The effect of safety factor profile on transport in steady-state, high-performance scenarios.

An analysis of the dependence of transport on the safety factor profile in high-performance, steady-state scenario discharges is presented. This is based on experimental scans of q95 and qmin taken with fixed βN, toroidal field, double-null plasma shape, divertor pumping, and electron cyclotron current drive input. The temperature and thermal diffusivity profiles were found to vary considerably with the q-profile, and these variations were significantly different for electrons and ions. With fixed q95, both temperature profiles increase and broaden as qmin is increased and the magnetic shear becomes low or negative in the inner half radius, but these temperature profile changes are stronger for the electrons. Power balance calculations show the peak in the ion thermal diffusivity (χi) at ρ=0.6-0.8 increases with q95 or qmin. In contrast, the peak in the electron diffusivity (χe) decreases as qmin is raised from ∼1 to 1.5, and it is insensitive to q95. This is important for fully non-inductive scenario development because it demonstrates that elevated qmin and weak or reversed shear allow larger electron temperature gradients and, therefore, increased bootstrap current density to exist at ρ=0.6-0.8. Chord-averaged measurements of long wavelength density fluctuation amplitudes (ñ) are shown, and these have roughly the same dependence on q-profile as χi. This data set provides an opportunity for testing whether theory based transport models can provide insight into the underlying transport physics of high performance scenarios and if they can reproduce observed experimental trends. To this end, we applied the trapped gyro-Landau fluid (TGLF) code to calculate the linear stability of drift waves and found that the resulting variation of growth rates with q-profile are mostly inconsistent with the observed trends of χi, χe, and ñ with q-profile. TGLF simulations of the temperature profiles consistent with heating sources also have mixed agreement with the measured profiles, such that the simulated electron and ion heat flux in low qmin discharges are too low and heat fluxes in high qmin discharges are too high. [ABSTRACT FROM AUTHOR]