CFD study of filtration process in moulded filters within a vacuum pump

Nausheen Sultana Mehboob Basha, Lloyd Cochrane, Stacey Culham, Faik Hamad, Zulfiqur Ali

Research output: Contribution to conferencePaperpeer-review

1793 Downloads (Pure)


Air/Oil filtration through filters is commonly utilised in the vacuum industry where oil lubricated pumps are used across a number of different applications such as food and packaging, industrial, pharmaceutical, R&D, forming and drying. The air/oil filters are crucial in the reduction of exhaust emissions, which, when suspended as fine particulate matter can cause great harm to the environment, climate, equipment life and public health. However, the behaviour of flow through the filters is not fully understood and much of the design and development work is based on historical know-how and experimental studies. Computational Fluid Dynamics (CFD) is a powerful tool to understand the flow characteristics and droplet trajectory through the filters which is challenging through experimental techniques. In this study, a CFD model is developed by using the commercial ANSYS FLUENT code. Oil droplets from the pump entering the filter are treated as a discrete phase. Euler-Lagrangian frame is used to characterise the multiphase flow, K-Ɛ as a turbulence model, Rosin-Rammler distribution of oil droplets, User Defined Functions (UDF) are written for droplet injection, distribution and deposition. Various methodologies and tests were developed to obtain the required data to feed into the model and validate the data predicted by the computational model. The obtained computational data agrees well with the experimental data.
Original languageEnglish
Publication statusPublished - 13 Oct 2016
EventFiltration and Separation technology, FILTECH2016 - , Germany
Duration: 11 Oct 201613 Oct 2016


ConferenceFiltration and Separation technology, FILTECH2016


Dive into the research topics of 'CFD study of filtration process in moulded filters within a vacuum pump'. Together they form a unique fingerprint.

Cite this