Construction and Building Materials 163 (2018) 277–285

Contents lists available at ScienceDirect

Construction and Building Materials

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

Effect of temperature on bond characteristics of geopolymer concrete

Hai Yan Zhang a, Venkatesh Kodur b,⇑, Bo Wu a, Jia Yan a, Zhen Sheng Yuan a
a
State Key Laboratory of Subtropical Architecture Science, South China University of Technology, Guangzhou, PR China
b
Department of Civil and Environmental Engieering, Michigan State University, East Lansing, MI, USA

h i g h l i g h t s

Bond between geopolymer concrete and rebar at elevated temperatures are quantified.

Bond characteristics of geopolymer concrete is compared with conventional concrete.

Temperature induced bond strength degradation in geopolymer concrete is evaluated.

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

Article history: This paper presents experimental results on the bond behavior between geopolymer concrete and rebar.
Received 3 July 2017 Pull-out tests on geopolymer concrete specimens embedded with plain and ribbed rebars were carried
Received in revised form 4 November 2017 out at ambient temperature and after exposure to 100, 300, 500 and 700 °C. Two batches of geopolymer
Accepted 6 December 2017
concrete with compressive strength of 48 and 64 MPa respectively, and five rebar diameters (of 10, 12,
Available online 18 December 2017
14, 18 and 25 mm) were used for preparing the test specimens. Comparative benchmark tests were also
conducted on ordinary Portland cement (OPC) concrete specimens. Results from these tests show that
Keywords:
geopolymer concrete exhibits insignificant reduction in bond strength till exposure to 300 °C, but under-
Bond characteristics
Geopolymer concrete
goes significant degradation beyond 300 °C. Data from the tests indicate that rate of bond strength degra-
Rebar dation in geopolymer concrete is close to that of splitting tensile strength, but higher than that of
High temperature compressive strength. Also, results infer that geopolymer concrete exhibits similar or better bond prop-
Pull-out test erties than OPC concrete, both at ambient temperature and after exposure to elevated temperatures. Thus
geopolymer concrete can be a practical alternative to OPC concrete in reinforced concrete structures
when fire resistance is one of the main design considerations.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction There is an estimated 44–64% reduction in greenhouse gas emis-

Concrete industry is believed to be one of the major contribu-
tors to global warming, and this is mainly attributed to the use
of Portland cement as the binder. Cement production is thought
to be responsible for about 8% of global CO2 emission, based on
production data in 2014 [1]. To reduce the environmental impact
of concrete industry, efforts are on to find alternatives to Portland
cement. Geopolymer, a new environmentally friendly inorganic
binder, derived by alkaline solution activating aluminosilicate
source material (such as metakaolin, fly ash and slag), has attracted
significant attention in recent years as a practical alternative to
Portland cement [2–5]. With efficient use of industrial by-
products, geopolymer binder greatly reduces greenhouse gas
(CO2) emissions and energy requirements during its production.
⇑ Corresponding author.
E-mail address: kodur@egr.msu.edu (V. Kodur).
sions for a typical Australian geopolymer product, compared with
ordinary Portland cement (OPC) [6].
Geopolymer concrete is reported to have comparable mechani-
cal properties as that of OPC concrete. A great amount of experi-
mental work in the literature has showed that fly ash based
geopolymer concrete, cured at 60–80 °C temperature, exhibit a
high early mechanical strength and low dry shrinkage [7], excel-
lent fire resistance [8,9], and good durability properties [10,11].
In the case of curing at ambient temperature, fly ash based nano-
silica modified geopolymer concrete [12], and metakaolin (MK)-
fly ash (FA) based geopolymer concrete [13], also show excellent
mechanical strength, compared to conventional heat cured
geopolymer concrete and OPC concrete.
For reinforced concrete to be effective as a composite material,
reinforcing bar is to be well bonded to the surrounding concrete.
Therefore, evaluating bond behavior between geopolymer concrete
and reinforcing bar is critical for the use of geopolymer concrete as

https://doi.org/10.1016/j.conbuildmat.2017.12.043
0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.
278 H.Y. Zhang et al. / Construction and Building Materials 163 (2018) 277–285
an alternative to OPC concrete in reinforced concrete structures.
Many attempts to evaluate bond characteristics between geopoly-
mer concrete and reinforcing bar are reported in the literature
[14–18], but the experimental data on bond behavior of geopoly-
mer concrete with reinforcing bar at elevated temperatures are
not established. A recent study [19] investigated the influence of
compressive strength of geopolymer concrete, diameter of
reinforcing bar, concrete cover thickness and anchorage length
on bond behavior between MK-FA based geopolymer concrete
and reinforcing bar at room temperature. The current study
presents experimental results on bond behavior of MK-FA based
geopolymer concrete with reinforcing bar after exposure to ele-
vated temperatures. These experimental results are of great signif-
icance for the use of geopolymer concrete in building applications
where provision of fire resistance is a major design requirement.
2. Experimental program
Direct pull-out tests were conducted on 90 geopolymer con-
crete cubic blocks embedded with reinforcing bars of different
diameters, at ambient and after exposure to elevated tempera-
tures, to evaluate the effect of temperature on bond characteristics
of geopolymer concrete. Comparative benchmark bond tests were
also conducted on 30 OPC concrete specimens.
2.1. Raw materials
Geopolymers used in this study are derived by alkaline-silicate
solution activating metakaolin (MK) and fly ash (FA) blend. The
chemical composition and particle sizes of MK and FA are detailed
in Ref. [20]. The alkaline-silicate activator with desired SiO2/K2O
molar ratio of 1.0 was formulated by blending commercial
potassium silicate solution with 15.8 wt% K2O, 24.2 wt% SiO2 and
60 wt% H2O (SiO2/K2O molar ratio is 2.4), and potassium hydroxide
flakes with 95% purity, and tap water. The alkaline-silicate activa-
tor was prepared one day prior to use.
For generating comparative benchmark data, specimens made
of ordinary Portland cement (OPC, Grade P.O.32.5) concrete were
also tested. To enhance the strength of OPC concrete, polycarboxy-
late superplasticizer was added in the preparation of OPC concrete
specimens.
The coarse aggregates for geopolymer and OPC concretes con-
sisted of graded gravel with sizes of 10–20 mm and fine aggregates
consisted of locally available river sand with a maximum size
of 2 mm.
Six groups of rebars, with different surface (plain and ribbed
rebars) and different nominal diameters (10, 12, 14, 18 and 25
mm), were used in geopolymer concrete blocks for pull-out tests.
The geometric characteristics of ribbed rebars were measured
through a vernier caliper. Samples of rebars were tested through
a universal material testing machine to obtain their actual yield
and ultimate strength. Full details of rebars, together with their
mechanical properties, are presented in Table 1.
Table 1
Geometric characteristics and mechanical properties of rebars.
Group Surface Rib height/spacing Rib phase angle (°)
P-10 Plain – –
R-10 Ribbed 0.14 45
R-12 Ribbed 0.14 62
R-14 Ribbed 0.14 45
R-18 Ribbed 0.15 45
R-25 Ribbed 0.16 45
2.2. Mix proportions
Two batches of geopolymer concretes (GC1 and GC2) and two
batches of OPC concretes (CC1 and CC2) were used for preparing
pull-out test specimens. The mix proportions of geopolymer con-
crete and OPC concrete are tabulated in Table 2. Geopolymer con-
cretes were derived by adding alkaline activator into metakaolin
(MK) and fly ash (FA) blend precursor, and then mixed with coarse
and fine aggregates. Details on preparing geopolymer concrete can
be found in Ref. [20]. Slumps of geopolymer concretes GC1 and GC2
are 173 mm and 97 mm respectively, measured through casting
geopolymer concretes into a standard truncated conical mold.
2.3. Specimen preparation and test procedure
Pull-out tests were carried out as per specifications in Chinese
GB 50152-92 [21] and RILEM standard [22]. The sizes of a typical
test specimen are shown in Fig. 1. The rebar was uniaxially embed-
ded in the center of a geopolymer or OPC concrete cubic block of
size 150 mm  150 mm  150 mm, with a bond length of 5d,
where d is the diameter of the rebar. Six groups of reinforcing bars,
listed in Table 1, were used in geopolymer concrete specimens, but
only R-12 and R-14 rebars were used in OPC concrete specimens.
To prevent local stress concentration at the loading end of the
pull-out rebar, the rebar near the loading end was encased in a
plastic tube with a length of (150-5d) (see Fig. 1).
It is reported that the chemical reaction in geopolymers is a
rapid polymerization process and geopolymers can gain high
strength at early curing age [20,23]. Fig. 2 presents the comparison
on compressive strength development with curing age of geopoly-
mer concrete tested in authors’ previous study and that of OPC
concrete calculated by the proposed empirical formula in Ref.
[24]. It can be seen from Fig. 2 that the compressive strength of
geopolymer concrete at 7-days reached 96% of that at 28-days,
which is much higher than the strength development rate of OPC
concrete at 7-days (account for 59% of 28-day strength). Therefore,
geopolymer concrete specimens were only cured for 7 days in a
tank at a constant 22 °C temperature and 95% humidity before
undertaking pull-out tests. OPC concrete pull-out specimens were
cured under same conditions for 28 days.
The reinforcing bar in test specimens was gripped at the bottom
by a UTM5205X universal material testing machine, and secured at
the top through a specially designed loading frame. Two displace-
ment meters were placed at the free end of the rebar and the sur-
face of concrete respectively, to monitor the relative displacement
of the rebar during loading. The reinforcing bar was pulled out
slowly, at a pulling rate of 1.2 mm/min. The pull-out force and dis-
placement data were recorded during the tests.
For undertaking pull-out tests on high temperature exposed
specimens, the specimens were first heated in an electrical furnace,
at an incremental heating rate of 5 °C per minute. Two thermocou-
ples were mounted on the surface and center of the specimen
respectively, to evaluate the difference between surface and center
temperatures during heat exposure. Once the predetermined
Diameter (mm) Yield strength (MPa) Ultimate strength (MPa)
10 312.00 502.36
10 420.64 606.31
12 470.40 620.15
14 437.37 600.04
18 365.15 520.00
25 445.78 596.79
 H.Y. Zhang et al. / Construction and Building Materials 163 (2018) 277–285 279

Table 2
Batch mix proportions of geopolymer concrete and OPC concrete (unit: kg/m3).

Concrete Type Alkaline activator MK FA OPC Water Fine aggregate Coarse aggregate Super-plasticizer
GC1 417 219 219 – – 616 1434 –
GC2 343 226 226 – – 634 1477 –
CC1 – – – 468 145 702 1076 1
CC2 – – – 470 143 636 1181 5

concrete and embed rebars would reduce to very low level at

Displacement meter 700 °C. However, the bond test results presented in the following
sections are better than expected. The difference in bond strength
of geopolymer concrete-rebars and geopolymer mortar-concrete
Bond length

20 substrate is induced by the difference in test method, specimen

size and material properties. A total of 90 geopolymer concrete
pull-out specimens and 30 OPC concrete pull-out specimens were
150

tested to evaluate bond characteristics of geopolymer concrete and

OPC concrete with reinforcing bar respectively. The details of these
specimens are listed in Table 3.
Plastic tube
To investigate the relationship between bond, compressive and
splitting tensile strengths of concrete, compressive and splitting
Reinforcing bar tensile strengths of geopolymer concrete and OPC concrete speci-
130

mens were also tested at ambient and after exposure to target tem-
peratures. The sizes of test specimens for compressive and splitting
F tensile strengths were 100 mm  100 mm  100 mm cubes.

Fig. 1. Typical pull-out test specimen for bond tests.

3. Results and discussion

Data from pull-out tests on geopolymer concrete specimens

60 with different types of concrete and reinforcing bars, at ambient
Compressive Strength (MPa)

and after exposure to high temperatures, is used to quantify the

50 effect of temperature on bond behavior of geopolymer concrete.
The data from tests on OPC concrete specimens were used as the
40 comparing benchmark of bond characteristics of geopolymer
concrete.
30
Geopolymer concrete
20 OPC concrete [24] 3.1. Compressive and splitting tensile strengths

10 Table 4 lists the comparative compressive and splitting tensile

0 at ambient and after exposure to elevated temperatures. It can
0 5 10
Curing age (day)
Fig. 2. Compressive strength development of geopolymer concrete and OPC
concrete with curing age.
target temperature (test temperature) was attained, temperature
of the furnace was kept constant for 120 min. From the measured
temperature data, thermal stability was achieved throughout the
specimens (temperature difference (gradient) between surface
and center was less than 10%), after 120-min exposure to the con-
stant target temperature. Then the heating in the furnace was
turned-off and the specimens were allowed to cool naturally. After
cooling down to the ambient temperature, the specimens were
taken out of the furnace and then pull-out tests were conducted,
using the same procedures as those at ambient tests.
The target temperatures for undertaking pull-out tests include
ambient temperature (about 25 °C), 100, 300, 500 and 700 °C. At
each temperature, three specimens with same concrete type and
same rebar group were tested. The pre-determination of the high-
est test temperature of 700 °C is based on authors’ previous study
[25]. It was reported that the bond strength of geopolymer mortar
with concrete substrate was almost lost at 700 °C [25]. Therefore, it
had been thought that the bond strength between geopolymer
strengths of geopolymer concrete (GC1) and OPC concrete (CC1)
15 20 25 30
be seen that the decrease in compressive and splitting tensile
strengths of GC1 and CC1 concretes is insignificant till the expo-
sure temperature is up to 300 °C. Especially, a slight increase in
strength of GC1 is seen at 300 °C. This slight strength increase in
geopolymer concrete is probably attributed to sintering reaction
of un-reacted fly ash particles in geopolymers [13,26]. When the
exposure temperature exceeds 300 °C, significant strength
decrease occurs on geopolymer and OPC concretes. Geopolymer
concrete GC1 exhibits similar level of compressive strength as that
of OPC concrete CC1 at ambient and after exposure to elevated
temperatures, but relatively higher splitting tensile strength than
that of CC1 concrete. Comparing the retention ratio of compressive
and splitting tensile strength after exposure to elevated tempera-
tures, a higher degradation rate occurred in the splitting tensile
strength of GC1 and CC1 concretes than that of compressive
strength. Geopolymer concrete GC2 and OPC concrete CC2 were
used to prepare pull-out specimens with slightly higher strength
concrete, to study the influence of compressive strength. The com-
pressive and splitting tensile strengths of GC2 and CC2 concretes at
ambient temperature are also presented in Table 4. Geopolymer
concrete GC2 exhibits similar compressive strength with OPC con-
crete CC2 at ambient temperature, but a slightly lower splitting
tensile strength than that of CC2. The strength variation of
geopolymer concrete GC2 with exposure temperature is not
280 H.Y. Zhang et al. / Construction and Building Materials 163 (2018) 277–285

Table 3
Specimen configurations and test parameters.

Group No. Concrete type Surface of rebar Rebar diameter, d (mm) Concrete cover, c (mm) c/d Test temp. (°C)
GC1-P-10 GC1 Plain 10 70 7.00 25, 100, 300, 500, 700
GC1-R-10 GC1 Ribbed 10 70 7.00
GC1-R-12 GC1 Ribbed 12 69 5.75
GC1-R-18 GC1 Ribbed 18 66 3.67
GC1-R-25 GC1 Ribbed 25 62.5 2.50
CC1-R-12 CC1 Ribbed 12 69 5.75
GC2-R-14 GC2 Ribbed 14 68 4.86
CC2-R-14 CC2 Ribbed 14 68 4.86

Table 4
Compressive and splitting tensile strengths of geopolymer concrete and OPC concrete.

Type of concrete Test temperature Comp. strength Comp. strength retention ratio Tensile Strength Tensile strength retention ratio
T (°C) fcu (MPa) (%) ft (MPa) (%)
GC-1 25 48.13 100 4.85 100
100 48.12 100 4.29 88
300 50.92 106 4.42 91
500 39.90 83 2.69 55
700 26.98 56 1.76 36
CC-1 25 49.78 100 3.09 100
100 46.69 94 2.53 82
300 49.65 100 2.98 96
500 39.9 80 1.79 58
700 27.84 56 0.96 31
GC-2 25 63.92 – 4.24 –
CC-2 25 66.66 – 5.24 –

presented herein, since it is similar with that reported in Ref. [20]. Table 5
Similar to the trend in strength development of high strength OPC Failure modes in specimens at ambient and after exposure to elevated temperatures.
concrete, the temperature-induced strength degradation of
Group No. Test temperature (°C)
geopolymer concrete with higher strength level is more significant
25 100 300 500 700
than that with lower strength level.
GC1-P-10 B B B B B
GC1-R-10 B B B B B
3.2. Failure modes in pull-out tests GC1-R-12 B B B B B
GC1-R-18 S S S-B S-B B
Four types of failure modes were seen during pull-out tests on GC1-R-25 S S S S-B S-B
CC1-R-12 B B B B Spalling
geopolymer concrete specimens: typical splitting failure, bond fail- GC2-R-14 F F S-B S-B S-B
ure (pull-out failure), a combination of splitting and bond failure CC2-R-14 B F, B F, S-B S-B S-B, B
(splitting-bond failure), and rebar fracture. Typical splitting failure
Note: S, B, S-B and F denote splitting, bond, splitting-bond and rebar fracture failure
is characterized by the formation of radial cracks from the rebar respectively.
surface, and quick progression of these cracks in longitudinal direc-
tion, leading to sudden breakage of geopolymer concrete without
much slippage of the rebar. Bond failure is characterized by the in exposure temperatures. However, for specimens in Group
interfacial shear failure between geopolymer concrete and rein- GC2-R-14, the failure modes changed from rebar fracture at ambi-
forcing bar, with significant slippage of the rebar, but without ent temperature to splitting-bond failure when exposure tempera-
any visible splitting cracks. If a specimen failed with significant ture exceeded 300 °C.
slippage of rebar and visible splitting cracks, it was termed as The failure modes in pull-out tests is governed by the minimum
splitting-bond failure. Some specimens, made from high strength of the splitting capacity of geopolymer concrete, the bond capacity
geopolymer concrete (GC2) or OPC concrete (CC2), failed through between geopolymer concrete and rebar, and the tensile capacity
slight slippage and fracture of rebar, but without visible splitting of rebar. The splitting capacity is dependent on the thickness of
cracks. This type of failure was termed as rebar fracture failure. concrete cover and splitting tensile strength of concrete [27], while
The failure mode in each specimen type is listed in Table 5. It the bond capacity is related to the geometric characteristic of the
can be seen from the table that at ambient temperature, bond fail- rebar and compressive strength of concrete. With an increase in
ure occurs on the specimens with smaller rebar diameter and exposure temperature, the compressive and splitting tensile
thicker concrete cover. However, specimens with larger rebar strengths of geopolymer concrete decrease, as discussed above
diameter and thinner concrete cover are more susceptible to split- (see Table 4). In addition, the mechanical properties of rebar also
ting failure. degrade after exposure to elevated temperatures, but the strength
With an increase in the exposure temperature, a change in fail- loss of rebar is lower as compared to that of concrete [28]. There-
ure modes was seen in Group GC1-R-18, GC1-R-25 and GC2-R-14 fore, specimens in Group GC2-R-14 failed through splitting-bond
specimens. Fig. 3 compares relative failure modes in GC1-R-18 failure, when exposed to high temperatures, but not through rebar
specimens after exposure to different temperatures. It can be seen fracture as that at ambient temperature. For specimens in Groups
that the failure modes of GC1-R-18 changed from splitting failure GC1-R-18 and GC1-R-25 after exposure to elevated temperatures,
to splitting-bond failure, and even bond failure, with an increase a significant difference in deformation levels between geopolymer
 H.Y. Zhang et al. / Construction and Building Materials 163 (2018) 277–285
(a) 25oC (b) 100oC (c) 300oC
Fig. 3. Failure patterns in Group GC1-R-18 specimens after exposure to elevated temperatures.
(a) 25oC
Fig. 4. Deformation state at rebar and geopolymer concrete interface after bond tests.
concrete and steel rebar were observed, as shown in Fig. 4. This is
due to incompatibility in deformations between geopolymer con-
crete and rebar during thermal expansion under heating state
and thermal shrinkage under cooling state. The deformation gap
between concrete and rebar, due to differential thermal expansion
and shrinkage, is also reported previously by other authors in pull-
out tests on ordinary concrete [29]. This difference in deformations
reduces the friction and mechanical bearing between geopolymer
concrete and rebar, leading to higher degradation in bond capacity.
The degradation in bond capacity possibly overwhelms that of
splitting capacity. Therefore, the failure modes in specimens from
Groups GC1-R-18 and GC1-R-25 changed from splitting failure at
ambient temperature to splitting-bond or bond failure after high
temperature exposure.
3.3. Bond behavior of specimens with different rebar surfaces
The interfacial bond stress developed between geopolymer con-
crete and rebar is assumed to be of uniform distribution along the
bond length. The average bond stress can be determined as:
F
s¼ ð1Þ
pdl
where s is the bond stress (MPa), F is the applied load (N), d is the
nominal rebar diameter (mm) and l is bond length (mm). The bond
strength (su) is the calculated average bond stress from the maxi-
mum applied loading during the pull-out tests. The free end slip
of the rebar is determined as the difference between the displace-
ment measured by the displacement meter on the free end of the
rebar and that on the surface of concrete.
Fig. 5 presents the relative bond strength of geopolymer con-
crete specimens embedded with plain rebar (GC1-P-10) and ribbed
rebar (GC1-R-10), as a function of exposure temperature. It can be
seen that a slight increase or decrease in bond strength occurred on
GC1-R-10 and GC1-P-10 specimens respectively, when the expo-
281
(d) 500oC (e) 700oC
(b) after exposure to 700oC
25
Bond strength (MPa)
20
15 GC1-R-10
10
GC1-P-10
5
0
0 100 200 300 400 500 600 700
o
Temperature ( C)
Fig. 5. Bond strength of specimens with different rebar surfaces.
sure temperature was below 300 °C. However, when the tempera-
ture exceeded 300 °C, a significant decrease in bond strength
occurred in GC1-R-10 and GC1-P-10 specimens. This trend is in
line with that of compressive and splitting tensile strengths of
geopolymer concrete. The significant strength degradation of
geopolymer concrete between 300 and 500 °C is mainly due to
temperature-induced dehydration and dehydroxylation reaction
of geopolymers, together with the thermal incompatibility
between geopolymers and aggregates [25].
Comparing the bond strength of GC1-R-10 and GC1-P-10 spec-
imens, it can be seen that the bond strength of geopolymer con-
crete with ribbed rebar (GC1-R-10) is more than twice of that
with plain rebar (GC1-P-10) at ambient temperature. When
exposed to temperatures above 300 °C, GC1-R-10 and GC1-P-10
specimens exhibit a significant degradation in bond strength, with
GC1-P-10 specimens having a higher degradation (rate). The bond
strength retention ratio, designated as the ratio of the residual
282 H.Y. Zhang et al. / Construction and Building Materials 163 (2018) 277–285
bond strength after exposure to high temperatures to the bond
strength tested at ambient temperature, is only 22% and 8% for
GC1-P-10 specimens at 500 °C and 700 °C respectively, but that
of corresponding GC1-R-10 specimens is 44% and 35% respectively.
The lower bond strength retention for specimens embedded
with plain rebars can be explained by the bond mechanism
between geopolymer concrete and plain rebar. The resistance of
a plain rebar against pull out is provided by chemical adhesion
and friction between rebar and concrete. Adhesion is lost at early
stages of loading once slip gets initiated. Therefore, the bond
strength between geopolymer concrete and plain rebar is mainly
determined by the friction. As the exposure temperature is raised,
severe deterioration occurs on concrete and significant deforma-
tion incompatibility between concrete and rebar gets developed,
which makes the friction weakened. Thus, the resistance of plain
rebars to pull out decreases significantly with an increase in tem-
perature. However, for the specimens embedded with ribbed
rebars, apart from the chemical adhesion and friction, the mechan-
ical bearing between rebar and concrete plays a prominent role in
the bond strength. Therefore, GC1-R-10 specimens experienced a
lower degradation (rate) in bond strength.
Fig. 6(a) and (b) present the typical bond stress-slip (s-s) curves
of specimens in Groups GC1-R-10 and GC1-P-10 respectively, at
different temperatures. GC1-R-10 and GC1-P-10 specimens all
failed through pull-out of rebars. As shown in Fig. 6, bond behavior
of these specimens can be characterized under three stages, which
are the ascending stage with a steep increase in bond stress but lit-
tle slippage, the descending stage with a quick decrease in bond
stress but a quick increase in slippage, and the steady stage with
Fig. 6. s-s curves of specimens in Groups GC1-R-10 and GC1-P-10.
no significant variation in bond stress but significant increase in
slippage.
From Fig. 6(a), it can be seen that the s-s curves for specimens
GC1-R-10 at 100 and 300 °C are close to that at ambient tempera-
ture (25 °C), but the s-s curves at 500 and 700 °C are much differ-
ent. At 500 and 700 °C, the slips of rebars initiated at a lower
loading, and the peak bond stresses are much lower than that at
25 °C, but the slips corresponding to the peak bond stresses are
greater.
The effect of temperature on s-s curves of GC1-P-10 specimens
is slightly different with that of GC1-R-10. With the increasing
exposure temperature, the peak bond stress and the residual stress
at the steady state stage decrease gradually. Compared to Fig. 6(a),
the peaks of the s-s curves of GC1-P-10 specimens in Fig. 6 (b) are
sharper, and the slips corresponding to the peak bond stresses are
much lower than those of GC1-R-10 specimens. This is due to the
fact that once the slip is initiated, the chemical adhesion between
geopolymer concrete and rebar is quickly lost, leading to the bond
stress of GC1-P-10 specimens reaching a peak value at a lower slip,
and then s-s curves abruptly enter the descending stage. However,
for specimens in Group GC1-R-10, the mechanical bearing is dom-
inant for bond capacity, while the loss of chemical adhesion has no
significant effect on bond capacity, thus the peaks of s-s curves are
flatter.
3.4. Bond behavior of specimens with different rebar diameters
Bond strength between concrete and rebar is generally believed
to decrease with an increase of rebar diameter, when the specimen
sizes are kept constant [28]. However, there are few studies which
reported that the bond strength increases with rebar diameter
increasing [30]. The bond strength of geopolymer concrete speci-
mens with different rebar diameters is presented in Fig. 7, as a
function of exposure temperature. It can be seen that the bond
strength of specimens with rebar diameters of 10 and 12 mm is
higher than that of 18 and 25 mm, both at ambient and after expo-
sure to elevated temperatures. The specimens with rebar diameter
of 10 and 12 mm failed through pull-out, thus the bond strength of
these specimens is related to the geometric characteristics of
rebars, such as rib spacing, rib height and phase angle. As listed
in Table 1, R-12 rebars have the same rib height/spacing ratio as
R-10 rebars, but higher rib phase angle, which leads to higher bond
strength with geopolymer concrete than that of R-10 rebars. Spec-
imens in Groups GC1-R-18 and GC1-R-25 experienced splitting or
splitting-bond failure prior to the occurrence of bond failure, due
to lower ratio of concrete cover thickness to rebar diameter (c/d).
Therefore, the real bond strength could not be obtained, and the
24
GC1-R-10
GC1-R-12
Bond strength (MPa)
20 GC1-R-18
GC1-R-25
16 d=10
d=18
12
d=25 d=12
8
0
0 100 200 300 400 500 600 700
o
Temperature ( C)
Fig. 7. Bond strength of specimens with different rebar diameters.
 H.Y. Zhang et al. / Construction and Building Materials 163 (2018) 277–285
calculated bond strength of GC1-R-18 and GC1-R-25 specimens is
lower than that of GC1-R-10 and GC1-R-12 specimens.
As illustrated in Fig. 7, the variations of bond strength with
increasing temperature are similar for specimens with different
rebar diameters. The bond strength in these specimens has no sig-
nificant reduction when the exposure temperature is below 300 °C,
but exhibit significant degradation above 300 °C.
Fig. 8 plots the typical s-s curves of specimens in Groups GC1-R-
18 and GC1-R-25, after exposure to elevated temperatures. The
trends of these curves are much different with those in Fig. 6,
due to different failure modes occurred in these specimens. The
specimens of GC1-R-18 at ambient temperature and after exposure
to 100 °C, and GC1-R-25 at ambient temperature and after expo-
sure to 100 °C and 300 °C, failed through splitting failure. Splitting
failure is characterized by brittle and abrupt failure, and thus the
s-s curves of these specimens have no descending segment, as well
as steady segment. In specimens that experienced splitting-bond
failure, such as GC1-R-18 specimens at 300 °C and GC1-R-25 spec-
imens at 500 and 700 °C, although the three stages (ascending,
descending and steady stage) can be observed on their s-s curves,
the descending stage in them developed in a very short time and
had higher decreasing rate than those specimens that failed
through pull-out mechanism. This is due to rapid growth of split-
ting cracks from interior concrete to the surface, which leads to
sudden decrease of bond stress and abrupt increase in slippage
after the attainment of peak bond stress. In comparison to speci-
mens failing through splitting, specimens experiencing splitting-
bond failure could still sustain a certain portion of loading due to
friction action, which leads to attaining a steady state on their
s-s curves.
Fig. 8. s-s curves of specimens in Groups GC1-R-18 and GC1-R-25.
283
3.5. Comparison of bond behavior of geopolymer concrete and OPC
concrete
To compare the bond behavior of geopolymer concrete and OPC
concrete, results from pull-out tests on two batches of geopolymer
concrete (GC1 and GC2) and two batches of OPC concrete (CC1 and
CC2) are utilized. As listed in Table 4, GC1 and CC1 concretes exhib-
ited similar level of compressive and splitting tensile strengths, but
GC2 and CC2 concretes had only similar compressive strength but
different splitting tensile strength. In addition, the compressive
strength of GC2 and CC2 concretes was higher than that of GC1
and CC1 concretes.
Fig. 9 illustrates the bond strength of four groups of specimens,
including GC1-R-12, CC1-R-12, GC2-R-14 and CC2-R-14, as a func-
tion of exposure temperature. It can be seen that although GC1 and
CC1 concretes exhibited similar compressive and splitting tensile
strengths, GC1 concrete exhibited higher bond strength than that
of CC1 concrete, both at ambient and after exposure to elevated
temperatures. It should be noted that the pull-out tests were not
conducted on CC1-R-12 specimens after exposure to 700 °C, due
to the spalling of OPC concrete in CC1-R-12 specimens during heat-
ing. No temperature-induced spalling was seen in any of geopoly-
mer concrete specimens. Specimens GC2-R-14 and CC2-R-14
exhibited close bond strengths. These strengths were higher than
those of GC1-R-12 and CC1-R-12 specimens, due to the fact that
GC2 and CC2 concretes had higher compressive strength than
GC1 and CC1 concretes.
Fig. 10(a) and (b) compare retention ratio of compressive, split-
ting tensile and bond strengths with ribbed rebar (R-12) of GC1
and CC1 concretes after exposure to elevated temperatures. It
can be seen that the trends of compressive, splitting tensile and
bond strengths of GC1 and CC1 concretes, with temperature, are
similar. When the exposure temperature is below 300 °C, there is
only slight degradation in compressive, splitting tensile and bond
strengths of GC1 and CC1 concretes; but when the temperature
exceeds 300 °C, significant degradation in corresponding strength.
Comparing the degradation rates of compressive, splitting tensile
and bond strengths, it can be seen that the degradation rates of
bond and splitting tensile strengths of GC1 and CC1 concretes are
higher than that of compressive strength. The degradation rates
of bond and splitting tensile strengths of GC1 concrete, after expo-
sure to elevated temperatures, are close, except that at 100 °C. Fol-
lowing 100 °C exposure, GC1 concrete has almost no reduction in
bond strength, but 12% reduction in splitting tensile strength.
Fig. 11(a) and (b) present the typical s-s curves of specimens in
Groups GC2-R-14 and CC2-R-14, at different temperatures. The
25
GC1-R-12
GC2 CC2 CC1-R-12
Bond strength (MPa)
20 GC2-R-14
CC2-R-14
GC1
15
CC1
10
0
0 100 200 300 400 500 600 700 800
o
Temperature ( C)
Fig. 9. Comparison of bond strength of geopolymer concrete with that of OPC
concrete.
284 H.Y. Zhang et al. / Construction and Building Materials 163 (2018) 277–285

Strength retention ratio (%)

100

80

60

40 Compressive strength
Splitting tensile strength
20 Bond strength

0
0 100 200 300 400 500 600 700 800
o
Temperature ( C)
(a) GC1
Strength retention ratio (%)

100

80

60

40
Compressive strength
Splitting tensile strength
20
Bond strength
0
0 100 200 300 400 500 600 700 800
o
Temperature ( C)
(b) CC1
Fig. 10. Strength retention ratio of GC1 and CC1 concretes after exposure to
Fig. 11. s-s curves of GC2-R-14 and CC2-R-14 test specimens.
elevated temperatures.

where T is the highest exposure temperature, sTu is the residual bond

rebars in GC2-R-14 specimens at ambient temperature and after
exposure to 100 °C fractured during pull-out tests, with little slip-
page. Therefore, s-s curves of GC2-R-14 specimens at 25 and 100 °C
abruptly dropped after the peak stress was reached. Similar trend
was seen on the s-s curve of CC2-R-14 at 100 °C. Based on the duc-
tility attained from yielding (Point A in Fig. 11(a)) to fracture of
rebars (Point B), an anchorage length of 5d can be regarded as
the critical anchorage length for this type of high strength geopoly-
mer concrete. Comparing Fig. 11(a) and (b), it can be seen that the
trends on s-s curves of GC2-R-14 and CC2-R-14 specimens at 300,
500 and 700 °C are similar, due to similar failure modes occurring
in these specimens (splitting-bond failure), but the slippage of
CC2-R-14 specimen corresponding to similar peak stress as that
strength between geopolymer concrete and rebar after exposure to
temperature T, sAu is the bond strength at ambient temperature. The
coefficients in the above expression are obtained through regression
analysis, with a correlation coefficient (R) of 0.966. The bond
strength between geopolymer concrete and ribbed rebar at ambient
temperature (sAu ) can be estimated by the proposed model in Ref.
[19].
Fig. 12 shows the comparison between experimental and pre-
dicted residual bond strength versus exposure temperature for
geopolymer concrete with ribbed rebars. It can be seen that the
predicted results fit well with the test data.
Bond strength retention ratio

of GC2-R-14 is larger. This infers that geopolymer concrete exhibits 1.2

similar or better bond behavior than OPC concrete with same
levels of compressive strength. 1.0
4. Prediction model of residual bond strength after high 0.8
temperature exposure Test data of d=10
0.6 Test data of d=12
Polynomial function is often adopted for expressing residual Test data of d=14
bond strength between rebar and OPC concrete exposed to ele- 0.4 Test data of d=18
vated temperatures [28,31]. To evaluate residual bond strength Test data of d=25
between geopolymer concrete and ribbed rebar, as a function of 0.2
Fitting curve
temperature, the following cubic expression on bond strength
retention ratio is developed, utilizing data generated in 0.0
experiments. 0 100 200 300 400 500 600 700
o
sTu Temperature ( C)
¼ 0:96 þ 1:39  103 T  7:06  106 T 2 þ 4:95  109 T 3 ð2Þ
sAu Fig. 12. Comparison of predicted and measured bond strength retention ratio.
 H.Y. Zhang et al. / Construction and Building Materials 163 (2018) 277–285
5. Conclusions
To evaluate the effect of temperature on bond characteristics of
geopolymer concrete, pull-out tests were conducted on large num-
ber of specimens with different concrete and rebar types, at ambi-
ent and after exposure to elevated temperatures. Based on the test
results, the following conclusions can be drawn:
(1) Geopolymer concrete exhibits significant temperature
induced degradation in bond strength, when exposed to
temperatures above 300 °C.
(2) Bond strength of geopolymer concrete with ribbed rebar is
more than twice of that with plain rebar of same diameter
at ambient temperature. Also, degradation in bond strength
of geopolymer concrete with ribbed rebar occurs at a lower
rate with temperature.
(3) Bond strength of geopolymer concrete decreases at the same
rate as that of splitting tensile strength with temperature,
but this degradation is at a higher pace than that of com-
pressive strength.
(4) Geopolymer concrete exhibits similar or better bond proper-
ties than ordinary Portland cement concrete with similar
levels of compressive strength.
Acknowledgements
The research presented in this paper is supported by National
Natural Science Foundation of China (Grant No. 51478195), the
Fundamental Research Funds for the Central Universities
(2015ZZ071), and State Key laboratory of Subtropical Architecture
Science, South China University of Technology (Grant No.
2016KB15) and Michigan State University. Any opinions, findings,
and conclusions or recommendations expressed in this paper are
those of the authors and do not necessarily reflect the views of
the sponsors.
References
[1] J.G.J. Olivier, G. Janssens-Maenhhout, M. Muntean, J.A.H.W. Peters, Trends in
Global CO2 Emissions; 2015 Report, PBL Netherlands Environmental
Assessment Agency, The Hague, Netherlands, 2015.
[2] F.U.A. Shaikh, Review of mechanical properties of short fibre reinforced
geopolymer composites, Constr. Build. Mater. 43 (2013) 37–49.
[3] J. Davidovits, Geopolymers geopolymeric materials, J. Therm. Anal. 35 (2)
(1989) 429–441.
[4] P. Duxson, A. Fernandez-Jimenez, J.L. Provis, G.C. Lukey, A. Palomo, J.S.J. van
Deventer, Geopolymer technology: the current state of the art, J. Mater. Sci. 42
(9) (2007) 2917–2933.
[5] S.A. Bernal, E.D. Rodriguez, R.M. de Gutierrez, M. Gordillo, J.L. Provis,
Mechanical and thermal characterisation of geopolymers based on silicate-
activated metakaolin/slag blends, J. Mater. Sci. 46 (16) (2011) 5477–5486.
285
[6] B.C. McLellan, R.P. Williams, J. Lay, A. van Riessen, G.D. Corder, Costs and
carbon emissions for geopolymer pastes in comparison to ordinary Portland
cement, J Clean Prod. 19 (2011) 1080–1090.
[7] A.M. Fernandez-Jimenez, A. Palomo, C. Lopez-Hombrados, Engineering
properties of alkali-activated fly ash concrete, ACI Mater. J. 103 (2) (2006)
106–112.
[8] R. Zhao, J.G. Sanjayan, Geopolymer and Portland cement concretes in
simulated fire, Mag. Concr. Res. 63 (3) (2011) 163–173.
[9] Prabir Kumar Sarker, Sean Kelly, Zhitong Yao, Effect of fire exposure on
cracking, spalling and residual strength of fly ash geopolymer concrete, Mater.
Des. 63 (2014) 584–592.
[10] M. Olivia, H. Nikraz, Properties of fly ash geopolymer concrete designed by
Taguchi method, Mater. Des. 36 (2012) 191–198.
[11] D. Hardjito, S.E. Wallah, D.M.J. Sumajouw, B.V. Rangan, On the development of
fly ash-based geopolymer concrete, ACI Mater. J. 101 (6) (2004) 467–472.
[12] D. Adak, M. Sarkar, S. Mandal, Structural performance of nano-silica modified
fly-ash based geopolymer concrete, Constr. Build. Mater. 135 (2017) 430–439.
[13] H.Y. Zhang, V. Kodur, B. Wu, L. Cao, S.L. Qi, Comparative thermal and
mechanical performance of geopolymers derived from metakaolin and fly ash,
J. Mater. Civil Eng. 28 (2) (2016) 1–10.
[14] M. Sofi, J.S.J. van Deventer, P.A. Mendis, G.C. Lukey, Bond performance of
reinforcing bars in inorganic polymer concrete (IPC), J. Mater. Sci. 42 (9) (2007)
3107–3116.
[15] P.K. Sarker, Bond strength of reinforcing steel embedded in fly ash-based
geopolymer concrete, Mater. Struct. 44 (5) (2011) 1021–1030.
[16] A. Castel, S.J. Foster, Bond strength between blended slag and Class F fly ash
geopolymer concrete with steel reinforcement, Cem. Concr. Res. 72 (2015) 48–
53.
[17] R.D. Moser, P.G. Allison, B.A. Williams, C.A. Weiss, A.D. Diaz, E.R. Gore, et al.,
Improvement in the geopolymer-to-steel bond using a reactive vitreous
enamel coating, Constr. Build. Mater. 49 (2013) 62–69.
[18] S. Songpiriyakij, T. Pulngern, P. Pungpremtrakul, C. Jaturapitakkul, Anchorage
of steel bars in concrete by geopolymer paste, Mater. Des. 32 (5) (2011) 3021–
3028.
[19] H.Y. Zhang, J. Yan, B. Wu, Study on bond behavior between geopolymer
concrete and steel bars, China Civil Eng. J. 49 (7) (2016) 107–115 (In Chinese).
[20] H.Y. Zhang, V. Kodur, S.L. Qi, L. Cao, B. Wu, Development of metakaolin–fly ash
based geopolymers for fire resistance applications, Constr. Build. Mater. 55
(2014) 38–45.
[21] National Standard of PRC, Standard Method for Testing of Concrete Structures
(GB 50152-92), China Building Industry Press, Beijing, 1992.
[22] RILEM/CEB/FIP, RC6: Bond Test Reinforcement Steel-Pull Out Test, Georgi
Publishing Company, 1983.
[23] D.L.Y. Kong, J.G. Sanjayan, Damage behavior of geopolymer composites
exposed to elevated temperatures, Cem. Concr. Compos. 30 (10) (2008) 986–
991.
[24] Z.R. Jiang, Handbook of Construction Calculation, China Architecture &
Building Press, 2007.
[25] H.Y. Zhang, V. Kodur, B. Wu, L. Cao, F. Wang, Thermal behavior and mechanical
properties of geopolymer mortar after exposure to elevated temperatures,
Constr. Build. Mater. 109 (2016) 17–24.
[26] D.L.Y. Kong, J.G. Sanjayan, K. Sagoe-Crentsil, Comparative performance of
geopolymers made with metakaolin and fly ash after exposure to elevated
temperatures, Cem. Concr. Res. 37 (2007) 1583–1589.
[27] K. Hertz, The anchorage capacity of reinforcing bars at normal and high
temperatures, Mag. Concr. Res. 34 (121) (1982) 213–220.
[28] A. Ergun, G. Kurklu, M.S. Baspinar, The effects of material properties on bond
strength between reinforcing bar and concrete exposed to high temperature,
Constr. Build. Mater. 112 (2016) 691–698.
[29] T. Pothisiri, P. Panedpojaman, Modeling of bonding between steel rebar and
concrete at elevated temperatures, Constr. Build. Mater. 27 (1) (2012) 130–
140.
[30] B. Singh, G. Ishwarya, M. Gupta, S.K. Bhattacharyya, Geopolymer concrete: a
review of some recent developments, Constr. Build. Mater. 85 (2015) 78–90.
[31] Z. Huang, Modelling the bond between concrete and reinforcing steel in a fire,
Eng. Struct. 32 (11) (2010) 3660–3669.