Crack detection in simply supported beams without baseline modal parameters by stationary wavelet transform

摘要: There are significant changes in the vibration responses of cracked structures when crack depth is comparison to structure. This fact enables identification cracks from their response data. However relatively small, it difficult identify presence by a mere observation A new approach for detection beam-like presented and applied simply supported beams this paper. The based on finding difference between two sets detail coefficients obtained use stationary wavelet transform (SWT) mode shape data These data, which constitute signal series, reconstructed modal displacement beam. They represent left half modified right SWT redundant that doubles number input samples at each iteration. It provides more accurate estimate variances scale facilitates salient features signal, especially recognising noise or rupture. well known beam containing small apparently single smooth curve like an uncracked However, actually exhibits local peak discontinuity region damage. Therefore, ‘signal’ can be approximately considered as contaminated ‘noise’, consists additional due crack. Thus, decomposed into curve, called approximation coefficient, coefficient. series includes information useful damage detection. damaged used computed using finite element method. For real cases, affected experimental noise. with normally distributed random also studied. results show proposed method has great potential does not require parameters baseline effects size, location, sampling interval examined.