Finite element analysis of CFRP-externally

strengthened reinforced concrete beams

subjected to three-point bending
Sabiha Barour and Abdesselam Zergua
Department of Civil Engineering, University of Constantine 1 Faculty of Engineering Sciences, Constantine, Algeria
Farid Bouziadi
Department of Civil Engineering, Laboratory of Materials Sciences and Environment, University Hassiba Benbouali of Chlef,
Chlef, Algeria, and
Waleed Abed Jasim
Projects Department, Ministry of Construction and Housing and Public Municipalities, Baghdad, Iraq

Abstract
Purpose – This paper aims to develop a non-linear finite element model predicting the response of externally strengthened beams under a three-
point flexure test.
Design/methodology/approach – The ANSYS software is used for modeling. SOILD65, LINK180, SHELL181 and SOLID185 elements are used,
respectively, to model concrete, steel reinforcement, polymer and steel plate support. A parametric study was carried out. The effects of compressive
strength, Young’s modulus, layers number and carbon fiber-reinforced polymer thickness on beam behavior are analyzed. A comparative study
between the non-linear finite element and analytical models, including the ACI 440.2 R-08 model, and experimental data is also carried out.
Findings – A comparative study of the non-linear finite element results with analytical models, including the ACI 440.2 R-08 model and
experimental data for different parameters, shows that the strengthened beams possessed better resistance to cracks. In general, the finite element
model’s results are in good agreement with the experimental test data.
Practical implications – This model will predict the strengthened beams behavior and can describe the beams physical conditions, yielding the
results that can be interpreted in the structural study context without using a laboratory testing.
Originality/value – On the basis of the results, a good match is found between the model results and experimental data at all stages of loading the
tested samples. Crack models obtained in the non-linear finite element model in the beams are also presented. The submitted finite element model
can be used to predict the behavior of the reinforced concrete beam. Also, the comparative study between an analytical model proposed by of
current code of ACI 440.2 R-08 and finite element analysis is investigated.
Keywords Reinforced concrete beams, CFRP, Numerical model, ANSYS, Shear, Flexural
Paper type Research paper

Nomenclature Es = modulus of elasticity of steel;

0
fc = specified compressive strength of concrete;
Af = cross-sectional area of fiber-reinforced polymer ffe = effective stress in the FRP; stress level attained at
(FRP) reinforcement; section failure;
As = cross-sectional area of tension steel reinforcement; ffu = ultimate tensile strength of CFRP;
c = distance from extreme compression fiber to the ffu = ultimate tensile strength of FRP reinforcement;
neutral axis; fs = stress in non-prestressed steel reinforcement;
CE = environmental reduction factor; fy = specified yield strength of non-prestressed steel
d = distance from extreme compression fiber to reinforcement;
centroid of tension reinforcement; Icr = moment of inertia of cracked section transformed
df = effective depth of FRP flexural reinforcement; to concrete;
Ec = modulus of elasticity of concrete; k = ratio of depth of neutral axis to reinforcement depth
Ef = modulus of elasticity of carbon FRP (CFRP); measured from extreme compression fiber;
MDL = dead-load moment;
Mn = nominal flexural strength;
The current issue and full text archive of this journal is available on Mnf = contribution of FRP reinforcement to nominal
Emerald Insight at: https://www.emerald.com/insight/1708-5284.htm flexural strength;

World Journal of Engineering

Received 19 April 2019
17/2 (2020) 183–202 Revised 15 June 2019
© Emerald Publishing Limited [ISSN 1708-5284] 15 September 2019
[DOI 10.1108/WJE-04-2019-0121] Accepted 19 October 2019

183
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Mns = contribution of steel reinforcement to nominal also compared the existing variation in behaviors of RC beams
flexural strength; with and without shear reinforcement to that obtained from
n = number of plies of FRP reinforcement; finite element (FE) analysis. The results have showed that all
tf = assumed thickness of the CFRP layer; types of web reinforcements have almost similar effect for a
« bi = strain level in concrete substrate at time of FRP static loading condition. The results of the FE model have
installation (tension is positive); showed a good matching with the experimental results.
«c = strain level in concrete; Also, Abu–Obeidah et al. (2015) and Taljsten (2003) have
«cu = ultimate axial strain of unconfined concrete; showed that the using FRP sheet/laminate for shear
«fd = debonding strain of externally bonded FRP strengthening RC beams can improve the ultimate load and
reinforcement; reduce deflection. Also, experimental studies have showed the
«fe = effective strain level in FRP reinforcement attained effectiveness of using externally bonded CFRP to improve the
at failure; strength of RC members (Hawileh et al., 2014). Similarly,
« fu = ultimate rupture strain of FRP reinforcement; Panigrahi et al. (2014) have worked on 12 RC beams, which
«fu = design rupture strain of FRP reinforcement; have been tested under four-point bending. Focus on
«s = strain level in non-prestessed steel reinforcement; retrofitting of the RC beams in shear has been investigated by
Wf = FRP strength reduction factor; using mechanically anchored bonded FRP fabrics. They have
0
a1 = multiplier on fc to determine intensity of an found that the FRP composites can significantly enhance the
equivalent rectangular stress distribution for strength of the RC beams. Musmar et al. (2006) have
concrete; investigated the physical response of shallow RC beams
b1 = ratio of depth of equivalent rectangular stress block subjected to transverse loading. The distribution of stresses
to depth of the neutral axis; and cracks and load–deflection relationship have been evaluated
1 = strength reduction factor. using ANSYS the FE software. Sasmal et al. (2012) have
focused on the numerical investigation using ANSYS non-
1. Introduction linear FE analysis software to study the response of flexural
strengthening of FRP on RC beams. Concrete, steel
Composite materials have good physical and mechanical
reinforcement, FRP and epoxy matrix have been modeled
properties, high resistance, good damping capacity, ease of
using SOLID65, LINK8, SHELL41 and SOLID45,
installation and high corrosion resistance (Wang et al., 2011).
respectively. The numerical results have indicated a significant
Numerous studies on carbon fiber-reinforced polymer
improvement in the displacement in the strengthened RC
(CFRP)-externally strengthened reinforced concrete beams
beams with the advancement of crack.
subjected to bending have investigated by several studies. The
Al-Zaid et al. (2012) have used an analytical approach to
results have shown that the CFRP composite materials increase
predict the response of RC beams strengthened for flexure with
the strength and ductility of the structural elements (Al-Hadithy
externally bonded FRP composites. They have proposed a
and Al-Ani, 2016). Ashour et al. (2004), Ahmed et al. (2011)
model based on cross-sectional analysis satisfying strain
and Aboutaha et al. (2003) have shown that reinforced concrete compatibility and equilibrium conditions. The proposed
(RC) beams have a higher load capacity than the non-RC analytical model has been compared with experimental results
beams by 20-45 per cent. Dong et al. (2013) have studied the from the literature. The model has been verified against test
behavior of RC with external flexural and flexural-shear data related to full composite action failure mode, and a good
strengthening by CFRP and Glass composed fiber-reinforced agreement has been found.
polymer (GFRP) sheets. The obtained results have Previous studies have used the ANSYS program to predict
demonstrated that both the flexural and flexural-shear the behavior of CFRP-retrofitting RC beams. The load–
strengthening capacity of the RC beams using externally deflection curves and failure modes have been consistent with
bonded of CFRP or GFRP sheets on bottom and/or lateral faces the laboratory tests (Ibrahim, 2009; Mohamed et al., 2017;
can significantly improve the flexural and shear capacity of the Abbas, 2010). Ibrahim and Mahmood (2009) have modeled
beams strengthened. The increase on the overall flexural the RC beams strengthened using externally bonded CFRP
capacity of the CFRP-strengthened beams varies between 41 laminates; they have used the smeared cracking approach. The
and 125 per cent over the control beam, and on the shear curve of load–deflection relationship matches well with the
capacity of the GFRP- or CFRP-strengthened beams between experimental results in the linear stage. Fathelbab et al. (2011)
31 and 74 per cent. Esfahani et al. (2007) have found that when and Jayajothi et al. (2013) have studied the effect of shear
low reinforcement ratios are used, the bending strength of the reinforcement on RC beams strengthened with different FRP
beams is overestimated. On the other hand, previous sheet patterns. Strengthened beams show a higher ultimate
experimental works have been carried out on the behavior of the capacity and a significant improvement in ductility compared
CFRP-externally strengthened RC beam under bending (Sharif to the control beam. Zidani et al. (2015) have presented an
et al., 1994; Basappa and Rajagopal, 2013). Dahmani et al. advanced FE model using an FE software to simulate the
(2010) have studied the crack propagation in reinforced complete history of the steps where the concrete beam is
concrete beams using FEA software ANSYS. It has been found initially loaded to introduce damage and then the beam is
there is a good agreement between the experimental and reloaded at rupture. Gopinath et al. (2017) have numerically
numerical results. In addition, Saifullah et al. (2011) have studied the response of RC beams loaded with basalt RC. The
studied the response of RC beams with different shear load–deflection plots obtained from numerical study show a
reinforcement patterns using finite element analysis. They have good agreement with the experimental plots. Atea (2018) has

184
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

used the ANSYS program to simulate the bending behavior of Figure 1 Details of the experimental program
four RC beams composed by adding steel plates with different
thicknesses.
The aim of this work is the development of numerical
models of CFRP-externally strengthened RC beams, without
stirrups and tested under three-point flexure. The obtained
results are compared with those of the experimental plots
reported by El-Sayed (2014). An important agreement is
reached between the experimental and FE results, and a high
level of accuracy is observed. The comparative study between
the current code of ACI 440.2 R-08 and FE analysis is also
investigated.

2. Modeling and validation of the experimental

studies
2.1 Experimental program
According to the investigations carried out by El-Sayed (2014),
an experimental program is achieved to study the behavior of
RC beams strengthened with layers of CFRP and subjected to
three-point bending. All beams have a rectangular cross-section
dimensions of 150  300 mm and a total length of 1,800 mm. Figure 2 SOLID65 elements
All specimens have been classified according to elastic modulus
of CFRP, thickness, number and width of layer. The beams
have been reinforced with two longitudinal bars of 16 mm
diameter at bottom zone, giving a reinforcement ratio ( r s) of 1
and 1.5 per cent. The retrofitted RC beams are externally
strengthened in flexural zone (bottom of the RC beam). The
details of the RC beams are presented in Table I, and Figure 1
shows the details of all tested specimens.

2.2 Numerical investigation

A non-linear FE method is a numerical technique for
obtaining approximate solutions to a wide variety of
engineering problems. ANSYS is software of FE modeling
used to solving a wide variety of problems, of which structural
analysis (linear or non-linear) is one (ANSYS, 2007).
ANSYS 12.0 software has developed to provide an FE model
tension and crushing in compression (ANSYS, 2007). As
of experimental works.
reported by El-Sayed (2014), the detailed mechanical
2.2.1 Non-linear finite element modeling characteristics of the materials used in the developed FE
The elements used in this study are divided into four types. models are given in Table II. The uniaxial stress-to-strain
2.2.1.1 Concrete. SOLID65 element is used for modeling relationship for the concrete model is that proposed by
concrete material (Figure 2). SOLID65 element has a total of Wolanski (2004).
eight nodes with three degrees of freedom in translation in For concrete, FEA software ANSYS requires input data for
directions X, Y and Z on each node.SOLID65 has the ability of material properties as shown in Tables III, IV and V. The real
simulating the non-linear behavior of concrete by cracking in constants for this model are shown in Table III. A value of zero

Table I Details of the laboratory tested beams

External strengthening of CFRP
Beams ID Ef (GPa) Thickness (mm) No. of layers Width of layer (mm)
Series I C-1.0 – – – –
PL-1.0-0.43 165 1.4 1 120
PL-1.0-0.85 165 1.4 2 120
PH-1.0-0.36 300 1.4 1 100
S-1.0-0.27 240 0.3 3 120
Series II C-1.5 – – – –
PL-1.5-0.85 165 1.4 2 120
Source: El-Sayed (2014)

185
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Table II Characteristics of concrete and steel reinforcement

Material Property Values
Concrete For beams of Series I – Compressive strength of concrete at 28 days 42.7 MPa
For beams of Series II – Compressive strength of concrete at 28 days 33.5 MPa
Poisson’s ratio 0.2
Steel reinforcement Modulus of elasticity 205 000 MPa
Poisson’s ratio 0.3
Yield stress 562 MPa
Source: El-Sayed (2014)

Table III Element type and real constants used in numerical study
Real constant set Element type Real constants
1 SOLID65 Real constants for Rebar 1
Material number 0
Volume ratio 0
Orientation angle 0
Orientation angle 0

Table IV Material properties of concrete FE model (first type)

Material model no. Element type Material properties
1 Soild65 Linear isotropic
EX 29658 MPa
PRXY 0.2
Multi-linear isotropic
Point strain stress (MPa) Point strain stress (MPa) Point strain stress (MPa)
1 0 0
2 0.0005244 15.5528
3 0.00065 18.92865
4 0.0008 22.69156
5 0.001 27.19506
6 0.0012 31.06854
7 0.0014 34.30191
8 0.0016 36.91679
9 0.0018 38.95744
10 0.002 40.48197
11 0.0022 41.55488
12 0.0027806 42.7
13 0.003 42.7
Open shear Coefficient transfer 0.5
Closed shear Transfer coefficient 0.9
Uniaxial cracking Stress 4.05
Uniaxial crushing Stress 42.7
Biaxial crushing Stress 0
Hydrostatic Pressure 0
Hydro biaxial Crushing stress 0
Tensile crack factor 0

is entered for all real constants. Because, in this study, the SOLID 65 element requires linear isotropic and multi-
concrete used is a non-fibrous concrete. linear isotropic material properties to properly model
All the material properties of two types of concrete used in concrete. Poisson’s ratio for concrete is assumed to be 0.2.
simulating the elements SOLID65 are illustrated in Tables IV Two shear transfer coefficients are required to be introduced
and V. for an open crack Bt and for a closed crack Bc. The value of

186
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Table V Material properties of the concrete FE model (second type)

Material model no. Element type Material properties
1 Soild65 Linear isotropic
EX 26,024 MPa
PRXY 0.2
Multi-linear isotropic
Point strain stress (MPa) Point strain stress (MPa) Point strain stress (MPa)
1 0 0
2 0.0005244 13.64708
3 0.00065 16.53073
4 0.0008 19.68565
5 0.001 23.35339
6 0.0012 26.38133
7 0.0014 28.78413
8 0.0016 30.60802
9 0.0018 31.91791
10 0.002 32.78675
11 0.0022 33.28763
12 0.0024629 33.5
13 0.003 33.5

Concrete
Open shear Coefficient transfer 0.5
Closed shear Transfer coefficient 0.9
Uniaxial cracking Stress 3.59
Uniaxial crushing Stress 33.5
Biaxial crushing Stress 0
Hydrostatic Pressure 0
Hydro biaxial Crushing stress 0
Tensile crack factor 0

these coefficients ranges from 0.0 to 1.0, with 0.0 (ANSYS, 2007). The steel bar is considered to be an elastic
representing a smooth crack (complete loss of shear plastic material with identical tensile and compressive behavior,
transfer) and 1.0 representing a rough crack (ANSYS, as shown in Figure 4.
2007). The adopted values for Bt and Bc are 0.5 and 0.9, LINK180 element is assumed to be a bilinear isotropic
respectively. (ANSYS, 2007). The bilinear model requires the yield stress
2.2.1.2 Steel reinforcement. LINK180 element is used for the (fy), as well as the tangent modulus of the plastic region of steel
flexural steel bar reinforcements. It has three degrees of freedom to be defined. Elastic modulus (EX) is also defined, and
to each node; translations in nodal X, Y and Z directions. Poisson’s ratio (PRXY) is assumed to be 0.3, as shown in
Plasticity, creep, rotation, large deflection and large strain Table VI. And also, the real constants for this model are shown
capabilities are included. This element is shown in Figure 3 in Table VII.

Figure 3 LINK180 elements

187
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Figure 4 Stress–strain relationships for steel reinforcement Table VII Element type and real constants used for FE model of steel
reinforcement
Real constant set Element type Real constants
2 LINK180 Cross-sectional area (mm2) 200.96

Figure 5 SOLID185 element

The perfect bond between the concrete and steel reinforcement

should be considered. To provide the perfect bond, the link
element for the steel reinforcement is connected between nodes Table VIII Material properties used for FE of steel plates
of each adjacent concrete solid element, so the two materials Material model no. Element type Material properties
shared the same nodes (Wolanski, 2004).
3 SOLID185 Linear isotropic
2.2.1.3 Loading and supporting steel plates. SOLID185
EX 200,000 MPa
element is used for the modeling of loading and supporting
PRXY 0.3
steel plates (Figure 5). SOLID185 is defined by eight nodes
having three degrees of freedom in each node. Translations in
the nodal directions x, y and z. SOLID185 is capable of
plasticity, hyperelasticity, stiffness of constraint, creep, materials consisting of one type of fiber material is
significant deflection and large strain capabilities (ANSYS, characterized by a linearly elastic stress–strain relationship until
2007). failure. This relation is shown in Figure 6 (Al-Zaid et al., 2012).
The SOLID185 element requires an elastic linear isotropic Table IX shows the material characteristics of CFRP composite
material. Table VIII shows the material properties of steel plates materials and Table X shows mechanical properties of the
used for simulation numerical using the ANSYS FEA software. CFRP plate. SHELL181element is used for modeling the
The surface interaction between SOLID185 (steel plates at CFRP sheet; it is an element of four-node with six degrees of
supports and at point of application of load) and SOLID65 freedom at each node; translations in directions of x, y and z;
(concrete) is created using CONTA174 and TARGET170 and rotations about the x, y and z. SHELL181 is well suited for
elements. A surface-to surface contact is created considering linear and large rotation and/or large strain nonlinear
the applied load and support as rigid (Banjara and applications. The geometry, node locations and the coordinate
Ramanjaneyulu, 2017). system are shown in Figure 7 (ANSYS, 2007).
2.2.1.4 Carbon fiber-reinforced polymer. When loaded in According to the previous investigations (Jayajothi et al.,
direct tension, FRP materials do not exhibit any plastic 2013; Hawileh et al., 2013; Banjara and Ramanjaneyulu,
behavior (yielding) before rupture. The tensile behavior of FRP 2017), a perfect bonding is considered to model the contact

Table VI Material properties used for FE model of steel reinforcement

Material model no. Element type Material properties
2 LINK180 Linear isotropic
EX 205,000 MPa
PRXY 0.3

Bilinear isotropic
Yield stress 562 MPa
Tangent modulus 20 MPa

188
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Figure 6 Stress–strain relationships for steel reinforcement and FRP contact between the concrete and steel reinforcement. The
composite materials epoxy adhesive is also modeled using the SOLID185 element.
The perfect bonding is assumed to model the interface between
the CFRP and adhesive material. The concrete–epoxy adhesive
interface is also assumed perfectly bonded (Banjara and
Ramanjaneyulu, 2017).
2.2.2 Meshing
In this step, the mesh beam test number of maximum meshing
dimension is 25  25 mm; the geometrical modeling and the
materials attributes for beam are shown in Figure 8.

3. Numerical simulation results and discussion

3.1 Validation of numerical model
Non-linear FE analysis of RC beams are strengthened with
layers of CFRP under the static incremental loads. The
deflections of the beams are measured at mid-span. In FE
models, the bond between the concrete and reinforcement steel
is assumed to be perfect, which could add some stiffness during
the non-linear behavior.
Table IX Material characteristics of CFRP composite materials
Micro cracks due to dry shrinkage and deficiency during the
experimental are present in the concrete to some extent. On the
CFRP type Ef (GPa) Thickness (mm) No. of layers other hand, the CFRP is modeled with the ultimate strength of
Low-modulus plates 165 1.4 1 the CFRP sheets assumed to be linear elastic until failure that
Low-modulus plates 165 1.4 2 shows the stress remains elastic. The FRP sheet test for the
High-modulus plates 300 1.4 1 mechanical properties has not been experimentally performed
Carbon sheets 240 0.3 3 in the laboratory. If the experimental mechanical properties of
Source: El-Sayed (2014) the FRP sheets were available, then the behavior of the
continuous failure of strengthened experimental beams could
be arrested by the FE model, which subsequently reduces the
Table X Mechanical properties of the CFRP plate ultimate load obtained from the FE analysis. The results
obtained are compared with El-Sayed’s experimental
Young modulus (GPa) Yield stress (MPa) (El-Sayed, 2014).
165 2,800 Figures 9(a)-(g) show a comparison between the load–
300 1,350 deflection relationship curves. It is clear that the FE analyses
240 4,000 are in good agreement with those of the laboratory-tested
beams. All the strengthened RC beams have a higher stiffness
than the un-strengthened ones. Table XI summarizes the
results of the experimental tests and numerical analysis for all
Figure 7 SHELL181 element specimens.
All the strengthened RC beams exhibit stiffer behavior in
comparison to the un-strengthened RC beams, resulting in a
change in the deflection and failure load values shown in
Tables XII and XIII.
The difference between the numerical and experimental
values, for the measured load, the load capacity and the final
bending of the tested specimens, is less than 10 per cent.

3.2 Numerical model behavior

The comparison between the experimental and numerical mid-
span deflection is presented in Figures 10 and 11, which show
the comparison between the experimental and numerical
ultimate forces at mid-span for all RC beams. It is clear that
there is a good agreement between the experimental and
numerical results of maximum deflections and load. On the
other hand, for deflections, the elastic modulus of the beam
between the concrete surface and CFRP. As debonding is not strengthened with a CFRP layer gives the minimum bending
observed between the concrete and CFRP (during value at mid-span.
experimental study), a perfect bond approach is adopted for Figure 12 shows the numerical results of first crack load.
this numerical study, which is the same approach used for Each beam has its first crack load. These differences are due to

189
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Figure 8 3D FE model of the CFRP-externally strengthened RC beam

the test parameters, such as the compressive strength, Young’s Figure 21 shows the FE model of CFRP-externally
modulus, number of layers and CFRP thickness on RC beam strengthened RC beams.
behavior. Figure 22 compares the load–deflection response curves
resulting from the FE analysis for the three strengthening
3.3 Evolution of crack configurations and control beam. While, Table XV represents
The ANSYS software reports the crack configuration at each the ultimate load for each strengthening configuration, as
loading step. Figure 13 shows the typical cracking signs calculated from the FE analysis.
occurring in FE models: (a) flexural cracks; (b) compressive From curves of load versus deflection presented in
cracks; (c) diagonal tensile cracks. Figures 14-20 show the Figure 22, it can be noticed that the maximum load (ultimate
development of crack patterns developed for each RC beam at load capacity) of strengthened RC beam PL-1.0-0.43 is equal
each applied loading step of the FE model with values first to 112.7245 KN, which leads to an increase in ultimate load
cracking at load and cracking at failure. Model validation is capacity about 21.33 per cent compared to control beam (C-
performed to compare the expected mid-span load deflection 1.0). The ultimate load capacity for configuration A is
response and ultimate load capacity with the experimental data 101.864 KN. This leads to an increase in ultimate load
for the RC beams. In all the beams, it is observed that the capacity about 9.64 per cent compared to control beam. For
bending cracks starts from the mid-span of the beam. The Configuration B, the ultimate load capacity is 98.868 KN,
appearance of cracks begins near the mid-span of all beams, which improves ultimate load capacity by 7.06 per cent.
which is observed in all RC beams with a CFRP layer. Based on these numerical results, it can be noticed that the
Hence, as the load increases, the bending cracks increase strength of RC beam increases with increasing of the CFRP
within the shear span. Obviously, the shear stresses multiply, length plate.
and the cracks gradually tilted and propagated near the middle Figure 23 shows the development of the maximum
of the span of all beams, leading eventually to diagonal tension deflection at mid-span of RC beams (C-1.0, PL-1.0-0.43,
failure. Configurations A and B). The percentage decrease in
maximum deflection of PL-1.0-0.43, Configurations A and
3.4 Parametric study B are found to be 6, 3 and 2 per cent, respectively, as
3.4.1 Effect of carbon fiber-reinforced polymer length compared to that of control beam (C-1.0). In fact, in the
To investigate the effectiveness of CFRP lengths on non-linear FE analysis formulations, the element stiffness
strengthened RC beam, three different strengthening schemes matrix is proportional to the material property matrix. The
modeled are investigated. Table XIV shows CFRP lengths material property matrix is composed of the contribution
used in this parametric study. from the material property of concrete, steel, epoxy adhesive
 Configuration (PL-1.0-0.43): One layer of CFRP of 1,400 and CFRP. Therefore, the increased CFRP length enhances
mm length; these material properties; consequently, the element
 Configuration A: One layer of CFRP of 1,125 mm length; stiffness matrix of an externally retrofitted RC beam is most
and robust with the greatest CFRP layer-length, which in turn
 Configuration B: One layer of CFRP of 1,050 mm length. restricts crack growth, thus improving the strength of the

190
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Figure 9 Comparison of load–deflection relationship of RC beams from test results and FE models
140 140

120 120

100 100

80 80
Load (kN)

Load (kN)
60 60

40 40
Experimental Experimental
Numerical Numerical
20 20

0 0
0 1 2 3 4 0 1 2 3 4
Deflection (mm) Deflection (mm)
(a) (b)

140 140

120 120

100 100

80 80
Load (kN)
Load (kN)

60 60

40 40
Experimental
Numerical
20 Experimental 20
Numerical
0 0
0 1 2 3 4 0 1 2 3 4
Deflection (mm) Deflection (mm)
(c) (d)
(continued)

RC beam and reducing the mid-span deflection. On the 4.1 Proposed model by ACI 440.2 R-08
other hand, the increase in CFRP layer-length serves to ACI 440.2 R-08 (ACI, 2008) uses a rectangular stress block
stiffen the RC beams, which results in lowering the similar to that used in normal RC beams. To prevent
interfacial shear stress and restricting crack propagation, debonding of the FRP laminate, a limitation should be placed
hence increasing the load-carrying capacity and decreasing on the strain level developed in the laminate.
the mid-span deflection. Table XVI shows the mechanical properties for concrete,
In general, the numerical results from the non-linear FE steel reinforcement and CFRP materials used in the analytical
analysis are in good matching with the experimental results method.
recorded in the experimental program (Obaidat, 2011). The following step-by-step algorithm is used to predict
the behavior of un-strengthened and strengthened RC
beams using ACI 440.2 R-08 code (ACI, 2008) for beam
PL-1.0-0.43:
4. Analytical investigation Step 1: FRP system design material properties:
In this section, a theoretical analysis is carried out
according to the analytical model ACI 440.2 R-08 (ACI, f fu ¼ CE f fu ¼ 2660 MPa
2008), which is the American code that tends to estimate
the behavior of FRP-externally strengthened reinforced
« fu ¼ CE « fu ¼ 0:01615
concrete beams.

191
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Figure 9
140 140

120 120

100 100

80 80
Load (KN)

Load (KN)
60 60

40 40
Experimental Experimental
Numerical Numerical
20 20

0 0
0 1 2 3 4 0 1 2 3 4

Deflection (mm) Deflection (mm)

(e) (f)

140

120

100

80
Load (KN)

60

40
Experimental
Numerical
20

0
0 1 2 3 4
Deflection (mm)
(g)

Notes: (a) C-1.0; (b) PL-1.0-0.43; (c) PL-1.0-0.85; (d) PH-1.0-0.36; (e) S-1.0-0.27; (f) C-1.5; (g) PL-1.5-0.85

Table XI Validation of numerical model

Ultimate load (KN) Difference (%) Deflection at failure load (mm) Difference (%)
Specimen Exp. No. (Exp-No.)  100/Exp Exp. No. (Exp-No.)  100/Exp
C-1.0 92.905 92.876 10.03 2.814 3.116 4.48
PL-1.0-0.43 117.760 112.724 14.27 2.497 2.757 12.08
PL-1.0-0.85 122.148 122.836 0.56 2.564 2.459 16.57
PH-1.0-0.36 116.216 119.840 3.11 2.243 2.2933 7.25
S-1.0-0.27 100.226 111.601 11.34 2.659 2.7000 0.241
C-1.5 87.834 97.744 11.28 2.180 2.121 12.70
PL-1.5-0.85 119.722 125.083 4.47 1.925 2.099 9.02

192
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Table XII Percentage change values of failure load obtained from experimental Figure 12 Numerical results of first crack load
approach and numerical simulation
Change of failure Change of failure
Beam load in the experiment (%) load in the numerical (%)
C-1.0 – –
PL-1.0-0.43 21.10 17.60
PL-1.0-0.85 31.47 24.39
PH-1.0-0.36 20.05 29.03
C-1.5 – –
PL-1.5-0.85 26.63 21.85

Table XIII Percentage change values of deflection obtained from experimental

approach and numerical simulation
Step 2: Preliminary calculations:
Change of deflection Change of deflection
Beam in the experiment (%) load in the numerical (%)
Concrete properties:
C-1.0 – –
PL-1.0-0.43 12.69 13.02 qffiffiffiffi
PL-1.0-0.85 9.75 26.71 Ec ¼ 4700
0
f c ¼ 30712:26 MPa
PH-1.0-0.36 25.45 35.87
C-1.5 – –
PL-1.5-0.85 13.24 1.04 Area of reinforcing steel:

As ¼ 401:92 mm2
Figure 10 Comparison between experimental and numerical mid-span
deflection for all RC beams Area of externally bonded CFRP:

Af ¼ 168 mm2

Step 3: Existing state of strain on the soffit:

MDl ð df  kdÞ
« bi ¼ ¼ 0:0124
Icr Ec

Step 4: Design strain of CFRP system:

sffiffiffiffiffiffiffiffiffiffiffiffiffi
0
fc
Figure 11 Comparison between experimental and numerical mid-span « fd ¼ 0; 41  0:9 « fu
n tf Ef
load for all RC beams

« fd ¼ 0; 00557  0:0145

Step 5: Estimation of the depth of neutral axis:

An initial estimate of the depth of neutral

axis 00 c00 is 0.2 d. However, the value of 00 c00 is
taken as 0.339 d. The value of 00 c00 is assumed
here, the value of 00 c00 obtained finally in Step
9. Here, taking C = 0.2 d.

Step 6: Effective level of strain in the CFRP ma-

terial composites:

193
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Figure 13 Typical cracking signs occurring in FE models

Figure 14 Numerical crack pattern of RC beam C-1.0

194
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Figure 15 Numerical crack pattern of RC beam PL-1.0-0.43

« s ¼ 0:0106
 
df  c Step 8: Stress level in steel reinforcement and
« fe ¼ « cu  « bi  « fd CFRP plate:
c

Hence,
f s ¼ Es « s  f y

« fe ¼ 0:000241
f s ¼ 2173 MPa > 562 MPa
As section failure is controlled by CFRP; con-
Hence,
crete strain at failure « c may be less than
0.003 and is calculated as follows:
f s ¼ 562 MPa
 
c Step 9: Internal force resultants and check for
« c ¼ ð« fe 1 « bi Þ
df  c equilibrium (as mentioned in Step 5):

« c ¼ 0:003 Force equilibrium is verified by checking the

initial estimate of 00 c00 with following
Step 7: Strain in existing steel reinforcement: equation:

  As f s 1 Af f fe
dc c¼ 0
« s ¼ ð« fe 1 « bi Þ a1 f c b 1 d
df  c

195
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Figure 16 Numerical crack pattern of RC beam PL-1.0-0.85

where, considered is equal to 0, 22 d, which yielded a

value of 57.08 mm and which matches with the
value of c = 57.08 mm obtained in Step 9.
d = 262 mm
Step 11: Flexural strength components con-
a1 and b 1 are stress block factors:
tribution of steel in flexure can be given as
follows:
where,
 
b c
a1 ¼ 0: 84 Mns ¼ As f s d 1
2
b 1 ¼ 0:7484
Mns ¼ 54:35 KN:m
c = 57.080 mm
Contribution of CFRP in flexure can be given by:
00 00
Step 10: Adjustment of value of c until force
equilibrium is satisfied:

Steps 6-9 are repeated several times with val-  

b c
ues of 00 c00 ranging from 0.2 to 0.4 until force Mnf ¼ Wf Af f fe df  1
equilibrium is achieved. The value of 00 c00 2

196
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Figure 17 Numerical crack pattern of RC beam PH-1.0-0.36

Figure 18 Numerical crack pattern of RC beam S-1.0-0.27

197
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Figure 19 Numerical crack pattern of RC beam C-1.5

Figure 20 Numerical crack pattern of RC beam PL-1.5-0.85

198
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Table XIV Different strengthening schemes used in parametric study Mnf ¼ 1:585 KN:m
(modulus of elasticity = 165 GPa, thickness = 1.4 mm)
Step 12: Design flexural strength of the section:
Beams ID CFRP length (mm)
   
C1 0.0 – b c b c
PL-1.0-0.43 1,400 Mn ¼ As f s d  1 1 Wf Af f fe df  1
2 2
Configuration A 1,125
Configuration B 1,050 where:

Wf ¼ 0:85

Mn ¼ 55:935 KN:m

Figure 21 Finite element model of CFRP-externally strengthened RC beams

199
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Figure 22 Effect of CFRP lengths on strengthened RC beam f Mn  Mu

140 where, f ¼ 0:9

120
f Mn ¼ 50:341 KNm  47:104 KN: m
100
The strengthened section is now capable to sustain the revised
80 moment.
Load (KN)

60 4.2 Evaluation of analytical model

The ultimate load capacity of the tested specimens is calculated
40
C-1.0 using ACI 440.2 R-08 code (ACI, 2008) for RC beams
Configuration B retrofitted with layers of CFRP, under three-point bending.
20 Configuration A
PL-1.0-0.43
Table XVII represents the comparison between the different
0 estimates of the ultimate load capacity from analytical and
0 1 2 3 4
numerical approaches. It is clear that there is a good agreement
Deflection (mm)
between the two approaches.

5. Conclusion
In the current study, FE simulation code ANSYS and
Figure 23 Development of the mid-span deflection of RC beams (C-
analytical model ACI 440.2 R-08 are used to numerically
1.0, PL-1.0-0.43, Configurations A and B)
evaluate the structural response of a CFRP-strengthened RC
beam. In addition, a parametric study consisting of four models
is conducted.
The validated FE model is used to evaluate the load–
deflection behavior and crack pattern of the control and
strengthened RC beams. It is found that the FE models could
predict ultimate load within (0.03-11.34 per cent) variation
with that of experimental study and deflection at failure within
(0.241-9.02 per cent) variation with that of experimental
investigation. However, the general behaviors of the FE models
are in good agreement with the experimental test data. On the
other hand, the increase in number of layers of CFRP leads to
reduce the mid-span deflection and the density of cracks in RC
beams. Similarly, increasing in elastic modulus of CFRP
composites tends to reduce the mid-span deflection and the
Table XV Values of failure load of different strengthening schemes using
density of cracks in RC beams. Therefore, the developed model
numerical simulation
gives very good results compared to the experimental results,
Beam ID Failure load (KN) Increasing strength (%) from the beginning to the end of the loading stages of cracks in
C-1.0 92.9054 – the RC beams are observed.
Configuration B 98.868 7.063 From numerical study, it can be noted that the load-carrying
Configuration A 101.864 9.642 capacity of RC beam increases with increasing of the CFRP
PL-1.0-0.43 112.7245 21.332 length. The percentage increase in maximum load-carrying
capacity of PL-1.0-0.43, Configurations A and B are found to
be 21.332, 9.642 and 7.063 per cent, respectively, as compared
to that of control beam (C-1.0). This is because of the
increased CFRP length that leads to enhance the material

Table XVI Mechanical properties for concrete, steel reinforcement and

Table XVII Comparison of NLFEA results with analytical results
CFRP materials
Ultimate Ultimate load
Specified compressive strength of concrete f’c 42.7 MPa
load (KN) analytical Difference
Modulus of elasticity of steel (Es) 2.05 105 MPa
Beam (KN) NLFEA approach (%)
Specified yield strength of steel reinforcement fy 562 MPa
Ultimate axial strain of unconfined concrete ecu 0.003 PL-1.0-0.43 112.7245 125.866 10.440
Modulus of elasticity of CFRP Ef 165,000 MPa PL-1.0-0.85 122.836 126.551 2.935
Ultimate tensile strength of CFRP f *fu 2,800 MPa PH-1.0-0.36 119.84 126.302 5.116
Rapture strain of CFRP ’e*fu 0.017 S-1.0-0.27 111.601 112.334 0.652
Assumed thickness of CFRP layer tf 1.4 mm PL-1.5-0.85 125.083 149.168 16.146

200
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

properties (concrete, steel, CFRP, epoxy adhesive); method”, Journal of Structural Engineering, Vol. 40 No. 2,
consequently, the element stiffness matrix of an externally pp. 169-184.
retrofitted RC beam is most robust with the greatest CFRP Dong, J., Wang, Q. and Guan, Z. (2013), “Structural behaviour
length, which in turn restricts crack growth, thus improving the of RC beams with external flexural and flexural-shear
strength of RC beam and reducing mid-span deflection. strengthening by FRP sheets”, Composites Part B: Engineering,
The use of an analytical model ACI 440.2 R-08 tends to Vol. 44 No. 1, pp. 604-612.
predict ultimate load particularly for RC beam S-1.0-0.27. Dahmani, L., Khennane, A. and Kaci, S. (2010), “Crack
Generally, the values predicted using equations of ACI identification in reinforced concrete beams using ANSYS
440.2 R-08 approach correlate well with the numerical software”, Strength of Materials, Vol. 42 No. 2, pp. 232-240.
simulation values. El-Sayed, A.K. (2014), “Effect of longitudinal CFRP
strengthening on the shear resistance of reinforced concrete
beams”, Composites Part B: Engineering, Vol. 58, pp. 422-429.
References Esfahani, M.R., Kianoush, M.R. and Tajari, A.R. (2007),
“Flexural behaviour of reinforced concrete beams
Abbas, A.L. (2010), “Non–linear analysis of reinforced concrete
strengthened by CFRP sheets”, Engineering Structures,
beams strengthened with steel and CFRP plates”, Diyala
Vol. 29 No. 10, pp. 2428-2444.
Journal of Engineering Sciences, Vol. First, pp. 249-256.
Fathelbab, F.A., Ramadan, M.S. and Al-Tantawy, A. (2011),
Aboutaha, R.S., Wattanadechachan, P. and Kim, S.H. (2003),
“Finite element modeling of strengthened simple beams
“Flexural ductility of CFRP strengthened concrete beams-
using FRP techniques”, Concrete Research Letters, Vol. 2
experimental investigation”, Special Publication, Vol. 213,
No. 2, pp. 228-240.
pp. 207-220.
Gopinath, S., Prabha, M., Murthy, A.R. and Iyer, N.R.
Abu-Obeidah, A., Hawileh, R.A. and Abdalla, J.A. (2015),
(2017), “Modelling of RC beams strengthened with basalt
“Finite element analysis of strengthened RC beams in shear
reinforced concrete”, Journal of the Institution of Engineers
with aluminum plates”, Computers & Structures, Vol. 147, (India): Series A, Vol. 98 No. 3, pp. 285-291.
pp. 36-46. Hawileh, R.A., Rasheed, H.A., Abdalla, J.A. and Al-Tamimi,
ACI (2008), “ACI 440.2-08 guide for the design and construction A.K. (2014), “Behavior of reinforced concrete beams
of externally bonded FRP systems for strengthening concrete strengthened with externally bonded hybrid fiber reinforced
structures”, ACI, Farmington Hills, MI. polymer systems”, Materials & Design, Vol. 53, pp. 972-982.
Ahmed, E., Sobuz, H.R. and Sutan, N.M. (2011), “Flexural Hawileh, R.A., Naser, M.Z. and Abdalla, J.A. (2013), “Finite
performance of CFRP strengthened RC beams with different element simulation of reinforced concrete beams externally
degrees of strengthening schemes”, International Journal of strengthened with short-length CFRP plates”, Composites
Physical Sciences, Vol. 6 No. 9, pp. 2229-2238. Part B: Engineering, Vol. 45 No. 1, pp. 1722-1730.
Al-Hadithy, L.K. and Al-Ani, M.M. (2016), “Influence of Ibrahim, A.M. (2009), “Finite element analysis of reinforced
soffit bonded CFRP strips on shear capacity and failure type concrete beams strengthened with CFRP in flexural”, Diyala
of RC beams without stirrups”, American Scientific Research Journal of Engineering Sciences, Vol. 2 No. 2, pp. 88-104.
Journal for Engineering, Technology, and Sciences, Vol. 26 Ibrahim, A.M. and Mahmood, M.S. (2009), “Finite element
No. 4, pp. 135-150. modeling of reinforced concrete beams strengthened with
Al-Zaid, R.Z., Al-Negheimish, A.I., Al-Saawani, M.A. and El- FRP laminates”, European Journal of Scientific Research,
Sayed, A.K. (2012), “Analytical study on RC beams Vol. 30 No. 4, pp. 526-541.
strengthened for flexure with externally bonded FRP Jayajothi, P., Kumutha, R. and Vijai, K. (2013), “Finite element
reinforcement”, Composites Part B: Engineering, Vol. 43 analysis of FRP strengthened RC beams using ANSYS”, Asian
No. 2, pp. 129-141. Journal of Civil Engineering, Vol. 14 No. 4, pp. 631-642.
ANSYS (2007), ANSYS, Release Version 12, a Finite Element Mohamed, R.A., Zana, A.A. and Jaza, H.M. (2017), “Nonlinear
Computer Software and User Manual for Nonlinear Structural finite element analysis of reinforced concrete beams strengthened
Analysis, ANSYS, Canonsburg, PA. with externally bonded steel plate using ANSYS”, Sulaimani
Ashour, A.F., EL-Refaie, S.A. and Garrity, S.W. (2004), Journal for Engineering Sciences, Vol. 4 No. 4, pp. 41-50.
“Flexural strengthening of RC continuous beams using Musmar, M.A., Rjoub, M.I. and Hadi, M.A. (2006), “Nonlinear
CFRP laminates”, Cement and Concrete Composites, Vol. 26 finite element analysis of shallow reinforced concrete beams
No. 7, pp. 765-775. using solid65 element”, Elastic, Vol. 25743, pp. 1-3.
Atea, R.S. (2018), “A case study of flexural performance of Panigrahi, A.K., Biswal, K.C. and Barik, M.R. (2014),
reinforced concrete beams bonded with steel plates with “Strengthening of shears deficient RC T-beams with
different thickness”, Journal of Materials Research and externally bonded GFRP sheets”, Construction and Building
Technology, Vol. 8 No. 1, pp. 22-32. Materials, Vol. 57, pp. 81-91.
Banjara, N.K. and Ramanjaneyulu, K. (2017), “Experimental Obaidat, Y.T. (2011), “Structural retrofitting of reinforced concrete
and numerical investigations on the performance evaluation beams using carbon fibre reinforced polymer”, PhD Thesis,
of shear deficient and GFRP strengthened reinforced Department of Construction Sciences, Structural Mechanics.
concrete beams”, Construction and Building Materials, Saifullah, I., Hossain, M.A., Uddin, S.M.K., Khan, M.R.A.
Vol. 137, pp. 520-534. and Amin, M.A. (2011), “Nonlinear analysis of RC beam for
Basappa, U. and Rajagopal, A. (2013), “Modeling of CFRP different shear reinforcement patterns by finite element
strengthened RCC beam using the nonlinear finite element analysis”, IJCEE, Vol. 11 No. 1, pp. 63-74.

201
 Finite element analysis World Journal of Engineering
Sabiha Barour et al. Volume 17 · Number 2 · 2020 · 183–202

Sasmal, S., Kalidoss, S. and Srinivas, V. (2012), “Nonlinear finite Wang, R.M., Zheng, S.R. and Zheng, Y.P. (2011), Polymer
element analysis of FRP strengthened reinforced concrete Matrix Composites and Technology, Woodhead Publishing Series
beams”, Journal of the Institution of Engineers (India): Series A, in Composites Science and Engineering, USA.
Vol. 93 No. 4, pp. 241-249. Wolanski, A.J. (2004), “Flexural behaviour of reinforced and
prestressed concrete beams using finite element analysis”,
Sharif, A., Al-Sulaimani, G.J., Basunbul, I.A., Baluch, M.H.
PhD thesis, Marquette University.
and Ghaleb, B.N. (1994), “Strengthening of initially loaded
Zidani, M.H.B., Belakhdar, K. and Tounsi, A. (2015), “Finite
reinforced concrete beams using FRP plates”, Structural
element analysis of initially damaged beams repaired with
Journal, Vol. 91 No. 2, pp. 160-168.
FRP plates”, Composite Structures, Vol. 134, pp. 429-439.
Taljsten, B. (2003), “Strengthening concrete beams for shear
with CFRP sheets”, Construction and Building Materials, Corresponding author
Vol. 17, pp. 15-26. Sabiha Barour can be contacted at: baroure18@hotmail.fr

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com

202