Space Power Resources, Manufacturing and Development Volume 9 Numbers 2/3 1990
SPACE POWER Published under the auspices of the SUNS AT Energy Council EDITOR Andrew Hall Cutler, Space Studies Institute, and the Space Engineering Research Center, University of Arizona ASSOCIATE EDITORS V. I. Auduyevsky, Soviet Academy of Sciences R. A. Binot, European Space Technology Centre Eleanor A. Blakely, Lawrence Berkeley Laboratory Richard Boudreault, Oerlikon Aerospace Lars Broman, Solar Energy Research Centre, Borlange, Sweden William C. Brown, Consultant Gay Canough, Extraterrestrial Materials, Inc. Lucien Deschamps, Electricite de France Ben Finney, University of Hawaii Josef Gitelson, BIOS Project, Krasnoyarsk, USSR Peter E. Glaser, Arthur D. Little, Inc Praveen K. Jain, Northern Telecomm Ltd Mikhail Marov, Soviet Academy of Sciences Gregg Maryniak, Space Studies Institute Michael Mautner, University of Canterbury, New Zealand Rashmi Mayur, Global Futures Network, Bombay Makoto Nagatomo, Institute of Space and Astronautical Science, Tokyo Mark Nelson, Institute of Ecotechnics John R. Page, University of New South Wales Geoffrey Pardoe, Brunel Science Park Tanya Sienko, Association for Research on Japanese Space Development Gerry Webb, Commercial Space Technologies, UK Ray A. Williamson, Office of Technology Assessment/US Congress Andrew R. Wolff, Cleveland, OH Space Power is an international journal for the presentation, discussion and analysis of advanced concepts, initial treatments and ground-breaking basic research on the technical, economic and societal aspects of large-scale, space-based solar power, space resource utilization, space manufacturing, and other areas related to the development and use of space for the long-term benefit of humanity. Papers should be of general and lasting interest and should be written so as to make them accessible to technically educated professionals who may not have worked in the specific area discussed in the paper. Editorial and opinion pieces of approximately one journal page in length will occasionally be considered if they are well argued and pertinent to the content of the journal. Submissions should represent the original work of the authors and should not have appeared elsewhere in substantially the same form. Proposals for review papers are encouraged and will be considered by the Editor on an individual basis. Editorial Correspondence: Dr Andrew Hall Cutler can be reached by telephone at (602) 322-2997, by Facsimile at (602) 326-0938 and by mail at 4717 East Fort Lowell, Tucson, AZ 85712, USA. Dr Cutler should be consulted to discuss the appropriateness of a given paper or topic for publication in the journal, or to submit papers to it. Questions and suggestions about editorial policy, scope and criteria should initially be directed to him, although they may be passed on to an Associate Editor. Details concerning the preparation and submission of manuscripts can be found on the inside back cover of each issue. Business correspondence, including orders and remittances for subscriptions, advertisements, back numbers and offprints, should be addressed to the publisher: Carfax Publishing Company, PO Box 25, Abingdon, Oxfordshire 0X14 3UE, UK. The journal is published in four issues which constitute one volume. An annual index and titlepage is bound in the December issue. ISSN 0883-6272 © 1990, SUNS AT Energy Council
SPACE POWER Volume 9 Numbers 2/3 1990 Daniel E. Hastings & Mengu Cho. Ion Drag on a Highly Negatively Biased Solar Array 99 Hiroshi Matsumoto, Nobuyuki Kaya & Makoto Nagatomo. Microwave Energy Transmission Experiment 113 Kunihisa Eguchi, Sachio Ogiwara & Tsutomo Fujiwara. An Experimental Stirling Engine for Use in Space Solar Dynamic Power Systems: Preliminary Tests 131 Sumio Kato, Hiroshi Oda, Yasuhiro Takeshita, Yoshinori Sakai, Tatsusaburo Nakamura & Osamu Muragishi. Study of Parabolic Solar Concentrators 149 B. D. Meredith, P. R. Ahlf & R. J. Saucillo. Space Station Freedom Growth Power Requirements 163 Krishna Kumar. Satellite Attitude Control through Solar Radiation: a New Approach 175 Karl A. Fay mon, Ira T. Myers & Denis J. Connolly. High Temperature Superconductivity Technology for Advanced Space Power Systems 185 Francis G. Graham. An Indirect Search for Lunar Polar Ices 195 F. Carre, E. Proust, S. Chaudoume, P. Keirle, Z. Tilliette & B. Vrillon. Overview of CNES-CEA Joint Programme on Space Nuclear Brayton Systems 205 Gerald Halpert & Subbarao Surampudi. Rechargeable Lithium Battery Technology: a Survey 221 Jack F. Mondt. SP-100 Power System Development Status 241 Book Reviews 275 Announcements 277
The first international scientific journal for microgravity research - a promising and interdisciplinary forum
Ion Drag on a Highly Negatively Biased Solar Array DANIEL E. HASTINGS & MENGU CHO Summary Highly biased solar arrays are found to have a number of significant interactions with the space environment. The negatively biased parts of the array undergo enhanced ion drag and also suffer from destructive arcing. The enhanced drag suffered by highly biased solar arrays is studied with the PIC code. A new model of the drag is developed and the results are compared to recent experimental work. The drag calculations contain the effect of having the conductor surrounded by dielectrics as well as the charging of the dielectric by electrons. Introduction Future space systems, such as the space station, call for solar arrays to deliver anywhere from 25 kW to several megawatts of power. At these power levels, weight and efficiency constraints will demand solar arrays which operate at several hundred volts. The greatest voltage used so far has been the 100 V used on Skylab [1], and it has been seen that at voltages above this, there is significant interaction between the solar array and the ambient plasma. Solar arrays typically have the interconnectors between the solar cells exposed to the space environment. The solar array is then connected electrically in some fashion to the space system on which it resides. In equilibrium, the net current drawn from the space plasma to the entire space system including the solar arrays must be zero. This is achieved when some part of the array floats positively with respect to space and the rest of the array floats negatively. The positive part of the array attracts electrons and the negative part of the array attracts ions. Beyond a critical negative voltage, solar arrays undergo arcing on their negative ends which can rapidly destroy the solar cells. Conversely, beyond a critical positive voltage, the solar array begins to collect anomalously high currents. This jump in current collection is called ‘snapover’. The parts of the array which are negatively charged will suffer from ion drag. The ion drag for a large three dimensional solar array has been studied numerically and experimentally [2, 3]. In the experimental and numerical work it is shown that for practical high-voltage solar arrays in low earth orbit that the iron drag could be as much as a quarter of the aerodynamic drag on the solar array. This is a significant drag and is much larger than would be estimated by treating the ion flow like a neutral flow. Daniel E. Hastings (Class of 1956 Career Development Associate Professor of Aeronautics and Astronautics, Member AIAA) & Mengu Cho (Research Assistant), Department of Aeronautics and Astronautics, MIT, Cambridge, MA 02139, USA.
However the experimental and numerical model [2] treats the solar array as a conductive plate completely covered with a sheath. The ions then interact with the whole plate and the electrons are completely repelled for large potentials. For real solar arrays where the sheath may not completely 'cover the array or where some parts of the array are positive, these assumptions on the ion and electron interaction with the array will not be valid. This motivates us to use a PIC (particle in cell) code to consider the problem of the ion drag on a model solar array consisting of a conductor surrounded by two dielectrics. The sheath from the conductor does not completely cover the array and the electrons can participate in the charging of the dielectrics. In this paper we discuss ion drag, its simulation by PIC codes and compare it to the results of Kuninaka & Kuriki [2]. Ion Drag The ion drag is considered in a control volume fixed on the charged body. At steadystate there shall be three components of the momentum flux. First, there will be the momentum flux of ions entering the control volume. These ions have momentum defined by the orbital velocity of the charged body. Second, there will be the momentum flux of ions going out of the control volume due to Coulomb scattering with the charged body. These ions have their trajectories modified by the electrostatic potential of the body but do not directly strike the body. Third, there will be the momentum flux of ions which are reflected from the body surface as neutrals and go out of the control volume. We assume all the ions which strike the surface are neutralized and are reflected after some momentum accommodation. Since the first and second components depend on the control volume, we combine these components into one and call them the ram-scattering component. We call the third component the reflection component. Interacting Limit A solar array is made up of interconnected solar cells with the interconnectors being biased and exposed to space. The model of Kuninaka & Kuriki [2] assumes that the electrostatic sheaths from all charged interconnectors overlap. Hence the entire array can be treated (as far as the incoming ions are concerned) as one charged body with a potential drop across the body equal to the potential difference across the array. We shall call this model the interacting limit since the sheaths from the connectors all overlap and interact with each other. In their analysis they neglected the electrons (except that they kept quasineutrality in the far field), took the ions as cold so that the ion flow is supersonic and took the plasma as collisionless. They showed that the force on the array (modelled as a charged plate) depends only on the parameters O = eA0/((l/2)m,72orbltal), ^=Larray/(Forbiul/wPi), P, (7n and a, where A0 is the potential drop across the array, Larray is the length of the array, Mpt is the ion plasma frequency, Forbita) is the orbital velocity of the array, p is the angle of attack measured from the uncharged side of the array and crn(t) are the normal and tangential momentum accommodation coefficients. The dimensionless potential <I> measures the energetic effect of the potential of the array on the ions relative to their incoming kinetic energy. The dimensionless length £ measures the spreading of the sheath relative to the conductor size. When then the sheath does not spread much beyond the array relative to the array length. For the experimental results of Kuninaka & Kuriki showed that the force became independent of This is a
statement of the fact that for these large lengths the force due to deflecting the ions has become small compared to the direct ram and reflected force. In their interacting limit they obtained drag coefficients as high as 10 as well as a significant coefficient of lift. It is easy to see that the interacting limit will give the maximum ion drag on the array since it suggests that the drag on parts of the array where the bias is low will be affected by the highly biased parts. Noninteracting Limit The key assumption of the interacting model is that all the sheaths overlap. However for a typical space station solar cell which is 8 X 8 cm with a 2 mm interconnector the aspect ratio of conductor length to dielectric length is 0.03. Only for very large potentials would one expect that the sheath from the conductor would overlap the dielectric and reach to the next conductor. This motivates a model which assumes that the sheaths do not overlap so that each segment of the array can be treated independently of the other segments. We shall call this the noninteracting limit. The noninteracting limit will give a useful lower bound on the force on a solar array. In the non-interacting limit the force on an array composed of ns segments and with total area Jarray is where d is the unit vector in the drag direction (parallel to the ion flow velocity) and I is the unit vector in the lift direction (perpendicular to the ion flow velocity). The key assumption that leads to the noninteracting limit is that the force on each segment of the array depends only on the potential of the conductor in that segment and not on the potential on the nearby conductors. The physical model is schematically indicated in Fig. 1. The sheath boundary can be defined physically by the condition that efy/\ w,F^rbital = 1. This is the potential at which the kinetic energy of the ions is just equal to their potential energy. As indicated in Fig. 1 the noninteracting limit will be true as long as the sheath boundaries from neighbouring conductors do not overlap and as long as the incoming ion flow can reach a given segment without passing through the sheath of any neighbouring segments. This will put a limit on the angle of attack for which the noninteracting model will be valid since it is clear that particles striking the array at very small angles of attack will pass through the sheaths of several conductors before hitting the array. Therefore for small angles of attack the force on any given segment will depend on the potentials on the other segments. Hence we have to know the ion flow field over the whole solar array which has many conductors and dielectrics.
Fig. 1. Ion Flow at Angle of Attack /I Interacting with Two Segments of a Solar Array. Fig. 2. PIC System Geometry.
In the noninteracting limit and with the following assumptions: (1) the sheath does not spread radially from the conductor; (2) the ions are cold; (3) all segments are identical; (4) the accommodation coefficients for the conductor and dielectric are the same; it is possible to evaluate the drag and lift coefficients analytically. The drag coefficient is then: and the dimensionless conductor area is ac = nsJinterconnect/Jarray. These analytic estimates will give an underestimate of the drag and lift in the noninteracting limit since thermal effects and sheath spreading are ignored. Use of a Particle in Cell Code to Simulate Ion Drag In order to obtain a more realistic estimate of the drag and lift coefficients in the noninteracting limit a 2-dimensional PIC code was used to simulate the detailed physics over the surface of an array segment composed of a conductor surrounded by a dielectric. The PIC code was developed from PDW1[4]. PDW1 is one-dimensional in real space and three-dimensional in velocity space. The geometry of the PIC code is shown in Fig. 2. On the right side of the system is a conductor with a dielectric on either side. This enables the surface charging to be modelled. The boundary condition on the potential over the dielectric plates is E-n = a/e0 where o is the charge on the plate. The potential over the conductor is taken to be the biased potential. The boundary condition over the rest of the boundary is taken to be 0=0 or S0/Sn=O. The computer code allows injection of electrons and ions at the boundaries at an arbitrary angle with respect to the plate. The ion to electron mass ratio was taken as 100, and both particles were injected from the left hand side, top and bottom boundary with Maxwellian distributions consisting of a drift velocity and a thermal velocity. They are given by ©eiectr„n thermal! ®drift: ©ionthermal= 1:0.7:0.07. Secondary electron emission from the dielectric was allowed. The system was taken to be a square of dimension 32 by 32 units or sometimes a rectangle of 32 by 48 units. The spatial grid was also chosen as 32 by 32. The conductor was taken to have a length of 3 units and to have a dielectric on each side of 13 units. The simulations were always started with 4500 electrons and 4500 ions. This choice of particle density gives the grid size as one Debye length.
The simulation voltage on the conductor was chosen so that the parameter d> for the simulation was the same as found in low earth orbit. In low earth orbit a solar array voltage of 200 V in an flow of 5 eV oxygen gives <I> = 40. In order to compare our results with the work of Kuninaka & Kuriki the choice £=48 was made. This corresponds to a array of about 1.6 m in length. The array was modelled as composed of five segments where the potential on segment j is obtained from equation (2). The force on an array of potential drop △ 0 was obtained as the result of the sum of five simulations with simulation voltage (j=l, 2... 5). For each simulation we had Aegmen/( ^orbita/^) = 4.09. Therefore to obtain the length of the array such that £=48 we calculated the force on the dielectric of length (Forbltal/(yp,) tai/^p,)) analytically (ns + 5) and included it in the total force. This approach is justified by the noninteracting limit since any dielectric outside the sheath of a conductor will be bombarded by ions which are only accelerated by the weak potential on the dielectric. This gives the ratio of conductor area to total array area as ac = 0.044. We chose the accommodation coefficients in the same way as in Kuninaka & Kuriki [2] i.e. crn=0.7 cos 0, and crt = 0.6 where 0,- is the incident angle of the ions (except for Fig. 10). Numerical Result Since we divide the 1.6 m array into five segments, each solar cell corresponds to 32 cm by 32 cm with 1.4 cm interconnector. In order to check the validity of the noninteracting assumption in actual scale of solar array, we performed one PIC simulation for a 8 cm X 8 cm solar cell with a 3.5 mm interconnector in addition to the large scale PIC simulations mentioned above. In Fig. 3 we show the potential contours in steady state for a conductor biased to —250 V with a dielectric on either side with an ion flow at P=~90. The e^=s»t,P’Jrbita] contour is seen to reach almost half way out over the dielectric. This indicates that for this voltage, dielectric length and angle of attack the noninteracting limit is the appropriate limit in which to evaluate the force on the array. Conversely, if the lowest voltage on the array is — 250 V then the interacting limit where all the sheaths are overlapping will clearly overestimate the force on the array. In Fig. 4 we show the ion velocity vectors at steady-state for the same conditions as in Fig. 3. This figure shows that the potential contour at which the streaming ions are affected is the j»z, P^rbital contour. The parts of the dielectric which fall outside the projected area of this contour are struck only by ions which stream directly to them and undergo almost no acceleration. In Fig. 5, we show the normalized components of the total force on each biased segment as a function of voltage. The reflection force on the dielectric is small and independent of the voltage on the conductor. This is because the dielectric takes a voltage very close to ground regardless of the conductor voltage. The ram-scattering force is independent of voltage from momentum conservation. The total force is mainly composed of the reflection force on the conductor and the ram-scattering force. The reflection on the conductor is a function of voltage since it is composed of neutralized ions which are emitted from the conductor after they have been accelerated by the potential. For voltages below — 200 V the ram-scattering force dominates while for voltages above this the reflection force dominates. This result was also observed in the work of Kuninaka & Kuriki.
Use of the PIC simulation enables us to estimate the bounds on use of a noninteracting limit. In Fig. 6 we show the sheath dimension normalized to Debye length as a function of conductor voltage for angle of attack of 90°. For voltages up to the maximum considered the sheath does not overlap all the dielectric. For a Debye length of 0.47 cm corresponding to low earth orbit conditions, the sheath size at — 480 V would be 7.0 cm. Hence for the present large cell size 32 cm by 32 cm, the noninteracting assumption would be valid. For a typical scale 8 cm X 8 cm, the noninteracting assumption is valid as shown in Fig. 3. In Fig. 7 the minimum angle of attack allowed for the noninteracting limit to be valid as a function of conductor voltage is shown. This angle is obtained at a given voltage by calculating from the PIC simulations the minimum angle at which ions can approach the dielectric and only pass through the sheath of that conductor. As can be seen from the figure as the voltage and hence sheath increases the range of allowable angles of attack decrease. Figure 8 contains the drag coefficient as a function of voltage for /?= —90. The numerical results of Kuninaka & Kuriki are shown along with the PIC code results. For comparison the analytic results are shown for the cases ac= 1 which corresponds to
a charged plate, ac = 0.044 which corresponds to the PIC simulations and ac = 0 which is the result for a neutral flow to the solar array. The analytic result for ac=l agrees well with Kuninaka & Kuriki for low voltages but deviates for larger voltages as the sheath spreading neglected in the analytic model becomes more important. The analytic result for ac=0.044 is very close to the neutral result. The PIC code results fall in between the analytic results and the interacting limit results. This is because the PIC code results include thermal effects and the effects of sheath spreading away from the conductor. In Fig. 9 we show the drag coefficient as a function of angle of attack for A0 = — 250 V. In addition, we show the numerical result of Kuninaka & Kuriki (ac=l) for the case of only one segment («s=l). In Fig. 9 the interacting limit overestimates the drag at large negative angle of attack, since the whole array is at one potential. As the angle of attack P decreases, cd increases in the PIC result, while it decreases in the interacting limit and the analytical results. In the PIC result the sheath structure is spherical rather than planar as shown in Fig. 3. Therefore, the projected area of the sheath to the plane normal to the ion flow is larger than A sin for small p. On the other hand, in the interacting limit the sheath structure is planar if we neglect the round sheath at the edge of the solar array
Fig. 5. Normalized Force on a Segment as a Function of Voltage for Angle of Attack of —90°. Fig. 6. Sheath Dimension Normalized to Debye Length as a Function of Conductor Voltage for Angle of Attack of —90°.
Fig. 7. Minimum Angle of Attack Allowed for the Noninteracting Limit to be Valid as a Function of Conductor Voltage. for £»1. Therefore, the projected area of the sheath is proportional to A sin p. Since cd is defined by cd=F-d/\piV2o^A sin P, cd increases for small P in the PIC result. This cd~P characteristic can not be found in the interacting limit where all the sheath interacts over the solar array and forms the planar sheath. However, from the limit of validity of the noninteracting limit we would expect that for P<— 45 we have to know the ion flow field over the whole array. In Table I we give the ion drag force relative to the neutral drag force as a function of altitute for two potential drops and for the numerical results of Ref. [2] and the PIC code results. We see that at high altitudes the ion drag force can significantly modify the total force on the array for large array voltages. This is important to understand since while the total force is small, an enhanced drag force acting for a long time will cause large changes in the orbit of the system carrying the solar array.
The results in Table 1 depend on the momentum accommodation coefficient an. In Table I ern=0.7 cos 0, was used for both the neutrals and ions to the conductor and the dielectric. In Fig. 10 we show the ion drag coefficient as a function of the normal accommodation coefficient crn = cr cos 6, of the conductor and the dielectric for P=— 90. As 0 changes only the reflection component changes. At high voltage, the change of o of the conductor gives a larger change than the change of a of the dielectric since the total force is dominated by the reflection on the conductor and the ram-scattering component. At the maximum cr=O (complete reflection), cd is as much as two times that at <7= 1 (complete accommodation). In reality, the accommodation coefficient crn depends on the ion species, surface material, incident energy and incident angle. The data of the momentum accommodation coefficient for O+ are limited. Boring & Humphris [5] showed the momentum accommodation of O+ was independent of the surface material since O+ interacted with the adsorbed gas layer. If we use their data cr=0.8~0.95 for /?= —90, it leads to low cd. But at high voltage the physics of momentum accommodation become complicated because we have to take into account sputtering of the surface material. Conclusion We have examined the enhanced ion drag on a highly negatively biased solar array identified in Ref. [2]. We have shown that the drag found in Ref. [2] is an overestimate of the drag and occurs because all the conductors are taken as interacting. We have formulated a model where the conductors do not interact and examined its predictions and validity with PIC simulations. We find drag coefficients lower than found before
although still larger than analytical predictions. This enables us to estimate the ratio of the ion drag force to neutral drag force on an array as a function of altitude and voltage. ACKNOWLEDGEMENTS The authors would like to acknowledge Dr William Lawson for providing the basis for the PIC code used in this work. The advice and help of Dr Kuninaka is also greatly appreciated. This work was supported by the Air Force Office of Sponsored Research under grant AFOSR-87-0340. REFERENCES [1] Stevens, N.J. (1980) Review of interactions of large space structures with the environment, in: Space Systems and Their Interaction with Earth’s Space Environment (American Institute of Aeronautics and Astronautics). [2] Kuninaka, H. & Kuriki, K. (1987) Numerical analysis of the interaction of a high-voltage solar array with ionospheric plasma. Journal of Spacecraft and Rockets, 24; 512-517. [3] Kuninaka H. & Kuriki, K. (1986) Interference of high voltage solar array with ionospheric plasma, in: Proceedings of the Fifteenth International Symposium on Space Technology and Science. [4] Crystal, T.L. & Kuhn, S. (1985) Particle simulations of the low-a pierce diode, Physics of Fluids, 28, 2116-2124. [5] Boring, J.W. & Humphris, R.R. (1973) Drag coefficient for spheres in free molecular flow in O at satellite velocities. NASA CR-2233.
digitized by thee space studies institute ssi.org
Microwave Energy Transmission Experiment HIROSHI MATSUMOTO, NOBUYUKI KAYA & MAKOTO NAGATOMO Summary A METS (Microwave Energy Transmission in Space) experiment using the future Space Flyer Unit is proposed. Two fundamental areas will be addressed: one is the development of a control system for the microwave beam enabling accurate pointing to the energy receiver; the other problem concerns the study of nonlinear propagation effects of the microwave beam as it passes through space plasmas as well as the effects of the microwave beam on the plasma environment. Introduction Space development has progressed from simple exploration to active utilization of space in such areas as crystal growth, pharmaceuticals and creation/processing of new materials. As this phase increases, power demands are expected to increase exponentially. At the same time, energy demands on Earth will also increase in the next century due to increased pollution and economic growth. The Solar Power Satellite (SPS) was proposed by P. Glaser in 1968 [1] to meet both space-based and earth-based power needs. The concept was examined in detail by NASA and DOE in 1980 [2]. The SPS generates electric power on the order of several hundreds to thousands of megawatts using solar cells of sizeable area, then transmits the generated power via a microwave beam to the receiving station. The establishment of microwave energy transmission technology is the key point. Even in the near future before the SPS is built it may be used for energy transmission in space on a much smaller scale, such as from an intermediate scale energy station to other spacecraft. It is also useful for many other ground-based applications, such as to provide a wireless energy supply to stratospheric aircraft. Since the environment of space is not a simple vacuum but is filled with plasma, propagation characteristics of the microwave beam are not as simple as in vacuum. The high intensity of the (microwave) electric field modifies the plasma medium in such a way that the beam parts into filaments [3-10]. This will be a serious problem for long-distance energy transmission since the propagation path is thus modified. In addition, nonlinear interactions will lead to plasma heating [11] and generation of plasma waves [12]. Such nonlinear interactions have not yet been studied in detail and no quantitative predictions are available. Much experimental as well as theoretical work is necessary. Hiroshi Matsumoto, Radio Atmospheric Science Center, Kyoto University, Uji, Kyoto 611, Japan; Nobuyuki Kaya, Dept, of Instrumentation, Kobe University, Rokkodai, Nada, Kobe 657, Japan; Makoto Nagatomo, Institute of Space and Astronautical Science (ISAS), Sagamihara, Kanagawa 229, Japan.
We propose a Microwave Energy Transmission in Space (METS) experiment using the future Space Flyer Unit (SFU) to establish basic technologies for microwave energy transmission and to examine nonlinear interactions between high power microwaves and the ionospheric plasma. Among all the technologies necessary for microwave energy transmission, control of the microwave beam is the most essential and the most complicated. An SPS system requires very accurate beam-pointing but a relatively limited angle range in scanning. Conversely, space-to-space energy transmission between an orbiting transmitter and orbiting receiver requires a wide scanning angle range with high speed control. Our METS system covers both possibilities with its versatile active-phased transmitting array. The system also uses a retrodirective antenna and computer controls. The retrodirective antenna uses two-tone pilot signals to unambiguously determine the phase of the transmitting wave. Computer control uses a newly developed neural network to identify input beam direction. To assess the effect of the high-power microwaves on the space plasma environment and vice versa, we will carry out an active experiment injecting microwave energy into the plasma. The beam will be focused on a nearby spot in space so that a very intense microwave field is produced which in turn causes a ‘hot spot’ in the plasma. Plasma response in the hot spot will be observed by diagnostic probes extended from the SFU. METS Experimental Objectives Research areas of the METS experiment are as follows: (1) Research and development of microwave power transmission technology: a microwave beam control, a semiconductor transmitter able to handle the high power involved, transmitting array antenna and a high-efficiency rectenna. (2) Study of possible interactions of microwave beam with the ionosphere and the neutral atmosphere. Beam Pointing Control of the energy beam is highly important as the power station should aim accurately at its moving target regardless of any distortion of the transmitting antenna structure. A retrodirective antenna appears to be the best bet to realize the very stringent beam pointing requirement. The retrodirective type of antenna—called a self-phasing system—uses a standard technique to control the main microwave beam using a pilot signal transmitted from the receiving site. Each element of the emitting array transmits an amplified microwave with a phase conjugate to the received pilot signal. Using this phase-conjugate transmission the beam is automatically directed towards the receiving site. It is, however, impossible to use the same frequency for the pilot signal and the retransmitted energy beam because then the pilot signal cannot be discriminated from it. Choosing a slightly different frequency for the pilot signal can solve this isolation problem, yet this induces a pointing error when the microwave beam is scanned at directions different from the normal to the antenna. The pointing error △ 0 is given by
where 9 is the scan angle (0=0 indicates the normal to the antenna) and Aw is the frequency difference between the pilot signal and the retransmitting wave. An exact phase conjugate circuit (PCC) was invented by Chernoff [13] in order to avoid the above pointing error. An n-fold phase ambiguity may occur due to dividing by n in the PCC, but this ambiguity can be removed if the PCC output frequency is multiplied by a multiple of (n —2). This in turn introduces complications in the PCC circuits. One of the main objectives of the METS project has been to develop a new simple PCC with no ambiguity. Microwaves and Plasmas The effects of the microwave on the space plasma environment are not negligible. If the intensity of the microwave were on the order of that used in standard communication, little effect would be expected since the characteristic frequencies of the
ionospheric and magnetospheric plasmas are orders of magnitude lower than the frequency of the microwaves, 2.45 GHz. Unfortunately, the (power) density of microwaves for power transmission is 12 orders of magnitude higher than that for
communication and thus strong nonlinear effects are expected. These effects, such as plasma heating and nonlinear scattering, will cause the excitation of various plasma waves [12]. The nonlinear excitation of plasma waves was experimentally studied in 1983 by the Japanese MINIX experiment (Microwave Ionospheric Nonlinear Interaction experiment) using a sounding rocket [14-17]. Figures 1 and 2 indicate an artist’s concept of MINIX and a power spectrum of the observed plasma wave in the HF frequency band, respectively. Since the phase velocity of these plasma waves is much slower than light speed, the energy of the wave is easily transferred to the ionospheric plasma (particles) through wave-particle interactions. An energy channel from the microwaves to the environmental plasma via plasma wave excitation has thus been created, giving rise to a sizeable modification of the earth’s plasma environment. METS will investigate these nonlinear effects in detail by creating a local hot spot via a very intense microwave field. The microwave beam will be focused about 2-3 m from the SFU; a spot which is then reachable by diagnostic probes mounted on the SFU platform.
METS System The METS system consists of five main components as follows: (1) a microwave transmitter (MWT) with an active-phased-array antenna (PAA); (2) a target satellite (TGS) for the energy transmission test; (3) a diagnostic package (DGP) to observe plasma phenomena excited by the high power microwave; (4) a neutral gas plume (NGP) for an active experiment on ohmic heating by the microwave beam; and (5) a control/data management system (CDMS). The METS configuration on board the SFU is shown in Fig. 3. 1. Microwave Transmitter (MWT) The microwave transmitter system consists of the active PAA, an oscillator of 2.45 GHz, a power supply and a controller. The main component of the transmitter is the circle-shaped PAA (diameter 2.7 m). A possible configuration of the PAA as installed on the SFU is shown in Fig. 4. All the antenna elements (indicated by dots and small circles) are placed on the corners of equilateral triangles. The subarrays (SAR) (shown as hexagons in the figure) are the most elementary components of the MWT. The PAA is composed of 109 SARs, each of which has seven antenna elements.
Center antenna elements (small circles) of the SARs use a phase conjugation circuit (PCC) for retrodirective control, while the other six elements (dots) use only phase shifters. The PCC gives an output of the conjugated phase of the received pilot signal. Each antenna element is a micro-strip antenna connected to the high power amplifier through a digital phase shifter. The HPA consists of seven transmission amplifiers, each with an output of 12 W at maximum.
The PAA consists of four components: module guides, a honeycomb, lower substrates and suspended lines, as shown in Fig. 5. The module guide is a support for the subarrays and is a hexagonal column made of CFRP. Suspended lines are used for the electrical power lines, the local RF signal feeders and the control signal lines. The PAA is reinforced by the honeycomb and the lower substrates, which are in turn connected to each other with an adhesive. The system is designed so as to control the top power of the transmitted microwave with 10 kW at its maximum. Since the SFU solar cells can only generate an electrical power of 1760 W, the microwave transmitter needs a battery system to radiate at 10 kW. Transmission duration in the high power transmmission mode (HPT) is designed to be 0.1 h (6 min). The DOD of the battery is about 82%, although this is restricted by battery weight and the efficiency of the thermal control systems. Using standard switching regulators to regulate electrical power to the HP A would result in an extra weight of over 200 kg—far too high for the SFU mission plan being considered. Power for the HPA will therefore be supplied directly from the battery. This means the effects of input voltage fluctuations on output gain and phase of the HPA must be carefully examined. Effects on the phase of the transmitted wave are especially critical from the point of view of the retrodirective system. On the other hand, electric power supplied to the PCC and PDR cannot be taken directly from the battery but must be regulated by switching regulators due to the PCC/PDR’s severe voltage deviation. A simulated antenna pattern is shown in Fig. 6, where the microwave is transmitted with identical output phases from all antenna elements. The microwave beam can be steered in any direction. The active phase array system also has the capability of focusing microwave energy at a specific point in space by controlling the 4-bit digital phase shifters. Figure 7 shows the distribution of the electric field strengths when the
beam has been focused on a point 20 A above the antenna plane in a direction perpendicular to the antenna normal. A two-tone system has been studied as a possible retrodirective antenna which solves the problem of induced phase error as mentioned in the last section [18], Two pilot signals (frequencies/t + 2A/and/+ A/;/ is the frequency of the energy beam) are used in this two-tone system. Subtraction of one of the pilot signals (/ + 2A_/) from doubled signal of the other (2/+ 2 A/) gives a signal with frequency / and phase of 0, (Fig. 8). 0, is the phase lag of the/ frequency wave along the path difference AL. The merit of this system is that the phase of the transmitting wave can be determined without ambiguity from the two pilot signals. A retrodirective antenna using such a system was described in detail elsewhere [18]. The computer monitors and corrects the direction of the microwave beam by adjusting the individual output phases of the transmitting antennas via commands from the receiving site. Many methods have been proposed to determine correct adjustment; one is a coherent multiple tone technique documented by Chie [19]. We have studied another method using a small neural net corrected by a back-propogation learning procedure. One merit of using a neural network is the high speed of computation due to parallel processing. This is especially suitable for very large array antennas with many elements, i.e., antennas of the kind liable to be found in SPS applications. A neural net consists of an input layer, several internal hidden layers and an output layer (Fig. 9) which are connected with values of weights. The network needs a learning procedure to determine the weights. In the back-propagation learning method, test cases—in the form of patterns—are fed into the system. The weights of the system are then adjusted so as to minimize the difference between actual output patterns and
desired output patterns. This is then repeated until a representative set of test cases has been surveyed. Due to the additive nature of the weights, ‘learning’ one pattern does not wipe out ‘knowledge’ of those already surveyed. For the present case of the beam pointing system, the input layer consists of several units of the microwave detectors (i.e. selected rectenna elements). The number of units in the output layer are determined according to the desired accuracy of the main beam direction. Each unit in the output layer is expected, ideally, to give 1 when the beam hits it, otherwise to give 0. Figure 10 shows the actual output of the five output units of the test system. Weights were determined by learning at five angles of the main beam ( — 20, —10, 0, 10 and 20). Responses of each unit were measured by swinging the direction of the main beam from —20 to 20. To simulate distortion of the antenna structure, random phases were added into the beam. Results show that a neural system can recognize the direction of the input beam quite accurately. 2. Target Satellite (TGS) The target satellite (Fig. 11) is to be used as a receiving target for the energy transmission test. It would be stored in the housing payload unit (PLU) until necessary, then released from the SFU on a tether wire. The TGS is roughly a sphere 30 cm in diameter, and consists of a micro wave receiver (rectenna), transmitter to transmit the pilot signal, and battery. Printed cross dipole antennas with rectifiers are arranged on the surface of the TGS. Six flexible monopole antennas for pilot signal transmission are placed perpendicularly to the other (cross) antennas for omnidirectional transmission of the pilot signal. A block diagram is shown in Fig. 12. The rectenna can measure the antenna pattern of the PAA by scanning the main microwave beam. This tests the tracking ability of the TGS. Since the rectenna can rectify the received microwaves and provide power to the TGS battery, it is only the pilot signal transmitter which needs extra electric power.
Even so, an on-off switch on the pilot signal transmitter is needed to avoid unnecessary power consumption. It is thus desirable to use optical fibre for the tether wire since optical fiber is suitable for data/command communications even in the strong electric fields of the microwaves. If this is not adopted, a simple command receiver is necessary, and a programmable timer is used as a back-up system for on-off control. One other interesting idea—now under study—is to use the high power microwave beam itself as a switch. In such a case, the microwaves would be radiated without the pilot signal before the transmission test. 3. Diagnostic Package (DGP) Ohmic heating and excitation of the plasma through scattering are expected to be the dominant nonlinear interactions. As mentioned in the introductory section, this will be investigated by producing a hot spot in the plasma. The microwave beam will be focused at one point to produce a very strong electric field. Sensors are extended upwards to approach the microwave focal point from the PLU with two rigid extension booms as shown in Fig. 13. Only key data measurements will be transmitted to the ground because of limitations on the transmission speed of the telemeter. All data, however, will be stored in the PLU’s data recorder and later recovered. The diagnostic package consists of six scientific instruments to observe plasma phenomena excited by high power microwaves. Parameters of the plasma and the neutral atmosphere (electron density, electron temperature and pressure) will be measured, as well as plasma waves over a wide frequency range from ELF to HF. The six sensors are a Langmuir probe (LP), an impedance probe (IMP), a density fluctuation detector (DFD), a vacuum meter (VAM), receivers from ELF to HF (VLF and HF) and a microwave detector (MWR). A block diagram of all the sensors is shown in Fig. 14.
The sensors are divided into two groups; one group (DFD, LP, MWR, and VAC) is installed in a case house on the top of the second stage of the deployed boom, while the other group (HF, VLF and IMP) is installed in another case housed on top of the first stage of the boom. The sensors of each case are assembled in one package in order to make it as small as possible. Shield cases are used to protect the sensors from interfering with each other. The antennas for the VLF, the HF and the IMP are extendable and initially restrained by a wire on the surface of the case, which is cut later on. After the METS experiment, the antennas are detached by being cut at the roots of the antennas and thrown away from the case. This is unfortunately unavoidable because their complete retraction (necessary before retrieval of the SFU) would require a complex mechanical system. Neutral Gas Plume (NGP) The frequency of collisions at the altitude of the experiment (about 400 km) between ionospheric electrons and neutral particles will not be large enough to effectively heat the plasma with the microwave. The neutral gas plume (NGP) injects a neutral gas of nitrogen into the ionospheric plasma in order to artificially enhance the collision frequency for ohmic heating. The gas is released via a simple system (blow-down type) which is composed of a gas tank (904 cc), an electrical valve, a flow regulator and a Laval nozzle. The nozzle has an aperture of 0.9 mm in diameter opening out into a 10 degree cone towards the focal point of the microwave. The spherical gas tank and the electrical valve must withstand the gas pressure
(50 atm) under the strong shock and vibration of the launch. Since the nozzle is located on the top surface of the PLU and is to point towards the microwave focal point, modification of the PLU is required for correct installation of the nozzle. METS Experimental Operation The experiment should be carried out according to the operation sequence outlined below: 1. Preparatory Functional Checkout These tasks must be performed right after launch. The following are checked and run: (1) Check of the system which charges the transmitter battery. (2) Test of microwave transmission (MWT) in the low power mode, with the phases of the power amplifiers all fixed by the computer. (3) Testing of the retrodirective microwave receivers by getting waves from other satellites or from the ground.
(4) All the instruments on board the TGS as well as the Diagnostic Package will be turned on except for the high voltage power supply of the DFD. (5) The target satellite is released to a distance of 100 m from the SFU by springs and/or thrusters. Individual antenna elements of the MWT are checked by switching on the HPA. Next, the pilot signal transmitter is turned on, and the MWT starts to examine the retrodirective systems and computer systems used for phase control. After all the preparatory checkouts, the TGS is further released to a distance of 1 km. A TV camera will monitor the release. 2. Microwave Beam Control Test (towards TGS) The attitude of the SFU is oriented so as to keep the solar panels in direct sunlight, and the battery is charged to the maximum. In this section, three different modes will be possible for control of the transmitted microwaves: (1) using the retrodirective system; (2) computer control mode; and (3) mixed mode with both retrodirective and computer systems. The following parameters will be measured while increasing transmitting power: (a) scanning angle range of the transmitting antenna; (b) steering accuracy of the microwave beam control; (c) microwave beam transmission efficiency. A schematic view of the microwave beam control test towards the target satellite is shown in Fig. 15.
3. Microwave Beam Control Test (Ground Directed) Microwave energy transmission to the ground is performed as outlined in Fig. 16. The pilot signal will be received from the ground receiving site. The same parameters will be measured as in the TGS test through remote control. Special attitude control will be required during this operation. 4. Plasma Experiment After the above technical tests, the scientific tests to monitor non-linear plasma interactions will be carried out (Fig. 17). The sensor package on the extension boom will be moved up to a distance of 2 m. The small sensor package will be released from main sensor package to approach the focal point of the microwaves, and the five
antennas for the wave receivers and the impedance probe are extended. The high voltage power supply is turned on for the DFD. The microwaves are concentrated at a focal point of 20 A away from the SFU with phase control under computer guidance. Measurements will be taken while gradually increasing the microwave transmission power in discrete steps. One experiment cycle will take anywhere from a few minutes to a day due to the differing charging times of the battery. The total number of experimental cycles will be at least 50. As mentioned above, sensor antennas will be detached and discarded after this experiment. Both the TGS and the diagnostic packages will be brought back and stored in the FLU. In case of an emergency, they can be completely jettisoned. Summary and Conclusions The objectives of the METS experiment are to develop necessary pointing technology for microwave transmission and to examine nonlinear interactions between high power microwaves and the ionospheric plasma. The METS experimental objectives are: (1) orbital tests of the hardware developed for microwave power transmission in space environments; (2) verification of the pointing technology with transmission tests towards both a target satellite and a receiving site on the ground; (3) accumulating experience in space power system operations where the complete system is composed of several separated elements in orbital motion; (4) finally, an environmental study of the effects on the ionospheric plasma of a very intense microwave field. ACKNOWLEDGEMENTS The authors wish to express their appreciation to Professor I. Kimura, Dr T. Sato and Dr M. Tsutsui of Kyoto University, as well as Mr. S. Miyatake of the Electrocommunication University for their useful discussions. The authors are also grateful to Mr T. Hashizume of Mitsubishi Electric Corp, for his earnest support. REFERENCES [1] Glaser, P.E. (1968) Power from the sun: its future, Science, 162, pp. 857-886. [2] Hanley, G.M. (1980) Satellite Power System (SPS) Concept Definition Study, NASA CR, pp. 3317-3324. [3] Duncan L.M. & Gordon, W.E. (1977) lonosphere/Microwave Beam Interaction Study, Final Report, NASA Contract NASA-15212, Rice University. [4] Duncan, L.M. & Zinn, J. (1978) lonosphere/Microwave Interactions for Solar Power Satellites, Final Report J-10-4306, Los Alamos Scientific Laboratory. [5] Gordon, W.E. & Carlson, H.C., Jr (1974) Arecibo heating experiments, Radio Science, 9, pp. 1041-1047. [6] Perkins, F.W., Oberman, C. & Valeo, E.J. (1974) Parametric instabilities and ionospheric modification, Journal of Geophysical Research, 19, pp. 1478-1496. [7] Perkins, F.W. & Roble, R.G. (1978) Ionospheric heating by radio-waves; predictions for Arecibo and the satellite power station, Journal of Geophysical Research, 83, pp. 1611-1624. [8] Perkins, F.W. & Goldman, M.V. (1981) Self-focusing of radio waves in an underdense ionosphere, Journal of Geophysical Research, 86, pp. 606-608. [9] Thome, G.D. & Perkins, F.W. (1974) Production of ionospheric striations by self-focusing of intense radio waves, Physical Review Letters, 32, pp. 1238-1240. [10] Utlaut, W.F. & Violette, E.J. (1974) A summary of vertical incidence radio observations of ionospheric modification, Radio Science, 9, pp. 895-903.
RkJQdWJsaXNoZXIy MTU5NjU0Mg==