On the trapping of stars by a newborn stellar supercluster

Numerical experiments conducted by Fellhauer et al. (MNRAS, 372, 338, 2006) suggest that a supercluster may capture up to about 40 per cent of its mass from the galaxy where it belongs. Nevertheless, in those experiments the cluster was created making appear its mass out of nothing, rather than from mass already present in the galaxy. Here we use a thought experiment, plus a few simple computations, to show that the difference between the dynamical effects of these two scenarios (i.e., mass creation vs. mass concentration) is actually very important. We also present the results of new numerical experiments, simulating the formation of the cluster through mass concentration, that show that trapping depends critically on the process of cluster formation and that the amounts of gained mass are substantially smaller than those obtained from mass creation.

💡 Research Summary

The paper revisits the claim that a newly formed massive star cluster can capture a substantial fraction of its host galaxy’s stellar mass. Fellhauer et al. (2006) reported that up to ~40 % of a cluster’s final mass could be acquired from surrounding stars, but their simulations assumed that the cluster’s mass appeared “out of nothing” – a mass‑creation scenario that does not reflect how clusters actually form. The authors first present a thought experiment contrasting two extreme cases: (1) instantaneous mass creation, which deepens the galaxy’s gravitational potential globally, and (2) mass concentration, where the cluster assembles from pre‑existing galactic material. By applying simple energy‑ and angular‑momentum conservation arguments they show that the capture radius and the amount of bound stellar mass are dramatically larger in the creation case than in the concentration case.

To test these analytic expectations, the authors conduct a suite of new N‑body simulations in which a massive cluster forms by gradually gathering gas particles from a realistic galactic disk‑halo system. The simulations explore a range of formation timescales, initial gas density profiles, and thermodynamic conditions. In all runs the fraction of pre‑existing stars that become bound to the nascent cluster remains below 5 % (even under the most favorable conditions it never exceeds ~10 %). This is an order of magnitude lower than the 40 % reported by Fellhauer et al. The modest increase in capture efficiency for very rapid mass‑concentration events still falls far short of the creation scenario.

The authors identify two key reasons for the discrepancy. First, creating mass instantaneously alters the global potential, allowing many stars to lose enough orbital energy to become bound. In contrast, concentrating mass locally produces only a modest, spatially limited deepening of the potential, so most stars retain sufficient kinetic energy to escape. Second, realistic cluster formation involves gas cooling, feedback, and angular momentum transport, which lengthen the assembly time and further reduce the chance of stellar capture.

The study concludes that the dynamical impact of cluster formation on surrounding stars is far less dramatic than previously thought. The amount of stellar mass that can be accreted by a newborn massive cluster through gravitational capture is likely only a few percent of the cluster’s final mass. Consequently, models of galaxy evolution that rely on large “self‑enrichment” of clusters via stellar capture must be revised, and the role of massive clusters in shaping their host galaxies should be reconsidered in light of these more modest capture efficiencies.