Dense Stellar Systems as Laboratories for Fundamental Physics

Abstract

Galactic nuclei and globular clusters act as laboratories in which nature experiments with normal stars, neutron stars and black holes, through collisions and through the formation of bound states, in the form of binaries. The main difference with the usual Earth-based laboratories is that we cannot control the experiments. Instead, we have no choice but to create virtual laboratories on Earth, in order to simulate all the relevant physics in large-scale computational experiments. This implies a realistic treatment of stellar dynamics, stellar evolution, and stellar hydrodynamics. Each of these three fields has its own legacy codes, workhorses that are routinely used to simulate star clusters, stars, and stellar collisions, respectively. I outline the main steps that need to be taken in order to embed and where needed transform these legacy codes in order to produce a far more modular and robust environment for modeling dense stellar systems. The time is right to do so: within a few years computers will reach the required speed, in the Petaflops range, to follow a star cluster with a million stars for ten billion years, while resolving the internal binary and multiple star motions. By that time simulation software will be the main bottleneck in our ability to analyze dense stellar systems. Only through full-scale simulations will we be able to critically test our understanding of the `microphysics' of stellar collisions and their aftermath, in a direct comparison with observations.