Corrosion Science 191 (2021) 109739

Contents lists available at ScienceDirect

Corrosion Science
journal homepage: www.elsevier.com/locate/corsci

Effects of heat treatment on corrosion fatigue and stress corrosion crack

growth of additive-manufactured Alloy 800H in high-temperature water
Jingfan Yang a, Miao Song b, Laura R. Hawkins c, d, Xiang Liu c, Lingfeng He c, Xiaoyuan Lou a, *
a
Materials Research and Education Center, Department of Mechanical Engineering, Auburn University, Auburn, AL, 36849, USA
b
Nuclear Engineering and Radiological Sciences, University of Michigan, Ann Arbor, MI, 48109, USA
c
Idaho National Laboratory, ID, 83402, USA
d
Department of Nuclear Engineering, Texas A&M University, TX, 77843, USA

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

Keywords: Crack growth behaviour of additively manufactured (AM) Alloy 800H were studied under cyclic and static load
Laser powder bed fusion additive in high-temperature water, focusing on heat treatment effects. Under cyclic loading, heat treatment gently
manufacturing affected corrosion fatigue. The growth rate is comparable to wrought 800H. As the load cycle frequency reduced
Alloy 800H
to below 0.001 Hz, AM 800H exhibited inconsistent stress corrosion cracking response as compared to wrought
Crack growth
Stress corrosion cracking
800H. Depending on heat treatment, inhomogeneous crack propagation and localized pinning were observed in
Corrosion fatigue stress-relieved and recrystallized AM 800H, attributed to carbides, microstructural inhomogeneity, and twin
Sensitization boundaries. Grain boundary sensitization at ̃10 wt% Cr caused fast crack growth.

1. Introduction been an on-going topic for three decades [10–12]. While SCC under the
static operation condition covers most reactor life, the occasional cyclic
Laser powder bed fusion (LPBF) represents the most widely used loading from mechanical and thermal fatigue can happen during the
metal additive manufacturing (AM) technology to fabricate complex reactor shut-down and re-start [13,14]. The SCC of austenitic alloys (e.g.
designs with customized structural and functional properties [1–4]. Its SS and nickel-based alloy/weld) in high-temperature water is predom­
increasing ability to produce high-quality end-use components has inantly intergranular (IG) cracking. With little variations in environ­
shifted its use from prototyping to large-scale manufactur­ mental and mechanical parameters during operation, the SCC
ing/production. A variety of structural alloys made by LPBF, including susceptibility of these alloys is affected by metallurgical contributions,
austenitic stainless steels (SSs) [5], nickel-based alloys [6,7], Co-Cr alloy more specifically alloy composition and microstructure. It is widely
[8], and Ti-6Al-4 V [9], have demonstrated good material density and accepted that higher Cr content in alloy chemistry can improve SCC
equivalent or better mechanical properties compared to their wrought susceptibility in high-temperature water [15,16]. However, compared
counterparts. Up to date, most applications of LPBF AM fall into the to the bulk alloy chemistry, the progressive understandings of SCC over
category of non-critical parts. After realizing the potential advantages of the past 20 years revealed the importance of the local chemical and
AM for reducing the manufacturing cost, shortening the deployment microstructural characteristics to the material’s EAC susceptibility [17,
schedule and condensing the supply chain, the nuclear industry is now 18], such as cold work (from bulk deformation, surface cold work, weld
exploring the possibility of employing AM to construct future reactors. residual strain, etc.) [15], grain boundary structure [19], grain boundary
The emergent needs of producing high-value nuclear-grade AM metal microchemistry (e.g. Si enrichment due to radiation [20], Cr depletion
components that are capable of servicing in light water reactors (LWRs) due to sensitization [17]), and precipitates [12]. It should be noted that
and future advanced reactors require better understandings of the ma­ carbide precipitation along the grain boundaries can have either a
terial’s properties in these conditions. positive impact (due to crack-tip blunting) [12,21] and a negative effect
Among many material properties that are critical to the safe opera­ (due to sensitization) [22] on SCC resistance. Since many reactor
tion of modern LWRs, environmentally-assisted cracking (EAC), structures were not designed for repair in their lifespan (60 years or
including stress corrosion cracking (SCC) and corrosion fatigue (CF), has more), the changes of grain boundary chemistry and structure over time

* Corresponding author.
E-mail address: xzl0092@auburn.edu (X. Lou).

https://doi.org/10.1016/j.corsci.2021.109739
Received 5 April 2021; Received in revised form 28 July 2021; Accepted 29 July 2021
Available online 31 July 2021
0010-938X/© 2021 Elsevier Ltd. All rights reserved.
J. Yang et al. Corrosion Science 191 (2021) 109739

can also affect the SCC behaviour of the alloys, mostly through grain annealing. While the annealing reduced AM 316 L’s pitting resistance
boundary chemical change [23] and migration [24]. For CF where cyclic [33], it improved its resistance to SCC [34].
stress is involved, the dependency of the crack growth behaviour on Alloy 800H (UNS N08810, also named as “800H” in this paper) is an
material factors relies on how much oxidation/corrosion can contribute, austenitic solid-solution alloy [41] with increased Cr and Ni contents
changing from the material/microstructure dependent crack growth over austenitic SSs. It is among the few structural alloys that are
behaviour at low ΔK and cyclic frequency to the mechanical dominant approved by ASEM Boiler & Pressure Vessel Code. It has been used
crack growth behaviour at high ΔK and cyclic frequency [25,26]. extensively in steam generator designs in pressurized water reactors
Due to the rapid cooling in the melt track, austenitic alloys produced (PWRs) and CANDU reactors [42,43]. Benefiting from its high carbon
by LPBF exhibit a far-from-equilibrium nature in microstructure and content, 800H can offer higher creep and rupture properties for
microchemistry heterogeneity [27]. In addition to the common metal­ high-temperature service. It is considered a bridging material between
lurgical defects, including the porosity from entrapped gas and austenitic SSs and high Ni alloys to balance the performance and cost,
lack-of-fusion, many microstructural characteristics that are less and a near-term candidate structural alloy for different GEN IV
observed in the conventional cast and wrought products often co-exist in advanced reactors [42]. However, this material was also known as
the AM materials, such as refined columnar grain structures [5], strongly susceptible to intergranular stress corrosion cracking (IGSCC) in reactor
textured grains [28], cellular dislocation sub-structures [29,30], nano­ environments [44,45]. Due to its inherent high carbon level, sensitiza­
scale second-phase particles (e.g., oxides) [31], high residual stress and tion (grain-boundary Cr depletion due to carbide formation) is expected
plastic strain [32]. Their effects on material properties, such as the in 800H at the temperature of 540–800 ℃. Although welding is the
uniaxial tensile and high-cycle fatigue properties, have been widely common cause of sensitization, carbide can nucleate and grow during
reported. In comparison, less work has been done to understand AM the service over a wide range of temperatures, even at low temperature
alloys’ corrosion behaviour, and most of which focused on AM 316 L SS (<400 ⁰C) (so-called “low-temperature sensitization”) [46]. It is
due to its broad applicability for corrosion resistance across industries. believed that the sensitized alloys are prone to IGSCC in
Nevertheless, AM alloys’ unique microstructures were confirmed to high-temperature water environments [22,47].
affect both pitting corrosion [33] and EAC [34,35], as summarized in This study evaluated the CF and SCC crack growth behaviour of AM
Table 1. Alloy 800H in high-temperature water and contributed to the emergent
Pre-existing porosity reduced re-passivation potential of AM 316 L SS needs for understanding the corrosion properties of AM alloys for nu­
in aqueous 3.5 wt% NaCl solution [36]. The negative impact from clear applications. By conducting different heat treatments on AM 800H,
porosity to pitting corrosion became more dominant with longer im­ including stress relief, sensitization, and hot isostatic pressing (HIP)
mersion. For SCC in high-temperature water, the negative effect of followed by solution annealing (SA), the work seeks to establish the
porosity in AM 316 L SS was found mediocre at high corrosion potential microstructure-EAC relationship for high-carbon austenitic alloys in
but more significant at low corrosion potential [33]. The textured col­ boiling water reactor (BWR) environments.
umnar/dendritic grains in AM SS induced the orientation-dependent
crack growth behaviour. Both SCC and CF cracks grew fastly along the 2. Experimental
build-up direction [34,35], but showed a tendency to arrest when ori­
ented perpendicular to the build-up direction. In comparison, pitting 2.1. Materials and heat treatment
behaviour in AM SS exhibited a similar trend along different orienta­
tions. Dislocation cellular structures increased the residual plastic strain Alloy 800H powder was produced by Carpenter Powder Products
and elevated the SCC growth rate in high-temperature water [34], (Bridgeville, PA, USA). The powder was sieved to the range of 15− 44 μm
consistent with the observation from the cold-worked material [38,39]. to ensure good flowability. AM 800H was fabricated using either
AM SS generally exhibits a higher oxygen level than the wrought ma­ Concept Laser Mlab Cusing LPBF AM system (90 W laser power, 600
terial. Nanoscale Si/Mn rich oxide inclusions were observed in the ma­ mm/s scan speed, 25 μm layer thickness, and 80 μm hatching distance)
terial [31,40]. For pitting, it was reported that the oxide formation or EOS M270 LPBF AM system (195 W laser power, 1200 mm/s scan
during AM trapped Mn and suppressed MnS precipitation, thus speed, 20 μm layer thickness, and 90 μm hatching distance). The ma­
increased pitting resistance [33]. However, in high-temperature water, terial produced by Concept Laser system was used for stress-relieved and
the presence of these oxides increased the local oxidation/dissolution, sensitization study, and the material produced by EOS was for the
and created severe crack branching [40]. Full recrystallization can be recrystallized material study by HIP and SA. This arrangement was
achieved on both AM SSs and Ni alloys through high-temperature mainly due to the limited accessibility to AM system. The printing pa­
rameters were chosen so that two materials showed a minimum differ­
ence in grain structure under as-built condition. We emphasize the study
Table 1 focused on establishing the microstructure-EAC relationship. Detailed
The effects of some basic microstructural characteristics in AM materials to characterization of microstructures at different length scale was pro­
pitting corrosion and SCC. vided. This arrangement did not affect this objective. More importantly,
Pitting Corrosion SCC the HIP + SA material exhibited a fully recrystallized grain structure,
Porosity Reduced re-passivation Increase CGRs in 2 ppm DO and which was quite different than stress-relieved and sensitized materials.
potential for long time 63 ppb DH [34] Thus, it can be treated as a separate study.
immersion [36] The laser parameters were selected to yield the best material density.
Cellular Decrease corrosion Increase CGRs in high- For the comparison purpose, the data of wrought 800H and AM 316 L
dislocations resistance [30] temperature water [34]
Oxides Increase pitting resistance Accelerate oxidation and cause
were also included in this manuscript, cited from Janssen et al. [17]’s
by reducing the formation of crack branching [37] and Lou et al. [34,35]’s work. The chemical compositions of the
other secondary particles, e. as-received powder and AM 800H were listed in Table 2.
g., MnS [33,37] As listed in Table 3, three heat treatment conditions were explored,
Grain structure Increase corrosion High CGR along build
including (1) stress relief (namely “SR 800H”), (2) HIP + SA (namely
resistance [33,36] direction, Low CGR ⊥ build
direction [34] “HIP 800H”), (3) sensitization (namely “sensitized 800H”). Per the
Recrystallization Decrease corrosion Similar CGR as wrought material specification of Special Metals®. Stress relief heat treatment
resistance [33,36] counterpart, minimal effect was done at 870 ℃ for 1 h, followed by air cooling [41]. Note that the
with partial uncrystallized stress-relief temperature of Alloy 800H is higher than that of 316 L SS
grains [34]
(̃650 ℃) [34]. HIP followed by SA treatment was performed to fully

2
J. Yang et al. Corrosion Science 191 (2021) 109739

Table 2
Chemical compositions of as-received, AM, and wrought 800H.
Element (wt%) Fe Cr Mn Si Ni Cu Al C P Ti S O (ppm)

AM Powder Bal. 21 0.88 0.62 34 <0.01 0.38 0.095 0.003 0.41 0.006 95
AM Part Bal. 21.4 0.87 0.63 32.5 <0.01 0.34 0.09 <0.005 0.42 0.003 200

2.2. CF and SCC crack growth test in BWR water environment

Table 3
A summary of heat treatment conditions and microstructural characteristics of
The specimens used for measuring crack growth rate (CGR) in high-
the AM Alloy 800H.
temperature water are 0.5-inch compact tension (CT) specimens with
Conditions Thermal Treatment Microstructure
5% side grooves on both sides (Fig. 1(a)), following ASTM-E647 [49].
Treatment 1 = 870 ℃ for 1 h in argon, followed Anisotropic, cellular For AM materials, the specimens were evaluated with two different
Stress relief by air cool dislocation structures, crack growth orientations, one is along the build direction (“Z” direction
carbides
in Fig. 1(b)), and the other one is perpendicular to the Z direction. The
Treatment 2 = Step 1: hot isostatic pressed for 4 Partial recrystallized, non-
HIP + SA h at 1204 ℃ and 14,500 psi in uniform and elongated notation “X-Z” indicates the load and crack propagation orientations,
argon; Step 2: solution annealing grains respectively.
at 1204 ℃ for 1 h, followed by The crack growth tests were performed in a 4-liter autoclave system
water quench
with a simulated BWR water environment at 288 ℃. The high-water
Treatment 3 = 621 ℃ for 24 h in vacuum, Anisotropic, cellular
Sensitization followed by air cool dislocation structures,
purity in the autoclave inlet (conductivity of ̃0.055 μS/cm at room
carbides temperature) was achieved by a closed-loop water system, which
employed a mixed bed demineralizer to purify the outlet water returning
from the autoclave. The NWC water with 2 ppm O2 was achieved by
recrystallize the material. HIP was done to densify the AM parts at 1204 continuously bubbling Ar-5% O2 mixture gas to the inlet water at room
℃ for 4 h under 14500 psi. After HIP, SA at 1204 ℃ for 1 h with water temperature.
quench was done to remove any possible carbides formed during the The in-situ monitoring of crack length was achieved by using the
slow HIP cooling. Compared to 316 L SS, Alloy 800H needs a higher direct current potential drop (DCPD) testing method. The specimen was
temperature for solution annealing [41]. AM 800H made by LPBF electrically insulated from the load frame and the load pins using the
required a high annealing temperature with an extended period. Thus, insulating washers and ceramic sleeves. The details of the DCPD set-up
1204 ℃ was selected for both HIP and SA to ensure full recrystallization. design can be found elsewhere [10,50]. An Agilent Technologies 6611C
Sensitization treatment was conducted at 621 ℃ for 24 h, followed by DC power supply was used to provide 2.5A current through the speci­
air cooling. This condition was picked based on the mens, and the current was reversed with̃0.5 Hz frequency to reduce the
temperature-time-sensitization (TTS) curve of wrought 800H [41] and thermocouple effects-related errors. During the tests, DCPD potentials,
our previous sensitization study of AM 800H, which corresponds to the along with some other signals (corrosion potentials, temperature, inlet
relatively higher susceptibility of sensitization (>25 mm/year) in the and outlet conductivities, and load.), were collected by a multiplexer
temperature range between 500 ℃ and 800 ℃. Cold work is known to and a digital voltmeter. The computer software automatically main­
increase steel’s SCC susceptibility. Cold forging is often applied as an tained a constant stress intensity factor by analyzing these signals [10,
accelerating method for testing material’s SCC susceptibility [12,34,48]. 47]. The averaged data was used to improve the measurement resolution
With a reasonable amount of work, the crack growth rate is more and reduce errors. Finally, the CGR was calibrated and corrected by
reproducible so that other variables can be studied. In this study, 20 % measuring the post-test fractography. The DCPD method can reach a
warm forging at 200 ⁰C (thickness reduction) was applied to HIP AM resolution as high as̃1 μm in crack length or̃10− 9 mm/s in CGR.
800H for the crack growth test. Prior to the tests, specimens need to be held in the testing environ­
ment for 24 h or longer to equilibrate the corrosion potential. Both CF
and SCC crack growth tests were carried out at Kmax = 27.5 MPa√m. For
each specimen, the test started with the cyclic loading to evaluate CF

Fig. 1. Schematic of (a) 0.5 compact-tension (CT) specimen used in this study. The measurements are in mm; (b) the orientations of CT specimens relative to powder
bed: “X” refers to the direction perpendicular to the material build direction of LPBF AM, and “Z” refers to the direction parallel to the material build direction of
LPBF AM. To designate the testing orientations, “Z-X” refers to the testing condition where the load is along “Z” direction while the crack advances along
“X” direction.

3
J. Yang et al. Corrosion Science 191 (2021) 109739

crack growth, and then transitioned to the static load SCC growth samples were prepared using the same metallographic procedures as the
evaluation. The K control with sine waveform was applied during the CF EBSD samples to ensure a flat surface without minimum roughness. A
CGR test. The low-frequency region of 10− 3 to 1 Hz was studied to LECO DM-400FT microhardness tester with Vickers indenter was used.
simulate the realistic reactor operation. The ΔK (Kmax-Kmin) was varied Each indentation was completed under 500 gf force for 15 s, and a total
from 11 to 22 MPa√m, and the load ratio R = Pmin/Pmax was changed of 10 indentations were measured from each sample.
from 0.2 to 0.6. All the specimens were tested in the same sequence,
transitioning from aggressive fatigue to mild fatigue conditions, to 3. Results and discussion
reduce the test errors between specimens. Sufficient crack extension was
allowed between loading variables to achieve a steady-state growth rate. 3.1. Microstructural evolution with the heat treatment
To ensure linear elastic fracture mechanics, the crack extension was
controlled within the required limit of ASTM E647 [49]. The Kmax and Fig. 2 shows the microstructures of AM 800H with different heat
ΔK values were calculated using the load and the measured crack length treatments. As-built (Fig. 2(a)), sensitized (Fig. 2(b)), and stress-relieved
derived from the DCPD. (Fig. 2(c)) AM 800H exhibited similar anisotropic grain structures that
Static-load SCC growth test was conducted immediately after the CF were oriented along the build direction, consistent with other reports
evaluation. This arrangement ensured the SCC test began with a crack [51–53]. The sensitization (621 ℃ for 24 h) and stress relief (870 ℃ for
front with sufficient depth and sharpness to establish the crevice 1 h) did not produce a visible change in grain structure. Large and
chemistry. A hold-time segment (0.001 Hz + 9000 s hold at Kmax) was discrete carbides were found along the HAGBs in the sensitized and
also be applied to facilitate the transition from transgranular (TG) CF stress-relieved samples, not in the as-built sample. Fine cellular/co­
crack to IG SCC crack. During SCC evaluation, the constant K of 27.5 lumnar dislocation structures were present inside the grains of as-built,
MPa√m was maintained by automatically lowering the load as the crack sensitized, and stress-relieved AM 800H. The cell spacing was̃300− 500
advanced. Besides the automatic load adjustment, the specimen was nm, while the cell length could reach up to ̃20 μm. The cellular dislo­
mostly tested under the static load condition. The duration of each test cation structures and elemental segregation on cellular walls were
condition was long enough so that the steady-state crack growth was commonly observed in AM materials and caused by fast solidification.
achieved. Each testing condition was repeated by either two separate The recent simulation confirmed that the cell spacing varied with alloy
specimens or two different measurements on the same specimen. composition and cooling rate [54]. It is worth noting that a noticeable
recovery of dislocation cellular structures happened in AM 316 L SS
2.3. Microstructure characterization made by LPBF at 800 ⁰C [55]. Thus, AM 800H by LPBF presented more
stable dislocation substructures than AM 316 L SS. Carbides were
The basic evaluation of microstructure and fractography were con­ observed in both sensitized and stress-relieved AM 800H. Detailed
ducted using a JEOL JSM-7000 F field emission scanning electron mi­ characterization of the carbides will be discussed later. After HIP and SA
croscope (SEM) with 20 kV voltage. Electron Back Scattered Diffraction at a high temperature (higher than typical SA temperature of wrought
(EBSD) was performed using an FEI Helios 650 system operating at a 800 H), recrystallization can be achieved but still with ̃13.2 % in area
voltage of 20 kV. EBSD scan was set to a scan area of 300 μm × 300 μm fraction of unrecrystallized region, as shown in Fig. 2(d). Different from
with 0.2 μm step size for SR AM 800H, and 1200 μm × 1200 μm with 0.5 AM 316 L SS [34,37,56], the recrystallized grains were not uniform in
μm step size for HIP AM 800H to accommodate different grain sizes. size across the material, with most grains slightly elongated along the
Samples were all prepared with the mechanical grinding up to 320-grit, build direction. Multiple SA conditions were studied for AM 800H, and
then followed by 0.9 μm, 0.3 μm, and 0.05 μm diamond or alumina the non-uniformity of grain structure persisted in AM 800H. The dislo­
polishing, and finalized with the vibration polishing with 0.03 μm cation substructures were removed by HIP and SA.
colloidal silica suspension. The electrolytic etching was done in 10 % Fig. 3(a) and (b) show the EBSD grain boundary maps for as-built and
oxalic acid for a few seconds to reveal grain/subgrain microstructures. HIP AM 800H, respectively. Most grain boundaries in the as-built ma­
Transmission electron microscopy (TEM) was used to identify the terial were HAGBs with the misorientation angles higher than 15 ◦ (̃91.6
carbides, characterize the carbides morphology and distribution, and % in total length). Note that the low-angle subgrain cellular boundaries
∑
measure the Cr depletion zone on high-angle grain boundaries (HAGBs) were not analyzed due to EBSD’s resolution. A low fraction of 3 twins
and cellular boundaries. The TEM lamellae were prepared from selected (̃2.3 %) was also observed. In contrast, HIP AM 800H exhibited a much
areas of interest using an FEI Quanta 3D focused ion beam/scanning higher fraction of twins (̃64.4 %), which was significantly higher than
electron microscope (FIB/SEM). After lift-out, the lamellae were the fully-recrystallized AM 316 L SS after HIP and SA [57]. It is worth
mounted onto the copper grids and thinned down to about 100 nm in noting that the fraction of the twin boundary in the total boundary
thickness using 30 keV Ga ions with decreasing current. The lamellae length in HIP AM 800H was approaching the grain boundary engi­
were then cleaned using 5 keV and 2 keV Ga ions for 1− 3 min on each neering (GBE) alloys [58–60]. Some works have shown that the IG
side. The TEM work was done with an FEI Talos F200X FEG STEM corrosion and IGSCC resistance can be improved by improving the
∑
machine at 200 kV voltage. The bright/dark field images and elemental fraction of 3 twins [19,61], while these special boundaries seem to
scans were conducted under the scanning-TEM (STEM) mode. The have minimal benefit to CF behaviour [61].
STEM-EDX mappings were done with 1024 by 1024 pixel size, 20 ms Fig. 3(c) and (d) showed the grain size distribution for as-built and
dwelling time, and zero degree tilt angle towards to the detector. The Cr HIP AM 800H. Both materials exhibited a large size variation, suggesting
depletion zone was measured by doing STEM-EDX linear scan across the the non-uniform microstructures. In the as-built material, the average
boundaries in the vicinity of a carbide. The carbide distance and the grain size was 17.83 ± 23.18 μm with an aspect ratio of 2.73 ± 2.3 (long
scanning line are maintained at̃50 nm on HAGBs and̃10 nm on cellular axis / short axis). The HIP material showed an average grain size of
boundaries. To get a representative Cr depletion zone profile for each 34.30 ± 40.68 μm with a 1.71 ± 1.86 aspect ratio.
condition, the EDX line scan was conducted near the carbides from at Fig. 3(e) and (f) show the inverse pole figure from the build direction
least three different grain boundaries. HRTEM and Fast Fourier Trans­ for the as-built and HIP AM 800H, respectively. The as-built material
form (FFT) were used to identify the type of carbides. FFTs and STEM- exhibited a strong texture of <001> along the build direction. < 001>
EDX were analyzed using Thermo Scientific Velox. texture was also found in LPBF AM 316 L SS reported by other re­
searchers [28], depending on the printing parameters and scanning
2.4. Hardness strategy. After the HIP treatment, a more random texture formed.

The microhardness test was conducted based on ASTM E384. The

4
J. Yang et al. Corrosion Science 191 (2021) 109739

Fig. 2. SEM secondary electron images of AM 800H: (a) as-built, cross-section parallel to the Z direction (build direction); (b) sensitized at 621 ℃ for 24 h, cross-
section parallel to the Z direction; (c) stress relief, cross-section parallel to the Z direction; (d) HIP + SA.

3.2. Carbide formation and sensitization is slightly higher than those in SR materials. However, the average
carbide size on the cellular boundary (35.83 ± 5.15 nm) was close to
In SR AM 800H, coarse and discrete Cr carbides were found on those in the SR material, suggesting the growth limit on cellular
HAGBs, with fine Ti carbides on cellular boundaries (Fig. 4). The boundaries was reached in both materials. Due to the slight misorien­
selected area diffraction patterns confirmed the Cr carbides on HAGB tation angle on the cellular boundaries (̃5− 7 ◦ ), the carbides were hard
were M23C6 Cr carbides (Fig. 5(a)), and those small carbides on cellular to grow [63]. In Fig. 6(c), Cr depletion appeared on both types of
boundaries were mainly MC Ti carbides (Fig. 5(b)). The average particle boundaries. More severe Cr depletion (̃10 wt% Cr) was measured from
size is 142.02 ± 58.53 nm for HAGB, and 36.70 ± 8.97 nm for the the HAGBs in the sensitized AM 800H. The full width at half maximum
cellular boundary. The large size variation of carbides on HAGB was (FWHM) of the depletion profile was smaller in the sensitized material
mainly due to the boundary misorientation angle. The boundary than in the SR material. Different from the SR material, a shallow Cr
misorientation with 3 ◦ to 15 ◦ is believed to be not as favourable as depletion was detected on the cellular boundaries in the sensitized
higher-angle boundaries to form large Cr carbides [62,63]. The linear material, with ̃16 wt% Cr in-depth, which was generally considered
number density of the HAGB particles is 2.38 ± 0.42 No./μm. immune to sensitization.
To quantify the Cr depletion level on boundaries, STEM-EDX line
scan was applied in the vicinity of a Cr carbide and crossed the boundary
3.3. Hardness
(Fig. 5(a) and (b)). The distance between carbide and the EDX line was
kept at 5̃0 nm on HAGBs and at 2 ̃0 nm on cellular boundaries. For each
Fig. 7 shows the room-temperature microhardness of AM 800H with
condition, over three different boundaries were measured to ensure a
different heat treatments. AM 316L SS under similar heat treatment
representative result. In SR 800H, Cr depletion was detected near the
conditions were also shown for comparison. The as-built AM 800H
HAGBs with ̃13 wt% Cr, while no depletion was found on the cellular
exhibited similar hardness as the as-built AM 316 L SS. After SR treat­
boundaries (Fig. 5(c)). This result suggests that the SR treatment can
ment, AM 800H showed a lower hardness than AM 316 L SS, primarily
result in carbide formation and Cr depletion on the HAGBs in AM 800H,
due to its more significant dislocation recovery at the higher SR tem­
even though the temperature is higher than the normal temperature
perature (870 ℃ for AM 800H vs. 650 ℃ for AM 316 L SS). With full
range for Cr carbide formation (̃500 to 800 ℃) [17].
recrystallization, HIP AM 800H showed the lowest hardness value,
In sensitized AM 800H, high-density M23C6 Cr carbides were found
comparable to HIP AM 316 L SS. The 20 % cold-rolled HIP materials
on both HAGBs and the cellular boundaries (Fig. 6(a) and (b)), which is
exhibited similar hardness values as their as-built conditions, suggesting
consistent with the observation that M23C6 is more stable than MC at the
a similar residual strain level.
temperature lower than 800 ℃ [64]. The selected area diffraction
pattern confirmed they were both M23C6 carbides with FCC structure.
The lattice constant is 10.8 Å, compared to 3.6 Å for the austenite ma­ 3.4. Effects of AM microstructure on the cyclic and static crack growth in
trix. The diffraction pattern also confirmed the FCC cube-on-cube AM 800H
coherent interface between the matrix and carbide on one side of the
boundary, consistent with other researchers’ observations [63,65]. The 3.4.1. Corrosion fatigue crack growth in high-temperature water
average carbide size on HAGBs is 75.03 ± 27.12 nm, half of the SR The CGRs as a function of ΔK for AM 800H with different crack
material carbide size. The carbide density is 3.44 ± 0.98 No./μm, which propagation directions are shown in Fig. 8(a). AM 316 L SS CGR data
were also included for comparison, which was obtained from the earlier

5
J. Yang et al. Corrosion Science 191 (2021) 109739

Fig. 3. EBSD analysis of AM Alloy 800H: (a) boundary misorientation map, as-built; (b) boundary misorientation map, HIP + SA; (c) grain size, as-built; (d) grain
size, HIP + SA; (e) inverse pole figure, as-built; (f) inverse pole figure, HIP + SA.

work [34,35]. All the tests were performed with a sine wave load cycling Pressure Vessel (BPV) code were also included in the plot for references.
at Kmax = 27.5 MPa√m under 0.5 Hz. The relationship between da/dN The ASME Section XI model (shown in Eq. (1)–(3)) has been used to
and ΔK can be described by Paris’ law, da/dN = CΔKn [66]. The “n” is reasonably describe the air fatigue crack growth of wrought austenitic
the slope of ln(da/dN) vs. ln(ΔK) and the “C” is an environmental SS for various conditions in the boiler and pressure vessel code. How­
related constant. For the SR AM 800H materials, “n” is around 1.2–1.4, ever, many researchers reported this model and its several revisions are
which is slightly lower than the range of 1.4–2.2 that researchers re­ non-conservative for evaluating fatigue crack growth in
ported from 316 L SS [26]. The CGRs increased with ΔK due to the high-temperature water [13,26,35,68].
increasing crack-tip strain rate [67]. For the SR AM 800H, X-Z direction /
da dNAir = C0 S(R)ΔK 3.3 (1)
was slightly faster than Z-X direction, suggesting that the crack grew
difficultly along the tortuous crack path through the close-packed
columnar grains where more boundary junctions were encountered log(C0 ) = − 8.714 + 1.34 × 10− 3 − 3.34 × 10− 6 T 2 + 5.95 × 10− 9 T 3 (2)
(Fig. 3(a)). A similar trend was also reported in the SR AM 316 L. In both
S(R) = 1 + 1.8R g for R < 0.79 (3)
crack orientations, SR AM 800H exhibited slower CGRs than AM 316 L
SS. American Society of Mechanical Engineers (ASME) Section XI curve da/dNAir is in mm/cycle, ΔK is in MP√m, and T is in ℃
for evaluating air fatigue CGR behaviour of austenitic SSs at 288 ℃ and Both SR AM 800H and SR AM 316 L SS showed higher CGRs than the
the Japanese Society of Mechanical Engineers (JSME) Boiler and ASME XI curve towards the lower ΔK, suggesting that the environmental

6
J. Yang et al. Corrosion Science 191 (2021) 109739

Fig. 4. STEM-EDX mappings showing the Cr carbides on HAGB and Ti carbides on cellular boundaries in SR AM800H.

Fig. 5. STEM bright field images and selected area diffraction patterns of carbides on (a) HAGBs; (b) cellular boundaries in SR AM 800H. (c) shows the Cr depletion
profile cross a HAGB and a cellular boundary collected by STEM-EDX in SR AM 800H.

effect increased when the strain amplitude decreased. In comparison, power-law relationship: da/dN = Cf m . For both SR AM 800H and SR AM
the CF model developed as JSME BPV code [69,70] (shown in Eq. (4)) 316 L SS, the CGRs increased as the load frequency decreased, indicating
provides a more conservative prediction, which all the CGR data points the increasing environmental effect. SR AM 800H exhibited lower CGRs
of these materials were below this line. than the SR AM 316 L SS in both orientations. Depending on the fre­
/ quency range, the CGR of AM 800H in X-Z orientation was 3–8 times
da dN = 8.17 × 10− 12 ΔtR0.5 (1 − R)− 2.12 ΔK 3.0 (4) higher than that in Z-X direction. Excluding the 0.001 Hz data of SR AM
800H Z-X, the m value of AM 800H along two orientations was slightly
da/dN is in mm/cycle, ΔK is in MP√m, and rise time ΔtR is in sec.
different. At 0.001 Hz, SR AM 800H along Z-X exhibited an unusual
Fig. 8(b) shows the effects of CGR as a function of load frequency.
increase in CGR, which will be discussed later.
The dependence of da/dN on loading frequency can be described by a
Fig. 9(a) shows the CF crack path of SR AM 800H tested along X-Z

7
J. Yang et al. Corrosion Science 191 (2021) 109739

Fig. 6. STEM bright field images and selected area diffraction patterns of Cr23C6 carbides on (a) HAGBs; (b) cellular boundaries in sensitized AM 800H. (c) shows the
Cr depletion profile cross a HAGB and a cellular boundary collected by STEM-EDX in sensitized AM 800H.

Fig. 7. Microhardness results for AM 800H and AM 316 L SS with different heat treatments.

orientation at 0.001 Hz. The crack was primarily TG with small sec­ behaviour in SR AM 800H could result in unexpected failure mode in AM
ondary TG cracks. The orientations of the columnar structure near the components.
crack path were marked in the red line in each grain. Most columns were GB particles, including carbides and intermetallic phases, were re­
aligned along the build direction (Z direction), as well as the crack path. ported to retard the growing crack in high-temperature water [71].
Thus, the less interaction with grain and columnar boundaries was However, in both X-Z and Z-X orientations, the CF crack was predomi­
responsible for the slightly less CF CGR along X-Z orientation. nantly transgranular. The contribution of carbides, both on GB or
Fig. 9(b) shows the detail of crack path at 0.001 Hz, and later tran­ cellular boundaries, to CF crack growth was insignificant.
sitioning to 9000 s hold time. Slower CGR along this orientation was
primarily due to the more frequent interaction with the grain and 3.4.2. Stress corrosion crack growth in high-temperature water
columnar boundaries. After encountering a high-density region, the Fig. 10 summarizes the CGRs of SR AM 800H, as compared to AM
crack slightly turned towards the build direction. The tendency of 316 L SS, under constant load/K at 27.5 MPa√m (no cyclic load) in
establishing faster CF CGR along the build direction explains the unusual NWC water. Due to the off-plane crack growth, the CGRs along Z-X
increase in CGR at 0.001 Hz along Z-X orientation in Fig. 8(b). After orientation are not discussed here. The SR AM 800H exhibited a lower
transitioning to 9000 s hold time, the crack growth direction was average CGR (1.43 × 10− 8 mm/s) than SR AM 316 L SS, with broader
completely reoriented along the build direction, growing outside of the data scattering in the range of 10-9–10− 8 mm/s. The crack propagation
original cracking plane. The off-plane crack growth behaviour affected was highly non-uniform in SR AM 800H, as shown in Fig. 11(a). The
DCPD, resulting in under-estimated CGR by DCPD. Similar phenomenon fractographic SEM image is shown in the crack front in Fig. 11(b),
was reported in AM 316 L SS as well [34]. The anisotropic cracking confirming the regional crack protrusion was IG cracking. Two attempts

8
J. Yang et al. Corrosion Science 191 (2021) 109739

Fig. 8. Effects of (a) ΔK; (b) loading frequency on corrosion fatigue crack growth rate of SR AM 800H and 316 L with different crack propagation orientations.

were made to transition the crack to IG by slowly reducing the aggres­ CW exhibited half of CGR than SR 800H tested along the X-Z orientation.
sivity of the fatigue condition to a constant load. Only a small portion of A similar trend was also observed by comparing HIP and SR AM 316 L
the crack front was transitioned to IG successfully, with most front SS, although AM 316 L showed higher CGR than AM 800H. The build
remaining TG. The observation suggests the SR AM 800H is more direction in SR AM materials was reported to accelerate fatigue crack
difficult to transition to IG SCC as compared to SR AM 316 L SS, and growth rate in the air [35], primarily attributed to its grain/columnar
requires a more careful assessment of its IG SCC CGR. We emphasize the orientation and high plastic strain. Through the hardness comparison,
protrusion region was located 4 mm away from the edge of the spec­ the 20 % forged HIP AM 800H possessed a similar amount of work as SR
imen, and the plane strain condition was maintained. DCPD response material. Thus, the microstructural anisotropy is likely the main reason
could not accurately predict the length of the local crack advance due to causing the different SCC growth susceptibility between HIP and SR
the averaging effect across the whole crack front. The under-estimated materials in high-temperature water.
crack length in the crack protrusion region resulted in a higher local K As the load frequency reduced, shown in Fig. 12(b), the environment
at the protruded crack tip than the pre-set value (27.5 MPa√m), thus contributed more to the crack propagation, in conjunction with the
higher CGR than other regions was sustained. cyclic loading. HIP AM 800H with 20 % CW exhibited the largest slope (|
It is worth noting that such non-uniform crack growth was not m| = 0.45) among all the materials. Although HIP AM 800H started with
observed in SR AM 316 L SS [34] tested under similar condition. Alloy the lowest CGR in the high-frequency region, it surpassed SR AM 800H
800H requires slightly higher SR heat treatment temperature than 316 L and HIP AM 316 L SS in CGR at 0.001 Hz, suggesting its higher envi­
SS (870 ⁰C for 800H vs. 650 ⁰C for 316 L SS). The hardness measurement ronment susceptibility, which could be due to the cold-work enhanced
in Fig. 7 showed the SR AM 800H was softer than SR AM 316 L SS, crack tip strain rate [12,35,72]. A similar cold work effect on
suggesting a more dislocation recovery in SR AM 800H. It was reported low-frequency CF was also found from AM 316 L SS. Overall, the AM
the residual plastic strain is responsible for the high SCC susceptibility in 800H materials exhibited better CF performance than AM 316 L SSs.
AM material [34]. Thus, the reduced plastic strain in AM 800H by SR
heat treatment contributed to its less susceptibility to IG SCC. However, 3.5.2. Stress corrosion crack growth
since the recrystallization did not take place in SR AM 800H, the strain Fig. 13 shows the SCC CGR of HIP AM 800H under constant load/K =
recovery was not uniform across the material. It is reasonable to expect 27.5 MPa√m. SR AM 800H and wrought 800H (from Janssen’s work
the presence of slightly higher strain region in the material promoted the [17]) along with AM 316 L SS variants were included for comparison.
local IG SCC crack protrusion. Besides plastic strain non-uniformity, GB Similar to the SR AM 800H, the measured CGR of HIP AM 800H with 20
carbides in SR AM 800H could also made the IG SCC transition harder. % CW also varied across a wide range from 10− 9 to 10− 7 mm/s. The data
Carbides were reported to arrest IG crack growth by blunting the crack were averaged at a ̃1 × 10-7 mm/s. Compared to SR AM 800H tested
tip and re-distributing the crack-tip strain field [12,22]. In a highly along X-Z orientation, HIP AM 800H with 20 % CW showed higher SCC
heterogeneous AM microstructure, carbides result in unexpected crack CGR under constant load/K condition. GB carbides in SR AM 800H could
arresting in the crack front. The contributions of intergranular particles be the reason for its slower CGR. Overall, AM 800H showed less SCC
on the SCC growth in AM 800H require further investigation. susceptibility than AM 316 L SS under constant load/K. The crack front
in HIP AM 800H, shown in Fig. 14(a), was highly non-uniform. The
3.5. Effects of recrystallization on the cyclic and static crack growth in crack pinning took place on one side of the specimen. A similar
AM 800H non-uniform crack front was often seen in materials with low SCC sus­
ceptibility or non-uniform grain structure like pressure vessel low alloy
3.5.1. Corrosion fatigue crack growth steels [73] or weld metals [67]. Fig. 14(b) shows the close look of the
Similar to AM 316L SS fabricated by LPBF [34], AM 800H can be crack pinning region of the HIP AM 800H, revealing that the crack front
recrystallized to form equiaxed grains when annealed at1204 ℃. After was initially pinned by several large grains. Most of the SCC region,
HIP + SA (Fig. 2(d)), AM 800H showed complete recrystallization with shown in Fig. 14(c), presented a mixture of IG and TG cracking. The
slightly elongated grains. Fig. 12(a) shows the CF CGRs of HIP AM 800H grain topology on the fracture surface in HIP AM 800H was not as well
as a function of ΔK. SR AM 800H and AM 316 L SS are included for distinguishable as those in the wrought 316 L SS and 800H reported in
comparison. Cold/warm forging was commonly applied to fully the literature. Grain inhomogeneity and high twin fraction may
annealed austenitic materials as an accelerating method for evaluating contribute to the TG crack propagation [58].
SCC susceptibility. With cold work, the CGRs of different materials can
be compared with reasonable reproducibility. HIP AM 800H with 20 %

9
J. Yang et al. Corrosion Science 191 (2021) 109739

Fig. 10. SCC growth rate of SR AM 800H and AM 316 L X-Z in 288 ℃ NWC
water with 2 ppm DO. AM 316L SS data is adopted from literature for com­
parison purpose.

GB Cr level for IG SCC in AM 800H can be slightly below 12 wt% in high

temperature water.
In Fig. 16, the CF crack path at 0.001 Hz in the sensitized AM 800H
exhibited a mixture of IG and TG cracking, representing the highest CGR
measured in this study. The orientation of cellular boundaries and inter-
cellular carbides showed minimal effect on the TG cracking.

3.6.2. Stress corrosion crack growth

Fig. 17 summarized the SCC CGR of the sensitized AM 800H, along
with other forms of Alloy 800H, under constant load/K in the oxygen­
ated high-temperature water. The sensitized AM 800H exhibited fast
CGR at the average of̃5.7 × 10− 6 mm/s, 2–3 orders of magnitude higher
than the SR material. Compared with the SR AM 800H that is primarily
dominated by TG SCC (Fig. 11(a)), the sensitized AM 800H revealed IG
crack path with several secondary IG crack branches (Fig. 18(a)), sug­
gesting the high susceptibility of HAGBs. The crack front of the sensi­
tized sample was uniform, and the CGR was highly reproducible. Fig. 18
(b) shows the morphology of the SCC fracture in the sensitized AM
800H. The fracture was predominantly intergranular. Fig. 18(c) shows
the closely spaced lines were mildly etched out on the exposed grains
Fig. 9. (a) CF crack path SEM secondary electron image of SR AM 800H X-Z at during the crack propagation. The careful examination confirmed that
0.001 Hz; (b) Cross-sectional images of the crack path in the etched micro­ this feature was related to the cellular/columnar structures in AM 800H.
structure of the SR AM 800H Z-X. The crack turned 90 ◦ once transferred the Slight Cr depletion (̃16 wt% Cr) was observed along the cellular
load mode from CF to SCC. The red lines represent the local directions of boundaries, as shown in Fig. 6(c). In general, Cr depletion at this level is
cellular structures. (For interpretation of the references to colour in this figure not considered susceptible to corrosion. Thus, severe intercellular
legend, the reader is referred to the web version of this article). dissolution was not observed.

3.6. Carbide, sensitization and their impacts on the cyclic and static crack
3.7. Further discussion on the roles of microstructures on the
growth behaviour
environmental cracking susceptibility
3.6.1. Corrosion fatigue crack growth
The foregoing results showed the crack growth responses of AM
Fig. 15(a) shows the fatigue CGRs of the sensitized AM 800H X-Z
800H tested varied depending on heat treatment, load condition, and
tested along X-Z orientation, along with HIP and SR AM 800H. The
crack orientation. No significant difference in CGR was shown among
sensitized material exhibited similar CGR like the SR AM 800H tested
stress-relieved, sensitized, and HIP’ed materials under cyclic load when
along the same orientation, suggesting the sensitization had a negligible
ΔK is above 10 MPa√m and the loading frequency is above 0.01 Hz. As
effect on the CF crack growth. However, as the load frequency reduced
the load frequency further dropped to below 0.001 Hz with the intro­
(Fig. 15(b)), the difference in CGR between sensitized and SR AM 800H
duction of hold time, as shown in Fig. 14(b), microstructural factors,
enlarged due to the increasing environmental effects. The sensitized and
such as the presence of GB carbides, sensitization, grain inhomogeneity,
SR AM 800H both suffered from Cr depletion at HAGBs, with the GB Cr
and GB type, contributed to the CGR differences among AM 800H under
level of̃10 wt% and 13 wt%, respectively. It is commonly agreed that the
different heat treatments.
threshold Cr level at GBs that may lead to intergranular corrosion is
Compared to SR AM 316 L SS, SR AM 800H presented a considerably
̃12− 13 wt% for austenitic SS [22]. This threshold value can vary with
slower SCC growth rate and broader CGR data scattering. While their
alloy composition, environment, cold work, etc. Considering SR AM
slight differences in grain inhomogeneity and residual plastic strain
800H experienced the more difficult transition to IG SCC, the threshold
existed, we believe carbides in SR AM 800H contributed to its hesitant

10
J. Yang et al. Corrosion Science 191 (2021) 109739

Fig. 11. (a) Optical fractography of the whole fracture surface; (b) CF and SCC fractographic SEM secondary electron image of SR AM 800H X-Z.

Fig. 12. Effects of (a) ΔK; (b) loading frequency on corrosion fatigue crack growth rate of SR/HIP AM 800H and AM 316L SS. AM 316L SS data is adopted from
literature for comparison purpose.

800H. This conclusion was also backed by the evidence (Fig. 11(a) and
(b)) that the SCC in SR AM 800H did not follow the HAGBs and pre­
sented transgranular characteristics.
There is no question GB sensitization can increase the CGR of AM
800H in high-temperature water. With 13 wt% Cr on GB in SR AM 800H,
TG cracking was still predominant. Thus, we confirmed the threshold Cr
level for sensitization was̃10 wt% for AM 800H. Fig. 19 shows the high-
magnification SEM image of an active crack tip. SEM-EDS confirmed the
presence of carbides at the crack opening. The crack path followed one
side of the carbides along the GB, leaving carbides mostly attached to the
other half of the boundary. It is known that M23C6 carbides nucleate
coherently to the matrix on GBs, and grow only towards one side of the
boundary with the same lattice orientation to its parent grain. Cr
depletion was present at the growing interface. The pinning effect of
carbides to crack growth is no longer effective under this circumstance.
We conclude the grain inhomogeneity and high percentage of twins
in HIP AM Alloy 800H with 20 % CW contributed to its non-uniform
crack growth and the larger CGR data variation. Grain inhomogeneity
is known to induce inconsistent crack path, resulting in non-uniform
Fig. 13. SCC growth rate of SR, HIP, and wrought 800H in 288 ℃ high tem­
crack front or crack arrest. This phenomenon is common in weld
perature water with 2 ppm DO. AM 316 L counterparts were also included as
metals or alloys that experience abnormal grain growth. HIP AM 800H
a comparison.
exhibited a large standard deviation in grain size (34.30 ± 40.68 μm),
which could produce a similar effect on SCC growth. More importantly,
crack propagation, resulting in a non-uniform crack front and enlarged
the high fraction of twins (̃64.4 %) was measured in the recrystallized
the data variation. Grain boundary carbide has long been found to affect
AM 800H after HIP and SA. We emphasize the material is generally
the SCC resistance in austenitic stainless steels or Ni base alloys by
considered as GBE material at such a high twin fraction [60]. Note AM
obstructing the emergence of the slip steps [17,21,47,74]. The length
316 L SS never achieved a high amount of twins even with the intention
percentage of carbide coverage on HAGB was calculated at ̃57 % in SR
to do GBE [57]. Extensive work has been done to understand the effects

11
J. Yang et al. Corrosion Science 191 (2021) 109739

Fig. 14. SCC fractographic images under constant load of HIP + SA AM 800H X-Z with 20 % CW tested in high temperature water. (a) optical fractography of the
whole fracture surface; (b) SEM secondary electron (SE) image of the crack pinning region; (c) representative SEM SE image from the SCC region.

Fig. 15. Effects of (a) ΔK; (b) loading frequency on corrosion fatigue crack growth rate in AM Alloy 800H with different heat treatments.

of GBE on material properties related to grain boundaries [58,75]. While whose severity generally increases as grain size decreases, also changes
most works agreed that GBE greatly enhances IG corrosion resistance of GB misorientation and affects IG SCC. Fig. 14(b) identified large grains
the austenitic alloys, its effects on IG SCC are mixed. GBE was less made the crack transition from TG to IG more difficult. At the same time,
effective in reducing IG cracking propensity when cold work was IG SCC fracture showed less distinguishable GB topology, with mixed TG
involved. The grain size was identified as dominating the IG cracking in characteristics. The grain inhomogeneity and twins should work
GBE austenitic SS and Ni alloy with cold work, primarily due to its in­ simultaneously on the cracking susceptibility. Their exact roles in sup­
verse relationship to GB triple junctions. Grain rotation under cold work, pressing or promoting crack advance under low-frequency load cycles

12
J. Yang et al. Corrosion Science 191 (2021) 109739

Fig. 16. CF crack path SEM secondary electron image of sensitized AM 800H X-
Z at 0.001 Hz in high temperature water (2 ppm DO). The red lines show the
local orientations of cellular structures. (For interpretation of the references to
colour in this figure legend, the reader is referred to the web version of
this article).

Fig. 17. SCC growth rate of AM and wrought 800H with different heat treat­
ments in NWC (2 ppm DO). Wrought 800H data is adopted from literature for
comparison purpose.

require further study.

4. Conclusions

This study evaluated the CF and SCC growth behaviour of LPBF

manufactured AM 800H in BWR NWC water, focusing on three heat
treatment conditions: stress relief, HIP + SA, and sensitization. Fig. 18. (a) SCC crack path SEM secondary electron (SE) image of sensitized
Depending on the heat treatment, AM 800H exhibited distinguishably AM 800H X-Z; (b) and (c) SCC fractographic SEM SE images of sensitized AM
different CF and SCC behaviour than wrought 800H [17] and AM 316 L 800H tested in high temperature water with 2 ppm DO. The material is sus­
SS [34,35] that were tested under the same condition, largely due to the ceptible to IGSCC.
microstructural variations in grain structure, carbides, sensitization, and
twin fraction. The differences in CGR among materials as well as the Different from AM 316 L SS, we conclude that AM 800H needs to be
data scattering with the same material were enlarged as the aggressivity used with caution even after HIP and SA, due to its inconsistent cracking
of the cyclic loading reduced to low frequency or even constant load. behaviour and high sensitivity to the local microstructural in­
The CF CGR under ΔK ≥ 11 MPa√m, frequency ≥ 0.01 Hz followed a homogeneity. Detailed findings are summarized as follows:
ranking of sensitized X-Z > SR X-Z > HIP + SA with 20 % CW > SR Z-X.
As the cyclic loading reduced to below 0.001 Hz or constant load, the (a) As-built AM 800H exhibited an elongated non-equilibrium grain
rank of CGR changed to Sensitized X-Z > HIP + SA with 20 % CW > SR structure with fine cellular/columnar structures and <001>
X-Z > SR Z-X. crystallographic texture. High SR temperature for 800H resulted
ASME Section XI code for wrought austenitic SS does not bound the in the partial recovery of the cellular dislocations in AM 800H,
CF crack growth data for AM 800H under all tested conditions. JSME but still preserved most grain structures. High-temperature HIP +
BPV code is conservative for both SR and recrystallized AM 800H, but SA completely recrystallized AM 800H. The recrystallized AM
still does not bound the sensitized AM 800H. 800H exhibited slightly elongated grains, large grain size

13
J. Yang et al. Corrosion Science 191 (2021) 109739

Fig. 19. SCC crack tip SEM secondary electron image of sensitized AM 800H X-Z tested in 288 ℃ high temperature water with 2 ppm DO.

variation, and a high fraction of annealing twins, which can be Author statement
considered as GBE.
(b) Stress relief of AM 800H produced M23C6 Cr carbides on both Jingfan Yang: Methodology, Validation, Investigation, Writing-
HAGBs and cellular boundaries, and fine MC Ti carbides on Original Draft, Visualization; Miao Song: Methodology, Investigation,
cellular boundaries. Sensitization only resulted in M23C6 Cr car­ Visualization; Laura Hawkins: Methodology, Investigation, Visualiza­
bides on both HAGBs and cellular boundaries. The HAGB car­ tion; Xiang Liu: Methodology, Investigation, Visualization; Lingfeng
bides in SR material are twice larger than the sensitized material, He: Methodology, Investigation, Visualization; Xiaoyuan Lou:
but with lower density. The size and density of cellular boundary Conceptualization, Methodology, Validation, Investigation, Writing-
carbides are similar in the two materials. Review&Editing, Project administration, Funding acquisition.
(c) For the SR AM 800H, the build direction is the preferred crack
path. The crack could not grow in perpendicular to the build Data availability
direction as the load condition reduced to ΔK = 11 MPa√m,
0.001 Hz, or constant load, primarily due to more frequent The raw/processed data required to reproduce these findings cannot
intersection with grain and cellular boundaries. The presence of be shared at this time. Data sharing requires the permissions from the
HAGB carbides was considered as the cause of large variations in research sponsors.
SCC CGR and non-uniform crack growth under constant load, as
compared to AM 316 L SS. Declaration of Competing Interest
(d) The transition to IG SCC was more difficult in SR AM 800H, as
compared to sensitized or recrystallized AM 800H. Localized No conflict of interest has been identified for this work.
protrusion of IG SCC was observed in SR AM 800H.
(e) With recrystallization, HIP AM 800H still showed non-uniform
Acknowledgments
crack growth and large CGR data scattering. The fracture was
primarily IG, with some mixed TG characteristics. The high
This research is primarily supported by the Idaho National Lab
amount of twins in HIP AM 800H was ineffective in mitigating
Laboratory Directed Research & Development (LDRD) Program under
crack propagation. However, the high twin fraction and grain
the US Department of Energy Idaho Operations Office contract DE-
inhomogeneity in HIP AM 800H can contribute to the CGR
AC07-05ID14517. The purchase of Alloy 800H powder and the earlier
variation, difficult local IG SCC transition, and non-uniform crack
material fabrication/evaluation was supported by the Nuclear Energy
front.
Enabling Technologies (NEET) program, the US Department of Energy
(f) The degree of sensitization in AM 800H that leads to fast IG SCC
under contract DE-NE0008428. The work was also partially supported
in high-temperature water was identified to be ̃10 wt% Cr.
by the National Institute of Standards and Technology (US) under con­
Sensitized AM 800H exhibited the fastest IG crack growth among
tract NIST-70NANB18H220. The authors gratefully acknowledge Dr.
all tested materials.
Bart Prorok and Dr. Pu Deng’s help with LPBF AM at Auburn University.
(g) Although cellular boundaries in AM 800H experienced minor Cr
depletion (̃16 wt% Cr) by sensitization, the cellular boundary
References
sensitization did not affect both CF and SCC behaviour of AM
800H in high-temperature water. [1] W.E. Frazier, Metal additive manufacturing: a review, J. Mater. Eng. Perform. 23
(2014) 1917–1928, https://doi.org/10.1007/s11665-014-0958-z.

14
J. Yang et al. Corrosion Science 191 (2021) 109739

[2] N. Guo, M.C. Leu, Additive manufacturing: technology, applications and research stainless steels: roles of heavy element segregation and stacking fault energy, Jom
needs, Front. Mech. Eng. 8 (2013) 215–243, https://doi.org/10.1007/s11465-013- (2020), https://doi.org/10.1007/s11837-020-04427-7.
0248-8. [30] J.R. Trelewicz, G.P. Halada, O.K. Donaldson, G. Manogharan, Microstructure and
[3] L.E. Murr, S.M. Gaytan, D.A. Ramirez, E. Martinez, J. Hernandez, K.N. Amato, P. corrosion resistance of laser additively manufactured 316L stainless steel, Jom 68
W. Shindo, F.R. Medina, R.B. Wicker, Metal fabrication by additive manufacturing (2016) 850–859, https://doi.org/10.1007/s11837-016-1822-4.
using laser and Electron beam melting technologies, J. Mater. Sci. Technol. 28 [31] P. Deng, M. Karadge, R.B. Rebak, V.K. Gupta, B.C. Prorok, X. Lou, The origin and
(2012) 1–14, https://doi.org/10.1016/S1005-0302(12)60016-4. formation of oxygen inclusions in austenitic stainless steels manufactured by laser
[4] D. Herzog, V. Seyda, E. Wycisk, C. Emmelmann, Additive manufacturing of metals, powder bed fusion, Addit. Manuf. 35 (2020) 101334, https://doi.org/10.1016/j.
Acta Mater. 117 (2016) 371–392, https://doi.org/10.1016/j.actamat.2016.07.019. addma.2020.101334.
[5] K. Saeidi, Stainless Steels Fabricated by Laser Melting: Scaled-down Structural [32] X.X. Zhang, H. Andrä, S. Harjo, W. Gong, T. Kawasaki, A. Lutz, M. Lahres,
Hierarchies and Microstructural Heterogeneities, Stockholm University, 2016. Quantifying internal strains, stresses, and dislocation density in additively
[6] K.N. Amato, S.M. Gaytan, L.E. Murr, E. Martinez, P.W. Shindo, J. Hernandez, manufactured AlSi10Mg during loading-unloading-reloading deformation, Mater.
S. Collins, F. Medina, Microstructures and mechanical behavior of Inconel 718 Des. 198 (2021) 109339, https://doi.org/10.1016/j.matdes.2020.109339.
fabricated by selective laser melting, Acta Mater. 60 (2012) 2229–2239, https:// [33] Q. Chao, V. Cruz, S. Thomas, N. Birbilis, P. Collins, A. Taylor, P.D. Hodgson,
doi.org/10.1016/j.actamat.2011.12.032. D. Fabijanic, On the enhanced corrosion resistance of a selective laser melted
[7] A. Mostafaei, E.L. Stevens, E.T. Hughes, S.D. Biery, C. Hilla, M. Chmielus, Powder austenitic stainless steel, Scr. Mater. 141 (2017) 94–98, https://doi.org/10.1016/j.
bed binder jet printed alloy 625: densification, microstructure and mechanical scriptamat.2017.07.037.
properties, Mater. Des. 108 (2016) 126–135, https://doi.org/10.1016/j. [34] X. Lou, M. Song, P.W. Emigh, M.A. Othon, P.L. Andresen, On the stress corrosion
matdes.2016.06.067. crack growth behaviour in high temperature water of 316L stainless steel made by
[8] B. Qian, K. Saeidi, L. Kvetková, F. Lofaj, C. Xiao, Z. Shen, Defects-tolerant Co-Cr-Mo laser powder bed fusion additive manufacturing, Corros. Sci. 128 (2017) 140–153,
dental alloys prepared by selective laser melting, Dent. Mater. 31 (2015) https://doi.org/10.1016/j.corsci.2017.09.017.
1435–1444, https://doi.org/10.1016/j.dental.2015.09.003. [35] X. Lou, M.A. Othon, R.B. Rebak, Corrosion fatigue crack growth of laser additively-
[9] L. Thijs, F. Verhaeghe, T. Craeghs, J. Van Humbeeck, J.P. Kruth, A study of the manufactured 316L stainless steel in high temperature water, Corros. Sci. 127
microstructural evolution during selective laser melting of Ti-6Al-4V, Acta Mater. (2017) 120–130, https://doi.org/10.1016/j.corsci.2017.08.023.
58 (2010) 3303–3312, https://doi.org/10.1016/j.actamat.2010.02.004. [36] M. Kazemipour, M. Mohammadi, E. Mfoumou, A.M. Nasiri, Microstructure and
[10] P.L. Andresen, C.L. Briant, Environmentally assisted cracking of types 304L/316L/ corrosion characteristics of selective laser-melted 316L stainless steel: the impact of
316NG stainless steel in 288◦ C water, Corrosion 45 (1989) 448–463, https://doi. process-induced porosities, Jom. 71 (2019) 3230–3240, https://doi.org/10.1007/
org/10.5006/1.3582044. s11837-019-03647-w.
[11] P.L. Andresen, Stress Corrosion Cracking (SCC) of Austenitic Stainless Steels in [37] X. Lou, P.L. Andresen, R.B. Rebak, Oxide inclusions in laser additive manufactured
High Temperature Light Water Reactor (LWR) Environments, Woodhead stainless steel and their effects on impact toughness and stress corrosion cracking
Publishing, 2010, https://doi.org/10.1533/9781845699956.2.236. behavior, J. Nucl. Mater. 499 (2018) 182–190, https://doi.org/10.1016/j.
[12] P.L. Andresen, M.M. Morra, IGSCC of non-sensitized stainless steels in high jnucmat.2017.11.036.
temperature water, J. Nucl. Mater. 383 (2008) 97–111, https://doi.org/10.1016/j. [38] J. Kuniya, I. Masaoka, R. Sasaki, Effect of cold work on the stress corrosion
jnucmat.2008.08.005. cracking of nonsensitized aisi 304 stainless steel in high-temperature oxygenated
[13] H.P. Seifert, S. Ritter, H.J. Leber, Corrosion fatigue initiation and short crack water, Corrosion. 44 (1988) 21–28, https://doi.org/10.5006/1.3582020.
growth behaviour of austenitic stainless steels under light water reactor conditions, [39] S.K. Pradhan, P. Bhuyan, S. Mandal, Influence of the individual microstructural
Corros. Sci. 59 (2012) 20–34, https://doi.org/10.1016/j.corsci.2012.02.008. features on pitting corrosion in type 304 austenitic stainless steel, Corros. Sci. 158
[14] H.P. Seifert, S. Ritter, Environmentally-assisted cracking in austenitic light Water (2019) 108091, https://doi.org/10.1016/j.corsci.2019.108091.
reactor structural materials. Final Report of the KORA-I Project, 2009. [40] X. Lou, P.L. Andresen, R.B. Rebak, Oxide inclusions in laser additive manufactured
[15] W.C. Moshier, Effect of cold work and processing orientation on stress corrosion stainless steel and their effects on impact toughness and stress corrosion cracking
cracking behavior of alloy 600, Corrosion 56 (2000) 307–320, https://doi.org/ behavior, J. Nucl. Mater. 499 (2018) 182–190, https://doi.org/10.1016/j.
10.5006/1.3287659. jnucmat.2017.11.036.
[16] K. Arioka, T. Yamada, T. Miyamoto, T. Terachi, Dependence of stress corrosion [41] INCOLOY Alloy 800. & 800HT, 2004. https://www.specialmetals.com/.
cracking of Alloy 690 on temperature, cold work, and carbide precipitation - role of [42] W. Ren, R. Swindeman, Status of alloy 800 H in considerations for the gen IV
diffusion of vacancies at crack tips, Corrosion 67 (2011) 1–18, https://doi.org/ nuclear energy systems, J. Press. Vessel Technol. 136 (2014) 054001, https://doi.
10.5006/1.3563640. org/10.1115/1.4025093.
[17] J.S. Janssen, M.M. Morra, D.J. Lewis, Sensitization and stress corrosion cracking of [43] M.D. Pandey, S. Datla, R.L. Tapping, Y.C. Lu, The estimation of lifetime distribution
alloy 800H in high-temperature water, Corrosion 65 (2009) 67–78, https://doi. of Alloy 800 steam generator tubing, Nucl. Eng. Des. 239 (2009) 1862–1869,
org/10.5006/1.3319121. https://doi.org/10.1016/j.nucengdes.2009.05.027.
[18] K. Arioka, T. Yamada, T. Terachi, G. Chiba, Influence of carbide precipitation and [44] S.Y. Persaud, A.G. Carcea, J. Huang, A. Korinek, G.A. Botton, R.C. Newman,
rolling direction on intergranular stress corrosion cracking of austenitic stainless Analytical electron microscopy of a crack tip extracted from a stressed Alloy 800
steels in hydrogenated high-temperature water, Corrosion 62 (2006) 568–575, sample exposed to an acid sulfate environment, Micron 61 (2014) 62–69, https://
https://doi.org/10.5006/1.3280670. doi.org/10.1016/j.micron.2014.02.011.
[19] V.Y. Gertsman, S.M. Bruemmer, Study of grain boundary character along [45] K. Arioka, T. Yamada, T. Miyamoto, M. Aoki, Intergranular stress corrosion
intergranular stress corrosion crack paths in austenitic alloys, Acta Mater. 49 cracking growth behavior of Ni-Cr-Fe alloys in pressurized water reactor primary
(2001) 1589–1598, https://doi.org/10.1016/S1359-6454(01)00064-7. water, Corrosion 70 (2014) 695–707, https://doi.org/10.5006/1205.
[20] A. Etienne, B. Radiguet, N.J. Cunningham, G.R. Odette, P. Pareige, Atomic scale [46] M.J. Povich, Low temperature sensitization of type 304 stainless steel, Corrosion
investigation of radiation-induced segregation in austenitic stainless steels, J. Nucl. 34 (1978) 60–65, https://doi.org/10.5006/0010-9312-34.2.60.
Mater. 406 (2010) 244–250, https://doi.org/10.1016/j.jnucmat.2010.08.043. [47] P.L. Andresen, Environmentally assisted growth rate response of nonsensitized aisi
[21] D. Van Rooyen, Review of the stress corrosion cracking of inconel 600, Corrosion 316 grade stainless steels in high temperature water, Corrosion 44 (1988)
31 (1975) 327–337. 450–460, https://doi.org/10.5006/1.3583961.
[22] S.M. Bruemmer, B.W. Arey, L.A. Charlot, Influence of chromium depletion on [48] Y. Murase, J. Nagakawa, N. Yamamoto, Fracture surface topography analysis of in-
intergranular stress corrosion cracking of 304 stainless steel, Corrosion 48 (1992) beam fatigue behavior for 20% cold-worked 316 stainless steel, J. Nucl. Mater. 322
42–49, https://doi.org/10.5006/1.3315917. (2003) 249–254, https://doi.org/10.1016/j.jnucmat.2003.08.027.
[23] Q. Peng, J. Hou, Y. Takeda, T. Shoji, Effect of chemical composition on grain [49] A.S.T.M. E647− 13, Standard test method for measurement of fatigue crack growth
boundary microchemistry and stress corrosion cracking in Alloy 182, Corros. Sci. rates, Am. Soc. Test. Mater. (2014) 1–50, https://doi.org/10.1520/E0647-13A.2.
67 (2013) 91–99, https://doi.org/10.1016/j.corsci.2012.10.012. [50] P.L. Andresen, SKI Report 98 : 27 IGSCC Crack Propagation Rate Measurement in
[24] L. Volpe, M.G. Burke, F. Scenini, Correlation between grain boundary migration BWR Environments Executive Summary of a Round Robin Study SKI Report 98 : 27
and stress corrosion cracking of alloy 600 in hydrogenated steam, Acta Mater. 186 IGSCC Crack Propagation Rate Measurement in BWR Environments Executive
(2020) 454–466, https://doi.org/10.1016/j.actamat.2020.01.020. Summary of a Round Robin Study, (n.d.).
[25] H.P. Seifert, S. Ritter, H.J. Leber, Corrosion fatigue initiation and short crack [51] K. Saeidi, X. Gao, F. Lofaj, L. Kvetková, Z.J. Shen, Transformation of austenite to
growth behaviour of austenitic stainless steels under light water reactor conditions, duplex austenite-ferrite assembly in annealed stainless steel 316L consolidated by
Corros. Sci. 59 (2012) 20–34, https://doi.org/10.1016/j.corsci.2012.02.008. laser melting, J. Alloys. Compd. 633 (2015) 463–469, https://doi.org/10.1016/j.
[26] D. Du, J. Wang, K. Chen, L. Zhang, P.L. Andresen, Environmentally assisted jallcom.2015.01.249.
cracking of forged 316LN stainless steel and its weld in high temperature water, [52] M.L. Montero-Sistiaga, M. Godino-Martinez, K. Boschmans, J.P. Kruth, J. Van
Corros. Sci. 147 (2019) 69–80, https://doi.org/10.1016/j.corsci.2018.10.032. Humbeeck, K. Vanmeensel, Microstructure evolution of 316L produced by HP-SLM
[27] T. DebRoy, H.L. Wei, J.S. Zuback, T. Mukherjee, J.W. Elmer, J.O. Milewski, A. (high power selective laser melting), Addit. Manuf. 23 (2018) 402–410, https://
M. Beese, A. Wilson-Heid, A. De, W. Zhang, Additive manufacturing of metallic doi.org/10.1016/j.addma.2018.08.028.
components – process, structure and properties, Prog. Mater. Sci. 92 (2018) [53] J. Suryawanshi, K.G. Prashanth, U. Ramamurty, Mechanical behavior of selective
112–224, https://doi.org/10.1016/j.pmatsci.2017.10.001. laser melted 316L stainless steel, Mater. Sci. Eng. A 696 (2017) 113–121, https://
[28] T. Niendorf, S. Leuders, A. Riemer, H.A. Richard, T. Tröster, D. Schwarze, Highly doi.org/10.1016/j.msea.2017.04.058.
anisotropic steel processed by selective laser melting, Metall. Mater. Trans. B [54] T. Pinomaa, M. Lindroos, M. Walbrühl, N. Provatas, A. Laukkanen, The significance
Process Metall. Mater. Process. Sci. 44 (2013) 794–796, https://doi.org/10.1007/ of spatial length scales and solute segregation in strengthening rapid solidification
s11663-013-9875-z. microstructures of 316L stainless steel, Acta Mater. 184 (2020) 1–16, https://doi.
[29] P. Deng, H. Yin, M. Song, D. Li, Y. Zheng, B.C. Prorok, X. Lou, On the thermal org/10.1016/j.actamat.2019.10.044.
stability of dislocation cellular structures in additively manufactured austenitic

15
J. Yang et al. Corrosion Science 191 (2021) 109739

[55] H. Yin, M. Song, P. Deng, L. Li, B.C. Prorok, X. Lou, Thermal stability and in austenite stainless steel, Scr. Mater. 65 (2011) 509–512, https://doi.org/
microstructural evolution of additively manufactured 316L stainless steel by laser 10.1016/j.scriptamat.2011.06.010.
powder bed fusion at 500–800 ℃, Addit. Manuf. 41 (2021) https://doi.org/ [66] N. Pugno, M. Ciavarella, P. Cornetti, A. Carpinteri, A generalized Paris’ law for
10.1016/j.addma.2021.101981. fatigue crack growth, J. Mech. Phys. Solids 54 (2006) 1333–1349, https://doi.org/
[56] M. Song, M. Wang, X. Lou, R.B. Rebak, G.S. Was, Radiation damage and 10.1016/j.jmps.2006.01.007.
irradiation-assisted stress corrosion cracking of additively manufactured 316L [67] L. Dong, E.H. Han, Q. Peng, W. Ke, L. Wang, Environmentally assisted crack growth
stainless steels, J. Nucl. Mater. 513 (2019) 33–44, https://doi.org/10.1016/j. in 308L stainless steel weld metal in simulated primary water, Corros. Sci. 117
jnucmat.2018.10.044. (2017) 1–10, https://doi.org/10.1016/j.corsci.2016.12.011.
[57] S. Gao, Z. Hu, M. Duchamp, P.S.S.R. Krishnan, S. Tekumalla, X. Song, M. Seita, [68] O.K. Chopra, W.J. Shack, A review of the effects of coolant environments on the
Recrystallization-based grain boundary engineering of 316L stainless steel fatigue life of LWR structural materials, J. Press. Vessel Technol. Trans. ASME 131
produced via selective laser melting, Acta Mater. 200 (2020) 366–377, https://doi. (2009), https://doi.org/10.1115/1.3027496.
org/10.1016/j.actamat.2020.09.015. [69] H. Kobayashi, K. Kashima, Overview of JSME flaw evaluation code for nuclear
[58] E.A. West, G.S. Was, IGSCC of grain boundary engineered 316L and 690 in power plants, Int. J. Press. Vessel. Pip. 77 (2000) 937–944, https://doi.org/
supercritical water, J. Nucl. Mater. 392 (2009) 264–271, https://doi.org/10.1016/ 10.1016/S0308-0161(01)00016-3.
j.jnucmat.2009.03.008. [70] M. Itatani, M. Asano, M. Kikuchi, S. Suzuki, K. Iida, Fatigue crack growth curve for
[59] M. Shimada, H. Kokawa, Z.J. Wang, Y.S. Sato, I. Karibe, Optimization of grain austenitic stainless steels in bwr environment, J. Press. Vessel Technol. Trans.
boundary character distribution for intergranular corrosion resistant 304 stainless ASME 123 (2001) 166–172, https://doi.org/10.1115/1.1358841.
steel by twin-induced grain boundary engineering, Acta Mater. 50 (2002) [71] K. Arioka, T. Yamada, T. Terachi, G. Chiba, Influence of carbide precipitation and
2331–2341, https://doi.org/10.1016/S1359-6454(02)00064-2. rolling direction on intergranular stress corrosion cracking of austenitic stainless
[60] M. Michiuchi, H. Kokawa, Z.J. Wang, Y.S. Sato, K. Sakai, Twin-induced grain steels in hydrogenated high-temperature water, Corrosion 62 (2006) 568–575,
boundary engineering for 316 austenitic stainless steel, Acta Mater. 54 (2006) https://doi.org/10.5006/1.3280670.
5179–5184, https://doi.org/10.1016/j.actamat.2006.06.030. [72] C. Hahin, R.M. Stoss, B.H. Nelson, P.J. Reucroft, Effect of cold work on the
[61] J. Gao, J. Tan, X. Wu, S. Xia, Effect of grain boundary engineering on corrosion corrosion resistance of nonsensitized austenitic stainless steels in nitric acid,
fatigue behavior of 316LN stainless steel in borated and lithiated high-temperature Corrosion 32 (1976) 229–238, https://doi.org/10.5006/0010-9312-32.6.229.
water, Corros. Sci. 152 (2019) 190–201, https://doi.org/10.1016/j. [73] X. Lou, P.L. Andresen, J. Yang, R. Pathania, T. Lian, R.G. Carter, Mechanical and
corsci.2019.01.036. metallurgical considerations on the effects of ppb-level chloride on stress corrosion
[62] B. Tang, L. Jiang, R. Hu, Q. Li, Correlation between grain boundary misorientation cracking of low alloy steels in high-temperature water, Corros. Sci. 179 (2021)
and M23C 6 precipitation behaviors in a wrought Ni-based superalloy, Mater. 109136, https://doi.org/10.1016/j.corsci.2020.109136.
Charact. 78 (2013) 144–150, https://doi.org/10.1016/j.matchar.2013.02.006. [74] W. Kuang, G.S. Was, The effects of grain boundary carbide density and strain rate
[63] E.A. Trillo, L.E. Murr, A TEM investigation of M23C6 carbide precipitation on the stress corrosion cracking behavior of cold rolled Alloy 690, Corros. Sci. 97
behaviour on varying grain boundary misorientations in 304 stainless steels, (2015) 107–114, https://doi.org/10.1016/j.corsci.2015.04.020.
J. Mater. Sci. 33 (1998) 1263–1271, https://doi.org/10.1023/A:1004390029071. [75] M. Shimada, H. Kokawa, Z.J. Wang, Y.S. Sato, I. Karibe, Optimization of grain
[64] R.G. Faulkner, J. Caisley, The kinetics of grain boundary carbide precipitation in boundary character distribution for intergranular corrosion resistant 304 stainless
Nimonic PE16, Met. Sci. 11 (1977) 200–207, https://doi.org/10.1179/ steel by twin-induced grain boundary engineering, Acta Mater. 50 (2002)
msc.1977.11.6.200. 2331–2341, https://doi.org/10.1016/S1359-6454(02)00064-2.
[65] K. Kaneko, T. Fukunaga, K. Yamada, N. Nakada, M. Kikuchi, Z. Saghi, J.S. Barnard,
P.A. Midgley, Formation of M23C6-type precipitates and chromium-depleted zones

16