Research:

Please feel free to contact me for more informations or/and more recent updates on these different research fronts.

Noble gas diffusion, the importance of crystal shape and diffusion anisotropy on thermochronology (with Bill Cassata and Paul Renne):

Ar or He diffusion in crystal is often approximated to 1D diffusion (diffusion out of a sphere, infinite cylinder or infinite slab) for simple analytical treatment. Recent studies have shown that for small anisotropy (shape and/or diffusion coefficients) diffusive flux of Ar/He out of crystal can be successfully approximated by a sphere with an equivalent Surface to Volume ratio provided the anistropy is small. In this work we derive a new analytical model to offers a better correction for a much greater range of anisotropy. We call this model Average Normalized Distance (AND). The manuscript will be submitted soon (summer 2010).

 

Additionally, I developed a new numerical model for diffusion that can deal with arbitrarily complex boundary topology with no additional cost in terms of algorithm. A separate paper for this model will be submitted shortly (summer 2010). This code allows users to basically scan thin section and solve for diffusion into/out of crystals. Here is an example of the method with Buzz the mascot of GTech:

 

Reactive porous media flow in pressure vs buoyancy-driven flows. (with Andrea Parmigiani and Joe Dufek):

In this study, we investigate the difference between buoyancy and pressure-driven flows in porous media. According to Darcy's theory, the two types of flow are equivalent when the piezometric head gradient is identical. Darcy's equation however describes the discharge through a porous medium at scales much greater than the pore-scale. We show however that the flow-fields (at the pore-scale) are, in most cases, different. The difference depends on the tortuosity of the porous medium (increases with more tortuous medium) and as such is often correlated with porosity. The difference in flow field at the pore-scale is shown to be large enough in many case to affect the discharge (and therefore the computed permeability), this effect becomes stronger when the flow advects reactants that dynamically affect the permeability of the porous medium.

Difference of velocity field magnitude between porous and buoyancy-driven flows (with same piezometric head gradients). The peak difference is ~ 20 times darcy velocity for scale.

Melt extraction from the mantle, continental growth and volcanic flare-up episodes (with Joe Dufek):

In this project, we investigate the role of crustal foundering in the generation of high melt fluxes from the mantle and test their ability to produce flare-up type magmatism. We use a multiscale numerical approach where the mantle flow response is calculated from a finite volume multiphase numerical model...

and the production and extraction of melts is calculated with a combined crystal nucleation and growth algorithm and a lattice Boltzmann 3D porous media flow (to parameterize the evolution of the permeability)...

Magma chamber dynamics (with Olivier Bachmann and Michael Manga):

I am working on several projects related to magma chamber processes trying to address the following questions:

Figure showing the different regimes controlled by the injection of volatiles exsolved by a cooling underplating magma (with various initial water contents x-axis) on the reactivation of crystal mushes with different composition (y-axis):

Conduit flow and eruption dynamics (with Joe Dufek and Andrea Parmigiani):

The explosivity of volcanic eruptions is controlled by the viscosity of the magma and the ability of volatiles to escape the magma during its ascent in the conduit. Together with Joe Dufek, we are developing a multiscale numerical approach to include effects such as bubble coalescence, deformation and the development of percolating gas pathways and outgassing. We parameterize the evolution of the bubble size distribution due to coalescence and bubble deformation with a 3D multiphase lattice Boltzmann model...

These parameterizations are then used in the 2D multiphase transient conduit flow of Dufek and Bergantz (2005).

 

I am currently approaching the problem differently numerically, using a lattice Boltzmann model developed by Dr Koerner (University of Erlangen) for multiphase flows. This model better handles capillary forces and therefore offers a better control of coalescence (a key concept in this project).

 

Multiphase fluid dynamics (multiple projects...):

Together with Jim Watkins and Michael Manga, we calculated the shape of a buoyant bubble ascending along a sloping boundary at low Reynolds number (and in the miscible limit). Experimentally, we found that the shape of the bubble at steady-state is independent of the slab angle (when >0 and < 40 degrees). We used a Boundary Integral Method and scaling arguments which confirmed the experimental results.

ascending bubble... (picture J. Watkins)

 

Capillary instabilities during the migration of buoyant non-wetting fluids in saturated porous media (with Andrea Parmigiani, Jonas Latt and Joe Dufek):

I am also investigating the development of capillary instability during the injection of a buoyant non-wetting fluid phase in a saturated porous medium. I predict from theoretical arguments that the non-wetting phase will get channelized even in an homogeneous porous medium because of capillary effects. Together with Andrea Parmigiani and Jonas Latt, I am developing a lattice Boltzmann model to test this instability on Teragrid super-computers.

Here is a snapshot of capillary instabilities (fingers). The porous medium is constructed such as to have channels with identical pore structure (separated by low porosity channels) and are subjected to the same volume flux of gas. The instability leads to heterogenous distribution of gas amongst the channels (some channels growing at the expense of others).

The numbers on each channel (second row of numbers) represent the gas saturation in each channel. Odd number channels have a high porosity and therefore control the mass transfer of gas.

Clathrate destabilization (with Wendy Mao and Bruce Buffett):

This project is about to start. We are interested to quantify the mobility of methane in a saturated porous medium during the destabilization of clathrate reservoirs during climate changes. To address this question we plan on using X-ray tomography images of real sediments and use a multiphase lattice Boltzmann model to quantify the motion of methane through the porous medium. Snapshots of multiphase flow in a porous medium - in red -, with the buoyant bubbles in blue:

Development of lattice Boltzmann numerical methods (with Andrea Parmigiani and Bastien Chopard):

The different projects I am involved in often require the development of new numerical techniques, moslty based on the lattice Boltzmann method. For instance, I developed a model for fluid flow and melting-solidification (see publications)... here convection melting in a square enclosure heated from the left (color coding for temperature)

Dispersion of diffusing chemical solutes (coupled through charge balance) in porous media flow (flow form the left)...

Compositional Rayleigh-Taylor instabilities triggered by solute concentration gradients (diffusing from the top). The different ions are coupled through charge conservation.