Abstract
In our earlier work, a stochastic model of multi-stage deformation at elevated temperatures was developed. The model was applied to calculate histograms of dislocation density and grain size at the onset of phase transformation. The histograms were used as input data for the simulation of phase transitions using the traditional deterministic model. Following this approach, microstructural inhomogeneity was predicted for different cooling conditions.The results obtained, showing the effect of dislocation density and inhomogeneity of austenite grain size on the microstructural inhomogeneity of the final product, can be considered reliable as they are based on material models determined in previous publications and validated experimentally. The aim of the present work was to extend the model by taking into account the stochastic nature of nucleation during phase transitions. The analysis of existing stochastic models of nucleation was performed, and a model for ferritic transformation in steels was proposed. Simulations for constant cooling rates as well as for industrial cooling processes of steel rods were performed. In the latter case, uncertainties in defining the boundary conditions and segregation of elements were also considered. The reduction of the computing costs is an important advantage of the model, which is much faster when compared to full field models with explicit microstructure representation.