Materials Science & Engineering A 739 (2019) 17–25

Contents lists available at ScienceDirect

Materials Science & Engineering A

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

The influence of microstructural characteristics on yield point elongation T

phenomenon in Fe-0.2C-11Mn-2Al steel
Z.H. Caia, , S.Y. Jinga, H.Y. Lib, , K.M. Zhanga, R.D.K. Misrac, H. Dinga, Z.Y. Tanga
⁎ ⁎

a
School of Materials Science and Engineering, Northeastern University, Shenyang 110819, China
b
School of Materials Science and Engineering, Taiyuan University of Science and Technology, Taiyuan 030024, China
c
Laboratory for Excellence in Advanced Steel Research, Department of Metallurgical, Materials and Biomedical Engineering, University of Texas at El Paso, El Paso, TX
79968, USA

ARTICLE INFO ABSTRACT

Keywords: We elucidate here the underlying reasons for the yield point elongation (YPE) phenomenon in medium-Mn steel.
Medium Mn steel The cold-rolled Fe-0.2C-11Mn-2Al steel was subjected to controlled annealing temperature and soaking time
Yield point elongation such that YPE of different lengths was obtained. It was observed that the ultra-fine grained or globular micro-
Austenite stability structure is not the critical factor responsible for YPE. YPE is a consequence of close competition between the
Dislocation density
work hardening (via martensitic transformation) and softening (by stress relaxation and transfer) induced by the
TRIP effect
TRIP effect. The ferrite grains with γ fiber have the potential to have high dislocation storage ability, with
consequent increase in dislocation density and work hardening exponent, leading to decrease of YPE.

1. Introduction
Medium Mn steels continue to receive significant attention because
of excellent mechanical properties and potential for automotive struc-
tural components. In the past decade, numerous studies have focused
on obtaining large fraction of austenite with proper stability by opti-
mizing chemical composition and adjusting heat treatment [1–4]. It
was reported that ductility could be increased simultaneously with
strength by transformation-induced plasticity (TRIP) mechanism, con-
tributing to superior tensile property > 30 GPa⋅%, when austenite was
present with ferrite or martensite in medium Mn steels. Especially,
ultra-high strength (> 1.5 GPa) combined with high ductility (> 20%)
were achieved in (austenite + martensite) duplex microstructure [5].
In order to achieve good strength-ductility balance, austenite should
gradually transform over a wide range of strain [6]. It has been proven
that Lüders bands directly results from the mechanical stability of
austenite [7]. Austenite stability is known to depend on chemical
composition [8], grain size [4,9], morphology [10], and crystal-
lographic orientation of austenite [11], etc, among which chemical
composition and grain size are the primary parameters [4,11]. Auste-
nite stability can be varied by controlling the intercritical annealing
(IA) temperature and the annealing time [2,12]. Moreover, Lüders
bands were influenced by IA temperature in medium Mn steels [13,14].
The phenomenon of Lüders bands characterized by a stress plateau
accompanied by serrations, which initiates at the onset of yielding and
spreads for a certain strain, is referred as yield point elongation (YPE).
The classical explanation for YPE phenomenon is static strain aging,
which means that dislocations can be pinned by clusters of interstitial
atoms, or Cottrell atmospheres [15]. A different mechanism that pro-
motes Lüders band formation is ultra-fine grained (UFG) microstructure
(grain size < 1 µm) [16]. With decreasing grain size, an increasing
number of dislocations are trapped at grain boundaries and become
immobile [17–19], which has a similar effect as the trapping of dis-
locations by interstitials. Medium-Mn steels commonly exhibit large
YPE, some of them easily exceeding 10% [14,20,21]. It was reported
that both low IA temperature and a globular microstructure supported
the formation of large YPE. Moreover, it was recommended to provide a
martensitic microstructure prior to intercritical annealing in order to
limit YPE [20].
Although there are some studies on the YPE, the micromechanical
origin of YPE is still unknown. The study described here concentrates
on the influence of microstructural characteristics on YPE. A detailed
characterization of microstructural evolution, including austenite
(fraction, size, stability) and ferrite (size, texture) characteristics, and
the resulting mechanical properties including YPE and work hardening
behavior are presented.
2. Experimental
The experimental steel had a nominal composition (wt%) of Fe-

⁎
Corresponding authors.
E-mail addresses: tsaizhihui@163.com (Z.H. Cai), lhyemail@163.com (H.Y. Li).

https://doi.org/10.1016/j.msea.2018.09.114
Received 13 September 2018; Received in revised form 27 September 2018; Accepted 29 September 2018
Available online 09 October 2018
0921-5093/ © 2018 Elsevier B.V. All rights reserved.
Z.H. Cai et al.
Table 1
Chemical composition (wt%) of steel and critical temperatures (°C).
Mn Al C Fe Ac1 Ac3
11.20 1.95 0.22 Bal. 630 730
0.2C-11Mn-2Al, and the actual chemical composition is listed in
Table 1. The ingot was heated at 1200 °C for 2 h, hot forged to rods with
a section size of 100 mm × 30 mm, followed by air cooled to room
temperature. Subsequently, the rods were soaked at 1200 °C for 2 h, hot
rolled to 4 mm in thickness and finally air cooled to room temperature.
The as-hot rolled sheets were then cold-rolled to 1 mm thickness.
In order to establish an appropriate heat treatment schedule, the
critical temperatures Ac1 and Ac3 of the experimental steel were ob-
tained by dilatometry and are listed in Table 1. For obtaining different
microstructural constituents and grain size, the as-cold rolled sheets
were annealed in the temperature range of 600–800 °C for 5 min, fol-
lowed by water quenching. Additionally, to obtain similar austenite
fraction but different austenite stability, the as-cold rolled sheets were
annealed at 650 °C for different periods (2 min~48 h), followed by
water quenching.
Tensile tests were conducted on specimens of 12.5 mm width and
gauge length of 25 mm, using a universal testing machine (SANSCMT
5000) at a constant crosshead speed of 3 mm·min−1 at room tempera-
ture. The samples were etched with 25% sodium bisulfite aqueous so-
lution. Microstructural examination was carried out using scanning
electron microscope (SEM) equipped with electron backscatter dif-
fraction (EBSD) and transmission electron microscope (TEM). The vo-
lume fraction of austenite was determined by X-ray diffraction (XRD)
with CuKα radiation using direct comparison method [22], involving
the use of integrated intensities of (200)α and (211)α peaks and those of
(200)γ, (220)γ and (311)γ peaks. The volume fraction of austenite VA
was calculated using equation [23]:
VA = 1.4I /(I + 1.4I ) (1)
where Iγ is the integrated intensity of austenite and Iα is the integrated
intensity of phases with body-centered cubic structure.
Fig. 1. SEM micrographs of the cold-rolled samples quenched from (a) 600 °C, (b) 650 °C, (c) 700 °C and (d) 750 °C, respectively and (e) high magnification of
austenite and ferrite strips. The bright phase is austenite, and the dark one is ferrite. (A: austenite, F: ferrite, M: martensite).
18
Materials Science & Engineering A 739 (2019) 17–25
3. Results and discussion
3.1. Microstructure and mechanical properties
SEM micrographs of cold-rolled sheets annealed at 600 °C, 650 °C,
700 °C,and 750 °C for 5 min, respectively, and then immediately quen-
ched in water, are presented in Fig. 1. As shown in Fig. 1a, the mi-
crostructural constituents consisted of ferrite and austenite. The mi-
crostructure can be described into two types: strip-like and equiaxed /
granular. The strip structure developed during cold rolling of partially
recrystallized, evolved into equiaxed structure on annealing at 600 °C.
For clarity, a mixture of austenite and ferrite strips are identified by a
rectangle in Fig. 1a, and a high-magnification micrograph of this region
is illustrated in Fig. 1e. It is distinct that austenite strips were present
between the adjacent ferrite strips. As shown in Fig. 1b and c, the
austenite and ferrite strips gradually developed into equiaxed micro-
structure with increase in annealing temperature. A small number of
strip structure continue to be retained on annealing at 650 °C. However,
it completely disappeared after annealing at 700 °C. When the samples
were annealed at 750 °C, as shown in Fig. 1d, the microstructure
comprised of martensite and austenite.
The measured austenite fraction as a function of annealing tem-
perature or soaking time was summarized in Fig. 2. As illustrated in
Fig. 2a, the austenite fraction increases from 55% to 89% in the tem-
perature range of 600–700 °C, followed by drastic decrease to 46%,
when quenching was carried out at 750 °C because of martensitic
transformation, consistent with microstructure illustrated in Fig. 1d.
Therefore, the austenite fraction is sensitive to annealing temperature.
Furthermore, it is confirmed that the austenite fraction remains nearly
constant when the samples were annealed at 650 °C for different per-
iods (as evidenced in Fig. 2b).
Fig. 3 shows the engineering stress-strain plots of the annealed
samples. As shown in Fig. 3a, it is noticed that YPE only appears in the
sample annealed at 650 °C for 5 min (henceforth nominated as
650–5 min sample). Moreover, as demonstrated in Fig. 3b, it is clear
that the length of YPE decreases with extended soaking time at 650 °C.
To interpret this phenomenon, the interplay between microstructural
characteristics and plastic deformation behavior including YPE and
TRIP effect should be elucidated.
Z.H. Cai et al. Materials Science & Engineering A 739 (2019) 17–25

Fig. 2. The measured austenite fractions of the cold-rolled samples (a) annealed at different temperatures, (b) annealed at 650 °C for different soaking time.

Fig. 3. The engineering stress-strain plots of the cold-rolled samples (a) annealed at different temperatures for 5 min, (b) annealed at 650 °C for different soaking
time.

Fig. 4. EBSD phase maps of the annealed samples. The phases colored in red and blue correspond to ferrite and austenite, respectively. The white lines are low-angle
boundaries with misorientation angles between 5° and 15°. The black lines are high-angle boundaries with misorientation angles of over 15°. (a) 650–5 min sample,
(b) 650–1 h sample, (c) 650–12 h sample, (d) 650–48 h sample, (e) 700–5 min sample.

19
Z.H. Cai et al. Materials Science & Engineering A 739 (2019) 17–25

Table 2 that, as shown in Fig. 3, YPE only appeared in the samples annealed at
The average grain size (μm) of austenite and ferrite of the annealed samples. 650 °C (referred as 650 samples). Therefore, the UFG or globular mi-
Annealed samples Austenite Ferrite crostructure is not the critical factor causing YPE in our case.
In order to decrease or eliminate the YPE, Li et al. [25] suggested
650–5 min 0.34 0.28 increasing austenite stability and dislocation density in ferrite. As
700–5 min 0.54 0.50
shown in Fig. 2, it is inferred that austenite stability decreases with
650–1 h 0.54 0.38
650–12 h 0.69 0.54
increase in annealing temperature or soaking time. Furthermore, the
650–48 h 0.85 0.66 average dislocation density in ferrite and austenite were determined by
XRD experiments [26,27]. The average dislocation density was eval-
uated by selecting the diffraction peaks of (220)γ and (211)α. The
3.2. The influences in YPE density of dislocations was estimated using Eq. (2) proposed by Dunn
[28,29].
Fig. 4 demonstrates EBSD phase maps of the annealed samples. The
phases colored in red and blue correspond to ferrite and austenite, re- = 2 /(4.35 × b 2) (2)
spectively. The white lines are low-angle boundaries with misorienta-
where ρ and β is dislocation density and full-width at half maximum
tion angles between 5° and 15°. The black lines are high-angle bound-
(FWHM) of diffraction peaks, respectively as determined by X-ray dif-
aries with misorientation angles of over 15°. The average grain size of
fraction profile, b is Burgers vector and the value is 2.48 × 10−8 cm.
austenite and ferrite determined via Channel 5 software is summarized
FWHM is primarily affected by crystallite size and microstrain. In our
in Table 2. Austenite and ferrite grain size were in the range of
present case, the increase of dislocation density of samples which were
0.1–0.9 µm, and increased with increasing annealing temperature and
subjected to interrupted tensile tests was mainly contributed from the
retention period. It is reported that UFG microstructure is likely to
increasing strain. Therefore, the variation of FWHM is primarily af-
trigger YPE, and YPE will be more distinct with decreasing grain size
fected by microstrain. In order to better understand the evolution of
[17,24]. Steineder et al. [20] reported that both low annealing tem-
dislocation density, the XRD patterns corresponding to increasing strain
perature and globular microstructure remarkably supported the for-
are shown in Fig. 5. Table 3 summarizes the estimated dislocation
mation of large YPE. As shown in Fig. 1 and Table 2, all the annealed
density of ferrite in the annealed samples. It is obvious that the dis-
samples had UFG and globular microstructure. However, it is curious
location density of ferrite was unchanged or decreased slightly with

Fig. 5. XRD patterns of the annealed samples subjected to interrupted tensile tests. (a) 650–5 min sample, (b) 650–1 h sample, (c) 650–48 h sample, (d) 700–5 min
sample.

20
Z.H. Cai et al. Materials Science & Engineering A 739 (2019) 17–25

Table 3
Dislocation density (1014) of ferrite in the annealed samples.
Annealed samples Dislocation density
600–5 min 2.22
650–5 min 2.49
700–5 min 2.21
650–1 h 2.48
650–12 h 2.27
650–48 h 2.23
increase in annealing temperature or soaking time. Based on the above
results, austenite stability, ferrite dislocation density and YPE of 650
samples decreased with increase in soaking time, which is contrary to Li
et al.′s results [25]. Thus, the question arises as to what is the under-
lying reason for YPE.
3.3. The underlying reason for YPE
It has been observed that the YPE and martensitic transformation
takes places simultaneously during tensile test [30]. Nevertheless, it
cannot be concluded that the YPE is induced by martensitic transfor-
mation of austenite. It has been found that the YPE also takes place
when there is no martensitic transformation [31]. Although He et al.
[21] suggested that the large ferrite grains were responsible for the
initiation of YPE, which was on the basis that the large ferrite grains
had a lower resistance to dislocation plasticity as compared to the
austenite phase and the small ferrite grains. However, Han et al. [32]
proposed that the coarsening of globular ferrite grains increased the
propagation velocity of Lüders bands, resulting in the reduction in the
YPE. In our present case, as evidenced in Table 2 and Fig. 3, 650–12 h
sample characterized by a long YPE, instead, had a larger ferrite grain
size than 700–5 min sample which had no YPE. In addition, 650–48 h
sample having a larger ferrite grain size than 650–12 h sample showed
continuous yielding (without YPE). Therefore, the above interpretation
on the relationship between ferrite and YPE is not applicable to our
results. There are some other important factors contributed to YPE.
As shown in Fig. 3b, 650–2 min sample was characterized by long
YPE, UFG structure and high austenite stability. A smaller grain size
reduces the probability of dislocation trapping in the grain interior and
austenite has higher stability, thereby reducing the work hardening
rate. It is conceivable that when the austenite grain size is large enough,
and the hardening contributed by TRIP effect overcomes the loss of UFG
structure, YPE will disappear. Combined with XRD results (Fig. 2),
stress-strain plots (Fig. 3) and grain size (Table 2), it is assumed that
YPE is a consequence of close competition between work hardening (via
martensitic transformation) and softening (by stress relaxation and
transfer) induced by the TRIP effect. In an attempt to clarify the hy-
pothesis, concerning the impact of grain size on YPE, 650–5 min,
650–1 h, 650–48 h and 700–5 min samples were subjected to inter-
rupted tensile tests.
Fig. 6 illustrates the volume fraction of austenite as a function of
engineering strain in the aforementioned four samples. For 650–5 min
and 650–1 h samples (Fig. 6a and b), similarly, the austenite fraction
decreases rapidly on straining to 5%, and then remains at a certain level
until the end of YPE, and finally decreases with increase in strain.
Martensitic transformation was accompanied by volume expansion,
therefore, the nearby retained austenite and ferrite was inevitably
squeezed by the transformed martensite, leading to additional dis-
location propagation. As shown in Fig. 7a and b, the dislocation density
in austenite first increases, and then remains nearly invariant, finally
increases with increasing strain, corresponding to the change in aus-
tenite fraction. In contrast, the dislocation density in BCC structure
involving ferrite and martensite increases slightly throughout straining.
Therefore, for 650–5 min and 650–1 h samples, a majority of the local

Fig. 6. The engineering stress-strain plots and the variation of austenite fraction as a function of strain. (a) 650–5 min sample, (b) 650–1 h sample, (c) 650–48 h
sample, (d) 700–5 min sample.

21
Z.H. Cai et al.
Fig. 7. The variation of dislocation density as a function of stain. (a) 650–5 min sample, (b) 650–1 h sample, (c) 650–48 h sample, (d) 700–5 min sample.
stress was transferred to austenite rather than ferrite. In contrast, as
shown in Fig. 6c and d, for 650–48 h and 700–5 min samples, the
austenite fraction decreased quickly with increase in strain, especially
during the initial stage of straining (< 5%), indicating a significant
TRIP effect, contributing to the disappearance of YPE. Correspondingly,
the dislocation density both in austenite and BCC structure increase
quickly with increasing strain (as illustrated in Fig. 7c and d).
Comparing the variation in dislocation density and austenite stabi-
lity among the three 650 samples, it is distinct that the dislocation
density of austenite decreases with decrease in austenite stability,
corresponding to a reduction in YPE. Because of higher austenite sta-
bility in 650–5 min and 650–1 h samples, more dislocations were re-
quired to stimulate martensitic transformation. Moreover, at similar
strain (< 10%), the fraction of transformed martensite in 650–5 min
and 650–1 h samples was less. Thus, the local stress can be effectively
transferred to the nearby austenite and ferrite grains, leading to local
softening and the consequent existence of YPE. Therefore, in spite of
higher dislocation density of austenite in the two samples, the work
hardening ability is still low, which was further confirmed in Fig. 8a
and b. In contrast, due to the lower austenite stability, more martensite
nucleated during the initial stage of straining in 650–48 h and
700–5 min samples, which was resulted in the local stress concentration
and rapid work hardening (as evidenced in Fig. 8c and d), leading to a
continuous yielding (without YPE).
22
Materials Science & Engineering A 739 (2019) 17–25
3.4. Work hardening behavior and texture evolution
The relationship between variation in dislocation density and YPE
can be further reinforced through the study of work hardening behavior
and evolution of texture in ferrite. Fig. 8 shows the work hardening
exponent of 650 samples and 700–5 min sample. It is distinct that the
work hardening exponent of 650–5 min and 650–1 h samples remained
at a low value until the finish of YPE, corresponding to low dislocation
density in BCC structure, but high dislocation density in austenite. In
contrast, the work hardening exponent of 650–48 h and 700–5 min
samples increased rapidly during the initial stage of deformation, cor-
responding to simultaneous increase in dislocations in both BCC
structure and austenite. Moreover, it is noted that, at similar strain, the
dislocation density of austenite in 650–48 h and 700–5 min samples was
much lower compared to 650–5 min and 650–1 h samples. Therefore, it
is proven that the increase of dislocation density in austenite con-
tributed little to work hardening which was primarily contributed from
BCC structure. Austenite itself could not provide effective hardening,
but the accumulation of dislocations in austenite promotes more mar-
tensitic transformation with consequent enhancement in hardening. For
another, as martensite was surrounded by the soft austenite and ferrite,
the increase in dislocation density in the BCC structure was primarily
from ferrite, especially during the initial stage of deformation. Con-
sidering Figs. 6–8, it is deduced that when dislocations in austenite
Z.H. Cai et al.
Fig. 8. Work hardening exponent of (a) 650–5 min sample, (b) 650–1 h sample, (c) 650–48 h sample, (d) 700–5 min sample.
accumulate to a critical extent, the hardening induced by TRIP effect
overcomes local softening, which results in the finish of YPE. On the
basis of aforementioned discussion, it is confirmed that YPE is a con-
sequence of equally close competition between work hardening (via
martensitic transformation) and softening (by stress relaxation and
transfer) induced by the TRIP effect.
The difference in the evolution of ferrite dislocation density illu-
strated in Fig. 7 could be clarified by studying the texture of ferrite.
Fig. 9 shows contour plots of the orientation distribution function
(ODF) for ϕ2 = 45° and texture intensity of γ fiber. For 650–1 h sample,
the ferrite grains are characterized by a strong α fiber (∥ < 110 >), and
γ fiber (∥ < 111 >) with decreasing intensity. Comparing Fig. 9a with
c, it is distinct that 650–48 h sample had a similar intensity of α fiber to
650–1 h sample. However, the intensity of γ fiber was stronger in
650–48 h sample. In contrast, in 700–5 min sample, the ferrite grains
had a strong γ fiber, and α fiber trends to be weaker or nearly absent. It
was reported that the stored energy of α fiber was lower, thus the
crystal defects such as dislocations, dislocation cell walls were difficult
to be stored, which implied that the work hardening ability will be low
[33,34]. In comparison, the stored energy of γ fiber was higher, dis-
locations pile-up, tangle and cross-slip easily, resulting in high dis-
location density and work hardening ability. Moreover, α fiber has a
lower Taylor factor than γ fiber [35], indicating that ferrite with α fiber
23
Materials Science & Engineering A 739 (2019) 17–25
will deform preferentially and play a dominant role during the initial
stage of deformation, resulting in low dislocation density and work
hardening exponent. Therefore, the ferrite grains with γ fiber have high
dislocation storage ability, which is also responsible for the increase in
dislocation density and work hardening exponent, leading to decrease
of YPE.
4. Conclusions
(1) The YPE only appeared in samples annealed at 650 °C, and the
length of YPE decreased with increase in soaking time. The UFG or
globular microstructure was not the critical factor responsible for
YPE in 650 samples. YPE is a consequence of close competition
between the work hardening (via martensitic transformation) and
softening (by stress relaxation and transfer) induced by the TRIP
effect.
(2) The length of YPE decreased with decrease in austenite stability.
With increase in austenite stability, more dislocations were re-
quired to stimulate martensitic transformation. Austenite itself
could not provide effective hardening, but the accumulation of
dislocations in austenite promotes more martensitic transformation
with consequent enhancement in hardening.
(3) The formation of YPE was not affected by ferrite grain size but
Z.H. Cai et al.
Fig. 9. The contour plots of the orientation distribution function (ODF) for ϕ2 = 45° and texture intensity of γ fiber. (a, b) 650–1 h sample, (c, d) 650–48 h sample, (e,
f) 700–5 min sample.
texture. The ferrite grains with γ fiber have high dislocation storage
ability, which was also responsible for the increase in dislocation
density and work hardening exponent, leading to decrease of YPE.
Acknowledgements
The authors acknowledge support from the National Natural Science
Foundation of China (51501035 and 51574077), the Fundamental
Research Funds for the Central Universities (N170204017) and
Shanxi International Science and Technology Cooperation Program
24
Materials Science & Engineering A 739 (2019) 17–25
(2015081011). R.D.K. Misra gratefully acknowledges support from the
National Science Foundation, USA through grant # DMR 1602080.
R.D.K. Misra also acknowledges collaboration with Northeastern
University as an Honorary Professor.
References
[1] H.W. Luo, J. Shi, C. Wang, et al., Acta Mater. 59 (2011) 4002–4014.
[2] M.J. Merwin, Mater. Sci. Forum 539–543 (2007) 4327–4332.
[3] J. Shi, X.J. Sun, M.Q. Wang, et al., Scr. Mater. 63 (2010) 815–818.
[4] S. Lee, S.J. Lee, B.C. De Cooman, Scr. Mater. 65 (2011) 225–228.
Z.H. Cai et al. Materials Science & Engineering A 739 (2019) 17–25

[5] H. Lee, M.C. Jo, S.S. Sohn, et al., Acta Mater. 147 (2018) 247–260. [20] K. Steineder, D. Krizan, R. Schneider, et al., Acta Mater. 139 (2017) 39–50.
[6] M.M. Wang, C.C. Tasan, D. Ponge, et al., Acta Mater. 111 (2016) 262–272. [21] B.B. He, Z.Y. Liang, M.X. Huang, Scr. Mater. 150 (2018) 134–138.
[7] J.H. Ryu, J.I. Kim, H.S. Kim, et al., Scr. Mater. 68 (2013) 933–936. [22] A.K. Srivastava, D. Bhattacharjee, G. Jha, (et a), Mater. Sci. Eng. A 445–446 (2007)
[8] J.G. Speer, E. De Moor, K.O. Findley, et al., Metall. Mater. Trans. A. 42A (2011) 549–557.
3591–3601. [23] A. Grajcar, H. Krzton, J. Achiev. Mater. Manuf. Eng. 35 (2009) 169–176.
[9] Z.H. Cai, H. Ding, R.D.K. Misra, et al., Acta Mater. 84 (2015) 229–236. [24] A. Vinogradov, Adv. Eng. Mater. 17 (2015) 1710–1722.
[10] X.C. Xiong, B. Chen, M.X. Huang, et al., Scr. Mater. 68 (2013) 321–324. [25] Z.C. Li, R.D.K. Misra, H. Ding, et al., Mater. Sci. Eng. A 712 (2018) 206–213.
[11] G.K. Tirumalasetty, M.A. Van Huis, C. Kwakernaak, et al., Acta Mater. 60 (2012) [26] Z.Y. Liang, Y.Z. Li, M.X. Huang, Scr. Mater. 112 (2016) 28–31.
1311–1321. [27] W. Li, W.Z. Xu, X.D. Wang, J. Alloy Compd. 474 (2009) 546–550.
[12] H.F. Xu, J. Zhao, W.Q. Cao, et al., Mater. Sci. Eng. A 532 (2012) 435–442. [28] C.G. Dunn, E.F. Koch, Acta Metall. 5 (1957) 548–554.
[13] M. Callahan, O. Hubert, F. Hild, et al., Mater. Sci. Eng. A 704 (2017) 391–400. [29] P. Gay, P.B. Hirsch, A. Kelly, Acta Metall. 1 (1953) 315–319.
[14] F. Yang, H.W. Luo, E.X. Pu, et al., Int. J. Plast. 103 (2018) 188–202. [30] X.G. Wang, L. Wang, M.X. Huang, Acta Mater. 124 (2017) 17–29.
[15] A.H. Cottrell, B.A. Bilby, Dislocation theory of yielding and strain ageing of iron, [31] X.G. Wang, M.X. Huang, J. Iron Steel Res. Int. 24 (2017) 1073–1077.
Proc. Phys. Soc. A 62 (1949) 49e62. [32] J. Han, S.H. Kang, S.J. Lee, et al., J. Alloy Compd. 681 (2016) 580–588.
[16] Y. Tomota, A. Narui, N. Tsuchida, ISIJ Int. 48 (2008) 1107–1113. [33] R.L. Every, M. Hatherly, Texture 1 (1974) 183–194.
[17] C.E. Carlton, P.J. Ferreira, Acta Mater. 55 (2007) 3749–3756. [34] I.L. Dillamore, C.J.E. Smith, T.W. Watson, Mater. Sci. J. 1 (1967) 49–54.
[18] A.A. Nazarov, A.E. Romanov, R.Z. Valiev, Nanostruct. Mater. 4 (1994) 93–101. [35] N. Rajmohan, Y. Hayakawa, J.A. Szpunar, et al., Acta Mater. 45 (1997) 2485–2494.
[19] K.T. Park, D.H. Shin, Metall. Mater. Trans. A 33 (2002) 705–707.

25