Subsurface polycrystalline microstructure reconstruction via immersion ultrasound full waveform inversion

Microstructural mapping of polycrystalline metallic alloys is essential for predicting their macroscopic mechanical performance. Among existing techniques, non-destructive subsurface imaging offers a promising but technically challenging pathway for advancing the characterization of metallic materials. This study introduces a water-immersion ultrasound full waveform inversion (FWI) framework for nondestructive reconstruction of subsurface polycrystalline microstructure. By integrating spectral-element wavefield simulations with adjoint-based inversion, this method simultaneously recovers anisotropic elastic coefficients and crystal orientations. Benchmark tests on Ti-6Al-4V alloy models demonstrate that the proposed method accurately images grain orientations and anisotropic elastic coefficient distributions. It achieves high-resolution reconstructions of polycrystalline microstructures and outperforms direct-contact ultrasound methods in accuracy and artifact suppression. Further applications to equiaxed polycrystals and microtextured regions verify the robustness and applicability of the approach.