Discrete Particle Tracking in Fluid Flows for Particulate Filter Simulations

Abstract

Filtration of particulate-laden airstreams is an important industrial process, undertaken for a combination of operational, environmental, safety and health reasons. The fundamentals of filtration are well-understood and these can serve as a basis for design where the filter is assumed clean, and where it has a regular geometry. However, after a filter is has undergone a period of service, or for filters with complex fibre geometries, the application of these basic principles is often unsuccessful. For this reason it is desirable to fully simulate a filter design prior to manufacture in order to gauge both the initial and long term filter performance.This paper presents a coupled fluid and particle solver that can accurately model particulate filters, and allows their performance to be predicted. The solver is built on the open-source OpenFOAM Computational Fluid Dynamics (CFD) software libraries, which are extended to include a physically accurate model for particle motion and capture. Results demonstrate that the particle motion is accurately modelled for a range of particle sizes, and allows the three major particle capturing mechanisms to be simulated - namely Brownian motion, interception, and inertial impact. A comparison between the performance calculated from Single Fibre Efficiency (SFE) theory and that predicted by the numerical model is presented and shows good agreement.To demonstrate the feasibility of conducting a realistic filter simulation using the developed code, the flow and particle capture behaviour of a representative filter geometry is presented. Future developments to the code are planned to combine discrete particle tracking with a multiphase volume of fluid model to predict filter behaviour of oil-mist filters, including coalescence and break-up.