EFFECTS OF GFRP REINFORCING REBARS ON SHRINKAGE AND

THERMAL STRESSES IN CONCRETE

Roger H. L. Chen
1
and Jeong-Hoon Choi
2
ABSTRACT
The use of Glass Fiber Reinforced Polymer (GFRP) rebars instead of conventional steel rebars as the
reinforcement in Continuously Reinforced Concrete Pavement (CRCP) gives solutions to the problems
caused by corrosion of reinforcement. However, it is necessary to know what effect this replacement has
on the development of concrete cracks, which is inevitable in CRCP. Concrete shrinkage and temperature
variations are known to be the principal factors for early-age crack formation in CRCP. By employing an
analytical model, this study presents the shrinkage and thermal stress distributions in concrete due to the
restraint provided by GFRP rebars in comparison with that provided by steel rebars. It reveals the
advantages of using GFRP rebars as reinforcement in CRCP in terms of internal tensile stress reduction in
concrete. Numerical calculation of the concrete stress distribution in a GFRP reinforced CRCP section
subjected to thermal change is also presented.
Keywords: GFRP rebars, CRCP, cracks, concrete shrinkage, thermal stress
INTRODUCTION
Other than the advantage of eliminating steel corrosion, thermal and stiffness compatibility
between GFRP and concrete may also offer possible advantages when GFRP rebars are used in
reinforced concrete pavements. Traditional CRCP reinforced by steel rebars has been applied for
a few decades (AASHTO, 1986), and the CRCP behavior has been reported (Won et al., 1991;
Kim et al., 2001). However, up to this point, there is no precedent for the use of GFRP as
reinforcement in CRCP, and little related work has been done. It is therefore necessary to study
the mechanical behavior of CRCP reinforced with GFRP rebars. The effects of the GFRP
reinforcing rebars on shrinkage and thermal stresses in CRCP were investigated at the onset of
this study, and will be discussed in this paper. The results of this study will eventually contribute
to the development of the design of CRCP with GFRP reinforcement.
Shrinkage and thermal stresses in concrete have been known to be principal factors for the
incipient cracking in concrete pavements or bridge decks. Understanding the development of
these stresses is essential to properly controlling cracking, which may ultimately determine the
performance and longevity of the concrete structure. In the case of a freely supported concrete
slab subjected to shrinkage or temperature variation, the concrete stresses are produced because

1
Prof., Dept. of Civ. and Envir. Engrg., West Virginia Univ., Morgantown, WV 26506. E-mail: [email protected]
2
Grad. Res. Asst., Dept. of Civ. and Envir. Engrg., West Virginia Univ., Morgantown, WV. E-mail: [email protected]
2
of the restraints provided by the reinforcements. While concrete shrinkage causes tensile stresses
in concrete (Zhang et al., 2000), the temperature variation can cause either tensile or compressive
stresses in concrete. The thermal stresses are depending on whether the temperatures drop or rise
and the relationship between the coefficients of thermal expansion (CTE) of the concrete and the
reinforcement used. The CTE of concrete varies with different coarse aggregate types, and the
CTE of the GFRP depends on the composite materials used. In this paper, analytical results are
presented to describe the effect of GFRP reinforcing rebars on shrinkage and thermal stresses in
concrete slabs.
ANALYTICAL MODEL
To approximate the developments of shrinkage and thermal stresses in a concrete slab, a
representative concrete prismatic model containing a longitudinal reinforcing rebar at its center
with width (or reinforcing space in CRCP) B, height (or thickness in CRCP) H, length L, and
rebar diameter 2r
r
is considered. Then, as a matter of analytical convenience, the model is
modified into an equivalent cylindrical one with the corresponding equivalent diameter 2R, where
/ HB R , accompanied by the same length and rebar diameter as those for the prismatic
model. The schematic details of the models are shown in Fig. 1.
Adopting the shear-lag theory (Cox, 1952), there are several assumptions made for this
analysis: 1) the concrete and reinforcement exhibit elastic behavior, 2) the bond between
concrete and reinforcement is perfect at an infinitely thin interface, 3) the stiffness of the concrete
and the reinforcement in the radial (r-) direction are the same, 4) the strain in the concrete,
c
at a
distance R from the x-axis is equivalent to the restraint-free concrete strain due to the shrinkage or
temperature variation, and 5) the temperature distribution in the concrete and reinforcement are
uniform in the radial direction. The effect of concrete radial shrinkage on the concrete stress
development in the longitudinal direction is neglected. Also, effects from the CTE discrepancies
between the concrete and the reinforcement in the radial direction are neglected.
When the concrete is subjected to a strain,
c
in the longitudinal (x-) direction, the rate of
transfer of load from concrete to reinforcement can be assumed as dP/dx = C
o
(u-v), where P is
the load of the reinforcement, and C
o
is a constant. v and u are the axial displacements at r = R
and r = r
r
, respectively. It is also known from force equilibrium that dP/dx = 2 r. Integrating
the shear strain along the radial direction one gets (u-v) = dP/dx ln(R/r
r
)/(2 G
c
), where G
c
is the
shear modulus of concrete. Hence,

) / ln(
2
r
c
o
r R
G
C

(1)
Also, dv/dx =
c
and du/dx =
r
. Therefore,
( )
c r o
C
dx
P d

2
2
(2)
The restraint-free concrete axial strain,
c
, in the above equation can be substituted with either
shrinkage strain at any time t (in days)
c,s
(t) or thermal strain
c,t
which are given by

ult s c s c
t
t
t ) (
35
) (
, ,

+

and
c t c
T
,
(3a)
where (
c,s
)
ult
= ultimate shrinkage strain for drying at 40% RH; T = temperature variation; and

c
= CTE of concrete.
c,s
(t) is an empirical equation (Mindess and Young, 1981) for moist cured
concrete. The reinforcement axial strain due to concrete shrinkage,
r,s
or due to temperature
3
variation,
r,t
are shown as:

r r
s r
E A
P

,
and
r r
r t r
E A
P
T +
,
(3b)
where A
r
= reinforcement cross-sectional area, E
r
= Youngs modulus of reinforcement, and
r
=
CTE of reinforcement.
Substituting Eq. (3) into Eq. (2) gives the governing differential equation for P, and then, by
solving the differential equation with boundary conditions that P = 0 at x = 0 and x = L, the
reinforcement force, P(x), can be obtained. The axial force equilibrium with the average axial
concrete stress, (
c
)
avg
must also be satisfied at any x location: 0 ) ( +
c avg c
A P , where A
c
is the
concrete cross-sectional area. Hence, the average axial concrete stress can be given as follows:

1
1
1
1
]
1

,
_

2
cosh
2
cosh
1 ) (
35
) (
,
L
x
L
t
t
E
ult s c r avg c

(Concrete Shrinkage) (4a)

and
( )
1
1
1
1
]
1

,
_

2
cosh
2
cosh
1 ) (
L
x
L
T E
r c r avg c

(Temperature Variation) (4b)

where = reinforcing ratio (A
r
/A
c
), and

) / ln(
/ 2
2
r r
r c
r R r
E G
(5)
The negative sign in front of Eq.(4b) indicates compressive axial stresses in concrete. The
maximum axial stress in concrete can be simply found at x = L/2.
MATERIAL PARAMETERS
In the shrinkage stress analysis, Youngs modulus of concrete, E
c
, as well as concrete
shrinkage strain,
c,s
(t), are employed as a time-dependent properties, and therefore, the elapsed
time, t (days) is the only variable. The time-dependent Youngs modulus of concrete can be
evaluated by (Mosley and Bungey, 1990)

[ ]
E t E t
c c
( ) . . ln( )
,
+
28
052 015 for t 28 (6a)
E t E
c c
( ) .
,
1019
28
for t > 28 (6b)
where E
c,28
= Youngs modulus of concrete at 28 days. In the thermal stress analysis, the concrete
stresses at 28 days are estimated for different temperature variations; a value of E
c,28
is employed
here. Table 1 lists a set of model parameters and material properties used in this study.
RESULTS AND DISCUSSION
In Fig. 1, the maximum average tensile stresses in the concrete due to the concrete shrinkage
are estimated over a period of time. The stresses with either steel or GFRP reinforcements are
shown in this figure, and they are compared with each other. In the comparison, # 5 rebars with a
radius of 0.3125 in. ( = 0.00519) are employed for the model length of L = 60 in. From the
4
figure, it can be seen that the maximum concrete stress level created by GFRP rebar is about one-
fifth of that by steel rebar. This ratio is about the same as that of the longitudinal GFRP rebars
elastic modulus to steel rebars elastic modulus.
TABLE 1. Model Parameters and Material Properties Used in Analysis
Parameter and Properties Value
Width, B, (in.) 6
Height, H, (in.) 10
Length, L, (in. ) 60
Ultimate Concrete Shrinkage Strain, (
c,s
)
ult
, () 800
Elastic Modulus of Concrete at 28 days, E
c,28
, (Msi) 4.8
(1)
, and 5
(2)
Poissons Ratio of Concrete,
c
0.2
Youngs Modulus of Steel Rebar (Msi) 29
Longitudinal Youngs Modulus of GFRP Rebar (Msi) 5.8
CTE of Concrete,
c
, (/
o
F) 5.7
(1)
and 8.0
(2)
CTE of Steel Rebar,
r,s
, (/
o
F) 6.6
CTE of GFRP Rebar,
r,g
, (/
o
F) 5.2
(note: (1) granite aggregate and (2) siliceous river gravel coarse aggregate used)
Fig. 1. Max. Avg. Tensile Stress in Concrete vs. Time ( = 0.00519 and L = 60 in.)
Different sizes of GFRP rebars, #3 through #6 ( = 0.00184 through 0.00739), are also studied in
Fig. 2, showing that the maximum concrete stress increases with an increase in the reinforcing
ratio. GFRP rebars with a lower Youngs modulus provide the concrete with less restraint than
0
0.02
0.04
0.06
0.08
0.1
0.12
0 50 100 150 200 250 300 350 400 450
Time (days )
M
a
x
.

A
v
g
.

T
e
n
s
i
l
e

S
t
r
e
s
s

(
k
s
i
)
Steel GFRP
L
r
x
r r R
B
L
H
x
y
z
Representative Prism
5
steel rebars do while the concrete shrinks. The maximum concrete stresses for different values of
L are examined in Fig. 3. The stresses increase with increases in length, L. For the set of
parameters (Table 1) used, the maximum concrete stress increasingly converges to E
r

c,s
(t) when
L approaches , and E
r

c,s
(t) is attained as L reaches about 12 in.
Fig. 2. Max. Avg. Tensile Stress in Concrete vs. Time
(L = 60 in. and different GFRP reinforcing ratios, )
Fig. 3. Max. Avg. Tensile Stress in Concrete vs. Time
( = 0.00519 GFRP and different slab lengths, L)
0
0.005
0.01
0.015
0.02
0.025
0 50 100 150 200 250 300 350 400 450
Time (days )
M
a
x
.

A
v
g
.

T
e
n
s
i
l
e

S
t
r
e
s
s

(
k
s
i
)
L = 3 in.
L = 12 in.
E
r

c,s
(t)
L = 6 in.
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0 50 100 150 200 250 300 350 400 450
Time (days )
M
a
x
.

A
v
g
.

T
e
n
s
i
l
e

S
t
r
e
s
s

(
k
s
i
)
= 0.00184 (#3)
= 0.00334 (#4)
= 0.00519 (#5)
= 0.00739 (#6)
6
The discrepancy in CTE between concrete and reinforcement causes the thermal stresses in
both concrete and reinforcement. Figs. 4 and 5 show the plots of the maximum average axial
stress in concrete versus temperature change for two different concrete CTEs.
Fig. 4. Max. Avg. Axial Stress in Concrete vs. Temperature Change
( = 0.00519 and L = 60 in.)
Fig. 5. Max. Avg. Axial Stress in Concrete vs. Temperature Change
( = 0.00519 and L = 60 in.)
-0.002
-0.001
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0 10 20 30 40 50 60
T (
o
F )
M
a
x
.

A
v
g
.

A
x
i
a
l

S
t
r
e
s
s

(
k
s
i
)
Steel GFRP

c
= 5.7 /
o
F

r,s
= 6.6 /
o
F

r,g
= 5.2 /
o
F
-0.012
-0.01
-0.008
-0.006
-0.004
-0.002
0
0 10 20 30 40 50 60
T (
o
F )
M
a
x
.

A
v
g
.

A
x
i
a
l

S
t
r
e
s
s

(
k
s
i
)
Steel GFRP steel(FEM) GFRP(FEM)

c
= 8.0 /
o
F

r,s
= 6.6 /
o
F

r,g
= 5.2 /
o
F
7
Fig. 6. Axial Stress in Concrete vs. Longitudinal Location by FEM
( = 0.00519 and L = 60 in.)
The CTE of granite aggregate concrete is lower than that of steel rebar and higher than that of
GFRP rebar (Table 1). As shown in Fig. 4, this CTE of concrete causes tensile concrete stress
when using steel rebar and compressive concrete stress when using GFRP rebar as temperature
increases; in the figure, the positive value represents the tensile stress, and the negative value the
compressive stress. The siliceous river gravel aggregate concrete has a CTE higher than the
CTEs of both steel and GFRP rebars, and temperature increases lead to compressive stress
development in the concrete for both cases, as shown in Fig. 5. These compressive and tensile
states are reversed as temperature decreases. As can be seen in both Figs. 4 and 5, the absolute
values of thermal concrete stresses from GFRP rebar are less than those from steel rebar, mainly
because of GFRP rebars lower elastic modulus. In addition, both the magnitude and direction of
the stress caused by temperature variation can be controlled by using different combinations of
concrete and reinforcement CTEs.
A finite element model (FEM) has also been created to compare the results with those from
the thermal analytical model by using the FE analysis program, ABAQUS. The concrete slab is
modeled as four-node, 2-D elements under plane stress conditions, and the reinforcement is
modeled as beam elements with circular cross-sections. The thickness of the concrete 2-D plane
stress elements is the width, B, of the concrete prismatic model. All the material properties are
assumed to be the same as those used in the thermal analytical model (Table 1). The comparison
(Fig. 5) shows that the FE results have a good agreement with the analytical results.
As can be seen in the figures, the tensile stress levels in concrete caused by both concrete
shrinkage and temperature variation are below the tensile strength of normal concrete. However,
other restraining forces act on a given concrete slab, such as friction from the subbase under
CRCPs, restraints from the girders underneath bridge decks, or restraints from reinforcement ties
to neighboring slabs in CRCPs. When these restraints are considered, the overall resulting tensile
stress level in the concrete will increase, most likely causing cracks in the concrete slab. Fig. 6
shows the FEM results of thermal tensile stresses at the top of a reinforced concrete pavement
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 5 10 15 20 25 30 35 40 45 50 55 60
x (in. )
A
x
i
a
l

S
t
r
e
s
s

(
k
s
i
)
steel GFRP
T at top of slab = -50
o
F
T at bottom of slab = -30
o
F
(Temperature changes linearly from top to bottom.)

c
= 8.0 /
o
F

r,s
= 6.6 /
o
F

r,g
= 5.2 /
o
F
8
along the longitudinal (x-) direction, considering the aforementioned restraining forces. The
friction from the subbase is modeled by using horizontal spring elements with a spring stiffness
of 1,350 lb/in., and the reinforcement at its ends is constrained in longitudinal and rotational
directions. For both cases, using the steel and using the GFRP reinforcement, the maximum
tensile stress caused by temperature variation increases after considering the restraints. The
maximum tensile stresses in concrete reinforced with steel rebars exceed the tensile strength of
normal concrete, while that for the GFRP reinforced concrete still stays below the concrete tensile
strength. The concrete pavement reinforced with steel rebars will crack at its middle in this case.
It is noted that in the above numerical examples, bond between concrete and reinforcement is
assumed to be perfect. When bond slip is considered, a smaller concrete tensile stress can be
expected. Furthermore, because of the low modulus of elasticity in the GFRP rebars, larger crack
spacings followed by wider crack widths in the GFRP reinforced concrete pavement can be
expected. If it is necessary to shorten the crack spacing or narrow the crack width, such as in a
CRCP case, a larger amount of GFRP reinforcement is needed.
SUMMARY
In this paper, the concrete stress produced in a GFRP reinforced concrete slab due to concrete
shrinkage or temperature variation is calculated. The analytical solution indicates that the lower
Youngs modulus of GFRP rebars results in the stress reduction in concrete. The thermal stress in
concrete can be either tensile or compressive, depending on temperature variation and the CTEs
of the concrete and reinforcement used. Lower tensile stresses developed in GFRP reinforced
concrete can also cause the crack spacing to be larger and crack width to be wider in a
continuously reinforced concrete pavement, when compared with the CRCP reinforced with the
same amount of traditional steel rebars.
ACKNOWLEDGMENTS
The authors acknowledge the support from USDOT/FHWA (DTFH61-99-X-00078).
REFERENCES
AASHTO (1986), Guide for Design of Pavement Structures, American Association of State
Highway and Transportation Officials.
Cox, M. A. (1952), The Elasticity and Strength of Paper and Other Fibrous Materials, British
Journal of Applied Physics, 3, 72-79.
Kim, S. M., M. C. Won, and B. F. McCullough (2001), CRCP-9 Computer Program Users
Guide, Research Report 1831-3, Center for Transportation Research, The University of Texas
at Austin.
Mindess, S. and J. F. Young (1981), Concrete, Prentice Hall.
Mosley, W. H. and J. H. Bungey (1990), Reinforced Concrete Design, Macmillan Education Ltd..
Won, M., K. Hankins, and B. F. McCullough (1991), Mechanistic Analysis of Continuously
Reinforced Concrete Pavements Considering Material Characteristics, Variability, and
Fatigue, Research Report 1169-2, Center for Transportation Research, The University of
Texas at Austin.
Zhang, J., V. C. Li, and C. Wu (2000), Influence of Reinforcing Bars on Shrinkage Stresses in
Concrete Slabs, ASCE Journal of Engineering Mechanics, 126(12), 1297-1300.