A two-dimensional wavelet transform for detection of cracks in plates

In this paper, a two-dimensional wavelet transform is used for identification of cracks in plate structures. The vibration mode of a cracked plate is transformed to the wavelet domain and both the location and extent of the crack are accurately displayed in the transformed response. Quantitative estimation of the crack depth is performed using both the maximum value and the energy content of the wavelet coefficients. The feasibility of the proposed method is demonstrated through simulation examples, which involve plates with cracks of various depths at different positions. The overall estimation errors for the crack location and length are found equal to 0 and 0.0673, respectively. The influence of the added noise on the estimation accuracy of the method was systematically investigated using a noise stress test. The proposed method is attractive due to the simplicity of its computational implementation and the accuracy of the results. (c) 2005 Elsevier Ltd. All rights reserved.