We provide a detailed description of the Chimera code, a code developed to model core collapse supernovae in multiple spatial dimensions. The core collapse supernova explosion mechanism remains the subject of intense research. Progress to date demonstrates that it involves a complex interplay of neutrino production, transport, and interaction in the stellar core, three-dimensional stellar core fluid dynamics and its associated instabilities, nuclear burning, and the foundational physics of the neutrino-stellar core weak interactions and the equations of state of all stellar core constituents -particularly, the nuclear equation of state associated with nucleons, both free and bound in nuclei. Chimera, by incorporating detailed neutrino transport, realistic neutrino-matter interactions, three-dimensional hydrodynamics, realistic nuclear, leptonic, and photonic equations of state, and a nuclear reaction network, along with other refinements, can be used to study the role of neutrino radiation, hydrodynamic instabilities, and a variety of input physics in the explosion mechanism itself. It can also be used to compute observables such as neutrino signatures, gravitational radiation, and the products of nucleosynthesis associated with core collapse supernovae. The code contains modules for neutrino transport, multidimensional compressible hydrodynamics, nuclear reactions, a variety of neutrino interactions, equations of state, and modules to provide data for post-processing observables such as the products of nucleosynthesis, and gravitational radiation. Chimera is an evolving code, being updated periodically with improved input physics and numerical refinements. We detail here the current version of the code, from which future improvements will stem, which can in turn be described as needed in future publications.

We use OpenMP to target hardware accelerators (GPUs) on Summit, a newly deployed supercomputer at the Oak Ridge Leadership Computing Facility (OLCF), demonstrating simplified access to GPU devices for users of our astrophysics code GenASiS and useful speedup on a sample fluid dynamics problem. We modify our workhorse class for data storage to include members and methods that significantly streamline the persistent allocation of and association to GPU memory. Users offload computational kernels with OpenMP target directives that are rather similar to constructs already familiar from multi-core parallelization. In this initial example we ask, "With a given number of Summit nodes, how fast can we compute with and without GPUs?", and find total wall time speedups of $\sim 12\mathrm{X}$. We also find reasonable weak scaling up to 8000 GPUs (1334 Summit nodes). We make available the source code from this work at https://github.com/GenASiS/GenASiS_Basics.

GenASiS Basics provides Fortran 2003 classes furnishing extensible object-oriented utilitarian functionality for large-scale physics simulations on distributed memory supercomputers. This functionality includes physical units and constants; display to the screen or standard output device; message passing; I/O to disk; and runtime parameter management and usage statistics. This revision---Version 3 of Basics---includes a significant name change, some minor additions to functionality, and a major addition to functionality: infrastructure facilitating the offloading of computational kernels to devices such as GPUs.

The large-scale computer simulation of a system of physical fields governed by partial differential equations requires some means of approximating the mathematical limit of continuity. For example, conservation laws are often treated with a `finite-volume' approach in which space is partitioned into a large number of small `cells,' with fluxes through cell faces providing an intuitive discretization modeled on the mathematical definition of the divergence operator. Here we describe and make available Fortran 2003 classes furnishing extensible object-oriented implementations of simple meshes and the evolution of generic conserved currents thereon, along with individual `unit test' programs and larger example problems demonstrating their use. These classes inaugurate the Mathematics division of our developing astrophysics simulation code GenASiS (General Astrophysical Simulation System), which will be expanded over time to include additional meshing options, mathematical operations, solver types, and solver variations appropriate for many multiphysics applications.

We present an initial report on 160 simulations of a highly simplified model of the post-bounce core-collapse supernova environment in three spatial dimensions (3D). We set different values of a parameter characterizing the impact of nuclear dissociation at the stalled shock in order to regulate the post-shock fluid velocity, thereby determining the relative importance of convection and the stationary accretion shock instability (SASI). While our convection-dominated runs comport with the paradigmatic notion of a `critical neutrino luminosity' for explosion at a given mass accretion rate (albeit with a nontrivial spread in explosion times just above threshold), the outcomes of our SASI-dominated runs are much more stochastic: a sharp threshold critical luminosity is `smeared out' into a rising probability of explosion over a $\sim 20\%$ range of luminosity. We also find that the SASI-dominated models are able to explode with 3 to 4 times less efficient neutrino heating, indicating that progenitor properties, and fluid and neutrino microphysics, conducive to the SASI would make the neutrino-driven explosion mechanism more robust.

Core-collapse supernovae are among the most powerful explosions in the Universe, releasing about $10^{53}~\mbox{erg}$ of energy on timescales of a few tens of seconds. These explosion events are also responsible for the production and dissemination of most of the heavy elements, making life as we know it possible. Yet exactly how they work is still unresolved. One reason for this is the sheer complexity and cost of a self-consistent, multi-physics, and multi-dimensional core-collapse supernova simulation, which is impractical, and often impossible, even on the largest supercomputers we have available today. To advance our understanding we instead must often use simplified models, teasing out the most important ingredients for successful explosions, while helping us to interpret results from higher fidelity multi-physics models. In this paper we investigate the role of instabilities in the core-collapse supernova environment. We present here simulation and visualization results produced by our code GenASiS.

GenASiS (General Astrophysical Simulation System) is a new code being developed initially and primarily, though by no means exclusively, for the simulation of core-collapse supernovae on the world's leading capability supercomputers. This paper---the first in a series---demonstrates a centrally refined coordinate patch suitable for gravitational collapse and documents methods for compressible nonrelativistic hydrodynamics. We benchmark the hydrodynamics capabilities of GenASiS against many standard test problems; the results illustrate the basic competence of our implementation, demonstrate the strengths and limitations of the HLLC relative to the HLL Riemann solver in a number of interesting cases, and provide preliminary indications of the code's ability to scale and to function with cell-by-cell fixed-mesh refinement.

In this paper, the second in a series, we document the algorithms and solvers for compressible nonrelativistic hydrodynamics implemented in GenASiS (General Astrophysical Simulation System)---a new code being developed initially and primarily, though by no means exclusively, for the simulation of core-collapse supernovae. In the Mathematics division of GenASiS we introduce Solvers, which includes finite-volume updates for generic hyperbolic BalanceEquations and ordinary differential equation integration Steps. We also introduce the Physics division of GenASiS; this extends the Manifolds division of Mathematics into physical Spaces, defines StressEnergies, and combines these into Universes. We benchmark the hydrodynamics capabilities of GenASiS against many standard test problems; the results illustrate the basic competence of our implementation, demonstrate the manifest superiority of the HLLC over the HLL Riemann solver in a number of interesting cases, and provide preliminary indications of the code's ability to scale and to function with cell-by-cell fixed-mesh refinement.

The stationary accretion shock instability (SASI) plays a central role in modern simulations of the explosion phase of core-collapse supernovae (CCSNe). It may be key to realizing neutrino powered explosions, and possibly links birth properties of pulsars (e.g., kick, spin, and magnetic field) to supernova dynamics. Using high-resolution magnetohydrodynamic simulations, we study the development of turbulence, and subsequent amplification of magnetic fields in a simplified model of the post-bounce core-collapse supernova environment. Turbulence develops from secondary instabilities induced by the SASI. Our simulations suggest that the development of turbulence plays an important role for the subsequent evolution of the SASI. The turbulence also acts to amplify weak magnetic fields via a small-scale dynamo.

We describe the initial implementation of magnetohydrodynamics (MHD) in our astrophysical simulation code \genasis. Then, we present MHD simulations exploring the capacity of the stationary accretion shock instability (SASI) to generate magnetic fields by adding a weak magnetic field to an initially spherically symmetric fluid configuration that models a stalled shock in the post-bounce supernova environment. Upon perturbation and nonlinear SASI development, shear flows associated with the spiral SASI mode contributes to a widespread and turbulent field amplification mechanism. While the SASI may contribute to neutron star magnetization, these simulations do not show qualitatively new features in the global evolution of the shock as a result of SASI-induced magnetic field amplification.

We extend our investigation of magnetic field evolution in three-dimensional flows driven by the stationary accretion shock instability (SASI) with a suite of higher-resolution idealized models of the post-bounce core-collapse supernova environment. Our magnetohydrodynamic simulations vary in initial magnetic field strength, rotation rate, and grid resolution. Vigorous SASI-driven turbulence inside the shock amplifies magnetic fields exponentially; but while the amplified fields reduce the kinetic energy of small-scale flows, they do not seem to affect the global shock dynamics. The growth rate and final magnitude of the magnetic energy are very sensitive to grid resolution, and both are underestimated by the simulations. Nevertheless our simulations suggest that neutron star magnetic fields exceeding $10^{14}$ G can result from dynamics driven by the SASI, \emph{even for non-rotating progenitors}.

We show that, even for extremely stiff systems, explicit integration may compete in both accuracy and speed with implicit methods if algebraic methods are used to stabilize the numerical integration. The required stabilizing algebra depends on whether the system is well-removed from equilibrium or near equilibrium. This paper introduces a quantitative distinction between these two regimes and addresses the former case in depth, presenting explicit asymptotic methods appropriate when the system is extremely stiff but only weakly equilibrated. A second paper examines quasi-steady-state methods as an alternative to asymptotic methods in systems well away from equilibrium and a third paper extends these methods to equilibrium conditions in extremely stiff systems using partial equilibrium methods. All three papers present systematic evidence for timesteps competitive with implicit methods. Because an explicit method can execute a timestep faster than an implicit method, algebraically-stabilized explicit algorithms might permit integration of larger networks than have been feasible before in various disciplines.

We begin an exploration of the capacity of the stationary accretion shock instability (SASI) to generate magnetic fields by adding a weak, stationary, and radial (but bipolar) magnetic field, and in some cases rotation, to an initially spherically symmetric fluid configuration that models a stalled shock in the post-bounce supernova environment. In axisymmetric simulations we find that cycles of latitudinal flows into and radial flows out of the polar regions amplify the field parallel to the symmetry axis, typically increasing the total magnetic energy by about two orders of magnitude. Nonaxisymmetric calculations result in fundamentally different flows and a larger magnetic energy increase: shearing associated with the SASI spiral mode contributes to a widespread and turbulent field amplification mechanism, boosting the magnetic energy by almost four orders of magnitude (a result which remains very sensitive to the spatial resolution of the numerical simulations). While the SASI may contribute to neutron star magnetization, these simulations do not show qualitatively new features in the global evolution of the shock as a result of SASI-induced magnetic field amplification.

By adding a weak magnetic field to a spherically symmetric fluid configuration that caricatures a stalled shock in the post-bounce supernova environment, we explore the capacity of the stationary accretion shock instability (SASI) to generate magnetic fields. The SASI develops upon perturbation of the initial condition, and the ensuing flow generates--{\em in the absence of rotation}--dynamically significant magnetic fields ($\sim 10^{15}$ G) on a time scale that is relevant for the explosion mechanism of core-collapse supernovae. We describe our model, present some recent results, and discuss their potential relevance for supernova models.