Materials Science & Engineering A 714 (2018) 75–83

Contents lists available at ScienceDirect

Materials Science & Engineering A

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

Stress relaxation in a nickel-base superalloy at elevated temperatures with in T

situ neutron diffraction characterization: Application to additive
manufacturing☆
⁎
Zhuqing Wanga, Alexandru D. Stoicab, Dong Mab, Allison M. Beesea,
a
Department of Materials Science and Engineering, Pennsylvania State University, University Park, PA 16802, USA
b
Chemical and Engineering Materials Division, Neutron Sciences Directorate, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA

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

Keywords: The complex thermal histories in additive manufacturing (AM) of metals result in the presence of residual
Inconel 625 stresses in the fabricated components. The amount of residual stress accumulated during AM depends on the
Stress relaxation high temperature constitutive behavior of the material. The rapid solidification and repeated thermal cycles with
Neutron diffraction each laser pass result in material contraction, and subject the surrounding, constrained material to both elevated
Additive manufacturing
temperatures and internal stresses, providing driving forces for stress relaxation. In this study, the stress re-
laxation behavior and mechanisms of conventionally processed and additively manufactured Inconel 625 (CP-
IN625 and AM-IN625) at 600 °C and 700 °C were investigated via compression tests up to an engineering strain
of 9% with in situ neutron diffraction characterization. The stress decayed to a plateau stress equivalent to 18%
of the peak stress in CP-IN625 and 16% in AM-IN625 at 600 °C, and 39% in CP-IN625 and 44% in AM-IN625 at
700 °C. At the same temperature, the stress relaxation rate in AM-IN625 was twice as high as that in CP-IN625,
and the magnitude of the plateau stress in AM-IN625 was slightly lower than that in CP-IN625, as the textured
AM-IN625 had much larger grains than the texture-free CP-IN625. The stress relaxation in CP- and AM-IN625
was deduced to be controlled by dislocation glide and climb, where dislocations interact with grain boundaries,
solute atoms, and secondary phases. The stress relaxation constitutive behavior reported here is a necessary
input for the development of accurate thermomechanical models used to predict and minimize residual stresses
and distortion in AM, as well as to predict the stress relaxation behavior of Inconel 625 in high temperature
structural applications.

1. Introduction deposited layer, and powder feedstock is delivered to the melt pool. As
the laser advances to continue depositing the layer, the melt pool ra-
Nickel-base superalloys (e.g., Inconel 625 and Inconel 718) are pidly solidifies and fuses to the layer below [7–9].
widely used in aerospace, marine, and petrochemical applications due During deposition, as subsequent layers are added, the surrounding
to their high mechanical strength, creep resistance, and corrosion re- material is subjected to thermal cycling with each additional laser pass.
sistance at elevated temperatures [1–3]. However, fabricating complex The material in the component being fabricated acts as a constraint,
components from nickel-base superalloys using traditional subtractive resulting in the generation of internal stresses during deposition
machining can be expensive, as the hard nickel-base superalloys wear [10–12]. The extent to which these stresses are relieved during fabri-
out machine tools easily. In addition, given the high cost of nickel-base cation depends on the stress magnitude and temperature as a function
superalloys, it is of interest to use additive manufacturing (AM) to of time at each point within the component during fabrication, and the
fabricate near net-shape components with minimal waste of material corresponding stress relaxation behavior [13]. Thermomechanical
[4–6]. In powder-based directed energy deposition (DED) AM with a modeling can be used to predict residual stresses that develop during
laser heat source, a laser beam melts the substrate or previously AM, as well as to optimize build parameters to minimize distortion in

☆
This manuscript has been authored 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).
⁎
Corresponding author.
E-mail address: [email protected] (A.M. Beese).

https://doi.org/10.1016/j.msea.2017.12.058
Received 11 July 2017; Received in revised form 26 October 2017; Accepted 13 December 2017
Available online 15 December 2017
0921-5093/ © 2017 Elsevier B.V. All rights reserved.
Z. Wang et al. Materials Science & Engineering A 714 (2018) 75–83

the fabricated components. However, for accurate modeling, the stress

relaxation behavior, on the timescale over which the temperature
during AM is elevated, is a required input. In addition, an accurate
description of stress relaxation is necessary for predicting the behavior
of components that will be loaded in service at elevated temperatures
[14]. For example, compressive stresses are intentionally introduced
into Inconel 718 blade fixtures to hold the blades tightly in turbine
engines, but these stresses may relax during a component's lifetime at
high service temperatures, leading to undesirable vibration or failure
[15].
Prior research has been performed to investigate the microstructure
and creep/stress relaxation behavior in conventionally processed
nickel-base superalloys at elevated temperatures [16–21]. The creep
resistance in nickel-base superalloy Inconel 625 is due to solid solution
strengthening by niobium and molybdenum solute atoms in the nickel-
chromium γ matrix. Additively manufactured Inconel 625 contains
small amounts of niobium-, titanium-, and molybdenum-rich carbides,
Fig. 1. Schematic of the experimental setup for compression with in situ neutron dif-
nitrides, and Laves phase in interdendric regions and along boundaries
fraction, in which diffraction patterns from the axial and transverse directions were re-
of columnar γ grains [4,5,22], while conventionally processed Inconel corded in bank 1 and bank 2, respectively.
625 contains small amounts of niobium-, titanium-, and molybdenum-
rich carbides primarily along boundaries of equiaxed γ grains [23,24].
powder was deposited using a powder feed rate of 16 g/min, an argon
Diehl and Messler [16] investigated the stress relaxation behavior of
flow rate of 9.4 L/min, a laser power of 2 kW, and a scanning speed of
annealed Inconel 625 welds under tension, with applied stresses ap-
10.6 mm/s [27]. Cylindrical specimens measuring 5 mm in diameter
proximately equal to the temperature-dependent yield stress, between
and 10 mm in length were extracted from AM-IN625 using wire elec-
566–982 °C. They found that the stress relaxed exponentially, with 13%
trical discharge machining (EDM) such that their loading axes in the
of the initially applied stress relieved at 566 °C, 87% at 871 °C, and 90%
subsequent compression tests were aligned with the laser scanning di-
at 982 °C all after 8 h.
rection. Specimens with the same geometry were also extracted from
Rodriguez et al. [17] studied creep in ultrafine-grained Inconel 625
the CP-IN625 substrate.
from 538 °C to 650 °C over a period of 36 h. Through investigation of
Stress relaxation tests under compression were performed at 600 °C
the microstructure of deformed samples after testing using scanning
and 700 °C on the conventionally processed and additively manu-
and transmission electron microscopy, they concluded that creep was
facturing cylindrical specimens. These tests were performed with in situ
controlled by dislocation glide and climb, as dislocation tangles were
time-of-flight neutron diffraction using the VULCAN instrument at the
found within grains. Mathew et al. [18] evaluated creep behavior and
Spallation Neutron Source of Oak Ridge National Laboratory [28,29]. A
rupture life of conventionally processed Inconel 625 from 760 °C to
schematic of the stress relaxation compression test with neutron dif-
815 °C. These data indicated that dislocation creep was active under
fraction is shown in Fig. 1. VULCAN has two detector banks that, being
these conditions. Evans et al. [19] studied microstructural changes in
positioned at −/+90° diffraction angles, simultaneously record dif-
conventionally processed Inconel 625 of varying thickness subjected to
fraction patterns from grains with hkl-specific lattice planes whose
creep tests at 750 °C for 1310 h. They showed that the creep resistance
plane normals are along the axial and transverse directions, respec-
decreased with decreasing thickness due to fewer grain boundaries in
tively. The grain-orientation dependent, or hkl-specific, lattice strain,
the thinner material. Scanning and transmission electron microscopy 0
dhkl − dhkl
analyses of tested samples showed secondary phases, including car- εhkl, was computed as εhkl = 0
dhkl
, where dhkl is the hkl-specific
bides, δ phase, and σ phase, precipitated along grain boundaries and stressed lattice spacing, and 0
the corresponding stress-free lattice
dhkl
inhibited grain boundary sliding. spacing. In addition to stress relaxation tests, uniaxial compression tests
While stress relaxation has been studied in conventionally processed were performed at 600 °C and 700 °C to determine appropriate dis-
Inconel 625 [16–21], it has not yet been reported in additively manu- placements to apply for stress relaxation tests.
factured Inconel 625. The microstructure of additively manufactured In stress relaxation tests, the specimens were first heated to 600 °C
materials, and therefore, potentially their stress relaxation behavior, or 700 °C and then compressed to an engineering strain of 0.09. The
differs from their conventionally processed counterparts. In addition, applied strain at each test temperature corresponded to an applied, or
the typical timescales studied for stress relaxation or creep in con- peak, stress of 75–85% of the ultimate tensile strength at that test
ventionally processed Inconel 625 are on the order of 10s–1000s of temperature [30]. The macroscopic strain was held constant, and the
hours, whereas the timescale during which stress relaxation may be an evolving stress was recorded. After the stress reached a steady state
active mechanism during AM processing is much shorter, as discussed plateau, defined by a decrease in stress magnitude less than 5 MPa in
in Section 3. The aim of this work was to quantify the stress relaxation 5 min, the specimen was unloaded and cooled to room temperature. A
behavior at 600 °C and 700 °C over a period up to 2.5 h, and to examine type K thermocouple welded to the center of each specimen was used to
possible stress relaxation mechanisms in conventionally processed and measure the temperature throughout the test. The test parameters are
additively manufactured Inconel 625 via in situ time-of-flight neutron given in Table 1, while Fig. 2 shows the applied strain, temperature,
diffraction. The stress relaxation behavior measured and reported here and resulting engineering stress, as a function of time during a stress
can be used in thermomechanical models of Inconel 625 to predict and relaxation test at 700 °C.
mitigate residual stresses and distortion in AM as well as to predict the In order to evaluate the microstructure, untested cylindrical speci-
behavior of Inconel 625 in high temperature structural applications. mens extracted from CP- and AM-IN625 were polished using 0.05 µm
colloidal silica and electrolytically etched using 10 wt% oxalic acid in
2. Materials and methods DI water at 4 V for 20 s. The specimens were then examined using a
scanning electron microscope (SEM, FEI Quanta 200).
A 102 mm long × 28 mm tall × 7 mm thick Inconel 625 wall was
deposited by DED AM onto a conventionally processed Inconel 625
substrate (ASTM B443 Grade 1 [25,26]). Pre-alloyed Inconel 625

76
Z. Wang et al. Materials Science & Engineering A 714 (2018) 75–83

Table 1
Test parameters for stress relaxation tests.

Temperature (°C) 600 600 700 700

Processing method Conventionally processed Additively manufactured Conventionally processed Additively manufactured
Heating rate (°C/s) 0.5 0.5 0.6 0.6
Strain rate (s−1) 2.8 × 10−5 2.8 × 10−5 1.7 × 10−5 1.7 × 10−5
Applied compressive strain 0.09 0.09 0.09 0.09

about 18% of the stress was relieved at 600 °C and 39% was relieved at
700 °C by the time the stress plateaued. In AM-IN625, about 16% of the
stress was relieved at 600 °C and 44% was relieved at 700 °C when the
stress plateaued. At the same temperature and applied strain, AM-
IN625 had lower peak and plateau stresses than CP-IN625. In the same
material, the peak and plateau stresses decreased as the temperature
increased.

3.2. Neutron diffraction measurements of stress partitioning

Neutron diffraction patterns collected from the axial direction of an

AM-IN625 specimen during heating, loading, and stress relaxation are
given in Fig. 4. This confirms that the lattice spacings increased during
heating and stress relaxation, and decreased during compressive
loading.
The hkl-specific lattice stresses, σhkl , in Inconel 625 along the axial
Fig. 2. Applied engineering strain, temperature, and resulting engineering stress as a
function of time for a stress relaxation test on an additively manufactured specimen he-
direction were computed as σhkl = Ehkl εhkl , where Ehkl is the hkl-specific
ated at 700 °C. This shows the testing procedure of (a) heating, (b) applying a com- Young's modulus [36]. The computed axial lattice stresses as a function
pressive load, (c) holding the displacement constant to allow for stress relaxation, (d) of time are plotted in Fig. 5. As Inconel 625 consists of a primary γ
unloading, and (e) cooling. matrix (typically greater than 98 vol% [22]), the relaxation of lattice
stresses in the γ phase were measured. Fig. 5 shows that the lattice
3. Results stresses in the grains with plane normals along the 〈220〉, 〈311〉,
〈200〉, and 〈111〉 directions of the γ phase relaxed simultaneously and
3.1. Macroscopic stress relaxation behavior followed a similar trend as the macroscopic stress versus time during
stress relaxation. In CP-IN625, the lattice stresses in grains with plane
Creep and stress relaxation in Inconel 625 have been found to be normals along 〈111〉 direction were highest, followed by almost the
controlled by dislocation glide and climb at similar levels of applied same lattice stresses in grains with normals along 〈311〉, 〈200〉, and
normalized tensile stresses (stress divided by shear modulus) and 〈220〉 directions. In AM-IN625, the lattice stresses in grains with nor-
homologous temperature as those in the present study [17,18]. In ad- mals along the 〈111〉 were the highest, but these measurements had
dition, creep deformation maps for pure nickel (with grain diameters significant scatter due to texture [1], followed by almost the same
spanning 10 µm to 1 mm compared to the average diameters of lattice stresses in grains with normals along the 〈311〉 and 〈200〉 di-
25–53 µm in this study, as discussed in Section 4) indicate that for the rections. The lattice stresses in grains with normals along 〈220〉 were
temperatures and stress levels applied here, the dominant creep me- lowest. In both CP- and AM-IN625, the macroscopic stress versus time
chanism is dislocation motion [31]. The stress relaxation mechanisms curve lies between lattices stresses in grains with normals along 〈111〉
in CP- and AM-IN625 in the present study are further discussed in direction and those with normals along the remaining directions.
Section 4. In CP-IN625, about 5% of the initial lattice stresses were relieved at
The relevant timescale for stress relaxation of nickel-base alloys 600 °C and 25–32% were relieved at 700 °C by the time the stresses
during AM, or the timescale during which the temperature remains plateaued. In AM-IN625, about 7–9% of the lattice stresses were re-
above 0.25 of the melting temperature, or 240 °C (the temperature lieved at 600 °C and 39–48% were relieved at 700 °C by the time the
range for which creep is primarily controlled by dislocation motion stresses plateaued.
[31]) during deposition, depends on the component geometry and
processing parameters. In simulations of DED of Inconel 718, it has 4. Discussion
been shown that for walls measuring similar sizes as that in the present
study, with linear heat inputs on the order of 30 J/mm, it takes To determine a characteristic stress relaxation time constant, the
5–20 min for the material to drop below 240 °C [32,33]. These results macroscopic stress relaxation curves in Fig. 3 were fitted using a simple
point to the fact that quantifying the stress relaxation behavior of In- rheological viscoplastic standard linear solid (SLS) model [37,38],
conel over a period of 10s of minutes is required for thermomechanical namely a spring, in parallel with a spring and a dashpot that are in
modeling of DED AM of this material. Additionally, with increasing series (Fig. 3a). For an increment of applied strain, the evolution of
linear heat input, as well as in builds where heat accumulates due to stress, σ(t), is expressed as:
faster scanning speeds or more layers, slower cooling rates will be
t
present, resulting in longer periods of time being relevant for stress σ (t ) = ε0 ⎡EA + EB exp ⎛− ⎞ ⎤
⎣ ⎝ τ ⎠⎦ (1)
relaxation [34,35]. Therefore, here we investigated the stress relaxation
behavior of Inconel 625 over a period of up to 2.5 h. where EA and EB are material constants that describe the linear elastic
The stress relaxation curves for CP- and AM-IN625 at 600 °C and behavior of the material, ε0 is the applied engineering strain during
700 °C are given in Fig. 3. The stress and the rate of stress relaxation, stress relaxation, and τ is the relaxation time constant, which is the time
defined as the slope of the normalized stress-time curves, decreased it takes for the stress to drop to 1/e of the peak stress.
continuously until the stresses reached a steady state. In CP-IN625, The calibrated parameters for the SLS model, given in Table 2, were

77
Z. Wang et al. Materials Science & Engineering A 714 (2018) 75–83

Fig. 3. (a) Engineering stress, and (b) engineering

stress normalized by peak stress, as a function of time
during stress relaxation in conventionally processed
and additively manufactured Inconel 625 at 600 °C
and 700 °C. Symbols represent experimental results
and lines represent the fitted standard linear solid
(SLS) model, which is schematically shown in the
bottom right corner of (a).

〈311〉 and 〈200〉 directions were the same, indicating no load or stress
transfer between these grains.
In order to examine any load transfer between grains, the relative
difference between lattice stresses, χi, at a given time, ti, was computed
as:
σhkl, ti − σh′k′l′, ti
χi =
(σhkl, ti + σh′k′l′, ti )/2 (2)

where 〈hkl〉 is 〈311〉, 〈200〉, or 〈220〉 in CP-IN625, and 〈311〉 or 〈200〉

in AM-IN625, and 〈h′k′l′〉 is 〈111〉 in CP-IN625 and 〈220〉 in AM-
IN625. Here, lattice stresses in grains with plane normals along the
〈111〉 direction in AM-IN625 are not considered.
The average value of χ in CP-IN625 at 600 °C was 0.4 ± 0.03 and at
700 °C was 0.6 ± 0.04. The average value of χ in AM-IN625 at 600 °C
was 0.1 ± 0.03 and at 700 °C was 0.2 ± 0.05. The small standard de-
viation of χ for both CP- and AM-IN625 at both temperatures examined
indicates that lattice stresses in differently oriented grains relaxed to-
gether. Therefore, stress relaxation is not due to load or stress transfer
Fig. 4. Neutron diffraction patterns of additively manufactured Inconel 625 corre- between differently oriented grains in the γ phase.
sponding to the axial direction during a stress relaxation test. Curves from the bottom to Previous research has shown that creep in IN625 is primarily con-
top showing pattern changes during heating (black to red), compressive loading (red to trolled by dislocation motion with applied normalized (by shear mod-
green), and stress relaxation (green to purple, in which the stress in the green pattern
ulus) tensile stresses of 2 × 10−3–7 × 10−3 and homologous tem-
corresponds to the peak stress). The inset shows the shifting of the (200) peak during a
stress relaxation test. (For interpretation of the references to color in this figure legend,
peratures of 0.47–0.64 [17,18]. In the present study, the homologous
the reader is referred to the web version of this article). temperatures ranged from 0.51 to 0.57, and the normalized compres-
sive stresses ranged from 7 × 10−3 to 10−2, which should promote
dislocation-based deformation mechanisms. In order to verify that
determined by fitting the curves in Fig. 3a using Eq. (1). As the stress in
dislocation glide and climb were responsible for the stress relaxation,
these tests did not drop to 1/e of the peak stress, a relaxation time
the full width at half maximum (FWHM) values of major peaks from the
constant, τ′, defined as the time for the stress to drop to 85% of the peak
γ phase were plotted as a function of time during stress relaxation at
stress, a stress level reached in all specimens studied, was computed.
700 °C, as shown in Figs. 6 and 7.
Since stress relaxation rate is time dependent, here the reciprocal of
Using a Gaussian approximation of diffraction peaks, the square
relaxation time constant, 1/τ′, is used to represent stress relaxation rate,
addition law can be used to estimate the contribution of relaxation to
r, as given in Table 2. Fig. 3 and Table 2 show that, for both con-
the peak width change, as shown in Figs. 6 and 7. Thus, ΔFWHM, the
ventionally processed and additively manufactured material, as the
change in the peak width caused by the relaxed inhomogeneity, follows
temperature increased, the stress relaxation rate increased and the peak
the equation:
and plateau stresses decreased. At the same temperature and applied
strain, AM-IN625 had a faster stress relaxation rate, and lower peak and ∆FWHM 2 = FWHMt20 − FWHMt2 (3)
plateau stresses than its conventionally processed counterpart.
The stress relaxation rate, r, was also computed for lattice stress where (FWHM )t0 is the initial full width at half maximum at the be-
evolution and the rates for the specimens at 600 °C and 700 °C are given ginning of stress relaxation and (FWHM )t is the final full width at half
in Table 3. As shown in Table 3, in a single specimen, the lattice stress maximum at the end of stress relaxation.
relaxation rates in grains with different orientations were similar and At 700 °C, ΔFWHM amounts to 21–34% of FWHMt0 in CP-IN625 and
on the same order of magnitude as the macroscopic stress relaxation 29–35% in AM-IN625. These values are too high to be explained solely
rate (Table 2). As shown Fig. 5, at 600 °C and 700 °C in CP-IN625, by the relaxation of intergranular strains [39]. It can be concluded that
lattice stress relaxation curves in grains with plane normals along the the decrease in FWHM, or peak sharpening, is due to dislocation an-
〈311〉, 〈200〉, and 〈220〉 lie on top of each other. At 600 °C and 700 °C nihilation during stress relaxation [40], which indicates that stress re-
in AM-IN625, lattice stresses in grains with plane normals along the laxation is controlled by dislocation motion. It is thus reasoned that
stress relaxation, with applied strains of 9%, at 600 °C and 700 °C, in

78
Z. Wang et al. Materials Science & Engineering A 714 (2018) 75–83

Fig. 5. Axial lattice stresses in grains with plane

normals along the 〈220〉, 〈311〉, 〈200〉, and 〈111〉
directions versus time in (a) conventionally pro-
cessed and (b) additively manufactured Inconel 625
during stress relaxation at 600 °C, and (c) con-
ventionally processed and (d) additively manu-
factured Inconel 625 during stress relaxation at
700 °C.

both CP- and AM-IN625 is controlled by dislocation glide and climb. minor axes parallel to the load direction. Therefore, if considering
Solute atoms, molybdenum and niobium, in the γ matrix in Inconel 625 compression along the length of the wall, the relevant grain dimension
introduce a frictional force that impedes gliding dislocations, and an along which dislocations may travel before encountering a grain
atmosphere that provides a dragging effect on climbing dislocations boundary is the minor axis. A histogram of the length of minor axes in
[41,42]. Secondary phases, including carbides, nitrides, and Laves AM-IN625 and grain diameter in CP-IN625 measured by EBSD are
phase, are also dispersed in the γ matrix in Inconel 625 [4,5,22–24], as given in Fig. 9. These show that the additively manufactured material
shown in Fig. 8. These features serve as obstacles to block dislocation had much larger grains with a larger grain size distribution than the CP-
glide and climb [42,43]. Therefore, secondary phases, solute atoms, as IN625. The average value for the minor axes in AM-IN625 was 53 µm
well as grain boundaries impede dislocation motion, which decreases with a maximum value of 187 µm, while the average grain diameter in
the rate of stress relaxation. CP-IN625 was 25 µm with a maximum value of 53 µm. Therefore, the
The higher rate of stress relaxation in AM-IN625 versus CP-IN625 number of grain boundaries that dislocations will eventually run into in
can be partially attributed to different grain size/morphology in these CP-IN625 is significantly higher than that in AM-IN625, meaning grain
differently processed materials. The AM-IN625 contained elongated size strengthening may contribute to the slower stress relaxation in CP-
grains oriented along the build direction, while CP-IN625 contained IN625 relative to AM-IN625.
smaller equiaxed grains [1]. By approximating the grains in AM-IN625 Additionally, the stress as a function of time was higher in CP-IN625
as elliptical, the grain morphology can be described by the aspect ratio, than AM-IN625 because the latter had a strong 〈100〉 texture along the
defined as the ratio of the major to minor axis of the grain, so that an loading axis. This texture, as indicated by electron backscatter diffrac-
equiaxed grain has an aspect ratio of 1 [44]. The major axes of the tion (EBSD; Oxford Nordlys Max2) measurements [1], led to texture
elliptical grains were roughly parallel to the build direction, with the softening in the additively manufactured material compared to the

Table 2
Parameters for the standard linear solid model of stress relaxation in conventionally processed and additively manufactured Inconel 625 at 600 °C and 700 °C.

Temperature (°C) 600 600 700 700

Processing method Conventionally processed Additively manufactured Conventionally processed Additively manufactured
ε0 −0.09 −0.09 −0.09 −0.09
EA (GPa) 5.9 4.6 3.9 3.1
EB (GPa) 1.5 1.5 2.0 1.8
τ (min) 95.2 66.9 19.7 12.4
τ′ (min) 140.5 59.4 5.2 2.8
r (min−1) 7.1 × 10−3 1.7 × 10−2 0.19 0.36

79
Z. Wang et al. Materials Science & Engineering A 714 (2018) 75–83

Table 3
Lattice stress relaxation rates in grains with plane normals along the 〈311〉, 〈220〉, 〈200〉, and 〈111〉 directions in conventionally processed and additively manufactured Inconel 625 at
600 °C and 700 °C. The rates for lattice stresses in grains with plane normals along the 〈111〉 direction in additively manufactured Inconel 625 are not computed due to significant scatter.

Temperature (°C) 600 600 700 700

Processing method Conventionally processed Additively manufactured Conventionally processed Additively manufactured
r (min−1) 〈311〉 6.6 × 10−3 1.7 × 10−2 0.08 0.12
〈220〉 6.4 × 10−3 7.0 × 10−3 0.11 0.16
〈200〉 3.6 × 10−3 1.5 × 10−2 0.07 0.12
〈111〉 4.6 × 10−3 NA 0.08 NA

texture-free conventionally processed counterpart [45]. relaxation rate increased, and the peak and plateau stresses de-
creased, for the same applied strain, in both conventionally pro-
5. Summary and conclusions cessed and additively manufactured Inconel 625.
• At the same temperature and applied strain, additively manu-
In this work, high temperature stress relaxation behavior of addi- factured Inconel 625 had a higher stress relaxation rate and lower
tively manufactured and conventionally processed Inconel 625 was peak and plateau stresses than conventionally processed Inconel
investigated at macro- and micro-scales by performing compression 625, due to different texture and grain sizes in these two materials.
tests with in situ neutron diffraction. The primary findings are as fol- • At 600 °C and 700 °C, the stress relaxation in additively manu-
lows: factured and conventionally processed Inconel 625 was found to not
result from load transfer between differently oriented grains in the γ
• The macroscopic stress and hkl-specific lattice stresses in con- phase. It is concluded that stress relaxation is primarily controlled
ventionally processed and additively manufactured Inconel 625 by dislocation glide and climb.
decreased exponentially and reached a steady state after 150 min at • With an applied compressive engineering strain of 9%, in additively
600 °C and 120 min at 700 °C. As temperature increased, the stress manufactured Inconel 625, 16% of the resulting stress was relieved

Fig. 6. Full width at half maximum (FWHM) of grains with plane normals along the (a) 〈311〉, (b) 〈220〉, and (c) 〈200〉 directions versus time in conventionally processed Inconel 625
during stress relaxation at 700 °C.

80
Z. Wang et al. Materials Science & Engineering A 714 (2018) 75–83

Fig. 7. Full width at half maximum (FWHM) of grains with plane normals along the (a) 〈311〉, (b) 〈220〉, and (c) 〈200〉 directions versus time in additively manufactured Inconel 625
during stress relaxation at 700 °C.

Fig. 8. (a) Back-scattered electron image of con-

ventionally processed Inconel 625 showing bright
molybdenum- and niobium-rich carbides and dark
titanium-rich carbides. (b) Secondary electron image
of additively manufactured Inconel 625 showing
bright molybdenum- and niobium-rich carbides, ni-
trides, and Laves phase.

at 600 °C and 44% was relieved at 700 °C by the time the stress capture the stress relaxation behavior of Inconel 625. It can be used
plateaued. With the same applied strain in conventionally processed in future thermomechanical models to predict and mitigate residual
Inconel 625, 18% of the resulting stress was relieved at 600 °C and stresses and distortion in additively manufactured components. In
39% was relieved at 700 °C by the time the stress plateaued. addition, the model can be used to predict stress relaxation behavior
• The calibrated standard linear solid rheological model is able to of Inconel 625 in high temperature structural applications.

81
Z. Wang et al. Materials Science & Engineering A 714 (2018) 75–83

Mater. Des. 113 (2017) 169–177, http://dx.doi.org/10.1016/j.matdes.2016.10.

003.
[13] Z. Wang, A.D. Stoica, D. Ma, A.M. Beese, Stress relaxation behavior and mechan-
isms in Ti-6Al-4V determined via in situ neutron diffraction: application to additive
manufacturing, Mater. Sci. Eng. A 707 (2017) 585–592, http://dx.doi.org/10.
1016/j.msea.2017.09.071.
[14] Z. Zhou, A.S. Gill, D. Qian, S.R. Mannava, K. Langer, Y. Wen, V.K. Vasudevan, A
finite element study of thermal relaxation of residual stress in laser shock peened
IN718 superalloy, Int. J. Impact Eng. 38 (2011) 590–596, http://dx.doi.org/10.
1016/j.ijimpeng.2011.02.006.
[15] O. Bapokutty, Z. Sajuri, J. Syarif, Stress relaxation behavior of heat treated inconel
718, J. Appl. Sci. 12 (2012) 870–875, http://dx.doi.org/10.1017/
CBO9781107415324.004.
[16] M.J. Diehl, R.W. Messler Jr, Using stress relaxation tests for evaluating and opti-
mizing postweld heat treatments of alloy 625 welds, Weld. J. 74 (1995) 109–114
http://aws.perusion.com/wj/supplement/WJ_1995_04_s109.pdf.
[17] R. Rodriguez, R.W. Hayes, P.B. Berbon, E.J. Lavernia, Tensile and creep behavior of
cryomilled Inco 625, Acta Mater. 51 (2003) 911–929, http://dx.doi.org/10.1016/
S1359-6454(02)00494-9.
[18] M.D. Mathew, K. Bhanu Sankara Rao, S.L. Mannan, Creep properties of service-
exposed Alloy 625 after re-solution annealing treatment, Mater. Sci. Eng. A 372
Fig. 9. Histogram of relevant grain dimensions in conventionally processed (grain dia- (2004) 327–333, http://dx.doi.org/10.1016/j.msea.2004.01.042.
[19] N.D. Evans, P.J. Maziasz, J.P. Shingledecker, Y. Yamamoto, Microstructure evolu-
meter) and additively manufactured (grain width) Inconel 625 measured by EBSD.
tion of alloy 625 foil and sheet during creep at 750 C, Mater. Sci. Eng. A 498 (2008)
412–420, http://dx.doi.org/10.1016/j.msea.2008.08.017.
[20] M.D. Mathew, P. Parameswaran, K. Bhanu Sankara Rao, Microstructural changes in
Acknowledgments
alloy 625 during high temperature creep, Mater. Charact. 59 (2008) 508–513,
http://dx.doi.org/10.1016/j.matchar.2007.03.007.
The authors gratefully acknowledge the financial support of the [21] V. Shankar, K. Bhanu Sankara Rao, S. Mannan, Microstructure and mechanical
National Science Foundation through award number CMMI-1402978. properties of Inconel 625 superalloy, J. Nucl. Mater. 288 (2001) 222–232, http://
dx.doi.org/10.1016/S0022-3115(00)00723-6.
Any opinions, findings, and conclusions or recommendations expressed [22] A.M. Beese, Z. Wang, A.D. Stoica, D. Ma, Absence of Dynamic Strain Aging in an
in this material are those of the authors and do not necessarily reflect Additively Manufactured Nickel-base Superalloy. , 2017 (Submitted for publica-
the views of the National Science Foundation. AMB also acknowledges tion).
[23] L. Ferrer, B. Pieraggi, J.F. Uginet, Microstructural evolution during thermo-
funding from the Oak Ridge Associated Universities Ralph E. Powe mechanical processing of alloy 625, in: Superalloys 718, 625 Various Derivatives,
Junior Faculty Enhancement Award. The additively manufactured 1991, pp. 217–228.
Inconel 625 wall was fabricated at the Center for Innovative Materials [24] S. Floreen, G.E. Fuchs, W.J. Yang, The metallurgy of alloy 625, in: Superalloys 718,
625, 706 Various Derivatives Mineral Metal Materials Society, 1994, pp. 13–37.
Processing through Direct Digital Deposition (CIMP-3D) in Penn State. 〈http://dx.doi.org/10.7449/1994/Superalloys_1994_13_37〉.
The neutron scattering work conducted at ORNL's Spallation Neutron [25] INCONEL ® Alloy 625, Spec. Met. Corp. SMC-020, 2006.
Source was sponsored by the Scientific User Facilities Division, Office of [26] ASTM B443, Standard Specification for Nickel-Chromium-Molybdenum-
Columbium Alloy (UNS N06625), Nickel-Chromium-Molybdenum-Silicon Alloy
Basic Energy Sciences, and U.S. Department of Energy. The authors (UNS N06219), and Nickel-Chromium-Molybdenum-Tungsten Alloy, ASTM Int,
thank Mr. Matthew Frost for the technical support at the VULCAN West Conshohocken, PA, 2014, pp. 1–7.
beamline. [27] E.R. Denlinger, J.C. Heigel, P. Michaleris, T.A. Palmer, Effect of inter-layer dwell
time on distortion and residual stress in additive manufacturing of titanium and
nickel alloys, J. Mater. Process. Technol. 215 (2015) 123–131, http://dx.doi.org/
References 10.1016/j.jmatprotec.2014.07.030.
[28] K. An, H.D. Skorpenske, A.D. Stoica, D. Ma, X.L. Wang, E. Cakmak, First in situ
[1] D. Ma, A.D. Stoica, Z. Wang, A.M. Beese, Crystallographic texture in an additively lattice strains measurements under load at VULCAN, Metall. Mater. Trans. A Phys.
manufactured nickel-base superalloy, Mater. Sci. Eng. A 684 (2017) 47–53. Metall. Mater. Sci. 42 (2011) 95–99, http://dx.doi.org/10.1007/s11661-010-
[2] D. Li, Q. Guo, S. Guo, H. Peng, Z. Wu, The microstructure evolution and nucleation 0495-9.
mechanisms of dynamic recrystallization in hot-deformed Inconel 625 superalloy, [29] G.M. Stoica, A.D. Stoica, M.K. Miller, D. Ma, Temperature-dependent elastic ani-
Mater. Des. 32 (2011) 696–705, http://dx.doi.org/10.1016/j.matdes.2010.07.040. sotropy and mesoscale deformation in a nanostructured ferritic alloy, Nat.
[3] A. Thomas, M. El-Wahabi, J.M. Cabrera, J.M. Prado, High temperature deformation Commun. 5 (2014) 1–8, http://dx.doi.org/10.1038/ncomms6178.
of Inconel 718, J. Mater. Process. Technol. 177 (2006) 469–472, http://dx.doi.org/ [30] INCONEL Alloy 625, Spec. Met. Corp., 2013. 〈http://www.specialmetals.com/
10.1016/j.jmatprotec.2006.04.072. documents/Inconelalloy625.pdf〉. (Accessed 25 April 2017).
[4] C.C. Silva, H.C. De Miranda, M.F. Motta, J.P. Farias, C.R.M. Afonso, A.J. Ramirez, [31] H.J. Frost, M.F. Ashby, A Second Report on Deformation Mechanism Maps, Harvard
New insight on the solidification path of an alloy 625 weld overlay, J. Mater. Res. University, 1973.
Technol. 2 (2013) 228–237, http://dx.doi.org/10.1016/j.jmrt.2013.02.008. [32] K.T. Makiewicz, Development of Simultaneous Transformation Kinetics
[5] F. Xu, Y. Lv, Y. Liu, F. Shu, P. He, B. Xu, Microstructural evolution and mechanical Microstructure Model with Application to Laser Metal Deposited Ti-6Al-4V and
properties of inconel 625 alloy during pulsed plasma arc deposition process, J. Alloy 718, The Ohio State University, 2013.
Mater. Sci. Technol. 29 (2013) 480–488. [33] A.M. Kamara, S. Marimuthu, L. Li, Finite element modeling of microstructure in
[6] M. Rombouts, G. Maes, M. Mertens, W. Hendrix, Laser metal deposition of Inconel laser-deposited multiple layer inconel 718 parts, Mater. Manuf. Process. 29 (2014)
625: microstructure and mechanical properties, J. Laser Appl. 24 (2012) 1–6, 1245–1252, http://dx.doi.org/10.1080/10426914.2014.930963.
http://dx.doi.org/10.2351/1.4757717. [34] V. Manvatkar, A. De, T. DebRoy, Spatial variation of melt pool geometry, peak
[7] D.D. Gu, W. Meiners, K. Wissenbach, R. Poprawe, Laser additive manufacturing of temperature and solidification parameters during laser assisted additive manu-
metallic components: materials, processes and mechanisms, Int. Mater. Rev. 57 facturing process, Mater. Sci. Technol. 31 (2015) 924–930, http://dx.doi.org/10.
(2012) 133–164, http://dx.doi.org/10.1179/1743280411Y.0000000014. 1179/1743284714Y.0000000701.
[8] X. Wu, A review of laser fabrication of metallic engineering components and of [35] R. Rai, J.W. Elmer, T.A. Palmer, T. DebRoy, Heat transfer and fluid flow during
materials, Mater. Sci. Technol. 23 (2007) 631–640, http://dx.doi.org/10.1179/ keyhole mode laser welding of tantalum, Ti–6Al–4V, 304L stainless steel and va-
174328407X179593. nadium, J. Phys. D. Appl. Phys. 40 (2007) 5753–5766, http://dx.doi.org/10.1088/
[9] Z. Wang, A.M. Beese, Effect of chemistry on martensitic phase transformation ki- 0022-3727/40/18/037.
netics and resulting properties of additively manufactured stainless steel, Acta [36] Z. Wang, A.D. Stoica, D. Ma, A.M. Beese, Diffraction and single-crystal elastic
Mater. 131 (2017) 410–422. constants of Inconel 625 at room and elevated temperatures determined by neutron
[10] R.J. Moat, A.J. Pinkerton, L. Li, P.J. Withers, M. Preuss, Residual stresses in laser diffraction, Mater. Sci. Eng. A 674 (2016) 406–412, http://dx.doi.org/10.1016/j.
direct metal deposited Waspaloy, Mater. Sci. Eng. A 528 (2011) 2288–2298, http:// msea.2016.08.010.
dx.doi.org/10.1016/j.msea.2010.12.010. [37] C. Zener, Elasticity and Anelasticity of Metals, University of Chicago Press, Chicago,
[11] P. Mercelis, J.-P. Kruth, Residual stresses in selective laser sintering and selective 1948.
laser melting, Rapid Prototyp. J. 12 (2006) 254–265, http://dx.doi.org/10.1108/ [38] B.J. Lazan, Damping of Materials and Members in Structural Mechanics, Pergamon
13552540610707013. Press, Oxford, 1968.
[12] Z. Wang, E. Denlinger, P. Michaleris, A.D. Stoica, D. Ma, A.M. Beese, Residual stress [39] P.R. Dawson, D.E. Boyce, R.B. Rogge, Correlation of diffraction peak broadening to
mapping in Inconel 625 fabricated through additive manufacturing: method for crystal strengthening in finite element simulations, Mater. Sci. Eng. A 399 (2005)
neutron diffraction measurements to validate thermomechanical model predictions, 13–25, http://dx.doi.org/10.1016/j.msea.2005.02.029.
[40] J. Pešička, R. Kužel, A. Dronhofer, G. Eggeler, The evolution of dislocation density

82
Z. Wang et al. Materials Science & Engineering A 714 (2018) 75–83

during heat treatment and creep of tempered martensite ferritic steels, Acta Mater. behavior of alloy 617 at intermediate temperatures, Metall. Mater. Trans. A 45A
51 (2003) 4847–4862, http://dx.doi.org/10.1016/S1359-6454(03)00324-0. (2014) 3010–3022, http://dx.doi.org/10.1007/s11661-014-2244-y.
[41] A. Wasilkowska, M. Bartsch, U. Messerschmidt, R. Herzog, A. Czyrska- [44] Z. Wang, T.A. Palmer, A.M. Beese, Effect of processing parameters on micro-
Filemonowicz, Creep mechanisms of ferritic oxide dispersion strengthened alloys, J. structure and tensile properties of austenitic stainless steel 304L made by directed
Mater. Process. Technol. 133 (2003) 218–224, http://dx.doi.org/10.1016/S0924- energy deposition additive manufacturing, Acta Mater. 110 (2016) 226–235,
0136(02)00237-6. http://dx.doi.org/10.1016/j.actamat.2016.03.019.
[42] H.J. Frost, M. Ashby, Rate-equations, in: Deform. Maps, Plast. Creep Met. Ceram., [45] G.Y. Chin, W.L. Mammel, Computer solutions of taylor analysis for axisymmetric
1st ed., Pergamon Press, 1982. flow, Trans. Metall. Soc. AIME 239 (1967) 1400–1405.
[43] J.K. Benz, L.J. Carroll, J.K. Wright, R.N. Wright, T.M. Lillo, Threshold stress creep

83