Introduction

Nickel-based alloys are extensivelyused in stationary components of gasturbines such as combustion chamber,casing, liner, exhaust ducting, andbearing housing (Refs. 1, 2). In this re-gard, Alloy 263 (UNS N07263), a precipitation-hardened alloy, and AlloyX (UNS N06002), a solid solutionstrengthened alloy, are highly on de-mand for good creep strength and out-standing oxidation resistance. The most important joining tech-nique in manufacturing of the compo-nents mentioned above is gas tungstenarc welding (GTAW). Welding opera-

tions should be carried out with materi-als in solution-treated condition. In thesolution treatment condition (1150ºC/2h/air cooling), Alloy 263 and Alloy X mi-crostructures consist of Ti-enriched MCcarbide and Mo-enriched M6C carbidedistributed in the matrix, respectively.It should be mentioned that Alloy X suf-fers from the heat-affected zone (HAZ)liquation phenomenon due to a reactionbetween M6C carbides and the matrix toform an interfacial liquid film that is ateutectic composition. This process hap-pens as a result of heating above the eu-tectic temperature, called constitutionalliquation (Refs. 3–5). In general, enhancement of me-chanical properties with Ni-based al-

loys is of great importance in weld-ments. A fundamental way to increasethe yield strength and toughness is torefine the structure of the fusion zone(FZ). In addition, formation of fine den-dritic structure in the FZ can cause a de-crease in the susceptibility of solidifica-tion cracking during welding (Refs. 5,6). It has been reported that the addi-tion of nitrogen in argon gas changesthe microstructure of the weld metal inaustenitic steels (Refs. 7, 8). Nitrogen is a strong solid solutionstrengthening element in austeniticstainless steels (Ref. 9). According to theprevious works, solubility of nitrogen innickel-based melts is less than that iniron-based alloys and depends stronglyon alloy composition (Refs. 10–12). Itwas known that Cr plays a more impor-tant role in increasing the nitrogen solu-bility than other elements in a nickel-based melt (Ref. 13). The existence of ahigh nitrogen amount in the melt cancause precipitation of nitride com-pounds (Refs. 14, 15). Moreover, the ni-trides may either be transferred directlyfrom the filler metal during the weldingprocess to the FZ or be formed by a heattreatment of alloys containing a suffi-cient amount of the nitrogen present insolid solution. Heat treatment has noinfluence on the nitrides (Refs. 11, 15).Ramirez et al. (Refs. 15, 16) showedthat the titanium nitride transferredfrom the filler metal to the molten weldpool of Ni-based Alloys 600/625, and690 dissolves significant amounts of Cand Cr whose reason is the equilibrium

WELDING RESEARCH

Nitrogen Effect on the Microstructure andMechanical Properties of Nickel Alloys

In Alloy 263 and Alloy X, the weld metal microstructures and mechanical propertiesdue to welding by different Ar­N2 shielding gases were observed

BY B. NABAVI, M. GOODARZI, AND V. AMANI

ABSTRACT In this research, the effects of nitrogen addition in Ar gas on weld metal microstruc­ture and mechanical properties of Alloy 263 (UNS N07263) and Alloy X (UNS N06002)were studied. Autogenous gas tungsten arc welding (GTAW) was employed by adding0–4 vol­% N2 in Ar. Welding speed and heat input rate were measured as functions of gascomposition. The weld metal microstructure was studied by optical and scanning elec­tron microscopy. Experimental results demonstrated that the dendritic structure of theweld was refined by increasing N2 in Ar for both alloys. An addition of 4 vol­% N2 to Ardecreased significantly the columnar region in Alloy 263 fusion zone (FZ), while no simi­lar change was observed in Alloy X. This difference is discussed based on microstructuralcharacterization. Finally, it was found that the tensile strength and hardness have beenaugmented with increased nitrogen in the shielding gas.

KEYWORDS • Nitrogen • Alloy 263 • Alloy X • Gas Tungsten Arc Welding (GTAW) • Equiaxed Dendrites • Ultimate Tensile Strength (UTS)

B. NABAVI (behrooz.nabavi@gmail.com) and M. GOODARZI are with the School of Metallurgy and Materials Engineering, Iran Universityof Science and Technology, Tehran, Iran. V. AMANI is with the School of Metallurgical and Materials Engineering, College of Engineering,University of Tehran, Tehran, Iran.

FEBRUARY 2015 / WELDING JOURNAL 53-s

NABAVI SUPP FEB 2015_Layout 1 1/9/15 3:56 PM Page 53

phase composition change through so-lidification temperature range from TiNto (MTi)(CN), where M stands for othermetallic elements. The beneficial impact of nitrogen onthe carbide morphology in unidirection-al and single crystal Ni-based alloys wasreported by several investigators (Refs.14, 17). It has been also revealed thatthe addition of nitrogen to CMSX-4, asingle crystal Ni-based superalloy, in-creases the number of solidification de-fects, especially high-angle boundaries(Ref. 18). According to Nage et al. (Ref.8), nitrogen has a positive effect on de-creasing the compositional differencebetween the dendrites and interden-drites regions in 904L stainless steels. Ithas been recognized that nitrogen im-proves the depth of welding penetra-tion, corrosion resistance, and mechani-cal properties in stainless steels (Refs. 7,8, 19–22). Unlike iron-based alloys, there is adearth of information on the mi-crostructural and mechanical behaviorof Ni-based weld metals obtained byAr-N2 shielding gases. Therefore, inthe current work, the effect of N2 addi-tion in Ar gas on the FZ microstruc-ture plus mechanical properties of Al-loy 263 and Alloy X were investigated.

Experimental Method

Materials and SamplePreparation

In the present study, solution an-nealed Alloy 263 and Alloy X sheets of100 x 70 x 1 mm size were prepared. To

ensure about heattreatment condition,specimens were solu-tionized at 1150ºCfor 2 h followed bywater quenching.The chemical compo-sitions and mechani-cal properties of thetest materials aregiven in Tables 1Aand 1B, respectively.Gas tungsten arcwelding was carriedout by an experi-enced specialistwelder without fillermetal. The shieldinggas was argon with amaximum 4 vol-%nitrogen fraction,prepared in a specialmixer. To protect theweld root against oxida-tion, to avoid the influ-ence of oxygen on thenitrogen dissolution inthe welds and promote structure unifor-mity during solidification, specimenswere welded on a water-cooled copperbackup bar with dimensions of 25 X 15 X5 cm. Argon was purged continuouslyfrom the bottom as the backing gas. Constant welding parameters in-clude the following: current 20 A; di-ameter of the 2% thoriated tungstenelectrode 1 mm; electrode tip angle 60deg; arc length 3 mm; and the shield-ing gas and backing gas flow rate 81/min and 6 1/min, respectively. Theheat input rate, welding voltage, andwelding speed as the functions of

shielding gas composition are listed inTable 2. The nitrogen content in thewelds was measured by inert gas fu-sion using a LECOanalyzer on theweld metal extracted by drilling. The welding speed was measured bythe following:

S = x/t (1)

where S is the welding speed (mm/s), xis the length of weld (x = 100 mm),and t denotes welding time (s). The heat input rate was estimatedby the following:

WELDING RESEARCH

WELDING JOURNAL / FEBRUARY 2015, VOL. 9454-s

Fig. 1 — Schematic diagram showing the dimensions of the weldedspecimens used in the tension tests.

Fig. 2 — Optical images of solution heat treated (1150°C/2h) base metal. A — Alloy 263; B — Alloy X showing inter­granular and intragranular particles within the matrix.

Table 1A — Chemical Compositions (wt­%) of Test Materials

Alloy C Si Mn Cr Co Mo Fe Al Ti Cu W S N Ni

Alloy 263 0.05 0.25 0.3 19.58 19.1 5.9 0.48 0.2 2.38 0.1 – 0.0017 0.0062 balAlloy X 0.1 0.15 0.75 22.17 1.3 9.02 18.5 0.15 – 0.3 0.6 0.0053 0.0137 bal

A

B

NABAVI SUPP FEB 2015_Layout 1 1/9/15 3:56 PM Page 54

H = VI/S (2)

where H, , I, and V stand for theheat input rate (J/mm), arc efficiencycoefficient ( = 0.65), current (I = 20A), and the arc voltage (V), respec-tively.

Microstructural Characterization

To examine microstructural fea-tures of the weld metal, optical andscanning electron microscope (SEM)were used. Elemental distribution ofthe precipitations in welds was deter-mined by energy dispersive X-ray ana-lyzer (EDX) coupled with SEM. Forsamples preparation, specimens werefirst machined in needed dimensions.Thereafter, they were mounted andground with SiC abrasive paper downto 1500 mesh grit. The polishing oper-ation was conducted with 3 and 1 mdiamond paste. Alloy 263 sampleswere etched in a solution with 2 parthydrochloric acid, 2 part nitric acid, 3parts acetic acid, and four drops ofglycerin for 4 to 8 s. A solution with 3parts HCl, 1 part HNO3, and 2 partsglycerin was used to etch Alloy X sec-tions for 25 to 30 s. The size of theequiaxed dendrites and secondarydendrite arm spacing (SDAS) were de-termined in the etched FZ sections us-ing Clemex image analysis software.

Mechanical Tests

Vickers microhardness was meas-ured using a load of 0.49 N. The ten-sile behavior of the welds was evaluat-ed for all samples welded with variousshielding atmosphere. Figure 1 illustrates the configura-tion of the tensile specimens made inaccordance with the specifications ofASTM E8, Standard Test Methods forTension Testing of Metallic Materials.Each specimen was inspected by X-raymethod prior to mechanical tests.

Results and Discussion

Base Metals

Figure 2A and B show the typicaloptical microstructure of preweld solu-tion heat treated base Alloy 263 andAlloy X, respectively. As seen from Fig.

2A and B, both alloys contained dis-persed particles in intergranular andintragranular regions of the matrix. Backscattered SEM micrographs ofthe alloys are shown in Fig. 3A and B.The particles distributed in the matrixof Alloy 263 appear in dark contrast(Fig. 3A) whereas the particles in thegrains of Alloy X appear in bright con-trast — Fig. 3B. The EDS spectra fromthe particles, shown in Fig. 3C and D,revealed that they are of the MC-typecarbide and M6C-type carbide in Alloy263 and Alloy X, respectively. Metallographic evidence also eluci-dated that liquation phenomenon oc-curred in the HAZ of Alloy X. One ofthese liquated M6C carbides was shownin bottom left inset of Fig. 3B at highermagnification. This liquated carbide canbe compared to a pristine carbideshown in the top right inset of Fig. 3B. What’s more, Fig. 4A shows an op-tical image of the HAZ in Alloy X.Two liquated M6C can be seen in Fig.4A. The liquation can be identified by

partial dissolution near the edges ofthe large carbides. Under the heatingconditions experienced in the HAZ,small carbides can totally dissolve inthe matrix. In the case of MC car-bides, they have been found with twoblocky and cylindrical morphologiesin the base Alloy 263. For example,one blocky shaped carbide approxi-mately 7 m in size and one cylindri-cal carbide, 10 m long and about 2m wide, were shown in Fig. 4B andC, respectively. Several investigators have reportedthat MC carbide forms generally withcylindrical and blocky morphologiesduring solidification, hot workingprocess, and aging treatment (Refs. 1,3). Owing to the large (> 5 m) andmedium (1–2 m) average size of thecarbides, it is reasonable to assumethat the carbides observed in the ma-trix of Alloy 263 are either primary so-lidification constituents or secondaryphases precipitated during the hotworking process.

WELDING RESEARCH

FEBRUARY 2015 / WELDING JOURNAL 55-s

Fig. 3 — SEM microstructure. A — Alloy 263; B — Alloy X weldments; C and D — EDS spec­tra from MC and M6C carbides distributed in Alloy 263 and Alloy X matrices, respectively.

Table 1B — Mechanical Properties of Test Materials in the Solution­Treated Condition

Material Yield Strength (MPa) Ultimate Tensile Strength (MPa) Elongation % Hardness (Hv)

Alloy 263 424 860 56 245Alloy X 330 721 47 240

A

C

B

D

NABAVI SUPP FEB 2015_Layout 1 1/9/15 3:56 PM Page 55

Nitrogen Content in the Welds

The nitrogen content in the weldsas a function of the shielding gas com-position was illustrated in Fig. 5. Theexperimental results demonstrate thatthe N content of the weld increaseswith the increasing N2 level in theshielding gas mixture for both alloys.Similar results were found in the pre-vious works on austenitic steels (Refs.7, 8, 19–22). From Fig. 5, it can beseen that the nitrogen solubility of Al-loy X welds is higher than that of Alloy263 welds. It was found that effective factors onnitrogen solubility in the weld includethe alloying element content of the FZ,primary nitrogen level, the surface ac-tive element concentration in the weld,and the welding parameters (Ref. 20).Regarding higher primary nitrogen level

in Alloy X, low sulfurconcentration forboth materials (Table1A) and higher weld-ing speed for Alloy X(Table 2), it is sug-gested that the differ-ence in nitrogen solu-bility in Alloy 263 and Alloy X may re-sult from the presence of various alloy-ing elements.

Welding Speed

Table 2 shows the effect of shield-ing gas composition on welding volt-age. As seen from Table 2, the weldingvoltage increases by an N2 rise in Ar. Itis consistent with findings by severalresearchers (Refs. 7, 19, 21, 22). Theinfluence of different shielding gas at-mospheres on the welding speed is

also shown in Table 2. Based on the experimental results,welding speed increases with a nitro-gen rise in the shielding gas. Doping of4% N2 to Ar leads to increased weldingspeed by about 20% compared withthe pure argon shielded process forboth Alloy 263 and Alloy X. The keyreason is associated with increasing inthe density of heat flux to anode (sam-ple) due to adding N2 to Ar. The heatintensity or heat flux at the anode sur-face is given by the following:

Equation 3 includes three terms:heating by electron capture, thermalconduction, and blackbody radiativecooling. Here je is current density, ϕwais work function, k is thermal conduc-tivity, T is the temperature, z is dis-placement in the direction perpendi-cular to the anode surface, εa is emis-

S j kTz

T (3)a e wa a4= ϕ −

∂

∂− ε α

WELDING RESEARCH

WELDING JOURNAL / FEBRUARY 2015, VOL. 9456-s

Fig. 4 — Optical images. A — M6C carbides in the HAZ of Al­loy X; B and C — MC carbides in base Alloy 263.

Fig. 5 — Nitrogen content in the weld metal as a function ofthe shielding gas composition.

Fig. 6 — Influence of shielding gas composition on the size ofequiaxed dendrites d and SDAS λ in welds.

Table 2 — Influence of Shielding Gas Composition on the Voltage and Heat Input Rate

Shielding Gas Voltage (V) Welding Speed (mm/s) Heat Input Rate (J/mm)Composition Alloy 263 Alloy X Alloy 263 Alloy X

Ar 10.3 1.25 2 107 ± 1 67 ± 1Ar­1% N2 10.5 1.3 2.05 105 ± 1 67 ± 1Ar­2% N2 11 1.35 2.15 105 ± 1 67 ± 1Ar­3% N2 11.3 1.4 2.3 105 ± 1 63 ± 1Ar­4% N2 11.4 1.5 2.4 99 ± 1 62 ± 1

A

B

C

NABAVI SUPP FEB 2015_Layout 1 1/15/15 10:07 AM Page 56

sivity of the anode, and finally de-notes the Stefan-Boltzmann constant.Some authors take into account an ad-ditional term, 5kB/2ejeTp, where kB isthe Boltzmann’ constant, and Tp is theplasma temperature at the edge of thesheath. If this term were included, theterm proportional to je would increaseby about 20%. This would not alter theconclusion of the total heat flux. Another term that has been neglect-ed is the anode fall voltage as well, be-cause the fall can be negative due tostrong electron diffusion to the anode,and its impact is in any case small com-pared with the electron capture term.According to Murphy et al. (Refs. 23,24), nitrogen addition to argon causes ahigher increase in current density andless increase in thermal conductivity ofarc, so the total heat flux increases, in-dicating arc constricts.

Microstructural Investigation

Based on the microstructural exami-nation, the FZ of Alloy 263 and Alloy Xsuperalloys consisted of equiaxed den-drites at the center and columnar den-drites around them. The effect ofshielding gas composition on the size ofthe equiaxed dendrites and the SDAS isshown in Fig. 6. Experimental resultsshow that increasing the N content inthe weld metal leads to reduction of theequiaxed grains size and the SDAS inboth alloys. According to Fig. 6, the size of theequiaxed dendrites decreases over 11%in Alloy 263 welds and 27% in Alloy Xwelds by addition of 4 vol-% nitrogen inthe shielding gas in comparison withthe pure argon shielded method. In ad-dition, the SDAS in Alloy 263 welds re-duced from 16 ± 1 m for pure argon to11 ± 1 m for Ar-4% N2 shielding gas. Asimilar decline in the SDAS occurred inAlloy X weld metals (Fig. 6).

Figure 7A–D displays the columnarzone of the weld metal of Alloy 263 andAlloy X specimens obtained by pure Arand Ar-4% N2 shielding gases. Compar-ing Fig. 7A–D indicates that the weldstructure has been altered due to thewelding by N2-containing shielding gas.It is known that the SDAS depends onthe cooling rate and mobility factor. In

Ni-base alloys, the SDAS correlateswith the cooling rate by the following:

~ (tf)1/3 ~ ()–1/3 ~ (G)–1/3 (4)

where tf is the solidification time, isthe cooling rate, G is the temperaturegradient, and is the solidificationrate (Refs. 6, 25). According to Equa-

WELDING RESEARCH

FEBRUARY 2015 / WELDING JOURNAL 57-s

Fig. 7 — Influence of nitrogen addition to argon gas on weld metal structure. A — PureAr and B — Ar­4% N2 for Alloy 263; C — pure Ar and D — Ar­4% N2 for Alloy X.

Table 3 —EDS Results of MC Precipitates Observed in Alloy 263 Welds Shielded by Arand Ar + 4% N2 Gases

Elements Shielding Gas (wt­%)Pure Ar Ar + 4% N2

C 10.2 13.5Ti 56.4 43.7Co 5.2 7.1Cr 8.1 9.7

Mo 3.2 5.1Ni 16.9 20.9

Fig. 8 — Influence of shielding gas composition on the microstructure of Alloy 263welds. A — Pure Ar; B — Ar­2% N2 ; C — Ar­3% N2; D — Ar­4% N2 .

A

A

C

C

B

B

D

D

NABAVI SUPP FEB 2015_Layout 1 1/9/15 3:56 PM Page 57

tion 4, the cooling rate is the productof the temperature gradient and solidi-fication rate. Since the sample geomet-rical shape and environmental condi-tion are the same for all tests, the tem-perature gradient and subsequentlythe cooling rate are constant. It wasfound that the mobility factor is afunction of surface tension and parti-tion coefficient (Ref. 8). Therefore, thereduction in the SDAS is attributed tothe change in surface tension andchemical composition due to the nitro-gen rise in the weld metal. It is neces-sary to mention that the decrease inthe SDAS corresponds with a more ho-mogeneous distribution of alloying el-ements (Ref. 25). Figure 8A–D illustrates typical opti-cal micrographs associated with the ef-fect of the nitrogen amount in theshielding gas on the weld metal mi-crostructure adjacent to the HAZ inAlloy 263. As seen from Fig. 8A–D,dendritic structure of the welds is sig-nificantly changed by increasing in thenitrogen content. As a case in point, itis clear that the area of columnar zoneis considerably decreased by a nitrogenrise in the shielding gas. That is, theratio of the columnar area to the FZarea decreased from 0.55 for speci-mens welded with Ar to less than 0.2

for weldments prepared with Ar-4% N2shielding gas; nitrogen addition leadsto a noticeable extending in theequiaxed zone in the weld metal. Optical micrographs of the Alloy 263weld metals obtained by differentshielding gas compositions are depictedin Fig. 9A–D. Two precipitates withblocky and cylindrical morphologies areevident in Fig. 9A and B, respectively. Table 3 represents the EDS results ofthe precipitates in Alloy 263 FZ weldedwith N2 free and 4% N2 containingshielding gases. According to Table 3,the observed precipitates in the Ar-shielded weld are MC-type carbide dueto the presence of C and Ti, which con-tain other elements. High amounts ofnickel, chromium, and cobalt detecteddefinitely indicates high interferencewith the matrix surrounding the precip-itate. Similarity of blocky and cylindricalphases in Fig. 4B and C, and Fig. 9A andB also supports the existence of MC-type carbides in the weld metal. Based on the metallographic investi-gation, it was found that with increas-ing the nitrogen content in the weldmetal, the number of medium size pre-cipitates (1–2 m) increases in the Alloy263 FZ. The quantity of these precipi-tates in the weld made by Ar-4% N2shielding gas exceeded four times as

much as the number of them in theweld made by pure Ar. The orange-likecolor of these secondary phases sug-gests that they are either nitride compo-sitions or carbides similar to Fig. 9A andB detected in the specimens weldedwith pure argon. In order to identifythese precipitates, SEM images of thedifferent N-bearing welds were studied. Figure 10 illustrates an SEM imageof the weld metal prepared with Ar-4%N2 shielding gas. EDS analysis related toone of these precipitates was listed inTable 3. The analysis result reveals highC concentration. However, nitrogen wasnot detected in any of the analyzed pre-cipitations; it doesn’t suggest the ab-sence of nitrogen because the precisionlimit of this analysis was 0.1 wt-%. Lowatomic weight of C and N elementsmakes it difficult to identify them in theEDS analysis, though the high wt-% ofC supports the presence of carbide orcarbo-nitride enriched in Ti and Cr. Itshould be noted that beside large andmedium size precipitates, some verysmall phases were also observed. It is likely that they are of eitherM23C6-type carbides precipitated alongthe welds grain boundary or MN-typenitrides like TiN particles formed inthe weld pool. Ramirez et al. (Ref. 16)identified both nanoscale M23C6 car-bides precipitated along grain bound-ary and very small nitrides (< 50 nm)in strain-to-fracture Ni-based samplestrained 1.6% at 956ºC for 10 s. Ofcourse, it is necessary to consider thatwe did not apply any heat treatmenton the weldments.

WELDING RESEARCH

WELDING JOURNAL / FEBRUARY 2015, VOL. 9458-s

Fig. 9 — Optical images of Alloy 263 specimens welded by different shielding gas composi­tions. A and B — Pure Ar; C — Ar­3% N2; D — Ar­4% N2. (Arrows show MC­type carbides.)

Fig. 10 — SEM image of Alloy 263 weld met­al prepared with Ar­4% N2 shielding gas.

A

C

B

D

NABAVI SUPP FEB 2015_Layout 1 1/9/15 3:56 PM Page 58

Nonetheless, regarding the presenceof Ti, a strong nitride former, and highamounts of nitrogen in the melt, it ispossible to precipitate nanoscale TiNparticles throughout the liquid metalduring solidification. Therefore, it issuggested that the heterogeneous nu-cleation of carbides on extremely fineTiN precipitates is the reason for the in-creasing in number of MC precipitatesresulted from the nitrogen rise in Alloy263 welds. The impact of TiN precipi-tates on motivating heterogeneous nu-cleation of carbides at a higher tempera-ture reported before (Ref. 14). It has been previously ascertained inGTA welds of Alloy 690 using thermo-dynamic calculations that TiN is formedin higher temperatures in the melt andis changed to the TiC phase (TiN iso-morph phase) during solidification inthe weld metal (Refs. 16, 17). Hence,this phase transformation is likely to bethe cause for nonidentification of TiNparticles in none of the specimens weld-ed with N2 containing shielding gas. On the other hand, investigation ofoptical micrographs of Alloy X welds ob-tained by different shielding gasesshowed that the area of the equiaxedgrains has not been changed with in-creasing N2 content in Ar. Figure 11A–D

shows the weld metal microstructure ofAlloy X adjacent to the HAZ resultedfrom different shielding gas composi-tion. As seen from Fig. 11A–D, nochange was observed in the area of thecolumnar region of the welds with anaddition of nitrogen to argon. Based on the above discussions, thisdifference between Alloy 263 and AlloyX welds can be described by the role ofTiN or carbonitrides formed in Alloy263 welds. Reasons confirming theaforementioned notion include an ab-sence of Ti in Alloy X (Table 1A); thepresence of N in Alloy X welds abouttwice as much as Alloy 263 weld metalsmade by similar shielding gases (Fig. 5);and a high affinity between Ti and N,plus no alteration in the area ofequiaxed grains in the Alloy X welds. Moreover, it should be stated thatno precipitation was observed in AlloyX welds prepared with all of the shield-ing gases. As a result, the main reasonfor significant grain refinement of Al-loy 263 weld metal resulted from 4%N2 added to Ar gas is based on the ideathat the dendrite nucleation was in-tensified heterogeneously on MC,(MTi)(CN), or TiN precipitates. Blackarrows in Fig. 9C and D show threeMC carbides nuclei at the center of

three equiaxed grains. The investigation of the optical im-ages shows that the number of cylindri-cal carbides decreases as the nitrogencontent in the weld increases. That is,increasing the nitrogen content of theweld leads the cylindrical carbides to be-come blocky. This finding is in goodagreement with all investigators (Refs.14, 17). A clear mechanism for thischange in morphology has not yet beenfound. In addition, with increasing ni-trogen content in the weld pool, thequantity of large size carbides signifi-cantly decreased.

Mechanical Tests Results

The effect of nitrogen content onthe ultimate tensile strength (UTS) isshown in Fig. 12. Based on the experi-mental results, UTS increases withaugmenting the amount of nitrogen inargon gas. The main reason is relatedto the microstructure refinement dis-cussed in the previous section. Thesecond reason is precipitationstrengthening caused by raising thenumber of MC carbides in the Alloy263 weld zone. Indeed, glide of dislo-

WELDING RESEARCH

FEBRUARY 2015 / WELDING JOURNAL 59-s

Fig. 12 — Effect of nitrogen content in ar­gon gas on the ultimate tensile strength ofweldments.

Fig. 13 — Vickers hardness of the weldmetal as a function of the shielding gascomposition.

Fig. 11 — Influence of shielding gas composition on the microstructure of Alloy X welds.A — Pure Ar; B — Ar­2% N2 ; C — Ar­3% N2 ; D — Ar­4% N2.

A

C

B

D

NABAVI SUPP FEB 2015_Layout 1 1/9/15 3:56 PM Page 59

cations could be hindered by MC pre-cipitates, so the dislocations pile-up atthese obstacles. Hence, more stress isneeded to move the dislocations. Figure 13 displays the microhard-ness of the weld as a function of theshielding gas composition. Microhard-ness measurements on the as-weldedspecimens indicate that the hardnessof the weld is less than that of the basemetal for both alloys (Table 1B andFig. 13). It is interesting to note thatVickers hardness reached HV 238 andHV 227, respectively, for Alloy 263and Alloy X welds resulting from 4%N2 addition to Ar. This finding wasconsistent with microstructural obser-vations; in other words, the refine-ment of the dendritic structure (Figs.7, 8) and interstitial solid solutionstrengthening and precipitation hard-ening (Fig. 9) have been obviouslyidentified as hardening mechanisms inthe weld zone.

Conclusions

The weld metal microstructuresand mechanical properties due towelding by different Ar-N2 shieldinggases were examined in Alloy 263 andAlloy X. The results are summarized asfollows: 1) In both alloys, with increasingthe amount of N2 in Ar gas, the N levelin the welds increased and dendriticstructures were refined. 2) A considerable decrease in thecolumnar region in Alloy 263 was ob-served due to an addition of 4% N2 inAr, while such a similar event did notoccur in Alloy X. 3) As the N content of the Alloy 263weld metal increases, the number ofMC precipitates increases, and theytend to precipitate in blocky form. 4) It is suggested that heteroge-neously promoted nucleation of thedendrites and MC carbides during so-lidification are the main reasons forthe microstructural modification in Al-loy 263 weld. 5) The UTS increased from 773MPa and 665 MPa for Ar gas to 815MPa and 715 MPa for Ar-4% N2 shield-ing gas in Alloy 263 and Alloy X weld-ments, respectively. 6) The hardness of the weld aug-mented with increasing the N contentof the weld zone in both superalloys.

The authors would like to acknowl-edge the MAPNA Group for financialsupport of this research and employ-ees of MavadKaran Co., especially Mo-hammad Cheraghzadeh and Amin Am-jadi, for supplying materials and kindassistance.

1. Wang, W. Z., Hong, H. U., Kim, I. S.,Choi, B. G., Jeong, H. W., Kim, M. Y., and Jo,C. Y. 2009. Influence of ' and grain bound-ary carbide on tensile fracture behaviors ofNimonic 263. Materials Science and Engineer-ing A 523: 242–245. 2. Zhao, J. C., Larsen, M., and Raviku-mar, V. 2000. Phase precipitation and time-temperature-transformation diagram ofHastelloy X. Material Science & Engineering:112–119. 3. Zhao, J. C., Ravikumar, V., and Bel-tran, A. M. 2001. Phase precipitation andphase stability in Nimonic 263. Metallurgicaland Materials Transactions A 32A:1271–1282. 4. Hong, H .U., Kim, I. S., Choi, B. G.,Jeong, H. W., and Jo, C. Y. 2008. Effects oftemperature and strain range on fatiguecracking behavior in Hastelloy X. MaterialsLetters 62: 4351–4353. 5. DuPont, J. N., Lippold, J. C., and Kiser,S. D. 2009. Welding metallurgy and weld-ability of nickel-base alloys. John Wiley andSons Co., New Jersey. 6. Kou, S. 2003. Welding Metallurgy. 2ndEd., John Wiley and Sons Co., New Jersey. 7. Shankar, V., Gill, T. P. S., Mannana, S.L., and Sundaresan, S. 2003. Effect of nitro-gen addition on microstructure and fusionzone cracking in Type 316L stainless steelweld metals. Materials Science and Engineer-ing A 343: 170–181. 8. Nage, D. D., Raja, V. S., and Raman,R. 2006. Effect of nitrogen addition on themicrostructure and mechanical behavior of317L and 904L austenitic stainless steelwelds. Journal of Material Science 41: 2097–2112. 9. Iorio, L., Cortie, M., and Jones, R.1994. Technical note: solubility of nitrogenin experimental low-nickel austeniticstainless steels. Journal of the South AfricaInstitute of Mining and Metallurgy:173–177. 10. Christ, H. J., Chang, S. Y., and Krupp,U. 2003. Thermodynamic characteristics andnumerical modeling of internal nitridationof nickel base alloys. Materials and Corrosion54(11): 887–894. 11. Penna, C. D. 2000. Development ofnew nitrided nickel-base alloys for high tem-perature applications. The Minerals, Metals &

Materials Society: 821–828. 12. Abdulrahman, R .F., and Hendry, A.2001. Solubility of nitrogen in liquid nickel-based alloys. Metallurgical and MaterialsTransactions B 32B: 1103–1112. 13. Kowanda, C., and Speidel, M. O.2003. Solubility of nitrogen in liquid nickeland binary Ni–Xi alloys (Xi ¼ Cr, Mo, W, Mn,Fe, Co) under elevated pressure. Scripta Ma-terialia 48: 1073–1078. 14. Cutler, E. R., Wasson, A. J., andFuchs, G. E. 2008. Effect of minor alloyingadditions on the carbide morphology in asingle crystal Ni-base superalloy. Scripta Ma-terialia 58: 146–149. 15. Ramirez, A. J., and Lippold, J. C.2004. High temperature behavior of Ni-baseweld metal Part II. Insight into the mecha-nism for ductility dip cracking. Materials Sci-ence and Engineering A 380: 245–258. 16. Ramirez, A. J., and Lippold, J. C.2004. High temperature behavior of Ni-baseweld metal Part I. Ductility and microstruc-tural characterization. Materials Science andEngineering A 380: 259–271. 17. Huang, X., Zhang, Y., Liu, Y., and Hu,Z. 1997. Effect of small amount of nitrogenon carbide characteristics in unidirectionalNi-base superalloy. Metallurgical and Mate-rials Transaction A 28A: 2143–2147. 18. Cutler, E. R., Wasson, A. J., andFuchs, G. E. 2009. Effect of minor alloyingadditions on the solidification of single-crys-tal Ni-base superalloys. Journal of CrystalGrowth 311: 3753–3760. 19. Huang, H. 2009. Effects of shieldinggas composition and activating flux onGTAW weldments. Materials and Design 30:2404–2409. 20. Toit, M. 2001. The behavior of nitro-gen during the autogenous arc welding ofstainless steel. PhD thesis, University of Pre-toria, Pretoria. 21. Lin, Y. C., and Chen, P. Y. 2001. Ef-fect of nitrogen content and retained fer-rite on the residual stress in austeniticstainless steel weldments. Materials Scienceand Engineering A 307: 165–171. 22. Tseng, K. H., and Chou, C. P. 2003.The study of nitrogen in argon gas on theangular distortion of austenitic stainlesssteel weldments. J. Mater. Process Technolo-gy 142: 139–44. 23. Murphy, A. B., Tanaka, M., Tashiro,S., Sato, T., and Lowke, J. J. 2009. A compu-tational investigation of the effectiveness ofdifferent shielding gas mixtures for arc weld-ing. Applied Physics 42: 115205 (14 pp). 24. Murphy, A. B., Tanaka, M., Ya-mamoto, K., Tashiro, S., Sato, T., andLowke, J. J. 2009. Modelling of thermalplasmas for arc welding: the role of theshielding gas properties and of metalvapour. Applied Physics 42: 194006 (20 pp). 25. Franke, M. M., Hilbinger, R. M.,Konrad, C. H., Glatzel, U., and Singer, R. F.2011. Numerical determination of second-ary dendrite arm spacing for IN738LC in-vestment castings. Metallurgical and Mate-rials Transactions A 42A: 1847–1853.

WELDING RESEARCH

WELDING JOURNAL / FEBRUARY 2015, VOL. 9460-s

Acknowledgments

References

NABAVI SUPP FEB 2015_Layout 1 1/9/15 3:56 PM Page 60