Joining Dissimilar Aluminum Alloys by Resistance Spot Welding for Aluminium-Intensive Mixed Constructions

An EN AW 6082-T6 and an EN AW 5083 H111 in 2 mm are used as wrought aluminium alloy. The joining partners are the heat-treatable casting alloys EN AC AlSi7MnMg and EN AC AlSi10MnMg T6/T7, and the naturally hardened alloy EN AC AlSi9Mn. The aluminium casting alloys have a product thickness of approximately 3 mm. The test materials differ in terms of their chemical composition. Table 2 shows the different alloy compositions of the test materials according to the manufacturer's specifications.

2.1 Welding tests

A mid-frequency direct current (MFDC) welding machine in C-type-design from Nimak with servo-motoric force unit and electromagnetic follow-up unit (magneticDrive) is used for the welding tests. The nominal frequency of the inverter is 1000 Hz. Electrode caps of the type A0 20 R100 made of CuCr1Zr are used. [3]. These are milled on the system to the desired working surface radius (100 mm). To utilize the heat field displacement in the direction of the anode resulting from the Peltier effect, the rolled sheet is aligned on the anode side for all of the experiments presented here.

The welding current profile selected at the beginning is the standard welding profile according to VDA 238 401 [18] with a constant electrode force of 7.5 kN, a 100 ms long 8 kA strong pre-current, followed by a 150 ms long main current pulse with a current strength between 20 kA and 65 kA to achieve critical weld spot diameters. The critical weld spot diameters are d_w,min = 4√t and d_w,nom = 5√t. In this context, I_min is the current strength at which d_w,min is achieved and I_nom is the current strength for d_w,nom. [18] Table 3 lists the I_min. and I_nom. currents determined for the various MDKs.

3 Results and Discussion

In the following, the selected conventional approach according to VDA 238 401 [18] is compared with an alternative approach based on the previously obtained results, in which, in particular, the electrode force profile was modified in a targeted manner.

3.1 Conventional Approach

The weld lens geometry generated by RSW results from the physical properties of the joining partners involved in the material combination to be joined. Similar to the joining of dissimilar joints between austenitic and ferritic steels by RSW, the weld depth that can be achieved in the respective workpieces of the aluminium dissimilar joints correlates with the respective thermal conductivity of the joining partner. Figure 1 and Figure 2 show cross-sections of example joints (left) and the welding depth p determined in the cross-section in both joining partners for the different MDKs depending on the welding current strength I_H, Table 3, with sample size n = 3.

Due to the high thermal conductivity of EN AW 6082 (λ_{(EN AW 6082)}: 170 – 220 W/(m∙K)), the weld nugget is clearly shifted into the casting material in the case of MDK 1, MDK 3 and MDK 5 (e.g. λ_{(EN AC AlSi10MnMg)}: approx. 150 W/(m∙K), Figure 1). In contrast, for MDKs with EN AW 5083, the weld nuggets at I_H = I_nom are almost symmetrical, Figure 2. Here, the thermal conductivities are closer (λ_{(EN AW 5083)}: approx. 140 W/(m∙K)). At I_H = I_min, the penetration depths p_{EN AW 6082} and p_{EN AW 5083} determined in the cross-section are not above the recommended penetration depth of at least 20 % (p_nom = 0.4 mm, Figure 1, Figure 2). Furthermore, the weld nuggets have a porosity in the center, which is outside the joining plane in the case of MDK 5 due to the localization of the weld lens. In the case of MDK 11, on the other hand, the porosity is within the joining plane,Figure 1, Figure 2. Figure 3 shows SEM images of the volume defects detected in the center of the welding lens, using the example of MDK 5 and MDK 9.

The detected porosity is mainly due to solidification, in which intact dendrites protrude into the the cavities. In MDKs with EN AW 5083, isolated pores with an approximately spherulitic shape and a flat surface can also be detected (Figure 3, right). This pore shape is typical of gas-related porosity. However, due to the cracks present in the pores, initial gas porosity is assumed, which is dominated by solidification porosity as solidification progresses due to the strong shrinkage of the melt.

Another significant difference between MDKs is the formation of cracks in the rolled plate-side HAZ. In the case of MDKs with EN AW 6082, these cracks are filled with material from the weld nugget. Figure 4 shows the results of an EDX mapping on MDK 5 with a higher silicon content of the material solidified in the crack lumen compared to the rolled alloy.

The cracks can be classified as liquation cracks. According to the current state of technology, this type of hot crack occurs as a result of melting in the presence of low-melting phases at the grain boundaries. These can occur with the additional local occurrence of thermomechanically induced tensile stresses. When in contact with the melt pool, the widened grain boundaries allow liquid melt to flow in, where it solidifies dendritically [6; 11]. In the case of MDKs with EN AW 5083, the corresponding cracks are not present to the same extent, Figure 2. Figure 5 shows the shear and cross tension forces determined for the different MDKs in quasi-static shear tensile test (STT) and quasi-static cross tension test (CTT).

In particular, the maximum shear forces are up to 20 % higher for I_H = I_nom than the minimum shear forces recommended in DIN EN ISO 18595. By contrast, for I_H = I_min, only MDK 5, MDK 9, MDK 10 and MDK 11 are above the minimum value of 4.84 kN, or 4.68 kN, Figure 5. The influence of the MDK on the joint strength under quasi-static cross tension loading is to be classified as marginal, apart from the larger standard deviation observed for MDK with EN AW 5083. The tendency for the cross tension force to decrease with increasing Si content of the cast alloy for MDKs with EN AW 5083 is not observed for MDKs with EN AW 6082. The different weld nugget geometries result in differences in failure behavior under quasi-static load. Figure 6 shows the different types of fracture and the characteristics of the heat-affected zones (HAZ) on the rolled sheet side, using MDK 5 and MDK 9 as examples.

For MDK 1, MDK 3 and MDK 5, the failure in STT occurs exclusively along the hot-rolled sheet-side fusion line (lens interface fracture) due to the low weld penetration depth on the hot-rolled sheet side. MDK 9, MDK 10 and MDK 11, on the other hand, fail with different types of fracture, Figure 6. In addition to shear fractures running through the weld nugget along the joining plane, nugget interface fractures occur with failure along the fusion line, both on the rolled sheet side and on the cast part side, as well as a mixed form of shear failure and nugget interface fracture on the casting side. In CTT, button pull-outs occur throughout the rolled sheet in MDKs with EN AW 6082. In contrast, the fracture behavior in MDKs with EN AW 5083 is again alternating. Button pull-out fractures from one of the joining partners, both cast-side and rolled-sheet-side fractures along the fusion line, shear fractures with a fracture path through the nugget and mixed fractures occur. In case of mixed fractures, the failure begins along the rolled-sheet-side fusion line and transitions into a button pull-out fracture (residual fracture). As a result of the more even heat distribution that occurs in MDKs with EN AW 5083 when the heat conductivities of the joining partners are close together, a wide PMZ (partially melted zone) with signs of grain boundary melting forms on the rolled sheet side Figure 6. These are indicative of local process temperatures between the solidus temperature and liquidus temperature, i.e. above the melting temperature of low-melting phases located at the grain boundaries. The resulting cracks can be linked to the failure in the heat-affected zone near the fusion line. The strength of the base material and the weld bead exceeds the strength of the HAZ, where the cracks resulting from the melting of the grain boundaries further weaken the material cohesion.

In MDKs with EN AW 6082, on the other hand, this phenomenon is less pronounced, which, in addition to the alloy composition, is mainly due to the heat field being strongly displaced into the casting, whereby the melting temperatures of the phases located at the grain boundaries are only exceeded in a locally limited area. The failure of the MDKs with EN AW-6082 along the fusion line is due to the low welding depth on the rolled sheet side and the resulting stress increase (strength jump at the fusion line) positioned close to the main load plane. A direct influence of the filled cracks occurring in the HAZ of the EN AW 6082, Figure 4, on the mechanical properties under quasi-static loading could not be determined. Based on the results, an optimization problem can be derived in which the need for an overall lower heat input into the joints can be determined. On the other hand, a reduction in current strength is likely to result in subcritical welding depths. In addition, the need for higher welding depths on the rolled sheet side is offset by the resulting shift in the solidification porosity into the joining plane, which in turn increases its significance for the failure of the joint.

3.2 Modification of the electrode profile

To achieve an increase in the welding depth while simultaneously reducing solidification porosity, the electrode force profile is modified based on the process control used in the die casting process with a final repressing phase through a dynamic increase in force. Figure 7 shows the formation of welding lenses documented in cross sections during current shut-off experiments with d_n and A_n as a function of the welding current time t_i when applying a constant electrode force of F = 7.5 kN.

At the beginning of the welding process, the weld pool is only present in the cast workpiece and grows in the direction of the cast part thickness with increasing welding time. As the welding time increases, the welding depth into the rolled sheet increases, too, Figure 7. Liquation cracks can already be determined at an early stage of the RSW process (t_i = 40 ms). The measurement of the weld nugget diameter d_n and the weld nugget area A_n in the cross-section analysis shows a linear increase for A_n with a logarithmic increase for d_n as the welding time ti increases. Due to the weld nugget porosity detectable in the cross-section from t_i ≥ 80 ms, an increase in force from t_i = 60 ms and after switching off the welding current, i.e. for compression, is selected. The previously determined time to realize an electrode force increase of 4 kN is 250 ms for the system used.

Figure 8 shows the achievable shear forces depending on the selected welded profiles in comparison to VDA 238-401. Figure 9 shows schematically the most expedient modifications of the electrode force profile (left) and the resulting welding lens shape, including the fracture types achieved after quasi-static shear loading (right).

While the use of a force increase from 7.5 kN to 9 kN after switching off the welding current has no significant influence on the mean shear tensile force F_{S, max,} this modification already leads to a significant reduction in the standard deviation, Figure 8. In the cross-section, a significant reduction in porosity can also be seen in the center of the welding lens, although no change in fracture behavior can be seen, Figure 9. With regard to the reduced scatter of measured values, it can be assumed that the repressing results in a more uniform hardening of the HAZ on the rolled sheet side.

On the other hand, greater welding depths in the rolled sheet can only be achieved by lowering the initial electrode force from 7.5 kN to 5 kN and initiating the increase in electrode force (to 9 kN) at t_i1 = 60 ms. The resulting changes in contact conditions and the higher contact resistances at the beginning of the welding process enable a greater heat generation and thus greater welding depths. At the same time, the compressive forces exerted on the joint as a result of the dynamic increase in force counteract the tensile forces caused by solidification (due to the high shrinkage of aluminum). As a result, weld nuggets with sufficiently high penetration depths and reduced solidification porosity are formed. Although these can withstand comparatively lower shear tensile forces, they still exceed the minimum forces required by DIN EN ISO 18595 and lower standard deviations can be achieved, Figure 8 and Figure 9. A pull-out fracture in the wrought aluminum sheet does not occur.

4 Summary and outlook

The research project showed that the currently available recommendations for welding parameters for the RSW of dissimilar joints between aluminium die-castings and aluminium wrought alloys are only partially suitable due to the different properties of the alloys. The weld nuggets have an atypical geometry with centrally located solidification porosity. At a constant electrode force of 7.5 kN, the greater the difference between the thermal conductivities of the joining partners, the more the weld lens is displaced into the joining partner with the higher thermal conductivity. Consequently, for MDKs with EN AW 5083, a more symmetrical weld nugget with greater welding depth into the wrought aluminium results.

In the HAZ on the wrought aluminum side, MDKs with EN AW 5083 primarily show cracks resulting from the melting of the phases near the grain boundaries, which subject the mechanical behavior and the type of fracture with which a connection fails under quasi-static load to statistical scattering. In MDKs with EN AW 6082, liquation cracks dominate in the HAZ on the rolled sheet side, which are filled with solidified material from the weld nugget. No significant influence of the cracks on the mechanical connection properties could be demonstrated in the tests. The shear tension forces that can be achieved are up to 20 % higher than the minimum values recommended in actual standards. Nevertheless, the weld nuggets do not meet the current geometric requirements, particularly with regard to the welding depth in the rolled sheet and the occurrence of cracks in the HAZ.

A key factor influencing the quality of the weld spot is the choice of electrode force profile. The use of electrode force profiles with a dynamic force increase offers the possibility of significantly improving both the welding depth and the reproducibility of the spot weld quality and of avoiding the volume defects in the center of the nuggets that result from volume contraction during solidification. Consequently, RSW-joined dissimilar joints between modern, dissimilar aluminium alloys show great potential for industrial applications. However, a key factor for the feasibility of corresponding process strategies with dynamic increases in electrode force in the future is the capability and modernization of the system technology to be used, which is currently limiting, particularly with regard to rising times of the electrode force.

(Source: Welding And Cutting 2/2025 DOI: https://doi.org/10.53192/WAC202502139)