2019 Volume 60 Issue 11 Pages 2435-2441
Al–Mg alloy is a representative non-heat-treatable aluminum alloy. The main strengthening mechanism of the alloy is solid solution hardening by magnesium. If we can increase Mg solubility much more in the solid solutions by using the excellent cooling capacity of the vertical-type high-speed twin-roll casting, further improvement of strength is expected. In this study, high Mg containing Al–8%Mg, 12%Mg and 23%Mg alloy strips were fabricated by using this method, and increase in Mg solid solubility and mechanical properties were investigated. Cold rolling was performed for Al–8%Mg and Al–12%Mg alloy strips. Although some porosities were observed in the mid-thickness region of the as-cast strips, they were reduced by optimizing the casting conditions. The slightly remaining porosities in the as-cast strip could be eliminated by the subsequent cold rolling. The lattice constant of the α-Al phase was calculated from the diffraction peak position of XRD profile, and by which the Mg solid solubility was estimated. The maximum solid solubility was about 12%. The β-Al3Mg2 phase particles which were observed in the original as-cast Al–12%Mg alloy strip were completely dissolved in to the matrix by the subsequent heat treatment. Compared with the typical commercial Al–Mg alloy, such as 5052 or 5083 alloy, the Al–12%Mg alloy sheets fabricated from the twin-roll cast strip had an excellent strength and ductility balance.
Al–Mg alloy is a one of the representative non-heat-treatable aluminum alloys with excellent workability and corrosion resistance, as well as a good balance between strength and ductility. Therefore, the alloy is widely used for various products including automobile parts. However, due to the wide solid-liquid coexistence temperature range, it is also known that the alloy is difficult to cast. In addition, as increasing Mg content, the amount of brittle intermetallic compounds (β-Al3Mg2 phase) also increases, which deteriorates mechanical properties. For these reasons, Mg content in commercial Al–Mg alloys is limited to about 5 mass% (hereafter abbreviated as %) for wrought alloys, to about 10% even for cast alloys. According to the Al–Mg binary equilibrium phase diagram, the maximum solid solubility of Mg in the Al parent phase is about 18.6 at% at the eutectic temperature of 450°C.1) Since the main strengthening mechanism of this alloy is solid solution strengthening by Mg, the Al–Mg alloys in which high concentration of Mg is solid-solved in the α-Al parent phase have been tried to make by using the various techniques. For example, Jie et al. reported that the solid solubility increases up to about 30 at% by the decrease of the diffusion coefficient of solute Mg in the α-Al parent phase due to the solidification under high pressure (3 GPa).2) Luo et al. achieved the solid solubility of 36.8 at% by quenching a small amount (about 20 mg) of molten metal on a copper plate.3) Mechanical alloying is also effective method, and the solid solubility of about 20 to 40 at% has been reported.4–7) Schoenitz et al. reported that the solid solubility of 20.8 at% was obtained by the mechanical alloying of aluminum and magnesium powders by using the shaker-mill.4) Calka et al. and Scudino et al. also achieved the solid solubility of 40 at% by using the planar-type ball-mill5) and the planetary-ball-mill,6,7) respectively. These reports are limited to the investigation for the small amount of samples. There is no report that Al–Mg alloy strip in which high concentration of Mg is solid-solved is fabricated in practical size.
The vertical-type high-speed twin-roll casting used in this research is a unique method to fabricate aluminum alloy strips directly from the molten metal. Further, it is possible to continuously cast a strip at a high speed of about 30 to 150 m/min. Since a thin strip can be produced directly from the melt, the hot-rolling process can be omitted. The cooling rate during solidification is about 1000°C/s, which is effective for refining the microstructure and to make the impurities elements harmless by rapid quenching.8–10) The excellent cooling ability is effective to achieve the supersaturation of solute atoms in the matrix. If we can increase Mg solubility much more in the solid solutions by this method, further improvement in the strength of Al–Mg alloy strips is expected.
In this study, high Mg containing Al–Mg alloy strips from Al–8%Mg to Al–23%Mg were fabricated by using this method, and increase in Mg solid solubility and mechanical properties of the cold-rolled and annealed Al–12%Mg alloy sheet were investigated.
Materials used in this study were Al–8%Mg, Al–12%Mg, and Al–23%Mg alloys. The chemical compositions of the alloys are listed in Table 1. The liquidus temperature, solidus temperature, and solid-liquid coexistence temperature range of each alloy obtained from the Al–Mg binary equilibrium phase diagram1) (as shown in Fig. 1) are listed in Table 2. Al–8%Mg and Al–23%Mg alloys were used to be revealed the optimal casting parameters (initial spring load and roll rotation speed) and the maximum solid solubility of Mg in the vertical-type high-speed twin-roll casting, respectively. Due to the maximum solid solubility of Mg was about 12%, mechanical properties of the cold-rolled and annealed Al–12%Mg alloy sheet were investigated.
Al–Mg binary phase diagram.
The vertical-type high-speed twin-roll caster used in this study consists of a pair of rotating water-cooled pure copper rolls, a spring for applying load to the strips, nozzles and side-dams for making a melt pool.11) The diameter and the width of pure copper rolls are 300 mm and 100 mm, respectively. One roll is firmly fixed to the pedestal, whereas the other roll is attached to the pedestal with springs, and it is possible to apply the load to the strips. The spring load is used so that a strong contact between the roll surface and the solidification shells can be kept to achieve an excellent heat extraction and high cooling rates. Casting parameters are listed in Table 3. The strip was fabricated at the speed of 30 m/min or 60 m/min. 11 kN or 40 kN load was applied to one of the rolls by springs in advance before the casting. The initial roll gap was 1 mm. The solidification length, the contact length between a tip of the nozzle and the roll gap along the roll surface, was fixed to 100 mm. The melt head was kept at about 100 mm during casting. The alloy placed in the graphite crucible was melted in an electric furnace under an argon gas atmosphere. The argon gas was also blown into the melt to perform a degasification treatment for 20 min. Thereafter, the crucible containing the molten metal was taken out from the furnace and the molten metal was poured into the nozzle at a temperature 20°C higher than the liquidus temperature. Since the melt head height is insufficient at the start and end of the strips fabricated by this method, the strip thickness is thin and the microstructure is not stable. For this reason, the specimen was cut out from the middle part of the strip with a constant thickness, and was provided for subsequent microstructural observation and tensile test. About 2.5 kg molten alloy was prepared for fabricating about a 3∼4 m-long and 100 mm-wide strip. When the melt head was stable at about 100 mm, thickness of the strip was constant at about 2∼4 mm depending on the casting parameters.
The cross-section of the specimens was ground with waterproof abrasive paper sheets up to #2000 and then polished with diamond paste (6, 3, and 1 µm) and colloidal silica suspension. The polished surfaces were etched by Keller’s reagent. The cross-sections (porosities and cracks) and fracture surface after tensile test were observed by an optical microscopy (OM) and a scanning electron microscope (SEM).
2.4 Phase identification and estimation of Mg solid solubilityAn X-ray diffractometer (XRD) was used to identify the composition phases of the strips. The measurement was carried out at room temperature under the parameters of a tube voltage of 40 kV, a tube current of 30 mA, a measurement range of 2θ = 30° to 90°, a step interval of 0.01°, and a scanning speed of 1.2 s/step. The lattice spacing “d” of (h k l) can be expressed by eq. (1) from Bragg’s law. In the case of cubic crystal, the lattice constant “a” can be calculated by eq. (2). The average of the lattice constant was obtained by the diffraction peak position “θ” from each lattice plane of the Al phase in the XRD profile.
| \begin{equation} d = \lambda/2\sin\theta, \end{equation} | (1) |
| \begin{equation} a = d\cdot\surd(h^{2}+k^{2}+l^{2}), \end{equation} | (2) |
The maximum solid solubility of Mg was about 12% in the vertical-type high-speed twin-roll casting. Therefore, mechanical properties of the cold-rolled and annealed Al–12%Mg alloy sheet were investigated. Homogenization (440°C - 6 h) was applied to the as-cast strip of Al–12%Mg alloy, followed by water cooling and cold rolling (reduction of about 70%) to a thickness of 1 mm. Small tensile test specimens (t = 1 mm) as shown in Fig. 2 were taken from the central part of the rolled sheet as the loading direction was parallel to the rolling direction. The specimens were annealed (460°C - 6 h) and water cooled. The tensile test was carried out at room temperature using an Instron type testing machine at a crosshead speed of 1 mm/min (N = 5).
Diagram of tensile test specimen.
The initial spring load and roll rotation speed are the most important parameters in the vertical-type high-speed twin-roll casting. Figure 3 shows the average thickness of steady part in the Al–8%Mg alloy strip fabricated by each parameter. When the roll rotation speed was the same, the strip thickness was almost the same even if the initial spring load was changed. On the other hand, when the initial spring load was the same, the strip thickness became thicker as the roll rotation speed became lower. The influence of roll rotation speed on strip thickness was greater than that of initial spring load. Because the roll rotation speed directly changes the growth time of the solidified shell growing from both roll surfaces.
Relationship between casting parameters and strip thickness.
Figure 4 shows the cross-sectional images of Al–8%Mg alloy strip fabricated under the initial spring load of 11 kN and the roll rotation speed of 60 m/min. On the surface of the Al–Mg alloy strips fabricated by the vertical-type high-speed twin-roll casting, periodic patterns were observed in the casting direction consisting of the shiny region and un-shiny region with cracks.11) The cross-sectional images just below the shiny region and un-shiny region are shown in Fig. 4(a) and (b), respectively. In both cross-sections, dark contrast was observed in the central part of the strip thickness. The magnified SEM images of each area are shown in Fig. 4(c) and (d), respectively. The dark contrast was found to be porosities, and the amount of porosities was larger in the un-shiny region than that of in the shiny-region. This is because the local cooling rate under the un-shiny region was lower than that of the shiny region.12) Such porosities are considered to be a shrinkage defect caused by insufficient supply of molten metal to fill for the volume reduction due to the solidification of the residual liquid phase at the central part of the strip thickness, which was the final solidified part. In order to fabricate the sound strips, it is necessary to suppress such defects. In the vertical-type high-speed twin-roll casting, the initial spring load and roll rotation speed are greatly influenced for suppression of such defects.13) Al–8%Mg alloy strips were fabricated by changing these two casting parameters. The cross-sectional images of these strips were compared with the images shown in Fig. 4 (initial spring load of 11 kN, roll rotation speed 60 m/min). Figures 5(a) and (b) are cross-sectional images just below the shiny-region and the un-shiny region of the strip fabricated under the initial spring load of 40 kN and the roll rotation speed of 60 m/min. Although the porosities in the central part were reduced just below the shiny-region, a large internal crack was observed just below the un-shiny region. The internal crack is thought to be caused by the solidification shrinkage of the residual liquid film formed between the central band region consisting of fine globular grains and the solidified shell at the central part of the strip thickness.14) Figures 5(c) and (d) are cross-sectional images of the strip fabricated under the initial spring load of 40 kN and the roll rotation speed of 30 m/min. No porosity or internal crack was observed in the central part in both the shiny-region and the un-shiny region. However, the surface cracks, that existed only in the un-shiny region, were also observed in the shiny region. In this casting condition, the initial spring load was increased to 40 kN. Furthermore, in the case of the roll rotation speed is 30 m/min, the strip thickness increases as shown in Fig. 3. Therefore, it is considered that the cracking on the surface occurred due to the larger applied load which was generated when the strip passed through and expanded the initial roll gap (1 mm). Thus, in the case of Al–Mg alloy, the increase in the initial spring load is not appropriate. Figures 5(e) and (f) are cross-sectional images of the strip fabricated under the initial spring load of 11 kN and the roll rotation speed of 30 m/min. Although the porosities were not completely eliminated, it was better than the strip microstructures shown in Fig. 4, and no internal crack was observed. Although the surface cracks remain even after cold rolling, the slight porosities may be eliminated by the cold rolling. Therefore, in this study, an initial spring load of 11 kN and a roll rotation speed of 30 m/min were concluded to be appropriate casting parameters and used for fabrication of strips.
Cross-sectional images of Al–8%Mg alloy strip (Initial spring load: 11 kN, Roll rotation speed: 60 m/min). (a) below shiny region, (b) below un-shiny region, (c), (d) magnified SEM images of (a), (b).
Cross-sectional images of Al–8%Mg alloy strips (various initial spring load and roll rotation speed). (a), (b) 40 kN, 60 m/min, (c), (d) 40 kN, 30 m/min, (e), (f) 11 kN, 30 m/min.
In order to investigate the maximum solid solubility of Mg by rapid solidification of vertical-type high-speed twin-roll casting, Al–23%Mg alloy strip was tried to cast. As a result, the strip was obtained even for such a high Mg content alloy. The XRD profile of the strip surface is shown in Fig. 6(a). In addition to the α-Al parent phase, many peaks of the β-Al3Mg2 phase were observed. Since the β phase is a brittle intermetallic compound, mechanical properties of the strip are deteriorated. Therefore, it is necessary to investigate the composition limit that the alloy does not crystallize the β phase under the cooling rate of the vertical-type high-speed twin-roll casting. Here, in order to estimate the Mg solid solubility, a linear relationship between the Mg solid solubility in the Al–Mg alloy and the lattice constant of the α-Al parent phase was used.3) The lattice constant was obtained from the peak position of the α-Al parent phase in the XRD profile shown in Fig. 6(a), and the estimated Mg solid solubility was about 12.0%. The maximum solid solubility of Mg was found to be about 12.0% in the vertical-type high-speed twin-roll casting. Figures 6(b) and (c) show the XRD profiles of Al–12%Mg and Al–8%Mg alloy strip surfaces. Although the peaks of β phase were observed in the Al–12%Mg alloy strip, the peak strength was very small, and the estimated Mg solid solubility was about 10.4%. No peak of β phase was observed in the Al–8%Mg alloy strip. The estimated Mg solid solubility was about 7.9%, indicating that all the contained Mg was completely solid-solved.
XRD profiles of (a) Al–23%Mg, (b) Al–12%Mg, and (c) Al–8%Mg alloy strip surface (as-cast).
Figure 7 shows the cross-sectional images of the as-cast material, as-rolled material, and as-annealed material of Al–12%Mg alloy strips. Although porosities were observed in the central part of the plate thickness in both the shiny region and the un-shiny region of the as-cast material ((a), (b)), such slight porosities were almost eliminated after cold rolling ((c)–(f)). The XRD profiles of the Al–12%Mg alloy strip surface (as-cast, as-rolled, as-annealed) are shown in Fig. 8. The peaks of the β phase found in the as-cast material were not observed in the as-rolled and the as-annealed material. The peaks of the α-Al parent phase in the as-rolled and the as-annealed material were shifted to the lower angle, and the estimated Mg solid solubility was about 12.0%. It is considered that the β phase slightly crystallized in the as-cast material was completely solid-solved into the parent phase by homogenization heat treatment (440°C - 6 h) and subsequent water cooling.
Cross-sectional images of Al–12%Mg alloy strips. (a), (b) as-cast, (c), (d) as-rolled, (e), (f) as-annealed.
XRD profiles of Al–12%Mg alloy strip surface (as-cast, as-rolled, as-annealed).
As mentioned above, the periodic patterns with cracks were formed on the surface of the Al–Mg alloy strips. It was possible to reduce the porosities in the central part of the plate thickness by optimizing the casting conditions. However, it was difficult to suppress the surface defects. Optimizing of the nozzle shape and pouring system will be required to reduce the surface defects.
The mechanical properties of cold-rolled and annealed Al–12%Mg alloy sheet were evaluated. Cold rolling (reduction rate was about 70%.) was possible for the strip. However, several surface cracks were observed in the cold-rolled sheet. Small specimens with no-surface cracks were taken from the annealed sheet and provided for the tensile test. Average values of the yield strength, the ultimate tensile strength, and the elongation were 205 ± 1.6 MPa, 448 ± 2.7 MPa, 29.0 ± 4.6%, respectively. The sheet with supersaturated 12% Mg exhibited excellent strength due to the solid solution strengthening. The fracture surface exhibited dimple morphology as shown in Fig. 9. It is known that both strength and ductility of Al–Mg alloy increases as increasing Mg concentration.15) Figure 10 shows the yield strength, the ultimate tensile strength, and elongation of the several commercial Al–Mg alloy (for wrought, O tempered). The Al–12%Mg alloy sheet fabricated in this study has an excellent strength and ductility balance.
Fracture surface of Al–12%Mg alloy sheet after tensile test.
Relationships between (a) yield strength and elongation, and (b) ultimate tensile strength and elongation of typical commercial Al–Mg alloy (for wrought, O tempered).
This work was supported by Japan Foundry Engineering Society young research grant and The Light Metal Educational Foundation research grant.
