2018MNRAS.480.5211N

Current constraints on gravity are relatively weak on galactic and intergalactic scales. Screened modified gravity models can exhibit complex behaviour there without violating stringent tests of gravity within our Solar System. They might hence provide viable extensions of the theory of gravity. Here, we use galaxy kinematics to constrain screened modified gravity models. We focus on chameleon f(R) gravity and predict its impact on galaxy rotation curves and radial acceleration relations. This is achieved by post-processing state-of-the-art galaxy formation simulations from the AURIGA PROJECT, using the MG-GADGET code. For a given galaxy, the surface dividing screened and un-screened regions adopts an oblate shape, reflecting the disc morphology of the galaxy's mass distribution. At the 'screening radius' - the radius at which screening is triggered in the disc plane - characteristic 'upturns' are present in both rotation curves and radial acceleration relations. The locations of these features depend on various factors, such as the galaxy mass, the concentration of the density profile, and the value of the background field amplitude {bar}f_R0. Self-screening of stars and environmental screening also play a role. For Milky Way-size galaxies, we find that a model with |{bar}f_R0|=10^–7 results in rotation curves that are indistinguishable from ΛCDM, while for |{bar}f_R0| ≥2 ×10^–6 the simulated galaxies are entirely unscreened, violating Solar System constraints. For intermediate values, distinct upturns are present. With a careful statistical analysis of existing samples of observed rotation curves, including lower mass objects, constraints on f(R) gravity with a sensitivity down to |{bar}f_R0|∼10^–7 should be possible.