Accepted Manuscript

Phase transformations and grain growth behaviors in superalloy 718

J.J. Ruan, N. Ueshima, K. Oikawa

PII: S0925-8388(17)34109-9
DOI: 10.1016/j.jallcom.2017.11.327
Reference: JALCOM 44020

To appear in: Journal of Alloys and Compounds

Received Date: 9 February 2017

Revised Date: 27 November 2017
Accepted Date: 28 November 2017

Please cite this article as: J.J. Ruan, N. Ueshima, K. Oikawa, Phase transformations and grain
growth behaviors in superalloy 718, Journal of Alloys and Compounds (2017), doi: 10.1016/
j.jallcom.2017.11.327.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 1473 K for 170 mins 1473 K for 170 mins
ACCEPTED MANUSCRIPT

The annealing temperatures TiN

GB NbC Calculated Activation energy was
were decided after GB
data from calculated when the main
the DSC measurement.

PT
8 mm 4 mm Thermo-Cal pinning controller was
endothermic

Heating Rate = 10 K / min;

RI
was introduced. identified.

SC
Atmosphere : argon gas flow Mechanism for
The microstructure of the alloy m=4 15.5

U
grain growth

AN
for DSC measurement -2.0
NbC 14.5

ln(K1T/CmNb )
M
γ 1473 K
-3.0 1373 K

D
13.5

TE
R2=0.97

ln(dG/dt)
exothermic

-4.0

EP
4 mm TiN 12.5 Q=417 kJ/mol
1555 K

C
-5.0 m=4

AC
ing 1495 K 11.5
Heat m=3
1323 K
1423 K 0.68 0.70 0.72 0.74 0.76
1073 1173 1273 1373 1473 1573 -6.0 m=2 1/T, 10-3 K-1
4.5 4.9 5.3 5.7 6.1
Temperature / K lnG
 ACCEPTED MANUSCRIPT
Phase transformations and grain growth behaviors in superalloy 718

J.J. Ruan, N. Ueshima*, K. Oikawa

Department of Metallurgy, Graduate School of Engineering, Tohoku University,

Aramaki-aza Aoba 6-6-02, Aoba-ku, Sendai 980-8579, Japan

* Corresponding author: n-ueshima@material.tohoku.ac.jp

PT
RI
Abstract: The phase transformation and grain growth behaviors of commercial

Ni-Fe-base superalloy 718 at 1323, 1373, 1423 and 1473 K were investigated through

SC
DSC and SEM methods. The transformation temperatures of γ + NbC → L and

γ → L were determined as 1495 and 1555 K, respectively, and TiN was shown to

U
AN
gradually dissolve into the formed L. The pinning effect caused by Nb-rich carbide

(NbC) and Ti-rich nitride (TiN) was observed, and the mean size of the grains
M

increased with holding time at each temperature. The grain growth exponent, m, was

confirmed to be 4, indicating that grain growth was controlled by the coarsening of

D

the precipitate through grain boundary diffusion. The grain boundary diffusion of Nb
TE

and N was confirmed to control the coarsening of NbC and TiN, respectively. The

major factor for γ grain growth was confirmed to be the coarsening of NbC, and the
EP

activation energy for grain growth was calculated as 417 kJ·mol-1. The equations for
C

grain growth were determined as G4-G04＝0.10×106×t (1323 K), G4-G04＝0.49×106×t

AC

(1373 K), G4-G04＝2.01×106×t (1423 K) and G4-G04＝3.89×106×t (1473 K), for all

four temperatures respectively. The predicted grain growth curves were in good

agreement with the experimental data.

Keywords: Superalloy 718, Grain growth, Microstructure, Modeling

1
 ACCEPTED MANUSCRIPT
1. Introduction

Superalloy 718 is a precipitation-strengthened Ni-Fe-base superalloy, notable for

its ability to maintain its attractive mechanical properties for lengthy periods of time

over an extensive temperature range, including cryogenic temperatures to 923 K [1-3].

In addition to this, due to the sluggish precipitation of the main strengthening phase γ''

PT
(DO22-ordered crystal structure), superalloy 718 also shows good weldability. It has

RI
been confirmed that the weld residual stresses are able to be released before the

precipitation of γ'', then the strain-age cracking usually occurred during followed

SC
aging heat-treatment can be resisted [1]. Thus, superalloy 718 is extensively used in

the gas turbine industry for rotating disks and shaft components, etc [1, 4].

U
Apart from the coherent dispersion strengthening caused by the metastable
AN
coherent γ'' phase, another metastable L12-ordered γ' with a smaller volume fraction
M

than that of γ'' also makes a small contribution to the reinforcement of superalloy 718

[1, 4-11]. The other phases reported in superalloy 718 are orthorhombic DOa-ordered
D

δ-Ni3Nb with a needle/plate-like morphology [3-5, 9, 12-15], island-like hexagonal

TE

close-packed Laves-C14 [4, 12, 16-18], discrete Nb-rich MC-type carbides (NbC) [4,

9, 12, 16, 19] and Ti-rich MN-type nitrides (TiN), etc [19-23].
EP

The microstructure, and particularly the size of the grains in the matrix, largely
C

determines the properties of the superalloys [24-26]. For instance, a large grain size
AC

has been shown to slow the crack growth rate in superalloy 718 under cyclic loading

at the elevated-temperature condition [26] where intergranular crack extension tends

to occur. To date, investigations into the grain growth behavior have mainly been

focused on nanocrystalline [27] and hot deformed [28-30] superalloy 718. However,

rather insufficient data regarding the grain growth of bulk superalloy 718 under

isothermal annealing is available in the literature [31-32]. Furthermore, the data

2
 ACCEPTED MANUSCRIPT
provided by Huda et al. [31-32] shows a considerable scattering when subject to the

mathematic calculations used in the present work: this is particularly invalid for lower

temperatures (1323 and 1373 K). This scatter is discussed in section 3.2. Therefore,

the main objective of the present work is to obtain data related to the growth kinetics

of the matrix γ during isothermal holding at temperatures higher than the solvus of the

PT
δ-Ni3Nb phase, and to provide essential information for the prediction of the

RI
microstructure evolution of superalloy 718.

SC
2. Experimental details

The material used in the present work was provided by the Hitachi Metals

U
company, and its composition is listed in Table 1. Samples with the dimensions of
AN
10mm × 5mm × 1.5mm were cut from the as-received alloy, polished using abrasive
M

paper, and then sealed in the quartz capsules backfilled with high-purity argon after

evacuation. They were then subjected to heat-treatment. Solution heat-treatment was

D

performed before isothermal heat-treatment to eliminate the possibility of the

TE

influences of γ', γ'' and the δ-Ni3Nb phases on γ grain growth. The samples were

quenched into ice water after heat-treatment. Details of the heat-treatments performed
EP

in the present work are given in Table 2.

C

A differential scanning calorimetry (DSC) test was performed under an argon

atmosphere at a heating rate of 10 K min-1 on the alloys annealed at 1473 K to

AC

determine the appropriate annealing temperature for the present work. Microstructure

observations were operated by a field emission scanning electron microscope

(FE-SEM), and the chemical compositions of carbide and nitride were analyzed by the

energy dispersive spectrometer (EDS) equipped on the FE-SEM. The alloys were

polished by a vibratory polisher in OP-S suspension before FE-SEM observations.

3
 ACCEPTED MANUSCRIPT
The commonly used two-dimensional microstructure was adopted instead of the

more complex and more time-consuming three-dimensional microstructure in

calculating the grain size [33]. Since the microstructures were simple, with only a

small amount of precipitates (NbC and TiN) [34], the Jeffries’ (or Planimetric)

procedure was considered suitable for measuring the mean grain size from the

PT
FE-SEM images. In the Jeffries’ method, the average area, Am, is calculated by

RI
dividing the total area by the number of grains counted (N): each grain clearly inside

the test area (a circle in the present work) was counted as 1 while those which were

SC
partially inside the test area were counted as 0.5 [34]. Meanwhile, the mean grain size

was represented by the diameter G of an equivalent circle with same area Am [33-36].

U
AN
3. Results and discussion
M

3.1. Phase transformation and determination of annealing temperature

The microstructure of the as-received alloy is shown in Figures 1(a)-(c) where

D

the tiny precipitates can be observed. Because the as-received alloy was taken from a
TE

large rod-like sample (~0.25 m×4.70 m) manufactured by the Vacuum Induction

Melting (VIM)-Electroslag Remelting (ESR)-Vacuum Arc Remelting (VAR)-Hot

EP

forging-Air cooling process, it is reasonable to assume the tiny precipitates in Figure

C

1(b) are the γ', γ'' and δ-Ni3Nb phases. The Nb-rich carbide NbC (white contrast) and
AC

the Ti-rich nitride TiN [19, 21] (grey contrast phase in Figure 1(c)) are confirmed by

the energy dispersion spectrum (EDS) results, as shown in Figure 1(d). The

morphology in Figure 1(c) is consistent with the results confirmed by other

researchers [20-22] with part of the NbC phase nucleating on pre-existing TiN.

Importantly, though the influence of possible γ', γ'' and δ-Ni3Nb on γ grain growth can

be avoided, the high solvus temperature of NbC and TiN means their influence on γ

4
 ACCEPTED MANUSCRIPT
grain growth must be considered.

The DSC heating curve obtained from the alloy annealed at 1473 K for 170 mins

is shown in Figure 2(a), where two transition peaks, 1495 and 1555 K, are identified.

The microstructure of the alloy before the DSC test is shown in Figure 2(b): the white

area is the NbC phase and the grey area is the TiN phase.

PT
In order to reveal the phase transitions during heating process, the alloys were

RI
heated at the same heating rate (10 K/min) as that of the DSC test, and then quenched

at 1537 (Figures 3(a)-(c)) and 1580 K (Figures 3(d)-(f)), respectively. A low

SC
magnification image obtained from the alloy quenched from 1537 K is shown in

Figure 3(a). The decomposing TiN and NbC phases are clearly visible in this sample,

U
as shown in Figures 3(b) and (c), respectively, and the eutectic microstructure derived
AN
from the cooling of the liquid phase during quenching is also evident. Thus, it was
M

confirmed that the microstructure of this sample consisted of NbC + TiN + γ +

eutectic (γ/ NbC) (derived from the liquid phase (L) after water quenching). The
D

eutectic microstructure (γ/ NbC) in superalloy 718 was also observed by Knorovsky
TE

[16]. The melting of the matrix γ in the alloy quenched from 1580 K is clear from the

image in Figure 3(d). The solidus temperature of the superalloy 718 was reported as
EP

1500 and 1593 K by Eiselstein [37] and Knorovsky [16], respectively. The melting
C

temperature of the matrix γ indicated in Figure 2(a) (1555 K) is between these two
AC

values. Considering the presence of solid-soluted alloying elements in superalloy 718,

the variation in the solidus temperature is reasonable. The decomposing TiN, eutectic

(γ/ NbC) and dendrite γ are also clear in this sample, as shown in Figures 3(e) and (f),

respectively. Therefore, it was confirmed that the microstructure of this sample

included TiN, some remaining γ, eutectic (γ/ NbC) and dendritic γ.

Since decomposing TiN is found near the eutectic microstructure (Figures 3(b)

5
 ACCEPTED MANUSCRIPT
and (e)) for both quenching temperatures (1537 and 1580 K), it can be assumed that

TiN gradually dissolved into the liquid phase as the temperature increased. This is in

good agreement with the finding by Cockcroft and Mitchell et al. [20-21] that the

saturation solubility of TiN in superalloy 718 increases with increasing temperature.

Considering the high N content (65 ppm) in the present alloys (Table 1), and the

PT
solubility of N in superalloy 718 at 1700 K, calculated by Cockcroft et al. [20] to be ~

RI
38 ppm, it is believed that some pockets of TiN exist above the liquidus temperature

even though most of the TiN has dissolved into the liquid phase. As such, the peak

SC
located at 1495 K in Figure 2(a) can be largely attributed to the formation of the liquid

phase due to the reaction between NbC and the γ phases, while the peak at 1555 K can

U
be attributed to the melting of the remaining matrix γ.
AN
A total of four annealing temperatures, 1323, 1373, 1423 and 1473 K, were
M

chosen to investigate grain growth behavior in the absence of influence from γ', γ'',

δ-Ni3Nb and the liquid phases. The holding times selected ranged from 0 to 281
D

minutes after solution heat-treatment. The microstructures of the alloys after solution
TE

heat-treatment are shown in Figures 4(a)-(d) using EBSD (Electron Backscattered

Diffraction) grain maps. The mean grain sizes of the alloys heat-treated at 1323, 1373,
EP

1423 and 1473 K were measured as 97, 147, 162 and 221 µm, respectively, and are
C

listed in Table 2. In the sample annealed at 1473 K for 170 mins after solution
AC

heat-treatment, grain boundary pinning microstructures resulting from the TiN and

NbC phases were observed. This can be seen in Figures 5(a)-(b).

3.2. Grain growth behavior

The grain growth curves in Figure 6 indicate an increase in the grain size as the

growth rate slowed down with longer annealing-time at each temperature. The

6
 ACCEPTED MANUSCRIPT
kinetics of isothermal grain growth can be described by the parabolic equation

presented by Burke and Turnbull [38], as shown in Equation (1)

G − G = KV σt (1)

where G is the average grain diameter; G0 is the mean grain size at t=0; σ is the grain

boundary energy; Vg is the gram atomic volume; t is the heating-time, and K is a

PT
temperature related parameter.

RI
This parabolic equation (Equation (1)), however, tends to be applied to the pure

single phase grain growth since it only considers the driving force caused by the

SC
reduction of the total amount of grain boundary energy and does not consider other

factors affecting grain growth, such as the solute atoms, solute segregation,

U
precipitates or texture [38-39]. Thus, a more general modified equation based on the
AN
parabolic equation, power law, was proposed as a mathematical expression of
M

universal grain growth, as described in Equation (2)

G − G = K  t (2)
D

where m is the grain growth exponent and K1 is a temperature related parameter.

TE

Usually, an m value of 2 indicates an ideal system controlled by diffusion, a value of 3

represents the coarsening of precipitates by diffusion in the grain and a growth

EP

exponent of 4 suggests the coarsening of precipitates through diffusion at the grain

C

boundary [39-40]. It should be noted that the m value is also affected by the void size
AC

and distribution, solute and segregation, the size and distribution of the second phase,

and texture in the alloys [39].

In this study, the average grain diameter of the alloys after solution

heat-treatment is not small enough to be neglected, as indicated in Table 2. Therefore,

by performing the natural logarithm on differentiated Equation (2), Equation (3) is

obtained.

7
 ACCEPTED MANUSCRIPT
dG K
( ) =    + (1 − m)G (3)
dt m

In Equation (3), dG/dt is calculated as below:

dG G − G
= ;
dt t − t

dG 1 dG dG"
= ∗ + ;

PT
dt 2 dt dt "

dG" 1 dG dG#
= ∗ + ;

RI
dt " 2 dt dt #
.
.

SC
.

dG% G% − G%&
=

U
dt % t % − t %&
AN
The m and K1 values at each temperature are determined from the linear

regression of the plot of ln(dG/dt) versus lnG displayed in Figure 7. A fixed m value
M

of 4 indicates the Ostwald ripening of the precipitates by grain boundary diffusion

[39-40], and K1 is calculated as 0.10 × 106 (1323 K), 0.49 × 106 (1373 K), 2.01 × 106
D

(1423 K) and 3.89 × 106 (1473 K) µm4·s-1, respectively. It was not possible to
TE

describe the present experimental data with other m values, such as 2 or 3, as can be
EP

seen in Figure 7. The pinning microstructures observed in Figures 5(a)-(b) confirm the

Ostwald ripening of the NbC-controlled and TiN-controlled grain growth of the

C

matrix. The plots in Figure 8 were computed by the same methods used in the present
AC

work using the grain growth data of superalloy 718 provided by Huda et al [31-32].

While m values of 3.76 (1323 K), 3.92 (1373 K), 3.72 (1423 K) and 3.61 (1473 K)

were obtained, it should be noted that these plots show poor coefficients of

determination (R2), especially at lower temperature (1323 and 1373 K).

According to the parameters obtained by Equation (2), the predicted grain

growth curves were obtained, and are displayed in Figure 9 with the measured grain
8
 ACCEPTED MANUSCRIPT
size. It can be seen that the predicted data is in good agreement with the experimental

data.

The relationship between the grain size of the matrix and particle size was

studied by many researchers [41-43], and can be finally described by Equation (4)

G K
= ) (4)

PT
r f

where r and f are the size and volume fraction of second-phase particles,

RI
respectively, K2 and k are constants for a given system. The value of f does not

SC
change during Ostwald ripening. The relationship between particle size and heating

time can be expressed in Equation (5) [44] if the Ostwald ripening of particle is

U
limited by the grain boundary diffusion.
AN
8 γC. ωv D·2
r# − r# = t (5)
9 ABRT

where r0 is the initial size of the second-phase particles; 8 is the interfacial energy
M

between particle and matrix; C∞ is the equilibrium concentration (molarity) of solute

D

atoms in the boundary without the influence of precipitates; ω is the thickness of the
TE

grain boundary; νm is the molar volume of the particle; Dg·b is the grain boundary

diffusion coefficient; and A and B are constants. By integrating Equations (2), (4) and
EP

(5), the Equation (6) is thus obtained.

C. D·2
C

K = K " (6)
T
AC

:;=< >ω?<@
where K " ＝ and is constant for the same particle, and Dg·b is described in
AB=C DEF

Equation (7)

Q·2
D·2 = D exp(− ) (7)
RT

where D0 is the diffusion coefficient at infinite temperature, and Qg·b is the activation

energy for grain boundary diffusion.

9
 ACCEPTED MANUSCRIPT
The product of C. and D·2 of the solute atom is necessary for determining the

controller element for the coarsening of the precipitate. Ideally, the grain boundary

diffusion data of Nb, Ti, C and N in superalloy 718 would be used to determine the

diffusion controller in the present work due to the pinning microstructures seen in

Figures 5(a)-(b), however, no such data is available. Therefore, the inter-diffusion

PT
coefficients of Nb [45] and Ti [46], and the impurity-diffusion coefficient of C [47] in

RI
Ni were used, and the related data is listed in Table 3 where the data related to

impurity-diffusion coefficients of C [48] and N [49] in γ-Fe (face-centered-cubic (fcc)

SC
crystal structure) are also provided. The diffusion coefficients were calculated by

Equation (8) using the data provided in Table 3, and are listed in Table 4.

D = D exp(−
Q
) U (8)
AN
RT

where D is the diffusion coefficient, and Q is the activation energy for diffusion. The
M

relationship between the diffusion coefficient and temperature are also plotted in
D

Figures 10(a)-(b): the inter- (Nb and Ti in Ni) and impurity-diffusion coefficients (C
TE

in Ni, C and N in γ-Fe) clearly increase with rising temperature. In Table 4, it can be

seen that the impurity-diffusion coefficient of C in Ni is slightly lower than that of C

EP

in Fe even though they show the same order of magnitude (-11). Thus, because both C

and N are interstitial atoms, it is reasonable to assume the same tendency for the
C

impurity-diffusion coefficients of N in Ni and N in Fe.

AC

The equilibrium molarity (Cm) of the solute atoms Nb, C, Ti and N in the matrix

γ was taken into account due to the lack of C. , despite the difference between them.

By introducing the molar percentage of Nb, C, Ti and N in the matrix γ as calculated

by the Thermo-Calc software [50] using the Ni-Data database [51] and the lattice

constant of the matrix γ, the Cm for the solute atoms Nb, C, Ti and N in the matrix γ

can be calculated. In the investigation performed by Dehmas et al. [3], the lattice
10
 ACCEPTED MANUSCRIPT
constant of the matrix γ slightly decreases with the loss of Nb in the matrix γ caused

by the precipitation of the δ-Ni3Nb phase. The difference between the lattice constant

of the matrix γ without the δ-Ni3Nb phase and that of the matrix γ which contains a

small amount of the δ-Ni3Nb phase is negligible for the calculation of the matrix

molar volume in the present work. As such, the lattice constant of the matrix γ (3.604

PT
Å) with the lowest volume fraction of the δ-Ni3Nb phase in the results reported by

RI
Dehmas et al. [3] was used for all the temperatures investigated in the present work.

The calculated Cms for the solute atoms Nb, C, Ti and N in the matrix γ and the

SC
related data are listed in Table 5.

The products of C and D for the solute atoms Nb, C, Ti and N were calculated

U
using the data listed in Tables 4-5, as seen in Table 6. Note that the DR2 Q% RQ R2
L%MNO&PQBB. ·C is
AN
lower than DVLSTOQMU&PQBB.
Q% RQ V
·C by about one order of magnitude, while DWQ Q% RQ WQ
L%MNO&PQBB. ·C
M

exceeds than DR Q% XN R
LSTOQMU&PQBB. · C by about one order of magnitude among the

temperatures investigated. This suggests that Nb is the controller for the Ostwald
D

ripening of NbC, and that N is the controller for the Ostwald ripening of TiN.
TE

The number densities of the NbC and TiN in the solution-treated samples are
EP

necessary for determining the main pinning phase on the matrix γ grain boundary

during the heating process. However, because they sparsely distributed in the sample,
C

it is difficult to count them. The molar percentage of NbC and TiN in the alloys after
AC

solution heat-treatment is believed to be close to that of NbC and TiN in the alloys

under equilibrium because of the small size of the pre-existent precipitates indicated

in Figure 1(b). Through the molar percentage of NbC and TiN in the matrix γ

calculated by the Thermo-Calc software [50] using the Ni-Data database [51] and the

lattice constant of these phases, the volume of NbC and TiN in the alloy with 1 mol

can be obtained. The lattice constants of NbCx and TiNx were calculated using the

11
 ACCEPTED MANUSCRIPT
lattice constants obtained by Storms [52] (aNbCx

(nm)=0.44704-0.00239(1-x)-0.03586(1-x)2) and Nagakura [53] (aTiNx

(nm)=0.41925+0.00467x), respectively. The x refers to the molar ratio of C/Nb and

N/Ti, respectively, and was calculated through the molar percentage of the solutes in

NbCx and the solutes in TiNx provided by the Thermo-Calc software [50] using the

PT
Ni-Data database [51]. The calculated volumes of NbC and TiN in the alloy with 1

RI
mol are listed in Table 7 where the related data is also displayed. The ratios of

VNbCx/VTiNx are found to be around 3.7~4.8 at the same temperature. Moreover, if we

SC
accept the results obtained by other researchers [20-22] that the TiN provides the

nucleation site for the precipitation of NbC, it is reasonable to assume that the number

U
density of NbC is higher than that of TiN. This speculation agrees well with the
AN
conclusion derived from the present volume ratio calculation on NbC and TiN that the
M

number density of NbC is higher than that of TiN if the mean sizes of these two

phases are close. This further indicates that the NbC is the main controller for γ grain
D

growth during heating.

TE

Equation (9) can be obtained from Equations (6) and (7).

K T Q·2
 = K # − (9)
EP

C. RT

where K4=K3·D0. As mentioned above, the γ grain growth is mainly controlled by the
C

coarsening of NbC, while the solute Nb is the controller for the coarsening of NbC,
AC

R2
therefore, the equilibrium molarity C of the Nb in the matrix γ was adopted for the

calculation of the activation energy for γ grain growth. Then, the relationship between

R2
ln(K1T/C ) and 1/T was plotted, as shown in Figure 11 where Qg·b is 417 kJ·mol-1.

4. Summary

The phase transitions during the heating process and grain growth behaviors of
12
 ACCEPTED MANUSCRIPT
the commercial Ni-Fe-based superalloy 718 were investigated, and the power law was

introduced for its mathematical expression. The predicted curve fits the present

experimental data well. The main experimental and calculated results can be

summarized as follows:

(1) Matrix γ starts to react with NbC to form L phase at 1495 K, and transforms

PT
into the L phase at 1555 K, while TiN gradually dissolves into the formed L phase.

RI
(2) The pinning effects caused by NbC and TiN phases on γ grain growth were

confirmed. The determined grain growth exponent m value of 4 indicates that γ grain

SC
growth was controlled by the Ostwald ripening of the NbC and TiN phases through

grain boundary diffusion during heating.

U
(3) The power laws for grain growth were determined to be G4-G04＝0.10×106×t
AN
(1323 K), G4-G04＝0.49×106×t (1373 K), G4-G04＝2.01×106×t (1423 K) and G4-G04
M

＝3.89×106×t (1473 K), respectively.

(4) The diffusion of Nb was confirmed to control the coarsening of NbC while
D

the coarsening of TiN is controlled by the diffusion of N. The coarsening of NbC was
TE

shown to be the largest factor controlling the γ grain growth, and the activation energy
EP

for γ grain growth was calculated as 417 kJ·mol-1.

C

Acknowledgements
AC

This work was supported by Council for Science, Technology and Innovation

(CSTI), Cross-ministerial Strategic Innovation Promotion Program (SIP), “Structural

Materials for Innovation” (Unit No. B21, Development of Innovative Forging Process

Technology and Construction of Material/Process Database with the Largescale and

Precise Forging Simulator).

13
 ACCEPTED MANUSCRIPT
References

[1] C.T. Sims, N.S. Stoloff, W.C. Hagel, Superalloys II, 1987, pp. 165-187.

[2] D. Fournier, A. Pineau, Low cycle fatigue behavior of Inconel 718 at 298 K and

823 K, Metall. Mater. Trans. A, 8A (1977) 1095-1105.

PT
[3] M. Dehmas, J. Lacaze, A. Niang, B. Viguier, TEM study of high-temperature

RI
precipitation of delta phase in Inconel 718 alloy, Adv. Mater. Sci. Eng., 2011 (2011)

1-9.

SC
[4] E.A. Loria, Recent developments in the progress of superalloy 718, JOM, 44

U
(1992) 33-36.
AN
[5] Y.P. Mei, Y.C. Liu, C.X. Liu, C. Li, L.M. Yu, Q.Y. Guo, H.J Li, Effects of cold
M

rolling on the precipitation kinetics and the morphology evolution of intermediate

phases in Inconel 718 alloy, J. Alloys Compd., 649 (2015) 949-960.

D

[6] S.J. Hong, W.P. Chen, T.W. Wang, A diffraction study of the γ″ phase in
TE

INCONEL 718 superalloy, Metall. Mater. Trans. A, 32A (2001) 1887-1901.

EP

[7] R. Cozar, A. Pineau, Morphology of γ′ and 8′′ precipitates and thermal stability of

inconel 718 type alloys, Met. Trans., 4 (1973) 47-59.

C
AC

[8] M. Sundararaman, P. Mukhopadhyay, S. Banerjee, Some aspects of the

precipitation of metastable intermetallic phases in Inconel 718, Metall. Trans. A, 23A

(1992) 2015-2028.
[9] S.H. Chang, In situ TEM observation of γ′, γ ″and δ precipitations on Inconel 718

superalloy through HIP treatment, J. Alloy. Compd, 486 (2009) 716-721.

14
 ACCEPTED MANUSCRIPT

[10] D.F. Paulonis, J.M. Oblak, D.S. Duvall, PRECIPITATION IN NICKEL-BASE

ALLOY 718, Amer. Soc. Metals., 62 (1969) 611-622.

[11] M.C. Chaturvedi, Y.F. Han, Strengthening mechanisms in Inconel 718

superalloy, Met. Sci., 17 (1983) 145-149.

PT
[12] J.F. Radavich, Metallography of alloy 718, JOM, 40 (1988) 42-43.

RI
[13] M. Sundararaman, P. Mukhopadhyay, S. Banerjee, Precipitation of the δ-Ni3Nb

SC
phase in two nickel base superalloys, Metall. Trans. A, 19A (1988) 453-465.

[14] L.S. Araujo, D.S. Santos, S. Godet, J. Dille, A.L. Pinto, L.H. De Almeida,

U
Analysis of Grain Boundary Character in a Fine-Grained Nickel-Based Superalloy
AN
718, JMEP, 23 (2014) 4130–4135.
M

[15] P.J.P. Kañetas, L.A.R. Osorio, M.P.G. Mata, M.D. La Garza, V.P. López,

Influence of the Delta Phase in the Microstructure of the Inconel 718 subjected to
D
TE

“Delta – processing” Heat Treatment and Hot Deformed, Procedia Mater. Sci., 8

(2015) 1160-1165.
EP

[16] G.A. Knorovsky, M.J. Cieslak, T.J. Headley, A.D. Romig, W.F. Hammetter,
C

Inconel 718: a solidification diagram, Metall. Trans. A, 20A (1989) 2149-2158.

AC

[17] J.J. Schirra, R.H. Caless, R.W. Hatala, The effect of Laves phase on the

mechanical properties of wrought and cast + HIP Inconel 718, TMS, (1991) 375-388.

[18] C.H. Radhakrishna, K.P. Rao, The formation and control of Laves phase in

superalloy 718 welds, J. Mater. Sci., 32 (1997) 1977-1984.

[19] Y.C. Zhang, Z.G. Li, P.L. Nie, Y.X. Wu, Carbide and nitride precipitation during

15
 ACCEPTED MANUSCRIPT

laser cladding of Inconel 718 alloy coatings, Opt. Laser Technol., 52 (2013) 30-36.

[20] S.L. Cockcroft, T. Degawa, A. Mitchell, D.W. Tripp, A. Schmalz, Inclusion

Precipitation in Superalloys, TMS, (1992) 577-586.

[21] A. Mitchell, Nitrogen in Superalloys, High Temp. Mater. Processes, 24 (2005)

PT
101-110.

RI
[22] A. Mitchell, Progress in understanding clean metal production for IN718, TMS,

SC
(1994) 109-123.

[23] A. Mitchell, T. Wang, Solidification and Precipitation in IN718, TMS, (2001)

81-90.
U
AN
[24] R. Jiang, S. Everitt, M. Lewandowski, N. Gao, P.A.S. Reed, Grain size effects in
M

a Ni-based turbine disc alloy in the time and cycle dependent crack growth regimes,

Int. J. Fatigue, 62 (2014) 217–227.

D
TE

[25] Y. Ono, Y. Yuri, N. Nagashima, T. Ogata, N.Nagao, Effect of microstructure on

high-cycle fatigue properties of Alloy718 plates, IOP Conf. Series: Materials Science
EP

and Engineering. 102 (2015) 012001.

C

[26] L.A. James, The effect of grain size upon the fatigue-crack propagation behavior
AC

of alloy 718 under hold-time cycling at elevated temperature, Eng. Fract. Mech., 25

(1986) 305-314.

[27] H.G. Jiang, M.L. Lau, E.J. Lavernia, Grain growth behavior of nanocrystalline

Inconel 718 and Ni powders and coatings, Nanostruct. Mater., 10 (1998) 169-178.

[28] X.M. Chen, Y.C. Lin, D.X. Wen, J.L. Zhang, M. He, Dynamic recrystallization

16
 ACCEPTED MANUSCRIPT

behavior of a typical nickel-based superalloy during hot deformation, Mater. Des., 57

(2014) 568-577.

[29] Y. Wang, W.Z. Shao, L. Zhen, L. Yang, X.M. Zhang, Flow behavior and

microstructures of superalloy 718 during high temperature deformation, Mater. Sci.

PT
Eng. A, 497 (2008) 479-486.

RI
[30] J.M. Zhang, Z.Y. Gao, J.Y. Zhuang, Z.Y. Zhong, Grain growth model of IN718

SC
during holding period after hot deformation, J. Mater. Process. Technol., 101 (2000)

25-30.

U
[31] Z. Huda, T. Zaharinie, S.H. Islam, Effects of annealing parameters on
AN
grain-growth behavior of HAYNES-718 superalloy, Int. J. Phys. Sci., 6 (2011)
M

7073-7077.

[32] Z. Huda, T. Zaharinie, I.H.S.C. Metselaar, S. Ibrahim, G.J. Min, Kinetics of grain
D
TE

growth in 718 Ni-base superalloy, Arch. Metall. Mater., 59 (2014) 847-852.

[33] E. Persson, Austenite grain growth in bearing steels: An investigation on steel

EP

grades 100Cr6 and 100 CRMnMoSi8-4-6, 2014.

C

[34] Standard, ASTM E112: Standard Test Methods for Determining Average Grain
AC

Size, ASTM International, West Conshocken, PA 19428-2959, United States, (1996)

1-27.

[35] M. Hillert, On the theory of normal and abnormal grain growth, Acta Metall., 13

(1965) 227-238.

[36] A. Giumelli, Austenite grain growth kinetics and the grain size distribution,

17
 ACCEPTED MANUSCRIPT

Doctoral dissertation, University of British Columbia, 1995.

[37] H.L. Eiselstein, Metallurgy of a Columbium-hardened nickel-chromium-iron

alloy, Advances in the Technology of Stainless Steels and Related Alloys, ASTM

STP 369 (1965) 62-79.

PT
[38] J.E. Burke, D. Turnbull, Recrystallization and grain growth, Prog. Met.

RI
Phys., (1952) 220-292.

SC
[39] H.V. Atkinson, Overview no. 65: Theories of normal grain growth in pure single

phase systems, Acta Metall., 36 (1988) 469-491.

U
[40] S. Illescas, J. Fernández, J.M. Guilemany, Kinetic analysis of the austenitic grain
AN
growth in HSLA steel with a low carbon content, Mater. Lett., 62 (2008) 3478-3480.
M

[41] C.S. Smith, Grains, phases, and interphases: an interpretation of microstructure,

Trans. Metall. Soc. AIME, 175 (1948) 15-51.

D
TE

[42] M. Hillert, Inhibition of grain growth by second-phase particles, Acta Metall., 36

(1988) 3177-3181.
EP

[43] D.J. Srolovitz, M.P. Anderson, G.S. Grest, P.S. Sahni, Computer simulation of
C

grain growth-III. Influence of a particle dispersion, Acta Metall., 32 (1984)

AC

1429-1438.

[44] R.D. Vengrenovitch, On the Ostwald ripening theory, Acta Metall., 30 (1982)

1079-1086.

[45] M.S.A. Karunaratne, R.C. Reed, Interdiffusion of Niobium and Molybdenum in

Nickel between 900-1300 ℃, Defect and Diffusion Forum Vol. 237-240, Trans. Tech.

18
 ACCEPTED MANUSCRIPT

Publications, (2005) 420-425.

[46] S.B. Jung, T. Yamane, Y. Minamino, K. Hirao, H. Araki, S. Saji, Interdiffusion

and its size effect in nickel solid solutions of Ni-Co, Ni-Cr and Ni-Ti systems, J.

Mater. Sci. Lett., 11 (1992) 1333-1337.

PT
[47] C.J. Smithells, C.E. Ransley, The Diffusion of Gases through Metals. III. The

RI
Degassing of Nickel and the Diffusion of Carbon Monoxide through Nickel, Proc. R.

SC
Soc. London. Ser. A, 155 (1936) 195-212.

[48] J. Ågren, A revised expression for the diffusivity of carbon in binary Fe-C

U
austenite, Scr. Metall., 20 (1986) 1507-1510.
AN
[49] P. Grieveson, E.T. Turkdogan, Kinetics of Reaction of Gaseous Nitrogen with
M

Iron Part I: Kinetics of Nitrogen Solution in Gamma Iron, Trans. Metall. Soc.

AIME, 230 (1964) 407-414.

D
TE

[50] J.O. Andersson, T. Helander, L. Höglund, P. Shi, B. Sundman, Thermo-Calc &

DICTRA, Computational Tools For Materials Science, Calphad, 26 (2002) 273-312.

EP

[51] Thermotech Ni-Data Version 6, Thermotech Ltd. / Sente Software, Surrey,

C

United Kingdom.
AC

[52] E.K. Storms, N.H. Krikorian, The variation of lattice parameter with carbon

content of niobium carbide, J. Phys. Chem., 63 (1959) 1747-1749.

[53] S. Nagakura, T. Kusunoki, F. Kakimoto, Y. Hirotsu, Lattice parameter of the

non-stoichiometric compound TiNx, J. Appl. Crystallogr., 8 (1975) 65-66.

19
 ACCEPTED MANUSCRIPT
Figure Captions

Fig. 1. BSE images (a)-(c) and energy spectrum results (d) obtained from the

as-received alloy.

Fig. 2. (a) DSC heating curve and (b) BSE image obtained from the alloy annealed at

1473 K for 170 mins after solution heat-treatment.

PT
Fig. 3. BSE images obtained from the alloys quenched from 1537 K (a)-(c), and 1580

RI
K (d)-(f).

Fig. 4. EBSD grain maps with same scale obtained from the alloys after solution

SC
heat-treatment: (a) 1323 K; (b) 1373 K; (c) 1423 K; (d) 1473 K.

Fig. 5. BSE images obtained from the alloy annealed at 1473 K for 170 mins after

solution heat-treatment.
U
AN
Fig. 6. Measured grain size versus heating-time for each isothermal temperature.
M

Fig. 7. Plots of ln(dG/dt) against lnG for different temperatures.

Fig. 8. Plots of ln(dG/dt) against lnG: calculated from the experimental data reported
D

by Huda et al [31-32].
TE

Fig. 9. Comparisons between predicted growth curves and experimental data at each

isothermal temperature.
EP

Fig. 10. Plots of lnD against 1/T (a), and D against T (b).

R2
C

Fig. 11. Plot of ln(K1T/C ) against 1/T.

AC

20
 ACCEPTED MANUSCRIPT
Table Captions

Table 1. Chemical composition of the superalloy 718 used in the present work

(mass %).

Table 2. Details of the heat-treatments performed in the present work.

Table 3. Data relating to the inter-diffusion of Nb [45] and Ti [46] in Ni, and the

PT
impurity-diffusion of C [47] in Ni, C [48] and N [49] in Fe.

RI
Table 4. Predicted diffusion coefficients using the reported diffusion data [45-49].

Table 5. Equilibrium molarity of the solute atoms in the matrix and the related data.

SC
Table 6. The D·Cm of the solute atoms in the matrix.

Table 7. Volumes of the NbC and TiN phases in the alloy with 1 mol and the related

data.
U
AN
M
D
TE
C EP
AC

21
 ACCEPTED MANUSCRIPT

Table 1. Chemical composition of the superalloy 718 used in the present work (mass %).
Ni Fe Cr Nb Mo Ti Al Co Si

PT
53.8 Bal. 18.12 5.42 2.97 0.93 0.46 0.27 0.09

RI
Mn C Ta Zr N P B Mg S
0.07 0.025 <0.01 <0.01 0.0065 0.006 0.003 0.0013 0.0001

U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
Table 2. Details of the heat-treatments performed in the present work.
T/K Time for solution Mean grain diameter after Holding time for isothermal
heat-treatment / solution heat-treatment / µm heat-treatment / mins
mins
1323 15 97 0, 10, 25, 96, 165
1373 15 147 0, 10, 25, 96, 165
1423 10 162 0, 6, 23, 101, 281
1473 10 221 0, 6, 23, 101, 170

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
Table 3. Data relating to the inter-diffusion of Nb [45] and Ti [46] in Ni, and the
impurity-diffusion of C [47] in Ni, C [48] and N [49] in Fe.
Inter-diffusion Impurity-diffusion
D0 / m2·s-1 Q / kJ·mol-1 D0 / m2·s-1 Q / kJ·mol-1
Nb in Ni 0.88 × 10-4 257 - -
Ti in Ni 4.1 × 10-4 275 - -
C in Ni
(Recalculated in - - 8.36 × 10-5 172

PT
the present work)
C in Fe
(Recalculated in - - 2.34 × 10-5 148

RI
the present work)
N in Fe
9.10 × 10-5

SC
(Recalculated in - - 169
the present work)

U
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
Table 4. Predicted diffusion coefficients using the reported diffusion data [45-49].
Dinter-diffusion / 10-14 m2·s-1 Dimpurity-diffusion / 10-11 m2·s-1
T/K
Nb in Ni Ti in Ni C in Ni C in Fe N in Fe
1323 0.63 0.57 1.35 3.35 1.93
1373 1.47 1.41 2.39 5.48 3.38
1423 3.24 3.30 4.06 8.64 5.69
1473 6.77 7.25 6.65 13.21 9.24

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Table 5. Equilibrium molarity of the solute atoms in the matrix and the related data.

PT
Molar volume of Molar percentage of the solute atoms in the Molarity of the solute atoms in the matrix γ, Cm /
Lattice constant of
T/K
the matrix γ / Å [3]
the matrix γ / matrix γ / % [50-51] mol·m-3
m3·mol-1

RI
Nb C Ti N Nb C Ti N
-3
1323 3.27 5.02 × 10 1.09 4.99 × 10-5 1.16 × 103 1.78 3.88 × 102 1.77 × 10-2
1373 3.28 8.08 × 10-3 1.09 8.59 × 10-5 1.16 × 103 2.87 3.87 × 102 3.05 × 10-2
a=b=c=3.604 2.82 × 10-5

SC
1423 3.29 12.56 × 10-3 1.09 14.22 × 10-5 1.17 × 103 4.46 3.87 × 102 5.05 × 10-2
1473 3.19 19.51 × 10-3 1.07 23.10 × 10-5 1.13 × 103 6.92 3.79 × 102 8.20 × 10-2

U
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Table 6. The D·Cm of the solute atoms in the matrix.

T/K  
Nb, D . ·C C, D
  
 . ·C Ti, D   
 . ·C
 
N, D  . ·C

PT
1323 0.73 × 10-11 0.24× 10-10 0.22 × 10-11 0.34 × 10-12
1373 1.71× 10-11 0.69× 10-10 0.55 × 10-11 1.03 × 10-12

RI
1423 3.77× 10-11 1.81× 10-10 1.27 × 10-11 2.87 × 10-12
1473 7.67× 10-11 4.60× 10-10 2.75 × 10-11 7.58 × 10-12

U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Table 7. Volumes of the NbC and TiN phases in the alloy with 1 mol and the related data.
Molar percentage of the main solutes Molar Volume of the phases in
Molar ratio x Lattice constant (a=b=c) / Å

PT
Phases T/K in phases / % [50-51] percentage / the alloy with 1 mol /
[52-53] -2
Nb C Ti N C/Nb N/Ti 10 % [50-51] 10-8 m-3

RI
1323 50.12 45.48 0.907 4.4651 24.78 1.660
1373 49.62 45.44 0.916 4.4658 23.98 1.607
NbC

SC
1423 49.09 45.37 0.924 4.4665 22.87 1.533
1473 48.47 45.32 0.935 4.4673 20.44 1.371

U
1323 49.19 43.75 0.889 4.2340 6.04 0.345

AN
1373 49.38 42.56 0.862 4.2327 6.16 0.351
TiN
1423 49.45 41.32 0.836 4.2315 6.28 0.358

M
1473 49.58 39.84 0.804 4.2300 6.45 0.367

D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

Highlights

(1) Phase transformations and γ growth behaviors in superalloy 718 were studied;

(2) The γ grain growth curves at four isothermal temperatures were determined;

PT
(3) Precipitate coarsening was confirmed to control γ grain growth;

RI
(4) Nb and N are the controller for the coarsening of NbC and TiN, respectively;

SC
(5) Activation energy for γ grain growth was calculated as 417 kJ·mol-1.

U
AN
M
D
TE
C EP
AC