Large scale experimental study on bond behavior between polymer modified cement mortar layer and concrete

The effectiveness of concrete structures reinforced by the high strength steel wire mesh-polymer mortar depends on the bond-slip behavior of the interface between the polymer modified cement mortar layer (referred as mortar layer) and the concrete. The areas of the mortar layer tested in the previous studies are generally very small and the obtained conclusions are with limited applications. In this study, the experimental approach was used to analyze the slip behavior of the interface between the concrete block and a larger scale mortar layer. 30 concrete blocks retrofitted with mortar layers were fabricated and the single shear tests on these specimens were conducted. The applied shear loads, slip displacement between the concrete block and the mortar, as well as the strain of the mortar were monitored during tests. The experimental results exhibit the specimens have two representative failure types: debond with and without cracks. Then the bonding performance of the interface was systematically explored by analyzing the failure modes of the specimens and the mechanical property of the mortar. Meanwhile, the slip behavior of the interface influenced by the bond area, the thickness of the mortar layer, the mortar strength and the roughness of the interface were discussed. It was presented that the interface roughness treatment and increasing the mortar strength could significantly improve the behavior of the interface.

Large scale experimental study on bond behavior between polymer

modified cement mortar layer and concrete

Beijing Higher Institution Engineering Research Center of Civil Engineering Structure and Renewable Material, Beijing Advanced Innovation Center for Future Urban Design & Beijing University of Civil Engineering and Architecture, Beijing 100044, China

h i g h l i g h t s

Tests on large-scale concrete blocks retrofitted with mortar layers

Two typical failure types of crack or debond

Layer thickness, bond length, mortar strength and interface roughness influencing the shear strength of interface

a r t i c l e i n f o

Article history:

Received 10 September 2018

Received in revised form 7 June 2019

Accepted 16 August 2019

Keywords:

Mortar layer

Interface

Single shear test

Bond-slip

Debonding

a b s t r a c t The effectiveness of concrete structures reinforced by the high strength steel wire mesh-polymer mortar depends on the bond-slip behavior of the interface between the polymer modified cement mortar layer (referred as mortar layer) and the concrete The areas of the mortar layer tested in the previous studies are generally very small and the obtained conclusions are with limited applications In this study, the experimental approach was used to analyze the slip behavior of the interface between the concrete block and a larger scale mortar layer 30 concrete blocks retrofitted with mortar layers were fabricated and the single shear tests on these specimens were conducted The applied shear loads, slip displacement between the concrete block and the mortar, as well as the strain of the mortar were monitored during tests The experimental results exhibit the specimens have two representative failure types: debond with and without cracks Then the bonding performance of the interface was systematically explored by ana-lyzing the failure modes of the specimens and the mechanical property of the mortar Meanwhile, the slip behavior of the interface influenced by the bond area, the thickness of the mortar layer, the mortar strength and the roughness of the interface were discussed It was presented that the interface roughness treatment and increasing the mortar strength could significantly improve the behavior of the interface

Ó 2019 Elsevier Ltd All rights reserved

1 Introduction

Retrofitting approaches were proposed in order to repair,

strengthening and updating existing reinforced concrete (RC)

structures to resist higher loads, improve load-carrying capacity

of structures[1–5] The commonly utilized systems are generally

made of fiber sheets embedded in an epoxy matrix (i.e., fiber

rein-forced polymers, FRP) due to the high strength to weight ratio, high

corrosion resistance, convenience application and minimal section

area change[6,7] The effectiveness of the FRP systems in

retrofit-ting RC structures has been proved by numerous experimental and

theoretical studies Good bond behaviors of CFRP sheets attached

to concrete have been validated[8] The externally bonded rein-forcement on grooves techniques could improve the FRP-concrete bond strength[9–13]and eliminate the debonding failure of the interface [14] However, due to the poor fire resistance of the organic matrix, poor thermal compatibility with concrete, suscep-tibility to radiations, the FRP is inapplicably applied on wet sur-faces and at low-temperatures[15] Relatively, the high strength steel wires mesh-polymer mortar (HSWM-PM) composite layer

as another structural strengthening and retrofitting system has been developed rapidly because of its good durability, material compatibility, adhesive properties and high-temperature resis-tance[16] Moreover, the effectiveness of the HSWM-PM system retrofitting columns[17–21], beams and other structures[22–33] were also evaluated

The HSWM-PM system consists of the high strength steel wires, mortar layer [34,35] and the interfaces, which are between the https://doi.org/10.1016/j.conbuildmat.2019.116751

0950-0618/Ó 2019 Elsevier Ltd All rights reserved.

⇑ Corresponding author.

E-mail addresses: liaoweizhang@bucea.edu.cn (W Liao), machao@bucea.edu.cn

(C Ma).

Contents lists available atScienceDirect

Construction and Building Materials

j o u r n a l h o m e p a g e : w w w e l s e v i e r c o m / l o c a t e / c o n b u i l d m a t

mortar layer and the concrete as well as between the steel wires

and mortar layer Generally, the high strength steel wires play

two roles in this system: like the longitudinal bars in RC structures

carrying the tensile forces, like hooping bars in RC structures

carry-ing shear forces and providcarry-ing confincarry-ing loadcarry-ing to improve the

shear strength of the mortar The load carrying capacity of the

structures retrofitted by the HSWM-PM system mainly depends

on the bonding behavior of the interfaces, and debond failure of

the interfaces would result in the loss of the reinforcement effect

[36,37]

The bonding behavior of the interface between the steel wires

and mortar layer influenced by the strength of mortar-based

matrix[38], the diameter and anchorage length of the steel strand

[39,40] and the density of the steel strand mesh[16] has been

studied in detail, and a series of valuable conclusions were

acquired Such as, the mineral mortar with natural kaolin and

bauxite can be used to enhance the bonding force of the interface

The grid density of the steel wire mesh is not directly proportional

to the bond strength of the interface The critical anchorage length

of the strand has a linear relationship with the diameter of steel

strand[40] Accordingly, to ensure enough bonding force, the bond

length of the strand should be greater than the critical anchorage

length[41] Theoretical and numerical approaches were also used

to analyze the stress distribution rule[41,42] and the bond-slip

distribution of the prestressing strands, furtherly, to analyze the

transmission and anchorage lengths of prestressing steel strands

in different mortar-based matrixes[43] The reasonable

arrange-ment of the steel strand will lead to a well-distributed bond stress

[16] These experimental data and theoretical analyses present that

the interface between the steel wires and mortar layer exhibits a

soundness behavior[38–44]

On the other hand, the reinforcement effect of the composite

layer also depends on the bond performance between the mortar

layer and the concrete The bonding performance between the

mortar layer and concrete not only involves the strength of the

bonding interface, but also involves the durability of the interface

The bond-slip mechanism of the bonding interface between the

mortar layer and the concrete is the same as the interface between

old concrete and new concrete[45,46], this is because the bonding

action in these interfaces is mainly achieved through the

mechan-ical interlock between the inorganic matrices However, there are

still differences in mechanical properties and microscopic

molecu-lar structures between the mortar and concrete, such as the age

between new and existing concretes and the relative stiffness

between the substrate concrete and added concrete layer

Conse-quently, the achievements of the bonding mechanism of the

concrete-concrete interface[47]could not be suitable for the

inter-face between the mortar and concrete directly

To analyze the bonding mechanism of the interface between the

mortar layer and the concrete, experiments [48–55] were

con-ducted on the mortar-based system Experiment data presented

that crack[56]and debond[57]of the mortar layer were the main

failure modes in the mortar-concrete interface Moreover, the

cur-ing time, strength of the concrete and mortar, interface roughness

and repair position also greatly influence the bonding behavior of

the interface The mechanical occlusion of the mortar layer and

the concrete interface are the important guarantee for the

rein-forcement effect of the whole composite layer The bonding

beha-viour between the magnesium potassium phosphate cement and

the concrete substrate was evaluated based on the pullout tests

[58] Besides, an excellent bonding property between the

magne-sium potasmagne-sium phosphate cement and the old concrete was

veri-fied [59]by the splitting tests For example, the Alkali-activated

mortar could also develop a fast and good adherence strength

(30 MPa after 24 h) to the concrete substrate[60–64] However,

during the past studies, the bond area of mortar layer was usually

less than 200 mm 200 mm and the deformation of mortar layer was small and almost with no damage occurring before the inter-face producing failure In fact, the contribution of the thickness and bond area of the mortar matrices to the bond performance need to

be investigated Moreover, shear tests on the interface between mortar layer and concrete at small area cannot be used to ade-quately explain the interface bonding mechanism, and the corre-sponding conclusions as the design reference directly in actual engineering are defective

In order to demonstrate the damage process of the mortar layer, and to analyze the difference of bonding performances influenced

by the bond length, mortar strength, interface treatment and the thickness of mortar layers, tests on the mortar layer and the con-crete interface were conducted in this study Then the bond mech-anism and failure mode of the larger bond area were systematically studied based on the experimental data

2 Test program 2.1 Test specimen preparation

30 large-scale plain concrete blocks retrofitted with mortar layer, which sketch is shown inFig 1, were designed and tested

to investigate the bond-slip behaviors influenced by the bond area, thickness and strength of the mortar layer, as well as the roughness

of the interface between the mortar layer and concrete Each spec-imen consists of a plain concrete block with the size of

300 mm 300 mm  900 mm and a mortar layer There are three steps to fabricate the concrete specimens: plain concrete block preparing, concrete surface chiseling and mortar layer laying The plain concrete blocks were fabricated using concrete with grade

of C30 and the largest aggregate size of 25 mm Then after 28 days curing, manually chiseling was conducted at the surface of the crete blocks to increase the mechanical occlusion between the con-crete block and the mortar layer The chiseled concon-crete block is illustrated inFig 2(a) Finally, the mortar layer was then equably laid on the concrete blocks after the chiseling procedure, the fabri-cated concrete specimen is shown inFig 2(b) To prevent the evap-oration of the water, the specimens were covered by plastic sheets during the curing of the mortar layer, shown inFig 2(c) Note that three blocks were not chiseled as the comparative cases The detailed information about the 30 specimens is listed in

Note: In the experiment, the thin layer of mortar fell off before the W1-1 and W4-1 specimens were loaded, therefore, the final quantity of test specimens is only one

In the process of the surface chiseling, the maximum depth is

controlled as 12 mm The sand filling method was used to estimate

uniformity of the interface roughness.Fig 3shows procedure of

the sand filling, and the roughness is calculated by the Eq.(1)

r¼V

where, r is the roughness at the bonding interface, Vmm2is the

vol-ume of standard sand filled in the chiseled holes, A is the area of the

bonding interface The standard sand particle size ranges from

0.08 mm to 2.0 mm The calculated roughness is around 1.27 mm

with variance of only 0.036 mm Moreover, in order to analyze the discreteness of interface chiseling, 8 specimens with band area

of 350 mm 300 mm presented inTable 1were designed The mixture ratio of the concrete is presented inTable 2, and the average compressive strengths of 25.8 MPa and 36.1 MPa were respectively obtained from cubic tests after 7 days and 28 days casting The major ingredients of the applied mortars are the high-performance cement, 20–40 mesh sand, 40–70 mesh sand, chopped fiber, additives and water The mortars are referred to

as the commercially TCPM polymer mortar with the strength grade

of M30, M40, M50 The component proportions of the polymer mortar are shown in Table 3 and the measured compressive strengths of the mortars are listed inTable 4 The interface treating

(a) Concrete blocks after chiseled

(b) Mortar layer

(c) Specimen during curing Fig 2 Preparation process of specimens.

(a) Interface treatment

(b) Laying standard sand

(c) Measuring standard sand

Fig 3 Evaluation of interface roughness.

agent was used in the concrete surface commercially known as

YT-302 type two-specimen treatment agent The 14-day shear bond

strength and tensile strength of the agent obtained from the

man-ufacturer are 1.7 MPa and 0.9 MPa, respectively

2.2 Test setup Tests were conducted by applying quasi-static single shear loading at the interface between the concrete blocks and mortar layer Considering the areas of the mortar of the specimen are rather large, to finish the tests smoothly, a hydraulic jack operated

by the manual pump (shown inFig 4) was used to provide the sin-gle shear loading by the horizontal thrust between the lifting jack and the concrete block The specimens were placed between the lifting jack and the force-transmitting steel plate The hydraulic jack and the force-transmitting steel plate are connected by four steel bars The force-transmitting steel plate is welded with the four steel bars, and the hydraulic jack is also fixed with the steel bars The movement of the mortar layer was restricted by the steel

Table 1

Test specimen mortar layer details.

Number Mortar layer thickness (mm) Bonding surface length (mm) Bond surface width (mm) Mortar strength grade Chiseling or not Quantity

Table 2

Mixture ratio of concrete (Unit: kg).

Material P.O42.5 cement Water Sand Crushed stone BM-PM001 water reducer F kind of Fly ash grade I Blast furnace slag powder S95

Note: the moisture content of sand is 7.8%, and the moisture content of crushed stone is 0.2%.

Table 3

Component proportion of polymer mortar.

Mortar strength grade Water (g) Cement (g) Emulsion powder (g) Early strength agent (g) Silica fume (g) Cellulose ether (g) Fly Ash (g) Sand (g)

Table 4

Strength properties of TCPM high performance mortars.

Mortar

strength

grade

3d compressive

strength (MPa)

7d compressive strength (MPa)

28d compressive strength (MPa)

plate During tests, the loading device and specimens were placed

horizontally as shown inFig 5, the horizontal thrust compels the

relative horizontal displacement and provides the single shear

force between the concrete block and mortar layer In the

follow-ing, the end of the concrete block close to the hydraulic jack is

referred as the loading end, the end of the mortar layer close to

the force-transmitting steel plate is referred as the constraint end and the other end of the mortar layer is referred as the free end,

as shown inFig 5

It should be emphasized that the applied shear loading should

be concentric with the longitudinal axis of the specimen in case

of avoiding the eccentric effect During the quasi-static loading,

Fig 5 Setup of the experimental device and specimen.

(a) Sketch of measuring points arrangement

(b) Layout of measuring points arrangement

the loading procedure will be stopped when any one of the

follow-ing conditions occurs: (1) obvious misalignment or mortar layer

slip occurs, (2) the local mortar layer is obviously cracked or

crushed; (3) the load bearing capacity of specimens decreases

obviously due to other uncertainties

2.3 Test observations

A total of 6 Sz120-100AA-type strain gauges were bonded at the

surface of the mortar to monitor the surface strains during testing

The gauges were glued along 3 rows transversely and 2 columns

longitudinally Locations and labels of the strain gauges are

illus-trated inFig 6 The gauges are glued 100 mm and 200 mm away

from the constraint end and with the transverse spacing of

150 mm The shear load was monitored by the hydraulic pressure

conversion Two rod-type displacement meters with the range

of ±25 mm were arranged on the upper left of the concrete block

and the free end of the mortar layer to respectively monitor the

horizontal displacements of the mortar layer and the concrete

Dis-placement meter 1 is used to measure the disDis-placement of the

con-crete block, and displacement meter number 2 is used to measure

the deformation of the mortar layer Then the difference between

the two measured displacement is the sliding displacement of

the interface

3 Test results and parameter studies According to the test observations, the failure types of all the specimens could be classified into two categories: uncracked and cracked types The uncracked type is defined as that there is no vis-ible crack in the mortar layer during the loading process, and the cracked type is defined as that there are visible cracks in the mor-tar layer during the process These two failure types of the mormor-tar layer mainly include five failure modes: (1) The failure mode of the complete debond with cracks is that the mortar layer experiences the formation and development of cracks and eventually produced debond failure, shown inFig 7 (2) The failure mode of the com-plete debond with no cracks is that the mortar layer was gradually debond but the mortar layer did not produce cracks, given inFig 8 (3) The failure mode of the local crushing with expansion is that crushing failure occurred in the mortar layer with expansion shown inFig 9 (4) The failure mode of the local debond is that the mortar layer produced local debond as shown inFig 10 (5) The failure mode of the local debond with the crush is that the mortar layer produced local debond with the crush at the con-straint end as given inFig 11 In the uncracked type, noticeable compression deformation, local crushing can be obtained during loading as shown in Figs 9 and 11 Besides, complete debond and local debond without cracks of the mortar layer can be observed as shown inFigs 8and10 In the cracked type, the mor-tar layer cracking includes three stages: the formation of cracks,

(a) Failure mode of W1-2-1

(b) Failure mode of W2-1-1

(a) Failure mode of W1-2-2

(b) Failure mode of W1-4-2 Fig 8 Failure modes of complete debond with no cracks.

the development of cracks and cracking failure Mortar embedded

in the concrete blocks can be seen inFig 12, it proves that the

char-acteristics of mortar in concrete interface after chiseling are

obvious

According to the loading scheme, the shear stress and strain at

the interface should not be even as presented inFig 13 The

max-imum value of the shear stress and strain depends on the applied

load, and the minimum value of the shear stress and strain depends on the bond length of the mortar layer Particularly, if the mortar has enough length, there might be no stress and strain

at the free end Even so, the applied load and the horizontal dis-placement between the mortar layer and the concrete could be applied as the indicators to discuss the shear performances of the bonding interface Obviously, the means shear stress along the interface could also be a reliable indicator to explore the shear strength of the interface

The failure modes and peak loads of each specimen in the tests are summarized and analyzed inTable 5 The local bond stresss

(MPa) is the average bond stress between the mortar layer and the concrete It can be calculated by dividing the applied force by the bond area as follows:

s¼F

where, F is the load applied at the bonding interface between mor-tar layer and concrete

The failure mode of the mortar layer and the interface bonding performance are greatly affected by the strength of the interface bond area, the thickness of the mortar layer and the roughness of the interface These factors will be discussed respectively in the fol-lowing sections

3.1 Effect of the mortar layer thickness on the bonding performances The failure loads of the specimens of Groups W1, W2, W3 and W4 with mortar layer thickness of 20 mm and 25 mm are con-trastively presented inFig 14 With the increase of the layer thick-ness, the ability of the mortar layer to resist the shear force will be enhanced when the strength grade of mortar and bond area are same Therefore, increasing the thickness of the mortar layer can result in the enhancement of the bond force and the reduction of the occurrence of the complete debond failure of the mortar layer Fig 14also presents the shear load differences of two specimens with same area and different layer thickness, which are values of 7.8%, 1.6% and 1.4% They are calculated by Fð 25 F20Þ=F20, herein,

F20is the failure load of the interface with the mortar layer thick-ness of 20 mm, and the F25is the failure load of the interface with the mortar layer thickness of 25 mm As presented, the differences are no more than 8.0% Hence, when enhance a structure, the mor-tar layer with thickness of 20 mm might be enough

3.2 2 Effect of the bond length on the bonding performances Fig 15presents the relation between the bond length L and the bond stresss.Fig 15(a) and (b) present all the experimental data

(a) Failure mode of W2-2-1

(b) Failure mode of W2-3-1 Fig 9 Failure modes of local crushing with expansion.

Fig 10 Failure mode of local debond of W2-2–2.

Fig 11 Failure mode of local debond with crush of W3-2–7.

on the mortar layer of 20 mm and 25 mm It exhibits the stability

and feasibility of the tests Fig 14(c) presents the relationship

between the mean bond stresss

and bond length L As presented, when the bond thickness is 20 mm,s

equals 2.47 MPa, with max-imum value of 2.49 MPa and minmax-imum value of 2.45 MPa The

maximum difference value is only 0.044 MPa and 1.78% deviation

Similarly, the average value ofs

is 2.55 MPa when the bond thick-ness is 25 mm, with the maximum value of 2.64 MPa and the

min-imum value of 2.50 MPa The maxmin-imum difference value is only

0.14 MPa and 5.6% deviation Experimental results indicate that

the bond stress at the interface changes slightly when the bond

length is in the range of 250 mm to 400 mm Therefore, the uneven

distribution of the bond stress can be ignored when the bond

length is larger than 250 mm to reinforce a concrete structural

component with the HSWM-PM

3.3 Effect of the mortar strength on the bonding performances

Essentially, the shear strength of the bond surface is the bond

strength between the concrete and mortar, obviously, the mortar

strength partly determines interfacial bond strength.Fig 16

pre-sents the average strength when the bond thickness is 25 mm

Experiment results show that with the increase of strength grade

of the mortar, the shear strength of the interface increases

remark-ably Therefore, the mortar with high strength results in the bond interface with high shear capacity and reducing the occurrence of the interfacial debond failure On the other hand, the failure modes

of the mortar layer with different strength grades are diverse The specimens retrofitted with mortar with lower mortar have the complete debond failure mode, as shown inFig 7(b) On the con-trary, the specimens retrofitted with the higher strength mortar have the crushing and local crushing failure modes, which is shown inFig 9(b)

3.4 Effect of the interface roughness on the bonding performances The bond strength between the concrete and mortar layer is also affected by the mechanical occlusal interaction between the mortar layer and the concrete Interface treatment can change the interface roughness; therefore, interface treatment can signifi-cantly improve the bond strength of the interface Certainly, inter-face treatment can also improve the frictional strength of the interface Interface treatment influencing the frictional strength will not be discussed in this study As presented inTable 5, the average peak shear force of Group W3-2 with the interface treat-ment is 263.75 kN, the average shear force of Group W4-2 without interface treatment is 161.62 kN This proves that interface treat-ment improves 63.2% shear load-carrying capacity of the interface Moreover, the specimens without interface treatment failed with

(a) W2-1-2

(b) W1-2-1

(c) W2-3-3

(d) W4-2-1 Fig 12 Failure mode of the interface.

complete debond mode, it also proves that the interface without treatment has lower shear strength from another perspective

4 Load-slip response 4.1 Interface bond-slip relationship Rod-type displacement meters were used to measure the dis-placements of the loading end and free end, which are respectively the deformations of the concrete blocks and the mortar The differ-ence value of the monitored displacements is the relative displace-ment between the free and constraint ends Because the maximum load applied during tests is about 300 kN, correspondingly, the maximum axial stress of the concrete block is smaller than 3.4 MPa which is far below the average compressive strength of 36.1 MPa This confirms that the deformation of the concrete block

is very small and can be ignored Further, the relative displacement between the free and constraint ends can be treated as the slip dis-placement between the mortar layer and the concrete block Fig 17 shows the load-slip relationship of representative speci-mens, and the behavior of each specimen can be divided into three stages

(1) Preloading stage During this stage, the deformation of the mortar layer increases quickly with the tight compress between the steel plate and the mortar layer of the constraint end because the constraint end is not absolutely smooth However, the shear load hardly increases during this stage The load-slip curves of this stage are nearly par-allel to the horizontal axis

L

Loading

end

(a) Distribution of the shear stress

L

Loading

end

(b) Distribution of the strain

Fig 13 Sketch of shear stress and strain distribution at the interface.

Table 5

Results of shear test.

Number A = L  b (mm  mm) t (mm) Mortar strength grade F(kN) sp = Fp/A(MPa) s

(MPa) Failure mode

Note: t denotes the thickness of the mortar layer; FP denotes peak load; Fdenotes mean failure load;sp denotes bond shear strength;sdenotes mean value of bond shear

(2) Bonding performing stage

During this stage, the bonding effect between the mortar layer

and concrete plays its function The shear force between the

mor-tar and concrete is resisted by the bond force of the interface With

the increase of the applied load, the displacement monitored by

the rod-type displacement meters also increases Because the

interface does not lose the bond strength, the shear load could

increase to its peak value

(3) Failure stage

The shear load decreases quickly after reaching its peak value,

and the interface loses the load bearing capacity during this stage

Moreover, the monitored displacement in this stage is mainly the

slip displacement between the mortar and concrete Meanwhile,

the specimens fail to various failure mode, including local debond,

complete debond, local crushing, etc

4.2 Surface strain response

As described above, all the specimens were instrumented by

strain gauges bonded to the surface of the mortar layer at

predeter-mined locations as shown inFig 6.Figs 18 and 19present the

rela-tion between the shear forces and surface strain of the

representative specimens which failure modes belong to the

cracked and uncracked types The solid lines represent the curve

of the strain close to the constraint end and the dash lines

repre-sent the curve of the strain close to the free end As shown in

Figs 18 and 19, during the early stage of loading, although the

stress and the strain are not even, the differences between each

curve shown inFigs 18 and 19are not too large With the increase

of the applied horizontal load, the compressive strains of S4, S5 and

S6 increase more rapidly than those of S1, S2, S3, then the uneven

distribution of the stress and strain becomes more and more

pronounced

The initial stage of the shear force versus strain relation curves

shown inFigs 18 and 19 is not parallel to the horizontal axis,

which is rather different from the bond-slip curve shown in

Fig 17 This confirms that the initial stage of the tests is the

preloading stage The shear force versus strain relation curves

can be divided into two groups: strain only increasing, strain

increasing firstly and then decreasing The shear force versus strain

relation curves of S1, S2 and S3, which locate far away from the

constraint end have no strain decreasing stage Whereas, the shear

force versus strain relation curves of S4, S5 and S6, which close to

the constraint end, have both the strain increasing and decreasing stage Moreover, the shear force versus strain relation curves of S4, S5 and S6, which specimens have crack failure type, have a

0

50

100

150

200

250

300

350

237.25

20mm 25mm

300.37 263.75

296.24 259.56

7.8%

220.11

Bond length (mm)

186.73

1.6%

1.4%

Fig 14 Shear force of bond with thickness of 20 mm and 25 mm.

0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2

2.20 2.46 2.54 2.81

1.612.00 2.01 2.12 2.27 2.37 2.52 3.05

2.50 2.56 2.71

Bond length L(mm)

2.78

(b) Relationship between the load bearing capacity and bond length with the W2-3 group and W3 group

2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.7

20 mm

25 mm

0.14MPa

Difference value < 5.6 %

Bond length, L (mm)

0.044MPa

(c) Relationship between the load bearing capacity and bond length

0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8

2.42

2.52 2.43

2.51 2.42

2.49

Bond Length L(mm)

2.48

(a) Relationship between the bond length and load bearing capacity of Group W1

Fig 15 Relationship between the load bearing capacity and bond length.