Article

Journal of Fire Protection Engineering

22(3) 183–196
Elevated temperature ! The Author(s) 2012
Reprints and permissions:

deflection behavior sagepub.co.uk/journalsPermissions.nav

DOI: 10.1177/1042391512447045

of concrete members jfe.sagepub.com

reinforced with
FRP bars
MA Faruqi
Department of Civil and Architectural Engineering,
Texas A&M University-Kingsville, Kingsville, TX, USA

S Roy
Department of Civil and Architectural Engineering,
Texas A&M University-Kingsville, Kingsville, TX, USA

A Salem
Department of Mechanical Engineering, Rochester Institute of
Technology, Dubai

Abstract
In recent years there has been an increased interest in the use of fiber reinforced
polymer (FRP) materials in concrete members. However, the behavior of such members
in fire is still relatively unknown. Since this is the main reason limiting the widespread
use of FRP in buildings, the use of FRP in fire vulnerable structures needs additional
study. In this article, a model is developed that incorporates the temperature dependent
progressive changes of Elastic-modulus of FRP in predicting the deflection behavior of
FRP reinforced concrete structures within the range of practical elevated temperatures.
Predictions from the model correlate well with experimental results from the literature.
The new approach provides an additional tool to evaluate the deflection of FRP rein-
forced concrete structures in fire.

Keywords
Concrete, fiber-reinforced polymer reinforcement, fire, glass-transition, practical ele-
vated temperatures, deflection

Corresponding author:
MA Faruqi, Department of Civil and Architectural Engineering, Texas A&M University-Kingsville, MSC 194,
Kingsville, TX 78363, USA
Email: [email protected]

Downloaded from jfe.sagepub.com at Heriot - Watt University on January 6, 2015

184 Journal of Fire Protection Engineering 22(3)

Introduction
Deterioration of infrastructure resulting from corrosion of steel reinforcement in con-
crete has led to the use of fiber-reinforced polymer (FRP) composites as an alternative.
High strength to weight ratio and corrosion resistance of FRP rebars provides a
significant advantage over steel [1]. FRP reinforcement is used in bridges, multi-storied
buildings, industrial structures and parking garages to name a few. In order to design
FRP reinforced concrete structures subject to elevated temperatures, it is important to
understand some of the thermo-mechanical properties (stiffness, strength) of the FRP
reinforcement within the range of practical temperatures.
The design behavior of FRP reinforced (rebar) concrete structures is well known
at ambient temperatures [2–4]. However, its behavior at elevated temperatures is
quite complex and as a result the codes do not specify any fire guidelines. The
American Concrete Institute (ACI) code merely recommends that FRP reinforced
concrete structures must meet all building and fire code requirements that apply to
a typical reinforced concrete structure [4]. The British Standard BS 476 [5] and BS
9999 [6] consider deflection and structural fire resistance as factors to determine
beam failure. These codes respectively provide a maximum allowable deflection of
L/20 and a minimum of 90 min of fire resistance.
At higher temperatures, the concrete gets de-moisturized rapidly and produces
cracks. This causes the FRP rebar to burn and eventually de-bond [7]. Thermo-mechan-
ical behavior of FRP rebar depends on the polymer resin [7]. The polymer resin will
soften and the FRP bar reaches its glass transition temperature rapidly. Elastic modulus
(E-modulus) of the polymer decreases significantly when the temperature reaches and
exceeds glass transition temperature [7]. The glass transition region is the most signifi-
cant practical region of FRP for design purposes. This is because the system undergoes
significant plastic deformations beyond this region resulting in structural collapse
[8–10]. At this transition temperature, the resin is no longer able to transfer stresses
from concrete to fibers. This leads to increased crack widths and deflections. The deflec-
tion is directly related to the progressive changes in the E-modulus of FRP composites
under elevated temperatures. Unfortunately, this topic has received very little attention
from the research community and therefore lacks design specifications. The current
study examines the effect of fire on the deflection behavior of FRP reinforced concrete
structures and accounts for the temperature-dependent progressive changes in the
E-modulus of FRP within the range of practical elevated temperatures.

Materials and experimental setup

Experimental data from fire tests of concrete beams reinforced with FRP were
obtained by Abbasi and Hogg [11]. A brief description of these tests is provided
here for the convenience of the reader. Two FRP reinforced concrete beams were
cast for testing at elevated temperatures. Cross sections of the beams were 350  400
mm2 and the total length of the specimens were 4250 mm. Glass fiber reinforced
polymer (GFRP) #4 rebar was used. The first beam and second beams used special
GFRP rebar containing a vinyl ester and polyurethane resin matrix respectively, with

Downloaded from jfe.sagepub.com at Heriot - Watt University on January 6, 2015

Faruqi et al. 185

a 75 mm concrete cover. This ensured integrity at elevated temperatures. The E-mod-

ulus of the rebar was 41 GPa. The GFRP surface was wrapped with a helical fiber
having a pitch of about 25.4 mm to produce traction on the bar surface. To avoid shear
failure traditional #3 U shaped stirrups were used. There were seven bars at the bottom
of the beam and two at the top. All beams were designed to fail through the mechanism
of concrete crushing as recommended by ACI 440 [2]. This was accomplished with a
reinforcement ratio greater than the balanced ratio. A rf/rb ratio of 1.98 was used. A
control beam with dimensions and reinforcement similar to the first beam was also cast
and tested at room temperature. Type I cement was used in concrete mix, which had a
water to cement ratio of 0.38. To increase the strength of the concrete, a water reducing
admixture composition that was 1% of the cement was used. The average size of
coarse aggregate was 20 mm and that of fine aggregate, 10 mm. Further details of
the experimental setup and testing can be found in [11].

Mechanical properties of FRP in a fire

The mechanical properties of FRP composites vary significantly when subjected to
elevated temperatures. This is due to their molecular bonds [12]. In the lower
temperature range, the bonds are intact and therefore correspond to a horizontal
segment of E-modulus. This is known as the glassy state. However, as the tem-
perature increases, more bonds are weakened and a new state is reached [12]. This
new state is referred to as the leathery state. The region between glassy and leathery
state is known as the glass transition zone. As the temperature rises, further break
down of bond leads to the rubbery state. The region between leathery and rubbery
states is known as the leathery zone [12]. When temperatures are further elevated, a
decomposed state is reached via a rubbery zone. The most significant practical region
of FRP for design purposes in fire is the glass transition temperature. This is because
the system undergoes significant plastic deformations at and beyond this region [7].
Figure 1 shows the approximate states and polymer transitions at elevated tempera-
tures [13]. Figure 2 shows a generalized glass transition zone.

Proposed approach
Formulation of E-modulus
The E-modulus of FRP can be calculated as follows: A unidirectional composite
can be modeled by assuming fibers to be continuous and parallel throughout the
composite. Also, assume a perfect bond exists between fibers and the matrix. This
model is shown in Figure 3.
The strains experienced by fiber, matrix, and composite are assumed to be equal.
Therefore, the load carried by the total system is shared between fibers and matrix.
This means

Pc ¼ Pfrp ¼ Pf þ Pm ð1Þ

Downloaded from jfe.sagepub.com at Heriot - Watt University on January 6, 2015

186 Journal of Fire Protection Engineering 22(3)

Figure 1. Approximate states and transitions of a polymer at elevated temperatures.

gs: glassy state 250 C; gt: glass transition zone (approximate), 230 C  gt  435 C; ls: lea-
thery state and glass transition temperature (&435 C); lt: leathery zone; rs: rubbery state; rt:
rubbery zone, 600 C  rt  900 C; ds: decomposed state >900 C.

Figure 2. Generalized glass transition zone.

Figure 3. Model of unidirectional composite.

Downloaded from jfe.sagepub.com at Heriot - Watt University on January 6, 2015

Faruqi et al. 187

Considering the fibers to be glass, the equation in terms of stresses becomes

frp ¼ g ðAg =Afrp Þ þ m ðAm =Afrp Þ ð2Þ

Equating area fractions to volume fractions, differentiating with respect to

strain, substituting E-moduli, and simplifying yields

dfrp =d" ¼ ðdg =d"ÞVg þ ðdm =d"ÞVm ð3Þ

Efrp ¼ Eg Vg þ Em Vm ð4Þ

Therefore

Efrp =Em ¼ ðEg =Em  1ÞVg þ 1 ð5Þ

Noting that E m Vm is a standard term for matrix, that E lt Vlt is the state of
matrix in the leathery zone, and that temperature at a particular section is con-
sidered constant, then substituting Elt Vlt for Em Vm in equation (4) yields:

Efrp ¼ Eg Vg þ Elt Vlt ð6Þ

Development of deflection model at elevated temperatures

At higher temperatures, the concrete gets de-moisturized rapidly, dries, shrinks,
and produces cracks. This concrete behavior causes the FRP rebar in the beam to
be compromised. The resultant deflection depends on the shrinkage curvature and
the support conditions of the FRP reinforced system exposed to elevated tempera-
tures. Figure 4 shows the basic temperature-based shrinkage causing deflection.
This type of deflection can be expressed as:
Deflection ¼ constant  shrinkage curvature due to elevated temperature  L.

et ¼ KS et L ð7Þ

et ¼ 2Th e=ðEconc Ieff Þ ð8Þ

Where e is the distance from the tension reinforcement to the neutral axis
 
Th ¼ Afrp þ A0frp "h Efrp ð9Þ

Substitution of Th and et into equation (7), and simplifying yields

 hn  o i
et ¼ 2KS eL=Ieff Afrp þ A0frp "h Efrp =Econc ð10Þ

Downloaded from jfe.sagepub.com at Heriot - Watt University on January 6, 2015

188 Journal of Fire Protection Engineering 22(3)

One of the important factors in determining the properties of a composite is the

relative proportion of volume at various states and temperatures. Consider a unit
volume of FRP material at a particular temperature. The volume of the material
[13] at a practical state can be obtained as

Vg ¼ ð1  g Þ and Vlt ¼ ð1  g Þð1  d Þ ð11Þ

Substituting the volumes at glass and leathery transition states into equation (6)
yields

Efrp ¼ Eg ð1  g Þ þ Elt ð1  g Þð1  d Þ ð12Þ

Equations provided in Saafi [14] are not accurate for the special GFRP rebar
with a glass transition zone of 230–435 C. Therefore, the equations in [14] are
modified to address this new situation, as follows. Eg & MrEfrp, 20 C where
Efrp, 20 C is approximately 4.10  104 N/mm2 and

Mr  1; 0  T  100 ð13Þ

Mr  1  0:00188T; 100  T  300 ð14Þ

Mr  1  0:0023T; 300  T  435 ð15Þ

Mr  0; T 4 435 ð16Þ

Substituting equation (12) into equation (10) then provides

hn   o i     
et ¼ 2KS eL Afrp þ A0frp "h =Ieff Eg 1  g þ Elt 1  g ð1  d Þ =Econc
ð17Þ

Figure 4. Temperature-based deflection in a FRP reinforced concrete beam.

Downloaded from jfe.sagepub.com at Heriot - Watt University on January 6, 2015

Faruqi et al. 189

hn  o i
Considering that Eg/Econc ¼ ng; Elt/Econc ¼ nlt and 2KS eL Afrp þ A0frp =Ieff ¼ ,
gives the result:
    
et ¼ "h ng 1  g þ nlt 1  g  d  g d ð18Þ

Since g d & 0 as the product is very small, the expression becomes

    
et ¼ "h ng 1  g þ nlt 1  g  d ð19Þ

Where "h  0g Tfrp . For practical purposes, only temperatures up to 435 C (glass
transition temperature) are considered.
Therefore, Elt can be approximated
  as zero. This implies that nlt(1–g–d) ¼ 0
Substituting "h and ng 0g 1  g ¼  into equation (19) provides the deflection
in the glass transition zone:

et ¼ Tfrp ð20Þ

Comparison of results
Figures 5 and 6 show a comparison of the model with experimental results from the
literature [11] for beams 1 and 2, respectively. In Figure 5, the approximate glass
transition zone ranges from 300 C to 435 C with approximate model values respect-
ively ranging from 33.1 to 159.4 mm ( & L/28) as compared to 33.9 and 178.1 mm.
Similarly, for beam 2 the approximate model values respectively range from 31.1 to
150.62 mm as compared to 32.09 to 157.91 mm (&L/29) for the approximate glass

Figure 5. Comparison of results for Beam 1.

Downloaded from jfe.sagepub.com at Heriot - Watt University on January 6, 2015

190 Journal of Fire Protection Engineering 22(3)

Figure 6. Comparison of results for Beam 2.

transitional temperature range of 230–335 C. The predicted deflections are lower

than the safe limit specified by BS 476. Sudden deflections in beams 1 and 2 are
respectively seen at approximate temperatures of 407 C and 280 C. This can be
attributed to the fact that the polymeric material underwent a change from hard
and brittle to viscous at approximate glass transition temperatures. At this tempera-
ture, the resin is no longer able to transfer stresses from concrete to fibers. Figure 7
shows a comparison of E-modulus with experimental results from the literature [10].
A similar trend of strength loss at glass transition temperatures is observed.

Application
FRPs are high performance materials and offer a wide range of applications.
However, when used in building applications, they need to conform to fire resist-
ance ratings stated in the building codes. There has been a limited effort to develop
solutions for understanding required fire endurance of FRPs, as the current
approaches to fire resistance through standard fire tests are expensive and time
consuming. The provisions provided in the building codes for fire resistance evalu-
ation are prescriptive and are not applicable to performance-based design that
provides rational solutions under realistic scenarios. The preceding predictive mod-
eling and evaluation of beams tests has led to a cost effective, efficient and practical
performance-based method for fire safety design.
Parametric estimation is an important aspect of a practical fire-based design.
The modulus of elasticity of FRP reinforcement is lower than that of steel.
Therefore, FRP reinforced members typically display larger deflections than
equivalent steel reinforced members at ambient temperatures. This means the sup-
port conditions need to be incorporated to reflect this, in particular, relatively large

Downloaded from jfe.sagepub.com at Heriot - Watt University on January 6, 2015

Faruqi et al. 191

Figure 7. Comparison of E-modulus.

Figure 8. A simply supported beam subjected to a four-point loading.

deflections due to elevated temperatures. For practical applications and depending

on the importance of the design, glass conversion factors can be estimated to be
1–4.5% of the unit glass volume up to glass transition temperature. Also, for
practical applications and based on the latest conclusions of ACI 435R [15], a
semi fixed support system can safely be assumed. An example together with explan-
ations to calculate beam deflection is presented in the following discussion.
A rectangular, doubly FRP reinforced concrete beam using a four point loading
system is shown in Figure 8. The two vertical loads are 1000 mm apart. The sup-
ported span is 4250 mm. End supports are semi-fixed. Other parameters are:
P ¼ 440 kN; beam cross-section is 350 mm  400 mm; cover ¼ 75 mm overall;
fc0 ¼ 42 N/mm2; Afrp ¼ 1013.90 mm2; A0frp ¼ 289.70 mm2; e ¼ 125 mm; Ks ¼ 0.49;
Econc ¼ 30.65 kN/mm2; Efrp, 20 C ¼ 4.15  104 N/mm2; Efrp, 435 C & 3.60  104 N/mm2;
g ¼ 0.0465; exp & 178.1 mm; Tfrp & 435 C, which is chosen as this is the approxi-
mate glass transition temperature.

Downloaded from jfe.sagepub.com at Heriot - Watt University on January 6, 2015

192 Journal of Fire Protection Engineering 22(3)

The beam cross-section and the dimensions of the transformed area are shown
in Figure 9.
In order to calculate the beam deflection at elevated temperature (et), the
deflection constant ( ) is needed. However, this constant is a function of effective
moment of inertia (Ie) and other parameters. The Ie value depends on actual
moment (Ma), cracked moment of inertia (Icr), gross moment of inertia (Ig), and
cracking moment (Mcr) of the beam. Calculations of Ma, Icr, Ig, Mcr, Ie, , and et
are shown in the following paragraphs.
Maximum actual moment in the beam occurs at point C in Figure 8. This
moment can be calculated as: Ma ¼ P/2 * distance AC & 220  (1.62) ¼ 356 kN-m.
Tensile cracks that develop in the beam will, in effect, cause the beam cross-
section to be reduced. The cracked beam starts to lose strength as these cracks
appear. For this reason, Icr is important. The Icr value can be calculated using the
transformed area diagram shown in Figure 9. In this diagram, the FRP is trans-
formed to an equivalent (i.e. same axial stiffness) area of concrete. The location of
the neutral axis (x), shown in Figure 9 is a function of modular ratio (n) and tension
reinforcement area (Afrp). The modular ratio for the transformed area ¼ Efrp/
Econc ¼ 41.50/30.65 ¼ 1.35. The first moment of the compression area about the
neutral axis must equal the first moment of the tensile area about the neutral
axis. Therefore, taking the moment of top and bottom areas about the neutral
axis, equating, and substituting the values of ‘n’ and ‘Afrp’ provides:
350*(x)*(x/2) ¼ nAfrp*(325–x). This quadratic equation in terms of neutral axis
location is solved and yields: x ¼ 46.32 mm.

Figure 9. Beam cross-section and dimensions of transformed area.

Downloaded from jfe.sagepub.com at Heriot - Watt University on January 6, 2015

Faruqi et al. 193

Using this value of neutral axis location, the cracking moment of inertia about
the neutral axis of the transformed section can now be calculated as: 1/3 b(x)3 +
nAfrp(d–x)2, where b and d are respectively beam width, and the effective
depth of beam cross-section (Figure 9). This provides: Icr ¼ 1/
3*(350)*(46.32)3 + 1.35*1013.90*(325–46.32)2 ¼ 10.89  107 mm4.
Similarly, the gross moment of inertia can be calculated from the beam cross-
section in Figure 9 as: Ig ¼ 1/12 bh3, where h is the height of the beam cross-section
in Figure 9. This yields: Ig ¼ (1/12)*(350)*(400)3 ¼ 18.67  108 mm4.
The cracking moment is the moment that, when exceeded, causes concrete to
begin cracking. The cracking moment is found by setting the elastic flexural stress
equation equal to the tensile stress capacity of the concrete, or, the value of
Mcr ¼ (fr Ig)/yt, where fr is the modulus of rupture. This is the stress developed in
the beam on the verge of tensile failure. By definition, fr ¼ 0.7 (fc0 )1/2 and yt ¼ h/2.
This is the distance to the tensile edge from the neutral axis in the beam
cross-section. Therefore, fr ¼ 0.7(fc0 ) 1/2 ¼ 0.7(41.8)1/2 N/mm2 ¼ 4.53 N/mm2, and
Mcr ¼ 4.53(18.67  108)/200 ¼ 4.2  107 N-mm.
The quantity, Ie, accounts for both the tension stiffening and the variation of
moment of inertia along the beam. It is based on an estimation of the probable
amount of cracking caused by the varying moment throughout the beam span. The
value of Ie is always less than Ig. The equation used to calculate Ie is Ie ¼ (Mcr/
Ma)3 Ig + [1–(Mcr/Ma)3]Icr. Substitution of required parameters in this equation
yields: Ie ¼ (4.2  107/35.6 
107)3  18.67  108 + {1–(4.2  107/35.6  107)3}  10.89  107 ¼ 10.1  107 mm4.
This value is less than Ig, as expected.
Calculated and provided parameters can now be used to obtain the deflec-
tion constant and beam deflection. Therefore, ¼ [2 Ks*e*L*(Afrp + A0frp )]/
7
Ieff ¼ [2*0.49*125*4250*(1013.9 + 289.70)]/(10.1  10 ) ¼ 6.71. Substitution of
this into the deflection equation provides the deflection as: et ¼
( *g*Tfrp*Efrp)/Econc ¼ (6.71*0.0465*435*3.6  104)/30650 ¼ 159.4 mm. Therefore,
% error ¼ (178.1–159.4)/178.1 & 10.3%.

Conclusions and future work

In this article, a new model is developed that incorporates the temperature-depen-
dent progressive changes of E-modulus of FRP in predicting the deflection of FRP
reinforced concrete structures within the range of practical elevated temperatures.
Information presented has led to the following conclusions:

1. The model incorporates temperature-dependent progressive changes of E-mod-

ulus of FRP. The model is capable of predicting the deflection of FRP reinforced
concrete structures within the range of practical elevated temperatures conser-
vatively. This can provide an additional tool to evaluate the deflection of FRP
reinforced concrete structures a in fire.

Downloaded from jfe.sagepub.com at Heriot - Watt University on January 6, 2015

194 Journal of Fire Protection Engineering 22(3)

2. The predicted larger deflection values for beams 1 and 2 are approximately L/28
and L/29, respectively. These are lower than the maximum allowable limit of L/
20 specified by BS 476.

Based on the mechanical property models for FRP materials proposed here,
further work will be undertaken on flanged beams and slabs.

Funding
This research received no specific grant from any funding agency in the public, commercial,
or not-for-profit sectors.

Nomenclature
Ag ¼ area of glass
Afrp ¼ tension reinforcement
A0frp ¼ compression reinforcement
Am ¼ area of matrix
b ¼ width of beam
d ¼ effective depth of the beam cross-section
e ¼ distance from the tension reinforcement to neutral axis
Elt ¼ Young’s modulus in the leathery zone
Econc ¼ Young’s modulus of concrete
Em ¼ Young’s modulus matrix
Eg ¼ Young’s modulus of glass
Efrp ¼ Young’s modulus of FRP rebar
fr ¼ modulus of rupture
fc0 ¼ 28 days compressive strength of concrete
h ¼ overall height of the beam cross-section
Ig ¼ gross moment of inertia
Icr ¼ cracking moment of inertia
Ieff ¼ effective moment of inertia
Ks ¼ support conditions of the FRP system exposed to elevated temperatures
L ¼ beam span
Ma ¼ maximum applied moment
Mcr ¼ cracking moment
Mr ¼ reduction modulus
ng ¼ modular ratio up to leathery state
nlt ¼ modular ratio in the leathery zone
Pc ¼ force in composite material
Pfrp ¼ force in FRP rebar
Pf ¼ force in fiber
Pm ¼ force in matrix
Th ¼ tension force due to shrinkage
Vg ¼ volume of glass
Vm ¼ volume of matrix
Vlt ¼ volume in the leathery zone

Downloaded from jfe.sagepub.com at Heriot - Watt University on January 6, 2015

Faruqi et al. 195

x ¼ location of neutral axis in the transformed area diagram

yt ¼ distance from neutral axis to extreme tension fiber
g ¼ conversion of FRP at glassy state
0g ¼ FRP expansion per degree variation of temperature
d ¼ coefficient of expansion of FRP at decomposed state
Tfrp ¼ temperature at which deflection is evaluated
et ¼ beam deflection at elevated temperature
exp ¼ experimental deflection at elevated temperature
" ¼ Strain
"h ¼ concrete strain due to elevated temperature
 ¼ a conversion factor constant at particular temperature regime
f/b ¼ ratio of GFRP to balanced GFRP
 frp ¼ stress in FRP rebar
 g ¼ stress in glass
 m ¼ stress in matrix
et ¼ shrinkage curvature due to elevated temperature
¼ deflection constant

References
1. Nadjai A, Talamona D and Ali F. Fire performance of concrete beams reinforced with
FRP bars. In: Proceedings of the First International Symposium on Bond Behavior of FRP
in Structures. BBFS 2005, Hong Kong, 2005, pp.401–410.
2. ACI 440.1R-06. Guide for the design and construction of structural concrete reinforced with
FRP bars. Farmington Hills, MI: American Concrete Institute, 2006, pp.1–45.
3. CSA, S806-02. Design and construction of building components with fiber-reinforced poly-
mer. Mississauga, Canada: Canadian Standards Association, 2002.
4. ACI Committee 440. Guide for the design and construction of externally bonded FRP
systems for strengthening concrete structures. ACI 440.2R-08. Farmington Hills, MI:
American Concrete Institute, 2008, pp.1–45.
5. British Standards Institution. Fire tests on building materials and structures part 21:
Method of determination of fire resistance of elements of construction. BS 476, United
Kingdom: BSI, 1987.
6. British Standards Institution. Code of practice for fire safety in the design, management
and use of buildings. BS 9999, United Kingdom: BSI, 2008.
7. Blontrock H, Taerwe L and Matthys S. Properties of fiber reinforced plastics at
elevated temperatures with regard to fire resistance of reinforced concrete
members. In: Proceedings of 4th International Symposium on Non-Metallic (FRP)
Reinforcement for Concrete Structures, American Concrete Institute, Baltimore, MD,
1999, pp.43–45.
8. Bank LC. Properties of FRP reinforcements for concrete: fiber reinforced plastic (FRP)
reinforcements for concrete structures. In: Nanni A (ed.) Properties and applications.
London: Elsevier Science, 1993, pp.59–86.
9. Bakis CE. FRP reinforcement: material and manufacturing: fiber reinforced plastic
(FRP) reinforcements for concrete structures. In: Nanni A (ed.) Properties and
Applications. London: Elsevier Science, 1993, pp.13–58.

Downloaded from jfe.sagepub.com at Heriot - Watt University on January 6, 2015

196 Journal of Fire Protection Engineering 22(3)

10. Wang YC, Wong PH and Kodur V. Mechanical properties of fiber reinforced polymer
reinforcing bars at elevated temperatures. Canada: National Research Council Canada,
2003, NRCC-46121, pp.1–9.
11. Abbasi A and Hogg PJ. Fire testing of concrete beams with fibre reinforced plastic
rebar. Compos Part A: Appl Sci Manuf 2006; Vol. 37(8): 1142–1150.
12. Ashby MF and Jones DRH. Engineering materials: an introduction to microstructures,
processing and design. Oxford: Pergamon Press, 1999.
13. Bai Y, Keller T and Valle T. Modeling of stiffness of FRP composites under elevated
and high temperatures. Compos Sci Technol 2008; Vol. 68: 3099–3106.
14. Saafi M. Effect of fire on FRP reinforced concrete members. Compos Struct 2008;
Vol. 58: 11–20.
15. ACI 435R. Control of deflection in concrete structures. Farmington Hills, MI: American
Concrete Institute, 1997, pp.1–77.

Downloaded from jfe.sagepub.com at Heriot - Watt University on January 6, 2015