Technology vs. Cost Issues

As in the United States, both Stirling and Gifford-McMahon (GM) cryocoolers are being manufactured and sold in Japan. The panel did not discern any major differences in the technology of these machines in the two countries. However, two issues related to cost might become important:

  1. To the best of the panelists' knowledge, all the compressors used by U.S. companies in their cryocoolers for the civilian market, such as GMs for cryopumps and research, are made in Japan. This should give Japanese companies a competitive advantage of lower price over U.S. companies when such coolers are used extensively in wireless base stations. The cost of the cooler is already a major fraction of the cost of the high temperature superconductor (HTS) wireless subsystem, and it is likely to remain so.
  2. In recent years, the low-temperature limit of the two-stage GM coolers has been extended well below 4 K by use of materials such as Er3Ni in the regenerator. This material has a large heat capacity peak near 10 K due to magnetic ordering. This innovation was made in Japan, and Japan remains the only source of this material in a readily usable form. If such coolers become important for cooling the rapid single flux quantum (RSFQ) chips of future petaflop computers, Japanese manufacturers would again be in a good position to compete on the price of the cooler, as they would manufacture both the compressor and the regenerator material.

At present, however, it appears that the price of refrigerators made in Japan, even when sold there, is unusually high. An average price of, say, a Stirling or GM refrigerator is about 50% higher in Japan than that of a similar machine made in the United States. As this only became apparent to the panel after its visits in Japan, we were unable to determine the reasons for the price difference. However, the higher price is consistent with the prices of many items in Japan, including scientific instrumentation.

Pulse-Tube Refrigerators

It was clear in the WTEC panel's visits that the level of R&D activity on pulse-tube cryocoolers in Japan is higher than in the United States. Already one company, Iwatani (which the panel did not visit), is selling a pulse-tube cooler. Figure 7.1 shows a copy of Iwatani's advertising literature. In the panel's discussions with representatives from Aisin Seiki, they indicated that their company would make a pulse-tube product available later in 1997. At AMTEL, the panel visited the laboratory where development of pulse-tube coolers was ongoing.

In the United States, pulse-tube technology is well advanced in the laboratory of Dr. R. Radebaugh at the National Institute of Standards and Technology (NIST), at TRW, and at other laboratories. Unfortunately, none of the U.S. manufacturers of commercial coolers have shown much interest in this technology, perhaps in part because of ongoing problems in the United States with technology transfer mechanisms. There is some evidence that U.S. manufacturers will soon have to change their outlook. The largest commercial refrigerator product is the cryopump, which currently uses GM coolers at 12 K or so for use on the vacuum systems of silicon fabrication lines. But as the linewidths of Si circuits continue to shrink, the vibration associated with the GM refrigerators is becoming undesirable. Perhaps, in order to retain market share of the cryopump market, the U.S. refrigerator manufacturers will have to make pulse-tubes using the NIST, TRW, or other technology. At present, however, the Japanese companies, with their existing pulse-tube cryocooler products and/or past investment in some years of development, would appear to be in a position to dominate the cryopump market, if and when GM coolers become unacceptable. Using the same technology, they could supply the coolers for HTS applications, if prices were competitive (or preferably much lower) than the GM machines being used at present.

Fig. 7.1. Iwatani's brochure for pulse-tube cryocooler; the new version has a heat lift of 10-11 watts at 77 K and an ultimate temperature of about 50 K (sold in the United States by Kelvin International Corp.).

[Previous Section][Top of Report][Welcome Page][Next Section]

Published: July 1998; WTEC Hyper-Librarian