A comparison study of discrete-ordinates and flux-limited diffusion methods for modeling radiation transport

The Center for Radiative Shock Hydrodynamics (CRASH) is investigating methods of improving the predictive capability of numerical simulations for radiative shock waves that are produced in Omega laser experiments. The laser is used to shock, ionize, and accelerate a beryllium foil into a xenon-filled shock tube. These shock waves, when driven above a threshold velocity of about 60 km/s, become strongly radiative and convert much of the incident energy flux into radiation.Radiative shocks have properties that are significantly different from purely hydrodynamic shocks and, in modeling this phenomenon numerically, it is important to compute radiative effects accurately. In this article, we examine approaches to modeling radiation transport by comparing two methods: (i) a computationally efficient, multigroup, flux-limited-diffusion approximation, currently in use in the CRASH radiation-hydrodynamics code, with (ii) a more accurate discrete-ordinates treatment that is offered by the radiation-transport code PDT. We present a selection of results from a growing suite of code-to-code comparison tests, showing both results for idealized problems and for those that are representative of conditions found in the CRASH experiment.