Construction and Building Materials 95 (2015) 448–456

Contents lists available at ScienceDirect

Construction and Building Materials

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

Experimental study on bond behaviour between BFRP bar

and engineered cementitious composite
Hailong Wang a, Xiaoyan Sun a,⇑, Guangyu Peng a, Yuejing Luo b, Qiming Ying c
a
Department of Civil Engineering, Zhejiang University, Hangzhou 310058, China
b
Guangxi Transportation Research Institute, Nanning 530007, China
c
Zhejiang Provincial Institute of Communication Planning, Design and Research, Hangzhou 310006, China

h i g h l i g h t s

 Effects of various parameters on bond performance between BFRP bar and ECC are presented.
 Change mechanisms of bond behaviours are discussed.
 PVA fibres in ECC matrix enhance the bond performances between BFRP bar and cementitious matrix.
 Some factors are recommended to provide sufficient bond between BFRP bar and ECC matrix.

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

Article history: To understand the bond behaviour between the basalt fibre reinforced plastic (BFRP) bar and the engi-
Received 27 April 2015 neered cementitious composite (ECC), pullout tests were performed in this paper to study the effects
Received in revised form 27 June 2015 of the bar diameter, the embedded length, the cover thickness, and the properties of matrix materials
Accepted 14 July 2015
on the bond performance. The test results indicate that the bond strength is generally controlled by
the shear resistance of the BFRP bar surface layer for most of the specimens with pullout failure. For
the specimens with a cover thickness of 5.5 mm, splitting failure occurred. The bond strength between
Keywords:
the BFRP bar and cementitious matrix decreases with an increase in the bar diameter, and the specimen
Bond strength
Slip
with a shorter embedment length achieves a higher bond strength. A linear equation can be used to
Failure mode describe the relationship between the average bond strength and the embedment length. The bond
BFRP bar strength increases with the increase in cover thickness; however, this increase can be neglected when
ECC the ECC cover thickness exceeds 20 mm. The addition of PVA fibres decreases the damage and enhances
Influence factors the bond performances between the bar and the matrix.
Ó 2015 Published by Elsevier Ltd.

1. Introduction
The performance of reinforced concrete (RC) structure in an
aggressive environment is generally determined by the corrosion
of steel rebar in concrete. To overcome this corrosion issue,
fibre-reinforced polymer (FRP) bar has been considered as a
promising alternative for the conventional steel rebar due to its
high strength-to-weight ratio, excellent corrosion resistance, low
cost, ease of handling, and good fatigue properties [1–3]. To evalu-
ate the crack width and spacing and to understand the load trans-
fer mechanism between FRP bar and concrete, various studies have
been performed to evaluate the bond performance between
⇑ Corresponding author at: Anzhong Bldg B717, Department of Civil Engineering,
Zhejiang University, Hangzhou, Zhejiang Province, China.
E-mail address: selina@zju.edu.cn (X. Sun).
carbon/glass FRP bar and concrete. It is concluded that the param-
eters such as the type of FRP bar, the bar surface, the bar diameter,
the embedment length, the test setup and the concrete strength
have a considerable effect on the bond behaviours [4–8].
As a new type of FRP material, basalt fibre reinforced plastic
(BFRP) bar was recently produced using basalt fibres. In addition
to the excellent corrosion resistance and mechanical performance,
the basalt fibre is environmentally harmless and has a good range
of thermal performance, superior electro-magnetic property, and
good resistance to UV light and impact, which makes it better than
the glass fibre and less expensive than the carbon fibre. Several
studies have been performed on the mechanical properties of
BFRP bar reinforced concrete elements, revealing a good integrity
of these two materials. Lin and Zhang used the four-point bending
beam test to study the flexural and bond–slip behaviour of
FRP-reinforced concrete beams reinforced with CFRP, GFRP, and

http://dx.doi.org/10.1016/j.conbuildmat.2015.07.135
0950-0618/Ó 2015 Published by Elsevier Ltd.
 H. Wang et al. / Construction and Building Materials 95 (2015) 448–456
BFRP bars with different surface conditions. They found that the
BFRP-reinforced concrete beams perform much better than CFRP-
and GFRP-reinforced concrete beams due to the strong bond in
the BFRP-reinforced concrete beam [9]. However, due to the much
lower modulus of BFRP bars compared to steel bars, the deflections
and the average crack widths of beams with BFRP reinforcement
were significantly higher than those of the RC beam [10].
Mahroug et al. developed an experimental study on the mechanical
behaviour of concrete slabs reinforced with BFRP bars, and they
reached the same conclusion as the beam study [11]. Therefore,
the cracks in concrete will result in a serviceability problem for
the BFRP bar reinforced concrete structures. To address this ser-
viceability problem, the addition of fibres in the cement matrix is
believed to be an effective solution on the basis of the existing
studies.
Engineered cementitious composite (ECC) is a new generation
of high performance fibre reinforced concrete with the characteris-
tics of high ductility [12–13]. Due to the bridging effect of struc-
tural fibres, ECC displays a characteristic of multiple stable
micro-cracks, unlike the single unstable cracking of normal con-
crete under tension. In addition, ECC’s high tensile ductility and
self-controlled micro-crack width lead to its good durability under
different mechanical and environmental conditions [14–16]. Some
available studies have provided the advantages of applying ECC
material into structures, which prevent the brittle fracture of ele-
ment. Because BFRP bar and concrete are brittle materials and
the crack width in BFRP bar concrete members is larger than that
in RC structures, ECC can be a good choice to replace the concrete
in the structural member. To develop a good performance of com-
posite structure, the assessment of bond behaviour between BFRP
bars and ECC matrix is a key aspect, but few relevant studies
reported on this issue.
However, several studies have been performed to examine the
effects of structural fibres on the bonding properties between
G/CFRP bars and concrete, which provides insight into the change
mechanism and regulations. Firas et al. conducted a pullout test to
reveal the bond of CFRP bars in ultra-high performance fibre rein-
forced concrete (UHPFRC). Their results clearly indicated that the
use of UHPFRC increased the CFRP bar/concrete interface shear
strength, and the bond failure was caused by the inter-laminar
rupture mechanism. Due to the high autogenous shrinkage of
UHPFRC, the ultimate bond strength was mostly expected during
early age (3 days) [17]. Kim et al. investigated the pullout beha-
viours of GFRP bars as well as steel bars in concretes reinforced
with steel, PP and PVA fibres. Their results indicated that the struc-
tural fibres in the interface changed the bond behaviours before
and after the maximum stress and resulted in significant improve-
ment of the relative bond strength [18], which agreed with the test
results reported by Won et al. [19] and Mazaheripour et al. [20].
However, Belabi and Wang found that the fibres in concrete had
a moderate improvement in the ductility but no distinct effect on
the bond strengths between C/GFRP bars and concrete [21].
To better understand the bond behaviour between the BFRP bar
and ECC, pullout tests were performed in this study to reveal the
bond change mechanism and regulations. The influences of varying
parameters, such as the bar diameter, the embedded length, the
cover thickness, and the properties of matrix materials, on the
bond performance are discussed on the basis of the relationship
between the average bond stress and slips.
2. Experimental programme
2.1. Materials
Three diameters of BFRP bars, namely 4 mm, 10 mm, and 16 mm, were consid-
ered in this study for the direct pullout test. These BFRP bars were manufactured by
the pultrusion process and are composed of basalt fibre and thermosetting resin. As
449
shown in Fig. 1, the outer layer of each BFRP bar was wrapped with a basalt fibre
braid to protect the surface and enhance the bonding strength between the BFRP
bar and the cementitious materials. The BFRP bars exhibit linear elastic behaviour
up to brittle failure. The elastic modulus of BFRP bar is 50.6 GPa, and the tensile
strength is 750 MPa.
To reveal the effect of the matrix performance on the bond behaviour, two types
of cementitious materials, ECC and cement mortar (CEM), were designed in the
study. The cement used for cementitious materials was a P.O.42.5R ordinary
Portland cement that complies with China code GB8076-1997. A type of local fine
sand with an average size of 150 lm was used as fine aggregate. The water–binder
ratio was 0.3 for all materials. To produce a high toughness ECC material, a certain
amount of fly ash (FA), silica fume (SF) and metakaolin (MK) was used in the
cementitious binder, and a type of polyvinyl alcohol (PVA) fibre was used to rein-
force the cementitious composites with a volume fraction of 2%. The properties of
PVA fibre are presented in Table 1. In addition, an SP8-CR water reducer was used
to optimise the workability of the ECC. The mixture proportions are presented in
Table 2.
Uniaxial tensile tests were conducted on the specimens with dimensions of
400 mm  100 mm  15 mm to characterise the tensile behaviours of the ECC
and cement mortar, as shown in Fig. 2. The tensile stress–strain curves of ECC spec-
imens are shown in Fig. 3. A strain-hardening characteristic is displayed for the ECC
material with a strain capacity of approximately 2.0%. The average tensile strength
of ECC is 3.97 MPa, while it is 3.46 MPa for the cement mortar. The ECC exhibits bet-
ter ductility than the cement mortar with a strain capacity of 0.024%. These ECC
materials also exhibit saturated multiple cracking with a crack width at ultimate
strain limited to below 175 lm and an average crack width below 115 lm. In addi-
tion, a number of cubic specimens with sizes of 150 mm  150 mm  150 mm were
also casted and tested in compression. The compressive strengths of ECC and CEM
are 45.8 and 49.3 MPa, respectively.
2.2. Specimens
The direct pullout test was used in this study to evaluate the bond behaviours
between the BFRP bars and the cementitious matrix. The prismatic specimens were
100 mm  15 mm  150 mm,100 mm  20 mm  150 mm,100 mm 30 mm  150
mm,100 mm  40 mm  150 mm,100 mm  50 mm  150 mm,100 mm 100 m-
m  150 mm, and 150 mm  150 mm  150 mm in size, with a single BFRP bar
embedded along the central axis, as shown in Fig. 4. The variables considered in this
experiment were the bar diameter d (4 mm, 10 mm, and 16 mm), the embedded
length la (2.5d, 5d, 10d and 15d), the cover thickness c (5.5 mm, 8 mm, 13 mm,
18 mm, 23 mm, 48 mm, 67 mm, 70 mm and 73 mm), and the properties of cementi-
tious materials (ECC and CEM). As summarised in Table 3, three specimens of each
type and a total of 51 specimens were prepared for the pullout tests.
The specimens were cast in prismatic moulds with a BFRP bar horizontally
placed at the centre. As shown in Fig. 4, the debonded parts of the specimen were
created using two soft plastic tubes to minimise the stress concentration near the
(a) 4mm (b) 10mm (c) 16mm
Fig. 1. Characteristics of BFRP bars.
Table 1
Properties of PVA fibre.
Length Diameter Nominal Elongation Elastic Density
(mm) (lm) strength (%) modulus (g/cm3)
(MPa) (GPa)
12 39 1620 7 42.8 1.3
450 H. Wang et al. / Construction and Building Materials 95 (2015) 448–456

Table 2
Mix proportions of ECC and cement mortar.

Mix type Cement (kg/m3) Water (kg/m3) Sand (kg/m3) FA (kg/m3) SF (kg/m3) MK (kg/m3) PVA fiver (kg/m3)
CEM 1233 370 617 0 0 0 0
ECC 432 370 617 617 62 124 26

Plastic tube Bond zone

150

Fig. 4. Test specimen (Unite:mm).

2.3. Test setup

The BFRP bar is sensitive to fracture under lateral pressures. Therefore, a

wedge-type anchorage system was developed to clamp the BFRP bar in the pullout
test, as shown in Fig. 5(a). The system consisted of three pieces of steel wedges and
an outer steel anchor cup. This anchorage system was installed at the loaded end of
the BFRP bar, as indicated in Fig. 5(b).
The setup for the pullout tests is shown in Fig. 6. The loads were applied to the
specimens by a 2500 kN capacity servo-hydraulic testing machine with a loading
rate of 0.1 mm/min. The load and the slips were recorded in the pullout test to
obtain the load–slip relationship. The load was measured by the load cell of the test
machine, while the loaded and free end slips were measured using four linear vari-
able differential transformers (LVDTs) as shown in Fig. 6. The relative displacement
between the BFRP rebar and the cementitious matrix can be calculated by the dif-
ferences in the displacements. A data acquisition system was used to collect the test
data automatically.

3. Test results and discussion

Fig. 2. Direct tensile tests of cementitious materials (Unite:mm).
The stress distribution is not constant along the embedment
length of the BFRP bar. Therefore, an average bond stress s during
the pullout test is defined as follows:
P
s¼ ð1Þ
pdla
where P is the pullout load, d is the BFRP bar diameter, and la is the
embedment length. The bond strength was calculated by Eq. (1)
using the ultimate pullout load Pmax. The experimental results are
summarised in Table 4.

3.1. Bond stress-global slip curve

The specimen with the strength most similar to the average

Fig. 3. Stress–strain relationship of ECC under tension.
boundaries. These specimens were demoulded after 24 h and then stored in a stan-
dard curing room for 28 days with a temperature of 20 ± 2 °C and a relative humid-
ity above 95%.
strength of three specimens was selected as the representative
specimen. The representative bond stress–slip curves for different
ECC-BFRP bar specimens are illustrated in Fig. 7. Generally, the
bond stress–slip curve displays three distinct stages: the linear,
nonlinear, and descending stages. At the beginning, the slips
increase linearly with the increase in the pullout stress, but the
bond stress increases rapidly while the slips increase slowly. At
this first stage, the adhesion of cementitious material and the
mechanical interlock between the bar and the matrix makes a
great contribution to this characteristic. The slip at the free end
lags the one at the loaded end as shown in Fig. 7, indicating that
the damage for debonding begins from the loaded end and then
 H. Wang et al. / Construction and Building Materials 95 (2015) 448–456 451

Table 3
Summary of tested specimens.

Specimens ID Matrix material Bar diameter d (mm) Specimen size (mm) Cover thickness c (mm) Embedment length (mm) Specimen number
ECC15-04-10 ECC 4 100  15  150 5.5 10(2.5d) 3
ECC15-04-20 ECC 4 100  15  150 5.5 20(5d) 3
ECC15-04-40 ECC 4 100  15  150 5.5 40(10d) 3
ECC20-04-20 ECC 4 100  20  150 8 20(5d) 3
ECC20-04-40 ECC 4 100  20  150 8 40(10d) 3
ECC20-04-60 ECC 4 100  20  150 8 60(15d) 3
ECC30-04-20 ECC 4 100  30  150 13 20(5d) 3
ECC40-04-20 ECC 4 100  40  150 18 20(5d) 3
ECC50-04-20 ECC 4 100  50  150 23 20(5d) 3
ECC100-04-20 ECC 4 100  100  150 48 20(5d) 3
ECC100-04-60 ECC 4 100  100  150 48 60(15d) 3
ECC150-04-20 ECC 4 150  150  150 73 20(5d) 3
ECC150-10-50 ECC 10 150  150  150 70 50(5d) 3
ECC150-16-80 ECC 16 150  150  150 67 80(5d) 3
CEM15-04-20 CEM 4 100  15  150 5.5 20(5d) 3
CEM50-04-20 CEM 4 100  50  150 23 20(5d) 3
CEM100-04-20 CEM 4 100  100  150 48 20(5d) 3

Fig. 5. A wedge-type anchorage at the loaded end of BFRP bar.

transfers gradually to the free end. The bond stress is not uniform
along the embedment length, which will cause some effects on the
bond behaviours. After the first stage, the bond stress–slip curve
exhibits a nonlinear response up to the peak bond strength. Due
to the development of debonding damage, the slip increases with
the increasing pullout load, which causes a change in the slope
of the curve. At this stage, the friction force and the mechanical
interlock produce the resistance to the pullout. Subsequently, the
post peak phase appears with a characteristic of a descending
branch of the bond stress with the increase in the slips. At this
descending stage, it is still the friction force and the mechanical
interlock causing the resistance to the external load. Due to the
good confinement and the strong shear resistance of the ECC cover,
a relatively high residual bond stress still exists at the post-peak
phase.
In the case of ECC specimens with the smallest thickness of
5.5 mm, splitting failure of the ECC matrix occurred, which causes
a slight difference in the bond stress–slip curve compared with
other BFRP bar-ECC specimens. Due to the cracking of surrounding
matrix, a longer linear ascending branch and a shorter nonlinear
response are exhibited at the first and second stages. After the peak
stress, a sharper decrease is exhibited in the curve compared with
the other ECC specimens.

3.2. Failure modes

Fig. 8 shows the bond failure modes in this study. It can be

observed from Fig. 8 that the bond failure mode is mainly related
to the cover thickness and the diameter of the BFRP bar. A pullout
failure mode without splitting the cover occurs in most specimens
Fig. 6. Test setup. when they have a sufficient cover thickness. As illustrated in
452 H. Wang et al. / Construction and Building Materials 95 (2015) 448–456

Table 4
Bond test results.

Specimens ID Average bond Free end slip at ultimate Loaded end slip at ultimate Failure modea
strength/MPa pullout load/mm pullout load/mm
ECC15-04-10 13.15 1.02 1.95 P&S
ECC15-04-20 12.34 1.12 2.34 P&S
ECC15-04-40 10.52 1.48 2.16 P&S
ECC20-04-20 12.41 2.59 3.14 PO
ECC20-04-40 10.68 3.10 3.67 PO
ECC20-04-60 9.11 3.54 4.06 PO
ECC30-04-20 13.16 2.34 3.08 PO
ECC40-04-20 13.67 3.16 3.96 PO
ECC50-04-20 14.61 3.91 4.24 PO
ECC100-04-20 14.02 2.99 3.59 PO
ECC100-04-60 – – – R
ECC150-04-20 14.29 3.06 3.73 PO
ECC150-10-50 13.34 3.03 4.01 PO
ECC150-16-80 11.33 3.34 3.90 PO
CEM15-04-20 9.06 1.22 2.05 P&S
CEM50-04-20 12.95 2.25 2.82 PO
CEM100-04-20 13.03 2.73 2.93 PO
a
Failure mode: PO—pullout; P&S—pullout with matrix splitting; R—bar rupture.

3.3. Bar diameter

The bond behaviours between ECC and 4-mm, 10-mm, and

Fig. 7. Representative bond stress-slip curves, (a) ECC100-04-20, (b) ECC30-04-20.
Fig. 8(a), the surface of the bar was peeled off. In the case of the
specimens with a cover thickness of 5.5 mm, splitting failure
occurs because of the insufficient shear resistance of the cover. In
this failure mode, a single crack appears in the mortar and the
ECC materials along the BFRP bar, as shown in Fig. 8(b), resulting
in a negative effect on the bond behaviour. However, the crack
width in the ECC material is smaller than that in the mortar due
to the bridging effect of PVA fibres in the crack. For the specimens
of ECC100-04-60, the BFRP bar was ruptured before pullout
occurred. It seems that the embedded length of 15d is sufficient
to anchor the 4-mm BFRP bar in the ECC matrix.
16-mm BFRP bars are shown in Fig. 9, and the change tendency
of bond strengths is illustrated in Fig. 10. The average bond
strength is found to decrease with the increase in the bar diameter.
This tendency agrees well with the results between CFRP/GFRP
bars and concrete reported by other researchers [6,22–23].
Compared with the 10-mm and 16-mm BFRP bars, the bond
strength of the 4-mm bar is increased by 5% and 23.7%,
respectively.
The average bond strength depends on the bond stress distribu-
tion in the BFRP bar. From Fig. 7, the bond stress is not uniform
along the BFRP bar during the pullout test in that the stress
migrates gradually from the loaded end to the free end of the
bar. Therefore, larger diameters require larger embedment lengths
to achieve the same normal bond stress due to this nonlinear stress
distribution phenomenon, which results in a lower bond strength
for the specimen with a larger diameter BFRP bar. Furthermore, a
shear lag phenomenon can also be used to explain the reduction
of the bond strength with the increase in bar diameter.
According to the results in other similar studies [23–24], a
non-uniform distribution of normal stresses appears through the
bar cross-section due to the different movement between the core
and the surface fibres under the surface tension load. The larger the
diameter, the more non-uniform the stress distribution appears.
The Poisson effect is also another dominant factor for the reduction
of bond strength. Due to the Poisson effect, a slight shrinkage in
diameter occurs when the bar is subjected to pullout stress. The
shrinkage (in absolute value) increases with the increase in bar
diameter, which leads to the reduction of the frictional or mechan-
ical locking loads during the pullout test [23]. In addition, the size
effect confirmed in the concrete interface research can also be
accepted to explain this reduction phenomenon. There are many
voids in the interface zone between the BFRP bar and cementitious
materials. The larger the diameter, the more and larger are the
voids that develop, which lowers the bond strength of the test
specimen.
3.4. Cover thickness
Fig. 11 compares the average bond strengths of ECC specimens
with different cover thicknesses. The figure indicates that the aver-
age bond strength increases with the cover thickness; this
 H. Wang et al. / Construction and Building Materials 95 (2015) 448–456 453

(a)

(b)

CEM ECC

(c)

Fig. 8. Bond failure mode (a) pullout, (b) pullout with matrix splitting, (c) bar rupture.

Fig. 9. Bond stress versus free and loaded end slips for 4 mm, 10 mm and 16 mm BFRP bar.

behaviour occurs because the greater cover thickness can promote
the confinement to the bar and also the shear resistance of the
cementitious cover. However, this increase can be neglected when
the ECC cover thickness is larger than 20 mm because the pullout
force is not affected by the splitting of the cover. As shown in
Fig. 8, the cementitious matrix for the specimen with a 5.5-mm
cover was split, but no obvious damage appears in the other spec-
imen’s cover. Therefore, to attain the maximum bond strength,
adequate confinement surrounding the bar should be provided,
and the recommend value for the ratio of c (the ECC cover thick-
ness) to d (the bar diameter) should be larger than 5.
3.5. Embedment length
Fig. 12 shows the bond stress–slip relationship for the ECC20
specimens with different embedment lengths (5d, 10d and 15d).
The free end slip at the peak stress indicates that the embedment
length in this study is not sufficient to utilise the tensile strength
of the BFRP bar. As shown in Fig. 12, the loaded end slip at the max-
imum pullout force increases with the embedment length, while
there are insignificant changes in the free end slip at the maximum
pullout force. Fig. 13 presents the evolution of the average bond
strength with the embedment length for the ECC20 and ECC15
454 H. Wang et al. / Construction and Building Materials 95 (2015) 448–456
Fig. 10. Effect of BFRP bar diameter on average bond strength.
Fig. 11. Effect of cover thickness on average bond strength.
specimens. The experimental result indicates that the specimen
with the shorter embedment length achieves a higher level of bond
strength. A linear relationship appears between the average bond
strength and the embedment length, which has been reported by
many studies [17,25].
Fig. 12. Bond stress versus free and loaded end slips for different embedment length.
Fig. 13. Effect of embedment length on average bond strength.
The reduction of bond strength with embedment length is
caused by the nonlinear distribution of the bond stress along the
embedment length and the Poisson effect [24]. A larger load is
required to pull out the BFRP bar for a longer embedment length.
As a result of the Poisson effect, this longer embedment length
leads to a greater reduction in the confinement and the mechanical
locking, followed by a reduction of the bond strength.
3.6. Matrix performance
Fig. 14 illustrates the influence of matrix performance on the
bond behaviours during the pullout test. The PVA fibres in the
ECC are observed to have an obvious effect on the pullout curve.
As is known, the damage in the specimen is accumulated with
the development of cracks under pullout force; however, the pres-
ence of PVA fibre slows down and limits the cracks’ growth by its
bridging effect. For the favourable contribution of the bridging
action and the confinement, the ECC specimen exhibits a good
toughness, in particular for the specimens with a cover thickness
of 5.5 mm. In this case, the softening curve of the CME specimen
slows down quickly, whereas there is a gradual slope in the ECC
specimen as shown in Fig. 15.
The comparison of the average bond strengths in the ECC and
CEM specimens is shown in Fig. 16. The bond strength of the ECC
specimen is found to be higher than that of the CEM specimen at
the same embedment length and cover thickness. Compared with
 H. Wang et al. / Construction and Building Materials 95 (2015) 448–456
Fig. 14. Bond stress versus free and loaded end slips for 100  50  150 specimen.
Fig. 15. Bond stress versus free and loaded end slips for 100  15  150 specimen.
Fig. 16. Effect of matrix material on average bond strength.
the C50-04-20 and C100-04-20 specimens, the bond strengths of
E50-04-20 and E100-04-20 are increased by 12.8% and 7.6%,
respectively. As shown in Fig. 3, the addition of PVA fibres in mor-
tar enhances its tensile strength, toughness and resistance to crack
propagation. Therefore, the bridging effect of PVA fibres decreases
the damage in the matrix and at the interface during the pullout
process, thus increasing the shear resistance and the friction force
to enhance the bond strength.
455
4. Conclusions
The pullout test was performed in this paper to evaluate the
influences of the bar diameter, the embedded length, the cover
thickness, and the properties of matrix materials on the bond beha-
viours between BFRP bars and cementitious materials. On the basis
of the experimental results and the analysis, the following conclu-
sions can be drawn:
The failure modes depend on the cover thickness. Splitting fail-
ure occurred in the case of specimens with a cover thickness of
5.5 mm. However, with the addition of PVA fibres, the width of
the single crack that appeared in the ECC matrix was much smaller
than that in the mortar. When the bar was surrounded with a suf-
ficient amount of matrix, pullout failure without splitting the cover
was observed. In this case, the surface of the bar was peeled off
because the bond strength is generally controlled by the shear
resistance of the BFRP bar surface layer.
The bond strength between the BFRP bar and the cementitious
matrix decreases with the increase in the bar diameter due to the
combined effect of the uniform stress distribution along the bar,
the shear lag through the bar cross-section, the Poisson effect,
and the size effect.
The specimen with the shorter embedment length achieves a
higher level of bond strength, and a linear equation can be used
to describe the relationship between the average bond strength
456 H. Wang et al. / Construction and Building Materials 95 (2015) 448–456
and the embedment length. It seems that the embedded length of
15d is sufficient to anchor the 4-mm BFRP bar in the ECC matrix.
The bond strength increases with the cover thickness; however,
this increase can be neglected when the ECC cover thickness is lar-
ger than 20 mm because the pullout force is not affected signifi-
cantly by the splitting of the cover. Therefore, a cover thickness
of 20 mm will be sufficient for the BFRP bar-ECC structure.
The bridging effect of PVA fibres limits the cracks in the matrix
and decreases the damage at the interface during the pullout pro-
cess, which enhances the bond performance between the BFRP bar
and the ECC matrix.
Acknowledgements
The authors are grateful for the financial support received from
the National Natural Science Foundation of China (51178417,
51378456), the Science Foundation of Zhejiang Province
(2015C33027), the Science Foundation of Guangxi Province and
Zhejiang Communication Department.
References
[1] H.A. Abdalla, Evaluation of deflection in concrete members reinforced with
fiber reinforced polymer (FRP) bars, Compos. Struct. 56 (1) (2002) 63–71.
[2] R. Nayal, H.A. Rasheed, Tension stiffening model for concrete beam reinforced
with steel and FRP bars, J. Mater. Civ. Eng. 18 (6) (2006) 831–841.
[3] H.A. Rasheed, R. Nayal, H. Melhem, Response pre-diction of concrete beams
reinforced with FRP bars, Compos. Struct. 65 (2) (2004) 193–204.
[4] M.A. Aiello, M. Leone, M. Pecce, Bond performances of FRP rebars reinforced
concrete, J. Mater. Civ. Eng. 19 (3) (2007) 205–213.
[5] J.J. Zheng, J.G. Dai, Analytical solution for the full-range pull-out behavior of
FRP ground anchors, Constr. Build. Mater. 58 (5) (2014) 129–137.
[6] K. Cho, J.R. Cho, W.J. Chin, B.S. Kim, Bond-slip model for coarse sand coated
interface between FRP and concrete from optimization technique, Comput.
Struct. 84 (7) (2006) 439–449.
[7] J.J. Zheng, J.G. Dai, Prediction of the nonlinear pull-out response of FRP ground
anchors using an analytical transfer matrix method, Eng. Struct. 81 (2014)
377–386.
[8] A. Katz, Bond mechanism of FRP rebars to concrete, Mater. Struct. 32 (12)
(1999) 761–768.
[9] X.S. Lin, Y.X. Zhang, Bond–slip behaviour of FRP-reinforced concrete beams,
Constr. Build. Mater. 44 (2013) 110–117.
[10] M. Urbanski, A. Lapko, A. Garbacz, Investigation on concrete beams reinforced
with basalt rebars as an effective alternative of conventional R/C structures,
Procedia Eng. 57 (2013) 1183–1191.
[11] M.E.M. Mahroug, A.F. Ashour, D. Lam, Experimental response and code
modelling of continuous concrete slabs reinforced with BFRP bars, Compos.
Struct. 107 (2014) 664–674.
[12] M. Li, V.C. Li, High-early-strength ECC for rapid durable repair: material
properties, ACI Mater. J. 108 (2011) 3–12.
[13] V.C. Li, Tailoring ECC for special attributes: a review, Int. J. Concr. Struct. Mater.
6 (3) (2012) 135–144.
[14] H.M.E.D. Afefy, M.H. Mahmoud, Structural performance of RC slabs provided
by pre-cast ECC strips in tension cover zone, Constr. Build. Mater. 65 (2013)
103–113.
[15] G. Fischer, V.C. Li, Intrinsic response control of moment resisting frames
utilizing advanced composite materials and structural elements, ACI Struct. J.
100 (2) (2003) 166–176.
[16] K. Yu, J. Dai, Z. Lu, C. Leung, Mechanical Properties of Engineered Cementitious
Composites Subjected to Elevated Temperatures, J. Mater. Civ. Eng. (2014),
http://dx.doi.org/10.1061/(ASCE)MT.1943-5533. 0001241, 04014268.
[17] S.A. Firas, F. Gilles, L.R. Robert, Bond between carbon fibre-reinforced polymer
(CFRP) bars and ultra high performance fibre reinforced concrete (UHPFRC):
experimental study, Constr. Build. Mater. 25 (2011) 479–485.
[18] B. Kim, J.H. Doh, C.K. Yi, J.Y. Lee, Effects of structural fibers on bonding
mechanism changes in interface between GFRP bar and concrete, Compos. Part
B 45 (2013) 768–779.
[19] J.P. Won, C.G. Park, H.H. Kim, S.W. Lee, C. Won, Effect of fibers on the bonds
between FRP reinforcing bars and high-strength concrete, Compos. Part B 39
(2008) 749–755.
[20] H. Mazaheripour, J.A.O. Barros, J.M. Sena-Cruz, M. Pepe, E. Martinelli,
Experimental study on bond performance of GFRP bars in self-compacting
steel fiber reinforced concrete, Compos. Struct. 95 (2013) 202–212.
[21] A. Belabi, H. Wang, Bond-slip response of FRP reinforcing bars in fiber
reinforced concrete under direct pullout, in: Proceeding of ICFRC-International
Conference on Fiber Composites, High Performance Concretes and Smart
Materials, Chennai, India, 2004, pp. 409–419.
[22] E. Cosenza, G. Manfredi, R. Realfonzo, Behavior and modeling of bond of FRP
rebars to concrete, J. Compos. Constr. 1 (2) (1997) 40–51.
[23] M. Baena, L. Torres, A. Turon, C. Barris, Experimental study of bond behaviour
between concrete and FRP bars using a pull-out test, Compos. Part B 40 (2009)
784–797.
[24] Z. Achillides, K. Pilakoutas, Bond behaviour of fiber reinforced polymer bars
under direct pullout conditions, J. Compos. Constr. 8 (2) (2004) 173–181.
[25] B. Tighiouart, B. Benmokrane, U.D. Gao, Investigation of bond in concrete
member with fibre reinforced polymer FRP bars, Constr. Build. Mater. 12 (8)
(1998) 453–462.