Probabilistic framework to evaluate scenario-based building vulnerability under landslide run-out impacts

Abstract

Quantifying building vulnerability under landslide run-out impacts is a pivotal aspect of landslide quantitative risk assessment (QRA). The degree of loss of buildings exposed to landslides (i.e., vulnerability) is dictated by the interplay between landslide run-out impacts and the resistance capacity of the building. However, the current link between landslide run-out and building capacity evaluations in landslide QRA is weak. Furthermore, uncertainties regarding both landslide run-out and building response must be quantified via suitable methods to enhance the trustworthiness of vulnerability evaluations. This study proposes a probabilistic framework to evaluate building vulnerability under landslide run-out impacts. The framework comprises a numerical model, which integrates dynamic run-out computations (for simulating the landslide run-out process) and finite element computations (for evaluating the resistance capacity of the building under lateral loading), aiming to assess the damage inflicted on buildings by landslide run-out impacts. Considering the uncertainties of the landslide run-out process and building structural response, a probability density evolution method (PDEM) is employed to calculate the exceedance probabilities of different limit states of building capacities. Ultimately, building vulnerability under landslide run-out impacts is ascertained by combining the probabilities of different damage states with their respective loss ratios. The proposed methodology is illustrated in a scenario where a flow-type loess landslide impacts a five-story reinforced-concrete (RC) frame building. The results indicate that the PDEM effectively estimates the exceedance probabilities with tolerable computational errors. Moreover, the vulnerability results link the failure probability of landslides and the cost of building damage within the QRA framework of individual landslides whose run-out affects nearby buildings.