Lang Lao attended and chaired the Seventh ITER Integrated Modeling Expert Group (IMEG) Annual Meeting at ITER Headquarters, Cadarache, France, held September 14-16, 2015. The IMEG consists of two representatives from each of the seven ITER partners. The two main goals of the meeting were to discuss progress in the ITER partners domestic integrated modeling programs and to review progress and advise ITER on its Integrated Modeling Program to develop an Integrated Modeling Analysis Suite (IMAS) and an infrastructure with a workflow management system and a data model to support ITER plasma operation and plasma research.

A new module, reOrbit, has been added to the OMFIT framework to calculate drift orbits for relativistic runaway electrons (REs) on 2D or 3D fields. The module uses the RE drift orbit calculation routines in NIMROD, but does not advance any MHD equations, and requires only an EFIT g-file and optional 3D fields file as input. For a large number of test electrons, the RE position is written at fixed time intervals, and the output can provide information about the region of RE losses and the loss rate when stochasticity is present. For a small number of test particles full orbit information is stored and can also provide information about orbit losses due to curvature drift as a function of RE energy. The orbits are integrated beyond the LCFS and striking points on the simulation boundary are also given as output. At present, the equations are valid only for suprathermal, passing runaway electrons with an assumed small pitch angle, although upgrades to the physics model are planned.

A new version of the TGLF transport model has been developed which dramatically improves energy, particle and momentum transport predictions for high poloidal beta discharges on DIII-D. Previous modeling of the fully non-inductive high-betap discharges developed on DIII-D, as a prototype for steady state operation on EAST, found that there was a large shortfall in the electron energy transport predicted by the TGLF transport model across most of the core region. New multi-scale ion and electron GYRO simulations of an Alcator C-Mod discharge performed by Nathan Howard (MIT) have shown a strong interaction of the ion and electron scales, such that the high-k electron transport depends on the low-k ion transport. Weaker ion transport leads to stronger electron transport. A new model of the saturated electric potential for these multi-scale simulations has been developed that explains this electron-ion energy transport coupling as being mediated by the zonal fluctuations driven by the low-k turbulence. Using this saturation model in TGLF gives an accurate calculation of the transport in the multi-scale GYRO simulations. Applying this new version of TGLF to the DIII-D high-betap discharges shows much better agreement, compared to the original TGLF, of the predicted profiles with the data for all four channels (density, electron and ion temperature and ion toroidal rotation). We now plan to explore whether the new TGLF model can also resolve the shortfall of energy transport in the L-mode edge, which was the original motivation for the multi-scale simulations.

The recently proposed explanation for the ideal-like m/n = 2/1 external kink instability observed in diverted discharges when q95 is just below 2 (see March 20 2015 Highlight at Theory Weekly Highlights for March 2015) as a resistive kink, has also yielded an explanation of a slightly puzzling experimental result in limiter discharges for which the ideal instability is predicted when qedge < 2. In the experiments, the instability was actually first observed when qedge was slightly above 2.0. For example, in the limiter discharge #154907 the instability onset was determined at qedge = 2.08, so that qedge > 2, significantly outside the measurement error estimated using a detailed Monte-Carlo analysis (see Hanson, et al, Phys. Plasmas, 21, 072107). Previously, the small discrepancy was attributed to a possible unknown systematic equilibrium reconstruction error. However, according to the proposed model, the resistive kink is also expected to be destabilized for qedge > 2 in the limiter configuration if the q=2 surface is in the high resistivity region, which is the case. Calculations with the resistive MHD code MARS, have, in fact, found the resistive 2/1 kink unstable if the Spitzer resistivity profile is assumed, with an enhancement in resistivity required only in the small region around q=2. In the Spitzer model, this corresponds to a lower edge temperature than measured in the outer one percent of the flux. Further work is in progress to identify and eliminate possible sources of systematic error in qedge and the edge electron temperature.