NASA TECHNICAL

M E M O R A N D U M

ootX

00

<C

NASA TM X-I874

COMPARISON OF TUNGSTEN AND DEPLETED• — — -- •* — -W 'm •»• -* w ^ Tfc ~r •»- » *" •»' T TR JT •«' 1 r 7 "F^ l" V~^ "» T TTT"1 T' A X / ° I""' T> "f"

1 f~*\

L K A IN 1LJ M 1 IN iVl 1 IN I iVl U jvi- w J. I Vj r11, i.n i i; iv i; i /

SHIELDS FOR A SPACE POWER REACTOR

by Gerald P. Lahti and Paul F. Herrmann

Leu'is Research Center

Cleveland, Ohio

N A T I O N A L A E R O N A U T I C S AND S P A C E A D M I N I S T R A T I O N • W A S H I N G T O N , D . C . - SEPTEMBER 1969

https://ntrs.nasa.gov/search.jsp?R=19690026439 2018-05-25T03:03:40+00:00Z

1. Report No.

NASA TMX-18742. Government Accession No. 3. Recipient's Catalog No.

COMPARISON OF TUNGSTEN AND DEPLETEDURANIUM IN MINIMUM-WEIGHT, LAYERED SHIELDSFOR A SPACE POWER REACTOR

5. Report DateAugust 1969

6. Performing Organization Code

7. Author(s)

Gerald P. Lahti and Paul F. Herrmann8. Performing Organization Report No.

E-50799. Performing Organization Name and Address

Lewis Research CenterNational Aeronautics and Space AdministrationCleveland, Ohio 44135

10. Work Unit No.124-09

11 . Contract or Grant No.

12. Sponsoring Agency Name and Address

National Aeronautics and Space AdministrationWashington, D.C. 20546

'

13. Type of Report and Period Covered

Technical Memorandum

14. Sponsoring Agency Code

15. SJJppJ,ernentary;Notes

16. Abstract

The weights of layered spherical shields for a fast spectrum 2.2-megawatt-thermalnuclear space power reactor have been calculated for two candidate gamma shieldmaterials, tungsten and depleted uranium, using natural lithium hydride as the ma-terial for neutron shielding. Each shield configuration was optimized to a dose rateconstraint of 2 millirem per hour at 20 meters. A one-dimensional discrete ordinatestransport program and a steepest-descent method optimization program were used.

17. Key Words (Suggested by Author(s))

Depleted Uranium Radiation ShieldLithium Hydride TungstenOptimization

18. Distribution Statement

Unclassified - unlimited

19. Security Classif. (of this report)

Unclassified

20. Security Classif. (of this page)

Unclassified21. No. of Pages

17

22. Price*

$3.00

*For sale by the Clearinghouse for Federal Scientific and Technical InformationSpringfield, Virginia 22151

COMPARISON OF TUNGSTEN AND DEPLETED URANIUM IN MINIMUM-WEIGHT,

LAYERED SHIELDS FOR A SPACE POWER REACTOR

by Gerald P. Lahti and Paul F. Herrmann

Lewis Research Center

SUMMARY

The weights of layered spherical shields for a spherical reactor equivalent to a fastspectrum 2. 2-megawatt-thermal space power reactor have been calculated for two can-didate gamma shield materials, tungsten and depleted uranium. Three layers of gammashield material were placed alternately between four layers of natural lithium hydridefor neutron attenuation. The dose rate constraint was 2 millirem per hour at a radius of20 meters. The reactor core of uranium-235 nitride and tantalum had a radius of26 centimeters and was reflected by 11 centimeters of molybdenum. A one-dimensionaldiscrete ordinates transport program and a steepest-descent method optimization pro-gram were used.

The minimum weight of the seven-layer shield was 28 500 kilograms using tungstenand 25 300 kilograms using depleted uranium. About 70 percent of the 2-millirem-per-hour dose rate for both shields is due to capture gammas produced in metals. Foreither shield, if the reactor power was increased by a factor of 2, or the dose rate con-straint was reduced by a factor of 2, the shield weight increased about 10 percent. Areduction in reactor core radius reduced shield weight about 800 to 1000 kilograms percentimeter.

INTRODUCTION

The radiation shield required for manned space missions that use a nuclear reactorfor a heat source will be one of the heaviest components in the space power system(ref. 1). Consequently, the designer of such a shield must strive for weight reductionsby proper selection of materials and proper choice of the number, thickness, and ar-rangement of the shield layers.

For space power reactor shields, lithium hydride is generally considered to be one

of the best neutron shield materials. It has a high hydrogen density and a dominant (n,oi)reaction in lithium-6. Metallic tungsten and depleted uranium (0. 23 percent uranium -235) are candidate high-density gamma-ray shield materials. Both materials generatesecondary gamma rays by neutron capture and neutron inelastic scattering reactions(i. e, (n, y) or (n, n'y) reactions). A method of reducing the production of these second-ary gamma rays is to alternate layers of lithium hydride with layers of metal so as toreduce the magnitude of the neutron flux in successive layers of metal.

Because of the strong effect of geometry on weight, the placement and thickness ofthe metal layers can be critical. Further, because of the complex relation between neu-tron flux and the secondary-gamma-ray production rates in layered configurations, de-tailed neutron transport calculations are required to accurately follow the spatial distri-bution of the neutron flux and subsequent secondary-gamma-ray transport. To optimizea given layered shield with respect to weight for a particular dose rate constraint (thatis, to determine the optimum placement and thickness of the gamma-ray shield layerssuch that the total shield weight is minimized), an iterative optimization procedure hasbeen used.

The purpose of this report is to compare, on a consistent basis, layers of eithertungsten or depleted uranium as gamma-ray shield materials using lithium hydride asthe neutron shield material. The criterion used for the comparison is the relativeweights of minimum-weight shield configurations for a given dose rate constraint.These weights are determined for initially selected spherical seven-layer shield con-figurations of tungsten-lithium hydride or uranium-lithium hydride wrapped around aspherical uranium-235 nitride and tantalum core that is 26 centimeters in radius and isreflected by 11 centimeters of molybdenum. The reactor operates at 2. 2-megawatt-thermal output. The dose rate constraint is 2 millirem per hour at a radius of20 meters from the center of the core. Possible shield-cooling requirements were notconsidered. Also determined are the effects of changes in dose rate constraint, reactorpower, and core size on the shield weight.

METHOD OF ANALYSIS

The reactor is represented by a spherical core-plenum-pressure vessel arrange-ment and is surrounded by a molybdenum reflector and a shield that consists of fourspherical layers of natural lithium hydride alternated with three layers of either fullydense tungsten or fully dense depleted uranium. The initial shield-layer thicknesseswere selected to give a near-optimum weight, and were based on the results of previousexploratory calculations. A schematic representation of the assembly is shown in fig-ure 1. Increasing the number of shield layers beyond seven did not significantly reduce

Figure 1. - Schematic representation of layered, spherical assembly.(Approx. to scale.)

the shield weight. The composition of each region or material is given in table I. Theinitial configurations (layer thicknesses), the dose rates calculated for them, and theirweights are listed in table II.

The GAM II (ref. 2) and GATHER II (ref. 3) computer programs were used to obtaina 26-Lroad group (25 fast, 1 thermal) set of neutron microscopic cross sections and thePg downscatter matrix for each of the nuclides in the reactor-shield configurations.Homogeneous resonance calculations were made for the isotopes of tungsten and foruranium-238 mixed with lithium hydride to obtain a conservative estimate of neutroncapture cross sections in the resolved resonance region. The neutron energy groupboundaries are listed in table III.

The one-dimensional discrete ordinates transport program ANISN (ref. 4) with Sjgfull-range Gauss-Legendre quadrature was used with the Pg cross sections to obtainthe neutron flux distribution by energy group throughout the reactor-shield assembly.The dose rate from core neutrons was obtained from the flux distribution using flux-to-dose-rate conversion factors derived from the fluence-to-kerma factors of reference 5.All of the kerma was assumed to be absorbed locally. This permits direct conversion toabsorbed rad dose rate. An average value of 1 was used for the relative biological effec-tiveness (RBE) to convert the neutron rad dose rate to a rem dose rate.

TABLE I. - COMPOSITION OF MATERIALS

Region and material

Core:Lithium- 7NitrogenTantalumTungstenUranium-235Uranium-238

Plenum:Lithium -7TantalumTungsten

Pressure vessel:TantalumTungsten

Reflector:Lithium -7Molybdenum

Shield:Natural lithium hydride:

HydrogenLithium- 6Lithium- 7

Depleted uranium:Uranium-235Uranium-238

Tungsten

Atom density,atoms /barn- Cm

0.01158.01316. 01249. 002041.012305. 000855

0. 022975. 025590. 0022385

0.05118. 004477

0. 004595. 05763

0.056837.004217. 052620

. 000109

.0472

. 063229

Voidfraction

0.109

0

0

0.100

0

TABLE II. - SHIELD CONFIGURATIONS AND DOSE RATES

(a) Tungsten gamma shield; total initial shield weight, 35 900 kilograms;total final shield weight, 31 100 kilograms

Region or layer

Core

Plenum

Pressure vessel

Reflector

Lithium hydride -1

Tungsten- 1

Lithium hydride -2

Tungsten- 2

Lithium hydride- 3

Tungsten- 3

Lithium hydride -4

Total

Initialthickness,

cm

a26.0

2.5

.6

11.0

17.9

7.0

14.0

5.0

10.0

3.5

59.5

a!57.0

Finaloptimizedthickness,

cm

a26. 0

2.5

.6

11.0

20.6

9.8

12.34

3.0

10.33

2.54

39.3

a!38.0

Sourcecomponent

NeutronsAll gammas

All gammas

All gammas

Mo(n, y)Mo(n, n'y)Total

--

W(n,y)W(n,n'y)Total

--

W(n, r)W(n,n'y)Total

--

W(n, y)

W(n,n'y)Total

--

Initial shieldthickness

Final optimizedshield thickness

Dose rate from each source,mrem/hr

0. 0243.0030

.0020

. 0022

.2040

.0050

.2090

--

. 0921

.0097

.1018

--

.0988

.0278

.1266

--

.2010

.0947

.2960

--

0.765

0.2927. 0049

.0031

.0035

.2876

.0082

.2958

--

.6000

.0716

.6716

--

.3001

.0707

.3708

--

.2381

.1197

. 3578

--

2 . 0002

Radius.

TABLE II. - SHIELD CONFIGURATIONS AND DOSE RATES

(b) Depleted uranium gamma shield; initial shield weight,28 600 kilograms; final shield weight, 27 900 kilograms

Region or layer

Core

Plenum

Pressure vessel

Reflector

Lithium hydride

Uranium -1

Lithium hydride -2

Uranium -2

Lithium hydride-3

Uranium -3

Lithium hydride- 4

.Total

Initialthickness,

cm

a26.0

2.5

.6

11.0

19.9

9.0

12.0

3.0

10.0

2.5

40.0

a!36.5

Finaloptimizedthickness,

cm

a26.0

2.5

.6

11.0

12.0

6.27

16.78

4.48

14.66

3.16

42.58

a!41.0

Sourcecomponent

NeutronsAll gammas

All gammas

All gammas

Mo(n,y)Mo(n,n'y)Total

--

NeutronsFission yU238(n,y)

U238(n,n'y)Total

--

NeutronsFission yU238(n,y)U238(n,n'y)Total

-- .

NeutronsFission yU238(n,y)U238(n,n'y)Total

--

Initial shieldthickness

Final optimizedshield thickness

Dose rate from each source,mrem/hr

0. 4954. 0025

.0016

.0019

.1590

.0040

.1630

--

.0186

.0592

.1977

.0323

. 3078

,

.0049

.0715

.2176

. 0377

.3317

--. '

.0043

. 1445

.2940

. 0695

.5123

--

1.8162

0.3535.0085

.0054

.0061

.4599

.0135

.4734

--

.0173

. 0666

.2547

.0323

.3709

—

.0037

.0846

.2048

.0466

.3397

. --

.0031

.1371

.2388

.0641

. 4431

--

2.0006

Radius.

TABLE HI. - ENERGY GROUP BOUNDARIES

Energygroup

123456789

101112131415

161718192021222324252627

Neutrons,eV

1.492X107

1.2211.0008. 19xl06

6.074.974.073.012.472.231.831.351.119. 07X105

5.504.08

1.111.50X104

3.35X103

5.83X102

1.012.90X101

1.073.06x10°1.134. 14X10"1

0

Gammas,MeV

8.05.55.04.54.03.53.02.62.21.81.35.9.4.26.15.08

From the multigroup flux distributions and the reaction cross sections for producingsecondary gamma rays from GAM H/GATHER n, the number and spatial distribution ofcapture and inelastic scattering events in each of the nuclides were calculated. Thegamma spectra associated with these events (from ref. 6) were then used to define thespatial and energy distribution of secondary gamma sources throughout the assembly.Multigroup gamma cross sections and PQ downscatter matrices were generated by the

«5computer program GAMLEG (ref. 7), using gamma absorption (photoelectric plus pairproduction) cross sections from reference 8. The gamma energy group boundaries arelisted in table III. ANISN was then used with the S.g full-range Gauss-Legendre quad-rature to calculate the dose rates from the primary gamma sources in the reactor coreand each of the secondary gamma sources. The gamma flux-to-kerma conversion fac-

tors were calculated from the mass energy absorption coefficients for muscle from ref-erence 9. All of the kerma was assumed to be absorbed locally.

A separate calculation of dose rate was made for each individual source region andfor each type of reaction (capture and inelastic scattering). The dose rate from neutronsand gammas produced by fissions in the depleted uranium of the shield layers was alsocalculated separately. Separate calculations for each source region are required by theoptimization computer program. Separate calculations for the different reactions in agiven region are desirable because of the dependence of the respective interaction rateson different portions of the rapidly changing neutron energy spectrum (i.e., the spatialdistribution of neutron captures and inelastic scattering reactions are quite differentbecause the reactions primarily involve different neutron energy groups).

The shields were optimized by using the OPEX II computer program described inreference 10. The assumed dose rate-thickness model is of the form

where

and D is the total dose rate, D. is the i component of the dose rate (i.e., capturegammas from the first layer, . . ., inelastic gammas from the last layer), t. is the

th •*thickness of the j region, H-? is an "attenuation coefficient" that describes the effectof a change in the thickness of the j layer on the i dose rate component, and C.is a fitted parameter. To obtain the coefficients for each dose rate component, eachlayer thickness is systematically perturbed, and dose rates are recalculated; the coef-ficients are then calculated from the perturbed dose rates (see ref. 10 for details).The transport calculations for the perturbed thicknesses constitute the bulk of thecomputation effort since approximately i • j calculations are required.

The complete set of coefficients calculated for and used with the initial configurationfor each combination of shield materials is shown in table IV. In addition to being usedto determine the optimized shield, the coefficients are used to estimate shield weight forother-than-design dose rate constraints. From this model the program first determinesa set of thicknesses that meets the dose rate constraint, and then proceeds to minimizethe weight by the method of steepest descent. Because the dose rate-thickness model isinexact and the coefficients M^ are somewhat dependent on layer thickness, a final setof transport calculations is necessary to confirm that the total dose rate from the opti-

8

N 2tj o

I i

.-̂ I

m

QJ

1asu

Sourc

e

IH

IHO

B

'Si

COI

hoC

CMi

1

I

TH

sbO

FHO

-i_>U(U

1

1u.

cu

a>

O

-«

c"

f

.cf

|

c"

°e

a

h ~> ̂ m

'S » to "2 S5 <0 to 0 PCO CU <D Jj 5a > | srta »2 raaCOco

<D

c-co oo o c- ^H 1-1 enc— in co co co i~^ eg

d ' •

^ ^ *H lO CO »H COCO IO OS O C>3 ̂ J1 NT-H CM TH CN] C>3 CM O

b

O i-too ^>* co CD tn «-* o>co m cr> eg CM m CMrH CM *H rH O CO O

b

O CMCM OS CD CO CO O COO CO CM t- CM i-l CMCM CM CM «-< O CO O

O

o m CDO i-H If5 t- ^* CM COCO CO CM CM CM i-l CMCM i-t O CO O CO O

b

O C- CMCO CM r-4 CO CO CM COCD C- CM Oi *H CO CMCO CM O C- O t- O

O

O CO OS CM•^ m o 1-1 co to mCM Cft CM CO CM t- CMO t- O t> O I> O

b

CM t- T-< CDi-H in CD O O 1/3 OiCM O iH OS CM O i-lO CO O C- O CO O

b

t- "*** O CD-* CM O CD CM CO CMT-< C- CM CD CM CD CMO D- O t> O t- O

O

OS OO O f̂CO ^ CM O ^ CO ̂TH C- CM C- CM CD CMO C" O £> O C- O

0

m oo c- CM CM i-t o>CO ^ CO CO CM CO TH»H CM »H CM TH CM iH

O

iH CM CO ^1 1 1 1

1*? |«? 1? 1a -S s .a s s B5 to 3 w 3 to 3

1 11 II II3 H 13 H 3 H 3

3cu!cto

Id(43

•OCU

Itt<s

soo.

o

Sourc

e c

t-,0)>,3t,oco'3>2

CO

raniu

m-

D

niu

m-2

rtfcl

U

raniu

m-:

s

hO

•*-»u(4-1

2

Ple

num

cuSO

0>fc.a

_N

"cc"

Na

§

1 *E

_N"cc

Nc"

gin N10

£

>."c

NcT

coM ?-CO

fe_N

"ce"

N

ti

£ - S «.« 3 « -3 S «^2 to to >3 5 >,c to to s S oj"* a; <u y d t<

tH > 0, k0

P. to

d

I

toao1G

COco CM en co in c- mCD CM CD O C- CD egi-H. CM iH CM i-H O O

d '

CD t- ^* ^ CM en OC- O Ci CO i-H CO COTH CM l-l >-l CM l-l O

o

CO i-l CD CM O CD Oin csj m 1-1 CD co coy~t CM ^H CNJ TH tH O6 • • • _ • • •

COt- CD C- CM TT Tj< Tj<t- o o> t- eg r- ca<H CM i-l O O CO O

O

O CO in O5 00 CM COO5 1O •* O OJ O5 CM*-l ^H CM CM O CO O

0

t- *-l *-H CO CO ^J* COCD i-H CO t- CM O CM»-( CM ^* *-* O OS O

0

m IT5 1T5m t- co CM co CM coO5 CM CM CO CM CO CMT-< TH O CO O CO O

0

m CDr- CD c- o c- o c~O CM CM in CM lf> CMCM i-t O CO O 00 O

d

mCO i-H C- O C- O C-co t-i CM in CM m CMi-l CM O CO O CO O

o

in CD in CD in CD inCM CM CM CM CM CM CMO CO O CO O CO O

d

i-H CD *-H CD «H CD *-<CM i-l CM i-H CM i-H CMO CO O CO O CO O

ei

CM CO CM CO CM CO CMCM •-! CM •-! CM T-I CMO CO O CO O CO O

d

^* OS ̂ O5 ^* G5 ^*pa i-i eg I-H e»a IH caO CO O CO O CO O

o

00 (O 00 CD CO CD Oco c- eo t- co t- CDCSJ O CX) O CVJ O COiH N >-t W »H W i-l

0

i-H CM 00 ^1 1 1 1<D 0) (D O>•3 -O -O TJ

i i i s>. rt >. CV1 >. CO >.£ 1 Si 1 £ 1 £

'** C '̂ * C5 -FH C* "~*5 rt £ a £ rt 5;; ^ S tn S h J5J D J P J D J

mized shield arrangement satisfies the constraint. If the total dose rate does not matchthe constraint, OPEX II is rerun using the dose rates from the final set of transport cal-culations and the same set of attenuation coefficients. The set of adjusted shield-layerthicknesses that gives a total dose rate of 2 millirem per hour (without further optimiza-tion) is then considered to be the optimized configuration. The uncertainties in calcu-lated dose rate are those normally present in any discrete ordinates calculation and in-clude those due to uncertainties in the cross sections themselves, multigroup averages,and the effect of truncation on anisotropic scattering approximations. Although it is be-yond the scope of this report to determine the magnitude of the uncertainty in dose rate,a test of the conservativeness of the GAM II capture cross sections for the resonanceregion is made and discussed in the RESULTS section of this report. The weight penaltyassociated with a factor of 2 uncertainty is also calculated.

RESULTS

Optimized Configurations

The use of tungsten as the gamma shield material resulted in a total weight of31 100 kilograms. For depleted uranium the total weight was 27 900 kilograms. Theweight of the common core-reflector assembly was 2600 kilograms.

Dose rates from every region and source type are listed in table II for the final,optimized shield configurations. The dose rate from core neutrons was about 15 percentof the total dose rate of 2. 0 millirem per hour in both cases. The dose rate from fissionneutrons produced in the three depleted uranium shield layers amounted to only about7 percent of the core neutron dose rate or slightly more than 1 percent of the total. Sec-ondary gammas produced by capture events in the reflector and the heavy layers of theshield account for more than 70 percent of the.total for tungsten and 55 percent for de-pleted uranium. In both cases, about 50 to 75 percent of the neutrons captured in thetungsten or depleted uranium layers have energies between 1 eV and 3 KeV, 20 percenthave energies between 3 KeV and 0. 9 MeV, and the remainder are either thermal or haveenergies greater than 0. 9 MeV. This distribution is typical of all of the tungsten anduranium shield layers. In the first case, nearly all of the remaining 15 percent of thedose rate is from inelastic scattering events with the tungsten in the shield layers. Inthe case of the depleted uranium, the remainder is divided between inelastics (10 percent)and fission gammas produced in the shield layers (20 percent). The core gammasamount to less than 0. 5 percent of the total dose rate.

The fixed dimensions of the core, plenum, pressure vessel, and reflector, as wellas the final thicknesses of the shield layers, total weight, and dose rate components for

10

r' r

S OO

«"H »— i ^-* CVJ c \O S P*-O O O

» O O «""" C^J rO p^ "-— — -

rZ tSI

~

jaqiunu

11

the optimized configurations are listed in table II. The thicknesses of the shield layersshould not be considered unique. The region of minimum weight for the thick shieldsconsidered here is quite broad. Figure 2 illustrates the changes in layer thicknesses forthe tungsten case during the course of optimization. Each combination of thicknesseswould result in about the same dose rate at the detector (approx. 2 mrem/hr at 20 m),and the weight differences are small when compared to the total weight of a shield thathas an optical thickness of 15 to 20 mean free paths. For a dose rate constraint that ishigher by 2 or 3 orders of magnitude (i. e., a thinner shield), the minimum weight con-figuration is more sharply defined.

Estimates of Effects of Some Uncertainties

A transport calculation was also made to determine the importance of; the essentiallyunshielded secondary gammas produced in the outer layer of lithium hydride, which wereassumed to be negligible. This calculation considered captures in hydrogen andlithium-7, and inelastic scattering events with lithium-6 and lithium-7. The dose rateat 20 meters was 0. 033 millirem per hour, less than 2 percent of the total, justifying theassumption that the secondary gamma rays are negligible.

The conservativeness of the resonance capture cross section from GAM II wastested by estimating the resonance capture rate in the tungsten and uranium layers of theoptimized shield configurations shown in table II, using cross sections generated by thecode CAROL (ref. 11). This iteration of capture cross section with layer thicknesscould not have been done in the course of the iteration procedure without unduly lengthen-ing the optimization procedure. In some resonance energy groups the CAROL-calculatedcapture cross sections were lower than the GAM II values by as much as a factor of 2.This would .result in capture rates in the resonance groups that are lower by about60 percent. And since resonance capture represents 50 to 75 percent of the total numberof captures, the total capture rate, and hence to first order, the capture gamma doserate, would be lower by about 25 percent. This difference was about the same for boththe tungsten and uranium shields. Also, since the capture gamma dose rate represents50 to 70 percent of the total dose rate at the detector, the cross-section correctionmeans, in effect, that the shields were optimized to a dose rate constraint which was 10to 15 percent lower than 2 millirem per hour. This would not affect the relative shieldweight difference as will be shown in the next section of this report.

12

Variable Parameters

The consequence of a factor of 2 increase in reactor power, or a factor of 2 de-crease in the dose rate constraint, or systematic inaccuracies in the dose rate calcula-tions such that they are too low by a factor of 2, is an increase in the shield thickness toproduce an additional factor of 2 attenuation. The approximate effect of such an increasein attenuation on the total weight of the configuration was determined by using the opti-mization program to adjust the thicknesses of the shield layers to meet a dose rate con-straint of 1.0 millirem per hour. Using the original set of attenuation coefficients andthe layer thicknesses and dose rates from the configuration optimized to 2 millirem perhour, the optimization program adjusted the layer thicknesses to meet the new constraint.The weight of the adjusted configuration was compared to the weight for the 2-milliremconstraint. This was done for both tungsten and uranium and was repeated for a con-straint of 3 millirem per hour. A change of about 10 percent in shield weight was re-quired to achieve a factor of 2 change in shield attenuation for both tungsten and uranium.These results are shown in figure 3.

35X10*

30

25

TungstenDepleted uranium

\\\\

Dose rate constraint, mrem/hr

Figure 3. - Variation of total assembly weight with dose rate constraint. Core radius,26 centimeters.

Over a small range of core radii, it can be assumed that the leakage of neutronsfrom the core is independent of core radius. To approximate the effect of core radius onthe total weight of the configurations, all of the layers outside the core from the config-urations optimized to 2 millirem per hour were wrapped around cores with different

13.

radii. The thickness of the plenum, pressure vessel, reflector, and each of the shieldlayers was identical to the thickness from the 2-millirem optimized configuration for thatcombination of shield materials. The total weight was then calculated for each differentcore radius.. The results are shown in figure 4. Again, the variation was about the samefor the tungsten and uranium shields, about 800 to 1000 kilograms per centimeter of coreradius.

35xl03

•S1 305 '

25

— Tungsten— Depleted uranium

20 25 30 35Core radius, cm

Figure 4. - Variation of total assembly weight with core radius. Dose rate constraint,2 millirem per hour.

CONCLUSIONS

The use of fully dense depleted uranium layered with lithium hydride in a sphericalshield configuration for a nuclear space power reactor with a thermal power of 2.2 mega-watts resulted in an optimized shield weight of 25 300 kilograms. The uranium-235nitride and tantalum core was 52 centimeters in diameter and was reflected by 11 centi-meters of molybdenum. The dose rate constraint was 2. 0 millirem per hour at a radiusof 20 meters. The use of fully dense tungsten with lithium hydride resulted in an opti-mized shield weight of 28 500 kilograms for the same core-reflector configuration. Thethickness of both shields was about 100 centimeters, and each consisted of four layers oflithium hydride alternated with three of uranium or tungsten. Fifteen to 20 percent of thetotal dose rate was from primary neutrons. Secondary gamma production in the uraniumor tungsten layers accounted for the remainder. The primary gamma dose rate was lessthan 0. 5 percent of the total.

14

The effect on the system weight of variations in the reactor power level, changes inthe dose rate constraint, and changes in the core radius were also determined. The con-sequence of a factor of 2 increase in reactor power or a factor of 2 decrease in the doserate constraint is an increase of about 10 percent in the total weight. The variation withcore radius was 800 to 1000 kilograms per centimeter of core radius.

Lewis Research CenterNational Aeronautics and Space Administration,

Cleveland, Ohio, June 26, 1969,124-09-11-01-22.

REFERENCES

1. Lahti, Gerald P.; Lantz, Edward; and Miller, John V.: Preliminary Considerationsfor Fast-Spectrum, Liquid-Metal Cooled Nuclear Reactor Program for Space-Power Applications. NASA TN D-4315, 1968.

2. Joanou, G. D.; and Dudek, J. S. : GAM-II. A B3 Code for the Calculation of Fast-Neutron Spectra and Associated Multigroup Constants. Rep. GA-4265, GeneralAtomics Div., General Dynamics Corp., Sept. 16, 1963.

3. Joanou, G. D.; Smith, C. V.; and Vieweg, H. A.: GATHER-II, An IBM-7090FORTRAN-II Program for the Computation of Thermal-Neutron Spectra and Asso-ciated Multigroup Cross-Sections. Rep. GA-4132, General Dynamics Corp., JulyJuly 8, 1963.

4. Engle, Ward W., Jr.: A User's Manual for ANISN: A One-Dimensional DiscreteOrdinates Transport Code with Anisotropic Scattering. Rep. K-1693, Union Car-bide Corp., Mar. 30, 1967.

5. Ritts, J. J.; Solomito, E.; and Stevens, P. N.: Calculations of Neutron Fluence-to-Kerma Factors for the Human Body. Rep. ORNL-TM-2079, Oak Ridge NationalLab., Jan. 1968.

6. Celnik, J.; and Spielberg, D.: Gamma Spectral Data for Shielding and HeatingCalculations. Rep. UNC-5140, United Nuclear Corp. (NASA CR-54794), Nov. 30,1965.

7. Lathrop, K. D.: GAMLEG - A FORTRAN Code to Produce Multigroup Cross Sec-tions for Photon Transport Calculations. Rep. LA-3267, Los Alamos ScientificLab., Mar. 11, 1965.

15

8. Hubbell, J. H.; and Berger, M. J.: Photon Attenuation and Energy Absorption Co-efficients Tabulations and Discussion. Rep. 8681, National Bureau of Standards,Sept. 28, 1966.

9. Anon.: Physical Aspects of Irradiation. Handbook 85, National Bureau of Standards,Mar. 31, 1964.

10. Lahti, Gerald P.: OPEX-II, A Radiation Shield Optimization Code. NASA TMX-1769, 1969.

11. Stevens, C. A.; and Smith, C. V.: CAROL - A Computer Program for EvaluatingResonance Absorption Including Resonance Overlap. Rep. GA-6637, GeneralAtomic, Aug. 24, 1965.

NASA-Langley, 1969 — 22 E-5079

VATIONAI. AERONAUTICS AND SPACE ABMJ.VJSTRATJON> WASHINGTON, D.C. 20546

OFHCJAI. BUSINFSS FIRST CLASS MAIL

POSTAGE AND fEES PAID'NATIONAL AERONAUTICS AND

SPACE A&MlNISTRATtON

!t *.'i".pOSfa|( Stxtt'tn 15

'Tfi-c jcroH.tirtifal a ,icttrilh'$ .»/ the 'Un'jc^- Sutfet shift be

edge "of f'foenowetiii iti'S4ljalitj>rt)t'ufc for-f-kc

•af htjf/

ttfttsosfrhefe attd space. The AdtntnhirjlronV jj»tl iipprvfruffif d't£f€m'm$jim „,

I'" ff i" f '4

TECHKTCAL REPORTS: Sciewific andrrchntcAl Infflllftatioh cart$Wcmfcfrn

t* .uid a lasting comrihuttort

Ififorrtwion l^s }>roitj >K\)pfe bur M-verrM'k-ss ot import jf«v AS .1

TCCIiNICAL MEMORANDUMS:

- ofjfsrr'rhi«fe<n

CDN'I&ACTOR REPORTS. S,cir,atu;« andtcthnicj] inJomT.irinn ^c-ncmtt-.! wnJc-r ,\ XAtaiuraa t'>r I;T,UIV and: aTts-iJctcd ,m injport,i

cs A«F> Sflfet-: Apr OF 1^58..-,«

f, !,'•• £*tJ'i\<r-~fl*fm<'rfr*Vt~t-r&i!t ji'^Yf^m" *»»*«SA SClElSrriFlC AND Tl

TECHNICAL TRANSLATIONS: InformationpuWishwl to a ^eigrt langtmg* omddeftd <to merit NASA distribution in English,

SPiClAL PUIMJCATJONS; Jolormari<>nfrom orstrf vabte to

- •

soiircc'h'iyiJcs. a raff spcr«l hih!'ioe,r.if>hic^.

TUBf ICATIOHS; InformatMn oil-r*^ by NASA tf^tr may h<5 dfinitTi>t rn vr.jnoicKui arid oihtr

i^y IJtliy.iriun Rt.pnrt-s aad Notes.y Stu'fl^.

M ?/je avjilablfity cf fftsse f?ubt':scttKn$ may t>& obtained from:

AND TECHMICAL INFORMATION DfVIStON

NATIONAL AERONAUTICS AND SPACE-ADMINiSTRATION