Particle Transport and Mixing: Effects of Shape, Fluid, and Cohesion

Abstract:
Although granular media are ubiquitous, their transport and mixing involve complex interactions between particles and with fluids and, consequently, remain poorly understood. My talk examines two of these complexities: one considers more realistic representations of particle shape and fluid interactions while the other examines the influence of interparticle cohesion.

The first study aims to develop a numerical tool to simulate fluid-driven transport of sediment particles with a realistic shape distribution, rather than using a single representative shape as in previous numerical studies. Hundreds of distinct natural gravel particles are scanned, reconstructed, and implemented in a fluid-coupled flow solver. The particle-resolved simulation results of natural gravel and spheres are benchmarked quantitatively with flume experiments. Furthermore, the numerical tool allows us to test a recently proposed sediment transport relation (Deal et al., 2023) by varying the material parameters and simulation geometry to extend its applicability from flumes to wide rivers. Additionally, particle-level details are upscaled to a two-phase continuum, enabling a workflow with different resolutions.

The second study considers the influence of cohesion on spontaneous size segregation in gravity-driven bounded heap flow using discrete element method simulations. Size segregation is influenced by enduring or transient agglomerates for strong or weak cohesion, respectively, resulting from the interplay between kinematic formation and shear-induced breakup of agglomerates. Breakup simulations of cohesive particle pairs in dense noncohesive-particle shear flows provide a minimal model for understanding the physics. The breakup time, nondimensionalized by the shear rate, depends only on a modified Bond number (Bo) comparing cohesion and the pressure-dependent contact force. Cohesive breakup time is constant and quite short for Bo<1, but increases rapidly for Bo>1. The Bo dependence is captured by a probabilistic model considering bed driven force fluctuations, which promises to be widely applicable as it accurately predicts erosion in previous studies of large cohesive agglomerates.

Bio:
Qiong Zhang is a postdoc of Mechanical Engineering at Northwestern University. He obtained his B.Eng. from Tsinghua University, and his M.Sc. and Ph.D. in Mechanical Engineering from MIT. His research focus falls primarily within the broad field of granular and fluid flows, with specific interests covering sediment transport, nonlocal granular rheology, scaling relations for locomotors on granular media, and segregation of cohesive particles.