Fully automated orientation relationship calculation and prior austenite reconstruction by random walk clustering

Two new methods, one for determining the experimentally observed Orientation Relationship (OR) and another for reconstructing prior austenite phase, are proposed. Both methods are based on the angular deviation of the OR at the grain boundaries. The first algorithm identifies the optimum OR using the misorientation distribution of the entire scan i.e. without manual selection of parent grains. The second algorithm reconstructs the parent phase using a random walk clustering technique that identifies groups of closely related grains based on their angular deviation of the OR