Correlation as a Source of Stochasticity in Continuum Models: Part 1

Keith Promislow
Michigan State University

Molecular dynamics models resolve material properties in terms of pairwise interactions between N particles. These can be characterized via a partial differential equation for the N-particle probability distribution function. The BBGKY hierarchy represents the coupled interaction of the partially integrated marginals of the probability distribution. Under ideal circumstances, as N grows the effective interactions can be modeled by coupling a single particle to a mean-field generated by the other particles – in this situation the joint probability distributions reduce to tensor products of a single particle distribution. This represents a tremendous reduction in complexity. This mean-field limit is not universal, and it generally breaks down when small groups of particles have successive interactions with each other such as between the water molecules in a hydration shell of an ion. These repeated interactions generate correlations described by 2 and 3 particle joint distributions. We will review the BBGKY framework, and the impact of correlation on the hydrodynamic limits that provide a further reduction to continuum systems. Our focus will be on models of interactions of particles with dipole moments, such as water.

Presentation (PDF File)
View on Youtube

Back to Bridging the Gap: Transitioning from Deterministic to Stochastic Interaction Modeling in Electrochemistry Tutorials