A binary black hole metric approximation from inspiral to merger

We present a semi-analytic binary black hole (BBH) metric approximation that models the entire evolution of the system from inspiral to merger. The metric is constructed as a boosted Kerr-Schild superposition following post-Newtonian (PN) trajectories at the fourth PN order in the inspiral phase. During merger, we interpolate the binary metric in time to a single black hole remnant with properties obtained from numerical relativity (NR) fittings. Different from previous approaches, the new metric can model binary black holes with arbitrary spin direction, mass ratio, and eccentricity at any stage of their evolution in a fast and computationally efficient way. We analyze the properties of our new metric and we compare it with a full numerical relativity evolution. We show that Hamiltonian constraints are well-behaved even at merger and that the mass and spin of the black holes deviate in average only ∼ 5 % compared to the full numerical evolution. We perform a General Relativistic magneto-hydrodynamical (GRMHD) simulation of uniform gas evolving on top of our approximate metric. We compare it with a full numerical relativity evolution of the fluid and Einstein's equations. We show that the properties of the gas such as the accretion rate are remarkably similar between the two approaches, exhibiting only ∼ 10 % differences in average. The approximate metric is five times more efficient among other computational advantages. The numerical implementation of the metric is now open-source and optimized for numerical work. We have also implemented this spacetime in the widely used GRMHD codes Athena++ and EinsteinToolkit.