metals

Article Effect of Post Weld Heat Treatment on the Microstructure and Mechanical Properties of a Submerged-Arc-Welded 304 Stainless

Tae-Hoon Nam 1,2, Eunsol An 1, Byung Jun Kim 1, Sunmi Shin 1, Won-Seok Ko 3, Nokeun Park 4, Namhyun Kang 2,* ID and Jong Bae Jeon 1,*

1 Functional components and Materials R&D Group, Korea Institute of Industrial Technology, Yangsan 50635, Korea; [email protected] (T.-H.N.); [email protected] (E.A.); [email protected] (B.J.K.); [email protected] (S.S.) 2 School of Materials Science and Engineering, Pusan National University, Busan 46241, Korea 3 School of Materials Science and Engineering, University of Ulsan, Ulsan 44610, Korea; [email protected] 4 School of Materials Science and Engineering, Yeungnam University, Gyeongsan 38541, Korea; [email protected] * Correspondence: [email protected] (N.K.); [email protected] (J.B.J.); Tel.: +82-55-367-9407 (N.K.); +82-51-510-3274 (J.B.J.)

Received: 8 November 2017; Accepted: 22 December 2017; Published: 2 January 2018

Abstract: The present study is to investigate the effect of post heat treatment on the microstructures and mechanical properties of a submerged-arc-welded 304 stainless steel. The base material consisted of austenite and long strips of delta-ferrite surrounded by Cr-carbide, and the welds consisted of delta ferrite and austenite matrix. For the heat treatment at 850 ◦C or lower, Cr-carbides were precipitated in the weld metal resulting in the reduction of elongation. The strength, however, was slightly reduced despite the presence of Cr-carbides and this could possibly be explained by the relaxation of internal stress and the weakening of particle . In the heat treatment at 1050 ◦C, the dissolution of Cr-carbide and disappearance of delta ferrite resulted in the lower yield strength and higher elongation partially assisted from deformation-induced martensitic transformation. Consequently, superior property in terms of fracture toughness was achieved by the heat treatment at 1050 ◦C, suggesting that the mechanical properties of the as-weld metal can be enhanced by controlling the post weld heat treatment.

Keywords: austenitic 304 stainless ; sub-merged arc ; post-weld heat treatment

1. Introduction Austenitic stainless steels are widely used in important parts of chemical plants such as pressure vessels, chemical machineries, and pipes because they have excellent corrosion resistance, acid resistance, formability, and weldability. They are also extensively applied in gas turbines, jet propellants, and nuclear power plants because of their excellent toughness in high-temperature and high-pressure environments [1–4]. Submerged Arc Welding (SAW) is widely used for joining ship structures and piping thick-wall stainless steels [5,6]. In SAW, a solid filler wire is automatically inserted to pre-dispersed granular flux and an arc generates between the base metal and the filler wire while being covered with flux. The flux in contact with the arc melts and then the molten slag covers and protects the molten metal. Generally high-efficiency welding is achieved by flowing a large current through a large-diameter wire. It has the advantage of improving welding productivity through high speed solidification during welding and improving weld quality through welding heat control [7].

Metals 2018, 8, 26; doi:10.3390/met8010026 www.mdpi.com/journal/metals Metals 20182018, ,8,8 ,26 26 2 of 13

Due to the heat input from the SAW welding process, there is a degradation problem of microstructuralDue to the and heat mechanical input from properties the SAW such welding as inhomogeneity process, there of the is microstructure, a degradation concentration problem of ofmicrostructural residual stress, and brittleness, mechanical and properties deterioration such of as toughness inhomogeneity in the ofwe thelded microstructure, material [8]. Post-weld concentration heat treatmentof residual (PWHT) stress, brittleness, is usually and introduced deterioration to solve of toughness such degradation. in the welded However, material if the [8]. delta Post-weld ferrite heat in austenitictreatment stainless (PWHT) steel is usually is exce introducedssively present to solve in the such weld degradation. metal during However, SAW welding, if the deltaembrittlement ferrite in ofaustenitic the welded stainless part may steel occur is excessively due to transformation present in the weldinto another metal during brittle SAWphase welding, such as a embrittlement sigma phase duringof the welded post heat part treatment. may occur In dueorder to transformationto solve these problems, into another it is brittle important phase to such investigate as a sigma post-heat phase treatmentduring post process heat treatment.to control the In order content to solveof delta these ferrite problems, and the itharmfu is importantl carbides to investigate[9]. Proper post-heat treatment can process improve to control the mechanical the content properties of delta ferrite by homogenizing and the harmful the carbidesmicrostructure [9]. Proper of the post-heat welded materialtreatment and can controlling improve the the mechanical formation of properties beneficial by and homogenizing harmful phases the [10–13]. microstructure of the welded materialAlthough and controlling extensive theresearches formation have of conducted beneficial and on the harmful effectphases of PWHT [10 –on13 ].the welded austenitic stainlessAlthough steels, extensiveindustrial researches field still haverequires conducted the comprehensive on the effect ofresearch PWHT on on the the resultant welded austenitic effect of PWHT,stainless especially steels, industrial for the field practical still requires and the the cost comprehensive effective process research condition on the resultantof PWHT. effect The of present PWHT, studyespecially thus for investigated the practical the and effect the cost of effectivePWHT on process the change condition of ofmicrostructural PWHT. The present and mechanical study thus propertiesinvestigated of thethe effectwelded of part PWHT of AISI on the STS change 304 st ofeel microstructural produced by SAW. and mechanicalThe microstructural properties changes of the duringwelded the part heat of AISI treatment STS 304 were steel analyzed produced and by the SAW. resultant The microstructural effect on mechanical changes properties during the were heat discussed.treatment were analyzed and the resultant effect on mechanical properties were discussed.

2. Materials Materials and Methods In this study, submerged-arcsubmerged-arc welding was performedperformed in austenitic stainlessstainless steel STS304 using STS308L as as a a welding welding electrode. electrode. The The chemical chemical compos compositionsitions of the of materials the materials used usedare shown are shown in Table in 1.Table A plate1. A with plate the with width the of width 100 mm, of 100 the mm, length the of length 1000 mm of 1000 and mmthe thickness and the thickness of 25 mm ofwas 25 fabricated mm was byfabricated the SAW. by the In SAW.order In to order evaluate to evaluate the micros the microstructuretructure and and mechanical mechanical properties, properties, PWHT PWHT was was ◦ ◦ performed from 650–1050 °CC with an interval of 200 °CC as shown in Figure1 1a.a. AfterAfter maintainingmaintaining atat these temperatures for 0.5–4 h, the specimensspecimens were thenthen quenchedquenched inin water.water. In order order to to observe observe the the microstructure, microstructure, the the test test pieces pieces were were grinded grinded using using sandpapers sandpapers and then and werethen micro-polished were micro-polished using a using diamond a diamond suspension suspension with a 1 μ withm diameter. a 1 µm According diameter. to AccordingASTM E-340, to theASTM polished E-340, surface the polished was etched surface using was a etchedmixed usingsolution a mixed of distilled solution water, of distilledhydrochloric water, acid hydrochloric and nitric acid with and nitrica ratio acid of 4:3:3. with Through a ratio ofthe 4:3:3. optical Through microscopy, the optical the microstructures microscopy, the of the microstructures base metal, the of weldthe base part metal,and the the heat weld affected part zone and (HAZ) the heat were affected analyzed. zone According (HAZ) were to ASTM analyzed. E-8, subsized According plate- to typeASTM tensile E-8, subsized specimens plate-type were taken tensile from specimens the middle were of taken the fromwelded the plate middle with of theperpendicular welded plate to with the weldingperpendicular direction to the as weldingshown in direction Figure 1b. as shown in Figure1b.

Figure 1. Schematic diagrams showing (a) post-weld heat treatment schedules and (b) ASTM-E8 Figure 1. Schematic diagrams showing (a) post-weld heat treatment schedules and (b) ASTM-E8 subsized tensile specimen. subsized tensile specimen.

The tensile tests were conducted to measure yield strength, tensile strength and elongation with −3 the strain rate of 1010−3/s/s according according to to ASTM ASTM A370 A370 using using universal universal tensile tensile machine machine (MTS, (MTS, Eden Eden Praire, Praire, MN, USA). The yield strength was determined by the 0.2% offset method and the elongation was measured by the travel distance of crosshead. Vickers hardness test was conducted with the load of

Metals 2018, 8, 26 3 of 13

MN, USA). The yield strength was determined by the 0.2% offset method and the elongation was measured by the travel distance of crosshead. Vickers hardness test was conducted with the load of 2.94 N for 15 s dwell time and the hardness values were measured 10 times in total, and the average value was calculated by excluding the maximum and the minimum value. To measure the content of delta ferrite in the weld metal, the mean value was measured 10 times per specimen using FERITSCOPE-FMP30 (Fischer, Windsor, CT, USA). Electron backscattered diffraction (EBSD) and Electron Probe Microanalysis (EPMA) was performed to observe the phase change after post heat treatment using JEOL FE-SEM 7200F (JEOL, Tokyo, Japan). For thermodynamic calculations to predict the volume fraction of equilibrium phase at given temperatures, ThermoCalc software (Thermo-Calc Software, 2017b, Solna, Sweden) package was adopted with the latest steel database (TCFE9).

Table 1. Chemical composition of the base metal and the filler wire (wt %).

Alloys C Mn Si Cr Ni Mo Al Co Nb Cu N STS 304 (Base) 0.046 1.19 0.42 18.2 8.02 0.15 0.003 0.164 0.01 0.23 0.05 STS 308L (Filler) 0.02 1.98 0.41 19.7 10.79 0.03 - - - 0.13 0.05

3. Results and Discussion Figure2 shows the microstructure of weld metal, heat-affected zone (HAZ) and base metal after sub-merged arc welding (SAW). The base metal consisted of austenite single phase and hot-rolling strips. These rolling strips correspond to delta ferrite surrounded by Cr-based carbides precipitated during hot rolling (Figure2d–f). The weld metal was composed of austenite (white) and delta ferrite (black). The delta ferrite is classified into lathy and vermicular type by shape (white circle and rectangle in Figure2c, respectively), and this shape difference is considered to be caused by welding heat input. In the sequential order of multi-pass SAW, the welded structure initially appeared in a lathy shape, and then the lathy shape was transformed into a twisty vermicular shape induced by the next welding pass. From the results of the Energy Dispersive Spectroscopy (EDS) analysis, it was confirmed that the formation of the primary Cr-carbide did not occur in the weld metal. Scanning electron microscope (SEM) analysis revealed that pores and cracks possibly formed during solidification were hardly found in the weld metal. Therefore, it was confirmed that the welding conditions such as welding power and electrodes were correctly selected. The heat affected zone of the austenitic stainless steel is generally prone to be Cr-deficient due to the precipitation of Cr-Fe rich carbides at phase boundaries or grain boundaries causing brittleness and welding decay [14]. From the EPMA analysis in Figure2d, there was no remarkable increase of Cr-carbide in HAZ, compared to the base metal where Cr-carbides initially present around long-elongated delta ferrite. Also the shape of Cr-carbides near HAZ still remained in a long elongated shape, which means that those carbides were likely to be formed during hot-rolling process. It is thus considered that the formation of Cr-carbide by welding heat effect hardly occurred in the present specimen. Figure3a,b show Vickers hardness measured over base metal, HAZ and welding metal. The hardness value is considered not to vary significantly over those regions despite that the hardness increased at the regions where austenite and carbides existing together. The hardness of the base material and the HAZ were similar with each other and the hardness of the weld metal was slightly lower than those of the base metal and the HAZ. The hardness results provide that microstructural changes such as grain coarsening/refining or Cr-carbides precipitation barely occurs in the HAZ. The as-welded specimens were always fractured in the middle of the weld metal, not in the HAZ during tensile tests. This also indicates that there is almost no formation of brittle phases in the HAZ. Therefore, the analysis of microstructural changes and mechanical properties during post weld heat treatment is focused on the weld metal, not on the HAZ. Metals 2018, 8, 26 4 of 13 Metals 2018, 8, 26 4 of 13 Metals 2018, 8, 26 4 of 13

FigureFigure 2. Optical2. Optical micrographs micrographs of of the the as-weld as-weld specimen specimen showing showing (a) the( abase) the metal, base (b metal,) the heat- (b) the heat-affectedaffectedFigure zone, zone, 2. Optical and and (c) ( micrographscthe) the weld weld metal metalof (White the (Whiteas-weld arrows arrows specimen and triangle and showing triangless are indicating (a) are the indicating base the pre-existing metal, the (b pre-existing) the strips heat- and delta ferrite, respectively.); (d) Electron Probe Microanalysis (EPMA) images showing the strips andaffected delta zone, ferrite, and respectively.); (c) the weld metal (d) (White Electron arrows Probe and Microanalysis triangles are indicating (EPMA) the images pre-existing showing strips the presenceand delta of Cr-carbides ferrite, respectively.); in the base metal; (d) Electron(e) Electron Probe backscattered Microanalysis diffr action(EPMA) (EBSD) images phase showing map on the presence of Cr-carbides in the base metal; (e) Electron backscattered diffraction (EBSD) phase map on thepresence fusion line of Cr-carbidesincluding base in the metal base (right) metal; and (e) Electronheat affected backscattered zone (HAZ) diffr (left).action Red, (EBSD) blue, phase and cyanmap on the fusion line including base metal (right) and heat affected zone (HAZ) (left). Red, blue, and cyan colorthe indicate fusion linedelta-ferrite, including austenite, base metal and (right) Cr23C and6, respectively. heat affected zone (HAZ) (left). Red, blue, and cyan color indicate delta-ferrite, austenite, and Cr23C6, respectively. color indicate delta-ferrite, austenite, and Cr23C6, respectively.

Figure 3. (a) Optical micrographs indicating the indenting positions; (b) Vicker hardness measured throughFigure the 3. base(a) Optical metal, micrographsthe HAZ, and indicating the weld metal.the indenting positions; (b) Vicker hardness measured Figure 3. (a) Optical micrographs indicating the indenting positions; (b) Vicker hardness measured through the base metal, the HAZ, and the weld metal. throughFigure the base4 shows metal, the the microstructure HAZ, and the of weld the base metal. material and the HAZ after heat-treated at 650– 850 °C forFigure 0.5–4 4 h.shows The size the ofmicrostructure the austenite ofgrain the boundariesbase material of theand base the HAZmaterial after and heat-treated the HAZ is at 40– 650– Figure 850 °C4 showsfor 0.5–4 the h. The microstructure size of the austenite of the grain base boundaries material of and the base the HAZmaterial after and heat-treatedthe HAZ is 40– at 650–850 ◦C for 0.5–4 h. The size of the austenite grain boundaries of the base material and the HAZ is 40–60 µm on average, which is not significantly changed, compared to the as-weld specimen. MetalsMetals 20182018,, 88,, 2626 55 ofof 1313

60 μm on average, which is not significantly changed, compared to the as-weld specimen. The strips Thealready strips existing already in existing the base in material the base still material remained, still remained, without withoutnoticeable noticeable changes changesin the shape in the and shape the ◦ ◦ andvolume the volume fraction fraction even after even the after heat the treatment heat treatment at 650 at °C. 650 In C.the In 850 the °C 850 heatC heattreatment, treatment, however, however, the thevolume volume of the of thestrips strips decreased decreased and and their their shapes shapes were were blurred blurred as compared as compared with with the theas-weld as-weld and and the ◦ theheat-treated heat-treated specimens specimens at at650 650 °C.C. It Itcan can be be infe inferredrred that that the the Cr-based Cr-based carbide carbide was decomposed intointo ◦ anan austeniteaustenite phasephase atat 850850 °C.C.

Figure 4. Optical micrographs on the base metal and the HAZ of the heat-treated specimens at (a) Figure 4. Optical micrographs on the base metal and the HAZ of the heat-treated specimens at (a) 650 ◦C 650 °C for 0.5 h, (b) 650 °C for 1 h, (c) 650 °C for 4 h, (d) 850 °C for 0.5 h, (e) 850 °C for 1 h, and (f) for 0.5 h, (b) 650 ◦C for 1 h, (c) 650 ◦C for 4 h, (d) 850 ◦C for 0.5 h, (e) 850 ◦C for 1 h, and (f) 850 ◦C for 850 °C for 4 h. Arrows indicating pre-existing strips of delta ferrite and Cr-carbide in the HAZ, 4 h. Arrows indicating pre-existing strips of delta ferrite and Cr-carbide in the HAZ, respectively. respectively.

ItIt hashas beenbeen reportedreported thatthat Cr-basedCr-based carbidecarbide isis precipitatedprecipitated andand aa deltadelta phasephase isis transformedtransformed intointo ◦ aa sigmasigma phasephase inin thethe temperaturetemperature rangerange ofof 600–850600–850 °CC inin 300-series300-series austeniticaustenitic stainlessstainless steelssteels [[11,15].11,15]. The results of the thermodynamic calculations in Figure5 show that at 650 ◦C, about 1% of the The results of the thermodynamic calculations in Figure 5 show that at 650 °C, about 1% of the Cr23C6 Crand23C 8%6 and of the 8% sigma of the sigmaphase phasecan be canprecipitated be precipitated in equilibrium in equilibrium from fromthe base the metal base metal composition. composition. This This Cr C is known to be precipitated in the austenite grain boundaries where segregated chromium Cr23C6 23is known6 to be precipitated in the austenite grain boundaries where segregated chromium atoms tend to be bound with carbon atoms. Therefore, the formation of Cr C makes chromium atoms tend to be bound with carbon atoms. Therefore, the formation of Cr2323C66 makes chromium concentrationconcentration inin thethe graingrain boundariesboundaries lowered,lowered, i.e.,i.e., chromiumchromium depletion,depletion, whichwhich resultsresults inin thethe lessless formationformation ofof thethe Cr-basedCr-based oxidesoxides andand thusthus providesprovides harmfulharmful effecteffect onon corrosioncorrosion propertiesproperties andand toughnesstoughness ofof graingrain boundaries.boundaries. OnOn thethe otherother hand,hand, sigmasigma phasephase ofof aatetragonal tetragonal structurestructure composedcomposed ofof Fe-Cr-Ni-MoFe-Cr-Ni-Mo isis generallygenerally reportedreported to to be be precipitated precipitated in in ferrite/austenite ferrite/austenite phasephase boundariesboundaries andand toto bebe growngrown intointo ferriteferrite phasephase [[16,17].16,17]. InIn thethe presentpresent thermodynamicthermodynamic calculations,calculations, thethe sigmasigma phasephase waswas expectedexpected toto havehave aa highhigh fractionfraction ofof 8%,8%, whichwhich couldcould bebe regardedregarded asas thethe muchmuch higherhigher valuevalue thanthan thethe actualactual experimentalexperimental value.value. TheThe reasonreason forfor thisthis overestimationoverestimation isis thatthat thethe thermodynamicthermodynamic calculationcalculation ◦ doesdoes notnot consider kinetics kinetics of of phas phasee transformation; transformation; alpha alpha ferrite ferrite was was calculated calculated to exist to exist in 650 in °C 650 fromC fromthe present the present thermodynamic thermodynamic calculation calculation but was but wasnot actually not actually present present in th ine thebase base metal metal where where the thesingle single austenite austenite phase phase only only existed. existed. In the In base the basemeta metal,l, the preferred the preferred nucleation nucleation site of site the ofsigma the sigmaphase, phase,i.e., ferrite, i.e., ferrite, is hardly is hardly existing, existing, and thus and the thus actual the actualvolume volume of the ofsigma the sigmaphase phasecould couldbe much be muchlower ◦ lowerthan the than calculated the calculated predictions. predictions. On the On other the hand, other hand,in the incase the of case the ofheat the treatment heat treatment at 850 at°C, 850 it wasC, it was predicted that 0.4% of Cr C phase would be precipitated and the sigma phase could not predicted that 0.4% of Cr23C6 phase23 6 would be precipitated and the sigma phase could not be be precipitated. The volume fraction of Cr C decreased with increasing temperature up to 900 ◦C precipitated. The volume fraction of Cr23C623 decreased6 with increasing temperature up to 900 °C above abovewhich which the Cr-carbide the Cr-carbide would would not be not precipitated. be precipitated.

Metals 2018, 8, 26 6 of 13 Metals 2018, 8, 26 6 of 13

Figure 5. Predicted volume fraction of equilibrium phases at given temperatures in (a,b) the base Figure 5. Predicted volume fraction of equilibrium phases at given temperatures in (a,b) the base metal metal (STS 304) and (c,d) the filler metal (STS 308L). (STS 304) and (c,d) the filler metal (STS 308L). Figure 6 shows the microstructure of the weld metal heat-treated at 650–850 °C for 0.5 and 4 h. ◦ SimilarFigure to the6 shows as-weld the microstructure, microstructure the of the austenit welde metal matrix heat-treated contained atdelta 650–850 ferrite Cdendrites. for 0.5 and After 4 h. heatSimilar treatment to the as-weld at 650–850 microstructure, °C, the lathy the type austenite delta fe matrixrrite changed contained to deltavermicular ferrite dendrites.type due to After the heat ◦ treatment and at 650–850 the size ofC, the the individual lathy type dendrite delta ferrite of delta changed ferrite to increased vermicular at 850 type °C due in comparison to the heat ◦ withtreatment 650 °C. and The the phase size of boundary the individual between dendrite austenite of delta and ferritedelta ferrite increased is considered at 850 C into comparisonmigrate in favor with ◦ of650 reducingC. The phasethe interfacial boundary energy. between The austenite initially andthin deltadendrites ferrite of is delta considered ferrite could to migrate thus inbefavor thicker of duringreducing heat the treatment interfacial energy.without The losi initiallyng their thin total dendrites volume offraction delta ferrite as confirmed could thus in beFigure thicker 6e,f. during The fractionheat treatment of delta without ferrite was losing measured their total to be volume slightly fraction decreased as confirmed during heat in Figuretreatment6e,f. at The 650–850 fraction °C. ◦ Figureof delta 7 ferriteshows wasthe volume measured fractional to be slightly change decreased of delta ferrite during with heat temperature treatment and at 650–850 time obtainedC. Figure from7 theshows magnetic the volume induction fractional method. change Because of delta possible ferrite ph withases temperaturesuch as austenite, and time Cr-carbide, obtained and from sigma the phasemagnetic are inductionall paramagnetic, method. only Because ferromagnetic possible phases ferrite such reacts as austenite,to the magnetic Cr-carbide, induction, and sigma so that phase the surfaceare all paramagnetic, volume fraction only of ferromagneticferrite can be precisely ferrite reacts measured to the magnetic[18]. At 650–850 induction, °C, the so that ferrite the fraction surface ◦ ofvolume the weld fraction metal of was ferrite reduced can be by precisely 1% compared measured to [the18]. as-weld At 650–850 specimen.C, the From ferrite the fraction thermodynamic of the weld calculationmetal was reducedresults shown by 1% in compared Figure 5c, to a the small as-weld amou specimen.nt of ferrite From has been the thermodynamic transformed into calculation austenite becauseresults shown the Ae3 in temperature Figure5c, a is small less than amount 650 of°C. ferrite However, has been as observed transformed from intooptical austenite photographs because in ◦ Figurethe Ae3 6,temperature the volume isfraction less than variation 650 C. of However, delta ferrite as observed seems to frombe insignificant optical photographs but rather in the Figure size of6, individualthe volume delta fraction ferrite variation is likely of to delta become ferrite larg seemser as to temperature be insignificant increases but rather from the650 sizeto 850 of individual°C. delta ferrite is likely to become larger as temperature increases from 650 to 850 ◦C.

Metals 2018, 8, 26 7 of 13 Metals 2018, 8, 26 7 of 13 Metals 2018, 8, 26 7 of 13

Figure 6. Optical micrographs (OM) on the weld metal of the heat-treated specimens for (a) 0.5 h at Figure 6. Optical micrographs (OM) on the weld metal of the heat-treated specimens for (a) 0.5 h at 650Figure °C, (6.b )Optical 4 h at 650 micrographs °C, (c) 0.5 h (OM) at 850 on °C, the and weld (d) me4 htal at 850of the °C, heat-treated respectively. specimens To show thefor thickening(a) 0.5 h at 650 ◦C, (b) 4 h at 650 ◦C, (c) 0.5 h at 850 ◦C, and (d) 4 h at 850 ◦C, respectively. To show the thickening of650 delta °C, (ferriteb) 4 h atdendrite, 650 °C, (markersc) 0.5 h at indicate 850 °C, the and thickn (d) 4 hess at of850 delta °C, respectively.ferrite dendri Tote showfor the the heat-treated thickening of delta ferrite dendrite, markers indicate the thickness of delta ferrite dendrite for the heat-treated specimensof delta ferrite for ( edendrite,) 4 h at 650 markers °C and indicate (f) 4 h at the 850 thickn °C. ess of delta ferrite dendrite for the heat-treated specimens for (e) 4 h at 650 ◦C and (f) 4 h at 850 ◦C. specimens for (e) 4 h at 650 °C and (f) 4 h at 850 °C.

Figure 7. Volume fraction of delta ferrite in the weld metal measured by magnetic induction method. Figure 7. Volume fraction of delta ferrite in the weld metal measured by magnetic induction method. Figure 8 is the EBSD phase map that shows the presence of Cr23C6 in the weld after at 23 6 650 °CFigure for 4 h. 8 isThe the Cr EBSD23C6, which phase didmap not that exist shows in the the as-weld presence specimen, of Cr C was in thefound weld to afterprecipitate annealing finely at Figure8 is the EBSD phase map that shows the presence of Cr 23C6 in the weld after annealing 23 6 in650 the °C austenite/delta-ferrite ◦for 4 h. The Cr C , which interface did not after exist 4 h in of the heat as-weld treatment specimen, at 650 was °C. Crfound23C6 tocarbide precipitate is reported finely at 650 C for 4 h. The Cr23C6, which did not exist in the as-weld specimen, was found to precipitate in the austenite/delta-ferrite interface after 4 h of heat treatment at 650 °C. Cr23C6 carbide is reported tofinely precipitate in theaustenite/delta-ferrite from austenite/ferrite interfacephase boun afterdary 4 h ofor heataustenite treatment grain at boundary 650 ◦C. Cr in C300-seriescarbide to precipitate from austenite/ferrite phase boundary or austenite grain boundary in23 300-series6 austeniticis reported stainless to precipitate steels [11,15,16]. from austenite/ferrite After precipitation, phase it boundary grows into or austenite matrix grain boundaryrather than in ferriteaustenitic matrix. stainless It is knownsteels [11,15,16]. that the precipitates After precipitation, are finely it locatedgrows intoin the austenite boundaries matrix with rather the thansize distributionferrite matrix. from It isseveral known hundred that the nm precipitates to few μm, are depending finely located on the in composition the boundaries and thewith formation the size temperaturedistribution from[15]. severalAs can hundredbe seen nmfrom to thefew therμm,modynamic depending oncalculations the composition shown and in Figurethe formation 5, the temperature [15]. As can be seen from the thermodynamic calculations shown in Figure 5, the

Metals 2018, 8, 26 8 of 13

300-series austenitic stainless steels [11,15,16]. After precipitation, it grows into austenite matrix rather than ferrite matrix. It is known that the precipitates are finely located in the boundaries with the size distribution from several hundred nm to few µm, depending on the composition and the Metals 2018, 8, 26 8 of 13 formation temperature [15]. As can be seen from the thermodynamic calculations shown in Figure5, ◦ temperaturethe temperature range range near near 650 650 °C couldC could have have 0.5% 0.5% volume volume fraction fraction of of Cr Cr2323CC6,6 which, which is is the the maximum maximum value in whole temperature range. Although Although it it is is a a small volume fraction, it it is presumed that if the precipitates are finely finely distributeddistributed along the phase boundaries, it could deteriorate the ductility and toughnesstoughness of of the the material. In In the the thermodynamic thermodynamic calc calculation,ulation, it it was was predicted predicted that about 12% of the ◦ sigma phase would be be precipitated in in the the weld metal at at 650 °C.C. However, However, the EBSD results hardly showed the presence of sigma phase. It It is is believed th thatat the transformation kinetics of sigma phase is relatively slow, so that sigma phase could not be transformed transformed from from ferrite ferrite with with the the given given temperature temperature and time. Even Even if if sufficient sufficient time time is is provided, provided, it is expected expected that the the sigma sigma phase phase would would not not be be present present as much as the predicted value since the volume fracti fractionon of ferrite, which is the preferred preferred nucleation site of the sigma phase, is relatively small in the actual weld metal.

Figure 8. (a) EBSD map showing presence of Cr23C6 at phase boundary; (b) OM image after color- Figure 8. (a) EBSD map showing presence of Cr23C6 at phase boundary; (b) OM image after etching showing the presence of Cr23C6 (dark). color-etching showing the presence of Cr23C6 (dark).

Figure 9a–d show the microstructures of the base material and HAZ of the specimens annealed Figure9a–d show the microstructures of the base material and HAZ of the specimens annealed at 1050 °C for 0.5–4 h. The grain size of the base metal after PWHT in 1050 °C for 4 h became larger at 1050 ◦C for 0.5–4 h. The grain size of the base metal after PWHT in 1050 ◦C for 4 h became (72 μm) than the as-weld specimen (52 μm), which shows the grain growth of austenite with larger (72 µm) than the as-weld specimen (52 µm), which shows the grain growth of austenite with increasing time and temperature. This grain growth is believed to occur because of the acceleration increasing time and temperature. This grain growth is believed to occur because of the acceleration of of thermally-assisted migration of the austenite grain boundary. The strip already existing in the base thermally-assisted migration of the austenite grain boundary. The strip already existing in the base metal was observed to disappear as the heat treatment time became longer. As can be seen from the metal was observed to disappear as the heat treatment time became longer. As can be seen from the thermodynamic calculation in Figure 5, the Cr-based carbide was considered to transform into the thermodynamic calculation in Figure5, the Cr-based carbide was considered to transform into the austenite phase at 1050 °C since the dissolution temperature of the Cr-based carbide is about 900 °C. austenite phase at 1050 ◦C since the dissolution temperature of the Cr-based carbide is about 900 ◦C. Figure 9e–h shows the microstructure of welded specimens after heat treatment at 1050 °C for Figure9e–h shows the microstructure of welded specimens after heat treatment at 1050 ◦C 0.5–4 h. The delta ferrite, which existed as lathy or vermicular shape in the as-weld specimen, lost its for 0.5–4 h. The delta ferrite, which existed as lathy or vermicular shape in the as-weld specimen, shape into the elliptical or the spheroidized after the heat treatment of 1050 °C. This is due to the lost its shape into the elliptical or the spheroidized after the heat treatment of 1050 ◦C. This is due to migration of austenite/delta-ferrite interface for decreasing the interfacial energy, similar to the the migration of austenite/delta-ferrite interface for decreasing the interfacial energy, similar to the mechanism of the shape change from the lathy to the vermicular type observed in the heat treatment mechanism of the shape change from the lathy to the vermicular type observed in the heat treatment at 650–850 °C. The volume fraction of delta ferrite at 1050 °C decreased by heat treatment time as at 650–850 ◦C. The volume fraction of delta ferrite at 1050 ◦C decreased by heat treatment time as confirmed in Figure 7 where delta-ferrite, which was 8% of volume fraction in the as-weld specimen confirmed in Figure7 where delta-ferrite, which was 8% of volume fraction in the as-weld specimen decreased down to 1.2% after heat treatment at 1050 °C for 4h. From the thermodynamic point of decreased down to 1.2% after heat treatment at 1050 ◦C for 4h. From the thermodynamic point view, the transformation of delta ferrite to austenite was believed to be promoted at 1050 °C, well of view, the transformation of delta ferrite to austenite was believed to be promoted at 1050 ◦C, above the Ae3 temperature. From the thermodynamic calculations, austenite single phase exists at well above the A temperature. From the thermodynamic calculations, austenite single phase exists at 1050 °C, without e3Cr-carbide or delta-ferrite phase, and well agree with the experimental observations. 1050 ◦C, without Cr-carbide or delta-ferrite phase, and well agree with the experimental observations. The residual delta-ferrite, which was not transformed into austenite by the rapid cooling after the The residual delta-ferrite, which was not transformed into austenite by the rapid cooling after welding, was gradually austenized by heat treatment at 1050 °C. In addition, at that temperature, the welding, was gradually austenized by heat treatment at 1050 ◦C. In addition, at that temperature, residual stress due to welding could be released and carbides could dissolve sufficiently, which could possibly lead to noticeable changes in the mechanical properties.

Metals 2018, 8, 26 9 of 13 residual stress due to welding could be released and carbides could dissolve sufficiently, which could possiblyMetals 2018 lead, 8, 26 to noticeable changes in the mechanical properties. 9 of 13

Figure 9. Optical micrographs on the base metal and the HAZ of the heat-treated specimens at 1050 °C Figure 9. Optical micrographs on the base metal and the HAZ of the heat-treated specimens at for (a) 0.5 h, (b) 1 h, (c) 2 h, and (d) 4 h, respectively. Optical micrographs on the weld metal of the 1050 ◦C for (a) 0.5 h, (b) 1 h, (c) 2 h, and (d) 4 h, respectively. Optical micrographs on the weld heat-treated specimens at 1050 °C for (e) 0.5 h, (f) 1 h, (g) 2 h, and (h) 4 h, respectively. Arrows and metal of the heat-treated specimens at 1050 ◦C for (e) 0.5 h, (f) 1 h, (g) 2 h, and (h) 4 h, respectively. triangles indicate the pre-existing strips in the base/HAZ and delta ferrite in the weld metal, Arrows and triangles indicate the pre-existing strips in the base/HAZ and delta ferrite in the respectively. weld metal, respectively.

Figure 10 summarized the yield and tensile strength, uniform and total elongation, and fracture toughnessFigure of 10 specimens summarized after the PWHT. yield and The tensile yield strength,and tensile uniform strengths and in total all elongation,range of heat and treatment fracture temperaturestoughness of specimensand times were after lower PWHT. than The those yield of andthe as-weld tensile strengths specimen. in The all yield range strength of heat treatmentgradually decreasedtemperatures as the and temperature times were increased lower than while those there of the was as-weld no significant specimen. change The yield depending strength on gradually the time. Thedecreased tensile as strength the temperature of the specimens increased heat-treated while there at was 650–850 no significant °C was 570–580 change depending MPa and a on noticeable the time. ◦ Thedifference tensile was strength hardly of found the specimens depending heat-treated on the heat at 650–850treatmentC time. was 570–580In the case MPa of andthe heat a noticeable treated specimendifference at was 1050 hardly °C, the found tensile depending strength was on the ma heatrkedly treatment lower and time. continuously In the case decreasing of the heat with treated the ◦ time.specimen As atshown 1050 inC, Figure the tensile 10b, strengththe uniform wasmarkedly and total lower elongation and continuously decreased decreasingwith increasing with temperaturethe time. As up shown to 850 in°C, Figure compared 10b, with the uniformthe as-wel andd specimen. total elongation On the other decreased hand, withthe specimen increasing at ◦ 1050temperature °C showed up tohigher 850 C,elongation compared than with the the as-wel as-weldd specimen, specimen. and On prolonged the other hand, its elongation the specimen as the at heat treatment time increased.

Metals 2018, 8, 26 10 of 13

1050 ◦C showed higher elongation than the as-weld specimen, and prolonged its elongation as the heatMetals treatment 2018, 8, 26 time increased. 10 of 13

Figure 10.10. ((aa)) YieldYield and and tensile tensile strength strength of of the the as-weld as-weld and and the the heat-treated heat-treated specimens; specimens; (b) uniform(b) uniform and totaland total elongation elongation of the of as-weld the as-weld and theand heat-treated the heat-treated specimens; specimens; (c) fracture (c) fracture toughness toughness of the of as-weld the as- andweld the and heat-treated the heat-treated specimens; specimens; (d) Vickers (d) hardnessVickers ofhardness the as-weld of the and as-weld the heat-treated and the specimens.heat-treated specimens. These changes of mechanical properties can be understood based on the microstructural evolution duringThese heat changes treatment. of mechanical In the case properties of reduction can in be strength understood and ductilitybased on shown the microstructural at 650–850 ◦C, relaxationevolution ofduring internal heat stress, treatment. decreased In the volume case fractionof reduction of delta in ferritestrength and and increased ductility size shown of individual at 650– delta850 °C, ferrite relaxation dendrite of internal could be stress, responsible decreased for suchvolu degradation.me fraction of As delta shown ferrite in Figureand increased8, the carbides size of wereindividual found delta to form ferrite in thedendrite austenite/ferrite could be responsible phase boundaries. for such degradation. It is expected As thatshown material in Figure strength 8, the wouldcarbides increase were found by compromising to form in the ductility austenite/ferrite with the formation phase boundaries. of Cr-carbide, It butis expected inversely that the material strength decreasedstrength would at the increase corresponding by compromising temperature. ductility This reduction with the information strength, of first Cr-carbide, of all, could but inversely probably bethe explained strength decreased by the stress at the relaxation corresponding due to temperatur the recoverye. ofThis the reduction hot-rolled in microstructure strength, first of in all, the could base metalprobably and be by explained the relaxation by the of stress the internal relaxation stress due in theto the base recovery and the of weld the hot-rolled metal initially microstructure caused by the in weldingthe base processmetal and [19 ].by Second, the relaxation as shown of in the Figure internal7, the stress slight in decrease the base in and volume the fractionweld metal of the initially delta ferritecaused in by the the weld welding metal process could possibly [19]. Second, cause theas show strengthn in reduction Figure 7, in the terms slight of modulusdecrease hardening.in volume Third,fraction as of can the be delta seen ferrite in Figure in the6, when weld the metal individual could possibly size of thecause delta the ferrite strength increased reduction without in terms losing of itsmodulus volume hardening. fraction substantially, Third, as can the be strength seen in canFigure be reduced6, when owingthe individual to the weakening size of the of delta dispersion ferrite strengtheningincreased without or precipitation losing its volume strengthening fraction substantially, [20]. Stochastically, the strength the probability can be reduced of the owing existence to the of dislocationweakening sourcesof dispersion is higher strengthening in the austenite or matrixprecipitation whose volumestrengthening fraction [20]. and sizeStochastically, is much higher the thanprobability delta ferrite. of the Givenexistence that of austenite dislocation has generally sources is lower higher yield in strengththe austenite than ferrite,matrix thewhose operation volume of dislocationfraction and is size likely is tomuch initiate higher in the than austenite delta matrix.ferrite. TheGiven dislocation that austenite movement has generally in the austenite lower couldyield thenstrength be impeded than ferrite, by stiff the deltaoperation ferrite of arrays dislocation and if theis likely individual to initiate size ofin thethe deltaaustenite ferrite matrix. increased, The particledislocation strengthening movement couldin the be austenite degraded. could then be impeded by stiff delta ferrite arrays and if the individualOn the size other of the hand, delta in theferrite case increased, of heat treatment particle strengthening above 1050 ◦ C,could the yieldbe degraded. and tensile strengths wereOn lower the butother elongations hand, in the were case higher of heat than treatm theent as-weld above 1050 specimen. °C, the Asyield shown and tensile in Figure strengths8, the decreasewere lower in thebut volume elongations fraction were of delta higher ferrite than at the 1050 as-weld◦C is considered specimen. to As be theshown cause in ofFigure the strength 8, the reduction.decrease in the Since volume the thickness fraction of of delta delta ferrite ferrite at was 1050 small °C is considered as about 1–2 to µbem, the it cause is believed of the strength that the reduction. Since the thickness of delta ferrite was small as about 1–2 μm, it is believed that the delta ferrite is strengthening the weld metal in terms of modulus hardening and particle hardening. Therefore, decreasing the fraction of delta ferrite may have affected the decrease of strength and increase of elongation inversely. In addition, as shown in Figure 9, the austenite grain size of the base and welds increased during the heat treatment at 1050 °C, and thus Hall-Petch effect due to the

Metals 2018, 8, 26 11 of 13 delta ferrite is strengthening the weld metal in terms of modulus hardening and particle hardening. Therefore, decreasing the fraction of delta ferrite may have affected the decrease of strength and increase of elongation inversely. In addition, as shown in Figure9, the austenite grain size of the base and welds increased during the heat treatment at 1050 ◦C, and thus Hall-Petch effect due to the increase of grain size also has an effect on the reduction in yield and tensile strengths. Furthermore, the stress relaxation due to the heat treatment at 1050 ◦C could make the base and weld metal softer. Figure 10c shows the fracture toughness obtained by integrating stress-strain curves. When heat treatment temperature was in the range of 650–850 ◦C, the fracture toughness was lower than that of the as-weld specimen. When the heat treatment temperature was more than 1050 ◦C, the fracture toughness was in overall similar or higher than the as-weld specimen. In the temperature range studied in the present work, the strengths were reduced as compared with the as-weld specimen, but ductility significantly increased at 1050 ◦C. This ductility increase is reflected in the increase of the fracture toughness, and the PWHT at a temperature of 1050 ◦C was found to provide positive effect on improving the toughness of the as-weld material. Figure 10d shows the hardness variation of the welds with the PWHT temperature and time. After 10 times of indentations, the values were averaged by excluding the maximum and the minimum. Hardness is determined by the magnitude of strengthening under compression, and thus it can be interpreted similarly to the strength. Vickers hardness showed a monotonic decrease with increasing the heat treatment temperature. The hardness change with the heat treatment time was negligible for 1 to 4 h, and temperature was more of an influential factor on the hardness. As described above, reduction of hardness with increasing temperature is also able to be explained by the relaxation of the residual stress, the decrease of delta-ferrite volume fraction and the increase of the individual delta ferrite thickness. Figure 11 contained the true stress-strain curves together with work hardening rate for analyzing the difference in work hardening behaviors after PWHT. In the as-weld specimen, the work hardening rate decreased rapidly after passing yield point and maintained in the steady state. On the other hand, the work hardening rate of the specimen at 650 ◦C for 2 h was generally higher than that of the as-weld specimen, and decreased continuously with further straining after having a short steady-state flow regime. The work hardening rate of PWHT 850 ◦C increased continuously with deformation after reaching the steady state. This is presumably due to the deformation-induced martensitic transformation (DIMT) occurring in both of 304 base metal and 308 weld metal. When the DIMT occurs, the martensite phase is produced with the deformation, and the volume fraction of martensite increases steadily as the deformation continues, leading to the gradual increase of the work hardening rate [18]. The inflection shown in true stress-strain curve is related to the initiation of DIMT. A number of researches reported that the DIMT indeed occurs in 300-series austenitic stainless steels and results in the continuous increase of the work hardening rate during deformation [18,21,22]. At 1050 ◦C, the work hardening rate gradually increased and an inflection was observed in true stress-strain curve. In this temperature, the initially present Cr-carbide arrays would dissolve and thus make austenite matrix denser with carbon atoms. The austenite is thus more stabilized and the DIMT kinetics could be slower, leading to the increase of elongation. Since the delta ferrite already existing in the weld metal was mostly transformed into austenite and also the grain size of austenite increased at 1050 ◦C, the yield strength became lower than the as-weld specimen. Metals 2018, 8, 26 11 of 13 increase of grain size also has an effect on the reduction in yield and tensile strengths. Furthermore, the stress relaxation due to the heat treatment at 1050 °C could make the base and weld metal softer. Figure 10c shows the fracture toughness obtained by integrating stress-strain curves. When heat treatment temperature was in the range of 650–850 °C, the fracture toughness was lower than that of the as-weld specimen. When the heat treatment temperature was more than 1050 °C, the fracture toughness was in overall similar or higher than the as-weld specimen. In the temperature range studied in the present work, the strengths were reduced as compared with the as-weld specimen, but ductility significantly increased at 1050 °C. This ductility increase is reflected in the increase of the fracture toughness, and the PWHT at a temperature of 1050 °C was found to provide positive effect on improving the toughness of the as-weld material. Figure 10d shows the hardness variation of the welds with the PWHT temperature and time. After 10 times of indentations, the values were averaged by excluding the maximum and the minimum. Hardness is determined by the magnitude of strengthening under compression, and thus it can be interpreted similarly to the strength. Vickers hardness showed a monotonic decrease with increasing the heat treatment temperature. The hardness change with the heat treatment time was negligible for 1 to 4 h, and temperature was more of an influential factor on the hardness. As described above, reduction of hardness with increasing temperature is also able to be explained by the relaxation of the residual stress, the decrease of delta-ferrite volume fraction and the increase of the individual delta ferrite thickness. Figure 11 contained the true stress-strain curves together with work hardening rate for analyzing the difference in work hardening behaviors after PWHT. In the as-weld specimen, the work hardening rate decreased rapidly after passing yield point and maintained in the steady state. On the other hand, the work hardening rate of the specimen at 650 °C for 2 h was generally higher than that of the as-weld specimen, and decreased continuously with further straining after having a short steady-state flow regime. The work hardening rate of PWHT 850 °C increased continuously with deformation after reaching the steady state. This is presumably due to the deformation-induced martensitic transformation (DIMT) occurring in both of 304 base metal and 308 weld metal. When the DIMT occurs, the martensite phase is produced with the deformation, and the volume fraction of martensite increases steadily as the deformation continues, leading to the gradual increase of the work hardening rate [18]. The inflection shown in true stress-strain curve is related to the initiation of DIMT. A number of researches reported that the DIMT indeed occurs in 300-series austenitic stainless steels and results in the continuous increase of the work hardening rate during deformation [18,21,22]. At 1050 °C, the work hardening rate gradually increased and an inflection was observed in true stress-strain curve. In this temperature, the initially present Cr-carbide arrays would dissolve and thus make austenite matrix denser with carbon atoms. The austenite is thus more stabilized and the DIMT kinetics could be slower, leading to the increase of elongation. Since the delta ferrite already existingMetals 2018 in, the8, 26 weld metal was mostly transformed into austenite and also the grain size of austenite12 of 13 increased at 1050 °C, the yield strength became lower than the as-weld specimen.

FigureFigure 11. 11. TrueTrue stress-strain stress-strain curves curves and and wo workrk hardening hardening rates rates in in terms terms of of ( (aa)) PWHT PWHT temperature temperature and and ((bb)) PWHT PWHT time. time.

4. Conclusions The effect of PWHT on the microstructure and mechanical properties of submerged-arc-welded 304 austenitic stainless steel was investigated. The results are summarized as follows:

(1) The base metal initially had a microstructure consisting of austenite matrix and strips of delta-ferrite surround by Cr23C6. The weld metal consisted of lathy and vermicular type of delta ferrite and austenite matrix. There was insignificant variation of microstructures and mechanical properties between HAZ and the base metal, implying that heat affection by welding was negligible. (2) Cr-carbides were precipitated at the delta-ferrite/austenite interface in the weld metal after the heat treatment at 650–850 ◦C. Despite of the presence of Cr-carbide, the yield strength was lower than the as-weld specimen, which could be possibly explained by the relaxation of internal stress accumulated during the hot-rolling and the welding. (3) After heat-treatment at 1050 ◦C, the pre-existing strips in the base metal was dissolved into austenite and the delta-ferrite of the weld metal was transformed into austenite, leading to the microstructure close to the austenite single phase. Although the yield strength was lower than that of the as-weld specimen due to the disappearance of hard phases and the relaxation of residual stress, the elongation became higher than that of the as-weld metal due to the compromising effect of strength and partially to the DIMT occurring during deformation. In terms of fracture toughness, the heat treatment at 1050 ◦C were suggested to provide superior property than as-weld material due to the trade-off between strength and ductility.

Acknowledgments: This study has been conducted with the support of the Korea Institute of Industrial Technology as “Development of metal powder large-area DED process including pre- and post- processes based on simulations (KITECH EO-17-0041)” and “Development of Manufacturing Technology of Cryogenic Stainless Steel Welded Steel Pipe (KITECH JG-17-0006)”. Author Contributions: Tae-Hoon Nam and Jong Bae Jeon conceived and designed the experiments; Tae-Hoon Nam and Eunsol An performed the experiments; Byung Jun Kim, Sunmi Shin, Nokeun Park, Won-Seok Ko, Namhyun Kang and Jong Bae Jeon analyzed and discussed the data; Tae-Hoon Nam and Jong Bae Jeon wrote the paper. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Weiss, B.; Stickler, R. Phase instabilities during high temperature exposure of 316 austenitic stainless steel. Metall. Trans. 1972, 3, 851–866. [CrossRef] Metals 2018, 8, 26 13 of 13

2. Chen, X.H.; Lu, J.; Lu, L.; Lu, K. Tensile properties of a nanocrystalline 316L austenitic stainless steel. Scr. Mater. 2005, 52, 1039–1044. [CrossRef] 3. Menthe, E.; Rie, K.T. Further investigation of the structure and properties of austenitic stainless steel after plasma . Surf. Coat. Technol. 1999, 116–119, 199–204. [CrossRef] 4. Altenberger, I.; Scholtes, B.; Martin, U.; Oettel, H. Cyclic deformation and near surface microstructures of shot peened or deep rolled austenitic stainless steel AISI 304. Mater. Sci. Eng. A 1999, 264, 1–16. [CrossRef] 5. Tabatabaeipour, S.M.; Honarvar, F. A comparative evaluation of ultrasonic testing of AISI 316L welds made by shielded metal arc welding and gas tungsten arc welding processes. J. Mater. Process. Technol. 2010, 210, 1043–1050. [CrossRef] 6. Li, X.R.; Zhang, Y.M.; Kvidahl, L. Penetration Depth Monitoring and Control in Submerged Arc Welding. Weld. J. 2013, 92, 48–56. 7. Gunaraj, V.; Murugan, N. Application of response surface methodology for predicting weld bead quality in submerged arc welding of pipes. J. Mater. Process. Technol. 1999, 88, 266–275. [CrossRef] 8. Gunaraj, V.; Murugan, N. Prediction and comparison of the area of the heat-affected zone for the bead-on-plate and bead-on-joint in submerged arc welding of pipes. J. Mater. Process. Technol. 1999, 95, 246–261. [CrossRef] 9. Kumar, S.; Shahi, A.S. Effect of heat input on the microstructure and mechanical properties of gas tungsten arc welded AISI 304 stainless steel joints. Mater. Des. 2011, 32, 3617–3623. [CrossRef] 10. Luo, J.; Dong, Y.; Li, L.; Wang, X. Microstructure of 2205 duplex stainless steel joint in submerged arc welding by post weld heat treatment. J. Manuf. Process. 2014, 16, 144–148. [CrossRef] 11. Hamada, I.; Yamauchi, K. Sensitization behavior of type 308 stainless steel weld metals after postweld heat treatment and low-temperature aging and its relation to microstructure. Metall. Mater. Trans. A 2002, 33, 1743–1754. [CrossRef] 12. Jang, D.; Kim, K.; Kim, H.C.; Jeon, J.B.; Nam, D.-G.; Sohn, K.Y.; Kim, B.J. Evaluation of Mechanical Property for Welded Austenitic Stainless Steel 304 by Following Post Weld Heat Treatment. Korean J. Met. Mater. 2017, 55, 664–670. [CrossRef] 13. Kang, N.-H. Development of Alloy Design and Welding Technology for Austenitic Stainless Steel. J. KWJS 2010, 28, 10–14. [CrossRef] 14. Barbosa Gonçalves, R.; Henrique Dias de Araújo, P.; José Villela Braga, F.; Augusto Hernandez Terrones, L.; Pinheiro da Rocha Paranhos, R. Effect of conventional and alternative solution and stabilizing heat treatment on the microstructure of a 347 stainless steel welded joint. Weld. Int. 2017, 31, 196–205. [CrossRef] 15. Tseng, C.C.; Shen, Y.; Thompson, S.W.; Mataya, M.C.; Krauss, G. Fracture and the formation of sigma phase, M23C6, and austenite from delta-ferrite in an AlSl 304L stainless steel. Metall. Mater. Trans. A 1994, 25, 1147–1158. [CrossRef] 16. Kington, A.V.; Noble, F.W. σ phase embrittlement of a type 310 stainless steel. Mater. Sci. Eng. A 1991, 138, 259–266. [CrossRef] 17. Vitek, J.; David, S. The sigma phase transformation in austenitic stainless steels. Weld. J. 1986, 65, 106s–111s. 18. Shin, H.C.; Ha, T.K.; Chang, Y.W. Kinetics of deformation induced martensitic transformation in a 304 stainless steel. Scr. Mater. 2001, 45, 823–829. [CrossRef] 19. Mújica Roncery, L.; Weber, S.; Theisen, W. Welding of twinning-induced plasticity steels. Scr. Mater. 2012, 66, 997–1001. [CrossRef] 20. Zhang, Z.; Chen, D.L. Consideration of Orowan strengthening effect in particulate-reinforced metal matrix nanocomposites: A model for predicting their yield strength. Scr. Mater. 2006, 54, 1321–1326. [CrossRef] 21. Huang, C.X.; Yang, G.; Deng, B.; Wu, S.D.; Li, S.X.; Zhang, Z.F. Formation mechanism of nanostructures in austenitic stainless steel during equal channel angular pressing. Philos. Mag. 2007, 87, 4949–4971. [CrossRef] 22. Lee, T.-H.; Oh, C.-S.; Kim, S.-J. Effects of nitrogen on deformation-induced martensitic transformation in metastable austenitic Fe–18Cr–10Mn–N steels. Scr. Mater. 2008, 58, 110–113. [CrossRef]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).