Thumbnail Image

The effects of gasdynamics, cooling, star formation, and numerical resolution in simulations of cluster formation

We present the analysis of a suite of simulations of a Virgo-mass galaxy cluster. Undertaken within the framework of standard cold dark matter cosmology, these simulations were performed at differing resolutions and with increasingly complex physical processes, with the goal of identifying the effects of each on the evolution of the cluster. We focus on the cluster at the present epoch and examine properties including the radial distributions of density, temperature, entropy, and velocity. We also map "observable" projected properties such as the surface mass density, X-ray surface brightness, and Sunyaev-Zeldovich signature. We identify significant differences between the simulations, which highlights the need for caution when comparing numerical simulations to observations of galaxy clusters. While resolution affects the inner density profile in dark matter simulations, the addition of a gaseous component, especially one that cools and forms stars, affects the entire cluster. For example, in simulations with gasdynamics but no cooling, improving the gravitational force resolution from 200 to 14 kpc increases the X-ray luminosity and emission-weighted temperature by factors of 2.9 and 1.6, respectively, and it changes the form of the X-ray surface brightness and temperature profiles. At the higher resolution, a simulation that includes cooling and star formation converts 30% of the cluster baryons into stars and produces a massive central galaxy that substantially alters the cluster potential well. This cluster has 20% higher X-ray luminosity and 30% higher emission-weighted temperature than the corresponding cluster in the no-cooling simulation. Its properties are reasonably close to those of observed X-ray dominant (XD) clusters, with conversion of cooled gas into stars greatly reducing the observational conflicts found by Suginohara & Ostriker in simulations with cooling but no star formation. We conclude that both resolution and included physical processes play important roles in simulating the formation and evolution of galaxy clusters. Therefore, physical inferences drawn from simulations that do not include gaseous components that can cool and form stars present a poor representation of reality.