A Lagrangian Stochastic Model for Droplet Deposition Simulations in Connection With Wall Function Approaches

Abstract

The deposition rate of droplets is strongly linked to their interaction with the boundary layer turbulence. In “industrial simulations”, droplets dispersion is usually modeled using Lagrangian stochastic simulations based on Reynold Average Navier Stokes (RANS) fluid calculations. Wall functions are also used to bound the number of mesh cells in the near wall region. But they also reduce the description of the boundary layer and lead to bad predictions of the droplets deposition rate. This study presents channel flow simulations using wall functions and run with the CFD code Fluent. In such configurations, the stochastic model of Fluent failed to represent the so-called “diffusion-impaction” regime of deposition. The “Concentration Wall Boundary Layer” model presented in this paper has been developed to predict deposition in simulations using industrial meshes with refinement such as y* > 20. This model calculates the deposition rate using only the intrinsic properties of the particles and the turbulent kinetic energy of the fluid expressed at the top of the boundary layer. The data provided by wall functions are then sufficient to calculate the deposition rate. This model is turned into a “Lagrangian stochastic wall boundary condition model” for the commercial CFD code Fluent. Various simulations have shown that this model improves remarkably the deposition predictions in channel flow. The dependence on the boundary cell size and the channel flow mean velocity has been tested. This model draws interesting perspectives to model deposition in complex configurations without requiring prohibiting mesh sizes.