Robust event-driven particle tracking in complex geometries

For most particle simulations, a time-dependent mapping between the particles’ positions and an underlying grid is an important component and is used, for example, to increase the eciency during the collision detection step. In the case of unstructured grids, which are frequently employed to handle domains of complex shape, obtaining this mapping is computationally expensive. The process can be accelerated by performing particle tracking, that is, the repeated localization of particles within a grid by means of tracking the trajectories of the particles. In fact, particle tracking is an application of event-driven particle dynamics (EDPD), hence, in this work, recent advances in stable EDPD algorithms are applied to the problem of particle tracking to address inconsistencies which arise due to numerical errors or imperfect meshes. It is illustrated how interactions of the particles with the system boundaries can be integrated into the new algorithm consistently. Additionally, it is demonstrated that the modeling of solid objects via constructive solid geometry can be combined with event-driven particle tracking algorithms to provide a fully analytical description of complex objects defining or embedded into the simulation domain. A robust particle tracking algorithm is presented, along with several optimizations with respect to the computational eciency. The capabilities of the developed method are exemplified via the simulation of a gas flow through a highly porous medium.