Low Density Fe–Mn–Al–C Steels: Phase Structures, Mechanisms and Properties

Abstract

This review introduces the structural phases, microstructural characteristics, and the most relevant room and cryogenic properties of low density Fe–Mn–Al–C steels. The combination of outstanding physical and mechanical properties while offering a weight reduction of up to 18% make low density Fe–Mn–Al–C steels attractive structural materials as lightweight crash-resistant car body structures and structural components in the cryogenic industry. In this review, the latest alloy design strategies are introduced. In particular, the novel aspects of the phase structures and their deformation behavior, in particular, those related to L’1₂ (Fe, Mn)₃AlC carbides (κ-carbides) and B2-type Ni/Cu-rich precipitates, are critically summarized. Future scientific and technical challenges are provided to establish these steels as structural materials for industrial applications.

1. Introduction

Low density Fe–Mn–Al–C steels are one of the emerging class of structural materials for applications in the automotive, chemical, and aircraft industries. These steels exhibit outstanding tensile mechanical performance at room and cryogenic temperatures while offering weight reduction of up to 18% owing to the high Al content (1.3% density reduction per 1 wt.% Al addition, Fig. 1.¹⁾) In addition, these steels exhibit attractive properties such as high energy absorption behavior, high strength, and toughness at room and low temperatures, good fatigue, and good oxidation resistance at elevated temperatures.^{2,3,4,5,6,7,8,9,10,11,12,13)} Fe–Al–Mn–C steels were initially developed in the ‘80s and ‘90s as inexpensive replacement of Fe–Cr–Ni–C stainless steel due to the valuable effect of Mn and Al on the mechanical performance and oxidation resistance. In the last decade, low density Fe–Mn–Al–C steels have attracted considerable attention as these steel grades can potentially be used for lightweight crash-resistant car body structures and structural components in the cryogenic industry. Owing to the occurrence of several disordered and ordered fcc and bcc phases, Fe–Mn–Al–C steels exhibit outstanding combination of mechanical and physical properties that can be tuned by selective microstructural control. In particular, the formation of ordered precipitates such as L’1₂ (Fe, Mn)₃AlC carbides (κ-carbides) and B2-type Ni/Cu-rich precipitates have provided new alloy design strategies for the development of advanced low density steels. The exceptional combination of strength and ductility exhibited by these steels is ascribed to distinctive strengthening and strain-hardening mechanisms, such as ordering strengthening, transformation induced plasticity (TRIP), twinning induced plasticity (TWIP), shear induced plasticity (SIP), and dynamic slip band refinement.^{2,5,8,10,11,14,15)} In this review, the relevant phase structures of Fe–Mn–Al–C steels revealed by cutting-edge experimental and theoretical approaches as well as the underlying strengthening and strain-hardening mechanisms are summarized. Mechanical properties are summarized and analyzed from a microstructural standpoint. Future scientific and technical challenges are provided to establish low density Fe–Mn–Al–C steels as structural materials for industrial applications. Steel composition and alloying content are hereafter given in weight percentage.

Fig. 1.

Density reduction in low density Fe–Mn–Al–C steels as a function of Al content.¹⁾

2. Phase Structures

Low density Fe–Mn–Al–C steels are characterized by a single-phase (bcc-ferrite/fcc-austenite) or duplex phase structure (ferrite + austenite) that depending on the alloying content and thermomechanical treatment, contain a complex distribution of carbides. The most relevant carbides in low density steels are MC, M₃C, M₂₃C₆, and M₇C₃-type carbides, and ordered fcc (L’1₂) (Fe, Mn)₃AlC carbides, so-called κ-carbides.^16,17,18,19) In general, calculated isothermal sections of phase diagrams for the Fe–Mn–Al–C system are strongly dependent on the thermodynamic database, and hence, different descriptions have been reported. Phase diagrams show general trends such as that Al content enlarges the phase field of ferrite and suppresses that of austenite. These diagrams also predict that for a given Al content, the austenite region is enlarged with C and Mn contents. Figure 2 shows examples of the isothermal phase sections of an austenitic Fe-30Mn-Al-C steel at 1000°C (a) and 600°C (b) calculated from the PrecHiMn-4 database.¹⁹⁾ These diagrams show that austenite is stable at low Al content (< 6 wt.%). The stability of austenite with Al content can be increased with C content (for instance, the addition of 1 wt.% C enlarges the stability of austenite up to 9–10 wt.% Al.) These diagrams also show that the stability of κ-phase in austenitic Fe–Mn–Al–C steels requires high Al and C contents. At 600°C, κ-carbides are formed when Al > 5 wt.% and C > 1 wt.%. At low Al content (Al < 6 wt.%), cementite and M₂₃C₆-type carbides can be formed instead of κ-carbides. Interestingly, this diagram also reveals that austenite, ferrite, cementite, and κ-carbide are stable at specific Al and C content ranges. For instance, at large C content (> 1 wt.%) combined formation of κ-carbides and cementite can occur. On the other hand, at high Al content (> 9 wt.%), ferrite, austenite, and κ-phase are stable within a limited C range. These features reveal that in order to promote the precipitation of κ-carbides in austenite, C content and annealing temperature must be carefully controlled. It can be seen that at 1000°C, κ-carbides are only stable at large C and Al contents, such as C > 1.5 wt.% - Al > 12 wt.%, and C > 3 wt.% - Al > 6 wt.%, which are not relevant compositional ranges for practical purposes.

Fig. 2.

Isothermal phase sections of the austenitic Fe-30Mn-Al-C steel at 1000°C (a) and 600°C (b) calculated from the PrecHiMn-4 database.¹⁹⁾ (Online version in color.)

3. Microstructure-based Classification of Low Density Fe–Mn–Al–C Steels

From the standpoint of matrix phase structure, low density Fe–Al–Mn–C steels can be classified into three categories, namely ferritic steels, austenitic steels, and duplex steels. Table 1 summarizes the most relevant microstructural and compositional characteristics. The microstructural features of Fe–Mn–Al–C steels are the following.

Table 1. Microstructural and compositional characteristics of low-density Fe–Mn–Al–C steels.

Fe–Al–Mn–C class	Ferritic	Austenitic	Ferrite-based duplex	Austenite-based duplex
Relevant phases	δ, B2, DO3, κ	γ, α, δ, κ, β-Mn	α, δ, B2, DO3, γ, α’, κ	α, B2, DO3, γ, κ, β-Mn
Typical composition range	Mn < 5% Al: 5–9% C < 0.1%	Mn: 15–30% Al: 2–12% C: 0.5–2.0%	Mn: 3–10% Al: 3–9% C < 0.4%	Mn: 10–25% Al: 5–12% C: 0.6–1.0%

3.1. Ferritic Fe–Mn–Al–C Steels^{20,21,22,23,24)}

Ferritic Fe–Mn–Al–C steels contain low Mn content (< 5 wt.%), Al content of 5–9 wt.%, and low C content (< 0.1 wt.%). These steels may also contain additions of elements such as Si, Nb, Ti, V, and Ta in order to promote the precipitation of MC-type carbides and promote grain refinement. At hot working conditions, these steels have elongated grain structure of δ-ferrite along the rolling direction resulting in a band-like grain structure. δ-ferrite is directly produced from the liquid state during the casting process. Depending on the alloying content, the austenitic phase undergoes a martensitic transformation γ → α’ upon thermomechanical treatment resulting in the formation of macroscopic martensitic bands in the ferritic matrix.^21,23) Al content strongly modifies the crystal structure of the ferritic matrix of these steels. When Al > 7 wt.%, the δ-phase undergoes ordering transformation to form complex cubic structures such as the B2 structured phase based on a FeAl stoichiometry and the DO₃ structured phase based on a Fe₃Al stoichiometry. Depending on the Al content, δ-matrix phase can be A2-disordered FeAl, B2-ordered FeAl, or DO₃-ordered Fe₃Al. Cold rolling and microalloying of elements such as Nb, Ti, V, and B are used for grain refinement. In these steels, κ-carbides are semicoherent resulting in elongated rod-like morphologies due to the high lattice misfit between δ and κ phases (lattice mismatch ~6%), Fig. 3.²⁵⁾ The orientation relationship between δ and κ-carbide corresponds to the Nishiyama–Wasserman relationship, i.e. (011)_δ//(111)_κ. Recent analysis of the chemical structure of κ-carbides by atom probe tomography (APT) in a ferritic Fe-3.2Mn10Al-1.2C steel has revealed the nonstoichiometric structure of these carbides.²⁵⁾ This report reveals that the nonstoichiometry structure of κ-carbide is due to the partial replacement of Al by Fe or Mn in the corner positions in its face-centered cubic-based unit cell. Depending on the C content and cooling rate, coarse κ-carbides can be formed along grain boundaries.

Fig. 3.

Bright-field TEM image and corresponding diffraction pattern of the lamellar α-phase/κ-carbide structure in a ferritic Fe-3.0 Mn-5.5Al-0.3C steel.²⁵⁾

3.2. Austenitic Fe–Mn–Al–C Steels^{9,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41)}

Austenitic Fe–Mn–Al–C steels are characterized by high Mn content (15–30 wt.%), Al content of 2–12 wt.%, and high C content (0.5–2.0 wt.%). This type of low density steels exhibits fully equiaxed-austenitic microstructure at hot working conditions. Carbide precipitation behavior and the characteristics of the γ → α transformation are determined by the cooling rate and Al and C contents. In particular, high Al content (> 9 wt.%) promotes the formation of disordered ferrite with ordered B2 and DO₃ domains along grain boundaries.⁹⁾ From a mechanical standpoint, κ-carbides (L1’₂ phase) are the most relevant precipitates in these steels although other carbides, namely M₂₃C₆, M₇C₃, and M₃C, can also influence the mechanical performance. The formation of κ-carbides in austenitic Fe–Mn–Al–C steels requires high Al and C contents, namely Al > 5 wt.% and C > 1 wt.% (Fig. 2). κ-carbides are coherent or semicoherent precipitates (lattice misfit < 2%) with a cube-on-cube orientation relationship with austenite, i.e. [100]κ//[100]γ and [010]κ//[010]γ. Recent analysis by atom probe tomography (APT) combined with calculations based on the density functional theory (DFT) have revealed the nonstoichiometric structure of κ-carbides in an austenitic Fe-29.8Mn-7.7Al-1.3C steel.^35,37) Specifically, APT has evidenced sub-lattice depletion of interstitial C and substitutional Al sites. This effect has been attributed to the low formation energies for $M{n}_{Al}^{\gamma }$ anti-sites (assuming thermodynamic equilibrium of $M{n}_{Al}^{\gamma }$ anti-sites with the surrounding matrix) that are energetically favored under the compressive elastic strain caused by the coherency of κ-carbides. κ-carbide distribution consists of a complex three-dimensional arrangement of cuboidal and plate-like precipitates with diameter of about 10–50 nm along the three orthogonal <001> directions. 2-D analysis of κ-carbide distribution by TEM typically shows a sideband distribution of nanosized precipitates along <100> directions, Fig. 4(a).¹⁵⁾ 3-D analysis of κ-carbide distribution by atom probe tomography (APT) has recently revealed that κ-carbides are compactly distributed into stacks with different inter-particle lengths, Figs. 4(b), 4(c).¹⁵⁾ Specifically, the inter-particle spacing of κ-carbides distributed along the same stack is about 2–5 nm (narrow γ channel in Fig. 4(c)). The inter-particle spacing of κ-carbides distributed along different stacks is about 10–40 nm (broad γ channel in Fig. 4(c)).

Fig. 4.

κ-carbide distribution in a Fe-30.4Mn-8Al-1.2C austenitic steel annealed at 600°C for 24 h.¹⁵⁾ (a): Dark-field TEM image; (b): Schematic illustration of the 3D morphology and arrangement of κ-carbides based on APT data; (c): 2D projections of the κ-carbide distribution along <001> directions highlighted in (b).

In austenitic Fe–Mn–Al–C steels, κ-carbides can be massively precipitated (volume fraction up to 40%) by spinodal decomposition, i.e. chemical modulation, of the metastable γ-phase followed by ordering. The latter reaction involves short-range ordering (SRO) of γ into the metastable (Fe, Mn)₃AlC_x (x < 1) L1₂ phase (κ’-carbide) and the subsequent ordering of κ’-carbide to κ-carbide.^26,27,29,31) κ’-phase has the same crystal structure as the κ phase but with uncompleted occupation of C atoms. The kinetics of the spinodal decomposition and the density of κ-carbides increase with Al and C contents, in particular when Al > 10 wt.% and C > 1 wt.%. Prolonged aging in the temperature range 450–650°C results in the increase of the lattice parameter of κ-carbide and the decrease of that of austenite. This effect promotes the loss of coherency of κ-carbides. At annealing temperatures of 650–800°C, coarse κ-carbides tend to precipitate along grain boundaries resulting in lamellar γ/κ structure, Fig. 5(a). Depending on steel composition and annealing conditions, these carbides can be formed through different types of reactions such as cellular transformations and eutectoid reactions.^28,30,41) Grain boundary κ-carbides have parallel orientation relationship with one of the neighboring grains. At the early precipitation stages, κ-carbides are formed as discrete particles along grain boundaries. With further annealing in the (γ + κ) or the (γ + α + κ) regions, these carbides transform into thin films distributed continuously along grain boundaries that grow into adjacent grains, as shown in the example of Fig. 5(b). With high Al content (> 6 wt.%), β-Mn precipitates and α-bcc particles can also be formed along grain boundaries through the phase transformation κ → α + β-Mn,^34,36,42) Figs. 5(c), 5(d). γ/β-Mn orientation relationship is (111)_γ//(221)_β; (01-1)_γ//(01-2)_β; (-211)_γ//(-542)_β.³⁹⁾ High temperature (800°C) aging of austenitic Fe–Mn–Al–C steels with high Al content results in the γ → α transformation.³²⁾ The resulting ferrite phase contains ordered B2 and DO₃ structures, which are detrimental to mechanical performance.

Fig. 5.

(a): Bright-field TEM image of κ-carbides along the grain-boundary between two adjacent austenitic grains γ₁ and γ₂ in a Fe-30.5Mn-8Al-1.2C austenitic steel annealed at 600°C for 96 h;³³⁾ (b): Bright-field TEM image of the α/γ/κ lamellar structure in a Fe-18Mn-7.1Al-0.85C steel annealed at 650°C for 100 h;³⁰⁾ (c): Bright-field TEM image of β-Mn precipitate in a Fe-31.4 Mn-11.4Al-0.9 C austenite steel annealed at 550°C for 500 h;³⁹⁾ (d) Diffraction pattern taken at the γ-matrix/β-Mn interphase boundary (red circle).³⁹⁾ (Online version in color.)

3.3. Duplex Fe–Mn–Al–C Steels^{3,8,9,10,11,12,32,43,44,45,46,47,48,49,50)}

Duplex Fe–Mn–Al–C steels are characterized by a complex multiphase microstructure consisting of δ-ferrite, α-ferrite, austenite, and κ-carbides. α’-martensite and bainite can be formed in ferrite-based duplex steels with high-Mn content. Due to the high alloying contents of Mn and Al, the solidification microstructure contains macroscopic segregations of these elements resulting in a macroscopic banded structure during the hot rolling process. The structure, size, volume fraction, and distribution of the structural phases can be controlled by adjusting the thermomechanical conditions. From the standpoint of matrix phase structure, low density duplex steels can be classified into ferrite- and austenite-based duplex steels. The microstructural features of these steels are the following.

Ferrite-based duplex steels contain medium Mn content (3–10 wt.%), Al content of 5–9 wt.% and low C content (< 0.4 wt.%).^10,44,48) These steels are characterized by a complex bimodal banded structure of δ-ferrite and austenite bands elongated along the rolling direction, Fig. 6(a). The phase structure of the austenite band is strongly dependent on the Mn content and annealing treatment. Austenite partially or completely transforms into lamellar colonies of (α + κ) that are formed along the parent γ-grain boundaries or at δ/γ interfaces. The lamellar (α + κ) structure is characterized by elongated rod-like semicoherent κ-carbides. At high Mn content, the decomposition γ → α + κ is suppressed and the martensitic transformation γ → α’ occurs.^46,47) Austenite-based duplex steels contain high Mn content (10–25 wt.%), Al content of 5–12 wt.% and C content of 0.6–1.2 wt.%.^{3,8,9,11,12,32,43,45,47,49,50)} Fe–Mn–Al–C steels with high Mn and Al contents (Mn > 25 wt.% and Al > 9 wt.%) can exhibit fully austenitic structure in the as-quenched state. Upon further aging at 500–800°C, complex ferritic band structure can be formed along austenitic grain boundaries. On the other hand, ferrite-based duplex steels are characterized by a complex bimodal banded structure of δ-ferrite and austenite bands elongated along the rolling direction, Fig. 6(b). As this figure shows, austenite can undergo strain induced α’-martensitic transformation. Together with the formation of L’1₂ κ-carbides, several precipitation strategies have been recently suggested to promote the formation of ordered B2-type precipitates. Additions of Ni and Cu of around 3–5 wt.% enable the formation of non-shearable B2-type precipitates in austenitic bands resulting in enhanced mechanical performance.^8,12,50) As an example, Fig. 6(c) shows the piling-up of dislocations at the interface of a B2 Cu-rich precipitate in a Fe-12Mn-7Al-0.5C-3Cu steel. Depending on the alloying content and annealing conditions, ferritic bands can contain B2 precipitates, κ-carbides, and austenite grains. κ-carbides are formed through the eutectoid decomposition γ → α + κ.

Fig. 6.

(a): Optical micrograph of the banded structure of a ferrite-based duplex Fe-3.5 Mn-5.8Al-0.35C steel;⁴⁴⁾ (b): EBSD phase map of a hot-rolled austenite-based duplex Fe-8.5Mn-5.6Al-0.3C steel;⁴⁷⁾ (c): Bright-field TEM image of a Cu-rich B2 precipitate in a Fe-12Mn-7Al-0.5C-3Cu annealed at 830°C for 1 minute and strained to 0.02.¹²⁾ (Online version in color.)

4. Thermomechanical Treatment

Cast ingots of Fe–Mn–Al–C steels are typically homogenized at a temperature of 1100°C–1250°C for 1–3 h and then hot-rolled to 2–5 mm thickness at a temperature of 850–1000°C. After hot rolling, the hot-rolled strips are cooled to a temperature range of 500–650°C for 1–5 h, and water- or air-cooled to room temperature. In ferritic steels, the hot-rolled strips are not used as the final product due to the large and elongated grain size of the δ-matrix. To control the grain size, texture, and precipitation behavior, hot-rolled strips undergo a cold rolling process followed by annealing at a temperature of 700–900°C. In the case of ferrite-based duplex steels, austempering process can be finally applied. In austenitic steels, the hot-rolled strips can be directly fast cooled to a temperature range of 500–750°C and subsequently slow cooled. Alternatively, the hot-rolled strips can be fast cooled to room temperature, followed by isothermal annealing. The cooling rate after hot rolling is critical to avoid the formation of deleterious intergranular κ-carbides. Cold-rolled products of austenitic steels are typically obtained by solution treatment of the cold-rolled strips in the temperature range of 900–1100°C and followed by fast quenching. Precipitation hardening is typically performed by annealing treatments in the temperature range of 500–700°C for 5–20 h.

5. Properties

In this section, the most relevant engineering properties of low density Fe–Mn–Al–C steels are reviewed. However, due to the novelty of these steels as potential structural materials, the available literature on several engineering properties is still limited. In this section, room-temperature and cryogenic tensile properties, impact toughness, energy absorption, and corrosion resistance of low density Fe–Mn–Al–C steels are analyzed.

5.1. Room-temperature Mechanical Properties

Fe–Mn–Al–C steels exhibit a wide range of mechanical properties at room temperature depending on the alloy composition and the processing route. Figure 7 plots the tensile elongation (TE) vs. ultimate tensile strength (UTS) diagram at room temperature for several types of Fe–Mn–Al–C steels.^{2,3,5,6,8,9,10,11,12,21,23,24,26,32,33,38, 43,44,45,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62)} The plot includes the tensile properties of conventional high strength steels for comparison. In particular, Fig. 7 shows the following characteristics:

Fig. 7.

Tensile elongation (TE) vs. ultimate tensile strength (UTS) diagram at room temperature for several types of low density Fe–Mn–Al–C steels. The plot includes the tensile properties of some conventional high strength steels, namely, IF (interstitial-free), Mild, BH (bake hardenable), CMn, HSLA (high-strength low-alloy), TRIP (transformation induced plasticity), DP (dual-phase) and MART (martensitic). (Online version in color.)

- Tensile properties of ferritic Fe–Mn–Al–C steels are similar to those of bake hardenable (BH), low-alloyed C–Mn, and HSLA steels. These steels exhibit UTS of 450–500 MPa and TE of 15–30%. When Al > 7 wt.%, the formation of the short-range ordering (K₁ state) induces brittle fracture in the sheet forming even during the process step of cold rolling.

- Ferrite-based duplex steels are located on the upper bound of the first generation advanced high strength steels, i.e. TRIP, dual-phase, and CP steels. They typically exhibit UTS of 600–800 MPa and TE of 15–40%.

- Austenite-based duplex steels exhibit superior tensile properties than the ferrite-based duplex steels. These steels typically exhibit UTS of 600–1400 MPa and TE of 20–60%. In particular, austenite-based duplex steels with Ni and Cu additions exhibit attractive combinations of mechanical strength (yield stress > 1.0 GPa; UTS > 1.5 GPa; moderated ductility: 20–30%.) Steels with high Al content (Al > 10 wt.%), contain high volume fraction of nanosized κ-carbides, coarse intergranular κ-carbides, and α-particles. This precipitation behavior results in steels with high UTS (~1400 MPa), limited TE (<10%), and high mechanical anisotropy.

- Austenitic steels exhibit attractive combination of tensile properties, i.e. UTS of 600–1100 MPa and TE of 40–100%.

Tensile properties of low density Fe–Mn–Al–C steels are associated with the underlying strengthening and deformation mechanisms. The main characteristics of these mechanisms are the following:

- Tensile properties of ferritic and ferrite-based duplex steels are mainly controlled by the strain-hardening capacity of ferrite,^3,9,43) and the formation of coarse intergranular κ-carbides. Evolving dislocation substructures in these steels are typical of ferritic steels.^9,11,59) In ferrite-based duplex steels, austenite grains may undergo strain induced α’-martensitic transformation and at high stress levels, deformation twins can be formed. These effects result in enhanced strain-hardening capacity due to the activation of the TRIP (transformation induced plasticity) and TWIP (twinning induced plasticity) effects. The efficiency of the TRIP effect is dependent on the crystalline orientation of the metastable austenite and its mechanical stability against the formation of strain-induced martensite.

- Tensile properties of austenite-based duplex steels are controlled by the strain-hardening capacity of austenite, the formation of shearable κ-carbides and non-shearable B2 precipitates, and the size of ferritic grains.

- Tensile properties of austenitic steels are strongly tunable by proper microstructural control. Yield stress values of up to 1200 MPa can be achieved by tuning the size distribution of these carbides. The strengthening behavior of these steels is ascribed to several strengthening mechanisms such as solid solution hardening, ordering strengthening due to particle shearing, and precipitation hardening, i.e. Orowan looping. Yao et al.¹⁵⁾ have recently analyzed the strengthening mechanisms in austenitic Fe–Mn–Al–C steels. Their analysis is summarized in Fig. 8(a). This plot shows the normalized strengthening (normalized with respect to the volume fraction, V_f) $\raisebox{1ex}{$N\mathrm{\Delta }{\sigma }_{prec}$}\!\left/ \!\raisebox{-1ex}{$\sqrt{V_f}$}\right.$ as a function of the mean particle radius, r, for two values of anti-phase boundary (APB) energy, γ_APB. These energies were calculated by ab-initio calculations for two different phase structures of κ-carbides, namely a C-containing structure with γ_APB ~700 mJ/m² and a C-free structure with γ_APB ~350 mJ/m². The diagram shows that as the particle size, r, increases, the strengthening mechanism shifts from weakly coupled particle shearing to strongly coupled particle shearing. Furthermore, the diagram reveals that for high r-values (r > 40 nm), the active strengthening mechanism is Orowan looping. The plot also shows that for high γ_APB-values, the strengthening peak stress increases, and the particle radii delimiting the different strengthening regimes decrease. It is worth pointing out that for conventional sizes of κ-carbides (10–40 nm), the diagram shows that particle-shearing is the dominant strengthening mechanism, which agrees with experimental observations.^{2,4,6,9,15,38,45)} Figure 8(b) shows an example of κ-carbide shearing in a Fe-30.4Mn-8Al-1.2 C steel tensile strained to 0.02 true strain.¹⁵⁾ Orowan looping is however favored in regions with large inter-particle spacing, i.e. broad γ-channels in Fig. 4(c).

Fig. 8.

(a): Normalized κ-carbide strengthening plotted as a function of the mean particle radius, r, for γ_APB of 350 (dashed line) and 700 (solid line) mJ/m²;¹⁵⁾ (b): Dark-field TEM image of κ-carbide shearing in a Fe-30.4Mn-8Al-1.2 C steel tensile strained to 0.02 true strain.¹⁵⁾ (Online version in color.)

The outstanding strain-hardening behavior of austenitic Fe–Mn–Al–C steels is ascribed to the formation and evolution of characteristic planar dislocation substructures^{3,5,6,14,38,45)} and planar slip bands^2,14,15,38) upon straining due to the dynamic slip band refinement mechanism.^14,15) In low-Al κ-carbide-free austenitic steels, deformation twinning can be also activated at high stress levels enabling further strain-hardening capacity.⁵⁾ The mechanisms controlling the formation of these dislocation structures are mainly associated with the shearing of short-range ordering (SRO) clusters and κ-carbides. Haase et al.³⁸⁾ have recently suggested that these shearing mechanisms have similar influence on the glide plane softening effect and hence, on the evolving dislocation configuration. In κ-carbide-free austenitic steels, planar dislocation configurations have been mainly ascribed to the effect of carbon content on the frequency of dislocation cross-slip⁵⁾ and the effect of short-range ordering (SRO) on dislocation glide.^5,14,38) In κ-carbide-containing austenitic steels, the formation of planar dislocation structures is ascribed to the shearing of κ-carbides, Fig. 8(b). The resulting strain-hardening rate produced by the evolving planar dislocation structure is high enough to compensate the negative strain hardening rate produced by shearing and fragmenting κ-carbides by dislocations and the additional effect of dragging out carbon and other elements by moving dislocations which leads to a reduction of γ_APB of κ-carbides.¹⁵⁾

5.2. Cryogenic Mechanical Properties

Fe–Mn–Al–C austenitic steels exhibit cryogenic mechanical performance comparable to that of conventional stainless steels. In general, the cryogenic mechanical properties of Fe–Mn–Al–C austenitic steels are mainly determined by the stability of austenite against the martensitic transformation γ → α’,⁶³⁾ carbide behavior, and dislocation and twin structures.⁶⁴⁾ Further exploration of recently calculated Fe–Mn–Al–C phase diagrams together with novel precipitation approaches may open the space design map of low density Fe–Mn–Al–C steels for cryogenic applications. For instance, Li et al.⁶⁴⁾ have recently suggested a thermomechanical approach where a heterogeneous distribution of nanosized κ-carbides and Nb-rich carbides is formed that yields superior cryogenic mechanical performance. Figure 9 plots the yield stress (YS), ultimate tensile strength (UTS), tensile elongation (TE), and Charpy impact energy for several types of Fe–Mn–Al–C steels tested at 77K.^{51,52,63,64,65)} Stress values correspond to true stress values. For comparison, the plots include the mechanical properties of conventional stainless steels. Figures 9(a), 9(b) show that Fe–Mn–Al–C steels exhibit YS ~1 GPa, UTS ~1.8–2.0 GPa, and true elongation of 40–50% at 77K, which are comparable to those exhibited by 316LN (15Cr–12.5Ni–1.5Mn–2.5Mo–0.2Si–0.15N) stainless steel. On the other hand, Fe–Mn–Al–C steels exhibit wide range of Charpy V-notch (CVN) impact energy values at 77K (6–140 J), which are lower than that of 316L stainless steel (~200 J), Fig. 9(c). In general, κ-carbide-free austenitic steels exhibit superior CVN values than those of κ-carbide-containing austenitic steels.

Fig. 9.

Cryogenic mechanical properties of low density Fe–Mn–Al–C steels. (a): Yield stress (YS) and ultimate tensile strength (UTS); (b): Tensile elongation (TE); (c): Charpy V-notch (CVN) impact energy.

5.3. Impact Toughness

Low density Fe–Mn–Al–C steels exhibit high energy absorption behavior compared to that of conventional deep drawing steels, such as DC04, HC300LA, and high strength IF steels, Fig. 10(a).¹⁾ When Al > 7 wt.%, E_spec value of Fe–Mn–Al–C steels is about 0.5 J/mm³, which is around fifty percent higher than that of the conventional deep drawing steels (0.16–0.25 J/mm³.) E_spec is the deformation energy per unit volume at a given temperature at a high strain rate. The enhanced energy absorption of Fe–Mn–Al–C steels is ascribed to severe shear band formation at high strain rates. Impact toughness of Fe–Mn–Al–C steels is strongly microstructural dependent and, in particular, it depends on the formation of κ-carbides and the presence of δ-ferrite.¹³⁾ Al content plays significant role to tune the impact toughness of these steels. When Al < 7 wt.%, grain boundary precipitation of Fe₃C and M₇C₃ carbides occurs, which has detrimental effect on ductility and toughness. When Al > 7 wt.%, mechanical strength increases but ductility and impact toughness decrease due to the increasing volume fraction of κ-carbides and the formation of δ-ferrite. The latter results in strong decrease of impact toughness due to the deformation incompatibility between γ and δ phases. Figure 10(b) shows an example of the evolution of the Charpy V-notch (CVN) impact energy with testing temperature in several austenitic Fe–Mn–Al–C steels at different annealing conditions.^13,66) The plot also shows the dependence of CVN impact energy with temperature for different steel grades. In the as-solution treated state (curve 3), Fe–Mn–Al–C steels exhibit high impact energy at room temperature, which is similar to that of austenitic stainless steels. In the aged states (curves 4 and 5), the CVN impact energy decreases with testing temperature moderately to values similar to those in duplex steels.

Fig. 10.

(a): Specific energy absorption, E_spec, in low density Fe–Mn–Al–C steels as a function of Al content.¹⁾ The plot also shows the value of E_spec for different steels grades; (b): Charpy V-notch (CVN) impact energy of several Fe–Mn–Al–C austenitic steels as a function of test temperature.^13,66) The plot also shows the dependence of CVN impact energy with temperature for different steel grades.

5.4. Corrosion Resistance

The corrosion resistance of low density Fe–Mn–Al–C steels in aqueous environments is comparable to that of the conventional high strength steels, although inferior to that of 304 stainless steel.⁶⁷⁾ The addition of Al, which is generally expected to provide active protective function, does not enhance the stress corrosion cracking behavior of Fe–Mn–Al–C alloys.⁶⁸⁾ On the other hand, the addition of Cr together with the reduction of C content enhances the corrosion resistance of Fe–Mn–Al–C alloys although limits their mechanical performance, mainly due to the duplex structure.^67,69)

6. Future Challenges

In the last decade, fundamental research on Fe–Mn–Al–C steels has been extremely active. Research fields such as alloy design,^{1,8,10,12,32)} thermodynamic database,^16,17,18,19) structural analysis,^{25,31,34,35,37,39,40)} precipitation behavior,^{26,27,28,29,41,42)} phase transformations,^30,36) and deformation and strengthening mechanisms^{2,3,4,5,6,11,14,15,33,38,46,50,55,60,61,64)} have been extensively investigated. However, there are still open questions that are still pending to establish Fe–Mn–Al–C steels as structural materials in industrial applications. In the following several scientific and technical challenges of Fe–Mn–Al–C steels are addressed.

6.1. Technical Challenges

- Steelmaking. Large additions of Mn and Al are difficult to handle during the steelmaking process of Fe–Mn–Al–C steels. Intensive chemical reactions can occur between the melt and the refractory materials which can lead to impurities and chemical alloy deviations. For instance, the formation of alumina in the liquid state can lead to clogging of the nozzles during continuous casting. Heavy and dense Al-oxides can form on the surfaces during the strip casting stage, as a result of the chemical reactions between the as-cast strip and the atmosphere. As a consequence, several phenomena such as the formation of surface defects, brittle phases and cracks, and decarburization can occur during the casting of Fe–Mn–Al–C steels. In order to overcome these difficulties, novel casting routes have to be implemented.

- Processing. Cold-rolled and recrystallized microstructure, texture evolution upon rolling and subsequent recrystallization, and grain size control of Fe–Mn–Al–C steels have not been evaluated in detail. These aspects are key to optimize the processing settings under industrial conditions.

- Mechanical properties. Analysis of mechanical properties such as fracture behavior (in particular, at cryogenic temperatures), mechanical anisotropy (in particular, duplex-based steels), and fatigue properties have not been yet performed in detail. The evaluation of these mechanical properties is critical for the establishment of low density Fe–Mn–Al–C steels as structural materials.

- Hydrogen embrittlement. Hydrogen embrittlement is one of the main technological issues to be addressed on high-strength steels. The analysis of the effect of hydrogen on the mechanical behavior of low density Fe–Mn–Al–C steels is critical for the application of these steels on an industrial scale. Al addition tends to mitigate hydrogen embrittlement in austenitic Fe–Mn–C steels.^{70,71,72,73,74,75)} This effect has been ascribed to the role of Al on the enhanced hydrogen solubility and reduced permeability and diffusivity of hydrogen in Fe–Mn–Al–C steels. Low density Fe–Mn–Al–C steels contain several types of microstructural interfaces and phase structures that may exhibit different hydrogen trapping efficiency resulting in complex hydrogen embrittlement behavior. For instance, calculations based upon the density functional theory (DFT) have predicted that the hydrogen trapping efficiency of κ-carbides is strongly dependent on C and Mn contents.⁷⁶⁾ These calculations show the potential of κ-carbides as hydrogen traps by proper precipitate control. Further investigations are required to elucidate the efficiency of microstructural interfaces such as twin interfaces, dislocations configurations, B2 precipitates, and κ-carbides on hydrogen trapping.

6.2. Scientific Challenges

- Short-range ordering (SRO). The occurrence of SRO has been frequently put forward to explain the planar character of the dislocation configuration formed in the early stages of deformation of κ-carbide-free austenitic Fe–Mn–Al–C steels. Features such as SRO cluster size and thermal and structural stability of these clusters are currently unknown. From an experimental standpoint, the analysis of SRO, as well as the role of SRO clusters on plasticity, is a challenging task that deserves to be investigated to further understand the role of SRO on plasticity and hence provide novel alloy design strategies.

- Phase structure, precipitation behavior, and strengthening mechanisms of Cu/Ni-rich B2 precipitates in duplex-based Fe–Mn–Al–C steels. These issues are critical for microstructural control of these precipitates and the establishment of novel alloy design strategies.

- Thermal and mechanical behavior of microstructural interfaces. Low density Fe–Mn–Al–C steels and, in particular duplex-based steels, contain different types of homophase (twin interfaces and dislocation configurations) and heterophase interfaces (ferrite/austenite, austenite/B2, austenite/L1’₂, ferrite/B2, ferrite/L1’₂.) Thermal and mechanical behavior of microstructural interfaces have not been explored yet in detail. Thermal behavior of microstructural interfaces is critical to understand the mobility of these interfaces and their role on the recrystallization behavior and grain growth. On the other hand, the mechanical behavior of microstructural interfaces (for instance, twin/dislocation interaction, precipitate/dislocation interaction, and crack propagation behavior along microstructural interfaces) is key for understanding the nucleation and propagation of plasticity, damage nucleation, and fracture behavior of these steels.

- Alloy control of strain-hardening mechanisms. Several strain-hardening mechanisms have been observed in low density Fe–Mn–Al–C steels which are associated with the complex interaction of plasticity with the different structural phases and the mechanical stability of these phases. In order to establish robust alloy design strategies, the alloying dependence of the strain-hardening mechanisms should be explored. This is a challenging task due to the presence of different types of microstructural interfaces, in particular heterophase interfaces, that can result in the occurrence of local chemical gradients.

7. Conclusions

Low density Fe–Mn–Al–C steels are potential structural materials for automotive, aircraft, and cryogenic industries. The combination of outstanding physical and mechanical properties while offering a weight reduction of up to 20% make these steels attractive structural materials as lightweight crash-resistant car body structures and structural components in the cryogenic industry. Owing to the occurrence of several disordered and ordered fcc and bcc phases, Fe–Mn–Al–C steels exhibit outstanding combination of mechanical properties at room and cryogenic temperatures that can be tuned by selective microstructural control. In the last decade, high research activity has been devoted to the analysis of the underlying deformation mechanisms and determination of crystal structures. These works have provided new understanding of the crystal structure of relevant phases such as L’1₂ (Fe, Mn)₃AlC carbides (κ-carbides) and B2-type Ni/Cu-rich precipitates, and their deformation behavior. These activities have led to new alloy design strategies for the development of advanced low density steels. Novel strain-hardening mechanisms ascribed to the formation of extended planar dislocation structures by shearing of short-range ordered clusters and κ-carbides have been put forward to explain the outstanding strain-hardening of these steels. There are still several scientific and technological challenges to establish low density Fe–Mn–Al–C steels as structural materials in industrial applications. In particular, further research on alloy design, steelmaking, processing, mechanical performance, and hydrogen embrittlement is required. These challenges make low density Fe–Mn–Al–C steels attractive steels for further scientific and technological research.

Acknowledgment

The author would like to thank the financial support of the NIMS funding Program “Principle elucidation of the relationship between nano-scale structures and dynamics at interfaces”.

References

• 1)   D.  Raabe,  H.  Springer,  I.  Gutierrez-Urrutia,  F.  Roters,  M.  Bausch,  J.-B.  Seol,  M.  Koyama,  P.-P.  Choi and  K.  Tsuzaki: JOM, 66 (2014), 1845.
• 2)   K.  Choi,  C.-H.  Seo,  H.  Lee,  S. K.  Kim,  J. H.  Kwak,  K. G.  Chin,  K.-T.  Park and  N. J.  Kim: Scr. Mater., 63 (2010), 1028.
• 3)   H.  Ding,  D.  Han,  Z.  Cai and  Z.  Wu: JOM, 66 (2014), 1821.
• 4)   G.  Frommeyer and  U.  Brüx: Steel Res. Int., 77 (2006), 627.
• 5)   I.  Gutierrez-Urrutia and  D.  Raabe: Acta Mater., 60 (2012), 5791.
• 6)   I.  Gutierrez-Urrutia and  D.  Raabe: Scr. Mater., 68 (2013), 343.
• 7)   H.  Kim,  D.-W.  Suh and  N. J.  Kim: Sci. Technol. Adv. Mater., 14 (2013), 014205.
• 8)   S.-H.  Kim,  H.  Kim and  N. J.  Kim: Nature, 518 (2015), 77.
• 9)   K.-T.  Park,  S. W.  Hwang,  C. Y.  Son and  J.-K.  Lee: JOM, 66 (2014), 1828.
• 10)   S. S.  Sohn,  K.  Choi,  J.-H.  Kwak,  N. J.  Kim and  S.  Lee: Acta Mater., 78 (2014), 181.
• 11)   S. S.  Sohn,  H.  Song,  B.-C.  Suh,  J.-H.  Kwak,  B.-J.  Lee,  N. J.  Kim and  S.  Lee: Acta Mater., 96 (2015), 301.
• 12)   H.  Song,  J.  Yoo,  S.-H.  Kim,  S. S.  Sohn,  M.  Koo,  N. J.  Kim and  S.  Lee: Acta Mater., 135 (2017), 215.
• 13)   S.  Chen,  R.  Rana,  A.  Haldar and  R. K.  Ray: Prog. Mater. Sci., 89 (2017), 345.
• 14)   E.  Welsch,  D.  Ponge,  S. M. H.  Haghighat,  S.  Sandlöbes,  P.  Choi,  M.  Herbig,  S.  Zaefferer and  D.  Raabe: Acta Mater., 116 (2016), 188.
• 15)   M. J.  Yao,  E.  Welsch,  D.  Ponge,  S. M. H.  Haghighat,  S.  Sandlobes,  P.  Choi,  M.  Herbig,  I.  Bleskov,  T.  Hickel,  M.  Lipinska-Chwalek,  P.  Shanthraj,  C.  Scheu,  B.  Gault and  D.  Raabe: Acta Mater., 140 (2017), 258.
• 16)   K.  Ishida,  H.  Ohtani,  N.  Satoh,  R.  Kainuma and  T.  Nishizawa: ISIJ Int., 30 (1990), 680.
• 17)   K.-G.  Chin,  H.-J.  Lee,  J.-H.  Kwak,  J.-Y.  Kang and  B.-J.  Lee: J. Alloy. Compd., 505 (2010), 217.
• 18)   H.-J.  Lee,  S. S.  Sohn,  S.  Lee,  J.-H.  Kwak and  B.-J.  Lee: Scr. Mater., 68 (2013), 339.
• 19)   B.  Hallstedt,  A. V.  Khvan,  B. B.  Lindahl,  M.  Selleby and  S.  Liu: Calphad, 56 (2017), 49.
• 20)   L.  Falat,  A.  Schneider,  G.  Sauthoff and  G.  Frommeyer: Intermetallics, 13 (2005), 1256.
• 21)   S. Y.  Han,  S. Y.  Shin,  H.-J.  Lee,  B.-J.  Lee,  S.  Lee,  N. J.  Kim and  J.-H.  Kwak: Metall. Mater. Trans. A, 43 (2012), 843.
• 22)   Y.-U.  Heo,  Y.-Y.  Song,  S.-J.  Park,  H. K. D. H.  Bhadeshia and  D.-W.  Suh: Metall. Mater. Trans. A, 43 (2012), 1731.
• 23)   S. Y.  Han,  S. Y.  Shin,  B.-J.  Lee,  S.  Lee,  N. J.  Kim and  J.-H.  Kwak: Metall. Mater. Trans. A, 44 (2013), 235.
• 24)   A.  Zargaran,  H. S.  Kim,  J. H.  Kwak and  N. J.  Kim: Scr. Mater., 89 (2014), 37.
• 25)   J.-B.  Seol,  D.  Raabe,  P.  Choi,  H.-S.  Park,  J.-H.  Kwak and  C.-G.  Park: Scr. Mater., 68 (2013), 348.
• 26)   K.  Sato,  K.  Tagawa and  Y.  Inoue: Mater. Sci. Eng. A, 111 (1989), 45.
• 27)   K.  Sato,  Y.  Inoue and  K.  Tagawa: Metall. Trans. A, 21 (1990), 5.
• 28)   C. N.  Hwang,  C. Y.  Chao and  T. F.  Liu: Scr. Metall. Mater., 28 (1993), 263.
• 29)   W. K.  Choo,  J. H.  Kim and  J. C.  Yoon: Acta Mater., 45 (1997), 4877.
• 30)   W.-C.  Cheng: JOM, 66 (2014), 1809.
• 31)   L. N.  Bartlett,  D. C.  Van Aken,  J.  Medvedeva,  D.  Isheim,  N. I.  Medvedeva and  K.  Song: Metall. Mater. Trans. A, 45 (2014), 2421.
• 32)   A.  Etienne,  V.  Massardier-Jourdan,  S.  Cazottes,  X.  Garat,  M.  Soler,  I.  Zuazo and  X.  Kleber: Metall. Mater. Trans. A, 45 (2014), 324.
• 33)   I.  Gutierrez-Urrutia and  D.  Raabe: Mater. Sci. Technol., 30 (2014), 1099.
• 34)   K.  Lee,  S.-J.  Park,  J.  Moon,  J.-Y.  Kang,  T.-H.  Lee and  H. N.  Han: Scr. Mater., 124 (2016), 193.
• 35)   M. J.  Yao,  P.  Dey,  J.-B.  Seol,  P.  Choi,  M.  Herbig,  R. K. W.  Marceau,  T.  Hickel,  J.  Neugebauer and  D.  Raabe: Acta Mater., 106 (2016), 229.
• 36)   X. P.  Chen,  Y. P.  Xu,  P.  Ren,  W. J.  Li,  W. Q.  Cao and  Q.  Liu: Mater. Sci. Eng. A, 703 (2017), 167.
• 37)   P.  Dey,  R.  Nazarov,  B.  Dutta,  M.  Yao,  M.  Herbig,  M.  Friak,  T.  Hickel,  D.  Raabe and  J.  Neugebauer: Phys. Rev. B, 95 (2017), 104108.
• 38)   C.  Haase,  C.  Zehnder,  T.  Ingendahl,  A.  Bikar,  F.  Tang,  B.  Hallstedt,  W.  Hu,  W.  Bleck and  D. A.  Molodov: Acta Mater., 122 (2017), 332.
• 39)   K.  Lee,  S.-J.  Park,  J.-Y.  Kang,  S.  Park,  S. S.  Han,  J. Y.  Park,  K. H.  Oh,  S.  Lee,  A. D.  Rollett and  H. N.  Han: J. Alloy. Compd., 723 (2017), 146.
• 40)   C. H.  Liebscher,  M.  Yao,  P.  Dey,  M.  Lipińska-Chwalek,  B.  Berkels,  B.  Gault,  T.  Hickel,  M.  Herbig,  J.  Mayer,  J.  Neugebauer,  D.  Raabe,  G.  Dehm and  C.  Scheu: Phys. Rev. Mater., 2 (2018), 023804.
• 41)   Y. H.  Tuan,  C. L.  Lin,  C. G.  Chao and  T. F.  Liu: Mater. Trans., 49 (2008), 1589.
• 42)   C. Y.  Chao,  C. N.  Hwang and  T. F.  Liu: Scr. Metall. Mater., 28 (1993), 109.
• 43)   M. C.  Ha,  J.-M.  Koo,  J.-K.  Lee,  S. W.  Hwang and  K.-T.  Park: Mater. Sci. Eng. A, 586 (2013), 276.
• 44)   S. S.  Sohn,  B.-J.  Lee,  S.  Lee,  N. J.  Kim and  J.-H.  Kwak: Acta Mater., 61 (2013), 5050.
• 45)   Z. Q.  Wu,  H.  Ding,  H. Y.  Li,  M. L.  Huang and  F. R.  Cao: Mater. Sci. Eng. A, 584 (2013), 150.
• 46)   C.-Y.  Lee,  J.  Jeong,  J.  Han,  S.-J.  Lee,  S.  Lee and  Y.-K.  Lee: Acta Mater., 84 (2015), 1.
• 47)   S. S.  Sohn,  H.  Song,  J.-H.  Kwak and  S.  Lee: Sci. Rep., 7 (2017), 1927.
• 48)   X.  Li,  R.  Song,  N.  Zhou and  J.  Li: Mater. Sci. Eng. A, 709 (2018), 97.
• 49)   D.  Han,  H.  Ding,  D.  Liu,  B.  Rolfe and  H.  Beladi: Mater. Sci. Eng. A, 785 (2020), 139286.
• 50)   M. X.  Yang,  F. P.  Yuan,  Q. G.  Xie,  Y. D.  Wang,  E.  Ma and  X. L.  Wu: Acta Mater., 109 (2016), 213.
• 51)   Y. G.  Kim,  J. M.  Han and  J. G.  Lee: Mater. Sci. Eng. A, 114 (1989), 51.
• 52)   R. K.  You,  P. W.  Kao and  D.  Gan: Mater. Sci. Eng. A, 117 (1989), 141.
• 53)   O.  Grässel,  L.  Krüger,  G.  Frommeyer and  L. W.  Meyer: Int. J. Plast., 16 (2000), 1391.
• 54)   Y.  Kimura,  K.  Handa,  H.  Hayashi and  Y.  Mishima: Intermetallics, 12 (2004), 607.
• 55)   J. D.  Yoo,  S. W.  Hwang and  K.-T.  Park: Metall. Mater. Trans. A, 40 (2009), 1520.
• 56)   K. M.  Chang,  C. G.  Chao and  T. F.  Liu: Scr. Mater., 63 (2010), 162.
• 57)   C. L.  Lin,  C. G.  Chao,  H. Y.  Bor and  T. F.  Liu: Mater. Trans., 51 (2010), 1084.
• 58)   S.-J.  Park,  B.  Hwang,  K. H.  Lee,  T.-H.  Lee,  D.-W.  Suh and  H. N.  Han: Scr. Mater., 68 (2013), 365.
• 59)   H.  Ding,  D.  Han,  J.  Zhang,  Z.  Cai,  Z.  Wu and  M.  Cai: Mater. Sci. Eng. A, 652 (2016), 69.
• 60)   S. S.  Sohn,  H.  Song,  M. C.  Jo,  T.  Song,  H. S.  Kim and  S.  Lee: Sci. Rep., 7 (2017), 1255.
• 61)   J.  Moon,  S.-J.  Park,  J. H.  Jang,  T.-H.  Lee,  C.-H.  Lee,  H.-U.  Hong,  H. N.  Han,  J.  Lee,  B. H.  Lee and  C.  Lee: Acta Mater., 147 (2018), 226.
• 62)   H.  Ishii,  K.  Ohkubo,  S.  Miura and  T.  Mohri: Mater. Trans., 44 (2003), 1679.
• 63)   S. S.  Sohn,  S.  Hong,  J.  Lee,  B.-C.  Suh,  S.-K.  Kim,  B.-J.  Lee,  N. J.  Kim and  S.  Lee: Acta Mater., 100 (2015), 39.
• 64)   Y.  Li,  Y.  Lu,  W.  Li,  M.  Khedr,  H.  Liu and  X.  Jin: Acta Mater., 158 (2018), 79.
• 65)   K. T.  Luo,  P.-W.  Kao and  D.  Gan: Mater. Sci. Eng. A, 151 (1992), L15.
• 66)   O.  Acselrad,  J.  Dille,  L. C.  Pereira and  J.-L.  Delplancke: Metall. Mater. Trans. A, 35 (2004), 3863.
• 67)   Y. H.  Tuan,  C. S.  Wang,  C. Y.  Tsai,  C. G.  Chao and  T. F.  Liu: Mater. Chem. Phys., 114 (2009), 595.
• 68)   S. C.  Chang,  J. Y.  Liu and  H. K.  Juang: Corrosion, 51 (1995), 399.
• 69)   G. D.  Tsay,  C. L.  Lin,  C. G.  Chao and  T. F.  Liu: Mater. Trans., 51 (2010), 2318.
• 70)   D. K.  Han,  Y. M.  Kim,  H. N.  Han,  H. K. D. H.  Bhadeshia and  D.-W.  Suh: Scr. Mater., 80 (2014), 9.
• 71)   E. J.  Song,  H. K. D. H.  Bhadeshia and  D.-W.  Suh: Scr. Mater., 87 (2014), 9.
• 72)   D. K.  Han,  S. K.  Lee,  S. J.  Noh,  S.-K.  Kim and  D.-W.  Suh: Scr. Mater., 99 (2015), 45.
• 73)   I.-J.  Park,  K.-H.  Jeong,  J.-G.  Jung,  C. S.  Lee and  Y.-K.  Lee: Int. J. Hydrog. Energy, 37 (2012), 9925.
• 74)   K. H.  So,  J. S.  Kim,  Y. S.  Chun,  K.-T.  Park,  Y.-K.  Lee and  C. S.  Lee: ISIJ Int., 49 (2009), 1952.
• 75)   M.  Koyama,  E.  Akiyama and  K.  Tsuzaki: ISIJ Int., 52 (2012), 2283.
• 76)   T. A.  Timmerscheidt,  P.  Dey,  D.  Bogdanovski,  J.  von Appen,  T.  Hickel,  J.  Neugebauer and  R.  Dronskowski: Metals, 7 (2017), No. 7, 1.