Microfluidic Modelling of Clogging in Filtration Membranes using 2-D coupled CFD-DEM

Abstract

In this dissertation, clogging in microfiltration membranes has been investigated using computer models. The main objective of this work was to identify clogging's key factors then derive a relationship to link them. Clogging in microfiltration membranes was modelled using 2-d symmetrical DEM-CFD coupled models in which Hertz-Mindlin and particle-wall links were utilized to describe particles’ interactions laminar incompressible models for fluid flow. Validity of the chosen models was tested by adopting and numerically modeling an experimental study then compare numerical results to experimental ones. When compared to the experiential data, the results from the chosen models have an error of roughly 25%. Then a parametric study was carried out to examine the effect of trans-membrane flux, flow velocity, confinement ratio (a ratio of pore size to particle size), and pore’s entrance angle (with reference to the vertical axes). Clogging in microfiltration membrane is believed to be a multi-dimension phenomenon with great contribution from these parameters. Clogging is found to be greater with high flow velocity and higher trans-membrane flux and highly proportional to geometric parameters since it is inversely proportional to both the confinement ratio and the cosine of pore’s entrance angle. Also, clogging in microfiltration membranes is categorized in two types, internal and external clogging. Internal clogs are more likely to happen to membranes with high confinement ratios while external clogs happen to membrane with low confinement ratios. Based on the nature of the relationships between each parameter and clogging chances, a dimensionless number (α) is proposed to describe clogging on microfiltration membrane quantitatively in accordance with the input of these parameters. This number is derived from a hypothesis that is based on simple probability analysis. For α values greater than unity clogging is significa