Materials Science & Engineering A 754 (2019) 29–37

Contents lists available at ScienceDirect

Materials Science & Engineering A

journal homepage: www.elsevier.com/locate/msea

Effect of heat treatment on mechanical property and microstructure of a T

powder metallurgy nickel-based superalloy
Hongyu Wua, Xiaoli Zhuangb, Yan Niec, Yunping Lia,∗, Liang Jianga,∗∗
a
State Key Laboratory for Powder Metallurgy, Powder Metallurgy Research Institute, Central South University, Changsha, 410083, China
b
State Key Laboratory for Advanced Metals and Materials, University of Science and Technology of Beijing, Beijing, 100083, China
c
Yuanmeng Precision Technology (Shenzhen) Institute, Shenzhen, China

A R T I C LE I N FO A B S T R A C T

Keywords: In this study, single and multi-step aging treatments (SAT and MAT) following a super-solvus solution treatment
Nickel-based superalloy were employed to tailor the mechanical property of a prototypical powder metallurgy (PM) nickel-based su-
Heat treatment peralloy. Vickers hardness and tensile test at room temperature were used to evaluate the mechanical properties
Single and multi-step aging treatment after different heat treatments. Microstructural evolution after heat treatments and tensile tests were revealed
Mechanical performance
through scanning electron microscope (SEM) and electron backscatter diffraction (EBSD) techniques. Our results
Microstructure
show that the MAT with a stabilization at 650 °C for 24 h between the solution and aging assists to achieve the
highest strength corresponding to microstructure of bimodal distribution of γ’ precipitates. The associated me-
chanism for variation in microstructures and mechanical properties are discussed based on thermodynamic
calculations and precipitation strengthening theory. In addition, an empirical relationship between Vickers
hardness and tensile strength of nickel-based superalloys is examined which is believed to be assistance for rapid
evaluating mechanical response of materials with different microstructures.

1. Introduction to meet the requirement of mechanical strength at elevated temperature

[3,13]. These physical or structure differences in turn leads to sig-
Powder metallurgy nickel-based superalloys are widely used as nificant interactions between each alloy element and variation in
turbine disc materials in aircraft engines and power generation turbines thermal properties (solid and liquid melting temperature, γ′ solvus
because of their excellent properties at elevated temperatures [1,2]. temperature) and thermodynamic behavior (nucleation, growth and
These alloys typically are composed of a face-centered cubic matrix (γ) coarsening behavior of γ’) are expected corresponding to composition
and a L12 structure precipitate (γ′) embedded in γ matrix [3]. The fluctuation in specific alloy. Thus a universal heat treatment strategy
conventional heat treatment, including solution treatment and aging which could be applicable to extensive alloys is unavailable. For in-
treatment, is considered as a powerful tool to tailor the microstructures stance, the typical heat treatment for IN100 is consisting of sub- or
and therefore to achieve the desired mechanical properties [3–5]. super-solvus solution, stabilization at 982 °C for 1 h and aging at 732 °C
Generally speaking, the solution temperature and soaking time can be for 8 h [14]. While for Udimet 720Li the maximized tensile properties
determined flexible to control the grain size [6]. Cooling rate plays an appeared under the following heat treatment: sub-solvus solution at
important role in tuning the features of γ′ precipitates, e.g. morphology, 1120 °C or 1080 °C and then aged at 760 °C for 16 h + 650 °C for 24 h
size distribution, composition and misfit between γ matrix and γ′ pre- [15]. The recommended heat treatment for Rene 88DT is first super-
cipitates (the misfit changes as a function of composition of γ and γ′) solvus solution treated and then aged at 760 °C for 8 h [16]. These in-
[7–11]. And aging temperature and time enables engineers to further vestigations indicate that it is necessary to develop specific heat treat-
improve the volume fraction and size distribution of γ’ precipitates ment procedure to optimize the mechanical performance when a new
[12]. superalloy has been designed.
Extensive alloying elements e.g., Mo, W, Nb, Ti and Ta with dif- Current work focuses on the development of an appropriate heat
ferent electronic structure and atomic radii compared with Ni [2] are treatment strategy for a prototypical PM nickel-based superalloy. The
added for solid solution strengthening and γ′ precipitates strengthening alloy was first super-solvus solution treated and then aged with SAT and

∗
Corresponding author.
∗∗
Corresponding author.
E-mail addresses: [email protected] (Y. Li), [email protected] (L. Jiang).

https://doi.org/10.1016/j.msea.2019.03.064
Received 19 December 2018; Received in revised form 12 March 2019; Accepted 13 March 2019
Available online 16 March 2019
0921-5093/ © 2019 Elsevier B.V. All rights reserved.
H. Wu, et al. Materials Science & Engineering A 754 (2019) 29–37

Table 1 argon gas protection and the heating rate was 10 °C/min.
Nominal chemical composition of the nickel base superalloy in wt.%. Field emission SEM (Quanta 650 FEG, FEI) was used to observe the
Ni Co Cr Mo W Al Ti Nb C B Zr Hf γ′ precipitates and fractography of the specimens after tensile test using
the secondary electron mode. Specimens for γ′ observation were po-
Bal. 26 13 4 4 3.2 3.7 0.95 0.05 0.025 0.05 0.2 lished with the standard metallographic procedure and then etched
using a H2O:HNO3:CH3COOH:HF = 33:33:33:1 solution. EBSD tech-
nique was performed using a step size of 0.3–0.6 μm and an accelerating
MAT. Corresponding mechanical responses were investigated using voltage of 20 KV to reveal the microstructure both before and after
Vickers hardness test and tensile test at room temperature. Scanning deformation. Vibration polishing was performed on specimens before
electron microscopy (SEM) and electron backscattered diffraction deformation for EBSD observation. While electrochemically polish was
(EBSD) were served to reveal the microstructural features produced applied to acquire the post mortem deformation microstructure [17].
under different heat treatments. Our experimental results show that the Statistical analysis for γ′ precipitates was carried out using the software
maximized mechanical performance occurs under the teat treatment ImageJ according to the method described in Ref. [18] and the reported
consisting of super-solvus solution at 1180 °C for 40 min, stabilization values for volume fraction and average diameter of γ′ precipitates were
at 650 °C for 24 h and aging at 760 °C for 16 h in the studied strategies. averaged form three measurements in each case. It should be noted that
The related strengthening mechanism is discussed based on the mi- due to the resolution limitation of SEM, the quantitative information
crostructure evolution. The developed features of γ’ precipitates are about tertiary γ’ was unpresented in this work.
analyzed using precipitation kinetic simulation. In addition, an em- Vickers hardness test at room temperature was carried out at the
pirical law between Vickers hardness and tensile strength is examined Buehler Micrometer 5100 hardness tester with a maximum load of 1 kg
which is expected to be assistance in fast assessing mechanical strength and dwell time of 15 s. Each value was obtained through averaging ten
of nickel-based superalloys. measurements.
Tensile tests were performed at room temperature using a com-
2. Experimental detail puter-controlled tensile testing machine on as-HIPed and heat treated
specimens with a constant initial strain rate of 10−3s−1. The dog bone-
The alloy used in this study is a newly developed polycrystalline shaped tensile specimen had a gauge length of 8 mm, a width of 3 mm
nickel-based superalloy by powder metallurgy (PM) processing route. and a thickness of 2 mm. The surface of the tensile specimens was po-
The nominal composition of this alloy is listed in Table 1. This alloy was lished down to a 1200-grit SiC paper to remove the recast layer re-
prepared by vacuum induction melting and then argon-atomization. sulting from the wire-based electrical discharge machining process.
Powder with a size range below 74 μm was collected and filled in a Three tensile specimens were examined for each condition for the re-
stainless steel container, which was outgassed and sealed by welding for producibility.
following hot isostatic pressing (HIP). The HIP process was carried out Thermodynamic calculation for driving force and number density of
at 1170 °C and 150 MPa for 4 h. The density of the alloy after HIP was γ’ was performed using the Pandat software with Panprecipitation
measured to be 8.22±0.01 g/cm3. Specimens were produced using module [19].
wire-based electrical discharge machining for the following heat
treatment. The schematic of the heat treatment strategies is presented 3. Results
in Fig. 1 and consisted of solution treatment at 1180 °C for 40 min (step
1), stabilization at 980 °C for 1 h (step 2) or 650 °C for 24 h (step 3) and 3.1. Mechanical properties evolution
aging at 760 °C for 8 h (step 4) or 16 h (step 5), as shown in Fig. 1. All
the cooling path applied for the above treatments was air cooling. The 3.1.1. Vickers hardness
heat treatment for all specimens was operated at a tube furnace with The Vickers hardness values of specimens under different heat

Fig. 1. Schematic showing the sequence of heat treatment procedures.

30
H. Wu, et al. Materials Science & Engineering A 754 (2019) 29–37

tensile strength (both ultimate tensile strength and yield strength) in-
crease but elongation reduce monotonically. Longer aging time (16 h,
specimen SA2) or stabilized at high temperature with a short time
(980 °C/1 h, specimen SSA1) achieve an increase in elongation com-
pared with specimen SA1 while maintain the tensile strength (specimen
SA2) or reduce the tensile strength slightly (for specimen SSA1) at the
same time. When a lower temperature (650 °C) and longer time (24 h)
stabilization process is applied, the highest tensile strength (among all
the aging treated specimens) is realized accompanied with an accep-
table elongation. These results show the advantage of stabilization
process at low temperature with a long duration in helping the alloy
reach an excellent tensile property at room temperature.

3.2. Microstructural evolution

3.2.1. γ’ precipitates
Fig. 4 shows the features of γ′ precipitates for as-HIPed and heat
treated specimens, it can be seen that heat treatment especially the
solution treatment show significant impact on tuning the features of γ′
Fig. 2. Vickers hardness evolution under different heat treatments (insert a precipitates. For the as-HIPed specimen, two type of γ′ are observed, the
typical indentation in the as-HIPed specimen). butterfly-shaped secondary γ’ (about 200 nm in diameter, detailed
statistical analysis for the γ′ will be presented in the discussion section)
and small spherical tertiary γ’ (< 50 nm, as indicated by yellow arrows
treatment are plotted in Fig. 2, where the insert is a typical indentation
in Fig. 4a). During the super-solvus solution treatment, γ′ precipitates
after hardness test in as-HIPed specimen. The diagonal length is about
dissolve and then re-produce in the solution cooling process [20]. Ac-
60 μm under the load of 1 kg. It can be seen that there is a significant
cordingly, all the specimens after solution treatment exhibit different
increase of hardness in solution treated (436 Hv) and solution + aging
morphologies and size distributions compared with that in as-HIPed
treated (466–479 Hv) specimens compared with the as-HIPed specimen
specimen. Fig. 4b shows the γ′ precipitates of specimen S with only
(401 Hv). In the cases of SAT (specimens SA1 and SA2), it seems that
solution treatment applied. Compared to as-HIPed specimen, all γ′
aging longer time offers limited help for reaching a higher hardness,
precipitates show near spherical or cubical morphology, no split of γ′
namely, a slight increment of 2 Hv is generated when aged additional
precipitates are observed in this case. The size of γ′ is much smaller and
8 h. However, the value of hardness drops from 473 to 466 Hv for the
the number density is higher. Direct aging at 760 °C for 8 (specimen
specimen SSA1 stabilized at 980 °C for 1 h and aged at 760 °C for 16 h.
SA1) and 16 h (specimen SA2) lead to a small increase in size of γ′
In contrast, stabilization at low temperature (650 °C for 24 h) followed
compared to specimen S, as shown in Fig. 4c and d, respectively. Both
by the same aging makes the specimen achieve the highest hardness
in these two cases, the γ′ precipitates present more cuboidal mor-
(479 Hv). These results clearly indicate that the solution and aging
phology. The specimens underwent stabilization between solution and
treatments play a vital role in improving the mechanical property in
aging treatments show lower number density of secondary γ′ pre-
nickel-based superalloys.
cipitates than specimen SA2 (Fig. 4e and f). However, the size of γ′ in
specimen SSA2 is smaller than SSA1 indicating stabilization at high
3.1.2. Tensile property temperature is helpful to the growth of γ’. From the SEM image (Fig. 4f)
Fig. 3 shows the tensile property obtained at room temperature of of SSA2, very fine spherical tertiary γ′ precipitates develop in the
the as-HIPed and heat treated specimens. After solution and aging channel between secondary γ’, as indicated by yellow arrows in the
treatment, higher tensile strength and lower elongation are observed in insert image of Fig. 4f.
all specimens. However, tensile property responses resulted by different
aging treatments are not always same. For example, solution (specimen 3.2.2. Grain size evolution
S) and aging treatment at 760 °C for 8 h (specimen SA1) make the Fig. 5 shows the grain size evolution during different heat treat-
ments. The as-HIPed specimen has the smallest grain size (with average
diameter of 8.5 μm, see Fig. 6a), while a small increase in grain size of
all specimens after solution and solution + aging treatment are ob-
served compared with as-HIPed specimen. The average grain size of
these specimens locates in the range of 9–11 μm based on the EBSD
measurements (Fig. 6). The largest grain size is corresponding to the
specimen SSA1, reaching an average diameter of ∼11 μm, as evident in
Figs. 5e and 6e. Interestingly, in all specimens the powder-like (or
circular) grains are developed whose morphology suggests that those
grains are inherited from the raw powders. In addition, all specimens
show bimodal grain size distribution, i.e., some large grains (indicated
by the black arrows in Fig. 6) are embraced by small ones.

3.3. Microstructural evolution upon deformation

In order to investigate the deformation details of the specimens, the

cross-sectional areas near the fractured tip are examined by EBSD IPF
and kernel average misorientation (KAM) maps, as shown in Fig. 7.
Fig. 3. The yield strength, ultimate tensile strength and total elongation of the Here, the KAM is calculated up to the first neighbor shell with a max-
as-HIPed and heat treated specimens. imum misorientation angle of 5°. The KAM map serves as a measure of

31
H. Wu, et al. Materials Science & Engineering A 754 (2019) 29–37

Fig. 4. Secondary electron SEM images showing the features of γ′ precipitates for specimens (a) as-HIPed, (b) S, (c) SA1, (d) SA2, (e) SSA1 and (f) SSA2. The insert in
(g) is a higher magnification image of specimen SSA2.

the deformation-induced local orientation gradients inside grain, and deformation at these regions [22], which mainly located in the grain
KAM value is higher in grains whose dislocation density is higher [21]. boundaries and small grains (IPF maps in Fig. 7). The KAM maps reveal
The unindexed areas (black area) appear due to the severe plastic similar results that KAM values increase at the grain boundaries and

Fig. 5. EBSD IPF maps of (a) as-HIPed, (b) S, (c) SA1, (d) SA2, (e) SSA1 and (f) SSA2 specimens.

32
H. Wu, et al. Materials Science & Engineering A 754 (2019) 29–37

Fig. 6. Grain size distribution of specimens (a) as-HIPed, (b) S, (c) SA1, (d) SA2, (e) SSA1 and (f) SSA2.

small grains. These observations illustrate that the plastic strain and Fig. 9. This result reveals that the weakened metallurgical bonding
lattice curvature are preferentially accommodated inside the small caused by PPBs could be nucleation site and propagation path for
grains and grain boundaries [21]. cracking and accelerates the failure of the alloy.

3.4. Examination of the fracture surface 4. Discussion

Fig. 8 shows the typical fracture morphologies of all the tested 4.1. Corresponding mechanism for microstructure and mechanical
specimens. All the specimens failed with a flat fracture surface that is performance
normal to the tensile direction (the insert in Fig. 8) which is believed to
be dominated by plane strain condition and suggests poor fracture 4.1.1. The role of solution cooling rate on tuning the size, morphology and
toughness [23]. From Fig. 8, crack initiation occurs mainly at the site size distribution of γ’ precipitates
near surface of specimen and then propagates by a transgranular mode From Fig. 4, the γ’ precipitates exhibit different features between as-
leading to the final fracture. In higher magnification images (Fig. 8), the HIPed specimen and solution treated specimen (S) which are strongly
fracture surface is covered with fully dimples suggesting the alloy related to the applied cooling approach [9,20]. The HIP process is si-
fractured in a ductile mode [24]. Besides, failure via inter-particle de- milar with solution treatment and due to the temperature applied in
bonding fracture is also observed which is primarily due to the cracking this work, 1170 °C, is such closer to the solution temperature, thus the
along the remained prior particle boundaries (PPBs) [25], as shown in difference observed in as-HIP and S specimens can be attributed to the

Fig. 7. EBSD (a, c, e, g, i, k) IPF maps and (b, d, f, h, j, l) KAM maps showing microstructural evolution of (a, b) as-HIPed, (c, d) S, (e, f) SA1, (g, h) SA2, (i, j) SSA1 and
(k, l) SSA2 specimens during tensile deformation.

33
H. Wu, et al. Materials Science & Engineering A 754 (2019) 29–37

Fig. 8. Secondary electron SEM images showing the fracture surfaces of as-HIPed specimens (a), specimen S (b), SA1 (c), SA2(d), SSA1 (e) and SSA2 (f). The insert is
a full view of the fracture surface.

different applied cooling rate reasonably. The as-HIPed specimen was Vα ΔG 2

Z=
cooled by furnace cooling, i.e., extremely slow cooling rate, after the 8π σ 3Rg T (3)
HIP process. In contrast, a much high cooling rate was used in the
specimen S. With the purpose to understand the underlying mechanism, 16πσ 2Deff
precipitation thermodynamic calculations are performed using the β∗ =
ΔG 2a4 (4)
Panprecipitation module in Pandat software under continuous cooling
process, in which the cooling rates are set as 1 °C/min and 360 °C/min 8Rg Tσa4
to simulate the current cases of as-HIPed and S specimens, respectively. τ=
Vα2 ΔG 2Deff (5)
In the calculations, the nucleation kinetics of precipitates is inferred
from classical nucleation theory (CNT) extended for multicomponent n −1
⎡ (Cpi − C0i )2 ⎤
systems [19]. Accordingly, the transient nucleation rate, J, is given by Deff = ⎢∑
C0i D0i ⎥ (6)
⎣ i=1 ⎦
∗
ΔG ⎞ −τ
J = Nv Zβ ∗ exp ⎜⎛− ⎟ exp ⎛ ⎞ where Vα is the molar volume of the matrix, a is the lattice constant of
⎝ R g T ⎠ ⎝ t ⎠ (1) γ′, Deff is the effective diffusivity as expressed as Eq. (6). In which Cpi
and C0i represent the mean concentration in γ′ and γ phases, respec-
in which Nv is the nucleation site density, Z is the Zeldovich factor, β ∗
tively.
denotes the atomic attachment rate, Rg is the gas constant, T is the
Fig. 10 shows the simulation results. For as-HIPed specimen soaked
temperature, τ is the incubation time for nucleation and t is the time.
at high temperature (1170 °C) and high pressure (150 MPa) for 4 h, a
And the activation energy for nucleation is calculated from
bimodal distribution consisting of butterfly-shaped secondary γ′ with
16πσ 3 lower number density and small spherical tertiary γ’ (Fig. 4a) were
ΔG∗ = developed. Fig. 10a shows the driving force for nucleation of γ′ in this
3(ΔGv + ΔGs )2 (2)
case (black line), the first γ′ nucleation burst occurs at higher tem-
ΔGv is the chemical driving force per volume for nucleation, ΔGs is the perature (compared with the high cooling rate case) as the driving force
elastic strain energy per volume of precipitate which is associated with decreases at higher temperature then these γ′ nuclei grow rapidly as a
the lattice misfit and σ is the interfacial energy of the γ/γ′ interface. For result of diffusion of the γ′ forming elements at higher temperature [26]
a spherical nucleus, which in turn reduces the supersaturation in matrix leading to the

Fig. 9. Fracture surface of the specimen SSA2. (a) a low magnification SEM image, (b) higher magnification images corresponding to the labelled areas b in (a).

34
H. Wu, et al. Materials Science & Engineering A 754 (2019) 29–37

Fig. 10. Calculated (a) driving force for nucleation and (b) number density of γ′ precipitates during continuous cooling from 1180 to 650 °C. The as-HIPed specimen
was cooled by furnace cooling (slow cooling rate) after the HIP process while the specimen S was air cooled following solution treatment (high cooling rate). Thus,
two cooling rates, 1 and 360 °C/min, were applied to simulate the cooling process of as-HIPed and S specimens, respectively. During the solution cooling process with
different cooling rate, variational features of γ′ precipitates can be obtained.

“shut-off” by soft impingement of the γ′ nucleation burst [27]. This is

supported by the prediction of driving force, which first increases with
decreasing temperature until the γ′ nucleation occurs. After that, it
fluctuates with cooling process going on and start to increase slowly
again blew 800 °C as a result of that higher undercooling is available. It
leads to the second burst of γ′ nucleation and corresponding fine ter-
tiary γ′ precipitates (or bimodal distribution of γ′) [26,27]. In the case
of fast cooling rate (corresponding to the air cooling, specimen S), the
driving force shows similar characteristic with that in as-HIPed spe-
cimen at the initial stage of first nucleation burst of γ’. However, it
increases within a much shorter period and in this context, the shut-off
of nucleation may not be activated resulting in the unimodal particle
size distribution in this case [28].
The number density of precipitates in these two specimens are also
consistent with the calculation results (Fig. 10b). It is obvious that the
first nucleation burst of γ′ occurs at lower temperature for specimen S
under a higher nucleation driving force (Fig. 10a) [29]. Thus, larger
nucleation rate of γ’ is expected to lead to a higher number density, see
Fig. 10b. Fig. 11. Variation for average diameter and volume fraction of secondary γ′
Due to the available time at higher temperature is adequate for precipitates up different heat treatments.
diffusion of γ′-forming elements, larger size of γ′ is observed in as-HIPed
specimen compared to specimen S [30,31]. Additionally, the higher closer to the cuboidal morphology and almost no split of γ′ precipitates
number density for γ′ limits the distance between individual pre- are observed. It implies the fast cooling rate strongly limits the element
cipitates, namely, the available solute atoms for growth in fast cooling diffusion process which maintains a moderate misfit between γ and γ’
case (S) is less than that in slow cooling one (as-HIPed specimen) [28]. leading to a restricted anisotropic split from spherical or cuboidal
As a result, the smaller size of γ′ is reasonable under these two com- morphology [33,34].
bined mechanisms. When stabilization procedure is applied in MAT, the γ′ precipitates
The developed split γ′ precipitates in as-HIPed specimen are also exhibit different features compared with the cases of SAT. For the case
attributed to the larger mobility of γ′-forming elements which produces of specimen SSA1 (stabilized at 980 °C for 1 h), the secondary γ′ pre-
a large misfit between γ matrix and γ′ precipitate [32] leading to ani- cipitates display larger size and lower number density than that in SA1
sotropic growth of γ’ [33,34]. and SA2 specimens. This is due to the fast diffusion rate of alloying
elements and lower undercooling at the stabilization temperature
4.1.2. The evolution of γ’ precipitates during SAT and MAT (980 °C) [28]. Besides, the supersaturation in matrix is reduced sig-
Fig. 11 shows the corresponding size and volume fraction of sec- nificantly at that higher temperature (980 °C). Similar phenomenon is
ondary γ′ under different heat treatments. As discussed in the above observed in the specimen SSA2, however, some fine tertiary γ′ pre-
section, the much smaller size and slightly decreased volume fraction in cipitates emerge in the γ channels between the secondary γ′ precipitates
γ′ for specimen S can be rationalized through much higher cooling rate. (Fig. 4f), which may be related to the slight lower volume fraction of
The SAT at 760 °C for 8 (specimen SA1, Figs. 4c) and 16 h (specimen secondary γ′ in this case. This phenomenon is attributed to the lower
SA2, Fig. 4d) result in a larger average size of γ′ accompanied with stabilization temperature (650 °C), the γ matrix keeps some super-
higher volume fraction compared to specimen S indicating growth and saturation assisting the nucleation of γ′ in the following cooling or
coarsening occur during this stage [35]. Both processes are controlled aging process to form the tertiary γ’ [36], as indicated by yellow arrows
by diffusion mechanism while the supersaturation drives the growth of in the insert image of Fig. 4f.
γ’ [31] and surface energy dominates the coarsening of γ’ [28]. The tiny
fluctuation between specimen SA1 and SA2 may be induced by statis-
tical error. Additionally, the γ′ precipitates in these two cases are much

35
H. Wu, et al. Materials Science & Engineering A 754 (2019) 29–37

4.1.3. Mechanical responses under different heat treatments

The corresponding hardness and tensile strength are strongly re-
lated to the γ′ precipitates. Extensively investigations [37–39] have
been carried out to establish the correlation between the features of γ′
precipitates and strength of nickel-based superalloys and provide deep
insight on this issue. There is a critical size for γ′ precipitates to obtain
the maximum strength, while away form (smaller or larger) this critical
value will deliver a weakened strengthening effect [1,37,38]. It should
be note that this critical size depends on the specific composition,
namely, the optimized size for γ′ changes upon different alloys with
varied compositions. In contrast, the enlarged volume fraction of γ′
precipitates always contributes to the strength positively. Take the
consideration of the similar grain size among all the specimens (Fig. 6),
it is reasonable to discuss the mechanical responses under different heat
treatments based on the variation of γ′ precipitates. The lowest strength
in as-HIPed specimen is largely attributed to the coarse-γ′ precipitates
[14] as well as the low volume fraction. When smaller γ′ precipitates
generated during the solution cooling process in specimen S, a higher Fig. 12. Relationship between tensile strength and Vickers hardness.
strength is achieved in this case. For SAT specimens, the enhanced
volume fraction of γ′ dominates over the increase size in γ′ precipitates σUTS = 3.14Hv (8)
resulting in a better mechanical strength than the specimen S accom-
where σYS and σUTS represent the yield strength and ultimate tensile
panied with lower ductility. On one hand, the higher volume fraction of
strength, respectively. Hv is the value of Vickers hardness. From Fig. 12,
γ′ precipitates improve the strength of the alloy, on the other hand, it
both points of the ultimate tensile strength and yield strength are lo-
increases the susceptibility of damage tolerance leading to an inferior
cated near the fitted line with moderate deviation. However, a litter
ductility [40]. This combined mechanical performance, referred as the
higher correlation coefficient (R2) for yield strength vs Vickers hardness
strength-ductility trade-off effect, is prevalent in metallic materials
is observed. It should be attributed to the different work hardening
[41]. It is interesting that higher total elongation is observed in SA2
effect in different specimens under variation plastic strains (15%–24%
specimen compared with SA1, which may imply that coarse-γ′ pre-
engineering tensile strain) [5]. It seems that the Vickers hardness has a
cipitates can recover some ductility in precipitate-strengthened nickel-
better proportional relationship with yield strength from the view of
based superalloys. Stabilized at high temperature with a short time
correlation coefficient. However, using Vickers hardness to assess the
(980 °C/1 h, specimen SSA1) slightly decreases the strength while its
ultimate tensile strength via Eq. (8) is also acceptable. By these above
elongation is maintained compared with specimen SA2. This combined
relationships, the tensile strength of nickel-based superalloys or the
performance can be attributed to the larger size of γ′ precipitates in this
influence of alloy element/grain size/γ′ precipitate features/heat
condition. Coarse-γ′ precipitates generate negative effect in strength-
treatment procedures on the tensile property can be rapidly assessed.
ening the alloy while benefit the ductility at the same time. This result
supports the above hypothesis that the high ductility is related to larger
γ′ precipitates. The most excellent property, i.e. the highest tensile 5. Conclusions
strength and an acceptable elongation, is achieved in specimen SSA2
which suffered a lower temperature stabilization process (650 °C/24 h) 1. Different aging treatments are applied to a prototypical PM nickel-
prior to aging. Compared its γ′ precipitates feature with that in spe- based superalloy and the two-step aging, namely, stabilized at
cimen SSA1, the higher strength should be attributed to the combina- 650 °C for 24 h followed by aged at 760 °C for 16 h, can make the
tion of smaller size of secondary γ’ (Fig. 11) and emergence of tertiary γ′ alloy achieve the highest hardness, tensile strength and an accep-
in the channel between secondary γ’ (Fig. 4f) [30]. These above results table ductility at room temperature.
show the advantage of stabilization process (low temperature and long 2. The solution cooling rate plays an important role in tailoring the
time) in helping the alloy reach an excellent tensile property at room morphology, size and size distribution of γ′ precipitates which is
temperature. clarified by the precipitation calculation enabled by multi-
component thermodynamic database.
3. Stabilization at high temperature (980 °C for 1 h) before aging leads
4.2. Relationship between Vickers hardness and tensile strength of nickel- to a coarsening of γ′ precipitates, which in turn results a lower
based superalloys corresponding tensile strength than the case of single aging treated
or stabilized at low temperature.
It is unambiguous that Vickers hardness is positively related to the 4. All the specimens fracture mainly in a transgranular ductile mode at
tensile strength for nickel-based superalloys and other metallic mate- room temperature tensile test. Besides, the PPBs act as weakened
rials [42]. Jackson et al. [5] reported that the hardness correlates better interfaces could be initial sites for cracking nucleation and propa-
with the ultimate tensile strength rather than the yield strength from gation resulting in earlier fracture of the material.
the perspective of indentation induced strain. However, based on nu- 5. Vickers hardness is examined has a linear relationship with both
merous experimental results, a more linear relationship has been ob- ultimate tensile strength and yield strength which is helpful in ra-
served between Vickers hardness and yield strength in structural ma- pidly assessing the strength of a nickel-based superalloy with dif-
terials [3]. In this work, we examine these relations using the data both ferent microstructures.
from our work and literature [3,5,43,44]. It should be note that only the
results of nickel alloys and nickel-based superalloys are considered in Acknowledgement
the current discussion. The corresponding outcomes are plotted in
Fig. 12. Using linear fitting method, the quantitative relations are de- This work was supported by the National Key Research and
scribed by the following equations, Development Program of China (2016YFB0700300). And H. W. ap-
preciates the financial support from the outstanding graduate project of
σYS = 2.34Hv (7)
Advanced Non-ferrous Metal Structural Materials and Manufacturing

36
H. Wu, et al. Materials Science & Engineering A 754 (2019) 29–37

Collaborative Innovation Center. 131 (2017) 323–335.

[22] M.J. Yao, K.G. Pradeep, C.C. Tasan, D. Raabe, A novel, single phase, non-equiatomic
FeMnNiCoCr high-entropy alloy with exceptional phase stability and tensile duc-
References tility, Scripta Mater. 72 (2014) 5–8.
[23] J.W. Hutchinson, Singular behaviour at the end of a tensile crack in a hardening
[1] R.C. Reed, The Superalloys: Fundamentals and Applications, Cambridge university material, J. Mech. Phys. Solids 16 (1) (1968) 13–31.
press, Cambridge, UK, 2008. [24] Z. Li, D. Raabe, Influence of compositional inhomogeneity on mechanical behavior
[2] T.M. Pollock, S. Tin, Nickel-based superalloys for advanced turbine engines: of an interstitial dual-phase high-entropy alloy, Mater. Chem. Phys. 210 (2018)
chemistry, microstructure, and properties, J. Propuls. Power 22 (2) (2006) 29–36.
361–374. [25] G.A. Rao, M. Kumar, M. Srinivas, D.S. Sarma, Effect of standard heat treatment on
[3] T. Osada, Y. Gu, N. Nagashima, Y. Yuan, T. Yokokawa, H. Harada, Optimum mi- the microstructure and mechanical properties of hot isostatically pressed superalloy
crostructure combination for maximizing tensile strength in a polycrystalline su- inconel 718, Mater. Sci. Eng., A 355 (1–2) (2003) 114–125.
peralloy with a two-phase structure, Acta Mater. 61 (5) (2013) 1820–1829. [26] P.M. Sarosi, B. Wang, J.P. Simmons, Y. Wang, M.J. Mills, Formation of multimodal
[4] H.Y. Li, J.F. Sun, M.C. Hardy, H.E. Evans, S.J. Williams, T.J.A. Doel, P. Bowen, size distributions of γ' in a nickel-base superalloy during interrupted continuous
Effects of microstructure on high temperature dwell fatigue crack growth in a cooling, Scripta Mater. 57 (8) (2007) 767–770.
coarse grain PM nickel based superalloy, Acta Mater. 90 (2015) [355]–[369]. [27] Y.H. Wen, J.P. Simmons, C. Shen, C. Woodward, Y. Wang, Phase-field modeling of
[5] M.P. Jackson, R.C. Reed, Heat treatment of UDIMET 720Li: the effect of micro- bimodal particle size distributions during continuous cooling, Acta Mater. 51 (4)
structure on properties, Mater. Sci. Eng., A 259 (1) (1999) 85–97. (2003) 1123–1132.
[6] D.M. Collins, B.D. Conduit, H.J. Stone H J, M.C. Hardy, G.J. Conduit, R.J. Mitchell, [28] R. Radis, M. Schaffer, M. Albu, G. Kothleitner, P. Pölt, E. Kozeschnik, Multimodal
Grain growth behavior during near-γ′ solvus thermal exposures in a polycrystalline size distributions of γ’ precipitates during continuous cooling of UDIMET 720 Li,
nickel-base superalloy, Acta Mater. 61 (9) (2013) 3378–3391. Acta Mater. 57 (2009) 5739–5747.
[7] J.Y. Hwang, S. Nag, A.R.P. Singh, R. Srinivasan, J. Tiley, G.B. Viswanathan, [29] S.S. Babu, M.K. Miller, J.M. Vitek, S.A. David, Characterization of the micro-
H.L. Fraser, R. Banerjee, Compositional variations between different generations of structure evolution in a nickel base superalloy during continuous cooling condi-
γ’ precipitates forming during continuous cooling of a commercial nickel-base su- tions, Acta Mater. 49 (20) (2001) 4149–4160.
peralloy, Metall. Mater. Trans. 40 (13) (2009) 3059–3068. [30] H. Wu, J. Li, F. Liu, L. Huang, X. Zeng, Q. Fang, Z. Huang, L. Jiang, A high-
[8] Y.Q. Chen, E. Francis, J. Robson, M. Preuss, S.J. Haigh, Compositional variations for throughput methodology search for the optimum cooling rate in an advanced
small-scale gamma prime (γ’) precipitates formed at different cooling rates in an polycrystalline nickel base superalloy, Mater. Des. 128 (2017) 176–181.
advanced Ni-based superalloy, Acta Mater. 85 (2015) 199–206. [31] Q. Chen, J. Jeppsson, J. Ågren, Analytical treatment of diffusion during precipitate
[9] J. Tiley, G.B. Viswanathan, R. Srinivasan, R. Banerjee, D.M. Dimiduk, H.L. Fraser, growth in multicomponent systems, Acta Mater. 56 (8) (2008) 1890–1896.
Coarsening kinetics of γ’ precipitates in the commercial nickel base Superalloy René [32] J.S. Van Sluytman, T.M. Pollock, Optimal precipitate shapes in nickel-base γ–γ′
88 DT, Acta Mater. 57 (8) (2009) 2538–2549. alloys, Acta Mater. 60 (4) (2012) 1771–1783.
[10] M. Zhang, G.Q. Liu, H. Wang, B.F. Hu, Stability of γ′ multimodal microstructure in a [33] Y. Chen, T.J.A. Slater, M. Bai, R. Mitchell, O. Ciuca, M. Preuss, S.J. Haigh, An in-
Ni-based powder metallurgy superalloy, Sci. China Technol. Sci. 61 (2018) vestigation of diffusion-mediated cyclic coarsening and reversal coarsening in an
1824–1828 https://doi.org/10.1007/s11431-018-9372-8. advanced Ni-based superalloy, Acta Mater. 110 (2016) 295–305.
[11] H. Wu, Z. Huang, N. Zhou, J. Chen, P. Zhou, L. Jiang, A study of solution cooling [34] X. Fan, Z. Guo, X. Wang, J. Yang, J. Zou, Morphology evolution of γ′ precipitates in
rate on γ’precipitate and hardness of a polycrystalline Ni-based superalloy using a a powder metallurgy Ni-base superalloy, Mater. Char. 139 (2018) 382–389.
high-throughput methodology, Mater. Sci. Eng. 739 (2019) 473–479. [35] A.J. Goodfellow, E.I. Galindo-Nava, K.A. Christofidou, N.G. Jones, T. Martin,
[12] R.S. Moshtaghin, S.I.R.O.U.S. Asgari, Growth kinetics of γ′ precipitates in superalloy P.A.J. Bagot, C.D. Boyer, M.C. Hardy, H.J. Stone, Gamma prime precipitate evo-
IN-738LC during long term aging, Mater. Des. 24 (5) (2003) 325–330. lution during aging of a model nickel-based superalloy, Metall. Mater. Trans. 49 (3)
[13] C.L. Jia, C.C. Ge, Q.Z. Yan, Microstructure evolution and mechanical properties of (2018) 718–728.
disk superalloy under multiplex heat treatment, Mater. Sci. Eng., A 659 (2016) [36] A.R.P. Singh, S. Nag, J.Y. Hwang, G.B. Viswanathan, J. Tiley, R. Srinivasan,
287–294. H.L. Fraser, R. Banerjee, Influence of cooling rate on the development of multiple
[14] R.W. Kozar, A. Suzuki, W.W. Milligan, J.J. Schirra, M.F. Savage, T.M. Pollock, generations of γ' precipitates in a commercial nickel base superalloy, Mater. Char.
Strengthening mechanisms in polycrystalline multimodal nickel-base superalloys, 62 (9) (2011) 878–886.
Metall. Mater. Trans. 40 (7) (2009) 1588–1603. [37] D.M. Collins, H.J. Stone, A modelling approach to yield strength optimisation in a
[15] J.R. Vaunois, J. Cormier, P. Villechaise, A. Devaux, B. Flageolet, Influence of Both γ′ nickel-base superalloy, Int. J. Plast. 54 (2014) 96–112.
Distribution and Grain Size on the Tensile Properties of UDIMET 720Li at Room [38] E.I. Galindo-Nava, L.D. Connor, C.M.F. Rae, On the prediction of the yield stress of
temperature[C]//7th International Symposium on Superalloys 718 and Derivatives, unimodal and multimodal γ’ Nickel-base superalloys, Acta Mater. 98 (2015)
(2010), pp. 199–213. 377–390.
[16] D.D. Krueger, R.D. Kissinger, R.G. Menzies, C.S. Wukusick, Fatigue crack growth [39] Q. Fang, L. Li, J. Li, H. Wu, Z. Huang, B. Liu, Y. Liu, P.K. Liaw, A statistical theory of
resistant nickel-base article and alloy and method for making, U.S. Patent, 4,957, probability-dependent precipitation strengthening in metals and alloys, J. Mech.
567, (1990). Phys. Solids 122 (2019) 177–189.
[17] L. Jiang, H. Wang, P.K. Liaw, C.R. Brooks, D.L. Klarstrom, Characterization of the [40] T. Yang, Y.L. Zhao, Y. Tong, Z.B. Jiao, J. Wei, J.X. Cai, X.D. Han, D. Chen, A. Hu,
temperature evolution during high-cycle fatigue of the ULTIMET superalloy: ex- J.J. Kai, K. Lu, Y. Liu, C.T. Liu, Multicomponent intermetallic nanoparticles and
periment and theoretical modeling, Metall. Mater. Trans. 32 (9) (2001) 2279–2296. superb mechanical behaviors of complex alloys, Science 362 (6417) (2018)
[18] M. Pinz, G. Weber, W.C. Lenthe, M.D. Uchic, T.M. Pollock, S. Ghosh, Microstructure 933–937.
and property based statistically equivalent RVEs for intragranular γ−γ' micro- [41] Z. Li, K.G. Pradeep, Y. Deng, D. Raabe, C.C. Tasan, Metastable high-entropy dual-
structures of Ni-based superalloys, Acta Mater. 157 (2018) 245–258. phase alloys overcome the strength–ductility trade-off, Nature 534 (7606) (2016)
[19] W. Cao, S.-L. Chen, F. Zhang, K. Wu, Y.A. Chang, R. Schmid-Fetzer, W.A. Oates, 227–230.
PANDAT software with PanEngine, PanOptimizer and PanPrecipitation for multi- [42] T. Osada, N. Nagashima, Y. Gu, Y. Yuan, T. Yokokawa, H. Harada, Factors con-
component phase diagram calculation and materials property simulation, Calphad tributing to the strength of a polycrystalline nickel–cobalt base superalloy, Scripta
33 (2) (2009) 328–342. Mater. 64 (9) (2011) 892–895.
[20] A.R.P. Singh, S. Nag, S. Chattopadhyay, Y. Ren, J. Tiley, G.B. Viswanathan, [43] C. Joseph, C. Persson, M.H. Colliander, Influence of heat treatment on the micro-
H.L. Fraser, R. Banerjee, Mechanisms related to different generations of γ’ pre- structure and tensile properties of Ni-base superalloy Haynes 282, Mater. Sci. Eng.,
cipitation during continuous cooling of a nickel base superalloy, Acta Mater. 61 (1) A 679 (2017) 520–530.
(2013) 280–293. [44] V.A. Popovich, E.V. Borisov, A.A. Popovich, V.S. Sufiiarov, D.V. Masaylo, L. Alzina,
[21] Z. Li, C.C. Tasan, K.G. Pradeep, D. Raabe, A TRIP-assisted dual-phase high-entropy Impact of heat treatment on mechanical behaviour of Inconel 718 processed with
alloy: grain size and phase fraction effects on deformation behavior, Acta Mater. tailored microstructure by selective laser melting, Mater. Des. 131 (2017) 12–22.

37