Are you sure?
This action might not be possible to undo. Are you sure you want to continue?
REINFORCED AND PRESTRESSED CONCRETE BEAMS
USING FINITE ELEMENT ANALYSIS
by
Anthony J. Wolanski, B.S.
A Thesis submitted to the
Faculty of the Graduate School,
Marquette University,
in Partial Fulfillment of
the Requirements for
the Degree of
Master of Science
Milwaukee, Wisconsin
May, 2004
PREFACE
Several methods have been utilized to study the response of concrete structural
components. Experimental based testing has been widely used as a means to analyze
individual elements and the effects of concrete strength under loading. The use of finite
element analysis to study these components has also been used.
This thesis is a study of reinforced and prestressed concrete beams using finite
element analysis to understand their loaddeflection response. A reinforced concrete
beam model is studied and compared to experimental data.
The parameters for the reinforced concrete model were then used to model a
prestressed concrete beam. Characteristic points on the loaddeformation response curve
predicted using finite element analysis were compared to theoretical (handcalculated)
results.
Conclusions were then made as to the accuracy of using finite element modeling
for analysis of concrete. The results compared well to experimental and hand calculated.
ACKNOWLEDGMENTS
This research was performed under the supervision of Dr. Christopher M. Foley. I am
extremely grateful for the guidance, knowledge, understanding, and numerous hours
spent helping me complete this thesis. Appreciation is also extended to my thesis
committee, Dr. Stephen M. Heinrich and Dr. Baolin Wan, for their time and efforts.
I would like to thank my parents, John and Sue Wolanski, my brother, John
Wolanski, and my sister, Christine Wolanski for their understanding, encouragement and
support. Without my family these accomplishments would not have been possible.
TABLE OF CONTENTS
PAGE
LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi
LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
CHAPTER 1 – INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Objectives and Outline of Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
CHAPTER 2 – LITERATURE REVIEW AND SYNTHESIS . . . . . . . . . . . . . . . . . 4
2.0 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.1 ExperimentBased Testing of Concrete . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Finite Element Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.3 Failure Surface Models for Concrete . . . . . . . . . . . . . . . . . . . . . . . . 10
2.4 FE Modeling of Steel Reinforcement . . . . . . . . . . . . . . . . . . . . . . . . 13
2.5 Direction for Present Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
CHAPTER 3 – CALIBRATION MODEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.0 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.1 Experimental Beam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.2 ANSYS Finite Element Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.2.1 Element Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.2.2 Real Constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2.3 Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2.4 Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.2.5 Meshing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.2.6 Numbering Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.2.7 Loads and Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . 35
3.2.8 Analysis Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.2.9 Analysis Process for the Finite Element Model . . . . . . . . . . . 40
3.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.3.1 Behavior at First Cracking . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.3.2 Behavior at Initial Cracking . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.3.3 Behavior Beyond First Cracking . . . . . . . . . . . . . . . . . . . . . . 43
3.3.4 Behavior of Reinforcement Yielding and Beyond . . . . . . . . 44
3.3.5 Strength Limit State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.3.6 LoadDeformation Response . . . . . . . . . . . . . . . . . . . . . . . . . 48
CHAPTER 4 – PRESTRESSED CONCRETE BEAM MODEL . . . . . . . . . . . . . . . 49
4.0 Introdution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.1 Finite Element Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.1.1 Real Constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.1.2 Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.1.3 Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.1.4 Analysis Process for the Finite Element Model . . . . . . . . . . . 53
4.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.2.1 Application of Effective Prestress . . . . . . . . . . . . . . . . . . . . 54
4.2.2 SelfWeight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.2.3 Zero Deflection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.2.4 Decompression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.2.5 Initial Cracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.2.6 Secondary Linear Region . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.2.7 Behavior of Steel Yielding and Beyond . . . . . . . . . . . . . . . . . 60
4.2.8 Flexural Limit State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
CHAPTER 5 – CONCLUSIONS AND RECOMMENDATIONS . . . . . . . . . . . . . . 63
5.0 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.2 Recommendations for Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . 64
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Theoretical Calculations for Calibration Model . . . . . . . . . . . . . . . . . . . . . . . 69
Theoretical Calculations for Prestressed Model . . . . . . . . . . . . . . . . . . . . . . . 73
LIST OF FIGURES
FIGURE PAGE
2.1 Typical Cracking of Control Beam at Failure (Buckhouse 1997) . . . . . . 5
2.2 Reinforced Concrete Beam With Loading (Faherty 1972) . . . . . . . . . . . 6
2.3 FEM Discretization for a Quarter of the Beam (Kachlakev, et al. 2001) . 8
2.4 Load vs. Deflection Plot (Kachlakev, et al. 2001) . . . . . . . . . . . . . . . . . . 9
2.5 Typical Cracking Signs in Finite Element Models
(Kachlakev, et al. 2001) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.6 Failure Surface of Plain Concrete Under Triaxial Conditions (Willam
and Warnke 1974) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.7 Three Parameter Model (Willam and Warnke 1974) . . . . . . . . . . . . . . . . 11
2.8 Models for Reinforcement in Reinforced Concrete (Tavarez 2001) . . . . 14
3.1 Loading and Supports for the Beam (Buckhouse 1997) . . . . . . . . . . . . . . 16
3.2 Typical Detail for Control Beam Reinforcement (Buckhouse 1997) . . . . 17
3.3 Failure in Flexure (Buckhouse 1997) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.4 Load vs. Deflection Curve for Beam C1 (Buckhouse 1997) . . . . . . . . . . 19
3.5 Solid 65 Element (ANSYS, SAS 2003) . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.6 Solid 45 Element (ANSYS, SAS 2003) . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.7 Link 8 Element (ANSYS, SAS 2003) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.8 Uniaxial StressStrain Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.9 Volumes Created in ANSYS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.10 Mesh of the Concrete, Steel Plate, and Steel Support . . . . . . . . . . . . . . . . 32
3.11 Reinforcement Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.12 Boundary Conditions for Planes of Symmetry . . . . . . . . . . . . . . . . . . . . . 35
3.13 Boundary Condition for Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.14 Boundary Conditions at the Loading Plate . . . . . . . . . . . . . . . . . . . . . . . . 37
3.15 1
st
Crack of the Concrete Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.16 Cracking at 8,000 and 12,000 lbs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.17 Increased Cracking After Yielding of Reinforcement . . . . . . . . . . . . . . . . 46
3.18 Failure of the Concrete Beam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.19 Load vs. Deflection Curve Comparison of ANSYS and Buckhouse (1997) 48
4.1 StressStrain Curve for 270 ksi strand . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.2 Load vs. Deflection Curve for Prestressed Concrete Model . . . . . . . . . . . . 55
4.3 Deflection due to prestress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.4 Bursting Phenomenon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.5 Localized Cracking From Effective Prestress Application . . . . . . . . . . . . 58
4.6 Cracking at 12,000 and 20,000 lbs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.7 Cracking at Flexural Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
A.1 Loading of Beam with Supports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
A.2 Transformed CrossSection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
B.1 Typical Prestressed Concrete Beam with Supports . . . . . . . . . . . . . . . . . 74
LIST OF TABLES
TABLE PAGE
3.1 Properties for Steel and Concrete (Buckhouse 1997) . . . . . . . . . . . . . . . . . 17
3.2 Test Data for Control Beam C1 (Buckhouse 1997) . . . . . . . . . . . . . . . . . . 19
3.3 Element Types for Working Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.4 Real Constants for Calibration Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.5 Material Models for the Calibration Model . . . . . . . . . . . . . . . . . . . . . . . . 25
3.6 Dimensions for Concrete, Steel Plate, and Steel Support Volumes . . . . . . 31
3.7 Mesh Attributes for the Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.8 Commands Used to Control Nonlinear Analysis . . . . . . . . . . . . . . . . . . . . 38
3.9 Commands Used to Control Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.10 Nonlinear Algorithm and Convergence Criteria Parameters . . . . . . . . . . . 39
3.11 Advanced Nonlinear Control Settings Used . . . . . . . . . . . . . . . . . . . . . . . . 39
3.12 Load Increments for Analysis of Finite Element Model . . . . . . . . . . . . . . . 40
3.13 Deflection and Stress Comparisons At First Cracking . . . . . . . . . . . . . . . 43
3.14 Deflections of Control Beam (Buckhouse 1997) vs. Finite Element Model
At Ultimate Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.1 Real Constants for Prestressed Beam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.2 Values for Multilinear Isotropic StressStrain Curve . . . . . . . . . . . . . . . . . 52
4.3 Load Increments for Analysis of Prestressed Beam Model . . . . . . . . . . . . 53
4.4 Analytical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
CHAPTER 1
INTRODUCTION
1.1 General
Concrete structural components exist in buildings and bridges in different forms.
Understanding the response of these components during loading is crucial to the
development of an overall efficient and safe structure.
Different methods have been utilized to study the response of structural
components. Experimental based testing has been widely used as a means to analyze
individual elements and the effects of concrete strength under loading. While this is a
method that produces real life response, it is extremely time consuming, and the use of
materials can be quite costly. The use of finite element analysis to study these
components has also been used. Unfortunately, early attempts to accomplish this were
also very time consuming and infeasible using existing software and hardware.
In recent years, however, the use of finite element analysis has increased due to
progressing knowledge and capabilities of computer software and hardware. It has now
become the choice method to analyze concrete structural components. The use of
computer software to model these elements is much faster, and extremely costeffective.
To fully understand the capabilities of finite element computer software, one must
look back to experimental data and simple analysis. Data obtained from a finite element
analysis package is not useful unless the necessary steps are taken to understand what is
happening within the model that is created using the software. Also, executing the
necessary checks along the way is key to make sure that what is being output by the
computer software is valid.
By understanding the use of finite element packages, more efficient and better
analyses can be made to fully understand the response of individual structural
components and their contribution to a structure as a whole. This thesis is a study of
reinforced and prestressed concrete beams using finite element analysis to understand the
response of reinforced and prestressed concrete beams due to transverse loading.
1.2 Objectives and Outline of Thesis
The objective of this thesis was to investigate and evaluate the use of the finite element
method for the analysis of reinforced and prestressed concrete beams The following
procedure was used to meet this goal.
First, a literature review was conducted to evaluate previous experimental and
analytical procedures related to reinforced concrete components. Second, a calibration
model using a commercial finite element analysis package (ANSYS, SAS 2003) was set
up and evaluated using experimental data. A mildsteel reinforced concrete beam with
flexural and shear reinforcement was analyzed to failure and compared to experimental
results to calibrate the parameters in ANSYS (SAS 2003) for later analyses.
Based on the results obtained from the calibration model and the
analysis/modeling parameters set by this model, a prestressed concrete beam was
analyzed from initial prestress to flexural failure. Deflections, stresses, and cracking of
the concrete beam were analyzed at different key points along the way. These key points
include initial prestress, addition of selfweight, zero deflection point, decompression,
initial cracking, yielding of steel, and failure.
Discussion of the results obtained for the calibration model and the prestressed
concrete beam model is also provided. Conclusions regarding the analysis are then
drawn and recommendations for further research are made.
CHAPTER 2
LITERATURE REVIEW AND SYNTHESIS
2.0 Introduction
To provide a detailed review of the body of literature related to reinforced and prestressed
concrete in its entirety would be too immense to address in this thesis. However, there
are many good references that can be used as a starting point for research (ACI 1978,
MacGregor 1992, Nawy 2000). This literature review and introduction will focus on
recent contributions related to FEA and past efforts most closely related to the needs of
the present work.
The use of FEA has been the preferred method to study the behavior of concrete
(for economic reasons). Willam and Tanabe (2001) contains a collection of papers
concerning finite element analysis of reinforced concrete structures. This collection
contains areas of study such as: seismic behavior of structures, cyclic loading of
reinforced concrete columns, shear failure of reinforced concrete beams, and concrete
steel bond models.
Shing and Tanabe (2001) also put together a collection of papers dealing with
inelastic behavior of reinforced concrete structures under seismic loads. The monograph
contains contributions that outline applications of the finite element method for studying
postpeak cyclic behavior and ductility of reinforced concrete columns, the analysis of
reinforced concrete components in bridge seismic design, the analysis of reinforced
concrete beamcolumn bridge connections, and the modeling of the shear behavior of
reinforced concrete bridge structures.
The focus of these most recent efforts is with bridges, columns, and seismic
design. The focus of this thesis is the study of nonprestressed and prestressed flexural
members. The following is a review and synthesis of efforts most relevant to this thesis
discussing FEA applications, experimental testing, and concrete material models.
2.1 ExperimentBased Testing Of Concrete
Buckhouse (1997) studied external flexural reinforcement of existing concrete beams.
Three concrete control beams were cast with flexural and shear reinforcing steel. Shear
reinforcement was placed in each beam to force a flexural failure mechanism.
All three beams were loaded with transverse point loads at third points along the
beams. Loading was applied to the beams until failure occurred as shown in Figure 2.1.
Figure 2.1 – Typical Cracking of Control Beam at Failure (Buckhouse 1997)
The mode of failure characterized by the beams was compression failure of the concrete
in the constant moment region (flexural failure). All failures were ductile, with
significant flexural cracking of the concrete in the constant moment region.
Loaddeflection curves were plotted for each beam and compared to predicted
ultimate loads. This thesis will utilize the experimental results of these control beam tests
for calibration of the FE models.
2.2 Finite Element Analysis
Faherty (1972) studied a reinforced and prestressed concrete beam using the finite
element method of analysis. The two beams that were selected for modeling were simply
supported and loaded with two symmetrically placed concentrated transverse loads
(Figure 2.2).
Figure 2.2 – Reinforced Concrete Beam With Loading (Faherty 1972)
The analysis for the reinforced concrete beam included: nonlinear concrete
properties, a linear bondslip relation, bilinear steel properties, and the influence of
progressive cracking of the concrete. The transverse loading was incrementally applied
and ranged in magnitude from zero to a load well above that which initiated cracking.
Because the loading and geometry of the beam were symmetrical, only one half of the
beam was modeled using FEA. The finite element model produced very good results that
compared well with experimental results in Janney (1954).
Faherty (1972) also analyzed a prestressed concrete beam that included: non
linear concrete properties, a linear bond slip relation with a destruction of the bond
between the steel and concrete, and bilinear steel properties. The dead load, release of
the prestressing force, the elastic prestress loss, the time dependent prestress loss, and the
loss of tensile stress in the concrete as a result of concrete rupture were applied as single
loading increments, whereas the transverse loading was applied incrementally. Only
three finite element models of the prestressed beam were implemented (or used): two
uncracked sections, and a partially cracked section. Symmetry was once again utilized.
These results for the prestressed beam showed that deflections computed using
the finite element model were very similar to those observed by Branson, et al. (1970).
However, the loaddeflection curve past the cracking point was not generated because
only three distinct cracking patterns were used for this analysis. It was recommended
that additional analysis of the prestressed concrete beam should be undertaken after a
procedure is developed for modeling the tensile rupture of the concrete. The model
utilized in this research required the beam to be unloaded and the finite element model
redefined as each crack is initiated or extended.
Kachlakev, et al. (2001) used ANSYS (SAS 2003) to study concrete beam
members with externally bonded Carbon Fiber Reinforced Polymer (CFRP) fabric.
Symmetry allowed one quarter of the beam to be modeled as shown in Figure 2.3.
Figure 2.3 – FEM Discretization for a Quarter of the Beam (Kachlakev, et al. 2001)
At planes of symmetry, the displacement in the direction perpendicular to the plane was
set to zero. A single line support was utilized to allow rotation at the supports. Loads
were placed at third points along the full beam on top of steel plates. The mesh was
refined immediately beneath the load (Figure 2.3). No stirruptype reinforcement was
used.
The nonlinear NewtonRaphson approach was utilized to trace the equilibrium
path during the loaddeformation response. It was found that convergence of solutions
for the model was difficult to achieve due to the nonlinear behavior of reinforced
concrete material. At certain stages in the analysis, load step sizes were varied from large
(at points of linearity in the response) to small (when instances of cracking and steel
yielding occurred). The loaddeflection curve for the nonCFRP reinforced beam that
was plotted shows reasonable correlation with experimental data (McCurry and
Kachlakev 2000) as shown in Figure 2.4.
Figure 2.4 – Load vs. Deflection Plot (Kachlakev, et al. 2001)
Also, concrete crack/crush plots were created at different load levels to examine the
different types of cracking that occurred within the concrete as shown in Figure 2.5.
The different types of concrete failure that can occur are flexural cracks,
compression failure (crushing), and diagonal tension cracks. Flexural cracks (Figure
2.5a) form vertically up the beam. Compression failures (Figure 2.5b) are shown as
circles. Diagonal tension cracks (Figure 2.5c) form diagonally up the beam towards the
loading that is applied.
Figure 2.5 – Typical Cracking Signs in Finite Element Models: a)Flexural Cracks,
b)Compressive Cracks, c)Diagonal Tensile Cracks (Kachlakev, et al. 2001)
This study indicates that the use of a finite element program to model
experimental data is viable and the results that are obtained can indeed model reinforced
concrete beam behavior reasonably well.
2.3 Failure Surface Models For Concrete
Willam and Warnke (1974) developed a widely used model for the triaxial failure surface
of unconfined plain concrete. The failure surface in principal stressspace is shown in
Figure 2.6. The mathematical model considers a sextant of the principal stress space
because the stress components are ordered according to
1 2 3
σ σ σ ≥ ≥ . These stress
components are the major principal stresses.
The failure surface is separated into hydrostatic (change in volume) and deviatoric
(change in shape) sections as shown in Figure 2.7. The hydrostatic section forms a
meridianal plane which contains the equisectrix
1 2 3
σ σ σ = = as an axis of revolution (see
Figure 2.6). The deviatoric section in Figure 2.7 lies in a plane normal to the equisectrix
(dashed line in Figure 2.7).
Figure 2.6 – Failure Surface of Plain Concrete Under Triaxial Conditions (Willam and
Warnke 1974)
Figure 2.7 – Three Parameter Model (Willam and Warnke 1974)
The deviatoric trace is described by the polar coordinates r , and θ where r is the
position vector locating the failure surface with angle, θ . The failure surface is defined
as:
1 1
1
( )
a a
cu cu
z f r f
σ τ
θ
+ = (2.1)
where:
a
σ and
a
τ = average stress components
z = apex of the surface
cu
f = uniaxial compressive strength
The opening angles of the hydrostatic cone are defined by
1
ϕ and
2
ϕ . The free
parameters of the failure surface z and r , are identified from the uniaxial compressive
strength (
cu
f ), biaxial compressive strength (
cb
f ), and uniaxial tension strength (
t
f )
The Willam and Warnke (1974) mathematical model of the failure surface for the
concrete has the following advantages:
1. close fit of experimental data in the operating range;
2. simple identification of model parameters from standard test data;
3. smoothness (e.g. continuous surface with continuously varying tangent
planes);
4. convexity (e.g. monotonically curved surface without inflection points).
Based on the above criteria, a constitutive model for the concrete suitable for FEA
implementation was formulated.
This constitutive model for concrete based upon the Willam and Warnke (1974)
model assumes an appropriate description of the material failure. The yield condition can
be approximated by three or five parameter models distinguishing linear from nonlinear
and elastic from inelastic deformations using the failure envelope defined by a scalar
function of stress ( ) 0 f σ = through a flow rule, while using incremental stressstrain
relations. The parameters for the failure surface can be seen in Figure 2.7.
During transition from elastic to plastic or elastic to brittle behavior, two
numerical strategies were recommended: proportional penetration, which subdivides
proportional loading into an elastic and inelastic portion which governs the failure surface
using integration, and normal penetration, which allows the elastic path to reach the yield
surface at the intersection with the normal therefore solving a linear system of equations.
Both of these methods are feasible and give stress values that satisfy the constitutive
constraint condition. From the standpoint of computer application the normal penetration
approach is more efficient than the proportional penetration method, since integration is
avoided.
2.4 FE Modeling of Steel Reinforcement
Tavarez (2001) discusses three techniques that exist to model steel reinforcement in finite
element models for reinforced concrete (Figure 2.8): the discrete model, the embedded
model, and the smeared model.
The reinforcement in the discrete model (Figure 2.8a) uses bar or beam elements
that are connected to concrete mesh nodes. Therefore, the concrete and the reinforcement
mesh share the same nodes and concrete occupies the same regions occupied by the
reinforcement. A drawback to this model is that the concrete mesh is restricted by the
location of the reinforcement and the volume of the mildsteel reinforcement is not
deducted from the concrete volume.
(a) (b)
(c)
Figure 2.8 – Models for Reinforcement in Reinforced Concrete (Tavarez 2001): (a)
discrete; (b) embedded; and (c) smeared
The embedded model (Figure 2.8b) overcomes the concrete mesh restriction(s)
because the stiffness of the reinforcing steel is evaluated separately from the concrete
elements. The model is built in a way that keeps reinforcing steel displacements
compatible with the surrounding concrete elements. When reinforcement is complex,
this model is very advantageous. However, this model increases the number of nodes and
degrees of freedom in the model, therefore, increasing the run time and computational
cost.
The smeared model (Figure 2.8c) assumes that reinforcement is uniformly spread
throughout the concrete elements in a defined region of the FE mesh. This approach is
used for largescale models where the reinforcement does not significantly contribute to
the overall response of the structure.
Fanning (2001) modeled the response of the reinforcement using the discrete
model and the smeared model for reinforced concrete beams. It was found that the best
modeling strategy was to use the discrete model when modeling reinforcement.
2.5 Direction for Present Research
The literature review suggested that use of a finite element package to model reinforced
and prestressed concrete beams was indeed feasible. It was decided to use ANSYS (SAS
2003) as the FE modeling package. A reinforced concrete beam with reinforcing steel
modeled discretely will be developed with results compared to the experimental work of
Buckhouse (1997). The loaddeflection response of the experimental beam will be
compared to analytical predictions to calibrate the FE model for further use. A second
analysis of a prestressed concrete beam will also be studied. The different stages of the
response of a prestressed concrete beam are computed using FEA and compared to
results generated using hand computations.
CHAPTER 3
CALIBRATION MODEL
3.0 Introduction
This chapter discusses the calibration of the finite element model using experimental
loaddeformation behavior of a concrete beam provided in Buckhouse (1997). The use of
ANSYS (SAS 2003) to create the finite element model is also discussed. All the
necessary steps to create the calibrated model are explained in detail and the steps taken
to generate the analytical loaddeformation response of the member are discussed.
3.1 Experimental Beam
Buckhouse (1997) studied a method to reinforce a concrete beam for flexure using
external structural steel channels. The study included experimental testing of control
beams that can be used for calibration of finite element models. The width and height of
the beams tested were 10 in. and 18 in., respectively. As shown in Figure 3.1, the length
Figure 3.1 – Loading and Supports for the Beam (Buckhouse 1997)
of the beam was 15 ft.6 in. with supports located 3 in. from each end of the beam
allowing a simply supported span of 15 ft. The mild steel flexural reinforcements used
were 3#5 bars and shear reinforcements included #3 Ustirrups. Cover for the rebar was
set to 2 in. in all directions. The layout of the reinforcement is detailed in Figure 3.2.
Figure 3.2 – Typical Detail for Control Beam Reinforcement (Buckhouse 1997)
The steel yield stress, 28day compressive stress of concrete, and area of steel
reinforcement are included in Table 3.1.
Table 3.1 – Properties for Steel and Concrete (Buckhouse 1997)
Area of Steel (in.
2
) 0.93
Yield Stress of Steel, f
y
(psi)
60,000
28Day Compressive
Strength of Concrete, f
c
'
(psi)
4,770
Two 50kip capacity load cells were placed at third points, or 5 ft. from each
support on steel bearing plates (Figure 3.1). Data acquisition equipment was used to
record applied loading, beam deflection at the midspan, and strain in the internal flexural
reinforcement. The beam was loaded to flexural failure (Figure 3.3).
Figure 3.3 – Failure in Flexure (Buckhouse 1997)
Vertical cracks first formed in the constant moment region, extended upward, and then
out towards the constant shear region with eventual crushing of the concrete in the
constant moment region as shown in Figure 3.3. Test data for the beam is summarized in
Table 3.2.
The theoretical ultimate load for the beam was calculated to be 14,600 lbs
(Buckhouse 1997). Table 3.2 shows the experimental ultimate load determined was
16,310 lbs. The ultimate loading corresponded to the nominal flexural capacity of the
crosssection being reached. A plot of load versus deflection for control beam C1
(Buckhouse 1997) is shown in Figure 3.4.
Table 3.2 – Test data for control beam C1 (Buckhouse 1997)
Avg. Load at 1st
Crack (lbs.)
4,500
Avg. Failure Load, P
(lbs.)
16,310
Avg. Centerline
Deflection at Failure
(in.)
3.65
Mode of Failure
compression failure of
concrete
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
Avg. Centerline Deflection (in.)
A
v
g
.
L
o
a
d
,
P
(
l
b
s
.
)
Figure 3.4 – Load vs. Deflection Curve for Beam C1 (Buckhouse 1997)
C1 theoretical
ultimate load
(14,600 lbs.)
A
B
C
Point A  First Cracking
Point B  Steel Yielding
Point C  Failure
Nonlinear Region
Linear Region
The plot shows the linear behavior before first cracking (point A). A second slope
corresponding to the cracked section is followed until point B where the flexural
reinforcement yields. The cracked moment of inertia with yielding internal
reinforcement then defines the stiffness until flexural failure at point C.
3.2 ANSYS Finite Element Model
The FEA calibration study included modeling a concrete beam with the dimensions and
properties corresponding to beam C1 tested by Buckhouse (1997). Due to the symmetry
in crosssection of the concrete beam and loading, symmetry was utilized in the FEA,
only one quarter of the beam was modeled.
To create the finite element model in ANSYS (SAS 2003) there are multiple tasks
that have to be completed for the model to run properly. Models can be created using
command prompt line input or the Graphical User Interface (GUI). For this model, the
GUI was utilized to create the model. This section describes the different tasks and
entries into used to create the FE calibration model.
3.2.1 Element Types
The element types for this model are shown in Table 3.3. The Solid65 element was used
to model the concrete. This element has eight nodes with three degrees of freedom at
each node – translations in the nodal x, y, and z directions. This element is capable of
plastic deformation, cracking in three orthogonal directions, and crushing. A schematic
of the element is shown in Figure 3.5.
Table 3.3 – Element Types For Working Model
Material Type ANSYS Element
Concrete Solid65
Steel Plates and
Supports
Solid45
Steel
Reinforcement
Link8
Figure 3.5 – Solid 65 Element (SAS 2003)
A Solid45 element was used for steel plates at the supports for the beam. This
element has eight nodes with three degrees of freedom at each node – translations in the
nodal x, y, and z directions. The geometry and node locations for this element is shown
in Figure 3.6. The descriptions for each element type are laid out in the ANSYS element
library (SAS 2003).
Figure 3.6 – Solid 45 Element (SAS 2003)
A Link8 element was used to model steel reinforcement. This element is a 3D
spar element and it has two nodes with three degrees of freedom – translations in the
nodal x, y, and z directions. This element is also capable of plastic deformation. This
element is shown in Figure 3.7.
3.2.2 Real Constants
The real constants for this model are shown in Table 3.4. Note that individual elements
contain different real constants. No real constant set exists for the Solid45 element.
Figure 3.7 – Link 8 Element (SAS 2003)
Table 3.4 – Real Constants For Calibration Model
Real Constant Set Element Type Constants
Real
Constants for
Rebar 1
Real
Constants for
Rebar 2
Real
Constants for
Rebar 3
Material Number 0 0 0
Volume Ratio 0 0 0
Orientation Angle 0 0 0
1 Solid 65
Orientation Angle 0 0 0
Crosssectional
Area (in.
2
)
0.31
2 Link8
Initial Strain
(in./in.)
0
Crosssectional
Area (in.
2
)
0.155
3 Link8
Initial Strain
(in./in.)
0
Crosssectional
Area (in.
2
)
0.11
4 Link8
Initial Strain
(in./in.)
0
Crosssectional
Area (in.
2
)
0.055
5 Link8
Initial Strain
(in./in.)
0
Real Constant Set 1 is used for the Solid65 element. It requires real constants for
rebar assuming a smeared model. Values can be entered for Material Number, Volume
Ratio, and Orientation Angles. The material number refers to the type of material for the
reinforcement. The volume ratio refers to the ratio of steel to concrete in the element.
The orientation angles refer to the orientation of the reinforcement in the smeared model
(Figure 2.8c). ANSYS (SAS 2003) allows the user to enter three rebar materials in the
concrete. Each material corresponds to x, y, and z directions in the element (Figure 3.5).
The reinforcement has uniaxial stiffness and the directional orientation is defined by the
user. In the present study the beam is modeled using discrete reinforcement. Therefore,
a value of zero was entered for all real constants which turned the smeared reinforcement
capability of the Solid65 element off.
Real Constant Sets 2, 3, 4, and 5 are defined for the Link8 element. Values for
crosssectional area and initial strain were entered. Crosssectional areas in sets 2 and 3
refer to the reinforcement of 3#5 bars. Due to symmetry, set 3 is half of set 2 because
onehalf the center bar in the beam is cut off. Crosssectional areas in sets 4 and 5 refer
to the #3 stirrups. Once again set 5 is half of set 4 because half of the stirrup at the mid
span of the beam is cut off resulting from symmetry. A value of zero was entered for the
initial strain because there is no initial stress in the reinforcement.
3.2.3 Material Properties
Parameters needed to define the material models can be found in Table 3.5. As seen in
Table 3.5, there are multiple parts of the material model for each element.
Table 3.5 – Material Models For the Calibration Model
Material Model
Number
Element
Type
Material Properties
Linear Isotropic
EX 3,949,076 psi
PRXY 0.3
Multilinear Isotropic
Strain Stress
Point 1 0.00036 1421.7
Point 2 0.0006 2233
Point 3 0.0013 3991
Point 4 0.0019 4656
Point 5 0.00243 4800
Concrete
ShrCfOp 0.3
ShrCfCl 1
UnTensSt 520
UnCompSt 1
BiCompSt 0
HydroPrs 0
BiCompSt 0
UnTensSt 0
TenCrFac 0
1 Solid65
Linear Isotropic
EX 29,000,000 psi
PRXY 0.3
2 Solid45
Linear Isotropic
EX 29,000,000 psi
PRXY 0.3
Bilinear Isotropic
Yield Stss 60,000 psi
Tang Mod 2,900 psi
3 Link8
Material Model Number 1 refers to the Solid65 element. The Solid65 element
requires linear isotropic and multilinear isotropic material properties to properly model
concrete. The multilinear isotropic material uses the von Mises failure criterion along
with the Willam and Warnke (1974) model to define the failure of the concrete. EX is
the modulus of elasticity of the concrete (
c
E ), and PRXY is the Poisson’s ratio (ν ). The
modulus was based on the equation,
'
57000
c c
E f = (3.1)
with a value of
'
c
f equal to 4,800 psi. Poisson’s ratio was assumed to be 0.3. The
compressive uniaxial stressstrain relationship for the concrete model was obtained using
the following equations to compute the multilinear isotropic stressstrain curve for the
concrete (MacGregor 1992)
2
0
1
c
E
f
ε
ε
ε
=
 
+

\ .
(3.2)
'
0
2
c
c
f
E
ε = (3.3)
c
f
E
ε
= (3.4)
where:
f = stress at any strain ε , psi
ε = strain at stress f
0
ε = strain at the ultimate compressive strength
'
c
f
The multilinear isotropic stressstrain implemented requires the first point of the curve to
be defined by the user. It must satisfy Hooke’s Law;
E
σ
ε
= (3.5)
The multilinear curve is used to help with convergence of the nonlinear solution
algorithm.
0
1000
2000
3000
4000
5000
6000
0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035
Strain (in./in.)
S
t
r
e
s
s
(
p
s
i
)
Figure 3.8 – Uniaxial StressStrain Curve
Figure 3.8 shows the stressstrain relationship used for this study and is based on
work done by Kachlakev, et al. (2001). Point 1, defined as
'
0.30
c
f , is calculated in the
linear range (Equation 3.4). Points 2, 3, and 4 are calculated from Equation 3.2 with
0
ε
obtained from Equation 3.3. Strains were selected and the stress was calculated for each
0
ε
'
0.30
c
f
Strain at Ultimate Strength
c
E
Ultimate Compressive Strength
'
c
f
5
4
3
2
1
strain. Point 5 is defined at
'
c
f and
0
.
0.003
.
in
in
ε = indicating traditional crushing strain
for unconfined concrete.
Implementation of the Willam and Warnke (1974) material model in ANSYS
requires that different constants be defined. These 9 constants are: (SAS 2003)
1. Shear transfer coefficients for an open crack;
2. Shear transfer coefficients for a closed crack;
3. Uniaxial tensile cracking stress;
4. Uniaxial crushing stress (positive);
5. Biaxial crushing stress (positive);
6. Ambient hydrostatic stress state for use with constants 7 and 8;
7. Biaxial crushing stress (positive) under the ambient hydrostatic stress state
(constant 6);
8. Uniaxial crushing stress (positive) under the ambient hydrostatic stress state
(constant 6);
9. Stiffness multiplier for cracked tensile condition.
Typical shear transfer coefficients range from 0.0 to 1.0, with 0.0 representing a
smooth crack (complete loss of shear transfer) and 1.0 representing a rough crack (no loss
of shear transfer). The shear transfer coefficients for open and closed cracks were
determined using the work of Kachlakev, et al. (2001) as a basis. Convergence problems
occurred when the shear transfer coefficient for the open crack dropped below 0.2. No
deviation of the response occurs with the change of the coefficient. Therefore, the
coefficient for the open crack was set to 0.3 (Table 3.4). The uniaxial cracking stress was
based upon the modulus of rupture. This value is determined using,
'
7.5
r c
f f = (3.6)
The uniaxial crushing stress in this model was based on the uniaxial unconfined
compressive strength (
'
c
f ) and is denoted as
t
f . It was entered as 1 to turn off the
crushing capability of the concrete element as suggested by past researchers (Kachlakev,
et al. 2001). Convergence problems have been repeated when the crushing capability
was turned on.
The biaxial crushing stress refers to the ultimate biaxial compressive strength
(
'
cb
f ). The ambient hydrostatic stress state is denoted as
h
σ . This stress state is defined
as:
1
( )
3
h xp yp zp
σ σ σ σ = + + (3.7)
where
xp
σ ,
yp
σ , and
zp
σ are the principal stresses in the principal directions. The biaxial
crushing stress under the ambient hydrostatic stress state refers to the ultimate
compressive strength for a state of biaxial compression superimposed on the hydrostatic
stress state (
1
f ). The uniaxial crushing stress under the ambient hydrostatic stress state
refers to the ultimate compressive strength for a state of uniaxial compression
superimposed on the hydrostatic stress state (
2
f ). The failure surface can be defined
with a minimum of two constants,
t
f and
'
c
f . The remainder of the variables in the
concrete model are left to default based on these equations: (SAS 2003)
' '
1.2
cb c
f f = (3.8)
'
1
1.45
c
f f = (3.9)
'
2
1.725
c
f f = (3.10)
These stress states are only valid for stress states satisfying the condition
'
3
h c
f σ ≤ (3.11)
Material Model Number 2 refers to the Solid45 element. The Solid45 element is
being used for the steel plates at loading points and supports on the beam. Therefore, this
element is modeled as a linear isotropic element with a modulus of elasticity for the steel
(
s
E ), and poisson’s ratio (0.3).
Material Model Number 3 refers to the Link8 element. The Link8 element is
being used for all the steel reinforcement in the beam and it is assumed to be bilinear
isotropic. Bilinear isotropic material is also based on the von Mises failure criteria. The
bilinear model requires the yield stress (
y
f ), as well as the hardening modulus of the steel
to be defined. The yield stress was defined as 60,000 psi, and the hardening modulus was
2900 psi.
Note that the density for the concrete was not added in the material model. For
the control beam in Buckhouse (1997), the LVDT’s used to measure deflection at mid
span were put on the beam after it was set in the test fixture. Deflections were taken
relative to a zero deflection point after the selfweight was introduced. Therefore, the
selfweight was not introduced in this calibration model.
3.2.4 Modeling
The beam, plates, and supports were modeled as volumes. Since a quarter of the beam is
being modeled, the model is 93 in. long, with a crosssection of 5 in. x 18 in. The
dimensions for the concrete volume are shown in Table 3.6. The zero values for the Z
coordinates coincide with the center of the crosssection for the concrete beam.
Table 3.6 – Dimensions for Concrete, Steel Plate, and Steel Support Volumes
ANSYS Concrete (in.) Steel Plate (in.) Steel Support (in.)
X1,X2 Xcoordinates 0 93 60 66 1.5 4.5
Y1,Y2 Ycoordinates 0 18 18 19 0 1
Z1,Z2 Zcoordinates 0 5 0 5 0 5
The 93 in. dimension for the Xcoordinates is the midspan of the beam. Due to
symmetry, only one loading plate and one support plate are needed. The support is a 3 in.
x 5 in. x 1 in. steel plate, while the plate at the load point is 6 in. x 5 in. x 1 in. The
dimensions for the plate and support are shown in Table 3.6. The combined volumes of
the plate, support, and beam are shown in Figure 3.9. The FE mesh for the beam model
is shown in Figure 3.10.
Figure 3.9 – Volumes Created in ANSYS
Steel Support
Steel Loading Plate
Concrete Beam
Figure 3.10 – Mesh of the Concrete, Steel Plate, and Steel Support
Link8 elements were used to create the flexural and shear reinforcement.
Reinforcement exists at a plane of symmetry and in the beam. The area of steel at the
plane of symmetry is one half the normal area for a #5 bar because one half of the bar is
cut off. Shear stirrups are modeled throughout the beam. Only half of the stirrup is
modeled because of the symmetry of the beam. Figure 3.11 illustrates that the rebar
shares the same nodes at the points that it intersects the shear stirrups. The element type
number, material number, and real constant set number for the calibration model were set
for each mesh as shown in Table 3.7.
Steel Plate Element
Width 1.25 in.
Concrete Element
Length 1.5 in.
Concrete Element
Width 1.25 in.
Concrete Element
Height 1.2 in.
Steel Plate Element
Length 1.5 in.
Steel Support Element
Width 1.25 in.
Steel Support
Element
Length 1.5 in.
Figure 3.11 – Reinforcement Configuration
Table 3.7 – Mesh Attributes for the Model
Model Parts
Element
Type
Material
Number
Real Constant
Set
Concrete Beam 1 1 1
Steel Plate 2 3 N/A
Steel Support 2 3 N/A
Rebar at Center of CrossSection 3 2 3
Rebar 2.5 in. of CrossSection 3 2 2
Stirrup at Center of Beam 3 2 5
Other Stirrups 3 2 4
3.2.5 Meshing
To obtain good results from the Solid65 element, the use of a rectangular mesh is
recommended. Therefore, the mesh was set up such that square or rectangular elements
#3 Shear Stirrups
#5 Bar Reinforcement located 2.5 in.
from the end of the CrossSection
Shared nodes of
Stirrups and Rebar
#5 Bar Reinforcement at
Plane of Symmetry
Stirrup at Plane of
Symmetry
were created (Figure 3.10). The volume sweep command was used to mesh the steel
plate and support. This properly sets the width and length of elements in the plates to be
consistent with the elements and nodes in the concrete portions of the model.
The overall mesh of the concrete, plate, and support volumes is shown in Figure
3.10. The necessary element divisions are noted. The meshing of the reinforcement is a
special case compared to the volumes. No mesh of the reinforcement is needed because
individual elements were created in the modeling through the nodes created by the mesh
of the concrete volume. However, the necessary mesh attributes as described above need
to be set before each section of the reinforcement is created.
3.2.6 Numbering Controls
The command merge items merges separate entities that have the same location. These
items will then be merged into single entities. Caution must be taken when merging
entities in a model that has already been meshed because the order in which merging
occurs is significant. Merging keypoints before nodes can result in some of the nodes
becoming “orphaned”; that is, the nodes lose their association with the solid model. The
orphaned nodes can cause certain operations (such as boundary condition transfers,
surface load transfers, and so on) to fail. Care must be taken to always merge in the order
that the entities appear. All precautions were taken to ensure that everything was merged
in the proper order. Also, the lowest number was retained during merging.
3.2.7 Loads and Boundary Conditions
Displacement boundary conditions are needed to constrain the model to get a unique
solution. To ensure that the model acts the same way as the experimental beam,
boundary conditions need to be applied at points of symmetry, and where the supports
and loadings exist.
The symmetry boundary conditions were set first. The model being used is
symmetric about two planes. The boundary conditions for both planes of symmetry are
shown in Figure 3.12.
Figure 3.12 – Boundary Conditions for Planes of Symmetry
Constraint in the zdirection
Constraint in the xdirection
Nodes defining a vertical plane through the beam crosssection centroid defines a plane
of symmetry. To model the symmetry, nodes on this plane must be constrained in the
perpendicular direction. These nodes, therefore, have a degree of freedom constraint UX
= 0. Second, all nodes selected at Z = 0 define another plane of symmetry. These nodes
were given the constraint UZ = 0.
The support was modeled in such a way that a roller was created. A single line of
nodes on the plate were given constraint in the UY, and UZ directions, applied as
constant values of 0. By doing this, the beam will be allowed to rotate at the support. The
support condition is shown in Figure 3.13.
Figure 3.13 – Boundary Condition for Support
Support roller condition
to allow rotation
The force, P, applied at the steel plate is applied across the entire centerline of the
plate. The force applied at each node on the plate is one tenth of the actual force applied.
Figure 3.14 illustrates the plate and applied loading.
Figure 3.14 – Boundary Conditions at the Loading Plate
3.2.8 Analysis Type
The finite element model for this analysis is a simple beam under transverse loading. For
the purposes of this model, the Static analysis type is utilized.
The Restart command is utilized to restart an analysis after the initial run or load
step has been completed. The use of the restart option will be detailed in the analysis
portion of the discussion.
Loading Applied
on the Plate
Boundary Conditions at Plate
The Sol’n Controls command dictates the use of a linear or nonlinear solution for
the finite element model. Typical commands utilized in a nonlinear static analysis are
shown in Table 3.8.
Table 3.8 – Commands Used to Control Nonlinear Analysis
Analysis Options Small Displacement
Calculate Prestress Effects No
Time at End of Loadstep 5120
Automatic Time Stepping On
Number of Substeps 1
Max no. of Substeps 2
Min no. of Substeps 1
Write Items to Results File All Solution Items
Frequency Write Every Substep
In the particular case considered in this thesis the analysis is small displacement and
static. The time at the end of the load step refers to the ending load per load step. Table
3.8 shows the first load step taken (e.g. up to first cracking). The sub steps are set to
indicate load increments used for this analysis. The commands used to control the solver
and output are shown in Table 3.9.
Table 3.9 – Commands Used to Control Output
Equation Solvers Sparse Direct
Number of Restart Files 1
Frequency Write Every Substep
All these values are set to ANSYS (SAS 2003) defaults. The commands used for the
nonlinear algorithm and convergence criteria are shown in Table 3.10. All values for the
nonlinear algorithm are set to defaults.
Table 3.10 – Nonlinear Algorithm and Convergence Criteria Parameters
Line Search Off
DOF solution predictor Prog Chosen
Maximum number of iteration 100
Cutback Control Cutback according to predicted number of iter.
Equiv. Plastic Strain 0.15
Explicit Creep ratio 0.1
Implicit Creep ratio 0
Incremental displacement 10000000
Points per cycle 13
Set Convergence Criteria
Label F U
Ref. Value Calculated calculated
Tolerance 0.005 0.05
Norm L2 L2
Min. Ref. not applicable not applicable
The values for the convergence criteria are set to defaults except for the tolerances. The
tolerances for force and displacement are set as 5 times the default values. Table 3.11
shows the commands used for the advanced nonlinear settings. The program behavior
upon nonconvergence for this analysis was set such that the program will terminate but
not exit. The rest of the commands were set to defaults.
Table 3.11 – Advanced Nonlinear Control Settings Used
Program behavior upon nonconvergence Terminate but do not exit
Nodal DOF sol'n 0
Cumulative iter 0
Elapsed time 0
CPU time 0
3.2.9 Analysis Process for the Finite Element Model
The FE analysis of the model was set up to examine three different behaviors: initial
cracking of the beam, yielding of the steel reinforcement, and the strength limit state of
the beam. The NewtonRaphson method of analysis was used to compute the nonlinear
response.
The application of the loads up to failure was done incrementally as required by
the NewtonRaphson procedure. After each load increment was applied, the restart
option was used to go to the next step after convergence. A listing of the load steps, sub
steps, and loads applied per restart file are shown in Table 3.12.
Table 3.12 – Load Increment for Analysis of Finite Element Model
Beginning
Time
Time at End
of Loadstep
Load Step Sub Step
Load
Increment
(lbs.)
0 5210 1 1 5210
5210 5220 2 10 10
5220 5300 3 16 5
5300 5400 4 20 5
5400 10000 5 92 50
10000 13000 6 30 100
13000 14000 7 10 100
14000 14500 8 50 10
14500 14700 9 20 10
14700 14800 10 20 5
14800 14900 11 100 1
14900 15000 12 10 10
15000 15100 13 10 10
15100 15200 14 50 2
15200 15300 15 20 5
15300 15600 16 150 2
15600 15900 17 150 2
15900 16200 18 150 2
16200 16300 19 50 2
16300 16382 20 41 2
The time at the end of each load step corresponds to the loading applied. For the first
load step the time at the end of the load step is 5210 referring to a load of, P, of 5210 lbs
applied at the steel plate.
The two convergence criteria used for the analysis were Force and Displacement.
These criteria were left at the default values up to 5210 lbs. However, when the beam
began cracking, convergence for the nonlinear analysis was impossible with the default
values. The displacements converged, but the forces did not. Therefore, the convergence
criteria for force was dropped and the reference value for the Displacement criteria was
changed to 5. This value is then multiplied by the tolerance value of 0.05 to produce a
criterion of 0.25 during the nonlinear solution for convergence. A small criterion must be
used to capture correct response. This criteria was used for the remainder of the analysis.
As shown in Table 3.12, the steps taken to the initial cracking of the beam can be
decresed to one load increment to model/capture initial cracking. Once initial cracking of
the beam has been passed (5220 lbs), the load increments increased slightly until
subsequent cracking of the beam (14,000 lbs) as seen in Table 3.12. Once the yielding of
the reinforcing steel is reached, the load increments must be decreased again. Yielding of
the steel occurs at load step 13,400; therefore, the load increment sizes begin decreasing
further because displacements are increasing more rapidly. Eventually, the load
increment size is decreased to 2 lb. to capture the failure of the beam. Failure of the
beam occurs when convergence fails, with this very small load increment. The load
deformation trace produced by the analysis confirmed the failure load.
3.3 Results
The goal of the comparison of the FE model and the beam from Buckhouse (1997) is to
ensure that the elements, material properties, real constants and convergence criteria are
adequate to model the response of the member. Figure 3.4 shows the different
components that were analyzed for comparison: the linear region, initial cracking, the
nonlinear region, the yielding of steel, and failure.
3.3.1 Behavior at First Cracking
The analysis of the linear region can be based on the design for flexure given in
MacGregor (1992) for a reinforced concrete beam. Comparisons were made in this
region to ensure deflections and stresses were consistent with the FE model and the beam
before cracking occurred. Once cracking occurs, deflections and stresses become more
difficult to predict. The stresses in the concrete and steel immediately preceding initial
cracking were analyzed. The load at step 5210 was analyzed and it coincides with a load
of 5210 lbs. applied on the steel plate.
Calculations to obtain the concrete stress, steel stress and deflection of the beam
at a load of 5210 lbs. can be seen in Appendix A. A comparison of values obtained from
the FE model and Appendix A can be seen in Table 3.13. The maximums exist in the
constantmoment region of the beam during load application. This is where we expect
the maximums to occur. The results in Table 3.13 indicate that the FE analysis of the
beam prior to cracking is acceptable.
Table 3.13 – Deflection and Stress Comparisons At First Cracking
Model
Extreme Tension
Fiber Stress (psi)
Reinforcing
Steel Stress
(psi)
Centerline
Deflection
(in.)
Load at First
Cracking
(lbs.)
Hand
Calculations
530 3024 0.0529 5118
ANSYS 536 2840 0.0534 5216
3.3.2 Behavior at Initial Cracking
The cracking pattern(s) in the beam can be obtained using the Crack/Crushing plot option
in ANSYS (SAS 2003). Vector Mode plots must be turned on to view the cracking in the
model.
The initial cracking of the beam in the FE model corresponds to a load of 5216 lbs
that creates stress just beyond the modulus of rupture of the concrete (520 psi) as shown
in Table 3.13. This compares well with the load of 5118 lbs calculated in Appendix A.
The stress increases up to 537 psi at the centerline when the first crack occurs. The first
crack can be seen in Figure 3.15. This first crack occurs in the constant moment region,
and is a flexural crack. Buckhouse (1997) reported first cracking at a load, P, of 4500 lbs
using visual detection.
3.3.3 Behavior Beyond First Cracking
In the nonlinear region of the response, subsequent cracking occurs as more load is
applied to the beam. Cracking increases in the constant moment region, and the beam
begins cracking out towards the supports at a load of 8,000 lbs.
Figure 3.15 – 1
st
Crack of the Concrete Model
Significant flexural cracking occurs in the beam at 12,000 lbs. Also, diagonal tension
cracks are beginning to form in the model. This cracking can be seen in Figure 3.16.
3.3.4 Behavior of Reinforcement Yielding and Beyond
Yielding of steel reinforcement occurs when a force of 13,400 lbs. is applied. At this
point in the response, the displacements of the beam begin to increase at a higher rate as
more load is applied (Figure 3.4). The cracked moment of inertia, yielding steel, and
nonlinear concrete material, now defines the flexural rigidity of the member. The ability
of the beam to distribute load throughout the crosssection has diminished greatly.
Therefore, greater deflections occur at the beam centerline.
1
st
Crack in
Concrete Beam
Figure 3.16 – Cracking at 8,000 and 12,000 lbs.
Figure 3.17 shows successive cracking of the concrete beam after yielding of the
steel occurs. At 15,000 lbs., the beam has increasing flexural cracks, and diagonal
tension cracks. Also, more cracks have now formed in the constant moment region. At
16,000 lbs., cracking has reached the top of the beam, and failure is soon to follow.
Flexural Cracks
Diagonal Tension Cracks
Figure 3.17 – Increased Cracking After Yielding of Reinforcement
3.3.5 Strength Limit State
At load 16,382 lbs., the beam no longer can support additional load as indicated by an
insurmountable convergence failure. Severe cracking throughout the entire constant
moment region occurs (see Figure 3.18). The deflections at the analytical failure load of
the control beam were compared with the finite element model as shown in Table 3.14.
Multiple cracks
occurring
Increasing Diagonal
Tension Cracks
Figure 3.18 – Failure of the Concrete Beam
Table 3.14 – Deflections of Control Beam (Buckhouse 1997) vs. Finite Element Model
At Ultimate Load
Beam Load (lb.)
Centerline
Deflection
(in.)
C1 16310 3.65
ANSYS 16310 3.586
The deflection of the finite element model was within 2% of the control beam at the same
load at which the control beam failed.
Excessive cracking and of the beam in
the constant moment region
Diagonal Tension
Cracking
3.3.6 LoadDeformation Response
The full nonlinear loaddeformation response can be seen in Figure 3.19. This response
was calibrated by setting the tolerances so that the loaddeformation curve fits to the
curve from Buckhouse (1997). The response calculated using FEA is plotted upon the
experimental response from Buckhouse (1997). The entire loaddeformation response of
the model produced compares well with the response from Buckhouse (1997). This gave
confidence in the use of ANSYS (SAS 2003) and the model developed. The approach
was then utilized to analyze a prestressed concrete beam.
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
Avg. Centerline Deflection (in.)
A
v
g
.
L
o
a
d
,
P
(
l
b
s
.
)
Figure 3.19 – Load vs. Deflection Curve Comparison of ANSYS and Buckhouse (1997)
Buckhouse (1997)
FEA
C1 theoretical
ultimate load
(14,600 lbs.)
CHAPTER 4
PRESTRESSED CONCRETE BEAM MODEL
4.0 Introduction
This chapter discusses the finite element modeling of a prestressed concrete beam. The
prestressed beam has the same dimensions as the experimental beam modeled by
Buckhouse (1997). In this case, the reinforcement in the beam is different. The shear
stirrups are included as discussed in the calibration model. However, no mildsteel
flexural reinforcement was used. The beam was prestressed using ½ in. diameter 270 ksi
7wire strands as opposed to the #5 60 ksi mildsteel reinforcement. All the necessary
steps taken to create the model and the analysis used for the prestressed beam are
explained in detail.
4.1 Finite Element Model
The finite element model that was used for analysis of the prestressed concrete beam is
very similar to the calibration model. Many of the steps taken to model the prestressed
concrete beam were the same as the calibration model and can be found in chapter 3.
The rest of this chapter will discuss the steps taken that were different from those
in the calibration model. These steps include definition of real constants, material
properties, and the parameters for the nonlinear analysis.
4.1.1 Real Constants
The real constants for the concrete element were left untouched because the modeling of
the concrete has not changed. Also, the constants for the Link8 element designated to the
stirrups used in the model were left unchanged. However, the real constants for the
Link8 element for the flexural reinforcement has changed. Since the beam is now using
prestressing steel, the crosssectional area of the steel has changed, and an initial strain
due to prestressing is now added to the element. The change in area and strain is shown
in Table 4.1.
Table 4.1 – Real Constants for Prestressed Beam
Real Constant Set Element Type Constants
Crosssectional
Area (in.
2
)
0.153
2 Link8
Initial Strain
(in./in.)
0.001903
Crosssectional
Area (in.
2
)
0.0765
3 Link8
Initial Strain
(in./in.)
0.001903
The crosssectional area for real constant set 2 is the area of an equivalent ½ in. diameter
7wire strand. The crosssectional area for real constant set 3 is half of real constant set
2, because it is at a point of symmetry on the model.
The initial strains for each real constant set were determined from the effective
prestress (
pe
f ) and the modulus of elasticity (
ps
E ). An example of this can be seen in
Appendix B. This prestress level is too low and not suitable for practical use. It was
utilized to prevent any convergence problems occurring from bursting and significant
cracking near the support as the prestrain is applied.
4.1.2 Material Properties
The material properties for the shear reinforcement steel, steel plates at loading points,
and steel support plates are the same. Density was added to the concrete material
property so the selfweight of the concrete beam could be taken into account.
The material properties for the prestressing steel have been changed from bilinear
isotropic to multilinear isotropic following the von Mises failure criteria. The
prestressing steel was modeled using a multilinear stressstrain curve developed using the
following equations,
0.008:
ps
ε ≤ 28, 000
ps ps
f ε = ( ) ksi (4.1)
>0.008:
ps
ε
0.075
268 <0.98 ( )
0.0065
ps pu
ps
f f ksi
ε
= −
−
(4.2)
The values entered into ANSYS (SAS 2003) for the stressstrain curve are given in Table
4.2 and Figure 4.1 shows the stressstrain behavior of the prestressing steel.
4.1.3 Solution
Most of the parameters used to control the solution are the same as those used for the
calibration model. The only changes made were as follows. First, no point loads are
applied in the first load step, due to the fact that the initial prestrain is applied. Second,
the time at the end of the prestrain application load step is 1. Third, the selfweight was
added in a load step. The addition of the selfweight is done by applying gravitational
acceleration of 386.4 in/s
2
in the global ydirection. A mass density was entered in as a
function of the gravitational acceleration (Density/g).
Table 4.2 – Values for Multilinear Isotropic StressStrain Curve
Strain Stress Strain Stress Strain Stress Strain Stress
0 0 0.0101 247.1667 0.0123 255.069 0.0145 258.625
0.008 224 0.0103 248.2632 0.0125 255.5 0.0147 258.8537
0.0083 226.3333 0.0105 249.25 0.0127 255.9032 0.0149 259.0714
0.0085 230.5 0.0107 250.1429 0.0129 256.2813 0.0151 259.2791
0.0087 233.9091 0.0109 250.9545 0.0131 256.6364 0.0171 260.9245
0.0089 236.75 0.0111 251.6957 0.0133 256.9706 0.0189 261.9516
0.0091 239.1538 0.0113 252.375 0.0135 257.2857 0.0215 263
0.0093 241.2143 0.0115 253 0.0137 257.5833 0.0259 264.134
0.0095 243 0.0117 253.5769 0.0139 257.8649 0.0301 264.822
0.0097 244.5625 0.0119 254.1111 0.0141 258.1316
0.0099 245.9412 0.0121 254.6071 0.0143 258.3846
0
50
100
150
200
250
300
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035
Strain (in./in.)
S
t
r
e
s
s
(
k
s
i
)
Figure 4.1 – StressStrain Curve for 270 ksi strand
4.1.4 Analysis Process for the Finite Element Model
Analysis for the prestressed concrete beam was very similar to the calibration model.
However, different load step and restart points were used. The first load step taken was
to produce the camber in the concrete beam due to the prestress. The second load step
was the addition of the selfweight. From that point on, the concrete was modeled to the
point of cracking, yielding of the prestress steel ( 0.85
pu
f ), and eventual failure. A listing
of the load steps, sub steps, and loads applied per restart file are shown in Table 4.3.
Table 4.3 – Load Increments for the Analysis of the Prestressed Beam Model
Beginning
Time
Time at
End of
Loadstep
Load Step Sub Step
Load
Increment
(lbs.)
0 1 1 1 Prestress
1 2 2 1 386.4
2 7840 3 1 7840
7840 7900 4 30 2
7900 8500 5 12 50
8500 10000 6 15 100
10000 20000 7 100 100
20000 25000 8 100 50
25000 27000 9 200 10
27000 28000 10 100 10
28000 28823 11 412 2
When the analysis reaches the point of initial cracking, the force convergence
criteria is dropped, and the reference value of the displacement criteria is 5. From this
point, the load increments are decreased to 100 lbs. to capture the initial cracking of the
beam. When yielding of the steel occurs, the load increments are decreased to 10 lb.
Finally, load increments are decreased to 2 lb. until unresolvable convergence failure of
the nonlinear algorithm occurs.
4.2 Results
The analysis yielded results for seven distinct points on the loaddeflection application of
the prestressed concrete beam and the full nonlinear loaddeformation response. As seen
in Figure 4.2, these distinct points are: effective prestress, addition of selfweight, zero
deflection, decompression, initial cracking, steel yielding ( 0.85
pu
f ), and failure.
4.2.1 Application of Effective Prestress
Calculation of the effective prestress for the beam can be found in Appendix B. The
deflections computed using hand calculations and FE analysis due to the prestress force is
shown in Table 4.4.
Table 4.4 – Analytical Results
ANSYS
Hand
Calculations
Centerline Deflection at
Effective Prestress (in.)
0.0323 0.0341
Top Fiber Stress at
Effective Prestress (psi)
156.05 163.04
Deflection at Application
of SelfWeight (in.)
0.0211 0.0229
Top Fiber Stress at Self
Weight (psi)
36.97 45.86
Zero Deflection Load
(lbs.)
2055 2126
Decompression Load
(lbs.)
2817 2858
Initial Cracking Load
(lbs.)
7850 7538
Prestress at Failure Load
(psi)
264,822 254,520
Failure Load (lbs.) 28,823 27,587
5000
0
5000
10000
15000
20000
25000
30000
35000
0.50 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00
Centerline Deflection (in.)
L
o
a
d
,
P
(
l
b
s
.
)
0
1000
2000
3000
4000
5000
6000
7000
8000
0.04 0.02 0.00 0.02 0.04 0.06 0.08 0.10
Centerline Deflection (in.)
L
o
a
d
,
P
(
l
b
s
.
)
Figure 4.2 – Load vs. Deflection Curve for Prestressed Concrete Model
Initial Cracking
Steel Yielding
Failure
Zero Deflection
Decompression
SelfWeight
Initial Cracking
Linear Region
Secondary
Linear Region
SEE FIGURE BELOW
Prestressing
The deflection in ANSYS (SAS 2003) corresponds well to the calculated value. The
stresses in the extreme fibers were looked at in the beam also. Since camber is occurring
in the beam (Figure 4.3), the controlling stress exists in the extreme top fiber.
Figure 4.3 – Deflection due to prestress
The handcalculated top fiber stress and the top fiber stress found using FEA are also
shown in Table 4.4. The values correlate very well.
A phenomenon that occurred when the prestress was added was bursting in the
concrete where the prestress force is being applied. This phenomenon can be seen in
Figure 4.4. The contour plot shows that at the time of prestressing, there is a maximum
Camber due to
prestressing force
Deflection (in.)
stress in the concrete where bursting occurs. The bursting phenomenon is explained in
detail in Nawy (2000).
Figure 4.4 – Bursting Phenomenon
Also, localized cracking occurs in the concrete, as shown in Figure 4.5. However, this
does not impact the solution because no other cracking occurs in that area after the initial
application of prestress. When a traditional level of prestress was applied
( 159, 840
pe
f psi = ), bursting zone cracking was extensive – a converged solution was not
possible to obtain. This is the reason for the reduction in
pe
f by the factor of 3.
Bursting Effect at
Center of Beam
Tendon Level
Figure 4.5 – Localized Cracking From Effective Prestress Application
4.2.2 SelfWeight
Adding the selfweight gives a deflection value which corresponds well to the calculated
deflection in Appendix B (Table 4.4). The FEA calculated top fiber stress at this stage
also correlates well with handcalculations.
4.2.3 Zero Deflection
Handcalculations (Appendix B) determined the load at which the deflection of the
centerline was zero. Table 4.4 indicates very good correlation between the hand
calculated values and the FEA.
Cracking that occurs
from bursting
4.2.4 Decompression
One definition of decompression denotes the point of loading where the stress at the
bottom fiber of the concrete beam moves from compression due to the prestress to
tension from superimposed loading. Table 4.4 indicates very good correlation at this
level of loading as well.
4.2.5 Initial Cracking
Initial cracking is defined to be the loading at which the extreme tension fiber reaches the
modulus of rupture. Initial cracking of the beam in the FE model occurs at load 7850 lbs.
The handcalculated load where cracking occurs is 7538 lbs. (Table 4.4 and Appendix B).
This is just past the modulus of rupture of the beam of 520 psi. The stress increases up to
539 psi at the centerline when the first crack occurs. This first crack occurs in the
constant moment region, and is a flexural crack.
4.2.6 Secondary Linear Region
In the secondary linear region of the response, significantly more cracking occurs as more
load is applied the beam as seen in Figure 4.6. Cracking increases in the constant
moment region, and the beam begins cracking out towards the supports with at a load of
12,000 lbs. Additional flexural cracking occurs in the beam at 20,000 lbs. Also, diagonal
tension cracks are beginning to form in the model.
Figure 4.6 – Cracking at 12,000 and 20,000 lbs.
4.2.7 Behavior of Steel Yielding and Beyond
Yielding of the prestress steel is defined as 0.85
pu
f for this model. Figure 4.2 illustrates
the load level at the point of yielding of the prestress steel. At this point in the response,
the displacements of the beam begin to increase at a higher rate as more load is applied.
The ability of the beam to distribute load throughout the crosssection has diminished
Flexural Cracks
Bursting Cracks
Diagonal Tension Cracks
greatly. Therefore, greater deflections occur at the beam centerline. Increasing flexural
cracks and diagonal tension cracks form as the beam approaches failure (Figure 4.6).
4.2.8 Flexural Limit State
At a load of 28,823 lbs. unresolvable nonconvergence of the nonlinear algorithm occurs,
indicating that cracking throughout the entire constant moment region has occurred.
Figure 4.7 illustrates the excessive cracking in the beam.
Figure 4.7 – Cracking at Flexural Capacity
Cracking at Failure
Hand calculations (Appendix B) predicted that the flexural capacity of the beam would
correspond to 27,587 lbs. (Table 4.4). The FEA prediction (28,823 lbs. in Table 4.4)
corresponds very well with the hand calculations. The stress in the prestressing steel at
failure predicted using FEA was 264,822 psi (Table 4.4). Using strain compatibility the
stress in the prestressing steel at failure was 254,520 psi (Appendix B), which
corresponds well to the FEA prediction.
The application of the moderate prestressing,
3
pe
f
, assumed in the present case
increased the limit load for the beam from 16,310 lbs. (Chapter 3) to 28,823 lbs. (near
doubling). The benefits of prestressing are apparent.
CHAPTER 5
CONCLUSIONS AND RECOMMENDATIONS
5.0 Introduction
The use of the finite element method to analyze reinforced and prestressed concrete
beams was evaluated. A reinforced concrete beam model was calibrated to experimental
data, and predictions of initial cracking, yielding of the steel and flexural failure of the
beam were compared to the experimental results. A prestressed concrete beam model
was also created. Initial prestress, addition of selfweight, zero deflection point,
decompression, initial cracking, yielding of steel, and flexural failure were then studied
and compared to theoretical values obtained via accepted methods of hand calculation.
5.1 Conclusions
The following conclusions can be stated based on the evaluation of the analyses of the
calibration model and the prestressed concrete beam.
(1) Deflections and stresses at the centerline along with initial and progressive
cracking of the finite element model compare well to experimental data obtained
from a reinforced concrete beam.
(2) The failure mechanism of a reinforced concrete beam is modeled quite well using
FEA, and the failure load predicted is very close to the failure load measured
during experimental testing.
(3) For the prestressed concrete beam, camber due to the initial prestress force and
after application of the selfweight of the beam compares well to handcomputed
values. Also, a bursting effect was seen in the FE model.
(4) Deflections and stresses at the zero deflection point and decompression are
modeled well using a finite element package.
(5) The load applied to cause initial cracking of the prestressed concrete beam
compares well with hand calculations.
(6) Flexural failure of the prestressed concrete beam is modeled well using a finite
element package, and the load applied at failure is very close to hand calculated
results.
5.2 Recommendations for Future Work
The literature review and analysis procedure utilized in this thesis has provided useful
insight for future application of a finite element package as a method of analysis. To
ensure that the finite element model is producing results that can be used for study, any
model should be calibrated with good experimental data. This will then provide the
proper modeling parameters needed for later use.
While modeling the prestressed beam, relaxation losses due to prestress, creep,
shrinkage, and elastic shortening were lumped together in a single load step. Individual
modeling of these losses could be included in future research. Many prestressed
structural components have nonsymmetric geometries and loadings. Therefore, non
symmetric geometries and loadings should be analyzed using finite element analysis with
prestress for further study. Higher strength concrete, the bursting effect, and the transfer
length of the prestressing steel are all candidates for future research.
REFERENCES
American Concrete Institute (1978), Douglas McHenry International Symposium on
Concrete and Concrete Structures, American Concrete Institute, Detroit, Michigan.
Branson, D.E.; Meyers, B.L.; and Kripanarayanan, K.M. (1970), “Loss of Prestress,
Camber and Deflection of Noncomposite and Composite Structures Using Different
Weight Concrete,” Iowa State Highway Comission, Report No. 706, Aug.
Buckhouse, E.R. (1997), “External Flexural Reinforcement of Existing Reinforced
Concrete Beams Using Bolted Steel Channels,” Master’s Thesis, Marquette University,
Milwaukee, Wisconsin.
Faherty, K.F. (1972), “An Analysis of a Reinforced and a Prestressed Concrete Beam by
Finite Element Method,” Doctorate’s Thesis, University of Iowa, Iowa City.
Fanning, P. (2001), “Nonlinear Models of Reinforced and Posttensioned Concrete
Beams,” Electronic Journal of Structural Engineering, University College Dublin,
Earlsfort Terrace, Dublin 2, Ireland, Sept.12.
Kachlakev, D.I.; Miller, T.; Yim, S.; Chansawat, K.; Potisuk, T. (2001), “Finite Element
Modeling of Reinforced Concrete Structures Strengthened With FRP Laminates,”
California Polytechnic State University, San Luis Obispo, CA and Oregon State
University, Corvallis, OR for Oregon Department of Transportation, May.
Janney, J.R. (1954), “Nature of Bond in Pretensioned Prestressed Concrete,” Journal of
the ACI, Proceedings, Vol.50, No.5, May.
MacGregor, J.G. (1992), Reinforced Concrete Mechanics and Design, PrenticeHall, Inc.,
Englewood Cliffs, NJ.
McCurry, D., Jr. and Kachlakev, D.I (2000), “Strengthening of Full Sized Reinforced
Concrete Beam Using FRP Laminates and Monitoring with Fiber Optic Strain Guages” in
Innovative Systems for Seismic Repair and Rehabilitation of Structures, Design and
Applications, Technomic Publishing Co., Inc., Philadelphia, PA, March.
Nawy, E.G., (2000), Prestressed Concrete: A Fundamental Approach, PrenticeHall,
Inc., Upper Saddle River, NJ
SAS (2003) ANSYS 7.1 Finite Element Analysis System, SAS IP, Inc.
Shing, P.B. and Tanabe, T.A., Ed. (2001), Modeling of Inelastic Behavior of RC
Structures Under Seismic Loads, American Society of Civil Engineers.
Tavarez, F.A., (2001), “Simulation of Behavior of Composite Grid Reinforced Concrete
Beams Using Explicit Finite Element Methods,” Master’s Thesis, University of
WisconsinMadison, Madison, Wisconsin.
Willam, K., and Tanabe, T.A., Ed. (2001), Finite Element Analysis of Reinforced
Concrete Structures, American Concrete Institute, Farmington Hills, MI.
Willam, K.J. and Warnke, E.P. (1974), “Constitutive Model for Triaxial Behaviour of
Concrete,” Seminar on Concrete Structures Subjected to Triaxial Stresses, International
Association of Bridge and Structural Engineering Conference, Bergamo, Italy, p.174.
APPENDICES
The following appendices have been included for reference:
PAGE
Appendix A: Theoretical Calculations for Calibration Model . . . . . . . . . . . . . . . . 69
Appendix B: Theoretical Calculations for Prestressed Model . . . . . . . . . . . . . . . . 73
Appendix A
Analysis of reinforced concrete beam for flexure at applied load of 5210 lbs.
Figure A.1 – Loading of Beam with Supports
Maximum Moment
The moment that occurs from the existing forces
max
(5210 .)(60 .) 312, 600 . . M lbs in lb in = = −
Material Properties
The gross moment of inertia
3 3 4
1 1
(10 .)(18 .) 4860
12 12
G
I bh in in in = = =
The modulus of elasticity of the concrete
'
57, 000 57, 000 4800 3, 949, 076
c c
E f psi = = =
The modulus of rupture
'
7.5 7.5 4800 520
r c
f f psi = = =
3”
60” 60”
3”
5210 lbs. 5210 lbs.
60”
Stresses in Concrete and Steel
The stresses at the extreme tension fiber are calculated using a transformed moment of
inertia of the concrete and steel reinforcement
Figure A.2 – Transformed CrossSection
Transformed area of steel
2
( ) 8 8(3)(0.31) 7.44
s t s
A A in = = =
2
3.72in distributed on each side of the concrete crosssection
Calculate the distance from the top fiber to the neutral axis of the transformed moment of
inertia
2
1 1 2 2
2
1 2
(18 .)(10 .)(9 .) (2)(3.72 . )(15.5 .)
9.258 .
(18 .)(10 .) (2)(3.72 . )
A y A y in in in in in
y in
A A in in in
+ +
= = =
+ +
The transformed moment of inertia
2 2
tr G
Concrete Steel
I I Ad Ad ( ( = + +
¸ ¸ ¸ ¸
4 2 2 2 4
4860 . (10 .)(18 .)(0.258 .) (2)(3.72 . )(6.242 .) 5162 . in in in in in in in ( ( = + + =
¸ ¸ ¸ ¸
The stress at the extreme tension fiber is then calculated
18”
3.72 in
2
15.5”
Y
bar
10”
4
(312, 600 . .)(8.742 .)
5162 .
b
ct
tr
My lb in in
f
I in
−
= = = 530 psi
The stress in the steel at this point is calculated
4
(312, 600 . .)(6.242 .)
(8)
5162 .
b
s
tr
My lb in in
f
I in
η
−
= = =3024 psi
Deflections
The deflection at the centerline of the concrete beam at load 5210 lbs.
2 2
max
(3 4 )
24
tr
Pa
l a
EI
∆ = −
2 2
4
(5210 .)(60 .)
(3(180 .) 4(60 .) )
24(3, 949, 076)(5162 . )
lb in
in in
in
= − = 0.0529 in.
Loads
The load at first cracking
b
ct
tr
My
f
I
=
4
(60 .)(8.742 .)
520
5162 .
P in in
psi
in
=
P =5117 lbs.
Appendix B
Analysis of Deflection for the Prestressed Concrete Beam
Figure B.1 – Typical Prestressed Concrete Beam with Supports
Effective Prestress
270, 000
pu
f psi =
0.74 0.74(270, 000 ) 199, 800
pi pu
f f psi psi = = =
0.80 0.80(199, 800 ) 159, 840
pe pi
f f psi psi = = =
159, 840 / 3 psi = 53,280 psi
Deflections
Deflection due to prestress
2 2
4
(24, 456 .)(6.6 .)(180 .)
8 8(3, 949, 076 )(4860 . )
c
c c
Pel lb in in
E I psi in
δ = − = − =0.03406 in.
Deflection due to selfweight
4 4
5 5(15.624 )(180 .)
384 384(3, 949, 076 )(4860)
wl pli in
EI psi
δ = = = 0.1113 in.
Deflection after prestress and selfweight is applied
0.03406 0.01113 0.02293 .
net c sw
in δ δ δ = + = − + = −
3”
60” 60”
3”
60”
cgs
cgc
e
c
P P
Stresses
Stress in extreme top fiber of the beam after initial prestress
2 2 2
24, 456 . (6.6 .)(9 .)
(1 ) (1 )
180 . 27 .
t e t
c
P ec lb in in
f
A r in in
= − − = − − = 163.04 psi (T)
'
0.45 0.45(4800 ) 2160 ( ), . .
c c
f f psi psi T O K < = = =
Stress in extreme top fiber of the beam after selfweight
2 2 2 3
24, 456 . (6.6 .)(9 .) 63277.2 . .
(1 ) (1 )
180 . 27 . 540 .
t e t T
t
c
P ec M lb in in lb in
f
A r S in in in
−
= − − − = − − −
= 45.86 psi (T)
'
0.45 0.45(4800 ) 2160 ( ), . .
c c
f f psi psi T O K < = = =
Stress in extreme bottom fiber after prestress and selfweight
2 2 2 3
24, 456 . (6.6 .)(9 .) 63, 277.2 . .
(1 ) (1 ) 317.6 ( )
180 . 27 . 540 .
e b T
b
c b
P ec M lb in in lb in
f psi C
A r S in in in
−
= − + + = − + + = −
'
12 12 4800 831.4 ( ), . .
t c
f f psi psi T O K < = = =
Loads
Load needed to cause zero deflection
2 2
4
(60 .)
0.02293 . (3(180 .) 4(60 .) )
24(3, 949, 076 )(4860 . )
P in
in in in
psi in
= −
P =2126 lbs.
Calculated Load of Application at Decompression
2
(1 )
e b T
b
c b
P ec M My
f
A r S I
= − + + +
(60 .)(9 .)
0 317.6
4
P in in
psi = − +
P =2858 lbs.
Determine the force of first cracking based on the cracked moment
2
( )
cr D L r b e
b
r
M M M f S P e
c
= + = + +
2
3
27 .
63, 277.2 . . (60 .) (520 )(540 . ) 24, 456 .(6.6 . )
9 .
in
lb in P in psi in lbs in
in
− + = + +
P =7538 lbs.
Ultimate Strength Design by Strain Compatibility
1
53, 280
.
0.0019029
.
28, 000, 000
pe
pe
ps
f
psi
in
in
E psi
ε ε = = = =
2
3(0.153 . )(53, 280 ) 24, 456 .
e
P in psi lbs = =
2 2
2 2 2 2
24, 456 . (6.6 .)
.
(1 ) (1 ) 0.0000899
.
(180 . )(3, 949, 076 ) 27 .
e
decomp
c c
P e lbs in
in
in
A E r in psi in
ε ε = = + = + =
Assume 205,000 psi for iteration process:
2
'
3(0.153 . )(205, 000 )
2.30625 .
0.85 0.85(4800 )(10 .)
ps ps
c
A f
in psi
a in
f b psi in
= = =
1
0.85 0.05 0.80 β = − =
1
2.30625
2.8828 .
0.8
a
c in
β
= = =
15.6 . d in =
3
15.6 . 2.8828 .
. .
( ) 0.003 ( ) 0.0132342
. .
2.8828 .
c
d c in in
in in
in in
c in
ε ε
− −
= = =
1 2 3
. . . .
0.0019029 0.0000899 0.0132342 0.015227
. . . .
ps
in in in in
in in in in
ε ε ε ε = + + = + + =
0.075 0.075
268 268 259, 406
.
0.0065
0.015227 0.0065
.
ps
ps
f psi
in
in
ε
= − = − =
−
−
Plugging 259,406 psi and iterating, the value of prestress obtained is:
ps
f = 254,520 psi
Using the prestress force, the nominal moment obtained is:
2
2.8633 .
( ) 3(0.153 . )(254, 520 )(15.6 . ) 1, 655, 213 . .
2 2
n ps ps p
a in
M A f d in psi in lb in = − = − = −
Failure
Determining the force needed to reach failure,
n
M Pa =
1, 655, 213 . .
60 .
n
M lb in
P
a in
−
= = = 27,587 lb.
Marquette University
This is to certify that we have examined
this copy of the masters thesis by
Anthony J. Wolanski
and have found that it is complete and
satisfactory in all respects.
This thesis has been approved by:
Director
Christopher M. Foley, Ph.D., P.E.
Stephen M. Heinrich, Ph.D.
Baolin Wan, Ph.D.
PREFACE Several methods have been utilized to study the response of concrete structural components. Experimental based testing has been widely used as a means to analyze individual elements and the effects of concrete strength under loading. The use of finite element analysis to study these components has also been used. This thesis is a study of reinforced and prestressed concrete beams using finite element analysis to understand their loaddeflection response. A reinforced concrete beam model is studied and compared to experimental data. The parameters for the reinforced concrete model were then used to model a prestressed concrete beam. Characteristic points on the loaddeformation response curve predicted using finite element analysis were compared to theoretical (handcalculated) results. Conclusions were then made as to the accuracy of using finite element modeling for analysis of concrete. The results compared well to experimental and hand calculated.
ACKNOWLEDGMENTS This research was performed under the supervision of Dr. Christopher M. Foley. I am extremely grateful for the guidance, knowledge, understanding, and numerous hours spent helping me complete this thesis. Appreciation is also extended to my thesis committee, Dr. Stephen M. Heinrich and Dr. Baolin Wan, for their time and efforts. I would like to thank my parents, John and Sue Wolanski, my brother, John Wolanski, and my sister, Christine Wolanski for their understanding, encouragement and support. Without my family these accomplishments would not have been possible.
TABLE OF CONTENTS
PAGE LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
CHAPTER 1 – INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1 1.2 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Objectives and Outline of Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
CHAPTER 2 – LITERATURE REVIEW AND SYNTHESIS . . . . . . . . . . . . . . . . . 4 2.0 2.1 2.2 2.3 2.4 2.5 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 ExperimentBased Testing of Concrete . . . . . . . . . . . . . . . . . . . . . . . 5 Finite Element Analysis .................................. 6
Failure Surface Models for Concrete . . . . . . . . . . . . . . . . . . . . . . . . 10 FE Modeling of Steel Reinforcement . . . . . . . . . . . . . . . . . . . . . . . . 13 Direction for Present Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
CHAPTER 3 – CALIBRATION MODEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 3.0 3.1 3.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Experimental Beam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 ANSYS Finite Element Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 3.2.1 3.2.2 Element Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Real Constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2.3 Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
. . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . 33 Numbering Controls . . . . . . . . . . . . . . . . . . . . . . . .1. . . . .4 3. . . . . 43 Behavior Beyond First Cracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Meshing . . . . . .6 Behavior at First Cracking . . .6 3. . . . . 54 SelfWeight . . . . . . . . . . . . . . . . . . . . .2 Material Properties . . .5 3. . . . .1 Introdution . . . . . . . . . . . . . . . . . . . . . . . 42 3. . . . . . . . 37 Analysis Process for the Finite Element Model . . . . . . . . . . . .3. . . . . . 34 Loads and Boundary Conditions . . . . . . . . 48 CHAPTER 4 – PRESTRESSED CONCRETE BEAM MODEL . . 51 Analysis Process for the Finite Element Model . . . . . . .3 3. . . . . . 46 LoadDeformation Response .1 4. . . . . . 43 Behavior of Reinforcement Yielding and Beyond . . 49 Finite Element Model . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . .0 4. . . . . . .1 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8 3. . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . .2 Solution . . . . . . .3 4. . . . . . . . . . . . . . . . . .1 Real Constants . . . . . . . . . . . . . . . . . . . . 35 Analysis Type . . . 49 4. . . . . . . . . .2. . . . . . . . . . . . . . . . . .5 3. .1. . . .3. . . . . . . . . . . . . .3. . . .2 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 Modeling . . 51 4.2. . . 40 Results .2 Application of Effective Prestress . . . 53 Results . 44 Strength Limit State . .4 4. . . . . . . . .9 3. . 54 4. . . . . .1. . . . . . . . . . . . . . . . 50 4. . . .3. . . . . . . . . . . .3. . . . . . . . . .2. . 49 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1. . . . . . . . . . .4 3. . . .2. . 42 Behavior at Initial Cracking . . . . 58 . .2. . . . . . . .2. . . . . . . . . . . . .7 3. . . . . . . . . . . . . . . . . . . . . .
. 73 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8 Zero Deflection . . . . . 66 APPENDICES . . . . . . . . . . . . . . . . . 63 5. . . . . . . . . . . . . . . . . . . . . . . . .0 5. . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . .4. . 59 Initial Cracking . . . . . . 61 CHAPTER 5 – CONCLUSIONS AND RECOMMENDATIONS . . . . . . . . .5 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Theoretical Calculations for Prestressed Model . . . . . 68 Theoretical Calculations for Calibration Model .2. . . .2. . . . . . . .1 5. . . . . 59 Behavior of Steel Yielding and Beyond . . . . . . . . . . . . . . . . . . . . 60 Flexural Limit State . . . . . . .6 4. . . . . . . . . . . . . . . . . . . . . . . . . . .2 Introduction . . . . . . . . . . . . 63 Recommendations for Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Decompression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . 59 Secondary Linear Region .4 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 Conclusions . . . . . . . . . . . . . . . . . . . .3 4. . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . .2. . . . . . . 64 REFERENCES . . . . . . . . . . . .7 4. . . . . . . . .
. 5 Reinforced Concrete Beam With Loading (Faherty 1972) . . . . . . . . . . . . . .10 Three Parameter Model (Willam and Warnke 1974) . . . . . . . .1 2. . . . SAS 2003) . . . . SAS 2003) . . . . . . . . et al. . . . . . . 31 Mesh of the Concrete. . . . . . . . 18 Load vs. . . . .LIST OF FIGURES FIGURE 2. . . . 10 2. . 6 FEM Discretization for a Quarter of the Beam (Kachlakev. . . . . . . . . . . . . . . . . . . .3 2. 11 2.8 3. . . . . .7 3. . . . . . . . 22 Link 8 Element (ANSYS. . . . .1 3. . . . . . . . . . . . . . .2 3. . . . . . . . . . . . . .6 3. . . and Steel Support . . . . . . . 9 Typical Cracking Signs in Finite Element Models (Kachlakev. . . .2 2. . . . . . . . . . 27 Volumes Created in ANSYS . . . . . . . . . . . . . . . . .5 PAGE Typical Cracking of Control Beam at Failure (Buckhouse 1997) . . . 8 Load vs. . . . . Deflection Plot (Kachlakev. 23 Uniaxial StressStrain Curve . . . 2001) . . . . . . . . . . . . . et al. . . . . . . . .4 2. . . . . . . . . . .3 3. . . . . . 17 Failure in Flexure (Buckhouse 1997) . . . . . . . 32 . . . . . . . . . et al. . . Steel Plate. . . . 19 Solid 65 Element (ANSYS. 2001) . . . . . . . . . . . . . . . . . . . . . . 21 Solid 45 Element (ANSYS. . . . . . . . . . . . . . . . .9 3. . . Deflection Curve for Beam C1 (Buckhouse 1997) . . . . .8 3. . . . . . . . 2001) . . . . . . . 16 Typical Detail for Control Beam Reinforcement (Buckhouse 1997) . . . . . . . . . . . . . 11 Models for Reinforcement in Reinforced Concrete (Tavarez 2001) . . . . 14 Loading and Supports for the Beam (Buckhouse 1997) . . . . SAS 2003) . . . . . . . .7 2. . . . . . . . . . . . . .5 3. . . .6 Failure Surface of Plain Concrete Under Triaxial Conditions (Willam and Warnke 1974) . . . . . . . . . . . . . . .4 3. . . . . . . . . . . . . . .
. . . . . . . . . .000 and 12. . . . . . . . . . . . . . . . . .000 lbs.7 A. . . . . . . . . . . . . . . . . . .11 3. Deflection Curve Comparison of ANSYS and Buckhouse (1997) 48 StressStrain Curve for 270 ksi strand . . . .4 4. . .13 3. . . . . Deflection Curve for Prestressed Concrete Model . . . . . . . . . . . . . . . 74 . .15 3. . . . . . . . . . . . . . . . . . . . 33 Boundary Conditions for Planes of Symmetry . . . . . . . . . . . . . . . . .16 3. . . 56 Bursting Phenomenon . . . . . . . . . . . . . . . . . . . . . . . 46 Failure of the Concrete Beam . . 52 Load vs. . . . . . . . . . . . . . . . . . . . .2 B. . .6 4. . . . . . 58 Cracking at 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5 4.18 3. . . . . . . . . . . 44 Cracking at 8. . . . . . . . . . . . . . . . . . . . . . . . 45 Increased Cracking After Yielding of Reinforcement . . . . . . . .14 3. . . . . . . . . . . . . . . . . . . . . . . . . . 57 Localized Cracking From Effective Prestress Application . . . . . . . . 36 Boundary Conditions at the Loading Plate . . . . . . . . . . . . . . . . . . . . . .1 A. . . . . . . . . . . . 55 Deflection due to prestress . . . 37 1st Crack of the Concrete Model .2 4.000 lbs. . . . . . . . . . . 70 Transformed CrossSection . . . . . . . . . . . . . .000 and 20. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 4. . 35 Boundary Condition for Support . .19 4. . . 60 Cracking at Flexural Capacity . . . . . . . . . . . . . 71 Typical Prestressed Concrete Beam with Supports . . . . . . . . . . . . . 47 Load vs. . . . . . . . . . . . . . . . . . . . 61 Loading of Beam with Supports . . . . . . . . . . . . . . . .12 3. . . . . . . .3 4. . . . . . . . . . . .3. . .1 Reinforcement Configuration . . . . . . . . . . . . . . . . . . . . . . .17 3. . .
. . . . . . . . . . . . . . . . . . 25 Dimensions for Concrete. . . . . . . . . . . . . . . . . . . . Steel Plate. . . . . . . . . . . . . . . . . . . . . . . . . Finite Element Model At Ultimate Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Test Data for Control Beam C1 (Buckhouse 1997) . . . . . 50 Values for Multilinear Isotropic StressStrain Curve . . .4 3. . . . . . 53 Analytical Results . . . . . . . . . .4 Real Constants for Prestressed Beam . . . . 33 Commands Used to Control Nonlinear Analysis . . . . . . . 39 3. . . . . . . . . 54 . . 47 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 4. .11 Advanced Nonlinear Control Settings Used . . . . . . . . . . . . . . .LIST OF TABLES TABLE 3. . . . 21 Real Constants for Calibration Model . 23 Material Models for the Calibration Model . . . . . . . . . . .13 3. 52 Load Increments for Analysis of Prestressed Beam Model . . . . . . . .8 3. 43 Deflections of Control Beam (Buckhouse 1997) vs. . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 3. . . 31 Mesh Attributes for the Model . . . . . . . . . . . . . . . . . . . . . 40 Deflection and Stress Comparisons At First Cracking . . . . . . .3 4. .10 PAGE Properties for Steel and Concrete (Buckhouse 1997) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 3. . . . . . . .1 4. . . . . .5 3. . . . . . . . . .3 3. . . . . . . . . . . . . . . .1 3. . . . . . . . .12 3. . 38 Commands Used to Control Output . . . . . . and Steel Support Volumes . . . . .9 3. . . . . . 19 Element Types for Working Model . .2 3. . . . . . . . . . . . . . .7 3. . . . .14 Load Increments for Analysis of Finite Element Model . 38 Nonlinear Algorithm and Convergence Criteria Parameters . . . . .
executing the . The use of finite element analysis to study these components has also been used. The use of computer software to model these elements is much faster. Data obtained from a finite element analysis package is not useful unless the necessary steps are taken to understand what is happening within the model that is created using the software. Experimental based testing has been widely used as a means to analyze individual elements and the effects of concrete strength under loading. and the use of materials can be quite costly. it is extremely time consuming. early attempts to accomplish this were also very time consuming and infeasible using existing software and hardware. It has now become the choice method to analyze concrete structural components.CHAPTER 1 INTRODUCTION 1. however. In recent years. While this is a method that produces real life response. and extremely costeffective.1 General Concrete structural components exist in buildings and bridges in different forms. Different methods have been utilized to study the response of structural components. one must look back to experimental data and simple analysis. Also. Understanding the response of these components during loading is crucial to the development of an overall efficient and safe structure. To fully understand the capabilities of finite element computer software. the use of finite element analysis has increased due to progressing knowledge and capabilities of computer software and hardware. Unfortunately.
a literature review was conducted to evaluate previous experimental and analytical procedures related to reinforced concrete components. This thesis is a study of reinforced and prestressed concrete beams using finite element analysis to understand the response of reinforced and prestressed concrete beams due to transverse loading. Second. Based on the results obtained from the calibration model and the analysis/modeling parameters set by this model.2 Objectives and Outline of Thesis The objective of this thesis was to investigate and evaluate the use of the finite element method for the analysis of reinforced and prestressed concrete beams The following procedure was used to meet this goal. A mildsteel reinforced concrete beam with flexural and shear reinforcement was analyzed to failure and compared to experimental results to calibrate the parameters in ANSYS (SAS 2003) for later analyses. 1. a prestressed concrete beam was analyzed from initial prestress to flexural failure. SAS 2003) was set up and evaluated using experimental data. more efficient and better analyses can be made to fully understand the response of individual structural components and their contribution to a structure as a whole. stresses. First. a calibration model using a commercial finite element analysis package (ANSYS.necessary checks along the way is key to make sure that what is being output by the computer software is valid. These key points . By understanding the use of finite element packages. and cracking of the concrete beam were analyzed at different key points along the way. Deflections.
decompression. addition of selfweight.include initial prestress. zero deflection point. and failure. Conclusions regarding the analysis are then drawn and recommendations for further research are made. Discussion of the results obtained for the calibration model and the prestressed concrete beam model is also provided. yielding of steel. . initial cracking.
Shing and Tanabe (2001) also put together a collection of papers dealing with inelastic behavior of reinforced concrete structures under seismic loads.CHAPTER 2 LITERATURE REVIEW AND SYNTHESIS 2. shear failure of reinforced concrete beams. The use of FEA has been the preferred method to study the behavior of concrete (for economic reasons). and the modeling of the shear behavior of reinforced concrete bridge structures. Nawy 2000). The monograph contains contributions that outline applications of the finite element method for studying postpeak cyclic behavior and ductility of reinforced concrete columns. MacGregor 1992. .0 Introduction To provide a detailed review of the body of literature related to reinforced and prestressed concrete in its entirety would be too immense to address in this thesis. This literature review and introduction will focus on recent contributions related to FEA and past efforts most closely related to the needs of the present work. This collection contains areas of study such as: seismic behavior of structures. and concretesteel bond models. the analysis of reinforced concrete beamcolumn bridge connections. there are many good references that can be used as a starting point for research (ACI 1978. the analysis of reinforced concrete components in bridge seismic design. However. Willam and Tanabe (2001) contains a collection of papers concerning finite element analysis of reinforced concrete structures. cyclic loading of reinforced concrete columns.
and concrete material models. All three beams were loaded with transverse point loads at third points along the beams.1 – Typical Cracking of Control Beam at Failure (Buckhouse 1997) .The focus of these most recent efforts is with bridges.1 ExperimentBased Testing Of Concrete Buckhouse (1997) studied external flexural reinforcement of existing concrete beams. experimental testing. The following is a review and synthesis of efforts most relevant to this thesis discussing FEA applications.1. Three concrete control beams were cast with flexural and shear reinforcing steel. Loading was applied to the beams until failure occurred as shown in Figure 2. and seismic design. 2. Figure 2. columns. The focus of this thesis is the study of nonprestressed and prestressed flexural members. Shear reinforcement was placed in each beam to force a flexural failure mechanism.
The mode of failure characterized by the beams was compression failure of the concrete in the constant moment region (flexural failure). Loaddeflection curves were plotted for each beam and compared to predicted ultimate loads. This thesis will utilize the experimental results of these control beam tests for calibration of the FE models. 2. Figure 2.2). The two beams that were selected for modeling were simply supported and loaded with two symmetrically placed concentrated transverse loads (Figure 2. All failures were ductile.2 Finite Element Analysis Faherty (1972) studied a reinforced and prestressed concrete beam using the finite element method of analysis.2 – Reinforced Concrete Beam With Loading (Faherty 1972) . with significant flexural cracking of the concrete in the constant moment region.
However. and the influence of progressive cracking of the concrete. Because the loading and geometry of the beam were symmetrical. the time dependent prestress loss. and a partially cracked section. (1970). . the loaddeflection curve past the cracking point was not generated because only three distinct cracking patterns were used for this analysis. the elastic prestress loss. It was recommended that additional analysis of the prestressed concrete beam should be undertaken after a procedure is developed for modeling the tensile rupture of the concrete. whereas the transverse loading was applied incrementally. Symmetry was once again utilized. bilinear steel properties. The model utilized in this research required the beam to be unloaded and the finite element model redefined as each crack is initiated or extended. et al. These results for the prestressed beam showed that deflections computed using the finite element model were very similar to those observed by Branson. and the loss of tensile stress in the concrete as a result of concrete rupture were applied as single loading increments. Faherty (1972) also analyzed a prestressed concrete beam that included: nonlinear concrete properties. release of the prestressing force. The transverse loading was incrementally applied and ranged in magnitude from zero to a load well above that which initiated cracking. Only three finite element models of the prestressed beam were implemented (or used): two uncracked sections.The analysis for the reinforced concrete beam included: nonlinear concrete properties. The finite element model produced very good results that compared well with experimental results in Janney (1954). and bilinear steel properties. only one half of the beam was modeled using FEA. a linear bond slip relation with a destruction of the bond between the steel and concrete. a linear bondslip relation. The dead load.
Kachlakev. et al.3 – FEM Discretization for a Quarter of the Beam (Kachlakev. Figure 2. (2001) used ANSYS (SAS 2003) to study concrete beam members with externally bonded Carbon Fiber Reinforced Polymer (CFRP) fabric. At certain stages in the analysis. No stirruptype reinforcement was used. Symmetry allowed one quarter of the beam to be modeled as shown in Figure 2.3. A single line support was utilized to allow rotation at the supports. The mesh was refined immediately beneath the load (Figure 2. the displacement in the direction perpendicular to the plane was set to zero. Loads were placed at third points along the full beam on top of steel plates. load step sizes were varied from large . 2001) At planes of symmetry. The nonlinear NewtonRaphson approach was utilized to trace the equilibrium path during the loaddeformation response. It was found that convergence of solutions for the model was difficult to achieve due to the nonlinear behavior of reinforced concrete material. et al.3).
Diagonal tension cracks (Figure 2. The loaddeflection curve for the nonCFRP reinforced beam that was plotted shows reasonable correlation with experimental data (McCurry and Kachlakev 2000) as shown in Figure 2. Figure 2. compression failure (crushing).5b) are shown as circles. concrete crack/crush plots were created at different load levels to examine the different types of cracking that occurred within the concrete as shown in Figure 2.(at points of linearity in the response) to small (when instances of cracking and steel yielding occurred). and diagonal tension cracks. Compression failures (Figure 2.4 – Load vs. The different types of concrete failure that can occur are flexural cracks.5a) form vertically up the beam. et al.4. .5c) form diagonally up the beam towards the loading that is applied. 2001) Also. Flexural cracks (Figure 2.5. Deflection Plot (Kachlakev.
b)Compressive Cracks. 2001) This study indicates that the use of a finite element program to model experimental data is viable and the results that are obtained can indeed model reinforced concrete beam behavior reasonably well. 2. The failure surface in principal stressspace is shown in Figure 2.5 – Typical Cracking Signs in Finite Element Models: a)Flexural Cracks.6. . These stress components are the major principal stresses. et al.Figure 2.3 Failure Surface Models For Concrete Willam and Warnke (1974) developed a widely used model for the triaxial failure surface of unconfined plain concrete. c)Diagonal Tensile Cracks (Kachlakev. The mathematical model considers a sextant of the principal stress space because the stress components are ordered according to σ 1 ≥ σ 2 ≥ σ 3 .
6).6 – Failure Surface of Plain Concrete Under Triaxial Conditions (Willam and Warnke 1974) Figure 2. The deviatoric section in Figure 2.The failure surface is separated into hydrostatic (change in volume) and deviatoric (change in shape) sections as shown in Figure 2. Figure 2.7).7 lies in a plane normal to the equisectrix (dashed line in Figure 2.7.7 – Three Parameter Model (Willam and Warnke 1974) . The hydrostatic section forms a meridianal plane which contains the equisectrix σ 1 = σ 2 = σ 3 as an axis of revolution (see Figure 2.
simple identification of model parameters from standard test data.The deviatoric trace is described by the polar coordinates r .g. continuous surface with continuously varying tangent planes). 2. The yield condition can . 3. convexity (e. a constitutive model for the concrete suitable for FEA implementation was formulated. smoothness (e. biaxial compressive strength ( f cb ).g. are identified from the uniaxial compressive strength ( f cu ). and uniaxial tension strength ( ft ) The Willam and Warnke (1974) mathematical model of the failure surface for the concrete has the following advantages: 1. close fit of experimental data in the operating range. monotonically curved surface without inflection points). Based on the above criteria. The free parameters of the failure surface z and r . 4. and θ where r is the position vector locating the failure surface with angle. The failure surface is defined as: 1 σa 1 τa + =1 z f cu r (θ ) f cu where: (2.1) σ a and τ a = average stress components z = apex of the surface f cu = uniaxial compressive strength The opening angles of the hydrostatic cone are defined by ϕ1 and ϕ2 . This constitutive model for concrete based upon the Willam and Warnke (1974) model assumes an appropriate description of the material failure. θ .
be approximated by three or five parameter models distinguishing linear from nonlinear and elastic from inelastic deformations using the failure envelope defined by a scalar function of stress f (σ ) = 0 through a flow rule. while using incremental stressstrain relations. Both of these methods are feasible and give stress values that satisfy the constitutive constraint condition. which allows the elastic path to reach the yield surface at the intersection with the normal therefore solving a linear system of equations.8a) uses bar or beam elements that are connected to concrete mesh nodes. and the smeared model. Therefore. and normal penetration. two numerical strategies were recommended: proportional penetration. since integration is avoided.4 FE Modeling of Steel Reinforcement Tavarez (2001) discusses three techniques that exist to model steel reinforcement in finite element models for reinforced concrete (Figure 2. The reinforcement in the discrete model (Figure 2. From the standpoint of computer application the normal penetration approach is more efficient than the proportional penetration method.7. The parameters for the failure surface can be seen in Figure 2. which subdivides proportional loading into an elastic and inelastic portion which governs the failure surface using integration. 2. the embedded model. A drawback to this model is that the concrete mesh is restricted by the .8): the discrete model. During transition from elastic to plastic or elastic to brittle behavior. the concrete and the reinforcement mesh share the same nodes and concrete occupies the same regions occupied by the reinforcement.
increasing the run time and computational cost. When reinforcement is complex. and (c) smeared The embedded model (Figure 2. this model is very advantageous. The model is built in a way that keeps reinforcing steel displacements compatible with the surrounding concrete elements. However. (b) embedded. therefore.location of the reinforcement and the volume of the mildsteel reinforcement is not deducted from the concrete volume. .8b) overcomes the concrete mesh restriction(s) because the stiffness of the reinforcing steel is evaluated separately from the concrete elements. (a) (b) (c) Figure 2.8 – Models for Reinforcement in Reinforced Concrete (Tavarez 2001): (a) discrete. this model increases the number of nodes and degrees of freedom in the model.
The different stages of the response of a prestressed concrete beam are computed using FEA and compared to results generated using hand computations. It was decided to use ANSYS (SAS 2003) as the FE modeling package. . This approach is used for largescale models where the reinforcement does not significantly contribute to the overall response of the structure. A reinforced concrete beam with reinforcing steel modeled discretely will be developed with results compared to the experimental work of Buckhouse (1997). A second analysis of a prestressed concrete beam will also be studied. 2.The smeared model (Figure 2.8c) assumes that reinforcement is uniformly spread throughout the concrete elements in a defined region of the FE mesh.5 Direction for Present Research The literature review suggested that use of a finite element package to model reinforced and prestressed concrete beams was indeed feasible. Fanning (2001) modeled the response of the reinforcement using the discrete model and the smeared model for reinforced concrete beams. It was found that the best modeling strategy was to use the discrete model when modeling reinforcement. The loaddeflection response of the experimental beam will be compared to analytical predictions to calibrate the FE model for further use.
1. The width and height of the beams tested were 10 in. The study included experimental testing of control beams that can be used for calibration of finite element models.0 Introduction This chapter discusses the calibration of the finite element model using experimental loaddeformation behavior of a concrete beam provided in Buckhouse (1997).1 Experimental Beam Buckhouse (1997) studied a method to reinforce a concrete beam for flexure using external structural steel channels. and 18 in.. All the necessary steps to create the calibrated model are explained in detail and the steps taken to generate the analytical loaddeformation response of the member are discussed. the length Figure 3. respectively.1 – Loading and Supports for the Beam (Buckhouse 1997) .CHAPTER 3 CALIBRATION MODEL 3. 3. As shown in Figure 3. The use of ANSYS (SAS 2003) to create the finite element model is also discussed.
Cover for the rebar was set to 2 in. 28day compressive stress of concrete. The layout of the reinforcement is detailed in Figure 3. from each end of the beam allowing a simply supported span of 15 ft. in all directions.2 – Typical Detail for Control Beam Reinforcement (Buckhouse 1997) The steel yield stress. with supports located 3 in.2) Yield Stress of Steel. fc' (psi) 0.of the beam was 15 ft. The mild steel flexural reinforcements used were 3#5 bars and shear reinforcements included #3 Ustirrups.1. Figure 3.93 60.2.6 in. and area of steel reinforcement are included in Table 3.1 – Properties for Steel and Concrete (Buckhouse 1997) Area of Steel (in.770 .000 4. fy (psi) 28Day Compressive Strength of Concrete. Table 3.
3).3. Figure 3.2. Data acquisition equipment was used to record applied loading. and strain in the internal flexural reinforcement. The theoretical ultimate load for the beam was calculated to be 14. Test data for the beam is summarized in Table 3. The ultimate loading corresponded to the nominal flexural capacity of the crosssection being reached.2 shows the experimental ultimate load determined was 16.310 lbs.1).3 – Failure in Flexure (Buckhouse 1997) Vertical cracks first formed in the constant moment region.4. The beam was loaded to flexural failure (Figure 3. extended upward. A plot of load versus deflection for control beam C1 (Buckhouse 1997) is shown in Figure 3.Two 50kip capacity load cells were placed at third points. and then out towards the constant shear region with eventual crushing of the concrete in the constant moment region as shown in Figure 3. beam deflection at the midspan. or 5 ft. .600 lbs (Buckhouse 1997). Table 3. from each support on steel bearing plates (Figure 3.
5 Avg.65 compression failure of concrete 18000 C 16000 14000 12000 Avg.) Avg. Load at 1st Crack (lbs.) C1 theoretical ultimate load (14.4 – Load vs.500 16.600 lbs. P (lbs.Steel Yielding Point C . Load.) Mode of Failure 4.5 3 3.) B 10000 8000 Nonlinear Region A Point A .Table 3.First Cracking Point B .) Avg.Failure 6000 4000 Linear Region 2000 0 0 0.) Figure 3.5 1 1.5 4 4. Centerline Deflection at Failure (in.2 – Test data for control beam C1 (Buckhouse 1997) Avg. P (lbs.5 2 2. Centerline Deflection (in.310 3. Failure Load. Deflection Curve for Beam C1 (Buckhouse 1997) .
The Solid65 element was used to model the concrete. This element is capable of plastic deformation. A second slope corresponding to the cracked section is followed until point B where the flexural reinforcement yields. 3. This section describes the different tasks and entries into used to create the FE calibration model.5. and crushing. To create the finite element model in ANSYS (SAS 2003) there are multiple tasks that have to be completed for the model to run properly.3. Due to the symmetry in crosssection of the concrete beam and loading. only one quarter of the beam was modeled. the GUI was utilized to create the model.2 ANSYS Finite Element Model The FEA calibration study included modeling a concrete beam with the dimensions and properties corresponding to beam C1 tested by Buckhouse (1997). symmetry was utilized in the FEA.1 Element Types The element types for this model are shown in Table 3. and z directions. For this model. . A schematic of the element is shown in Figure 3.The plot shows the linear behavior before first cracking (point A). 3. cracking in three orthogonal directions. The cracked moment of inertia with yielding internal reinforcement then defines the stiffness until flexural failure at point C.2. Models can be created using command prompt line input or the Graphical User Interface (GUI). This element has eight nodes with three degrees of freedom at each node – translations in the nodal x. y.
and z directions. The geometry and node locations for this element is shown .Table 3.3 – Element Types For Working Model Material Type ANSYS Element Concrete Solid65 Steel Plates and Solid45 Supports Steel Reinforcement Link8 Figure 3.5 – Solid 65 Element (SAS 2003) A Solid45 element was used for steel plates at the supports for the beam. This element has eight nodes with three degrees of freedom at each node – translations in the nodal x. y.
This element is shown in Figure 3. No real constant set exists for the Solid45 element. .2.6 – Solid 45 Element (SAS 2003) A Link8 element was used to model steel reinforcement. Note that individual elements contain different real constants.7. Figure 3.4.2 Real Constants The real constants for this model are shown in Table 3. This element is a 3D spar element and it has two nodes with three degrees of freedom – translations in the nodal x. y.in Figure 3.6. This element is also capable of plastic deformation. and z directions. 3. The descriptions for each element type are laid out in the ANSYS element library (SAS 2003).
2) Initial Strain (in.) Crosssectional Area (in.2) Initial Strain (in.2) Initial Strain (in.7 – Link 8 Element (SAS 2003) Table 3.2) Initial Strain (in./in.) 0 0./in./in.) Crosssectional Area (in./in.11 0 0.4 – Real Constants For Calibration Model Real Constant Set Element Type Constants Real Real Real Constants for Constants for Constants for Rebar 1 Rebar 2 Rebar 3 Material Number 0 0 0 Volume Ratio 0 0 0 Orientation Angle 0 0 0 Orientation Angle 0 0 0 Crosssectional 0.055 0 1 Solid 65 2 Link8 3 Link8 4 Link8 5 Link8 .Figure 3.155 0 0.31 Area (in.) Crosssectional Area (in.
5. The reinforcement has uniaxial stiffness and the directional orientation is defined by the user. A value of zero was entered for the initial strain because there is no initial stress in the reinforcement.3 Material Properties Parameters needed to define the material models can be found in Table 3.2. and z directions in the element (Figure 3. y. and Orientation Angles. Therefore. As seen in Table 3. 3. It requires real constants for rebar assuming a smeared model.Real Constant Set 1 is used for the Solid65 element. In the present study the beam is modeled using discrete reinforcement. . and 5 are defined for the Link8 element. Once again set 5 is half of set 4 because half of the stirrup at the midspan of the beam is cut off resulting from symmetry.5. Real Constant Sets 2. 3.5). Volume Ratio. there are multiple parts of the material model for each element. ANSYS (SAS 2003) allows the user to enter three rebar materials in the concrete. set 3 is half of set 2 because onehalf the center bar in the beam is cut off. a value of zero was entered for all real constants which turned the smeared reinforcement capability of the Solid65 element off. Values can be entered for Material Number. Crosssectional areas in sets 4 and 5 refer to the #3 stirrups. Crosssectional areas in sets 2 and 3 refer to the reinforcement of 3#5 bars. The material number refers to the type of material for the reinforcement. The volume ratio refers to the ratio of steel to concrete in the element. Each material corresponds to x. Values for crosssectional area and initial strain were entered.8c). 4. The orientation angles refer to the orientation of the reinforcement in the smeared model (Figure 2. Due to symmetry.
5 – Material Models For the Calibration Model Material Model Element Number Type Material Properties EX PRXY Linear Isotropic 3.0019 0.7 2233 3991 4656 4800 ShrCfOp ShrCfCl UnTensSt UnCompSt BiCompSt HydroPrs BiCompSt UnTensSt TenCrFac 0.00243 Concrete 1 Solid65 Point 1 Point 2 Point 3 Point 4 Point 5 Stress 1421.3 EX PRXY 3 Link8 Linear Isotropic 29.000 psi 0.3 Bilinear Isotropic Yield Stss 60.000.900 psi .000 psi 0.000 psi Tang Mod 2.Table 3.000.0013 0.0006 0.3 1 520 1 0 0 0 0 0 2 Solid45 EX PRXY Linear Isotropic 29.949.076 psi 0.00036 0.3 Multilinear Isotropic Strain 0.
3) f ε (3.1) with a value of f c' equal to 4. The Solid65 element requires linear isotropic and multilinear isotropic material properties to properly model concrete. Poisson’s ratio was assumed to be 0.2) ε0 = Ec = where: (3.4) f = stress at any strain ε .3. psi ε = strain at stress f ε 0 = strain at the ultimate compressive strength f c' The multilinear isotropic stressstrain implemented requires the first point of the curve to be defined by the user. .Material Model Number 1 refers to the Solid65 element.800 psi. EX is the modulus of elasticity of the concrete ( Ec ). Ec = 57000 f c' (3. The compressive uniaxial stressstrain relationship for the concrete model was obtained using the following equations to compute the multilinear isotropic stressstrain curve for the concrete (MacGregor 1992) f = Ecε ε 1+ ε0 2 f c' Ec 2 (3. and PRXY is the Poisson’s ratio (ν ). It must satisfy Hooke’s Law. The multilinear isotropic material uses the von Mises failure criterion along with the Willam and Warnke (1974) model to define the failure of the concrete. The modulus was based on the equation.
8 shows the stressstrain relationship used for this study and is based on work done by Kachlakev.001 0.30 f c' 1000 0 0 0. and 4 are calculated from Equation 3. defined as 0.0005 0.4).0025 0.E= σ ε (3. (2001).5) The multilinear curve is used to help with convergence of the nonlinear solution algorithm. Points 2. is calculated in the linear range (Equation 3.2 with ε 0 obtained from Equation 3. 6000 5000 f c' 4000 Stress (psi) Ultimate Compressive Strength 3 4 5 Ec 3000 2000 2 1 0. 3.) ε0 Strain at Ultimate Strength 0.3.0035 Figure 3. et al.0015 0. Point 1./in.30 f c' .8 – Uniaxial StressStrain Curve Figure 3.003 0. Strains were selected and the stress was calculated for each .002 Strain (in.
Shear transfer coefficients for a closed crack. Uniaxial crushing stress (positive).4). Shear transfer coefficients for an open crack. for unconfined concrete. 8. Biaxial crushing stress (positive) under the ambient hydrostatic stress state (constant 6). These 9 constants are: (SAS 2003) 1. 6.strain. Uniaxial crushing stress (positive) under the ambient hydrostatic stress state (constant 6). Point 5 is defined at f c' and ε 0 = 0. Stiffness multiplier for cracked tensile condition. . No deviation of the response occurs with the change of the coefficient. indicating traditional crushing strain Implementation of the Willam and Warnke (1974) material model in ANSYS requires that different constants be defined. Biaxial crushing stress (positive). Convergence problems occurred when the shear transfer coefficient for the open crack dropped below 0. 9. with 0. The uniaxial cracking stress was based upon the modulus of rupture.0 to 1. 4. 3. the coefficient for the open crack was set to 0.0. 5.003 in. This value is determined using. et al. Typical shear transfer coefficients range from 0.2. Ambient hydrostatic stress state for use with constants 7 and 8. The shear transfer coefficients for open and closed cracks were determined using the work of Kachlakev.0 representing a rough crack (no loss of shear transfer).3 (Table 3. Therefore. in. 7. 2.0 representing a smooth crack (complete loss of shear transfer) and 1. (2001) as a basis. Uniaxial tensile cracking stress.
ft and f c' . It was entered as 1 to turn off the crushing capability of the concrete element as suggested by past researchers (Kachlakev. 2001).f r = 7.725 f c' (3.8) f1 = 1. The remainder of the variables in the concrete model are left to default based on these equations: (SAS 2003) ' f cb = 1.7) where σ xp . and σ zp are the principal stresses in the principal directions.9) (3. The failure surface can be defined with a minimum of two constants.5 f c' (3. The biaxial crushing stress under the ambient hydrostatic stress state refers to the ultimate compressive strength for a state of biaxial compression superimposed on the hydrostatic stress state ( f1 ).10) . et al.2 f c' (3.45 f c' f 2 = 1. The ambient hydrostatic stress state is denoted as σ h .6) The uniaxial crushing stress in this model was based on the uniaxial unconfined compressive strength ( f c' ) and is denoted as ft . The uniaxial crushing stress under the ambient hydrostatic stress state refers to the ultimate compressive strength for a state of uniaxial compression superimposed on the hydrostatic stress state ( f 2 ). The biaxial crushing stress refers to the ultimate biaxial compressive strength ' ( f cb ). This stress state is defined as: σ h = (σ xp + σ yp + σ zp ) 1 3 (3. Convergence problems have been repeated when the crushing capability was turned on. σ yp .
These stress states are only valid for stress states satisfying the condition σ h ≤ 3 f c' (3. and the hardening modulus was 2900 psi. plates. and supports were modeled as volumes.11) Material Model Number 2 refers to the Solid45 element. x 18 in. as well as the hardening modulus of the steel to be defined. 3.3). Material Model Number 3 refers to the Link8 element. Since a quarter of the beam is being modeled. the model is 93 in. The dimensions for the concrete volume are shown in Table 3. long. Therefore. Therefore.2. and poisson’s ratio (0. with a crosssection of 5 in.000 psi. the selfweight was not introduced in this calibration model. Bilinear isotropic material is also based on the von Mises failure criteria. Note that the density for the concrete was not added in the material model.4 Modeling The beam. the LVDT’s used to measure deflection at midspan were put on the beam after it was set in the test fixture. The bilinear model requires the yield stress ( f y ). The zero values for the Zcoordinates coincide with the center of the crosssection for the concrete beam. this element is modeled as a linear isotropic element with a modulus of elasticity for the steel ( Es ). The Link8 element is being used for all the steel reinforcement in the beam and it is assumed to be bilinear isotropic. The Solid45 element is being used for the steel plates at loading points and supports on the beam. The yield stress was defined as 60.6. . For the control beam in Buckhouse (1997). Deflections were taken relative to a zero deflection point after the selfweight was introduced.
) 0 93 0 18 0 5 Steel Plate (in.) 60 66 18 19 0 5 Steel Support (in.) 1. and Steel Support Volumes ANSYS X1. x 5 in.Table 3. The dimensions for the plate and support are shown in Table 3. x 1 in. Steel Plate. x 1 in.X2 Xcoordinates Y1.5 4.5 0 1 0 5 The 93 in.6.9. x 5 in. steel plate.10. while the plate at the load point is 6 in. The FE mesh for the beam model is shown in Figure 3. Due to symmetry.9 – Volumes Created in ANSYS . only one loading plate and one support plate are needed.6 – Dimensions for Concrete. Concrete Beam Steel Loading Plate Steel Support Figure 3. dimension for the Xcoordinates is the midspan of the beam. The combined volumes of the plate. and beam are shown in Figure 3.Z2 Zcoordinates Concrete (in.Y2 Ycoordinates Z1. support. The support is a 3 in.
5 in.25 in.25 in. The area of steel at the plane of symmetry is one half the normal area for a #5 bar because one half of the bar is cut off.5 in. Shear stirrups are modeled throughout the beam. Reinforcement exists at a plane of symmetry and in the beam. Steel Support Element Width 1.7.10 – Mesh of the Concrete. Steel Plate. Steel Plate Element Length 1. Figure 3.5 in. . Steel Plate Element Width 1. Concrete Element Height 1. The element type number.11 illustrates that the rebar shares the same nodes at the points that it intersects the shear stirrups. and real constant set number for the calibration model were set for each mesh as shown in Table 3. Steel Support Element Length 1. Figure 3.25 in. Only half of the stirrup is modeled because of the symmetry of the beam. and Steel Support Link8 elements were used to create the flexural and shear reinforcement. Concrete Element Length 1. material number.2 in.Concrete Element Width 1.
#3 Shear Stirrups #5 Bar Reinforcement at Plane of Symmetry Stirrup at Plane of Symmetry #5 Bar Reinforcement located 2. from the end of the CrossSection Shared nodes of Stirrups and Rebar Figure 3. Therefore.7 – Mesh Attributes for the Model Model Parts Concrete Beam Steel Plate Steel Support Rebar at Center of CrossSection Rebar 2.5 Meshing To obtain good results from the Solid65 element.5 in.11 – Reinforcement Configuration Table 3.2. the use of a rectangular mesh is recommended. the mesh was set up such that square or rectangular elements . of CrossSection Stirrup at Center of Beam Other Stirrups Element Material Type Number 1 2 2 3 3 3 3 1 3 3 2 2 2 2 Real Constant Set 1 N/A N/A 3 2 5 4 3.5 in.
The meshing of the reinforcement is a special case compared to the volumes. Caution must be taken when merging entities in a model that has already been meshed because the order in which merging occurs is significant. Care must be taken to always merge in the order that the entities appear. 3. plate. . and so on) to fail. surface load transfers.10. The necessary element divisions are noted. and support volumes is shown in Figure 3.were created (Figure 3. However. the nodes lose their association with the solid model.2.10). All precautions were taken to ensure that everything was merged in the proper order. Merging keypoints before nodes can result in some of the nodes becoming “orphaned”. The overall mesh of the concrete. the lowest number was retained during merging. the necessary mesh attributes as described above need to be set before each section of the reinforcement is created. These items will then be merged into single entities. The orphaned nodes can cause certain operations (such as boundary condition transfers.6 Numbering Controls The command merge items merges separate entities that have the same location. This properly sets the width and length of elements in the plates to be consistent with the elements and nodes in the concrete portions of the model. that is. Also. No mesh of the reinforcement is needed because individual elements were created in the modeling through the nodes created by the mesh of the concrete volume. The volume sweep command was used to mesh the steel plate and support.
7 Loads and Boundary Conditions Displacement boundary conditions are needed to constrain the model to get a unique solution. The symmetry boundary conditions were set first.3. and where the supports and loadings exist. Constraint in the zdirection Constraint in the xdirection Figure 3. boundary conditions need to be applied at points of symmetry. To ensure that the model acts the same way as the experimental beam.12.12 – Boundary Conditions for Planes of Symmetry .2. The boundary conditions for both planes of symmetry are shown in Figure 3. The model being used is symmetric about two planes.
The support condition is shown in Figure 3. By doing this. These nodes were given the constraint UZ = 0. applied as constant values of 0. Support roller condition to allow rotation Figure 3. all nodes selected at Z = 0 define another plane of symmetry. therefore.13 – Boundary Condition for Support .13. A single line of nodes on the plate were given constraint in the UY. have a degree of freedom constraint UX = 0. and UZ directions. These nodes.Nodes defining a vertical plane through the beam crosssection centroid defines a plane of symmetry. The support was modeled in such a way that a roller was created. nodes on this plane must be constrained in the perpendicular direction. Second. the beam will be allowed to rotate at the support. To model the symmetry.
. applied at the steel plate is applied across the entire centerline of the plate. Figure 3.The force. For the purposes of this model. P. The Restart command is utilized to restart an analysis after the initial run or load step has been completed. Loading Applied on the Plate Boundary Conditions at Plate Figure 3. The use of the restart option will be detailed in the analysis portion of the discussion. The force applied at each node on the plate is one tenth of the actual force applied.14 illustrates the plate and applied loading.14 – Boundary Conditions at the Loading Plate 3.2. the Static analysis type is utilized.8 Analysis Type The finite element model for this analysis is a simple beam under transverse loading.
8 shows the first load step taken (e.8. Typical commands utilized in a nonlinear static analysis are shown in Table 3. The commands used for the nonlinear algorithm and convergence criteria are shown in Table 3.The Sol’n Controls command dictates the use of a linear or nonlinear solution for the finite element model.10. The commands used to control the solver and output are shown in Table 3. Table 3. Table 3. All values for the nonlinear algorithm are set to defaults. The time at the end of the load step refers to the ending load per load step. Table 3.9. of Substeps 1 Write Items to Results File All Solution Items Frequency Write Every Substep In the particular case considered in this thesis the analysis is small displacement and static.8 – Commands Used to Control Nonlinear Analysis Analysis Options Small Displacement Calculate Prestress Effects No Time at End of Loadstep 5120 Automatic Time Stepping On Number of Substeps 1 Max no. of Substeps 2 Min no. .g. up to first cracking). The sub steps are set to indicate load increments used for this analysis.9 – Commands Used to Control Output Equation Solvers Sparse Direct Number of Restart Files 1 Frequency Write Every Substep All these values are set to ANSYS (SAS 2003) defaults.
Plastic Strain 0. The program behavior upon nonconvergence for this analysis was set such that the program will terminate but not exit. The rest of the commands were set to defaults.005 0.11 shows the commands used for the advanced nonlinear settings. Table 3. Table 3.11 – Advanced Nonlinear Control Settings Used Program behavior upon nonconvergence Terminate but do not exit Nodal DOF sol'n 0 Cumulative iter 0 Elapsed time 0 CPU time 0 . The tolerances for force and displacement are set as 5 times the default values.10 – Nonlinear Algorithm and Convergence Criteria Parameters Line Search Off DOF solution predictor Prog Chosen Maximum number of iteration 100 Cutback Control Cutback according to predicted number of iter. Equiv.05 Norm L2 L2 Min.Table 3.15 Explicit Creep ratio 0. not applicable not applicable The values for the convergence criteria are set to defaults except for the tolerances.1 Implicit Creep ratio 0 Incremental displacement 10000000 Points per cycle 13 Set Convergence Criteria Label F U Ref. Value Calculated calculated Tolerance 0. Ref.
the restart option was used to go to the next step after convergence.) 5210 5220 5300 5400 10000 13000 14000 14500 14700 14800 14900 15000 15100 15200 15300 15600 15900 16200 16300 16382 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 1 10 16 20 92 30 10 50 20 20 100 10 10 50 20 150 150 150 50 41 5210 10 5 5 50 100 100 10 10 5 1 10 10 2 5 2 2 2 2 2 . and the strength limit state of the beam. A listing of the load steps. The application of the loads up to failure was done incrementally as required by the NewtonRaphson procedure. yielding of the steel reinforcement.12 – Load Increment for Analysis of Finite Element Model Beginning Time 0 5210 5220 5300 5400 10000 13000 14000 14500 14700 14800 14900 15000 15100 15200 15300 15600 15900 16200 16300 Load Time at End Load Step Sub Step Increment of Loadstep (lbs.9 Analysis Process for the Finite Element Model The FE analysis of the model was set up to examine three different behaviors: initial cracking of the beam. The NewtonRaphson method of analysis was used to compute the nonlinear response. and loads applied per restart file are shown in Table 3. After each load increment was applied.12.2. sub steps. Table 3.3.
For the first load step the time at the end of the load step is 5210 referring to a load of.12. but the forces did not. the convergence criteria for force was dropped and the reference value for the Displacement criteria was changed to 5. the steps taken to the initial cracking of the beam can be decresed to one load increment to model/capture initial cracking. the load increments increased slightly until subsequent cracking of the beam (14. Yielding of the steel occurs at load step 13. therefore.400. This criteria was used for the remainder of the analysis. P.12. convergence for the nonlinear analysis was impossible with the default values. These criteria were left at the default values up to 5210 lbs. A small criterion must be used to capture correct response. The displacements converged. Failure of the beam occurs when convergence fails.000 lbs) as seen in Table 3. The two convergence criteria used for the analysis were Force and Displacement. Once the yielding of the reinforcing steel is reached. However. the load increment size is decreased to 2 lb. Therefore.05 to produce a criterion of 0. of 5210 lbs applied at the steel plate.25 during the nonlinear solution for convergence. . the load increment sizes begin decreasing further because displacements are increasing more rapidly. when the beam began cracking. Eventually. with this very small load increment.The time at the end of each load step corresponds to the loading applied. This value is then multiplied by the tolerance value of 0. to capture the failure of the beam. The load deformation trace produced by the analysis confirmed the failure load. the load increments must be decreased again. As shown in Table 3. Once initial cracking of the beam has been passed (5220 lbs).
Figure 3. A comparison of values obtained from the FE model and Appendix A can be seen in Table 3.4 shows the different components that were analyzed for comparison: the linear region. The load at step 5210 was analyzed and it coincides with a load of 5210 lbs. real constants and convergence criteria are adequate to model the response of the member.13 indicate that the FE analysis of the beam prior to cracking is acceptable. material properties.13. applied on the steel plate.3. and failure. can be seen in Appendix A. 3. Comparisons were made in this region to ensure deflections and stresses were consistent with the FE model and the beam before cracking occurred. .3 Results The goal of the comparison of the FE model and the beam from Buckhouse (1997) is to ensure that the elements. deflections and stresses become more difficult to predict. Calculations to obtain the concrete stress. This is where we expect the maximums to occur.3. The results in Table 3. steel stress and deflection of the beam at a load of 5210 lbs.1 Behavior at First Cracking The analysis of the linear region can be based on the design for flexure given in MacGregor (1992) for a reinforced concrete beam. The maximums exist in the constantmoment region of the beam during load application. The stresses in the concrete and steel immediately preceding initial cracking were analyzed. Once cracking occurs. the nonlinear region. initial cracking. the yielding of steel.
0529 0. Cracking increases in the constant moment region.3. subsequent cracking occurs as more load is applied to the beam. This compares well with the load of 5118 lbs calculated in Appendix A. and is a flexural crack. Vector Mode plots must be turned on to view the cracking in the model.13. P. The stress increases up to 537 psi at the centerline when the first crack occurs. and the beam begins cracking out towards the supports at a load of 8.13 – Deflection and Stress Comparisons At First Cracking Model HandCalculations ANSYS Reinforcing Centerline Load at First Extreme Tension Steel Stress Deflection Cracking Fiber Stress (psi) (psi) (in. . This first crack occurs in the constant moment region.3 Behavior Beyond First Cracking In the nonlinear region of the response. 3. Buckhouse (1997) reported first cracking at a load.) (lbs.2 Behavior at Initial Cracking The cracking pattern(s) in the beam can be obtained using the Crack/Crushing plot option in ANSYS (SAS 2003).3.) 530 536 3024 2840 0.0534 5118 5216 3. The first crack can be seen in Figure 3. of 4500 lbs using visual detection. The initial cracking of the beam in the FE model corresponds to a load of 5216 lbs that creates stress just beyond the modulus of rupture of the concrete (520 psi) as shown in Table 3.000 lbs.15.Table 3.
and nonlinear concrete material. Therefore. This cracking can be seen in Figure 3. The ability of the beam to distribute load throughout the crosssection has diminished greatly.16. 3.400 lbs. . the displacements of the beam begin to increase at a higher rate as more load is applied (Figure 3. diagonal tension cracks are beginning to form in the model.3. At this point in the response. is applied.4 Behavior of Reinforcement Yielding and Beyond Yielding of steel reinforcement occurs when a force of 13. now defines the flexural rigidity of the member.1st Crack in Concrete Beam Figure 3.15 – 1st Crack of the Concrete Model Significant flexural cracking occurs in the beam at 12. yielding steel.000 lbs. greater deflections occur at the beam centerline.4). The cracked moment of inertia. Also.
000 and 12.. and diagonal tension cracks.000 lbs.17 shows successive cracking of the concrete beam after yielding of the steel occurs. and failure is soon to follow. more cracks have now formed in the constant moment region.Flexural Cracks Diagonal Tension Cracks Figure 3. At 16.16 – Cracking at 8.000 lbs.. . At 15.000 lbs. the beam has increasing flexural cracks. cracking has reached the top of the beam. Figure 3. Also.
3.18). The deflections at the analytical failure load of the control beam were compared with the finite element model as shown in Table 3. .382 lbs. Severe cracking throughout the entire constant moment region occurs (see Figure 3.14.Multiple cracks occurring Increasing Diagonal Tension Cracks Figure 3.17 – Increased Cracking After Yielding of Reinforcement 3.. the beam no longer can support additional load as indicated by an insurmountable convergence failure.5 Strength Limit State At load 16.
Finite Element Model At Ultimate Load Centerline Load (lb.18 – Failure of the Concrete Beam Table 3. .) C1 16310 3.65 ANSYS 16310 3.14 – Deflections of Control Beam (Buckhouse 1997) vs.Excessive cracking and of the beam in the constant moment region Diagonal Tension Cracking Figure 3.586 Beam The deflection of the finite element model was within 2% of the control beam at the same load at which the control beam failed.) Deflection (in.
19.) C1 theoretical ultimate load (14.5 3 3.) 10000 8000 6000 4000 2000 0 0 0. 18000 FEA 16000 14000 12000 Avg.600 lbs.6 LoadDeformation Response The full nonlinear loaddeformation response can be seen in Figure 3. The approach was then utilized to analyze a prestressed concrete beam.19 – Load vs.3. P (lbs. This response was calibrated by setting the tolerances so that the loaddeformation curve fits to the curve from Buckhouse (1997).3.5 Avg.) Buckhouse (1997) Figure 3. This gave confidence in the use of ANSYS (SAS 2003) and the model developed. Centerline Deflection (in. The response calculated using FEA is plotted upon the experimental response from Buckhouse (1997). The entire loaddeformation response of the model produced compares well with the response from Buckhouse (1997). Load.5 1 1.5 2 2.5 4 4. Deflection Curve Comparison of ANSYS and Buckhouse (1997) .
These steps include definition of real constants.0 Introduction This chapter discusses the finite element modeling of a prestressed concrete beam.CHAPTER 4 PRESTRESSED CONCRETE BEAM MODEL 4. In this case. The beam was prestressed using ½ in.1 Finite Element Model The finite element model that was used for analysis of the prestressed concrete beam is very similar to the calibration model. The shear stirrups are included as discussed in the calibration model. . However. no mildsteel flexural reinforcement was used. Many of the steps taken to model the prestressed concrete beam were the same as the calibration model and can be found in chapter 3. The rest of this chapter will discuss the steps taken that were different from those in the calibration model. 4. The prestressed beam has the same dimensions as the experimental beam modeled by Buckhouse (1997). and the parameters for the nonlinear analysis. All the necessary steps taken to create the model and the analysis used for the prestressed beam are explained in detail. diameter 270 ksi 7wire strands as opposed to the #5 60 ksi mildsteel reinforcement. material properties. the reinforcement in the beam is different.
153 0. and an initial strain due to prestressing is now added to the element./in.1. Also.) Crosssectional Area (in. The initial strains for each real constant set were determined from the effective prestress ( f pe ) and the modulus of elasticity ( E ps ). the constants for the Link8 element designated to the stirrups used in the model were left unchanged./in. because it is at a point of symmetry on the model.4.1 – Real Constants for Prestressed Beam Real Constant Set Element Type Constants Crosssectional Area (in.2) 3 Link8 Initial Strain (in.001903 0. the real constants for the Link8 element for the flexural reinforcement has changed. The crosssectional area for real constant set 3 is half of real constant set 2. However. Since the beam is now using prestressing steel.0765 2 Link8 The crosssectional area for real constant set 2 is the area of an equivalent ½ in.) 0.1.1 Real Constants The real constants for the concrete element were left untouched because the modeling of the concrete has not changed. diameter 7wire strand.001903 0. An example of this can be seen in Appendix B. Table 4.2) Initial Strain (in. the crosssectional area of the steel has changed. It was . This prestress level is too low and not suitable for practical use. The change in area and strain is shown in Table 4.
000ε ps (ksi ) 0.1. 4. The addition of the selfweight is done by applying gravitational .2) f ps = 268 − The values entered into ANSYS (SAS 2003) for the stressstrain curve are given in Table 4.98f pu (ksi ) ε ps − 0. The only changes made were as follows. Density was added to the concrete material property so the selfweight of the concrete beam could be taken into account.utilized to prevent any convergence problems occurring from bursting and significant cracking near the support as the prestrain is applied. Second.0065 (4. 4.1 shows the stressstrain behavior of the prestressing steel. and steel support plates are the same. no point loads are applied in the first load step. steel plates at loading points.2 and Figure 4. The prestressing steel was modeled using a multilinear stressstrain curve developed using the following equations.075 <0.008: f ps = 28. First. due to the fact that the initial prestrain is applied. The material properties for the prestressing steel have been changed from bilinear isotropic to multilinear isotropic following the von Mises failure criteria. the selfweight was added in a load step. the time at the end of the prestrain application load step is 1.008 : ε ps >0. Third.1.3 Solution Most of the parameters used to control the solution are the same as those used for the calibration model. ε ps ≤ 0.1) (4.2 Material Properties The material properties for the shear reinforcement steel.
0099 Stress 0 224 226.0085 0.1111 254.0115 0.0151 0.0143 Stress 255.0171 0.0117 0.0133 0.8537 259.375 253 253.1 – StressStrain Curve for 270 ksi strand .822 300 250 200 Stress (ksi) 150 100 50 0 0 0.3846 Strain 0.9545 251.015 0.2791 260.0111 0.0103 0.5833 257.2857 257. A mass density was entered in as a function of the gravitational acceleration (Density/g).5 233.6957 252.9091 236.5 255.0131 0.0097 0.1667 248.0149 0.0135 0.0107 0.9032 256.0147 0.2 – Values for Multilinear Isotropic StressStrain Curve Strain 0 0.02 0.0109 0.0093 0.1429 250./in.625 258.0141 0.3333 230.1316 258.035 Strain (in.4 in/s2 in the global ydirection.0101 0.9516 263 264.0105 0.9245 261.5625 245.6364 256.acceleration of 386.9706 257.069 255.0125 0.03 0.2632 249.0137 0.0129 0.0091 0.134 264.) Figure 4.9412 Strain 0.025 0.25 250.0083 0. Table 4.0259 0.0113 0.0095 0.0145 0.0301 Stress 258.0215 0.75 239.0714 259.2813 256.5769 254.2143 243 244.0119 0.8649 258.0089 0.01 0.0127 0.0123 0.0189 0.005 0.0087 0.0121 Stress 247.1538 241.6071 Strain 0.008 0.0139 0.
A listing of the load steps. yielding of the prestress steel ( 0. From this point.1.4. to capture the initial cracking of the beam.4 Analysis Process for the Finite Element Model Analysis for the prestressed concrete beam was very similar to the calibration model. The second load step was the addition of the selfweight.85 f pu ). different load step and restart points were used.3. and the reference value of the displacement criteria is 5. Finally. The first load step taken was to produce the camber in the concrete beam due to the prestress. From that point on. When yielding of the steel occurs.4 7840 2 50 100 100 50 10 10 2 When the analysis reaches the point of initial cracking. the force convergence criteria is dropped. . load increments are decreased to 2 lb. sub steps. and eventual failure. the load increments are decreased to 10 lb. Table 4.3 – Load Increments for the Analysis of the Prestressed Beam Model Beginning Time 0 1 2 7840 7900 8500 10000 20000 25000 27000 28000 Time at Load End of Load Step Sub Step Increment Loadstep (lbs. However.) 1 2 7840 7900 8500 10000 20000 25000 27000 28000 28823 1 2 3 4 5 6 7 8 9 10 11 1 1 1 30 12 15 100 100 200 100 412 Prestress 386. and loads applied per restart file are shown in Table 4. the concrete was modeled to the point of cracking. the load increments are decreased to 100 lbs. until unresolvable convergence failure of the nonlinear algorithm occurs.
initial cracking.823 .85 f pu ).4 – Analytical Results ANSYS Centerline Deflection at Effective Prestress (in.) Top Fiber Stress at SelfWeight (psi) Zero Deflection Load (lbs.04 0. these distinct points are: effective prestress. decompression.520 27. Table 4.0341 163.) Top Fiber Stress at Effective Prestress (psi) 0.2.1 Application of Effective Prestress Calculation of the effective prestress for the beam can be found in Appendix B.05 Hand Calculations 0.587 Deflection at Application 0.97 2055 2817 7850 Prestress at Failure Load 264. The deflections computed using hand calculations and FE analysis due to the prestress force is shown in Table 4. and failure.) 36.0211 of SelfWeight (in. steel yielding ( 0.4.2 Results The analysis yielded results for seven distinct points on the loaddeflection application of the prestressed concrete beam and the full nonlinear loaddeformation response.) Decompression Load (lbs. As seen in Figure 4.0229 45.) Initial Cracking Load (lbs. zero deflection.822 (psi) Failure Load (lbs. addition of selfweight. 4.0323 156.) 28.86 2126 2858 7538 254.4.2.
04 0.) 0. P (lb s .50 Linear Region 3.02 0.00 Decompression Zero Deflection 0.04 0.50 0 0.) Figure 4.) 15000 Steel Yielding Initial Cracking Secondary Linear Region 10000 5000 SEE FIGURE BELOW 0.00 5000 Centerline Deflection (in.2 – Load vs.02 0.10 Centerline Deflection (in.35000 30000 25000 Failure 20000 Load. Deflection Curve for Prestressed Concrete Model .06 0.00 2.00 3.08 0.50 1.00 1. P (lbs.00 8000 7000 L o a d .50 4.50 2.) 6000 5000 4000 SelfWeight Prestressing Initial Cracking 3000 2000 1000 0 0.
The contour plot shows that at the time of prestressing.4. This phenomenon can be seen in Figure 4. A phenomenon that occurred when the prestress was added was bursting in the concrete where the prestress force is being applied.3). there is a maximum . Since camber is occurring in the beam (Figure 4.3 – Deflection due to prestress The handcalculated top fiber stress and the top fiber stress found using FEA are also shown in Table 4.4. the controlling stress exists in the extreme top fiber. The values correlate very well. The stresses in the extreme fibers were looked at in the beam also.) Camber due to prestressing force Figure 4. Deflection (in.The deflection in ANSYS (SAS 2003) corresponds well to the calculated value.
bursting zone cracking was extensive – a converged solution was not possible to obtain. However.5.stress in the concrete where bursting occurs. Bursting Effect at Center of Beam Tendon Level Figure 4. The bursting phenomenon is explained in detail in Nawy (2000). When a traditional level of prestress was applied ( f pe = 159. this does not impact the solution because no other cracking occurs in that area after the initial application of prestress. localized cracking occurs in the concrete.840 psi ). This is the reason for the reduction in f pe by the factor of 3. .4 – Bursting Phenomenon Also. as shown in Figure 4.
4 indicates very good correlation between the hand calculated values and the FEA.Cracking that occurs from bursting Figure 4. 4.4).2.5 – Localized Cracking From Effective Prestress Application 4.2. Table 4.3 Zero Deflection Handcalculations (Appendix B) determined the load at which the deflection of the centerline was zero.2 SelfWeight Adding the selfweight gives a deflection value which corresponds well to the calculated deflection in Appendix B (Table 4. The FEA calculated top fiber stress at this stage also correlates well with handcalculations. .
4. diagonal tension cracks are beginning to form in the model.2. Additional flexural cracking occurs in the beam at 20. (Table 4. significantly more cracking occurs as more load is applied the beam as seen in Figure 4.4 indicates very good correlation at this level of loading as well.6. 4.2. This is just past the modulus of rupture of the beam of 520 psi.2.6 Secondary Linear Region In the secondary linear region of the response. This first crack occurs in the constant moment region. Also. Table 4. Cracking increases in the constant moment region. 4.5 Initial Cracking Initial cracking is defined to be the loading at which the extreme tension fiber reaches the modulus of rupture.4 Decompression One definition of decompression denotes the point of loading where the stress at the bottom fiber of the concrete beam moves from compression due to the prestress to tension from superimposed loading.4 and Appendix B). Initial cracking of the beam in the FE model occurs at load 7850 lbs. The handcalculated load where cracking occurs is 7538 lbs. .000 lbs. and the beam begins cracking out towards the supports with at a load of 12.000 lbs. and is a flexural crack. The stress increases up to 539 psi at the centerline when the first crack occurs.
000 lbs.Bursting Cracks Flexural Cracks Diagonal Tension Cracks Figure 4. Figure 4. the displacements of the beam begin to increase at a higher rate as more load is applied.85 f pu for this model. 4.6 – Cracking at 12.000 and 20.2 illustrates the load level at the point of yielding of the prestress steel. At this point in the response.7 Behavior of Steel Yielding and Beyond Yielding of the prestress steel is defined as 0.2. The ability of the beam to distribute load throughout the crosssection has diminished .
2. indicating that cracking throughout the entire constant moment region has occurred. unresolvable nonconvergence of the nonlinear algorithm occurs.6).greatly.7 illustrates the excessive cracking in the beam.7 – Cracking at Flexural Capacity .8 Flexural Limit State At a load of 28. Figure 4. greater deflections occur at the beam centerline.823 lbs. Cracking at Failure Figure 4. Therefore. Increasing flexural cracks and diagonal tension cracks form as the beam approaches failure (Figure 4. 4.
in Table 4. (near doubling). The FEA prediction (28.Hand calculations (Appendix B) predicted that the flexural capacity of the beam would correspond to 27.4). Using strain compatibility the stress in the prestressing steel at failure was 254. which corresponds well to the FEA prediction. f pe 3 .4). (Table 4.823 lbs. .822 psi (Table 4. assumed in the present case increased the limit load for the beam from 16.310 lbs.823 lbs.587 lbs.520 psi (Appendix B). (Chapter 3) to 28. The stress in the prestressing steel at failure predicted using FEA was 264. The benefits of prestressing are apparent.4) corresponds very well with the hand calculations. The application of the moderate prestressing.
decompression. yielding of the steel and flexural failure of the beam were compared to the experimental results. A reinforced concrete beam model was calibrated to experimental data. yielding of steel. and flexural failure were then studied and compared to theoretical values obtained via accepted methods of hand calculation.1 Conclusions The following conclusions can be stated based on the evaluation of the analyses of the calibration model and the prestressed concrete beam.CHAPTER 5 CONCLUSIONS AND RECOMMENDATIONS 5. and predictions of initial cracking. . (2) The failure mechanism of a reinforced concrete beam is modeled quite well using FEA. Initial prestress. zero deflection point. (1) Deflections and stresses at the centerline along with initial and progressive cracking of the finite element model compare well to experimental data obtained from a reinforced concrete beam. A prestressed concrete beam model was also created.0 Introduction The use of the finite element method to analyze reinforced and prestressed concrete beams was evaluated. 5. initial cracking. addition of selfweight. and the failure load predicted is very close to the failure load measured during experimental testing.
creep. Many prestressed structural components have nonsymmetric geometries and loadings. and the load applied at failure is very close to hand calculated results. nonsymmetric geometries and loadings should be analyzed using finite element analysis with . (6) Flexural failure of the prestressed concrete beam is modeled well using a finite element package. any model should be calibrated with good experimental data.2 Recommendations for Future Work The literature review and analysis procedure utilized in this thesis has provided useful insight for future application of a finite element package as a method of analysis. 5. Also. To ensure that the finite element model is producing results that can be used for study. camber due to the initial prestress force and after application of the selfweight of the beam compares well to handcomputed values. and elastic shortening were lumped together in a single load step. (5) The load applied to cause initial cracking of the prestressed concrete beam compares well with hand calculations. Therefore. Individual modeling of these losses could be included in future research. While modeling the prestressed beam. shrinkage. (4) Deflections and stresses at the zero deflection point and decompression are modeled well using a finite element package.(3) For the prestressed concrete beam. relaxation losses due to prestress. a bursting effect was seen in the FE model. This will then provide the proper modeling parameters needed for later use.
prestress for further study. . the bursting effect. Higher strength concrete. and the transfer length of the prestressing steel are all candidates for future research.
E. T. (2000). Dublin 2. Miller. Ed. “An Analysis of a Reinforced and a Prestressed Concrete Beam by Finite Element Method. Prestressed Concrete: A Fundamental Approach. (1992). Milwaukee. NJ. PrenticeHall.. “Finite Element Modeling of Reinforced Concrete Structures Strengthened With FRP Laminates. and Kachlakev. (2001).L.” Master’s Thesis. MacGregor. PrenticeHall.G.. J. D. Meyers. Faherty. Inc.. Potisuk. Ireland. “Nature of Bond in Pretensioned Prestressed Concrete. (1972).1 Finite Element Analysis System.I (2000).M. Proceedings. Jr. Shing. J. (2001). Douglas McHenry International Symposium on Concrete and Concrete Structures. Detroit. Branson. Aug. “Loss of Prestress. E. S. March. May. (1997). T. 706.. and Kripanarayanan. (2001). Kachlakev. “Strengthening of Full Sized Reinforced Concrete Beam Using FRP Laminates and Monitoring with Fiber Optic Strain Guages” in Innovative Systems for Seismic Repair and Rehabilitation of Structures. Modeling of Inelastic Behavior of RC Structures Under Seismic Loads. Technomic Publishing Co. K.REFERENCES American Concrete Institute (1978). Philadelphia. CA and Oregon State University. Yim.R. Marquette University. San Luis Obispo.5. Englewood Cliffs. Inc. Sept. Fanning. P. May.R.12. Upper Saddle River. Design and Applications. Earlsfort Terrace. No.. D.50. “External Flexural Reinforcement of Existing Reinforced Concrete Beams Using Bolted Steel Channels. University of Iowa..” Doctorate’s Thesis. B. Report No.F. Wisconsin. E. Inc.G. Inc.. P. University College Dublin..” Iowa State Highway Comission. Buckhouse. D.” Electronic Journal of Structural Engineering... K. T. (1970). (1954). Corvallis.” California Polytechnic State University.. Janney. Vol. and Tanabe.” Journal of the ACI. K. Reinforced Concrete Mechanics and Design. . Michigan. Nawy. Camber and Deflection of Noncomposite and Composite Structures Using Different Weight Concrete... American Society of Civil Engineers. McCurry. NJ SAS (2003) ANSYS 7. Chansawat. OR for Oregon Department of Transportation. “Nonlinear Models of Reinforced and Posttensioned Concrete Beams. PA. D.I. American Concrete Institute.B. Iowa City. SAS IP.A.
T. Finite Element Analysis of Reinforced Concrete Structures..P. MI.A.174.. Ed. F. Willam. American Concrete Institute.” Master’s Thesis. (2001). Bergamo. K.A. . E. University of WisconsinMadison. p.J. (2001). “Constitutive Model for Triaxial Behaviour of Concrete. Willam. Italy. and Tanabe. Farmington Hills. and Warnke..Tavarez. K. (1974).” Seminar on Concrete Structures Subjected to Triaxial Stresses. Wisconsin. “Simulation of Behavior of Composite Grid Reinforced Concrete Beams Using Explicit Finite Element Methods. Madison. International Association of Bridge and Structural Engineering Conference.
. . . . .APPENDICES The following appendices have been included for reference: PAGE Appendix A: Theoretical Calculations for Calibration Model . . . . . 73 . . . . . . 69 Appendix B: Theoretical Calculations for Prestressed Model . . . . . . . . . . . . . . . . .
Appendix A .
1 – Loading of Beam with Supports Maximum Moment The moment that occurs from the existing forces M max = (5210lbs. 5210 lbs. 076 psi The modulus of rupture f r = 7. 000 4800 = 3.)(18in.949.5 f c' = 7. 5210 lbs.) = 312.)3 = 4860in 4 12 12 The modulus of elasticity of the concrete Ec = 57. 600lb. 000 f c' = 57. Material Properties The gross moment of inertia IG = 1 3 1 bh = (10in. 60” 3” 60” 60” 3” Figure A. − in.)(60in.Analysis of reinforced concrete beam for flexure at applied load of 5210 lbs.5 4800 = 520 psi .
44in 2 3.72in 2 distributed on each side of the concrete crosssection Calculate the distance from the top fiber to the neutral axis of the transformed moment of inertia y= A1 y1 + A2 y2 (18in.) 2 = 5162in.) + (2)(3.31) = 7.72in.4 + (10in.) = = 9.)(9in.242in.)(18in.2 )(6.)(10in.2 )(15.)(10in.4 The stress at the extreme tension fiber is then calculated .5in.5” Ybar 18” 3.) 2 + (2)(3.72 in2 Figure A.72in.72in. (18in.258in.Stresses in Concrete and Steel The stresses at the extreme tension fiber are calculated using a transformed moment of inertia of the concrete and steel reinforcement 10” 15.) + (2)(3.2 – Transformed CrossSection Transformed area of steel ( As )t = 8 As = 8(3)(0.2 ) A1 + A2 The transformed moment of inertia 2 I tr = I G + Ad 2 Concrete + Ad Steel = 4860in.258in.)(0.
∆ max = Pa (3l 2 − 4a 2 ) 24 EI tr (5210lb.)(8.)(8.4 The stress in the steel at this point is calculated fs = Myb (312. . 4 24(3.f ct = Myb (312.4 520 psi = P = 5117 lbs.) 2 − 4(60in.) 5162in.) (8) = 3024 psi η= 5162in.4 I tr Deflections The deflection at the centerline of the concrete beam at load 5210 lbs.) (3(180in. − in.)(6. 600lb.949. 076)(5162in.742in.) = = 530 psi I tr 5162in.)(60in.242in.742in.) 2 ) = 0. 600lb.0529 in. ) = Loads The load at first cracking f ct = Myb I tr P(60in. − in.
Appendix B
Analysis of Deflection for the Prestressed Concrete Beam
cgc P ec cgs P
60” 3”
60”
60” 3”
Figure B.1 – Typical Prestressed Concrete Beam with Supports
Effective Prestress
f pu = 270, 000 psi f pi = 0.74 f pu = 0.74(270, 000 psi ) = 199,800 psi f pe = 0.80 f pi = 0.80(199,800 psi ) = 159,840 psi
159,840 psi / 3 = 53,280 psi
Deflections
Deflection due to prestress
δc = −
Pel 2 (24, 456lb.)(6.6in.)(180in.) 2 =− = 0.03406 in. 8Ec I c 8(3,949, 076 psi )(4860in.4 )
Deflection due to selfweight
δ=
5wl 4 5(15.624 pli)(180in.) 4 = = 0.1113 in. 384 EI 384(3,949, 076 psi )(4860)
Deflection after prestress and selfweight is applied
δ net = δ c + δ sw = −0.03406 + 0.01113 = −0.02293in.
Stresses
Stress in extreme top fiber of the beam after initial prestress
ft =−
Pe ec 24, 456lb. (6.6in.)(9in.) (1 − 2t ) = − (1 − ) = 163.04 psi (T) 2 180in. 27in.2 Ac r
< f c = 0.45 f c' = 0.45(4800 psi ) = 2160 psi (T ), O.K .
Stress in extreme top fiber of the beam after selfweight
ft =−
Pe ec 24, 456lb. (6.6in.)(9in.) 63277.2lb. − in. M (1 − 2t ) − tT = − (1 − )− 2 180in. 27in.2 540in.3 Ac r S
= 45.86 psi (T)
< f c = 0.45 f c' = 0.45(4800 psi ) = 2160 psi (T ), O.K .
Stress in extreme bottom fiber after prestress and selfweight
fb = −
Pe ec 24, 456lb. (6.6in.)(9in.) 63, 277.2lb. − in. M (1 + 2b ) + T = − (1 + )+ = −317.6 psi (C ) 2 180in. 27in.2 540in.3 Ac r Sb
< ft = 12 f c' = 12 4800 psi = 831.4 psi (T ), O.K .
Loads
Load needed to cause zero deflection 0.02293in. =
P(60in.) (3(180in.) 2 − 4(60in.) 2 ) 4 24(3,949, 076 psi )(4860in. )
P = 2126 lbs.
Calculated Load of Application at Decompression
fb = −
Pe ec M My (1 + 2b ) + T + Ac r Sb I
0 = −317.6 psi +
P(60in.)(9in.) 4
P = 2858 lbs.
Determine the force of first cracking based on the cracked moment
M cr = M D + M L = f r Sb + Pe (e +
r2 ) cb
27in.2 ) 9in.
63, 277.2lb. − in. + P(60in.) = (520 psi )(540in.3 ) + 24, 456lbs.(6.6in. +
P = 7538 lbs.
Ultimate Strength Design by Strain Compatibility
ε1 = ε pe =
f pe E ps
=
53, 280 psi = 0.0019029 in. in. 28, 000, 000 psi
Pe = 3(0.153in.2 )(53, 280 psi ) = 24, 456lbs.
ε 2 = ε decomp =
Pe 24, 456lbs. (6.6in.) 2 e2 (1 + 2 ) = (1 + ) = 0.0000899 in. in. (180in.2 )(3,949, 076 psi ) 27in.2 Ac Ec r
Assume 205,000 psi for iteration process: a= Aps f ps 0.85 f c'b = 3(0.153in.2 )(205, 000 psi ) = 2.30625in. 0.85(4800 psi )(10in.)
β1 = 0.85 − 0.05 = 0.80
c= a
β1
=
2.30625 = 2.8828in. 0.8
d = 15.6in.
ε3 = εc (
d −c 15.6in. − 2.8828in. ) = 0.003 in. ( ) = 0.0132342 in. in. in. c 2.8828in.
− in. + 0.0000899 in. 60in.075 0.587 lb.015227 in.075 = 268 − = 259. in. Plugging 259. in.6in.0065 0. 2 2 Failure Determining the force needed to reach failure. M n = Pa P= M n 1. 406 psi ε ps − 0.520 psi )(15. = 0. 213lb. − in. 655.406 psi and iterating. f ps = 268 − 0. M n = Aps f ps (d p − ) = 3(0. − ) = 1.015227 in.ε ps = ε1 + ε 2 + ε 3 = 0.8633in.2 )(254. in. = = 27.0132342 in.153in. a . the value of prestress obtained is: f ps = 254.0019029 in. + 0. in.0065 in.520 psi Using the prestress force. 655. 213lb. the nominal moment obtained is: a 2. − 0.
. . Stephen M. Ph. Foley.D. Wolanski and have found that it is complete and satisfactory in all respects.Marquette University This is to certify that we have examined this copy of the masters thesis by Anthony J.D. P. Ph. Baolin Wan. Heinrich.E. This thesis has been approved by: Director Christopher M.D. Ph.
This action might not be possible to undo. Are you sure you want to continue?
Use one of your book credits to continue reading from where you left off, or restart the preview.