Optimal Noise Filtering Method for Lamb Waves in Identification of Crack in Cylindrical Structures

The presence of narrowband coherent noise and random noise, which are induced by extraneous modes and reflections, etc., poses the challenge in wave-based online crack detection testing. The goal of this investigation is to set up the powerful multi-feature methods of signal examination for issues of guided wave-based nondestructive technique (NDT) in barrel-shaped shell structures. Filtration of time-frequency signal of elastic waves through the noisy signal is explored in the present examination utilizing matched filtering technique (MFT) and wavelet denoising strategies. We likewise propose wavelet matched filter method (WMFM), a blend of the wavelet denoising and matched filtering technique, which can altogether enhance the exactness of signal peaks and distinguish relatively small damage, particularly in enormously noisy data. Also, the capabilities of wavelet and matched filtering methods are compared. To provide better insight for performance evaluation of denoising methods, processing of received Lamb wave signals with a low signal-to-noise ratio (SNR) is addressed. The interaction of guided waves with circumferential damage is analyzed by utilizing this method. Guided wave excitation and the communication law with the round indent in the hollow cylinder are considered through associating Lamb wave theory with finite element method.