EFFECT OF REINFORCEMENT RATIO ON CONCRETE

MEMBERS REINFORCED WITH FRP BARS

Brahim Benmokrane, Michele Theriault, Radhouane Masmoudi, and Sami Rizkalla

ISIS Canada. Departement of Civil Engineering, Faculty of Applied Sciences
University of Sherbrooke, Sherbrooke, Quebec, Canada, J1K 2Rl

ABSTRACT

Non-corrosive Fiber Reinforced Plastic (FRP) rebars are being used as an alternative to steel
reinforcement to overcome the corrosion problem in bridge decks, parking garages, water and
wastewater treatment facilities. marine structures and chemical plants. This paper presents
test results of concrete beams reinforced with FRP C-BAR and conventional steel
reinforcements. The beams were tested under static loading conditions to investigate the
effects of reinforcement ratio on cracking, deflection, ultimate capacities and modes of failure.
Based on this investigation, theoretical correlations for the prediction of crack width,
maximum deflection, and ultimate load carrying capacity are proposed.

KEY WORDS: Fiber Reinforced Plastic (FRP), Composite, Concrete.

1. INTRODUCTION

Deterioration of reinforced concrete structures has become a serious problem in the last
decade. This situation is mainly due to corrosion of steel rebars embedded in concrete. In
North America, this phenomenon has been accelerated in parking garages and bridge decks
due to the use of deicing salts and significant fluctuations of temperature. The construction
industry is urgently in need of alternative materials to steel reinforcement that do not corrode.
Although various solutions has been tried in the past to counter the threat of corrosion in steel
reinforcement by using epoxy coatings, cathodic protection, increased concrete cover
thiclaless, and polymer concrete, none of the measures has provided long-tenn solution (1).
Government and private owners are realizing that better cost-effective materials are needed to
maintain and improve the infrastructure.

- 1-
9"t·d 17c:'..8-.L II~I~ - s ~p n 30
This paper presents experimental resclts of concrete beams reinforced with FRP C-BAR rods
tested under static loading conditions up to failure. The study focuses on the effects of the
reinforcement ratio on crac1cing, deflection, ultimate capacities, and modes of failure. The
study also attempts to establish a theoretical basis for the development of simple and rational
design procedures inspired by the Canadian code (2) and the ACI code (3) provisions for
concrete beams reinforced with FRP bars.

2. EXPERIMENTAL PROGRAM

2.1 Test Specimens Four series of reinforced concrete beams were tested in flexure. The
beams were 200 rom wide. 300 nun high, and 3300 mm long. Each series included two
identical beams reinforced with C-BAR rods. As shown in Figure I, beams were reinforced
with two. three, four and six rods depending on whether they pertain to series 1. 2, 3, or 4,
respectively. Also for comparison purposes, an identical series 1 beam reinforced with two
conventional 15.9 rom-diameter steel rebars and another series 4 beam reinforced with six
conventional 15.9 mm-diameter steel rebars were tested. Details of test specimenS including
reinforcement are shown in Figure 1.

The beams of all series were made from a commercial ready-mixed nOITnal weight concrete.
The measured average compressive strength and modulus of elasticity of the concrete varied
from 45 to 52 MPa and 30 to 33 GPa, respectively (Table 1). Type 1 grade B C-BAR 14.9
mm-diameter reinforcing rod was used in this study. Unlike other FRP rebars. C-BAR
reinforcing rod looks identical to epoxy coated steel rebar. Based on tests carried out by the
authors (4), a mean tensile strength of 773 ± 54 MPa and a mean modulus of elasticity of
37.6 ± 1.0 GPa were obtained for the rebar used.

2.2 Test Setup and Test Procedures Schematic of the test set-up used is shown in
Figure 2. The beam spanning 3000 rom was subjected to four-point flexural testing (pure
bending). The beams were instrumented with a linear variable differential transformer
(L VDT) at the midspan to monitor deftection, as well as strain gauges bonded on concrete
and reinforcement surfaces to measure deformations and slippage. An automatic data
acquisition system was used to monitor loading, midspan deflection and defonnations in the
concrete and in the reinforcement. At cracking and at each load increase of 5 kN, cracks were
sketched and a near midspan crack-width was measured using a microscope.

3. TEST RESULTS AND DISCUSSION

3.1 Cracking Distribution Average crack spacing in C-BAR beams was similar to
corresponding steel beams at low loading (25% Mu). However, at moderate (50% M,..) and
high (90% M,,) loading, it was approximately 65% that of steel beams, as shown in Table 2.
From this table, it can also be seen t!tat the effect of the reinforcement ratio on the crack-
spacing is negligible and that the crack-spacing slightly decreases as the load increases.
Average crack-spacing in C-BAR beams was around 139, 104 and 97 rnm, respectively. at
low. moderate, and high load.

3.2 Crack-Width Figure 3 presents the effect of the reinforcement ratio on the
experimental crack-width. It is observed that the crack-width in concrete beams reinforced
with C-BAR rods is three to five times that of identical beams reinforced with steel bars.

-2 -
.!..l·d 17'!"8-.l II~IJ - 5 ap n 30
3.2.1 Theoretical Correkztion Figure 4 shows a typical experimental and theoretical
correlation of observed crack-width using the well-known Gergely-Lutz equation (1),
adopted by the Canadian code (2) and the ACI code (3) fOl: the prediction of crack-width for
concrete members reinforced with steel:

[ 1]

Kg is a coefficient equal to 11 x 10·6mrn 2/N for conventionally reinforced concrete flexural

members. However, using the least-square method and the experimental observed crack-
6
width, this coefficient has been recalculated. A mean value of 40.9 x 10- ± 9.5 x 10-6mm 2JN
has been found for C-BAR beams.

3.3 Deflection Behaviour Figure 5 traces the experimental load-deflection history of a

typical beam. Initially, the beams were uncraked and stiff. With further loading, cracking
occurs at midspan as the applied moment exceeds the cacking moment Mer causing a reduction
in stiffness. For the C-BAR reinforced concrete beams, the load-deflection response, in post-
cracking stage, is linear until the crush of concrete. However, the residual deflection of FRP
reinforced concrete beams is three to four times that of identical conventional beams. This
ratio is mainly attributed to the difference of modulus of elasticity and other physical and
mechanical characteristics such as the bond properties and the Poisson's coefficient of the
reinforcement. Figure 6 presents the effect of the reinforcement ratio on the maximum
midspan deflection.

3.3.1 Prediction Model-l For a simply supported beam subjected to a four-point flexural
test, the maximum midspan deflection, including the shear component. is given by Eq. [21-

I!1midspan = Ma [5(3L2 _ 4z2)+ 24h2(1 +v )] [2)

rrun.. 120Ec I e 1 c

According to the Canadian code (2) and the ACI code (3), the immediate deflection of a
cracked member can be estimated using a constant effective moment of inertia Ie. given by
Branson's equation (Eq. [3]):

[3]

However, based On previous (5) and current study (6), it is found that Eq. [3] overestimates
the effective moment of inertia of a beam reinforced with FRP rebars, and therefore the
maximum deflection is underestimated. In order to take into account the effect of the
modulus of elasticity E/ of FRP tensile reinforcement, a modified expression for the effective
moment of inertia of a simply supported beam reinforced with FRP bars is proposed (Eq. [4]):

[4]

where, ~ is a reduction coefficient, which equals 0.6 for C-BAR reinforcing rods.

81 'd 17L.8-.l II~IJ - S ap n 3a

3.3.2 Prediction Model-2 For comparison purposes, the model proposed by Faza and
GangaRao (7) for the prediction of the deflection of FRP reinforced concrete beam is also
used. The model is based on the assumption that the concrete section between the load points
is fully cracked, while the end sections are partially cracked. Therefore, the new derived
deflection equation to predict the maximum midspan deflection of the beam shown in
Figure 2, by using this model, is given by Eq. [5].

[5J

Both models are plotted for a typical beam in Figure 7 and compared with the experimental
recorded midspan deflections. It can be seen that these models have excellent agreement with
experimental results.

3.4 Ultimate Load.Carrying Capacities and Modes of Failure For a beam section
reinforced in tension with FRP C-BAR rebars and in compression with conventional steel
rebars, the actual and balanced reinforcement ratios are defined by Eq. [6] and Eq. [7],
respectively.

[6]

Pbalanced =
0.85~lfc(
f fu
0.0035
0.0035+£ fu
J [7 ]

where 131 =0.85 - 0.008 (fc - 30).

Table 3 presents, for every tested bearn, the actual versus the balanced reinforcement ratios, as
well as the expected modes of failure. Note that for an under-reinforced concrete beam,
where tension reinforcement is made with FRP rebars and compression reinforcement with
steel rebars, the depth of concrete stress block Ca,) and the ultimate moment capacity (M!) are
defined by Eq. [8] and Eq.(9], respectively.

[8]
0. 85 fc b

[9 ]

However, for an over-reinforced conCrete beam, the depth of the concrete stress block (a,,,)
and the ultimate moment capacity (M/)are given by Eg. [10] and Eq.[ll], respectively.

61'd 17",8-.1 II~IJ -

'/--n
S ap 30
-------------------------------

a c2 ..
b-.., - A d
s
[12]

Table 4 presents the experimental versus the theoretical1.lltimate moment capacities, as well as
the observed modes of failure. It can be seen that, as the reinforcement ratio increases, the
ultimate moment capacity increases, but for the FRP reinforced concrete beams, this increase
is limited by the concrete compressive failure strain. Table 4 shows a good agreement
between experimental and theoretical results. Also, when comparing the actual versus
balanced reinforcement ratios (Table 3), rhe observed modes of failure can be predicted.

It should be noted that, for an over-reinforced FRP concrete beam. the failure of me beam is
of two stages (Figure 6). The flISt represents rhe crush of concrete followed immediately by
the decrease and then the increase of the moment-resistance strength. The second represents
the initiation of failure of the FRP rebars. N ate that depending on the level of the
reinforcement, a partial or a total failure of the FRP rebars can occur. When a partial failure
occurs, the beam regains the elastic deflection component, as seen in Figure 6. This figure also
shows that over-reinforced beams offer a kind of an energy dissipation despite the large
deform ability that FRP reinforced concrete sections have.

4. CONCLUSIONS

Eight beams reinforced with Type 1 Grade B C-BAR rods and two companion beams
reinforced with conventional steel bars were tested under static loading conditions. Behaviour
of the beams was monitored, to measure crack widths, crack spacing. midspan deflections,
strains and loading. The results of this investigation led to following conclusions:
• Average crack-spacing in beams reinforced with FRP C-BAR rods is similar to
corresponding beams reinforced with steel rebars at low loading (25%Mu). However at
moderate and high loading, it was approximately, on average 65% that of beams
reinforced with steel rebars.
• The effect of the 'reinforcement ratio on the crack-spacing is negligible. Average crack-
spacing for the beams reinforced with C-BAR rods is around 139, 104 and 97 rom,
respectivelY, at low, moderate, and high load.
• The maximum observed crack-width in beams reinforced with C-BAR rods is three to five
times that of identical beams reinforced with steel bars.
• Experimental strain distributions have clearly demonstrated the perfect bond between
FRP reinforcement rods and the surrounding concrete.
• Crack-width in FRP C-BAR reinforced concrete beams can be predicted using the
modified Gergely-Lutz equation (Eq. [1]). For practical prediction, the coefficient Kg
equals 41 for the C-BAR rod used in this study.
• Prediction of maximum deflection of concrete beams reinforced with C-BAR rods can be
made using either the prediction model-lor the prediction model-2.
• As the reinforcement ratio increases, the ultimate moment capacity increases. but this
increase is limited b:y the concrete compressive failure strain for the C-BAR reinforced
concrete beams.

- :J-
0c'd vL.8-.L II~IJ - 5 dP n 30
 5. NOTATIONS

A Effective tension area of concrete surrounding the flexural tension

reinforcement and having the same centroid as that reinforcement.
2
divided by the number of bars, mm
2
Section of compressive steel reinforcement, mm
2
Section of tensile FRP reinforcement, mm
Depth of the concrete stress block for an over reinforced beam, mm
Depth of the concrete stress block for an under reinforced beam, mm
Width of the beam, mm
Effective depth. mm
Depth of the compressive steel reinforcement. mm
Thickness of concrete cover measured from extreme tension fiber to the
center of the longitudinal bar located closest thereto, mm
Ec Compressive modulus of elasticity of concrete, MPa
Ef Modulus of elasticity of FRP tensile reinforcement, MPa
f'e Compressive strength of concrete, MPa
ftu Ultimate strength of FRP C-BAR reinforcing rods, MPa
is Calculated stress in reinforcement at specified loads, MPa
iy Yield strength of steel, MPa
h Height of the beam, mm
hI Distance from centroid of tension reinforcement to neutral axis. rom
h2 Distance from extreme tension fiber to neutral axis, mm
leT Cracked moment of inertia of the beam section, mm 4
Ie Effective moment of inertia of the beam section, mm 4
4
Ig Moment of inertia of the gross section of the beam, mm
Kg Correlation coefficient for Gergely-Lutz relationship
L Beam span, mm
II Distance between the applied force and the support, mm
12 Distance between the applied forces, mm
M,. Applied moment, kN.m
MeT Cracking moment. kN.m
M! Ultimate moment capacity, kN.m
p Point load, kN
W Crack-width, mm
Eju Ultimate tensile strain of FRP C-BAR reinforcing rods
Vc Poisson's coefficient of concrete
p Tensile reinforcement ratio, (Afl bd)
p' Compressive reinforcement ratio, (.4.; I bd )

6. REFERENCES

1. J.L. Clark in J.L. Clark, ed., Alternatiye Materials for the Reinforcement and
Prestressin~
of Concrete, Blackie, London, 1993, pp 1-33.
2. Canadian Standards Association (CSA), Design of CQncrete Structures for Buildings,
(CAN-A23.3-M84), Rexdale, Ontario, Canada 1984.
3. ACI Committee 318, Building Code Requirements for Reinforced Concrete and
CommentaIy (ACI318-89IACI318R-89), American Concrete Institute. Detroit. 1989.

-6-
"tc'd 17L8-1. II~I~ - S dP n 3a
4. B. Berunokrane and R. Masmoudi, Tensile Properties of Deformed FRP C-BAR
Reinforcing Rod. Technical Report No 1, Department of Civil Engineering, Universite
de Sherbrooke, Sherbrooke. Canada, 1996, 10 p.
5. R. Masmoudi, l'heQmtical ane! Experimental Eyaluation of the Flexural Behaviour of
Beams Reinforced with ERr Rebars. (in French) Doctoral Thesis, Department of Civil
Engineering, Universit6 de Sherbrooke, Sherbrooke, Canada, 1996, 273 p.
6. R. Masmoudi, M. Theriault and B. Benrnokrane, flexural Behaviour of Concrete Beams
Reinforced with Defonned FRP C-Bar Reinforcin& RQds, Technical Report No 2,
Department of Civil Engineering, Universite de Sherbrooke, Sherbrooke. Canada, 1996,
64 p.
7. S. Faza and H.V.S.Gangarao in K.W. Neale and P. Labossiere. eds., 1st International
Conference on Advanced Composite Materirus in aridges and Structures, Canadian
Society for Civil Engineering, Montreal, 1992, pp 151-160.

Table 1. Designation of beams and characteristics of longitudinal reinforcement and concrete

Series Beam Longitudinal reinforcement Concrete

No. No.
type number of reinforcement fa Ee
rebars ratio, (%) (MPa) (OPa)

CB2B-1 C-BAR 0.56 52 33

1 CB2B-2 C-BAR 2 0.56 52 33
STIB Steel 0.42 46 30
2 CB3B-l C-BAR 3 0.91 52 33
CB3B-2 C-BAR 0.91 52 33
3 CB4B-l C-BAR 4 1.38 45 30
CB4B-2 C-BAR 1.38 45 30
CB6B-1 C-BAR 2.15 45 30
4 CB6B-2 C-BAR 6 2.15 45 30
ST6B Steel 2.00 46 30

cc'd 17L.8-J. II~I~ - S dP n 30

 Table 2. Average crack-spacing (rom)

Series Beam Reinforcement Load level

No. No. ratio. %
25% M~ 50%Mu 90%M.

1 eB2B 0.56 155 99 99

2 CB3B 0.91 153 115 96

3 CB4B 1.38 145 106 99

4 CB6B 2.15 107 97 97
Average; 139 ± 22 104± 8 97±2
1 ST2B 0.42 - 248 146
4 ST6B 2.00 - 144 143

Table 3. Actual versus balanced reinforcement ratios and expected modes of failure

Beam Actual reinforcement Balanced reinforcemenl Reinforcement level

No. ratio. % ratio. % ~e~ected mode of failure)
CB2B-l 0.56 0.57 balanced (comp. or tens.)
CB2B-2 0.56 0.57 ,balanced (camp. or tens.)
STIB 0.42 4.49 under-reinforced (tension)
CB3B-1 0.91 0.57 over-reinforced (compression)
CB3B-2 0.91 0.57 over-reinforced (compression)
CB4B-l 1.38 0.53 over-reinforced (compression)
CB4B-2 1.38 0.53 over-reinforced (compression)
CB6B-l 2.15 0.53 over-reinforced (compression)
CB6B-2 2.15 0.53 over-reinforced (compression)
ST6B 2.00 4.94 under-reinforced (tension)

Table 4. Experimental versus theoretical ultimate moment capacities

Beam Mup. MI/oJor, Observed mode of failure

No. kN.m kN.m
CB2B-l 57.9 57.0 compression
CB2B-2 59.8 57,0 compression
ST2B 61.7 69.9 tension
CB3B-l 66.0 70.0 compression
CB3B-2 64.8 70.0 compression
CB4B-I ,
75.4 69.7 compression
CB4B-2 71.7 69.7 comj:>ression
CB6B-l 84.8 80.7 compression
CB6B-2 85.4 80.7 compression
ST6B 110.4 86.4 (108.2)" tension
A: value calculated usmgj.. = 500 MPa instead of jy ;; 400 MPa for conventional steel

-cf-
£C:'d 17",8-.1 II~I~ - 5 ~p n 30
 80m~ ~ mm
0
Stimlp
'pacin& =
\D 11111111 611300 11111111

a) Test specimen and reinforcement details

Steel stinup, <> = 10 nun

Steel compressive
reinforcement, <p = 10mm
h =300 mm
2 C-BAR rods,
Ib = 14.9 mm

30 rom
t I-.- .\
b :::: 200 mm

b)Typical section of beams c)TypicaI section of beams of

of series 1 serles 2 (any other detail is
identical to Fig. 3b)

j30mm r-I--II_I-I j 30 IIUD

t
1...-.4--4-+-4--'-

d)Typical scction of beams of

series 3 (any other detail is e)Typical section of beams of
identical to Fig. 3b) series 4 (any other detail is
identical to Fig. 3b)

Figure 1. Details of reinforcement and test specimens

II = 1.25 m II = 0.5 m

L=3rn

Figure 2. Schematic of the test set-up

-2 -
vc'd vL.8-.L II~I~ - 5 ~p n 3Q
 90
-e 80
BeamST6B

-~=
. .:
Q,I
70
60
§
e SO
--Ii'
"CI
Q,I

i5.. 40

S 30
e= 20
-=
~ 10

0
0 O.S 1 1.5 Z 2.S 3
Crack.wtdth, (mm)

Figure 3. Effect of reinforcement ratio on the crack width

50
-e 45 -M- Experimental

~
.J'
40 - S - GERGELY ·UJ'TZ modified

~
35
§ 30
e
'2 25
==0.
0.
20
•
~
IS
Beam No. CB2B-2
'~a 10
5
0
0 0.5 1 1.5 2 2.5 3
Crack-width, (mm)

Figure 4. Theoretical correlation of experimental crack-width

--/0 -
Sc"d 17LS-.l II~I~ - s ~p n 3a
 70

~ 60 . '" --
z.:t: ':':~---',\
-..... 50
\
\

!e 40
\
\
I

.--
"1:1
CIoI

~ 30 I
I
I

~ I
I

I 20

"=~ 10

0
0 20 40 60 80 100 120
Mid-span deflection (mm)

Figure 5. Experimental mid-span deflection versus applied moment

100
1 ; !ieam No. CB2B-2
e 90
2: Beam No. CB3B-2
~
... 80 3 : Beam No. CB4B·2
4 : B~ No. CBtSB·l
7(}
0
~ 60
Ei
"f 50
:.=
Q"
40
i'
30

J 20
10
0
0 20 40 60 80 100 120
Mid-span denection (mm)

Figure 6. Effect of the reinforcement ratio on the experimental mid-span deflection

-1/-
9c·d 17L.8-.L II~I~ - S ~p n 30