Obtaining synergy of strength and ductility in low-density steel via two-step annealing treatment

Adding light elements such as Mn and Al to steels can significantly reduce the density and obtain good strength-ductility of steels, making it a suitable choice for the development of lightweight materials. In this study, the microstructure and mechanical properties of Fe-30Mn-10Al-1C low density steel after cold rolling and annealing heat treatment were investigated. The results demonstrated that nano-sized κ-carbides are unevenly precipitated during the low temperature (500 °C) annealing process, which pin austenite grain boundaries and restrain the growth of austenite grains during the subsequent high temperature (900 °C) annealing process. As a consequence, the two-step annealed sample exhibits a fine and heterogeneous grain structure, leading to improvements of approximately 30 MPa in the yield strength to 752 MPa while keeping almost the same uniform elongation. Compared to the one step annealed sample (900 °C annealed for 3 minutes), the two-step annealed sample can achieve a high strength-ductility product (58 GPa·%) and an excellent impact toughness (128 J/cm2 at -40 °C).


Introduction
Fe-Mn-Al-C low density steels have received great attention due to their high strengths and large tensile elongations [5].Due to the addition of a large amount of Al element, the densities of the steels are reduced, and -carbides are generated in the matrix, which significantly enhances the strength of the steels [5].Zhang et al. [6] reported that the yield strength of the steels can be improved to 1216 MPa by precipitating nano-carbides through aging process.However, excessive precipitation of -carbides may damage the ductility and toughness by reducing the work hardening abilities of the steels [7].
In order to reduce the adverse effects of -carbides on ductility and toughness, the combination of recrystallization-induced softening and the precipitation of a small amount of -carbides was proposed to improve the mechanical properties of the steels through.Zhang et al. [8] obtained fine recrystallized austenite grains with a small amount of -carbides through cold rolling and annealing, and the prepared Fe-27Mn-9Al-1C low-density steel had a tensile strength of 1125 MPa and a total elongation of 30.8%.Liu et al. [9] tailored the local recrystallization and intra-grain -carbide strengthening of Fe-11Mn-10Al-1.2Csteels, and a tensile strength to 1089 MPa and a total elongation of 35.8% were obtained.
As mentioned above, traditional processing methods can only compromise between the strength and plasticity of -carbides.This study designs a two-step annealing strategy to make reasonable use of carbides, avoiding their harmful effects on ductility and toughness while improving the material's strength.In the first step of low-temperature annealing, the formation of -carbides is controlled, and they precipitate unevenly within and between grains.In the second step of high-temperature annealing, -carbides pin the austenite grains and ultimately dissolved, resulting in a single-phase austenite structure with heterogeneous grain structure and an excellent combination of strength, plasticity, and toughness.

Experiment
The experimental steels used in this study were Fe-30Mn-10Al-1C, which were produced by vacuum induction melting to ingot.The ingot was heated to 1150 C for 4 h and then hot forged, followed by cooling at 850 C for 1 h.The forged ingot was cut and cold rolled to a reduction of 80% in thickness.Two annealing processes were adopted for the cold rolled samples: the first process involved a short annealing at 900 C for 3 min (denoted as sample A).The second process involved a two-step annealing, with the first annealing at 500 C for 3 min and the second annealing at 900 C for 3 min (denoted as sample B).The initial density of the samples was measured by the Archimedes method and was tested to be 6.75 g/cm 3 .Dog-bone shaped tensile specimens and Charpy V-notch specimens were prepared by wire-electrode cutting.Tensile specimens were tested (with a gauge length of 10 mm, width of 4 mm, and thickness of 1.5 mm) at room temperature using a universal tensile machine at a constant strain rate of 1 × 10 -3 s -1 .A pendulum impact tester (maximum energy 450 J) was used to test the impact toughness of samples (10 mm × 5 mm × 55 mm) at -40 C.
A Bruker D8 Advance X-ray diffractometer (XRD, Cu Kα radiation, 5° min -1 scan rate, 0.02° scan step) was used for the analysis of the phase composition of the samples.The annealed structures were observed in a Zeiss Gemini SEM500 scanning electron microscope (SEM), combined with electron backscatter diffraction (EBSD) for orientation analysis.EBSD samples were prepared by electrolytic polishing using a mixture of perchloric acid and ethyl alcohol (5:95 by volume) at 30 V. TEM foils were prepared by conventional twin-jet technique in the solutions in a mixture of perchloric acid and ethyl alcohol (5:95 by volume) at -20 °C.The annealed samples were observed using a JEOL 2100F transmission electron microscope (TEM) at 200 kV.

Mechanical properties
Figure 1(a) shows the engineering strain-stress curves for the two annealed samples.Samples A and B yield continuously in tension and no significant yielding terrace is observed.The yield strength of sample A is 721 MPa, the tensile strength is 1065 MPa and the elongation is 53.6%.Sample B shows an increase in strength with a yield strength of 752 MPa, a tensile strength of 1076 MPa and an elongation of 53.5%.The tensile properties and impact toughness of samples A and B are recorded in table 1. Sample B with the two-step annealed process has essentially the same elongation and impact toughness and possesses a high strengthen and plasticity product of 58 GPa% compared to sample A with the one-step annealing.However, the yield strength and ultimate tensile strength are increased by about 30 MPa and 10 MPa respectively.Figure 1(b) shows the work-hardening curves for the two annealed samples.Sample A and sample B have similar work hardening behaviour and both show a continuous decrease, mainly caused by strain localization [7].
There are three typical stages of work hardening, associated with the slip band refinement mechanism (DBSR).Stage 1: When the accumulation of dislocations on the slip surface reaches saturation, slip bands penetrate the entire grain and work-hardening rate decreases rapidly (from the true stain of 0.004-0.007).Stage 2: As strain increases, new dislocation sources open up to accommodate further plastic deformation and new slip bands are formed, causing the spacing between slip bands and the work-      The grain size (including twins) of sample A and sample B were identified by Channel 5 and the grain size distribution is shown in figure 4. The average grain size of austenite for sample A and sample B are 2.9 μm and 2.3 μm, respectively, and the maximum grain size is around 20 μm for both samples, but the proportion of grains smaller than 1 μm is significantly higher for sample B than for sample B.
Figure 5 shows the TEM results of sample B when annealed at 500 °C for 3 min.As seen from the bright field images, a large number of dislocation entanglements and dislocation cells due to cold rolling can be observed, the sample does not undergo recrystallisation at this state.From the electron-selected diffraction, a clear super-dot formation can be observed.In combination with the dark-field image, a large number of nanoscale (<5 nm) granular κ-carbides are found to be unevenly distributed within the austenite grain and grain boundaries.Figure 6 shows the final state TEM results for sample B. The dislocation substructure has largely disappeared and no obvious hyperpitting of the κ carbide is observed.The κ-carbide formed by annealing at 500 °C for 3 min is almost dissolved.

Discussion
Combining the EBSD and TEM results, the main reason for the finer austenite grain size in the two-step annealed sample B compared to the one-step annealed sample A is that a large amount of κcarbide precipitates in sample B after annealing at 500 °C .These κ-carbides pin the austenite grain boundaries and the growth of austenite grain recrystallisation is thus inhibited during the second step of annealing process at 900 °C, which eventually leads to an inhomogeneous grain size distribution in sample B [16].The grain refinement results in a higher work-hardening rate and improved yield strength    in the low strain state (<0.05) of sample B compared to sample A. During tensile deformation, the slip band structure is accelerated for refinement in sample B with small grains, i.e. small grains at low strains produces dynamic microstructure refinement more quickly leading to higher initial work hardening rates [17].The strain distribution between the soft coarse and hard fine grains is an important factor in improving the mechanical properties of the two-step annealed samples, while avoiding premature fracture due to severe strain concentration in the finer grains [7,18], which leads to an increase in the yield strength of the specimen while maintaining excellent plasticity.The yield strength of this lowdensity steel can be calculated using the following strengthening mechanism: where σ0 is Peierls stress, σss is solid solution strengthening, σHP is grain boundary strengthening, σdis is dislocations strengthening.As samples A and B are of the same composition and are both fully recrystallized, the contributions of σ0, σss and σdis are approximately the same.The difference in yield strength between samples A and B is mainly due to grain boundary strengthening.The grain boundary strengthening contribution is described by the Hall-Petch equation [16] where d is average grain size and the Hall-Patch constant Ky, 461 MPaμm 1/2 , was used in this study [6].The calculated grain boundary strengthening contribution of sample A is approximately 271 MPa and that of sample B is approximately 304 MPa, with a difference of 33 MPa, which is very close to the experimental difference in the yield strength.
In addition, the grain size of the samples prepared by the two-step annealing method has a bimodal distribution and is characterized by a laminar distribution of coarse and fine crystalline regions, which is heterogeneous.The geometrically necessary dislocations resulting from the coordinated deformation of the fine and coarse crystalline regions in the heterostructure increase the plasticity of the steels.
Figure 7 shows a comparison of the properties of the steel in this study and other reported lowdensity steels.Figure 7a shows the plasticity strengthen product-impact toughness distribution and figure 7b shows the specific strength-impact toughness distribution.It is observed that the properties of the two-step annealed sample B are at the top right of the distribution.Compared to other low-density steels reported, an excellent combination of large plasticity strength product of 58 GPa%, good impact toughness of 128 J/cm 2 at -40 °C and high specific strength of 159.4 MPag -1 cm 3 can be obtained in the present low-density steel by two-step annealing of the cold rolled materials.

Conclusions
The effect of two-step annealing on the tensile properties and impact toughness of Fe-30Mn-10Al-1C austenitic low-density steel was investigated in this study.Based on EBSD, XRD and TEM characterizations, the effects of the two-step annealing process on the organization and grain size of the experimental steels were analysed.The conclusions are as follows: (1) The microstructures of annealed samples are single austenitic phase.After the two-step annealing process, the austenite grains are refined and show a bimodal distribution with fine and coarse grains exhibiting a lamellar structure.
(2) The two-step annealed sample has good overall mechanical properties.A yield strength of 752 MPa, a tensile strength of 1076 MPa, an elongation of 53.5% and specific strength of 159.4 MPag -1 cm 3 can be achieved while maintaining an ultrahigh strength plastic product of 58 GPa% and an excellent impact toughness of 128 J/cm 2 at -40 °C.

Figure 2
shows the XRD patterns of sample A and sample B. Both samples are austenitic single phase with no peak lines of κ-carbide.

Figure 3
Figure 3 shows the EBSD orientation maps for sample A and sample B. The microstructures of both samples are of fully recrystallised austenitic equiaxed grains, with very fine sizes of recrystallised

Figure 3 .
Figure 3. EBSD results of annealed samples (a) grain distribution of sample A, (b) inverse pole figure (IPF) plot of sample A, (c) grain distribution of sample B, (d) IPF plot of sample B.

Figure 4 .
Figure 4. Grain size distribution of annealed samples.(a) Grain size diagram of sample A austenite, (b) Grain size diagram of sample B.

Figure 5 .
Figure 5. TEM images of sample B after first step annealing (with a [011] beam direction, BD = [011]).(a) bright field image inside the austenite grain, (b) dark field image inside the austenite grain, (c) bright field image of the boundary of austenite grain, (d) dark field image of the boundary of austenite grain.

Figure 6 .
Figure 6.TEM image of sample B after the second annealing step with dislocation substructure largely disappeared and no obvious hyperpitting of the -carbide observed (BD = [011]).

Figure 7 .
Figure 7. Performance comparison with the values for other low-density steels reported in literature.(a)PSE-Ak, (b) Specific strength-Ak.

Table 1
Mechanical properties of annealed samples.