Filtration and clogging behaviour: A systematic definition in the pore network

This study utilises graph concepts to provide a novel framework that characterises a key element of the filtration process: the clogging of finer base particles within the pore space of coarser filter particles. By utilising particle-scale data, a hyper-graph is generated that combines the filter pore network and the base particle contact network, through the use of graph components. This hyper-graph representation enables the identification of individual constriction status as empty, clogged, or percolating depending on the component membership. Using the constriction status and the connectivity of the pore network, the status of individual pores was defined as empty, clogged or non-clogged, and at capacity or with remaining capacity. This framework was developed in a highly generalised way to enable application to any numerical or experimental investigation of filtration that could generate the required particle-scale data. The proposed framework was applied to an extensive set of gravity-driven filtration simulations using the Discrete Element Method. This demonstrated that it was essential to account for constriction clogging and percolation, as well as, separately accounting for pore clogging and pore capacity status. An improved clogging criterion that accounts for base particle group effects is presented, and the existence of meso-scale relationships was demonstrated. These findings not only provide an improved understanding of the filtration mechanisms, but will also inform better particulate transport modelling and improved filter criteria that capture underlying mechanisms of filtration.