Anda di halaman 1dari 10

Color profile: Generic CMYK printer profile

Composite Default screen

125

Concrete flexural members reinforced with fiber


reinforced polymer: design for cracking and
deformability
John Newhook, Amin Ghali, and Gamil Tadros

Abstract: Fiber reinforced polymer (FRP) bars have lower modulus of elasticity than steel bars. For this reason when
FRP bars are used as flexural nonprestressed reinforcement in concrete sections, the stress in the FRP is limited to a
relatively small fraction of its tensile strength. This limit, necessary to control width of cracks at service, governs design of the required cross-sectional area of the FRP. Parametric studies on rectangular and T-sections are presented to
show that the design based on allowable strain in the FRP results in sections that exhibit large deformation before failure. The concept of deformability, given in the Canadian Highway Bridge Design Code, as a requirement in the design
of sections is discussed and modifications suggested. Using the new definition, it is shown that when, in addition to the
crack control requirement, an upper limit is imposed on the cross-sectional area of the FRP, no calculations will be
necessary to check the deformability.
Key words: fibre reinforced polymer, reinforcement, concrete, design, deformability.
Rsum : Les barres en polymres renforcs de fibres (FRP : fiber reinforced polymer ) ont un module dlasticit
plus bas que celui de barres dacier. Pour cette raison, lorsque des barres en FRP sont employes en tant que renforcement non prcontraint en flexion dans des sections en bton, il est ncessaire de limiter la contrainte dans le FRP
une fraction relativement petite de sa rsistance en tension. Le respect de cette limite, ncessaire au contrle de la largeur des fissures durant le service, dtermine quelle est laire de la section transversale du renforcement en FRP qui
est requise. Des tudes paramtriques sur des sections rectangulaires et en T sont prsentes afin de montrer que la
conception base sur le respect de la tension limite permise dans les FRP rsulte en des sections qui prsentent de larges dformations avant la rupture. Le concept de ltat de dformation, donn dans le Code canadien sur le calcul des
ponts routiers en tant quexigence dans la conception de sections, est discut et des modifications sont suggres. En
utilisant la nouvelle dfinition, il est montr que, si une limite suprieure est impose sur laire de la section transversale de FRP en plus de lexigence du contrle des fissures, alors il nest pas ncessaire de procder des calculs afin
de vrifier ltat de dformation.
Mots cls : polymres renforcs de fibres, renforcement, bton, conception, tat de dformation.
[Traduit par la Rdaction]

Newhook et al.

134

Introduction
The design of concrete sections in flexure that are reinforced with fibre reinforced polymers (FRP) is different
from that of sections reinforced with steel because of the difference in mechanical properties of FRP and steel. Generally, the FRP bars used as reinforcement in concrete have
Received 1 May 2001. Revised manuscript accepted
22 November 2001. Published on the NRC Research Press
Web site at http://cjce.nrc.ca on 10 February 2002.
J. Newhook. Department of Civil Engineering, Dalhousie
University, 1360 Barrington Street, Halifax, NS B3H 3J5,
Canada.
A. Ghali.1 Department of Civil Engineering, The University
of Calgary, 2500 University Drive N.W., Calgary,
AB T2N 1N4, Canada.
G. Tadros. ISIS Canada, 43 Schiller Crescent N.W., Calgary,
AB T3L 1W7, Canada.
Written discussion of this article is welcomed and will be
received by the Editor until 30 June 2002.
1

Corresponding author (e-mail: aghali@ucalgary.ca).

Can. J. Civ. Eng. 29: 125134 (2002)

I:\cjce\cjce_29\cjce-01\L01-085.vp
Tuesday, February 05, 2002 11:24:38 AM

tensile strength varying between 500 and 2200 MPa and


modulus of elasticity varying between 40 and 150 GPa. The
stressstrain relationship for FRP is linear up to rupture
when the ultimate strength is reached. Unlike steel reinforcing bars, FRP bars do not undergo yield deformation or
strain hardening before rupture. For this reason, design of
sections in flexure has been based upon consideration of ultimate strength, serviceability, and deformability. The purpose of design for deformability is to ensure that failure of a
section in flexure occurs only after development of sufficiently large curvature.
Because of the relatively low modulus of elasticity of FRP
bars compared with that of steel reinforcing bars, it is necessary to limit the stress in FRP bars in service to a relatively
small fraction of their strength in order to control crack
width. This paper demonstrates that this serviceability requirement controls the design, without the need in practice
for special calculations to verify the deformability as specified in Section 16 of the Canadian Highway Bridge Design
Code (CHBDC) (CSA 2000).
The importance that sufficient deformation occurs before
failure of nonprestressed members reinforced with FRP has

DOI: 10.1139/L01-085

2002 NRC Canada

Color profile: Generic CMYK printer profile


Composite Default screen

126

Can. J. Civ. Eng. Vol. 29, 2002


Table 1. Typical properties of reinforcing bars.
Reinforcement

Ef or Es
(GPa)

ffu or fy
(MPa)

Glass (GFRP)
Aramid (AFRP)
Carbon (CFRP)
Steel

40
80
150
200

550
1200
2000
400

been discussed by Jaeger et al. (1997), Kakizawa et al.


(1993), and Gangarao and Vijay (1997). The deformability
of members prestressed with FRP has been discussed by
Naaman and Jeong (1995), Abdelrahman et al. (1997), and
Grace and Sayed (1997). The present paper is concerned
only with internally reinforced, nonprestressed members.

Reinforcement properties
Three types of FRP bars used as reinforcement for concrete are considered in this paper: glass, aramid, and carbon
(GFRP, AFRP, and CFRP). A linear elastic stressstrain relationship will be assumed for the three types of FRP in tension, with rupture of the FRP occurring when ultimate
strength is reached. FRP bars are weak in compression and
thus the presence of FRP bars in the compression zone of a
concrete section in flexure is ignored. The commonly used
bilinear stressstrain relationship is assumed for steel reinforcement with a linear elastic portion up to yield followed
by an extensive perfectly plastic portion. Table 1 lists the
modulus of elasticity, Ef, and the tensile strength, ffu, for
each of the three most common types of FRP together with
the modulus of elasticity, Es, and nominal yield strength, fy,
for steel bars.
Flexural strength
Failure of a reinforced concrete section in flexure can be
caused by rupture of FRP or yielding of steel reinforcement
in tension or by crushing of the concrete. Research
(Gangaroa and Vijay 1997; Jaeger et al. 1997) has established that the ultimate flexural strengths for both failure
types can be calculated using the same equations when the
reinforcement is steel or FRP bars. This approach is adopted
in design guidelines by both the American Concrete Institute
(ACI 2001) and ISIS Canada (2001) and in the CHBDC design criteria (CSA 2000). These equations are reviewed here
and used in the discussions in this paper. The actual deformation and flexural strength of sections is investigated in
this paper; therefore, the material resistance and strength reduction factors required by codes are not included in the
equations.
Failure of tension reinforcement
Figure 1a shows the stress and strain distributions in a
section at failure by rupture of the FRP reinforcement. Such
a section is said to be under-reinforced. The theory for sections under-reinforced with steel bars is well documented in
textbooks. Initially the steel yields and the curvature increases rapidly until the strain in concrete, c, at the extreme
compressive fibre reaches an ultimate value cu = 0.0035 and

thus failure occurs. The stress in concrete is idealized by the


rectangular stress block shown in Fig. 1b. However, when a
section is under-reinforced with FRP, no yielding occurs and
the failure is caused by rupture of the FRP. The strain in the
reinforcement will be fu = ffu/Ef, where ffu and Ef are the
tensile strength and the modulus of elasticity of the FRP.
The corresponding strain c at the extreme compressive fibre
will be less than cu. Thus, the distribution of compressive
stress on the concrete cannot be idealized by the traditional
rectangular block in Fig. 1b.
One way to determine the distribution of stress on the
concrete is to use a stressstrain relationship such as the one
proposed by Todeschini et al. (1964) and adjusted by
MacGregor (1997) (Fig. 2):
[1]

fc = 1.8 fc

/ 0
1 + ( / 0) 2

where 0 = 1.71 fc /Ec; fc is the specified compressive


strength of concrete; Ec is the modulus of elasticity of concrete, which may be taken as 4750 fc MPa (Todeschini et
al. 1964); fc is the magnitude of the concrete stress corresponding to a strain at any fibre. The assumption that plane
cross section remains plane after deformation leads to the
linear strain distribution shown in Fig. 1a, from which
[2]

c
c
=
d c + fu

where c is the depth of the compression zone and d is the


distance from the extreme compressive fibre to the tension
reinforcement. The absolute value of the resultant tensile
force, T, in the reinforcement and the resultant compression
force in the concrete, C, are equal. Thus,
[3]

Af ffu = fc da

where da is an elemental area of the compressive zone and


Af is the cross-sectional area of the tension reinforcement.
Equations [1][3] can be used to determine the strain c at
the extreme compressive fibre, the stress distribution, and
hence the location of the resultant compressive force and yCT
(using numerical integration). The nominal moment resistance of the under-reinforced section and the corresponding
curvature are
[4]

M u = Af ffu y CT

[5]

u = ( c + fu )/d

where yCT is the distance between the resultant concrete


compressive force and the tension reinforcement.
Failure by crushing of concrete
When the flexural failure is induced by crushing of concrete, without rupture of FRP or yielding of steel, the section
is said to be over-reinforced. For a T-section to be overreinforced, it must have a large amount of reinforcement,
which is considered impractical.
Figure 1b shows the strain and stress distribution at ultimate for an over-reinforced rectangular section. The nonlinear distribution of concrete in the compression zone is
replaced by an equivalent uniform stress over a part of the
2002 NRC Canada

I:\cjce\cjce_29\cjce-01\L01-085.vp
Tuesday, February 05, 2002 11:24:38 AM

Color profile: Generic CMYK printer profile


Composite Default screen

Newhook et al.

127

Fig. 1. Stressstrain distributions in flexure: (a) failure by rupture of FRP; (b) failure by crushing of concrete; and (c) service condition with strain in FRP equal to fs.

compression zone such as is defined in Fig. 1b in accordance with Standard A23.3-94 (CSA 1994). The nominal
moment resistance of over-reinforced sections is given by
[6]

M u = Af ff y CT

[7]

y CT = d 1c / 2

Fig. 2. Stressstrain relationship adopted in analysis of underreinforced sections (MacGregor 1997).

The stress in the reinforcement at failure, ff, which has a


value smaller than ffu (or smaller than the yield stress when
steel is used), is given by
[8]

1/ 2

f
ff = 0.5 Ef cu 1 + 41 1 c 1

f Ef cu

where cu is the strain in concrete at the extreme compression fibre and f = Af/bd is the reinforcement ratio. Equation
[8] can be derived by equating the resultants of tensile and
compressive stresses and using eqs. [6] and [7]. According
to A23.3-94 (CSA 1994), failure by concrete crushing is
considered to have occurred when cu = 0.0035 and the values of 1 and 1 are empirically taken equal to

[9]

1 = 0.85 0.0015 fc 0.67


1 = 0.97 0.0025 fc 0.67

The curvature at ultimate is given by


2002 NRC Canada

I:\cjce\cjce_29\cjce-01\L01-085.vp
Tuesday, February 05, 2002 11:24:39 AM

Color profile: Generic CMYK printer profile


Composite Default screen

128

Can. J. Civ. Eng. Vol. 29, 2002


Table 2. Balanced and minimum reinforcement ratios for FRP reinforced rectangular sections.
Balanced reinforcement ratio (%)

[10]

u =

Reinforcement

fc = 30 MPa

fc = 60 MPa

fc = 30 MPa

fc = 60 MPa

Glass (GFRP)
Aramid (AFRP)
Carbon (CFRP)

0.80
0.34
0.23

1.38
0.59
0.39

0.42
0.19
0.11

0.59
0.27
0.16

c + ( ff / Ef )
d

with c = 0.0035. For T-sections having c > hf (where hf is


the depth to flange), eqs. [7] and [8] need to be adjusted
(McGregor 1997).
Balanced condition
The reinforcement ratio, fb, for which failure of reinforcement and crushing of concrete occur simultaneously is
referred to as the balanced reinforcement ratio. For a rectangular section, fb can be determined by the equation
[11]

c f
fb = 11 b c
d ffu

with
cb
Ef cu
=
d
Ef cu + ffu
The values of fb for various reinforcement types are given in
Table 2.

Minimum reinforcement ratio for Mu to be


greater than the cracking moment
Cracking occurs when the stress in concrete at the extreme tension fibre reaches the tensile strength of concrete in
tension, fr; the moment value, Mcr, at which cracking occurs
is
[12]

M cr = fr

I
yt

where I is the second moment of area of the uncracked


transformed section about its centroid axis. The transformed
section is composed of the areas of concrete plus (Ef/Ec)Af.
yt is the distance between the centroid of the transformed
section and the extreme tension fibre. After cracking, the
tension is assumed to be resisted entirely by the reinforcement. To avoid sudden failure by rupture or yielding of the
reinforcement, its amount must be sufficient to have a moment of resistance, Mu (calculated by eq. [4]), greater than
Mcr by an appropriate safety factor.
Thus,
[13]

Minimum reinforcement ratio (%)

Af
f
I 1
> r
(safety factor)
bd
ffu y CTy t bd

where yCT is given by eq. [14] or eq. [18].

Many design codes such as CSA A23.3-94 (1994) and ACI


318-99 (1999) transform eq. [13] to empirical expressions for
the minimum steel reinforcement, Afmin , permitted in rectangular
and in T-sections, in terms of fc divided by the specified yield
strength. For simplicity, the moment of inertia of gross concrete
section about its centroidal axis is used in lieu of I in eq. [13].
The minimum ratio of FRP, fmin, has to be greater than
the value given by eq. [13], with a safety factor relevant for
this material. The appropriate value to be used is not within
the scope of this paper; this should be given in codes on design of concrete structures reinforced with FRP.
Table 2 gives fmin for a rectangular section for the three
types of reinforcement in Table 1 and two concrete strengths.
The values in the table are equal to fmin = 5 fc /(12ffu), in
accordance with the guidelines of American Concrete Committee 440 (ACI 2001). These are approximately equal to the
values that can be calculated by eq. [13], assuming fr =
0.6 fc , d = 0.9 (height of section), and yCT 0.9d, with a
factor of safety = 3.

Crack width control


FRP bars have higher strength, ffu, than the specified yield
strength fy of steel reinforcing bars commonly used in North
America. However, because the modulus of elasticity of FRP
bars, Ef, is lower than that for steel, Es, the higher strength
cannot be effectively exploited. This is because of the need
to control the width of flexural cracks in the tension zone.
The main parameters that influence crack width include the
crack spacing, the quality of bond between the concrete and
reinforcing bars, and, above all, the strain in the reinforcement. Through the use of a crack control parameter, CSA
A23.3-M94 (1994) implicitly limits the maximum crack
width for flexural members with steel reinforcement to 0.40
and 0.33 mm respectively for exterior and interior exposure.
ACI 318-99 (1999) also limits crack widths to 0.40 mm for
all exposure conditions. In design, CSA A23.3-M94 requires
that the nominal moment multiplied by a material resistance
factor, 0.85 for steel reinforcement, be equal to or greater
than the applied service moment multiplied by an appropriate combination of dead load and live load factors, 1.25 and
1.5 respectively. Taking an average load factor of 1.38, the
service stress in the steel is approximately (0.85/1.38) times
the yield stress, or 0.6fy. CSA A23.3-M94 permits the use of
this 60% value directly in the calculation of crack control
parameters. Similarly, ACI 318-99 allows the use of 0.6fy as
the permissible reinforcement stress in calculations related
to crack control. When fy = 400 MPa, the corresponding reinforcement strain is 1200 106.
An advantage of FRP bars over steel reinforcement is that
there is no risk of corrosion. Thus, CHBDC (CSA 2000) rec 2002 NRC Canada

I:\cjce\cjce_29\cjce-01\L01-085.vp
Tuesday, February 05, 2002 11:24:40 AM

Color profile: Generic CMYK printer profile


Composite Default screen

Newhook et al.

129

Table 3. Reported crack widths


(Kobayashi et al. 1997).

Amount of reinforcement required to


control cracking

Specimen

Crack width
(mm)

f
(%)

C
C
C
C
C
C

0.50
0.50
0.60
0.55
0.60
0.70

0.40
0.28
0.23
0.23
0.19
0.16

120
160
190
190
225
260

ommends limiting the crack width to 0.71 and 0.50 mm for


interior and exterior exposure, respectively, while JSCE
(1997) recommends maximum crack width of 0.50 mm for
both interior and exterior exposure. The limits adopted by
ACI 440 (ACI 2001) are similar to those of CHBDC. From
this it is seen that the width of cracks allowed for FRP reinforced members is 1.8 or 1.5 times the value allowed for
steel-reinforced members. In the following discussion, it is
assumed that this ratio between allowable crack widths is
5/3. Furthermore, it is assumed that the width of cracks is
approximately proportional to the strain in the reinforcement. Thus, the corresponding permissible strain in the FRP
reinforcement at service is 5/3 1200 106, or 2000 106.
For a specified strain in the reinforcement, the width of
cracks can vary substantially from member to member, depending on parameters such as duration or repetition of
loading, shape and dimensions of the cross section and
cover. Even identical members may exhibit different crack
widths for the same load. In experiments (Kobayashi et al.
1997) on beams in flexure of cross section 400 mm wide
and depths 120260 mm, reinforced with carbon FRP, the
crack width listed in Table 3 have been observed when the
strain in the reinforcement reached 2000 106. The table
also includes the reinforcement ratio f.
The values listed above indicate that for this test series,
the crack width is 0.50.7 mm. This test series shows crack
widths not substantially different from 0.7 mm, which is anticipated when f = 2000 106. However, in other tests
more difference can occur between the observed and the anticipated crack width. The parametric study described below,
therefore, examines the effect of choosing either a higher or
a lower value of permissible strain associated with the acceptable crack width.
For a permissible strain of 2000 106, the corresponding
stresses at service, ffs, are 80, 160, and 300 MPa for GFRP,
AFRP, and CFRP, respectively. With these relatively low
stresses, it will be shown that the deformability requirements
will be satisfied. In other words, when the design does not
allow the stress in service in FRP bars to exceed these values, no special calculations are necessary to verify that the
requirements for deformability are satisfied.
It is noted that certain types of FRP, particularly glass FRP,
require that service strains in the FRP be limited to a fraction
of ultimate strain to avoid problems such as creep rupture,
static fatigue, and alkali deterioration (ACI 2001; Benmokrane
and Rahman 1998). The allowable limits suggested above are
more stringent than those service strain limits.

To ensure that the stress in the FRP does not exceed the
allowable limit for crack control, the cross-sectional area of
the bars should be determined by
[14]

Af = Ms/(ffs yCT)

where yCT = d c/3, with c being the depth of the compression zone (Fig. 1c). In service, the stress in concrete is commonly sufficiently low to justify the use of linear stress
strain relationship for concrete. Thus the stress in concrete
varies linearly over the depth of the compression zone
(Fig. 1c). With these assumptions, the neutral axis passes
through the centroid of area of the transformed section composed of the area of the concrete in compression and nAf,
where n = Ef/Ec. The distance, c, between the extreme compression fibre and the neutral axis is given by the solution of
the quadratic equation (Ghali et al. 2002):
[15]

c2 + a1c + a2 = 0

[16]

c=

1
a1 + a12 4a 2
2

where
a1 =

2hf
2nAf
(b bw) +
bw
bw

a2 =

hf2
2nAf d
(b bw)
bw
bw

[17]

The dimensions b, bw, d, and df are defined in Fig. 1c. When


c < hf, set bw = b in eq. [17]. In this case, the resultant compressive force is located at a distance c/3 from the extreme
compressive fibre and yCT is given by eq. [14]. The same
equation may also be used when c > hf to give an approximate value of yCT. More accurately, when c > hf, the distance
between the resultant compressive and tensile forces is given
by
[18]

y CT = d

c bc2 (b bw) (c hf ) 2 (c + 2hf )/ c

3
bc2 (b bw) (c hf ) 2

For a rectangular or T-section with a given value of the reinforcement ratio f (= Af/bd) and the strain fs, the corresponding values of the moment at service Ms and curvature
s are given by
[19]

Ms = f bdfsEf yCT

[20]

s =

fs
dc

Deformability requirements
Failure in flexure of concrete sections reinforced by steel
bars is accompanied by large curvature. The relative amount
of curvature is largest when the failure is by yielding of the
reinforcement in tension (under-reinforced sections) and reduces markedly when the failure is by crushing of the concrete (over-reinforced sections). For this reason, design
2002 NRC Canada

I:\cjce\cjce_29\cjce-01\L01-085.vp
Tuesday, February 05, 2002 11:24:41 AM

Color profile: Generic CMYK printer profile


Composite Default screen

130

Can. J. Civ. Eng. Vol. 29, 2002

Fig. 3. Deformability factors, DF, for steel-reinforced sections.

codes have generally recommended that flexural members


be designed as under-reinforced, or that the reinforcement
ratio be significantly less than b. However, CHBDC (CSA
2000) recommends, and other code committees are considering recommending, that the reinforcement ratio of FRP for
sections in flexure be greater than fb to ensure that at ultimate the failure will be caused by crushing of concrete without rupture of the FRP bars. For T-sections with wide
flange, the reinforcement areas Af > fbbd is too large to be
practical. Furthermore, it will be shown below that both Tand rectangular under-reinforced sections, with FRP bars,
will also exhibit sufficiently large curvature at failure. Thus,
it is proposed to allow reinforcement ratio f to be less than
fb provided that the product (moment curvature) in service be an acceptable fraction of the same product at ultimate. This is referred to as the deformability requirement
and is measured by the following dimensionless factors:
u
s

[21]

Curvature factor =

[22]

Moment factor =

[23]

Deformability factor, DF =

Mu
Ms
u Mu
sM s

CHBDC (CSA 2000) and Jaeger et al. (1997) check the


deformability by a performance factor defined by an equation similar in form to eq. [23], but with sMs replaced by
the product of the curvature and the moment corresponding
to a maximum concrete compressive strain of 0.001.
CHBDC requires that the performance factor be greater than
4 or 6 for rectangular and T-sections, respectively. In the
present paper the deformability factor, DF, is defined as the
ratio of uMu to sMs, with the subscripts u and s referring
to actual ultimate and service states, respectively.
For a member in flexure, the strain energy per unit length
is equal to the integral of the area below the M graph.

When the M graph is linear, the area is equal to (1/2)M


and the ratio of M at any two load levels is equal to the ratio of the strain energy values. By the use of the deformability factor as defined in eq. [23], the ratio of M at
ultimate and at service is used as a practical means to indicate approximately the ratio of the strain energy values at
the two load levels.
Using the definition of deformability provided in eq. [23],
the deformability factors for a rectangular section with steel
reinforcement are shown in Fig. 3. The strain in the steel at
service is taken as 1200 106 with fy = 400 MPa, Es =
200 GPa, and fc = 30 and 60 MPa. Each curve is extended
to the respective balanced reinforcement ratio, bal, as calculated by eq. [11] with fy replacing ffu. Due to the ductility of
steel, large curvature factors (eq. [21]) and hence high values
of DF are achieved at low reinforcement ratios. However, the
values drop rapidly as the reinforcement ratio increases
reaching a value of 4 at approximately 85% of bal.
An allowable deformability factor, DF 4, is adopted here
for all concrete sections in flexure, because as will be
shown, there is no substantial difference between T- and
rectangular sections in terms of their deformability. It will
also be shown that DF is greater than 4 for FRP-reinforced
sections except when the reinforcement ratio is impractically
high, that is, when f > fmax. An empirical equation will be
given for the maximum allowable reinforcement ratio, fmax.
It is noted, however, that even the value of fmax given here
may be impractically high.

Parametric study
The magnitude of the deformability factor, DF, is determined below for rectangular and T-sections in flexure reinforced with GFRP, AFRP, or CFRP. The allowable strain at
service in the FRP, fs = 2000 106, is adopted. The parametric study shows that DF decreases with the increase in f
and the decrease of fc . The value f is varied, starting by a
minimum value, fmin, given in Table 2. The maximum value
is taken as the smaller of 0.04 and the value that corresponds
to DF = 4. A relatively low value, fc = 30 MPa, and a relatively high value, fc = 60 MPa, are adopted. The range of f
is extended to 4%, although this reinforcement ratio may be
considered impractical; this is done to cover the range where
DF approaches the allowable limit, DF = 4.0.
Rectangular sections
Table 4 gives the ratios Mu/Ms, u/s, and DF for rectangular sections in flexure; b is the width of the section; d is
the distance from the extreme compression fibre to the centroid of tension reinforcement. GFRP, AFRP, and CFRP are
used as reinforcement for the sections analyzed in Table 4.
Results are presented for fc = 30 MPa and fc = 60 MPa for
each FRP type. The lower limits of the reinforcement ratio,
f, adopted in the tables are 0.40%, 0.19%, and 0.11% for
GFRP, AFRP, and CFRP, respectively, with fc = 30 MPa.
For the high strength concrete, with fc = 60 MPa, the lower
limits adopted for f are 0.60%, 0.28%, and 0.16%, respectively, for GFRP, AFRP, and CFRP. These lower limits approximate the minimum reinforcement ratios given in
Table 2. The values of Mu and u are calculated by eqs. [4]
2002 NRC Canada

I:\cjce\cjce_29\cjce-01\L01-085.vp
Tuesday, February 05, 2002 11:24:41 AM

Color profile: Generic CMYK printer profile


Composite Default screen

Newhook et al.

131
Table 4. Representative values of deformability factors for FRP reinforced rectangular sections.
f
(%)

Service condition
f = 2000 106
Ms/bd2

s*d

Ultimate condition
Mu/bd2

Factors

u*d

Mu/Ms

u/ s

DF

0.01564
0.01725
0.01561
0.01163
0.00885

6.8
6.6
5.8
3.8
2.4

7.0
7.3
6.5
4.5
3.1

47.6
48.3
37.6
17.1
7.6

0.01542
0.01725
0.01553
0.01466
0.01096

6.8
6.7
5.9
5.4
3.6

6.9
7.2
6.4
5.9
4.1

46.9
48.4
37.2
32.1
14.5

0.01705
0.01850
0.01561
0.01163
0.00734

7.4
7.2
5.8
3.8
1.7

7.6
8.0
6.5
4.5
2.3

56.4
57.2
37.5
17.1
4.0

0.01689
0.01850
0.01662
0.01003
0.00734

7.4
7.3
6.4
3.1
1.8

7.5
7.8
6.9
3.6
2.2

56.0
57.4
44.1
11.2
4.0

0.01556
0.01691
0.01483
0.01194
0.00733

6.8
6.4
5.4
4.0
1.7

6.9
7.2
6.1
4.7
2.3

47.3
45.7
33.0
18.5
4.0

0.01542
0.01683
0.01508
0.00957
0.00734

6.9
6.5
5.6
2.9
1.8

6.9
7.0
6.1
3.3
2.2

47.0
45.6
34.5
9.7
4.0

(a) GFRP reinforcement, f c = 30 MPa


0.40
0.80
1.00
2.00
4.00

0.3086
0.6065
0.7563
1.4808
2.8796

0.00224
0.00235
0.00239
0.00258
0.00286

2.1059
3.9873
4.3592
5.6285
7.0381

(b) GFRP reinforcement, f c = 60 MPa


0.60
1.38
1.75
2.00
4.00

0.4624
1.0440
1.3158
1.4978
2.9230

0.00225
0.00239
0.00244
0.00248
0.00271

3.1671
7.0018
7.7032
8.1176
10.4904

(c) AFRP reinforcement, f c = 30 MPa


0.19
0.34
0.50
1.00
3.35

0.2937
0.5208
0.7570
1.4822
4.7046

0.00224
0.00232
0.00240
0.00258
0.00318

2.1746
3.7430
4.3592
5.6285
8.1042

(d) AFRP reinforcement, f c = 60 MPa


0.28
0.59
0.75
2.50
5.80

0.4325
0.8967
1.1338
3.6229
8.0535

0.00224
0.00236
0.00241
0.00281
0.003334

3.2174
6.5656
7.2434
11.3213
14.5965

(e) CFRP reinforcement, f c = 30 MPa


0.11
0.23
0.30
0.50
1.80

0.3180
0.6546
0.8479
1.3909
4.7319

0.00225
0.00236
0.00242
0.00256
0.00318

2.1691
4.1820
4.5615
5.5029
8.1193

( f ) CFRP reinforcement, f c = 60 MPa


0.16
0.39
0.50
1.50
3.10

0.4623
1.1013
1.4065
4.0492
8.0598

0.00225
0.00240
0.00246
0.00286
0.00333

3.1669
7.1603
7.9154
11.7697
14.6049

and [5] when f < fb and by eqs. [6] and [10] when f fb.
The values of Ms and s are determined by eqs. [19] and
[20] when the strain in the FRP is taken as fs = 2000 106.
Figures 46 show the variation of DF with the reinforcement
ratio, f.
T-sections
The equations used in preparing Table 4 for rectangular
sections apply also to T-sections when the distance c from
the extreme compression fibre to the neutral axis is less than
or equal to the thickness of the flange, hf. When c > hf, the
distance yCT between the extreme compression fibre and the

tension reinforcement will be different from that for a rectangular section. However, this has only a small influence on
the value of DF. Figure 7 shows the variation of DF with
hf/d for a T-section with CFRP. The T-section used for this
figure has the ratio b/bw = 9. The highest reinforcement ratio
used is f = Af/(bd) = 0.44%; the cross-sectional area Af is
equal to 4.0% of bwd. Any larger reinforcement ratio is considered impractical.
Effect of varying the allowable strain, fs
The allowable strain in service in the FRP, fs = 2000
106, adopted in preparation of Figs. 46 corresponds to al 2002 NRC Canada

I:\cjce\cjce_29\cjce-01\L01-085.vp
Tuesday, February 05, 2002 11:24:43 AM

Color profile: Generic CMYK printer profile


Composite Default screen

132
Fig. 4. Deformability factors, DF, for GFRP-reinforced sections.

Can. J. Civ. Eng. Vol. 29, 2002


Fig. 7. Deformability factors, DF, for T-sections with variable
hf/d and b/bw (CFRP, fc = 30 MPa, fs = 2000 106).

Fig. 5. Deformability factors, DF, for AFRP-reinforced sections.


Fig. 8. Effect of varying fs on the deformability factors, DF
(CFRP, rectangular section fc = 30 MPa).

Fig. 6. Deformability factors, DF, for CFRP-reinforced sections.

lowable crack width of 5/3 times the width implicitly given


by CSA A23.3-M94 or ACI 318-99 for steel-reinforced
members. It may be desirable in design to tolerate wider or
to require narrower cracks by the selection of the value for
fs greater or less than 2000 106. As well, other factors,
including bond characteristics of individual bars, bar spacing, and bar size, may result in a variation in the value of fs.
It can be verified that the deformability factor is approximately proportional to (fs)2. This can be seen from the
graphs in Fig. 8, which are prepared for CFRP with fc =
30 MPa and fs = 2400 106, 2000 106, and 1600 106.
Discussion of results of parametric study
Examining the graphs in Figs. 46, it can be seen that the
deformability factor, DF, has an almost constant very high
value when f fb. The value of DF decreases with the increase of f and the decrease of fc . With the allowable value
of strain in service in the FRP, ffs = 2000 106, used to pre 2002 NRC Canada

I:\cjce\cjce_29\cjce-01\L01-085.vp
Tuesday, February 05, 2002 11:24:43 AM

Color profile: Generic CMYK printer profile


Composite Default screen

Newhook et al.

133

pare Figs. 46, it is seen that DF 4 except in the following


unusual situations:
For AFRP, when fc = 30 MPa and f > 3.4%.
For CFRP, when fc = 30 MPa and f > 1.75% or when
fc = 60 MPa and f > 3.0%.
The parametric study in which DF is calculated for variable f, fc , and ffs enabled to determine the value of f for
which DF is equal to 4.0. These calculations gave the empirical inequality:
[24]

ffs
< 0.19
1 fc

This can be used to set a maximum FRP reinforcement ratio


to be used in design:
[25]

fmax = 0.2

1 fc
ffs

As f approaches fmax, the amount of FRP reinforcement becomes difficult to accommodate. Thus, for economy and
ease of accommodating the FRP bars, f << fmax.
The graphs in Figs. 47 indicate that under-reinforced
sections and over-reinforced sections have DF > 4.0. Thus,
there is no need to set a lower limit for f to achieve
deformability. The condition that f be greater than fb is not
necessary.
Figures 46 indicate that rectangular sections with GFRP,
AFRP, or CFRP bars have DF > 4 except when f is extremely high. Comparison of the three figures indicates that
with CFRP bars the values of DF are smaller than with the
other two types of bars. Figure 7 shows that a T-section with
CFRP bars, having a flange width = 9bw also has DF > 4
even when the cross-sectional area of the FRP, Af, is as high
as 4% of bwd. As the ratio b/bw approaches 1, the area properties of the section approach those of a rectangle. Thus, the
parametric study performed for rectangular sections and for
T-sections with b/bw = 9 covers a wide range of the T- or
rectangular sections that can occur in practice, with the three
kinds of FRP bars.

Design of sections in flexure


Based on the above discussion and the parametric study,
the amount of FRP reinforcement for a section in flexure,
for which the concrete dimensions have been selected,
should be determined as follows:
(1) Due to service loads, the stress in the reinforcement
should not exceed the allowable value, ffs. The reinforcement ratio, f = Af/(bd), should be between the limits fmin f < fmax, where fmax = 0.21 fc / ffs (eq. [25]).
The minimum reinforcement should be selected to ensure that the calculated nominal moment exceeds the
moment causing flexural cracking by an acceptable margin. The maximum reinforcement ratio is introduced to
ensure that the section exhibits sufficient deformability
before failure; however, it may be necessary not to approach the maximum to avoid congestion of the reinforcement.
(2) The calculated moment resistance should be greater
than the required demand due to the factored external
loads. The strength reduction and load factors required

by codes or technical committees dealing with FRP


reinforcement should be used in calculating the two moments.
The moment resistance is given by eq. [4] or eq. [6] for
under-reinforced or over-reinforced sections, respectively.
Comments: The concrete dimensions of the section should
be selected such that the deflection not be excessive. Deflection of flexural members can be more critical for members
with FRP than with steel reinforcement. Selection of minimum thickness to control deflection is beyond the scope of
the present paper; however, guidance is provided by ACI
440 (ACI 2001), the Japan Society of Civil Engineers (JSCE
1997), and Ghali et al. (2002).
The purpose of the permissible stress ffs is to control the
width of cracks. The permissible stress in the FRP in service
may be taken as Effs with fs selected according to the tolerable crack width. A value of fs = 2000 106 is proposed
here. This value permits a strain in FRP reinforcement of 5/3
times the strain allowed for steel reinforcement by various
codes. In practice, f << fmax and the condition that f be less
than fmax is easily satisfied. If this is not so, increasing the
depth of section or reducing ffs can satisfy the condition.

Conclusions
A design procedure is presented to give the crosssectional area of FRP reinforcement required for a section
subjected to flexure. It is assumed that the concrete dimensions have been selected and deflection criteria will be satisfied at service. The procedure is based on controlling the
width of cracks in service and ensuring that adequate
deformability of the member occurs before failure. To control the width of cracks of concrete sections in flexure reinforced with FRP, it is necessary to limit the stress in the
reinforcements to a relatively small fraction of the ultimate
strength. This design requirement governs the amount of reinforcement.
Parametric studies are presented for rectangular and Tsections reinforced with glass, aramid, and carbon FRP.
These studies show that with the amount of reinforcement
necessary to control the width of cracks, the sections will
have a deformability factor exceeding a minimum value of 4
as specified by CHBDC (CSA 2000). Thus, there is no need
to check the deformability factor in design. It is also shown
that the deformability factor, as defined in this paper, is
much greater than 4 for under-reinforced sections but reduces for over-reinforced sections.
After selection of concrete dimensions of a section in
flexure, the amount of required reinforcement area, Af,
should be such that, due to service loads, the stress in the
FRP does not exceed an allowable stress, ffs, given by Effs,
where fs is the strain in the FRP at service and Ef is the
elastic modulus of the FRP. The paper proposes to limit the
strain in the FRP due to service loads to (5/3) times the
strain allowed for steel reinforcement, yielding fs = 2000
106. The designer may increase or decrease the allowable
strain in the FRP depending upon the tolerable crack width.
The ratio of the FRP reinforcement, f = Af/(bd), should
be within the limits fmin f < fmax, where fmax =
0.21 fc /ffs. The value of fmin should be selected to ensure
that the flexural strength provided by the FRP after cracking
2002 NRC Canada

I:\cjce\cjce_29\cjce-01\L01-085.vp
Tuesday, February 05, 2002 11:24:44 AM

Color profile: Generic CMYK printer profile


Composite Default screen

134

adequately exceeds the cracking moment, Mcr, in accordance


with relevant codes.

Acknowledgements
This research is part of the work of Intelligent Sensing for
Innovative Structures (ISIS Canada), Network of Centres of
Excellence, Natural Sciences and Engineering Research
Council of Canada (NSERC). Jaeger, L.G. kindly reviewed
the first draft of the paper and suggested improvements.

References
Abdelrahman, A., Rizkalla, S.H., and Tadros, G. 1997.
Deformability of flexural concrete members prestressed with
FRP. Proceedings of the 3rd International Symposium on NonMetallic (FRP) Reinforcement for Concrete Structures, Sapporo,
Japan, Vol. 2, pp. 767774.
ACI Committee 318. 1999. Building code requirements for structural concrete. ACI 318-99, American Concrete Institute,
Farmington Hills, Mich.
ACI Committee 440. 2001. Guideline for the design and construction of concrete reinforced with FRP bars. ACI 440.1R-01,
American Concrete Institute, Farmington Hills, Mich.
Benmokrane, B., and Rahman, H. (Editors). 1998. Durability of fibre reinforced polymer (FRP) composites for construction. Proceedings of the First International Conference on Durability of
Composites in Construction, CDCC98, Sherbrooke, Que.
CSA. 1994. Design of concrete structures. Standard CSA-A23.394, Canadian Standards Association, Rexdale, Ont.
CSA. 2000. Canadian highway bridge design code (CHBDC), Section 16, fibre reinforced structures. Standard CSA-S6-00, Canadian Standards Association, Rexdale, Ont.
Gangarao, V.S., and Vijay, P.V. 1997. Design of concrete members
reinforced with GFRP bars. Proceedings of the 3rd International
Symposium on Non-Metallic (FRP) Reinforcement for Concrete
Structures, Sapporo, Japan, Vol. 1, pp. 143149.
Ghali, A., Favre, R., and Elbadry, M. 2002. Concrete structures:
stresses and deformations. 3rd ed. E&FN Spon, London, U.K.
Grace, F.N., and Sayed, G.A. 1997. Ductibility of prestressed concrete bridges using internal/external CFRP strands. Proceedings
of Conference on Faults and Repairs, July, Scotland, U.K.
ISIS Canada. 2001. Reinforced concrete structures with fibre reinforced polymers (FRPs). ISIS-M04-01, The University of Manitoba, Winnipeg, Man.
Jaeger, L.G., Mufti, A.A., and Tadros, G. 1997. The concept of the
overall performance factor in rectangular-section reinforced concrete members. Proceedings of the 3rd International Symposium
on Non-Metallic (FRP) Reinforcement for Concrete Structures,
Sapporo, Japan, Vol. 2, pp. 551559.
JSCE. 1997. Recommendation for design and construction of concrete structures using continuous fibre reinforced materials.
Concrete Engineering Series, No. 23, Japan Society of Civil Engineers, Tokyo, Japan.
Kakizawa, T., Ohno, S., and Yonezawa, T. 1993. Flexural behaviour and energy absorption of carbon FRP-reinforced concrete
beams. American Concrete Institute, SP138-35, pp. 585598.
Kobayashi, K., Rahman, H., and Fujikaki, T. 1997. Crack width
prediction for NEFMAC reinforced flexural members. Proceedings of the 3rd International Symposium on Non-Metallic
(FRP) Reinforcement for Concrete Structures, Sapporo, Japan,
Vol. 2, pp. 447454.

Can. J. Civ. Eng. Vol. 29, 2002


MacGregor, J.G. 1997. Reinforced concrete, mechanics and design.
Prentice Hall, Upper Saddle River, N.J.
Naaman, A.E., and Jeong, S.M. 1995. Structural ductility of concrete beams prestressed with FRP tendons. In Non-metallic
(FRP) reinforcement for concrete structures. Edited by L.
Taerwe. E&FN Spon, London, pp. 379386.
Todeschini, C.E., Bianchini, A.C., and Kesler, C.E. 1964. Behaviour of concrete columns reinforced with high strength steels.
ACI Journal, 61(6): 701716.

List of symbols
Af cross section of FRP reinforcement
Afmin minimum cross-sectional area of FRP
b width of flange of T-section, or width of rectangular
section
bw width of web of T-section
c distance between extreme compression fibre and neutral
axis
cb distance between extreme compression fibre and neutral
axis for balanced section
d distance between extreme compression fibre and centroid of tension reinforcement
DF deformability factor defined by eq. [23]
Ec modulus of elasticity of concrete
Ef modulus of elasticity of FRP
Es modulus of elasticity of steel
fc specified compressive strength of concrete
ff calculated stress in FRP reinforcement
ffs allowable stress in FRP reinforcement in service
ffu tensile stress that ruptures FRP
fr modulus of rupture (tensile strength of concrete in flexure)
fy specified strength of steel reinforcement
hf thickness of flange of T-section
I moment of inertia of transformed section
Mcr moment causing flexural cracking at a section due to externally applied load
Ms moment producing the allowable tensile stress ffs in FRP
Mu nominal moment strength
n modular ratio of elasticity, n = Ef/Ec
yCT distance between resultant compression and centroid of
tension reinforcement
1, 1 dimensionless factors defined by eq. [9]
c calculated strain at extreme compressive fibre
cu assumed value of strain at extreme fibre when crushing
of concrete occurs, cu = 0.0035
fs allowable strain in FRP due to service loads
bal ratio of steel reinforcement that produces balanced conditions
f ratio of tension reinforcement, f = Af/bd
fb ratio of FRP tension reinforcement that produces balanced conditions
fmax maximum ratio of reinforcement for FRP reinforced
sections
fmin minimum ratio of reinforcement for FRP reinforced sections
s curvature when strain in FRP is fs
u curvature when moment is Mu

2002 NRC Canada

I:\cjce\cjce_29\cjce-01\L01-085.vp
Tuesday, February 05, 2002 11:24:45 AM

Anda mungkin juga menyukai