Viscous hydrodynamic evolution of neutron star merger accretion disks: a code comparison

TL;DR

This study compares two viscous hydrodynamic simulation codes for neutron star merger accretion disks, assessing their agreement and impact on ejecta, nucleosynthesis, and kilonova predictions.

Contribution

It provides a detailed comparison of ALCAR and FLASH codes, highlighting their similarities, differences, and implications for modeling neutron star merger disks.

Findings

Good agreement (~10%) in most quantities between codes.

Outflow velocity varies with treatment of nuclear binding energy.

Code differences influence nucleosynthesis yields and kilonova light curves.

Abstract

The accretion disk formed after a neutron star merger is an important contributor to the total ejecta from the merger, and hence to the kilonova and the $r$-process yields of each event. Axisymmetric viscous hydrodynamic simulations of these disks can capture thermal mass ejection due to neutrino absorption and in the advective phase -- after neutrino cooling has subsided -- and are thus likely to provide a lower-limit to the total disk ejecta relative to MHD evolution. Here we present a comparison between two viscous hydrodynamic codes that have been used extensively on this problem over the past decade: ALCAR and FLASH. We choose a representative setup with a black hole at the center, and vary the treatment of viscosity and neutrino transport. We find good overall agreement ($\sim 10\%$ level) in most quantities. The average outflow velocity is sensitive to the treatment of the nuclear binding energy of heavy nuclei, showing a larger variation than other quantities. We post-process trajectories from both codes with the same nuclear network, and explore the effects of code differences on nucleosynthesis yields, heating rates, and kilonova light curves. For the latter, we also assess the effect of varying the number of tracer particles in reconstructing the spatial abundance distribution for kilonova light curve production.