5th Russian-German Conference on Electric Propulsion
Dresden, Germany, September 7-12, 2014
Development and Testing of Field Emission Thrusters at TU Dresden D. Bock1, M. Ebert2, F. Roesler3, M. Koessling4 and M. Tajmar5 Institute of Aerospace Engineering, Technische Universität Dresden, Germany
Abstract The principle of field ionization allows the development of highly miniaturized electric propulsion thrusters that require a minimum of electronics compared to other competing technologies. At TU Dresden, two field emission thrusters are currently being developed. The first concept is a FEEP thruster called NanoFEEP [1] using gallium as propellant that is small enough to be integrated into a CubeSat enabling formation flying and orbit control on a widely available small satellite platform. Also the power requirement is so small that up to 8 thrusters can be fired simultaneously enabling full 3-axis attitude and orbit control with maximum thrusts in the range up to 10 µN. The second concept uses gas as propellant using a porous MEMS silicon chip with carbon nanotubes for field ionization [2]. The thruster requires only a single power supply with a few hundred volts DC and no magnetic field. A first prototype of the MEMS ion thruster demonstrated a thrust level of 150 µN with a chip size of only 25 mm. Both thrusters can be tested on a new torsion thrust balance that promises thrust resolutions down to the Nano-Newton range, which is currently being integrated into a large vacuum chamber suitable for micropropulsion testing and plasma diagnostics. Our paper will review and summarize the latest electric propulsion activities at TU Dresden.
Keywords:
Miniaturized FEEP thruster – MEMS ion thruster – Nano-Newton thrust balance
1. Introduction Small satellites require highly miniaturized propulsion systems, but the efficiency is often compromised during down-scaling resulting in low lifetime, bulky electronics or non-ideal specific impulse. Micro propulsion is also of interest for providing highly controllable thrust down to the µN level for precision attitude- and orbit control (AOC) e.g. applied to formation flying (LISA/NGO, DARWIN, etc.). Here we will propose two different thruster concepts, a highly miniaturized FEEP thruster and a MEMS ion thruster, currently being developed at TU Dresden and facing the downscaling requirements of miniaturized propulsion systems. The first presented concept is a highly miniaturized Field Emission Electric Propulsion (FEEP) thruster [1], called NanoFEEP, which enables redundant 3-axis AOC on small satellites, even on CubeSats. Small satellites like CubeSats are very attractive as they enable universities to launch their own satellites to qualify 1
PhD student MSc student 3 MSc student 4 MSc student 5 Professor and Head of Space Systems Chair, Email:
[email protected] 1 2
5th Russian-German Conference on Electric Propulsion
Dresden, Germany, September 7-12, 2014
new technologies or to use them for educational purposes. Due to their small size, applications especially for science-driven missions are highly limited as they often need precision pointing and orbit manoeuvring skills. In addition, there is no possibility to de-spin the wheels which limits the lifetime of the overall system crucially [1]. Various types of miniaturized propulsion systems have been proposed like μPPT, micro vacuum-arc-thrusters (VATs), micro-resistojets, etc., however, only one propulsion system has been successfully flown aboard a CubeSat up to now [2]. Moreover, thrusters such as PPTs or VATs produce a strong electromagnetic pulse during operation that can interfere with the onboard computer. As opposed to this, FEEP thrusters require only a DC heater and DC high voltage power supply. In addition, the FEEP thruster is a truly proportional thruster with very small impulse bits for high accuracy. It will be shown that up to 8 of these highly miniaturized thrusters can be easily accommodated on a double-unit CubeSat. This would enable 3-axis redundant AOC and formation flying with CubeSats. The second presented concept is a gas field emission thruster based on MEMS technology. This type of thruster uses gas as propellant and a porous MEMS silicon chip with carbon nanotubes (CNTs) for field ionization. Applying circuit board manufacturing technology to propulsion sounds appealing from the viewpoint of reducing mass and volume and on the other hand increasing reliability through thousands of micro-emitters on a chip, however, MEMS also has its limitations mostly with respect to the used materials (mostly silicon) and geometries. Some MEMS propulsion systems have been proposed so far, e.g. a MEMS electrothermal propulsion system (Vaporizing Liquid Microthruster) first built by JPL and later in India [1,2]. Thrust values of about 0.5 mN have been achieved at specific impulses of 100-150 s. Although simple in concept, the performance is similar to a warm-gas thruster and shows a very poor performance for electric propulsion. Here we are suggesting a new MEMS thruster based on field emission that uses gas as propellant to avoid propellant droplet related lifetime issues common to all other field emission thrusters (colloid, FEEP and nanoFET thrusters). This concept also enables to emit both polarities (positive ions and negative electrons) and can eliminate the need for a separate neutralizer. A novel manufacturing approach, the fabrication process and results of performance tests of a first prototype will be presented in the following. To verify operational conditions of the newly developed thrusters, a highly precise thrust balance is needed. We therefore developed a horizontal torsion thrust balance with sub-Nanonewton resolution that allows testing not only the thruster but also the whole CubeSat assembly (max. 12 kg payload weight). The setup and important characteristics of our new balance will be introduced as well.
2. Miniaturized FEEP thruster - NanoFEEP The first presented ion thruster is a highly miniaturized FEEP thruster, called NanoFEEP. The operational principle of a FEEP thruster is illustrated in Fig. 1 [4]. If a high voltage potential is applied between a sharp needle (or capillary) wetted with liquid metal and an extractor electrode, a so-called Taylor cone forms at the needle tip. This Taylor cone is caused by the interplay of the liquid metal’s surface tension and the applied electric field. If the evaporation field strength of about 1010 V/m is reached at the needle tip, the liquid metal is evaporated and ionized. The generated ions are accelerated towards the extractor electrode by the same electric field and exit the thruster with velocities of more than 100 km/s. With this concept thrusts of the range of µN with high specific impulse of several thousand seconds can be achieved what makes FEEP thrusters unique for ultra-precise AOC capabilities. Our miniaturized FEEP thruster design is compatible with both capillary and porous needle designs [5]. Here we will concentrate on the performance of porous needle emitters, as shown in Fig. 2, which promise higher mass efficiencies compared to capillary emitters. Gallium is used as propellant due to its low melting point (30 °C) for low heater power consumption.
2
5th Russian-German Conference on Electric Propulsion
Dresden, Germany, September 7-12, 2014
Fig. 1: Basic principle of Field Emission Electric Propulsion (FEEP), [4]
Fig. 2: NanoFEEP Emitter: Tantalum reservoir including a porous tungsten needle
The porous tungsten needles are manufactured using Micro Powder Injection Molding (µPIM) and are sharpened by electrochemical etching with sodium hydroxide. With this process tip radii of sub µm can be achieved, shown in Fig. 3. The porous needles have an open porosity of approximately 30%. Thus, the liquid metal can be hold at the tip of the needle by capillary forces during emission and a sufficient structural robustness is guaranteed. Additionally, the self-feeding propellant is able to flow inside the porous tungsten needle towards the needle tip and is consequently protected from possible contamination. To fill the pores of the porous needle with the propellant the tantalum reservoir including needle and propellant is heated up in high vacuum. The result of this wetting procedure is shown in Fig. 4.
Fig. 3: Porous tungsten needle (tip radius 0.8µm), sharpened by electrochemical etching
Fig. 4: Porous tungsten needle after wetting
The tantalum reservoir can hold about 42 mm3 of propellant (equivalent to 0.25 g Gallium). This amount of propellant enables operation of several hundred hours up to a few thousand hours, dependent on the provided thrust. The size of the reservoir and consequently the possible operation time can be enhanced without any problem, if needed. Though, the dimensions of the thruster need to be adjusted in that case.
2.1. Module design of NanoFEEP As CubeSats are limited in mass and in available power, miniaturization and a low power demand are the key requirements for the NanoFEEP design. Fig. 5 shows a cut away view of the CAD model of NanoFEEP. The module features labyrinth shielding to avoid electrical short circuits during long term operation. The porous Liquid Metal Ion Source (LMIS), which is at high voltage potential during operation, is electrically isolated from all other components. The emitter and the extractor electrode are electrically connected via thin wires to the thruster board at the bottom. In this way the design of NanoFEEP could be kept very compact (dimensions: 13x21mm, weight per thruster: < 6g). 3
5th Russian-German Conference on Electric Propulsion
Dresden, Germany, September 7-12, 2014
Fig. 6 shows one manufactured NanoFEEP thruster compared to the size of a one Euro coin. Moreover, NanoFEEP provides a modular setup. This makes it possible to easily change any part of the thruster. Extractor electrode
Porous LMIS
Thermal/Electrical insulation Heater assembly
Thruster board
Fig. 5: Cut away view of NanoFEEP CAD model
Fig. 6: Manufactured highly miniaturized NanoFEEP thruster; size comparison to a 1 € coin
To meet the requirement of a low power demand we decided to use Gallium as propellant on NanoFEEP. Due to the low melting point of Gallium (approximately 30°C), a much lower heating power for liquefying the propellant is needed, compared to commonly used propellants like Indium (melting point ~157°C). In addition, thermal simulations were performed during the design phase to optimize the thermal design and to predict the heating power demand assuming an operating temperature of 50°C. The simulations showed a heating power demand of 50 to 90 mW, depending on the satellites structure temperature (simulated range: -10 to +20°C). First heating tests of the thruster module inside the vacuum chamber showed a very good agreement with the simulated predictions. Only a difference of less than 10 mW was observed between simulations and thermal performance tests.
2.2. Performance of NanoFEEP First performance tests have been performed to characterize operation of the NanoFEEP module. Fig. 7 shows NanoFEEP operating inside our vacuum chamber at a pressure of 10-6 mbar. 100 Mass efficiency of NanoFEEP (tip radius 1.4µm) curve fitting
90 80
Mass efficiency [%]
70 60 50 40 y = 135.59x-0.30
30 20 10 0
0
Fig. 7: Operating NanoFEEP thruster
20
40
60
80 100 120 140 160 180 200 220 240 Emitter current [µA]
Fig. 8: Mass efficiency of NanoFEP run with Gallium; Needle tip radius 1.4 µm
Ramp tests up to the maximum emitter current of 250 µA (limited by used high voltage supply) were performed to characterize the NanoFEEP thruster module. This maximum current is approximately equivalent to a thrust of 22 µN. Higher thrusts would be theoretically possible, if a different power supply is used. The current-voltage characteristic of NanoFEEP is shown in Fig. 9 and the mass efficiency as a 4
5th Russian-German Conference on Electric Propulsion
Dresden, Germany, September 7-12, 2014
8000
9000
7000
8000 7000
6000
Specific Impulse [s]
Emitter Voltage [V]
function of emission current is illustrated in Fig. 8. Mass efficiency was determined by measuring the change in weight during several tests at different constant emitter currents. After determining mass efficiency as a function of emitter current the specific impulse as a function of thrust could be calculated, shown in Fig. 10. A constant beam divergence factor of 0.8 was assumed throughout the whole thrust range.
5000 4000 3000 2000
6000 5000 4000 3000 2000
1000
1000
0
0
0
50
100 150 Emitter Current [µA]
200
250
0
Fig. 9: Current-Voltage Characteristic of NanoFEEP
5
10 Thrust [µN]
15
20
Fig. 10: Specific Impulse over thrust range
2.3. Potential CubeSat configuration Fig. 11 shows schematically potential thruster arrangements for different AOC configurations of a 2-unit CubeSat. By using four thrusters (1 – 4) in the back of the satellite, 2-axis control (around x- and y-axis) and orbit control (along z-axis) is possible. The minimum configuration for 3-axis attitude control is obtained, if two more thrusters (5 and 6) are added to the front of the satellite. Using all marked thrusters (1 - 8) a configuration with redundancy is possible. With that configuration the failure of any single thruster could be compensated by the remaining seven thrusters. 6
2 1 x
5 Flight direction
z
8
y 4 3
7
Thrusters
Fig. 11: Potential thruster arrangements: 1 - 4: orbit control (thrust) + 2-axis attitude control 1 - 6: 3-axis attitude control, minimum configuration 1 - 8: 3-axis attitude control with redundancy
An exemplary integration of eight NanoFEEP thrusters (redundant 3-axis thruster configuration) is shown in Fig. 12 with a CAD model of TU Dresden’s 2U CubeSat SOMP2. As it can be seen, the compact thruster design of NanoFEEP allows space-saving thruster integration in the corners of a CubeSat. One of the two illustrated main boards in Fig. 12 is able to provide space for the electronics of four NanoFEEP thrusters. The main board location can be chosen freely, but would be most reasonable near the related thrusters. Fig. 13 gives a detailed view of the integration of the thrusters in the back of CubeSat (analogous to thruster 1-4 in Fig. 11). It can be seen that the thrusters are able to be integrated partially into the officially required CubeSat rails [7]. Moreover, due to the protruding thrusters the solar arrays are protected against 5
5th Russian-German Conference on Electric Propulsion
Dresden, Germany, September 7-12, 2014
possible contamination with the metallic propellant. As the casing design of the NanoFEEP thrusters can be chosen freely other integration options are possible as well.
Fig. 12: CAD model of 8 NanoFEEP thrusters integrated in a 2U CubeSat (SOMP2)
Fig. 13: Detail view of NanoFEEP thrusters partially integrated into CubeSat rails
2.4. Power demand of NanoFEEP As the available power of CubeSats is strongly limited, the power demand of an active AOC system such as it would be with NanoFEEP is a crucial issue. According to [6] the average orbit power of a 2U CubeSat is about 3 W. Due to the optimized thermal design of NanoFEEP and the use of Gallium as propellant with its low melting point, the power demand for keeping the propellant liquid could be minimized. In addition to heating, the power demand for ion beam generation and power consumption of the electronics needs to be considered to estimate the total power demand. Based on first power measurements of the real breadboard electronics the total power demand including heating (50 mW), beam generation and electronic efficiencies were calculated, shown as a function of thrust in Fig. 14. The standby power (no thrust generation) is 90 mW, which is dominated by the heater power demand. The power consumption of some ICs is independent of the applied thrust (e. g. microcontroller and power monitors). However, the predominant part of the power demand depends on the desired thrust. The most critical factor is the efficiency of the DC/DC-converter which significantly decreases for small output power (down to only 20 %). More efficient converters may significantly reduce the present power budget which will be investigated in the near future. In summary, the estimated power values in Fig. 14 show that our miniaturized NanoFEEP thruster can be used on CubeSats for attitude and orbit control. 2,5
Table 1: Requirements of different thruster arrangements (one thruster active, others in stand-by)
Power [W]
2
Number of thrusters
1,5
System weight [g]
1
Power [W] (1.5 µN thrust) Power [W] (7.5 µN thrust)
0,5 0 0
2,5
5
7,5 10 12,5 Thrust [µN]
15
17,5
20
4
6
8
150
250
300
0.98
1.15
1.31
1.53
1.7
1.86
Fig. 14: Total realistic power demand of one thruster
Table 1 shows the mass, volume and weight requirements of different thruster arrangements. The system weight includes the weight of the electronics, the thrusters and the cables needed for the stated amount of thrusters. For the power demand it was assumed that one thruster is operating and the remaining thrusters are 6
5th Russian-German Conference on Electric Propulsion
Dresden, Germany, September 7-12, 2014
in standby mode. This shall simulate an active attitude control. In stand-by mode, a thruster can be switched on/off within a µs (only limited by the capacity of the electronics). Table 2 summarizes all important characteristics of our NanoFEEP thruster: Table 2: Overview of NanoFEEP characteristics ~13 mm L ~ 21 mm Weight ~ 6g
Thruster Size Total mass for 4 thrusters (including electronics)
Delta-v on a 2U CubeSat (4 thrusters)
150 g
~ 30 m/s * 0.05 – 22 µN
Thrust
Specific Impulse
~ 6,000 s *
Propellant
Operating Time
~ 1,800 h *
Feeding
Self Feeding
Beam Power-toThrust Ratio
60 W/mN* – 90 W/mN**
Total Impulse (4 thrusters)
~ 60 Ns *
Gallium
*
At average thrust of 1-2 µN ** At max. thrust of 22 µN
3. MEMS Gas Field Ionization Source (GFIS) thruster concept The second field emission thruster which we are currently developing is a gas thruster using a porous MEMS silicon chip with carbon nanotubes (CNTs) for field emission. The gas is flowing along a CNT forest opposite of an extractor grid which generates the high electric field for ionization (Fig. 15). In fact, so-called gas field ion sources (GFIS) are a relatively new research topic and are becoming more and more important as they are already replacing Liquid-Metal-Ion Sources in several applications. A Xenon-based GFIS would be a hybrid between a classical ion and field emission thruster without the use of magnetic fields and liquid propellants. Such a concept looks very appealing to be investigated for MEMS electric propulsion as it avoids the usual small electrode clogging lifetime issue completely. The lifetime critical factor in this design will be similar to the classical ion engines: due to ionization inefficiencies, neutral gas atoms will collide with fast beam ions and produce slow chargeexchange ions that can flow back and cause sputtering of grids and tips. A multi-grid structure with low voltages at the top is necessary to limit the backflow energies to a value below the sputter threshold. This has already been done for electron field emission sources. Another advantage is the fact that the CNT chip can emit both polarities (positive ions and negative electrons) depending on the polarity of the power supply which can eliminate the need for a separate neutralizer.
Fig. 15: MEMS Gas Field Ionization Source (GFIS) Thruster Concept
One of the main research areas will be to obtain good ionization efficiencies. Some GFIS already use large carbon nanotube (CNT) arrays on their cathode through which the gas is flowing [8,9]. Using Xenon, ionization efficiencies of up to 30% have been reported. This is about half of the value from conventional ion 7
5th Russian-German Conference on Electric Propulsion
Dresden, Germany, September 7-12, 2014
thrusters and still leaves a lot of room for improvement. With extraction voltages of a few hundred Volts up to 1 kV, the specific impulse will be several thousands of seconds in such a device. As an example a MEMS GFIS was recently developed at MIT using MEMS manufacturing and carbon nanotubes as field ionizers [10]. Here the anode was a porous silicon wafer working as a flow through electrode with CNTs deposited on the surface. Though a maximum ion current of only 7.3 nA was achieved, the possibility of creating such a device was proven and the required voltage of less than 300V proved to be a step in the right direction. However, the CNT homogeneity was poor also corresponding to a rather poor ionization efficiency. Following these results, a master thesis by T. C. Hicks conducted at the NASA Ames Research Center investigated the feasibility of gas field ion sources for space propulsion applications [11]. Using a different etching procedure and attempting to generate a patterned CNT-growth, Hicks’s goal was to improve the CNT layer at the cost of hole quantity. But continuous breaking of the etching mask using the new approach made the generation of a flow through electrode complicated. For this reason and due to lack of time the fabrication of a flow through electrode (gas passing through the wafer/CNTs) was abandoned and the focus shifted on the fabrication and testing of flow past electrodes (no porous wafer, gas flows along CNTs). Ion currents of up to 100µA as well as electron currents of up to 400µA were successfully generated. Though the flow through electrode fabrication had to be abandoned, very high currents were achieved, as well as valuable knowledge of electrode life times collected.
3.1. Novel manufacturing approach Reviewing the newly introduced designs, several aspects showed further potential for significant improvement in performance of the ion source. The first aspect of our investigation was the carbon nanotube layer. As field ionization only occurs in the strong electric field region close to the CNT-tips, a high CNTcount and thus density is generally more advantageous. Though sparse CNT-growth improves the field enhancing properties and thus reduces the required voltage, it also limits the amount of generated ion current, as in the case of Velásquez-García [10]. This led to the goal to manufacture a flow through electrode with a denser and more uniform CNT-layer, similar to the results from Hicks’s flow past electrodes [11]. Another aspect was the shape and fabrication of the porous substrate. In the work of Velásquez-García [10] the average hole-diameter was 50µm generated through dry etching. As the hole-diameter drives the distance between the carbon nanotubes, and thus is supposed to be as small as possible for a maximum ion current, the design should have a significant lower hole-diameter. A potential alternative to dry etching of the porous substrate is laser drilling. The usage of laser drilling would enable easier alteration of shape parameters to investigate different arrangements, a less risky fabrication process and has the potential of generating smaller hole-diameters. As the manufacturing of such a structure through laser processing has not yet been investigated, it was chosen as the manufacturing process for the new substrates. To be able to test the ion source in an open and closed architecture assembly, the case design was chosen to be modular, to easily adapt for different test procedures. To achieve these goals, we partnered up with experts in CNT deposition (Fraunhofer-Institut für Keramische Technologien und Systeme IKTS) and laser manufacturing (Laser Institute of the Hochschule Mittweida).
3.2. Fabrication process As base material highly n-doped, 550 µm thick silicon wafers were obtained and the substrate size was set to be 26 mm in diameter with a 20 mm diameter region in the centre for the CNT-layer and the holes. The wafers were first coated by a sapphire, Al2O3 layer and afterwards with a Fe-catalyst layer. These layers are crucial for the CNT-deposition. The Al2O3 improves the overall quality of the deposition result, while the Fecatalyst, consisting of nm-sized Fe-particles, defines the growth site for the carbon nanotubes. 8
5th Russian-German Conference on Electric Propulsion
Dresden, Germany, September 7-12, 2014
We then chose to first perform laser drilling and afterwards CNT deposition in order to not damage CNTs during the production process. To minimize pollution of the coated surface and to minimize the achieved hole-diameter at the exit side, the drilling was performed through the back side of the wafers using a pulsed UV-laser. This was followed by the laser-cutting of the substrates and erosion of the excess layers. The laser drilling achieved hole-diameters as little as 10 µm with a maximum substrate transparency of 19% and a total of 450.000 holes drilled in average per substrate – achieving an aspect ratio of close to 50:1.
Fig. 16: Results of the Laser Drilling and CNT-
Fig. 17: Fabrication Process of the Flow Through
Deposition
Electrode
After the laser processing, the substrates were handed to the IKTS institute where the CNTs were deposited using atmospheric pressure chemical vapour disposition, APCVD. This process involved the placement of the substrates inside an oven, which is filled with a carbonaceous gas. When heating up the oven to temperatures above 700°C, the gas molecules break apart resulting in free atomic carbon. Eventually, the carbon atoms settle around the Fe-particles forming tube like structures, carbon nanotubes. The resulting CNT-layers achieved a maximum CNT-length of 130µm with high density as well as close growth near the holes. Fig. 16 illustrates an angular view of the layer, with a section being whipped of for better vision. The whole fabrication process is shown in Fig. 17. In summary, the flow through electrode manufacturing generated very promising results, having achieved very small hole-diameters as well as high CNT-density and uniformity. In addition to the CNT-electrodes, the Laser Institute also manufactured grid electrodes to serve as cathodes. These were manufactured using similar laser processing parameters. The entire design of the first prototype can be seen in Fig. 18. It consists of a modularly designed PEEK case, copper contact electrodes, a Teflon pipe serving as gas supply and it can be altered for usage as an open and closed architecture ion source.
Fig. 18: Test Module Design and Actual Laboratory Model
3.3. Test results 9
5th Russian-German Conference on Electric Propulsion
Dresden, Germany, September 7-12, 2014
500
500
400
400
Ion Current [µA]
Electron Current [µA]
While all preceding research work had focused on open architecture gas field ion sources, the new ion source was the first device capable of performing as a closed architecture ion source as well. Argon gas was chosen as the initial gas for all tests, because of its availability in the lab. To determine the field enhancement properties of the electrodes using the Fowler-Nordheim model, the first tests performed were electron emission tests at 10-6 mbar. The tests revealed an average field enhancement factor of the electrodes of 4000. The maximum currents achieved were 1mA, but to prevent early damage to the electrodes and the rest of the device, no higher electron currents were attempted, though are possible. The resulting I-V plot of one of the CNT electrodes is shown in Fig. 19.
300
200
300
200
0.06 mln/min 0.3 mln/min
100
100
0.9 mln/min 2.2 mln/min
0 220
240
260
280
300
320
340
360
0 250
380
300
350
400
450
500
550
Voltage [V]
Voltage [V]
Fig. 19: Electron Emission Test Results
Fig. 20: Ion Emission Test Results with Argon at
different Gas Flow Rates By switching the polarity between the electrodes, the CNT electrode serves as an anode for field ionization. The field ionization tests were performed at different flow rates between 0.06 mln/min and 2.2 mln/min. As higher flow rate increases the probability of micro discharges inside the plasma generator, the ion source was operated at low voltages, as the first goal was to prove its functionality. The resulting I-V plot can be seen in Fig. 20. Another aspect of interest was the maximum achievable ionization efficiency. The higher the used voltage, the more gas particles are being field ionized on their passage through the plasma generator. However, a higher voltage also increases the probability of micro discharges. Thus a maximum voltage can be determined below which no discharges appears. This maximum voltage was found to be 740 V at 0.06 mln/min. At this voltage, an ion current of 1.2 mA was achieved, which equals to an ionization efficiency of approximately 28%. The related I-V plot can be seen in Fig. 21 as well as the maximum achievable ion current at the lowest flowrate. 1200
30
1000
25
1800
800 600 400 200 0 300
400
500
600
700
1400
20
1000
15 800
10
Voltage [V]
600 400
5 200
0 300
800
1200
Specific Impulse [s]
Thrust [µN], Efficiency [%]
Ion Current [µA]
1600
0
400
500
600
700
800
Voltage [V]
Fig. 21: Maximum achievable Current with Argon
Fig. 22: Thruster-Related Performance Data of the MEMS GFIS Chip, flow rate: 0.06 mln/min Argon
at 0.06 mln/min
10
5th Russian-German Conference on Electric Propulsion
Dresden, Germany, September 7-12, 2014
The thruster related performance data is shown in Fig. 22 for the lowest flow rate of 0.06 mln/min. A total thrust of close to 30 µN at a specific impulse of 1700 s could be achieved with a single chip active area of 20 mm diameter with this first prototype. The overall efficiency was also 30%. This is similar to the performance of FEEP thrusters but with a gas propellant. Different propellants like Xenon and higher voltages can easily create thrusts in the several hundred µN range similar to other µN ion thruster technologies like the µN-RIT. However this GFIS thruster concept requires only a single DC power supply, no magnetic fields and its thrust is highly controllable from µN to several hundred of µN. This makes it also a very interesting candidate for precise attitude and orbit control applications.
4. Novel thrust balance with nanonewton thrust resolution 4.1. Basic setup and characteristics The here presented balance is designed as a horizontal torsion balance similar to Refs. [8] to [11]. Such balances allow thrust measurements without the influence of gravity as the measured forces are perpendicular to the weight force. As opposed to this, vertical balances or pendulums always depend on gravity. These are using gravity either by measuring directly the weight force or by using gravity as a reacting force. By using a horizontal balance though, gravitational influences on the measurement result are eliminated. Our balance can handle very small forces with high precision at the same time. The intended measuring range spans from sub-nanonewton (
