Umberto D'Ortona

An immersed boundary-lattice Boltzmann method for single- and multi-component fluid flows

The paper presents a numerical method to simulate single- and multi-component fluid flows around moving/deformable solid boundaries, based on the coupling of Immersed Boundary (IB) and Lattice Boltzmann (LB) methods. The fluid domain is simulated with LB method using the single relaxation time BGK model, in which an interparticle potential model is applied for multi-component fluid flows. The IB-related force is directly calculated with the interpolated definition of the fluid macroscopic velocity on the Lagrangian points that define the immersed solid boundary. The present IB-LB method can better ensure the no-slip solid boundary condition, thanks to an improved spreading operator. The proposed method is validated through several 2D/3D single- and multi-component fluid test cases with a particular emphasis on wetting conditions on solid wall. Finally, a 3D two-fluid application case is given to show the feasibility of modeling the fluid transport via a cluster of beating cilia.

Influence of rough and smooth walls on macroscale granular segregation patterns

Size bidisperse granular materials in a spherical tumbler segregate into two different patterns of three bands with either small particles at the equator and large particles at the poles or vice versa, depending upon the fill level in the tumbler. Here we use discrete element method simulations with supporting qualitative experiments to explore the effect of the tumbler wall roughness on the segregation pattern, modeling the tumbler walls as either a closely packed monolayer of fixed particles resulting in a rough wall or a frictional geometrically smooth wall. Even though the tumbler wall is in contact with the flowing layer only at its periphery, the impact of wall roughness is profound. Smooth walls tend toward a small-large-small (SLS) band pattern at the pole-equator-pole at all but the highest fill fractions; rough walls tend toward a large-small-large (LSL) band pattern at all but the lowest fill fractions. This comes about because smooth walls induce poleward axial drift of small particles and an equator-directed drift for large particles, resulting in an SLS band pattern. On the other hand, rough walls result in both sizes of particles moving poleward at the surface of the flow. Due to radial segregation, small particles percolate lower in the flowing layer and when arriving near the pole are caught in the return current drift that carries them back toward the equator incrementally with each passage through the flowing layer, while large particles remain at the surface near the pole, resulting in an LSL band pattern. The tendency toward either of the two segregation patterns depends on the fill level in the tumbler and the roughness of the tumblers bounding wall.

Influence of rough and smooth walls on macroscale flows in tumblers

Walls in discrete element method simulations of granular flows are sometimes modeled as a closely packed monolayer of fixed particles, resulting in a rough wall rather than a geometrically smooth wall. An implicit assumption is that the resulting rough wall differs from a smooth wall only locally at the particle scale. Here we test this assumption by considering the impact of the wall roughness at the periphery of the flowing layer on the flow of monodisperse particles in a rotating spherical tumbler. We find that varying the wall roughness significantly alters average particle trajectories even far from the wall. Rough walls induce greater poleward axial drift of particles near the flowing layer surface but decrease the curvature of the trajectories. Increasing the volume fill level in the tumbler has little effect on the axial drift for rough walls but increases the drift while reducing curvature of the particle trajectories for smooth walls. The mechanism for these effects is related to the degree of local slip at the bounding wall, which alters the flowing layer thickness near the walls, affecting the particle trajectories even far from the walls near the equator of the tumbler. Thus, the proper choice of wall conditions is important in the accurate simulation of granular flows, even far from the bounding wall.