Lamb Wave Anomaly Detection by Ensembling Spatial and Wavenumber Domains

Abstract

Abstract Guided wave field anomaly detection has proven to be a feasible tool for nondestructive evaluation and structural health monitoring. These anomaly detection methods are often based on the fact that waves in damaged regions and undamaged regions propagate differently. Yet, most anomaly detection approaches are performed over a single domain, resulting in poor damage discrimination under certain conditions. By analyzing wave propagation in multiple domains, we should be able to achieve more robust detection. We present two methods for analyzing wave propagation in the spatial domain and wavenumber domain to detect and locate damaged regions. To validate our approach, we generate simulated guided wave fields and varied wave speed in two square-shaped regions, representing damage. In the spatial domain, we utilize pixel intensity change to detect anomalies. In the wavenumber domain, we cluster wave modes in polar coordinates using insights from solutions to the Helmholtz equation. By ensembling results from our detection methods in both domains, we can achieve more robust damage detection. We employ the intersection over union (IOU) score to quantitatively assess this approach. Our approach achieves an IoU score of 47% when utilizing only the spatial domain, with improvements to 78.5% upon model ensembling.