ABSTRACT
Nowadays, the operational speed of the trains is increasing significantly. In this paper, a train-bridge stochastic system is built to assess the reliability of railway bridges for high-speed traffic. The track quality and system parameters are considered in the assessment by considering the acceleration response of the bridge deck. The uncertainties within this coupling stochastic system are identified and modeled by using appropriate distributions, and the Latin Hypercube sampling is adopted to generate the random samples. The uncertainty of track quality is considered via random rail profiles. Specifically, the probabilistic rail irregularity samples are generated based on the rail irregularity power spectrum density function. Subsequently, a developed train-bridge model is adopted to calculate the stochastic dynamic responses of the coupling system. A representative railway bridge subjected to high-speed trains is investigated to assess the efficiency and accuracy of the built system. The system parameters, e.g., the concrete elasticity modulus, the mass of the car body, and the vertical stiffness of the 1st suspension of the train, are selected to evaluate the effects of random variables and rail irregularity on structural reliability.
