Neural classification of Lamb wave ultrasonic weld testing signals using wavelet coefficients

This paper presents an ultrasonic nondestructive weld testing method based on the wavelet transform (WT) of inspection signals and their classificatio n by a neural network (NN). The use of Lamb waves generated by an electroma gnetic acoustic transducer (EMAT) as a probe allows us to test metallic mel ds. In this work, the case of an aluminum weld is treated. The feature extr action is made by using a method of analysis based on the WT of the ultraso nic testing signals; a classification process of the features based on a ne ural classifier to interpret the results in terms of weld quality concludes the process. The aim of this complete process of analysis and classificati on of the testing ultrasonic signals is to lead to an automated system of w eld or structure testing. Results of real-world ultrasonic Lamb wave signal analysis and classifications for an aluminum weld are presented; these dem onstrate the feasibility and efficiency of the proposed method.