Mesoscopic coarse-grained representations of fluids rigorously derived from atomistic models

Area of Science:

• Computational fluid dynamics
• Mesoscopic modeling
• Statistical mechanics

Background:

• Mesoscopic models are crucial for simulating complex phenomena in chemical and biological systems.
• Establishing a rigorous link between fine-grained (FG) atomistic models and coarse-grained (CG) mesoscopic models for fluids remains a challenge.
• Existing CG models often lack clear connections to their underlying FG descriptions.

Purpose of the Study:

• To introduce a novel CG mapping scheme for constructing mesoscopic fluid models from FG data.
• To address the challenge of creating dynamically self-consistent CG representations.
• To develop a parameterization protocol for these CG models.

Main Methods:

• Dynamically self-consistent smooth centroidal Voronoi tessellation for CG mapping.
• Application to supercritical Lennard-Jones fluids under quiescent and shear flow conditions.
• Parameterization using the Mori-Zwanzig formalism and the Generalized Langevin Equation.

Main Results:

• The new method generates continuous, stable, and ergodic CG trajectories.
• It quantitatively captures slow collective motions of FG fluids.
• The parameterized CG models accurately reproduce structural and dynamical correlations.

Conclusions:

• The developed dynamical mapping scheme provides a direct bottom-up approach for constructing mesoscopic fluid models.
• This method facilitates the creation of Lagrangian-based mesoscopic simulations.
• It offers a robust framework for bridging atomistic and mesoscopic scales in fluid dynamics.