Automated reconstruction of pre-transformation microstructures in zirconium

An automated reconstruction of the pre-transformation microstructure from the microtexture data of the post-transformation product phase is proposed. The method involves identifying triplets of neighboring product grains with a common variant and linking such neighboring triplets via a generalized misorientation criterion. The approach is non-iterative and extremely efficient computationally. The method was tested successfully for different post-transformation microstructures in zirconium.