PoreLab lecture series with Daniel Lester on Fluid Mixing in Porous Media Flows of Arbitrary Complexity

Welcome to the next lecture in the PoreLab Lecture Series for the semester.

Who: Associate Professor Daniel Lester, from RMIT University, Australia

When: Wednesday 3 May at 13:00 (Oslo time)

Where: on zoom  https://uio.zoom.us/j/65837085049?pwd=WjZianUyN3FJa2liQkxBbzQrOCtGdz09

Title: Fluid Mixing in Porous Media Flows of Arbitrary Complexity

Abstract:

Mixing, dispersion and reaction of fluids and solutes in heterogeneous porous media is a fundamental problem in nature and man-made systems, ranging from e.g., geophysical processes in the subsurface to poro-elastic flows in brain tissue. Despite over a century of research, conventional approaches to fluid mixing and dispersion are largely based upon a macro-dispersion paradigm that does not properly resolve the underlying physics.

In recent years, significant theoretical advances have been made to better understand and quantify solute transport in porous media across scales using a combination of Lagrangian methods (dynamical systems theory, Hamiltonian chaos) and stochastic modelling (continuous time random walks, Markov models). In parallel, an explosion of novel experimental techniques can now quantify these processes with unprecedented resolution. In principle, these datasets are rich enough to facilitate ab initio predictions of fluid mixing, but until now it has been unclear how to utilise these.

In this talk I will present a general stochastic framework for mixing and dispersion in porous media flows of arbitrary complexity, i.e. from simple model flows (such as steady, non- chaotic flows) to complex flows (unsteady and chaotic flows in poro-elastic media) in heterogeneous media. This framework honours the topological constraints associated with simple flows, whilst providing flexibility to accommodate porous media flows of arbitrary complexity. I demonstrate application of this framework to a wide range of porous media flows and show how experimental data can be used to generate ab-initio predictions of fluid mixing and dispersion across a broad range of length scales.