Microstructure and Microhardness of a Nanostructured Nickel-Iron Based Alloy

Shamil Mukhtarov1, a, Xavier Sauvage 2, b 1Institute for Metals Superplasticity Problems, Russian Academy of Sciences, 39 Khalturin St, Ufa

450001, Russia 2University of Rouen, CNRS UMR 6634, Groupe de Physique des Matériaux, Faculté des

Sciences, BP12, 76801 Saint-Etienne du Rouvray, France ashamil@anrb.ru, bxavier.sauvage@univ-rouen.fr

Key words: severe plastic deformation, high pressure torsion, multiple isothermal forging, nanostructure, superalloy, atom probe tomography, microhardness.

Abstract. This paper presents an overview and some original results about the mechanical properties and phase analysis of a nanostructured (NS) nickel-iron based alloy INCONEL 718. This structure was obtained by severe plastic deformation (SPD) via high pressure torsion (HPT) and multiple isothermal forging (MIF) of the alloy with an initial coarse-grained (CG) structure. Materials before and after SPD were analyzed by scanning, transmission electron microscopes and atom probe tomography (APT). Experimental data indicate that after HPT at room temperature γ′′-phase was partly dissolved and that precipitation of the δ-phase occurs during post deformation aging. A hardness up to 8 GPa was recorded for the NS alloy after SPD and annealing at 600°C.

Introduction

Among nickel-iron based superalloys, INCONEL® alloy 718 is widely used and especially in aircraft engine building [1]. It is a high-strength and corrosion-resistant alloy for applications in a large range of temperature from -253°C to 760°C [2]. It is important to note that this superalloy can be applied for superplastic forging and forming if the grain size is in a range of 5 to 10 µm which is usually required for superplasticity [1, 3]. Huang and co-authors have shown that with such a structure the maximum value of the strain rate sensitivity was ~0.37 for an elongation of ~200% (testing temperature 965°C, strain rate of 10-4 s-1 [4]). In similar conditions (grain size 6-8 µm at 982°C and strain rate 5×10-4 s-1) Mahoney reports a strain rate sensitivity of 0.5 for an elongation of δ=500% [1]. It is well known that grain refinement down to NS condition can be achieved in metals and alloys using SPD via HPT, equal channel angular extrusion and MIF [5]. In previous studies it was demonstrated that grain refinement of INCONEL 718 can be achieved by MIF (grain size down to 80 nm) and by HPT (mean grain size of about 30 nm) [6, 7]. Such NS state is characterized by large internal elastic stresses and non-equilibrium grain boundaries [5], but also unique properties. Indeed, the NS alloy 718 with a grain size of about 80 nm exhibited an elongation of 580% at 700°C for a strain rate of 3×10-4 s-1 [8]. It is also important to note that this alloy exhibits some superplastic properties at a temperature as low as 600°C with a strain rate sensitivity m=0.4 corresponding to the so called low temperature superplasticity. The thermal stability investigations of this microstructure upon annealing have shown that increasing the temperature from 600 to 700°C (2 hours), or heating at 800°C (5 minutes) gives rise to a structure closer to the equilibrium state with a grain size of about 0.5 µm, still in the ultrafine range. The high strength of INCONEL 718 results from the large density of nanoscaled ordered precipitates embedded in the austenitic matrix based on Ni, Fe, Cr with an face-centered cubic (FCC) structure. The major strengthening comes from the metastable γ′′- phase (Ni3Nb, body-centered tetragonal), while some additional strength arises from precipitates of the γ′- phase (Ni3Al(Ti), FCC), carbides and the incoherent δ-phase (Ni3Nb, orthorhombic) [9].

Materials Science Forum Vol. 683 (2011) pp 127-135Online available since 2011/May/17 at www.scientific.net© (2011) Trans Tech Publications, Switzerlanddoi:10.4028/www.scientific.net/MSF.683.127

All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of TTP,www.ttp.net. (ID: 68.181.176.15, University of Southern California, Los Angeles, United States of America-03/04/14,02:53:32)

SPD induced phase transformations, like precipitate or carbide decomposition have been reported in the literature for various alloys [10-14]. In the present alloy the mechanical properties do not only depend on the grain size but also on second phase nanoscaled particles. Thus, it is important to characterize these precipitates (density, morphology, size) accurately at each step of the process. Therefore, in the present study the microstructure of the alloy was systematically investigated by transmission Electron Microscopy (TEM) and Atom probe Tomography (APT) with a special emphasis on the relationship between observed microstructural features and the hardness mechanical behavior. Model (HPT) and larger scale (MIF) experiments were performed.

Material and Experimental Methods

The material selected for this study was a hot-deformed nickel based alloy 718. The chemical composition of the alloy is given in Table 1. Bulk NS samples were obtained by SPD via HPT or MIF. Samples with a diameter of ∅4 mm and a thickness of 0.7 mm were machined for HPT experiments. Torsion was carried out on 5 turns at room temperature. At the end of the HPT process, the dimensions of samples were about 9 mm in diameter with a thickness of 0.25 mm. Specimens for microstructure observations and hardness measurements were cut a distance of 2-3 mm from the HPT disc center corresponding to a plastic strain of about 90.

Table 1. Chemical composition of the INCONEL 718 that was studied

Element Cr Al Ti Fe Nb Mo Co B C Ni

Concentration, [wt. %]

19 0.5 0.9 18.5 5.1 3.0 0.1 0.025 0.04 Bal.

Concentration, [at. %]

21.15 1.07 1.09 19.18 3.18 1.81 0.1 0.13 0.19 Bal.

MIF is a process in which a billet is compressed between flat dies along three orthogonal

directions during a sequence of forging passes. This process provides an accumulation of high plastic strains with little change in billet size or shape. MIF was carried out using hydraulic press of 630 ton-force equipped with isothermal die-stack units with flat dies. Deformation was performed with decreasing the deformation temperature from 950 to 575°С at strain-rates ranging from 8×10-4 to 10-1 s-1. The last deformation was upsetting of the sample at 575°C down to 10-12 mm in one direction. Strains were treated as additive over successive forging passes and total true strain was of about 50. Initial sample dimensions were 40-60 mm in diameter and 10-12 mm height [6-8].

Annealing at 600 °C during 1, 2, 5, 10 hours of thin samples were carried out using closed container with small volume of air.

Microhardness and hardness were carried out by standard methods. Atom probe tomography specimens were prepared by standard electropolishing methods.

Analyses were carried out using an Energy Compensated Atom Probe (80K - 20% pulse fraction - 2 kHz) and Laser Assisted Tomographic Atom Probe (λ = 515 nm, 2mW, 10 kHz, 20K) but only for the SPD processed material that exhibits an extremely high rate of specimen failure during field evaporation. Both instruments have a sufficient mass resolution (FWHM M/∆M ~ 1000 in electric mode and ~ 200 in laser mode) to allow precise composition measurements. However, due to the large amount of alloying elements, there are several overlaps on mass spectra (25 a.m.u. 50Ti2+/50Cr2+/50V

2+ ; 27 a.m.u. 54Cr2+/54Fe2+ ; 29 a.m.u. 58Fe2+/58Ni2+ ; 31 a.m.u. 93Nb3+/62Ni2+ ; 32 a.m.u. 64Ni2+/96Mo3+). Compositions were estimated by peak deconvolution on the basis of the natural abundance of each isotope. The precipitate composition was estimated by filtering the data the following way: %Cr > 15at.% for the matrix, %Al > 4at.% for γ′ precipitates, %Al < 2at.% and %Cr < 5at.% for γ′′ precipitates). It should be noted that during field evaporation with the laser

128 Advanced Mechanical Properties and Deformation Mechanisms of BulkNanostructured Materials

pulses, the lower electric field does not allow a quantitative measurements of Al content because of the overlap between Al+ and 54Fe2+ ions (leading also to a slight overestimation of Fe). Data processing was performed using the GPM 3D Data software ®.

The microstructure of the specimens was characterized using a JEOL JEM-2000EX and a JEM-2100 transmission electron microscopes (TEM), JXA-6400 scanning electron microscope (SEM) and Axiovert-100А optical microscope (OM). TEM foils were made by Tenupol-5. OM and SEM images were taken from specimens etched with Marble`s reagent (4 g CuSO4, 20 ml water, 20 ml HCl).

Experimental Results and Discussion

Initial state. The initial microstructure of the investigated alloy is fully recrystallized CG with a mean grain size of 40 µm (Fig. 1a). There are also large carbides (MC) with an average size of 5 µm. TEM studies have shown that coherent disk-type γ′′-precipitates are uniformly distributed within grains. The diameter of γ′′-phase disks is about 60 nm, and their thickness is 20 nm (Fig. 1b) in agreement with other published studies [15]. These precipitates were also analyzed by APT (Fig. 2). As already observed by other authors [16] they are sometime surrounded by a layer of γ′ phase (see the Al rich layer around the Ti and Nb rich particles in Fig. 2b and 2c). Carbides and δ phase were not analyzed but the composition of all other phases (namely the FCC matrix, γ′ and γ′′ precipitates) are given in Table 2, they are in good agreement with data reported in the literature for similar alloys [16].

(a) (b)

Fig. 1. Microstructure of the as-received INCONEL 718: OM (a); bright field TEM image (b.) INCONEL 718 after SPD. After HPT, the INCONEL 718 alloy exhibits an average grain size

of only about 30 nm (Fig. 3). Because of the small grain size and the complicated contrast resulting from internal elastic stresses it was impossible to image any precipitates (namely γ′ or γ′′), therefore the deformed material was analyzed by APT. However, unfortunately in such a nanostructured state samples were prone to early specimen failure during the field evaporation process. Therefore, it was possible to collect only small data sets corresponding to small volumes. In all these volumes, Nb (respectively Al, Ti) rich and Cr, Fe poor regions corresponding to the γ′′ phase (respectively to the γ′ phase) were never detected. For comparison with the undeformed state, the largest collected volume is displayed in the Fig. 4. The composition of all analyzed volumes is given in Table 3 and it should be noted that the Al, Nb and Ti concentration is often higher than the composition of the matrix before deformation (see Table 2). A composition profile was also plotted along the horizontal direction of the volume displayed in the Fig. 4. This profile clearly shows that there is a Nb concentration gradient (about 1 at.% Nb on the left versus about 3 at.% on the right, i.e. close to

Materials Science Forum Vol. 683 129

the nominal composition, see Table 1). Thus, these features seem to indicate that γ′ and γ′′ may have been partly decomposed during the HPT process and the grain fragmentation.

Fig. 2. 3D reconstruction of a volume analyzed in the INCONEL 718 before deformation (as-received state). Density map of Fe, Cr and Nb showing the distribution of nanoscaled γ′′ precipitates

(a). Selected part showing the distribution of Al and Ti atoms (b), the distribution of Nb and Al atoms (c) and the distribution of Fe and Cr atoms (d).

Table 2. Composition of the volume analyzed in the INCONEL 718 before deformation (see Fig. 2),

composition of the γ′, γ′′ precipitates and of the matrix [at. %]

Elem. Si P Cu V Co Al Cr Mo Mn Ni Fe Ti Nb all 0.16 0.02 0.04 0.04 0.54 1.01 19.35 1.96 0.1 52.02 19.9 1.33 3.53 ± 0.01 0.005 0.006 0.007 0.02 0.03 0.14 0.05 0.01 0.17 0.14 0.04 0.06 γγγγ′′′′ 0.13 0.04 0.04 0.05 0.32 7.74 7.36 0.99 0.14 61.29 8.24 5.38 8.28 ± 0.06 0.03 0.03 0.04 0.1 0.5 0.5 0.2 0.07 0.9 0.5 0.4 0.5 γγγγ′′′′′′′′ 0.05 0.03 0.03 0.04 0.24 0.46 1.8 1.55 0.06 71.7 1.87 5.41 16.76 ± 0.03 0.02 0.02 0.03 0.07 0.1 0.2 0.2 0.03 0.7 0.2 0.3 0.5

matrix 0.16 0.02 0.04 0.04 0.57 0.76 22.26 2.01 0.11 49.16 22.32 0.71 1.84 ± 0.01 0.06 0.007 0.008 0.03 0.03 0.2 0.05 0.01 0.2 0.15 0.03 0.05

In bulk samples of INCONEL 718 subjected to MIF with a deformation temperature decreased

down to 575°C, a uniform NS structure with grain size of about 80 nm was achieved (Fig. 5a). Most of grain boundaries exhibit high angle misorientations and locally high dislocation densities can be observed. Moreover, it is known that deformation at high temperatures results in the transformation of γ′′-phase into orthorhombic δ-phase with non-coherent boundaries [15]. As indicated on the SEM image, grains of δ-phase are uniformly distributed (Fig. 5b) giving rise to a duplex structure consisting of γ+δ-phases.

130 Advanced Mechanical Properties and Deformation Mechanisms of BulkNanostructured Materials

(a) (b)

Fig. 3. INCONEL 718 nanostructured by HPT (5 turns): TEM bright field image and corresponding electron-diffraction pattern (inset) (a); Grain size distribution (b).

Table 3. Compositions measured in the volumes analyzed by APT in the INCONEL 718 processed by HPT (only major elements with a strong partionning between the matrix and γ′ and γ′′

precipitates are listed) Element /

concentration [at. %]

Al Cr Ni Fe Ti Nb Collected

atoms Volume [nm3]

vol. 1 0.9 20.1 50.2 22.3 1 2.2 12000 4×4×25

± 0.1 0.7 0.9 0.7 0.1 0.2 vol. 2 0.7 22.1 48.5 23.7 0.4 1.6

600 4×4×15 ± 0.2 1.0 1.2 1.0 0.1 0.3

vol. 3 1.1 20.2 50.8 21.9 0.75 2.12 14500 5×5×20

0.2 0.7 0.8 0.7 0.1 0.2 vol. 4 0.76 20.4 50.8 21.5 1 1.9

69700 7×7×44 ± 0.06 0.3 0.4 0.3 0.07 0.1

vol. 5 (laser) - 21.5 51.1 21.5 1.5 2.3 126000 11×11×30

± 0.2 0.3 0.2 0.07 0.08 Evolution of the microstructure upon annealing at 600°C. During annealing at 600°C of the

INCONEL 718 nanostructured by HPT, the proportion of equilibrium grain boundaries increases. After 10 h at 600°C, the mean grain size has increased up to about 60 nm but it was still impossible to observe γ′′- precipitates (Fig. 6a). However, splitting of some γ-phase (FCC matrix) diffraction spots and superlattice reflections typical of the δ-phase were found on electron-diffraction pattern (set in Fig. 6b). These features indicate that during such long time annealing (10 h) at 600°C, the equilibrium δ-phase nucleates and grows as proposed earlier in [17].

Similar features were observed in the INCONEL 718 nanostructured by MIF and subsequently aged at 600°C during several hours. The proportion of high-angle boundaries increases while a significant decrease of non-equilibrium boundaries is observed. A typical image of the microstructure after 2 hours aging is shown on the Fig. 7. Banded contrasts do appear clearly at grain boundaries. Such features are typical of equilibrium high-angle boundaries. There is also a significant grain growth during annealing at 600°C. After 10 hours of aging, the average grain size is about 170 nm (versus 80 nm just after MIF).

Materials Science Forum Vol. 683 131

Fig. 4. 3D reconstruction of a volume analyzed in the INCONEL 718 processed by HPT (5 turns). Density map of Fe, Cr and Nb showing the absence of nanoscaled γ′′ precipitates (a), distribution of Nb and Ti atoms (b), and concentration profile computed along the horizontal direction (c) showing

a Nb concentration gradient that may result from SPD induced partial decomposition of γ′′ precipitates.

(a) (b)

Fig. 5. INCONEL 718 subjected to MIF: TEM bright field image and corresponding electron-diffraction pattern (inset) (a) and SEM image (b).

132 Advanced Mechanical Properties and Deformation Mechanisms of BulkNanostructured Materials

(a) (b)

Fig. 6. TEM bright field images and corresponding electron-diffraction patterns (inset) of INCONEL 718 subjected to HPT (5 turns) and annealed at 600°C during 10 hours at

low (a) and high (b) magnification.

Fig. 7. TEM bright field image and corresponding SAED pattern of the INCONEL 718 subjected to MIF followed by 2 hours at 600°C.

Microhardness of the NS INCONEL 718 alloy. The microhardness of the INCONEL 718 alloy

in the as-received state and nanostructured by HPT and MIF is reported in Fig. 8. As expected, the reduction of the grain size leads to a significant strengthening. Indeed, according to the Hall-Petch relationship the strength properties of metals and alloys increase with decreasing grain size [5]. The tensile properties of alloy 718 with different grain sizes usually exhibit the same behavior [18]. However, the comparison between nanostructured alloys after HPT and MIF shows that both precipitates and grain size affect the hardness. Indeed, after MIF the microhardness is higher than after HPT (6.6 versus 6.1 GPa) while the grain size is larger (80 nm versus 30 nm). This difference could be explained by the partial decomposition of γ′ and γ′′ precipitates during HPT (Table 3 and Fig. 4) while in the alloy processed by MIF, there is a high density of δ-phase precipitates (Fig. 5) that could significantly affect the properties [9].

Materials Science Forum Vol. 683 133

(a) (b)

Fig. 8. Effect of annealing time at 600°C on the hardness of the NS INCONEL 718 alloy obtained by HPT (a) and MIF (b).

However, more surprisingly, after 2 hour annealing at 600°C, it is interesting to note that both samples, processed by HPT and MIF exhibit a higher hardness, but it is now much higher in the HPT material (from 6.1 to 8.3 GPa) than in the MIF (from 6.6 to 6.9 GPa). This difference might result from the smaller grain size after HPT while the age hardening might result from precipitation. However, it should be noted that the as-received material aged in similar conditions exhibits a significant decrease of its hardness (from 5.1 to 4.8 GPa). Thus, this unexpected increase of hardness may be connected with polygonization occurring during pre-recrystallization annealing or the generation of short-range ordered regions [19], but it is most probably related to the re-precipitation of nanoscaled γ′ and Nb-rich (γ′′ or δ) precipitates that were dissolved during SPD. APT analyses of the aged NS state will be soon carried out to fully clarify this point.

Summary

1. γ′′ precipitates were not observed by TEM in the NS alloy obtained by HPT and neither after further annealing at 600°C during 10 hours. 2. APT data show that γ′ and γ′′ precipitates of alloy 718 are probably at least partly dissolved during HPT. 3. Nanostructuring of INCONEL 718 by HPT and 2 hour annealing at 600°C gives rise to a microhardness up to 8.3 GPa at room temperature. 4. The maximum hardness was obtained after annealing during 2 hours at 600°C after the SPD process. Thus it is though that the nanoscaled γ′ and γ′′ precipitates dissolved during HPT have re-precipitated from the super saturated solid solution. 5. A larger grain size was achieved by MIF leading to a larger hardness (about 7 GPa), but a similar evolution of the microhardness upon annealing at 600°C is observed.

Acknowledgements

This study was supported by Russian Foundation for Basic Research, Projects No. 07-08-00287 а.

134 Advanced Mechanical Properties and Deformation Mechanisms of BulkNanostructured Materials

References

[1] M.W. Mahoney, in: Superalloy 718 - Metallurgy and Applications, edited by E.A. Loria TMS (1989), p. 391

[2] R.P. Jewett and J.A. Halchak, in: Superalloys 718, 625 and Various Derivatives, edited by E.A. Loria TMS (1991), p. 749

[3] G.D. Smith and H.L. Flower, in: Superalloys 718,625,706 and Various Derivatives, edited by E.A. Loria TMS (1994), p. 355

[4] Y. Huang, M. Strangwood and P.L. Blackwell: Mater. Sci. Technol. Vol. 16 (2000), p. 1309

[5] R.Z. Valiev and I.V. Aleksandrov: Nanostructured materials processed by severe plastic deformation (Logos, Russia 2000).

[6] V.A. Valitov, O.A. Kaibyshev, Sh.Kh. Mukhtarov, B.P. Bewlay and M.F.X. Gigliotti: Mater. Sci. Forum Vol. 357-359 (2001), p. 417

[7] Sh. Mukhtarov, V. Valitov and N. Dudova, in: Superalloys 718, 625, 706, and Various Derivatives, edited by E.A. Loria TMS (2005), p. 507

[8] V.A. Valitov, Sh.Kh. Mukhtarov and Yu.A. Raskulova: The Physics of Metals and Metallography Vol. 102, No. 1 (2006), p. 97

[9] Y. Huang and T.G. Langdon: J. Mater. Sci. Vol. 42 (2007), p. 421

[10] V.V. Sagaradze and V.A. Shabashov: Nanostruct. Mater. Vol. 9 (1997), p. 681

[11] M. Muruyama, K. Hono and Z. Horita: Mater. Trans. JIM Vol. 40 (1999), p. 938

[12] Y. Ivanisenko, I. MacLaren, X. Sauvage, R.Z. Valiev and H.-J. Fecht: Acta Mater. Vol. 54 (2006), p. 1659

[13] X. Sauvage, C. Genevois, G. Da Costa and V. Pantsyrny: Scripta Mater. Vol. 61 (2009), p. 660

[14] X. Sauvage, A. Chbihi and X. Quelennec: Journal of Physics: Conference Series Vol. 240 (2010), 012003

[15] C.T. Sims, N.S. Stoloff, and W.C. Hagel: Superalloys II: High Temperature Materials for Aerospace and Industrial Power (Wiley-Interscience, USA 1987)

[16] M.K. Miller, S.S. Babi and M.G. Burke: Mater. Sci. Eng. A Vol. 327 (2002), p. 84

[17] Sh. Mukhtarov, V. Valitov and N. Dudova: Rev. Adv. Mater. Sci. Vol. 25 (2010), in press.

[18] Sh. Mukhtarov: Mater. Sci. Forum Vol. 633-634 (2010), p. 569

[19] I.I. Novikov: Theory of metals heat treatment (Metallurgiya, Moscow 2000).

Materials Science Forum Vol. 683 135

Advanced Mechanical Properties and Deformation Mechanisms of Bulk Nanostructured Materials 10.4028/www.scientific.net/MSF.683 Microstructure and Microhardness of a Nanostructured Nickel-Iron Based Alloy 10.4028/www.scientific.net/MSF.683.127

DOI References

[4] Y. Huang, M. Strangwood and P.L. Blackwell: Mater. Sci. Technol. Vol. 16 (2000), p.1309.

doi:10.1179/026708300101507154 [6] V.A. Valitov, O.A. Kaibyshev, Sh. Kh. Mukhtarov, B.P. Bewlay and M.F.X. Gigliotti: Mater. Sci. Forum

Vol. 357-359 (2001), p.417.

doi:10.4028/www.scientific.net/MSF.357-359.417 [8] V.A. Valitov, Sh. Kh. Mukhtarov and Yu.A. Raskulova: The Physics of Metals and Metallography Vol.

102, No. 1 (2006), p.97.

doi:10.1134/S0031918X06070143 [9] Y. Huang and T.G. Langdon: J. Mater. Sci. Vol. 42 (2007), p.421.

http://dx.doi.org/10.1007/s10853-006-0483-z [10] V.V. Sagaradze and V.A. Shabashov: Nanostruct. Mater. Vol. 9 (1997), p.681.

doi:10.1016/S0965-9773(97)00150-5 [12] Y. Ivanisenko, I. MacLaren, X. Sauvage, R.Z. Valiev and H. -J. Fecht: Acta Mater. Vol. 54 (2006),

p.1659.

doi:10.1016/j.actamat.2005.11.034 [13] X. Sauvage, C. Genevois, G. Da Costa and V. Pantsyrny: Scripta Mater. Vol. 61 (2009), p.660.

doi:10.1016/j.scriptamat.2009.06.007 [14] X. Sauvage, A. Chbihi and X. Quelennec: Journal of Physics: Conference Series Vol. 240 (2010),

012003.

doi:10.1088/1742-6596/240/1/012003 [16] M.K. Miller, S.S. Babi and M.G. Burke: Mater. Sci. Eng. A Vol. 327 (2002), p.84.

http://dx.doi.org/10.1016/S0921-5093(01)01881-0 [18] Sh. Mukhtarov: Mater. Sci. Forum Vol. 633-634 (2010), p.569.

http://dx.doi.org/10.4028/www.scientific.net/MSF.633-634.569