Shear response of brick masonry small assemblages strengthened with bonded FRP laminates for in-plane reinforcement

Construction and Building Materials 24 (2010) 1372–1384

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Shear response of brick masonry small assemblages strengthened with bonded FRP laminates for in-plane reinforcement P. Roca a,*, G. Araiza b a b

Department of Construction Engineering, Universitat Politècnica de Catalunya, Jordi Girona, 1-3, 08034 Barcelona, Spain Autonomous University of Aguascalientes, Avenida Universidad 940, CP 20100, Aguascalientes, Ags, Mexico

a r t i c l e

i n f o

Article history: Received 24 May 2006 Received in revised form 8 January 2010 Accepted 11 January 2010

Keywords: Brick masonry Shear behavior Shear strengthening FRP laminates Cyclic load

a b s t r a c t Modern composite materials are receiving increasing attention as reinforcing solutions applicable to the repair and strengthening of concrete and masonry structures. Aiming towards a better characterization of the possibilities offered by these materials, the research work reported here investigates the shear response of small masonry assemblages strengthened externally with sheets made of glass and aramid fiber reinforced polymer laminates. An alternative strengthening approach provided by microlaminated wood is also investigated. The assemblages, consisting of masonry couplets, were subjected to combined shear and axial loading and were laid to failure through monotonic and cyclic loading processes. The efficiency of the different strengthening materials is investigated by comparison with measures obtained for unreinforced assemblages subjected to similar load conditions. The research is aimed at characterizing the contribution of the different strengthening materials and arrangements to increase both the peak shear strength and the residual (post-peak) shear strength. An attempt is made to analytically describe the strengthening effect of reinforcement. For this purpose, two different effects provided by the reinforcing laminates – the increase of friction at the brick–mortar interface and the shear strength of the laminate itself – are considered. Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction. Shear strengthening of masonry A significant part of the building stock of many areas of the world is still constituted by historical or traditional constructions structurally consisting of brick masonry load-bearing walls. This includes both residential buildings still in use and heritage constructions of large cultural value. The evaluation and the retrofitting of the masonry shear and load-bearing walls is a key issue to ascertain the adequate safe response of these buildings under lateral actions such as hurricane winds and earthquakes. This aspect is of particular interest in the Mediterranean countries, where a massive stock of masonry buildings, erected during the 20th c. or before, coexists with very significant seismicity. Among possible strengthening strategies, external reinforcement consisting of externally chemically bonded resistant materials, such as fiber reinforced polymers (FRP), provides an interesting possibility because it can be implemented easily, only requires minor preparation works, and preserves the material integrity of the masonry wall. However, external chemically bonded reinforcement involves complex mechanical and strength phenomena, such * Corresponding author. Tel.: +34 934017381; fax: +34 4054135. E-mail address: [email protected] (P. Roca). 0950-0618/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2010.01.005

as: (1) the possible peeling off of the brick surface, (2) the brittle behavior of FRP both in shear and tension, (3) the effective resisting response of the reinforcement with respect to its theoretical capacity, (4) the influence of friction and dilatancy in the brick–mortar interface on the response of the strengthening and (5) the coupling of the different strength mechanisms activated by the reinforcing, which include an increase of friction due to the generation of normal anchoring forces and the contribution of the shear strength of the laminates. A number of experimental or analytical research works have been previously carried out on in-plane response of masonry with externally bonded reinforcement (Fig. 1). Most of these works deal with FRP applications. Triantafillou [1] derived a set of analytical expressions for the prediction of the ultimate response of masonry structures using epoxy-bonded FRP laminates and compared them with experimental results obtained by testing a set of small wall specimens in out-of plane and in-plane bending (Fig. 1d). Valluzzi et al. [2,3] chose the diagonal compressive test to investigate the in-plane shear response of brick masonry panels strengthened with FRP laminates and compared the experimental results with the predictions yielded by different analytical models (Fig. 1a). A number of investigations have been carried out on the inplane response of walls strengthened with FRP sheets or laminates.

P. Roca, G. Araiza / Construction and Building Materials 24 (2010) 1372–1384

(b)

(a)

(c)

1373

(d)

Fig. 1. Different in-plane tests configurations utilized to investigate the in-plane response of reinforced wallets and specimens: (a) Valluzzi et al. [2,3], (b) ElGawady et al. [4,5], (c) Eshani and Saadatmanesh [20], (d) Triantafillou [1].

ElGawady et al. [4,5] have investigated the response of large masonry panels strengthened with FRP laminates applied diagonally to the joints (Fig. 1b) subjected to both static and cyclic loading. Similar tests, either for monotonic or cyclic loading, have been carried out by Laursen et al. [6], Santa Maria et al. [7] using CFRP and by Fam et. al. [8], Mahmood et al. [9], Al-Salloum and Almusallam [10], Wang et al. [11] Liu et al. [12] and Stratford et al. [13], among other, using GFRP. Fam et al. [8] have specifically addressed the case of deteriorated walls repaired by means of injection and GFRP sheets; as shown by these authors, the combination of both repair measures permits to fully recover and even enhance the capacity of the walls. Marcari et al. [14] has investigated the use of FRP strengthening to improve the strength capacity of tuff masonry specimens simulating the mechanical properties of buildings in Italian historic centers. Aprile et al. [15] have investigated the influence of FRP reinforcement on the seismic reliability of ordinary masonry wall systems and have found that effective safety increment of the repaired or strengthened systems is limited due to loss of ductility. Benedetti and Steli [16] propose analytical shear-displacement curves for masonry panels, including the case of FRP reinforced ones, to be used in the evaluation of the seismic capacity of wall systems. In lack of more specific analytical approaches, the expressions provided by the Eurocode 6 [17] or by other authors [18] for masonry conventionally reinforced with embedded steel bars, have also been used to assess externally FRP reinforced masonry. More recently, CNR-DT 200/2004 [19] has also provided an analytical approach for the estimation of the ultimate capacity of masonry walls strengthened with FRP laminates. In fact, the analytical approach given by Triantafillou [1] and CNR-DT200/2004 [19] stems from the conventional analytical treatment of shear strength of reinforced concrete using the well-known truss analogy. Because of this, these expressions can only be used for strengthened masonry members developing modes of failure equivalent to those of reinforced concrete in shear. Among the possible modes of failure shown by masonry walls subjected to in-plane forces (Fig. 2),

namely overturning, sliding, brick cracking, compression and their mixed forms, only the third one (brick cracking) can be, to a certain extent, assimilated to the response of a concrete beam. The mentioned analytical expressions are only intended to assess the reinforcement of walls experiencing this third mode of failure. However, the other modes are also possible in common masonry buildings subjected to horizontal forces. As opposed to other works, the present experimental study is carried out on elementary shear masonry assemblages rather than in entire masonry components such as large walls or panels. This is so because the study is mainly oriented to the identification of the elementary mechanisms involved in the strength response of reinforced masonry, while other studies focus on the evaluation of the overall efficiency of the strengthening. Another difference with respect to some previous research works lays on the consideration of strengthening strips applied perpendicular to the mortar joints. This type of arrangement may be necessary to provide strengthening not only in the brick cracking range but also in the sliding mode of failure (Fig. 2). A similar approach, also using simple shear assemblages strengthened by means of overlay reinforcement placed through the mortar joints, has been previously considered by Eshani and Saadatmanesh [20] (Fig. 1c). Haroun et al. [21] have also carried out shear tests on small wall strengthened transversely to the mortar joints. Experimentation and numerical modeling at the elementary level is at present receiving increasing interest. Aiello and Sciolti [22] have proposed a test procedure for analysis of bond performance between FRP sheets and natural stones allowing evaluation of bond stress and slip values. The bond behavior of FRP reinforcement on clay bricks has been investigated by Liu et al. [23] and Willis et al. [24], among other. Numerical approaches to model the masonryFRP interface behavior have been recently proposed by Maruccio et al. [25] and Grande et al. [26]. Certain problems or limitations of external reinforcement are not to be ignored and deserve further research. Among the possible problems is fire-resistance, especially when epoxi-based materials

1374

P. Roca, G. Araiza / Construction and Building Materials 24 (2010) 1372–1384

(a)

(b)

(c)

(d)

Fig. 2. Failure modes of masonry walls subjected to in-plane shear forces: (a) overturning (rocking), (b) joint sliding, (c) brick cracking and (d) brick cracking with crushing in compression.

are used as part of the strengthening or the bonding material. Another problem may be found in causing undesired water-proofing effect incompatible with the natural perspiration of the ceramic material. Possible problems related to the durability of the FRP strengthened masonry should not be ignored. Karbhari [27] provides a detailed analysis on the durability of FRP materials; according to this author, a satisfactory durability of FRPs results when appropriate constituent materials and manufacture processes are used in combination with adequate environmental protection. A more specific research on the durability of GRPF composite exposed to different environmental conditions can be found in [28]. It should be noted that some solutions based on FRP may be inadequate for heritage or historical constructions because of lack of compliance with conservation principles resulting from excessive invasivity, non-removability or obtrusiveness. External strengthening composed of more traditional materials, such as wood or ceramics, applied to the wall surface with adequate gluing materials or anchoring mechanical devices, might be more advantageous with regard to these aspects. In the experimental work here presented, two different strengthening composite materials, namely aramid fiber reinforced polymer (AFRP) and glass fiber reinforced polymer (GFRP), were

investigated. A possible strengthening technique based on externally glued plywood reinforcement is also considered.

2. Test arrangement and execution Experiments on small masonry assemblages subjected to combined shear and axial loading have been previously carried out using different test configurations Hofmann and Stockl [29], Atkinson et al. [30], van der Pluijm [31], Riddington et al. [32], Lourenço and Ramos [33] (Fig. 3). For the present research, involving tests on over 140 specimens, it was necessary to envisage an efficient test configuration in order to keep the cost and time required within reasonable limits. Based on the previous experience, a specific device was designed allowing a simpler preparation and execution of tests. The device consists of a combination of two stiff boxes which can move relative to each other thus permitting the application of forces along two axes (Fig. 4). This device can be used to carry out a couplet test similar to the one used in [29], with the possibility of carefully controlling the axes of the forces applied on the specimen. Other possible arrangements, such as the triplet test (Fig. 3d) adopted by CEN/TC125 [34] or van der Pluijm’s [31]

1375

P. Roca, G. Araiza / Construction and Building Materials 24 (2010) 1372–1384

(b)

(a)

(c)

(d)

(e)

teflon

neoprene shims

Fig. 3. Different test arrangements for masonry specimens subjected to combined shear and axial loading: (a) couplet test, Hofmann and Stockl [29], (b) Atkinson et al. [30], (c) Van der Pluijm [31], (d) triplet test CEN/TC125 [34], (e) Lourenço and Ramos [33].

a f

c

e

b d

g

h

Fig. 4. Left: device used for the execution of test: (a) and (b) actuators, (c) masonry specimen (couplet or triplet), (d), (e) and (f) stiff elements, (g) free sliding surfaces, (f) fixed connection. Right: picture of the equipment.

(Fig. 3c) have been disregarded because, in spite of possible advantage in terms of experimental consistency, they demand more delicate preparation and time-consuming execution and would hardly permit the efficient testing of the large amount of specimens required by the present investigation. Dynamic actuators were utilized to carry out displacement-controlled tests. The designed device was installed in a 1000 kN Instron press machine combined with a MTS 250 kN servocontrolled actuator. Normal load was gradually introduced on the device through the horizontal actuator. Keeping the normal load constant, either monotonic or cyclic shear loading, producing a gradual increase of the vertical relative deformation between the bricks, was provided until reaching the peak response and beyond, thus allowing the characterization of the post-peak branch.

3. Description of tested assemblages Using the test arrangement described, simple assemblages consisting of two solid clay bricks measuring 280  140  50 mm with

a mortar joint 10 mm thick, were tested. The chosen solid bricks, hand-made, are currently produced to build exterior façades due to their aesthetical quality. They are similar to the solid bricks that were utilized for the construction of structural walls during the 19th and most part of 20th c. in many countries. The specimens were externally strengthened using the different materials mentioned (AFRP, GFRP and plywood, in different arrangements). The properties of the different materials used are referred in Table 1. Complementary tests were carried out to characterize the main properties of the masonry components. Average brick compression strength of 24.9 MPa was determined by carrying out a uniaxial compression test on 6 units according to EN772-6: 2001 [35]. Similarly, flexural tests carried out on six specimens yielded an average flexural strength ftb of 7.1 MPa. Using the approach given by ENV 1992-1-1: 1991 [36] for concrete, the uniaxial tensile strength of brick (ft) is estimated as 0.5ftb, resulting in 3.5 MPa. Standard tests carried out according to EN1015–11:1999 [37] on six mortar specimens measuring 40  40  160 mm, after a period of 28 days of curing, revealed an average flexural strength equal to 0.8 MPa

1376

P. Roca, G. Araiza / Construction and Building Materials 24 (2010) 1372–1384

Table 1 Properties of FRP laminates.

a b

Material

GFRP

AFRP

Fiber Material Meshing Density (g/cm3) Equivalent thickness (mm) Tensile strength (MPa) Young modulus (MPa) Ultimate strain (%)

Glass Bidirectional 2.68 0.085a,b 3000 65,000 4.5%

Aramid Unidirectional 1.45 0.2a 2900 120,000 2.5%

Matrix Binding material Approximate thickness (mm) Tensile strength (MPa) Young modulus (MPa) Ultimate strain in tension Bonding stress to concrete (MPa)

Epoxi resin 1.6 54 3034 2.5% >2.5

Epoxi resin 0.8 54 3034 2.5% >2.5

Gluing material to masonry Binding material Approximate thickness (mm) Tensile strength (MPa) Young modulus (MPa) Ultimate strain in tension Bonding stress to concrete (MPa)

Epoxi resin 0.3 12 717 3% >2.5

Equivalent thickness of fiber layers. In each direction.

(and thus an estimated tensile strength of 0.65 MPa) and a compressive strength of 7.3 MPa. The AFRP and GFRP laminates used are real commercial products and were applied on the specimens following the manufacturer’s instructions. The AFRP laminates are unidirectional, while the GFRP are bidirectional. Both were glued to the brick surface by means of an epoxi resin. Information on the more significant mechanical properties of the used FRP laminates, provided by the manufacturers (i.e. extracted from producer data sheet), is summarized in Table 1. As mentioned, strengthening consisting of externally glued plywood panels were also investigated. The plywood panels, with thickness of 24 mm, were made of glued microlaminated wood composed of multiple thin layers with fibers oriented in two normal directions. The plywood panels had a tensile and shear

strength of 27 and 5.5 MPa respectively and a Young’s modulus of 10.000 MPa in the direction parallel to the fibers. They were bonded to the masonry specimens by means of a polyurethane adhesive composite normally used for the application of wood or ceramic tiles on steel or polyester surfaces. The application of GFRP involved the following operations: (1) cleaning of the specimen’s surfaces by compressed air; (2) cutting of the glass fiber sheets; (3) application of an epoxy primer; (4) application of a smoothing thixotropic mortar layer. The smoothing mortar layer had an average thickness of 2 mm and compression strength of 50 MPa at 24 h; (5) application of a first layer of saturating resin, on which the GFRP sheet was placed and covered with a second layer of saturating resin. The application of AFRP only differed in the fact that no smoothing mortar was used. The plywood panels were applied using a similar process including the (1) cleaning of the specimen’s surfaces, (2) cutting the wood panels, (3) application of a layer of polyurethane adhesive about 4 mm thick to both the brick and wood surfaces, and (4) placing the wood panel on the brick surface. A total number of 141 specimens, corresponding to 11 different groups, were tested (Tables 2 and 3). The different groups are characterized by (1) the type of material used as strengthening, (2) the orientation of the fibers in the case of AFPR and GFRP laminates, (3) the number of sides with reinforcement applied (one or two), and (4) the type of loading process (monotonic or cyclic). In turn, each group included various series subjected to different normal stresses. Each series includes different specimens subjected to the same normal stress. In Tables 2 and 3, the different series are designated according to the type of reinforcement (U for unreinforced, G for GFRP, A for AFRP, W for wood), number of sides strengthened (D for double side), the orientation of the fibers with respect to the mortar joint (‘‘45” if 45°), the type of loading condition (M for monotonic and C for cyclic), and the applied axial force (20, 40, 60 or 80 kN). Four different amounts of axial load were applied, corresponding to values of 20, 40, 60 and 80 kN, which in turn produced average normal stresses of 0.51, 1.02, 1.53 and 2.04 MPa. These values cover the range of expectable work compression stresses experienced by masonry walls in real buildings. In fact, real work stress values normally fall in the lower interval comprised between 0.5 and 1.0 MPa.

Table 2 Experimental results of monotonic tests.

a

Group

Series (number of specimens)

Strengtheninga

Normal stress (MPa)

Average ultimate shear stress (MPa) (S. deviation)

Average residual shear stress (MPa) (S. deviation)

Failure mode

UM

UM20 UM40 UM60 UM80

(No strengthening)

0.51 1.02 1.53 2.04

0.84 1.50 1.94 1.86

0.42 0.79 1.19 1.50

Frictional

GM

GM20 (5) GM40 (5)

Single-side GFRP at 90°

0.51 1.02

1.33 (0.19) 1.64 (1.40)

0.53 (0.11) 0.75 (0.13)

Peeling / Shear

GMD

GMD20 GMD40 GMD50 GMD60

Double-side GFRP at 90°

0.51 1.02 1.53 2.04

1.92 2.12 2.73 2.80

(0.10) (0.12) (0.16) (0.27)

0.50 1.13 1.93 1.76

(0.09) (0.10) (0.17) (0.05)

Peeling / Shear

AM

AM20 AM40 AM60 AM80

Single-side AFRP at 90°

0.51 1.02 1.53 2.04

1.30 1.66 2.29 2.30

(0.15) (0.20) (0.22) (0.44)

0.50 1.03 1.23 1.99

(0.23) (0.14) (0.02) (0.25)

Peeling

WM

WM20 (3) WM40 (3) WM80 (3)

Single-side Wood at 45°

0.51 1.02 2.04

1.03 (0.09) 1.46 (0.48) 2.40 (0.02)

(5) (5) (5) (5)

(5) (5) (5) (5)

(3) (3) (3) (3)

(0.18) (0.27) (0.31) (0.39)

Strengthening material, number of sides reinforced and orientation of fibers with respect to mortar joint.

(0.02) (0.06) (0.17) (0.06)

0.50 (0.06) 0.97 (0.10) 1.58 (0.17)

Debonding

1377

P. Roca, G. Araiza / Construction and Building Materials 24 (2010) 1372–1384 Table 3 Experimental results of cyclic tests. Series (number of specimens)

Strengtheninga

Normal stress (MPa)

Average ultimate shear stress (MPa) (S. deviation)

Average residual shear stress (MPa) (S. deviation)

Failure mode

UC

UC20 UC40 UC60 UC80

(No strengthening)

0.51 1.02 1.53 2.04

1.05 1.34 1.77 2.37

0.35 0.90 1.27 1.36

Frictional

GCD

GCD40 (3) GCD60 (3) GCD80 (3)

Double-side GFRP at 90°

1.02 1.53 2.04

2.24 (0.14) 2.97 (0.45) 2.15 (0.14)

1.12 (0.23) 1.73 (0.13) 1.47 (0.35)

Peeling off/ Shear

GC45

GC4520 GC4540 GC4560 GC4580

Single-side GFRP at 45°

0.51 1.02 1.53 2.04

1.57 1.70 2.01 2.50

(0.16) (0.24) (0.21) (0.55)

0.85 1.00 1.35 1.40

(0.10) (0.20) (0.10) (0.26)

Peeling off / Shear

AC

AC20 AC40 AC60 AC80

Single-side AFRP at 90°

0.51 1.02 1.53 2.04

1.35 1.95 1.97 3.04

(0.28) (0.26) (0.43) (0.47)

0.48 0.97 1.30 1.76

(0.08) (0.11) (0.08) (0.30)

Peeling off

ACD

ACD20 ACD40 ACD60 ACD80

Double-side Double layer AFRP at 90°

0.51 1.02 1.53 2.04

2.70 2.67 3.04 3.21

(0.31) (0.35) (0.42) (0.60)

0.74 1.03 1.34 1.75

(0.20) (0.21) (0.02) (0.11)

Peeling off

WC

WC20 (3) WC40 (3) WC60 (3)

Single-side wood at 45°

0.51 1.02 1.53

1.52 (0.25) 2.21 (0.10) 2.16 (0.21)

0.76 (0.05) 1.44 (0.06) 1.58 (0.06)

Debonding

(5) (5) (5) (5)

(5) (5) (5) (5)

(3) (3) (3) (3) (3) (3) (3) (3)

(0.35) (0.24) (0.21) (0.25)

(0.05) (0.03) (0.03) (0.10)

Strengthening material, number of sides reinforced and orientation of fibers with respect to mortar joint.

4. Experimental results 4.1. Unreinforced assemblages As shown in Tables 2 and 3, the groups labeled UM and UC consisted of four series composed of unreinforced specimens. These specimens were used to determine the frictional properties of the mortar-unit interface subjected either to monotonic or cyclic loading. Fig. 5 shows experimental diagrams relating the average shear stress with the relative vertical displacement experienced by the bricks across the mortar joint. The diagrams obtained in the monotonic tests typically exhibit a very pronounced peak followed by a brusque post-peak decreasing branch and then an almost horizontal branch corresponding to a residual shear strength (Fig. 5a). As expected, both the peak and the residual shear strengths increase

Shear force (kN)

(a)

with the applied average normal stress. A very similar trend is observed in the cyclic tests (Fig. 5b). In Fig. 5, N is the axial load applied on the specimen. The deformations associated to the ultimate strength (or peak values) show a significant scattering and even some randomness. In spite of this, a certain dependency can be recognized between the shear deformation at peak (c = d/e, Fig. 6) and the applied normal stress. The monotonic tests produce a more uniform set of values in the vicinity of c = 0.25. In the cyclic tests, the deformation at peak varies with the normal stress in a more pronounced way; the maximum deformation at peak, around 0.40, is reached for the smaller and larger values of the normal stress, while the minimum one (around 0.15) is obtained for the intermediate values. Using these results, diagrams comparing the peak shear strength with the normal stress (in fact, a set of failure envelope curves) can be elaborated for both the monotonic and the cyclic

(b)

100

100

90

90

80

80

70

2.04 MPa

60 50 1.53 MPa

40 30

1.02 MPa 0.51 MPa

20

Shear force (kN)

a

Group

σ=0.51 MPa σ=2.04 MPa

70 60 50 40 30 20 10

10

0

0 0

2

4

Displacement (mm)

6

0

2

4

6

8

10

Displacement (mm)

Fig. 5. Experimental shear force (V) vs. displacement (d) diagrams of unreinforced assemblages for different compression stresses, subjected to: (a) monotonic and (b) cyclic loading.

1378

P. Roca, G. Araiza / Construction and Building Materials 24 (2010) 1372–1384

0.45 0.4

γ=δ/e

δ

τ

Shear strain

0.35

e

0.3 0.25 0.2 0.15 0.1 0.05

Monotonic

Cyclic

Trend (Monotonic)

Trend (cyclic)

0 0

0.5

1

1.5

2

2.5

Normal stress (MPa) Fig. 6. Shear deformation (d) at peak strength vs. applied normal compressive stress for unreinforced assemblages.

For normal stresses above a certain threshold comprised between 1.5 and 2 MPa, a more complex and scattered response is obtained due to the influence of compression effects causing local yielding or crushing in compression. The values of the cohesion (c) and the tangent of the friction angle (tan /) can be estimated by carrying out linear regression on the experimental values. A cohesion of 0.32 MPa and a friction angle of 47.2° (tan / = 1.08), with a regression coefficient R2 of 0.78, result when all the values are considered. Ignoring more anomalous values obtained for the shear strength peaks (points market with symbol ‘‘  ” in Fig. 7a) leads to a similar estimation of the cohesion and friction angle; the resulting values are c = 0.39 MPa and / = 44.7° (tan / = 0.99), with an improved regression coefficient R2 of 0.96. A lower envelope of the experimental values is given by a friction angle of 40.4 (tan /min = 0.85). The values corresponding to the residual response show a much lesser scattering. In that case, null cohesion is assumed, so that the frictional envelope reads

(a)

3

Shear stress (MPa)

s ¼ r tan /r

2.5

where /r is the residual friction angle. Using all the available experimental results /r is estimated as 37.2° (tan /r = 0.76) with a regression coefficient R2 of 0.95. This result is in agreement with van der Pluijm’s [31] experiments showing that a value of 0.75 after peak load is typically obtained independently of the materials used. The cyclic tests show a similar trend. Considering all the experimental peak values yields c = 0.54 MPa and tan / = 0.86 for R2 = 0.79. When the more anomalous values are disregarded, a cohesion of 0.40 MPa and friction angle of (tan / = 0.91), with a regression coefficient of 0.94, are obtained. A residual friction angle of 40° (tan / = 0.84) is estimated with R2 = 0.98. Like in the static case, a friction angle of 40.4° (tan /min = 0.85) provides an adequate lower envelope. As can be observed, very similar values of the initial friction angle are obtained for both the monotonic and cyclic tests. Conversely, in the case of the residual friction angle, a larger value results for the cyclic tests. This has been already noted by Lourenço and Ramos [33] for tests on dry joint specimens and is attributed to the increase of surface roughness under cyclic load. The shape of the envelope of the cyclic curves is qualitatively similar to that of the monotonic curves. In spite of this, and due to the quantitative differences mentioned (among which the different residual angle and the different shear deformation at peak) the monotonic curves do not accurately model a possible cyclic envelope.

Peak Peak (disregarded) Residual lower bound Regression (peak) Regresion (residual)

2

τ = 0,99 σ + 0,39

1.5

R2 = 0,96

1

τ = 0,76σ R2 = 0,95

0.5 0 0

0.5

1

1.5

2

3

Shear stress (MPa)

Normal stress (MPa)

(b)

2.5

Peak Peak Residual minimum Regression (residual) Regression (peak)

2

τ = 0,91 σ + 0,40 R2 = 0,94

1.5 1

τ = 0,84σ R2 = 0,98

0.5

ð2Þ

4.2. Assemblages strengthened with FRP laminates

0 0

0.5

1

1.5

2

Normal stress (MPa) Fig. 7. Experimental failure criteria: shear strength of bed joints as a function of the applied normal compressive stress for: (a) monotonic and (b) cyclic loaded assemblages.

tests. Similar diagrams can be elaborated to relate the residual shear strength with the applied normal stress (Fig. 7). The relationship between normal and peak or residual shear stresses shows a significant scattering. This scattering is mainly attributed to the irregularity (and variable roughness) of the surfaces of the hand-made clay bricks. In spite of this, the response of the assemblages can be described acceptably by means of a linear relationship corresponding to a Mohr–Coulomb frictional law,

s ¼ c þ r tan /

ð1Þ

where s is the shear strength, r is the applied normal stress, and c, / are the cohesion and friction angle at the brick–mortar interface.

The assemblages strengthened with FRP laminates produce, in all cases, qualitatively similar stress–strain diagrams. They are characterized, as in the case of the unreinforced ones, by a gradual, almost linear increase of stress up to the peak value, followed by a loss of strength to a residual value which keeps almost constant until the maximum deformation is reached. Compared with the unreinforced specimens, a more gradual transition between the peak and residual ranges is generally obtained. This trend is observed in both the monotonic and the cyclic tests (Fig. 8). An exhaustive description of the experimental results is given in [38]. The strengthening produces an increase of the peak shear strength ranging between 0% and 160% depending on the type of reinforcement (Fig. 9) and the applied compression stresses. As expected, the peak shear strength increases with the tensile strength and the total section of fibers. The highest strengthening effect, ranging between 20% and 160%, is observed for moderate compression stresses (up to 1 MPa). The maximum (160%) is obtained for the series provided with a larger section of reinforcing fibers

1379

P. Roca, G. Araiza / Construction and Building Materials 24 (2010) 1372–1384

(a)

140

(b)

σ=0.51 MPa 120

120

σ=2.04 MPa

100

Shear force (kN)

100

Shear force (kN)

140

2.04 MPa

80

60 1.53 MPa

80

60

40

40 1.02 MPa

20

20 0.51 MPa

0

0 0

2

4

0

6

5

(c)

160 σ=2.04 MPa 140

(d)

90 σ=2.04 MPa 80

σ=0.51 MPa

σ=0.51 MPa

70

Shear force (kN)

120

Shear force (kN)

10

Displacement (mm)

Displacement (mm)

100 80 60 40

60 50 40 30 20

20

10 0

0 0

5

0

10

2

4

6

Displacement (mm)

Displacement (mm)

Fig. 8. Experimental shear force vs. displacement diagrams of strengthened assemblages belonging to groups: (a) GMD, (b) GCD, (c) ACD and (d) WC, subjected to different compression stresses.

sophisticate testing technique, such as the one proposed by van der Pluijm [31]. As can be seen in Fig. 12, the specimens strengthened with GFRP or AFRP experience, on average, larger deformations at peak shear

Shear strength gain (%)

(ACD). For larger compressive stresses, the strengthening effect becomes significantly less effective and decreases as the compressive stress increases. The effect of the reinforcement on the residual shear strength is less remarkable and even negligible in some cases. The mode of failure normally observed in the case of FRP reinforcements has been peeling off (delamination) experienced by the brick surface (Fig. 10c and d). In some specimens reinforced with GRPF, rupture of the fibers in shear, in combination with possible brick surface peeling off, has also been observed (Fig. 10a and b). In the case of the wood reinforcement, the failure has normally been caused by the debonding of the microlaminated panel from the brick surface. The diagrams plotting the relationship between the shear stress peak values and the normal compression stress show again significant scattering (Fig. 11). In spite of this, a roughly linear dependency between the applied normal stresses and shear peak strength can be observed. As in the unreinforced specimens, a reduction of the shear peak strength is observed in some cases for the maximum applied compression stresses. This effect seems connected to a certain uneven distribution of compression stresses in the bricks and could probably be attenuated by using a more

Group of specimens Fig. 9. Average increase of peak shear strength for different strengthening arrangements under monotonic and cyclic loading.

1380

P. Roca, G. Araiza / Construction and Building Materials 24 (2010) 1372–1384

Fig. 10. Condition of GFRP specimens after failure. Rupture of fibers in shear for fibers oriented at 90°: (a) and 45° (b) with respect to the mortar joint. Peeling off of brick surface (c,d).

strength than the unreinforced ones (d ranging between 0.2 and 0.5). 5. Analysis of results An attempt is made to quantitatively model the strength envelope curves of the strengthened assemblages. Subjected to the combination of shear and normal forces, the strengthened specimen can develop two different resisting contributions in failure (Fig. 13). Firstly, the specimen will develop a frictional mechanism similar to the one described in Section 4.1. The contribution of this mechanism to the shear strength is estimated as

s1 ¼ ðr þ rr Þ tan /

tar interfaces, causing, in turn, the destruction of the existing cohesion. Because of this, no cohesive term is included in Eq. (3). The increment of compression produced by the strengthening can be calculated as

rr ¼

Pr Ac

ð4Þ

where Ac is the area of the face of the brick and Pr is the maximum tensile force that can be developed by the reinforcement. Pr is obtained as

 Pr ¼ min

Ptr ¼ qAtr ftr Pu

ð5Þ

ð3Þ

where r is the external compression applied on the specimen in the direction normal to the joint and rr is the increment of compression produced by the strengthening (Fig. 13a). The activation of the strengthening (and the appearance of a resisting term rr) is granted by the dilatancy associated to friction and sliding across the joint. This mechanism will require a significant opening of the brick–mor-

where Ptr is the maximum tensile force which can be resisted by the strengthening material and Pu is the maximum tensile force which can be resisted by the anchorage of the reinforcement to the brick surface. In Eq. (5), Atr and ftr are the section and tensile strength of the fibers and q is an efficiency factor. In the experiments carried out, the term Pu has normally been determined by the peeling off (delamination) of the ceramic brick surface.

1381

P. Roca, G. Araiza / Construction and Building Materials 24 (2010) 1372–1384

4

3

GM

GMD, GCD

3.5

Shear stress (MPa)

Shear stress (MPa)

2.5 2 1.5 1 0.5

3 2.5 2 1.5 1 0.5 0

0 0

0.5

1

1.5

2

0

2.5

0.5

Normal compression stress (MPa)

GC45

2

2.5

AM, AC

3.5

Shear stress (MPa)

3

Shear stress (MPa)

1.5

4

3.5

2.5 2 1.5 1 0.5

3 2.5 2 1.5 1 0.5 0

0 0

0.5

1

1.5

2

0

2.5

0.5

1

1.5

2

2.5

Normal compression stress (MPa)

Normal compression stress (MPa) 4

4.5 4

ACD

WM, WC

3.5

Shear stress (MPa)

3.5

Shear stress (MPa)

1

Normal compression stress (MPa)

3 2.5 2 1.5 1

3 2.5 2 1.5 1 0.5

0.5

0

0 0

0.5

1

1.5

2

2.5

0

Normal compression stress (MPa)

0.5

1

1.5

2

2.5

Normal compression stress (MPa)

Peak monotonic

Peak cyclic

Residual cyclic

Friction

( 1)

Residual monotonic Friction + shear

(

1+ 2)

Fig. 11. Shear peak and residual strength of strengthened assemblages as a function of applied normal compressive stress for reinforced assemblages.

Secondly, the strengthening material can contribute with its own shear strength (Fig. 13 and b). In fact, even unidirectional FRP composites develop certain, even if small, shear strength when subjected to shear forces perpendicular to the fibers. This second contribution to shear strength can be estimated as

s2 ¼

Vu Ac

ð6Þ

where Vu is the maximum shear force developed by the reinforcement. This term can be estimated as

V u ¼ qsh fsh Atr

ð7Þ

where qsh is a second efficiency factor and fsh is the shear strength of the reinforcing material in the direction parallel to the mortar joint (perpendicular to the FRP strip axis). In principle, the two contributions, s1 and s2 cannot be simply added because they may occur at different deformation. As illustrated by Figs. 5 and 8, the phenomena involved (fiber shear strength and peak frictional strength) are not plastic and cannot be simply superposed. Nevertheless, a lower boundary of the

1382

P. Roca, G. Araiza / Construction and Building Materials 24 (2010) 1372–1384

0.6

GM

0.5

Shear strain

GMD GCD

0.4

GC45 AM

0.3

AC ACD

0.2

Trend (UM) 0.1

Trend (UC)

0 0

0.5

1

1.5

2

2.5

Normal stress (MPa) Fig. 12. Shear deformation at peak strength vs. applied normal compressive stress in assemblages strengthened with FRP.

(a)

σr

(b)

σ

Fig. 13. Strength contributions in reinforced assemblages: (a) friction and (b) shear strength of reinforcement.

resulting shear strength su can be obtained by only considering the first and main contribution, s1. In turn, an upper boundary can be estimated by simply adding both terms (su = s1 + s2). Consistently with the aim of obtaining a lower bound, the value s1 has been calculated by considering /min instead of / (see Section 4.1). In fact, these two boundaries seem to actually confine the experimental values (Fig. 11). The first approach (su = s1) seems more adequate for the intermediate values of the applied normal stress. The second approach (su = s1 + s2) seems more accurate for the extreme values considered for the applied normal stress (r = 0.5 or r = 2 MPa). A possible explanation lies in the fact that for those extreme values the shear strain associated with the frictional peak strength becomes larger and more similar to the deformation associated with the shear strength of the laminate; this larger proximity between both deformations might enable a certain superposition of both contributions. In turn, this provides an understanding of the scattering obtained as a result of the fluctuations in the distance between the two peak deformations. A detailed prediction of Vu would require a previous measurement of the actual shear strength of the fibers or material compos-

ing of the reinforcement. In this study, the term s2 has been adjusted from the experimental results. The average efficient shear strength estimated for the different reinforcements considered is given in Table 4. The term Vu is evaluated by considering the strength fsh on the total section of fibers. No well established criterion is known for the evaluation of the maximum force resisted by the anchorage of the strengthening to ceramic brick surface. As a tentative approach, expressions derived for concrete may be used. It must be remarked, however, that no previous and direct measurement of their real applicability of such expressions to ceramic brick surfaces has been undertaken. A summary of available approaches for concrete can be found in Chen and Teng [39] and Lu et al. [40]. A very simple approach is given by van Gemert’s [41] expression,

Pu ¼ 0:5bp Lfctm

ð8Þ

where fctm is the concrete surface tensile strength, bp is the width of the reinforcement and L is the bond length. Chen and Teng’s [39] proposal has the advantage of considering the influence of the stiffness of the strengthening strips on the ultimate bond strength. According to this approach,

Pu ¼ 0:427bp bL

qffiffiffiffi fc0 bp Le

ð9Þ

where

( rffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffi Ep tp 2bp =bc Le ¼ pffiffiffi0 bp ¼ 1þbp =bc bL ¼ fc

1

if L P Le

pL sin 2L e

if L < Le

ð10Þ

being Ep and tp the Young’s modulus and the thickness of the bonded reinforcement, respectively, and f0 c the concrete cylinder compressive strength. For the present application, Eptp is taken as Eftf + Ertr, where tf and tr are the equivalent thickness of the reinforcing fibers and the thickness of the polymer laminate, and Ef and Er are the corre-

Table 4 Calculation of the force carried by the strengthening material. The meaning of the variables in the table is defined in Section 5.

a b c

Strengthening material

Type and orientation

Eptpa (MN/mm)

Le(mm)

bL

Pu(kN)

qsh fsh(MPa)

Vu(kN)

GFRP (1 layer) GFPR (1 layer) AFRP (1 layer) Microlaminated wood

Bidirectional, 90° Bidirectional, 45° Unidirectional, 90° Bidirectional, 45°

10,354 10,354 26,427 240,000

46 46 73 219

1.00 1.00 0.88 0.35

19.2 19.2 27.1 324.9

670b 1300b 420b 5.5c

15.9 30.9 23.5 37.0

For the composite section including fibers and binding resin. Estimated from experimental tests. For qsh = 1 and fsh as declared by manufacturer.

P. Roca, G. Araiza / Construction and Building Materials 24 (2010) 1372–1384

sponding Young’s moduli. The strength f0 c is assimilated to the measured brick compression strength. Units of megapascal and millimeters shall be used in Eq. (8)– (10). The calculation of the maximum forces carried by the strengthening is summarized in Table 4. In this calculation, the bond length is taken as the width of the brick, namely 50 mm.

6. Parameters influencing on the response As shown by diagrams of Figs. 9 and 10, the attained strengthening effect, measured as the increase of shear capacity of the specimens, varies significantly with the characteristics of the reinforcement utilized. Among the different parameter analyzed, the ones showing larger influence on the shear capacity of are, as expected, the equivalent thickness of the reinforcement and the shear strength of the reinforcing material. The influence of the equivalent thickness is clearly seen in the fact that the highest gains in strength are obtained for the specimens including more than one layer of reinforcement, as in the case of double side reinforcement (ACD, GMD). For a given reinforcing material, the strengthening effect increases roughly in proportion with the equivalent thickness of the reinforcement. AFRP, characterized by larger tensile and shear strength, provided the larger strengthening effect, reaching up to 160% of gained strength for double side reinforcement. In turn, GFRP provided a gain up to 130% for double side reinforcement. Compared to AFRP and GFRP, the wood reinforcement has shown a limited strengthening capacity due to a weaker adherence developed at the interface between the wood microlaminate and the brick specimen. The maximum gain obtained has been of 65% for cyclic loading. It must be noted that the percentages of gain mentioned refer to the case of specimens subjected to moderate compression strength (about 0.5 MPa). As aforementioned, for larger compression stresses in the range between 0.5 and 1.0 MPa, the strengthening effect reduces significantly. As can be seen in Fig. 9, for compression strength of 1.0 MPa the gain on shear capacity is reduced to more than one half of the mentioned percentages. Beyond this limit, the gain on shear capacity becomes less sensitive to the compression level or even does not depend with it. Generally, cyclic tests have yielded higher capacity gains that monotonic ones. In some cases the difference in gain between both tests has been very significant. In the case of wood reinforcement, cyclic tests have provided gains of about twice those obtained in the monotonic ones. The potential influence of the equivalent thickness (through the available section of reinforcement) and shear strength of the reinforcing material is clear from Eq. (7). The tensile strength of the reinforcing material has shown little direct influence on the experimental results because the failures have been normally caused by the peeling off of the bricks.

7. Conclusions The experiments carried out on simple masonry assemblages allowed the characterization of the strengthening effect provided by externally bonded reinforcing laminates, including GFRP, AFRP and microlaminated wood, under monotonic and cyclic loading. It has been observed that reinforcing laminates applied perpendicular to the mortar joints can be effective in upgrading the response of the masonry in the sliding mode of failure. The upgrading capacity increases to a certain extent with both the strength and the stiffness of the laminates. Some of the experiments carried out yielded a total increase of shear capacity of 160% with respect to the unreinforced case. It should be noted, however, that reinforce-

1383

ment applied parallel to the mortar joints can be more effective in the brick cracking failure. This upgrading effect of the reinforcement is more effective for moderate applied compression stresses (around 0.5 to 1.0 MPa), for which the maximum shear capacity has reached increments between 20% and 160% depending on the type of strengthening. For higher compression stresses, the load increments decrease gradually. The strengthening produced only a slight or even null improvement of the residual (post-peak) shear strength. The mode of failure normally observed in the case of FRP reinforcements has been peeling off (or delamination) experienced by the brick surface. Rupture of the fibers in shear has also been observed in some cases. The influence of tension capacity (in terms of tensile strength and total section of reinforcement) in the strengthening effect is particularly noticeable for moderate stresses (around 0.5 to 1.0 or even 1.5 MPa). An attempt has been made to mathematically model the effect of the external reinforcement. For that purpose, two different strength contributions were considered, namely the unit-mortar friction mechanism, enhanced by the strengthening action, and the shear strength of the laminate. The maximum responses provided by the two mechanisms cannot be simply superposed because they occur for different deformations. In order to obtain safe estimations of the ultimate shear capacity of the reinforced members, only the contribution of the frictional mechanism should be taken into account. However, the contribution of the shear strength of the laminate is perceptible and even significant (depending on the amount of reinforcement) for moderate values of applied normal compression stress (up to 0.5–1.0 MPa). Acknowledgements The studies presented here were developed within the Research Projects ARQ2002-04659 and BIA2006-04127, funded by DGE of the Spanish Ministry of Education and Science, whose assistance is gratefully acknowledged. References [1] Triantafillou TC. Strengthening of masonry laminates using epoxi-bonded FRP laminates. J Compos Constr 1998;2(2):96–104. [2] Valluzzi MR, Tinazzi D, Modena C. Shear behavior of masonry panels strengthened with FRP laminates. Constr Build Mater 2002;16:409–16. [3] Valluzzi MR, Modena C, Marchetti M. Shear strengthening of masonry panels using epoxy-bonded FRP. In: Proceedings of the 12IB2 MaC conference, Madrid, 2002. p. 1297–308. [4] ElGawady M, Lestuzzi P, Badoux M. In-plane response of URM walls upgraded with FRP. J Compos Constr 2005;9(6):524–35. [5] ElGawady MA, Lestuzzi P, Badoux M. Static cyclic response of masonry walls retrofitted with fiber-reinforced polymers. J Compos Constr 2007;11(1):50–61. [6] Laursen PT, Seible F, Hegemier GA, Innamorato D. Seismic retrofit and repair of masonry walls with carbon overlays. In: Taerwe L, editor. Non-metallic (FRP) reinforcement for concrete structures, RILEM, 1995. P. 616–27. [7] Santa Maria H, Alcaino P, Luders C. Experimental response of masonry walls externally reinforced with carbon fiber fabrics. In: Proceedings of the 8th US national conference on earthquake engineering; 2006. Paper no. 1402. [8] Fam A, Musiker D, Kowalsky M, Rizkalla S. In-plane testing of damaged masonry wall repaired with FRP. Adv Compos Lett 2008;11(6):275–81. [9] Mahmood H, Russell AP, Ingham JM. Laboratory testing of unreinforced masonry walls retrofitted with glass FRP sheets. In: 14th International brick and block masonry conference (14IBMAC), Sidney, 2008. [10] Al-Salloum YA, Almusallam TH. Walls strengthened with epoxy-bonded GFRP sheets. J Compos Mater 2005(39):1719–45. [11] Wang Q, Chai Z, Huang Y, Zhang Y. Seismic shear capacity of brick masonry wall reinforced by GFRP. Asian J Civ Eng (Building and Housing) 2006;7(6):563–80. [12] Liu J, Ming L, Song Y. Experimental investigation on flexural performance of masonry walls reinforced with GFRP. Journal of Wuhan University of Technology – Mater Sci Ed 2007;22(1):82–4. [13] Stratford T, Pascale G, Manfroni O, Bonfiglioli B. Shear strengthening masonry panels with sheet glass–fiber reinforced polymer. J Compos Constr ASCE 2004;8(5):434–43. [14] Marcari G, Manfredi G, Prota A, Pecce M. In-plane shear performance of masonry panels strengthened with FRP. Composites: Part B 2007;38:887–901.

1384

P. Roca, G. Araiza / Construction and Building Materials 24 (2010) 1372–1384

[15] Aprile A, Benedetti A, Cosentino N. Seismic reliability of masonry structures strengthened with FRP materials. In: Proceedings of the 8th US national conference on earthquake engineering, 2006. Paper no. 1677. [16] Benedetti A, Steli E. Analytical models for shear-displacement curves of unreinforced and FRP reinforced masonry panels. Constr Build Mater 2008;22:175–85. [17] ENV 1996-1-1 Eurocode 6. Design of masonry structures. Brussels: European Committee for Standardization; 1996. [18] Tomazevic M, Lutman M, Petrovic L. In plane behavior of reinforced masonry walls subjected to cyclic lateral loads. Report to the Ministry of Science and Technology of Republic of Slovenia, parts 1 and 2. Liubljana: Institute for Testing and Research in Materials and Structures; 1993. [19] CNR-DT 200/2004. Code for the design, execution and control of static strengthening by FRP. Rome: Consiglio Nazionale delle Richerche; 2004 [in Italian]. [20] Eshani MR, Saadatmanesh H. Shear behavior of URM retrofitted with FRP overlays. J Compos Constr 1997;1(1):17–25. [21] Haroun MA, Mosallam AS, Allam KH. In-plane shear behaviour of masonry walls strengthened by FRP laminates. In: 2nd International workshop on structural composites for infrastructure applications, Cairo, 2003. p. 17– 8. [22] Aiello MA, Sciolti SM. Bond analysis of masonry structures strengthened with CFRP sheets. Constr Build Mater 2006;20:90–100. [23] Liu Y, Dawe J, McInerney J. Behaviour of GFRP sheets bonded to masonry walls. In: Proceedings of the international symposium on bond behaviour of FRP in structures (BBFS 2005), International Institute for FRP in Construction (IIFC), 2005. p. 473–80. [24] Willis CR, Yang Q, Seracino R, Griffith MC. Bond behavior of FRP-to-clay brick masonry joints. Eng Struct 2009;31:2580–7. [25] Maruccio C, Oliveira D, Lourenço P. Development of a constitutive model for the masonry–FRP interface behaviour. In: 8th World congress on computational mechanics (WCC8) and 5th European congress on computational methods in applied sciences and engineering (ECCOMAS 2008), 2008. [26] Grande E, Imbimbo M, Sacco E. Modelling of the bond behaviour of CFRP laminates glued on clay bricks. In: 8th World congress on computational

[27] [28] [29] [30] [31] [32] [33] [34]

[35]

[36] [37]

[38]

[39] [40] [41]

mechanics (WCC8) and 5th European congress on computational methods in applied sciences and engineering (ECCOMAS 2008), 2008. Karbhari VM. Durability of FRP composites for civil infrastructure – myth, mystery or reality. Adv Struct Eng 2003;6(3):243–55. Kim HY, Park YH, You YJ, Moon CK. Durability of GFRP composite exposed to various environmental conditions. KSCE J Civ Eng 2006;10(4):291–5. Hofmann P, Stockl S. Tests on the shear-bond behavior in the bed-joints of masonry. Masonry Int 1986;9:1–15. Atkinson RH, Amadei BS, Saeb S, Sture S. Response of masonry bed joints in direct shear. J Struct Eng 1989;115(9):2276–96. Van der Pluijm R. Shear behaviour of bed joints. In: Proc. 6th North American masonry conf., Philadelphia, 1993. p. 125–36. Riddington JR, Fong KH, Jukes P. Numerical study of failure initiation in different joint shear tests. Masonry Int 1997;11(2):33–64. Lourenço PB, Ramos LF. Characterization of cyclic behaviour of dry masonry joints. J Struct Eng 2004;130(5):779–86. CEN/TC125. Methods of test for masonry – part 3: determination of initial shear strength, EN 1052-3:2002. Brussels: European Committee for Standardization; 2003. CEN/TC125. Methods of test for masonry units – part 1: determination of compressive strength. EN 772-1:2000. Brussels: European Committee for Standardization; 2001. ENV 1992-1-1. Eurocode 2. Design of concrete structures, part 1: general rules for buildings. Brussels: European Committee for Standardization; 1993. CEN/TC125. Methods of test for mortar for masonry – part 11: determination of flexural and compressive strength of hardened mortar, EN 1015-11:1999. Brussels: European Committee for Standardization; 2000. Araiza G. Repair and strengthening of masonry walls. Experimental investigation of shear strength response. Ph.D. Dissertation, Universitat Politècnica de Catalunya, Barcelona; 2005 [in Spanish]. Chen JF, Teng JG. Anchorage models for FRP and steel plates bonded to concrete. J Struct Eng 2001;127(7):784–91. Lu XZ, Teng JG, Ye LP, Jiang JJ. Bond-slip models for FRP sheets/plates bonded to concrete. Eng Struct 2005;27:920–37. Van Gemert D. Force transfer in epoxy-bonded steel–concrete joints. Int J Adhes Adhes 1980;1:67–72.