GYRO has been used to identify global eigenmodes for the first time. Global ITG modes have been observed in GYRO initial-value simulations in a variety of experimental plasmas using Miller model equillibria. A “proof-of-principle” investigation of a DIII-D discharge found multiple unstable global ITG modes across the n-spectrum. At large n, the most unstable global eigenfunction is identified in a straightforward way with the long time state of the system. At lower n, multiple modes with close growth rates tend to exist in different regions of the global simulation domain. A de-convolution of these modes is possible if each complex eigenfrequency is known. De-convolution of co-dominant ITG modes has been shown for n = 5 modes in this DIII-D discharge. In another beam-heated DIII-D discharge, a similar analysis at n = 5 resolved a co-dominant ITG and an unidentified very-low-frequency instability when the beam was neglected. Preliminary results including the high-energy beam show high-frequency global Alfvén instabilities. Work is underway to resolve separate TAE and RSAE branches expected from experimental observations.
The new APS09 version of TGLF (see Highlight from January 15 2010 at Theory Weekly Highlights for January 2010) has been applied to examine the impact of the new collision model on the ITER baseline scenario fusion performance and the results obtained are close to those previously found using the GLF23 transport model. The predictions using TGLF with the original collision model are pessimistic compared to the GLF23 results. Using the new collision model the fluxes are reduced and the fusion projections are closer to those found using GLF23. The impact of collisions is more prevalent for the very low-k modes which survive in ITER because E×B shear effects are weak. This is not typically not the case in modeling studies of DIII-D and JET where E×B shear quenches the very low-k modes and changing the collision model has very little impact on the predicted profiles. Additional DIII-D cases with very low E×B shear are needed to test the very low-k transport in TGLF.
The previous tedious and inefficient process of improving statistics in ORBIT-RF simulations by summing and averaging results from separate simulations performed at several different time scales has been greatly improved by instead dumping fast ion information at certain time steps from a single simulation. Typically, ORBIT-RF simulations use at most 150K test particles to speed up the simulations on the NERSC Franklin Cray XT4 massively parallel processing system. The new dumping technique is more efficient and saves performing separate simulations. For example, by dumping fast ion data at 20 different time steps during one simulation, one can obtain fast-ion data for effectively 3 million particles (20 steps Χ 150K particles) at the end of the simulation. This may also be a more natural way of extracting fast ion data for the purpose of comparing the simulations with FIDA since FIDA data is also extracted and averaged over a few slowing down time scales. Tests for NSTX (#128739) HHFW heating discharges indicate very good agreement with the previous result obtained with the old fast-ion extracting technique. Testing against DIII-D and comparison with FIDA is underway.
Residual stress refers to the remaining toroidal angular momentum (TAM) flux (divided by major radius) when the shear in the parallel velocity (and parallel velocity itself) vanishes. Previously we demonstrated with gyrokinetic (GYRO) simulations that TAM pinching from the diamagnetic level shear in the ExB velocity could provide the residual stress needed for spontaneous toroidal rotation. We have now shown that the shear in the diamagnetic velocities themselves provide comparable residual stress (and level of stabilization). The sign of the residual stress, quantified by the ratio of TAM flow to ion power flow (M/P), depends on the signs of the various velocity shears as well as ion (ITG) versus electron (TEM) mode directed turbulence. The residual stress from these temperature and density gradient diamagnetic velocity shears is demonstrated in global gyrokinetic simulation of “null” rotation DIII-D discharges by matching M/P profiles within experimental error.
Resistive wall boundary conditions have been implemented and tested in M3D-C1. The resistive boundary conditions are calculated using the VACUUM code, which determines the magnetic response outside the wall given the normal magnetic field at the wall. This new capability will allow calculations of RWM stability and nonaxisymmetric plasma response using a resistive, two-fluid model of both the open and closed field line regions.
These highlights are reports of research work in progress and are accordingly subject to change or modification