An experimental study of friction stir welding parameters effect on joint properties of aluminium alloy and copper plate

Joining dissimilar metals like copper and aluminium is a complex area of research for the manufacturing industry. To effectively join the AA7075 with copper plates, an experimental investigation on friction stir welding was carried out. A straight cylindrical pin profile tool of H13 tool steel was chosen for this study. On the advancing direction, the AA7075 plates were fixed under the condition of zero tool offset. The tool was positioned at a 2° tilt angle for the entire experiment. Tensile test, microhardness test, SEM and EDS analysis were completed to investigate the effectiveness of the process parameters (tool traverse and rotational speed) in combination with a modified set-up on the microstructure, tensile strength and microhardness of the joint. Fractography was conducted on fractured specimens to identify the mode of failure. Complex microstructures of swirl and vortex-type material flow patterns were created in the stir zone. As a result of SEM analysis, one of the samples (S5) had finer grains at intermediate heat input, which produced excellent microhardness and tensile strength. Fractography identified ductile failure as the predominant mode of failure in the joints. The results of this study include an ultimate tensile strength value of 221.091 MPa with a joint efficiency of 95.05%, yield strength of 159.035 MPa, maximum elongation of 9.335% and a microhardness value of 271 HV for specimen welded at a traverse speed of 60 mm min−1 and a rotational speed of 1400 rpm.


Introduction
Aluminium and copper assemblies are broadly used in numerous industries, including aerospace, chemical, electrical, and electronic systems. Electrical and thermal applications can benefit greatly from this combination of materials. The high electrical conductivity of these joints makes them ideal for electrical power transmission [1]. Combining these materials (Al and Cu) reveals the benefits of properties from both materials [2][3][4]. Typically, traditional joining processes like fusion welding are incompatible with joining these kinds of dissimilar metals since the melting temperatures of both metals are vastly different and their mechanical and physical properties do not match. They are also chemically unstable at higher temperatures. As a result of these variations, complications like cracking, residual stresses, and the huge number of intermetallic compounds (IMC) formation can occur in the weld [5][6][7]. Mechanical bolted joints of Al-Cu plates are also not sustainable joints [8]. Although Al and Cu can be welded with explosion and friction welding, the applications of these processes are limited by the joint geometry and the welded plate thickness.
Friction stir welding (FSW) is an enhanced joining approach compared to other joining techniques. It does not cause distortion or shrinkage of the joint and offers excellent mechanical properties, with fewer defects as observed in other joining methods [9]. In addition, FSW is also capable of joining metals and alloys whose melting temperatures are very low like aluminium, copper, and magnesium. The joining of these alloys by any other method is extremely challenging. A suitable tool equipped with pins and shoulders is inserted between plates to be welded and moves through them while stirring. Friction resulting from the tool and the plate produces localized heat that does not melt the workpiece. Instead, it causes remarkable deformation of the surrounding material. Plastically deformed material can be stirred by rotating the tool and mixed to form a joint. Also, brittle intermetallic phase formation can be reduced by plastic deformation, thus reducing cracking in the joints. The rotational speed, traverse speed, tool material, plate positioning and tool offset are the crucial variables that have an impact on the dissimilar FSW [9].
Materials are mixed and stirred around the rotating pin as a result of the tool's rotation and the by moving forward, the tool transfers fluidized material from one side to the other. Heat input increases with higher rotational speeds or lower traversal speeds in FSW processes, which in turn improves the circumstances for Al and Cu diffusion reactions and offers a fine-grained zone [10,11]. Material and characteristics of the tool are crucial in FSW of distant alloys such as Al and Cu. The most vital characteristics of tool material for successful welds include strength at atmospheric temperatures, stability at elevated temperatures, machinability, wear resistance, microstructure uniformity, fracture toughness, chemical reactivity, and a density that is appropriate [9]. Tool material selection should be based on the workpiece material and expected tool life [7,12]. Since more frictional heat is engendered on the advancing direction, when welding dissimilar materials, positioning of plates in relation to tool rotation is a concern. It has been observed that welding integrity in dissimilar FSW is clearly influenced by the plate location during the welding process [13][14][15]. Several researchers observed that with tool offsetting, excessive intermetallic structures during FSW can be minimized, which adversely affects the weld strength and surface morphology [3,14,16]. At the same time some researchers stated that providing offset to the tool is complex and excellent mechanical properties can be obtained with zero tool offset [14,[17][18][19].
Baghdadi A H et al [20] fabricated AZ31 and AA6061 (T6) of 4 mm thickness with AZ31 on the advancing direction. This study showed that better mixing of Mg and Al was accomplished with zero pin offset. Higher tool traverse speed result in enhanced tensile properties. Highest value of tensile strength of 180 MPa was recorded at 40 mm min −1 traverse speed and 600 rpm tool rotational speed. Gotawala et al [21] welded AA1050 and Cu with tool offsets into the AA1050 and positioning the Cu on the advancing direction. It was concluded that dragging the Cu particle into the stir zone was encouraged at a high tool rotation speed and small tool offset which leads to larger interface shifts. Celik S et al [22] welded AA1050 and Cu plates of 4 mm thickness with a high-speed steel tool and achieved 89.5% joint efficiency at a tool offset of 1 mm into Al side while positioning the Cu on the advancing direction. Xue P et al [14] joined 5 mm thick plates of AA1060 and Cu with the Al plate on the advancing direction. Tensile strength was found very poor at large pin offsets and high rotation speeds. AA1100-H14 aluminium alloy and Cu were joined with FSW by Muthu M et al [23] and obtained 70.62% joint efficiency with 113 MPa tensile strength at 2 mm tool offset into Al side and 80 mm min −1 traverse speed. Heat input was found optimum at this traverse speed, which plasticizes and enables the material to flow to obtain defect free joint. Liu H et al [6] welded AA5052 aluminium alloy and Cu using high speed steel tool. Maximum tensile strength value of 127 MPa was attained. Cu plates were positioned on the advancing direction and fine quality welds were observed with zero tool offset. Al-Roubaiy et al [3] welded 5083-H116 and Cu plates of 6.3 mm thickness by providing a tool pin offset towards aluminium side. Chromium alloy steel tool was used for this study. Maximum tensile strength of 206.7 MPa was obtained at 0.2 mm pin offset. Karrar G et al [17] fabricated AA5083 and pure copper plates of thickness 3 mm and achieved joint efficiency of 94.8% with 203 MPa tensile strength with zero tool offset and AA5083 at the advancing direction. Composite like structures were observed at higher rotational speed and dispersed at lesser rotational speed. Bakhtiari A F et al [16] noticed rise in the grain size and IMC due to ample heat at rapid rotational speed and lower traverse speed during joining AA5754 and pure Cu using 1 mm tool offset to Al. Akinlabi E T et al [24] welded AA5754 and C11000 and found intermediate or hot welding conditions favorable for FSW. Heat input process resulted in recrystallized grains and the existence of IMCs in these zones. Tunnel defect was reported by Mehta K P et al [25] at a larger tool pin offset. It has been suggested by Das Chowdhury I et al [19] that grain refinement in stir zones and the development of IMC layers in the joint line significantly affect the microhardness and tensile strength of welds, whereas the IMC thickness was largely depending on temperature and holding time. The FSW experiments were conducted on AA6063 and C26000 with zero tool offset. Safi S et al [26] fabricated preheated AA7075 and pure copper by FSW while keeping the Cu at advancing direction. Maximum tensile strength reported was 84 MPa of the sample preheated up to 75°C and fabricated at the rotational speed of 1300 rpm and traverse speed of 75 mm min −1 . Heat input which was influenced by rotational and traverse speed shown significant impact on the mechanical characteristics of the welds. Osman N et al [27] joined AA6061 and C2801P plates of thickness 1.5 mm using multi pass FSW and studied the impact of process parameters on interfacial bonding characteristics. According to the study, the shear and peel load factor rises up to a specific level with increases in rotational speed and welding speed. After this level, further increases in traverse speed and rotational speed lead to a reduction in the shear and peel load factor. Same effect was observed on Al-Cu bonding. Baghdadi A H et al [18] inspected the impressions of tool offset on tensile strength of AZ31B and AA6061 Friction stir welded (FSWed) joint and realized that the tensile strength was maximum at zero tool offset.
Based on careful consideration of various research papers published concerning FSW of aluminium and copper, it was found that the joining ability through FSW of many aluminium alloys (specially 7xxx series) and copper which are very useful in a variety of industries, requires more investigation. Additionally, not enough research has examined the possibility of FSW of aluminium and copper using an economical and simple design tool. Limited research has focused on conducting FSW of Al and Cu with aluminium on the advancing direction. Therefore, it is possible to assert that there is still a great deal of room for research in this field.
In this experimental investigation, the 3 mm thick plates of AA7075 were welded with pure copper plates of the same thickness using H-13 tool steel cylindrical pin profile tool. Experiments were conducted with placing AA7075 on the advancing direction without the use of tool offsets. Dependency of the mechanical and metallurgical characteristics of the FSWed assemblies on the process parameters of the FSW were investigated.

Materials
To carry out the experiments in this research, AA7075 and pure copper plates that are commercially available were used as the base materials. Table 1 shows the compositions of base metals. Table 2 shows the mechanical properties of the AA7075 and pure copper. The plates selected were of thickness of 3 mm, length of 200 mm and width of 75 mm.
A tool made from H13 tool steel was considered for the FSW process. The composition of the tool materials is presented in table 3. In the experiment, a straight cylindrical pin tool was considered. In terms of the tool dimension, its pin diameter was 5 mm, its shoulder diameter was 16 mm, its pin length was 2.7 mm, and the tool diameter was 20 mm. There was a total length of 90 mm for the tool (figure 1).

FSW process
A linear FSW machine with 10 kN capacity was used for the welding process (figure 2). The parameters rotational and traverse speed in which the FSW was conducted were taken as 1200 to 1600 rpm and 40 to 80 mm min −1 correspondingly [17,22]. In this experiment, the tool tilt angle was unchanged at 2°throughout the experiment.
Even though previous studies have found that the soundest joints can usually be obtained when the material having high melting point are fixed on the advancing direction, this has not always been the case. In order to reduce tool wear during these welding processes, the AA7075 plates were kept on the advancing direction in the interest of preventing higher stirring of the Cu plates during these welding processes (figure 3). Nine samples were welded according to the parameters specified in table 4 of this study [17,22].

Tensile and microhardness test
The tensile specimens were extracted as per the standard dimensions of ASTM E8M-03, by wire cut electric discharge machining (EDM) in the transversal (perpendicular to the direction of welding) direction while maintaining the weld line in the middle of the tensile specimen ( figure 4). A uniaxial universal testing machine was used for conducting the tensile tests, at 1 mm min −1 strain rate. In Vicker's microhardness test, a 200 g load was applied to the specimen by means of a hardness testing machine. In order to assess the microhardness variation in the welded specimen, it was measured every 2 mm from the center line of the specimen up to 12 mm on both sides, i.e. −12, −10, −8, −6, −4, −2, 0, 2, 4, 6, 8, 10 and 12 (all distances are in mm). The center of the stir zone was taken as a zero point when these readings were taken (figure 5).

Microstructure analysis
As part of the microstructure analysis, samples of the stir zone were extracted by wire cut EDM and then polished with 320, 800, 1000, and 2000 silicon grit carbide followed by cloth polishing by applying ferrous alumina. As soon as the samples had been prepared, they were then etched in the solutions of 10 ml HCl, 100 ml H 2 O, 1 g FeCl 3 on the Cu side and 75 ml H 2 O, 25 ml HNO 3 on the Al side by swabbing it on the transverse section. Microstructure, grain refinement, and elemental distribution were assessed using scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS).

Visual observation
After FSW of AA7075 and pure copper, the surface condition of the welds was inspected visually (figure 6). The results of the FSWed specimens that were subjected to different traverse and rotational speeds are presented in table 5. In samples 4, 7 and 8, excessive flash formation has been observed. This was due to the fact that there was    a very high generation of heat [28,29]. There was an issue with the quality of sample 3 welded at 80 mm min −1 traverse speed and 1200 rpm rotational speed. This was because of incorrect material flow due to small heat generation during the weld process [30]. Sound weld quality has been obtained at optimum process parameters for samples 1, 2, 5, 6 and 9. Due to poor quality weldings sample numbers 3, 4, 7 and 8 were rejected for further investigations.

Tensile strength and microhardness
Tensile test results for specimens selected after visual inspection are presented in table 6. Figure 7 further illustrates the stress and strain curves for samples. It was observed that S1, S2, S6, and S9 exhibited minor deviations in ultimate tensile strength. In contrast, S5 welded at 1400 rpm and 60 mm min −1 reported the ultimate tensile strength of 221.091 MPa and yield strength of 159.035 MPa. It also showed the highest maximum elongation of 9.335%. The ultimate tensile strength value is approximately 263% higher than that previously reported when the Cu was kept on the advancing direction [26]. Weld joint efficiency designated to the ratio of tensile strengths of welded joint to the base metal in identifying an acceptable welded joint. This calculation takes into account the lower value of tensile strength in both base metals. Consequently, the joint efficiency of dissimilar aluminium and copper is generally reported to be lower than 100% in the FSW process [31]. The maximum joint efficiency that was achieved in this experiment was approximately 95.05%. When it comes to FSW, the welding conditions have a prodigious impact on the joint's tensile strength [32,33]. The results of this study suggest that S5 was welded using the optimum parameters of the process. It has been observed that the thermomechanically affected zone (TMAZ) has a low strength, resulting in almost all the tensile specimens breaking at the TMAZ. Tensile test results of present study are compared with previous work in figure 8.
In this study, three specimens from each sample were tested for tensile strength. The average value of these three results was taken into account when calculating the tensile strength value.
In order to determine the microhardness of a sample, three specimens from each welded sample were checked and the average of these three results was taken into account. The graph for variations in microhardness for different samples is shown in figure 9. When compared to base metals, all of the samples had a much higher    High heat input causes increased flash generation S8 High heat input causes increased flash generation S9 Sound weld obtained microhardness value approaching the stir region. This is likely as an outcome of the existence of brittle and hard IMC in the stir region [24]. By recrystallizing and forming IMC, TMAZ increases hardness through grain refinement. Hardness decreases in heat affected zone (HAZ) as an outcome of the mild influence of recrystallization. A maximum microhardness was recorded in the stir zone of the S5 of 271 HV.

Microstructure examination
There is a very significant difference between the material flow profile in similar and distinct FSW. The onion ring pattern occurs frequently in similar FSWs [34]. While the microstructures of FSWed dissimilar materials are characterized by swirl patterns, lamella structures, and vortex types within the stir zones. These same patterns can also be found within TMAZs and the HAZs [31]. There are irregular-sized particles of Cu in the Al matrix that clearly visible as Cu islands. For the FSW of Al to Cu, it's quite challenging to differentiate the different sections of the joint in comparison to similar aluminium alloys. Additionally, the weld nugget exhibits very irregular characteristics. The microstructure results of samples 1 and 5 obtained by SEM are shown in figure 10. Cu particles can be observed to be scattered randomly throughout the Al matrix as well as forming a variety of Cu islands within it. It is possible that in the Al matrix of the stir zone large and brittle IMC were generated by randomly shaped Cu particles. This is due to an unusual plastic combination of both materials [9,35,36]. The IMC like CuAl, CuAl 2 , and Cu 9 Al 4 may be present in the weld nugget [9,36]. It is also evident from microstructure analysis that appropriate mixing between Cu and Al has taken place. Swirl and vortex-type material flow patterns were identified in the stir region. As visible in figure 10(f) and 11(l).
To determine the variations in the aluminium and copper content of the samples, EDS analysis was performed. The percentage distribution of aluminium and copper in the welded samples is shown in figure 11. EDS Distribution map shown in figure 12 also demonstrate that the appropriate mixing of Cu and Al has been achieved and the Cu particles were evident in the Al matrix. The predominant elements in the nugget region  were aluminium and copper as evidenced by elemental distribution. Figure 13 shows the EDS line analysis of Al and Cu for samples 1 and 5. Microstructure of the base metals can be seen in figure 14(a) and (b). Response of rotational and traverse speed on the grain size at weld zone of S1 (1200 rpm, 40 mm min −1 ), S2 (1200 rpm, 60 mm min −1 ) and S5 (1400 rpm, 60 mm min −1 ) are shown in figure 15(a), (b) and (c). Grain refinement can be noted in the stir region of the welded samples. It can be recognized that the average grain size in S1 is lesser than in S2, considering the higher  heat generation due to lower traverse speed. While finer grains can be observed in S5 as compared with S1, due to higher rotational speed. It may be concluded that the fine grains in the stir zone are therefore substantially influenced by rotational and traversal speeds. It was also resulted in previous studies [26,33].
It is universally acknowledged that grain size has a drastic influence on the mechanical characteristics of welded materials [17,19]. When the frictional heat continues to increase, the thermoplastic layer will gradually expand and flow, decreasing the pores within the weld and preventing the development of weld defects [27]. This defect free microstructure containing refined grains further contributes to the superb mechanical and microstructural characteristics. However, excessive heat can result in flash formation. In S5 finer grains were obtained at the optimum heat generation were resulted in excellent tensile strength and microhardness.

Fracture analysis
Using microscopic fractography to examine the fractured surfaces of the object could provide a tremendous amount of information about the nature of the fracture. SEM impressions of the fractured faces are shown in    detected that all the fractured surfaces had a dimpled surface appearance created by the coalescence of microvoid. This indicates that the failure mode of the material was ductile in nature [38,39]. There is a common feature found in failures at the weld zone that was excellent in terms of the mixing of the materials [40]. Also,  some of the dimples were larger and deeper than others which indicates a larger force before fracturing than shallow dimples. Fractography also revealed that considerable plastic deformation and necking had took place in all specimens before the fracture. Based on SEM fractography, it can be concluded that the fracture was primarily ductile under current welding conditions.

Conclusions
FSW has been successfully used to weld dissimilar AA7075 and copper plates, and the responses of process parameters on mechanical and metallurgical properties have been studied experimentally. This study has brought to the following conclusions: • The AA7075 and Cu plates with a thickness of 3 mm were successfully welded using a cylindrical pin profile tool made of H-13 tool steel with a pin diameter of 5 mm and a shoulder diameter of 16 mm.  • It was determined that it was possible to obtain sound welds by FSW with zero tool offset in which the AA7075 plates were placed on the advancing direction as opposed to the conventional procedure where high melting point materials are placed on the advancing direction when FSW is done with dissimilar metals.
• Due to the low amount of heat generated, at 80 mm min −1 of traverse speed and 1200 rpm tool of rotational speed, tunnel defects and inappropriate mixing of Al and Cu was detected. A high amount of flash formation and excessive melting was observed due to the high heat creation, at a traverse speed of 40 mm min −1 and a tool rotational speed of 1600 rpm, • According to the results of this study, the ultimate tensile strength of FSWed joints of AA7075 and Cu was found to be 221.091 MPa with maximum elongation of 9.335%. This was obtained at a traverse speed of 60 mm min −1 , a tool rotation speed of 1400 rpm and tool tilt angle of 2°. The maximum joint efficiency achieved was 95.05%. The tensile strength obtained in this FSW experiment is approximately 263% higher than the previously reported tensile strength which is obtained when the FSW is conducted with positioning the Cu on the advancing direction.
• Highest hardness of 271 HV was reported at the mid of the stir zone of the specimen welded at a traverse speed of 60 mm min −1 and tool rotation speed of 1400 rpm. An obvious cause of the hard stir zone was the existence of IMC.
• The specimens that failed in tensile tests were found to have failed because of ductile failure. This is because all fractured surfaces of the tensile specimens showed dimpled surfaces as a result of microvoid coalescence.
• In the weld zone, it was also evident that there had been a proper mixing of aluminium and copper, which was confirmed by the SEM analysis. In sample S5 excellent gain refinement was evident. Also the grain refinement was found increasing with increment in the heat input.