New approach for multi-material design: Combination of laser beam and electromagnetic melt pool displacement by induced Lorentz forces

The trend toward lightweight vehicles is leading to a mix of materials to produce function and weight-optimized car bodies.¹ This so-called multimaterial design is a promising solution to reduce the weight of the vehicle, to protect the environment, and to extend the range of electric cars.^2,3 To realize the multimaterial design, the joining of dissimilar materials is necessary. For the automotive industry, the material combination of steel and aluminum alloy is of interest.⁴ However, thermal joining of steel and aluminum is difficult.² The different material properties, e.g., melting temperatures and thermal expansion coefficients, complicate the joining process. In addition, due to the insolubility between Fe and Al, intermetallic phases form, which reduce the quality and the load capacity of the joint.⁵ Several studies have investigated the metallurgical reaction between solid steel and liquid aluminum alloys. Mainly, a two-phase reaction layer consisting of a needle-shaped Al₃Fe (also known as Al₁₃Fe₄⁶) and a compact phase seam of Al₅Fe₂ is formed.^7,8 These phase compositions are mainly formed since they have the lowest free enthalpies of all Fe-Al phases.⁹ To obtain sound joints, the thickness of the intermetallic reaction layer should not exceed the critical value of 10 μm.¹⁰ These two intermetallic phase compositions are very brittle with hardness values of 820–1100 HV.¹¹ The formation of the intermetallic phases is a diffusion-controlled process and cannot be prevented. It is only possible to control the thickness of these phases below the critical value of 10 μm. For this, the maximum temperatures and the interaction time within the joining process should be limited as much as possible.^9,12 Solid-phase welding and joining processes with a low heat input such as welding-brazing are ideal for this task.¹³ Laser beam welding-brazing has been studied for joining steel and aluminum alloy. Wang et al. analyzed the microstructure of a dissimilar overlap joint between 1 mm galvanized DP590 steel and 1 mm 6061 aluminum alloy by experiments and numerical simulation. They show that with increasing line energy density the thickness of the intermetallic phases increases, and the presence of these phases affects the tensile strength of the welded joint. They detected the phases Al₅Fe₂ and Al₁₃Fe₄ with a maximal thickness of 20 μm at 162 J/mm. They achieved the best tensile strengths at 162 J/mm with 1.7 kN.⁸ Huang et al. investigated the process window and its influence on the interfacial microstructure and mechanical properties of laser welded-brazed joints between 1.9 mm 22MnB5 steel and 1.5 mm 6061 aluminum alloy. In addition, they used a Zn-Al₁₅ filler material. They show that the additional Zn creates new compositions of intermetallic phases. They detected Fe₂(Al, Zn)₅ and FeZn₁₀, which are more ductile than the conventional intermetallic phases produced without a Zn containing filler material. The thickness of the intermetallic phases was between 0.47 and 11.13 μm. They reached a max. tensile force of 2793 N.¹⁴ In the study of Windmann et al., 1.5 mm sheets of aluminum-coated high-strength 22MnB5 steel and 6016 aluminum were joined by laser beam welding-brazing. They used AlSi₃Mn as the filler metal and analyzed the resulting microstructure at different welding energies. They confirm that the thickness of the intermetallic phases increases at higher heat input. The growth of the intermetallic phases is influenced by the Si-containing filler material. They observed the formation of Al₈Fe₂Si and Al₁₃Fe₄ phases. The thickness ranged around 2–15 μm and resulted in maximum tensile strengths of 1900–2800 N with a wetting length of about 2.4 mm at a laser power of 3.3 kW.¹⁵ Yang et al. also investigated the formation of intermetallic phases and their influence on the load capacity of a lap joint of 1 mm galvanized DP980 steel and 1.5 mm 5754 aluminum alloy. They used Zn₂₂Al as the filler material. By forming a ductile reaction layer consisting of FeZn₁₀+ amorphous al-rich phase and Fe₂Al_5-xZn_x, the joint reaches a maximum tensile force of 1200 N, although the thickness of the intermetallic phases is between 20 and 45 μm. They observed two failure modes: the failure at the reaction layer and the failure within the fusion zone. The authors explain the failure of the fusion zone not by the thickness of the intermetallic phases but rather by the geometry of the fusion zone.¹⁶ All these investigations produced a load-bearing material-fit joint with or without a filler material. An additional form fit could increase the strength of the joints. To achieve a form fit, the filling of a conical cavity is necessary. To improve the lightweight balance and to save production costs, it is advantageous to use a joining process that does not require any filler material or auxiliary joining elements. Therefore, a controlled movement of the melt is needed. Liquid metals are electrically conductive media and can be influenced by magnetic fields during a welding process. The use of the electromagnetically melt pool influence could already be demonstrated for different applications. For example, when welding thick plates with filler wire, unsatisfactory mixing of the filler metal in the weld pool can occur. It results in a nonuniform element distribution in the weld seam. Üstündag et al. show that an oscillating magnetic field improves the element distribution over the whole thickness of a 20 mm thick laser beam welded S355J2 steel plate.¹⁷ Another example is the melt pool support to prevent drop out when welding thick sheets. Fritzsche et al. demonstrate this for a 20 mm thick ferromagnetic duplex and mild steel sheet.¹⁸ In the described examples, the magnetic field influences the melt pool flows. This offers a basis for a new approach to join dissimilar materials.

The new approach is schematically shown in Fig. 1. This joining method combines melt pool generation by a laser beam and a controlled melt pool displacement by an AC magnetic system. The joining partners are placed in an overlap configuration, whereby the upper joining partner must have a significant higher melting temperature and a conical cavity. The laser beam melts the lower joining partner through this cavity (Fig. 1 left side). The AC magnetic system is placed below the overlap configuration and induces upward directed electromagnetic forces (Lorentz force F_L). This Lorentz force causes a movement of the melt, so that the cavity of the upper joining partner is filled (Fig. 1 right side). After the displacement process, the laser beam is turned off. The moved melt solidifies in a material- and form-fitting joint.

With this method, the formation of intermetallic phases can be reduced because only one joining partner was melted and wets the solid joining partner. This approach can be used for spot- and line-shaped joints and needs no filler materials or auxiliary joining elements. In the work of Heßmann et al., more details of the displacement process can be read.¹⁹ They tested this new joining method to create spot-shaped joints between 2 mm 5754 aluminum alloy and 1 mm 1.0330 steel. They find out that an equilibrium is achieved between the downward and upward directed forces, which limits the height of the melt displacement. The forces are the hydrostatic pressure, the Laplace pressure, and the electromagnetic pressure. The formation of many spatters could be observed at high-speed camera recordings due to the known melt pool dynamic of an aluminum alloy during laser beam deep welding. As a result, a part of the molten material is lost, which sometimes leads to a cavity that is not completely filled. In some cases, cracks are visible in the transverse sections, which reduce the load-bearing capacity of the joints. They found a magnetic field power of about 300–2000 W to be sufficient for a displacement process. They also show that the displacement needs less time at higher electromagnetic forces compared to lower electromagnetic forces.¹⁹ 

This study tests the new approach for line-shaped joints. Based on the findings of the spot-shaped joints, in this case the process is spatially decoupled from the keyhole to optimize the displacement process. Thus, it is to be expected that the influence of the keyhole collapse can be reduced. The created linear joints were microstructurally and mechanically characterized.

The experimental set up is shown in Fig. 2. A 3 mm aluminum alloy (EN AW 5754) and a 1 mm steel sheet (1.0330) were used and placed in an overlap configuration, whereby the aluminum sheet was below the steel sheet. The measured chemical compositions of the used materials are shown in Table I. To displace the generated melt pool, a self-constructed AC magnetic system was placed 2 mm below the overlap configuration. More details of the used AC magnetic system and the calibration procedure can be read in the work of Heßmann et al.¹⁹ and Fritzsche et al.¹⁸ To limit the range of the magnetic field to 4 mm, a frequency of 3750 Hz was used. This reduced the interactions with the ferromagnetic steel.

The cavity of the steel sheet was 30° beveled, 3 mm wide, and 100 mm long. This conical cavity is important to create an additional form fit to the material fit. If the cavity is not conical, it is not possible to produce a joint due to the different thermal expansion coefficients of the joining partners. The sheets were fixed in a clamping system to ensure the technical zero gap. An IPG YLR 20 000 fiber laser was used, and the optics was tilted by 10° against the weld direction. The position of focus was 10 mm above the upper sheet, and argon with a flow rate of 20 l/min was used as the shielding gas. The Argon gas nozzle was tilted by 60°. The laser spot was positioned on the outer edge of the magnetic pole shoes to spatially decouple the keyhole and the displacement process. Details of the laser system and the used process parameters are listed in Table II. The laser parameters were evaluated by preliminary tests to ensure a stable reproducible melt pool. The magnetic field power of 670 W was selected based on the findings of Heßmann et al.¹⁹ The joining method was repeated twice. Three transversal sections of the created line joints were prepared to analyze the filling of the cavity (etching: 2% Nital and 2% NaOH). The grain size was also determined by the line intersection method. In addition, laser profile scans were carried out to evaluate the complete displacement height and quality of the line joints. A SEM LEO Electron Microscopy Inc. type 1530 VP was used to identify the intermetallic phases at the interface by electron backscatter diffraction (EBSD). Therefore, the specimen was tilted by 70°. The voltage of the electron beam was 20 kV, and the detector was tilted by 4°. The exposure time was 150 ms and the size of the analyte was 200 × 150 pixel. In addition, energy dispersive x-ray spectroscopy (EDX) measurements were carried out to identify the intermetallic phases (IMC). Hardness profile test according to Vickers (HV 0.01) was done to analyze the development of hardness in the area of the displaced aluminum melt and the steel sheet. Tensile shear tests were carried out for the magnetic field of 670 W to quantify the load capacity of the line joints. These five additional tensile specimens with a weld length of 40 mm and an overlap of 35 mm were produced and tested at a test speed of 1 mm/min. Before the tensile shear test, the specimens were x ray tested for internal cracks and pores in a nondestructive way (resolution, 25 μm). The local strains of the specimens were optically monitored by the GOM ARAMIS system and checked by an extensometer measurement. For this, the surface of the tensile shear specimens was sprayed with a black and white pattern. Figure 3 shows the top view and side view of an exemplary tensile shear specimen.

Figure 4 shows the top and bottom sides of the created line-shaped joint by photographs and laser profile scans. The top side of the joint has a uniform filling of the cavity without spatters and edge overlaps of the displaced melt. The laser profile scan of the top side shows a height of the displaced melt pool of about 0.5 mm over the upper edge of the steel sheet. The bottom side of the displaced melt is wavy shaped. The bottom side shows also a uniform displacement of the aluminum melt in the laser profile scan.

An exemplary transverse section is shown in Fig. 5. A magnetic field power of 670W is sufficient for a complete filling of the cavity. In some cases, some pores can be observed in the cross sections around the bottom side of the steel sheet. The pores could be air injections coming from the gap between the steel and the aluminum sheet. The air bubbles could not escape during the joining process. Rarely, cracks have been found. These cracks are very rare and small. Thus, the cracks are acceptable.

Compared to the spot joints investigated by Heßmann et al.,¹⁹ the joining process for line joints is significantly more stable and effective. It was possible to achieve significantly higher displacements with a low magnetic field power of 670 W compared to the spot joints. The melt was displaced minimally above the upper edge of the steel sheet. The influence of the melt pool dynamics was successfully reduced by spatial decoupling. The intermetallic phases were identified using EBSD- and EDX-analyses and are shown in Fig. 6. A compact phase seam at the interface toward the steel side has formed. This is followed by an acicular phase toward the aluminum side. The compact seam could be identified as the phase composition Al_5.6Fe₂ with about 73 at. % Al and 27 at. % Fe. The acicular phase consists of Al₁₃Fe₄ with about 83 at. % Al and 17 at. % Fe. The thickness of the compact phase seam is in average 5.7 ± 1.9 μm. The needle-shaped phases have an average length of 1.3 ± 0.6 μm. Locally, the formation of significantly different dimensions of IMC was observed. Overall, the formed intermetallic phases are below the critical value of 10 μm. In agreement with the findings from Refs. 6–8, a two-phase reaction layer was also formed in this process. Normally, the Al₅Fe₂ phase is detected. It is known from the work of Szczepaniak et al. that high cooling rates exist in the case of laser beam welding. Due to these high cooling rates, intermetallic phase compositions can form different from the equilibrium state.²⁰ Thus, in our study, a non-equilibrium phase of Al₅Fe₂ has been formed.

The results of the hardness profile measurement are shown in Fig. 7. The hardness of the steel is about 150 HV 0.01 and increases to about 200 HV 0.01 in the heat-affected zone (HAZ). The measured chemical composition of the used steel sheet of Table I shows that the steel contains a very low carbon content of 0.02 mass-%. So, martensite formation is not possible. The steel sheet is cold rolled. Thus, the steel already contains a high dislocation density. Furthermore, additional deformations were added by the production of the long conical cavity. During the displacement process, temperatures of up to about 1300 °C occur in the steel sheet.¹⁹ Thus, a recrystallization process is possible. Grain size analyses using line intersection methods have shown that the average grain size is reduced from about 17 μm to about 9 μm. The recrystallization has resulted in a very fine microstructure in the HAZ compared to the base metal, thereby increasing its hardness by grain boundary strengthening. The base metal aluminum has a hardness of about 60 HV 0.01. The fusion zone (FZ) shows hardness values of about 85 HV 0.01. In the SEM images, only at the interface, intermetallic phases (IMC) were found but not in the middle of the displaced aluminum melt. Thus, the increase in hardness can also be explained by a fine grain resulting from the solidification process. A typical dendritic structure with a dendrite size of about 4–6 μm has formed. The base material has an average grain size of about 12 μm. The highest hardness value of about 840 HV 0.01 was achieved directly at the interface due to the formed intermetallic phases. It agrees with the reported values of the phase composition of Al₅Fe₂ and Al₁₃Fe₄.¹¹ 

The force-elongation curves of the tensile shear tests are shown in Fig. 8. In the previously performed x-ray tests, no cracks or pores could be detected. The different curve profiles are caused by the locally differently formed thicknesses of the IMC. The dimension of the IMC has a significant influence on crack initiation and propagation. In the work of Windmann et al., also a significant range of thickness of the IMC (2–15 μm) could be observed and resulted in a significant variation of the maximum tensile strengths of 1900–2800 N.¹⁵ 

In our study, the line joints can bear a maximum force of 2300 N with an elongation of 2 mm. All the samples failed in the same way: The displaced aluminum melt did not fail or shear off during the entire tensile shear test. Also, the steel sheet did not break. At first, the material fit failed. The formed intermetallic phases at the interface are the weakest region in the joint due to their brittle properties. In addition, the intermetallic phases have a low ductility as well as a high crack sensitivity. The results of the hardness measurements show that the highest hardness in the joint is achieved in the IMC. The two different IMCs were identified as Al-rich phases.

These are known to be very brittle and sensitive for cracking. As a result, the material fit fails first under the maximum tensile force. The form fit still holds the sheets together. During tensile loading, the 40 mm long cavity becomes wider due to the elongation of the steel sheet and the end of the steel sheet bends upward. When the maximum elongation is reached, the form fit finally also breaks. The described steps of the failure of tensile shear specimen can be exemplarily seen in Fig. 9.

The measured tensile shear values correspond to the values reported by Wang et al. for welding-brazing without the use of a filler metal.⁸ However, the authors did not specify the tested joint length. The results of this work cannot be compared exactly with the values of the literature because in other studies filler materials or other base materials were used. The tested specimens in this research without the use of a filler material show that it is possible to produce a similarly strong joint only by a controlled melt pool displacement.

In this study, a new joining approach for line-shaped joints was tested. The spatial decoupling between the displacement process and the keyhole dynamic improves the melt pool displacement process. The microstructure and the mechanical properties of the created line joints were analyzed. The results can be summarized as follows:

• With a magnetic field power of 670 W, a uniform aluminum melt displacement is possible so that the cavity is filled completely.

• Compared to previous studies on the creation of spot joints, the displacement process is more stable and effective. Higher displacement heights are possible.

• A two-phase reaction layer was detected, and it consists of a 5.7 μm compact phase seam on the steel side and a following needle-shaped structure toward the aluminum side. The needles are on average 1.3 μm long and consist of the Al₁₃Fe₄ phase. The compact phase seam was detected as Al_5.6Fe₂. The thickness of the intermetallic phases is below the critical value of 10 μm.

• The hardness values compared to the base metals increase in the HAZ and FZ. In the HAZ, a very fine grain structure with an average grain size of 9 μm is formed. The FZ consists of a cast structure with a size of dendrites of about 4–6 μm.

• Tensile shear tests show a maximum load capacity of 2300 N. The first failure starts in the material fit due to the brittle material properties of the intermetallic phases, where cracks can easily form and grow. Due to the elongation of the steel sheet, the cavity widens, so that finally the form fit also fails.

The authors have no conflicts to disclose.

Jennifer Heßmann: Investigation (equal); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal). Kai Hilgenberg: Writing – review & editing (equal).