Modeling Equal and Unequal Mass Binary Neutron Star Mergers Using Public Codes

TL;DR

This paper presents open-source simulations of binary neutron star mergers, exploring various mass ratios and using different numerical methods, providing a reproducible framework for future research in neutron star merger dynamics.

Contribution

First comprehensive simulation study of BNS mergers using only publicly available software, with detailed comparison of numerical methods and open data sharing.

Findings

High-resolution agreement between methods

Superiority of BSSN-NOK with WENO at lower resolutions

Simulations cover a range of mass ratios and post-merger outcomes

Abstract

We present three-dimensional simulations of the dynamics of binary neutron star (BNS) mergers from the late stage of the inspiral process up to $\sim 20$ ms after the system has merged, either to form a hyper-massive neutron star (NS) or a rotating black hole (BH). We investigate five equal-mass models of total gravitational mass $2.207$, $2.373$, $2.537$, $2.697$ and $2.854 M_\odot$, respectively, and four unequal mass models with $M_{\mathrm{ADM}}\simeq 2.53\ M_\odot$ and $q\simeq 0.94$, $0.88$, $0.82$, and $0.77$ (where $q = M^{(1)}/M^{(2)}$ is the mass ratio). We use a semi-realistic equation of state (EOS) namely, the seven-segment piece-wise polytropic SLyPP with a thermal component given by $\Gamma_{th} = 1.8$. We have also compared the resulting dynamics (for one model) using both, the BSSN-NOK and CCZ4 methods for the evolution of the gravitational sector, and also different reconstruction methods for the matter sector, namely PPM, WENO and MP5. Our results show agreement and high resolution, but superiority of BSSN-NOK supplemented by WENO reconstruction at lower resolutions. One of the important characteristics of the present investigation is that, for the first time, this has been done using only publicly available open source software, in particular, the Einstein Toolkit code deployed for the dynamical evolution and the LORENE code for the generation of the initial models. All of the source code and parameters used for the simulations have been made publicly available. This not only makes it possible to re-run and re-analyze our data; it also enables others to directly build upon this work for future research.