Efficient Post Compression Techniques for ACA Algorithm Accelerated Eddy Current NDT Solver

This article proposes the efficient post compression techniques to address redundancies existing in adaptive cross approximation (ACA) algorithm accelerated eddy current nondestructive testing (ECNDT) solver. The techniques include singular value decomposition (SVD) and kernel truncation (KT) algorithms, where SVD algorithm optimizes non-orthogonal columns after ACA compression and KT algorithm truncates the trivial interactions, both in the distant block interactions of the impedance matrix. Taking the ECNDT benchmark problem as the testing case, the correctness and efficiency of the suggested method are confirmed by comparing the results anticipated by the post compression techniques for the ACA algorithm-based solver with those predicted by other methods