Work

3D Defect Image Reconstruction of Eddy Current Magneto-Optical Non-destructive Testing Based on Digital LIA

Year: 2025

Type: conference-paper

Abstract: The eddy current magneto-optical non-destructive testing (ECMO NDT) system offers advantages such as high detection speed, precision and resolution, making it widely used in the NDT of weakly conductive materials. In recent years, ECMO NDT technology based on digital lock-in amplifier (LIA) signal extraction has become increasingly mature. However, it is limited to revealing the surface contours of defects, while reconstructing the 3D morphology of defects remains a key research focus. In this work, we propose a method for reconstructing 3D images of material defects using polynomial fitting combined with cubic spline interpolation. The experiments were conducted on an ECMO NDT platform, which measured the amplitude and phase information of the eddy current magnetic field of the material. The 3D images of the defects were reconstructed based on the detected amplitude and phase information. The amplitude and phase data corresponding to the 3D shapes of different defects were treated as prior information for the fitting model. We incorporated the electromagnetic patterns derived from simulation phenomena, which describe the effect of defects on eddy currents, as constraints for the fitting process and introduced a regularization term to prevent model overfitting when prior data is limited. The cubic spline interpolation method was employed to enhance image resolution, allowing the system to reconstruct higher-resolution 3D images within a shorter detection time. Experimental results demonstrate that the reconstructed images of various defects in different materials are clear, providing detailed depth information. Compared to traditional 2D imaging methods that rely solely on the amplitude or phase of the eddy current, our method, which combines both amplitude and phase information, offers a more comprehensive representation of defect size and shape.

DOI: 10.58286/30973

Authors: Jin Shi, Kun Xu, Li Wang

Institutions: University of Electronic Science and Technology of China, Henan University of Technology

Language: English

Topic: Non-Destructive Testing Techniques

Subfield: Mechanical Engineering

Field: Engineering

Domain: Physical Sciences

Open Access status: gold

Funder: National Natural Science Foundation of China

Award: G6974142195

OpenAlex ID: W4409730792