Research on the Microstructures and Properties of AA5052 Laser-Welded Joints with the ER4043 Filler Wire

Abstract

Researches were conducted on the laser welding of 3 mm sheet-thickness lap joints of AA5052 with ER4043 filler wires. The effects of laser power on the joint morphology, microstructure, mechanical properties, and corrosion resistance were investigated. The results indicate that both increased heat input and the addition of filler wires make the molten pool more instable, which results in more process pores. Circular pores are observed in the upper part of the weld, while chain-like pores are distributed in the middle of the weld. The highest tensile strength of the weld joint is 192.61 MPa with an elongation of 10.1% at a laser power of 3.5 kW. The microhardness at the center of the weld is approximately 25% higher than the base material, which is probably because the addition of ER4043 filler wires brings more Si element to the weld. Moreover, the weld joints display superior corrosion resistance compared to the base material. These outcomes enhance the understanding of AA5052 laser welding with fillers wire and provide valuable in-sights for engineering applications.

1. Introduction

With the continuous deepening of energy conservation and emission reduction, aluminum alloys are considered ideal materials for lightweight automotive applications [1]. The 5xxx-series aluminum alloys belong to the Al-Mg alloy type, which is a non-heat-treatable strengthening aluminum alloy. It has advantages such as being lightweight and having a high specific strength, an easy formability, and a high corrosion resistance [2]. Welded structural components made of this alloy are widely used in automotive body panels, wheel hubs, and other parts.

It is widely acknowledged that laser welding has advantages over conventional welding processes, including a narrow heat-affected zone (HAZ), minimal distortion, a low environmental impact, non-contact processing, a high control precision, and an increased efficiency [3]. In laser welding with filler wire, the addition of filler wire can adjust the composition and microstructure of the weld, thereby enhancing the properties of the weld seam or inhibiting the formation of cracks [4]. However, aluminum alloys pose a greater welding challenge compared to other materials due to their excellent thermal conductivity, high solidification shrinkage rate, oxide formation, and wide solidification temperature range [5]. The high magnesium (Mg) content in 5xxx-series aluminum alloys, coupled with their low boiling point (1090 °C), leads to porosity defects in the weld [6]. Additionally, cracking is a significant welding challenge faced with aluminum alloys. ER4043 belongs to the Al-Si-series welding wire and has an excellent fluidity and optimal anti-solidification cracking properties [7]; it can serve as a solution for the solid solution strengthening of the welded joint, thus improving the joint’s mechanical properties [8].

The influence of the laser welding process parameters on the performance of welded joints is crucial. The main process parameters include the laser power, welding speed, and defocus amount. Chen et al. [9] studied the effects of laser power on the grain size and tensile strength of 5A90 Al-Li alloy T-joints welded with double laser beams on both sides synchronously. Their results indicated that laser power was a key process parameter affecting the mechanical properties of welded joints. He et al. [10] demonstrated that a lower welding speed or increased laser power resulted in wider welds with coarser grain structures. During the welding process, the filler material (welding with filler wire) is a critical factor influencing the microstructure, mechanical properties, and corrosion performance of the weld seam. Omprakasam et al. [11] used two different filler materials, ER4043 and ER5356, for the TIG welding of AA5052. The results showed that the weld seam’s center had an increased hardness compared to the base material. And when the ER4043 filler was used, the weld seam was free from hot cracking.

In this study, an AA5052 alloy was utilized for multi-layer laser welding with filler wire, employing ER4043 as the filler wire. The influence of different laser powers on the macroscopic appearance of the weld bead was studied, and the types, distribution patterns, and mechanisms of porosity formation were analyzed. The properties of the weld were studied by examining the weld’s microstructure, microhardness, tensile strength, and electrochemical corrosion properties. This study aimed to analyze the specific effects of the filler wire and laser power on the weld’s microstructure and properties, which provides guidance for engineering applications.

2. Materials and Methods

The base material (BM) was an AA5052 aluminum alloy with dimensions of 150 mm × 50 mm × 3 mm. ER4043 filler wire with a diameter of 1.2 mm was used in the present study. The chemical compositions of the AA5052 aluminum alloy and ER4043 filler wire are presented in

Table 1. Chemical compositions of AA5052 and ER4043 (wt%).

	Mg	Si	Cu	Zn	Mn	Cr	Fe	Al
AA5052	2.6	0.05	0.00	0.03	0.06	0.20	0.14	Bal.
ER4043	0.10	4.5	0.04	/	/	/	0.04	Bal.

.

The equipment for laser welding with filler wire and the welding schematic are shown in Figure 1. The equipment used was an alta 6000 (nLight, Shanghai, China) fiber laser whose rated power was 6 kW. A fiber-optic cable with a core diameter of 100 µm channeled the laser beam to the ALO3 laser-processing head (Scansonic, Berlin, Germany). A KUKA KR210 (KUKA, Augsburg, Germany) robot, a TransPuls Synergic 5000 CMT (Fronius, Shanghai, China) welding machine (which controlled the speed of the filler wire), and a VR 1500 robotic (Fronius, Shanghai, China) wire-feeding system were employed. Argon with a purity of 99.99% and a flow rate of 12 L/min was used as the shielding gas. The samples were overlapped and welded with front-side wire feeding. The detailed parameters for the welding process with the filler wire are listed in

Table 2. Laser-welding parameters.

Parameters	Value	Units
Laser power	2.5~4.5, ∆ = 0.5	kW
Welding speed	30	mm/s
Wire-feeding speed	3.2	m/min
Defocus distance	0	mm
Distance between laser and welding wire	3	mm

.

Samples with dimensions of 18 mm × 6 mm were cut from the weld cross-section using a wire-cutting machine. These samples were firstly ground with metallographic sandpaper to 5000#, and were then polished using a MgO polishing agent to obtain a mirrored surface and ultrasonically cleaned in an anhydrous ethanol solution for 5 min. Finally, the etching experiment was performed using Keller’s solution (1% HF + 1.5% HCl + 2.5% HNO₃ + 95% H₂O). The microstructure of the sample was examined with an optical microscope (OM, VHX-500F KEYENCE, Osaka, Japan) and a scanning electron microscope (SEM, Quanta 450, FEI, Hillsboro, OR, USA). The EBSD test was performed to obtain the grain size and orientation distribution.

The microhardness of the joint was measured along the cross-section with a microhardness-testing machine (DHV-1000Z, CANYTEC, Shanghai, China). The load and holding time were, respectively, 100 g and 15 s. The interval between the test points was 0.5 mm. Tensile tests were carried out at room temperature on a 50 KN universal testing machine (Zwick/Roell Z050, Ulm, Germany) with a cross-head velocity of 1 mm/min. The tensile samples were prepared perpendicular to the weld seam according to ISO 6892-1:2009 [12], as shown in Figure 2.

Electrochemical corrosion tests were conducted on a Chi660e electrochemical workstation. A standard three-electrode electrochemical cell, wherein a saturated calomel electrode (SCE) served as the reference electrode, a platinum wire as the counter electrode, and the sample as the working electrode, was employed. The tests were performed in a 3.5% NaCl solution at room temperature. The sample was metallurgically ground and ultrasonically cleaned before the test. The testing started from the open-circuit potential (OCP), and the Tafel polarization curves were measured after the OCP stabilized.

3. Results and Discussion

3.1. Weld Appearance and Internal Quality

Table 3. Surface morphology and longitudinal section of the welds with different laser powers.

Power (kW)	Weld Surface Morphology	Longitudinal Section	Porosity
2.5			9.84%
3.0			14.30%
3.5			15.31%
4.0			6.54%
4.5			4.60%

illustrates the macro-morphology of the welds with different laser powers. The weld surfaces were well formed with clear fish-scale patterns, and no obvious cracks were observed. The jagged morphology that was observed at the bottom of the welding seam was primarily influenced by the instability of the keyhole structure. This irregularity was chiefly attributed to the velocity of the wire feed [13]. The Image Pro Plus software 6.0 was used to measure the penetration width and the average penetration depth in the longitudinal cross-sections of the weld, as shown in Figure 3. The change in the penetration depth was due to an increase in the heat input causing slower cooling rates and longer solidification times, resulting in an improved penetration depth [14]. When the power reached 4.0 kW, partial full penetration was achieved. Thus, the optimal weld formation quality was realized with a laser power of 3.5 kW, where the weld bead exhibited the maximal upper weld width and minimal excess height.

There were two main types of pores in the weld joints: metallurgical pores and process pores. Figure 4 presents SEM images of the two types of pores. Metallurgical pores are characterized by their small size (diameters of around 50 µm) and smooth inner walls with minimal impurity particles and orderly arranged grains. During the melting of the base material, a large amount of hydrogen dissolves into the liquid Al, while the solubility of hydrogen varies with the state of the aluminum alloy (0.69 mL/100 g in the molten state; approximately 0.036 mL/100 g in the solid state) [13]. The reduced solubility in the solid aluminum causes hydrogen to precipitate in the weld pool, which finally forms hydrogen pores (process pores) [15]. The hydrogen mainly originates from the oxide film of the base material surface and the moisture in the surrounding atmosphere [16].

The process pores exhibited irregular shapes and rough inner walls, with numerous irregular small holes on the walls (as shown in Figure 4). Two types of process pores were observed in the laser-welded joints: circular pores, which were distributed in the upper portion of the weld, and chain-like pores, which were distributed in the middle portion of the weld. The diameter of the circular process pores was approximately 1 mm, which was 20 times larger than that of the metallurgical pores (as shown in Table 3). It is known that a high laser power exacerbates the molten pool oscillation and keyhole oscillation frequency, which results in pore formation [17]. The formation mechanisms of process pores in this research are illustrated in Figure 5.

The reason for the formation of spherical pores is the interaction between the vapor induced by the incident laser and the locking hole located in the middle, which results in a local bulge forming on the rear wall of the locking hole. As the locking hole moves forward, the gas cavity transforms into a bubble, which enters the weld pool. Given the rapid solidification of the aluminum alloy, the captured bubbles rise and form circular pores within the weld [18]. With increasing laser power, the number of circular process pores in the upper portion of the weld gradually decreased, but their volume increased. This was because a higher laser power resulted in more heat input and a greater force exerted by the metal vapor, which led to larger gas cavities on the rear wall of the locking hole.

The chain-like process pores in the middle of the weld were caused by the collapse of the rear wall of the locking hole due to its instability, trapping the gas inside the molten fluid to form bubbles [19]. Because of the surface tension of the surrounding molten metal, the process pores tended to shrink into spherical shapes. Several spherical pores aggregated and ultimately turned into chain-like spherical process pores. These pores, being relatively large in volume, experience significant buoyancy and, therefore, can be distributed in the middle and upper portions of the weld at the root of the molten pool [20]. When the weld was not fully penetrated, the filler wire exacerbated the instability of both the keyhole and the molten pool. This was because the molten filler wire flowed downward along the front side of the keyhole, which accelerated the collapse of the rear wall of the keyhole. Additionally, the presence of the filler wire shortened the solidification time of the molten pool, and the pores were unable to rise and escape in time before the molten pool fully solidified, causing them to remain in the weld.

When the weld was not fully penetrated, the porosity rate reached as high as 15.31% with a laser power of 3.5 kW. As the laser power increased from 2.5 kW to 3.5 kW, there was a 5.47% rise in the porosity rate. The average depth of weld penetration increased by 50% compared to that at 2.5 kW. This inevitably led to the formation of more pores, which were positioned deeper within the weld. When the power reached 4.0 kW, the percentage of chain-like spherical process pores significantly decreased to about 6.54%. At a laser power of 4.5 kW, the process pores located at the middle of the weld were ellipsoidal, with the lowest porosity rate being about 4.6%. This was because, when full penetration of the weld occurred, the keyhole collapse was reduced and process pores could escape rapidly from the bottom of the weld seam, which contributed to the reduced porosity rate.

Compared with other studies [21], the laser welding of AA5052 with filler wire in this study resulted in a relatively high porosity, primarily due to the increased instability of the molten pool caused by the filler wire.

3.2. Microstructures

The microstructures of the AA5052 alloy welded joints with different laser powers are shown in Figure 6. The grain size of the fusion zone (FZ), counted by the three-wire method, is shown in Figure 7. The HAZ and the FZ were selected for the analysis. During the welding process, the dendritic crystals were perpendicular to the fusion line, that is, they tended to grow in the direction of maximum heat flow from the molten pool to the BM (the grain growth direction is plotted in Figure 6c). This was because the thermal gradient (G), along with the solidification growth rate (R), plays a pivotal role in defining the microstructure of the FZ [22]. G is the highest when R reaches its minimum close to the fusion line [23]. Such conditions facilitate heterogeneous nucleation on the surfaces of partially melted base metal grains (i.e., HAZ) at the interface (i.e., the fusion zone boundary), which grows toward the center of the weld bead in the form of columnar dendrites [24].

However, the temperature gradient was the lowest and multidirectional heat transfer to the surrounding environment occurred in the central region of the weld. The solidification velocity then gradually dropped, the solute content increased, and the constitutional supercooling zone significantly expanded accordingly [22]. New nuclei formed near the center of the molten pool. As a result, typical equiaxed crystals (fine equiaxed grains and dendritic equiaxed grains) formed, which can be clearly observed in Figure 6. As the laser power increased, the greater heat input prolonged the solidification time and allowed more time for the crystals to grow. Therefore, the higher the laser power, the larger the grain size [14].

EBSD was performed on the heat-affected zone of the 3.5 kW sample, as shown in Figure 8. In the inverse pole figure (IPF) map, different colors represent different crystallographic orientations, and the map clearly shows the difference in the grain size and orientation among the BM, HAZ, and FZ. At the interface of the fusion boundary, the upward growth of the columnar grains at an angle was distinctly evident, a characteristic attributable to the preferential alignment of columnar grains along the heat transfer vector during solidification [3]. The grain size of the HAZ was 70 ± 15 µm, while the grain size of the BM was 40 ± 15 µm. This was primarily due to the grain growth occurring in the HAZ as a result of thermal exposure. The kernel average misorientation (KAM) map indicates that the level of strain accumulation in the weld zone was lower than that in the base metal region. This suggests that the addition of the ER4043 filler wire reduced the stress concentration in the weld zone and decreased the tendency for crack formation.

An SEM-EDS elemental analysis was conducted on the weld zone and base metal of the 3.5 kW sample, as shown in Figure 9. It can be noticed that the content of Mg in the BM was slightly reduced compared with that in the FZ. Mg is a highly volatile element and is subject to burnout from the BM during the welding process. Furthermore, the FZ contained a higher Si content, which primarily originated from the filler wire. It is inferred that more hypoeutectic structures (α-Al + Si binary eutectic) are more likely to generate between the primary α-Al grain boundaries.

3.3. Microhardness

The welding hardness distribution is shown in Figure 10. It was observed that the weld center had the highest hardness value, the HAZ had the lowest hardness value, and the hardness value of the BM fell in between. The maximum hardness value at the weld center was approximately 25% higher than that of the BM, which indicated the occurrence of hardening in the FZ. The hardness of the HAZ decreased to a certain extent, reaching about 90% of the BM. Therefore, the HAZ was the weakest region of the welded joint. As the laser power increased, the microhardness of the weld center did not change much, which indicated that the change in laser power did not have a significant effect on the hardness of the FZ.

The occurrence of recrystallization led to the formation of an equiaxed dendritic microstructure at the center of the weld, which resulted in higher hardness values. Conversely, a drop in the hardness of the HAZ was caused by thermal cycling, which induced grain growth and over-aging [25]. This is consistent with the EBSD results. It can be inferred that the failure occurred in the HAZ, where the hardness was minimized.

Additionally, the addition of the Si element by the ER4043 wire compensated for the loss of elements in the weld, and the Si element itself had a high wear resistance and increased the microhardness of the intermetallic compounds, which led to joint hardening. Compared to laser remelting welding [3,26], the laser welding of AA5052 with filler wire can compensate for joint-softening defects, which result in a weld seam with a higher hardness.

3.4. Tensile Tests

Figure 11 illustrates the tensile engineering stress–strain curves of welded joints with different laser powers, while

Table 4. Tensile strength and elongation of welded joints with different laser powers.

Power (kW)	Tensile Strength (MPa)	Elongation (%)
BM	232.20	14.98
2.5	119.91	9.65
3.0	180.00	8.44
3.5	192.61	10.10
4.0	205.81	8.53
4.5	190.84	9.02

summarizes the statistics of the tensile strength and elongation. The tensile strength increased with a rise in the laser power, reaching a maximum of 205.81 MPa with a laser power of 4 kW. However, there was still a decline of 11.37% compared to the base material. Although AA5052 is known to achieve elongation rates of up to 14.98%, the highest ductility observed in the experiments was only 10.1%. It is inferred that AA5052 presents a favorable weldability because of the relatively small disparity between the tensile strength of the weld metal and that of the BM. This characteristic can be attributed to the nature of the non-heat-treatable alloy, on which the thermal cycling induced by welding exerted a negligible impact [26].

It is obvious that the depth of fusion and the presence of porosity affected the tensile strength of the weld joint. The existence of pores within the weld reduced the effective area, ultimately reducing the joint’s performance. Fine equiaxed crystals at the center of the weld exhibited higher mechanical properties, which was attributed to the increased grain boundary density impeding dislocation movement (i.e., hindering the plastic deformation of the metal).

3.5. Electrochemical Corrosion

Figure 12a shows the variation in the open-circuit potential with time for the base metal and weld zone. While the open-circuit potential’s magnitude is not a definitive measure of a sample’s corrosion resistance, it can nevertheless offer an indication of its proclivity for corrosion within an electrolyte. A more negative open-circuit potential indicates the loss of electrons by the metal, which makes it more susceptible to oxidation reactions and corrosion [27]. By comparing the open-circuit potentials of the base metal and the weld with different laser powers (as shown in

Table 5. Corrosion properties.

Region	EOCP (V)	Ecorr (V)	Icorr (A/cm2)
BM	−0.751	−0.716	7.949 × 10−6
P = 2.5 kW	−0.727	−0.699	1.116 × 10−7
P = 3.0 kW	−0.725	−0.676	1.377 × 10−7
P = 3.5 kW	−0.720	−0.637	1.400 × 10−7
P = 4.0 kW	−0.719	−0.651	2.055 × 10−6
P = 4.5 kW	−0.675	−0.677	2.669 × 10−8

), it can be concluded that the corrosion tendency of the FZ was smaller than that of the BM.

The polarization curves of the samples after electrochemical testing are shown in Figure 12b. The corrosion potential in the center region of the welded joint showed positive displacement compared to that in the BM. The cathodic and anodic current densities of the welded joints were lower than those of the base material, which indicates that both the cathodic and anodic processes in the welded joints were inhibited. The common chemical reactions of an aluminum alloy in a NaCl solution are as follows [28]:

Anodic reaction: Al → Al³⁺ + 3e⁻,

(1)

$$\mathrm{Cathodic}\mathrm{reaction}:3{\mathrm{H}}^{+}+3{\mathrm{e}}^{-}\to {\displaystyle \frac{3}{2}}{\mathrm{H}}_2$$

(2)

The corrosion potential and the corrosion current density were determined by the Tafel extrapolation technique, and the results are shown in Table 5. The trend in the corrosion potential mirrored that of the open-circuit potential, displaying the lowest corrosion potential in the base metal (BM) and registering an upward shift in the center of the weld with increasing laser power. It is known that the presence of Mg can lower the corrosion potential of the Al substrate, whereas elements such as Si, Cu, and Fe are inclined to elevate the corrosion potential [29], which can lead to the destruction of the protective layer on the surface of the aluminum alloy and accelerate the cathodic reaction. Therefore, it can be concluded that the improved corrosion resistance in the weld zone was due to the lower Mg content and higher Si content compared to the base metal. Additionally, the presence of finer equiaxed grains at the center of the weld bolstered its resistance to anodic dissolution, which led to an improved overall corrosion resistance.

An SEM examination of the surfaces of the base metal and the weld after the corrosion tests with a laser power of 3.5 kW was carried out, as shown in Figure 13. The base metal’s surface exhibited numerous uniformly distributed pitting corrosion pits, and the magnified image revealed crystal tunnels [30]. The weld surface after corrosion showed a small number of corrosion pits. This is consistent with the results of the potentiodynamic polarization curve. The accelerated anodic process in the base metal resulted in more Al atoms losing electrons and being oxidized, leading to the formation of pitting corrosion pits.

4. Conclusions

In the present study, the laser welding of a 3 mm thick AA5052 alloy with ER4043 filler wire under various laser powers was performed. The morphology of the welded joints and the microstructures were observed and the mechanical properties of the weld joint were tested. The main conclusions can be summarized as follows:

• As the laser power increased, the rise in the heat input resulted in a deeper molten pool and an increased depth of fusion in the weld seam. The addition of filler wire enhanced the instability of the keyhole, leading to the presence of pores in the weld. Metallurgical porosity was primarily caused by hydrogen pores, while process porosity was due to the instability of the keyhole.
• The grain structure at the center of the weld seam was refined, featuring fine equiaxed grains and dendritic equiaxed crystals, while that of the HAZ exhibited a columnar grain structure with coarser grains.
• A significant amount of the Si element from the ER4043 filler wire was incorporated into the weld bead, acting as a strengthening agent. The hardness in the FZ was the highest. The optimal parameter with a laser power of 3.5 kW resulted in a maximum tensile strength of 192.61 MPa and an elongation of 10.1%.
• In the electrochemical corrosion experiments, the center of the weld seam exhibited a higher open-circuit potential and corrosion potential than that of the base material. After corrosion occurred, the base material’s surface had more pitting corrosion pits and presented crystalline tunnels, while the welded area contained fewer corrosion pits. Therefore, the FZ was found to be more corrosion-resistant compared to the BM.