IEPC-93-216 2008

MEASUREMENT OF ENERGY DEPOSITION MODESIN AN INTERMEDIATE POWER HYDROGEN ARCJET

W.A. Hoskins*, A. E. Kull**, W. M. Nessert and G W. Butler*Rocket Research Company

Redmond, WA 98073

Abstract Pi = density of species i

p = total densityIonization losses, thermal losses and thrust efficiency =

a = Stefan-Boltzmann constantwere measured for a hydrogen arcjet at powers from 6 = Stefan-Boltzmann constantto 14 kW and at specific energies of 150 and 200 MJ/kg. T = thrust

Ionization losses were calculated from electron Ti = temperature of species or node i

densities obtained from emission spectroscopy at the ui,, uzi,U = total, axial, radial velocity of iexit plane. Thermal losses were calculated from arcjet V = arcjet voltagesurface temperatures obtained with thermocouples and ) = vibration constant of species iimaging optical pyrometry. Ionization rates werefound to be relatively constant with different Introductionoperating conditions, while thermal losses increasedsignificantly with decreasing power or increasing As arcjet propulsion systems near deployment forspecific energy. north-south stationkeeping, interest in improving

performance and developing arcjets for other missionsAn analysis of energy loss modes derived from a has intensified. For example, intermediate powernumerical simulation of the arcjet reveal random ammonia and hydrogen arcjets are being considered fortranslational motion and dissociation to be the largest orbit transfer missions. To be competitive with currentenergy sinks at over 20% and 15%, respectively, chemical propulsion systems and attractive to users forThermal losses, ionization. rotation and profile losses such missions, an arcjet system will have to deliverwere also found to be signi:lcant. 1000 to 1200 s specific impulse at 35 to 50% efficiency,

depending upon the mission(l). However, efficienciesNomenclature approaching 50% at high specific impulses have yet to

be achieved.Ai = area of node i

c = speed of light In order to increase the specific impulse of an arcjet

dA = integral surface area thruster at a fixed power-to-flow-rate ratio (specificEiD n n energy), it is necessary to increase the thrust efficiency.Ei" = dissociation energy of species iSDuring the rapid expansion of the propellant in theEi = ionization energy of species i nozzle, much of the energy in the propellant is frozenEj = emissivity of node i in such non-translational energy modes as ionizationh = Planck's constant and dissociation and is not recovered as useful thrust.

hi 0 = enthalpy of i referenced to 0 K In addition, significant amounts of energy are lost as

hfi0 = heat of formation of i at 0 K thermal energy to the arcjet electrodes and structure.The objective of this work is to draw conclusions about

I = arcjet currentI = arcjet rnt the dominant modes of energy loss in arcjets in order tok = Boltzmann's constant gain insight into designs that minimize these losses.mh = mass flow rate

ii = unit vector normal to dA While much excellent work is being done to measure

ni = number density of species i various plasma parameters in arcjets with ever more

q = rate of heat transfer from node i sophisticated diagnostics, only a few papers havefocused on characterizing the energy deposition modesin arcjets. A comprehensive study of 30 kW classhydrogen arcjets in 1965 by the McDonnell Aircraft

* RRC Sr. Development Engineer Corp.(2 ) measured thrust, flow rate, thermal losses by

** RRC Development Engineer calorimetry, and exit plane profiles of severalt RRC DevelopmentEngineer

SRRC Staff Engineer

1

2009 IEPC-93-216

parameters, including mass flux and stagnation conditions to highlight the major energy lossenthalpy. Thrust efficiency was generally 20 to 25% mechanisms, as well as for comparison with :heand thermal losses ranged from 3 to 25%. An exit measurements presented.plane power balance of a water-cooled arcjet indicatedthat roughly 7% of the input power remained in the Arciet Power Balanceexhaust plume as random translational energy,molecular rotation and vibration. Another 50% was To appropriately account for all energy loss modes ininferred to be lost to unrecovered ionization and an arcjet, it is necessary to first define energydissociation energy. These figures were based on deposition modes such that they form a self-consistentindirectly calculated exit plane temperatures on the power balance. An energy flow diagram is shown inorder of 1200 K. Fig. 1. In addition to the input electrical power,

More recently, work by Crofton et al.( 3 ,4 ) with low PInE = IV, (1)

power hydrogen and ammonia arcjets indicated thatpower rogen and ammonia arjets icated that the propellant also carries with it its own enthalpy,dissociation and thermal losses through the arcjet including heats of formation. For the calculationsincluding heats of formation. For the calculationselectrodes were major loss mechanisms, althoughelectrodes were major loss mechanisms, although presented, enthalpy was referenced to 0 K because themeasurements with XUV absorption spectroscopymeasurements with XUV absorption spectroscopy theoretical maximum thrust energy is extracted fromindicated dissociation losses as low as 4%. In addition, the propellant by reducing its temperature to 0 K. The

th e propellant by reducing its temperature to 0 K. Thetheir measurements of exit plane temperatures input power due to propellant enthalpy will thus be,. input power due to propellant enthalpy will thus beindicated that the energy tied up in translational, defined as:

rotational and vibrational modes was roughly 17%.Finally, calorimetry measurements by Sankovic andCurran(5) for 1 to 4 kW hydrogen arcjets found 5 to Pnh =mrh i-(h0(Tin)+hO) (2)20% of the input energy was lost to heating the ielectrodes.

Thermal loss is defined as the net power that goes intoIn a previous paper by the present authors,(6) heating the arcjet electrodes. This loss category arisesionization loss calculations were presented from principally from the current attachment and netspectroscopic measurements of an intermediate power conduction/convection from the flow. A morehydrogen arcjet. A qualitative comparison of practical definition of the thermal loss is the heationization losses with measured anode temperatures transferred at steady state across a control surfacesuggested a possible trade between frozen flow losses surrounding the arcjet. Primarily, this heat transferand thermal losses. This paper presents thermal loss, as occurs by radiation, although conduction through solidwell as ionization loss measurements, for the same connections and transfer to the background gas inarcjet for 6 operating conditions over 5 power levels ground test facilities must also be included.and 2 specific energies. The ionization losses werederived, as before, from spectroscopic measurements of The rest of the power balance is comprised of energythe Hydrogen Balmer series taken at the nozzle exit. deposition modes in the exhaust stream at the nozzleThe thermal losses were determined from arcjet exit plane. All integrations in the following equationssurface temperatures by calculating the heat transfer occur across the exit plane. Although the net plasmaacross a control surface around the arcjet. The radiation across the nozzle exit is another loss, it hasmaximum energy lost to other modes, such as been found to be insignificant in arcjets( 8 ) and will bedissociation, can be inferred from an energy balance of neglected in this paper. The loss to unrecoveredionization and thermal losses, along with the measured ionization is defined as:thrust, flow rate and arc power.

lon = JE+ niui indA (3)To better characterize the magnitude of other energyloss mechanisms, the results of a single fluid two-dimensional performance model of the arcjet were where i is summed over the ionized species. The loss toemployed. This model( 7 ) yields species densities, dissociation is likewise defined as:temperature and velocity information throughout thearcjet flow field. The numerical results were broken PDs = ED nii u -i dA (4)down into a detailed power balance for three operating

2

IEPC-93-216 2010

where i here is all dissociated species. A rough (Uri 2 -expression for the power lost to electronic excitation PRaV = JPi "i idA (9)can be obtained by subtracting the contributions ofvibration, rotation and translation from the exhaust

enthapy: summed over all species. Profile "loss" arises due tothe conventional definition for thrust power. It is the

PEE = fp ih i (T) ui dA difference between axially directed kinetic energy andEEx thrust power and is analogous to the difference

PVib -Rot RTr (5) between the square of the sums and the sum of the-Pvib -PRot -PRTr (5) squares:

This expression implicitly assumes equilibration 2between these 4 energy modes, which spectroscopic Ppro = fPi Li ,ii dA - PThr (10)measurements have shown not to be the case( 3 ,9 ).i 2However, it is a convenient expression for comparisonwith numerical results. where thrust power is defined as:

Since the characteristic temperature for the vibration T2 1 (fiu zii - ndA) 2

mode in H2 is 6400 K, vibration may not be "fully PThr (11)excited" at the exit plane and the equipartition value 2 2 i Piji -

for specific energy of vibration, kT/mi, cannot be used.However, with the approximation that the molecules The overall power balance is obtained by summing theare harmonic oscillators, an analytical expression for two input terms, PInE and PInh, and setting themthe specific energy of vibration can be obtained(10) and equal to the loss terms plus the thrust power.the loss due to molecular vibration can be written:

Experimental Apparatus

Pvib ni h o ui fi dA (6)Ji[ The arcjet used for this study was the same

e -1 intermediate power (5-15 kW) development model

arcjet used in the study reported previously.( 6 ) Allwhere i is summed over the vibration modes of the data in this study was obtained using a single anodemolecules. Without a well established equilibrium configuration with a 200 half angle nozzle, shown inbetween energy modes, a vibration temperature would Fig. 2. The tests were conducted inside a 2.4 m diameterbe substituted for Ti. vacuum cell in which background pressures ranged

from 80-210 mTorr (11-28 Pa). The arcjet wasAt exit plane conditions, the rotational mode should powred by a Hypertherm MAX100 DC ar cutterbe both fully excited and well equilibrated with the power s y separate hh vtae capacitive startpower supply. A separate high voltage capacitive starttranslational energy,(10) making the rotational loss circuit started the are directly on hydrogen.term:

PRot = ni(kTi)i iidA Hydrogen gas was used as the propellant for alli (7) testing. The flow rate was measured with a

Micromotion coriolis mass flow meter. Arcjet

where i is summed over molecular species. The power performance was measured on a swinging arm, null

lost to random translational motion, including unused balance thrust stand. Thrust, flow rate, voltage,

pV work, is similarly: current, pressures and thermocouple temperatures wererecorded on a PC based data acquisition system. During

PRTr = I f ni j(kTi) Ui - i dA (8) a typical test run, the arcjet was operated for at least 20i minutes before collecting spectroscopic and thermal

imaging data to allow the arcjet to come to thermal

for i summed over all species. The remaining power equilibrium. Steady state thrust, flow rate and power

exits the nozzle in the form of kinetic energy. The loss data were averaged over an end of run sequence to

due to the radial components of the exhaust velocity is: subtract out zero shifts in the data.

3

2011 IEPC-93-216

Data from two test series are reported here. In the first, electron densities resulting from uncorrected data.data was collected at 5 power levels ranging from 6 to Therefore, no corrections have been used for the data14 kW, all at a specific energy of 150 MJ/kg. In the reported in this paper. It should also be noted thatsecond series, two different specific energies, 150 and intensities of the upper 2 lines taken with the wider200 MJ/kg, both at 10 kW, were studied. slits exhibited interference from neighboring lines and

were not used in the following analysis.Ionization Losses

AnalysisIonization losses were estimated from exit planeelectron number density profiles measured spectro- The Balmer series intensity measurements werescopically. The spectroscopic system, measurement and converted to radial profiles of atomic hydrogen excitedanalysis procedures were described in detail in an state number densities through the intensityearlier paper(6 ) and will only be summarized here. calibration, transition probabilities and an Abel

Inversion. Next, a collisional-radiative model ofSpectroscopy System atomic hydrogen was used to relate the state

populations to electron density and temperature at eachThe spectroscopy system is shown schematically in radial location. Finally, ionization losses wereFig. 3. Inside the test cell, emitted radiation from a 0.5 calculated from the flux of electron-ion pairs acrossmm diameter volume of the arcjet plume is focused the exit assuming an average exhaust velocity derivedonto the end of a fiber optic cable. The collection optics from the measured specific impulse. This assumptionare secured to a translation stage that allows spatial tends to underpredict the ionization losses since thescanning in a plane perpendicular to the arcjet plume. highest velocities occur near center line where theThe optical fiber transmits the collected radiation to electron densities are also high.the entrance slit of the SPEX Industries model 500Mhalf-meter spectrometer outside the test cell. The ThermalLossesdetector is a thermoelectrically cooled photo-multiplier tube. A 386 personal computer controls Temperature Measurementswavelength selection and translation stage position, aswell as providing data acquisition. An absolute To obtain estimates for the thermal losses, tempera-intensity calibration was made with a tungsten ribbon ture measurements were made over the entire surface oflamp radiance standard for each wavelength the arcjet by a combination of two methods. The firstmeasured, method consists of a series of chromel-alumel

thermocouples spot-welded along the length of theExperimental Procedure arcjet at the positions noted in Fig. 4. With an

operating range up to 23000 F (1570 K), theseAt each operating condition, a series of measurements thermocouples were used to measure the cooler regionswere taken in a plane 2.75 mm downstream of the of the arcjet.nozzle exit plane. The intensities of the first 9hydrogen Balmer series line were measured in 0.25 mm The second method uses an image analysis system,increments from at least 12 mm above to 12 mm below shown schematically in Fig. 5, to measure anodethe arcjet center line. temperatures with single color optical pyrometery. A

CIDTech Model 3170 CID video camera was used inTo measure the entire line intensity, the exit slits were conjunction with a 9.5 nm bandpass interference filterset to 500 pm in the first test series, corresponding to a centered at 951.8 nm and assorted neutral densitybandwidth of 8 A. Preliminary analysis of this test filters to make surface temperature measurements ofseries indicated that Stark broadening at the electron 1700 OF (1230 K) and above. The camera images weredensities present in the plume could cause a significant captured by an ITI overlay framegrabber board with 8-fraction of the upper lines to be missed by the 8 A bit data resolution and stored on a 386 personalbandwidth. To remedy the problem, a correction for the computer for later analysis.line broadening was initially applied to the data. Widerslits, corresponding to a 32 A bandwidth, were used in The CID camera was mounted outside the test cell forthe second test series. However, a comparison between either side or end viewing. The side view allowsthe Series 1 and 2 data taken at the same operating external surface measurements axially along theconditions demonstrated less than 5% difference in the outside of the anode. The end view was obtained by

4

IEPC-93-216 2012

imaging the inside of the anode nozzle through a qi = OAi Fi E(Ti4 - Tchamber 4 ),reflecting prism placed on arcjet center line about 30"downstream of the nozzle exit. where the gray body view factors, Fi, were calculated

using Thermal Model Generator (TMG), a commercialThe stored images were converted to radiance values via software package. These view factors account for directan absolute intensity calibration obtained using a radiation to the chamber walls, as well as intermediatetungsten ribbon lamp spectral radiance standard. These reflections from neighboring arcjet nodes. The totalradiance values were then converted to temperatures hemispherical emissivities, e,, were assumed for thisusing Planck's Distribution Law and published values analysis to be a constant for each material, althoughof spectral emissivity. ( l 1,12 ) The effective emissivity future analysis will incorporate temperatureof the inside of the nozzle was approximated as double dependent emissivities. Tchamber was assumed to bethe true emissivity due to the effect of reflections. 150°F (340 K), the measured cell wall temperature.Work is being conducted on a view factor model thatmore accurately accounts for the effect of reflections Transfer to the background gas was also calculated forinside the nozzle. each node, i, except those inside the anode nozzle using:

The uncertainty in the thermocouple measurements is qi = heff Ai (Ti - Tgas),estimated at 5%, primarily due to the effect of thethermocouple attachment to the surface. The imaging where heff was chosen to be 41 W/m 2 K, which issystem temperatures are accurate to within 2%, except consistent with the free molecular flow conditionsinside the nozzle where the uncertainty is somewhat present in the test cell outside the plume. Tgas washigher. estimated conservatively to be 150°F (340 K), the

measured cell wall temperature. Background gasThermal Loss Calculations transfer accounted for under 10% of the total thermal

loss and under 1% of the input electrical power. TheseTo determine the thermal loss from the measured

values roughly correspond to the change in the thermalarcjet temperatures, the heat transfer across a control les o a arcet at dieret cl

losses of a 1-4 kW arcjet at different cell pressuressurface surrounding the arcjet was calculated, taking measured by Sankovic and Curran( 5 ). They observed ainto account conduction to the solid connections,

ito a n to te od i drop in thermal losses equivalent to 2% of input powerradiation, and transfer to the background gas. For the when going from 230 mTorr to 2 mTorr, while we

when going from 230 mTorr to 2 mTorr, while wepurposes of the calculation, the arcjet surface was r o of

. . rcalculate a transfer to the background gas of 0.8% ofdivided into a series of nodes. The nodes surroundingwer at 150 mTorr.

the input power at 150 mTorr.the thermocouples were assumed to have roughly thesame temperature as the nearby thermocouple (Fig. 4).The image analysis system provided a more detailed Overall, the uncertainty in the thermal loss values is

estimated to be better than 20%, with uncertainties intemperature map of the anode from which most of theheat transfer occurs. the total emissivities, temperatures and the background

gas transfer coefficient being the biggest contributors.

Three areas of solid conduction were considered: theanode power connection, the cathode power connection. Numerical Loss Estimatesand the mounting bracket. The heat transfer rate wascalculated using the conductivities of the materials and The numerical model used in these studiestemperature gradients as measured by pairs of (KARNAC), has been described in a number ofthermocouples. Total conduction losses were small, previous publications.( 7 ,13 , 14 ) It is a nonsteady, finiteamounting to approximately 2% of total thermal loss. volume code which solves the single fluid equationsand only 0.1% of total input energy. for a plasma with finite rate chemistry and high

temperature transport properties. In addition, theRadiation accounted for the majority of the heat solutions for the anode and plasma energy have beentransfer. Radiative transfer from all surfaces of the physically coupled so that anode and plasmaarcjet, including the end of the anode and the inside temperatures as well as energy fluxes at the anode-surface of the nozzle was determined with the standard plasma and the anode-space boundaries are self-transfer equation: consistent.(15)

5

2013 IEPC93-21 6

It must be noted here, however, that KARNAC does flow rates than on total input power. This idea isnot yet model either swirl or diffusive transport of reinforced by the strong increase in temperature withmass or energy, any of which may be important in increasing specific energy, which is obtained byregions where high gradients or long fluid transit reducing the flow rate at the same power.times exist. Although recent comparisons with laserinduced fluorescence (LIF) measurements at Stanford Ionization and thermal losses, as well as measuredUniversity have shown excellent agreement between thrust efficiency are plotted in Fig. 8 for all operatingpredicted and measured exhaust velocity profiles of a conditions tested. While the absolute ionization losslow power H2 arcjet, significant differences between showed an increase with increasing power, its fractionmeasured and predicted exhaust temperature profiles of the total energy actually dropped slightly fromremain unresolved.( 16) In particular, the measured 9.2% at 6 kW to 7.6% at 14 kW. At the same time theretemperature profiles at high specific energies was a dramatic decrease in thermal losses from 22% atdemonstrate significantly greater centerline values 6 kW to less than 5% at 14 kW. Meanwhile, the thrustthan the predictions. The neglect of mass and energy efficiency went up slightly, and then dropped, but diddiffusion, particularly at the arc root, may be the cause not change greatly.of such discrepancies.

As specific energy increases, ionization losses increaseCurve fits are used throughout the code to estimate slightly due to the increased exhaust velocity. Again,thermodynamic properties as a function of the local the thermal loss show a dramatic increase from 8% totemperature and species populations. The use of such 17%. This increase in thermal loss partially appears as afits is equivalent to the assumption that the internal 4% drop in thrust efficiency.energy modes instantaneously relax to their equi-librium values at the local temperature. Working Overall, however, the ionization losses do not appearwithin the limits imposed by the model uncertainties, to change dramatically with operating condition for awe have used the predicted exhaust product concen- given arcjet geometry, while the thermal losses do.trations, velocities, and temperature distribution to Coupled with the widely observed fact that thrustestimate the energy residing in different modes at the efficiency is not easily increased, there appears to be aarcjet exit plane and to estimate the anode thermal trade off between thermal losses and other losslosses due to conduction, convection, and arc categories that inhibits significant improvements inattachment. thrust efficiency.

Results To gain insight into the major energy deposition modesother than ionization and electrode heating, the results

Electron density profiles for three of the operating of the numerical model were analyzed according to theconditions are shown in Fig. 6. As power, and therefore categories in the arcjet power balance section. Theflow rate, increases for the same specific energy, the fractions of input electrical energy for each mode arecenter line values remain unchanged, but density shows plotted on Fig. 9. The fractions total slightly greateran increase at larger radii, indicating a change in the than 1, in part because of the enthalpy of the propellantinteraction with the nozzle wall with increasing flow at the inlet, which is typically about 2% of the inputrate. The densities remain relatively unchanged electrical power. The power fractions for thrust,between the different specific energies at 10 kW, which ionization and thermal losses agree very closely withis in contrast to the dramatic decrease in densities with experiment for the two 10 kW cases. However, for theincreasing specific energy previously reported in a 100 14 kW case, the model overpredicts all three fractionsnozzle. The increased interaction of the flow in the 100 by 2 to 4%.nozzle may be responsible for greater sensitivity of thedensities to changes in flow rate. Overall, however, the As expected, dissociation is a significant losselectron densities showed relatively little dependence mechanism, accounting for about 15% of the power inon operating condition in the 200 nozzle, all cases. However, the largest single loss mechanism

is random translational energy at 20 to 23% of inputIn contrast, the anode temperature profiles shown in power. This value is dependent upon the predictedFig. 7 demonstrate clear dependence on operating temperatures at the exit plane, which peak at overconditions. The anode temperatures tend to drop with 4500 K. While measured rotational temperatures haveincreasing power at the same specific energy, indicating been reported at roughtly 2000 K for low powera stronger dependence on the cooling effects of higher arcjets( 3 ,9 ), recent LIF measurements, also of a low

6

IEPC93-216 2014

power arcjet, indicate peak translational temperatures Referencesof 4000 to 5000 K or centerline ( 17 ). However, even auniform exit tempeature as low as 2000 K would stillproduce a translational power loss as high as 15%. 1. Vaughan, C. E. and Cassady, R. J. "An Updated

Moreover, since the rotational temperature is closely Assessment of Electric Propulsion Technology

equilibriated with translation, the rotational power for Near-Earth Space Missions", AIAA paper 92-

fraction would account for another 5% of the losses. 3202, 28th JPC, Nashville, TN 1992.

2 Van Camp, M. N. et al. "Study of Arc-JetThe energy lost to electronic excitation, vibration, and Propulsion Devices", NASA CR-54691, Marchradial components of the exhaust velocity was found 1966.to be relatively small. Although the calculation ofelectronic excitation power was relatively crude, it can 3. Crofton, M. W., Welle, R. P., Janson, S. W. andbe shown with Boltzmann and Saha relations that the Cohen, R. B. "Rotational and Vibrationalpower in electronic excitation cannot exceed about 2% Temperatures in the Plume of a 1-kW Ammoniaof that in ionization, and is therefore negligible. In this Arcjet", AIAA paper 91-1491, 22nd Fluidanalysis, vibrational temperatures were assumed to be Dynamics, Plasma Dynamics and Lasersequilibriated with local rotational and translational Conference, June 1991.temperatures. However, measurements have typicallyshown vibrational temperatures to be somewhat below 4. Janson, S. W. "The Impact of Advanced Diagnosticthe predicted temperature used in this analysis. The Techniques on Electrothermal Thruster Design",remaining loss mode, profile losses, was a moderate 6 AIAA paper 92-3242. 28th JPC, July 1992.to 9% of input power, which is comparable to therotational power lost. 5. Sankovic, J. M. and Curran, F. M., "Arcjet Thermal

Characteristics." AIAA paper 91-2456, 27th JPC,

Unfortunately, this analysis did not demonstrate a Sacramento, CA, June 1991.

strong trade off between thermal losses and any other 6. Hoskins, W. A, Kull, A. E. and Butler, G. W.,single loss category. Rather, changes in thermal losses "Measurement of Population and Temperatureappeared to be offset by small changes in several of the Profiles in an Arcjet Plume." AIAA paper 92-other major loss categories. 3240, 28th JPC, Nashville, TN, 1992.

Conclusions 7. Butler, G, W. and King, D. Q., "Single and TwoFluid Simulations of Arcjet Performance", AIAA

Ionization and thermal losses were measured for an paper 92-3104, 28th JPC, Nashville, TN, 1992.intermediate power hydrogen arcjet and found toincrease with decreasing power or increasing specific 8. Crofton, M. W., "Spectral Irradiance of the 1 kWenergy. Ionization losses tended to be only slightly Arcjet Thruster from 80 to 500 nm." AIAA papersensitive to changes in operating conditions, while 92-3237. 28th JPC, Nashville, TN, July 1992.thermal losses tended strongly decrease with increasesin flow rate. 9 Zube, D. M. and Myers R. M., "Nonequilibrium in

a Low Power Arcjet Nozzle", AIAA Paper 91-

Analysis of the results of a numerical model of the 2113, 27th JPC, Sacramento, CA, 1991.

arcjet was conducted to examine the complete power 10. Vincenti, W. G. and Krueger, C. H., Jr.,balance for some of the conditions tested. This analysis "Introduction to Physical Gas Dynami,

Introduction to Physical Gas Dynamics",did not discover a particular loss mode that offsets Malabar, FL: Robert E. Krieger Publishing Co.,Malabar, FL: Robert E. Krieger Publishing Co.,changes in the thermal losses to keep the thrust 1982.efficiency relatively unchanged. However, it didindicate that, in addition to thermal losses, ionization I1. De Vos J. C., "A New Determination of theand dissociation, random translational motion at the Emissivity of Tungsten Ribbons", Physica Vol.exit consumes a major fraction of the input power. This XX, pp. 690-714, 1954.result suggests the importance of measuring heavyparticle temperature, as well as dissociated species 12. Gubareff, G. G., Janssen, J. E. and Torborg, R. H.,densities and velocities, in understanding energy loss "Thermal Radiation Properties Survey",mechanisms in arcjets. Minneapolis: Honeywell Research Center, 1960.

7

2015 IEPC-93-216

13. King, D. Q. and Butler, G. W. "Modeling and 15. Butler, G. W., Kull, A. E. and King. D. Q.,Measurement of N2 Arcjet Performance". AIAA "Numerical Simulation of Hydrogen Arcjetpaper 93-2616, 21st IEPC, Orlando, FL, 1990. Performance". IEPC paper IEPC 93-249, 23rd

IEPC, Seattle, WA, 1993.14. Butler, G. W., Kashiwa, B. A. and King, D. Q.,

"Numerical Modeling of Arcjet Performance". 16. Cappelli, M. A., et.al., "A Comparison of ArcjetAIAA paper 90-1474, 21st Fluid Dynamics, Plume Properties to Model Predictions". AIAAPlasma Dynamics and Lasers Conf., Seattle, WA, paper 93-0820, 31st Aerospace Sciences Meeting1990. and Exhibit, Reno, NV, 1993.

REGENERATIVE HEAT TRANSFER TO FLOW

OHMIC LOSSES IN ELECTRODECURRENT ATTACHMENT -- N, 10CONVECTION TO WALL THERMAL LOSSESRADIATION TO WALL

NET RADIATION ACROSS NOZZLE EXIT

IONIZATION

FROZEN DISSOCIATIONFLOWLOSSES ELECTRONIC EXCITATION

MOLECULAR VIBRATION

MOLECULAR ROTATION

ELECTRICALPOWER RANDOM THERMAL ENERGYIN

RADIAL COMPONENTS OF EXHAUST VELOCITY

PROFILE LOSSES

USEFUL THRUST

PROPELLANT ENTHALPYAT INLET

Fig. 1. Arcjet Energy Flow Diagram

SPECTROSCOPICMEASUREMENTREGION

S 200

1.27 mm 1.27 mmS(0.050")) II

L/ J1.27 mm IIi(0.050") 0 - I

CATHODE 11 20.1 mm'/ 0 1 (0.791") 0

^ 0.5 mm -ANODE II

AREA RATIO 250:1- 2.75 mm

Fig. 2. Arcjet Electrode Configuration

8

IEPC-9316 2016Load I Intensity DataLoad0

PMT andCoupling .CoolerOptics

SPEX 500M

S -- Spectrometer

386 PCWavelength

SPEX CD2A ControlController

Position Control

Test Chamber

Vertical AretTranslation

Stage

Fiber Optic Cable

CollectionOptics

Fig. 3. Schematic Diagram of Spectroscopic System

181rF lO*F

15)F q AN DE 243F 46P°F 9 0°F 1840°F 2500*F - 4700F

A N O D E

2060F - 2380OF

q CATHODE '

q MOUNTARCJET BODY ANODE

210°F

190°F

THERMOCOUPLE LOCATION

FILTERED CAMERA MEASUREMENTS

Fig. 4. Thermal Loss Calculation Control Surface

9

2017 IEPC-93-216

RGB PCMONITOR MONITOR

SRGB SOFWARECAMERA OUTPUTAY

VIDEO WINDOWSOUTPUT

OVERLAYFRAMEGRABBER \

-- BOARDSYNC (OPTIONAL

CID CAMERA

386 PC

Fig. 5. CID Image Analysis System

3.50E+14 0 10 kW; 150 MJ/kg3.oo00E+14 ooo3.0E+14 o0 o 10 kW; 200 MJ/kg2.50E+14 a w

C?...E. n 14kW ; 15 0 MJ/kgS2.00E+14 o ;E -

S1.50E+14

S1.00E+14 ^ a^5.00E+13 O g0000

0.00E+00

0 2 4 6 8 10

Radial Position (mm)

Fig. 6. Exit Radial Profile of Electron Number Density

10

IEPC-93-216 2018

3000

S28000

. 2600 r I 10kW,200MJ/kg

* 2400I 2400 0 1OkW,150MJ/kgE 2200I- 2A 14kW,150MJ/kg$ 2000

to

: 1800

S1600 __

0.0 1.0 2.0 3.0 4.0 5.0

Axial Distance From Exit Plane (cm)

Fig. 7. Arcjet Anode Temperatures

150 MJ/kg 10 kW

1.00 1.00

CL 4 0.80 0. 0.80

o a. 0.60 o o 0.60

0s 0.40 - 0 0.40

=e o 0Si 0.20E u 0.20Uu.

0.00 .. 0.006 8 10 12 14 150 200

Input Electric Power (kW) Input Specific Energy (MJ/kg)

* Thrust E Thermal U Ionization D Other

Fig. 8. Arcjet Power Balance

11

. 0.40

- -10 kW; 200 MJ/kg2019 IEP-93-216-0.31

o 0.30-9

* 0.25W 0.20' 0.20 0.170. 0.15S0.15o 0.10c 0.10o 0.05 0.061 0.05 0.007 0.02 0.008a-

u- 0.00

0.40 0.36

o 0.35 10 kW; 150 MJ/kg

o 0.30a-

, 0.25 0.22wS 0.20a. 0.16S0.15

0 0.09C 0.10 0.08 0.06 0.070 0.06o 0.050.05 0.003 0.02 0.009u. 0.00

0.40 0.370.35 . 14 kW; 150 MJ/kg

. 0.30

* 0.25 0.23w5 0.20

. 0.150.15 0.12

0 0.09C 0.10o 0.06 0.06o 0.05

hO 0.003 0.01 0.01IL 0.00

Fig. 9. Arcjet Power Balance From Numerical Analysis

12