Barred spiral galaxies in modified gravity theories

TL;DR

This paper compares galaxy bar dynamics in standard dark matter models and extended gravity theories, finding that extended gravity models naturally produce faster bars consistent with observations, unlike dark matter simulations.

Contribution

It demonstrates that extended gravity theories predict faster, weaker bars and resolve the tension with observed bar pattern speeds, unlike standard dark matter models.

Findings

Extended gravity models predict weaker, faster bars.

Dark matter simulations show slower bars due to dynamical friction.

Extended gravity models align better with observed fast bar pattern speeds.

Abstract

When bars form within galaxy formation simulations in the standard cosmological context, dynamical friction with dark matter (DM) causes them to rotate rather slowly. However, almost all observed galactic bars are fast in terms of the ratio between corotation radius and bar length. Here, we explicitly display an $8\sigma$ tension between the observed distribution of this ratio and that in the EAGLE simulation at redshift 0. We also compare the evolution of Newtonian galactic discs embedded in DM haloes to their evolution in three extended gravity theories: Milgromian Dynamics (MOND), a model of non-local gravity, and a scalar-tensor-vector gravity theory (MOG). Although our models start with the same initial baryonic distribution and rotation curve, the long-term evolution is different. The bar instability happens more violently in MOND compared to the other models. There are some common features between the extended gravity models, in particular the negligible role played by dynamical friction $-$ which plays a key role in the DM model. Partly for this reason, all extended gravity models predict weaker bars and faster bar pattern speeds compared to the DM case. Although the absence of strong bars in our idealized, isolated extended gravity simulations is in tension with observations, they reproduce the strong observational preference for `fast' bar pattern speeds, which we could not do with DM. We confirm previous findings that apparently `ultrafast' bars can be due to bar-spiral arm alignment leading to an overestimated bar length, especially in extended gravity scenarios where the bar is already fast.