d-Ferrite Formation and Its Effect on the Mechanical

Properties of Heavy-Section AISI 316 Stainless Steel

Casting
QIYU WANG, SHENGHU CHEN, and LIJIAN RONG

d-Ferrite formation across the thickness of a heavy-section AISI 316 casting slab and the effect
of its decomposition on the bending properties have been studied. Cellular-shaped d-ferrite with
the faction of less than 3 pct is observed at the surface position of the cast while skeletal-shaped
d-ferrite with the faction of up to 8 pct is present at the center position, which is attributed to the
combined effects of macro-segregation and cooling rate. Extensive decompositions of d-ferrite
through direct transformation (d fi r) and eutectoid transformation (d fi r + c2) are
demonstrated. Micro-cracks could be observed within the decomposed d-ferrite on the convex
surface of bending samples from both the surface and center position of the cast, but occurrence
of cracking is only present in the sample from center position of the cast. Removal of the brittle
r phase within d-ferrite by a preheat treatment at 1050 C could effectively improve the bending
performance.

https://doi.org/10.1007/s11661-020-05717-0
 The Minerals, Metals & Materials Society and ASM International 2020

I. INTRODUCTION ductility,[13,14] creep properties,[15] and intergranular

corrosion resistance.[16,17]
AUSTENITIC stainless steels are widely used in Afterwards, grade 316 austenitic stainless steels are
components designed for high-temperature applications developed for the high-temperature application by the
because of their superior mechanical properties and proper balance of ferrite/austenite stabilizing elements.
excellent corrosion resistance.[1–6] High contents of Cr They are predicted to be fully austenitic microstructure
and Mo are responsible for the favorable high-temper- according to the Schaeffer diagram.[18,19] Unfortunately,
ature properties of austenitic stainless steel, but simul- the residual d-ferrite could still be observed in these
taneously increase the propensity for the formation of steels,[20] and a higher amount of d-ferrite has been
d-ferrite. Although a certain amount of d-ferrite (5 to 8 found in the type 316L steel in comparison with the type
vol. pct) is deliberately introduced into weld metals of 316H steels,[21–23] which is due to its higher Creq/Nieq
austenitic stainless steel to prevent hot cracking during value. Besides, different amounts of d-ferrite were
welding,[7,8] existence of d-ferrite in the base metal would introduced by different processing methods in the grade
deteriorate the mechanical properties, especially at 316 stainless steel of similar composition. A wide range
temperature above 550 C due to the decomposition of of solidification cooling rate for different processing
d-ferrite.[9,10] Investigations showed that d-ferrite could methods might affect the solidification condition, i.e., a
decompose into M23C6 carbides, r phase, v phase, and g higher cooling rate will result in microstructures that are
phase when exposed to temperature in the range of 500 far from equilibrium, and complex morphologies of the
C to 900 C, and the presence of these phases was residual d-ferrite would be developed. For the heavy-sec-
detrimental to impact toughness,[11,12] tensile tion casting component, the total solidification time and
cooling rate are different at the different positions across
the thickness, which might affect the microstructural
development and the residual d-ferrite characters during
solidification. Furthermore, it is noteworthy that it
QIYU WANG is with the CAS Key Laboratory of Nuclear Materials would be so hard to completely remove the residual
and Safety Assessment, Institute of Metal Research, Chinese Academy
of Sciences, Shenyang 110016, China and also with the School of d-ferrite through solid solution treatment.[24] Therefore,
Materials Science and Engineering, University of Science and elucidating the d-ferrite formation mechanism in the
Technology of China, Hefei 230026, China; SHENGHU CHEN and heavy-section cast is vital to propose methods to remove
LIJIAN RONG are with the CAS Key Laboratory of Nuclear Materials the residual d-ferrite.
and Safety Assessment, Institute of Metal Research, Chinese Academy
of Sciences. Contact e-mail: [email protected], [email protected]
In this study, the fraction and morphologies of
Manuscript submitted November 8, 2019. d-ferrite across the thickness of a heavy-section AISI
316 casting slab have been investigated. Furthermore,

METALLURGICAL AND MATERIALS TRANSACTIONS A

formation mechanism of d-ferrite and its effect on the diffraction (EBSD) maps are acquired by the attached
bending properties are provided here. Oxford NordlysNano EBSD detector to evaluate the
strain distribution after deformation.

II. EXPERIMENTAL
Slab with a thickness of 180 mm was produced by III. RESULTS AND DISCUSSION
continuous casting and finished with air-cooling to room
A. Distribution of d-Ferrite in Cast Slab and Its
temperature. Chemical compositions are present in
Formation Mechanism
Table I. In order to reveal the distribution of d-ferrite
across the slab thickness, 18 samples were finely sliced Figure 1(a) presents the d-ferrite fraction across the
across the thickness, and space interval of each sample thickness of cast slab. Apparently, d-ferrite fraction at
was set as 10 mm. Besides, samples near the centerline, the center position is much higher than that at the
marked with micro-hardness, were sealed in the tube surface, showing that d-ferrite fraction is less than 3 pct
first and then heat-treated at 1050 C for 1 hour, near the surface up to a depth of 70 mm and then
followed by water quenching to observe the dissolution increases sharply to 8 pct within 20 mm in the center.
behavior of r phase in d-ferrite. Furthermore, in order Corresponding OM images are displayed in Figures 1(b)
to reveal the critical precipitation temperature of r and (c), which correspond to the surface and center
phase in d-ferrite, these samples after eliminating r position of the cast, respectively. It is found that
phase were sealed in different tubes and heat-treated at d-ferrite mainly locates at the interdendritic regions
900 C, 950 C, 1000 C, and 1030 C for 1 hour, and cellular-shaped d-ferrite is dominant at surface
respectively, which are followed by water quenching. position of the cast (Figure 1(b)). However, prominent
The oxygen-free atmosphere in the tube is used to dendritic morphologies could hardly be observed at
prevent the oxidation of the sample, so that the center position of the cast (Figure 1(c)). Contrary to the
observation on the dissolution and re-precipitation of morphology at surface position, d-ferrite here appears in
r phase at the same position becomes possible. The aim a skeletal shape and is interlinked with each other to
of water quenching process is to avoid the occurrence of present continuous morphologies. It is found that
re-precipitation during cooling process. significant differences in the morphologies and amounts
Samples (15 9 10 9 10 mm) for metallographic of d-ferrite occur from the surface to the center of the
observation here were mechanically ground with SiC cast.
abrasive papers up to 2000 grit, polished with 0.5 lm Further observations were performed by EPMA to
diamond powder, and electrolytically etched in a 65 pct reveal the element distributions at surface position of
HNO3 solution with 1.2 V for about 5 seconds to show cast, as shown in Figure 2. The dendritic structures
up d-ferrite. The easily etched d-ferrite appears darker could be observed, and Cr and Mo, which act as ferrite
colored than the austenite by OM observation stabilizers, would segregate towards the interdendritic
(Figures 1(b) and (c)). The distribution of d-ferrite was regions due to their solute partition coefficients of k < 1.
observed by optical microscopy (OM, Olympus GX51). In addition, d-ferrite is found exactly in the interden-
Among the methods for measurement of d-ferrite dritic regions with the high content of Cr and Mo, while
fractions, image analysis is considered to be one of the the content of Ni in d-ferrite is depleted. EPMA results
most accurate methods.[25] Micrographs in this study suggest that micro-segregation of ferrite-forming ele-
were captured by optical microscopy at 9 100 magni- ment Cr and Mo during solidification is responsible for
fication to clearly resolve the microstructure, and at least the formation of d-ferrite.
5 fields (1.42 9 1.06 mm each) were recorded to increase In addition, chips are obtained from the surface and
the accuracy of measurements. Image Pro Plus software center of the cast to determine the compositions by
was used to calculate the fraction of the specific area and chemical analysis, and the results are summarized in
the average fraction is obtained. Furthermore, detailed Table II. It shows that concentrations of Si, C, and Mn
microstructural observations were performed by a are similar at both the surface and center position, while
scanning electron microscope (SEM, Zeiss Merlin Com- variation in concentrations of Cr, Ni, and Mo is
pact FE-SEM), and phase identifications were further significant. It is known that Cr and Mo are the
performed with attached energy-dispersive X-ray spec- ferrite-stabilizing elements while Ni is the austenite
trometry (EDS, Oxford X-MaxN). Segregation of solute stabilizer, and the ratio Creq/Nieq would affect the
elements was evaluated by Electron probe microanalysis solidification mode. In accordance with the formula of
(EPMA-1610). In addition, thin foils were prepared by a Cr and Ni equivalents[20,26]:
twin-jet electropolishing apparatus in an electrolyte of
10 pct HClO4 solution at 20 V and  20 C and Creq ¼ pct Cr þ 1:5 ðpct SiÞ þ pct Mo þ 0:5 ðpct NbÞÞ
examined on transmission electron microscopy (TEM, ½1
Tecnai G2 F30) at 200 kV. EDS line analyses were
performed to study the elemental distribution using
Nieq ¼ pct Ni þ 0:5 ðpct MnÞ þ 30 ðpct CÞ ½2
drift-correcting programs (TIA, FEI). Specimens with
the dimension of 80 9 10 9 3 mm were cut from
different positions of slab and machined subsequently The ratio of Creq/Nieq at the surface and center
for three-point bending tests, and electron backscatter position of the cast can be estimated as 1.45 and 1.52,

METALLURGICAL AND MATERIALS TRANSACTIONS A

 Table I. Chemical Composition of AISI 316 Stainless Steel

Element C N Si Ni Cr Mo Mn S P
Weight Percent 0.039 0.075 0.38 12.4 17.57 2.55 1.46 < 0.001 0.013

Fig. 1—Distribution of d-ferrite across the slab thickness: (a) area faction, (b, c) OM images at the surface and center position, respectively.

respectively, indicating an increase in the Creq/Nieq ratio estimated that the cooling rate decreases from 1.3 C/s
from surface towards the center. According to the near the surface to 0.2 C/s at the center of the cast
vertical section of the Fe-Cr-Ni ternary diagram at a (Figure 3). It is found that the amount of d-ferrite is not
constant of ~ 70 wt pct Fe, alloy with Creq/Nieq ratio affected by a change in the cooling rate from 1.3 C/s
above 1.48 would exhibit a change in solidification near the surface to 0.2 C/s up to a depth of 70 mm,
mode.[20] In the present work, sample at the surface with which corresponds to the microstructures developing
a Creq/Nieq ratio of 1.45, solidifies as primary austenite from AF mode. In the AF solidification mode, positive
with secondary d-ferrite (AF mode) (Figure 1(b)), while micro-segregation of Cr and Mo at the interdendritic
sample at the center position with a Creq/Nieq ratio of regions changes the local compositions (Creq/Nieq) and
1.52 solidifies as primary d-ferrite with secondary induces the cellular d-ferrite formation during solidifi-
austenite (FA mode) (Figure 1(c)). Therefore, a change cation (Figure 1(b)). Thus, a change of austenite den-
in solidification mode from AF mode at the surface to drite morphologies caused by the reduction of cooling
FA mode at the center is present as the Creq/Nieq ratio rate has little effect on the amount of d-ferrite. However,
increases from 1.45 to 1.52. macro-segregation of the solute caused by the different
What is more, a large difference in the secondary cooling rates (from 1.3 C/s at surface to 0.2 C/s at the
dendrite arm spacing (k) is observed in across the depth of 70 mm) during solidification increases the Creq/
thickness of cast slab, which reveals changes in the Nieq at center position and induces the transition to FA
solidification cooling rate (e) according to the equation mode. A higher amount of d-ferrite could solidify from
below[27,28]: the melts primarily and a relatively slow cooling rate
increases the time available for the solidification reac-
k ¼ 44e0:38 ½3 tion. Therefore, combined effects of macro-segregation
and cooling rate are responsible for the significant
Based on the measured secondary dendrite arm differences in the morphologies and amounts of d-ferrite
spacing (k) across the thickness of cast slab, it is across the thickness of cast slab.

METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 2—EPMA micrographs of dendrite structure at surface position of the cast: (a) second electron (SE) image, (b to d) elemental distribution
maps of Cr, Mo and Ni.

Table II. Chemical Compositions at the Surface and Center shows a case of precipitates appear at the d/c interface,
Position of Cast (Weight Percent) and corresponding EDS line-scan analysis reveals that
the newly formed precipitates are enriched with Cr and
Element Cr Si Ni C Mn Mo Creq/Nieq Mo in comparison with the d-ferrite (Figure 4(c)). In
Surface 17.62 0.37 12.4 0.048 1.41 2.87 1.45 addition, complete decomposition of d-ferrite could be
Center 17.80 0.37 11.9 0.042 1.40 2.67 1.52 found, and the co-existed products in the interior of
d-ferrite exhibit a eutectoid structure (Figure 4(d)). EDS
point analysis is conducted to determine the chemical
composition of precipitates, as shown in Figure 4(e). It
is depicted that the compositions of precipitate at the
B. Decomposition Mechanism of d-Ferrite marked position #1 are much closer to that in the
Magnified SEM examinations of d-ferrite at the austenite matrix (marked position #3), regardless of the
surface and center position reveal the extensive decom- slight difference in Fe and Ni content, so that this kind
position of d-ferrite in the cast (Figure 4). As can be of precipitate is considered as regenerated austenite.
seen, precipitates could be observed in both of the Moreover, precipitate at marked position #2 has higher
cellular-shaped d-ferrite at the surface of the cast content of Cr and Mo and lower content of Ni than the
(Figure 4(a)) and the skeletal-shaped d-ferrite at the regenerated austenite, which is close to the compositions
center of the cast (Figures 4(b) and (d)). Figure 4(b) of r phase as reported elsewhere.[29,30]

METALLURGICAL AND MATERIALS TRANSACTIONS A

 more, a higher concentration of Cr (~ 27 wt pct) is
found in the residual d-ferrite prone to direct transfor-
mation compared to that (~ 25 wt pct) in the residual
d-ferrite prone to eutectoid transformation. It is
reported that Cr content had an important effect on
the accelerated precipitation of the r phase.[32,33]
CALPHAD-based thermodynamic calculation also
showed that residual ferrite was unstable at high
temperature due to its high Cr value, and the decom-
position mechanism of d-ferrite depended on the chem-
ical compositions of d-ferrite-stabilizing elements in
316Nb stainless steel, displaying that a critical content
of Cr of ~ 27 wt pct in d-ferrite is required for the
occurrence of direct transformation of d-ferrite during
thermal treatment.[29] Several experimental studies also
provided favorable results for the present work.[34,35] As
a consequence, eutectoid transformation of d-ferrite
takes place for the d-ferrite with the insufficient Cr
Fig. 3—Combined effects of cooling rate and solidification mode on concentration of ~ 25 wt pct (Figure 5(d)).
the d-ferrite content across the thickness of the cast.

C. Bending Behaviors of the Cast

TEM samples are obtained from the center position In order to reveal the variation of mechanical
(with 8 pct d-ferrite) of the cast to explore the decom- properties in the thickness direction, bending samples
position mechanism of d-ferrite, and two different are extracted from the surface and center of the cast for
morphologies of d-ferrite are found, as shown in 180 deg bending tests, as shown in Figure 6. No cracks
Figure 5. Apparently, residual d-ferrite still could be are detected on the convex surface of bending samples
found, indicating a non-completed decomposition pro- from surface position of the cast (Figure 6(a)), while
cess. In Figure 5(a), only one type of precipitate is found numerous cracks are found on the bending samples
at the d/c interface, which is induced by a direct from center position of the cast (Figure 6(b)). In
transformation from d-ferrite. Selected area electron addition, detailed observations on the microstructure
diffraction (SAED) pattern of the new precipitate beneath the convex surface are conducted to reveal the
presents a tetragonal structure (inserted in cracking behaviors. As can be seen, no cracks are
Figure 5(a)), which is further identified as r phase detected on the bending samples from surface position
(a = b = 0.88 nm, c = 0.46 nm). Another kind of of the cast under OM observation (Figure 6(c)), but
d-ferrite decomposition is characterized by a eutectoid numerous micro-cracks within decomposed d-ferrite
transformation at the d/c interface, and the identified could be found by higher magnified SEM observation
pattern of r phase is also attached in Figure 5(c). Based (Figure 6(e)). By contrast, obvious cracks extending
on the above SEM and TEM observations, two different along the continuous d-ferrite network is found on the
decomposition mechanisms of d-ferrite are present in the bending samples from center position of the cast under
heavy-section cast: direct transformation (d fi r) and OM observation (Figure 6(d)), and micro-cracks within
eutectoid transformation (d fi r + c2). decomposed d-ferrite could also be observed
It is reported that the precipitation of r phase in (Figure 6(f)).
d-ferrite was much faster than in c-austenite in austenitic Micro-cracks could be observed within the decom-
stainless steels,[31] which is consistent with the preferen- posed d-ferrite on the convex surface of bending samples
tial precipitation of r phase within d-ferrite in the from the surface position and center position of the cast,
present work. In order to further elucidate the decom- but occurrence of cracking is only present in the sample
position mechanisms of d-ferrite, compositional profiles from center position of the cast. Compared with the
of the residual d-ferrite adjacent to the decomposed austenite and residual d-ferrite, it is known that r phase
d-ferrite are determined by TEM-EDS line-scan analysis exhibits a lower deformability,[36,37] thus plastic defor-
(Figures 5(b) and (d)). Steady-state concentrations of mation preferentially occurs in the regenerated austenite
Cr, Ni, and Mo are present in the residual d-ferrite and residual d-ferrite during bending deformation. The
adjacent to the r phase produced from direct transfor- strain incompatibility would be resulted at the r phase/
mation (Figure 5(b)), while fluctuations in Cr and Mo austenite or r phase/d-ferrite interface, which is evalu-
concentration are observed in the residual d-ferrite ated by the EBSD local misorientation component.[38–40]
adjacent to the eutectoid structure produced from It is found that higher strain concentrations are present
eutectoid transformation (Figure 5(d)). The residual at the r phase/austenite interface (Figure 7(b)).
d-ferrite adjacent to the r phase precipitates contains Accordingly, micro-cracks are initiated at these inter-
higher concentrations of Cr and Mo than those adjacent faces. Once the local stress exceeds the critical shear
to the regenerated austenite, demonstrating the parti- strength of the brittle r phase, micro-cracks would
tioning of Cr and Mo between r phase and regenerated extend along the r phase, as shown in the schematic
austenite during eutectoid transformation. What is illustrations of Figures 8(a) and (c). The presence of

METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 4—SEM morphologies of d-ferrite at (a) surface and (b, d) center position of the cast. (c) EDS line analysis marked in (b), (e) EDS point
analysis at positions marked in (d).

micro-cracks in the decomposed d-ferrite could affect the suffered maximum stress of the bending sample from
the suffered maximum stress (rmax) at this region during center position is significantly larger than that of the
bending test, which can be expressed in terms of the bending sample at surface position according to Eq. [4],
applied force (F), the area of cross-section of the which would result in a higher crack propagation rate in
specimen (A0), and the total volume fraction of the the bending sample at center position. What is more,
micro-cracks (f) [41]: continuous d-ferrite network in the center position of
cast is favorable for the crack propagation (Figure 8(d)).
KF
rmax ¼ ½4 By contrast, discontinuous cellular-shaped d-ferrite at
A0 ð1  fÞ the surface position of cast is much more resistant to the
crack propagation during bending deformation
where K is the stress concentration factor of the
(Figure 8(b)).
micro-cracks, which is related to the size and shape of
the cracks, and is considered to be a constant value here.
In this case, the fraction of micro-cracks in individual
decomposed d-ferrite is related to the fraction of D. The Dissolution of r Phase and the Improvement
d-ferrite. In the heavy-section cast, a higher fraction of of Bending Performance
d-ferrite at center position of the cast results in a higher Since r phase formation in d-ferrite is detrimental to
fraction of micro-cracks during bending test. As a result, the bending properties of cast, a proper heat treatment is

METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 5—TEM images showing decomposition behaviors involving (a) decomposition of d-ferrite into r phase and (c) decomposition of d-ferrite
into r phase and regenerated austenite at the center of the cast. (b) EDS analysis marked in (a) and (d) EDS analysis marked in (c). (Inserted
electron diffraction patterns refer to the r phase.).

needed to modify the microstructures. Based on the chemical composition of the reappeared precipitates
thermodynamic calculations by Thermos-Calc soft- within d-ferrite (marked #1 position in Figure 9(d)) is
ware (2019) with TCFE7 database, the solubility tem- very close to the EDS results of r phase obtained from
perature of r phase is close to 900 C. Thus, 1050 C was the center position of cast (Figure 4(d)). In addition, the
chosen as the heat treatment temperature to eliminate residual d-ferrite, which presents lower contents of Cr
the r phase. In order to reveal the dissolution behaviors and Mo than the r phase is observed (marked #2
of r phase, the phase transformation within d-ferrite at position in Figure 9(d)). The verification experiments on
the same position before and after heat treatment at the dissolution and re-precipitation of r phase within
1050 C for 1 hour is conducted on the samples at the d-ferrite through heat treatment processes demonstrate
center position of the cast, as shown in Figures 9(a) and that decomposed d-ferrite present in the heavy-section
(b), where micro-hardness marks allow the determina- cast is generated during the slow cooling stage after
tion of the location of d-ferrite. It can be seen that solidification. Investigations also found that r phase
precipitates in d-ferrite have been eliminated with a could be formed quickly in d-ferrite due to the high
slight reduction in the size of d-ferrite. diffusion rate of Cr in d-ferrite[31]
In addition, samples after the removal of r phase were Based on the dissolution experiment of r phase,
reheated at different temperature for 1 hour to reappear samples extracted from the center position of the cast
the precipitation of r phase. After reheating at 900 C, with 1050 C heating are prepared to re-evaluate the
950 C, 1000 C, and 1030 C, respectively, precipitates bending properties. Interestingly, suppression of crack-
within d-ferrite would reappear when temperature falls ing on the convex surface is achieved after heat
below 1000 C, as shown in Figures 9(c) and (d). treatment, as shown in Figure 10(a). Microstructure
Considering the spatial resolution of SEM-EDS mea- observations beneath the convex surface indicate the
surements is ~ 1 lm, chemical compositions of the removal of r phase and disappearance of micro-cracks
reappeared precipitates within d-ferrite could be deter- within d-ferrite (Figures 10(b) and (c)). Therefore,
mined by point EDS analysis. It is found that the removal of the brittle r phase within d-ferrite by a

METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 6—Observation of convex surface of samples extracted from (a), (c), (e) surface and (b), (d), (f) center position of the cast after bending test.

Fig. 7—(a) EBSD phase identification map (red, blue, yellow, white area refer to austenite, r phase, d-ferrite and micro-cracks, respectively), and
(b) local misorientation map (phase boundaries are depicted with black line) in the center position of the cast after bending (Color figure online).

METALLURGICAL AND MATERIALS TRANSACTIONS A

 preheat treatment at 1050 C could effectively improve
the bending performance of the heavy-section cast, even
though a continuous morphology of d-ferrite still exists.

IV. CONCLUSIONS
d-Ferrite formation across the thickness of a
heavy-section AISI 316 casting slab and the effect of
its decomposition on the bending properties are studied
in the present work. The main conclusions can be
drawn:
1. Cellular-shaped d-ferrite with the faction of less
than 3 pct is observed at the surface position of the
cast while skeletal-shaped d-ferrite with the fraction
of up to 8 pct is present at the center position.
Micro-segregation of Cr and Mo during solidifica-
tion is responsible for the formation of d-ferrite.
Macro-segregation of Cr, Mo, and Ni across the
slab thickness results in the increased Creq/Nieq
ratio from surface to center, ultimately altering the
solidification mode from AF mode to FA mode. In
addition, the amount of d-ferrite from AF mode is
not affected by the change in cooling rate, and a
relatively slow cooling rate might be favorable for
Fig. 8—Schematic of cracking behaviors in (a, b) cellular-shaped
d-ferrite at surface position of the cast and (c, d) skeletal-shaped
the formation of d-ferrite from FA mode.
d-ferrite at the center position of the cast during bending test: (a, c) 2. Extensive decomposition of d-ferrite occurred
crack initiation and (b, d) crack propagation. across the thickness of the cast. Two different

Fig. 9—Observation on (a, b) dissolution of r phase after heating to 1050 C for 1 h, and (c, d) re-precipitation of r phase after reheating at
1000 C for 1 h. Inserted chart presents EDS point analysis (wt pct) at positions marked in (d).

METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 10—Micrographs of bending samples preheat treated at 1050 C: (a) convex surface, (b) OM and (c) SEM microstructures beneath the
convex surface.

decomposition mechanisms of d-ferrite involving REFERENCES

direct transformation (d fi r) and eutectoid trans- 1. M. Mamivand, Y. Yang, J. Busby, and D. Morgan: Acta Mater.,
formation (d fi r + c2) are found. 2017, vol. 130, pp. 94–110.
3. Micro-cracks within the decomposed d-ferrite could 2. A. Hermant, E. Suzon, P. Petit, J. Bellus, E. Georges, F. Cortial,
be observed on the convex surface of bending M. Sennour, and A.F. Gourgues-Lorenzon: Metall. Mater. Trans.
A, 2019, vol. 50A, pp. 1625–42.
samples from the surface position and center 3. S. Chen, Y. Miyahara, A. Nomoto, and K. Nishida: Acta Mater.,
position of the cast. Occurrence of cracking is only 2019, vol. 179, pp. 61–69.
present in the sample from center position of the 4. T.G. Lach, T.S. Byun, and K.J. Leonard: J. Nucl. Mater., 2017,
cast due to a higher crack propagation rate. vol. 497, pp. 139–53.
5. A.S. Alomari, N. Kumar, and K.L. Murty: Metall. Mater. Trans.
Removal of the brittle r phase within d-ferrite by A, 2019, vol. 50A, pp. 641–54.
a preheat treatment at 1050 C could effectively 6. KH Lo, CH Shek, and JKL Lai: Mater. Sci. Eng. R, 2009, vol. 65,
improve the bending performance of the heavy-sec- pp. 39–104.
tion cast. 7. J.A. Brooks and A.W. Thompson: Int. Mater. Rev., 1991, vol. 36,
4. The dissolved r phase within d-ferrite after heat pp. 16–44.
8. T. Ogawa and E. Tsunetomi: Welding. J., 1982, vol. 61, pp. 82–93.
treatment at 1050 C will reappear after reheat 9. A.F. Padilha and P.R. Rios: ISIJ Int., 2002, vol. 42, pp. 325–37.
treatment at 1000 C, demonstrating that decom- 10. A.F. Padilha, D.M. Escriba, E. Materna-Morris, M. Rieth, and
posed d-ferrite present in the heavy-section cast is M. Klimenkov: J. Nucl. Mater., 2007, vol. 362, pp. 132–38.
generated during the slow cooling stage after 11. C Gennari, L Pezzato, E Piva, R Gobbo, and I Calliari: Mater.
Sci. Eng. A, 2018, vol. 729, pp. 149–56.
solidification. 12. YL He, NQ Zhu, XG Lu, and L Li: Mater. Sci. Eng. A, 2010,
vol. 528, pp. 721–29.
13. AV Kington and FW Noble: Mater. Sci. Eng. A, 1991, vol. 138,
pp. 259–66.
14. C.C. Tseng, Y. Shen, S.W. Thompson, M.C. Mataya, and
G. Krauss: Metall. Mater. Trans. A, 1994, vol. 25A, pp. 1147–58.
ACKNOWLEDGMENTS 15. Y Kumar, S Venugopal, G Sasikala, PK Parida, and A Moitra:
Mater. Sci. Eng. A, 2018, vol. 731, pp. 551–60.
This work is financially supported by the National 16. G. Bai, S. Lu, D. Li, and Y. Li: Corros. Sci., 2015, vol. 90,
Natural Science Foundation of China (Nos. 51871218 pp. 347–58.
and 51401215), Youth Innovation Promotion Associa- 17. M. Matula, L. Hyspecka, M. Svoboda, V. Vodarek, C. Dagbert,
tion, CAS (No. 2018227), and Structural Materials for J. Galland, Z. Stonawska, and L. Tuma: Mater. Charact., 2001,
vol. 46, pp. 203–10.
Nuclear Reactor Vessels and Major Component (No. 18. A.L. Schaeffler: Met. Prog., 1949, vol. 56, pp. 680–680B.
2016-DGB-I-KYSC-0024). 19. A.L. Schaeffler: Welding J., 1947, vol. 26, pp. 601–20.

METALLURGICAL AND MATERIALS TRANSACTIONS A

20. PL Ferrandini, CT Rios, AT Dutra, MA Jaime, PR Mei, and 33. D. Peckner and I.M. Bernstein: Handbook of Stainless Steels, 1st
R Caram: Mater. Sci. Eng. A, 2006, vols. 435–436, pp. 139–44. ed., McGraw-Hill, New York, NY, USA, 1977, pp. 1–76.
21. R. Song, J. Xiang, and D. Hou: J. Iron Steel Res. Int., 2011, 34. D.E. Villanueva, F.C.P. Junior, R.L. Plaut, and A.F. Padilha:
vol. 18, pp. 53–59. Mater. Sci. Technol., 2006, vol. 22, pp. 1098–1104.
22. G.I. Zelada-Lambri, O.A. Lambri, and G.H. Rubiolo: J. Nucl. 35. A.V. Kington and F.W. Noble: Mater. Sci. Technol., 1995, vol. 11,
Mater., 1999, vol. 273, pp. 248–56. pp. 268–75.
23. M. Mataya, E. Nilsson, E. Brown, and G. Krauss: Metall. Mater. 36. A. Pineau, A.A. Benzerga, and T. Pardoen: Acta Mater., 2016,
Trans. A, 2003, vol. 34A, pp. 1683–1703. vol. 107, pp. 424–83.
24. J.Y. Cho, F. Czerwinski, and J.A. Szpunar: J. Mater. Sci., 2000, 37. M. Hajian, A. Abdollah-zadeh, S. Rezaei-Nejad, H. Assadi,
vol. 35, pp. 1997–2003. S. Hadavi, K. Chung, and M. Shokouhimehr: Mater. Des., 2015,
25. J. Niagaj and L. Mazur: Weld. Int., 2014, vol. 28, pp. 345–53. vol. 67, pp. 82–94.
26. Y.H. Kim, D.J. Lee, J.C. Byun, K.H. Jung, J.I. Kim, H.J. Lee, 38. Z. Wang, Q. Meng, M. Qu, Z. Zhou, B. Wang, and W. Fu: Metall.
Y.T. Shin, S.H. Kim, and H.W. Lee: Mater. Des., 2011, vol. 32, Mater. Trans. A, 2016, vol. 47A, pp. 1268–79.
pp. 330–36. 39. X Jin, S Chen, and L Rong: Mater. Sci. Eng. A, 2018, vol. 722,
27. C. Cicutti and R. Boeri: Scripta Mater., 2001, vol. 45, pp. 1455–60. pp. 173–81.
28. S. Fukumoto, Y. Oikawa, S. Tsuge, and S. Nomoto: ISIJ Int., 40. W. Feng, S. Yang, and Y. Yan: Metall. Mater. Trans. A, 2018,
2010, vol. 50, pp. 445–49. vol. 49A, pp. 2257–68.
29. A. Perron, C. Toffolon-Masclet, X. Ledoux, F. Buy, T. Guilbert, 41. X.Y. Jiao, J. Wang, C.F. Liu, Z.P. Guo, G.D. Tong, S.L. Ma,
S. Urvoy, S. Bosonnet, B. Marini, F. Cortial, G. Texier, Y. Bi, Y.F. Zhang, and S.M. Xiong: J. Mater. Sci. Technol., 2019,
C. Harder, V. Vignal, Ph. Petit, J. Farre, and E. Suzon: Acta vol. 35, pp. 1099–1107.
Mater., 2014, vol. 79, pp. 16–29.
30. H Sieurin and R Sandstrom: Mater. Sci. Eng. A, 2007, vol. 444, Publisher’s Note Springer Nature remains neutral with regard to
pp. 271–76. jurisdictional claims in published maps and institutional affiliations.
31. J. Barcik: Mater. Sci. Technol., 1988, vol. 4, pp. 5–15.
32. K. Shinohara, T. Seo, and K. Kumada: Trans. Jpn. Inst. Met.,
1979, vol. 20, pp. 713–23.

METALLURGICAL AND MATERIALS TRANSACTIONS A