A new approach, using stationary wavelet transform (SWT), is proposed for crack detection in beam-like structures. SWT is a redundant transform that doubles the number of input samples at each iteration, which can provide a more accurate estimation of the variances and facilitate the identification of salient features in a signal, especially for recognizing noise or signal rupture. The mode shape of a cracked beam with a small crack depth, while apparently a single smooth curve, actually exhibits a local peak or discontinuity in the region of damage. The mode shape ‘signal’ can be approximately considered as that of the intact beam contaminated by ‘noise’ which consists of response noise and the additional response due to the crack. The signal can be decomposed by SWT into a smooth curve, called approximation coefficient, and a detail coefficient curve, which includes crack information, respectively. In this paper, the modal responses of damaged simply supported beams are computed using the finite element method in conjunction with some experimental tests. The effect of noise on the proposed method is also studied. It is shown that SWT of the modal data of cracked simply supported beams provides a better crack indication than conventional discrete wavelet transform. The relationship between SWT detail coefficient and crack size (depth and width) are also discussed. A new method based on the average difference of the SWT detail coefficient of vibration bending modes of a cracked beam and an intact beam is proposed as a damage index and verified.
展开▼
