A Novel Ultrasonic Guided Wave-Based Method for Railway Contact Wire Defect Detection

Abstract

Detection of defects in contact wire is significantly important to railway system. Due to the insensitivity to the deep-seated defects, traditional optics-based methods fail to detect the internal defects of contact wire. To overcome this shortcoming, a new defect detection method based on ultrasonic guided wave is proposed in this paper. Theoretical dispersion curves and wave structures for the TCG-110 contact wire are derived using a special modeling technique called the semi-analytical finite element (SAFE) method. It is shown that the longitudinal-like M3 mode in the frequency range of 50-100 kHz is attractive for the proposed application. Then, the experimental scheme is established based on the pulse-echo configuration to detect the defects in contact wire. The echoes overlapping and drown problems in guided wave signal leading to the time resolution of original signal is too low to extract useful position information of defects. Hence, the sparse deconvolution method is introduced based on the convolution model of guided wave signal and sparse theory, and the time resolution of guided wave signal is significantly improved. The experimental results show that the proposed method can detect the defects in contact wire effectively with high accuracy.