Dynamic Modelling and Acceleration Signal Analysis of Rail Surface Defects for Enhanced Rail Condition Monitoring and Diagnosis

Abstract

Early detection of rail surface defects (RSDs) is crucial for timely rail maintenance, repair, or replacement to prevent potential risk of rail breaks and train derailments. Axle box acceleration (ABA) measurement is commonly used to identify rail and track irregularities. This paper aims to (1) model and simulate three common RSDs (rail corrugation, spalling, and squat) with three severity levels (light, moderate, and severe); (2) process and analyse simulated RSD-driven ABA signals; and (3) ascertain relationships between RSDs and their corresponding ABA signals. Explicit dynamic finite element models of wheel-rail interaction were created for rails with and without RSDs. Simulated RSD-driven ABA signals were subsequently processed and analysed using various signal processing algorithms including short-time Fourier transform, continuous wavelet transform, empirical mode decomposition, and power spectral density. Results reveal the likelihood of automatically detecting and classifying different types and different severity levels of RSDs from ABA signals, thereby enhancing the efficiency and effectiveness of rail condition monitoring and diagnosis.