In-Plane Behavior of cavity masonry infills and strengthening with textile reinforced mortar

Transcription

1 In-Plane Behavior of cavity asonry infills and strengthening with textile reinforced ortar Farhad Akhoundi a, Graça Vasconcelos b, Paulo Lourenço b, Luis M. Silva b, Fernando Cunha c, Raúl Fangueiro c a faculty of architecture and urbanis, Islaic Art University of Tabriz b ISISE, Departent of Civil Engineering, University of Minho, Azuré, Guiarães c C-TAC, Departent of Civil Engineering, University of Minho, Azuré, Guiarães Abstract The seisic vulnerability of asonry infilled reinforced concrete (rc) fraes observed during past earthquakes in soe south European countries resulted in losses of huan lives and huge repair or reconstruction costs, justifies the need of deeper study of the seisic behavior of asonry infills enclosed in rc fraes. Therefore, the ain goals of this study are related to: (1) better understanding of the cyclic in-plane behavior of traditional brick infills built in the past decades as enclosures in rc buildings in Portugal; (2) analysis of a strengthening technique based on textile reinforced ortar (TRM) aiing at enhancing the in-plane behavior. To accoplish the objectives, an extensive experiental capaign based on in-plane static cyclic tests on seven reduced scale rc fraes with asonry infill walls was carried out. The perforance of strengthening of asonry infill based on textile reinforced ortar was also evaluated experientally. Aong the conclusions of this research, it should be stressed that: (1) the presence of infill inside the bare frae could significantly enhance the in-plane stiffness and resistance of bare frae; (2) TRM technique could enhance the in-plane behavior of infilled fraes by iproving the lateral strength and by reducing significantly the daage of the brick infill walls. 1 Introduction Masonry infills have been widely used in the building construction as enclosure walls in reinforced concrete (rc) or steel structures for any decades due to their good theral and acoustic insulation properties and also reasonable fire resistance. Nowadays, asonry infills are still typically used in odern buildings as partition and also as enclosure walls in reinforced

2 concrete fraes. Generally, they are assued as non-structural eleents and are not considered in the design of the buildings. Although the infill panels are assued as non-structural eleents, their daage or collapse is not desirable, given the consequences in ters of huan life losses and repair or reconstruction costs. Past earthquakes such as Mexico City earthquake in 1985 [1], Kocaeli (Turkey) earthquake in 1999 [2] Bhuj earthquake in 2001 [3], L Aquila earthquake in 2009 [4] have confired that asonry infills can affect the global and local behavior of the asonry infilled reinforced concrete (rc) or steel fraes. This influence can be positive or negative. When it is positive it eans that the presence of asonry infills increases the strength and stiffness of the structure to resist the lateral loads due to earthquakes. The negative influence ainly relates to the foration of soft story and short colun phenoena, which can result in the global or local failure of the structure. In-plane interaction of infill panel with its surrounding frae was studied by different researchers, either considering reinforced concrete [5-7] or steel [8, 9] as structural fraes. Fro these experiental researches, it was concluded that for low levels of in-plane loading, the infilled frae acts as a onolithic load resisting syste but when the lateral load increases, the infill tends to partially separate fro its bounding frae and a copression strut echanis develops. The assuption that infill wall behaves as a copression strut sees to be reasonable and has been supported in several experiental researches [10, 11]. It was also concluded that the added infills significantly iprove the lateral strength and initial stiffness of the bare frae and also change its dynaic properties [5] [12], which results in a relevant change in the seisic deand of the structure. Another contribution of the asonry infill within the frae is the enhanceent of the energy dissipation capacity during earthquake due to the cracking of the asonry infill, being possible to increase the daping ratio fro 4-6% to 12% [12]. Based on experiental observations, five ain failure echaniss were considered by Mehrabi et al [13] as probable failure odes that can develop in rc fraes with asonry infills. Those failure odes depend on the relative strength and stiffness of the surrounding frae with respect to the infill wall, the geoetric configuration of the frae, as well as the loading conditions. The high seisic vulnerability of the asonry infilled frae structures observed during the last decades has prooted research on techniques and aterials to strengthen the asonry infill walls and, thus, to iprove their seisic perforance. With this respect, conventional techniques or innovative aterials for in-plane strengthening have been presented. The strengthening can

3 change the behavior of the structure by changing its fundaental period as well as the center of ass and stiffness [14, 15]. Coposite aterials have been received large attention fro the research counity and they have been already applied in real context. With this regard, different researchers investigated the effectiveness of using FRPs on enhancing the stiffness, strength and energy dissipation capacity of reference speciens in the in-plane direction [16, 17]. In spite of any advantages associated with use of FRPs, this retrofitting technique is not proble-free. Soe of its drawbacks are related to the poor behavior of epoxy resins at high teperatures, relatively high cost of epoxy, non-applicability of FRPs on wet surfaces or at low teperatures and incopatibility of epoxy resins with soe substrate aterials such as clay. Specific properties of clay such as porosity and roughness, which affects the epoxy-brick bond behavior ay inhibit the use of FRP [18]. One possible solution to the above entioned probles can be the replaceent of organic binders with inorganic ones such as ceent based ortars. Seared fibers can be replaced by reinforcing eshes such as textile eshes with different continuous fibers. This results in the textile reinforced ortar technique (TRM) which is relatively new (it was started to use in early 1980s) [18-41]. The first studies on TRM technique were ainly carried out on concrete speciens. In the research carried out by Triantafillou et al. (2006) [19] TRM is used as a tool for increasing the axial capacity of RC coluns through confineent. It was concluded that using TRM jacketing resulted in substantial gain in copressive strength and deforability of the speciens. Bournas et al. [21, 25, 27] investigated the influence of TRM technique on confineent of RC coluns and Triantafillou et al. [20] conducted a research on shear strengthening of RC ebers using TRM technique as an alternative to fiber reinforced polyers (FRP). Also in the tests carried by Brückner et al. (2008) [23] and by Al-Sallou et al. (2012) [30] [23, 30] the effectiveness of textile reinforced ortar (TRM) as a ean of increasing the shear resistance of reinforced concrete beas was investigated. The effectiveness of TRM technique on flexural resistance of two-way RC slabs was investigated by Koutas LN, Bournas (2017) [41] by considering different paraeters such as nuber of TRM layers, the strengthening configuration, the textile fibers aterial (carbon versus glass) and the role of initial cracking in the slabs. In the scope of asonr strsutures, Papanicolaou et al. (2008) [24] copared the effectiveness of TRM technique with respect to FRP and NSM ethods and concluded that TRM technique coprised an extreely proising solution for the structural upgrading of asonry structures under out-of-plane loading. In the siilar study [26], asonry speciens were tested under static

4 and cyclic out-of-plane loading and the coresponding flexural behavior was analyzed. The paraeters investigated included the type of asonry wall (concrete block, sandstone, and brick), ortar (natural lie and ceent-based), and textile esh (bituen coated E-glass, basalt, or coated basalt fibers). It was concluded that the TRM speciens developed a substantial outof-plane strength and displaceent ability under static loading. Under cyclic loadind the asonry speciens were able to present stiffness and strength degradation after peak load and considerable ultiate displaceent capacitiy. Papanicolaou et al. (2007) [22] conducted also a research by testing 22 ediu-scale asonry walls subjected to cyclic in-plane loading. Fro the results it was concluded that TRM technique hold strong proise as a solution for the structural upgrading of asonry structures under in-plane loading. In a recent study carried out by Da Porto et al [36], the effectiveness of different strengthening solutions for light asonry infills were investigated by testing eight full-scale one-bay one-storey clay asonry infilled fraes, naely special lie-based plaster with geo-polyer binder, bidirectional coposite eshes applied with inorganic aterials and textile reinforced ortar (TRM) with anchorage of the esh to the rc frae. The speciens were subjected to the cobined in-plane/out-of-plane loading. It was concluded that the application of special plasters or TRM strengthening systes does not significantly change the initial stiffness or axiu in-plane resistance of the reference frae. The ain contribution can be related to the reduction of daage in the infill. It was also concluded that TRM strengthening systes iprove the out-of-plane behavior of infill walls. The research conducted by Koutas et al [35, 37] focused on TRM strengthening of one-bay three-storey infilled fraes. Connectors were used to fix the textile eshes to the upper and botto rc beas. The speciens were tested quasi-statically and the results showed that the retrofitting schee enhances the lateral strength by 56% and dissipates 22% ore energy than the control unstrengthened specien. It was also concluded that the presence of custofabricated textile-based anchors was proved to be particularly effective in delaying or even precluding the debonding of TRM. Martins et al [38] proposed an innovative reinforcing esh to be used in the TRM strengthening technique for brick asonry infills. The textile eshes are coposed of braided coposite rods (BCR) anufactured fro a braiding process. Fifteen wallets of asonry strengthened with different coercial textile eshes and with new esh with braided coposite rods were tested under four-point bending tests. It was concluded that the speciens strengthened with anufactured reinforcing eshes of BCRs with carbon fibers exhibit higher resistance to bending than other retrofitted speciens. It should be also entioned that the speciens retrofitted with anufactured eshes of braided coposite aterials with a

5 core of glass fibers presented rearkably better post-peak behavior than the other retrofitted speciens. The evaluation of the perforance of the above entioned textile eshes coposed of braided coposite rods in asonry infill tested under in-plane loading is presented and discussed in this paper as part of an extensive research carried out on the echanical behavior and strengthening of traditional brick asonry infills used in Portuguese reinforced concrete buildings and that can be also representative of the construction practice in other south European countries in the past decades. This paper presents and discusses the experiental results on in-plane static cyclic tests of double leaf (cavity) brick asonry infill walls before and after strengthening with textile reinforced ortar. Additionally, the asonry construction quality is also taking into consideration by using two different asons in the construction of asonry infills. Although the effect of infills inside rc fraes were studied by large nuber of researchers taking into account different configurations, there are soe factors in this research that akes it relevant to strengthen the state of the art in this field of study: (1) the typical Portuguese asonry infilled rc frae fro 1960s are considered to be also typical construction practice in other south European countries. The infill consists of double-leaf asonry with horizontally perforated clay bricks of low copressive strength; (2) the strengthening based on textile reinforced eshes needs to be better studied, particularly when there is a need to study the strengthening of this type of constructive eleents and when the perforance of a new esh is evaluated. 2 Experiental Progra The experiental progra for the characterization of the in-plane behavior of traditional brick asonry infill walls typical of south European countries was based on static cyclic in-plane tests. For this, five brick asonry infilled rc fraes were considered. Additionally, two additional rc fraes with strengthened asonry infills were tested by using the sae experiental setup and protocol. The strengthening of the asonry infilled fraes was carried out by adding textile eshes ebedded in rendering ortar (textile reinforced ortar TRM) to the brick asonry infilled fraes. Taking into account the liited facilities at the laboratory of Civil Engineering at University of Minho and to avoid difficulties in handling full scale speciens, it was decided to design reduced scale speciens to represent the full scale rc frae with infill wall. The design of the reduced scale speciens was based on Cauchy s siilitude law and test the according to the loading pattern that coplies with FEMA-461 guidelines [42]. Details about the prototype walls, the design of the reduced scale speciens, tests setups and loading pattern are represented in the next sections.

6 2.1 Characterization of Prototype and designing reduced scale speciens The prototype of an rc frae with asonry infill was defined based on a study carried out for the characterization of the typical rc buildings constructed in Portugal since 1960s [43]. About 80 buildings were analyzed to identify the cross section of beas and coluns, the reinforcing schees of those eleents, the geoetry of the brick asonry walls and the nuber and typology of the openings within the walls and also their position. Fro this study, a prototype with length of 4.50 and height of 2.70 was defined. The cross section of the rc coluns was 0.3x0.3 and of the rc beas was 0.3x0.5. The asonry infills were ostly built as cavity walls coposed of two leaves with horizontal perforated clay brick units. The external leaf has a thickness of 15c and the internal leaf has a typical thickness of 11 c, being separated by an air cavity of about 4 c. An overview of the scaled reinforceent schee of the rc frae and of the cross sections of coluns and beas are shown in Figure 1 and Figure 2. For the asonry infills, horizontally perforated bricks of 175x115x60 and of 175x115x80 were used to have siilar height to length ratio of the bricks (0.67) with the one used in the prototype. For the design of reduced scale speciens, an allowable stress design approach was followed. A scale factor of 0.54 was adopted to scale all eleents, including the diensions of the bricks. In the first step, the sections of the real scale rc coluns and beas of prototype were analyzed based on ACI [44] guidelines in order to obtain the axiu resisting forces and flexural oents in the coluns and beas. After this, Cauchy s siilitude was applied to calculate the axiu allowable forces and bending oents of reduced scale cross sections fro the axiu allowable forces and flexural oents of real scale sections obtained in the first step. Finally, cross-sections and reinforceent of the reduced scale structural eleents were designed based on the sae allowable stress design approach.

7 Table 1 Relation between different paraeters of prototype and odel based on Cauchy s Siilitude Law Paraeter Scale Factor Paraeter Scale Factor LP P 3 Length (L) Mass () L E P Elasticity Modulus (E) 1 E P Specific Mass (ρ) 1 Area (A) Volue (V) Displaceent (d) A A V V d d P P P 2 3 vp Velocity (v) 1 v Acceleration (a) a a P 1 Weight (W) Force (F) Flexural Moent (M) W W F F P P M M P P Stress (σ) 1 P Strain (ε) 1 t P Tie (t) t Frequency (f) f f P 1 In total, seven speciens were considered in the experiental capaign (Table 2), naely one bare frae (specien BF-I), four unstrengthened infilled fraes and two strengthened speciens with TRM technique. It is stressed that the unstrengthened speciens were tested until distinct values of iposed lateral drift to siulate different in-plane daage levels: one specien was tested until collapse (specien SIF-I-A), one specien was tested until iposed lateral drift of 0.3% (specien SIF-I(0.3%)), one specien was tested until lateral drift of 0.5% (specien SIF- I(0.5%)) and one specien was tested until lateral drift of 1.0% (specien SIF-I(1.0%)).

8 In the strengthened speciens two different types of reinforcing eshes were used, naely a coercial esh (specien SIF(CTRM)-I-B) and the textile esh developed at University of Minho (specien SIF(DTRM)-I-B). Figure 1 Geoetry and reinforceent schee of the reduced scale rc frae Figure 2 Cross-sections of coluns and beas in reduced scale rc fraes Table 2 Designation of the speciens for in-plane static cyclic loading Nae Type of specien Type of loading Nuber of leaves during construction Group of Mason BF-I Bare frae In-plane - SIF-I-A Solid infilled frae In-plane Double leaf A SIF-I(0.3%)-B Solid infilled frae Prior In-plane drift of 0.3% Double leaf B SIF-I(0.5%)-B Solid infilled frae Prior In-plane drift of 0.5% Double leaf B SIF-I(1%)-B Solid infilled frae Prior In-plane drift of 1% Double leaf B SIF(DTRM)-I-B SIF(CTRM)-I-B Solid infilled frae strengthened with TRMdesigned esh Solid infilled frae strengthened with TRMcoercial esh In-Plane Double leaf B In-Plane Double leaf B

9 It should be entioned that the workanship used in the construction of the unstrengthened specien loaded until collapse was different fro the workanship used in the construction of the reaining speciens. For the specien loaded until collapse, the workanship is considered as type A and for the other speciens the workanship is denoted as type B. This designation is considered in the last part of the labeling of the speciens as presented in Table 2. The need of a new ason for the construction of the reaining speciens was derived fro the poor workanship available for the construction of unstrengthened speciens. The steel used for the construction of rc frae was of class A400NR, with a yielding tensile strength of 400MPa and for the concrete, a C20/35 class was adopted. The coercial and developed reinforcing eshes used in the strengthened speciens consist of bi-directional glass fiber eshes. The developed esh is coposed of a set of coposite rods with an external polyester helicoidally braided with a reinforcing nucleus of glass fibers. The idea is that the braided rod can protect the reinforcing fibers and provide ductility after the rupture of the fibers [38], see Figure 3a. The bond between the external braid and the reinforcing fibers can be ensured in the anufacturing process by adding polyester resin during the braiding process or after the production of the coposite braided rods by adding the resin over the external polyester anually [38]. For the application on the asonry infills, the coposite rod was coposed of a braided structure with 15 ultifilaent of polyester with 11 Tex and one braided eleent with a siple structure consisting of 8 braided polyester yarns produced at the axiu speed of the production equipent (1.07/in). To have anufactured eshes that are coparable with coercial eshes, 5 glass ultifilaent of the 544 Tex were required, corresponding to a density of 207 g/ 2 (about 92% copared to coercial esh). As the coercial eshes present a spacing of 25, the anufactured eshes were anufactured with the sae spacing. The anufacturing of the eshes is carried out by interlacing the coposite rods in two directions, assuing that the configuration of the connections of rods in two directions ay proote an additional interlocking and can work as additional roughness, iproving the bond adherence between the eshes and the rendering ortar, see Figure 3b. The esh has a tensile strength of about 55kN/ and stiffness of about 154kN/. The coercial esh consists of resistant glass fibres in both directions, see Figure 3c. The esh density is 225 g/2 with spacing between the fibres of 25. Fro the technical inforation, it is seen that the tensile strength is 45 kn/ and the ultiate elongation is about 1.8%. The odulus of elasticity is 72GPa and the equivalent thickness is about

10 (a) (b) (c) Figure 3 Details of braided rods and eshes; (a) cross section of a braided esh [38]; (b) designed esh; (c) coercial esh 2.2 Mechanical properties of aterials Brick units and ortar The construction of the asonry infills was carried out by using a preixed ortar of class M5, given that the copressive strength shall be close to the one used in the past decades in the construction of infills in rc buildings. In addition, in order to control the quality of ortar, it was decided to use a preixed ortar. The sapling of the ortar was carried out during the construction of the asonry infill walls (internal and external leaves) for the assessent of the construction quality of asonry walls and echanical characterization. The consistency of ortar was based on flow table tests [45] and the copressive and flexural strength was obtained according to EN :1999[46]. In spite of the special attention ade for ixing the ortar, soe scatter on the results of the flow table test was achieved, ranging fro 165 in specien SIF(CTRM)-I-B to 180 in specien SIF-I-A. This resulted also in soe variation on the flexural and copressive strength of ortar. The average copressive strength ranges fro 2.76MPa with a coefficient of variation (COV of 5.0%) to 6.18MPa (COV of 5.8%). The average flexural strength, ff, ranges fro 1.20MPa (COV of 9.0%) to 2.51MPa (COV of 6.0%). As entioned before, the units used in the construction of the brick infills were considered with reduced diensions to coply with the requireents of the reduced scale of asonry infills. Taking into account that the typical thickness of double leaf brick infills is 15c for the external leaf and 11 c for the internal leaf, it was decided to use coercial brick units with theoretical diensions of 24.5cx11.5cx8c and 24.5cx11.5cx6c (length x height x thickness) for the external and internal leaves respectively. To have the siilar height to length ratio of the units of the prototype in the reduced scale units, the length of the bricks was cut to The copressive strength of bricks was obtained according to the European Standard EN-772-1:2000 [47] and by considering three directions of loading, naely the direction parallel to the

11 height, parallel to the horizontal perforation and parallel to the thickness. According to the average copressive strength of the bricks, fb, shown in Table 3, the copressive strength is clearly different aong the loading directions. The higher value is obtained in the parallel direction to the length, corresponding to the direction of perforations. It should be noted that the resistance of the bricks used in the external leaf is lower than the copressive strength obtained in the bricks used in the internal leaf, and lower than the value provided by the producer. The average odulus of elasticity, obtained according to EN-772-1:2000 [47], in direction of parallel to the perforations is MPa (COV of 22.9%) and MPa (COV of 20.3%) in the bricks of external and internal leaf respectively. Table 3 Copressive strength of bricks. Coefficient of variation is indicated inside brackets. f b (MPa) External leaf Internal leaf Parallel to the height 1.57 (15.1) 2.75 (27.7) Parallel to length 2.53 (22.8) 8.17 (3.4) Parallel to the thickness 1.83 (28.6) 4.86 (15.3) The quality of the rendering ortar was controlled by testing the consistency of ortar according to EN , 1999 [45] in the fresh state and by obtaining the copressive and flexural strength in hardened ortar according to EN , 1999 [46]. For the latter case, three saples of ortar were taken per each retrofitting layer to characterize the flexural and copressive strength according to European standard. The flow table test results on ortar used for rendering with designed and coercial eshes were calculated as 162 and 160 respectively. The ortar used in rendering with designed esh had a copressive strength of 9.11MPa (COV of 2.31%) and a flexural strength of 3.50 MPa (COV of 2.59%). Those values for the ortar used in coercial esh were MPa (COV of 4.28%) and 3.87 MPa (COV of 6.62%) respectively Brick asonry The echanical properties obtained in the echanical characterization of brick asonry include the copressive strength, tensile and shear strength, flexural strength and the echanical shear properties of the unit-ortar interface. The copressive strength and the elastic odulus of brick asonry used in the construction of the infill walls were obtained based on uniaxial copressive loading following the European standard of EN1052-1:1999 [48]. Three speciens were adopted to represent the external leaf of the asonry and only one specien was used to represent the internal leaf due to the liitation

12 of brick units. An average copressive strength of 1.17MPa (COV of 4.8%) and an average elastic odulus of MPa (COV of 29.8%) were obtained for the brick speciens representing the external leaf. The copressive strength and the elastic odulus for the brick asonry representing the internal leaf were 1.59MPa and MPa respectively. Although the brick units of the internal leaf were clearly stronger than the bricks of the external leaf, its copressive strength is closest to the copressive strength of the external leaf. The shear resistance of brick asonry was characterized through diagonal copression tests, following the recoendations of ASTM standard [49]. Diagonal copression tests were perfored on three asonry speciens representing the external leaf and on one specien representing the internal leaf. An average shear strength of 0.24MPa (COV of 9%) and shear odulus of MPa (COV of 1.1%) were obtained for the brick speciens representing the external leaf. The shear strength and shear odulus for the brick asonry representing the internal leaf were 0.17MPa and MPa respectively. The flexural resistance of the brick asonry was obtained in two different directions, naely in directions parallel and perpendicular to the bed joints. The diensions of the speciens adopted for flexural testing were defined according to European standard EN (1999) [50]. In the test setup, the specien was placed vertically and the flexural load was applied in the horizontal direction. This testing configuration was decided based on the test setup facilities and on the fragility of the speciens, due to their reduced thickness. Three speciens of the brick asonry representing the external leaf and one specien of the brick asonry representing the internal leaf were tested for each direction. The flexural tests were carried out under displaceent control by onotonically increasing the displaceent by 0.1/sec. For the external leaf, the average flexural strength of 0.053MPa (COV of 6.4%) in the direction parallel to the bed joints and of 0.29MPa (COV 14%) in the direction perpendicular to the bed joints were obtained. In case of the specien representative of the internal leaf, the flexural strength in direction parallel to the bed joints was 0.059MPa, whereas in the direction perpendicular to the bed joints it was calculated as 0.23MPa. The in-plane initial shear strength of unit-ortar interface was deterined by testing nine asonry priss representative of the external leaf wall according to the European standard EN1052-3:2003[51]. As the copressive strength of the units is less than 10MPa, the precopression loads were defined so that they represent reasonable values without any type of copression failure. Therefore, confining copressive stresses of 0.1MPa, 0.3MPa and 0.5MPa were adopted. Average values of about 0.18MPa and 0.58 were calculated for the cohesion and

13 tangent of friction angle, calculated through the linear fitting to the experiental results (r 2 =0.87). 2.3 Construction and strengthening of the speciens The construction process of the speciens was divided into two phases, naely casting of the rc fraes and construction of the brick asonry infills. In order to ake the casting of the rc fraes easier, it was decided to put the wooden olds in the horizontal position. The reinforceent was ounted separately and then placed inside the wooden old. As there was little space between the longitudinal reinforceent of the beas in the reduced scale speciens, a concrete with a axiu aggregate of 9 with a characteristic copressive strength of 25MPa was ordered to ensure unifor concrete in the speciens, even in the places where little space exists between the reinforceents. The construction of the asonry infills was carried out in the storage area when the fraes were placed and after finishing the construction of the walls, they were covered to be protected fro the rain. After 28 days of curing tie for asonry infills, the rc faes with brick asonry infills were carefully transported to the testing place by eans of a crane to avoid any cracking. The rendering ortar used in the strengthened speciens was a pre-ixed coercial ortar indicated to be applied with the selected coercial textile esh and was applied in both external surfaces of internal and external leaves. A ultipurpose latex additive was added to the pre-ixed ortar aiing at iproving its workability and consequently enhancing the echanical and adhesive characteristics of ceent-based rendering ortar. Additionally, L- shaped glass fiber connectors were used both in the asonry infill and in the rc frae aiing at avoiding any detachent of the rendering ortar (Figure 4a). The application of reinforced rendering to the asonry infills was carried out in the following steps: (1) definition of the pattern for pilot holes (Figure 4a) to place the connectors aiing at iproving the adherence of the rendering ortar to the asonry infill; (2) drilling and cleaning the holes and insertion of plastic row plugs shown in Figure 4b in the holes (Figure 5a); (3) application of the first thin layer of about 5 of ortar (Figure 5b); (4) injection of a cheical anchor into the holes and inserting the L-shaped glass fiber connectors; (5) positioning of the textile esh on the first layer of ortar; (6) application of the second layer of ortar and rectifying the rendered surface, see (Figure 5c,d). The total thickness of the rendering was easured as approxiately 20 in all the speciens. The application of the rendering in two successive layers enables the involveent of the textile esh by the ortar and also adequate developent of the adherence between the.

14 (a) (b) Figure 4 Details of the esh connectors; (a) pattern of the connectors (b) plastic row plug and glass fiber connector (a) (b) (c) (d) Figure 5 Application of the reinforced rendering; (a) drilling the pilot holes (b) applying the first layer of ortar (c) positioning of the textile esh and application of the second layer of ortar; (d) final aspect after rendering 2.4 Experiental setup and instruentation The test setup designed for the static cyclic in-plane testing of the rc fraes with asonry infills is shown in Figure 6. The rc frae with asonry infill was placed on two steel beas (HEA300) that were firly attached to the strong floor to avoid its sliding and uplifting. Additionally, the sliding of the rc frae was prevented by bolting an L-shape steel profile (L200x200x20) to each side of the steel bea. In turn, the uplifting was additionally prevented by bolting two tubular steel profiles (two welded UNP140 steel profiles) to the steel beas. The out-of-plane oveent of the enclosure frae was restrained by putting L-shaped steel profiles (L100x100x10) at each side of the upper concrete bea, which instead were bolted to the top steel frae, see Figure 6b. Three rollers were placed on upper L-shaped profiles to iniize or even eliinate the friction between the and the upper reinforced concrete bea during in-plane loading.

15 Two vertical jacks were placed on the top of the coluns to apply the vertical load of 160 kn, corresponding to 40% of the colun s axial force capacity. Each jack was pinned to the lower steel bea by eans of four vertical rods with a diaeter of 16 (two at each side). A hydraulic actuator with a capacity of 250kN was attached to the reaction wall and connected at id height of the top concrete bea of the rc frae to apply the horizontal cyclic load. Two steel plates (400x300x30 3 ) were ounted on the left and right sides of top rc bea and were connected by two 2φ50 steel rods confining the top rc bea. The right steel plate was connected to the horizontal actuator by eans of a hinge, enabling the application of reversed cyclic loading. (a) (b) Figure 6 a) Test setup for in-plane cyclic loading b) out-of-plane support of the upper bea in the in-plane testing The instruentation adopted for the easureent of the ost relevant displaceents during the static cyclic in-plane testing is shown in Figure 7. Twenty-two linear variable differential transforers (LVDT) were used to record the key displaceents. Fro these LVDTS, four LVDTs were ounted on the asonry infill to easure the diagonal deforation of both leaves (L1, L2, L21 and L22). Two LVDTs were placed on the reinforced concrete frae to easure the diagonal deforation of the surrounding frae (L19 and L20). Eight LVDTs were used to easure the relative displaceent of the infill with respect to the surrounding frae in the corners (L3, L4, L5, L6, L7, L8, L9 and L10). LVDTs L11 and L12 easured the sliding and uplifting of the infilled frae with respect to the steel profile. Four LVDTs of L13, L14, L15 and L16 easured the sliding and uplifting of the steel profiles with respect to the strong floor. These easureents were taken to control the reliability of the test setup. The LVDTs L17 and L18 easured the horizontal displaceent of the upper bea of the reinforced concrete frae.

16 The instruentation of the strengthened speciens was defined to have the response of both leaves recorded, see Figure 8. Four LVDTs (L1 to L4) were placed along the diagonals to onitor the diagonal deforation of the external and internal leaves. Twelve LVDTS (L5 to L16) were placed on the rendering layer to capture its possible detachent in relation to the rc frae in the out-of-plane direction. Six LVDTs (L17 to L22) were positioned to onitor the possible uplifting and sliding of the specien fro ground and steel profiles and thus control the reliability of the test setup. Finally, two LVDTs, L23 and L24, were placed on the top bea to easure the horizontal displaceent at the top rc bea in the direction of the applied load. (a) (b) Figure 7 Instruentation for in-plane loading; (a) external leaf (b) internal leaf (a) (b) Figure 8 Instruentation of the specien for in-plane loading; (a) external leaf (b) internal leaf 2.5 Loading Protocol The in-plane static cyclic tests were perfored in displaceent control by iposing increasing pre-defined levels of displaceents through an LVDT connected to the horizontal hydraulic actuator. The loading protocol adopted for in-plane static cyclic testing, which is in accordance with the guidelines provided by FEMA 461[52], is shown in Figure 9. The loading protocol includes sixteen different sinusoidal steps, starting fro a displaceent of 0.5, representing

17 0.03% drift, calculated as the ratio between the top lateral displaceent and the height at which the horizontal load is applied fro the base of the frae. Figure 9 Displaceent protocol for in-plane testing The axiu lateral displaceent in the protocol is 66.68, corresponding to a lateral drift of 3.5%. Each step was repeated two ties unless the first step that was repeated for six ties. The displaceent aplitude (ai+1) of step i+1 is 1.4 ties the displaceent aplitude (ai) of step i: Eq. 1 It should be stressed that the aboveentioned loading protocol is a general protocol used for all speciens that enables to follow the in-plane response until a lateral drift of 3.5%. However, in case of speciens tested until certain lateral drifts to induce different levels of in-plane daage, the tests were stopped at the corresponding displaceent. The wall subjected to the axiu lateral drift of 0.3% was tested until a lateral displaceent of 5.7, the wall subjected to the axiu lateral drift of 0.5% was tested until a lateral displaceent of 9.5, and the wall subjected to the axiu lateral drift of 1.0% was tested until a lateral displaceent of The acquisition rate was two per second. 3 Experiental results and discussion 3.1 Force-displaceent diagras The lateral force-displaceent diagras obtained for the different unstrengthened speciens tested under cyclic in-plane loading are shown in Figure 10. The positive direction is considered to be the direction in which the hydraulic actuator pushes the specien whereas the negative

18 direction is the direction in which the actuator pulls the specien through two plates that were connected with two thick steel rods. It is iportant to note that both lateral strength and initial stiffness of the rc frae with asonry infill are significantly higher than the values found for the bare frae, which confirs the role of the asonry infill in the lateral strength of the rc fraes. The lateral displaceent at which the lateral resistance is attained is uch lower in case of the rc fae with asonry infill, which is related to the higher stiffness of the infilled frae. The force-displaceent diagras are characterized by an initial linear behavior corresponding to the elastic behavior of the structure. After the onset of cracking, the nonlinear behavior is visible both through the nonlinearity of the force-displaceent envelop and through the higher hysteresis corresponding to the developent of daage and dissipation of energy. Given the shape of the hysteretic force-displaceent diagra, it is expected that the energy dissipation is higher in the rc frae with asonry infill, being associated to the cracking developent in the asonry wall. The strength degradation in the second cycle starts after cracking and increases as the daage accuulation increases. This is particularly evident in the coplete force-displaceent diagra of rc frae with asonry infill built with ason A tested until failure. The post-peak behavior is characterized by a progressive but sooth reduction on the lateral strength and stiffness, eaning that the daage is controlled and progressive. The lateral strength of the rc frae with asonry infill corresponding to the ultiate displaceent is higher than the lateral strength of the bare frae, which appears to indicate that in spite of the severe cracking of the asonry infill, it still contributes for the lateral strength of the coposite structure until the end of the test. (a) (b)

19 (c) (d) (e) Figure 10 Force-displaceent diagra; (a) bare frae; (b) specien SIF-I-A; (c) specien SIF-I(0.3%)-B; (d) specien SIF-I(0.5%)-B; e) specien SIF-I(1%)-B The cyclic behavior of the rc fraes with asonry infills is characterized by relatively low values of peranent deforation, being great part of the deforation corresponded to a certain iposed lateral drift recovered during the release of the lateral force. This feature is attributed to the opening and closing of the cracks that developed through unit-ortar interfaces by applying and releasing the force respectively. The lateral strength of the rc frae with asonry infill built with ason B tested until 0.5% and 1% lateral drifts is higher than the strength of the rc frae with asonry infill built with ason A. The increase is about 26% and 16% in the speciens tested until lateral drifts of 0.5% and 1% respectively. Note that it is considered that in both cases the axiu lateral strength is attained, even if at different lateral drifts, enabling the coparison in ters of lateral strength. It is also interesting to ention that the displaceent corresponding to the axiu lateral force

20 tends to be higher in the rc fraes with asonry walls built with ason B. These results show the relevance of the quality of the workanship in the in-plane response of the rc infilled fraes. The cyclic force-displaceent diagras obtained for the asonry infilled fraes strengthened with textile reinforced ortar are shown in Figure 11. The ost relevant difference with respect to the coplete response of the rc frae with asonry infill built with ason A is the increase of the stiffness and lateral strength. The nonlinear behavior before the peak is ore liited in the strengthened asonry speciens, when copared with the rc fraes with asonry infill. The deforation capacity of these speciens is higher than the deforation attained in the unstrengthened speciens but the plastic deforation is higher, which should be associated to ore peranent daage. (a) (b) Figure 11 Force-displaceent diagra (a) specien SIF(CTRM)-I-B; (b) specien SIF(DTRM)-I-B The post-peak behavior is very controlled, eaning that very progressive and sooth reduction of the lateral loadbearing capacity of the coposite structure is observed. The ultiate lateral drift is slightly higher in case of the specien strengthened with the designed textile esh (SIF(DTRM)-I-B) when copared to the specien strengthened with the coercial esh (SIF(CTRM)-I-B), but it should be stressed that there are no significant differences between both strengthened speciens. This appears to indicate that the role of the textile eshes is very close. 3.2 Crack patterns The final cracking pattern developed in the cavity walls during the cyclic in-plane tests are shown in Figure 12.

21 (a) (b) (c) (d) (e) Figure 12 Final cracking pattern of the bare frae; (b) specien SIF-I-A at end of the test; (c) specien SIF- I(0.3%)-B at end of the test; (d) specien SIF-I(0.5%)-B; (e) specien at lateral drift of 1%; (f) specien SIF(CTRM)-I-B It is clear that the final cracking patterns of the bare frae show that all the cracks are ostly concentrated in the coluns, indicating the developent of plastic hinges at the top and botto part of the coluns. This behavior should be associated to the design criteria of rc fraes used before 1980s, when no specific seisic design was used for rc buildings. It can be concluded that based on the experiental results, the concept of weak colun-strong bea characterizes this rc frae odel representing a prototype built before 1980s. It is iportant to stress that the crack patterns observed on the rc fraes with the brick asonry infill walls are clearly dependent on the axiu lateral drift iposed to the coposite structure. In fact, the ost cracked specien is, as expected, the one tested until the lateral drift of 2.5% (SIF-I-A), corresponding practically to a near collapse liit state. In this case, the cracking pattern of the asonry infill is characterized by: (1) separation of the infill fro its rc

22 frae. This separation starts fro very early stages of lateral deforation; (2) diagonal cracking along the unit-ortar interfaces. After the separation between the asonry infill and the rc frae, the diagonal strut fors and diagonal cracking develops; (3) crushing of brick units adjacent to the top interface between the rc frae and asonry infill. It is observed that the cracking of the rc coluns is ore controlled in case of presence of the asonry infill. The daage observed in the infilled fraes is increasing with increasing iposed lateral displaceent. In the specien tested until lateral drift of 0.3%, part of the rc frae-asonry interfaces were separated, together with very thin cracks along the ain brick-ortar interfaces. For the specien tested until 0.5% lateral drift (SIF-I(0.5%)-B), the asonry infill was copletely separated fro the rc frae and the density of stair-stepped cracks along the unitortar interfaces is higher, being visible soe horizontal sliding cracks. Few cracks in the connection of the rc coluns to the top bea were also observed. The specien subjected to the lateral drift of 1% (SIF-I(1%)-B) presented ore intense daage when copared to the daage observed in the specien tested until lateral drift of 0.5% (SIF-I(0.5%)-B), ainly at the level of cracking density on the asonry infill and at the rc coluns. It is iportant to note that the cracking shown in Figure 12 is regarding to the external leaf but the cracking in the internal leaf is very siilar, indicating that that both leaves are effective in the resistance to the in-plane loading. This siilar behavior exhibited by the internal and external leaves is also confired by the average shear distortion calculated for both internal and external leaves that presented in Figure 13. (a) (b) Figure 13 Average shear distortions (a) asonry infill built with ason A; (b) rc frae with asonry infill tested until lateral drift of 1%

23 The final cracking patterns observed in the speciens strengthened with textile reinforced ortar are presented in Figure 14. It is noticed that seared cracking pattern is observed in the ortar layers of the specien strengthened with the coercial glass fiber esh. The cracks developed ostly along the diagonals and soe horizontal cracks at the level of the upper and botto interfaces between the asonry infill and rc frae were also observed. These horizontal cracks should indicate the trend of separation of the asonry infill fro the rc frae. In case of the speciens strengthened with bi-directional esh coposed of braided coposite rods, only few cracks were developed in the reinforced ortar layer. Horizontal cracking was visible along the infill-rc frae interfaces. Soe sall cracks were also observed in the areas where the shear connectors were totally failed. The strengthening ortar layers started to detach fro the rc frae at early stages of loading (lateral drift of 0.07% in both directions) as observed fro the evolution of the displaceent easured at the LVDTs placed to easure eventual debonding of the ortar layer, see Figure 15. Fro the results obtained, it sees that other type of connectors should be used in the rc fae. Besides, it should be entioned that the failure of the connectors of the rc fraes is brittle and, thus, ore ductile aterial should be selected. On the other hand, it is entioned that the connectors behaved in appropriate way in case of brick infill, as no detachent of the reinforced ortar layer fro the asonry infill was detected. It is observed that the displaceents easured by different LVDTS are very siilar, which indicates that the separation of the ortar layer fro the rc frae was practically unifor along the height of the specien. The appropriate preparation of concrete surface before applying the rendering ortar according to what is pointed out by Escrig et al (2015) [53] can be assued as an alternative solution for enhancing the adherence of the rendering ortar to the reinforced concrete frae. Figure 14 Cracking pattern in the rendering ortar at the end of the test; (a) specien SIF(CTRM)-I-B; (b) specien SIF(DTRM)-I-B

24 (a) (b) Figure 15 Detachent of the reinforced ortar layer fro the rc frae (a) specien strengthened with coercial esh at lateral drift of 0.27% b) (a) specien strengthened with developed esh at lateral drift of 0.20% After the test, the reinforced ortar layer was reoved echanically to obtain inforation about the daage of the brick infill. It was seen that the asonry units crushed along the rc frae-infill interfaces, see Figure 16. (a) Figure 16 View of the crushing of bricks of the infill after reoving soe parts of the retrofitting layer in speciens a) SIF(CTRM)-I-B b) SIF(DTRM)-I-B This could be related to the higher concentration of copressive stresses along the edge of the brick infill. It was also seen that no additional cracks were fored in the brick infill apart fro the crushing in the bricks adjacent to the rc frae, indicating that the reinforced ortar layer acts as a daage controller by liiting the developent of cracks in the brick infill. The crushing of the bricks along the interface is very siilar between both strengthened speciens, even if it appears to be ore significant in the specien strengthened with the coercial textile esh. (b)

25 3.3 Evaluation of in-plane perforance Key paraeters The force-displaceent envelops of all speciens tested under cyclic in-plane loading are presented in Figure 17. These envelops were defined for the first cycle of in-plane loading protocol. The coparative analysis aong the curves enables to identify clearly the differences aong the distinct speciens. These envelops are used to define the key paraeters for systeatic coparison of the in-plane response of the rc fraes with asonry infills, naely: (1) the lateral force corresponding to the crack initiation, Hcr, and the corresponding lateral displaceent, dcr,; (2) the secant stiffness at the first cracking point, calculated as the ratio between the crack initiation force and crack initiation displaceent; (3) the axiu lateral force, Hax, and the corresponding lateral displaceent, dhax. The key paraeters characterizing the in-plane response of the different speciens are shown in Table 4. (a) (b) Figure 17 Monotonic in-plane force-displaceent diagra of all speciens It is clear that the presence of infill significantly increases the initial stiffness and lateral strength of the rc bare frae. This increase in the initial stiffness and lateral strength is about 5.2 and 1.25 ties in case of brick infill built with ason A. In case of brick infill built with ason B, the iproveent on the initial stiffness and lateral strength is even higher, being approxiately 14.9 and 1.9 ties respectively. Additionally, it is observed that the initial stiffness of speciens built with asonry B is ore than the double of the initial stiffness recorded for the specien built by the first ason (ason A). It is considered that the axiu lateral force was attained in speciens tested until lateral drifts of 0.5% and 1%. In case of the specien tested until the lateral drift of 0.5%, the axiu force was attained at the lateral drift of 0.38% and 0.27% in the positive and negative directions respectively. Based on this, it is possible to calculate the