An Evaluation of Hollow Cathode Scaling to Very Low Power and Flow Rate

As Hall and ion thrusters are scaled to power levels below 300 W to operate on smalland micro-satellites, hollow cathodes must be designed to accommodate the new operating regimes with efficient use of power and propellant. The conclusions of previous cathode scaling efforts by other authors are reviewed, and the importance of heat transfer and orifice processes are emphasized. Performance data for 3.2 mm orificed hollow cathodes are presented. Spot mode operation was achieved on Xe at the same equivalent mass flow rate as the SERT II ion thruster neutralizer. An insert with a diameter on the order of the orifice size was tested and required ten times the mass flow of the SERT II neutralizers to generate enough ions within the cavity for charge neutralization. Design considerations are discussed for improvement of the power and propellant consumption of the cathodes presented in this work. Copyright  1997 by Matthew Domonkos. Published by the Electric Rocket Propulsion Society with permission. Nomenclature AR = orifice to insert area ratio d = orifice diameter D = insert diameter I = current, A Jo = equivalent flow rate, A l = orifice length L = orifice plate to keeper gap length W = atomic mass, amu σ = cross-section, m Introduction The advent of small satellite constellations for cellular communications and their need for efficient propulsion has driven electric thruster design toward low power levels, typically 300 W or less. Candidate thrusters for these missions include resistojets, arcjets, ion thrusters, and Hall thrusters. Both ion and Hall thrusters use hollow cathodes, and scaling these thrusters to power levels below the kilowattclass devices currently planned for use involves reduction in thruster component size. The 6.4 mm diameter cathodes currently employed in ion and Hall thrusters in the United States consume needlessly large fractions of the power and propellant for low power thrusters. Such thrusters will require between 0.2 and 1.5 A of current from their discharge cathode or neutralizer. Consequently, an investigation is underway to examine the scaling of hollow cathodes to minimize the consumption of power and propellant. This paper presents the initial experimental investigation of 3.2 mm diameter orificed hollow cathodes and discusses the issues relevant to further reductions in cathode power and propellant consumption. Orificed hollow cathodes generate a plasma within a cylindrical cavity by the thermionic emission of electrons from a low work function insert. The plasma then flows through an orifice and couples the cathode to the rest of the discharge. Since the plasma is largely confined within the hollow cavity and little power is required to sustain the thermionic emission relative to a more open geometry, this configuration permits low cathode fall voltages. Further, the plasma facilitates electron emission by neutralizing the effects of space charge. If the thermal flux of electrons from the orifice is sufficient to carry the discharge, the electric field between the keeper electrode or anode accelerates those electrons to maintain the current. This type of operation is called spot mode. As the mass flow rate decreases, the thermal flux becomes less than the discharge current, and additional ionization events must occur as the electrons traverse the gap between electrodes. This is termed plume mode operation. The ions created in the gap have some fraction of the cathode fall to accelerate them toward the cathode orifice plate which exhibits damage, presumably from ion sputtering, when operated in plume mode. Whether this erosion prohibits cathodes operating in plume mode from fulfilling mission requirements has yet to be established. However, extended lifetime operation has been demonstrated to more than 25,000 hours in spot mode which is consequently the preferred operating condition. As a result of these phenomena, cathode efficiency is dependent upon scaling of the flow rate necessary to maintain spot mode emission. It is also necessary to reduce the power consumption of current generation 6.4 mm diameter cathodes. These cathodes typically operate at several tens of watts. Power deposition to the cathode acts to maintain thermionic emission by balancing electron cooling, thermal conduction along the cathode tube, and radiative transfer to the surroundings. Clearly a reduced current requirement decreases the electron cooling. However, to minimize the heat loss due to conduction and radiation, geometric modification of the cathode becomes necessary. This forms the basis for the drive toward physically smaller cathodes than the 6.4 mm diameter cathodes currently in use. This paper begins by reviewing the previous work performed in cathode scaling. The experimental apparatus is discussed, followed by a summary of the performance of the cathodes. The implications of these results on the design of even more efficient hollow cathodes is discussed, and finally, the conclusions are summarized. Cathode Scaling Background Both 6.4 mm and 3.2 mm outer diameter hollow cathodes have been under investigation since the middle 1960’s, although most of the recent work in the United States has focused on the 6.4 mm cathodes. The SERT II space test of mercury ion thrusters formed the basis for most of the hollow cathode work in the late 1960’s and early 1970’s in the U.S. One of the first attempts to develop a scaling law for orificed hollow cathodes was introduced by Kaufman. In this work, spot mode operation was observed provided that the ratio of the equivalent mass flow to the orifice diameter in a mercury hollow cathode exceeded an empirically determined constant: J o d ≥ 0.14 to 0.40 (1) This equation gives a means by which to minimize the mass flow rate for a given cathode. As inert gas thrusters began to be preferred, research on hollow cathodes also gradually moved toward use with the noble gases.25 Rehn and Kaufman expanded the dimensional analysis shown by Equation 1 to apply to inert gas hollow cathodes. A generalized criterion for spot mode operation that is dependent upon the propellant species was presented: J oσ W d ≥ 13.9x10 (A −m −amu 0.5 mm ) (2) The cross-section used in this analysis was the maximum value for the ionization of neutrals by electrons. In addition, some effort was made to correlate the coupling voltage and the electron emission to the propellant type. The voltage parameter scaled as the inverse of the ionization potential. Consequently, the discharge voltage and power were shown to increase with the propellant ionization potential. Geometric considerations were absent from the voltage parameter. A more rigorous phenomenological model for orificed hollow cathode operation was introduced by Siegfried and Wilbur in 1981, and then refined by the same authors in 1984. This model examined the physics of the plasma within a mercury hollow cathode. While the model demonstrated qualitative agreement with experiments, little discussion was presented about using the model to scale the cathodes. Work by Salhi and Turchi expanded on the phenomenological model of Siegfried and Wilbur to reduce the number of experimental inputs to the model and to include two dimensional variation in the heavy particle temperature. The model was used to establish some scaling relations for high current (greater than 20 A) cathodes. All of their results were in the limit of high power density where the dominant cooling mechanism was electron emission. Although this limit may be ill-matched to operation of low current cathodes, the trends predicted by this model are summarized here. The maximum discharge current for a given cathode was shown to be proportional to the insert inner diameter; i.e., I ∝ D. (3) As the area ratio between the orifice and insert, AR = D ( ) , decreases, so does the maximum current that the cathode can carry. A smaller area ratio or a smaller insert diameter leads to a reduction in the cathode fall voltage for a given discharge current. This result shows that the smallest cathode capable of carrying a given current is also the most efficient since the power deposition is the smallest in this case. The model also predicted that larger cathodes tend to operate at low temperatures relative to small cathodes. These results provided insight into the complex relationship between the orifice and insert diameters. Modeling of the physical processes within the orifice has been comparatively less wide-spread. Mandell and Katz have developed a model describing the mechanism behind the transition from spot to plume mode based on a control volume approach in the orifice. This model treats the insert region as a plasma source. It balances ion production and loss, and energy within the orifice to determine electron temperature and number density given the discharge current and flow rate. The thermal flux of ions from the insert region is neglected. The model facilitates optimization of cathode geometry versus operating conditions, and illustrates that the ratio of the orifice length to diameter, l/d, should be optimized to provide a high gas utilization without driving up the cost of ion production. Both power consumption and gas utilization were found to scale with the ratio, l/d. Additionally the model provided an alternative method for determining the transition to plume mode by equating the thermal flux of electrons to the keeper with the discharge current. This condition defined the lower limit necessary for spot mode operation. The most recent iteration on the phenomenological model was developed by Capacci, et al. This work builds on the work by Siegfried and Wilbur by including models of both the orifice region and the expansion toward the keeper. The voltage-current characteristics in this work accurately model the transition to plume mode operation, and the trends are in good qualitative agreement with experiment. The predictions of these models will be compared with the results of the current experimental investigation, and the consequences of these data for very low power and flow rate cathodes will be discussed. Experimental Apparatus All cathode tests were performed in port 2 of Tank 11 at the NASA Lewis Research Center. This port was isolated from the tank using a vacuum gate valve, and the vacuum was maintained using a cryopump. The base pressure in this tank was approximately 4 x 10 Torr, and the pressure in the vessel during testing ranged from 9 x 10 to 4.6 x 10 -5 Torr on Xe. A flow system using a 0 to 5 sccm volume flowmeter and a needle valve supplied 99.999 percent Xe to the cathode with an accuracy of +/0.1 sccm. The experimental configuration is illustrated in Figure 1. A swaged heater was friction fit around the outside of the cathode tubes. For these tests, only a molybdenum keeper was used to evaluate performance. Table 1 gives the area ratio and l/d. The two mechanically identical cathodes with BaO:CaO:Al2O3 impregnated porous tungsten inserts were tested and referred to as SC.001 and SC.002. Additionally, a cathode with a rolled-foil insert coated with R500 low work function compound was used to evaluate the effect of insert diameter on the performance. This was referred to as SC.002rfi since it was tested in the SC.002 tube prior to the tests with the porous tungsten insert. Consequently, the orifice geometry was identical to the other cathodes within the design tolerances. However, the insert diameter was on the order of the orifice diameter. This was accomplished by rolling a long piece of foil until a small inner diameter was achieved. The application of the R500 further reduced the insert inner diameter. This cathode had l/d and area ratios of 9 and approximately 1, respectively. The large area ratio made the orifice length, l, somewhat ambiguous.