Singal Processing Method Based on Wavelet Transform for Testing of 9% Ni Steel Welds Using Ultrasonic Time-of-Flight-Diffraction Technique

Abstract

This paper reports a signal processing application to signals obtained by the ultrasonic Time-of-Flight-Diffraction technique for 9% nickel steel weld joints. Obtained signals were decomposed into several components using the wavelet transform. The frequency bands of these components vary in width. To reduce grass (noise from diffraction through the dendrite), products of these components were calculated. The components were thus weighted according to these products, and A-scope signals were reconstructed. Flaw indications in D-scan images were found to be remarkably clear using this signal processing method for all signals obtained by D-scanning.