A High-Fidelity Texture Discretization Method for Polycrystalline Aggregates Considering Grain Size Distributions

Area of Science:

• Materials Science
• Crystallography
• Computational Modeling

Background:

• Accurate orientation distribution function (ODF) discretization is crucial for microstructural modeling of polycrystalline aggregates.
• Conventional methods face challenges with discretization errors and adaptability to non-uniform grain size distributions (GSDs).

Purpose of the Study:

• To propose a novel texture discretization method for high-fidelity ODF approximation.
• To improve adaptability to non-uniform GSDs and suppress sampling-induced errors.
• To provide a robust framework for texture modeling considering GSDs.

Main Methods:

• Extended inverse transform sampling by reconstructing source samples before inversion.
• Introduced spatial shuffling and unscrambling of grain positions to preserve texture diversity.
• Utilized total variation distance (TVD) as a global metric for discretization error quantification.

Main Results:

• Achieved one order of magnitude lower TVD compared to conventional methods for typical grain numbers (10^2-10^3).
• Demonstrated a two orders of magnitude smaller standard deviation in TVD.
• Showcased improved adaptability to non-uniform grain size distributions.

Conclusions:

• The proposed method offers a significant reduction in discretization errors for ODF approximation.
• It provides a robust and efficient framework for texture modeling in materials science.
• The method enhances microstructural modeling reliability by accurately representing texture with GSD considerations.