In massive gravity the so-far-found black hole solutions on Minkowski space happen to convert horizons into a certain type of singularities. Here we explore whether these singularities can be avoided if space-time is not asymptotically Minkowskian. We find an exact analytic black hole (BH) solution which evades the above problem by a transition at large scales to self-induced de Sitter (dS) space-time, with the curvature scale set by the graviton mass. This solution is similar to the ones discovered by Koyama, Niz and Tasinato, and by Nieuwenhuizen, but differs in detail. The solution demonstrates that in massive GR, in the Schwarzschild coordinate system, a BH metric has to be accompanied by the Stückelberg fields with nontrivial backgrounds to prevent the horizons to convert into the singularities. We also find an analogous solution for a Reissner-Nordström BH on dS space. A limitation of our approach, is that we find the solutions only for specific values of the two free parameters of the theory, for which both the vector and scalar fluctuations loose their kinetic terms, however, we hope our solutions represent a broader class with better behaved perturbations.

We study Galileon theories that emerge in ghost-free massive gravity. In particular, we focus on a sub-class of these theories where the Galileons can be completely decoupled from the tensor Lagrangian. These Galileons differ from generic ones -- they have interrelated coefficients of the cubic and quartic terms, and most importantly, a non-standard coupling to external stress-tensors, governed by the same coefficient. We show that this theory has no static stable spherically symmetric solutions that would interpolate from the Vainshtein region to flat space; these two regions cannot be smoothly matched for the sign of the coefficient for which fluctuations are stable. Instead, for this sign choice, a solution in the Vainshtein domain is matched onto a cosmological background. Small fluctuations above this solution are stable, and sub-luminal. We discuss observational constraints on this theory, within the quantum effective Lagrangian approach, and argue that having a graviton mass of the order of the present-day Hubble parameter, is consistent with the data. Last but not least, we also present a general class of cosmological solutions in this theory, some of which exhibit the de-mixing phenomenon, previously found for the self-accelerated solution.Comment: 17 pages; Typos correcte

It was previously found that in a certain parameter subspace of scalartensor theories emerging from massive gravity, the only stable field configuration created by static spherically symmetric sources was one with cosmological asymptotics. Moreover, these backgrounds were shown to be sub-luminal everywhere in the space; in contrast to the common believe that these theories are necessarily superluminal in the vicinity of a static source. In this work we complete that analysis by extending it to cover the whole parameter space of these scalar-tensor theories. We find that the stability argument renders the asymptotically flat backgrounds unrealizable, forcing once again for cosmological asymptotics. In the case of pressureless sources these backgrounds are stable. However, they get destabilized in the presence of positive pressure, larger than a critical density. Even on the self-accelerated background, on which the scalar mode decouples from sources, in the region occupied by the source it acquires an elliptic equation of motion. Therefore, we conclude that the only parameter space which is not ruled out, by solar system measurements, is the one considered in Berezhiani et al. (arXiv:1302.0549), namely the one for which the scalar and tensor modes can be diagonalized via local transformations.We also reinvestigate the scale at which perturbation theory breaks down in a general Galileon theory. We show that the Vainshtein mechanism successfully redresses the strong coupling scale to a small one, just like in the cubic Galileon, despite the cancellations occurring in the special spherically symmetric case. We emphasize that even if these tests were performed at scales at which perturbation theory broke down, these could not be interpreted as a lower bound for the graviton mass.

No abstract