# Experimental and Localisation Method for Non-Destructive Testing of Cable Corrosion Based on Weak Magnetic Imaging

## TL;DR

This paper introduces a new non-destructive method using weak magnetic imaging to detect and locate corrosion in steel cables.

## Contribution

A novel weak magnetic imaging method using adaptive threshold K and linear interpolation for corrosion detection in cables.

## Key findings

- Axial corrosion length shows a strong linear relationship (R2 = 0.998) with the peak-to-trough distance of Bz.
- Circumferential corrosion width correlates with the circumferential distribution range of Bz.
- Adaptive threshold K effectively delineates corrosion boundaries for precise localization.

## Abstract

In order to address the challenge of accurately identifying the extent of corrosion in parallel steel wire cables, a series of corrosion detection tests were performed on parallel wire bundles with varying degrees of corrosion. Following the collection of weak magnetic signals from a 12-channel Hall array, the influence patterns of corrosion severity on the distributions (Bx, Bz) of leakage magnetic signals were analysed. The experimental results were validated by means of a three-dimensional finite element model, leading to the proposal of a novel weak magnetic imaging method based on the fusion of adaptive threshold K and linear interpolation. The findings of the study demonstrate a strong linear relationship (R2 = 0.998) between axial corrosion length and the peak-to-trough distance of the normal component Bz. Furthermore, it was determined that a positive correlation exists between the circumferential corrosion width and the circumferential distribution range of Bz. The utilisation of an adaptive threshold K for the purpose of threshold segmentation has been demonstrated to be an effective method for the delineation of corrosion boundaries, thereby enabling precise localisation. The present research provides technical support for the visualisation and quantitative assessment of cable corrosion.