Measurements of impurity concentrations and transport in the Lithium Tokamak Experiment

Abstract

This thesis presents new measurements of core impurity concentrations and transport in plasmas with lithium coatings on all-metal plasma facing components (PFCs) in the Lithium Tokamak Experiment (LTX). LTX is a modest-sized spherical tokamak uniquely capable of operating with large area solid and/or liquid lithium coatings essentially surrounding the entire plasma (as opposed to just the divertor or limiter region in other devices).

Lithium (Li) wall-coatings have improved plasma performance and confinement in several tokamaks with carbon (C) PFCs, including the National Spherical Torus Experiment (NSTX). In NSTX, contamination of the core plasma with Li impurities was very low (<0.1%) despite extensive divertor coatings. Low Li levels in NSTX were found to be largely due to neoclassical forces from the high level of C impurities. Studying impurity levels and transport with Li coatings on stainless steel surfaces in LTX is relevant to future devices (including future enhancements to NSTX-Upgrade) with all-metal PFCs.

The new measurements in this thesis were enabled by a refurbished Thomson scattering system and improved impurity spectroscopy, primarily using a novel visible spectrometer monitoring several Li, C, and oxygen (O) emission lines. A simple model was used to account for impurities in unmeasured charge states, assuming constant density in the plasma core and constant concentration in the edge. In discharges with solid Li coatings, volume averaged impurity concentrations were low but non-negligible, with~2-4% Li, ~0.6-2% C, ~0.4-0.7% O, and Z_eff<1.2.

Transport was assessed using the TRANSP, NCLASS, and MIST codes. Collisions with the main H ions dominated the neoclassical impurity transport, unlike in NSTX, where collisions with C dominated. Furthermore, neoclassical transport coefficients calculated with NCLASS were similar across all impurity species and differed no more than a factor of two, in contrast to NSTX where they differed by an order of magnitude. However, time-independent simulations with MIST indicated that unlike NSTX, neoclassical theory did not fully capture the impurity transport and anomalous transport likely played a significant role in determining impurity profiles.