Electric Propulsion Activities at ONERA

Electric Propulsion Activities at ONERA IEPC-2015-92094 Presented at Joint Conference of 30th International Symposium on Space Technology and Science 34th International Electric Propulsion Conference and 6th Nano-satellite Symposium, Hyogo-Kobe, Japan Denis Packan 1, Paul-Quentin Elias 2, Julien Jarrige 3, Félix Cannat 4, Clément Zaepffel 5, Julien Labaune 6 and Trevor Lafleur 7 ONERA-The French Aerospace Lab, Palaiseau, 91120, France Abstract: ONERA is the French Aerospace Research Center, under the tutelage of the minister of Defense, and has been involved in electric propulsion studies since 1961. It works with contractual funding and internal funding, and develops scientific collaborations with several establishments. Activities cover the whole range of TRL, from fundamental developments at TRL 2 to flight hardware at TRL 9. The recent activities in electric propulsion and the perspectives will be described. I. Introduction NERA (Office National d Etudes et Recherches Aérospatiales) is the French national aerospace research Ocenter. It is a public research establishment, with eight major facilities in France and about 2,000 employees, including 1,500 scientists, engineers and technicians, and 260 doctoral students and post-docs. ONERA was originally created by the French government in 1946. The research efforts performed there are keyed to applications, and whether it has short, medium or long-term goals, it is designed to support the competitiveness and creativity of the aerospace and defense industries. The research carried out at ONERA results in computation codes, methods, tools, technologies, materials and other products and services which are used to design and manufacture everything to do with aerospace. ONERA s funding comes from two sources: 60% from contract research for industry and agencies, and 40% from annual subsidy from the French government. The subsidy primarily finances long-term research, which lays the groundwork for future developments. Research contracts finance medium and short-term work, closer to applications. The strategic challenge for ONERA is to organize this broad knowledge stream, ranging from the acquisition of knowledge to transferring it to industry. Space was obviously a domain of interest since the inception of the space age, and electric propulsion of satellites has been studied at ONERA from 1960s (the oldest vacuum tank still in operation dates from 1961). During the 1970s an emphasis was placed on cesium and mercury thrusters, in particular cesium contact ionization thrusters. In 1 Research Scientist, Physics and Instrumentation Department, denis.packan@onera.fr. 2 Research Scientist, Physics and Instrumentation Department, paul-quentin.elias@onera.fr. 3 Research Scientist, Physics and Instrumentation Department, julien.jarrige@onera.fr. 4 PhD student, Physics and Instrumentation Department, felix.cannat@onera.fr. 5 Research Scientist, Physics and Instrumentation Department, clement.zaepffel@onera.fr. 6 Research Scientist, Physics and Instrumentation Department, julien.labaune@onera.fr. 7 Research Scientist, Physics and Instrumentation Department, trevor.lafleur@lpp.polytechnique.fr. 1

the 1990s the renewed interest in electric propulsion resulted in new works in Hall Effect thrusters and micronewton thrusters. Today the electric propulsion team is in the FPA unit (Lightning, Plasmas and Applications) within the DMPH department (Physics and Instrumentation). It inherited a strong diagnostics and laser culture from the other teams in the department, and is focused on three development activities regarding electric propulsion: innovative diagnostic methods, thruster technologies and plasma models. Some examples are given below. II. Facilities ONERA currently has 3 vacuum tanks dedicated to electric propulsion, shown in Figure 1. The balance tank is 0.5x0.7 meters in dimensions, with a 1000 l/s air pumping speed. It holds the micronewton balance. The most recent tank is the B09, it is 0.8x2 meters large, and has three 2000 l/s turbopump, which results in 2000 l/s pumping speed for Xenon. The use of turbopumps allows rapid opening and pumping of the tank, and along with its multiple feedthroughs and optical viewports it is ideal for developments in the sub-mn range. Finally the B61 vacuum tank is the oldest and largest: it is 1x4 meters in dimension, with a hatch 0.6x0.8 meters equipped with a 50cm gate valve. Pumping is performed by a 2000 l/s turbopump (air speed) and a 7000 l/s (Xe speed) cryo pump with 3 GiffordMacMahon cold heads. The total pumping speed for Xenon is 8000 l/s. A commercial Hiden PSM mass/energy spectrometer (0-300 amu, 0-1000 ev) can also be used to measure ion beam energy and content. Figure 1. Vacuum tank for the micronewton balance (left) and B09 vacuum tank (right). Figure 2. B61 vacuum tank. 2

Different intrusive probes have been developed in-house and are used for measurements: Faraday, RPA, emissive, Langmuir, hairpin resonator (microwave), wave phase-shifting (microwave), sputtering probe. They will not be detailed here. III. LIF (Laser Induced Fluorescence) Lase Induced Fluorescence has been studied at ONERA since the late 1990s. The first measurements were performed on an SPT-50 and SPT-100 1, and were the first LIF measurements in France on an electric thruster (Figure 3). Figure 3. Axial velocity of Xenon ions in a SPT-100, and reference versus Doppler-shifted signal (inset). LIF was further developed 2 in the framework of the Cesium FEEP thruster program at ESA. Field Effect Electric Propulsion thrusters were being developed at the micronewton level (150 μn max thrust) for different scientific missions, and the question arose of the mass efficiency of the thruster. Since mass consumption was minute with respect to the weight of the thruster, it was difficult to measure and compare to the integrated ion current measured. Thus we proposed to measure the mass efficiency of the thruster optically. Since ion mass flow rate is known from the total current, what is needed is neutral cesium mass flow rate. Profiting from the work that had been performed at ONERA on cesium cold atoms interferometry, we built a setup (Figure 4) under ESA contract that comprised 3 laser beams for the 852 nm transition of cesium: since each beam measures one component of the velocity in the probed volumes, the actual velocity vector is obtained. The intensity of the signal gives the absolute density of cesium (since the transition starts from the ground state). Integrating the density and velocity maps (Figure 5) we found the total mass flow rate of neutral cesium. Comparison to the ion current showed for the first time that the mass utilization efficiency of the thruster was about 50%. 3

Figure 4. Three-laser LIF setup with a Cesium FEEP thruster from ALTA mounted. Figure 5. Measured absolute density map (left) and velocity vector map (right) of neutral cesium. More recently, a new LIF method was devised3 to go even further. One should note that standard LIF, as applied most of the time on thrusters, has limitations because it leads to only a partial appreciation of the ion velocity distribution. Indeed it can be shown that with a fixed laser beam orientation, only a projection of the ion velocity distribution is measured. Therefore the use of 3 laser beams provides only 3 projections of the velocity distribution, which obviously cannot describe complex distributions, such as in crossing beams. Having a technique able to retrieve the full IVDF might give a new picture of current thruster physics. Thus, under an ESA contract, a new method and setup has been devised to reveal the full extent of the content of thrusters plumes. The method proposed is based on LIF tomography in phase space. The setup (Figure 6) allows the laser beam optics to scan a whole halfsphere, in front of the thruster, thus testing many different angles. The full velocity distribution in the 3D velocity space of the ions in the probed volume can then be retrieved. Measurements are ongoing on a SPT-20 from SPASE GmbH. 4

Detection 1 XYZ Translation stage Carriage Laser injection Thruster mount point Detection 2 Circular Translation stage (θ) Figure 6. Velocity-space tomography 3D LIF setup. Rotation stage (ϕ) IV. Micronewton thrust measurements At the end of the 90s, it appeared that several incoming scientific space missions, among them the MICROSCOPE mission from CNES which payload is built by ONERA, would need micronewton thrusters (other missions include LISA-Pathfinder, GAIA, Euclid, etc..). But there did not exist a thrust balance sensitive enough to measure a 1 μn thrust in Europe at the time. A program at ONERA was thus set up to develop such a balance on internal funding, and the micronewton balance was devised. It is based on a vertical pendulum, the movement detection is made using a capacitive sensor and absolute calibration of the thrust is obtained with calibrated weights. Further improvements were done under CNES contract, and now the ONERA balance boasts the best performance in Europe, particularly for cold gas thrusters. Tests were performed on an Indium FEEP thruster and on a Helicon thruster, but most of the measurements were done on cold gas thrusters because of the high demand for such measurements (cold gas thruster become baseline for most scientific missions). The ONERA micronewton balance was thus selected by ESA and CNES for the performance tests of GAIA, LISA-Pathfinder and MICROSCOPE flight models cold gas thusters. In particular, the thrust vs flowrate curve (or equivalently Isp vs thrust) was measured 4 for all 13 GAIA flight model thrusters from Selex ES. Other types of measurement that can be performed include thrust noise measurements and resolution measurements. Examples of performance and measurements with the balance are shown in Figure 7, Figure 8 and Figure 9. A second μn/mn balance, that can accommodate larger and heavier thrusters, was developed in the framework of an FP7 project (HPH.com). It was able to measure 100 μn thrust steps from a 50Watt helicon thruster from Khai in Ukraine. 5

Figure 7. Balance with a cold gas thruster mounted (left) and example of performance (right): floor nois of the balance (top) and 0.1 μn thrust steps (bottom). Figure 8. Left: 0.5 μn thrust steps on top of a 500 μn thrust for a Euclid EM thruster (black) compared to the mass flow rate steps, (red) which were converted into thrust using the Isp. Right: Noise measurement of a LISA-Pathfinder FM thruster. 6

Figure 9. Isp vs thrust with 1s error bars for a GAIA FM (left) and comparison of Isp measurements for all 13 GAIA trusters and the Euclid EM thruster (right). V. EBF (Electron Beam Fluorescence) Electron beam fluorescence (EBF) is a technique for measuring the absolute density of atomic or molecular gas. The principle is based on the injection of high energy electrons in a gas by an electron gun: some particles are excited by direct electron impact (pumping) and emit light by radiative de-excitation (fluorescence). In appropriate conditions, the intensity of the fluorescence signal is proportional to the density of gas. One of the advantages of EBF is its ability to analyze all types of gases: all atoms and molecules can indeed be excited effectively by inelastic collisions with electrons. Unlike the LIF technique, there is no selectivity to species or to energy levels. Moreover, it is possible to analyze all types of atoms or molecules in the ground-state energy level. The EBF technique was first developed at ONERA for use in low density wind tunnels (super- and hypersonic) when Shlieren techniques are not useable. It has then been adapted, under a CNES contract, for the visualization of the plume of a cold gas thruster 5. The idea is to scan the electron beam using a coil actuator (Figure 10), which by proper calibration, results in an image of a section of the plume. This image can be used to compare with numerical codes, are can be used directly to show an asymmetry in the plume (tilted thrust). -15 Density Image -10 20.5-5 20 mm 0 5 19.5 10 19 15 0 5 10 15 20 18.5 mm Figure 10. Principle of the EBF method scanning beam (left) and resulting absolute density section of the plume, in m -3 (right). It should be mentioned that the technology used for the electron gun is proprietary from ONERA: it is a secondaryemission electron gun. It uses a plasma instead of a filament, and thus can operate in degraded vacuum with robustness. 7

The EBF technique was enhanced with the LIF technique, where a laser is used on an excited state created by the electron beam. The technique is called EBLIF, and is useful in determining velocities and temperatures in a cold gas plumes. The EBF technique has also been recently adapted for the measurement of gas injection uniformity by the anode of a Hall effect thruster 6. Indeed one of the main causes of lifetime limitations in Hall Effect Thrusters is the erosion of ceramic walls of the discharge channel due to ion impingement, and azimuthal inhomogeneity of the plasma density in the discharge channel (due to gas injection non uniformity) would result in non- axisymmetrical plume which could potentially decrease thruster performance and generate a dissymmetry of the erosion process. Therefore it is important to assess the azimuthal distribution of neutral gas in the discharge channel. However very few if any- non-intrusive techniques are capable of measuring accurately the gas density at such low pressure (typically: 10-3 mbar) with a good spatial resolution. We developed a diagnostic based on electron beam fluorescence for the characterization of the distribution of propellant gas into the discharge channel through the anode, and its demonstration on the anode of a Snecma PPS 5000 Thruster. The thruster is operated is cold gas mode (without discharge) for the measurements of the gas density. Figure 11 presents the configuration of the experiment. By rotating the channel over several angles, the density over the whole gas plume can be obtained, and measurements are shown in Figure 12. Figure 11. Configuration of electron beam scanning in front of the annular ceramic channel (left), image of the electron beam from the front, with the channel fluorescing from the X rays created at the e-beam impact on the vacuum tank (middle) and resulting EBF image at one rotation angle (right). Figure 12. Density of cold gas measured for a perfect anode (left) and for an anode with and added defect (right). 8

VI. ECR thruster development ONERA develops a cathodeless, magnetic-nozzle ECR plasma thruster that is under a patent. It is based on the resonant heating of electrons immersed in the magnetic field (created by permanent magnets) by a microwave source in a coaxial geometry, and the acceleration/ejection of quasi-neutral current free plasma by the magnetic nozzle (Figure 13). Figure 13. Principle of the ECRA thruster (left) and picture of the plume (right). The intrinsic advantage of the thruster compared to other technologies is that it does not need a neutralizer (similarly to helicon thruster) and is naturally magnetically shielded, hence its simplicity, reduced cost, higher reliability and potential long lifetime. An added advantage is the possibility to do magnetic beam steering. The originality of this particular ECR thruster lies in its coaxial geometry that allows a size reduction compared to typical waveguide ECR sources, and in the low power consumption (around 50 W). Several experimental characterization of ECR plasma thruster have already been done in the past few years at ONERA using argon and xenon as propellant gases 7,8,9,10. The effect of the thruster diameter is investigated in a more recent publication 11. The study of such a magnetic nozzle thruster is challenging, both experimentally and theoretically. Experimentally, unlike usual technologies (Hall, gridded), there is no a priori indication of even the order of magnitude of the ion current and ion energy, since the only input are the gas and microwave feeding (no DC voltage). Ion current and energies must be measured with electrical probes, which must be calibrated. And it is notorious than magnetic nozzle thrusters are particularly sensitive to the vacuum level, so a good vacuum (< 10-5 mbar) is needed, both for the thruster operation and for the measurements. Theoretical difficulties also arise because the ECRA thruster exhibits wave coupling, ionization and acceleration simultaneously, which is difficult to model. But very encouraging results have already been obtained on a low power development model: 16% total efficiency has been achieved at the 1 mn point (see table below). Examples of energy and current density distributions are shown in Figure 14. Future works on the ECR plasma thruster will include direct thrust measurements on a thrust balance. Finally, it should be noted that the efficiency level reached (16 % at 1mN) is already not far from the state of the art of plasma thrusters in that thrust range, and the simplicity of the ECR thruster (one single power supply and gas feed line, no cathode) makes it already a potentially interesting candidate for microsatellite propulsion. Among all parameters that need further investigation (magnetic field topology, wave frequency, etc ), studying an increase in size would probably help improve further the efficiency and reach the thrust levels that are needed for the larger satellite platforms. 9

Figure 14. Angular profile of the ion current density and the ion energy distribution function measured in a 13mm diameter thruster. (a) and (b) using argon at 35W (c) and (d) using xenon at 40W. VII. Modeling Several modeling efforts have been made at ONERA as applicable to electric propulsion. A 2D3V PIC code has been developed for low pressure plasma source modeling, and it could be modified for thruster modeling. Since it is not a parallel code, other options are also envisaged to tackle massively parallel kinetic simulations. Dedicated fluid codes have also been developed. One-fluid and two-fluid approximations have been used to model a helicon thruster, and results have been encouraging (Figure 15). Figure 15. Image of a helicon thruster from KHAI, Ukraine, firing in the ONERA B09 vacuum tank (left) and 1-fluid modeling of the plasma (right). 10

But for fluid modeling, a major new tool is now available at ONERA. The numerical code KRONOS that couples plasmas with hypersonic flow fields has been developed during the past 4 years. Under the 1-fluid and quasi-neutral approximations for the plasma, and based on the already existing CEDRE CFD platform (hydrodynamic) and SATURNE code (electric solver), it is able to simulate plasmas from 0D to 3D, with non-equilibrium electron energy distribution (Boltzmann equation) or at LTE, with complex plasma chemistry, MHD (including Hall effect), and the interaction with complex flow conditions ranging from subsonic to hypersonic. It is also fully parallel, allowing to compute plasma simulations with high refinement and complex geometries, and runs on the 5000-core computer of ONERA. Given its capabilities, it is a world-level numerical tool. The first goal of KRONOS is to be able simulate gas discharges for various applications, such as aerodynamic actuators, plasma-assisted combustion, decontamination, arcs, lightning or reentry (Figure 16). But it could be successfully used to model electric thrusters where the fluid approximation can be valid, such as Hall effect thrusters and MPD thruster. Full 3D simulations would then be accessible. Figure 16. Hall currents during the reenntry of a magnetized sphere, calculated by the KRONOS code (70 km altitude, 6.5 km/s, β e =1). Finally, analytical models are also being developed to better understand ECR thrusters 10,12. VIII. Perspectives ONERA has developed a recognized know-how in electric propulsion, and the development of this activity is one of its priorities. In the past 10 years it has built a strong and young team on electric propulsion, which will be the basis for future work as continuity of the competences is assured. Efforts will continue unabated in diagnostic development and thruster technologies, and specific efforts will be made in the modeling area, both in the PIC and fluid models, in particular with the use of KRONOS. Finally, ONERA s facilities currently do not allow to test thrusters of more than 20 mn thrust, which is not in phase with the growing demand in higher power, therefore a project is being studied to acquire a larger facility. 1. Dorval, N. et al. Determination of the Ionization and Acceleration Zones in a Stationary Plasma Thruster by Optical Spectroscopy Study: Experiments and Model. J. Appl. Phys. 91, 4811 4817 (2002). 2. P.Q. Elias et al. Optical Measurements of Neutral Cesium Mass Flow Rate in Field Emission Thrusters, JOURNAL OF PROPULSION AND POWER, Vol. 27, No. 2, March April 2011 3. P.Q. Elias et al. Full Ion Velocity Distribution Function measurement in an Electric Thruster, using LIFbased tomographic reconstruction, IEPC-2015-90315 4. J. Jarrige et al. Thrust Measurements of the Gaia Mission Flight-Model Cold Gas Thrusters, Journal of Propulsion and Power, Vol. 30, No. 4 (2014), pp. 934-943. 5. D. Packan et al. Number Density and Velocity Measurement in the Plume of a Micronewton Cold Gas Thruster by Electron Beam Fluorescence and Laser Induced Fluorescence, IEPC-2009-184 6. J. Jarrige et al. Assessment of the Azimuthal Homogeneity of the Neutral Gas in a Hall Effect Thruster using Electron Beam Fluorescence, IEPC-2015-91059 11

7. J. Jarrigeet al., Characterization of a coaxial ECR plasma thuster, Proceedings of the 44th AIAA Plasmadynamics. 8. J. Jarrige et al., Performance Comparison of an ECR Plasma Thruster using Argon and Xenon as propellant Gas, Proceedings of the 33th International Electric Propulsion Conference, vol. 420, 2013. 9. F. Cannat et al., Experimental investigation of magnetic gradient influence in a coaxial ECR plasma thruster, Proceedings of the Space Propulsion conference, Cologne, Germany, 2014. 10. F. Cannat et al., Optimzation of a coaxial electron cyclotron resonance plasma thruster with an analytical model, Physics of Plasmas, 22, 053503 (2015). 11. F. Cannat et al., Experimental geometry investigation of a coaxial ECR plasma thruster, IEPC-2015-90492 12. Lafleur T., Helicon plasma thruster discharge model, Physics of Plasmas, vol. 21, no. 043507, 2014. 12