Time-dependent fatigue reliability analysis of heavy-haul railway steel bridges based on coupled train-track-bridge dynamic analysis

Abstract

Due to the severity and sudden onset of damage, fatigue failure in heavy-haul railway bridges has become a critical concern for all stakeholders. This study develops a comprehensive time-dependent fatigue reliability framework based on stress range analysis, where stress ranges, regarded as the primary drivers of fatigue damage, are systematically characterized for their stochastic and dynamic nature. The developed framework integrates three key innovations: (1) implementation of a coupled train-track-bridge system-based stochastic dynamic analysis, enhancing conventional static methods to more accurately characterize the dynamical development of the stress range; (2) development of a continuous nonstationary stochastic process model for the stress range that surpasses traditional stationary assumptions in realism; and (3) application of a fourth-moment transformation method for time-dependent reliability analysis involving nonstationary stochastic processes. The proposed framework is validated through application to an actual heavy-haul railway steel bridge, revealing that both the stress range and the fatigue limit state function exhibit time-dependent, nonstationary, and non-Gaussian characteristics. Notably, incorporating dynamic analysis under a nonstationary assumption proves to be critical, as conventional methods tend to overestimate fatigue reliability. This study provides valuable insights for engineers in the design, assessment, and management of railway bridges under stochastic fatigue loads.