University of Wollongong

Research Online

Faculty of Engineering and Information Faculty of Engineering and Information

Sciences - Papers: Part A Sciences

1-1-2012

Axial and flexural performance of square RC columns wrapped with CFRP

under eccentric loading
Muhammad N. S Hadi
University of Wollongong, [email protected]

Ida Bagus Rai Widiarsa

University of Wollongong, [email protected]

Follow this and additional works at: https://ro.uow.edu.au/eispapers

Research Online is the open access institutional repository for the University of Wollongong. For further information
contact the UOW Library: [email protected]
Axial and flexural performance of square RC columns wrapped with CFRP under
eccentric loading

Abstract
The majority of studies on fiber-reinforced-polymer (FRP) strengthened concrete columns deal with
columns of a circular cross section. However, most concrete columns in the field have square or
rectangular cross sections and resist eccentric loads. This paper presents the results of an experimental
study on the performance of carbon-fiber-reinforced-polymer (CFRP) wrapped square reinforced concrete
(RC) columns under eccentric loading. The influence of the number of CFRP layers, the magnitude of
eccentricity, and the presence of vertical CFRP straps were investigated by testing 16 specimens. The
specimens had the dimensions 200 × 200 × 800 mm and round corners with a radius of 34 mm. Twelve
specimens were tested as columns and four specimens as beams. The results of this study showed that
CFRP wrapping enhanced the load-carrying capacity and ductility of the columns under eccentric loading.
Furthermore, the application of the vertical CFRP straps significantly improved the performance of the
columns with large eccentricity

Keywords
cfrp, under, wrapped, columns, rc, square, performance, eccentric, flexural, axial, loading

Publication Details
Hadi, M. N. S. & Widiarsa, I. (2012). Axial and flexural performance of square RC columns wrapped with
CFRP under eccentric loading. Journal of Composites for Construction, 16 (6), 640-649.

This journal article is available at Research Online: https://ro.uow.edu.au/eispapers/8

 Axial and Flexural Performance of Square RC Columns
Wrapped with CFRP under Eccentric Loading
Muhammad N. S. Hadi, M.ASCE1; and Ida Bagus Rai Widiarsa2
Downloaded from ascelibrary.org by UNIVERSITY OF WOLLONGONG on 12/02/12. Copyright ASCE. For personal use only; all rights reserved.

Abstract: The majority of studies on fiber-reinforced-polymer (FRP) strengthened concrete columns deal with columns of a circular cross
section. However, most concrete columns in the field have square or rectangular cross sections and resist eccentric loads. This paper presents
the results of an experimental study on the performance of carbon-fiber-reinforced-polymer (CFRP) wrapped square reinforced concrete (RC)
columns under eccentric loading. The influence of the number of CFRP layers, the magnitude of eccentricity, and the presence of vertical
CFRP straps were investigated by testing 16 specimens. The specimens had the dimensions 200 × 200 × 800 mm and round corners with a
radius of 34 mm. Twelve specimens were tested as columns and four specimens as beams. The results of this study showed that CFRP
wrapping enhanced the load-carrying capacity and ductility of the columns under eccentric loading. Furthermore, the application of
the vertical CFRP straps significantly improved the performance of the columns with large eccentricity. DOI: 10.1061/(ASCE)CC
.1943-5614.0000301. © 2012 American Society of Civil Engineers.
CE Database subject headings: Fiber reinforced polymer; Reinforced concrete; Concrete columns; Eccentric loads; Ductility.
Author keywords: CFRP; Square reinforced concrete columns; Eccentric loading; Ductility.

Introduction investigations of columns under eccentric loads. Consideration of

applying eccentric loads was undertaken due to the reality that most
High-strength concrete (HSC) has been used widely in civil columns in the field are loaded under a combination of axial com-
construction projects due to its high compressive strength and elas- pression load and bending moment. Bending moment always exists
tic modulus. The use of HSC instead of normal-strength concrete in a column, even though in columns nominally carrying only axial
(NSC) would reduce significantly member size and material con- compression load, bending moment is introduced by unintentional
sumption. HSC members behave differently than those made of load eccentricities and by out-of-straightness of the constructed col-
NSC; HSC is more brittle than NSC. The failure of a HSC member umn (Warner et al. 2007).
is more explosive and sudden. A review of the literature shows that The effect of applying eccentric loads on FRP-strengthened
the majority of studies on fiber-reinforced-polymer (FRP) confined concrete columns has been studied by a few researchers. Other than
concrete columns investigated columns with NSC (e.g., Teng et al. the magnitude of eccentricities, the shape of column cross section
2002). Only a few studies investigated FRP confined columns with is an important factor that influences the confining behavior of
HSC (Li and Hadi 2003; Hadi 2007b; Eid et al. 2009). FRP-strengthened concrete columns. Some researchers have inves-
The use of FRP composites for strengthening reinforced tigated FRP-strengthened concrete columns under eccentric loads,
concrete members has been investigated experimentally by many focusing mostly on columns having a circular cross section (Li and
researchers, and FRP has been applied in construction. Strengthen- Hadi 2003; Hadi 2006a, b, 2007a, b; Bisby and Ranger 2010).
ing reinforced concrete columns using FRP composites is preferred Although most concrete columns in the field have square or rec-
to using other materials like steel due to the high strength-to-weight tangular cross sections rather than circular ones and may also resist
ratio and high corrosion resistance of FRP composites. Several in- eccentric loads, only a few investigations have studied concrete col-
vestigations on FRP-strengthened concrete columns have been umns with square or rectangular cross sections (Parvin and Wang
undertaken over the years. Most of these studies investigated the 2001; Maaddawy 2009). Both of these studies tested small-scale
behavior of columns under concentric loads (Kumutha et al. square concrete columns constructed with normal strength con-
2007; Wang and Wu 2008; Lam and Teng 2002; Teng and Lam
crete. Columns tested by Parvin and Wang (2001) had no internal
2004; Rousakis et al. 2007), whereas only a few studies presented
reinforcement and were wrapped with no layer, one layer, or two
layers of CFRP. The columns were tested under concentric and
1
Associate Professor, School of Civil, Mining and Environmental eccentric loads. Columns tested by Maaddawy (2009) had no
Engineering, Univ. of Wollongong, Wollongong NSW 2522, Australia round corners but had end corbels. The columns were unwrapped,
(corresponding author). E-mail: [email protected] fully wrapped, and partially wrapped with CFRP and were tested
2
Lecturer, Dept. of Civil Engineering, Univ. of Udayana, Bali 80361, eccentrically.
Indonesia; presently, Ph.D. Scholar at the School of Civil, Mining and En- Consequently, research on FRP-strengthened square or rectan-
vironmental Engineering, Univ. of Wollongong, Wollongong NSW 2522, gular HSC columns under eccentric loads must still be undertaken
Australia. E-mail: [email protected] to understand their behavior and performance. This paper is
Note. This manuscript was submitted on November 20, 2011; approved
directed toward this purpose. Parameters investigated in this study
on April 24, 2012; published online on April 26, 2012. Discussion period
open until May 1, 2013; separate discussions must be submitted for indi- include the magnitude of eccentricity and the number of FRP
vidual papers. This paper is part of the Journal of Composites for Con- layers. Three different eccentricities—0, 25, and 50 mm—were
struction, Vol. 16, No. 6, December 1, 2012. © ASCE, ISSN 1090-0268/ investigated. In relation to the number of FRP layers, unwrapped
2012/6-640-649/$25.00. columns, columns wrapped with one layer, and columns wrapped

640 / JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / NOVEMBER/DECEMBER 2012

J. Compos. Constr. 2012.16:640-649.

 with three layers of CFRP were tested. Due to the significant The general properties of the specimens are shown in Table 1.
effect of longitudinal fiber in eccentric loading (Chaallal and The specimens were divided into four groups: unwrapped, wrapped
Shahawy 2000; Hadi 2007a), one layer of CFRP straps was applied with one layer of CFRP, wrapped with three layers of CFRP, and
longitudinally in combination with two layers of CFRP wrapped wrapped with one layer of vertical (along specimen axis) CFRP
circumferentially. straps and two layers of horizontal (transverse) CFRP. Each group
consisted of four specimens; one specimen was tested concentri-
cally, one was tested under a 25 mm eccentric load, one was tested
Experimental Program under 50 mm eccentric load, and the final one was tested under
flexural load. The label used for the specimens was composed of
The experimental program consisted in testing a total of 16 RC a combination of numbers and letters. The numbers 0, 1, 2, and 3
specimens, of which 12 were tested under compression loading denote the number of CFRP layers. The letters H and V denote the
and four under flexural loading. All specimens were tested at orientation of the CFRP: horizontal and vertical, respectively. The
Downloaded from ascelibrary.org by UNIVERSITY OF WOLLONGONG on 12/02/12. Copyright ASCE. For personal use only; all rights reserved.

the laboratories of the School of Civil, Mining and Environmental letters C and B denote column and beam, respectively. The last
Engineering at the University of Wollongong, Australia. numbers, 0, 25, and 50, denote the eccentricity under which the
specimens were tested. For example, Specimen 1V2HC50 had
one layer of vertical CFRP straps and two layers of transverse
Design of Specimens CFRP and tested as a column with a 50-mm eccentricity.
Short RC specimens were cast and tested in this study. Each speci-
men had a square cross section with a side dimension of 200 mm Preliminary Testing
and a height of 800 mm. Short columns were designed to avoid the
The concrete of the specimens was ready-mix HSC supplied by a
formation of secondary moments due to the slenderness effect.
local concrete supplier and had an average 28-day compressive
Moreover, the dimensions were chosen to be adaptable with the
strength of 79.5 MPa. The tensile strength tests were conducted
condition and capacity of the available testing machine in the
on the reinforcing steel bars to determine their tensile strength.
laboratory. The four corners of the specimens were rounded for
Average tensile yield strengths of 564 and 516 MPa were obtained
the purpose of avoiding premature failure and to prepare sufficient
for N12 and R8 reinforcing bars, respectively. For wrapping the
effect of confinement of the columns (Wang and Wu 2008;
specimens, in this study carbon fiber (CFRP) was used that was
Mirmiran et al. 1998). The concrete cover was 20 mm on each side
available in the form of rolls, 100 m in length and 75 mm in width.
of the specimen and at top and bottom as well. A corner radius of
The fiber had a nominal thickness of 0.45 mm. Tensile tests were
34 mm was applied to the specimens. The longitudinal and trans-
carried out on three CFRP test coupons of one layer, two layers, and
versal reinforcement of the specimens was designed in accordance
three layers of CFRP. In conducting the tests, specimen loading
with AS3600-2009 (Standards Australia 2009). All specimens were
and strain measurement were performed according to ASTM D
designed inadequately as their internal steel reinforcement ratio was
3039-08 (ASTM 2008). The tensile strength, tensile strain, and
around the lowest ratio of that specified by the standard. The pur-
elastic modulus of the CFRP were found to be 1399 MPa, 1.86%,
pose of this design was to produce the condition of an old column
and 75.4 GPa, respectively. These values were the average values
that has deteriorated and needs to be strengthened. In addition, the
calculated using the actual width of the test coupons and the nomi-
shear reinforcement provided was also at the minimum shear
nal thickness of the fiber. The nominal thickness for the two and
reinforcement required by the design standard (Asv: min ). Therefore,
three layers of CFRP was assumed to be equal to 0.90 mm and
the specimens had four N12 (12 mm diameter deformed bars) as
1.35 mm, respectively.
longitudinal steel reinforcement and R8 (8 mm diameter plain bars)
spaced at 100 mm as transverse steel reinforcement (ties). The R8
ties spaced at 50 mm were applied at both ends of the specimens Specimen Preparation
to prevent premature failure at the locations. Fig. 1 shows the The specimens were cast in a wooden formwork from one batch of
details of the reinforcement. concrete. A slump value of 165 mm was measured, which means

Table 1. Configuration of Test Specimens

Side Number Test
Test width Height Internal of FRP eccentricity
specimen (mm) (mm) reinforcement layers (mm)
0C0 200 800 4N12 and None 0
0C25 R8 100 mm 25
0C50 50
0B Bending
1HC0 200 800 4N12 and 1 layer 0
1HC25 R8 100 mm 25
1HC50 50
1HB Bending
3HC0 200 800 4N12 and 3 layers 0
3HC25 R8 100 mm 25
3HC50 50
3HB Bending
1V2HC0 200 800 4N12 and 2 layers 0
1V2HC25 R8 100 mm with 25
1V2HC50 1-layer 50
Fig. 1. Details of specimen reinforcement 1V2HB straps Bending

JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / NOVEMBER/DECEMBER 2012 / 641

J. Compos. Constr. 2012.16:640-649.

 the concrete had proper workability. The concrete was poured into a
formwork in three stages; in each stage the concrete was compacted
adequately using a vibrator to ensure an even dispersion of con-
crete. After casting, all the specimens were placed under wet hes-
sian rugs and covered with plastic sheets to maintain their moisture
conditions. The specimens and hessian rugs were watered twice a
day until the formwork was removed at 14 days after casting. The
curing process lasted 28 days.
Surface preparation was done by cleaning and grinding the sur-
face of the specimens. Twelve specimens were wrapped with CFRP
using a wet layup technique. A mixture of epoxy resin and hardener (a)
at a 5∶1 ratio was used as an adhesive. The surface was coated first
Downloaded from ascelibrary.org by UNIVERSITY OF WOLLONGONG on 12/02/12. Copyright ASCE. For personal use only; all rights reserved.

with epoxy resin then the first CFRP layer in predefined orienta-
tion. The layer was saturated with resin before the subsequent layer
was applied to ensure a perfect bond between the layers. Similarly,
the final layer was also saturated with epoxy resin. The column
was wrapped strap by strap in either the transverse or longitudinal
direction, with no overlap between straps. An overlap of 100 mm
was applied only at the end of the final layer of each strap in the
transverse direction. The wrapped specimens were then placed in
room temperature for a minimum of 14 days to allow curing of the
epoxy resin.

Specimen Testing (b)

Twelve specimens were tested as columns under axial compression Fig. 2. Developed eccentric loading mechanism: (a) plan view;
load, and four specimens were tested as beams under flexural load. (b) Section A-A
The Denison 5000 kN compression testing machine was used to
test all the specimens until failure. The column ends were leveled
first for compression testing to obtain a uniform distributed load
applied to the entire face. For this purpose, both ends of the col-
umns were capped with high-strength plaster. To apply an eccentric
load on the column, a loading mechanism (Fig. 2) was designed
and a new set of loading heads made with high-strength steel plate
(Fig. 3) was manufactured. The loading heads were attached at both
ends of the column. The loading heads consisted of two parts: a
25 mm thick steel plate, called a bottom plate, that had a ball joint,
and a square 50 mm thick steel plate, called an adaptor plate. The
bottom plate transferred the load generated by the machine to the
adaptor plate through the ball joint, which had a designed eccen-
tricity with the column. Then the adaptor plate exerted the load to
the column. For concentric loading, only adaptor plates were used
to apply the load.
Two different monitoring systems were used to measure the Fig. 3. Loading head: (a) adaptor plate; (b) bottom plate with ball joint
displacement of the columns. For concentric loading, one LVDT
was connected directly to the testing machine to measure the axial
displacement of the column during the test. Data read from this
LVDT were recorded at the same time as load data were recorded
by the testing machine. A second LVDT, a laser LVDT, was also
used in addition to the first one for eccentric load to measure the
lateral deflection (δ) of the column. The second LVDT was placed
horizontally near the mid-height of the column. When the specimen
and the instrumentation were placed in position and initial calibra-
tion was done, the compression testing then started. The column
was tested under displacement control with a loading rate of
0.5 mm=min, and the end point position was set at 50 mm. A photo
of a typical compression testing is shown in Fig. 4.
As mentioned previously, four specimens were tested as beams
under flexural loading with a span of 700 mm. Pure bending was
applied to the specimens by means of a four-point loading to
determine the flexural capacity of the specimens without axial load.
Two rigs, top and bottom, were used to exert the load generated by
the Denison compression testing machine to the beam specimens.
Fig. 4. Typical compression testing setup
The bottom rig was placed diagonally on the bottom plate (loading

642 / JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / NOVEMBER/DECEMBER 2012

J. Compos. Constr. 2012.16:640-649.

Downloaded from ascelibrary.org by UNIVERSITY OF WOLLONGONG on 12/02/12. Copyright ASCE. For personal use only; all rights reserved.

Fig. 5. Typical flexural testing setup

cell) of the machine, and the beam specimen was then placed on the
rig. The typical setup of flexural testing of the beams is shown in Fig. 6. Typical failure of unwrapped columns
Fig. 5. The testing machine recorded the applied load simultane-
ously during the test. A laser LVDT that was placed vertically
underneath the bottom rig shot the beam through the provided slot
to measure the midspan deflection of the beam during the test. The
beam was tested under displacement control, the end point position
was set at 50 mm, and the loading rate was set at 0.3 mm=min.

Experimental Results and Analysis

Column Failure Modes

All column specimens were tested to failure. The failure of columns
without CFRP wrapping was generally marked by sudden loss of
concrete cover, followed by the buckling outward of the longitu-
dinal reinforcement and the rupture of the ties. The failure occurred
near the mid-height of the columns. Meanwhile, the failure of the
columns wrapped with CFRP was initiated by the appearance of
FRP ripples at some places on the column sides, followed by a
snapping sound when the load approached the maximum load.
A strap-by-strap rupture of the FRP was subsequently observed.
The first rupture of CFRP occurred at the maximum load, resulting
in a decrease in the load. The columns continued supporting the
load until the rupture of the other CFRP while experiencing a large
displacement. The rupture of the CFRP and debonding between
the CFRP layer and the concrete in the column test revealed the
concrete expansion at the place where the CFRP failed. The CFRP Fig. 7. Typical failure of wrapped columns
rupture occurred at the corner of the column. Buckling of longitu-
dinal reinforcement and crushing of concrete in compression were
also observed. Typical failure of unwrapped and wrapped columns the load-axial displacement curve up to the ultimate displacement
under concentric and eccentric loading is shown in Figs. 6 and 7, to the area under the curve up to the yield load. The yield load
respectively. was assumed to be the load corresponding to an approximation
of the limit of the elastic behavior. The ultimate displacement
was assumed to be the displacement at 85% of the maximum load
Behavior of Columns (GangaRao et al. 2007).
The ultimate load and corresponding axial and lateral displace- Fig. 8 shows the load-displacement curves for concentric
ments were recorded during the compression testing; the results columns. The columns had a similar behavior before reaching
are summarized in Table 2. Two methods were used to calculate the maximum load. It is clear that the biggest maximum load
the ductility of the specimens. The first method used the ratio and maximum axial displacement among the four columns was
of axial displacement at ultimate load to the axial displacement achieved by wrapping the column with three layers of CFRP.
at yield load. The second method used the ratio of area under The maximum load did not increase significantly with the CFRP

JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / NOVEMBER/DECEMBER 2012 / 643

J. Compos. Constr. 2012.16:640-649.

 Table 2. Column Test Results
Displacement at
ultimate load (mm) Axial displacement (mm) Ductility
Test Ultimate
specimen load (kN) Axial Lateral At yield At 85% ultimate load Maximum Method 1a Method 2a
0C0 3,248 4.58 — 3.21 4.71 5.20 1.47 2.09
1HC0 3,279 4.53 — 3.20 4.75 13.88 1.48 2.18
1V2HC0 3,522 5.00 — 3.57 8.57 19.57 2.40 4.39
3HC0 3,585 5.29 — 3.63 13.47 18.51 3.71 7.16
0C25 1,950 3.91 1.87 2.93 4.12 6.25 1.41 1.87
1HC25 2,076 4.45 2.25 3.11 5.07 9.90 1.63 2.68
1V2HC25 2,296 4.76 2.44 3.19 8.30 15.94 2.60 5.17
Downloaded from ascelibrary.org by UNIVERSITY OF WOLLONGONG on 12/02/12. Copyright ASCE. For personal use only; all rights reserved.

3HC25 2,269 4.48 2.11 3.18 11.17 14.95 3.52 7.45

0C50 1,336 3.86 2.65 2.93 4.01 4.66 1.37 1.90
1HC50 1,433 4.05 2.32 3.02 4.98 13.20 1.65 2.84
1V2HC50 1,533 3.99 2.52 2.88 9.22 14.55 3.20 7.06
3HC50 1,534 3.99 3.19 2.89 6.64 13.09 2.30 4.66
Refer to the section “Behavior of Columns” for a definition of the methods.
a

4000 2400
3500
2000
3000
Loaad (kN)

N)
2 00
2500 1600

Loaad (kN
2000
1500
1200
1000 800
500
0 400
0 5 10 15 20
A i l di
Axial displacement
l t ((mm)) 0
-35 -30 -25 -20 -15 -10 -5 0 5 10 15 20
0C0 1HC0 1V2HC0 3HC0 Lateral displ. (mm) Axial displ. (mm)
Fig. 8. Load-displacement curves for columns tested under concentric 0C25 1HC25 1V2HC25 3HC25
loading Fig. 9. Load-displacement curves for columns tested under 25-mm
eccentricity

wrapping. However, wrapping columns with CFRP enhanced the

cracks in the tension face near the mid-height of the columns.
performance of the columns by increasing their displacement at
FRP rupture occurred after the maximum load was reached, fol-
failure, meaning more ductility. Increases in the maximum load
lowed by an increase in displacement. Column 1HC50 had
of 1, 8.4, and 10.4% relative to the unwrapped column were
7.3% increase in maximum load and a better performance than
achieved for Columns 1HC0, 1V2HC0, and 3HC0, respectively.
the unwrapped column. A similar increase in maximum load of
This insignificant increase in strength was obtained due to the
14.8% was achieved for Columns 1V2HC50 and 3HC50. Column
use of HSC, where in nature HSC is brittle. When approaching
1V2HC50 had a higher ductility than Column 3HC50 (Table 2).
the ultimate load, only a small number of microcracks developed
Tests of the 12 column specimens revealed an increase in the
in the concrete and the failure occurred suddenly with the appear-
maximum load of the columns resulting from wrapping columns
ance of a big crack.
The axial and lateral displacements versus the applied load with CFRP. An important advantage was achieved: all the wrapped
curves for columns tested under 25 mm eccentric load are shown columns showed a better performance than the unwrapped col-
in Fig. 9. All columns failed in compression. FRP rupture occurred umns, as indicated by their ductility. From the load-displacement
in all wrapped columns before failure. Column 1HC25 had 6.5% 1600
higher maximum load than the unwrapped column. Increases in
maximum load of 17.8 and 16.4% were achieved for Columns
1200
1V2HC25 and 3HC25, respectively. However, Column 3HC25
Loadd (kN))

showed a better performance than Column 1V2HC25 by having

a higher ductility (Table 2). Theoretically, Column 3HC25 should 800
have had a higher maximum load than Column 1V2HC25. The
presence of vertical straps that were bonded first to the concrete 400
and then wrapped with two layers of CFRP also contributed to
the confinement effect. Similarly, in concentric columns, Column 0
1HC0 had only a 1% higher maximum load than Column 0C0, -45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20
whereas Column 3HC0 had a 2% higher maximum load than Laterall displ.
di l (mm)
( ) A i l di
Axial l ((mm))
displ.
Column 1V2HC0. 0C50 1HC50 1V2HC50 3HC50
Fig. 10 shows the axial and lateral displacements versus the
Fig. 10. Load-displacement curves for columns tested under 50-mm
applied load curves for columns tested under 50 mm eccentric load.
eccentricity
Failure in compression was also observed in all columns with

644 / JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / NOVEMBER/DECEMBER 2012

J. Compos. Constr. 2012.16:640-649.

 3500 3500
0C0 1HC0
3000 3000
0C25 1HC25
2500 2500
Load (kN)

Load (kN)
0C50 1HC50
2000 2000
1500 1500
1000 1000
500 500
0 0
0 2 4 6 8 0 5 10 15
Downloaded from ascelibrary.org by UNIVERSITY OF WOLLONGONG on 12/02/12. Copyright ASCE. For personal use only; all rights reserved.

(a) Axial displacement (mm) (b) Axial displacement (mm)

4000 4000
3500 3HC0 3500 1V2HC0
3000 3HC25 3000 1V2HC25
2500 3HC50 2500 1V2HC50
Load (kN)

Load (kN)
2000 2000
1500 1500
1000 1000
500 500
0 0
0 5 10 15 20 0 5 10 15 20
(c) Axial displacement (mm) (d) Axial displacement (mm)

Fig. 11. Load-displacement curves of columns with different eccentricities

curves of the columns, it can be seen that the displacement 600

increased with increasing load until the maximum load was ap-
500
proached. Then, for unwrapped columns the load suddenly dropped
with loss of concrete cover. For wrapped columns, the first CFRP 400
Loadd (kN)

rupture occurred when the maximum load was reached, followed 300
by a decrease in load and a sharp increase in displacement. At this
stage, the effect of confinement was active to the controlling of the 200
column behavior. Some CFRP rupture still occurred until the col- 100
umn failed. Rupture of longitudinal straps was also observed in the
columns wrapped longitudinally with CFRP straps. 0
0 5 10 15 20 25 30
Midspan deflection (mm)
Influence of Eccentricity 0B 1HB 1V2HB 3HB
To describe the influence of eccentricity on the behavior of the col-
Fig. 12. Load-midspan deflection curves of beams tested under
umns, load-displacement curves of the 12 columns in relation to the
flexural loading
eccentricity were plotted as shown in Fig. 11. It can be clearly seen
that the eccentricity of loading reduced the load carrying capacity
and performance of the columns.
straps occurred until the beam failed (Fig. 12). Yielding of the ten-
sile reinforcing bars was observed in all beams. Table 3 summarizes
Behavior of Beams
the analysis results of the beams. It can be seen that Beam 1V2HB
All beams were tested to failure. Fig. 12 shows the load-midspan had the highest load and ductility of the four beams. Wrapping a
deflection of the beams tested under flexure. It can be seen that beam in the hoop direction with one layer of CFRP had no influ-
wrapping the beam in the hoop direction with CFRP did not im- ence on beam ductility. However, wrapping with three layers of
prove significantly the load carrying capacity and performance of CFRP increased beam ductility slightly. Calculations of beam duc-
the beam under flexural loading. However, the presence of CFRP tility were done using methods similar to those used for columns.
straps applied longitudinally in Beam 1V2HB showed a large im-
provement in its load carrying capacity compared to the other
beams. As in columns with vertical straps, a consequent strap- Axial Load-Bending Moment Interaction Diagrams of
by-strap rupture of the longitudinal strap was also observed in beam Columns
1V2HB. The first rupture of the strap resulted in a sudden decrease
in load, and then the beam still resisted the load under increasing Axial load-bending moment interaction diagrams were constructed
displacement until the second peak load was reached. The next rup- to describe the axial load and bending moment capacity of the col-
ture of the strap resulted in a drop of load. Several ruptures of the umns tested in this study. Any point on the diagrams may represent

JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / NOVEMBER/DECEMBER 2012 / 645

J. Compos. Constr. 2012.16:640-649.

 Table 3. Beam Test Results Table 4. Compressive Strength of Wrapped Concrete Columns
Test specimen 0B 1HB 1V2HB 3HB 0 (MPa)
Compressive Strength, f cc
Ultimate load (kN) 241 247 518 260 Theoretical Experimental Difference
Deflection at ultimate load (mm) 11.94 11.03 11.85 14.95 Columns 0
(fcc:the )a 0
(f cc: exp ) (%)
Yield deflection (mm) 4.06 4.20 4.15 3.89
1HC0 88.8 105.6 15.9
Maximum deflection (mm) 20.33 20.24 25.47 22.78
3HC0 107.3 118.6 9.5
Ductility, Method 1a 5.01 4.82 6.13 5.86
1V2HC0 98.0 116.3 15.7
Ductility, Method 2a 9.74 10.41 15.12 12.50
a
Calculated using Eq. (4).
Refer to the section “Behavior of Columns” for a definition of the
a

methods.
2f frp tfrp
fl ¼
Downloaded from ascelibrary.org by UNIVERSITY OF WOLLONGONG on 12/02/12. Copyright ASCE. For personal use only; all rights reserved.

(2)
d
a condition of a column under a concentric load, an eccentric load,
or a pure bending. The procedure described in Warner et al. (2007) where f l = lateral confining pressure; ffrp = tensile strength of FRP
for developing the interaction diagram of conventional RC columns in hoop direction; tfrp = total thickness of FRP; and d = diameter of
was followed. Full details are given in Warner et al. (2007). The confined concrete.
analysis to construct the interaction diagram of an FRP-confined The following formula, proposed by Richart et al. (1928), has
RC column was carried out based on principles of equilibrium been used by many researchers for predicting the compressive
and strain compatibility applied to a conventional RC column. strength of FRP-confined circular concrete columns (Lam and Teng
The only difference in the analysis of confined column was the 2002):
use of an FRP-confined stress-strain curve to replace the stress-
strain curve of unconfined concrete. The stress-strain curve of 0
fcc f
confined concrete proposed by Teng et al. (2002) was adopted here. ¼ 1 þ k1 l0 (3)
fc0 fc
The model is divided into two portions. The first portion is the
ascending branch of the curve from zero stress to stress equal to 0 = the compressive strength of the confined concrete and
where fcc
f c0 , which describes the stress-strain relationship of unconfined
k1 = the confinement effectiveness coefficient.
concrete. In the second portion, the stress beyond fc0 , the concrete
FRP-wrapped rectangular concrete columns have a different re-
stress increases linearly until reaching the stress of the confined
0 ) with the corresponding confined-concrete strain sponse than FRP-wrapped circular concrete columns, where the
concrete (fcc
0 ). The second portion describes the stress-strain relationship concrete is not uniformly confined by the FRP jacket. The nonun-
(εcc
iformity of confinement reduces the confinement effectiveness of
of confined concrete.
the FRP, and, furthermore, the compressive strength of confined
In developing axial load-bending moment interaction diagrams,
concrete varies over a given section (Mirmiran et al. 1998). To pre-
the axial load and the bending moment were calculated first. The
dict the confined concrete strength of rectangular column, the shape
axial concentric load capacity of an unconfined RC column (N uo )
factor (ks ) is taken into account to consider the effect of nonuni-
was calculated using the following formula (Warner et al. 2007):
formity of confinement. Thus, Eq. (3) is modified by Teng et al.
(2002) as follows:
N uo ¼ α1 fc0 Ag þ f sy As (1)
0
fcc f
¼ 1 þ k1 ks l0 (4)
where fc0 = compressive strength of the unconfined concrete; Ag = f c0 fc
cross-sectional area of column; f sy = yield strength of longitudinal
steel reinforcement; and As = total cross-sectional area of longitu-
dinal steel reinforcement. Because theoretical values were used to
Table 5. Experimental Bending Moment Capacity of Columns
develop the interaction N-M diagrams (which were really design
diagrams), the factor α1 was used, as recommended by Australian Deflection
standard AS3600-2009 (Standards Australia 2009). The standard Maximum at maximum
provides a value of α1 ¼ 1.0 − 0.003fc0 within the limits load, load, M1 ¼ N max e M 2 ¼ N max ðe þ δÞ
0.72 ≤ α1 ≤ 0.85. The value of α1 for concrete with compressive Specimen N max (kN) δ (mm) (kN-m) (kN-m)
strength up to 50 MPa is 0.85, which is similar to that recom- 0C0 3,248 0 0 0
mended by ACI318-2008 [American Concrete Institute (ACI) 0C25 1,950 1.87 48.75 52.39
2008]. The value of α1 used in the axial load calculation for the 0C50 1,336 2.65 66.80 70.33
case of wrapped columns was different than that used for un- 0B 241 11.94 28.30a 28.30a
wrapped columns. For unwrapped columns (Column 0C0) with 1HC0 3,279 0 0 0
1HC25 2,076 2.25 51.90 56.58
a compressive strength of 79.5 MPa, the value of α1 is 0.762.
1HC50 1,433 2.32 71.65 74.97
Meanwhile, for wrapped columns, i.e., Columns 1HC0, 3HC0, 1HB 247 11.03 29.01a 29.01a
and 1V2HC0, with compressive strength as shown in Table 4, 3HC0 3,585 0 0 0
the corresponding values of α1 are 0.734, 0.72, and 0.72, respec- 3HC25 2,269 2.11 56.73 61.52
tively. A minimum value of α1 was used for Columns 3HC0 and 3HC50 1,534 3.19 76.70 81.60
1V2HC0. 3HB 260 14.95 30.52a 30.52a
When an FRP-wrapped concrete column is subjected to com- 1V2HC0 3,522 0 0 0
pression, FRP confinement resists the expansion of the concrete 1V2HC25 2,296 2.44 57.40 63.01
core. Thus, the FRP is subjected to tension in the hoop direction. 1V2HC50 1,533 2.52 76.65 80.51
For a circular column, the lateral confining pressure is uniformly 1V2HB 518 11.85 60.87a 60.87a
a
distributed and can be calculated as follows (Lam and Teng 2002): Calculated using Eq. (10).

646 / JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / NOVEMBER/DECEMBER 2012

J. Compos. Constr. 2012.16:640-649.

  
b Ae area; Ac = total area of concrete; Rc = corner radius; and ρsc =
ks ¼ (5) cross-sectional area ratio of longitudinal steel. A value of
h Ac
k1 ¼ 2.98 was proposed for predicting the strength of FRP-
confined concrete columns having a rectangular or square cross
Ae 1 − ½ðb=hÞðh − 2Rc Þ2 þ ðh=bÞðb − 2Rc Þ2 =ð3Ag Þ − ρsc section. f l in Eq. (4) becomes an equivalent confining pressure
¼
Ac 1 − ρsc that is interpreted as the confining pressure provided by the
(6) FRP jacket to an equivalent circular column (with the diagonal
length D) and proposed as follows:

Ag ¼ bh − ð4 − πÞR2c (7)
2ffrp tfrp 2f frp tfrp
fl ¼ ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi (8)
where b and h = width and height of column section, respectively D h2 þ b2
Downloaded from ascelibrary.org by UNIVERSITY OF WOLLONGONG on 12/02/12. Copyright ASCE. For personal use only; all rights reserved.

(in the case of a square section, b ¼ h); Ae = effective confinement

The theoretical compressive strengths of the CFRP-wrapped
columns investigated in this study are presented in Table 4 and
4000 compared with the experimental results. The 28-day concrete com-
0C pressive strength was taken as the unconfined concrete compressive
3500
1HC strength to calculate the compressive strength of the confined
3000
3HC concrete. It can be seen that the confined compressive strength
Load N (kN)

2500 obtained from the theoretical results were 15.9, 9.5, and 15.7%
1V2HC
2000 smaller than that obtained from the experimental results for Col-
1500 umns 1HC0, 3HC0, and 1V2HC0, respectively. Since the compres-
sive strengths of the CFRP-wrapped columns were obtained, the
L

1000
axial concentric load capacity of the wrapped column (N uc ) were
500 calculated using Eq. (1) by replacing f c0 with f cc
0
, as formulated in
0 Eq. (4). As a comparison, another strength model, proposed by Wu
0 20 40 60 80 100
and Wang (2009), was used to predict the strength of confined col-
0
Bending moment M2 (kN-m) umns. Values of f cc =f c0 of 1.089, 1.173, and 1.255 for Columns
1HC, 1V2HC, and 3HC, respectively, were obtained, which were
Fig. 13. Experimental N-M diagram of columns
smaller than those calculated using Eq. (4). Eq. (4) yielded values

3500 3500
Experimental N-M1 Experimental N-M1
3000 3000
Experimental N-M2 Experimental N-M2
2500 Calculated N-M 2500 Calculated N-M
N (kN)
N (kN)

2000 2000
1500 1500
1000 1000
500 500
0 0
0 20 40 60 80 0 20 40 60 80
(a) M (kN-m) (b) M (kN-m)

4000 4000
Experimental N-M1 Experimental N-M1
3500 3500
Experimental N-M2 Experimental N-M2
3000 3000
Calculated N-M Calculated N-M
2500 2500
N (kN)
N (kN)

2000 2000
1500 1500
1000 1000
500 500
0 0
0 20 40 60 80 100 0 20 40 60 80 100
(c) M (kN-m) (d) M (kN-m)

Fig. 14. Experimental and theoretical N-M diagrams of columns

JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / NOVEMBER/DECEMBER 2012 / 647

J. Compos. Constr. 2012.16:640-649.

 0
of fcc =f c0 of 1.125, 1.213, and 1.355 for the corresponding 6. Using an existing strength model to estimate the capacity of
columns. CFRP column yielded strength magnitudes that were lower
The bending moment capacity (M) of the eccentrically loaded than the experimental ones. This observation shows that
columns, including the secondary moments, was calculated by the model yields good predictions of the strength of CFRP-
multiplying the maximum load capacity (N max ) and the sum of wrapped columns.
eccentricity (e) and lateral deflection (δ) at the maximum load 7. The theoretical axial load-bending moment interaction dia-
as follows: gram shows values that are close to those of the experimental
result, except at the point under concentric load.
M ¼ N max ðe þ δÞ (9) Finally, the use of longitudinal FRP layers can be recommended
in combination with FRP wrapped circumferentially to enhance the
However, the bending moment capacity of beams was calculated performance of columns loaded eccentrically.
using
Downloaded from ascelibrary.org by UNIVERSITY OF WOLLONGONG on 12/02/12. Copyright ASCE. For personal use only; all rights reserved.

P
M¼ a (10) References
2
American Concrete Institute (ACI). (2008). “Building code requirements
where P = maximum load applied to the specimen with four-point
for structural concrete and commentary.” ACI 318-08, Farmington Hills,
loading apparatus and a = shear span length, or the distance MI.
between the support and the closer loading point (a ¼ 235 mm ASTM. (2008). “Standard test method for tensile properties of polymer
in this study). matrix composite materials.” ASTM D 3039-08, Philadelphia.
The bending moment capacity of the investigated columns is Bisby, L., and Ranger, M. (2010). “Axial-flexural interaction in circular
reported in Table 5, and the experimental interaction diagram of FRP-confined reinforced concrete columns.” Constr. Build. Mater.,
the columns is shown in Fig. 13. It can be seen that wrapping a 24(9), 1672–1681.
column with one layer of CFRP increased slightly the axial load Chaallal, O., and Shahawy, M. (2000). “Performance of fiber-reinforced
and bending moment capacity of the eccentrically loaded columns. polymer-wrapped reinforced concrete column under combined
A larger increment than that was obtained in column wrapped with axial-flexural loading.” ACI Struct. J., 97(4), 659–668.
three layers of CFRP and in the columns wrapped with one layer Eid, R., Roy, N., and Paultre, P. (2009). “Normal- and high-strength
of vertical straps and two layers of CFRP. The only increment concrete circular elements wrapped with FRP composites.” J. Compos.
in bending moment capacity was obtained in columns that had Constr., 13(2), 113–124.
vertical CFRP straps. The presence of the CFRP straps applied lon- GangaRao, H. V. S., Taly, N., and Vijay, P. V. (2007). Reinforced concrete
design with FRP composites, CRC, London.
gitudinally contributed an important effect to the bending capacity
Hadi, M. N. S. (2006a). “Behaviour of FRP wrapped normal strength
of the column. concrete columns under eccentric loading.” Compos. Struct., 72(4),
Fig. 14 shows the calculated and experimental interaction 503–511.
diagrams of Columns 0C, 1HC, 3HC, and 1V2HC, respectively. Hadi, M. N. S. (2006b). “Comparative study of eccentrically loaded FRP
It can be seen from those figures that under eccentric loading wrapped columns.” Compos. Struct., 74(2), 127–135.
the calculated interaction diagram gave values that were close to Hadi, M. N. S. (2007a). “Behaviour of FRP strengthened concrete columns
the experimental results. However, the calculated axial loads of under eccentric compression loading.” Compos. Struct., 77(1), 92–96.
all columns under concentric loading were smaller than those of Hadi, M. N. S. (2007b). “The behaviour of FRP wrapped HSC columns
the experimental results. under different eccentric loads.” Compos. Struct., 78(4), 560–566.
Kumutha, R., Yaidyanathan, R., and Palanichamy, M. S. (2007). “Behav-
iour of reinforced concrete rectangular columns strengthened using
GFRP.” Cem. Concr. Compos., 29(8), 609–615.
Conclusions
Lam, L., and Teng, J. G. (2002). “Strength model for fiber-reinforced
plastic-confined concrete.” J. Struct. Eng., 128(5), May, 612–623.
A total of 16 square RC specimens with CFRP wrapping were
Li, J., and Hadi, M. N. S. (2003). “Behaviour of externally confined high
tested in this study to investigate the influence of eccentricity,
strength concrete columns under eccentric loading.” Compos. Struct.,
number of layers, and presence of vertical straps on specimen 62(2), 145–153.
behavior and load carrying capacity. Twelve columns were tested Maaddawy, T. E. (2009). “Strengthening of eccentrically loaded reinforced
under compression loading and four beams were tested under concrete columns with fiber-reinforced polymer wrapping system: Ex-
flexural loading. Based on the experimental work carried out in this perimental investigation and analytical modeling.” J. Compos. Constr.,
study, the following conclusions were drawn: 13(1), 13–24.
1. CFRP wrapping had a more significant effect on the maximum Mirmiran, A., Shahawy, M., Samaan, M., Echary, H. E., Mastrapa, J. C.,
load of eccentrically loaded columns compared to concentri- and Pico, O. (1998). “Effect of column parameters on FRP-confined
cally loaded columns. concrete.” J. Compos. Constr., 2(4), 175–185.
2. No significant increase in maximum load was obtained when Parvin, A., and Wang, W. (2001). “Behavior of FRP jacketed concrete
the columns were wrapped with one layer of CFRP. columns under eccentric loading.” J. Compos. Constr., 5(3), 146–152.
3. Increasing the number of the CFRP layers resulted in an Richart, F. E., Brandtzaeg, A., and Brown, R. L. (1928). “A study of
increase in the load and the performance of the columns. the failure of concrete under combined compressive stresses.” Bulletin
No. 185, Engineering Experimental Station, Univ. of Illinois at Urbana-
4. The CFRP wrapping enhanced the performance of the col-
Champaign, Urbana-Champaign, IL.
umns by postponing the rupture of the concrete and reinforce- Rousakis, T. C., Karabinis, A. I., and Kiousis, P. D. (2007). “FRP-confined
ment, which means it increased the column ductility. Similar concrete members: Axial compression experiments and plasticity
results were obtained in beams tested in flexure. modeling.” Eng. Struct., 29(7), 1343–1353.
5. In columns with a large eccentricity, which means with a large Standards Australia. (2009). “Concrete structures.” AS3600-2009, Sydney,
bending moment, the presence of CFRP straps produced Australia.
higher ductility than in columns wrapped horizontally with Teng, J. G., Chen, J. F., Smith, S. T., and Lam, L. (2002). FRP-strengthened
a similar number of CFRP layers. RC structures, Wiley, West Sussex, UK.

648 / JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / NOVEMBER/DECEMBER 2012

J. Compos. Constr. 2012.16:640-649.

 Teng, J. G., and Lam, L. (2004). “Behavior and modeling of fiber Warner, R. F., Foster, S. J., and Kilpatrick, A. E. (2007). Reinforced
reinforced polymer-confined concrete.” J. Struct. Eng., 130(11), concrete basics: Analysis and design of reinforced concrete structures,
1713–1723. Pearson Education, Sydney, Australia.
Wang, L. M., and Wu, Y. F. (2008). “Effect of corner radius on the Wu, Y. F., and Wang, L. M. (2009). “Unified strength model for square and
performance of CFRP-confined square concrete columns: Test.” Eng. circular concrete columns confined by external jacket.” J. Struct. Eng.,
Struct., 30(2), 493–505. 135(3), 253–261.
Downloaded from ascelibrary.org by UNIVERSITY OF WOLLONGONG on 12/02/12. Copyright ASCE. For personal use only; all rights reserved.

JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / NOVEMBER/DECEMBER 2012 / 649

J. Compos. Constr. 2012.16:640-649.