Bayesian time-frequency analysis of the vehicle-bridge dynamic interaction effect on simple-supported resonant railway bridges

Monitoring the conditions or damages of bridges under train passages demands a high accuracy modal-characteristic identification method that separates the apparent fluctuations caused by vehicle-bridge dynamic interaction (VBI) effects from other fluctuations. This study proposes a novel method based on a time-varying autoregressive model, which is solved using a hierarchical Bayesian estimation approach. The VBI effect is estimated from the displacement response of the railway bridges as temporal fluctuations of the natural frequency and modal damping ratio. The exogenous variable is the train load, expressed as an external force. Numerical experiments verified the higher accuracy of the proposed method than the existing method. The influences of train speed and rail irregularity on the VBI effects are clarified by the application of the proposed method to various VBI simulations. The proposed method was applied to the measured resonance responses of actual bridges and succeeded in empirically demonstrating the decreased natural frequency and the increased modal damping ratio under train passage. Additionally, using the proposed method, modal characteristics variation due to VBI effect calculated using VBI model simulation was verified by comparing with those estimated from the measured results. (C) 2019 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).