General Relativistic Simulations of the Quasicircular Inspiral and Merger of Charged Black Holes: GW150914 and Fundamental Physics Implications

We perform general-relativistic simulations of charged black holes targeting GW150914. We show that the inspiral is most efficient for detecting black hole charge through gravitational waves and that GW150914 is compatible with having charge-to-mass ratio as high as 0.3. Our work applies to electric and magnetic charge and to theories with black holes endowed with U(1) (hidden or dark) charges. Using our results, we place an upper bound on the deviation from general relativity in the dynamical strong-filed regime of Moffat’s modified gravity.

Figure 1

Mismatch between the strains from uncharged black holes and from charged ones with chirp mass rescaled by $\mathcal{M}/{\mathcal{M}}_{00}$. Solid curves are the mismatch including all available frequencies (the entire signal), dashed ones are only restricting to the frequency range (23,55) Hz (inspiral) and dotted ones have frequencies restricted to (55,1024) Hz (merger and ringdown). The solid horizontal line is the detection threshold for GW150914 ($8\times {10}^{-4}$). The top (bottom) row is the largest (smallest) value of $\lambda$ (in our survey) compatible (incompatible) with GW150914. Red curves (left column) are for the simulation with ${\lambda }_{+}^{+}=0.3$ (top) and ${\lambda }_{+}^{+}=0.4$ (bottom), blue (central column) for ${\lambda }_{-}^{+}=0.1$ and ${\lambda }_{-}^{+}=0.2$, and green (middle column) for ${\lambda }_0^{+}=0.3$ and ${\lambda }_0^{+}=0.35$. The error bars shown are estimated comparing the standard resolution simulation with the one at higher resolution. We report the error bar only at minimum mismatch, but each point along the curve has the same level of error.

Figure 2

Comparison between the (2,2) mode of the detector-response strain for Hanford for the simulations with no charges (dashed curves) and the ones with it, but with chirp mass $\mathcal{M}$ rescaled with respect to ${\mathcal{M}}_{00}=28.095M_{\odot }$. Time and phase shifts are applied to minimize the mismatch between the two signals. All of these waveforms have two-detectors mismatch larger than the detection threshold for GW150914 of $8\times {10}^{-4}$, mostly coming from the inspiral phase.