Additive Manufacturing 60 (2022) 103280

Contents lists available at ScienceDirect

Additive Manufacturing
journal homepage: www.elsevier.com/locate/addma

Using post-processing heat treatments to elucidate precipitate

strengthening of additively manufactured superalloy 718☆
Stephen Taller a, *, Ty Austin b, c
a
Nuclear Energy and Fuel Cycle Division, Oak Ridge National Laboratory, USA
b
Manufacturing Sciences Division, Oak Ridge National Laboratory, USA
c
University of Tennessee, Knoxville, USA

A R T I C L E I N F O A B S T R A C T

Keywords: The poor machinability and extensive work hardening of Ni-based superalloys makes additive manufacturing an
Alloy 718 attractive option for producing geometrically complex components with distinct microstructures. Although
Heat treatment previous studies show recovery of high strength at room temperature, very few studies demonstrate successful
Precipitation
properties at elevated temperatures required for industrial applications. The objectives of this study are to
Tensile properties
Laser powder bed fusion
present a post-build heat treatment for high strength across a wide temperature range, determine the strength
contribution of nanoscale precipitating phases to the overall mechanical properties of superalloy 718, and from
these, provide a comprehensive microstructure-property relationship for wrought and AM 718 to guide efforts to
simulate the properties of AM components. Laser powder bed fusion–produced superalloy 718 was characterized
at multiple length scales using scanning electron microscopy and transmission electron microscopy in the as-built
condition and with multiple heat treatments designed to form combinations of γ′ , γ′′ , and δ precipitates. Uniaxial
tensile tests performed from room temperature to 600 ◦ C on subsize specimens determined the yield strength,
elastic modulus, ultimate tensile strength, fracture stress, and uniform elongation. Precipitates in this work
proved to be weak barriers to dislocation motion through a dispersed barrier model, but they provided strength
to the alloy through their consistent high density. The relative contribution to the yield strength from γ′′
remained consistent between 48% and 57% of the total strength up to 600 ◦ C, the primary influence on the high
temperature strength of superalloy 718. The strength factors for γ′′ and δ precipitates were found to trend
inversely with tensile test temperature and may be attributable to the differences in precipitate coherency. A
post-build heat treatment is recommended to maintain high strength at elevated temperatures. A quantitative
microstructure-property relationship, dependent on precipitate size, density, and morphology, was derived and
can estimate the yield strength across a wide temperature range applicable to the operational regimes for su­
peralloy 718.

1. Introduction their strength primarily through solid solution strengthening and/or

secondary precipitating phases in the lattice [3], such as the interme­
Ni-based superalloys are a primary candidate alloy class for high- tallic phases δ, γ′ , or γ′′ . The poor machinability and extensive work
temperature applications in the petrochemical, aerospace, and nuclear hardening of Ni-based superalloys makes additive manufacturing (AM)
power industries because of their intrinsic resistance to creep, their an attractive option to produce geometrically complex components.
adequate corrosion resistance, and the ability to tailor the microstruc­ Powder bed fusion – laser beam (PBF-LB) is an AM process used to build
ture for high strength [1,2]. These high-strength Ni-based alloys gain parts through sequential application of thin layers (20–100 µm) of metal

☆
This manuscript has been authored in-part by UT-Battelle, LLC under Contract No. DE-AC05–00OR22725 with the U.S. Department of Energy. The United States
Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a non-exclusive, paid-up,
irrevocable, world-wide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government purposes. The
Department of Energy will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (http://energy.gov/
downloads/doe-public-access-plan).
* Correspondence to: 1 Bethel Valley Rd., Oak Ridge, TN 37830, USA.
E-mail address: [email protected] (S. Taller).

https://doi.org/10.1016/j.addma.2022.103280
Received 3 August 2022; Received in revised form 26 October 2022; Accepted 11 November 2022
Available online 14 November 2022
2214-8604/© 2022 Elsevier B.V. All rights reserved.
S. Taller and T. Austin Additive Manufacturing 60 (2022) 103280

feedstock powder melted by a scanning laser, resulting in rapid solidi­ 2. Methodology

fication and thermal cycling, and has been used to produce parts from a
Ni-based superalloy, 718 [4–6]. 2.1. Composition, fabrication and heat treatments
Despite the continual improvement in processing parameters, the
microstructures of AM materials typically contain a high density of pores The additively manufactured superalloy 718, designated AM718,
and other nonequilibrium features that encourage the use of post- used in this investigation was printed in one build cycle by using PBF-LB
fabrication heat treatments [5]. Like alloy 718, many nickel superal­ on the Concept X-Line 2000R at the Manufacturing Demonstration Fa­
loys have compositions that allow for the formation of many different cility (MDF) at Oak Ridge National Laboratory (ORNL, build ID:
precipitating phases, all of which may result from PBF-LB. Several of 20200206XL) using a powder with the composition listed in Table 1
these appear to be beneficial for the alloy at high temperatures (γ′ and with process parameters of a laser power of 370 W, spot size of 300 µm,
γ′′ ), whereas alternative phases (δ and Laves) are detrimental [7,8]. scanning speed of 500 mm/s, and hatch spacing of 0.16 mm with 67◦
Post-build heat treatments typically include direct aging, solution rotation between layers. The plate was sectioned into four pieces: three
annealing, hot isostatic pressing (HIP), or a combination [9–13] in an for subsequent heat treatments, and one for the as-built (ASB) micro­
attempt to produce a high strength component [14]. The literature of structure. Three heat treatments, summarized in Table 2 and shown
post-build heat treatments show varying degrees of success [6,15,16] to schematically in Fig. 1 with the time-temperature-transformation (TTT)
achieve a yield strength in the range of 800–1300 MPa at room tem­ diagram for superalloy 718, were designed to reduce the complexity of
perature but with visible δ and Laves particles [17]. In many cases, an the AM microstructure and to allow for comparative analysis of the
additional step of a high temperature homogenization at 1100 ◦ C or microstructure and mechanical properties. Based on the work performed
greater was needed to result in high strength [18,19] and does not al­ by Poole et al. [25] to identify the δ and Laves phase solvus tempera­
ways eliminate the Laves phase [20–22] or porosity [23]. For a tures, a two-step homogenization treatment of 1174 ◦ C for 2 h followed
comprehensive review, the reader is referred to a recent article [24] on by 1204 ◦ C for 6 h was chosen to remove all secondary phases, including
laser bed fusion of Ni-based superalloys highlighting many of the pro­ the detrimental Laves phase that typically forms during AM and com­
cessing parameter studies, heat treatments, and properties from which petes with δ, γ′ , and γ′′ precipitates for Nb [14]. Annealing further at
this study was pursued, and the challenges with unpredictable heat 1093 ◦ C for 1 h with no additional heat treatments provides a reference
treatments for precipitation hardening. Although high strength can be case solution-annealed alloy 718 with minimal precipitation, designated
recovered at room temperature, very few studies demonstrate success as AM718-HM. The extensive knowledge of the precipitation in alloy
over the wide operating temperature regime from room temperature to 718 [26–30] led to the adoption of a standard heat treatment to produce
650 ◦ C for superalloy 718. There is still a need to determine a δ precipitates that consists of solution annealing at 945 ◦ C followed by
post-processing heat treatment of additively manufactured superalloy aging at 718 ◦ C and then 621 ◦ C for 8 h at each step, shortened to
718 that results in high strength at the elevated temperatures expected 945 ◦ C/718 ◦ C/621 ◦ C [30–32], and designated as AM718-HT1. A
during service and eliminates detrimental phases. higher temperature heat treatment of 1093 ◦ C/718 ◦ C/621 ◦ C was
Furthermore, as the strength of the 3D-printed component depends chosen to form a large density of small γ′ and γ′′ precipitates based on the
on the nanoscale particles that develop during the repetitive heating and work performed in [8,30,31], designated as AM718-HT2. A traditional
cooling cycles associated with AM, there is a significant need to un­ wrought alloy 718, designated W718 heat Z1653 was prepared ac­
derstand how each precipitate that can be formed from AM contributes cording to ASTM specifications [33], was used as a control, and was
to the overall strength of the alloy. The objectives of this study are 1) to compared with the AM specimens. However, as demonstrated further in
demonstrate the combination of a two-step post-build high temperature the article, the same heat treatment for AM718-HT2 and W718 resulted
normalization and aging heat treatments on the microstructure and in distinct microstructures and mechanical properties, and the reader is
mechanical properties of alloy 718, and 2) use the post-heat treatment referred to the Appendix for more details. After heat treatment, speci­
properties to determine the strength contribution of nanoscale precipi­ mens were fabricated into the form of SS-J2 subsize tensile specimens
tating phases to the overall mechanical properties of Superalloy 718 [34], displayed in Fig. 2, for uniaxial tensile tests with a gauge 5 mm
through a combination of electron microscopy and uniaxial tensile long, 0.5 mm thick, and 1.2 mm wide using electrical discharge
straining. The results of this study will be used to recommend a post- machining (EDM). The SS-J2 geometry was chosen to minimize the
build heat treatment and provide a physically informed equation con­ amount of material needed for testing, provide adequate tensile property
necting the microstructure and yield strength across elevated tempera­ measurements [35], and allow for future studies in which subsize
tures for modeling and simulation. Superalloy 718 was characterized in specimens will be used. Each specimen was individually engraved via a
the as-built condition, and with multiple heat treatments, to form δ laser with a unique ID for tracking from the original build plate to the
precipitates, as well as γ′ and γ′′ precipitates at multiple length scales tensile specimen. Previously obtained results for room temperature
using scanning electron microscopy (SEM) and transmission electron properties and preliminary characterization were reported by Taller
microscopy (TEM) to correlate the microstructure to the bulk properties. et al. [36], from which this manuscript builds.
Uniaxial tensile tests were performed under conditions ranging from
room temperature to 600 ◦ C on subsize specimens to determine the 2.2. Microscopy
common engineering properties of yield strength, elastic modulus, ul­
timate tensile strength, uniform elongation, and total elongation. An Prior to characterization, standard metallographic grinding using
Orowan dispersed barrier hardening model was used to assess the con­ successively finer grits of SiC paper and polishing using diamond solu­
tributions of microstructural features to the yield strength as a function tions with successively decreasing particle sizes of 6 µm, 3 µm, and
of tensile test temperatures, and from the calculated trends, to elucidate 0.25 µm were performed until a mirror-like surface was achieved with
the role of precipitate coherency on tensile strength. The microstructure- ~0.02 µm silica particles. Surface characterization was performed on a
property relationship derived from this work can be used to estimate the Tescan MIRA3 GHM scanning electron microscope (SEM) using 20 keV
yield strength of wrought and AM superalloy 718 components across a electrons. Electron backscatter diffraction (EBSD) Kikuchi band patterns
wide temperature range applicable to the operational regimes for nickel were collected on an Oxford Instruments Symmetry EBSD detector with
superalloys. AZtec software using a step size of 1 µm. At each point, the AZtec soft­
ware indexes the Kikuchi band structure and determines the Euler angles
of the lattice. The open source MTEX toolbox [40] was used to generate
orientation maps and inverse pole figures and to determine grain sizes
from the Euler angles exported from AZtec for grains consisting of more

2
S. Taller and T. Austin Additive Manufacturing 60 (2022) 103280

Table 1
Composition of superalloy 718 powder used for PBF-LB in wt% provided by the vendor for Lot 119 and a commercially purchased wrought superalloy 718 for
comparison as measured with x-ray fluorescence.
Designation Ni Fe Cr Nb Mo Ti Al Si C N O Co B Mn P S

AM718-Lot 119 Bal. 18.22 18.99 5.15 3.0 0.93 0.50 0.04 0.04 0.012 0.016 0.1 0.001 0.02 0.006 0.001
W718-Z1653 Bal. 18.77 17.56 5.18 2.89 0.94 0.49 0.08 0.02 – – 0.05 0.004 0.03 0.008 <
0.0003
ASTM-B670 50.0 – Bal. 17.0 – 4.75 – 2.8 – 0.65 – 0.2 – 0.35 0.08 – – 1.0 0.006 0.35 0.015 0.015
Reference 55.0 21.0 5.50 3.3 1.15 0.8 max max max max max max max
[37]

Table 2
Heat treatments for AM718 and W718 to produce increasingly simpler microstructures for characterization and mechanical property measurement.
Designation Homogenization (◦ C/ Cooling Solution anneal Cooling Aging #1 Cooling rate to reach aging Aging #2 Cooling rate to
h) (◦ C/h) (◦ C/h) #2 (◦ C/h) (◦ C/h) room temp.

W718- 1174/2 + 1204/6 WC 1093/1 AC 718/8 55 621/8 AC

Z1653
AM718-HT1 1174/2 + 1204/6 WC 945/1 AC 718/8 55 621/8 AC
AM718-HT2 1174/2 + 1204/6 WC 1093/1 AC 718/8 55 621/8 AC
AM718-HM 1174/2 + 1204/6 WC 1093/1 AC N/A N/A N/A N/A

The cooling between steps was performed using either water cooling (WC) or air cooling (AC).

Fig. 1. The (a) time-temperature diagram for the heat treatments in this study with (b) a TTT diagram for alloy 718 derived from [8,38,39].

than 5 pixels. The equivalent grain diameter was calculated by taking

the area of the grain determined by the orientation map and assuming a
circular area through Eq. (1):
√̅̅̅
A
deq = 2 . (1)
π
TEM lamellae were prepared on an FEI Versa™ Nanolab™ 650 or an
Fig. 2. Schematic of the subsize tensile specimens with appro­ FEI Quanta™ Nanolab™ focused ion beam (FIB) system by using stan­
priate dimensions. dardized lift-out procedures. Each lamella was made with a thick win­
dow frame to minimize foil bending and twisting during sample thinning
below a thickness of ~300 nm. Low-energy Ga ion beams at 2 and 5 keV
were used to thin the ~150 nm thick lamella to a final thickness of

3
S. Taller and T. Austin Additive Manufacturing 60 (2022) 103280

around 80–100 nm, which effectively eliminated the TEM-visible FIB 1024 × 1024 pixels with a resolution of ~1.3 nm/pixel, a probe full-
damage induced at high-beam energy. For each condition, two lamellae width half-maximum of ~1.5 nm, and a beam current of around 3 nA.
were extracted from different grains comprising of at least 25 µm2 of Each scan had a duration of 1 h, with more than 35,000 counts/s and
electron transparent area to assess uniformity. dead times from 1% to 6%. The qualitative x-ray counts were converted
The nanoscale microstructure for each heat treatment of AM718 and to quantified weight percentages by using the Cliff-Lorimer [41] method
W718 was characterized by using the FEI Talos F200X scanning trans­ for calculation at each pixel.
mission electron microscope (STEM) instrument equipped with high Each corresponding collection of STEM images and EDS spectrum
counting rate electron dispersive spectroscopy (EDS). Before imaging images was characterized for nanoscale features. Precipitates were
and EDS spectra collection, the lamella was tilted to a low-order zone identified by using the composition expected for each phase [38] and
axis such as the (001), and the collection angles were optimized. Each overlaying the corresponding EDS maps for the elements of interest in
STEM image set consisting of Bright Field (BF), Dark Field (DF), and the phase. Phase identification was assisted using a trained dynamic
High Angle Annular Dark Field (HAADF) images was collected with a segmentation convolutional neural network [42]. Each feature was an­
region of interest size of 2048 × 2048 pixels with a resolution of notated by the neural network and designated as belonging to a class.
~0.7 nm/pixel. Additional images were collected as needed at a higher These annotations were used to produce an image in which the type of
resolution of 0.16 nm/pixel to confirm the precipitate structure and feature present at any given pixel was labeled by a class number. Each
composition. EDS-based spectrum images for precipitate measurement feature class was individually filtered out of the image and analyzed by
were taken over a broad area by using a region-of-interest size of using the particle analysis function in FIJI [43] to obtain parameters for

Fig. 3. Representative forescatter electron images and corresponding EBSD orientation maps for (a) W718, (b) AM718-ASB, (c) AM718-HT1, (d) AM718-HT2, and
(e) AM718-HM.

4
S. Taller and T. Austin Additive Manufacturing 60 (2022) 103280

the area, perimeter, minimum width, maximum width, and circularity

for each feature present. When features were found to overlap in the
two-dimensional image, the features were separated manually and
confirmed using hand counting procedures. An equivalent diameter for
each feature to use with the dispersed barrier hardening model was
calculated using Eq. (2). The mean and standard error of the mean for
the equivalent diameter were calculated for each identified feature of
the microstructure.
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
deq = major axis × minor axis. (2)

2.3. Uniaxial tensile testing

Subsize SS-J2 specimens were tested for common engineering

properties by using uniaxial tensile straining. Five specimens of each
heat treatment of the AM718 and W718 were strained at room tem­
perature, 300 ◦ C, 450 ◦ C, or 600 ◦ C. Tensile straining was conducted
according to ASTM E8/E8M for room temperature and ASTM E21–20 for
elevated temperatures on an MTS Systems uniaxial test frame. The
straining was maintained at a crosshead displacement of 0.15 mm/min
for a strain rate of 5 × 10− 4 s− 1 to avoid complications typical at high
strain rates [44]. Specimens were strained to failure at the test
temperature.

3. Results

3.1. Grain size determination

Representative forescatter SEM images and corresponding colorized

EBSD orientation maps are shown in Fig. 3. The wrought alloy 718
displayed the expected microstructure consisting of large grains with an
average equivalent grain diameter of 106 ± 21.8 µm gathered from
three areas containing 229 grains and with interior annealing twins. For
AM718, the as-built microstructure consisted of elongated grains that
showed a preferred orientation along the build direction, resulting in an
average equivalent grain diameter of 8.3 ± 0.22 µm from 10 areas
containing 3732 grains. The grain size distributions normalized by the
total number of grains observed for the AM718 in the as-built and heat- Fig. 4. Grain size distributions for AM718 normalized by the total number of
treated conditions are shown in Fig. 4. With the application of the ho­ grains captured within the EBSD map areas.
mogenization heat treatment and subsequent solution anneal, incom­
plete recrystallization occurred, leading to a mixture of large and small of γ′′ precipitates was observed in the vicinity of larger δ phase particles
grains in the heat-treated AM conditions and average equivalent grain (Fig. 6), which is consistent with previous observations from
diameters of 24.6. ± 2.5 µm gathered from three areas containing 342 AM-fabricated superalloy 718 [47] and as-cast wrought alloy 718 [46].
grains for AM718-HT1, 29.4. ± 2.8 µm gathered from three areas con­ The time-temperature-precipitation diagram suggests that at higher
taining 315 grains for AM718-HT2, and 37.9. ± 2.6 µm gathered from temperatures such as those present during AM, the formation of inter­
three areas containing 442 grains for AM718-HM. However, the skew­ dendritic γ′′ was favored over γ′′ interior to the grain [47]. Interspersed
ness of the distributions may make the median a more relevant param­ throughout the grain interiors was a mixture of fine γ′ and γ′′ precipitates
eter which results in median equivalent diameters of 3.1, 9.9, 12.5, and of a very high density (Fig. 7) with the frequency size distributions in
18.8 µm for AM718-ASB, AM718-HT1, AM718-HT2, and AM718-HM, Fig. 8. Two nonoverlapping areas were examined for precipitate for­
respectively, and 62.7 µm for W718. With each heat treatment, the mation closer to the part interior to attempt to capture representative
tail of the grain size distributions extended to larger sizes, but the sub­ microstructures and determine the local heterogeneity in the micro­
sequent aging steps did not significantly increase the grain size further. structure. A quantitative summary of the observed microstructure is
included in Table 3, including the number of precipitates observed, the
3.2. Precipitate characterization using STEM percentage that each precipitate phase contributes to the total precipi­
tation, the mean equivalent diameter, the standard error of the mean
The first characterization performed was on the ASB superalloy 718. equivalent diameter, and the number density for each phase. Between
An overview of the microstructure at the micron scale is shown in Fig. 5, the two areas examined, the size and density of γ′′ , Laves, and oxide
in which the print direction is out of the page. The grain structure precipitates were found to be nearly identical. However, both γ′ and δ
consisted of fine lath-like boundaries decorated with Nb-rich pre­ showed a large variation in density (~2 ×) between ASB-1 and ASB-2,
cipitates. Toward the part exterior, the Laves phase formed as a dendrite, with no carbides being observed in ASB-2. Although mechanical prop­
and the δ phase formed on boundaries nearer to the part’s interior. For erties could be obtained for comparison with this microstructure, it is
volumes of the part consistently above the δ solvus line at 1010 ◦ C [45, unlikely to yield correlative results because of the high level of
46], Laves phase formation is preferred as a metastable phase from so­ complexity and heterogeneity found in the ASB superalloy 718. There­
lidification at high cooling rates, whereas the δ phase forms latter in the fore, ASB superalloy 718 was excluded from the mechanical testing
fabrication process, likely from the longer time at elevated tempera­ campaign and is not included in subsequent sections.
tures, to form the stable phase. At a higher magnification, a high density

5
S. Taller and T. Austin Additive Manufacturing 60 (2022) 103280

Fig. 5. STEM images of ASB AM Superalloy 718 showing dendritic Laves phase, large δ precipitates, TiC, pores, oxide particles, and γ′ and γ′′ precipitates as a
function of distance from the part surface. Note the build direction vector is directed out of the page towards the viewer.

Fig. 6. STEM images of ASB AM Superalloy 718 with corresponding EDS maps for precipitate-forming elements showing large δ precipitates circled in yellow
(Ni3Nb), Laves particles circled in red(Nb-Mo-Si rich), TiC, and γ′′ precipitates circled in blue (Ni3Nb), and oxide particles circled in light blue. For interpretation of
the references to color in this figure, the reader is referred to the web version of this article.

The microstructure found in the heat-treated specimens was not as 3.3. Tensile test results up to 600 ◦ C
complex as the ASB AM718. Representative STEM HAADF images for
the wrought alloy 718 and the three heat treatments of AM718 are Representative engineering stress-strain curves for each heat treat­
shown in Fig. 9. The wrought alloy contained the expected microstruc­ ment of AM718 and the reference wrought alloy 718, labeled as W718,
ture that consists of a high density of fine γ′ and γ′′ precipitates, strained at each test temperature, are shown in Fig. 11, and the stress-
consistent with the ASTM-prescribed microstructure for precipitate- strain curves for each strained specimen are provided in the supple­
hardened superalloys [33]. Following the homogenization treatment mentary materials. Engineering stress-strain curves for each strained
to produce AM718-HM, the microstructure consisted of a low density of specimen from each condition are displayed in the appendix as
remaining oxide and carbide particles and dissolved all Laves, δ, γ′ , or γ′′ Figure A.14 through Figure A.17 for transparency. These curves were
phases present in the ASB AM718. After aging, the AM718-HT1 and used to calculate the common engineering properties of elastic modulus,
AM718-HT2 treatments appeared qualitatively similar and contained a 0.2% offset yield strength, uniform elongation, ultimate tensile stress,
high density of fine γ′ , γ′′ , and δ precipitates, with AM718-HT2 having total elongation, and fracture strength. These properties are included in
greater precipitate densities overall. However, AM718-HT1 contained a Table 5. Although acoustic methods are preferred for determining the
higher fraction of δ phase, as expected from the heat treatment. The elastic modulus, the elastic modulus here was calculated by fitting a line
quantitative analysis of γ′ , γ′′ , and δ particles is provided in Table 4, and to the elastic portion of each tensile curve, minimizing the least squares
the size distribution of γ′ , γ′′ , and δ precipitates from each heat treatment error, and extracting the slope of the line through a simple Hooke’s Law
is shown in Fig. 10. approximation. Fig. 12.

6
S. Taller and T. Austin Additive Manufacturing 60 (2022) 103280

Fig. 7. Higher magnification STEM images of ASB AM Superalloy 718 with corresponding EDS maps for precipitate forming elements showing γ′ (Ni3[Ti,Al]) circled
in white and γ′′ (Ni3Nb) precipitates circled in blue. For interpretation of the references to color in this figure, the reader is referred to the web version of this article.

Fig. 8. Precipitate size distributions for γ′ , γ′′ , and δ phases in ASB AM Superalloy 718.

Table 3
Summary of characterization results for precipitate phases observed in ASB AM718.
Alloy Feature Number of features Percent of total precipitation Mean eq. diameter Standard error of mean Number density
designation observed (%) (nm) (nm) (m¡3)

AM718-ASB-1 γ’ 571 16.08 8.34 0.11 2.68 ± 0.3 × 1021

γ’’ 2484 69.94 9.47 0.14 1.17 ± 0.1 × 1022
δ 467 13.15 9.89 0.55 2.19 ± 0.2 × 1021
Laves 7 0.20 77.25 9.9 3.29 ± 0.6 × 1019
Carbides 15 0.42 27.52 2.7 7.04 ± 0.9 × 1019
Pores 10 – 12.49 1.3 4.70 ± 0.8 × 1019
Oxide 7 0.20 25.65 6.2 3.29 ± 0.8 × 1019
AM718-ASB-2 γ’ 247 8.27 10.59 0.21 1.16 ± 0.2 × 1021
γ’’ 2570 86.01 10.05 0.09 1.21 ± 0.1 × 1022
δ 156 5.22 14.34 1.9 7.33 ± 0.9 × 1020
Laves 9 0.30 51.56 14.9 4.23 ± 0.4 × 1019
Carbides N.O. – – – –
Pores 26 – 9.24 0.44 1.27 ± 0.3 × 1020
Oxide 7 0.20 14.3 6.7 2.82 ± 0.6 × 1019

The error in number densities is the minimum of error from counting statistics [48] or from the TEM foil thickness measurement. N.O. indicates that the feature was not
observed.

7
S. Taller and T. Austin Additive Manufacturing 60 (2022) 103280

Fig. 9. STEM HAADF images of wrought alloy 718, AM718-HT1, AM718-HT2, and AM718-HM showing the mixture of γ′ , γ′′ , and δ as bright features and oxides as
dark features.

4. Discussion recrystallization regime from 900◦ to 1066◦ C [50]. Likely as a result, the
yield strength and ultimate tensile strength of AM718-HT1 and
4.1. Comparison of wrought and AM 718 AM718-HT2 (Table 2) increased relative to W718 and were well within
ASTM specifications of superalloy 718. The AM718-HT2 condition dis­
The wrought superalloy 718 displayed the expected behavior for a played a larger uniform elongation after yield compared with the
precipitate-hardened alloy consisting of a large yield stress and a high AM718-HT1 condition. Comparison of the microstructures of these two
ultimate tensile stress. However, both the yield and ultimate tensile conditions as presented in Table 4, both exhibited a high density of γ′′
stress were 15–20% lower than expected for the ASTM heat treatment precipitates on the order of 1022 m− 3, with AM718-HT-2 having a larger
[1,8], and both had an increased uniform elongation and fracture strain density. However, AM718-HT1 had a higher fraction of the δ-phase,
compared with results presented in the literature [8,46,49]. The which likely contributed to the decrease in uniform elongation. The
increased fracture strain is consistent with the expectation of subsize agreement in the uniform elongation between AM718-HT-1 and W718
specimens and may result from the decrease in scale factor [35]. A lends support to the detrimental effects of δ-phase precipitates observed
fraction of the metastable γ′′ transformed into intragranular δ-phase in W718.
precipitates as predicted by the time-temperature-transformation dia­ When the precipitates were removed through homogenization and
gram [45], and over-aging as discussed in Appendix A.2. Although δ solution annealing (AM718-HM), the yield stress and ultimate tensile
precipitates share the composition of Ni3Nb with γ′′ , the strength decreased with a significantly greater uniform elongation and
lattice-precipitate interface is incoherent for the δ-phase, thus reducing fracture strain, which is consistent with solution-annealed alloy 718
its contribution to strength [8]. The high density of δ precipitates [51]. However, the elastic modulus was significantly lower than the
(Table 4) in the wrought superalloy 718 likely reduced the strength of 190–220 GPa range expected for polycrystalline pure Ni [52], stemming
the alloy compared to the expected ASTM-designated strength. from the decrease in yield stress. In the literature, a decreased elastic
Prior to heat treatments, there was a clear preferred grain orientation modulus was found when annealing at high temperatures for 1 h to
in the as-built condition relative to the build direction. This likely would achieve values of 165–172 GPa [53–58]. Increasing the annealing
result in anisotropic properties. However, the first step of the heat temperature reduced the elastic modulus linearly and correlated with an
treatment was a high temperature solution anneal meant to remove all increase in the grain size [55]. Compared with these previous works, the
previous precipitation. This resulted in significant grain growth, as ex­ homogenization treatment in this work was performed for a much
pected for annealing above 982 ◦ C [50], and a transition from lath-like longer time and at higher temperatures. In this work, the recrystallized,
elongated grains to more equiaxed grains, as expected for the homogenized AM718-HM still had a reduced grain size compared to that

8
S. Taller and T. Austin Additive Manufacturing 60 (2022) 103280

Table 4
Summary of characterization results for precipitate phases observed in heat-treated AM718 and W718.
Alloy Feature Number of features Percent of total precipitation (%) Mean eq. diameter Standard error of mean Number density
designation observed (nm) (nm) (m¡3)

AM718-HT1 γ’ 407 7.39 8.16 0.09 1.91 ± 0.2 × 1021

γ’’ 4615 83.74 8.93 0.04 2.17 ± 0.2 × 1022
δ 487 8.84 8.17 0.08 2.29 ± 0.2 × 1021
Laves N.O. – – – –
Carbides 2 0.04 35.2 14.4 9.39 ± 1.2 × 1018
Pores N.O. – – – –
Oxide N.O. – – – –
AM718-HT2 γ’ 846 9.93 6.80 0.073 3.97 ± 0.4 × 1021
γ’’ 7208 84.59 7.72 0.024 3.39 ± 0.3 × 1022
δ 466 5.47 7.06 0.062 2.19 ± 0.2 × 1021
Laves N.O. – – – –
Carbides 1 0.01 63.59 N/A 4.70 ± 2.4 × 1018
Pores N.O. – – – –
Oxide N.O. – – – –
AM718-HM-1 γ’ N.O. – – – –
γ’’ N.O. – – – –
δ N.O. – – – –
Laves N.O. – – – –
Carbides 3 75.00 45.79 13.7 1.41 ± 0.5 × 1019
Pores N.O. – – – –
Oxide 1 25.00 135.89 N/A 4.70 ± 2.4 × 1018
AM718-HM-2 γ’ N.O. – – – –
γ’’ N.O. – – – –
δ N.O. – – – –
Laves N.O. – – – –
Carbides 4 66.67 81.83 49.5 1.88 ± 0.6 × 1019
Pores 1 – 11.10 N/A 4.70 ± 2.4 × 1018
Oxide 2 33.33 47.33 40.9 9.39 ± 1.8 × 1018
W718-Z1653 γ’ 528 4.38 5.6 0.065 2.38
± 0.25 × 1021
γ’’ 7735 64.28 5.9 0.015 3.48
± 0.38 × 1022
δ 3770 31.33 5.9 0.019 1.69 ± 1.8 × 1022
Laves N.O. – – – –
Carbides N.O. – – – –
Pores N.O. – – – –
Oxide N.O. – – – –

The error in number densities is the minimum of error from counting statistics [48] or from the TEM foil thickness measurement. N.O. indicates that the feature was not
observed. N/A indicates there was not enough of the feature to calculate this value.

of the wrought alloy, but without precipitates, making the implied constant, and d is the grain diameter; and (3) the dislocation friction
relationship between grain size and elasticity unclear. Additionally, at stress, also known as the Peierls-Nabarro stress, τP− N [63]. The impact of
600 ◦ C, the elastic modulus appeared to increase for AM718-HM. While defects and precipitates is generally described by a dispersed barrier
√̅̅̅̅̅̅̅̅̅
precipitation is not expected based on the TTT diagram (Fig. 1b), solute hardening model, as shown in Eq. (3): αi Mμb Ni di , where α represents
clustering is likely occurring. Many precipitating phases proceed along a the strength of the barrier, M is the Taylor factor, μ is the shear modulus,
mechanism consisting of clustering of the constituent elements and b is the Burgers vector, N is the density of the precipitates, and d is their
accumulation of solutes until a critical concentration then transition to a size described as an equivalent diameter:
distinct phase [59–61]. Therefore, the elastic modulus increase may ∑δ,γ′ ,γ′′ ,MX,oxide
k √̅̅̅̅̅̅̅̅
stem from dislocation pinning on solute clusters that have not transi­ σ Y = σ SS + τP− N + √y̅̅̅ + αi M μb Ni di . (3)
i
tioned into γ′ or γ′′ precipitates. These clusters were not visible in the d
STEM EDS datasets and thus may reside below the resolution of the A linear superposition was utilized because the contributions from
technique estimated at approximately 2 nm for the parameters pre­ each component are not expected to be similar in strength. Because
sented in the Methodology. An approach is needed to deconvolute the W718 and each heat treatment of AM718 all exhibited similar micro­
contributions of grain size, precipitates, and other features contributing structure features, all yield stress measurements were made into a linear
to the yield strength. system of equations to solve for the contributions of each component of
the microstructure to the strength. Because of the low density of carbides
4.2. Microstructure-property relationships and oxides, these features were assumed to contribute a negligible
amount of strength compared to the other factors and were excluded
There is an inherent link between a material’s microstructure and from the analysis. The solid solution strengthening and the dislocation
macroscopic mechanical properties. In the broadest sense, the yield friction stress were grouped together and assumed to act together line­
strength of a material is the sum of the inherent strength from the crystal arly. The median grain diameter was used instead of the average to
lattice, σY,0 , with additional strength from the effects of defects and reduce the influence of outliers. Rather than assume an average density
secondary phases, σD , interpreted through a classical Orowan disloca­ and diameter, the precipitate size distributions shown in Fig. 10 were
tion bowing process [62]. The inherent strengthening shown in Eq. (2) integrated to be included in the dispersed barrier calculations. M is 3.06
stems from (1) the solid solutioning strength from substitutional ele­ [64], whereas the Burgers vector was assumed to be equal to the lattice
ments in the face-centered cubic (FCC) Ni lattice, σ SS ; (2) the grain parameter for FCC nickel. This results in a linear system of equations
k
with 5 unknown variables (σSS + τP− N , ky, αγ’, αγ’’ and αδ) for each of the
boundaries through the Hall-Petch relationship, √y̅̅,
d
where ky is a

9
S. Taller and T. Austin Additive Manufacturing 60 (2022) 103280

Fig. 10. Precipitate size distributions for γ′ , γ′′ , and δ phases in (a,b) heat-treated AM alloy 718 and (c) W718.

16 yield stress measurements given in Table 5. Assuming that the microstructural features to primarily stress-driven interstitial disloca­
inherent strengthening and grain boundary contributions are the same tion motion. The results in Table 6 suggest that the main contributions to
for each measurement of the same heat treatment, the system reduces yield strength at room temperature are grain boundaries, γ′′ , and δ
the system to 16 equations with 28 unknowns—an underdetermined precipitates, with the γ′′ and δ exhibiting the same obstacle strength.
system. In many similar analyses, the barrier strength is assumed to be However, when the tensile temperature is increased above the onset of
independent of temperature for each feature, which would reduce the vacancy mobility, the strength coefficient for γ′′ precipitates remain
system to 16 equations with 7 unknowns and would become over­ about the same as at room temperature but decreases for δ phase.
determined. An overdetermined system almost always yields inconsis­ Because 300 ◦ C is still a relatively low temperature and is unlikely to
tent solutions when constructed with randomly determined starting have a significant thermal concentration of vacancies, this reduction in
points for each variable. Although an underdetermined system cannot the strength of δ can be attributed to the increase in interstitial motion
be solved for a unique solution, the Moore-Penrose pseudoinverse of the between room temperature and 300 ◦ C. With increasing temperature up
linear coefficient matrix of the unknowns can be used to construct a to 600 ◦ C and corresponding increase in thermal contributions to
linear solution that minimizes the least-squares approximation. The interstitial motion and vacancy concentration, the strengthening coef­
coefficients found from this solution are displayed in Table 6. An ficient for γ′′ decreases, and δ increases until these precipitate phases
alternative approach using theoretical strength factors from Tan and become equal strength. Both phases are the same chemical composition
Busby [65] is presented in the Appendix for reference and suggests of Ni3Nb, but the disk-shaped γ′′ precipitates are coherent with the
similar values for α as this discussion. nickel lattice on {100} planes, maintaining the c-axis perpendicular to
Lucas [66] summarizes the reported α values of different types of the disks, whereas the δ phase is semi-coherent [49], possibly driving the
defects and categorizes them into three groups: (1) weak barriers such as differences in the observed strengthening factors when thermal motion
small bubbles/loops and dislocations with α < 0.25, (2) intermediate is considered. Coherent precipitates produce larger long-range strain
barriers such as Frank loops and small precipitates with fields compared to semi-coherent precipitates, and thus they also in­
0.33 < α < 0.45, and (3) strong barriers such as voids and large pre­ crease the combined mechanical and thermal work required to over­
cipitates with α~1 obstacles. According to this interpretation, the pre­ come a barrier. If the required work is sufficiently small, then thermal
cipitates examined are all weak barriers with a variable strength vibrations of the lattice may assist dislocations to pass through a weak
coefficient based on temperature. All materials exhibit several major strain field and overcome the precipitate [68]. Atomistic simulations of
temperature regimes separated by the onset of migration of point de­ point defect and dislocation interactions with these precipitates would
fects. In pure nickel, vacancy defects are immobile below about provide strong additional insight into the role of precipitate coherency
77–100 ◦ C, whereas small interstitial clusters and interstitial-solute on strength at high temperatures.
clusters may migrate [67]. Thus, the tensile tests at room temperature With these coefficients, the relative contribution of each term to the
can be considered as testing the resistance of each class of total yield strength can be estimated and is displayed as a percentage in

10
S. Taller and T. Austin Additive Manufacturing 60 (2022) 103280

Fig. 11. Representative engineering stress-strain curves for heat–treated AM alloy 718 and wrought alloy 718 at (a) room temperature, (b) 300 ◦ C, (c) 450 ◦ C, and
(d) 600 ◦ C.

Table 5
Engineering properties derived from room temperature tensile tests of wrought alloy 718 and heat-treated AM alloy 718.
Designation Test temp. Elastic modulus 0.2% offset yield Uniform elongation Ultimate tensile Total Elongation Fracture Strength
(◦ C) (GPa) (MPa) (%) strength (MPa) (%) (MPa)

AM718-HT1 22 244 ± 19.9 1056 ± 13.6 14.6 ± 1.9 1187 ± 17.7 26.1 ± 1.7 1007 ± 45.0
300 223 ± 11.5 843 ± 68.0 9.99 ± 1.9 1011 ± 17.9 19.6 ± 2.1 930 ± 48.3
450 230.6 ± 21.3 836 ± 53.5 11.7 ± 2.2 959 ± 23.1 23.2 ± 0.8 834 ± 118
600 227 ± 22.1 810 ± 17.0 7.1 ± 1.2 896 ± 9.22 16.3 ± 1.2 789 ± 15.1
AM718-HT2 22 234 ± 11.8 978 ± 33.2 18.5 ± 2.7 1144 ± 18.6 28.6 ± 3.3 1008 ± 24.3
300 240 ± 7.6 852 ± 22.7 15.9 ± 3.3 987 ± 12.7 24.6 ± 3.4 888 ± 22.9
450 226.1 ± 11.9 825 ± 45.5 11.0 ± 1.9 945 ± 12.9 22.3 ± 1.5 806 ± 102
600 230 ± 8.2 789 ± 33.0 7.9 ± 2.2 876 ± 23.4 17.1 ± 1.5 770 ± 26.1
AM718-HM 22 131 ± 6.1 338 ± 14.8 54.1 ± 1.2 717 ± 20.3 67.5 ± 1.5 552 ± 26.1
300 119 ± 44 303 ± 14.7 58.3 ± 5.8 610 ± 12.4 63.8 ± 4.4 463 ± 36.8
450 100.3 ± 15.4 294 ± 48.0 47.6 ± 5.1 563 ± 9.22 66.7 ± 3.7 478 ± 10.8
600 148 ± 27 250 ± 9.39 53.6 ± 4.6 517 ± 16.4 62.5 ± 3.5 423 ± 51.4
W718- 22 215 ± 23.0 759 ± 114 14.9 ± 1.8 928 ± 37.2 24.4 ± 1.3 734 ± 77.7
Z1653 300 217 ± 3.5 711 ± 28.6 13.8 ± 4.8 802 ± 25.3 22.6 ± 6.0 585 ± 97.8
450 190.5 ± 18.5 666 ± 14.0 8.4 ± 4.1 719 ± 30.8 17.5 ± 3.8 574 ± 21.9
600 179 ± 18.8 653 ± 13.5 6.0 ± 1.9 697 ± 21.8 14.3 ± 2.3 521 ± 40.7

Values reported are the averages of all tested specimens and their associated standard deviations.

Fig. 13. Based on this analysis, several trends are visible, and general Therefore, motion around solutes is comparatively trivial and results in
processes can be proposed. At room temperature, the grain boundaries an apparent lack of strengthening from these features. As temperature
appear to be responsible for most of the strength from the lattice increases to 300 ◦ C, interstitials have more thermal motion, and solutes
(AM718-HM), whereas the precipitates contribute more than half of the in the matrix provide pinning sites and increase the relative contribution
yield strength (AM718-HT1, AM718-HT2, and W718). However, this to the yield strength. However, as temperature increases even further, it
result is counterintuitive because some contribution from solutes in the becomes easier to bypass the solutes, and the grain boundaries once
lattice would be expected. Assuming primarily stress-driven dislocation again become a relatively stronger barrier to overcome. The main
motion, these results suggest that the stresses necessary to eject dislo­ assumption in the Hall-Petch slope (ky ) is that it is proportional to the
cations from grain boundaries are significantly higher than those work required to eject dislocations from grain boundaries [69], and thus
necessary to overcome small interstitial solutes or solutes in solution. the trends in Table 6 and Fig. 13 suggest a strong reliance on interstitial

11
S. Taller and T. Austin Additive Manufacturing 60 (2022) 103280

Fig. 12. Engineering properties for heat-treated AM718 and W718 as a function of testing temperature: (a) elastic modulus, (b) yield stress, (c) uniform elongation,
(d) ultimate tensile strength, (e) total elongation, and (f) fracture strength.

motion, similar to the discussion surrounding δ phase precipitates in the

Table 6
preceding paragraph. At all temperatures, the relative contribution to
Coefficients determined to correlate microstructure to yield strength with linear
the yield strength from γ′′ remained consistent between 48% and 57% of
least squares error.
the total strength, indicating the utility of this phase to the
Temperature T/ Solid Hall-Petch α for γ′′ α for δ high-temperature strength of superalloy 718. One drawback from this
(◦ C) Tm solution coefficient, (unitless) (unitless)
+ friction ky (MPa×m1/
analysis is that the contributions of other microstructural features such
stress 2
) as carbides, oxides, and γ′ precipitates could not be assessed because of
modifier, C1 their lower density and apparent strength.
(×1 MPa) While many researchers investigated the relationships between post-
22 0.17 0 (see text) 1716,711 0.129 0.129 build heat treatments, mechanical response, and microstructure of su­
300 0.33 236.24 484,541 0.131 0.057 peralloy 718, the results tend to focus more on the room temperature
450 0.42 134.46 882,908 0.118 0.078 properties and semi-empirical relationships. However, superalloy 718 is
600 0.50 56.35 1000,045 0.112 0.111
used in many applications up to its high creep-rupture strength, which
limits its use at high temperatures near 650–700 ◦ C. This is believed to

12
S. Taller and T. Austin Additive Manufacturing 60 (2022) 103280

Fig. 13. Fractional contribution to the yield strength from solid solution strengthening and friction strength, grain boundaries through the Hall-Petch relationship, γ′′
precipitates, and δ precipitates for each heat treatment of superalloy 718 examined in this work.

stem from the coarsening of the γ′′ precipitates [8] expected from the 5. Conclusions
TTT diagram (Fig. 1b). Because of the wide temperature range in the
derivation and relative simplicity of the dispersed barrier model pro­ Microstructure characterization and baseline mechanical testing
duced in this work, it can easily integrate into modeling and simulation were performed on additively manufactured superalloy 718 produced
workflows to estimate the yield strength of AM superalloy 718 compo­ via PBF-LB. The key findings and conclusions derived from this work are
nents after fabrication. Such projects, for example the ExaAM initiative summarized as follows.
[70], are expected to transform AM through high performance The microstructure of as-built superalloy 718 contained many of the
computing (exascale) simulation of the complex interplay between the phases predicted from the isothermal phase diagram with strong vari­
physical phenomena (e.g., heat transfer, melting, solidification, fluid ation, depending on the location in the part at nano- to micro-scales. The
flow) involved in the additive process to simulate the microstructure and formation of Laves and on-boundary δ phase suggests that the me­
resulting properties. Because the precipitation in superalloy 718 is chanical properties would result in poor performance of the as-built
strongly sensitive to the thermal history, having a component, whereas the lath-like structure suggests that the mechani­
microstructure-strength relationship dependent on the precipitate size, cal properties would strongly depend on orientation. Purely based on the
precipitate density, and morphology will significantly improve the as-built microstructure, superalloy 718 is not recommended for struc­
quality of efforts to simulate the properties of traditional and AM tural use without post-build heat treatments using these processing
components. parameters.
Three heat treatments consisting of two step homogenization, solu­
tion treatment, and aging were employed to control the precipitate
morphology in additively manufactured superalloy 718 after

13
S. Taller and T. Austin Additive Manufacturing 60 (2022) 103280

fabrication. The yield strength and ultimate tensile strength of the heat References
treated additively manufactured superalloy 718 increased relative to the
wrought 718 and were well within ASTM specifications for superalloy [1] T. Yonezawa, Nickel-based alloys, in: Compr. Nucl. Mater., Elsevier, 2020,
pp. 319–354, https://doi.org/10.1016/B978-0-12-803581-8.00676-7.
718. Both HT-1 and HT-2 treatments resulted in a high density of γ′′ [2] M.J. Donachie, S.J. Donachie, Superalloys: A Technical Guide. 2nd Editio, 2nd
precipitates on the order of 1022 m− 3, with AM718-HT-2 having a larger edition.,, ASM International,, 2002.
density and a lower fraction of intragranular δ-phase. Interpreting the [3] M. Griffiths, Ni-based alloys for reactor internals and steam generator applications,
in: Struct. Alloy. Nucl. Energy Appl., Elsevier Inc, 2019, pp. 349–409, https://doi.
yield strength through an Orowan dislocation mechanism determined org/10.1016/B978-0-12-397046-6.00009-5.
that all precipitates analyzed in this work proved to be weak barriers to [4] E. Paccou, M. Mokhtari, C. Keller, J. Nguejio, W. Lefebvre, X. Sauvage, S. Boileau,
dislocation motion, but they provided strength to the alloy through their P. Babillot, P. Bernard, E. Bauster, Investigations of powder reusing on
microstructure and mechanical properties of Inconel 718 obtained by additive
consistent high density. At temperatures ranging from 22◦ to 600◦ C, the manufacturing, Mater. Sci. Eng. A. 828 (2021), 142113, https://doi.org/10.1016/j.
relative contribution to the yield strength from γ′′ remained consistent msea.2021.142113.
between 48% and 57% of the total strength, indicating the utility of this [5] A. Hilaire, E. Andrieu, X. Wu, High-temperature mechanical properties of alloy 718
produced by laser powder bed fusion with different processing parameters, Addit.
phase to the high-temperature strength of superalloy 718. The strength
Manuf. 26 (2019) 147–160, https://doi.org/10.1016/j.addma.2019.01.012.
factors for γ′′ and δ precipitates were found to trend inversely with [6] E.M. Fayed, M. Saadati, D. Shahriari, V. Brailovski, M. Jahazi, M. Medraj, Effect of
tensile test temperature and may be attributable to the differences in homogenization and solution treatments time on the elevated-temperature
precipitate coherency. From the homogenized condition, it is suggested mechanical behavior of Inconel 718 fabricated by laser powder bed fusion, Sci.
Rep. 11 (2021) 2020, https://doi.org/10.1038/s41598-021-81618-5.
that the contributions in strength from solutes and grain boundaries are [7] R. Firoz, S.K. Basantia, N. Khutia, H.N. Bar, S. Sivaprasad, G.V.S. Murthy, Effect of
strongly temperature dependent. Based on the results of this study, a microstructural constituents on mechanical properties and fracture toughness of
high-temperature homogenization followed by a traditional ASTM Inconel 718 with anomalous deformation behavior at 650 ◦ C, J. Alloy. Compd. 845
(2020), 156276, https://doi.org/10.1016/j.jallcom.2020.156276.
aging, designated HT-2 in this work, is recommended to recover the [8] D.D. Keiser, H.L. Brown, Review of the physical metallurgy of Alloy 718, U. S. Dep.
properties and microstructure from additively manufactured superalloy Energy (1976), https://doi.org/10.2172/4016087.
718 necessary for high-temperature strength. Because of the wide tem­ [9] K. Moussaoui, W. Rubio, M. Mousseigne, T. Sultan, F. Rezai, Effects of Selective
Laser Melting additive manufacturing parameters of Inconel 718 on porosity,
perature range in the derivation and relative simplicity of the dispersed microstructure and mechanical properties, Mater. Sci. Eng. A 735 (2018) 182–190,
barrier model produced in this work, it can easily be integrated into https://doi.org/10.1016/j.msea.2018.08.037.
modeling and simulation workflows to estimate the yield strength of [10] Y. Gao, D. Zhang, M. Cao, R. Chen, Z. Feng, R. Poprawe, J.H. Schleifenbaum,
S. Ziegler, Effect of δ phase on high temperature mechanical performances of
traditional and AM components. Inconel 718 fabricated with SLM process, Mater. Sci. Eng. A. 767 (2019), 138327,
https://doi.org/10.1016/j.msea.2019.138327.
CRediT authorship contribution statement [11] L. Huang, Y. Cao, J. Zhang, X. Gao, G. Li, Y. Wang, Effect of heat treatment on the
microstructure evolution and mechanical behaviour of a selective laser melted
Inconel 718 alloy, J. Alloy. Compd. 865 (2021), 158613, https://doi.org/10.1016/
Stephen Taller: Conceptualization, Methodology, Software, Formal j.jallcom.2021.158613.
analysis, Investigation, Visualization, Writing − original draft, Project [12] J.R. Zhao, F.Y. Hung, T.S. Lui, Microstructure and tensile fracture behavior of
administration. Ty Austin: Software, Formal analysis, Investigation, three-stage heat treated inconel 718 alloy produced via laser powder bed fusion
process, J. Mater. Res. Technol. 9 (2020) 3357–3367, https://doi.org/10.1016/j.
Writing − review & editing. jmrt.2020.01.030.
[13] M.M. Kirka, F. Medina, R. Dehoff, A. Okello, Mechanical behavior of post-
Declaration of Competing Interest processed Inconel 718 manufactured through the electron beam melting process,
Mater. Sci. Eng. A. 680 (2017) 338–346, https://doi.org/10.1016/j.
msea.2016.10.069.
The authors declare that they have no known competing financial [14] E. Hosseini, V.A. Popovich, A review of mechanical properties of additively
interests or personal relationships that could have appeared to influence manufactured Inconel 718, Addit. Manuf. 30 (2019), 100877, https://doi.org/
10.1016/j.addma.2019.100877.
the work reported in this paper. [15] F. Liu, F. Lyu, F. Liu, X. Lin, C. Huang, Laves phase control of inconel 718
superalloy fabricated by laser direct energy deposition via $δ$ aging and solution
Data availability treatment, J. Mater. Res. Technol. 9 (2020) 9753–9765, https://doi.org/10.1016/j.
jmrt.2020.06.061.
[16] S. Sui, C. Zhong, J. Chen, A. Gasser, W. Huang, J.H. Schleifenbaum, Influence of
Data will be made available on request. solution heat treatment on microstructure and tensile properties of Inconel 718
formed by high-deposition-rate laser metal deposition, J. Alloy. Compd. 740
(2018) 389–399, https://doi.org/10.1016/j.jallcom.2017.11.004.
Acknowledgments [17] X. Li, J.J. Shi, C.H. Wang, G.H. Cao, A.M. Russell, Z.J. Zhou, C.P. Li, G.F. Chen,
Effect of heat treatment on microstructure evolution of Inconel 718 alloy fabricated
This research was sponsored in part by the Laboratory Directed by selective laser melting, J. Alloy. Compd. 764 (2018) 639–649, https://doi.org/
10.1016/j.jallcom.2018.06.112.
Research and Development Program of Oak Ridge National Laboratory,
[18] X. Yu, X. Lin, F. Liu, L. Wang, Y. Tang, J. Li, S. Zhang, W. Huang, Influence of post-
managed by UT-Battelle, LLC, for the US Department of Energy under heat-treatment on the microstructure and fracture toughness properties of Inconel
contract number DE-AC05-00OR22725. This research was supported by 718 fabricated with laser directed energy deposition additive manufacturing,
the Transformational Challenge Reactor/Advanced Materials and Mater. Sci. Eng. A. 798 (2020), 140092, https://doi.org/10.1016/j.
msea.2020.140092.
Manufacturing Technologies program supported by the US Department [19] C.E. Seow, H.E. Coules, G. Wu, R.H.U. Khan, X. Xu, S. Williams, Wire + Arc
of Energy, Office of Nuclear Energy. T. Austin’s contributions are based Additively Manufactured Inconel 718: Effect of post-deposition heat treatments on
on work supported under a US Department of Energy Office of Nuclear microstructure and tensile properties, Mater. Des. 183 (2019), 108157, https://doi.
org/10.1016/j.matdes.2019.108157.
Energy Integrated University Program Graduate Fellowship. SEM and [20] S. Sui, J. Chen, Z. Li, H. Li, X. Zhao, H. Tan, Investigation of dissolution behavior of
STEM characterization were performed in the Low Activation Materials laves phase in inconel 718 fabricated by laser directed energy deposition, Addit.
Development and Analysis laboratory at ORNL. Additional acknowl­ Manuf. 32 (2020), 101055, https://doi.org/10.1016/j.addma.2020.101055.
[21] S. Sui, H. Tan, J. Chen, C. Zhong, Z. Li, W. Fan, A. Gasser, W. Huang, The influence
edgments are extended for Keith Carver for operating the Concept X- of Laves phases on the room temperature tensile properties of Inconel 718
Line 2000R, Tom Geer and Gavin Mattingly for sample preparation, Eric fabricated by powder feeding laser additive manufacturing, Acta Mater. 164
Manneschmidt for performing the tensile tests, and T. S. Byun and (2019) 413–427, https://doi.org/10.1016/j.actamat.2018.10.032.
[22] M.J. Sohrabi, H. Mirzadeh, M. Rafiei, Solidification behavior and Laves phase
Holden Hyer for thoughtful feedback. dissolution during homogenization heat treatment of Inconel 718 superalloy,
Vacuum 154 (2018) 235–243, https://doi.org/10.1016/j.vacuum.2018.05.019.
Appendix A. Supporting information [23] B.B. Babamiri, J. Indeck, G. Demeneghi, J. Cuadra, K. Hazeli, Quantification of
porosity and microstructure and their effect on quasi-static and dynamic behavior
of additively manufactured Inconel 718, Addit. Manuf. 34 (2020), https://doi.org/
Supplementary data associated with this article can be found in the 10.1016/j.addma.2020.101380.
online version at doi:10.1016/j.addma.2022.103280.

14
S. Taller and T. Austin Additive Manufacturing 60 (2022) 103280

[24] G.M. Volpato, U. Tetzlaff, M.C. Fredel, A comprehensive literature review on laser [47] A. Segerstark, J. Andersson, L.E. Svensson, O. Ojo, Microstructural characterization
powder bed fusion of Inconel superalloys, Addit. Manuf. (2022), 102871, https:// of laser metal powder deposited Alloy 718, Mater. Charact. 142 (2018) 550–559,
doi.org/10.1016/j.addma.2022.102871. https://doi.org/10.1016/j.matchar.2018.06.020.
[25] J.M. Poole, K.R. Stultz, J.M. Manning, The Effect of INGOT Homogenization [48] G.F. Knoll, Radiation Detection and Measurement. 3rd Editio, John Wiley & Sons,
Practice on the Properties of Wrought Alloy 718 and Structure, in: Superalloys 718 Inc,, 2000.
Metall. Appl., TMS, 1989, pp. 219–228, https://doi.org/10.7449/1989/ [49] S.G.K. Manikandan, D. Sivakumar, M. Kamaraj, Physical metallurgy of alloy 718,
Superalloys_1989_219_228. Weld. Inconel 718 Superalloy (2019) 1–19, https://doi.org/10.1016/b978-0-12-
[26] H.L. Eiselstein, Metallurgy of a Columbium-Hardened Nickel-Chromium-Iron 818182-9.00001-3.
Alloy, in: Adv. Technol. Stainl. Steels Relat. Alloy., ASTM International, 100 Barr [50] D.D. Keiser, H.L. Brown, Review of the physical metallurgy of Alloy 718, U. S. Dep.
Harbor Drive, PO Box C700, West Conshohocken, PA 19428–2959, 1965: pp. Energy (1976), https://doi.org/10.2172/4016087.
62–62–18. https://doi.org/10.1520/STP43733S. [51] T.S. Byun, K. Farrell, Tensile properties of Inconel 718 after low temperature
[27] M.C. Chaturvedi, Y. Han, Strengthening mechanisms in Inconel 718 superalloy, neutron irradiation, J. Nucl. Mater. 318 (2003) 292–299, https://doi.org/10.1016/
Met. Sci. 17 (1983) 145–149, https://doi.org/10.1179/030634583790421032. S0022-3115(03)00006-0.
[28] M. Sundararaman, P. Mukhopadhyay, S. Banerjee, Precipitation of the δ-Ni3Nb [52] H.M. Ledbetter, R.P. Reed, Elastic Properties of Metals and Alloys, I. Iron, Nickel,
phase in two nickel base superalloys, Metall. Trans. A. 19 (1988) 453–465, https:// and Iron-Nickel Alloys, J. Phys. Chem. Ref. Data. 2 (1973) 531–618, https://doi.
doi.org/10.1007/BF02649259. org/10.1063/1.3253127.
[29] M. Sundararaman, P. Mukhopadhyay, S. Banerjee, Some aspects of the [53] S. Umekawa, On Young’s Moduli of Nickel-Copper and Nickel-Cobalt Alloys,
precipitation of metastable intermetallic phases in INCONEL 718, Metall. Trans. A. J. Jpn. Inst. Met 18 (1954) 387–390, https://doi.org/10.2320/jinstmet1952.18.7_
23 (1992) 2015–2028, https://doi.org/10.1007/BF02647549. 387.
[30] M. Wang, M. Song, G.S. Was, J.L. Nelson, The roles of thermal mechanical [54] Y. Shirakawa, K. Numakura, Elasticity Constants of Single Crystals of Nickel-
treatment and δ phase in the stress corrosion cracking of alloy 718 in primary Copper Alloys, Sci. Rep. Res. Inst., Tohoku Univ. Ser. A, Phys., Chem. Metall. 10
water, Corros. Sci. 160 (2019), 108168, https://doi.org/10.1016/j. (1958) 51–57. 〈http://hdl.handle.net/10097/26859〉.
corsci.2019.108168. [55] Y. Shirakawa, K. Numakura, On Young’s Modulus and Grain Size in Nickel-Copper
[31] S. Mahadevan, S. Nalawade, J.B. Singh, A. Verma, B. Paul, K. Ramaswamy, Alloys, Sci. Rep. Res. Inst., Tohoku Univ. Ser. A, Phys., Chem. Metall. 10 (1958)
Evolution of δ phase microstructure in alloy 718, 7th Int. Symp. Superalloy 718 110–119. 〈http://hdl.handle.net/10097/26866〉.
Deriv. 2010 (2) (2010) 737–750, https://doi.org/10.1002/9781118495223.ch57. [56] Y. Shirakawa, Y. Tanji, H. Moriya, I. Oguma, The Elastic Constants of Ni and Ni-Fe
[32] C. Silva, M. Song, K. Leonard, M. Wang, G. Was, J. Busby, Characterization of alloy (fcc) Alloys, Sci. Rep. Res. Inst., Tohoku Univ. Ser. A, Phys., Chem. Metall. 21
718 subjected to different thermomechanical treatments, Mater. Sci. Eng. A. 691 (1969) 187–200. 〈http://hdl.handle.net/10097/27496〉.
(2017) 195–202, https://doi.org/10.1016/j.msea.2017.03.045. [57] M. Yamamoto, On elastic constants of nickel crystals, Phys. Rev. 77 (1950) 566,
[33] Standard Specification for Precipitation-Hardening Nickel Alloy (UNS N07718) https://doi.org/10.1103/PhysRev.77.566.
Plate, Sheet, and Strip for High-Temperature Service, Annu. B. ASTM Stand. 07, [58] H. Masumoto, H. Saito, S. Sawaya, Thermal expansion and temperature
2010: 7–10. https://doi.org/10.1520/B0670–07. dependence of young’s modulus of nickel-copper alloys, Trans. Jpn. Inst. Met. 11
[34] M.N. Gussev, R.H. Howard, K.A. Terrani, K.G. Field, Sub-size tensile specimen (1970).
design for in-reactor irradiation and post-irradiation testing, Nucl. Eng. Des. 320 [59] S. Li, Y. Wang, X. Wang, F. Xue, G-phase precipitation in duplex stainless steels
(2017) 298–308, https://doi.org/10.1016/j.nucengdes.2017.06.008. after long-term thermal aging: A high-resolution transmission electron microscopy
[35] M.N. Gussev, J.T. Busby, K.G. Field, M.A. Sokolov, S.E. Gray, Role of scale factor study, J. Nucl. Mater. 452 (2014) 382–388, https://doi.org/10.1016/j.
during tensile testing of small specimens, ASTM Spec. Tech. Publ. STP 1576 (2015) jnucmat.2014.05.069.
31–49, https://doi.org/10.1520/STP157620140013. [60] Y. Matsukawa, T. Takeuchi, Y. Kakubo, T. Suzudo, H. Watanabe, H. Abe,
[36] S. Taller, T. Austin, V. Paquit, K. Terrani, Report on Properties and Microstructure T. Toyama, Y. Nagai, The two-step nucleation of G-phase in ferrite, Acta Mater. 116
of 3D Printed Inc-718, ORNL/TM-2021/2149, Oak Ridge, TN (U. S. ) (2021), (2016) 104–113, https://doi.org/10.1016/j.actamat.2016.06.013.
https://doi.org/10.2172/1820785. [61] Y. Fan, T.G. Liu, L. Xin, Y.M. Han, Y.H. Lu, T. Shoji, Thermal aging behaviors of
[37] S. Specification, B670: Standard Specification for Precipitation-Hardening Nickel duplex stainless steels used in nuclear power plant: A review, J. Nucl. Mater.
Alloy (UNS N07718) Plate, Sheet, and Strip for High-Temperature Service, Annu. (2020), 152693, https://doi.org/10.1016/j.jnucmat.2020.152693.
B. ASTM Stand 07 (2010) 7–10, https://doi.org/10.1520/B0670-07R18.2. [62] P.B. Hirsch, F.J. Humphreys, Plastic deformation of two-phase alloys containing
[38] C. Kumara, A.R. Balachandramurthi, S. Goel, F. Hanning, J. Moverare, Toward a small nondeformable particles, in: A.S. Argon (Ed.), Phys. Strength Plast., M. I. T.
better understanding of phase transformations in additive manufacturing of Alloy Press, 1969: p. 189.
718, Materialia 13 (2020), 100862, https://doi.org/10.1016/j.mtla.2020.100862. [63] F.R.N. Nabarro, Fifty-year study of the Peierls-Nabarro stress, in: Mater. Sci. Eng.
[39] A.K. Sahu, S. Bag, Design of a double aging treatment for the improvement of A, 234–236, 1997, pp. 67–76, https://doi.org/10.1016/s0921-5093(97)00184-6.
mechanical and microstructural properties of pulse micro-plasma arc welded alloy [64] R.E. Stoller, S.J. Zinkle, On the relationship between uniaxial yield strength and
718, J. Mater. Sci. 56 (2021) 13400–13415, https://doi.org/10.1007/s10853-021- resolved shear stress in polycrystalline materials, J. Nucl. Mater. 283–287 (2000)
06121-8. 349–352, https://doi.org/10.1016/S0022-3115(00)00378-0.
[40] F. Bachmann, R. Hielscher, H. Schaeben, Texture analysis with MTEX- Free and [65] L. Tan, J.T. Busby, Formulating the strength factor α for improved predictability of
open source software toolbox, Solid State Phenom. 160 (2010) 63–68, https://doi. radiation hardening, J. Nucl. Mater. 465 (2015) 724–730, https://doi.org/
org/10.4028/www.scientific.net/SSP.160.63. 10.1016/j.jnucmat.2015.07.009.
[41] G. Cliff, G.W. Lorimer, The quantitative analysis of thin specimens, J. Microsc. 103 [66] G.E. Lucas, The evolution of mechanical property change in irradiated austenitic
(1975) 203–207, https://doi.org/10.1111/j.1365-2818.1975.tb03895.x. stainless steels, J. Nucl. Mater. 206 (1993) 287–305, https://doi.org/10.1016/
[42] S. Taller, L. Scime, K. Terrani, Utilizing a Dynamic Segmentation Convolutional 0022-3115(93)90129-M.
Neural Network for Microstructure Analysis of Additively Manufactured Superalloy [67] B.L. Eyre, Transmission electron microscope studies of point defect clusters in fcc
718, Microsc. Microanal. 27 (2021) 3110–3112, https://doi.org/10.1017/ and bcc metals, J. Phys. F. Met. Phys. 3 (1973) 422–470, https://doi.org/10.1088/
s143192762101076x. 0305-4608/3/2/009.
[43] J. Schindelin, I. Arganda-Carreras, E. Frise, V. Kaynig, M. Longair, T. Pietzsch, [68] D. Hull, D.J. Bacon, Strength of Crystalline Solids, Introd. Dislocations 1 (2011)
S. Preibisch, C. Rueden, S. Saalfeld, B. Schmid, J.Y. Tinevez, D.J. White, 205–249, https://doi.org/10.1016/b978-0-08-096672-4.00010-4.
V. Hartenstein, K. Eliceiri, P. Tomancak, A. Cardona, Fiji: An open-source platform [69] V. Bata, E.V. Pereloma, An alternative physical explanation of the Hall-Petch
for biological-image analysis, Nat. Methods 9 (2012) 676–682, https://doi.org/ relation, Acta Mater. 52 (2004) 657–665, https://doi.org/10.1016/j.
10.1038/nmeth.2019. actamat.2003.10.002.
[44] K. Yuan, W. Guo, P. Li, Y. Zhang, X. Li, X. Lin, Thermomechanical behavior of laser [70] J.A. Turner, J. Belak, N. Barton, M. Bement, N. Carlson, R. Carson, S. DeWitt, J.
metal deposited Inconel 718 superalloy over a wide range of temperature and L. Fattebert, N. Hodge, Z. Jibben, W. King, L. Levine, C. Newman, A. Plotkowski,
strain rate: Testing and constitutive modeling, Mech. Mater. 135 (2019) 13–25, B. Radhakrishnan, S.T. Reeve, M. Rolchigo, A. Sabau, S. Slattery, B. Stump, ExaAM:
https://doi.org/10.1016/j.mechmat.2019.04.024. Metal additive manufacturing simulation at the fidelity of the microstructure, Int.
[45] A. Oradei-Basile, J.F. Radavich, A Current T-T-T Diagram for Wrought Alloy 718, J. High. Perform. Comput. Appl. 36 (2022) 13–39, https://doi.org/10.1177/
2012: 325–335. https://doi.org/10.7449/1991/superalloys_1991_325_335. 10943420211042558.
[46] J.F. Radavich, The Physical Metallurgy of Cast and Wrought Alloy 718, in:
Superalloys 718 Metall. Appl., TMS, 2004, pp. 229–240, https://doi.org/10.7449/
1989/Superalloys_1989_229_240.

15