Anomaly Imaging using Decomposed Lamb Wave Modes

Abstract

This paper presents a tomography-based method for structural health monitoring (SHM) using guided Lamb waves. The method assumes an active SHM system in which the structure is excited sequentially using a set of piezoelectric transducers attached to it. The resulting waveforms are collected at the remaining sensors on the structure for further analysis. The algorithm compares waveforms collected at inspection time to baseline signals corresponding to the pristine state of the structure and defines an appropriate anomaly index. The anomaly index used in this work is based on the attenuation of the first arriving mode component. The damage indices associated with all actuatorsensor pairs form the input to an algebraic reconstruction tomography algorithm that reconstructs an anomaly map for the structure. The reconstruction is performed using a sparse recovery method known as basis pursuit denoising assuming, that the anomalies are relatively small compared to the size of the structure. The reconstructed map is further processed to estimate the boundaries of the anomaly on the structure. Experimental results for estimating the outline of aluminum masses attached to a unidirectional composite plate are presented in this paper. These results indicate that our approach works better than traditional approaches in accurately estimating the extend of anomalies outline in composite structures.