Abstract:
Flow phenomena in porous media are relevant in many technological applications including wastewater filtration, chemical separations, and oil recovery. Nonwoven fibrous membranes are widely used as filtration media, and tailoring the microstructure of the pore space is a key to improving filtration efficiency and preventing membrane fouling. However, predicting effective properties such as permeability and tortuosity for complex porous microstructures remains a challenging task as most available models do not incorporate the influence of random fiber arrangement and orientation on the structure-property relations. This study presents a computational analysis of the permeability and tortuosity of realistic nonwoven fibrous media with regular and random fiber arrangements over a wide range of porosities. We conduct pore-scale lattice Boltzmann simulations to predict the fluid flow through nonwoven fibrous media and use the results to test the accuracy of semiempirical scaling relations. Combining morphological properties of the pore space with pore-scale flow simulations enables us to determine the influence of the statistical pore size distribution on the effective fluid transport properties over a wide range of macroscopic porosities. We find that semiempirical coefficients such as the Kozeny-Carman coefficient depend markedly on the statistical distribution of pores and throats. This dependence can be quantified as a constriction factor that incorporates the pore size distribution to yield accurate predictions of permeability of random fibrous media. In combination with experimental imaging techniques, the ability to quantify the connection between microstructure and fluid transport paves the way to computational characterization and design of porous media with tailored properties for filtration and separation applications.