Reaction of massive clusters to gas expulsion – The cluster density dependence

Abstract

Context. The expulsion of the unconverted gas at the end of the star formation process potentially leads to the expansion of the just formed stellar cluster and membership loss. The degree of expansion and mass loss depends largely on the star formation efficiency and scales with the mass and size of the stellar group when stellar interactions can be neglected. Aims. We investigate the circumstances in which stellar interactions between cluster members become so important that the fraction of bound stars after gas expulsion is significantly altered. Methods. The Nbody6 code is used to simulate the cluster dynamics after gas expulsion for different star formation efficiences. Concentrating on the most massive clusters observed in the Milky Way, we test to what extent the results depend on the model, i.e. stellar mass distribution, stellar density profile etc., and the cluster parameters, such as cluster density and size. Results. We find that stellar interactions leading to ejections are responsible for up to 20% mass loss in the most compact massive clusters in the Milky Way. Therefore, ejections are the prime mass loss process in these systems. Even in the loosely bound OB associations, stellar interactions are responsible for at least ∼5% mass loss. The main reason why the importance of encounters for massive clusters has been largely overlooked is because of the often used approach of a single-mass representation instead of a realistic distribution for the stellar masses. The density dependence on the encounter-induced mass loss is shallower than expected because of the increasing importance of few-body interactions in dense clusters compared to sparse clusters where 2-body encounters dominate.

The bound fraction of young star clusters

“…Should ε SFE be higher in the central regions, we could expect a higher stellar density compared to a constant ε SFE over the whole cluster. Simulations discussed in Pfalzner & Kaczmarek (2013) compare King-type and Plummer-type clusters, showing a significant difference for the bound fraction around ε SFE = 0.33, whereas other ε SFE has matching results. This high sensitivity to the density profile, and more generally to the phase-space density function (Boily & Kroupa 2002, 2003, should be addressed in future studies.…”

“…Notably, the relative bound mass after 20 Myr, as a function of ε SFE and initial cluster density, is studied by Pfalzner & Kaczmarek (2013), whereas our results give an upper limit on the bound fraction for a given initial condition, without focusing on a certain point in time. Their bound fractions after 20 Myr would consequently be smaller than those determined with our method.…”

Cluster Dynamics Largely Shapes Protoplanetary Disk sizes

“…Clusters depicted,e.g., in Lada & Lada (2003) usually have somewhat smaller radii (<1 pc) as they still form stars. There are strong indications that clusters' sizes increase with age during the star formation process and are typically about 1-2 pc by the time star formation is finished (Kroupa 2005;Pfalzner & Kaczmarek 2013a;.…”

Does the mass distribution in discs influence encounter-induced losses in young star clusters?

“…Therefore, a set of random initial stellar positions, velocities, and masses has been achieved for each run, which are analysed and averaged in a subsequent step. It turned out that tracking at least 500 000 stellar trajectories (Pfalzner & Kaczmarek 2013) leads to an error of <3% for each of the presented results. For an initial number of stars N star = 1 000, around 500 simulations were performed while 16 simulations were performed when N star = 32 000.…”