You are on page 1of 48

N]NCLASSI FIED

k

1 Ž ,.

-

fArmed Services Technical Information Agency
ARLINGTON HALL STATION
ARLINGTON 12 VIRGINIA
FOR
MICRO-CARD
CONTROL ONLY

NOTICE: WEN GOVERNMENT OR OTHER DRAWINGS, SPECIFMCATIONS OR OTHER DATA
MMM
FOR ANY PURPOSE OTHER THAN IN CONNECTION WITH A DEFINITELY RELATED
GOVERNMENT PROCURREMENT OPERATION. THE V. S. GOVERNMENT TREREBY INCURS
NO RESPONSMMILrrY, NOR ANY OBLIGATION WHATSOEVER; AND THE FACT THAT THE
GOVERNMENT MAY HAVE FOPMULATIED, FURNISHED, OR IN ANY WAY SUPPLID THE
SAID DRAWINGS. SPECIFICATIONS. OR OTHER DATA IS NOT TO BE REGARDED BY
IMPLICATION OR CrHERWISE AS IN ANY MANNER LICENSING THE HOLDER OR ANY OTHER
PFARSON OR CORPORATION, OM CONEING ANY RIGHTS OR PERMISSION TO MANUFACTURE,
tý2' OR SELL ANY PATENTED BWENTION THAT MAY IN ANY WAY BE RELATED THERCTO

- -

OMIC RUp*e t08
Match 1. 1959

-

ASTIA AD 210464

OTS PB 151063

WELDING AND BRAZING
OF MOLYBDENUM

FL•L

COPY

ASTIA

HG¶O'ON •

~~API

T&SAI1•ON

1

£+uut

1O"O

4

12., VIO*N&1A

DEFENSE METALS INFORMATION CENTER
Batlllo Memorial Institute
Columbus

ASTIA

1, Ohio

t..',j
I

•5S -•. +-- -. +_ +2
• •

•m +m

••

l

,

,

u

m

•-

6X

2

21 .
r

DMIC Report 1(

ASTIA AD 210486

March 1# 19!

OTS PB 151063

*

|I

WELDING AND BRAZING
OF MOLYBDENUM

by

N. E. Weare and R. E. Monroe

OFFICE OF THE DIRECTOR OF DEFENSE
RESEARCH AND ENGINEERING

DEFENSE METALS INFORMATION CENTER
Battelle Memorial ILititute
Columbus 1, Ohio
____

q

£

___
___

____

___

___

____

___

___

TABLE or

I
ISTRODIUC"O.04
A
ME
t

RG

cCf•ETE+S

I

. .

-A

C

ME

RA

IN

.

.

.

.+ .

•.

. .

.

.

. .
,o

..i

.

. . . . ._. . . . . . . ....... ....
. . . ..
M~ro•"•:•4•

.o

.

. .

. .

. .

oo,

. .

......
.o

.

.,

. 1.41
v.
.

.

.. . . .

.,
. ..

.

. . . . . . . . . .I.........
E+ew.*

CLEA.

.....

.

.

.

.

.

. . .

.

.

.

.

... .

.. .

M.ES
........................................

Mi ,.•V+•.

........

.......

l-sJal
44

sol

brazti

.. .

.

........

.

.

.

.

.Z?. .............

.

.

.

.

.

... .

.

.

.

.

.

.

.

.

.

.

.

....................
.

ucPab se.k r...l Co

iderations

.

33

.

4

.

.................................

.AS.......

JOI ur..I oG
.....................
h t5J.SL

.

....

143

.

............
.

. . ...

.

.
.... .....
..........................
. . . . . . .......
m.•..
.....
...
. . ...
.
...
. . .
-VG-APof
olybdenwn . .
.
.
of M.lybd

3.

...

braz•ng ofSoly•,'W=m

.................

. .
. . .

.

.

.........................

40
40
4
41
40

-i
-ra'-oi.yte3gn

*

*4

iii

+

.. I_-

...

-

-.

.i,.-

--

:

--

-l

i+t-~

-

-

,:

-

+:

-

-_

These zones are very sens.WELDING AN. These latter processes do not lend themselves readily to the j of conmplex *tructures.rom the welding atmosphere. ThL* effect is a problem to be kept in mizid particularly h-en fusion welding molybdenum. or carbon. from the surface of the -eldmenpt. At roosr. nitrogen. - 71 . when brazing. pressure welding.perature. (2) The extent of fused or recrystallized zores in molybdenum should be kept to a. (4) Molybdenum is highly notch sensitive. (5) The use of brazing filler metals that alloy with molybdenum to form brittle intermetallics. The most widely studied and most adsptable methods are inert-gas-shielded arc welding and brazing. and the properties of joints made by these gethods may leave much to be desired. nitrogen. or alloys in the mnolybdenumn with low recrystallization temperatures bhould be avoided if molybdenum is to be brazed for structural use at elevato.D BRAZN MOLYBDENUM Or SUMMARY Molybdenum can be welded to itself by any of several methods. the following considera- (1) Minute quantities of oxygen. molybdenum may be brittle.-ratu-es slightly above this molybdenum exhibits sufficienit ductility to be formed. Contamination can occ-ur ýfrom' residuals ini the base metal. tions should be remembered. te. Resistance wdding ultraonic %elding. To obtain the most useful molybdenum joint. but at temp. a.d temperatures. if po**ible. Umw-rnelting etuectics. and similur processes also are used on molybdenum. Recrystallization also greatly lowers the strength of wrought mrolybdenum.tive to embrtitlement by residual oxygen.d stress raisers should be avoided in molybdenum joints. (3) Molybdenum possesses a transition temperature typical of body-centered cubic metals. or f. and carbon greatly lower the ductility of molybdenum.midnimum in fusion welding and recrystallized zones should be avoided.

These include: (1) type of molybdenum. jet-engine. which can be directly related to improvements in the commercially produced metal.one produced by pow ler-metallurgy techniques and another produced by arc canting. there has been a noticeable improvement in the ductility of fusion welds in molybdenum. although only a limited amount of this information has been reported as standardized engineering data.. L. Arc-cast molybdenum has the advantage of higher S~density.wlybdenum-base alloys are available commercially i in two types . and joining procedures. In order to utilize fully the unique properties of molybdenum. This is not true of most powder-metallurgy alloys. methods must be found to join molybdenum to itself and to other heat-resisting materials. . cleaning. During the past few years. (5) physical properties. (4) alloying elements. there has been an emphasis on the application of molybdenum as a structural material for use at high temperatures. and (6) notches. In preparing this report. larger ingot size. Successful fabrication and joining of molybdenum is largely governed by these basic metallurgical factors. Types of Molybdenum Molybdenum and r. it was felt that only data obtained during the last 5 or 6 years are of value. the problem of producing ductile joints in molybdenum has been studied. mflssile. For some years. Fnsion velds in are-cast r •- material can be made without cracks and porosity. although a recent patent has been issued for a weldable powder-metallurgy alloy containing a small percentage of titanium. (12)* The primary reason for thA improved fusion-welding onetatea .INTRODUCTION * In recent years. Vin= at end of tsRepm%. (Z) microstructure. there are certain metallurgical considerations that must be kept in mind. A survey of the published literature reveals considerable information of value to designers and materials and process engineers associated with the defense industry. METALLURGICAL CONSIDERATIONS When joining molybdenum. This report covers most of the published work on joining molybdemm to itself. and lower gas content. : 9 . because of the improvements in purity over that of commercial molybdenum produced prior to this time. as well as testing. and nuclear-reactor fields.. The metallurgical considerations involved in joining molybdenum axe discussed. (3) impurities. Xhis has been brought about by the future material requirements in the aircraft.

are given iA Table I for both types of molybdenum.00020. ($) TABLE 1.5. Impurities have much less effect an wrought molybdenum because of its fibered grain structure. This has been primarily due to improvements in the quality of commercial molybdenum. In provements in arc-casting technique have steadily lowered the amount of residual carbon in molybdenum ingots and still maintained low percentages of oxygen and nitrogen.004 0. characteristics of arc-cast molybdenum its its exceptionally low gas content.4 I :*•. Not only does recrystallization of wrought molybdenum lower its strength. Molybdenum receives its strength through strain hardening. These efforts have shown highpurity molybdenum to be very ductile when compared with commercialpurity molybdenum.-num in the wrought condition.00010.0100. Mic rostructure The microstructure of molybdenum has considerable influence on its strezigth and ductility. and carhvin. such high-purity molybdenum is not now commercilly available and is not discussed in this report.00050. NITROGEN. 0002 0. therefore.00S Commerci. Recrystallization raises the ductile-to-brittle transition temperature because a given amount of impurity element can form a more continuous network of grain-boundary films in the recrystallized structure having large - I n .5D 18) Composition. Working of molybdenum breaks up the grain-boundary films that cause brittle failure through the grain boundaries.0022 0. nitrogen. The approximate amounts of impurity elements. especially oxygen.030 Since around 1953 there has been a scady increase in the ductility of molybdenum fusion welds. weight per cent Oxygen Nitrogen Carbon Type of Molybdenum Commercial powder metallurgy 0.. However. oxygen.18) Pcwder-metallurgy molybdenum can be Joired by solid-state bonding or by brazing without porosity formation provided it Is not heated to temperatures near its melting point.0071 0. since the tolerance for impurities decreases as the grain size increases.00010.(2.0030. AND CARBON ANALYSES OF MOLY BDENUMt 2 . OXYGEN. but it may result in embrittlement by the impurities present in commercial alloys. arc cast 0. Considerable research effort has also gone into determining methods "ofproducing ultra-high-purity molybdenaizn. it is desirable to keep molybd.

"• •. Fusion welds in molybdenum contain a coarse-grain stractuze In the weld metal and a coarse-grained recrystallized structure in the heat-affected zone. 2X. the weld metal and the heat-affected zone in molyb. The very large-grain cast materiaf has an extremely wide transition range and an even higher transition temperature. The prior history of each material should be studied carefully to determine the proper time- temperature cycle for brazing or solid-state welding operations.•m •O•. As can be seen from this figure. The effects of these impurities on the ductility of molybdenum have been determined by studying the ductile-tobrittle transition temperature of cast molybdenum. When joining by brazing or solid-state bonding. oxygen. The reported minimum temperatures for recrystallization in 1 hour for naterials previous Ly-reduced 97 per cent at 2200 to 1900 F are as follows:1 2 ) Unalloyed molybdenum .w1. 5 per cent titanium . For this reason. o. Impurities iOxygen. etemelyn smal erenags o oygn av denum.2450 F Mtolybdenum-1 per cent vanadium. nitrides.S grains. . 21-50 F The maxiumr temperature that can be tolerated without causing any detrimental effects associated with recrystallization may be higher or lower than the above values and should be determined experimentally.2150 F S" tMolybdenum-0. The temperature range is relatively narrow for wrought molybdenum (the lower graphs). nitrogen. or carbides form . Increasing grain size also broadens the temperature ran e ia vIhch completely brittle to completely ductile behavior is obtained. Larger weld-metal grains increase the transition range and transition temperature (middle graphs). (1) The results of this study are shown in Figure Z. . M This is illustrated in Figure 1.2200 F " Molybdenum-0. . . temperatures and time should be held below the required minimum for recrystallization. - agreat ebitngfec A vcatmolyb- Nitrogen and carbon havre much less effect on ductility than does Ernbrittlement occurs when the oxides. The above tabulation shows the alloy containing titanium to have the highest recrystallization temperature and. • •_ . e. 3 per cent niobium . . denurn are less ductile than base material of the same compositiom. Contamination may have played a role in decreasing the ductility and increasing the breadth of the transitiontemperature range of the welded samples. it is receiving much attention at the present time. for this reason.. . and carbon are the most common impurity elementsi that remain in commercial molybdenum.

7/ eqiae gais 1-- I:. -C- .1"A C12[ _ 00046 w/o carbon 0carbo 0. 016 in. fibered groins) FIGURE 1.sees TRANSITION TEMPERATURES AND RANGES OF MOLYBDENUM WITH VARIOUS GRAIN SIZES(17) From bend tests at 0.015at crbon0.C 0 100 Wrought Carbon-Deoxidized Molybdenum Sheet (Fine. 20 -o00 4a. /in.• m lJ 0. A-.2 w/ cor o f " 20- TemperaturwC "CostCafob-Oeoxidized Molybdenum (very coarse equiaxed grains) :•:-::' • so 0 /. 0 100 -100 Temperature.te.o 0.o .• -o -.c ~grinding the surroce face (cure 20 0 -o 1.0"6 80 o w/b caron t~40. /sec strain r.

Hydrogen embrittlement in molybdenum does not occur.grain-boundary film. In the above-mentioned study of cast molybdenum. therefore.O -. For this reason. a Si FIGURE 2. - . Figures 2. Hydrogen contamination is. "•"results S-"' . and (3) strain rate.l 3 .AND . not a problem when joining molybdenum by any of the processes to be discussed. 3 5) Grain-boundary films form upon cooling from the melting point or from high temperatures because of the decreasing solubility of these elements as temperature decreases. the investigators found that ingots melt-i under partial pressure of hydrogen did not pick up additional hydrogen.- * Some relationships between the amounts of oxygen and nitrogen and ductility have been found for molybdenum fusion weldments. molybdenum is used as a barrier against hydrogen diffusion in certain types of apparatus. and 4 indicate that impurities have somewhat less effect on the ductility of fusion weldment than they do on cast molybdenum. CARBON UPON BRITTLE TRANSITION TEMPERATURE OF S-~ •CAST MOLYBDENUM(l) From bend tests at 0.& EFFECTS OF OXYGEN. but the are discussed in relation to impurities in the welding atmosphere. 3. Metallic impurities are present in only minor amounts and apparently have no noticeable effect on the ductility of molybdenum. One study has shown that contamination by tungsten that may occur in tungsten arc welding lowers the ductility of molybdenum welds slightly. (15) _ _ _ . hydrogen through molybdenum is not perceptible at elevated temperatures. (2) grain size. (1) Diffusion o. .I sec strain rate. /in. 038 in. NITROGEN. thus causing failure to occur intergranularly.14) These effects were studied by adding oxygen or nitrogen to an argon welding atmosphere. (11. However. The same results for fusion weldments are discussed in the section on Fusion Welding.11. direct comparisions of the results of these two studies are not possible because of: (1) differences in base materials to which the addirio:s were made. 13.

0 ' I 3 AND THE BRITTLE-FAILURE TEMPERATURE(I ) From bend tests at 0. . -•:- 100 i :i Oign (from figur 2) OX* -- ? 200 -30019 1 - 0.04 0006 0008 0060 0012 0014 0016 Oxygen inWe'd Bead.0203 in.rJ . /sec strain rate.04 0.10 *1.. - A130320 BEND RELATIONSHIP BETWEEN NITROGEN IN THE WELD 14 AIND THE BRITTLE-FAILURE TEMPERATURE( ) From bend tests at 0. /in. .08 CX12 0.16 0.i v wt. 'sec strain rate.. 4r- -± 2002 S0 E I-A -200 0 .18 0 0 per cent Nitrogen inU. /in.weight FIGURE 4.weight per cent OXYGEN IN THE WELD BEAD BETWEEN 1RELATIONSHIP 0 .Wed Bead. 0002 O0.I. -- FIGURE Ch3.0003 in.- .

Weare.wnable -- 00 V elec"rodes Spread of djcttftw ofi Contral in"o 0 _1-0 Titorwurn Content. 0 weight per cent improved the ductility of cast molybdenum. Alloying of molybdenum to overcome the deleterious effects of oxygen on fusion welds offers promise in improving the ductility of welds in molybdenum. (1) The Battelle investigators found that additions of titanium in the range of 0. Titanium presently appears to be the most promising alloying element for this purpose.( 17 ) Weldments containing 0. /in. 5 to 1. * *. 0 At" me - Wm a Alloys meltd in oamn fmn is.7 weight per cent titanium had the 4 Ia 1 - . This is illustrated in Figure 5. EFFECT OF TITANIUM ON TRANSITION TEMPERATURE OF CAST MOLYBDENUM From bend tests at 0. (4) Additions of titanium raised the recrystallization temperature of wrought molybdenum and. Monroe. /sec strain rate. and Martin found a relationship similar to that shown in Figure 5 for welds and alloys prepared from purified molybdenum. in this manner. Research at Climax Molybdenum Company also has shown an improvement in ductility of molybdenumn with small additions of titanium and/or zirconium. These additions have been made with the idea of overcoming the deleterious effects of oxygen on ductility. also reduced the possibility of embrittlement.Alloying Elements Many investigators have concentrated their efforts on determining the effects of small additions of alloying elements on the ductility of molybdenum.038 in. weight per Cent FIGURE 5. Work at Battelle Memorial Institute has centered on studies of additions to cast molybdenum to overcome the effects o1 oxygen.

The coefficient of thcrmal conductivity for molybdenum is very high at room temperature. (15) The effect of carbon on weldment ductility has not been resolved. ( 17 ) In the above-mentioned study by Platte. and could be determined only by a controlled experiment in which several carbon contents were studied and all sheet material was fabricated in an identical manner. as determined by a slow-bend test. (1 7 ) Another study by these investigators indicated significant differences in the ductilities of weldments in different lots of arc-cast molybdenum of about the "samecarbon content. 5 per cent titanium. Monroe. "o' ' :. In unalloyed molybdenum. A cornmercial arc-cast carbon-deoxidized alloy is now available that contains a nomizal 0.termined that aluminum lowers the temperature for ductile behavior in a slow-bend test but caused porosity in tungsten-arc welds. and Martin found no improvement in weldment ductility over unalloyed material in a similar comnaercial alloy containing 0..y. who found increated ductility with decreasing carbon content from commercially fabricated rmolybdenum in heats ranging from 0. and this drops slowly as the temperature is increased to ZI00 F. (21) This is completely opposite the result obtained by Monroe.'-.(20 ) Tungsten-arc welds in this alloy were ductile at room temperahate. "Y" :- Improvemerts in fusion-weldment ductility have been made by some investigators by the use of a 55 weight per cent molybdenum-45 weight per "centrienium filler-metal composition. -' It was also surmised during Platte's study on the basis of the carbon contents in several different heats of material that an increased carbon con- tent from 0. -'- Physical Proprties The thermal conductivity and thermal expansion are important factors to bt considered in brazing molybdenum. w-in•-g rhenium -Aire around the molybdenum wire. Weare. which represents a significant improvement over the 6T radii obtained when molybdenum filler was used.. 015 to 0. 02 to 0.. Platte found the most ductility' in molybdenum weldments in the cornmercial arc-cast alloy containing 0.-(2) The filler wire was made by.to lowest brittle-to-ductile transition temperature. . However.. 029 per cent carbon.46 per cent titanium and 0. et al. the ductility of welds produced by these investigators in unalloyed arc-cast-carbondeoxidized molybdenum that was available in 1954 was comparable with the ductility of welds in the high-purity molybdenum-titanium -alloys. 70 Btu/sq ft/ftl hr/*F. 06 per cent increased weldment ductility.05 per cent "c"trbon. -. Transverse bend radii of 3T were obtained at room temperature.. Mýl-ybdenum-rhenium alloys of the composition given are ductile. 46 per cent titanium and 0.' .- - ." . it is beneficial to have some carbon present to react with residual oxygen in welding atmospheres to produce a deoxidizing effect..-'. 042 per cent carbon. This seemed to indicate some effect of sheetfabrication history on weldment ducu. it also was d.

' . Notches All joints in molybdenum should be designed to avoid notch effects at the joints. (20) The geometry of the specimen used by Platte produced a high stress at this notch. In welding molybdenum. Transverse cracking occurred in weld metal and propagated into the base metal if reduced weld cross-sectional areas occurred. Platte experienced severe crater cracking in many weldments because of the notch remaining as a resuit of shrinkage during solidification. (18) Longitudinal cracks also can occur with cornplete penetration if a notch is left at the start or end of a weld. The thermal expansicn of molybdenum is low compared with that of most metals. Monroe. In this test. (6) These high coefficients of thermal conductivity permit rapid heating and a minimum of time at the brazing temperature. Olds and Fisher demonstrated that molybdenum in the fibered condition is notch sensitive in bending. Failure to design properly for this low value of thermal expansion can lead to trouble when molybdenum is brazed to other metals that may have thermal expansion coefficients three or four times higher. and Martin have showed longitudinal weld cracking to result when a notch remained at the root of a butt weld because of incomplete penetration. PF. should be taken to eliminate craters at the ends of fusicn. Fusion-weld ductility is normally determined by a bend test. the mean thermal expansion coefficient in the temperature range from 3Z to 2200 F being only 3./in.where the value is 56. Detrimental effects from joint notches on assemblies that are not recrystallized during brazing also would be expected. ". Most researchers have used the transverse bend test. (7) The presence of a notch surrounding resistance or ultrasonic spot welds probably contributes to the low strength and the brittleness of these joints "* *- * TESTING OF JOINTS I -" " * Ductility is a property of primary interest when testing joints in molybdenum. (41) Redesign of the joint to minimize the notch effect still did not prevent this type of failure at room temperature. Fracture nearly always occurs in the weld metal. although the greatest stress is at the centerline of the weld. A transverse bend specimen is illustrated in Figure 6. p. steps. Weare. For comparison purposes. . "C Jacobson and Martin found it impossible to avoid tensile failure through recrystallized base metal at low loads in brazed double-shear lap joints. welds. since this indicates the ductility in the least ductile direction for a fusion weld due to weld-metal grain orientation. the greatest amount of deformation generally occurs in the heataffected zone.8 x 10-6 in. thereby minimizing the danger of uneven heating.5 Btu/sq ft/ft/hrl*F.

and reduction in area.. (0) Transition temperatures are increased as strain rate increases. Relatively small changes in strain rate can shift the fracture behavior from ductile to brittle at room temperature. and base metal must undergo the same strain during bending. The constantmoment bend test has been used by some investigators to determine the ductility of upset welds in molybdenum. The three for whi-h the greatest amount of data are presently available are: (1) Determination of the ductile-to-brittle transition 1 . (I1) In testing upset or flash welds. Upset butt welds. . (15) When attempting to compare the results of studies carried out by different investigators. This effect is shown in Figure 7.'ar the ductilities of various welds can be determined in various manners. Resistance and ultrasonic spot welds have been tested by tension-shear and cross-tension tests.. beat-affected zone. All molybdenum weldments should have their surfaces ground to produce a smooth surface for testing. elongation. The removal of surface defects allows better comparisons to be made of different weldrnents. it must be kept in mind that strain rate has an effect on the ductility of molybdenum. the upset or flash must be removed in order to test the weld properly.( 1 7 ) Still others use a transverse guided-bend test. Also shown in Figure A is a tensile or stress-rupture test specimen for fusion vlds in molybdenu. Surface grinding removes any microcracks or reduced cross-sectional areas that would tend to produce brittle failures.( 2 7 ) In this test. Battelle investigators found an improvement in fusion-weldment ductility when the molybdenum weldment surfaces had been ground. sheet. and brazed butt joints have been tension tested to determine strength. investigators have used several methods. -- . since weld. Some investigators measure the total deflection during testing.tudinal bend test (Figure 6).37) This test indicates whether the weld or heat-affected zones were ductile or brittle. (27) For the purpose of comparing weldments made ir different lots or alloys of molybdenum or welds made in different welding atmospheres. primarily because of an increase in the temperature range in which the yield strength increases rapidly (Figure 7). Some investigators use the long-.917) temperature( (Z) Determination of the temperature of ductile failure for the test (an arbitrary value.11I) This measures both elastic and plastic deformation. Other investigators determine the plastic deformatio r by measuring the permanent bend angle after fracture. (36. flash welds. the weldment is made to conform to the radius of the punch. usually the limit of the test equipment) (3) Determination of the room-temperature ductility.

' " .•.'. Transvese. YIELD STRENGTH..FIGURE . / . TEST SPECIMENS FOR FUSION OR UPSET WELDS IN MOLYBDENUM Sir"V or-~ oU s . AND BRITTLEFRACTURi. STRENGTH OF ANNEALED MOLYBDENUM Grain size 900 grains sq rnm.)F TEMPERATURE AND STRAIN RATE ON DUCTILITY.- Yom V Tmowieroture.l a-mis. ¢ %...'" . (I 9 ) . _--- .. b"n test consfant-smomitM single powt boded bond 109 ~ ': "" ' -l•" Lonitudinol berid test single point loaded ~~0250w -..rupture saple FIGURE 6.-. * 7.. EFFECT . -. 19°m~ tf44 Tension and Stress.V I•----•• ---------II• 1- L- Load Load .."C *D 2'V..

are acceptable methods of oxide removal.( CLEANI. double-lap shear specimens and slotted sp-cimens have been used to study the shear strength of brazed ýoints. but are not always satisfactory for c. the determination of the room-temperature shear strengths has not been possible. The most popular method employed in cleaning molybdenum for joining involveE the use of chemical etchants. ( ) The first bath requires close control because the molten s~alt attacks the molybdenum. Mechanical means of removing oxide. Because "ofthe notch effect present in the specimens for shear testing. ) V PIVW t-% ~ ~ ¶ ¶i 7 - ~ ~ ~ V ~K - -. such as sandblasting.um was severely attacked at the grain boundaries. F . the :olyLder. Two fused-salt compositions used are: (1) 70 per cent NaOH and 30 per cent NaNOZ operating at 500 to 700 F(0 0 ) (2) Commercial Martempering salt. There are several methods that will clean molybdenum satisfactorily for joining. Many of these etchants attack the molybdenum. Electrolytic etchants may be used to remove surface oxides..G PROCEDURES C The removal of surface films from molybdenum before joining by any method is maandatory.~ 14 "Sincethe shear strength of lap-brazed joints is important. but can be used only on simple parts. Saltbath cleaning should be followed by another treatment to remove light oxide films before joining. especially if complex shapes are involved. successful short-time tests have been carried out at 1800 F. Oxide removal by chemical means is a more satisfactory method of cleaning molybdenum. consisting of a mixture of sodium and potassium nitrites and nitrates operating 17 at 700 F.- '-~. but if the time of immrersion in the solution is controlled. when possible. Three cleaning methods used successfully are: 4 . ( 4 1 ) Tensile tests of brazed butt joints also have been performed. / The removal of heavy oxides from molybdenum is most easily accomplished by immersion in a molten salt bath. or abrasion. all can be used satisfactorily. In all cases studied. vapor blasting. Cleaning should be done immediately before joining. How4 1 ) ever. (4-) These test specimens are shown in Figure 8.earing complex shapes. No gross attack was noted when using the second bath.:' h 1.

- -F 7 .4 C 4 4 2r C-7 3SUX N)~lAO vaoa~3 72-7.W7 - ef ý .MR1TX z T -- a V MaI '-. FT7 - -.

5 per cent HF. operating at S150 to 180 F.5 per cent titanium alloy for welding:( 2 7 ) A (a) Degrease 10 minuteds in tricblor.r r= = :• : :•= = 16 -' i r (1) Immersion In a glass-cleaning etch composed of 95 per cent H SOO 4. . 5. although precleaning is recommended. 8 g/ CrZOj.17. (g) Electropolish with 8. 70 per cent HZO. (d) Buff and vapor blast. 15 per cent HCI.-•.to 10-min•te .to IZ-arn-pere current in a bath of 80 per cent H2 SO 4 at 130 F. . (43) This is of i=Portance for brazing operations. . (i) Wrap each piece in a clean paper toweL. plus 6 to 10 weight per cent per unit voiane chromic acid. -• •w . (b) Immersion for 5 to 10 minutes in-a bath to remove the smut resulting from t4e first treatment.. (3) A nine-step procedure to prepare rnolybdenum-0. by weight. i =T•.25T: •V • 'r.oethylenvapor degreaser.=mersion time.lkaline cleaner (sam-e as b)..5 per cent HNO 3 .CW WX ..-- •-a . . 18 ) (2) A recently suggested two-step treatmene( 10 ) (a) Immersion in an alkaline bath ceasisting of of 10 per cent NaOH. (h) Rinse in cold water. (b) Immerse in alkaline cleaner for 2 to 3 minutes (commercial cleaner). The second bath consists of 15 per cent HZSO 4 . and 85 per cent HZO..0 5 .J•2• • ... V (f) Rinse. • • • . 5 per cer KCMnO 4 .e or fingerprints can be removed by scrubbing under a solvent such as acetouc or by vapor degreasing.. operating at 120 to 140 F. Molybdenum dioxide can be reduced by hydrogen with a dew point of 80 F when heated above 600 F. • S. G-ea. 0. _ This etchant has been used with a great deal of success for cleanig molybdenum for welding and braxing. . e : 770=a -.S.Wr. •+ . (c) Rinse in cold water. . and 18. (e) Immerse in .

g sections.or more oxygen were present in the welding atmosphere.of additions of oxygen and/or nitrogen to the weldirg a-. and the results of tese studies were dependent upon the degree of weld restraint provided by the bead-on-plate specimr~ens used. Fus ion Welding F=. discussions of this process are not included in this report. submerged-arc welding.. He also established a relationship between oxygen in the atmosphere and the occurrence of centerbead cracking in fusion welds.-e----• Studies have made using the inert-gas-shielded tungsten-arc and cons•-ale-electrode processes.c. the arc is maintained between the -Arkp-ece -A a molybdenun electrode.rt. "--'4' These processes will be discussed briefly. Nitrogen in the welding atmosphere was fotun to have less effect on weldment ductility.& welded. Inert-gas-sh•ielded welding of molybdenum can be carried out with a nonconsumable or a consumable electrode. t .. four general classifications . The joining of molybdenum by these methods is discussed in the folloinr. atornic hydrogen. Of these processes. Several organizatinns have fabricated molybdenumi assembl:es by using the tungsten-arc process.- --. and brazing. Platte deter-mined the effect. aas not been used extensively. ( 3 1 4 ) The results of these studies are shown in Figures 9 and 10. the provided between a tungsten electrode welding process be!=.. the tungs:en-ar. ? per ce-.e most widely used methods of joini"g molyb.-ogen welding of mrolybdenu-. For this reasc-. cracking wo"ld occur in welds in l'16-inch-thick unalloyed arc-cast molybdenum. Heat for fusion for the nonconsumnableandpool. . and addit:ocs of -"4. The molybdenum electrode is consumed and filer metal is proelled through the arc to the weld joint. divided "- These methods can be . the ductility of molybdenum weldmentse . Brief "investg-tgaons have been made to study submerged-arc welding of rnolybdenumn.- When a consu•nable electrode is employed. L I . (13) He found that if 0. and presently molybdenum is being welded using an electron beam. and.ic welding is one of t.fusion welding.!ybdenun can be joined by several methods. the electron-beam processes.rnosphere on the ben-d-test ductile-brittle transition temperature. Fillerismet~al S~metal is added directly to the electrode moltenavId .g might not have occurred in alloyed molybdenum.ng that the rnolybdenumn has been sufficiently cleaned prior to welding.e-ic contarninarts or. process has been used most extensively and will be discussed in the greatest detail. resistance weld- ing. atomic--y-r. L - _________ _ 7. This crack-n. "*" " -- Ass--n-.major source of contamination is the welding atmosphere. Cousiderable research has been cornidcted to determine quantitatively the effects of atmosp.17 JOItLNG PROCESSES . Since the development of the inert-g-s-shielded arc-welding process. soliZd-jtate welding. recently.

. . per cent FIGURE 9. /in. - ..3 2 EFFECT OF OXYGEN LNq WELDING ATMOSPHERE ON BRITTLE TRANSITION TEMPERATURE OF WELDS(13) • . /sec strain rate. 0 025 0.t Q 4. _ a'L' 0 005 010 0. 0003 in.15 _ Oxygen in Welding Atmosphere. bend tests at 0.20 A-a~30. From . E_0s 100' cg I"=' 4i c 100 ""2.

. 0..-3 o II -- -- . . EFFECT OF NITROGEN IN THE WELDING ATMOSPHERE ON "THE TEMPERATURE FOR BRITTLE . /sec strain rate..( ) 0.. 0003 in.5 0-3 0.020 "Nairogen o mosplhre.'S 0 •.062-INCH-THICK MOLYBDENUM 1 4 From bend test at 0. /in. z.(bond• og%* oscods 1189 •'" So " 2 400 0 17 Si. 0- Brittle faoilure ci-: 1ZLI 2100 S 0 5 25 20 15 10 30 Nitrogen in Welding Atmosphere .. percent 75 NZ 10 a NZ A -0-06 o N2 5 *0. oo---0009-- 2 I. per cent A-30324 OXYGEN ACQUIRED DURING WELDING AS A FUNCTION OF OXYGEN AND NITROGEN IN THE WELDING ATMOSPHERE01 4 ) . _..4 0.012 . _ . CD •... ..5 - • - _ 0004 %C O"YLN FIGURE 11.-ND DUCTILE FAILURE FOR FUSION WELDS IN 0..2 00 of Oxygen inWetding Atmosphere.per cent FIGURE 10.

995 per cent and helium is supplied with a purity excecding this value. argon is supplied in tank form with a guaranteed minimum purity of 99. the addition of nitrogen to the welding atmosphere to lessen the pickup of oxygen does not improve the weldment ductility over that of welds made in pure inert atmosphere.( 16 ) Purification of the gas. Here. once a dry box has been filled with the shielding gas. pr-obably is beneficial. rather than the 0. nor was ductility so consistent. The system used by Platte is shown in Figure 12. (15) They found the ductility of welds to be decreased when the dry box was evacuated to 100 microns of mercury. 1 micron normally used in their studies. Weare. and purifying trains may not lower the amount of impurities present sufficiently to justify their use. They also made several welds outside the dry box using various shielding devices. Monroe. There are several methods of drying and purifying gases for welding molybdenum.8 and 42 degrees were found when the chamber was purged to 100 and 0. However.( 1 3) However.20 up to 50 per cent nitrogen to the welding atmosphere did not cause hot cracking. These welds were not so ductile.( 1 4 ) This same study also indicated an interaction between oxygen and nitrogen in the welding atmosphere that reduced the pickup of oxygen to sorm degree. and Martin studied the effect of ultimate vacuum obtained in vacuum purging a dry box prior to welding molybdenum. Charcoal purifier at liquid N2temperaturit 0 •N ond hydrcarbon removal WELDING . Average room-temperature bend angles of Z. 1 micron of mercury. This is illustrated in Figure 11.ATMOSPHERE 20 PURIFICATION SYSTEM( ) *| •. shielding gases are available in high purities. ( 1 7 ) Either argon D ion of flow 02 ond N2 removol Zr chips of acetonea• Fromatank o ue00 d o ananddyrb -Chrcoa lSowenough to hiquify Exhaust FIGURE 12.-.•F - A-30325 . as for welds made in the dry box. respectively. Some investigators believe that the shielding gas used for welding molybdenum must be purified prior to use. since it will remove residual contaminants and absorbed and adsorbed gases released when welding. This can be accomplished by welding on titanium or "zirconiumfor several minutes before welding molybdenum.

ecte shielding is provided. the porosirTy is probably due to the carbon-oxygen reaction.bhdenum available at the time of their study. "usedas close to the weld as possible..% • K % Various studies hare :rdicated that molybdenum welds should be made in a dry box to insure the best shielding. (15) Since this porosity was not found "i bead -on-plate welds made with adequate shielding. which may contaminate the welds. Monroe.~~~ 47 .!ybdenum weldments is no sign they have not become embrittled by conaminamts.(3) Weare. Surface o. . Sufficient purging time also must be provided for each shield to remove entrapped air. However. . It has been shown that welds with some ductili.in conjunction with a shieldLzg crp and gas backup are shown in Figure 13. Udin. and Martin( 1 5 ) fo. if adequate shielding is used. and proper shielding-gas flows should be maintained. These shields should be . ductile welds can be made . (15) Shields must provide shielding for all areas heated above a temperature at which oxitdes form on the unwe2ded. molybdenum. (15) Some persons believe that thoriated-tungsten electrodes should not be used because of their oxide content. (15. and Wulft reported some duct-1ay of fusion welds made in high-purity molybdenum (0.deoxidized molybdenum. Johnston.or helium may be used t sAield molybdenum during welding..01 per cent carbc. Monroe.p17918 The only method of determining whether shielding is adequate for welding c-olybdenum is to make sample welds for testing. In carbon. whereas in powder-metall:rgy molybdenum the porosity could be due to the vapor pressure of 111oO 3 at the melting point of molybdenum. Both vacuum-purged and flow-purged dry boxes have been used successfully. welds can be made outside dry boxes. Helium Is preferred by many beca=se of the greater heat available for fusing the molybdenum. .xda:cion of a completed molybdenum weld does not appear to have an effect o.. it was surmised thar the porosity was due to contaminants in sheet laminations that had been opened ap during edge preparation.. ( 5) "-• Shigh . 0. This is almost mandatory when welds must be made nan-afly. high-purity molybdenum.nd no effect of the fusion-line porosity on weldment ductility in arc-cast carbon-deoxidized molybdenum. and Martin along the fusion line of many tungsten-arc b=• welds in arc-cast molybdenum.. "Tungsten contaminatEio• during tungsten-arc welding Will lower the ductility of the welds. Several types of shield used. (8) Thoriated-tungsten electrodes are generally .. providing ad. Porosity was noted by Weare. Lack of discoloration of -.001 per cent oxygen) prepared by the hydrogen reduction of molybdenum pen-achloride when only brittle welds could be pro3 duced in comme rcial rwoy.-" corld be made using either standard shielding cups or leadirg-trailing shields La conjunction with a gas backup. the presence of porosity in molybdenum welds is a sign of the presence of oxygen and should be avoided if possible.019) All gas lines and shield parts should be dried before welding to prevent moisture pickup by the sh•-eld•ing gas. It has been known for some time that.the ductility of the weld. However..

5 ! ~ina .LN3WNU3AO L ~nO U) . I U) 49: IXI 3SN~dX3.0 J I ~~ii r. I.\.

*~the N Another possible method of improving weldment ductility would be the reduction of the weld-metal grain size. (20) However. (3) Many of these postweld heat treatments. . with the exception of stress relieving. This treatment would have to be carried out at an elevated ternperature.I 23 * preferred. however.pneroidal 3xide inclusions. This also improves the bend ductility. Battelle investigators( 5 ) studied the effect on grain size of small additions of oxides to cast mo/. Platte attempted to improve weldments embrittled by nitrogen by diffusing the nitrogen from the weld into the base plate. (24) This -method consists of cold working.'e only a slight amount and.ybdenum. possibly 1800 to 2000 F. he found that the slight improvement in ductility was not justified by the expense. ( 2 0 ) Perry. but indicated the presence of boundary precipitates in material treated in this manner. Monroe and ?v1atin(22) and Boas(2 3 ) mention the usefulness of a 400 F preheat to prevent cracking in the welding of complex structures of molybdenum. and Wulff found that annealing low-carbon molybdenum in ' Jeratu-e range 1950 to 2100 C (3542 to 3812 F) led to the formatiox -1 . (25) The tungsten carbide segregated at the root of the weld. Possible improvement of weld-metal ductility and strength would occur if welds "were roll planished after welding. (21) Increasing the speed decreased the weld-metal grain "r'. where it decreased weld grain size. because of ease of arc starting. Platte attempted to decrease the grain . but the'effect on ductility was not determined. Castings containing some oxides did have smaller grains. Several persons have attempted to improve the ductility of molybdenum welds by some p6stweld treatment. A stress relief of 1800 F for I hour in hydrogen has been used. indicated that it might S* be possible to ir. Platte. less tendency for electrode sticking to the work. (21) These studies indicated that polygonization was not successful in improving weldment ductility. The effect on ductility is shown in . in a study of precipitates in molyiadenum weld metal. Cold working of molybdenum is the only method of strengthening present molybdenum-base alloys.ze by increasing the welding speed. ductile material.mprove the properties by a postweld heat treatment at 1200 C.strength. Special equipment is available for this postweld treatment. but the effect of this treatment on the bend ductility of simple butt welds used for determining weld ductility has not been Stress Sdetermined.. He also attempted to polygonize the coarse weld metal by stressing the weld at elevated temperature. Roll planishing. involve recrystallization of the base metal a•rid thus a distinct lowering of the strength. retraction of the weld from the plate occurred. the weld metal by rolling. A similar method was attempted on a tungsten"arc weld in which tungsten carbide powder was added to the weld metal. which resulted in a low. and the extended current range in which they can operate without melting. Spacil. could not be "usedon complex structures. relief did lessen the degree of crater cracking expg:rienced by Platte. and it is felt that such a stress relief is beneficial when complex structures are welded. at the highest speeds studied. . however.

/sec strain rate.!- I 0.0325 EFFECT OF TRAVEL SPEED ON BEND DUCTILITY OF . -*00 - Jo 100 20 - 00 "_ .jt o WELDS IN MOLYBDENUM(ZJ) From bend tests at 0.r * 24 0. F FIGURE 14. 0003 in. /n. per min Iff V co 0Z " •I I 9ninpri in.4 a k0.per min or Test Temperature.3 U I 48 in. A.

definite conclusions could not be drawn regarding the effect of fabrication procedure on the ductility of molybdenum fusion weldments.7 34. • .7 57. -. how- the temperature for ductile behavior was greatly lowered by increased weld speed.. . - .=.1 29. Atmosphere AVERAGE BEND ANGLE AND FACTORS FOR DETERMINIMNG SIGNIFICANT DIFFERENCES FOR HEATS TESTED IN 15 CONSISTENCY RU1U( ) 75 C(a) Z5 C(a) Avg Bend Avg Bend Rank He. Because of the high melting point and good thermal conductivity of molybdenum. = • .5 38. . . However.T z$ Figure 14. 1 pet cent) 17..2 40. Cry ..4 deg. Z 57.4 Z8. This sheet had a coarse grain structure and was not cross rolled.ent ductiLty did exist in welds in different heats of arc-cast molybdenum.6 55. L4t Average(b) Avg Bend Heat Angle.-• '. bil:ty level of risk.3 53.7 44.0 47.1 43. It was found that significant differences in weldn."r"!Y. as for welds made in a high-purity dry-box atmosphere. deg Heat Angle.5 Air.8 55.0 8. e\ -r. Some of the least ductile welds were also made in the least ductile sheet.0 67. . Boam noted that weld cracking occasionally could be traced to poor yield from the ingot.6 (a) Significant diffeence (probalbiity level of bihly level of risk.7 17. 5 per crAt) 13..7 51.8 31.0 C21 C23 CZZ CZ5 CZ6 CZ4 52. • •-• • :'-• . ..(I 5 ) The data obtained from a statistical analysis of this study a:. additional chilling of the weld probably has little effect on. These data show that welds made outside the dry box using a standard shielding cup and gas backing were not so ductile.r r.6 39.8 17.6 CU5 C24 36.• SW?=U .c:ven in Table 2. .per cent) 12. deg Dry box 1 2 3 4 5 6 CZ3 Czz CZI C24 C25 C26 64. Monroe.1 66. helium shield I Z 3 4 5 6 CZ1 CZ3 CZZ C. and Martin studied the consistency in weldment ductility of welds made in different lots of molybdenum containing nearly the same amount of 'carbon (0. 3 19.6 deg. Hi•ghy significant differcece (ptoba- (b) Significant diffesence (pobabihty level of risk.2 CZ1 CZ3 C2z C25 C26 C24 67. nucleation and grain growth.5 74.4 C23 Czz C21 CZ4 C25 C26 69.-•. deg H-ighly significant difference (ptoba- r. Weare. The most ductile welds were made in sheet from rolled sheet bar.0 60. .w.t Angle.021 per cent carbon). . '9r•.6 CZZ C23 C25 C21 C24 C26 74. (23) TABLE Z. --.0 37. rr . The temperature for brittle behavior was slightly affected.. 5 pet cent) 9.) 58.4 deg. whereas sheet in which welds were less ductile was made from forged sheet bar.8 44. without complete knowledge of the fabrication procedures.• A .3 deg.0 68.015 to 0.u..'r.9 66. 5 21. nor were their ductilities so consiistent.

Data obtained by different investigators are difficult to compare because of differences in testing procedures. no worthwhile comparisons could be made between consumable-electrode welds and tungsten-arc welds. Molybdenum probably reacted with oxides of the flux to produce molybdenum oxide. (18) This process would eliminate tungsten contamination. fusion welding is carried out in a high vacuum. Some tensile data obtained by Kearns for 0. but it is probable that the welds were extremely brittle. Ccnsiderable alloying was noted. Most investigators used a bend test to study the ductility of weldments. Inert-gas-shielded consumable-electrode welding may be useful in joining heavier sections of molybdenum. there are few data available of any engineering significance. Because good arc stability and metal transfer were not obtained. 061-inch-thick molybdenum.5 per cent titanium alloy tungsten-arc welds are *ivenin Figure 15(27) and stress-rupture data are given in Figure 16. No weldment properties were reported. This should result in high-purity weld metal. Best results were obtained by welding in helium with direct-current stra:ght polarity (electrode negative) with an ernissive coating on the electrode. Poor arc and metal transfeT were problems that required. For this process. and it was felt that a finer weld-metal grain size might result from the more rapid energy input available with this process. ¶19_ _ _ . Some investigators have determined the tensile strength of molybdenum fusion welds. notably strain rate. Some of the results obtained by these investigators are given in Table 3. ( 2 3) The failure occurred in the heat-affected zone. (28) The welding was done in a nitrogen atmosphere preheated to 400 to 1000 F. solution before any evaluation of the process could be made. and Martin studied the inert-gas-shieldel consumableelectrode welding of molybdenum. Fusion is accomplished by bombardment of the jcint by an electron beam. A good spray-type metal transfer was never obtained. residual carbon would be the best deoxidant.126) Sstability Weare. A more recent process that has promise for joining molybdenum is the electron-beam welding process. Monroe. This study pointed out that notches or reduced cross-sectional areas will cause cracks to propagate because of the thermal stresses induced by welding.000 psi for a weld in wrought molybdenum possessing a 98. (2 9 ) In this process.0. Goodman studied submerged-arc welding of molybdenum using commercial steel welding fluxes as the basis for additions of metals or oxides. Boam reported a weld tensile strength of 53.000-psi tensile strength. since its reaction product with oxygen is gaseous.26 Although many persons have investigated and used the tungsten-arc process to weld molybdenum. It was found that coating the arc cathode with an emissive compound improved the arc stability.

4C. n--n Vo~a~3 .4 -- 3SU3I3*iA.zI aa PV P .

-. see I t sheet 1 welded spr•og•tsh treno t +40 speien sp 4%welded ]-++" 00.25 0 - +%.v - .0.5S PER CENT TrTITANIUM ALLOY SHEET( TepeaurF 33Z . - I. •'..' A ~.20 wedod 3 ecimens0 -% 30--- -A-30327_ 10 1600 1700 1800 1900 2000 2100 2.. •"0.oe•.200 0 ~Temperature -F -/.062-INCH-THICK A-OCP• TENSILE PROPERTIES OF WELDED AND UNWELDED MOLYBDENUM . !__ I.l m te -r i ""60 wrought L-- 50 -P (o.bose alloy) 45 bO '+wrugt Usiaesrn . . - - - - . FIGURE 15.U N. m••:. . • --.e.

100f - j10 2 100.~.F. __25i t Igo E".20 SesR~.- ttnw- 0 . _ 6 -IO(It m 134dsg F461~ T .Z9 Temporoture.000 40- 440 ~30 00 Fully maystoaiszod ber -.A_ . uim~ .Daid ai I Tim S.~hu To" _ El 60 94 lam 4..-.) 8 --10) _ w" X1 4s.

One important factor to keep in mind when flash welding is that the platen acceleration during welding should be as high as possible and the upset should be held in the criti. They showed fairly good bend ductility. and the notch found in all lap welds rules out spot or seam welding as a method of joining molybdenum structures where strength and some ductility are required.. (30) All joints made by these methods must be considered undesirable for structural uses.nce welding is of little interest in producing structures of molybdenum. Flash welds in arc-cast molybdenum made in air were fo--d to be more ductile than welds made in an atmosphere of argon or hydrogen. . ••-. which depends upon the size and shape of the parts being welded. 10 to 0. "Flashwelding probably would have ILnited use in joining molybdenum . ." . (30. This was attributed by "Nippes and Chang to decarburization of the eld zone by reaction with oxygen. si. Spot-welded joints in molybdenum are brittle. . . V .. and carbide precipitation in the heat-affected zone. (30) Nippes and Chang( 3 2 3 3 ) report some success in flash welding molybdenum rods.-. In these applications.. although the latter has been alleviated by placing lead foil between the molybdenum and the copper electrodes. • •. but bend tests were carried out at extremely low strain rates. % "components.- ' • . .30 Resistance WeldigSpot welding has been used for several years to join molybdenum for electronic use. Current requirements were two to three times those required for mold steel. Z1 inch for the rods flash welded in this study. Flash welding has shown some promise 'n joining rods or bars of molybden~rn. Weldment ductility was decreased by entrapped oxides at the interface... Several investigators have made upset-butt welds using resistancewelding equipment. but. or hydrogen. . argon. .ce these represent solid-state bonds. . (35) Seve-ra.al range. "Molybdenum is difficult to resistance weld because of its high melting point and thermal conductivity. . It was found that welds possess"ingsome ductility could be made in air. -. "excessive upset... or some brazing alloy between the sheets of molybdenum being welded. .elds in molybdenum are brittle and generally weak. . . Spot . tantalum. Some results of their work are shown in Figure 18.. •• • -•. .35) Resistance spot welding also is improved if projections are made on the tw sheets being joined. 31. the =echanical propertieb of the joints are of little importance.investigators have joined molybdenum by placing nickel. " V ". Spot welding will be discussed briefly and the results of flash-welding studies will be given in this section.. and for this reason resista. This range was 0. Tensile-shear strengths of molybde=-nn spot welds made by a condenser-discharge welder are given in Figure 17..-•-. Extreme deformation of electrodes and sticking of the electrodes to molybdenum also are problems in spot welding molybdenum. they will be discussed in a later section.

.000 Energy input. ~~-.04 shear strength. 0002 4" %Weld 0.%.:]. 31 * // * 12.08 Sheet Thickness .K.10 A-3O32I CONDITIONS USED FOR AND RESULTS OBTAINED IN SPOT WELDING MOLYBDENUM SHEET(3 0 ) - *-.seconds 8000 6000 .inch FIGURE 17.000 2000- 0000 0 -0. pounds 06 0. 0. = . .'"f .rW. - 4000 ~pounds I092 - Welding apocity. watt.4-Welding force.000 10.

.~25- 15- to -I t- 5 range 0 FIGURE 18. 33) . inch 030 035 A-30330 EFFECT OF UPS7:TTING CONDITIONS ON BEND ANGLE OF ARC-CAST-MOLYBDENUM FLASH WELDS (ALL WELDS 32 MADE IN AIR)( .n/S2) 3 0060 4 000 5 0120 w0 50 I'•" 0.35 30 ! .Alsed by oxides ond poromty -. II I/ a! Optimum-• Q05 010 015 020 025 Upset Per Specimen.150 0200 0 0260 1 severe f 1 wrg of NOmw4 B~.tneness -~ 45 ON40 * S-• 6 7 I Brtleness coised by 4o.3Z I 65 Group PA (.

33 Solid-State Welding The brittleness of molybdenum fusion welds has caused some investigators to study solid-state welding.-s as low as 1274 F (690 C) but recommended using the highest "temperature that can be used without recrystallizing the molybdenum. and low strength were reported for pressure welds made in a vacuum.57.000 psi Ultimate tensile strength Elongation - - 77. helium. The best welds were made in dry hydrogen. When intermetallics b-ittle joints resulted. Fair ductility at room temperature. nickel.. carbon tetrachloride. . chlorine. (3) Fast.zyact -pressure welding of commercial molybdenum in hydrogen and in a vacumn. austenitic stainless steel. This method would eliminate the coarse grain structure and might prevent precipitation of oxides or carbides in the . Moss studied the u. Insert materials studied included: platinum-13 per cent rhodium. These investigators found that the success of the process was based on control of three key metallurgical variables: (1) Peak temperatures had to be carefully regulated to allow forging of the embrittled zor e (recrystallized metal). or water. columbium.000 psi 14 per cent "Reductionof area . because* testing conditions and base-material properties were not reported. and chromium plate.11 per cent "Johnston. -' * * Sformed. The data are of little value. and Wulff carried out upset welding of molybdenum in hydrogen. (35) Best results were obtained by welding under carbon tetrachloride or water. tantalum. (2) Time at temperature had to be minimized. The use of a softer metal insert in the interface decreased the temperature and deformation required to produce a bond. Some welds were produced having ductilities approaching that of the parent metal. Udin. powdered iron. at temperatur. extensive upset was required to insure adequate forging of the embrittled zone and render beneficial deformation. but without excessive fusion at the interface. Spressure grain boundaries. Lithium or sodium in the joint offered some improvement when welding under carbon tetrachloride.(34}) He found that welds could be mad. Roomtemperature mechanical-property data reported by Moss for rod-type welds made in a vacuum are: Bend Test Brealring angle - 21 degrees Tensile Test Yield point .

Present emphasis is on the use of molybdenum for structural parts that will operate at high temperatures. Not all alloys appear to be equally satisfactory under slightly differing conditions. Discussion of these processes will dealwith factors affecting the selection of the best process for a given application. hydrogen with a dew point of 80 F (27 C) will reduce molybdenum dioxide. the ultrasonic %'eldiuag of molybdenum has been studied. Satisfactory brazed joints in molybdenum have been made using all common brazing processes.37) found that ductile upset welds could be made In purified molybdenum. When using hydrogen atmnospheres. Disadvantages are limited temperature range and relatively low heating . Brazing has been studied for joining molybdenum to itself and to join oxidation-resistant coatings to molybdenum. Before any brazing operation can be carried out. et al. For this reason. but did not require high strengths. provided adequate shielding was provided during welding. Some of the high-temperature filler metals are li:ted in Table 4. using one of the procedures mentioned in the section on cleaning. only brazing alloys with melting points exceeding 1800 F will be discussed here. Recently. Conflicting evaluations regarding the suitability of various filler metals for brazing molybdenum in a vacuum or hydrogen atmosphere are not uncommon. (36. the dew-point requirement is -100 F (-73 C) at 2200 F (1205 C) to prevent formation of titanium dioxide. when brazing molybdenum-titanium alloys. the molybdenum surfaces must be cleaned. Furnace brazing has the advantage of being adaptable to large and complex parts. Monroe( 4 0 ) made a survey of the brazing of molybdenum. 3 9 o Ultrasonic spot welds in molybdenum possess low tensile shear strength and no cross-tension strength. Molybdenum has been brazed to itself and other metals using various filler metals. Metallurgical considerations regarding" brazing of molybdenum were given in a previous section. only moderately low dew points are required for unalloyed molybdenum.( 4 3) above 600 F (315 C). One group of investigators reports cracking of the molybdenum during welding at room temperature to be a serious problem. (38) Brazing *. However. ( 38 . According to Chang. The best results were achieved by welding in a high vacuum. rhese joints involved small parts that must be leak tight.34 Kearns. Furnace brazing of molybdenum assemblies can be carried out in inert or hydrogen atmospheres or in a vacuum. The electronics industry has brazed molybdenum to itself and other metals for a number of years. He found 9he work done by Jacobson( 4 1) and Dike( 4 2 ) to be particularly valuable sources of materida on brazing of molybdenum.

17Mo. 6 Nickel-titanium Hastelloy C Monel Inconel 78Ni.aluminum Palladium-nickel. 80Ni.14Cr-6Fe J 1950 2350 2380 2460 2540 Iron Base 18-8 stainless steel 25-20 stainless steel l8Cr-8Ni-Ba1." %' • * "Molybdenum Base Molybdenum-boride Molybdenum.reuthenium. Platinum 100 Pt 3225 -• "- Nickel Base Wall Colmonoy No.3OFe 50Pd-'IFe 2400 2400 Palladium.16Ti 57Ni. per centF Pure Mrtals Llqwidus.3S TABLE 4. • . Mo-B eutectic Mo-Ru S~~-p 34S0 3450 . " Copper Nickel 100 Cu 100 Ni 1980 2650 I Palladium 100 Pd 2860 . ~ $ .35Cu 70Pd-3OCu 2200 2300 2400 Palladium-silver Ditto 5OPd-5OAg 60Pd-40Ag 2400 2500 Palladium-iron Ditto 70Pd.Fe 2600 2650 Cobalt Base Haynes Alloy 21 Haynes Alloy 25 54Co-27Cr-6Mo-3Ni 55Co-ZOCr-I5W-IONI 25S0 2600 Columbium Base *Colurmbiwn-nickel S2Nb-48Ni 2175 Palladium Base 93Pd-7AI 53Pd-47Ni 2150 Z200 Ditto 60Pd-4ONi 2300 Palladium copper Ditto 60Pd-40Cu 65Pd.30Cu..18Cr-4B 84Ni.16Cr-6Fe-4W 67Ni. HIGH-TEMPERATURE BRAZING ALLOYS FOR MOLYBDENUM Alloy Name Composition.Fe ZSCr-ZONi-Bal.

Exact determination of the shear strength of brazed joints at room temperature has not been possible because of the notch sensitivity of molybdenum. has a substantial background of usage for electronic components. which provides fast and reliable heating cycles. ( 4 1) Fair protection was offered for brazing molybdenum in the temperature range 800 to 2600 F (427 to 1427 C) by a flux containing 50 per cent "Eutector 189" °and 50 per cent CaF 2 mixed with alcohol and used in conjunction with a preco. This behavior is common for many brazed materials. Additional work is felt necessary to develop completely satisfactory fluxes for brazing alloys melting above 2000 F (1093 C). . and relative inefficiency in production. Where diffusion of molybdenum into the braze occurs (typical of nickel-base alloys). The advantages of torch brazing are the availability of the necessary equipment. which may result in recrystallization of the molybdenum..36 rates. Disadvantages are inadequate fluxes for high temperatures. but the shape and size of components that can be brazed this way are limited by the available sizes of resistance -heating equipment. The highest braze shear strength was approximately 19. However. to small or simple shapes.. Induction brazing. This process is easily adapted to brazing in vacuum or protective atmospheres. will require more knowledge of the mechanical properties of molybdenum brazes and studies of the properties that determine suitability for operation in various environments. however. Brazed joints are generally designed for shear loading because preferred joint designs usually call for some overlap to facilitate the brazing operation. the results of short-time tensile tests at 1800 F 1982 C) that have been published(4 1 ) indicated that the braze shear strength could be higher or lower than the tensile strength of the filler metal. Similar specimens brazed with fb.es suitable for use with silver-base and copper-base alloys can be used. "Handy Flux" is silver-soldering flux that melts at 800 F and is active up to 1600 F. a strengthening effect was observed. Experi"mentalfurnace-brazed molybdenum assemblies weighing up to 50 pounds have been made succesifully. The high investment required for equipment generally limits its use to production items and. Commercial fl.000 psi at 1800 F when using Inconel as the brazing alloy.elting point of calcium fluoride is 2480 F. Strengths lower than those of the filler metals are probably caused by the formation of brittle intermetallics at the bond line. Jacobson and Martin developed flixes with melting points up to 2600 F (1427 C). "Eutector 189" is a commercial borate-base brazing flm that is active in the range 1400 to 2000 F. Oxyacetylene-torch brazing can be used tý join molybdenum. possible flux entrapment. versatility. and simplicity. The rr. The major current uses of brazed molybdenum components have not required extensive compilations of property data. Resistance brazing -!so can be used for molybdenum parts.ting of "Handy Flux". to some extent. Future use in hightemperature structures.

Rupture Time. • -6 ----* Shear Strength. shear and stress-rupture data are given in Table 5 and Figure 19.750 12.1500 Room 70Pd.Z00-13. psi Type of Joint 1800 1800 1800 1800 1800 1800 1800 1800 . Z50 12. F Inconel Ilaynes Alloy 25 )ionel Hastelloy C Haynes Stellite 21 53Pd-47Ni 60Pd-4OCu 60Pd-40AX 60Pd-4ONi 65Pd-35Cu ."WTI Inconel had a 10G-hour stress-rupture atrength of 5000 psi at 1800 F.400 4. N [-__ v.500 15.900-18. STRESS-RUPTURE CURVES AT 1800 F FOR BRAZES "MADE WITH INCONEL AND HAYNES ALLOY 25(41) I "% _ _ __ __ .400 1"4.800 14.000-33.3OCu 2 A- Tect Tcmperature.800-17.400 6. ':0 p.100-16.. hours A-30•1 -4 FIGURE 19. " I I I0 t 0 "w I 50 j1II• )(0000. Some 20 r 0 lnconel (induchon) I" )ne oalloy 25 (induction) Inconsl (oxyacetylene) N-.200 14.1000 5.100 29.000 12.800-9.000 13. * SHEAR STRENGTHS OF BRAZED JOINTS IN MOLYBDENUM Brazing Alloy Slotted Slotted Slotted Slotted Slotted Slotted Slotted Slotted Single lap Ditto 1500 " 700 1000 1500 " " 93Pd-7Al ~ 18. r--spectively.. InConel broze (oxyocetyltne) Haeynes Alloy 25 braze (cduCtio.000 9.500 O----O Inconel braze (induction I L L V V 7 . TABLE S.

Vacuum"tubeapplications require the use of brazing alloys that do not contain elements that volatilize easily or outgas excessively. Brazing Alloy TENSILE STRENGTHS OF BRAZED BUTT JOINTS IN MOLYBDENUM Test Temperature. 0.38 42 Studies of the tensile strength of vacuum-brazed joints( ) 0. therefore.300-19. 84 per cent nickel.. 52 per cent colwnbium-48 per cent nickel.800 ZZ.900 "1800 S18 0 50Pd-50Fe 0(a) Room Room(a) 1800 S1800(a) g- 49.700-13.400 17. Three alloys. alloys .002 inch thick showed strengths as high as 20. A recent study has been made of brazed joints in the i-olybdenum0.. In addition to strength. were subjected to 24-hour treatment at 1800 F to determine the effects of operation at this temperature on joint strength at 1800 F. -4. F Tensile Strength. (44) The joints were furnace brazed in a hydrogen atmosphere.300 6.400 43.16 per cent titanium.500 17. o-" .100 17.600-7.150 -.-v or procedure with molybdenum are those properties influenced by the service environment. . the braze properties that are most likely to determine the suitability of using a given ali.900 84Ni.'- '"% - '- -.500-39. 001 to "0. TABLE 6.300 52Nb-48Ni Room Room(a) 21.6Ti Room Room(a) 1800 26.000 psi at 1800 F (982 C) (see Table 6). •A.100 6. (a) Heated ix 24 bouz at 1800 7 in vacuum prict to testing.200-28..600-18. and 50 per cent iron-50 per cent palladium.800 9.900 17.700 25. but the strength of the other brazes was not - "affected.300-18.200-34. Palladium-base brazing alloys were found to be the most promising for high-temperature service. psi 70Pu-3OCu 1500 1800 2000 Z5. per cent titanium alloy.200-27.700 18. The strength of the iron-palladium-alloy brazes was reduced to about one-third "ofits previous strength.

Navy Department (Contract No.. "Brazes in parts subjected to oxidizing environments . f:sed salts.. I.-ust be oxidation resistant or must not interfere with the application of ezdation-resista. • . "Development and Properties of Arc-Cast Molybdenum-Base Alloys". Ohio._ .. .••.uid-metal systems are generally nuot severe except in polythermal systems. < -. Ohio. ... References are found under the section 6-hadings to which they have the greatest application.. Battelle Memorial Institute Report to the (•ice of Navýl Research. (2) Climax Molybdenum Company. American Society for Metals. American Society for Metals. "Effec. (5) Jaffe. . most of the high-temperatre alloys are generally suitable.Not only must the braze filler metal have good corrosi.r -. T. . Cleveland. • BIBLIOGRAPHY The bibliography lists all the source material used in the writing of this report.Ga resistance. The behavior of brazed parts in liquid metals... -. N9onr 82100). -. The Metal Molybden. "Arc-Cast Molybdenum and Its Alloys" (1955). -". (4) Semchyshen. 442s-448s (1954). R.J - •••- -r . J. Spacil. and Wulff. Effects such as mass trarnsfer in i•. . The Metal Molybdenum. S. Research Supplement.39 containing cadmium and zinc and the use of fluxes a. 281-329 (1958). . October 15. Cleveland. L.mt coatings.. 330-364 (1958).. 9 4• . ...* not recommended. "Powder Metallurgy Molybdenum-Base Alloys". S. Metallurgical Consideratio=s (13 Case. but there should no: be any corrosive effects caused by *he presence of a bimetallic system. or glasses where molybdenum has excellent corrosion resista-re is difficult to predict.= - - .-~ • - " :L. M. . In the first instance. whi'reas the latter case may require brazing alloys with very high melting temperatures to allow coating with s-licides or other ce ramic coatings.information applicable to mo re than one subject heading.. . Welding Journal. . 33 (9). 1954.. of Oxygen on Welding and Brazing Molybdenum". '" r' - 21 -" • . (3) Perry. . Many references contai. "Summary Report on a Metallurgical Study of Molybdenum".••. . " _ . H..

T. IA (15) Weare. . *. New York (1956). D. R-W. . (16) .40 (6) Mikol. .819. W. Welding Journal. 291s-300s (1957). June 24. W. R. F.. E. "Ductility c-' Tungsten-Arc Welds in Molybdenum". 35 (8).. American Society for Metals.. 2.•Z-% "Testing of Molybdenum Joints (9) Bechtold. . (7) Olds. Western Metal. and Martin. 61-6Z (1957).. 241-261 (1958). (13) Platte. Fusion Welding of Molybdenum (11) Platte. E. . N. L. . United States Patent No. N.. Welding Journal. 36 (6). 151-191 (1958). L. N. (14) Platte.. 1958. N9onr-82100). Research Supplement. (12) Platte. Cleveland. ------ - _ _ _ _ _ _ _ •- .839.•' TFW 0 . February Z8. "The Thermal Conductivity of Molybdenum Over the Temperature Range 100-2100 F".--.. P. W. Molybdenum. Monroe.. Welding Journal. American Society for Metals.. 1953. . 369s-381s (1956). The Metal Molybdenum. and Wessel. The Metal Molybdequm. 195Z. E. 36 (6). 37 (9). Cleveland. Welding Jcurnal. and Fisher. B. 15 (12). Ohio... W. C.. N. - . H. "Effect of Nitrogen on the Soundness and Ductility of Molybdenum We'ds". (8) Northcott. Ohio. "A Metallurgical Study of Molybdenum". E.. E. N. "Influence of Oxygen on the Soundness and Ductility of "Molybdenum Welds". Research Supplement. . "Ductile-to-Brittle Transition in Molybdenum". Cleaning of Molybdenum (10) "Found Deoxidizing Process to Clean Molybdenum for Welding". "Welding of Molybdenum". J. 882-889 (1958). "Inert Gases for Controlled-Atmesphere Processes". Academic Press. E.. 301s-306s (1957). Research Supplement. October 31. -.. R. ORNL-1131. Battelle Memorial Institute Report to the Office of Naval Research (Contract No. F_ Gorman.

Research Supplement. _4__ L "•- . 1957.. 145-148 (July 10. (28) Goodman. W. Weare.. "High-Vacuum Electron-Beam Fusion Welding". Iron (23) Age. .. C. E.. 1 l8s-124s (1958). M. W. . 195Z).roe R. Materials and Methods. Research Supplement. W. 87-89 (December 16. 488s-4"s (i7o1. N. Iron Age. 1169-1175 (1958). . 162. (29) Wyman. Research Supplement. "Jet Engine Push Welded Molybdenum Study". Welding _ournal. 113-117 (1957). Welding "Journal. L.Z and 3. IZZZ-IZZ5 (1956).. W. Noven-ber. .... J. F. L S. 170. Welding Journal. N. Welding Journal. 37 (2). 36 (2). L. N. Welding Journal. 45-48 (1950). (27) Kulju. Welding Journal. Welding Journal. . and Martin. 37-44 (1950)." r %%. 1956. E. D. H.. and Martin. L. W. (21) * Platte.. N.20! Platte. 123) Boam.- ". E. 35S 10). ~(31) Chelius.. N. E. 37 (3). H. Mo-. (30) Heuschhel. J. B. Welding Journal. "Joining of Molybrienurn". 35 (12). . "Welding High Purity Molybdeaum". WADC Technical Report 54-1li. Welding". (26) Kearns. R. Restarch Supplement. "Inert-GasShielded Ccnaruble-Elec'rode Welding of Molybdenum".309. unpublished data. "RcU Planishing Improves WeldJoint Efficiency and Quality". . and Russel. C. August. (19) Weare. private communication. E. N. . "Controlled Atmosphere Arc- . (18) Weare.. D. ."flow to Fabricate Molybdenum". 3Z (7). . ((24) Meredith... and Kearns. "Joining of Molybdenum". 37 (101. K. I1. M. and Monroe. . E. Parts 1. 37 (12). R. -4 -'. C. "Welding of Molybdenum Alloy Sheet". E. "Fabrication and Welding of Arc-Cast Molybdenum". D. 49s-53s (1958). 29 (1). "U'elding of Molybdenum Heat Exchangers".. Resistance Welding of Molybdenum it i'. (25) Weare. "Arc Welding of Molybdenum". WADC Technical Report 57. - (22) Monroe.41 41 (17) Monroe. 1"48). R. and Martin. E. 440s-444s (1958). W.).

251s-264s (1955). J. Welding Journal.. 1954. E. Journal of the Institute of Metals. 65s-74s (1955). and Wulff. 1951. Research Supplement. 1952. May. "The Flash Welding of Commercial Molybdenum . "The Flash Welding of Commercial Molybdenum . F. E. 82. (41) Jacotson. C. C. (38) Weare. B. 34 (2). and Chang. (37) Kearns. American Society for Metals. (42) Dike. Brazing of Molybdenum (40) Monroe. Ohio. "Brazing of Molybdenum". W. Martin. . D. and Martin.Uppes. H. Cleveland. H. I -WTWTWl'7. E. Eichen. WADC Technical Report 58-479. J. * - . December 9. 34 (3). 33 (9). . "Ultrasonic Welding of High-Temperature and CorrosionResistant Material". and Voldrich.' ~~ ~ ~ ~ ~ w'~- ~ -T ~ ~ 1 Tu ýw I~ I. R. Martin. . 1957. (33) Nippes. D. K. A. B.. N. . Antonevich. H. . . . "Welding of Mclybdenum". . R..Part II". July. H. (39) Jones. "The Pressure Welding of Molybdenum". D. E. Washington. . B.pp. (35) Johnston. . Chapters of the American Society for Metals and the American Welding Society. C.. . . 34 (5). Welding Journal. D. and Voldrich. "Joining of Molybdenum"'. Research Supplement. D. B. 374-378 (1954). BMI-703. and Chang. . . M. J. H. . . F. . "Research and Development of Procedures for the Joining of Similar and Dissimilar Heat-Resisting Alloys by Ultrasonic Welding". R. C. I. and Martin. . Monroe. Udin. .~~~ 42 (32) . Welding Journal. W. September 1. Research Supplement. E. "Evaluation of Alloys for Vacuum Brazing of Sintered Advisory Committee for Aeronautics Technical Note 3148. "Brazing Molybdenum for High Temperature Service".Part 1"p Welding Journal. H. C. The Metal Molybdenum. 1958. W. BMI-788. C. Goodwin. . Research Supplement. 13Zs-140s (1955). 3. 192-198 (1958). W. . . C. H. C. paper presented at joint meeting of. E. "The Effect of Purification in Welds in Molybdenum". "(36) Kearns. 449s-458s (1954). . N. Solid-State Welding of Molybdenum (34) Moss. December 1.

I. . III. "Brazing of Molybdenum".. V 4 •' . . "A Dew Point Temperatures Diagram for Metal Metal Oxide Equilibrium in Hydrogen Atmospheres". S. paper presented at the ASM Molybdenum Fabr:cat. *NEW/REM:pa V-s. 1958. 35 (7). 4 ['. .- - (44) Hoppin. H. Welding Journal. May 8.on Conference in Los Angeles. 622s-624s (1956).43 and 44 (43) Chang. G. W. .. Research Supplement.

pe~lsci of Titan. D. bwt PS numbers atid prices iln paMnhlAeae..-ivtac-:5 su cuntrA~i. DrPA:tItk~lt Of M)fCItSC.C MegnXay12 L-atIIuic Columbits 1. DIMC %m~i ur bet 46D Tktlt repartirrdnt of D~efcnre T~tanium Sheet-RILing Program for Shied Mateiihjj (PD 12164.u= Alko~s it Low Temperatures W7) I.''S I I-T 4It IIMI'.)r5 Anidtheir supplsers.wI I Al- .o~a U ~(DT161 - TLemial Stabilit) ofItbe 1~4 - Berylliumn for Situclumna Applications (PS 121648 S3. I C' INICAL ANIR)~Ts JlLll) In tININA U TAL. paie through the Officc I TechnicalSiM. tOao The :~rAI :tp-h ll~lCd bcl~w A:v Aalaasble f&-idust~abilwu to Government cza~cniv. Wa~hsrtuqw 25.ý) - Undo.4lain Program TiIAn'iuPm Shcet-RAU r-P.* CNrtAT ILI CENIER BAli I. Oilier' n.4) eq~acst c.CC.:rn Testing Procedure 4EDepansuiacat of lkfcrrse Tritaijum Sbht't-R.00) Tcus~ke Pro. Gvverui'vi c. C.

.i -de o f4 . A -44ý 4 P~~~j.43. 9%... t..... S*. . oft: P. I . ....ai ful.?. It I.I £tSU i . rob .. - ... legibfle f..f~ 4 44. .. .a%....re j:.r. 4 ..*t S'tp . .*. ... ft.1va -Cop av.f-i.. *.4....14 1.. ..y TI * to%.. . 4.V b. A4-. 7.~~A.I~. f 4 V 447 44.. .w 4..aAri 4.-..p. ..P.6. t I a. !. .L.. .