Construction and Building Materials 186 (2018) 790–800

Contents lists available at ScienceDirect

Construction and Building Materials

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

Evaluation of the bond strength of a novel concrete for rapid patch

repair of pavements
Tianxiong Guo ⇑, Yuchen Xie, Xingzhong Weng
Department of Airport Building Engineering, Air Force Engineering University, Xi’an 710038, China

h i g h l i g h t s

Combined effects of surface roughness, bonding agents and curing age were investigated through orthogonal experiments.

Bond strength was evaluated using a self-designed test device (Patent Number: ZL 2017 2 1680364.4).

Freeze-thaw cycle test was conducted to validate the results of orthogonal experiments.

RRM concrete appears to be a promising material for rapid repair of concrete pavements.

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

Article history: Patch repair is an important category in the repair or restoration of pavements. The effectiveness of the
Received 8 June 2018 repair largely depends on the bond strength between the repair material and the concrete substrate. In
Received in revised form 29 July 2018 the present work, surface roughness, bonding agents and curing age of the repair materials have been
Accepted 3 August 2018
taken into consideration to evaluate the applicability and serviceability of a novel rapid repair material
Available online 9 August 2018
for concrete pavement. A three-factor and four-level orthogonal experiment was designed for systematic
investigation. The bond strength of the combined specimens was measured using splitting tensile test
Keywords:
and direct shear test. It is found that the splitting tensile strength varied from 1.21 MPa to 2.43 MPa
Bond strength
Patch repair
and the direct shear strength varied from 2.84 MPa to 5.92 MPa and the bond strength has a good corre-
Splitting tensile test lation with surface roughness and curing age. Different failure modes and the shear-slip characteristics
Direct shear test were analyzed to understand the actual bond performance of combined specimens. It can be concluded
Bonding agent that the slip distance of high surface roughness concrete was almost three times that of low surface
Surface roughness roughness. In addition, the freeze-thaw cycle test of the combined specimen was conducted to validate
Freeze-thaw cycle test the compatibility of the designed repair material with the concrete substrate. It is found that the he split-
ting tensile strength and direct shear strength of high surface roughness specimens reduced to 74.5% and
66.3% of its original bond strength after 100 times of freeze-thaw cycles, respectively. The rapid repair
material studied in this study can act as a promising alternative for rapid patch repair of concrete
pavements.
Ó 2018 Elsevier Ltd. All rights reserved.

1. Introduction Road Information Program, 32 percent of major roads in the US

are in either poor or mediocre condition and require immediate
Cement concrete has been widely used in the construction of repair or maintenance [2]. Consequently, rapid repair of the deteri-
pavements, municipal roads and airport runways due to the char- orated pavements is essential to eliminate congestions and envi-
acteristics of high strength, longevity and good stability [1]. How- ronmental impacts induced by poor road conditions, and most
ever, the transportation infrastructures suffer from premature importantly, for safety concerns.
deterioration or damage due to the combined effects of environ- In recent years, researchers have developed many rapid repair
mental and mechanical loadings. The rehabilitations and restora- materials (RRM) for the rehabilitation of concrete pavements,
tions cost of the infrastructures was close to and even exceeded which can be divided into three categories: cement-based materi-
new construction cost [1]. Similarly, according to a survey by The als, polymer modified cement-based materials, and polymer or
resin materials [3–6]. However, variation in results obtained by dif-
ferent researchers could be attributed to the difference in raw
⇑ Corresponding author.
materials, specimen geometry and test methods.
E-mail address: 1205793029@qq.com (T. Guo).

https://doi.org/10.1016/j.conbuildmat.2018.08.007
0950-0618/Ó 2018 Elsevier Ltd. All rights reserved.
 T. Guo et al. / Construction and Building Materials 186 (2018) 790–800 791

Some works [7,8] indicated that Ordinary Portland Cement It is well recognized that the effectiveness of a repair is strongly
(OPC) often led to interface failure stemming from large autoge- influenced by the quality and behavior of the interface created
nous shrinkage stress. Subsequently, numerous studies have been between the repair material and the old concrete [34]. Regardless
conducted focusing on the preparation of rapid repair materials of the rapid repair material, it is well documented that the bond
using various sort of admixtures, for example, fly ash [9,10], silica strength between the substrate and the repair material plays cru-
fume [10,11], recycled aggregates [12], industrial or agricultural cial role for carrying capacity and long-term performance of the
by-products [13]. Due to low dry shrinkage, short setting time, repaired pavement [3,35–39]. Many factors including interface
good compatibility and high bond strength with the concrete sub- stiffness [40], surface roughness [36,37,41,42], adhesive agents
strate, magnesium phosphate cement (MPC) are suitable for rapid [43–46], curing regime [47,48] and the age of the repair materials
repair of concrete structures used in pavements, airport runways, influence the bond strength. For example, some researchers claim
bridge decks and key municipal roads, etc [14–17]. However, due that both an increase in surface roughness and the application of
to the strong heat release in the reaction process, MPC is sensitive bonding agents are imperative to good bond performance
to the ambient temperature and reaction heat and rate in hot [41,42,44,49–51].
weather need to be properly investigated [16]. It is reported that the measured bond strength is greatly depen-
Polymer modified concrete as an important class of rapid repair dent on the test method [35]. The existing test methods to evaluate
material for concrete infrastructure, on the other hand, has many the bond strength between substrate and repair materials can be
outstanding advantages: low thickness, superior chemical resis- divided into three categories [3,35,41,52,53]. The first category is
tance, low cure shrinkage and permeability and adhesion charac- to evaluate the bond strength under tensile stress, for example,
teristics [18–20]. Ghasan Fahim Huseien et al. [21] provided an pull-off [54], the flexural test [55] and splitting [38,56]. The second
exhaustive and informative overview on the past development category of the test methods measures the bond strength under
and recent progress of geopolymer mortars (GPMs). They summa- shear stresses, for example, the direct shear test [39]. The third cat-
rized the properties of GPMs as repair materials from the aspects of egory is to determine the bond strength under the combination of
strength, addition, setting time, mix proportion, bond strength and compressive stress and shear stress, which is commonly referred to
microstructure and concluded that GPMs are effective repair mate- the slant shear test [8,57,58]. The difference between these test
rials for deteriorated concrete and structure protection as they pos- methods is that the stress state of the test specimens is different.
sess the characteristics of far lower carbon footprint, less cracking, In this study, a three-factor and four-level orthogonal experi-
more resistant to corrosive elements such as sea salt, excellent ment was conducted to evaluate the bond strength between the
frost resistance and particularly, bonding strength. In this type of concrete substrate and a novel rapid repair material for concrete
material, epoxy-based polymer concretes/mortars are the most pavements. The three factors adopted herein are surface rough-
commonly used [22]. Saccani et al. [23] investigated the behavior ness, bonding agents and curing age. The bond strength was mea-
and durability of epoxy-modified mortar and indicated that the sured by splitting tensile test and direct shear test, for comparison.
repair material performed well when exposed to thermal and envi- In addition, freeze-thaw cycle test was carried out to validate the
ronmental stresses and provided higher durability and bonding compatibility of the repair material and assess the influence of sur-
strength when combined with epoxy-modified primer. Rodrigo H. face roughness on bond strength.
Geraldo et al. [24] investigated alkali-activated mortars (AAM) as
repair material to reinforced concrete and found that AAM devel-
oped a fast and good strength (30 MPa after 24 h) and had good 2. Experimental procedure
adherence to the substrate. Jung et al. [25] studied the serviceabil-
ity of polymer concrete by measuring basic mechanical properties 2.1. Materials
using embedded FBG (fiber bragg grating) optic sensors and
reported that the thermal expansion coefficient of polymer con- 2.1.1. Cement
crete was much higher than that of cement concrete. Based on The chemical compositions of the cement and RRM used in the
analysis of the geopolymerization process, Yanping Li et al. [26] study are shown in Table 1.
discussed the mechanism of how raw material formulation affects
fresh and mechanical properties of geopolymer concrete. Besides, a 2.1.2. Rapid repair material (RRM)
design of experiment (DOE) approach was proposed to quantify the The rapid repair material is a kind of polymer modified cemen-
effects of silica concentration, hydroxide concentration, and slag titious material with the characteristics of rapid-setting, high early
replacement percentage on setting time, 7-day, and 28-day com- strength, small shrinkage, excellent workability and good durabil-
pressive strength. What’s more, it had reported that the applicabil- ity developed by Haimu Transportation Technology Co. Ltd,
ity of polymer concrete could be improved through the Chongqing, China [59,60]. It has been successfully applied in the
supplement of silicone rubber [27,28] and using some other poly- emergency repair of airport pavement and municipal roads bridge
mers, for example, latex [29], high calcium fly ash geopolymer [30] decks.
and ultra-fine slag based geopolymer [31].
There are few literatures available about the use of polymer or
2.1.3. Bonding agents
resin materials for rapid patch repair, which is classified as third
Three bonding agents were chosen in this study, (a) RRM paste
type of rapid repair material. However, these materials have been
(RRMP) (water: cement = 0.4), (b) polyvinyl formaldehyde
successfully applied in the case of full-depth repair. Priddy et al.
adhesive (commonly known as 108 glue) modified RRM paste
[32,33] investigated the applicability of commercially available,
(GM-RRMP) (adhesive: water: cement = 2:4:15) and (c) epoxy resin.
rigid polyurethane foams for the expedient airfield repairs. They
mentioned that the pavement repair technique can be completed
within 4–6 h and can support threshold and objective aircraft pass 2.2. Specimens preparation
levels. However, foam backfill materials used in these studies are
reported to be sensitive to ambient moisture, which has a great 2.2.1. Mix proportions
influence on the quality of the expedient pavement repair [32]. Ordinary concrete was used to prepare the concrete substrate.
Besides, this technique relies heavily on proprietary devices, which The proportions of concrete substrate and RRM concrete mixtures
may increase logistic burden. are given in Table 2.
792 T. Guo et al. / Construction and Building Materials 186 (2018) 790–800

Table 1
Chemical compositions of cement.

Chemical composition (%) CaO SiO2 Al2O3 Fe2O3 MgO SO3 TiO2 Ignition loss
Cement 62.9 21.6 5.5 3.5 2.3 2.44 – 0.62
RRM 58.69 21.45 6.28 3.85 3.01 1.96 0.45 0.11

Table 2 After the surface preparation of the concrete substrate, the sur-
Mix proportions of concrete substrate and RRM concrete. faces were rinsed with water and different bonding agents were
Materials Substrate (kg/m3) Repair concrete brushed on the prepared interface. Afterwards, the concrete sub-
(kg/m3) strate was placed into the 150 mm  150 mm  150 mm cube
Ordinary Portland Cement (OPC) 320 – mould and the RRM concrete was cast to form a combined specimen.
Rapid repair material – 400
River sand 598 620
Coarse aggregate 5–10 mm 378 195
2.2.3. Experimental method
10–20 mm 882 378 The basic mechanical properties of the RRM concrete such as
20–40 mm 860 – compressive strength and flexural strength were investigated in
Water 140 160 accordance with the test method described in test specifications
Super plasticizer – 2%
[65]. The cube specimen with a dimension of 150 mm  150 mm
 150 mm and the prismatic specimen of 100 mm  100 mm 
400 mm are used to test the compressive strength and flexural
strength of RRM concrete, respectively. In the orthogonal experi-
2.2.2. Specimen design ments, the table of L16(45) was adopted to assess the splitting ten-
Semi-cube specimens (150 mm  150 mm  75 mm) cast using sile strength and the shear strength. Splitting tensile strength and
ordinary Portland cement concrete were used as concrete sub- direct shear strength were calculated using Eqs. (2) and (3),
strate. The concrete substrate was cured under standard conditions respectively.
(temperature 20 ± 2 °C, relative humidity 95%) for 28 days.
After curing, the surface of concrete substrate was processed to 2F
f ts ¼ ð2Þ
a certain degree of roughness. Surface preparation process includes pA
sandblasting, dill holes, surface chiseling, steel brushing, high- where: fts represents splitting tensile strength of concrete (MPa), F
pressure water jet spraying and surface grooving [34,38]. Sand- denotes failure load (N) and A is the area of splitting surface.
blasting is demonstrated as the best interface preparation method
for achieving high bond strength [10,61] whereas high-pressure Fs
f ¼ ð3Þ
water jet is acknowledged as the cleanest method. However, there As
are few dedicated devices available for rapid repair projects, for
where: f means direct shear strength of concrete (MPa), Fs repre-
example, jetting machine. Therefore, the methods of surface chisel-
sents failure load (N), and As is the area of shear surface (mm2).
ing (using chisel and electronic breaker) and surface grooving
The factors and levels of the orthogonal experiments are listed
(using chainsaw) were adopted in this study as these approaches
in Table 3.
do not require professional operators. The processed interfaces of
The loading rates of compressive strength test, tensile strength
concrete substrate are shown in Fig. 1. In order to quantify the
and splitting tensile test were used to be 0.5–0.8 MPa/s, 0.02–0.05
roughness of the artificial interface, several methods have been
MPa/s and 0.08 MPa/s, respectively. The arithmetic mean value of
proposed [61–64]. In this work, the sand filling method was used
three tests was evaluated as the final strength. The direct shear test
to quantify the surface roughness by measuring the average sand
was carried out on a self-designed experimental device (Patent
filling depth using Eq. (1).
Number: ZL 2017 2 1680364.4). The test device is mainly com-
Vs posed of four parts: reaction force system, shearing system, loading
Hs ¼ ð1Þ system, and measurement and recording system. The test pro-
As
cesses are shown in Fig. 2.
where: Hs represents the average sand filling depth (mm); Vs Based on the orthogonal experiments results, the optimum bond-
denotes the volume of sand (mm3) and As is the area of the treated ing agent was selected to conduct the freeze-thaw cycle test. The
concrete substrate (mm2). process modes of the concrete substrate are mentioned in Fig. 1

Fig. 1. Processed interface of the concrete substrate using (a, b) chiselled and (c, d) grooved methods. The width and depth of the grooves in (c) 10 mm and 15 mm and (d) 15
mm and 20 mm. The measured average sand filling depth of (a)–(d) is 2.4 mm, 5.6 mm, 6.8 mm and 8.7 mm, respectively.
 T. Guo et al. / Construction and Building Materials 186 (2018) 790–800 793

Table 3 As can be seen from Fig. 4, the compressive strength and the flex-
The factors and levels of the orthogonal experiments. ural strength gradually increased with the increase of curing age.
Levels Factors However, the growth rate of the strength of RRM concrete was sig-
Process methods (A) Bonding agents (B) Curing age (C) nificantly faster than ordinary concrete. The compressive strength
and flexural strength of RRM concrete after curing for 7d were
1 Chisel-low roughness RRMP 4h
2 Chisel-high roughness GM-RRMP 1d
33.2% and 27.5% higher than that of ordinary concrete, respectively.
3 Groove-low roughness Epoxy resin 3d Besides, the compressive strength and flexural strength of RRM con-
4 Groove-high roughness None 7d crete after curing for 4 h reached to 50.4 MPa and 5.3 MPa, respec-
Note: letters after each factor correspond to the letters in the orthogonal experi-
tively, accounting for 80.8% and 88.9% of the corresponding values
mental design table L16 (45). of ordinary concrete after curing for 28d. The high early strength
of RRM concrete makes it possible to reopen the traffic quickly.

and Table 2. After curing for 7 days, the combined specimens were 3.2. Results of orthogonal experiments
placed in the freeze-thaw tank with the concrete substrate upwards.
Water was added until the water reached 1 cm above the interface 3.2.1. Bond strength
level. A 12-hour cycle was repeated to 25, 50, 75 and 100 times According to the orthogonal experimental arrangement, a total
and afterwards the splitting tensile strength and the direct shear of 16 groups of tests were carried out. Three specimens were pre-
strength were measured to assess the influence of freeze-thaw cycles pared for each group. The results of the orthogonal experiments
on the compatibility and durability of RRM concrete. The test process and range analysis are listed in Tables 4 and 5, respectively.
and the temperature control parameters are shown in Fig. 3. As can be seen from the range analysis, bonding agent has
greatest influence on both splitting tensile strength and direct
3. Results and discussion shear strength, followed by curing aging time and surface process
methods. Besides, the ratio of direct shear strength to splitting ten-
3.1. Basic mechanical properties sile strength varied from 2.15 to 3.26. However, the ratio of most
samples was within the range of 2.4–2.5. The influences of each
The mechanical properties of concrete substrate and RRM factor in the orthogonal experiments on the bond strength are
concrete at different curing ages are shown in Fig. 4. shown in Fig. 5.

(a) (b)

Repair Substrate

(c) (d)

Fig. 2. Images of equipment and corresponding set up used for (a) compression test, (b) flexural strength test (c) splitting tensile test and (d) direct shear test.
794 T. Guo et al. / Construction and Building Materials 186 (2018) 790–800

25

20

15

Tenoerature (degree)
10

0
0 2 4 6 8 10 12 14 16 18 20 22 24
-5
Time (h)
-10

-15

-20

-25

(a) freeze-thaw cycles (b) temperature control parameters

Fig. 3. Freeze-thaw cycle tests.

Fig. 4. (a) the compressive strength and (b) the flexural strength of RRM concrete and ordinary concrete.

Table 4
Results of the orthogonal experiments.

Serial number Factors Splitting tensile strength (MPa) Direct shear strength (MPa)
Process mode (A) Bonding agents (B) Curing age (C)
1 1 1 1 1.60 3.84
2 1 2 2 1.37 3.25
3 1 3 3 1.64 4.40
4 1 4 4 1.32 2.84
5 2 1 2 2.12 5.31
6 2 2 1 1.21 2.94
7 2 3 4 1.85 5.01
8 2 4 3 1.27 3.21
9 3 1 3 2.36 5.92
10 3 2 4 1.57 4.04
11 3 3 1 1.43 3.65
12 3 4 2 1.35 3.26
13 4 1 4 2.43 7.92
14 4 2 3 1.58 4.10
15 4 3 2 1.62 3.94
16 4 4 1 1.28 3.08

As can be seen from Fig. 5(a), the bond strength of the combined increased by 8.7, 13.1, and 16.5% and the direct shear strength
specimens increased with the increase of surface roughness. The increased 14.9, 17.7 and 32.8%, respectively, compared to the
splitting strength of the specimens with chisel-high roughness, strength of chisel-low roughness type combined specimens. This
groove-low roughness and groove-high roughness type interfaces indicates that the roughness of the concrete substrate affect the
 T. Guo et al. / Construction and Building Materials 186 (2018) 790–800 795

Table 5
Range analysis of the orthogonal experiments.

Range Splitting tensile test Direct shear test

Factors A B C A B C

K1 1.4825 2.1275 1.380 3.5825 5.7475 3.3775

K2 1.6125 1.4325 1.615 4.1175 3.5825 3.940

K3 1.6775 1.635 1.7125 4.2175 4.250 4.4075

K4 1.7275 1.305 1.7925 4.760 3.0975 4.9525
R 0.245 0.8225 0.4125 1.1775 2.650 1.575

increased 63.0, 9.8 and 25.3% and the direct shear strength
increased 85.6, 15.7 and 37.2%, respectively, which means the
application of bonding agents improved the bond strength signifi-
cantly compared to the specimens with no bonding agents. For
rapid patch repair, the interface was believed as the weakest part
due to the massive cracks and pores generated in the process of
surface preparation. However, the bonding agents can incorporate
into the pores of concrete substrate, resulting in less porosity
between new and old concrete, which is the reason why the bond
strength was improved by using bonding agents [36]. Furthermore,
among the bonding agents, RRMP displayed best bonding result,
which may attribute to its’ good compatibility with the concrete
substrate, followed by epoxy resin and GM-RRMP. It was revealed
that the moisture level of the concrete substrate affect the bond
strength significantly. For instance, as the moisture level of GM-
RRMP was much lower than that of RRM concrete and concrete
substrate, the bonding agents may absorb some water from the
repair material, which may change the properties of interfacial
transition zone to some extent [36]. When applied with epoxy
resin, the moisture level of both concrete substrate and repair
material were changed as epoxy resin is hydrophilic. As for the
RRMP, the w/c of was close to that of concrete substrate, in this
case, the change in humidity had the least effect on the bond
strength. These inferences are consistent with the results of the lit-
erature [66,67] and may explain the difference in the performance
of the three bonding agents.
As can be seen from Fig. 5(c), the bond strength increased with
the increase of curing age. The direct shear strength increased
approximately linearly with curing age while the splitting tensile
strength increased rapidly during 4 h–1 d and the growth rate
decreased after curing for 1d. However, the splitting tensile
strength and the direct shear strength of the combined specimens
after curing for 4 h reached to 1.38 and 3.38 MPa, respectively,
which meet the requirements of rapid repair and reopen the traffic
fast.

3.2.2. Interfacial shear characteristics

In order to analyze the interfacial shear characteristics, the
shearing force and the slip distance were continuously measured
and recorded from the start of loading to the destruction of the
combined specimen during the direct shear test. The shearing force
Fig. 5. The effects of the various factors e.g. (a) effect of surface roughness (b) effect was then converted to direct shear strength using Eq. (3). The
of bonding agents and (c) effect of curing age on the bond strength. RRMP was chosen as the bonding agent and the tests were carried
out after curing for 7 d. The shear-slip curves of combined speci-
mens with different surface process methods are shown in Fig. 6.
bond strength significantly. The enhancement of the bond strength As can be seen from Fig. 6, the curves mainly show two stages.
of high surface roughness specimens can be explained by the In the initial stage of deformation, the slopes of the curves were
increase of effective contact area between the concrete substrate small, which indicates that the slip distance increased rapidly
and the RRM concrete. Besides, the interlocking effect between while the direct shear strength increased slowly. The interlocking
the aggregates was gradually improved as the surface roughness effect between the interfacial layers was enhanced and the stiff-
of concrete substrate increased. ness of the combined specimen increased continuously in this
As can be seen from Fig. 5(b), the splitting tensile strength of stage. Afterwards, the shear strength increased sharply and the
RRMP, GM-RRMP and epoxy resin-treated combined specimens slopes of the curves gradually slowed down after approaching
796 T. Guo et al. / Construction and Building Materials 186 (2018) 790–800

was present in the curves. Besides, the slip distance of concrete

with groove-high roughness was almost three times than that of
surface with chisel-low roughness. With respect to the results
mentioned above, it can be inferred that both the slip distance
and the shear strength increased with the increase of surface
roughness. By increasing the surface roughness, for one hand, the
anchoring of RRM concrete into the irregularities or the grooves
increases; on the other hand, the effective contact area between
the bonding agents with the concrete substrate increases, resulting
in better bonding effect. The high roughness of the substrate delays
the destruction of the combined specimen, which is consistent
with the results of bond strength.

3.2.3. Failure mode

Various failure modes were seen due to the differences in sur-
face process methods, bonding agents and curing time. In this sec-
tion, the typical failure modes were discussed to understand the
Fig. 6. Shear-slip curves of combined concrete.
actual conditions of the interface between concrete substrate and
the repair material, and to evaluate the bonding properties of the
combined specimen. Figs. 7 and 8 display the failure modes of
the peak. The combined specimen suffered from brittle failure the combined specimens in the splitting tensile test and the direct
when it reached to ultimate shear strength. It is worth noting that shear test, respectively. The format of the comment below the pic-
there was no residual shear strength as the combined specimen ture is: surface process method-bonding agent-curing age.
was not subjected to vertical stress during the direct shear test. The interface of chisel-low roughness type specimen was found
Hence, no decline phase after arriving at ultimate shear strength to be flat where the brushed slurry was uniformly distributed (the

Chisel-low roughness; RRMP; curing for 1d Chisel-high roughness; RRMP; curing for 1d

Groove-high roughness; RRMP; curing for 1d Groove-high roughness; GM-RRMP; curing for 1d

Chisel-low roughness; GM-RRMP; curing for 4h Chisel-low roughness; GM-RRMP; curing for 1d

Fig. 7. Failure modes in splitting tensile tests.

 T. Guo et al. / Construction and Building Materials 186 (2018) 790–800 797

Chisel-low roughness; Epoxy Resin; curing for 1d Chisel-high roughness; GM-RRMP; curing for 1d

Chisel-high roughness; RRMP; curing for 1d Groove-high roughness; RRMP; curing for 1d

Fig. 8. Failure modes in direct shear tests.

top two pictures in Fig. 7). The artificial chiseled surface was was smooth and covered with a mixture of dry bonding agent
clearly visible and no obvious fracture on the aggregates of both and cement slurry (the picture in the top left corner of Fig. 8).
concrete substrate and RRM concrete could be observed. The inter- The interface of GM-RRMP treated combined specimen was cov-
face of chisel-high roughness type specimen was uneven where the ered with dry bonding agent (the picture in the top right corner
aggregates were either pulled out or broken and the characteristic of Fig. 8). The fracture occurred in the interface indicating unsatis-
of chiseled surface was no longer obvious. We can therefore con- factory bond strength. The fracture of chisel-high roughness sur-
clude from the comparison that the failure modes are consistent face coated with RRMP and after curing for 1d occurred in repair
with the results of the bond strength tests. As the surface rough- concrete with a large amount of aggregates scattered on the shear
ness increased, the proportion of the aggregates of the repair con- failure surface (the picture in the bottom left corner of Fig. 8). As
crete embedded into the concrete substrate interior increased, for the picture in the bottom right corner of Fig. 8, most repair con-
which in fact increased the effective bonding area and thus crete embedded into the grooves underwent shear failure and a
improving the bonding effect. small amount of repair concrete was pulled out. The fracture sur-
As can be seen from the two pictures in the middle of Fig. 7, the face was smooth and the grooves remained relatively intact. The
grooves were still visible with different interfacial failure modes shearing force in the interface was shared by the grooves and the
characteristics. The fracture of RRMP treated specimen occurred bond stress provided by bonding agents, thus exhibiting the best
in the protruding part of the concrete substrate. This signified that bonding performance.
the failure did happen in concrete substrate rather than the inter- Generally, the findings are in agreement with the failure modes
face, indicating that the bonding effect was impressive. However, classified by Gadri et al. [36]. Results indicate that the failure
the grooves of GM-RRMP treated surface were clear and distinct, modes are highly dependent on the surface roughness, bonding
with some GM-RRMP remained in the grooves. The failure mode agents and curing age. The bond strength and monolithic charac-
presented obvious characteristic of shedding with a small propor- teristic of the combined specimen can be considerably improved
tion of aggregates that were broke. The failure modes reflect the by high surface roughness and proper bonding agent. Actually, it
bond strength of the combined specimen and the effectiveness of is the interlock effect between the aggregates that contribute to
the bonding agents to some extent. It can be inferred from the high bond strength.
observation of the failure modes that the speed of coagulation
and solidification of the GM-RRMP was too fast, resulting in insuf- 3.3. The freeze-thaw cycle tests
ficient bond strength of the interface.
As can be seen from the two pictures at the bottom of Fig. 7, Based on the orthogonal experiments outcomes, the RRMP was
after curing for 4 h, clearly display the bonding agent attached to selected as the bonding agent. The splitting tensile tests and the
the repaired concrete and the apparent chiseled surface of concrete direct shear tests were conducted to evaluate the compatibility
substrate. The bonding agent was observed on both concrete sub- between the repair concrete and concrete substrate and to investi-
strate and the repair concrete after curing for 1d, which indicates gate the influence of freeze-thaw cycle on the bond strength. The
that curing age has a great influence on the bond strength, espe- influences of freeze-thaw cycle on bond strength are shown in Fig. 9.
cially in the process of coagulation and penetration of the bonding It is clear that both splitting tensile strength and direct shear
agents. strength decreased with the increase of freeze-thaw cycle number
Fig. 8 shows the damage degree that increased successively. In (Fig. 9).
fact, the direct shear strength increased gradually as introduced in After the freeze-thaw cycles reached to 100, the splitting tensile
Table 4. The interface of epoxy resin treated combined specimen strength and direct shear strength of chisel-low roughness type
798 T. Guo et al. / Construction and Building Materials 186 (2018) 790–800

Fig. 9. Influence of freeze-thaw cycle on bond strength (a) splitting tensile strength and (b) direct shear strength.

specimens decreased to 44.2 and 43.8% of its original bond (3) The bond strength measured by direct shear tests was 2.15
strength before freeze-thaw cycle, while those of groove-high to 3.26 times of that measured by splitting tensile tests.
roughness type specimens reduced to 74.5 and 66.3% of its original The bond strength measured by these methods can provide
bond strength, respectively. This indicates that groove-high rough- reference for the bond performance of repair concrete under
ness type prepared specimens showed good bond strength after different stress conditions.
100 times of freeze-thaw cycles, thus validating the compatibility (4) The surface roughness had a great influence on the shear-
of the rapid repair material. It implies that excellent surface rough- slip properties of the combined concrete. High surface
ness is an important parameter for showing good bond perfor- roughness delayed the destruction of concrete under shear
mance between the repair concrete and concrete substrate. stress, thus improved the bond strength.
To summarize, for rapid patch repair, the high early strength of (5) The results of the orthogonal experiments were supported
RRM allows the concrete pavement repair to open up traffic by the failure modes analysis. An appropriate bonding agent,
quickly and the excellent bond strength and good compatibility a well prepared surface with high surface roughness and
with the concrete substrate permit the repair system to cope well adequate curing age were essential for the satisfactory bond-
with the combined effect of repeated vehicle load and stress ing performance. The failure for the combined specimens
caused by temperature and humidity changes. with low bond strength mostly occurred at the interface
and the corresponding damage surface was smooth with
polling out of or broken rare aggregate. The specimens with
4. Conclusions high bond strength were seriously damaged at the repair
concrete or the protruding part of the grooves, rather than
The bond strength between a novel rapid repair material and ordi- the interface, indicating excellent bonding performance.
nary concrete substrate has been investigated using orthogonal (6) The bond strength decreased with increasing the freeze-
experiments. The influence of surface roughness, bonding agents thaw cycle number. However, the splitting tensile strength
and curing age on the bond strength were thereafter analyzed. and the direct shear strength of (d)-type combined speci-
Besides, the bond strength was assessed through splitting tensile mens showed 74.5 and 66.3% of its original values after
tests and self-designed direct shear tests. In addition, the freeze- 100 cycles of freeze-thaw, representing that the compatibil-
thaw tests were employed to evaluate the durability of the interface ity between repair concrete and concrete substrate is
and the compatibility of the rapid repair material. Based on the acceptable.
results obtained, the following conclusions can be drawn:
Finally, the experimental results obtained reveal that the novel
(1) The compressive strength and flexural strength of RRM con- rapid repair material is a promising alternative for the repair or
crete after curing for 4 h reached at 50.4 and 5.3 MPa, rehabilitation of concrete pavements, the durability of the repaired
respectively, accounting for 80.8 and 88.9% of the corre- system will be studied in future.
sponding values of ordinary concrete at aged for 28d. The
characteristic of high early strength is very beneficial for Conflict of interest
rapid repair and reopen traffic fast.
(2) In the orthogonal experiments, the bonding agents displayed None.
greatest influence on the bond strength, followed by curing
age and surface process methods. Among the levels of these Acknowledgements
factors, RRMP showed best bonding effect; high surface
roughness enhanced the bond strength greatly, the bond The authors would like to acknowledge the support provided by
strength had a good correlation with curing age. the Faculty of Airport Pavement Laboratory for giving the access to
 T. Guo et al. / Construction and Building Materials 186 (2018) 790–800 799

the Laboratories for performing experiments and particularly, Pro- [25] Kyung-Chae Jung, In-Taek Roh, Seung-Hwan Chang, Evaluation of mechanical
properties of polymer concretes for the rapid repair of runways, Compos.: Part
fessor Xingzhong Weng, for his continuous technical support and
B 58 (2014) 352–360.
valuable advice on the data processing. The support from the Nat- [26] Yanping Li, Lin Shen, Reza Mirmoghtadaei, Li Ai, A design of experiment
ural Science Foundation of China (grant No. 51608526) is approach to study the effects of raw material on the performance of
appreciated. geopolymer concrete, Adv. Civil Eng. Mater. 6 (1) (2017) 526–599, https://
doi.org/10.1520/ACEM20160007.
[27] In-Taek Roh, Kyung-Chae Jung, Seung-Hwan Chang, Yoon-Ho Cho,
References Characterization of compliant polymer concretes for rapid repair of runways,
Constr. Build. Mater. 78 (2015) 77–84.
[28] Kyung-Chae Jung, In-Taek Roh, Seung-Hwan Chang, Thermal behavior and
[1] Y. Ge, X.C. Zhu, Y.X. Zhu, Y. Li, Reinforcement Corrosion and Cathodic
performance evaluation of epoxy-based polymer concretes containing silicone
Protection in Concrete (in Chinese), China Machine Press, Beijing, 2007.
rubber for use as runway repair materials, Compos. Struct. 119 (2015) 195–
[2] TRIP, Key Facts about America’s Surface Transportation System and Federal
205.
Funding, The Road Information Program (TRIP), Washington, DC, 2012.
[29] Jong-Pil Won, Joon-Mo Kim, Su-Jin Lee, Sang-Woo Lee, Sung-Ki Park, Mix
Available at: www.tripnet.org/docs/TRIP_National_Fact_Sheet_April_2012.pdf.
proportion of high-strength, roller-compacted, latex-modified rapid-set
[3] Jueshi Qian, Chao You, Qingzhen Wang, Hongtao Wang, Xingwen Jia, A method
concrete for rapid road repair, Constr. Build. Mater. 25 (2011) 1796–1800.
for assessing bond performance of cement-based repair materials, Constr.
[30] Tanakorn Phoo-ngernkham, Vanchai Sata, Sakonwan Hanjitsuwan,
Build. Mater. 68 (2014) 307–313, https://doi.org/10.1016/
Charoenchai Ridtirud, Shigemitsu Hatanaka, Prinya Chindaprasirt, High
j.conbuildmat.2014.06.048.
calcium fly ash geopolymer mortar containing Portland cement for use as
[4] I.-T. Roh, K.-C. Jung, S.-H. Chang, et al., Characterization of compliant polymer
repair material, Constr. Build. Mater. 98 (2015) 482–488.
concretes for rapid repair of runways, Constr. Build. Mater. 78 (2015) 77–84.
[31] Sulaem Musaddiq Laskar, Sudip Talukdar, Preparation and tests for
[5] H. Alanazi, M. Yang, D. Zhang, et al., Bond strength of PCC pavement repairs
workability, compressive and bond strength of ultra-fine slag based
using metakaolin-based geopolymer mortar, Cem. Concr. Compos. 65 (2016)
geopolymer as concrete repairing agent, Constr. Build. Mater. 154 (2017)
75–82.
176–190.
[6] Y.-L. Chen, C.-J. Lin, M.-S. Ko, et al., Characterization of mortars from belite-rich
[32] Lucy P. Priddy, Mariely Mejias-Santiago, Jeb S. Tingle. Development of foam
clinkers produced from inorganic wastes, Cem. Concr. Compos. 33 (2011) 261–
backfill repair techniques for airfield pavement repairs, Transportation
266.
Research Record: Journal of the Transportation Research Board, No. 2569,
[7] M.H. Fiebrich, Scientific aspects of adhesion phenomena in the interface
Transportation Research Board, Washington, D.C., 2016, pp. 53–61. DOI:
mineral substrate-polymers, in: F.H. Wittman (Ed.), Adherence of young and
10.3141/2569-06.
old concrete, Proceedings of 2nd Bolomey Workshop, Aedificatio Verlag,
[33] L.P. Priddy, J.K. Newman, Full-scale field testing for verification of mechanical
Unterengstringen, 1994.
properties of polyurethane foams for use as backfill in PCC repairs, ASCE J.
[8] B. Rabehi, Y. Ghernouti, A. Li, Comparative behavior under compression of
Mater. Civil Eng. 22 (3) (2010) 245–252.
concrete columns repaired by fiber reinforced polymer (FRP) jacketing and
[34] Cristina Zanotti, Nemkumar Banthia, Giovanni Plizzari, A study of some factors
ultra high-performance fiber reinforced concrete (UHPFRC), J. Adhes. Sci.
affecting bond in cementitious fiber reinforced repairs, Cem. Concr. Res. 63
Technol. 28 (22–23) (2014) 2327–2346.
(2014) 117–126.
[9] A. Ghazy, M.T. Bassuoni, A. Shalaby, Nano-modifed fly ash concrete: a repair
[35] A. Momayez, M.R. Ehsani, A.A. Ramezanianpour, H. Rajaie, Comparison of
option for concrete pavements, ACI Mater. J. 113 (1–6) (2016) 1–12.
methods for evaluating bond strength between concrete substrate and repair
[10] Wang Kejin, Lomboy Gilson, Developing green, highly flowable, rapid set,
materials, Cem. Concr. Res. 35 (2005) 748–757.
high-performance concrete for pavement patch repair, 2016. InTrans Project
[36] Yan He, Xiong Zhang, R.D. Hooton, Xiaowei Zhang, Effects of interface
Reports. 214. http://lib.dr.iastate.edu/intrans_reports/214.
roughness and interface adhesion on new-to-old concrete bonding, Constr.
[11] Masih Mohammadi, Reza Mir Moghtadaei, Nasibe Ashraf Samani, Influence of
Build. Mater. 151 (2017) 582–590.
silica fume and metakaolin with two different types of interfacial adhesives on
[37] Bassam A. Tayeh, B.H. Abu Bakar, M.A. Megat Johari, Yen Lei Voo, Evaluation of
the bond strength of repaired concrete, Constr. Build. Mater. 51 (2014) 141–
bond strength between normal concrete substrate and ultra high performance
150.
fiber concrete as a repair material, Proc. Eng. 54 (2013) 554–563.
[12] Anna Sidorova, Enric Vazquez-Ramonich, Marilda Barra-Bizinotto, Joan Josep
[38] B.A. Tayeh, B.H. Abu Bakar, M.A. Megat Johari, Yen Lei Voo, Mechanical and
Roa-Rovira, Emilio Jimenez-Pique, Study of the recycled aggregates nature’s
permeability properties of the interface between normal concrete substrate
influence on the aggregate–cement paste interface and ITZ, Constr. Build.
and ultra high performance fiber concrete overlay, Constr. Build. Mater. 36
Mater. 68 (2014) 677–684.
(2012) 538–548.
[13] Luxia Song, Zhen Li, Ping Duan, Min Huang, Xiaofei Hao, Yu. Yongsheng, Novel
[39] H.M. Elbakry, A.M. Tarabia, Factors affecting bond strength of RC column
low cost and durable rapid-repair material derived from industrial and
jackets, Alexandria Eng. J. 55 (1) (2016) 57–67.
agricultural by-products, Ceram. Int. (2017), https://doi.org/10.1016/j.
[40] YooSeok Jung, Wuguang Lin, Han Hao, Yoon-Ho Cho, Interface behavior of
ceramint.2017.07.106.
partial depth repair for airport concrete pavement subjected to differential
[14] Jiusu Li, Wenbo Zhang, Yong Cao, Laboratory evaluation of magnesium
volume change, Constr. Build. Mater. 143 (2017) 515–521.
phosphate cement paste and mortar for rapid repair of cement concrete
[41] Karima Gadri, Abdelhamid Guettala, Evaluation of bond strength between
pavement, Constr. Build. Mater. 58 (2014) 122–128.
sand concrete as new repair material and ordinary concrete substrate (The
[15] Quanbing Yang, Durability and applications of magnesium phosphate material
surface roughness effect), Constr. Build. Mater. 157 (2017) 1133–1144.
for rapid repair of pavements, Adv. Mater. Res. 857 (2014) 81–89, https://doi.
[42] Luc Courard, Tomasz Piotrowski, Andrzej Garbacz, Near-to-surface properties
org/10.4028/www.scientific.net/AMR.857.81.
affecting bond strength in concrete repair, Cem. Concr. Compos. 46 (2014) 73–
[16] Fei Qiao, C.K. Chau, Zongjin Li, Property evaluation of magnesium phosphate
80.
cement mortar as patch repair material, Constr. Build. Mater. 24 (2010) 695–
[43] R. Sen, Developments in the durability of FRP-concrete bond, Constr. Build.
700.
Mater. 78 (2015) 112–125.
[17] Dongshuai Hou, Handong Yan, Jinrui Zhang, Penggang Wang, Zongjin Li,
[44] J.R. Cromwell, K.A. Harries, B.M. Shahrooz, Environmental durability of
Experimental and computational investigation of magnesium phosphate
externally bonded FRP materials intended for repair of concrete structures,
cement mortar, Constr. Build. Mater. 112 (2016) 331–342.
Constr. Build. Mater. 25 (5) (2011) 2528–2539.
[18] D.R. Morgan, Compatibility of concrete repair materials and systems, Constr.
[45] A. Garbacz, M. Gorka, L. Courard, Effect of concrete surface treatment on
Build. Mater. 10 (1) (1996) 57–67.
adhesion in repair systems, Mag. Concr. Res. 57 (1) (2005) 49–60.
[19] N.F. Ariffin, M.W. Hussin, A.R.M. Sam, et al., Strength properties and molecular
[46] R. Mirmoghtadaei, M. Mohammadi, N. Ashraf Samani, S. Mousavi, The impact
composition of epoxy-modified mortars, Constr. Build. Mater. 94 (2015) 315–
of surface preparation on the bond strength of repaired concrete by
322.
metakaolin containing concrete, Constr. Build. Mater. 80 (2015) 76–83,
[20] E.-H. Hwang, J.-M. Kim, J.H. Yeon, Characteristics of polyester polymer
https://doi.org/10.1016/j.conbuildmat.2015.01.018.
concrete using spherical aggregates from industrial by-products, J. Appl.
[47] P.W. Chen, X. Fu, D.D.L. Chung, Improving the bonding between old and new
Polym. Sci. 129 (2013) 2905–2912.
concrete by adding carbon fibers to the new concrete, Cem. Concr. Res. 25 (3)
[21] Ghasan Fahim Huseien, Jahangir Mirza, Mohammad Ismail, S.K. Ghoshal,
(2005) 491–496.
Ahmed Abdulameer Hussein, Geopolymer mortars as sustainable repair
[48] S. Pedro, J. Eduardo, Factors affecting bond strength between new and old
material: a comprehensive review, Renew. Sustain. Energy Rev. 80 (2017)
concrete, ACI Mater. J. (2011).
54–74.
[49] D. Snoeck, J. Wang, D.P. Bentz, N. De Belie, Applying a biodeposition layer to
[22] Xu Zhang, Xiaoqun Wang, Xuehong Xu, Yueqing Zhao, Microwave curing
increase the bond of a repair mortar on a mortar substrate, Cem. Concr.
process and mechanical properties study of epoxy mortars for repairing
Compos. 86 (2018) 30–39.
concrete pavement rapidly, J. Reinf. Plast. Compos. 36 (6) (2017) 443–451.
[50] D.J. Cleland, Adrian E. Long, The pull-off test for concrete patch repairs, Proc.
[23] Andrea Saccani, Vittorio Magnaghi, Durability of epoxy resin-based materials
Inst. Civil Eng. Struct. Build. 122 (4) (2015) 451–460.
for the repair of damaged cementitious composites, Cem. Concr. Res. 29 (1999)
[51] Caroline Talbot, Michel Pigeon, Denis Beaupre, D.R. Morgan, Influence of
95–98.
surface preparation on long-term bonding of shotcrete, ACI Mater. J. 91 (6)
[24] Rodrigo H. Geraldo, Otacisio G. Teixeira, Samile R.C. Matos, Francisco G.S. Silva,
(1994) 560–566.
Jardel P. Gonçalves, Gladis Camarini, Study of alkali-activated mortar used as
[52] S. Austin, P. Robins, Y.G. Pan, Shear bond testing of concrete repairs, Cem.
conventional repair in reinforced concrete, Constr. Build. Mater. 165 (2018)
Concr. Res. 29 (7) (1999) 1067–1076.
914–919.
800 T. Guo et al. / Construction and Building Materials 186 (2018) 790–800

[53] A. Momayez, M.R. Ehsani, H. Rajaie, A. Ramezanianpour, Cylindrical specimen [60] Zheng Xuguang, in: Research on high bridge expansion joint transition zone
for measuring shrinkage in repaired concrete members, Constr. Build. Mater. rapid repair material, Master dissertation of Chong Qing Jiaotong University,
19 (2) (2005) 107–116. 2016, pp. 21–28.
[54] E. Bonaldo, J.A. Barros, P.B. Lourenço, Bond characterization between concrete [61] D.S. Santos, P.M.D. Santos, D. Dias-da-Costa, Effect of surface preparation and
substrate and repairing SFRC using pull-off testing, Int. J. Adhes. Adhes. 25 (6) bonding agent on the concrete-to-concrete interface strength, Constr. Build.
(2005) 463–474. Mater. 37 (2012) 102–110.
[55] B.A. Tayeh, B.H. Abu Bakar, M.A. Megat Johari, A.M. Zeyad, Flexural strength [62] P.M.D. Santos, E.N.B.S. Julio, V.D. Silva, Correlation between concrete-to-
behavior of composite UHPFC-existing concrete, Adv. Mater. Res. 701 (2013) concrete bond strength and the roughness of the substrate surface, Constr.
32–36. Build. Mater. 21 (8) (2007) 1688–1695.
[56] A.D. Espeche, J. León, Estimation of bond strength envelopes for old-to-new [63] L. Courard, Adhesion of repair systems to concrete: influence of interfacial
concrete interfaces based on a cylinder splitting test, Constr. Build. Mater. 25 topography and transport phenomena, Mag. Concr. Res. 57 (5) (2005) 273–
(3) (2011) 1222–1235. 282.
[57] B.A. Tayeh, B.A. Bakar, M.M. Johari, Characterization of the interfacial bond [64] Xiong Zhang, Lei Zhang, Characterization and properties of artificial roughness
between old concrete substrate and ultra high performance fiber concrete of young on old concrete bonding interface, J. Tongji Univ. (Nat. Sci.) 41 (5)
repair composite, Mater. Struct. 46 (5) (2013) 743–753. (2013) 753–759.
[58] Ahmed M. Diab, M. Abd Elmoaty, Mohamed R. Tag, Eldin, Slant shear bond [65] Test methods of cement and concrete for highway engineering, JTG E30-2005.
strength between self compacting concrete and old concrete, Constr. Build. [66] P.H. Emmons, Concrete Repair and Maintenance, in: Part Three: Surface Repair,
Mater. 130 (2017) 73–82. Section 6: Bonding Repair Materials to Existing Concrete, R.S. Means Company,
[59] Yang Yang, in: Research of rapid repair materials for bridge expansion joints MA, 1994. pp. 154–163.
rehabilitation, Master dissertation of Chong Qing Jiaotong University, 2017, pp. [67] S. Austin, P. Robins, Y. Pan, Tensile bond testing of concrete repairs, Mater.
34–40. Struct. 28 (5) (1995) 249–259.