High-resolution imaging of microdefects detected by laser ultrasound with grouped adjacent wave correlation subtraction

Abstract

Laser ultrasound technology has become a key focus of nondestructive testing due to its non-contact, high-precision and suitability for harsh environments, but its multimode nature tends to lead to artifacts and noise in the imaging, reducing the resolution of defect detection. To meet the demand for high-resolution detection of sub-millimeter micro-defects, a laser ultrasound multi-mode total focusing method (TFM) imaging based on grouped adjacent wave correlation subtraction is proposed. Firstly, the full matrix capture (FMC) data are preprocessed by grouped adjacent wave correlation subtraction, grouped according to the relative positions of excitation-reception points, and the Pearson correlation coefficients of neighboring signals in the same group are calculated for signal phase subtraction, which effectively suppress the coherent noise caused by multimode characteristics. Further, the phase entropy coherence factor (ECF) is proposed to weight the TFM images to enhance the defect signals by quantifying the concentration of the phase distribution, which further reduce the effects of mode artifacts, background noise, and residual surface waves on the defect images, and further improves the defect resolution. Experiments are performed on 7075 aluminum alloy specimens with 0.3 mm and 0.5 mm side drilled hole (SDH) and blind hole (BH) defects. The imaging results show that multi-modal fusion imaging combined with ECF weighting after grouped neighbor wave coherent phase subtraction preprocessing significantly improve the defect resolution, and the relative error of defect size quantification is less than 10 %. The method proposed in this paper provides a reliable means for the quantitative detection of sub-millimeter micro-defects, which has important engineering application value.