2023 Volume 64 Issue 2 Pages 479-484
A5052 aluminum alloy and SPCC steel plates were welded using magnetic pulse welding, and interfacial microstructure and strength of the lap joint were examined. The A5052 aluminum alloy and the SPCC steel plates were used for a flyer plate and a parent plate, respectively. Charging energy stored in capacitor and gap between the A5052 aluminum alloy and the SPCC steel plates were changed. Microstructure was examined by using an optical microscope, a scanning electron microscope and a scanning transmission electron microscope. Tensile-shear test was used for evaluating the strength of the joint. The magnetic pulse welding of A5052 aluminum alloy and SPCC steel was achieved at the charging energy above 6.0 kJ and at the gap between the plates above 1.0 mm. The range of charging energy in which the strong welding is accomplished was different at each gap and the range increased with decreasing the gap. The lap joint was not fractured at the welding interface by the tensile-shear test whereas the fracture occurred at a part of the aluminum base metal. The welding interface exhibited characteristic wavy morphology. An intermediate layer was produced along the wavy interface. Scanning transmission electron microscope observation revealed that the intermediate layer is composed of fine Al–Mg grains and dispersed dendritic Al–Fe intermetallic compound particles.
The importance of welding technology for dissimilar metals has been increasing, since multi-material structures are useful for improving functionality of products. In particular, dissimilar-metal welding of aluminum alloy and steel is a technology which contributes to reducing the weight of transportation equipment represented by a vehicle, and the weight reduction results in improved fuel efficiency and reduced exhaust emissions, thus contributing to the construction of a “carbon neutral society”.1,2) However, when the conventional fusion welding such as arc welding,3) resistance welding4) and laser welding5,6) is applied to the welding of aluminum and steel, it is known to be difficult to obtain sufficient joint strength due to the formation of thick layers of brittle Al–Fe intermetallic compounds such as Al3Fe and Al5Fe2 phases at the welding interface. Therefore, solid-state welding method is demanded for suppression of growth of the intermetallic compound layer.
Magnetic pulse welding is a kind of impact welding represented by explosive welding.7) Figure 1 shows schematic diagrams of the principle of magnetic pulse welding. A discharge circuit consisting of a capacitor, a discharge gap switch and a one-turn coil is used (Fig. 1(a)). The two or multiple metal plates with a small gap are fixed above the coil. The metal plate of the coil side is called the flyer plate and the other metal plate is called the parent plate. A high-density magnetic flux is generated around the coil by discharging the electrical energy stored in the capacitor (Fig. 1(b)). Eddy currents are induced on the surface of the flyer plate by intersecting of the magnetic flux and the flyer plate. The correlation between the magnetic flux and the eddy current produces the electromagnetic force in the upward direction, i.e., from the flyer plate to the parent plate, according to Fleming’s left-hand rule. The flyer plate is deformed toward the parent plate by the electromagnetic force and collides with the parent plate at a certain angle at high speed to form a joint (Fig. 1(c)). The welding is normally completed within 10 microseconds with a negligible temperature rise.8–10)
Schematic diagrams of magnetic pulse welding. (a) Discharge circuit. (b) Generation principle of magnetic flux, eddy current and electromagnetic force. (c) After welding.
It has been shown that magnetic pulse welding can form various combinations of similar- and dissimilar-metal joints due to effect of the high-speed oblique collision.11) The welding interface often exhibits characteristic wavy morphology regardless of the combination of metals, and depending on the combination of metals an intermediate layer may be produced along the welding interface.12–16) There have been several reports on magnetic pulse welded aluminum alloy/steel joint, and the strong joint can be obtained.17–20) However, the correlation among welding condition, interfacial microstructure and joint strength has not yet been systematically clarified. In the present study, we investigated the welding condition under which aluminum alloy and steel can be welded by magnetic pulse welding. In addition, the strength and interfacial microstructure of the obtained joint were examined.
A 0.5-mm-thick A5052-H32 aluminum alloy (hereafter Al) plate and a 0.8-mm-thick SPCC-SD steel (hereafter steel) plate were used. The chemical compositions were shown in Table 1 and Table 2. The length and width of each plate were 200 mm and 70 mm, respectively. The surface of the plate was polished by a #1000 abrasive paper and then was ultrasonically cleaned in acetone before welding.
A magnetic pulse system (Bmax, MP 12.5/25, capacitance of a capacitor: 40 µF) was used for performing magnetic pulse welding of Al and steel plates. A one-turn coil with a trapezoidal cross section (length of upper surface: 3 mm) was used. The Al plate was fixed by overlapping the upper surface of the coil with a width of 3 mm, and then the steel plate was set with a small gap over the Al plate. This means that the Al plate and the steel plate were used for the flyer plate and the parent plate, respectively. The charging energy and the gap between the plates was changed in the range from 6.0 to 8.0 kJ and 1.0 to 2.0 mm, respectively.
Tensile-shear test was used for evaluating the joint strength. The specimens used for the test were cut from the obtained joints by using an electric discharge machine to have a shape which conformed to Japanese Industrial Standard 13B (length of parallel part: 58 mm, width of parallel part: 12.5 mm). The test was performed at room temperature at a rate of 1.0 mm/min. Three tests were conducted for each welding condition.
Microstructure of the welding interface was observed using an optical microscope (BX51M), a field emission scanning electron microscope (FE-SEM, S-4500, acceleration voltage: 15 kV) and a scanning transmission electron microscope (STEM, Technai Osiris, acceleration voltage: 200 kV). The chemical composition was examined using an energy-dispersive X-ray spectrometer (EDX) equipped with SEM and STEM. Samples for microstructural observation and analysis were cut out from the obtained joints and their cross section was mechanically polished to a mirror surface. STEM samples were prepared by using focused ion beam system (FB-2100, acceleration voltage: 40 kV).
Result of welding experiment of Al and steel plates is shown in Table 3. An open circle and a cross mark indicate the success and failure of the welding, respectively. The result of welding was differed by combination of the charging energy and the gap between the plates. Any charging energy was available to weld them with the gap of 1.0 mm. In contrast to that, when the gap of 1.5 mm was used, the welding was achieved at the charging energy in the range from 7.0 to 8.0 kJ. Also, in case of the gap of 2.0 mm, the welding was accomplished at the charging energy of 8.0 kJ.
Figure 2 shows (a) a typical macroscopic appearance and (b) a schematic illustration of the joint. The plate on the right in the figure is the Al used as the flyer plate, and the plate on the left is the steel used as the parent plate. Also, a width of approximately 15 mm from the tip of the Al plate near the center overlaps the steel plate. The Al plate was deformed toward the steel side by a width of approximately 10 mm from the edge near the center (indicating by the gray hatched line in Fig. 2(b)) and was pressed to the steel plate. This indicates that electromagnetic force acted on a part of the Al plate which overlapped the steel, deforming the edge of the Al plate toward the steel plate. In order to examine the joint strength, the specimens were cut so that the deformed part of the Al plate was located in the center of the parallel part in the specimen, as shown by the dashed line in Fig. 2(b), and tensile-shear tests were performed. The result is shown in Fig. 3. Figure 3(a) is a typical macroscopic appearance of the specimen after the tensile-shear test and Fig. 3(b) is relationship among average fracture load obtained by using the tensile-shear test, charging energy and gap between the plates. An arrow in Fig. 3(a) indicates the welded area in the specimen. The average fracture load of the Al and steel base metal plates obtained using tensile tests are also shown as dashed lines in the Fig. 3(b). In all specimens welded under each welding condition, fracture did not occur at the welded area, but at the Al base metal. No significant difference was observed in the fracture load of each specimen depending on the welding conditions, and the fracture load of the joint was equivalent to that of the Al base metal. This result indicates that strength of the welded area is stronger than that of the Al base metal.
(a) Macroscopic appearance of a joint observed from Al side. (b) Schematic diagram of a joint indicating collection locations of specimens for tensile-shear tests.
(a) Macroscopic appearance of a specimen after tensile-shear test. Arrow indicates location of welded area. (b) Relationship among fracture load, charging energy and gap between plates.
As mentioned above, strong welding of Al and steel plates was accomplished by using magnetic pulse welding, and the range of charging energy in which the strong welding is accomplished decreased as the gap between the plates increased. This is considered to indicate that the formation of the welding interface is affected by the collision velocity and the collision angle. Magnetic pulse welding is one of the impact welding method represented by explosive welding. For explosive welding, welding feasibility is systematized by the welding window consisting of collision velocity and collision angle.21,22) In case of magnetic pulse welding, since the charging energy stored in the capacitor is discharged to the coil as a pulse current by closing the discharge gap switch and is converted into electromagnetic force, it is considered that it mainly affects the collision velocity. Also, since the gap between the plates affects the deformation distance of the flyer plate, it is considered to mainly affects the collision angle. Therefore, the obtained result in the present study may be rephrased that the range of collision velocity which accomplished the welding differs depending to the collision angle. In the present study, the welding was not accomplished at relatively low charging energy as the gap became larger. This is considered to indicate that the welding of Al and steel is accomplished in the relatively slow collision velocity range for a relatively small collision angle and in the relatively fast collision velocity range for a relatively large collision angle.
3.2 Microstructure of strong welding interfaceFigure 4(a) shows an optical micrograph of a cross section of the area where the Al plate was pressed into the steel plate in the joint formed with the charging energy of 8.0 kJ and the gap between the plates of 2.0 mm. The upper plate in the figure is the steel plate used as the parent plate and the lower plate is the Al plate used as the flyer plate. As mentioned in Fig. 2(a), a part of the Al plate was deformed to the steel side and pressed to the steel plate. Detailed observation of the interface between the Al plate and the steel plate in the pressed area revealed that the entire width of the pressed area was not welded but a part of it was welded, as indicated by the arrow in Fig. 4(a). Relationship among the width of the welded area, the charging energy and the gap between the plates are shown in Fig. 4(b). The width of the welded area was measured as the horizontal distance at the welded area when surface of the steel plate was horizontal in the pressed area. The width of the welded area increased with increasing the charging energy. On the other hand, they were almost constant with increasing the gap between the plates.
(a) Optical micrograph of cross section of a joint. (b) Charging energy or gap dependence on width of welded area.
Figure 5(a) shows an optical micrograph of the welding interface formed with the charging energy of 8.0 kJ and the gap between the plates of 2.0 mm. The welding interface exhibited characteristic wavy morphology. The interfacial wavy morphology was observed through the entire welded area. Such periodic wavy morphology at the welding interface has been observed in the similar- and dissimilar-metal joints fabricated by using magnetic pulse welding, and is known to be one indicator of a strong welding in impact welded joints. A SEM image of the welding interface are shown in Fig. 5(b). The upper bright-contrast side is steel plate used as the parent plate and the lower dark-contrast side is Al plate used as the flyer plate. An intermediate layer with intermediate contrast between the two base metals was produced at the region between the Al and the steel plates. The thickness and width of the intermediate layer were not constant in the welded area. The intermediate layer was continuously formed in some places larger than twice the amplitude or wavelength of the wavy interface and in other places smaller than half the wavelength or amplitude of the wavy interface. The intermediate layer was composed of particles with various diameters less than 5 µm. The contrast of the particle was not constant and various intermediate contrasts were observed in the intermediate layer. The coarser particles had brighter contrast similar to that of the steel base metal, indicating that the coarser particles are fragments of the steel generated by the high-speed oblique collision of the Al and steel plates. These various contrasts of the particles are considered to indicate that the intermediate layer is not a single phase, but is composed of phases with various chemical compositions.
(a) Optical micrograph of welding interface. (b) SEM image of welding interface.
Figure 6(a) shows a high-angle annular dark-field (HAADF) STEM image of the intermediate layer produced at the charging energy of 8.0 kJ and the gap of the plates of 2.0 mm. Bright-contrast particles with diameters of approximately 1 µm or less were dispersed in the matrix. An enlarged HAADF-STEM image of a bright-contrast particle is shown in Fig. 6(b). The bright-contrast particle was found to have a dendritic shape. The bright-contrast particles observed in the intermediate layer were of various sizes but had a similar dendritic shape. STEM-EDX result of the dendritic particle is shown in Table 4. Each analysis was performed for a single dendritic particle, as a typical analysis area is shown by a dashed circle in Fig. 6(b). The particles exhibited a variety of composition ratio of Al and Fe. No Mg intensity was detected in the particles. Comparing the atomic ratios of Al and Fe obtained from each particle with those of the equilibrium Al–Fe intermetallic compounds, the atomic ratios of “Particle 1”, “Particle 2” and “Particle 3” were corresponded to those of Al3Fe, Al5Fe2 and Al2Fe phases, respectively. Itoi investigated interfacial microstructure of a magnetic pulse welded A6061 aluminum alloy/DP590 steel joint, and reported formation of fine grain of Al–Fe intermetallic compounds such as Al3Fe and Al5Fe2 phases in the intermediate layer.23) The results of this report and the present STEM-EDX results of the dendritic particles are assumed to support that the dendritic particles observed in the intermediate layer are Al–Fe intermetallic compounds. Figure 7 shows (a) a HAADF-STEM image, (b) an Al map, (c) a Mg map and (d) a Fe map of the matrix in the intermediate layer. Table 5 shows chemical composition of the matrix in the intermediate layer analyzed by using STEM-EDX. The analyzed region in the matrix was a region containing multiple grains, as a typical analysis area is shown by a dashed circle in Fig. 7(a). The matrix was composed of fine grains with size of several hundred nanometers. Al was almost uniformly distributed in the matrix, and the distribution of Mg was similar to that of Al. The composition analysis revealed that Al is main element and a few atomic percentage of Mg is also detected. These results indicate that Mg is dissolved in Al and the chemical composition of the fine grains of the matrix in the intermediate layer is Al–Mg. The Al–Mg grains were surrounded by bright-contrast phase in the HAADF-STEM image. When the Fe map was observed, high Fe intensity was detected at the area corresponded to the bright-contrast area. These results indicate that the intermediate layer is composed of fine Al–Mg grains surrounded by Fe and dispersed dendritic Al–Fe intermetallic compound particles.
(a) HAADF-STEM image of intermediate layer. (b) Enlarged HAADF-STEM image of bright-contrast particle. A dashed circle indicates a typical composition analysis area of particle.
(a) HAADF-STEM image, (b) Al map, (c) Mg map and (d) Fe map of matrix in intermediate layer. A dashed circle in (a) indicates a typical composition analysis area of matrix in intermediate layer.
Formation of the intermediate layer resulted in strong Al/steel welding interface because the intermediate layer was produced along the welding interface. STEM observation and analysis revealed that the microstructure of the intermediate layer is fine Al–Mg grains with dispersed dendritic Al–Fe intermetallic compound particles. These results are considered to indicate that the formation of the intermediate layer is due to local melting and subsequent solidification. Li et al. investigated the magnetic pulse welded Al/steel interface using both experimental and numerical analyses and reported the temperature change at the welding interface.24,25) These reports showed that at the welding interface, the temperature rises above the melting point of base metal and then cools rapidly. In addition, the dendritic structure at welding interface was reported in the joints formed by using explosive welding which is a kind of the impact welding method, and it is concluded that the formation of the dendritic structure resulted from local melting and successive solidification.26) Therefore, it is assumed that the intermediate layer consisted of the dendritic Al–Fe intermetallic compound particles obtained in the present study was produced by local melting of the base metals by high-speed collision and then rapidly solidification.
In the present study, strength and interfacial microstructure of magnetic pulse welded Al and steel lap joint were investigated. Welding of Al and steel plates was achieved with the condition of the charging energy of 6.0 to 8.0 kJ and the gap between the plates of 1.0 to 2.0 mm. The obtained joints were strong enough and they did not fracture at the welding interface by tensile-shear test. The welding interface exhibited characteristic wavy morphology. The intermediate layer was produced along the wavy interface. STEM observation and analysis revealed that the intermediate layer is composed of fine Al–Mg grains and dispersed dendritic Al–Fe intermetallic compound particles.
A part of this study was financially supported by research grant from JGC-S Scholarship Foundation (2017 to 2018).
