Fault Diagnosis of Steel Wire Ropes Based on Magnetic Flux Leakage Imaging Under Strong Shaking and Strand Noises

Abstract

Because of its flexibility, high strength, and durability, steel wire rope (SWR) is widely used in irrigation works, bridges, harbors, tourism, and many industrial fields as a vital component. Thus, it can cause accidents and economic losses if local flaws (LFs) of the SWR in service are not detected in time. This article points out two major problems in magnetic flux leakage (MFL) imaging-based nondestructive testing for fault diagnosis of SWR and proposes an integrated signal-processing method specifically designed for addressing the two problems. In this article, the MFL signals are collected by a detector that is formed by a set of permanent magnets and a Hall sensor array. Based on these multichannel MFL signals obtained from the Hall sensor array, we use the principle of multichannel signal fusion to determine rich information from all MFL signals. We solve the strand noise problem by an oblique-directional resampling and filtering method, which avoids severe attenuation in the LF signal. Moreover, the shaking noise is effectively removed by the proposed antishaking filtering based on the median filter. According to our simulation and experiment, the proposed fault diagnosis method for SWR significantly improves the performance of LF detection and localization under strong shaking and strand noises.