Prediction of the Permeability of Packed Beds of Non-Spherical Particles

This paper details the ongoing development of a virtual permeameter which will have value in the design and performance assessment of filters used in a variety of chemical and process engineering applications. Having previously established the basic simulation requirements for such a permeameter with spherical glass beads, further experimental investigation and associated simulations are reported for non-spherical particles, namely mono-disperse sand and arbitrary poly-disperse, polymorphous minerals. To test the validity and applicability of the previously established guidelines with regard to sample size, resolution and accuracy, the micro-structural details of representative porous media samples are acquired using x-ray micro-tomography. Small sample arrays of such media are then used as input into lattice Boltzmann method (LBM) simulations for predicting their bulk permeability and related properties under laminar flow conditions. It is established that LBM is able to predict the flow rates through the beds at varying fluid pressures, with average error margins between experimental data and simulation predictions of 28% for glass beads, 27% for silica sand and 31% for polymorphous particles.