Radiation-Hydrodynamic Simulations of the Formation of Orion-Like Star Clusters II. The Initial Mass Function from Winds, Turbulence, and Radiation

Abstract

[abridged] We report a series of simulations of the formation of a star cluster similar to the Orion Nebula Cluster (ONC), including both radiative transfer and protostellar outflows, and starting from both smooth and self-consistently turbulent initial conditions. Each simulation forms >150 stars and brown dwarfs, yielding a stellar mass distribution from < 0.1 to > 10 Msun. We show that a simulation that begins with self-consistently turbulence embedded in a larger turbulent volume, and that includes protostellar outflows, produces an initial mass function (IMF) consistent both with that of the ONC and the Galactic field. This is the first simulation published to date that reproduces the observed IMF in a cluster large enough to contain massive stars, and where the result is determined by a fully self-consistent calculation of gas thermodynamics. This simulation also produces a star formation rate that, while still somewhat too high, is much closer to observed values than if we omit either the larger turbulent volume or the outflows. Moreover, we show that the combination of outflows, self-consistently turbulent initial conditions, and turbulence continually fed by motions on scales larger than that of the protocluster yields an IMF that is in agreement with observations and invariant with time, resolving the "overheating" problem in which simulations without these features have an IMF peak that shifts to progressively higher masses over time. The simulation that matches the observed IMF also reproduces the observed trend of stellar multiplicity strongly increasing with mass. We show that this simulation produces massive stars from distinct massive cores whose properties are consistent with those of observed massive cores. However, the stars formed in these cores also undergo dynamical interactions that naturally produce Trapezium-like hierarchical multiple systems.