Ultrasonic Defect Characterization in Coarse-Grained Materials Using SHADE-A-TFM With Subarray-Based Parameter Optimization

This paper proposes an innovative ultrasonic imaging methodology, SHADE-A-TFM, for accurately characterizing crucial defect parameters such as size and orientation angle in complex materials by integrating success-history adaptation differential evolution (SHADE) with the adaptive total focusing method (A-TFM). Firstly, the proposed A-TFM shows enhanced capabilities over conventional TFM, providing greater flexibility to handle grain noise, which lays a foundation for subsequent defect characterization. Secondly, a subarray-based parameter reduction technique is employed, streamlining the optimization process and offering a more physically plausible simplification than existing methods such as the receiver-optimized TFM (ROTFM). Thirdly, the problem is formulated as a constrained optimization problem aiming to maximize the defect signal-to-noise ratio (SNR), with SHADE serving as the key optimization algorithm. The SHADE-A-TFM approach proves highly effective in reconstructing clear TFM images from noise-distorted ones, and its performance is evaluated through simulations and experiments. In simulations, the results obtained using the SHADE-A-TFM for imaging and the 6-dB method for defect characterization demonstrated outstanding accuracy and robustness. Experimental results from 12 crack-like defects further validated the superiority of this method, as it consistently achieved lower error rates in estimating defect parameters, providing a reliable and accurate solution for defect analysis in complex materials.