Composites Part B 163 (2019) 96–106

Contents lists available at ScienceDirect

Composites Part B
journal homepage: www.elsevier.com/locate/compositesb

Prismatic RC columns externally confined with FRP sheets and pre- T

tensioned basalt fiber ropes under cyclic axial load
Theodoros C. Rousakis∗, Georgios D. Panagiotakis, Emmanouela E. Archontaki,
Alexandros K. Kostopoulos
Laboratory of Reinforced Concrete, Department of Civil Engineering, Democritus University of Thrace (DUTH), Xanthi, 67131, Greece

A R T I C LE I N FO A B S T R A C T

Keywords: This experimental investigation addresses the use of transversely placed composite ropes made of basalt as
A. Fibres external standalone confining reinforcement in cases of reinforced concrete columns with square section or
A. Hybrid combined with already applied FRP sheets. Non bonded, non impregnated basalt fiber rope is wrapped around
B. Buckling the columns by hand in 1 or 2 layers or is pre-tensioned with special mechanical devices. The efficiency of
D. Mechanical testing
wrapped columns is compared with identical columns confined with polypropylene fiber ropes. Further, the
E. Pretension of confinement
columns are wrapped with externally bonded GFRP sheets and then with non bonded basalt fiber ropes in a
hybrid scheme. The columns are subjected to multiple cycles of increasing compressive deformation to simulate
seismic actions. The columns wrapped with two layers of basalt composite ropes (pre-tensioned or not) present
an ever increasing stress-strain response for concrete axial strains higher than 5.6%, even for very high rope
pretension level, without rope fracture. Both the polypropylene and basalt ropes reveal a high potential for strain
redistribution around the variably damaged concrete core. In the cases of hybrid FRP – basalt rope confinement
the strain redistribution potential of the rope is similarly remarkable, controlling the dilation of the column even
after the multiple fracture of the GFRP sheet. The rope could increase the axial strain sustained by the concrete
from 2.21% to beyond 5.1%, without rope fracture.

1. Introduction scale RC columns of circular section [14,15,19] by hand and provide

adequate axial strength and axial strain ductility enhancement. FRP
Deficient reinforced concrete (RC) columns in old structures de- confinement has been successfully applied on non-circular small and
signed and built according to previous generation of codes, are common large scale RC columns with slender steel bars and inadequate stirrup
cause of high vulnerability and of accumulation of severe damages. detailing as well [14,16–18,20–22]. Partial FRP wrapping was proved
Confinement with FRP materials is an attractive technique to increase to be very effective for plain and RC columns with deficient detailing of
adequately the axial strain ductility of concrete. Plain concrete columns steel reinforcement [15,23,24]. On the other hand research efforts ex-
confined with FRPs have been extensively studied during the last 35 plored the effects of pre-tensioned FRP filaments made of carbon
years either as encased in glass FRP tubes [1] or wrapped with FRP [25,26] or basalt [28]. The use of pre-tensioned FRP filaments, sheets
sheets and subjected to monotonic [2–4] or cyclic compressive load or straps [27,29] for concrete confinement, further prolonged the lin-
[5,6] to address technique-related or mode of load-related effects. The earity limit in stress-strain behavior of concrete columns under axial
geometry of the concrete section affects significantly the efficiency of load and increased the bearing load at lower strain levels at the expense
FRP confinement. Therefore, several strength and strain prediction of reduced deformability. Despite the high efficiency of FRP confine-
models have been proposed for FRP confined concrete columns of cir- ment, several studies have concluded that unfavorable effects of the
cular or square or rectangular sections [1,3,4,7–10,13], validated internal steel reinforcement in columns with old detailing have to be
against experimental results. Advanced constitutive models and nu- taken into account (Pessiki et al., 2001 [14], Matthys et al., 2006 [15],
merical approaches enabled for the utilization of finite element ap- Rousakis and Karabinis 2008 and 2012 [16,18], Bai et al., 2017 [22],
proaches [3,6,7,11,12] in predicting the stress-strain behavior of con- Wang et al. [19,20] among others), especially in large scale reinforced
fined concrete. FRP sheets have been successfully used to confine large concrete columns (commonly met in practice [21]). There should be a

∗
Corresponding author. Department of Civil Engineering, Democritus University of Thrace, Vas. Sofias 12, 67131, Xanthi, Greece.
E-mail address: trousak@civil.duth.gr (T.C. Rousakis).

https://doi.org/10.1016/j.compositesb.2018.11.024
Received 16 August 2018; Received in revised form 4 October 2018; Accepted 3 November 2018
Available online 05 November 2018
1359-8368/ © 2018 Elsevier Ltd. All rights reserved.
T.C. Rousakis et al. Composites Part B 163 (2019) 96–106

further reduced effective strain of the FRP jacket at failure for design fracture of the FRP sheet. Then, there was load regaining, up to the
[21,22] and even revise the considered stress-strain curves for confined early stop of the cyclic testing at levels higher than 4% without PPFR
concrete as the efficiency is lower than in small scale plain concrete fracture. The proposed hybrid scheme could extend the axial deform-
ones, if the confinement is inadequate [18]. These issues are very sig- ability of the plain or RC column beyond the fracture of the GFRP jacket
nificant especially if performance - based design is followed and seismic [38] or of the basalt fiber tape [34]. It could provide maximum utili-
resistant structures are involved. In that case, unexpected overloading zation of the potential of the column to bear axial load, up to the col-
of the structure due to seismic excitations may jeopardize the failure lapse because of global second order effects. In the latter case, the RC
sequence in the building and trigger undesirable collapse and loss of column revealed no temporary load drop while it was already tested to
human lives [30]. axial strain levels higher than 5.5% in Ref. [40], it was repaired with
Several studies have highlighted the interesting features of concrete high strength mortar and then retrofitted with hybrid basalt fiber tape-
columns externally wrapped with non impregnated, non bonded tapes PPFR confinement (no resin at all has been used).
[31–35] or composite fiber ropes (FR) [34,36–41] (even in RC ones This paper, explores the efficiency of the external transverse
with deficient detailing). Suitably designed composite FR wrapping strengthening of deficient RC columns with composite basalt fiber ropes
could extend remarkably the compressive axial strain of RC columns (BSFR). The columns had square section with slender bars. The BSFR
beyond 5% strain with no fracture of the FR [34,40], up to the early were non-impregnated, non bonded, applied by hand or pre-tensioned
stop of the testing to avoid unsafe global-instability-related failures. by mechanical devices, similarly to PPFRs in Ref. [40]. Therefore, low
Even in cases pre-tensioned composite rope confinement had suitable deformability BSFR confined columns are used in this study and are
detailing according to [40], the axial strain ductility of deficient RC compared against very high deformability PPFR ones directly. Further,
columns was remarkably high, reaching strain higher than 5% with hybrid GFRP – BSFR confining schemes are applied to deficient RC
hardening stress-strain behavior and no FR fracture up to the early stop columns of square section. The columns were subjected to suitable
of testing. successive cycles of increasing compression – decompression – re-
Hybrid FRP strengthening may be a very effective approach to uti- compression, through displacement control.
lize the unique characteristics of different materials and techniques. So
far, only limited studies refer to the application of different FRP sheets
2. Materials and methods
in combination such as: hybrid glass-basalt FRP (De Luca et al. [17]) in
a large scale RC column. Further, plain concrete cylinders confined with
The tests included eight (8) columns of square section with di-
combined high modulus (and very low deformability) carbon FRP sheet
mensions of 150 × 150 mm, corner radius of 17 mm and a height of
(540 GPa) with high strength carbon FRP sheet or with PBO FRP sheet
750 mm, confined with glass FRP sheets or composite ropes (preten-
or with aramid FRP sheet or with glass FRP sheet or further combined
sioned or not). Seven out of the eight specimens had internal steel re-
high modulus carbon FRP and high strength carbon FRP and aramide
inforcement. Another two concrete prisms were left unwrapped for
FRP together in the study by Wu et al., 2008 [42]. They suggest [42]
comparison.
that a suitable combination of high strength inner FRP with high de-
formability outer FRP sheet at adequate ratio should be provided to
improve the ultimate strength. Then, the axial strain ductility of con- 2.1. Mechanical properties of materials
crete is governed by the outer FRP sheet with the high axial deform-
ability. In another study [43], hybrid glass – sisal fiber FRP confinement The columns were constructed with ready–mixed concrete con-
was used as a viable and cost effective solution to retrofit plain concrete forming to low strength C12 category. The plain concrete prism had
cylinders. Recently, Ispir et al., 2018 [44] combined carbon FRP sheet strength of 15.24 MPa at an axial strain of 0.291%. All reinforced col-
or glass FRP sheets as inner layers with PET FRP sheets as outer layers umns had S220 (nominal yield strength fy = 220 MPa) transverse
(rupture strain of 10%) to retrofit concrete cylinders. Further, hybrid smooth steel stirrups of 5.5 mm diameter, spaced at 100 mm centers
carbon FRP sheet and glass FRP sheet confinements have been con- and ribbed longitudinal bars of 8 mm diameter and B500C quality
structed. They [44] concluded that by using a combination of low- (nominal yield strength fy = 500 MPa). As mentioned above, their
rupture strain FRP as inner layer (i.e. carbon or glass) and high rupture stirrup spacing to bar diameter ratio s/ØL was equal to 12.5.
strain FRP (i.e. PET) as outer layer of confinement, the axial strain and The FRP confinement included unidirectional glass sheets of
dissipated energy of confined concrete increase enormously after the SikaWrap®-430 G/25 type. The GFRP sheets had a nominal tensile
fracture of the inner layer. They reach values around the axial strain of modulus (Ef) 76 GPa, elongation at failure (εju) 2.8% and a layer
concrete at failure for PET-FRP confined concrete alone. thickness (tj) of 0.172 mm (Sika, Products Data sheets 2010) as shown
In the study by Rousakis 2013 [38] hybrid FRP – PPFR strength- in Table 1. Treatment of any undesirable cavities and pores needed the
ening was proposed for circular plain concrete columns under cyclic use of a two-component epoxy paste (Sikadur 31 CF Normal). The
axial compressive load, combining glass FRP sheets with non bonded composite materials were applied by hand on the concrete columns
non impregnated composite ropes of very high deformability. In Rou- with the use of a two – component epoxy resin Sikadur 330.
sakis 2016 [34] hybrid basalt fiber tape and PPFR was applied in de- The polypropylene fiber rope (tPPFR 11gr/m, product of Thrace
ficient RC column subjected to cyclic axial compressive load. In all Plastic Co. S.A, Greece) is designed with two Z-twisted strands. The
cases the hybrid confining scheme resulted in enormous increase of the rope has a tensile modulus of elasticity of 2.0 GPa, an ultimate stress of
concrete axial strain, after a temporary load drop because of the 405.3 MPa and 20.36% elongation at failure. This study also includes a
braided basalt fiber rope, which has a tensile modulus of elasticity of

Table 1
Tensile properties of glass FRP, of polypropylene fiber rope and of basalt fiber rope.
Label Type of reinforcement Area of cross section, AFR, (mm2) Tensile strength, fFRu, (MPa) Tensile modulus of elasticity, EFR, (GPa) Strain at failure
εFRu, (%)

SikaWrap 430 G/25 Unidirectional GFRP 0.172a 2300 76 2.8

PPFR 11 gr/m Z-twisted 2-strand PPFR 11.9 405.3 2.0 20.36
Βasalt FR Braided BFR 7.81 1282.8 69.3 1.85

a
Per 1 mm width of GFRP sheet.

97
T.C. Rousakis et al. Composites Part B 163 (2019) 96–106

Table 2
Specimens detailing.
Specimens label Concrete fck Maximum stress at first stress Confining layers Internal B500C steel bars Pretension (%) Failure
(MPa) reduction fcc1 (MPa) material and S220 stirrups mode

PC 15.24 19.05 – – – –
RC 16.79 20.99 – – Yes –
RCGL1 16.79 27.74 GFRP 1 Yes – F
RCPPL4 16.79 25.41 PPFR 4 Yes – ES
RCPPL13PRE1.98%a 16.79 34.36 PPFR 1 + 3PRE Yes 1.98 ES
RCBSL1 16.79 23.75 BS 1 Yes – CRF
RCBSL2 16.79 29.44 BS 2 Yes – ES
RCBSL12PRE1.87%a 16.79 38.22 BS 1 + 2 PRE Yes 1.87 ES
PCGL1BSL2 15.24 25.06 GFRP + BS 1 GFRP + 2 BS No – F, ES
RCGL1BSL2 16.79 29.14 GFRP + BS 1 GFRP + 2 BS Yes – F, ES

No composite rope fracture for PPFR.

F = FRP jacket fracture.
CRF = Composite Rope Fracture (for 1 layer of basalt rope).
ES = Early Stop because of no composite rope fracture.
a
The 1st layer of composite rope was wrapped around the concrete surface and the remaining 3 or 2 layers were wrapped around the metal devices and were
pretensioned.

69.3 GPa, an ultimate stress of 1282 MPa and 1.85% elongation at

failure (as provided by the manufacturer Juan Gili S.L.).

2.2. Layout of composite strengthening

One reinforced concrete (RC) column was confined with GFRP. The
600 mm width sheets were placed at the mid-height area of the column,
leaving a 75 mm gap on the top and on the bottom of the column that
was covered by additional 150 mm width straps. Thus, the upper areas
were further reinforced with an additional glass sheet layer in order to
ensure failure at the mid-height of the specimen. Also, a 10 mm gap was
left between the loading plates of the compression machine and the
composite jacket, to avoid direct loading of the jacket. Wrapping had an
orientation perpendicular to the column axis. The RC column with glass
fiber (G) sheet in one layer (L1) was labeled RCGL1 (see Table 2). The
wrapped column remained in lab conditions for more than 10 days after Fig. 1. Metal pretensioning device and attachment of the devices on the column
gluing. (a), RCPPL4PRE wrapping detailing (b and c).
Three reinforced concrete columns were confined with fiber ropes
(FR) made of polypropylene or basalt, respectively. The composite for more details). At first the composite ropes were fully wrapped
ropes were wrapped around the section of the columns and along the around the column for 1 layer. Then, the metal devices were attached
height by hand, while no mechanical means for the stretching of the FR on the concrete surface (Fig. 1a) using one of their holes. Finally, the
were used. The external wrapping of concrete columns with FR needs composite rope was wrapped around the specimen by hand as in non-
no special treatment of the concrete surface. Yet, large concave-con- pretensioned columns (Fig. 1b and c). After the end of the wrapping, the
crete substrates need suitable treatment since they may reduce the two bolts in each metal device were screwed gradually with a simple
confinement effectiveness significantly. Moreover, the FRs were not wrench, by hand. After each turn, the metal bar rises and the FRs are
impregnated with resin and they were not bonded on the concrete subjected to pretension. This results in a reaction force on the grove
surface. The ropes may have been reused several times in cases with no region of the plate which is spread all over the plate surface. The pre-
FR damage. The two ends of the ropes were anchored mechanically tension scheme was designed suitably in order to apply prestress on the
through two steel collars placed near the top and the base of the square concrete core both at the middle of the sides through the plates and at
column, by pushing them on the concrete surface. the corners of the sections where the FR was in contact [40]. The ex-
The wrapping of external reinforcement was done manually. The clusive contact regions of the pretensioned rope were the metal bars
continuous wrapping stress on each coil plays a crucial role. The and the corner regions. The technique enables for the full utilization of
wrapping requires special care, as the rope needs to be placed with the pretension since no friction forces between the core and the PPFR
great detail at the edges of the test specimen, remaining stretched by are generated. It ensures a uniform pretension of the rope around the
hand at all times. The reinforcement should be wound tightly around core. Finally, it enables for the uniform pretension of multiple layers of
every edge, without any looseness or void between the coils. More FRs [40].
details on the technique can be found in Refs. [37,38]. Reinforced The column RCPPL13PRE1.98% was wrapped with 1 layer of con-
concrete columns RCPPL4, RCBSL1 and RCBSL2 were wrapped with 4 tinuous PPFR according to Fig. 1a prior to the attachment of the metal
layers (L4) of continuous polypropylene rope (PP) or 1 layer (L1) of devices. Then, another 3 layers of PPFR were applied on the metal
continuous basalt rope (BS), or 2 layers (L2) of continuous basalt rope devices and were pretensioned. The column had three pretension de-
respectively (see Table 2). vices per column face (see Fig. 1b). The column had 3 layers of PPFR
Another two columns RCPPL13PRE1.98% and RCBSL12PRE1.87% and lower pretension level on the middle pretension devices, around
had pretensioned composite ropes, fully wrapped along the height of 1.98% strain. The other two pretension devices had a higher pretension
the columns according to Fig. 1. Special mechanical devices have been level to ensure that the failure would happen at the middle-height
used in order to apply pretension to the PPFR and BSFR (see Ref. [40]

98
T.C. Rousakis et al. Composites Part B 163 (2019) 96–106

region of the specimen (see Ref. [40]).

The column RCBSL12PRE1.87% was wrapped with 1 layer of con-
tinuous basalt fiber rope according to Fig. 1a prior to the attachment of
the metal devices. Then, another 2 layers of BSFR were applied on the
metal devices and were pretensioned. Similarly to PPFR pretensioned
column, the BSFR pretensioned one had 2 layers of BSFR and lower
pretension level on the middle pretension devices, around 1.87% strain.
The remaining 2 columns had hybrid GFRP – basalt fiber rope
confinement. Plain concrete column PCGL1BSL2 had 1 layer of GFRP
sheet glued on the concrete surface as mentioned above (see RCGL1).
Afterwards, 2 layers of continuous basalt rope were applied on the
column, non bonded and non impregnated with resin (as in RCBSL2).
Column RCGL1BSL2 had identical external confinement and ad-
ditionally internal steel reinforcement.

2.3. Equipment and loading of specimens

The specimens were instrumented with a linear variable displace-

ment transducer (LVDT) to record the displacement between the
loading platens. A wire displacement meter was attached between the Fig. 3. Specimen RCGL1 after the fracture of GFRP sheet (a), after the removal
two steel collars to measure the axial deformation on the concrete core. of the damaged GFRP sheet (b) and of severely cracked concrete (c).
The steel collars were positioned with a relative distance of 550 mm
centers (gap of 100 mm from specimen ends). Two linear variable dis- Then they were gradually increased up to 0.8% compressive strain. The
placement transducers (LVDT), placed opposite one another, provided tests were executed with a loading machine of 3000 kN capacity. All
the lateral deformations of the concrete core. Fig. 2 presents the typical specimens were subjected to axial compressive loading– unloading –
instrumentation of the specimens. A special laser displacement meter reloading, under a displacement control of 3*10-5 strain per sec. Most
gave the relative deformation of the outer FR spiral by measuring the of the tests of columns wrapped with FR (with or without FRP jackets)
relative distance between attached targets. Specimen RCGL1 had also stopped for safety reasons because of severe concrete related damages
strain gauge attached one the FRP jacket. without FR fracture or axial load drop to avoid global instability failure
Previous tests on plain concrete cylinders [37] and on PC and RC (see the following sections).
prismatic columns confined with FR [40], have presented a temporary
load drop in some cases. Herein the square section specimens were
subjected to close cycles of approximately 0.05% compressive strain. 3. Experimental results

3.1. Audiovisual remarks

The column RCGL1 with GFRP jacketing presented an abrupt failure

marked by the fracture of the sheet. The fracture of the GFRP jacket was
located at the middle height of the specimen (Fig. 3a). Fig. 3b and c
illustrate the column after the removal of the GFRP.
The PPFR wrapped RC column presented no failure despite the ex-
tensive cracking of concrete during loading – unloading even for mul-
tiple cycles and severe bar buckling in different locations. Fig. 4a shows
the reinforced concrete column RCPPL4 with 4 layers of PPFR after the
end of the test. The FR showed a very high stress redistribution capacity
and preserved full contact with the dilating and variably damaged
concrete core (see Fig. 4b and c).
The reinforced concrete column RCBSL2 presented negligible tem-
porary load decrease despite the obvious lateral deformation of the
concrete core, the buckling of the steel bars and the pronounced
member buckling. Fig. 5 shows the behavior of the basalt FR wrapped
reinforced concrete column. Similarly, the composite rope showed a
very high stress redistribution capacity and preserved full contact with
the dilating and variably damaged concrete core.
On the other hand, column RCBSL1 suffered severe damage of the
concrete core and multiple buckling of the steel bars, followed by
fracture of the composite rope during the 24th cycle. The specimen
could sustain another cycle up to 0.9 of maximum bearing load and
then to final failure marked with the fracture of the rope in 3 different
places at the middle height region (see Fig. 6).
Specimen RCPPL13PRE1.98% suffered severe global buckling
without polypropylene rope fracture (Fig. 7). Fig. 7b and c shows re-
latively bulged concrete core outside the region of the metal plate
Fig. 2. Instrumentation of specimens RCGL1 (a), RCPPL4 (b), RCBSL2 (c), contact. The pretension could maintain the integrity of the concrete
PCGL1BSL2 (d), RCPPL13PRE1.98% (e), RCGL1BSL2 (f), RCBSL12PRE1.87% core up to high axial strain ductility.
(g), RCBSL1 (h). Similarly, in column RCBSL12PRE1.87% no fracture of the basalt

99
T.C. Rousakis et al. Composites Part B 163 (2019) 96–106

Fig. 4. Wrapped specimen RCPPL4 after the end of the test (a). Specimen after Fig. 7. Specimen RCPPL13PRE1.98% after the end of the test. Global buckling
unwrapping and removal of severely cracked concrete (b). Extensive disin- of the column (a), intact PPFR (b), relatively bulged concrete core at corner
tegration of concrete core (c). regions (c).

Fig. 5. Specimen RCBSL2 after the end of the test. Dilation of the concrete core
and member buckling (a), variable deformation of concrete after column un-
wrapping and removal of highly cracked concrete (b), (c).

Fig. 8. Specimen RCBSL12PRE1.87% after the end of the test. Intact BSFR (a),
cracked concrete core (b).

contributed to the better integrity of the concrete core.

The column PCGL1BSL2 was without steel reinforcement. The test
was terminated early because the column exhibited severe global
buckling (Fig. 9a), while no breakage occurred in the basalt rope
(Fig. 9b). Fig. 9c and d shows the FRP jacket fractured in different re-
gions, after the removal of the intact basalt rope. The variable FRP
damages (Fig. 9c and d) are related to the variable tensile and com-
pressive strains developed due to the interaction with the concrete
through full bonding. Multiple FRP fracture occurs due to the interac-
tion with the BSFR (see also [38]).
Fig. 6. Specimen RCBSL1 after the end of the test. Dilation of the concrete core The specimen RCGL1BSL2 had internal steel reinforcement. The test
and rope fracture (a), extensive shear cracking of concrete and bar buckling
was terminated early similarly, because the column exhibited severe
(b–d).
global buckling (Fig. 10a), while no breakage occurred in the basalt
rope (Fig. 10b). Figs. 10c and d shows the variable FRP jacket damage,
rope occurred (see Fig. 8a) despite the fact that it was pretensioned at related also to the location of the steel stirrups and buckled steel bars.
1.87% strain prior to loading. Again, the pretension of the rope The tests that involved FR wrapping were stopped before reaching

100
T.C. Rousakis et al. Composites Part B 163 (2019) 96–106

coupling of different parts through simple knots. In several cases the

polypropylene or basalt ropes have been reused multiple times.

3.2. Stress - strain curves

The mechanical behavior of the confined columns was evaluated in

terms of their stress - axial and lateral strain response. The evaluation
concerned qualitative aspects of the behavior depicted in Figs. 11–18.
Also, stress and strain recordings at several specific column shortening
points and at test termination were gathered in Table 3.
Test specimen PC1 was subjected to axial compressive load reaching
maximum axial compressive stress of 19.05 ΜPa for an axial strain of
0.262%. Failure occurred for fcu = 15.24 ΜPa and a corresponding
strain of εcu = 0.2908%. Test specimen RC1 had maximum axial com-
pressive stress of 20.99 ΜPa for an axial strain of 0.502%. Failure oc-
curred for fcu = 16.79 ΜPa and εcu = 0.8524% strain.
Fig. 9. Specimen PCBSL2GL1 after the end of the test. Global buckling of Column RCGL1 developed a maximum axial stress of 27.74 MPa at
column (a), intact BSFR and bulged concrete core (b), variable fracture of FRP 1.53% strain. Then a degrading stress-strain behavior followed (Fig. 11)
jacket due to tensile and compressive strains (c–d).
with an ultimate axial strain of εcu = 21.3‰ at 22.12 MPa (see Table 3)
upon the fracture of FRP. The energy absorbed by the column was
measured by the area below the stress-strain curves during loading
through the load-unload-reload cycles, estimated at 0.586 MJ/m3.
Column RCPPL4 developed a maximum axial stress of 25.41 MPa at
an axial strain of εcu = 6.930% (Fig. 12), absorbing energy of
2.156 MJ/m3. Then, the test was early terminated without fracture in
any part of the propylene rope. The load dropped temporarily while the
experiment was being carried out (negligible drop of 5.6% from
fcc1 = 25.41 MPa to fcc2 = 24 MPa), but following a redistribution of
the lateral stresses, the load was stabilized and, subsequently, increased
(see also [30,34,40]).
Fig. 13 shows the pretensioned column RCPPL13PRE1.98% devel-
oped a maximum axial stress of 41.52 MPa at axial strain of
εcu = 6.071%, absorbing energy of 2.519 MJ/m3. The test stopped early
without fracture in any part of the polypropylene rope. The temporary
load drop was negligible, ensuring ever-increasing bearing stress re-
sponse for the column. Pretension of the confining PPFR resulted in
Fig. 10. Specimen RCGL1BSL2 after the end of the test. Global buckling (a), remarkable increase of the bearing stress at the elastic limit of the
intact BSFR (b), variable FRP fracture (c), cracked concrete core and buckled stress-strain response of the column, at 1% axial strain level as well as at
bars (d). 3% axial strain level: from 23.61 MPa to 34.03 MPa, increase of 44%
when compared with RCPPL4 column.
failure (except for RCBSL1). The main reason for terminating the tests Fig. 14 shows that test specimen RCBSL2 developed a maximum
was the severe global buckling of the columns as well as potential un- axial stress of 29.44 MPa at a strain of εcu = 7.341%, absorbing energy
safe dislocations of the steel collars. During the testing of the column, of 2.208 MJ/m3. The two layers of basalt rope restricted the dilation of
sounds could be heard because of the concrete cracking, the rope concrete core efficiently as the bearing load was ever increasing up to
stretching and GFRP fracture. These sounds were more pronounced early termination. The general stress-strain behavior is similar to that of
during the temporary load drop and load regaining phase and around RCPPL4 column as well as the bearing stress at 1% and 3% axial strain
the maximum load per cycle. The ropes presented no sensitivity to levels. Similar were also the lateral strain levels for the two columns,
around 3.5% strain. It should be noted that this lateral strain level

Fig. 11. Stress versus axial and lateral strain curves of RC column, confined by 1 layer of GFRP (RCGL1), under cyclic axial load.

101
T.C. Rousakis et al. Composites Part B 163 (2019) 96–106

Fig. 12. Stress versus axial and lateral strain curves of RC column, confined by 4 layers of PPFR (RCPPL4), under cyclic axial load.

Fig. 13. Stress versus axial and lateral strain curves of RC column, confined by 1 plus 3 pretensioned layers of PPFR (RCPPL13PRE1.98%), under cyclic axial load.

Fig. 14. Stress versus axial and lateral strain curves of RC column, confined by 2 layers of basalt FR (RCBSL2), under cyclic axial load.

before early stop, is far higher than the strain at failure of the basalt rope developed a maximum axial stress of 38.22 MPa at axial strain of
fiber rope, denoting the high redistribution capacity of strains at rope εcu = 5.617% (see Fig. 16), absorbing energy of 1.771 MJ/m3. The test
level, through friction with the dilating concrete core, despite the stopped early without fracture in any part of the basalt rope and
square shape of the section. without temporary load drop. Pretension of the confining BFR resulted
Column RCBSL1 developed a maximum axial stress of 23.75 MPa at similarly in remarkable increase of the bearing stress at the elastic limit
axial strain of εcc1 = 0.634%. Fig. 15 suggests that despite column of the stress-strain response of the column, at 1% axial strain level as
suffered a severe temporary load drop around 21% of fcc1, it managed to well as at 3% axial strain level: from 23.11 MPa to 31.66 MPa, increase
bounce back at 22.33 MPa bearing stress at εcu = 3.991%, absorbing of 25% when compared with RCBSL2 column.
energy of 1.033 MJ/m3 before basalt rope fracture. The gradual fracture Column PCGL1BSL2 with hybrid GFRP – basalt rope confinement
of the rope during the two last cycles is also depicted in the stress-strain developed a maximum axial stress of 30.76 MPa at axial strain of
diagram of Fig. 15. εcu = 6.428% absorbing energy of 2.087 MJ/m3 as shown at Fig. 17.
The test specimen RCBSL12PRE1.87% with pretensioned basalt Then, the test was early stopped. The load dropped twice during the

102
T.C. Rousakis et al. Composites Part B 163 (2019) 96–106

Fig. 15. Stress versus axial and lateral strain curves of RC column, confined by 1 layers of basalt FR (RCBSL1), under cyclic axial load.

Fig. 16. Stress versus axial and lateral strain curves of RC column, confined by 1 plus 2 pretensioned layers of basalt FR (RCBSL12PRE1.87%), under cyclic axial load.

Fig. 17. Stress versus axial and lateral strain curves of concrete column, confined by 2 layers of basalt FR and 1 layer of GFRP (PCGL1BSL2), under cyclic axial load.

experiment (the maximum load drop was 18.1%), corresponding to the test without rope fracture at εcu = 5.170% (Fig. 18). The corresponding
variable fracture of the GFRP jacket. However, the basalt rope managed stress was as high as 28.64 MPa, and the absorbing energy 1.912 MJ/
to control the variably damaged concrete core after the FRP fracture. m3. The load dropped twice during the experiment because of the
The very high recorded lateral strains of the column suggest that the multiple fracture of the GFRP jacket. Intense buckling of the long-
non bonded rope could maintain the confining action despite the local itudinal bars was evident in various regions after the removal of the
fracture and bulging of the FRP jacket (see Fig. 9). rope and fractured FRP jacket. However, the two layers of BSFR man-
The specimen RCGL1BSL2 with internal steel reinforcement and aged to control the variably damaged column. During the test there was
external hybrid confinement developed a maximum axial stress of a transposition of the two linear variable displacement transducers
29.14 MPa at εcc1 = 1.180%. The additional confinement with basalt (LVDT) which provided the lateral deformations of the concrete core
rope resulted in higher and earlier development of the maximum and the measurements were unreliable and thus excluded.
bearing stress than in the RCGL1 column (27.74 MPa at 1.530%). The
bearing stress was maintained at high levels up to the early stop of the

103
T.C. Rousakis et al. Composites Part B 163 (2019) 96–106

lateral strain,
Ultimate

εlu (%)

1.860
3.477

3.592

6.142
3.91
–

–
Ultimate axial
strain, εcu (%)

6.071
3.991
7.341
5.617
6.428
5.170
2.13
6.93
stress, fcu
Ultimate

(MPa)

22.12
29.72
41.52
22.33
29.44
38.22
30.76
28.64
Stress at 3%
axial strain
fcc3 (MPa)
Fig. 18. Stress versus axial strain curves of RC column confined by 1 layer of

23.61
34.03
18.41
23.11
31.66
28.86
25.76
GFRP and 2 layers of BSFR (RCGL1BSL2), under cyclic axial load.

–
strain at fcc2,
3.3. Assessment of confining effects and discussion

εlc2 (%)
Lateral

2.067
2.118
2.167
3.91

0.13
This section explores the effectiveness of the different confining

–
techniques following the approach by Rousakis (2018) [30]. All col-

Axial strain
at fcc2, εcc2
umns are assessed based on their maximum effective retrofit confine-
ment rigidity ratio CRj. It is the maximum effective lateral stress

3.871
3.038
1.654
7.341
5.617
1.743
1.796
(%)
(fljmax) divided by the lateral strain (εl), multiplied by the effectively

–
confined concrete core coefficient (αj) and divided by the plain con-

fcc2, (fcc1-fcc2)/fcc1
between fcc1 and
crete strength (fco), namely CRj=(fljmax/εl)*αj/fco (more details can
be found in Refs. [30,41]. The maximum effective lateral stress depends
Stress drop

on the effective confinement volumetric ratio and the modulus of

21.09

18.12
elasticity and the deformability of the confinement. When the rope is

5.59

5.76
(%)

0.2
pretensioned, the initial strain of the confinement is subtracted from the

0
0
–

lateral strain of the confinement at the maximum bearing stress level.

during load drop
Minimum stress

When hybrid confinement is used, the two different CRj are added.
Based on previous experimental and analytical elaborations of the
fcc2 (MPa)

first of the authors, accurate prediction of the load and deformation at

23.99
34.10
18.74
29.44
38.22
20.52
27.46
failure for rope confined columns or for pretensioned ropes or for hy-
–

brid FRP – rope confinement requires special attention as the delay of

activation of the rope, the pretension level of the rope, the absence of
fcc1,εlc1 (%)
strain at

bond among different interfaces (concrete-rope, FRP-rope etc) and the

Lateral

1.143
0.100
1.972
0.065
3.910

0.598

specific detailing of the different techniques (pretension application or

–

hybridization of confinement) have to be taken into account. The latter

Axial Strain

is of great significance when accurate estimation of deformation at

at fcc1, εcc1

failure is needed. Three-dimensional finite element analysis of the

2.006
2.630
0.634
7.341
5.617
0.987
1.53

1.18

specific retrofit solutions could be utilized to address the above-

(%)

mentioned issues and produce reliable design expressions. However,

first stress reduction

such analyses are out of the scope of this paper focusing mostly on the
Maximum stress at

experimental behavior. Therefore, CRj is utilized to assess the varying

effectiveness of the different columns presented herein with respect to
fcc1 (MPa)

their bearing axial load.

27.74
25.41
34.36
23.75
29.44
38.22
25.06
29.14

Fig. 19a shows the ratio of maximum bearing stress of confined

concrete divided by the plain concrete strength versus CRj for columns
RCGL1, RCPPL4 and RCPPL13PRE1.98%. Retrofit confinement rigidity
confinement volumetric

ratio displays a strong correlation with the concrete strength en-

Effective FR and FRP

hancement for different strengthening techniques such as glass FRP

0.00445 + 0.091
0.00445 + 0.091

jacketing, PPFR confinement and pretensioned PPFR confinement. The

correlation of the CRj with the increased concrete strength for retrofits
ratio ke*ρF

including the basalt fiber rope seems satisfactory (Fig. 19b). However,
0.00445
0.2216
0.2216
0.0466

0.1376
0.091

the delay in activation of the basalt fiber rope in BSFR confinement,

pretensioned BSFR confinement or hybrid GFRP-BSFR wrapping is
Experimental results.

significant.
RCPPL13PRE1.98%

RCBSL12PRE1.87%
Specimen label

RCGL1BSL2
PCGL1BSL2

4. Conclusions
RCPPL4

RCBSL1
RCBSL2
RCGL1
Table 3

The study investigated the experimental performance of externally

strengthened RC columns of square section with low concrete strength

104
T.C. Rousakis et al. Composites Part B 163 (2019) 96–106

Fig. 19. Diagrams of strength increase ratio fcc/fco versus CRj for (a) GFRP and PPFR confined RC columns and (b) for BSFR confined columns.

and slender bars, under pseudoseismic, cyclic compressive load. It ex- pretension detailing and finally to the initial lateral pressure exerted
plored the confining effects of PPFR and BSFR wrapping with and on the concrete core.
without pretension. The pretension of the composite ropes was realized
with the use of special metal pretension devices, enabling for high le- Further, hybrid GFRP – BSFR strengthening was investigated for
vels of pretension without development of friction between the com- plain concrete and reinforced concrete columns. The additional non
posite rope and the concrete core. The pretension technique can de- bonded non impregnated basalt composite rope wrapping of RC column
velop significant lateral restriction of the concrete core of non-circular in RCGL1BSL2 resulted in higher and earlier development of the max-
sections in the region between the corners, where the shape effect re- imum bearing stress than in the case of RCGL1. The bearing stress was
duces the efficiency of passive confinement techniques. The pretension maintained at high levels of 28.6 MPa up to early stop at 5.1% strain
of the wrapping can be easily applied in situ. This paper further extends because of the dual confinement, without basalt rope failure, compared
the application of the pretension technique to low deformability BSFR. with the collapse of the RCGL1 at 2.13% strain. The BSFR could control
Concerning the mechanical behavior of the strengthened columns, the temporary load drop during the fracture of the GFRP jacket
the following main conclusions are drawn: throughout the loading process and reveal high redistribution potential.

- PPFR and BSFR could confine adequately square concrete sections Acknowledgements
with 17 mm corner radius and slender bars, without fracture of the
rope, achieving concrete axial strains higher than 5%. Only column Thanks are owed to Zarras S.A., Thrace Plastic Co. S.A., Juan Gili
RCBSL1 failed with the gradual rupture of the composite rope at 4% S.L. and Sika Hellas ABEE for providing concrete, polypropylene fiber
axial strain, after a temporary load drop of 21% and unique load rope, basalt fiber rope and glass fiber sheets - epoxy resins respectively.
recovery. The analytical elaborations suggest that PPFR confine- Arvanitis L., Mourelatos G., Koukourikos S. and Tzimoulidis E. are also
ment of RC columns reveals efficiency similar to that of GFRP thanked for their help in the construction of the specimens as well as
jacketing in terms of strength, while BSFR confinement is inferior. Tsepis V. and Giouleas M. for their help in tests respectively. Finally,
This is mainly attributed to the delay in activation of the specific thanks are owed to DUTh RC lab staff.
BSFR confinement to provide adequate lateral pressure and re-
striction to the concrete core. However the axial strain enhancement Appendix A. Supplementary data
is remarkable. Both PPFR and BSFR showed in all cases a very high
stress redistribution capacity and preserved full contact with the Supplementary data to this article can be found online at https://
dilating and variably damaged concrete core throughout loading, doi.org/10.1016/j.compositesb.2018.11.024.
achieving remarkable enhancement of axial strain. This feature is
attributed to the nonbonded multi-strand rope wrapping technique. References
- The special devices for the pretension of the composite ropes had
satisfactory performance throughout loading even for high axial [1] Fardis MN, Khalili H. Concrete encased in fibreglass-reinforced- plastic. J. Am.
strains of concrete. The novel pretension technique could achieve Concr. Inst. Proc. 1981;78(6):440–6.
[2] Mirmiran A, Shahawy M, Samaan M, El Echary H, Mastrapa JC, Pico O. Effect of
far higher levels of pretension, limited by the axial strain capacity of columns parameters on FRP-Confined Concrete. J Compos Construct ASCE
the confining reinforcement because they stretch the composites 1998;2(4):175–85.
against the concrete core without interface friction. The pretension [3] Karabinis AI, Rousakis TC. Concrete confined by FRP material: a plasticity ap-
proach”. Eng Struct J 2002;24:923–32. Elsevier.
levels of the BSFR were close to the ultimate deformability provided [4] De Lorenzis L, Tepfers R. Comparative study of models on confinement of concrete
by the manufacturer (stretching of loose rope included). Despite cylinders with fiber-reinforced polymer composites. ASCE J Compos. Constr
this, column RCBSL13PRE1.87% could reach 5.6% axial strain up to 2003;7(3):219–34.
[5] Valdmanis V, De Lorenzis L, Rousakis T, Tepfers R. Behavior and capacity of CFRP-
the early stop of the test without FR fracture. This feature is at- confined concrete cylinders subjected to monotonic and cyclic axial compressive
tributed to nonbonded pretensioned rope confinement as well as to load. Struct Concrete J fib December 2007;8(4). [Thomas Telford, London].
the regions along the column in between the metal devices that [6] Rousakis TC, Karabinis AI, Kiousis PD. FRP-confined concrete members: axial
compression experiments and plasticity modeling. Eng Struct 2007;29(7):1343–53.
could deform more than the pretensioned regions and further extend
[7] Rousakis TC, Karabinis AI, Kiousis PD, Tepfers R. Analytical modelling of plastic
the redistribution mechanisms. behaviour of uniformly FRP confined concrete members. Elsevier. Journal of
- For pretensioned composite rope, the elastic part of the stress-strain Compos B Eng October-December 2008;39(7–8):1104–13.
behavior was remarkably prolonged as well as the bearing stresses [8] Jian-Guo Dai, Yu-Lei Bai, Teng JG. Behavior and modeling of concrete confined
with FRP composites of large deformability. J Compos Construct 2011;15(No.
at 1% and at 3% axial strain levels (25% for BSFR and 40% for 6):963–73.
PPFR) marking the increase of the inelastic stress-strain modulus. [9] Wie Y, Wu Y. Unified stress–strain model of concrete for FRP-confined columns.
This feature is attributed to the earlier activation of the ropes, to the Construct Build Mater 2012;26(1):381–92.
[10] Nisticò N, Monti G. RC square sections confined by FRP: analytical prediction of
increase of the effectively confined concrete core because of the peak strength. Compos B Eng 2013;45(1):127–37.

105
T.C. Rousakis et al. Composites Part B 163 (2019) 96–106

[11] Nisticò N. RC square sections confined by FRP: a numerical procedure for predicting polymer reinforcement for concrete structures. ACI; 2005. p. 1631. SP-230—92.
stress–strain relationships. Compos B Eng 2014;59:238–47. [28] Ciniņa I, Zīle E, Zīle O. Mechanical behavior of concrete columns confined by basalt
[12] Gambarelli S, Nisticò N, Ožbolt J. Numerical analysis of compressed concrete col- FRP windings. Mech Compos Mater November, 2012;48(5). (Russian Original Vol.
umns confined with CFRP: microplane-based approach. Compos B Eng 48, No. 5, September-October, 2012).
2014;67:303–12. [29] Janke L, Czaderski C, Ruth J, Motavalli M. Experiments on the residual load-bearing
[13] Nisticò N, Pallini F, Rousakis T, Wu YF, Karabinis A. Peak strength and ultimate capacity of prestressed confined concrete columns. Eng Struct October
strain prediction for FRP confined square and circular concrete sections. Compos B 2009;31(10):2247–56.
Eng December 2014;67:543–54. [30] Rousakis TC. Inherent seismic resilience of RC columns externally confined with
[14] Pessiki S, Harries KA, Kestner JT, Sause R, Ricles JM. Axial behavior of reinforced nonbonded composite ropes. Compos B Eng 15 February 2018;135:142–8.
concrete columns confined with FRP jackets. J Compos Construct [31] Karantzikis M, Papanicolaou CG, Antonopoulos CP, Triantafillou TC. Experimental
2001;5(4):237–45. investigation of nonconventional confinement for concrete using FRP. J Compos
[15] Matthys S, Toutanji H, Taerwe L. Stress–strain behavior of large-scale circular Construct 2005;9(6):480–7.
columns confined with FRP composites. J Struct Eng 2006;132(1):123–33. [32] Triantafillou TC, Papanicolaou CG. Textile reinforced mortars (TRM) versus fiber
January 1. reinforced polymers (FRP) as strengthening materials of concrete structures. SP230-
[16] Rousakis TC, Karabinis AI. Substandard reinforced concrete members subjected to 6, FRPRCS-7: 7th int. Symp. On FRP reinforcement for concrete structures, ACI SP-
compression - FRP confining effects. RILEM Materials and Structures 230. Farmington Hills, MI: ACI; 2005. p. 99–118.
15.01.2008;41(9):1595–611. Springer Netherlands. [33] Triantafillou TC, Papanicolaou CG, Zissimopoulos P, Laourdekis T. Concrete con-
[17] De Luca A, Nardone F, Matta F, Nanni A, Lignola G, Prota A. Structural evaluation finement with textile-reinforced mortar jackets. ACI Struct J 2006;103(1):28–37.
of full-scale FRP-confined reinforced concrete columns. J Compos Construct [34] Rousakis TC. Reusable and recyclable nonbonded composite tapes and ropes for
2011;15(1):112–23. concrete columns confinement. Compos B Eng 2016;103(15 October):15–22.
[18] Rousakis TC, Karabinis AI. Adequately FRP confined reinforced concrete columns [35] Kwon M, Seo H, Kim J. Seismic performance of RC-column wrapped with velcro.
under axial compressive monotonic or cyclic loading. RILEM Mater Struct July Structural engineering and mechanics. An Int'l Journal 2016;58(2)https://doi.org/
2012;45(Issue 7):957–75. Springer Netherlands. 10.12989/sem.2016.58.2.379.
[19] Wang Z, Wang D, Smith ST, Lu D. Experimental testing and analytical modeling of [36] Shimomura Τ, Phong ΝΗ. Structural performance of concrete members reinforced
CFRP-confined large circular RC columns subjected to cyclic axial compression. Eng with continuous fiber rope. FRPRCS-8 conference university of patras, patras,
Struct 2012;40:64–74. Greece, July 16-18. 2007.
[20] Wang Z, Wang D, Smith ST, Lu D. CFRP-confined square RC columns. I: experi- [37] Rousakis TC. Elastic fiber ropes of ultra high extension capacity in strengthening of
mental investigation. J Compos Construct 2012;16(2):150–60. concrete through confinement. J Mater Civ Eng 2013. https://doi.org/10.1061/
[21] Wang DY, Wang ZY, Smith ST, Yu T. Size effect on axial stress-strain behavior of (ASCE)MT.1943-5533.0000796.
CFRP-confined square concrete columns. Construct Build Mater 2016;118:116–26. [38] Rousakis T. Hybrid confinement of concrete by FRP sheets and fiber ropes under
[22] Bai Y-L, Dai J-G, Teng JG. Buckling of steel reinforcing bars in FRP-confined RC cyclic axial compressive loading. J Compos Construct 2013. https://doi.org/10.
columns: an experimental study. Construct Build Mater 2017;140:403–15. 1061/(ASCE)CC.1943-5614.0000374.
[23] Barros JAO, Ferreira DRSM. Assessing the efficiency of CFRP discrete confinement [39] Rousakis TC, Kouravelou KB, Karachalios TK. “Effects of carbon nanotube enrich-
systems for concrete cylinders. J Compos Construct 2018:134–48. https://doi.org/ ment of epoxy resins on hybrid FRP - FR confinement of concrete.” Elsevier. Journal
10.1061/(ASCE)1090-0268(2008)12:2(134). of Composites Part B: Engineering February 2014;57:210–8.
[24] Triantafyllou GG, Rousakis TC, Karabinis AI. Axially loaded reinforced concrete [40] Rousakis TC, Tourtouras IS. RC columns of square section – passive and active
columns with a square section partially confined by light GFRP straps. ASCE confinement with composite ropes Elsevier Journal of Composites Part B:
Composites for Construction 2015;19(1):04014035https://doi.org/10.1061/ Engineering March 2014;58:573–81.
(ASCE)CC.1943-5614.0000496. [41] Rousakis, T. C., Tourtouras, I. S.. Modeling of passive and active external confine-
[25] Rousakis T.C, Chi-Sang You, Laura de Lorenzis, Tamuzs V, Tepfers R. Concrete ment of RC columns with elastic material. ZAMM Journal, by Wiley-VCH Verlag
cylinders confined by prestressed carbon FRP filament winding, subjected to GmbH & Co. KGaA. Article first published online: 6 AUG 2015. DOI: 10.1002/
monotonic and cyclic axial compressive load. 6th international symposium on fibre- zamm.201500014.
reinforced polymer (FRP) reinforcement of concrete structures (FRPRCS-6), 8 – 10 [42] Wu G, Wu ZS, Lu ZT, Ando YB. Structural performance of concrete confined with
July 2003, Singapore. hybrid FRP composites. J Reinforc Plast Compos 2008;27(12):1323–48.
[26] Tamuzs V, Tepfers R, You Chi-Sang, Rousakis T, Repelis I, Skruls V, Vilks U. [43] Padanattil A, Karingamanna J, Mini KM. Novel hybrid composites based on glass
Behavior of concrete cylinders confined by carbon-composite Tapes and prestressed and sisal fiber for retrofitting of reinforced concrete structures. Construct Build
yarns 1. Experimental data. Mechanics of Composite Materials 2006;42(1):13–32. Mater 2017;133:146–53.
Springer Science+Business Media, Inc.0191–5665. [44] Ispir M, Dalgic KD, Ilki A. Hybrid confinement of concrete through use of low and
[27] Nesheli KN, Meguro K. External prestressing concrete columns with fibrous com- high rupture strain FRP. Compos B Eng 2018;153:243–55. ISSN 1359-8368.
posite belts. SP-230: 7th international symposium on fiber-reinforced (FRP)

106