Prediction of initial filtration performance in porous filters using granular bed theory

Baghouses are widely used to remove particulate matter (PM). However, predictive equations for porous filters with irregular pore networks still remain limited. This study proposes a morphology-aware framework that maps the polymer network to an analogous granular bed using a volume-equivalent sphere diameter and a shape factor. Under clean filter, low- Re conditions, the Ergun correlation is employed to predict pressure drop, and a modified log-penetration model—based on single-sphere theory and incorporating shape factor, an inhomogeneity factor, and the pore size distribution—is used to predict collection efficiency. Experiments with two porous filters show that Ergun equation provides reliable pressure drop predictions with a small, consistent overprediction attributable to pore size dispersion and local packing nonuniformity. For collection efficiency, using only the mean volume-equivalent diameter overestimates measurements, whereas including the full pore size distribution substantially improves agreement across particle sizes. These results provide valuable insight for predicting initial filtration performance of irregular porous filters, linking measurable geometric characteristics to both pressure drop and size-resolved efficiency—thereby supporting design and evaluation of baghouse media. • A predictive method for the filtration performance of porous filters is proposed. • Porous filters are modeled as granular beds with simplified hexagonal cell structures. • The Ergun equation and log-penetration equation were applied to predict filtration performance, and good agreement was observed.