CONCRETE LIBRARY OF JSCE NO.

26, DECEMBER 1995

EVALUATION OF SHEAR STRENGTH OF CONCRETE BEAMS

REINFORCED WITH FRP

(Translation from Proceedings of JSCE, No.508/V-26, 1995)

- a;;;._I'.-1,-I_;.-~'--II_, 3-

,' i1f;II3;'3{l' ;1.i;-25.1-Z.

'- 'u.'h%"
-Ii- 1.
=,'1-.'.r.-i ".=1_

1-1-

I"

,
1

__

" '~*"?' -

_::'T?

'.1;*=:,-.
.~=y~-..,l- -

-. -.

it
~

" '
. -.
.
-

- -. '
:.. . ~I

' .,;.

'
_

H _ _

I .

- T-11,211 . "fr.

:,,.""-=1':"=v;r'
it

_:_

f_'. 1 _
2._. .~ _ .

_,.j%_V\M.

"
._,,,.,_.,,.l.

:= _ "
-i

-.1-~.- . ,

i**\lZ>/We'r'iil'i
;,.+r=s;,vi;r
r , ii=
., ., - ..-an ,- 1 ~r
'.-_-.\. _'1v'
iii * *i-sit
Ir" ~ "~\

gin, ~i

Hikaru NAKAMURA

..'i*:?
~ 4.

"'?::
;.-; j-,;~;' J"|."*'.."",*'I_Q
1'ai;=>rc:."=
.-..~ '"
,
.:.-".-..'-.'.='-"=.,1';
i, 1;:-.
1,-' "- Ii i.
-=1.-,F"'='*f"-w
:1:-an-_~_"'- -_ "

.-'

,_

"'

-_

'-mi-'~;

' -I

"

'

rare:

_,: -,_\_,_,,'.=.;r__!__

V%%g;_."ii{T_iE_f%I?.,"'Jr}
.,,,r,,..._.,___,,,,,___

-=1-.
V .,, n -:.--3:-; .~ .=-,~,;;
~."- v.
i5i*1 V F. xi-~.,", tit,r"~:;~\@i~.-5.,-r"i
I -_
"4=..!,Q*5t5':!!i.i

_;=='.~5f-f"'f5*?

' .-i .1.;';,.ri'3",~,-

. *"'tl"'

:'i_jzZ

Takeshi HIGAI

The shear strength of concrete beams reinforced with FRP was studied through experiment and
analysis. The effect of longitudinal tensile reinforcement stiffness on the diagonal tension failure
strength was calculated using the extended modied compression elds theory, and an evaluation
method was proposed based on the analytical results. Furthermore, shear strength of beams failing
due to FRP stirrup rupture was well evaluated by considering the lower stiffness and the strength
reduction of FRP at bent corners.
A
Key Words : FRP, shear strength, bent bar strength, stijjiaess of reinforcement, modied
compression elds theory

Hikaru Nakamura is an associate professor in the Department of Civil and Environmental

Engineering at Yamanashi University, Kofu, Japan. He received his Doctor of Engineering Degree
from Nagoya University in 1992. His research interests include failure mechanism of RC
structures and application of nite element analysis. He is a member of ISCE and JCT.
Takeshi Higai is a professor in the Department of Civil and Environmental Engineering at
Yamanashi University, Kofu, Japan. He received his Doctor of Engineering Degree in 1972 from
University of Tokyo. His research interests relate to the shear failure of reinforced concrete
structures, dynamic behavior of concrete structures and maintenance and strengthening existing
concrete structures. He is a member of JSCE and ICI.

11l

1.

NTRODUCTION

In recentyears, the use of FRPqiber ReinforcedPlastic)rods in place of steel bars or prestressing

bars as reinforcementfor concretestructureshas been activelystudied. As a result, many useful
reports were presented[1],and both deformationand strength were evaluated for the nexure
behaviorof the concretebeams reinforcedwith FRP[2].However,the shear behaviorof concrete
beamsusing FRP has not everbeen established.
Many reports have made clean that the shear strength of concrete beams without shear
reinforcement.reinforcedwith FRP as the longitudinaltensile reinforcementis less than those with
steel bars[3][4].Tsuji et al.[3] proposeda methodto evaluatethe shear strengthfor suchbeams by
consideringa ratio of Young's modulusof FRP to that of steelbar in a term of the reinforcement
ratio of shear strengthequationof JSCE.Sincethe methodwas, however,obtainedon the basis of
a small number of experimentalresults, its applicabilityis not always clear. One report[5]show
that the shear strengthcan be calculatedclosely or safely in comparisonwith the experimental
results using the method proposedby Tsuji et al. Onthe contrary,anotherreport[6]showsthat the
correction considering a ratio of Young's modulus does not necessarily evaluate the actual
strength.
At the same time, it is reportedthat the shear strengthof concretebeams using FRP for shear
reinforcement is small, as compared with the calculated value by a modified truss theory
(vc+vs)[2]. Althoughseveralevaluationmethodsof the strengthof such beams are proposed, it
seems that, at present, no method gives satisfactorysolutions.The main reason why the shear
strength of such beams has never actuallybeen evaluatedappearsto be thaf the rupture of FRP
stirrups correspondingto yielding of the steelbars occurs at a crackintersectionOrbent corner?f
a beam due to local stress intensitybeforethe uniaxialstrengthis reached[7]-[12].Therefore,1tlS
importantto considersuchphenomenawhen evaluatingthe shear strength.
The purpose of this paper is to evaluatethe shear strengthof concretebeamsreinforcedwith FRP
rods as the tensile and shear reinforcement.First, the shear resistance behavior of the concrete
beamsreinforcedwith FRP rod is confirmedby experimentalobservation.Second,the bent corner
strengthand the diagonaltensionstrengthof FRP rods,whichare typical mechanicalpropertiesof
FRP in concretebeams, are investigatedtheoretically,and the effectof Young's modulus of the
reinforcements on shear behavior is analyzed using an extended modified compressionfields
theory. Finally, the shear strength of the beams is evaluated-using the results of the above
investigations.
2. OUTLINEOF EWERmNTS

GR4afgiai&
FRP used consistsof continuousglass fibersimpregnatedwith resin (GFRP).GFRP was used so
that the effect of Young's modulus of the reinforcementon the shear strength could be
investigated,since the Young's modulusof GFN is very small comparedwith the steel bars.

(sGdlA)p;Ss6)foE#a
i:,t:ssgnied:tl:ith.fSP.Sc;ncgm
o2fioercemu:endd,
tAed
aEeea
rZ5iuC:OSsf
=cain.nre!nff2T.c;Tln2tS.

the bent comer was 5mm or 10mm.The materialproperties of GFRP and steel bars are shownin
Table-1. The tensile strength and Young's modulusof GFRP were obtainedby the test proposed
by Idemitsu[14].The compressivestrength of the concrete(f'c)is shownin Table-2.

b
An outline of the beam specimenstested in this study is shown in Table-2. The cross secdon,
without stirrups, was 30cm wide, 20cm high and with an effective beam depth of 15cm.The
beams were tested on a two-pointloadingsystemwith monotonicallyincreasingloads.

-112-

The cross section with stirrupswas a width of 20cm, height of 30cm, and an effectivebeam depth
of 25cm. Parameters of the experimentare (1) material of longitudinalreinforcement(GFRP,steel
bar), (2) material of stirrup(GFRP,steel bar), (3) spacing of the stirrups and (4) the radius of the
bent corner of GFRP stirrups. The beams were tested on a one-point loading system with
monotonically increasing loads. Fig:1 gives details and its loading condition of specimens of
GGO5-20and GG10-20. This experiments are characterizedby the radius of the bent corner of
GFRP stirrups and allow investigation of the effect of the bent bar strength on beam shear
strength.
3. ENERWNT

RESULTS

The experimentalresults are summarizedin Table-3.All test specimensfailed in shear mode and
the longitudinalreinforcementsdid not yield or break at failurein any specimen.

&
ub

For beams withoutstirrups,the nemral crackpropagatedto the upper part of the crosssectionas
soon as the crack initiated. The ultimatefailure was determinedby the diagonaltension crack
which occurredfrom the flexural crack at a distanceof about 1.5d fromthe loadingpoint. The
failure modeof the specimenswas diagonaltensionfailure.

Crack formationin the specimenswith stirrupswas, before failure, almost the same as in the
specimenswithout stirrups.The diagonal crack occurredfTromthe nexural crack at a distanceof
1.5d- 2.Od from the loading point. The crack formationof a specimenof GGO5-20at the
ultimate is shown in Fig:2. The load carrying capacity increasedby the effect of the stirrups
after the diagonal tension crack occurred.The specimensreinforced with steel bars as stirrups
failed after the stirrupsyielded,while the failure of specimensreinforcedwith GFRPstirrups,was
causedby the breakingof the stirrups.GFRP breakingoccurredat the upper or lowerbent corner
near the diagonaltension crack. One purpose of this experimentsis invesdgationof the effect of
the bent bar strengthon shear strength.It is understoodfrom the experimentalresults that the
shear strength of the specimenswith a bent comer radiusof 5mmis obviouslylowerthan that of
10mm;the strengthof the bent corneris an importantfactor for evaluadngthe shear strengthof
the beam reinforcedwithFN.

Table-2 Outline of Specimens

S p e c im e n
Table-1 MaterialProperties

L o n gi tu d in a l

sp a c in g

re in f o r c e m e n t

o f s tirr u p
Cm

A re a o f

Y o u n g 's

S t rai n a t

c ro s s s e c tio n

T e n sil e
s tre ,n g th

m o d ul u s

b re a k h g

MP a

MPa

Cm
G 16

2. Ol

75 l.4

2 .94x lO 4

8 28.0

3. l4x lO4

2.8 7

37 l.6

i . 8 O x lO 5

D 10

0. 7l

324. 6

i . 9 0 x lO 5

D 6

0. 32

370. i

G6
D 19

O.35

i.8 0xlO 5

2.5 5
2.64

C ol

Pt

r
mn

3-G IG

a/ d

f'c

1.34

4. 0

22. 7

XP a

'G O 2

4-G 16

i.79

4. 0

27.8

G GO5 -10

4 -G l6

G 6 (10)

I. 6 l

3. 0

35.4

4 -G 16

C 6 ( 10)

10

i. 6 1

3. 0

33.4

GG O5 -20

4 -G 16

G 6 (2 0)

I, 61

3.0

35.2

GG 10-2 0

4 -G1 6

C6 (2 0)

lO

1. 6l

3 ,0

3 5.2

GG lO -10

DG O5- 15

3 -D1 9

G6 (15 )

1.7 2

3 .0

34 .7

D C lO- l5

3- Dl9

G6 (15 )

10

i.7 2

3 .0

34 .4

D G O5-2 5

3-D 19

G6 (25 )

1.7 2

3. 0

35. 6

D G 10-25

3-D 19

C6 (25 )

i.72

3. 0

35. 8

CD -15

4-G IG

D6 (l5 )

1.6 l

3. 0

38. 6

CD -25

4-G 16

D6 (25 )

1.6 1

3. 0

37. i

-113-

lO

Table-3ExperimentalResults
S p e c im e n

D ia g o n a l c ra c k in g lo a d
V c.eq '
(a

V c.)
(a )

V c,2
(b ?

S h e a r str e n g th

v c,ca[i ,c.c,p

V exp
(fO

V c2+ V s)
(a )

V c2+ V s2
(a)

v co[/v eg

Col

6.75

9.9 1

6. 13

0. 9 1

G O2

7.4 l

l1.67

7.2 2

0. 97

G G O 5 -l O

10.00

l3 .21

8. 18

0. 82

l7 . 0 0

33 .38

16.08

0.95

G G lO -1 0

9.50

l3 . 0 4

8.0 7

0.8 5

20.4 0

33 .27

17.48

0 ,9 1

G G O 5 -2 0

7.50

13. 20

8. l7

1. 09

l l. 4 6

2 0.77

12.0 1

1. 0 5

G G 10 -2 0

8 .50

13. 20

8.17

0. 96

13. 50

20 .77

12 . 8 8

0. 95

D G O5-15

16.50

l3. 61

0.8 2

l9. 66

30 .4 1

l8 . 8 7

0. 96

D G lO - l 5

14 . 9 0

13.5 7

0.9 l

2l. 68

30 .37

1 9. 8 4

0. 92

D G O5-25

13. 90

13. 73

0.9 9

16.2 7

23 .81

16. 72

1. 03

D G l O _2 5

14. 00

13.7 6

0.9 8

16.2 7

23. 84

l7. 52

i. 08

G D - l5

10. 00

14.04

8 .69

0.8 7

15.73

14. 7l

0 . 94

G D -2 5

8. 50

l3.8 6

8 .58

1.0 1

l l . 16

12. l9

1. 09

vv::
clip;BTla;;,eSqtueaeti
obna,rs,
vm::f
:fE.?.
t83RP
Vs,):Truss TheoryVyAs(U1.15)/s),
Vs,2:Eq.(1l)
Vco/:Vs.I for steel bar, vs.2 for stirrup

lp

20

4@
Cn

.i.

3@ 20

75

i?_.
ls
ItL)

l5

Figure-i Detail and Loading Condition of Specimen(GGO5-20,GGlO-20)

/ R u pture of G FR P

"

Ji#
_J

l l

5
I

Figure-2 Crack Fomution at Ulumate SLate(GGO5-20)

a&

As shown in experimentalobservations,bent bar strengthis a major factor for shear strength.

Therefore,we must first Clarifythe bentbar strengthin orderto evaluatethe shear strengthof the
concrete beams reinforced with FRP. The evaluation equation of the bent bar strength is
theoretically.derivedas follows.
We assume an bent cornerFRP cast in concreteas shownin Fig:3. We considerthe deformation
at the cross section varied from straight part to bend, since it is experimentallyobserved that
breakingat the bent corneralwaysoccurat that location[8].When a tensileforceis appliedto the
bar as shown in Fig.-3, and there is no bond between concrete and FRP, the FRP stretch 8x
uniformlyat the cross secdonin the straightpart subjectedto uniformaxial force. It is assumed
that the cross sectiondeformrotationangleof 4)(fad)maintaininga radiusof the bent cornerof r.
Then, if Bemoulliassumptionis used,the straindistributionin the crosssectionis representedby
a followinghyperboliccurve at any point y in the cross section sinceFRP should deform6x at
the bent part.

-114-

i
0.9

i.'

After
Deformation

Before

0.8

>o.7
+

0.6
0.5

- -+TTI/I

0_4

0.3

/
/

t i pE;x&ns%ent

0.2
0.1

Deformadon
FiglEe-3Model of Bent Bar in Concrete

rn1

Figure-4 Evaluadon of Bent Bar Strent,dh

a(y) -8x/((r+y)0)

(1)

Stress distributionis obtainedby multiplyingYoung's modulusi to strain,whenFRP behavioris

strictly elastic. Then, the stress distributionat breakingis representedby the following equation.
Here it is assumed that the breaking of FRP occur when a stress in cross section reach the
breakingstress of q.
o-oy.r/(r+y)

(2)

Therefore,the evaluationequation of bent bar strength can be derived by integratingthe stress

distributionof Eq.(2) in the cross section.
T

(3)

-JAG,.r/(r.y)dA

The equationis rewrittento Eq.(4) as the function of the radius of the bent.corner,r, and the
height of the cross section,h, whenthe cross sectionis rectangularwith the height ofh.
i7= Tu.,/hlm(1+h/r)

(4)

in which Tu is the uniaxialstrength,r is the radius of the bent corner,and h is the heightof the
crosssection.
Fig:4 shows the comparisonbetweenthe evaluatedvalues using the proposedequadonand the
experimentalresultsobtainedby Miyataet al.[9].h the figure,the lateralaxis showsa ratioof r
to h and the vertical axis showsthe reductionratio of the strengthof the bent cornerfrom the
uniaxialstrength.A solid line showsthe Eq.(4) and a + indicates an experimentallyobtained
value. It is understoodthat the proposedequationcan satisfactorilyevaluatethe bent bar strength.
QiBi494aiW
The effect of the diagonaltensileforce due to a diagop_al
crack,as shownin Fig.-5,is considered
a factor of shear strength.Next, we investigatethe evaluationmethodfor diagonaltensile strength
of FRP as anisotropicmaterial.
Considerthat FRP with a lengthL is appliedto diagonaltensile force T with angle 0. Then,the
stressresultantacted in the crosssectionis the axialforceis T.cosOand the bendingmomentis T.
L.sine. A strainin the cross sectionreach breaking strainfrom a smallertensile forcethan with
uniaxialstate becauseof the effectof the straingradient.Therefore,the diagonaltensilestrengthis
less than the uniaxialtensilestrengthsincethe cross sectionof FRPwithoutplasticregion,breaks
as soon as one part of the cross sectionis broken, Then, the diagonaltensile strengthof FRP is
representedby the followingequationsfor a rectangularcrosssectionwithheighth.

-115-

=0.9
rF

0.8

t1' 1S

0.7

tt

0.6

0.5

-tS

: Carbon
A : Aramid
I : Glass

0.3
0.2

01
0

Figure-5DiagonalTension due to DiagonalCrack

T-Tu/(cosO+6L.sine/h)

1
^

0.4

Jt

10

20

Eq.(6)

7b

tt`

Eq.(7)
30

40

degrea

Figure-6ComparisonwithExperiment

(5)

in which Tu is uniaxial tensile strength, 0 is the direction of the tensile force, and h is the cross
section height.
h the above equation, L is the unknown value. Now, we assume that L is the length of
intersection between FRP and the diagonal crack with angle of 0. Then, L-h.tanO and the
diagonaltensile strengthis derivedas,
T= Tu/(cosO+6tan0.sine)

(6)

For circular.sections,
T- Tu/(cosO+8tan0.sine)

(7)

Figure-6 comparesthe experimentalresultsfrom Maruyama et al.[10] and the proposed equations.

The experimentalresults were obtainedfor carbon(CFRP),aramid(AFRP),and.glass(GFRP)with
the directionof diagonaltensileforce variedfrom 0 to 30 degree.In this figure, a solid or broken
line correspondrespectively to Eq.(6) and Eq.(7), while the symbols + , A and I showthe
experimentalresults for carbon, aramid, and glass fibers. It is understoodthat strength reduction
due to diagonal tensile force is reasonably evaluated by the proposed equations. However, the
equationscan not considerthe effect of fiber type, even though the strengthreduction is different
for differentflbers in experiment.

5. EVjuUATION OF SmAR STRENGTHOF CONCRETEBEAMSRBWORCED WITH

FRP

uM
It is known from many experimentalobservationsthat the shear strength of a beam reinforced
with FRP, which has a lower Young7smodulus than steel bar, is smallerthan those reinforced
with the steel bar[3][4],This fact is confirmedby this experiment(Got and GO2specimensin
Table-3) and the strength is about 30 - 40% smaller than the value predicted by the equation
proposedby Niwa et a1.[15].It is reportedthat the reason is mainly the differencein Young's
modulus of reinforcement.Therefore, we will analytically investigate the effect of Young's
modulus of the mainreinforcementon the shear strength.
Analysis is a method based on the extended modified compressionfields theory which can
accuratelyevaluatethe shear strengthof the concretebeams[13].The model used in the analysis
has a cross sectionof 20 x 20cm, a beam depth of 16cm,and a longitudinalreinforcementratio
of 2.68%, as shown in Fig.-7. The material properties are that the compressivestrengthof the
concreteis 28MPa and the tensile strengthis 2.8MPa.Analysisis performedin which Young's

-116-

V :2.00

`
S6
3
LL

20

D:i.00

A:0.50

0..o.25

o5
o5
o5

t`
3
i)

5;4

16

4cm

u
1
2
M3.men((Gem)
Figure-8 Effect of Young's Modulus of Main Reinforcement

Figtlre-7 An(1l_yLicalModel

^1.2

modulus of the reinforcementis varied from

o.25 x 105fo

2.0 x 105MPa.

The

iul

(n6) I/4

lL)

relationshipsbetween the shear force and the

moment in failure state obtained from the
analysis are illustrated in Fig.-8. The results

O
Il

o.8

X
S. o,6
llil

0 :a/d:i

i?6 xyolu.ng',S.?50dxullu.S
5Vaaln3es..20f
2.o1.X
51Xi
0:a/d=2
S- o.4
are respectivelymarkedwith V , D , A and
^
A:a/d=3
0. The range in which the shear force
Luv0.2
V:dd=4
>
gradually decreases as the moment increases,
correspondsto diagonaltension failure in the
0
0.5
.5
2
2.5
analysis. On the other hand, the range in
Ei X 105(MPa)
which the moment is constant and the shear
force decrease rapidly corresponds to the FiglEe-9yR.eahbie
T?Mh.pdbulesTe.efnb?%arRSepnfan.
r?eaEoenntand
nexural failure region. The load carrying
capacity for two-pointloading and simplyT
suTpported
beams can be evaluatedfor both shear and
nemral failure from this figureusing the relationof a/d-M/(V.d).However,this analysiscan not
be applied to shear compression failure, since the analysis does not consider the effect of
transverselocal compressivestress.
i

As shownin Fig:8, the shear strengthdecreasealongwithYoung's modulus.It is understoodthat

the Young's modulusof the main reinforcementis an importantfactor influencingshear strength
and it can be seen that the Young's moduluscurves are almost parallel.This implies that the
effect of Young's moduluson shear strengthis independentto a ratio of the shear force to the
moment and is a functionof Young's modulusonly. Figure-9showsthe reductionratio of shear
,,.+.^n^L

C`^-4_

._J.__A.,J1_

I_

.L1__

____

_i.

1T_

_)_

I1^

1^

Jrn

nl1

..

.1

rT

from strength in the case ofyoung's moddlus oTf2.0 x 10 5 MPa. The symbols 0 , a
ArTagtnhd
V correspond to a/d-1, a/d-2, a/d-3, and a/d-4. The difference of a/d does not appear.

It was confirmed and similarly concluded the effects of longitudinal reinforcementA;alio,

compressivestrength of the concrete,and effectivedepth of the cross secdon in the analysis.
Therefore, the effect.of Young's modulus on the shear strength can be introducedby simply
multiplying the prev10uSlyintroduced shear strength equations by the effect of the Young,s
modulusof the longitudinalreinforcement.

?hhoeeuaTag:lag,leynEfglu!ru;:tuh;Si?cf,ce;eesb;adltl?inee::;n.eracfid
Sa?
oecnvu?=fa.o:fhdeelsi;;::T8.,?F
naa;1ayb,
othOef
v - vc(Ei /Es)1/4

(8)

in which Vc is the shear strength of the concrete beam reinforcedwith steel bar, Ei is the Young,s
modulus ofFRP, and Es is the Young's modulus of the steel bars.

-117-

oO
oO

1.6

1
H:

tu i.4
3
>
1.2

! :.TsouFsed

0 :Niwa

0
0
0

V c a lN e x D

_a

&

0.8

0.6

Table-4 Comparison betweenEstirnatedand ExperimentalValues

:rK

0OA

0.8

C o e ffic ie n t o f v a r ia t io n %

N iw a

T s u ii P ro p o s e d

N iw a

G , G + C (7 )

i.50

0 .8 O

T s u 1'i

P rop osed

0 .94

7 .4

7 .2

(u )

I.09

6.9

0 .8 6

0 .9 l

lO . I

7 .3

A ll s p e c im e n s

i.25

8.0

0 .8 4

0 .92

l8 . 0

8 .0

7.8

1.2

pwoi(res)(
%)
Figure-10ComparisonbetweenEstimatedand
ExperimentalValues
6

.`

Y)
u
t<

Et4
.1

C13

g5

i)
U
t`

a4
i-

3
4}

i?;3

0.004

A :2.00
D:1.00

x
x

A
0

x
x

0.008

:o.50
:o.25

0.012

.i

0.0002

ShearStrain

:2.00

D:I.00
A :o.50

x
x

0..o.25

o5

os
o5
o5

0.0004

CurvaLure(1/cm)

FiglEe-ii (a)f.eiaceoannsdhighbeearhvseEZiSheaI Figure-11O')FR.efgEoAsdhi8E%ereen

shear

Tsuji et al. proposeda methodof evaluatingthe shearstrengthof concretebeams reinforcedwith

FRP using the transformed area of reinforcementAsPinEs), consideringa ratio of Young's
modulus of FRP to that of the steel bar[3]. The results in our analysis are identical with the
method proposedby Tsuji et al., since the effect of the longitudinalreinforcementratio on the
shear strength is also proportionalto the 1/4 power in the analysis[13].However,the effect of
young's modulus as proposedby Tsuji et a1. differsfrom this analysis,The effect on the shear
strengthis representedby a 1/3 power,sincethey use the JSCE equation.
Figure-10 showsthe ratio of esdmatedand experimentalvalues[3][4][6][7]of shear strength.The
estimatedvalues are obtainedfrom the equationproposedby Niwa et al.[15Lwhich is proposed
the method of Tsuji et
for the conpcretebeams reinforced with steel bars(indicatedby
a1.(indicatedby A ), and the proposedmethod of Eq.(8)(indicatedbyc'l,). It is assumedthat Vc is
evaluated by Niwa's equatiod add that the Young's modulus of the steel bars in Eq.(8) is 2.0 x
10 5 wa.

the number-of

specimenswere hybrids of ea?aesrsimazndtaia,rbeSun1(tLlc8)?aosr

n7ef,r&iin-f20.r,Cexd
1%t4h-gl3?S3S.(Gx),
1Tbi&1a1)
a

And 18 used Garb-on(C)OEi-6v.73x 10 4-

1.4tx 16 4 MPa). table-4

shows the average and a

coefficientof variationfor the ratio of estimatedand the experimentalvalues for each method.
The equationfor concretebeams reinforcedwith steel bar overestimatedthe experimentalresults
and the differenceappearsremarkablyfor lower Young's modulus.Tsuji's methodunderestimated
the test results. On the other hand, the proposed methodcan satisfactorilyestimatetest results,
regardlessof the differenceof Young's modulus.The coefficientof variationis also satisfactory.
Figure-11 show the relationshipsbetween shear force and shear strain and between shear force
and curvaturefor M/(V.d)-3 when the Young's modulusof the main reinforcementis varied in

-118-

10
I'''T

Rtthtt

f'8

Ei=2.Ox100kgcmi

20

/-<1
A/

Ei=2.94xlO5kgucm2

25

?
+

0.5

Eq.(8)
1

1.5

?_.5

3.5

Vs(tf)

Cm'

Figure-13 AnalyticalModel

Figure-12 EMfFu?;tROefH.EcnegisenTodulus
of
7__5

2,5
1
>:
u

1
y.
lJ

>t2

>3?-

.1
(J

'J

>

+
0

1.5

>
1.5

A
+ : r-5mm(GFRP)
0 : FlOmm(GFRP)
A : steel barstirrup

0.5

10

15

0.5

20

25

30

Spachgof Shp(cm)
Figue-14 Estimated een%orfsE&L?negthqe
nt Dan
(ignoring !

+ : r-5mm(GFRP)
0 : r-lOrrm(GFRP)

10

15

20

?_5

30

Spacingof Sdrrup(cm)
Figure-l5 EsdmatedResultof ExperimentData
(consideringbent bar Strength)

t2?Saxnall.ys5i
s,.TT?.! ,1.]5,,A.,.5anxd
1? 5STnbdOl
s.?z;exsp10.nd5
tXha?anrealsy,seicsbfvoerlyYou;gh'.SuThOduBuq
(gf
satisfactorilyestimatesthe effect of Young's modulus on shear strength,the mechanicalmeaning
was not clear. However, as shown in Fig.-ll, the shear strain which is identical with shear
deformationincreasesrapidlyin proportionas Young's modulusbecomessmall and the effect of
young's modulus on shearstrain is great comparedwiththat on the curvature.Therefore,it seems
that the inctease of the shear strain due to the lower Young's modulus probablyinfluencesshear
strength.

Qi*p&

The shear failure modes of the concretebeams are distinguishedinto failure causedby yieldingor
breaking of stirrups and compressivefailure of the concrete near the loading point, even if
diagonalcrackspredominate.Both failure modes are basicallydifferentand the shear strengthfor
both modes can not be evaluatedwith the same estimatingmethod. In this paper, only shear
failure causedby the breakingof FRPstirrupsor the yieldingof steel bar is considered.
Figure-12shows the analyticalresult of M/(V.d)-3 obtainedfrom analysisbased on the extenqed
modified compressionfields theory.The model used in the analysishad rectangularcross section
of20 x 30cm, beam depth of 25ch, longitudinalreinforcementr-atioof 1.61%,stirrup area of 2 x
0.35cm 2,and
stirrup spacing of 10cm, as illustrated in Fig:13. Analysis was_pe5rformed
in which
v'rvyAAA

>

YYA-Y

UY-J.-Y,I

IT"I---O

y-

T--I7

"-

MIIlr----TTTI

-T

1^

the Young,s modulus of the longitudinal reinforce&ehtu&as eifhe; 2.0 x Ilo5 or 2.94 x 10
U

Jl

force

2TdaihT%1tee,talhatax7f
sstiTwS
Eaes
s2h
;5a,xrels?
s4tl
nTfa&cTeh
e.fV
Shnei
csatll
Ln,ss.
shTOhTssythmeb
:FpiedinS?Z
:rfigure
represents the shear strength without stirrupTascalculatedby EqM.(8).
The shear resisting force of

-119-

the stirrups is calculatedby the followingequation.

vs - Bwh/n2G,i /tanOE.

(9)

in which oyi is the stirrupstress obtainedfrom the analysis,Oiis the directionof principalstress,
n is the numberof subdividedlayers in cross section, and Bw and h are the width and the height
of the cross section.
As shown in Fig.-12, the shear resisting force of the stirrup yields before the diagonaltension
crack occurs and increases graduallyaRer that. Also, it increasesin linear proportion after the
applied force reachesthe shear strengthof Eq.(8). This tendencyis the same, independentof the
Young's modulusof the longitudinalreinforcementand/orthe stirrup.This result impliesthat the
shear resisdng force of the concretebeam with stirrupis approximatelyevaluatedby the modified
truss theory. h this paper, we tried to evaluatethe shear strengthof such beams by using the
modifiedtruss theoryin which yielding or breakingof the.stirrup at the ultimate state is assumed.
V =Vc+Vs

(10)

in which Vc is the shear strengthconsideringthe effectof Young's modulus of the longitudinal

reinforcement,Eq.(8),and Vs is the shear force carriedby the stirrupscalculatedby a trusstheory.
Figure-14showthe estimatedresults of the experimentdata by Eq.(10).Here, Vs is calculatedby
assumingthat the compressivestress declinesto 45 degree and the stirrup stress is the uniaxial
yieldingstress or the breakingstress.The A symbolsin the figure show the results with steel bar
stirrups. The shear strength can now be accuratelyevaluatedby the truss theory. Therefore,the
shear strength of the concrete beam with FN as longitudinalreinforcementand steel bar as
stirrup can be calculateconsideringonly the effect of the Young's modulus of the longitudinal
reinforcement on Vc. On the contrary,the shear strength of that with FRP as the stirrup is
overestimatedby Eq.(10).Therefore,in this case,Vs can not be evaluatedusinguniaxial strength
ofFRP in the same way as with the steel bar stirrup.One reasonis that the strengthof FRP in the
concrete deterioratesfrom the uniaxial strengthdue to the local stress intensity or the combined
effect of the tension and the shear at the bent and a crack intersection part. Therefore,we
attemptedto modifyVs consideringthe innuencingfactors.

i&

In Fig.-14, the shear strength of the specimenswith differingradius of bent corner is different.It
is understood that the strength of the bent corner must be consideredcorrectly to evaluateVs.
Therefore, we re-evaluateVs consideringthe strength of the bent corner using Eq.(4). Figure-15
compares estimated values using Eq.(4) with the experimental results. The differing shear
strengthsfor specimenshaving differentradius of bent comerthen disappears.By using Eq.(4) in
Vs, the effect of the bent corner on the shear strengthis satisfactorilyevaluated.However,since
the estimatedvalues are larger than experimentvalues,we showfurther modiflCationof Vs.

hM
The factor relatedto the shear strength,besidesthe strengthof the bent corner,is the effectof the
diagonaltensile force. h this case, Eq.(6) and Eq.(7) proposedfor diagonaltensile strengthmay
be applicable.However,the equationsrequirethe crack angleat crack intersectionand the angle
is not always uniquely determined.Therefore,the evaluationof the shear strength using Eq.(6)
and Eq.(7)is not consideredin this study.
The other effecton the shear strengthis Young's modulusof the stirrup. Satoh et a1.[17]report
from the analyticalresult of FEM that the shear strengthbecomessmall when Young's modulus
of the stirrup is low. We therefore investigatethe effect of Young's modulus of the sdrrup
analyticallyand evaluatethe shearstrengthconsideringthe effectin this section.

-120-

8
i_1

\7

S7
O

t-

'=6

,?` 6

J4

cn-.5

Ld
Q

a) 3

4
3

withoutstirrup

?o.i2h&Xli.Z:kEgcccm:222

0.2

0.3

0.4

2.00x 100kgucrn2

0.5

0.002

Curvature(10-3/cm)

0.004

0.006

0.008

Shear Strain

Figtue-I6 Raneiasdzensarhip.?cegween
curvahre
(young's

'-

;v5Bjo.su6!E%&h:c

1
0.1

g3Z
A

modulus of mah reirdoecement of 29400MPa)

Figtue-I7 RaneiagE:ihip.brceehveen
ShearSwain
(Young's modulus of mah rehdoecement of 29400MPa)

10

10

ql
U

//

G=

88
t<

/I

\8

O
(I

L<

`
56
a)

LJ

r.

=76
Cn

A
`)

?without

sdrrup

2::a-&x
iso6;kEgg:gg2Z
?

0.05

0.1

0.15

0.2

Curyature(1 0p3/cm)

0.001

_ithoutsdmp

gis?eq:SigSSg
0.002

0.003

0.004

ShearStrain

Figure-18Reladonshipbetween Curvature
and ShearForce

Figure-i9 Raneiagz=arhiF.brce!ween
ShearSBain

(Young's rnodulus of main reirJoecement of 200000MPa)

(Young's modulus of main reinfoecement of 200000MPa)

The analyticalresults based on the extendedmodifiedcompressionfields theory are shown in

Fig.-16 - Fig.-19, for the RC cross section shown in Fig.-13. Figure-16 shows the relationship

between the curvatureand the shearforce, while Fig:17 shows the relationshipbetweenthe shear

;arhSaegn:
a:ded
tth:ae.tuSonhfge,a!refmAr.ie:1fioesri
A;r'c:e';'e==3tali;h2,?e;i4ntfh.xe,c1:0:Yenn%5:Hfofd.g?-:
d;effsoult3
Young's modulus of the sdrrup does not appearfor the curvatureand shearforce relationshipsas
shown in Fig:16 and Fig:18. This means that the Young,s modulus of the sdrrup is not a
predominantfactor for nexural deformation.On the other hand, the shear strain and the shear
force relationshipsas shown in Fig.-17and Fig.-19are innuencedby the Young's modulusof the
stirrup. The shear strain increasesremarkablyfor lower Young's modulus of the stirrup. If the
shear strain is identicalwith the shear deformation,the deformationof the beam is dominatedby
the shear deformationand directly dependson the Young's modulus of the stirrups. Therefore,
assumingthat the shear strengthis closelyrelatedto the shear deformation,Vs may be evaluated
by consideringthe effectof Young's modulusof the stirrup on the sheardeformation.The method
consideringthe effect of the shear deformationin Vs is alreadyproposedby Kobayashiet al.[6].
Consideringthe shear deformationindirectlyrepresentsthe strengthreductionof FRP due to the
combinedeffect of the tensionand the shear causedby the local shearstrain in beam.
Since the shear strain and the shear forcerelationshipsare almost linearfor every case,the effect
of Young's modulus of the stirrup is investigatedby the shear stiffness.Figure-20 show the
reduction ratios of the shear stiffnessfrom the results of Young's modulus of 2.0 x 10 5MPa.

-121-

2.5
ill

1
`

'd:ao.8

>1_
2
n
U

>

cEiG)0.4

Ba 0.6

1.5

S2 L.

Young's Modulus ofMain Reinforcement

;gao.4
0.?_

0.5

1.5.

: r-10mm(GFRP)

young,s Modulus of Sdrrup( x lO5MPa)

Figue-20 Effect of Young's Modulus of Stirrup
on Shear Stiffhess

The symbols with + and A show the results

when the Young's modulus of the main
reinforcement is 2.94 x 104wa

+ : r-5mm(GFRP)

0.5

:2.Ox106kgucn2

and 2.0 x 10 5

MPa, respectively. The reduction of the shear

stiffnessis independentof Young's modulusof the
main reinforcementand depends on that of the
stirrup. The effect is proportionalto the 0.4 power
of the Young's modulusof the stirrup. Assuming
that the effect of the stirrupis related to the shear
deformationdirectly,Vsis modifiedas
vs* - vs(Ei /Es)0.4

10

15

20_

25

30

Spacingof Stimp(cm)
Figue-21 Esdmated Result of E:(9eeiment Data
Table-5 ComparisonwithExperiment[1
1]
S p e c im e n

fy
MPa

C -A 2-0

l5 4 0 . 0

C - C 2 -0

14 4 0 . 0

C - C S 2 -0

20 80. 0

E L.
MPa
7 .00x lO 4
I.l2x lO 5
i.47x lO 5

V eg
tf

V oa/
tf

v cd /v a c

8 .65

9.5 9

i. ll

8 .50

9.8 3

i . l6

lO.0

9.6 9

0.97

(ll)

in which Ei is the Young's modulus of the stirrup and Es is the Young's modulus of the steel
bar(2.0 x 10 5wa). Figure-21showthe comparisonbetweenthe experimentaland the estimated
valuesusing Eq.(11). The applicabilityof the proposedmethod is found satisfactory.
Since the above investigationsare obtained from concrete beams reinforced with GFRP, the
applicabilityof this methodwas then checkedby comparingwith test results carried out by Wakui
et al.(a/d-3, s-25cm) using other FRP[11]. The mechanicalpropertiesof these FRP stirrups are
greatly different from those of our experiment,as shown in Table-5. The estimatedresults are
summarizedin Table-5.The accuracyof estimationby the proposedmethod is satisfactory.
6. CONCLUSIONS

(1) Experimentalinvestigationshowsthat the radiusof the bent comeris an importantfactor for

the shearstrengthof a concretebeamwith FRP as the stirrup.
(2) Equations to evaluatethe bent bar strength and the diagonal tensile strengthof FRP are
proposedfrom the theoreticalinvestigation.They are useful for evaluatingthe shearstrengthof
concretebeamsreinforcedwithFRP.
(3) For the shear strengthof concretebeams reinforcedwith FRP as longitudinalreinforcement,
the Young's modulusof the longitudinalreinforcementis a factorinnuencedto the shearstrength
and the effect can be evaluatedaccuratelyby the term of 1/4 power of the ratio of Young's
modulus.The effect of Young'smodulusratioon the shearstrengthis similarlyrepresentedto the
effect of the longitudinalreinforcementratio.
(4) The shearstrengthof a concretebeam reinforcedwith FRP as longitudinalreinforcementand
with steelbar stirrupcanbe accuratelyevaluatedby the modifiedtruss theory.

-122-

(5) The strengthof FRP in concretedeterioratesfrom the uniaxial strengthdue to the local stress
intensity or the combinedeffect of the tension and the shear at a bend and a crack intersection.
The shear strengthof the concretebeamreinforcedwith FRP stirrups,whichfails by breakingof
FN, is satisfactorilycalculatedby the proposedmethod when the bent corner strength and the
effect of shear deformationis consideredin Vs.
References
[1]JSCE Subcommitteefor Studieson ContinuousFiber : Present state of technologyconceming
applicationto the field of civil engineeringstructuresof concrete-basecompositematerial using
continuousfiber, 1992(inJapanese).
[2]Kakuta Y., et a1. : Establishmentof design method of concrete structuresusing continuous
fiber, Report of Ministry of Educationin the form of Grant-in-Aidfor co-operativeresearch(A)
No.04302040, 1994(in Japanese).
[3]Tsuji Y., Saito H., Sekijima K. and Ogawa H. : Flexural and shear behaviors of concrete beams
reinforced with FRP, Proc. of JCI, Vol.10, No.3, pp.547-552, 1988(in Japanese).
[4]Kanekura S., Maruyama K., Shimizu K. and Nakamura Y. : Shear behavior of concrete beams
reinforced by CFRP rods, Proc. ofJCI, Vo1.15, No.2, pp.887-892, 1993(in Japanese).

[5]Yokoi K., Shima H. and Mizuguchi H. : Applicability of shear strength equation of RC beam
for reinforced concrete beams with FRP rods, Proc. of JCI, Vo1.14, No.2, pp.713-716, 1992(in
Japanese).
[6]Kobayashi T., Maruyama K., Shimizu K. and Kanekura S. : Shear behavior of concrete beam
reinforced by CFRP rods, Proc. ofJCI, Vol.14, No.2, pp.70l-706, 1992(in Japanese).
[7]Wakui H., Tottori S., Terada T. and Hara C. : Shear resisting behavior of PC beams using FRP
as tendons and spiral hoops, Proc. of JCI, Vol.1 1, No.1, pp.835-838, 1989(in Japanese).

[8]MaruyamaT., Honma M. and Okamura H. : Experimentalstudy on the tensile strength at the

bending part of fiber reinforced plastic rods, Proc. of JCI, Vol.12, No.1, pp.1025-1030, 1990(in
Japanese).
[9]Miyata S., Tottori S., Terada T. and Sekijima K : Experimental study on tensile strength of
FRP bent bar, Proc. of JCI, Vol.1 1, No.1, pp.789-794, 1989(in Japanese).
[10]Maruyama T., Horma M. and Okamura H. .. Some experiments on diagonal tension
characteristics of various flber reinforced plastic rods, Proc. of JCI, Vo1.ll, No.1, pp.771-776,
1989(in Japanese).
[1 1]Wakui H. and Tottori S. : Reinforcing effect of spiral FRP rods as shear reinforcement, Proc.
of JCI, Voi.12, No.1, pp.1 141-1146, 1990(in Japanese).

[12]Terada T., Tottori S., Wakui H. and Miyata S. : Study on behavior of shear failure of concrete
beams reinforced with FRP bars, Proc. of JCI, Vo1.10, No.3, pp.541-546, 1988(in Japanese).
[13]Nakamura H. and Higai T. : Evaluation of shear strength of RC beam section based on
extended modified compression field theory, Concrete Library of JSCE, No.25, pp.93-105, 1995.
[14]Idemitsu T., Yamasaki T., Harada T. and Yoshioka T. : An experimental study on the method
oftension test for FN tendons, Proc. ofJCI, Vo1.13,No.1, pp.795-800, 1991(in Japanese).
[15]Niwa J., Yamada K., Yokozawa K. and Okamura H. : Revaluation of the equation for shear
strength of concrete beams without web reinforcement, Proc. of JSCE, No.372N-5, pp.167-176,
1986(in Japanese).
[16]Yokoi K., Shima H., Mizuguchi H. and mrohata H. : Stirrup stress of concrete beams using
FN rods as main reinforcement, Proc. of JCI, Vol.13, No.2, pp.771-776, 1991(in Japanese).
[17]Sato Y., Ueda T. and Kakuta Y. : Qualitative evaluation of shear resisting behavior of
concrete beams reinforced with FRP rods by finite element analysis, Concrete Library of JSCE,
No.24,

pp.193-209,

1994.

-123-