Micro-Scale Modelling of Flow and Particle Transport In Porous Media Via CFD

Porous media are ubiquitous in chemical engineering. They are involved in many separation processes (e.g., filtration, chromatography) and they play an important role in numerous chemically reacting systems such as in catalysis and in remediation of contaminated aquifers. Very often the fluids involved are characterized by very diverse rheological properties, ranging from classical Newtonian to non-Newtonian behaviors. Although the final applications are very different from each other and are typically characterized by very different features there are some common issues that despite the huge amount of experimental data still need to be addressed. These are generally related to the need of describing the porous medium at the macro-scale as a continuum defined in terms of a set of parameters that quantify its flow properties (e.g., porosity, permeability) and its particle transport characteristics (e.g., collection efficiency). This work (co-funded by European Union project AQUAREHAB FP7 - Grant Agreement Nr. 226565) aims at tackling these issues from a fundamental perspective by the use of pore-scale (or micro-scale) simulations via computational fluid dynamics. Porous media with different characteristics have been analyzed and used to extract some representative two-dimensional geometries of a few millimeters in size. Grids were created with Gambit whereas the flow of Newtonian and non-Newtonian (i.e., Carreau and Cross models) fluids through these media was investigated with Fluent 12.0. Data were analyzed in terms of the flow field and pressure drops across the media for different superficial velocities ranging from laminar to turbulent flows. Simulation predictions were then analyzed with different classical approaches (i.e., Darcy, Ergun, Darcy-Forchheimer equations) and the corresponding parameters were eventually extracted. Results show that the considered geometry are large enough to model and represent a fraction of the porous medium and allowed to correct and extend the Darcy-Forchheimer law to non-Newtonian fluids. In the second part of the work the presence of solid particles suspended in the fluid was considered. The collection efficiency of the different porous media was then quantified with computational fluid dynamics by using an Eulerian approach. The gathered simulation predictions were then compared with classical approaches for the estimation of the collection efficiency of the porous medium (i.e., classical filtration theory). Results show that these theories generally under-predict the overall efficiency and a correction was eventually formulated. Future work includes the use and implementation of these corrections in macro-scale simulations.