FINITE ELEMENT ANALYSIS AND MODEL VALIDATION OF SHEAR

DEFICIENT REINFORCED CONCRETE BEAMS STRENGTHENED WITH

GFRP LAMINATES
Damian I. Kachlakev, Ph.D., P.E.
California Polytechnic State University

Abstract
An ANSYS finite element model is used to study the effects of shear strengthening by
comparing the behaviors of two full-scale reinforced concrete beams (a reinforced concrete beam with
no shear stirrups; and a reinforced concrete beam externally reinforced with Glass Fiber Reinforced
Polymer (GFRP) on both sides of the beam). Experimental beams replicated the transverse members
from the Horsetail Creek Bridge, which are deficient in shear reinforcement. Three-dimensional finite
element models are developed using a smeared cracking approach for the concrete and three-
dimensional layered elements for the FRP composites.
It was found that the general behaviors through the linear and nonlinear ranges up to failure
of the finite element models show good agreement with observations and data from the experimental
full-scale beam tests. The addition of GFRP reinforcement to the control beam shifts the behavior of the
actual beam and model from a sudden shear failure near the ends of the beam to flexure failure by steel
yielding at the midspan. The shear reinforcement increases the load carrying capacity by 45% for the
experimental beam and by 15% for the finite element model. This finite element model can be used in
additional studies to develop design rules for strengthening reinforced concrete bridge members using
FRP.

KEY WORDS: Finite Element Analysis; Reinforced Concrete Beams; Fiber-Reinforced Polymers

Introduction
In the last decade, fiber reinforced polymer (FRP) composites have been used for strengthening
structural members of reinforced concrete bridges, which are deficient or obsolete due to changes in
their use or consideration of increased loadings. Many researchers have found that FRP composites
applied to those members provide efficiency, reliability and cost effectiveness in rehabilitation ([10],
[13], [18], [21]). Currently in the U.S., ACI Committee 440 is working to establish design
recommendations for FRP application to reinforced concrete [2].
The Horsetail Creek Bridge was constructed in 1913. The bridge is in current use, located east
of Portland, Oregon along the Historic Columbia River Highway, and is a historic structure. The
transverse bridge beams were constructed without the presence of shear reinforcement [11]. External
reinforcement with FRP composites was used to increase the strength of the beams due to the historic
nature of the bridge, limited funding and time restrictions.
In this paper, the ANSYS finite element program [3] is used to simulate the behavior of two full-
scale reinforced concrete beams, which replicated the transverse members from the Horsetail Creek
Bridge [11]. The finite element model uses a smeared cracking approach and three-dimensional layered

1
elements to model FRP composites. This model can help to confirm the theoretical calculations as well
as to provide a valuable supplement to the laboratory investigations of behavior.

Experimental Beams
Four full-scale reinforced concrete beams (similar to the transverse beams of the Horsetail Creek
Bridge) were fabricated and tested at Oregon State University (a control beam, a shear-strengthened
beam, a flexural-strengthened beam, and a beam with both shear and flexural strengthening) in order to
compare their behaviors in the laboratory. Glass Fiber Reinforced Polymer (GFRP) was applied on the
sides of the actual Horsetail Creek Bridge beams to provide increased shear strength, and Carbon Fiber
Reinforced Polymer (CFRP) to the bottom of the beams to increase the flexural strength.
This paper focuses on the performance of the shear-strengthened beam; therefore the behaviors
of the control reinforced concrete beam (no shear stirrups and no FRP reinforcement) and the shear-
strengthened beam (added GFRP reinforcing on the sides) are compared. The experimental beams were
tested in third point bending. Figure 2.1 illustrates these two experimental beams.

P P
/2 /2
2135 1825

#5 Steel rebar 305

50 #5 Steel rebar
65

770
510

65
#6 Steel rebar #7 Steel rebar
305 100
6095

(a)
P P
/2 /2

770 745

4 layers
150 Unidirectional GFRP (2layers)
1675

2895
6095

(b)

Figure 2.1: Experimental beams (dimensions shown in mm): (a) Control Beam;
(b) Shear-Strengthened Beam [14]

2
 Finite Element Models

Element Types
A solid element, SOLID65, is used to model the concrete in ANSYS. The solid element has
eight nodes with three degrees of freedom at each node, translations in the nodal x, y, and z directions.
The element is capable of plastic deformation, and cracking in three orthogonal directions. A LINK8
element is used to model the steel reinforcement. Two nodes are required for this element. At each
node, degrees of freedom are identical to those for the SOLID65. The element is also capable of plastic
deformation. A layered solid element, SOLID46, is used to model the GFRP composite. The element
allows for up to 100 different material layers with different orientations, and orthotropic material
properties in each layer. The element has three degrees of freedom at each node, translations in the
nodal x, y, and z directions.

Material Properties
Concrete: SOLID65 elements are capable of predicting the nonlinear behavior of concrete
materials using a smeared crack approach [22]. The smeared crack approach has been adopted widely in
the recent decades ([16], [19], [4], [20]). Concrete is a quasi-brittle material and has very different
behaviors in compression and tension. The tensile strength of concrete is typically 8-15% of the
compressive strength. For the full-scale beam models, the elastic modulus for each beam was estimated
using a pulse velocity method. The ultimate concrete compressive and tensile strengths for each beam
model were calculated by Equations 1, and 2, respectively [1].

2
 E 
fc '=  c  (1)
 4730 
1
f r = 0.623( f ' c ) 2 (2)
where:
E c = elastic modulus of concrete, MPa
f c ' = ultimate compressive strength, MPa
f r = ultimate tensile strength (modulus of rupture), MPa

Next, equations 3 and 4 [7] are used along with Equation 5 to construct the uniaxial compressive
stress-strain curve for concrete in this study.

Ec ε
f = 2
(3)
ε 
1 +  
 ε0 
2 f 'c
ε0 = (4)
Ec

f
Ec = (5)
ε

3
 where:
f = stress at any strain ε , MPa
ε = strain at stress f
E c = concrete elastic modulus, MPa
ε 0 = strain at the ultimate compressive strength f c '

In tension, the stress-strain curve for concrete is assumed to be linearly elastic up to the ultimate
tensile strength. After this point, the concrete cracks and the strength decreases to zero. Figure 3.1
shows the simplified uniaxial stress-strain relationship that is used in this study.

-σ

ultimate compressive strength

fc’

Ec

Compression

0. 30 f’c

strain at ultimate strength

+ε -ε
ε0
Tension 0.623 f’c1/2 ultimate tensile strength
+σ
Figure 3.1: Simplified Uniaxial Stress-Strain Curve for Concrete

Poisson’s ratio for concrete is assumed to be 0.2 and is used for all beams. The value of a shear
transfer coefficient, representing conditions of the crack face, used in many studies of reinforced
concrete structures varied between 0.05 and 0.25 ([5], [8], [9]). The shear transfer coefficient used in
this study is equal to 0.2.
Steel Reinforcement and Steel Plates: Steel reinforcement in the experimental beams was
constructed with typical steel reinforcing bars (fy = 415 Mpa). Elastic modulus and yield stress for the
steel reinforcement used in this FEM study follow the design material properties used for the
experimental investigation. The steel for the finite element models is assumed to be an elastic-perfectly
plastic material and identical in tension and compression. A Poisson’s ratio of 0.3 is used for the steel
reinforcement. Figure 3.2 shows the stress-strain relationship used in this study. Material properties for
the concrete and steel reinforcement are summarized in Table 3.1.
For the finite element models, each load is distributed over a small area as for the experimental
beams. A 25 mm thick steel plate, modeled using SOLID45 elements, is added at the support location in
order to avoid stress concentration problems. This provides a more even stress distribution over the
support area. An elastic modulus equal to 200 GPa and Poisson’s ratio of 0.3 are used for the plates.
The steel plates are assumed to be linear elastic materials.

4
 -σ

- fy

Es Compression
εy
+ε -ε
-εy
Tension
fy

+σ
Figure 3.2: Stress-Strain Curve for Steel Reinforcement

Table 3.1: Summary of Material Properties for Reinforced Concrete

Concrete Steel Rebar
Beam
Ec (GPa) fc’ (MPa) fr (MPa) ν βt Es (GPa) fy (MPa) ν

Control beam 19.3 16.7 2.55 0.2 0.2 200 415 0.3
Shear beam 18.2 14.7 2.39 0.2 0.2 200 415 0.3

FRP Composites: For this study, the GFRP is assumed to be a specially orthotropic and
transversely isotropic material, where the properties of the FRP composites are the same in any direction
perpendicular to the fibers. GFRP is applied on the sides of the beams for increased shear strength, due
to its superior strain at failure. Linear elastic properties of FRP composites are assumed throughout this
study.

Table 3.2: Summary of Material Properties for GFRP Composite [11]

FRP Elastic Major Tensile Shear Thickness
composite modulus Poisson’s strength modulus of laminate
(GPa) ratio (MPa) (GPa) (mm)
Ex = 20.7 νxy = 0.260 Gxy = 1.52
GFRP 1.3
Ey = 6.89* νxz = 0.260 600 Gxz = 1.52
Ez = 6.89* νyz = 0.300* Gyz = 2.65**
*[10]
E y or z
** G yz =
2(1 + ν yz )

5
Modeling Methodology
By taking advantage of the symmetry of the beams, a quarter of the full beam is used for
modeling with proper boundary conditions. This approach reduces computational time and computer
disk space requirements significantly. The steel reinforcement is simplified in the model by ignoring the
inclined portions of the steel bars present in the test beams. Ideally, the bond strength between the
concrete and steel reinforcement should be considered. However, in this study, perfect bond between
materials is assumed.
The various thickness of the FRP composites create discontinuities, which are not desirable for
the finite element analysis. These may develop high stress concentrations at local areas on the models,
yielding difficulties in convergence of the solutions. Therefore, a consistent thickness of FRP
composites is used in the models to avoid discontinuities, by compensating with changes in the elastic
and shear moduli in each layer. For example, if the thickness of FRP laminates is doubled, the elastic
and shear moduli are both reduced by 50%. Note that the relationship between elastic and shear moduli
is linear. Perfect bond between concrete and FRP laminates is assumed. A convergence study was
carried out to determine an appropriate mesh density. Minor modification of dimensions for the FRP
reinforcing was made due to geometric constraints from the other elements in the models, i.e. meshing
of concrete elements, steel rebar locations and required output locations. Figure 3.3 shows the finite
element models.
P
P
/4 /4
2135
100
1690
1525
CL
1220 50
90

770
510

100 65
305
3050 150

(a) P
/4
CL

770 680

4 layers
150 1525 Unidirectional GFRP (2 layers)
2885

3050

(b)

Figure 3.3: Finite Element Models (dimensions shown in mm): (a) Control Beam;
(b) Shear-Strengthened Beam

6
 Comparison of Results
Tensile Strain in Main Steel Reinforcing
Comparisons of the load-tensile strain plots from the finite element analyses and the
experimental data for the main steel reinforcing at midspan are shown in Figure 4.1. For both the
control and shear-strengthened beams in the linear range (before concrete cracking) the strains from the
finite element analysis correlate well with those from the experimental data. In the nonlinear range, the
trends of the finite element and the experimental results are generally similar. The finite element
analysis supports the experimental results that the main steel rebar at midspan for the control beam has
not yielded at failure, while the steel rebar for the shear-strengthened beam yields but at a lower load.

700 700

600 600

500 500

Load, P (kN)
Load, P (kN)

400 400
Steel yielding
300 300

200 200
Experiment Experiment
100 100
FEM FEM
0 0
0 1500 3000 4500 6000 7500 9000 10500 12000 13500 0 1500 3000 4500 6000 7500 9000 10500 12000 13500
Microstrain Microstrain

(a) (b)

Figure 4.1: Load-Tensile Strain Plot for Main Steel Rebar: (a) Control Beam;
(b) Shear-Strengthened Beam

Compressive Strain in Concrete

The load-compressive strain (in concrete) plots collected at midspan at the center of the top face
from the experiment are compared with results from the finite element analysis in Figure 4.2. For the
control beam, the load-compressive strain plots from the finite element analysis and the experimental
data have excellent agreement. For the shear-strengthened beam, the load-strain plots from the finite
element and experimental results do not correlate well. As shown in the figure, the experimental beam
shows unexpected nonlinear behavior for applied loads from 0 to 470 kN. Either erroneous test data or
local material imperfections may be the cause. Cracks occurring at the interfaces between the cement
and aggregate due to their differences in elastic modulus, thermal coefficient, and response to change in
moisture content when the concrete is hardened could be the source of the local material imperfections.
At about 490 kN, large strains occur for the finite element model, whereas at a load of 535 kN similar
behavior takes place for the experimental beam. These loads are close to the yielding loads of the steel
as shown in Figure 4.1. The yielding of the steel explains the large concrete strains.

7
 700

600
Shear Beam
500
Load, P (kN)
400 Control Beam

300

200
Experiment
100
FEM
0
0 300 600 900 1200 1500 1800 2100 2400 2700 3000
Microstrain
Figure 4.2: Load-Compressive Strain Plot for Concrete

Load-Deflection Plots
Deflections are measured at midspan at the center of the bottom face of the beams. Figure 4.3
shows the load-deflection plots for the control and shear-strengthened beams. In general, the load-
deflection plots for the beams from the finite element analyses agree quite well with the experimental
data. The finite element load-deflection plots in the linear range are somewhat stiffer than the
experimental plots. After first cracking, the stiffness of the finite element models is again higher than
that of the experimental beams. There are several effects that may cause the higher stiffnesses in the
finite element models. First, microcracks are present in the concrete for the experimental beams, and
could be produced by drying shrinkage in the concrete and/or handling of the beams. On the other hand,
the finite element models do not include the microcracks. The microcracks reduce the stiffness of the
experimental beams. Next, perfect bond between the concrete and steel reinforcing is assumed in the
finite element analyses, but the assumption would not be true for the experimental beams. As bond slip
occurs, the composite action between the concrete and steel reinforcing is lost. Thus, the overall
stiffness of the experimental beams is expected to be lower than for the finite element models (which
also generally impose additional constraints on behavior).

700

600
Shear
500 Beam
Load, P (kN)

400

300 Control Beam

200

100
Experiment
0
0 5 10 15 20 25 30 35 40
Mid-span deflection (mm)

Figure 4.3: Load-Deflection Plot

8
Loads at Failure
Table 4.1 shows comparisons between the ultimate loads of the experimental beams and the
final loads from the finite element models. The final loads for the finite element models are the last
applied load steps before the solution diverges due to numerous cracks and large deflections. It is seen
that the ANSYS models underestimate the strengths of the beams, as anticipated.
One explanation is that the inclined portions of the steel reinforcement are excluded from the
finite element models. Toughening mechanisms at the crack faces [17], i.e. the grain bridging process,
interlocking between the cracked faces, crack tips blunted by voids, and the crack branching process,
may also slightly extend the failures of the experimental beams before complete collapse. The finite
element models do not have these mechanisms. Finally, the material properties assumed in this study
may be imperfect.
In the experiment, the failure modes for the beams were as predicted. The control beam failed in
shear. The shear-strengthened beam failed in flexure at the midspan, with yielding of the steel
reinforcing followed by a compression failure at the top of the beam. Crack patterns obtained from the
finite element analyses at the last converged load steps and the failure modes of the experimental beams
agree very well. For the finite element model of the control beam, smeared cracks spread over the high
shear stress region and occur mostly at the ends of the beam from the support toward the loading area.
The finite element program accurately predicts that the control beam fails in shear. For the shear-
strengthened beam, numerous cracks occur at midspan rather than underneath the loading location. The
crack patterns and steel yielding at the midspan for the finite element shear strengthened beam support
the experimental results that the beam fails in flexure.

Table 4.1: Comparisons Between Experimental Ultimate Loads and Finite Element Final Loads
Ultimate load (kN) Final load
Beam from Experiment, (kN) from %Difference
Failure Mode* FEA
Control beam 475, Shear 455 -5
Shear beam 690, Flexure 525 -24
Percent Gain over Control Beam 45% 15%
*[14]

Evolutions of Crack Patterns for Concrete

The ANSYS program records a crack pattern at each applied load step. Figure 4.4 shows
evolutions of crack patterns developing for each beam. The cracks appear underneath the loading
location on the control beam model. For the shear strengthened beam model, there are no compressive
cracks underneath the loading location. The appearance of the cracks reflects the failure modes for the
beams.

9
 CL CL

107 kN 107 kN

267 kN 267 kN

454 kN 525 kN

(a) (b)

Figure 4.4: Evolution of Crack Patterns: (a) Control Beam; (b) Shear-Strengthened Beam

Conclusions
The general behaviors of the finite element models show good agreement with observations and
data from the experimental full-scale beam tests. The addition of GFRP reinforcement to the control
beam shifts the behavior of the actual beam and model from a sudden shear failure near the ends of the
beam to flexure failure by steel yielding at the midspan. The shear reinforcement increases the load
carrying capacity by 45% for the experimental beam and by 15% for the finite element model. This
finite element model can be used in additional studies to develop design rules for strengthening
reinforced concrete bridge members using FRP.

References
1. ACI 318-99 (1999), American Concrete Institute, Building Code Requirements for Reinforced
Concrete, American Concrete Institute, Farmington Hills, Michigan.
2. ACI 440 (2001), American Concrete Institute, Guide for the Design and Construction of Externally
Bonded FRP Systems for Strengthening Concrete Structures, American Concrete Institute,
Farmington Hills, Michigan.
3. ANSYS (1998) ANSYS User’s Manual Revision 5.5, ANSYS, Inc., Canonsburg, PA.
4. Arduini, M., Di Tommaso, A., and Nanni, A., (1997) “Brittle Failure in FRP Plate and Sheet Bonded
Beams,” ACI Structural Journal, 94(4), pp. 363-370.
5. Bangash, M. Y. H. (1989) Concrete and Concrete Structures: Numerical Modeling and
Applications, Elsevier Science Publishers Ltd., London, England.
6. Chansawat, K., (2000) “Nonlinear Finite Element Analysis of Reinforced Concrete Structures
Strengthened With FRP Laminates,” personal communication, Corvallis, Oregon..

10
7. Desayi, P. and Krishnan, S., (1964) “Equation for the Stress-Strain Curve of Concrete,” Journal of
the American Concrete Inst., 61, pp. 345-350.
8. Hemmaty, Y., (1998) “Modelling of the Shear Force Transferred Between Cracks in Reinforced and
Fibre Reinforced Concrete Structures,” Proceedings of the ANSYS Conference, Vol. 1,
Pittsburgh, Pennsylvania.
9. Huyse, L., Hemmaty, Y., and Vandewalle, L., (1994) “Finite Element Modeling of Fiber Reinforced
Concrete Beams,” Proceedings of the ANSYS Conference, Vol. 2, Pittsburgh, Pennsylvania.
10. Kachlakev, D. I., (1998) “Strengthening Bridges Using Composite Materials,” FHWA Report No.
OR-RD-98-08, Corvallis, Oregon.
11. Kachlakev, D.I. and McCurry, D., Jr., (2000) “Simulated Full Scale Testing of Reinforced Concrete
Beams Strengthened with FRP Composites: Experimental Results and Design Model
Verification,” United States Department of Transportation, Federal Highway Administration.
12. Kachlakev, D. I., Miller, T. H., Yim, S., Chansawat, K., and Potisuk, T., (2000)“Linear and
Nonlinear Finite Element Modeling of Reinforced Concrete Structures Strengthened With FRP
Laminates,” United States Department of Transportation, Federal Highway Administration.
13. Marshall, O. S. and Busel, J. P., (1996) “Composite Repair/Upgrade of Concrete Structures,” 4th
Materials Conference, “Materials for the New Millennium”, editor K. Chong, American Society
of Civil Engineers, Washington, D. C.
14. McCurry, D., Jr. and Kachlakev, D. I., (2000) “Strengthening of Full Sized Reinforced Concrete
Beam Using FRP Laminates and Monitoring with Fiber Optic Strain Gauges,” in “Innovative
Systems for Seismic Repair and Rehabilitation of Structures,” Design and Applications,
Technomic Publishing Co., Inc., Philadelphia.
15. Potisuk, T., (2000) “Finite Element Modeling of Reinforced Concrete Beams Externally
Strengthened by FRP Composites,” MS Thesis, Oregon State University, Corvallis, Oregon.
16. Rashid, Y. R., (1968) “Ultimate Strength Analysis of Prestressed Concrete Pressure Vessels,”
Nuclear Engineering and Design, 7(4), pp. 334-344, Amsterdam, Netherlands.
17. Shah, S. P., Swartz, S. E., and Ouyang, C., (1995) Fracture Mechanics of Concrete, John Wiley &
Sons, Inc., New York, New York.
18. Steiner, W., “Strengthening of Structures with CFRP Strips, (1996) ”Advanced Composite Materials
for Bridges and Structures, 2nd International Conference, Proceedings, editor M.El-Bardry,
Montreal, Canada.
19. Suidan, M. and Schnobrich, W. C., (1973) “Finite Element Analysis of Reinforced Concrete,”
Journal of the Structural Division, ASCE, ST10, pp. 2109-2122.
20. Tedesco, J. W., Stallings J. M., and El-Mihilmy, M., (1999) “Finite Element Method Analysis of a
Concrete Bridge Repaired with Fiber Reinforced Plastic Laminates,” Computers and Structures,
72, 379-407.
21. Tedesco, J. W., Stallings J. M., El-Mihilmy, M., and McCauley, M., (1996) “Rehabilitation of
Concrete Bridges Using FRP Laminates,” 4th Materials Conference, “Materials for the New
Millennium”, editor K. Chong, American Society of Civil Engineers, Washington, D. C.
22. William, K. J. and Warnke, E. P., (1975) “Constitutive Model for the Triaxial Behavior of
Concrete,” Proceedings, International Association for Bridge and Structural Engineering, Vol.
19, ISMES, Bergamo, Italy, pp. 174.

11