Improvements to Stochastic Approaches for Simulating Agglomeration in Particle‐Laden Flows

Abstract

AbstractDirect numerical simulation, facilitated by the spectral element method, has been used to study collisions and agglomeration in particle‐laden fluid flows through a channel at shear Reynolds number. The particulate phase is simulated by deterministic Lagrangian particle tracking alongside a stochastic technique to resolve collisions. Interestingly, the implementation of agglomeration in the deterministic approach caused a major reduction in the particle collision rate. To determine the cause of this effect, analysis of the collision rate and location across the channel for different particle properties was performed. For systems without agglomeration, collisions between particles tend to be located on consistent streamlines, occurring between single pairs of particles. A numerical analysis of this effect confirmed the result, where inter‐particle collisions across the channel mostly took place between consistent particle pairs in close proximity. Repeat collisions are shown to be almost eliminated with the addition of the agglomeration mechanism. The impact and implications of this effect on the accuracy of the stochastic technique is discussed, and modifications are suggested to offer improvements while accounting for the repeat collisions.