SUBJECT: Ph.D. Dissertation Defense
   
BY: Jordan Rader
   
TIME: Friday, June 21, 2013, 9:30 a.m.
   
PLACE: Love Building, 210
   
TITLE: Thermal Performance of Finger-Type Gas-Cooled Divertors
   
COMMITTEE: Dr. Said. I. Abdel-Khalik, Co-Chair (ME)
Dr. Minami Yoda, Co-Chair (ME)
Dr. S. Mostafa Ghiaasiaan (ME)
Dr. Michael Schatz (PHYS)
Dr. Thorsten Stoesser (Cardiff Univ.)
 

SUMMARY

A significant factor in the overall efficiency of the balance of plant for a future magnetic fusion energy (MFE) reactor is the thermal performance of the divertor. A significant fraction of the reactor power is delivered to the divertor as plasma impurities and fusion products are deposited on its surface. For an advanced MFE device, an average divertor heat load of 10 MW/m^2 is expected at steady-state operating conditions. Helium cooling of the divertors is one of the most effective ways to accommodate such a heat load. Several helium-cooled divertor designs have been proposed and/or studied during the past decade including the T-Tube divertor, the helium-cooled flat plate (HCFP) divertor, the helium-cooled multi-jet (HEMJ) divertor, the helium-cooled modular divertor with integral fin array (HEMP), and the helium-cooled modular divertor with slot array (HEMS). All of these designs rely on some form of heat transfer enhancement via impinging jets or cooling fins to help improve the heat removal capability of the divertor. For all of these designs very large heat transfer coefficients on the order of 50-60 kW/m^2-K have been predicted. As the conditions of a fusion reactor and associated helium flow conditions (600 °C and 10 MPa) are difficult to achieve safely in a controlled laboratory environment, the study of these divertors often relies on computer simulations and experimental modeling at non-prototypical, albeit dynamically similar, conditions. Earlier studies were based on the assumption that, for geometrically similar divertor test modules, dynamic similarity can be achieved by matching only the Reynolds number. Experiments conducted in this investigation using different coolants and test module materials have shown this assumption to be false. Modified correlations for the Nusselt number and loss coefficients for the HEMJ and HEMP-like divertor modules have been developed. These have been used to develop generalized performance curves to predict the divertor performance, i.e. the maximum allowable heat flux and corresponding pumping power fraction, at prototypical conditions. Additionally, a numerical study has been performed to optimize the fin array geometry of the HEMP-like divertor module. The generalized correlations and performance curves developed in this investigation can be incorporated into system design codes, thereby allowing system designers to optimize the divertor geometry and operating conditions.