- Email: [email protected]
Effect of Mo on the thermal stability of γ′ precipitate in Inconel 740 alloy
M A TE RI A L S C HA RACT ER I ZA TI O N 95 ( 20 1 4 ) 1 8 0– 1 8 6
Available online at www.sciencedirect.com
ScienceDirect www.elsevier.com/locate/matchar
Effect of Mo on the thermal stability of γ′ precipitate in Inconel 740 alloy Gyeong Su Shin, Jae Yong Yun, Myung Chul Park, Seon Jin Kim⁎ Division of Materials Science and Engineering, Hanyang University, 17 Haengdang-dong, Seongdong-gu, Seoul 133-791, Republic of Korea
AR TIC LE D ATA
ABSTR ACT
Article history:
The effect of Mo on gamma prime (γ′) coarsening kinetics of Ni–25Cr–20Co–1.8Ti–0.9Al–
Received 9 January 2014
2Nb–0.03C–xMo (x = 0.5, 3, 6 wt.%) alloys was investigated with respect to the activation
Received in revised form 18 June 2014
energy for γ′ precipitate coarsening. The coarsening rates were measured at temperatures
Accepted 20 June 2014
of 760 °C, 780 °C, and 810 °C up to 500 h. Coarsening rate decreased with increasing Mo, and the activation energies for coarsening were determined to be 245 kJ/mol, 261 kJ/mol, and 278 kJ/mol for 0.5 wt.% Mo, 3 wt.% Mo, and 6 wt.% Mo, respectively.
Keywords:
© 2014 Elsevier Inc. All rights reserved.
Nickel based superalloys Coarsening
1. Introduction Nickel-based superalloys are widely studied as a candidate material within coal-fired power plants due to their high creep strength and excellent corrosion and oxidation resistance at high temperatures. Among the nickel-based superalloys, Inconel 740 was developed for steam boiler tubing in advanced ultra-supercritical fossil power plants [1]. The high temperature properties of Inconel 740 are known to be the result of fine dispersions of the γ′ precipitate. Hence, after exposure of this alloy to high temperatures, the mean radius of the γ′ precipitates increases due to coarsening, and this coarsening may result in degradation of the thermal stability. To achieve high thermal stability, it is necessary to study the kinetics of γ′-precipitate coarsening behavior. In previous investigations, the coarsening rate was controlled by volume diffusion [2,3]. For that reason, numerous studies have examined the influence of alloying elements affecting the diffusivity of solute atoms so that the coarsening rate can be reduced [4–6].
⁎ Corresponding author. E-mail address: [email protected] (S.J. Kim).
http://dx.doi.org/10.1016/j.matchar.2014.06.019 1044-5803 © 2014 Elsevier Inc. All rights reserved.
In previous investigations, Fährmann et al. [7] observed that the coarsening rate decreased and the activation energy for coarsening increased with increasing Mo content in the Ni–Al–Mo system. It was also reported that the solute controlling the coarsening rate was Al in the Mo concentration range from 8 wt.% to 13 wt.% and Mo in the Mo concentration of 20 wt.%. Namely, the solute controlling the coarsening rate changed to Mo with increasing Mo content. As a consequence of this, it was reported that the activation energy for coarsening increased and the coarsening rate decreased. Thus, the Mo played a role in the higher thermal stability. Hence, it was considered that Inconel 740 with added Mo increased the thermal stability. However, it was confirmed that the solute controlling the coarsening rate for Inconel 740 was Al or Ti [8]. Therefore, the effect of the Mo content on the coarsening rate, the activation energy for coarsening and partition coefficient of Ni–25Cr–20Co–1.8Ti–0.9Al–2Nb–0.03C– xMo (x = 0.5, 3, 6 wt.%) alloys were investigated in this study over a temperature range of 760 °C to 810 °C.
181
M A TE RI A L S C HA RACT ER I ZA TI O N 95 ( 20 1 4 ) 1 8 0– 1 8 6
2. Experimental procedure Ni–25Cr–20Co–1.8Ti–0.9Al–2Nb–0.03C–xMo (x = 0.5, 3, 6 wt.%) alloys were arc melted under an argon atmosphere using a water-cooled copper crucible and a non-consumable tungsten electrode. Each alloy was inverted and remelted at least 10 times to ensure homogeneity. The chemical composition of alloys was analyzed using an optical emission spectrometer, and the results are summarized in Table 1. Specimens were solution-annealed at 1200 °C for 1 h and water quenched. Then, solutions were aged at 800 °C for 16 h and cooled in air. In this study, specimens were aged at 760 °C, 780 °C, and 810 °C for 50, 100, 150, 200, and 500 h to examine the effect of aging temperature and time on the coarsening of γ′ precipitates. The morphological evolution was observed using field emission scanning electron microscopy (FE-SEM). Specimens were prepared for FE-SEM by mechanical grinding and polishing followed by electro-etching in a solution of 170 mL H3PO4 + 10 mL H2SO4 + 10 g CrO3 at 3–4 V for 8–12 s at room temperature. The average precipitate size was measured via FE-SEM images using semi-automatic image analysis software. The radius of γ′ precipitate, r, was calculated from the pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi traced areas using a circular-equivalence, i.e. r ¼ area=π , where the measurements were approximately equivalent to those determined from a planar section. Individual outlines of the precipitates' edges were labeled with a unique grayscale value employing a subroutine, BinaryLabel8, contained in a morphology collection plug-in developed for ImageJ [6]. To minimize the measurement error, more than 400 γ′ precipitates were measured in different areas of the samples. Transmission electron microscopy (TEM) samples were fabricated by punching 3-mm diameter discs from 100- to 150-μm thick foils, which were produced by mechanical thinning. The 3-mm diameter discs were then jet electropolished employing a solution of 5 vol.% perchloric acid in methanol at −20 °C. Scanning transmission electron microscopy (STEM) and energy dispersive X-ray spectroscopy (EDS) were performed utilizing a JEOL 2100 operating at 200 kV. EDS was collected on 2048 channels and the acquired data was quantified using the Cliff– Lorimer approximations. XRD was performed between 20° and 90° (2θ) at room temperature using a diffractometer with a Cu target and monochromator, a step size of 0.05°. A current of 50 mA and a voltage of 40 kV were employed. XRD pattern was shown Fig. 1. It was observed that there is no carbide.
Table 1 – Chemical compositions of the specimens (wt.%). Specimen Mo Co 0.5 Mo 3 Mo 6 Mo
0.4 2.7 5.8
Cr
Al Ti Nb Si Mn
19.8 24.9 0.9 1.8 2.1 0.5 19.9 25.4 0.8 1.9 2.0 0.5 20.7 26.4 0.9 1.8 2.0 0.5
0.3 0.3 0.2
C
Ni
0.02 0.7 Bal. 0.02 0.8 Bal. 0.03 0.5 Bal.
of precipitate shape from spherical to cubic. It was reported that the transition of γ′ precipitate shape depended on the total matrix/precipitate strain present due to the lattice mismatch [9]. The size of the γ′ precipitate became larger maintaining the spherical shape before the effects of strain were sufficient to influence the precipitate shape. Namely, it should be noted that for alloys of low misfit, the γ′ precipitate maintained the spherical shape for long. In our experiment, it was observed that the γ′ precipitate maintained for long with increasing Mo contents. This could suggest that the misfit decreased with increasing Mo contents, which is consistent with that of Wang et al. [10]. In addition, no coalescence behavior was evident in this work.
3.2. Coarsening kinetics of the precipitate The sizes of the γ′ precipitates for the alloys aged at 760 °C, 780 °C, and 810 °C for different times are summarized in Table 2. At 760 °C, the mean size of the γ′ precipitates for the 0.5 Mo alloy grew from 19.1 nm after 50 h to 35.5 nm after 500 h. At the same temperature, the mean size of the γ′ precipitates for the 3 Mo alloy grew from 17.6 nm after 50 h to 27.8 nm after 500 h, and the mean size of the γ′ precipitates for the 6 Mo alloy grew from 17.1 nm after 50 h to 26.6 nm after 500 h. At 500 h, the mean size of the γ′ precipitates for the 0.5 Mo alloy grew from 35.5 nm at 760 °C to 54.0 nm at 810 °C. At same time, the mean size of the γ′ precipitates for the 3 Mo alloy grew from 27.8 nm at 760 °C to 41.9 nm at 810 °C, and the mean size of the γ′ precipitates for the 6 Mo alloy grew from 26.6 nm at 760 °C to 41.0 nm at 810 °C. It was observed that the growth of the γ′ precipitate increased as the temperature and time increased, but most notably when
3. Results and discussion 3.1. Microstructural characterization The microstructures of alloys after aging are shown in Figs. 2–4. The γ′ precipitate shape for 3 Mo and 6 Mo alloys exhibited mainly a spherical morphology during aging. However, it was evident that the shape of γ′ precipitate consists of a cubic morphology over 200 h at 810 °C. For the 0.5 Mo alloy, the γ′ precipitate shape exhibited mainly a cubic shape during aging. This result implied that there was a transition
Fe
Fig. 1 – XRD pattern of the 6 Mo alloy.
182
M A TE RI A L S C HA RACT ER I ZA TI O N 95 ( 20 1 4 ) 1 8 0– 1 8 6
Fig. 2 – FESEM images of γ′ precipitates of the 0.5 Mo alloy aged at 760 °C, 780 °C and 810 °C for five different times.
temperature increased. It was found that the mean size of the γ′ precipitate for the 6 Mo alloy was slightly smaller than for the 3 Mo alloy at equal temperature and time. However, for 0.5 Mo, it was observed to be 30% larger than for other alloys. This indicated that the coarsening of the γ′ precipitate was suppressed by adding Mo content. The coarsening kinetics can be written as a linear volumetric growth law [8]: r3 −r30 ¼ kt
ð1Þ
where k is the coarsening rate coefficient, and r and r0 are the average precipitate radii at times t and 0, respectively. The coarsening rate can be determined from the slope of the r3 −r30 versus t plot, which is shown in Fig. 5. The values obtained for the coarsening rate are summarized in Table 3. It was evident that the coarsening of the γ′ precipitates obeys the linear relationship, r3 −r30. This result suggests that the coarsening of the γ′ precipitate follows the cubic kinetics of diffusion-controlled particle growth. As shown in Table 3, it was found that the coarsening rate decreased with increasing Mo content for all temperatures. It was
Fig. 3 – FESEM images of γ′ precipitates of the 3 Mo alloy aged at 760 °C, 780 °C and 810 °C for five different times.
183
M A TE RI A L S C HA RACT ER I ZA TI O N 95 ( 20 1 4 ) 1 8 0– 1 8 6
Fig. 4 – FESEM images of γ′ precipitates of the 6 Mo alloys aged at 760 °C, 780 °C and 810 °C for five different times.
observed that the coarsening rate for the 0.5 Mo alloy was about twice higher than other alloys.
3.3. Activation energy for coarsening and partition coefficient The activation energy for the coarsening and partition coefficient was investigated to confirm the effect of increasing Mo content on the coarsening rate. The activation energy for coarsening can be determined from the slope of the ln k versus 1 / T plot by consulting a previous study [11], which is shown in Fig. 6. The values obtained for the activation energy for coarsening are summarized in Table 4. As shown in Table 4, the activation energies for coarsening were approximately 245 kJ/mol for 0.5 the Mo alloy, 262 kJ/mol for the 3 Mo alloy and 278 kJ/mol for the 6 Mo alloy. The activation energy for coarsening of the 0.5 Mo alloy was consistent with the results by Zhao et al. [8], which have shown the activation energy for coarsening to be 247 kJ/mol in Inconel 740. The activation energy for coarsening was observed to increase by approximately 15 kJ/mol as the Mo concentration increased from 3 wt.% to
6 wt.%. This indicated that additions of Mo increased the activation energy for coarsening. The line profile across the γ′ precipitate and mapping with respect to the γ′ precipitate were investigated to confirm the partition coefficient, as shown in Figs. 7 and 8. It was observed that Al, Ti, and Nb were preferred to the γ′ precipitate, whereas Cr and Co were preferred to the γ matrix. Mo was uniformly distributed between the γ matrix and γ′ precipitate. The value for the partition coefficient is summarized in Table 5. Along with the activation energy for coarsening and the partition coefficient results in this work, the solute controlling the coarsening rate was considered to be Al. This result is consistent with the solute controlling the coarsening rate, as investigated by Fährmann et al. [7]. The decrease in the coarsening rate with increasing Mo contents can be attributed to the slow diffusion of Al in the γ matrix. According to Wang et al. [10], it was observed that the lattice parameter of the γ matrix increased with the increasing Mo content. This behavior affected the increase in the saddle-point energy, resulting in a decrease in the diffusion of
Table 2 – Mean γ′ sizes for alloys after long-term aging at 760 °C, 780 °C, and 810 °C. Specimen
γ′ size by treatment time (nm)
Temperature (°C) 50 h
0.5 Mo
3 Mo
6 Mo
760 780 810 760 780 810 760 780 810
19.1 21.31 25.2 17.67 19.28 22.04 17.16 18.46 21.56
± ± ± ± ± ± ± ± ±
0.02 0.20 0.23 0.24 0.05 0.86 0.55 0.83 0.18
100 h 23.2 27.5 33.3 19.60 21.73 26.04 18.73 20.84 25.46
± ± ± ± ± ± ± ± ±
0.13 0.31 0.54 0.44 0.67 0.13 0.33 0.54 0.96
150 h 26.0 30.6 38.1 20.96 23.82 29.09 20.08 22.78 28.44
± ± ± ± ± ± ± ± ±
0.58 0.20 0.74 0.71 0.83 0.52 0.36 0.33 0.4
200 h 28.4 33.9 42.1 22.29 25.01 31.38 21.26 24.43 30.89
± ± ± ± ± ± ± ± ±
0.51 0.71 0.34 0.33 0.56 0.77 0.15 0.02 0.43
500 h 35.5 42.8 54.0 27.75 32.43 41.94 26.56 31.46 40.95
± ± ± ± ± ± ± ± ±
0.61 0.16 0.61 0.28 0.64 0.78 0.43 0.21 0.24
184
M A TE RI A L S C HA RACT ER I ZA TI O N 95 ( 20 1 4 ) 1 8 0– 1 8 6
Fig. 5 – Average γ′ precipitate size versus aging time at 760 °C, 780 °C, and 810 °C; (a) 0.5 Mo alloy, (b) 3 Mo alloy and (c) 6 Mo alloy.
Al in the γ matrix. This was agreeable with the result of Stalker et al. [12]. They found that Mo reduces the diffusion of Al in Ni alloys. This fact indicated that the activation energy
Table 3 – k (x temperatures.
nm3 h−1)
values
at
various
aging
Temperature (°C)
0.5 Mo
3 Mo
6 Mo
760 780 810
65.6 ± 0.21 119.8 ± 0.71 247.1 ± 0.73
32.6 ± 0.22 57.7 ± 0.51 133.6 ± 0.29
28.6 ± 0.11 53.1 ± 0.11 128.2 ± 0.18
for coarsening increased and the coarsening rate decreased. Therefore, it was concluded that the decrease in the coarsening rate and the increase in the activation energy for coarsening with increasing Mo content were due to the slow diffusion of Al in the γ matrix.
4. Conclusions The coarsening kinetics of Ni–25Cr–20Co–1.8Ti–0.9Al–2Nb– 0.03C–xMo (x = 0.5, 3, 6 wt.%) alloys were investigated with respect to the activation energy for coarsening and the partition coefficient. The following conclusions could be made: 1. In the alloys, it was observed that the growth of the γ′ precipitate increased as the temperature and time increased, but most notably when temperature increased. It was found that the mean size of the γ′ precipitate for the 6 Mo alloy was slightly smaller than for the 3 Mo alloy at equal temperature and time. For 0.5 Mo, it was observed to be approximately 30% larger than other alloys.
Table 4 – Activation energy for coarsening for the alloys. Q (kJ/mol) Fig. 6 – Arrhenius plot of the coarsening rate coefficient k vs. the reciprocal of absolute temperature.
0.5 Mo 3 Mo 6 Mo
245.6 ± 5.3 261.6 ± 7.1 277.8 ± 11.0
M A TE RI A L S C HA RACT ER I ZA TI O N 95 ( 20 1 4 ) 1 8 0– 1 8 6
185
Fig. 7 – Line profile as a function of distance (nm) with respect to the γ′ precipitate for 6 Mo alloy aged at 760 °C for 500 h.
Fig. 8 – Mapping with respect to the γ′ precipitate for 6 Mo alloy aged at 760 °C for 500 h.
2. In the alloys, it was evidenced by the fact that the coarsening kinetics of the γ′ precipitate was controlled by the volume diffusion. The coarsening rate was shown to be slower for the 6 Mo alloy than for the 3 Mo alloy. For the 0.5 Mo alloy, the coarsening rate was shown to be approximately twice as fast as other alloys. 3. The activation energy for coarsening was approximately 245 kJ/mol for the 0.5 Mo alloy, 262 kJ/mol for the 3 Mo alloy and 278 kJ/mol for the 6 Mo alloy. It was observed that Al, Ti,
Table 5 – Measured composition (at.%) and partition coefficient for 6 Mo alloy aged at 760 °C for 500 h.
Cγ Cγ′ Cγ′/γ
Ni
Cr
Co
Mo
Al
Ti
Nb
38.89 56.03 1.44
34.46 13.01 0.38
21.48 11.35 0.53
2.05 1.78 0.87
1.26 7.48 5.94
0.93 8.53 9.20
0.38 1.82 4.77
and Nb were preferred to the γ′ precipitate while Cr and Co were preferred to the γ matrix. Mo was uniformly distributed between the γ matrix and γ′ precipitate. Along with the activation energy for coarsening and the partition-coefficient results in this work, the solute controlling the coarsening rate was considered to be Al. It was concluded that the decrease in the coarsening rate and the increase in the activation energy for coarsening with increasing Mo content were due to the slow diffusion of Al in the γ matrix.
Acknowledgment This work was supported by the Human resource Development program (No. 20114010203020) of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) grant funded by the Korea government Ministry of Trade, Industry and Energy.
186
M A TE RI A L S C HA RACT ER I ZA TI O N 95 ( 20 1 4 ) 1 8 0– 1 8 6
REFERENCES [7] [1] Zhao S, Dong J, Xie X, Smith GD, Patel SJ. Thermal stability study on a new Ni–Cr–Co–Mo–Nb–Ti–Al superalloy. Superalloys 2004:63–72. [2] Lifshitz M, Slozov VV. The kinetics of precipitation from supersaturated solid solutions. J Phys Chem Solids 1961;19:35–50. [3] Wagner C. Theorie der alterung von niederschlägen durch umlösen (Ostwald-reifung). Z Elektrochem 1961;65:581–91. [4] Wang WZ, Jin T, Liu JL, Sun XF, Guan HR, Hu ZQ. Role of Re and Co on microstructures and γ′ coarsening in single crystal superalloys. Mater Sci Eng A 2008;479:148–56. [5] Devaux A, Naze L, Molins R, Pineau A, Organista A, Guedou JY, Uginet JF, Heritier P. Gamma double prime precipitation kinetic in Alloy 718. Mater Sci Eng A 2008;486:117–22. [6] Sudbrack CK, Ziebell TD, Noebe RD, Seidman DN. Effects of a tungsten addition on the morphological evolution, spatial
[8]
[9] [10]
[11]
[12]
correlations and temporal evolution of a model Ni–Al–Cr superalloy. Acta Mater 2008;56:448–63. Fährmann M, Fährmann E, Pollock TM, Johnson WC. Element partitioning during coarsening of (γ–γ′) Ni–Al–Mo alloys. Metall Trans A 1997;28:1943–5. Zhao S, Xie X, Smith GD, Patel SJ. Gamma prime coarsening and age-hardening behaviors in a new nickel base superalloy. Mater Lett 2004;58:1784–7. Ricks RA, Porter AJ, Ecob RC. The growth of γ′ precipitates in nickel-base superalloys. Acta Metall 1983;31:43–53. Wang Tao, Sheng Guang, Liu Zi-Kui, Chen Long-Qing. Coarsening kinetics of γ′ precipitates in the Ni–Al–Mo system. Acta Mater 2008;56:5544–55. Ges AM, Fornaro O, Palacio HA. Coarsening behaviour of a Ni-base superalloy under different heat treatment conditions. Mater Sci Eng A 2007;458:96–100. Stalker MK, Morral JE, Romig Jr AD. Application of the square root diffusivity to diffusion in Ni–Cr–Al–Mo alloys. Metall Trans A 1992;23A:3245–9.
Available online at www.sciencedirect.com
ScienceDirect www.elsevier.com/locate/matchar
Effect of Mo on the thermal stability of γ′ precipitate in Inconel 740 alloy Gyeong Su Shin, Jae Yong Yun, Myung Chul Park, Seon Jin Kim⁎ Division of Materials Science and Engineering, Hanyang University, 17 Haengdang-dong, Seongdong-gu, Seoul 133-791, Republic of Korea
AR TIC LE D ATA
ABSTR ACT
Article history:
The effect of Mo on gamma prime (γ′) coarsening kinetics of Ni–25Cr–20Co–1.8Ti–0.9Al–
Received 9 January 2014
2Nb–0.03C–xMo (x = 0.5, 3, 6 wt.%) alloys was investigated with respect to the activation
Received in revised form 18 June 2014
energy for γ′ precipitate coarsening. The coarsening rates were measured at temperatures
Accepted 20 June 2014
of 760 °C, 780 °C, and 810 °C up to 500 h. Coarsening rate decreased with increasing Mo, and the activation energies for coarsening were determined to be 245 kJ/mol, 261 kJ/mol, and 278 kJ/mol for 0.5 wt.% Mo, 3 wt.% Mo, and 6 wt.% Mo, respectively.
Keywords:
© 2014 Elsevier Inc. All rights reserved.
Nickel based superalloys Coarsening
1. Introduction Nickel-based superalloys are widely studied as a candidate material within coal-fired power plants due to their high creep strength and excellent corrosion and oxidation resistance at high temperatures. Among the nickel-based superalloys, Inconel 740 was developed for steam boiler tubing in advanced ultra-supercritical fossil power plants [1]. The high temperature properties of Inconel 740 are known to be the result of fine dispersions of the γ′ precipitate. Hence, after exposure of this alloy to high temperatures, the mean radius of the γ′ precipitates increases due to coarsening, and this coarsening may result in degradation of the thermal stability. To achieve high thermal stability, it is necessary to study the kinetics of γ′-precipitate coarsening behavior. In previous investigations, the coarsening rate was controlled by volume diffusion [2,3]. For that reason, numerous studies have examined the influence of alloying elements affecting the diffusivity of solute atoms so that the coarsening rate can be reduced [4–6].
⁎ Corresponding author. E-mail address: [email protected] (S.J. Kim).
http://dx.doi.org/10.1016/j.matchar.2014.06.019 1044-5803 © 2014 Elsevier Inc. All rights reserved.
In previous investigations, Fährmann et al. [7] observed that the coarsening rate decreased and the activation energy for coarsening increased with increasing Mo content in the Ni–Al–Mo system. It was also reported that the solute controlling the coarsening rate was Al in the Mo concentration range from 8 wt.% to 13 wt.% and Mo in the Mo concentration of 20 wt.%. Namely, the solute controlling the coarsening rate changed to Mo with increasing Mo content. As a consequence of this, it was reported that the activation energy for coarsening increased and the coarsening rate decreased. Thus, the Mo played a role in the higher thermal stability. Hence, it was considered that Inconel 740 with added Mo increased the thermal stability. However, it was confirmed that the solute controlling the coarsening rate for Inconel 740 was Al or Ti [8]. Therefore, the effect of the Mo content on the coarsening rate, the activation energy for coarsening and partition coefficient of Ni–25Cr–20Co–1.8Ti–0.9Al–2Nb–0.03C– xMo (x = 0.5, 3, 6 wt.%) alloys were investigated in this study over a temperature range of 760 °C to 810 °C.
181
M A TE RI A L S C HA RACT ER I ZA TI O N 95 ( 20 1 4 ) 1 8 0– 1 8 6
2. Experimental procedure Ni–25Cr–20Co–1.8Ti–0.9Al–2Nb–0.03C–xMo (x = 0.5, 3, 6 wt.%) alloys were arc melted under an argon atmosphere using a water-cooled copper crucible and a non-consumable tungsten electrode. Each alloy was inverted and remelted at least 10 times to ensure homogeneity. The chemical composition of alloys was analyzed using an optical emission spectrometer, and the results are summarized in Table 1. Specimens were solution-annealed at 1200 °C for 1 h and water quenched. Then, solutions were aged at 800 °C for 16 h and cooled in air. In this study, specimens were aged at 760 °C, 780 °C, and 810 °C for 50, 100, 150, 200, and 500 h to examine the effect of aging temperature and time on the coarsening of γ′ precipitates. The morphological evolution was observed using field emission scanning electron microscopy (FE-SEM). Specimens were prepared for FE-SEM by mechanical grinding and polishing followed by electro-etching in a solution of 170 mL H3PO4 + 10 mL H2SO4 + 10 g CrO3 at 3–4 V for 8–12 s at room temperature. The average precipitate size was measured via FE-SEM images using semi-automatic image analysis software. The radius of γ′ precipitate, r, was calculated from the pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi traced areas using a circular-equivalence, i.e. r ¼ area=π , where the measurements were approximately equivalent to those determined from a planar section. Individual outlines of the precipitates' edges were labeled with a unique grayscale value employing a subroutine, BinaryLabel8, contained in a morphology collection plug-in developed for ImageJ [6]. To minimize the measurement error, more than 400 γ′ precipitates were measured in different areas of the samples. Transmission electron microscopy (TEM) samples were fabricated by punching 3-mm diameter discs from 100- to 150-μm thick foils, which were produced by mechanical thinning. The 3-mm diameter discs were then jet electropolished employing a solution of 5 vol.% perchloric acid in methanol at −20 °C. Scanning transmission electron microscopy (STEM) and energy dispersive X-ray spectroscopy (EDS) were performed utilizing a JEOL 2100 operating at 200 kV. EDS was collected on 2048 channels and the acquired data was quantified using the Cliff– Lorimer approximations. XRD was performed between 20° and 90° (2θ) at room temperature using a diffractometer with a Cu target and monochromator, a step size of 0.05°. A current of 50 mA and a voltage of 40 kV were employed. XRD pattern was shown Fig. 1. It was observed that there is no carbide.
Table 1 – Chemical compositions of the specimens (wt.%). Specimen Mo Co 0.5 Mo 3 Mo 6 Mo
0.4 2.7 5.8
Cr
Al Ti Nb Si Mn
19.8 24.9 0.9 1.8 2.1 0.5 19.9 25.4 0.8 1.9 2.0 0.5 20.7 26.4 0.9 1.8 2.0 0.5
0.3 0.3 0.2
C
Ni
0.02 0.7 Bal. 0.02 0.8 Bal. 0.03 0.5 Bal.
of precipitate shape from spherical to cubic. It was reported that the transition of γ′ precipitate shape depended on the total matrix/precipitate strain present due to the lattice mismatch [9]. The size of the γ′ precipitate became larger maintaining the spherical shape before the effects of strain were sufficient to influence the precipitate shape. Namely, it should be noted that for alloys of low misfit, the γ′ precipitate maintained the spherical shape for long. In our experiment, it was observed that the γ′ precipitate maintained for long with increasing Mo contents. This could suggest that the misfit decreased with increasing Mo contents, which is consistent with that of Wang et al. [10]. In addition, no coalescence behavior was evident in this work.
3.2. Coarsening kinetics of the precipitate The sizes of the γ′ precipitates for the alloys aged at 760 °C, 780 °C, and 810 °C for different times are summarized in Table 2. At 760 °C, the mean size of the γ′ precipitates for the 0.5 Mo alloy grew from 19.1 nm after 50 h to 35.5 nm after 500 h. At the same temperature, the mean size of the γ′ precipitates for the 3 Mo alloy grew from 17.6 nm after 50 h to 27.8 nm after 500 h, and the mean size of the γ′ precipitates for the 6 Mo alloy grew from 17.1 nm after 50 h to 26.6 nm after 500 h. At 500 h, the mean size of the γ′ precipitates for the 0.5 Mo alloy grew from 35.5 nm at 760 °C to 54.0 nm at 810 °C. At same time, the mean size of the γ′ precipitates for the 3 Mo alloy grew from 27.8 nm at 760 °C to 41.9 nm at 810 °C, and the mean size of the γ′ precipitates for the 6 Mo alloy grew from 26.6 nm at 760 °C to 41.0 nm at 810 °C. It was observed that the growth of the γ′ precipitate increased as the temperature and time increased, but most notably when
3. Results and discussion 3.1. Microstructural characterization The microstructures of alloys after aging are shown in Figs. 2–4. The γ′ precipitate shape for 3 Mo and 6 Mo alloys exhibited mainly a spherical morphology during aging. However, it was evident that the shape of γ′ precipitate consists of a cubic morphology over 200 h at 810 °C. For the 0.5 Mo alloy, the γ′ precipitate shape exhibited mainly a cubic shape during aging. This result implied that there was a transition
Fe
Fig. 1 – XRD pattern of the 6 Mo alloy.
182
M A TE RI A L S C HA RACT ER I ZA TI O N 95 ( 20 1 4 ) 1 8 0– 1 8 6
Fig. 2 – FESEM images of γ′ precipitates of the 0.5 Mo alloy aged at 760 °C, 780 °C and 810 °C for five different times.
temperature increased. It was found that the mean size of the γ′ precipitate for the 6 Mo alloy was slightly smaller than for the 3 Mo alloy at equal temperature and time. However, for 0.5 Mo, it was observed to be 30% larger than for other alloys. This indicated that the coarsening of the γ′ precipitate was suppressed by adding Mo content. The coarsening kinetics can be written as a linear volumetric growth law [8]: r3 −r30 ¼ kt
ð1Þ
where k is the coarsening rate coefficient, and r and r0 are the average precipitate radii at times t and 0, respectively. The coarsening rate can be determined from the slope of the r3 −r30 versus t plot, which is shown in Fig. 5. The values obtained for the coarsening rate are summarized in Table 3. It was evident that the coarsening of the γ′ precipitates obeys the linear relationship, r3 −r30. This result suggests that the coarsening of the γ′ precipitate follows the cubic kinetics of diffusion-controlled particle growth. As shown in Table 3, it was found that the coarsening rate decreased with increasing Mo content for all temperatures. It was
Fig. 3 – FESEM images of γ′ precipitates of the 3 Mo alloy aged at 760 °C, 780 °C and 810 °C for five different times.
183
M A TE RI A L S C HA RACT ER I ZA TI O N 95 ( 20 1 4 ) 1 8 0– 1 8 6
Fig. 4 – FESEM images of γ′ precipitates of the 6 Mo alloys aged at 760 °C, 780 °C and 810 °C for five different times.
observed that the coarsening rate for the 0.5 Mo alloy was about twice higher than other alloys.
3.3. Activation energy for coarsening and partition coefficient The activation energy for the coarsening and partition coefficient was investigated to confirm the effect of increasing Mo content on the coarsening rate. The activation energy for coarsening can be determined from the slope of the ln k versus 1 / T plot by consulting a previous study [11], which is shown in Fig. 6. The values obtained for the activation energy for coarsening are summarized in Table 4. As shown in Table 4, the activation energies for coarsening were approximately 245 kJ/mol for 0.5 the Mo alloy, 262 kJ/mol for the 3 Mo alloy and 278 kJ/mol for the 6 Mo alloy. The activation energy for coarsening of the 0.5 Mo alloy was consistent with the results by Zhao et al. [8], which have shown the activation energy for coarsening to be 247 kJ/mol in Inconel 740. The activation energy for coarsening was observed to increase by approximately 15 kJ/mol as the Mo concentration increased from 3 wt.% to
6 wt.%. This indicated that additions of Mo increased the activation energy for coarsening. The line profile across the γ′ precipitate and mapping with respect to the γ′ precipitate were investigated to confirm the partition coefficient, as shown in Figs. 7 and 8. It was observed that Al, Ti, and Nb were preferred to the γ′ precipitate, whereas Cr and Co were preferred to the γ matrix. Mo was uniformly distributed between the γ matrix and γ′ precipitate. The value for the partition coefficient is summarized in Table 5. Along with the activation energy for coarsening and the partition coefficient results in this work, the solute controlling the coarsening rate was considered to be Al. This result is consistent with the solute controlling the coarsening rate, as investigated by Fährmann et al. [7]. The decrease in the coarsening rate with increasing Mo contents can be attributed to the slow diffusion of Al in the γ matrix. According to Wang et al. [10], it was observed that the lattice parameter of the γ matrix increased with the increasing Mo content. This behavior affected the increase in the saddle-point energy, resulting in a decrease in the diffusion of
Table 2 – Mean γ′ sizes for alloys after long-term aging at 760 °C, 780 °C, and 810 °C. Specimen
γ′ size by treatment time (nm)
Temperature (°C) 50 h
0.5 Mo
3 Mo
6 Mo
760 780 810 760 780 810 760 780 810
19.1 21.31 25.2 17.67 19.28 22.04 17.16 18.46 21.56
± ± ± ± ± ± ± ± ±
0.02 0.20 0.23 0.24 0.05 0.86 0.55 0.83 0.18
100 h 23.2 27.5 33.3 19.60 21.73 26.04 18.73 20.84 25.46
± ± ± ± ± ± ± ± ±
0.13 0.31 0.54 0.44 0.67 0.13 0.33 0.54 0.96
150 h 26.0 30.6 38.1 20.96 23.82 29.09 20.08 22.78 28.44
± ± ± ± ± ± ± ± ±
0.58 0.20 0.74 0.71 0.83 0.52 0.36 0.33 0.4
200 h 28.4 33.9 42.1 22.29 25.01 31.38 21.26 24.43 30.89
± ± ± ± ± ± ± ± ±
0.51 0.71 0.34 0.33 0.56 0.77 0.15 0.02 0.43
500 h 35.5 42.8 54.0 27.75 32.43 41.94 26.56 31.46 40.95
± ± ± ± ± ± ± ± ±
0.61 0.16 0.61 0.28 0.64 0.78 0.43 0.21 0.24
184
M A TE RI A L S C HA RACT ER I ZA TI O N 95 ( 20 1 4 ) 1 8 0– 1 8 6
Fig. 5 – Average γ′ precipitate size versus aging time at 760 °C, 780 °C, and 810 °C; (a) 0.5 Mo alloy, (b) 3 Mo alloy and (c) 6 Mo alloy.
Al in the γ matrix. This was agreeable with the result of Stalker et al. [12]. They found that Mo reduces the diffusion of Al in Ni alloys. This fact indicated that the activation energy
Table 3 – k (x temperatures.
nm3 h−1)
values
at
various
aging
Temperature (°C)
0.5 Mo
3 Mo
6 Mo
760 780 810
65.6 ± 0.21 119.8 ± 0.71 247.1 ± 0.73
32.6 ± 0.22 57.7 ± 0.51 133.6 ± 0.29
28.6 ± 0.11 53.1 ± 0.11 128.2 ± 0.18
for coarsening increased and the coarsening rate decreased. Therefore, it was concluded that the decrease in the coarsening rate and the increase in the activation energy for coarsening with increasing Mo content were due to the slow diffusion of Al in the γ matrix.
4. Conclusions The coarsening kinetics of Ni–25Cr–20Co–1.8Ti–0.9Al–2Nb– 0.03C–xMo (x = 0.5, 3, 6 wt.%) alloys were investigated with respect to the activation energy for coarsening and the partition coefficient. The following conclusions could be made: 1. In the alloys, it was observed that the growth of the γ′ precipitate increased as the temperature and time increased, but most notably when temperature increased. It was found that the mean size of the γ′ precipitate for the 6 Mo alloy was slightly smaller than for the 3 Mo alloy at equal temperature and time. For 0.5 Mo, it was observed to be approximately 30% larger than other alloys.
Table 4 – Activation energy for coarsening for the alloys. Q (kJ/mol) Fig. 6 – Arrhenius plot of the coarsening rate coefficient k vs. the reciprocal of absolute temperature.
0.5 Mo 3 Mo 6 Mo
245.6 ± 5.3 261.6 ± 7.1 277.8 ± 11.0
M A TE RI A L S C HA RACT ER I ZA TI O N 95 ( 20 1 4 ) 1 8 0– 1 8 6
185
Fig. 7 – Line profile as a function of distance (nm) with respect to the γ′ precipitate for 6 Mo alloy aged at 760 °C for 500 h.
Fig. 8 – Mapping with respect to the γ′ precipitate for 6 Mo alloy aged at 760 °C for 500 h.
2. In the alloys, it was evidenced by the fact that the coarsening kinetics of the γ′ precipitate was controlled by the volume diffusion. The coarsening rate was shown to be slower for the 6 Mo alloy than for the 3 Mo alloy. For the 0.5 Mo alloy, the coarsening rate was shown to be approximately twice as fast as other alloys. 3. The activation energy for coarsening was approximately 245 kJ/mol for the 0.5 Mo alloy, 262 kJ/mol for the 3 Mo alloy and 278 kJ/mol for the 6 Mo alloy. It was observed that Al, Ti,
Table 5 – Measured composition (at.%) and partition coefficient for 6 Mo alloy aged at 760 °C for 500 h.
Cγ Cγ′ Cγ′/γ
Ni
Cr
Co
Mo
Al
Ti
Nb
38.89 56.03 1.44
34.46 13.01 0.38
21.48 11.35 0.53
2.05 1.78 0.87
1.26 7.48 5.94
0.93 8.53 9.20
0.38 1.82 4.77
and Nb were preferred to the γ′ precipitate while Cr and Co were preferred to the γ matrix. Mo was uniformly distributed between the γ matrix and γ′ precipitate. Along with the activation energy for coarsening and the partition-coefficient results in this work, the solute controlling the coarsening rate was considered to be Al. It was concluded that the decrease in the coarsening rate and the increase in the activation energy for coarsening with increasing Mo content were due to the slow diffusion of Al in the γ matrix.
Acknowledgment This work was supported by the Human resource Development program (No. 20114010203020) of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) grant funded by the Korea government Ministry of Trade, Industry and Energy.
186
M A TE RI A L S C HA RACT ER I ZA TI O N 95 ( 20 1 4 ) 1 8 0– 1 8 6
REFERENCES [7] [1] Zhao S, Dong J, Xie X, Smith GD, Patel SJ. Thermal stability study on a new Ni–Cr–Co–Mo–Nb–Ti–Al superalloy. Superalloys 2004:63–72. [2] Lifshitz M, Slozov VV. The kinetics of precipitation from supersaturated solid solutions. J Phys Chem Solids 1961;19:35–50. [3] Wagner C. Theorie der alterung von niederschlägen durch umlösen (Ostwald-reifung). Z Elektrochem 1961;65:581–91. [4] Wang WZ, Jin T, Liu JL, Sun XF, Guan HR, Hu ZQ. Role of Re and Co on microstructures and γ′ coarsening in single crystal superalloys. Mater Sci Eng A 2008;479:148–56. [5] Devaux A, Naze L, Molins R, Pineau A, Organista A, Guedou JY, Uginet JF, Heritier P. Gamma double prime precipitation kinetic in Alloy 718. Mater Sci Eng A 2008;486:117–22. [6] Sudbrack CK, Ziebell TD, Noebe RD, Seidman DN. Effects of a tungsten addition on the morphological evolution, spatial
[8]
[9] [10]
[11]
[12]
correlations and temporal evolution of a model Ni–Al–Cr superalloy. Acta Mater 2008;56:448–63. Fährmann M, Fährmann E, Pollock TM, Johnson WC. Element partitioning during coarsening of (γ–γ′) Ni–Al–Mo alloys. Metall Trans A 1997;28:1943–5. Zhao S, Xie X, Smith GD, Patel SJ. Gamma prime coarsening and age-hardening behaviors in a new nickel base superalloy. Mater Lett 2004;58:1784–7. Ricks RA, Porter AJ, Ecob RC. The growth of γ′ precipitates in nickel-base superalloys. Acta Metall 1983;31:43–53. Wang Tao, Sheng Guang, Liu Zi-Kui, Chen Long-Qing. Coarsening kinetics of γ′ precipitates in the Ni–Al–Mo system. Acta Mater 2008;56:5544–55. Ges AM, Fornaro O, Palacio HA. Coarsening behaviour of a Ni-base superalloy under different heat treatment conditions. Mater Sci Eng A 2007;458:96–100. Stalker MK, Morral JE, Romig Jr AD. Application of the square root diffusivity to diffusion in Ni–Cr–Al–Mo alloys. Metall Trans A 1992;23A:3245–9.