UNCLASSIFIED

AD NUMBER

ADB182385

NEW LIMITATION CHANGE

TOApproved for public release, distributionunlimited

FROMDistribution authorized to U.S. Gov't.agencies and their contractors;Administrative/Operational Use; 12 AUG1960. Other requests shall be referred toNational Aeronautics and SpaceAdministration, Washington, DC.

AUTHORITY

NASA TR Server Website

THIS PAGE IS UNCLASSIFIED

AD-B182 385 .....

5EOHRE C

ELECTE N~5LE LBM

FEB2219 TIBiAy cov6 A AUG6 29 196o

AVSust 12, 190

LI~~~~~y ANLSIWTUuND31C 0=01

An pwainL oWan. mnM rca ASim a? huJaM01 gmw

By compS C. amober

94- 05477 DI SR IL

NATIOAL AERONAUTICS AND SPACE ADINISTRATIBN

*~~2 18Omg=6o'

Augu~t 12, 1900 NNI.-.-rPs4O

All AIMIYSU OF TKK TWXWNDTNAIEC CYCLS AND

POSI VOKIIG PLUMS FOR A

SPACE NZA! RIJECTION SYSTEM

by

George Cs Biicher

Accesion Fof

NTIS CR'4DTIC TAR3

Justificationi

Dish ibatiot. I

Avail an~d i orDist I Special

'ODTIC USERS OtILV

TBCHNICAL AND SCIENTIVIC AESMTANTRESEARCH PROJ3CTS DIM ON

AW_ _ ___E

ABSTRACT

Within -the next few years, nuclear reactor - thermalsystems will most logically supply the large power requirementsfor space vehicles utilizing electric propulsion units. An In-tegral part of the system Is the radiator for the dissipationof waste heat.

The ideal Carnot thermodynamic cycle is analyzed to de-termine the relationship between radiator crharacteristics andthe over-all system. For minimum radiator area, the heatrejection temperature Is. shown to be three-fourths of theheat addition temperature. An expression is derived forminimum radiator area as a function of power output of thecycle and radiator temperature and the Carnot cycle efficiencyis shown to be 25 percent for minimum radiator area. The an-alysis Is then extended to the actual Rankine thermodynamiccycle, taking Into account properties of working fluids andthe inefficiencies of system: components. An expression is de-rived to show the relationship between the radiator tempera-ture and heat addition temperature for the case of optimumcycle work and minimum radiator area. Another equation ex-presses the radiator area as a function of power output andthe heat addition and heat rejection temperatures of theactual Rankine cycle. A figure illustrates that increased heataddition temperatures result in substantial reductions in theradiator size.

Suitable working fluids for a nuclear reactor - thermalsystem are discussed. Elements are more attractive thanalloys or compounds. Liquid metals have desirable properties;rubidium, potassium and sodium are found to be most suitablefor the system described.

l

.. I

-*, -II- I I I V

TAILz OF COITINTS

Page

ABSTRACT 0 0 . . . . . . . .*e** * i

LIST OP SYMBOLS, . , . • 0 , , , , . . • , . . . . iv

1 INTiRODUCTION... . • o o o • • . . . . . I

2. CYCLE ANALYSIS TO DBTZitNII RADIATORCRARACTBRISTICS . . . . . .. .. .. . . 1a. Ideal Carnot Cycle . . . . . . . . . . . . . 3b. Actual lankine Cycle o.. ......... 6

3o VORKING PLUIDS. . , . . . , , . .. . . . . . • . 15

RIF IN Ci . . , . . . . . , . . . . . . . . . . . 24

il

ILIl

J-1

LIST OF SYMBOLS

Qi heat radiated

QA heat added

a Stefan-Boltzmann constant

9 eaissivity

A1 area of radiator

S entropy

h enthalpy

Ti temperature of radiator

TA heat addition temperature

Wc work of Carnot cycle

Wr work of actual Rankine cycle

k work factor

Inc efficiency of Carnot cycle

't efficiency of turbine

' ir efficiency of actual Rankine cycle

Jv

______ n______ I_____________________

1. INTRODUCTION

Space vehicles utilising electric propulsion units will requirelarge amounts of power. Within the next few years the nuclearreactor- thermal system appears to be the most attractivesystem to mot these requirements.

Mackay (1] has compared thermal poverplant cycles forspace vahlcles and has concluded that vapor cycles are farsuperior to other thermal cycles, Including the Brayton gascycle. He found that the efficiencies of the vapor cycles arenot greatly different from those obtainable with the Ideal Car-not cycles and therefore are probably the beat that can beobtained from a thermal cycle.

In this paper, consideration will be given to a thermodynamicvapor cycle, as may be envisioned for use in a space vehicle.The power, propulsion and heat rejection system is shown sche-matically In PIgure 1. The heat source is a nuclear reactorwherein the liquid working fluid is transformed to a vapor whichpasses through a turbine and into a radiator where it Is con-denod back to a liquid, accompanied by a rejection of wasteheat. The direct condenser- radiator may not necessarily bethe best scheme for condensing vapor in an actual space ve-hicle, but It Is considered feasible.

The purpose of this paper Is to present an analysis of thethermodynamic cycle to determine the relationship between theradiator characteristics and the over-all system, and to dis-cuss possible working fluids for a nuclear reactor - thermalsystem.

2. CYCLE ANALYSIS TO DITIRIIN RADIATOR CHARACTZRISTICS

The system shown in FIgureo 1 utillsos an actual lankinecycle for the working fluid. The Rankine cycle closely approachesthe Carnot cycle In principle. The Carnot cycle is consideredthe ideal thermodynamic cycle. By comparing the performance ofthe actual ansinse cycle to the Carnot cycle, one ay see hownear the actual cycle is to perfection.

Therefore, an analysis will first be made of the Carnotcycle to determine the characteristics of the radiator. Theanalysis will then be eztemded to the actual Rankine cycle, todetor n the radiator charactorlsties a Influenced by an ac-tuel (rather than an Ideal) working fluid &ad by inefficienciesof system components.

I -

zz

IL-

zza 0 I2

44 0 4

00

la.

z2

a. Idl Carot cycle

The Carnot cycle Is shown in Figure 2. It producesthe maximum amont of work obtainable from a heat engine op-erating between two fixed temperature iaits. In this Idealisedcyolel, the workla fluid is isentropically compressed betweenpoints d and a heat Is added Isothermally at temperatureTbetween points a and b g the fluid to isentropicaUy ox-pnded betwen points b and c; and heat is rejected fromthe fluid isothermally between points c and d . The efficiencyof the id*&sod Carnot cycle is independent of the propertiesof the working fluid.

Uing the temperature - entropy diagram of Figure 2,the efficiency of the cycle Is derived by Faires E21 as follewe

OA = Beat Added a area under curve a-b

TaA( N - a )

OR a Seat Rejected - area under curve c-d

a TI(Sb - sa )

V0 a alet work ef the Carmet cycle M 2A - 2

- TA(S-Sa)- 1 (%b-Sa)

- (TATaI)(%S a )

a Thermal efftloeey a w o - 2A-01

RA 2A

a (TA TO (1) S) TA I

TA TA

3

II

U tt

(A

w

-JC)

*~~~

0z

C.,

C.)U

'I

£wU

a

KIlL

I -~

h Sq. (2)D

w. USL AU a* A TA

we (3)

The heat reoteds tg o D shem by the Stefan-moemam oqustlea

OR Alt21 (4)

Cealag I. (4) mad (3), ,w ebtalna o ez "emiefor We*k o tom oie per Uit of radate area as a fuactisOf 2 A Oml 2as

0 9 Alt lt t T S

Pear naxmm work par alt radiator area coaslierlag T asthe eoml Waia ,eI - a 0. The.e/ore, to Red tA op-

imm rea N o .hp olrg N t T I t q. (S) is diffeenUatd withrepot, t ao Ta d tho is et .qual to Soros

1 -M a 3 Ta" TA4TI'" 0a AR Of&

TA,. ! 'A(6)

SI _ _ _ _ _ _ _ _ _ _

t -V

Iq, (6) sim that for ram Cas"e cycle werkussraitor aea# the heat lejoetle teoetue hemw

thve om'h @ the heat aditien temoa'aturs. !Ms feetban beow 1 ited at by seeuul Iveeatoa'e [Is 3, 43 -*0the wf ift umu, -two a513 miame that the rati ofadttw o to oysle whas hes udmieds or that the

asimm raator arneaem occur the Wweiater 10615"iotee-feinthe of the heat Maitie teieuature. Sq. (5)

the fenoeia oftee er be obtaimed tor the raatorarome

Lam (tw A

For mhd vaater ame# goo (6) saW (7) are sem-

bhed~~g Tiecaa s

This Owne that tw hel adiator armea "a be eel.oulatod N thoer oa ut Of the eyse sad the radiator

te~eua or ow h =etm Imlet -oestwao ame hoove Itim i that the heat aeeo MBeotim the Power'ouuspor that the Camet asle efflulemey is twenty-ftw Per-

It sheuM be noed that anml oe" deared fr theCaimet elsie are 1-1epolmiero the &a--elo of the vealagfihS sad do met semid"Or the I II-I~ of system ammpemeats sush as the pow dina eep'aht or the turhimo zua;a

ozpadss 2Ue aulyme WU be extemiled to the acteadlamlme yce, to esasliow th e amiea teeters wbich are

pwomsmt Is as actedl woring cyele.

b. skua mo ml

The acteud Mltime evole Io sm Is now*. 3. tevo~a fluld at elt4 Is Is a 1q54 state. ailis peessuedby pOQ to p so umah s st h atematba,pweine at pdsat I sad a,.ttI de to the 1q4 frepahl S mN It wOaSebe the Mat MMe tempOrstmr at pelst 1.

Mi~b eat Is adIded as the1 n fin -- methsey vIts"m ftpamin tws ela Ito pelat S.e I% 9" at pelat i I

S

mumu-

~'EWWEs I-S

Ig~

*1

mu-IU

0

a.,zI-zw z4

I.'.vi ______

7

expanded In the turbine to point '3. If the turbine were 100per cent efficient, the expansion' would be Isentropic to point3 . The fluid i then condensed isothermaily to point 4 andrejects heat during the process* At point 4, the cycle Isstarted again. The path between points 4 and 1 Is not isen-tropic as in the Carnot cycle because heat must be added tothe fluid in a liquid state between points 5 and 1. Therefore,It Is apparent that the absorption of heat by the liquid attemperatures less than the maximum temperature, TA, decreasesthe efficiency to less than that of the corresponding Carnotcycle.

An actual vapor-liquid Rankine cycle differs from theideal Rankine cycle because the actual pumping and expansionprocesses would be accompanied by entropy Increases, and theflow processes would involve pressure losses due to fluid fric-tion and changes in fluid momentum during phase changes. How-ever, a preliminary cycle analysis can be made by neglecting theeffects of all of these deviations except that for the turbineexpansion process. Also, the pump work may be neglected with-out serious results. Therefore, in the following analysis, theactual Rankine cycle will be considered as the process 1-2-3-4-1illustrated in Figure 3.

Referring to Figure 3, the heat added to the workingfluid may be expressed as the heat necessary to raise thetemperature of the liquid f ro. T to TA, plus the heat neces-sary to completely vaporize the Fluid. In other words, it Inthe change in enthalpy between point 4 and point 2:

QA - ha- h (9)

The heat rejected by the working fluid Is that heatwhich Is released during the condensation of the wet turbineexhaust vapor at point 3 to the saturated liquid at point 4, i.e.:

QR - ha - h, (10)

The work of the actual Rankine cycle, Wr , Is the dif-ference between the heat added and the heat rejected:

W r Q A - R "he- ha 11

8

~~t

Referring to Figure 3, the work Is the difference in enthalpybetween point 2 and point 3, or the change in enthalpy whenthe wrking fluid undergoes expansion in the turbine. If theturbine were 100 percent efficient, the expansion would beisentropic, I.e., the work would be the difference in enthalpybetween point 2 and point 3%. But because the actual turbineIs lose than 100 percent efficient, the expansion takes placewith an increase in entropy, to point 3. Thus, the work ofthe cycle may also be expressed in the following manner to takeinto account the turbine efficiency q :

W r - 11t (h - h3) (12)

The efficiency of the actual Rankine cycle is the ratioof the work of the cycle to the heat added to the cycle:

a-- h a- W t(h - h.) (13)r hs - hs ha- h4

To obtain a picture of the differences in the efficien-cies of the Carnot cycle, the ideal Rankine cycle (turbine 100percent efficient), and the actual Rankine cycle (turbine, lessthan 100 percent efficient), Figure 4 has been prepared. Thisfigure is based upon the use of potassium as the working fluid,with a turbine inlet temperature of 2100"R, and varying radi-ator temperatures. For the actual Rankine cycle, a turbineefficiency of 85 percent was assumed.

At this time, it Is convenient to introduce the workfactor, k , which may be defined as

k =!r . Wr (14)

in other words, the work factor, k, is the ratio of efflciency,and of the work, of the actual Rankine cycle to that of theCarnot cycle. The work factor will be useful in determining theoptimum temperature ratios and minimum radiator area for theactual Rankine cycle.

The work of the actual Rankine cycle may also be ex-pressed as the product of the heat added and the cycle

9

w ~0 m

0,

LL a

0 0

1- i0(1 w-

cr 0

wI - 0

(DW

'II-

IM - 0

IX w W

(L02

> F4

w

too

efficiency:

wr A Ia• 15

Combining Eqs. (11) and (15),

QA OR- QA 'qr

or

QA(1 - q) - QR (16)

The heat rejected by the radiator is:

Rx R"RcA' T" (4)

which may be combined with Sq. (16) to obtain

QA (1- q.) -=0AR TR"

or

Q AR T 1 ' (17)1 - qr

Combining Eqs. (1) and (14), one obtains

'1r -k (18)ItTX

which may be used in Eq. (17) to obtain the following ezpres-sion for the heat added to the actual Rankine cycle:

at AR TR" (19)QA IA - T1

TA1

Since the work of the actual Rankine cycle is expressedby Eq. (15), it may be combined with Eq. (19) to obtain:

o AR VT(20Wr"-g 1 -k :TI= (20)

(TAI

Using the value of 'I. from Eq. (18) in combination with Eq.(20), the work of the actual Rankine cycle per unit radiatorarea may be finally expressed by:

Wr k T TR

- k

TA R

As in the case of the Carnot cycle, the work of theactual Rankine cycle per unit radiator area can be optimizedby differentiating Eq. (21) with respect to T and settingthe result equal to zero, as shown in the foliLng steps:

Wr OCA R k -T R"

TA - k

TA -T R

a T (cAR k) T WA - Ti- )- TR A- k .) 0

TA

12U IJ

oI

4 k TT ( TA

~TA ~( ) TA -TR)'

4 kTR - TRTA ( 8 i-S)-4 TA(1-k) -0

Therefore,

TA (8 k-S) I4TA(8k-5Sf+ 64 kT A (1-k)8k

Rearranging, and disregarding the minus root, which does nothave a physical meaning:

TAT A 8k .8kV

This equation expresses the relationship between theradiator temperature and the heat addition temperature forthe case of optimum work and minimum radiator area of the ac-tual Rankine cycle. For the Carnot cycle, k - 1 , and it isinteresting to note that TR - 3/4 TA , which Is the same valueshown by 3q. (6).

It Is enlightening to visualize the effect that varyingvalues of the work factor, k, have upon the optimum temper-ature ratio T./T A * Figure S illustrates the relationship be-tween k and Tn7T, when values of k from 0.3 to 1.0 are in-sorted in Sq. (227. It may be seen that the value of T /Tlies between 0.75 and 0.80, and does not differ too greftlPfrom 0.75O, which Is the value determined for optimum Carnotcycle performance. This fact has been pointed out by Electro-Optical Systems, Inc. [5] . Thus, it may be concluded thatthe radiator temperature for the actual Rankine cycle will alwaysbe approximately three-fourths of the turbine Inlet temperature.

Equation (21) can be rearranged to provide an expres-sion for the radiator area of the actual Rankine cycle:

!13

0I--

IL)

0

It~

Jw

U.x

0

10

SC'a.

Wj/W1OLLV ~01

- S.

-,~. I

* -

(23)a z kTa4

Thus, with Eq. (22), It Is possible to determine theoptimum radiator temperature for any given value of TA , andusing these temperature values, to determine the minimum radi-ator area by using Eq. (23).

The turbine Inlet temperature of the working fluid hasa great effect upon the radiator area, as may be expected.Since the heat radiated varies as the fourth power of theradiator temperature, any increase In turbine Inlet temperaturewill result in an increased radiator temperature, and a conse-quent substantial Increase in the heat radiated. Figure 6 In-dicates the decrease In radiator area for progressively higherturbine inlet temperatures. For this figure, the actual Rankinecycle was used, with a system power output of 1 megawatt, aturbine efficiency of 85 percent, and potassium as the workingfluid. It should be noted that for minimum radiator area theradiator temperature Is always approximately three-fourtUs ofthe turbine Inlet temperature, as previously determined.

Thus far, the thermodynamics of the ideal Carnot cycleand the actual Rankao cycle have boem investigated, the dif-ferences in efficiencies of the cycles have been illustrated,expressions have been determined for the optimum relationshipbetwon turbine inlet temperature and radiator temperature andfor the minimum radiator area, and the effect of increasedtemperatures upon radiator area has been Illustrated. In thenext sectlen, suitable working fluids will be evaluated.

3. WORKIN FLUMS

Proper selection of the working fluid for the system i ofutmost Importance. The properties of the fluid will affect theefficiency of the cycle, the weight of the system, and theperformance temperature. The primary functions of the workingfluid are to absorb heat from the nuclear reactor heat source,provide driving force for the turbine, and carry waste heatfrom the turbine to the heat rejection system. In addition, itmay provide lubrication for moving parts In the system.

is

4k * a

1600 ACTUAL RANKINE CYCLE - I MEGAWATT POWERtL POTASSIUM WORKING FLUID - TURBINE EFFICIENCY 85%

/ Es 1.0wTL TURBINE INLET

oCi40 TEMP 2100R

100

1200 TURBINE INLETTEMP 2400*R

1000

600 -INLET TEMP

400

1400 1600 1800 2000 2200 2400

RADIATOR TEMPERATURE ,R

FIG. 6RADIATOR AREA VS RADIATOR TEMPERATURE FOR

VARYING TURBINE INLET TEMPERATURES

- .

Working fluids may be elements, alloys or compounds. Theelements are more favorable than the alloys or compounds be-cause of their greater stability during long-term operation.Alloys and compounds may decompose or change properties whenundergoing boiling or during high temperature operations orwhen subjected to nuclear radiation. Therefore, In this report,consideration will be given only to elements as possible workingfluids.

The following physical properties of the working fluid areof Interest when the fluid is considered for use in a turbineand radiators

(1) Melting point - low melting points are desirable forease of start-up,

(2) Boiling point - boiling points should be low enoughto permit condensation at reasonable pressureswithout requitring heavy structures.

(3) vapor pressure - vapor pressure should be lowenough at operating temperatures to avoid arequirement for heavy plumbing.

(4) Specific volume of vapor - specific volumesshould be low, or conversely, densities shouldbe highs to -inlmiso the sises of vapor passages.

(S) Density of liquid - the density of the liquidshould be low to reduce the weight of the fluidinventory.

(6) Viscosity - the viscosity of the fluid should below to reduce pressure drops in flow passages.

In addition to favorable physical properties as just de-scribed, the working fluid should also have favorable heattransfer properties, such as:

(1) Thermal conductivity - the thermal conductivityof the fluid should be high to assure adequateheat transfer with a minimum area.

(2) Specific heat - the specific heat of the fluidshould be high to minimise the fluid inventory andstructure.

17

(3) Latent heat of vaporization - the latent heatof vaporization should be high to mInimise thefluid inventory and structure.

For the power cycle as considered In this paper, the fol-lowing properties of the working fluid are also of interest:

(1) Neutron absorption cross section - the neutronabsorption cross section should be low to assurereliable long-term operation with the nuclear re-actor heat source.

(2) Corrosiveness - the fluid should be as non-corrosive as possible to assure compatibility withavailable structural materials.

(3) Long-time stability under high temperatures andnuclear radiation - the fluid must not lose Ithproperties during extended exposure to nuclearradiation and continued high temperatures.

A number of fire have investigated fluids suitable forpower cycles and cooling systems in space applications. One ofthe most recent studios was by Southwest Research Institute,under sponsorship of the Wright Air Development Center. Theirtechnical report by Weatherford, Tyler and Ku [6] states thatseveral considerations are necessary in selecting a suitable fluid:

"As design temperatures increase in the future, It maybe necessary to consider less volatile coolants than alkali metalsin order to avoid ocessive Rankine cycle pressures. With thisthought in mind, (the following table) presents a summary ofselected metals with boiling points ranging from that of mercuryat loss than 7007F to that of tin at nearly 5000°F. Two cr1-teria have been used for selecting the specified metals. Oneis that the eltiag point is less than 640*F so that start-updoes not present unnecessary problems. The other criterionis that the temperature range in which metal Is liquid at atmos-pheric pressure Is greater than 60 percent of the absoluteboiling temperature in order that cycle temperatures and, honce,cycle efficiencies will not be unnecessarily restricted by theavoidance of sold-phase formation or excessive vapor pressures.It should be noted that the described slecation criteria ruledout comelderatlon of aluminum cadmium, sine, magnesium, and as-timony which have bee oom previously as potential liquid-metal coolants.

18

- . .- .

metal Symbol Normal melting Melting Point-Boiling PoInt R Boiling PointPolnt,R Ratio

mercury 8g 1134 422 0.37Coum Cs 1720 543 0.32Raubidium Rb 1730 562 0.32Potassium K 1860 606 0.32Sodium Ia 2093 668 0.32Uthim U 2897 817 0.28Thalium TI 3132 1039 0.33Bimuth al 3298 981 0.30Lad Pb3643 1082 0. 30Indim n 4176 772 0.27Gallum Ga 4S18 U4S 0.12Tin Sn S328 909 0.17•

The discussion thus far In this section Indicates thatthe working fluid should preferably be an element, and thatliquid metals are attractive. In addition to the liquid metals,sulfur and phospho are In the correct temperature rangefor application in heat rejection systems. However, verylittle Is known about the corrosiveness or material compatibilityof sulfur and phosphorus at elevated temperatures, except forson* data showing that sulfur ts compatible with selected ceramicmaterials up to 1600 F E 6 ] . Therefore, sulfur and phosphorswill not be considered further in this paper.

As a first stop in selecting suitable working fluids, an

upper temperature limit can be Imposed. It may be reasonablyexpected that approximately 2OOP will be the temperaturelimitation of structural materials within the next decade or so.Using 2500°' as an upper limit, and selecting from the tablejust shown, the liquid metals then suitable for a working fluidare mercury, cesium, rubidium, potassium, sodium and lithium.

As a second stop in choosing the working fluid, consider-ation is given to the vapor pressures of the six suitable liquidmetals. Figure 7 shows the vapor pressures plotted againstthe temperature. This figure may be used to impose furtherrestrictions upon the fluid. A lower temperature lit of 600"Pwi be set to assure efficient operation of the cgcle and ofthe radiator. An upper tomperature limit of 2500 F has alreadyboon selected. A lower pressur limit of 3 psla will be selectedto avoid bulky and heavy flow passages due to the associatedhigh specific volumes. An upper pressure limit of several hun-dred psia will be selected to avoid heavy pressurixed structures.

19

10Vt

n33

pr4--

. . . . . . . . . . .

. . . . . .

When these limits are shown on Figure 7, It becomes apparentthat the most attractive fluids are rubidium, cesium, potassium,and sodim. Utbium appears to be more attractive for highertemperature system, and mercury for lover temperature systems.

A third step In narrowing the selection of the working fluidis to consider the neutron absorption crose sections of the sixliquid metals. The thermal neutron C22aR4 ) absorption crosssections of' the naturally occurring elements are given by (6]as follows:

mercury 360 ± 20 barnsCesim 29 ± .5 barms

Rubidium 0.70 ± 0.07 bars

Potassium 1.97 ±0.06 barns

aodlum 0Oj05s 0.010 barns

Uthum 71.00 1.0 barns

The high neutron absorption cross sections for mercury, cesium,and lIthium make them usltable for use in a vapor-condensingcycle utlisig a nuclear reactor heat source. An Isotope ofmercury, 3,804, has an absorption cross section of 0.43 1 0.10barns, and thus could be Considered for use in a nuclear re-actor If Its other propertes were more attractive than rubidium,

poasum or sodiu. e eisg that mercury, cesium, and lith-Lu are net well suited for the nuclear reactor - thermal systemconsidered here because of their high croe sections, the re-maiming fluids are rubiddiumt potassium, and sodium.

The physical and thermal properties of rubidium, potassium,,and sodium are shown in Table 1. Comparing: the properties ofthe flulds, several differences are significant. The density ofrubidum vapor at the boiling point Is approximately twice thatof potassium and approzimately four times that of sodium.* Thiswould Indicate, that smaller vapor passages would be required forrubidium. However, the latent heat of vaporination of rubidium,is less than that of potassium and sodium, so for a given powerrating, a greater mass of rubdum must be condensed, thus off-setting Its apparent advantage ofl higher vapor density. Thelewer viscosity of potassium vapor makes potassium appear ad-vantageous.* The higher specific heat and higher thermal con-ductivity of sodium, makes that fluid attractive.

The difference, in boiling point. for the three fluids re-quire different operating temperatures to assure optimum cycle

21

\0

V' C"N 0 CY'\I N 0 r-4t n c

.- I \~ 0 (vl 0 0 V-1 0l CC\ N -

c N CJC0 00000 ~OC\ \0 \0 "10

'-44

*'N Vr0 ' trON VIO El('4- 0 kn n xt' 4, \0 \0

+)C 0 N' (C'C\i 0 0 00NH- 0 -4 O r" -I-(CN0C)0o00 i-W\r

o o +4. jv ....

4-) 'd o 0 0\ r 0C000 0 C) UC12 .-- :4C1D 0 _: r *

-1 ztC' 00 n'C 0 0,00 0-

p.44

-- o(r C o 0 C oo - "- p4 V;0looc\''C =,0 ;C 4C l

m 04-)

-1

0d POO 0r-I -4 '-.r- 4

*9 4-4 4-' 4 -or, i-I

~4-30 C+4 4D.r4.-\0-0202 4- - ,44-4-rd - 4o

E-4 -4 r4 4 -.) 0 H j ~ -

22 ; 0-

efficiency and optimum radiator area for each fluid. Thus, itmay be seen that it Is difficult to select the most suitableworking fluid for the system described herein without making acomplete cycle performance analysis for each of the fluids, andthen comparing the analyses. Such a procedure Is beyond thescope of this report.

Potassium has physical and thermodynamic properties general-ly between those of rubidium and sodium, so may be consideredto be a reasonable choice. Also, the properties of potassiumare well known, which will enhance the accuracy of thermodynamiccalculations. Potassium is commercially available, and thus hasan advantage over rubidium.

Reference [6 J contains figures for the foregoing fluidsshowing the vapor pressure, liquid density, vapor density, liquidviscosity, vapor viscosity, and materials compatibility, as wellas a Mollier diagram for each fluid. In addition, Ref. [6] con-tains tables showing the physical, thermal, and thermodynamicproperties of the fluids, and should be consulted if furtherdata is desired.

23

_ _t

REFERENCES

1. D. Mackay : Powerplant Heat Cycles for Space Vehicles(paper presented at the Institute of Aeronautical SciencesSummer Meeting, Los Angeles, California, 16-19 June 1959).

2. V. Faires : ARolied Thermodvnamcs, The MacMillan Company,New York, 1948.

3. E. Pitkin : Optimum Radiator Temperature for Space PowerSystems, ARS Journal, Vol. 29, No. 8, p. 596, August 1959.

4. Study of Heat Rejection In Space, Interim Report 250-IR-1,Electro-Optical Systems, Inc., Pasadena, Calif., July 1959.

5. A Proposal for the Analysis and Development of a SpaceHeat Rejection System and its Optimum Integration into aSpace Vehicle, Electro-Optical Systems, Inc., Pasadena,Calif., September 1958.

6. W. Weatherford, Jr., J. Tyler, P. M. Ku: Properties ofInorganic Working Fluids and Coolants for Space Applications,Southwest Research Institute, WADC Tech. Rpt. 59-598, 1959.

24

DISTRIBUTION

GEORGE C. MARSHALL SPACE FLIGHT CENTER

M-DIRH-DEP-RKD

M-FUT-DIR

M-MS-I

M-G&C-DIRM-GMC-EM-G&C-NM-GMC-AM-GMC-R

M-LOD-D (Mr. von Tiesenhausen)

M-SWM-F (Mr. Barker)M-S&M-FE (Mr. R. Callaway)M-S&M-FE (Mr. W. Jordan)M-SM-EM-SWM-SM-S&M-SR (Mr. Blumrich)M-S&M-TSM (Dr. Kuettner)

M-RP-DIRM-RP-P (Dr. Lundquist) M-RP-S (Mr Thompson)M-RP-T (Mr. Holier) M-RP-R Mr. Mile.)M-RP-T Dr. Schocken) M-RP-R( Mr. Downey)M-RP-T Mr. Jones) M-RP-TS (Mr. Buchor)(6)M-RP-T Mr. Snoddy) M-RP-PCA (Mrs. Christopher (10)M-RP-N Dr. Shelton)M-RP-N Dr. Weber)M-RP-N (r. Seitz)

NASA HEADQUARTERS

Dr. W. C. Cooley LPDr. H. Harrison LPMr. H. Finger LPDr. F. Schulman LPMr. A. von Doenhoff DLMr. H. Rochen RPP

25

4 . '

LEIS RESEARCH CENTER

Dr. John EvvardMr. Henry SloneMr. Richard P. Geyo

LANGLEY RESEARCH CENTER

Mr. Joseph Hallisy, Jr.

JET PROPULSION LABORATORYMr. Herman Bank

26

-4

N . . , , I J .• . . . . .. .

4, r