBS method solution algorithm, in which compliance function is
approximated by Dirichlet series. Hence, the most widely recognized creep prediction models
could be implemented.
The CCStructureCreep module currently supports the following models:
1. CCModelACI78 (ACI_Committee_209 1978), recommended by ACI,
2. CCModelCEB_FIP78 (Beton 1984), recommended by CEB committee, by now already
obsolete,
3. CCModelB3 (Bazant and Baweja 1999), developed by Bazant and Al Manaseer in 1996,
very efficient model recognized world-wide,
4. CCModelB3Improved, same as the above, improved to account for temperature history,
probably the best model available in ATENA,
5. CCModelCSN731202, model developed by CSN 731202 Code of practice in Czech
Republic,
6. CCModelBP1_DATA (Bazant and Panula 1978; Bazant and Panula 1978; Bazant and
Panula 1978; Bazant and Panula 1978), relatively efficient and complex model; now it is
superseded by CCModelBP_KX or CCModelB3,
7. CCModelBP2_DATA (Bazant and Panula 1978), simplified version of the above model,
8. CCModelBP_KX (Bazant and Kim 1991; Bazant and Kim 1991; Bazant and Kim 1991;
Bazant and Kim 1991), a powerful model with accounts for humidity and temperature
history etc., for practical use it may-be too advanced,
9. CCModelGeneral general model into which experimentally obtained (t , t ') and  r0
function can be input.
10. CCModelEN1992- Eurocode model for creep, (EN1992),
11. CCModelFIB_MC2010- creep model based on CEB-FIP FIB Model Code 2010.

The following data summarized input parameters for the supported models. Note that some
models allow improved prediction based on laboratory data. If it is the case, the model input the
corresponding experimentally measured values. Also, some models can account for material
point history of humidity h(t ) and temperature T (t ) . Again, a model supports this feature if it
can input adequate data.

248
 Table 5.5-1 : List of material parameters for creep and shrinkage prediction – definition and
description

Parameter name Description Units Default

Concrete. type Type of concrete according to ACI. Type 1 1

is Portland cement etc. Types 1,3 accepted
for static analysis, types 1-4 accepted for
transport analysis.

Cement class Type of cement, see e.g. 42,5

http://www.cis.org.rs/en/cms/about-
cement/standardization-of-cement :
Strength classes of cement
Cements are according to standard strength
grouped into three classes, they being:
• Class 32,5
• Class 42,5
• Class 52,5
Three classes of early strength are defined
for each class of standard strength:
• Class with ordinary early strength –
N
• Class with high early strength – R
• Class with low early strength – L
Class L can be applied only on CEM III
cements.
Aggregate Type of aggregate. One of QUART
BASALTDENSELIMESTONE, ZITE
QUARTZITE, LIMESTONE,
SANDSTONE ,
LIGHTWEIGHTSANDSTONE
Thickness V / S Cross section thickness defined as ration of length 0.0767m
section's volume to surface
Strength f cyl 28 Material cylindrical strength in compression stress 35.1MPa
at time 28 days
Strength f cyl 0,28 Strength at onset of nonlinear behaviour in stress Constant
compression at time 28 days from the
base
material

ATENA Theory 249

 Fracture energy G f ,28 Fracture energy at time 28 days stress Constant
from the
base
material
Strength ft 28 Material tensile strength at time 28 days stress Constant
from the
base
material
Young m. E28 Short-term material Young modulus at 28 stress F ( f cyl 28 )
days, i.e. inverse compliance at 28.01 days
loaded at 28 days
Ambient humid. h Ambient relative humidity. Accepted range 0.78
(0.4..1).
Ratio ac Total aggregate/cement weight ratio. 7.04

Ratio wc Water/cement weight ratio. 0.63

Ratio as Total aggregate/find sand weight ratio. 2.8

as  sa 1 .

Ratio sa Fine/total aggregate weight ratio. sa  as 1 0.4

Ratio g s Coarse gravel/fine aggregate weight ratio. 1.3

Ratio sc Fine aggregate/cement weight ratio. 1.8

Shape factor Cross section shape factor. It should be 1, 1.25

1.15, 1.25, 1.3, 1.55 for slab, cylinder,
square prism, sphere, cube, respectively.
Slump Result of material slump test. length 0.1 m
Air content Material volumetric air content. % 5
Cement mass Weight of cement per volume of concrete mass/ 320kg/
length3 m3
Concr. density Material density used to evaluate strength mass/ 2125kg/
and Young modulus at 28 days.. length3 m3
Curing type Curing conditions. It can be either in water AIR
(i.e. WATER) or air under normal
temperature (i.e. WATER) or steamed
curing (i.e. STEAM).

250
 Thermal expansion Thermal expansion coefficient T 1/temp Constant
coefficient T erature from he
base
material
End of curing Time at beginning of drying, i.e. end of days 7
curing.
 a, Autogenous shrinkage at infinity time, - 0
(typically negative!)

 t  ts 
 a   a ,  (0.99  min(0.99, ha ,  ) tanh  
 a 
a Half-time of autogenous shrinkage. days 30

ts Time of final set of cement days 5

ha ,  Final self-desiccation relatibe humidity - 0.8

Current time t Current time days 0

I/D

Load time t' Load time days 0

Tot.water loss w Total water loss (up to zero humidity and kg N/A
infinite time). It is measured in an oven in a
laboratory and it is used to enhance
prediction of shrinkage infinite tsh 2 
(Bazant and Baweja 1999). This value is in
turn used to elaborate drying creep and
shrinkage prediction of the model. If it is
not specified, the model prediction
enhancement is not activated. It can be used,
if water loss w(t) are input as well.
Water loss w(t) Water losses at time t; measured at a kg N/A
laboratory. It is used to enhance
drying creep and shrinkage
prediction. See also description of
total water loss w.
Improvem.

Shrink.  0 (t ) Measured shrinkage at time t. It is used to N/A

enhance drying creep and shrinkage
prediction. See also description of total
water loss w.
Compl. (t , t ') Measured material compliance at time t. It is 1 N/A
used to improve overall creep and shrinkage /stress
prediction of the model.

ATENA Theory 251

 Humidity h(t ) History of humidity in a material point. N/A
Value at time t. Some material models can
use these values to account for real temporal
humidity and temperature conditions.
Although the data can be input manually,
i.e. to group material points with similar
humidity and temperature history into a
group and dedicate a distinct material for
Hist.

that group, it is prepared for full automatic

processing being currently in development.
It will automatically link heat and humidity
transport analysis with the static analysis
using one of available creep and shrinkage
prediction model. Applicable range (0.4..1).
Temperat. T (t ) History of temperature in a material point. Celsia
See also description of h(t )

Compl. (t , t ') Measured compliance at time t loaded at 1/

time t'. This and the next two parameters stress
should be used, if known (measured)
compliance functions are to be employed
in ATENA creep and shrinkage analysis.
Hence, no prediction is done and the given
Direct

data are only used to calculate the

parameters of Dirichlet series
approximation.

Shrink.  0 (t ) Measured shrinkage at time t . See the N/A

parameter above.
Strength f cyl (t ) Measured shrinkage at time t . See the
parameter above

252
 Table 5.5-2: Input parameters needed by individual creep and shrinkage prediction models

Model name B3 B3- BP- CEB ACI CSN BP1 BP2 Gen EN MC
impr KX eral 1992 2010
Model No. 3 4 8 2 1 5 6 7 9 10 11

Concrete. Type x x x x x x x
Cement class x x
Aggregate x x
Thickness x x x x x x x x x x
S /V
Strength f cyl 28 x x x x x x x x x x

Strength x x x
f cyl 0,28

Fracture x x x
energy G f ,28

Strength ft 28 x x x

Young m. E28 x x x x x x x

Ambient x x x x x x x x x x
humid. h
Ratio ac x x x x x x

Ratio wc x x x x x x

Ratio as

Ratio sa x x

Ratio g s x x

Ratio sc x x

Shape factor x x x x x
Slump x
Air content x
Cement mass x
Concr. density x x x x x x

ATENA Theory 253

 Curing type x x x x x x
End of curing x x x x x x x x x x
Thermal x x x
expansion
coefficient T

 a, x

a x

ts x

ha ,  x

Current x x x x x x x x x xx x
I/D

time t
Load time t’ x x x x x x x x x xx x
Tot.water loss x x x
w
Water loss x x
w(t)
Shrink. x x x x x x x x xx x
 0 (t )
I

Compl. x x x
(t , t ')

Humidity x x x xx x
h(t )
Hi

Temperat. x x x xx x
T (t )

Compl. x
(t , t ')

Shrink. x
 0 (t )
Di

Strength x
f cyl (t )

The above parameter "Concrete type" actually referes to a cement type according to the ACI
classification. It used in the creep analysis. The following table brings description of widely
recognized cement types. Note that only types 1,3 are supported in Atena static analysis. The

254
transport analysis in Atena recognizes types 1-4. The remaining types are described just for
information.
Table 5.5-3: Cement types according to ACI classification

ATENA
Concrete Cement type Description
type

General purpose cements suitable for all uses where the

1 I and Type IA 6
special properties of other types are not required.
Type II cements contain no more than 8% tricalcium
aluminate (C3A) for moderate sulfate resistance. Some
2 II and Type IIA6
Type II cements meet the moderate heat of hydration
option of ASTM C 150.
Chemically and physically similar to Type I cements
6
3 III and Type IIIA except they are ground finer to produce higher early
strengths.
Used in massive concrete structures where the rate and
amount of heat generated from hydration must be
4 IV
minimized. It develops strength slower than other cement
types.
Contains no more than 5% C3A for high sulfate
5 V
resistance.
6 IS (X) 7 Portland blast furnace slag cement
7 IP (X)7 Portland-pozzolan cement.
GU 8
8 General use

9 HE8 High early strength

10 MS8 Moderate sulfate resistance
11 HS8 High sulfate resistance
12 MH8 Moderate heat of hydration

6
Air-entraining cements
7
Blended hydraulic cements produced by intimately and uniformly intergrinding or blending two or more types of
fine materials. The primary materials are portland cement, ground granulated blast furnace slag, fly ash, silica fume,
calcined clay, other pozzolans, hydrated lime, and pre-blended combinations of these materials. The letter “X”
stands for the percentage of supplementary cementitious material included in the blended cement. Type IS(X), can
include up to 95% ground granulated blast-furnace slag. Type IP(X) can include up to 40% pozzolans.
8
All portland and blended cements are hydraulic cements. "Hydraulic cement" is merely a broader term. ASTM C
1157, Performance Specification for Hydraulic Cements, is a performance specification that includes portland
cement, modified portland cement, and blended cements. ASTM C 1157 recognizes six types of hydraulic cements.

ATENA Theory 255

 13 LH8 low heat of hydration

5.6 References
ACI_COMMITTEE_209 (2008). Prediction of Creep, Shrinkage and Temperature Effects in
Concrete Structures. Detroit, 209R-92, ACI.
BATHE, K. J. (1982). Finite Element Procedures in Engineering Analysis. Englewood Cliffs,
New Jersey 07632, Prentice Hall, Inc.
BAZANT, Z. AND T. SPENCER (1973). "Dirichlet Series Creep Function for Aging Concrete."
ASCE Journal of Engineering and Mechanical Division: 367-387.
BAZANT, Z. P. (1988). Mathematical Modeling of Creep and Shrinkage of Concrete. New
York, John Wiley & Sons.
BAZANT, Z. P. AND S. BAWEJA, EDS. (1999). Creep and Shrinkage Predicition Model for
Analysis and design of Concrete Structures: Model B3. Creep and Shrinkage of Concrete,
ACI Special Publicatino.
BAZANT, Z. P. AND J. K. KIM (1991). "Improved Prediction Model for Time-Dependent
Deformation of Concrete: Part 1- Shrinkage." Materials and Structures 24: 327-345.
BAZANT, Z. P. AND J. K. KIM (1991). "Improved Prediction Model for Time-Dependent
Deformation of Concrete: Part 2- Basic Creep." Materials and Structures 24: 409-421.
BAZANT, Z. P. AND J. K. Kim (1991). "Improved Prediction Model for Time-Dependent
Deformation of Concrete: Part 3- Creep at Drying." Material and Structures 25: 21-28.
BAZANT, Z. P. AND J. K. KIM (1991). "Improved Prediction Model for Time-Dependent
Deformation of Concrete: Part 4- Temperature Effects." Material and Structures 25: 84-
94.
BAZANT, Z. P. AND L. PANULA (1978). "Practical Prediction of Time-dependent
Deformations of Concrete; Part 1: Shrinkage." Material and Structures 11 (65): 301-316.
BAZANT, Z. P. AND L. PANULA (1978). "Practical Prediction of Time-dependent
Deformations of Concrete; Part 3: Drying Creep." Material and Structures 11 (65): 415-
423.
BAZANT, Z. P. AND L. PANULA (1978). "Practical Prediction of Time-dependent
Deformations of Concrete; Part 4: Temperature Effect on Basic Creep." Material and
Structures 11 (66): 424-434.
BAZANT, Z. P. AND L. PANULA (1978). "Practical Prediction of Time-dependent
Deformations of Time-dependent Deformation of Concrete; Part 2: Basic Creep."
Material and Structures 11 (65): 317-328.
BAZANT, Z. P. AND L. PANULA (1978). Simplified Prediction of Concrete Creep and
Shrinkage from Strength and Mix. Struct. Engng. Report No. 78-10/6405. Evanston,
Illinoins, Northwestern University, Dep. of Civ. Engng.
BAZANT, Z. P. AND F. H. WHITTMAN (1982). Creep and shrinkage in Concrete Structures.
New York, John Wiley & Sons.
BAZANT, Z. P. AND Y. XI (1995). "Continous Retardation Spectrum for Solidification Theory
of Concrete Creep." Journal of Engineering Mechanics 121(2): 281-287.

256
BETON, C. E.-I. D. (1984). CEB Design Manual on Structural Effects on Time Dependent
Behaviour of Concrete. Saint Saphorin, Switzerland, Georgi Publishing Company.
HAGEMAN, L. AND D. YOUNG (1981). Applied Iterative Methods. New York, Academic
Press.
JENDELE, L. AND D. V. PHILLIPS (1992). "Finite Element Software for Creep and Shrinkage
in Concrete." Computer and Structures 45 (1): 113-126.
REKTORYS, K. (1995). Přehled užité matematiky. Prague, Prometheus.
SEAGER, M. K. AND A. GREENBAUM (1988). A SLAP for the Masses, Lawrence Livermore
National Laboratory.

ATENA Theory 257

6 DURABILITY ANALYSIS9
The durability analysis in ATENA can currently assess the deterioration of structures due to
carbonation and chlorides ingress. It is available for static and creep analyses. At each time step,
an appropriate 1D transport analysis is carried out to investigate how far the pollution (i.e.,
carbonation and/or chlorides) penetrate from loaded surfaces inside the structure. The main
results of the analyses are induction times, i.e., times at which the pollution concentration
reaches critical values that are already for the structure unacceptable (e.g., the reinforcement
corrosion begins etc.). They are always given with respect to time t0  0 . In addition, pollution
concentration at times (corresponding to the individual steps) is also computed.
Note that static analysis in ATENA typically does not care about time (or more precisely, each
analysis step increments the structural age by unit time). At each step, it yields a sort of artificial
age of the structure. Hence, if the durability analysis is carried out, this artificial age must be
somehow mapped onto real structural age. It is done in ATENA with the help of a multilinear
function. Such a function corresponds to loading functions used to define variable BCs and it is
input in exactly the same way.
The following text describes the theory behind the 1D transport analysis of the carbonation and
chlorides pollution, and, in the end, some information regarding the transport parameters is
given.
The service life of a structure tl usually has the form of
tl  tc  ti  t p  tr (4.76)

where tc is the construction phase, ti initiation (induction) period, tp propagation period, and tr
post-repair period.
We aim at predicting the initiation period without going into propagation or post-repair phases.
Carbonation and chloride ingress are two leading mechanisms contributing to reinforcement
corrosion. Both of them are described further. The initiation phase ends with the beginning of
reinforcement corrosion. Fig. 6-1 brings a more detailed description of initiation and propagation
phases and their relationship to concrete events. Prediction of the initiation period represents a
preventive measure that is affected above all by concrete cover thickness, concrete composition,
and environment. It makes sense to change the design at the beginning rather than mitigating
reinforcement corrosion later. Acceleration of carbonation and chloride ingress on crack
appearance is taken into account.

9
Not available in ATENA version 5.1 and older. Development/testing implementation of CARBONATION,
CHLORIDES, and ASR in version 5.3.

ATENA Theory 259

 Fig. 6-1 Important events in service life (III 2000).

6.1 Carbonation
Carbonation depth of a sound (uncracked) concrete reads (Papadakis and Tsimas 2002)

2 De, CO 2CO2
xc  t  A1 t (4.77)
0.218(C  kP)
where xc is the carbonation depth, De,CO2 is the effective diffusivity for CO2, C is the Portland
cement content in kgm-3, k<0.3,1.0> is the efficiency factor of supplementary cementitious
material (SCM-slag, silica, fly ash), P is the amount of SCM in kgm-3, CO2 is the volume
fraction of CO2 in the atmosphere taken as 3.6e-4 and t is the time of exposure. The effective
diffusivity in m2s-1 is given by the empirical equation (Papadakis and Tsimas 2002)
3
 
 (W  0.267  C  kP ) / 1000) 
De , CO 2  6.1106   (1  RH ) 2.2 (4.78)
 C  kP W 
  
 c 1000 
where W is the water content in concrete in kgm-3, c is the cement density in kgm-3 assumed as
3150 kgm-3 and RH is the relative humidity of ambient air. Eqs. (4.77)(4.78) allow predicting
either carbonation depth or induction time of uncracked concrete. Relative humidity must be
higher than 0.50 for carbonation to proceed.
Cracked concrete leads to faster carbonation. This acceleration is given in the form (Kwon and
Na 2011)

xc (t )  (2.816 w  1) A1 t (4.79)
where w is the crack width in mm, A1 is the carbonation velocity according to Eq.(4.77).
Eq. (4.79) allows computing carbonation depth and induction time. Note that crack 0.3 mm
increases carbonation depth by a factor of 2.54. This also means that induction time is 6.46 times
shorter compared to a sound concrete.

260
In reality, cracks may grow during any service time. Thus, Eq. (4.79) needs to be recast to
incremental form. An increment of carbonation depth in a given time step t is evaluated from
the total derivative by differentiating Eq. (4.79)

xc  t  
 2.816 
wi 1  1 A1
t 
2.816 A1 t
w (4.80)
2 ti  0.5 2 wi  0.5
where wi+1 is the crack width at the end of the time step, ti+0.5 is the mid-time. It is assumed that
nonzero w at a frozen time t has no effect on carbonation depth; thus the term w can be left
out. Eq. (4.80) allows predicting either carbonation depth or induction time of gradually cracking
concrete.

6.1.1 Example of Carbonation

Let us consider first a regular concrete made from ordinary Portland cement, w/b=0.45, C=400
kgm-3, W=202.5 kgm-3, P=50 kgm-3. The supplementary cementitious material is fly ash with
almost zero calcium content hence k=0.5. Concrete is exposed to relative humidity 0.60.
Consider a concrete cover of 30 mm. A crack is always introduced at the beginning of the
exposure.
The second concrete is made from ordinary Portland cement, w/b=0.45, C=200 kgm-3,
W=90 kgm-3, P=0 kgm-3. Table 6.1-1 compares both concretes in terms of induction time.
Crack width Induction time for concrete Induction time for concrete
(mm) w/b=0.45, C=400 kgm-3, w/b=0.45, C=200 kgm-3, P=0 kgm-3
-3
P=50 kgm (years) (years)
0 246 157
0.1 69.9 44.5
0.2 49.2 31.4
0.3 39.1 24.9
Table 6.1-1. Induction time for carbonation, two concretes, cover thickness 30 mm.

6.2 Chlorides
Implemented model for chloride ingress is based on (Kwon, Na et al. 2009). Let us consider 1D
transient problem of chloride ingress in concrete with initially free chloride content

  x 
C  x, t   CS 1  erf   (4.81)
  2 Dm  t  f ( w)t 
  
where CS is the chloride content at surface in kgm-3, Dm is the averaged diffusion coefficient at
time t in mm2 s-1, x is the position from the surface in mm, and f(w) gives acceleration by
cracking and equals to one for a crack-free concrete. Cs and C can be related to concrete volume
or to binder volume; however, the units must be kept consistently through the computation.
The diffusion coefficient D(t) is assumed to decrease over time t according to the power-law

ATENA Theory 261

 m
t 
D  t   Dref  ref  (4.82)
 t 
where m is a decay rate (sometimes called an age factor). If m=0, a constant value of D(t)=Dref is
recovered. This model was proposed by (Collepardi, Marcialis, et al. 1972). Nowadays, it
became clear that this assumption is too conservative and is not generally recommended. The
mean diffusion coefficient Dm is obtained by averaging D(t) over time of interest
m m
1
t
t  D t 
Dm (t )  Dref  ref  d  ref  ref  , t  t R (4.83)
t0    1 m  t 
m
 t  m    tref 
Dm (t )  Dref 1  R     , t  tR (4.84)
 t  1  m    t R 
where tR is the time when diffusion coefficient is assumed to be constant and is generally taken
as 30 years. tref corresponds to the time when the diffusion coefficient was measured. Fig. 6-2
shows the characteristic evolution of diffusion coefficients over time.
The mean diffusion coefficient increases when cracks are present in the concrete. Based on
recent results, the following scaling function is proposed (Kwon, Na et al. 2009)
f ( w)  31.61w2  4.73w  1 (4.85)
where w stands for crack width in mm. The crack width 0.3 mm increases the mean diffusion
coefficient by a factor of 5.26. In reality, crack width evolves, and incremental solution needs to
be formulated. The mean coefficient Dm,w(t) incorporating crack width is evaluated from a crack
increment
Dm , w (t ) 
w n
  w(ti 1 )  w(ti )   (4.86)
Dm (t )  63.22 w  4.73 dw  Dm  t    63.22  w  ti     4.73 {w(ti 1 )  w(ti )}
0 i 0   2  
If last values of f(w) and w are stored, Eq. (4.86) can be evaluated only in the actual time step.
This speeds up the solution.

262
 Fig. 6-2. Evolution of actual and mean diffusion coefficients for standard concrete, based on data
from (Kwon, Na et al. 2009).

6.3 Diffusion coefficient for chlorides

Proper determination of diffusion coefficient is not a trivial subject, considering various
concretes, cement types, models, and exposure conditions. (Papadakis 2000) presented a model
for estimating intrinsic effective diffusivity for concretes made from blended cements; however,
recalculation to Da is not straightforward. DuraCrete model (III 2000) provides useful data for
estimating apparent diffusion coefficient in the form
m
t 
Da  t   ke kc DCl (t0 )  0   Da (4.87)
t 
where ke<0.27,3.88> is the environment factor, kc<0.79,2.08> is the curing factor, Dcl(t0) is
the measured diffusion coefficient determined at time t0, m<0.2,0.93> is the age factor and
Da<1.25,3.25> is the partial factor. In our notation, Da(t)=Dm(t) and t0=tref.
To our opinion, the most relevant and well-documented field data come from 10 years exposure
tests (Luping, Tang et al. 2007). Fig. 6-3 shows the apparent diffusion coefficient in dependence
of water-binder ratio. In this particular case, tref=10 years, m is unknown, Dref=(1-m)Da, tR can be
assumed as 30 years.

ATENA Theory 263

 Fig. 6-3. Fitted apparent diffusion coefficients from 10-years exposure of concrete (Luping, Tang et
al. 2007).
The next figure shows the apparent diffusivity coefficient at 10 years from Fig. 6-3. They can be
used as a starting point for estimating Dref.

264
 Fig. 6-4. Apparent diffusion coefficients from 10-years exposure of concrete (Luping, Tang et al.
2007).

3.2. Example of chloride ingress

Let us consider regular concrete made from ordinary Portland cement, w/b=0.45. According to
Fig. 6-3, Da is about 2e-12 m2s-1 at tref=10 years. According to the Duracrete model, the age
factor for concrete submerged in salt water corresponds to m=0.30 (Table 8.6 in DuraCrete). In
such case, Dref=(1-m)Da=1.4e-12 m2s-1. Fig. 6-5 shows the evolution of diffusion coefficients for
this particular case.

ATENA Theory 265

 Fig. 6-5. Evolution of diffusion coefficients for chlorides in an example.

Let us assume characteristic value Cs 10.3% of chlorides per binder for submerged concrete
without further reductions (Table 8.5 in DuraCrete). The critical level for corrosion is 1.85 % per
binder (Table 8.7 in DuraCrete). The concrete cover is taken as 100 mm. Computed induction
time according to Eq. (4.81) is summarized in Table 6.3-1. Crack width is considered since the
beginning of the exposure.

Table 6.3-1 Induction time for chloride corrosion of submerged concrete, in dependence on original
crack width.

Crack width Induction time (years)

(mm)
0 74.58
0.1 36.02
0.2 15.70
0.3 7.76

6.4 MODELS for PROPAGATION PHASE

6.4.1 Carbonation during propagation phase

The corrosion rate for the carbonation depends on the corrosion current density icorr [µA/cm2],
which ranges between 0.1-10 (passive corrosion-high corrosion) and depends on the quality and
the relative humidity of the concrete (Page CL, 1992). This model predicts the amount of
corroded steel during the whole propagation period tp. The corrosion rate is based on Faraday's
law (Rodriguez, 1996), determined as follows:
xcorr (t )  0.0116icorr  t  (4.88)

266
where xcorr is the average corrosion rate in the radial direction [m/year], icorr is corrosion current
density [µA/cm2], and t is the calculated time after the end of the induction period [years].
By integration of Eq. (1), it is obtained the corroded depth for 1D propagation:
t
xcorr (t )   0.0116icorr  t  Rcorr dt (4.89)
tini

where xcorr is the total amount of corroded steel in radial direction [mm] and Rcorr is parameter,
depends on the type of corrosion [-]. For uniform corrosion (carbonation) Rcorr = 1, for pitting
corrosion (chlorides) Rcorr = <2; 4> according to (Gonzales at.al., 1995) or Rcorr = <4; 5.5>
according to ( Darmawan &, 2007).
Effective bar diameter for both types of corrosion is obtained from:
d  t   dini   2 xcorr (t ) (4.90)

where d(t) is the evolution of bar diameter in time t, d ini is initial bar diameter [mm], ψ is
uncertainty factor of the model [-], mean value ψ = 1 and xcorr is the total amount of corroded
steel according to (2).

6.4.2 Chloride ingress during propagation phase

The corrosion rate for chlorides is more complicated because it is affected by the concentration
of chlorides in the concrete. Calculation of corrosion current density was formulated by Liu and
Weyer's model (Liu, Weyers, 1998):

 3006 
icorr  0.926 * exp 7.98  0.7771ln 1.69Ct    0.000116 RC  2.24t 0.215  (4.91)
 T 
where icorr is corrosion current density [µA/cm2], Ct is total chloride content [kg/m3 of concrete]
on reinforcement which is determined from 1D nonstationary transport, T is temperature at the
depth of reinforcement [K] and Rc is ohmic resistance of the cover concrete [Ω] (Liu, 1996) and t
is time after initiation [years]:

RC  exp 8.03  0.549ln 1  1.69Ct   (4.92)

The average corrosion rate in radial direction is determined further when plugging(4.93),(4.94)
to (1). The total amount of corroded steel in radial direction stems from (2) and the effective bar
diameter from (3).

6.4.3 Cracking of concrete cover

The cracking of concrete cover for both carbonation and chlorides can be estimated from
DuraCrete model, which provides realistic results (DuraCrete, 2000). The critical penetration
depth of corroded steel xcorr,cr is formulated as:
C
xcorr , cr  a1  a2  a3 f t , ch (4.95)
dini

ATENA Theory 267

where parameter a1 is equal 7.44e-5 [m], parameter a2 is equal 7.30e-6 [m], a3 is
[-1.74e-5 m/MPa], C is cover thickness of concrete [m], dini initial bar diameter [m], ft,ch is
characteristic splitting tensile strength of concrete [MPa].

6.4.4 Spalling of concrete cover

The critical penetration depth of corroded steel xcorr,sp for both carbonation and chlorides is
calculated from (DueaCrete, 2000) as:

wd  w0
xcorr , sp   xcorr , cr (4.96)
b
where parameter b depends on the position of the bar (for top reinforcement 8.6 µm/µm and
bottom 10.4 µm/ µm), wd is critical crack width for spalling (characteristic value 1 mm), w0 is
the width of initial crack (known from previous ATENA computation) and xcorr,cr depth of
corroded steel at the time of cracking [m].
After spalling of concrete cover, corrosion of reinforcement takes place in direct contact with the
environment. To determine the rate of corrosion of reinforcement after spalling, (Spec-net, 2015)
gives rates of reinforcement corrosion.

Table 2: Corrosion rates of steel under atmospheric exposition

Corrosivity zone (ISO 9223) Typical environment Corrosion rate for first year (µm/yr)
Category Description Mild steel Zinc
C1 Very low Dry indoors ≤1,3 ≤0,1
C2 Low Arid/Urban inland >1,3 a ≤25 >0,1 a ≤0,7
C3 Medium Coastal and >25 a ≤50 >0,7 a ≤2,1
industrial
C4 High Calm sea-shore >50 a ≤80 >2,1 a ≤4,2
C5 Very High Surf sea-shore >80 a ≤200 >4,2 a ≤8,4
CX Extreme Ocean/Off-shore >200 a ≤700 >8,4 a ≤25

6.5 Alkali-Aggregate Reaction

6.5.1 Introduction of alkali-aggregatea model for concrete

268
In most concrete, aggregates are more or less chemically inert. However, some aggregates react
with the alkali hydroxides in concrete, causing expansion and cracking over a period of many
years. This alkali-aggregate reaction has two forms: alkali-silica reaction (ASR) and alkali-
carbonate reaction (ACR).
Alkali–silica reaction (ASR), one of those common deleterious mechanisms, consists of a
chemical reaction between "unstable" silica mineral forms within the aggregate materials and the
alkali hydroxides (Na, K–OH) dissolved in the concrete pore solution. It generates a secondary
alkali-silica gel that induces expansive pressures within the reacting aggregate material(s) and
the adjacent cement paste upon moisture uptake from its surrounding environment, thus causing
micro cracking, loss of material's integrity (mechanical/durability), and, in some cases,
functionality in the affected structure.
Several aggregate types in common use, particularly those with a siliceous composition, may be
attacked by the alkaline pore fluid in concrete. This attack, essentially a dissolution reaction,
requires a certain level of moisture and alkalis (leading to high pH) within the concrete to take
place. During the reaction, a hygroscopic gel is produced. When imbibing water, the gel will
swell and thus cause expansion, cracking, and in the worst case, disruption of the concrete
(Lindgart 2012).
Thus, the degree of reaction of an aggregate is a function of the alkalinity of the pore solution.
For a given aggregate, a critical lower pH-value exists below which the aggregate will not react.
Consequently, ASR will be prevented by lowering pH of the pore solution beneath this critical
level where the dissolution of alkali-reactive constituents (silica) in the aggregates will be
strongly reduced or even prevented, as discussed in (Rodriguez at.al, 1996). No "absolute" limit
is defined because the critical alkali content largely depends on the aggregate reactivity [3], but
from many experimental tests we can estimate threshold value (Lindgart 2012), (Poyet , 2003).
Many studies carried out over the past few decades have shown that ASR can affect the
mechanical properties of concrete as a "material." Usually, ASR generates a significant reduction
in tensile strength and modulus of elasticity of concrete. These two properties are much more
affected than compressive strength, which begins to decrease significantly only at high levels of
expansion.
Several ASR models were developed over the years to predict expansion and damage on both
ASR affected materials (microscopic models) (Multon at.al., 2009), (Bazant, Steffens, 2009),
(Comby-Perot, 2009) and ASR affected structures/structural elements (macroscopic models)
(Ulm at.al., 1999), (Saouma, Perotti,2006), (Comi, Fedele, Perego, 2009). The first group has a
goal of modeling both the chemical reactions and the mechanical distress caused by ASR or even
the coupling of the two phenomena. The second group aims at understanding the overall distress
of structures/structural concrete elements in a real context, simulating their likely in situ behavior
(Farage et al.,2000) seems to have finally bridged the gap between scientific rigor and practical
applicability to real structures.
In terms of mechanical effects, it is known that ASR expansions occur over long time periods.
During this process, ASR-affected concretes are subjected to a progressive stress built up that is
very likely to cause creep on the distressed materials.

ATENA Theory 269

AAR depends on the availability of three factors: alkalis liberated from cement during hydration,
siliceous minerals present in certain kinds of aggregates, and water. Several microscopic and
random factors are involved in AAR expansion, such as concrete porosity, amount and location
of reactive regions in the material, and permeability (Farage et al.,2000). These parameters,
added to concrete's intrinsic heterogeneity, turn simulating the AAR expansion into a rather
complex task.
Even though the AAR process has not been well explained so far, the commonly accepted theory
for describing it is two distinct phases that need to be considered: gel formation and water
absorption by the gel, causing expansion. According to this mechanism, the reaction does not
always lead to expansion. As long as there is enough void space to be filled by the gel, i.e., pores
and cracks, concrete volume remains unchanged.
Due to the lack of a model, which is able to incorporate effects of relative humidity, alkali/silica
content in the mixture, ambient temperature, authors suggest to combine ASR kinetics proposed
by (Ulm et. al., 1999) with the influence of moisture, published by (Léger et al., 1996) and
influence of alkali/silica content proposed by Multon et al.
Implementation of modeling expansion due to ASR consists of modeling engeinstrains in time-
steps t on the entire structure. Function for volumetric eigenstrain reads
 ASR  t    cal

  t  FM (4.97)

where  cal

is the volumetric strain of ASR swelling at infinity time,   t   0,1 is the chemical
extent of ASR, and FM is the coefficient reflecting moisture influence. It is described later in the
text. In the case of varying the relative humidity, eq. (4.97) changes to the incremental form, for
time ti

 ti  ti 1 
 ASR  ti    ASR  ti 1    cal

  ti     ti 1   FM   (4.98)
 2 

6.5.2 Model for ASR kinetics

For the complete 3D constitutive model, we consider the first-order reaction

1  ξ  tc  ,ξ  ξ (4.99)

where tc  ,ξ   kd / A0 is the characteristic time. It has been found that tc depends on

temperature  [ K ] and the ASR extent ξ . Referring to (4.99) the implementation of the
chemoelastic material law in the constitutive laws is relatively straightforward and a suitable
integration scheme is given in (Ulm ea., 1999).
Consider an isothermal stress-free ASR expansion test carried out at constant temperature
   0 . In this test, the volumetric strain  ASR is recorded as a function of time that and ASR
extent is calculated as
 ASR  t 
ξ t   (4.100)
 ASR   

270
For macroscopically stress-free sample, (4.99) in (4.100) yields
ASR  t 
 ASR    (1  ξ)   ASR    tc  ,ξ 
 ASR   
 ASR     ξ ASR     tc  ,ξ  ASR  t  (4.101)
 ASR      ASR  t   tc  ,ξ  ASR  t 

With  ASR (t ) and ASR (t ) being measurable functions of time, the characteristic time tc can be
determined from a stress-free expansion test. In a recent extensive series of stress-free expansion
tests carried out at different constant temperatures( Larive, 1998), tc has been found to depend on
both temperature  [ K ] and reaction extent ξ [-] in the form

tc   c     ,  (4.102)

1  exp   L   /  C   
  ,   (4.103)
ξ  exp   L   /  C   

In this experimentally determined kinetics function,  c   is a characteristic time [day] and

 L   is a latency time [day]. The use of (4.103),(4.102) in (4.101) yields after integration

1  exp  t /  C 
ξ t   (4.104)
1  exp  t /  C   L /  C 

For variable temperature, cracking etc., it is difficult to solve for ξ  t  analytically and numerical
integration is needed. A suitable solution scheme is derived in (Ulm at.al., 2006), which is
implemented in in ATENA. Fig. 6-6 shows the shape of (4.100), together with the time
constants,  c  C and  L , which stand for the characteristic time and the latency time of ASR
swelling, respectively. Furthermore, proceeding as in physical chemistry (Atkins, 1994), we
explore the temperature dependence of the time constants  c  C and  L from stress-free
expansion tests carried out at different constant temperatures. The plots of ln( c )  C nd ln( L )
against 1 /   C are given in Fig. 6-7. It is remarkable that the experimental values align (almost)
perfectly along a straight line, matching the Arrhenius concept.

 1 1 
 C     C  0  exp U c     (4.105)
    0  

 1 1 
 L     L  0  exp U L     (4.106)
    0  
where
U C  5400  500 K ; U L  9400  500 K (4.107)

ATENA Theory 271

It is explored (Atkins, 1994) that the temperature dependence of the time constants  c  C nd
 L carried out at different constant temperatures (23, 33, 38, and 58 °C), see Fig. 6-6. Default
values are τ C  311,15K  ays and  L (311.15C)  145 days [20], see Fig. 6-8. , Fig. 6-9.
According to Larive's experimental data from water-saturated tests [14] τ C  288,15K  days and
τ L  288,15K  ays, τ C  281,15K  ys and τ L  288,15K  ays. Under drying conditions, the values for
 L τ L roughly increase by a factor of 4; and  C by 2.5 (Larive , 1998), (Ulm at.al, 1999)

Fig. 6-6. Larive's test data of temperature dependency of ASR time constants τ C nd τ L Slope of
trendlines represents activation energy constants Uc = 5,400 K and UL = 9,700 K.

272
Fig. 6-7. Definition of Latency Time τ L d Characteristic Time τ C Normalized Isothermal Expansion

Curve ξ   t  /    

Fig. 6-8. Parameter Analysis of Characteristic Time τ C ( 311 K) of ASR Swelling with Regard to Hydral
Ambient Conditions, reproduced from 0.

Fig. 6-9. Parameter Analysis of Latency Time τ L ( 311 K) of ASR Swelling with Regard to Hydral
Ambient Conditions, reproduced from 0.

6.5.3 Prediction of ASR swelling  cal



 cal

[-] is the predicted volumetric expansion at infinity time obtained by model proposed by
(Multon et al., 2008). It is calculated based on reactive aggregates, amount of reactive silica in
the aggregates, and value of measured stress-free expansion test done in Poyet's study (Lindgart ,
2012) on samples containing reactive particles only.  cal

is defined as follows

ATENA Theory 273

 AC
 cal

 t   s  p  AC  F (4.108)
AR
where  F [m3/kg] is measured ASR strain expansion per kg of aggregate in m3 of the concrete
mixture on samples containing reactive particles only with enough sufficiency of alkali.
Typically it ranges in 8.93e-7 ... 1.34e-5 [m3/kg]. See Table 3 for more details. AC kg/m3
Na2Oeq] and AR kg/m3 Na2Oeq] are amounts of consumed and required alkali, respectively. AC is
total aggregate content in [kg/m3]. One of the main assumptions of the model is that the
maximum expansion of mortar is achieved if there is enough alkali to react with all the reactive
silica of the mixture. This amount of required alkali content AR kg/m3 Na2Oeq] is defined as

AR  r  s  p  AC (4.109)
where s is the proportion of quantity of soluble silica [-], p is the proportion of reactive aggregate
[-]. r states for the amount of required alkali per kg of reactive silica, and it is a constant value r
= 15.4 %. Value AC s defined as min AR (, AA s the available amount of alkali for ASR reaction.
AA s defined as the difference between the initial amount of available alkali AT kg/m3 Na2Oeq]
and alkali content threshold A0 kg/m3 Na2Oeq] when ASR reaction starts.

AA  AT  A0 (4.110)
It should be noted that this model does not consider any alkali flow through boundaries inside
the structure during the service life. By default, A0 s equal to 3.7 kg/m3 Na2Oeq (Poyet, 2003), but
other values in the range of 3 – 5 kg/m3 Na2Oeq can be found in the literature (Lindgart, 2012)

Table 3: Mixtures and ASR expansions of mortars studied by (Poyet, 2003) and (Multon, 2008). F1-F3
are size fractions 80 μm-3.15 mm.

Value of p depends on the mix ratio of reactive aggregate. Value s depends on amount of
reactive silica in aggregates, moreover common values are: p = 11,1% (Multon, 2008) 0or 9,4%
and 12,4% (Multon, 2009).

6.5.4 Influence of moisture FM

Approximately 75% relative humidity (RH) within concrete is necessary to initiate significant
expansion, which is assumed to vary linearly between 75% RH and 100% RH as shown in Fig.
6-10.

274
Fig. 6-10. Parameter Factor of RH influencing ASR concrete expansion, reproduced from (Multon,
Toutlemonde, 2010).

The coefficient FM eflects influence of moisture h. The function for FM is approximated as

1
FM (h)  (h  hmin ) (4.111)
1  hmin

where hmin is relative humidity threshold where ASR begins to appear, 0.75 by default. Other
variables will be explained in further text.

6.5.5 ASR for 3D conditiions

Expansion of free concrete specimens due to ASR has been summarized in (Červenka, Jendele,
Šmilauer, 2016). It predicts ASR under unrestrained conditions, i.e., under free expansion. The
expansion model takes into account reaction kinetics, alkali content, reactive amount of
aggregates, relative humidity, and temperature. The model has been validated on 4 examples
found in the literature.
Degradation of material due to ASR reaction (Saouma, 2016, eqs. 18,19)

E  t ,   E0 1- 1-  E    t ,   (4.112)

ft  t ,   f t ,0 1- 1-  f    t ,   (4.113)

G f  t ,   G f ,0 1- 1-  G    t ,   (4.114)

ATENA Theory 275

where βE,f,G are residual values of E/E0, f/f0, Gf/Gf0. Default values are  E  0.1 ,
 f  0.6 (Esposito, Hendriks, 2012) and G  0.6 is estimated.

The general equation for the incremental volumetric AAR strain is given by (Saouma, 2016, (5))
V  t   I  t   II  t   III  t  



t  c  , f c FM  h    t   cal

 
 t  c  , f c FM  h    t   cal

   t , , fc, FM , (4.115)
Only considered in implementation

where  c reflects the effect of compressive stresses (Saouma, 2016, eq. 10), t accounts for the
influence of tensile cracking ( assumed here as 1), FM is the effect of relative humidity, which is
already accounted for in (4.111) and equals to one. (4.115) considers further only the most
relevant first term and is rewritten in incremental form as

 
 V  ti   t   c  i 1 , f ci1  FM  hi    cal

   ti     ti 1  
(4.116)
i  (i  (i  1)) / 2

Reduction  c due to compressive stress is considered as follows:

 1 if   0 Tension

c   e
1 if  <0 Compresion
 1  1  e   (4.117)

   II   III
 I
3 f c

where the shape factor  is ‐2 by default (Saouma, 2016, Tab.2) and f c' is the compressive strength.

Under constrained conditions, ASR expansion develops depending on the stress state. It is
known that compressive stress beyond approximately -10 MPa stops ASR expansions, which
needs to be reflected for strain redistribution into the principal directions. Similarly to (Saouma,
2016, Fig. 5), weight factors are assigned to three directions. Let us assume that directions of
principal stresses σI, σII, σIII are known. Expansion is then assigned to each principal stress
direction according to the weight factors W1 ', W2 ', W3 '. When compressive stress reaches -0.3
MPa, the weight factor decreases until maximum stress -10 MPa is reached in that direction.
This situation is depicted in Fig. 6-11.

276
Fig. 6-11. Weight factor for ASR expansion

For compressive stress σi under -0.3 MPa, the following decay function is used, according to
(Leger, Coté, Tinawi, 1995), where σL ≈ -0.3 MPa and σu ≈ -10 MPa, see Fig. 6-11. :

 
 
 1  1
 log  i /  L   for  i  0.3MPa
Wi  i     u   (4.118)
  log    
   L  
 1 for  i  0.3MPa

Weight factors need to be normalized as

Wi '
Wi  (4.119)
 i 1Wi '
3

Three principal strains from ASR are assigned as

 ASR ,i  Wi   V  ti  (4.120)

This new approach simplifies the procedure outlined by (Saouma, 2016, Fig. 5) where several stress
state cases were treated individually.

6.5.6 Validation on free expansion

The following Fig. 2-12 and Fig. 6‐13 validate experimental data for free expansion. The
following material parameters were used, summarized in Table 6.5‐4.

Variable Symbol Value Source

(Multon, Cyr,
REQUIRED ALKALI PER REACTIVE Sellier,
r 15.4 %
SILICA Leklou, &
Petit, 2008)

ATENA Theory 277

 (Multon, Cyr,
Sellier,
PROPORTION REACTIVE SILICA s 21.8 %
Leklou, &
Petit, 2008)
(Multon, Cyr,
PROPORTION REACTIVE PARTICLES Sellier,
p 30 %
IN SAND Leklou, &
Petit, 2008)
(Kagimoto,
Yasuda, &
SAND MASS AC 833 kg/m3
Kawamura,
2014)
ASR MEASSURED ASR STRAIN ƐF 0.0525 %/kg (Poyet, 2003)
AMOUNT OF REQUIRED ALKALI AR 8.39 kg/m3 (Poyet, 2003)
(Kagimoto,
TOTAL ALKALI IN MORTAR for Ca-5.4 Yasuda, &
AT 5.4 (9) kg/m3
(for Ca-9.0) Kawamura,
2014)
THRESHOLD ALKALI IN CONCRETE A0 3.7 kg/m3 (Poyet, 2003)
CHARACTERISTIC TIME τC 20 day
LATENCY TIME for Ca-5.4 (for Ca-9.0) τL 55 (45) day
ELASTIC MODULUS E (Kagimoto,
Yasuda, &
27 GPa
Kawamura,
2014)
COMPRESSIVE STRENGTH fc (Kagimoto,
Yasuda, &
26 MPa
Kawamura,
2014)

278
 Table 6.5‐4. Summarized parameters for validation.

Fig. 6‐12. Validation of free expansion (Kagimoto, Yasuda, & Kawamura, 2014)

Fig. 6‐13. Validation of free expansion, (Kagimoto, Yasuda, & Kawamura, 2014)

6.5.7 Implementation in Atena

Differential Equation (4.99) represents kinetics of development of ASR extent ξ . In the case of
constant temperature  in the structure, it can be solved analytically, see (4.104). Otherwise, it
must be solved numerically. The following lines and equations describe the procedure to solve ξ
that is implemented in ATENA.
Let's start from (4.99) and rewrite the equation into its differential form. We expect to now all at
the time i and solve for time i  1 . We do it in an iterative manner, i.e., we know all at iteration
k and compute ξ at iteration k  1 :

ATENA Theory 279

 ξ ik11  ξ i
tck, i 1  1  ξ ik11  = 0 (4.121)
ti  1  ti

The unknown ξ ik11 ASR extent is searched for in the form ξ ik11  ξ ik1   ξ , where  ξ is the
correction of ξ resulting from the k -th iteration. Denoting ti 1  ti  t and ξ ik1  ξ i  ξ the
above equation can be written

tck,i 1  ξ  ξ   t 1  ξ ik1   ξ  = 0 (4.122)

from which, after some mathematical manipulation, we can calculate  ξ

ξ tc , i 1  t 1  ξ ik1 
ξ   (4.123)
tck, i 1  t

and ξ ik11  ξ ik1   ξ . Note that in (4.121) thru (4.123) we used tck, i 1 although tck,i 11 should be
employed, as tc  tc ( , ) is a nonlinear function. Therefore, after each iteration, k  1 we update
tck, i 1 to tck,i 11 and recalculate (4.122)

tck,i 11  ξ  ξ   t 1  ξ ik1   ξ  = E k 1 (4.124)

It yields an error E k 1 that is further compared against some maximum acceptable error. If it is
too high, the next iteration is carried out; otherwise, the iteration process is finshed.
Note, however, that for the sake of convergency speed, the third and further iterations are in
ATENA computed in a different way. Using linear interpolation between iteration k and k  1
requiring error E k  2  0 in iteration k  2 value ξ ik12 is calculated by

1
E k 2  k 
 E k 1  ξ ik12  ξ ik1   E k  ξ ik12  ξ ik11    0
ξ k 1
 ξ i 1 
i 1

E k 1  ξ ik12  ξ ik1   E k  ξ ik12  ξ ik11   0 (4.125)

ξ ik1E k 1  ξ ik11E k
ξ ik12 
E k 1  E k
and checked by (4.124) written for iteration k  2 . The iterating process continues this (latter)
way until a sufficient accuracy is obtained.
The time step t is input by the user, but it is automatically limited by t  0.01tc requirement
to ensure reasonable accuracy and convergence of the solution.

ASR loading results in the development of ASR strain and deterioration of material properties
like Young modulus E, tension strength ft, and fracture energy Gf. For each step i, we can write
 i   i 1  Ei 1 ( i   i 1 )   i 1 ( Ei  Ei 1 ) (4.126)

280
The above equation calculates stress  i at (current) time step i based on stress  i 1 from the
previous time step and current changes of Young modulus E and strains  . The strains 
represents "mechanical strains," i.e., strains producing stresses in an unrestrained material. They
are total geometrical strains minus initial strain that corresponds to ASR expansion strains  ASR ,i .
The differential formulation corresponds to the incremental solution used in Atena and the case
of linear elastic material law. (More advanced materials are treated in a similar way). Using
Ek  E0 cf EASR ( k ) , (4.126) can be written

 i   i 1  E0 cf EASR (i 1 )  i   i 1 E0  cf EASR ( i 1 , i )

 i   i 1  E0 cf EASR (i 1 )( i   i 1 )   i 1 E0 (cf EASR (i )  cf EASR (i 1 ))
 cf EASR (i ) 
 i   i 1  E0 cf EASR (i 1 )( i   i 1 )   i 1 E0 cf EASR (i 1 )   1 (4.127)
 cf E (i 1 ) 
ASR

 cf EASR (i ) 
 i   i 1  E0 cf ASR
(i 1 )( i   i 1 )   i 1  ASR  1
 cf E (i 1 ) 
E

Note that strains  are strains that are facilitated in material law, i.e., geometrical strains after
subtracting ASR swelling strains. The ASR strains are implemented by initial element strains,
  
and the term  i 1  i  1 is incorporated in the solution in the form of element initial stresses.
 i 1 
Also, at each step, we update ft and Gf.
An alternative solution to (4.127) is
 i   i 1  E0 cf EASR (i 1 )( i   i 1 )   i 1 E0 (cf EASR (i )  cf EASR (i 1 ))
 cf EASR (i ) 
 i   i 1  E0 cf (i 1 )( i   i 1 )   i 1 E0 cf (i 1 )  ASR
ASR ASR
 1
 cf E (i 1 ) 
E E

(4.128)
  cf EASR (i )  
 i   i 1  E0 cf E (i 1 ) ( i   i 1   i 1 1  ASR
ASR

  cf E (i 1 )  

 cf EASR (i ) 
The term  i 1  1  ASR  is then added to ASR swelling initial element strains (4.120)
 cf E (i 1 ) 
calculated earlier. For linear material law, both equations (4.127) and (4.128) are equivalent. For
the case of nonlinear law, they can slightly differ. By default, Atena prefers approach according
to (4.128).
For the sake of simplicity, the above derivation has been presented for uniaxial stress-strain
conditions. Its extension to 3D conditions is obvious.

6.5.8 Comments

The proposed model is derived from free expansion tests. The model works in 2D and 3D stress
state by limiting expansion when a compressive load is present in a principal direction. In the

ATENA Theory 281

case of hydrostatic compression above -10 MPa, no ASR expansion occurs, and no reduction of
mechanical properties happens (E, ft, Gf). This is justified by the fact that ASR gel grows into
cracks and no macroscopic cracks occur.
The majority of structures are exposed to the thermal field; hence ASR usually proceeds faster
close to the surface due to higher average temperature. Since the surface is often unloaded, the
main expansion happens perpendicular to the surface, which induces a small compressive load
and delamination of layers.

6.6 References

Atkins, P. W. (1994). Physical chemistry, 5th Ed., Oxford University Press, Oxford, U.K.
Berra, M. et al., Influence of stress restraint on the expansive behavior of concrete affected by
alkali-silica reaction, Cement and concrete research 40, 1403-1409, 2010
Bazant, Z.P., Steffens, A. Mathematical model for kinetics of alkali–silica reaction in concrete.
Cem Concr Res 2000;30:419–28.M. S. Darmawan & M. G. Stewart, Effect of Pitting
Corrosion on Capacity of Prestressing Wires, Magazine of Concrete Research, 59(2),
131-139, 2007.
Collepardi, M., A. Marcialis, et al. (1972). "Penetration of Chloride Ions into Cement Pastes and
Concrete." Journal of the American Ceramic Society 55: 534-535.
Comby-Perot, I., Bernard F., Bouchard P.O., Bay, F., Garcia-Diaz, E. Development and
validation of a 3D computational tool to describe concrete behaviour at mesoscale.
Application to the alkali–silica reaction. Comput Mater Science 2009;46(4):1163–77.
Červenka, J, Jendele, L., Šmilauer, V: Report I-05-01-2016 TAČR - TA04031458, 2016
Comi, C., Fedele, R., Perego, U. A chemo–thermo-damage model for the analysis of concrete
dams affected by alkali–silica reaction. Mech Mater 2009;41:210–30.
III, T. E. U. B. E. (2000). DuraCrete Final Technical Report. General Guidelines for Durability
Design and Redesign. Doc. BE95-1347/R17, 2000
Esposito, R., Hendriks, M.A.N., Degradation of the mechanical properties in ASR-affected
concrete: overview and modeling, Strategies for Sustainable Concrete Structures, 2012
Farage, M. C. R. Modelagem e implementaca˜o nume´rica da expansa˜o por reaca˜o a´lcali–
agregado do concreto, DSc thesis, Department of Civil Engineering, COPPE,
Universidade Federal do Rio de Janeiro, Rio de Janeiro, 2000 (in Portuguese).
J. A. Gonzales, C. Andrade, C. Alonso & S. Feliu, Comparison of Rates of General Corrosion
and Maximum Pitting Penetration on Concrete Embedded Steel Reinforcement, Cement
and Concrete Research, 25(2), 257-264, 1995.
Kwon, S.-J. and U.-J. Na (2011). "Prediction of Durability for RC Columns with Crack and Joint
under Carbonation Based on Probabilistic Approach." Int. Journal of Concrete Structures
and Materials 5(1): 11-18.
Kwon, S. J., U. J. Na, et al. (2009). "Service Life Prediction of Concrete Wharves with Early-
aged Crack: Probabilistic Approach for Chloride Diffusion." Structural Safety 31(1): 75-
83.

282
Larive, C. (1998). ‘‘Apports combinés de l’expérimentation et de la modélisation á la
comprehension de l’alcali-reaction et de ses effets mécaniques.’’ Monograph LPC, OA
28, Laboratoires des Ponts et Chaussées, Paris (partially translated into English).
Leger, P., Coté, P., Tinawi, R., Finite element analysis of concrete swelling due to alkali-
aggregate reactions in dams, Computer and structures, 1995
Léger, P., Coté, P. and Tinawi, R. (1996) Finite Element Analysis of Concrete Swelling due to
Alkali-aggregate Reactions in Dams, Comparers & Structures Vol. 60. No. 4. pp. 601-
611.
Lindgart, J., (2012), Alkali–silica reactions (ASR): Literature review on parameters influencing
laboratory performance testing. Cement and Concrete Research, 42, 223-243.
Liu Y, Weyers R.E., Modeling the Dynamic Corrosion Process in Chloride Contaminated
Concrete Structures, Cement and Concrete Research, 28(3), 365-367, 1998.
Liu Y., Modelling the Time-to-corrosion Cracking of the Cover Concrete in Chloride
Contaminated Reinforced Concrete Structures, Ph.D. dissertation, Virginia Polytechnic
Institute, 1996.
Luping, Tang, et al. (2007). Chloride Ingress and Reinforcement Corrosion in Concrete under
De-Icing Highway Environment – A Study after 10 Years' Field Exposure, SP Sveriges
Tekniska Forskningsinstitut. Vol. SP Report 2007:76.
Multon, S., Cyr, M., Sellier, A., Leklou, N., Petit L. (2008) Coupled effects of aggregate size and
alkali content on ASR expansion, Cement and Concrete Research 38, 350-359.
Multon, S., Toutlemonde, F. (2010) Effect of moisture conditions and transfers on alkali silica
reaction damaged structures, Cement and Concrete Research 40, 924–934.
Multon, S., Sellier, A., Cyr, M. Chemo–mechanical modeling for prediction of alkali silica
reaction (ASR) expansion. Cement Concrete Research 2009;39:490–500.
Lindgart, J., (2012), Alkali–silica reactions (ASR): Literature review on parameters influencing
laboratory performance testing. Cement and Concrete Research, 42, 223-243.
Papadakis, V. G. (2000). "Effect of Supplementary Cementing Materials on Concrete Resistance
Against Carbonation and Chloride Ingress." Cement Concrete Research 30(2): 291-299.
Papadakis, V. G. and S. Tsimas (2002). "Supplementary Cementing Materials in Concrete. Part
I: Efficiency and Design." Cement and Concrete Research 32(10): 1525–1532.
Page CL, Nature and properties of concrete in relation to environment corrosion, Corrosion of
Steel in Concrete, Aachen, 1992.
S. Poyet, Etude de la dégradation des ouvrages en béton atteints par la réaction alcali-silice:
Approche expérimentale et modélisation numérique multi-échelles des dégradations dans
un environnement hydro-chemomécanique variable, Ph.D. Thesis (in French), Université
de Marne-La-Vallée, 2003
J. Rodriguez, L. M. Ortega, J. Casal & J. M. Diez, Corrosion of Reinforcement and Service Life
of Concrete Structures. In Proc. of Int. Conf. on Durability of Building Materials and
Components, 7, Stockholm, 1:117-126, 1996.
Saouma, V., Perotti, L., 2006, Constitutive model for alkali-aggregate reaction. ACI Material
Journal 103, 194-202.
Spec-net, Corrosivity zones for steel construction [online] available from: http://www.spec-
net.com.au/press/1014/gaa_081014/Corrosivity-Zones-for-Steel-Construction-
Galvanizers-Association, 2015.

ATENA Theory 283

Ulm, F. J., Coussy, O., Li, K., Larive, C. Thermo–chemo–mechanics of ASR expansion in
concrete structures. J Eng Mech 1999;126(3):233–42. Saouma, V., Perotti, L.
Constitutive model for alkali-aggregate reactions. ACI Mater J 2006;103(3):194–202.

284
7 TRANSPORT ANALYSIS
As pointed out in the previous section, creep material behavior of concrete strongly depends on
moisture and temperature conditions. Some constitutive models for creep in ATENA can pay
regards to these factors and based on previously computed moisture and temperature histories
within the structure they can predict concrete behavior more accurately. This section describes a
module called CCStructuresTransport that is used to calculate the histories. A more accurate
creep analysis then typically consists of two steps: firstly execute CCStructuresTransport
module and calculate the moisture and humidity histories of the structure and secondly execute
CCStructuresCreep module to carry out the actual static analysis. Of course, for both analyses,
we have to prepare an appropriate model. Export/Import of the results between the modules is
already done by ATENA automatically.
To be exact, both the transport and static analysis should be executed simultaneously, but as
moisture and temperature transport does not depend significantly on structural deformations, i.e.,
coupling of the analyses is low, the implemented “staggered” solution yields sufficiently
accurate results.
The governing equations for moisture transport read (for representative volume REV] :
w  ( we  wn )
  div( J w ) (5.1)
t t
where:
w is total water content defined as a ratio of weight of water at current time t to weight of
water and cement at time t0  0 in REV, [mass/mass], e.g. [kg/kg]

we , wn = stands for the amounts of free and fixed (i.e. bound) water contents, [mass/mass],

J w = moisture flux, [length*mass/ (time*mass)]. e.g. [m/day],

t =time, [time], e.g., [day].

The moisture flux is computed by
J w   Dwwe (5.2)
where
Dw is moisture diffusivity tensor of concrete [m2/day],

 is gradient operator.
Note that in (5.2) only diffusion of water vapor is considered. Moisture advection is negligible.
The equations (5.1) and (5.2) can also be written as being dependent on w or relative moisture h .
A relationship between h and w is given by
w  w( h ) (5.3)
Using (5.3) Equation (5.2) can be written as follows
J w   Dh h (5.4)

ATENA Theory 285

A special attention must be paid to the calculation of the above time derivatives and integration
of the governing equations. For example, in the case of usual Gauss integration and use of exact
time derivatives the solution may suffer from mass losses. To remedy the problem the
CCStructuresTransport module integrates the structure, i.e., all the individual finite elements in
nodes and time derivatives are calculated numerically (Jendele 2001). This integration is similar
to use of finite volume method, which is also known to be robust against the mass losses.
Heat transfer is governed by similar equation
Q  T
  CT (T  Tref )   CT   div( J T ) (5.5)
t t t
where
Q is total amount of energy in a unit volume [J/m3]
CT is heat capacity [J/(K.m3)],

JT is heat flux [J/(day.m2)].

   
If hydration we want to add heat Qh (t ) , which expr esses amount of hydration heat
t t t t
 J 
within unit volume i.e Qh ,  3  , Equation (5.5) changes to
m 
 T Qh
t
 CT (T  Tref )  Qh   CT
t

t
 div( J T ) (5.6)

 J 
Heat flux JT ,  2  is calculated by
m s
JT   KT grad (T ) (5.7)
and KT stands for heat conductivity, e.g. [J/(day.m.K)].

Note that Equation (5.5) accounts for heat transport due to conduction only. Heat advection is
negligible. In (5.5) we can also neglect hydration heat because in large times, its impact for heat
transfer is small. On the other hand, we cannot neglect concrete moisture consumption due to the
hydration process. According to (Bazant and Thonguthai 1978; Bazant 1986) hydration water
content wh can be calculated by:
1
 t 3
wn  wh  0.21 c  e  (5.8)
  e  te 
where
 e = 23 days, te is equivalent hydration time in water at temperature 25 0C that corresponds to
the same degree of hydration subject to real age, moisture and temperature conditions of the
material. The parameter c relates to the amount of cement and is calculated by(5.53). If
temperature ranges from 0 to 100 0C , te is computed by

286
 te    h T dt (5.9)

where dt is time increment after the mold has been removed and coefficients T ,  h are
calculated by
1
h  (5.10)
1  (3.5  3.5h) 4

U  1 1  
T  exp  h
   (5.11)
  T0 T  
R

Uh
In the fraction the symbol U h stands for the activation energy of hydration and R is gas
R
Uh
constant. According to (Bazant 1986)  2700 0 K . T, T0 are real and reference concrete
R
0
temperature is expressed in K . The reference temperature is given by

T0  273.15  25 (5.12)

The following figure depicts the relationship between real t and equivalent time te for the case
of constant temperature T  15 0C and moisture h  0.8 . In practice, this relationship is rarely
linear because with increase of time the amount of fixed water (due to hydration) wh is
increasing as well and it involves a gradual decrease of relative moisture h .

Fig. 7-1 Equivalent vs. real time relationship

ATENA Theory 287

The amount of water that was needed for hydration of concrete according to Equation (5.8) for
the case of c  300 kg is shown below:

Fig. 7-2 Moisture consumed by hydration as a function of equivalent time

7.1 Numerical Solution of the Transport Problem – Spatial

Discretisation
The transport governing equations for a typical engineering problem are too complex for
analytical solution. Hence, similar to other ATENA engineering modules, the finite element
method is also used for the CCStructuresTransport module. The transport problem gets spatially
and temporarily discretized and then the resulting set of nonlinear algebraic equations is solved
by a special iterative solver. This section is dedicated to the detailed description of the former
type of discretization.
The solution is based on Equations (5.1) thru(5.7). Note that the unknown variables are
h  h(t ); T  T (t ); w  w(h, T ); wh  wh (h, T , t ) (5.13)
and they are to be discretized. Let the left-hand side part of (5.1) and (5.4) is denoted
LHSh , LHST , respectively. The subscript h and T indicates moisture and temperature fluxes.
Similar subscripts are also used for the right-hand-side of the equations, RHS h , RHST . Notice
that RHS expressions do not include the divergence operator!

LHS h   w  wh  (5.14)
t

T Qh
LHST  CT  (5.15)
t t

288
 RHSh   J w   J h (5.16)

RHST   J T (5.17)
The strip over an entity in the above equations means that the entity is vector. (Scalar entities do
not have the strip). The fluxes J w  J h are identical, i.e., the subscript w indicates also moisture
phase. Using the above notation Equations (5.1) and (5.5) can be written as follows
LHSh  div( RHSh )
(5.18)
LHST  div( RHST )
The LHSh includes time derivative of moisture. It is computed using the following expressions:

wh  wh (te )

te
  h T (5.19)
t

wh wh te wh

   h T
t te t te
For the next derivation, let us write Equations(5.14), (5.15) in a general form:
h w T
LHS h  chh  chw  chT  ch 0
t t t
(5.20)
h w T
LHST  cTh  cTw  cTT  cT 0
t t t
and equations(5.16), (5.17)

RHSh   khh  h   khw  w   khT  T  kh 0

(5.21)
RHST   kTh  h   kTw  w   kTT  T  kT 0
where square bracket indicates that the enclosed entity is a matrix [ ].
Comparing (5.20) with (5.1) and (5.5) we find that
chh  chT  0; chw  1; ch 0  0
(5.22)
Qh
cTh  cTw  0; cTT  0; cT 0  0
t

ATENA Theory 289

The parameter cTT is in ATENA an input material parameter, ch 0 is computed from(5.19), i.e.
wh w w
ch 0   h T . The solution also includes expressions  0; . Their values depend on a
te h T
constitutive model being used in the solution. For more information, please refer to Section
Material Constitutive Model.
For right-hand sides, we can write in a similar manner:

 khw    khT   0 ;  khh   0; kh0  0

(5.23)
 kTh    kTw   0 ;  kTT   0; kT 0  0
The parameter  kTT  is a material input parameter,  khh  is calculated from a constitutive model,
see the next section.
For the next derivation, let us assume discretization of the unknown variables as follows.
Remind that these are in the governing equations integrated in finite nodes, (Celia, Bouloutas et
al. 1990; Celia and Binning 1992).
T
h  NT h; h  N  h

T
w  N T w; w  N  w (5.24)

T
T  NT T ; T  N  T

where
h , w, T stands for vectors of the corresponding entities. The vectors have dimension n equal
to number of finite nodes of the problem.
N is vector of interpolation, (i.e., shape) functions,

 N1 N 2 N n 
 x ...
 x x 
N N 2 N n 
N    1
T
...
y y y 
 
 N1 N 2 N n 
 z ...
z z 

Using (5.24) Equations (5.20) and (5.21) can be written in the form

290
 h w T
LHSh  chh N T  chw N T  chT N T  ch 0
t t t
(5.25)
h w T
LHST  cTh N T  cTw N T  cTT N T  cT 0
t t t
and

RHS h   khh  N  h   khw  N  w   khT  N  T  kh 0

T T T

(5.26)
RHST   kTh  N  h   kTw  N  w   kTT  N  T  kT 0
T T T

The resulting set of equations are solved iteratively using finite element method, see
(Zienkiewicz and Taylor 1989), (weak formulation, in which the shape functions N are used as
weight function):

 N  LHS
V
h 
 div( RHSh ) dV  0

(5.27)

 N  LHS
V
T 
 div( RHST ) dV  0

where V is volume of the analyzed structure. Each of the above equations represents a set of
equations with dimension equal to number of finite nodes n. Note that div( RHSh ) and
div ( RHST ) are scalars !

In the next derivation, the two parts of (5.27) are dealt with separately.

 h w T 
 N  LHS  dV   N  c
V
h
V
hh NT
t
 chw N T
t
 chT N T
t
 ch 0  dV 


h w
c
V
hh NN T dV   chw NN T dV
t V t
 ... c
V
h0 NdV  (5.28)

h w
cchh    cchw   ...cc h 0
t t

 T h T w T T 
V N  LHST  dV  V 
N  cTh N  cTw N  cTT N  cT 0  dV 
t t t 
(5.29)
h w
 ccTh    ccTw   ...ccT 0
t t

and the matrices  cc  are calculated by

ATENA Theory 291

 cchh    chh NN T dV ; cchw    chw NN T dV ; ... cc h 0   ch 0 NdV
V V V

(5.30)
ccTh    cTh NN T dV ; ccTw    cTw NN T dV ; ... ccT 0   cT 0 NdV
V V V

The second part of (5.27) are calculated using Green theorem (5.36):

 N  div( RHS )  dV    N  n 
RHS h dS   N  RHSh dV 
T
h s
V S V

   N nsT
S
k  N  h  k  N  w  k  N  T  k  dS 
hh
T
hw
T
hT
T
h0
(5.31)

  N    k  N  h   k  N  w   k  N  T  k dV

T T T
hh hw hT h0
V

where S is the structural surface (with possibly defined boundary conditions).

In the case of heat transfer, we can derive all the equations in a similar way. In analogy to (5.30)
let us introduce matrices  kk 

 kkhh    N   khh  N 

T
dV
V

 kkhw    N   khw  N 

T
dV
V

kk h 0   N  kh 0 dV
V (5.32)
...

 kkTT    N   kTT  N 

T
dV
V

kk T 0   N  kT 0 dV
V

and also

292
  J hh    N nsT  khh  N  dS
T

 J hw    N nsT  khw  N  dS

T

...

 JTT    N nsT  kTT  N  dS

T

J h 0   N nsT kh 0 dS
S

JT 0   N nsT kT 0 dS
S
(5.33)

Using (5.28) to (5.33) the original governing equations (5.27) can be written as follows:
h w T
cchh    cchw    cchT   cc h 0   kkhh  h   kkhw  w   kkhT  T  kk h 0 
t t t
  J hh  h   J hw  w   J hT  T  J h 0
(5.34)
h w T
ccTh    ccTw    ccTT   ccT 0   kkTh  h   kkTw  w   kkTT T  kk T 0 
t t t
  J Th  h   J Tw  w   J TT  T  J T 0

After sorting the unknown variables h , T by finite nodes into a single vector , Equation (5.34)
will read

cc   kk   cc 0  kk 0   J   J 0 (5.35)
t
The right-hand side (5.35) is non-zero only for non-zero prescribed boundary conditions and
hence it has character of “load” vector in a static analysis.
In (5.31) we used Green theorem. It states:

 u div(v )dV   u n v dS   u  v dV

T
s
V S V

(5.36)

 u div(v )dV   u n v dS   u  v dV

T
s
V S V

where

ATENA Theory 293

  u u u 
u    x y z 


 u1 u1 u1 

 x y z 
 
 u2 u2 u2 
u    x y z 
 
 ... ... ... 
 
 un un un 
 x y z  (5.37)

7.2 Numerical Solution of the Transport Problem – Temporal

Discretisation
The heat and moisture transfer governing equations (5.35) can be written in the form:
 t t
t t
K t t + t t C
t
   t t J (5.38)

where t t K , t t

C = are unsymmetrical problem matrices defined in the previous section,
t t
J =vector of concentrated nodal fluxes (both moisture and heat) and t t
 is the vector of
unknown variables. All of these apply for time t  t . Equation (5.38) is solved iteratively. i.e.,
the vector t t
 is searched for in the incremental form:
t t
 t t (i )  t t (i 1)  t t (i )  (5.39)

where index (i )
indicates the number of iteration and t t ( i )
 is the increment of the unknowns
for time t  t and iteration ( i ) :
t t ( i )  -1 t t (i ) J
 t t (i 1) K (5.40)

The matrix and vector t t ( i 1)  -1 and

K t t ( i )
J is derived from t t ( i 1)
K -1 , t t (i 1)C-1 and t based
on temporal integration method being used:
CCStructureTransport module currently supports  Crank Nicholson (Wood. 1990) and Adams-
Bashforth (Diersch and Perrochet 1998) integration scheme. The former scheme is linear, i.e.,
it’s a first-order integration procedure. The latter scheme is a second-order integration procedure.
It is supposed to be more accurate; however, it is also more CPU and RAM expensive and it is
more difficult to predict its real behavior. Hence, the  Crank Nicholson scheme is typically
preferred. It has been more tested and verified in the CCStructureTransport module, and thereby
it is more recommended.

294
7.2.1  -parameter Crank Nicholson Scheme
This scheme comprises a number of well established integration procedures. It depends, what
value of the parameter  is used. The set of equations (5.38) is solved for time t  t  ,
whereby the vector of unknown variables is calculated as a linear combination of the
corresponding vectors at a time t and t  t . Hence
t t
 t  (1   ) t t  (5.41)

Depending on a particular value of the parameter  we get the well known Euler implicit
integration (for  =1), trapezoidal Crank Nicholson scheme (for  =0.5), Galerkin integration
method (for  =2/3) or even Euler explicit scheme (for  =0), which is only conditionally stable.
Solution predictor:
 t
t t
 t   t (5.42)
t
Solution corrector:

 t t 1
  t t t   (5.43)
t t
 J .
Using the above, after some mathematical manipulation, we derive final expressions for K,
These read:

 =  K  1 C 
K  
 t 

1
J  J  K  t t  1    t   C  t t  t  (5.44)
t

 
1

  K J

7.2.2 Adams-Bashforth Integration Scheme

Solution predictor:

t    t t   prev 
t
t
t t
     2 
t
   (5.45)
 t prev
2   t t prev t 
where
index prev indicates that the entity comes from time preceding time t Note that we assume that all
entities from time t are already known and we solve for their values at time t  t .
Solution corrector:

 t t 2  t
  t t t    (5.46)
t t t

ATENA Theory 295

  t t prev  t t t   t  t t  prev 
     (5.47)
t t  t prev  t  t  t prev  t prev 
 J :
Similar to (5.44) we have here for K,
 = t  Kt  t  t   C  2t  t  
K n 1 n n n 1 n n 1

J  K t t  t  t  t  t   
n
2
n 1 n n 1
2

 C     2t t   t      2t
t t
n 1 n n 1
2 t
n 1 
tn   tn 1    tn   t t  tn 
2 2 2
 (5.48)
 J   t  t  t  t  
2 2
n n 1 n n 1

 
1

  K J

7.2.3 Reduction of Oscillations and Convergence Improvement

The transport governing equations are prone to suffer from oscillations. As reported in (Jendele
2001) this can be improved by introducing a sort of Line Search method damping  . The basic
idea is that Equation (5.39) gets replaced by
t t
 t t (i )  t t (i 1)  t t (i )   (5.49)

where  is a new damping factor. The factor is typically set to something in range  0.3...1 
depending on the current convergence behavior of the problem.

7.3 Material Constitutive Model

The previous section referred to a material constitutive model, i.e., it was assumed that we know
how to compute material diffusivity matrix Dh , (see(5.4)), and material capacity w  w(h) ,
(see(5.1). Calculation of these entities is described here.
Currently, ATENA has only two constitutive models available for transport analysis. The first
one, i.e., CCModelBaXi94 is characterized as follows and the second one, i.e.,
CCTransportMaterial is briefly described later in this section.
CCModelBaXi94
For heat transport, a simple constant linear model is implemented. For moisture transport, a
nonlinear model based on the model (Xi, Bazant et al. 1993; Xi, Bazant et al. 1994) has been
developed.
It can be used for temperatures in range T  5 ...75 0C  and moisture H  0 ...1  . It is
important to note that the model was originally written only for mortar; hence, it is inaccurate for
concrete with an aggregate having higher permeability (i.e., diffusivity) and/or absorption. The
model has the following main parameters

296
  Type of cement
w
 Water-cement ratio wc 
c
As already pointed out, the model does not account for aggregate, i.e., it predicts moisture move
only in pores filled by water-cement paste.
w
The main entity of the model is water content w  w(h, t , T , ) . It is defined as follows:
c
Gw
w (5.50)
Gw,0  G c

where

 kg 
Gw is the water content in mortar at time t ,  3 ,
 m of morter 

 kg 
Gw,0 is the water content at time zero,  3 ,
 m of morter 

 kg 
Gc is the amount of cement at time zero,  3 .
 m of morter 
Mortar here stands for a mixture of water and cement. If concrete material is to be considered,
then w can be calculated by
Vconcrete
Gw w
Vmortar G
w  (5.51)
V V  w,0  G
G c
Gw,0 concrete  G c concrete
Vmortar Vmortar
Vconcrete
where is the ratio of total volume to (only) volume of mortar (i.e., water and cement) and
Vmortar
 are corresponding amounts of water and cements in concrete, (i.e., not only in
G
 kg 
mortar!)  3 .
 m of concrete 
w
The model itself already accounts for moisture used by the hydration process. i.e.,  0 . As a
t
result, wh according to (5.19) need not be implemented.

On the other hand, if moisture losses due to hydration are to be computed by the model based on
w
(5.19), it is important to fix  0 and to modify wh , so that it predicts “relative” moisture
t
content w used throughout whole derivation CCStructuresTransport. The original function for
wh was written for absolute weight of water and hence, for “relative” content of water Equations
(5.8) must be rewritten into

ATENA Theory 297

 1

 c  te 
3
0.21 G   1 1

  e  te  c
G  te  3 Gc  te  3
wh   0.21    0.21   (5.52)
 w,0  G c  w,0  G
c
G G   e  te  Gw,0  Gc   e  te 
and the constant c from (5.8) becomes equal to
Gc c
G
c  (5.53)
Gw,0  Gc G w,0  G
c

More detailed description of the model is beyond the scope of this document and the reader is
referred to in (Xi, Bazant et al. 1993; Xi, Bazant et al. 1994).
CCTransportMaterial
CCTransport material is a simple constitutive law that allows users to enter laboratory-measured
moisture and heat characteristics. Referring to Equations (5.1) and (5.5) heat and moisture flow
governing equations can be written in the following general form:

Heat :
Q h T w 
 CTh  CTT  CTw  CTt    KTh grad (h)  KTT grad (T )  KTw grad ( w)  KTgrav 
t t t t x

Moisture :
w h T w 
 Cwh  CwT  Cww  Cwt    Dwh grad (h)  DwT grad (T )  Dww grad ( w)  Dwgrav 
t t t t x
(5.54)
The parameters CTh , CTT … K wgrav are calculated as:

298
 CTh  CTh0 fChTh (h) f CTTh (T ) fCtTh (t )
CTT  CTT
0
fChTT (h) f CTTT (T ) fCtTT (t )
CTw  CTw
0
f ChTw (h) fCTTw (T ) fCtTw (t )
CTt  CTt0 f ChTt (h) f CTTt (T ) f CtTt (t )
Cwh  Cwh
0
fChwh (h) fCTwh (T ) fCtwh (t )
CwT  CwT
0
f ChwT (h) fCTwT (T ) fCtwT (t )
Cww  Cww
0
fChww (h) fCTww (T ) fCtww (t )
Cwt  Cwt0 fChwt (h) f CTwt (T ) f Ctwt (T )

KTh  KTh0 f KhTh (h) f KTTh (T ) f Kt Th (t )

KTT  KTT
0
f KhTT (h) f KTTT (T ) f Kt TT (t )
KTw  KTw
0
f KhTw (h) f KTTw (T ) f Kt Tw (t )
KTgrav  KTgrav
0
f KhTgrav (h) f KTTgrav (T ) f Kt Tgrav (t )
Dwh  Dwh
0
f Dhwh (h) f DTwh (T ) f Dt wh (t )
DwT  DwT
0
f DhwT (h) f DTwT (T ) f Dt wT (t ) (5.55)
Dww  Dww
0
f Dhww (h) f DTww (T ) f Dt ww (t )
Dwgrav  Dwgrav
0
f Dhwgrav (h) f DTwgrav (T ) f Dt wgrav (t )
0
and the constant parameters CTh0 thru Dwgrav and functions fChTh (h) thru f DTwgrav (T ) are input
parameters, (to be possibly obtained from some experiments). The functions are defined as
multilinear functions and only their ids are input into CCTransportMaterial model definition.
Note that gravity terms in RHS of (5.54) have a little physical justification in heat and moisture
diffusion gathered transports; nevertheless, they are included to allow using this material law for
the solution of other kinds of transport problems.

CCTransportMaterialLevel7 material
CCTransport materialLevel7 is an extension of the above CCMaterialTransport material in the
way it automatically computes moisture and temperature capacity and conductivity/diffusivity
incl. "sink" terms regarding hydration (i.e., rate of hydration heat and moisture consumption
during concrete hydration). In terms of the above nomenclature, this upper material level
calculates CTT , KTT , CTt , Cwh , Dwh , Cwt . As already mentioned, the presented material adds on its
bottom level, i.e., CCMaterialTransport. All parameters and characteristics from the bottom
level, (i.e., those from CCMaterialTransport) can still be input and used. They typically serve for
a refinement/addition of parameters generated by the upper material level. The result from the
bottom and upper levels are simply added to form the final characteristics of the material model
CCTransportMaterialLevel7. Note that default values of CTT , KTT , CTt , Cwh , Dwh , Cwt in the bottom
level are by default set to zero.
Hydration heat and affinity hydration model

ATENA Theory 299

The most important part of the presented model is the computation of the concrete hydration maturity
factor. It is accompanied by the calculation of generated hydration heat and consumed hydration
moisture. The analysis is based on the affinity hydration model, which provides a framework for
accommodating all stages of cement hydration.

Consider hydrating cement under isothermal temperature 25oC a relative humidity h  1 . At this
temperature, the rate of hydration maturity factor  ,   0...1 can be expressed by chemical affinity
A  A ( ) :
25 25

 
 A25 (5.56)
t
where A stands for the chemical affinity, [ s 1 ], The expression already includes coefficient
 E 
exp   a  . Hence A 25 is not normalized and refers to temperature 25oC. For different
 RT 
J
temperatures it is replaced by A , see (5.60). R is gas constant 8314.41 , T is
kmol K
temperature, [K] and Ea is 40 kJ/mol. It is worthy to note the incorporation of the maturity

method into (5.56). A characteristic time might be introduced to express an affinity A (Bernard,
Ulm et al. 2003).
The affinity property can be obtained experimentally or analytically. Using experimental
approach, heat flow q (t ) that corresponds to the hydration heat Qh  Qh (t ) is measured by
isothermal calorimetry.
Alternatively, the hydration material parameters are computed by an analytical micro-scale
model that accounts for the majority of underlying chemical reactions as well as the topology of
cement grains (with the consequence to hydration kinetics). The solution stems from (Smilauer
and Bittnar 2006), and it employs discrete hydration model CEMHYD3D (Bentz 2005), allowing
to account for the particle size distribution of cement, the chemical composition of cement,
temperature and moisture history in concrete, etc.

Having history of Qh (for T  273.15  25,  1 ), the approximation of  parameter is given by

Qh
 (5.57)
Qh , pot

1 Qh  
  A25 (5.58)
Qh , pot t t
Qh
where Qh, pot is potential hydration heat, [J/kg]. Hence the normalized heat flow under
Qh , pot
isothermal 25oC equals to chemical affinity A 25 .

Cervera et al. (Cervera, Oliver et al. 1999) developed an analytical form of the affinity which
was refined in (Gawin, Pesavento et al. 2006). A slightly modified formulation is proposed here

300
 B    
A 25  B1  2         exp    (5.59)
     
where B1 ,[ s 1 ], B2 , [-] are coefficients to be calibrated,   is the ultimate hydration degree, [-],
and  represents microdiffusion of free water through formed hydrates, [-]. The parameters in
(5.59) express isothermal hydration at 25◦C.
When hydration proceeds under varying temperature, maturity principle expressed via Arrhenius
equation scales the a nity to arbitrary temperature T
E  1 1 
Ar  exp  a   
 R  273.15  25 T   (5.60)
A  A A
T 25 r

For example, simulating isothermal hydration at 35oC means scaling A 25 with a factor of 1.651
at a given time. This means that hydrating concrete for 10 hours at 35oC 35◦C releases the same
amount of heat as concrete hydrating for 16.51 hours under 25◦C. Note that setting Ea  0
ignores the e ect of temperature and proceeds the hydration under 25◦C.
Gawin et al. (Gawin, Pesavento et al. 2006), among others, added the effect of relative humidity.
The extension of (5.58) leads to
1 Qh  
  AT  h (5.61)
Qh , pot t t

1
h  (5.62)
1   a  ah 
4

where  h   h (h) accounts for the reduction of capillary moisture. h is relative humidity r,
(Bazant and Najjar 1972). a is material parameter, typically a  7.5 . Depending on curing
conditions   is calculated as follows:

Sealed curing:
w/c
  ,   1 (5.63)
0.42
Saturated curing:
w/c
  ,   1 (5.64)
0.36
w / c is the water-cement ratio.
Substituting (5.59) and (5.62) into (5.61) yields final equation to predict the development of
hydration heat. As it is difficult to express  function analytically (from (5.59), (5.61)), the
above equations are integrated numerically.

ATENA Theory 301

  (t ) 
 A25 (t25 ) Ar  h
t (5.65)
tend
 (tend )   (tstart )   A 25 ( 25 ) Ar  h d
tstart

Substituting d 25  Ar  h d
t25,end
 (tend )   (tstart )   A 25 ( 25 )d 25 (5.66)
t25 start

If the function DoH 25 (t )   25 (t ) at reference temperature is known, (e.g. it was meassured in a

calorimeter), Ar  h is constant within  t25 start ...t25 end  and it is acceptable to use linear (Taylor)

approximation of  25 (t25 )   25 (t25 start )  (t25  t25 start ) within  t25 start ...t25 end  , we can write:
t25
 25 (t25 ) 
 A25 (t25 )
t25
(5.67)
 25 ( 25 )
d 25   tstart   25 (t25end )   25 (t25 start ) 
t25,end
 (tend )   (tstart )  
t25 start  25

In the above t25 start   25

1
( start ), t25 end  t25 start  (tend  tstart ) Ar  h are equivalent time for the case
of reference temperature.  25
1
(...) is inverse function to  25 (...) so that  25 ( 251 ( ))   .

Note that Qh is calculated in the same unit as is entered the parameter Qh, pot . If the governing
equations are written for unit volume and Qh, pot is given per cement unit weight, then Qh must
be multiplied by fraction of cement mass mcement and total volume of concrete Vtot .

Heat capacity
The model assumes the following components of concrete: aggregate, filler, water, and cement.
The total mass of concrete in one cubic meter results from individual masses of components:
mconcr  maggregate  m filler  m paste
(5.68)
m paste  mcement  mwater

where mconcr is the mass of concrete per a unit volume. Similarly, for the mass of aggregate
maggregate , the mass of filler m filler , the mass of water mwater , and mass of cement mcement .
Corresponding volumes are Vaggregate  maggregate /  aggregate , V filler  m filler /  filler etc. i stands for
the mass density of the phase i. Having total volume Vconcr  Vaggregate  V filler  Vwater  Vcement , we
can calculate phase fractions f aggregate  Vaggregate / Vconcr and similarly for the remaining phases.

Heat capacity and its evolution of cement paste (cement+water) were studied in (Bentz 2007) at
230C for w/c between 0.3 and 0.5. The capacity of fresh cement paste yields

302
 Cconcrete  f aggregateCaggregate  f filler C filler  Cˆ paste (5.69)

where Cconcrete is the concrete capacity (per unit volume) and akin for aggregate, filler, and cement
paste. The last term, i.e., C paste also depends on the degree of hydration  and is calculated by

Cˆ paste  ( f cement Ccement  f water Cwater ) 1  0.26(1  exp(2.9 ))  (5.70)

where Ccement is cement capacity at time zero.

The heat capacity of structural concrete spans the range between 0.8 and 1.17 Jg-1K-1. A former
Czech standard CSN 731208 declares 840 and 870 Jkg-1K-1 for dry and saturated mature concrete,
respectively. Caggregate is approximately 840 Jkg-1K-1 for basalt and limestone, 790 Jkg-1K-1 for
granite, 800 Jkg-1K-1 for sand. Ccement is about 750 Jkg-1K-1 and Cwater is 4180 Jkg-1K-1 .

Heat conductivity
The thermal conductivity of cement paste was found to remain in the range 0.9-1.05 Wm-1K-1 for
an arbitrary degree of hydration, for both sealed and saturated curing conditions, and for w/c
from 0:3 to 0.4 (Bentz 2007). Water in the capillaries has a thermal conductivity 0.604 Wm-1K-1
(Bentz 2007). The thermal conductivity of hardened concrete varies between 0.85 and 3.5 Wm-
1 -1
K (Neville 1997) p.375, depending strongly on an aggregate type.
Thermal conductivity also depends on the saturation state of concrete. For example, a structural
concrete made from normal-weight aggregate with a unit mass of 2240 kg/m3 yields  = 1.696
Wm-1K-1 for protected and 1.904 for weather-exposed conditions (Neville 1997), p. 376.

Figure 7-1. Thermal conductivity of concrete according to the Czech code CSN 731209.
Figure 7-1 summarizes thermal conductivities for ordinary concrete depending on concrete unit
mass and saturation conditions, according to (Neville 1997) and a former Czech standard CSN

ATENA Theory 303

731208. The latter considers 1.5 for a dry concrete and 1.7 Wm-1K-1 for a water-saturated
concrete.
Faria et al. (Faria, Azenha et al. 2006) applied the evolution of concrete conductivity with
regards to 

  0 1.0  0.248 
where  is the conductivity of fully hardened concrete, i.e., in infinite time.

The model implemented in Atena, i.e., CCTransportMaterialLevel3 stems from homogenization

theories. Consider conductivity of cement paste  paste and aggregates aggregate such that
 paste  aggregate . Corresponding volume fractions are f paste , f aggregate . Hashin-Shtrikman lower
concrete,low and upper bounds concrete,upper are (Bentz 2007)

3 f aggregate  paste  aggregate   paste 

concrete,low,   paste 
3 paste  f paste  aggregate   paste 
(5.71)
3 f paste aggregate   paste  aggregate 
concrete,upper ,  aggregate 
3aggregate  f aggregate   paste  aggregate 

The presented model uses average conductivity, i.e.

concrete.low,  concrete,upper ,
concrete  1.33  0.33  (5.72)
2
Figure 7-2 considers  paste =1.0 Wm-1K-1 and aggregate = 2.0 Wm-1K-1 . Volume fraction of
aggregates varies from 0 to 1. Important thermal conductivities: limestone 1.26 - 1.33, sandstone
1.7, granite 1.7 - 4.0 Wm-1K-1 .
The above equations for homogenization are written for phases paste-aggregates. In ATENA, the
homogenization is carried out as follows:
1. homogenize phases cement - water -> phase paste.
2. homogenize phases paste - filler -> phase paste with filler
3. homogenize phases paste with filler - air -> phase paste with filler and air
4. homogenize phase paste with filler and air - aggregates -> concrete
Note that filler and aggregate are in this case treated as one component, and the same applies for
water and cement (being the component paste). The volume averaging technique is used to
calculate the corresponding properties of paste and mixed aggregate.

304
 Figure 7-2 Predicted thermal conductivities of concrete from bounds.
Moisture consumption due to hydration
It is assumed that 1 kg of cement (in concrete) approximately consumes during the full hydration
process about Qw, pot of water. Typically Qw, pot =0.42 kg of water per 1 kg of cement. Thus, e.g.
concrete mixture with 300kg cement per 1m3 of concrete needs 300*0.42=126kg o water per 1m3 of
concrete. Assuming linear dependence of water hydration consumption wh on concrete hydration
level  , (   0 for fresh concrete and   1 for fully hydrated concrete) the water sink term due to
hydration is

wh wh 
Ch , t   (5.73)
t  t
wh  Qw, pot c  ,[kg] (5.74)

where c stands for weight of cement in 1m3 of concrete.

Moisture capacity
The moisture content at unit volume w,[kgm -3 ] is calculated a simple expression

w  wf
 b  1 h (5.75)
bh
where w f ,[kgm-3 ] is the free water saturation and b is the dimensionless approximation factor,
which must always be greater than one. It can be determined from the equilibrium water content
w80 at relative humidity h  0.8 by substituting the corresponding numerical values in equation
(5.75):
h( w f  w80 )
b (5.76)
w f h  w80

ATENA Theory 305

Moisture capacity C ,  kgm -3  is calculated as derivative of moisture content with respect to h :

w w f  b  1 b
Ch   (5.77)
h b  h 
2

The above expression is applicable for analyses using reference unit volume. If reference unit
weight of the structure is preferred, then we employ moisture capacity C  C /  ,  kg/kg  , where
 is concrete density,  kg/m 3  .

Moisture diffusion
The present model accounts for the diffusivity mechanism of moisture transport. It is valid for
dense concrete, which has not mutually connected pores and moisture convection thru pores
 kg 
(being driven by water pressure) can be neglected. Hence, moisture flux qh ,  2  is calculated
m s
 kg 
by the equation qh  Dh h , where total moisture diffusivity Dh ,   is calculated as sum of
ms
water Dhw and water vapor Dhwv diffusivity:

Dh  Dhw  Dhwv (5.78)

Water liquid diffusivity Dhw is calculated

w
Dhw  Dww (5.79)
h
where water diffusivity Dww ,  m 2 / s  is

 w 
 1
2  wf 
 
3.8 A 1000
Dww  (5.80)
 wf 
2

 kg 
and A is the water absorption coefficient  2 0.5  .
m s 
Water vapor permeability is computed from water vapor pressure-driven diffusivity
 kg 
Dpwv ,  :
 m s Pa 

D pwv  (5.81)

where  is the water vapor diffusion resistance factor and  is the vapor diffusion coefficient
 kg 
in air  
 m s Pa 

306
 1.81
0.00002306 pa  T  273.15 
   (5.82)
R
T  273.15 pa  273.15 
Mw

Atmospheric pressure pa  101325Pa , gas constant R  8314.41Jkmol-1K -1 and molar mass of

water is M w  18.01528kgkmol-1

As in the presented model, relative humidity h is the primary variable used to analyze moisture
transport, Dpwv must be transformed to Dhwv . This is done by:

p  ( psat h)
Dhwv  D pwv  D pwv  D pwv psat (5.83)
h h
Any expression to calculate the pressure of saturated water vapor can be used. The presented
model uses
 aT 
 
psat  611e  T0 T 
,  Pa  (5.84)

In the above T is temperature [ o C ] and the remaining parameters are

T  0 : T0  234.18o C , a  17.08; T  0 : T0  272.44o C , a  22.44

Some guidelines towards the model's parameters

Fitted parameters for cement paste hydration need to be considered for each concrete separately.
Due to high cement variability, it is impossible to assign one particular cement to one concrete
grade. The user needs first to select the cement parameters from the following table:

Table 7.3-1 Parameters of affinity hydration model used for CEM I.

The above table is based on fitting predicted results from CEMHYD3D analysis by (5.59), see
Table 7.3-5 and Figure 7-3. The simulations were carried out on CEMHYD3D’s microstructures
50 × 50 × 50 µm and with the activation energy 38.3 kJ/mol. Saturated curing conditions were

ATENA Theory 307

assumed since sealed conditions will be obtained from coupling with moisture transport. Table
7.3-5 specifies input data for selected Portland cements.

Figure 7-3 Fit of selected cements to the affinity model, w/c = 0.4
The majority of concretes is produced from blended cements (CEM II - CEM V); hence it is
necessary to scale down Q pot by approximately 30 %. This is a common Portland clinker
substitution in the majority of blended cements in Europe.
There are other default parameters, which are not specified here: QW POT= 0.42, TH INIT = 0,
ALPHA INIT = 0, TEMPERATURE INCR MAX =0.1, H80 = 0.8, TEMP0 = 234.18, A WV =
17.08, TEMP0 ICE = 272.44 ,A WV ICE = 22.44
The parameter A ≈ 7.5 expresses hydration slow-down with regards to relative humidity. The
hydration practically stops at  ≈ 0.8.
Parameters in Figure 7-1 are computed for saturated state. When  = 1, the hydration proceeds
as there is saturated water environment around concrete. Under standard circumstamces,
hydration consumes water, which decreases relative humidity in the calculation. Three
parameters are related to moisture transport and are given for ordinary structural concrete:
 W80 expresses total mass of free water at  =80%. Standard value is 50 kg/m 3 for
structural concrete.
 A W is water absorption coe cient, whose value spans the range 0.25-0.846 kgm− 2 h
0.5 ].
 MI WV is the water vapor di usion resistance factor, spanning 210-260 [-] for structural
concrete.

308
Parameters specifying specific heat capacity for concrete components are summarized in Table
7.3-2. Values are obtained from http://www.engineeringtoolbox.com/density-solids-
d_1265.html, http://www.engineeringtoolbox.com/specific-heat-solids-d_154.html
Parameters specifying specific heat conductivity for concrete components are summarized in
Table 7.3-3. Sources from http://www-odp.tamu.edu/publications/192_SR/109/109_5.htm
Concrete strength classes strongly depend on the amount of cement in concrete. Table 7.3-4
specifies approximate compositions for major concrete classes used in EN 206-1. The
calculation assumes 5 % of entrained air in the concrete, cement density 3220 kg/m 3 and
aggregate density 2800 kg/m 3 .
Table 7.3-2 Parameters specifying density and specific heat capacity for concrete components

Table 7.3-3 Parameters specifying specific heat conductivity for concrete

components

Ready-mix concrete is assumed, which requires rather higher w/c due to workability and
pumping issues. The parameters CEMENT DENSITY, WATER DENSITY, AGGREGATE
DENSITY, FILLER DENSITY are provided in Table 7.3-2 in the units [kg/m 3 ].
Table 7.3-4 Approximate composition for major concrete classes used in EN206-1

ATENA Theory 309

310
 Table 7.3-5 CEMHYD3D parameters for fittiong of affinity hydration model.

ATENA Theory 311

7.4 Fire Element Boundary Load
When undertaking heat transfer calculations, it is important that relevant thermal properties of
materials and heat transfer coefficients at the boundaries are defined for the entire temperature
interval of the load.

7.4.1 Hydrocarbon Fire

Hydrocarbon fires are those sustained by hydrocarbon-based products, such as chemicals, gas,
and petroleum. Depending on the heat load, different HC-curves can be derived in accordance
with Equation (5.85). The magnitude of the maximum temperature of the radiation source ( T1 ) is
crucial for the time temperature development. The nominal HC-curve is represented by the heat
load 200 kW/m2 and reaches maximum temperature of 1100 °C. The curve representing 345
kW/m2 is called the "modified" or "increased" HC-curve for tunnel applications. It reaches at
maximum 1300 °C.
T (t )  T1 (1  0.325e 0.167 t  0.204e 1.417 t  0.471e 15.833t ) (5.85)
where:
T (t ) = temperature of radiation source as function of time [°C],
T1 = maximum temperature of radiation source [°C] according to (5.85)
t = time [minutes]
Time development of temperature of the radiation source is depicted in the figure below. For
time t  0 Equation (5.85) yields T (0)  0 and hence, it is necessary to supplement (5.85) by
requirement T (t )  Tambiant ,ini , where Tambiant ,ini is initial ambient temperature prior the fire broke
up, (typically something about 20 °C).

Fig. 7-3 Temperature of radiation source

312
7.4.2 Fire Exposed Boundary
The nature of the structural ambient conditions is essential for the determination of the
temperature fields. Depending on the geometry, view factors, and ambient conditions, various
types of boundary conditions may be considered.
Fire exposed boundary
The heat is transferred from the fire gas to the exposed structure through radiation and
convection. At high temperatures, the radiation dominates. The radiation is expressed by the
resulting emissivity factor, which takes into account the emissivity of the fire source  , and
absorptivity of the heated surface  . The convection is calculated from the temperature
difference between the structure and ambient gas, depending on the gas velocity. Emissivity and
convection factors used for exposed surfaces are shown below
 r  0.56, []
 W  (5.86)
hc  50,  m 2 K 

The convection and emissivity heat flux on a boundary exposed to fire is calculated as follows:
qn  hc (Tg  Tb )   r (Tg4  Tb4 ) (5.87)

where
 = Stefan-Boltzmann constant [5.67x10-8 W/m2 K4],
Tg = absolute temperature of radiation source [K],

Tb =boundary temperature of the structure,

 r = resulting emissivity factor of the radiation source and the heated surface [-],

qn = heat flow at the fire exposed boundary [W/m2],

hc = convection heat transfer coefficient [W/m2K].

Adiabatic boundary
Adiabatic boundary surface refers to a boundary surface, where no heat can pass in (and/or out)
the structure. Structural symmetry lines and areas are good examples of this boundary
conditions.

7.4.3 Implementation of Fire Exposed Boundary in ATENA

The described fire boundary load conditions are ATENA modeled by
CCFireElementBoundaryLoad load. It is essentially an element boundary load that applies the
heat flow qn at the element boundary, i.e., at a surface exposed to fire. Unlike other loads in
ATENA (that are of incremental nature and constant within one load step), this load is
considered variable and has kind of a total load.
Four types of heat source definitions are implemented:

ATENA Theory 313

  Nominal HV fire – Temperature of the heat source is calculated by (5.85) and T1 (unless
it is manually inputted as temp_g_ref) is set to 1100 [°C].
 Modified HC fire – This definition is much the same as the above with the only
difference that default value for T1 is 1300 [°C].
 Generic fire (also refered to as User curve fire) - Temperature of the heat source is
assumed constant and is set value of temp_g_ref . If temp_g_ref is not inputted, then
1100 [°C] is used.
In any case, the generated (or directly inputted) curve for T (t ) can be additionally modified in
time by a user-supplied function time_id. The function takes one parameter, which is time of the
fire and it specifies a coefficient by which the generated initially (or inputted) boundary
conditions should be multiplied. Of course, load variation in space can be modified by coeff_x,
coeff_y coefficients etc. in the same way as for any other generated element load, (for more
details see Atena Input file manual).

7.5 Moisture-Heat Element Boundary Load 10

This type of boundary load is used for modeling heat and moisture fluxes from the structure to
the ambient air. Hence, it is typically applied as a boundary element load on external surfaces of
the structure. It resembles the fire boundary load described above and is implemented in a
similar way.
The moisture flux consists of three parts.
qh ,1  hcw (hg  hb )
qh ,2  ( xg  xb )
 9 
2,5
9  
2,5

q 'h ,3  1.38 E 9   TCb  32   hg  TCg  32   1  0.9000000000 v  [kg/m 2 s ] (5.88)

  
 5  5
qh ,3  transform _ units(q 'h ,3 )
qh  qh ,1  qh ,2  qh ,3

The first part qh,1 includes moisture flux driven by the gradient between ambient and surface
relative humidity hg and hb . hcw stands for moisture convection coefficient of the concrete-air
interface. The second part qh ,2 , accounts for moisture flux due to evaporation driven by the
gradient of humidity air ratio at the interface, i.e., xb and xg with the evaporation coefficient  .
 kg 
By default   (25  19v),  2  , where v is ambient air velocity, [m/s]. For more information,
m s
see http://www.engineeringtoolbox.com/evaporation-water-surface-d_690.html.

10
Available starting from ATENA version 5.

314
The humidity air ratio, [-] is calculated as follows (i reflects conditions in ambient air, i.e., i=g,
or in the surface of the structure i.e., i=b):
mwv , i  wv ,i
xi   (5.89)
ma a

It is calculated at state variables hi , Ti , i.e., relative humidity and temperature at i conditions.

In the above mwv , i ,  wv ,i , ma ,  a is mass and density of water vapor in REV corresponding to i
conditions and mass and density of dry air, [kg/m3], respectively.
pa
a  (5.90)
R
(Ti  273.15)
Ma
where M a is weight of 1kmol of dry air (assumed M a  28.96 kg/kmol). R is gas constant,
(R=8313JK-1), Ti is the temperature in oC. pa is the partial pressure of dry air, [Pa]

pa  p  hi pvw.sat (5.91)
Here p stands for total air pressure (typically normal air pressure p=101325Pa), hi is relative
humidity and pvw.sat is the partial pressure of saturated water vapor at Ti, (see
http://en.wikipedia.org/wiki/Density_of_air)
 7.5Ti 
 
 Ti  237.3 
pwv , sat  610.78 10 (5.92)

The density of water vapor at i th conditions is calculated similar to (5.90):

pwv , sat
 wv , sat  (5.93)
R
(Ti  273.15)
M wv
In the above M wv is the weight of 1kmol of saturated water vapor, (assumed M a  18.06 kg/km.

The third part is moisture flux evaporated from concrete calculated by CEMSTONE, see
http://www.cemstone.com/concrete-evaporation-forecast-engineers.cfm. It yields nearly the
same values as provided by ACPA calculator; see
http://www.apps.acpa.org/apps/EvaporationCalculator.aspx. TCg , TCb is the ambient and surface
temperatures in Celsia.

The heat flux also consists from two parts.

qT 1  hcT (Tg  Tb )   rT  (TKg4  TKb4 )
qT 2  qh hwe (5.94)
qT  qT 1  qT 2

ATENA Theory 315

The first part of the heat flux qT represents the usual flux due to heat convection and emission.
Its computation resembles (5.87). hcT stands for heat convection coefficient of the concrete-air
 W 
interface  2  ,  rT is heat emissivity coefficient [-], TKg , TKb are ambient and surface
m K 
 W 
temperatures in Kelvins and  is Stephan-Boltzmann constant ,   5.67 E  8,  2 4  . The
m K 
second part takes into account heat consumption due to the evaporation of moisture flux. By
 kJ 
default hwe  2270,   is assumed. More information available at
 kg 
http://www.engineeringtoolbox.com/evaporation-water-surface-d_690.html.
Both moisture and heat fluxes are typically computed using only their first or second part.
Therefore, the related ATENA input commands allow reading some boolean flags that specify,
which parts of the above fluxes should by accounted for and which should be skipped. For more
information, refer to the ATENA input file manual.

7.6 References
BAZANT, Z. P. (1986). Mathematical Modelling of Moisture Diffusion and Pore Pressure,
Chapter 10. Concrete at High Temperature. Z. P. Bazant: 198-237.
BAZANT, Z. P. and W. THONGUTHAI (1978). Pore Pressure and Drying of Concrete at High
Temperature. Proceedings of the ASCE.
CELIA, M. A. and P. BINNING (1992). "A Mass Conservative Numerical Solution for Two-
Phase Flow in Porous Media with Application to Unsaturated Flow." Water Resour. Res
28(10): 2819-2828.
CELIA, M. A., T. BOULOUTAS, et al. (1990). "A General Mass-Conservative Numerical
Solution for the Unsaturated Flow Equations." Water Resour. Res 27(7): 1438-1496.
DIERSCH, H. J. G. and P. PERROCHET (1998). On the primary variable switching technique
for simulating unsaturated-saturated flows, http://www.wasy.de/eng/prodinfo/flow/
swpool/swpool.htm#fef_manuals.
HUGHES, J. R. (1983). Analysis of Transient Algorithms with Particular Reference to Stability
Behaviour. Computational Methods for Transient Analysis, Elsevier Science Publishers
B.V.
JENDELE, L. (2001). ATENA Pollutant Transport Module - Theory. Prague, Edited PIT, ISBN
80-902722-4-X.
JENDELE, L. and D. V. PHILLIPS (1992). "Finite Element Software for Creep and Shrinkage
in Concrete." Computer and Structures 45 (1): 113-126.
REKTORYS, K. (1995). Přehled užité matematiky. Prague, Prometheus.
SEAGER, M. K. and A. GREENBAUM (1988). A SLAP for the Masses, Lawrence Livermore
National Laboratory.

316
WOOD., W. L. (1990). Practical-Time Stepping Schemes. Oxford, Clarenton Press.
XI, Y., Z. P. BAZANT, et al. (1993). "Moisture Diffusion in Cementitious Materials, Adsorbtion
Isotherms." Advn. Cem. Bas. Mat. 1: 248-257.
XI, Y., Z. P. BAZANT, et al. (1994). "Moisture Diffusion in Cementitious Materials, Moisture
Capacity and Diffusivity." Advn. Cem. Bas. Mat. 1: 258-266.
ZIENKIEWICZ, O. C. and R. L. TAYLOR (1989). The Finite Element Method, Volume 1:
Basic Formulation and Linear Problem. London, McGraw-Hill, 4th edition.

ATENA Theory 317

8 DYNAMIC ANALYSIS

ATENA software support four methods to execute dynamic analyses. These are:
 Newmark's  method,

 Hughes  method (Hughes 1983),

 Wilson 

 Modified Wilson  .

Note that Hughes  method with   0 reduces to Newmark's  method and Modified Wilson
 is just an extension to Wilson  .
The governing equations for dynamic analysis read:
Hughes  method:
M ut  t  C  1  α  u t  t  αu t   K  1  α  u t  t  αu t   1  α  R t  t  α  R t

Newmark  method:
M ut  t  Cu t  t  Ku t  t  R t  t

(Modified) Wilson  method:

M ut t  Cu t t  Ku t t  R t t
(5.95)
where
ut  t , u t  t , u t  t is acceleration, velocity, and displacement at a time t  t , (similar for time t
and t  t ),
M, C, K is mass, damping, and stiffness matrix, respectively,
R is the vector of external forces, i.e., concentrated loads,
 is the Hughes damping parameter.
They are is solved for displacement at time t  t . The displacement, acceleration and velocity
at time t  t is calculated as functions of (already known) ut , u t , u t and displacement
increments u t  t  u . If l-th iteration is solved, then we solve for displacement increment
l
u and u t  t   uk
k 1

ATENA Theory 319

 Hughes  method:
Newmark  method:
u t  t  u t  u t  t  u

u t  t   1 
 2βt 2  t 2  γ  1  2u t  2u  2u  γ  t
t t  t

 t 1  γ   u 
 t
u
2 tβ  2 tβ
 
1 2u βt  u t  2u t  2u  2u t  t
t 2 t 2 t
ut  t 
2 t 2β

Modified Wilson  method:

Wilson  method:
u t  t  u t  u t  t  u
3u t 1 3(u t  u t  t  u )
u t  t    2u t  ut t 
t 2 t
6u t
6v _ t 6(u  u t  t  u )
t
ut  t   2   2ut 
t t t 2 (5.96)

Substituting (5.96) into (5.95) and after some mathematical manipulation, the requested
displacement increment at iteration l can be calculated:
u  K inv
eff Reff (5.97)

where (for using structural damping C   M M   K K ) effective stiffness and RHS vector are:

K eff  M M  K K
(5.98)
Reff  M ( M1   M2 )  K ( K1   K2 )   0

The coefficients above are calculated using the following expressions. They are summarized (by
solution method):

320
Hughes  method:
  1 1  
   α   M γ  1   1  α   M γ 1  t
  
2 2
M   α  1  M  t  1   ut    M  u 
 β  2β   β tβ 
  
  
u t  t  αu  u t  t  γ M
t  t


t 2β βt
1 1 
  2 α  2  K γ   1  α   K γ   αu t  t  u t  t  γ K
K      α  1  K  tut     K  u t 
 β   β  βt
 
 
  1 1  
   α   M γ  1   1  α   M γ 1  t
 M1    
2 2
  α  1  M  t  1   ut    M  u
 β  2β   β tβ 
  
  

 M2  
 u t  t
 αγt M  γ t M  1 
t β
2

1 1 
  2 α  2  K γ   1  α   K γ 
 K1       α  1  K  tut     K  u t
 β   β  (5.99)
 
 
u t  t 1  α  γ K
 K2  
βt
0  R t  t (1  α)  R t α  F t  t (1+α)+F t α

Newmark  method:
  M γ  1   γ 1  t
 M     M  t  1   ut   M   M  u 
  2β  2β   β tβ 

u t  t  u  γ M
t  t


t 2β β t

K  
 K γ   γ 
  K  tut   K   K  u t 
 u  γ K t  t

 2β   β  βt
  M γ  1   γ 1  t
 M1      M  t  1   ut   M   M  u
  2β  2β   β tβ 
u t  t  γt M  1
 M2  
t 2β
 K γ   γ 
 K1     K  tut   K   K  u t
 2β   β 
u γ K
t  t
 K2  
β t (5.100)
0  R t  t
F t  t

ATENA Theory 321

Wilson  method and Modified Wilson  method:
3 M 6
M   2 2
θt θ t
3
 K 1 K
θt
 1 6θ  2 1  3θ 2  2θ  t M   6 1  6θ  2   M   t
2

M  
1
 
u   2 
t
u
 2 θ3 2 θ3   θ t 2 θ3 
   
3u  θt M  2 
t  t

 M2  
θ 2 t 2
 1  3θ 2  2θ  t K 1  θ3  θ 2  t 2   1  6θ 2  2   K 1  2θ3  2θ  t 
K  
1
 
u  
t
  u t
2 θ 3
2 θ 3
  2 θ 3
2 θ 3

   
3u  K
t  t
 K2  
θt
Wilson  method :
R t  t  1 1  t  1
0  2
   2  3  R  F t  t  1  3  F t
θ  θ θ   θ 
Modified Wilson  method
(5.101)
1 1  F t t R t t
0    3  F t   3
θ θ  θ θ

The parameters  ,  are the integration parameters used by Newmark  and Hughes 
method. Their value is essential for convergence of this time marching scheme. It can be shown
1 1 1 1
that   ,   corresponds to linear acceleration within the time step. Values   ,  
2 6 2 4
yield constant acceleration. The integration scheme is unconditionally stable, if
1 1 1 1
  ,   0.25(   ) 2 and it is only conditionally stable for   ,   0.25(   ) 2 provided
2 2 2 2
that the stability limit is fulfilled:

1   1  2
2
 2
             
 2   2  2  
tcrit  (5.102)
 
 
2 
where  is the modal damping parameter.

The above defines the condition for time increment t for a linear conditionally stable case:

322
 t 1

Tn 2
 2  (5.103)
3
 0.551

As for Wilson  and Modified Wilson  method, they use  parameter. Its value is   1 and
the scheme is unconditionally stable for   1.4 . It essentially specifies the time, for which time
we calculate the governing equations (5.95), i.e t  t . For   1 Wilson  and Modified
Wilson  method yield the same solution expressions and equations, and these are also the same
1 1
as those for Newmark and Hughes methods with   ,   ,  0 .
2 6
Modified Wilson  method assembles the governing equations for time t  t . As a result, all
Von Neumann boundary conditions must be given for t  t , (e.g., concentrated load, load by
MASS_ACCELERATION etc.). It does not apply to Dirichlet boundary conditions that are (as
usually) input for t  t (e.g., prescribed displacement, acceleration etc.).
The fact that the Modified Wilson  method executes for t  t also affects output/draw of
results in structural material points. Within iterations (e.g., for monitors at iterations), they are
printed/drawn for t  t . After the iterations process has been completed, they are
printed/drawn for t  t as usual. Internal forces are always printed for t  t and the same for
external forces.
As described above Modified Wilson  method behaves in a bit nonstandard way. Particularly
input for R t t is unpractical. To alleviate these difficulties and inconvenience, Atena also offers
Wilson  method. Although it still solves the governing equations for time t  t , it uses
several extrapolations (e.g., R t t  R t   ( R t  t  R t ), F t t  F t   ( F t  t  F t ) ) so that it
suffices with R t  t and F t  t only. Consequently, it inputs all boundary conditions and
print/draw all result for t  t akin to any other solution method for dynamic analysis. On the
other hand, it is at price of accuracy because the extrapolation is linear, whereby the loading and
internal forces are not!
Remind that for dynamic analysis, concentrated forces, element body/boundary load, etc., is
input in the incremental form, and it is "cumulated" in the structure. The same applies to
prescribed displacements.
Prescribed velocities, accelerations, etc., must be input as total load. MASS_ACCELERATION
must also be input in total values (and in each step, it is also recalculated from scratch).

More details on the methods' convergency can be found in (Hughes 1983) and (Wood. 1990).

8.1 Structural Damping

As far as damping matrix C is concerned, Atena uses the well known proportional damping:

ATENA Theory 323

 C  M M   KK (5.104)
where  M ,  K are user-defined damping coefficients. These coefficients can be directly set as
user input data, or they can be generated based on knowledge of modal damping parameters  .
The parameters  are defined by

iT Ci  iT ( M M   K K )i  2ii (5.105)

where:
i is i-th structural eigenvector,

i is i-th structural eigenmode,

i is modal damping parameter associated with i and i .

Using the fundamental properties of eigenmodes iT M   1, iT K   i , we can rewrite (5.105)

 M   K i  2ii (5.106)
Equations (5.104) introduces 2 parameters for damping and, thus, if only 2 values of i are to be
used, they are directly substituted in (5.105), (resp. (5.106)) and solved for from this set of
equations.
However, in practice, structural damping is more complicated and some sort of compromise
must be done. In this case, structural damping properties are typically measured for more
eigenmodes, and optimal values of coefficients  M ,  K are calculated by the least square method,
i.e., we are seeking a minimum of the expression  M   K i  2ii  . It yields the following
2

set of equations

wi2 wi2i
M   K  wi2  2
i i2 i i i
(5.107)
 M  wi2  K  wi2i2  2 wi2ii
i i i

which is used to calculate the required damping parameters  M ,  K .

There exist other assumptions to account for structural damping; however, their use is typically
significantly more complex and more costly in terms of both RAM and CPU.

324
8.2 Spectral analysis
A proper selection of the solution time increment dt is essential for each dynamic analysis. If it is
too large, the computed results will suffer from unacceptable inaccuracies. We will probably
miss some important peaks in the loading history, and the analysis as a whole may even diverge.
On the other hand, the use of a too-small value of dt will yield an analysis that is pointlessly
expensive in terms of execution time and its demands towards CPU/RAM resources. In addition,
its postprocessing is more laborious and prone to errors.
The spectral analysis described in this section is designed to assist the engineer in setting suitable
dt. The main idea of the procedure is based on approximation of the structural loading f (t ) by
Fourier series f FT (t ) , i.e. f (t )  f FT (t ) , refer e.g., to http://en.wikipedia.org/wiki/Fourier_series
. Both f (t ), f FT (t ) have one independent variable, which is structural time t .

The function f FT (t ) is assembled in the following form:

a0 N  2   2 
f FT (t )    an sin  n t   bn cos  nt  (5.108)
2 n 1  T   T 

where N denotes the number of harmonics used for the approximation, n is harmonic-th id and
 2   2 
sin  n t  and cos  n t  are n-th approximation functions, (i.e., n-th harminics). Eqn.
 T   T 
(5.108) is suitable for approximation of a function (e.g. f (t ) ) in interval t  0...T  . Its
Fourier coefficients are calculated as follows, see
http://stelweb.asu.cas.cz/~slechta/fourier/fourier.html ,
http://www.mathstools.com/section/main/fourier_series_calculator#.VCFKkhZIpKI:

T
2
a0 
  f ( ) d
0

 2
T
2 

an  f ( ) sin  n  d  (5.109)
0  T 
 2
T
2 

bn  f ( ) cos  n  d
0  T 

Now let us introduce a coefficient cn  an2  bn2 and create a spectrum diagram of the loading.
 2 
For each harmonic from (5.108), plot a point, whose coordinates are  n, cn'  . Such a point
T 

ATENA Theory 325

shows how much important is the nth harmonic (i.e., the harmonic with circular frequency
2
n ) for the loading function, i.e., how much it is excited by the load function f (t ) .
T
The recommended solution time increment should be set so that the highest important harmonics
are integrated in about 10 steps, i.e.,
T
dt  min(nsignificant ) (5.110)
10

By default, the FFT analysis uses a full modal spectrum, i.e., n  1..N in (5.108). However, the
modal spectrum can be filtered, e.g. n  n1..m1 , n2 ..m2 ,...nk ..mk ,...nL ..m1 . In this case, only values n
from within the L intervals are used. This technique can be used to filter out some noise signals,
etc.

Let's take an example: Assume a simplified ElCentro accelerogram loading conditions, whose
acceleration in time are depicted by the green line in the figure below:

Let's approximate this function by the Fourier series. In the first case, we use 300 harmonics, i.e.,
N  300 . The approximated accelerations are shown by the blue line, as seen in the figure
above. In the second case, we use only 50 harmonics, and the corresponding approximation
function is drawn by the red line. Plotting the functions in more detail, it can be seen that the
approximation with N  300 is fairly accurate whilst the approximation with N  50 is rather
crude, see the figure below. This conclusion is endorsed by the calculated average relative
absolute error of the approximations. These are respectively 0.0314256 and 0.878789

326
The spectrum diagram below shows the contribution of individual approximation harmonics. It
detects what harmonics are or are not important. Looking at the diagram, we see that the highest
important harmonics is the one with log10(Tn )  1 , i.e. Tmin  0.1 . Therefore, the recommended
T
solution time increment is dt  min  0.01 . This dt should ensure reasonably accurate results
10
from dynamic analysis of a structure that is loaded by the investigated accelerogram.

ATENA Theory 327

The described spectrum analysis is fully supported by Atena (incl. all the plots). Its use is simple
as it requires only a few input commands. For more details, please refer to the examples of
commands for the input of a multilinear function (in the Atena input file documentation).

328
9 EIGENVALUES AND EIGENVECTORS ANALYSIS
This section describes methods used by ATENA software to calculate structural eigenvalues
and eigenvectors. Good knowledge of eigenmodes of a structure is, in many cases, essential
for understanding its behavior and selection of a method for its further analysis. It applies to
statics and particularly to dynamic analyses, in which case it helps to choose a proper time
increment in subsequent loading steps. It also helps in avoiding or reducing oscillation of the
structure.

9.1 Inverse Subspace Iteration

Currently, ATENA uses the Inverse subspace iteration method to compute the eigenvalues
and eigenvectors. The method is in detail described in (Bathe 1982), and hence, only its main
features are presented here. The current implementation can be used only for symmetric
matrices. The same applies to Jacobi and Rayleigh-Ritz methods that are mentioned later in
this section.
It consists of three methods; each of them is capable of solving the eigenvalue problem on its
own. However, if they are used simultaneously, they yield a very efficient scheme for solving
eigenvalues and eigenvectors of large sparse structural systems. The significant advantage of
this approach is that it is possible to search for a selected number of the lowest eigenmodes
only. The lowest eigenmodes are typically the most important for the behavior of the structure
because they represent the highest energy that the structure can absorb. On the other hand, the
highest eigenmode is of low importance, can be neglected, and thereby save a lot of CPU time
and other computational resources.
The Inverse subspace iteration consists of
 Inverse iteration method
 Rayleigh-Ritz method
 Jacobi method
It solves general eigenvalues and eigenvector problem of the following form:

Ku   2Mu (8.1)
where
K, M is stiffness and mass matrix of structure,
u is the vector of eigenvector’s nodal displacements,
 is circular eigenfrequency
We are looking for a non-trivial solution, so that we solve for  2 that comes from
det(K   2 M )  0 (8.2)

ATENA Theory 329

9.1.1 Rayleigh-Ritz Method
This method is used to transform the original eigenproblem of dimension n into an associated
eigenproblem of dimension m<<n. The solution is to search in space Vm  Vn . Let vectors
 k constitute linearly independent bases in Vn . An eigenvector ui is computed as a linear
combination ci of the base vectors  k , i.e.

ui  Ψci (8.3)
where
Ψ is the matrix of base vectors  k , k  1..m ,

ci is the vector of coefficients of the linear combination.

Rayleigh quotient is defined as

uiT Kui
 (ui )  (8.4)
uiT Mui
It can be proved that  (ui ) converges from the upper side to the corresponding circular
frequency i2 . The condition of a minimum of  (ui ) yields:

 (ui )
 0, k  1..m (8.5)
ci ,k

where ci ,k is k component of the vector ci

If we introduce
A  Ψ T KΨ , B  Ψ T MΨ (8.6)
the condition (8.5), after substituting (8.6), results in
Aci  i2 Bci (8.7)
which is an equation of eigenproblem of matrices A,B. This problem has dimension m, which
is significantly smaller than the original dimension n.

9.1.2 Jacobi Method

Jacobi method is used for the solution of full symmetric eigensystems of lower dimension. In
the frame of the Inverse subspace iteration method, it is used to solve (8.7). (Note, however,
that that the eigenproblem (8.7) can be used by any other method).
The Jacobi method is based on the property that if we have a matrix A, an orthogonal matrix
C, and a diagonal matrix D, whereby

CT AC  D (8.8)

330
then the matrices A and D have identical eigenvalues, and they are diagonal elements of the
matrix D. The transformation matrix C is calculated in an iterative manner
C  S1 S 2 .......S k , k  1.. (8.9)
where the individual S k has the following form

1 0 0 0
 1 
 
 1 
 
Sk  0 cos( ) 0  sin( ) 0 (8.10)
 0 1 0 
 
0 sin( ) cos( ) 0
0 0 0 1 

The entries cos( ),  sin( ) are put in i,j rows and columns, and they are constructed in the
way that they will zeroize aij after the transformation. The other diagonal elements are equal
to 1 and the remaining off-diagonal elements are 0.
In the case of a general eigenproblem, the whole procedure of constructing S k is very similar.
The matrices S k now adopt the shape

1 0 0 0
 1 
 
 1 
 
S k  0 1 0 a 0 (8.11)
 0 1 0 
 
0 b 1 0
0 0 0 1 

Notice that the matrix S k is not orthogonal anymore. The two variables a,b are calculated to
zeroize off-diagonal elements i,j of both matrices K and M. Eigenmodes of the problem are
then calculated as

aii'
  '
i
2
(8.12)
bii

where aii' , bii' are diagonal elements of transformed (and diagonalized) matrices A, B.

Eigenvectors of the problem are columns of the transformation matrix C.

9.1.3 Inverse Iteration Method

Inverse iteration method is carried out as follows: Starting with an initial transformation of
eigenvector ui ,1 , we calculate a vector of corresponding inertia forces (step 1)

ATENA Theory 331

 fi ,1  M ui ,1 (8.13)

Knowing f i ,1 , we can compute a new approximation of ui , (step 2)

ui ,2  K 1 fi ,1 (8.14)

and repeat the step 1. Hence, for iteration k we have

fi , k  M ui ,k
(8.15)
ui ,k 1  K 1 fi ,k

and the iterating is stop, when ui ,k 1  ui ,k . The above-described algorithm tends to converge
to the lowest eigenmodes. If any of these are to be skipped, the initial eigenvector ui ,1 must be
orthogonal to the corresponding eigenvectors. In practice, the vector ui ,k must be
orthogonalized with respect to the skipped eigenvectors even during the iterating procedure,
as the initial orthogonality may get (due to some round-off errors) lost.

9.1.4 Algorithm of Inverse Subspace Iteration

Having briefly described the above three methods, we can now proceed to the actual solution
algorithm of the Inverse subspace iteration method itself:
Step1- Inverse iteration method:
KU  MU
k 1 k

Step 2 - Raylegh quotient method:

A U  T KU

k 1 k 1 k 1
 T MU
B k 1  U 
k 1 k 1
(8.16)

Step 3  Jacobi method:

A k 1Ck 1  B k 1Ck 1Δ 2

Step 4 - Correct the eigenvectors:

UT  U T C
k 1 k 1 k 1

In the above
m is the number of projection eigenmodes (reasonably higher than the number of required
eigenmodes),
U k is the matrix of columnwise arranged m eigenvectors after k- th iteration,

A k 1 , B k 1 are transformed stiffness and mass matrices of the problem, (having dimension
m<<n),
Ck 1 is the matrix of eigenvectors of A k 1 , B k 1 , see (8.9)

332
Δ 2 is a matrix with eigenmodes (on its diagonal). Notice that eigenmodes for transformed and
the original eigenmode problem are the same.

The steps 1 thru 4 are repeated until the difference between the two subsequent operations is
negligible.
The solution algorithm (8.16) is in ATENA a bit modified in order to reduce CPU time and
RAM resources and is described below:
Step1- Inverse iteration method:
ˆ  MU
U k 1 k

 U
KU ˆ
k 1 k 1

Step 2 - Raylegh quotient method:

A U  T KU
 U T U ˆ
k 1 k 1 k 1 k 1 k 1

ˆ  MU
U 
k 1 k 1


B  U MU T  T U
U ˆ
k 1 k 1 k 1 k 1 k 1

Step 3  Jacobi method:

A k 1Ck 1  B k 1Ck 1Δ 2

Step 4 - Correct the eigenvectors:

UT  U T C (8.17)
k 1 k 1 k 1

The advantage of this procedure over the one defined in (8.16) is that now you don’t need to
store the original and factorized form of the matrix K. Only the factorized form is needed
during the iterations.
A special issue in this method is how to set up the initial vectors U1 . This is what we do in
ATENA. The first vector contains the diagonal elements of M. The next vectors are
constructed in the way that they have zeros everywhere except one entry. This entry
m
corresponds to maximum ii and is set to 1.
kii

The procedure as it is (because of the Inverse iteration method) cannot solve for zero
eigenmodes. This may be a problem, especially if we want to analyze structural rigid body
motions or spurious energy modes. If this is the case, shift matrix K by an arbitrary value s ,
i.e., solve the associated eigenproblem
(K  s M )us  s2 Mus (8.18)
The original eigenvalues and eigenvectors are then calculated by

ATENA Theory 333

 u  us
(8.19)
 2  s2  s
Another problem of Inverse subspace iteration is to compute multiple eigenvectors.
Unfortenatly, it is not that rare case and it happens, e.g., if the structure has an axis of
symmetry. The occurrence of multiple eigenmodes in the structure may yield non-orthogonal
eigenvectors, and thus, some eigenmodes can be missed. There are some techniques for
resolve this problem (Jendele 1987); however, they have not been implemented in ATENA
yet. Good news is that in reality, no eigenmodes are usually quite identical due to some
round-off errors. The case of multiple structural eigenmodes thus typically causes only some
worsening of accuracy and no eigenmode gets missed.
Nevertheless, if we want to be sure that no eigenmode was missed, we can assess it by Sturm
sequence property test.

9.1.5 Sturm Sequence Property Check

This property says (Bathe 1982) that if we have an eigenproblem (8.1), perform a shift s and
factorize that matrix (i.e., D is a diagonal matrix, L is a lower triangular matrix),
K  s M  L D LT (8.20)
then the number of negative diagonal elements in D equal to the number of eigenvalues
smaller than the shift  . This way, we can simply test, whether we missed an eigenvalue
with the calculated set of m eigenmodes or not
There are other methods that can be used to compute eigenvalues and eigenvectors of large
sparse eigensystems. Particularly popular is e.g., Lanczosh method (Bathe 1982). There exist
also several enhancements for the present Inverse subspace iteration method. For instance,
using a shifting technique may significantly improve the convergency of the method
(especially if some eigenvalues are close to each other).
These improved techniques may be implemented in the future. In any case, the current
ATENA implementation of eigenmodes analysis proves to solve the eigenmodes problem in
most cases quite successfully.

9.2 References
BATHE, K. J. (1982). Finite Element Procedures in Engineering Analysis. Englewood Cliffs,
New Jersey 07632, Prentice Hall, Inc.
JENDELE, L. (1987). The Orthogonalization of Multiple Eigenvectors in Subspace Iteration
Method. IKM - XI. Internationaler Kongress ueber Anwendungen der Mathematik in
der Ingenieurwissenschaften, Weimar.
WOOD., W. L. (1990). Practical-Time Stepping Schemes. Oxford, Clarenton Press.

334
10 GENERAL FORM OF DIRICHLET BOUNDARY CONDITIONS
A unique feature of ATENA software is the way in which it implements Dirichlet boundary
conditions. It supports to constraint any degree of freedom (DOF) by a linear of any number
of other structural DOFs. The proposed method of applying and processing the boundary
conditions is computationally efficient and memory economical because all constraint degrees
of freedoms (DOFs) are eliminated already during assembly of structural global stiffness
matrix and load vectors. The adopted concept has a wide range of use, and several of its
possibilities are discussed. At the end of the Section, a few samples are given.

10.1 Theory Behind the Implementation

A crucial part of a typical finite element analysis (whether linear or nonlinear) is the solution
of a set of linear algebraic equations in the following form
n

K
j 1
ij u j  ri , i  1..n (9.1)

where Kij is an element i, j of a predictor matrix K, (i.e., usually structural stiffness matrix),
ri is an external force (or unbalanced force), applied into i-th structural degree of freedom
(DOF), and finally ui is displacement (or displacement increment) at the same DOF. Such a
set of equations is always accompanied by many boundary conditions (BCs). They can be one
of the following:
Von-Neumann boundary conditions, (also called right-hand side (RHS) BCs). Number and
type of these BCs have no impact on dimension n of the problem (9.1). They are accumulated
in the vector r . This vector is assembled on the per-node basis for concentrated nodal forces
and/or per-element basis for nodal forces being equivalent to element loads.
The second type of boundary conditions are Dirichlet boundary conditions (also called left-
hand side (LHS) BCs). ATENA implementation of this type of BCs is now described. A
simple form of such BCs reads
ul  0, l  1, n 
(9.2)
ul  ul 0 , l  1, n 
These kinds of BCs typically represent structural supports with no displacements (the first
equation) or with prescribed displacements ul 0 , (the second equation). Although most LHS
BCs are of the above form (and only a few finite element packages offer anything better),
there are cases when a more general LHS BC is required. Therefore, ATENA software
provides a solution for implementing a form of Dirichlet BCs, where each degree of structural
freedom can be a linear combination of any other degrees of freedom. Mathematically, this is
expressed by
ul  ul 0  
k1, n 
lk uk , l  1, n  (9.3)

ATENA Theory 335

There are many cases in which the above form of Dirichlet conditions proves helpful. Some
examples are discussed later in the Chapter. The important point about implementing
Equations (9.3) is that they are utilized already during the assembling of the problem (9.1). It
means that if we have m of these BCs, then the final dimension of the matrix K becomes only
(n  m) . This fact significantly reduces requirements for computer storage.
In the following, we shall call such boundary conditions as “Complex Boundary Conditions”,
or CBCs, (see also ATENA Input file manual, where the same name is used).

10.1.1 Single CBC

The procedure of implementing Dirichlet BCs of the form (9.3) is now presented. Let us start
with just one BC equation (9.4). It says that ul equals to a constant prescribed displacement
ul 0 plus  lk multiple of a displacement u k .

ul  ul 0   lk uk (9.4)
Substituting (9.4) into the Equation (9.1) yields
n n


j 1, j  l
Kij u j  Kil ul  
j 1, j  l
Kij u j  Kil (ul 0   lk uk )  ri , i  1..n (9.5)

which after some manipulation can be simplified into the form

n

K
j 1
ij  K il lk  kj  u j  ri  Kil ul 0 , i  1..n (9.6)

The above set of equations could be already used to solve for the unknown displacements (or
displacement increments) u j .  kj stands for k. j Kronecker delta tensor. The trouble is,
however, that even though the matrix K might be symmetric, the set of equations (9.6) is not
symmetric anymore. Thus, to preserve the symmetry, add an  lk multiple of the row l , i.e.,

 n 
 lk    K lj  K ll lk  kj  u j    lk  rl  K ll ul 0  (9.7)
 j 1 
to the row k, i.e.,
n

 K
j 1
kj  K kl lk kj  u j  rk  K kl ul 0 (9.8)

This results in the row k getting the form

K 
n

kj  K kl lk  kj   lk  K lj  K ll lk  kj  u j 
j 1
n

K
j 1
kj   lk K lj   K kl lk   lk lk K ll   kj  u j  (9.9)

rk  K kl ul 0   lk  rl   lk K ll ul 0 

336
Hence, the final form of the governing set of equations will read
n

K
j 1
ij  K il lk  kj   ik lk K lj   ik  kj lk2 K ll  u j 
(9.10)
ri  K il ul 0   ik lk  rl  K ll ul 0 

The above equations can be written as

n

 K
j 1
ij u j  ri , i  1..n (9.11)

where
K 

 K11 ... K1i ... K1k  K1l lk ... K1 j ... K1n 

 ... ... ... ... ... ... ... ... ... 
 
 K i1 ... K ii ... K ik  K il lk ... K ij ... K in 
 
 ... ... ... ... ... ... ... ... ... 
 K k 1   lk K l1 ... K ki   lk K li ... K kk  2 K kl lk   kk  kk K ll2
lk ... K kj   lk K lj ... K kn   lk K ln 
 
 ... ... ... ... ... ... ... ... ... 
  (9.12)
K j1 ... K ji ... K jk  K jl lk ... K jj ... K jn
 
 ... ... ... ... ... ... ... ... ... 
 K n1 ... K ni ... K nk  K nl lk ... K nj ... K nn 
 

r 
 r1  K1l ul 0 
 ... 
 
 ri  Kil ul 0 
 
 ...  (9.13)
 rk  K kl ul 0   lk  rl  K ll ul 0  
 
 ... 
 rj  K jl ul 0 
 
 ... 
 rn  K nl ul 0 
 

 is now also
Providing the original matrix K is symmetric, i.e. K ij  K ji , then the matrix K
symmetric, i.e. Kij  K ji .

The displacement ul constrained by Equation (9.4) has a constant part ul 0 and a variable part
 lk u k , in which ul depends only on a single u k . A more general form of this BC would be if
ul depends on more displacements. It corresponds to the following form of the boundary
condition:

ATENA Theory 337

 ul  ul 0    lk uk (9.14)
k

In this case, the displacement ul is calculated as a constant part ul 0 plus a linear combination
 lk of displacements u k . k can be any displacement, i.e. k  1..n  . Replacing BC defined
by Equation (9.4) by the above Equation (9.14), the equation will change to the form
n
 
 K    K ij il lk  kj   ik lk Klj   ik  kj lk2 Kll   u j 

j 1 k ,k l  (9.15)
ri  Kil ul 0    ik lk  rl  K ll ul 0  
k ,k l

10.1.2 Multiple CBCs

The previous paragraph derived all the necessary relations for implementing a single
boundary condition. Now we will proceed to the case of multiple boundary conditions. Each
particular BC is again written in the form (9.14).
ul  ul 0    lk uk , l  1, n , l  {l1 , l2 ,...lr } (9.16)
k

The problem is, however, that displacements u k in (9.16) need not be free but fixed by
another BC, k can also run through l, (resulting in a recursive formulation), more BCs can be
specified for the same ul , a particular BC can be specified more times and in more forms etc.
For example, we may have a set of boundary equations that contains BCs
u1  u2 , u2  u1 (9.17)
or it can contain
u1  u2 , u2  u1 , u1  0.003 (9.18)
Both of these are correct. Unfortunately, the set can also contain
u1  u2 , u2  0.003, u2  u1 , u1  0.003 (9.19)
which is definitely wrong. Therefore, before any use of such set of BCs it is necessary to
detect and fix all redundant and contradictory multiple BCs present in it. It is easily done in
case of a simple set of BCs like the one above, but in real analyses with thousands of BCs in
the form (9.16), (some of them quite complex, i.e., k runs through many DOFs) the only way
to proceed is to treat (9.16) as a set of equations to be solved prior their use in (9.13).
Redundant BCs are ignored, and contradictory BCs are fulfilled after their summation. Let us
suppose that all structural constraints are specified in the set of equation (9.16). This can be
written in matrix form
ul  ul 0  A u k (9.20)
The above relationship represents a system of algebraic linear equations. The system is
typically non-symmetric, sparse and has a different number of rows (i.e., the number of BCs)
and columns, (i.e., the number of master and slave DOFs). Moreover, it is often ill-

338
conditioned, with a number of equations being linear combinations of the others, e.g., see the
example in (9.17). In the beginning, it is often not known which DOF is dependent, (i.e.,
slave) and which is independent, (i.e., master), (e.g., see also (9.17)).
Based on the above properties, the following procedure has been developed to solve the
problem (9.20):
1. Allocate "columns" for all slave and master DOFs, i.e., loop through all BCs in (9.16)
and allocate DOFs for both slave (i.e., LHS) and master (i.e., RHS) displacements ui .
2. Allocate storage for the matrix A and vectors ul , ul 0 in (9.20). The matrix has lr the
number of rows (see (9.16)) and lc the number of columns. lc is the dimension of the
DOFs map created in the point add. 1.
3. Assemble the matrix A and the vectors ul , ul 0 .
4. Detect constant BCs, i.e., ul  ul 0 and swap rows of A so that the rows corresponding
to constant BCs are pushed to the bottom.
5. Detect constant fixed DOFs, i.e., those with  lk  0 and variable fixed DOFs, i.e., that
are those dependent on other (master) DOFs and having  lk  0 .
6. Swap columns of A , so that the former DOFs are pushed to the right and the latter
DOFs to the left. The operations described at the point 5 and 6 are needed to assure
order, in which the constrained DOFs are eliminated. This is important for later
calculation of the structural reactions.
7. Using the Gauss method to triangulate the set of BC equations. The triangulation is
carried out in the standard way with the following differences.
a. Before eliminating entries of A located in column below akk , check for a non-
zero entry in the row k. If all its entries are zero, then ignore this row and
proceed to the next one. (It is the case of multiple BCs having the same
content).
b. Check, whether the row k specifies BC for constant or variable DOF, (see
explanation in the point 5 above). In the former case push the row k to the
bottom and proceed to the next row.
c. Swap columns  k ...lc  so that abs ( akk ) becomes maximum.
d. If akk  0 , swap lines  k ...lr  to find a non-zero entry in akk . Thereafter,
swap columns  k ...lc  to find maximum akk .
e. Eliminate entries below akk as usually.
As it was already mentioned, the matrix A is typically very sparse. Hence, a special storage
schemes are used that stores only non-zero entries of A. The data are stored by rows. Each
row has a number of data series, i.e., sequences or chunks of consecutive non-zero data
(within the row). The data are in a three-dimensional container. For each such chunk of data,
we also need to store its first position and length. This is done in two two-dimensional
containers.

ATENA Theory 339

As an example, suppose that we have the following matrix A:

 a11 0 0 0 0 0 0
0 a22 a23 0 0 a26 a27 

0 0 a33 0 0 0 0
 
A0 a42 0 a44 0 0 0 (9.21)
0 0 0 0 a55 a56 a57 
 
0 a62 0 0 0 a66 0
0 0 0 0 0 0 a77 

It is stored as follows ( A.data stores the actual data, A.rowbase stores base indices for non-
zero entries in rows, A.rowlength contains dimension of non-zero data chunks; all arranged
by rows):
A.data (1)(1)(1)  a11
A.data (2)(1)(1)  a22 , A.data(2)(1)(2)  a23 , A.data (2)(2)(1)  a26 , A.data (2)(2)(2)  a27
A.data (3)(1)(1)  a33 ,
A.data (4)(1)(1)  a42 , A.data(4)(2)(1)  a44
.......
A.rowbase(1)(1)  1
A.rowbase(2)(1)  2, A.rowbase(2)(2)  6
(9.22)
A.rowbase(3)(1)  3
A.rowbase(4)(1)  2, A.rowbase(4)(2)  4
.....
A.rowlength(1)(1)  1
A.rowlength(2)(1)  2, A.rowlength(2)(2)  2
A.rowlength(3)(1)  1
A.rowlength(4)(1)  1, A.rowlength(4)(2)  1
A number of optimisation techniques are used to speed up the process of triangularization of
the matrix A. These are summarized below:
The data are stored by rows and the elimination is also carried out by rows. (Row-based
storage is also more convenient during assembling the A from (9.16)). All the operations
needed for the elimination are carried out only for nonzero data. Their horizontal position is
stored in A.rowbase and A.rowlength , hence it is no problem to skip all zero entries. A
typical total number of columns lc , see (9.16), is of order from thousands to hundred
thousands DOFs. On the other hand a.rowlength is on average only of order of tens. This is
where the CPU savings comes from.
By the way, the same mapping of non-zero entries is also used for columns. This is achieved
by additional arrays A.columnbase and A.columnlength that are also included in the storage
scheme A. (Their construction is similar to A.rowbase and A.rowlength ; instead by rows

340
they are arranged by columns). These two additional arrays make possible to skip all zero
entries during column-base operations. The resulting significant increase of triangularization
speed pays off for a small amount of an extra RAM that is needed to store A.columnbase and
A.columnlength .
The adopted procedure of triangularization many times swaps lines and/or columns of A. In
view of the adopted storage scheme, it can be a quite expensive procedure. To alleviate this
problem, the storage scheme includes four additional arrays, namely A.rowindex ,
A.rowinverseindex , A.columnindex and A.rowinverseindex . In the
beginning, A.rowindex(i)  i and similarly A.rowinverseindex (i )  i, i  1...lc . When a row r1
should be swapped with a row r2 , the data in A.data remains unchanged and we swap only
corresponding row indices in A.rowindex , (and accordingly also entries in the array for
inverse mapping A.rowinverseindex ). The same strategy is used for swapping the columns.
As a result, any swapping operation does not require any moving of actual data (except of
swapping corresponding indices for mapping the rows and columns) and thus it is extremely
fast. On the other hand, in order to access aij we must use ai ' j ' , where i '  rowindex(i) and
j '  columnindex( j ) . The incurred CPU overhead is well acceptable, because the matrix A is
very sparse.

10.2 Application of Complex Boundary Conditions

This section presents several examples where the developed Dirichlet boundary conditions are
advantageously used. In each case, the corresponding finite element model exploits the
general form of BC defined by Equation (9.16).

10.2.1 Finite Element Mesh Refinement

Suppose we need to refine a mesh as shown in Fig. 10-1. The mesh should refine from 5
elements per row to 10 elements per row. The figure depicts three possible techniques to
achieve the goal.
In the case A, the fine and coarse parts of the mesh (consisting of quadrilateral elements) are
connected by a row of triangular elements. This way of mesh refinement is used the most
often. However, mixing quadrilateral and triangular elements is not always the best solution.
In the case B, the refinement is achieved by using hierarchical finite elements, see (Bathe
1982). The coarse mesh near the interface employs five nodes hierarchical elements. This
refinement is superior to the others; however, it requires special finite elements and special
mesh generator; both of these rarely available in a typical finite element package.
In the case C, the fine and coarse parts of the mesh are generated independently. After the
generation of all nodes and elements, the interface nodes are connected by complex boundary
conditions. For example, we can use ui  um , uk  un , u j  0.5um  0.5un . The main advantage
of this approach is that it is simple for both finite element pre/postprocessor and finite element
modeler (namely its finite element library). Hence it is preferable!

ATENA Theory 341

 Fig. 10-1 Mesh refinement
Note that all the above techniques are supported in ATENA finite element package, the last
one requiring implementation of CBCs in the form (9.14).

10.2.2 Mesh Generation Using Sub-Regions

This example demonstrates another advantage of using the proposed CBCs: It is possible to
generate meshes within sub-regions without requirement of nodes coincidence on their
interfaces. Because mesh structure on the sub-regions’ surfaces is not prescribed, this
approach provides more flexibility to mesh generation. This feature is heavily used by
ATENA 3D pre-processor.

Compatible meshes on the contact between the blocks

Fig. 10-2 Mesh generation from simple blocks

342
 Incompatible meshes on the contact between the blocks using CBCs
Fig. 10-2 (cont) Mesh generation from simple blocks
In the above example, two blocks are connected to form a structure, where the top (smaller)
block is placed atop of the bottom (larger) block. The position of the top block is arbitrary
with respect to the bottom block. Unless the concept of CBCs is used, the meshes on the
interface of the two blocks must be compatible (see top of Fig. 10-2). On the other hand, the
proposed CBCs allow using of incompatible meshes (see the bottom of Fig. 10-2). In this
case, the mesh in each block is generated independently, which is significantly simpler. After
they are done, the proposed CBCs are applied to connect the interface nodes. (Typically, the
surface with the finer mesh is fixed to the surface with the coarse mesh). The latter approach
also demonstrates the possibility of a mesh refinement while still using well-structured and
transparent meshes. This is particularly useful in the case of complex numerical models.

10.2.3 Discrete Reinforcement Embedded in Solid Elements

In this example, the described boundary conditions are used to simplify the modeling of the
reinforced concrete beam, see Fig. 10-3. The procedure to create the model is as follows.
Firstly, the mesh for solids, i.e., concrete elements are generated. It poses no problem, as it is
a regular mesh consisting of 48 quadrilateral elements. At this point, no attention needs to be
paid to the geometry of reinforcing bars present in the beam. Thereafter, the reinforcing bars
are inserted and their meshes are generated based on the existing mesh of solid elements. The
first step is to find all nodes, where the bar changes direction. These nodes are called principal
nodes; see e.g., node n in Fig. 10-3. Then, the intersection of all straight parts of the bar with
underlying solid elements are detected, e.g., the nodes m,p. Thus, all end nodes of embedded
bar elements are defined. The last step is to link displacements of the nodes of the bar to the
underlying solid elements.

ATENA Theory 343

 i j
m n
l p k

bar 1
bar 2

Fig. 10-3 Discrete bar reinforcement

For example, if we want to connect the node n to the an embedding solid element, i.e., to
4
nodes i,j,k,l, see Fig. 10-3, we use the standard interpolation u (r , s)   hi (r , s ) U i , where
i 1

hi ( r , s ), U i are element interpolation function and Ui are nodal displacements for the
underlying solid element, respectively. For displacement at the node n we can write
4
u (rn , sn )   hi (rn , sn ) U i . ( rn , sn ) are coordinates of the node n. Comparing this formula with
i 1

(9.3), it is obvious that  ni  hi ( rn , sn ), un 0  0 . Consequently, the bar DOFs are always

kinematically dependent on the DOFs of underlying solid elements.
This technique can also be applied when bond elements are inserted between solid and
embedded bar elements. This is treated in a separate paper by authors in ref. (Jendele, 2003).
Currently, ATENA software can generate discrete reinforcement to all 2D and 3D linear and
nonlinear elements (triangles, quads, tetrahedral elements, wedges, bricks…). The user only
draws the position of the principal nodes of reinforcement bars and the rest is done
automatically.

10.2.4 Curvilinear Nonlinear Beam and Shell Elements

In the following text, another possible use of the present boundary conditions is presented. A
curvilinear nonlinear beam from Chapter 3.17 is discussed. A particular feature we would like
to point out here is that although it originally has only three displacements and three rotations
in the nodes 13,14,15, see Fig. 3-40, its implementation in ATENA has also 3 displacements
in the nodes 1 to 12. However, these DOFs are linked to the original DOFs in the nodes 13 to
15 by the proposed CBCs. This concept has several advantages.
 The beam finite element has native 3D geometry and its pre- and post-processing
visualization is more realistic than using its original 1D geometry.
 It is simple to connect such beam elements to any adjacent 3D finite elements, e.g.,
brick elements.
 Mesh generation is easily done by any 3D solid element generator that can pull off
nonlinear hexahedral elements. It suffices to generate only the nodes 1 to 12 (with 3
displacement DOFs) and the three original beam nodes (each beam node has 3
displacement and 3 rotation DOFs) are generated automatically. The pre-processor
need not to support rotational DOFs.

344
  The post-processing of this element and an ordinary nonlinear hexahedral element is
the same. Consequently, this element does not need any extra support for the
visualisation of the results. It makes its implementation and use simple.
Derivation of all  ij coefficients and ui 0 constants for all nodes 1 to 12 is beyond the scope
of this document. Nevertheless, a similar procedure is used, as it was in the previous example.
ATENA package also covers Ahmad element for curved shell structures, see Chapter 3.12.
The usual 2D shape of the shell element is in the same manner, replaced by geometry of a 3D
nonlinear hexahedral element. Originally, the shell element has 3 displacements and 2
rotations at each node located in the middle thickness of the shell. These 5 DOFs are in by use
of CBCs replaced by 3 displacements at the top and 2 displacements at the bottom at the
respective nodes from the hexahedron, (i.e., brick) geometry. Advantages of this approach
are the same as those in the case of the curvilinear beam above: simpler pre/post-processing,
simpler connection to the adjacent 3D elements, no need to support rotational DOFs during
pre/post-processing, no need for extra support for geometry of the shell element.

10.3 References
BATHE, K.J. (1982), Finite Element Procedures In Engineering Analysis, Prentice-Hall, Inc.,
Englewood Cliffs, New Jersey 07632, ISBN 0-13-317305-4.
JENDELE, L, CERVENKA, J, CERVENKA V., " Bond in Finite Element Modelling of
Reinforced Concrete", Proceedings Euro-C 2003, Computatinal Modelling of Concrete
Structures, Swets & Zeitlinger, Lisse, The Netherlands, ISBN 90 809 536 3, 793-8036

ATENA Theory 345

INDEX
Microplane ................................................... 76
A
peak stress .................................................... 22
Adam-Bashfoth integration ............................ 295
Rankine fracturing........................................ 35
axis
reinforcement ............................................... 67
hydrostatic ................................................... 38
reinforcement bond ...................................... 72

B SBETA model .............................................. 17

biaxial......................................................... 17, 25 SBETA model parameters............................ 33

boundary condition........................................... 12 shear and stiffness in crack .......................... 29

complex ....................................................... 13 size effect ..................................................... 24

simple........................................................... 13 smeared cracks ............................................. 27

smeared cracks-fixed.................................... 27
C smeared cracks-rotated................................. 28
Clapeyron divergent theorem ............................. 3 tensile failure................................................ 26
confinement...................................................... 55 tension stiffening.................................... 30, 47
constitutive model ............................................ 15 transformation .............................................. 16
after peak behaviour .................................... 23 variants......................................................... 44
Bigaj............................................................. 74 Von Mises .................................................... 59
CC3DCementitious...................................... 34 constitutive tensor............................................... 6
CC3DNonLinCementitious ......................... 34 constutive model
CC3DNonLinCementitious2 ....................... 34 plasticity and crushing ................................. 38
CC3DNonLinCementitious2Fatigue ........... 48 convergence35, 42, 43, 80, 224, 225, 227, 231,
CC3DNonLinCementitious2User................ 34 232, 246, 296, 322

CC3DNonLinCementitious2Variable.......... 34 convergency............................................ 323, 334

CEB-FIP 1990 ............................................. 72 crack

compressive failure...................................... 25 fixed ............................................................. 48

compressive stress ....................................... 29 rotated .......................................................... 48

crack opening law ........................................ 20 cracking .................................... 17, 25, 35, 41, 53

crack spacing ............................................... 48 Cracking ........................................................... 19

definition.................15, 85, 215, 241, 285, 335 Crank-Nicholson integration .......................... 295

Drucker Prager............................................. 62 creep ............................................................... 241

equivalent uniaxial law ................................ 18 basics.......................................................... 241

fracture process............................................ 25 constitutive models .................................... 247

Hooke’s law................................................. 15 Dirichlet series ........................................... 243

HORDIJK law ............................................. 20 parameters needed by models ............ 253, 255

interface model ............................................ 63 retardation times......................................... 245

localization limiters ..................................... 24 solution parameters .................................... 249

Step by Step Method .................................. 244

ATENA Theory 347

creep model shell element .............................................. 130
CCModelB3............................................... 248 spring element ............................................ 110
CCModelB3Improved ............................... 248 triangular element ........................................ 97
CCModelBP_KX....................................... 248 truss 2D/3D .................................................. 87
CCModelBP1_DATA................................ 248 Fire analysis.................................................... 312
CCModelBP2_DATA................................ 248 flux
CCModelCEB_FIP78 ................................ 248 heat............................................................. 286
CCModelCSN731202................................ 248 moisture...................................................... 285
CCModelGeneral....................................... 248 formulation ......................................................... 2
creep model Euler............................................................... 2
CCModelACI78......................................... 248 Lagrange ...................................... 2, 7, 58, 149
Updated Lagrange .......................................... 2
D
damping...................296, 319, 322, 323, 324, 325 G
Dirichlet conditions ........................................ 335 governing equations............................................ 7
discretisation Green theorem ................................................ 293
spatial......................................................... 288
H
temporal ..................................................... 294
heat ................................................................. 286
dynamic .......................................................... 319
Heterosis element ........................... 130, 142, 149
E hierarchical formulation.................................... 85
eigenvalues..............................329, 331, 333, 334 high performance fiber reinforced concrete...... 52
eigenvectors.............................329, 332, 333, 334 HPFRCC........................................................... 52
equibiaxial ........................................................ 57 hydration................................................. 286, 297
Hydrocarbon fires ........................................... 312
F
fatigue............................................................... 48 I
fiber reinforced concrete ............................ 21, 52 integration points
finite element CCBeamNL element .................................. 183
Ahmad element.......................................... 130 CCIsoBrick element ................................... 103
axisymmetric element................................ 128 CCIsoQuad element ..................................... 92
brick quadrilateral element .......................... 99 CCIsoTetra element ................................... 101
extwernal cable element ............................ 117 CCIsoTrianle element .................................. 99
hexahedral element ...................................... 99 CCIsoWedge element ................................ 108
interface element........................................ 124 Q10/Q10Sbeta element .............................. 114
nonlinear 3D beam element ....................... 173 shell/Ahmad element.................................. 140
plane quadrilateral element.......................... 91 truss 2D/3D element..................................... 87
Q10 element............................................... 112 integration shell element-summart.................. 149
Q10Sbeta element...................................... 112 interpolation function ....................................... 85
reinforcement bar with prescribed bond .... 118 CCBeamNL element .................................. 175

348
 CCIsoQuad element..................................... 92 formulation..................................................... 2
CCIsoTetra element ................................... 101 general............................................................ 2
CCIsoTriangle element ................................ 98
R
CCIsoWedge element ................................ 106
Rayleigh Ritz method ..................................... 330
problem discretisation.................................... 9
Q10/Q10Sbetaelement............................... 113 S
shell/Ahmad element ................................. 140 Serendipity element ........................ 130, 141, 149
truss 2D/3D element .................................... 87 shape function................................................... 85
introduction ........................................................ 1 SHCC................................................................ 52
Inverse Iteration method................................. 331 S-N curve.................................................... 48, 50
Inverse Subspace Iteration method................. 329 solver .............................................................. 215
isoparametric formulation ................................ 85 Arc length................................................... 226
Arc Length step.......................................... 232
J
Consistently Linearized Method ................ 229
Jacobi method................................................. 330
Crisfield Method ........................................ 231
L direct sparse................................................ 217
Lagrangian element .........................130, 141, 149 Explicit Orthogonal Method ...................... 230
Cholesky .................................................... 216
M
iterative ...................................................... 217
moisture.......................................................... 285
linear .......................................................... 215
multiaxial........................................................ 242
Linear search .............................................. 232
Multipoint constraint ...................................... 335
Modified Newton Raphson ........................ 225

N Newton Raphson ........................................ 223

Newmark ........................................................ 319 nonlinear .................................................... 223

nonlinearity ........................................................ 1 Normal Update Method.............................. 229

types............................................................... 1  scaling parameter .................................... 233

nonlinearity classification .................................. 1 strain ................................................................... 5

Almansi .......................................................... 6
O engineering..................................................... 5
oscilations....................................................... 296 fracturing...................................................... 36
Green Lagrange.............................................. 5
P
Strain Hardening Cementitious Composite ...... 52
Palmgren-Miner hypothesis ............................. 49
Strain Hardening Cementitious Composites..... 52
plasticity ........................................................... 41
stress ................................................................... 4
principle of virtual displacements .................. 3, 6
2nd Piola-Kirchhoff ....................................... 4
principle of virtual forces ................................... 3
Cauchy ........................................................... 4
problem .............................................................. 2
stress/strain smoothing ..................................... 10
configuration.................................................. 3
lumped.......................................................... 11
FEM discretisation......................................... 9

ATENA Theory 349

 variational .................................................... 11 U
Sturm sequence check .................................... 334
uniaxial ....................... 17, 26, 39, 57, 59, 67, 118

T W
time equivalent ............................................... 286
Wöhler curve .................................................... 50
Transport analysis .......................................... 285
triaxial ........................................................ 55, 57 X
Xi-Bazant model............................................. 296

350