PSI - Issue 78

Giuseppe Brandonisio et al. / Procedia Structural Integrity 78 (2026) 2162–2168

2165

The most critical vulnerability, both under vertical loads and seismic actions, is related to the widespread presence of vaulted floor systems at the first and second levels. These vaults often exert lateral thrusts that are not consistently counteracted b y tie rods or chains. The masonry walls are of substantial thickness, ranging from a minimum of 70 cm for the external walls on the top floor to a maximum of 130 cm at ground level —values that exceed those typically prescribed by traditional construction standards for this type of masonry (Rondelet, 1814). 2.3. FEM modeling The seismic vulnerability assessment of the case study was carried out using nonlinear static (pushover) analyses. In accordance with the Italian Building Code (NTC 2018) (CM2019), the masonry walls were modelled using one dimensional beam elements, following the so-called Equivalent Frame approach. The walls were subdivided into vertical elements (piers), horizontal elements (spandrels), and rigid regions at the intersections between piers and spandrels (rigid joints). Piers and spandrels were modelled respectively as columns and beams within a two dimensional frame, while rigid joints were represented using rigid links. This modelling strategy allows for the use of a lumped plasticity model, with plastic hinges defined at specific locations—typically at the ends of beams and at the base and top of columns—to simulate both bending and shear failure mechanisms. Such an approach enables the performance of nonlinear incremental collapse analysis of the masonry walls. Plastic hinges can be activated in the piers to account for combined axial compression and bending actions, as well as for shear.

To evaluate the influence of structural complexity and inter-body interaction (Brandonisio and De Luca, 2024)., eight finite element (FE) models were developed and analysed (see Fig. 3 ). In detail, six structural models hypothesized isolated from the adjacent bodies were analysed, i.e.: Body A; Body B; Body C; Block 1; Block 2; Block 3. To take into account the good connection that characterizes the building, the following two global models were also analysed: AS-IS global model (Body A + Body B + Body C + Block 1 + Block 2 + Block 3) and the complete global model (Body A + Body B + Body C + Body D + Block 1 + Block 2 + Block 3 + Block 4) (see Figure 3c). Block D overlooking Piazza d'Armi and Block 4 were not analysed as they were in a state of partial collapse and therefore, structurally unsafe. They were modelled only in the complete Global Model to simulate the behaviour of the building in the hypothesis of complete reconstruction of these portions of the building as well Fig. 3 Global and locals FE models by CDS-WIN software