Site effects and soil-foundation-structure interaction: derivation of fragility curves and comparison with Codes-conforming approaches for a masonry school

The paper investigates the role of site-amplification and soil-foundation-structure (SFS) interaction on the seismic response of masonry structures. The aim is twofold: to quantify the potential impact of such effects on fragility curves numerically derived through nonlinear dynamic analyses; and to verify the reliability of conventional code-conforming approaches. These goals are pursued by referring to a prototype masonry building inspired by a school whose response was proven to be affected by SFS interaction during the Central Italy 2016/ 2017 earthquake. The 3D structural model was generated through the equivalent frame approach, able to simulate the in-plane nonlinear response and the hysteretic behaviour of masonry panels. The model was firstly fully restrained at its base to simulate the fixed-base conditions and then endowed with springs simulating the soil compliance. To account for the energy dissipated by the foundation, the latter model was characterized by a damping ratio higher than that assumed for the fixed base. The sensitivity of the structural response to various formulations of the soil-foundation stiffness and damping was investigated. Nonlinear dynamic analyses were performed on both models, firstly, under the acceleration time histories of 49 natural earthquakes recorded on stiff rock outcrop, and under the same set of accelerograms propagated through a site response analysis in a 1D soil model reproducing the actual soil profile under the examined school. The resulting fragility curves confirmed the expected negative impacts of site amplification, while highlighted a potential beneficial role of the SFS interaction enhanced by soil hysteresis. In the latter case, the probability of failure reduces more significantly at severe and very severe to near collapse damage levels. Finally, the average damage associated with the fragility curves calculated in this study was compared with that obtained from the fragility curves of a fixed base system in which site effects are considered through the conventional coefficients proposed in three different international Standards. The comparison highlighted that, when only site effects are considered, the code-approach underestimates the damage up to three damage levels. The beneficial effect of the SFS interaction reduces this gap but a difference of one/two damage levels still results.