To main content

Implementation and Demonstration of a Boundary Condition Wall Function for Industrial Scale Particulate Fouling CFD Modeling

Abstract

Modeling of mass transfer of particles from a gas/liquid bulk flow to solid surfaces is a considerable challenge in most industrial processes, including heat exchanger applications. Using computational fluid dynamics (CFD) directly on these applications is very challenging and costly due to the grid refinement required to capture the complex physical processes that dominate in the near-wall region. In two previous papers, Johnsen and Johansen (2009a) and (2009b), a mass transfer wall function, for coarse grid CFD models, was proposed. The model can be employed to calculate particle phase mass-transfer coefficients, for the turbulent boundary layer, by coupling the detailed physics of the near-wall region with the external flow. It can thus be employed as a boundary condition mass sink in industrial scale CFD models, at reasonable computational costs. An Eulerian-Eulerian two-fluid, particle-liquid, model is considered. The model equations are derived from volume and ensemble averaged Navier-Stokes equations coupled with heat transport equations, for an incompressible liquid and a mono-disperse incompressible inert particle phase. The boundary layer model (equations for axial velocity, temperature and volume fraction) is solved numerically in a one-dimensional grid that is capable of resolving the near-wall XDLVO force length scales. The proposed wall function incorporates gravity, turbulence and hydrodynamic lift and drag. Steady turbulent flow is assumed, and stochastic turbulent fluctuations are handled by another layer of ensemble averaging. In addition the effects of Brownian diffusion, thermophoresis, XDLVO forces and inter-particle collisions are included. Shear-induced re-entrainment is modeled by modifying the deposition flux by an effective adhesion probability, caused by the adhesion forces and the turbulence wall stress statistics. Thermodynamic and chemical effects, such as phase change or precipitation, are not included in the model. The model performs w

Category

Academic lecture

Language

English

Affiliation

  • SINTEF Industry / Process Technology

Presented at

7th International Conference on Multiphase Flow

Place

Tampa, Florida, USA

Date

30.05.2010 - 04.06.2010

Organizer

University of Florida

Year

2010

View this publication at Cristin