Comparison between experimental data and a cellular automata simulation of martensite formation during cooling

Computer simulations of steel microstructural development provide a powerful tool, which can form the basis of mechanical property predictions. However, in order to create detailed understanding of the factors affecting the properties, the model should predict microstructural evolution during cooling. The present study compares the results of cellular automata simulations with experimental data for two distinct austenite conditions, recrystallized and deformed. Detailed microstructural features were studied using a laser scanning confocal microscope, FESEM and FESEM-EBSD. The two-dimensional cellular automata (CA) model for simulating the formation of lath martensite was parameterized using fitted Johnson-Mehl-Avrami-Kolmogorov and Koistinen-Marburger equations. The parent austenite microstructure for the CA model was determined from the final martensitic microstructure using austenite grain reconstructions based on the use of MATLAB software and the MTEX toolbox. The results of this cellular automata simulation can be used to estimate the shapes and sizes of martensite blocks, which offers new possibilities for the qualitative estimation of the mechanical properties of high-strength steels formed from recrystallized or deformed austenite. (paper)