A new algorithm for extended nonequilibrium molecular dynamics simulations of mixed flow

Area of Science:

• * Computational physics
• * Fluid dynamics
• * Statistical mechanics

Background:

• * Simulating fluids under mixed flow (planar elongational and Couette flow) using nonequilibrium molecular dynamics (NEMD) has been limited by irreversible transformations.
• * Previous NEMD methods for mixed flow required repeating short simulations and increasing box dimensions due to lattice space violations.
• * These limitations hindered long-term simulations and accurate statistical analysis of fluid behavior under mixed flow conditions.

Purpose of the Study:

• * To develop a novel algorithm for simulating fluids under planar mixed flow using NEMD.
• * To enable continuous and extended simulations of fluid dynamics under mixed flow conditions.
• * To overcome the limitations of previous NEMD techniques that imposed irreversible transformations.

Main Methods:

• * Developed a new algorithm for NEMD of fluids under planar mixed flow.
• * Implemented a method where the simulation box deforms with the flow streamlines.
• * Incorporated a mapping procedure to reset the box to its initial shape without physical property discontinuity.

Main Results:

• * The new algorithm allows the simulation box to deform and recover its initial shape, enabling indefinite simulation times.
• * This method avoids the need for repeating short simulations or increasing box dimensions.
• * Demonstrated that general mixed flow can be simplified into a canonical form by adjusting field parameters.

Conclusions:

• * The developed algorithm significantly advances the simulation of fluids under mixed flow using NEMD.
• * Continuous and extended simulations are now feasible, improving statistical accuracy and enabling longer-term dynamic studies.
• * The ability to cast mixed flow into a canonical form simplifies analysis and broadens applicability.