Theory Weekly Highlights for November 2014

November 21, 2014

A set of global, linear GYRO simulations, including an isotropic slowing-down distribution of beam ions, has been used to study beam ion transport due to Alfvén eigenmodes (AEs) in DIII-D. Two recent DIII-D shots, which differ principally in the q profile, with minimum q of approximately 1 for shot 153071 and approximately 2 for shot 153072, are observed to have very different AE-induced beam transport. The lower-q case has virtually no AE-induced beam transport, while the higher-q case shows significant beam profile flattening due to active AEs. GYRO simulations have confirmed, AE’s are more unstable in the shot with greater anomalous beam ion transport (153072). Moreover, the dominant AE eigenfunction in 153072 extends to a larger radius (r/a>0.8 vs. r/a=0.6), suggesting a greater likelihood of edge loss of beam ions. Preliminary GYRO global gyrokinetic eigenvalue solver results show 2 sub-dominant AEs in each case, with the first subdominant mode in 153072 having nearly three times the growth rate than its counterpart in 153071. Study of this case with the local critical-gradient profile prediction model ALPHA is scheduled for the near future.

November 14, 2014

The GLF23 transport model (R.E. Waltz et al., 1997) has been converted to the Fortran 90 standard and added to the GACODE suite. The new version outputs quasi-linear fluxes of energy, particles, and momentum, rather than effective diffusivities, in the same gyro-Bohm units as used by the quasi-linear model TGLF and the gyro-kinetic turbulence code GYRO. A new transport code interface and stand alone test driver for GLF23 have been written with the same architecture as used for TGLF and the neoclassical code NEO. The TGYRO transport solver now has GLF23 as an option. Although GLF23 is not as accurate a quasi-linear representation of the gyro-kinetic turbulent fluxes as TGLF, it is much faster and is still in use worldwide for transport modeling.

November 07, 2014

In an attempt to explain the m/n = 2/1 disruption observed in the diverted L-mode DIII-Discharge #150513 when q95 reached 1.9, MHD stability calculations using MARS with a resistivity profile strongly increasing in the edge region revealed fast growing resistive kink modes strongly peaked at the edge that appear to be likely candidates. Previously, ideal stability calculations all yielded complete stability despite the ideal-like character of the observed disruption since the reconstruction for the diverted equilibrium yield an edge q well above 2.0. The resistive kink modes have the character of an ideal mode, with no sign of tearing, provided the resistivity near the q95 surface is sufficiently large - generally larger than the Spitzer model would predict using the reconstructed temperature profile. However, the enhancement required to obtain instability is in the low temperature edge where it seems likely that both the measured temperature and the model itself are probably inaccurate. This result also yields insight into the longstanding puzzle of why q95 in a diverted discharge appears to play the role of the edge q value in a limiter discharge. Effectively, the rapid increase in resistivity makes the edge behave more like a vacuum. While this has been the standard conjecture, the conventional wisdom has been that the mode would be a tearing mode, rather than the resistive kink that was found.

These highlights are reports of research work in progress and are accordingly subject to change or modification