Coupling a DG-IMEX method for two-moment neutrino transport with FLASH

To simulate neutrino transport in core-collapse supernova (CCSN) explosions, we develop numerical methods for a spectral two-moment model. In the two-moment model, the evolved variables are the neutrino number density and flux — angular moments of the neutrino distribution function. Our numerical scheme is based on the high-order discontinuous Galerkin (DG) method for phase space discretization, implicit-explicit (IMEX) time stepping, and the two-moment model is closed with the maximum entropy procedure of Cernohorsky & Bludman (1994). Importantly, the scheme is designed to remain consistent with Fermi-Dirac statistics. The IMEX method treats streaming terms explicitly and the collision terms, which model the interaction between neutrinos and matter, implicitly. It is second-order accurate in the streaming limit and works well in the diffusion limit, characterized by frequent collisions and long timescales. Neutrino-matter interactions and a nuclear equation of state are tabulated and provided by the WeakLib library, which currently includes emission, absorption and scattering on nucleons and nuclei. The DG-IMEX scheme has been implemented in the Toolkit for High-Order Neutrino-Radiation Hydrodynamics (thornado), and an interface to the adaptive mesh refinement code FLASH has been created. As an initial step, we run one-dimensional stellar core collapse simulations, including electron-type neutrinos and antineutrinos. Preliminary numerical results are presented and future plans are discussed.