Effect of system size on particle-phase stress and microstructure formation

Abstract

In this paper, we investigate the effect of particle number, or system size, on three-dimensional (3D) particle dynamic simulation results. Specifically, we simulate conditions with varying e (coefficient of restitution) and φ (solids volume fraction), containing particle numbers ranging from 250 to 300 000. Various algorithmic improvements are implemented in the simulation to efficiently handle these large numbers of particles. We observe, for the first time for 3D simulations, particle-phase microstructure formation at high coefficients of restitution. Furthermore, we show the onset of the particle-phase microstructure formation at various threshold system sizes, depending on the value of e and φ, and relate it to an increase in particle-phase stress. Visual observations are used to conduct a preliminary investigation of the nature of the microstructure formation. Multiple well-defined bands are observed for simulations of at least 100 000 particles.

Simulation of cohesive fine powders under a plane shear

“…Such clustering processes may be understood by a set of hydrodynamic equations of granular gases [56,57]. When we apply a shear to the granular gas, there exist various types of clusters such as 2D plug, 2D wave, or 3D wave for three dimensional systems [58][59][60][61][62]. In this paper, we try to characterize nonequilibrium pattern formation of cohesive fine powders under the plane shear by the three dimensional molecular dynamics (MD) simulations of the dissipative LJ molecules under the Lees-Edwards boundary condition [64].…”

“…Let us consider cohesive powders under a plane shear. So far there exist many studies for one or two effects of the shear, an attractive force, and an inelastic collision [44][45][46][47][48][49][50][51][52][53][54][55][56][57][58][59][60][61][62], but we only know one example for the study of the jamming transition to include all three effects [63]. On the other hand, when the Lennard-Jones (LJ) molecules are quenched below the coexistence curve of gas-liquid phases [44][45][46][47][48][49], a phase ordering process proceeds after the nucleation takes place [50][51][52].…”

Modeling particle‐laden flows: A research outlook

“…This point is illustrated in Figure 2, which shows a plot of the dimensionless normal stress (stress component perpendicular to the shearing plane) as a function of both the number of simulated particles and the relative system size, L/d p , with L the characteristic length of the system and d p the particle diameter (Lasinski et al, 2004). Increases in the particle-phase stress are evident, provided a sufficiently large number of particles are simulated.…”

“…Simulations also show that the particle-phase stress increases less rapidly with system size as the particle-size ratio is increased. Lasinski et al (2004) have conducted DEM simulations involving up to 300,000 particles. Both the nature of clustering and the particle-phase stress continue to change with system size.…”

Process Modeling in the Pharmaceutical Industry using the Discrete Element Method

“…The hard‐particle approach16–20 assumes that particles are rigid so that collisions are instantaneous and binary. As a result, hard‐particle models are generally best suited for dilute, collisional flows where these assumptions are good approximations.…”