Early Warning of Abnormal Train-Induced Vibrations for a Steel-Truss Arch Railway Bridge: Case Study

Abstract

Considering the new challenges for high-speed railway bridges, the early warning of abnormal train-induced vibrations is necessary for ensuring the operation safety of both the bridge structures and the trains on the bridge. In this study, an online monitoring system for detecting abnormal train-induced vibration responses is developed, and the Dashengguan Yangtze River Bridge is used for illustration. First, to accurately investigate the influence of different train lanes and the number of carriages on train-induced vibrations, the speed-acceleration (train speed-bridge acceleration) correlations under different loading cases are obtained using an online identification method. Then, a two-stage method for early warning of abnormal train-induced acceleration responses of the bridges is developed using wavelet packet decomposition and interval estimation theory. Finally, the early warning method for identifying abnormal train-induced transverse vibrations is presented. The results show that (1) the train lane and the number of carriages affect the speed-acceleration correlations, and the identification of loading cases is needed for the accurate monitoring of speed-acceleration correlations; (2) by using wavelet packet decomposition, the median line of speed-acceleration correlations can be optimally extracted, and the early warning thresholds for abnormal train-induced acceleration responses can be properly determined using the interval estimation theory compared with the point estimation theory; and (3) the train running parameters of the Dashengguan Yangtze River Bridge are all within safe limits, but the wheel unloading rate and derailment coefficient have reached 60% of the limits due to the train-induced transverse vibrations. The effects of train-induced transverse vibration on the train running stability is worthy of attention.