3D MHD simulation of dwarf novæ disc produced with the Idefix code.
Dwarf novæ are a type of outbursting binary systems composed of an accreting white dwarf and a solar-type secondary star. Because the two stars are so close, the secondary stars overflows from its Roche lobe and continously replenishes a disc around the white dwraf. Thanks to their short recurence time, of the order of one month, they are the ideal objects to study accretion. Theoretical and numerical models can then be strongly constrained by observations.
During the luminous phase, the outburst, the accretion disc gas is hot and well-ionised. The magneto-rotational instability can produce enough turbulence in the disc to drive the observed accretion. During the dim phase, the quiescence, however, the disc is cold and poorly ionised. This instability can not be sustained and the observed accretion level remain to be explained.
In my PhD work, I showed that in very cold quiescent disc, the spiral waves excited by the tidal potential of the companion star alone can not explain the observed accretion. They become increasingly more tightly wound with decreasing temperature and the angular momentum transport effieciency drops significantly. I also showed that during this phase, even with a weakly ionised disc, a global magnetic wind is launched from the disc and increases the accretion rate by a large amount. As is usually the case, the effieciency of the wind is related to amplitude of the magnetic field.
Accretion in TDE
Artist view of a star getting disrupted by a supermassive black hole. (Credit : ESO, ESA/Hubble, M. Kornmesser).
Accretion is a phenomena that plays a role in variety of astrophysical objects. The first type of object I was interrested on are Tidal Disruption Events, or TDE for short. TDE are violent phenomena that arise when a star wanders too close to a super-massive black hole. When the tidal forces exerted by the black hole become larger than the self-gravity of the star, the star is disrupted. Some of its matter will be gravitationnally bound to the black hole and form an acretion disc.
In these discs, matter will slowly fall towards the black hole while the angular momentum of the gas will be transported outwards. In doing so, the gas will be grealty heated by its viscosity and will thus emit light (the simplest model is to assume this emission to be black-body emission). More refined model can also take into account specific line emission of molecules whithin the hot gas, and even synchrotron emission from electron in the plasma.
However accretion is a phenomena arising in more systems than just TDE formed discs. Larger, brighter accretion discs in the centre of galaxies, around their supermassive black hole are also refered to as Active Galactic Nuclei (AGN). Even though the scales of those discs are larger than those formed by TDE, the physics at hand is subtentially the same. However more violent phenomena may also occur in those larger disc, for instance ultra relativistic jets. In a less violent fashion, but not simpler by any mean, protoplanetary discs (like what happened in our solar system), also are accretion dominated discs. Discs in binary systems like cataclysmics variable stars (or CV for short), where a white dwarf progressively absorbs its companion star, are also accretion discs. The main difference between these types of accretion discs is the mass of the object in their centre. Subject to different gravitationnal potential, the discs will reach more or less dense state, thus higher or lower temperature, on which their emission direclty depends.
Galaxy formation
Example of a simulated galaxy obtained with the FIRE code (from FIRE website).
The best way that we have to probe our understanding of the theory behind the formation of complex sturctures like galaxies is to use numerical simulation. We implement in codes like FIRE (Feedback In Realistic Environments) all the physics that we know of and let them run. Then we compare the galaxies formed by those simulations and actual observed galaxies. It is a way to probe our models : if the galaxies we obtain look very different to what we observe, then we might have forgotten some important phenomena.
I have been interrested in specific line emission in early galaxies. They are a very precise probe of the condition in both simulated and observed galaxies as lines can only be emitted in precise conditions. For instance CO line can not be observed if the temperature and radiation conditions are such that the C-O bond is not stable. Hence one has to compute chemical abundances in the simulated galaxies depending on those conditions in order to have a realistic simulated line emission.
