[Paper Review] The transport and deposition of heavy particles in complex terrain: insights from an Eulerian model for large eddy simulation

Scott T. Salesky, Marco G. Giometto+3 more|arXiv (Cornell University)|Mar 8, 2019

Aeolian processes and effects96 references4 citations

TL;DR

This paper presents a novel Eulerian large eddy simulation (LES) framework coupled with an immersed boundary method and finite volume cut-cell discretization to model heavy particle transport and deposition over complex terrain. It demonstrates that particle inertia, driven by mean flow acceleration, enhances deposition on windward slopes and suppresses it on leeward sides, with mean inertial effects being six times larger than turbulent components.

ABSTRACT

The transport and deposition of heavy particles over complex surface topography by turbulent fluid flow is an important problem in a number of disciplines, including sediment and snow transport, ecology and plant pathology, aeolian processes, and geomorphology. This article presents a framework to simulate the transport and deposition of heavy particles over complex surfaces using the large eddy simulation (LES) technique. An immersed boundary LES code is coupled with an Eulerian particle code that solves the advection-diffusion equation for the resolved particle concentration field. The mass conservation equation for the particle phase is discretized in a finite volume framework using a Cartesian cut cell method that reshapes finite volume cells intersected by the immersed boundary surface and conserves mass accurately. The proposed numerical model is compared with data from wind tunnel experiments of heavy particle deposition over topography and is found to have good agreement with observed deposition patterns. An LES case study of snow deposition over idealized topography leads to several new insights. Particle inertia leads to relative velocities between the particles and fluid in regions of mean flow acceleration, thereby enhancing deposition on the windward side of obstacles and suppressing deposition on the leeward side. In addition, it is found that the mean components of particle inertia are a factor of 6 or more larger than the turbulent components, indicating that the enhancement/suppression of deposition by topography can be modeled in terms of mean flow quantities.

Motivation & Objective

• To develop a high-fidelity numerical model for simulating turbulent transport and deposition of heavy particles over complex topography.
• To address the challenge of accurately conserving mass in complex geometries with sharp particle concentration gradients.
• To investigate how topographic features influence particle deposition patterns through turbulent flow interactions.
• To quantify the relative contributions of mean flow and turbulent components to particle inertia in deposition processes.
• To validate the model against wind tunnel experiments and apply it to idealized snow deposition cases.

Proposed method

• An immersed boundary method is used to represent complex terrain within a Cartesian structured grid, enabling accurate boundary resolution.
• A finite volume cut-cell method reshapes cells intersected by solid boundaries and ensures strict mass conservation of the particle phase.
• The advection-diffusion equation for particle concentration is solved in an Eulerian framework using a second-order accurate time integration scheme.
• Geometric quantities such as face areas and normal vectors on cut cells are computed using de Gua’s theorem and the divergence theorem.
• A conservative mixing model is applied to stabilize small cut cells and maintain numerical stability without reducing the time step.
• The mixing fractions are based on normal vector components and target cell volume fractions, with a power-law weighting for numerical robustness.

Experimental results

Research questions

• RQ1 How does topography influence the spatial distribution of heavy particle deposition under turbulent flow?
• RQ2 To what extent do mean flow accelerations and particle inertia govern deposition patterns on windward versus leeward slopes?
• RQ3 What is the relative contribution of mean versus turbulent components of particle inertia to deposition enhancement or suppression?
• RQ4 Can the deposition patterns induced by topography be accurately captured using only mean flow quantities?
• RQ5 How well does the proposed LES model reproduce experimental deposition patterns observed in wind tunnel studies?

Key findings

• The model shows good agreement with wind tunnel experiments, accurately reproducing observed deposition patterns over complex topography.
• Particle inertia leads to relative velocities between particles and fluid in regions of mean flow acceleration, enhancing deposition on windward slopes.
• Deposition is suppressed on leeward slopes due to reduced relative velocity and flow deceleration.
• The mean component of particle inertia is found to be at least six times larger than the turbulent component, indicating that deposition patterns can be modeled using mean flow quantities alone.
• The conservative mixing model successfully maintains numerical stability and mass conservation in cut cells with small volume fractions.
• The framework enables high-resolution simulation of snow deposition over idealized topography, revealing the dominant role of mean flow in shaping deposition patterns.