REVTEX 3.1 Released September 2 Development of a Vibration Measurement Method for Cryocoolers Takayuki Tomaru, Toshikazu Suzuki, Tomiyoshi Haruyama, Takakazu Shintomi, Akira Yamamoto High Energy Accelerator Research Organization (KEK), 1-1 Oho, Tsukuba, Ibaraki, 35-81, Japan Takashi Uchiyama, Shinji Miyoki Institute for Cosmic Ray Research, University of Tokyo, 5-1-5 Kashiwanoha, Kashiwa, Chiba, 277-8582, Japan Tsutomu Shimonosono, Yasumi Ohtani, Toru Kuriyama Toshiba Corporation, 2-4, Suehiro, Tsurumi, Yokohama, 23-45, Japan Tomohiro Koyama, Rui Li Sumitomo Heavy Industries Ltd., 2-1-1 Yato, Nishitokyo, Tokyo 188-8585, Japan A vibration measurement method for cryocoolers has been developed. The method can measure the vibration of the cold stage at the thermally steady state after cool-down as well as the measurement at room temperature. It can also separately monitor not only the motion of the cold stage, but also the acceleration of the whole cold head caused by inertial forces of the cryocooler. From the result of vibration measurements for a pulse tube cryocooler, we found that the vibration of the cryocooler has different profiles, at the steady state operation after the cool-down from that at the beginning of the cool-down at room temperature. This result showed an advantage of the method. I. INTRODUCTION A well-developed Gifford-McMahon (GM) cryocooler is widely used in various fields of science and industries because of its convenient handling. However, since a GM cryocooler has a large vibration due to the displacer motion, there is a demand for vibration-free cryocoolers in many fields, such as a cryogenic interferometric gravitational wave detector [1], a superconducting mixer for submillimeter waves, a magnetic resonant imaging device and an electron microscope. A pulse tube (PT) cryocooler is expected to be quieter than a conventional GM cryocooler, because a PT cryocooler has no mechanical displacer in the cold head. However, there are a few reports of vibration evaluations for cryocoolers [2,3]. Most of these evaluations were performed in an early stage of the cool-down process of cryocoolers, far from a steady state. Therefore, we developed a vibration measurement method in which it is possible to measure the vibrations of the cold stage and the whole cold head separately at the thermally steady state condition of the cryocooler. This method can evaluate the vibration coming from different vibration sources in a more realistic situation during its steady state operation. We report on the vibration measurement method for the cryocoolers in this paper. Corresponding author. E-mail: tomaru@post.kek.jp 7-1 c 23 American Institute of Physics
REVTEX 3.1 Released September 2 II. MEASUREMENT METHOD Figs. 1 and 2 show the developed vibration measurement method for cryocoolers. A cold head of the cryocooler was set on a top flange of a vacuum chamber. The size of the vacuum chamber was about φ45 mm in diameter and 67 mm in length, its weight was 1 kg, and there was no radiation shield in the chamber for simplicity. It was set on small rubber sheets to isolate the vibration coming through the ground. The rotary valve unit and the cold head were connected by a rigid copper-tube of 4 m, and the tube was mechanically anchored onto a block of 24 kg at about 1 m apart from the cold head to reduce the transmitted vibration from the valve unit and the compressor directly. The compressor was set in the next room, and a steel door was partitioned between the measurement apparatus and the compressor to reduce the effect of sounds. The vibration of the cold stage and the whole cold head should be measured separately, because their vibration source is different. As shown in Fig. 1 (a), the vibration of the cold stage was measured by an optical displacementsensor and the vibration of the whole cold head (vacuum chamber) was measured by an accelerometer. To measure the vibration of the cold stage in a high-frequency range, the resonant frequency of the table on which the optical sensor was set had to be sufficiently high. This table consisted of four stainless-steel rods in 38 mm diameter and in 5 mm length, and an aluminium plate in 3 mm diameter and in 2 mm thickness. The resonant frequency of the table was 21 Hz for the vibration measurement as given in section III. Fig. 1 (b) shows a picture of the PT cold head and the measurement apparatus. Fig. 3 (a) shows a schematic diagram of the optical sensor. The light emitted from a LED is reflected by a reflector attached under the cold stage, and detected by photo-detectors. The vibration of the cold stage is detected as the change of the light power corresponding to the change of the distance between the sensor and the reflector. The LED is a high-power GaAlAs diode with a peak wavelength of light of 89 nm, and the photo-detector is two InGaAs PIN photo-diodes. This optical sensor can work even at liquid-helium temperature. 1 The reason that two photo-detectors were used was to reduce the effect of the tilt between the sensor and the reflector. The cut-off frequency of the pre-amplifier of the sensor was 1.6 khz. The reflector was made of a polished aluminum block (75 g). The noises of the sensor mainly came from the line noise, the intensity noise of the LED and the noise of the pre-amplifier for the 1 Finally, the sensor was hardly cooled in this measurement, since no radiation shield was used. 7-2
REVTEX 3.1 Released September 2 photo-detectors. Since the thermal contraction and tilt due to the thermal stress of the cold stage occurred at cryogenic temperature, and also the reflectivity of the reflector changes by absorption of residual gas, the sensor had to be calibrated under this situation. Therefore, we set the optical sensor on a motor-driven X-stage with an encoder, and calibrated the sensor just before each measurement. Fig. 3 (b) show the relationship between the displacement and the output signal of the optical sensor. The sensing position of the sensor was chosen to be at about 5 mm from the reflector, since this position has an almost linear response. The conversion coefficients from the sensor-output to the displacement were calculated from the gradient of the graph. We adopted the accelerometer to measure the vibration of the vacuum chamber, since it needs no reference of position for the measurement. A commercial laser accelerometer 2 was set on the top flange of the vacuum chamber. The resonant frequency of the accelerometer was about 8 Hz. Since the accelerometer was set on outside of the vacuum chamber, the effect of sound transmitted through the air could not be eliminated. Although a principle sensitivity limitation for the acceleration measurement was ground-motion, the pre-amplifier noise of the accelerometer limited the sensitivity in the measurement as given in section III, because a wide dynamic range of the measurement was used. III. EXPERIMENT We measured the vibrations of a 4 K PT cryocooler produced by Sumitomo Heavy Industries Ltd. [5] using this measurement method. In this measurement, we compared the vibration amplitude at the thermally steady state at the beginning of the cool-down. The typical cooling capacity of the cryocooler was about 2 W at 4 K for the first stage and.5 W at 4.2 K for the second stage. The weight of the cold head was 12.4 kg and its length was 625 mm. An air-cooling type compressor of 7 kw was used in this experiment. The filling pressure was 1.7 MPa, and the operating pressure was measured at the outlet of the rotary valve unit. The operating frequency was 1 Hz. The vacuum pumps (a rotary and a turbo pump) were stopped during the measurements. Typical pressure in the chamber was 1 2 Pa. The measurement was performed for a vertical and two horizontal directions. 2 The accelerometer was a LA5 produced by RION Co. Ltd. The system of typical laser accelerometer is described in reference [4]. 7-3
REVTEX 3.1 Released September 2 Fig. 4 is the experimental result for the vertical vibration of the cryocooler. The output of the optical sensor was low-pass filtered at 1 Hz to eliminate the large line noise of 5 Hz. The optical sensor was calibrated just before each measurement. The temperature of the cold stage at the thermally steady state was 32 K, which is higher than the nominal temperature, since no radiation shield was used. We found that the vibrations of the cold stage, the chamber and the pressure at the thermally steady state were smaller than those at the beginning of the cool-down. Also, we found that the vibration of the cold stage at the beginning of the cool-down was very complicated, as shown typically in Fig. 5. Since the pressure oscillation had a different profile from the cold stage vibration, the complicated motion of the cold stage can come from the thermal stress of the stage. On the contrary, the vibrations in the thermally steady state were stable. The vibration amplitude of the cold stage was 12 µm in peak to peak, and the acceleration of the chamber was below.2m/s 2. Although there was a vertical-horizontal coupling, the horizontal vibration of the cold stage was smaller than that in the vertical direction. The vibrations of the vacuum chamber were almost same for all directions. IV. CONCLUSION We developed a vibration measurement method for the cryocoolers. This method can measure the vibrations of the cold stage and the whole cold head separately at the thermally steady state of the cryocooler with calibration of the optical sensor. By using this method, we found that the vibration of the cryocooler has different profiles, at the steady state operation after the cool-down from that at the beginning of the cool-down at room temperature. ACKNOWLEDGMENTS This study was supported by the Joint Research and Development Program of KEK and by a Grant-in-Aid prepared by Ministry of Education, Science, Sports and Culture. We express our appreciation to Dr. R. Takahashi at National Astronomical Observatory of Japan for useful advice during our measurement. [1] K. Kuroda et al., Int. J. Mod. Phys. D 8 (1999) 557. [2] C. Wang et al., Adv. Cryo. Eng. 47 (22) 641. [3] R. Lienerth et al., Proceedings of ICEC 18, (2) 555. [4] A. Araya et al., Rev. Sci. Instrum. 64 (5), (1993) 1337. [5] M.Y.Xu et al., Cryocoolers 12, Proceedings of the 12th International Cryocooler Conference, (23) 31. 7-4
REVTEX 3.1 Released September 2 FIG. 1. Vibration measurement apparatus for the cryocoolers. An optical displacement sensor monitors the motion of the cold stage and an accelerometer measures the acceleration of the whole cold head (vacuum chamber), separately. The optical sensor is calibrated by moving with a motor-driven X-stage. FIG. 2. Whole setup of the vibration measurement apparatus. The rotary valve unit and the cold head were connected by a rigid copper-tube. The tube was mechanically anchored onto a block of 24 kg. The compressor was set in the next room, and a steel door was partitioned between the measurement apparatus and the compressor to reduce sound effect. 7-5
REVTEX 3.1 Released September 2 FIG. 3. Optical displacement-sensor (reflective type). (a): schematic diagram of the sensor, (b): output signal of the sensor versus displacement. (a) 2 Displacement [m] 1-1 -2x1-6 (b) 2 4 6 8 1 12 14 16 Acceleration [m/s 2 ].4.2. -.2 -.4 (c) 1. 2 4 6 8 1 12 14 16 Pressure [MPa].5. -.5-1. 2 4 6 8 FIG. 4. Experimental results of the vibration of the 4 K PT cryocooler for the vertical direction. (a): vibration of the cold stage (displacement), (b): vibration of the chamber (acceleration) and (c): vibration of the pressure. The bright gray lines show noise of the sensor when the cryocooler stopped, the dark gray lines show the vibration at the beginning of the cool-down and the black lines show the vibration at the thermally steady state (32 K). The graph of (c) has no bright gray line of sensor noise. The vibration data of the cold stage at the beginning of the cool-down is a typical value, since its amplitude was not constant as shown in Fig. 5. 1 12 14 16 7-6
REVTEX 3.1 Released September 2 Displacement [m] (a) 15 1 5-5 -1-15x1-6 (b) Acceleration [m/s 2 ].4.2. -.2 -.4 2 4 6 8 (c) 2 4 6 8 Pressure [MPa].5. -.5 2 4 FIG. 5. Long-term data of the vibration of the 4 K PT cryocooler for the vertical direction at the beginning of the cool-down. (a): vibration of the cold stage (displacement), (b): vibration of the chamber (acceleration) and (c): vibration of the pressure. 6 8 7-7