Global reliability analysis of running safety of a train traversing a bridge under crosswinds

The derailment of any wheel of a running train results in failure of the safe operation of the train. Therefore, the running risk of each wheel must be comprehensively considered when calculating the running safety and reliability of a train. In this study, a methodology for calculating the global reliability of the running safety of a train on a bridge under crosswinds is proposed using the probability density evolution method (PDEM)-based equivalent extreme value principle. First, a refined 3-D train-track-bridge interaction model is constructed to accurately obtain the dynamic response of a train on a bridge under crosswinds. Second, a dimension-reduced scheme is developed to simulate the representative samples of turbulent winds and track irregularities so that the stochastic processes can be expressed using only two random variables. Third, the PDEM-based equivalent extreme value principle is proposed to efficiently calculate the global reliability of the running safety of trains, where four running safety indexes, namely the derailment coefficient, wheel unloading rate, and wheel-rail vertical and lateral relative displacements, are used to evaluate the running safety of the train. Using numerical examples, the accuracy of the proposed methodology is verified through comparison with the field measurement data and Monte Carlo method, and the influence of the number of representative points on the accuracy is investigated. Furthermore, the differences between the global reliability and single-wheel reliability as well as those between the time-dependent and time-independent reliabilities of running safety are discussed in detail.

Automated summary

This study proposes a methodology for calculating the global reliability of the running safety of a train on a bridge under crosswinds using the probability density evolution method (PDEM)-based equivalent extreme value principle. The accuracy of the proposed methodology is verified through numerical examples, and the differences between the global reliability and single-wheel reliability, as well as the differences between time-dependent and time-independent reliabilities, are discussed in detail.