A novel pulse eddy current algorithm for steel defect detection under large lift-off variations

As a versatile non-destructive evaluation technique, pulsed eddy current (PEC) inspection is widely used in metal defect detection for structural health monitoring. However, for ferromagnetic conductors covered with non-uniform insulating layers, the stability of the eddy current response will be threatened by large lift-off variations. The main purpose of this research is to achieve reliable PEC defect detection by developing new signal analysis strategies, so as to enrich the application of the PEC technology. In this paper, the diffusion behaviour of eddy currents is tracked by the dynamic attenuation coefficient (DAC) of the PEC signal. By analysing the perturbation mechanism of defects in eddy current diffusion, a novel DAC method is proposed to investigate the geometric dimensions of the flaws. Moving beyond the conventional single-value attenuation coefficient β, the DAC strategy provides additional features for defect parameter identification. This paper assesses the critical performance metric of lift-off robustness for the identified features, and analyzes the adaptability of the proposed DAC strategy and linear fitting methods with low signal-to-noise ratio (SNR) PEC data, which becomes significant as the lift-off parameter increases. The feasibility of the DAC method is supported by experimental data under significant lift-off variations.