Development and application of a fully coupled CFD-DEM framework for particle-laden flows

다국어 초록 (Multilingual Abstract)

Optimization of industrial processes involving particle-laden flows requires a fundamental understanding of fluid-particle interactions. This dissertation develops and applies a fully coupled computational fluid dynamics-discrete element method (CFD-DEM) framework for particle-laden flows, accounting for interactions among the carrier fluid, dispersed particles, and solid boundaries. The fluid solver employs a direct-forcing immersed boundary method (IBM) on a structured, staggered Cartesian mesh, which represents complex geometries without body-fitted grids. The discrete phase is modeled using a soft-sphere approach extended with a rolling resistance model. Additionally, to account for flexible non-spherical particles, the DEM solver incorporates a bonded particle model (BPM) in which inter-particle bonds are governed by the Timoshenko beam theory, allowing the representation of axial, shear, torsional, and bending deformations.
The accuracy of the coupled solver is verified against analytical solutions, cross-checked with commercial software, and validated against experimental data, showing good agreement. The validated CFD-DEM framework is then applied to two channel-cavity problems: particle trapping and particle removal.
First, regarding the particle trapping problem, a parametric study shows a non-monotonic relationship between the trap ratio and both the particle-to-fluid density ratio and the cavity aspect ratio, associated with the formation, merging, and escape of two dominant vortices inside the cavity. The trap ratio decreases with the fluid Reynolds number and increases with the volumetric particle loading, with four-way coupling effects becoming significant in dense suspensions due to flow modulation.
Second, for particle removal, results indicate that increasing Reynolds numbers provides only marginal improvements in cleaning efficiency, as particles in the deeper regions of the cavity remain less affected. However, the introduction of harmonic inflow excitation enhances particle removal because the oscillatory flow generates new vortices that penetrate deeper into the cavity and interact with existing flow structures, thereby achieving near-complete removal under optimal conditions.

목차 (Table of Contents)

• Chapter I Introduction
• 1.1 Motivation
• 1.2 Research objectives
• 1.3 Thesis structure
• Chapter II Numerical Methods
• 2.1 Coupled CFD-DEM algorithm
• 2.2 Mathematical descriptions for dispersed phase
• 2.2.1 Contact force model
• 2.2.2 Fluid-particle interaction forces
• 2.2.3 Bonded particle model
• 2.2.4 Time integration schemes
• 2.3 Mathematical descriptions for continuous phase
• 2.4 Verification and validation
• 2.4.1 Verification of the DEM solver
• 2.4.2 Verification of the CFD solver
• 2.4.3 Validation of the coupled CFD-DEM framework
• Chapter III Particle Trapping in Channel-Cavity Systems
• 3.1 Problem description
• 3.2 Baseline case
• 3.3 Effect of particle-to-fluid density ratio
• 3.4 Effect of fluid Reynolds number
• 3.5 Effect of volumetric particle loading ratio
• 3.6 Effect of cavity aspect ratio
• 3.7 Chapter summary
• Chapter IV Particle Removal in Channel-Cavity Systems
• 4.1 Problem description
• 4.2 Baseline case
• 4.3 Effect of fluid Reynolds number
• 4.4 Effect of inflow excitation frequency
• 4.5 Chapter summary
• Chapter V Conclusion