Trans. JSASS Aerospace Tech. Japan Vol. 12, No. ists29, pp. Tb_31-Tb_35, 2014 Topics Number Density Measurement of Neutral Particles in a Miniature Microwave Discharge Ion Thruster By Yuto SUGITA 1), Hiroyuki KOIZUMI 2), Hitoshi KUNINAKA 3), Yoshiki YAMAGIWA 1) and Makoto MATSUI 1) 1) Department of Mechanical Engineering, Shizuoka University, Hamamatsu, Japan 2) Department of Aeronautics and Astronautics, The University of Tokyo, Tokyo, Japan 3) Institute of Space and Astronautical Science, JAXA, Sagamihara, Japan (Received June 28th, 2013) Laser absorption spectroscopy (LAS) is a useful technique for sensitive quantitative measurement of plasma parameters. We used LAS to obtain plasma profiles inside the 1 miniature ion thruster developed by the Institute of Space and Astronautical Science of the Japan Aerospace Exploration Agency; this thruster is intended for installation on small spacecraft weighing up to 50 kg. It operates at low microwave power (1 W) and has been developed as a complete system, including a neutralizer. At present, the thrust efficiency of the 1 thruster is not as high as that of standard ion thrusters and optimization is required. A detailed profile of plasma inside the discharge chamber is necessary for further improving performance. We developed an experimental LAS setup with spatial resolution of 0.1 mm for a modified 1 thruster model. The number density distribution of neutral particles in a metastable state was measured and found to be about 10 17 m -3. Key Words: Miniature Ion Thruster, Plasma Measurement, Laser Absorption Spectroscopy Nomenclature A : Einstein coefficient g : degeneracy factor of lower level g : degeneracy factor of higher level I 0 : intensity of incident laser beam I t : intensity of transmitted laser beam k : absorption coefficient l : absorption path length n : number density : laser wavelength : laser frequency 1. Introduction Electric propulsion systems for spacecraft are being developed toward two goals: greater power and smaller size. On the one hand, more powerful systems are required for space exploration and quasi-zenith satellite deployment. On the other hand, smaller systems are required for the microsatellites and nanosatellites expected to be deployed instead of conventional large satellites to reduce launch costs and the risk associated with single-spacecraft missions 1). The 1 ion thruster was developed by the Institute of Space and Astronautical Science of the Japan Aerospace Exploration Agency (ISAS/JAXA). This thruster is feasible for installation on small spacecraft weighing up to 50 kg. A schematic diagram and a photograph of the 1 thruster are shown in Fig. 1. Microwaves with a frequency of 4.2 GHz are introduced into the chamber by a ring-shaped antenna. Two ring-shaped permanent magnets are installed on the bottom of the chamber to generate plasma by electron cyclotron resonance (ECR) heating. A working gas (xenon) is fed through holes in the yoke plate and into the space between the two magnets. The 1 thruster can operate at low microwave power (1 W) and has been developed as a complete system, including a neutralizer. The thruster was named 1 ( mu-1 ), because it utilizes a 1-cm-class beam and a 1-W-class microwave power source 2). However, the performance of the 1 ion thruster is not as high as that of standard or large ion thrusters, which have a long research history (since the 19s) and thus have a mature design, especially the discharge chamber. The thrust efficiency of large thrusters reaches 40 % depending on the power level. In contrast, research on miniature ion thrusters commenced in the 2000s, and the thrust efficiency has not yet reached 20%, leaving ample room for further optimization. In particular, ECR plasma has a characteristic length determined by the microwave wavelength, and so smaller sizes require more sophisticated modeling, specifically tuned for the small scale, as well as plasma sizes different from those in standard thrusters 3). (a) Yoke Gas inlet Microwave Grid system Outer magnet Magnetic field line Antenna Inner magnet Fig. 1. (a) Schematic diagram and (b) photograph of the miniature ion thruster (b) Copyright 2014 by the Japan Society for Aeronautical and Space Sciences and ISTS. All rights reserved. Tb_31
Trans. JSASS Aerospace Tech. Japan Vol. 12, No. ists29 (2014) Our goal is to obtain the plasma profile inside the ion thruster and revise our plasma model to increase the thruster s performance. There are several requirements for the measurement to achieve this goal. First, the measurement must be non-intrusive because the microwave pattern is easily disturbed by the insertion of electrodes. Second, quantitative measurements, particularly of the absolute number density of plasma, are essential for accurate plasma modeling. Third, the small physical size of the thruster requires high-resolution measurements. In this study, our objective was to develop a number density measurement system fulfilling these requirements. Thus, we designed a basic laser absorption spectroscopy setup and a high-resolution 2D scanning system 3) and measured the number density of neutral xenon particles in a metastable state (Xe I 823.16 nm) to demonstrate its effectiveness. visualized model) are shown in Fig. 2. There were three differences between the original 1 ion thruster and the 1 visualized model. The first modification was the replacement of the metal walls with glass walls, which made it possible to transmit a laser beam through the discharge chamber. The second modification concerned the discharge chamber itself, whose shape was changed from cylindrical to cuboid. The third modification was the substitution of the grid in the acceleration mechanism with an orifice plate in the 1 visualized model. The magnet rings and microwave antenna were unchanged. The visualized model was designed for validation of the basic measurement system. Therefore, in this study, we did not consider differences in plasma profile between the original 1 thruster and the visualized model caused by the modification of the discharge chamber. (a) (b) 2. Laser Absorption Spectroscopy Laser absorption spectroscopy (LAS) is a useful technique for sensitive quantitative measurement of the number density and translational temperature of plasma particles. A laser beam with a wavelength adjusted to the energy level of the particles targeted in the measurement is swept around the wavelength and irradiated onto a plasma source, in which the target particles transition to a higher excitation state. Laser absorption occurs as a function of the number density of the target particles. When a laser beam passes through a medium, the relationship between the intensities of the incident and transmitted beams obeys the Beer Lambert law: I t I0 exp kl (1) where l is the absorption path length (plasma length) and k is the absorption coefficient. From Eq. (1), the absorption coefficient is 1 I t k ln. (2) l I 0 The absorption coefficient is a function of frequency, and the relationship between absorption and number density is 2 g' An k d. (3) 8g That is, the number density is obtained by integrating the absorption profile over the frequency. In this study, we measured the number density of neutral xenon particles in a metastable state (Xe I 823.16 nm). Particles in the metastable state have a long lifetime (about 40 s) and high density. Therefore, their absorption coefficient is high and absorption is relatively easy to measure. Neutral xenon particles in a metastable state are thus suitable for demonstrating the basic characteristics of the measurement system 4-6). 3. Experimental Setup 3.1. 1 visualized model A 1 ion thruster was modified to allow the internal structure to be observed directly. A photograph and a schematic illustration of this model (referred to as the 1 Orifice plate Antenna Xe Microwave Ring magnets 20 mm 20 mm Ring magnets Antenna Glass wall 3 mm Fig. 2. Photograph and illustration of the 1 visualized model: views from (a) the left and (b) the front, combined with a schematic diagram of the inner structure. 3.2. LAS setup A schematic diagram of the setup for the LAS experiment is shown in Fig. 3. The infrared single-longitudinal-mode diode laser (OPNEXT 8325G) has a nominal wavelength of 830 nm at 25 C and a typical output power of 40 mw. The laser diode was placed on a thermoelectrically cooled mount (Thorlabs TCLDM9), and its current and temperature were controlled by precision diode laser drivers (Thorlabs LDC202 and TED200C). The wavelength was adjusted toward the target value of 823.16 nm by temperature tuning. Around this wavelength, fine wavelength scanning was conducted by current tuning using triangular waveforms generated by a function generator. One part of the emitted laser beam was transmitted into a wavemeter to measure the beam wavelength, after which it was transmitted into a confocal Fabry-Perot etalon (free spectral range: 0.375 GHz) for frequency calibration of the absorption spectrum. Another part of the emitted laser beam was sent to a photodetector (Thorlabs Tb_32
Y. SUGITA et al.: Number Density Measurement of Neutral Particles in a Miniature Microwave Discharge Ion Thruster PDA8GS) through a xenon discharge tube, and the resulting absorption profile was used as a reference. The rest of the beam was transmitted through the vacuum chamber of the 1 visualized model by a single-mode optical fiber (Thorlabs SM0-5.6-125) and reached a photodetector 3). scanning the wavelength while moving the stage continuously. The signal from the stage controller was recorded together with other experimental data, which were automatically analyzed with a specially developed program to obtain the number density distribution. Fig. 3. Experimental setup for number density measurement in a 1 visualized model by laser absorption spectroscopy. 3.3. Laser probe path The laser probe path in the vacuum chamber of the 1 visualized model is shown in Fig. 4. The single-mode optic fiber had a mode field diameter of 5.6 m and a numerical aperture of 0.12. The other end of the optic fiber was installed inside the vacuum chamber and emitted a beam that was collimated by a molded glass aspheric lens (Thorlabs C150TME-B). The lens had a focal length of 2.0 mm, and the expected diameter of the collimated beam was 0.48 mm. The collimated beam entered the 1 visualized model, and the beam passing through it was transmitted to a multi-mode optic fiber (Thorlabs GIF50). This optic fiber had a core diameter of 50 m and received the beam directly through its end surface. Hence, only the part of the beam which arrived at the core of the multi-mode optic fiber end was propagated. Thus, the maximum resolution of this laser probe system was equivalent to the diameter of the optic fiber core, or 50m. The emitting port of the single-mode fiber, the beam collimation lens, and the detecting port of the multi-mode fiber were all fixed onto a rigid aluminum stage installed on a two-axis liner stage system driven by stepping motors 3). Fig. 4. 1 visualized model and laser probe path. 3.4. 2D number density distribution measurement system The 1 visualized model was fixed on one of the walls of its vacuum chamber. The density distribution in the y-z plane was obtained by moving the two-axis stage relative to the fixed model. The stepping motor used to drive the two-axis stage was controlled by a controller that moved the stage according to a sequence control program. In this way, the number density distribution was measured automatically by Fig. 5. Experimental setup for 2D number density distribution measurement system. 4. Experimental Results 4.1. Absorption profile The experimental conditions are shown in Table 1. The photodetector signals for the beam passing through the 1 visualized model and the xenon discharge tube are shown in Fig. 6. The latter was used as a reference absorption profile, which was clearly different from the profile obtained from the photodetector signal for the beam passing through the 1 visualized model. This is attributed to the effect of the magnetic field in the 1 visualized model. The signal for the xenon discharge tube is almost linear in regions outside the absorption wavelength, whereas the signal for the 1 visualized model is non-linear. The photodetector signal for the 1 visualized model without microwave power input (and hence no plasma) is shown in Fig. 7. The photodetector signal for the 1 visualized model was markedly different from the reference profile, even though there was no plasma in the discharge chamber. This phenomenon is attributed to interference caused by the laser beam passing through the glass wall. To obtain accurate measurement results, we eliminated the effect of the glass wall by conducting the calculation based on the intensity of the transmitted laser beam without microwave power input instead of using the beam incident to the 1 visualized model. Figure 8 shows two plots of the intensity of the transmitted laser beam, where the continuous line represents the intensity of the transmitted laser beam with microwave power input, in which case the signal was affected by both the plasma and the glass wall. In contrast, the intensity of the transmitted laser beam without microwave power input (broken line) was affected by only the glass wall. Therefore, we were able to calculate the absorption coefficient precisely using the difference of the two data sets (Fig. 9). Table 1. Experimental conditions. Microwave power 2 W Xenon flow rate 0.15 sccm Sweep frequency 2 Hz Tb_33
Trans. JSASS Aerospace Tech. Japan Vol. 12, No. ists29 (2014) µ1 visualized model with plasma Xenon discharge tube Fig. 6. Absorption profile for the 1 visualized model and the xenon discharge tube. µ1 visualized model without plasma Xenon discharge tube Fig. 7. Absorption profiles for the 1 visualized model without plasma and the xenon discharge tube. 4.2. Number density distribution in the y-z plane The number density distribution was measured at multiple points using the method described above. The experimental conditions are shown in Table 2. Due to the shorter experiment duration, the sweep frequency was different from that in the preceding experiment. However, this did not affect the results of the experiment. The measurement region was a rectangular section of 3 20 mm, and the measurement interval was 0.1 mm. The number density distribution of metastable Xe I 823.16 nm is shown in Fig. 10, where the measurement region (outlined in red in the left panel) includes the base of the antenna and the downstream wall. The black sections in the right panel correspond to regions where analysis was impossible because the laser beam did not pass thorough the discharge chamber due to collision with the antenna or the wall. Hence, the black sections appear in the shape of the cross-section of the antenna and the wall. Near the antenna, the number density was about 10 16 10 17 m -3. This is a reasonable result based on comparison with previous studies on ECR plasma sources 6). However, near the wall, the number density is nearly zero or even negative. Thus, it is considered that LAS has low sensitivity for regions near walls, where the number density is low. Table 2. Experimental conditions. Microwave power 2 W Xenon flow rate 0.15 sccm Sweep frequency 10 Hz Number of measurement points 6231 µ1 visualized model with plasma µ1 visualized model without plasma Fig. 8. Absorption profiles for the 1 visualized model with and without plasma. The calculation is based on the intensity of the transmitted laser beam without microwave power input instead of the incident beam to the 1 visualized model. 0.12 1 visualized model Discharge tube 0.36 0.1 0.3 Absorption ratio ( visualized model) 0.08 0.06 0.04 0.02 0.24 0.18 0.12 0.06 Absorption ratio (Discharge tube) Fig. 10. Measurement region and density distribution (Xe I 823.16 nm). 0 0-8 -6-4 -2 0 2 4 6 8-0.02-0.06 Fig. 9. Absorption profile (Xe I 823.16 nm). 5. Conclusion In this study, LAS was used to analyze the number density Tb_34
Y. SUGITA et al.: Number Density Measurement of Neutral Particles in a Miniature Microwave Discharge Ion Thruster distribution of neutral particles in a metastable state (Xe I 823 nm) in a modified 1 model. Measurements were taken at over 00 points with a 0.1 mm spatial resolution. Near the antenna, the obtained number density was found to be reasonable when compared with the value obtained in related studies on ECR plasma sources. Acknowledgments This study was supported by a Grant-in-Aid for Young Scientists (A), No. 2268, provided by the Ministry of Education, Culture, Sports, Science and Technology, Japan. References 2) Koizumi, H. and Kuninaka, H.: Performance of the Miniature and Low Power Microwave Discharge Ion Engine, Proceeding of Joint Propulsion Conference, AIAA-2010-6617, 2010. 3) Koizumi, H., Nishiyama, K. and Kuninaka, H.: Study of Frequency Modulation Laser Absorption Spectroscopy for the Plasma Measurement of a Miniature Microwave Discharge Ion Thruster, 28 th International Symposium on Space Technology and Science, 2011-b16, 2011. 4) Matsui, M., Komurasaki, K. and Arakawa, Y.: Laser Absorption Diagnostics of High Temperature Plasma Flow, Institute of Applied Plasma Science, 9 (2001), pp.17-22. 5) Aymar, M. and Coulombe, M.: Theoretical Transition Probabilities and Life Times in Kr I and Xe I Spectra, Atomic Data and Nuclear Data Tables, 21 (1978), pp.537-566. 6) Tsukizaki, R., Koizumi, H., Hosoda, S., Nishiyama, K. and Kuninaka, H.: Numerical Density Measurement of Neutral Particles in Microwave Discharge Ion Engine mu10, Proceedings of Space Transportation Symposium, STEP2010-059, 2010. 1) Nakayama, Y., Funaki, I. and Kuninaka, H.: Sub-Milli-Newton Class Miniature Microwave Ion Thruster, Journal of Propulsion and Power, 23 (2007), pp.495-499. Tb_35