A two-stage regenerative cryocooler has been selected to produce and maintain the low temperatures required by the superconducting magnet system in an active magnetic regenerative liquefier (AMRL). By using a mechanical cryocooler, the need for liquid cryogens has been eliminated, thereby simplifying the overall system design. The operation of practical AMRLs requires relatively large magnetic fields (e.g., 8 T). Currently, these fields can only be produced via low-temperature superconducting magnets that typically operate at liquid helium temperatures (4.2 K) .

Obtaining non-zero cooling powers below 10 K is difficult when regenerative cryocoolers are used. This is mainly because the volumetric heat capacity of pressurized helium (the working fluid) is much higher than the corresponding value for lead (the material commonly used in low temperature regenerators). A commercial two-stage Gifford-McMahon cryogenic refrigerator has been successfully modified to reduce its minimum, no-load temperature from 6.1 ± 0.1 K to 3.42 ± 0.05 K at a nominal operating frequency of 1.2 Hz. The cooling power at 4.2 ± 0.1 K was measured to be 0.430 W with zero load at the first stage. The superconducting magnets in the AMRL have been designed to operate at 4.5 K. The refrigeration power available at this temperature was measured to be 0.504 W with a simultaneous second-stage load of 20W at 42.8 ± 0.5 K.

The required modifications to the GM cryocooler included the selection, preparation and characterization of appropriate magnetic regenerator materials, their fabrication into highly efficient and reliable cryogenic regenerators, and the modification of the frequency of operation. A novel and proprietary manufacturing technique was also developed to ensure the structural integrity of the second-stage regenerator without the use of spherical particles. This technique used irregularly-shaped particles made of rare-earth/transition metal alloys held together by a cryogenic epoxy matrix. The bonding of the entire regenerator into a monolithic yet porous bed makes it highly stable and resistant to mechanical and thermal stresses. In addition, the high cost associated with the production of spherical particles has been eliminated. To the author's knowledge, this is the first time that such a technique has been used to achieve sub liquid helium temperatures with a two­ stage GM cryocooler.

Lastly, a general purpose, cryogen free materials characterization system was conceptually designed as a potential application for the modified GM cryocooler. The modularity implemented in this design makes it possible to exchange material samples or entire experimental inserts without interrupting the operation of the superconducting magnet system. The conceptual design of an experimental insert to measure heat capacity has also been included as part of this work. The temperature range covered by this insert ranges from 4 to 300 K with magnetic field intensities up to 8 T.