GenASiS: General Astrophysical Simulation System. II. Self-gravitating Baryonic Matter

TL;DR

This paper extends GenASiS to include Newtonian self-gravity and an updated, GPU-accelerated fluid dynamics solver on a single-level spherical mesh, complemented by a multipole Poisson solver and tabulated baryonic equations of state. It validates the methods across five tests, including analytic spheroid potentials and self-similar collapse scenarios, and demonstrates robust, high-resolution results for adiabatic core-collapse models of eleven pre-supernova progenitors, revealing non-monotonic dependencies of shock speed and kinetic energy on stellar mass and compactness. The study reports substantial GPU speedups and argues for adiabatic core-collapse, with its associated explosion benchmark, as a standard cross-code test for the community, alongside a public GenASiS_II_Dataset. Together, these results establish a rigorous, accelerator-friendly framework for self-gravitating baryonic simulations relevant to core-collapse physics and beyond.

Abstract

GenASiS (General Astrophysical Simulation System) is a code being developed initially and primarily, though not exclusively, for the simulation of core-collapse supernovae on the world's leading capability supercomputers. This paper -- the second in a series -- documents capabilities for Newtonian self-gravitating fluid dynamics, including tabulated microphysical equations of state treating nuclei and nuclear matter (`baryonic matter'). Computation of the gravitational potential of a spheroid, and simulation of the gravitational collapse of dust and of an ideal fluid, provide tests of self-gravitation against known solutions. In multidimensional computations of the adiabatic collapse, bounce, and explosion of spherically symmetric pre-supernova progenitors -- which we propose become a standard benchmark for code comparisons -- we find that the explosions are prompt and remain spherically symmetric (as expected), with an average shock expansion speed and total kinetic energy that are inversely correlated with the progenitor mass at the onset of collapse and the compactness parameter.

Figures (20)

• Figure 1: An example 2D spherical coordinate mesh, in coordinate space (left) and physical space (right), showing the full mesh extent (top) and a region closer to $r = 0$ (bottom). For $r > r_\mathrm{core} = 1.25$ the radial cell width $\Delta r \propto r$, yielding a constant polar/radial cell aspect ratio $r \, \Delta \theta / \Delta r$ (here $\approx 2.45$). For $r < r_\mathrm{core}$ the cell radial width $\Delta r_\mathrm{min}$ is uniform and the polar/radial aspect ratio rapidly decreases with decreasing $r$. Approaching the origin, neighboring cells at a given radius with $r \, \Delta \theta < \Delta r _\mathrm{core}$ are grouped into polar-angle 'coarsening blocks' (randomly colored) consisting of $2, 4, 8, \dots$ cells as needed until the block width exceeds $\Delta r _\mathrm{core}$. Averaging over these blocks to suppress small-wavelength perturbations allows explicit time steps to be limited only by $\Delta r _\mathrm{core}$.
• Figure 2: An example 3D spherical coordinate mesh, in coordinate space (top) and physical space (bottom), showing the full mesh extent (left) and a region closer to $r = 0$ (right), with the $r = 0.25$ plane exposed in coordinate space (upper right). The coarsening blocks (randomly colored) now appear along the polar axis as well as near the origin and are now two-dimensional, with the block size in each angular dimension determined by comparing $r \, \sin \theta \, \Delta \phi$ and $r \, \Delta \theta$ with $\Delta r _\mathrm{core}$.
• Figure 3: An example field, sinusoidal in polar (2D, 3D) and azimuthal (3D) angles and displayed in coordinate space, has been coarsened by averaging over the blocks displayed in Figures \ref{['Fig:CoarseningBlocks_2D']} and \ref{['Fig:CoarseningBlocks_3D']}.
• Figure 4: Three-slice of the computed gravitational potential (left) and its relative error (right) for a three-dimensional oblate spheroid with $e = 0.9$. For both plots, the shape of the spheroid is outlined in black.
• Figure 5: $L_1$ error of the potential as a function of resolution (left) computed with $L=20$ and number of multipole terms $L$ (right) computed with $N_\theta = 384$ for spheroids with different eccentricities. The solid and dashed black lines are references for first- and second-order convergence respectively.