constitutional liquation of the laves phase in virgin fb2 steel€¦ · creep-resistant martensitic...

Introduction Elevating steam parameters is themain key to enhancing the efficiencyof fossil power plants to reduce fuelconsumption and noxious emission(Refs. 1, 2). Therefore, a lot of newcreep-resistant martensitic steels havebeen developed in the frame of the Eu-ropean Cooperation in Science andTechnology (COST) program. Amongthem, FB2 steel is the most promisingcandidate to be widely used in turbinerotors operated at a temperature rangefrom 600º to 650ºC (Ref. 3). FB2 steelis a forged, boron-added, and 9% Cr-containing martensitic stainless steel(Ref. 4). Welding is now widely appliedin manufacturing large componentsoperated at high temperatures to over-come the limit of forging capacity(Refs. 5, 6). Up to now, there is still alimited number of reports on welded

FB2 steam turbine rotors, so it is nec-essary to carry out experiments onFB2 steel welding, gaining experiencesto guide the practical production. On the other side, the Laves phaseusually occurs in martensitic stainlesssteels after long-term, high-tempera-ture exposure (Refs. 7–9), but there isstill a limited number of reports aboutthe Laves phase in 9–12% Cr marten-sitic steels in the as-received (or vir-gin) condition (Refs. 10, 11). Amongthe numerous martensitic stainlesssteels, CB2 is a cast steel with a similarchemical composition to FB2 steel,mainly used for steam turbine compo-nents. The occurrence of a micronsized (Fe, Cr)2Mo-type Laves phase invirgin CB2 steel was reported by Kasl(Ref. 11) and Jandová (Ref. 12). How-ever, the evolutionary behavior of theLaves phase during welding and theinfluence of the Laves phase evolu-

tionary behavior on weld joints of9–12% Cr martensitic stainless steelshave not been clarified in publishedarticles. In the present work, some (Fe,Cr)2Mo-type Laves phase was alsofound in virgin FB2 steel, which led to constitutional liquation duringwelding thermal cycles, suggesting a li-quation crack tendency in the heat-affected zone (HAZ) of FB2 steel. Thenthe evolutionary mechanism of theLaves phase during heating and cool-ing was analyzed. This work is impor-tant in guiding the practical produc-tion of welded turbine rotors made ofFB2 steel.

Materials and ExperimentProcedure

The specimens for scanning elec-tron microscopy (SEM) observation ofthe Laves phase were sampled fromvirgin FB2 steel (tempered at 700ºCfor 4 h), of which the chemical compo-sition is listed in Table 1. The specimens were electrolyticetched at 5V in oxalic acid for 10 s fol-lowing mechanical polishing. Weldingthermal cycle simulation was carriedout on dilatometer DIL 805A, andpeak temperatures (Tp) were 1150º,1200º, 1250º, and 1350ºC with a hold-ing time of 0.5 s. Both the heating andcooling rates were 100ºC/s. After ther-mal simulation, all the specimens wereelectrolytic etched followed by SEMobservation. In specimens experienc-ing peak temperatures of 1350ºC,some eutectic structures with net-likemorphology were found, and focusedion beam (FIB) was employed to sam-

WELDING RESEARCH

Constitutional Liquation of the Laves Phase in Virgin FB2 Steel

The evolutionary behavior of the Laves phase during welding was examined

BY K. LI, Z. CAI, Y. LI, AND J. PAN

ABSTRACT In 9–12% Crcontaining martensitic stainless steel, the Laves phase usually occursafter longterm, hightemperature exposure. However, in the present work, some sparse,relatively large particles of the Laves phase were observed in virgin FB2 steel (a new 9%Cr martensitic stainless steel). It is concluded that the large Laves phase particles formedduring casting due to dendritic segregation. Then constitutional liquation resulting froma eutectic reaction between the Laves phase and (Fe) during welding was found,suggesting a liquation crack tendency in FB2 steel. Subsequently, hot ductility testsfurther confirmed the existence of a liquation crack. The evolutionary behavior of theLaves phase during welding thermal simulation was analyzed, and the Chi phase wasfound as a eutectic constituent during cooling. In specimens experiencing a peak temperature of 1350˚C, some grain boundaries were surrounded by eutectic microstructures,forming “ghost boundaries,” in which way the grain boundary strength was severelyweakened.

K. LI, Z. CAI ([email protected]), Y. LI, and J. PAN are with the Department of Mechanical Engineering, Tsinghua University, Beijing, China.CAI is also with the Collaborative Innovation Center of Advanced Nuclear Energy Technology and the State Key Laboratory of Tribology, Tsinghua University, Beijing, China.

JULY 2016 / WELDING JOURNAL 257-s

KEYWORDS • Laves Phase • FB2 Steel • Constitutional Liquation • Liquation Crack • Chi Phase

LI ET AL SUPP 2015107_Layout 1 6/9/16 4:26 PM Page 257

ple a specimen for transmission elec-tron microscopy (TEM) observation ofthe eutectic structures. Selected areaelectron diffraction (SAED) was car-ried out to identify the types of eutec-tic constituents. Hot ductility tests were conductedon a Gleeble-3500 thermal simulator,and protected by an argon atmospherefrom oxidation of the fracture appear-ance. The rod specimens for hot duc-tility tests were 121.5 mm in lengthand 10 mm in diameter. The peak tem-peratures were 1350ºC (labeled 1) and1250ºC (labeled 2), and the displace-ment load was 0.5 mm with a velocityof 0.1 mm/s at peak temperatures.The fracture appearance of rupturedspecimens was observed by SEM.

Results

Observation of the Laves Phasein Virgin FB2 Steel

In virgin FB2 steel, some micron-sized particles were found, as shown inFig. 1. All the particles were distributedrandomly and inhomogeneously, andmost parts of the matrix were free ofthe particles. Based on our experience,the particles could only be found whenthe entire specimen was traversed bySEM with about 1000 magnification.Most of the particles were locatedwithin austenitic grains and were muchlarger than conventional precipitates,such as M23C6 or the MX phase. The results of energy-dispersive x-ray (EDX) show the particles wereenriched in boron, oxygen, molybde-num, vanadium, and chromium com-pared with the nominal chemical com-position of the FB2 steel listed in Table1. The main alloy elements in the par-ticles were molybdenum, chromium,and ferrite, suggesting they may besome kind of intermetallic compound.In addition, the mole ratio of (Fe + Cr)and Mo in the particles is nearly 2. Allof the characteristics mentioned were

WELDING RESEARCH

WELDING JOURNAL / JULY 2016, VOL. 95258-s

Table 1 — Chemical Composition of FB2 Steel (wt%)

C Si Mn Ni Cr Mo V Nb N Co B Fe

0.13 0.05 0.40 0.15 9.3 1.5 0.2 0.05 0.02 1.0 0.01 Balance

Fig. 1 — A — The Laves phase in virgin FB2 steel observed by SEM. B — Chemical composition of the Laves phase obtained by EDX.

Fig. 2 — Microstructures observed by SEM in FB2 steel after thermal simulations: A —Tp = 1150˚C; B — Tp = 1200˚C; C and D — Tp = 1250˚C; E and F — Tp = 1350˚C.

A

A

C D

E F

B

B

LI ET AL SUPP 2015107_Layout 1 6/9/16 4:26 PM 58

consistent with the Laves phase inCB2 steel reported by Kasl (Ref. 11).Therefore, it can be speculated thatthe particles observed in virgin FB2steel were the Laves phase.

Evolution Behavior of the Laves Phase during the Welding Thermal Cycle The results of thermal simulationshow the Laves phase in specimens ex-

periencing peak temperatures below1200ºC (including 1200ºC) stayed un-changed, as shown in Fig. 2A and B,while in specimens experiencing higherpeak temperatures (1250º and1350ºC), some microstructures withmorphologically distinct forms insteadof the Laves phase particles were found,as shown in Fig. 2C–F. Based on themorphology, it can be speculated theywere eutectic structures. It is notewor-thy that in Fig. 2C–F, the eutectic mi-crostructure observed by SEM is justone eutectic constituent and the othereutectic constituent located in the in-tervals of the eutectic microstructuresmight be electrolytic etched. In specimens experiencing a peaktemperature of 1250ºC, the net-likeeutectic microstructures (indicated bysolid arrows) are surrounded by someseparated particles (indicated bydashed arrows), and they all locatewithin austenitic grains, while in spec-imens experiencing a peak tempera-ture of 1350ºC, the morphology of eu-tectic microstructures is more typical,and the size was much larger than thatin Fig. 2C and D. The net-like eutecticmicrostructures (indicated by solid ar-rows) locate at triple grain junctionswith “tails” along grain boundaries(indicated by dashed arrows). The chemical composition of theeutectic microstructures listed in Fig. 3 shows the content of Mo waslower than that in the Laves phase,and the contents of other elementswere similar to those of the Lavesphase. Based on the differences inmorphology and chemical composi-tion, it can be speculated that the eu-tectic microstructure was a new phaserather than the Laves phase.

The Results of Hot Ductility Tests In the hot ductility tests, specimen1 (Tp = 1350ºC) ruptured withoutnecking, and the peak force was16,000 N. Specimen 2 (Tp = 1250ºC)did not rupture with a slight neckingof about 0.4 mm. The result showsthat the zero ductility temperature(ZDT) of FB2 steel is between 1250ºand 1350ºC. The fracture appearance of speci-men 1 presents a typical liquefiedmorphology, as shown in Fig. 4A. InFig. 4B, some microstructures exist on

WELDING RESEARCH


Fig. 3 — The chemical composition of eutectic microstructures obtained by EDX.

Fig. 4 — The fracture appearance of a ruptured specimen in hot ductility tests: A —Macroscopic fracture appearance; B and C — microscopic fracture appearance; D —chemical composition of eutectic microstructures in B and C obtained by EDX.

A

C D

B


the fracture surface, indicated by blackarrows. The microstructures distributealong grain boundaries or at triplegrain junctions, because the trace ofgrain boundaries can be seen on themicrostructures, indicated by white ar-rows in Fig. 4B. The microstructures present a typi-cal eutectic morphology, as shown inFig. 4C, with a similar chemical compo-sition to that shown in Fig. 3. It can beconcluded the microstructures foundon the fracture appearance were thesame with those in thermal simulatedspecimens shown in Fig. 2E and F.

DiscussionThe Origin of the LavesPhase in FB2 Steel

In 9% Cr martensitic stainlesssteels, the Laves phase usually occursafter long-term, high-temperature ex-posure, and it was found the Lavesphase nucleates at martensitic lathboundaries or around M23C6 carbides,coarsening at the expense of alloyatoms in the matrix or M23C6 (Refs. 7,8). Some researchers found the forma-tion of the Laves phase at 650ºC need-ed at least hundreds of hours (Ref. 7).As to virgin FB2 steel, it was impossi-ble for the Laves phase to form duringtempering due to the too short time. There is only one possibility left —that the Laves phase formed in priorprocessing, such as casting and forg-ing. As mentioned in the introduction,a similar Laves phase was also foundin virgin CB2 steel, a cast steel withsimilar chemical composition to FB2.In virgin CB2 steel, sparse, relativelylarge particles of the Laves phase (Fe,Cr2)Mo are present in interdendriticareas (Ref. 12), indicating that theLaves phase forms during the castingprocess. Considering that the FB2specimens observed also went throughthe casting process, the Laves phasewas most likely formed in this process. The results of our experimentshowed the Laves phase remained sta-ble at temperatures below 1200ºC, sothe forging process following castingcould not eliminate it. In forging, allthe dendrites recrystallized, and largeLaves particles remained within theaustenitic grains. In addition, it wasnoted that the formation of the Lavesphase should be attributed to dendrit-ic segregation in the casting process.

Constitutional LiquationResulting from the EutecticReaction between the LavesPhase and Austenite

The occurrence of eutectic mi-crostructures strongly suggests a li-quation phenomenon in specimens ex-periencing peak temperatures above1250ºC (including 1250ºC), implying aliquation crack tendency in the HAZ ofFB2 steel during welding. The resultsof hot ductility tests further con-firmed the existence of liquation crackin FB2 steel at 1350ºC. A prerequisite for liquation crack inthe HAZ was the formation of discreteliquid regions within the solid metalexperiencing enough high tempera-tures during welding. If this liquid ex-hibits a tendency to wet, and therebyform a continuous or semicontinuousthin film of liquid along grain bound-aries, and if sufficient strains are pres-ent in this weld region, then inter-granular separation will occur. The ori-gin of such liquation in the HAZ wasmost often attributed to the “constitu-tional liquation” phenomenon, whichwas originally proposed by Savage andcoworkers (Ref. 13). This phenomenoninvolves a eutectic reaction between asecondary constituent phase and the matrix, and mostly occurs inaustenitic stainless steels such as Alloy 718, A-286, GH150, and so on.During the past decades, HAZ liqua-tion in austenitic stainless steels con-taining Nb and Ti has been attributedto the constitutional liquation of Nb-

rich Nb(C, N) carbonitrides (Refs.14–16), Ti-rich TiC carbides, and theLaves phase (Refs. 17, 18). In the present study, results of thethermal simulation with differentpeak temperatures suggest that con-stitutional liquation of the Lavesphase in virgin FB2 steel accounts forthe liquation phenomenon. The heat-ing rate of the welding thermal simula-tion was nonequilibrium, whichcaused rapid decomposition of theLaves phase. This decomposition high-ly enriched the region adjacent to thedissolving Laves particles in solutessuch as Cr and Mo. The equilibriumphase for this high-solute compositionwas a liquid that surrounded the dis-solving Laves particles. The liquid re-gion remained until the Laves particlesdissolved completely. It was noted that the Ac3 of FB2steel at a heating rate of 100ºC/smeasured previously was about 960ºC,so the matrix had all transformed intoaustenite at peak temperatures ofthermal simulation. Therefore, theconstitutional liquation in the heatingprocess of thermal simulation resultedfrom the eutectic reaction between theLaves phase and austenite. In specimens experiencing a peaktemperature of 1250ºC, every eutecticstructure was separated, and the sizewas similar to that of the Laves phasein virgin FB2 steel, as shown in Fig.2A–D. This phenomenon can be ex-plained by the poor flowing ability ofliquid at 1250ºC. In the present work,the Laves phase remained stable at1200ºC, and when the peak tempera-

WELDING RESEARCH


Fig. 5 — Equilibrium phase diagram of the eutectic constituent obtained by JMatPro®.


ture arrived at 1250ºC, every separat-ed particle of the Laves phase dis-solved and formed some separated liq-uid cells owing to the eutectic reactionbetween the Laves phase and austen-ite (or the matrix). At this tempera-ture, the mobility of the liquid wasvery limited, so every liquid cell stayedwhere the Laves particle originally wasinstead of flowing and merging to-gether. In cooling, every separated liq-uid cell transformed into a solid phase,as shown in Fig. 2C and D. While inspecimens experiencing a peak tem-perature of 1350ºC, the mobility ofliquid was greatly promoted by thehigher temperature, which means thatthe liquid could flow in long distancesand combine together. In addition, the growth ofaustenitic grains promoted the im-pingement of grain boundaries withthe liquid regions, so most liquid dis-tributed along grain boundaries or attriple grain junctions, as shown in Fig.2E and F. Based on the liquation mechanismdiscussed, the results of hot ductilitytests can be explained. In these, thedisplacement load was applied to thespecimens at peak temperatures of1250º and 1350ºC. At 1250ºC, the limited mobility ofliquid retarded wetting grain bound-aries, so the grain boundaries’

strength was not impaired. As a result,the specimen exhibited some plasticityat 1250ºC. In specimens experiencingpeak temperatures of 1350ºC, somegrain boundaries were wet by the liq-uid and lost strength. When the loadwas applied, the areas losing grainboundary strength acted as crack initi-ation, thereby intergranular fractureoccurred without any plasticity. Owing to distribution characteris-tics of the Laves phase, it was men-tioned in the “observation of Lavesphase in virgin FB2 steel” section,most of the grain boundary area re-mained free of liquid to maintainsome strength, so the specimen at1350ºC exhibited some load-bearingcapacity.

The Evolutionary Behavior ofLiquid during Cooling

It has been mentioned in the sectiontitled “evolution behavior of the Lavesphase during the welding thermal cy-cle,” the eutectic microstructure ob-served by SEM in Fig. 2C–F is just oneof the eutectic constituents, and theother eutectic constituent in the inter-vals might be etched. Both the mor-phology and chemical composition ofthe eutectic constituent were differentfrom those of the Laves phase in virginFB2 steel, suggesting the eutectic con-

stituent is a new phase and anothertype of eutectic reaction might occurduring cooling. To clarify the type of the eutecticconstituent, thermodynamic softwareJMatPro® was employed to get all thepossible phases in equilibrium condi-tion based on the chemical composi-tion listed in Fig. 3. Although weldingis a nonequilibrium process, and theresults obtained by JMatPro® werespeculative, they can provide some in-formation for reference. The equilibrium phase diagram ofthe eutectic constituent is shown inFig. 5. It was evident there are two ma-jor phases at room temperature; one isferrite and the other is the Chi phase.The existence of ferrite in the eutecticconstituent can be excluded becauseferrite is not corrosion resistant. Sothe Chi phase is most likely the eutec-tic constituent. The Chi phase is an intermetalliccompound containing primarily Fe, Cr,and Mo. It is a body-centered-cubicphase (-Mn structure) with a latticeparameter of a0 = 0.892 nm (Ref. 19).The Chi phase is often found inaustenitic and ferritic stainless steelscontaining Mo (Refs. 20, 21). Kautzand Gerlach have reported finding theChi phase as a eutectic constituent inType 316 stainless steel, which washeated to 1380ºC and then waterquenched (Ref. 22). Cieslak and Ritteralso found the Chi phase as a eutecticconstituent along solidification grainboundaries in CF-8M welds, and theypointed out that the kinetics of theChi phase formation are greatly en-hanced by the presence of Mo, but themechanism of eutectic reaction wasnot clarified in his work (Ref. 23). InCieslak and Ritter’s research, the eu-tectic Chi phase occurred along the hotcrack in welds (Ref. 23). The chemical compositions of theChi phase obtained by Cieslak (Ref.23) and Weiss (Ref. 24) are the sameas that obtained in the present workto a large extent. All three chemicalcompositions are summarized in Table2. Based on morphology and chemicalcomposition already discussed, it ispreliminarily speculated that the eu-tectic constituent observed by SEM isthe Chi phase. The speculation is fur-ther supported by the results of TEManalysis. The TEM specimen shown in

WELDING RESEARCH


Table 2 — Chemical Composition of the Chi Phase Obtained by Two Other Authors and in the Present Work (wt%)

Material Fe Cr Mo Ni

Cieslak (Ref. 23) CF8M austenitic stainless steel 45 26 20 4 Weiss (Ref. 24) 316 austenitic stainless steel 52 21 22 5Eutectic constituent FB2 martensitic stainless steel 51–54 23–25 18–24 0 in the present work

Fig. 6 — The results of TEM analysis of the eutectic microstructures: A — Bright fieldimage; B — SAED results of the different areas in A.

A B


Fig. 6A was sampled by FIB, and in thebright field image, the eutectic con-stituent observed by SEM can also beseen clearly, indicated by white arrows.Because the specimen was not etched,the other eutectic constituent in inter-vals can also be seen, indicated byblack arrows. The four SAED patternsin Fig. 6B labeled 1–4 correspond tothe areas in Fig. 6A labeled by 1–4, respectively. The results of SAED indicate thatthe eutectic constituent observed bySEM had a body-centered-cubic (BCC)structure, consistent with the Chiphase, and the lattice parameter of theeutectic constituent was 0.906 nm(see the appendix for the calculatingprocess), close to that of the Chiphase. The other eutectic constituentalso had a BCC structure, and the lat-tice parameter was 0.271 nm. It’s acknowledged that martensitehas a body-centered-tetragonal (BCT)structure, of which the lattice parame-ter ‘a’ is equal to ‘b,’ but is not equal to‘c’ because of the distortion duringmartensitic transformation. However,the lattice parameter ‘a’ is almostequal to ‘c,’ so the diffraction patternsof martensite are almost the same asthose of BCC, such as ferrite. In addition, the lattice parameter ofmartensite was almost equal to that offerrite at 0.2866 nm. Based on thestructure and lattice parameter, theother eutectic constituent can be pre-liminarily judged as martensite. Con-sidering that martensite was trans-formed from austenite, the eutectic re-action in cooling was most likely to beL → Chi + (Fe).

Based on this analysis, the mor-phology difference of eutectic mi-crostructures in Fig. 2C–F can be ex-plained. As to the specimen experienc-ing a peak temperature of 1250ºC, itwas mentioned in the constitutional li-quation resulting from the eutectic re-action between the Laves phase andaustenite section, the interface be-tween the Laves particles and matrixdissolved first, as shown in Fig. 7B. The chemical composition of theliquid deviated from the eutectic com-position, biased toward austenite, be-cause the solute (Cr and Mo atoms)was diluted by the matrix. Subse-quently, the rest of the Laves particlesdissolved, as shown in Fig. 7C. Because there was no direct contactbetween the central part of the Lavesphase and matrix, and because themobility of Cr and Mo atoms was lim-ited, the solute in the central partcould not be diluted adequately likethat at the interface. Thereby, thechemical composition of liquid in thecentral part was biased toward theLaves phase enriched in Mo and Cr.During cooling, austenite precipitatedfrom liquid at the interface as the pre-dominant phase and merged with thematrix, leaving a very small amount ofChi phase alone, which formed di-vorced eutectic. The divorced eutectic is indicatedby dashed arrows in Fig. 2C and D.While in the central part, the amountof liquid was larger than that at the in-terface, so the Chi phase and (Fe)could precipitate alternatively, form-ing net-like symbiotic eutectic, labeledby solid arrows in Fig. 2C and D.

The divorced and symbiotic eutecticis illustrated in Fig. 7D. In the case of the specimen cooledfrom 1350ºC, solute atoms diffusedadequately in the liquid because ofthe higher temperature. At grainboundaries, the liquid film was thinand (Fe) precipitated from the liquidfilm and merged with the matrix,forming tail-like divorced eutectic,which is labeled by dashed arrows inFig. 2E and F. The amount of liquid attriple grain junctions was larger, so (Fe) and the Chi phase could precipi-tate alternatively, forming symbioticeutectic, which is labeled by solid ar-rows in Fig. 2E and F. The grainboundaries surrounded by the eutec-tic structures became “ghost bound-aries,” of which the strength was im-paired and crack initiation would pre-fer to take place at this site.

Conclusions In the present work, constitutionalliquation resulting from the eutecticreaction between the Laves phase and (Fe) was found, suggesting a liqua-tion crack tendency in FB2 steel dur-ing welding. Then the origin and evo-lutionary behaviors of the Lavesphase in the welding thermal cyclewere analyzed. There are four conclu-sions obtained as follows: 1) The large particles of the Lavesphase in virgin FB2 steel formed inthe casting process, which was attrib-uted to dendritic segregation. 2) The fast heating rate of weldingled to constitutional liquation in ar-eas where the temperature was above1250ºC (including 1250ºC), and theeutectic reaction between the Lavesphase and (Fe) should account forthe constitutional liquation. 3) The areas where constitutional li-quation occurred would act as crack ini-tiation when some load was present at1350ºC, leading to intergranular frac-ture. At 1250ºC, the limited mobility re-tarded liquid wetting grain boundaries,so hot plasticity was present. 4) The Chi phase was found as aeutectic constituent during cooling.The mobility of atoms and liquid in-fluenced the final morphology of eu-tectic microstructures, and divorcedeutectic as well as symbiotic eutecticwere observed in specimens experi-encing different peak temperatures.

WELDING RESEARCH


Fig. 7 — Schematic of the Laves phase evolution during welding, Tp = 1250˚C: A — Beforeheating; B — the interface between the Laves phase and (Fe) melted first during heating; C — the central part of the Laves phase melted subsequently; D — during cooling,the Chi phase and austenite precipitated from the liquid.

A B C D


The authors would like to thankSenior Engineers R. Wang and W. Yangfor their guidance in SEM and EDXanalysis. Dr. Guo at the University ofScience and Technology Beijing is alsothanked for her help in thermal simu-lation work.

1. Blaes, N., Donth, B., and Bokelmann,D. 2007. High chromium steel forgings forsteam turbines at elevated temperatures.Energy Materials 2: 207–213. 2. Lu, F., Liu, P., Ji, H., Ding, Y., Xu, X.,and Gao, Y. 2014. Dramatically enhancedimpact toughness in welded 10% Cr rotorsteel by high temperature post-weld heattreatment. Mater Charact. 92: 149–158. 3. Di Gianfrancesco, A., Cipolla, L.,Paura, M., Vipraio, S. T., Venditti, D., Neri,S., et al. The role of boron in long termstability of a CrMoCOB (FB2) steel for ro-tor application. Advances in materialstechnology for fossil power plants. Pro-ceedings from the Sixth International Con-ference, August 31–September 3, 2010,Santa Fe, N.Mex. ASM International, p.342. 4. Nakano, M., Kawano, K., and Mika-mi, M. Manufacturing of trial rotor forgingof 9% Cr steel containing Co and B (X13Cr-MoCoVNbNB9-2-1) for ultrasupercriticalsteam turbines. Advances in materialstechnology for fossil power plants. Proceed-ings from the Seventh International Confer-ence, October 22–25, 2013, Waikoloa,Hawaii. ASM International, p. 321. 5. Shige, T., Magoshi, R., Itou, S.,Ichimura, T., and Kondou, Y. 2001. Devel-opment of large-capacity, highly efficientwelded rotor for steam turbines. Mit-subishi Heavy Industries Technical Review38: 6–11. 6. Kern, T. U., Almstedt, H., Thiemann,T., Brussk, S., and Niepold, K. 2013. TheRole of Rotor Welding Design in Meeting Fu-ture Market Requirements. ASME, pp.V5B–V25B. 7. Cui, H., Sun, F., Chen, K., Zhang, L.,Wan, R., Shan, A., et al. 2010. Precipita-tion behavior of Laves phase in 10% Crsteel X12CrMoWVNbN10-1-1 duringshort-term creep exposure. Materials Sci-

ence and Engineering A 527: 7505–7509. 8. Xu, Y., Wang, M., Wang, Y., Gu, T.,Chen, L., Zhou, X., et al. 2015. Study onthe nucleation and growth of Laves phasein a 10% Cr martensite ferritic steel afterlong-term aging. Journal of Alloy Com-pounds 621: 93–98. 9. Shi, R., and Liu, Z. 2012. Growth be-haviour of Laves phase of -ferrite in P92steels. Iron and Steel 47: 55–59. 10. Dimmler, G., Weinert, P.,Kozeschnik, E., and Cerjak, H. 2003.Quantification of the Laves phase in ad-vanced 9–12% Cr steels using a standardSEM. Mater Charact. 51: 341–352. 11. Kasl, J., and Mikmeková, JandováD. 2014. SEM, TEM and SLEEM (scanninglow energy electron microscopy) of CB2steel after creep testing. IOP ConferenceSeries: Materials Science and Engineering. 12. Jandová, D., Kasl, J., and Chvos-tová, E. 2014. Microstructure of CB2 steelbefore and after long-term creep tests.Materials Science Forum: 782. 13. Pepe, J. J., and Savage, W. F. 1967.Effects of constitutional liquation in 18Nimaraging steel weldments. Welding Jour-nal 46(9): 411-s to 422-s. 14. Lee, C. H., and Lundin, C. D. 1998.Relationship between hot ductility behav-ior and microstructural changes in TP347stainless steel. Welding Journal 77(1): 29-sto 37-s. 15. Thompson, R. G., and Genculu, S.1983. Microstructural evolution in theHAZ of Inconel 718 and correlation withthe hot ductility test. Welding Journal62(12): 337-s to 345-s. 16. Andersson, J., Sjöberg, G. P.,Viskari, L., and Chaturvedi, M. C. 2012.Effect of solution heat treatments on su-peralloys. Part 1 — Alloy 718. Mater. Sci.Tech-Lond. 28: 609–619. 17. Brooks, J. A. 1974. Effect of alloymodifications on HAZ cracking of A-286stainless steel. Welding Journal 53(11):517-s to 523-s. 18. Baeslack, W. A., Lata, W. P., andWest, S. L. 1988. A study of heat-affectedzone and weld metal liquation cracking inAlloy 903. Welding Journal 67(4): 77-s to87-s. 19. Kasper, J. S. 1954. The ordering ofatoms in the Chi-phase of the iron-chromium-molybdenum system. ActaMetallurgica. 2: 456–461. 20. Okafor, I., and Carlson, O. N. 1978.Equilibrium studies on a Chi phase-strengthened ferritic alloy. MetallurgicalTransactions A 9: 1651–1657.

21. Song, Y., McPherson, N. A., andBaker, T. N. 1996. The effect of weldingprocess on the Chi phase precipitation inas-welded 317L weld metals. ISIJ Interna-tional 36: 1392–1396. 22. Kautz, H. R., and Gerlach, H. 1968.Mechanical and corrosion-resistance —Properties of unstabilized fully austeniticsteels used in reactors and steam-boilerplants. Arch Eisenhuttenw. 39: 151–158. 23. Cieslak, M. J., Ritter, A. M., andSavage, W. F. 1984. Chi-phase formationduring solidification and cooling of CF-8M weld metal. Welding Journal 63(4): 133. 24. Weiss, B., and Stickler, R. 1972.Phase instabilities during high tempera-ture exposure of 316 austenitic stainlesssteel. Metallurgical Transactions 3:851–866.

From Fig. 6B (number 1), we canget the interplanar distance in thethree directions. Now, as an example,take the interplanar distance of (–1,2, –1) d(–1, 2,–1). It is easy to get the length from thecentral spot to this spot (–1, 2, –1) R(–1, 2, –1) = 2.7034 1/nm on the recipro-cal space. Then the interplanar distanced(–1, 2, –1) = 1/R(–1, 2, –1) = 0.3699 nm. The relationship between the lat-tice parameter a0 and interplanar dis-tance d satisfies the equation below

where (h, k, l) is the corresponding in-dices of the crystal face. In this way, it is easy to get the lat-tice parameter a0.

In the same way, we can get the lat-tice parameter in Fig. 6B (numbers 2through 4).

d = a0 / h2 + k2 + l2

a0 = d h2 + k2 + l2

=0.3699 (=1)2 +22 +(=1)2

=0.906 nm

WELDING RESEARCH


Acknowledgments

References

Appendix


constitutional liquation of the laves phase in virgin fb2 steel€¦ · creep-resistant martensitic...

Documents