A Novel Method for Permeability Estimation from Micro-tomographic Images

Abstract

A methodology has been developed to create a pore network model (PNM) from the geometrical/topological information extracted from the micro-tomographic images of a polyurethane (PU) foam sample. By solving fluid-flow equations on this model, we could estimate the permeability of PU foams and validate with experimental findings. Previous literature suggested that the permeability of open-cell PU foams was overestimated using the traditional Carman–Kozeny and Duplessis–Masliyah models. Inter-cell connectivity was deemed as the potential cause of this difference. Thus, taking into consideration the effects of spatial arrangement of pores of different sizes, the throat constriction between pores, and the percentage open-cell content, a PNM was employed where the entire void space is treated as a network of pores and throats of varying dimensions. To simulate flow through the network, the mass conservation equations were solved at each pore, the Darcy’s equation was applied globally, and thus, the overall permeability of the foam was estimated. Using our PNM, we were able to estimate the permeability with very good agreement with experimentally observed results. This indicates that pore connectivity has a significant effect on permeability. Parametric studies with PNM also revealed that the permeability depends on the square of pore size while is quite sensitive to throat constrictions with cubic dependence.