Abstract: A genetic algorithm (GA) optimized TRL sparse total focusing method was proposed to address the challenges of low efficiency in ultrasonic total focusing method of thick-walled austenitic stainless steel welds, complex acoustic field of transmitter-receiver longitudinal wave (TRL) dual matrix array probe, and multiple factors affecting array sparsity. Based on the 2D array structure of the probe and the pseudo-focusing characteristics of the sound beam, combined with the acoustic radiation theory, this study analyzed the impact of sparse conditions along the active and passive axes on the directivity of the sound beam. A mathematical optimization model was established with the objectives of pointing function side-lobe level and main lobe width, and GA was used to design the array element layout of TRL probe on a 16×4 (row×column)constrained grid. The detection object was transverse through-hole defects with depth range of 10~70 mm and diameter of 3.2 mm on a 76 mm thick austenitic stainless steel weld test block. The phased array detection signals were collected for sparse total focusing method. The results showed that compared to full array, the defects detection signal-to-noise ratio of GA sparse imaging with 25% sparsity decreased by up to 3.29 dB, the image API only increased by 1.26, the imaging time decreased from 512.93 s to 56.63 s, and the efficiency improved by nearly 9 times.