Menu:

Research

My research interests focus on accretion discs physics, turbulence, dynamo action and planet formation. You'll find in these pages some informations about these researches plus some flash movies (just click on the snapshots to see the videos). The links below will direct you to a specific topic.
Turbulent convection in accretion discs
Subcritical baroclinic instability
Vortex stability
Magnetorotational instability
Accretion disc dynamo

Turbulent convection in accretion discs

It is often assumed that the magnetorotational instability (MRI) is responsible for the angular momentum transport in accretion discs. Recently, we have shown (with Gordon Ogilvie) that turbulent convection, driven by the disc vertical entropy profile, could also transport angular momentum outward. This result is new and interesting since previous studies found only inward angular momentum transport for convection in discs (e.g. Cabot 1996, Stone and Balbus 1996). We have shown that outward angular momentum transport in convectively active discs was due to small non axisymmetric turbulent motions, which are excited when the dissipation coefficients (viscosity, thermal diffusivity) are small enough. The animations below show a convectively active disc in the high and low dissipation regimes. Details can be found in our paper on turbulent convection.

Below: High dissipation case similar to Stone and Balbus (1996) transporting angular momentum inward (Ra=3e5, Ri=0.4). The flow is essentially axisymmetric (very elongated structures in the y direction).

Get the Flash Player to see this player.

Below: Low dissipation case transporting angular momentum outward (Ra=1.4e7, Ri=0.4). Small non axisymmetric structures appear in this case.

Get the Flash Player to see this player.

The subcritical baroclinic instability (SBI)

Baroclinic instabilities in discs have regain interest during the past few years (see our paper on the SBI for a more extensive introduction). I present here the results of a 3D simulation in which 2D vortices are produced by the subcritical baroclinic instability (SBI). In this simulation, 3D parametric instabilities are also present in the core of these vortices, producing 3D turbulence in each vortex. This animation represents the vertical vorticity (left) and the vertical velocity (right). Vortices are easely identified as negative regions of vorticity (blue regions in the left plot). Chaotic vertical motions are observed in the core of these vortices, demonstrating the presence of turbulence in these structures...

Get the Flash Player to see this player.


Vortex stability

The existence of vortices in accretions discs has been debated since the first discs models, about 30 years ago. Today, we think that these vortices can help planet formation and the accretion process itself if discs are too cold to be linked to magnetic fields. It is a very appealing idea, and it's actually quite easy to get vortices in a 2D disc. However, discs are 3D, and some simulations have shown that 2D vortices were highly unstable to 3D perturbations. I'm currently studying these instability to find whether or not accretion disc vortices were always unstable... You'll find below an example of an unstable elliptical vortex in a disc. The colour volume rendering represents the vertical velocity and the grey surface defines the vortex boundaries (it's an isocontour of vorticity). As you can see, 3D vertical motions grow inside the vortex, and the vortex structure is destroyed after about 5 orbit by a very powerfull instability...

Get the Flash Player to see this player.


The Magnetorotational instability

A part of my PhD was about the magnetorotational instability (or shortly MRI). This instability leads to strong turbulent motions in discs, and we think that this turbulence is responsible for the accretion process. During my PhD, we found (in collaboration with P-Y. Longaretti and S. Fromang) that the intensity of the MRI was partly controled by non ideal effects (viscous dissipation and ohmic resistivity). Why this happens is still a very open question... You'll find below a simulation showing the MRI in a shearing box with a mean vertical field (it's a Pm=1 Re=3200 simulation from the 2007 MNRAS paper). It shows the radial velocity fluctuations during 5 orbits (50 shear times).

Get the Flash Player to see this player.


Dynamo in discs

Since discs are basically differentially rotating MHD objects, one may wonder if dynamo action (generation of magnetic field from turbulent motions) can take place. It seems that this process works, at least in local simulations known as zero net flux shearing boxes. With G. Ogilvie, we have found that this dynamo was due to a cyclic process involving an axisymetric azimuthal magnetic field, regenerated by non axisymmetric waves. The exact process is described extensively in two papers (see Publications). Below is an example of such a cycle. It shows the azimuthal field, initially random. Progressively, we see that a large vertical mode appears, which corresponds to the maximum of the cycle. Then, non axisymmetric waves (unstable for the MRI) destroys the structure, and the cycle start again. Interestingly, this dynamo works because the MRI amplifies strongly the non axisymmetric waves. It is therefore a non linear dynamo, since the MRI requires a non negligible Lorentz force.

Get the Flash Player to see this player.