Guangqi: A two-dimensional radiation hydrodynamic code with realistic equation of states

TL;DR

Guangqi addresses the challenge of simulating multidimensional radiation hydrodynamics with realistic equation of state physics by combining a fully implicit FLD solver with a general EoS in a 2D framework. The code uses a two-step operator-splitting scheme (hydrodynamics with a general EoS Riemann solver followed by implicit radiation transport), and introduces PSAMA to conserve angular momentum accurately in spherical-polar coordinates. It supports AMR/SMR, global domain decomposition via Hilbert and Z-order curves, and a monolithic multilevel radiation matrix, allowing robust testing across 0D, 1D, and 2D benchmarks, including hydrogen and helium EoS. The results demonstrate accurate energy exchange, reliable radiative shocks, and scalable performance, making Guangqi a capable testbed for axisymmetric astrophysical systems with complex thermodynamics and optically thick regions, while acknowledging FLD limitations and guiding future 3D extensions.

Abstract

We present Guangqi, a new two-dimensional, finite-volume radiation hydrodynamics code designed for high-performance astrophysical simulations. The code simultaneously resolves the hydrodynamic equations for complex equations of state (EoS) and implicit radiation transport under the flux-limited diffusion approximation. Written in Fortran and parallelized via the Message Passing Interface. Guangqi supports analytic hydrogen and helium EoS under the assumption of local thermal and chemical equilibrium. The framework is compatible with both Cartesian and spherical-polar geometries -- utilizing non-uniform grid spacing -- and incorporates static (SMR) and adaptive mesh refinement to optimize computational efficiency. To address the inherent challenges of angular momentum conservation in spherical-polar coordinates, we implement a robust and consistent "passive scalar angular momentum algorithm" (PSAMA). Domain decomposition is managed through both Z-order and Hilbert space-filling curves to ensure scalability. The code has been rigorously verified against a suite of standard benchmarks and newly designed test cases specifically intended to diagnose the non-linear coupling between gas dynamics, intricate EoS, radiation transport, and angular momentum conservation.

Figures (14)

• Figure 1: The specific heat capacity $C_{V}/\rho$ of hydrogen gas as a function of temperature. The yellow and purple bands indicate the dissociation and ionization region, respectively, where the specific heat capacity shows rapid variations.
• Figure 2: A mesh refinement structure showing refined cells in the bottom-left next to the unrefined cells. Values of interfaces (purple dots labeled I) and cell centers (black hollow circles) when there is a level difference.
• Figure 3: Example of domain decomposition with the same mesh refinement. The computational domain is distributed into 9 sub-domains, represented by different colors. The upper panel shows the domain decomposition of Hilbert curve, and the bottom panel shows the domain decomposition of Z-order curve.
• Figure 4: The temporal change of $e_{\text{g}}$ in the radiation and perfect gas coupling tests with three different initial gas conditions.
• Figure 5: Temporal evolution of $T_{\text{g}}$. Top panel: hydrogen gas heating test; bottom panel: hydrogen gas cooling test. The radiation temperature is held constant at $10^4$K and $4000$K for the heating/cooling test. The initial gas temperature is $4000$K and $10^4$K for the heating/cooling test.