Binary black holes in circular orbits. II. Numerical methods and first results

TL;DR

This paper introduces a new numerical method for modeling corotating binary black holes in circular orbits, achieving results consistent with post-Newtonian calculations and identifying the innermost stable circular orbit.

Contribution

The authors develop a novel approach solving a set of five Einstein equations simultaneously using spectral methods, improving accuracy over previous models.

Findings

Constant apparent horizon area along the sequence

Identification of a turning point indicating the ISCO

Better agreement with post-Newtonian results

Abstract

We present the first results from a new method for computing spacetimes representing corotating binary black holes in circular orbits. The method is based on the assumption of exact equilibrium. It uses the standard 3+1 decomposition of Einstein equations and conformal flatness approximation for the 3-metric. Contrary to previous numerical approaches to this problem, we do not solve only the constraint equations but rather a set of five equations for the lapse function, the conformal factor and the shift vector. The orbital velocity is unambiguously determined by imposing that, at infinity, the metric behaves like the Schwarzschild one, a requirement which is equivalent to the virial theorem. The numerical scheme has been implemented using multi-domain spectral methods and passed numerous tests. A sequence of corotating black holes of equal mass is calculated. Defining the sequence by requiring that the ADM mass decrease is equal to the angular momentum decrease multiplied by the orbital angular velocity, it is found that the area of the apparent horizons is constant along the sequence. We also find a turning point in the ADM mass and angular momentum curves, which may be interpreted as an innermost stable circular orbit (ISCO). The values of the global quantities at the ISCO, especially the orbital velocity, are in much better agreement with those from third post-Newtonian calculations than with those resulting from previous numerical approaches.