Page: 1 of 6 Conference naam International Cryocooler Conference 12 Conference year 2002 Title of paper T. Trollier, A. Ravex and P. Crespi(1) J. Mullié, P. Bruins and T. Benschop (2) (1) Air Liquide Advanced Technology Division, AL/DTA Authors of paper Sassenage, France (2) THALES Cryogenics B.V. Eindhoven, The Netherlands ABSTRACT A high capacity Stirling cryocooler has been demonstrated at Development Model level during the year 2001 under AL/DTA and THALES Cryogenics co-funding. This development is based on a commerciallyoff-the-shelf LSF9320 type cryocooler from THALES Cryogenics (flexure bearing compressor and a standard wearing Stirling cold finger). It is now featuring a dual opposed piston compressor modified in order to drive pneumatically a Stirling cold finger also implementing flexure bearing technology. The pneumatically driven cold finger does not use any motor to obtain the movement and correct phase shift between the Stirling displacer and the pressure wave. The absence of this motor enhances the reliability of the system and simplifies the electronic control required to drive the system. This reliable and powerful cooler concept has been selected as the cooling system for the ESA / CRYOSYSTEM vial freezers to be delivered by AL/DTA to ASTRIUM for use on board the International Space Station in 2006. The CRYOSYSTEM is a set of facilities for ultra-rapid cooling, preservation and storage of biological samples and protein crystals at -180 C. The actual performances are presented for various water heat sinking locations taking into account the benefit of the Medium Temperature Loop (MTL) available on-board. Future performance improvements are discussed. INTRODUCTION AL/DTA has been selected by ASTRIUM in February 2002 for the delivery of the vial freezers for the ESA / CRYOSYSTEM project. The CRYOSYSTEM is a set of facilities for ultra-rapid cooling, preservation and storage of biological samples and protein crystals at -180 C on-board. Cryocoolers are required to cool down the dewar magazine of the vial freezers as shown in the figure 1 herein. Six Flight Models are needed to support the CRYOSYSTEM vial freezers flow between, Kennedy Space Center and Prime facilities. The delivery of the vial freezers is planned to start in 2006.
Page: 2 of 6 Cryocooler Outer shell Cold Figure 1. Cryosystem vial freezer dewar preliminary design, to be flown on-board. The figure 2 below shows the present day flexure bearing Stirling cryocooler. This stainless steal version weighs 7,5 kg. The outer diameter of the compressor halves is 90 mm and the total length is approximately 200 mm. The compressor design is built around a moving-magnet linear motor that drives the pistons in dual opposed configuration into the same compression chamber 1. The moving magnet linear motor offers big advantages over the conventional moving-coil design. This innovative concept allows the coils that are the main source of gas contamination to be placed outside the working gas. Additional advantages are the absence of flying leads and glass feed-throughs to supply current to the coils. Thus, moving magnet technology is applied in our compressor design to improve the reliability of the complete system. The main disadvantages of this configuration are the losses and the EMI which are higher than in a conventional moving-coil design. High performance axially magnetised NdFeB magnets are used in the motor. Figure 2. Cryosystem cooler Development Model (present day stainless steel version). Flexure-bearings are used in order to have a radial clearance between the piston and the cylinder. These flexure-bearings are round discs made of spring steel, with 3 arms. With this kind of flexure bearings, a
Page: 3 of 6 very high radial stiffness can be reached. By changing the shape, the length and the thickness of the arm, the ratio between the axial and the radial stiffness can be changed without increasing the maximum stresses in the flexures. The fatigue limit of the spring steel is 800N/mm 2. To have enough safety margin the design limit for the VonMisses stresses is set to 600N/mm 2 as presented in the figure 3 below. Flexure arm Figure 3. Flexures used in the compressor and the cold finger design. CRYOCOOLER PERFORMANCES Water Cooled Heat Sinks In this paragraph, we report some experiments that have performed on the water heat sinking of the cooler at various dissipation locations. Some aluminium brackets with internal water flow have been implemented alternatively around the compressor centre part, or around each coil and simultaneously around the centre part and the coils. The cold finger was equipped with a water cooling bracket at the heat rejection path of the warm end as depicted schematically on the figure 4 below. The experimental results are shown in the figure 5 hereafter. Coils Centre part Warm end Heat rejection path Figure 4. Critical heat sinks locations of the compressor and the cold finger.
Page: 4 of 6 7,5 Compressor Heat Sinks Locations Centre Part 7,0 Coils Centre Part + Coils Applied Load @ 75 K, W 6,5 6,0 5,5 5,0 100 110 120 130 140 150 Elec. Input Power, Wac Figure 5. Cryosystem cooler performances for various heat sinks locations. As shown in figure 5, the heat sinking of the centre part of the compressor alone is much more efficient than the heat sinking of the coils so that the heat sinking of the entire compressor casing (centre part + coils) is not necessary to provide very good performances. This will allow for implementing the mechanical brackets at the coil locations and for integrating the water circuit internally to the stainless steel centre part of the compressor. In the figure 6 below, we plot the load curves for various electrical input powers. This informations are required to support the cool down simulations of the vial freezers. At maximal input power of 150 W, a slope of about 180 mw/k is reached which allow for 7,3 W at 75 K or 10 W at 90 K. With half of maximal input power during start up, the cooler is expected to provide more than 30 W of cooling to the cold magazine. 18 16 14 12 Applied load, W 10 8 6 4 2 Pinput (Watts) 50 75 100 125 150 0 50 100 150 200 Cold tip temperature, K Figure 6. Cryosystem cooler load curves for various electrical input powers.
Page: 5 of 6 Compressor skin and internal coil temperatures During the cooling power measurements presented above, the skin temperatures were measured at different places on the compressor case with thermocouples. Heat insulator was put around the compressor coils to avoid additional natural convection and the centre part was surrounded by the water cooled heat sinks. Although the surface temperatures already give a good indication that the central heat sinking is very efficient, it is still interesting to complete the experiment with the measurements of the actual coil temperature in the same conditions. To get an indication of the temperature (T coil ) of a coil during operation, the resistance of the coil has been measured immediately after switch off. Knowing this resistance (R Ta ) at room temperature (Ta) we find the temperature with the following relationship [1]: 1+ α.( T T 0) Rcoil = RTa. [1] 1+ α.( Ta T 0) Where α is the temperature coefficient at the temperature To (for copper α = 0.0038 K -1 for To=20 C). At room temperature (Ta=25.2 C) R Ta =3.316Ω (for a single coil). The calculated coil temperatures for various input powers and for 75 K cold tip temperature operation are also gathered into the table 1. As shown, the temperature of the coils is very close to the temperature of the casing. This demonstrates again that the water cooling of the centre part is very efficient and sufficient. Input Power Cooling power @ 75 K Mean case temperature Coil resistance Coil temperature Pin = 100 W Qe = 5,2 W T case = 29,5 C R coil = 3,382 Ω T coil = 30,5 C Pin = 125 W Qe = 6,3 W T case = 35,6 C R coil = 3,458 Ω T coil = 36,5 C Pin = 150 W Qe = 7,3 W T case = 38 C R coil = 3,500 Ω T coil = 40,1 C Table 1. Temperature measurements for various input powers and for 75 K operation. FUTURE PERFORMANCE IMPROVEMENTS The coming work is dedicated to the reduction of the parasitic heat losses which can have a big impact during the power off on-board. In the present day design, the cold finger is entirely made of stainless steel in order to ease the manufacturing of prototypes. As reported in the table 2 behind, the total parasitic heat losses of the cold finger reach 1400 mw between 75 K and 300 K, which is about one third of the total heat entries to the cold magazine of the freezer. Some simulations are presented herein which consist of replacing stainless steel by high performance titanium alloy and then allow to reduce the thickness of the cold finger and the regenerator tubes. The impact of the length of the regenerator and cold finger tubes is also depicted. Thus, with a length of 80 mm and the use of thin Ti6Al4V tubes, it is expected to reduce the parasitic heat losses of about 70% and to provide 1,4 W extra cooling at 75 K with 80 W mechanical input power (about 125 W electrical input).
Page: 6 of 6 Cold finger version Conduction losses through the tubes Conduction losses through the He gas and the matrix Cooling power @ 75 K for 80 W mechanical input L = 56 mm, SS, δ = 0,2 mm Present day version L = 56 mm, Ti6Al4V, δ = 0,15 mm L = 70 mm, Ti6Al4V, δ = 0,15 mm L = 80 mm, Ti6Al4V, δ = 0,15 mm 1230 mw 170 mw 6,65 W 425 mw 170 mw 7,45 W 340 mw 136 mw 7,92 W 297 mw 119 mw 8,10 W Table 2. Cooler performance improvements with implementation of Ti6Al4V and for various tube thickness (δ) and regenerator/tube length (L). The balance between thermal performance and mechanical resistance is under analysis at dewar level with respect to the induced loads onto the cold finger during launch and landing phases. In this framework, flexible thermal link solutions are presently under design. CONCLUSIONS A compact, reliable and efficient flexure bearing Stirling cooler is under optimisation for use on-board the ISS within the CRYOSYSTEM vial freezers program. The displacer is supported by flexure bearing and is pneumatically driven which enhances the reliability of the system and simplifies the electronic control required to drive the overall system. With traditional materials a cooling capacity of 7,3 W @ 75 K (or 10 W @ 90 K) has been reached with 150 W max input power and with water cooling heat sinks. The future implementation of high performance titanium alloys will allow for a drastic reduction of the parasitic heat losses of the cold finger doubled with an increase of the cooling capacity. REFERENCES 1 M. Meijers, A. A. J. Benschop and J. C. Mullié High Reliability Coolers under development at Signaal-USFA, Cryocoolers 11, Keystone, Kluwer Academic / Plenum Publishers (NY), p 111-118.