Content area
Abstract
The transport of neutral atoms and molecules at the edge of a magnetically confined plasma plays an important role in the overall plasma performance. This transport problem is characterized by geometrical complexity, widely varying mean free paths, and sharp particle density variations. Most of the methods used to study neutral particles in plasmas are limited either by excessive computational time (Monte Carlo), inability to treat complex geometries, failure to treat widely varying mean free paths (diffusion theory, discrete ordinates) or lack of accuracy in some regimes (diffusion theory).
Two related computational methods for neutral particle transport in the outer regions of a diverted tokamak plasma have been recently introduced. These methods subdivide the computational domain into a number of relatively large regions, calculate transmission and escape probabilities for these regions using first flight integral transport methods, and then balance the partial currents or fluxes across the surfaces bounding these regions. While implementing the methods, a number of approximations were made in order to simplify the treatment of the angular distribution of particles and to characterize the transport probabilities. This work reports on the evaluation of the accuracy of these various approximations which was based on a detailed comparison with Monte Carlo. In addition, a number of corrections factors were introduced to improve accuracy. When dealing with the transport of a single species of monoenergetic neutrals, this model was shown to be capable of achieving accuracies comparable to Monte Carlo calculations at a fraction of the computational time. In addition, this work is to report the comparison of these methods and Monte Carlo calculations and experimental results for neutral densities in the DIII-D edge plasma.
