The document describes an experimental study on the behavior of fiber reinforced polymer (FRP)-confined normal and high-strength concrete cylinders under cyclic axial compression loading. 24 concrete cylinders wrapped with aramid or carbon FRP jackets were tested monotonically or cyclically. Test results showed that FRP-confined concrete failed by FRP rupture. Higher FRP layers led to higher confined concrete strength but strength enhancement was dependent on a threshold confinement. Existing stress-strain models were found to predict the behavior of normal strength concrete better than high-strength concrete under cyclic loading.

1. BEHAVIOUR OF FRP-CONFINED NORMAL
AND HIGH STRENGTH CONCRETE UNDER
CYCLIC AXIAL COMPRESSION
BY TOGAY OZBAKKALOGLU1AND EMRE ALEIN2
1SENIOR LECTURER,SCHOOL OF CIVIL
ENGINEERING,UNIVERSITY OF ADELAIDE
2RESEARCH FELLOW, DEPT OF CIVIL ENGINEERING,
SELUCK UNIV.,KONYA , TURKEY
JOURNAL OF COMPOSITES FOR CONSTRUCTION,ASCE,JULY
2012
BETHU PRAVEEN KUMAR(12CE65R11)
STRUCTURAL ENGINEERING
DEPARTMENT OF CIVIL ENGINEERING
IIT KHARAGHPUR

2. OVERVIEW

INTRODUCTION

EXPERIMENTAL PROGRAMME

TEST SETUP

TEST RESULTS

DISCUSSIONS

REFERENCES

3. INTRODUCTION

We know that concrete is strong in compression and
fails due to tension.

FRP(fiber reinforced polymers) are used as
confinement for concrete

Application of FRP as confining materials is used for
retrofitting of existing columns and new column as EQ
resistant

4. INTRODUCTION

Monotonic stress-strain behavior of FRP
confined has been studied past two decades

HSC members are known to exhibit brittle
behavior so their use in seismically active
regions is restricted

However by providing sufficient confinement we
can increase their ductility

5. EXPERIMENTAL PROGRAMME

24 FRP cylinders of dia 152.5mm and height
305mm are used

TEST PARAMETERS : compressive strength ,type of
FRP ,amount of confinement ,type of loading

Mix consists of crushed blue stone of max size
10mm as coarse aggregate and 8% of binder by
weight is replaced by silica

6. EXPERIMENTAL PROGRAMME

28 days strength of NSC was found to be 39MPa
and HSC is 103 MPa the stress(fIc0) and strain(єco)
at failure was recorded

AFRP was used as confinement for both NSC and
HSC where as CFRP are used for HSC

FRP jackets of 22 specimens were formed by
manually wrapping FRP sheets around concrete
and other 2 are confined by formerly
manufactured AFRP

7. TEST SETUP

Axial deformations of the specimens were
measured by LVDT as shown in fig

Transverse shear strain were measured from 3
unidirectional strain gauges of gauge length
20mm that were bounded on FRP jackets

For elastic loading the loading rate is 3KN/s and
after softening displacement control is 0.01mm/s

9. TEST SETUP

For 12 of the specimens the load was
monotonically increasing for other 12 specimens
cyclic compression involving unloading and loading
at 0.15% axial strain

Specimen are labeled as follows H-A-4L-M1 where
H is HSC and A is aramid polymer ,4 layer of
confinement , under Monotonic loading the final
number 1 is to denoted a difference between 2
similar specimens

10. TEST RESULTS

Failure Mode: All the specimens fail due to
rupture of FRP jacket

Concrete shear cones were formed in NSC
specimens due to gradual crushing of concrete

In HSC specimens the failure is highly localized
around a major shear crack

11. FAILURE PATTERNS OBSERVED

12. TEST RESULTS

Axial σ-Є behavior :The σ-Є curves of
monotonically loaded specimens exhibit an
ascending first branch that is followed by an
ascending or almost flat second branch

Where as HSC specimens experiences a
sudden drop in strength starting right at the
transition point this is due to initial softening

13. TEST RESULTS OF FRP CONFINED CONCRETE
CYLINDERS
Specimen
f’cc(MPa)
Єcu(%)
Єh,rup(%)
flu/f’co
ЄCU/єCO
flua/f’co
K1(avg)
K2(avg)
K3(avg)
N-A-2L-M1
69.2
2.32
1.71
0.40
10.9
0.28
2.82
13.8
0.68
N-A-2L-M2
67.1
2.30
1.56
0.40
N-A-3L-M1
85.0
2.86
1.66
0.61
2.86
11.2
N-A-3L-M2
87.6
3.11
1.84
0.61
H-A-4L-M1
122.3
1.45
1.18
0.32
1.26
8.4
H-A-4L-M2
118.7
1.29
1.29
0.31
H-A-6L-M1
154.7
1.70
1.10
0.45
2.33
9.7
H-A-6L-M2
153.2
1.70
1.07
0.45
H-C-4L-M1
98.9
0.93
0.89
0.23
H-C-4L-M2
103.3
0.96
0.81
0.21
H-C-6L-M1
122.3
1.13
0.94
0.31
H-C-6L-M2
124.4
1.16
0.78
0.37
N-A-2L-C1
64.3
2.25
1.50
0.42
N-A-2L-C2
64.3
2.25
1.56
0.40
N-A-3L-C1
97.4
4.04
1.76
0.61
N-A-3L-C2
104.5
4.43
2.02
0.61
H-A-4L-C1
136.4
1.82
1.24
0.32
H-A-4L-C2
125.4
1.63
1.10
0.31
H-A-6L-C1
157.2
1.87
1.16
0.46
H-A-6L-C2
170.9
2.13
1.45
0.45
H-C-4L-C1
102.3
1.07
0.69
0.23
H-C-4L-C2
96.0
1.06
0.81
0.23
H-C-6L-C1
123.7
1.14
0.64
0.31
H-C-6L-C1
129.9
1.16
0.81
0.33
0.25
14.1
0.40
0.45
4.0
0.15
0.46
0.16
4.9
0.20
0.19
2.7
0.13
5.5
0.55
0.11
3.4
0.19
1.17
5.8
2.67
14.6
3.46
14.9
2.0
13.1
2.39
9.5
0.18
10.7
0.25
0.68
0.25
20.0
0.43
0.49
5.1
0.16
0.50
0.14
5.8
0.21
0.26
3.2
0.10
8.9
0.12
3.3
00.13
0.17
1.21
7.6
0.48

14. TEST RESULTS

At this point hoops strains recorded on FRP
jacket increases rapidly but the confinement
pressures generated by FRP are sufficient to
confine concrete

17. DISCUSSIONS

Envelope curve of concrete represents the
upper boundary of the response under cyclic
axial compression

The envelope curve is drawn by connecting the
initial unloading points on the σ-Є curve of
cyclically loaded specimen

18. DISCUSSIONS

Unloading Reloading and Plastic Strain :To define
complete σ-Є of cyclically loaded specimen
,unloading and reloading paths are required in
addition to envelope curve

Unloading path intersects the strain axis at a value
referred to as residual plastic strain

The relationship between єpl and єun,enve is an
important aspect of cyclic loading

19. DISCUSSIONS

Lame et al(2006) demonstrated that the
relationship between єpl and єun,enve is linear for
CFRP confined NSC cylinders for єcu>0.0035

Trend lines as shown are drawn for every
specimen and the following observations are
made

Trend lines of specimen with same concrete
strength and confinement material coincide

22. DISCUSSIONS

Trend lines of AFRP confined cylinders and CFRP
confined HSC shows that they does not
significantly depend on type of FRP confinement

Comparison of trend lines of HSC and NSC
indicates that it does not significantly depend on
unconfined compressive strength

Which was against to lames and teng’s model
which suggests reduction in plastic strain with
increase in unconfined strength

23. DISCUSSIONS

There is a relationship between єpl and єun,enve given by shao et al
σ-Є model given by
 pl   un ,enve 
 un ,enve
Esec
 1 for 0 
Ec
 0.004
Esec
 un ,enve
 un ,enve
f 'co
 0.34 for
f 'c 0
1
 1.44 for 1 
 un,enve
f'co
 cr
f 'co
 2.5
 2 .5

Comparison of experimental and that obtained by above equations
are plotted in the fig below

The fig shows that the єpl of the present study is over estimated by
shao et al model is due to Esec

25. DISCUSSIONS

At any stage of loading history beyond initial
elastic portion with increasing deformation the
unloading stiffness of NSC and HSC specimens
decreases much more significantly than predicted
by shao’s models

The observations and discussions presented
suggests the variation of unloading stiffness can
be accurately predicted by using єun,enev/єcu while
giving due consideration to unconfined concrete
strength

26. DISCUSSIONS

The ultimate condition of FRP is referred to as the
ultimate strength and strain of concrete recorded
just before failure

The nominal confinement ratio flu/fIco is calculated
from equations assuming uniform confinement
distribution
2E f t f  fu
f lu
f 'co


Df 'co
The value obtained above is a theoretical value
and does not represent actual confining pressure
developed in FRP at failure

27. DISCUSSIONS

Which is because the ultimate hoop strain
reached in FRP is much smaller than ultimate
tensile strain in fiber which necessitates a
strain reduction factor kξ for finding actual
confining pressure at failure

28. DISCUSSIONS

Effect of loading pattern:kξ values does not
depend on the load cycles so is the hoops
strain єh,rupt

But lam et al proposed an observed increase in
єh,rupt with increase in loading/unloading cycles

29. DISCUSSIONS

Effect of unconfined concrete strength: By
comparing samples of same fIlu/fIco ,indicate
that the ultimate strength is lower for HSC than
NSC under cyclic loading

Kξ values of HSC are consistently lower than
that of NSC which suggests that it is strength
dependent

30. DISCUSSIONS

Stress and strain enhancement coefficients K1
and k2 are calculated by using lame and terg’s
expression shown below
f 'cc
 1  K1
f 'co
 cu
 co

f lu ,a
f 'co
f lu ,a   h ,rup 

 1.75  K 2
  
f 'co  co 

0.45
K1 and k2 are lower for HSC when compared to
NSC

31. DISCUSSIONS

By comparison of H-A-4L and H-C-6L layer with
same concrete strength and same confining

pressure shows that  for AFRP confine is
more then CFRP confinement
cu
co

Comparison of Kξ for these 2 specimens
suggest that it does not depend on type of FRP

32. DISCUSSIONS

It is well understood that by increasing confinement
both strength and strain enhancement ratios increases

However closer inspection of relation between fIcc/fIco
and flu,a/fIco suggests that it is not linear for HSC
specimens

This can be explained by tread of σ-є curve shown, there
is a descending second branch of H-C-4L specimen
suggests that confinement provided was not sufficient
to provide enhancement

33. STRESS-STRAIN CURVE OF H-C-4L-C1 AND H-C6L-C1

34. DISCUSSIONS

When CFRP layers are increased to 6L the second
branch has an ascending or almost flat trend

From this we can say the strength enhancement is
observed when concrete is confined by a certain
minimum confinement which is known as
Threshold Confinement

Threshold Confinement is sensitive to unconfined
concrete

35. DISCUSSIONS

The better prediction of ultimate strength of
FRP confined HSC has to be developed which
can accurately predict Threshold confinement
as a function of unconfined compressive
strength

36. COMPARISON WITH EXSISITNG STRESS-STRAIN
MODELS

Both envelope curves of shao’s and lam and
teng’s was in accordance with the results obtained

Fig illustrates that lam and teng’s model is highly
accurate in predicting both loading and unloading
curves of FRP confined NSC

As we have already discussed shao’s model over
predicts so it deviates from unloading and
reloading curves obtained

38. COMPARISON WITH EXSISITNG STRESS-STRAIN
MODELS

Further more lam and teng’s model accurately
predicting the reloading curve which is linear initial
and becomes parabolic as it moves to envelope
stress

Where as shao’s model is fully linear which is not
happening in practical

Application of both models to FRP confined HSC
leads to large errors in estimation σ-є curve which
can be seen from above mentioned graphs

39. COMPARISON WITH EXSISITNG STRESS-STRAIN
MODELS

In case of lam and teng’s the deviation is due to
inaccuracies in predicting plastic strain due to
limited load cycle data

But shao’s model performance does not degrade
for HSC when compared with lams but even does
it improve

Finally it is not possible to accurately predict σ-є of
FRP confined HSC under cyclic compressive
loading

40. CONCLUSIONS

The envelope curve of cyclically loaded FRP
confined concrete closely follows the σ-є curve
of same concrete under monotonic loading

The residual plastic strain єpl of FRP confined is
linearly related to unloading strain and this
relation does not depend on amount of
confinement, type of FRP used ,unconfined
strength of concrete

41. CONCLUSIONS

For a given confinement ratio fIcc/fIco both strength
and strain enhancement ratio decreases with
increase in unconfined concrete strength

Concrete experiencing similar level of confinement
when confined with AFRP and CFRP jackets
provides same confinement pressures but єcu of
concrete with AFRP is significantly high

42. REFERENCES
Lam, L., and Teng, J. G. (2009). “Stress-strain
model for FRP-confined concrete under cyclic
axial compression.” Eng. Struct., 31(2), 308–
321
 Shao, Y., Zhu, Z., and Mirmiran, A. (2006).
“Cyclic modeling of FRP confined concrete with
improved ductility.” Cem. Concr.
Compos.,28(10), 959–968..


43. THANK YOU