Rail defect detection based on mode wave packet separation algorithm of sparse point array

Abstract

The number of guided wave modes in irregular and complex cross-sectional structures is so abundant that it is easy to fail to separate the wave packets of each mode and to detect the reflected echoes during damage detection. In order to meet the needs of non-destructive testing and structural health monitoring for single mode analysis, this paper innovatively proposes a wave packet separation algorithm based on a sparse point array, establishes a practical structural defect detection workflow, and realizes for the first time the accurate identification and localization of rail defects under a few received points. In this paper, the frequency response matrix of the wave packet is constructed by the propagation characteristics of the waveguide, and the mode separation formula based on the sparse point matrix is derived by combining the pseudo-inverse algorithm. At the same time, the selection rule of the single-point excitation point position is determined, and the accurate propagation characteristics of the waveguide structure are obtained by the two-dimensional Fourier transform, and an optimal receiving distance calculation method is proposed according to the principle of waveform information discrimination. The structural defect detection workflow based on mode wave packet separation is further formed, and the separation of each mode wave packet and the reflection wave packet is realized. Finally, in the case where all the receiving points are arranged in the rail web, the simulation realizes the localization of the defects in the rail head, rail web, and rail base, as well as the identification of the defect size and longitudinal position, and is verified by experiment.