Wavelet-transform-based method of analysis for Lamb-wave ultrasonic NDE signals

A new acoustic nondestructive method using Lamb waves as a probe is presented. These waves are generated and received by an ElectroMagnetic Acoustic Transducer (EMAT). The position of flaws in the structure under test is computed from the time of arrival of the main peak of the reflected signal. Due to the noisy nature of the received signal, we use a wavelet transform algorithm to extract the required time information. The main advantage of such a multi-scale method of signal analysis is to be suitable for peak detection problems especially in highly noisy environments. We explain how we proceed to do the feature extraction, and we propose two methods for reconstructing the image of the inspected structure. Results of real-world ultrasonic Lamb waves signal analysis are presented. In addition, to test the noise robustness of the method, the case of synthetic noisy signals is also treated.