See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/245886536

Design of Concrete Flexural Members Strengthened in Shear with FRP

Article  in  Journal of Composites for Construction · November 2000

DOI: 10.1061/(ASCE)1090-0268(2000)4:4(198)

CITATIONS READS
248 382

2 authors, including:

Thanasis Triantafillou
University of Patras
143 PUBLICATIONS   7,014 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

TRM as Strengthening Materials of Existing Structures View project

Innovative seismic isolation of infill walls View project

All content following this page was uploaded by Thanasis Triantafillou on 19 February 2018.

The user has requested enhancement of the downloaded file.

 DESIGN OF CONCRETE FLEXURAL MEMBERS STRENGTHENED IN
SHEAR WITH FRP
By Thanasis C. Triantafillou,1 Member, ASCE, and Costas P. Antonopoulos2

ABSTRACT: The present study describes a simple design model for the calculation of the fiber-reinforced
polymer (FRP) contribution to the shear capacity of strengthened RC elements according to the design formats
of the Eurocode, American Concrete Institute, and Japan Concrete Institute. The key element in the model is
the calculation of an effective FRP strain, which is calculated when the element reaches its shear capacity due
to concrete diagonal tension. Diagonal tension failure may be combined with FRP debonding or tensile fracture,
and the latter also may occur at a stage beyond the ultimate shear capacity. An upper limit (maximum) to the
FRP effective strain also is defined and aimed at controlling crack opening. The effective strain, obtained through
calibration with >75 experimental data, is shown to decrease with the FRP axial rigidity divided by the concrete
shear strength. It also is demonstrated that the contribution of FRP to shear capacity is typically controlled by
either the maximum effective strain or by debonding and, for a given concrete strength, it increases linearly
with the FRP axial rigidity until the latter reaches a limiting value beyond which debonding controls and the
gain in shear capacity is relatively small. However, proper anchoring (e.g., full wrapping) suppresses the de-
bonding mechanism and results in considerable increases in shear capacity with the FRP axial rigidity. Finally
it is demonstrated that, when compared with others, the proposed model gives better agreement with most of
the test results available.

INTRODUCTION AND BACKGROUND tive arguments, Triantafillou (1998) derived a polynomial

The use of fiber-reinforced polymers (FRPs) as a shear function that relates the strain in the FRP at shear failure of
strengthening material for RC elements is becoming increas- the member, defined as effective strain ε f,e to the axial rigidity
ingly popular (see Fig. 1 for a practical application and Fig. of externally bonded strips or sheets. The axial rigidity was
2 for a schematic typical of FRP configurations) due to the expressed by the product Ef ␳f , where Ef = elastic modulus of
outstanding combination of properties (lightweight, easy han- FRP in the principal fiber orientation and ␳f = FRP reinforce-
dling, corrosion resistance, excellent mechanical strength, etc.) ment ratio equal to 2tf /bw for continuously bonded shear re-
offered by FRPs. However, detailed investigations on shear inforcement of thickness tf (bw = minimum width of the con-
strengthening of RC members have been relatively limited crete cross section over the effective depth) or (2tf /bw)(bf /sf)
and, to a certain degree, controversial. With a couple of excep- for FRP reinforcement in the form of strips of width bf at a
tions, most researchers have idealized FRP materials in an anal- spacing sf . This polynomial was derived through curve fitting
ogy with internal steel stirrups, assuming that the contribution on about 40 test data published by various researchers and
of FRP to shear capacity emanates from the capacity of fibers showed a clear tendency of ε f ,e to reduce with Ef ␳f . The mod-
to carry tensile stresses at a more or less constant strain, which eling approach of Triantafillou had the following three short-
is equal either to the FRP ultimate tensile strain ε f,u or to a comings: (1) FRP fracture was assumed to occur simultane-
reduced value (e.g., 0.005 or a fixed fraction of ε f,u). ously with shear failure (concrete diagonal tension), whereas
A review of the relevant literature up to early 1997 may be in reality it may occasionally appear after the peak load (shear
found in Triantafillou (1998), the first systematic attempt to capacity) is reached; (2) one equation was used to describe
characterize the contribution of externally bonded FRPs to the both FRP fracture and debonding, regardless of the type of
shear capacity of RC members based on both analysis and all FRP material [Carbon FRP (CFRP) versus aramid FRP
experimental results available. It was reported in that study (AFRP) or Glass FRP (GFRP)]; and (3) the concrete strength,
that the effectiveness of the external FRP shear reinforcement which is expected to affect debonding, was not introduced as
a design variable. All shortcomings were rather obvious and
and its contribution to the shear capacity of RC members de-
known, but the set of experimental results available at that
pends on the mode of failure, which may occur either by peel-
time did not justify a more rigorous treatment.
ing off through the concrete near the concrete-FRP interface,
In a recent study, Khalifa et al. (1998) used a slight modi-
or by FRP tensile fracture at a stress that may be lower than
fication of Triantafillou’s equation (1998) (calibrated with a
the FRP tensile strength (e.g., because of stress concentrations
few more test results) to describe shear failure combined si-
at debonded areas or at rounded corners). Whether peeling off
multaneously with FRP fracture and the bond model of Maeda
or fracture will occur first depends on the bond conditions,
et al. (1997) to describe shear failure combined with FRP de-
available anchorage length, type of attachment at the FRP cur-
bonding. These two models were then presented in American
tailment, FRP thickness, FRP elastic modulus, concrete
Concrete Institute (ACI) shear design format. This approach
strength, and other factors. It also was stated that, in many is more correct, as it treats the two failure mechanisms sepa-
cases, the actual mechanism is a combination of peeling off at rately, but it is based on the peculiar assumption that FRP
certain areas and fracture at others. Based on mostly qualita- fracture is described by an equation that has been derived from
1
Asst. Prof., Dept. of Civ. Engrg., Univ. of Patras, Patras 26500, test data corresponding to both FRP fracture and debonding.
Greece. E-mail: ttriant@upatras.gr Yet, Item (1) listed above was not treated properly.
2
Grad. Res. Asst., Dept. of Civ. Engrg., Univ. of Patras, Patras 26500, Finally, it is at least worth mentioning the ‘‘Japanese’’ ap-
Greece. proach to shear strengthening with FRP. A recent state-of-the-
Note. Discussion open until April 1, 2001. To extend the closing date art report of the Japan Concrete Institute (JCI) (1998) stated
one month, a written request must be filed with the ASCE Manager of that the contribution of FRP to the shear capacity of columns
Journals. The manuscript for this paper was submitted for review and
possible publication on January 29, 1999. This paper is part of the Jour-
(where FRP is wrapped all around) may be calculated using
nal of Composites for Construction, Vol. 4, No. 4, November, 2000. the truss model and an effective strain equal to 2/3 or 1/3 times
䉷ASCE, ISSN 1090-0268/00/0004-0198–0205/$8.00 ⫹ $.50 per page. the ultimate tensile strain for CFRP or AFRP, respectively.
Paper No. 20200. Other Japanese reports, described in JCI (1998), have sug-
198 / JOURNAL OF COMPOSITES FOR CONSTRUCTION / NOVEMBER 2000
 available to date and is presented in three design formats: Eu-
rocode, ACI, and JCI. Comparisons of the model with others
also are provided.

DESIGN APPROACH—EUROCODE FORMAT

According to the model of Triantafillou (1998) in the Eu-
rocode design format [Comité Euro-International du Béton
(CEB) 1992], the external FRP reinforcement may be treated
in analogy to the internal steel if one accepts that FRP carries
only normal stresses in the principal FRP material direction
and assumes that, at the ultimate limit state in shear (concrete
diagonal tension), FRP develops an effective strain ε f,e in the
principal material direction that is, in general, less than the
FIG. 1. Shear Strengthening Application in Elounda Beach tensile failure strain ε f,u. Hence, the FRP contribution to shear
Hotel, Greece, with CFRP Sheets
capacity Vfd can be written in the following form:
ε f k,e
Vfd = 0.9 Ef ␳f bw d(1 ⫹ cot ␤)sin ␤ (1)
␥f
where ε f k,e = characteristic value of the effective FRP strain;
␥f = partial safety factor for FRP; bw = minimum width of
cross section over the effective depth; d = effective depth of
cross section; ␳f = FRP reinforcement ratio; Ef = elastic mod-
ulus of FRP in the principal fiber orientation; and ␤ = angle
between the principal fiber orientation and longitudinal axis of
the member.
The characteristic value of the effective FRP strain may be
approximated by multiplying the mean value of the effective
FRP strain ε f,e by a reduction factor, say ␣. Yet, ε f k,e should be
limited to ε max to ensure that the shear integrity of concrete is
maintained sufficiently so that other mechanisms, such as the
aggregate interlock, may be activated too.
The effective FRP strain ε f,e may be established based on a
combination of qualitative arguments and calibration of (1)
with experimental results. Central to the estimation of ε f,e is
the identification of the status of the FRP material at shear
failure of the RC element, which is always defined by concrete
diagonal tension. As Fig. 3 illustrates in a qualitative manner,
this may occur either prematurely, as a result of FRP debond-
ing, or after FRP has been stretched considerably. In the latter
case FRP may fracture either exactly at the peak load or a
little after due to overstressing in the vicinity of the diagonal
cracks.
It may be argued that at the ultimate limit state a certain
degree of FRP debonding at the concrete-FRP interface is al-
ways expected, even if the ultimate failure does not occur si-
multaneously with debonding. This is attributed to excessive
straining in FRP that results in strain incompatibilities with the
substrate material (concrete) and subsequent cracking. Crack-
ing, in turn, causes stress concentrations that produce local
debonding [e.g., Rostásy and Neubauer (1997)]. Hence, one
FIG. 2. Schematic Illustration of Shear Strengthening with: (a)
FRP Sheets Bonded to Web; (b) Wrapped or U-Shaped Sheets or may argue that ε f,e depends heavily on the FRP bonded length,
Strips

gested values between 2/3 and 1.0 for CFRP, and 0.2 and 0.6
for AFRP. For beams, no design equation is provided, but it
is recommended that some sort of mechanical anchoring of the
side FRP jackets should always be provided to avoid prema-
ture debonding failures. One of the conclusions made in the
JCI report (1998) for both column- and beam-type members
is that ‘‘the contribution of continuous fiber reinforcing ma-
terials to the shear strength has not been correctly estimated
yet.’’
The present study, which draws heavily on the model of
Triantafillou (1998), aims at describing an improved, yet sim-
ple, modeling approach for the contribution of externally
bonded FRP shear reinforcement to the shear capacity of RC
flexural members. The model is calibrated with 75 test data FIG. 3. Schematic Illustration of Shear Failures

JOURNAL OF COMPOSITES FOR CONSTRUCTION / NOVEMBER 2000 / 199

its relation to the ‘‘effective bond length’’ through which FRP- that is meant to be as simple as possible for the sake of design
concrete interface shear bond stresses develop, and the relation calculations may not be desirable.
of the latter to the ‘‘development length,’’ which is defined as
that length necessary to reach FRP tensile fracture before de- RESULTS ON SHEAR STRENGTHENING OF
bonding. Apart from the bond conditions (surface preparation, CONCRETE WITH FRP
execution, etc.), the development length depends on the FRP
axial rigidity, which may be expressed by the product Ef ␳f , Here one proceeds to a synthesis and evaluation of the pub-
and is inversely proportional to the shear (i.e., tensile) strength lished experimental results on shear strengthening of RC mem-
of concrete. It is well established today that the tensile strength bers with FRP strips or sheets. Of the test results reported in
of concrete is proportional to f 2/3
c , where fc is the compressive
the literature, only those that correspond to shear failures of
strength, so one may finally argue that ε f,e depends on the RC members strengthened with FRP in shear are reported. The
quantity Ef ␳f /f c2/3. This dependence may be established from published data are summarized in Tables 1 [mainly taken from
experimental results, as described in the next section. One Triantafillou (1998)] and 2. The mean effective FRP strain ε f,e
drawback of the rather qualitative approach described above in these tables was calculated based on the following equation:
is that the FRP bonded length is not taken explicitly into ac-
Vexp = 0.9ε f,e Ef ␳f bw d(1 ⫹ cot ␤)sin ␤ (2)
count. But this may be at least partially justified in view of
the following arguments: (1) For real size structures, the ef- where Vexp = experimentally derived contribution of FRP to
fective bond length is typically a small fraction of an RC el- shear capacity.
ement’s depth; hence, the partially ineffective FRP comprises
a rather small fraction of the total; (2) the effect of short FRP MODEL CALIBRATION
bonded lengths may be taken into account to a certain extent
through the experimental data fitting and calibration procedure The experimental results provided in Tables 1 and 2 were
presented below, and (3) additional complexity to an approach used to produce the plots, in Fig. 4 for premature shear failure

TABLE 1. Experimental Data on Shear Strengthening of RC with FRP

Specimen a
bw(m) d(m) fc(MPa) FRPb ε f,u(⫻10⫺3) ␳f(⫻10⫺3) Ef(GPa) ␤(degrees) FRP at failurec ε f,e(⫻10⫺3)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
B(3) 0.114 0.085 42.9 G, s 20.5 11.0 16.8 45 Debonded 6.6
B(4) 0.114 0.085 42.9 G, s 20.5 27.0 16.8 45 Debonded 5.6
U(3) 0.1 0.17 24.6 C, w 11.5 1.94 230 90 Fracture 5.0
U(5) 0.1 0.17 24.6 C, s 11.5 1.94 230 90 Debonded 3.0
U(6) 0.1 0.17 27.4 C, s 11.5 1.94 230 56 Debonded 3.4
U(7) 0.1 0.17 27.4 C, s 11.5 3.9 230 90 Debonded 1.5
AS(WO) 0.15 0.113 37.7 G, s 12.5 40.0 16 90 Debonded 0.8
AS(SO) 0.15 0.113 37.7 G, s 12.5 16.0 16 90 Debonded 1.8
OO(BS12) 0.18 0.36 27.8 C, w 12.0 1.2 230 90 Fracture 8.4
OO(BS24) 0.18 0.36 27.8 C, w 12.0 2.4 230 90 Fracture 6.2
OO(BM06) 0.18 0.36 27.8 C, w 12.0 0.6 230 90 Fracture 11.7
OO(BM12) 0.18 0.36 27.8 C, w 12.0 1.2 230 90 Fracture 9.3
OO(BM18) 0.18 0.36 27.8 C, w 12.0 1.8 230 90 Fracture 7.8
OO(BM24) 0.18 0.36 27.8 C, w 12.0 2.4 230 90 Fracture 6.0
OO(BL06) 0.18 0.36 27.8 C, w 12.0 0.6 230 90 Fracture 8.4
OO(BL12) 0.18 0.36 27.8 C, w 12.0 1.2 230 90 Fracture 7.8
OO(BMW06) 0.18 0.36 21.0 C, w 12.0 0.6 230 90 Fracture 8.4
OO(BMW012) 0.18 0.36 21.0 C, w 12.0 1.2 230 90 Fracture 6.9
OO(BMW24) 0.18 0.36 21.0 C, w 12.0 2.4 230 90 Fracture 4.6
OO(2) 0.4 0.34 27.8 C, w 12.0 0.29 230 90 Fracture 12.0
OO(3) 0.4 0.34 27.8 C, w 12.0 0.58 230 90 Fracture 10.3
CJ(4) 0.063 0.152 46.0 A, s, b 22.5 33.0 11.0 90 No fracture 4.9
CJ(E) 0.063 0.152 46.0 G, s, b 13.4 21.0 14.3 90 Fracture 6.3
CJ(G) 0.063 0.152 46.0 C, s, b 9.5 18.0 21.0 90 Fracture 5.2
CJ(45G) 0.063 0.152 46.0 C, s, b 9.5 18.0 21.0 45 Fracture 5.1
SU(S2) 0.2 0.26 45.2 C, s 15.1 0.6 230 90 Debonded 10.0
SU(S3) 0.2 0.26 41.3 C, s, b 15.1 0.6 230 90 Debonded 16.0
SU(S4) 0.2 0.26 37.5 C, s 15.1 1.2 230 90 Debonded 5.0
SU(S5) 0.2 0.26 39.7 C, s, b 15.1 1.2 230 90 Debonded 8.0
A(a) 0.07 0.1 30.0 C, s 13.0 2.2 235 90/0 Debonded 4.1
A(b) 0.07 0.1 30.0 C, s 13.0 2.2 235 90/0 Debonded 3.4
A(c) 0.07 0.1 30.0 C, s 13.0 2.2 235 90/0 Debonded 3.3
A(45) 0.07 0.1 30.0 C, s 13.0 2.2 235 45/135 Debonded 3.0
MI(AN-1/5 Z-3) 0.125 0.165 35.1 C, w 15.1 0.35 230 90 Fracture 12.5
MI(AN-1/2 Z-3) 0.125 0.165 32.4 C, w 15.1 0.88 230 90 Fracture 7.85
MI(CN-1/2 Z-2) 0.125 0.165 39.1 C, w 15.1 0.88 230 90 Fracture 9.2
TK(BS2) 0.2 0.42 35.1 C, s, b 12.5 0.17 280 90 Debonded 10.9
TK(BS5) 0.2 0.42 36.8 C, s, b 12.5 0.13 280 90 Debonded 11.9
TK(BS6) 0.2 0.42 35.8 C, s, b 12.5 0.088 280 90 Debonded 16.1
TK(BS7) 0.2 0.42 34.7 C, w 12.5 0.13 280 90 No fracture 34.9
a
B = Berset (1992), U = Uji (1992), AS = Al-Sulaimani et al. (1994), OO = Ohuchi et al. (1994), CJ = Chajes et al. (1995), SU = Sato et al. (1996),
A = Antonopoulos (1996), MI = Miyauchi et al. (1997), TK = Taerwe et al. (1997). Symbols for each beam appear in ( ), as assigned by those who
performed tests.
b
G = GFRP, C = CFRP, A = AFRP, s = bonded to sides, b = bonded to bottom, w = wrapped around.
c
Fracture = fractured at shear failure, no fracture = fractured after shear failure.

200 / JOURNAL OF COMPOSITES FOR CONSTRUCTION / NOVEMBER 2000

 TABLE 2. Additional Experimental Data on Shear Strengthening of RC with FRP

bw d fc ε f,u ␳f Ef ␤ ε f,e
Specimena (m) (m) (MPa) FRPb (⫻10⫺3) (⫻10⫺3) (GPa) (degrees) FRP at failurec (⫻10⫺3)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
FS(S-2) 0.6 0.51 30.0 C, w 16.0 0.56 240 90 Fracture 6.5
FS(S-3) 0.6 0.51 30.0 C, w 16.0 1.1 240 90 No fracture 4.7
FS(S-4) 0.6 0.51 30.0 C, w 16.0 1.67 240 90 Fracture 4.4
UF(CS1) 0.3 0.257 40.5 C, w 17.5 0.74 244 90 No fracture 6.9
UF(CS2) 0.3 0.257 40.5 C, w 17.5 0.37 244 90 Fracture 5.1
UF(CS3) 0.15 0.272 44.8 C, w 17.5 0.47 244 90 No fracture 12.0
UF(AS1) 0.15 0.272 43.0 A, w 37.0 0.587 73 90 No fracture 17.5
UF(AS2) 0.15 0.272 43.0 A, w 37.0 0.293 73 90 Fracture 33.0
UF(AS3) 0.15 0.272 44.8 A, w 37.0 1.17 73 90 No fracture 15.9
UF(AB1) 0.15 0.253 41.9 A, w 37.0 0.587 73 90 No fracture 43.0
UF(AB2) 0.3 0.253 45.6 A, w 37.0 0.293 73 90 No fracture 31.0
UF(AB3) 0.3 0.253 41.9 A, w 37.0 0.587 73 90 No fracture 28.0
UF(AB4) 0.3 0.253 41.9 A, w 37.0 0.587 73 90 No fracture 33.0
UF(AB5) 0.3 0.253 42.7 A, w 37.0 0.96 73 90 No fracture 26.0
UF(AB8) 0.6 0.253 43.5 A, w 37.0 0.48 73 90 No fracture 29.0
UF(AB9) 0.45 0.4 39.9 A, w 37.0 0.64 73 90 No fracture 21.0
UF(AB10) 0.55 0.5 39.9 A, w 37.0 0.53 73 90 Fracture 30.0
UF(AB11) 0.55 0.5 40.6 A, w 37.0 1.05 73 90 No fracture 20.0
AM(CF045) 0.2 0.34 24.8 C, w 15.1 0.26 230 90 No fracture 9.0
AM(CF064) 0.2 0.34 24.9 C, w 15.1 0.45 230 90 No fracture 9.6
AM(CF097) 0.2 0.34 25.2 C, w 15.1 0.77 230 90 No fracture 9.8
AM(CF131) 0.2 0.34 25.4 C, w 15.1 1.1 230 90 No fracture 10.3
AM(CF243) 0.2 0.34 25.6 C, w 15.1 2.2 230 90 No fracture 6.6
AM(AF060) 0.2 0.34 25.8 A, w 28.1 0.59 87 90 No fracture 12.0
AM(AF090) 0.2 0.34 25.9 A, w 28.1 1.0 87 90 No fracture 11.0
AM(AF120) 0.2 0.34 26.1 A, w 28.1 1.4 87 90 No fracture 14.0
SUK(2) 0.15 0.24 35.7 C, s, b 15.1 1.46 230 90 Debonded 3.78
SUK(3)d 0.15 0.24 35.3 C, s, b 15.1 1.46 230 90 Debonded 5.97
OM(SB2), Ax 0.3 0.26 24.3 C, w 13.8 0.37 248 90 NC 3.4
OM(SB3), Ax 0.3 0.26 24.3 C, w 13.8 0.74 248 90 NC 8.76
OM(SC2), Ax 0.6 0.54 25.2 C, w 12.1 0.37 230 90 NC 5.0
OM(SC3), Ax 0.6 0.54 25.2 C, w 12.1 0.73 230 90 NC 5.4
OM(SC4), Ax 0.6 0.54 25.2 C, w 12.1 1.85 230 56 NC 2.1
T(S4) 0.18 0.46 48.5 C, s 10.0 8.8 70.8 45 Fracture 3.23
T(SR1) 0.18 0.46 53.8 C, s 10.0 4.4 70.8 45 Debonded 5.42
T(SR2) 0.18 0.46 52.7 C, s 10.0 8.8 70.8 45 Debonded 3.73
Note: Ax = axial stress = 0.062fc. NC = not clear if FRP fracture occurred simultaneously with shear failure or a little later.
a
FS = Funakawa et al. (1997), UF = Umezu et al. (1997), AM = Araki et al. (1997), SUK = Sato et al. (1997), OM = Ono et al. (1997), T = Taljsten
(1997).
b
G = GFRP, C = CFRP, A = AFRP, s = bonded to sides, b = bonded to bottom, w = wrapped around.
c
Fracture = fractured at shear failure, no fracture = fractured after shear failure.
d
Mechanical anchors on top part of side FRP.

FIG. 4. Effective FRP Strain in Terms of Ef ␳f /f 2/3

c —Shear Fail-
FIG. 5. Normalized FRP Strain in Terms of Ef ␳f /f 2/3
c —Shear

ure Combined with FRP Debonding Failure Combined with or Followed by FRP Fracture

due to FRP debonding and Fig. 5 for shear failure combined ever, the type of fibers should not affect ε f,e considerably in
with or followed by FRP fracture, which both give ε f,e in terms the case of debonding (see discussion above on parameters
of the quantity Ef ␳f /f c2/3. In agreement with the qualitative ar- expected to control debonding).
guments made above, it can be seen that ε f,e clearly decreases If debonding is not dominant, the effective strain appears to
as Ef ␳f /f c2/3 increases. The results in Fig. 4 are useful in char- depend on the type of FRP (different trends are shown in Fig.
acterizing CFRP-strengthened elements, as there are no data 5 for CFRP and AFRP strengthening materials) due to the very
points for AFRP and points for GFRP are very limited. How- different fracture strains of these materials (Tables 1 and 2).
JOURNAL OF COMPOSITES FOR CONSTRUCTION / NOVEMBER 2000 / 201
 Finally, the experimental results suggest that debonding is
not likely to dominate in the case of fully wrapped CFRP or
AFRP (for instance, in the case of columns or rectangular
beams). The best fit power-type expressions for the test data
(dashed lines in Figs. 4 and 5) are summarized below. Clearly,
the scatter of these data is considerable (especially in Fig. 5),
and the following expressions could be improved in the future
when more test results become available:
Premature shear failure due to debonding (for CFRP only, to
be used with caution for other FRPs):

冉 冊
0.56
f 2/3
ε f,e = 0.65 ⫻ 10⫺3
c
(3)
Ef ␳f
Shear failure combined with or followed by CFRP fracture:

冉 冊
0.30
ε f,e f 2/3
c
= 0.17 (4a)
ε f,u Ef ␳f
Shear failure combined with or followed by AFRP fracture: FIG. 6. FRP Contribution to Shear Capacity in Terms of Ef ␳f

冉 冊
for Two Concrete Strengths and Wrapped Versus Unwrapped El-
0.47
ε f,e f 2/3
c
ements
= 0.048 (4b)
ε f,u Ef ␳ f
strengthening material with ε f,u = 0.015 and ␤ = 90⬚ (fibers
Note that in all equations and figures above, fc is in MPa and perpendicular to the long axis of the member).
Ef is in GPa. A careful examination of Figs. 4–6 leads to the conclusion
that, for values of Ef ␳f below a limiting value (Ef ␳f)lim, the
SUMMARY OF DESIGN APPROACH design is governed by the limiting FRP strain ε max. That is, no
FRP failure mechanism is activated; hence, the contribution of
The design equations for shear strengthening of RC flexural FRP to shear capacity is proportional to Ef ␳f . For values of
members with externally bonded FRP materials are (1) and Ef ␳f exceeding (Ef ␳f)lim, failure is governed by (1) debonding
ε f k,e = ␣ε f,e ⱕ ε max (5) combined with shear failure, if FRP is not properly anchored
(e.g., side or U-shaped jackets); or (2) shear fracture combined
where ␣ = 0.8 and ε max = 0.005 (as an example, see solid lines with or followed by CFRP fracture, if the composite material
in Figs. 4 and 5 corresponding to ε f,u = 0.015 for CFRP and is anchored properly (e.g., fully wrapped). In the first case the
ε f,u = 0.035 for AFRP). The proposed values ␥f = ␥f,f = partial increase in shear capacity with Ef ␳f is relatively small, but the
safety factor if failure is combined with or followed by FRP concrete strength plays an important role, whereas in the sec-
fracture = 1.20 for CFRP, 1.25 for AFRP, and 1.30 for GFRP; ond case the increase in shear capacity with Ef ␳f becomes quite
␥f = ␥f,d = partial safety factor if FRP debonding dominates = substantial, but the role of concrete is of secondary impor-
1.30; and ␥f = 1.30 if ε f k,e = ε max. Note that, when the proposed tance. It may therefore be reasonable to suggest that Ef ␳f
value for ε max is divided by the material safety factor, it yields should not exceed (Ef ␳f)lim unless debonding may be prevented
a value approximately equal to 0.004, which has been pro- or the attachment of FRP to concrete may be improved by
posed by other researchers as an appropriate upper limit to specially designed mechanical anchorages. The limiting value
maintain the integrity of concrete and secure activation of the (Ef ␳f)lim for debonding to be suppressed is given by the fol-
aggregate interlock mechanism [e.g., Priestley and Seible lowing expression:

冉 冊
(1995) and Khalifa et al. (1998)]. Finally it should be empha- 1/0.56
sized that the other values given above are subject to possible 0.65 ⫻ 10⫺3␣
(Ef ␳f)lim = f 2/3 = 0.018f 2/3 (7)
ε max
c c
future revisions as more statistical analysis data become avail-
able for the tensile strength of FRP materials. Finally the ex-
As an additional recommendation toward the proper design
pressions for ε f,e are as follows:
of RC elements strengthened in shear with FRP, the spacing
Fully wrapped CFRP: sf of strips, if they are used vertically, should not exceed 0.8d,

冉 冊
0.30 so that no diagonal crack may be formed without intercepting
f 2/3 a strip. Finally, full wrapping appears to be far more effective
ε f,e = 0.17 ε f,u
c
(6a)
Ef ␳f than partial jacketing (U-shaped or side jackets). When full
wrapping is not feasible (for instance, when there is no access
Side or U-shaped CFRP jackets: to the top side of T-beams), it is recommended that FRP strips

冋 冉 冊 冉 冊 册
0.56 0.30
f 2/3 f 2/3 be attached to the compressive zone of the RC member
ε f,e = min ⫻ 10⫺3, 0.17 ε f,u
c c
0.65 (6b) through the use of simple mechanical anchors.
Ef ␳f Ef ␳ f
Fully wrapped AFRP: ACI AND JCI CODE FORMATS

冉 冊
0.47
f 2/3 In the ACI code format, the factored FRP contribution to
ε f,e = 0.048 ε f,u
c
(6c) shear capacity may be expressed as
Ef ␳f
␸fVf = ␸f ε f,e,A Ef ␳f (sin ␤ ⫹ cos ␤)bd (8)
DISCUSSION AND RECOMMENDATIONS
with ε f,e,A = 0.9ε f,e ⱕ ε max,A; and ␸f = strength reduction factor
To obtain a better insight into the contribution of FRP to (the subscript A is used to characterize the ACI code format).
shear capacity, the above equations are plotted in Fig. 6 for Note that 0.9ε max /␣ = 0.9 ⫻ 0.005/0.8 ⬇ 0.006, so ε max,A may
two concrete strengths, fc = 20 and 40 MPa, assuming a CFRP be taken (tentatively) equal to 0.006. Moreover, ␸f = ␸f,d if
202 / JOURNAL OF COMPOSITES FOR CONSTRUCTION / NOVEMBER 2000
 TABLE 3. Comparison of Predictions for Various Modeling Approaches
FRP at
failureb ε f,e ⫺ exp ε f,e ⫺ T ε f,e ⫺ K ε f,e ⫺ J ⌬ε ⫺ T ⌬ε ⫺ K ⌬ε ⫺ J
Specimena (m) (⫻10⫺3) (⫻10⫺3) (⫻10⫺3) (⫻10⫺3) (%) (%) (%)
(1) (2) (3) (4) (5) (6) (7) (8) (9)
B(3) Debonded 6.6 6.81 NA(⫺) NA(o) 3.15 — —
B(4) Debonded 5.6 4.12 NA(⫺) NA(o) ⫺26.5 — —
U(3) Fracture 5.0 4.72 3.98 7.67 ⫺5.5 ⫺20.4 53.4
U(5) Debonded 3.0 3.38 0.61 NA(o) 12.5 ⫺79.7 —
U(6) Debonded 3.4 3.51 0.65 NA(o) 3.4 ⫺80.9 —
U(7) Debonded 1.5 2.38 1.57 (fracture) NA(o) 58.5 4.7 —
AS(WO) Debonded 0.8 3.23 NA(m) NA(o) 304.5 — —
AS(SO) Debonded 1.8 5.40 NA(m) NA(o) 200.3 — —
OO(BS12) Fracture 8.4 5.84 5.81 8.00 ⫺30.5 ⫺30.8 ⫺4.8
OO(BS24) Fracture 6.2 4.74 3.32 8.00 ⫺23.5 ⫺46.5 29.0
OO(BM06) Fracture 11.7 7.18 7.44 8.00 ⫺38.6 ⫺36.4 ⫺31.6
OO(BM12) Fracture 9.3 5.84 5.81 8.00 ⫺37.2 ⫺37.5 ⫺14.0
OO(BM18) Fracture 7.8 5.17 4.44 8.00 ⫺33.7 ⫺43.1 2.6
OO(BM24) Fracture 6.0 4.74 3.32 8.00 ⫺21.0 ⫺44.7 33.3
OO(BL06) Fracture 8.4 7.18 7.44 8.00 ⫺14.4 ⫺11.4 ⫺4.8
OO(BL12) Fracture 7.8 5.84 5.81 8.00 ⫺25.2 ⫺25.5 2.6
OO(BMW06) Fracture 8.4 6.79 7.44 8.00 ⫺19.1 ⫺11.4 4.8
OO(BMW012) Fracture 6.9 5.51 5.81 8.00 ⫺20.0 ⫺15.8 15.9
OO(BMW24) Fracture 4.6 4.48 3.32 8.00 ⫺2.6 ⫺27.8 73.9
OO(2) Fracture 12.0 8.94 8.39 8.00 ⫺25.5 ⫺30.1 ⫺33.3
OO(3) Fracture 10.3 7.26 7.50 8.00 ⫺29.5 ⫺27.2 ⫺22.3
CJ(4) No fracture 4.9 4.79 (debonded) NA(m) NA(o) ⫺2.3 — —
CJ(E) Fracture 6.3 5.32 (debonded) NA(m) NA(o) ⫺15.5 — —
CJ(G) Fracture 5.2 4.65 NA(⫺) NA(o) ⫺10.6 — —
CJ(45G) Fracture 5.1 4.65 NA(⫺) NA(o) ⫺8.8 — —
SU(S2) Debonded 10.0 8.18 3.75 NA(o) ⫺18.2 ⫺62.5 —
SU(S3) Debonded 16.0 7.90 5.43 NA(o) ⫺50.6 ⫺66.1 —
SU(S4) Debonded 5.0 5.17 3.31 NA(o) 3.4 ⫺33.8 —
SU(S5) Debonded 8.0 5.28 5.28 NA(o) ⫺34.0 ⫺34.0 —
A(a) Debonded 4.1 3.34 NA(⫺) NA(o) ⫺18.4 — —
A(b) Debonded 3.4 3.34 NA(⫺) NA(o) ⫺1.6 — —
A(c) Debonded 3.3 3.34 NA(⫺) NA(o) 1.4 — —
A(45) Debonded 3.0 3.34 NA(⫺) NA(o) 11.5 — —
MI(AN-1/5 Z-3) Fracture 12.5 11.14 10.32 10.07 ⫺10.9 ⫺17.4 ⫺19.4
MI(AN-1/2 Z-3) Fracture 7.85 8.31 8.37 10.07 5.9 6.6 28.3
MI(CN-1/2 Z-2) Fracture 9.2 8.63 8.37 10.07 ⫺6.2 ⫺9.0 9.5
TK(BS2) Debonded 10.9 10.70(*) 5.24 NA(o) ⫺1.8 ⫺51.9 —
TK(BS5) Debonded 11.9 11.75(*) 5.41 NA(o) ⫺1.2 ⫺54.5 —
TK(BS6) Debonded 16.1 13.14(*) 5.31 NA(o) ⫺18.3 ⫺67.0 —
TK(BS7) No fracture 34.9 11.61 9.17 8.33 ⫺66.7 ⫺73.7 ⫺76.1
FS(S-2) Fracture 6.5 9.80 9.99 10.67 50.8 53.7 64.2
FS(S-3) Fracture 4.7 8.01 7.93 10.67 70.4 68.7 127.0
FS(S-4) Fracture 4.4 7.06 6.08 10.67 60.6 38.2 142.5
UF(CS1) No fracture 6.9 10.42 10.08 11.67 51.0 46.1 69.1
UF(CS2) Fracture 5.1 12.83 11.77 11.67 151.6 130.8 128.8
UF(CS3) No fracture 12.0 12.12 11.30 11.67 1.5 ⫺5.8 ⫺2.7
UF(AS1) No fracture 17.5 25.36 NA(m) 12.30 44.9 — ⫺29.7
UF(AS2) Fracture 33.0 35.16 NA(m) 12.30 6.5 — ⫺62.7
UF(AS3) No fracture 15.9 18.58 NA(m) 12.30 16.8 — ⫺22.6
UF(AB1) No fracture 43.0 25.16 NA(m) 12.30 ⫺41.5 — ⫺71.4
UF(AB2) No fracture 31.0 35.81 NA(m) 12.30 15.5 — ⫺60.3
UF(AB3) No fracture 28.0 25.16 NA(m) 12.30 ⫺10.1 — ⫺56.1
UF(AB4) No fracture 33.0 25.16 NA(m) 12.30 ⫺23.7 — ⫺62.7
UF(AB5) No fracture 26.0 20.08 NA(m) 12.30 ⫺22.7 — ⫺52.7
UF(AB8) No fracture 29.0 27.98 NA(m) 12.30 ⫺3.5 — ⫺57.6
UF(AB9) No fracture 21.0 23.79 NA(m) 12.30 13.3 — ⫺41.4
UF(AB10) Fracture 30.0 26.00 NA(m) 12.30 ⫺13.3 — ⫺59.0
UF(AB11) No fracture 20.0 18.96 NA(m) 12.30 ⫺5.2 — ⫺38.5
AM(CF045) No fracture 9.0 11.36 10.68 10.07 26.2 18.7 11.9
AM(CF064) No fracture 9.6 9.64 9.93 10.07 0.43 3.4 4.9
AM(CF097) No fracture 9.8 8.23 8.75 10.07 ⫺16.0 ⫺10.7 2.8
AM(CF131) No fracture 10.3 7.40 7.63 10.07 ⫺28.1 ⫺25.9 ⫺2.2
AM(CF243) No fracture 6.6 6.02 4.61 10.07 ⫺8.7 ⫺30.2 52.6
AM(AF060) No fracture 12.0 15.08 NA(m) 9.37 25.7 — ⫺21.9
AM(AF090) No fracture 11.0 11.78 NA(m) 9.37 7.1 — ⫺14.8
AM(AF120) No fracture 14.0 10.08 NA(m) 9.37 ⫺28.0 — ⫺33.1
SUK(2) Debonded 3.78 4.55 3.45 NA(o) 20.4 ⫺8.7 —
SUK(3)c Debonded 5.97 4.53 3.43 NA(o) ⫺24.1 ⫺42.5 —
OM(SB2) NC 3.4 9.09 9.26 9.20 167.4 172.4 170.6
OM(SB3) NC 8.76 7.38 7.91 9.20 ⫺15.7 ⫺9.7 5.0
OM(SC2) NC 5.0 8.21 8.21 8.07 64.3 64.2 61.4
OM(SC3) NC 5.4 6.70 7.13 8.07 24.0 32.0 49.4
OM(SC4) NC 2.1 5.07 4.37 8.07 141.3 108.1 284.3

JOURNAL OF COMPOSITES FOR CONSTRUCTION / NOVEMBER 2000 / 203

 TABLE 3. (Continued )
(1) (2) (3) (4) (5) (6) (7) (8) (9)
T(S4) Fracture 3.23 3.61 (debonded) 2.37 NA(o) 11.7 ⫺26.6 —
T(SR1) Debonded 5.42 5.35(*) 4.53 (fracture) NA(o) ⫺1.2 ⫺16.4 —
T(SR2) Debonded 3.73 3.72 2.37 (fracture) NA(o) ⫺0.2 ⫺36.5 —
Note: NA(⫺) = model is not applicable because it predicts negative ε f,e. NA(m) = model is not applicable to other than CFRP materials. NA(o) =
model is not applicable to open FRP jacket. Failure mode is given in ( ) only if model’s prediction is different from observed mode. (*) = other failure
mode than debonding was predicted. NC = not clear if FRP fracture occurred simultaneously with shear failure or a little later.
a
B = Berset (1992), U = Uji (1992), AS = Al-Sulaimani et al. (1994), OO = Ohuchi et al. (1994), CJ = Chajes et al. (1995), SU = Sato et al. (1996),
A = Antonopoulos (1996), MI = Miyauchi et al. (1997), TK = Taerwe et al. (1997), FS = Funakawa et al. (1997), UF = Umezu et al. (1997), AM =
Araki et al. (1997), SUK = Sato et al. (1997), OM = Ono et al. (1997), T = Taljsten (1997).
b
Fracture = fractured at shear failure, no fracture = fractured after shear failure.
c
Mechanical anchors on top part of side FRP.

FRP debonding dominates and ␸f = ␸f,f if shear failure is com- RC strengthened elements. The proposed model, which im-
bined with or followed by FRP fracture. Tentative values for proves earlier work performed by Triantafillou (1998), gives
␸f may be taken as follows: ␸f,d = 0.75, ␸f,f = 0.80, and ␸f = the FRP contribution in an analogy to conventional shear re-
0.75 if ε f,e,A = ε max,A. inforcement, according to the design format of the Eurocode,
The JCI code format is identical to the Eurocode format ACI, and JCI. A key element in the model is the calculation
except that 0.9 in (1) is replaced by 1/1.15. Hence the for- of an effective FRP strain, which is taken as the minimum of
mulation of the Eurocode applies here too, with the slight three values: maximum strain to control crack opening, strain
modification that ε f,e given by (6) is multiplied by 1.035. corresponding to premature shear failure due to FRP debond-
ing, and strain corresponding to shear failure combined with
COMPARISON WITH OTHER MODELS or followed by FRP tensile fracture. The last two strains, ob-
tained through calibration with >75 experimental data avail-
The effective FRP strain predicted by the model proposed able to date, were shown to be functions of the quantity
in this study for all test results listed in Tables 1 and 2 is Ef ␳f /f c2/3 (decreasing functions as this quantity increases). Fur-
compared with the strain calculated from test data and strains thermore, it was shown that the contribution of FRP to shear
predicted by other modeling approaches in Table 3. In this capacity is typically controlled by the first two of the above
table, Column 3 gives the strain calculated from test results, strains and, for a given concrete strength, it increases linearly
Column 4 gives the present model’s predictions (this model, with Ef ␳f until this product reaches a limiting value (Ef ␳f)lim
hence the letter T), Column 5 gives the predictions of the beyond which debonding controls and the gain in shear ca-
model of Khalifa et al. (1998) (hence the letter K), and Col- pacity is small unless the FRP is fully wrapped or properly
umn 6 gives the predicted strain according to the JCI proposal anchored through the use of mechanical anchoring devices.
(FRP strain at shear failure = 2/3 or 1/3 of the ultimate strain Finally it was demonstrated that, when compared with two
for CFRP and AFRP, respectively). Note that the scope of the other models, the proposed model gives better agreement with
comparison here is to compare predictions of analytical models most of the test results available.
and not those of various code formats. Another reason that the
latter has been avoided is that the writers have not found any
ACKNOWLEDGMENTS
published information regarding partial safety factors within
the JCI design format for shear strengthening of flexural mem- Partial support for this research has been provided by the European
bers. However, the reader may easily compare the design Commission Training and Mobility of Researchers Network Con-
model in the present study with the ACI-based formulation of FibreCrete, the Research Committee of the University of Patras, and the
Hellenic Secretariat for Research and Technology, Breca.
Khalifa et al. (1998) by introducing safety factors. This would
require multiplication of the values in Column (4) by ␣/␥f
(equal to 0.8/1.20 = 0.67, if CFRP is used and failure is not APPENDIX I. REFERENCES
controlled by debonding, or to 0.8/1.3 = 0.61 in all other cases) Antonopoulos, C. P. (1996). ‘‘Shear strengthening of reinforced concrete
and Column 5 by 0.7. It is obvious that the resulting changes structures using composite materials.’’ Diploma thesis, Dept. of Civ.
in terms of comparisons will be negligible. Engrg., University of Patras, Patras, Greece.
A careful examination of the data in Table 3 shows that the Araki, N., Matsuzaki, Y., Nakano, K., Kataoka, T., and Fukuyama, H.
comparison is, in most cases, favorable for the proposed ap- (1997). ‘‘Shear capacity of retrofitted rc members with continuous fiber
sheets.’’ Non-metallic (FRP ) reinforcement for concrete structures, Vol.
proach. This also is supported by the results shown in Columns 1, Japan Concrete Institute, Tokyo, 515–522.
7–9, which list the percentage difference between the three Berset, J.-D. (1992). ‘‘Strengthening of reinforced concrete beams for
predictions and test data. The prediction, in terms of the ef- shear using FRP composites.’’ MSc thesis, Dept. of Civ. and Envir.
fective FRP strain, of the model by Khalifa et al. (1998) is Engrg., Massachusetts Institute of Technology, Cambridge, Mass.
better than that of the proposed model in about 10% of the Chajes, M. J., Januska, T. F., Mertz, D. R., Thomson, T. A., and Finch,
cases [7 out of 76; note that the Khalifa et al. (1998) model W. W. (1995). ‘‘Shear strengthening of reinforced concrete beams using
externally applied composite fabrics.’’ ACI Struct. J., 92(3), 295–303.
is not applicable in 27 cases]. The prediction of the JCI ap- Comité Euro-International du Béton (CEB). (1992). ‘‘Eurocode No. 2.’’
proach is better in about 20% of the cases (17 out of 76; note Design of concrete structures, Lausanne, France.
that the JCI model is not applicable in 26 cases). These results Funakawa, I., Shimono, K., Watanabe, T., Asada, S., and Ushijima, S.
should not be surprising given the fact that the basis of com- (1997). ‘‘Experimental study on shear strengthening with continuous
parison has been data already used for calibration of the pres- fiber reinforcement sheet and methyl methacrylate resin.’’ Non-metallic
ent model. But these are all available test data, so no better (FRP ) reinforcement for concrete structures, Vol. 1, Japan Concrete
Institute, Tokyo, 475–482.
comparison may be made. Japan Concrete Institute (JCI). (1998). Technical Report on Continuous
Fiber Reinforced Concrete, Tokyo.
CONCLUSIONS Khalifa, A., Gold, W. J., Nanni, A., and Aziz, A. M. I. (1998). ‘‘Contri-
bution of externally bonded FRP to shear capacity of RC flexural mem-
A simple design approach is presented in this study for the bers.’’ J. Compos. for Constr., ASCE, 2(4), 195–202.
calculation of the FRP contribution to the shear capacity of Maeda, T., Asano, Y., Sato, Y., Ueda, T., and Kakuta, Y. (1997). ‘‘A study

204 / JOURNAL OF COMPOSITES FOR CONSTRUCTION / NOVEMBER 2000

 on bond mechanism of carbon fiber sheet.’’ Non-metallic (FRP ) rein- forcement for concrete structures, Vol. 1, Japan Concrete Institute, To-
forcement for concrete structures, Vol. 1, Japan Concrete Institute, To- kyo, 491–498.
kyo, 279–286.
Miyauchi, K., Inoue, S., Nishibayashi, S., and Tanaka, Y. (1997). ‘‘Shear APPENDIX II. NOTATION
behavior of reinforced concrete beam strengthened with CFRP sheet.’’
The following symbols are used in this paper:
Trans. Japan Concrete Inst., Tokyo, 19, 97–104.
Ohuchi, H., et al. (1994). ‘‘Seismic strengthening design technique for bf = width of FRP strip;
existing bridge columns with CFRP.’’ Seismic design and retrofitting bw = minimum width of cross section over effective depth;
of reinforced concrete bridges, R. Park, ed., 495–514.
d = effective depth of cross section;
Ono, K., Matsumura, M., Sakanishi, S., and Miyata, K. (1997). ‘‘Strength
improvement of rc bridge piers by carbon fiber sheet.’’ Non-metallic
Ef = elastic modulus of FRP in principal fiber direction
(FRP ) reinforcement for concrete structures, Vol. 1, Japan Concrete (GPa);
Institute, Tokyo, 563–570. (Ef ␳f)lim = limiting value of product Ef ␳f ;
Priestley, M. J. N., and Seible, F. (1995). ‘‘Design of seismic retrofit fc = compressive strength of concrete (MPa);
measures for concrete and masonry structures.’’ Constr. and Build. sf = spacing of FRP strips;
Mat., U.K., 9(6), 365–377. tf = thickness of FRP on each side of RC element;
Rostásy, F. S., and Neubauer, U. (1997). ‘‘Bond behaviour of CFRP- Vexp = experimentally derived shear capacity provided by
laminates for the strengthening of concrete members.’’ Proc., Compos. FRP;
Constr.—Conventional and Innovative, IABSE, Zurich, 717–722. Vf = shear capacity provided by FRP—ACI code format;
Sato, Y., Ueda, T., Kakuta, Y., and Ono, S. (1997). ‘‘Ultimate shear ca- Vfd = design shear capacity provided by FRP;
pacity of reinforced concrete beams with carbon fiber sheet.’’ Non- ␣ = reduction factor;
metallic (FRP ) reinforcement for concrete structures, Vol. 1, Japan ␤ = angle between principal fiber orientation and longitu-
Concrete Institute, Tokyo, 499–506.
dinal axis of member;
Sato, Y., Ueda, T., Kakuta, Y., and Tanaka, T. (1996). ‘‘Shear reinforcing
effect of carbon fiber sheet attached to side of reinforced concrete
␥f = partial safety factor for FRP;
beams.’’ Advanced composite materials in bridges and structures, M. ␥f,d = FRP material-independent partial safety factor if FRP
M. El-Badry, ed., The Canadian Society of Civil Engineering, Mon- debonding dominates;
treal, 621–627. ␥f, f = partial safety factor for FRP in uniaxial tension if FRP
Al-Sulaimani, G. J., Sharif, A., Basunbul, I. A., Baluch, M. H., and fracture dominates;
Ghaleb, B. N. (1994). ‘‘Shear repair for reinforced concrete by fiber- ε f,e = effective FRP strain in principal fiber direction (mean
glass plate bonding.’’ ACI Struct. J., 91(3), 458–464. value);
Taerwe, L., Khalil, H., and Matthys, S. (1997). ‘‘Behaviour of rc beams ε f,e,A = effective FRP strain in principal fiber direction—ACI
strengthened in shear by external CFRP sheets.’’ Non-metallic (FRP ) code format;
reinforcement for concrete structures, Vol. 1, Japan Concrete Institute, ε f k,e = characteristic effective FRP strain in principal fiber di-
Tokyo, 483–490. rection;
Taljsten, B. (1997). ‘‘Strengthening of concrete structures for shear with ε f,u = ultimate FRP tensile strain;
bonded CFRP fabrics.’’ Recent advances in bridge engineering, U.
ε max = limiting value of characteristic effective FRP strain;
Meier and R. Betti, eds., EMPA, Switzerland, 67–74.
Triantafillou, T. C. (1998). ‘‘Shear strengthening of reinforced concrete
ε max, A = limiting value of effective FRP strain —ACI code for-
beams using epoxy-bonded FRP composites.’’ ACI Struct. J., 95(2), mat;
107–115. ␳f = FRP shear reinforcement ratio;
Uji, K. (1992). ‘‘Improving the shear capacity of existing reinforced con- ␸f = shear strength (provided by FRP) reduction factor;
crete members by applying carbon fiber sheets.’’ Trans. Japan Concrete ␸f,d = shear strength reduction factor if FRP debonding dom-
Inst., Tokyo, 14, 253–266. inates; and
Umezu, K., Fujita, M., Nakai, H., and Tamaki, K. (1997). ‘‘Shear behav- ␸f, f = shear strength reduction factor if FRP fracture domi-
ior of rc beams with aramid fiber sheet.’’ Non-metallic (FRP ) rein- nates.

JOURNAL OF COMPOSITES FOR CONSTRUCTION / NOVEMBER 2000 / 205

View publication stats