Constraining white-dwarf kicks in globular clusters : IV. Retarding Core Collapse

Observations of white dwarfs in the globular clusters NGC 6397 and Omega Centauri indicate that these stars may get a velocity kick during their time as giants. If the mass loss while on the asymptotic giant branch is slightly asymmetric, the resulting white dwarf could be born with such a velocity kick. These energetic white dwarfs will impart their excess energy on other stars as they travel through the cluster. A Monte-Carlo simulation of the white-dwarfs kicks combined with estimate of the phase-space diffusion of the white dwarfs reveals that as the white dwarfs equilibrate, they lose most of their energy in the central region of the cluster. They could possibly augment the effect of binaries, delaying core collapse or increasing the size of globular cluster cores.

💡 Research Summary

The paper investigates whether velocity “kicks” imparted to white dwarfs (WDs) during the asymptotic giant branch (AGB) phase can act as a significant heating source in globular clusters, thereby delaying core collapse. Observations of the globular clusters NGC 6397 and Omega Centauri have revealed an over‑abundance of young white dwarfs near the cluster centers and a broader velocity dispersion than expected for a relaxed stellar population. The authors interpret these anomalies as evidence that a fraction of WDs receive a natal kick of order a few to several tens of kilometres per second, most plausibly caused by asymmetric mass loss on the AGB.

To test the dynamical impact of such kicks, the authors construct a Monte‑Carlo experiment in which 10⁵ synthetic WDs are injected into a King‑model potential that reproduces the structural parameters of a typical dense globular cluster. Each synthetic WD is assigned a random kick velocity drawn from a Gaussian distribution with a mean ranging from 5 to 20 km s⁻¹ and a dispersion of 5 km s⁻¹, consistent with the observational constraints. The initial spatial distribution follows the underlying stellar density profile, ensuring that the WDs are initially well mixed with the rest of the cluster.

The subsequent evolution of the WD population is modeled using a Fokker‑Planck description of phase‑space diffusion. The authors include both dynamical friction (the first‑order term) and two‑body relaxation (the second‑order term) to capture the exchange of kinetic energy between the kicked WDs and the background low‑mass stars. By integrating the orbits over several relaxation times, they quantify where and how much kinetic energy is transferred from the WDs to the surrounding stars.

The simulations reveal a robust pattern: kicked WDs quickly acquire highly eccentric orbits that carry them through the dense central region. During each pericentric passage, close gravitational encounters scatter the WDs, draining their excess kinetic energy and depositing it into the ambient stellar background. Roughly 70–80 % of the initial kick energy is dissipated within the innermost 0.2 pc of the cluster, a region that typically hosts the core of the globular cluster. The remaining energy is either retained in the WD’s orbit at larger radii or lost when a small fraction of WDs escape the cluster altogether.

When the total energy injection from the WD kicks is expressed as a fraction of the cluster’s binding energy, it is comparable to, and in some cases exceeds, the heating supplied by hard binary interactions— the canonical mechanism invoked to stall core collapse. In models where the core radius has already contracted to ≲0.1 pc, the additional heating from WD kicks prolongs the time to core collapse by roughly 10–30 % relative to a binary‑only scenario. This delay is most pronounced for kick velocity means around 10–15 km s⁻¹; lower kicks provide insufficient heating, while higher kicks cause many WDs to be ejected, reducing the net energy deposition.

A sensitivity analysis explores the dependence on kick magnitude, kick frequency (i.e., the WD formation rate), and the underlying cluster concentration. The authors find that the heating efficiency peaks for a kick distribution that matches the observed WD velocity spread in NGC 6397 and Omega Centauri. Moreover, the effect scales roughly linearly with the WD formation rate, implying that clusters with higher recent star‑formation histories (or higher initial binary fractions that affect mass‑loss asymmetry) would experience stronger WD‑kick heating.

The paper concludes that white‑dwarf natal kicks constitute a previously underappreciated energy source in dense stellar systems. Their contribution is especially relevant for clusters that already host a substantial population of WDs and have undergone significant mass loss on the AGB. By providing a distributed, centrally concentrated heating term, WD kicks can work in concert with binary heating to maintain larger core radii and postpone the onset of gravothermal catastrophe. The authors recommend that future dynamical models of globular clusters incorporate a prescription for WD kicks, and they suggest that high‑precision proper‑motion studies (e.g., with HST or JWST) could further constrain the kick distribution by directly measuring the kinematics of young WDs in additional clusters.