Subsurface polycrystalline reconstruction based on full waveform inversion - A 2D numerical study

Microstructural mapping of polycrystalline metallic alloys is a key component in predicting macroscopic material behavior, and non-destructive subsurface polycrystalline imaging is a valuable yet challenging field that offers promise for metallic material characterization. This paper proposes a computational ultrasonic imaging method for high-resolution, subsurface polycrystalline reconstruction using signals measured on a sample's boundary and a full waveform inversion technique. For this 2D numerical study, the anisotropic elastic coefficient parameterization is introduced and connected to the corresponding polycrystalline orientation. The forward wave propagation simulation is performed using a spectral finite element method. An inversion framework is developed for inverting the anisotropic elastic coefficients, and the crystalline orientations. Reconstruction performance benchmarking is systematically conducted for both rectangular regions and a representative polycrystal. Overall, the reconstructions have broad correspondence to the scanned, unknown models. We further discuss artifact mitigation strategies and the potential to extend this work.