Prediction of microstructure by multi-scale modeling of recrystallization based on cellular automata and finite element method

Abstract

High-strength metals are gaining greater importance in the metal industry due to the recent requirements of the market. Several strengthening mechanisms such as substructural, solid solution, precipitation, grain boundary, and phase transformation strengthening have been used to develop high-strength materials. In general, most of these methods increase the strength but decrease the toughness of a material. However, grain refinement increases both the strength and toughness of a material. Thus, it is expected that techniques to control the microstructure in industrial manufacturing processes will be developed; this will require better prediction and modeling of the microstructure evolution. In this study, a multiscale model using cellular automata and the finite element method was developed to predict microstructural changes due to dynamic recrystallization during hot deformation. The modified cellular automata model was developed to simulate the dynamic recrystallization process. In this model, Moore’s neighboring rule was applied with partial fractions and time step control to represent the grain growth kinetics more accurately. The cellular automata model consists of work hardening, nucleation, and growth modules. To model the work hardening process, the dislocation model proposed by Kocks and Mecking, which contains kinetic and evolution equations, was used. The constant nucleation model was used for nucleation modeling, and Turnbull’s rate equation was used for growth modeling. For validation, the cellular automata model developed in this study was applied to a simulation of the dynamic recrystallization of pure copper during hot deformation. A set of isothermal hot compression tests were conducted using a Gleeble machine to acquire flow stresses. The compressed specimen was investigated by electron backscattered diffraction (EBSD) after mechanical and chemical polishing to obtain grain size distributions. The predicted results were in reasonably good...