Multigroup Radiation Diffusion on a Moving Mesh: Implementation in RICH and Application to Tidal Disruption Events

TL;DR

This work extends the moving-mesh radiation-hydrodynamics code RICH to a multigroup flux-limited diffusion solver, enabling broadband, frequency-dependent radiation transport on an unstructured, dynamically evolving mesh. The authors derive and discretize the multigroup diffusion equations, implement a fully implicit coupling to the material energy and momentum, and validate the approach against Marshak waves, radiative shocks, and DENSMORE multigroup benchmarks. They also introduce convergence-acceleration techniques to manage stiffness in optically thick regions. As a proof of principle, they apply the multigroup RHD module to a tidal disruption event by an intermediate-mass black hole, producing self-consistent, time-dependent spectra and sky maps across ten frequency groups, including an early X-ray flash tied to nozzle shocks. The results demonstrate the feasibility and scientific value of self-consistent multigroup radiation on moving meshes for high-energy astrophysical phenomena and spectral predictions.

Abstract

Radiation-hydrodynamics (RHD) determines the bulk evolution and observable emission in a wide variety of high-energy astrophysical phenomena. Due to their complexity, RHD problems must usually be studied through numerical simulation. We have extended the publicly available RICH code, which previously solved the equations of RHD in the limit of grey flux-limited diffusion (FLD), to operate with a multigroup FLD solver. RICH is a semi-Lagrangian code that solves the equations of RHD on an unstructured moving mesh, and is the first multigroup RHD moving mesh code, making it uniquely applicable to problems with extreme dynamic range and dynamically important radiation forces. We validate our multigroup module against multiple analytic benchmarks, including a novel test of the RHD Doppler term. The computational efficiency of the code is aided by a novel scheme to accelerate convergence in optically thick cells. Finally, we apply multigroup RICH in a pilot study of a stellar tidal disruption event (TDE), using a $10^4 M_\odot$ intermediate-mass black hole. Our simulations self-consistently produce a bright early-time X-ray flash prior to peak optical/UV light, in qualitative agreement with post-processing of (grey) RICH simulations of supermassive black hole TDEs, as well as X-ray observations of the TDE AT 2022dsb.

Figures (15)

• Figure 1: Comparison of the radiation temperature and material temperature at $t=1\ \text{ns}$ between the rich simulation and the analytic solution for Marshak problems 1 (left) and 2 (right). In problem 2 the absorption coefficient is large enough to enforce LTE, so $T_{\rm rad}\simeq T_{\rm mat}$.
• Figure 2: Comparison of the radiation temperature and material temperature at $t=1\ \text{ns}$ between the rich simulation and the analytic solution for Marshak problems 3 (left) and 4 (right).
• Figure 3: The gas temperature (left) and radiation energy density (right) for the Mach 2 benchmark along with the semi-analytic solution.
• Figure 4: Comparison of the gas temperature at $t=1$ ns between rich and the data taken from DENSMORE20126924 for the cases $\kappa_0 = 10\;\text{keV}^{3.5}$ (left) and $\kappa_0 = 100\;\text{keV}^{3.5}$ (right)
• Figure 5: Comparison of the gas temperature at $t=1$ ns between rich and the data taken from DENSMORE20126924 for the medium opacity case $\kappa_0 = 1000\;\text{keV}^{3.5}$ (left) and the step function opacity as presented in Eq. \ref{['eq:densmore_step_function_opacity']} (right).