A machine learning approach to map crystal orientation by optical microscopy

Abstract Mapping grain orientation in crystalline solids is essential to investigate the relationships between local microstructure and crystallography interpret materials properties. One of main techniques used perform these studies electron backscatter diffraction (EBSD). Due limited measurement throughput, however, EBSD not suitable for characterizing samples with long-range heterogeneity, nor building large material libraries that include numerous specimens. We present a machine learning approach high-throughput crystal mapping, which relies on optical technique called directional reflectance microscopy. successfully apply our method Inconel 718 specimens produced by additive manufacturing, exhibit complex, spatially-varying microstructures. These results demonstrate mapping metal alloy achievable. Since data-driven, it can be easily extended different systems using manufacturing processes.