Macroscopic and molecular fluid descriptions merge for better hydrodynamic modeling

Modeling complex interfaces and boundaries well requires molecular-scale accuracy in hydrodynamic simulations at multiple length scales, a feat achieved with computational efficiency by hybrid molecular-continuum simulation methods. Current hybrid methods can be difficult to implement, however, and have so far been largely limited to single-component fluid systems.

In an article published in The Journal of Chemical Physics, researchers uncovered a novel hybrid approach for multicomponent flow simulations of miscible binary mixtures. The results provide a means for linking one simulation’s spatial domain to another, resolved using molecular dynamics, described by coarse-grained continuum hydrodynamics.

A key aspect to the novelty of the method was in representing the continuum solution using fluctuating particle-based fluid models to easily interface with the inherently particle-based world of molecular modeling. In this approach, larger particles representing fluid volume elements, with continuous concentration profiles, split into smaller particles, with discrete solvent or solute identities. These then transform into the respective species when they cross into the molecularly resolved domain.

A similar process works in reverse, with a combination of finer particles creating larger ones. The splitting and combining events are designed such that the concentration fluctuations satisfy local equilibrium probabilities. Ultimately, the method was demonstrated in non-equilibrium transport simulations whereby concentration profiles over an entire multiscale simulation domain evolve collectively with time. This basic strategy is also readily extensible to complicated, 3-D-varying concentration profiles.

This innovative methodology bridges the gap, fairly literally, between macroscopic and molecular fluid descriptions, allowing easier simulations of solute transport phenomena across or near complex interfaces. These kinds of simulations have a variety of applications across scientific fields, such as, for example, in modeling intravascular delivery of porous drug particles to dynamically capture time-dependent distributions of active therapeutics in body fluids.

Source: “Coupling discrete and continuum concentration particle models for multiscale and hybrid molecular-continuum simulations,” by Nikolai D. Petsev, L. Gary Leal, and M. Scott Shell, Journal of Chemical Physics (2017). The article can be accessed at https://doi.org/10.1063/1.5001703 .