You are on page 1of 16

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

net/publication/257389441

SIFCON strengthening of concrete cylinders in comparison with conventional


GFRP confinement method

Article  in  Construction and Building Materials · November 2012


DOI: 10.1016/j.conbuildmat.2012.06.021

CITATIONS READS

8 887

5 authors, including:

Bahareh Abdollahi Mehdi Bakhshi


University of Nevada, Reno Arizona State University
2 PUBLICATIONS   18 CITATIONS    29 PUBLICATIONS   261 CITATIONS   

SEE PROFILE SEE PROFILE

Zanyar Mirzaee Masoud Motavalli


ETH Zurich Empa - Swiss Federal Laboratories for Materials Science and Technology
7 PUBLICATIONS   21 CITATIONS    197 PUBLICATIONS   3,495 CITATIONS   

SEE PROFILE SEE PROFILE

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

Near-surface mounted, ribbed, iron-based shape memory alloy bars for prestressed strengthening View project

Bond behavior of CFRP-to-steel substrate View project

All content following this page was uploaded by Mehdi Bakhshi on 09 May 2018.

The user has requested enhancement of the downloaded file.


(This is a sample cover image for this issue. The actual cover is not yet available at this time.)

This article appeared in a journal published by Elsevier. The attached


copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy

Construction and Building Materials 36 (2012) 765–778

Contents lists available at SciVerse ScienceDirect

Construction and Building Materials


journal homepage: www.elsevier.com/locate/conbuildmat

SIFCON strengthening of concrete cylinders in comparison with conventional


GFRP confinement method
B. Abdollahi a, M. Bakhshi b,⇑, Z. Mirzaee c, M. Shekarchi a, M. Motavalli d
a
College of Civil Engineering, University of Tehran, Tehran, Iran
b
School of Sustainable Engineering and the Built Environment, Arizona State University, Tempe, AZ, USA
c
Department of Structural, Environmental and Geomatic Engineering (D-BAUG), ETH Zürich, Switzerland
d
Empa, Swiss Federal Laboratories for Materials Testing and Research, Structural Engineering Research Laboratory, Dübendorf, Switzerland

h i g h l i g h t s

" Performance of SIFCON jackets as a new confinement method was studied.


" SIFCON jackets were compared experimentally with FRP wraps as retrofit materials.
" Behavior models of GFRP-confined concrete were verified with experimental results.
" A constant passive type confinement behavior model for SIFCON jacket is proposed.

a r t i c l e i n f o a b s t r a c t

Article history: This paper presents SIFCON strengthening as a new method of confinement in comparison with GFRP
Received 9 June 2011 confinement. The experimental part of the work is focused on the investigation of confinement in con-
Received in revised form 27 May 2012 crete specimens with unconfined strength ranging from 15 to 40 MPa, enclosed with GFRP sheets or SIF-
Accepted 4 June 2012
CON jacket. Several parameters such as number of layers and ply configuration for GFRP method and
Available online xxxx
jacket strength, fiber length and jacket thickness for SIFCON method have been investigated. Further-
more, the ultimate strength models of FRP-confined concrete have been classified and evaluated with test
Keywords:
results and the first step in modeling the ultimate strength of SIFCON-confined concrete has been taken.
Composite
Concrete
The results show that SIFCON confinement method can be regarded as a competitive method with respect
Confinement to the well-known FRP confinement method and some of the available models of FRP confinement can
GFRP also give good estimation of the ultimate strength of concrete confined by SIFCON method.
SIFCON Ó 2012 Elsevier Ltd. All rights reserved.
Strengthening

1. Introduction Recently many types of High Performance Fiber Reinforced Ce-


ment (HPFRC) composites have been developed [11]. These mate-
Column rehabilitation is often critical to the seismic perfor- rials have high strength, ductility and toughness which make
mance of a structure. To prevent the column failure mechanism them ideal materials for structural repair and retrofit [12]. Among
during earthquakes, columns should never be the weakest compo- several types of HPFRC composites, Slurry Infiltrated Fiber CON-
nents in the whole structure [1]. Therefore, column jacketing may crete (SIFCON) is a type of concrete in which formwork molds
be used to improve flexural and shear strength as well as concrete are filled to capacity with fibers and then the resulting fiber net-
confinement so that columns are not damaged [2]. Steel and rein- work is infiltrated by a cement-based slurry [13]. Infiltration is
forced concrete jackets [3] and Fiber Reinforced Polymers (FRPs) usually accomplished by gravity flow aided by a light vibration
wraps [4–7] are often used for this purpose. Results of finite ele- or pressure grouting [14]. The volume fraction of fibers in SIFCON
ment analysis and full-scale tests show that the strength, ductility is 4–25% compared to 0.5–3% in conventional fiber reinforced con-
and the energy dissipation capacity of the rehabilitated columns crete. Infiltrated grout in SIFCON with very fine aggregate makes it
are enhanced greatly using these methods [8–10]. a ductile and a high performance material [15,16]. Elastic modulus
and maximum strain capacity as much as 13,800 MPa and 1.5%
were reported for SIFCON materials with 10% of steel fibers. Also,
⇑ Corresponding author. Tel.: +1 480 370 1685; fax: +1 480 965 0557. ultimate strain capacity of SIFCON can reach and even exceed the
E-mail address: mbakhshi@asu.edu (M. Bakhshi). yield strain of mild steel rebars [17].

0950-0618/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.conbuildmat.2012.06.021
Author's personal copy

766 B. Abdollahi et al. / Construction and Building Materials 36 (2012) 765–778

Superior properties of SIFCON have been the driving force of GFRP-confined concrete were verified with experimental results
several studies on this material in the field of strengthening and using confined concrete strength, fcc0 and strain at the maximum
rehabilitation of concrete structures. Results of a study done by stress, ecc [27]. Finally, a simple behavior model for SIFCON-con-
Elnono et al. [18] on SIFCON corner connections subjected to open- fined concrete is presented based on a constant passive type of
ing bending moments show that joint capacity and ductility were confinement and then compared with some FRP confinement
enhanced by as much as 66% and 173%, respectively. Abdou [19] models.
tested cast-in-place SIFCON joint as a connection between precast
beams under quasi-static load reversals and found an energy dissi- 2. Experimental program
pation up to five times higher than that of a reinforced concrete
joint. Results of studies on precast SIFCON flexural hinges by Wood 2.1. Scope of test program
[17] show that these hinges absorb 30% more energy than conven-
Unconfined concrete, SIFCON-confined concrete and FRP-confined concrete
tional fiber reinforced concrete hinges. Shannag et al. [20] used SIF- were tested under uniaxial compression test. The scope of the test program is
CON jackets in the structural repair of shear-deficient reinforced shown in Table 1. Two replicate specimens were tested for each test series. Strength
concrete beams to eliminate the brittle shear failure and increase of unconfined concrete cylinders was the common variable in both FRP and SIFCON
the ultimate shear strength of the repaired beams. Naaman and confinement methods in order to study confinement efficiency of different concrete
cylinders. Three strength levels of 15, 25 and 40 MPa were studied for this purpose.
Baccouche [15] showed that shear strength and energy absorption
150 mm  300 mm cylinders were confined by different FRP wraps and SIFCON
capacity of dowel reinforced SIFCON under monotonic and re- jackets. For FRP-confined cylinders, effects of different variables including number
versed cyclic loading is three times and 5–12 times higher than of GFRP layers and GFRP ply configuration were studied. Three GFRP layer numbers
that of reinforced concrete. Dogan and Krstulovic [21] showed that of 1, 2 and 4, and different fiber orientations of 0°, 45° and 90° with respect to the
use of SIFCON in the anchorage zone of the beam–bottom rein- hoop direction were used. On the other hand, effects of another set of variables
including, jacket thickness, fiber length and jacket strength were studied for SIF-
forcement eliminated reinforcement pull-out failure. This result CON-confined cylinders. Jacket thicknesses of 10, 15 and 20 mm, fiber lengths of
is in accordance with the results of Hamza and Naaman [22] who 30 and 50 mm and jacket strengths of 85 and 38 MPa were used as SIFCON jacket
observed the total pull-out work for debonding steel bars embed- variables.
ded in SIFCON is more than twenty times than that in plain con-
crete. Sharma and Singh [23], and Yazici et al. [24] found 2.2. Materials, mixing, placing and curing procedures
strength and ductility related properties of SIFCON plates and
2.2.1. Unconfined concrete
beams improved drastically due to multicrack behavior of the
The materials used in this study for casting unconfined concrete cylinders were
material. Among few studies performed on column strengthening Portland cement type II, aggregates with maximum size of 12.5 mm and superplast-
by HPFRC, Coskun [25] studied axial load–moment interaction icizer with polycarboxilate base. As shown in Table 2, three different mixture pro-
and moment–curvature diagrams of HPFRC-confined reinforced portions (C1, C2 and C3) were used to produce low, normal and higher strength
concrete columns. However, his works were limited to analytical concrete with cement content of 200, 250 and 400 kg/m3 and w/c ratios of 0.6,
0.5 and 0.35, respectively. After making unconfined cylinders, specimens were cov-
works and the enhancement in concrete confinement was not ered with a wet burlap cloth for 1 day, before demolding and curing in the water at
the focus of his study. the room temperature. The specimens needed for the GFRP and SIFCON confine-
In this paper, effect of SIFCON jackets in enhancing concrete ment were first cured for 7 days before applying jackets, while concrete cores were
confinement and increasing compressive strength of concrete cyl- cured for 28 days and tested immediately.
inders were studied. SIFCON jackets were compared with FRP
wraps as reliable retrofit materials for the confinement of rein- 2.2.2. SIFCON confinement
SIFCON jacket was made by placement of the unconfined concrete cylinder at
forced concrete elements [26]. Experimental investigation of
the center of a 170  300, 180  300, or 190  300 mm cylinder mold. First, the
Glass-FRP (GFRP) wraps included effects of unconfined concrete mold was filled to capacity by steel fibers. 30 mm or 50 mm long steel fibers with
(concrete core) strength level and fiber specifications on the a round cross section and 0.3 mm diameter with a tensile strength of 950 MPa were
stress–strain curves. In order to study SIFCON jackets, parameters used for fiber reinforcement. The resulting fiber content was approximately 5% and
such as jacket thickness and strength, fiber type, and unconfined 6% volumetric fractions for longer and shorter fibers, respectively. The fiber network
was then infiltrated by low and high strength slurries using a light vibration. The
concrete strength level were studied. The compressive test results
procedure was done in four stages along the height of specimens from bottom to
of GFRP-confined cylinders were compared with the test results of top. To ensure an even dispersion of fibers at each stage, the volume content of
SIFCON-confined cylinders. In addition, the behavior models of fibers was controlled in a way that the same amount of fibers was used for each

Table 1
Scope of the test program.

Sample type Test Unconfined concrete SIFCON jacket SIFCON fiber SIFCON slurry FRP layers FRP ply
series strength (MPa) thickness (mm) length (mm) strength (MPa) no. configuration
Unconfined concrete C1 15 N.A. N.A. N.A. N.A. N.A.
cylinder C2 25 N.A. N.A. N.A. N.A. N.A.
C3 40 N.A. N.A. N.A. N.A. N.A.
SIFCON-confined- S1 15 15 30 85 N.A. N.A.
cylinder S2 25 15 30 85 N.A. N.A.
S3 40 15 30 85 N.A. N.A.
S4 25 20 30 85 N.A. N.A.
S5 25 10 30 85 N.A. N.A.
S6 25 15 50 85 N.A. N.A.
S7 25 15 30 38 N.A. N.A.
FRP-confined cylinder F1 15 N.A. N.A. N.A. 1 0°
F2 25 N.A. N.A. N.A. 1 0°
F3 40 N.A. N.A. N.A. 1 0°
F4 25 N.A. N.A. N.A. 2 0°/0°
F5 25 N.A. N.A. N.A. 4 0°/0°/0°/0°
F6 25 N.A. N.A. N.A. 2 90°/0°
F7 25 N.A. N.A. N.A. 2 45°/+45°
Author's personal copy

B. Abdollahi et al. / Construction and Building Materials 36 (2012) 765–778 767

Table 2
Mix proportions of unconfined concrete and SIFCON jacket.

Test series Portland cement Silica fume Water Coarse aggregate Fine aggregate SP (%) w/c
C1 200 – 120 832 1248 – 0.6
C2 250 – 125 815 1213 – 0.5
C3 400 – 140 741 1111 2 0.3
S1–S6 jacket 900 270 350 – 700 3.5 0.3
S7 jacket 830 – 500 – 830 – 0.6

quarter of specimens. In order to make slurry, Portland cement type II, water, Poly- failure mode was observed in most of GFRP-confined samples
carboxilate based superplasticizer, silica fume and silica sand passing sieve No. 30
which exhibited failure due to both hoop fiber rupture and delam-
were used. Mixture proportions for low and high strength slurries, as presented in
Table 2, resulted in measured compressive strength of 39 and 81 MPa, respectively.
ination, with intact concrete and FRP jacket above and below the
After casting SIFCON jacket around the concrete cylinders, specimens were steam failure region. These results are in agreement with the results of
cured for 1 day in a room with 98% relative humidity and 50 °C temperature. SIF- Au and Buyukozturk [30] who investigated effect of fiber orienta-
CON-confined specimens were then stripped and cured in the water at the room tion on FRP-confined concrete. Effects of studied parameters on
temperature for 21 days before testing.
the results of confined samples are addressed in the following
sections.
2.2.3. GFRP confinement
For the FRP wrapped specimens, glass FRP sheets of SikaWrap Hex 430G were
used. The tensile strength, tensile modulus, tensile elongation and layer thickness 3.1. Effects of different parameters on SIFCON jacket performance
of the sheets are 537 MPa, 26493 MPa, 2.21% and 0.508 mm, respectively. Concrete
cylinders were wrapped by GFRP sheets with 0° fiber orientation with respect to the
As shown in Fig. 3, results of S1, S2 and S3 tests series were
hoop direction, except in two cases that ply configurations were 0°/90° and 45°/
+45°. Throughout the study, three sets of number of layers including 1, 2 and 4 lay- compared to determine SIFCON confinement efficiency on low-,
ers of GFRP sheets were used, resulting in the jacket thickness equal to 0.508, 1.016 mid- and higher-strength concrete cores. The increase in the com-
and 2.032 mm, respectively. GFRP jackets consisted of discrete wraps with one pressive strength of SIFCON-confined with respect to unconfined
overlapping splice length of 100 mm for each layer (1/4–1/5 of the perimeter). cylinders was 104%, 127% and 54% by increasing the core strength
For jackets made of more than one layer of GFRP composites, splices were located
from 15 to 25 and 40 MPa, respectively. Increasing core strength
at different points along the perimeter. The application of the composite sheets re-
quired a coat of a two-component epoxy resin (primer) on the concrete surface. also resulted in 219%, 667% and 319% increase in the ductility
After curing the primer, the composite sheets were glued using a two-component and 19.29, 27.00 and 19.51 times increase in energy absorption
epoxy resin of Sikadur 330. The specimens were left in the laboratory for 21 days capacity. These results indicate that this technique is more effec-
to be completely cured before testing.
tive for strengthening of mid-strength concrete (25 MPa), followed
by low- and higher-strength concrete. On the other hand, results of
2.3. Testing procedures
S5, S2 and S4 samples show that the increase in jacket thickness
All specimens were tested by an MTS machine under displacement control at a from 10 to 15 and 20 mm translated into 69%, 127% and 195% in-
rate of 1 mm/min, which corresponds to a strain rate of 0.003 per min. Uniaxial crease in the compressive strength of cylinders with respect to
compression loads were applied on total concrete area of unconfined cylinders, their unconfined strengths. These results are shown in Fig. 4. The
GFRP-confined and SIFCON-confined cylinders using a rigid plate to transfer the
same trend was observed for the ductility which increased by
load from the loading platens only to the core concrete. Loads and displacements
between the loading platens were used to obtain compressive stress–strain curves. 294%, 667% and 917% and the dissipation energy which increased
The load was divided by the loading area to obtain the axial stress. Hydraulic piston by 15, 27 and 37 times. These results show proportional relation-
travel was used for displacement measurements and therefore, effect of machine ships of strength, ductility and toughness with increasing SIFCON
compliance has to be removed from recorded displacements to obtain actual sam- jacket thickness from 10 to 15 and 20 mm.
ple displacements. This was done based on a relationship between the apparent
compliance and machine compliance presented by Kalidindi et al. [28]. Once cross
Results of S2 and S6 test series were compared in Fig. 5a to cap-
sectional area, length of samples, elasticity modulus calculated by the ACI code [29], ture effect of fiber length. Confinement by SIFON jackets with fiber
and apparent compliance of plain concrete samples are known, the machine com- lengths of 30 and 50 mm resulted in the strength enhancement of
pliance can be readily calculated. The values of machine compliance were concrete core by 127% and 100%, and ductility enhancement by
5.59  107, 3.55  107 and 4.75  107 mm/N for low, medium and higher
667% and 722%. The toughness enhancement of both test series
strength concrete in this study, which leads to an overall average machine compli-
ance of 4.63  107 mm/N. The machine compliance was used to obtain actual sam- is very similar. These values exhibited no significant fiber length
ple displacements. Finally, corrected displacement data were divided by the initial effect on the SIFCON jacket performance. However, effect of slurry
length of specimens as the gauge length to obtain axial strains. strength on compressive response evaluated by comparison of S2
and S7 test series was significant. As shown in Table 5b, 127%
3. Test results and discussion and 35% increase in compressive strength of S2 and S7, and 667%
and 511% increase in ductility for these samples are reported.
The stress–strain responses of GFRP-confined, SIFCON-confined Two times energy absorption enhancement for S2 comparing to
and unconfined cylinders are shown in Fig. 1. The stress–strain re- S7 samples indicates a much better confinement performance for
sponse of SIFCON-confined concrete shows a strain softening higher-strength jacket comparing to mid-strength jacket.
behavior after peak stress, while the GFRP-confined concrete dem-
onstrates a plateau response or a strain hardening with a distinct 3.2. Effects of different parameters on FRP wrap performance
bilinear behavior in which a reduction in stiffness is experienced
after reaching an axial stress higher than the unconfined strength. For determining effect of concrete core strength, F1, F2 and F3
However, the results show that both confinement methods signif- samples confined with 1 layer of GFRP with 0° fiber orientation
icantly enhanced concrete properties by increasing strength, duc- with respect to the hoop direction were studied. As shown in
tility and the energy dissipation capacity as shown in Tables 3 Fig. 6 and Table 4, the ratio of confined strength to unconfined
and 4. Fig. 2 shows confined specimens before and after testing. strength was 1.87, 1.42 and 1.24 for low-, mid- and higher-
Experimental observations indicate that the SIFCON-confined strength cores, respectively. Similar trend were observed for duc-
cylinders fail due to tensile failure of the jacket along with surface tility and energy dissipation capacity parameters. Ratio of confined
delaminations between the core concrete and the jacket. Similar to unconfined peak strain decreased from 6.90 to 6.56 and 6.12,
Author's personal copy

768 B. Abdollahi et al. / Construction and Building Materials 36 (2012) 765–778

100
(a) (b) 100
F7
S7
F6
S6
80 80 F5
S5
F4
S4
Axial Stress, MPa

Axial Stress, MPa


F3
S3
60 60 F2
S2
F1
S1

40 40

20 20

0 0
0 1 2 3 4 5 0 1 2 3 4 5
Axial Strain, % Axial Strain, %

(c) 50
C3
C2
40 C1
Axial Stress, MPa

30

20

10

0
0 0.2 0.4 0.6 0.8
Axial Strain, %

Fig. 1. Compressive stress–strain responses of: (a) GFRP-confined, (b) SIFCON-confined, and (c) unconfined cylinders.

and energy dissipation ratio decreased from 12.71 to 10.97 and the specimens with 45°/+45° wrap were higher than the speci-
6.12 by increasing concrete core strength from 15 to 25 and mens with 0°/90° wrap.
40 MPa, respectively. These results indicate that the efficiency of
FRP wrap in confining concrete columns decreases by increasing
3.3. Comparison of SIFCON and GFRP confinement methods
core concrete strength. On the other hand, effect of different num-
bers of confining GFRP layers were evaluated by studying F1, F4
The stress–strain responses of GFRP-confined concrete speci-
and F5 samples. As shown in Fig. 7, increases in the number of
mens show bilinear forms before starting failure, while SIFCON
GFRP layers from 1 to 2 and 4 layers translated into 87%, 121%
curves show nonlinear behaviors throughout the test. While failure
and 233% strength increase compared to the unconfined speci-
in GFRP-confined cylinders was diagnosed with a drastic reduction
mens. The same trends were observed for the ductility which in-
in the stress level, such significant stress decay was not observed in
creased by 590%, 917% and 1141% and absorbed energy which
SIFCON-confined cylinder. Therefore, the ultimate strain was not
increased by 13, 25 and 48 times.
used as a comparative parameter between these two confinement
Effect of GFRP wrap fiber orientation on the confinement perfor-
methods. In the case of GFRP-confined cylinders, the ultimate
mance was examined using three sets of fiber orientation of 0°/0°,
strain can be obtained from the stress–strain diagram. However,
0°/90° and 45°/+45° on 2-layer wrapped specimens. Compressive
the ultimate strain of SIFCON-confined specimens is calculated
responses of F4, F6 and F7 samples are shown in Fig. 8. The F4 sam-
based on the strain corresponding to 85% of peak strength [31].
ple with 0°/0° wrap is sufficiently confined and the second part of
The energy absorption capacity of SIFCON specimens can be mea-
the stress–strain curve is ascending, while this ascending part
sured based on the area under the stress–strain diagram up to a
shifts to a plateau in F6 and F7 samples. This different behavior
specified strain equal to 2%. These two methods were compared
of F4 sample with ply orientation of 0°/0° resulted in 121%, 917%
in three different categories based on the strength level of the
and 2435% increase with respect to the unconfined specimen in
unconfined concrete cylinder.
maximum stress, maximum strain and energy absorption parame-
ters, respectively. This increase was 52%, 444% and 862% for the 0°/
90° ply configuration and 34%, 867% and 1492% for the 45°/+45° 3.3.1. Comparisons of confinement methods for low-strength concrete
ply configuration, respectively. The 0°/90° wrap provided a higher cores
confined strength than the wrap with 45°/+45° ply configuration. Stress–strain diagrams and comparative results of confined
However, the ultimate confined strain and the absorbed energy for concrete cylinders of F1 and S1 series are shown in Fig. 9. As shown
Author's personal copy

B. Abdollahi et al. / Construction and Building Materials 36 (2012) 765–778 769

Table 3
Material properties of all unconfined and confined specimens.

Test series Peak strain, ecc (%) Maximum stress (MPa) Toughness, Ucc (kJ/m3)
C1-A 0.22 15.40a 27
C1-B 0.15 14.49a 16
C2-A 0.17 24.84a 32
C2-B 0.18 25.71a 37
C3-A 0.16 42.75a 42
C3-B 0.16 40.77a 42
S1-A 0.60 33.11b 493
S1-B 1.03 27.66b 435
S2-A 1.06 58.98b 896
S2-B 1.59 58.73b 938
S3-A 0.64 66.20b 873
S3-B 0.68 62.28b 803
S4-A 1.85 81.87b 1356
S4-B 1.60 72.02b 1173
S5-A 0.72 41.23b 488
S5-B 0.70 44.57b 553
S6-A 1.32 55.86b 903
S6-B 1.59 47.56b 781
S7-A 1.17 34.33b 515
S7-B 0.98 34.21b 450
F1-A 1.75 30.09b 404
F1-B 1.35 27.26b 268
F2-A 1.18 36.28b 383
F2-B 1.38 35.96b 434
F3-A 1.03 56.63b 411
F3-B 0.97 49.17b 387
F4-A 1.89 53.97b 825
F4-B 1.81 57.66b 851
F5-A 2.72 89.53b 1700
F5-B 2.82 79.61b 1671
F6-A 0.98 37.33b 322
F6-B 1.13 40.28b 392
F7-A 1.74 33.04b 330
F7-B 1.80 34.99b 345
a 0
The values are considered as fco parameters.
b
The values are considered as fcc0 parameters.

Table 4
Average material properties of both unconfined and confined specimens.

Test series Peak strain, ecc (%) Maximum stress (MPa) Toughness, Ucc (kJ/m3)
C1 0.21 14.82a 24
C2 0.18 25.20a 34
C3 0.16 41.69a 43
S1 0.67 30.25b 463
S2 1.38 57.19b 918 Fig. 2. Concrete specimens before and after testing confined by: (a) FRP wrap and
S3 0.67 64.22b 839 (b) SIFCON jacket.
S4 1.83 74.27b 1258
S5 0.71 42.72b 521
S6 1.48 50.56b 842
S7 1.10 34.25b 483
3.3.2. Comparisons of confinement methods for mid-strength concrete
F1 1.45 27.67b 305
F2 1.18 35.87b 373 cores
F3 0.98 51.51b 392 Behavior of S5 sample which is a 25 MPa concrete core confined
F4 1.83 55.59b 828 with 10 mm thick SIFCON jacket with the slurry strength of 85 MPa
F5 2.72 84.02b 1648 and fiber length of 30 mm is similar to F2 sample which is confined
F6 0.98 38.48b 327
F7 1.74 33.90b 541
with 1 layer of GFRP wrap with 0° fiber orientation. As shown in
Table 5 and Fig. 10a, the confinement in both cases led to similar
a 0
The values are considered as fco parameters. increases in compressive strength and absorbed energy. The ductil-
b
The values are considered as fcc0 parameters.
ity parameter however shows a higher value for GFRP wrapped
sample.
For concrete cores with the average strength of 25 MPa, a
15 mm thick SIFCON jacket such as S2 jacket enhanced the com-
in Table 5, the increase in compressive strength and energy pressive behavior considerably more than one layer of GFRP wrap.
absorption capacity of these two samples due to confinement are As shown in Fig. 10b and Table 5, similar compressive responses
similar to each other. However, the ratio of strain at maximum were obtained for S2 and F4 cylinders with two GFRP layers with
stress of confined to unconfined cylinder for GFRP wrapped sam- 0° fiber orientation. While the ratios of confined to unconfined
ples are as twice as the one for SIFCON jacketed samples. strength and energy absorption parameters for both S2 and F4
Author's personal copy

770 B. Abdollahi et al. / Construction and Building Materials 36 (2012) 765–778

100
(a) 100
S3 (Confined, f'co: 40 MPa)
S2 (Confined, f'co: 25 MPa)
80 S1 (Confined, f'co: 15 MPa) 80
S6 (Confined, Fiber Length: 50 mm)
C3 (Unconfined) S2 (Confined, Fiber Length: 30 mm)
Axial Stress, MPa

C2 (Unconfined)

Axial Stress, MPa


C2 (Unconfined)
60 C1 (Unconfined) 60 Replicates
Replicates Jacket Thickness: 15 mm
f'co : 25 MPa
40 40 Slurry Strength: 85 MPa

20 20

0 0
0 1 2 3 4 5 0 1 2 3 4 5
Axial Strain, % Axial Strain, %

Fig. 3. Effect of concrete core strength level on the compressive stress–strain


response of SIFCON-confined cylinders.
(b) 100

S2 (Confined, Slurry Strength:85MPa)


80
100 S7 (Confined, Slurry Strength:38MPa)
S4 (Jacket Thickness: 20 mm) (Unconfined)
Axial Stress, MPa

S2 (Jacket Thickness: 15 mm) Replicates


S5 (Jacket Thickness: 10 mm) 60
80 Jacket Thickness: 15 mm
Unconfined f'co : 25 MPa
(f'co : 25 MPa) Fiber Length: 30 mm
Axial Stress, MPa

40
60
Replicates

20
40

0
20 0 1 2 3 4 5
Axial Strain, %

Fig. 5. Effect of fiber length and slurry strength of SIFCON jacket on compressive
0
0 1 2 3 4 5 response of confined cylinders.

Axial Strain, %

Fig. 4. Effect of jacket thickness on compressive stress–strain response of SIFCON-


confined cylinders.

60
samples are in the range of 2.2–2.3 and 24–27, the ratio of ultimate F3 (Confined, f'co: 40 MPa)
strain for these two samples compared to unconfined sample are F2 (Confined, f'co: 25 MPa)
also in the close range of 8–10. F1 (Confined, f'co: 15 MPa)
Stress–strain diagrams and comparative results of 25 MPa con-
C3 (Unconfined)
crete cores confined with 20 mm SIFCON jacket identified by S4
Axial Stress, MPa

40
C2 (Unconfined)
can be compared with F5 series which refers to the confinement
C1 (Unconfined)
by 4 GFRP layers in direction of the perimeter. The comparison is
Replicates
shown in Fig. 10c and Table 5. The results show that strength
enhancement is about 3 times for both samples. The ultimate
strain and energy dissipation capacity are in the range of 10–15
20
and 37–48 times the unconfined, with the advantage for GFRP-con-
fined cylinders.

3.3.3. Comparisons of confinement methods for higher-strength


concrete cores 0
0 1 2 3 4 5
In order to investigate the confinement effect on high strength
Axial Strain, %
concrete cores, two types of confinements including S3 referring
to a 15 mm thick SIFCON jacket and F3 for a single layer GFRP Fig. 6. Effect of unconfined concrete strength on the behavior of GFRP confined
wrapped sample were compared. As shown in Fig. 11 and Table 5, concrete cylinders.
Author's personal copy

B. Abdollahi et al. / Construction and Building Materials 36 (2012) 765–778 771

120 ultimate stress enhancement using GFRP wraps was 1.24 which is
F5 (4 Layers) Unconfined less than 1.54 for SIFCON-confined sample. In this strength level,
F4 (2 Layers) (f'co : 25 MPa) the general form of the stress–strain curve of GFRP wrapped con-
100
F2 (1 Layer) crete cores is similar to SIFCON-confined samples considering a
Replicates descending part after initial elastic range. This behavior may have
happened due to insufficient confinement of the higher strength
Axial Stress, MPa

80
concrete cores using a single GFRP layer. The ultimate strain at
the maximum stress enhanced by 6.12 and 4.19 times for GFRP
60 and SIFCON-confined samples, respectively. However, the
enhancement of energy dissipation for SIFCON jacket is twice as
GFRP wrap.
40

4. Confinement mechanisms and models


20
4.1. Confinement mechanisms
0
0 1 2 3 4 5 For the concrete columns with steel reinforcement or external
Axial Strain, % jackets, the nature of confinement is of passive type. In other
words, the confining pressure is induced by the transverse dilation
Fig. 7. Effect of number of GFRP layers on the behavior of GFRP-confined concrete of concrete or the Poisson effect. In some cases of passive confine-
cylinders. ment, an initial active confining pressure may also be present, for
instance by injecting an expansive grout between a column and
an external jacket. However, the induced active pressure is small
100
relative to the additional passive confinement engaged by the con-
F4 (Confined, Ply Orientation: 0o/0o) crete dilation [32,33].
F6 (Confined, Ply Orientation: 90o/0o) The constant passive confinement is induced in case the confin-
80
F7 (Confined, Ply Orientation: -45o/+45o) ing material behaves in the plastic range, such as the one provided
C2 (Unconfined) by transverse steel reinforcement. On the other hand, variable con-
Axial Stress, MPa

Replicates fining pressure is provided in case the confining material has a con-
60
siderable stiffness. FRP jackets and steel reinforcement behaving in
Number of Layers: 2
f'co: 25 MPa an elastic manner provide a variable passive confinement depen-
dent on the axial and transverse concrete behavior and the con-
40 finement level [34]. For FRP-confined concrete under the axial
loading, the lateral expansion is partially restrained by FRP, which
results in a radial pressure at the concrete and FRP interface
20 depending on FRP stiffness in the circumferential direction. By
increasing the load, the lateral expansion and thus, the confining
pressure also increase resulting in a variable confinement [35].
0 HPFRC composite jackets with high stiffness values fall in this cat-
0 1 2 3 4 5
egory as well.
Axial Strain, %

Fig. 8. Effect fiber orientation on the behavior of GFRP-confined concrete cylinders. 4.2. Confinement models

The proposed behavior models for variable passive confinement


50 can be classified into two general categories: design-oriented mod-
els, and analysis-oriented models. In the first category, the com-
S1 (SIFCON Confined) pressive strength, ultimate axial strain and stress–strain behavior
F1 (GFRP Confined, Orientation: 0 o) of FRP-confined concrete are predicted using closed-form equa-
40
C1 (Unconfined, f'co: 15 MPa) tions directly based on the interpretation of the experimental re-
Replicates sults. In the second category, an active confinement method for
Axial Stress, MPa

concrete is used to evaluate the axial stress and strain of passively


30 SIFCON Properties
confined concrete at a given confining pressure. The interaction be-
Jacket Thickness: 15 mm
Fiber Length: 30 mm
tween concrete and confining material is explicitly accounted for
Slurry Strength: 85 MPa by equilibrium and radial displacement compatibility consider-
20 ations [36].
Many existing models have followed the idea of Richart et al.
[37], originally proposed for the strength of concrete confined by
10 lateral steel reinforcement, to predict the strength of FRP-confined
concrete. The general form of these models is presented in Eq. (1).
  
fl
0 fcc0 ¼ fco0 1 þ kc 0 ð1Þ
0 1 2 3 4 5 fco
Axial Strain, %
where fcc0 is the confined concrete strength, fco0 is the unconfined
Fig. 9. Comparison of stress–strain responses of a 15 MPa concrete core confined concrete strength, fl is the lateral confining pressure and kc is the
with the SFICON jacket and the GFRP wrap. confinement coefficient.
Author's personal copy

772 B. Abdollahi et al. / Construction and Building Materials 36 (2012) 765–778

Table 5
Comparison of confined parameters for SIFCON-and GFRP-confined cylinders.

Comparisons between GFRP and SIFCON Test series 0


fcc (MPa) fcc0 =fco
0 ecc (%) ecc/eco Ucc (kJ/m3) Ucc/Uco

Low-strength concrete core


F1 27.7 1.87 1.45 6.90 305 12.71
S1 30.2 2.04 0.67 3.19 463 19.29
Mid-strength concrete core
Low thickness jacket F2 35.9 1.42 1.18 6.56 373 10.97
S5 42.7 1.69 0.71 3.94 521 15.32
Medium thickness jacket F4 55.6 2.21 1.83 10.17 828 24.35
S2 57.2 2.27 1.38 7.67 918 27.00
High thickness jacket F5 84.0 3.33 2.72 15.11 1648 48.47
S4 74.3 2.95 1.83 10.17 1258 37.00
High-strength concrete core
F3 51.5 1.24 0.98 6.12 392 9.12
S3 64.2 1.54 0.67 4.19 834 19.51

(a) 100 (b) 100


S2 (SIFCON Confined)
F4 (GFRP Confined, Orientation: 0o/0o)
80 80 C2 (Unconfined, f'co: 25 MPa)
S5 (SIFCON Confined) Replicates
Axial Stress, MPa

F2 (GFRP Confined, Orientation: 0 o)


Axial Stress, MPa

C2 (Unconfined, f'co : 25 MPa) SIFCON Properties


60 60
Replicates Jacket Thickness: 15 mm
Fiber Length: 30 mm
SIFCON Properties Slurry Strength: 85 MPa
40 Jacket Thickness: 10 mm 40
Fiber Length: 30 mm
Slurry Strength: 85 MPa

20 20

0 0
0 1 2 3 4 5 0 1 2 3 4 5
Axial Strain, % Axial Strain, %

160
(c)
SIFCON Jacket Thickness: 20 mm
SIFCON Fiber Length: 30 mm
SIFCON Slurry Strength: 85 MPa
120 S4 (SIFCON Confined)
F5 (GFRP, Orientation: 0 o/0o/0o/0o)
Axial Stress, MPa

C2 (Unconfined, f'co: 25 MPa)


Replicates
80

40

0
0 1 2 3 4 5
Axial Strain, %

Fig. 10. Comparison of stress–strain responses of a 25 MPa concrete core confined with SFICON jackets with different thicknesses and GFRP wraps.

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi !
Richart et al. [37] suggested kc as 4.1 for the confinement by lat- fcc0 fl fl
eral steel reinforcement. Other researchers [27,38–46] suggested ¼ 2:254 1 þ 7:94 0  2 0  1:254 ð2Þ
fco0 fco fco
several different confinement coefficients for FRP-confined con-
crete. As shown in Table 6, kc was proposed either as a constant va-
lue, as an exponential function of the ratio of the lateral confining In some FRP confinement models [35,48,49], the confinement mod-
pressure to the unconfined concrete strength, or as an exponential el of Mander et al. [47] with some modifications were used to pre-
function of the lateral confining pressure. dict the strength of confined concrete by analysis-oriented models.
Mander et al. [47] proposed a model of confinement by lateral A summary of these models are presented in Table 7.
steel reinforcement based on a constant confining pressure accord- Other available confinement models as shown in Table 8
ing to Eq. (2). [40,48,50–52] are not directly based on the Richart or Mander
Author's personal copy

B. Abdollahi et al. / Construction and Building Materials 36 (2012) 765–778 773

100 Table 7
Confinement models of FRP-confined concrete based on Mander model [35,48,49].
S3 (SIFCON Confined)
Confinement model Confined concrete strength
F3 (GFRP Confined, Orientation: 0 o)
80 Spoelstra and Monti (I) [48] h i1EEsec
C3 (Unconfined, f'co: 40 MPa) fcc0 ¼ Esec;u ecc
Esec ðEc Esec;u Þ c
Esec;u ðEc Esec Þ
Replicates Esec;u ¼ Eco
Axial Stress, MPa

1þ2ðbf
Ecom Þ
com

0
60 SIFCON Properties Esec ¼ efcc
cc
h  i
Jacket Thickness: 15 mm 0
fcc
ecc ¼ eco 1 þ 5 0  1
fco
Fiber Length: 30 mm  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 
Slurry Strength: 85 MPa fcc0 ¼ fco
0
2:254 1 þ 7:94 ff0l  2 ff0l  1:254
co co
40 0
Fam and Rizkalla [35] ðfc0 Þi ¼
ðfcc Þi ðxÞi ðrÞi
ðrÞi
ðrÞi 1þðxÞi
 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 
ðf Þ ðf Þ
ðfcc0 Þi ¼ fco
0
2:254 1 þ 7:94 f l0 i  2 f l0 i  1:254
co co

20 ðfl Þi ¼
t
ð c Þi
ðec Þi
R 1ðtc Þi
Es tcom þ ðEc Þi
 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 
ACI Committee 440 [49] fcc0 ¼ fco
0
2:25 1 þ 7:9 ff0l  2 ff0l  1:25
co co

0
0 1 2 3 4 5
Axial Strain, % Table 8
Confinement strength models based on regression analysis [40,48,50–52].
Fig. 11. Comparison of stress–strain responses of a high-strength concrete core
(40 MPa) confined with the SFICON jacket and the GFRP wrap. Confinement model Confined concrete strength
Karbhari and Gao [40] fcc0 ¼ fco
0 0
þ 3:1fco t0 2tDcom EEcom þ 2f comDtcom
h qffiffiffiffii co
equations. Such confinement models use the composite analysis, Spoelstra and Monti [48] fcc ¼ fco 0:2 þ 3 ff0l
0 0
co
regression analysis of experimental results or other analytical Xiao and Wu [50]
h 02 :D
fco
 i
fl
fcc0 ¼ fco
0
1:1 þ ð4:1  0:75 2Ecom :t com Þ f 0
methods to predict the properties of FRP-confined concrete. co

CSA S806 [51] fcc0 ¼ 0:85fco


0
þ 6:7fl0:83
fcom,r = 0.005Ecom, fcom,r = 0.75fcom
5. Evaluation of confinement models with the results of GFRP- Mandal et al. [52] 0
 2  
fcc
f0
¼ 0:0017 EcomRtcom fcom
f0
þ 0:0232 EcomRtcom fcom
f0
þ1
confined cylinders co co co

Different confinement models described in the previous section low-strength concrete cores (see Fig. 12a). However, as shown in
were evaluated based on the results of the tests conducted on Fig. 12b and c, underestimated results have been observed in med-
GFRP-confined concrete cylinders. Comparison of root-mean- ium- and higher-strength concrete cores for the same level of con-
square of normalized errors in predicting the strength of confined finement (1 layer of GFRP). On the other hand, increasing the
concrete leads to the recognition of 7 most accurate models, confinement level by increasing GFRP layers from 1 to 2 and 4
namely Fardis and Khalili [39], Miyauchi et al. [7], Samaan et al. led to underestimation of confined strength results for medium-
[41], Toutanji [42], Spoelstra and Monti [48], ACI Committee 440 strength concrete cores. Also, the accuracy of models to predict
[49] and Li et al. [27]. Fig. 12 shows the differences between the the confined strength was reduced by increasing confinement le-
experimental confined strength and the predicted strengths with vel, as for 4 layers of GFRP, only three models show perdition errors
theses confinement models. As one can see, the predicted values less of 10% or less comparing to five models within such accuracy
are in the range of 20% to +20% of the experimental data. for 2 layers of GFRP.
For low level of confinement, i.e., 1 layer of GFRP, the models
show conservative results to predict the confined strength in
6. Behavior modeling of SIFCON confinement

Table 6 A simple model is presented to predict the compressive


Confinement coefficients of different FRP confinement strength of SIFCON-confined concrete based on a passive confine-
models based on Richart model [27,38–46]. ment model with a constant lateral confining pressure. The con-
Confinement model Factor kc finement model of Richart et al. [37] is suggested to predict the
Fardis and Khalili [38] 2.05 strength of SIFCON-confined concrete. The lateral confining pres-
Fardis and Khalili [39]  0:14 sure applied to the concrete can be calculated according to the free
3:7 ff0l
co
 0:13 body diagram of confined section based on a static analysis, equi-
Karbhari and Gao [40]
2:1 ff0l librium of forces and deformation compatibility [35,36,42,45,46]
co

Miyauchi et al. [7] 3.485 notated with properties of SIFCON confinement.


Samaan et al. [41] 6:0fl0:3
Toutanji [42]  0:15 fot tsifcon
3:5 ff0l fl ¼ ð3Þ
co
 0:16
R
Saafi et al. [43]
2:2 ff0l where fot and tsifcon are the tensile strength and thickness of the SIF-
co

Miyauchi et al. [7] 2.98 CON material and R is the radius of the concrete column.
Lam and Teng [36] 2.00
In this study, tensile stress–strain of SIFCON materials was not
ISIS Canada [44] /frp
/c obtained directly. Therefore, equations proposed by Naaman and
Li et al. [27] tan2 ð45 þ u=2Þ
 0:15 Homrich [16] were used to predict the tensile strength of SIFCON
Matthys et al. [45]
2:3 ff0l material from its compressive strength.
co

Bibsy et al. [46] 2.425


Bibsy et al. [46]  0:089
fl
fot ¼ afpl < V f f pu ð4Þ
2:217
l pffiffiffiffi
0
fco
Bibsy et al. [46] 3:587fl0:16 fpl ¼ V f 1:4 þ 0:66 fo ð5Þ
d
Author's personal copy

774 B. Abdollahi et al. / Construction and Building Materials 36 (2012) 765–778

(a) 30 (b) 30
F2: f'co = 25 MPa
Fardis & Khalili (1982)
Difference between Prediction and

Difference between Prediction and


Miyauchi et al. (2000) n=1
20 Samaan et al. (1998) 20
Toutanji (1999)
Spoelstra and Monti (1999)
10 10
Test Data (%)

Test Data (%)


ACI 440 (2002)
Li et al. (2003)
0 0
Fardis & Khalili (1982)
Miyauchi et al. (2000)
-10 -10 Samaan et al. (1998)
Toutanji (1999)
Spoelstra and Monti (1999)
-20 F1: f'co = 15 MPa -20
ACI 440 (2002)
n=1 Li et al. (2003)
-30 -30

(c) 30 (d) 30
F3: f'co = 40 MPa Fardis & Khalili (1982)
Difference between Prediction and

Difference between Prediction and


n=1 Miyauchi et al. (2000)
20 20 Samaan et al. (1998)
Toutanji (1999)
Spoelstra and Monti (1999)
10 10
Test Data (%)

Test Data (%) ACI 440 (2002)


Li et al. (2003)
0 0
Fardis & Khalili (1982)
Miyauchi et al. (2000)
-10 Samaan et al. (1998) -10
Toutanji (1999)
Spoelstra and Monti (1999)
-20 -20 F4: f'co = 25 MPa
ACI 440 (2002)
Li et al. (2003) n=2
-30 -30

(e) 30
Fardis & Khalili (1982)
Difference between Prediction and

Miyauchi et al. (2000)


20 Samaan et al. (1998)
Toutanji (1999)
Spoelstra and Monti (1999)
10
Test Data (%)

ACI 440 (2002)


Li et al. (2003)
0

-10

-20
F5: f'co = 25 MPa
n=4
-30

Fig. 12. Comparing experimental confined strength with different confinement models.

where fot is the tensile strength of the SIFCON material, calculated that the confinement coefficient, kc, is not constant in different
based on the post peak compressive strength, fpl, fiber tensile confinement plans and depends on the parameters of the uncon-
strength, fpu and fiber volume content, Vf. Factor a is approximately fined concrete strength and the level of provided confining pres-
equal to 1/3 [16] and f0 is the compressive strength of the SIFCON sure. Confinement coefficient as a function of the ratio of the
material. lateral confining pressure to the unconfined concrete strength is
Average Compressive strength of high strength SIFCON cylinders displayed in Fig. 13. As shown in the figure, Eq. (6) is suggested
was 129 MPa. Having the value of f0, the lateral confining pressure for the confinement coefficient based on the regression analysis
for S1 through S6 samples can be obtained using parameters dis- of the experimental results which is applicable to the confinement
0
played in Table 9. These results can be compared with the lateral ratio range of 0:1 < fl =fco < 0:32
confining pressure of GFRP wraps presented in Table 10. In this ta-  0:46
fl
ble, tcom is the thickness of GFRP wrap. Ecom, ecom and fcom are the kc ¼ 1:91 ð6Þ
elastic modulus, ultimate strain and tensile strength of GFRP wraps.
fco0
The general form of Richart confinement model (Eq. (1)) re- Inserting above equation in Eq. (1), the final equation to estimate
quires calculation of a confinement coefficient, kc, for the SIF- the strength of SIFCON-confined concrete is proposed as shown
CON-confined cylinders, as shown in Table 11. The results show below.
Author's personal copy

B. Abdollahi et al. / Construction and Building Materials 36 (2012) 765–778 775

"  0:54 #
fl 12
fcc0 ¼ fco0 1 þ 1:91 0 ð7Þ
fco Confinement with SIFCON Jackets

The above proposed model for SIFCON confinement was compared f'co: 14.8 MPa
with FRP confinement models mentioned earlier in this paper. f'co: 25.1 MPa
Fig. 14 shows a comparison between model predictions and exper- 8 f'co: 41.7 MPa
imental data for the proposed SIFCON model and most accurate

Factor Kc
Exponential Fit
models of FRP confinement applied to the SIFCON confinement.
These models with the best prediction of SIFCON-confined concrete
strength have been found to be Fardis and Khalili [39], Samaan et al.
[41], Toutanji [42] and ACI Committee 440 [49], which are among 4
the best models of FRP confinement as well.
Generally, in predicting strength of SIFCON confinement, the K c = 1.91 (fl / f'co )-0.46
2
R = 0.91
accuracy is lower for medium-strength concrete core with relative
medium and high SIFCON jacket thickness. On the other hand, the
highest accuracies go to medium-strength concrete core with rela-
0
tive low thickness and higher-strength concrete core with relative 0 0.1 0.2 0.3 0.4 0.5
medium jacket thickness. Samaan et al. [41] model overestimated fl /f'co
the experimental data of SIFCON-confined concrete strength for
low strength concrete, but underestimated the SIFCON-confined Fig. 13. Relation between the confinement coefficient factor and the ratio of the
concrete strength for normal and higher strength concrete. The lateral confining pressure to the unconfined concrete strength.
other three top models underestimated the experimental data of
SIFCON-confined concrete strength for normal strength concrete
with different levels of jacket thickness and overestimated SIF- SIFCON-confined specimens with the values lower than 0.23 are
CON-confined concrete strength for low and higher strength con- shown in Fig. 15. As shown in the figure, the proposed model along
crete. A rather similar trend has been observed for the proposed with some of the available models to predict the FRP-confined con-
model which underestimated SIFCON-confined concrete strength crete strength, i.e., ACI Committee 440 [49], Toutanji [42], Fardis
for low and normal strength concrete, but overestimated the con- and Khalili [39], Samaan et al. [41], and CSA 806 [51] can be reli-
fined strength for higher strength concrete core. Furthermore, ably used to estimate the strength of the concrete confined by
comparison of root-mean-square of normalized errors for all the SIFCON method.

Table 9
Calculated mechanical properties of specimens confined by SIFCON jackets.

SIFCON Jacket Fiber Concrete slurry SIFCON compressive SIFCON ultimate SIFCON tensile Lateral confining
confinement thickness, tsifcon length, l strength, fc0 (MPa) strength, fo (MPa) compressive stress, fpl strength, fot (MPa) pressure, fl (MPa)
plan (mm) (mm) (MPa)
S1 15 30 85 129 71.17 23.72 4.74
S2 15 30 85 129 71.17 23.72 4.74
S3 15 30 85 129 71.17 23.72 4.74
S4 20 30 85 129 71.17 23.72 6.33
S5 10 30 85 129 71.17 23.72 3.16
S6 15 50 85 129 118.62 39.54 7.91

Table 10
Calculated mechanical properties of specimens confined by GFRP jackets.

GFRP confinement plan GFRP layer thickness, GFRP tensile modulus, GFRP ultimate strain, GFRP tensile strength, Lateral confining pressure,
tcom (mm) Ecom (MPa) ecom (%) fcom (MPa) fl (MPa)
F1 0.508 24591 1.93 504 3.21
F2 0.508 24591 1.93 504 3.21
F3 0.508 24591 1.93 504 3.21
F4 1.016 24591 1.93 504 6.43
F5 2.032 24591 1.93 504 12.86

Table 11
Calculation of factor kc for concrete specimens confined by SIFCON jackets.

SIFCON confinement plan tsifcon (mm) l (mm) fc0 (MPa) 0


fco (MPa) fl (MPa) fcc0 (MPa) 0
fl =fco kc

S1 15 30 85 14.8 4.74 30.2 0.32 3.25


S2 15 30 85 25.2 4.74 57.2 0.19 6.74
S3 15 30 85 41.7 4.74 64.2 0.11 4.47
S4 20 30 85 25.2 6.33 74.3 0.25 7.76
S5 10 30 85 25.2 3.16 42.7 0.13 5.53
S6 15 50 85 25.2 7.91 50.6 0.31 3.21
Author's personal copy

776 B. Abdollahi et al. / Construction and Building Materials 36 (2012) 765–778

Prediction by Model of Fardis and Khalili (MPa)


80 80
Prediction by the Proposed Model (MPa)
Ultimate Strength of Ultimate Strength of
SIFCON Confined Cylinders SIFCON Confined Cylinders
60 60

40 40
S1 S1
S2 S2
S3 S3
20 S4 20 S4
S5 S5
S6 S6

0 0
0 20 40 60 80 0 20 40 60 80
Experimental Confined Strength (MPa) Experimental Confined Strength (MPa)

80 80
Prediction by Model of Samaan et al. (MPa)

S1 Prediction by Model of Toutanji (MPa)


Ultimate Strength of
S2 SIFCON Confined Cylinders
60 S3 60
S4
S5
S6
40 40
S1
S2
S3
20 20 S4
Ultimate Strength of S5
SIFCON Confined Cylinders S6

0 0
0 20 40 60 80 0 20 40 60 80
Experimental Confined Strength (MPa) Experimental Confined Strength (MPa)

Fig. 14. Comparison of experimental strength of SIFCON-confined cylinders with the proposed model and some confinement models for FRP jackets.

7. Size effects and practical applications of confinement [64]. They presented confined strength of concrete columns as a
methods function of confinement ratio, as well as characteristic length de-
fined by column length minus its diameter. As a result, compres-
Core size and length-to-diameter ratio are among parameters sive strength of a confined cylinder with 300 mm height and
which may affect compressive strength of confined concrete col-
umns. Some researchers found that increasing the core size de-
creases the measured strength [53]. However, others found the 0.4
core diameter has no significant effect on the confined strength ACI 440 (2002)
Root-Mean-Square of Normalized Error

for a specified length-to-diameter ratio [54,55]. Madandoust Toutanji (1999)


et al. [56] studied variations of the correction factor for prediction Fardis & Khalili (1982)
of confined concrete strength and concluded a negligible effect of 0.3 Proposed Model
the core size for a given length-to-diameter ratio. For a specific Samaan et al. (1998)
core diameter, the measured value of compressive strength de- CSA S806 (2002)
pends on its length due to effect of shape on stress distribution Miyauchi et al. (2000)
[57]. This is in agreement with the correction factors presented 0.2
by ASTM C 42 [58] and BS 1881 [59] for different length-to-diam-
eter ratios.
For FRP-confined concrete, a number of researchers stated that
specimen size has no significant effect on the strength improve- 0.1
ment of confined columns [7,60,61], while some other researchers
emphasized significance of the size effect [62,63]. Masia et al. [62]
found that the confinement effectiveness is reduced when the
specimen size is increased. Wang and Wu [64] indicated from 0
experimental studies that the specimen size have a significant ef- Fig. 15. Comparison of root-mean-square of normalized errors in strength predic-
fect on the confined concrete strength, a less significant effect on tion of SIFCON-confined specimens by the proposed model and most accurate FRP
the stress–strain behavior and a slight effect on failure modes confinement models.
Author's personal copy

B. Abdollahi et al. / Construction and Building Materials 36 (2012) 765–778 777

150 mm diameter is predicted to be about two times the confined confining concrete columns decreased by increasing core con-
strength of a practical column with 3000 mm height and 600 mm crete strength. Therefore, initial concrete strength should be
diameter with the same confinement ratio. Consequently, the con- among other criteria to choose a better confinement technique.
finement ratio has to be doubled in practice to achieve the same 4. Confinement performance for higher-strength SIFCON jacket is
performance as the results of this study for the compensation of much better than mid-strength jacket, while fiber length has
the size effect. On the other hand, special attention must be taken no significant effect on the performance of the jacket.
to confinement ratio boundary value which differentiates between 5. GFRP behavior in confinement is so sensitive to the fiber orien-
two types of responses of FRP-confined concrete, namely strain tation with respect to the hoop direction, as the change of fiber
softening and strain hardening. Wu and Wu [65] proposed a orientation from 0° weakened the GFRP performance.
boundary value of 0.13 which is in agreement with results of this 6. The stress–strain responses of GFRP-confined concrete show
study for sample F2. While exceeding this boundary value with bilinear forms before starting failure, while SIFCON curves have
GFRP is not usually practical in large columns, experimental stud- nonlinear behaviors throughout the test.
ies with insufficient confinement on small cylinders (such as F3 in 7. In many cases, the ultimate strength and absorbed energy of
this study) are suggested to better simulate these cases with soft- SIFCON-confined concrete is very similar to the results of
ening response. Nonetheless, the enhancement of ductility of small GFRP-confined cylinders. The comparison shows that SIFCON
diameter cylinders using FRP confinement method would be simi- confinement method can be regarded as a competitive method
lar to large columns. Thus, ductility improvement of large columns with respect to the well-known FRP confinement technique.
would still be considerable which has been known to play a more 8. Passive confinement models can be used to predict compressive
significant role on improving seismic performance of columns [65]. strength of SIFCON-confined columns. Confinement coefficient
Size effects on SIFOCN-confined columns are predicted to be depends on the unconfined concrete strength and the level of
similar to GFRP jackets. However, the ratio of SIFCON fiber length the provided confining pressure. In addition to the confinement
to shell thickness may also affect the behavior of jacket through model proposed in this study, applying some FRP confinement
changing the tensile strength of the material. Tensile strength of models result in relatively accurate prediction of SIFCON con-
a SIFCON composite can be presented by Eq. (8) [66,67]: finement strength.
 
L 9. Caution must be exercised in application of the proposed mod-
rt ¼ rm ð1  V f Þ þ kV f sf ð8Þ els and methods in practice due to size effects and related
D
parameters including characteristic length or length-to-diame-
where rm is the tensile strength of the plain matrix, Vf is the volume ter ratio for both techniques, and jacket thickness to fiber length
percentage of fibers in the composite, K is the orientation factor, sf is ratio for the SIFCON confinement method.
the fiber–matrix bond strength, and L/D is the fiber aspect ratio de-
fined as length/diameter. Stroeven [67] and Soroushian and Lee [68]
studied numerically effects of fiber length and specimen thickness
Acknowledgements
on the orientation factor K, considering a 3-D space with two
boundaries for simulation of a jacket. Their results indicate that at
The authors are profoundly grateful to the late Professor Mehdi
thicknesses smaller than the fiber length, the fiber orientation fac-
Ghalibafian (University of Tehran) for his constructive advice and
tor is close to a 2-D condition, and for thicknesses greater than
valuable discussions during conducting this research. The authors
two times the fiber length, there is a gradual approach to 3-D fiber
gratefully acknowledge financial support for this research from
orientation. In this study, jacket thickness to fiber length ratio
the RADYAB Engineering and Construction Company. This work
ranges from 0.3 to 0.67 which according to Soroushian and Lee
was conducted in the Construction Materials Institute (CMI) at
[68] results in an orientation factor of 0.63, instead of a lower orien-
University of Tehran and as such the authors would like to
tation factor of about 0.45 for practical cases with similar fiber
acknowledge the support that has made these laboratories and
length and jacket thicknesses larger than 90 mm. This shows a
their operation possible. The contents of this paper reflect the
30% reduction in the orientation factor and therefore in tensile
views of the authors, who are responsible for the accuracy of the
strength of SIFCON. Overall, caution must be exercised in applica-
data presented herein.
tion of the proposed models and methods in practice due to size ef-
fects and related parameters including characteristic length or
length-to-diameter ratio for both techniques, and jacket thickness References
to fiber length ratio for the SIFCON confinement method.
[1] Jianhua L. Rehabilitation and repair of reinforced concrete short columns with
external steel collars, PhD dissertation. University of Alberta; 2008.
8. Conclusions [2] Bai JW, Hueste MB. Seismic rehabilitation for reinforced concrete building
structures, Consequence-Based Engineering (CBE) Institute final report. Mid-
America Earthquake Center, Texas A&M University; 2003.
Based on the results of experiments and analyses in this study, [3] FIB – International Federation for Structural Concrete. Seismic assessment and
following conclusions may be drawn: retrofit of reinforced concrete buildings. State-of-art, report. Stuttgart: Sprint-
Digital-Druck; 2003.
[4] Mirmiran A, Shahawy M. Behavior of concrete columns confined by fiber
1. Both GFRP and SIFCON confinement methods enhanced composites. J Struct Eng 1997;123(5):583–90.
strength, strain and energy absorption capacity of concrete [5] Mirmiran A, Shahawy M. Dilation characteristics of confined concrete. Mech
cylinders. Cohes – Frict Mater 1997;2:237–49.
[6] Mirmiran A, Shahawy M, Samaan M, Echary HE, Mastrapa JC, Pico O. Effect of
2. The increase in the thickness of the SIFCON jacket as well as
column parameters on FRP-confined concrete. J Compos Constr
increasing the GFRP layer numbers led to a significant improve- 1998;2(4):175–85.
ment of stress–strain response of confined concrete samples. [7] Miyauchi K, Inoue S, Kuroda T, Kobayashi A. Strengthening effects of concrete
Thickness of the confining layer is found to be the most effective column with carbon fiber sheet. Trans Jpn Concr Inst 2000;21:143–50.
[8] Parent S, Labossière P. Finite element analysis of reinforced concrete columns
parameter to improve the confined concrete behavior in both confined with composite materials. Can J Civil Eng 2000;27(3):400–11.
methods. [9] Perera R. A numerical model to study the seismic retrofit of RC columns with
3. SIFCON confinement technique is more effective for strengthen- advanced composite jacketing. Composites Part B 2006;37(4–5):337–45.
[10] Maa R, Xiao Y, Lib KN. Full-scale testing of a parking structure column
ing of the mid-strength, followed by the low- and the retrofitted with carbon fiber reinforced composites. Constr Build Mater
higher-strength concrete. However, efficiency of GFRP wrap in 2000;14(2):63–71.
Author's personal copy

778 B. Abdollahi et al. / Construction and Building Materials 36 (2012) 765–778

[11] Naaman AE. HPFRCCs: properties and applications in repair and rehabilitation. [40] Karbhari VM, Gao Y. Composite jacketed concrete under uniaxial
ACI SP 85. American Concrete Institute, MI, USA; 2000. p. 1–16. compression–verification of simple design equations. ASCE J Mater Civil Eng
[12] Shao Y, Srinivasan R, Shah P. Parameters affecting high-performance response Constr 1997;9(4):185–92.
in fiber-reinforced concrete. ACI-SP 185. American Concrete Institute, [41] Samaan M, Mirmiran A, Shahawy M. Model of concrete confined by fiber
Michigan; 2000. p. 17–34. composites. ASCE J Struct Eng 1998;124(9):1025–31.
[13] Lankar DR. Properties and applications: slurry infiltrated fiber concrete [42] Toutanji HA. Stress–strain relationship of concrete cylinders confined with FRP
(SIFCON). Concr Int 1984;6(12):44–7. composites. ACI Mater J 1999;96(3):397–404.
[14] Bakhshi M, Abdollahi B, Ghalibafian M, Shekarchi M. Effects of material [43] Saafi M, Toutanji HA, Li Z. Behavior of concrete columns confined with fiber
parameters on compressive behavior of HPFRC strengthened concrete reinforced polymer tubes. ACI Mater J 1999;96(4):500–9.
columns, CONSEC’07; 2007. p. 1777–88. [44] ISIS Canada, Design manual no. 4: strengthening reinforced concrete
[15] Naaman AE, Baccouche MR. Shear response of dowel reinforced SIFCON. ACI structures with externally bonded fiber reinforced polymers, Intelligent
Struct J 1995;92(5):587–96. Sensing for Innovative Structures (ISIS), Canada, Winnipeg; 2001.
[16] Naaman AE, Homrich JR. Tensile stress–strain properties of SIFCON. ACI Mater [45] Matthys S, Toutanji HA, Audenaert K, Taerwe L. Axial load behavior of large-
J 1989;86(3):244–51. scale columns confined with fiber-reinforced polymer composites. ACI Struct J
[17] Wood BT. Use of slurry infiltrated fiber concrete (SIFCON) in hinge regions for 2005;102(2):258–67.
earthquake resistant structures, PhD thesis. North Carolina State University, [46] Bisby LA, Dent AJS, Green MF. Comparison of confinement models for fiber-
Raleigh, NC; 2000. reinforced polymer-wrapped concrete. ACI Struct J 2005;102(1):62–72.
[18] Elnono M, Salem HM, Farahat A, Elzanaty H. Use of slurry infiltrated fiber [47] Mander JB, Priestley MJN, Park RJT. Theoretical stress–strain model for
concrete in reinforced concrete corner connections subjected to opening confined concrete. ASCE J Struct Eng 1988;114(8):1804–23.
moments. J Adv Concr Technol 2009;7(1):51–9. [48] Spoelstra MR, Monti G. FRP-confined concrete model. ASCE J Compos Constr
[19] Abdou H. Precast construction using slurry infiltrated fiber concrete joints 1999;3(3):143–50.
under load reversals. PhD thesis. University of Michigan, MI; 1990. [49] ACI Committee 440. ACI 440.2R-02 Guide for the design and construction of
[20] Sahannang MJ, Barakat S, Jaber F. Structural repair of shear-deficient externally bonded FRP systems for strengthening concrete structures.
reinforced concrete beams using SIFCON. Mag Concr Res 2001;53(6):391–403. American Concrete Institute, Farmington Hills, Mich.; 2002.
[21] Dogan E, Krstulovic-Opara N. Seismic retrofit with continuous slurry- [50] Xiao Y, Wu H. Compressive behavior of concrete confined by carbon fiber
infiltrated mat concrete jackets. ACI Struct J 2003;100(6):713–22. composite jackets. ASCE J Mater Civil Eng 2000;12(2):139–46.
[22] Hamza AM, Naaman AE. Bond characteristics of deformed reinforcing steel [51] CAN/CSA-S806. Design and construction of building components with fiber-
bars embedded in SIFCON. ACI Mater J 1999;93(6):578–88. reinforced polymers. Canadian Standard Association, Ottawa; 2002.
[23] Sharma HK, Singh VP. Performance characteristics SIFCON plates. In: Adv top [52] Mandal S, Hoskin A, Fam A. Influence of concrete strength on confinement
geol seismolog. Proceedings of the 2nd IASME/WSEAS international conference effectiveness of fiber-reinforced polymer circular jackets. ACI Struct J
on geology and seismology (GES’08). Cambridge, UK; 2008. p. 149–54. 2005;102(3):383–92.
[24] Yazici H, Aydin S, Yiğiter H, YardimcI MY, Alptuna G. Improvement on SIFCON [53] Campbell RH, Tobin RE. ACI J, Proc 1967;64(4):190–5.
performance by fiber orientation and high-volume mineral admixtures. J [54] Meininger RC. J Mater 1968;3(2):320–36.
Mater Civil Eng 2010;22(11):1093–101. [55] Lewis RK. Effect of core diameter on the observed strength of concrete cores.
[25] Coskun H. Construction of SIMCON retrofitted reinforced concrete columns. Research report no. 50. Commonwealth Scientific and Industrial Research
PhD thesis. North Carolina State University, NC; 2002. Organization, Division of Building Research; 1976. 13 p.
[26] Pessiki S, Harries KA, Kestner JT, Sause R, Ricles JM. Axial behavior of reinforced [56] Madandoust R, Bungey JH, Ghavidel R. Prediction of the concrete compressive
concrete columns confined with FRP jackets. J Compos Constr strength by means of core testing using GMDH-type neural network and ANFIS
2001;5(4):237–45. models. Comput Mater Sci 2012;51:261–72.
[27] Li Y, Lin C, Sung Y. A constitutive model for concrete confined with carbon fiber [57] Bungey JH, Millard SG, Grantham MG. Testing of concrete in structures. 4th
reinforced plastics. Mech Mater 2003;35(3–6):603–19. ed. London, New York: Taylor & Francis; 2006. 352 p.
[28] Kalidindi SR, Abusafieh A, EI-Danaf E. Accurate characterization of machine [58] ASTM C 42-12. Standard test method for obtaining and testing drilled cores
compliance for simple compression testing. Exp Mech 1997;37(2):210–5. and sawed beams of concrete, ASTM Standard Book, 04(02); 2012.
[29] ACI 318-08. Building code requirements for structural concrete, and [59] BS 1881-108. Testing concrete. Method for making test cubes from fresh
commentary – an ACI standard. American Concrete Institute, Farmington concrete, British Standards Institution; 1983.
Hills, MI, USA; 2008. [60] De Lorenzis L, Micelli F, La Tegola A. Influence of specimen size and resin type
[30] Au C, Buyukozturk O. Effect of fiber orientation and ply mix on fiber reinforced on the behavior of FRP-confined concrete cylinders. In: Proc ACIC-2002,
polymer-confined concrete. J Compos Constr 2005;9(5):397–407. Thomas Telford, UK; 2002.
[31] Saatcioglu M, Razvi SR. Strength and ductility of confined concrete. ASCE J [61] Theriault M, Neale KW, Claude S. Fiber-reinforced polymer-confined circular
Struct Eng 1992;118(6):1590–637. concrete columns: Investigation of size and slenderness effects. J Compos
[32] Harries KA, Kharel G. Experimental investigation of the behavior of variably Constr 2004;8(4):323–31.
confined concrete. Cem Concr Res 2003;33(6):873–80. [62] Masia MJ, Gale TN, Shrive NG. Size effects in axially loaded square-section
[33] Carey SA, Harries KA. Axial behavior and modeling of confined small-, concrete prisms strengthened using carbon fiber reinforced polymer
medium-, and large-scale circular sections with carbon fiber-reinforced wrapping. Can J Civil Eng 2004;31(1):1–13.
polymer jackets. ACI Struct J 2005;102(4):596–604. [63] Silva MA, Rodrigues CC. Size and relative stiffness effects on compressive
[34] Harries KA, Kharel G. Behavior and modeling of concrete subject to variable failure of concrete columns wrapped with glass FRP. J Mater Civil Eng
confining pressure. ACI Mater J 2002;99(2):180–9. 2006;18(3):334–42.
[35] Fam AZ, Rizkalla SH. Confined model for axially loaded concrete confined by [64] Wang Y-F, Wu H-L. Size effect of concrete short columns confined with aramid
circular fiber-reinforced polymer tubes. ACI Struct J 2001;98(4):451–61. FRP jackets. ASCE J Compos Constr 2011;15(4):535–44.
[36] Lam L, Teng JG. Design-oriented stress–strain model for FRP-confined [65] Wu G, Lu ZT, Wu ZS. Strength and ductility of concrete cylinders confined with
concrete. Constr Build Mater 2003;17(6–7):471–89. FRP composites. Constr Build Mater 2006;20(3):134–48.
[37] Richart M, Grandtzaeg A, Brown RL. The failure of plain and spirally reinforced [66] Lankard SD. Properties, applications: slurry infiltrated fiber concrete (SIFCON).
concrete in compression, bulletin no. 190. Engineering Experiment Station, Concr Int 1984;6(12):44–7.
University of Illinois, Urbana; 1929. [67] Stroeven P. Stereological principles of spatial modeling applied to steel fiber-
[38] Fardis MN, Khalili HH. Concrete encased in fiberglass-reinforced plastic. ACI J reinforced concrete in tension. ACI Mater J 2009;106(3):213–22.
1981;78(6):440–6. [68] Soroushian P, Lee C-D. Distribution and orientation of fibers in steel fiber
[39] Fardis MN, Khalili HH. FRP-encased concrete as a structural material. Mag reinforced concrete. ACI Mater J 1990;87(5):433–9.
Concr Res 1982;34(121):191–201.

View publication stats

You might also like