Third-Order Effective Properties for Random-Packing Systems Using Statistical Micromechanics Based on a GPU Parallel...

Area of Science:

• Materials Science
• Computational Physics
• Statistical Mechanics

Background:

• Effective properties of particulate systems are crucial for understanding macroscopic performance.
• Accurate estimation requires considering microstructural details beyond simple averages.
• Third-order approximations offer a more refined approach to these estimations.

Purpose of the Study:

• To accurately evaluate third-order approximations for particulate systems.
• To develop and utilize a GPU-based parallel algorithm for computing n-point correlation functions.
• To investigate the influence of particle shape on effective thermal conductivity and diffusion coefficients.

Main Methods:

• Developed a GPU-based parallel algorithm for computing n-point correlation functions.
• Calculated third-order approximations using a triple integral involving 1-, 2-, and 3-point correlation functions.
• Modeled random-packing systems of super-ellipsoids to analyze particle-shape effects.

Main Results:

• The GPU algorithm demonstrated high accuracy, aligning well with analytical solutions.
• Third-order approximations were successfully applied to predict effective thermal conductivity and diffusion coefficients.
• Super-ellipsoid parameters were varied to show their impact on predicted material properties.

Conclusions:

• Third-order approximations, combined with efficient computational methods, provide accurate estimations of effective properties.
• Particle shape is a significant factor influencing thermal conductivity and diffusion in particulate systems.
• The developed methodology enables prediction of shape-dependent properties for various particulate materials.