PSI - Issue 78

Mariano Di Domenico et al. / Procedia Structural Integrity 78 (2026) 1237–1244

1238

1. Introduction Roads play a crucial role not only in everyday transportation but also during emergency situations. Earthquakes can interrupt road functionality either directly, due to the collapse of road elements like bridges or viaducts, or indirectly, due to debris falling onto the pavement (Anelli et al., 2020). Such interruptions can hinder both evacuation of civilians and rescue operations. For instance, Lu et al. (2020) showed that when debris covers more than 25% of a street's surface, pedestrians can barely pass, while Domaneschi et al. (2019) demonstrated that a car cannot drive along a street if debris volume exceeds 15 m³ or if debris height is over 0.30 m. In Italy, the awareness of this issue has led to specific guidelines such as the CLE Handbook (TCSM 2014), which proposes a methodology to assess the vulnerability of urban roads during seismic events. It introduces a road survey form collecting geometric and morphological data, as well as the strategic importance of each road. However, it does not directly evaluate the causes of road blockage, only highlighting critical situations like narrow streets flanked by tall buildings. Earlier, Goretti and Sarli (2006), within the SAVE Project, developed a probabilistic approach to estimate the likelihood of a road segment being blocked. This method considers the number and typology of adjacent buildings and the probabilities of specific damage mechanisms (propping, overturning, collapse) leading to road obstruction. Probabilistic values were derived from post-earthquake observations (Cherubini, 2006). However, their methodology does not provide intermediate outputs such as debris amount or location, which could be valuable for planning recovery actions. More recently, Hirokawa and Osaragi (2016) proposed a buffer-debris model simulating omnidirectional debris spread from a collapsed building. It allows to assess whether a street remains passable, depending on the overlap of debris zones. Garcia-Torres et al. (2017) estimated debris generation using the HAZUS model (FEMA 2003), by linking the structural damage level to debris percentages for different materials and typologies. Domaneschi et al. (2019) proposed a simplified analytical model to estimate the debris extent from masonry building collapse, calibrated with AEM simulations and experimental tests. Anelli et al. (2020) introduced a methodology for deriving fragility curves of road networks, based on the debris distribution from nearby buildings. Their estimation starts from assumptions in the Italian Code NTC 2018 (MIT 2018), where buildings at Immediate Occupancy (IO) state produce negligible debris, while Near Collapse (NC) for masonry corresponds to maximum debris extent. Between IO and NC, debris volume increases linearly with seismic intensity. In the case of RC buildings, the Life Safety (LS) limit state is taken as the point of maximum debris area, beyond which only volume continues to grow due to further non-structural damage. Amini et al. (2024) quantified the effect of debris on pedestrian speed by computing reduction coefficients based on debris geometry and weight. Debris was estimated using HAZUS and represented through a triangular prism model developed by Argyroudis et al. (2015). Lin et al. (2024) used LS-DYNA combined with Houdini to simulate building collapse and debris accumulation, evaluating its impact not only on traffic but also on post-earthquake hospital efficiency. From this overview, it is evident that the topic is widely investigated, although most studies concentrate on masonry buildings. Debris estimation is usually derived via simplified models or simulations, with limited application to RC buildings. This study focuses on reinforced concrete frames with masonry infill walls and investigates the debris generated by out-of-plane collapse of these elements. Using nonlinear dynamic analyses on buildings differing in number of storeys, construction age, and infill typology, the debris volume and distance of projection are calculated for increasing seismic intensity. 2. Case-study buildings This study investigates six existing reinforced concrete (RC) moment-resisting frame buildings with uniformly distributed masonry infills. The buildings are not designed according to current codes but represent typical Italian construction practices from the 1950s – 1960s, 1970s, and 1980s – 1990s. For each period, two buildings were designed in L’Aquila: a 3 -storey and a 6-storey building, labelled as 3P50/6P50, 3P70/6P70, and 3P90/6P90. All