FAD I

Report R-1769

IZTETZIMMIOTN OF 12MERFACIAL ENERGIES BETWEKIS SOLID

COPPER AND BIMARY BISUTH-THALLIMJ4 LIUID M1TAL

CODY $ -

"MiCROFICHE $ o 5by -

DDCB. J. RUGUS D

DDC-IRA E

AMCMS Cod& 5016.il 84400.01

July 1965

UNITED STATES A .lMYFRANKFORD ARSENALV PHILADELPHIA, PA.

MOIR /

DVC AVAIlABILT NUOIC

Qualified requesters may obtain copies of this report fromDefense Documentation Center, Cameron Station, Alexandria,Virginia 22314.

Release to CFSTI is authorized.

DISPOSITIOR INSTRUCTIONIS

Destroy this report wben ý.t ib no longer needed. Do notreturn it to the originator.

The findings in this report are not to be construed as anofficial Department of the Army position, unless so designatedby other authorized documents.

Report R-1769

IDJMINATI5 Of INRFACIAL I=RrIIS BETWEEN SOLID

COPPER AND BINALRY BISMH-TEALLIUK LIQUID METAL

CGIPOS ITIONS

BY

B. J. ROGUS

AMCMS Code 5016.11 34400.01

Quality Assurance Directorate"FRANKFORD ARSENAL

Philadelphia, Pa. 19137

July 1965

ABSTRACT

Interfacial energies were determined at 1250 7, 950 F, and 650 1between solid copper and binary bismuth-thallium liquid metal alloys.Interfacial energies vere determined between the copper and the liquidmetal environment as the relative proportions of bismuth and thalliumin the liquid metal were varied.

Results showed that progressive additions oi thallium to bismuthraises the interfacial energies between the resulting liquid metalenvironments and the solid copper. Interfacial energy values calculatedfor 1250 F ranged from 278 ergs/cm2 for a liquid metal ccmposition of100 per cenL bismuth to 360 ergs/cm2 for a composition of 100 per centthallium. Interfacial energies between solid copper and liquid bismuth-thallium euvironmer..zs increase as the temperature decreases in therange of 1250 F to 650 7. The effect of thallium on increa3ing theinterfacial energy becomes more pronounced as the temperature decreases.

TABLI OF CONTENTSPage

ABSTRACT i

gMILIlNTAL &O3IDURZ

A. lMaterial 1

I . Test Procedure . .. . . . 2

RISULTS 3

DISCUSSION 11

CmcwSICSOtNS 13

C 7ZRZC.S . . . . . 14

Ii

InTR0DUCT ION

Certain gaseous and liquid environments have an emabrittling effecton metals( 1, 2 , 3 , 4 , 5 , 6 ). One such embrittling enviro Pnt that is re-ceiving increased attention is that of liquid metals. A review of workconcerned with the embrittlement of metals by an envirorment of liquidmetals has been presented by Rostoker, McCaughey, and Markus( 7 ). One

factor that is related to embrittlment occurrence is the Interfacialenergy between the solid and the liquid metal( 8 ) 9). However, aich ofthe work on liquid metal embrittlement reported to date has not includedquantitative determinations of interfacial energies because of theexperimental difficulties involved. It has been shown by Morgn(8)that the degree of embrittlement of copper in the presence of liquidbismth-lead alloys increases as the interfacial energy decreases. Inthis system the interfacial energy decreases as the bismuth content ofthe liquid metal enviromPnt increases. Another system that eubrittlescopper is bismuth : thallium. However, no quantitative data on theinterfacial energies of liquid bismth-thallium alloys on solid copptrare available. Therefore, the purpose of this investigation was todetermine the interfacial energy for liquid bismuth-thallium alloyson copper. The interfacial energies were obtained by using the dihedralangle technique that has been explained by Smith(l 0 ).

KXPERIMENTAL PROCEDURE

A. Material

Base metal

An annealed sheet of OFHC copper was used for the solid metal.Thickness of the sheet was 0.194 1 .001 in. The chemical analysis ofthe material is given in Table I.

Table I

Chemical Analysis of Copper used in Interfacial Energy

Determination Studies.

Per Cent

Copper 99.97Iron Less than 0.005Lead Less than 0.01Silver Less than 0.01Tin Less than 0.00,Other element& No"n detected

Liouid Metal Alloy gCc~ositions

Sloven different alloy combinations were prepared from bisamthand thallium reagent chemicals. Prepared coapositions ranged in 10 percent weight increments between 100 per cent bismuth and 100 per centthallium. Solid bismuth and thallium were mixed together in vycor %esttubes and the atmosphere within the tubes was evacuated. The tubes werethen sealed and heated in a furnace at 750 F and were agitated after onehour at temperature to facilitate the formation of homogeneous liquidsolutions. The samples were retained in the furnace at 750 F for anadditional hour and were then allowed to cool. The tubes were thenbroken and the cylindrical metal pieces removed. These homogenizedalloy pieces weighed approximately 20 grams each.

B. Test Procedure

Preparation of Specimens

Circular disks, 1/2 in. in diameter, were punched from thecopper sheet. The disks were ann.aled at 1300 F for one hour to elimin-ate effects of possible cold wori present in the copper as a result ofthe punching operation. After anneal the pieces were mounted in bakeliteand polished metallographically to ensure that the flat circular faceswere free of scratches or mirtr defects. The disks were then broken outof the mounting medium, cleaned chemically, and placed flat on a stain-laws steel tray with the ,clished faces up. Using a jeweler's saw,approximately 0.2 g. of each prepared alloy combination of bismuththallium was cut from the homogenized samples. The alloy pieces were1.id on the polished faced of the disks. The test was perfo,-ted inquad•-uplicate for each composition of bismuth : thallium. The stainlessrtep: tray was then placed into a vacuum furnace. Prior to heating, the:urcace wag evacuated and back filled with argon to a slight positive;ecssure. Three groups of specimens were prepared. The first groupwas held at 1250 F, the second at 950 F, and the third at 650 F.

After 24 hours equilibration at temperature, the pieces wereallowed to cool in the argon atmosphere until the liquid metal solidified.The pieces were than quenched in water. Ueing a jeweler's saw the diskswere cut in half along a diameter so that the copper-metal interfacecould be examined on the cross section. The pieces were then mouzardin bakelite and the sectioned faces were polished metallographicalil andthen were et :ed with potascium dichromate solution. Microscopic ex-amination was made of the copper-alloy interfaces using a magnificationof I000%.

2

Dihedral Angl. Measurements

Grain boundary dihedral angles were measured on external anglesformed where the copper grain boundaries came to the surface at thecopper-alloy interface. Typical dihedral angles are shown in Figure 1.Neasurimnts were made by rotating the calibrated stage of a Bausch andLab metallograph to align the grain boundaries with a cross hair locatedwithin the eyepiece. Measurements were made of all clearly delineatedangles along the copper-alloy interface for each disk section and wererecorded to ± 0.5". Preliminary measurements were made of representntiveangles in order to check the reproducibility of the measuring process.Three typical angles were selected and each was measured 10 separatetines. The reproducibility was such that the measured values for anindividual angle were within ± 1.5" of the mean of the 10 measurementsrecorded for that angle.

A minimum of 60 argles were read for each condition of cim-position at the two higher temperatures. For the specimens equilib~itedat 650 F, a fewer number of measurements could be recorded since manyangles were poorly formed.

RESULTS

The equilibrium grain boundary dihedral angle for an individualcondition of temperature and composition was taken to be the median ofthe angles measured for that condition. These angles are given inTable II. The table also shows the number of angles measured as wellas the minim= an, maximm angle -values recorded for each condition.Interfacial energies were calculated by using the expression of equili-brium as presented by Smith(O/). The expression is of the form:

YB u 2YSL COs

whereT B - grain boundary energy of copper (ergs/cm2 )

'rSL - interfacial energy between solid copper and theliquid metal (ergs/ca 2 )

* - dihedral angle at equilibrium (degrees)

A constant value of 3-) J. ..s used for z. This value isthe mean value calcul.J "- c ' sber I-.:i••i) fron interface angleratios exrerimentally devc • b', .L>(>) Van Vack(12 ), Bailey andWatkins(13), and FulIr." • Il .

3

Copper Copper

L~quJA drp m mtal sztwmce. Kquilib?±um diagram foc extexualIz~m mfs nl djLWvral augla.

1007. Bisouth 1002, ThA~l1ttm~Mag. 500X Hag. 500X

Figjure 1. Diagrrcn and photomicrographs showing method ofmeasuring dihedral angles

4

100

0 % 00 0' 0- 0nfn4 04 1'

-4.

0 440 0 00 en CV 0n C 0

C4 0n 4l to 0- 0 ti-* N ^ cc

H 03 0 P4 1

00

01 0 0 00 0 0 00

0kC4 Go

-4 -T-4

A- 0%Ai " 0 0

44 ýa 4-4N N N

.0 CV) %0 '. 0

04 04 0 0

9 0 0 0 0In co4 N CO)~' O

C,

0 0 00 00 00 0

.0 C) 0 0 0 00 00 0 0 0 0

-' Ok N0 N4 -Z N r- C 4 0-- 4

An accurate evaluation of dihedral angles requires that the surfacecf the specimeu, the plane of the microsection, and the plane of thegrain boindary be mutually perpendicular. It should be noted that compli-ance tith the condition that the plane of the grain boundary must bep•ependicular to the microsection is clearly impractical when poly-crystallie s'%ecimena are used. Non-perpendicularity between the planeor the Sraio boundary and that of the microsection causes the measuredangle to be either larger or smaller than the true dihedral angle.

Rieggers and Van Vlack( 1 5 ) made statistical analysis studies ofthe relatienship between observed anglcs and true dihedral angles.They piopored a method by which thi nedian of the measured angles couldbe used as thu truz dihedral angle for an individual conditicn of com-position and tampcrat-_re. Their computations showed that the probableerror between the obtained median angle and the true dihedral anglebecame greateL as the dihedrLl -ngles increased toward 900. Moreover,the confidence level which could be attributed to selected limits ofprobable error was shown to increase as a greater number of angles wasmeasured to determine the median. For example, Rieggers and Van Vlackshboec that the measurement of 25 angles assured an accuracy of ± 5*in S`6 per cent of the cases if the true dihedral angle is 30' or less,and ± 5* in 68 Der cent of the cases for angles close to 90". In orderto increase the confidence level to 96 per cent for angles close to 900would reluire the measurement of approximately 90 angles.

TYnterfacial energies were calculated for each composition ofimiswuth : thallium for each of the three ejtilibrium temperatures

inveowiated. Calculations were made using the median dihedral anglesrecordcl in Table II and the value of 550 ergs/cm2 for YB" Interfacialenergy values for the 1250 F equilibra:.iott temperature are plotted inFigure 2. Values for the 950 F condition are plotted in Figure 3 andthe values for IS50 F are plotted in F.gice 4. The ranges oi valuesshown in these plots were obtained by considering the probable limitoof error of the median angle in the manner diccribed by Rieggers andVan Vlack. For the two higher temperatures, the 96 per cent confidencelevel for the probable limits of error was based on the measurement of60 angles. For the 650 F ccndition the same limits of error were maiatainedbut the confidence level was thereby reduced to 68 per cent since thenumber of angles measured was less. The intc.rfacial energy curves obtainedfor the three individual test temperatures are shown together in Figure 5.

For each of the three temperatures investig..ted the interfacialenergiesYSL) between the solid copper and trhe liquid bismuth-thalliumalloy compositions increased as the thallium content increasecd. Inter-facial energies calculated for the 1250 F temperature ranged from 273ergs/cm2 for 100 per cent bismuth to 360 ergs/cm2 for 100 per CeU Lb'• ,It can be seen from Figure 5 that the interfacial energies increiise.ýgradually as the equilibrium temperature decreaved in the range ofto 650 F. As can be seen from Table II the spread of measured diill.'-angies for an individual liquid metal composition was smallest Xor ýIe1250 F temperature condition. As the temperature decreased, the spreadof measured dihedral angles increased.

6

425 425

A )0 400

A 350 . ...- 350

I

30 • 300

Bi 100 80 60 40 20 0T1 0 20 40 60 80 100

Coposit io

Figure 2. Interfacial Energies between Copper aný. Liqu_'

Metal Alloys of Bismuth-Thallium. Specimens Equilibratid24 Hours at 1250F. (Confidence Level of 96 percent.)

7

42525

400 4... 00

I_ /S -

S350 •040

..- /.

)0000/ --

300 300

12501L 100 8o 60 40 20 0TI 0 20 40 60 80 100

Compoeitioc

Jigure 3. Interfacial Energies between Copper and LiquidMetal Alloys of Bismuth-Thallium. Specimens Equilibrated24 Hours at 9507. (Confidence Level of 96 percent.)

425 .25. _ Im I I• I i II1

400 . .00

/ 0/

S350 3, 50

04a

300 1'-- 300

* -_

Bi 100 80 60 40 20 0

T1 0 20 40 60 80 100

composit ion

Figure 4. Initerfacial Energies between Copper and Liquid

Metal Alloys of Bismuth-Thallium. Speoimons Equilibrated

24 flours at 650F. (Confidence LeAvel of 68 percent.)

.9

425- 425

400 - 400

........... 650 V

-- ---- 950 V ...

1250 V * SN

.335035

300 do,30

000

000

2501 250 I I IBi 80 60 40 20 0TI 0 20 40 60 80 100

Camposition

Figure 5. Interfaoial Energies between Copper and Liquid

Metal Alloys of Bismuth-Thallium. Specimens Equilibrated24 Hours at Temperatures Indicated.

• i0

DISCUSSION

The interfacial energy value of 282 ergs/cm2 calculated for aliquid metal composition of 100 per cent bismuth at 6L0 F agrees fairlywell with the value of 280.5 erg'!cm2 reported by Mc:rgan( 8 ) on copper:bismuth. Mcrgan, however, heated his specimens for 10 hours at 662 Fand used the most frequently observed dihedral angl as the equilibriumgrain boundary angle. In the present study the epeciwtns were heatedfor 24 hours at 650 F and the m-.dian of te measured angles was usedas the equilibrium angle.

The addition of lead to bismuth increasee the iq terfacial energybetween the liquid bismuth-lead allay and the copper%). This studyhas showna that the adcition of thallium to bismuth also increases theinterfacial energy between copper and the liquid metal environment.Morgan's work showed a value of 390.5 ergs/cm2 for a liquid metalcomposition of 100 per cent lead. In our experiments a value of402 ergs/cm2 was obtained for a liquid metal composition of 100 per centthallium at 650 F.

The values reported by Morgan for the copper-bismth-lead svstemwere for one temperature condition. With regard to the present study,examination of Figure 5 showed that the interfacial energies increasegradually as the equilibration temperature decreases. Moreover, theaddition of thalliu•r was seen to have a greater effect in increasingthe intertacial energy as the te.iperature decreases.

Interfacial ene.-rgy values tetermined from the measurement ofdihedral anglee =.ay b- infflurnced by factors such as the tiwe andtemperaturt of specimen equilibration as well as by possible effectson the grain boundary energy cf copper Nf temperature and relativegrain orientatiw). Scme. wention should be. made of the possible effectsof these factcra.

In thf cirrent srtudy, the 1':Iding of spec imnns for 24 hours attemperature wa3 cnsidered adequate to ensure equilibrium of anglesat grain boundary intersections. Smith(I 0 1, for his measurements ofdihedral angle5 cf th- copper-lead system, fo-ind 16 hcrrs at 1382 Fto be sufficient for equilibrium purposes. Sears(16), for similarstudies of lead dro'p -r coppe r, heated his ipecimens for 8 hours at1472 F. Bailey and Watki s'J 3 ) measured groove angles at t-e inter-section of grain boundari-s and a free copper surface, which weretlermally etched in a lead attoo3phere and repcrted identical resultsfor specimens held 2 hours ot 1652 F and those heId 5 hur at: :I472 F.

11

The equilibrium adjustment of interfaces probably is the result ofdiffusion(12). Therefore, comparisons of the effect of different holdingtimes to achieve equilibrium on the depth of penetration of the liquidmetal can be made in terms of reaction-rate principles. Application ofthe Arrhenius reaction rate equation would tend to indicate that compar-abi results would be expected between the work of the other investigat-orslO,13,16) and that of the present study for the 1250 F equilibriumtemperature. To achieve comparable depths of penetration for the 950 Fand 650 F temperature conditions wou'ld have required much longer holdingtimes.

In the opinion of Bailey and Watkins( 13 ), it is reasonable tosuppose that the equilibrium dihedral angle is estabilshed as soon asthermal etching begins, and that, as etching proceeds, the groovesmerely become deeper, while maintaining the same dihedral angle. Withrespect to the present study, no noticeable differences in dihedralangle were observed between holding times of 20 and 24 hours at 1250 F.It was noticeable, however, that the depth of penetration of the liquidmetal into the grain boundary intersections was less at 20 hours thanat 24 hours.

Under certain conditions the depth of penetration tends to influencethe measurement of dihedral angles. As the depth of penetration decreasesit tends to become more difficult to judge the exact limits of the di-hedral angles. This factor was believed responsible for the observationthat as the equilibration temperature was decreased the spread of themeasured dihedral angles increased. However, even though the angles forthe 650 F condition were more difficult to measure, the obtained inter-facial energy values appeared similar in nature with those for the highertemperatures. Values for all three temperatures are considered repre-sentative ofequilibrium conditions for those temperatures.

The grain boundary energy of copper, Y B, probably varies somewhatwith temperature. The interfacial energies calculated in this studywere obtained using a grain boundary energy value of 550 ergs/cm2 .Ikeuye and Smith(l), from studies of the ratio of the solid-liquidinterface energy to that of the grain boundary, for the copper-leadsystem, reported that the major change in the ratio with temperaturewas due to a change in the liquid-solid interface energy. This changewas chiefly the result of the change in composition of the liquid anddid not become pronounced until the monotectic temperature was nearlyreached, at which point the copper content of the liquid discontinuouslyincreased. Below a temperature of 1470 7 the ratio of the interfaceenergies for the copper-:.:- system was seen to be constant. Therefore,interfacial energy calculations made in this study based on Y beingconsidered constant over a temperature rang2 of 1250 F to 650P werebelieved to be meaningful.

1?

The grain boundary energy, Y B, of copper polycrystals alsodepends on the orientation of the boundary with respect to the crystalaxis of the adjoining grains. H-nsever, as noted by Cottrell(18), thisdependenc of the boundary energy - n relative grain orientation ispronounced only for small-angle boundaries. It is a fair approxinationto regard the energies of large-angle boundaries as being independentof orientation. The value of 550 ergs/cm2 for the grain boundary energyof copper was adopted by Fisher and Dunn(l) based on large-angle bound-aries. With respect to our experiments. microscopic exaaination of thecopper sheet material showed no pronounced indicacions of the presenceof textures. Moreover, after the preliminary anneal of 1300 F for onehour the grains were essentially equiaxed with lar1e-angle grainboundaries. If a different value were used for FBI the calculatedvalues of YTSL would change but the relative shape of the T SL vs.composition curves would remain the same.

CONCWLSIONS

1. At temperatures nf 1250 F, 950 F, and 650 F, progressive additionsof thallium to bismuth increases the interfacial energy between solidcopper and the resulting bis.uth-thallium alloys in the liquid state.

2. Interfacial energy values between solid copper and liquid bismuth-thallium compositions increase slightly as the temperature decreasesfrom 1250 F to 650 F.

3. The effect of thallium on increasing the interfacial energy betweensolid copper and liquid bismuth-th-allium compositions becomes sore pro-nounced as the temperature decreases.

13

RE FERE NCIS

1. Buzzard.* R. W.,. "Hydrogen Embrittlement of Steel" NIB Circular511., U. S. Government Printing Office (1951).

2. Hears, R. I., Brown, R. H., and Dix, 1. H.., "A GeneralizedTheory of Stress Corrosion of Alloys" ASTE Symposium:Stress-Carroeion Cracking of Meta~ls,, AST)( (1945).

3. Speller, F., Corrosion-Causes and Prevention McGraw Hill (1935).

4. $MQui on Cgorroaon Fundmntals. Symposium-University ofTennessee, March 1955, University of Tennessee Press (1956)

5. Otto,. H. I., Leighly., H. P., and Blockledge, J. P.., "Embrittlementof Alimintu Alloys by Mercury" Transactions ASh Vol. 55 No. 3 (1962).

6. Likhtmasn V. I., Rebinder, P. A.,. and Karpenko, G. I... Effect ofa jurfacce Active Medium on the Deformation of Solids Translationfrom Russian (1954).

7. Roetoker, W., McCaughey, J. M., and Markus, H. Iubrittl~ftnt byjAi-eLi X..t#..1 Reinhold (1960).

8. Morgan,. W. A., Thesis, Cambridge University (1954).

9. Zborall, R.., and Gregory, P., Journal of Institute of Metals,84, (1955).

10. Smith., C. S.., Transactions AIM, 175 (1948).

11. Fisher, J. C., and Dunn, C. G., Imperfectious in Nearly Perfeit

Ca~l John Wiley and Sons, New York (1952).

12. Van Viaak, L. R., Thesis, University of Chicago (1950).

13. Sailey, G. L. J., and Watkins, R. C.., Proc. Phys. Soc., London,631, (19501.

14. Fulasn, R. L., Journal of Applied Physics, 22, (1951).

15. Rieggrer. 0. K., and Van VLack, L. K., Transactions ADS, 218, (1960).

16. Sears, 0. W., Journal of Applied Physics, 21, (1950).

17. Ikeuy., K. K., and Smith, C. S., Metals Transactions, 185, (1949).

18. Cottrell,, A. H., Tbooretical Structural Metallurgy St. MartinsPress, New York, (1955).

14

DISTRIBUTIOI LIST

2 Defense Metals Information Center, Battelle Ma rial Institute,Columbus 1, Ohio

20 Cimanding Officer, Defense Documentation Center, Cmeron Station,Alexandria, Virginia 22314

1 Cinanding Officer, U. S. Army Research Office, Office ChiefResearch and Development, 3045 Columbia Pike, Arlington, Va. 22204

1 Commanding Officer, Army Research Office (Durham). Sm C0,Duke Station, Durham, North Carolina

1 Comanding General, U. S. Army Aviation Materiel Comand,St. Louis, Missouri 63166

1 Camanding General, U. S. Army Ilectronics Cmeand, Fort Monmouth,Now Jersey 07703

C~ nding General, U. S. Army Materiel Cosmand, Washington, D. C.2 &TTI: AMCPP-PS-QC1 AHRD2 ACRKD-RS-CM

3 Cosmandlig General, U. S. Army Missile Cand, Redstone Arsenal,Alabama 35809ATTI: AN5MI-RB Redstone Scientific Information Centor

2 Csianding General, U. S. Army Mobility Cciand, 28251 Van Dyke Ave.,Warren, Michigan 48090

2 CiwC ding General, U. S. Army Munitions Cooand, Dover, New JerseyATIN: Research and Ingineering Directorate

2 Cinanding General, U. S. Army Test and Ivaluation 'CZmand,Aberdeen Proving Ground, MarylandATTI: ANSTI

2 Cinanding General, U. S. Army Weapons Cosand, Rock Island,Illinois 61202ATTN: ANSW-RD Research Divi~tou

Cieanding General, U. S. Army 'ank-Au 1otiv Center,

28251 Van Dyke Avemne, Warren, Michigan 48090

1 Coanding General, Detroit Arsenal, Warren richiga• 48090

15

1 Conmndiun Officer, Ft. Dietrick, Frederick, Maryland 21701

1 Canding Officer, Harry Diamond Laboratories, Connecticut Ave.asd Van gess Street, N.W., Washington, D. C. 20438

1 Cinnding Officer, Idpwood Arsenal, Edgewood, Maryland 21010ATTN: Cbsuical R& Laboratories

2 Conoading Officer, Picatinny Arsenal, Dover, New Jersey 07801

2 Cm ndtng Officer, Rock Island Arstnal, Rock Island, Illinois 61202ATTN: 9320 Research and Development

2 Cnanding Officer, Springfield Armory, Springfield, Massachusetts,01101ATTN: SWSP-1 Research and Development Division

SWSP-KG Engineering Division

I Coomanding Officer, U. S. Army Auunition Procurement and SupplyAgency, Joliet, Illinois 60436

2 Comanding Officer, U. S. Army Engir -r Research and DevelopmentLaboratories, Ft. Belvoir, Virginia 22060

1 Comnding Officer, U. S. Army Electronics Research and Development

Laboratories, Fort Moiaouth, New Jersey 07703

2 Conanding Officer, Watertown Arsenal, Watertown, Massachusetts 02172

2 Cmsinding Officer, Watervliet Arsenal, Watervliet, New York 12189

1 National Aeronautics and Space Administration, Washington, D. C.

1 National Aeronautics and Space Administration, Levis FlightPropulsion Laboratory, 2100 Brookpark Road, Cleveland 11, OhioATTN: Library

1 Army Reactor Branch, Division of Reactor Developme'-t Atouic EnergyComission, Washington, D. C.

I Oq. Air Research and Development Command, Andrews Air Force BaseWashington 25, D. C. ATT11: RDDFK

1 Cinsnding Officer, Chemical Corps Engiatering Comand, ,'rzy ChemicalCenter, MarylanaATTN: Dir. Products Eng.

1 Cmander. Naval Ordrance Labs, Silver Spring 19, Maryland

16

1 Chief, Bureau of Ships, Departmout of the Navy, Washington, D. C.

1 Chief, Bureau of Weapons, Department of the Wavy, Washingtou, D. C.

1 Chief, Office of Naval Research,, Departuent of the Navy,, Washington, D.C.

I Director, Naval Research Laboratory, Anacostia Station, Washington, D.C.

2 Comanding General, Wright Air Development Division, Wright-Patterson2 Air Force Base, Ohio 45433

ATTN: ASRIC

Coanding Officer, U. S. Army Materials Research Agency, Watertown,Massachusetts 021.72

2 ATTN: AMXMR-ATL5 A1M-I1 AMS -ThS

1 University of Pennsylvania, Philadelphia, Pa. 19104ATTN: Dr. N. Brown, Dept. of Metallurgical Engineering

Cc.anding Officer, Frankford Arsenal, Philadelphia, Pa. 191373 ATTN: Pitman-Dunn Research Laboratories, L-10001 SMUFA-L7000, Mr. H. Markus2 Quality Assurance Directorate, Q1000I Test and Evaluatio, Division, Q6000

72 1 Basic Materials Evaluation Laboratory, Q61001 Engineering Mechanics Section, Q61205 SMUFA-Q6120, Mr. B. J. Bogus8 Technical Writing SeCLion, L-3100

17