Search

It is shown that numerical modelling of edge plasma physics may be successfully parallelized in time. The parareal algorithm has been employed for this purpose and the SOLPS code package coupling the B2.5 finite-volume fluid plasma solver with the kinetic Monte-Carlo neutral code Eirene has been used as a test bed. The complex dynamics of the plasma and neutrals in the scrape-off layer (SOL) region makes this a unique application. It is demonstrated that a significant computational gain (more than an order of magnitude) may be obtained with this technique. The use of the IPS framework for event-based parareal implementation optimizes resource utilization and has been shown to significantly contribute to the computational gain. [ABSTRACT FROM AUTHOR]

Abstract: Linear kinetic Monte Carlo particle transport models are frequently employed in fusion plasma simulations to quantify atomic and surface effects on the main plasma flow dynamics. Separate codes are used for transport of neutral particles (incl. radiation) and charged particles (trace impurity ions). Integration of both modules into main plasma fluid solvers provides then self-consistent solutions, in principle. The required interfaces are far from trivial, because rapid atomic processes in particular in the edge region of fusion plasmas require either smoothing and resampling, or frequent transfer of particles from one into the other Monte Carlo code. We propose a different scheme here, in which despite the inherently different mathematical form of kinetic equations for ions and neutrals (e.g. Fokker–Planck vs. Boltzmann collision integrals) both types of particle orbits can be integrated into one single code. We show that the approximations and shortcomings of this “single sourcing” concept (e.g., restriction to explicit ion drift orbit integration) can be fully tolerable in a wide range of typical fusion edge plasma conditions, and be overcompensated by the code-system simplicity, as well as by inherently ensured consistency in geometry (one single numerical grid only) and (the common) atomic and surface process modules. [Copyright &y& Elsevier]