ECCM15 - 15TH EUROPEAN CONFERENCE ON COMPOSITE MATERIALS, Venice, Italy, 24-28 June 2012

EFFECTIVE STIFFNESS APPROACH OF FRP REINFORCED

CONCRETE BEAMS

S. Khalfallah 1*
1
Department of Civil Engineering, University of Jijel, Algeria, B.P 98 Ouled Aissa University of Jijel,
18000 Jijel, Algeria.

Keywords: Effective moment of inertia, analytical expression, service loading conditions,

tension stiffening effect.

Abstract
Current provisions in ACI Committee for calculating flexural structures deflections
reinforced with fiber reinforced polymer (FRP) bars or reinforcements require the use of an
effective moment of inertia. Many equations have already been developed and they have
shown significant difference between them. In this study, CEB-FIP and Benmokrane formula
have been compared and the variation between them is principally due to bond character and
tension stiffening effects.
For this object, an empirical expression is proposed quantifying the effective moment of
inertia for flexural members. Obtained results show a favorite accuracy using the presented
effective moment of inertia model in-service loading conditions.

1 Introduction
The use of FRP bars to reinforce concrete structures has received a great deal of attention in
recent years. The effectiveness of FRP composites for strengthening beams is attributed for
various advantages, such as: high efficiency, high resistance to corrosion attacks under long-
term service loading, high tensile capacity, low weight, no conductivity of electricity and ease
of application and transport.
FRP reinforcements have been developed to replace steel bars in particular applications
mainly in corrosion-prone RC structures. Among rheological disadvantages, compared to
steel bars, FRP reinforcements possess a low modulus of elasticity that leads to higher
reinforcement strains, wider cracks and larger deflections. Therefore, the serviceability of
limit state may often used in the design of FRP reinforcement and steel bars as amount of
structure reinforcements. Furthermore, bond and tension stiffening characteristics of FRP re-
bars are not well quantified according to experimental data.
The design of FRP- bars reinforced beams requires an important attention in serviceability
conditions due to the cracking phenomenon and low stiffness of FRP-bars. For these reasons,
this category of structures design is often based on the evolution of stresses of the FRP-
reinforcing bars. In the cracking range of the behavior, the low stiffness of FRP-bars, in fact,
involves an intensive initiation and propagation of cracks therefore larger widths and greater
deflections can be obtained compared to steel reinforced concrete beams.
In this subject, several methods are available in literature for computing deflections in FRP
reinforced flexural structures taking into account the nonlinear effects, such as: the cracking
of concrete, the tension stiffening propriety and the bond character. Due to the complexities of

1
ECCM15 - 15TH EUROPEAN CONFERENCE ON COMPOSITE MATERIALS, Venice, Italy, 24-28 June 2012

computations to be performed, these procedures can’t be adopted by engineers. Many works

have published methods that more accurately estimate responses of structures and can be
easily executed by engineers in this field. Many engineers inspire beams capacities from the
fundamental structural analysis and conservative material assumptions. In general, these
procedures underestimate the really capacities of beams.
The aim of this contribution presents an empirically expression based on the Branson’s
equation. Obtained results were carried out to show the behaviour of concrete beams
reinforced by FRP-bars. Furthermore, predicted service load-deflections using selected
models are compared with maximum deflections at mid-span for four different beams and
their corresponding predicted values using CEB-FIP model [1] and Benmokrane model [2]
largely used in the literature.

2. Comparing effective moment of inertia relationshipS

The expression of the effective moment of inertia (Ieff) is the commonly approach used to
compute RC members deflections. Among, two relationships largely used in literature are
selected for comparison in this paper: the CEB-FIP approach and the Benmokrane’s method.
Firstly, the computed deflection according the CEB-FIP model uses the bilinear equation
expressed by:
 c  (1   b )1   b  2 (1)
where  b 11 2 M cr , which 1 , 2 are coefficients that depict the bond nature of bars and
Ma
the time of loading, eventually 0.8 and 0.5 for short and long term loading; respectively. Ma
and Mcr are the maximum applied moment and the cracking moment, respectively.
1  ks1 c and  2  k s 2 c are coefficients of the bilinear law and ks1 and ks2 depend on
EFRP
parameters :   ,   As and  '  As' . It should be noted that tension-stiffening
Ec bd bd
contribution in equation (1) is taken into account with an explicitly manner. The cracking
moment is computed when the maximum tensile stress of concrete reaches its strength value
in tension.
f Ig
M cr  t (2)
yt
Where Ig is the moment of inertia of the gross section, yt is the distance from the neutral axis
to the extreme tension fiber and ft is the tensile strength of concrete.
Computed deflections are primordially based on previsions intended for steel reinforced
concrete elements. Thus, ACI 440 code [3] recommends the use of Branson’s equation to be
consistent with the existing ACI Building Code Requirements [4] for Structural Concrete.
The expression used in the cracking range of the behavior is computed by:
Pa
 max  (3L2  4a 2 ) (3)
24 Ec I eff
with P is the applied load, a is the slenderness of the point load, L is the span of the beam, Ec
is the elastic modulus of concrete and Ieff is the effective moment of inertia.
However, the primary interest is given to the rational evaluation of the effective moment of
inertia, Ieff. The deflection relationship was empirically derived to represent a gradual
digression from the gross un-cracked moment of inertia Ig to the cracked moment of inertia
Icr. The following approach was originally proposed by Branson [4], as:
3 3
Ieff  M cr  I g (1 M cr  )Icr  I g (4)
 Ma   Ma 

2
ECCM15 - 15TH EUROPEAN CONFERENCE ON COMPOSITE MATERIALS, Venice, Italy, 24-28 June 2012

with Ma is the applied service load moment.

Some investigators have already observed that the previous equation was not generally in an
agreement with experimental data of FRP reinforced concrete structures. For this object,
Benmokrane et al. [2] have introduced two empirical factors in the equation (4) to correlate
results with experimental data of reinforced concrete beams reinforced with GFRP bars.
3
Ig  M cr 
Ieff  Icr (   Icr)    Ig (5)
  Ma 
where  and  were determined to be 0.84 and 7, respectively.
Later, Gao [6] further modified the equation (5) to agree obtained deflections with experiment
results. This modification feels the object of ACI guides [4] and [7], both codes subsequently
recommended modifications to be reconciled with test data.
The new expression of the effective moment of inertia is then given by:
3 3
M  M 
I eff   cr   r I g  (1   cr  ) I cr  I g (6)
 Ma   Ma 
E frp
with  r  0.5(  1) , Efrp and Es are elastic modulus values of the FRP and steel bars. The
Es
coefficient  r was initially set equal to 0.6 for glass FRP bars [8] and soon after it has been
E
expressed by  r   b ( fpr  1) while  b is the bond coefficient assumed equal to 0.5.
Es
However, Yost et al. [9] have neglected the dependence on bond character and set it equal to

 b  0.064  0.13 to account the reinforcing bars ratio  and the balanced reinforcing
b
ratio  b in consideration. Recent changes by ACI 440 [3] recommend using the reduction

factor with  r  0.2 based on a comprehensive statistical fit of existing beam deflection
b
data.
An alternative approach improving computed deflections of FRP beams has been proposed to
stiff the member by increasing the cubic power in Branson's original equation (Equ.4) to
higher value [10]. However, the prediction of results for lower load levels is not further
considered [11].
The challenge consists of developing an expression to be satisfactorily applied to different
types of FRP beams reinforced with various reinforcement ratios. To improve analytical
results, it must be reducing the tension stiffening contribution in Branson's model. This
procedure takes into account some variables incorporated in the effective moment of inertia
formula. However, different models for the effective moment of inertia, including the
effective contribution of cracked concrete and the participation of the tension stiffening
option, have been proposed to define flexural behavior of RC beams. The object can be
reached by multiplying the gross moment of inertia by the reduced factor  r (equ.6) that
I
effectively reduces the I g ratio to  r g [12-13]. The literature has already presented many
Icr I cr
attempts to modify the Branson's expression basing on the consistence of the tension
stiffening and bond effects.

3
ECCM15 - 15TH EUROPEAN CONFERENCE ON COMPOSITE MATERIALS, Venice, Italy, 24-28 June 2012

This study develops an empirical expression inspired from the Branson’s approach based on
the affectation to an appropriate value of the I g ratio, which can control the tension stiffening
Icr
in the Branson model. In addition, the paper identifies the limitations of both the existing and
past methods ACI 440 [3,4,7]. Results are also compared to a more general unified approach
that can be equally applied for both steel and FRP reinforced concrete [11].

3. Analytical approach
For the moment value less than the cracking one, the beam is assumed un-cracked and
behaves homogeneously and elastically. In this range of the behavior, the slope of the applied
moment-mid span deflection curve is proportional to the moment of inertia of the un-cracked
section. The member firstly cracks when the applied moment reaches the cracking moment
value, Mcr, then the tensile stress in the extreme fiber reaches the flexural tensile strength of
the concrete.
After the formation of the first crack, concrete between adjacent cracks can still resist to
tensile forces. This phenomenon known as the tension stiffening effect is primordially
characterized by an important change in the curvature of the beam. Really, it’s reasonable that
there was a balance between two members of the relationship giving the effective moment of
inertia.
From the typically response curve of applied moment-mid span deflection, it’s seen that is
possible to introduce parameters quantifying effective moment of inertia to improve the
predicted response of flexural beams reinforced with FRP bars. In the same manner, it’s also
possible to bring nearer predicted response curves to the experimental ones, which can be
obtained considering an increase of the intact section and simultaneously a decrease of the
cracking section. In this sense, the Benmokrane’s expression has empirically modified
Ec
introducing two following coefficients of correction that are:  1  (1  ) and
E FRP
Ec
 2  (1  ) , respectively.
E FRP
In this study, an analytical relation of the effective moment of inertia is established addicting
these terms of correction. The approach most commonly used to reduce tension stiffening to
realistic levels where the gross moment of inertia Ig is multiplied by a correction factor that
effectively reduced (Mcr/Ma) ratio. Since tension stiffening is directly related to applied
moment (Mcr/Ma) and the reducing of this ratio will also decreases the tension stiffening
effect in the cracking stage. Most attempts to improve the effective moment of inertia formula
have been done but not to be consistent with the principal mechanisms underlying the tension
stiffening effect. The aim of this paper is to develop an expression for reducing factor
(Mcr/Ma) based on the quantification of the (Mcr/Ma) ratio and controlling the tension
stiffening effect contribution. The corresponding equation can be written as,
3 3
Ieff   1  M cr  I g   2 (1 M cr  ) Icr  I g (7)
 Ma   Ma 
The use of this expression, it’s appeared that the balance between components, Ig and Icr of the
effective moment of inertia relation is more quantified. The contribution of the cracking
behavior is herein very considered and the contribution cracked concrete is strictly evaluated
in the cracking range. Eventually, this process leads to the minimization of the tension
stiffening contribution.

4. Results of equations

4
ECCM15 - 15TH EUROPEAN CONFERENCE ON COMPOSITE MATERIALS, Venice, Italy, 24-28 June 2012

This work is intended to develop an analytical solution for calculating the load-deflection
response of FRP reinforced concrete beams using the above expression. Comparisons are
made between analytical formula and experimental results of beams subjected to four-point
bending.
To show the accuracy of the three analytical approaches, it’s necessary to compare their
results with experimental data. Geometrical and mechanical properties are summarized in
Tables 1 and 2, respectively (Fig. 1). Beams are tested as: Beam 1 [14], Beam 2 [15], Beam 3
[16] and Beam 4 [17].
The Predicted load versus mid-span deflection using bilinear (CEB-FIP) model,
Benmokrane’s approach, this study and experimental data for beams are shown in Figs. 2-5.
The interest of this comparison is to confirm the accuracy of the approach presented and its
performance based on the evaluation of ultimate loads is regrouped in the table (3).

d P P
h

Afrp a L-2a a
●
● ●

Figure 1. Cross section and loading of beams studied.

Beam N° L [m] a[mm] b[mm] h[mm] d[mm] Ig[mm4]

1 3.4 1200 500 180 145 2.43E+8
2 2.1 700 154 254 222 2.10E+8
3 3.05 1067 127 304 273 3.00E+8
4 1.55 625 150 300 250 3.38E+8
Table 1. Geometrical properties of beams.

Beam N° f'c Ec Efrp ffrp Afr  (%)

[MPa] [MPa] [MPa] [MPa] [mm²]
1 30 24650 42000 886 887 1.22
2 42 29160 34000 586 530 1.55
3 29 27240 26220 724 724.5 2.09
4 34 29000 100000 1200 390 1.04
Table 2. Mechanical properties of beams.

Figure 2. Load-mid span deflections for beam 1. Figure 3. Load-mid span deflections for beam 1and beam 2.

5
ECCM15 - 15TH EUROPEAN CONFERENCE ON COMPOSITE MATERIALS, Venice, Italy, 24-28 June 2012

The figure 2 shows in general an acceptable agreement with the experimental curve.
Particularly in this case, analytical predictions overestimate the stiffness of the beam beyond
the applied load equal to 10 kN. But, figures 3-5 show in general a good agreement between
experimental and analytical load-deflection curves. The Benmokrame’s approach for the
beam 1 presents an accurate solution compared to experimental deflections. Curves compared
using this approach show stiffer post-cracking of the beams. The analytical load-deflection
curves were presented a well concordance up about 70% of the ultimate load. It is evident
from figures 3-5 that deflections of this procedure are in excellent agreement with the
experimental results during the cracking range.
The CEB-FIP model and Benmokrane’s approach show an under prediction of experimental
results of FRP reinforced concrete beams (Fig. 3), particularly at the level of application of
limit load, which occur a problem for the structural integrity and safety. As the service load is
increased the accuracy of the deflection predicted becomes very significant. The figure 4
reveals that CEB-FIP model and Benmokrane’s method become consistent basing on the
predicting deflection values. The model presented predicts the response of the beam 3 up to
67%. The CEB-FIP model and Benmokrane’s equation are seen to slightly overestimate the
stiffness resulting with less deflection (Fig. 4).

Figure 4. The load-mid span deflection for beam 3. Figure 5. The load-mid span deflection for beam 4.

In this case, analytical deflections are predicted up to 80% and an improvement is widely
observed compared to results obtained by CEB-FIP model and by that Benmokrane’s
approach (Fig. 5).
In this case, this approach shows an excellent concordance by the comparison made on the
convergence between results obtained by this procedure and experimental data. Particularly,
the following table regroups the ultimate load for the four beams studied.

Beam N° 1 2 3 4
CEB-FIP model 72.50 69.82 75.84 73.75
Benmokrane approach 87.00 75.86 78.56 81.25
This study 98.00 88.56 91.98 92.64

Table 3. Comparison in % of calculated /experimental load.

5. Summary and conclusions

In this study, a simple analytical approach is derived to predict load-deflection response of
FRP bars reinforced concrete beams. Obtained results of the four beams with different
mechanical properties revealed that the current models for predicting the deflection of beams
reinforced by FRP bars under-estimate deflections of beams. The formulation introduces
correction factors in Benmokrane’s equation reducing the tension stiffening effect to
reasonable level by decreasing of the (Ig/Icr) ratio. The developed model can be applied to

6
ECCM15 - 15TH EUROPEAN CONFERENCE ON COMPOSITE MATERIALS, Venice, Italy, 24-28 June 2012

both RC and FRP reinforced beams, indifferently and a wide range of analytical load-
deflection curves are compared, the present model, the CEB-FIP, Benmokrane et al. equations
with the experimental data. The comparison showed that this approach can lead to a favorable
agreement with experimental results.
The response established by this procedure is overestimated experimental results for the beam
1. Probably, it’s due to geometrical properties that having the rapport h/b <0.36 but the rest of
beams introduce an admirable concordance between analytical results and those experimental
ones. Two equations, largely used in literature, are compared and shown a greatly difference
between them. This variation is well modeled by considering the tension stiffening effect and
the cracking effect, respectively. These parameters are directly included in the novel
relationship for predicting beam deflections and analytical results have confirmed the
accuracy of the model and estimates better deflections of beams reinforced by FRP bars
opposite to test data.

References
[1] Comité Euro-International du Béton et la Fédération Internationale de la Précontrainte
(CEB-FIP). (1990). Model code for concrete structures, MC-90, Thomas Telford House,
London.
[2] Benmokrane B., Chaallal O., Masmoudi R. (1996). ‘’Flexural response of concrete beams
with FRP reinforcing bars’’. ACI Struct Eng., 93(1), pp. 46-55.
[3] American Concrete Institute (ACI). (2006). ‘’Guide for the design and construction of
concrete reinforced with FRP bars’’. ACI 440.1R-06, ACI Committee 440, Farmington
Hills, Mich.
[4] American Concrete Institute (ACI). (2003). ‘’Guide for the design and construction of
concrete reinforced with FRP bars’’. ACI 440.1R-03, ACI Committee 440, Farmington
Hills, Mich.
[5] Branson, D.E. (1965). ‘’Instantaneous and time-dependent deflections of simple and
continuous reinforced concrete beams’’, HPR report N° 7, Part 1: Alabama Highway
Dept. Bureau of public roads, Auburn University.
[6] Gao D., Benmokrane B., Masmoudi R. (1998). ‘’A calculating method of flexural
properties of FRP-reinforced concrete beam’’. Technical report, Department of Civil
Engineering, University of Sherbrooke, Canada, 24 p.

[7] American Concrete Institute (ACI). (2001). ‘’Guide for the design and construction of
concrete reinforced with FRP bars’’. ACI 440.1R-01, ACI Committee 440, Farmington
Hills, Mich.
[8] Thériault, M., and Benmokrane, B. (1998). ‘’Effects of FRP reinforcement ratio and
concrete strength on flexural behavior of concrete beams’’. J. Compos. Constr., 2(1), pp.
7–16.
[9] Yost, J. R., Gross, S. P., and Dinehart, D. W. (2003). ‘’Effective moment of inertia for
glass fiber-reinforced polymer-reinforced concrete beams’’. ACI Struct. J., 100(6), pp.
732–739.
[10] Brown V.L., Bartholomew (1996). ‘’Long-term deflections of GFRP-reinforced concrete
beams, Fiber composite in infrastructures’’: Proc., 1st Int. Conf. On Composites in
Infrastructures, ICCI’96, H. Saadatmanesh and M.R. Ehsani, eds., Tucson, Ariz., pp. 389-
400.
[11] Bischoff, P. H. (2005a). ‘’Reevaluation of deflection prediction for concrete beams
reinforced with steel and fiber reinforced polymer bars’’. J. Struct. Eng., 131(5), pp. 752–
767.

7
ECCM15 - 15TH EUROPEAN CONFERENCE ON COMPOSITE MATERIALS, Venice, Italy, 24-28 June 2012

[12] Bischoff, P. H. (2005b). ‘’A Rational proposal for predicting beam deflection’’. Proc. 33
rd annual conference of the Canadian Society for Civil Engineering, Toronto, GC-299-
1/10.
[13] Bischoff, P. H. (2007). ‘’Deflection calculation of FRP reinforced concrete beams based
on modification to the existing Branson equation’’. Journal of Composites for
construction, ASCE, 11(1), pp. 4–14.
[14] Cosenza, E., Manfredi, G., and Realfonzo, R. (1997). ‘’Behavior and modeling of bond of
FRP rebars to concrete’’, J. Compos. for Constr., ASCE, 1(2), pp. 40–51.
[15] Swamy, N., and Aburawi, M. (1997). ‘’Structural Implications of Using GFRP Bars as
Concrete Reinforcement’’. Proceedings of the Third International Symposium on Non-
Metallic (FRP) Reinforcement for Concrete Structures (FRPRCS-3), Japan Concrete
Institute, V.2, 503-510.
[16] Nawy E.G., Neuwerth G.E. (1977). ‘’Fiberglass reinforced concrete slabs and beams’’.
Journal of the Structural Division, ASCE, Vol. 103, N° ST2, pp. 421-428.
[17] Maruyama, K., and Zhao, W. (1996). ‘’Size Effect in Shear Behavior of FRP Reinforced
Concrete Beams’’. Advanced Composite Materials in Bridges and Structures, Montreal,
Quebec, pp. 227-234;