Abstract
The Helming method is employed to improve the definition of the orientation classes used to define the crystallographic texture in abnormal grain growth simulation. It involves fitting the experimental texture measured by X-ray or neutron diffraction by several orientations modelled by Gaussian functions. The orientation classes are then used to describe the distribution of grain boundary energy on experimental microstructures characterized by EBSD (Electron BackScattered Diffraction) and introduced as input data into the simulation. The approach is tested in the case of the Monte Carlo simulation of abnormal growth of Goss grains in a Fe3%Si sheet. The results are compared with those obtained through classical methods which arbitrarily impose spreading around the main texture components. Finally, it appears that the evolution of their volume fraction during the simulation is very sensitive to the definition of the orientation classes.

Key words
texture components, abnormal grain growth, Monte Carlo simulation, grain boundary energy

Full text (268 KB)