Lamb waves have long been recognized to be a valuable tool in the nondestructive testing of plates and shells. The patterns of dispersion which characterize the propagation of Lamb waves through a material may be used to determine the thickness of a plate, the elastic properties of materials, as well as the condition and integrity of the sample. Unfortunately, the analysis of Lamb wave signals is complicated by the existence of at least two modes at any given frequency. To efficiently determine the dispersive characteristics of a plate, it is therefore necessary to have a technique for analyzing broadband multimode signals. Methods for the time-frequency analysis of signals are especially useful in determining dispersion curves, since they reveal the time variation of each frequency component of the signal. The wavelet transform (WT) is one method for time-frequency analysis which has proven to be an excellent alternative to the Short Time Fourier Transform (STFT), particularly in the analysis of extremely broadband signals. It has, in fact, been used to determine the dispersion curve of a flexural wave in a beam from a single broadband signal. Aside from time-frequency analysis, the wavelet transform has also been used with great success for other applications such as denoising, data compression and feature detection.
展开▼
