Coarse brownian dynamics for nematic liquid crystals: Bifurcation, projective integration, and control via stochastic simulation

A computer-assisted approach that aspires to bridge macrocopic numerical analysis techniques for the cloed equations directly with microscopic/stochastic simulation codes was demonstrated in the Brownian dynamics context. This system-identification-based, computational "closure-on-demand" approach sidesteps the necessity of deriving good explicit closures. It enables state-of-the-art microscopic codes, by wrapping a computational superstructure around them, to perform tasks they were not in principle designed for.