A novel hybrid wavelet transform for detecting damages in laminated composite cylindrical panels

This paper proposes a hybrid wavelet transform (HWT) to detect local damages in the laminated composite cylindrical panels. This version of wavelet transform is developed to compensate for the weakness of two-dimensional continuous wavelet transforms (2D-CWTs) that are not sensitive to the local abrupt changes in higher scales. The idea is to combine two layers of the 2D-CWT and one layer of the image edges extracted from the first layer of the 2D-CWT. Thus, the novelty of the present work is to propose a new architecture to improve the accuracy of 2D-CWT in damage detection tasks. The finite element method is used to obtain the numerical vibration mode shapes as two-dimensional signals of the damaged laminated composite cylindrical panels. Experimental investigations are conducted to verify the performance of the proposed numerical model. The results indicate that using edge detection algorithms significantly contributes to increasing the accuracy of damage detection using 2D-CWT. The proposed hybrid wavelet transform method greatly enhances damage detection accuracy, particularly for low-level damage. Findings show that the hybrid wavelet transforms can improve the resolution and accuracy of damage detection results of the laminated composite cylindrical panels.