The first theoretical model of the penetration for a pellet cluster stream formed when a super large, high-speed D₂ pellet is shattered by collision against a striker plate configuration was developed. Injection of such a spray of pellet fragments is being investigated as a new method of quenching avalanche runaways during fast plasma terminations. The model assumes that all pellets are same size and uniformly distributed in the stream, which is assumed to have the shape of a long cylinder of specified dimensions. The two coupled differential equations describing the penetration and ablation of the stream into the plasma, while the stream is simultaneously cooling the plasma were cast in dimensionless form, using only a single independent variable formed by the combination of the rescaled time and space variables. The solutions will be applied to the recent experiment carried out on DIII-D and will be presented at the upcoming JET ITPA meeting from October 6 through 8 in Culham, UK.

Stability studies with ELITE, and pedestal predictions with EPED1, find that hybrid discharges in which the RMP was used for ELM suppression follow a similar pattern to other RMP ELM-suppressed discharges (before RMP, ELMs triggered by peeling- ballooning instabilities; with RMP and ELM suppression, the edge is stable to peeling- ballooning modes). In a separate investigation, shape studies conducted with ELITE in preparation for the Super H-Mode experiment found that “peapodness” has a relatively small effect on edge stability, but increasing triangularity by pulling the X-points in improves edge stability significantly.

Alan Glasser from the University of Washington visited GA this week to discuss collaboration on various aspects of MHD theory and numerical simulations.

To improve portability of the EFIT equilibrium reconstruction code, an EFIT source-code management system under the IMFIT Integrated Modeling and Fitting tool was developed and tested. This IMFIT utility provides a convenient tool to check out the EFIT source code from the SVN version control system and convert it to a specific tokamak device. An ITER configuration has been successfully tested. Additionally, work to convert EFIT arrays to allow dynamic memory allocation under F90/95 has also been completed. All grid-size related arrays have been modified to allow a single EFIT version for different grid sizes. The work was carried out by a summer intern student.

Simulations of a DIII-D current quench after rapid Argon cooling have been carried out with NIMROD using a newly added single particle orbit following model for runaway electrons (REs). A comparison of cases with and without applied n=3 I-Coil fields have produced the counter-intuitive result that resonant magnetic perturbation (RMP) fields may increase RE confinement in the current quench. In both cases, n>0 instabilities arise as the current profile evolves, which destroy flux surfaces and rapidly redistribute the current. The RMP case shows non-linear growth of the n=3 mode early, but the growth rate and amplitude of the other n>0 modes is reduced relative to the no-RMP case, producing less destruction of flux surfaces, and more confined REs following the MHD crash. Validation of these results against DIII-D experiments is underway.

Howard Wilson, of the University of York, visited GA from August 10 through September 2 to work on adding a self-consistent treatment of diamagnetic effects to the edge ballooning formalism and to ELITE.