Damage quantification in composite laminates based on sparsity-difference estimation of the normal time–frequency representation

Abstract

Damage index (DI) is of vital importance for the health monitoring of composite structures. However, most of the existing DI construction methods cannot reveal the rich information hidden in Lamb waves, such as the time-varying characteristics of the signal amplitude and phase, and thus cannot well characterize the evolutionary trend of damage. To address this limitation, this study proposes a novel DI construction method for damage identification in composite structures. The normal time–frequency transform is first applied to obtain the unbiased time–frequency distribution of the received Lamb wave signal, and then the singular value decomposition method is used to obtain the singular value sequence of the two-dimensional time–frequency matrix to characterize its matrix sparsity. Next, the dynamic time warping method is used to measure the distance between singular value sequences of the current and baseline signals. A quantitative DI is finally defined using these distances. Two experimental studies are utilized to validate the effectiveness of the proposed method. Results show that the proposed method can reveal the rich information hidden in the Lamb wave signals and has better trend-matching performance in fitting the actual damage propagation trend than traditional methods.