A 3D cellular automaton with inhomogeneous nucleation for simulating dynamic recrystallization of low-alloy steel with mixed-grain microstructure

Abstract

Cellular automaton (CA) was widely applied in predicting the dynamic recrystallization (DRX) behavior of metals. However, the current CA models cannot accurately simulate the DRX evolution of mixed-grain microstructure. In this work, the limitations of traditional CA in simulating DRX of mixed-grain microstructure of low-alloy steel were analyzed, and the inhomogeneous nucleation based on the relationship between nucleation probability and dislocation amounts of grains was used for the first time to overcome this challenge. The initial microstructure was defined based on average grain sizes and grain size distributions of the sample. Isothermal compression experiments were conducted at different temperatures with different strain rates to validate the models’ reliability. The APRGE software was employed to investigate the most suitable orientation relationship (OR) for martensitic transformation and revealed that the Greninger-Troiano (GT) OR was the best one. The reconstructed maps from EBSD data showed that the simulation results based on inhomogeneous nucleation matched experiments better as the boundaries of coarse grains tended to act as preferential nucleation sites. What is more important, the difference in average grain sizes on different sections made it inappropriate for mixed-grain microstructure to choose deformation parameters based on 2D, which exactly reflected the advantages of 3D CA. According to experiments and 3D simulations, the initial mixed-grain microstructure became uniform during hot deformation when the natural logarithm of the Zener–Hollomon parameter was 34.59 or less, indicating that the models could help the selection of suitable deformation parameters for mixed-grain microstructure.