Damage detection in laminated composite plates via an optimal wavelet selection criterion

Abstract

Delamination is a potential risk of failure considered as one of the failure modes and frequently occurs in composites due to its relatively low inter-laminar fracture toughness. In recent years, the majority of activities in this field have been focused on raising the level of sensitivity of these devising methods for detecting tiny damages. In this article, damage detection method via wavelet transform has been examined, and an appropriate procedure has been proposed to increase sensitivity of this transform for damage detection. Among the inherent impediments of classical wavelet transforms, the generality of these transforms and ignoring the studied signal can be mentioned. Consequently, various wavelet selection algorithms leading to provide appropriate wavelet functions with respect to the characteristics of the signal have been examined. As a novelty in the field, the correlation between wavelet and strain energy signal is considered as a criterion for optimal wavelet selection. In wavelet transforms, in addition to original wavelet functions, the signals used for damage detection are also of high importance. To achieve this goal, the frequency-weighted strain energy ratio signals resulting from intact and damaged forms have been exploited. Also, the edges’ effects were removed through stringing of plane mode shape signals. Moreover, by summing wavelet coefficients in all scale factors plus natural frequencies, the focus can bring to the detection of one or more damages in a laminated composite plate with symmetric layup. Finally, a quantitative measure to compare different wavelets has been presented.