Orbiting binary black hole evolutions with a multipatch high order finite-difference approach

TL;DR

This paper introduces a high-order multipatch finite-difference method for simulating orbiting binary black holes, demonstrating excellent scalability and potential for complex astrophysical systems involving matter flow.

Contribution

The paper develops and validates a multipatch high-order finite-difference approach for black hole simulations, offering an alternative to adaptive mesh refinement with improved scalability.

Findings

Simulated twelve orbits of binary black holes with high accuracy.

Achieved near-perfect scaling on thousands of processors.

Established the method's potential for future neutron star and black hole-neutron star simulations.

Abstract

We present numerical simulations of orbiting black holes for around twelve cycles, using a high-order multipatch approach. Unlike some other approaches, the computational speed scales almost perfectly for thousands of processors. Multipatch methods are an alternative to AMR (adaptive mesh refinement), with benefits of simplicity and better scaling for improving the resolution in the wave zone. The results presented here pave the way for multipatch evolutions of black hole-neutron star and neutron star-neutron star binaries, where high resolution grids are needed to resolve details of the matter flow.