ENGINEERING AND INDUSTRIAL EXPERIMENT STATION College · PDF fileENGINEERING AND INDUSTRIAL EXPERIMENT STATION College of Engineering ... Criticality calculations, ... capillary and

Technical Status Report

on

Experimental Investigations of a Uranium Plasma

Pertinent to a Self Sustaining Plasma Source

ENGINEERING AND INDUSTRIAL EXPERIMENT STATION

College of Engineering

University of Florida

Gainesville

https://ntrs.nasa.gov/search.jsp?R=19720013030 2018-05-05T17:39:23+00:00Z

Technical Status Report

on

Experimental Investigations of a Uranium Plasma

Pertinent to a Self Sustaining Plasma Source

by

Richard T. Schneider

Contract NGL 10-005-089

University of Florida

Department of Nuclear Engineering Sciences

Gainesville, Florida

December, 1971

Table of Contents

I . Introduction

II. Uranium Plasma Emission Coefficient in the Visible

and Near U.V.

III. Generation of a Uranium Plasma Near Gaseous Core

Reactor Conditions

IV. Ballistic Piston Fissioning Plasma Experiment

V. COj Laser Experiment Using Nuclear Reactions as

the lonization Source

VI. Temperature Profile Determination in an Absorbing

Plasma

VII. Publications Generated under Grant NGL 10-005-089

I. INTRODUCTION

This research carried out under NASA Grant NGL 10-005-089

is pertinent to the eventual realization of a self-sustained

fissioning plasma for applications such as nuclear propul-

sion, closed cycle MHD power generation using a plasma core

reactor, other heat engines such as the nuclear piston engine,

as well as the direct conversion of fission energy into

optical radiation (Nuclear Pumped Lasers).

Operation of plasma core nuclear reactors for propulsion

or other applications requires knowledge of the radiation

emitted by uranium plasmas as a function of operating pres-

sures, temperatures, wavelengths, and geometry. The knowledge

of partial pressures of the various uranium ionic species and

the electrons is mandatory. Criticality calculations, pre-

diction of radiative heat transfer from the gaseous fuel to

the working fluid, and design of system startup and control

devices require this information.

In order to measure the basic optical radiation proper-

ties needed, diagnostic measurement methods and experimental

devices simulating plasma core reactor conditions must be

developed. The program described herein is aimed toward

this goal.

Due to the complexity of the procedures involved inp-i r

handling the U isotope, an effort is made to do all

research connected with the optical properties of a uranium

plasma with natural uranium and to use enriched uranium only

235for in-core experiments, where the use of U no longer can

be avoided.

The program consists of several experiments, each one

being specialized to obtain a certain class of information.

These experiments are:

235A. Ballistic Piston Compressor (U )

B. High Pressure Uranium Plasma (natural uranium)

C. Sliding Spark Discharge (natural uranium)

3 235D. Fission Fragment Interaction (He and U )

3 235E. Nuclear Pumped Lasers Experiments (He and U )

These individual experiments shall be described

briefly.

A. Ballistic Piston Compressor

The ballistic compressor consists of a pressure

reservoir, a gun barrel and a high pressure test section.

A free traveling piston is accelerated by the high pressure

in the pressure reservoir and compresses the test gas into

the high pressure test section. Pressures up to 6000 atm

and temperatures up to 10,000°K can be reached using helium

as test gas. For the experiment in this program a mixture

of UF, and helium is used. Effective y* temperature and

density of this mixture is determined for a wide range of

parameters. The high pressure test section will be subjected

to neutron bombardment and the non-equilibrium effects caused

by the fission fragment will be investigated. (See also

Section IV. Ballistic Fissioning Plasma Experiment).

B. High Pressure Uranium Plasma Experiment

This experiment involves the investigation of optical

properties of uranium. A high temperature, high pressure

uranium plasma arc device is used. The arc is contained in a

high pressure cell capable of withstanding pressures up to

100 atmospheres. The arc can be operated in two different

modes. In one mode a helium cover gas is applied; the

other mode is a vacuum arc. In the high pressure mode the

emitted line radiation originates primarily from singly-

ionized uranium. In case of the vacuum arc, the line

radiation observed stems from neutral uranium. Employing

different spectroscopic diagnostic techniques, temperatures,

particle densities, and emission and absorption coefficients

of the uranium plasma for different operating conditions are

measured. (See also Section II. Uranium Plasma Emission

Coefficient in the Visible and Near U.V.).

C. Sliding Spark Discharge

The goal of this experiment is the generation of uranium

ion lines. A capacitor bank is discharged through a capil-

lary. The capillary is formed by a U0_ sinter body. The

sliding spark will evaporate the inner surface of the

capillary and subsequently ionize the vapor. The power

density is very high due to the constriction of the discharge

by the capillary. Therefore multiple ionization is observed.

Some of the components of the experiment were given to us by

NASA-Langley on a loan basis. (See also Section III.

Generation of a Uranium Plasma Near Gaseous. Core Reactor

Conditions).

D. Fission Fragment Interaction Experiment

This experiment involves a glow discharge in argon, helium

or the C02 laser gas mixture. In the case of He-gas fill, the

He isotope is used. Under influence of an external neutron

flux protons and tritons are generated by the He (n,p)T

reaction. For other gas fills, the plasma is surrounded by

a quartz tube, which is in the inside coated with U Oo'

The fission fragments formed close to the surface of the

235U 02 layer will escape and interact with the glow dis-

charge in the latter case, while in the first case the high

energy protons and tritons will fulfill this function. It

is expected that this will result in enhanced ionization.

This can be detected by analysis of the I-V curve of the

glow discharge. Also, Boltzmann plots of the intensities

of the detected spectrum lines can be used to measure

deviations of the excited states from a Maxwell-Boltzmann

distribution.

E. Nuclear Pumped Laser Experiment

This experiment involves in-core testing of a C02 gas

laser to demonstrate the feasibility of nuclear enhancement4

of the laser power output. The He of the laser gas mixture

is replaced by the He isotope. The He (n,p)T reaction is

employed to yield high energy protons and tritons, which

may enhance the ionization of the laser gas in direct or

indirect ways. Further testing with fission fragments are

planned. Also testing with liquid uranium compounds as laser

medium are planned. (See also Section V. CC>2 Laser Experi-

ment Using Nuclear Reactions as the Ionization Source).

II. URANIUM PLASMA EMISSION COEFFICIENT IN THE VISIBLE AND NEAR U.V.*

J. M. Mack, Jr., J. L. Usher, R. T. Schneider, and H. D. CampbellDepartment of Nuclear Engineering Sciences

University of FloridaGainesville, Florida

Abstract

Measurements of the specific emissioncoefficient in the near ultra-violet andvisible region of a uranium arc plasma arereported. Spatial unfolding of theintensity profile is used to determine theemission.coefficient in the spectral rangeof 2000 A to 6000 A. The uranium partialpressure is estimated to range between.001 and .01 atmosphere, and the corre-sponding temperature range is 5000 -10,000°k.

Introduction

To develop the technology necessaryfor the implementation of gas-corereactors, it is necessary to determine theemission coefficient of uranium plasmas atdifferent temperatures, partial pressures,and wavelengths. This property is funda-mental to the design considerationsrelated to criticality studies and radia-tive heat transfer from the plasma core tothe working fluid. The interface condi-tions which are likely to exist at thecore-fluid interface can be simulated witha D.C. uranium arc. Uranium, being one ofthe most complex elements, does not easilylend itself to a theoretical treatment ofthe emission coefficient (See References(1) and (2) for the most recent theore-tical attempts to describe the uraniumatom). Thus, experimental investigationsof uranium plasma emission are not onlyfundamental to gas-core reactor design butalso serve to support or deny theoreticaltreatments.

Average uranium plasma emission coef-ficients of a D.C. uranium arc have beendetermined previously(3) with an arc inwhich the emitted radiation was due pri-marily to singly-ionized uranium. Uraniumemission coefficients due primarily toexcited atomic uranium are presented inthis paper. These measurements arecorrelated to temperature and particledensity of the plasma.

Uranium Plasma Generation

The uranium plasma is generated withina chamber identical with that of Figure 1,Reference (3). Internally, however,arranged concentrically with the cathode-anode configuration, are several annularwater-cooled copper segments (disks).These segments are distributed the lengthof the cathode and arc column to form awall-stabilized arc. The arc is also runin low vacuum on the order of 3 Torr.

Wall-stabilization and evacuation of thearc chamber greatly enhance arc stability.Power input to the arc is limited to about1000 watts. The length and diameter ofthe arc are typically 1 cm and .9 cm,respectively.

The operation of the water-cooled seg-ments in the vicinity of the arc columnproduces a cooling effect on the arc as awhole. This cooling results in an arcwhich emits primarily by neutral uraniumrather than by singly-ionized uranium.Although singly-ionized uranium lines arefound, the atomic uranium lines appear tobe the dominant radiation. This is con-firmed by the spectral line identificationperformed with this arc.

Data Acquisition System (DAS)

There are two basic sections of the DAS:(1) that which measures the line profile(density), plasma temperature, and lineabsorption, and (2) that which measuresthe arc intensity. The componentsnecessary to measure arc intensity areshown in Figure 1.

LEGEND:

0, :OSCILt.OSCOM I1

FT.FAiKI-TIK SIIIIAL AVOun> :AKC POSITION

FIGURE I DATA ACQUISITION SYSTEM

*This work performed under NASA Grant NGL-10-005-089

A fast-scan spectrograph S2 receives arclight reflected from the front surface ofa beam splitter. The phototube responseis then monitored and stored digitallyby the Fabri-Tek signal-averager. TheFabri-Tek is a time-averaging device whichintegrates over the small arc fluctuations.This averaging results in very reproducta-ble arc intensity traces as a function ofwavelength. Four storage areas on theFabri-Tek are used for storage of thespectral intensity, I(X,x),where x is aparticular position in the arc traverse.Thus, a four point Abel unfolding forspatial resolution of the arc intensitiescan be performed to obtain the spectralemission coefficient, e(X,r), where r isthe distance from the arc center.Oscilloscope 02 is used as a Fabri-Tektrigger delay unit which enables thesignal-averager to be triggered at anywavelength. The photomultiplier used withS2 is an EMI-9514 with a sodium salicylatewindow which acts as a wavelength shifterfrom the ultraviolet to the visible. Thisserves to expand the phototube systemresponse down to 1900 A with highsensitivity.

Also shown in Figure 1 are thosecomponents necessary for the measurementof temperature, line absorption, andparticle density. The beam splitterpasses a portion of the arc radiation toSi. A rotating refractor plate (quartz)sweeps a particular line of interestacross the exit slit plane of SI. An RCAIP28 phototube placed behind the exit slitresponds to the spectral line as it isswept past the exit slit. The resultingline profile is recorded by 01 and 02. Atthat instant in time when the line isswept to its peak value, a linear Xenonflashtube (EGG FX-12-.25) is triggered(see Figure 2). Timing is accomplished byvarious electronic delay circuitry. Fromthe oscillogram of 02 is determined theparticle density by analysis of the abso-lute line profile. Oscillograms from 01allow the determination of line absorptionand plasma temperature.

Two tungsten calibration lamps areused to assign absolute units to intensi-ties. One of the lamps is calibrated bythe National Bureau of Standards, and itspurpose is to periodically cross-check thesecond lamp which is used in the actualcalibration of arc intensities for eacharc burn.

Measurement of Plasma Parameters

I. Line Absorption and Temperature

The absorption coefficient of aspectral line is determined by using abackground source of constant intensityand measuring the amount it is attenuatedas it passes through the plasma. Byassuming the nonhomogeneous plasma to becomposed of homogeneous rings, anunfolding technique is used to calculate

the line absorption coefficient,from the following equation:

• d)kk

where V , V.T and V.p are voltage signalsfrom a photomultiplier tube. The voltagescorrespond to the background source, thetotal attenuated intensity, and the plasmaintensity, in that order. Figure 2 showsa typical oscillogram of the photomulti-plier output.

FIGURE 2

OSCILLOGRAM OF PHOTOMULTIPLIER OUTPUT

The background source, in this case aXenon flashtube, is triggered on the lineof interest as shown on the upper trace.The lower trace shows the flash tubesignal spread to facilitate measurementof its intensity. The oscillograms arerecorded at different chordal positionsof the plasma and Equation (1) is used todetermine the line-absorption' profile.The fcn,m iR Equation (1) represent thelength segments in the mtn ring along thentn line-of-sight position. Figure 3is the absorption profile determined for auranium arc operating with 1000 wattspower input.

3S53 I i3CS9 1 ' UI

Lfls5I—*

§

LINE ABSORPTION COF.FFICIF.NT PRCFILF.

The temperature is determined by usingan extension of the brightness-emissivitymethod to the nonhomogeneous case^1*^.The technique makes use of the voltages

measured previously as well as the calcu-lated line-absorption values. Thetemperature method requires that thetemperature of the background source beknown. The average temperature in the ktn

ring is determined from the followingequation,

C2T

where:

r Bsloge[l + sf -I)

-1(2)

Bg, B = the Planck functions representing* the intensities of the background

source and the kfch plasma ring,respectively, at wavelength

T. = the brightness temperature of thebackground source

Bfc is calculated using the measuredvoltages and the calculated absorptionprofile. The complicated expression forBfc is not written here. A*"uranium tempera-ture profile is displayed graphically inFigure 4 where the power input is 1000watts.

inui

•U25 0.25 •:,.375 M.S ; 625 '...75 '.i.875 i

: 4 R/(;;o

:./ IT.MTWJATl IR F'RTFln

II. Determination of Ul Number Density

The absolute line-intensity methodis used to estimate the Ul number density.The general equation for a spectral lineis:

hvui.4TT

N Su (3)

Utilizing the BoltEmann factor gives thefollowing for the case of negligibleionization:

N4ir U ( T )

s (4)

where :

u = upper stateI = integrated line intensity in energy

per area and timeI = lower state

h

NU* =

No =

U(T) =

gu =skT =

transition probability in I/timePlanck's constantfrequency of emitted photonnumber density of particles in theexcited statenumber density of particles in theground statepartition function of the particlesin the ground statestatistical weight of state Uplasma depth in cmthermal energyexcitation energy

The0integrated line intensity of the3654 A Ul spectral line is found bysweeping the line across a phototube andrecording the response with an oscillo-scope (see Figure 2) . The same procedurewith identical optics is done with anaccurately calibrated tungsten strip lamp.The two recorded responses are then com-pared thus defining the line intensityprofile in absolute units. The tempera-ture is determined by the modifiedbrightness-emissivity method, and "guAu£"values are found in (5); U(T) isapproximated with the ground statestatistical weight.

Solving Equation (4) for No for severalarc burns at 1000 watts power inputresults in UI number densities rangingfrom 1015-1016 t/cm3 for the vacuumsegmented arc.

III. Determination of Uranium PlasmaEmission Coefficient

To determine the emission coefficientit is first necessary to experimentallymeasure the arc intensity. For the homo-geneous case the emission coefficient isfound by considering the depth of theplasma. However, the uranium arc is anonhomogeneous plasma, and so it isnecessary to measure the arc intensityat several different line-of-sightpositions. Due to the nonhomogeneousnature of the arc, spatial resolution isapplied to the intensity to obtain theemission coefficient.

Although some uranium lines havesignificant absorption, the plasma as awhole will have -a small absorption co-efficient. Spatial resolution isperformed by application of the familiarAbel transform given by Equations (5) and(6) to the measured arc intensities:

Ky)

x.

L e(r ) dx,

and

Ky>> e(r) r dr

/r2 - y2

(5)

(6)

where:I(y) = intensity from plasma at particular

lihe-of-sight observation,e(r) = plasma emission coefficient at

chorda1 locations.

A four-point finite-difference form ofEquation (6) is used to transform themeasured arc intensities into emissioncoefficient values at four arc positions.The bandpass of the intensity measurementis approximately 1 A. The resultingcalibrated intensities are shown inFigure 5. These intensities are thencomputer-averaged over a handwidth of100 A. The results for a power input tothe arc of 1000 watts are shown in Figure6 for 2000-6000 A. The average emissioncoefficient is shown in Figure 7 for thesame conditions. Figure 8 is the center-line intensity of a Helium-uranium arcoperated at 15 amperes. The emissioncoefficients are compared with pastresults(3) and with theory (1) in Figure9. Those points with total pressure (Pip)refer to the He-U arc, while those withpartial pressure (PUI^ refer to thevacuum segment arc.

MTUKLIZATICN = 0-305E 05

NCRMLIZATIO4 = 08

100

9000

URANIUM ARC INTENSITY VS WAVELENGTH

KJMALIZATICN = OB

100

50

3000 2500 4000 4500 "5000 "5500 GOOO

ANGSTROMS

AVE- U-ARL INTENSITY VS WAVELENGTH

• .

3a

$in»

*.*-

aooo 25oo 3000 3500 4000 4500 5000 gsbo GOOD

ANGSTROMS

EMISSION COEFFICIENT

OS

AVE-

Ml

tn

75!

50

35

°aooo S3ss Z75S xer 3450 3Bia 4175FIQ : 8 A*3STRCM5

HE-URANIUM INTENSITY VS WAVELENGTH

430O

KM

b.HIOU

oCO

w

LEGEND:X»Vocuum ore

9*N«w MtoiurtmtntiH«— Uranium

3OOO 4OOO 500C MOO

WAVELENGTH. I

FIGURE 9 COMPARISON OF RESULTS

Concluding Comments

I. Accuracy

The accuracy of measurements of plasmaproperties is frequently difficult toassess, and this is certainly the casewith many uranium plasmas. Of the threeplasma properties determined, -e^» T, N,certainly the particle density was themost difficult to measure while theemission coefficient was determined with ahigh degree of certainty.

The temperature measurements by themodified brightness-emissivity methodresulted in values somewhat lower thanmight have been expected. However exami-nation of the spectra revealed that UIwas the dominant radiating species ratherthan UII, as had been the case in earlierwork*6). AS explained earlier this lowertemperature was attributed to the coolingeffect of the water-cooled segments on thearc. It should be mentioned that thevalidity of the present temperaturemeasurements rests upon the assumption ofpartial L.T.E. (local thermal equilibrium)down to the lower levels of the transi-tions studied. It still remains a ques-tion as to whether or not the vacuumsegmented arc satisfies this condition.

There are three factors which couldaffect the accuracy of the particledensity determination. As in the tempera-ture measurement, it was necessary toassume that L.T.E. conditions prevailedin the plasma. Self-absorption frequentlyplays an important role in the determina-tion of the total intensity of spectrallines. Fortunately however, the tempera-ture measurement technique used hereyields a value of the optical depth whichwas used to account for this effect.Finally, there is the uncertainty asso-ciated with the transition probabilitieswhich had to be used in this measurement.

II. UI Emission

Figure 5 for the UI vacuum arc clearlyshows that there-is no significant radia-tion below 3500 A. On the other hand,the intensity from the higher temperatureUII arc (see Figures ) is strong in thespectral range of 3000-3500 A. Hence,it is evident that most of the U.V.radiation is generated by ionized uranium.

III. Future Work

It was found that operation of ashorter vacuum arc at 1000 watts resultedin a much more intense uranium plasma.Since the arc column was shorter, thewater-cooled segments did not have a sig-nificant arc-cooling effect and, thus, thearc became predominantly a UII plasma —but one with instabilities. Future planswill be to make a longer arc with higherpower input to hopefully obtain a very

stable UII vacuum arc. Similar measure-ments to those reported in this paper willbe taken on this arc.

Acknowledgements

The authors would like to thankstudents N.A. Smith, D. Baker, and J.Dixon for their help throughout the year.Special appreciation is in order for theassistance given by K. Fawcett in realizingthe complex circuitry necessary for thecompletion of this experiment. Finallythanks are in order for the help Dr. C. D.Kylstra has given in building the segmentedarc.

Bibliography

1. D. E. Parks, G. Lane, J. C. Stewart,and S. Peyton, "Optical Constants ofUranium Plasma," Report GA-8244(NASA-CR-72348), Gulf General Atomic

. Incorporated, February, 1968.

2. A. S. Keston and N. L. Krascella,"Theoretical Investigation of RadiantHeat Transfer in the Fuel Region of aGaseous Nuclear Rocket Engine," ReportNASA-CR-695, United Aircraft Corpora-tion, January, 1967.

3. C. D. Kylstra, R. T. Schneider, andH. D. Campbell, "Uranium PlasmaEmission Coefficients," AIAA PaperNo. 70-692, presented at the AIAA 6thPropulsion Joint Specialist Conference,San Diego, California, June, 1970.

4. J. L. Usher, "Temperature ProfileDetermination in an Absorbing Plasma,"Master's Thesis, University ofFlorida, August, 1971.

5. C. H. Corliss and W. R. Bozman,"Experimental Transition Probabilitiesfor Spectral Lines of Seventy ElementsDerived from the NBS Tables of SpectralLine Intensities," National Bureau ofStandards, Monograph 53, Superintendentof Documents, U. S. Government PrintingOffice, Washington, 1962.

6. H. D. Campbell, R. T. Schneider, andC. D. Kylstra, "Properties of aUranium Plasma," presented at thePlasma Dynamics Session of the AIAA8th Aerospace Sciences Meeting,.NewYork, 1970, reprint #70-43.

III. GENERATION OF A URANIUM PLASMA AT NEAR GASEOUS CORE REACTOR CONDITIONS*

J. F. Davis, III, D. G. Schnitzler, and R. T. SchneiderDepartment of Nuclear Engineering Sciences

University of FloridaGainesville,Florida

Abstract

A constricted sliding spark dischargeis used to generate a high density, hightemperature uranium plasma. Uraniumparticle densities up to 1020 cm"3 areobtained over a temperature range of30,000 K - 50,000°K. The device consistsof a capillary discharge channel linedwith pressed and sintered UOZ . A 250joule capacitor bank is discharged intothe channel, producing a plasma of 10-20Usec duration. Spectroscopic observationsare made over the spectral range of 1300 Ato 2500 A.

Introduction

The properties of a high temperatureuranium plasma are of great interest tomany researchers, particularly for devel-opment of the Plasma Core Reactor includ-ing the Nuclear Light Bulb EngineI1'*• 3'.For the latter case, information is neededto determine the amount of radiationemitted at spectral regions where thetransparent wall is opaque^*'5'. Theoperating conditions of a uranium plasmain the plasma core reactor are expectedto be in the range of 20,000°K to 40,000°Kwith pressures of 500-1000 atmospheres.Several experiments have been performedto produce and/or simulate a modeluranium plasma at these conditions ofhigh temperature and pressure.

The DC arc has the advantage of pro-ducing a plasma that is stable atsteady-state conditions (6f7).Campbell, et al., have produced a stableHelium-Uranium arc at pressures up to 10atmospheres and temperatures up to10,000°K(6'. Such operating conditionssimulate the properties of the coreworking fluid interface of a plasma corereactor.

An R-F discharge has also been appliedto simulate the conditions in a gaseouscore reactorI6/9). A temperature of upto 7,000°K has been obtained at pressuresup to 20 atmospheres.

Another method under development toform a model plasma is piston compressionof uranium hexa'fluoride gas to form ahigh density plasma*10). Pressures up to500 atmospheres of a He-UF6 mixture havebeen reported; however the temperature isexpected to be of the order of 1000°K.

The sliding spark discharge asdescribed in this paper has a shortlifetime and small volume, but has theadvantage of producing a high density-high temperature uranium plasma,approaching gaseous core reactor condi-tions.

The vacuum spark has been found to bean excellent source for studying thespectra of metals in the extreme ultra-violet region. Vodar and Astoin dis-covered that a high vacuum spark dischargecould be produced at lov.'er voltages andwith less sputtering of electrode materialif the electrodes were separated by. butin good contact with, an insulator^ ).The spark discharge occurs on the surfaceof the insulator, therefore the name"sliding spark discharqe". Manymodifications of the slidingspark have been published"12''3>l"•'5>.The emitted spectrum is characteristic ofthe electrodes and insulation material.The plasma produced consists of a welldefined and relatively long dischargechannel. Very high light intensities canbe obtained. Spectra with up to 500 eVexcitation energies have been observedwith a sliding spark.

Description of Apparatus

Conrads has performed considerablework in the area of sliding spark dis-charge*13'l "•1S'*7) . A similar apparatusis employed in the present experimentwith some modifications to introduce theuranium.

The experiment consists of two carbonelectrodes held firmly against a highdensity polyethylene insulator throughwhich a discharge channel is drilled.The carbon electrodes, polyethyleneinsulator and the uranium insert areillustrated in Figure 1. The uraniuminsert is composed of U02 powder whichhas been pressed and sintered to form ahard, brittle ceramic of the desiredsize.

The carbon electrodes are positioned2 cm apart and the discharge channel is1.4 to 2 mm in'diameter. These dimen-sions define the plasma volume which isviewed axially through the electrodesas shown in Figure 1.

*This work was performed under NASA Grant NGL 10-005-089.

U02 INSERT

lithium fluoride window also separates thetwo vacuum systems so that the spark pres-sure can be varied. The short wavelengthcutoff of the window is approximately1200 A. The film used is Kodak SpecialType 101-01.

Description of a Discharge

Typically, the capacitor bank is chargedto 12 kv and the residual gas pressure is100 to 140 microns. The vacuum spectro-graph is operated at a pressure of lessthan 5xlO~ microns. The spark dischargesthe 250 joule bank energy into the capil-lary channel. After one full cycle(approximately 10 ysec) most of the capa-citor bank energy is dissipated. A plotof current and photomultiplier outputversus time for a typical shot is shownin Figure 3.

Fig I. Schematic of Sliding Spark Discharge

A schematic of the circuit is shown inFigure 2. The capacitor bank consists ofseven parallel plate capacitors rated at0.5 pF, 30 kV each. A very low self-inductance of 0.3 nanohenries per capaci-tor permits a very fast discharge. Thebank is connected through a spark gapswitch to the electrodes by coaxial cables.The discharge is initiated by triggeringthe spark gap switch with a 30 kV triggerpulse. After one cycle the spark gapbecomes non-conducting breaking thecircuit.

SLIDING SPARK

P ^"^

SPARK GAP SWITCHi 1

1

HV POWER

SUPPLY

1

30 KVTRIGGER

DUAL BEAMOSCILLOSCOPE

Fig 2. C i r cu i t Diogrom

Spectroscopic observations are madewith a vacuum spectrograph set on a 1meter Rowland circle. The spectrographhas a concave grating with 1200 lines/mmyielding a dispersion of about 8 A/mm.The spectrograph slit and grating areprotected from the high velocity particlesthat ablate off the electrodes and the U02insert by a lithium fluoride window. The

FIG. 3 UPPER TRACE IS CURRENT (20 kA/cm)VS. TIME (20 usec/cm)LOWER TRACE IS PHOTOMULTIPLIER(10 volts/cm) VS. TIME (20usec/cm)

The photomultiplier output indicates aplasma duration of approximately 200 usec.It has not yet been determined if fluores-cence of the surrounding insulation andstructural materials or a true afterglowcausing the long pulse duration.

Results

An estimate of the plasma temperaturecan be made from the ionization stagespresent. In the 1500-2500 A region, thethree most intense lines observed were1548.18 A (CIV), 1550.71 A (CIV), and2296.87 A (CHI). The plasma temperaturecan then be approximated by the typicalnormal temoerature of the carbon IVspecies, 40,000 °K< 1 8>.

The maximum pressure produced in thedischarge channel is assumed equal to themagnetic pressure. The magnetic pressure,P , was calculated from Equation 1.

P =8ir 2 R 2

(1)

With a discharge channel radius of 0.75nun and a typical peak current of 28 kamp,the magnetic pressure obtained is 220atmospheres. This pressure correspondsto a particle density of 4 x 1019 cm"3

at 40,000 OK.

The capacitor bank discharges with afull cycle time of approximately 10 ysec.The peak plasma temperature is rapidlyachieved. Maximum pressure occurs at themaximum current condition. At thisinstant plasma can then be observed atnear plasma core reactor conditions of40,000 °K and 220 atmospheres.

Complete analysis of the spectraproduced involves position and densitydeterminations of a large number oflines. Manual analysis with a measuring .microscope requires a prohibitive invest-ment of manpower. The spectrum analysisfor this paper was made with the Univer-sity of Florida Automatic SpectrumAnalysis System^19'. The plate wasscanned by a Jarrell-Ash scanningdensitometer interfaced with an AD-80analog computer and an IBM-1800 DataAcquisition System. The collected datawas processed by programs written at theUniversity of Florida which convert theraw data from the densitometer to trans-missions which can be analyzed by thecomputer. Dispersion relations areautomatically calculated from the locationof a number of user selected and identi-fied emission lines. Line positions areautomatically identified and wavelengthsassigned. Transmission plots and listsof lines with corresponding wavelengthsare among the available output informa-tion.

Plates were scanned at approximately40 microns/sec and sampled every 4 micronsof plate movement. Relative plate posi-tion is indicated by a precision linearencoder attached to the plate carriage.Approximately 32,000 positions weresampled while scanning the 1500-2500 Aregion.

The 70 mm Kodak Special Film Type 101-01used has a 5.25 mil thick backing.Because of the thin backing, the film hasa tendency to curl away from the filmholder under high vacuum conditions. Thefailure of the film to conform exactly tothe Rowland circle introduces irregulari-ties in.the calculated dispersion rela-tion. This appears to be the main sourceof error in wavelength assignment. Tominimize these effects, the spectra wasanalyzed in short segments to which adispersion relation could be fitted frompreviously observed uranium lines(20).The detailed analysis of one such regionis rresented here.

The transmission plot shown inFigure 4 extends from 2242 A to 2338 £.This region was analyzed first because

of the abundance of new lines appearingwith the addition of uranium to the dis-charge. QIn addition, the prominent2296.87 A Carbon III line and severalknown Uranium I lines contained in thisregion made this spectral region an attrac-tive starting point.

The parameters in the line findingprogram were selected to locate as manylines as possible automatically. Of the134 lines located by the program, 16 wereeliminated as too weak to be significantor as possible noise. Visual examinationof the transmission plots revealed threelines not found by the program. Seven ofthe remaining lines were identified ascarbon and eight as oxygen*21'. Seventy-four previously observed uranium lineswere found. Thirty-two lines remainunidentified. A complete listing ofthese lines and corresponding wavelengthsis shown in Table I.

A 50 Angstrom region extending from1530 to 1580 A is shown in Figure 5. Theoprominent emission lines are the 1548.18 Aand 1550.77 A Carbon IV lines. Alsopresent are Caorbon I absorption lines andsignificant continuum radiation.

The emission spectrum below 2000 A is ofspecial interest-. Addition of uranium tothe spark dischargeQcaused significantchanges belov; 2000 A. A large number oflines were observed which had not beenobserved in the carbon calibration shotsunder various conditions. The lack ofinformation on uranium emission below2000 A prevents rapid analysis of theobtained spectra.

Since the film integrates the incidentlight, radiation from the more highlyionized uranium species is extremelyweak or can even be obscured by theradiation emitted as the plasma cools.

Conclusions

Uranium emission below the transparentwall material cutoff is significant tothe Nuclear Light Bulb design and is alsoimportant for radiative power transfer inthe coaxial flow plasma core concept.This experiment shows that at near plasmacore reactor conditions a complicatedspectrum of line radiation exists.

Future work will be directed towardobtaining a comparison between emissionsbelow 2000 A and at longer wavelengths.A high speed rotating shutter will beused to observe the plasma at peaktemperature and pressure in contrast tothe present time integrated measurements.A higher pressure discharge can be pro-duced by reducing the diameter of thedischarge capillary or by raising thevoltage. Higher voltages will also allowlower initial residual gas pressure andthus reduction of impurities. The use

21'Jl HOURS 23 SEPTE^eefl 1971f-ARHJ-j . LFANILW

PLATE VS19 Lf> P.

O.QOOT FM- = 0-50000001 CM. • 0-01000000

G

1

03 = 3

WAVELENGTH,

VJA

FIGURE «t: TYPICAL CARBON PLUS URANIUM SPECTRA BETWEEN 22«t2 X - 2338

TABLE l: TABLE OF LINES CONTAINED IN FIGURE *»

1 22k2.1 C II 61 2285.66 U2 22k2.65 U 62 2286.U 0 III3 22k3.37 U 63 2287.2 0 Tilk 22kk 1 U ^ 22£Z'8e U

5 22kk.95 U 65 2288.96 22k6.l5 - 66 2289.32 U7 22k6.63 - 67 2289.78 22k7.08 U 68 2290.67 U9 22k7 33 - 69 2291.610 2218i03 U 70 2293.62 U11 2218.82 U 71 2296.87 C-III12 2?k9.31 U 72 2297.72 , U13 2250.2 - 73 2298.37 UIk 2250.5 - ^ 2299.2 U15 2251.09 U 75 i|°?'§? 216 2251.3 - 76 2301.51 U17 2251 7 - 77 2302.018 2252.US U 78 2302.69 U19 2252.90 U 79 2303.93 U20 2255.15 - 2° 230k.k U21 2255.68 C II 81 2305.65 U22 2256.19 C II £2 2306.91 U23 2256.88 - 83 2308.7 0 III2k 2257.15 - & 2308.9 U25 2258.02 U *$ 2309.3 U26 2258.8 - 86 2310.36 U27 ' 2259 6k U 8? 2310.67 U§8 2260.2 - *8 2311.68 0 III29 2260.5 - fi9 2312.56 TJ30 2261.U7 U 90 2313.2 U31 2262.36 U 91 23114-. 332 2262.79 U 92 231k.76 a III33 2263.26 U 93 2315.52 0 III3k 2263.67 • - 9k 2315.88 U35 226k.3 U ' ' 9? 2317.16 U36 2265.5 - 96 2317.37 0 III37 2266.0 - 97 23l8.k7 U38 2266.k - q8 2319.52 0 III39 ' 2266.97 TJ 99 2320.2kO 2268.53 U 100 2321.1kl 2269.7 C II 101 2322.5k2 2270.2 C II 1°2 232k.03 Uk3 2271.81 U 103 232k.8 Ukk 2272. k - 10^ 2325. k6 Itk5 2272.68 U 105 2326.k5 U

$ %i? u. 10° mi:!*k8 I27kk7 U 108 2328.52 Uk9 2275.12 U 109 232B.89 TT50 2276.05 U H° 2329.k6 IT<1 2277 02 U HI 2330.22 U;,i s si.I tig-* ". S|g 2281! 8k U 116 2335.8k II^7 2282 78 U 117 2336.k2 U58 2283 3k U 118 2336.9k Ucl 2281 72 U 119 2338.056? 22Rk83 U 120 2338.1,8 UOU ^ 121 2338.92 . U

5119 MXRS 34 SEPTEMBER 1371CATO* • i AMLId

PLATE VJ18 TJf 8

f 5A/vtR_E.. NUMBER

FIGURE 5: TYPICAL CARBON PLUS URANIUM SPECTRA BETWEEN 1530 A - 1580 A

of uranium electrodes will eliminate thecarbon. An attempt will be made to .determine the emission coefficient as afunction of wavelength at variouspressures.

Acknowledgements

The authors gratefully acknowledgeLou Ayers for his assistance with thecomputer analysis and Wayne Tuohig for hishelp in producing the UO2 inserts.

References

1. J. D. Clement and J. R. Williams,"Gas-Core Reactor Technology,"Reactor Technology, Vol. 13, No. 3,Summer 1970.

2. F. E. Rom, "Comments on the Feasibilityof Developing Gas Core Nuclear Reac-tors," Report NASA-TM-X-52644, NASALewis Research Center, October 1969.

3. G. A. McLafferty, "Investigation ofGaseous Nuclear Rocket Technology,Summary Technical Report," ReportH-9100093-46, United Aircraft ResearchLab, November 1969.

4. D. E. Parks, G. Lane, J. C. Stewartand S. Peyton, "Optical Constants ofUranium Plasma," Report GA-8244(NASA-CR-72348) Gulf General AtomicIncorporated, February 1968.

5. N. L. Krascella, "TheoreticalInvestigation of the Radiant Emission •Spectrum from the Fuel Region of aNuclear Light Bulb Engine, ReportH-9100092-12, UNRL, October 1969.

10.

11.

12.

H. D. Campbell, R. T. Schneider andC. D. Kylstra,- "Properties of aUranium Plasma," presented at thePlasma Dynamic Session of the AIAA8th Aerospace Sciences Meeting, NewYork, 1970, reprint #70-43. .

J. M. Mack, J. L. Usher, H. D. Camp-bell and R. T. Schneider, "UraniumPlasma Emission Coefficient in theVisible and Near UV," paper presented2nd Symposium on Uranium Plasmas:Research and Applications.

W. C. Roman, "Experimental Investiga-tion of a High-Intensity R-F RadiantEnergy Source to Simulate the ThermalEnvironment in a Nuclear Light BulbEngine," UARL Report J-910900-4, alsoNASA Contract No. SNPC-70,September 1970.

P. J. Marteney, A. E, Mensing andN. L. Krascella, "Experimental Inves-tigations of the Spectral EmissionCharacteristics of Argon-Tungstenand Argon-Uranium Induction HeatedPlasmas, Report G-910092-11, UARL,September 1968.

B. E. Miller, R. T. Schneider,K. Thorn and G. T. Lalos, "BallisticPiston Fissioning Plasma Experiment,"paper presented 2nd Symposium onUranium Plasmas: Research andApplications.

B. Vodar and N. Astoin, Nature 166,1029(1950).

H. Damany, J. Y. Roucin and N.Damany-Astoin, Appl. Opt. S_ (1966)297.

13. II. Conrads and H. Hartwig, Z.Angew Phys. IT. (1964) 18.

14. P. Bogen, H. Conrads and D. Rusbiildt,2. Physik, 186 (1965) 240.

15. G. H. C. Freeman, Proc. Phys. Soc.,86_ (1965) 117.

16. H. Conrads, Z. Physik, 200 (1967)444.

17. H. Conrads, Thesis T. H. Aachen 1966;Berichte der KernforschungsanlageJulich Nr. 360.

18. R. H. Fowler and E. A. Milne, MonNat Roy Ast Soc £3, 403 (1923); 8 4,499 (1924).

19. C. D. Kylstra and R. T. Schneider,"Computerized Spectrum Analysis,"Appl. Spec., 24, No. 1, Jan-Feb. 1970.

20. G. R. Harrison, "M.I.T. WavelengthTables," the M.I.T. Press, 1969 ed.

21. A. R. Striganoy and N. S. Sventitskii,"Tables of Spectral Lines of Neutraland Ionized Lines," IFI/Plenum DataCorporation, 1968.

IV. BALLISTIC PISTON FISSIONING PLASMA EXPERIMENT*

B. E. Miller and R. T. SchneiderUniversity of FloridaGainesville,Florida

K. ThornAEC/NASA Space Nuclear Systems Office

U. S. Atomic Energy CommissionWashington, D. C.

G. T. LalosNaval Ordnance Laboratory

Washington, D. C.

Abstract

The production of fissioning uraniumplasma samples such that the fission frag-ment stopping distance is less than thedimensions of the plasma is approached byusing a ballistic piston device for thecompression of UF$. The experimentalapparatus is described. At room tempera-ture the gun can be loaded.up to 100 torrDF« partial pressure, but at compressiona thousand fold increase of pressure canbe obtained at a particle density in theorder of 1019/ccm. In order to under-stand the effects of fission fragmentinteractions, the thermodynamics isstudied, for a comparison with subsequentin core measurements involving * 3SUF6.

I. Introduction

A research program has been initiatedto investigate the interaction of fissionfragments with a uranium plasma. Theultimate goal is a critical fissioningplasma, perhaps preceded by a subcriticalassembly.

For such a system a number of conditionshave to be fulfilled. The uranium particledensity has to be high enough to ensurecriticality. Neutron leakage dictates aminimum volume. The temperature should behigh enough to create sufficient ioniza-tion and excitation to ascertain plasmaconditions. These conditions areapproached in a device called"ballisticpiston compressor"using UF$ as the gas tobe compressed.

Other devices used for uranium plasmaresearch approach the required specifica-tions, but not all at the same time. Forexample, uranium arcs have high enoughtemperatures but volume and particledensity are too small to allow appreciablenuclear interactions. Arcs are quiteuseful for optical studies however.King's Furnace type plasmas have therequired volume and temperature but theparticle density is quite low. They areuseful for precision optical measurements,especially measurement of transitionprobabilities. Pulsed capacitor dis-charges have high temperature and a high

enough particle density. Their volume istoo small though. Glow discharges, highor low pressure, have the required volumeand temperature. However, a high enoughparticle density will be hard to achieve.

In the ballistic compressor, the volumeand particle density are sizeable and thetemperature, although somewhat low, can beimproved. It is conceivable that such adevice could be built large enough toachieve criticality.

In the present form, the ballisticcompressor is a useful research tool togenerate information for applicationslike the gaseous core reactor, the nuclearpiston engine (using UF6 as a workingfluid) , nuclear pumped lasers and otherdevices which may depend on the non-equilibrium properties of a uranium plasmaor UF» seeded plasma.

With its many degrees of internalfreedom the UFg could exhibit an extremelylow ratio of specific heats, Y» such thatat heating, no appreciable increase ofpressure would occur. Therefore, a firststep in this research program is themeasurement of the effective y of UF».

II. Ballistic Piston Compressor

The device used for the rapid com-pression of UF6 is shown in Figure 1.Since this device is described in detailin References 1, 2, and 3, only a briefdescription of the compressor and theprinciple of operation is given.

The compressor consists of five mainparts, i.e., the reservoir, pistonrelease section, piston, barrel and highpressure section. The reservoir, whichcontains the driver gas for the piston,has a volume of 60 liters and wasdesigned for a maximum operating pressureof 136 atmospheres. Sealing this reser-voir from the test section prior to firingof the piston is achieved using theplunger arrangement in the piston releasesection shown in Figure 2. The barrel isapproximately 4 meters long with a 50millimeter bore. Figure 3 shows the highpressure section with its three data

This work was supported by the National Aeronautics and Space Administration

FIG. 1 BALLISTIC PISTON COMPRESSOR

FIG. 2 BALLISTIC PISTON COMPRESSOR(RESERVOIR END)

FIG. 3 BALLISTIC PISTON COMPRESSOR(HIGH PRESSURE END)

acquisition ports. This section wasdesigned to accommodate pressures up to5000 atmospheres. The piston body is madeof phosphor-bronze with a molybdenum headfor high temperatures and the one usedweighs 4 kilograms. Rulon cup seals areattached to the piston to minimize gasleakage between the reservoir and thetest section. The entire device issupported with roller bearings beneaththe reservoir and high pressure testsection to allow for recoil during firing.

Operation of the compressor can best beunderstood by examining the gas handlingsystem shown in Figure 4. The reservoiris first sealed from the barrel byapplying a high pressure, relative to thereservoir pressure,on the plunger to seat-it against a teflon gasket. The reservoiris then filled to the desired drivingpressure. Next, the piston is seated inthe breech by pressurizing the barrel andtest section. A vacuum is drawn on bothsides of the piston before filling thebarrel with the test gas. After fillingthe test gas, the piston is fired bybleeding the pressure on the plunger tothe atmosphere. Thi's causes the reser-voir gas to drive the plunger back andthe reservoir gas rushes into the breechdriving the piston down the barrel.

The high chemical reactivity of UFsnecessitates additional gas handlingsystem design and construction than thatrequired for most gases. For instance,UF6 reacts readily with water and becausecommercial helium is used as the carriergas for the UFs as well as the driver gas,it must be purified by passing it throughfirst an oil scrubber then a cold trap.Sealing the diagnostic ports, the barrelto the test section, and the barrel to thereservoir is also a problem. UF6 attacksrubber and neoprene, thus viton o-ringswith the chemical inertness of teflon andthe flexibility of rubber must be used.Another problem is in the type of windowsto be used. Both quartz and glass areetched by UFg rendering them unsuitablefor spectroscopic studies. Sapphirewindows have held up quite well andalthough a film does form on them,it iseasily removed. For more details inhandling UF6 see Ref. 5 and 6.

The toxicity of the UF$ requires aunique gas disposal system. The residualgas after firing is bubbled through asodium hydroxide solution to remove uraniumcompounds and any hydrogen fluoridepresent. Bubbling the gas through anadditional solution of potassium iodideremoves any free fluorine. The oil in thepumps used for this operation is changedregularly to remove the dissolved uraniumcompounds. When enriched UFs is used, abellows type compressor and a liquidnitrogen cold trap will be used to reclaimthe gas after firings.

FIG. k BALLISTIC COMPRESSOR GAS HANDLING SYSTEM

III. Experimental Techniques

The diagnostic ports in the highPressure test section are presently beingused for pressure and volume measurementsat maximum compression. These measurementsare being made for pure He and He-UFsmixtures to determine the thermodynamicsof UFe under the rapid compression condi-tions.

Pressure Measurements

The initial filling pressure of thetest section is measured using two Wallaceand Tiernan absolute pressure gauges.In all data presented, the UF6 fillingpressure is its vapor pressure at roomtemperature which is approximately 100 .torr. To determine the maximum pressureof the test gas during compression, a highpressure quartz transducer is used in oneof the side windows of the test section.This transducer is coupled to a chargeamplifier which produces a calibratedoutput signal that can be read directly inpounds per square inch.

If the reservoir pressure and the test.gas initial and maximum pressures areknown, an effective isentropic exponentfor the test gas can be determined. Thisis accomplished by equating the work done

by the reservoir gas on the piston to thework done by the piston on the test gas.The assumptions made are: (1) ideal testgas, (2) adiabatic process, (3) no gasleakage across piston, and (4)P,constant. res

V - P V

where P and Vpressure and volume and Pmax and

'

o are initial fillingV are. in

our, case the maximum test 'gas pressureand the corresponding volume. Rewritingthis expression

(2)

- 1

and substituting the pressure-volumeisentropic relation,

max o

the resultant equation becomes

(3)

res

Pmax fPmax \~T~~ ~ \ P /

1/Y

(Y-D I/Y(4)

- 1

If Pmax/po is large and Y is reasonably

greater than 1, the second terms in thenumerator and denominator of the rightside of equation 4 can be dropped.Solving the resultant equation for Pmaxresults in

(5)

P == <Y-1>(Y/Y~1)P«iav * * '

(-1/Y-Dmax res

By rearranging and taking logarithms, thisequation becomes

log

(6)

log CY-!])

A value of Y can then be determined bygraphical means. Knowing an effectivevalue for Y gives an indication of thedegrees of freedom excited in the com-pressor test gas (in our case, the UFjmolecule). The effective value of Y canalso be used to give an indication of thetemperature at maximum compression if thecompression ratio is known. This is doneusing the isentropic pressure temperaturerelation,

max . T.(Y/Y-DT~~/O

(7)

Volume Measurements

Volume measurements are made using apin in the test section end plate permit-ting pressure and volume measurements tobe made simultaneously. The pistonpresses the pin into a cylinder and thepin length remaining outside the cylinderafter firing reveals the minimum volumereached.

IV. Results

The results of the measurements ofpressure and volume are shown in Figs. 5and 6. For a given filling pressure asubstantially higher maximum compressionpressure is reached in UFs than in He.Such a behavior is expected according toequation 5 because the expected Y of UFgis smaller than that of.He. Similarlythe volume at maximum compression of UFsis smaller than the one of He. In orderto determine the effective Y/ tne com-pression ratio (Pmax/

po) is plotted versusthe reservoir to barrel pressure ratio(Pr/P0) in Figure 7. The slope of theresulting straight lines is related to the.effective Y/ as can be deduced from eqn.6. The resulting values for Y are 1.74for He and 1.59 for He-UF6 mixtures.

* «00

izs •I5 »

»0 400 450

INITIAL FILLING PRESSURE (mm H« )

FIG. 5 PRESSURE MEASUREMENTS

f,,,' 100 nig

uFt-miw» MIXTURE

100 UO 400

INITIAL FILLING PRESSURE (•«. H«)

FIG. 6 VOLUME MEASUREMENTS

The smaller effective Y of the He-UF6mixture indicates that a large part ofthe input energy is transferred to therotational and vibrational degrees offreedom of the UF6 . The effective Yvalue for pure He is larger than 5/3,due to leakage past the piston duringcompression.

The temperature of the test gas can bedetermined using the isentropic pressure-temperature relation (eqn. 7) and theresults are shown in Fig. 8. The tempera-ture values for He appear in this graphhigher than those of an adiabaticallycompressed monatomic gas, because of thetoo large Y for He as deduced from experi-ment, due to leakage around the piston.

V. Summary

The experimental results show that somebut probably not all of the rotationaland vibrational levels of UF6 areexcited under rapid compression. Thisindicates that by compression of He-UFsmixtures,temperatures may be achieved,higher than expected for equilibriumconditions.

Limited spectral studies of UFs wereperformed while obtaining the pressure-volume data. The light output

<o:

env>uK0.

OU

«t\

, IDEAL ISENTROPES. Htlium MEASUREMENTS, REF. 4

Htlium MEASUREMENTS, U. of FLORIDA.H«-UF6 MEASUREMENTS , U. of FLORIDA

10 20 SO 40 90RES. TO BARREL PRES. RATIO (P*/P0)

FIG. 7 EFFECTIVE y DETERMINATION

MONOATOIM Ola .COUPREMCO AOUBATIC4LLY FROM IM*KHCUUU UCiSURtMtNTSN< - Uf» Mt>SURtM(NT9

I"

has been weak due to the minimal compressorconditions used and there has also been aproblem with the piston blocking the sidewindows during compression. These prob-lems are overcome by increasing the testgas filling pressure and the reservoirgas pressure. The amount of UF6 in theHe-UFs mixture may have to be changedalso.

After determining the compressor andtest gas conditions required to obtainmeaningful UF6 spectroscopic data andafter obtaining such data, the testsection will be inserted in the thermalcolumn of the University of FloridaTraining Reactor. The compressor will befired with 93% enriched UF6 with helium asthe carrier gas and a comparison will bemade of the fissioning UF6 spectra withthe non-fissioning spectra.

VII. Acknowledgements

The authors would like to thankDr. Richard D. Dresdner for his valuableassistance in the safe handling ofuranium hexafluoride.

References

1. Lalos, G. T., NOLTR 63-96, 1963.

2. Lalos, G. T. and Hammond, G. L.,NOLTR 66-202, 1967.

3. Lalos, G. T. and Hammond, G. L.,NOLTR 70-15, 1969.

4. Hammond, G. L. and Lalos, G. T. ,NOLTR report in preparation, 1971.

5. DeWitt, R., Uranium Hexafluoride: ASurvey of the Physico-chemicalProperties. Goodyear Atomic Corpora-tion:fGAT-280). August 12, 1960.

6. Hudlicky, M., Organic FluorineChemistry, Plenum Press, New York,1971.

100 MO 40O MO

COMMISSION RATIO

FIG. 8 TEMPERATURE MEASUREMENTS

V. CO, LASER EXPERIMENTS USING NUCLEAR REACTIONS AS THE IONIZATION SOURCE*

H. S. Rhoads and R. T. SchneiderUniversity of FloridaGainesville,Florida

F. AllarioNASA - Langley

Hampton, Virginia

Abstract

Experimental studies show that the out-put of a COj laser is significantlyincreased by products of the nuclearreaction HeMn,p)T.

Helium-3 was used in lieu of thenatural helium normally present in the1:1:8 COz:N2:He laser gas mixture (pres-sure = 6 torr). The laser assembly wasthen exposed to a reactor thermal neutronflux of about 108 neutrons cm"2 sec"1.Power output of the laser doubled whilethe electrical power input decreased;electrical efficiency was thus more thandoubled. Results indicate that additionalionization by the energetic chargedparticles may be responsible for theimproved laser performance.

Introduction

Shortly after the first demonstrationof laser action, papers speculating on thepossibility of coupling the laser and thenuclear reactor began to appear. Theresearch stimulated by such speculationhas begun to show great promise; thispaper reports on some aspects of thatresearch.

Motivation

There are many theoretical advantagesin pumping lasers directly with energeticnuclear reaction products. The powerdensities attainable with nuclear reac-tions are very large compared to those ofchemical or electrical processes; anadvanced nuclear-powered laser could bequite compact but still immenselypowerful.

Such a system could also yield highoverall efficiencies. Nuclear-electricpower plants might one day operate at 40%efficiency in converting nuclear powerto electricity, and the advanced lasersmight convert that electrical power tolaser light with 25% efficiency (notcounting the power requirements of vacuumpumps, compressors and cooling systems).A nuclear laser which converts 10% ofnuclear reaction energy to coherent light.would thus be quite an attractive prospect.

While the goal of direct nuclear pump-ing of high-power lasers might be far in

the future, nuclear enhancement of elec-trically-pumped-laser performance asreported in this paper and in otherrecent publications^1'2'3) are ofimmediate interest to the industry. Alaser whose efficiency is bolstered bynuclear reaction products — say, alphaparticles from a radioactive coating onthe walls of a C02 laser tube — wouldrequire & smaller power supply and a lessrobust gas-handling system for a givenbeam power.

The advantages of compactness andreduced weight would be quite valuable inany airborne or space laser system.

History

Nuclear induced laser action has notbeen positively reported at this time,though several teams are pursuing thatgoali"). Experiments with solid lasermedia have not been encouraging, probablydue to the degradation of optical qualityin crystals and glasses exposed to nuclearradiation. Little nuclear work has beendone with liquid laser media.

Gases are not so subject to radiationdamage, so that most work has concentratedon gas lasers.

Andriakhin, et al., at the MoscowState University have reported a threefoldincrease in the power of a COj' laserexposed to a beam of protons from anaccelerator ''. The proton beam was usedto simulate the products of a nuclearreaction such as 3

KeV.He 760

This reaction was used in a later experi-ment in which a Hg-3He gas mixture in alaser tube was exposed to neutrons from apulsed reactor. The results were ambigu-ous, but 10 mW peak light pulses wereproduced'5'.

A University of Illinois group directedby G. H. Miley has been active in thisfieldl2'6~"). Most notable of theirrecent results is a 20% enhancement inpower and efficiency of a pulsedC02-rNa-He laser, reported by Ganley, etal. (2f The laser tube was lined withboron-10, which reacts with neutrons togive energetic lithium nuclei and alphaparticles. Neutrons were supplied in

*This work was supported by NASA Grant NGL-10-005-089

bursts of 5 x 10''neutrons/cm2sec peakflux from a, pulsed TRIGA reactor.

Experiments with a continuous COZlaser have previously been reported byAllario and Schneider*3'. Using aprocedure similar to that outlined laterin this paper, but with a lower neutronflux, an increase in laser power of a fewpercent was observed.

Other experimental results are summa-rized in the survey paper of Thorn andSchneider ("*).

Experiment

The basic idea of this experiment wasto replace the natural helium normallypresent in the CO2-N2-He electrically-pumped laser with helium-3. When thelaser was exposed to thermal neutrons inthe University of Florida Training Reactorshield tank, the reaction 3He + 'n ->•*H + 3H + 760 KeV altered laserperformance.

The laser was a conventional water-cooled internal-mirror design, with acavity length of 1 m and active dischargelength of 66 cm. Tube diameter was 2.5cm. Mirrors used were a 99% reflectiveflat and an 85% reflective germaniumoutput coupler with a 10 m radius ofcurvature. Maximum laser output outsidethe reactor was about 15 watts with flow-ing gas, with excitation provided by aDC power supply at (typically) 5 KV and 50ma. Best performance without gas flow wasabout 1 watt at 3.8 KV and 30 ma.

The laser assembly and a thermopile formonitoring beam power were enclosed in awatertight aluminum canister and loweredinto the UFTR shield tank (a 14-ft. deeptank of deionized water) to a locationnear the reactor core. Gas, coolingwater and electrical power were providedby hoses and cables from outside theshield tank. Thermopile output wasmonitored with a Keithley millivoltmeter.

Most experimental difficulties arosedue to the underwater location of theapparatus. Water leaks were annoyingduring initial checkouts, and the longvacuum hoses into the tank slowed pumpingtime. The beam could not be observeddirectly so that wavelength and resonantmodes could not be determined. Mirroralignment could not be optimized after thelaser container was sealed; care wasrequired while lowering the assembly intothe reactor to avoid shocks that coulddisturb alignment. Fortunately, radiationproblems were not as severe as expected;the optics were apparently not damaged bythe neutron, gamma and charged-particlebombardment over several hours of opera-tion. Neutron activation problems werereduced by careful selection of materials,so that overnight "cooling" was usuallyenough to allow gamma radiation from the

apparatus to decay.

The laser could not be operated in theflowing gas mode with helium-3 due to thevery high cost ($300 for 2 liters at STP,or 'vlO6 dollars/pound) of the gas. Aclosed loop system in which the helium-3could be recycled was considered, but thiswould permit the buildup of carbon monoxidefrom COj dissociation and defeat the pur-pose of the gas flow.

The procedure followed began with evacu-ating the system and premixing the gases(1:1:8 COi:Nj:He) for several minutes in a2-liter flask to insure uniformity of gasmixtures. With the reactor at the desiredpower level, the gas mixture was releasedto the laser system and the DC glow dis-charge ignited. Current and voltage weremonitored on meters and recorders at thepower supply panel.

Care was taken to isolate effects due tothe helium-3 reaction. Runs were made withnatural helium and with helium-3 at zeroreactor power; negligible differences inlaser output were observed. Some enhance-ment (about 20%) was seen with naturalhelium at full reactor power (60 KW,neutron flux $ = 108.n/cm*sec), possiblydue to the high gamma field near thereactor core (Table 1). The thermopile wasnot affected by nuclear radiation enoughto introduce error. Slight deviations fromthe desired gas partial pressures and totalpressures were found to cause only slightchanges (decreases) in laser output.

Results

The effects of nuclear reactions onlaser output and efficiency are summarizedin Figure 1 and Table 1. In the absenceof nuclear reactions, laser output andefficiency improved as discharge currentwas decreased to the minimum at which adischarge could be maintained (32 ma).With the reactor at full power (4> = 109

n/cm2sec), the current could be reduced

MINIMUM DISCHARGE CURRENTt OPTIMUM PERFORMANCE IWITHOUT NUCLEAR REACTIONS

LASER EFFICIENCYPERCENT

LASER POWER. WATTS

DISCHARGE CURRENT HA

FIGURE 1; EFFECT OF NUCLEAR REACTIONSON LASER PERFORMANCE.1:1:8 CO2:N : -toe, 6 torr

C02:N2lHe

«He8

6 torr

3He

€ torr

i:r.a3He+4He6 torr

i:i:e

14 torr

REACTOR POWER

ZERO

1.0 W32 ma0.86% efficiency

1.05 W33 ma0.85 %

1.0 W33 ma0.85 %

UNSTABLEDISCHARGE(OVERCURRENT)

10 KW

1.28 W27.5 ma1.18 %

60 KW ($»108)

1.2 W32 ma0.97 %

1.95 W25 mo1.97 %

1.5 W29 ma1.4 %

0.7 W46 ma , 5.2 KV0.3 %

TABLE i : E f f e c t s of Nuclear Reac t i ons

still further to about 25 ma withaccompanying increase in output to almosttwice the original power. Dischargevoltage remained approximately constantas the current decreased, so that elec-trical power input decreased and electricalefficiency went up by a factor of 2.2.

A few runs were made at lower reactorpower (10 KW, $ = 1.7 x 107 n/cm2sec). Inanother run at 60 KW, 50% helium-3 and 50%natural helium were used for the heliumfraction, giving half the nuclear reac-tions available with undiluted helium-3.As shown in the table, the enhancementeffect was reduced in these runs.

All the above data were taken at atotal pressure of 6 torr, which gave adischarge voltage of 3.8 KV. Since one ofthe great hopes for nuclear lasing is thathigher gas pressures might be used, a runat 14 torr was made, without nuclearenhancement, it was difficult to get thelaser discharge to operate at pressuresas high as 10 torr. The high currentpulse on ignition of the discharge almostalways tripped the "overcurrent" protec-tion circuit on the power supply (noload resistor was placed in series withthe laser). As the table shows, it waspossible with nuclear ionization toachieve laser action in a reasonablywell-behaved glow discharge at a pressure40% above the normal operating limit.Higher pressures than 14 torr were not

tried in these experiments, but are plannedfor a later laser in a higher-flux reactorlocation.

Some enhancement was observed in 4:1 and9:1 3He-COa mixtures, but the output powerwas quite low (<0.3 W). The power in-creased perhaps 80% to 100% at fullreactor power, but data at such low laserpower is not felt to be completelyreliable.

Midway through the experimental programdescribed above, a Soviet reference to alaser based on boron trichloride11

prompted interest in this gas as a possi-ble nuclear laser medium. The 10B(n,a)7Lireaction has been used in other nuclearlaser-work*2', but never with the boronin gaseous form. Natural boron contains alarge fraction of l°B.

Some experiments with BClj were attemptedbut no useful information was obtained.BClj is highly corrosive and the dischargewas run for only a short time before itbecame evident that a new laser designedto handle such corrosive gases as BClj,BFs and UF6 should be built. Such alaser is now being checked out.

Discussion

An estimate of the power PN beingreleased in nuclear reactions within thelaser shows that PN is quite trivial

compared to the 100-200 watt electricalpump power:

PN = (NHe o <>) .Q V < 1'mW

Hea

QV

= number density of *He

= neutron cross section for 3He(n,p)T, 5300 barns

=. neutron flux, = 108 n/cm2sec= energy per reaction, 760 KeV= laser volume, = 5300 cm3

However, only a small amount of poweris required for the ionization that main-tains the discharge O. The DC electricfield ionizes the gas by providing anumber of energetic (>3 eV) electrons,but these energetic electrons are notneeded for establishing the populationinversion. The N2 and COj vibrationallevels are excited by electron impactat a fraction of an eV, so laser outputincreases with the density of low-energyelectrons. Nuclear reaction products mayallow an increase in the number of low-energy electrons by directly producingmore electrons in ionizing collision andby relieving the DC power supply of theionization task so that the electricfield strength E may be reduced as fewerhigh-energy electrons are required.

Perhaps the most important result ofnuclear reaction ionization in this laseris the reduction in gas temperaturebrought about by the lower currents atwhich the laser would operate. Gastemperature is perhaps the single para-meter most important to CO* laserefficiency <9 >.

Nuclear enhancement apparently has aneffect similar to the addition of Xe orCs or other low-ionization-potentialatoms to the laser gas: it lowers theeffective ionization potential of thelaser medium, allowing the electricfield (and thus the average electron .energy) to be reduced'9! At higherneutron fluxes, or with the addition ofradioactive materials to the walls orgas in the laser, it may be possible tocompletely relieve the external powersupply of the maintenance of the electricdischarge and use it solely to acceleratenuclear-generated electrons to the lowenergies required for molecular vibra-tional excitation.

Speculation

The properties of such a possible lasersystem are rather interesting. Theproduction of CO by electron impactionization would cease since only veryfew electrons would have the 2.3 eVrequired to dissociate CO2. The nuclearreaction products would cause some dis-sociation, but the net dissociation ratemight decrease enough to reduce or elimi-nate the need for gas flow to remove CO.

Catalytic recombination of CO and O2 backto CO2 might then be sufficient to giveflowing-gas performance in a sealedsystem.

Another reason for gas flow is the con-vective removal of heat, particularly ingas dynamic lasers. If the efficiency ofa laser can be raised by nuclear effectsto the point that more of the electricalpower is converted to laser light, lesswaste energy will need to be removed asheat.

Perhaps most intriguing of all is thefact that high-pressure systems, in whichmore of the nuclear reaction energy couldbe absorbed by the gas, probably wouldshow the greatest effects from nuclearenhancement.

The use of a complex, expensive nuclearreactor as a radiation source for nuclearparticles is obviously impractical outsidethe laboratory (unless a way is found toconvert fission energy directly to laserlight) , but it provides a variable sourceconvenient for experiments. Fast ionfluxes equivalent to and greater thanthose in this experiment are readilyavailable from isotope sources such aspolonium-210 once the important processesare understood.

It would be unreasonable to expectoptimum performance in nuclear-pumpedlasers at the same gas mixtures andpressures found to be best for electric-discharge systems. In particular,Bullis, et al. , 9 ' report that fractionalelectron power transfer directly to theupper C02 laser level increases withdecreasing electron energy, whilevibrational excitation of N2 falls off.Nitrogen is important in dischargelasers where average electron energiesare high, but it might be unnecessary ordetrimental in a nuclear-ionized laser.

Work in Progress

As previously mentioned, new lasershave been constructed to investigate awide range of nuclear laser media. Theirmodular design allows easy interchange-ability of components (laser tube,mirrors, Brewster windows and electrodes),giving a significant advantage in flexi-bility over the laser used in the experi-ments reported here. The lasers are some-what more compact to allow investigationsin the higher neutron flux of the UFTRthermal column.

In addition to gases such as BClj,and UFe and solid coatings of uranium andboron, the lasers will allow investiga-tions with liquid media. In addition toliquid scintillator solutions, there area number of uranium-bearing solutions(e.g. of uranyl nitrate, sulfate,fluoride, phosphate and acetate) (I2)which fluoresce to some extent and might

exhibit laser action. This is an 10.interesting prospect, since criticalreactors of uranium-bearing solutions arewell within the grasp of present nuclearengineering technology; this cannot as 11.yet be said for plasma-core reactors.

Summary

Experimental evidence indicates that 12.further nuclear-pumped laser experimentsare well worth the experimental difficul-ties involved. As more data arecollected and the physical processesinvolved become better understood, it isnot unreasonable to expect that the goalof direct conversion of nuclear power tocoherent light may be realized in the nottop distant future.

References

1. Andriakhin, V. M., et__al. , "Increaseof COj Laser Power Under the Influenceof a Beam of Fast Protons," JETPLetters, IB, 214-216, 1968.

2. Ganley, T., et_al., "Enhancement ofCO2 Laser Power and Efficiency byNeutron Irradiation," Applied PhysicsLetters, June 15, 1971.

3. Allario, F., and Schneider, R. T.,"Enhancement of Laser Output byNuclear Reactions,'" NASA SP 236.

4. Thorn, K. and Schneider, R. T.,"Nuclear Pumped Gas Lasers," AIAA 9thAerospace Sciences Meeting, January,1971; AIAA Paper No. 71-110.

5. Andriakhin, V. M., et al., "Radiationof Hg-He3 Gas Bombarded by a NeutronStream," ZhETF Pis. Red. 12, No. 2,83-85, 20 July 1970.

6. Guyot, J. C., etal., "ExperimentsUsing Nuclear Radiation to Pump aHe-Ne Laser," Proceedings of FifthAnnual Review of Electronics,University of Illinois, 1967.

7. Miley, G. H. and Verdeyen, J. T.,"Advanced Methods for NuclearReactor-Gas Laser Coupling," AECContract AT(11-1)-2007.

8. Guyot, J. C., et al., "On Gas LaserPumping via Nuclear Radiations,"Symposium on Research on UraniumPlasmas and Their TechnologicalApplications, Gainesville, Florida,January 1970, NASA SP-236 (Supt. ofDocuments, U. S. Gov't Printing Office$3.75).

9. Bullis, R. H. et al., "Physics of COZElectric Discharge Lasers," Proceed-ings of AIAA 9th Aerospace SciencesMeeting, January, 1971; AIAA paperNo. 71-64.

Wiegand, w. J., et al., "CarbonMonoxide Formation in CO2 Lasers",Appl.Phys.Lett. 16_:6,237-9, 15 Mar 70.

Karlov, N. V., et al., "Laser Basedon Boron Trichloride," JETP Vol. 8 fl(ZhETF Pis. Red. 8, 22-25, 5 July1968).

Burris, L., "Chemistry and ChemicalEngineering" in Etherington (editor),Nuclear Engineering Handbook (NewYork, 1958; McGraw-Hill), section#11-37.

VI. TEMPERATURE PROFILE DETERMINATION IN AN ABSORBING PLASMA*

J. L. Usher and H. D. Campbell

A method is available for temperature profile deter-

mination in the optically-non-thin case which requires

only knowledge of the line absorption coefficient of the

plasma. It is known as the brightness-emissivity

(12) • •method ' ' (BEM) and is essentially a line-reversal

technique. The two stringent prerequisites for this

technique are (a) partial local thermodynamic equili-

brium (LTE) down to the lower level of the transition

in question, and (b) homogeneity of the plasma. As this

diagnostic technique is independent of the atomic proper-

ties of the plasma constituents, it can be very useful in

plasmas composed of constituents having large atomic

numbers. This is so because there is frequently a serious

shortage of available and accurate data on the atomic

properties of elements with high atomic numbers, e.g.,

uranium.p

The spectral intensity, I , emitted by a homogeneousA

(4)layer of plasma may be expressed as

(1)

where T is the optical depth of the plasma layer andA

S is the source function. The requirement of partial

LTE allows one to replace the source function by the

Planck function, B (T ) , where T is the electronA 6 • S

temperature of the plasma layer.

If one considers a non-homogeneous plasma to be

comprised of concentric rings of homogeneous plasma

(Fig. 1) , prerequisite (b) above may effectively be

removed in the following manner. By placing a

standard light source of known intensity behind the

plasma, one may calculate the optical depth of the

j line-of-sight position for the plasma from the

following expression for the total intensity:

cwhere I. is the intensity of the background source as

Ap

determined previously, and I (X) is the plasma intensity

(at the j position) , which also has to be independently

determined. The optical depth along the j line-of-sight

position, T . (X) , can be determined directly from Equation

(2) , and it may also be expressed as

One makes measurements of T . at several line-of-sightD

positions (Fig. 1) and then computes the average absorption

coefficient for the k ring, <k(X), from the following

expression:

(4)

Equation (4) is obtained directly from Equations (2) and

(3). The I term in Equations (3) and (4) represents

the chordal length segment present in the k concentric

ring along the j line-of-sight position. It has also

been assumed that the intensity of the background source

may be represented by B?(T ), where T is the brightnessA B 5

temperature.

This representation of the background source intensity

is chosen to simplify the mathematical calculations; it is

not necessary. The choice of sufficiently bright sources

may pose problems in the case of optically thick plasmas.

One now proceeds to calculate the average tempera-

ture, T , of the k concentric ring. A combination ofJC

Equations (1) and (4) produces the following value for the

p — thPlanck function, B (X,T, ), of the k ring:

_ _w ,j -

(5)

The average temperature for the ring is now determined

from the inversion of the above Planck function:

=?. _It- (6)

Once one has obtained values for K and T profiles in theA

plasma, it is also possible to calculate values for the

":}>'•" ''"' emission coefficient, e . One makes use of the LTEA

.approximation (Kirchoff's law):

Previous work ' ' in determining the temperature

profiles of optically-non-thin plasmas has made use of

Equation (7) to determine T(r) and thus requires the

determination of both the K and e. profiles of the plasma;A A

however, the present method requires only the K. profile.

It should be noted here that the same experimental data is

necessary for both methods. The use of the double-path

(7)method by Birkeland and Braun represents an excellent

solution to the problem of matching intensities of the

plasma and background source. A sensitivity analysis has

verified that the errors in the determination of these

profiles propagate in a manner similar to those in Abel

inversion techniques. Hence more accurate temperature

profiles can be determined with the BEM unfolding tech-

nique, which only requires the K profile.

REFERENCES

1. E. M. REEVES and W. H. PARKINSON, "TemperatureMeasurements for Shock Heated Powdered Solids",Scientific Report No. 1, Harvard CollegeObservatory.

2. W. LOCHTE-HOLTGREVEN, "Plasma Diagnostics",North-Holland, Amsterdam, John Wiley and Sons,Inc., New York (1968) .

3. H. R. GRIEM, "Plasma Spectroscopy", McGraw-Hill,New York (1964).

4. S. CHANDRASEKHAR, "Radiative Transfer", DoverPublications, Inc., New York (1960).

5. W. G. BRAUN, Rev. Sci. Instr., 36, No. 6, 802(1965).

6. P. ELDER, T. JERRICK and J. W. BIRKELAND, Appl.Optics, 4_, 589 (1965) .

7. J. W. BIRKELAND and W. G. BRAUN, "The TemperatureDistribution in a Vortex-Cooled Hydrogen Arc",Paper Presented at the Seventh InternationalConference on Phenomena in Ionized Gases, Belgrad(1965).

Fig. I Ring Division of a CylindricallySymmetric Plasma

Nomenclature

B - Planck functionA

B (T ) - Planck function representation of backgroundsource intensity, T is brightness temperature

BP —B (A,T ) - Planck function representation of intensity of

k concentric ring of plasma, T is averagetemperature of kth ring

BEM - Brightness -emissivity method

Ci . - First radiation constant

Cj - Second radiation constant

e. - Specific emission coefficientpI - Intensity of plasma layerA

P thI.(X) - Intensity of plasma along j line-of-sight3 positiongI. - Intensity of background sourceA

T ' thI.(X) - Total intensity observed along j line-of-sight3 position

K - Line absorption coefficientA

K. (X) - Average absorption coefficient in k concentricring of plasma

&.. - Chordal length segment present in k concentric•* ring along jth line-of-sight position

LTE - Local thermodynamic equilibrium

X - Wavelength

S - Source function*

T - Brightness temperatureB

T - Electron temperature

T, - Average temperature in k concentric ringI\.

T(r) - Temperature profile

r - Radius

T - Optical depthA

T.(X) - Optical depth along j line-of-sight position

VII. PUBLICATIONS GENERATED UNDER GRANT NGL 10-005-089

The following papers and journal publications were given

or are accepted for publication.

1. "Spectroscopic Diagnostics of a Uranium Plasma," by R. T.

Schneider, Seventeenth Symposium on Spectroscopy, January

8-10, 1969, Gainesville, Florida.

2. "Radiation Transfer Study on a Uranium Plasma," by R. T.

Schneider, Fifth Annual Southeastern Seminar on Thermal

Sciences, April 21-22, 1969, Gainesville, Florida.

3. "Spectroscopic Measurements on a High Pressure Uranium

Arc," by R. T.. Schneider, G. R. Shipman and A. G. Randol,

III, 20th Annual Mid-American Symposium on Spectroscopy,

May 12-15, 1969, Chicago, Illinois.

4. "Temperature Measurement on a Uranium Plasma," by R. T.

Schneider, G. R. Shipman and J. M. Mack, 8th National

Meeting, Society for Applied Spectroscopy, October 6-10,

1969, Anaheim, California, and Applied Spectroscopy, 23,

671 (1969).

5. "Radiation from a Uranium Plasma," by R. T. Schneider,

C. D. Kylstra, A. G. Randol III and M. J. Ohanian,

American Nuclear Society 15th Annual Meeting, June 15-19,

1969, Seattle, Washington, and ANS Transactions, 12, 3 (1969)

6. "Measurement of the Emission Coefficient of a Uranium

Plasma," by R. T. Schneider, A. G. Randol III, C. D. Kylstra

and M. J. Ohanian, American Nuclear Society Winter Meeting,

November 30 - December 5, 1969, San Francisco, California,

and ANS Transactions 12, 413 (1969).

7. "Experimental Determination of the Boiling Point of

Uranium," by A. G. Randol III, R. T. Schneider, and C. D.

Kylstra, American Nuclear Society Winter Meeting,

November 30 - December 5, 1969, San Francisco, California,

and ANS Transactions 12, 541 (1969).

8. "Properties of a Uranium Plasma," by H. D. Campbell, R. T.

Schneider and C. D. Kylstra, 8th Aerospace Science Meeting,

AIAA, January 19-27, 1970, New York.

9. "Measurement of the Temperature and Partial Pressure of a

Uranium Plasma," by R. T. Schneider, G. R. Shipman and

A. G. Randol III, Applied Spectroscopy 24, 253-258 (1970).

10. ' "Experimental Investigations of a Uranium Plasma Pertinent

to a Self Sustaining Plasma Source," Annual Report,

September 1969.

11. "Boiling Point of Uranium," by A. G. Randol III, R. T.

Schneider and C. D. Kylstra, Symposium on Research on

Uranium Plasmas and Their Technological Application,

January 7-10, 1970, Gainesville, Florida.

12. "Spectroscopic Study of a Uranium Arc Plasma," by H. D.

Campbell, R. T. Schneider, C. D. Kylstra and A. G. Randol

III, Symposium on Research on Uranium Plasmas and Their

Technological Application, January 7-10, 1970, Gainesville,

Florida.

13. "Generation of a Fissioning Plasma," by C. D. Kylstra and

R. T. Schneider, Symposium on Research on Uranium Plasmas

and Their Technological Application, January 7-10, 1970,

Gainesville, Florida.

14. "Uranium Plasma Research at the University of Florida/" by

R. T. Schneider, C. D. Kylstra and M. J. Ohanian, Sixth

Intercenter and Contractors Conference on Plasma Physics ,

December 8-10, 1969, Langley.Research Center.

15. "Review of Uranium Plasma Research," by R. T. Schneider,

Meeting of the NASA Research and Technology Advisory

Subcommittee on Electrophysics, NASA Lewis Research Center,

Cleveland, Ohio, April 13-14, 1970.

16. "Uranium Plasma Emission Coefficients," by C. D. Kylstra,

R. T. Schneider and H. D. Campbell, AIAA 6th Propulsion

Joint Specialist Conference, San Diego, California,

June 15-19, 1970.

17. "Experimental Investigations of a Uranium Plasma Pertinent

to a Self Sustaining Plasma Source," Annual Report, May 1970.

18. "Plasma Diagnostics of a Uranium Plasma," by R. T. Schneider,

C. D. Kylstra and H. D. Campbell, International Conference on

Gas Discharges, London, U.K., September 15-18, 1970.

19. "Generation of a'Uranium Plasma at Near Gaseous Core Reactor

Conditions," by J. F. Davis III, B. G. Schnitzler and R. T.

Schneider, 2nd Symposium on Uranium Plasmas r Research and

Application, Atlanta, Georgia, November 15-17, 1971.

20. "Ballistic Piston Fissioning Plasma Experiment," by B. E.

Miller, K. Thorn, R. T. Schneider and G. T. Lalos, 2nd Symposium

on Uranium Plasmas: Research and Application, Atlanta, Georgia,

November 15-17, 1971.

21. "Excitation and lonization of a CO- Laser by Nuclear Reaction

Products," by H. S. Rhoads and R. T. Schneider, Bull, of the

Amer. Phys. Soc-. (in print) ..

22. "Plasma Properties of a D.C. Uranium Arc," by J. M. Mack, Jr.,

J. L. Usher, R. T. Schneider and H. D. Campbell, Bull, of the

Amer. Phys. Soc. (in print).

23. "Uranium Plasma Emission Coefficient in the Visible and

Near U.V.," J. M. Mack, Jr., J. L. Usher, H. D. Campbell and

R. T. Schneider, 2nd Symposium on Uranium Plasmas; Research

and Application, Atlanta, Georgia,. November 15-17, 1971.

24. "c°2 Laser Experiments Using Nuclear Reactions as the

lohization Source," by H. S. Rhoads, R. T. Schneider and

F. Allario, 2nd Symposium on Uranium Plasmas: Research and

Application, Atlanta, Georgia, November 15-17, 1971.

25. "Nuclear Pumped Gas Lasers," by K. Thorn and R. T. Schneider,

AIAA Journal (in print).

26. "Research on Uranium Plasmas," by K. Thorn and R. T. Schneider,

NASA SP-236 (1971).

27. "Temperature Profile Determination in an Absorbing Plasma,"

by J. L. Usher and H. D. Campbell, submitted to JQSRT.

The following dissertations and theses were generated under

the program.

1. "A Determination of High Pressure, High Temperature Uranium

Plasma Properties," by Arthur G. Randol III, Dissertation,

University of Florida, 1969.

2. "Temperature Profile Determination of a Uranium Plasma in

a Helium Atmosphere," by J. M. Mack, Jr., Master's Thesis,

University of Florida, June 1969.

3. "Temperature Profile Determination in an Absorbing Plasma", by

John L. Usher, Master's Thesis, University of Florida, August

1971.

4. "Thermodynamic Study of Rapid Compression of UF6"/ bY Barry E.

Miller, Master's Thesis, University of Florida, November 1971.

5. "Generation of Uranium Plasma with a Sliding Spark," by

John F. Davis, Master's Thesis, University of Florida,

March 1972 (tentative)

6. "pirect Nuclear Excitation of a CO- Laser," by Harold S.

Rhoads, Ph.D. Dissertation, University of Florida, January

1972.