Modelling of ultrasonic guided waves for circular cylindrical structures using finite element approach and selection of noise filtering technique

Abstract

Development of effective multi-feature techniques of signal analysis is essential for guided wave based structural health monitoring in cylindrical shell structures; particularly in enormously noisy signals. The presence of random noise and narrow-band coherent noise, which are induced by extraneous modes and reflections, etc poses the challenge in all real guided wave based online crack detection testing. In the present investigation, interaction of Lamb waves with circumferential flaw is examined. Noisy signals are analyzed in the paper by using wavelet based denoising methods and matched filtering technique which 'sniffs' the signal buried in noise. The optimal wavelet is selected on the basis of mean square error (MSE) of noisy signal of multimode Lamb waves. Wavelet method is compared with matched filtering technique to establish the noise robustness. Various signal-to-noise ratio (SNR) are used for performance evaluation of filters. © Copyright (2016) by British Institute of Non-Destructive Testing All rights reserved.