A new multifluid method for dusty astrophysical flows

Context. Stars and planets form in collapsing clouds of gas and dust. The presence of dust grains and their local distribution play a significant role throughout the protostellar sequence, from the thermodynamics and the chemistry of molecular clouds to the opacity of collapsing protostellar cores and the coupling between the gas and the magnetic field and down to planet formation in young and evolved disks.Aims. We aim to simulate the dynamics of the dust, considering the whole range of grain sizes, from few nanometers to millimeters.Methods. We implemented a neutral pressureless multifluid that samples the dust size distribution in the RAMSES code. This multi-fluid is dynamically coupled to the gas via a drag source term and self-gravity, relying on the Eulerian approach.Results. We designed a Riemann solver for the gas and dust mixture that prevents unphysical dust-to-gas ratio variations for well coupled grains. We illustrated the capacities of the code by performing simulations of a protostellar collapse down to the formation of a first hydrostatic core, both for small and large dust grains. Grains over 100 microns significantly decouple from the gas. The spatial maps and the probability density functions indicate that dust enrichment within the first hydrostatic core and in some locations of the envelope increases as a function of the grain size and the level of initial turbulence.Conclusions. Thanks to the novel Riemann solver, we recovered the terminal velocity regime, even at low resolution. Moreover, we successfully extended it to regimes where the grain inertia matters. The multifluid module performs the coupling between the dust and the gas self-consistently all through the dynamical scales. The dust enrichment in the first hydrostatic core and the envelope have been revised here, assuming the initial turbulence and grain sizes. This enables us to probe new potential locations, epochs, and initial conditions for planet formation.