
Niveau de raffinement et carte de densité sur un grumeau dans une des simulations.
Primordial galaxies are very different from the galaxies that surround us. The disks of these galaxies are not composed of spirals, but of sustained star-forming lumps that are highly turbulent. Yet it is this turbulence that generates the supersonic motions capable of strongly compressing the gas and triggering its gravitational collapse and conversion into stars.
.
The origin of this turbulence is still debated: is it generated by galaxy dynamics or by young stars, through their ionizing radiation or supernova explosions? However, local dynamics depend on global morphology, and young stars will not have the same effect if they are localized in lumps or spread out in spiral arms. These two processes are therefore likely to differ between galaxies in the young and current Universe.
The aim of my work is to understand what feeds this turbulence and what effect it has on star formation in galaxy disks over the course of their cosmic history. To do this, we need to model the interaction between the overall dynamics of the galaxy and star formation, on the one hand, and the gas in the disk, on the other. The numerical challenge comes from the dynamic range of spatial scales that come into play and the associated memory hit.
To do this, I used GENCI resources to develop a zoom method inside a lumpy galaxy.
Thus, the entire dynamics of the galaxy is captured, together with the effect of star formation, at a maximum resolution of 0.38 parsec.
In order to differentiate the effect of dynamics from that of stars, I studied the consequences of numerically stopping the effects due to young stars. According to these preliminary studies, the turbulent energy would decrease by only 20%, implying a predominance of dynamical effects over stellar effects.