Seediscussions,stats,andauthorprofilesforthispublicationat:http://www.researchgate.net/publication/262791034

Effectofcoolingratesfrompartialsolutiontemperatureandagingony'precipitationinIN792superalloyARTICLEinMATERIALSSCIENCEANDTECHNOLOGYJUNE2013ImpactFactor:0.8

DOWNLOADS21

VIEWS24

1AUTHOR:

HosseinArabiIranUniversityofScienceandTechnology148PUBLICATIONS153CITATIONS

SEEPROFILE

Availablefrom:HosseinArabiRetrievedon:17June2015

Effect of cooling rates from partial solutiontemperature and aging on c9 precipitation inIN792 superalloy

H. Arabi1, S. Rastegari2, M. Mirhosseini3 and B. M. Sadeghi*2

One of the important factors affecting the microstructure of Ni base superalloys is the cooling rate

from partial solution temperature; therefore, the prime purpose of the present research is to

investigate the effects of this factor on some microstructural features of superalloy IN792. The

results showed that by increasing the cooling rate from 5 to 160uC min21 after partial solutiontreatment at 1120uC for 2 h, the size and volume fraction of primary c9 decreased, but during thesubsequent aging the volume fraction of secondary c9 increased initially exponentially and later

linearly. When high cooling rates .90uC min21 were applied, the formation of secondary c9 wastotally suppressed. However, after aging at 845uC for a period of 24 h, the amount of secondary c9increased with a decelerating rate up to 60uC min21 and then with a linear rate up to 160uC min21.The later observation has not been reported elsewhere.

Keywords: IN792, Superalloy, Cooling rate, c9 phase, Heat treatment

IntroductionNi base superalloys are very suitable materials for hightemperature applications due to their ability to keeptheir high strength at a temperature as high as 80% oftheir melting point. Considering the complexity of nickelbased superalloy microstructures, it is essential tocontrol the effective processing parameters on themicrostructures of these types of alloys in order toprevent the formation of harmful phases and be able todevelop useful features for specific application. Themicrostructure of this type of alloy in general includes asolid solution matrix of nickel known as c with FCCstructure, primary and secondary c9 phases with Ni3 (Al,Ti) composition having L12 structure; eutectic of c9/c;and some strengthening carbides such as MC, M6C andM23C6.

1 Other harmful needle shape or Chinese scriptshape phases such as laves and s might be present intheir microstructure if processing is not controlledcarefully.2

The high strength of these alloys is basically due to thestrength of solid solution c matrix and coherent c9particles that usually appear in the form of large cubicprimary c9 and small spherical secondary c9.2 Thecharacteristics of the c9 phase can directly affect themechanical properties of these alloys.14 These char-acteristics are a function of the composition and type of

processing parameters, such as heat treatment appliedon these types of alloys.16 The principle objective of fullsolution treatment is to put hardening phases intosolution and dissolve some carbide. However, since theutilisation of higher solution treating temperature fora long time will result in some grain growth, moreextensive dissolving of carbides, melting of eutecticphase particularly in the vicinity of grain boundaries,and possible formation of continuous MC carbideswithin grain boundaries, it would be better to performthis operation in two steps. One should bear in mindthat when more than one phase is capable of precipitat-ing from the alloy matrix during heat treatment ajudicious selection of assign solution and aging treat-ments that produce different sizes and types and betterdistribution of precipitates at different temperature isvery important. Therefore, in the present research, inaddition to the application of a full solution for a shorttime to homogenise the microstructure, partial solutionat lower temperature was also used to stabilise themicrostructure and form discontinuous carbides at grainboundaries.

Some researches79 have pointed out the importanceof cooling rate from partial solution temperature on themicrostructure of some Ni base superalloys. Forexample, Koul et al.7 reported by increasing the coolingrate from partial solution temperature the size of c9decreases, while other researchers8 on Udimet 720Lishowed that the sizes of c9 precipitates are insensitive tothe cooling rate after solution treatment. It has also beenreported9 that the presence of elements Re and W in thesuperalloy CMSX-4 reduces the sensitivity of c9 growthto cooling rate because of the delay that occurs in thediffusion process. In addition to the size of c9, a severe

1Center of Excellence for High Strength Alloys Technology (CEHSAT),School of Metallurgy and Materials Engineering, Iran University of Scienceand Technology (IUST), Narmak, Tehran, Iran2School of Metallurgy and Materials Engineering, IUST, Narmak, Tehran,Iran3IUST, Narmak, Tehran, Iran

*Corresponding author, email bmsadeghi@yahoo.com

2013 Institute of Materials, Minerals and MiningPublished by Maney on behalf of the InstituteReceived 14 January 2013; accepted 11 June 2013DOI 10.1179/1743284713Y.0000000325 Materials Science and Technology 2013 VOL 29 NO 12 1513

reduction in volume fraction of c9 precipitates withincreasing cooling rate has been reported by Sajjadiet al.10 for Ud500 and Mao et al.11 for Rene88DT.Therefore, one may conclude for the above argumentthat the cooling rate from solution or partial solutiontemperatures can affect the microstructure of super-alloys after aging in different ways that depend on thecomposition of these alloys.The superior tensile and stress rupture properties of

IN792 relative to IN738LC made this alloy a goodcandidate for use in gas turbine blades.12 However, tothe best knowledge of the authors, there is nocomprehensive published report on the effect of coolingrate on the microstructure of IN792 during cooling withvarious rates from partial solution temperature and theproceeding aging. Thus, the aim of the present researchis to clarify these effects to some extent.

ExperimentalThe mean chemical composition of the IN792 superalloyused in the present research is Ni0?08C3?4Al0?14Fe0?03Si8?8Co12?6Cr1?8Mo4Ti4?23W4?1Ta (wt-%).The initial bars of IN792 were cast in vacuum 1023 bar inan inductive furnace. These bars were solution treated at1200uC for 2 h. Then they were subjected to partialsolution treatment at 1120uC for 2 h, before cooling themto room temperature with controlled cooling rates of 5,20, 30, 60, 90,120 and 160uC min21. Then, severalsamples were prepared from each partially treated barbefore subjecting them to aging at 845uC for 24 h andcooling them to room temperature in air. The heattreatment cycles applied to the cast bars are shownschematically in Fig. 1.After preparing samples in a routine metallographic

way, the microstructures were examined by scanningelectron microscopy (SEM). The etchant used in thepresent study was 25 mL lactic acidz15 mLHNO3z1 mL HF. Quantitative measurement of c9precipitates was performed by an image analyser;however, the size distribution of the tiny particles wasnot established due to the lack of access to transitionelectron microscopy.

ResultsThe typical as cast microstructure of IN792 superalloy ispresented in Fig. 2. The microstructure of the as castsample contained basically cmatrix and cubic primary c9precipitates in a butterfly shape (e.g. see inside inserts inFig. 2). The cast microstructure after being subjected to

partial solution treatment and cooled with differentcooling rates to room temperature changed enormouslyas shown in various micrographs in Fig. 3. Thesemicrographs show at a very low cooling rate of5uC min21 a few number of secondary c9 particlesprecipitated and grown together with primary c9 duringcooling from partial solution temperature. By increasingthe cooling rate from 10 to 60uC min21, the number ofsmall spherical secondary c9 precipitates increased, whileat higher cooling rates up to 160uC min21 only fewsecondary c9 precipitated during cooling from partialsolution temperature. In addition, at cooling rateshigher than 60uC min21, the shape of primary c9 wasdifferent to those observed earlier in samples cooled withlower rates. In other words, cubic c9 was the dominantshape observed.

The SEM microstructures of the samples after agingare shown in Fig. 4. This figure shows that the mean sizeof the c9 particles in all of the samples slightly increasedrelative to the samples in partial solution conditions. Inaddition, a large number of secondary c9 appeared inthe matrix of the samples that had cooling rates.60uC min21.Figures 5 and 6 show by increasing the cooling rate

that the mean size of the primary c9 particles decreased,while during subsequent aging it increased substantially.Increasing the cooling rate from 5 to 160uC min21

caused the mean size of the primary c9 particles toreduce from 680 to 510 nm, but during aging moregrowth was observed on the samples that had highercooling rate from partial solution temperature. Forexample, the mean size of primary c9 particles afteraging for cooling rate of 5uC min21 was 730 nm, whichmeans 7% increase in their mean size, while the meansize of these particles after aging for cooling rate of160uC min21 was 620 nm, which means 17% increase intheir mean size.

Similarly, the mean size of the secondary c9 particlesafter aging decreased with increasing aging temperaturebut with a decelerating rate, as shown in Fig. 6. Forexample, the sizes of secondary c9 particles after agingfor cooling rates of 5 and 160uC min21 were 180 and96 nm respectively.

Changes in volume fraction of secondary c9 particlesafter aging are shown in Fig. 7. This figure shows thatincreasing the cooling rate from partial solutiontemperature severely affected the precipitation beha-viour of c9 particles after aging. The higher the coolingrate, the higher the volume fraction of secondary c9

1 Schematic of heat treatment cycles2 Microstructure of IN792 cast alloy

Arabi et al. Effect of cooling rates on c9 precipitation in IN792 superalloy

1514 Materials Science and Technology 2013 VOL 29 NO 12

particles after aging. However, the increasing rate ofvolume fraction of secondary c9 particles after agingreduces with increasing cooling rate, so that the use ofcooling rates .60uC min21 caused the volume fractionof secondary c9 particles to increase linearly with a verysmall slope rather than exponentially, as shown by thebroken line in Fig. 7.

DiscussionThe formation of secondary c9 phase within the matrixrequires, first of all, some parts of the matrix to becomerich in specific elements, and then these rich areastransform to new phase out of the host matrix, as theclassical nucleation theory states.13 The critical radiusfor nucleation depends basically on two factors: 1) freechemical energy of the super saturated matrix and 2)interface energy of c and c9 particles. Increasing thecooling rate from solution temperature causes the degreeof supersaturation to increase, as has been reported byMao et al.,11 and the coherent strain energy between cand c9 to decrease according to Mitchell et al.14 Both ofthe abovementioned phenomena can cause a reductionin the radius of the critical nuclei, so the nucleation of c9precipitates from the cmatrix with smaller critical radiusbecomes possible. Therefore, when the rate of coolingfrom solution temperature was low (i.e. 520uC min21),the energy barrier for nucleation was high, so that theelements required for the formation of new c9 particlesdiffuse preferably into the existed primary c9 particlesand caused their growth rather than be used for theformation of new c9 particles (see Figs. 2 and 5). Thus,as it can be seen in the microstructures presented in

Fig. 3 and the graph in Fig. 7, very limited number ofsecondary c9 particles were formed during cooling fromsolution temperature.

When the cooling rate increases up to 60uC min21,higher nucleation of secondary c9 particles occurs duringcooling, as Figs. 3 and 7 indicate; however, furtherincrease in the cooling rate severely reduced thediffusion time required for the elements, so no newlysecondary c9 particles can be precipitated during coolingfrom partial solution temperature. Under these condi-tions, a very limited amount of growth in the primary c9particles during cooling was possible. Therefore, whena very high cooling rate of .90uC min21 from thesolution temperature was used, an ignorable number ofsecondary c9 particles were detected within the micro-structure of the matrix (see Fig. 3eg). This seems torelate to the very limited time available for diffusion ofelements required for the formation of c9 particles withinthe matrix. Therefore, due to the low diffusion rate, therequired critical Gibbs free energy for nucleation was sohigh in value that neither formation of new secondary c9particles nor the growth of existing primary c9 particlesduring cooling was noticeable. Instead, for high coolingrates of 120 and 160uC min21, highly contrasted zonesdue to segregation of some elements [which can distortthe surrounding lattice (without changing the matrixstructure), causing local variation in the intensity ofelectron diffraction (similar to the image of GuinierPreston zones in AlCu alloys, or with some reservationsthe images produced during diffusionless transformationof austenite to martensite, see section 6 in Ref. 15),which in turn shows up as a variation in image intensity]can be observed (see Fig. 3f and g). These zones can

3 Microstructures of IN792 after partial solution treatment at 1120uC for 2 h and different cooling rates of a 5uC min21,b 20uC min21, c 30uC min21, d 60uC min21, e 90uC min21, f 120uC min21 and g 160uC min21

Arabi et al. Effect of cooling rates on c9 precipitation in IN792 superalloy

Materials Science and Technology 2013 VOL 29 NO 12 1515

provide the required conditions for precipitation ofsecondary c9 particles during the subsequent aging.15

Since after aging there is a driving force for precipitationof the equilibrium c9 phase, secondary c9 particlesprecipitate, causing a substantial reduction in the matrixlattice distortion, so that the zone contrast observedafter cooling disappeared after aging.

As indicated above, high cooling rates, particularlythe cooling rates of 120 and 160uC min21, causedsuppression of elemental diffusion within the matrix;hence, the formation of secondary c9 particles during

cooling was suspended. In other words, enough elementsrequired for precipitation during aging remained withinthe supersaturated matrix. On the other hand, since thesubstrate contains a large number of heterogeneouslydispersed in-grown dislocations as well as a great deal ofsolute atoms, one expects that the nucleation ofsecondary c9 particles occurs heterogeneously as thekinetic of precipitation of particles varies with disloca-tion densities within the matrix. The size and numbers ofnuclei depended on the degree of supersaturation and

5 Variation of primary gamma prime size with different

cooling rates from partial solution temperature

6 Variation of secondary gamma prime size with different

cooling rates from partial solution temperature

a 5uC min21; b 20uC min21; c 30uC min21; d 60uC min21; e 90uC min21; f 120uC min21; g and h 160uC min21

4 Microstructures of IN792 after aging samples cooled at different rates at 845uC for 24 h

Arabi et al. Effect of cooling rates on c9 precipitation in IN792 superalloy

1516 Materials Science and Technology 2013 VOL 29 NO 12

the number of stress concentration points (e.g. disloca-tion density in this case) exiting in the matrix. Where thesaturation degree and the stress concentration pointswere high, the number of secondary c9 particles formedduring aging was also very high (see Fig. 4). Thisargument is visualised in Fig. 7. These figures show atcooling rates .60uC min21 that the combination ofhigher degree of elemental saturation zones and highnumber of stress concentration zones caused theformation of a large amount of very small secondaryc9 particles within the supersaturated matrix duringaging.Another phenomenon observed in the microstructure

of this alloy under all conditions, i.e. cast, solutiontreated and aged, was that the primary c9 particlesappeared with a butterfly shape morphology. Theobservation of this type of c9 outlook is said16 to bedue to the partitioning phenomenon that occurs duringheat treatment, but in the current research this config-uration of primary c9 was also observed in the cast alloy(Fig. 2). This observation confirms the idea that thepartitioning phenomenon that occurs during heattreatment is not necessary the only partitioning mechan-ism for creating such morphology for primary c9. Theparticles formed during casting had no enough time forgrowth and partitioning; therefore, other mechanismsmay have been operative for the formation of this typeof morphology. A few years ago Zhang et al.17 reportedthat elements such as Ti, Ta and Cr severely segregatewithin the last portion of the melt during solidificationof IN792. Considering this argument, it seems thesegregation of the element, such as Al, Ti, Ta and Cr,which is required for the formation of primary c9particles during solidification, caused the formation of

very rich zones from those elements during this process.Therefore, during cooling of the melt, the chance ofnucleation of more than one nucleus in any segregatedzone was more; thus, several primary c9 particlesnucleated close to each other and grew as far as therequired species for growth were available within thesegregated matrix. The above argument requires furtherinvestigation.

Conclusions

1. The microstructural features of IN792 are sensi-tive to the cooling rate from partial solution, so that byincreasing this cooling rate from 5 to 160uC min21 theamounts of small secondary c9 particles increase with adecelerating rate during aging of this alloy.

2. By increasing the cooling rate, the sizes of primaryand secondary c9 particles decrease in the microstructureof superalloy IN792.

Acknowledgements

The authors like to thank engineers M. Cheraghzadehand Y. Milladi from Mavadkaran Eng. Co for support-ing this project.

References1. N. Das: Trans. Indian Inst. Met., 2010, 63, 265274.

2. M. J. Donachie and S. J. Donachie: Superalloys: a technical

guide; 2002, Materials Park, OH, ASM International.

3. M. Jackson and R. Reed: Mater. Sci. Eng. A, 1999, A259, 8597.

4. R. Sharghi-Moshtaghin and S. Asgari: J. Mater. Process. Technol.,

2004, 147, 343350.

5. S. A. Sajjadi, S. M. Zebarjad, R. Guthrie and M. Isac: J. Mater.

Process. Technol., 2006, 175, 376381.

6. H. Monajati, M. Jahazi, R. Bahrami and S. Yue: Mater. Sci. Eng.

A, 2004, A373, 286293.

7. A. K. Koul, R. Castillo, V. P. Swaminathan and N. S. Cheruvu:

Advanced materials and coatings for combustion turbines, Proc.

ASM 1993 Materials Cong. Materials Week, ASM International,

Pittsburg, PA, USA, October 1993, 1721.

8. F. Torster, G. Baumeister, J. Albrecht, G. Lutjering, D. Helm and

M. Daeubler: Mater. Sci. Eng. A, 1997, A234, 189192.

9. F. Blum, J. Benson and C. Stander: J. Mater. Sci. Lett., 1994, 13,

12131214.

10. S.A. Sajjadi, H. Elahifar and H. Farhangi: J. Alloys Compd, 2008,

455, 215220.

11. J. Mao, K. M. Chang, W. Yang, D. U. Furrer, K. Ray and S. P.

Vaze: Mater. Sci. Eng. A, 2002, A332, 318329.

12. G. L. Erickson: Metals handbook, 10th edn, Vol. 1, 981994;

1990, Materials Park, OH, ASM international.

13. R. Reed-Hill and R. Abbaschian: Physical metallurgy principles;

1994, Boston, MA, PWS Publishing Company.

14. R. Mitchell, M. Preuss, M. Hardy and S. Tin: Mater. Sci. Eng. A,

2006, A423, 282291.

15. D. A. Porter and K. E. Easterling: Phase transformations in metals

and alloys, Vol. 5, 290300; 1983, Berkshire, Van Nostrand

Reinhold Co. Ltd.

16. M. Doi, T. Miyazaki and T. Wakatsuki:Mater. Sci. Eng., 1984, 67,

247253.

17. J. Zhang and R. Singer: Acta Mater., 2002, 50, 18691879.

7 Variation of volume fraction of secondary gamma

prime size after aging with different cooling rates from

partial solution temperature

Arabi et al. Effect of cooling rates on c9 precipitation in IN792 superalloy

Materials Science and Technology 2013 VOL 29 NO 12 1517