,,_C"

;.... "_ NASA CR-.7"28 2B%,, !l'¢

i_,. ._ " ....._ WANL-PR(VVV).O02

< ' J '".d: '. ! November,1970

'_. _

'"_'_ Ir

.-,_

;.7 v.:. .,- 4 ._ t"

i_ :

:'_, 6."

,:,:.,,. [ THE VARES/RAINI TEST FOR REFRACTORY MEI"ALS

'> "- By

.... ,.,<! £-" t "

.: .., " . G.G. Lessmann And R.E. Gold

; _ _'(NASA-C_I-7282B) _HE VABESZ_AINT TEST FOR N72-15518 i

,_, }i_j i RE_'_ACTO_Y _BTA_S, _ASK 2 Final [leport. G.G. Lessi_a_n, et al (Westinghouse _lect£ic i

:_ '_ ,,Corp.) Nov. 1970 42 p CSCL IIF Unclas :,G3117 11925 ';__ \-..,_..<_:,,.._. !......_.........

_. " "Westinghouse Astronuclear Laboratory

I

:'"..,, r.-_"_ Prepared For

,, ; National Aeronautics And Space Administration

!::': _,, NASA Lewis Research Centerf._

i:" _ Task.II of Contract NAS 3-11827._ , Paul Moorhead, Project Manager

1972007868

https://ntrs.nasa.gov/search.jsp?R=19720007868 2018-06-10T07:08:15+00:00Z

f

/" (_) AstronuclearLaboratory

NASA CR-72828WANL-PR(VVV)-002

FINAL REPORT

THEVARESTRAINTTESTFOR REFRACTORYMETALSi

• by

G. G. Lessrnannand R. E. Gold

WESTINGHOUSEASTRONUCLEARLABORATORY

Pittsburght Pennsylvania 15236

preparedfor

NAT! ONAL AERONAUTICS AN D SPACEADMIN ISTRATION

Novembert 1970s

Task II of ContractNAS 3-11827

Fractureand Hot Crack Resistanceof Welds i_ T-111 and ASTAR-811C

e •

NASA LewisResearchCenterCleveland_ Ohio

Paul E. Moorhead_Project ManagerMaterials and StructuresDivision

1972007868-002

(_) AstronuclearLaboratory

FOREWORD

Thisprogramwasconductedby the WestinghouseAstronuclear Laboratory under NASA Contract

NAS 3-11827. Mr. P. E. Moorhead of the NASA Lewis ResearchCenter, Materials and

i StructuresDi_;islon_was the NASA Project Manager for the program.

The objectives delineated and the results reported herein represent the requirements of Task II

of Contract NAS 3-11827. Additional investigations which were performedasa part of this

programare the subjects of additional reports. The final reports for this contract are the fol-

lowing:

Tas.___k Titl_.__ee .Report Number

I Influence of Restraint and Thermal Exposureon NASA CR=72858Welds in T-111 and ASTAR-811C

II The Varestraint Test for Refractory Metals NASA CR-72828

III Investigation of High TemperatureFractureof NASA CR-72859T-111 and ASTAR-81lC

1972007868-003

.f

TABLEOF CONTENTS

Section Pag__e

ABSTRACT 1 I1.0 INTRODUCTION 2

2.0 TECHNICAL PROGRAM 5

2.1 ProgramMaterials 5

2.2 ApparatusDesign 8

2.3 Testing Procedure 10

2.4 Apparatus Qualification 14

3.0 RESULTSAND DISCUSSION 15

3.1 Screening and PerformanceEvaluation Tests 15

3.2 T-111 andASTAR-811C 22

4.0 CONCLUSIONS 33

5.0 REFERENCES 36

i!

1972007868-004

.f

• . . .... • • •" . •

(_ AstronuclearLaboratory

LISTOF ILLUSTRATIONS=

Figure Titl___e_e Pag.__e

1. Schematic:BendingTechniquefor AugmentingStrain by 3Varestraint Testing

I2. Varestralnt Test Apparatusfor Refractory Metals 9

t

3. ASTAR-811CSpecimenafter Varestraint Testing 12

4. Total Crack Lengthvs. Augmented Strain for Refractory Metal Alloy 13B-66Varestraint Test Spe:imens

5. Comparisonof Total Crack Lengthvs. Augmented Strain for Refractory 16Metal Alloys

6. BendAngle VSoTime for 1% Augmented Strain 17

7. BendAngle vs. Timefor 4% Augmented Strain 18

8. BendAngle vs. Time for 4% AugmentedStrain with 60% Bending 19- - - Force

9 SchematicShowingAnalysisof Varestralnt Test Usedto Calculate 20'; the ApproximateStrain Rate

10. Pre-test Microstructureof T-111 24 ,

11. Pre-test Microstructureof ASTAR-811C 25

12. Total Crack Lengthvs. AugmentedStrain for T-11| Alloy Varestralnt 28. TestSpecimens

, 13. Total Crack Lengthvs. AugmentedStrain for ASTAR-811CAlloy 29VarestralntTest Specimens

14. FusionZone Hot Cracksin T-111 SpecimenTestedat 4% Augmented 31Strain

ili

1972007868-005

LISTOF ILLUSTRATIONS(Cont.)

Figure Titl_e Pag___._.e

15, FusionZone Hot Cracks in ASTAR-811CSpecimenTestedat 4% 32Augmented Strain i

16, Comparlsonof Total Crack Lengthvs. Augmented Strain for All 34 ,Refractory Metal Alloys Evaluated This Program

|V

1972007868-006

(_) AstronuclearLaboratory

/

LISTOF TABLES

Tabl____e Titl...___e Pag_.___e

1 AlloysEvaluatedin VarestralntTestingProgram 6t

2 VendorAnalysesof ProgramMaterials 7

" 3 VarestraintDataforT-111 26

4 VarestraintDataforASTAR-811C 27

V

1972007868-007

AstronuclearLaboratory

/ ABSTRACT

Theprlnclple_ of the Varestralnt test developed by Savageand Lundin(1) were applied to the

evaluation of refractory metal alloys. Thisrequired developmentof a special tes_apparatus

employinga modestspecimenslze for improvedeconomyand compatiblewlth the hlgh purity

inert atmosphererequired for welding thesealloys(2'3). Minimumsr.,ecimenslze is desirable

becauseof the hlgh costof thesealloys andalso to permit timely testWngof limited production

developmentalalloys. The Varestralnt apparatusdescribedin this report satisfiestheserequire-

mentsproviding a timely andeconomic test for hot crack sensitivity. This test can be usedboth

for alloy developmentandproductionsurveillance. The testapparatuswasdesignedfor use in

vacuumpurgeddry boxesand is completely compatiblewlth high purlty inert atmosphere.Aug-

mentedstrain can be varied continuouslyfrom 1/4% thru4%. In addition to varying the

augmentedstrain - the primary Varestralnt variable - the effect on hot cracking of varying

other weld parameterssuchasunlt weld lengthheat input have alsobeen studied. Test

reproducibility, in termsof bendingor strain rate, wasdemonstratedusinghigh speedphotography.

Alloys evaluated in demonstratingthls apparatus inclucJedthe columblum-base alloys FS-85,

B-66, and SCb-291, and the tantalumalloys Ta-10W, T-111, T-222, and ASTAR-811C. All of

theseare commerciallyavailable except ASTAR-811C (Ta-8W-1Re-0,7Hf-0.025C) which isan

advancedalloy scaledup to productlonfor the first time in thls program. Primaryemphasis

wasdirected towardthe tantalum-basealloys T-111 and ASTAR-811Cdue to their currente

_mportancefor spacepowersystemappllcatlonsrequiring a combinationof excellent weldabillty,

ductility, andhlgh temperaturestrength. B-66 is a particularly usefulalloy for demonstrating

the flexlbility and ra0geof thls apparatusin defining hot ,rack sensitivity as a function of

Varestralnttestparameters.The propensitythls alloy exhlblts for hot cracking has been explalned

in termsof an emplrically-derlved correlationbasedon the relationshipof Mo-V-Zr contents.

In contrastto B-66 the otheralloys are relatively insensitive to compositionalvariations or weld

parameterselectlons. Excellent correlation wasachieved using thls test.

1

1972007868-008

*5

,f

.': .,

1;0 iNTRODUCTION0

Thisreport describesthe application of the Varestralnt Test to the evaluation of hot crack

propensity in refractory metal alloys. This test wasdeveloped by Savageand Lundln(1) and

hasbeen applied by others in the evaluation of moreconventional materials• For testing

refractory metal all_'s, constraints peculiar to thesematerials had to be satisfied asdescribed

in this report•

The Varestra:nt Test provides a meansof intentionally augmenting shrlr.kage strain during

welding as a meansof screening materials for hot crack sensitivity. Although hot cracking is

associatedwith shrinkage strain, in practice shrinkage is not readily measuredor calculated

because it is a complex function of weld parametersand weldment configuration. Hehce,

independent control of weld strain during welding is an attractive approach in testing for hot

crack sensitivity• Conversely, becauseof the complexity of the weld strains, useof the Varestralnt

data is ultimately dependent on correlation with field welding results.

Varestraint testing is accomplishedby bendinga test specimenduring welding to achieve a given

strain ucrosstile weld. This strain is determined by selection of the radius of curvature of the

bending die. This is shownschematically in Figure 1. Bending is initiated immediately prior

to the time the welding electrode is a_ the point of tangency of the specimenand die block•

Each level of augmentedstrain requires an individual test specimen• Bead-on-plate GTA welds

are normally employed in this test.!l

_ ' This test hasseveral intrinsic advantagesover other methodsof hot crack testing.

- • Thespecimenconfiguration is simpleand inexpensive to prepareand the test is inex- 'pensiveto run.

. • "a ' ' • '• The test is performedwhile weldlng'justas snrlnKge strainsoccurduringwelding.

• An infinite variety of weld thermalcycles andgradientsare screeneds|multaneously.• t 'Hence, screen=ngisachievedwlth minimal estlng comparedwith other hot ductility

test methods.

• The augmentedstraincan be varied independentlyof the other test variables.

-'" 2

1972007868-009

(_) AstronuclearLaboratory

L__ I I I In I ----_ .....

_,/CLAMP PLATE

WELDING,,,_RECTION,_

TOP VIEW li

J SPECIMEF OI)A_C r_n

T-_ r-flj-;.L,.- X

SPF.CIMEN(After ending)SIDEVIEW

FIGURE1 - Schematic:BendingTechniqupl_orAugmentingStrain by VarestralntTesting"'.

1972007868-010

. .

For thesereasons,this test'is even more attractive for exotic or developm_ 'al alloys

suchas refractory metal alloys. Theseare usually quite expensive. Hence, a maximum

amountof information mu_tbe obtained during testing with m_n_malmaterial consumptk_n.

Further, since these are usually of very I_mlted production, i_"!*:_sential that quallty standards

be developed independentof exhaustive field testlng. Finally, since hot crackTng tendencies are

sensitive to a wide spectrumof variables such as contamination, a discriminating hot cracking

test_h very attractive as a quality assurancescreenlng technique.

We have designedthe Varestra[nt test describedin this report specifically for refractory

metal alloys. The major design features accommod:'tedwere:

• Modestspecimensize due to the extreme cost of refractory metal alloys. Experimentalheats may cost several hundred doliars per pound• A_I x 6 x 1/8 inch specimen sizewasselected.

• All sourcesof weld atmospherecontamination were avoTded. The unit operates invacuumpurged weld chamberof highest qualify. The structural mater_aismustnotoutgas,and the potential leaks in the pneumatic drive systemrequired positive sealing.

• All sourcesof metallic contamlnaHon were eliminated. HoldTngfixtures and dTeblockswere madefrom molybdenumfor this purpose.

• The test unit wasdesignedas compactly as possible to permit maximumflexlb;lity inte.#ing, The unit is portable. It provides an unobstructedvlew of the specimenandfull :..:cessibili_y from one side so that specimenscan be changed in glove box operations.

• The speclmenwas located soas to permit electron beamwelding as well as GTA weld;ng.Only GTA welding hasbeen employed to date.

• Thestroke linkage wasdesigned to maintain simple bending throughout the stroke or tointentionally augmentbendingwith a proportionatet_nsile componentshouldthis '

become necessaryto achieve conformanceto the bendlngdle.

All of theseobjectivesweresuccessfullyachieved. Qualificatlon of this unlt consistedof the

following:

4

1972007868-011

® Operation in a monitoredinertgas dry box without compromlslngthe quality of the weldchamberatmosphere.

e Proofof discrimination by evaluating refractory metal alloys with differing field weldlngrecords=

1• Proof of reliability by testing numeroussamplesin one test run without breaklng the

weld chamber to air.

• Proof of reproduclbillty by utlllzing high speedphotography to measureconsistency ofbending rates for various dies.

2.0 TECHNICALPROGRAM

2.1 ProgramMaterlals

Thealloys evaluated in the'course of this investlgatlon are listed in Table 1 wlth their nominal

chemical compositions. Actual chemical analysessupplled by the vendors for the various alloys

are presented in Table 2. Check analyseswere performed for the interstitial elements C, O,

and N becauseof thelr pronounced influence on al'oy propertles, and, hence, asa quality

assurancemeasure•i

All materials were tested as0.125 inch thick sheet in the fully recrystalllzed condltlon. Re-

fractory metal alloys are nearly always usedin the recrystalllzed condition due to the obvious

_ advantagesoffered in termsof long time thermal stability.

• _ Theoverall weldability and elevated temperaturestabillty of thesealloys hasbeen the subject

. of a recently-concluded NASA-sponsoredresearchprogram(4). In termsof overall weldability

° thesealloys maybe "ranked" fromeasiestto mostdifficult as:

:-:, I.To-low,SCb-,9

i 2. FS-85 'iiiiiiiii _ 4.3'T-111t T-222_B_66 AST'6R'811C(being evaluated in a c_jrrentprogram)

1972007868-012

/

f

Alloy Nominal Composltion (w/o) °

T-111 To-8W-2Hf

ASTAR-811C Ta - 8W - 1Re - 0.7Hf - 0.025C

FS - 85 Cb - 27Ta - 10W - 1Zr

T - 222 Ta - 9.6W - 2.4Hf - 0.01CTa - lOW Ta - lOW

B - 66 Cb - 5Mo - 5V - 1Zr

SCb - 291 Cb - i OW - 10Ta

........ i ii

TABLE 1 - Alloys Evaluated |n Varestralnt Testing Program.

!

°

'_ 6

1972007868-013

Theabove ranklng is of coursesuk_ecttothe conditions that all welding be done in an ultra-

high quality inert welding atmosphere. Thlsrequirement is not unique for the programalloys

but is general for all refractory metal alloys. Thlsranking doesnot imply any particular dlffi-

culty exists in welding thesealloys but ralher that somealloys tend to be moresensitive to weld Iparameter variations than others. In fact, hot cracking hasonly been observed in one of these

alloys, B-66, and ther; only when the (Mo + V)/Zr ra.tlo is allowed to exceed 10.2(5).

2.2 Apparatus Design

TheVarestralnt test apparatus designedand built for testing refractory metal alloys is shown in

Figure 2. The apparatusis shown in the post test (specimenbent) posltlon. It incorkorates the

following features:

• All stainlesssteel construction except for specimenclampsand dle block which aremolybdenum.

• A double sealed pneumatic drive. Thepneumatic cylinder is seal_,dwffhln a stainlesscan. Theplunger is sealed with a welded bellows and the can utilizes a static O-ringflange seal.

• The plunger is equipped with an adjustable stop so that specimensare not bent overthe end of the radius blocks, ' ,

• Thestroke arm is mountedto a pivot arm which servesthe purposeof maintaining simplebending throughout the stroke. An alternate'connectlon is provided for the stroke armon the pivot arm to augmentbending with tensile stressif requlr_d.

• Triggering occurs at the specimenso that the entire stroke mechanismcan be preloaded.Thiswasdone to achieve minimumdelay in triggering and maximumreproducibility.Before firing the trigger restsagainst the trigger release. It isshown locked in thereleased position.

• During welding the thorlated tungstenelectrode traversesthe test specimenfromright to left. The bare electrode is mounted in a molybdenumchill block of suf-ficient massto provide cooling of the electrode during the short run time.

• Rollingelementbearingsare usedat all pivot pointsto mlnlmlze strokeresistance.Linearbearingsare usedon the electrode traversemechanlsm.

8

1972007868-015

(_ AstronuclearLaboratory

/

_ FIGURE 2 Varestralnt Test Apparatus for Refractory Metals

[' t

1972007868-016

!

!iI

• In the flnal conflguratlon of thls unit the electrode traverse mechanlsmwas mountedindependently to ellmlnate shock transmlsslonand consequentelectrode vlbratlonduring firing. !

I

o Flexible stainlesssteel convolutedgaslines (not shown)are attached to the pneumaticconnectlonsfor charging the cylinder. Heliumor argon is usedfor this purpose. 0

• The strokemechanismis activated automatically by meansof a mlcroswltchattached .to the drive mechanism( not shown).

2=3 Testlng Procedure

Testspecimens0.125 inch x 1 inch wlde x 6 inches long are loadedcantilever fashion into

the testlng fixture with two 0.125 inch x 0.5 inch wlde x 6 inches long molybdenum "bending

bars." The bending barsare located approxlmately 0.5 inchesapart on each side of, and

parallel to, the Iongltudlnal dlmenslonof the specimen. The useof bendlng bars prevents

kinking or preferential bending of the specimenand guaranteesconformanceof the speclmen

to the die block. The actual test is accomplished by the suddenapplicatlon of the bending

load as the gas tungsten-arc welding electrode approachesthe polnt of tangency between the

curved surface of the die block and the cantilevered speclmen. Thearc is allowed to travel

an additional 3/4 to 1 inch before being extingulshed.

The magnitudeof the augmentedstraln to which the specimenis subjected at the instant of

bendingis:

' augmented(tangentlal) strain = t/2R

where t = specimenthickness,andR = radiusof curvatureof the block '

Hence, by varying the radiusof curvatureof the die block one can vary theaugmentedstrain.

For thls programdie blocks were machlnedout of arc cast molybdenumhavlng radii of curvature

soas to allow testingof 0.125 inch thlck specimensat augmentedstrain levels of 1/4, 1/2,

3/4, 1, 2, 3, and 4%. Since the magnitudeof the augmentedstrain is not a functlon of the

10

1972007868-017

AstronuclearLaboratory

welding parameters,the effects of welding process,alloy composition,andother parameters

which influence the metallurgical features of the weld can be separately evaluated from those

due to externally imposedmechanical restraint.

!For the purposesof this programand in order not to introduce too great a numberof variables,

welding parameterswere selected whlch would give a reproducible weld width and weld pud-

dle geometry. The bulk of the GTA welding wasdone at 15 ipm, a weld speed found to be

practical for virtually all refractory metal alloys (4). Where the numberof specimenspermitted,

testswere conducted at additional welding speeds.Specific weld parametersused for the various

alloys are provided later in the Discussionsection of this report.

Following testing, the as-welded speclmensare examined for cracksusinga low power bench

microscope. The total combined crack length in the weld fusion zone is usedas the index for

hot cracking sensitivity of the weld metal(1). The total crack length is determined by meanst-4 of a calibrated reticule located in the eyepiece of a metallurgical microscope. In general,

._ crack counting was performeddirectly on the as-tested specimensurface although for purposes

t of comparisonseveralspecimenswere examinedafter approximately 0.01 inches had been removedh"• from the as-tested surface by polishing and etc rag. The crack length values resulting from

the two inspectlon techniquesvaried little, so long as the amountof material removedwasnot

excessive. In general, only thosespecimensintended to be examined _nmoredetail metallo-

graphically were polishedand etched.

J

As-tested and lightly polished and etched photographsof the sameareasof an ASTAR-811C

- ' specimenare shownin Figure3. Cracking in the weld zoneand in the heat-affected zone (HAZ)

4 are readily seenfollowing testing at 4% augmentedstrain during GTA welding at 15 ipm.

A typical plot of total crack length in the weld fusionzone vs. augmentedstrain level isshown

in Figure 4 (for B-66). Figure 4 also showsthe effect of varying weld parameters(i.e the weld

, speed- hence the heat input/unit weld length). The effect seenfor B-66 of moreseverecrack-

ing at higher welding speedswasobservedfor all metalsand isdiscussedlater in this report.

- 11

1972007868-018

:_:_:_I_.I.JUCIB!LITY _'THE ORIGINAL PAGE IS '_

mJ--.mlV.., • ,w........ w,, . _ , .,._ .... ., je v w I " '. .- p , _'el,, ."" f" "'.,..,r • • , . .' . ."i,. ,C' _" " ', ..... ' ; =) ; . , ' " , ......

"." "_,* _ "._'.,.'," ' ''_ _'_ 'r. • "6"¢ --- _ ..... :"_ 'I

I. ,-_;_ ",".,.,_-,,_ • ' ¢ ( ' • ] , * '"

Ik::,.._,_"'" ' (." ,__-;. ". : ". _ : .... - . -.;

. " / ".,k", . " .:_.: -_.".,".'

:, , . . - _- .. , .' . ' |• "" " "'' Fusion

: "" "_ " "-," :"t", "'" "" "J, ". ' "*:'" ; , " ---_- • " Zone '!, ' ,_. . ' _, _, . ,'. . -..,: ..... ., ',..._: . • / : --: ..'-'.,"-. :, ,/' _" .:.:'-,-_A,_,..". :..":-.'..'-_,.:

t ' / ._' :I# ..---,. ' "'_.:.._ ..". _"; _:

. , . • ._. ___.,'.-'._. ..... -,:'.,-- . .--;.,-:

-_ '._..... .... ._: ...... :•._- - -.._....-;.-..... :

t..: • _': "" " _ .... ;'_1""--:- " " _-_ _ ' " ,_' "_ "_'_ k _" '_ "'_._._ '_' "t_:-_

AS TESTED 15X

_Welding Dffectioh

V .'-" '-* • " -, ' .'"_' / I ,..," . _'.......r.':_

_.:.'-': ' '-'"-.:/:,... .." •,":*--_.,,'A1,:'-,,:,'., . " : .,, . : • _>.-._.k .'''-' '" )*- ." ' :"!:. . -.-,:,,.-..- .:_,:-,. . --:.I ,: J.,.- --..

" ., :,_.: X . , _• . , .." .

_::."• ....- " - " :.-_\? '' i.t • ' s ..k'< '.I:-' "' '. " " ....." "":.".::.i!.-,"_'.• ,, '.,,,.'"_ .

LIGHTLY POLISHED AND ETCHED 15X

FIGURE3 - ASTAR-811CSpecimenafter VarestralntTesting.

12

1972007868-019

,f

@ AstronuclearLaboratory

1000 I I i •

900 B-66 . _ I_'"- Very Narrow 15 ipm weld -(seetext )

I 800 - FusionZone Cracks Only,

' E 700 -

-rI-.,0 600 -Z14.1.,--J

500 - Welding Speeduev,

•u 400 --15 ipm

O 7.5 ipm300 _3 ipm

200- _ -

100- "" _ "_

0 _ I I I I ,0 1 2 3 4

AUGMENTED STRAIN, PERCENT

e

FIGURE4 - Total Crock L'engthvs. AugmentedStrain for B-66 Al'loyVarestralnt Test Specimens.

13

1972007868-020

2.4 Apparatus .Qual_fication

The Varestralnt test apparatuswasqualified for this programby determln_ng:

• Its effect on weld atmospherequality.I

• Its performance in correlating test results wffh field experience for several refractorymetal alloys.

• The reproducibility of resultsas discussedin a later section of thTsreport, and

• The actual bendlng rate using hlgh speedphotography.

Excellent resultswere ach _evedin all the qualification tests conducted. Useof the apparatus

in a hlgh purity hellum atmosphereproduced no detrimental effects. After six hoursof exposure,

during which tlme four testswe_'emade, a helium samplecontained the following impurities.

Elementor Compound ppm(Volume)

N2 5.18

0 2 1.03Ar 0.11

CO2 0.43

CH4 0.06

C2H6 0.14

C3, C4 hydrocarbons 0.10 .

C5, C7 hydrocarbons 0.09Ne ,5.94 ,

H20 <1.0

Lessthan 8 ppmtotal active impuritieswere presentafter six hoursoperatlondemonstrating

that the testapparatushadno detrimentaleffect on theweldingatmosphere_2)."

14

1972007868-021

.f-

(_.,) AstronuclearLaboratory

Several alloys were chosenspecifically to assessthe ability of the Varestralnttest to discrim-

inate betweenalloysof knownhot crack sensltlvlty. Resultswere in excellent agreem,ntwith

practice. Theseare summarized in Figure5. Thealloy sensitivity is in full agreementwlth ob-

servedbehavlor(3) as discussedlater underResults.

tResultsof high speedphotography of the operation of this apparatusare summarizedin Figures

6, 7, and 8. Figures 6 and 7 depict the effect of using different bend radii while Figure 8 shows

the effect of lowered firing pressure.All three figures correlate bend angle and time with Io-

cation on the test specimenrelative to bend initiation. Approximately the first 3/4 inch from

the start of the bend is relevant in this test. Hence, the substandardpressureresulted in stretch-

out of the stroke by a factor af approximately five. For these testssteel bending bars were us=d

with aluminumspecimensto simulate the hot refractory metal specimensand molybdenumbend-

ing bars used in the actual tests.The photographic data wasaccumulatedusing a cameraspeed

of 4500 frames/second.The photographic evaluation revealedentlrely satisfactory operation of

the Varestralnt tester with the tester providing a very fasb nearly instantaneousbend. The photo-

graphic results were usedto calculate the approximate strain rate by the methodshown in Fig-

ure 9. As seen in that figure, to s_mplify the analysis the deformation is taken to occur in a

"plastic hinge" region having a width approximately equal to the specimenthickness. The strain

rate calculated in this manner is 103 mln"1. This high strain rate, per se0.did not seemto affect

the resultsand there did not seemto be any extraneous effects on the ranking of the alloys tested.

3.0 RESULTSAND DISCUSSIONL. _

3.1 Scree.nlngand Pe.rformanceEvaluation Tests

. Preliminary tests were performedon the alloys B-66, Ta-10W, SCb-291, FS-85, and T-222.

The purposesof these,testswer_ several. First, by using a serlesof alloys having well-deflned

general welding characteristicswe were able to assessthe ability of the Varestralnt test apparatus

to dlscrlmlnate between them. In addition, whereasmostof the experimental data available

regarding the welding characteristicsof thesealloys hasbeen accrued using specimensof such

size and geometry that little mechanical restraint exists, the Varestraint test is capable of

imposingan exact, known longitudinal strain on the outer surfaceof the weld specimen. Hence,

15

1972007868-022

1000I I 1 1

900 - All data for welds madeat a welding speedof 15 ipm.

Data for fusionzonecracksonly. I800 -

e

I/I

. 00--

600 -

_" 500 -- B-66 --.

iv,

u 40O--...I._ T-222I--O_" 300-

200

100SCb-291 ,

0 Ta- 10W0 I 2 3 4

AUGMENTED STRAIN,PERCENT

FIGURE5 - Comparisonof Total Crack Lengthvs. AugmentedStrainfor SeveralRefractoryMetal Alloys,

16

1972007868-023

.f

,., .." . . . ,p.

'i (_ AstronuclearLaboratory

Distance from start of bend, inches

0.2 0.4 0,6 1.0,:v I , i , , I I - ,

15-

"o lu- • -

c<

_ "10" mented Strain

"'_ 5-_ _" 80 psi Cylinder Pressure

..__._. / . (325 lb. Bending Force )

o f I I I , ,,0 5 10 15 20

Time _ seconds x 103

,_ ,

FIGURE 6 - Bend Angle vs. Time for 1% AugmentedStrain.

1

I

1972007868-024

°f

/ -

D_stancefrom start of bend, _nches

0.2 0.40.6 0.9 1.0 .TI I I I I

40 _e

30(D

"O

20_: - AugmentedStrain -"O

.......... _1_ - ' . 4% AugmentedStrain ....10 i 80 psi Cylinder Pressure -

. (325 lb. Bending Force )0 I i I 1

0 ! 0 20 30 40 50 60 70 80 90

Ti.me,secondsx 103

FIGURE7 - BendAngle vs. Time for 4% AugmentedStrain.

18

1972007868-025

--_ AstronoclearLaboratory

D_s.tancefrom start of bend, inchesI

0.1 0.2 0.4 0.5 0.6" 0.7 0.8 0.940 , I , ',' , , , , ,

30 -

.@

20

<

"_ .... 4% AugmentedStrain10 50 psl Cylinder Pressure =

(210 lb. BendingForce )

J

.0 I I I I I I I . I I I I I0 10 20 30 40 50 60 70 80 90 100 110 120 130-

Time , secondsX 103

FIGURE8 - BendAngle vs. Time for 4% AugmentedStrain with60% BendingForce.

19

1972007868-026

4--

t

_L

[_o°/ Total Strain_Outer FiberI -/ _ t ... t .-.

V (:- t_:'_ " 2R ,n/,n

P

.'_.,;;.._','-:_-.:.5:'.-:.

/_- .jplastically deformedhinge area

° e"t"_ t = assumedeffecHve length of plastic h0ng

= angularveloclty of bending(slopeofc_vs, time curve)

t

ApproximateStrain Kate

l l,i J , ii ,

I/2...t_ I

FIGURE9 - Schematic ShowingAnalysisof VarestralntT.estUsedto Calculatethe Approximate Strain Rate.

2O

1972007868-027

_, _ Aslrenuclear::_:, Laboralory

the informationobtained is "specific regarding the abillty of each of thesealloys to accommodate

various levels of strain durlng welding.

The resultsof all prellmlncry testshave already been presentedin Figure 5. All of the data

represented in Figure 5 are for testsmadeon welds produced at 15 ipm. The weld currents1

used in each case were as follows:

Allu___Z.y Weld Current (amps)T-222 210FS-85 180Ta-10W 220SCb-291 180•B-66 180

The total crack length is plotted asa function of augmentedstrain for each of the alloys used

in the screening tests. The hot cracking sensltlvlty reflected the exact order of weldability

established previously, ranging from Ta-10W which did not exhibit hot cracking at even the

mostsevere test conditions to B-66 which contalned hot cracks at every condltlon evaluated.

-_ _ All alloys except Ta-10W displayed a measurabledegree of hot crack sensitivity. Thehot

:'_ I cracking tendency displayed by B-66 correlates well with the observed behavior of this alloy.

i Field welding of T-222 and FS-85 hasshownthese to be sensitive to underbead cracking in:j_ : multipassplate welds but not sensltive to classichot cracking. Hence, this test mayprovide ar_a "

_ indication of senslti_,ity to failure modesother than classichot cr ckmg.3

:__ • In addition to total crack length a useful index of cracking sensitivity is provided by the.- "cracking threshold". This representsthe minimum level of augmentedstrain required to pro-

duce cracking for a g.iven set of weld parameters.For the alloys represented in Figure 5 the

• : ,_ cracking threshold values, for a weld speedof 15 ipm, w_re :,S

_ Cracking ThresholdL, B-66 <I/2%

.. T-222 .,-*I%, , FS-85 ," 1%

- SCb-291 2%_"!':_ Ta*.10W >4%

1972007868-028

Although mosttestswere madeusing a weld speedof 15 ipm the llm_ted n0mberof testsmade

at other welding speedsindicates the metallurglcal factors which influence hot cracking are

somewhattempered by reducing the welding speed. Thls wasseen for the B-66 test results in

Figure4.

/,

During the preliminary testing it was found desirable to maintain a nearly constant weld width

in the specimensof a given alloy. The inadvertent useof too low a weld current resulted in test- '

ing a very narrow weld in one of the B-66 specimens.This is the data point in Figure 4 which is

seen to be exceptionally high relative to similarly tested wider welds. The narrow weld forces

a relatively small total volume of solidifying weld metal to accommodatea considerableamount

of strain. Interaction between the mechanical effects due to changing soecimen/weld geometries

andmetallurgical effects arising from changesin heat input and freezing rate are analytically

very complex. Fortunately, test results to data indicate that small variations in weld width can

be tolerated. Hence, this variable is readily accommodated.

In summary,performanceof thls unit proved to be excellent in all respects. The cracking index

developed accounts well for classic hot cracking which _:ccursat temperaturesin the vicinity

of the effective solidus of the alloy as the weldment cools. In addition, there is evidence the

test may also provide informat;onregarding hot ductility limitations of both the basemetal and0

weld metal. The latter evidence isobtained by noting the occurrence andseverity of heat '

affected zone cracking. HAZ cracking wasobserved in manyof the specimenstested but

detailed analysis of this phenomenonwas beyond the scopeof the presentstudy. Where duplicate

• tests were conducted the variation in test results was lessthan 15%, indicating excellent

reproduclbillty.

3.2 T-111 and ASTAR-811C

Primary emphasisin the testing programwasdirected toward the tantalum-base alloys T-111

and ASTAR-811Cdue to their current importancefor spacepower systemapplTcatlonsrequiring

a combination of excellent weldabillty, ductility, and high temperaturestrength.

.--/" 22

1972007868-029

_ Astr0nuclearLaboratory/

t

All T-111 and ASTA.R-81]C material used in this program was tested as fully recrystalllzed

1/8 inch thick plate. Figures 10 and 11 show the pre-test microstructure of the "1"-1]1 and

ASTAR-811C evaluatedt respectively. Both heats of ASTAR-81 ]C are represented on that

Figure.

)

The test results for T-111 and ASTAR-81 |C are presented in tabular form in Tables 3 and 4_

respectively, and plotted in Figures 12 and 13_ respectively_ as a function of total crack length

in the weld fusion zone vs. augmented strain. The correlation appears to be excellent for all

conditions evaluated. The previously noted effect of weld width was taken into consideration

during the testing of T-| 11 and AbTAR-8| 1C. An effort was made to select weld parameters

which would produce a uniform weld width -approximately 0.165 to 0.180 inches. For the data

of Figures 12 and 13 this necessitated use of the following weld currents vs. welding speed

(_he same weld currents were used for both alloys);

Welding Speed Weld Current

30 ipm 260 amps.15 2107.5 1803 140

It shouldbe notedthat the resultingweldswere somewhatnarrowerthan thoseusedfor the

preliminaryalloy testing;hence_the conditionsof testingwere probablya little moresevere

for T-111 andASTAR-811C. Individual data pointsare notshownon Figures12 and 13. How-

ever_ in order to evaluate test reproducibilityseveralduplicate testswere conductedat the

• mostsevereconditionsevaluated (4% augmentedstrain - 15 ipm weldingspeed).The resultlng

scatterin the data wasfound to be no greater than 10%.

Thedata plotted in Figure 13 for ASTAR-811Crepresentstestson two different healsof this

alloy. Theperformanceof the low rheniumHeat 650056 is seento be considerablybetter than

that of the high rheniumHe,at 650078. At the time of the purchaseof ASTAR-811Cfor this

programthe specification for rheniumwas 1.00 to 1.50 weightpercent. Since that tlme_ the

23

1972007868-030

(_ AstronuclearLaboratory

23,075 100X

-- Avg. Grain Size34.8 lain '

ASTM 6.5

FIGURE10 - Pre-testMicr¢_tructureof T-111.

24

1972007868-031

.f-

TABLE3 - Varestralnt Data for T-111

Weld Conditions Results• d

Specimen Percent Speed Current No. of Max. Crack Lgth_ Total Crack Lgth.Number Strain (ipm) (amps.) Cracks (mils) (mils)

1 4 15 210 17 7 105

2 4 7.5 I 190 17 5 60

3 1 15 210 - - -

4 3 15 210 30 6 100

5 2 15 210 9 5 40

6 3 7.5 180 26 8 70

7 2 7.5 180 - - -

8 4 3 140 21 6 65

9 3 3 140 19 5 50

10 3 15 210 18 14 80

11 4 15 210 30 8 1!5

12 4 15 245 38 18 180

13 4 7.5 180 31 7 110

14 4 30 260 23 48 220imln

26

1972007868-033

(_ Astronuclear i,.: Laboratory '

TABLE4 -,.'VarestralntData for ASTAR-811C

WeldConditions Results, ,,,, .... _, ,,, ,,

Specimen Percent Speed Current No. of Max. CrackLgth. Total CrackLgth.Number Strain. (ipm) (amps.) Cracks (mils) (mils). ,., , ,, ., ,,,,

t

1 4 15 210 39 9 140

. 2 4 7.5 210 29 9 110

,.'. 3 1 15 210 - -

_ 4 3 15 210 30 8 120

! 5 2 15 210 18 4 45,.: 6 4 7.5 180 25 10 1t5 '

_ 7 2 7,5 ]80 6 3 20?,. 8 4 3 140 32 8 90

-f, 9 3 3 140 18 5 55; 10 4 15 210 40 12 160_o

."_[" 11 > 4 15 210 38 32 240

_ 12 4 15 245 50 26 16513 4 7.5 180 28 9 135

14 4 30 260 30 ' 43 195 ,1re' L , ,, ,, , • ,-K.

i 15 4 15 210 6 10 34

.: 16 2 15 210 - - -• 17 1 15 210 4 4 12

18 3 15 210 9 6 40

• 19 3 7.5 180 8 3 16

20 4 7.5 180 10 6 38

_' Specimens 1 through14werefromHeat650078Specimens15through20 werefromHeat650056

27

1972007868-034

.f

1000 I I i ...... I /

900 -- T,,111 ._Data for fusionzone cracksonly.

800 - - .

700 -I

'E

-r. 600 -.

ZuJ 500 --.J

"_ 400 --Welding Speed

U 30 ipm_J

300 - 15 ipm.

7.5 ipm200 - -__ -

3 ipm100 -

0 1 2 3 4

AUGMENTED STRAIN, PERCENT

FIGURE12-Total Crack Lengthvs. AugmentedStrain forT-1ll Alloy Varestraln_TestSpecimens.

28

1972007868-035

f

4

'_AstronuclearLaboratory/

/

lOOO I I I I|

ASTAR-811C_oo-

Data for fusionzone cracksonly.

800--

•_ 700-

'I"I--0 600--Z...J

v 500 -- Heat 650078 30 ipm _U

_, 400- 15ipm_\ --

200 -- Heat 15ipm _ ._\_._._ --

650056 __

7.5 ipm100 --

0 I-" "0 ! 2 3 4

AUGMENTED STRAIN, PERCENT

FIGURE13 - Total Crack Lengthvs. AugmentedStrain forASTAR-811CAI Ioy Varestralnt Test Specimens.

integration of experiencefactorsaccruedin the processingand mechanical testing of ASTAR-

811C hasresultedin the lowering of the rheniumspecificationto therange 0.80 to 1o20weight

percent. Hence,the Heat 650078 testdata representedon Figure 13 shouldbe viewed as being

for "out-of-spec" material andare providedfor comparisonandfor thesake of completeness°

Significantly, the Varestraint test wasable to discriminatebetween thesetwo heatsof ASTAR-

811C despitethe rather minorvariations in composition.

Figure 14 showsthe fusionzonecracking in a T-111 specimenfollowing testingat a 4% aug-

mentedstrainlevel. Thisis nota typical specimenbut rather representsthe worstincidence

of cracking observedin T-111 testedduring 15 ipm GTA welding. Thehot cracks in the fusion

zone are invariably located at the trailing edge of the weld puddleandare seento lie perpend-

icular to the locationof the solld-liquid interface at the instantof straining. Included in Figure

14 is a 400X photographof the teglcn arounda typical fusionzone hot crack. Theslightly

different orientationsof thevarioussuE,grainsaroundthe crack confirm the fact hot cracking

propagatesalong solute-enrichedgrain andsubgrainboundaries.A similarly tested(high rhenium)

ASTAR-811Cspecimenis shownin Figure 15° (A highrheniumspecimenwasusedmerely be-

causeof thescarcity of cracks in the low rheniumspecimens).Again, the dominantsites for

hot cracks are the subgrainboundaries.

J

C.omparisonof Figures12 and 13 indicates that ASTAR-811Cis lessproneto hot cracking than "

is T-11 I. Althoughthe cracking thresholdfor bothT-111 and ASTAR-811Cis in the vicinity

of 1% augmentedstrain_atthe strain levelsgreater than 2% the superiorityof ASTAR-811Cis

evident. This is in agreementwith the general behaviorobservedin multipassGTA plate weld-

ing whereASTAR-811Cconsistentlydisplaysa lessertendencytowardunderbeadgrain boundary

cracking thanT-111.

HAZ cracking wasnot observedin either T-111 or the low rheniumHeat 650056 ASTAR-811C.

The behaviorof the highrhen|umHe_ 650078ASTAR-811Cwaserratic in this respect,however,

displayingsevere HAZ cracking in somespecimensand noneat all in others.Neither the occur-

3O

1972007868-037

FIGURE14,- FusionZone Hot Cracksin T-111 SpecimenTestedat 4%AugmentedStrain. SpecimenLightly PolishedandEtchedPrior to Examination•

Y I

i 31

1972007868-038

15X

1972007868-039

.f ,,

(_ Astr0nuclearLaboratory

rence nor the severity of the HAZ crackingshoweda definite relationship to % augmented

strain duringtesting. The ASTAR-811Cspecimenshownin Figure3 displaysextensiveHAZ

crackingwhile in Figure 15 a single HAZ crack is seen.In addition1to the moretypical cracks

of Figure 3 numerousinstanceswere found which lookedmorelike "pits". In mc_tcasesthese

' appearedto have likely been the result of weld "splatter" but this general conclusiondid not

adequately accountfor all instanceswherethis wasseen=A microprobescanof oneof these

regionsindicated hafniumenrichmentand tungstendei_letionshowingthat there wassomelocal

inhomogeneltyin thismaterial=

To providea final perspectiveall of the test resultson welds producedat 15 ipm are shown

in Figure 16 for all sevenrefractory metal alloys evaluated. Only the test resultsfor the low

rheniumHeat 650056ASTAR-811Care included in this figure t for thosereasonspresented

previously.

The correlation betweentheseresultsand the knownweld variability of thesealloys (see prior

"ranking" llst ) is excellent. TheT-222 data appearsslightly higherthan mightbe anticipated

but it doesfall in the expected order. Thehigherhafniumcontentof T-222 relative to T-111

mightreasonablybe expectedto lowerthe effective solidustemperaturein regionsof micro-

segregation(asin welds)therebyenhancingthe possibiiity of hot cracking.A slmilar argument,

mightbe offered (i.e, higherhafniumcontent) for the observeddifference betweenT-111 and

ASTAR-811C.

w

4.0 CONCLUSIONS

1. TheVarestralnttest conceptof applyingaugmentedstrain to a bead-on-plate GTA weld

hasbeenmodifiedandadaptedto providea meansof evaluating hot crackingsensitivity

in the refractory metalalloys. Augmehtedstrain levels from 1/4% through4% are possible.

2. TheVarestralnt testcapability hasbeendevelopedwlthoutcompromiseof the stringent

environmentalcontrolsrequiredfor highrellabillty welds in refractorymetal alloys.

33

1972007868-040

AUGMENTED STRAIN , PERCENT

"=.'_:_ FIGURE16 " Comparisonof Total Crack Lengthvs. AugmentedStrqln

":_i:i"_ for All Refractory Metal Alloys EvaSuatedThis Program.

:' 34

' 1972007868-041

(_ Astr0nuclearLaboratory

Theseprocedures.shouldbe logically applicable to studieson the reactive metals (Zr_ Hf,

Ti ) and their alloys.

3. Excellent correlation wasachieved betweentest resultsandpreviously observedweld

quality variabi&ity.f

404. Refractorymetal alloys_wlfh the exception of B-66 andpossiblyT.-'_.2, are not proneto

• hot cracking.

5. Hot cracking in refractory metal alloys is related to _4crosegregationin the welds, being

invariably locatedalong fusionzone grainand suI- ,_in boundaries.

6. Theenvironmentalcontrol capability developedshouldbe of partlcular value in evaluating

the effectsof atmosphericcontaminantson the weidabillty of both conventionalanddevel-

opmentaI materlaIs.

7. Within the somewhatlimited frameworkof the presentinvestigationtest reproducibility

appearsexceIlent.

Acknowledgments

Informationreportedin this reportresultedfrom researchsponsoredby the National Aeronautics

and SpaceAdmlnlstratlon-LewlsResearchCenter Contract NAS 3-11827, "Fracture and Hot

Crack Resistanceof Weldsin T-111 andASTAR-811C". The authorswish to expresstheir ap-

reclatlon to NASA Project Managemenb/vtssrs.P. E. MoorheadandR. E. Davies for their great

interestandconfidencein this project. We would also llke to thank Dr. W. F. Savagefor his

interestand helpful discussionsduringthe planningof thisprogram.

Special recognitionis glven Mr. A. R. Keetonwhoreducedour general requirementsfor the

Varestralnttest to practice by designinga first class_functionalandreliable apparatus.

_" 35

1972007868-042

"4"

i 5.0 REFERENCES

1. Savage t W. F._and Lundln, C. D., UThe Yarestralnt Tesb" Welding Journal Research

SupplemenbVol. 44 (10), October, 1965.

2. Stoner_D. R._and Lessmann_G. G.t "Measurementand Control of Weld ChamberAtmospheres," TheWelding Journal ResearchSupplemenbVol. 44 (8), Augusb 1965.

:3. Lessmann,G. G., "The Comparative Weldabillty of Refractory Metal AIIoys_" TheWelding_ Journal ResearchSupplement_Vol. 45 (12), December_1966.

.f

"i, _ 4. Lessmann_G. G._ "Determination of Weldal_ility and Elevated TemperatureStability of• . ._ Refractory Metal Alloys/' Volume I, NASA CR-1607_August, 1970.

,s

5. Lessmann_G. G., "Welding Evaluation of Experimental ColumblumAlloys," The WeldingJournal, Rese,-ch Supplement,Voi. 43 (3), March, 1964

:._

."?-..._

I

i ,• i

".._,_'.,-;

"1g72007868-043