Estimation of large domain Al foam permeability by Finite Difference methods. Issue 1 (December 2013)

• Record Type: Journal Article
• Authors: Osorno, Maria; Steeb, Holger; Uribe, David; Ruiz, Oscar
• Is Part Of: Proceedings in applied mathematics and mechanics. Volume 13:Issue 1(2013)
• Page Start: 247
• Page End: 248
• Publication Date: 2013-12
• Subjects: Applied mathematics -- Periodicals; Engineering mathematics -- Periodicals; Mathematical physics -- Periodicals; 519
• DOI: 10.1002/pamm.201310119
• Languages: English
• ISSNs: 1617-7061

Abstract

<abstract abstract-type="main" xml:lang="en"> <title>Abstract</title> <p>Classical methods to calculate permeability of porous media have been proposed mainly for high density (e.g. granular) materials. These methods present shortcomings in high porosity, i.e. high permeability media (e.g. metallic foams). While for dense materials permeability seems to be a function of bulk properties and occupancy averaged over the volume, for highly porous materials these parameters fail to predict it. Several authors have attacked the problem by solving the Navier‐Stokes equations for the pressure and velocity of a liquid flowing through a small domain (Ω<sub><italic>s</italic></sub>) of aluminium foam and by comparing the numerical results with experimental values (prediction error approx. 9%). In this article, we present calculations for much larger domains (Ω<sub><italic>L</italic></sub>) using the Finite Difference (FD) method, solving also for the pressure and velocity of a viscous liquid flowing through the Packed Spheres scenario. The ratio <italic>Vol</italic>(Ω<sub><italic>L</italic></sub>)/<italic>Vol</italic>(Ω<sub><italic>s</italic></sub>) is around 10<sup>3</sup>. The comparison of our results with the Packed Spheres example yields a prediction error of 5% for the intrinsic permeability. Additionally, numerical permeability calculations have been performed for Al foam samples. Our geometric modelling of the porous domain stems from 3D X‐ray tomography, yielding voxel information, which is particularly appropriate for FD. Ongoing work concerns the reduction in computing times of the FD method, consideration of other materials and fluids, and comparison with experimental data. (© 2013 Wiley‐VCH Verlag GmbH & Co. KGaA, Weinheim)</p> </abstract>