Magnetorotational Core-Collapse Supernovae in Three Dimensions

Philipp Mösta, Sherwood Richers, Christian D. Ott,
Roland Haas, Anthony L. Piro, Kristen Boydstun, Ernazar Abdikamalov, Christian Reisswig, and Erik Schnetter
ApJ, 785, L29 , arXiv:1403.1230
 

Abstract

We present results of new three-dimensional (3D) general-relativistic magnetohydrodynamic simulations of rapidly rotating strongly magnetized core collapse. These simulations are the first of their kind and include a microphysical finite-temperature equation of state and a leakage scheme that captures the overall energetics and lepton number exchange due to postbounce neutrino emission. Our results show that the 3D dynamics of magnetorotational core-collapse supernovae are fundamentally different from what was anticipated on the basis of previous simulations in axisymmetry (2D). A strong bipolar jet that develops in a simulation constrained to 2D is crippled by a spiral instability and fizzles in full 3D. While multiple (magneto-)hydrodynamic instabilities may be present, our analysis suggests that the jet is disrupted by an m = 1 kink instability of the extremely strong toroidal field near the protoneutron star. Instead of an axially symmetric jet, a completely new, previously unreported flow structure develops. Highly magnetized spiral plasma funnels expelled from the core push out the shock in polar regions, creating wide secularly expanding polar lobes. No runaway explosion obtains by the end of the full 3D simulation at ∼185 ms after bounce. At this time, the lobes have reached maximum radii of ∼900km.

Simulation Source Code

The simulations were carried out with the Zelmani core collapse simulation package, which is based on the open-source Einstein Toolkit, which, in turn, uses the Cactus Computational Toolkit and the Carpet adaptive mesh refinement driver. While not all components of Zelmani are open source yet, we can already provide as open source the leakage/heating module used in this study, ZelmaniLeak. Other components of the code that have modifications that are not yet publicly available as part of the Einstein Toolkit are provided below as well. These should all build smoothly in the latest regular Einstein Toolkit build.

Please contact Philipp Moesta (pmoesta # tapir.caltech.edu) should you have any questions.

The parameters used for the electron fraction parameterization used in the collapse phase are as follows:

# ye of rho parameters
ZelmaniLeak::yeofrho_rho1 = 1.0e7
ZelmaniLeak::yeofrho_rho2 = 1.0e13
ZelmaniLeak::yeofrho_ye1 = 0.5e0
ZelmaniLeak::yeofrho_ye2 = 0.29e0
ZelmaniLeak::yeofrho_yec = 0.035e0
ZelmaniLeak::yeofrho_do_high_correction = yes
ZelmaniLeak::yeofrho_high_correction_ye = 0.2717e0
ZelmaniLeak::yeofrho_high_correction_rho = 2.55e14

Computational Grid Structure

The grid structure of the simulations initially consists of four levels of nested adaptive mesh-refinement (AMR) boxes. During collapse we add 5 more AMR boxes as the central density of the inner core increases. The grid structure at core bounce consists of 9 levels of AMR boxes with half extends of [5674.0 km, 3026.1km, 2435.1, 1560.3, 1283.7, 212.8, 144.8, 59.1, 17.7 km]. The coarsest resolution is h = 94.6 km and refined meshes differ in resolution by factors of 2. We use adaptive shock tracking to ensure that the shocked region is always covered by the AMR box with resolution h = 1.48 km.

The parameters used for adding additional AMR boxed during the collapse phase of the simulations are:

corecollapsecontrol::rho_max_list[0] = 8.0e10
corecollapsecontrol::rho_max_list[1] = 3.2e11
corecollapsecontrol::rho_max_list[2] = 1.28e12
corecollapsecontrol::rho_max_list[3] = 5.12e12
corecollapsecontrol::rho_max_list[4] = 2.048e13