The numerical impact on the star formation rate of GMC collisions
Glen Hunter
Wednesday, Dec. 7th, 16:30CET
Giant molecular clouds (GMCs) are the predecessors of star forming regions, spanning tens of parsecs and containing thousands of solar masses worth of molecular gas. These clouds are shaped and guided by the large scale motions of their host galaxy. Simulations have shown that these GMCs can collide in disk galaxies and are a fairly common occurrence. When they collide, a region of much denser gas is formed resulting in a subsequent burst of star formation. Observations of these collisions have found massive stars near the compressed gas layer. As such, cloud collisions are believed to be one mechanism in which high mass stars can form. Due to the burst of star formation, the star formation rate (SFR) of the GMCs increase. How much the SFR increases by the collision of clouds continues to be debated in the literature. Smooth particle hydrodynamics (SPH) based simulations of cloud-cloud collisions have shown an increase of a factor of less than two whilst grid based simulations have shown an increase of a factor of ten. It then becomes important to understand why such a difference arises. We run a group of simulations with the same initial conditions as Wu et al. 2017 (a grid based simulation) with AREPO (a quasi-Lagrangian code) with variations in collisional velocity, magnetic field inclination and resolutions to determine which parameter has the greatest impact on the SFR. We find that increasing the collisional velocity initiates star formation sooner, and non-parallel magnetic fields to the collisional axis delays the onset of star formation. In our simulations, the collisions provide an increase in the SFR of a factor of around 2-3, consistent with previous SPH studies, compared to Wu et al study, that sees up to a factor of 10 increase in the SFR for the same collision velocities. We demonstrate that much of the dense gas that forms does not become gravitationally bound, and so stars do not form by our sink particle recipe. In contrast, the simple “density threshold” star formation prescription adopted by Wu et al. will turn this unbound gas into stars. We conclude that while density threshold prescriptions for star formation may be appropriate for galaxy-scale simulations, they are not a valid model at the scales of GMCs and below.