1 Introduction

In recent years, the automotive industry has focused its efforts on reducing energy consumption and CO2 emissions through weight reduction of vehicles.[1] In the literature, Fe-Mn-Al-C steels have been investigated as candidate materials for light-weight automotive steel over the past several decades.[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15] Steels of these grades still have many problems that must be addressed before they can be applied in the automotive industry. For example, the high carbon content has made it difficult to weld automotive structures and its high production cost is also an obstacle to its broader application. Nevertheless, many studies have attempted to develop high-performance light-weight Fe-Mn-Al-C steel due to its excellent mechanical properties and low density.[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16] A previous study reported that the addition of 1 wt pct Al has the effect of reducing the density by approximately 0.1 g/cm3.[1]

In addition, high-performance light-weight Fe-Mn-Al-C steels may be considered as structural materials for military vehicles because mobility is very important for such vehicles. The use of light-weight steel can improve the mobility of vehicles through weight reductions.

Fe-Mn-Al-C steels can be categorized as austenitic, duplex, or ferritic base steels, according to their matrix phase constituents.[1] Austenitic Fe-Mn-Al-C steels cover Fe-(15-30)Mn-(8-12)Al-(0.5-1.2)C (in wt pct)[1] and show a combination of high strength and good total elongation.[2,3,4,5,6,7,8] Lin et al.[2] investigated the effect of C on the microstructures and tensile properties of Fe-30Mn-8.5Al-C steels and reported that the yield strength was clearly improved by an increase in the C content, with nearly equivalent ductility. From a TEM analysis, they confirmed that this was due to the precipitation of the fine κ-carbide of (Fe,Mn)3AlC.[2,8] Yoo and Park[3] studied the correlation between the room-temperature tensile behavior and the deformed microstructure of Fe-27.8Mn-9.1Al-0.79C steel and showed that high ductility with continuous strain hardening occurred due to microband-induced plasticity (MBIP). Gutierrez-Urrutia and Raabe[4] introduced a type of high-performance austenitic Fe-30.5Mn-8.0Al-1.2C steel, known as Simplex steel. This steel showed an outstanding strain-hardening capacity resulting from dislocation substructure refinement and the subsequent activation of deformation twinning. Research on improvements of the mechanical properties utilizing B2 phase, a FeAl type of hard intermetallic compound, is another important issue related to Fe-Mn-Al-C steel.[5,6,7] Kim et al.[5] developed an advanced light-weight steel with a higher specific strength than Ti alloys using the FeAl-type hard intermetallic compound (B2). Meanwhile, some researchers have attempted to find the mechanism of κ-carbide precipitation in austenitic light-weight steels.[9,10,11] Moon et al.[10] investigated the precipitation mechanism of κ-carbide using an atom probe tomography (APT) analysis and first-principle calculations, finding that the addition of Si accelerated κ-carbide precipitation, whereas an addition of Mo delayed it.

Ferritic Fe-Mn-Al-C steels cover Fe-(0-8)Mn-(5-8)Al-(0-0.3)C (in wt pct).[1,12] This type of alloys can be strengthened by B2-ordered FeAl or D03-ordered Fe3Al.[1] Rana et al.[12] investigated the effect of the Al content on microstructural and mechanical properties of ferritic Fe-Al steels and reported that the deep drawability and elongation of 6.8 wt pct Al-containing Fe-Al steel were lower than those of IF grades but better than those of DP grades.

Duplex Fe-Mn-Al-C steels cover Fe-(5-30)Mn-(3-10)Al-(0.1-0.7)C (in wt pct)[1,13,14] and show intermediate properties between austenitic and ferritic steels.[13,14] Lee et al.[13] and Park et al.[14] studied the deformation behavior of duplex Fe-8.1Mn-5.3Al-0.23C steel using nanoindentation and in situ electron backscattered diffraction (EBSD). Meanwhile, Frommeyer and Brüx[1,15] reported high-strength light-weight steels with the generic composition of Fe-(18-28)Mn-(9-12)Al-(0.7-1.2)C, also known as TRIPLEX steels. TRIPLEX steels consist of three major phases of an austenite matrix, 5 to 15 vol pct ferrite, and nano-sized κ-carbide at less than 10 vol pct.

Most previous research in this area used fine κ-carbide as a strengthener of Fe-Mn-Al-C steels.[2,3,4,5,6,7,8,9,10,11,12,13,14,15,16] Meanwhile, few studies have devoted effort to the study of strengthening steels using microalloying elements such as Nb and V.[17,18] The results thus far have clearly shown that the addition of Nb and/or V has an effect on grain refinement and precipitation hardening, but the researchers did not present a detailed microscopic analysis of the precipitation behavior and its effect on the mechanical properties. In addition, it is somewhat difficult to find work on the effect of an aging treatment on the microstructural and mechanical properties of Fe-Mn-Al-C light-weight steels in References 2,3,4,5,6,7,8,9,10,11,12,13,14,15, through 16.

Taking all of these considerations, here we develop austenitic Fe-30Mn-9Al-0.9C light-weight steels containing Nb and V for possible application to military vehicles and investigated the microstructural evolution and mechanical properties during an aging treatment. The microstructural evolution and precipitation behavior during aging were carefully analyzed using scanning electron microscopy (SEM), EBSD, and transmission electron microscopy (TEM). In addition, we evaluated the mechanical properties using Vickers hardness tests and tensile tests, and discussed the correlation between the microstructural evolution and the age-hardening behavior.

2 Experimental Procedures

The chemical composition of the light-weight steel examined in this investigation is presented in Table I. Ingots of the three Fe-Mn-Al-C steels were fabricated using a commercial vacuum induction melting (VIM) furnace. Compared to steel A, steels B and C additionally included microalloying elements of Nb and/or V. Note that steel B had only V, while steel C had both V and Nb. Ingots were homogenized for 2 hours at 1423 K (1150 °C) and then hot-rolled into plate samples of 8 mm in thickness. Figure 1 shows the schematic schedule used for the heat treatment of the samples after hot rolling. Each sample was solution-treated for 2 hours at 1323 K (1050 °C) and then quenched with water. Here, the solution treatment temperature of 1323 K (1050 °C) was selected in consideration of the equilibrium phase diagram reported by Ishida et al.,[19] who investigated the phase constituents of Fe-(20-30) wt pct Mn-(0-10) wt pct Al-C alloys and reported that the temperature range of 1173 K to 1473 K (900 °C to 1200 °C) is the single phase region of austenite in the Fe-30 wt pct Mn-9 wt pct Al-0.9C-based alloy, corresponding to chemical composition of the tested alloys in this study.

Table I Chemical Composition of Tested Alloys and Austenite Grain Size of Solution-Treated Samples
Full size table
Fig. 1
figure 1

Schematic illustration of the heat treatment of the tested alloy

Full size image

The true stress–strain relation of the solution-treated samples (ASTM E8M) was obtained using a tensile test machine (INSTRON 5882, Canton, MA) at a nominal strain rate of 1.33 × 10−3 s−1, and the austenite grain size of the solution-treated samples was measured by the linear intercept method according to the ASTM standard. Subsequently, the solution-treated samples were aged at 823 K (550 °C) for up to 10,000 minutes. The mechanical properties of the aged samples were examined using a Vickers hardness test machine (FM-700, Future-Tech Corp., Japan) under a load of 200 g (0.2 kgf).

The microstructures were observed using a SEM (JSM-7001F, JEOL, Japan) equipped with an EBSD (HKL Nordlys Channel 5). Specimens for the SEM microstructure observation were prepared by mechanical polishing and chemical etching in a mixed solution of 90 ml ethanol and 10 ml nitric acid. Specimens for the EBSD analysis were mechanically polished by a colloidal silica suspension for the final polishing stage. The EBSD and X-ray diffraction (XRD; D/Max 2500; Rigaku Corporation, Tokyo, Japan) analyses were carried out to identify the microstructural evolution during the aging process. Particles were observed by TEM (JEM-2100F, JEOL, Japan) and identified by selected-area diffraction pattern (SADP) analysis. Thin foil specimens for the TEM analysis were prepared by twin-jet electrolytic polishing at 20 V and 200 mA with a mixed solution of 10 pct perchloric acid and 90 pct methanol at 243 K (−30 °C).

3 Results and Discussion

3.1 Microstructure and Tensile Properties After a Solution Treatment

Figure 2 shows SEM micrographs of the tested alloy after a solution treatment. The matrix consisted of austenite, and some annealing twins were also observed. Compared to steel A, the particles were observed with a high fraction in the matrix of steels B and C, and they were randomly distributed regardless of the grain boundaries or grain interior. In Figure 3, the particles in steels B and C were identified as VC and V-enriched (V,Nb)C particles of MC-type carbides, respectively, through the TEM observation and SAD pattern analysis. The diffraction patterns in Figure 3 show that the VC and (V,Nb)C exhibit a cube–cube orientation relationship [001] γ //[001]MC, (001) γ //(001)MC with respect to the austenite matrix.[20] As shown in Table I, steel B and steel C contain V and/or Nb, which are strong carbide formers. Considering earlier published results[17] and the chemical compositions of the tested alloys, it is predicted that VC and (V,Nb)C particles may be nucleated during solidification in VIM process and subsequently coarsened during homogenization at 1423 K (1150 °C) and a solution treatment at 1323 K (1050 °C) before and after the hot rolling. In addition, it is expected that the fine particles shown in Figure 3 were newly precipitated during the rapid cooling after the solution treatment. Meanwhile, the austenite grain sizes of steels B and C were much smaller than that of steel A, as shown in Figure 2 and Table I. This may have arisen because VC and (V,Nb)C inhibited austenite grain growth during the solution treatment.

Fig. 2
figure 2

SEM micrographs after the solution treatment: (a) steel A, (b) steel B, and (c) steel C

Full size image
Fig. 3
figure 3

TEM micrographs showing particles after the solution treatment: (a, b) bright-field image and SAD pattern analysis of VC particles in steel B and (c, d) bright-field image and SAD pattern analysis of the V-enriched (V,Nb)C particles in steel C, respectively

Full size image

Figure 4 shows the true stress–true strain curves with the corresponding strain-hardening rate of the solution-treated samples, and all samples exhibited continuous yielding. Steel A showed the lowest yield and tensile strength, while the strength gradually increased with the addition of V and Nb. This may be due to the effects of grain refinement and precipitation hardening, as shown in Figure 2.

Fig. 4
figure 4

True stress–strain curves of solution-treated samples with the corresponding strain-hardening rate as a function of the true strain (Color figure online)

Full size image

Here, we analyzed the grain refinement and precipitation hardening contribution of steel B separately. The yield strength of polycrystalline materials can be calculated using Eq. [1]:

$$ \sigma_{\text{y}} = \sigma_{0} + \sigma_{\text{ss}} + \sigma_{{ 0 {\text{ro}}}} + \sigma_{\text{HP}}, $$
(1)

where σ y is the overall yield strength, σ 0 is the inherent yield strength of the matrix, σ ss is the solid solution contribution, σ Oro is the precipitation strengthening contribution by the Orowan mechanism, and σ HP is the grain boundary contribution according to the Hall–Petch relationship. In Eq. [1], the increase in the strength (σ Oro) by the Orowan mechanism can be expressed as follows[21]:

$$ \sigma_{{ 0 {\text{ro}}}} = \frac{G \cdot b \cdot \sqrt 3 }{\lambda }, $$
(2)

where G and b are the shear modulus of the matrix and the Burgers vector, respectively. λ is the interparticle spacing between the precipitates which is expressed by[21]

$$ \lambda = \frac{4 \cdot (1 - f) \cdot r}{3 \cdot f}, $$
(3)

where f is the precipitate volume fraction and r is the precipitate size. Therefore, by applying Eq. [3], Eq. [2] can be rewritten as follows:

$$ \sigma_{{ 0 {\text{ro}}}} = \frac{3 \cdot G \cdot b \cdot f \cdot \sqrt 3 }{4 \cdot (1 - f) \cdot r}. $$
(4)

In order to calculate the Orowan strengthening effect of Eq. [4], we selected the values of G and b to 80 GPa and 2.58 × 10−10 m, respectively, from References 13 and 22. In addition, we measured the precipitate size (r = 49.9 nm) and volume fraction (f = 2.12 pct) of steel B, using Image Analyzer software. From Eq. [4], the calculated result of Orowan strengthening in steel B is 11.6 MPa.

Meanwhile, in Eq. [1], the strength increase (σ HP) according to the Hall–Petch relationship can be expressed as[23]

$$ \sigma_{\text{HP}} = \frac{{k_{\text{y}} }}{\sqrt d }, $$
(5)

where k y is the Hall–Petch coefficient for the flow stress and d is the grain size. In Eq. [5], we selected the value of k y to be 14.55 MPa mm1/2 from Reference 23. In order to calculate the effect of the grain refinement of steel B as compared to that of steel A, we used the measured grain size data shown in Table I. That is, from Eq. [5], the calculated results of the grain boundary strengthening in steels A and B are 51.3 and 112.2 MPa, respectively. Thus, the difference in the grain boundary strengthening between steels A and B is close to 60.9 MPa. In summary, it is estimated that the yield strength of steel B is 72.5 MPa higher than that of steel A, and this estimation is acceptable as compared to the experimental results, i.e., the yield strengths of steels A and B in Figure 4 are 352.1 and 436.4 MPa, respectively, with a difference of 84.3 MPa.

Compared to steel B, steel C had higher strength, indicating that the combined addition of V and Nb improved the precipitation hardening by promoting the additional nucleation of carbides. Meanwhile, steel A showed extensive strain-hardening behavior. Previous research[3,24,25] studied the deformation mechanisms of compositions similar to that of steel A in the present study. According to these results,[3,24,25] the continuous strain hardening of steel A may be caused by microband-induced plasticity in that the formation of microbands and their intersections occurred during the tensile deformation process, leading to a state of high dislocation density and resulting in continuous strain hardening. Figure 5 shows TEM micrographs of the microbands which developed at local areas during tensile deformation of 30 pct. The SAD patterns in Figures 5(b) and (d) indicate that the deformed samples in Figures 5(a) and (c) consist of full austenite. That is, the SAD patterns show only austenite spot patterns, implying that the band structures in Figures 5(a) and (c) were not microtwins.

Fig. 5
figure 5

TEM micrographs showing the microbands developed during tensile deformation of 30 pct: (a, b) bright-field image and SAD pattern analysis of steel A and (c, d) bright-field image and SAD pattern analysis of steel C, respectively (Color figure online)

Full size image

Interestingly, during the process of tensile deformation, the strain-hardening rate of steel A increased steadily, as shown in Figure 4, due to the microband-induced plasticity (MBIP) from the onset of plastic deformation to ε = 25 pct, while such behavior was weakened and not observed in the Nb- and V-added steels despite MBIP. This may have arisen because MC carbides with a high fraction had a dominant effect on the strengthening of steels B and C, despite the fact that the strengthening by MBIP was also expected.

3.2 Microstructural Evolution and Precipitation Behavior During Aging

Figure 6 shows SEM images of the samples after aging at 823 K (550 °C) for 100, 1000, and 10,000 minutes. While fine particles were mainly formed in the austenite matrix at the initial stage of aging, coarse secondary phases (marked by the yellow arrows in Figure 6) were found after 10,000 minutes of aging. Figure 7 shows TEM micrographs of steel A after aging at 823 K (550 °C) for 1000 minutes. The dark-field image in Figure 7(b) shows that nano-sized particles were homogeneously precipitated in the matrix. From the SAD pattern analysis shown in Figures 7(c) through (f),[8] nano-sized particles in Figure 7(b) were identified as κ-carbides, which have an E21 perovskite crystal structure with a composition of (Fe,Mn)3AlC.[1,8,15] The crystallographic relationship between the austenite (γ) matrix and the κ-carbide was confirmed to be [011] γ //[011] κ and [001] γ //[001] κ . Figure 8 shows TEM micrographs of the precipitation behavior of steel B after aging at 823 K (550 °C) for 1000 minutes. The dark-field images in Figures 8(b) and (c) indicate that two types of particles exist in steel B after aging. Figure 8(b) shows VC particles, which were precipitated during the VIM and solution treatment process, shown in Figure 3, and which remained stable during aging. The fine particles shown in Figure 8(c) were newly formed κ-carbides during aging, showing a uniform distribution. Figures 8(d) and (e) are the corresponding SAD patterns showing the orientation relationship among the austenite matrix, the VC particles, and the κ-carbides. We also observed TEM micrographs of steel C after aging at 823 K (550 °C) for 1000 minutes, confirming the same SAD pattern noted with steel B, i.e., κ-carbide was newly precipitated during aging in addition to the formation of (V,Nb)C particles during the VIM and solution treatment process.

Fig. 6
figure 6

SEM micrographs of samples aged at 823 K (550 °C) until 10,000 min: (a) steel A/aged for 100 min, (b) steel A/1000 min, (c) steel A/10,000 min, (d) steel B/100 min, (e) steel B/1000 min, (f) steel B/10,000 min, (g) steel C/100 min, (h) steel C/1000 min, and (i) steel C/10,000 min, respectively (Color figure online)

Full size image
Fig. 7
figure 7

TEM micrographs of steel A after aging at 823 K (550 °C) for 1000 min: (a) bright-field image, (b) axial dark-field image of the κ-carbides, (c, d) SAD pattern of Z = [011] γ , and (e, f) SAD pattern of z = [001] γ

Full size image
Fig. 8
figure 8

TEM micrographs of steel B after aging at 823 K (550 °C) for 1000 min: (a) bright-field image, (b) axial dark-field image of the VC particles, (c) dark-field image of the κ-carbides, and (d, e) SAD pattern of Z = [011] γ , respectively

Full size image

As mentioned above, coarse secondary phases were formed after 10,000 minutes of aging, as shown in Figure 6. Here, we confirmed that this secondary phase was a mixture of ferrite and β-Mn,[26] through SEM and EBSD analyses. Figures 9(a) and (b) show a SEM micrograph and EBSD phase map of steel A after aging for 10,000 minutes, respectively. In addition, Figure 9(c) presents the results of the EDS line scanning data for each element. From Figure 9, it can be indicated that a secondary phase was formed along the austenite grain boundaries. This formation of ferrite and β-Mn can be predicted from the phase diagram reported in previous work.[27] According to the phase diagram of Fe-30 wt pct Mn-9. pct Al-1.0C-based alloy, which is very similar to the present tested alloys, developed by Kim et al.,[27] the four phases of austenite, κ-carbide, ferrite, and β-Mn are stable at 823 K (550 °C). The formation of ferrite and β-Mn was closely related to the precipitation of the carbides shown in Figures 7 and 8. First, κ-carbides were nucleated at an early stage of aging, resulting in a reduction of the solute C content in the matrix. With the progress of aging, the solute C content gradually decreased, and this reduced the austenite stability. Next, ferrite was able to form along the grain boundary of the metastable austenite, after which strong austenite-stabilizing elements such as Mn were enriched at the phase boundaries between the austenite and the ferrite. As a result, β-Mn was formed in the Mn-enriched region around the ferrite and grew. In order to provide a qualitative assessment of the precipitation sequence, XRD analyses were performed. Figure 10 shows the results, which indicate that κ-carbide was initially precipitated during aging, and ferrite and β-Mn were formed after aging for 10,000 minutes. Meanwhile, Table II shows the results of the X-ray analysis of the constituent phases in the aged samples.

Fig. 9
figure 9

β-Mn observed in steel A after aging for 10,000 min: (a) SEM image, (b) EBSD phase map, and (c) EDS line scanning profile. Face-centered cubic ( FCC, austenite), body-centered cubic ( BCC, ferrite), and β-Mn ( β-Mn); misorientation angle > 5 deg (Color figure online)

Full size image
Fig. 10
figure 10

X-ray diffraction patterns of the samples aged at 823 K (550 °C): (a) steel A, (b) steel B, and (c) steel C

Full size image
Table II X-ray Analysis for Constituent Phases
Full size table

3.3 Changes in the Mechanical Properties During Aging

Figure 11 shows the Vickers hardness as the aging time elapsed at 823 K (550 °C). Before the aging treatment, steels B and C had a higher hardness as compared to steel A due to the effects of grain refinement and precipitation hardening by the MC carbides, as shown in Figure 3, revealing some consistency with the tensile test results in Figure 4. At an early stage of aging, the hardness of all alloys gradually increased due to the precipitation of nano-sized κ-carbides, as shown in Figures 7 and 8. It is interesting that the age hardening of steel A was greater than that of steels B and C, with the hardness of all samples eventually becoming similar after prolonged aging above 300 minutes. This may have arisen because newly formed κ-carbide exerted greater precipitation hardening as compared to MC carbide. According to Reference 21, the precipitation hardening effect improves when the fraction of fine particles increases. As shown in Figures 2, 7, and 8, the size of the κ-carbide was much finer than that of the MC carbides. Therefore, the hardness difference among the solution-treated samples by MC carbides may be compensated by the precipitation of nano-sized κ-carbide during aging. In addition, it is expected that the κ-carbide in steel A was precipitated more than that in steels B and C because the solute C in steels B and C was consumed separately for the precipitation of MC and κ-carbide, while most of the solute C in steel A was able to be produced entirely for the precipitation of κ-carbide. The hardness remained stagnant between the aging times of 1000 and 3000 minutes, most likely due to the saturation of κ-carbide. Hardness increased again after 3000 minutes of aging, which is due to the formation of ferrite and brittle β-Mn. Previous work reported that the formation of β-Mn led to age hardening and a loss of ductility.[28] Figure 12 shows the strain–stress curves of the solution-treated and aged samples. The strain–stress curves present evidence of identical behavior with regard to the Vickers hardness as the aging time elapses, i.e., the yield strength increased with the aging time due to the precipitation of κ-carbide.

Fig. 11
figure 11

Vickers hardness vs aging time at 823 K (550 °C)

Full size image
Fig. 12
figure 12

Changes of the strain–stress curves during aging at 823 K (550 °C): (a) steel A, (b) steel B, and (c) steel C (Color figure online)

Full size image

4 Concluding Remarks

The aging behavior and mechanical properties of microalloyed austenitic Fe-30Mn-9Al-0.9C light-weight steels were explored and the following conclusions were drawn:

  1. (1)

    The addition of V and/or Nb contributed to grain refinement and greater precipitation strength due to the precipitation of MC carbides. In addition, the behavior of the strain-hardening rate was changed by the addition of Nb and/or V. The strain-hardening rate of Fe-30Mn-9Al-0.9C steel steadily increased during tensile deformation due to MBIP from the onset of plastic deformation to ε = 25 pct, while such behavior was weakened and not observed in V- and/or Nb-added steels despite the presence of MBIP.

  2. (2)

    At an early stage of aging, nano-sized κ-carbides were precipitated and the crystallographic relationship between the austenite (γ) matrix and the κ-carbide was confirmed to be [011] γ //[011] κ and [001] γ //[001] κ .

  3. (3)

    Coarse secondary phases of the mixture of ferrite and β-Mn were formed along the austenite grain boundaries after 10,000 minutes of aging, depending on the precipitation of the carbide. That is, ferrite was formed due to the weakening of the stability of austenite caused by carbide precipitation, and subsequently β-Mn was formed in the Mn-enriched regions around the ferrite.

  4. (4)

    At the initial stage of aging, the hardness of all alloys gradually increased with an increase in the aging time, due to the precipitation of nano-sized κ-carbide. The hardness remained stagnant between the aging times of 1000 and 3000 minutes, as the precipitation of the κ-carbide became saturated. The hardness increased again after 3000 minutes of aging, due to the formation of ferrite and brittle β-Mn.