2021 年 62 巻 8 号 p. 1184-1193
A 1050-O aluminum plate and a low carbon steel (SPCC) plate were seam welded by magnetic pulse welding (MPW). Experimental and numerical analysis methods were combined to investigate the wavy interface formation and local melting phenomena at the Al/Fe joint interface. The metallographic observation results showed that Al/Fe joint interface presents a trigger-like wave shape with two types of intermediate layers intermittently formed at the interface. Collision angle between two plates and impact velocity of Al sheet driven by electromagnetic force was numerically solved by coupled mechanical and electromagnetic fields analysis. The metal jet emission behavior, wavy interface formation process and the temperature change during MPW collision process was analyzed by SPH method. The numerical results showed that the temperature near the interface exceeds the melting point of both Al and Fe at extremely high pressure, which leads to the formation of a local melting zone (LMZ) where Al and Fe were mixed in melted state. The numerically reproduced wavy interface morphology showed a good consistency with the experimentally observed results.
Magnetic pulse welding (MPW) is a novel technique that uses electromagnetic force to accelerate and collide metals at extremely high speed under room temperature to form a robust metallurgical bonding. MPW is categorized as the solid-state joining method and shares some similarities with other impact welding, such as explosive welding (EXW)1) and vaporizing foil actuation welding (VFAW).2) Figure 1(a) illustrates the principle of the MPW process.3) The blue plate is fixed and named as parent plate. The green plate is fixed on one side, while another side will be accelerated by the magnetic force and fly onto parent plate, therefore it is called flyer plate. The parent plate is fixed at an initial distance with the flyer plate. At the onset of the MPW process, a discharged impulse current from the capacitor bank passes through the one-turn coil to generate a high-density magnetic flux around the coil. The high-density magnetic flux lines induce eddy current within the flyer plate. Eddy current flows in the opposite direction of the impulse current in the coil. The interaction between these two opposing currents activates the repulsive electromagnetic force upward. This electromagnetic force drives the flyer plate to collide with the parent plate with extremely high speed obliquely. Figure 1(b) shows the collision behavior between the flyer plate and the parent plate. During this oblique collision process, the metal jet is emitted from the collision point to clean the metal surface to facilitate a metallurgical bonding. The joint interface usually presents a characteristic wavy morphology.
Schematic diagram of (a) magnetic pulse welding process; (b) collision behavior during MPW.
When a metal plate collides with a metal plate at high speed of several hundreds to thousands m/s by the impact welding method, the pressure at the collision surface will be in the order of GPa, exceeding the Hugoniot Elastic Limit (HEL) of the metal and causing the solid metal to exhibit fluid behavior. Furthermore, partial vaporization begins to occur at the collision surface when the collision occurs at 7000 m/s or higher. In this way, high speed collisions are a complex phenomenon in which a fluid and a solid coexist.4) In addition, an intermediate layer (IML) may be formed at the joint interface of impact welding materials such as Al/Fe (Steel).5) If IML is excessively formed, the mechanical and physical properties of the joint such as bonding strength and thermal conductivity will deteriorate, so its control is important. One of the causes of the formation of IML is local melting due to the temperature rise of the joint interface during collision.6) However, it is difficult to directly measure the temperature increase near the joint interface in impact welding where joining is completed in tens to hundreds of microseconds.
In recent years, attempts have been made to reproduce the temperature change at the joint interface during collision in impact welding by numerical method. Hokamoto et al. using a one-dimensional finite difference method calculated the velocity change of a flyer plate during explosive welding of Al alloy/stainless steel and reported that the kinetic energy of the flyer plate during the collision is consumed by a large deformation near the interface, resulting in a temperature increase at the joint interface.7) Nassiri et al. using both experimental analysis and numerical simulations investigated the melting phenomena at the vaporizing foil actuation welded Al/Fe joint interface.8) They reported that the formation of microstructural defects within IML were deduced as the combined result of ultra-fast melting followed by rapid cooling. Bataev et al. also using SPH method calculated the temperature distribution at the explosive welded steel/steel joint interface and presented that due to the effect of high pressures, the melting temperature of steel significantly increased.9)
To discuss the local melting that occurs at the joint interface during collision, it is necessary to consider both the change in temperature near the interface and the change in melting point due to pressure change. In this study, we reproduced the formation process of the wavy interface in magnetic pulse welded Al/Fe material, temperature distribution near the interface and the change in melting point due to pressure change during collision using ANSYS AUTODYN to investigate the causes of local melting at the joint interface.
In the present study, the pure aluminum 1050-O plate (200 × 70 × 0.4 mm) was set as the flyer plate. Low carbon steel SPCC plate (200 × 70 × 0.3 mm) was fixed as the parent plate. Their chemical compositions are shown in Table 1 and Table 2. They are referred to as the Al and Fe hereafter. The plate surface subjected to the lap welding was cleaned with alcohol. No special surface treatment was made before welding.
Welding experiment was performed using a magnetic pulse welding system, Bmax MP12.5/25 (capacitance: 40 µF). The experimental setup for magnetic pulse welding is shown in Fig. 2. The initial distance between the two plates is called ‘gap’ that affects the collision angle and accelerating distance of the flyer plate. The discharge energy (DE) mainly determines the kinetic energy or the impact velocity of the flyer plate. In the current experiment, the gap is set as 2 mm and DE is set as 6 kJ.
Setup for magnetic pulse welding.
Figure 3(a) shows the appearance of the Al/Fe magnetic pulse welded lap joint. The assessment of bonding strength was conducted by a universal tester (SHIMADZU AG-X). The tensile shear test specimen was collected from the middle of the joint along the welding direction, and the dimensions are shown in Fig. 3(b). The tensile shear tests were performed two times under a crosshead speed of 2 mm/s at room temperature. Figure 3(c) shows the lap joint appearance before and after the test. The joints broke at the base Al plate shows that the bonding strength of joint is higher than the base metal.
(a) Appearance of MPWed Al/Fe joint and position of the tensile shear test specimen. (b) Dimension of the tensile-shear test specimen. (c) Al/Fe joint specimens before and after the tensile shear test.
Specimens for microstructure observation were collected from the tensile shear tested joints. They were cut in cross-section along the welding direction and then polished and mirror finished. The microstructural observation of the joint interface was carried out using a scanning electron microscope (SEM: KEYENCE VE-9800) and a field emission scanning electron microscope (FE-SEM: JEOL JSM-7500F).
The formation process of joint interface of MPW is reproduced by two 2D models based on the purposes and functionality of the simulation software. Model 1 (Emag-Mechanical model) using finite element method to reproduce the deformation behavior of flyer plate in the electromagnetic field.10) Model 2 (SPH model) using SPH method to reproduce metal jet emission and wavy interface formation process.11)
3.1 Model 1 (Emag-Mechanical model): Reproduction of the deformation behavior of flyer plate in the electromagnetic fieldModel 1 is a coupled mechanical and electromagnetic fields model constructed in ANSYS Emag-Mechanical that reproduces the deformation behavior of the Al sheet in the electromagnetic field. This model is used to calculate the collision angle β between flyer plate and parent plate and the impact velocity of flyer plate v. Figure 4 shows the schematic diagram of Model 1. Model 1 consists of FEM circuit and FEM model. FEM circuit consists of inductance (L), capacitor (C) and resistance (R) and coil (2D metal conductor). FEM model consists of flyer plate, parent plate, coil, air, and infinite boundary. The coil in FEM circuit is coupled with the coil element in FEM model. The true stress-strain curves obtained at a high strain rate condition are selected as the constitutive models for reproducing the deformation behavior of metallic materials in the electromagnetic field.12) The properties of flyer plate, parent plate, coil and anvil are shown in Table 3.
Schematic diagram of Emag-Mechanical model.
In the Model 1, discharge current, magnetic flux and electromagnetic force were calculated by Emag. The deformation process of the flyer plate was calculated by Mechanical. Figure 5 shows the calculation cycle of β and v in Model 1. In one cycle of the calculation, the discharge current running through the coil was firstly calculated by FEM circuit in Emag. Secondly, by coupling the coil element, the discharge current was input into the FEM model in Emag to calculate the magnetic flux around the coil and the eddy current generated in the flyer plate. The electromagnetic force was calculated from the interaction between magnetic flux and eddy current in the flyer plate and then import into the FEM model in Mechanical. Thirdly, the deformation process of the flyer plate in the electromagnetic field was reproduced by FEM model in Mechanical. The geometry shape of flyer plate will be updated at the end of each cycle and reloaded for next cycle until the flyer plate collided with the parent plate. When flyer plate collides with parent plate, the angle between flyer plate and parent plate is defined as initial collision angle β0, the velocity of flyer plate is defined as initial impact velocity of flyer plate v0. β0 and v0 will be further import into Model 2 (SPH model) as initial conditions to reproduce the wavy interface formation process during the collision in MPW. The governing equations of electromagnetic field can be found in the previous report.10)
Flowchart of the calculation cycle in the Model 1.
Model 2 is an analytical model that reproduces the collision process between metal plate and parent plate and the accompanying temperature rise using the SPH method. Figure 6 shows the schematic diagram of Model 2. This analytical model consists of flyer plate, parent plate and anvil. The symmetric system of Model 2 is a two-dimensional flat plate system assuming that the depth direction is infinite. The unit system was set to mm, mg, and ms. To reduce the computing elements, the space other than the flyer plate, parent plate and anvil was assumed as vacuum in this study. The dimensions of the flyer plate, parent plate and anvil in the analysis model are 3 mm long × 0.4 mm wide, 3.5 mm long × 0.3 mm wide and 4.5 mm long × 0.5 mm wide, respectively. The impact velocity of flyer plate v (it is referred to as impact velocity hereafter) and the collision angle β obtained from the analysis of Model 1 were set as the initial conditions for Model 2. A fixed layer was set as a boundary condition (v = 0) for SPH particles located at the top of Anvil. The parameters used in equation of state and constitutive model for flyer plate, parent plate and anvil are Al 1100-O, Steel 1006 and SS304 from ANSYS material library and listed in Table 4.
Schematic diagram of SPH model.
The smoothing length h of the SPH method is an important factor that significantly affects the accuracy of analysis results and analysis time. In this study, for the flyer plate and parent plate, the smoothing length of the area within 0.05 mm from the surface, where corresponds to the collision surface, was set as 1 µm to reproduce more detail of the collision behavior. The smoothing length of the region further away from the interface was divided into two parts and set as 2 µm and 4 µm to reduce the analysis time. The smoothing length of anvil was set as the largest of 10 µm due to no detailed collision behavior is needed to reproduce.
Gauge point is a function of AUTODYN, by installing a gauge point on a specific SPH particle, it is possible to know the time history change of the specific parameters (temperature, pressure, and specific internal energy, etc.) related to that particle. Each SPH particle can only be installed for one gauge point. In this study, we focused on the range 2.6∼2.9 mm away from the first impact point and installed 200 gauge points on the SPH particles at the collision surface region (0∼0.004 mm from metal surface) and each base metal (0.02 mm away from the metal surface) (as shown in Fig. 6).
Figure 7 shows the SEM backscattered electron images of Al/Fe joint interface. Figure 7(b) and (c) show the enlarged image of white rectangle in Fig. 7(a). In Fig. 7(a), the left edge of the downside Al plate sharpened distinctly due to the dramatic deformation. This area is defined as the first impact area. Along the welding direction, the joint interface presents an inconstant characteristic. The interface morphology shows a trigger-like morphology with IML, as indicated by the gray area, located near the vicinity of interface area. Away from the first impact area, the joint interface firstly shows a fractured area with continuous IML adhered to the Fe surface, as shown in Fig. 7(b). It is considered that in this area, during the collision process, a weak bond is formed briefly and then broke. A sound bonding is usually cannot formed near the first impact area. Continuously, along the welding direction, the IML shows a gradually decreased tendency. Figure 7(c) shows the onset of the welding area. In this area, the IML at the interface becomes intermittent and thinner.
SEM backscattered electron images of Al/Fe joint interface.
Figure 7(c) shows the high magnification FE-SEM image of the white dotted rectangle region in Fig. 7(b). The IML presents two different contrast areas: the brighter contrast zone is surrounded by Fe, and the darker contrast zone is located between the Fe and Al. The IML with bright contrast has a quite homogeneous structure, while the IML with darker contrast shows a heterogeneous structure and contains many Fe-rich fragments. The embedded Fe-rich fragments in IML may be firstly ejected out of the Fe surface by the intense impact force, and then wrapped into the local melted compound of Al and Fe by an intense shear force.13,14) Depending on the formation location and composition difference, the two kinds of IML were defined as Tail-side layer (TSL) and Front-side layer (FSL).6) TSL is formed at the rear of the wave region and enveloped by the Fe at the collision surface, while the FSL is formed at the front of the wave region and adjoined with the Al matrix.
4.2 Reproduction of electromagnetic field and calculation of collision angle and impact velocity of flyer plateIn the Emag-Mechanical model, the value of impulse current is needed to solve the FEM circuit. In this study, the impulse current is obtained by fitting the current waveform obtained from the experiment with that of analytical analysis results. Figure 8 shows the fitting of the impulse current waveform by experiment with numerical analysis. The numerical analysis was conducted from the onset of discharge to the moment when flyer plate collides with parent plate. The current waveform obtained by the numerical analysis shows good agreement with the experimental result.
Comparison of the impulse current waveform by experiment and numerical analysis.
Figure 9 shows the magnetic flux distribution in electromagnetic field and deformation behavior of flyer plate reproduced by Emag-Mechanical model. The maximum magnetic flux density concentrates on the region between the coil and flyer plate. The forefront of flyer plate corresponding the top of coil received the largest electromagnetic force and deformed radially up towards parent plate. The elapsed time between the onset of discharge to the moment when flyer plate collides with parent plate is around 8 µs. The location where the flyer plate starts to collide with parent is defined as first impact point.
Magnetic flux distribution in electromagnetic field and deformation behavior of flyer plate reproduced by Emag-Mechanical model. The color bar represents the magnitude of magnetic flux density.
From the analysis results of Model 1, the initial impact velocity v0 and initial collision angle β0 when flyer plate collides with parent plate can be obtained as shown in Fig. 10. Here, the subscript 0 indicates the status when flyer plate starting to collide with parent plate. The obtained v0 = 619 m/s and β0 = 15° will be import into Model 2 (SPH model) as the initial conditions. It is worth noting that in the present SPH model, the impact velocity and collision angle are considered to remain constant during the collision process.
Initial impact velocity v0 (vx0, vy0) of flyer plate and initial collision angle β0 at first impact point.
Figure 11 shows the overview of the SPH model at 0 µs (at the start of the collision), 0.6 µs (during the collision), and 1.4 µs (at the end of the collision). The wavy interface begins to form at about 2 mm from the first impact point and becomes constant after 2.5 mm. In this study, we focused on the region (length 0.3 mm × width 0.1 mm) where located 2.8 mm from the first impact point (as shown in the red frame) and investigated the formation process of the wavy interface in this region.
Overview of the collision process between flyer plate and parent plate.
Figure 12 shows the formation process of the wavy interface between 1.05∼1.17 µs in the region inside the dashed line in Fig. 11. While Al obliquely collides with Fe, metal jets are continuously emitted from the collision point. A trigger-like wavy interface is formed behind the collision point. In the emitted metal jets, Al and Fe were randomly mixed and dispersed. In addition, six gauge points (A–F) are selected to investigate the change history of material parameters at the specific location. Gauge points A, C are installed on the surface region (0∼4 µm away from metal surface) of Fe while B, D are installed on the surface region of Al. E and F are installed 20 µm away from the metal surface. During the collision process, A and B were drawn into the apex of the wavy interface. C and D moved to the foot of wave. E and F installed at locations far from the interface did not move much from the initial position compared to other gauge points and remained in each base metal away from the interface throughout the collision process.
The emission behavior of the metal jet and wavy interface formation process during collision.
To further confirm the component of metal jet, two different colors are assigned to the surface region of each metal plates (as shown in Fig. 13). The yellow and purple particles are indicating the region 5 µm from the surface of parent plate and flyer plate, respectively. The numbers of Al and Fe particles in the emitted metal jets in Fig. 13 were counted and the particle number ratio of Al and Fe was about 26:1. It clearly shows that the emitted metal jets mainly composed of Al. Kakizaki et al.15) compared the composition of the metal jet at different metal density differences using both simulation and experiment methods. They proved that when the density difference of the two metal plates was large (Al/Cu, Al/Fe, Al/Ni), the metal jet mainly composed of the lower density metal. When the density between each metal is close (Al/Al, Cu/Ni), the metal jet composed of both metals with a similar amount. This can be explained by the conservation of momentum during collision process in MPW. An ideal situation describing the impact process at one unit distance ΔX is assumed without considering the friction loss and air resistance (Fig. 14). In this ideal situation, the total momentum is assumed to be conserved for both the whole flyer-parent plate system and local collision at one unit distance ΔX. The flyer plate firstly impacts onto the parent plate with an impact velocity of v, then the two plates are joined and keep moving with a buffered velocity v′ in a short time. The points A and B are the center of gravity of the flyer plate and parent plate before the impact, while A′ and B′ show the center of gravity after the impact. The thickness of each plate is reduced due to compression caused by the plastic deformation. The center of gravity of each plate is moving upward and the distance changes are indicated by df and dp, respectively. According to the conservation of momentum, for unit distance ΔX during the collision process in MPW, there is:
| \begin{equation} (m_{\text{f}} + m_{\text{p}})v' - m_{\text{f}}v = 0 \end{equation} | (1) |
| \begin{equation} (m_{\text{p}})v' = F\varDelta t, \quad m_{\text{f}}(v' - v) = F \varDelta t \end{equation} | (2) |
| \begin{equation} \rho_{\text{f}}\Delta v_{\text{f}} + \rho_{\text{p}}\Delta v_{\text{p}} = 0 \end{equation} | (3) |
The emission behavior of metal jet with different color assigned to the surface region.
Schematic illustrations of the ideal impact process of flyer plate and parent plate at unit distance ΔX.
The relationship between the emission behavior of the metal jet near the collision point and the formation of the wavy interface will be investigated in more detail. Figure 15 shows the emission behavior of metal jet near the collision point from 1.07∼1.12 µs. At 1.07 µs, the metal jets were emitted from the Al side toward the Fe surface in front of the collision point. The emitted metal jets collide with the Fe surface in front of the collision point at 1.08 µs, and most of them were emitted as an Emitted flow along the Fe surface in front of the collision point, and a part of them were incorporated into the interface as an Entrapped flow. Furthermore, as shown by the yellow arrow, Fe overhangs to the Al side, causing a flow of Fe particles in the opposite direction to the Entrapped flow. As a result, an apex between Al and Fe was formed behind the collision point at 1.09 µs. At 1.10 µs, the overhang of Fe occurred in the region indicated by the yellow arrow, and the emission direction of the metal jet changed toward the Fe surface in front of the collision point. At 1.11 µs, the metal jet collides with the Fe surface again. At this time, as in 1.08 µs, the metal jets divided into the Emitted flow emitted along the Fe surface in front of the collision point and the Entrapped flow incorporated into the interface, and the latter formed an apex between Al and Fe. At 1.12 µs, as in 1.09 µs, an apex between Al and Fe was formed behind the collision point due to the flow of Fe particles in the opposite direction to the Entrapped flow as indicated by the yellow arrow. In this way, in the Al/Fe magnetic pulse welding, the metal jets are emitted from the Al side with low density and repeatedly collide with the Fe surface with high density in front of the collision point, resulting in a trigger-like wavy interface.
The behavior of metal jet emission and wave interface formation during collision in MPW.
Figure 16 shows the pressure change at each gauge point during the formation of a wavy interface. The pressure at gauge points A and B located at the apex of wave within Al and Fe increase rapidly to 25 GPa and 9 GPa around 1.1 µs. After that, the pressure decreases gradually until 1.4 µs. The rapid increase in pressure is expected to cause large compression. The pressure of gauge point C and D, where located at the foot of wave, reached around 18 GPa and 5 GPa, respectively.
Pressure change during collision at each gauge point.
The maximum pressure of A, C are higher than B, D, which indicates that the location of A and C in Fe experienced more compression than B and D in Al. During the collision, the pressure at the outermost surface of the colliding metal plate should be the same due to the interaction, but due to the low density of Al, more particles on the outermost surface of the Al plate are emitted as metal jets. Therefore, the Fe particles left at the apex of wavy interface after the collision is almost the same Fe on the outermost surface of the Fe plate, while the Al left at the apex of wavy interface is not the same Al on the outermost surface of the Al plate. Therefore, the maximum pressure of Fe in the apex of wave within Fe is considered higher than that of Al.
In addition, the maximum pressure at E and F where located far away from interface were smaller than that of A–D, and the maximum pressure was about 3∼4 GPa. Therefore, the compression at E and F are smaller than that of A–D. It should be noted that the pressure increased rapidly due to the collision, but then also dropped rapidly. At 1.4 µs, the pressure at all gauge points dropped to almost atmospheric pressure. In this way, it was clarified that the time when the high pressure is reached near the joint interface during collision is extremely short.
As the pressure increases, the melting point of the material is expected to increase as well.16,17) Therefore, to discuss local melting phenomenon at the joint interface, it is necessary to consider not only the temperature increase at the joint interface during collision, but also the increase in the melting point of the material due to the increase in pressure at the collision point. In 1966, Gilvarry simplified Lindemann’s equation of melting by approximating it based on data from previous studies and showed that the pressure dependence of the melting point can be expressed by eq. (4).16)
| \begin{equation} T_{m} = T_{m0}(V_{0}/V)^{2 \varGamma - 2/3} \end{equation} | (4) |
| \begin{equation} (V_{0}/V)_{\textit{Al}} = 0.0086p + 1 \end{equation} | (5) |
| \begin{equation} (V_{0}/V)_{\textit{Fe}} = 0.0055p + 1 \end{equation} | (6) |
Temperature and the estimated melting point at each gauge point.
The temperature at each gauge point reached its maximum around 1.1 µs. The maximum temperatures of A and B increased rapidly to about 2600 K and 1730 K at around 1.1 µs, respectively. The maximum temperature of C and D reached 1930 K and 1130 K. The temperatures of E and F located away from the joint interface hardly changed. According to Gauss’s law, the magnetic flux generated in the coil is attenuated by the square of the distance, so it is considered that the direct effect of the leakage magnetic field of the coil on the temperature rise at the interface is extremely small.
The melting point increased up to about 2610 K and 1060 K for A and B, respectively, as the pressure increased near 1.1 µs. The melting points at C and D increased to about 2340 K and 1030 K, respectively. And the melting points at E and F increased to about 1910 K and 980 K, respectively. Among these selected gauge points, the temperature at A and B has clearly exceeded the melting point of Fe and Al, respectively. Therefore, it is considered that melting occurs at the location of A and B. Although the highest temperature at C (1930 K) passed the atmospheric melting point of Fe (1811 K), it did not exceed the melting point at the corresponding high pressure (2340 K). So, the local melting did not occur at C. The temperature at C exceeded the melting point by a little amount, but not as much as B. Therefore, local melting is also considered occur at location D to a lesser degree than B. The temperatures at E and F have not reached the melting point, and it is considered that melting does not occur at these locations.
In this way, the melting points of Fe and Al increase due to the increase in pressure caused by the collision. However, these increased melting points decreased rapidly when the pressure returns to atmospheric pressure, and at 1.4 µs, they became almost equal to the melting point under atmospheric pressure.
4.5 Temperature distribution and location of local melting zone at the joint interfaceFigure 18(a) shows the temperature distribution of the wavy interface at 1.4 µs. The temperature of the base metal of Fe remained at room temperature of 300 K and hardly increased at more than 30 µm from the interface. On the Al side, the temperature rose to about 400∼700 K within 40 µm from the interface, but the temperature of the base metal further away from the interface did not rise and remained at 300 K, as in the case of Fe. In addition, high temperature regions exceeding 1500 K have been observed near the apex region, where the melting of Fe and Al may occur. Therefore, such region is defined as local melting zone (LMZ). Figure 18(b) shows the location of LMZ (as indicated in red color). Local melting phenomenon is limited to the extreme vicinity of the joint interface. Therefore, it is considered that a large temperature difference occurs between the joint interface and the base metal where temperature has hardly risen, and this large temperature difference causes rapid heat removal.
(a) Temperature distribution near the interface; (b) Location of local melting zone.
To clarify the formation mechanism of IML formed at the joint interface and its formation process, it is necessary to investigate the cooling process of the joint interface in detail in addition to the temperature and substance distribution at the joint interface clarified in this study. However, the simulation method used in this study cannot reproduce the cooling process of the joint interface. To achieve this purpose, it is necessary to introduce an analysis method that can analyze heat conduction. Therefore, we also investigated the cooling process of the joint interface using OpenFOAM. The results of the study on the formation process of IML based on the cooling process of the joint interface will be reported in a separate report.
It is worth noting that the SPH method can calculate the local composition from the ratio of particles in a certain region, and from the obtained temperature change and material, it can calculate the composition of the IML, which is considered to be formed by partial melting and rapid cooling, such as TSL. However, it cannot reproduce the fragments ejected from Fe surface by the impact as far as the SPH model assuming a smooth sample surface is analyzed. It is possible to reproduce even finer parts of the wavy interface by reducing the size of h. In addition, more precise observation is required to trace the movement of particles and their time-dependent changes in location, temperature in the jet and the heavily deformed area surrounding the collision point.
Magnetic pulse welded joint of 1050-O/SPCC were fabricated at initial gap 2 mm and discharge energy 6 kJ. The obtained joint interface was investigated by both experimental and numerical analysis methods.
The Emag-Mechanical model was used to reproduce the electromagnetic field and the collision process between the flyer plate and the parent plate. The impact velocity of flyer plate v (vx, vy) and the collision angle β were calculated from the results of Emag-Mechanical model. The SPH model was carried out to investigate the formation process of the wavy interface, temperature change, and melting point change during collision at the joint interface of Al/Fe magnetic pulse welding materials. With the formation of the wavy interface, a rapid temperature rise occurred near the interface. However, the temperature rise occurred only in the vicinity of the interface, and the temperature of the base material 40 µm away from the interface hardly rose. At the collision point, the melting point increased as the pressure increased, but the duration only last for several microseconds. A local melting phenomenon may occur at the location where the temperature exceeds the melting point.
In LMZ obtained from the SPH model, Al and Fe particles were mixed, and the shape of LMZ was in good agreement with the shape of IML observed at the joint interface. This suggests that IML at the joint interface was formed mainly by rapid solidification of LMZ where the molten Al and Fe was generated and mixed at the interface during the collision.
A part of this work was financially supported by the Japan Society for the Promotion of Science (JSPS) 19K05028 (Shinji Kumai).
