Axial Compressive Behavior of Circular

High-Strength Concrete-Filled FRP Tubes

Togay Ozbakkaloglu 1 and Thomas Vincent 2
Downloaded from ascelibrary.org by M S Ramaiah Institute of Technology on 06/25/18. Copyright ASCE. For personal use only; all rights reserved.

Abstract: Concrete-filled fiber-reinforced polymer (FRP) tubes (CFFTs) have received significant research attention over the last two
decades. However, experimental studies on the behavior of CFFTs filled with high-strength concrete (HSC) remain very limited. This paper
presents the results of an experimental study on the axial compressive behavior of 83 monotonically-loaded circular CFFTs. The effects of
fiber type, concrete strength, specimen size, and manufacturing method on the compressive behavior of CFFTs were investigated. The CFFTs
were manufactured with carbon FRP (CFRP), high-modulus CFRP (HMCFRP), or aramid FRP (AFRP) tubes, and their average unconfined
concrete strengths ranged between 34–110 MPa. The diameters of the test specimens ranged from 75–300 mm with all specimens maintaining
a 2∶1 height-to-diameter ratio. The effect of the CFFT manufacturing method was investigated through AFRP specimens that were
manufactured through either an automated filament winding or manual wet layup technique. The experimentally recorded stress-strain
relationships are presented graphically and the ultimate axial stresses and strains and hoop rupture strains are tabulated. The large quantity
of the results presented in this paper allows for a number of significant conclusions to be drawn. The results clearly indicate that over a certain
confinement threshold, high-strength CFFTs (HSCFFTs) exhibit a highly ductile behavior. However, for the same nominal confinement ratio,
compressive behavior of CFFTs degrades as concrete strength increases. The results also indicate that the compressive behavior of CFFTs is
significantly influenced by the manufacturing method and fiber type with an improvement in compressive behavior linked to an increase in
fiber rupture strain. Finally, the influence of specimen size was found to be negligible for the range of diameters tested in this study. Further
experimental observations on these and other key parameters are presented and discussed in the paper. DOI: 10.1061/(ASCE)CC.1943-5614
.0000410. © 2013 American Society of Civil Engineers.
Author keywords: Concrete; High-strength concrete (HSC); Fiber-reinforced polymer (FRP); Columns; Concrete-filled FRP tube (CFFT);
Confinement; Compressive strength; Stress-strain relationship.

Introduction columns (Seible et al. 1996; Yamakawa et al. 2003; Zhu et al.
2006; Ozbakkaloglu and Saatcioglu 2006, 2007; Saatcioglu et al.
It is now well-understood that the confinement of concrete with 2008; Idris and Ozbakkaloglu 2013) have demonstrated the ability
fiber-reinforced polymer (FRP) composites can lead to significant of CFFTs to develop very high inelastic deformation capacities,
improvements in both compressive strength and ductility. A recent which makes them a highly attractive alternative for construction
comprehensive review study (Ozbakkaloglu et al. 2013) revealed of new high-performance columns.
that over 200 experimental studies have been conducted over The popularity of higher strength concretes in the construction
the last two decades on the compressive behavior of FRP-confined industry has been on a steady incline during the last two decades
concrete, resulting in the developments of over 80 axial stress-strain because of the superior performance and economy offered by
models (e.g., De Lorenzis and Tepfers 2003; Lam and Teng 2003; high-strength concrete (HSC) over normal-strength concrete
Binici 2005; Bisby et al. 2005; Youssef et al. 2007; Ilki et al. (NSC) in a large number of structural engineering applications.
2008; Wei and Wu 2012; Yazici and Hadi 2012). The majority These beneficial characteristics result in more cost-effective con-
of these studies focused on FRP-wrapped specimens, and rela- struction of bridges and multistory buildings. However, the use of
tively few studies have been reported on the behavior of concrete- higher strength concretes in seismically active regions poses dif-
filled FRP tubes (CFFTs) (Mirmiran et al. 1998; Saafi et al. 1999; ficulties because of the inherently brittle nature of the material.
Fam and Rizkalla 2001; Hong and Kim 2004; Fam et al. 2005; Confining HSC with FRP tubes is an attractive option because
Ozbakkaloglu and Oehlers 2008a, b; Mohamed and Masmoudi of the efficient combination of two high-strength materials to form
2010; Park et al. 2011; Ozbakkaloglu 2013a, b, c). Nonetheless, a high-performance member that benefits from a substantial
these early studies on the compressive behavior of CFFTs together increase in ductility compared with unconfined HSC members.
with a few studies reported on the seismic behavior of CFFT Research on the compressive behavior of FRP-confined HSC,
in general, and on high-strength CFFTs (HSCFFTs), in particular,
1
Senior Lecturer, School of Civil, Environmental and Mining Engineer- remain very limited with only a handful of studies reported on
ing, Univ. of Adelaide, SA 5005, Australia (corresponding author). E-mail: FRP-wrapped specimens (Rousakis 2001; Berthet et al. 2005;
[email protected] Mandal et al. 2005; Almusallam 2007; Eid et al. 2009; Wu et al.
2
Ph.D. Candidate, School of Civil, Environmental and Mining Engi- 2009; Cui and Sheikh 2010; Xiao et al. 2010; Ozbakkaloglu and
neering, Univ. of Adelaide, SA 5005, Australia.
Note. This manuscript was submitted on January 30, 2013; approved on
Akin 2012) and only one on CFFTs (Ozbakkaloglu 2013a).
June 27, 2013; published online on July 1, 2013. Discussion period open The aforementioned review study by Ozbakkaloglu et al. (2013)
until April 6, 2014; separate discussions must be submitted for individual revealed that the majority of existing experimental studies on the
papers. This paper is part of the Journal of Composites for Construction, compressive behavior of FRP-confined concrete in circular sections
© ASCE, ISSN 1090-0268/04013037(11)/$25.00. have been concerned with FRP-wrapped specimens and very few

© ASCE 04013037-1 J. Compos. Constr.

J. Compos. Constr., 2014, 18(2): 04013037

 studies have been reported on circular CFFTs. Furthermore, the Table 1. Details of Test Specimens
majority of the studies on the axial compressive behavior of CFFTs Concrete Dimensions Number of
have been concerned with specimens manufactured with glass type (mm) Fiber type FRP layers specimens
FRP (GFRP) tubes (Mirmiran et al. 1998; Fam and Rizkalla 2001;
NSC 74 × 150 CFRP 1 4
Mohamed and Masmoudi 2010; Park et al. 2011) with a limited 100 × 200 AFRP 1 3
number of studies reported on carbon FRP (CFRP)-confined AFRP-FW 0.3a 3
CFFTs (Davol et al. 2001; Hong and Kim 2004) and a single 152 × 305 CFRP 1 2
study, which only provided three test results, on high-modulus 2 3
CFRP (HMCFRP)-confined CFFTs (Saafi et al. 1999). As recently HMCFRP 1 3
demonstrated in Ozbakkaloglu and Akin (2012), the behavior of 302 × 605 CFRP 2 1
CFFTs is significantly influenced by the properties of the confining 4 1
HSC1 74 × 150 CFRP 1 3
Downloaded from ascelibrary.org by M S Ramaiah Institute of Technology on 06/25/18. Copyright ASCE. For personal use only; all rights reserved.

fibers, and a certain fiber type could prove to be more suitable for
2 3
an intended application of a given CFFT. Therefore, for a success-
100 × 200 AFRP 2 3
ful design of CFFTs, it is important to understand and be able to 3 3
model the influence of fiber properties on CFFT behavior. To this AFRP-FW 0.6a 3
end, the study reported in this paper investigated the behavior of 152 × 305 AFRP 3 3
circular CFFTs confined with FRP materials that received minimal CFRP 1 3
attention in previous studies, namely aramid FRP (AFRP), CFRP, 2 3
and HMCFRP. 3 3
Small scale specimens have been the focus of existing experi- HMCFRP 1 3
mental research on the compressive behavior of FRP-confined 2 3
concrete, with limited data available on full-scale columns. HSC2 74 × 150 CFRP 1 3
2 2
The aforementioned review study by Ozbakkaloglu et al. (2013)
3 4
revealed that researchers performed the majority of existing exper- 100 × 200 AFRP 3 3
imental studies on specimens with nominal diameters of 150 mm. If 4 3
the behavior examined in these studies is to be applied to the design AFRP-FW 0.9a 3
of large-scale structures, then an understanding of the influence of 152 × 305 AFRP 6 3
specimen size is essential. To investigate the influence of specimen CFRP 3 3
size, this study examined the compressive behavior of test speci- 4 3
mens with four different sizes, ranging in diameters from 75 6 3
to 300 mm. Total 83
As the first experimental study on the axial compressive behav- Note: HSC1 = lower grade high-strength concrete; HSC2 = higher grade
ior of circular HSCFFTs confined by AFRP, CFRP, or HMCFRP high-strength concrete; NSC = normal-strength concrete.
a
tubes, the study reported in this paper was aimed at investigating Indicates total nominal fiber thickness (mm) for filament wound
the influence of key confinement parameters and address the afore- specimens.
mentioned research gaps. First, this paper provides a summary of
the test program, including specimen properties and the testing pro- with compressive strengths below 55 MPa and they were included
cedure. Following this, results of the test program are presented, in the test matrix to establish reference values to allow a compari-
where the ultimate conditions of the test specimens are tabulated son between NSC and higher strength concrete specimens. It
and their stress-strain responses are graphically illustrated. Finally, should be noted that due to the differences in the specimen size
a detailed discussion on the results of the test program is provided, 0
and curing conditions, in-place concrete strengths (f co ) given in
where the influence of fiber type, concrete strength, specimen size, Table 2 for some of the specimens differed from the control cyl-
and FRP tube manufacturing methods are examined along with inder strengths used to classify the mixes, and in some cases,
0
other key experimental outcomes. fco values reported for these specimens fell slightly outside of
the aforementioned ranges.
The majority of the specimens (i.e., 44) were constructed with
Test Program tubes manufactured with CFRP. Another 30 specimens were pre-
pared with AFRP tubes and nine with HMCFRP tubes in order to
examine the influence of fiber type on specimen axial behavior. The
Test Specimens
influence of tube manufacturing method was investigated with
Table 1 summarizes the research program, which included a total of AFRP specimens manufactured with either automated filament
83 CFFTs set up to investigate the influence of the concrete com- winding or manual wet layup techniques. The influence of speci-
pressive strength, specimen size, manufacturing method, and fiber men size was examined by designing the specimens with nominal
type of FRP tubes. Concrete mixes were designed for target diameters of 75, 100, 150, and 300 mm, with each specimen main-
strengths ranging from 30–120 MPa, with the compressive taining a 2∶1 height-to-diameter ratio.
strengths of these mixes established by testing 100-mm diameter The number of FRP layers of the CFFTs was established accord-
control cylinders. The majority of the CFFTs (i.e., 63) had com- ing to the concrete strength and specimen size, with higher strength
pressive strengths greater than 55 MPa, and they were classified and larger specimens receiving proportionally more layers to
as HSC specimens. Within this category, 33 of the specimens ensure adequate confinement. This was done through the use of
had compressive strengths less than 85 MPa and were classified 0 as the performance criterion,
the nominal confinement ratio f lu =f co
as lower grade HSCs, labeled HSC1-FFT; whereas, 30 of the which is the ratio of the lateral confinement pressure at ultimate
HSC specimens had compressive strengths greater than 85 MPa 0
(flu ) to the in-place concrete compressive strength (fco ). The
0
and were classified as HSC2-FFTs to represent the higher grade nominal confinement ratio (flu =fco ) was calculated from
HSC. The remaining 20 specimens were classified as NSCFFTs Eq. (1), assuming a uniform confinement distribution:

© ASCE 04013037-2 J. Compos. Constr.

J. Compos. Constr., 2014, 18(2): 04013037

 Table 2. Test Results
Concrete 0
fco εco tf 0
f cc 0
fcu εcu εh;rup
0 0 0 0
type Group Specimen (MPa) (%) (mm) f lu =f co flu;a =f co (MPa) (MPa) (%) (%) f cc =f co εcu =εco kε
NSC 75-mm CFRP N-75C1L-1 43.0 0.22 0.117 0.28 0.19 67.4 — 1.35 1.07 1.48 5.09 0.67
N-75C1L-2 43.0 0.22 0.117 0.28 0.24 71.0 — 1.44 1.32
N-75C1L-3 43.0 0.22 0.117 0.28 0.16 61.1 — 0.92 0.91
N-75C1L-4 47.8 0.23 0.117 0.25 0.13 60.9 — 0.84 0.83
100-mm AFRP N-100A1L-1 37.0 0.21 0.200 0.31 0.28 70.6 — 2.06 2.22 1.87 9.34 0.87
N-100A1L-2 35.5 0.20 0.200 0.33 0.27 65.5 — 1.75 2.08
N-100A1L-3 34.0 0.20 0.200 0.34 0.31 62.8 — 1.88 2.25
N-100AFW0.3-1 37.2 0.21 0.300 0.47 0.34 89.1 — 3.10 2.11 2.44 15.31 0.77
N-100AFW0.3-2 37.2 0.21 0.300 0.47 0.39 91.9 — 3.31 2.39
Downloaded from ascelibrary.org by M S Ramaiah Institute of Technology on 06/25/18. Copyright ASCE. For personal use only; all rights reserved.

N-100AFW0.3-3 35.4 0.20 0.300 0.49 0.37 86.7 — 3.04 2.21

150-mm CFRP N-150C1L-1 37.3 0.21 0.117 0.16 0.12 42.0 — 0.79 1.20 1.16 3.55 0.64
N-150C1L-2 34.6 0.20 0.117 0.17 0.08 41.6 — 0.66 0.77
N-150C2L-1 35.5 0.20 0.234 0.33 0.28 59.1 — 1.43 1.32 1.67 7.17 0.84
N-150C2L-2 36.3 0.21 0.234 0.32 0.28 60.9 — 1.53 1.36
N-150C2L-3 37.3 0.21 0.234 0.31 0.25 61.7 — 1.45 1.23
150-mm HMCFRP N-150HM1L-1 36.3 0.21 0.190 0.18 0.05 46.4 — 0.28 0.12 1.25 1.34 0.34
N-150HM1L-2 36.3 0.21 0.190 0.18 0.05 46.0 — 0.30 0.11
N-150HM1L-3 36.3 0.21 0.190 0.18 0.08 43.3 — 0.25 0.18
300-mm CFRP N-300C2L-1 36.3 0.21 0.234 0.16 0.11 38.6 — 0.80 1.08 1.06 3.90 0.70
N-300C4L-1 36.3 0.21 0.468 0.32 0.24 57.0 — 1.52 1.17 1.57 7.41 0.75
HSC1 75-mm CFRP H1-75C1L-1 62.0 0.26 0.117 0.19 0.06 69.9 — 0.63 0.50 1.11 2.56 0.34
H1-75C1L-2 66.6 0.27 0.117 0.18 0.04 71.5 70.1 0.57 0.36
H1-75C1L-3 55.0 0.25 0.117 0.22 0.10 62.0 56.5 0.80 0.72
H1-75C2L-1 55.0 0.25 0.234 0.44 0.32 96.0 — 1.43 1.13 1.91 7.61 0.68
H1-75C2L-2 50.3 0.24 0.234 0.48 0.29 98.1 — 1.71 0.95
H1-75C2L-3 52.0 0.24 0.234 0.46 0.32 105.7 — 2.41 1.07
100-mm AFRP H1-100A2L-1 85.9 0.31 0.400 0.27 0.19 121.3 — 1.65 1.76 1.36 5.27 0.74
H1-100A2L-2 82.4 0.31 0.400 0.28 0.21 107.3 — 1.58 1.84
H1-100A2L-3 82.4 0.31 0.400 0.28 0.22 112.3 — 1.65 1.92
H1-100A3L-1 85.9 0.31 0.600 0.41 0.26 148.2 — 1.92 1.62 1.79 6.96 0.74
H1-100A3L-2 85.9 0.31 0.600 0.41 0.28 154.3 — 2.23 1.76
H1-100A3L-3 85.9 0.31 0.600 0.41 0.35 159.7 — 2.38 2.17
H1-100AFW0.6-1 85.9 0.31 0.600 0.41 0.33 176.2 — 2.89 2.36 1.99 8.91 0.75
H1-100AFW0.6-2 83.0 0.31 0.600 0.42 0.25 154.9 — 2.53 1.74
H1-100AFW0.6-3 85.9 0.31 0.600 0.41 0.34 176.6 — 2.89 2.42
150-mm AFRP H1-150A3L-1 79.6 0.30 0.600 0.29 0.24 105.0 — 1.67 2.12 1.39 6.21 0.73
H1-150A3L-2 77.2 0.30 0.600 0.30 0.19 102.0 — 1.64 1.59
H1-150A3L-3 77.0 0.30 0.600 0.30 0.21 118.0 — 2.23 1.79
150-mm CFRP H1-150C1L-1 59.0 0.26 0.117 0.10 0.06 58.8 45.2 0.72 0.89 1.01 2.48 0.64
H1-150C1L-2 59.0 0.26 0.117 0.10 0.07 60.1 39.0 0.56 1.08
H1-150C1L-3a 59.0 0.26 0.117 0.10 0.07 57.3 43.3 0.61 1.03
H1-150C2L-1 59.0 0.26 0.234 0.20 0.15 68.4 — 0.95 1.14 1.11 3.63 0.72
H1-150C2L-2 59.0 0.26 0.234 0.20 0.15 65.4 — 1.05 1.19
H1-150C2L-3 62.0 0.26 0.234 0.19 0.13 66.8 — 0.84 1.03
H1-150C3L-1 59.0 0.26 0.351 0.30 0.21 79.2 — 1.24 1.07 1.31 5.18 0.60
H1-150C3L-2 65.0 0.27 0.351 0.27 0.13 78.0 — 1.30 0.77
H1-150C3L-3 59.0 0.26 0.351 0.30 0.18 81.6 — 1.54 0.92
150-mm HMCFRP H1-150HM1L-1 59.0 0.26 0.190 0.11 0.07 70.0 — 0.50 0.26 1.19 1.93 0.61
H1-150HM1L-2 55.6 0.25 0.190 0.12 0.07 66.6 — 0.50 0.22
H1-150HM1L-3 59.0 0.26 0.190 0.11 0.07 69.9 — 0.47 0.26
H1-150HM2L-1 59.0 0.26 0.380 0.22 0.06 70.8 — 0.47 0.11 1.25 1.71 0.29
H1-150HM2L-2 59.0 0.26 0.380 0.22 0.08 77.3 — 0.45 0.14
H1-150HM2L-3 59.0 0.26 0.380 0.22 0.06 73.5 — 0.40 0.10
HSC2 75-mm CFRP H2-75C1L-1 75.0 0.29 0.117 0.16 0.06 86.2 — 0.66 0.62 1.08 2.39 0.44
H2-75C1L-2 77.0 0.30 0.117 0.16 0.08 83.4 — 0.78 0.83
H2-75C1L-3 83.1 0.31 0.117 0.14 0.06 84.5 78.4 0.70 0.62
H2-75C2L-1 83.1 0.31 0.234 0.29 0.18 104.4 96.9 1.31 0.95 1.30 4.02 0.61
H2-75C2L-2 83.1 0.31 0.234 0.29 0.18 111.2 — 1.16 0.95
H2-75C3L-1 93.8 0.33 0.351 0.38 0.21 141.4 — 1.29 0.85 1.48 3.68 0.50
H2-75C3L-2 99.9 0.34 0.351 0.36 0.22 121.2 119.8 1.26 0.93
H2-75C3L-3 77.0 0.30 0.351 0.47 0.22 131.8 — 1.14 0.73
H2-75C3L-4 82.5 0.31 0.351 0.44 0.16 122.6 — 0.97 0.57
100-mm AFRP H2-100A3L-1 110.1 0.35 0.600 0.32 0.17 154.8 — 2.11 1.35 1.40 5.36 0.62
H2-100A3L-2 110.1 0.35 0.600 0.32 0.19 150.9 — 1.71 1.54
H2-100A3L-3 110.1 0.35 0.600 0.32 0.23 156.6 — 1.87 1.78

© ASCE 04013037-3 J. Compos. Constr.

J. Compos. Constr., 2014, 18(2): 04013037

 Table 2. (Continued.)
Concrete 0
fco εco tf 0
f cc 0
fcu εcu εh;rup
0 0 0 0
type Group Specimen (MPa) (%) (mm) f lu =f co flu;a =f co (MPa) (MPa) (%) (%) f cc =f co εcu =εco kε
H2-100A4L-1 110.1 0.35 0.800 0.42 0.25 183.8 — 2.21 1.47 1.70 6.63 0.61
H2-100A4L-2 110.1 0.35 0.800 0.42 0.27 190.9 — 2.47 1.57
H2-100A4L-3a 110.1 0.35 0.800 0.42 — 198.8 — — —
H2-100AFW0.9-1 110.1 0.35 0.900 0.47 0.33 232.4 — 3.22 2.01 2.12 8.97 0.73
H2-100AFW0.9-2 110.1 0.35 0.900 0.47 0.34 224.1 — 2.81 2.11
H2-100AFW0.9-3 110.1 0.35 0.900 0.47 0.37 244.6 — 3.48 2.26
150-mm AFRP H2-150A6L-1 104.5 0.34 1.200 0.44 0.21 164.3 — 1.98 1.19 1.63 6.01 0.58
H2-150A6L-2 104.5 0.34 1.200 0.44 0.27 168.7 — 2.18 1.53
H2-150A6L-3 104.5 0.34 1.200 0.44 0.29 178.9 — 2.05 1.63
Downloaded from ascelibrary.org by M S Ramaiah Institute of Technology on 06/25/18. Copyright ASCE. For personal use only; all rights reserved.

150-mm CFRP H2-150C3L-1 92.7 0.32 0.351 0.19 0.09 101.5 84.8 0.81 0.75 1.09 2.59 0.53
H2-150C3L-2 94.7 0.33 0.351 0.19 0.10 103.7 99.2 0.89 0.86
H2-150C3L-3 90.1 0.32 0.351 0.19 0.11 96.0 86.7 0.82 0.84
H2-150C4L-1 93.0 0.33 0.468 0.25 0.12 97.9 95.8 0.92 0.71 1.08 2.91 0.55
H2-150C4L-2 100.0 0.34 0.468 0.23 0.13 107.9 98.9 0.96 0.88
H2-150C4L-3 97.5 0.33 0.468 0.24 0.15 107.2 — 1.01 0.97
H2-150C6L-1 102.5 0.34 0.702 0.34 0.20 131.1 — 1.27 0.89 1.26 3.53 0.50
H2-150C6L-2 96.0 0.33 0.702 0.37 0.18 124.2 — 1.16 0.78
H2-150C6L-3 93.0 0.33 0.702 0.38 0.16 112.1 — 1.09 0.66
a
Indicates specimens that experienced problems with loading or instrumentation.

f lu 2Ef tf εfu prepared using a manual wet layup process that involved wrapping
0 ¼ 0 ð1Þ
fco Df co epoxy resin impregnated fiber sheets around precision-cut, high-
density Styrofoam templates in the hoop direction. The epoxy resin
where flu = the confining pressure at ultimate; Ef = the modulus of was allowed to cure at room temperature for at least 24 h before the
elasticity; tf = the total nominal thickness; εfu = the ultimate tensile FRP tubes were removed from their molds. The group of specimens
strain of the fibers; and D = the internal diameter of the CFFT. that were not prepared manually were manufactured using an au-
Three nominally identical specimens were manufactured and tomated filament winding process with the majority of the fibers
tested for each unique specimen configuration, unless marked aligned along the hoop direction through the use of a winding
otherwise in Table 1. angle of 88 degrees. Stacking sequences of [þ88], [þ88, −88], or
[þ88, −88, þ88] were used for the manufacturing of the tubes with
0.3 mm, 0.6 mm, and 0.9 mm thicknesses, respectively. These auto-
Material Properties matically manufactured tubes were all made from aramid fibers and
are designated as aramid fiber-reinforced polymer for the filament
Concrete wound specimens (AFRP-FW) in Table 1, whereas the manually
All of the concrete mixes were prepared in the laboratory, and these prepared tubes are designated as AFRP. The AFRP-FW tubes
mixes consisted of ordinary portland cement as the binder and had a fiber volume fraction of 63%, and they were manufactured
crushed limestone, with a 10-mm nominal maximum diameter, at the University of Alberta in Canada.
as the coarse aggregate. The HSC mixes were prepared with silica All manually prepared FRP tubes were manufactured from uni-
fume dosed at 8% of the binder content by weight. Superplasticizer directional sheets and were finished with a 100 mm overlap. Spec-
was added to the HSC mixes in different amounts to ensure a work- imens with one to three layers of FRP were wrapped with a single
able concrete with measured slumps of more than 200 mm. length of FRP continuously, whereas specimens with four to six
layers of FRP were wrapped with two FRP sheets, creating two
overlap zones of 100 mm length. The manufacturer supplied prop-
FRP erties of the unidirectional fiber sheets used in the fabrication of the
The specimens were manufactured by casting concrete into FRP tubes are provided in Table 3. In addition, the material proper-
preformed circular FRP tubes. The majority of these tubes were ties of the FRP composites were established through flat coupon

Table 3. Material Properties of Fibers and Epoxy Resin

Fiber properties
Epoxy resin properties provided
Provided by manufacturers Experimentally determineda by manufacturer
Nominal
fiber Ultimate Ultimate Elastic Ultimate Ultimate Elastic Ultimate Ultimate Elastic
thickness tensile stress, tensile strain, modulus, tensile stress, tensile strain, modulus, tensile stress, tensile strain, modulus,
Type (mm/ply) f fu (MPa) εfu (MPa) Ef (GPa) f Fu (MPa) εFu (MPa) EF (GPa) f Eu (MPa) εEu (MPa) EE (GPa)
Aramid 0.200 2,900 2.50 116 2,663 2.12 125.7 — — —
Aramid FW — 2,930 2.90 99 — — — — — —
Carbon 0.117 3,800 1.55 240 3,626 1.44 251.0 — — —
HM carbon 0.190 2,650 0.40 640 2,365 0.36 657.0 — — —
Epoxy resin — — — — — — — >50 2.5 >3
a
Obtained from flat FRP coupon tests and calculated based on nominal thickness of fibers.

© ASCE 04013037-4 J. Compos. Constr.

J. Compos. Constr., 2014, 18(2): 04013037

 tests, in which the coupon specimens were loaded in accordance Specimen Designation
with ASTM standard D3039M-08 (ASTM 2008). Along with
the manufacturer-supplied properties of the epoxy resin, Table 3 The specimens were identified according to the concrete grade and
reports the material properties obtained from the coupon tests, FRP tube details, where the latter included the internal tube diam-
calculated on the basis of nominal fiber thicknesses. eter, FRP type, and number of FRP layers. Concrete grade is des-
ignated as either N (normal-strength), H1 (high-strength level 1), or
H2 (high-strength level 2). FRP tube details are then given starting
Instrumentation and Testing Procedure with the tube diameter in millimeters. The symbols C, A, and
HM are used to represent the types of FRP-confinement, namely:
The specimens were instrumented with four linear variable differ- carbon (C), aramid (A), or high-modulus carbon (HM). The num-
ential transformers (LVDTs) mounted at the corners between the ber of FRP layers is then provided accompanied by the letter ‘L’
loading and supporting steel plates of the test machine as shown and a final number was provided to identify between nominally
Downloaded from ascelibrary.org by M S Ramaiah Institute of Technology on 06/25/18. Copyright ASCE. For personal use only; all rights reserved.

in Fig. 1. The recorded deformations were used in the calculation identical specimens. For the filament wound specimens, the letter
of the average axial strains along the height of the specimens. In that indicated the FRP type was followed by the acronym FW, and
addition to the LVDTs, the specimen was also instrumented at the the total nominal fiber thickness of the tube was provided in place
midheight with two unidirectional axial strain gauges installed on of the FRP layers. For example, a specimen that consists of level 1
opposing sides. Lateral strains were measured by three strain high-strength concrete, with a nominal diameter of 150 mm, en-
gauges installed at 120 degree intervals around the perimeter. Both cased by two layers of carbon FRP, and is the third in its group
axial and lateral strain gauges were 20 mm in length and were in- would be labeled as H1-150C2L-3.
stalled at locations outside of the overlap region. The axial strain
gauge readings were used to correct the LVDT measurements at the
early stages of loading, in which LVDT measurements are consid- Test Results and Discussion
ered inaccurate as they record additional displacements attributable
to the closure of the gaps in the setup.
The specimens were tested under monotonic axial compression Ultimate Condition
using a 5000-kN capacity, universal testing machine. Prior to test- The ultimate condition, which consists of the ultimate axial
ing, all specimens were capped at both ends to ensure uniform dis- 0
strength (f cu ) recorded at failure of the specimen, the correspond-
tribution of the applied pressure and the load was applied directly to ing axial strain (εcu ), and the FRP hoop rupture strain (εh;rup ), of
the concrete core through the use of precision-cut, high-strength each specimen is reported in Table 2. Fig. 2 presents the
steel end plates. During the initial stage of elastic behavior, the representative stress-strain relationships of the specimens, with a
loading was applied at 3 kN per second, whereas displacement con- full set of curves presented in Figs. S1–S3, which are available
trol was used at approximately 10 microstrain per second beyond online in the ASCE Library (www.ascelibrary.org). When the
the initial softening until specimen failure. Fig. 1 provides a tech- stress-strain relationship contained a descending branch so that
nical illustration of the instrumentation and testing equipment used 0
the determined ultimate strength (f cu ) was lower than the recorded
in the experimental study. 0 0 0
peak strength (f cc ), then both the ultimate (fcu ) and peak (f cc )
The specimens were tested in three groups which began after the strengths were reported in Table 2. When the stress-strain relation-
28-day strength of concrete was attained and continued for approx- ship contained an ascending branch so that the ultimate strength
imately three weeks for each group. Table 2 reports the in-place 0 0
(fcu ) was the peak strength (f cc ), this value was reported under
0 ) at the time of testing, 0
strengths of the unconfined concrete (fco fcc in Table 2. The ultimate axial strain (εcu ) of each specimen
along with the axial strain at peak strength (εco ). εco values were was calculated by averaging the four LVDT readings recorded
not measured directly and were calculated using the expression at failure. FRP hoop rupture (εh;rup ) was averaged from the three
given by Tasdemir et al. (1998). lateral strain gauge readings prior to FRP tube rupture. Previous
research has established that the hoop rupture strains (εh;rup ) are
often lower than the ultimate tensile strain of the material (εfu )
(e.g., Pessiki et al. 2001; Ilki and Kumbasar 2003; Lam and Teng
2003; Ozbakkaloglu and Oehlers 2008b; Ozbakkaloglu and Akin
2012). For each group of nominally identical specimens, Table 2
presents the average strain reduction factors (kε ) determined from
Eq. (2):
εh;rup
kε ¼ ð2Þ
εfu

Axial Stress-Strain Behavior

Fig. 2 illustrates the effect of important confining parameters on the
compressive behavior of NSCFFTs and HSCFFTs. The stress-
strain curves show that HSCFFTs can exhibit highly ductile behav-
ior when sufficiently confined by FRP tubes. However, a loss of
axial stress after the stress-strain curve reaches its initial peak is
evident for some HSCFFT specimens as shown in Fig. 2(b). This
postpeak loss of strength is usually temporary and is followed
by strength recovery for specimens with adequate confinement.
Fig. 1. Technical illustration of test setup and instrumentation
This behavior is in agreement with that previously reported in

© ASCE 04013037-5 J. Compos. Constr.

J. Compos. Constr., 2014, 18(2): 04013037

 (a) (b)
Downloaded from ascelibrary.org by M S Ramaiah Institute of Technology on 06/25/18. Copyright ASCE. For personal use only; all rights reserved.

(c) (d)

Fig. 2. Representative axial stress-strain curves of CFFTs: (a) NSC CFRP; (b) HSC CFRP; (c) HSC AFRP; (d) HSC AFRP-FW

Lim and Ozbakkaloglu (2013), in which a model is provided to The influence of specimen size on the ultimate conditions of
determine the adequate confinement levels for FRP-confined HSC. adequately confined NSCFFTs and HSCFFTs was examined for
specimens featuring stress-strain curves with ascending branches,
and the results are presented in Table 4. This table presents a com-
Influence of Specimen Size parison of strength and strain enhancement coefficients (k1 and k2 ),
The influence of specimen size was investigated in this study by which were established from Eqs. (4) and (5), respectively:
comparing the axial performance of CFFTs of four different sizes. 0
f cc f
All four sizes maintained the same height to diameter ratio of 2∶1, 0 ¼ 1.0 þ k1 lu;a
0 ð4Þ
with the following nominal diameters: 75, 100, 150, and 300 mm. f co f co
To allow for a meaningful comparison between specimens of differ-
0 ) was kept constant
ent size, the nominal confinement ratio (flu =fco εcc f
¼ 1.0 þ k2 lu;a
0 ð5Þ
for the companion specimens with different diameters. However, εco f co
as noted previously, the hoop rupture strains (εh;rup ) reported at ul-
timate conditions are regularly lower than the ultimate tensile On the basis of the comparison given in Table 4, for both
strains of fibers reported by manufacturers (εfu ). Assuming a uni- strength and strain enhancement, specimens ranging in nominal
form confinement pressure for a circular section, the actual confine- diameters from 75–300 mm demonstrate a similar behavior. There-
0
ment ratio (f lu;a =fco ) can be calculated by using the following fore, the results in the present paper indicate that specimen size has
equation: no significant influence on the axial compressive behavior of
CFFTs independent of the concrete strength.
flu;a 2tf Ef εh;rup The influence of specimen size on the strain reduction factor
0 ¼ 0 ð3Þ
f co f co D (kε ) was assessed by comparing the values of kε for the specimens
presented in Fig. 3. This assessment reveals relatively close kε
Fig. 3 illustrates the influence of the specimen size on the axial values. For instance, Table 2 shows that specimens in the groups
stress-strain curve by comparing the specimens with similar actual of N-150C1L and N-300C2L had kε values of 0.64 and 0.70,
0
confinement ratios (flu;a =fco ). Because of the differences in re- respectively. Fig. 3 presents the same trend for HSC1 and HSC2
corded FRP hoop rupture strain (εh;rup ) and individual in-place specimens, in which a maximum difference in kε values between
concrete strength (f co 0 ), values of the actual confinement ratios
comparable specimens was 0.03. This comparison indicates that
0
(flu;a =fco ) differed slightly among the companion specimens. specimen size does not have a major influence on the strain reduc-
The comparison presented in Fig. 3 shows that small, medium, tion factor (kε ).
or large scale test specimens with similar actual confinement ratios
0 ) behave similarly in terms of the trend of the curves and
(flu;a =fco
Influence of Concrete Compressive Strength
ultimate conditions. This observation is consistent for comparisons
made among each strength range (i.e., NSC, HSC1, and HSC2) for The influence of concrete strength was investigated by comparing
specimens with nominal diameters ranging from 75–300 mm and the axial performance of CFFTs manufactured with three different
confined with CFRP or AFRP tubes. This finding is in agreement concrete strength ranges (i.e., NSC, HSC1, and HSC2). To main-
with those previously reported on the size effect of FRP-wrapped 0 ),
tain comparable values of the nominal confinement ratio (f lu =f co
NSC cylinders (Carey and Harries 2003; Wang and Wu 2011; the specimens of the present study were designed with FRP
Liang et al. 2012). layers adjusted relative to the concrete strength. Because of slight

© ASCE 04013037-6 J. Compos. Constr.

J. Compos. Constr., 2014, 18(2): 04013037

 (a) (b)
Downloaded from ascelibrary.org by M S Ramaiah Institute of Technology on 06/25/18. Copyright ASCE. For personal use only; all rights reserved.

(c) (d)

Fig. 3. Influence of specimen size on compressive behavior of CFFTs: (a) NSC CFRP group 1; (b) NSC CFRP group 2; (c) HSC1 AFRP; (d) HSC2
AFRP

Table 4. Influence of Specimen Size on Ultimate Condition of CFFTs differences between target and test day in-place concrete strengths
0
Specimen Strength Strain (fco ) and differences in recorded strain reduction factors (kε ), as
diameter, enhancement enhancement influenced by concrete strength, the values of the actual confine-
0
Group D (mm) coefficient, k1 coefficient, k2 ment ratio (flu;a =f co ) differed slightly within each group.
CFRP confined NSC 75 2.67 22.8 Fig. 4 illustrates the influence of the concrete strength on the
150 2.37 22.9 axial performance of the CFFT specimens, in which normalized
300 2.33 26.2 axial stress (fcc =fco 0 ) is plotted against normalized axial strain
AFRP confined HSC1 100 2.40 21.9 (εcc =εco ). Fig. 4 presents comparisons separately for groups of
150 2.47 19.4
specimens that have the same size, fiber type, and similar actual

(a) (b)

(c)

Fig. 4. Influence of concrete strength on stress-strain behavior of CFFTs: (a) 100-mm diameter AFRP specimens; (b) 100-mm diameter AFRP-FW
specimens; (c) 150-mm diameter CFRP specimens

© ASCE 04013037-7 J. Compos. Constr.

J. Compos. Constr., 2014, 18(2): 04013037

 Table 5. Influence of Concrete Strength on Ultimate Condition of CFFTs CFFTs. This comparison indicates that the strain reduction factor
Group Concrete type k1 k2 (kε ) decreases as the strength of concrete increases. An example of
this is a comparison of 150-mm CFRP specimens in Table 2, which
100-mm AFRP NSC 3.02 29.1
HSC1 2.40 20.1
shows that kε decreases from 0.84 to 0.50 for comparable speci-
HSC2 2.35 21.9 mens N-150C2L and H2-150C6L, respectively. This observation
100-mm AFRP FW NSC 3.91 38.9 is also apparent when examining companion 100-mm AFRP spec-
HSC1 3.22 25.6 imens, and it is in agreement with Ozbakkaloglu and Akin (2012),
HSC2 3.23 22.9 in which this observation was first reported. This outcome can also
150-mm CFRP NSC 2.37 22.9 be attributed to the change in the concrete brittleness and cracking
HSC1 1.80 23.6 patterns with a change in the concrete strength, as previously
HSC2 1.44 14.0 discussed. However, this trend is not as clear for CFFTs with
Downloaded from ascelibrary.org by M S Ramaiah Institute of Technology on 06/25/18. Copyright ASCE. For personal use only; all rights reserved.

HMCFRP or filament wound AFRP tubes, for which only limited

0
experimental results are available.
confinement ratios (flu;a =fco ) from each of the concrete strength
ranges (i.e., NSC, HSC1, and HSC2). Fig. 4 shows that, in general,
an increase in concrete compressive strength (fco0 ) leads to an over-
Influence of Fiber Type
all decrease in both the strength enhancement ratio (f cc 0 =f 0 ) and
co
Fig. 5 illustrates the influence of the fiber type on the axial stress-
strain enhancement ratio (εcu =εco ). This outcome can be attributed
strain behavior of CFFTs, comparing selected specimens with sim-
to the increased concrete brittleness with increasing concrete 0
strength, which alters the concrete crack patterns from heterogenic ilar actual confinement ratios (flu;a =f co ). This comparison reveals
microcracks to localized macrocracks. This change in the cracking substantial differences between second branch behaviors of com-
pattern results in a more rapid and uncontrolled expansion of HSC parable CFFTs. These differences cause only a slight difference on
compared with the progressive expansion of NSC, which gradually strength enhancement ratios. Conversely, as expected, significant
activates the confinement mechanism. As a result, a HSC confine- differences exist in the ultimate axial strain of the CFFTs, with
ment mechanism gets activated only after a significant amount of AFRP-confined CFFTs developing the highest ultimate strain
damage is sustained by the concrete. and HMCFRP-confined CFFTs developing the lowest. These ob-
Table 5 presents an assessment of the influence of concrete servations illustrate the significant correlation between the ultimate
strength on the ultimate conditions of CFFT specimens and com- rupture strain of fibers (εfu ) and the ultimate axial strain (εcu )
pares the values of k1 and k2 . Each comparison is performed for of CFFTs.
groups of specimens with the same size and fiber type and only Table 6 presents a comparison of the influence of the fiber type
specimens featuring ascending second branches are included. As on the ultimate conditions of the NSC-FFTs and HSC2-FFTs fea-
shown in Table 5, in general, the axial strength and strain enhance- turing ascending second branches. This table shows that similar
ment ratios (fcc 0 0
=fco and εcu =εco ) of CFFTs decrease as the values of k1 exist for comparable specimens confined by different
strength of concrete increases. fibers. In contrast, the comparisons of k2 values reveal a clear dif-
The influence of the concrete compressive strength on the strain ference between the strain enhancement ratios (εcu =εco ) of these
reduction factor kε was examined for comparable AFRP- and specimens, which supports the aforementioned observation regard-
CFRP-confined CFFTs. It was observed that there is a strong cor- ing the influence of the ultimate tensile strain of fibers (εfu ) on the
0
relation between kε and fco in both AFRP- and CFRP-confined ultimate axial strain (εcu ) of CFFTs.

(a) (b)

(c)

Fig. 5. Influence of fiber type on the compressive behavior of CFFTs: (a) NSC; (b) HSC1; (c) HSC2

© ASCE 04013037-8 J. Compos. Constr.

J. Compos. Constr., 2014, 18(2): 04013037

 Table 6. Influence of Fiber Type on Ultimate Condition of CFFTs Table 7. Influence of FRP Tube-Manufacturing Method on Ultimate
Condition of CFFTs
Group Fiber type Ef (GPa) k1 k2
Group Manufacturing method k1 k2
NSC AFRP 116 3.02 29.1
CFRP 240 2.46 23.3 NSC Automatic 3.91 38.9
HMCFRP 640 3.64 4.9 Manual 3.02 29.1
HSC1 AFRP 116 2.43 20.8 HSC1 Automatic 3.22 25.6
CFRP 240 2.27 14.1 Manual 2.34 20.1
HSC2 Automatic 3.23 22.9
Manual 2.40 21.9
The influence of fiber type on the strain reduction factor (kε )
was examined for specimens with similar actual confinement ratios
Downloaded from ascelibrary.org by M S Ramaiah Institute of Technology on 06/25/18. Copyright ASCE. For personal use only; all rights reserved.

0
(flu;a =fco ). Table 2 shows that the strain reduction factor (kε ) of filament wound specimens can be attributed to several factors,
decreases as the fiber elastic modulus (Ef ) increases. This is shown including better precision in fiber alignment, higher fiber vol-
when examining comparable NSC specimens (N-150C1L and ume fraction, a reduced void volume fraction, and reduced fiber
N-150HM1L), HSC1 specimens (H1-150A3L and H1-150C3L), waviness.
or HSC2 specimens (H2-150A6L and H2-150C6L). This compari- Table 7 presents a comparison of the influence of the
son indicates that specimens confined by HMCFRP developed the CFFT manufacturing method on the ultimate conditions of AFRP-
lowest kε values and those confined by AFRP developed the high- confined CFFTs with 100-mm diameters. This comparison indi-
est kε values. This suggests that the influence of the increased fiber cates that the tube-manufacturing method influences the ultimate
brittleness associated with an increase in fiber elastic modulus (Ef ) condition of CFFTs, with filament wound specimens exhibiting
resembles the aforementioned influence of the concrete brittleness higher values of both k1 and k2 .
on the same factor. The influence of CFFT manufacturing method on the strain re-
duction factor, kε , was examined for companion 100-mm diameter
specimens. Table 2 shows that kε values for NSCFFTs are greater
Influence of CFFT Manufacturing Method for manually prepared FRP tubes compared with automatically pre-
Fig. 6 shows the influence of the tube-manufacturing method and pared FRP tubes. Similar values of kε are reported for HSC1-FFTs
the normalized stress-strain curves of HSCFFTs that were prepared with different manufacturing methods, whereas the HSC2-FFTs
with either the automated filament winding technique or manual display the opposite trend to that observed for the NSCFFTs, with
wet layup technique. Although the properties of the aramid fibers automatically prepared FRP tubes developing higher kε values than
used in the fabrication of these CFFT systems were slightly differ- manually prepared ones. This comparison indicates that the CFFT
ent, each of the two companion CFFTs compared had very close manufacturing method has some influence on the strain reduction
0 0
actual confinement ratios (f lu;a =fco ). It is apparent in this compari- factor (kε ), which appears to be concrete strength (f co ) dependent.
son that the slope of the second branch is not influenced signifi- When assessing the influence of the manufacturing method, a
cantly by the manufacturing method. However, the second branch large number of additional variables exist and they may affect
is substantially longer for filament wound specimens, resulting the performance of a given method. Therefore, additional system-
in higher ultimate strengths and strains. The higher performance atic experimental studies are required to better understand the

(a) (b)

(c)

Fig. 6. Influence of tube manufacturing method on the stress-strain behavior of AFRP-confined CFFTs: (a) HSC1 specimens group 1; (b) HSC1
specimens group 2; (c) HSC2 specimens

© ASCE 04013037-9 J. Compos. Constr.

J. Compos. Constr., 2014, 18(2): 04013037

 influence of the FRP tube-manufacturing method on kε and the way Berthet, J. F., Ferrier, E., and Hamelin, P. (2005). “Compressive behaviour
0
this is influenced by fco. of concrete externally confined by composite jackets. Part A: Experi-
mental study.” Constr. Build. Mater., 19(3), 223–232.
Binici, B. (2005). “An analytical model for stress–strain behavior of con-
fined concrete.” Eng. Struct., 27(7), 1040–1051.
Conclusions
Bisby, L. A., Dent, A. J. S., and Green, M. F. (2005). “Comparison of
This paper has presented the results of an experimental study on the confinement models for fiber-reinforced polymer-wrapped concrete.”
ACI Struct. J., 102(1), 62–72.
behavior of circular NSC- and HSC-FFTs under axial compression.
Carey, S. A., and Harries, K. A. (2003). “Shape and ‘gap’ effects on
In this paper, the findings on the influences of the specimen size
the behavior of variably confined concrete.” Cem. Concr. Res., 33(6),
and type of fibers are in agreement with those reported in the ma- 881–890.
jority of the previous studies. The findings show that the specimen Cui, C., and Sheikh, A. (2010). “Experimental study of normal- and high-
size does not have a significant effect on the compressive behavior
Downloaded from ascelibrary.org by M S Ramaiah Institute of Technology on 06/25/18. Copyright ASCE. For personal use only; all rights reserved.

strength concrete confined with fiber-reinforced polymers.” J. Compos.

of CFFTs and a strong correlation exists between the ultimate Constr., 10.1061/(ASCE)CC.1943-5614.0000116, 553–561.
tensile strain (εfu ) of the fibers and the ultimate axial strain Davol, A., Burgueno, R., and Seible, F. (2001). “Flexural behavior of cir-
(εcu ) of the CFFTs. On the basis of the results reported in this paper, cular concrete filled FRP shells.” J. Struct. Eng., 10.1061/(ASCE)0733-
the following additional conclusions can be drawn: 9445(2001)127:7(810), 810–817.
• Adequately confined high-strength CFFTs can exhibit highly De Lorenzis, L., and Tepfers, R. (2003). “Comparative study of models
ductile behavior. However, the behavior of these CFFTs is on confinement of concrete cylinders with fiber reinforced polymer
highly sensitive to the level of confinement, and lightly confined composites.” J. Compos. Constr., 10.1061/(ASCE)1090-0268(2003)
HSCFFTs may not be able to maintain their load-carrying ca- 7:3(219), 219–237.
pacity after their initial peak strengths are attained. Eid, R., Roy, N., and Paultre, M. (2009). “Normal- and high-strength con-
0
• For a given nominal confinement ratio (flu =fco ), an increase in crete circular elements wrapped with FRP composites.” J. Compos.
0 Constr., 10.1061/(ASCE)1090-0268(2009)13:2(113), 113–124.
concrete compressive strength (fco ) and associated brittleness
leads to an overall decrease in the strength enhancement ratio Fam, A., Schnerch, D., and Rizkalla, S. (2005). “Rectangular filament-
0
(f cc 0
=fco ), the strain enhancement ratio (εcu =εco ), and the strain wound GFRP tubes filled with concrete under flexural and axial load-
ing: Experimental investigation.” J. Compos. Constr., 10.1061/(ASCE)
reduction factor (kε ).
1090-0268(2005)9:1(25), 25–33.
• A significant correlation exists between the fiber elastic modu-
Fam, A. Z., and Rizkalla, S. H. (2001). “Behavior of axially loaded
lus (Ef ) and the strain reduction factor (kε ). An increase in the concrete-filled circular fiber-reinforced polymer tubes.” ACI Struct. J.,
elastic modulus of fibers used to manufacture CFFTs results in a 98(3) 280–289.
decrease in the strain reduction factor (kε ). The increased fiber Hong, W. K., and Kim, H. C. (2004). “Behavior of concrete columns con-
brittleness associated with an increase in the fiber elastic mod- fined by carbon composite tubes.” Can. J. Civ. Eng., 31(2), 178–188.
ulus (Ef ) resembles the aforementioned influence of the con- Idris, Y., and Ozbakkaloglu, T. (2013). “Seismic behavior of square
crete brittleness on the same factor. high-strength concrete-filled FRP tube columns.” J. Compos. Constr.,
• FRP tubes manufactured automatically by using a filament 10.1061/(ASCE)CC.1943-5614.0000388.
winding technique provide greater strength and strain enhance- Ilki, A., and Kumbasar, N. (2003). “Compressive behaviour of carbon fibre
ment (fcc 0 =f 0 and ε =ε ) for CFFTs compared with tubes composite jacketed concrete with circular and non-circular cross-
co cu co
manufactured manually using a wet layup process. sections.” J. Earthquake Eng., 7(3), 381–406.
Ilki, A., Peker, O., Karamuk, E., Demir, C., and Kumbasar, N. (2008).
“FRP retrofit of low and medium strength circular and rectangular re-
Acknowledgments inforced concrete columns.” J. Mater. Civ. Eng., 10.1061/(ASCE)0899-
1561(2008)20:2(169), 169–188.
The authors wish to extend their gratitude to Ms. Chan, Kashyap, Lam, L., and Teng, J. G. (2003). “Design-oriented stress-strain model for
Staak, Verma, Wang, and Wang, and Messrs. Haw, Miller, Puan, FRP-confined concrete.” Constr. Build. Mater., 17(6–7), 471–489.
and Rohrlach, who performed the experimental procedures pre- Liang, M., Wu, Z. M., Ueda, T., Zheng, J. J., and Akogbe, R. (2012).
“Experiment and modeling on axial behavior of carbon fiber reinforced
sented in this paper. The authors also wish to thank Drs. Cagri
polymer confined concrete cylinders with different sizes.” J. Reinf.
Ayranci and Pierre Mertini, who manufactured the tubes of the
Plast. Compos., 31(6), 389–403.
filament wound CFFTs reported in this paper, at the University Lim, J. C., and Ozbakkaloglu, T. (2013). “Confinement model for
of Alberta. The experimental study presented in this paper is part FRP-confined high-strength concrete.” J. Compos. Constr., 10.1061/
of an ongoing experimental program at The University of Adelaide (ASCE)CC.1943-5614.0000376.
on FRP-concrete composite columns. Mandal, S., Hoskin, A., and Fam, A. (2005). “Influence of concrete
strength on confinement effectiveness of fiber-reinforced polymer
circular jackets.” ACI Struct. J., 102(3), 383–392.
Supplemental Data Mirmiran, A., Shahawy, M., Samaan, M., El Echary, H., Mastrapa, J. C.,
and Pico, O. (1998). “Effect of column parameters on FRP-confined
Figs. S1–S3 are available online in the ASCE Library (www concrete.” J.Compos.Constr.,10.1061/(ASCE)1090-0268(1998)2:4(175),
.ascelibrary.org). 175–185.
Mohamed, H., and Masmoudi, R. (2010). “Axial load capacity of concrete-
filled FRP tube columns: Experimental versus predictions.” J. Compos.
References Constr., 10.1061/(ASCE)CC.1943-5614.0000066, 231–243.
Ozbakkaloglu, T. (2013a). “Axial compressive behavior of square and rec-
Almusallam, T. H. (2007). “Behaviour of normal and high-strength con- tangular high-strength concrete-filled FRP tubes.” J. Compos. Constr.,
crete cylinders confined with e-glass/epoxy composite laminates.” 10.1061/(ASCE)CC.1943-5614.0000321, 151–161.
Compos. Part B, 38(5–6), 629–639. Ozbakkaloglu, T. (2013b). “Compressive behavior of concrete-filled FRP
ASTM. (2008). “Standard test method for tensile properties of polymer tube columns: Assessment of critical column parameters.” Eng. Struct.,
matrix composites materials.” D3039M-08, West Conshohocken, PA. 51, 151–161.

© ASCE 04013037-10 J. Compos. Constr.

J. Compos. Constr., 2014, 18(2): 04013037

 Ozbakkaloglu, T. (2013c). “Concrete-filled FRP tubes: Manufacture and Saatcioglu, M., Ozbakkaloglu, T., and Elnabelsy, G. (2008). “Seismic
testing of new forms designed for improved performance.” J. Compos. behavior and design of reinforced concrete columns confined with
Constr., 10.1061/(ASCE)CC.1943-5614.0000334, 280–291. FRP stay-in-place formwork.” ACI Special Publications, 257, 149–170.
Ozbakkaloglu, T., and Akin, E. (2012). “Behavior of FRP-confined normal- Seible, F., Burgueño, R., Abdallah, M. G., and Nuismer, R. (1996).
and high-strength concrete under cyclic axial compression.” J. Compos. “Development of advanced composite carbon shell systems for concrete
Constr., 10.1061/(ASCE)CC.1943-5614.0000273, 451–463. columns in seismic zones.” Proc., 11th World Conference on Earth-
Ozbakkaloglu, T., Lim, J. C., and Vincent, T. (2013). “FRP-confined quake Engineering, Elsevier Science, Oxford, UK, Paper No. 1375.
concrete in circular sections: Review and assessment of stress-strain Tasdemir, M. A., Tasdemir, C., Jefferson, A. D., Lydon, F. D., and Barr,
models.” Eng. Struct., 49, 1068–1088. B. I. G. (1998). “Evaluation of strains at peak stresses in concrete: A three-
Ozbakkaloglu, T., and Oehlers, D. J. (2008a). “Concrete-filled square and phase composite model approach.” Cem. Concr. Res., 20(4), 301–318.
rectangular FRP tubes under axial compression.” J. Compos. Constr., Wang, Y. F., and Wu, H. L. (2011). “Size effect of concrete short columns
10.1061/(ASCE)1090-0268(2008)12:4(469), 469–477. confined with aramid FRP jackets.” J. Compos. Constr., 10.1061/
Ozbakkaloglu, T., and Oehlers, D. J. (2008b). “Manufacture and testing of a
Downloaded from ascelibrary.org by M S Ramaiah Institute of Technology on 06/25/18. Copyright ASCE. For personal use only; all rights reserved.

(ASCE)CC.1943-5614.0000178, 535–544.
novel FRP tube confinement system.” Eng. Struct., 30(9), 2448–2459. Wei, Y. Y., and Wu, Y. F. (2012). “Unified stress-strain model of concrete
Ozbakkaloglu, T., and Saatcioglu, M. (2006). “Seismic behavior of high- for FRP-confined columns.” Constr. Build. Mater., 26(1), 381–392.
strength concrete columns confined by fiber reinforced polymer
Wu, H. L., Wang, Y. F., Yu, L., and Li, X. R. (2009). “Experimental and
tubes.” J. Compos. Constr., 10.1061/(ASCE)1090-0268(2006)10:6(538),
computational studies on high-strength concrete circular columns con-
538–549.
fined by aramid fiber-reinforced polymer sheets.” J. Compos. Constr.,
Ozbakkaloglu, T., and Saatcioglu, M. (2007). “Seismic performance of
10.1061/(ASCE)1090-0268(2009)13:2(125), 125–134.
square high-strength concrete columns in FRP stay-in-place formwork.”
J. Struct. Eng., 10.1061/(ASCE)0733-9445(2007)133:1(44), 44–56. Xiao, Q., Teng, J. G., and Yu, T. (2010). “Behavior and modeling of con-
Park, J. H., Jo, B. W., Yoon, S. J., and Park, S. K. (2011). “Experimental fined high-strength concrete.” J. Compos. Constr., 10.1061/(ASCE)CC
investigation on the structural behavior of concrete filled FRP tubes .1943-5614.0000070, 249–259.
with/without steel re-bar.” KSCE J. Civ. Eng., 15(2), 337–345. Yamakawa, T., Zhong, P., and Ohama, A. (2003). “Seismic performance of
Pessiki, S., Harries, K. A., Kestner, J., Sause, R., and Ricles, J. M. (2001). aramid fiber square tubed concrete columns with metallic and/or non-
“The axial behavior of concrete confined with fiber reinforced metallic reinforcement.” J. Reinf. Plast. Compos., 22(13) 1221–1237.
composite jackets.” J. Compos. Constr., 10.1061/(ASCE)1090-0268 Yazici, V., and Hadi, M. N. (2012). “Normalized confinement stiffness
(2001)5:4(237), 237–245. approach for modeling FRP-confined concrete.” J. Compos. Constr.,
Rousakis, T. C. (2001). “Experimental investigation of concrete cylinders 10.1061/(ASCE)CC.1943-5614.0000283, 520–528.
confined by carbon FRP sheets under monotonic and cyclic axial com- Youssef, M. N., Feng, M. Q., and Mosallam, A. S. (2007). “Stress-strain
pression load.” Research Rep. No. 01:2, Division of Building Technol- model for concrete confined by FRP composites.” Compos. Part B,
ogy, Chalmers Univ. of Technology, Göteborg, Sweden. 38(5–6), 614–628.
Saafi, M., Toutanji, H. A., and Li, Z. (1999). “Behavior of concrete col- Zhu, Z., Ahmad, I., and Mirmiran, A. (2006). “Seismic performance of
umns confined with fiber reinforced polymer tubes.” ACI Struct. J., concrete-filled FRP tube columns for bridge substructure.” J. Bridge
96(5), 500–509. Eng., 10.1061/(ASCE)1084-0702(2006)11:3(359), 359–370.

© ASCE 04013037-11 J. Compos. Constr.

J. Compos. Constr., 2014, 18(2): 04013037