RANDOM DYNAMIC ANALYSIS OF TRAIN-BRIDGE SYSTEM INVOLVING SYSTEM PARAMETERS WITH PROBABILITY DENSITY ENVOLUTION METHOD

Random dynamic characteristic of simple supported girder bridges is important in high speed railway, the dynamic vibration problems caused by random excitation (i.e. track irregularity) and structural parameters (i.e. Young's modulus) could not be neglected. Considering the randomness of track irregularity and Young's modulus of the concrete, a 3D train-bridge model is established to study the system vibration characteristic. The probability density evolution method (PDEM), which is capable of capturing the instantaneous probability density function (PDF) of the dynamic response and its evolution, is employed at the train-bridge system vibration analysis. Newmark-beta integration method and the bilateral difference method of TVD (Total Variation Diminishing) schemes are proposed to solve the train-bridge dynamic equation. Results show that, compared to the Monte Carlo method (MCM), PDEM is capable for the train-bridge vibration analysis and has higher accuracy and computational efficiency. By parametric analysis, track irregularity is important factor causing random train-bridge vibration. Young's modulus of bridge have greater influence than track irregularity on the random bridge dynamic analysis, while it can be almost neglected on the vehicle dynamic analysis.