Pressure drop in fibrous filters

Abstract

Coalescing filtration is a mechanical process, which is employed to remove dispersed aerosol particles from a gas stream. This kind of filtration is a depth filtration process and it is widely used in process industries to remove particulate matter from exhaust gases, or in compressed air applications to filter oil particles introduced during compression. Fibrous filters are often used due to low cost, high capture efficiency and low pressure-drop, where droplets are first captured on fibres, then coalesce, and eventually drain out. The performance of a filter medium is judged based on its capture efficiency and its pressure drop characteristics. Estimating these parameters without setting up experimental investigations of each filter medium would be beneficial for choosing and developing optimal filters. In the present study, numerical simulations using ANSYS CFX© are used to predict the pressure drop caused due to the air flow through the randomly oriented fibrous filter medium. A fictitious domain approach is used to simulate solid fibres without the need to create a case-specific mesh for each different fibre alignment. The method is compared to a benchmark simulation and a mesh analysis is carried out to find a balance between mesh refinement and computational effort required. This method is extended to model 3D fibres with random orientations. Multiple simulations, each with a different randomized fibre alignment, is carried out, varying both Reynolds number and solidity, and the results are compared with theoretical and 2D simulation results. It is seen that both available theoretical models and 2D simulation results overestimate the pressure drop caused by a real fibrous filter which is attributed to the inherent random orientation of the fibres.