An efficient ultrasonic full-matrix imaging method for industrial curved-surface components defect detection

Abstract

Curved-surface components are ubiquitous in various industries. However, defects inside the components badly degenerate the mechanical properties and lead to premature failure. Ultrasonic full-matrix imaging is a prominent method in nondestructive testing, but there is still a lack of efficient imaging methods for cases with curved interfaces. To amend this problem, this article proposes an efficient ultrasonic full-matrix imaging method to detect the defects in the industrial curved-surface components. The wavefield extrapolation for full matrix capture (FMC) dataset is taken as the combined effect of the downward migration of the source and receiver wavefields in the wavenumber-frequency domain. Compensation terms corresponding to the source and receiver wavefields act on the wavefield of FMC in the spatial-frequency domain to make up for lateral variations of the sound speed. After the wavefield reconstruction, an efficient imaging condition is implemented to transform the wavefield into the defect image. The best compensation term is determined and the influence factors is investigated in the simulation experiments. Laboratory experiments furtherly validate the superiority of the proposed method over the existing methods, especially for the imaging efficiency which is twice of the phase shift migration (PSM) based method.