PSI - Issue 44

Carlo Vienni et al. / Procedia Structural Integrity 44 (2023) 2270–2277

2273

4

Vienni et al. / Structural Integrity Procedia 00 (2022) 000–000

a c Fig. 2. (a) Test setup of the single-lap shear test; (b) failure mechanism for detachment at the mortar-to-masonry interface; (c) mortar cracking. b

Table 2. Results of single-lap shear bond tests. Specimen F u_B [kN] σ u = F u_B /Ay [MPa]

τ u = F u_B /A b [MPa]

τ u = F u_B /A b [MPa]

σ u = F u_B /A y [MPa]

Specimen

F u_B [kN]

BT_X_01 BT_X_02 BT_X_03

7,68 9,57 7,51 8,26

388 483 379

0,21 0,27 0,21

BT_Y_01 BT_Y_02 BT_Y_03

10,25

431 308 371

0,28 0,20 0,25

7,33 8,83 8,80

Average (CoV)

417 (11,3%)

0,23 (11,3%)

Average (CoV)

370 (13,5%)

0,24 (13,5%)

The reinforcement was modeled considering the real geometry of the grid, with tension-only elastic-brittle truss elements. The tensile strength of the fibers, in warp and weft direction, was set equal to the minimum value from tensile tests on dry yarns and tensile tests on CRM coupons: f t,x = 503 MPa in weft direction and f t,y = 485 MPa in the warp direction. The Young modulus of the fibers was obtained from tensile tests on dry yarns (Section 2) and was set equal to E x =56.5 GPa in weft and E y =35.1 GPa in the warp direction. Regarding the simulation of detachment phenomenon, differently from other composite reinforcements that provide only tensile resistance where delamination can be reproduced by limiting the tensile strength of the reinforcement as shown in Ernesto Grande et al. (2008), it is necessary to consider CRM contribution and possible detachment even for compressive stresses. Hence, a nonlinear interface at the matrix-to-support interface was modeled. The interface was modeled with combined cracking shearing-crushing constitutive law; interface stiffness G and tangential strength τ were set. These two parameters, together with the tangential fracture energy G f II , were calibrated considering the results of shear bond tests (Table 2). The reliability of the modeling approach to reproduce the shear response of masonry walls was also verified using the results of shear-compression tests carried out by Gattesco et al. (2013), considering the same panel geometry, test setup, and mechanical parameters described in the work. The comparison between experimental and numerical results in terms of capacity curves is shown in Fig. 3. 3.1. Shear-compression parametric simulations A parametrical analysis was carried out on masonry walls of dimensions of 1500x2000x t mm 3 , where the thickness t was considered as a parametrical variable, in double bending configuration, to study the CRM effect on shear compression behavior of the walls. Nonlinear static analyses were carried out in shear compression configuration: compressive load was applied at the beginning of each test and set constant during all the loading process; the horizontal load was applied at the top of the panel in displacement control. Regarding reinforced walls, CRM was applied on both faces of the walls. An important modeling hypothesis was to consider compressive load at the bottom and top faces applied to the wall only. In this way, the two CRM layers were not directly loaded and the actions acting on the reinforcement were transferred only through tangential stresses, simulating a more realistic loading condition, as also observed in Donnini et al. (2021). Analyses were carried out by varying vertical compression levels, masonry typology, and wall thickness. The considered parameters are shown in Table 3. All the walls were analyzed in unreinforced and reinforced configurations, for a total of 90 numerical analyses.