The loss of the intracluster medium in globular clusters

Abstract

Stars in globular clusters (GCs) lose a non-negligible amount of mass during their post-main-sequence evolution. This material is then expected to build up a substantial intracluster medium (ICM) within the GC. However, the observed gas content in GCs is a couple of orders of magnitude below these expectations. Here, we follow the evolution of this stellar wind material through hydrodynamical simulations to attempt to reconcile theoretical predictions with observations. We test different mechanisms proposed in the literature to clear out the gas such as ram-pressure stripping by the motion of the GC in the Galactic halo medium and ionization by UV sources. We use the code ramses to run 3D hydrodynamical simulations to study for the first time, the ICM evolution within discretized multimass GC models including stellar winds and full radiative transfer. We find that the inclusion of both ram pressure and ionization is mandatory to explain why only a very low amount of ionized gas is observed in the core of GCs. The same mechanisms operating in ancient GCs that clear the gas could also be efficient at younger ages, meaning that young GCs would not be able to retain gas and form multiple generations of stars as assumed in many models to explain ‘multiple populations’. However, this rapid clearing of gas is consistent with observations of young massive clusters.

On the role of Type Ia supernovae in the second generation star formation in globular clusters

“…It has to be noted that in all our simulations we do not include ionizing feedback from SG stars which could lead to a steeper decrease of the SFR (Chantereau et al 2020). Massive stars belonging to the SG, if any, should contribute to heat up the ISM further, reducing the amount of gas eligible for star formation.…”

“…The low density case is characterized by 𝜌 pg = 10 −24 g cm −3 while for the high density case we assume a gas density 10 times greater. Ours is one of the first "wind tunnel" experiment on cluster scale (Priestley et al 2011, C19), in which the effects of feedback are investigated in 3D (together with Chantereau et al 2020, who included photoionization). In our setup, the cluster is maintained fixed in its position and, at time 𝑡 inf , gas is allowed to flow into the computational box from one of the boundaries.…”

Star Cluster Ecology: Revisiting the Origin of Iron and Age Complex Clusters

“…Typically, other than nuclear star clusters (Neumayer et al 2020), clusters (open clusters, YMCs or GCs) are not able to accrete significant amounts of gas from their surroundings. This is due to a combination of ram pressure and stellar feedback (winds, SNe, and photionisation), which act to expel the ejecta from stars within the cluster and stop accretion from the surroundings (e.g., Chantereau et al 2020). Additionally, most GCs have orbits away from the disk, both spatially and kinematically.…”

Simulations of globular clusters within their parent galaxies: multiple stellar populations and internal kinematics

“…This gas must be expelled in order to explain the lack of neutral ISM observed in Galactic GCs. Simulations by Chantereau et al (2020) found that both ram pressure stripping and ionisation is mandatory to explain the small amount of ionized gas in the core of GCs. We intend to investigate these channels of gas expulsion in future simulations of our Galactic GCs.…”