A general framework to test gravity using galaxy clusters VI: Realistic galaxy formation simulations to study clusters in modified gravity

TL;DR

This paper develops a retuned baryonic physics model for large-scale cosmological simulations, enabling the study of galaxy clusters in modified gravity scenarios with high accuracy and large sample sizes.

Contribution

It introduces a new, lower-resolution simulation model that accurately reproduces galaxy and cluster properties, and applies it to large-scale Hu-Sawicki $f(R)$ gravity simulations with full baryonic physics.

Findings

The retuned model matches observed galaxy and cluster properties despite lower resolution.

The large simulation sample includes about 500 galaxy clusters and 8000 groups.

Scaling relation mappings between $f(R)$ and $ m extLambda$CDM are accurate within a few percent.

Abstract

We present a retuning of the IllustrisTNG baryonic physics model which can be used to run large-box realistic cosmological simulations with a lower resolution. This new model employs a lowered gas density threshold for star formation and reduced energy releases by stellar and black hole feedback. These changes ensure that our simulations can produce sufficient star formation to closely match the observed stellar and gas properties of galaxies and galaxy clusters, despite having $\sim$160 times lower mass resolution than the simulations used to tune the fiducial IllustrisTNG model. Using the retuned model, we have simulated Hu-Sawicki $f(R)$ gravity within a $301.75h^{-1}{\rm Mpc}$ box. This is, to date, the largest simulation that incorporates both screened modified gravity and full baryonic physics, offering a large sample ($\sim$500) of galaxy clusters and $\sim$8000 galaxy groups. We have reanalysed the effects of the $f(R)$ fifth force on the scaling relations between the cluster mass and four observable proxies: the mass-weighted gas temperature, the Compton $Y$-parameter of the thermal Sunyaev-Zel'dovich effect, the X-ray analogue of the $Y$-parameter, and the X-ray luminosity. We show that a set of mappings between the $f(R)$ scaling relations and their $\Lambda$CDM counterpart, which have been tested in a previous work using a much smaller cosmological volume, are accurate to within a few percent for the $Y$-parameters and $\lesssim7\%$ for the gas temperature for cluster-sized haloes ($10^{14}M_{\odot}\lesssim M_{500}\lesssim10^{15}M_{\odot}$). These mappings will be important for unbiased constraints of gravity using the data from ongoing and upcoming cluster surveys.