Determination of sheath heat transmission coefficient in NSTX discharges with applied lithium coatings

Abstract

Recycled particle flux can be a significant contributor to tokamak edge plasma density, and lead to reductions in edge temperature. Previous measurements have shown that solid evaporated lithium coatings can lead to lowered edge recycling, corresponding decreases in edge plasma density, and a radial broadening of the electron temperature profile. During the 2010 run campaign, The National Spherical Torus Experiment operated with both solid and liquid lithium coatings on its plasma-facing components. A section of graphite outer divertor tiles was replaced with a substrate designed to hold liquid lithium coatings. While the lithium coatings on the graphite remain solid, the plates can be heated to render the lithium into a liquid state.

In preparation for this campaign, a 99-tip dense Langmuir probe array was installed in the outboard divertor to measure scrape-off layer density and temperature. The probe array was located so as to radially span these two different divertor surfaces and measure their respective effects on the temperature and density. A dual-band fast IR camera was also installed to provide surface temperature and heat flux measurements. The use of two-color IR thermography allows for an assessment of effects due to the uncertain, phase- and purity-dependent emissivity of the lithium coatings.

The present study compares the derived heat fluxes from these diagnostics to determine an effective classical sheath heat transmission coefficient γ, a measure of the heat flow reaching the device surfaces as a function of plasma parameters, namely the edge temperature and density - quantities which should be modified by lithium coatings. The sheath heat transmission is of great interest to future devices, which can expect to see large steady-state and transient heat fluxes to material surfaces. The value of γ was measured to be 2.49 +/- 0.04, a factor of ∼3 smaller than the expected classical result of ∼7. The implications of this measurement and the changes to edge plasma profiles that are possible causes of the lower observed value are discussed. Supported by US-DOE Contract DE-AC02-09CH11466.