PSI - Issue 78

Sara Mozzon et al. / Procedia Structural Integrity 78 (2026) 646–653

647

1. Introduction Geological, hydrological, and hydraulic instability phenomena are widespread across Italy, causing considerable damage and posing serious risks to people, infrastructure, built environment, and buildings (Poljanšek et al. (2017); Spano et al. (2020)). In recent decades, several catastrophic events have led to significant destruction and fatalities. According to global data EM-DAT Europe recorded over 500 major damaging floods between 1998 and 2023. Floods rank among the most frequent and economically burdensome natural hazards worldwide (Capparelli et al. (2023), underscoring the need for coordinated risk mitigation strategies (Bangalore et al. (2016); Capparelli et al. (2023)). Flash floods and debris flows are particularly dangerous in mountainous areas, where they may be triggered by intense rainfall or dam failures, leading to severe downstream impacts (Iverson (2004); Milanesi and Pilotti (2021)). These risks are further intensified by climate-driven hydrological changes, infrastructure deterioration, and urban expansion (Fuchs et al. (2019); Milanesi and Pilotti (2021)). Thus, comprehensive assessment of flood and debris flow risk is essential to support both structural and non-structural mitigation measures. Effective risk evaluation requires an integrated approach encompassing hazard, exposure, and vulnerability (Milanesi and Pilotti (2021)). Yet, hazards are often assessed independently, limiting the capacity to address their compound effects. A shift towards multi-risk frameworks is necessary to evaluate interactions among different natural processes within a common methodology (IPCC (2014); Marzocchi et al. (2012)). As an initial step, multi layer vulnerability assessments can be adopted, treating each hazard separately but with harmonized procedures (Poljanšek et al. (2017) ). Assessing the vulnerability of the built environment is especially important in multi hazard-prone areas, where structural damage − up to building collapse − can result in severe human and economic losses (Cantelmo and Cuomo (2021); Milanesi et al. (2018)). A better understanding of structural vulnerability supports more effective emergency planning and disaster risk reduction strategies (Papathoma-Köhle et al. (2011)). Several studies have investigated the vulnerability of both masonry (Asad et al. (2024); Capparelli et al. (2023); Herbert et al. (2018); Lonetti and Maletta (2018); Milanesi et al. (2018)) and reinforced concrete (Luo et al. (2020); Petrone et al. (2017)) structures under hydraulic and debris impact. In Italy, residential buildings are mainly composed of unreinforced masonry (URM) or reinforced concrete (RC) frames with brick infills. URM structures rely on load-bearing walls, where the failure of a single wall may compromise the entire system. RC buildings, while structurally more redundant, are vulnerable to out-of-plane (OOP) horizontal forces acting on non-structural infill panels. A localized stability assessment of both structural (load-bearing walls) and non-structural (infills) elements offers a rational basis for evaluating the effects of hydrological hazards such as flash floods and debris flows, whose frequency is rising in connection with climate change (Cantelmo and Cuomo (2021)). This paper presents an analytical model developed to assess the OOP response of masonry infill panels. The model was implemented in a Monte Carlo simulation, incorporating variations in geometric and mechanical properties across different infill classifications. Based on the Monte Carlo results, polynomial surrogate vulnerability models were derived to describe the OOP behavior of panels, providing an efficient tool for vulnerability assessment. As the final step, fragility curves relevant for territorial-scale risk assessments were developed, accounting for different infill types and load conditions. 2. Presentation of the analytical model The analytical model proposed in this study simulates a plate resistance mechanism by incorporating the formation of a double arching system − both vertical and horizontal − within the wall thickness. This is coupled with an incremental procedure that governs OOP displacements. The algorithm iteratively applies horizontal loads and recalculates force and moment equilibrium until collapse occurs in segments. As illustrated in Fig. 1, red cracks indicate the development of five internal fracture lines, segmenting the panel into four distinct blocks. The model is highly versatile, allowing for the analysis of infill panels under various horizontal load distributions − triangular, rectangular, or trapezoidal − with different heights ( h w ), thereby simulating a range of triggering events. Furthermore, it can be embedded within a Monte Carlo simulation framework to generate capacity models for different infill typologies, accounting for uncertainties in geometry and material properties. From a multi-risk