Marginal Stability of Microturbulence near ITB Onset on Alcator C-Mod

Baumgaertel, J.A. ; Redi, M.H. ; Budny, R.V. ; McCune, D.C. ; Dorland, W. ; Fiore, C.L.

Princeton University. Plasma Physics Laboratory.

关键词: Critical Temperature;    Transport Computational Physics;    70 Plasma Physics And Fusion Technology;    Simulation;    Temperature Gradients;   

DOI  :  10.2172/834530
RP-ID  :  pppl-3999
RP-ID  :  AC02-76CH03073
RP-ID  :  834530

美国|英语

来源: UNT Digital Library

【 摘 要 】

Insight into microturbulence and transport in tokamak plasmas is being sought using linear simulations of drift waves near the onset time of an internal transport barrier (ITB) on Alcator C-Mod. Microturbulence is likely generated by instabilities of drift waves and causes transport of heat and particles. This transport is studied because the containment of heat and particles is important for the achievement of practical nuclear fusion. We investigate nearness to marginal stability of ion-temperature-gradient (ITG) modes for conditions in the ITB region at the trigger time for ITB formation. Data from C-Mod, analyzed by TRANSP (a time-dependent transport analysis code), is read by the code TRXPL and made into input files for the parallel gyrokinetic model code GS2. Temperature and density gradients in these input files are modified to produce new input files. Results from these simulations show a weak ITG instability in the barrier region at the time of onset, above marginal stability; the normalized critical temperature gradient is 80% of the experimental temperature gradient. The growth rate increases linearly above the critical value, with the spectrum of ITG modes remaining parabolic up to a multiplicative factor of 2. The effect of varying density gradients is found to be much weaker and causes the fastest growing drift mode to change from ITG to trapped-electron mode character. Simulations were carried out on the NERSC [National Energy Research Supercomputer Center] IBM 6000 SP using 4 nodes, 16 processors per node. Predictive simulations were examined for converged instability after 10,000-50,000 time-steps in each case. Each simulation took approximately 30 minutes to complete on the IBM SP.