Wavelet analysis to detect nonstationary and nonlinear behaviours in railway bridge

Abstract

Listen

Translate article

This paper demonstrates the potential of using the Continuous Wavelet Transform (CWT) for modal identification of an existing bridge. The dynamic response of a railway bridge crossed by high speed trains is investigated. First, the theoretical dynamic response of the bridge is studied and found to have a very distinctive shape due to the periodic spacing of the train axles. A CWT-based procedure is then proposed for processing the global signal, with emphasis on the choice of wavelet function. An application is given using experimental accelerometer data collected during the crossing of a high-speed train on the bridge. In a first step, the signal has been properly partitioned since it is composed of different time intervals containing specific modal information (in particular, bridge + train and bridge alone when the train has left the bridge). The dynamic response of the "bridge + train" system was found to have a very distinct non-stationary shape. The modal parameters of the bridge were then successfully calculated and non-linearities in the system behaviour were detected. It illustrates the ability of CWT to detect and characterise non-linear behaviour.