Adaptive total focusing ultrasonic imaging of multidirectional CFRP laminates

Accurate defect detection in multidirectional carbon-fibre-reinforced polymer (CFRP) laminates remains challenging because of their strong anisotropy and heterogeneity. Conventional imaging approaches, such as the total focusing method (TFM), often suffer from degraded image quality due to inaccurate velocity estimation. In this study, an adaptive TFM imaging method for multidirectional CFRP laminates was proposed using limited a priori information. During imaging, the mean backwall amplitude is extracted from TFM images reconstructed using the anisotropic heterogeneous model (AHM), and the corresponding elastic constants are estimated to update the depth- and angle-dependent velocity profiles. The method was validated experimentally using a multidirectional CFRP laminate containing delamination defects at different depths. Compared with conventional backwall reflection method (BRM)-based calibration, the proposed approach accounts for depth-dependent variations in equivalent velocity, enabling improved defect imaging from the same full matrix capture (FMC) dataset. For reliably detected near-surface delaminations at depths of 2.5–3.3 mm, the defect amplitude increased by up to 29.0%, and the SNR improved by up to 3.3 dB. By avoiding the need for independent elastic characterisation of unidirectional laminates fabricated from the same material, the proposed method offers practical advantages for ultrasonic non-destructive evaluation of multidirectional CFRP structures.