Effective transport parameters of porous media from 2D microstructure images

• Published in: International journal of heat and mass transfer 2021-08, Vol.175, p.121371, Article 121371
• Main Authors: Maier, Lukas, Scherle, Marc, Hopp-Hirschler, Manuel, Nieken, Ulrich
• Format: Article
• Language: English
• Subjects: Asymptotic homogenization; Asymptotic methods; Chemical engineering; Diffusion; Diffusion coefficient; Digital image-based modeling; Lumping; Mathematical models; Membrane, gas transport; Parameters; Porous materials; Porous media; Transport properties; Up-scaling
• ISSN: 0017-9310; 1879-2189
• DOI: 10.1016/j.ijheatmasstransfer.2021.121371

•An approach to calculate transport parameters of the dusty-gas model from a cross-sectional image of porous material is presented.•The method of asymptotic homogenization is applied to real porous material.•Effective binary and Knudsen-diffusion coefficients of porous material are determined by simulation.•Effective Knudsen-diffusion coefficients are determined by experiment.•The simulations are compared with experiments. Transport in porous media is central in chemical engineering. Effective transport properties are required for evaluating and improving many processes and applications. A widely used model to describe transport in porous media is the Dusty-gas model. This model accounts for permeation, Knudsen-diffusion and binary diffusion. The model parameters are commonly determined from experiments, for instance, performed in a Wicke-Kallenbach cell. The parameters derived in this way are spatially averaged effective parameters. In this contribution, we present an approach to calculate transport parameters of the Dusty-Gas model from a cross-sectional image of the porous material, which often reduces experimental effort. The approach is also applicable to virtually designed materials for which structural information is available from simulations, but no physical sample has yet been synthesized. The method applied here is asymptotic homogenization, which allows lumping structural information on the micro-scale into effective transport properties. In a previous contribution, we successfully demonstrated the calculation of effective permeability for a porous hollow fiber from image data. Here we extend the work to Knudsen diffusion and binary diffusion. To validate our results, we compare model predictions for Knudsen-diffusion with experimental data from measurements in a Wicke-Kallenbach cell and predictions of the effective binary diffusion coefficient by comparison to literature data for a sphere packing.