Available online at www.sciencedirect.

com

ScienceDirect
Acta Materialia 78 (2014) 369–377
www.elsevier.com/locate/actamat

The effects of the initial martensite microstructure

on the microstructure and tensile properties of
intercritically annealed Fe–9Mn–0.05C steel
Jeongho Han, Seung-Joon Lee, Jae-Gil Jung 1, Young-Kook Lee ⇑
Department of Materials Science and Engineering, Yonsei University, Seoul 120-749, Republic of Korea

Received 1 July 2014; accepted 4 July 2014

Available online 30 July 2014

Abstract

The effects of the initial microstructure of a0 martensite on the microstructural evolution during intercritical annealing and the tensile
properties of annealed specimens were investigated for Fe–9Mn–0.05C (wt.%) steel. The hot-rolled specimen with fully a0 martensitic
microstructure showed a mixed microstructure of lath-shaped ferrite (aL) and austenite (cL) after intercritical annealing. The aL grains
had a high density of dislocations due to inactive recovery, and also had a low Mn concentration. The cL grains had a low density of
dislocations and high Mn and C concentrations. The aL and cL grains were deformed simultaneously during the tensile test because the
aL grains were as hard as the cL grains due to their high dislocation density, resulting in continuous yielding. The cold-rolled specimen
with a deformed a0 martensite microstructure exhibited a mixed microstructure of globular-shaped ferrite (aG) and austenite (cG) after
intercritical annealing. The aG grains had a low dislocation density due to active recovery, and also had a low Mn concentration. The cG
grains had a low dislocation density and high Mn and C concentrations. The soft aG grains with a low dislocation density were easily
deformed at the early stage of the tensile test, resulting in discontinuous yielding and a large yield point elongation.
Ó 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: TRIP-assisted steel; Ultrafine-grained material; Phase transformation; Yield phenomena; Medium-Mn steel

1. Introduction The volume fraction of cR ranges from 20% to 40%. Inter-

Recently, medium-Mn transformation-induced plastic-
ity (TRIP) steels with less than approximately 10 wt.%
Mn have been actively investigated as advanced
high-strength steels that offer reasonable material and
production costs as well as excellent mechanical properties
[1–13]. While medium-Mn TRIP steels are essentially sin-
gle-phase a0 martensite after hot and cold rolling, these
steels have a mixed microstructure consisting of a ferrite
and retained c austenite (cR) after intercritical annealing.
⇑ Corresponding author. Tel.: +82 2 2123 2831; fax: +82 2 312 5375.
E-mail address: yklee@yonsei.ac.kr (Y.-K. Lee).
1
Present address: Light Metal Division, Korea Institute of Materials
Science, Changwon 642-831, Republic of Korea.
critically annealed medium-Mn TRIP steels exhibit remark-
able combinations of high strength (over 800 MPa) and
ductility of up to 40% due to the TRIP phenomenon that
occurs in cR during plastic deformation [1].
Previous studies on medium-Mn TRIP steels focused on
the effects of alloying elements [2–5] and intercritical
annealing conditions, e.g. isothermal holding temperature
[6–9], time [10–12], heating rate [13] and cooling rate [3],
on the microstructure and mechanical properties. Higher
Mn and C concentrations resulted in greater volume frac-
tions of cR after intercritical annealing and higher tensile
strength and ductility for the Fe–(4–8)Mn–0.1C (wt.%)
steels [2] and Fe–5Mn–(0.01–0.4)C (wt.%) steels [3]. The
addition of Al to medium-Mn TRIP steels allowed a higher

http://dx.doi.org/10.1016/j.actamat.2014.07.005
1359-6454/Ó 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
370 J. Han et al. / Acta Materialia 78 (2014) 369–377
intercritical annealing temperature because Al increased
the equilibrium reverse transformation start (Ae1) and fin-
ish (Ae3) temperatures, leading to a short intercritical
annealing time of approximately 2 min [4].
Regarding the effects of intercritical annealing condi-
tions, the volume fraction of cR and the total elongation
increased with increasing annealing temperature [6–9] and
time [10–12], and then decreased with further annealing.
This phenomenon is closely related to the volume fraction,
grain size and chemical stability of reverted c during inter-
critical annealing. A high annealing temperature and long
holding time result in a large fraction of reverted c with
coarse grain size and low chemical stability. The reverted
c undergoes martensitic transformation during cooling,
resulting in a small volume fraction of cR at room temper-
ature. Han and Lee [13] reported that diffusive reverse
transformation occurred in medium-Mn TRIP steels
during continuous heating at slow rates of less than
15 °C s1, giving rise to stable reverted c, particularly when
the holding time was short. Whereas the tensile properties
of medium-Mn TRIP steels with 0.1 wt.% or less C were
insensitive to the cooling rate after annealing, those of
steels with over 0.1 wt.% C were improved by rapid cooling
because of suppressed cementite precipitation [3].
The microstructures and tensile properties of intercriti-
cally annealed medium-Mn TRIP steels were also greatly
influenced by the initial microstructure of a0 martensite
before intercritical annealing. Whereas the athermal a0
martensite, which formed during cooling after hot rolling,
decomposed to the lath-shaped a and c phases during inter-
critical annealing [10,12], the deformed a0 martensite chan-
ged to the globular-shaped a and c phases [4,7,8,11].
Medium-Mn TRIP steels with lath-shaped a and c phases
exhibited continuous yielding [10,12], and the steels with
globular-shaped a and c phases revealed discontinuous
yielding and large yield point elongation [4,7,8,11]. How-
ever, until now, no reports have explained how the initial
microstructure of a0 martensite affected the microstructure
and tensile properties of intercritically annealed medium-
Mn TRIP steels.
Therefore, the objective of the present study was to
examine the following two phenomena in Fe–9Mn–0.05C
(wt.%) steel. First, we systematically studied how the
microstructure of the athermal a0 martensite changed to
the lath-shaped microstructure during intercritical anneal-
ing, and how the microstructure of the deformed a0 mar-
tensite changed into the globular-shaped microstructure.
Second, we elucidated the reasons why the specimen with
lath-shaped microstructure showed continuous yielding
and why the specimen with globular-shaped microstructure
exhibited discontinuous yielding and large yield point
elongation.
2. Experimental procedure
A 30 kg ingot of Fe–9Mn–0.05C (wt.%) steel was cast
using a vacuum induction furnace. The chemical
composition of the ingot was Fe–8.51Mn–0.05C–0.29Si–
0.08Al–0.0085P–0.005S (wt.%). After solution treatment
at 1200 °C for 2 h, the ingot was hot-rolled at approxi-
mately 900 °C to form a 3 mm thick plate, held at 600 °C
for 1 h, and then furnace-cooled to room temperature for
the coiling simulation. After surface descaling, the
hot-rolled plate was cold-rolled to form a 1.2 mm thick
sheet at room temperature.
For observation of the microstructural evolution during
intercritical annealing, dilatometric specimens measuring
3  1  10 mm3 were heated to 620 °C at a rate of
10 °C s1, held for 10, 30 and 600 s, and then gas-quenched
to room temperature at 50 °C s1 using a quench dilatom-
eter (Theta, Dilatronic III). The annealing temperature
(620 °C) was between the Ae1 and Ae3 temperatures, which
were calculated using Thermo-Calc software with the
TCFE7 database.
The microstructures of the hot-rolled, cold-rolled and
annealed specimens were observed using by optical micros-
copy (OM; Olympus, BX41M), field-emission scanning
electron microscopy (FE-SEM; JEOL, JSM-7001F), and
field-emission transmission electron microscopy (FE-
TEM; JEOL, JEM-2100F, operated at 200 kV) combined
with energy-dispersive X-ray spectroscopy (EDXS; Oxford,
INCA Energy). OM samples were etched with a 4% nital
etchant and a mixed solution of 0.5% hydrochloric acid,
3% dodecylbenzene sulfonic acid and 96.5% saturated pic-
ric acid at 60 °C. The SEM samples were electroetched for
1 min with a mixed solution of 90% glacial acetic acid
(CH3COOH) and 10% perchloric acid (HClO4). The
TEM samples were mechanically ground to 100 lm in
thickness and then punched to make disks 3 mm in diame-
ter. The disks were electropolished in the same mixed solu-
tion used for the SEM samples using a twin-jet polisher
(Struers, Tenupol-5).
The constituent phases and the misorientation angles
between phases and grains were examined by SEM (Hit-
achi, Su-70) combined with electron backscattered diffrac-
tion (EBSD; Hikari, EDAX-TSL, operated at 20 kV with
a step size of 0.06 lm). The EBSD samples were polished
using a colloidal silica suspension and then ion-milled to
remove the damaged layer.
The volume fraction and lattice parameter of each phase
in the annealed specimens were determined by X-ray
diffraction (XRD; Rigaku,0 Ultima-IV) using a Cu-Ka radi-
ation source (k = 1.5405 Å ). (The scanning range, rate and
step size were 40–100°, 2° min1 and 0.02°, respectively.
Phase fractions were calculated using the integrated inten-
sities of all diffracted peaks. The lattice parameter of cR
was calculated using each diffracted peak and then
averaged [14].
The dog-bone-shaped tensile specimens were machined
along the rolling direction from both the hot-rolled and
cold-rolled sheets. The size of the gauge portion of tensile
specimens was 1.2 mm in thickness, 6.0 mm in width and
25 mm in length. The tensile specimens annealed in a
tubular furnace at 620 °C for 600 s. The annealed tensile
 J. Han et al. / Acta Materialia 78 (2014) 369–377
specimens were mechanically ground, electropolished and
deformed at a constant strain rate of 1  104 s1 at room
temperature using an Instron 3382 machine.
The change in the volume fraction of strain-induced a0
martensite (a0S ) during the tensile tests was measured using
a Feritscope (Fischer, MP30E), which was calibrated in
advance with the volume fraction of a0S measured by XRD
before and after the tensile test [15,16]. The volume fraction
of a0S was measured every 10 s at the middle of the gauge
portion of the tensile specimens using the Feritscope.
The Lüders band and its propagation during the tensile
test were observed at the middle of the gauge portion by
OM using an Olympus BX51 with a differential interfer-
ence filter (Olympus, U-DICRHC) and two kinds of polar-
izing filters: a perpendicular filter (Olympus, MM6-PO)
and a horizontal filter (Olympus, U-AN360-3).
3. Results and discussion
3.1. Microstructural evolution during intercritical annealing
The hot-rolled (HR) specimen showed fully a0 martens-
itic microstructure without any precipitates (Fig. 1a and d).
The average grain size of the prior c in the HR specimen
was approximately 25 lm (Fig. 1b). The cold-rolled (CR)
specimen revealed deformed a0 martensite microstructure
(Fig. 1c and d).
Fig. 2a–c show SEM micrographs of the HR specimen
annealed at 620 °C for 10, 30 and 600 s, respectively. The
Fig. 1. OM micrographs showing (a) the a0 martensite and (b) prior austenite grain boundaries for the hot-rolled specimen and (c) the deformed a0
martensite for the cold-rolled specimen. (d) XRD patterns of both the hot- and cold-rolled specimens.
371
a0 martensite changed to tempered a0 martensite (a0T ) with
cementite particles (h, (Fe,Mn)3C) after 10 s (Fig. 2a)
[13]. The h particles precipitated at the boundaries of prior
c grains and at the laths and packets of a0 martensite,
where the Mn and C atoms were segregated. The dark pits
are considered to be cR based on the EBSD results (Fig. 4b
and c) and on a previous study [13]. A small amount of
globular-shaped reverted c (cG) was first nucleated at the
high-angle boundaries, e.g. at prior c grain and a0 packet
boundaries, most likely due to their high interfacial
energies.
When the isothermal holding time was 30 s (Fig. 2b), the
matrix was lath-shaped a (aL), and the h particles were
almost dissolved because the isothermal holding tempera-
ture of 620 °C was higher than the equilibrium dissolution
temperature of the h particles (520 °C). Whereas a small
amount of cG formed at the high-angle boundaries, a large
amount of lath-shaped reverted c (cL) was observed along
the aL boundaries, where the Mn and C atoms were enriched
by the dissolution of the pre-formed h [17]. The volume
fraction of cL increased greatly to approximately 30% after
600 s (Figs. 2c and 4a).
Fig. 2d and e show bright- and dark-field TEM micro-
graphs of the HR specimen annealed at 620 °C for 600 s.
The lath widths of aL and cL were approximately 250
and 180 nm, respectively, and the diameter of cG was
approximately 200 nm. The aL still had a high density of
dislocations due to inactive recovery, and it possessed a
lower concentration of Mn (approximately 7.0 wt.%) than
372 J. Han et al. / Acta Materialia 78 (2014) 369–377

Fig. 2. SEM micrographs of hot-rolled specimens annealed at 620 °C for (a) 10 s, (b) 30 s and (c) 600 s. (d) Bright-field and (e) dark-field TEM
micrographs of the hot-rolled specimen annealed at 620 °C for 600 s. cG is globular-shaped austenite, cL is lath-shaped austenite, a0T is tempered
martensite, aL is lath-shaped ferrite, and h is cementite.

the average Mn concentration of the whole specimen
(approximately 8.5 wt.%). However, both cL and cG had
low densities of dislocations and high Mn concentrations
(approximately 11.7 wt.%). The average C concentrations
of both cL and cG were calculated to be approximately
0.12 wt.% using the following equation [18]:
ac ¼ 3:556 þ 0:0453C þ 0:00095Mn
0 a0 martensite. The volume fraction of cG increased, and the
where ac is the lattice parameter (Å ) of cL or cG, and C and
Mn are the concentrations of C and Mn (wt.%), respec-
tively, in cL or cG. The calculated C concentration was
much higher than the average C concentration of the whole
specimen (0.05 wt.%). The reason for the high concentra-
tions of Mn and C in both cL and cG was because cL and
cG were nucleated at the boundaries, where the Mn and C
atoms were enriched by the dissolution of the h particles
[13], and because the Mn and C atoms were partitioned
from aL to cL and cG during intercritical annealing [7].
Fig. 3a–c show SEM micrographs of the CR specimen
annealed at 620 °C for 10, 30 and 600 s, respectively. When
the isothermal holding time was 10 s (Fig. 3a), the
deformed a0 martensite decomposed to a0T with h and cG.
The cG grains were observed near the h particles, indicating
that they were nucleated after the partial dissolution of the
h particles. After holding at 620 °C for 30 s (Fig. 3b), the a0T
matrix changed to the globular-shaped a (aG), most likely
ð1Þ due to active recovery and recrystallization in the deformed
h particles were almost dissolved. The CR specimen that
had been annealed for 600 s exhibited a mixed microstruc-
ture of aG and cG (Fig. 3c). The CR specimen exhibited a
higher volume fraction (37%) of cG (Figs. 3c and 4a), com-
pared to the HR specimen (30%) annealed under identical
conditions. This result indicates that the deformation by
cold rolling accelerated the reverse transformation from
a0 martensite to c.
Fig. 3d shows a bright-field TEM micrograph of the CR
specimen annealed at 620 °C for 600 s. The grain sizes of
aG and cG were nearly the same at approximately 300 nm.
The aG had a relatively lower density of dislocations than
 J. Han et al. / Acta Materialia 78 (2014) 369–377 373

Fig. 3. SEM micrographs of cold-rolled specimens annealed at 620 °C for (a) 10 s, (b) 30 s and (c) 600 s. (d) Bright-field TEM micrograph of the cold-
rolled specimen annealed at 620 °C for 600 s. aG is globular-shaped ferrite.

Fig. 4. (a) XRD patterns and EBSD phase maps of (b) hot-rolled and (c) cold-rolled specimens annealed at 620 °C for 600 s. On the EBSD phase maps,
austenite is in red, ferrite is in white, the green lines indicate low-angle boundaries with misorientation angles between 3° and 15°, and the blue lines
indicate high-angle boundaries with misorientation angles of over 15°. (For interpretation of the references to color in this figure legend, the reader is
referred to the web version of this article.)

the aL in the HR specimen (Fig. 2d) due to active recovery,
and the cG also had a low density of dislocations. The aG
and cG had lower (6.15 wt.%) and higher (12.15 wt.%)
Mn concentrations, respectively, than the average Mn
concentration of the whole specimen (8.51 wt.%). The C
concentration in cG was calculated from Eq. (1) to be
0.19 wt.%.
To confirm the phase fractions and the behaviors of
recovery and recrystallization, XRD (Fig. 4a) and EBSD
(Fig. 4b and c) analyses were conducted using the HR
and CR specimens annealed at 620 °C for 600 s. Analysis
of the XRD patterns shows that the volume fraction of
cR was 30% in the HR specimen and approximately 37%
in the CR specimen, as noted above. The full-width at half-
maximum (FWHM) values of the (1 1 0)a, (2 0 0)a and
(2 1 1)a peaks of the HR specimen were 0.31°, 0.53° and
0.42°, respectively; the FWHM values of the three peaks
of the CR specimen were 0.22°, 0.45° and 0.38°, respec-
tively (Fig. 4a). These results imply that the HR specimen
had a higher density of crystal defects, such as dislocations,
than the CR specimen, which shows good agreement with
the TEM observations (Figs. 2d and 3d).
374 J. Han et al. / Acta Materialia 78 (2014) 369–377

Fig. 4b and c show the phase maps of the HR and CR
specimens, respectively. The phases in the red and white
colors correspond to cR and a, respectively, and the green
and blue lines represent low-angle boundaries with misori-
entation angles of less than 15° and high-angle boundaries
with misorientation angles of over 15°, respectively. The
volume fraction of cR was approximately 38% in the HR
specimen and 34% in the CR specimen. The volume frac-
tion of cR measured by EBSD was different from that mea-
sured by XRD, most likely due to the localized observation
that is a characteristic of the EBSD. Whereas the HR spec-
imen had lath-shaped microstructure with few subgrains
and recrystallized grains (Fig. 4b), the CR specimen exhib-
ited globular-shaped microstructure with subgrains and
recrystallized grains (Fig. 4c), implying good agreement
with the XRD (Fig. 4a) and SEM (Figs. 2 and 3) analyses.
3.2. Tensile properties
Fig. 5a shows the engineering stress–strain curves and
variations in the volume fraction of cR measured during
tensile tests in HR and CR specimens that were annealed
at 620 °C for 600 s. The flow curve of the HR specimen
exhibited continuous yielding. The yield strength (YS), ulti-
mate tensile strength (UTS), and total elongation (El.) were
830 MPa, 1034 MPa, and 27%, respectively. After yielding,
the volume fraction of cR gradually decreased with increas-
ing strain, indicating that TRIP occurred steadily and uni-
formly during the tensile test.
The flow curve of the CR specimen showed discontinuous
yielding and a large yield point elongation of
approximately 10.7%. One Lüders band propagated with a
velocity of 0.095 mm s1 from one end of the gauge portion
to the other end during the yield point elongation, indicating
localized deformation (Fig. 5c). The width of the Lüders
band was approximately 0.83 mm and the angle between
the tensile direction and the Lüders band was
approximately 65° (Fig. 5d). After the yield point elonga-
tion, no more bands propagated and the specimen was uni-
formly deformed throughout the entire gauge portion. The
upper and lower YSs, UTS, and El. were 1110, 1060,
1193 MPa and 25%, respectively (Fig. 5a). The CR specimen
exhibited higher strength compared to the HR specimen.
The volume fraction of cR in the CR specimen dropped
abruptly from approximately 37% to 25% at the middle of
the yield point elongation, and then remained relatively
steady until the end of the yield point elongation. Consider-
ing the fact that the change in the volume fraction of cR was
measured using the Feritscope at the middle of the gauge
portion and that one Lüders band propagated during yield

Fig. 5. (a) Engineering stress–strain curves and variations in the volume fraction of retained austenite during the tensile tests and (b) true stress–strain
curves and strain-hardening rate curves for the hot- and cold-rolled specimens annealed at 620 °C for 600 s. VcR is the volume fraction of retained
austenite. (c) OM micrographs showing the propagation of a Lüders band with tensile strain in the cold-rolled specimen annealed at 620 °C for 600 s, and
(d) a polarizing filtered OM micrograph showing the Lüders band.
 J. Han et al. / Acta Materialia 78 (2014) 369–377
point elongation, this result meant that the TRIP occurred
only at the locally deformed area caused from Lüders band
passage.
Just after the yield point elongation, the specimen start
to be deformed uniformly (Fig. 5c). The strength increased
steeply, and the volume fraction of cR dropped rapidly
from approximately 25% to 18%, indicating that the TRIP
occurred actively throughout the entire gauge portion.
Afterwards, the strength increased slightly, and the volume
fraction of cR dropped slowly for the engineering strain
range from 11% to 13%. With additional strain, the
strength increased actively again and then decreased after
necking, and the volume fraction of cR decreased gradually
until the specimen failed, indicating that the TRIP occurred
steadily and uniformly.
Fig. 5b shows the true stress (r)-strain (e) curves and the
strain-hardening rate (SHR, dr/de) curves of HR and CR
specimens annealed at 620 °C for 600 s. The SHR of the
HR specimen decreased rapidly until the strain reached
approximately 0.02, most likely due to fast dynamic recov-
ery [19]; it then increased with further strain to 0.033 due to
the active TRIP phenomenon, and then decreased again
until failure. The SHR of the CR specimen decreased rap-
idly at the early stage of deformation and remained rela-
tively steady until the strain reached approximately 0.1
due to yield point elongation. Then, the SHR curve exhib-
ited two peaks due to the double TRIP. The SHR increased
rapidly until reaching a strain of approximately 0.105,
decreased with further strain until reaching 0.116,
increased again up to the strain of 0.137, and then
decreased until failure. The more unstable cR with coarse
Fig. 6. (a) SEM micrographs of the hot-rolled specimens that were annealed at 620 °C for 600 s and deformed with five different strains and (b) changes in
the tensile strain at the grains of the lath-shaped ferrite (aL) and austenite (cL) and the volume fraction of retained austenite with the engineering strain.
375
grains or low concentrations of Mn and C underwent TRIP
at relatively low strain values, and then the more stable cR
with fine grains or high concentrations of Mn and C was
transformed to martensite later, giving rise to the double
peaks on the SHR curve.
To examine the difference in yielding behavior between
the HR and CR specimens, microstructural evolutions dur-
ing the tensile tests were observed using the HR and CR
specimens annealed at 620 °C for 600 s. Fig. 6a shows the
change in the microstructure of the HR specimen with
increasing tensile strain up to 18.5%. The SEM micro-
graphs were taken at the same location at the middle of
the gauge portion. Although some of the cL grains were
already transformed to a0 martensite (a0S ) just after yielding
(Fig. 5), cL and a0S were not distinguishable on the SEM
micrographs due to the chemical etching.
The amounts of deformation for the aL grains and the
mixed grains of cL and a0S were measured on the SEM
micrographs at a given strain (Fig. 6a) and are plotted
along with the cR volume fractions against the tensile strain
in Fig. 6b. The tensile strain for a given grain (ep) was cal-
culated using the following equation:
Ls  Li
ep ð%Þ ¼  100 ð2Þ
Li
where Li and Ls are the initial and elongated widths of the
grain (nm), respectively. The aL grains and the mixed
grains of cL and a0S were linearly elongated with increasing
tensile strain. For strain values of over 9%, the mixed
grains of cL and a0S were more strongly elongated than
the aL grains, because it was thought to be mainly due to
376 J. Han et al. / Acta Materialia 78 (2014) 369–377

the transformation strain caused by the strain-induced a0
martensitic transformation from face-centered cubic to
body-centered tetragonal.
Fig. 7a shows microstructural changes in the CR
specimen with increasing tensile strain up to 19.2%. The
SEM micrographs were taken at the same location in the
middle of the gauge portion. Although some of the cG
grains started to transform to a0S after yielding (Fig. 5),
the cG and a0S phases were not distinguishable on the
SEM micrographs. The microstructure of the specimen
strained by 4.2% was insignificantly changed compared to
that of the non-deformed specimen because the propaga-
tion of the Lüders band was not yet completed at the
observed region.
Fig. 7b shows the variations in the tensile strain of both
the aG grains and the mixed grains of cG and a0S and the cR
volume fraction as a function of tensile strain in the CR
specimen. The aG grains were more severely deformed than
the mixed grains of cG and a0S during the yield point
elongation (up to engineering strain of 10.7%). This
phenomenon is referred to as strain partitioning [20]. How-
ever, when the strain increased from 9.5% (just before the
end of the yield point elongation) to 11.2% (just after the
yield point elongation), whereas the mixed grains of cG
and a0S were greatly elongated by 6.2%, the aG grains were
slightly strained by 0.8% because the aG was already strain
hardened. This result indicates that the cG was primarily
deformed just after the yield point elongation, leading to
the large reduction in volume fraction of cR. Meanwhile,
the reason why there was no significant increase in strain,
despite of the large elongation of mixed grains of cG and
a0S , is because the volume fraction of cR was only
approximately 25%. With further strain, the mixed grains
of cG and a0S were more strongly elongated than the aG
grains, showing the similar results of HR specimen.
Strain partitioning is considered one of the reasons for
the yield point elongation during the tensile test of the
CR specimens. The aG grains in the CR specimen were
softer than the cG grains due to active recovery and the
low concentrations of Mn and C. The soft aG grains were

Fig. 7. (a) SEM micrographs of the cold-rolled specimens that were annealed at 620 °C for 600 s and deformed with five different strains and (b) changes in
the tensile strain at the grains of the globular-shaped ferrite (aG) and austenite (cG) and the volume fraction of retained austenite with the engineering
strain.
 J. Han et al. / Acta Materialia 78 (2014) 369–377
easily deformed at the early stage of yielding, and Lüders
bands formed and propagated by the interaction between
mobile dislocations and the C atoms in the aG grains,
resulting in the large yield point elongation.
However, in the HR specimen, although the cL grains
had high concentrations of Mn and C, the aL grains were
as hard as the cL grains due to a high density of disloca-
tions caused by inactive recovery. Therefore, the aL and
cL grains began to be deformed almost simultaneously at
the early stage of yielding. The active TRIP improved the
strain hardening rate at the early stage of yielding
(Fig. 5b) and prevented localized deformation, eliminating
the yield point elongation.
4. Conclusions
(1) The HR Fe–9Mn–0.05C (wt.%) specimen showed a
fully a0 martensitic microstructure without precipi-
tates. The HR specimen exhibited a mixed microstruc-
ture of lath-shaped ferrite (aL) and retained austenite
(cL) phases after intercritical annealing at 620 °C for
600 s. The cL volume fraction was 30%. The aL grains
had a low Mn concentration (approximately 7.0 wt.%)
and a high density of dislocations due to inactive
recovery. The cL grains had high concentrations of
Mn (11.7 wt.%) and C (0.12 wt.%) and a low density
of dislocations. The intercritically annealed HR spec-
imen exhibited continuous yielding during tensile test-
ing. The aL and cL grains were simultaneously
deformed at the early stage of the tensile test because
the aL grains were as hard as the cL grains due to their
high density of dislocations. The TRIP effect
improved the strain-hardening rate in the early stage
of the tensile test and prevented localized deforma-
tion, eliminating the yield point elongation.
(2) The CR Fe–9Mn–0.05C (wt.%) specimen showed a
fully deformed a0 martensite microstructure without
any precipitates. The CR specimen had a mixed
microstructure of globular-shaped ferrite (aG) and
retained austenite (cG) after intercritical annealing
at 620 °C for 600 s. The cG volume fraction was
37%. The aG grains had low Mn concentrations
(approximately 6.15 wt.%) and low densities of dislo-
cations due to active recovery. The cG grains had
high concentrations of Mn (12.15 wt.%) and C
(0.19 wt.%) and low densities of dislocations. The
377
intercritically annealed CR specimen exhibited dis-
continuous yielding and large yield point elongation
during the tensile test. The soft aG grains with low
densities of dislocations were easily deformed at the
early stage of the tensile test, and Lüders bands
formed and propagated, resulting in the large yield
point elongation.
Acknowledgments
This research was supported by Basic Science Research
Program through the National Research Foundation of
Korea (NRF) funded by the Ministry of Education (Grant
no. 2013R1A1A2060558).
References
[1] Miller RL. Metall Trans 1972;3:905.
[2] Hong H, Lee OY, Song GH. J Korean Soc Heat Treat 2003;16:205.
[3] Furukawa T, Huang H, Matsumura O. Mater Sci Technol
1994;10:964.
[4] Suh DW, Park SJ, Lee TH, Oh CS, Kim SJ. Metall Mater Trans A
2010;41:397.
[5] Han Y, Shi J, Xu L, Cao WQ, Dong H. Mater Sci Eng A
2011;530:643.
[6] Gibbs PJ, De Moor E, Merwin MJ, Clausen B, Speer JG, Matlock
DK. Metall Mater Trans A 2011;42:3691.
[7] Lee S, Lee SJ, Lee SJ, Santhosh Kumar S, Lee K, Cooman BCD.
Metall Mater Trans A 2011;42:3638.
[8] Suh DW, Ryu JH, Joo MS, Yang HS, Lee K, Bhadeshia HKDH.
Metall Mater Trans A 2013;44:286.
[9] Lee S, De Cooman BC. Metall Mater Trans A 2013;44:5018.
[10] Luo H, Shi J, Wang C, Cao W, Sun X, Dong H. Acta Mater
2011;59:4002.
[11] Jang JM, Kim SJ, Kang NH, Cho KM, Suh DW. Met Mater Int
2009;15:909.
[12] Cao WQ, Wang C, Shi J, Wang MQ, Hui WJ, Dong H. Mater Sci
Eng A 2011;528:6661.
[13] Han J, Lee Y-K. Acta Mater 2014;67:354.
[14] Cullity BD, Stock SR. Elements of X-ray diffraction. Englewood
Cliffs, NJ: Prentice Hall; 2001.
[15] Jacques PJ, Allain S, Bouaziz O, De A, Gourgues AF, Hance BM,
et al. Mater Sci Technol 2009;25:567.
[16] Talonen J, Aspegren P, Hänninen H. Mater Sci Technol
2004;20:1506.
[17] Liu L, Yang ZG, Zhang C, Liu WB. Mater Sci Eng A 2010;527:7204.
[18] Van Dijk NH, Butt AM, Zhao L, Sietsma J, Offerman SE, Wright JP,
et al. Acta Mater 2005;53:5439.
[19] Kuhlmann-Wilsdorf D. Mater Sci Eng A 1989;113:1.
[20] Ryu JH, Kim DI, Kim HS, Bhadeshia HKDH, Suh DW. Scripta
Mater 2010;63:297.