As part of the ARIES study, the Georgia Tech group has experimentally studied the thermal performance of a helium-cooled `finger-type' tungsten divertor design that uses jet impingement and a fin array to cool the plasma-facing surface. These studies were performed using air at Reynolds numbers Re, spanning those for prototypical operating conditions. A brass test section heated with an oxy-acetylene torch at incident heat fluxes up to 2 MW/m2 was used. Recently, data obtained with room-temperature helium suggests that dynamic similarity between the air and helium experiments cannot be achieved by only matching Re because of the difference in the relative contributions of convection and conduction through the annular side walls of the divertor. Numerical simulations suggest that achieving dynamic similarity requires matching the ratio of the thermal conductivity of the divertor module material to that of the coolant under operating conditions, as well as Re.

Studies were performed to verify that experiments at the prototypical Re and thermal conductivity ratio using helium at room temperature give Nusselt numbers Nu that are dynamically similar to those at prototypical operating conditions. Given that the thermal conductivity of helium decreases as temperature decreases, matching of the thermal conductivity ratio required a carbon steel test section with a thermal conductivity much lower than that of the brass alloy previously used. The resulting ratio of the test section to coolant thermal conductivities is similar to that of the tungsten alloy and helium at prototypical conditions. The data were used to verify generalized correlations for Nu, as a function of Re and the thermal conductivity ratio. The correlations can be used to determine the maximum heat flux that can be accommodated by the divertor at prototypical conditions.