Plasma flows in the Alcator C-Mod scrape-off layer

Abstract

Near-sonic parallel plasma flows are persistently observed in the scrape-off layer (SOL) of tokamaks, at locations far from material surfaces. Ballooning-like transport asymmetries are thought to be a principal driver for the strong parallel flows, a hypothesis supported by the observation of steep high-field side pressure profiles in double-null discharges. Yet parallel flow can also arise as a result of toroidal plasma rotation and/or neoclassical Pfirsch-Schliiter currents. In addition, the mechanism that closes the mass-flow loop back onto itself has remained elusive. To investigate these phenomena, a novel magnetically-actuated scanning probe has been deployed on the high-field side in Alcator C-Mod. This probe, along with two other scanning probes on the low-field side, measure the total plasma flow vector at these locations: parallel flows, perpendicular E_r x B drifts and radial fluctuation-induced particle fluxes. Boundary layer flows have been systematically examined as magnetic topology (upper versus lower-null) and plasma density were changed. It is found that the plasma flow pattern can be decomposed into two principal parts: (1) a drift-driven component, which lies within a magnetic flux surface and is divergence-free and (2) a transport-driven component which gives rise to parallel flows on the high-field side scrape-off layer.

(cont.) Toroidal rotation, Pfirsch-Schlilter and transport-driven contributions are unambiguously identified. Parallel flows are found to dominate the high-field particle fluxes; the total poloidally-directed flow carries one half of the particle flux arriving on the inner divertor. As a result, convection is also found to be an important player in high-field side heat transport. In contrast, E_r x B plus parallel flows yield a mostly-toroidal flow component in the low-field SOL. The magnitude of the transport-driven flow component is found to be quantitatively consistent with radial fluctuation-induced particle fluxes measured on the low-field side, identifying this as the primary driver. In contrast, fluctuation-induced flux measurements on the high-field side midplane are found to be essentially zero, thereby excluding an 'inward pinch' effect as the mechanism that closes the mass-flow loop in this region.