Fatigue Strength Assessment of Laser Beam Welded Joints Made of AA7075 and Magnetic Pulse Welded Joints Made of AA7075 and 3D-Printed AlSi10Mg

: laser beam and magnetic pulse welding. Herein, laser beam welding is successfully used to manufacture a roll-formed and longitudinally welded pipe made of AA7075 and joined by magnetic pulse welding with a 3D-printed lug-tube made of AlSi10Mg. The fatigue strength of these pipe joints and of laser beam welded butt joint specimens is determined using load-controlled fatigue tests. For the characterization of the specimens, cross sections are prepared and examined metallographically, which re ﬂ ect the local weld seam geometry in the joining area. A fatigue assessment is made using linear-elastic approaches. The reference radius concept is applied to map the in ﬂ uence of geometric notches on the fatigue strength, assuming linear-elastic stress – strain behavior. It is shown that the recommended notch stress fatigue class FAT 178 (von Mises stress) can be applied for a safe and reliable fatigue assessment.

mentioned primarily address the problem of instability in the keyhole due to fluctuating vapor pressures. The formation of hot cracks during LBW of AA7075 due to the alloying element copper can only be insufficiently influenced. Modern laser beam sources or mirror-optical manipulations of the laser beam to adjust the intensity of the beam in the welding area dominate recent developments for process stabilization. For example, Dittrich et al. [15] were able to significantly reduce instabilities in the keyhole and at the same time significantly reduce hot cracking in their investigations on AA7075 by using scanner optics with high beam deflection (up to 4 kHz). Another approach, which also aims to minimize the tendency to hot cracking, is described in the work of Norman et al. [16] and Sokoluk et al. [17] In an alternative to the approaches described so far, the weldability is not improved via the laser beam, but via a metallurgical approach by using special filler materials. In Norman's work, scandium was added to the fusion zone via the filler material, resulting in an extremely fine-grained grain structure (%10 μm). In his investigations, Sokoluk used a filler wire made of AA7075, which was doped with titanium carbide. Here, too, the result is an extremely fine-grained structure.
In general, the fusion welding of aluminum is much more unstable than the welding of steel due to its low viscosity and surface tension. Therefore, pressure welding processes, such as friction stir welding (FSW), are a promising alternative to the classic fusion welding processes. The reason for this is the comparatively lower welding temperature (480°C), and consequently the nonexistence of a molten phase, which results in a higher weld quality compared to fusion welding. Detailed investigations on the influence of various process parameters on the obtainable weld seam qualities during FSW of AA7075 can be found, for example, in the work of Rajakumar et al. [18] Magnet pulse welding (MPW), as a representative of the pressure welding process, is characterized by the capability of forming joints without the input of external thermal energy. In MPW, a material-locking joint is created by a collision of two joining partners. [19,20] The collision is initiated by an MPW tool coil, which is passed through by a high-frequency electric current pulse. Consequently, a primary magnetic field is generated. Now, if an electrically well-conducting joining partner (flyer) is positioned near this magnetic field, eddy currents are induced in it due to this magnetic field formation. In turn, these eddy currents generate a second magnetic field, which is oppositely directed to the primary magnetic field. Electromagnetic interactions of both magnetic fields now result in a force acting on the flyer and accelerating it away from the tool coil until it collides with positioned second static joining partner (target). [19] During the collision, extreme conditions such as high pressures and very high temperatures prevail in the collision area, which causes the boundary layers of both joining partners to be brought into a plastic viscous state. [21,22] Furthermore, components of both joining partners, base material, oxide layer, and impurities, together with elements of the ambient atmosphere are ejected from the joining area as a mass flow, called jet. The whole collision process in MPW is characterized by unsteady collision velocities and collision angles, which must be within specific ranges for sound welds. [21] These process characteristics of MPW are the reason why high-strength dissimilar joints are realizable. MPW technology can be applied to three joining geometries: sheet metal forming, pipe expansion, and pipe compression. [19] In this study, pipe compression was used. Nevertheless, there are limitations regarding the weldability of some materials using MPW. High-strength aluminum alloys used as flyers challenge the MPW process. Consequently, most investigations focus aluminum alloys of the 1000, 5000, and 6000 series. [21,23,24] Sufficient joint strengths, including good long-term properties, were obtained. [23] The MPW of high strength 7000 series aluminum was investigated by Stern [25] and Pourabbas. [26] Main findings were that MPW of AA7075 is possible-both similar AA7075 joints [25] and dissimilar AA7075 (target) with AA4014 (flyer) joints are feasible. [26] MPW-induced interface hardening has been shown and the interface showed MPW typical characteristics such as wave-like structures including melt pockets. However, previous studies insufficiently address the material characteristics of the AA7075 base material and the joint characteristics of the welded joint. MPW of tube-to-tube joining partners in the context of AA7075 is also uninvestigated, as previous studies analyzed tube-to-rod welds. This is advantageous for sufficient MPW because, due to this welding geometry, no backing is required. [25,26] In recent years, additive manufacturing (AM) processes have been continuously developed and improved. The laser powder bed fusion (LPBF), also known as selective laser melting (SLM), is a powder-based AM process for components made of metal, which offers a high design flexibility and near-net-shape parts up to 99.9% relative density. [27] Nevertheless, porosity is an important influence on the mechanical properties of welded structures, which are based on Al-Si powders and formed in the SLM process. It was shown that LBW of similar AlSi10Mg joints by Mäkikangas et al. [28] led to increased porosity and pore diameter up to 0.7 mm. Nahmany et al. used electron beam welding (EBW) to join SLM-built AlSi10Mg specimens. The porosity in the weld metal tended to decrease with smaller weld speeds and with conductive welding mode instead of the key-hole welding mode. The weld seams welded the slowest showed almost no porosity. It was also stated that the weld metal's porosity was dependent on the base metal porosity. Moeini et al. studied the microstructure and mechanical behavior of welded butt joints made by SLM-built AlSi12 using FSW. [29] Hardness and tensile strength were lower for FSW parts compared to the AlSi12 base material, which coincided with the dissolution and rearrangement of secondary Si-rich phases. The pore size was reduced by severe plastic deformation by the FSW process. Low cycle fatigue (LCF) tests revealed that especially for higher strain amplitudes of 0.6 % the fatigue lives of the FSW specimens were remarkably lower, which was also attributed to secondary phases. Crack initiation at surface-near pores was observed in both SLM-built and FSW specimens. Further, Moeini et al. studied the influence of different build orientations of FSW SLM-built AlSi10Mg butt joints. [30] The ultimate tensile strengths of FSW parts were about 25% reduced compared to the SLM-built condition. Generally, FSW is not suitable for joining of pipes because a counter holder is always needed and accessibility through the tool is often insufficient.
The structural integration of AM structures into conventional sheet metal structures is also the subject of research. [31][32][33][34][35] In addition to high-strength aluminum alloys, additively manufactured joining partners are a research area of MPW. Shribman [36] showed that a defect-free joint between wrought AA6060-T6 and 3D-printed AlSi10Mg was created. The interface showed MPW typical characteristics like melting pockets and a conducted leaking test revealed low leaking rates. The analysis of MPW joints of advanced wrought aluminum alloys with AM materials as well as mechanical characterizations of these joints, however, require further research. LBW of dissimilar aluminum joints made of additively manufactured AlSi10Mg and AA6082 T6 sheet metal resulted in a high porosity up to 16 % and weld undercuts that negatively affect the fatigue strength. [33] In ref. [35] the influence of metallurgical defects on the fatigue strength of AlSi10Mg alloy specimens produced by LPBF was investigated. It was found that the near-surface defects serve preferentially as crack initiation locations in almost all specimens tested, and that large oblate defects are oriented in the build plane, leading to a strongly anisotropic fatigue strength.
In the work of Mucci et al., [37] the fatigue strength of LBW and FSW overlap specimens made of AA7075 was investigated experimentally. The evaluation based on the notch stress approach, reference radius concept with r ref = 0.05 mm, leads to conservative results with reference to the notch stress fatigue class FAT 160. Based on the results obtained in this work with the consideration of additional literature data of overlap and butt joint specimens, the fatigue strength can be evaluated more reliably using the effective stress approach. The fatigue assessment of Electro-Magnetic Pulse Technology (EMPT) welded joints using the reference radius concept was investigated by Baumgartner et al. [38] A direct comparison between EMPT and conventional welded joints shows no significant difference. This indicates that the fatigue strength of EMPT aluminum welded joints can be estimated using approved fatigue assessments.
A reliable fatigue strength assessment of LBW as well as MPW joints made of AA7075 is essential to use the two technologies in future lightweight design concepts. For this reason, the extent to which the recommended notch stress fatigue classes can be used for a reliable fatigue strength assessment is investigated. In this study, LBW was successfully used to manufacture butt joint specimens and longitudinally welded roll-formed pipes made of AA7075. The pipes then were joined by MPW with 3D-printed lug-tubes made of AlSi10Mg. Both types of specimens were experimentally investigated under axial loading in fatigue tests to compare their fatigue strength potential compared to literature data. For that finite element (FE) models of the butt joint specimens were created based on measured cross-sections, while an idealized FE model based on a 3D-scan was created for the pipe joint specimen. An evaluation is made using the reference radius concept with r ref = 0.05 mm to map the influence of geometric notches on the fatigue strength, assuming linear-elastic stress-strain behavior.

LBW
High-strength aluminum sheets made of AA7075 in T6 condition were used for the investigations. The sheet dimensions were 125 mm Â 250 mm with a material thickness of t = 1.5 mm. According to the test certificate, the maximum tensile strength of the plates used is R m = 591 MPa. The main alloy components of the high-strength aluminum alloy AA7075 are zinc, magnesium, and copper. As already explained, the alloy AA7075 is not qualified for welding due to the mentioned main alloying elements, which is shown in the form of strong instabilities during the LBW process as well as in a strong softening of the weld metal and the heat-affected zone. Furthermore, due to the high copper content, the alloy tends to develop hot cracks during the welding process.
The formation of hot cracks can be prevented by a low welding speed (≤5 m min À1 ). The instabilities in the vapor capillary can be reduced by a correspondingly strong defocusing (≥ þ 8 mm) of the laser beam from the workpiece surface, but this leads to a widening of the weld seam and consequently a strong weld seam droop. To counteract both the weld seam droop and the strong softening, two different filler materials were used in the scope of the investigations. One is a filler metal available on the market with the type description "S Al 5087 (AlMg4,5MnZr)" according to EN ISO 18 273. [39] In addition, an experimental filler material made of AA7075 was used, which contains titanium carbide and will therefore be referred to as "S Al 7075TiC" in the following. Both filler materials are spooled on a 300 mm basket coil and have a wire diameter of 1.2 mm. The welding wire is fed into the process by means of a cold wire feeder (FD101 LS, DINSE GmbH, Norderstedt, Germany). The exact chemical composition of the two filler metals used can be found in Table 1.
A 10 kW high-power fiber laser (IPG YLS-10 000-S4, IPG Laser GmbH, Burbach, Germany) with a wavelength of 1070 nm and a top-hat intensity profile was used for the LBW tests. The generated laser beam was transferred via a laser light guide cable with a core diameter of 200 μm to a fixed focal length optical system (Reis Lasertec MWO 54, KUKA Industries GmbH & Co. KG, Aachen, Germany). This has a fixed focal length of 300 mm and an imaging ratio of 2:1, resulting in an achievable focus diameter of 400 μm.
First, LBW tests were carried out on sheets in the butt joint. The basic test setup for welding tests in the butt joint is shown in Figure 1A. To minimize the risk of weld seam imperfections due to contaminated sheets or joint gaps, the joint edges were milled off prior to the welding tests and cleaned with acetone immediately before LBW. In order to protect the liquid melt from the ambient air, the inert shielding gas argon with a purity level of 4.6 was used and coaxially guided into the process via the cold wire feeder. The welding optics were set at an angle of 6°with relation to the sheet surface, on the one hand, to protect the optics from radiation reflection and, on the other hand, to enable a sluggish welding arrangement. The wire feeder was positioned leading with an angle between 25°and 30°to the workpiece surface (cf., Figure 1B).
After carrying out LBW tests on aluminum sheets in the butt joint, the test results were transferred to roll-formed aluminum sheets, which were also welded in the butt joint via a longitudinal weld seam to form a pipe. A device was developed and integrated into the experimental setup to pretension and position the roll-formed sheets for LBW (cf., Figure 1C).
In the context of preliminary tests in the form of a small parameter study, the optimum welding parameters were determined for both welding filler materials for the joining application at hand. The laser power was modified depending on the focus position and wire feed speed. The angle of beam incidence as well as the welding speed itself was kept constant during the parameter study. The final parameter sets selected for the two filler metals are listed in Table 2.
The analysis of the generated LBW joints regarding weld seam geometry and microstructure was carried out based on cross sections using a reflected light microscope (DM 2700, Leica Microsystems GmbH, Wetzlar, Germany). For better contrast, the cross sections were prepared metallographically in advance and etched using Keller's etching medium. When comparing the microstructure between the weld metal with the standard filler metal S Al 5087 and the experimental filler metal S Al 7075TiC, there is a clearly visible difference in the grain structure (cf., Figure 2).
In particular, the extremely fine-grained structure of the experimental filler metal S Al 7075TiC (cf., Figure 2B), which can be explained by the addition of titanium carbide, can be expected to improve the mechanical-technological material properties. The formation of hot cracks can also be prevented by such a finegrained weld structure. Based on the microscopic analysis, no significant amount of pores could be identified that could negatively influence the fatigue strength. For this reason, their influence on the fatigue strength is not discussed in this article.
For the fatigue tests, LBW butt joint specimens with (LBW-MetaLi) and without filler material (LBW-batch-1) were used. S Al 7075TiC was used as supplementary filler wire. For both batches, a defocused laser beam was used. The specimens were then removed by waterjet cutting and the cut edges were subsequently machined to avoid crack initiation starting from the edge region ( Figure 3).

MPW
For the fatigue tests of MPW joints, a pipe joint specimen was developed, which was manufactured from a roll-formed and laser beam longitudinally welded pipe made of AA7075 and a 3D-printed lug-tube made of AlSi10Mg using MPW process. The pipe was used as flyers and the lug-tube as target. As for the butt joint specimens, a defocused laser beam and a filler wire made of S Al 7075TiC were used for the longitudinal welding ( Figure 4).
For the manufacturing, a PS 48-16 MPW system of the manufacturer PSTproducts GmbH was used. This system allows the application of 48 kJ discharge energy. For the welding operation 45 kJ were applied and the gap between the AA7075 pipe (flyer) and the AlSi10Mg lug-tube (target) was 1.5 mm due to the given geometry of the flyer and target (see Figure 5). The edge of the flyer was positioned centrally in the MPW circular coil to force a one-sided collision motion. No backing was used, so the target had to withstand the MPW collision. Two alternatives were considered for the manufacturing of the specimens. Specimens were produced with and without a centering ring. This centering ring is used for adjusting the joint gap prior MPW because the joining partners are not ideally straight and problems regarding a uniform joint gap were expected. It should be noted that the centering ring affects the kinematic MPW process variables (collision angle and collision velocity) during the collision process between flyer and target. This occurs as the centering ring limits the pipe length available for deformation (see Figure 5), which directly limits the maximum length of the weld seam. Effects on the collision angle are as well to be expected. The weld length is particularly important because it directly affects the size of the joined area and therefore stresses that occur during loading. Regardless,  www.advancedsciencenews.com www.aem-journal.com welds could be achieved both with and without centering ring by means of MPW.
To analyze how the weld formation occurred between the AA7075 pipe and the 3D-printed lug-tube, cross sections at six locations were utilized to conclude the overall weld seam formation. The position of the cross sections and the weld lengths obtained can be seen in Figure 6. Potentially, there are three weak locations: the LBW seam (pos. 1) and both positions of the field shaper cut because lower electromagnetic forces act on the AA7075 pipe here (pos. 3 and 7). Positions 6 and 8 were not analyzed using cross sections because they are symmetrical to 2 and 4, respectively.
The determined weld lengths show that the MPW seam was poorest or not formed at all at the three potential weak points. Accordingly, the MPW of the LBW-welded AA7075 pipe with the 3D-printed lug-tube is not a fully circumferential welded joint.
A cross section of the MPW joint is shown in Figure 7 (position 4 of Figure 6), which reveals typical MPW characteristics. At the beginning of the weld seam, a rather laminar interface is   www.advancedsciencenews.com www.aem-journal.com present. Later, it becomes wavier with local pore inclusions. Toward the end of the weld seam, a large porous pocket is present. This could indicate that the joint-forming MPW collision overruns and traps the jet. As the jet is extremely hot and is rapidly cooling when rolled over, the very porous microstructure is formed.   www.advancedsciencenews.com www.aem-journal.com

Butt Joint Specimen
Load-controlled fatigue tests were performed using the butt joint specimens with constant load amplitudes F a and a load ratio of R F = 0.1 until total fracture. The tests were conducted with a test frequency of f = 50 Hz and at room temperature. The results are plotted in a S-N curve diagram (Woehler diagram) (Figure 8). For this purpose, the nominal stress σ N was calculated in each case in relation to the smallest square specimen cross section, measured with a caliper gauge. The maximum likelihood method was used to statistically evaluate the S-N curves. [40] Compared to the specimens without filler material (LBWbatch-1), specimens with filler material (LBW-MetaLi) show a higher fatigue strength by a factor of 1.1 at cycles to fracture of N f = 2·10 4 and a flatter slope of k = 7. At cycles to fracture of N f = 2 Â 10 6 , the factor is 2.1. Specimens without filler material have a slope of k = 3 and an identified knee point at N k = 4 Â 10 5 . Both S-N curves show approximately identical scatter T S . Due to the limited number of specimens with filler material, no knee point could be identified. All specimens without filler material show specimen fractures in the weld seam center. This can also be observed predominantly in the specimens with filler material. Three specimens show an off-center fracture at the weld seam notch. These specimens also tend to show higher fatigue strength (Figure 9).

Pipe Joint Specimen
The fatigue strength potential was investigated experimentally and numerically under cyclic axial loading using the developed pipe joint specimen (Figure 4). For this purpose, load-controlled axial fatigue tests were performed until fracture, corresponding to 1 mm increase in cylinder displacement amplitude x a , at a load ratio of R F = 0, constant force amplitudes F a , a test frequency of f = 10-35 Hz, and at room temperature. The results are plotted in a F-N curve diagram (Woehler diagram) ( Figure 10). The evaluation of the F-N curves was performed according to the maximum likelihood method. [40] Compared to the pipe joint specimens with centering ring, the specimens without centering ring show higher fatigue strength by a factor of 1.25 at N f = 1 Â 10 6 , with comparable slope of k = 3.8 and scatter of T F = 1:1.1 of the F-N curves. Due to the limited number of specimens, no clear knee point N k could be identified and was therefore fixed at N k = 1 Â 10 6 for both batches. Crack initiation was always found in the additively manufactured lug-tube made of AlSi10Mg. It is assumed that this is due to the lower fatigue strength compared to AA7075 and different degrees of residual stresses in the joining area, caused by the joining process. In a few exceptions, additional failure occurred in the lug-tube in the restraint area at low force amplitudes F a and, in the case of very high force amplitudes F a , additional failure occurred in the pipe ( Figure 11).

FEA
To characterize the pipe joint and butt joint specimens, cross sections were made and analyzed metallographically, which determined the local weld seam geometry in the joining area. Based on the cross sections, 3D-CAD models were created with reference radii of r ref = 0.05 mm modeled in the relevant notches as recommended by Sonsino. [41] These models were then transformed into FE models. Elements with a quadratic approach function (Abaqus 2020: C3D20) and linear-elastic material behavior were used. A Young's modulus of E = 70 GPa and a Poisson's ratio of ν = 0.34 were applied for aluminum. The procedure for meshing of the weld notches was based on ref. [42] to ensure that the convergence rate deviation of the simulation results is less than 2%.

Butt Joint Specimen
The use of the filler metal resulted in an improvement of the weld seam geometry, which was manifested in reduced weld seam root fallback ( Figure 12). The boundary conditions in the FE model result from the test setup of the fatigue tests. The axial load is applied via a reference point in the center of the left clamping surface. The fixed boundary conditions, on the other hand, are applied via a reference point in the center of the right clamping surface. Both reference points are connected to the clamping surfaces with rigid elements (Abaqus 2020: kinematic coupling) ( Figure 12). The strain comparison between experimentally determined and simulated strains shows high agreement. The maximum local notch stress σ vM,max or notch stress amplitude σ vM,a was determined for the butt joint specimens with and  www.advancedsciencenews.com www.aem-journal.com without filler material. This results in a stress concentration factor of K t = 4.3 for the specimens without filler material and K t = 2.7 for specimens with filler material, which is due to the local weld geometry (Figure 12).

Pipe Joint Specimen
Using high-resolution 3D-scans and subsequent "reverse engineering" of a pipe joint specimen, a detailed 3D-CAD model was built to serve as a realistic representation of the pipe specimen ( Figure 13). Due to the symmetric shape of the pipe joint specimen and to save calculations time, a FE half model was used. For this purpose, symmetry boundary conditions were set. Two reference points were modeled on the center axis for the axial loading and the fixed boundary conditions, each located in the middle of the clamping surfaces, which are connected to the clamping surfaces with rigid elements (Abaqus 2020: kinematic coupling) ( Figure 13). Two competing hot spots were identified in the analysis of the simulation results ( Figure 14). The maximum local notch stress σ vM,max was determined in notch Z on the pipe side. In contrast, the local notch stress on the lug-tube side in notch Y shows lower stresses by a factor of about 1.4. Despite higher local notch stresses in the pipe, cracks occur in the lug-tube. It is assumed that this is due to varying degrees of residual stress in the joining area, which were not considered in the simulation.

Fatigue Strength Assessments
The calculated local stresses were subsequently scaled with the load amplitudes F a of the fatigue tests and plotted in a S-N curve diagram (Woehler diagram, stress-based) ( Figure 15). In addition, a reference S-N curve from ref. [32] has been included, which was evaluated for overlap specimens between AlSi10Mg and AA6082-T6 joined by LBW. Figure 15 also shows results from tests on LBW overlap joints made of similar aluminum wrought alloys from refs. [43] resp. [44] Compared to the reference S-N curve, the test results show similar stress amplitudes, but tending to higher endurable stresses and are partly outside the scatter band T σ . Compared to a von Mises fatigue class FAT 178 (0.89 FAT 200 from the principal stress hypothesis [41] ) recommended for weld seam root failures in ref. [45], the test points of the LBW butt joint specimens and the MPW pipe joint specimens are in the conservative range, i.e., show higher notch stress amplitudes. The recommended notch stress fatigue class FAT 178 can thus be applied for a safe and reliable fatigue assessment.

Summary and Discussion
In this article, it was shown that LBW can successfully be used to manufacture a longitudinally welded pipe made of AA7075, followed by MPW to join the pipe to a 3D-printed lug-tube made of additively manufactured AlSi10Mg.  The fatigue strength of MPW pipe joints and LBW butt joint specimens was determined using load-controlled fatigue tests. Compared to the butt joint specimens welded without filler material, specimens welded with filler material show a higher fatigue strength related to nominal stresses σ N . This is due to an improvement of the weld seam geometry, which was manifested in reduced weld seam root fallback. All butt joint specimens welded without filler material show fractures in the weld seam center. This can also be observed predominantly in the specimens applying filler material and is due to high local stress peaks in the area of weld seam fallback. Three specimens show an off-center fracture at the weld seam notch. These specimens also tend to show higher fatigue strength and thus indicate a higher weld seam quality.
Compared to the pipe joint specimens where a centering ring is used for the manufacture, the specimens manufactured without centering ring show higher fatigue strength with comparable slope k and scatter T F of the F-N curves. This can be justified by the fact that the centering ring allows precise adjustment of the distance between the AA7075 tube and the AlSi10Mg lug-tube, but at the same time limits the maximum length of the weld seam.
For the characterization of the specimens, cross sections were prepared and examined metallographically, which reflect the local weld seam geometry in the joining area. Using highresolution 3D-scans and subsequent "reverse engineering" of a pipe joint specimen, a detailed 3D-CAD model was built to serve as a realistic representation of the pipe joint specimen.  www.advancedsciencenews.com www.aem-journal.com The maximum local notch stress σ vM,max was determined for the butt joint specimens welded with and without filler material. This results in a higher stress concentration factor of K t = 4.3 for joints without filler material and a lower K t = 2.7 for joints with filler material, which is due to the local weld seam geometry. In the case of the pipe joint specimen, two competing hot spots were identified in the analysis of the simulation results. The maximum local notch stress σ vM,max was determined in the root notch on the pipe side. In contrast, the local notch stress at the lug-tube has a lower value by a factor of 1.4. It is assumed that due to the lower fatigue strength of the additively manufactured lug-tube made of AlSi10Mg compared to the roll-formed pipe made of AA7075 and due to different degrees of residual stresses in the areas of the notches, caused by the joining process, cracks occur in the lug-tube despite higher local notch stresses in the pipe.
A fatigue assessment is made using linear-elastic approaches. The reference radius concept with r ref = 0.05 mm is applied to map the influence of geometric notches on the fatigue strength, assuming linear-elastic stress-strain behavior. It was shown that the fatigue strength can be estimated sufficiently well compared to the recommended notch stress fatigue class FAT 178 (von Mises stress). No difference in endurable notch stresses between the two highly different welding processes, LBW and MPW, could be determined. Consequently, only the geometrical features and notches in combination with the metallurgical properties determine the fatigue strength.

Main Conclusions
The main conclusions drawn from the presented investigations are summarized as follows: Using MPW, roll-formed and laser beam longitudinally welded pipes were successfully joined to AM fabricated lug-tubes. The weld seam formation, however, did not occur at all positions around the circumference. This is attributable, on the one hand, to the root of the LBW seam, which locally prevents successful MPW of the AA7075 pipe with the AlSi10Mg lug-tube. On the other hand, there are weak points in the magnetic field due to the field former cut, resulting in a negative effect on the weld seam length located there. It remains to be noted that the acceleration distance between the flyer and the target was not always constant due to the nonregular shape of the pipe, causing the weld seam length to be formed unevenly even in areas other than the weak points described.
Using centering rings within the manufacture of pipe joint specimens improved the geometric tolerance of the joint but led to reduced fatigue strength by a factor of 1.25. This is due to the fact that the centering ring provides a precise adjustment of the distance between the AA7075 pipe and the AlSi10Mg lugtube, but at the same time limits the maximum length of the weld seam.
From FE analysis, two competing hot spots of pipe joint specimens have been identified by applying the reference radius r ref = 0.05 mm in the notch stress approach showing a maximum on the pipe-side root notch. The evaluated notch stresses versus service life of butt joint and pipe joint specimens are in good agreement with LBW lap joints from refs. [32,43,44] Consequently, only the geometric features and notches in combination with the metallurgical properties determine the fatigue strength.
Crack initiation was always found in the additively manufactured lug-tube made of AlSi10Mg, although higher notch stresses were evaluated in the pipe. It is assumed that due to the lower fatigue strength of the additively manufactured lugtube compared to the roll-formed and laser beam longitudinally welded pipe and due to different degrees of residual stresses in the areas of the notches, caused by the joining processes, cracks occur in the lug-tube.
The recommendation of FAT 178 with von Mises stresses is suitable for the fatigue strength assessment of the MPW pipe joint specimen with r ref = 0.05 mm and can be applied safely and reliably. This indicates that in the case of MPW specimens, only the geometric features and notches in combination with the metallurgical properties determine the fatigue strength, what should be investigated by further studies.
Further investigations should be carried out with additional specimen types and component-like specimens with complex stress conditions to validate the fatigue assessment with the reference radius concept.