Probabilistic evaluation approach for nonlinear vehicle-bridge dynamic performances

Railroad vehicle and bridge coupled lateral vibration problems are traditionally solved through detailed nonlinear models in time domain using limited samples to represent rail irregularity. Ideally, a random vibration and reliability based approach should be implemented because of the random nature of the excitation process. In this study, vehicle-bridge coupled dynamic equation was derived using the principle of virtual work utilizing a linearized wheel-rail contact equation. This simplification enables the calculation of the system random lateral responses through the pseudo-excitation method. By applying rail irregularity as random excitations to the system, this study utilized an explicit linearization method to avoid iterative solution at each time step of the integration. The results from the linearized method were validated through comparison with results obtained from Monte-Carlo simulations. By applying the linearized approach to probabilistic assessment of the vehicle-bridge system reliability, it was shown that system probability of exceedance of admissible limits increases with train speed and reduces with increased bridge self-weight. It is concluded that the proposed approach provides a viable efficient alternative to investigate the random dynamic characteristics of vehicle-bridge system especially in the lateral direction, which is dominated by the random rail irregularities.