Austenite residual stress and lath martensite variant selection in low carbon steels

Abstract

Controlled thermomechanical processing, in the austenite phase of two different steel grades, produced 15 different martensitic microstructures. The variant selection and the hierarchy of martensite were affected. Both, hierarchy and variant selection of the predominantly lath martensite, scaled with experimental bulk martensite residual strains (${\alpha }_{RS}^{\prime }$). Local ${\alpha }_{RS}^{\prime }$ and martensite crystallography, on the other hand, appeared to provide an elastic strain energy minimization. A novel methodology of estimating residual strain in the ‘invisible’ austenite phase was then proposed and validated. This approach combined pixel-by-pixel prior austenite reconstruction with dynamical high-resolution Kikuchi pattern simulations. Pattern simulation-based Phantom strain(s), considering only pattern rotation, scaled linearly with experimental, micro-Laue diffraction and microtexture based, lattice strain(s). This study was then extended to a residual stress assisted, and interaction energy based, martensite variant selection model. It was clearly shown that the austenite residual stress, estimated from pattern simulations of reconstructed austenite, controlled the martensite variant selection.