Mechanical Properties and Microstructures of Highly Fe-Containing Al–Mg–Si Alloys Processed by Severe Plastic Deformation under High Pressure

Abstract

In this study, A6022-based Al–Mg–Si alloys with three additional Fe contents are processed by high-pressure torsion (HPT) and high-pressure sliding (HPS). Both processes yield a similar tensile strength exceeding 400 MPa. Fe intermetallics were finely and homogeneously fragmented to an average size of ∼2 µm by the HPT process. The high tensile strength is attributed to such a fine and homogeneous fragmentation of Fe intermetallics. It is also demonstrated that the finely fragmented Fe intermetallics play an important role to maintain finer grain size even after solution treatment.

1. Introduction

Based on economic and environmental demand which calls for the reduction of fuel consumption and CO₂ emission, aluminum alloys are extensively used in transportation industry to improve the fuel efficiency owing to its excellent mechanical properties, low weight, excellent corrosion resistance and good formability. Specially, aluminum alloys are suitable candidates for various applications in the automotive industry. For example, Al–Mg–Si alloys are well used for the automotive body panels due to their bake-hardenability. As one of the most environmentally friendly metals, aluminum has another advantage of 100% recyclable. However, the presence of high impurities such as Fe is inevitable during recycling processes. Fe impurities can significantly degrade mechanical properties of the Al–Mg–Si alloys.^1–4) This is mainly because the solubility of Fe in Al is very low and thus Fe exits as brittle Fe intermetallics.^5–10) Many studies have examined the effects of Fe on the microstructures and mechanical properties of the Al–Mg–Si alloys. Wu et al. reported that Fe impurities decreases the strength and ductility of a cast Al–Mg–Si alloy due to the presence of Al₈Si₆Mg₃Fe phases.¹⁾ Eisaabadi et al. found that the formation of β–Al₅FeSi phase decrease the strength and ductility of an Al–Mg–Si alloy.²⁾ Wang et al. confirmed that the formation of β–Al₅FeSi phase reduce not only the elongation but also fatigue properties of an Al–Mg–Si alloy.³⁾

However, it was reported that these brittle intermetallics can be fragmented and dispersed or even be dissolved within the matrix when Fe-containing Al alloys are subjected to severe plastic deformation (SPD).^11–16) Phongphisutthinan et al. showed that the negative effect of Fe is reduced in an Al–Mg–Si alloy using a caliber-rolling process.¹⁷⁾ Kim et al. reported that the hardness is increased due to the enhanced nucleation of precipitates on the fragmented β–Al₅FeSi phase in Al–Si–Cu–Mg–Fe alloy.¹⁸⁾ Duchaussoy et al. found that the thermal stability is improved in an Al–Fe alloy by the formation of Fe intermetallics when the alloy is processed by SPD through high-pressure torsion (HPT).¹⁹⁾ Cubero-Sesin et al. consolidated an Al–10% Fe alloy from a powder mixture using high-pressure torsion (HPT) and showed that the tensile strength is significantly enhanced as the strain is more intensely imparted.²⁰⁾

It is noted that the HPT process is well known a typical SPD process and has been recognized as a useful technique for grain refinement in many bulk metallic materials.^21–24) Since the HPT process is operated under high pressure with constrained conditions,^25,26) it is applicable to hard and less ductile materials^25–34) including intermetallics,^35–37) ceramics^38–45) and semiconductors.^46–49) Considering the industry application, although upscaling the HPT process is reported as described in a recent overview paper,^22,50) the form of sample shape is limited to disk for the HPT process. High-pressure sliding (HPS) was proposed as an alternative SPD process for grain refinement under high pressure.⁵¹⁾ The HPS process has a similar feature to the HPT process but the advantage is that it can be used with a bulk form of rectangular sheet^52–56) including consolidation of powders into the sheet form.⁵⁷⁾ Another advantage of the HPS process is that it is applicable to rod^58–60) and pipe⁶¹⁾ rather than round disk.

In this study, high Fe contents between 0.5–2 mass% (commonly up to 0.1 mass% in commercial aluminum alloys) were intentionally added to the newly proposed Al–Mg–Si alloys in order to investigate the effects of Fe impurities. HPT and HPS processes are performed not only to refine the grain size but also to fragment harmful Fe intermetallics into favorable finer sizes and homogeneous distributions to improve the mechanical properties. Mechanical properties are also evaluated after additional annealing treatment, and microstructural evolution is examined by transmission electron microscopy (TEM) and scanning electron microscopy (SEM) with electron backscatter diffraction (EBSD) analysis.

2. Experimental

The chemical compositions of Al–Mg–Si–Fe alloys examined in this study are shown in Table 1, which correspond to a commercial A6022 alloy and three alloys with 0.5%, 1.0% and 2.0% extra additions of Fe, (named hereafter 6022, 6022-0.5Fe, 6022-1.0Fe, and 6022-2.0Fe, respectively). Figure 1 shows a flow chart for how the alloys are processed and examined in this study. The numbers in circles in Fig. 1 represent the states where the samples were examined.

Table 1 Chemical compositions (in mass%).

Fig. 1

Flow chart of sample processing. Numbers in circles represent sample states for examination.

Disks with 10 mm diameter and rectangular strips with 100 mm length and 10 mm width were cut from cold-rolled sheets with 1.0 mm thickness for the HPT and HPS processings using a wire-cutting electrical discharge machine (EDM), respectively. Disks with 10 mm diameter and 1.0 mm thickness were also cut from an as-cast ingot for the HPT processing.

HPT process was then carried out at room temperature under a pressure of 2 GPa and concurrent rotation of the upper anvil with respect to the lower anvil for either 1 or 10 turns with a rotation speed of 1 rpm.

To compare the processing effect on the mechanical properties, strips were also subjected to the HPS process at room temperature under an applied pressure of 2 GPa, using an HPS machine developed recently with the maximum capacity of 500 ton.⁵⁵⁾ The HPS machine was operated with a sliding speed of 1 mm/sec for sliding distances of 5, 10 and 15 mm. The sliding direction was made perpendicular to the original rolling direction.

To investigate the effect of SPD on the mechanical properties compared to the as-received state, the SPD-processed samples were subjected to solution treatment (S.T.) at 550°C for 30 min followed by quenching in water.

After the disk was polished with a #2000 abrasive paper, Vickers microhardness was measured from the disk center to edge with 1.0 mm apart each other using a microhardness tester (Mitutoyo HM-102) with an applied load of 4.9 N for a dwell time of 15 s.

The equivalent strain (ε_eq) introduced by HPT processing was estimated using the following equation:   

\begin{equation} \varepsilon_{\textit{eq}} = \frac{2\pi rN}{\sqrt{3}t} \end{equation} (1)

where, r is the distance from the disk center, N is the number of revolution and t is the thickness of the disk after HPT processing. The tensile specimens were extracted as indicated in Fig. 2(a) with a parallel gauge section in 1.5 mm length, 0.65 mm width and 0.7 mm thickness for HPT-processed samples, and in Fig. 2(b) with a parallel gauge section in 18.0 mm length, 6.0 mm width and 0.7 mm thickness (following JIS14B standard) for HPS-processed samples using EDM. Tensile tests were conducted with an initial strain rate of 2.0 × 10⁻³ s⁻¹ at room temperature.

Fig. 2

(a) Positions for Vickers microhardness measurements, tensile test, and SEM and TEM observations in HPT-processed sample and (b) tensile specimen with dimensions extracted from HPS-processed sample.

X-ray diffraction (XRD) analysis was conducted using monochromatic X-rays at the BL04B1 beamline of SPring-8 in JASRI. Microstructural observations were performed by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). SEM was conducted at an accelerating voltage of 5 kV for backscattered electron imaging and of 15 kV for electron backscatter diffraction (EBSD) analysis using a JEOL JSM-7900F. For TEM thin specimens were prepared by mechanical thinning and Ar ion milling. TEM was conducted at 200 kV using a HITACHI HF-2200.

3. Results and Discussion

3.1 Microstructural observations

Figures 3 and 4 show micrographs taken from the 6022 alloy and 6022-2.0Fe alloy, respectively, in (a) the cold-rolled state (②) and (b), (c) the states processed by HPT plus subsequent S.T. (⑨). Here, in each of Figs. 3 and 4, SEM was used for (a) and (b) so that areas of bright contrasts contain Fe, while STEM was for (c) where dark contrasts correspond to Fe-enriched regions as it was operated in a bright-field mode. More fractions of bright contrast areas (Fe-enriched regions) are visible in the 6022-2.0Fe alloy in Fig. 4 than in the 6022 alloy in Fig. 3. It is apparent that the Fe-enriched regions are finely and homogeneously distributed in Fig. 4(b) with an average size of ∼2 µm by the HPT processing (⑧) when this is compared with the distribution in Fig. 4(a) after cold rolling (⑩) where the average size is estimated to be ∼5 µm. It is confirmed that the HPT process is effective for the fragmentation of Fe intermetallics. The STEM images in Fig. 3(c) and Fig. 4(c) reveal that the grain size is smaller in the 6022-2.0Fe alloy than in the 6022 alloy and this is due to the presence of fine particles as indicated by arrows in Fig. 4(c) which consequently inhibited grain growth. The features of fragmented Fe intermetallic particles are similar to those reported earlier on Al–Fe alloys after processing by HPT.^62–64)

Fig. 3

SEM micrographs of 6022 alloy in (a) cold-rolled state, and (b) HPT-processed state plus subsequent S.T. (c) STEM image in same state for (b).

Fig. 4

SEM micrographs of 6022-2.0Fe alloy in (a) cold-rolled state, and (b) HPT-processed state plus subsequent S.T. (c) STEM image in same state for (b).

Close observation of the fragmented particles using XRD analysis confirmed the presence of a β–Al₅FeSi phase in both the 6022 and 6022-2.0Fe alloys, but the α–Al₈Fe₂Si phase was also observed in the 6022-0.2Fe alloy as shown in Fig. 5. Here, Fig. 5(a) is a micrograph containing an α–Al₈Fe₂Si particle and Fig. 5(b) a high-resolution lattice image of the particle with a diffractogram obtained by fast Fourier Transform (FFT) analysis from the dashed area. Furthermore, the presence of the α–Al₈Fe₂Si phase is evidenced from an XRD profile given in Fig. 5(c) with enlargement of the region from 2θ = 3.5 to 4.5 degrees in Fig. 5(d). The presence of the α–Al₈Fe₂Si phase in the 6022-2.0Fe alloy is reasonable because β–Al₅FeSi is replaced by α–Al₈Fe₂Si with increasing Fe content as reported earlier⁶⁵⁾ and this is consistent with the phase diagram reported by Kuijpers et al.⁶⁶⁾

Fig. 5

(a) TEM bright-field image and (b) high-resolution lattice image with diffractogram of α–Al₈Fe₂Si in 6022-2.0Fe alloy processed by HPT and subsequent S.T. (c) X-ray profiles of 6022-2.0Fe alloy processed by HPT and subsequent S.T. (d) enlargement of region from 3.5 to 4.5 degrees in (c).

Figure 6 shows results of EBSD analysis for (a) the 6022 alloy and (b) the 6022-2.0Fe alloy after processing by HPS plus S.T. (⑨). The distribution of grain size is also shown in Fig. 6(c) and (d) for the corresponding alloys. It should be noted that the grain boundaries in the images in Figs. 6(a) and (b) were decided based on the conventional criteria where the difference in adjacent pixels was more than 15 degrees. It is apparent that the grain size is smaller in the 6022-2.0Fe alloy than the 6022 alloy. The average grain size is ∼15 µm in the 6022-2.0Fe alloy, while it is ∼25 µm in the 6022 alloy. This difference should be attributed to a pinning effect of fragmented Fe intermetallics on grain boundaries as indicated by the arrows in Fig. 4(c). It is suggested that the fine dispersion of Fe intermetallic particles play an important role for the thermal stability of the fine grain structure.¹⁹⁾

Fig. 6

Crystal orientation images by EBSD for (a) 6022 alloy and (b) 6022-2.0Fe alloy after processing by HPS plus subsequent S.T. Samples before HPS processing are in cold-rolled state. (c), (d) Grain size distributions in corresponding images.

3.2 Effect of Fe contents on mechanical properties

Figures 7 and 8 plot the Vickers microhardness as a function of the radial distance from the disk center and the equivalent strain calculated through eq. (1), respectively, where (a) is for the 6022 alloy and (b) the 6022-2.0Fe alloy. The results are summarized as follows.

Fig. 7

Plots of Vickers microhardness against distance from disk center for (a) 6022 alloy and (b) 6022-2.0Fe alloy after processing by HPT for 1 and 10 turns, including those after cold rolling and after S.T.

Fig. 8

Plots of Vickers microhardness against equivalent strain for (a) 6022 alloy and (b) 6022-2.0Fe alloy after processing by HPT for 1 and 10 turns, including those after cold rolling and after S.T.

First, the hardness level significantly increases for all the samples processed by HPT (④) when compared with the cold-rolled samples (②). The hardness level is higher in the sample after processing for 10 turns than 1 turn. While it is almost constant irrespective of the radial distance after the 10 turns, it is lowered when the position is closer to the center after the 1 turn. This is understood because the strain is less introduced as the position is closer to the disk center and the rotation number is smaller following eq. (1).^67–72)

Second, when all hardness values are plotted against the equivalent strain in Fig. 8, it is apparent that the hardness gradually increases with the strain and reaches saturations around the strains of 50. The level of the hardness saturation is 115 ± 2 HV and 125 ± 2 HV for the 6022 and 6022-2.0Fe alloys, respectively.

Third, Figs. 7 and 8 also include the hardness values measured after S.T. at 550°C for 30 min for all the processed samples (⑧) including cold rolling (⑩). The hardness levels are considerably lowered. Further comments are added below with reference to Fig. 9.

Fig. 9

Effect of Fe contents on (a) Vickers microhardness and (b) increment of microhardness above cold-rolled state.

Figure 9(a) provides summary of the hardness measurements in all samples including the 6022-0.5Fe and 6022-1.0Fe alloys. The hardness level at the cold-rolled state (②) appears to take almost the same as 80 ± 2 HV except for the 2.0%Fe addition where the hardness is slightly higher as 82 ± 2 HV. The HPT processing (④) thus leads to a significant increase in the hardness in comparison with the cold rolling (②). The increase is more prominent for the 10 turns than 1 turn and it appears that the hardness increases with the increasing addition of the Fe content.

The importance of the HPT processing is further demonstrated by inspecting the hardness levels after the S.T. in Fig. 9(a). Although the hardness level decreases by S.T. (⑧, ⑩), it is still higher in the samples after being processed by HPT (⑧) than the cold rolling (⑩). Figure 9(b) plots ΔHV_N1 and ΔHV_N10 which are defined as the hardness increment above the hardness level of the cold-rolled plus S.T. state (⑩) for the samples processed through 1 and 10 turns by HPT plus S.T. (⑧), respectively. Both ΔHV_N1 and ΔHV_N10 increases with increasing the Fe contents, suggesting that the HPT process is effective to increase the hardness and this is considered to be due to fine fragmentation of Fe intermetallics. The microstructural observations verifies this as shown in section 3.1. Because ΔHV_N1 and ΔHV_N10 exhibit similar behavior in terms of not only the dependence of Fe contents but also their magnitudes, it is suggested that the HPT processing through even 1 rotation is sufficient to the fragmentation of Fe intermetallics.

3.3 Effect of initial states on mechanical properties

Figure 10 delineates nominal stress-strain curves of the 6022-2.0Fe alloy after processing by HPT for (a) the cold-rolled state (②) and (b) the as-cast state (①). For both initial states, the yield strength (σ_y) and the ultimate tensile strength (σ_uts) significantly increase while the elongation to failure (E_f) decreases by the HPT process (③, ④). Strengthening is better achieved for the as-cast state (③) than the cold-rolled state (④) despite the fact that the strength is markedly lowed in the as-cast state (①).

Fig. 10

Nominal stress-nominal strain curves after tensile testing at initial strain rate of 2 × 10⁻³ s⁻¹ in 6022-2.0Fe alloy processed by HPT. Samples prior to HPT processing are in (a) cold-rolled state and (b) as-cast state. Solid lines represent as-received states ((a) cold-rolled and (b) as-cast states), while dashed lines represent states subjected to S.T. after HPT processing and after casting.

Figures 11 and 12 show comparisons between the sample states in terms of (a) σ_y, (b) the increment of the yield stress above the as-received states (cold-rolled (②) or as-cast (①)) and (c) E_f. As found from Figs. 11(a) and 12(a), the following three points should be noted. (1) The strength significantly increases by the HPT process (③, ④) in comparison with the as-received states (cold-rolled (②) or as-cast (①)), and this increase is far remarkable for the as-cast state (③) as shown in Fig. 12(a). (2) The strengthening by the HPT process is almost the same for up to the Fe additions of 1.0% but is appreciably higher for the 2.0%Fe addition in the as-cold-rolled state (④). (3) However, the strength for the 2.0%Fe addition decreases in the as-cast state (③). Further inspection of Figs. 11 and 9 shows that the dependence of σ_y as well as Δσ_N1 and Δσ_N10 on the Fe content is quite similar to the one for the hardness in Fig. 9(a) and (b). Here, Δσ_N1 and Δσ_N10 represent the difference between the values of σ_y in the cold-rolled states plus S.T. (⑩) and the HPT-processed states through N = 1 and N = 10 plus S.T. (⑧).

Fig. 11

Effect of Fe contents on (a) yield stress, (b) increment of yield stress above cold-rolled states and (c) elongation to failure.

Fig. 12

Effect of Fe contents on (a) yield stress, (b) increment of yield stress above as-cast states and (c) elongation to failure.

Close comparison between Fig. 11(b) and Fig. 12(b) leads to the following three points. (1) Both Δσ_N1 and Δσ_N10 are higher in the HPT-processed plus S.T. state (⑧) than the cold-rolled plus S.T. state (⑩) and this suggests that the HPT processing is more effective than the cold rolling for fine fragmentation of Fe intermetallics and thus for producing higher strength even after S.T. (2) There is no essential difference between Δσ_N1 and Δσ_N10 as the hardness results shown in Fig. 9(b) when the HPT process is applied for the alloys in the cold-rolled state (⑧), but there is appreciable difference between Δσ_N1 and Δσ_N10 especially for the Fe contents of 1.0% and 2.0% in the as-cast state (⑦). This suggest that the HPT processing promotes fragmentation of Fe intermetallics for the as-cast state (⑦). (3) Both Δσ_N1 and Δσ_N10 saturate for more than the Fe content of 1% in the cold-rolled state (⑧) but they keep increasing with the Fe addition, while the increase in Δσ_N10 is more prominent than the one in Δσ_N1, in the as-cast state (⑦).

In Fig. 6, the EBSD analysis have shown that there is a clear difference in grain size between the 6022 and 6022-2.oFe alloys in the states after HPS plus S.T. (⑨): ∼25 µm in the 6022 alloy and ∼15 µm in the 6022-2.0Fe alloy. Thus, it turns out that the strengthening due to grain boundaries is estimated to be 18 MPa and 24 MPa, respectively, using the following Hall-Petch relationship:^73,74)   

\begin{equation} \sigma_{GB} = \frac{k}{\sqrt{d}} \end{equation} (2)

where k is the Hall-Petch coefficient and d is the average grain size. k = 0.09 was used for age-hardenable Al alloys as reported earlier.⁷⁵⁾ However, these values are much less than the yield strengths, 131 MPa and 140 MPa, for the 6022 and 6022-0.2Fe alloys (Fig. 14), respectively. Therefore, the contribution of grain boundaries to the strengthening is minor in both alloys and the strength is dominated by other factors arising from the presence of intermetallic particles. Meanwhile, difference of the yield stress between the two alloys is 9 MPa and this is fairly comparable to the difference between the stresses estimated from the Hall-Petch relation, which is 6 MPa. It is considered that the inclusion of Fe is responsible for the difference through the pinning effect of Fe-rich particles.

Concerning E_f shown in Fig. 11(c) and Fig. 12(c), the following three points should be worth noting. (1) The value of E_f tends to be larger when the HPT process is applied for the samples in the cold-rolled state (②) than the as-cast state (①). (2) The value of E_f decreases with increasing the Fe content and this decrease is more prominent for the 2.0%Fe in the as-cast state (①). (3) The HPT process is effective to improve E_f in the as-cast state for the 2%Fe (②), when compared to the E_f in the state without processing (⑥).

3.4 Effect of HPS processing on the mechanical properties

Figure 13 shows stress-strain curves of the 6022-2.0Fe alloy after processing by HPS (⑤) and after processing by HPS plus S.T. (⑨) for the cold-rolled state (②). The strength increases with increasing the sliding distance: while the sample in the cold-rolled state (②) exhibits σ_y = 265 MPa and σ_uts = 270 MPa, the HPS processing for the sliding distance of 15 mm (⑤) gives rise to σ_y = 390 MPa and σ_uts = 410 MPa, respectively. Figure 14 shows summary of the tensile testing of the samples after the HPS processing (⑤) and the subsequent S.T. (⑨). For comparison, Fig. 14 includes the results after the HPT processing (④) and the cold rolling (②) together with the subsequent S.T. (⑧, ⑨, ⑩). The HPS processing exhibits σ_y well consistent with the ones obtained by the HPT processing for N = 1 (④), but E_f is appreciably less. This is also the case for E_f even after the S.T. (⑧, ⑨). It is reasonable to consider that the difference in E_f is attributed to the difference in the gauge length of the tensile specimen, because the total elongation is affected by the gauge length as discussed in an earlier report.⁷⁶⁾

Fig. 13

Nominal stress-nominal strain curves after tensile testing at initial strain rate of 2 × 10⁻³ s⁻¹ in 6022-2.0Fe alloy processed by HPS. Samples prior to HPS processing are in cold-rolled state. Solid lines represent cold-rolled states, while dashed lines represent states subjected to S.T. after HPS processing and after cold rolling.

Fig. 14

Yield stress and elongation to failure plotted against equivalent strain in 6022-2.0Fe alloy after processing by HPS and HPT and after subsequent S.T. Samples before processing by HPS and HPT are cold-rolled state.

It should be noted that the σ_uts of 410 MPa after sliding of X = 15 mm is in agreement with the value of σ_uts = 415 MPa attained by HPT processing through N = 1. Since the equivalent strain (ε_eq) introduced by HPS processing can be estimated using the following equation:⁵¹⁾   

\begin{equation} \varepsilon_{\textit{eq}} = \frac{x}{\sqrt{3}t} \end{equation} (3)

where x is the sliding distance and t is the sample thickness after HPS processing. For X = 15 mm, the ε_eq is calculated to be 8.7 and this is well similar to 9.1 imparted by the HPT process for N = 1 when ε_eq is estimated using eq. (1) with r = 2 mm and t = 0.8 mm where the tensile specimen was extracted and thickness measured after the processing. It is thus concluded that the HPS process can produce the tensile strength essentially the same as the HPT process.

Close examination of the stress-strain curves after S.T. (⑨) reveals the following three points. (1) There is no systematic difference in the flow stress levels but σ_uts lies in the range of 210 to 220 MPa and σ_y = ∼150 MPa. (2) E_f increases with increasing the sliding distance. The largest is achieved after processing for the 15 mm sliding (⑨) and the smallest is after the cold rolling (⑩). It is thus suggested that the Fe intermetallics are fragmented to a uniform distribution of smaller sizes, which can be more favorable to maintain better elongation to failure.

4. Summary and Conclusions

1. (1)    Al–Mg–Si alloys based on the compositions of A6022 and extra additions of 0.5%Fe, 1.0%Fe and 2.0%Fe were processed using HPT and HPS at room temperature under a pressure of 2 GPa.
2. (2)    Processing by both HPT and HPS led to a dramatic increase in the tensile strength (σ_uts), reaching 415 MPa and 410 MPa, respectively, and thus both processes yield almost the same σ_uts. The elongation to failure (E_f) was larger in the HPT process than the HPS process and this was because the shorter gauge length was adopted for the HPT process due to the limitation of the sample size.
3. (3)    Fe intermetallics are finely fragmented and homogenously dispersed in the Al matrix by processing through HPT and HPS in comparison with cold rolling. Consequently, the hardness, σ_uts, σ_y and E_f were improved by such a modification of microstructures.
4. (4)    Grain growth during solution treatment (S.T.) was inhibited by the formation of fine and homogeneous distributions of Fe intermetallic particles after processing by HPT and HPS.

Acknowledgements

This work was conducted under a project subsidized by the New Energy and Industrial Technology Development Organization (NEDO) and in part by a Grant-in-Aid for Scientific Research (A) from the MEXT, Japan (J19H00830).

REFERENCES

• 1)   X.  Wu,  H.  Zhang,  Z.  Ma,  T.  Tao,  J.  Gui,  W.  Song,  B.  Yang and  H.  Zhang: J. Alloy. Compd. 786 (2019) 205–214. doi:10.1016/j.jallcom.2019.01.352
• 2)   G.  Eisaabadi B.,  P.  Davami,  N.  Varahram and  S.K.  Kim: Mater. Sci. Eng. A 565 (2013) 278–284. doi:10.1016/j.msea.2012.12.022
• 3)   Y.  Wang,  Y.  Deng,  Q.  Dai,  K.  Jiang,  J.  Chen and  X.  Guo: Mater. Sci. Eng. A 803 (2021) 140477. doi:10.1016/j.msea.2020.140477
• 4)  N.A. Belov, A.A. Aksenov and D.G. Eskin: Iron in Aluminum Alloys: Impurity and Alloying Element, (Taylor and Francis, London, 2002) pp. 1–7.
• 5)   A.  Lendvai: J. Mater. Sci. Lett. 5 (1986) 1219–1220. doi:10.1007/BF01729370
• 6)  P. Villars: Pearson’s Handbook of Crystallographic Data for Intermetallic Phases, (Materials Park, OH, ASM International, 1997) pp. 371–373.
• 7)   P.J.  Black: Acta Crystallogr. 8 (1955) 43–48. doi:10.1107/S0365110X5500011X
• 8)   P.J.  Black: Acta Crystallogr. 8 (1955) 175–182. doi:10.1107/S0365110X55000637
• 9)   R.M.K.  Young and  T.W.  Clyne: Scr. Metall. 15 (1981) 1211–1216. doi:10.1016/0036-9748(81)90301-X
• 10)   P.  Skjerpe: Acta Crystallogr. Sec. B 44 (1988) 480–486. doi:10.1107/S0108768188005245
• 11)   R.Z.  Valiev,  R.K.  Islamgaliev and  I.V.  Alexandrov: Prog. Mater. Sci. 45 (2000) 103–189. doi:10.1016/S0079-6425(99)00007-9
• 12)   R.  Valiev: Nat. Mater. 3 (2004) 511–516. doi:10.1038/nmat1180
• 13)   R.Z.  Valiev,  Y.  Estrin,  Z.  Horita,  T.G.  Langdon,  M.J.  Zehetbauer and  Y.T.  Zhu: JOM 58 (2006) 33–39. doi:10.1007/s11837-006-0213-7
• 14)   Y.  Estrin and  A.  Vinogradov: Acta Mater. 61 (2013) 782–817. doi:10.1016/j.actamat.2012.10.038
• 15)   T.G.  Langdon: Acta Mater. 61 (2013) 7035–7059. doi:10.1016/j.actamat.2013.08.018
• 16)   K.  Edalati and  Z.  Horita: Mater. Sci. Eng. A 652 (2016) 325–352. doi:10.1016/j.msea.2015.11.074
• 17)   C.  Phongphisutthinan,  H.  Tezuka,  T.  Sato,  S.  Takamori and  Y.  Ohsawa: Mater. Trans. 53 (2012) 885–892. doi:10.2320/matertrans.MBW201121
• 18)   D.  Kim,  J.  Kim,  S.  Wenner,  E.  Thronsen,  C.D.  Marioara,  R.  Holmestad and  E.  Kobayashi: Mater. Charact. 173 (2021) 110863. doi:10.1016/j.matchar.2020.110863
• 19)   A.  Duchaussoy,  X.  Sauvage,  K.  Edalati,  Z.  Horita,  G.  Renou,  A.  Deschamps and  F.D.  Geuser: Acta Mater. 167 (2019) 89–102. doi:10.1016/j.actamat.2019.01.027
• 20)   J.M.  Cubero-Sesin and  Z.  Horita: Mater. Sci. Eng. A 558 (2012) 462–471. doi:10.1016/j.msea.2012.08.029
• 21)   P.W.  Bridgman: Phys. Rev. 48 (1935) 825–847. doi:10.1103/PhysRev.48.825
• 22)   A.P.  Zhilyaev and  T.G.  Langdon: Prog. Mater. Sci. 53 (2008) 893–979. doi:10.1016/j.pmatsci.2008.03.002
• 23)   Z.  Horita,  Y.  Tang,  T.  Masuda and  Y.  Takizawa: Mater. Trans. 61 (2020) 1177–1190. doi:10.2320/matertrans.MT-M2020074
• 24)   Y.  Tang,  W.  Goto,  S.  Hirosawa,  Z.  Horita,  S.  Lee,  K.  Matsuda and  D.  Terada: Acta Mater. 131 (2017) 57–64. doi:10.1016/j.actamat.2017.04.002
• 25)   A.P.  Zhilyaev,  G.  Nurislamova,  B.K.  Kim,  M.  Baró,  J.  Szpunar and  T.  Langdon: Acta Mater. 51 (2003) 753–765. doi:10.1016/S1359-6454(02)00466-4
• 26)   A.P.  Zhilyaev,  T.  McNelley and  T.  Langdon: J. Mater. Sci. 42 (2007) 1517–1528. doi:10.1007/s10853-006-0628-0
• 27)   Q.  Wei,  H.T.  Zhang,  B.E.  Schuster,  K.T.  Ramesh,  R.Z.  Valiev,  L.J.  Kecskes,  R.J.  Dowding,  L.  Magness and  K.  Cho: Acta Mater. 54 (2006) 4079–4089. doi:10.1016/j.actamat.2006.05.005
• 28)   B.J.  Bonarski,  E.  Schafler,  B.  Milkulowskei and  M.J.  Zehetbauer: Mater. Sci. Forum 584–586 (2008) 263–268. doi:10.4028/www.scientific.net/MSF.584-586.263
• 29)   S.W.  Lee,  K.  Edalati and  Z.  Horita: Mater. Trans. 51 (2010) 1072–1079. doi:10.2320/matertrans.M2009375
• 30)   S.W.  Lee and  Z.  Horita: Mater. Trans. 53 (2012) 38–45. doi:10.2320/matertrans.MD201131
• 31)   Y.  Takizawa,  K.  Otsuka,  T.  Masuda,  T.  Kajita,  M.  Yumoto,  Y.  Otagiri and  Z.  Horita: Mater. Sci. Eng. A 648 (2015) 178–182. doi:10.1016/j.msea.2015.09.023
• 32)   J.  Xu,  X.  Wang,  M.  Shirooyeh,  G.  Xing,  D.  Shan,  B.  Guo and  T.G.  Langdon: J. Mater. Sci. 50 (2015) 7424–7436. doi:10.1007/s10853-015-9300-x
• 33)   M.  Mito,  H.  Matsui,  K.  Tsuruta,  T.  Yamaguchi,  K.  Nakamura,  H.  Deguchi,  N.  Shirakawa,  H.  Adachi,  T.  Yamasaki,  H.  Iwaoka,  Y.  Ikoma and  Z.  Horita: Sci. Rep. 6 (2016) 36337–36344. doi:10.1038/srep36337
• 34)   Á.  Révész and  Z.  Kovács: Mater. Trans. 60 (2019) 1283–1293. doi:10.2320/matertrans.MF201917
• 35)   A.V.  Korznikov,  G.  Tram,  O.  Dimitrov,  G.F.  Korznikova,  S.R.  Idrisova and  Z.  Pakiela: Acta Mater. 49 (2001) 663–671. doi:10.1016/S1359-6454(00)00345-1
• 36)   K.  Edalati,  J.  Matsuda,  H.  Iwaoka,  S.  Toh,  E.  Akiba and  Z.  Horita: Int. J. Hydrogen Energ. 38 (2013) 4622–4627. doi:10.1016/j.ijhydene.2013.01.185
• 37)   K.  Edalati,  E.  Akiba and  Z.  Horita: Sci. Technol. Adv. Mater. 19 (2018) 185–193. doi:10.1080/14686996.2018.1435131
• 38)   P.W.  Bridgman and  I.  Simon: J. Appl. Phys. 24 (1953) 405–413. doi:10.1063/1.1721294
• 39)   C.H.  Bates,  W.B.  White and  R.  Roy: Science 137 (1962) 993. doi:10.1126/science.137.3534.993.a
• 40)   R.O.  Miller,  F.F.  Dachille and  R.  Roy: J. Appl. Phys. 37 (1966) 4913–4918. doi:10.1063/1.1708164
• 41)   K.  Edalati and  Z.  Horita: Scr. Mater. 63 (2010) 174–177. doi:10.1016/j.scriptamat.2010.03.048
• 42)   V.I.  Levitas,  Y.  Ma,  E.  Selvi,  J.  Wu and  J.A.  Patten: Phys. Rev. B 85 (2012) 054114. doi:10.1103/PhysRevB.85.054114
• 43)   K.  Edalati,  M.  Arimura,  Y.  Ikoma,  T.  Daio,  M.  Miyata,  D.J.  Smith and  Z.  Horita: Mater. Res. Lett. 3 (2015) 216–221. doi:10.1080/21663831.2015.1065454
• 44)   H.  Razavi-Khosroshahi and  M.  Fuji: Mater. Trans. 60 (2019) 1203–1208. doi:10.2320/matertrans.MF201916
• 45)   J.  Hidalgo-Jimenez,  Q.  Wang,  K.  Edalati,  J.M.  Cubero-Sesín,  H.  Razavi-Khosroshahi,  Y.  Ikoma,  D.  Gutiérrez-Fallas,  F.A.  Dittel-Meza,  J.C.  Rodríguez-Rufino,  M.  Fuji and  Z.  Horita: Int. J. Plast. 124 (2020) 170–185. doi:10.1016/j.ijplas.2019.08.010
• 46)   Y.  Ikoma,  K.  Hayano,  K.  Edalati,  K.  Saito,  Q.  Guo and  Z.  Horita: Appl. Phys. Lett. 101 (2012) 121908. doi:10.1063/1.4754574
• 47)   Y.  Ikoma,  K.  Kumano,  K.  Edalati,  M.R.  McCartney,  D.J.  Smith and  Z.  Horita: Mater. Charact. 132 (2017) 132–138. doi:10.1016/j.matchar.2017.08.011
• 48)   Y.  Ikoma: Mater. Trans. 60 (2019) 1168–1176. doi:10.2320/matertrans.MF201907
• 49)   Y.  Ikoma,  T.  Yamasaki,  T.  Shimizu,  M.  Takaira,  M.  Kohno,  Q.  Guo,  M.R.  McCartney,  D.J.  Smith,  Y.  Arai and  Z.  Horita: Mater. Charact. 169 (2020) 110590. doi:10.1016/j.matchar.2020.110590
• 50)   G.  Faraji and  H.  Torabzadeh: Mater. Trans. 60 (2019) 1316–1330. doi:10.2320/matertrans.MF201905
• 51)   T.  Fujioka and  Z.  Horita: Mater. Trans. 50 (2009) 930–933. doi:10.2320/matertrans.MRP2008445
• 52)   K.  Tazoe,  S.  Honda and  Z.  Horita: Mater. Sci. Forum 667–669 (2011) 91–96. doi:10.4028/www.scientific.net/MSF.667-669.91
• 53)   K.  Tazoe and  Z.  Horita: J. JILM 62 (2012) 454–458. doi:10.2464/jilm.62.454
• 54)   S.  Lee,  K.  Tazoe,  I.F.  Mohamed and  Z.  Horita: Mater. Sci. Eng. A 628 (2015) 56–61. doi:10.1016/j.msea.2015.01.026
• 55)   Y.  Takizawa,  T.  Masuda,  K.  Fujimitsu,  T.  Kajita,  K.  Watanabe,  M.  Yumoto,  Y.  Otagiri and  Z.  Horita: Metall. Mater. Trans. A 47 (2016) 4669–4681. doi:10.1007/s11661-016-3623-3
• 56)   Y.  Takizawa,  T.  Kajita,  P.  Kral,  T.  Masuda,  K.  Watanabe,  M.  Yumoto,  Y.  Otagiri,  V.  Sklenicka and  Z.  Horita: Mater. Sci. Eng. A 682 (2017) 603–612. doi:10.1016/j.msea.2016.11.081
• 57)   Y.  Tang,  T.  Komatsu,  T.  Masuda,  M.  Arita,  Y.  Takizawa,  M.  Yumoto,  Y.  Otagiri and  Z.  Horita: Materialia 14 (2020) 100916. doi:10.1016/j.mtla.2020.100916
• 58)   T.  Masuda,  K.  Fujimitsu,  Y.  Takizawa and  Z.  Horita: Lett. Mater. 5 (2015) 258–263. doi:10.22226/2410-3535-2015-3-258-263
• 59)   Y.  Tang,  K.  Sumikawa,  Y.  Takizawa,  M.  Yumoto,  Y.  Otagiri and  Z.  Horita: Mater. Sci. Eng. A 748 (2019) 108–118. doi:10.1016/j.msea.2019.01.071
• 60)   Y.  Tang,  Y.  Takizawa,  M.  Yumoto,  Y.  Otagiri and  Z.  Horita: Mater. Trans. 61 (2020) 1378–1389. doi:10.2320/matertrans.L-M2020828
• 61)   Y.  Tang,  K.  Matsuda,  Y.  Takizawa,  M.  Yumoto,  Y.  Otagiri and  Z.  Horita: Mater. Sci. Technol. 36 (2020) 877–886. doi:10.1080/02670836.2020.1746538
• 62)   O.N.  Senkov,  F.H.  Froes,  V.V.  Stolyarov,  R.Z.  Valiev and  J.  Liu: Nanostruct. Mater. 10 (1998) 691–698. doi:10.1016/S0965-9773(98)00107-X
• 63)   O.N.  Senkov,  F.H.  Froes,  V.V.  Stolyarov,  R.Z.  Valiev and  J.  Liu: Scr. Mater. 38 (1998) 1511–1516. doi:10.1016/S1359-6462(98)00073-6
• 64)   J.M.  Cubero-Sesin and  Z.  Horita: Mater. Trans. 53 (2012) 46–55. doi:10.2320/matertrans.MD201127
• 65)   L.  Sweet,  S.M.  Zhu,  S.X.  Gao,  J.A.  Taylor and  M.A.  Easton: Metall. Mater. Trans. A 42 (2011) 1737–1749. doi:10.1007/s11661-010-0595-6
• 66)   N.C.W.  Kuijpers,  F.J.  Vermolen,  C.  Vuik,  P.T.G.  Koenis,  K.E.  Nilsen and  S.  van der Zwaag: Mater. Sci. Eng. A 394 (2005) 9–19. doi:10.1016/j.msea.2004.09.073
• 67)   K.  Edalati and  Z.  Horita: Mater. Sci. Forum 667–669 (2011) 683–688. doi:10.4028/www.scientific.net/MSF.667-669.683
• 68)   M.  Kawasaki,  H.J.  Lee,  B.  Ahn,  A.P.  Zhilyaev and  T.G.  Langdon: J. Mater. Res. Tech. 3 (2014) 311–318. doi:10.1016/j.jmrt.2014.06.002
• 69)   A.P.  Zhilyaev,  K.  Oh-ishi,  T.G.  Langdon and  T.R.  McNelley: Mater. Sci. Eng. A 410–411 (2005) 277–280. doi:10.1016/j.msea.2005.08.044
• 70)   Z.  Horita and  T.G.  Langdon: Mater. Sci. Eng. A 410–411 (2005) 422–425. doi:10.1016/j.msea.2005.08.133
• 71)   C.  Xu,  Z.  Horita and  T.G.  Langdon: Acta Mater. 56 (2008) 5168–5176. doi:10.1016/j.actamat.2008.06.036
• 72)   C.  Xu and  T.G.  Langdon: Mater. Sci. Eng. A 503 (2009) 71–74. doi:10.1016/j.msea.2008.04.083
• 73)   E.O.  Hall: Proc. Phys. Soc. B 64 (1951) 747–753. doi:10.1088/0370-1301/64/9/303
• 74)   N.J.  Petch: J. Iron Steel Inst. 174 (1953) 25–28.
• 75)   J.W.  Wyrzykowski and  M.W.  Grabski: Philos. Mag. A 53 (1986) 505–520. doi:10.1080/01418618608242849
• 76)   Y.H.  Zhao,  Y.Z.  Guo,  Q.  Wei,  A.M.  Dangelewicz,  C.  Xu,  Y.T.  Zhu,  T.G.  Langdon,  Y.Z.  Zhou and  E.J.  Laverni: Scr. Mater. 59 (2008) 627–630. doi:10.1016/j.scriptamat.2008.05.031