Two improvements in heat-transfer design have been investigated with a view toward increasing the efficiency of refrigerators used to liquefy gases. The improvements could contribute to the development of relatively inexpensive, portable oxygen liquefiers for medical use.
A description of the heat-transfer problem in a pulse-tube refrigerator is prerequisite to a meaningful description of the first improvement. In a pulse-tube refrigerator — in particular, one of in-line configuration — heat must be rejected from two locations: an aftercooler (where most of the heat is rejected) and a warm heat exchanger (where a small fraction of the total input power must be rejected as heat). Rejection of heat from the warm heat exchanger can be problematic because this heat exchanger is usually inside a vacuum vessel.
When an acoustic-inertance tube is used to provide a phase shift needed in the pulse-tube cooling cycle, another problem arises: Inasmuch as the acoustic power in the acoustic-inertance tube is dissipated over the entire length of the tube, the gas in the tube must be warmer than the warm heat exchanger in order to reject heat at the warm heat exchanger. This is disadvantageous because the increase in viscosity with temperature causes an undesired increase in dissipation of acoustic energy and an undesired decrease in the achievable phase shift. Consequently, the overall performance of the pulse-tube refrigerator decreases with increasing temperature in the acoustic-inertance tube.
In the first improvement, the acousticinertance tube is made to serve as the warm heat exchanger and to operate in an approximately isothermal condition at a lower temperature, thereby increasing the achievable phase shift and the overall performance of the refrigerator. This is accomplished by placing the acoustic-inertance tube inside another tube and pumping a cooling fluid (e.g., water) in the annular space between the tubes. Another benefit of this improvement is added flexibility of design to locate the warm heat-rejection components outside the vacuum vessel.
The second improvement is the development of a compact radial-flow condenser characterized by a very high heat-transfer coefficient and a small pressure drop. The solid heat-transfer medium in this condenser is a core of aluminum foam with a mean pore diameter of ˜100 µm and a very high surface-area/volume ratio. At its radially innermost surface, the aluminum foam core is in contact with a cold head.
The vapor (e.g., oxygen) that one seeks to condense enters the condenser through a feed tube, then flows into an annular inlet plenum that surrounds the foam. The vapor then flows radially inward through the foam, toward the cold head, condensing along the way as it encounters colder foam. At the inner radius, the condenser, the subcooled liquid enters axial holes that lead out of the condenser.
The narrowness of the pores and the high surface-area/volume ratio of the foam give rise to an extremely high volumetric heat transfer coefficient (of the order of 106W/m3K); as a result, the condenser volume needed to obtain a given degree of cooling is very small. Another advantage of the high heat-transfer coefficient is that little subcooling is needed to condense the vapor and, therefore, the amount of cooling power is less than would otherwise be needed.
This work was done by Jerry L. Martin of Mesoscopic Devices, LLC for Johnson Space Center. For further information, contact the Johnson Innovative Partnerships Office at (281) 483-3809. MSC-23021/22