2020 Volume 61 Issue 8 Pages 1657-1662
In this study, pure Al powder was compression molded on the surface of flame-retardant Mg alloy plate, and integration of pure Al and Mg alloy was attempted by melting and solidification in air. In order to integrate different materials by liquid diffusion of metal elements, the structure, composition phase and hardness of the formed integrated materials were investigated.
It was found that heating the specimens for 180 s results in melting and liquid diffusion between Al and Mg. When heated and melted with pure Al down and Mg alloy up (Type B), shrinkage cavity was formed on the former Mg alloy side of the former Al/former Mg alloy interface. When heated at 450 s, the specimen was composed of Al–Mg system stable phases of Al3Mg2 and Al12Mg17, and Al–Mg system metastable phases of Al0.37Mg0.63 and Al0.1Mg0.9. In Type A with pure Al up and Mg alloy down and Type B, stable phases of Al3Mg2 and Al12Mg17 were formed on the former Al side. The constituent phase of the former Mg alloy side was composed of metastable phases of Al0.37Mg0.63 and Al0.1Mg0.9, but the structure morphologies of Type A and B differed near the surface of the former Mg alloy side, and the primary phase was Al0.37Mg0.63 and Al0.1Mg0.9 respectively. The final solidification phase consisted of eutectic phases of Al0.37Mg0.63 and Al0.1Mg0.9. The hardness on the former Al side was about 250 HV, and the hardness on the former Mg alloy side had decreased. Type B was evaluated to be lower than Type A.
This Paper was Originally Published in Japanese in J. JFS 92 (2020) 69–74.
Surface modification or composite of materials is performed by various methods.1,2) These composites are intended to impart the necessary properties to the material surface, and industrially, an atmosphere control device and advanced equipment for heating are required. The authors have found a method of generating a coating layer consisting of particle stacks by cold rolling of composite materials.3,4) Based on this result, we have been studying the development of a method to inexpensively and efficiently coat the surface of metallic materials.
Mg is the lightest metal for practical use, and has a higher specific strength and specific rigidity than both steel and Al. On the other hand, although its use is spreading, there are problems with corrosion resistance and plastic workability. Therefore, the authors have been investigating methods for forming an Al coating on Mg. Mg is not suitable for rolling because of its poor plastic workability. Therefore, the authors have attempted to apply a melting and solidification process. In this process, disturbance and convection based on specific gravity differences with liquid diffusion and Marangoni convection based on concentration differences and convection due to heating are dominant. Vacancy-based mechanisms such as solid diffusion are rare in liquids, and there is no specific energy barrier.5)
In this study, pure Al powder was compression-molded on a flame-retardant Mg alloy plate surface, and an attempt was made to integrate pure Al and Mg alloy by atmospheric melting and solidification processes. Aiming at the integration of metallic elements by liquid diffusion, the structure, phase structure, and hardness of the integrated materials were investigated.
A plate-like substrate with a thickness of 2 mm and lateral dimensions of 30 mm × 10 mm was cut from a flame-retardant Mg alloy AMX602 ingot. One side of the substrate was blasted and a pure Al powder layer (particle size: 45 to 10 µm) was formed on the substrate to produce a specimen with a surface pressure of 163 MPa using a hydraulic press (Sansho Industry: Mighty Press MT-100H) (Fig. 1). The target thickness of the powder molding layer was about 0.7 mm. In addition, when a pure Al plate was used, it was confirmed by preliminary experiments that it could not be integrated with the Mg alloy, so powder was used as the pure Al materials. Table 1 shows the chemical composition of the Mg alloy. A high-frequency furnace and an electric furnace were examined as the heating method for the specimen. In the case of the high-frequency furnace, heat is generated by Joule heating by eddy currents due to free electrons in the substrate and melts. It was found that when heated at a rate of 700 K/min, vibration due to thermal convection occurred in the substrate during melting, resulting in a lump without retaining its original shape. On the other hand, it was confirmed that heating with an electric furnace caused melting and solidification while maintaining the original shape. Therefore, priority was given to reducing the deformation risk for the specimen, and an integrated material was produced by heating with an electric furnace. The electric furnace used was a tubular furnace manufactured by Asahi Rika Seisakusho, which has an atmosphere size of 80 mm in diameter and 300 mm in length. A specimen was placed on a SUS304 steel plate (25 × 32 × 2.8 t) at the tip of a quartz jig (Fig. 2), and the electric furnace was slid and inserted into the center of the furnace. The temperature changes from the start of heating of the specimen in the electric furnace held at 993 K was measured with a JIS-K type thermocouple welded to SUS304 steel plate (Fig. 3). In preliminary experiments, it was confirmed that the specimen went through the process of heating, melting, deformation, and ignition with increasing heating time. Therefore, the heating and holding time was changed from 90 s, which is the process of heating and melting, to a maximum of 500 s. The specimen was cooled outside the furnace by sliding the electric furnace. The specimen was then devised and air-cooled so that no physical motion such as vibration occurred. The solidus and liquidus temperatures for the AMX602 alloys and the melting point of the pure Al are 799 K,6) 881 K6) and 933 K, respectively. In order to investigate the effect of the specific gravity difference between Al and Mg on interdiffusion, Type A and Type B experiments were conducted. Here, the case of heating with the Mg alloy substrate below is referred to as Type A, and the case of heating with the Al layer below is referred to as Type B. Appearance and cross-section observations were performed using scanning electron microscopy combined with energy-dispersive X-ray spectroscopy (SEM-EDS; Elionix: ERA-8800), and cross-sectional elemental analysis was carried out using electron probe micro-analysis (EPMA; JEOL: JXA-8530F) during heating and melting. Also, the constituent crystal phases were analyzed by X-ray diffraction using a Cu target (Rigaku: SmartLab). Furthermore, the hardness of the cross section (load: 100 gf) was measured with a micro Vickers hardness tester (MATSUZAWA: MMT-X7).
Schematics of preforming.
Schematics of top of quartz tube.
Change in temperature as function of heating time.
Figure 4 shows an optical micrograph of the Mg alloy. The microstructure consists of the primary αMg, the eutectic phase of Al2Ca, and αMg in the final solidification phase. Figure 5 shows the change in appearance of the specimen on the Al side when heated from 90 s to 270 s. In Type A and Type B after 90 s, it is the same as before heating. In the Type A specimen after 180 s, it can be seen that the surface color has changed. At 270 s, the surface has turned black. On the other hand, the Type B specimen is not different at 180 s than it was at 90 s. At 270 s, the surface turns black as in the case of the Type A specimen. From the temperature profile in Fig. 3, there is a temperature difference of about 25 K for the Type A specimen after 180 s and 270 s, which is considered to be the cause of the discoloration.
Microstructure of AMX602.
Appearance of specimens heated for each times.
Here, the relationship between the heating time and the melting of the specimen is considered assuming that there is no temperature difference or concentration difference in the thickness direction of the specimen. Based on the solidus and liquidus temperatures for the Mg alloy, the melting point of pure Al and the temperature profile in Fig. 3, the time to reach the melting temperature in Mg alloy and Al is 50 s and 140 s, respectively. Therefore, it is expected that in this experiment, the Mg alloy side and the Al side melt on heating for 90 s or more and 180 s or more, respectively.
Figure 6 shows an EPMA image of the specimen cross section. Figures 6(a) to 6(c) are characteristic X-ray images of Mg in the Type A specimen. The image in Fig. 6(a) shows that Mg is not distributed on the surface of the Al layer, which is a black region. On the other hand, it can be seen that Mg is distributed in the Mg alloy side region of the Al layer. In addition, newly formed dendrite is observed on the Al side of the Mg alloy. This indicates that it has melted and solidified. Therefore, it is considered that Al diffused into the Mg alloy and the melting point decreased. At 180 s in Fig. 6(b) and 270 s in Fig. 6(c), it can be seen that Mg is distributed over the entire thickness of the specimen, and a structure with new dendrite is seen. Furthermore, when an external force was applied to the specimen with tweezers, it deformed fluidly. Therefore, it is assumed that the specimen melted and liquid diffusion occurred. Figure 6(d) shows a backscattered electron image of the region shown in Fig. 6(a), which confirms the surface side of the Al layer, the interdiffusion region of Al and Mg, the newly formed dendrite region, and the unmelted Mg alloy. Figures 6(e) and 6(f) are characteristic X-ray images of Mg in the Type B specimen. At 180 s in Fig. 6(e), it can be seen that the interdiffusion of Al and Mg is not so advanced on the Al side. On the Al side of the Mg alloy, newly formed dendrites were observed, as in Fig. 6(a). This also indicates that it had melted and solidified. As with the Type A specimen, it is thought that Al diffused in the Mg alloy and the melting point decreased. Interdiffusion progresses with increasing temperature at 270 s in Fig. 6(f), but compared to 270 s in Fig. 6(c) for Type A, there is a difference in Mg concentration on the Al side. Thus, diffusion of elements occurs during heating of the specimen, creating a concentration gradient. In addition, the melting point drops and selective melting occurs. Liquid diffusion due to melting then also occurs.
EPMA images of specimens heated at different times. (a)–(c) Mg-Kα images of Type A, (d) BSE image of (a), (e)–(f) Mg-Kα images of Type B specimen.
Figure 7 shows continuous optical micrographs of the entire cross section at the center of each the Type A and Type B specimens by heating for 450 s. The specimen melts and liquid diffusion of Al and Mg alloy elements occurs. In addition, the boundary between the former Mg alloy and the former Al layer can be recognized. Comparing Type A and Type B specimens, the Type B specimen has many voids on the former Mg alloy side of the former Al layer/former Mg alloy interface. In these voids, primary dendrites crystallize, solidification of the residual melt progresses, the residual melt disappears, and solidification is completed. In other words, it is considered that shrinkage occurs. Shrinkage cavity form in the final solidified region. Therefore, it is considered that the Type B specimen contained more melt than the Type A specimen with the final solidifying composition. When the inside of a shrinkage cavity was observed with SEM, primary dendrite was observed. Solidification occurs from the free and bottom surfaces. As a result of observing the macrostructure of the entire longitudinal section of the specimen, a shrinkage cavity was found to be generated near the center of the specimen.
Transvers-section of specimens heated at 450 s.
Figure 8 shows the distribution of Al and Mg from the former Al side to the former Mg alloy by quantitative analysis using EPMA on the cross section of the specimen heated and held for 450 s. An electron beam with a diameter of 50 µm was scanned along the thickness direction of the specimen cross section at 50 µm intervals. The concentration distribution of each element changes due to interdiffusion from the concentration distribution before heating. In the Type A specimen, the concentration from the former Al side is distributed in the range of 40 to 30 at% and 50 to 70 at% for Al and Mg, respectively. On the other hand, in the Type B specimen, Al and Mg are distributed in the range of 50 to 20 at% and 50 to 80 at%, respectively. Therefore, the Type A specimen has stronger interdiffusion than the Type B specimen. In the Type B specimen, the diffusion of Al to the Mg alloy due to the difference in specific gravity between Al and Mg is considered to be weak. Therefore, the difference in concentration of Al and Mg in the thickness direction of the Type B specimen is larger than in the Type A specimen around 800 µm from the surface. In other words, in the Al–Mg system equilibrium diagram,7) the Type B specimen may have a composition in the region where the Mg concentration is higher than that in the Type A specimen. Therefore, it can be inferred that in the region of 800 µm or more from the surface in Fig. 8, the composition in which voids are formed, that is, the melt has the final solidifying composition.
Change in content of Mg and Al in Type A and B specimens.
Figures 9 and 10 show X-ray diffraction patterns for sequential surface layers removed from the former Al layer side of the Type A and Type B specimens heated and held for 450 s, respectively. From the analysis of these diffraction patterns, the constituent phases in the specimen are stable and metastable Al3Mg2, Al12Mg17, Mg0.63Al0.378) and Al0.1Mg0.9.9,10) The constituent phases in the former Al side consist of all crystal phases confirmed by X-ray diffraction in the Type A and Type B specimens. The constituent phases in the former Mg alloy side are Al0.37Mg0.63 and Al0.1Mg0.9 crystal phases in the Type A and Type B specimens. These crystal phases are metastable phases that are not seen in the Al–Mg system binary equilibrium diagram. Here, Al0.37Mg0.63 and Al0.1Mg0.9 are called MS-α and MS-β, respectively. Stable phases of Al3Mg2 and Al12Mg17 are generated in the former Al side of the Type A specimen with a scale layer of about 30 to 40 µm thick peeled off. This is probably because Mg diffused quickly into Al. On the other hand, there is no Al12Mg17 phase on the former Al surface in the Type B specimen, and Al3Mg2 and Al12Mg17 stable phases are formed on the surface with the scale layer removed. This is probably because Mg diffused into Al more slowly than in the case of the Type B specimen. In order to make the Mg concentration distribution in Fig. 8 easy to understand for interdiffusion in a liquid, it is converted to a straight line by the least squares method. As a result, the concentration range of Mg on the former Al side of the Type A and Type B specimens is 44 at% to 63 at%. This is the range where Al3Mg2 and Al12Mg17 are formed at room temperature in the Al–Mg system equilibrium phase diagram. The concentration range for Mg on the former Mg alloy side of Type A and Type B is 63 at% to 80 at%. This is the range where Mg and Al12Mg17 are generated at room temperature in the Al–Mg system equilibrium phase diagram. It is considered that nonequilibrium solidification occurs at the cooling rate associated with air cooling11) in the Mg concentration range on the former Mg alloy side. As a result, primary dendrite of MS-α or MS-β was formed, and then the eutectic phase was formed by MS-α and MS-β.
XRD pattern of Type A specimen.
XRD pattern of Type B specimen.
Figure 11 shows the microstructure near the center of Type A and Type B specimens heated and held for 450 s. In both specimens, the upper part is the former Al layer and the lower part is the former Mg alloy. When the structure of the Mg alloy substrate shown in Fig. 4 is integrated with pure Al by melt solidification, it shifts to the structure shown in Fig. 11. The horizontal arrow in the figure indicates the boundary position between the former Al layer and the former Mg alloy. In both specimens, dendrites (bright images) surrounded by broken lines from the vicinity of the interface between the former Al layer and the former Mg alloy to the former Mg alloy are seen. It can be seen that shrinkage cavity is generated in the Type B specimens. In the Type A specimen, dendrites grown from the surface of the former Mg alloy toward the former Al can be seen. On the other hand, dendrite (dark image) different from Type A surrounded by the solid line grows from the surface of the former Mg alloy in the Type B specimen. The region adjacent to the dendrite was confirmed to be a eutectic phase by high-magnification optical microscopy. Figure 12 shows the structure near the surface of the former Mg alloy side of the Type A and Type B specimens, which have different primary phases. Therefore, the Type A and Type B specimens have the same constituent phases on the former Mg alloy side but the microstructure is different. The surface near the former Mg alloy in the primary phases in the Type A and Type B specimens was analyzed by SEM-EDS. The results are shown in Table 2. From the analysis results, the primary phase in the Type A specimen is MS-α, while the primary phase in the Type B specimen is MS-β, and the eutectic phase consists of MS-α and MS-β.
Optical micrograph of transvers-section of specimens heated at 450 s. Horizontal arrow indicates boundary between former Al and former Mg alloy.
Microstructure of cross-section of Type A (a) and Type B (b) specimens close to surface in former Mg alloy.
Figure 13 shows the micro Vickers hardness of the cross section of the specimen heated and held for 450 s. It can be seen that the hardness of the former Al side is high for both the Type A and Type B specimens, and the hardness is decreased to the former Mg alloy side. The hardness of the former Al side is about 250 HV, which decreases on the former Mg alloy side, but the hardness of the Type A specimen is maintained from around 1500 µm from the former Al surface. On the other hand, it can be seen that the hardness of the Type B specimen continues to decrease. Within this range, the Type A and Type B specimens consist of primary phases of MS-α and MS-β. From Fig. 12, it can be seen that the Type A and Type B specimens have larger volume fractions of MS-α and MS-β, respectively. This is thought to be due to the hardness of the region.
Change in hardness of Type A and B specimens as function of distance from surface of former Al layer.
We would like to express our sincere gratitude to the technical staff of the Center for Instrument Analysis of Kyushu Institute of Technology, for their cooperation in this analysis. We would also like to thank Yurika Omori (currently Sankyu Inc.) and Daisuke Hotogi (currently Advanex Inc.) of Kyushu Institute of Technology, School of Engineering, for their cooperation in this experiment.
