Cost-effective liquefaction of gases at cryogenic temperatures requires a combined approach of designing efficient refrigeration cycles to generate cooling capacity and designing efficient liquefaction processes to utilize that capacity. This dissertation addresses both approaches for improvement of the liquefaction process. Magnetic refrigeration employing the magnetocaloric effect of ferromagnetic materials has been identified as potentially more efficient and cost-effective than conventional refrigeration systems. One magnetic cycle that shows promise for efficiently achieving cooling over large temperature spans is active magnetic regenerative refrigeration (AMRR). In this cycle the magnetic material serves the dual functions of work input and thermal regeneration, however the operation is complex with coupled fluid, thermal, and magnetic phenomena and a clearer understanding of the regenerative operation is required. Models to elucidate the flow characteristics of the regenerative heat exchangers of rotary AMRR and Gifford-McMahon (GM) systems using a commercially available computational fluid dynamics (CFD) software package are described. Theoretical results are presented to qualify and quantify the effect of maldistributed flow within regenerators. Experimental results of an improved regenerator for the GM system based on the CFD flow simulations are presented. Efforts to develop an improved AMRR thermal model using the commercial package are also described.

The second approach for the design of efficient liquefiers addresses a drawback of the usual embodiment of the AMRR cycle, namely, the provision of cooling at a single temperature which necessitates that cryogenic designs have multiple stages providing cooling over a range of discrete temperatures for an efficient liquefaction process. Use of multiple stages leads to increased expense and complexity. A simple, inexpensive plumbing change of the flow through the regenerator of a single-stage device can significantly increase the overall liquefaction capacity compared to the usual flow configuration making additional staging unnecessary. This dissertation describes the alternative flow arrangement, known variously as “bypass flow,” “permanent flow,” or “DC flow,” which is suitable for all passive and active regenerative refrigeration cycles used as liquefiers. Theoretical results showing increased liquefaction capacity when bypass flow is employed are given for active magnetic regenerative and Gifford-McMahon systems. Experimental results are presented for a single-stage GM refrigerator modified for bypass flow which demonstrates increased liquefaction capacity.