Ultrasonic Testing Signal Enhancement by Piece-wise Deconvolution

A piece-wise deconvolution (PWD) technique for processing non-stationary ultrasonic testing (UT) signals is proposed in this paper. The signal was divided into a number of segments, and each segment, assumed to be semi-stationary, was processed by its corresponding reference wavelet. The reference wavelets were obtained through a multi-Gaussian beam modeling technique. The number of segments required in the PWD technique was estimated by finding the time-varying correlation between the modeled and measured signals. The effectiveness of the proposed approach was examined by applying it to both simulated and experimental data. The results indicate that, compared to conventional deconvolution techniques, PWD can significantly improve the time resolution of UT signals.