Probabilistic assessment of vehicle derailment based on optimal ground motion intensity measure

Abstract

Railway vehicles frequently have derailed during major earthquakes. Although a probabilistic approach based on fragility analysis was established in structural engineering, it has never been applied to railway vehicles. Thus, the aim of this study was to select optimal intensity measures (IMs) and develop fragility curves for the derailment of vehicles. Eight commonly seen IMs were examined to determine their suitability for use in analyzing the derailment of vehicles. The railway vehicle response was analyzed using a nonlinear vehicle-track simulation, and 22 far-field and 28 near-field ground motion records were used as input. Based on the results of the simulation, the correlations of these IMs with the responses of vehicles were evaluated in terms of commonly accepted criteria in the literature, such as efficiency, practicality and sufficiency. It was found that spectral IMs were superior to peak-type IMs. In addition, the acceleration spectral intensity (ASI) and the velocity spectral intensity (SI) were shown to be optimal IMs for the evaluation of derailments. Using the selected IM (ASI), derailment fragility curves were developed for three levels of vehicle performance, which facilitated the probabilistic assessment of vehicle derailment in a performance-based manner.