Plasma simulation algorithms fall into two categories: fluid methods, which discard detailed information about the plasma’s velocity distribution in favor of computational efficiency; and kinetic methods, which retain this information in order to have a wider regime of validity. Recent large-scale simulations of plasmas have exposed the need for hybrid methods, which combine the speed of fluid methods with the wide applicability of kinetic methods. To this end, we have developed a numerical framework for the hybridization of kinetic and fluid methods. This was implemented in the hPIC2 plasma simulation code, which was designed to perform hybrid simulations on modern supercomputers. In addition, an algorithm was developed that identifies when a kinetically modeled portion of a plasma could be better treated as a fluid and dynamically converts it into a fluid. Together, these methods permit the efficient simulation of collisional plasmas spanning many scales.
