Movies of a 3D MHD simulation of stellar core collapse based on the isotropic diffusion source approximation (IDSA) for spectral neutrino transport
After several improvements of our code ELEPHANT (ELEgant Parallel Hydrodynamics with Approximate Neutrino Transport), we are posting here some movies from the 3D simulation of the core collapse and explosion of a 15 solar mass star (s15s7b2, Woosley & Weaver 1995). Slow rotation and a mainly toroidal magnetic field was imposed on the originally non-rotating progenitor star according to the estimates in (Heger, Woosley & Spruit 2005). The run is based on the FISH hydrodynamics (Käppeli et al., ApJS 195, 20, 2011) and the IDSA for neutrino transport (Liebendörfer, Whitehouse & Fischer, ApJ 698, 1174). Some GR effects are taken into account by an effective potential in an otherwise Newtonian simulation (Marek et al. 2006) and the energy loss by the emission of heavy neutrinos is treated by a leakage scheme implemented by A. Perego. The three-dimensional simulation has a resolution of 2 km with 432 zones in each of the three Cartesian directions, and was carried out at CSCS, the Swiss National Supercomputing Center. Later code versions and simulations at 1 km resolutino are documented in (R. Cabezon et al., A&A 619A.118C, 2018).
The movies illustrate the evolution of the velocity, density, entropy, and magnetic field in the equatorial xy-plane. Overlayed is a plot in black lines showing the emission of electron neutrinos (solid line) and electron antineutrinos (dashed line) as a function of time. The little triangles move with the fluid and point always into the direction of the local velocity. The movies show either entropy and electron fraction or entropy and magnetic field. The magnetic field grows in the postbounce phase. Spherically integrated profiles of the velocity, density, electron fraction and entropy are compared to the corresponding quantities in a spherically symmetric general relativistic simulation with three-flavour Boltzmann neutrino transport (Model G15 in Liebendörfer et al. 2005). The two simulations agree well as long as multi-dimensional effects play a little role and start to diverge when the convection develops in the 3D model.