Modelling ripple morphodynamics driven by colloidal deposition

TL;DR

This study uses numerical simulations to analyze how colloidal silica particles deposit and form ripple-like structures on heat exchanger surfaces, revealing effects of surface roughness on deposition patterns.

Contribution

It introduces a combined multiphase and dynamic mesh model to simulate colloidal deposition and surface morphodynamics in heat exchangers, validated against experimental data.

Findings

Silica deposition is faster at ripple peaks.

Surface roughness reduces silica accumulation by 20%.

Mesh quality is critical in deforming simulations.

Abstract

Fluid dynamics between a particle-laden flow and an evolving boundary are found in various contexts. We numerically simulated the morphodynamics of silica particle deposition from flowing water within geothermal heat exchangers using the arbitrary Lagrangian-Eulerian method. The silica particles were of colloidal size, with submicron diameters, which were primarily transported through the water via Brownian motion. First, we validated the Euler-Euler approach for modelling the transport and deposition of these colloidal particles within a fluid by comparing our simulation results with existing experiments of colloidal polystyrene deposition. Then we combined this multiphase model with a dynamic mesh model to track the gradually accumulated silica along the pipe walls of a heat exchanger. Surface roughness was modelled by prescribing sinusoidally-shaped protrusions on the wall boundary. The silica bed height grew quickest at the peaks of the ripples and the spacing between the protrusions remained relatively constant. The rough surface experienced a 20 % reduction in silica deposition when compared to a smooth surface. We also discuss the challenges of mesh deforming simulations with an emphasis on the mesh quality as the geometry changes over time.