Time-Frequency Analysis of Axle Box Accelerations in Railway Systems: Experimental Insights on Ballasted Tracks and Bridge Transition Zone

Railway track structural health monitoring (SHM) is becoming increasingly important due to the growing demand for intensive train operations and higher standards of passenger comfort. Axle box acceleration (ABA) signal, being one of the most commonly acquired data sources during train operations, has been widely used to monitor track substructures. Yet, the underlying mapping relationship between ABA signals and railway structural dynamics remains unclear. The contribution of this paper lies in exploring the connection between the frequency components of ABA signals and eigenfrequencies of train-track dynamics using the time-frequency analysis method. To be specific, this paper first introduces the short-time Fourier transform (STFT) and continuous wavelet transform (CWT) methods and applies them as time-frequency analysis tools to the ABA signals. It then outlines the approach for assessing frequency components associated with the train-track dynamics and explores the muti-tone-harmonics interaction behaviors. Finally, the paper presents case studies based on experiments conducted on a railway line. The results demonstrate a noticeable interaction between the muti-tone-harmonics of sleeper passing frequency fsp and wheelset rotation frequency fwr. In the cases presented in this paper, the observed frequency couplings follow the empirical relationship fc= C1fsp±0.5C2fwr, and this coupling principle remains consistent regardless of varying train speeds and substructure types.