Rail hidden damage detection based on the modal curvature differences method

Abstract

The rail hidden damage includes rail bottom and rail jaw cracks, which cannot be easily detected during the early development stage, and pose a significant danger and risk of rail-breaking accidents. This paper proposed an effective method to detect rail hidden damage by establishing a rail crack extension analysis model based on modal curvature difference theory and verified by the self-developed device. The results showed the distance of the rail bottom crack from the central axis of the rail bottom and the peak value of the modal curvature difference represented an approximate one-dimensional Gaussian correlation. The crack depth and crack length showed an approximately linear correlation with the peak value of the modal curvature difference. Setting cracks at rail bottom and rail jaw, the detection errors were below 5% at a 6 mm crack depth and lower than 3% at a 10 mm crack depth. The results of signal-to-noise ratio during rail bottom crack detection indicated that crack could be effectively detected before rapid expansion when the crack distance from the rail bottom central axis exceeded 35 mm, and crack parameters could be quantitatively determined before rapid expansion at a crack distance exceeding 45 mm.