An edge-reflection minimum variance method of Lamb waves for damage contour evaluation

Array imaging methods combined with a Lamb wave transducer network are widely used for damage localization in NDT (Nondestructive Testing) and SHM (Structural Health Monitoring). Conventional Lamb wave imaging methods including the delay-and-sum, sparse reconstruction and MUSIC (Multiple Signal Classification) are based on a priori assumption of small-scale damage. However, since many types of structural damages (e.g. delamination in composite laminates) are large with multiple discretized edges, the unreasonable assumption will result in the performance reduction of damage imaging. To improve imaging performance, an edge-reflection imaging method is proposed in this study. For damages with non-negligible sizes, reflections occur at different edge positions from different transmitter-receiver pairs in transducer network. On this basis, the amplitude superposition of array signals is calculated by a modified time shifting rule to highlight damage edges for both damage localization and contour evaluation. During amplitude superposition, minimum variance algorithm is used to further improve the edge recognition accuracy. Compared to conventional methods, imaging performance is improved in proposed method. An experimental investigation is conducted on composite laminates, and the results demonstrate the high accuracy and performance improvement of the proposed method.