Interfacial Microstructure Evaluation of Precisely Controlled Friction Stir Welding Joints of Al/Mg Dissimilar Alloys with Zn Interlayer

Abstract

In this study, a penetration-controlled friction stir welding (FSW) technique was employed to lap weld dissimilar Al/Mg alloys, incorporating a Zn interlayer. The joint’s microstructure, interfacial reaction, and phase composition were analyzed through optical microscopy, scanning electron microscopy, and X-ray diffraction. The findings demonstrate the formation of a hybrid joint comprising a FSWed region and a diffusion bonding region achieved by introducing a pure Zn interlayer at the Al/Mg interface. Within the FSWed region, the zinc was fully extruded, leading to favorable interface bonding. In contrast, the diffusion bonding region exhibited an aluminum–zinc diffusion reaction layer, an incompletely reacted zinc layer, and a zinc–magnesium diffusion reaction layer. Notably, no Al-Mg intermetallic compounds (IMCs) were observed in either the FSWed or diffusion bonding regions of the hybrid joint. This study further explored the underlying mechanism behind the joint’s formation.

1. Introduction

Aluminum alloys are extensively utilized in various industries, including aerospace, chemical, and automotive applications, owing to their exceptional properties, such as high strength, corrosion resistance, and excellent electrical and thermal conductivity [1]. Similarly, magnesium alloys, known for their high specific strength, specific stiffness, seismic resilience, and electromagnetic shielding properties, are employed as lightweight structural materials in aerospace and automotive applications [2]. However, the welding of dissimilar aluminum and magnesium alloys to produce complex shapes and geometries often leads to the formation of undesirable defects, such as pores, cracks, and brittle intermetallic compounds (IMCs) [3].

Friction stir welding (FSW) has emerged as a solid-state joining technology capable of addressing the challenges associated with dissimilar material welding [4]. Nevertheless, in Al/Mg dissimilar joints, the elevated peak temperature during friction stir welding surpasses the Mg-Al eutectic temperature, resulting in the formation of brittle IMCs, including Al₁₂Mg₁₇ and Al₃Mg₂ [5]. Furthermore, the direct mixing of Al and Mg alloys within the weld nugget zone under the influence of the welding tool pin contributes to the formation of these brittle IMCs. Consequently, the control of IMC formation in Al/Mg dissimilar joints assumes paramount importance in achieving joints with desirable mechanical properties.

Two approaches have been employed to regulate the formation of brittle IMCs in the friction stir welding of Al/Mg dissimilar alloys. The first involves the use of modified techniques, such as welding tool pin offset [6], ultrasonic vibration assistance [7], and forced cooling [8], to regulate heat generation and distribution. The second approach involves the application of metallurgical methods, including the addition of interlayers such as Ni [9], Zr [10], and Zn [7,11,12], to control the metallurgical reactions leading to IMC formation. Previous studies have demonstrated the effectiveness of these methods in controlling IMC formation and improving the mechanical properties of Al/Mg dissimilar joints. However, it is important to note that the direct contact and mixing of Al/Mg alloys within the nugget zone during friction stir welding are unavoidable, consequently resulting in the formation of Al-Mg series IMCs [9].

Hence, it is intriguing to explore the possibility of mitigating the direct contact between Al/Mg dissimilar alloys and, subsequently, inhibiting the formation of brittle Al-Mg series IMCs within the welded joint. This objective can be achieved by exerting precise control over the penetration depth of the friction stir pin during the friction stir welding process. The relevant literature, such as the work of Liu [13], proposes a method to control the formation of IMCs in the friction stir additive manufacturing of Al/steel dissimilar alloys by precisely regulating the distance between the tool pin and the steel. However, due to the lower reaction temperature between Al/Mg compared to that between Al/steel, employing Liu’s method results in the inevitable reaction between Al/Mg in the shoulder-affected zone of the welding tool, leading to the formation of brittle intermetallic compounds. Introducing an intermediate layer to overcome the aforementioned limitations of this method is a viable solution. In a similar vein, Gan et al. [14] conducted a friction stir-induced diffusion bonding of 6061-T6 Al and AZ31B Mg dissimilar alloys, using a pure Zn foil as a barrier layer. Their findings indicated that the Zn interlayer effectively impeded the reaction between Al and Mg, thereby preventing the formation of Al-Mg IMCs in the joint. Nevertheless, the study observed the presence of a molten Zn interlayer in the stirred zone between the Al and Mg alloys, which was attributed to inadequate extrusion effects due to the tool pin located at the interface between Al and Zn.

To address these limitations, the present study implemented the precise control of the tool pin penetration at the interface between Zn and Mg to prevent the direct mixing of Al/Mg dissimilar metals in the stirred zone. Additionally, a pure Zn foil interlayer was introduced to impede the reaction between the dissimilar metals in the shoulder-affected zone, effectively regulating the formation of brittle intermetallic compounds between aluminum and magnesium throughout the entire friction stir welding joint. This paper offers a systematic investigation of the microstructures, interfacial reactions, and phase formations in Al/Mg dissimilar joints featuring a Zn foil interlayer and the precise control of the tool pin penetration.

2. Experimental Procedure

This study utilized 2014-T4 aluminum alloy with dimensions of 3 mm × 200 mm × 80 mm and AZ31 magnesium alloy with dimensions of 5 mm × 200 mm × 80 mm as base materials.

Table 1. The chemical compositions of the aluminum and magnesium alloys (weight fraction, wt%).

Chemical Compositions	Al	Mg	Cu	Zn	Mn	Si
2014-T4	Bal.	0.55	4.3	0.08	0.73	1.0
AZ31	3.1	Bal.	-	0.88	0.48	0.01

lists the chemical compositions of the base materials. A commercial pure zinc interlayer with dimensions of 0.25 mm × 40 mm × 200 mm was used. Prior to welding, the oxide film on the base materials and the interlayer were removed using sandpaper (Hunan SHARPNESS Sandpaper Belt Co., Ltd., Changsha, China) and then cleaned with acetone (Nanchang Jinsha Chemical Raw Materials Co., Ltd., Nanchang, China).

The FSW tool consisted of a concave shoulder with a diameter of 18 mm and a left-threaded surface pin with a diameter of 5 mm and a length of 3 mm. The aluminum alloy was positioned at the top, the magnesium alloy at the bottom, and the zinc interlayer between the aluminum and magnesium alloys. The lap width was 40 mm. The welding configuration is depicted in Figure 1. Based on the results of the previous welding process parameter optimization experiments, we selected the following welding process parameters. The selected welding parameters consisted of a rotation speed of 750 rpm and a welding speed of 30 mm/min. In order to avoid the mixture of the Al and Mg alloys at the stirring zone, the penetration of the welding tool was controlled at 0.2 ± 0.05 mm.

The metallographic samples were sectioned perpendicular to the welding direction and then polished and chemically etched. The microstructures of the dissimilar joints were analyzed using optical microscopy (OM, Shanghai Yonghui Industrial Development Co., Ltd., Shanghai, China) and scanning electron microscopy (SEM, Thermo Fisher Scientific Inc., Waltham, MA, USA). X-ray diffraction equipment (PANalytical B.V., Almelo, The Netherlands) was used to identify the phase composition formed in the dissimilar joints. Microhardness tests were conducted on the metallographic samples, with the test location perpendicular to the plate thickness direction in the FSW tool’s shoulder-affected area.

3. Experimental Results

3.1. Macrostructure and Microstructure Analysis

Figure 2 illustrates the cross-sectional macrostructure of the Zn-added Al/Mg FSW joint. As depicted in Figure 2a, the joint was formed without any defects, and the added Zn interlayer remained at the joint interface. Figure 2b,c shows the magnified interfacial microstructure of the joint. As seen in Figure 2b, a straight interface was observed at the stirred zone of the joint, where no Zn interlayer was detected due to the extrusion of the FSW tool during welding. Figure 2c presents the interface image of the area located in the shoulder-affected zone of the dissimilar joint, where a four-layer structure was identified at the Zn interlayer. The interfacial microstructures in the diffusion zone were different from those in the stirred zone. A four-layer structure could be clearly identified due to the diffusion reaction between the aluminum alloy, the Zn foil, and the magnesium alloy. To analyze the microstructure of the four-layer structure more clearly, magnified images of these areas were taken, as shown in Figure 2d,e. Specifically, Figure 2d is the magnified image of area A marked in Figure 2c, while Figure 2e is the magnified image of area B marked in Figure 2c. The formation of the four-layer structures at the interface was attributed to the diffusion reaction between the aluminum alloy and the Zn interlayer, and the magnesium alloy and the Zn interlayer. The width of the diffusion reaction area at the interface varied with the location. The width of the diffusion reaction area at the aluminum alloy side was about 100 μm, which was smaller than that at the magnesium alloy side, with a width of about 400 μm. The diffusion reaction area presented various phases with different morphologies, which are further discussed in the following sections.

Line scan mode EDS (Thermo Fisher Scientific Inc., Waltham, MA, USA) was employed to further analyze the area shown in Figure 2b, and the results are shown in Figure 3a. The EDS analysis revealed that no zinc was detected at the interface of the stirred zone, indicating that the inserted Zn foil was completely melted and extruded by the FSW tool during welding. Figure 3b shows the SEM image of the diffusion area presented in Figure 2c, where a four-layer structure was clearly identified.

Table 2. The EDS analysis results (at %) in the diffusion area.

Point	Al	Mg	Zn	Possible Phases
1	12.24	-	87.76	Zn solid solution
2	-	-	100	Zn
3	-	29.27	70.73	MgZn2 + Zn solid solution
4	8.55	71.08	20.37	α(Mg) + MgZn2

presents the EDS results of the four layers. As mentioned earlier, the first layer was the diffusion layer between the Al base material and the Zn foil. The EDS analysis showed that the first layer reaction products were composed of an Al-Zn solid solution with an atom ratio of Zn:Al at 88:12. The second layer of the four-layer structure was the incompletely diffused pure Zn foil. The third and fourth layers of the four-layer structure were composed of Zn-Mg and Mg-Zn-Al, respectively. The atomic ratio of Mg: Zn was 29:71, and the ratio of Al:Mg:Zn was 9:71:20.

3.2. Phase Formation Analysis

To further investigate the reaction products within the four-layer structure, X-ray diffraction (XRD) analysis was performed on the dissimilar alloy joint. The obtained results, depicted in Figure 4, unveiled the presence of a MgZn₂ phase within the diffusion region, alongside the two base materials and the Zn foil. This XRD analysis provided valuable insights into the composition of the joint and the formation of specific phases.

4. Discussions

Brittle IMCs such as Al₁₂Mg₁₇ and Al₃Mg₂ are frequently observed in Al/Mg dissimilar alloy joints and are detrimental to the mechanical properties of the dissimilar joint. There are two main methods that are often used to control the formation of IMCs. One is controlling the heat generation and distribution by using a modified welding process including a solid-state welding process, forced cooling method, and so on. Another one is using selected interlayers such as Zn, Zr, and Ni materials to regulate the metallurgical reaction between the two base alloys. In the present study, an optimized friction stir welding process with precisely controlled welding tool pin penetration and a Zn interlayer were employed to join Al/Mg dissimilar alloys. The results show that hybrid joints that consisted of friction stir welding and diffusion bonding between the Al and Mg dissimilar alloys were obtained. Moreover, a hybrid dissimilar alloy joint free of Al-Mg IMCs was achieved by using the optimized friction stir welding process. Two reasons can be used to explain the results. The obtained friction-stirred zone free of Al-Mg IMCs resulted from the non-mixture of the Al and Mg alloys because of the controlled tool pin penetration. The added Zn interlayer acted as a diffusion barrier between the Al and Mg alloys, which is responsible for the formation of the diffusion area and the absence of Al-Mg IMCs.

Previously published work showed that the temperature during friction stir welding can reach over 420 °C [14], which is higher than the melting point of pure Zn foil. Therefore, it can be deduced that the Zn foil at the stirred zone was melted and extruded under the influence of the welding tool pin. This reasoning can explain the results shown in Figure 2b and Figure 3a, where the Zn element was not found at the interface of the stirred zone of the joint. The results are in contrast to earlier findings made by Gan [14] that the Mg and Zn elements were major components at the joint interface. The difference in the penetration amount of the tool pin was responsible for the contrary results. In the present study, the welding tool pin penetrated at the interface of the Zn foil and Mg base materials, while in the study performed by Gan, it penetrated at the interface of the Al and Zn. The extrusion effect towards the melted Zn by the welding tool pin was more significant than that of the Al alloy. Thereafter, the Zn element was not detected by EDS in Figure 3a. This was also confirmed by Niu [15], who proposed that the distance between the zinc interlayer and the rotating pin tip is the key factor to ensure the performance consistency of the lap joint by friction stir-induced diffusion bonding.

In addition, precise control of the tool pin penetration prevented direct contact between the pin and the bottom Mg alloy, thus avoiding the mixture of the Al and Mg alloys. This resulted in the formation of a distinct interface line between the Al and Mg alloys in the stirred zone. Although the aluminum alloy and magnesium alloy were in direct contact in the stirred zone, the XRD results in Figure 4 show that no Al-Mg intermetallic compounds were formed in the joint. The melted Zn interlayer can serve to clean the interface between the aluminum and magnesium, promoting their reaction. However, friction stir welding is a low heat input welding process, and the melting of the Zn foil consumes most of the welding heat. Thus, the rest of the welding heat is not sufficient to contribute to the formation of large numbers of Al-Mg intermetallic compounds. This is believed to be the reason responsible for the absence of Al-Mg intermetallic compounds in the area.

At the diffusion zone, the existence of the pure Zn foil layer strongly suggests that the Zn foil was not melted, but instead was replaced by a diffusion reaction between the base materials and the foil, which contributed to the formation of the four-layer structure. The first layer was formed by the diffusion reaction between the aluminum alloy and the Zn foil. The reaction layer between aluminum and zinc was a solid solution layer, and no intermetallic compound was formed between the two metals. This was confirmed by the EDS test results in Table 2 and XRD results in Figure 4. The findings match those observed in earlier studies [16,17]. At the interface between the Zn foil and the magnesium alloy, Zn-Mg diffusion reaction layers were formed. The width of the Zn-Mg reaction layer was four times larger than that of the Al-Zn reaction layer. Obviously, the reaction extent on the Mg side was greater than that of on the Al side.

A similar phenomenon was also observed by Zhang [16] and Gu [17]. The higher solubility and diffusion velocity of Zn in Mg, compared to that of Zn in Al, were responsible for the larger diffusion dissolution layer of Zn to Mg compared to that of Zn to the Al side in the same amount of time.

5. Conclusions

To regulate the formation of intermetallic compounds (IMCs) in Al/Mg dissimilar joints, a pure Zn foil was employed as an interlayer material in the friction stir welding process, utilizing a rotation speed of 750 rpm and a welding speed of 30 mm/min. This study conducted a systematic investigation on the interfacial microstructures and phase formation of the resultant Al/Mg dissimilar joints with the inclusion of a Zn foil. The following conclusions were drawn:

By incorporating a pure Zn interlayer at the Al/Mg interface, hybrid joints comprising both a friction stir welding joint and a diffusion joint were formed.

Through the precise control of the penetration depth of the welding tool pin, the mixing of Al and Mg was effectively prevented, resulting in the absence of Al-Mg IMCs in the nugget zone of the hybrid joint.

Within the diffusion region, an aluminum–zinc diffusion reaction layer, an incompletely reacted zinc layer, and a zinc–magnesium diffusion reaction layer were identified. Notably, Al-Mg IMCs were not observed at the interface and were replaced by Zn-Mg IMCs.