My scientific interests are located at the intersection between statistical physics, fluid mechanics and scientific computing. I am interested in fluid flow phenomena in the presence of complex fluid-fluid or fluid-solid interfaces. I gain insight into these by combining high-performance numerical simulations and methods from statistical physics. Here follows more information on the overarching themes:

Flow, fracture and stress in geophysical systems

The application of stress to heterogeneous disordered solid-liquid systems leads to a feedback loop between flow and deformation. Regions of high stress are prone to dissolution and/or crack formation, while regions of low stress are generally more prone to mineral precipitation. This changes the geometry, and subsequently the fluid flow paths which control the distribution of precipitate – closing the feedback loop between flow and deformation. We have studied the fracture and creep of idealized materials on the microscopic scale (using molecular dynamics). Further, using data from real rock samples, we studied the stress formation in evolving porous microstructures. We have also studied the channeling effects in charged model microfractures.

Transitional flow in a self-affine fracture.

Current research is focused the role of inertia for flow through self-affine fractures, and the role of roughness on the laminar-turbulent transition in wall-bounded shear flows.

Collaborators: Joachim Mathiesen, François Renard, Luiza Angheluta.

Computational microfluidics

Flow above a dead-end pore.

Micro- and nanofluidics are areas of rapid growth which are expected to continue to yield technological progress. At such small scales, electrokinetic effects may be essential. Such flow of ion-laden fluids under influence of electric fields is called electrohydrodynamics. For example, electrohydrodynamic effects alter the wetting properties of two-phase systems, which also may have consequences in geophysical systems.

Reliable computational methods are necessary to complement physical experiments, to validate theoretical models, and to enable rapid prototyping of devices. As much of computational microfluidics (especially with electrohydrodynamics) relies on black-box software, we have developed the open-source phase-field solver Bernaise (built on top of the finite-element framework FEniCS). Current research efforts include developing new, stable schemes for the strongly coupled problem of electrohydrodynamics.

Models for non-equilibrium multiphase flow

We have used and developed phase-field models for direct numerical simulation of multiphase fluid flows. At the scale of pipelines (meters to kilometers), it is, however, unfeasible to resolve the entire flow field. For such applications, which are highly relevant with regard to CO2 transport, the multiphase flow is represented as a system of (hyperbolic) first-order partial differential equations, wherein the complicated fluid-fluid and fluid-solid interactions are represented by effective relaxation source terms. Building on results from the literature, we have completed a hierarchy of simplified models. In particular (at least in the limit of equal phase velocities) every equilibrium assumption reduces the wave velocities of the relaxation system.

Collaborator: Tore Flåtten

General interests

  • Physics of complex systems
  • Statistical mechanics
  • Interface dynamics
  • Instabilities in fluid mechanics
  • Multiscale simulation
  • CO2 transport and storage (CCS)