A continuum-based multiphase DNS method for studying the Brownian dynamics of soot particles in a rarefied gas

In the mitigation of particulate matter (PM) using filters, the interplay between particle motion, geometry and flow conditions determines the overall performance. In such flows, molecular effects on the particle motion manifest as impeded momentum transfer and a meandering Brownian motion. Other particles and walls induce hydrodynamic interactions, which may further influence PM nucleation, growth and aggregation. Here, we formulate a multiphase direct numerical simulation framework to investigate these complex flows by including the molecular interactions in a consistent manner. The basis for this framework is a coupling between the Langevin description of particle motion with a mirroring immersed boundary method. We show that the method is able to capture the diffusion dynamics of a Brownian particle, including its transition from a particle-inertia dominated, correlated ballistic regime to a non-correlated diffusive one, and that the method also can be used to reproduce the meandering motion of realistic soot particles