Comparison of Grain-Growth Mean-Field Models Regarding Predicted Grain Size Distributions.

Area of Science:

• Materials Science
• Metallurgy
• Computational Modeling

Background:

• Mean-field models are crucial for predicting material microstructure evolution under thermomechanical conditions.
• Understanding grain size distribution is key to material property prediction.
• Existing models vary in their representation of microstructure and neighborhood topology.

Purpose of the Study:

• To compare different mean-field models for predicting grain size distribution during grain growth.
• To evaluate the impact of microstructure representation, particularly neighborhood topology, on model predictions.
• To validate models using experimental data from 316L austenitic stainless steel.

Main Methods:

• Comparison of various mean-field models under grain-growth conditions.
• Investigation of different microstructure representations, including neighborhood topology.
• Parameter identification using experimental heat treatment data on 316L steel.
• Application of models to both monomodal and bimodal initial grain size distributions.

Main Results:

• Topological mean-field models demonstrated improved prediction accuracy for monomodal grain size distributions.
• In bimodal cases, relative model comparisons showed minor differences in predictions.
• The inclusion of neighborhood topology generally yielded only marginal improvements over classical models.

Conclusions:

• Topological considerations in mean-field models offer limited advantages for grain growth prediction.
• Classical mean-field models remain effective, especially when implementation complexity is a factor.
• Further research may be needed to fully leverage topological information in microstructure modeling.