A time-frequency analysis algorithm for ultrasonic waves generating from a debonding defect by using empirical wavelet transform

Time-frequency distributions (TFDs) are key indicators for inspecting debonding defects using ultrasonic nondestructive testing (NDT) or nondestructive examination (NDE). A novel algorithm combining empirical wavelet transform (EWT), Hilbert transform (HT) and short time Fourier transform (STFT) is presented to obtain the TFD of ultrasonic testing waves. First, we decompose the signal into mono-components by means of EWT to satisfy the mono-component requirement of HT. Next, correlation coefficients between the empirical modes and the decomposed signal are computed to extract valuable components. Further, the instantaneous amplitude (IA) of the extracted valuable components are calculated using HT. Finally, using the computed IAs, the peak time representing the arrival time of the component is obtained. To calculate the corresponding frequency, a method based on STFT is applied on the valuable components. The principle of the proposed method is illustrated by a numerical signal. The effectiveness of the proposed method in extracting the TFD of ultrasonic testing wave is validated by an experimental analysis of the ultrasonic NDT used for the detection of debonding defect in a composite construction. In comparison with STFT, smoothed pseudo Wigner-Ville distribution (SPWVD), and empirical mode decomposition (EMD), a higher accuracy of the TFD can be obtained using the proposed algorithm.