Shock waves and star formation using multigroup radiation hydrodynamics
Radiative transfer plays a major role in the field of astrophysics since it provides all the information we know about the universe. In many cases, as for example in stellar atmospheres, the radiation is considered as a physical probe which provides access to the thermodynamical properties of the flow through the spectrum of emission and absorption lines. However, the radiation often has a very important dynamical role in the system. It cannot only be considered as a passive probe, but as an integral part of the equations governing the system dynamics.
As the radiation intensity depends on seven variables in 3D, solving the full transfer equation coupled to the hydrodynamics to tackle radiation hydrodynamics problems is still out of reach of modern computational architectures. In order to overcome this difficulty, much effort has been spent in recent years developing mathematically less complicated, yet accurate approximations to the equations of radiative transfer. Such approximations use frequency and/or angle-integrated variables. Yet, the absorption and scattering coefficients almost always depend strongly on frequency, and the grey approximation is no longer appropriate. I will present a multigroup model for radiation hydrodynamics to account for variations of the gas opacity as a function of frequency.
We have applied the method to the study of the structures of radiative shocks as well as simulations of gravitational collapse in the context of star formation. We show that using even a small number of frequency groups can significantly influence the dynamics and morphologies of subcritical and supercritical radiative shocks, and in particular the extent of the radiative precursor.
As for our simulations of star-formation, we used a non-ideal gas equation of state as well as an extensive set of spectral opacities to model the first and second phases of the collapse of a molecular cloud core to form the first and second Larson cores. We find that the first core accretion shock remains supercritical while the shock at the second core border is strongly subcritical with all the accreted energy being transfered to the core, and that the size, mass and temperature of the second cores are independent of the parent cloud properties.
Finally I will present early results from 3D simulations we have performed with the AMR code RAMSES.