RAIL CRACK MONITORING USING ACOUSTIC EMISSION TECHNIQUE

Abstract

Fatigue cracks are the primary defects in rail track. Aiming at a smooth and safe operation of the railway system, this PhD research is focused on the rail crack monitoring using AE technique. In order to acquire various AE signals related to rail crack, fatigue tests were carried out on rail steel specimens in the laboratory, pencil lead break (PLB)-simulation and train-running tests were carried out on rail tracks in the field. The discrimination among different types of AE waves was achieved by a classification index based on wavelet power (WP). A novel crack quantification method using crack closure-induced AE waves was then brought forward. Meanwhile, the propagation characteristics of AE waves in rail track were investigated through wavelet analysis. Finally, a strategy based on synchrosqueezed wavelet transform (SWT) and Tsallis entropy was proposed to efficiently identify and locate cracks in the filed rail track with complicated crack conditions and high operational noise. The performance of proposed crack monitoring strategy was validated by the analysis results of field tests.