Institution(s): 1. Princeton
Radiation feedback from massive clusters is expected to play a key role in setting the rate and efficiency of star formation on the scale of Giant Molecular Clouds (GMCs). However, due to the extreme cost of implementing full radiative transfer in 3D hydrodynamic simulations, the influence of radiation feedback on GMCs has been poorly understood. We employ the recently developed Hyperion extension of the Athena code, which solves the equations of radiation hydrodynamics (RHD) using the Reduced Speed of Light (RSL) approximation and M1 closure of the moment equations, to investigate the effects of direct, non-ionizing UV radiation on cloud dynamical evolution and star formation. Our model GMCs span a range of surface densities between 10 and 500 solar masses per square parsec, making them optically thick to UV and thin to reprocessed IR.
We find that radiation feedback has little effect on the density structure in the cloud or its star formation rate, both of which are set by the interaction between turbulence and gravity. Instead, the main effect of radiation is to truncate star formation and disperse gas rapidly when
a sufficiently luminous cluster has formed. We show that our numerical results can be explained by a simple paradigm of feedback-limited star formation that operates across a wide range of cloud surface densities. In this model, stars form steadily in a turbulent medium with log-normally distributed surface and volume densities, and successively larger portions of the original cloud become unbound when the forces on successively denser local patches of gas become super-Eddington. The global stellar efficiency in a GMC is therefore set not by the radiative force at the mean cloud surface density, but by the Eddington ratio in the high surface density tail of the gas distribution.