Formation of satellites from cold collapse

TL;DR

This paper investigates how cold, initially irregular clouds of particles collapse under gravity, forming a main object and satellite structures, with the process influenced by initial conditions and leading to complex angular momentum and virialization behaviors.

Contribution

It introduces a detailed analysis of satellite formation during gravitational collapse, highlighting the role of initial asymmetries and fluctuations in shaping satellite properties.

Findings

Satellites form with finite lifetimes and can be longer than the relaxation time.

Initial asymmetries and density fluctuations significantly influence satellite abundance.

Satellites may not be virially relaxed despite being gravitationally bound.

Abstract

We study the collapse of an isolated, initially cold, irregular (but almost spherical) and (slightly) inhomogeneous cloud of self-gravitating particles. The cloud is driven towards a virialized quasi-equilibrium state by a fast relaxation mechanism, occurring in a typical time $\tau_c$, whose signature is a large change in the particle energy distribution. Post-collapse particles are divided into two main species: bound and free, the latter being ejected from the system. Because of the initial system's anisotropy, the time varying gravitational field breaks spherical symmetry so that the ejected mass can carry away angular momentum and the bound system can gain a non-zero angular momentum. In addition, while strongly bound particles form a compact core, weakly bound ones may form, in a time scale of the order of $\tau_c$, several satellite sub-structures. These satellites have a finite lifetime that can be longer than $\tau_c$ and generally form a flattened distribution. Their origin and their abundance are related to the amplitude and nature of initial density fluctuations and to the initial cloud deviations from spherical symmetry, which are both amplified during the collapse phase. Satellites show a time dependent virial ratio that can be different from the equilibrium value $b\approx -1$: although they are bound to the main virialized object, they are not necessarily virially relaxed.