[2A4] High-accuracy array ultrasonic imaging of internal defects in multi-layered curved components based on frequency-domain full wave inversion and circular coherence factor
T Zhu, J Yu, Y Zhou, Z Zhou and W Zhou
Beihang University, China
Multi-layer curved structural components are critical load-bearing elements in power equipment and aerospace systems, where accurate detection of internal defects is essential to ensure structural integrity and operational safety. However, conventional array ultrasonic testing methods often suffer from limited imaging resolution and low signal-to-noise ratio (SNR), primarily due to acoustic impedance mismatches at interlayer interfaces and wavefront distortion induced by surface curvature. This study presents a novel defect imaging approach that combines frequency-domain full waveform inversion (FWI) with a circular coherence factor (CCF) enhanced total focusing method (TFM) for multi-layer curved structures. A logarithmic objective function based on Green’s function reconstruction is formulated, enabling blind inversion of acoustic velocity fields without requiring a priori knowledge of the source wavelet. Additionally, a modified pseudo-Hessian matrix incorporating Green’s function correction is introduced to enhance imaging fidelity and convergence stability, particularly in deeper regions. Following structural inversion, image thresholding and acoustic velocity modelling are employed and CCF-TFM is applied to suppress multiple reflection artefacts and improve SNR in the final image. Experimental validation using a three-layer curved medium demonstrates that the proposed method achieves accurate velocity reconstruction and high-resolution defect imaging. Compared to conventional techniques, the proposed method improves inversion accuracy by 59.55% and enhances image contrast by a factor of 3.98. This integrated approach offers a promising pathway for high-fidelity, non-destructive evaluation of complex multi-layer curved components in advanced engineering systems.
