The Mechanism of Ductility Dip Cracking in Nickel-Chromium Alloys

7/30/2019 The Mechanism of Ductility Dip Cracking in Nickel-Chromium Alloys

1/13

ABSTRACT. High-chromium (~30 wt-%) nickel-alloy filler metals are desirable

for use in nuclear power systems due totheir outstanding resistance to corrosionand stress corrosion cracking. However,these alloys are susceptible to welding de-fects, especially to subsolidus intergranu-lar cracking commonly known as ductilitydip cracking (DDC). In order to develop ahigh-chromium filler metal that is resis-tant to as-welded defects, a series of Ni-Cralloys between 16 wt-% and 34 wt-%chromium were assessed for their suscep-tibility to cracking. Each alloy was evalu-ated by fabricating a restrained, multipass,automatic gas tungsten arc, V-groove

weld, and counting the number of cracksper unit area observable at 50. The typeof cracking (subsolidus DDC or solidifica-tion cracking) was further differentiated

via scanning electron microscopy. The re-sults from these welds, coupled with mi-crostructural characterization, chemicalanalyses, mechanical testing, microstruc-tural modeling, and finite element model-ing indicate that DDC in Ni-Cr alloys iscaused by the combination of macroscopicthermal and solidification stresses in-duced during welding and local grainboundary stresses generated during pre-cipitation of partially coherent(Cr,Fe)23C6 carbides. Cracking can bemitigated by alloying to minimize(Cr,Fe)23C6 precipitation (e.g., by Nb andTi additions), lessening the misfit betweenthe matrix and these precipitates (lower-ing the Cr and Fe concentration), and byminimizing welding-induced stresses. Thismechanism of precipitation-inducedcracking (PIC) via misfit stresses is consis-tent with subsolidus cracking in other alloy

systems including superalloys, nickel-cop-per alloys, titanium alloys, and ferritic

steels where ductility loss corresponds tothe time/temperature regime where par-tially coherent or fully coherent secondphases form.

Introduction

Nickel-chromium-iron alloys are usedextensively in nuclear power systems fortheir resistance to general corrosion, lo-calized corrosion, and environmentally as-sisted cracking. However, concerns withstress corrosion cracking of moderatechromium (1422 wt-%) alloys such as

Alloy 600 and its filler metals (E-182 andEN82) have driven the application ofhigher chromium (2830 wt-%) alloys like

Alloy 690 (Refs. 14). While Alloy 690 andits filler metals show outstanding resis-tance to environmentally assisted crackingin most water-reactor environments (Refs.5, 6), these alloys are prone to welding de-fects, most notably to ductility dip cracking(DDC) (Ref. 4). Ductility dip cracks are in-tergranular and can be surface connected,

which is often an unacceptable conditionin components where fatigue, corrosion fa-tigue, or other forms of environmentallyassisted cracking can occur.

Ductility dip cracking is a solid-statephenomenon, typically occurring in re-

heated weld metal or in base metal heat-affected zones at homologous tempera-tures between 0.4 and 0.9. The name duc-tility dip comes from the correlation ofthis cracking to a decrease in ductility, asdetermined from elevated-temperaturetensile testing. This tensile ductility diphas been reported in several alloy systemsincluding austenitic stainless steels (Refs.7, 8), nickel-based alloys (Refs. 917) (in-cluding age-hardenable nickel-copper al-loys (Refs. 18, 19)), and in both near- and+ titanium alloys (Refs. 2023). In thenickel-chromium alloys of interest, DDC

typically manifests itself as intergranularcracks of one grain or less length (Fig. 1)with a minimum in tensile ductility near870C (Ref. 24).

While there has been considerable re-search into DDC (Refs. 8, 1113, 16, 25),the mechanism or mechanisms responsi-ble for this cracking remain poorly under-stood (Refs. 7, 8, 12, 13, 1618, 2629).Factors suggested to influence DDC haverecently been summarized by Collins,Ramirez, and Lippold (Refs. 9, 10, 12, 13)and include grain boundary sliding, sec-ond-phase precipitation, impurity ele-ment (S, P) segregation, grain boundarycharacter (Ref. 16), degree of grainboundary migration (Refs. 12, 13) and hy-drogen embrittlement. The purpose of thepresent research is to better understandthe mechanism of DDC and to developstrategies to mitigate its occurrence.

Experimental Procedure

In order to assess the resistance toDDC and solidification cracking in the Ni-Cr alloy system, welds were made with 24different filler metals ranging from 15 to34 wt-% chromium. Complete composi-

WELDING RESEARCH

-s31WELDING JOURNAL

SUPPLEMENT TO THE WELDING JOURNAL, FEBRUARY 2008

Sponsored by the American Welding Society and the Welding Research Council

KEYWORDS

Alloy 690Chromium CarbidesDuctility Dip CrackingEN52Intergranular CrackingNickel-Based AlloysSubsolidus CrackingWeldability

The Mechanism of Ductility Dip Cracking in

Nickel-Chromium AlloysSubsolidus cracking results from global stresses produced during

fusion welding and local stresses generated whencoherent or partially coherent second phases form

BY G. A. YOUNG, T. E. CAPOBIANCO, M. A. PENIK, B. W. MORRIS, AND J. J. McGEE

G. A. YOUNG, T. E. CAPOBIANCO, M. A.PENIK, B. W. MORRIS, and J. J. McGEE arewith Lockheed Martin Corp., Schenectady, N.Y.


2/13

WELDING RESEARCH

FEBRUARY 2008, VOL. 87-s32

tions of each weld are detailed in Table 1and include both commercial alloys(EN62, EN82, and EN52) as well as labo-ratory fabricated heats of material. Thedesignation commercial in the notescolumn indicates that the material was

commercially available. Details of thewelding procedure, microstructural char-acterization, hot ductility testing, andcomputational analyses are given below.

Weldments

The susceptibility of a given alloy toDDC was assessed by counting the cracksobserved in metallographic sections takenfrom a multipass, V-groove, gas tungstenarc weld with dissimilar metal side rails(carbon steel and 304L stainless steel).This weld is essentially a linear version of

a piping safe-end weld common to manynuclear power systems and is shownschematically in Fig. 2. The welding para-meters used for this mockup are given inTable 2. Furthermore, the stresses andstrains developed in this weld were inves-tigated via finite element modeling.

Microscopy

Each weld was sectioned into ninepieces for crack counts: eight end cut sec-tions perpendicular to the welding direc-tion and one diagonal slice from root to

crown of the weld. These nine pieces weremetallographically polished to a 0.05-micron finish and examined for defects at50. Any crack-like observations greaterthan 75 microns (0.003 in.) in length wererecorded and were further examined byhigher magnification light optical mi-croscopy and by scanning electron mi-croscopy to differentiate DDC from solid-ification cracking. This procedureexamined approximately 16 square inchesof material per weld and the alloys wereranked via the number of cracks persquare inch.

The dilution in this weld joint was in-

vestigated via scanning electron mi-croscopy and wavelength dispersive spec-troscopy (WDS). The iron concentrationof a cross section from one weld (Heat A3)

was mapped to illustrate the degree ofmixing between the iron-based siderailmaterials and the nickel-alloy weld. Themeasurements were made on a JEOL8200 microprobe operated at 15 kV accel-erating voltage, with a 90-nA probe cur-rent, and at 60-micron step intervals.

Lastly, the grain boundary microstruc-ture from multipass welds of EN82H(Heat A5) and EN52 (Heat B1) were ex-amined via transmission electron mi-croscopy. For this work, grain boundariesin the reheated portion of a weld bead(i.e., in a location that was reheated butnot remelted by subsequent weld beads)

were first identified via light optical mi-croscopy, and then mechanically sec-tioned, ground, and electropolished. Care

was taken to investigate the reheated por-tion of the beads since DDC is observed inbase metal or weld metal heat-affectedzones (Ref. 30). The TEM was performed

on a VG 603 dedicated scanning transmis-sion microscope operated at 300 keV ac-celerating voltage. Phase identification

was determined via selected area diffrac-tion and semiquantitative energy disper-sive spectroscopy.

Hot Ductility Testing

Hot ductility testing of EN52 was as-sessed at both Lockheed Martin and

Lehigh University. In both bases, the sam-ples used were standard round bars sec-tioned from multipass GTAW buildupsand the Gleeble testing employed a heat-ing rate of 93C/s (200F/s), a cooling rateof 32C/s (90F/s), and a stroke rate of 2in./s. At Lockheed-Martin, 6.3-mm-(0.250-in.-) diameter tensile bars wereused to investigate both the on-heatingand on-cooling hot ductility in a Gleeble1500D thermomechanical simulator. AtLehigh, 5-mm- (0.20-in.-) diameter tensilebars were used to confirm the on-coolinghot ductility behavior and to explore theeffect of hold time at the ductility mini-

Fig. 1 Example of a ductility dip crack in a Ni-29Cr-9Fe weld metal (Heat B5). Note that the crackis along the primary (i.e., not migrated crystallo-graphic grain boundary, and is


3/13

WELDING RESEARCH

-s33WELDING JOURNAL

mum temperature. For the Lehigh tests, aGleeble HAZ 1000 thermomechanicalsimulator was used. The peak tempera-tures for the on-cooling tests were similar,i.e., 1282C (2340F) for the Lockheed

Martin tests and 1288C (2350F) for theLehigh tests.

Modeling

Three modeling efforts were used tobetter understand the experiments andinterpret the data from the small bead V-groove welds: 1) finite element modelingof the weld joint, 2) Scheil solidificationmodeling of each composition tested inthe V-groove weld, and 3) microstruc-tural modeling of phase stability, phasetransformation kinetics, and second-

phase misfit.

The finite element modeling was per-formed via commercially availableSysweld software (Ref. 31). The purposeof this work was to explore the experi-mental observation that, if the siderail ma-

terials in the V-groove weld were identical(e.g., both Alloy 600 or both carbon steel),DDC was suppressed, while dissimilarmetal siderails (e.g., one austenitic stain-less steel and one carbon steel) promotedDDC. In order to investigate this notion,a simplified two-dimensional finite ele-ment model consisting of two siderails anda single weld bead was utilized. The modelconsisted of three sections correspondingto the left and right siderails and a single

weld bead. Linear quadrilateral and trian-gular elements were used since linear ele-ments are generally better suited for

analyses involving plasticity. A single fi-

nite element model with identical thermaland structural boundary conditions wasused to investigate the effects of siderailmaterial and filler metal. The model tem-

perature was driven by ramping from

Table 1 Summary of the Compositions Tested in the V-Groove Weld (wt-%) and Resulting Number of Cracks per Square Inch (a)

Heat Notes Cr Fe C Al Ti Nb Mn B Zr N S P Si Mg Cracks/in.2

IDA1 1EN62 15.80 6.90 0.030 --- --- 2.27 0.83 --- --- --- 0.007 0.011 0.17 --- 0

(commercial)A2 Modified EN82 25.50 1.25 0.038 0.04 0.01 2.40 2.67 0.0003


4/13

WELDING RESEARCH


Fig. 3 Illustration of the stresses developed in a single-pass V-groove weld as a function of siderail ma-terials and filler metal. Note that dissimilar metal siderails produce higher maximum principal stresseswith EN82H (C) and EN52 (D) than similar metal siderails (A) and (B).

Fig. 4 Summary of V-groove weld cracking data vs. chromium concentra-tion in the alloy. In general, lower Cr alloys and EN82H-like compositions (i.e.,alloys with ~3 wt-% Nb and 3 wt-% Mn) are resistant up to 29 wt-% Cr.

Fig. 5 Correlation between the number of ductility dip cracks and the ratioof MC-type carbide formers (Nb+Ti) and Cr, which promotes M23C6-type car-bides in high-chromium (> 20 wt-% Cr) nickel alloys.

room temperature to 2912F (1600C) inthe first second, holding constant for fourseconds, and cooling to ambient tempera-ture by convection and radiation. For thestress analysis, the model was constrainedhorizontally and vertically at the bottom-left siderail, vertically at the bottom-right

siderail, and evaluated in the plain straincondition. Loading for the structuralanalysis is due to thermal expansion andbased on the temperature distribution andthe interference strain due to phase trans-formation in the carbon steel end rail. Thematerial combinations investigated were1) Alloy 600 siderails with EN82 and EN52filler metals and 2) 304SS and ASTM 516Grade 70 siderails with EN82 and EN52filler metals. In all cases, alloy-specific,temperature-dependent material proper-ties were used.

The solidification behavior of each V-groove weld was investigated viaJMatPro,

version 4.0 software (Ref. 32). This analysiswas performed to estimate the amount ofcarbon in fcc gamma phase at the end of so-lidification. The calculations utilized the Ni-Fe alloy database and considered the fol-lowing elements for each alloy (Ni, Al, Cr,Fe, Mn, Nb, Ti, B, C, and N) and the fol-

lowing phases (liquid, fcc nickel, M23C6,M7C3, MC, and Laves phase). Note thatJMatPro 4.0 utilizes a Scheil-Gulliver solid-ification model, modified for fast diffusionof carbon and nitrogen (Ref. 33).

The microstructural modeling usedbothJMatPro andFactSage (Ref. 34) soft-

ware.JMatProwas used to assess the effectof alloying elements on the precipitationkinetics of M23C6-type carbides, to gener-ate time-temperature-transformation dia-grams for select alloys, and to estimate themisfit between (or ) and the matrix ofselected nickel-based superalloys. For theTTT calculations, the solvus temperature

for 100% was first determined, and curvefor 0.1% precipitation of or calcu-lated. For these calculations, the nominalcompositions of the alloys, supplied by thesoftware, were used.

The FactSage calculations were per-formed to assess the effects of tempera-

ture, chromium, and iron concentrationon chromium-rich carbide stability. Thesecalculations used the SGTE 2004 data-base, considered the , M7C3, and M23C6phases, and were performed assuming 1atm of pressure for a temperature rangeof 10002000F (5381093C),chromium concentrations between 15 and35 wt-%, and at two different iron levels(1 and 10 wt-%).

The elastic constants of the Cr23C6 car-bide phase were calculated via first-princi-ples methods. To give a sense of the accu-racy of this methodology, the elasticconstants of nickel were also calculated at

A B

C D

CONTOURSmax principal stresstimeDeformed shape X 1Comput. Ref Global


5/13

WELDING RESEARCH

-s35WELDING JOURNAL

the same level of theory and with identicalinput parameters. Geometry optimizationsand total energies were calculated via den-sity functional theory as implemented inthe Vienna Ab Initio Simulation Package(VASP) version 4.6 and the MedeA version2.1 interface (Ref. 35). The computationalprocedure for these calculations was torelax the structure to


6/13

WELDING RESEARCH


cracks/in.2, and D heats greater than 30cracks/in.2. With the exceptions of Heat B2,

where cracking was biased adjacent to the

stainless steel siderail, DDC occurred uni-formly throughout the weldment. This weldproduced repeatable results, i.e., multipleheats of EN82 (A4, A5, A6) did not crackand similar heats of EN52 (B1 and B5) pro-duced the same extent of cracking. Further-more, this weld differentiates the extent ofcracking in the alloys from no cracks inHeats A1A8, to >30 cracks/in.2 in HeatsD1D3. Note that in addition to DDC, hotcracking was observed in two heats of ma-terial (C3 and C4), which confounds thecrack counts. These heats were not includedin subsequent correlations of DDC to

composition.

Finite Element Modeling

Finite element modeling helps explain

the experimental observation that this V-groove weld geometry with dissimilarmetal siderails promotes DDC while thesame geometry and welding parameters

with similar metal siderails minimizesDDC (see footnote C in Table 1). Asshown in Fig. 3, the maximum principalstress developed in the dissimilar metal

weld is approximately 5 ksi (35 MPa)higher than the Alloy 600 siderail weld.

Also of note is that the stresses developedin EN82 and EN52 in the dissimilar metal

weld are comparable [~35 ksi (241 MPa)].The similar maximum stress developedbetween DDC-resistant (EN82) and

DDC-prone (EN52) alloys suggests fac-tors other than the macroscopic stressmust cause the different DDC resistancebetween the two alloys.

The Effect of Alloy Composition on DDC

In general, lower chromium alloys suchas EN62 and EN82 are immune from DDCin this weld, while higher chromium alloys

such as EN52 are susceptible to DDC asshown in Fig. 4. While EN82 did not showDDC in this weld, it is notable that anEN82-like heat without niobium exhibitedDDC (Heat B8). Furthermore, EN82-likecompositions with typical Nb and Mn levels(~3 wt-%) but higher chromium levels didnot DDC, until reaching a level of 29.8 wt-% chromium. Lastly, lowering the titaniumof EN52-like heats appears to exacerbatecracking (Heats D1 and D2). Since Nb andTi are both strong MC-type carbide-form-ing elements, these observations suggestthat carbides may have an important role in

DDC. The tendency to form MC-type car-bides (vice chromium-rich carbides) is as-sessed in Fig. 5 which plots the number ofductility dip cracks vs. the (Nb+Ti)/Cr ratio.

As shown in Fig. 5, heats with a high(Nb+Ti)/Cr ratio, i.e., a strong tendency toform MC-type carbides vs. Cr-rich carbides,show resistance to DDC.

The occurrence and extent of DDCshows little correlation with bulk levels oftramp elements (S, P) or with sulfide-forming elements (Mn and Mg) as shownin Fig. 6. While sulfur embrittlement ofnickel and nickel alloys is well established

(Refs. 3638), it is important to note thatvery well desulfurized heats (e.g., HeatsB8, B9, and C1 with 0.002, 0.00056, and


7/13

WELDING RESEARCH

-s37WELDING JOURNAL

(low Nb) are susceptible to DDC. More-over, DDC in low Ti variants of EN52show increased cracking susceptibility(e.g., D1 and D2). These observations sug-gest that these elements, which are knownstrong MC-type carbide and carbo-nitrideformers (where M = Nb, Ti) may play arole in DDC resistance. In fact, the pre-

dominant type of carbide is one key dif-ference between EN82, which typicallydisplays a mixture of Nb-rich, MC-type,and M23C6-type grain boundary carbidesand EN52, which forms primarily(Cr,Fe)23C6-type carbides (with occa-sional Ti-rich, MC-type carbides) asshown in Fig. 7.

The observation of semicontinuousprecipitation of (Cr,Fe)23C6-type carbidesalong the grain boundary in a DDC sus-ceptible alloy (EN52) is consistent withthe work of Capobianco and Hanson whoperformed scanning Auger microscopy on

in-situ fractured ductility dip cracks and

found extensive (Cr,Fe)23C6 precipitationon the fracture surface (Ref. 11). It is alsonotable that the carbides are partially- co-herent with one side of the grain boundary

with a cube-on-cube ((100)M23C6 (100)fcc alloy) orientation relationship(Refs. 39, 40). Lastly, it is important tonote that since M23C6 carbides are stable

to high temperatures and their precipita-tion displays fast kinetics in Ni-30 wt-% Cralloys, the microstructural observations ofthis work and the work of Capobianco andHanson represent a complex thermal his-tory from multipass welds.

A backscatter scanning electron micro-graph and the corresponding composi-tional map of the iron concentration in the

weldment is shown in Fig. 8. As expected,there is significant enrichment in iron ad-

jacent to the carbon steel and 304L stain-less steel siderail materials. The iron con-centration near the root of the weld is on

the order of 30 wt-%, while the iron en-

richment adjacent to the siderails is on theorder of 15 wt-% and falls off to the fillermetal iron concentration in approximately0.4 in. (10 mm). As discussed previously,

while Heat B2 displayed DDC biased to-ward the stainless steel siderail, ductilitydip cracks were observed uniformlythroughout the weldment for the bulk of

the alloys that were susceptible to DDC.The uniformity of cracking indicates thatthis weld is a good test for the DDC sus-ceptibility of the filler metal.

Gleeble Hot Ductility Testing

Results of the Gleeble hot ductilitytesting on EN52 are shown in Fig. 9. Con-sistent with previous studies on the hotductility of austenitic alloys, EN52 showslittle evidence of ductility loss on-heatingbut a significant tensile ductility dip on-cooling (Refs. 7, 26, 41). The minimum in

ductility as determined by % reduction in

Fig. 8 Backscatter electron image (top) and quantitative wavelength dispersive spectroscopy compositional map of the iron dilution in the weld. The filler

metal was A3 (27.3 wt-% chromium).

Fig. 9 The tensile ductility of EN52 filler metal on-heating (circles) and on-

cooling (squares). Two heats of material and slightly different peak tempera-

tures 1282C filled squares and 1288C open squares) were used for the on-cooling tests. The ductility data are compared to the calculated TTT kinetics of

M23C6carbide precipitation (dashed-dot line), showing very good correlation

between the nose of the curve and the ductility minimum near 860C.

Fig. 10 Effect of hold time at the temperature of the tensile ductility mini-

mum in EN52, illustrating the transient nature of ductility loss.


8/13

WELDING RESEARCH


area occurs at approximately 860C

(1580F). As suggested by Nippes and oth-ers (Refs. 7, 17, 41), this ductility mini-mum corresponds to a temperatureregime where second-phase precipitationis occurring. Note the very good corre-spondence between the minimum in duc-tility and the nose of the (Cr,Fe)23C6 TTTcurve near 860C. Additionally, the recov-ery in ductility at temperatures above~900C (1652F) and below ~750C(1382F) is also in good agreement withthe temperature regime of (Cr,Fe)23C6carbide precipitation. One implication ofincreased ductility at temperatures below

750C is that sulfur embrittlement is un-

likely, i.e., if sulfuris segregating toand embrittlinggrain boundaries at1600F, ductilityloss should persistto lower tempera-tures, if not to roomtemperature, in theon-cooling Gleebletesting.

AdditionalGleeble testing at

the ductility mini-mum temperatureshows that ductilityis recovered withhold time at 871C(1600F), as illus-trated in Fig. 10.The recovery ofductility with holdtime at tempera-ture has importantmechanistic impli-cations because inthis temperature

regime (i.e., at homologous temperatures~0.60.7) it would be expected that trampelements such as sulfur are segregating tothe grain boundary and exacerbating em-brittlement, not moving away from thegrain boundary and improving ductility(Refs. 36, 37, 42, 43).

The Mechanism and Modeling of DDC

The results of the V-groove weld, theTEM study in front of a ductility dip crack,and the Gleeble testing indicate that DDCis associated with the tendency to form

grain boundary (Cr,Fe)23C6-type car-

bides. As shown by the V-groove welds,high levels of Cr tend to promote DDC(Fig. 4), while strong MC-type carbideforming elements such as Nb and Ti im-part resistance to DDC Fig. 5. Further-more, microstructural investigations re-

veal semicontinuous (Cr,Fe)23C6-typecarbides in a DDC susceptible alloy(EN52) (Fig. 7) and on DDC fracture sur-faces (Ref. 11). Additionally, Gleeble test-ing shows that regime of ductility loss cor-responds to the temperature range of(Cr,Fe)23C6 carbide precipitation (Fig. 9)

and that ductility loss recovers with holdtime at temperature Fig. 10. Note thatall of the alloys susceptible to DDC willform M23C6-type carbides in the ductilitydip temperature range determined forEN52 (14001800F) as shown by thethermodynamic stability of the carbides asa function of temperature and bulkchromium concentration in Fig. 11. No-tably, lower-chromium, DDC-resistant al-loys (e.g., Heat A1, EN62) tend to formpseudo-hexagonal M7C3-type carbides(which are typically incoherent) ratherthan partially coherent M23C6-type car-

bides (Refs. 39, 44, 45).The tendency of a given alloy to form

M23C6-type carbides can be assessed byplotting the bulk (Cr + Fe) concentrationand the calculated free carbon in solutionafter solidification. The chromium andiron concentrations are summed becauseFe can substitute for Cr in the M23C6 car-bide. Thermodynamic-based models indi-cate that approximately 12 atomic % ofFe can substitute for Cr in the DDC tem-perature range of 14001800F (Ref. 32).The Scheil solidification calculation of thefree carbon in the fcc gamma phase inte-

grates the effects of MC-type carbide and

Fig. 11 Illustration of the effects of chromium and iron on the thermody-

namic stability of M7C3and M23C6-type carbides. Increasing chromium and

iron both stabilize the M23C6phase.

Fig. 12 Results from the V-groove weld ductility dip crack counts plotted

against the measured (Cr+Fe) and calculated free carbon after solidification.

Note three distinct regions: DDC free compositions in the lower left, moving up

and to the right a transition region where DDC is observed near siderail dilu-

tion followed by a region of general susceptibility as (Cr+Fe) and free carbon

are increased.

Fig. 13 Illustration of the effect of iron on the time-temperature-transfor-

mation kinetics of (Cr,Fe)23C6carbides in Ni-30Cr-XFe-0.03 wt-% C alloys. As

iron is increased, precipitation is shifted to shorter times, which helps explain

the observation of general DDC in heat B6 (8.75 wt-% Fe) but only localized

DDC in regions of siderail dilution in Heat B2 (2.91 wt-% Fe). The curves were

calculated with JMatPro 4.0 software.


9/13

WELDING RESEARCH

-s39WELDING JOURNAL

carbo-nitride forming elements (e.g., Tiand Nb).

The correlation of these compositionalparameters with the results of the V-groove welds are shown in Fig. 12, whichseparates the data into three distinct re-gions: 1) low Cr+Fe and free carbon

where the alloys tested are immune toDDC in this weld, 2) a region of increasedsusceptibility as Cr+Fe and free carbon

are increased. In this region, DDC was ob-served near the stainless steel weld siderailin one heat (B2), possibly due to increasediron and carbon from base metal dilution,and finally 3) a region of general suscepti-bility to DDC as Cr+Fe and free carbonare further increased. In this regime, DDCoccurs uniformly in the weld joint and isnot biased by siderail dilution. The dashedlines on Fig. 12 give theapproximateextentof DDC, which increases up and to theright (i.e., with increased tendency for(Cr,Fe)23C6 precipitation).

While Fig. 12 provides important in-

sight into interpreting the results of the V-groove weld and offers a practicalmethodology to develop DDC-resistantalloys, other factors likely affect the ten-dency for DDC. One notable comparisonis between Heats B6 and B2, which havesimilar compositions with the exception ofthe iron level (8.75 wt-% in B6 and 2.91 wt-% in B2). Heat B6 with higher iron dis-played 2 cracks/in.2, throughout the weld,

while Heat B2 displayed fewer cracks (1crack/in.2), that were biased toward thestainless steel siderail dilution area. In thesiderail dilution areas, iron and free car-

bon in the weld metal can be enriched (seeFig. 8 and note that the 304L side contains~0.02 wt-% C and >65 wt-% Fe), likely in-creasing the driving force for (Cr,Fe)23C6precipitation. The effect of increased ironincreasing the time-temperature-transfor-mation kinetics of model Ni-30Cr-XFe-0.03C (wt-%) alloys is shown in Fig. 13.

Additionally, research on Ni-Cr-Fe ternar-ies indicates that increasing iron levels willincrease the carbide/matrix misfit (Ref.46), as discussed below.

The correspondence of DDC with thetime/temperature regime of the onset ofM

23C

6precipitation and the noted effect

of welding-induced stresses indicates thatDDC in Ni-Cr alloys results from the com-bination of global stresses induced from

welding (or imposed via tensile testing)and a local effect associated with precipi-tation of grain boundary M23C6-type car-bides. One difference between M23C6 andcarbides in DDC-resistant alloys is thatM23C6 carbides typically nucleate withpartial coherency ((100)M23C6 (100)fccalloy), while MC-type and M7C3-type car-bides are typically incoherent with the fccmatrix (Refs. 39, 40, 44, 47, 48).

The cube-on-cube orientation rela-

tionship between the carbide and onegrain at a grain boundary likely generatesappreciable elastic stresses and, in combi-nation with global stresses from welding(or applied during tensile testing), is pos-tulated to result in local intergranular

cracking, i.e., ductility dip cracking. Pre-cipitation-induced grain boundarystresses in Ni-30Cr alloys are likely a max-imum near the onset of precipitation anddecreases with time as: 1) the chromiumconcentration near the grain boundary isdepleted by carbide precipitation and sub-sequent growth (discussed further below),2) the carbide/matrix interface grows offthe original plane of the grain boundary(i.e., as the precipitation-induced stressesare spatially removed from the grainboundary), and 3) dislocations are gener-ated to help accommodate the interfacialmisfit. The mechanism of precipitation-in-

duced cracking is outlined in Figure 14,which shows A the formation of par-tially coherent, grain boundary M23C6 car-bides and a schematic of the lattice misfit,B how local tensile stresses develop be-tween carbides and promote crack nucle-

ation, and C the result of these inter-carbide cracks linking up with subsequentstressing to form a typical ductility dipcrack along the crystallographic grainboundary.

The effect of the misfit betweenM23C6-type carbides and the matrix canbe further assessed by considering thespacing of {100} planes in each phase.The difference between the room-tem-perature lattice parameters of the Cr23C6carbide and selected alloys are summa-rized in Table 3. The room-temperaturelattice parameter is a good estimate forthe d-spacing in the DDC temperature

Fig. 14 Illustration of the mechanism of ductility dip cracking in Ni-Cr alloys. A Partially coherentgrain boundary M23C6carbides form in reheated weld metal with significant misfi t between the carbideand the alloy as shown above and in Fig. 15; B precipitation generates local grain boundary tensilestresses between carbides promoting crack nucleation; and C macroscopic stresses generated on-cool-ing (e.g., thermal and solidification stresses) often link the inter-carbide cracks and result in ductility dipcracking along the crystallographic grain boundary.

Table 3 Summary of Lattice Parameters and Cr23C6Carbide/Matrix Misfit for SelectedNi-Cr-Fe Alloys

Material Lattice Parameter d(100)(Angstroms) (Angstroms) Reference

Cr23C6 10.6595 3.5532 --- (62)A600 3.553 3.553 0.006EN82H 3.570 3.570 0.473 (63)

A690 3.5757 3.5757 0.633

% *( )

d d

d

Cr C a lloy

Cr C

23 6

23 6

100

A

B

C


10/13

WELDING RESEARCH


range (~14001800F), since the thermalexpansion of these both the alloys andCr23C6 carbides is similar and small in mag-nitude (


11/13

WELDING RESEARCH

-s41WELDING JOURNAL

carbide is significantly stiffer (E100 ~ 334GPa) than nickel (E100 ~112 GPa).

The elastic stress produced in thenickel alloy matrix during nucleation canthen be estimated from Equation 2, whereE, and are taken as the modulus and thestrain in the direction. This esti-mation (E=112 GPa, = 0.64%) results inelastic stresses near 690 MPa (100 ksi),which is far in excess of typical yieldstrengths for A690 or EN52 [typically inthe range of 207414 MPa (3060 ksi) atroom temperature]. While this is a crudeapproximation of the actual stresses at thecarbide/grain boundary interface, it doesshow that the elastic misfit stresses are ap-preciable and could result in cracking, es-pecially when additional global stressesare present as in the case of welding or me-chanical testing.

= E (2)

While misfit strains will partially offset the

energy benefit of nucleating with partialcoherency on the grain boundary, recentwork on Alloy 690 shows that even on highangle grain boundaries, M23C6-type car-bide precipitate with partial coherency(Ref. 47). In other words, the misfit strainbetween the carbide and matrix is not solarge as to negate the benefit of the cube-on-cube precipitation. However, grainboundary misorientation does affect therate of precipitation as discussed by Lim etal. (Ref. 47). Notably, the misfit betweenthe M23C6 carbide and Alloy 690 quotedby Lim (0.67%) is similar to the present

study (0.63%).The effects of coherent precipitate in-terfacial stresses on second-phase stress,strain, and compositional profiles havebeen studied in more detail by W. C. John-son (Ref. 50). Johnsons analyses showthat long range stresses can develop fromthe interface of a coherent second phaseand that, using reasonable parameter esti-mates, the magnitude of the interfacialstress is on the order of ~101000 MPa(1.45145 ksi). While mainly driven bythermodynamic modeling of phase equi-libria, Johnsons work gives further evi-dence that misfit strains from coherent orpartially coherent precipitates can lead toappreciable local stresses and potentiallyto cracking of the interface.

As shown schematically in Fig. 15, the{100}d-spacings of the Cr23C6 carbide aresmaller than in high-chromium nickel al-loys indicating that the misfit stresses arecompressive at the carbide/grain bound-ary interface. As multiple carbides nucle-ate along a grain boundary, this causes re-gions of tension between the carbides andcould produce intermittent grain bound-ary cracks, consistent with some experi-mental observations of DDC shown in Fig.

14. Similarly, failure could also occur atthe partially coherent side of the car-bide/matrix interface. Of course, the exactductility dip crack morphology is likely acomplex function of a given materials ten-

dency to form M23C6 carbides, local grainboundary orientation, and the magnitudeand direction of the global stresses andstrains.

Ductility Loss in Other Alloy Systems

The good correspondence betweenductility dip cracking in high-chromiumnickel alloys and (Cr,Fe)23C6 precipita-tion, suggests that the same mechanism ofprecipitation-induced cracking may affectother alloys. Review of nickel-based alloyliterature data shows that Alloys X-750,

718, and Monel K-500 all display signif-icant intermediate temperature ductilitydips as shown in Fig. 16 (Refs. 15, 19,5153). In the case of X-750, ductility losscorresponds to precipitation of partiallycoherent M23C6-type carbide, similar toEN52 and A690. In Alloy 718, the tensileductility dip correlates to the temperaturerange of precipitation of M6C-type car-bides. Like M23C6-type carbides, M6Chave a complex cubic structure, a latticeparameter similar to M23C6-type carbides(11.26 vs. 10.66 ) and likely have a(100)M6C (100)fcc alloy orientation rela-tionship with the matrix (Ref. 54). Monel

K-500 exhibits a large ductility dip, both interms of breadth of temperature range(~620 to 1610F) and ductility loss (~25%to 2% elongation). The tensile ductility dipcorresponds to the temperature regime

where will precipitate (Ref. 19), consis-tent with precipitation-induced cracking. It is also notable thatMonel 400, which does not form , doesnot exhibit this large ductility loss (Ref. 52).

The correspondence of ductility losswith precipitation in Monel K-500 issuggestive of strain-age cracking in nickel-based superalloys (Ref. 55). As shown byHughes and Berry (Refs. 56, 57), and laterby Franklin and Savage (Ref. 58), strain-age cracking occurs in the time/tempera-ture regime of precipitation, analogousto the correspondence of DDC with

M23C6 precipitation in EN52-type alloys.Furthermore, studies on the superalloyRene 41, show that strain-age cracking iscaused by the combined action of globalrestraint stresses and precipitation-induced stresses (Refs. 59, 60), the samebasic mechanism that is postulated forductility dip cracking. This assertion issupported by the JMatPro calculations ofmisfit strain and precipitation time (Ref.32), i.e., alloys that are prone to strain-agecracking (e.g., Udimet 700, IN 100, andAstroloy) have fast precipitation kinet-ics and large negative / mismatch rela-tive to alloys that are resistant to strain-

Table 4 Comparison of the Elastic Moduli of Nickel and Chromium Carbide (Cr23C6)

Elastic Modulus GPa (Msi)

ReferenceNickel - Experiment 150 (21.8) 250 (36.2) 320 (46.4) (64)Nickel - Calculated 112 (16.3) 196 (28.4) 260 (37.7) This workCr23C6 - Calculated 334 (48.4) 327 (47.4) 325 (47.1) This Work

Table 5 Comparison of C-curve Kinetics, Precipitate Mismatch, and Strain Age CrackingResistance for Selected Superalloys

Susceptibility to Calculated Nose of (") CalculatedAlloy Strain Age Precipitation Curve Time and Mismatch at

Cracking (Ref. 55) Temperature the Nose of C-(seconds/C (F))(a) curve(%)(a)

Inconel 718 Low 197/871 (1600) +0.44Inconel X-750 Low 600/832 (1530) +0.58Waspaloy Low 16.9/960 (1760) +0.11Rene 41 Medium 8/990 (1814) 0.28

Udimet 700 High 1.7/1070 (1958) 0.28Astroloy High 1.8/1070 (1958) 0.39IN 100 High 0.4/1149 (2100) 0.45

(a) TTT curves and precipitate mismatch were calculated withJMatPro 4.0 software, assuming all phases form and0.1% fraction transformed.


12/13

WELDING RESEARCH


age cracking (e.g., Inconel 718, In-conel X-750, and Waspaloy) which dis-play longer times before nucleation oc-curs and positive mismatch (i.e., matrixcompression on precipitation). A sum-mary of precipitation times, temperatures,misfit strains, and tendency to strain-agecrack are given in Table 5.

Lastly, it should be noted that themechanism of precipitation-induced

cracking presented herein is very similarto those proposed by Rath et al. for thesubsolidus cracking of/ titanium alloysand by Swift and Rodgers for stress-reliefcracking of 214Cr-1Mo steels (Refs. 20,61). In titanium alloys system, crackingoccurs in the temperature regime of the phase transformation and Rath andcoworkers have postulated that the stressassociated with this transformationcauses cracking. Similar to the presentstudy, grain boundary typically nucle-ates with the Burgers orientation rela-tionship (110) (0001), [111]

[1120], and is believed to generate sig-nificant local stresses (Ref. 20). Whilethere is no general agreement on themechanism or mechanisms of subsoliduscracking in titanium alloys, the precipita-tion-based mechanism of Rath et al. isgenerally consistent with the susceptibil-ity of a wide range of model alloys (Ref.20). Similarly, the research of Swift andRodgers indicates that reheat cracking in214Cr-1Mo steels occurs when partiallycoherent {(110) (001)2 C, and[100] [110] 2 C} Mo2C carbidesform. Resistance to cracking is reestab-

lished when precipitates grow in size andare no longer able to impart coherencystrains on the matrix (Ref. 61).

The correspondence of ductility lossto precipitation of a partially or fully co-herent precipitates in other alloys helps

validate the findings of the present studyand indicates that DDC, strain-age crack-ing, reheat cracking, and observations ofintermediate temperature ductility lossare often caused by precipitation-in-duced cracking. In alloys where precipi-tation of partially coherent or coherentprecipitates with large misfit occurs atshort times (e.g., A690, EN52, Astroloyand IN-100), the result is a weldabilityproblem. In alloys where precipitation ofthe misfitting second phase is sluggish,ductility loss is often observed during hottensile testing or deformation processing(e.g., X-750, Inconel 718, and Monel K-500). The key point is that ductility lossoccurs in the time/temperature regime

where the global stresses overlap with thelocal, precipitation0induced stresses.Welding-induced cracks are promoted bylarge precipitation-induced stresses,large global stresses (e.g., high weld con-straint), and with increasing volume frac-

tion of the detrimental second phase.

Conclusions

Ductility dip cracking in Ni-Cr alloys islikely a form of precipitation-inducedcracking similar to intermediate tempera-ture tensile ductility loss noted in severalother nickel-based alloys, strain-agecracking in superalloys, subsolidus crack-ing in titanium alloys, and stress-reliefcracking in ferritic steels. In Ni-(2030 wt-%)Cr alloys, DDC is caused by the combi-nation of macroscopic thermal and solidi-fication stresses induced during weldingand local stresses generated during grainboundary precipitation of partially coher-ent (Cr,Fe)23C6 carbides. The strong cor-relation between the time/temperaturedependence of DDC with precipitation,but not with grain boundary impurity seg-regation (e.g., sulfur) indicates that for thecompositions tested, segregation is not asignificant factor. Cracking can be miti-gated by alloying to minimize grainboundary (Cr,Fe)23C6 precipitation (e.g.,

by Nb and Ti additions), lessening the mis-fit between the matrix and these precipi-tates (lowering the Cr concentration), andby minimizing welding-induced stresses(e.g., by avoiding dissimilar metal welds).The alloying strategies outlined herein,offer promising avenues to improve the

weldability of high-chromium nickel-alloyfiller metals while maintaining the out-standing corrosion resistance required formany power generation applications ofthese materials.

Acknowledgments

The authors wish to thank Dr. RickNoecker, Dr. Masashi Watanabe, andProf. John DuPont of Lehigh Universityfor their important contributions to this

work. Additionally, at Lockheed Martinwe gratefully acknowledge the assistanceof Mr. Dominick Amedio, Dr. Cathy Jor-dan, Mr. Nathan Lewis, Dr. Reza Na-

jafabadi, Mr. Paul Sander, and Dr. JohnSutliff for their help in this research.

References

1. Gorman, J. A. 2001. Corrosion problems

in the nuclear industry and implications for thefuture in chemistry and electrochemistry ofstress corrosion cracking. A symposium honor-ing the contributions of R. W. Staehle, TMS.

2. Bamford, W., and Hall, J. 2003. A reviewof Alloy 600 cracking in operating nuclearplants: historical experience and future trends.11th International Conference on EnvironmentalDegradation of Materials in Nuclear Power Sys-tems-Water Reactors. Stevenson, Wash., NACE.

3. Young, G. A., et al. 2005. The mechanismand modeling of intergranular stress corrosioncracking of nickel-chromium-iron alloys ex-posed to high purity water. 12th InternationalConference on Environmental Degradation ofMaterials in Nuclear Power Systems. Salt LakeCity, Utah, TMS.

4. Young, G. A., and Capobianco, T. E. 2007.Development of a highly weldable and corro-sion resistant Ni-Cr filler metal. 13th Confer-ence on Environmental Degradation of Materialsin Nuclear Power Systems. Whistler, BC,Canada, CNS.

5. Inconel Alloy 690. 1980, HuntingtonAlloys.

6. Jacko, R. J., Gold, R. E., and Kroes, A.2006. Accelerated corrosion testing of Alloy52M and Alloy 182 weldments. 11th Interna-

tional Conference on Environmental Degrada-tion in Nuclear Power Systems - Water Reactors.Stevenson, Wash., TMS.

7. Nippes, E. F., Savage, W. F., and Grotke,G. 1957. Further studies of the hot-ductility ofhigh-temperature alloys, Welding ResearchCouncil, New York, N.Y.

8. Nissley, N. E., and Lippold, J. C. 2003.Development of the strain-to-fracture test.Welding Journal 82(12): 355-s to 364-s.

9. Ramirez, A. J., and Lippold, J. C. 2004.High temperature behavior of Ni-base weldmetal, Part I. ductility and microstructural char-acterization. Mat. Sci. & Eng. A, 2004. 380:259271.

10. Ramirez, A. J., and Lippold, J. C. 2004.High temperature behavior of Ni-base weldmetal, Part II. Insight into the mechanism forductility dip cracking.Mat. Sci. & Eng. A, 380:245258.

11. Capobianco, T. E., and Hanson, M. 2005.Auger spectroscopy results from ductility dipcracks opened under ultrahigh vacuum. Pro-ceed ings of the 7th International ConferenceTrends in Welding Research. Callaway Gardens,Ga., ASM International.

12. Collins, M. G., Ramirez, A. J., and Lip-pold, J. C. 2003. An investigation of ductility dipcracking in nickel-based weld metals Part 2.Welding Journal 82: 348-s to 354-s.

13. Collins, M. G., Ramirez, A. J., and Lip-

pold, J. C. 2004. An investigation of ductility-dip cracking in nickel-based weld metals Part3. Welding Journal 83: 39-s to 49-s.

14. Inconel Alloy X-750. 1998. Inco AlloysInternational, Huntington, W.Va. pp. 125.

15. Inconel Alloy 718. 1985, InternationalNickel Co., Inc.

16. Dave, V. R., et al. 2004. Grain boundarycharacter in Alloy 690 and ductility-dip crack-ing. Welding Journal 83: 1-s to 5-s.

17. Weiss, B., Grotke, G. E., and Stickler, R.1970. Physical metallurgy of hot ductility test-ing. Welding Journal 49: 471-s to 487-s.

18. Rhines, F. N., and Wray, P. J. 1961. In-vestigation of the intermediate temperatureductility minimum in metals. Trans ASM 54:

17128.19. Engineering Properties of MonelAlloy K-500. 1965. International Nickel Co.

20. Rath, B. B., et al. 1994. High Tempera-ture Ductility Loss in Titanium Alloys A Re-

view, DE-AC04-94AL85000. 1994, Sandia Na-tional Laboratories.

21. Damkroger, B. K., Edwards, G. R., andRath, B. B. 1989. Investigation of subsolidus

weld cracking in alpha-beta titanium alloys .Welding Journal 68: 290-s to 302-s.

22. Hayduk, D., et al. 1986. Cracking sus-ceptibility of Ti-6al-2Nb-1Ta-0.8Mo as deter-mined by the Varestraint test. Welding Journal,65: 251-s to 260-s.

23. Bowden, D. M., and Starke, E. A. 1984.

The effect of microstructure and deformation


13/13

WELDING RESEARCHbehavior on the hot ductility of Ti-6Al-2Nb-1Ta-0.8Mo.Met. TransA, 15A: 16871698.

24. Noecker, F. F., and DuPont, J. N. 2005.Microstructural characterization and hot duc-tility of nickel based weldments. AmericanWelding Society, Dallas, Tex.

25. Collins, M. G., and Lippold, J. C. 2003.An invest igation of ducti lity dip cracking innickel-based filler materials Part 1. WeldingJournal 82: 288-s to 295-s.

26. Yeniscavich, W. 1987. Joining, in Super-

alloys II, C. T. Sims, N. S. Stoloff, and W. C.Hagel, editors. John Wiley & Sons: New York,N.Y., pp. 495516.

27. Patriarca, P. 1967. Effects of minor ele-ments on the weldability of high nickel alloys.Houston, Tex., Welding Research Council.

28. Yeniscavich, W., and Fox, C. W. 1967. Ef-fects of minor elements on the weldability ofHastelloy alloy X. Effects of Minor Elementson the Weldability of High-Nickel Alloys. Hous-ton, Tex., Welding Research Council.

29. Canonico, D. A., et al. 1967. Effects ofminor additions on weldability of Incoloy 800.Effects of Minor Elements on the Weldability ofHigh-Nickel Alloys. Houston, Tex., Welding Re-search Council.

30. Welding, Brazing, and Soldering. OnlineEdition 1993.ASM Handbook. Vol. 6.

31. Group, E., Sysweld. 2005.32. JMatPro version 4.0. 2005, Sente Soft-

ware: Guildford, UK.33. Saunders, N., et al. 2003. Modeling of

thermo-physical and physical properties rele-vant to solidification. Advanced Solidificat ionProcessing X., TMS.

34. Bale, C. W., et al. 2005. FactSage.Thermfact, Montreal, Canada.

35.MedeA Version 2.1. 2005, Materials De-sign, Inc.

36. Mulford, R. A. 1983. Grain-boundaryembrittlement of Ni and Ni alloys. Materials

Science and Technology Embrittlement of En-gineering Alloys, C. L. Briant and S. K. Banerji,editors. Academic Press: New York. pp. 119.

37. Bruemmer, S. M., et al. 1983. Influenceof sulfur, phosphorus, and antimony segrega-tion on the intergranular hydrogen embrittle-ment of nickel.Met. Trans.A, 14A: 223232.

38. Young, G. A., et al. 2003. An atomisticmodeling study of alloying element, impurity el-ement, and transmutation products on the co-hesion of a nickel 5 {001}twist grain bound-ary. 11th Conference on EnvironmentalDegradation of Materials in Nuclear Power Sys-tems - Water Reactors. Stevenson, Wash., ANS.

39. Edington, J. W. 1976. Practical ElectronMicroscopy in Materials Science, Hendon, Va.,

Tech Books.40. Yonezawa, T., Yamaguchi, Y., andIijima, Y. Electron micro-autoradiographic ob-servation of tritium distribution on alloy X750.Proceedings of the International Symposium onPlant Aging and Life Predictions of CorrodibleStructures. Sapporo, Japan.

41. Yeniscavich, W. 1966. A correlation ofNi-Cr-Fe alloy weld metal fissuring with hotductility behavior. Welding Journal 45: 344-s to356-s.

42. Mulford, R. A. 1983. Grain bound-ary segregation in Ni and binary Ni alloys doped

with sulfur.Met. Trans A., 14A: 865870.43. Jones, R. H., and Bruemmer, S. M. 1988.

Environment-induced crack growth processes

in nickel-base alloys, Envi ronment-Induced

Cracking of Metals, R. P. Gangloff and M. B.Ives, editors, NACE. pp. 287310.

44. Sims, C. T., Stoloff, N. S., and Hagel, W.C., eds. 1987. Superalloys II, John Wiley & Sons:New York, p. 615.

45.Inconel Alloy 600, SMC 027. 2000. Spe-cial Metals Corp., Huntington, W.Va., p. 19.

46. Marucco, A. 1994. Atomic ordering inthe Ni-Cr-Fe system. Mat. Sci. & Eng., A189:267276.

47. Lim, Y. S., et al. 2004. The effect of grain

boundary misorientation on the intergranularM23C6 precipitation in thermally treated Alloy690.J. Nucl. Mat., 335: 108114.

48. Norring, K., Stiller, K., and Nilsson, J. O.1991. Grain boundary microstructure, chem-istry, and IGSCC in alloy 600 and Alloy 690.5thInternational Symposium on EnvironmentalDegradation of Materials in Nuclear Power Sys-tems - Water Reactors. Monterey, Calif.

49. Hinnuber, J., and Rudiger, O. 1954.Chromium carbide in hard metal alloys. Sym-posium on Powder Metallurgy.

50. Johnson, W. C. 2000. Superficial stressand strain at coherent interfaces.Acta Mat., 48:433444.

51.Huntington Nickel Alloys. 1968, Interna-tional Nickel Co., Huntington, W.Va.

52. Monel: Nickel-Copper Alloys. 1969. In-ternational Nickel Co., Huntington, W.Va.

53. Dix, A. W., and Savage, W. F. 1973. Shorttime aging characteristics of Inconel X-750.Welding Journal 52: 135-s to 139-s, 144-s.

54. Kuo, K. 1953. The formation of Eta car-bides.Acta Met., 1: 301304.

55. Prager, M., and Shira, C. S. 1968. Weld-ing Research Council Bulletin 128.

56. Hughes, W. P., and Berry, T. F. 1967. Astudy of the strain-age cracking charateristics in

welded Rene 41 Phase 1. Welding Journal46: 361-s to 370-s.

57. Berry, T. F., and Hughes, W. P. 1969. A

study of the strain-age cracking characteristicsin welded Rene 41 Phase II. Welding Jour-nal 48: 505-s to 513-s.

58. Franklin, J. E., and Savage, W. F. 1974.Stress relaxation and strain-age cracking inRene 41 weldments. Welding Journal 53: 380-s to 387-s.

59. Messler, R.W., Principles of Welding .1999, New York: John Wiley & Sons.

60. Schwenk, W., and Trabold, A. F. 1963.Weldability of Rene 41. Welding Journal42(10): 460-s to 465-s.

61. Swift, R. A., and Rodgers, H. C. 1973.Embrittlement of 214Cr-1Mo steel weld metalby postweld heat treatment. Welding Journal 52:145-s to 153-s, 172-s.

62. Yakel, H. L. 1987. Atom distributions intau carbide phases.Acta Crystallographica, B43:230238.

63. Sander, P. C. 2005. Lattice parameters ofselected nickel-based alloys as determined byX-ray diffraction. G. A. Young, Schenectady,personal communication.

64. Kittel, C. 1996. Introduction to SolidState Physics. 7th ed., New York: John Wiley &Sons. p. 673.

An Important

Eventon Its Way?

Send information on upcom-ing events to the WeldingJournal Dept., 550 NW LeJe-une Rd., Miami, FL 33126.Items can also be sent via FAXto (305) 443-7404 or by [email protected].

Dear Readers:

The Welding Journal encouragesan exchange of ideas throughletters to the editor. Please sendyour letters to the Welding JournalDept., 550 NW LeJeune Rd.,Miami, FL 33126. You can alsoreach us by FAX at (305) 443-7404or by sending an e-mail to KristinCampbell [email protected].

Documents

The Mechanism of Ductility Dip Cracking in Nickel-Chromium Alloys