Materials Science & Engineering A 923 (2025) 147741

Contents lists available at ScienceDirect

Materials Science & Engineering A

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

Excellent mechanical properties and underlined deformation twinning

mechanism of a low-carbon low-density cryogenic steel
Qingfeng Kang , Zexi Zhang , Ziyuan Gao , Cunyu Wang , Jianxiong Liang , Zhengdong Liu ,
Yuqing Weng , Wenquan Cao *
Special Steel Department, Central Iron and Steel Research Institute (CISRI), Beijing, 100081, China

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

Keywords: The mechanical properties and deformation behaviors of a low-carbon low-density steel of Fe35Mn6Al0.1C are
Low-carbon low-density steel studied under both room temperature (RT) and liquid nitrogen temperature (LNT). It is found that the yield
Excellent cryogenic toughness strength, tensile strength, total elongation are 251 MPa, 580 MPa and 52 % at RT, which increase 497 MPa, 945
Deformation twinning
MPa and to 75.5 % at LNT. The Charpy V-notch impact energy is 331±4 J at RT, which remains an excellent
Short-range ordering
High stacking fault energy
value of 254 ± 6 J at LNT, indicating an excellent combination of strength and ductility at LNT. The micro-
structural evolutions are examined by EBSD in SEM and TEM, it is shown that the RT tensile deformation is
mainly controlled by the dislocation substructure refinement accompanied by deformation twinning just before
the necking. However, the LNT deformation is controlled by both dislocation substructure refinement and
deformation twinning with twinning started just after yielding. The multiple strain hardenings are divided into
four stages with strain increasing at LNT, rather than two stages phenomena at RT. Deformation twinning is
found in the specimen after ~28 % tensile strain at RT, whereas it is found in the specimen after ~10 % tensile
deformation at LNT, which is seldom reported in low-density steel in literature. Based on the analysis of the
relation between work hardening behaviors and microstructure evolution, the high stacking fault energy (SFE)
twinning and its orientation dependence are mainly attributed to the stress concentration induced by planar
dislocation substructure. The exceptional combination of strength and toughness at LNT makes the low-carbon
low-density steel of Fe35Mn6Al0.1C to be a promising cryogenic steel.

1. Introduction strength and ductility [5,6,18–20].

The austenitic low-density steel with chemical composition of
Light weight steel as one of the new generation materials has been 0.5–1.5%C, 25–35%Mn and 6–11%Al has been thoroughly demon-
paid much more attention in recent two decades due to its combination strated in literature for the application in the automobile industry [2,5,
of excellent mechanical properties and low-density, with the aim to 15,19,21]. It is shown that the yield strength varied from 400 to 1000
understand the deformation mechanism in scientific research and MPa, tensile strength varied from 600 to 1000 MPa with total elongation
explore the possibility of industry application [1–6]. However, much in the range of 40–80 %. The excellent mechanical properties are mainly
work has been focused on the chemical design [3,7–10], the deforma- attributed to the thinning of the slip band spacing enclosed by the planar
tion mechanism [11–14] and the strengthening and toughening be- dislocations in the easy gliding plans [2,14,21]. It is reported that the
haviors of the low-density steel [3,15–18], such as the austenitic tensile deformation behavior of austenite-based duplex Fe-Mn-Al-C
low-density steel with high carbon, high manganese and high steels is primarily dominated by the behavior of the austenitic matrix
aluminum, the dual-phase low-density steel with relative lower contents [5,12]. Due to the deformation incompatibility between the two phases,
of carbon, manganese and aluminum, and the ferritic low-density steel the main crack sources originate from the FCC/BCC interfaces in duplex
with lowest content of carbon, manganese and aluminum. Amongst steels [4,18]. In addition, it is reported that the austenite grains can be
these three kinds of low-density steel, the austenitic low-density steel is refined by a small quantity of ferrite [1] and the mechanical properties
widely studied due its lowest density and the excellent combination of of austenite-based duplex low-density steels can be improved through

* Corresponding author.
E-mail address: [email protected] (W. Cao).

https://doi.org/10.1016/j.msea.2024.147741
Received 26 August 2024; Received in revised form 25 December 2024; Accepted 25 December 2024
Available online 28 December 2024
0921-5093/© 2024 Published by Elsevier B.V.
Q. Kang et al. Materials Science & Engineering A 923 (2025) 147741

solution treatment [1,2,5]. It is reported that the addition of carbon and Table 1
aluminum would results in a significant increase of stacking fault energy The compositions (wt.%) of the steel measured by wet-chemical analysis.
(SFE) [22–24], which is usually beneficial to the formation of the Element concentration (wt.%)
dislocation cell by the cross glide of the screw component of [25]. In
Mn Al C Nb V Mo W Fe
order to resolve this contrary, the term “glide plane softening” is pro-
posed and well interpreted by the interaction between dislocation and specimen 35.14 5.97 0.095 0.11 0.10 1.24 0.95 Bal.
γ 35.92 5.88 0.098 0.09 0.09 1.12 0.86 Bal.
short-range order (SRO) or clustering [14,26]. Another interesting α 32.44 6.83 – 0.09 0.18 1.54 1.63 Bal.
phenomenon is the twinning in the austenitic low-density steel, which
however usually take places in the austenite steel with relative lower
SFE of 20–45 mJ⋅m− 2, such as the TWIP steel. It is reported [13,27] that g⋅cm− ³.
deformation twinning is found in the region near the grain boundary at The phases fraction was determined by X-ray diffraction analysis
high strain during the tensile deformation of the austenitic low-density (Bruker D8 Focus) to confirm the stability of austenite after different
steel with SFE about 79–108 mJ⋅m− 2, which is explained by the local tensile strain. The microstructure was characterized in details by scan-
stress concentration enabling the critical stress for dislocation gliding ning electron microscope (SEM) and electron back-scattered diffraction
larger than that of twinning. However, other factors could also result in (EBSD) in a scanning electron microscope (JSM-7800 F). The original
the formation of the deformation twinning, such as deformation tem- data obtained from EBSD were analyzed by the HKL channel 5. The step
perature, grain orientation, Schmid factor, SRO and planar slip [27–33]. size of 1 μm, 300 nm, 100 nm and 30 nm were adopted for EBSD test.
Comparing with the extensive work has been done in the low-density The deformation substructures were investigated with the transmission
steel, few studies has been found for the low-density steel with low- electron microscopy operated at 200 kV (FEI-Tecnai G2 F20).
carbon content. This research trend may be explained by the fact that The geometrically necessary dislocation density can be estimated
high carbon content could result in much more reduction of the density with Eq. (1) from the EBSD data of the specimens after tensile defor-
of the low-density. Low temperature mechanical behavior of micro- mation with different tensile strain [33]:
alloyed and controlled-rolled Fe-Mn-Al-C-X alloys steel and also the /
ρGND = 2θ ub (1)
relative high strength and ductility as demonstrated in literature.
However, with increasing of the carbon content, the tendency of the in which u is the unit length for EBSD test, b is the Burgers vector, and
precipitation of κ-carbide is enhanced, which would impair the tough- θ is the average KAM value of the selected area.
ness and be detrimental to the welding properties of the high-carbon In order to measure the dislocation density in the specimens after
low-density steel. It is reported that a very high toughness about 200 J tensile deformation with different tensile strain, the average values of
[34] in the cryogenic temperature could be obtained in the low-density crystallite size D and micro-strain 〈ε2 〉1/2 were estimated according to
steel with a carbon content of 0.3 %, which making these low-carbon the Williamson–Hall equation [36,37] by using the profiles of (111)γ,
low-density steel suitable to the application as a cryogenic steel. How- (200)γ and (220)γ diffraction peaks after Gaussian curve fitting. The
ever, no detailed research on the microstructure evolution and me- dislocation densities ρ could be estimated by the following equation [36,
chanical responses have been done, which is essential to the application 37]:
of the low-carbon low-density steel in the cryogenic temperature. √̅̅̅̅̅̅
In this study, the austenite-based duplex low-density steel with of 3 2π〈ε2 〉1/2
ρ= (2)
0.1%C, 35%Mn and 6.0%Al with SFE (γSFE ) of 62.64 mJ⋅m− 2 is designed Db
and fabricated. The mechanical properties and the microstructure at
both room temperature and cryogenic temperature are explored for this where ε is the average micro-strain, D is average crystallite size, b is the
austenitic low-density steel. It could be expected that both extensive Burgers vector (0.26 nm).
deformation twinning and dislocation gliding could take place due to the Tensile tests were carried out on standard tensile specimens, with
high stack fault energy and short-range ordering as found in the high gauge length of 25 mm and diameter of 5 mm, cut from the rods of the
carbon low-density steel. The deformation microstructure of deforma- water quenched steel samples with the tensile axis direction parallel to
tion twinning and dislocation structure will be characterized by the the rod axis. The stress strain responses and corresponding strain
EBSD, TEM and XRD and the mechanical response of both tensile hardening rates of the solution-treated specimens were investigated
properties and impact toughness will be examined by the tensile test and from tensile tests carried out at an initial strain rate of 1.0 × 10− 3 s− 1 at
Charpy impact test. Our objective is to reveal the microstructure evo- room temperature (RT) and liquid nitrogen temperature (LNT) using an
lution during tensile deformation process at both ambient temperature MTS tensile machine. At least three tensile specimens were deformed
and cryogenic temperature and its effects on the mechanical properties. until fracture to confirm reproducibility. To reveal the microstructural
Both strengthening and toughening mechanism of the low-carbon low- evolution during straining, interrupted tensile tests to engineering
density steel will be discussed in this study. strains of 3 %, 10 %, 20 %, 28 %, 38 % and 55 % were performed. The
toughness of the samples was measured by the Charpy V-notch impact
test on the standard samples (55 mm in length, 10 mm in width, 10 mm
2. Materials and experimental design
in thickness, and a V-notch depth of 2 mm) and the test recording the
force-displacement curves and energy-displacement curves at RT and
Ingot with nominal compositions of Fe-35Mn-6Al-0.1C-0.1Nb-0.1V-
LNT.
1Mo-1W (in wt.%) was prepared by vacuum induction melting with
capacity of 50 Kg under high purity Ar atmosphere. The as-cast ingot
3. Results
was homogenized at 1150 ◦ C for 4 h, and subsequently hot-rolled (HR)
into rods with a diameter of 16.5 mm. The HR steel rods were then
3.1. Initial microstructures
solution-treated at 1050 ◦ C for 2 h followed by oil quenching to obtain
equiaxed grain structures with nearly equivalent grain sizes. Table 1
The microstructure of the starting materials after solution-treated at
presents the compositions of the solution-treated samples, austenite, and
1050 ◦ C for 2h is characterized by EBSD to reveal the phases, grain size,
ferrite. The composition of the samples was determined using wet-
dislocation and texture, which is shown in Fig. 1(a–c). The phase map is
chemical analysis. The compositions of austenite and ferrite were esti-
shown in Fig. 1(a) with face-centered cubic (FCC) γ austenite as the
mated using electron probe microanalysis, with the carbon content
matrix embedded with a small fraction of body-centered cubic (BCC) α
calculated using the lever rule [35]. The density of the investigated steel,
phase. The volume fraction of the α phase is approximately 3 %, with an
measured using the principle of Archimedes, is reported to be 7.18

2
Q. Kang et al. Materials Science & Engineering A 923 (2025) 147741

Fig. 1. Microstructure of the solution-treated steel. (a) EBSD phase map. (b) EBSD KAM map. (c) EBSD IPF map along the rolling direction. (d, e, f) Selected area
diffraction patterns along the [100]γ, [110]γ, and [110]MC zone axis, respectively.

average grain size of about 4.7 μm, uniformly distributed along the modulation caused by concentration fluctuations of solute Al and C
austenite grain boundaries. It can be seen that equiaxed γ grain struc- atoms [21,39]. Furthermore, based on examination of the precipitated
tures included many annealing twin boundaries depicted by white color. phase by the electrolytic extraction method, it is found that V is
The average grain size defined with misorientation angle no less than completely dissolved in the steel, Nb, Mo and W is partially dissolved in
15◦ is about 10.8 μm. The dislocation density in this starting material is the steel, with approximately 0.047 (wt.%) Nb, 0.017 (wt.%) Mo, and
revealed by the KAM values in Fig. 1(b), which is related to the geo- 0.024 (wt.%) W existing in MC carbide.
metric necessary dislocation (GND) and its average density is measured
to be is 1.6 × 1013 m− 2. The inverse pole figure (IPF) map along the
forging direction is shown in Fig. 1(c), indicating a very weak texture in 3.2. Mechanical properties
the starting material.
The microstructure of the starting materials after solution-treated at Fig. 2 (a) shows the typical tensile engineering stress-strain curves of
1050 ◦ C for 2 h is characterized by TEM to reveal the possible nano- the steels deformed 20 ◦ C and − 196 ◦ C. It is illustrated that both the
precipitation. Selected area electron diffraction patterns along the strength and ductility of the specimens tensile deformed − 196 ◦ C steel is
[100]γ, [110]γ, and [110]MC zone axis, are shown in Fig. 1 (d), (e) and significantly higher than that deformed at room temperature, indicating
(f), respectively. It can be seen from Fig. 1(d) that the extremely weak that the strength-ductility trade-off is evaded via low temperature. The
superlattice spots near the [0-11] diffraction patterns are observed, tensile mechanical properties of these steels were given in Table 2. As
which cannot be observed in the corresponding dark field image. This shown in this table, the 0.2 % offset yield strength (YS) increases from
weak superlattice diffraction spots have been revealed to be an L12 type 251 to 497 MPa, and the ultimate tensile strength (UTS) increases from
short-range ordering (SRO) in previous studies [3,14,21,27,38]. Satellite 580 MPa to 945 MPa as the temperature decreases from room temper-
reflections can be observed near some of the main γ diffraction spots. ature to liquid nitrogen temperature. It is also displayed that the low
Some previous studies indicated that satellite reflections adjacent to the temperature leads to the increase in the uniform elongation (UEL) from
{200}γ and {220}γ primary diffraction spots originate from lattice 38.5 % to 56.0 % and the enhancement in the total elongation (TEL)
from 52.0 % to 75.5 %. The Charpy V-notch (CVN) impact energy

Fig. 2. Mechanical properties at RT and LNT: (a) Tensile engineering stress-strain curves. (b) True stress-strain curves and corresponding strain hardening rate
curves. (c) Force-displacement curves and energy-displacement curves.

3
Q. Kang et al. Materials Science & Engineering A 923 (2025) 147741

Table 2
Mechanical properties at RT and LNT.
Test temperature Ultimate tensile strength, MPa Yield strength, MPa Uniform elongation, % Total elongation, % Reduction of area, % CVN impact energy, J

RT 580 ± 6 251 ± 3 38.5 ± 0.5 52.0 ± 0.5 78 ± 0.5 331 ± 4

LNT 945 ± 8 497 ± 5 56.0 ± 0.5 75.5 ± 0.5 80 ± 1 254 ± 6

reflects the toughness of the material, which is 331 ± 4 J at room weak texture of the austenite as shown in Fig. 3(a) before the tensile test,
temperature and 254 ± 6 J at cryogenic temperature, indicating an but the texture strength increases gradually with increasing of the ten-
excellent toughness at both RT and LNT for the low-carbon low-density sile deformation strain. Finally, a very strong <111>//TA texture and
steel. weak <100>//TA texture were observed in the samples tensile
The true stress-strain curves along with the corresponding strain deformed at RT as shown in Fig. 3(a). Similar texture evolution is found
hardening rate curves at RT and LNT were displayed in Fig. 2 (b). As in the samples tensile deformed at cryogenic temperature as shown in
shown in Fig. 2 (b), the strain hardening curves at RT and LNT can be Fig. 3(b). It should be pointed out, the texture strength of <111>//TA
divided into two and four stages, respectively, which indicates that the texture component at room temperature is slightly higher than that
hardening behavior at RT and LNT is governed by different mechanisms, deformed at cryogenic temperature. However, the texture strength of
which will be interpreted based on the microstructure examination in <111>//TA texture component at the maximum uniform elongation at
the discussion section. To further understand the differences in impact cryogenic temperature is significantly higher than that deformed at
toughness at RT and LNT, the instrumented impact test was applied to room temperature, which is clearly resulted from the higher tensile
reveal the different impact energy contribution from the crack initiation deformation strain at cryogenic temperature.
and crack propagation. The results were displaced in Fig. 2(c) by the Fig. 4 displays the black, red, blue and green lines representing the
impact force-displacement curves and energy-displacement curves, in
which the impact energy was calculated by the integration of the force-
displacement curve, which was taken as the impact energy. The crack
initiation energy Wm is defined by the integration of the force-
displacement curve from zero force to the maximum force Fm, and the
crack propagation energy Wp is defined by the integration of the force-
displacement curve from the maximum force Fm to the fracture force.
The total impact energy is calculated as the sum of crack initiation en-
ergy and the crack propagation energy as shown in Fig. 2(c). It is very
interesting that crack initiation is almost the same of ~80–100 J for both
room temperature deformation and cryogenic temperature deformation.
However, the crack propagation energy is changed significantly from
251 J at room temperature to 156 J at cryogenic temperature. This
means that the Charpy impact at both temperature is governed by crack
propagation energy.

3.3. EBSD analysis of tensile microstructural evolution

The texture evolutions of austenite along tensile axis (tensile axis is

shorten as TA) during tensile deformation at RT and LNT are examined
by EBSD and the results shown in Fig. 3(a) and (b). The IPF of the tensile
Fig. 4. Schmid factor evolution as a function of tensile deformation strain at
samples at room temperature are calculated from the orientation data
both RT and LNT.
measured by an in-situ EBSD tension experiments. It can be seen that a

Fig. 3. IPFs of austenite along the TA direction after different tensile strain and examined by EBSD. (a) 0 %, 3 %, 28 % and 44 % engineering strains at RT. (b) 3 %,
10 %, 28 % and 55 % engineering strains at LNT.

4
Q. Kang et al. Materials Science & Engineering A 923 (2025) 147741

evolution of the average {111} <110>γ Schmid factor and {111} Fig. 6 further illustrates the relationship between glide traces and
<112>γ Schmid factor with deformation at RT, {111} <110>γ Schmid grain subdivision. Interestingly, two types of dislocation glide traces as
factor and {111} <112>γ Schmid factor with deformation at LNT, shown in Fig. 6(a–c) and (d-f), double dislocation glide traces and single
respectively. It can be seen that the average {111} <110>γ Schmid dislocation slip traces, are formed. It can be seen that type II grains
factor at RT achieved its maximum value near 3 % tensile strain, while exhibit a propensity for the single slip traces, Type III grains demonstrate
the average {111} <112>γ Schmid factor at RT showed a decrease at the a tendency to develop double glide traces, whereas only a small fraction
same strain level. This means that in the early stage of deformation, the of type I grains exhibit slip traces. The study of Fe-30.5Mn-2.1Al-1.2C
activation of dislocation slip is easier than that of twinning. Upon (wt.%) steel [41] at 0.1 true strain demonstrates that grains with ori-
increasing the strain, the average {111} <112>γ Schmid factor sur- entations close to <111>//TA and <001>//TA are prone to forming a
passed the {111} <110>γ average Schmid factor at RT, which attributed Taylor lattice, while grains with orientations oriented along the line
to the development of <111>//TA and <100>//TA textures. It is very between <001>//TA and <111>//TA directions are inclined to form
interesting that both the average {111} <110>γ Schmid factor and the high-density dislocation walls. However, in low-carbon low-density
{111} <112>γ Schmid factor at LNT reach their maximum values at steel, the deformation microstructure of grains close to <001>//TA is
approximately 10 % engineering strain. This means that dislocation slip featured by the formation of the dislocation cells. This microstructure
and twinning are easier to occur in the early stage of deformation at difference between present steel and Fe-30.5Mn-2.1Al-1.2C (wt.%) steel
LNT. may be resulted from low-carbon content and high deformation rate,
Fig. 5 shows the orientation dependence of deformation mechanism which promote the cross glide in grains with <100>//TA.
of our studied steel deformed at both RT and LNT. It can be seen that the
deformation mechanisms are strongly dependent on the crystal orien-
tation, such as <001>//TA orientation with angle deviation of 3.4. TEM analysis of tensile microstructural evolution
approximately 15◦ (type I), <111>//TA orientation with angle devia-
tion of approximately 15◦ (type III) and the rest apart from type I and III For the understanding of the deformation behaviors of our studied
(type II) as defined by I. Gutierrez-Urrutia [40]. Fig. 5(a1) illustrates the steel at RT, the deformation microstructures are examined by TEM after
correlation between deformation twinning and grain orientation after different deformation strain and the results are displayed in Fig. 7. It can
38 % tensile deformation at room temperature. Similar to the study by I. be seen, when the strain reaches to 3 %, complex dislocation configu-
Gutierrez Urrutia et al., twins are more likely to occur in the type II and rations are formed, such as dislocation nodes, multiple dislocations and
type III grains. Due to the very small fraction of the grains with defor- dislocation pairs (Fig. 7(a)), which are supposed to be related to planar
mation twinning, the tensile deformation of our studied steel at RT is slip [21]. After 10 % deformation, Lomer-Cottrell (LC) locks and the
controlled by dislocation gliding. high-density dislocation walls (HDDWs) are formed (Fig. 7(b and c)),
Fig. 5(b1, c1, d1) illustrates the correlation between deformation which are regularly observed in other FCC materials [21,42]. After 20 %
twinning and grain orientation at 28 %, 38 %, and 55 % engineering strain, the so-called Taylor lattices, cell blocks (CBs) and dislocation
strain at LNT, respectively. Similar to room temperature deformation, cells (DCs) are formed along {111} slip plane (Fig. 7(d–f)). It should be
twins are more readily formed in type II and type III grains. As shown in noted that Taylor lattice is not popular compared to dislocation cell and
Fig. 5(b1, b2, c1, c2, d1, d2), plenty deformation twins could be found in cell block. This implies that dislocations are more prone to cross slip.
the microstructure when the studied steel is deformed at LNT. And the Furthermore, grains (type I) oriented close to <100>//TA are more
deformation twinning fraction increased with the increasing of the prone to forming dislocation cells (Fig. 7(f)), while grains (type II and
deformation strain. Thus, the deformation twinning plays an important III) oriented away from <100>//TA are more likely to develop cell
role in the deformation behaviors of our studied low-carbon low-density blocks (Fig. 7(e)). Fig. 7(g–j) shows that the CBs and DCs are refined as
steel at LNT temperature, which is strongly dependent on the grain the strain increases to 28 %. The refinement of cell blocks also has been
orientations. It can be observed in Fig. 5(c2) that type II grains tend to observed in TWIP and lightweight steels [40,41]. A small amount of
form single orientation twinning, while type III grains tend to form deformation twinning has been formed along the cell block walls
multiple orientation twinning. (CBWs), as shown in Fig. 7(g and h). Furthermore, a few microbands
(Fig. 7(j)) are found, which is an indication of stress localization.

Fig. 5. The dependence behaviors dependence on the crystal orientation of the low-carbon low-density steel during tensile deformation at both RT and LNT. (a1, b1,
c1, d1) Inverse pole maps along the TA direction displaying grain orientations with red, green and blue dots corresponding to type I, II and III grains, respectively, in
which the orientations with mtw < ms are labeled. (a2, b2, c2, d2) Inverse pole maps along the TA direction. (For interpretation of the references to color in this figure
legend, the reader is referred to the Web version of this article.)

5
Q. Kang et al. Materials Science & Engineering A 923 (2025) 147741

Fig. 6. Dislocation glide induced grain subdivision in samples of 28 % engineering strains at RT (a–c) and LNT (d–f). (a, d) IPF maps along the TA direction. (b, e)
KAM maps. (c, f) Inverse pole figure along the TA direction displaying sub-grain orientation map with dislocation glide traces (green dots: single glide; black dots:
double glide). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 7. TEM images of deformation induced microstructure at different tensile strains at RT, indicating occurrence of deformation twinning at 28 % tensile strain: (a)
3 %, (b, c) 10 %, (d–f) 20 %, (g–j) 28 %, (k, l) 38 %.

Although deformation twins paralleling to the primary slip plane are complex planar slip features, dislocation pairs, nodes, and multiple
further developed at 38 % strain level, the proportion of deformation dislocations emerged, as displayed in Fig. 8 (a). Unlike at room tem-
twins is still very low (Fig. 7(k, l)), which is consistent with the finding perature, the occurrence of extended dislocation pairs at low-
by EBSD as shown in Fig. 5(a1, a2). temperature indicates that low temperature facilitates planar glide. It
Deformation-induced microstructures at LNT are also examined by is particularly intriguing that stacking faults are formed at the annealed
TEM and the results are shown in Fig. 8. Upon reaching a strain of 3 %, twin boundaries, as displayed in the inset of Fig. 8(b). This implies that

6
Q. Kang et al. Materials Science & Engineering A 923 (2025) 147741

Fig. 8. TEM images of deformation induced microstructure at different tensile strains at LNT, indicating the occurrence of deformation twin at 10 % tensile strain: (a,
b) 3 %, (c–e) 10 %, (f–i) 20 %, (j–o) 28 %, (p, q) 38 %, (r–t) 55 %.

low temperature has facilitated the formation of stacking faults at early stacking faults formed at the annealed twin boundaries. Upon straining
deformation. At a strain of 10 %, the features of complex planar slip are to 20 %, dislocation cell blocks (Fig. 8(e)), and dislocation cells (Fig. 8
further pronounced, as depicted in Fig. 8(c). Fig. 8(d and e) illustrates (g)), are produced respectively in grains oriented away from <100>//
that deformation twins have been nucleated at annealed twin bound- TA and close to <100>//TA, which is similar to observation at room
aries upon reaching an approximate strain of 10 %. Typically, the width temperature. Furthermore, at 20 % strain, deformation twins and
of twin bundles is about 5 nm for 10 % strain (Fig. 8(e)), which cannot be dislocation cells have been developed in the interior of the grains, but
detected by EBSD. It should be noted that, unlike the deformation twins the proportion of twins is still very low (Fig. 8(h and i)), indicating a
formed at room temperature through high-density dislocations, these competition between dislocation gliding and deformation twinning in
deformation twins formed at low strain are closely related to the this stage.

7
Q. Kang et al. Materials Science & Engineering A 923 (2025) 147741

Similar to the dislocation substructure at RT, the dislocation cell The EBSD characterizations are carried out in in the typical crack
blocks are refined at 28 % strain level at LNT (Fig. 8(j)). Table 3 displays propagation zone, as indicated by the black dots in the inset of Fig. 9
the average sizes of the dislocation substructures at room temperature (b–g). To gain a deeper understanding of the substructure following
and under liquid nitrogen. It can be observed that at equivalent strains, impact deformation, TEM analysis is also performed, as shown in Fig. 9
the sizes of dislocation substructures are reduced under liquid nitrogen (d–i). Fig. 9(b–g) illustrates that the samples are highly stable under
compared to room temperature conditions (Table 3). This implies that both room temperature and liquid nitrogen temperature, with no phase
the dislocation density is higher at liquid nitrogen temperature under transformation occurring in the austenite. Dislocation substructures at
the same strain conditions. Upon reaching an engineering strain of 28 %, both temperatures are characterized by the dislocation cells or cell
while twins are further activated, their prevalence remains relatively blocks (Fig. 9(d–i), which is similar to the dislocation substructure under
low, as illustrated in Fig. 8(j and k). In addition, a small amount of wide tension. It can be observed that low temperature can promote the for-
microbands (Fig. 8(l, m)) and fine microbands (Fig. 8(n, o)) are mation of deformation twins, as shown in Fig. 9(c, d, h, i). Although
observed. The broad microbands can attain widths of approximately twins are more readily activated under low-temperature impact, they do
390 nm, and the fine microbands measure around 137 nm in width, as not burst as in TWIP steels, leading to a non-significant contribution of
shown in Fig. 8(l, n), respectively. It can be observed that these broad twins to the impact energy. Additionally, the red arrows in Fig. 9(g)
microbands and fine microbands originate from grain boundaries and display the presence of distinct fine cracks around the voids generated
within the grains, respectively, as illustrated in the inset of Fig. 8(l) and by the detachment of α for cryogenic impact sample. This indicates that
(o). Some studies indicate that the Fe-Mn-Al-C low-density steel [20,41, the γ/α interface is served as the crack initiation site for cryogenic
43], high manganese steel [29] and 316L stainless steel [44] exhibited impact sample, which may contribute to a decrease in crack propagation
plastic localization phenomena related to microbands during energy at low temperature. The impact energy is reduced about 30 %
room-temperature or liquid nitrogen temperature deformation. As the from 331J at room temperature to 254 J at liquid nitrogen temperature
strain is further increased to 38 %, a significant number of deformation due to the existence of 3.0 % percent of the α-phase in the austenite
twins are activated, as shown in Fig. 8(p, q). When the strain is reached matrix. Fig. 9 (e, j) shows that the fracture surfaces are mainly composed
55 %, closing to the maximum force strain, a large number of secondary of dimples of micrometer or sub-micrometer at RT and LNT,indicating
twins are activated, as displayed in Fig. 8(r, s). Fig. 8 (t) shows that, that both temperatures are ductile fracture. Therefore, although the
compared to primary deformation twinning, secondary twins contain presence of the γ/α interface leads to a slight decrease in impact energy
more stacking faults. at low temperatures compared to room temperature, the refined
Based on the microstructure evolution examination as shown in austenite grains [1,29], absence of κ-carbides [5,32], and stability of
Fig. (7, 8), it could be reasonably assumed that the deformation twin- austenite [29] endow this austenite-based duplex lightweight steel with
ning takes place at about 10 % tensile strain at LNT. However, the high impact toughness at the both temperatures.
deformation twinning is taken place at tensile strain of 28 % at RT,
indicating a strong dependence of deformation twinning on both 4. Discussion
deformation strain and deformation temperature.
4.1. Advantages to be one of the cryogenic steels and stacking fault energy
3.5. Deformation structure of impact specimens at both RT and LNT assessment of the low-carbon low-density steel

To further understand the suppression or elimination of the ductile to The cryogenic steel is a special steel applied under the condition with
brittle transition in impact toughness from RT to LNT, the surface very low temperature, specially the LNT. The selection of the cryogenic
morphology of the impact fracture and the structure in the vicinity of the steel is very conservative, requiring not only the excellent cryogenic
fracture at RT (Fig. 9(a–e)) and LNT (Fig. 9(f–j)) are analyzed, as shown toughness but also the high ductility to assure the safety of the cryogenic
in Fig. 9. For the RT sample, a large number of subordinate cracks and equipment during shipping and storage [45]. Thus, the asthenic steel
large necking are seen (Fig. 9(a)), whereas only a small number of such as the high-Mn TWIP-steel rather than the conventional 9Ni steel is
subordinate cracks and small necking are seen in the cryogenic sample much more welcome due to its high cryogenic toughness and excellent
(Fig. 9(f)). This means that a multitude of subordinate cracks and large ductility at LNT [45,46].
necking contribute to the high crack propagation energy at room For the revealing of the advantages of our designed low-carbon low-
temperature. density steel, the properties of the cryogenic steel were collected from
literature and compared with our new designed steel as shown in Fig. 12.
Table 3 It can be seen from Fig. 10(a) that the Charpy impact energy and specific
Stress, dislocation density and dislocation substructure size in different strains at yield strength of the steel at liquid nitrogen temperature and those of
RT and LNT. other steels (converted to full size 10 × 10 × 55 mm). It can be observed
Engineering 3.0 10.0 20.0 28.0 38.0 55.0 that, compared to high carbon lightweight steel [47,48] and TWIP steel
strain (%) [46], the designed low-carbon lightweight steel exhibits a higher Charpy
True strain, % 3.0 9.5 18.2 24.7 32.2 43.8
impact energy. In contrast to low-carbon high manganese steel [29] and
True stress (RT), 339 483 635 723 800 – 304-stainless steels [49], the designed low-carbon lightweight steel
MPa demonstrates a superior specific yield strength. 304-stainless steel is
True stress 598 778 1007 1148 1280 1463 mechanically unstable and undergoes martensitic transformation at low
(LNT), MPa
temperatures, leading to a significant reduction in its impact energy.
ρGND (RT-EBSD- 0.71 3.05 5.64 7.92 8.99 –
α), 1014 m− 2 Although the duplex structure plays a crucial role in the fracture
ρ (RT-XRD), 1014 0.64 2.42 4.96 6.96 8.03 ​ behavior of steel, the presence of a small amount of ferrite can signifi-
m− 2 cantly refine the austenite grains, which substantially enhances its
ρGND (LNT-EBSD- 0.97 2.59 5.95 6.93 9.17 9.73 impact toughness [29]. Moreover, compared to martensite, ferrite is
α), 1014 m− 2
more readily deformed, which may alleviate stress concentration at the
ρ (LNT-XRD), 0.72 2.94 5.66 7.90 9.31 13.32
1014 m− 2 γ/α interface to some extent. The strength and toughness of our designed
CBs or DCs size – – 448 282 – – lightweight steel are comparable to those of the Al0.1CoCrFeNi
(RT), nm ± 74 ± 56 high-entropy alloy [50]. The ductility and the specific yield strength of
CBs or DCs size – – 429 242 – – the steel at liquid nitrogen temperature and those of other steels are
(LNT), nm ± 76 ± 54
shown in Fig. 10(b). The ductility of our studied steel is in the same level

8
Q. Kang et al. Materials Science & Engineering A 923 (2025) 147741

Fig. 9. Deformation microstructure of samples after impact testing at RT (a–e) and LNT (f–g). (a, f) Macroscopic fracture appearance after testing. (b, g) Phase maps
near the fracture surface. (c, h) IPF map along the impact direction near the fracture surface. (d, i) TEM maps of the region approximately 1000um from the fracture
surface. (e, j) Fracture surface maps.

Fig. 10. Comparison of our studied lightweight steel with reported low-carbon low-density steel [34], high carbon lightweight steel [47,48], low-carbon high
manganese steel [29], TWIP steel [46], 304 stainless steels [49], 9Ni steel [51], and HEAs [50] at liquid nitrogen temperature. (a) Charpy impact energy versus
specific yield strength. (b) Total elongation versus specific yield strength.

comparing with the austenite TWIP steel, but nearly 3 times higher than However, deformation twinning is not only found in the specimens
that of the 9Ni-steel, indicating a strong resistance of deformation tensile deformed at room temperature as shown in Figs. 4 and 8, in the
induced rupture. Furthermore, although 9Ni-steel has been well recog- specimens tensile deformed at liquid nitrogen temperature as shown in
nized as a cryogenic alloy [51], the price of our designed low-carbon Figs. 5 and 9, and in the impact deformed specimens at both RT and LNT
low-density steel is much lower that of the conventional 9Ni-steel, due as shown in Fig. 11. However, due to the low content of carbon of 0.10 %
to its low-cost elements alloying and simple high temperature solution and relative low aluminum content of 6 %, the stacking fault energy and
heat treatment as revealed in the experiment section in this research. the austenite stability should be assessed for the tendency of deforma-
There are still challenges of high Ni alloy cost and complex heat treat- tion twinning and or deformation induced phase transformation during
ment processes. Similarly, other stainless steel and high-entropy alloy is deformation process.
more expensive because of their high Co or Ni alloying design. Based on The SFE (γ SFE ) and ΔGγ→ε of the austenite in our low-carbon light-
above comparison, the advantages of low-density, excellent cryogenic weight steel were evaluated based on the well-accepted thermodynamic
toughness and ductility and low-cost of our designed low-carbon low-- approach [52,53]:
density steel are very impressive comparing with other cryogenic steels.
γ SFE = 2ρΔGγ→ε + 2σ (3)
These advantages make this steel to be one of the promising materials as
cryogenic application. where γSFE is for stacking fault energy, ρ is the molar surface density
In order to understand the excellent cryogenic properties of the low- along {111} planes, ΔGγ→ε is the molar Gibbs energy of the phase
carbon low-density steel as mentioned above, the stacking fault energy transformation from γ to ε, and σ is the interfacial energy per unit area of
and austenite stability are calculated and discussed as follows. It is well the γ/ε phase boundary. According to the approximate model of solid
recognized that the addition of aluminum into the low-density steel solution, The Gibbs energy difference of the ΔGγ→ε in Fe-Mn-Al-C alloy
could increase the stack fault energy significantly, which in turn sup- can be expressed as [24,53]:
pressed the occurrence of twins in the low-density steel [22,24].

9
Q. Kang et al. Materials Science & Engineering A 923 (2025) 147741

Fig. 11. Austenite phase stability of our studied low-carbon low-density steel checked by XRD examination and phase diagram calculation. (a) XRD profiles of the
specimens after different tensile strains at LNT and (b) Phase diagram calculated by Thermal-Calc software (using TCFE12 database).

Table 4
γSFE and ΔGγ→ε at RT and LNT.
2 1
Temperature γSFE /mJ⋅m− ΔGγ→ε /J⋅mol−

RT 62.64 804.22
LNT 49.46 576.90

The SFE (γSFE ) and ΔGγ→ε of the designed lightweight cryogenic steel at
LNT are 49.46 mJ⋅m− 2 and 576.90 J mol− 1. Austenitic steels with me-
dium stacking fault energy (SFE) of 20–45 mJ⋅m− 2 are typically prone to
the formation of deformation twins [2,24]. When the SFE is below 20
mJ⋅m− 2 and above 45 mJ⋅m− 2, twins are replaced by martensitic
transformation and dislocations slip, respectively. Based on this calcu-
lation, the SFEs are higher than that of the requirement of the taking
Fig. 12. Taylor factor M distribution sketch along the TA direction in this steel. place of the deformation twinning.

FeMnAlC = XFe ΔGFe +XMn ΔGMn +XAl ΔGAl + XC ΔGC +XFe XMn ΩFeMn +XFe XAl ΩFeAl +XFe XC ΩFeC +XMn XAl ΩMnAl +XMn XC ΩMnC +XAl XC ΩAlC +ΔGMag (4)
ε γ→ε γ→ε γ→ε γ→ε
γ→ε γ→ε γ→ε γ→ε γ→ε γ→ε γ→ε
ΔGγ→

where Xi is the atomic molar fraction and Ω is the mixing energy. In the In order to check the austenite stability of our designed low-carbon
current lightweight steel, the possible interactions of Mn-Al and Al-C are low-density steel, the phase identification is done by XRD and reex-
negligible [24]. ΔGγ→mag is the magnetic transformation Gibbs energy,
ε
amined by the calculation of the phase diagram using thermal-Calc
which can be solved by formulas (5) ⁓ (8) [53]. software. Both results are demonstrated in Fig. 11. It can be seen from
Fig. 11(a) that the XRD profiles of the LNT deformed specimens do not
ΔGγ→ε ε γ
mag = ΔGmag − ΔGmag (5) vary with the deformation strain, showing the volume fraction of about
97 % and the existence of the BCC-phase with a small fraction. It can be
or ε
ΔGγmag = RTln(β(γ or ε) + 1)f(τ(γ or ε) ) (6) seen that the volume fraction of both phase in the low-carbon low-
density steel are very stable even when it was deformed at LNT. The
where R is the gas constant, T is the test temperature (T = 293 or 77K), stability of the phase is also proved that a small fraction of the BCC phase
TNγ or ε is Nèel temperature, β is the magnetic moment, and f(τ(γ or ε) ) is a is presented in phase diagram as shown in Fig. 11(b).
function of the scaled Nèel temperature (τ(γ or ε) = T/ TNγ or ε ). Based on the calculation of the stacking fault energy, the phase
( −5 ) fraction at LNT and the phase diagram as shown above, the stability of
τ τ− 15 τ− 25 our designed low-carbon low-density steel is very high, which theoret-
f(τ) = − + + D− 1 , if τ > 1 (7)
10 315 1500 ically cannot be deformed by deformation twinning and phase trans-
[ ( )] formation from austenite to martensite. However, deformation twins are
79τ− 1 474 ( − 1 ) τ3 τ9 τ15
f(τ) = 1 − + p − 1 + + D− 1 , if τ⩽1 (8) taken place both at RT and LNT, indicating a controversy between SFE
140p 497 6 315 600
prediction twins and experimental observation for our studied low-
carbon low-density steel, which is strongly dependent on the deforma-
where p = 0.28 and D = 2.34 for FCC [38]. By using thermodynamic
tion strain, deformation temperature (as shown in Figs. 9 and 10),
functions and parameters listed in literature [38].
crystal orientation of the grains (as shown in Fig. 7) and will be dis-
The calculated γSFE and ΔGγ→ε at RT and LNT were recorded in
cussed in section 4.2.
Table 4. It can be seen that the SFE (γSFE ) and ΔGγ→ε of the designed
lightweight cryogenic steel at RT are 62.64 mJ⋅m− 2 and 804.22 J mol− 1.

10
Q. Kang et al. Materials Science & Engineering A 923 (2025) 147741

4.2. Deformation twinning mechanism of low-density steel with high SFE model cannot interpret the twinning dependence on not only orientation
of <111>//TA and <110>//TA but on the deformation strain. The
Austenitic steel with medium SFE (20–45 mJ⋅m− 2) is typically prone Taylor factor M distribution sketch along the TA direction in this steel is
to the formation of deformation twins [2,24]. The spontaneous forma- shown in Fig. 12. It can be seen that twinning is primarily activated in
tion of deformation twinning is unlikely occurred in the low-density high Taylor factor grains (type II and III). Therefore, it is reasonable to
steel because of its high stacking fault energy, especially at room tem- speculate that SFE is not the only decisive factor controlling the twin-
perature [11,12,14]. Interestingly, our designed high-stacking fault ning process for our designed low-carbon low-density steel.
energy lightweight steel, with a room temperature SFE of 62.64 mJ⋅m− 2 The grain orientation dependence of twinning has also been
and a liquid nitrogen temperature SFE of 49.46 mJ⋅m− 2, has been observed in TWIP steel and high carbon low aluminum lightweight steel
generated deformation twins at both room temperature and liquid ni- [28,40,41]. Deformation twinning of type III grain is commonly
trogen temperature. The deformation twinning from this study is explained via Schmid law of twin dislocation slip [58]:
demonstrated strongly dependent on the deformation temperature and
τtw/s = σ cos Φ cos λ (11)
crystal orientation. In order to understand this interesting point, the
critical stress for deformation twinning is calculated based on different
where τtw/s is twin stress/dislocation slip stress, mtw/s = cos Φ cos λ is the
deformation principles as follows. The critical stress of twinning (σt ) can
twin/dislocation slip Schmid factor, σ is the macroscopic stress in MPa,
be estimated via a mean-field model [13,33].
Φ is the angle between the twinning plane normal/dislocation slip plane
( )
Γ Gb normal and the tensile axis, and λ is the angle between the twinning
σt = M + (9) shear direction/dislocation slip direction and the tensile axis. It can be
b D
observed that massive twinning is activated in the grains with orienta-
where M is the Taylor factor for a randomly textured polycrystal (M = tion type III assume higher twinning Schmid factor than that of dislo-
2.2 based on the Sachs models [54], M = 2.7 based on the crystal cation Schmid factor. However, Schmid law cannot explain the twinning
plasticity finite element model [55] and M = 3.06 based on the Taylor behavior of type I and II grains. Only a small number of deformation
model under an assumption of random orientation [54] and M = 3.64 twinning is formed at grain boundary in type I grains, though their
for both orientation of <111>//TA and <110>//TA, Γ is the SFE Schmid factor for dislocation gliding is lower than that for deformation
(62.64 mJ⋅m− 2 and 49.46 mJ⋅m− 2 at RT and LNT, respectively), G is the twinning activation. Type II grains do not obey Schmid law, but they do
polycrystal shear modulus (68 GP [56]), b is the magnitude of the Bur- exhibit a large amount of deformation twinning. Gutierrez-Urrutia and
gers vector of the partial dislocation (0.147 nm [33]), and D is the grain Raabe attributed the deformation twinning in these unfavorably ori-
size (~10.8 μm). The critical twinning stress is calculated according to ented grains (type II) in TWIP steel to stress concentrations induced by
Eq. (9) with different models and orientations and the results are given the impact of twins in neighboring grains [58]. However, they did not
in Table 5. Furthermore, Table 5 also displays the critical twinning stress consider the influence of dislocation structures on the flow stress. In fact,
estimated by a Hall-Petch-type relation which is well applied in TWIP type II grains, which generate twins, are scarcely impacted by the twins
steel [57,58]. of adjacent grains, as displayed in Fig. 5(c1-c3).
( ) Such twinning observed in this high SFE steel is similar to that found
1 γ K
σt = + 0.5 (10) in Cu–10Mn (at. %) alloy [31], Fe-22Mn-x (x = 0,3,6) Al-0.6C [62] and
m b D
Fe30Mn11Al1.2C(wt.%) steels [27]. Twinning cannot be activated for
copper in traditional deformation processes, but the incorporation of 10
Where m is the average Schmid factor (0.326) [58], b is the Burgers
at. % Mn can induce twin formation. The authors attributed this to the
vector (2.5 × 10− 10 m), γ is the SFE, K is Hall-Petch constant (0.350
presence of SRO, which promotes planar slip causing that critical
MPa⋅m0.5), K is estimated by linear fitting of yield strength (267, 258,
twinning stress can be easily reached by stress concentration [31]. Park
251 and 245 MPa) and D10.5 (D is average grain size (8.2, 9.3, 10.8 and
et al. [62] also attributed planar slip to the glide softening associated
12.1 μm)) using solid solution temperatures of 1000 ◦ C, 1025 ◦ C,
with SRO. In addition, they have qualitatively proved that the interac-
1050 ◦ C and 1075 ◦ C. The KGB(γ) value we estimated is very close to those
tion between dislocation and SRO can lower the critical stress for
of TWIP steels (0.357 MPa⋅m0.5 [59] and 0.330 MPa⋅m0.5 [60]). twinning in the Fe-22Mn-x (x = 3,6) Al-0.6C steel [62]. Ren et al. [27]
It is quite interesting that the experimental critical twinning stresses obtained similar findings that the high flow stress induced by dislocation
at RT is about 723 MPa, which is far smaller than the critical twinning planar slip and the reduction in critical twinning stress due to the
stress of 939/1153/1306/1096 MPa calculated from Sachs models/ presence of SRO may be the causes of the twinning phenomenon. In
crystal plasticity finite element model/Taylor model/Hall-Petch-type current work, the microstructure contains SRO (Fig. 1(d)). Furthermore,
relation. Similarly, experimental critical twinning stresses at LNT is no typical mechanical antiphase domain boundaries (APBs) bounded by
about 778 MPa, which is very close to the critical tinning stress calcu- two partials are observed in this steel. The short-range order in our
lated from the Sachs model, but far smaller than that of 910/1032/934 designed steel is similar to that of Fe-22Mn-x (x = 3,6) Al-0.6C steel
MPa calculated from crystal plasticity finite element model/Taylor [62]. Thus, the short-range order in our steel may reduce the stacking
model/Hall-Petch-type relation. It seems that the deformation twinning fault energy and critical twinning stress rather than increasing them.
can be predicted by the Sachs model, which is mainly determined by the Type II and III grains form cell block walls with planar slip character-
stacking fault energy at LNT. However, this explanation based on Sach istics, which lead to stress concentration required for twinning nucle-
ation and growth (Fig. 7(g–k)). Type I grains form uniform dislocation
cells, which are not conducive to the formation of the high stress
Table 5 required for twinning growth. The shear stress arising from stress con-
Estimated and experimental critical twinning stresses in this steel at both
centration at these cell block walls significantly exceeds the macroscopic
temperatures.
shear stress. The homogeneous dislocation cell structure in type I grains
Temperature σt (M = σt (M = σt (M = σt = Experimental effectively disperses the macroscopic shear stress. Thus, type II and III
(
2.2)/ 3.06)/ 2.7)/ 1 γ σt /MPa
MPa MPa MPa
+ grains, which are not favorable for twin activation, paradoxically
m b
K
) generate a substantial number of twins. Conversely, type I grains, which
D0.5 are conducive to twin initiation, only produce a small number of twins at
RT 939 1306 1153 1096 723
the grain boundaries, which are governed by the twin nucleation stress.
LNT 742 1032 910 934 778 Therefore, the high stacking fault energy (SFE) twinning and its

11
Q. Kang et al. Materials Science & Engineering A 923 (2025) 147741

orientation dependence are mainly attributed to the stress concentration ferrite content constitutes about 3 %. Therefore, the flow stress is esti-
induced by planar dislocation substructure. mated by Eq. (12):
σ mod = 0.97σmod(γ) + 0.03σ mod(α) (12)
4.3. Multistage hardening behaviors and flow stress contribution
The flow stress of σmod(γ) and σmod(α) after yielding can be predicted by
As it is shown in Fig. 2(b), the work hardening rate in stage A is Eqs. (13) and (14) [30,40,58]:
sharply decreased with increasing of tensile strain at both RT and LNT.
This sharp decreasing phenomenon could be easily interpreted by the σ mod(γ) = σf(γ) + σGB(γ) + σ twin(γ) + σ εdis(γ) − σ0dis(γ) (13)
dislocation accumulation and recovery [30,61]. However, the work
hardening rate at LNT is much higher than that at RT, which could be σ mod(α) = σ YS(α) + σεdis(α) − σ 0dis(α) (14)
attributed to the increased lattice friction stress with decreasing of
deformation temperature as reported found in TWIP steel [30]. This where σmod(γ) is the predicted flow stress of austenite, σf(γ) is austenite
explanation is consistent with our calculation that the lattice friction lattice friction stress, σGB(γ) is austenite initial grain boundary
stresses are 143 MPa for RT deformation but 389 MPa for LNT defor- strengthening contribution, σtwin(γ) is austenite deformation twinning
mation as revealed in subsequent calculations. contribution, σ εdis(γ or α) and σ 0dis(γ or α) are the strengthening contribution
Stage B hardening at LNT is characterized by almost constant in the from dislocations at ε strain and 0 strain, σ YS(α) is the ferrite yield
strain hardening rate. For low aluminum, high carbon lightweight steels, strength.
the Taylor lattice observed at low strains is ascribed to the solute carbon Grain-boundary strengthening is calculated with the classical Hall-
effect on diminishing the frequency of dislocation cross-glide [41]. Fig. 8 Petch formula (Eq. (15)) [65,66], the Hall-Petch coefficient KGB(γ)
(a–c) shows that the cross-slip of dislocations is notably enhanced, (0.350 MPa m0.5) is used. Initial grain boundary strengthening is about
which diminishes the propensity for the formation of a Taylor lattice. 108 MPa.
Therefore, the formation of a Taylor lattice is not favored in the designed
lightweight steel. Consequently, the approximately constant strain σ GB(γ) = KGB(γ) d− 0.5
(15)
hardening rate exhibited during this phase cannot be explained by the
Lattice friction stress is calculated by Eq. (16) [30], in which σYS(γ) is
Taylor lattice. At 3 % engineering strain, distinct stacking faults are
yield strength. The lattice friction stresses at RT and LNT are 143 MPa
produced at the annealed twin boundaries (Fig. 8(b)). Upon reaching 10
and 389 MPa, respectively.
% engineering strain, pronounced deformation twins are evident at the
annealed twin boundaries (Fig. 8(c)), a result of the previously formed σ f(γ) = σYS(γ) − σGB(γ) (16)
stacking faults (Fig. 8(b)). As detailed in Section 4.2, the true stress at 10
The contribution of twin stress can be described by the following
% engineering strain has surpassed the calculated lower bound of the
equation [30,67], where F is the volume fraction of twins in type II and
critical twinning stress. Furthermore, during this phase, an increasing
III grains, 0.8 is the average fraction of type II and III grains at 55 %
trend in the {111} <112>γ Schmid factor is observed, which augments
engineering strain, l is the average twin spacing, Ktwin(γ) is the Hall-Petch
the twinning partition stress and favors twin activation. Consequently,
the nearly constant high strain hardening rate observed in this stage is coefficient of twinning (0.350 MPa⋅m0.5 [58]).
primarily attributed to the activation of twin at the annealed twin Δσtwin(γ) = 0.8FKtwin(γ) l− 0.5
(17)
boundary.
Stage C is characterized by a continuous decrease in strain hardening F in Eq. (16) can be estimated by the stereological analysis of Full-
rate associated with the refinement of the dislocation substructure. As man (Eq. (18)) [30,68], where e is the average thickness of twins. The
discussed in Section 4.2, typical dislocation cells are formed in type I twin boundary volume fraction is usually estimated by TEM [30,68,69].
grains, while dislocation blocks are produced in Type II and Type III 1 1 F
grains. This hardening mechanism, driven by the refinement of the = (18)
l 2e 1 − F
dislocation substructure, manifests as a sustained reduction in the strain
Based on TEM image, l and e were measured to be 182 nm and 72 nm,
hardening rate [41]. Although this stage has reached the lower limit of
respectively, when engineering strain is 55 % at LNT. Utilizing Eqs. (17)
critical stress under liquid nitrogen condition, the continuous decrease
and (18), twin strengthening contribution is calculated to be 290 MPa.
in the {111} <112> γ Schmid factor results in less active twinning
The dislocation hardening (σdis(γ or α) ) can be calculated by Taylor
development. Consequently, the strain hardening behavior in this stage
equation (Eq. (19)) [70], where M is the Taylor factor, α is a constant
is primarily governed by the refinement of the dislocation substructure.
used to measure the interaction strength of dislocations, G is the shear
During the initial phase of Stage D, the true stress at 38 % engi-
modulus, b is the magnitude of the Burgers vector, and ρ is the dislo-
neering strain under liquid nitrogen temperature (LNT), reaching 1280
cation density. For ferrite, M is 2.8, α is 0.2, G is 80.7 GPa, b is 0.248 nm
MPa, has substantially surpassed the critical twinning stress of Taylor
[71,72]. Due to the ferrite content being only about 3 %, it is not possible
model. As a result, primary twins are substantially activated under these
to estimate its dislocation density using the XRD method. Therefore,
elevated stress conditions. Upon further straining to 55 %, the primary
deformation twins are further enhanced, and the secondary twins are ferrite dislocation density is estimated using ρGND (Table 3) and contri-
also significantly activated. The twin boundaries serve as effective bar- bution of ferrite dislocation strengthening to the flow stress is displayed
riers to dislocation motion [28,30,33], significantly contributing to the in Table 6.
sustained hardening response observed during Stage D. Therefore, the
continuous strain hardening in Stage D is ascribed to the extensive
activation of primary and secondary twins. Table 6
The strengthening contribution of σdis and σ twin strengthening to the flow stress.
It is widely known that during the deformation process, twin
boundary effectively acted as barriers to dislocation glide by reducing Engineering strain (%) 3.0 10.0 20.0 28.0 38.0 55.0
their mean free path, known as the “dynamic Hall Petch effect” [28,30, True strain (%) 3.0 9.5 18.2 24.7 32.2 43.8
33,59]. However, Liang and Zhi et al. [63,64] demonstrated that σdis(γ) , (RT), MPa 48 207 347 434 474 –
dislocation evolution dominated the work hardening rate of TWIP steels σdis(α) , (RT-GND), MPa 1 4 6 8 8 –
instead of twins. Tang et al. [30] held that dislocation hardening and σdis(γ) , (LNT), MPa 58 240 379 469 504 646
σdis(α) , (LNT-GND), MPa 2 4 6 7 8 9
TWIP effects worked synergistically. Therefore, it is necessary to further
σtwin(γ) , (LNT), MPa – – – – – 290
clarify the contribution of twins and dislocations to work hardening. The

12
Q. Kang et al. Materials Science & Engineering A 923 (2025) 147741

Table 7 mechanical response at both RT and LNT were examined systematically.

σdiff at two strain levels using different α values at RT. Different from the extensively studied low-density steel with a high
α Engineering strain (%) σdiff (MPa) carbon content around 1.0 %, massive deformation twinning was found
not only in the tensile deformation mode but in the impact deformation
0.136 10 152
0.136 20 280
mode, which gives the combination of excellent strength, ductility and
0.26 10 88 V-notch impact toughness at both RT and LNT. Based on the study of the
0.26 20 146 microstructure evolution at both RT and LNT, the factors affecting the
0.40 10 15 deformation twinning, including SFE, temperature, crystallographic
0.40 20 25
orientation, Schmid factor, and planar glide were analyzed thoroughly.
The main conclusions can be summarized as follows.
σ dis(γ or α) = MαGbρ0.5 (19)
1 In present low-carbon low-density steel, a two-stage deformation
For austenite, M is 3.06, α is 0.4 [63], G is 68 GPa [56], b is 0.26 nm, mechanism is observed for RT deformation and four-stage defor-
and the dislocation density ρ is shown in Table 3. It should be noted that mation mechanism is revealed for LNT deformation. Its RT behavior
different α values have been used for calculation, such as 0.136 [64], is dominated by dislocation planar slip, cells or cell blocks forming
0.26 [37,73], and 0.4 [63]. Therefore, different α Values of 10 % and 20 and refinement, followed with twinning occurring just before the
% engineering strain at RT are verified by Eqs. 18 and 19 in our study, as necking at RT. However, its LNT deformation behavior is found to be
shown in Table 7. σdiff is the difference between the experimental started with the dislocation planar slip firstly, primary twinning
measured stress (σ exp ) and modelling stress (σ mod ), which can be calcu- secondly, cells or cell blocks forming and refinement and finally
lated by Eq. (20) [30]: second deformation twinning at LNT, indicating a strong deforma-
tion twinning dependence on deformation temperature and defor-
σ diff = |σexp − σ mod | (20)
mation strain.
According to the results in Tables 7 and it is evident that when α is 2. Regarding to the deformation twinning, the type III grains with
0.4, the σdiff is the smallest among the three commonly used α values. orientation <111>//TA direction are more inclined to develop
The contributions of σtwin , and σ dis to flow stress calculated with the double orientation twins. But few deformed twinning is activated in
above method are shown in Table 6. It can be seen that the contribution type I grains with orientation <001>//TA and the type II grains with
of a small amount of ferrite to the flow stress is very minimal, less than orientation far from type III and I are more susceptible to the for-
10 MPa. Therefore, in the calculation of the flow stress, the contribution mation of single orientation twins. Regarding to the dislocation
of ferrite is neglected. The flow stress is calculated based on the fully substructure, the type III grains are characterized by a tendency to
austenitic structure,as shown in Fig. 13(a and b). It can be observed that form double dislocation glide traces. The type I grains are distin-
the calculated contributions of dislocations and twins to the flow stress guished by the formation of typical dislocation cells. The type II
are in excellent agreement with the measured values (Fig. 13(a and b)). grains exhibit a preference for the development of single dislocation
At room temperature, only a small number of deformation twins are glide traces.
generated at high strains; hence, deformation is primarily controlled by 3. The working hardening behaviors and strengthening contribution
dislocation hardening, which is the main driving force for strain hard- are strongly affected by the microstructure evolution at both RT and
ening. In contrast, under liquid nitrogen conditions, a synergistic effect LNT. Multiple strain hardening mechanisms of the lightweight
between dislocation hardening and twinning mechanisms is observed cryogenic steel were observed at LNT. It is revealed that an almost
(the shaded area in Fig. 13(b) represents the contribution of twins), with constant hardening rate was obtained in the early stage of defor-
both contributing to the material’s strain hardening behavior. In sum- mation due to the primary deformation twinning. Then a continuous
mary, this study provided a valuable insight into the TWIP effect and decrease of the hardening rate resulting from the heavily activated
dislocation substructure inducing complex deformation mechanisms in cross slip. For high strain, a constant work hardening is reached
cryogenic lightweight steels. again due to the extensive activation of primary twinning and sec-
ondary twinning, which in turn facilitates multi-stage strain hard-
5. Conclusion ening. During deformation at room temperature, dislocation
hardening but not the deformation twinning is the principal
In this research, the low-carbon low-density steel of Fe-35Mn-6Al- contributor to strain hardening.
0.1C (wt.%) was designed and its microstructure evolution and

Fig. 13. Contributions of various strengthening mechanisms to flow stress at different temperatures: (a) Room temperature (RT). (b) Liquid nitrogen tempera-
ture (LNT).

13
Q. Kang et al. Materials Science & Engineering A 923 (2025) 147741

4. An excellent combination of the strength, ductility and impact [7] L. Bartlett, D. Van Aken, High manganese and aluminum steels for the military and
transportation industry, J. Miner. Met. Mater. Soc. 66 (1989) 1770–1784, https://
toughness is obtained in the designed low-carbon low-density steel.
doi.org/10.1007/s11837-014-1068-y, 2014.
It is shown that yield strength, tensile strength, ductility and impact [8] L. Zhang, R. Song, C. Zhao, F. Yang, Work hardening behavior involving the
tough are 251 MPa, 580 MPa, 50 % and 331J at RT, and 497 MPa, substructural evolution of an austenite–ferrite Fe–Mn–Al–C steel, Mater. Sci. Eng.,
945 MPa, 75.5 % and 254J at LNT, which is superior than that not A 640 (2015) 225–234, https://doi.org/10.1016/j.msea.2015.05.108.
[9] S.S. Sohn, B.-J. Lee, S. Lee, N.J. Kim, J.-H. Kwak, Effect of annealing temperature
only the conventional 9Ni-steel due to both excellent mechanical on microstructural modification and tensile properties in 0.35 C–3.5 Mn–5.8 Al
properties but the low-cost, but better than that of the conventional lightweight steel, Acta Mater. 61 (2013) 5050–5066, https://doi.org/10.1016/j.
austenitic steel due to it low-density. Thes advantages make the actamat.2013.04.038.
[10] R. Rana, C. Liu, R.K. Ray, Low-density low-carbon Fe–Al ferritic steels, Scripta
designed low-carbon low-density steel of Fe–35Mn–6Al–0.1C (wt.%) Mater. 68 (2013) 354–359, https://doi.org/10.1016/j.scriptamat.2012.10.004.
to be one of the promising candidates of the cryogenic steel in the [11] J.D. Yoo, S.W. Hwang, K.-T. Park, Origin of extended tensile ductility of a Fe-
future. 28Mn-10Al-1C steel, Metall. Mater. Trans. A 40 (2009) 1520–1523, https://doi.
org/10.1007/s11661-009-9862-9.
5. The high stacking fault energy (SFE) twinning and its orientation [12] G. Frommeyer, U. Brüx, Microstructures and mechanical properties of high-
dependence are mainly attributed to the stress concentration strength Fe-Mn-Al-C light-weight TRIPLEX steels, Steel Res. Int. 77 (2006)
induced by planar dislocation substructure. Type II and III grains 627–633, https://doi.org/10.1002/srin.200606440.
[13] Z. Wang, W. Lu, F. An, M. Song, D. Ponge, D. Raabe, Z. Li, High stress twinning in a
form cell block walls with planar slip characteristics, which led to compositionally complex steel of very high stacking fault energy, Nat. Commun. 13
stress concentration required to activate twinning. Type I grains (2022), https://doi.org/10.1038/s41467-022-31315-2.
formed uniform dislocation cells, which are not conducive to the [14] E. Welsch, D. Ponge, S.M. Hafez Haghighat, S. Sandlöbes, P. Choi, M. Herbig,
S. Zaefferer, D. Raabe, Strain hardening by dynamic slip band refinement in a high-
formation of high stress required to activate twinning. The interac-
Mn lightweight steel, Acta Mater. 116 (2016) 188–199, https://doi.org/10.1016/j.
tion between dislocation and SRO may lower the critical twinning actamat.2016.06.037.
stress and promote planar slip of dislocations. [15] P. Ren, X.P. Chen, C.Y. Wang, Y.X. Zhou, W.Q. Cao, Q. Liu, Evolution of
microstructure, texture and mechanical properties of Fe–30Mn–11Al–1.2C low-
density steel during cold rolling, Mater. Char. 174 (2021) 111013, https://doi.org/
CRediT authorship contribution statement 10.1016/j.matchar.2021.111013.
[16] F. Wang, M. Song, M.N. Elkot, N. Yao, B. Sun, M. Song, Z. Wang, D. Raabe,
Shearing brittle intermetallics enhances cryogenic strength and ductility of steels,
Qingfeng Kang: Writing – original draft, Methodology, Formal
Science 384 (2024) 1017–1022, https://doi.org/10.1126/science.ado2919.
analysis, Data curation. Zexi Zhang: Validation, Methodology, Investi- [17] Y.F. An, X.P. Chen, L. Mei, X.N. Tan, P. Ren, X.Y. Zhang, W.Q. Cao, Overcoming the
gation. Ziyuan Gao: Supervision, Methodology, Investigation. Cunyu strength-ductility trade-off dilemma in austenitic lightweight steel via stepwise
controllable intragranular dual nanoprecipitation, Mater. Sci. Eng., A 908 (2024)
Wang: Software, Resources, Data curation. Jianxiong Liang: Software,
146773, https://doi.org/10.1016/j.msea.2024.146773.
Resources, Project administration. Zhengdong Liu: Validation, Re- [18] J. Chen, J. Ren, Z. Liu, Deformation microstructures as well as strengthening and
sources, Methodology. Yuqing Weng: Resources, Methodology, toughening mechanisms of low-density high Mn steels for cryogenic applications,
Conceptualization. Wenquan Cao: Writing – review & editing, Super- J. Mater. Res. Technol. 13 (2021) 947–961, https://doi.org/10.1016/j.
jmrt.2021.05.018.
vision, Resources, Methodology, Funding acquisition, [19] Z. Gao, Q. Kang, X. An, H. Wang, C. Wang, W. Cao, Enhanced mechanical
Conceptualization. properties of a Fe-Mn-Al-C austenitic low-density steel by increasing hot-rolling
reduction, Mater. Char. 204 (2023) 113237, https://doi.org/10.1016/j.
matchar.2023.113237.
Declaration of competing interest [20] I. Gutierrez-Urrutia, A. Shibata, Effect of deformation temperature on strain
localization phenomena in an austenitic Fe-30Mn-6.5Al-0.3C low-density steel,
Acta Mater. (2023) 119566, https://doi.org/10.1016/j.actamat.2023.119566.
The authors declare that they have no known competing financial [21] H. Zhi, J. Li, W. Li, M. Elkot, S. Antonov, H. Zhang, M. Lai, Simultaneously
interests or personal relationships that could have appeared to influence enhancing strength-ductility synergy and strain hardenability via Si-alloying in
medium-Al FeMnAlC lightweight steels, Acta Mater. 245 (2023) 118611, https://
the work reported in this paper.
doi.org/10.1016/j.actamat.2022.118611.
[22] W. Song, T. Ingendahl, W. Bleck, Control of strain hardening behavior in high-Mn
Acknowledgement austenitic steels, Acta Metall. Sin. 27 (2014) 546–556, https://doi.org/10.1007/
s40195-014-0084-9.
[23] W. Song, W. Zhang, J. von Appen, R. Dronskowski, W. Bleck, κ-Phase formation in
This research is supported by the Natural Nature Science Foundation Fe-Mn-Al-C austenitic steels, Steel Res. Int. 86 (2015) 1161–1169, https://doi.org/
of China with granted item number of 51871062 and 51371052. 10.1002/srin.201400587.
[24] A. Saeed-Akbari, J. Imlau, U. Prahl, W. Bleck, Derivation and variation in
composition-dependent stacking fault energy maps based on subregular solution
Data availability model in high-manganese steels, Metall. Mater. Trans. A 40 (2009) 3076–3090,
https://doi.org/10.1007/s11661-009-0050-8.
[25] D. Kuhlmann-Wilsdorf, Q: dislocations structures — how far from equilibrium? A:
Data will be made available on request.
very close indeed, Mater. Sci. Eng., A 315 (2001) 211–216, https://doi.org/
10.1016/s0921-5093(01)01204-7.
References [26] V. Gerold, H.P. Karnthaler, On the origin of planar slip in f.c.c. alloys, Acta Metall.
37 (1989) 2177–2183, https://doi.org/10.1016/0001-6160(89)90143-0.
[27] P. Ren, X.P. Chen, M.J. Yang, S.M. Liu, W.Q. Cao, Effect of early stage of κ-carbides
[1] H. Wang, C. Wang, J. Liang, A. Godfrey, Y. Wang, Y. Weng, W. Cao, Effect of
precipitation on tensile properties and deformation mechanism in high Mn–Al–C
alloying content on microstructure and mechanical properties of Fe–Mn–Al–C low-
austenitic low-density steel, Mater. Sci. Eng., A 857 (2022) 144132, https://doi.
density steels, Mater. Sci. Eng., A 886 (2023) 145675, https://doi.org/10.1016/j.
org/10.1016/j.msea.2022.144132.
msea.2023.145675.
[28] H. Beladi, I.B. Timokhina, Y. Estrin, J. Kim, B.C. De Cooman, S.K. Kim, Orientation
[2] H. Wang, Z. Cao, Z. Gao, C. Wang, J. Liang, A. Godfrey, L. Zhang, G. Wu, W. Cao,
dependence of twinning and strain hardening behaviour of a high manganese
Synergetic strengthening from dynamic slip band-grain boundary interaction in a
twinning induced plasticity steel with polycrystalline structure, Acta Mater. 59
low-density FeMnAlC steel, Mater. Sci. Eng., A 862 (2023) 144498, https://doi.
(2011) 7787–7799, https://doi.org/10.1016/j.actamat.2011.08.031.
org/10.1016/j.msea.2022.144498.
[29] Y. Wang, Y. Zhang, A. Godfrey, J. Kang, Y. Peng, T. Wang, N. Hansen, X. Huang,
[3] K.H. Yang, W.K. Choo, Evidence of carbon ordering and morphology change in a
Cryogenic toughness in a low-cost austenitic steel, Commun Mater 2 (2021),
cubic carbide phase, Phil. Mag. Lett. 62 (1990) 221–226, https://doi.org/10.1080/
https://doi.org/10.1038/s43246-021-00149-8.
09500839008215062.
[30] L. Tang, L. Wang, M. Wang, H. Liu, S. Kabra, Y. Chiu, B. Cai, Synergistic
[4] F. Yang, R. Song, Y. Li, T. Sun, K. Wang, Tensile deformation of low density duplex
deformation pathways in a TWIP steel at cryogenic temperatures: in situ neutron
Fe–Mn–Al–C steel, Mater. Des. 76 (2015) 32–39, https://doi.org/10.1016/j.
diffraction, Acta Mater. 200 (2020) 943–958, https://doi.org/10.1016/j.
matdes.2015.03.043.
actamat.2020.09.075.
[5] S. Chen, R. Rana, A. Haldar, R.K. Ray, Current state of Fe-Mn-Al-C low density
[31] X. Ma, J. Chen, Y. Liu, X. Wang, S. Huang, Z. Chen, Effect of short-range order on
steels, Prog. Mater. Sci. 89 (2017) 345–391, https://doi.org/10.1016/j.
microstructure, texture and strain hardening of cold drawn Cu-10 at.%Mn alloy,
pmatsci.2017.05.002.
Mater. Char. 135 (2018) 32–39, https://doi.org/10.1016/j.matchar.2017.11.024.
[6] S. Bai, Y. Chen, X. Liu, H. Lu, P. Bai, D. Li, Z. Huang, J. Li, Research status and
development prospect of Fe–Mn–C–Al system low-density steels, J. Mater. Res.
Technol. 25 (2023) 1537–1559, https://doi.org/10.1016/j.jmrt.2023.06.037.

14
Q. Kang et al. Materials Science & Engineering A 923 (2025) 147741

[32] I. Kalashnikov, O. Acselrad, A. Shalkevich, L.C. Pereira, Chemical composition [53] S. Curtze, V.-T. Kuokkala, A. Oikari, J. Talonen, H. Hänninen, Thermodynamic
optimization for austenitic steels of the Fe-Mn-Al-C system, J. Mater. Eng. Perform. modeling of the stacking fault energy of austenitic steels, Acta Mater. 59 (2011)
9 (2000) 597–602, https://doi.org/10.1361/105994900770345430. 1068–1076, https://doi.org/10.1016/j.actamat.2010.10.037.
[33] B.C. De Cooman, Y. Estrin, S.K. Kim, Twinning-induced plasticity (TWIP) steels, [54] R. Asaro, V. Lubarda, Mechanics of Solids and Materials, Cambridge University
Acta Mater. 142 (2018) 283–362, https://doi.org/10.1016/j.actamat.2017.06.046. Press, 2006, https://doi.org/10.1017/cbo9780511755514.
[34] Y.G. Kim, Y.S. Park, J.K. Han, Low temperature mechanical behavior of [55] Y. Tadano, M. Kuroda, H. Noguchi, Quantitative re-examination of Taylor model
microalloyed and controlled-rolled Fe-Mn-Al-C-X alloys, Metall. Trans. A 16 (1985) for FCC polycrystals, Comput. Mater. Sci. 51 (2012) 290–302, https://doi.org/
1689–1693, https://doi.org/10.1007/bf02663026. 10.1016/j.commatsci.2011.07.024.
[35] J.H. Hwang, T.T.T. Trang, O. Lee, G. Park, A. Zargaran, N.J. Kim, Improvement of [56] Z. Wang, W. Lu, H. Zhao, C.H. Liebscher, J. He, D. Ponge, D. Raabe, Z. Li,
strength – ductility balance of B2-strengthened lightweight steel, Acta Mater. 191 Ultrastrong lightweight compositionally complex steels via dual-nanoprecipitation,
(2020) 1–12, https://doi.org/10.1016/j.actamat.2020.03.022. Sci. Adv. 6 (2020) eaba9543, https://doi.org/10.1126/sciadv.aba9543.
[36] G.K. Williamson, R.E. Smallman III, Dislocation densities in some annealed and [57] M.A. Meyers, O. Vöhringer, V.A. Lubarda, The onset of twinning in metals: a
cold-worked metals from measurements on the X-ray debye-scherrer spectrum, constitutive description, Acta Mater. 49 (2001) 4025–4039, https://doi.org/
Phil. Mag. 1 (1956) 34–46, https://doi.org/10.1080/14786435608238074. 10.1016/s1359-6454(01)00300-7.
[37] G. Dini, R. Ueji, A. Najafizadeh, S.M. Monir-Vaghefi, Flow stress analysis of TWIP [58] I. Gutierrez-Urrutia, S. Zaefferer, D. Raabe, The effect of grain size and grain
steel via the XRD measurement of dislocation density, Mater. Sci. Eng., A 527 orientation on deformation twinning in a Fe–22wt.% Mn–0.6wt.% C TWIP steel,
(2010) 2759–2763, https://doi.org/10.1016/j.msea.2010.01.033. Mater. Sci. Eng., A 527 (2010) 3552–3560, https://doi.org/10.1016/j.
[38] J.D. Yoo, K.-T. Park, Microband-induced plasticity in a high Mn–Al–C light steel, msea.2010.02.041.
Mater. Sci. Eng., A 496 (2008). [59] F. de las Cuevas, M. Reis, A. Ferraiuolo, G. Pratolongo, L.P. Karjalainen, J. Alkorta,
[39] W.K. Choo, J.H. Kim, J.C. Yoon, Microstructural change in austenitic Fe-30.0wt% J. Gil Sevillano, Hall-petch relationship of a TWIP steel, KEM 423 (2009) 147–152.
Mn-7.8wt%Al-1.3wt%C initiated by spinodal decomposition and its influence on https://doi.org/10.4028/www.scientific.net/kem.423.147.
mechanical properties, Acta Mater. 45 (1997) 4877–4885, https://doi.org/ [60] K.M. Rahman, V.A. Vorontsov, D. Dye, The effect of grain size on the twin initiation
10.1016/s1359-6454(97)00201-2. stress in a TWIP steel, Acta Mater. 89 (2015) 247–257, https://doi.org/10.1016/j.
[40] I. Gutierrez-Urrutia, D. Raabe, Dislocation and twin substructure evolution during actamat.2015.02.008.
strain hardening of an Fe–22wt.% Mn–0.6wt.% C TWIP steel observed by electron [61] U.F. Kocks, H. Mecking, Physics and phenomenology of strain hardening: the FCC
channeling contrast imaging, Acta Mater. 59 (2011) 6449–6462, https://doi.org/ case, Prog. Mater. Sci. 48 (2003).
10.1016/j.actamat.2011.07.009. [62] K.-T. Park, K.G. Jin, S.H. Han, S.W. Hwang, K. Choi, C.S. Lee, Stacking fault energy
[41] I. Gutierrez-Urrutia, D. Raabe, Multistage strain hardening through dislocation and plastic deformation of fully austenitic high manganese steels: effect of Al
substructure and twinning in a high strength and ductile weight-reduced addition, Mater. Sci. Eng., A 527 (2010) 3651–3661, https://doi.org/10.1016/j.
Fe–Mn–Al–C steel, Acta Mater. 60 (2012) 5791–5802, https://doi.org/10.1016/j. msea.2010.02.058.
actamat.2012.07.018. [63] H. Zhi, C. Zhang, S. Antonov, H. Yu, T. Guo, Y. Su, Investigations of dislocation-
[42] J. Wang, J. Zou, H. Yang, X. Dong, P. Cao, X. Liao, Z. Liu, S. Ji, Ultrastrong and type evolution and strain hardening during mechanical twinning in Fe-22Mn-0.6C
ductile (CoCrNi)94Ti3Al3 medium-entropy alloys via introducing multi-scale twinning-induced plasticity steel, Acta Mater. 195 (2020) 371–382, https://doi.
heterogeneous structures, J. Mater. Sci. Technol. 135 (2023) 241–249, https://doi. org/10.1016/j.actamat.2020.05.062.
org/10.1016/j.jmst.2022.06.048. [64] Z.Y. Liang, Y.Z. Li, M.X. Huang, The respective hardening contributions of
[43] I. Gutierrez-Urrutia, A. Shibata, K. Tsuzaki, Microstructural study of microbands in dislocations and twins to the flow stress of a twinning-induced plasticity steel,
a Fe-30Mn-6.5Al-0.3C low-density steel deformed at cryogenic temperature by Scripta Mater. 112 (2016) 28–31, https://doi.org/10.1016/j.
combined electron channeling contrast imaging and electron backscatter scriptamat.2015.09.003.
diffraction, Acta Mater. 233 (2022) 117980, https://doi.org/10.1016/j. [65] H. Conrad, G. Schoeck, Cottrell locking and the flow stress in iron, Acta Metall. 8
actamat.2022.117980. (1960) 791–796, https://doi.org/10.1016/0001-6160(60)90175-9.
[44] C. Zhang, D. Juul Jensen, T. Yu, Microstructure and texture evolution during cold [66] A. Etienne, V. Massardier-Jourdan, S. Cazottes, X. Garat, M. Soler, I. Zuazo,
rolling of 316L stainless steel, Metall. Mater. Trans. A 52 (2021) 4100–4111, X. Kleber, Ferrite effects in Fe-Mn-Al-C triplex steels, Metall and Mat Trans A 45
https://doi.org/10.1007/s11661-021-06367-6. (2013) 324–334, https://doi.org/10.1007/s11661-013-1990-6.
[45] O. Grässel, L. Krüger, G. Frommeyer, L.W. Meyer, High strength Fe–Mn–(Al, Si) [67] R.S. Ganji, P. Sai Karthik, K. Bhanu Sankara Rao, K.V. Rajulapati, Strengthening
TRIP/TWIP steels development — properties — application, Int. J. Plast. 16 (2000) mechanisms in equiatomic ultrafine grained AlCoCrCuFeNi high-entropy alloy
1391–1409, https://doi.org/10.1016/s0749-6419(00)00015-2. studied by micro-and nanoindentation methods, Acta Mater. 125 (2017) 58–68,
[46] S.S. Sohn, S. Hong, J. Lee, B.-C. Suh, S.-K. Kim, B.-J. Lee, N.J. Kim, S. Lee, Effects of https://doi.org/10.1016/j.actamat.2016.11.046.
Mn and Al contents on cryogenic-temperature tensile and Charpy impact properties [68] R.L. Fullman, Measurement of particle sizes in opaque bodies, J. Miner. Met. Mater.
in four austenitic high-Mn steels, Acta Mater. 100 (2015) 39–52, https://doi.org/ Soc. 5 (1989) 447–452, https://doi.org/10.1007/bf03398971, 1953.
10.1016/j.actamat.2015.08.027. [69] P. Zhou, Z.Y. Liang, R.D. Liu, M.X. Huang, Evolution of dislocations and twins in a
[47] R.K. You, P.-W. Kao, D. Gan, Mechanical properties of Fe-30Mn-10Al-1C-1Si alloy, strong and ductile nanotwinned steel, Acta Mater. 111 (2016) 96–107, https://doi.
Mater. Sci. Eng., A 117 (1989) 141–148, https://doi.org/10.1016/0921-5093(89) org/10.1016/j.actamat.2016.03.057.
90095-6. [70] O. Bouaziz, N. Guelton, Modelling of TWIP effect on work-hardening, Mater. Sci.
[48] K.T. Luo, P.-W. Kao, D. Gan, Low temperature mechanical properties of Eng., A 319–321 (2001) 246–249, https://doi.org/10.1016/s0921-5093(00)
Fe–28Mn–5Al–1C alloy, Mater. Sci. Eng., A 151 (1992) L15–L18, https://doi.org/ 02019-0.
10.1016/0921-5093(92)90195-7. [71] D.J. Dyson, S.R. Keown, A study of precipitation in a 12 %Cr-Co-Mo steel, Acta
[49] C. Zheng, W. Yu, Effect of low-temperature on mechanical behavior for an AISI 304 Metall. 17 (1969) 1095–1107, https://doi.org/10.1016/0001-6160(69)90054-6.
austenitic stainless steel, Mater. Sci. Eng., A 710 (2018) 359–365, https://doi.org/ [72] T. Liu, Z. Cao, H. Wang, G. Wu, J. Jin, W. Cao, A new 2.4 GPa extra-high strength
10.1016/j.msea.2017.11.003. steel with good ductility and high toughness designed by synergistic strengthening
[50] D. Li, Y. Zhang, The ultrahigh charpy impact toughness of forged AlxCoCrFeNi high of nano-particles and high-density dislocations, Scripta Mater. 178 (2019)
entropy alloys at room and cryogenic temperatures, Intermetallics 70 (2016) 285–289, https://doi.org/10.1016/j.scriptamat.2019.11.045.
24–28, https://doi.org/10.1016/j.intermet.2015.11.002. [73] K. Jeong, J.-E. Jin, Y.-S. Jung, S. Kang, Y.-K. Lee, The effects of Si on the
[51] J.R. Strife, D.E. Passoja, The effect of heat treatment on microstructure and mechanical twinning and strain hardening of Fe–18Mn–0.6C twinning-induced
cryogenic fracture properties in 5Ni and 9Ni steel, Metall. Trans. A 11 (1980) plasticity steel, Acta Mater. 61 (2013) 3399–3410, https://doi.org/10.1016/j.
1341–1350, https://doi.org/10.1007/bf02653488. actamat.2013.02.031.
[52] S. Curtze, V.-T. Kuokkala, Dependence of tensile deformation behavior of TWIP
steels on stacking fault energy, temperature and strain rate, Acta Mater. 58 (2010)
5129–5141, https://doi.org/10.1016/j.actamat.2010.05.049.

15