Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams

Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. HYBRID STRENGTHENING AND FLEXURAL BEHAVIOUR OF POST-TENSIONED 1 LAMINATED GLASS BEAMS 2 3 Jorge Rocha 1 , Eduardo Pereira 2 , Julien Michels 3 , José Sena-Cruz 4 4 5 ABSTRACT: The strengthening of glass for structural applications is challenging. EBR (External Bonded 6 Reinforcement) reinforcement is often preferred, although limitations result from stress concentrations especially 7 in the case of post-tensioning. This study explores the use of CFRP (Carbon Fiber Reinforced Polymers) and/or 8 Fe-SMA (Iron-based Shape Memory Alloys) as reinforcement in hybrid and alternative configurations. Five full 9 size laminated glass beams were tested in flexure, combining NSM (Near-Surface Mounted) and EBR techniques 10 and exploring the efficiency of different alternative strengthening layouts (i.e. reinforcement material versus 11 application technique) to overcome the current shortfalls of glass strengthening for structural applications. Even 12 for lower reinforcement ratios, hybrid strengthening systems performed better and increased load-carrying 13 capacity while delaying crack-induced deboning failure, when compared to conventional EBR systems. Residual 14 strength ratio increased from 87 % to 160.8 % and ductility from 407 % to 971 %. The hybrid system combining 15 NSM-CFRP and EBR-SMA reinforcements showed the best performance, increasing the initial fracture stress of 16 glass and maintaining sufficient residual strength capacity, while allowing the safe post-tensioning. 17 18 Keywords: Fe-SMA strips; CFRP laminates; laminated glass; post-tensioning; recovery stress; hybrid 19 strengthening system 20 1 PhD Student, ISISE, IB-S, University of Minho, School of Engineering, Campus de Azurém, 4800-058 Guimarães, Portugal. Email: a618[email protected].pt 2 Assistant Professor, ISISE, IB-S, University of Minho, School of Engineering, Campus de Azurém, 4800-058 Guimarães, Portugal. Email: eduar[email protected]minho.pt 3 Managing Director, re-fer AG, Riedmattli 9, 6423 Seewen, Switzerland. Email: [email protected] 4 Associate Professor with Habilitation, ISISE, IB-S, University of Minho, School of Engineering, Campus de Azurém, 4800-058 Guimarães, Portugal. Email: [email protected] Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. 1. INTRODUCTION 21 Glass has been used for centuries because of its transparency and aesthetics (e.g. windows). However, in 22 recent decades, glass has also been used in more demanding applications including structural functions (e.g. floors, 23 beams and frames), being generally designed to withstand flexural loads. Following the design methodologies 24 used in the aeronautic industry, glass structures have been designed according to the concepts of hierarchy, 25 robustness and redundancy [1]. In the case of structural applications laminated glass is preferred, because it 26 satisfies the concept of redundancy by dividing the glass panel into thinner glass plies joined by transparent 27 interlayers. 28 1.1. Challenges for glass composite systems 29 Currently the design of glass structural elements consists on verifying whether the tensile strength of glass 30 is high enough to withstand the anticipated actions. However, its long-term tensile strength is unreliable due to the 31 growth of surface flaws under varying humidity conditions [2]. Glass requires sufficient structural redundancy 32 after breakage in order to accomplish the robustness requirements, however glass lamination is not sufficient to 33 provide the desirable robustness for unpredictable actions, e.g. vandalism and earthquakes, or imperfections 34 derived from fabrication and assembly, including stress concentrations. Alternative load carrying mechanisms are 35 required to provide ductility and residual strength capacity after glass cracking. Materials such as timber (e.g. 36 [3,4]), Carbon Fiber Reinforced Polymers – CFRP (e.g. [5–8]), Glass Fiber Reinforced Polymers – GFRP (e.g. 37 [9–12]) and steel (e.g. [13–17]) have been combined with glass aiming at enhancing the post-failure redundancy 38 of glass structures. 39 Annealed glass is better at providing residual strength capacity after cracking than thermally toughened 40 glass [17,18]. However, without the residual stress field generated by tempering, its tensile strength is time41 dependent [19]. Thus, in analogy to prestressed concrete, the mechanical post-tensioning of glass has been 42 investigated. This approach does not affect glass fragmentation and contributes to the overall behaviour, unlike 43 thermal toughening which is only effective before glass cracking [17]. In the few studies found in the literature 44 Steel (e.g. [13,20–22]), CFRP (e.g. [8]) and SMA (e.g. [23]) have been used in glass post-tensioning. Most 45 frequently, in these systems the reinforcement is mechanically anchored (e.g. [20]) and/or adhesively bonded (e.g. 46 [13]) to the glass substrate. 47 Among the different strategies adopted for strengthening glass, the EBR technique (e.g. [5,13,24]) and 48 reinforcement lamination (e.g. [7,17,25]) are the most frequently used. EBR systems often fail by crack-induced 49 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. interfacial debonding, both in concrete (e.g. [26]) and glass (e.g. [27]). As shear reinforcement is not easy to 50 implement in glass structures due to obvious aesthetic constraints, shear cracks induce high interfacial stresses at 51 the bonded interfaces, which in turn generate delamination cracks propagating by mixed-mode I + II fracture [13]. 52 When the interlayer is used as a bonding agent, the behaviour of glass composite systems strongly depend on the 53 loading and temperature history due to the viscoelastic behaviour of the interlayer, even when using the stiffest 54 interlayers (e.g. SentryGlas Plus – SGP) [28], as well as on the ratio between the bond perimeter and cross-section 55 area of the reinforcement [25]. Therefore, glass composite systems may not show sufficient load-bearing capacity 56 under certain environmental conditions (e.g. high temperature). Greater reinforcement ratios may be required when 57 using structural adhesives, given the low composite action provided by the interlayer. 58 On the other hand, the post-tensioning of annealed glass using traditional reinforcement materials vs. 59 prestressing methodologies is still a major challenge, since glass may fail due to the stress concentrations induced 60 near the loaded end section after applying the prestress [29]. Its time-dependent tensile strength and inherent 61 variability require the smooth transfer of the prestressing force to the glass substrate. Two strategies have been so 62 far successful at increasing the allowable prestressing in concrete: (i) fixing the ends of the prestressed 63 reinforcement, e.g. using end-anchorages (e.g. [30,31]) and (ii) increasing the reinforcement bond perimeter by 64 adopting the NSM technique. 65 Studies comparing the EBR and NSM techniques have shown that the latter is better at preventing the 66 premature debonding due to the larger bonding surface between adherends, as well as the confinement effect 67 provided by the grooves [32,33]. Furthermore, the NSM technique is also better at protecting the reinforcement 68 material against corrosion, fire, vandalism, mechanical damage and aging [34]. Therefore, the efficiency of the 69 FRP strengthening systems in flexural members is related to the NSM groove size [33]. Unlike the reinforcement 70 lamination, the NSM technique allows structural adhesives to be chosen based on the materials involved and the 71 exposure environment (e.g. temperature range), as well as their ability to promote the composite action. While the 72 interlayer is obviously unable to transfer the prestressing force due to its viscoelastic properties, and the EBR 73 technique promotes excessive stress concentrations in the glass substrate, the NSM seems promising for post74 tensioning of glass members. However few studies (e.g. [35]) have been found in the literature addressing this 75 topic. 76 Based on the redundancy principle behind laminated glass, the benefits of strengthening concrete using 77 hybrid systems, combining the NSM and EBR techniques, have been addressed by some studies (e.g. [36–38]). 78 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. These have shown that the structural element can still withstand a significant flexural load after debonding of the 79 externally bonded reinforcement. 80 1.2. CFRP and SMA in glass industry 81 As a result of its wide application for strengthening existing concrete structures [39], CFRP has been 82 recently used in investigations on glass composite systems, as its high resistance and stiffness maximize the 83 transparency of glass structures by reducing the reinforcement ratio required to reach a certain load-carrying 84 capacity. However, few experimental (e.g. [7,8,35,40]) and numerical (e.g. [7,41]) studies focusing on the 85 behaviour of CFRP-reinforced glass structural elements are found in the literature. CFRP has been applied 86 according to the EBR technique (e.g. [27,40]), embedded into the interlayer (e.g. [7]) and introduced inside 87 recessed grooves (e.g. [35]). The feasibility of prestressing CFRP has also been addressed (e.g. [8]). 88 In contrast to the traditional prestressing methodologies using FRPs, the activation of SMAs (Shape 89 Memory Alloys) prevents stress concentrations in the glass substrate [42]. Nitinol (Ni-Ti) is the most popular 90 SMA, but iron-based (Fe-) SMAs show better properties for the construction industry, such as (i) lower cost, 91 (ii) easier manufacturing process, (iii) higher modulus of elasticity, and (iv) lower activation temperature [43]. 92 SMAs exhibit shape memory effect, i.e. the ability to return to its initial shape through heating after mechanically 93 induced deformation. Thereby, these have been used for post-tensioning of existing structural elements (e.g. [44– 94 49]). Under certain temperature conditions, SMAs can also exhibit superelasticity, i.e. the ability to fully recover 95 its initial shape just by unloading. The Fe-17Mn-5Si-10Cr-4Ni-1(V, C) alloy is a superior Fe-SMA developed by 96 Dong et al. [50] for post-tensioning existing structures, such as concrete (e.g. [46,49,51–53]) and steel (e.g. 97 [44,45,54]). It can be produced at atmospheric conditions, not requiring expensive high-vacuum processing 98 facilities or thermo-mechanical training [47,55]. 99 Currently, few studies addressing the performance of SMA-reinforced glass systems are found in the 100 literature, probably due to the novelty of both materials in the construction industry. Such studies have focused on 101 the feasibility of activating externally bonded Fe-SMA strips (e.g. [23,42]), the bond behaviour of glass-to-SMA 102 joints (e.g. [56,57]), as well as exploratory numerical studies (e.g. [58]). 103 1.3. Research significance 104 The long-term evolution of the tensile strength of glass structures, their lack of ductility and their 105 insufficient reliability still raises concerns. The premature failure of glass composite systems due to unrestrained 106 opening of shear cracks, as well as the deleterious effect of stress concentrations induced by different strengthening 107 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. systems including the mechanical post-tensioning, are limiting factors for a more widespread application of 108 structural glass. The studies focusing on the feasibility of strengthening glass using CFRP and Fe-SMA 109 reinforcements are scarce, and given its promising features this research is aimed at investigating the post110 tensioning of laminated glass beams by prestressing CFRP and/or activating Fe-SMA. A total of 5 laminated glass 111 beams were tested adopting a four-point bending configuration. All specimens were strengthened by combining 112 the NSM and EBR techniques, using one or both reinforcement materials. Different strengthening layouts (i.e. 113 reinforcement material versus application technique) were also adopted in each specimen, in order to maximize 114 the allowable post-tensioning level and avoid premature glass breakage during the prestress application, while 115 improving the post-cracking performance. Two epoxy adhesives were used to join the components. The finite 116 elements software ABAQUS 6.14 [59] was used to numerically simulate the activation of the Fe-SMA strips. 117 Digital Image Correlation (DIC) method was used to support the analysis of results obtained from flexural tests, 118 therefore contributing to extend the current knowledge on structural glass strengthening using hybrid approaches. 119 2. MATERIALS AND SPECIMENS 120 2.1. Mechanical characterization 121 2.1.1. Annealed glass 122 This work aims to investigate the feasibility of mechanically post-tensioning the annealed glass as an 123 alternative to the thermal toughening to increase its initial fracture strength, but without compromising its post124 failure performance. Thus, annealed glass layers were used to manufacture tailored laminated glass panels. The 125 modulus of elasticity (Eg) and tensile strength (fg,t) of the annealed glass indicated in Table 1 were previously 126 assessed by Rocha et al. [5], through inverse analysis of the structural responses obtained from tensile tests on 127 double-lap joint specimens. 128 2.1.2. CFRP laminate 129 The CFRP laminates used in this study, with a cross section of 20  1.2 [mm], were produced by S&P® 130 Clever Reinforcement Company by pultruding unidirectional carbon fibres with a vinyl ester resin matrix. The 131 mechanical properties of the CFRP laminates were characterized according to ISO 527-5:2009 [60], by testing 132 five specimens of 250  20  1.2 [mm] (length  width  thickness) in tension at ambient temperature and at a 133 constant tensile displacement rate of 1.0 mm/min until failure. The longitudinal deformation of the specimens was 134 measured installing a clip gauge (type: MFA 12; linearity: 0.1 %; sensitivity: 2.0 mV/V; resolution: 1.0 pm; 135 precision: ±1.5 μm) with stroke of 50 mm at the central region of each specimen, while the applied load was 136 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. registered using a load cell with a maximum capacity of 200 kN and precision of 0.01 kN. Subsequently, the 137 modulus of elasticity was determined from the slope of the linear trend line of the stress-strain response between 138 strain values of 0.05 % and 0.25 % [60]. Table 1 presents the average properties of the CFRP laminate, namely 139 the modulus of elasticity (Er), tensile strength (fr,t) and ultimate strain (εr,ult). 140 2.1.3. Fe-SMA strip 141 The Fe-17Mn-5Si-10Cr-4Ni-1(V,C) (mass%) alloy is commercialized by re-fer AG Company and its 142 production process is detailed by Leinenbach et al. [43]. The tensile behaviour of the Fe-SMA was previously 143 characterized by Rocha et al. [42], including the average values of the modulus of elasticity (Er), tensile 144 strength (fr,t) and ultimate strain (εr,ult), as presented in Table 1. 145 2.1.4. Adhesives 146 Based on previous studies on glass composite systems with CFRP [27] and Fe-SMA [42], the two147 component epoxy adhesives SikaDur®-330 and 3M Scotch-Weld DP490 – later called as SD and 3M – were 148 adopted to bond CFRP and Fe-SMA to glass, respectively. Compared to 3M adhesive, which must be used when 149 the adhesive layer is thinner than 0.3 mm, according to the supplier, the SD adhesive can be used to produce 150 adhesive layers up to 4 times thicker. The long-term performance of both adhesives was not considered in this 151 study. Table 1 presents the modulus of elasticity (Eadh), tensile strength (fadh,t) and ultimate strain (εadh,ult) of each 152 adhesive, which were obtained by Rocha et al. [5] from material characterization according to EN ISO 527-2:2012 153 [61]. In contrast to the SD adhesive, which exhibits linear elastic behaviour until failure, 3M adhesive shows a 154 slight loss of stiffness with increasing load. 155 2.2. Beam geometry 156 This study is the last phase of a wider research project aimed at improving the flexural behaviour of 157 composite glass systems using CFRP and Fe-SMA as reinforcement. Monolithic glass beams with externally 158 bonded CFRP (SDur series) and Fe-SMA (P_T120 series) reinforcements were previously tested by Rocha 159 et al. [27] and Rocha et al. [42], respectively. Their characteristics are detailed in Table 2, such as the beam 160 geometry, materials (adhesive and reinforcement) and reinforcement ratio (ρr). Excluding the beams unloaded 161 before the collapse due to excessive deformation, all the others failed due to premature debonding of the 162 reinforcement when a critical shear crack appeared near the supports, as typically observed when EBR systems 163 are used. 164 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. Glass structures are typically very thin (slender), which poses challenges for inserting the reinforcement 165 inside the glass panel. Although EBR systems fail prematurely. EBR reinforcement can be useful in preventing 166 the growth of existing flaws in glass tensile zones due to unexpected actions (e.g. vandalism) and aggressive 167 environmental exposures. Therefore, the specimens were strengthened with two reinforcement elements. One of 168 them was introduced inside the laminated glass panel, according to the NSM technique, while the other was 169 externally bonded to the bottom glass edge. 170 The laminated glass panels were manufactured according to the geometry shown in Fig. 1. All panels 171 consisted of joining three layers of annealed glass with polished edges using two polyvinyl butyral (PVB) 172 interlayers with a thickness of 0.76 mm. The difference in height between the two outer layers – 220 (height)  173 10 (thickness) [mm] – and the inner layer – 198 (height)  3 (thickness) [mm] – created a groove of 22 (depth)  174 4.5 (thickness) [mm] for subsequent insertion of the NSM reinforcement. 175 As SMA activation is better at smoothing stress concentrations and CFRP is better at providing post176 cracking stiffness, different strengthening systems were adopted in terms of reinforcement material versus 177 application technique, as summarized in Table 3. A total of 5 specimens were produced and tested using the 178 nomenclature n-i-j, where i and j identify the reinforcement material applied according to the NSM and EBR 179 techniques, respectively. The parameter n was adopted to distinguish specimens strengthened with NSMand 180 EBR-CFRP laminates, representing the prefixes “R” or “P” the cases where passive or prestressed reinforcement 181 were applied, respectively. 182 3. TEST METHODS 183 3.1. Beam fabrication 184 All composite glass beams were reinforced according to Fig. 1c, by bonding CFRP laminates – 185 20 (width)  1.2 (thickness) [mm] – and/or Fe-SMA strips – 20 (width)  1.5 (thickness) [mm] – using both 186 adhesives mentioned in Section 2.1.4. Small PVC spacers, occupying less than 1.0 % of the bond surface area, 187 were used to ensure the symmetrical positioning of the NSM reinforcement inside the longitudinal groove. 188 Considering the manufactures’ suggestions, the EBR reinforcement was bonded to the glass adopting adhesive 189 layers with a thickness of 0.9 mm and 0.3 mm for the SD and 3M adhesives, respectively. Table 3 describes the 190 main characteristics of the beams tested in this research, including the adhesives used for each strengthening 191 system and the respective layer thickness. 192 The specimens were prepared according to the following general procedure: 193 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. i. Cutting of the reinforcement elements and manufacturing of the laminated glass panels; 194 ii. Prestressing of the CFRP laminate against the reaction frame versus pre-straining the Fe-SMA 195 strip (further details in Sections 3.1.1 and 3.1.2, respectively); 196 iii. Bonding of the reinforcement to the glass. First, the bonding surfaces were cleaned with acetone. 197 Then, the adhesive was prepared and applied according to the manufacturer’s specifications. 198 Afterwards, the components were carefully assembled. Finally, in order to obtain a further cure of 199 the adhesives and to avoid effects of different ambient temperature among the tested beams during 200 curing, all specimens were subjected to post-curing conditions before releasing the CFRP laminate 201 or activating the Fe-SMA strip. The post-curing was achieved by using heating fans aimed at the 202 specimens (zones in the vicinity of the loaded end section), producing the rise of the temperature 203 measured in the specimens to approximately 45 ºC for 2 hours (see Fig. 2a); 204 iv. Releasing of the CFRP laminate by removing the prestressing hydraulic jacks versus activating 205 the Fe-SMA strip by resistive heating. This step took place at least 96 hours after bonding to 206 guarantee sufficient adhesive toughness to transfer the post-tensioning force from the 207 reinforcement to the glass. 208 In the specimens R-C-C and P-C-C, the two reinforcement elements were simultaneously bonded to the 209 glass, since only the SD adhesive was used for this purpose. Regarding the R-C-C beam, it was prepared by 210 executing only the first and third steps. For the remaining specimens, each reinforcement element was bonded 211 individually and in a separate stage, either because different adhesives were used to bond each reinforcement, or 212 because in the case of the Fe-SMA strips these were not entirely activated, as further explained in Section 3.1.2. 213 Therefore, all specimens strengthened with Fe-SMA were manufactured by applying twice the general procedure 214 described above, for each of the reinforcement elements individually. 215 3.1.1. CFRP prestressing 216 Glass edges show lower apparent tensile strength than glass surfaces, because the former contain deeper 217 surface flaws induced during the production, cutting, polishing and handling operations [62,63]. In addition, it is 218 reasonable to assume that the glass corners show an even lower tensile strength, as they are typically the most 219 unprotected zones when glass pieces are handled. As a result, preliminary studies (e.g. [29]) have shown that 220 prestressing externally bonded reinforcement is often unsuccessful because the glass fails due to the stress 221 concentration at the loaded end sections. Accordingly, in this case, only the NSM-CFRP laminates were 222 prestressed. 223 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. After being mechanically anchored at both ends using metal clamps (see Fig. 2b), the NSM-CFRP 224 laminates were prestressed up to an average strain of 2.0 ‰ (σ ≈ 367.7 MPa) using a hydraulic jack. The axial 225 strain was recorded by means of a strain gauge (type: PFL-10-11-3LJC-F by TML; measuring length: 10 mm; 226 gauge factor: 2.12 ± 1 %)) previously installed in the middle of the CFRP laminate. The prestressing level was 227 continuously monitored until the NSM-CFRP laminate was released. 228 3.1.2. Fe-SMA activation 229 Fe-SMAs have two distinct crystal structures, called (i) austenite phase, which is stable at higher 230 temperatures, and (ii) martensite phase, which is stable at lower temperatures. The martensitic transformation 231 consists of modifying the lattice from austenite to detwinned martensite through mechanical deformation or 232 temperature variation. As schematized in Fig. 3, the martensitic transformation does not involve any slippage 233 between atoms (neighbors remain neighbors). It takes place at temperatures between Ms (martensite start 234 temperature) and As (austenite start temperature). When the Fe-SMA is heated at temperatures above As, the 235 detwinned martensite is reversed to austenite and its initial shape is retrieved. 236 Post-tensioning with Fe-SMAs involves three phases: (i) pre-straining, (ii) activation and (iii) service 237 loading (see Fig. 3). First, the Fe-SMA must be mechanically loaded at room temperature (Ms < T < As). Once 238 unloaded, it shows a permanent deformation that can be partially recovered through heating (T > As). Thereby, 239 post-tensioning (recovery stress) is generated when the Fe-SMA is adhesively bonded and/or mechanically 240 anchored to the target element prior to heating it. Activation is complete when the Fe-SMA reaches the room 241 temperature again. 242 According to the technical specifications provided by the producer, Fe-SMA strips were pre-strained to 243 2.0 % at room temperature [47]. A hydraulic jack was used to apply the load and a clip gauge (technical 244 specifications described in Section 2.1.1) to measure the longitudinal deformation. 245 Adhesive damage is an inevitable consequence of activating the Fe-SMA reinforcement. However, this 246 apparently deleterious effect can be used to prevent stress concentrations in the glass substrate. A favourable 247 damage gradient is produced along the adhesive connection, on both sides of the activated Fe-SMA strip zone, 248 which helps to smooth the stress transfer between adherends. Desirably, the adhesive bond regions closest the 249 beam ends should remain undamaged to efficiently transfer the post-tensioning force between the adherends. 250 Following the recommendations proposed by Rocha et al. [42], the Fe-SMA strips were heated symmetrically with 251 respect to the mid-span section and the activated length (la) was set to 1400 mm, as schematized in Fig. 4a. 252 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. NSM technique, even designing wider grooves to smooth stress concentrations in the glass substrate or adopting 420 adhesives with low viscosity (e.g. acrylic adhesive) to join the components. Regarding the latter requirement, it 421 should be noted that the initial pre-stress reduces as the reinforcement ratio increases. When reinforcement 422 materials such as steel are used to apply prestress, its tensile strength reserve after post-tensioning must be large 423 enough to ensure Fult > Fcr. Given the above, reinforcement ratios much higher (ρr > 2.8 %) than those adopted in 424 this study are generally found in the literature (e.g. [13,15,20]), while seeking to limit the initial pre-stress in steel 425 reinforcement to 50 % of its yield strength. In the case of stronger reinforcement materials (e.g. CFRP), increasing 426 the bond surface area seems to be the best strategy to exploit the full tensile capacity of such materials. 427 5.2. After first cracking 428 5.2.1. Strengthening system 429 In comparison to the monolithic beams previously tested (SDur series presented in Table 8), the R-C-C 430 beam exhibited a much better post-cracking performance. In spite of a slight reduction in the reinforcement ratio 431 from 1.17 %, in the SDur series, to 0.96 %, in the R-C-C beam, the RSi increased from 87 % to 160.8 %, 432 respectively. Such results indicate that the post-cracking performance of glass composite systems can be 433 significantly enhanced when all or part of the tensile reinforcement is applied according to the NSM technique. 434 Therefore, the hybrid strengthening systems adopted in this study are clearly more efficient than EBR systems in 435 delaying premature debonding of the reinforcement after the formation of large shear cracks. Unlike the SDur 436 series, which presented asymmetric crack propagation governed mainly by dynamic effects arising from glass 437 cracking (see Fig. 14b), the R-C-C beam showed a uniform and dense crack pattern, with vertical cracks in the 438 pure bending zone and increasingly inclined shear cracks towards the supports. As a result, Di increased from 439 407 %, in the SDur series, to 971 %, in the R-C-C beam (~ 2.5 times higher). Finally, taking the SDur series as a 440 reference, εr (Fult) was 38.1 % higher in the case of the R-C-C beam (4.45 ‰). Thus, hybrid strengthening systems 441 are better at exploiting the tensile capacity of the CFRP, allowing to reduce the reinforcement percentage and, 442 simultaneously, to obtain better flexural responses in comparison to the EBR systems. 443 On the other hand, although the SMA_SMA beam did not show the ability to exceed Fcr after crack 444 initiation, as opposed to the P_T120 series, some improvements in the post-cracking behaviour were observed. 445 Even slightly, the tensile capacity of Fe-SMA was better exploited in the SMA_SMA beam (εr (Fult) = 48.7 ‰) 446 than in the P_T120 series (εr (Fult) = 40.3 ‰). Moreover, unlike the monolithic beams, no progressive debonding 447 of the SMA reinforcement was observed in the SMA_SMA beam. 448 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. The hybrid strengthening system adopted in this study showed to be able to avoid premature debonding of 449 the reinforcement before the beam collapse. Moreover, in the case of premature debonding of the EBR 450 reinforcement, it is reasonable to assume that the NSM reinforcement can still transfer load to the supports, 451 generating a sequential failure process, as observed in previous studies on hybrid strengthening systems for 452 concrete, and preventing the sudden failure of the glass structural element. Due to the brittleness of annealed glass 453 and its unreliable tensile strength, redundancy at different levels is required to fulfil the ductility and safety 454 requirements. 455 5.2.2. Load bearing capacity 456 The beams R-C-C and P-C-C showed similar flexural responses during the post-cracking stage. For similar 457 mid-span deflections in both beams, the prestressing of the NSM-CFRP laminate resulted in a shift of the F – δ 458 response upwards, to higher load levels. This increase of load carrying capacity is equal to the additional bending 459 moment generated by the eccentricity between the post-tensioning force and the neutral axis (cracked section 460 analysis). However, it should be noted that the prestressing of the NSM-CFRP laminate reduced the RSi from 461 971 % (P-C-C beam) to 675 % (R-C-C beam), as well as the respective Di from 161 % to 146 %. Neglecting the 462 tensile pre-stress applied at the top edge of glass by post-tensioning, the Fult depends only on the material 463 properties. Thereby, the higher the post-tensioning level, the higher the Fcr and, consequently, the lower the RSi 464 and Di values. Hence, the R-C-C beam (passive reinforcement) exhibited the highest RSi and Di among the 465 specimens tested. 466 After glass cracking, the flexural stiffness of the CFRP_SMA beam gradually decreased due to the 467 progression of cracks towards the supports. At the end of the post-cracking stage, its F – δ response was 468 unexpectedly similar to that of the P-C-C beam. The main reason for this relies on the non-linear behaviour of the 469 Fe-SMA in tension. As the CFRP presents linear elastic behaviour until failure, when it is prestressed, the 470 increment of loading carrying capacity is approximately constant throughout the post-cracking stage (e.g. P-C-C 471 versus R-C-C). As the Fe-SMA exhibits plastic behaviour, the post-tensioning effect (shifting of the F – δ response 472 upwards) seems to have been lost when the yield strength was attained. Accordingly, the neutral axis gradually 473 moved upwards, towards the top edge of the glass, due to the yielding of the Fe-SMA, thus increasing the 474 compression stress in the upper uncracked zone. Hence, the Fult (CFRP_SMA beam) was 6.5 % lower than Fult (P475 C-C beam). 476 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. A series of consecutive load drops occurred in the SMA_CFRP beam for 9.5 mm < δ < 13.0 mm (see 477 Fig. 11), and two phases may be distinguished in the post-cracking response. During the first phase, with 478 δ < 9.5 mm, flexural cracks appeared between the load points. Then, for 9.5 mm < δ < 13.0 mm, increasingly 479 inclined shear cracks formed suddenly only towards one of the supports. Minor deviations in the cross-section 480 geometry of the laminated glass panel seem to the cause of this asymmetric behaviour, since the beam height 481 ranged between 219.2 mm and 222.6 mm. As expected, new cracks appeared first towards the stiffer support 482 section (monitored by the DIC technique). Nevertheless, this difference in beam height is not large enough to 483 create such a discrepancy between the crack patterns in both shear spans (see Fig. 11c at δ = δult / 2). The low 484 tensile stiffness of Fe-SMA after activation made it unable to restrain the crack opening, promoting damage 485 concentration and, as a consequence, high interfacial stress (mixed mode-I+II fracture) between the CFRP 486 reinforcement and the glass substrate. This seems to have resulted in the growth of existing surface flaws at the 487 bottom edge of glass, reducing the required tensile stress to open new cracks, why explains the large number of 488 cracks appearing for 9.5 mm < δ < 13.0 mm. Then, for δ > 13.0 mm, additional shear cracks formed in the non489 activation region and, consequently, the strengthening system shifted to a passive-like behaviour (no post490 tensioning effect), which explains the similarity between the F – δ responses of the beams SMA_CFRP and R-C491 C at this stage. This is probably the cause why Therefore, stiffer materials should be used as NSM reinforcement 492 to As a result, the SMA_CFRP beam showed the lowest Fmax among the tested specimens. Strengthening systems 493 show great difficulties in restraining crack opening when NSM systems are not adopted (e.g. monolithic glass 494 beams) or flexible materials are used as reinforcement (e.g. Fe-SMA after activation). Hence, 495 Concerning the SMA_SMA beam, it did not achieve the Fcr again during the post-cracking stage 496 (RSi < 100 %), in contrast to other specimens. Considering the Fcr,a (fg,eff) indicated in Table 9, the RSi of the 497 SMA_SMA beam would be equal to 116.5 % (Fult vs. Fcr,a ratio), which, in any case, would still be the lowest 498 value among the five tested specimens. No shear cracks appeared in the SMA_SMA beam. V-shaped cracks 499 prevailed in the pure bending zone (see Fig. 12b), unlike crack patterns observed in the other specimens. Both 500 aspects emphasized above are a result of the mechanical behaviour of the Fe-SMA. The higher the activation 501 temperature, the higher the recovery stress and the lower the tensile strength reserve of the Fe-SMA before the 502 yielding (martensitic transformation). When the first crack appeared, the tensile stress in the reinforcement 503 suddenly increased, resulting in yielding in the Fe-SMA and significantly reducing the flexural stiffness of the 504 SMA_SMA beam. This sudden loss of tensile stiffness resulted in large crack opening and extensive horizontal 505 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. crack propagation (crack branching), as observed in Fig. 12. Thereafter, the further yielding of the Fe-SMA 506 prevented crack propagation towards the supports. 507 In general, the post-tensioning resulted in obvious improvements in the overall structural response of 508 composite glass beams, both before and after crack initiation. Some aspects observed in this investigation are 509 highlighted below: (i) unlike the Fe-SMA, the CFRP reinforcement can be entirely prestressed and it does not 510 damage the adhesive connection; in addition, it prevents significant load drops when the cracking progresses 511 towards the non-activated regions (e.g. SMA_CFRP beam); (ii) heating the Fe-SMA smoothens the post512 tensioning force transfer from the reinforcement element to the glass, reducing the risk of peeling-off failure of 513 the strengthening system and allowing the post-tensioning of externally bonded reinforcement; (iii) a significant 514 tensile strength reserve before the yielding must be ensured when Fe-SMA reinforcement is activated, to create 515 sufficiently stiff post-cracking responses (e.g. SMA_SMA beam); and (iv) the post-tensioning level must be 516 defined by balancing the glass fracture strength and the post-cracking safety, ensuring in any case a residual 517 strength capacity above 100 %. 518 Based on the referred requirements, the best strengthening system seems to be the one adopted in the 519 CFRP_SMA beam. As experimentally observed, premature debonding of the reinforcement is much less likely 520 when NSM systems are adopted. Accordingly, the stiffer reinforcement material must be introduced into the 521 groove in order to prevent high interfacial stresses at the glass substrate, as well as to ensure sufficient post522 cracking stiffness, even in the case of premature debonding of the externally bonded reinforcement. When the 523 NSM-CFRP composite systems are prestressed, a favourable compression pre-stress can be applied along the 524 entire beam length in glass tensile zones. On the other hand, externally bonded Fe-SMA reinforcement can be 525 safely activated because the inevitable adhesive damage generates a damage gradient in the stress transfer zone 526 and prevents high stress concentrations at the glass substrate. 527 5.2.3. Failure modes 528 All specimens ruptured by shear-compression failure and showed the typical glass crushing in the 529 compression zone (see Fig. 15). According to Table 7, the εg,t (Fult) varied between 1.15 ‰ (R-C-C beam) and 530 2.65 ‰ (SMA_SMA beam). Despite this difference (2.3 times larger), all beams presented almost similar Fult, 531 which varied between 17.65 kN (in the SMA_CFRP beam) and 20.27 kN (in the P-C-C beam). The scatter in 532 εg,t (Fult) is probably related to the characteristics of each strengthening system (e.g. tensile stiffness and post533 tensioning level). The stiffer the strengthening system, the deeper the neutral axis and the lower the εg,t (Fult). As 534 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. the post-tensioning induced favourable tensile pre-stresses in the upper glass zone, the stiffer the strengthening 535 system and the higher the post-tensioning level, the higher the Fult should be, in line with the experimental 536 observations. 537 Excluding the SMA_SMA beam, where the compression failure by crushing occurred approximately at 538 mid-span section (see Fig. 15a), in the other specimens glass crushing was observed in the vicinity of the load 539 point sections. In the SMA_SMA beam, it is clear that the height of the compression glass zone progressively 540 decreased during the post-cracking stage due to the yielding of the Fe-SMA. In the remaining specimens, the 541 deformation was a result of the progressive crack propagation towards the supports. Without transverse 542 reinforcement, the shear cracks inevitably propagated towards the load point sections (see Fig. 12), triggering 543 shear-compression failure in glass. These shear cracks induced shear and compression stresses around the load 544 application zones and, when the maximum principal stress reached the compression strength of glass (fc,g), glass 545 crushing occurred at the tip of critical shear cracks. As a result, these shear cracks penetrated into the glass 546 compression zone and the glass zones around both load points shattered. As noticed in other studies on glass 547 composite systems (e.g. [65]), lateral-torsional buckling may have anticipated the beam failure by reducing the 548 Fult values. 549 Mixed shear-compression failure has been observed in reinforced concrete (e.g. [66]) and glass (e.g. [67]). 550 However, these failure modes have been rarely reported in previous studies, where glass composite systems often 551 fail due to premature debonding and tensile rupture of the reinforcement. Based on observations from previous 552 studies (e.g. [8,13,20]) as well as on the analytical models developed by Louter [68], the reasons why glass 553 crushing occurred in all tested beams are related to (i) the low reinforcement ratio, as well as (ii) the low interlayer 554 stiffness and thickness. The neutral axis moves away from the top edge of glass as the reinforcement ratio 555 increases, thus delaying the beam failure by glass crushing or lateral-torsional buckling. While the ρr adopted in 556 this study between 0.96 % and 1.20 % (see Table 3), values above 2.80 % are generally found in the literature. 557 Unlike these previous studies, the laminated glass panels used in this investigation were manufactured using PVB 558 as interlayer. As SGP is much stiffer than PVB within the temperature range registered during the test, the former 559 would be more effective in moving the neutral axis away from the top edge of the glass. Such effect has been 560 observed in some studies (e.g. [13]). Furthermore, while a 0.76 mm thick PVB interlayer was used to produce the 561 laminated glass panels, 1.52 mm thick SGP interlayers have been adopted in previous studies. The thicker the 562 interlayer, the higher its contribution for the resistant mechanism formed after glass breakage. 563 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. As referred in Section 5.2.2, no shear cracks appeared before the collapse of the SMA_SMA beam. 564 Therefore, neglecting any buckling effects on the Fult, it is reasonable to assume that glass crushing occurred only 565 when the fc,g was achieved at the top edge of glass. This explains why a much higher εg,t (Fult) was recorded in the 566 SMA_SMA beam compared to other specimens. The compression strength of glass is generally much higher than 567 its tensile strength [2]. Although DIN 1249-10:1990 [69] indicates values between 700 MPa and 900 MPa, 568 Campione et al. [70] found values of ~ 200 MPa, in line with the experimentally measured εg,t (σg,t = 196.1 MPa) 569 at failure. 570 6. CONCLUSIONS 571 This research work focused on the structural performance of laminated glass beams post-tensioned with 572 Fe-SMA and CFRP reinforcements, both applied according to the NSM and EBR techniques. First, laminated 573 glass panels were designed to create a groove on the tensile zone. Then, both reinforcement elements were 574 adhesively bonded to the tensile zone of the glass panels using two epoxy adhesives. After that, Fe-SMA and 575 CFRP reinforcements were activated and released, respectively, inducing compressive stress in the tensile glass 576 zone due to the prestress action. Finally, all specimens were tested under a four-point bending configuration. In 577 addition, numerical investigations were performed to determine the recovery stresses in the Fe-SMA strips from 578 the strain gauge measurements. 579 The main conclusions are summarized as follows: 580 • All specimens were able to maintain their integrity after crack initiation, then exhibiting a pseudo-ductile 581 behaviour until failure. Relatively safe and ductile failure mechanisms were observed in all beams; 582 • Post-tensioning increased the initial fracture stress of glass between 22.7 and 94.7 %, showing the 583 influence of the strengthening layout on the post-tensioning level; 584 • Compared to the EBR systems, the hybrid strengthening systems resulted in an increase in residual 585 strength, from 87 to 160.8 %, and in ductility, from 407 to 971 %, even with a slightly reduction of the 586 reinforcement ratio from 1.17 to 0.96 %; 587 • When using the hybrid strengthening system, the axial strain in the reinforcement at failure was at least 588 38.1 % higher than in the SDur series and 20.8 % higher than in the P_T120 series, and the specimens 589 failed before reinforcement rupture or debonding; 590 • Specimens reinforced with Fe-SMA presented lower post-cracking flexural stiffness than the beams R591 C-C and P-C-C, since the higher the activation temperature adopted, the lower the tensile strength reserve 592 before Fe-SMA yields; 593 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. • The main advantage of using Fe-SMA reinforcement is to improve the pre-cracking behaviour of glass 594 composite systems, i.e. increasing the initial fracture stress as much as possible, while the CFRP 595 reinforcement is more efficient at covering the post-cracking design requirements; 596 • Due to the absence of transverse reinforcement in glass, shear cracks suddenly propagated towards the 597 loading sections, inducing high shear and compression stresses that crushed the glass above the tip of the 598 critical shear crack; 599 • SGP interlayer should be preferred to join glass layers, since it is more effective than PVB in providing 600 load-carrying capacity after glass cracking and in delaying glass crushing. 601 The results obtained from this exploratory study show the advantages of post-tensioning laminated glass 602 beams using CFRP and Fe-SMA, both before and after crack initiation. Further studies are required to investigate 603 aspects such as (i) the influence of the reinforcement ratio on the failure mechanisms of composite beams, as well 604 as the ratio between both reinforcement materials; (ii) the influence of the tensile strength reserve of the Fe-SMA 605 reinforcement after activation; (iii) the long-term behaviour of the post-tensioning force transfer zones; and (iv) 606 new strategies to increase the prestressing level without causing the premature peeling-off failure of the CFRP 607 reinforcement, in order to better exploit the high tensile strength of this material. 608 ACKNOWLEGMENTS 609 The first author wishes to acknowledge the grant SFRH/BD/122428/2016 provided by Fundação para a 610 Ciência e a Tecnologia, IP (FCT), financed by European Social Fund and by national funds through the 611 FCT/MCTES. This work was partly financed by FCT / MCTES through national funds (PIDDAC) under the R&D 612 Unit Institute for Sustainability and Innovation in Structural Engineering (ISISE), under reference 613 UIDB/04029/2020, and under the Associate Laboratory Advanced Production and Intelligent Systems ARISE 614 under reference LA/P/0112/2020. Finally, the authors also like to thank the COVIPOR – Companhia Vidreira do 615 Porto Lda., S&P Clever Reinforcement Iberica Lda., Sika and re-fer AG for supplying the materials. 616 REFERENCES 617 [1] CNR-DT 2010/2013. Guide for the Design, Construction and Control of Buildings with Structural Glass 618 Elements. CNR - Advis. Comm. Tech. Recomm. Constr., Rome: National Research Council of Italy; 2013. 619 [2] Haldimann M, Luible A, Overend M. Structural use of glass. LABSE - Lanka Association of Building 620 Services Engineers; 2008. 621 [3] Furtak K, Rodacki K. Experimental investigations of load-bearing capacity of composite timber-glass I622 beams. Arch Civ Mech Eng 2018;18:956–64. https://doi.org/10.1016/j.acme.2018.02.002. 623 [4] Kozłowski M, Serrano E, B. E. Experimental investigation on timber-glass composite I-beams. 624 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. Challenging Glas. 4 COST Action TU0905 Final Conf., Taylor & Francis Group; 2014, p. 261–8. 625 [5] Rocha J, Sena-Cruz J, Pereira E. Tensile behaviour of CFRP-glass adhesively bonded connections: double626 lap joint tests and numerical modelling. Eng Struct 2022;260:114212. 627 https://doi.org/10.1016/j.engstruct.2022.114212. 628 [6] Cagnacci E, Orlando M, Salvatori L, Spinelli P. Four-point bending tests on laminated glass beams 629 reinforced with FRP bars adhesively bonded to the glass. Glas Struct Eng 2021;6:211–32. 630 https://doi.org/10.1007/s40940-021-00147-9. 631 [7] Bedon C, Louter C. Structural glass beams with embedded GFRP, CFRP or steel reinforcement rods: 632 Comparative experimental, analytical and numerical investigations. J Build Eng 2019;22:227–41. 633 https://doi.org/10.1016/j.jobe.2018.12.008. 634 [8] Louter C, Cupać J, Debonnaire M. Structural glass beams prestressed by externally bonded tendons. Glas. 635 Glob. Conf. Proc., Philadelphia, EUA: 2014, p. 450–9. https://doi.org/10.14296/deeslr.v5i0.1848. 636 [9] Neto P, Alfaiate J, Valarinho L, Correia J, Branco F, Vinagre J. Glass beams reinforced with GFRP 637 laminates: Experimental tests and numerical modelling using a discrete strong discontinuity approach. Eng 638 Struct 2015;99:253–63. https://doi.org/10.1016/j.engstruct.2015.04.002. 639 [10] Achintha M, Balan B. Characterisation of the mechanical behaviour of annealed glass – GFRP hybrid 640 beams. Constr Build Mater 2017;147:174–84. https://doi.org/10.1016/j.conbuildmat.2017.04.086. 641 [11] Achintha M, Balan B. Mechanical prestressing of annealed glass beams using pretensioned GFRP: 642 Characterisation and potentiality. Structures 2019;20:11–9. https://doi.org/10.1016/j.istruc.2019.02.017. 643 [12] Rocha J, Pereira E, Sena-cruz J. Alternative Approaches for the Numerical Simulation of Glass Structural 644 Beams Reinforced with GFRP Laminates. In: Louter, Bos, Belis, Veer, Nijsse, editors. Challenging Glas. 645 6 - Conf. Archit. Struct. Appl. Glas., vol. 14, Delft: 2018. 646 [13] Cupać J, Louter C, Nussbaumer A. Flexural behaviour of post-tensioned glass beams: Experimental and 647 analytical study of three beam typologies. Compos Struct 2021;255. 648 https://doi.org/10.1016/j.compstruct.2020.112971. 649 [14] Bedon C, Louter C. Finite Element analysis of post-tensioned SG-laminated glass beams with adhesively 650 bonded steel tendons. Compos Struct 2017;167:238–50. 651 https://doi.org/10.1016/j.compstruct.2017.01.086. 652 [15] Cupać J, Martens K, Nussbaumer A, Belis J, Louter C. Experimental investigation of multi-span post653 tensioned glass beams. Glas Struct Eng 2017;2:3–15. https://doi.org/10.1007/s40940-017-0038-5. 654 [16] Firmo F, Jordão S, Costa Neves L, Bedon C. The effect of adhesive joints on the performance of hybrid 655 steel-glass beams – An analytical and experimental study. Challenging Glas Conf Proc - Challenging Glas 656 5 Conf Archit Struct Appl Glas CGC 2016 2016:171–86. 657 [17] Louter C, Belis J, Veer F, Lebet J. Structural response of SG-laminated reinforced glass beams; 658 experimental investigations on the effects of glass type, reinforcement percentage and beam size. Eng 659 Struct 2012;36:292–301. https://doi.org/10.1016/j.engstruct.2011.12.016. 660 [18] Martens K, Caspeele R, Belis J. Development of Reinforced and Posttensioned Glass Beams: Review of 661 Experimental Research. J Struct Eng 2016;142:4015173. https://doi.org/10.1061/(asce)st.1943662 541x.0001453. 663 [19] Haldimann M. Fracture Strength of Structural Glass Elements - Analytical and Numerical Modelling, 664 Testing and Design. PhD thesis. École Polytechnique Fédérale de Lausanne, Switzerland, 2006. 665 [20] Louter C, Cupać J, Lebet J. Exploratory experimental investigations on post-tensioned structural glass 666 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. beams. J Facade Des Eng 2014;2:3–18. https://doi.org/10.3233/FDE-130012. 667 [21] Bos F, Veer F, Hobbelman G, Louter C. Stainless steel reinforced and post-tensioned glass beams. 12th 668 Int. Conf. Exp. Mech., Bari, Italy: 2004, p. 1–9. 669 [22] Engelmann M, Weller B. Post-tensioned glass beams for a 9 m Spannglass Bridge. Struct Eng Int 670 2016;26:103–13. https://doi.org/10.2749/101686616X14555428759000. 671 [23] Silvestru V, Deng Z, Michels J, Taras A. Enabling a Ductile Failure of Laminated Glass Beams with Iron‐ 672 Based Shape Memory Alloy (Fe-SMA) Strips. Int. Colloq. Stab. Ductility Steel Struct., Aveiro, Portugal: 673 Ernst & Sohn GmbH; 2022. https://doi.org/https://doi.org/10.1002/cepa. 1839. 674 [24] Valarinho L, Correia JR, Branco F. Experimental study on the flexural behaviour of multi-span transparent 675 glass-GFRP composite beams. Constr Build Mater 2013;49:1041–53. 676 https://doi.org/10.1016/j.conbuildmat.2012.11.024. 677 [25] Louter C, Leung C, Kolstein H, Vamberský J. Structural glass beams with embedded glass fibre 678 reinforcement. Challenging Glas 2 - Conf Archit Struct Appl Glas CGC 2010 2010:439–48. 679 [26] Sena-Cruz J, Barros J, Coelho M, Silva L. Efficiency of different techniques in flexural strengthening of 680 RC beams under monotonic and fatigue loading. Constr Build Mater 2012;29:175–82. 681 https://doi.org/10.1016/j.conbuildmat.2011.10.044. 682 [27] Rocha J, Sena-Cruz J, Pereira E. Influence of adhesive stiffness on the post-cracking behaviour of CFRP683 reinforced structural glass beams. Compos Part B Eng 2022;247. 684 https://doi.org/https://doi.org/10.1016/j.compositesb.2022.110293. 685 [28] Callewaert D. Stiffness of Glass / Ionomer Laminates in Structural Applications. PhD thesis. Faculty of 686 Engineering and Architecture, Ghent University, Belgium, 2012. 687 [29] Cupać J, Louter C, Nussbaumer A. Post-tensioning of glass beams: Analytical determination of the 688 allowable pre-load. Glas Struct Eng 2021;6:233–48. https://doi.org/10.1007/s40940-021-00150-0. 689 [30] You YC, Choi KS, Kim J. An experimental investigation on flexural behavior of RC beams strengthened 690 with prestressed CFRP strips using a durable anchorage system. Compos Part B Eng 2012;43:3026–36. 691 https://doi.org/10.1016/j.compositesb.2012.05.030. 692 [31] Correia L, Sena-Cruz J, Michels J, França P, Pereira E, Escusa G. Durability of RC slabs strengthened 693 with prestressed CFRP laminate strips under different environmental and loading conditions. Compos Part 694 B Eng 2017;125:71–88. https://doi.org/10.1016/j.compositesb.2017.05.047. 695 [32] Bilotta A, Ceroni F, Di Ludovico M, Nigro E, Pecce M, Manfredi G. Bond Efficiency of EBR and NSM 696 FRP Systems for Strengthening Concrete Members. J Compos Constr 2011;15:757–72. 697 https://doi.org/10.1061/(asce)cc.1943-5614.0000204. 698 [33] Abdulla NA. A state-of art-review of materials, methods, and applications of PVC-FRP-confined concrete. 699 Constr Build Mater 2023;363:129719. https://doi.org/10.1016/j.conbuildmat.2022.129719. 700 [34] Dias SJE, Barros JAO. Performance of reinforced concrete T beams strengthened in shear with NSM 701 CFRP laminates. Eng Struct 2010;32:373–84. https://doi.org/10.1016/j.engstruct.2009.10.001. 702 [35] Cagnacci E, Orlando M, Spinelli P. Experimental campaign and numerical simulation of the behaviour of 703 reinforced glass beams. Glas. Perform. Days, Tempere, Finland: 2009. 704 [36] Rezazadeh M, Cholostiakow S, Kotynia R, Barros J. Exploring new NSM reinforcements for the flexural 705 strengthening of RC beams: Experimental and numerical research. Compos Struct 2016;141:132–45. 706 https://doi.org/10.1016/j.compstruct.2016.01.033. 707 [37] Obaydullah M, Jumaat M, Alengaram U, Kabir M, Rashid M. Combining EBR CFRP sheet with 708 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. prestressed NSM steel strands to enhance the structural behavior of prestressed concrete beams. J Civ Eng 709 Manag 2021;27:637–50. https://doi.org/https://doi.org/10.3846/jcem.2021.14905. 710 [38] Lim D. Combinations of NSM and EB CFRP strips for flexural strengthening of concrete structures. Mag 711 Concr Res 2009;61:633–43. https://doi.org/10.1680/macr.2008.61.8.633. 712 [39] Belarbi A, Dawood M. Sustainability of fiber-reinforced polymers (FRPs) as a construction material. 713 Houston, USA: Woodhead Publishing; 2016. https://doi.org/10.1016/b978-0-08-100370-1.00020-2. 714 [40] Palumbo M. A New Roof for the XIIIth Century ‘Loggia de Vicari’ (Arquà Petrarca – PD Italy) Based on 715 Structural Glass Trusses: A Case Study. Glas. Process. Days, Tempere, Finland: 2005. 716 [41] Bedon C, Louter C. Numerical analysis of glass-FRP post-tensioned beams – Review and assessment. 717 Compos Struct 2017;177:129–40. https://doi.org/10.1016/j.compstruct.2017.06.060. 718 [42] Rocha J, Pereira E, Sena-Cruz J. Feasibility of mechanical post-tensioning of annealed glass beams by 719 activating externally bonded Fe-SMA reinforcement. Constr Build Mater 2023;365. 720 https://doi.org/10.1016/j.conbuildmat.2022.129953. 721 [43] Leinenbach C, Kramer H, Bernhard C, Eifler D. Thermo-mechanical properties of an Fe-Mn-Si-Cr-Ni-VC 722 shape memory alloy with low transformation temperature. Adv Eng Mater 2012;14:62–7. 723 https://doi.org/10.1002/adem.201100129. 724 [44] Hosseini A, Michels J, Izadi M, Ghafoori E. A comparative study between Fe-SMA and CFRP 725 reinforcements for prestressed strengthening of metallic structures. Constr Build Mater 2019;226:976–92. 726 https://doi.org/10.1016/j.conbuildmat.2019.07.169. 727 [45] Izadi M, Hosseini A, Michels J, Motavalli M, Ghafoori E. Thermally activated iron-based shape memory 728 alloy for strengthening metallic girders. Thin-Walled Struct 2019;141:389–401. 729 https://doi.org/10.1016/j.tws.2019.04.036. 730 [46] Michels J, Shahverdi M, Czaderski C. Flexural strengthening of structural concrete with iron-based shape 731 memory alloy strips. Struct Concr 2017;19:876–91. https://doi.org/10.1002/suco.201700120. 732 [47] Shahverdi M, Michels J, Czaderski C, Motavalli M. Iron-based shape memory alloy strips for 733 strengthening RC members: Material behavior and characterization. Constr Build Mater 2018;173:586– 734 99. https://doi.org/10.1016/j.conbuildmat.2018.04.057. 735 [48] Shin M, Andrawes B. Experimental investigation of actively confined concrete using shape memory 736 alloys. Eng Struct 2010;32:656–64. https://doi.org/10.1016/j.engstruct.2009.11.012. 737 [49] Rojob H, El-Hacha R. Self-prestressing using iron-based shape memory alloy for flexural strengthening 738 of reinforced concrete beams. ACI Mater J 2017;114:523–32. https://doi.org/10.14359/51689455. 739 [50] Dong Z, Klotz UE, Leinenbach C, Bergamini A, Czaderski C, Motavalli M. A novel Fe-Mn-Si shape 740 memory alloy with improved shape recovery properties by VC precipitation. Adv Eng Mater 2009;11:40– 741 4. https://doi.org/10.1002/adem.200800312. 742 [51] Shahverdi M, Czaderski C, Motavalli M. Iron-based shape memory alloys for prestressed near-surface 743 mounted strengthening of reinforced concrete beams. Constr Build Mater 2016;112:28–38. 744 https://doi.org/10.1016/j.conbuildmat.2016.02.174. 745 [52] Shahverdi M, Czaderski C, Annen P, Motavalli M. Strengthening of RC beams by iron-based shape 746 memory alloy bars embedded in a shotcrete layer. Eng Struct 2016;117:263–73. 747 https://doi.org/10.1016/j.engstruct.2016.03.023. 748 [53] Czaderski C, Shahverdi M, Brönnimann R, Leinenbach C, Motavalli M. Feasibility of iron-based shape 749 memory alloy strips for prestressed strengthening of concrete structures. Constr Build Mater 2014;56:94– 750 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. Table 4. Results obtained during the prestressing procedure of the NSM-CFRP reinforcement in the P-C-C and CFRP_SMA beams. Specimen Reinforcement εr,p[‰] Fp [kN] P-C-C NSM 1.926 8.50 CFRP_SMA NSM 1.923 8.49 Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. Table 5. Results obtained during the activation (prestressing) of Fe-SMA strips, as well as the comparison between the experimental measurements and the numerical results. Experimental measurements Specimen Reinforcement Ta [ºC] dexp [mm] σg,t [MPa] CFRP_SMA EBR 121.9 0.413 -2.997 SMA_CFRP NSM 178.9 0.552 -3.759 SMA_SMA NSM 200.8 0.615 -4.174 EBR 122.0 0.443 -3.138 Numerical results Specimen Reinforcement dnum [mm] σrec [MPa] Fp [kN] CFRP_SMA EBR 0.418 (1.2%) 257.3 7.12 SMA_CFRP NSM 0.557 (0.9%) 373.5 11.21 SMA_SMA NSM 0.619 (0.7%) 414.6 12.44 EBR 0.439 (-0.9%) 268.2 8.05 Note: The difference between dnum and dexp is indicated in parentheses Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. Table 6. Final post-tensioning force applied to the NSM (FNSM) and EBR (FEBR) reinforcement elements, as well as the compressive pre-stress at the bottom edge of glass, both estimated from the numerical simulations. Specimen FNSM [kN] FEBR [kN] σg,b [MPa] P-C-C 8.50 - -6.17 (+15.4 %) CFRP_SMA 8.14 7.12 -12.16 (+30.4 %) SMA_CFRP 11.21 - -8.28 (+20.7 %) SMA_SMA 12.24 8.05 -15.21 (+38.0 %) Note: The theoretical increase in initial cracking load considering fg,t = 40 MPa (see Table 1) is indicated in parentheses. Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. Table 7. Summary of the main properties extracted from the F – δ experimental responses of laminated glass beams, as well as the failure mode observed and the strain gauge measurements at failure. Property Units Specimen R-C-C P-C-C CFRP_SMA SMA_CFRP SMA_SMA Structural response K [kN/mm] 3.49 3.53 3.63 3.57 3.47 Fcr [kN] 11.31 13.88 15.48 14.02 22.02 δcr [mm] 3.24 3.93 4.26 3.93 6.34 Fmax [kN] 18.19 20.27 18.95 17.79 19.30 Fult [kN] 17.65 20.27 18.95 17.79 18.13 δult [mm] 31.45 26.54 22.24 34.59 35.21 Di [%] 971.2 675.3 521.7 881.2 555.0 RSi [%] 160.8 146.0 122.4 126.9 87.6 Strain gauge measurements εg,t [‰] -1.09 -1.43 -1.22 -1.25 -2.65 εr [‰] 4.45 4.35 4.72 4.71 48.72 Failure modes - - GC a) GC a) GC a) GC a) GC a) Note: a) Collapse triggered by glass crushing Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. Table 8. Main parameters derived from the post-cracking stage of the series SDur [27] and P_T120 [42], as well as the respective failure modes. Series Di [%] RSi [%] Failure mode εr [‰] SDur 407 87 CSC a) 3.22 P_T120 1584 111 UN b) 40.3 Note: a) Collapse triggered by critical shear crack formation b) Specimen unloaded before failure Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. Table 9. Modulus of elasticity adopted for each reinforcing element, as well as the comparison between the experimental cracking loads and those calculated analytically considering the fg,t obtained from the mechanical characterization and the fg,eff derived from the bending tests. Property Units Specimen R-C-C P-C-C CFRP_SMA SMA_CFRP SMA_SMA Reinforcement tensile stiffness ENSM [GPa] 183.87 183.87 183.87 59.94 52.13 EEBR [GPa] 183.87 183.87 101.68 183.87 99.18 Cracking load Fcr [kN] 11.31 13.88 15.48 14.02 22.02 Fcr,a (fg,t) [kN] 14.73 (30.2 %) 17.16 (23.6 %) 19.05 (23.1 %) 17.39 (24.0 %) 19.81 (10.0%) Fcr,a (fg,eff) [kN] 11.40 (0.8 %) 13.79 (-0.6 %) 15.74 (1.7 %) 14.13 (0.8 %) 16.56 (24.8 %) Notes: Values in parenthesis correpond to the difference between the theoretical values in relation to the experimental ones Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. LIST OF FIGURES Fig. 1. Schematic representation of the cross section of the laminated glass beams: (a) exploited view; (b) assembled view after the lamination process; and (c) detailing the glass groove for inserting the reinforcement. Units in [mm]. Fig. 2. Overview of the experimental setup used for prestressing the CFRP laminates: (a) prestressing bed; and (b) external reaction frame, hydraulic jack and metal clamps used to fix the CFRP laminates. Fig. 3. Schematic activation procedure of Fe-SMAs under constraint stress recovery (red colour), also including the phase behaviour. Adapted from Michels et al. [31]. Fig. 4. Activation of the Fe-SMA reinforcement: (a) schematic representation of the experimental procedure; (b) overview of the experimental setup adopted; and connection between the power supply clamps and the (c) NSMSMA and (d) EBR-SMA strips, respectively. Fig. 5. Four-point bending tests carried out in this study: (a) general layout; and (b) experimental setup. Fig. 6. Post-tensioning of the P-C-C beam, namely the evolution of the pre-strain in the CFRP laminate, of the compressive pre-stress at the bottom glass edge and the temperature over the time. Fig. 7. Measurements recorded during the activation of the NSM-SMA strips in the SMA_SMA and SMA_CFRP beams: (a) displacement at the mid-span section (da,exp), (b) axial strain at the top edge of the glass panel (εg,t) and (c) temperature in the Fe-SMA (T). Fig. 8. Measurements recorded during the activation of the EBR-SMA strips in the CFRP_SMA and SMA_SMA beams: (a) displacement at the mid-span section (da,exp), (b) axial strain at the top edge of the glass panel (εg,t) and (c) temperature in the Fe-SMA (T). Fig. 9. Phased analysis adopted in numerical simulation to model the post-tensioning procedure and determine the recovery stress in the Fe-SMA strips, including the finite element model and the Fe-SMA strip length (red colour) subjected to temperature variation. Fig. 10. Flexural behaviour of the beams R-C-C and P-C-C: (a) load deflection curves; and (b-c) DIC crack patterns at different stages. Fig. 11. Flexural behaviour of the beams CFRP_SMA and SMA_CFRP: (a) load deflection curves; and (b-c) DIC crack patterns at different stages. Fig. 12. Flexural behaviour of the SMA_SMA beam: (a) load deflection curves; and (b) DIC crack patterns at different stages. Fig. 13. Flexural responses and crack patterns at failure of monolithic glass beams reinforced with (a) CFRP and (b) Fe-SMA. Fig. 14. Typical failure modes observed in (a) SMA_SMA beam and (b) in all other glass composite beams. Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. (a) (b) (c) (d) Fig. 1. Schematic representation of the cross section of the laminated glass beams: (a) exploited view; (b) assembled view after the lamination process; (c) detailing the glass groove for inserting the reinforcement; and (d) image showing the strengthening system adopted. Units in [mm]. ~24.5 198 22 aab b c 1010 3 a - Outer glass layer [220 * 10 mm2] b - interlayer [ti = 0.76 mm] c -Inner glass layer [198 * 3 mm2] Strengthening system (see Fig. 1c) ~24.5 20 STRENGTHENING SYSTEM 22 20 Reinforcement element (NSM) Adhesive Annealed glass 1.2 or 1.5 1.2 or 1.5 0.9 or 0.5 1.65 or 1.5 0.76 10 103 Interlayer 220 Reinforcement element (EBR) ~24.5 198 22 aab b c 1010 3 a - Outer glass layer [220 * 10 mm2] b - interlayer [ti = 0.76 mm] c -Inner glass layer [198 * 3 mm2] Strengthening system (see Fig. 1c) ~24.5 20 STRENGTHENING SYSTEM 22 20 Reinforcement element (NSM) Adhesive Annealed glass 1.2 or 1.5 1.2 or 1.5 0.9 or 0.5 1.65 or 1.5 0.76 10 103 Interlayer 220 Reinforcement element (EBR) ~24.5 198 22 aab b c 1010 3 a - Outer glass layer [220 * 10 mm2] b - interlayer [ti = 0.76 mm] c -Inner glass layer [198 * 3 mm2] Strengthening system (see Fig. 1c) ~24.5 20 STRENGTHENING SYSTEM 22 20 Reinforcement element (NSM) Adhesive Annealed glass 1.2 or 1.5 1.2 or 1.5 0.9 or 0.5 1.65 or 1.5 0.76 10 103 Interlayer 220 Reinforcement element (EBR) Outer glass layers Inner glass layer EBR reinforcement NSM reinforcement Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. (a) (b) Fig. 2. Overview of the experimental setup used for prestressing the CFRP laminates: (a) prestressing bed; and (b) external reaction frame, hydraulic jack and metal clamps used to fix the CFRP laminates. Laminated glass beam Heating fan CFRP laminate Reaction frame Bonding zone Monitoring CFRP laminate Metal clamp Hydraulic jack Reaction frame Hydraulic system Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. Fig. 3. Schematic activation procedure of Fe-SMAs under constraint stress recovery (red colour), also including the phase behaviour. Adapted from Michels et al. [46]. Stress Strain Stress Af 1 2 3.1 3.2 4 Pre-straining1 2Unloading Activation Temperature Thermal expansion 3.1 4 Service load 3.2 Cooling Heating Recovery stress 3 3 Austenite Martensite Austenite 1 2 3 4 Ms < T < AsT > As Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. (a) δ = δcr δ = δult / 2 δ = δult (b) δ = δcr δ = δult / 2 δ = δult (c) Fig. 10. Flexural behaviour of the beams R-C-C and P-C-C: (a) load deflection curves; and (b-c) DIC crack patterns at different stages. 0 5 10 15 20 25 0 10 20 30 40 50 Load, F[kN] Deflection, δ[mm] R-C-C P-C-C CFRP_SMA SMA_CFRP SMA_SMA Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. (a) δ = δcr δ = δult / 2 δ = δult (b) δ = δcr δ = δult / 2 δ = δult (c) Fig. 11. Flexural behaviour of the beams CFRP_SMA and SMA_CFRP: (a) load deflection curves; and (b-c) DIC crack patterns at different stages. 0 5 10 15 20 25 0 10 20 30 40 50 Load, F[kN] Deflection, δ[mm] R-C-C P-C-C CFRP_SMA SMA_CFRP SMA_SMA Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. (a) δ = δcr δ = δult / 2 δ = δult (b) Fig. 12. Flexural behaviour of the SMA_SMA beam: (a) load deflection curves; and (b) DIC crack patterns at different stages. 0 5 10 15 20 25 0 10 20 30 40 50 Load, F[kN] Deflection, δ[mm] R-C-C P-C-C CFRP_SMA SMA_CFRP SMA_SMA Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. (a) (b) Fig. 13. Force vs. mid-span strain measured at (a) the top edge of the glass panel and (b) at the bottom edge of the reinforcement element. 0 5 10 15 20 25 -0,5 0,0 0,5 1,0 1,5 2,0 2,5 3,0 Load, F[kN] Mid-span glass strain, εg,t [] R-C-C P-C-C CFRP_SMA SMA_CFRP SMA_SMA 0 5 10 15 20 25 0,0 10,0 20,0 30,0 40,0 50,0 Load, F[kN] Deflection, δ[mm] R-C-C P-C-C CFRP_SMA SMA_CFRP SMA_SMA Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. δ = δult δ = δult (a) (b) Fig. 14. Flexural responses and crack patterns at failure of monolithic glass beams reinforced with (a) CFRP and (b) Fe-SMA. 0,0 1,5 3,0 4,5 6,0 0 10 20 30 40 50 Load, F[kN] Deflection, δ[mm] SDur-I SDur-II 0,0 1,5 3,0 4,5 6,0 0 10 20 30 40 50 Load, F[kN] Deflection, δ[mm] P_T120-I P_T120-II Rocha, J.; Pereira, E.; Michels, J.; Sena-Cruz, J. (2023) “Hybrid strengthening and flexural behaviour of post-tensioned laminated glass beams.” Construction and Building Materials, 408: 133637, 22 pp. (a) (b) Fig. 15. Typical failure modes observed in (a) SMA_SMA beam and (b) in all other glass composite beams.