PSI - Issue 78

Francesca Mattei et al. / Procedia Structural Integrity 78 (2026) 2169–2175

2173

on the other, to the rigid steel transverse beam transferring the load to the top RC beam of the masonry wall. The instrumentation layout was designed to measure and monitor both load magnitude and deformation during the test. Vertical, horizontal, and diagonal LVDT transducers, together with load cells, were employed. Load cells were placed at each hydraulic jack, while in-plane displacements at the top, bottom, and mid-height of the specimen were recorded using LVDT transducers of adequate measuring length. Possible relative slip between the laboratory strong floor and the steel basement frames, due to adjustments of the shear contrasts during loading, was checked. Diagonal deformations, as well as vertical deformations between the retrofit panel and both the foundation and the top RC beam, were also measured. Tests’ execution and results The vertical load was applied prior to the displacement history to replicate the actual conditions of existing buildings and was kept constant for the entire duration of all tests. The axial loads applied at the top of the masonry wall and the RC panel — estimated from preliminary numerical simulations, were about 90 kN and 22 kN, respectively. The monotonic test was carried out until a significant failure mode was reached.

Fig. 3. Cracks at the base of the walls, following tests.

Fig. 3 shows the cracks observed at the base section of the masonry wall, while the sudden load drops visible in Fig. 4 correspond to the rupture of the rebars at the connection surface between the RC foundation system and the specimen. Fig. 4 also compares the experimental results with the numerical modelling presented in Caprili et al. (2024) The model consisted of two one-dimensional nonlinear elements representing the retrofit panel and the existing masonry wall, respectively, and was developed in OpenSees®, McKenna et al. (2000). For both concrete and masonry, the uniaxialMaterialConcrete04 model was adopted, which reproduces the stress – strain compression behaviour proposed by Popovics (1973). Mechanical properties, such as compressive stre ngth and Young’s modulus, were calibrated on the basis of material test results. The connection between the two panels at the top level was modelled using rigid link elements capable of transferring only horizontal actions.