(EX/P3-9) Edge Profile Stiffness and Insensitivity of the Density Pedestal to Neutral Fueling in Alcator C-Mod Edge Transport Barriers

Abstract.  Mechanisms determining the structure of edge ``pedestals" in temperature and density, which are associated with edge transport barrier (ETB) formation in tokamaks, are investigated on Alcator C-Mod, using techniques that include empirical scaling studies and both experimental diagnosis and modeling of neutral fueling effects on the density pedestal. Experiments suggest a strong role for critical gradient behavior in setting profile characteristics of edge plasma. Maximum pressure gradient scales as the square of plasma current in both H-modes without edge-localized modes, and in the near scrape-off layer (SOL) in ohmic discharges, demonstrating a ballooning-like scaling. In either case, the obtained pressure gradient, normalized to the square of plasma current, is a function of local collisionality, hinting that common physics may contribute to setting profile gradients in both confinement regimes. The near SOL findings connect well with first-principles numerical results that suggest an underlying physical explanation based on electromagnetic fluid drift turbulence. The electron density pedestal in H-mode shows a linear dependence on plasma current, and inferred effective cross-field transport coefficients increase markedly as current is lowered. Varying the neutral fueling source alone has little effect on density gradient scale lengths in the ETB and a relatively weak impact on the height of the density pedestal, even during aggressive deuterium puffing. Altogether, these data indicate a substantial role for plasma transport in determining the density pedestal and gradient, with details of the neutral fueling source being less important. A modeled response of the density pedestal to perturbations to the edge neutral source couples a kinetic neutral treatment with a diffusive model for the plasma transport. The modeled response to small source perturbations is qualitatively consistent with experimental measurements, though the response to large perturbations does not reproduce the typically clamped density gradients seen in experiment during H-mode puffing. Together these results suggest that a simple diffusive model for plasma transport is deficient, and that a critical gradient assumption for transport may be essential for pedestal modeling.

Full paper available (PDF)