Placing stars within cosmological simulations

Abstract

I investigate the process of converting gas into stars within the framework of a standard cosmological model. By examining the set of objects grown in a combined N-body plus smoothed particle hydrodynamics simulation with those obtained in similar models where some of the cold, dense gas was replaced by collisionless ``star'' particles I show that it is possible to make this substitution without affecting the subsequent gas cooling rate. With even the most basic star forming criteria the masses of isolated objects are nearly identical to the mass of cold, dense gas found within the same objects in a non-star forming run. No evidence is found to support the contention that converting gas into stars might affect the amount of cold gas obtained in a simulation by retarding the cooling rate within those objects where stars have already formed. In practice, because cold gas can be reheated by shocks but stars remain as such whatever happens the masses of the largest objects found in the star forming runs are generally higher than those in the standard run. Finally, I demonstrate that an excellent match to the observed star formation rate can be achieved with even a very basic star formation prescription.