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

2075

2

1. Introduction The seismic hazard on the Italian territory is medium to high, highlighting the need for risk reduction measures especially when paired with the high vulnerability of the existing building stock. Recent seismic events have exposed the extreme vulnerability of existing unreinforced masonry (URM) structures in historic centers (e.g., D'Ayala and Paganoni (2011), Carocci (2012), Sorrentino et al. (2019)). The vulnerability sources within existing URM aggregate are various and mainly depend on the type of construction, weakening of the masonries due to improvised renovations, and the typical building density of the old towns. A common way of URM aggregate construction was to add structural components to pre-existing ones. As a result, adjacent structural units (SUs) are linked together through a structural connection that is variable in its effectiveness. Indeed, SUs can be found with shared mid-walls or built independently but in contact with pre-existing ones. The complicated structural configuration, as well as the utilization of a variety of materials and building processes, made the seismic assessment of URM aggregate challenging. Several approaches for assessing seismic vulnerability have been developed in recent decades, such as: • Large-scale approaches based on empirical evaluations obtained from post-earthquake data that identifies “vu lnerability classes" based on a few key morpho-typological and structural properties. Therefore, a vulnerability model for each recognized class can be calibrated. This is accomplished using statistical elaborations, which have the benefit of requiring limited data and processing quickly, making them ideal for analyzing a large number of structures or an entire city of big size (Del Gaudio et al. (2019), Penna et al. 2022a, Sisti et al. (2019); • Holistic approach is a multi-scale approach based on several components (i.e., hazard, vulnerability, exposure, and consequences) (Cardinali et al. (2021), Pirchio et al. (2021)); • Heuristic approach which is based also on the expert judgment (e.g. Vicente et al. (2014), Brando et al. (2017), Lagomarsino et al. (2021), Sandoli et al. (2022)); • Analytical/mechanical methodologies, which determine a vulnerability function that connects structural capacity and seismic demand using numerical models and building simulation procedures. Those methods necessitate in depth understanding of the building's properties and time-consuming structural calculations (particularly high for large-scale applications) (Cocco et al. 2019, Cima et al. (2021), Leggieri et al. (2021), Nale at al. (2021); • Analytical/numerical approach (Ramos and Lourenço (2004), Senaldi et al. (2010), Formisano and Massimilla (2019), Bernardini et al. (2019), Degli Abbati et al. (2019), Valente et al. 2019, Greco et al. 2020, Angiolilli et al. (2021), Battaglia et al. (2021), Bernando et al. (2022)); • Hybrid techniques integrating the different procedures by individuating representative building classes (Kappos et al. (2006), Maio et al. (2015). This method is useful when there is large observational data is missing or when calibrating or validating the results of analytical models is problematic. Despite the considerable effort put out in the aforementioned studies, some shortcomings remain. In particular, the aggregate-effect as well as the explicit consideration of the interaction between local mechanisms (OOP or pounding mechanisms) and the IP behavior is missing in most of the studies. For example, in empirical methods, only the unit location is treated as an extra vulnerability source making very difficult investigating in depth if the interaction among SUs is detrimental or beneficial varying the features of SUs. Furthermore, several analytical methods merely account for the IP or OOP response separately, without taking into consideration their interaction. It is worth mentioning that IP damage affects OOP strength decrease, as well documented in Dolatshahi and Yekrangnia (2015). One of the strategy to obtain reliable results at large scale, it is represented by the development of analytical- methodologies (able to take into account the problems described above) applied to a consistent number of representative case studies that, by varying their building features (e.g. geometry of buildings, layout of openings, structural details and restraints given by the structure, mechanical properties, diaphragm typologies, …), it allows to obtain more general results able to cover the much greater variability of the Italian building stock within small historical centres. Hence, the final deliverables of such studies (e.g. Lagomarsino et al. 2022) will improve and enrich the current database and increase the level of accuracy and reliability of large scale approaches. This leads to the possibility of obtaining a more accurate damage scenario specific for buildings with different peculiarities, such as different era (modern or historic) or context (within small or big historical centres). The mutual interaction between adjacent SUs is one of the main issue to be considered within the seismic assessment of historic URM aggregate. The influence of the SU compounds on the seismic behavior of the individual buildings is known as "aggregate-effect". A misleading aspect regarding this effect is represented to the fact that in