PSI - Issue 44

Michele Angiolilli et al. / Procedia Structural Integrity 44 (2023) 2074–2081 M. Angiolilli et al./ Structural Integrity Procedia 00 (2022) 000 – 000

2080

7

The identification of different OOP mechanisms was defined based on the geometry of buildings, layout of openings structural details, expert judgment and with the support of evidence on most recurring mechanisms provided by past earthquakes. The overturning of entire panels around a hinge at the base (i.e. the so-called cantilever mechanism) was considered for the SUs of Catania and Visso. For the latter case, the overturning of tympanum was also considered (see Angiolilli at al. (2021, 2022b) for further details). Fig. 3 shows the comparison between fragility curves of the SUs under the isolated or aggregated conditions (orange and blue curves, respectively) as well as fragility curves affected by local mechanisms (blue square markers). In particular, the aggregate-effect positively influences the global IP building performance at low seismic intensity (UPD or DL2) and, in particular, at high seismic intensity (GC or DL4); however, the entity of such effect significantly varies for the different SUs. Furthermore, one can see that curve of combined mechanisms (IP and local mechanisms) are more conservative, especially at high seismic intensity, even if the shift on the left of the fragility curve is more or less pronounced. It highlights the potential impact of considering in an integrated way both mechanisms.

Fig. 3. Fragility curves of SUs belonging to Catania’ and Visso’s aggregate .

3. Conclusions The study herein presented aimed at comparing the seismic performance of existing unreinforced masonry building in aggregate located in different Italian historical centres. In particular, the aggregate-effect was investigated through nonlinear static analyses (NSA; i.e. pushover) and nonlinear dynamic analyses (NDA), the latter processed by fragility curves. Moreover, the study provided the fragility curves of the instigated building accounting also for local mechanisms, through an integrated process. Despite the different way in processing NDA data, the fragility curves show the same main outcomes highlighting the importance of the integrated procedure in considering both global and local behaviors. An important outcome of this study regards the inefficacy of capturing the actual seismic behaviour of masonry buildings in aggregate through NSA, at least with common load patterns proposed for ordinary isolated buildings. Indeed, by observing the capacity curves obtained by NSA, one cannot generalize whether the aggregate effect positively or negatively affects the seismic behaviour of the individual buildings. The aggregate-effect was instead much clearer by observing the results of the fragility curves from NDA. Actually, for the investigated SUs, it emerged a benefit offered by the confinement of the adjacent structural units under seismic action. However, the entity of this effect appears variable and thus a parametrical study appears to be necessary as a future development to provide more general recommendations. Finally, the approach adopted to integrate IP and OOP responses proved to be quite effective in investigating also how these two failure modes interact. Acknowledgements A part of the study presented in this article was developed within the activities of the ReLUIS-DPC 2019 – 2021 research programs, funded by the Presidenza del Consiglio dei Ministri — Dipartimento della Protezione Civile (DPC).