Computational Fluid Dynamics Model of a Swirler Separator for Gas Cleaning

Abstract

This work is concerned with the development of a computational fluid dynamics model for a two-phase, turbulent, swirling flow produced by stationary guide vanes. The swirling flow causes separation of particles in the air stream and hence the device is called swirler separator. The Reynolds-averaged continuity and Navier-Stokes equations are solved along with the Boussinesq hypothesis to describe the stress distribution throughout the flow field in a body-fitted coordinate system. The κ-ε model is used to determine turbulent viscosity. Finite volume methodology is adopted to discretize the system of governing partial differential equations and the semi-implicit method for pressure linked equations consistent to deal with the pressure-velocity coupling. The dilute phase is accounted for by following a Lagrangian methodology in which a Newtonian force balance tracks the particles throughout the flow field. A stochastic method is employed to model the dispersion of particles due to turbulence of the fluid-phase. The phenomenological model is then successfully used to predict velocity and pressure fields created by the guide vanes as well as particle classification curves brought about by the swirler separator.