Numerical Modeling and Experimental Testing of a Mixed Gas Joule-Thomson Cryocooler

Under the supervision of Greg Nellis and John Pfotenhauer; 186pp.Mixed gas Joule-Thomson (MGJT) systems have been shown to provide order of magnitude improvements in efficiency relative to JT systems that use pure working fluids. This thesis presents theoretical and experimental work related to using a single- stage, low power (< 1 W) MGJT system for cooling the current leads required by high- temperature superconducting electronics. By thermally integrating the current leads with the recuperative heat exchanger of a MGJT cycle, it is possible to intercept the electrical dissipation and conductive heat leak of the wires at a relatively high temperature which provides a thermodynamic advantage. Also, directly cooling the leads rather than indirectly cooling the chips may provide some advantages relative to thermal integration. To design the recuperative heat exchanger for the MGJT cycle, the composition of the gas mixture was optimized using a robust genetic optimization technique. Following mixture selection, the optimization model was modified so that it included the effect of frictional pressure drop, axial conduction through the heat exchanger, and the overall conductance available from the heat exchanger on the performance of the MGJT cycle. The individual influences of these loss factors on the refrigeration power of the MGJT cycle were investigated parametrically and conceptually in order to determine the target values for a low power system and develop some insight into the relative importance of each effect. A detailed model of the specific Hampson-style heat exchanger geometry was developed and used to obtain a design for an initial demonstration device. The demonstration device was fabricated and integrated with a thermal vacuum test facility, gas handling equipment, and the appropriate instrumentation. Several tests were carried out. First, the heat exchanger alone was tested (outside of a JT cycle) using helium as the working fluid. These data provided some experimental verification of the detailed model. Next, the test facility was modified through the installation of a fixed orifice expansion valve to allow open cycle testing of the device using high pressure (9.745 MPa) pure Argon. These measurements provided further insight into the performance of the device. The test facility was subsequently integrated with a compressor in order to allow measurements of the Device's performance using gas mixtures in a closed loop configuration. These test results ultimately revealed issues relative to contamination, which were addressed through the installation of a liquid nitrogen trap, as well as liquid management. The liquid management issue is thought to be related to inadequate vapor kinetic energy which does not provide sufficient momentum transfer to the liquid to move it through the system. The liquid management issue constrains the performance of the MGJT cycle at low mass flow rates and was explored over a very limited range of conditions. Further testing is suggested which will allow the liquid management constraint to be explored more completely in order to guide future designs.Sponsored by the Office of Naval Research