Mechanics of Materials
Microstructure and Mechanical Study on Laser-Arc-Welded Al–Zn–Mg Alloy
2020 Volume 61 Issue 1 Pages 119-126
Details
2020 Volume 61 Issue 1 Pages 119-126
Al–Zn–Mg alloy is widely used as lightweight material. Laser-arc hybrid welding is considered to be a suitable joining process for aluminum alloy. In this paper, 3-mm-thick 7075 aluminum alloy was welded by laser-arc hybrid welding method. The microstructure of the welded joint was analyzed. Microhardness measurement and tensile test were conducted. The fatigue test of the welded joint and base metal was carried out. And the feature of fatigue crack initiation and propagation were discussed.
Fig. 6 S–N curve of the welded joint.
Al–Zn–Mg series aluminum alloys, the mechanical properties of which are considered to be the most enhanced among all aluminum alloys.1) 7075 is an typical Al–Zn–Mg series alloy widely used in aircraft structures, mold processing, and other mechanical equipment. Traditionally, metal inert gas (MIG) welding is usually used in joining aluminum alloy. However, due to the large heat input, defects such as pores, cracks, and deformations are generated easily; therefore, welded joints (WJs) have low mechanical performance.2) Laser beam welding (LBW) has been used as an advanced joining technology due to its many advantages including high energy density, high welding speed, and narrow heat-affected zone (HAZ). However, due to aluminum’s high reflectivity to light, LBW becomes a big challenge.3)
In order to overcome the disadvantages of laser welding and arc welding, laser-arc hybrid welding (LAHW) was developed. It has lots of advantages,4,5) such as deeper welding penetration, more stable and better gap tolerance, which raised increasing interests in the research field. In recent years, studies on the hybrid laser-MIG welding are a hot topic in the welding of Al-alloys. Wu et al.6) examined the microstructures and mechanical properties in hybrid 7075-T6 welds. Hu et al.7) examined the weldability of 7075 using a hybrid laser-GMA welding.
Welding is one of the most important way of connecting materials. Most of the damage is well known to be caused by fatigue fracture, which is usually the process of formation, propagation, and fracture of fatigue cracks. Therefore, studying the fatigue fracture behavior of materials and welded joints is of great significance in both theory and practice.
Shiraiwa et al.8) examines machine learning methods to predict fatigue strength with high accuracy. Liu et al.9) proposed a fatigue crack initiation life-based model and investigated fatigue characteristics of A7N01 alloy welded joint. Zhang et al.10) pointed out that hybrid welding produced the welds with low percent porosity, grain refinement, thus improving the fatigue properties. Ghosh et al.11) thought that the fatigue crack of Al–Zn–Mg weldment was initiated along to the weld center due to the plenty of pores formation during the welding process. Qiao et al.12) pointed that MgZn2 precipitation in the weld zone can make the dislocation movement slowdown, thus effectively prevented fatigue crack growth. Wu et al.13) thought that 7000 series hybrid weld shows a serious softening in hardness and poor mechanical properties due to the evaporation loss of Zn and Cu and precipitate growth together with the porosity.
This work uses a fiber laser-MIG welding method to weld 3-mm-thick 7075 aluminum alloy. An analysis of the microstructure of the WJs was conducted. The local-zone microhardness and tensile strength of the WJs were investigated intensively to analyze the joint properties. A fatigue test of the WJs was also carried out. The fracture morphology and crack initiation and propagation of the WJs were observed by metallographic microscope and scanning electron microscope (SEM), and the initiation and propagation features of the cracks were analyzed. The aim is to provide a theoretical and experimental basis for practical engineering applications.
The base metals (BMs) used in this study were 7075-T6 plates with 3-mm thickness. The filler material was ER5356 with a 1.2-mm diameter. The chemical compositions of 7075 and ER5356 are listed in Table 1.
The welding experiments were carried out by an IPG YLS-5000 fiber laser machine and a Fronius TPS4000 electric arc welder controlled by a KUKA industrial robot. A diagram of the experimental process is illustrated in Fig. 1(a). Before welding, a laser cleaning machine was used to remove surface oil, other metal elements, and oxide film from the surface to avoid the creation of pores during welding. Throughout the experiments, argon gas with a flow rate of 20 L/min was applied to protect the molten pool.
(a) Schematic diagram of the welding process; (b) Dimensions of the general extracted specimen; (c) Sketch of extraction location; (d) Dimensions of local-zone test specimens.
To optimize the welding parameters, several hybrid laser-MIG hybrid welding experiments were performed. According to the weld appearance, suitable welding process parameters are presented in Table 2. Figure 2 shows the postweld appearance of the WJs. The weld bead was well formed, the fusion with the BM was good, the transition was gentle, and no defects were generated, such as incomplete penetration and welding cracks.
Macroscopic appearance of the welded joint and hardness measurement position and direction.
After welding, the cross-section of the weld was polished and etched by Keller’s reagent (HF:HCl:HNO3 = 1:1.5:2.5). Microstructural examinations were carried out using Keyence VHX-6000 and Nikon Epiphot 300 optical microscope.
Standard tensile and fatigue specimens (Fig. 1(b)), were extracted both from the BM and the WJs according to the National Standard GB2561-88, tested at room temperature by a Zwick/Roell Amsler HB250 electrohydraulic servo-testing machine to determine the mechanical properties of the WJs. The loading frequency was 15 Hz and the stress ratio was 0.1. A Hitachi S-3400N SEM was used to observe the surface morphology of the fractures.
Extensive microhardness measurements were conducted as shown in Fig. 2. A 0.981 N force was applied for 15 s with the adjacent indention distance of 0.5 mm.
The local-zone samples were sliced parallel to the weld seam every 0.9 mm from the weld center (Fig. 1(c)). Considering the size of the BM and weld seam, these specimens had a thickness of 0.6 mm after polishing (detailed sizes are shown in Fig. 1(d)). The local-zone tests were performed at room temperature.
Figure 3 shows typical metallographic images of different regions of the hybrid WJ. Figure 3(b) shows the microstructure in the fusion zone (FZ), which is a complex transition zone between the weld and the unmelted BM. Therefore, the chemical composition and microstructure in the FZ are very complicated. The crystal grains close to the BM are only partially melted and the liquid phase coexists with the remaining unmelted solid phase, forming the partially melted zone (PMZ). The grains are all melted in the weld zone (WZ) near the PMZ and the original composition remains after melting due to the extremely short residence time. In the subsequent crystallization process, since the edge of the weld is sufficiently heat dissipated and the cooling rate is fast, the molten metal is directly nucleated on the surface of the unmelted crystal grains, thereby forming a columnar crystal structure growing along the heat-dissipation direction.
Metallographic image of different regions: (a) Weld center; (b) Fusion zone; (c) OSZ in HAZ; (d) BM.
The grain in the center of the weld is obviously equiaxed (Fig. 3(a)), indicating that free crystallization occurs. In this area, the temperature gradient is small and it is less affected by the edge heat-dissipation conditions. A wide component supercooling zone can be formed in the liquid phase. Coarse and coarse dendrites can be formed at the crystallization front. As the degree of subcooling continues to increase, new grains will be produced inside the liquid phase. These grains are not hindered by the surrounding fluid and can grow freely, resulting in equiaxed crystals in the middle of the weld.
Figure 3(c) and (d) shows the microstructure of the overaging softened zone (OSZ) and BM, respectively. The 7075-T6 aluminum alloy used is a rolled deformed aluminum alloy; therefore, the BM grains are elongated. Fine dispersed black particles are observed in the strengthening phase. The grains in the OSZ become coarse and the strengthening phase also gathers and grows up, due to the influence of welding heat, which weakens the original strengthening effect of the BM. Also, according to the grain boundary-strengthening theory, the finer the grain size of the metal material, the more grain boundaries, the strong hindrance of dislocation motion, and the higher the material strength. Thus, the strength of OSZ is much lower than that of the BM, which will be analyzed below.
3.2 Hardness distributionHardness is an important index to measure the performance of metal materials. It can be understood as the ability of materials to resist elastic deformation, plastic deformation, or damage, and can be expressed as the antidestructive ability of materials to resist residual deformation. At the same time, the change of hardness also reflects the change in the WJs’ microstructure. Figure 4 describes the distribution of the microhardness of both top and bottom positions of the transverse section of the hybrid WJs. The main alloying elements of 7075 aluminum alloys are Mg and Zn, which are aging-strengthened aluminum alloys, and the main strengthening phase is MgZn2. The main components of the welding wire are Al and Mg, and the Zn content in the weld is small; thus, the main strengthening phase MgZn2 is difficult to form. Therefore, the hardness of the welding seam is 92.4 HV, which is significantly lower than the HAZ and the BM, which is 179.8 HV.
Microhardness profiles of the welded joint.
The temperature field in the HAZ is unevenly distributed during welding. The closer to the weld, the higher the peak temperature. Because the welding thermal cycle experienced by different parts of the HAZ is different, the behavior of dissolution and aggregation growth of the second-phase particles is different, and the microstructure and properties in different regions of the HAZ are changed. Near the FZ, the peak temperature is high, and the second-phase particles are largely dissolved into the matrix. Then, the extremely fast cooling rate suppresses the precipitation of the second-phase particles, resulting in solid solution strengthening. The solid solution strengthening effect is weaker than the age strengthening, making the hardness of the zone lower than the BM, but higher than the weld. The peak temperature and cooling rate become lower as distance from the WZ increases. The dissolution of the second-phase-enhancing particles is insufficient and solid solution strengthening does not substantially occur, which reduces the microhardness. Because they are close to the BM, they are less affected by the welding heat. Therefore, because the second-phase particles do not substantially dissolve, the tendency of aggregation growth is gradually weakened, and the hardness loss is also smaller until it reaches a stable value.
3.3 Tensile performanceGenerally speaking, there is a certain relationship between hardness and the strength of the metal. The area with high hardness is also high in strength, but its plasticity and toughness will decrease accordingly.
Figure 5 shows the tensile test results of 7075 aluminum alloy BM, the overall WJs and different zones, including WZ and OSZ. The test results show that the WJs have an ultimate tensile strength of 355 MPa, a yield strength of 240 MPa, and an elongation after fracture of 2.35%. The tensile strength and yield strength of the BM are 518 MPa and 416 MPa respectively, which is much higher than that of the WJs. The elongation after fracture is 11.3%, which is also significantly higher than that of the WJs, indicating that the strength and plasticity change significantly after welding. In the test, the WJ specimen was generally broken in the WZ. Thus, the WZ is the weakest mechanical region of the WJ. The main reason is that the mechanical properties of Al–Mg-based filler metals are much lower than those of 7075 aluminum alloy BMs. The WZ results showed that the tensile strength, yield strength, and elongation are 317 MPa, 218 MP, and 3.37%, compared to 418 MPa, 323 MPa, and 5.16% of the OSZ, respectively. When it reaches the tensile strength limit of the WZ, the yield strength of OSZ and BM have not been reached; therefore, plastic deformation in only the WZ occurred during the WJ test, resulting in lower elongation. Wu et al. showed that the tensile strength of the WJ is 333 MPa using a hybrid fiber laser and pulsed arc heat source system. According to Hu et al., a hybrid weld subjected to 1-month natural aging, but without solution treatment, shows a maximum strength of 343 MPa.
Tensile properties of different zones.
According to the yield strength of the WJ, the fatigue performance test of the 7075 aluminum alloy laser-MIG hybrid WJ was carried out under different stress levels. The cyclic stress corresponding to N = 1 × 107 is taken as the fatigue limit of the WJ.
The power function of the S–N curve is expressed as:
| \begin{equation} \mathrm{c}(\mathrm{N}_{\text{f}})^{\text{k}} = \sigma_{a} \end{equation} | (1) |
Take the logarithm of eq. (1), then the double logarithmic equation is:
| \begin{equation} \lg (\sigma_{a}) = \lg (N_{f}) + \mathrm{C} \end{equation} | (2) |
Thus, in the logarithm coordinates, lg(σa) and lg(Nf) shows a linear relationship. The S–N curve fitted with the measured experimental data is illustrated in Fig. 6. According to eq. (2), the fitting formulae of the curves for the WJ and BM are calculated, respectively:
| \begin{equation} \lg (\sigma_{a}) = - 0.0644*\lg(N_{f}) + 2.1443 \end{equation} | (3) |
| \begin{equation} \lg (\sigma_{a}) = - 0.0644*\lg(N_{f}) + 2.3723 \end{equation} | (4) |
S–N curve of the welded joint.
Figure 6 shows that the distribution of fatigue life under different stresses is relatively discrete. The fatigue limit of the WJ and BM are 110 MPa and 190 MPa, respectively, under 1 × 107 cycles.
Even the small size of pores can have a great impact on the fatigue properties of materials. Therefore, the initiation and propagation mechanism of microcracks in pores are analyzed in the following paragraph.
4.1 Fatigue crack initiationFigure 7 shows two cracks initiated from two sides of the pore on the sample surface. According to the SEM images of the fractured fatigue specimens, the crack initiation from pores was the predominant cause of fracture for the WJs.
OM image of crack initiation and propagation direction near the pore.
As we can see from Fig. 7, the size of the pore is about 100 µm. According to Fig. 1(b), the fatigue specimen parallel section is 12 mm. The width of the hole size is far less than the size of the sample; therefore, we can use the infinite plate with a pore defect as the theoretical model to analyze the crack-initiation mechanism. Based on the above assumptions, the force model of a circular hole in an infinite plate is established (Fig. 8).
Diagram of an infinite plate containing a pore.
The plate is subjected to uniform force σ, and there exists a small circular hole with radius a. The existence of the circular hole must affect the stress distribution. The stress at any point P(r, θ) near the hole will be much greater than that without the hole, and much greater than that farther away from the hole. This phenomenon is called stress concentration. If r is sufficiently far from the center of the hole, the stress distribution should be the same as σ. Based on the above analysis, it becomes an external circular force problem of a thick-walled cylinder with inner diameter a and outer diameter r.
According to the elastic mechanics knowledge, three forces are applied to the outer circumference. They are the radial normal stress σr, ring normal stress σθ and ring shear stress τrθ, respectively (Fig. 9). The magnitude of the three forces are:14)
| \begin{equation*} \left\{ \begin{array}{l} \sigma_{r} = \dfrac{\sigma}{2}\left(1 - \dfrac{a^{2}}{r^{2}} \right) + \dfrac{\sigma}{2}\cos 2\theta \left(1 - 4\dfrac{a^{2}}{r^{2}} + 3\dfrac{a^{4}}{r^{4}} \right)\\ \sigma_{\theta} = \dfrac{\sigma}{2}\left(1 + \dfrac{a^{2}}{r^{2}} \right) - \dfrac{\sigma}{2}\cos 2\theta \left(1 + 3\dfrac{a^{4}}{r^{4}} \right)\\ \tau_{r\theta} = - \dfrac{\sigma}{2}\sin 2\theta \left(1 - 3\dfrac{a^{4}}{r^{4}} + 2\dfrac{a^{2}}{r^{2}} \right) \end{array} \right. \end{equation*} |
Diagram of stress distribution near the pore in the plate: (a) Radial normal stress σr; (b) Ring normal stress σθ; (c) Ring shear stress τrθ.
At the edge of the hole, r = a. When θ = ±π/2, σθ = 3σ. At the section perpendicular to the load direction, the maximum stress at the edge of the hole is three times higher than that without the hole.
However, according to Fig. 7, there are still some deviations between the actual crack initiation location at the stoma and the above model. In addition to the stress concentration at the edge of the pore, the unevenness of grain structure and lacking of smoothness at the edges of pore also affects the crack initiation.
4.2 Fatigue crack propagationFigure 10 shows two cracks initiated and propagated from both sides of a pore perpendicular to the load direction. A radial plastic zone can be observed at the tip of the cracks. During the crack-propagation process, the high stress concentration at the crack tip will cause the material to locally yield and form a plastic deformation zone of a certain size. The plastic deformation zone is caused by the slip along the maximum shear stress plane caused by the shear stress. The main function of plastic deformation at the crack tip is to absorb the plastic deformation, passivate and relax the crack tip, reduce the stress level of the crack tip, and prevent crack growth.
OM image of plastic zone at fatigue crack tip.
During each cycle, the fatigue crack propagation is caused by plastic passivation at the crack tip.15) The mechanism is shown in Fig. 11. Under the minimum load, the crack is closed and the tip is sharp. As we have discussed above, under a tensile load, the crack tip produces a plastic zone due to stress concentration. Plastic deformation causes the crack to gradually change from a closed state to an open state. When the tensile load increases to the maximum value, the crack-opening amount is the largest and the corresponding plastic deformation is also the largest. The crack tip slips along the direction of maximum shear stress by a double slip mechanism, causing plastic passivation. This passivation process allows the crack to extend forward a distance. When the stress is reduced, the plastic deformation will decrease, the crack tip will re-sharp and passivate again during subsequent loading process. Thus, fatigue cracks propagate forward.
Plastic passivation mechanism of fatigue crack propagation: (a) Crack closure; (b) Crack starts to open; (c) Significant plastic deformation; (d) Max plastic deformation; (e) Crack closure.
From Fig. 12(b), the propagation of the crack is not straightforward during the fatigue process, but propagates forward in a snake shape. The mechanism of its formation is illustrated in Fig. 12(a). For the crack type I, there exists two slip sources τ1 and τ2 at the crack tip, which can be activated on the intersecting slip plane. The crack propagation can be divided into four steps: microcrack is generated when a slip source (τ1) at the crack tip is first activated. After passivation, it is difficult to expand the microcrack in the original direction. A tear is generated in the other direction. The microcrack is then passivated again. After multiple cycles, a snake-shaped crack is produced.
The propagation of the crack and its formation mechanism: (a) Snake-shaped crack formation step; (b) OM image of fatigue the snake-shaped crack.
The propagation of the crack is mainly through transcrystalline form. Metallographic observation on the polished and corroded surface reveals the morphology of the crack across the crystal (Fig. 13).
Metallographic image of a transcrystalline crack.
Figure 14 shows the SEM images of fatigue fracture of WJs. Fatigue crack initiation occurs at the surface porosity defect. After initiation, the fatigue crack propagation zone is formed by extending into the sample. The specimen breaks when the fatigue crack grows to a critical size.
SEM image of fatigue fracture. (a) Fatigue initiation zone; (b) Quasi-cleavage steps; (c) Feathery extension features; (d) Fatigue striation.
There are three kinds of fracture forms for aging aluminum alloy: namely, slippage zone cracking, intergranular cracking, and dimple cracking. Figure 14(a) shows that although there is a river-like pattern on the fracture, the “river” extends to the periphery of the plane and the “river” is relatively short and discontinuous with small obvious convergence features, which is obviously different from the “river pattern” of the cleavage fracture. There are many relatively flat quasicleavage surfaces on the fracture surface, the small planes are connected by tearing, and obvious tearing edges can be seen. Therefore, from the perspective of fracture mechanism, the main mode of the WJ is quasicleavage fracture.
The quasicleavage steps parallel to the crack growth direction can be clearly seen in Fig. 14(b). The feathery extension features can be seen in Fig. 14(c). Its formation is due to the extension of fatigue crack propagation from one grain to another grain, and the dislocation movement is easily obstructed at the grain boundary. New cracks are generated by the tensile stress caused by dislocation plug accumulation. In pursuit of minimal energy expenditure, the cracks expand to slightly different surfaces, which causes a feathery feature. These are the obvious characteristics of brittle fatigue fracture, which occurs with quasidissociation.
As we discussed above, since the crack closure cannot eliminate the passivation caused by the maximum stress completely, the crack will extend a distance further in the subsequent process. The traces of such crack propagation can generally be observed on the fatigue fracture, that is, the so-called fatigue striation (Fig. 14(d)).
In the present study, the microstructural and mechanical properties of laser-MIG welded 7075-T6 aluminum alloy using ER5356 filler were investigated and the mechanism of fatigue initiation and propagation was analyzed. The main conclusions are summarized as follows:
This project is sponsored by the National Natural Science Foundation of China (51971129) and the Shanghai Natural Science Foundation of China (19ZR1421200), also supported by the Research Innovation Program for Graduate Student of Shanghai University of Engineering Science (E3-0903-19-01144).
