An improved sink particle algorithm for SPH simulations

TL;DR

This paper introduces a new sink particle algorithm for SPH simulations that improves the physical accuracy and stability of sink creation and evolution, especially in optically thick, heated regions.

Contribution

The paper presents a novel sink particle algorithm that reduces spurious creation, manages accretion and angular momentum transfer, and enhances simulation robustness.

Findings

Reduces spurious sink formation.

Mitigates non-physical perturbations during accretion.

Produces sink properties independent of user parameters.

Abstract

Numerical simulations of star formation frequently rely on the implementation of sink particles, (a) to avoid expending computational resource on the detailed internal physics of individual collapsing protostars, (b) to derive mass functions, binary statistics and clustering kinematics (and hence to make comparisons with observation), and (c) to model radiative and mechanical feedback; sink particles are also used in other contexts, for example to represent accreting black holes in galactic nuclei. We present a new algorithm for creating and evolving sink particles in SPH simulations, which appears to represent a significant improvement over existing algorithms {\refrpt -- particularly in situations where sinks are introduced after the gas has become optically thick to its own cooling radiation and started to heat up by adiabatic compression}. (i) It avoids spurious creation of sinks. (ii) It regulates the accretion of matter onto a sink so as to mitigate non-physical perturbations in the vicinity of the sink. (iii) Sinks accrete matter, but the associated angular momentum is transferred back to the surrounding medium. With the new algorithm -- and modulo the need to invoke sufficient resolution to capture the physics preceding sink formation -- the properties of sinks formed in simulations are essentially independent of the user-defined parameters of sink creation, or the number of SPH particles used.