PSI - Issue 78

Beatrice Travasoni et al. / Procedia Structural Integrity 78 (2026) 1111–1118

1117

Fig. 6. Kinematic representation of the three out-of-plane failure mechanisms analysed. (a) Overturning mechanism with a single hinge at the base (point O) with tensile connections; (b) Two-block mechanism with hinges at the base (O) and at mid-height (C), including tensile ties at first floor levels (points T1); (c) Combined mechanism with the same tie configuration as case (b), but allowing for a different rotation condition of the upper block. The top segment is free to displace horizontally, resulting in a combined overturning–bending behaviour. 4. Conclusion This study has investigated the out-of-plane seismic response of unreinforced masonry (URM) façade walls, with particular attention to the influence of timber diaphragms and wall-to-diaphragm connection systems. A nonlinear finite element model of a two-storey benchmark structure was calibrated using experimental data on masonry, timber diaphragms, and connection components. The results confirm that both the in-plane stiffness of the diaphragm and the mechanical properties of the connections significantly influence the overall seismic response. Rigid diaphragms improve global interaction and reduce out-of-plane displacements, but also increase the tensile demand on the connections. The nonlinear tensile behaviour of wall-to-diaphragm ties plays a key role in governing the evolution of local mechanisms. The numerical analyses allowed for the identification of a combined failure mechanism, in which initial overturning is followed by vertical bending around an intermediate section of the wall. This mechanism, which reflects the progressive degradation of multiple connection levels, cannot be captured by simplified models assuming rigid restraints or single-mode behaviour. These findings highlight the necessity of explicitly modelling nonlinear connection response, even in the presence of diaphragms commonly assumed as rigid. The results contribute to the understanding of diaphragm–wall interaction and support the development of more accurate vulnerability assessment and retrofitting strategies for existing URM buildings. Acknowledgements The present investigation was developed in the framework of the Research Program FAR 2024 of the University of Ferrara. Moreover, the analyses were developed within the activities of the (Italian) University Network of Seismic Engineering Laboratories (ReLUIS) in the research program funded by the (Italian) National Civil Protection (DPC) – Progetto DPC-ReLUIS 2024/26 – WP4.