NIMROD simulations of DIII-D, Alcator C-Mod and ITER undergoing a similar rapid shutdown scenario (massive Ar injection) have been compared directly, in order to investigate how runaway electron (RE) confinement during disruptions scales with machine size. Poor RE confinement during the current quench in ITER is desirable as a loss term to compete against avalanche growth. Drift-orbits for a test population of higher energy REs are integrated as the MHD equations are evolved. When MHD fluctuations arise, some REs escape and strike the simulation boundary. The fraction of REs that remain confined when the MHD fluctuations have subsided is seen to increase with device major radius, R. Radial profiles of fluctuating field amplitudes (δB_r/B) show no clear trend in maximum amplitude versus R, but values near the edge (r/a=0.8) appear to scale approximately as 1/R, producing minimal stochasticity in the edge for ITER relative to C-Mod or DIII-D. Further simulations will seek to isolate the effects of major radius from other differences in the three plasma equilibria.

Global, linear GYRO simulations of beam-heated, shear-reverse DIII-D shot 142111 have definitively identified global energetic particle modes (EPMs), toroidal Alfvén eigenmodes (TAEs), and reverse shear Alfvén eigenmodes (RSAEs) at multiple values of toroidal mode number n. The EPM is seen to peak on singular surfaces, as observed in previous local GYRO simulations. When the TAE is observable, the eigenfunction peak is always seen in the gap between singular surfaces lying closest to the minimum in the q profile. In all simulations run to date, the RSAE is sub-dominant to an EPM or TAE. The RSAE frequency and approximate eigenfunction can be extracted from analysis of initial-value simulations at early times (see figure)). While the RSAE frequency “upsweep” (as the q profile evolves) is usually seen in experiments, the “downsweep” phase has proven easier to resolve in GYRO.

CER profile signals (CERFTI, CERQROTT, CERAROTC, etc.) in MDSplus have been enhanced to include new rho-based signals, brightness profiles, integration time subnodes, and chord name information. Data retrieval has also been made faster and less prone to error. CER profiles for old shots will be redone over the next few weeks. These changes do not affect CER data, GAProfiles, or Zipfit.

The two-fluid, linear plasma response to nonaxisymmetric fields has been successfully calculated using the initial value extended-MHD code M3D-C1. The analysis was performed for a low-beta, single-null DIII-D discharge with an n=1 I-coil field applied. The physical model in M3D-C1 includes the two-fluid generalized Ohm's law, gyroviscosity, and electron thermal convection. The computational domain extends across the separatrix to include the open field-line region. Throughout the plasma and SOL, the Spitzer resistivity is used. M3D-C1 also has the capability to include toroidal rotation in a self-consistent equilibrium. This will allow the numerical study of rotational and two-fluid screening of nonaxisymmetric fields in realistic geometry.

Professor Francois Waelbroeck from the University of Texas at Austin is visiting from September 1st through 7th . He participated in the Meeting to discuss ELM Control Physics Hypotheses held at GA on September 1st and presented a seminar on “Order versus Chaos: Response of the Pedestal to Resonant Magnetic Perturbations”. He also participated in further discussions on aspects and ideas on ELM suppression and density pumpout in DIII-D discharges with applied RMPs.