PSI - Issue 44

Ingrid Boem et al. / Procedia Structural Integrity 44 (2023) 2238–2245 I. Boem, B. Patzák, A. Kohoutková / Structural Integrity Procedia 00 (2022) 000–000

2244

7

of the sample was applied. The simulations were performed by increasing monotonically the horizontal displacement of the control point, in the middle of the panel, so that CRM resulted on the tensed side of the panel. The characteristics of the masonry layer were those calibrated for the in-plane shear-compression tests, since the same type of masonry was built. In Fig. 8a it is reported the numerical capacity curve representing the applied horizontal load varying the horizontal deflection at the mid span, with CRM applied of the tensed side; the experimental cyclic curve is also plotted, for comparison. Moreover, in Fig. 8b the principal tensile strains at collapse are reported, in comparison with the experimental crack pattern. The sample performed almost elastically till the attainment of the first cracking in the mortar coating, on the tensed side, at the mid-height of the sample (cracking load 19.2 kN, span 1.1 mm). Then, a sudden drop of the stiffness occurred and the damage progressively diffused in a wider area, according to the experimental evidences. The failure was reached for the attainment of the rupture in the GFRP reinforcement, at about 40.4 kN (36.6 mm). In respect to the experimental envelope curve, the numerical one performed some higher initial stiffness, likely due to small gaps at the upper and lower constrains, not accounted in the simplified simulations. However, the first cracking load and the post-cracking stiffness were predicted with accuracy. The collapse point resulted underestimated by the numerical model (-22% resistance, -39% displacement). However, it has to be considered that the average ultimate strain value of single GFRP yarns was set for the simulations, neglecting possible beneficial “group-effect” that occurs when pulling simultaneously several wires, as in these tests. Moreover, the numerical model did not account of some debonding of the mortar coating from the masonry, as well as of some shear damaging in the masonry, that were monitored in the final cycles of the experimental test and the may have increased the monitored deflection.

a b Fig. 8. Capacity curves and principal tensile strain at ultimate displacement (CRM applied on the tensed side).

5. Conclusions The multi-layer numerical model was recently developed with the OOFEM code for the analysis of masonry elements strengthened with the CRM technique. In this paper, it has been applied for the simulation of the performances of masonry piers subjected to in-plane shear-compression tests and to out-of-plane bending tests. The models were based on 20-nodes brick elements composed by a sequence of through-the-thickness plies representing the masonry, the mortar coating and the GFRP reinforcement. The analyses were performed in the nonlinear-static range, considering the nonlinearities of the different materials (cracking and crushing of both the masonry and the mortar coating and GFRP brittle failure in tension), whose characteristics were set on the basis of previous experimental evidences. The results of a new set of tests, recently performed, served for the model validation. In general, the simulations resulted capable to predict the experimental performances in terms of both capacity curves and collapse modes. In particular, good predictions of the in-plane performances of piers strengthened with CRM applied at one or at both sides emerged in terms of stiffness, resistance and post-peak behavior; it was also possible to properly distinguish diagonal cracking and in-plane bending failure mechanisms. Also the simulation of the out-of-plane behavior of a masonry sample with the CRM applied on the tensed side provided satisfying results in terms of global trend and failure mode. It should however be remembered that the