Welding Characteristics and Effect of Gap Length on 2000 Series Aluminum Alloy Sheet Lap Joints Welded by Magnetic Pulse Welding
2017 Volume 58 Issue 12 Pages 1629-1635
Details
2017 Volume 58 Issue 12 Pages 1629-1635
Effect of gap length on welding characteristics of 2000 series aluminum alloy sheets in magnetic pulse welding (MPW) was investigated. Collision time for the 2017-T3 sheet welded at gap length of 1.2 mm and 4.6 mm were 6.16 μs and 16.9 μs respectively, which were increased with increasing of gap length. The collision speed calculated from collision times increased as the gap length increased, showed a maximum value of 380 m/s with a gap length of 2.0 mm, and then the speed declined. This result indicates that it is possible to join sheets at a higher collision speed even with the same discharge energy by adjusting an appropriate gap length. Strong lap joint was achieved for the 2024-T3/2024-T6 and the 2024-T3/7075-T6 lap joint sheets by welding condition of gap length more than 1.0 mm. Thus, it could be possible to improve the weld strength by widening the gap length and performing MPW at high collision speed. SEM and HAADF-STEM observations showed no clear oxide films formed on the weld interface. From microstructure observation, it was considered that lap joints fabricated by MPW were welded at solid state.
This Paper was Originally Published in Japanese in J. JILM 67 (2017) 423–429.
It is well known that heat treated type aluminum alloys from the 2000 and 7000 series are classified as difficult materials to weld together via the conventional fusion welding method due to the weak mechanical properties of the molten portion of the welded zone.1) In the recent years, solid-state welding such as friction stir welding (FSW) has been applied in joining these heat-treated type aluminum alloys. Previously, the 5000 and 6000 series, also, the 2000 and 7000 series aluminum alloy which are comparatively higher in strength were successfully joined together using this superior solid-state welding technique.2–6)
Magnetic pulse welding (MPW) is also known as one type of solid-state welding technique.7) MPW is a joining method whereby magnetic flux is applied abruptly to a flyer sheet, forcing it to collide to the fixture sheet at high speed. The impact energy from the generated electromagnetic force join the two metal sheets together, and this welding mechanism is similar to those in explosive welding (EXW).8) Additionally, wavy pattern can be observed at the weld interface of a good lap joint state, both in MPW and EXW. From welding perspective, the anchoring effect of both welding sheets is a preferable condition for strong bonding lap joint. Successful bonding of similar and dissimilar pure metal sheets, as well as the bonding between 6000 series aluminum alloy and SPCC have been done in the past, and details on the structure of the weld interface have been reported.9–14)
Previous reports stated that similar metal joining of 2017-T3/2017-T3 and 2024-T3/2024-T3 lap joints by MPW were possible whereby the weld width increased when the discharge energy W (hereafter abbreviated as W) applied increased, and strong bonding (base metal fracture) were also obtained.15) The increase in W results in the increase in collision speed and pressure, and this is one of a crucial joining conditions in MPW. On the contrary, the gap length between the fixture sheet and the flyer sheet is also one of the joining conditions which should be taken into consideration. The relationship between the gap length and the collision speed were investigated using industrial used pure-aluminum sheet (Plate thickness: 0.5 mm). It was found that the deformation speed of the flyer sheet at a gap length of 0.2 mm to 2.0 mm was 100 m/s or more, and gradually decreases after reaching at 300 m/s.16) This phenomenon is caused by electromagnetic force which continuously accelerate the flyer sheet during the welding process. Therefore, the deformation speed varies depending on the balance between the electromagnetic force and the mechanical properties such as work hardening rate of the flyer sheet. Numerous experiments have also been conducted on pure copper sheet (Plate thickness: 0.2～1.0 mm) and it was reported that the deformation speed increases until the gap length reaches a certain distance, and when the thickness of the metal sheet increases, deformation resistance increases, thus, collision speed decreases.17) This simply means that the change in the gap length directly affects the collision speed, hence this is one of the important joining conditions in MPW.
In the present study, MPW was conducted to join 2000 series aluminum alloy. 2024-T3 sheet was used as flyer sheet, 2024-T6 and 7075-T6 sheets with a much higher strength were used as fixture sheet. The gap length between the flyer sheet and fixture sheet were adjusted to change the collision speed and the effect of this gap length on the lap joint were investigated. Microstructure observation on the weld interface was carried out and the evaluation of weld characteristics were reported.
In MPW, two metal sheets are set on a coil with a gap in between provided by a spacer. The metal sheet near the coil is called the “flyer sheet”, whereas the metal sheet fixed above it is called the “fixture sheet”. Electromagnetic force applied deforms the flyer sheet forcing it to collide to the fixture sheet at high speed, thus joining both in a seam condition. The capacitance was set at 400 μF, and chromium copper was used to make the flat E-shaped one-turn coil for MPW. Reference on MPW principle can be found in papers reported by Okagawa et al.9) For the experiment, 2024-T3 sheet was used as the flyer sheet whereas 2024-T6 and 7075-T6 sheets were used as fixture sheets. The dimensions of the 2024-T3 and 2024-T6 sheets, both were 80 × 100 × 1.0 mm, whereas for 7075-T6 sheet was 80 × 100 × 3.0 mm. Table 1 shows the chemical composition of 2024-T3, 2024-T6, and 7075-T6 sheets, respectively. Experiment was conducted with W set to 3.0 kJ and 3.5 kJ with the gap length (d) adjusted from 1.0 to 1.4 mm. Figure 1(a) shows the appearance of the lap joint 2024-T3 and 2024-T6 (hereafter abbreviated as 2024-T3/2024-T6). Figure 1(b) and (c) shows the cross-sectional photograph of 2024-T3/2024-T6 and 2024-T3/7075-T6 lap joints respectively. As shown in the figure, the joining of thicker metal sheet is possible provided the metal sheet is set as fixture sheet during MPW process. The welding sheets were exposed to electromagnetic force results in the deformation of the flyer sheet which can be clearly seen in the figure.
| Si | Fe | Cu | Mn | Mg | Cr | Zn | Al | |
|---|---|---|---|---|---|---|---|---|
| 2024-T3 | 0.50 | 0.50 | 3.8–4.9 | 0.30–0.9 | 1.2–1.8 | 0.10 | 0.25 | Bal. |
| 2024-T6 | 0.04 | 0.13 | 4.0 | 0.47 | 1.4 | 0.01 | 0.04 | Bal. |
| 7075-T6 | 0.40 | 0.50 | 1.2–2.0 | 0.30 | 2.1–2.9 | 0.18–0.28 | 5.1–6.1 | Bal. |
(a) Macroscopic appearance of the 2024-T3/2024-T6 lap joint fabricated by MPW. Cross-sectional views of (b) the 2024-T3/2024-T6 lap joint and (c) the 2024-T3/7075-T6 lap joint. (d) Macroscopic appearances of specimens before and after tensile shear test.
Tensile shear tests were conducted using an Instron type universal testing machine to evaluate the weld strength of the lap joint. Magnetic pulse welded lap joints were cut into tensile test pieces of JIS 13B (1/2 reduction in size) using a wire electrical discharge machine. The tensile shear tests were performed at room temperature with crosshead speed set at a constant speed of 1.0 mm/min. In addition, in order to apply pure shear force at the weld joint, auxiliary plate was clamped and fixed together with the tensile shear test piece before conducting the tensile test. Figure 1(d) shows the appearance of the tensile shear test pieces before and after the tensile shear test. The tensile shear test was conducted three times for lap joints welded by several discharge energies. Low weld strength results in the test piece peeled off at the weld joint, whereas high weld strength causes the base metal (2024-T3 sheet) to fracture.
2.3 Measurement of collision time, and calculation for collision speed and collision pressure2017-T3 sheet (80 × 100 × 1.0 mm) was used to measure the collision time where W was set to 2.0 kJ, and d was adjusted from 0 to 4.6 mm. Figure 2 shows an example where oscilloscope was used to investigate the relationship between discharged current and collision time signal (d = 1.0 mm). The discharge current (upper part) reaches the maximum current value of 233 kA at 6.0 μs after the start of the current flow and attenuates while oscillating. The t0 is the first current zero value time, which is 15.9 μs. The period T is 30.0 μs. High frequency noise of 3 MHz generated at the start of the discharge circuit appears in the collision time signal (lower part) but it disappears in about 4 μs after which the pulse signal generated by the collision appears. The collision time of 5.6 μs is obtained by the detection time tc of the collision time signal when the rise time of the discharge current is 0. The collision speed was obtained by time-differentiating the curve obtained by approximating the plot with the gap length on the vertical axis and the collision time on the horizontal axis by a fourth order polynomial. For details on the measurement of the collision time in MPW, please refer to the papers reported by Okagawa et al.17) The collision pressure P was calculated from the density ρ of the 2017-T3 sheet, the collision speed V of the flyer sheet and the sound speed Vs as shown in the eq. (2.1). The equation can be applied to both EXW and MPW.8,12,15)
| \[P = \frac{1}{2}{\rm \rho}VVs\] | (2.1) |
Waveform of discharge current (upper) and collision time signal (lower) at discharge energy of 2.0 kJ. (d = 1.0 mm).
Microstructure of the weld interface was conducted on a cross-section perpendicular to the seam direction of the joint using optical microscope, scanning electron microscope (SEM: H6600). A mixed solution of 50 ml of distilled water, 2 ml of hydrofluoric acid (HF), 2.5 ml of nitric acid (HNO3) and 1 ml of hydrochloric acid (HCl) was used as an etching solution for the lap joint samples. Microstructure observation using high angle dark field scattering (HAADF-STEM) method was also carried out using a transmission electron microscope (TEM: JEM2100F). TEM samples were prepared using an ion slicer.
According to the previous reports, the strong lap joint of 2024-T3/2024-T3 was possible at W = 3.0 kJ and above.15) Furthermore, the joining of 2024-T3/2024-T6 was performed at W = 3.0 kJ, but the joining was unsuccessful. Therefore, W was then increased to 3.5 kJ, and the joining of 2024-T3/2024-T6 was possible. However, the fabricated lap joint peeled off at the joint section during tensile shear test. From this result, it can be said that it is impossible to achieve strong bonding (base metal fracture)-by only focusing on increasing W. Table 2 shows the mechanical properties of 2024-T3, 2024-T6 and 7075-T6 sheets. From this table, it was found that 0.2% proof stress and hardness value of 2024-T6 sheet was higher than that of 2024-T3 sheet, also, the elongation was about half. This seems that it is difficult to join materials with high tensile strength and poor ductility via MPW. In order to achieve ideal bonding state, there is a method whereby increasing collision pressure by increasing W and increasing collision speed are necessary. However, the increase in W burdens the coil, and deformation of the coil adversely affect the coil's life-span.
| 0.2% proof stress (MPa) |
Maximum stress (MPa) |
HV0.5 | Elongation (%) |
|
|---|---|---|---|---|
| 2024-T3 | 348 | 434 | 136 | 11 |
| 2024-T6 | 448 | 476 | 164 | 6.2 |
| 7075-T6 | 520 | 560 | 197 | 15 |
Therefore, MPW was performed by changing the gap length between the flyer sheet and the fixture sheet. Figure 3(a) shows the relationship between collision time, collision speed and gap length when 2017-T3 sheet was used as flyer sheet and W is set to 2.0 kJ. Also, Table 3 shows collision pressure calculated from gap length, collision time and collision speed. The collision time increased with the increasing gap length; 6.16 μs at d = 1.2 mm, and 16.9 μs at d = 4.6 mm. As shown in Fig. 3(a), the collision speed becomes faster as the gap length increases where it reaches a maximum of 380 m/s at d = 2.0 mm. After that, the speed gradually decreases. The reason why the collision speed increases as the gap length increases is that it takes time for the flyer sheet to sufficiently accelerate when it is deformed towards the fixture sheet via electromagnetic force. Figure 3(b) shows the relationship between collision speed and collision time. From this figure, it can be observed that the acceleration of the flyer sheet increase to a maximum value at 8.4 μs, and then decelerates becoming a negative value. This decreasing in collision speed after reaching a maximum value as described above is due to the decrease in influence of electromagnetic force on the flyer sheet as the gap length increases. Furthermore, at the same time, deformation on the flyer sheet increases, work hardening taking into effect and deformation resistance of the material becomes higher.9) The collision pressure can be estimated to be around 2.65 GPa at gap length of 2 mm. This result indicates that it is possible to manipulate the collision speed by adjusting the gap length. Since the collision time and collision speed at similar W hardly changes between 2017-T3 and 2024-T3 sheets, the effect of the gap length is considered to be almost similar for 2024-T3 sheet.15) Based on the above, 2024-T3/2024-T6 was fabricated by changing the gap length, and tensile shear test was conducted at room temperature.
(a) Relationship between collision time and collision speed with increasing gap length. (b) Relationship between traveling velocity and collision time.
| Gap length d (mm) |
Collision time tc (μs) |
Traveling velocity v (ms−1) |
Collision pressure P (Gpa) |
|---|---|---|---|
| 0.0 | 0.00 | 0.0 | 0.00 |
| 0.4 | 3.40 | 242.6 | 1.69 |
| 0.8 | 5.04 | 324.5 | 2.26 |
| 1.2 | 6.16 | 357.8 | 2.49 |
| 1.6 | 7.28 | 375.5 | 2.61 |
| 2.0 | 8.40 | 380.0 | 2.65 |
| 2.5 | 9.64 | 372.3 | 2.59 |
| 3.1 | 11.24 | 347.8 | 2.42 |
| 3.5 | 12.60 | 318.8 | 2.22 |
| 4.1 | 14.64 | 271.1 | 1.89 |
| 4.6 | 16.90 | 227.5 | 1.58 |
Figure 4(a) shows the relationship between the fracture load of 2024-T3/2024-T6 lap joint and gap length. The W was kept at 3.5 kJ. Samples peeled at the joint are indicated in white, whereas samples experienced base metal fracture (maximum load) are indicated in black. As shown in the graph, as the gap length increases, the fracture load also increases. For 2024-T3/2024-T6 welded at d = 1.0 mm and 1.2 mm, the samples peeled at joint at loads of 0.8 kN and 1.2 kN, which are 30% and 77% of failure load in the base metal respectively. At d = 1.3 mm, the sample experienced base metal fracture. Additionally, 2024-T3/7075-T6 lap joint was also conducted. Based on the results of the previous studies, strong bonding was achieved when a joint welded at d = 1.0 mm with W = 3.5 kJ.18) From this, we investigate the possibility of bonding 2024-T3/7075-T6 lap joint at a lower discharge energy at W = 3.0 kJ with different gap lengths. Figure 4(b) shows the relationship between fracture load 2024-T3/7075-T6 lap joint and gap length. At d = 1.0 mm, 2024-T3/7075-T6 lap joint peeled at the joint section with load of 70% of the base metal. At d = 1.4 mm, the same sample experienced base metal fracture.
Fracture load of (a) the 2024-T3/2024-T6 lap joint (W = 3.5 kJ) and (b) the 2024-T3/7075-T6 lap joint (W = 3.0 kJ) welded with various gap lengths.
From the above results, it can be said that it is possible to improve the weld strength by widening the gap length and performing MPW at high collision speed. This method is effective in joining materials that are difficult to bond via MPW. Furthermore, by setting the gap length to an appropriate value, it is possible to obtain the ideal bonding state at a low discharge energy.
3.2 Weld interface of the magnetic pulse welded lap jointFigure 5(a) and 5(b) show SEM images of cross-sections of 2024-T3/2024-T6 and 2024-T3/7075-T6 lap joint respectively. The upper side is the fixture sheet, the lower side is the flyer sheet. Similar to EXW, the joining via MPW can be achieved when proper welding condition, particularly when collision speed and collision angle are satisfied. In the case of using a flat E-shape one-turn coil, the joining occur at two area which are bilaterally symmetrical to the center line, which also corresponds to the coil width. Figure 6(a) and 6(b) show the optical microscopy cross-sectional images of 2024-T3/2024-T6 (W = 3.5 kJ, d = 1.0 mm) and 2024-T3/7075-T6 (W = 3.0 kJ, d = 1.4 mm) lap joints. The right side is the center region, the left side is the outer end region of the lap joint. The weld width is indicated by the solid line in the figure. A wavelike pattern caused by the plastic flow during the joining process can be observed at the weld interface.
Cross-sectional SEM images of (a) the 2024-T3/2024-T6 lap joint and (b) the 2024-T3/7075-T6 lap joint.
Low-magnification optical microscope images for (a) the 2024-T3/2024-T6 lap joint (W = 3.5 kJ, d = 1.0 mm) and (b) the 2024-T3/7075-T6 lap joint (W = 3.0 kJ, d = 1.4 mm). Positions of the initial and the end points of weld area in (a) the 2024-T3/2024-T6 lap joint and in (b) the 2024-T3/7075-T6 lap joint with various gap lengths.
Figure 6(c) and Fig. 6(d) show the positions of the initial and the end point of weld area of both 2024-T3/2024-T6 and 2024-T3/7075-T6 lap joints with various gap length. The observation results show the left side of the weld area with respect to the center portion of the lap joint sample. Looking at the distance from the central part on the vertical axis in both lap joint samples, it was found that as the gap length increases, the initial point of weld area shifts towards the central part, and the weld at the end point of weld area shifts further away from the center. This indicates that the weld width increases as the gap length increases. The weld width (sum of the left and right bond area) for 2024-T3/2024-T6 are 1.29, 2.65, and 3.38 mm at d = 1.0, 1.2, and 1.3 mm respectively. Additionally, weld width for 2024-T3/7075-T6 fabricated at d = 1.0 and 1.4 mm are 1.41 and 2.14 mm respectively. The obtained results show that regardless of the type of material used for lap joining, weld width increase with the increase of the gap length when discharge energy was set to the a constant value.
Hiroi et al. conducted FEM simulation on the deformation of the flyer sheet by electromagnetic force during MPW process (flat one-turn coil type).16) Based on the result, at the initial stage of deformation, the central portion of the flyer sheet remained flat (horizontal position) as the sheet project towards the fixture sheet. At the same time, the boundary with inclined surface through which the bending wave passes becomes the shape of plastic hinge. As the deformation progresses due to the electromagnetic force, the flat portion of the flyer sheet deforms into a conical shape, thus the result reported was consistent with experimental results.
Figure 7 shows a schematic cross-sectional view of the changes in initial weld point at different gap lengths during MPW process. As described above, deformation of the flyer sheet increases as the gap length widened, also, the flyer sheet overhangs more convexly. In this state where the curvature is large, the initial collision occurs at the center of the fixture sheet (center line in the drawing). From the initial collision point at the center, the gradient of the flyer sheet becomes steeper and the increase of the collision angle becomes more prominent while projecting towards the left and right side of the outer joint end. As a result, since the collision angle β' (collision angle at which bonding of both sheets occur) can be achieved earlier, as the gap length increases, it can be considered that the initial weld point shifts towards the center portion of the lap joint as shown by the arrow in the figure. Generally, β' is known to be 5 degrees or more.19,20) In MPW and EXW where impact force is utilized for joining, the collision angle between the bonding sheets and collision speed are important joining conditions. In the case where the collision speed at the collision point is fast enough and also the collision angle is relatively large, it is suggested that end point of weld area move to outer side from the center under the viewpoint of the joining condition. Thus, it is considered that the joining condition is maintained over a wide range of collision surface. Therefore, as the gap length increases, after the initial collision of the flyer sheet, the joinable condition is reached in a relatively shorter time, and this shifts the initial weld point towards the center portion, thus increasing the weld width of the lap joint.
Cross-sectional schematic diagram showing the change in initial point of weld area due to the difference in gap lengths.
Figure 8 shows the outer joint end section of 2024-T3/2024-T6 lap joint (W = 3.5 kJ, d = 1.3 mm), (a) SEM image of the section, also EDS analysis of (b) Al, (c) O, (d) Cu and (e) Mg respectively from the same field of view. Base on the observation on 2024-T3 sheet, precipitates of 2 to 10 μm were observed, and it was found that the precipitate was MgCuAl2 phase (Al-23at%Cu-20at%Mg). Since this precipitate overlaps with the element distribution of Cu and Mg, the boundary between 2024-T3 and 2024-T6 can be roughly determined from the presence or absence of precipitates. The portion indicated by the arrow in the SEM image is the joint end section and the black contrast area observed from the right end to the center is the unbonded region. From Fig. 8(c), it can be observed that O distribution is concentrated at the unbonded region. However, concentration of O at the weld region cannot be visibly seen. In other words, the amount of oxygen is small at the weld region as compared to the unbonded region thus, it is considered that no oxide can be detected by EDS analysis or by SEM observation.
(a) SEM image around the end point of weld area and energy-dispersive x-ray spectroscope (EDS) mappings for (b) Al, (c) O, (d) Cu, and (e) Mg, taken from same area of (a) in the 2024-T3/2024-T6 lap joint.
Figure 9(a) shows the BE image of outer joint end section of 2024-T3/7075-T6 lap joint.-From Table 1, it can be seen that 7075-T6 sheet contains a relatively large amount of Zn than 2024-T3 sheet. Therefore, as shown by the arrow in the BE image, the weld interface can be clearly seen. Wave pattern can be observed at the weld interface, but no evidence of oxide and diffusion of Zn can be confirmed. In order to investigate the weld interface in detail, HAADF-STEM observation was carried out. Figure 9(b) shows HAADF-STEM image and EDS mapping of (c) Al, (d) O, (e) Cu, (f) Zn and (g) Mg obtained from the same field of view. From the HAADF-STEM image, brighter contrast is observed in the 7075-T6 plate above the welding interface indicated by the arrow, since Zn content is higher than that in the 2024-T3 sheet. Furthermore, as shown in Fig. 9(d), the concentration of O and fine oxide cannot be confirmed at the weld interface. Up till now, the welding of pure-aluminum or aluminum alloy sheet have been successfully conducted using MPW method and it has been reported that the oxide film on the surface is discharged as metal jet during the welding process.15,21) Similar to the experiment in this study, it is considered that the oxide film on the surface was discharged during the welding process, resulting in a newly formed clean metal surfaces, suitable for joining. In Fig. 9(e), as indicated by 〇, particles of about 100 nm which turned out to be Cu was detected on the 2024-T3 sheet, indicating that MgCuAl2 phase is precipitated. From Table 1, it can be observed that the 7075-T6 sheet contains not only Zn, but also a large amount of Mg as in comparison to the 2024-T3 sheet. As shown in Fig. 9(f) concentrations of 20 to 30 nm of Zn can be observed on 7075-T6 side, but they are uniformly distributed like the distribution of Mg in Fig. 9(g). From this, it can be considered that fine precipitates (MgZn2) uniformly exist even after welding process. Also, as shown by the arrow, concentration differences of Zn and Mg element distribution can be clearly observed at the weld interface, and Zn cannot be confirmed on the 2024-T3 side. Based on this fact, it is suggested that the weld interface was not melted but bonded in a solid state.
(a) BE image around the end point of weld area in the 2024-T3/7075-T6 lap joint. (b) HAADF-STEM image and EDS mappings for (c)Al, (d) O, (e) Cu, (f) Zn and (g) Mg taken from same area of (b) at joint interface of the 2024-T3/7075-T6 lap joint.
In present study, it was suggested that the joining of materials with high mechanical strength and poor ductility is difficult using MPW method. However, by setting the gap length to be wider than 1 mm, also, by increasing the collision speed and utilizing the change in collision angle caused by the deformation mechanism of the flyer sheet by electromagnetic force in MPW process, a wide range of welding condition can be taken into consideration to achieve the desire product. As a result, it is possible to increase the weld width in any of the 2024-T3/2024-T6 and 2024-T3/7075-T6 lap joints. Furthermore, strong bonding can also be achieved for both of the lap joints fabricated by MPW.
The authors gratefully acknowledge the financial support of Grant-aided Project (2015–2016) from the Light Metal Educational Foundation of Japan. The authors also want to express gratitude to UACJ Co. Ltd. for preparing 2024-T6 sheet for this experimental.
