9.2 Development of Devices for Common Use

9.2.1 R&D Goals

The large-scale liquid hydrogen pump requires not only good performance, operability, security but also long term stability in supplying liquid hydrogen. Since liquid hydrogen, which is used as working fluid, is liquid with very low temperature (temperature:20K, density:71kg/m3) as well as very low viscosity (viscosity coefficient: below 1.3^-5Pa・s), the shaft bearing will be exposed to severe lubricating environment. As the shaft bearing, magnetic floating bearing was selected and development started.

Up to 1998, revolution test was conducted with LH₂, and verified up to 26,000 rpm. However, axial control was affected by magnetic hysteresis and became unstable. In 1999, magnet was improved and revolution test was conducted with 36,000 rpm in view. Field through for sensor cable was designed and developed to conduct revolution test of LH₂ and verify its practicability.

9.2.2 Results in FY 1999

9.2.2.1 Design Change

(1) Revision of magnetic bearing (specimen)

Last year, revolution test of LH₂ was conducted and shaft vibration was excessive with over 26,000 rpm. To solve that problem, material and thickness of the magnet for bearing were revised. Changed points are shown in Figure 9.2.2.1-1. As sensor gland developed last year was non-insulated type, there were problems in measurement of field through. It was newly developed by insulating sensor signal gland. Its shape of cross section is shown in Figure 9.2.2.1-2.

(2) Revision of special controller

Up to last year, analog PID control system was adapted to control magnetic bearing. In this year, digital control system controller was designed and manufactured due to easiness in changing control constant and its potential.

9.2.2.2 Tests

LH₂ test was conducted with the use of revised specimen, newly developed digital controller and insulated type field through. The testing device is shown in Figure 9.2.2.2-1.

(1) Room temperature test

After collecting characteristic data of magnetic bearing at room temperature, idling test was conducted.

(2) LH₂ test

After collecting characteristic data of displacement sensor and electromagnet at LH₂ temperature, revolution test was conducted.

9.2.2.3 Test Results

After combining magnetic bearing and controller at room temperature or LH₂ temperature, with rotor shaft lifted off to the center position, input signal was disturbed and frequency characteristics were verified. As a result, it was confirmed that there is no deterioration caused by hysteresis loss was.

In idling test at room temperature, analog controller made revolution up to 30,000. Around the points of 7,000 rpm (the first critical velocity) and 12,000 rpm (the second critical velocity), an increase in shaft vibration was observed. However, the amplitude was curbed to less than 20mmp-p even when it was at the critical velocity, which confirmed good control characteristic. Figure 9.2.2.3-1 shows correlation between the shaft displacement and the number of revolutions.

However, when it was approaching the point of 30,000 rpm, the shaft vibration increased rapidly and became out of control, which led to sudden stop. Before reaching the point, control characteristic was very good. Disturbing factors that cannot be controlled by analog controller exist at more than 30,000 rpm.

After that, LH₂ test was conducted with digital controller. Control amplitude was curbed to less than 20μmp-p, and control characteristic was not changed by very low temperature environment.

Figure 9.2.2.3-2 shows correlation between the pump discharge pressure and the number of revolutions. When it reaches the point of 33,000rpm, discharge pressure of the pump declined suddenly and the number of revolutions rose rapidly. Judging from this phenomena, it is possible that there was malfunction of the pump and that the shaft received excessive external force.

9.2.2.4 Conclusion

In 1999, the liquid hydrogen pump, which was manufactured in 1998, was revised and revolving test was conducted at LH₂ temperature environment. Following results were obtained.

(1) By changing magnetic material and reducing thickness of rotor yoke material, hysteresis loss and current loss can be curbed. It verified that deterioration of frequency characteristics can be prevented in that way.

(2) By changing the method of the controller from analog to digital with the use of DSP, tests were conducted in combination with the specimen. The test results verified the effectiveness of the digital control method.

(3) Even at the shaft critical speed, shaft vibration level could be curbed to less than 20μm p-p.

(4) While the target number of revolution is 36,000 rpm, revolving tests were conducted up to 33,000 rpm. When it crossed the point of 33,000rpm, shaft vibration level increased rapidly. It may have been caused by excessive fluid external force due to malfunction of the pump.

(5) Insulated field through was newly developed, with which combination tests were conducted. With the tests, good control performance was verified.

(6) In order to verify the performance of the pump, with the vacuum vessel filled with water after the LH₂ test, revolving test with GN2 was conducted up to 10,000 rpm. As in the LH₂ test, the targeted pumping function could not be realized. Improvement in next year on is under discussion.

9.2.3 Research Plan for FY 2000

(1) By changing the current magnetic bearing to (magnetic and liquid) bearing, its load carrying capacity and disturbance control performance will be improved, and pumping power at the target number of 36,000 rpm will be achieved.

(2) By launching development of high-pressure small capacity pump, its application to small-scale distributed system will be discussed.