Vol.:(0123456789)1 3

Arabian Journal for Science and Engineering (2020) 45:1081–1095 https://doi.org/10.1007/s13369-019-04244-4

RESEARCH ARTICLE - MECHANICAL ENGINEERING

Investigation of Heat‑Assisted Dissimilar Friction Stir Welding of AA7075‑T6 Aluminum and AZ31B Magnesium Alloys

Musa Bilgin1 · Şener Karabulut1  · Ahmet Özdemir2

Received: 4 August 2019 / Accepted: 8 November 2019 / Published online: 12 November 2019 © King Fahd University of Petroleum & Minerals 2019

AbstractThis study investigates the successful joining methods of dissimilar metals of AZ31B Mg and AA7075-T6 Al alloys. The specimens were joined by using open-flame heating system and investigated the effect of the friction stir welding (FSW) process parameters on the welding quality, transverse rupture strength (TRS), tensile strength, microhardness, and microstruc-ture of the resulting specimens. The welding experiments were conducted based on the Taguchi mixed-orthogonal-array for tests, L16 (42 × 22), at a constant spindle speed of 500 rpm using a K10 grade carbide mixing tool with two different tilt angles (0° and 3°), and two tool tip geometries (square and triangle). The specimens were successfully welded, and the test results revealed that the welding temperatures and forces had the strongest effect on the welding quality. The achieved maximum tensile strength and TRS were approximately 122.1 MPa and 712.1 MPa, respectively, at a tool travel speed of 20 mm/min, tool offset value of 0.75 mm, tool tilt angle of 3°, and using a triangle tool tip. The microstructural analysis indicated that a dynamically recrystallized and grain refinement structure formed in the welded zone depending on the temperature and on the excessive deformation. The increase in the microhardness values depends on the intermetallic compounds ratio and fine-grained structure in the welded zone.

Keywords Friction stir welding · Heat-assisted dissimilar welding · Mechanical properties · Microstructure · AZ31B · AA7075-T6

1 Introduction

Magnesium (Mg) and aluminum (Al) alloys are widely used in construction and in the transportation vehicles industry and have significant application as engineering materials owing to their low specific density, good formability, good corrosion resistance, hot formability, recycling potential, and superior mechanical properties. Weight reduction is the most important issue in aerospace and automobile industry because the increasing customer demand for convenience,

luxury, performance, and safety is expected to result in the increase in the average weight of structural components in vehicles. Therefore, the utilization of light metals, such as magnesium and aluminum, is expected to be more impor-tant in industrial applications to address the weight problem and to provide improved fuel efficiency in the future [1–4]. The robust joining of aluminum and magnesium in a hybrid structure can provide a possible solution for the desired weight saving. Nevertheless, the joining of these dissimilar light structural materials poses several difficulties. There-fore, several researchers have studied the welding proper-ties of aluminum and magnesium alloys using conventional welding techniques such as electron beam, laser beam, and arc welding. As reported, intermetallic formations reducing the welding strength were observed at the joint area of the welded workpieces [5–9]. Dorbane et al. [2] investigated the welding behaviors of Al6061 and AZ31B and analyzed the microstructural, mechanical, and fracture properties of the joined specimens. They obtained better quality weld values by placing the aluminum on the advancing side (AS) during the welding process. The joint strength values were in the

* Şener Karabulut senerkarabulut@hacettepe.edu.tr

Musa Bilgin musabilgin@hacettepe.edu.tr

Ahmet Özdemir ahmetoz@gazi.edu.tr

1 Department of Mechanical Program, Hacettepe University, 06935 Ankara, Turkey

2 Department of Manufacturing Engineering, Gazi University, 06500 Ankara, Turkey

1082 Arabian Journal for Science and Engineering (2020) 45:1081–1095

1 3

range of 18–55% of the base materials, and the combined specimens fractured along the weld joint. Gao et al. [10] joined the Mg and Al alloys using plasma electrolytic oxida-tion coating. They achieved a tough plastic deformation at the middle of the joint interface. Mirsalehi et al. [11] stud-ied the mechanical and microstructural properties of FSW welded AZ31B-O Mg and AA6063-T4 Al alloys using TiO2 nanoparticles. They reported that the generated input had the remarkable effect on the microstructure and the parti-cle distribution of TiO2 and measured a joint strength of 132 MPa. Chen et al. [12] studied the deformation proper-ties of AA5083 and AA2024 alloys during the dissimilar FSW and reported the obtained failure and strain formation in the lowest hardness region. Li et al. [13] joined AZ31B magnesium alloys and investigated the joint strength and microstructural properties. The mixing tool rotation speed indicated a significant effect on the grain size and the joint strength and the lowest micro hardness values were meas-ured in the thermo-mechanically affected zone (TMAZ) on the AS. Yu et al. [14] performed dissimilar FSW of high-pressure die casting of A383 Al and AZ91 Mg alloys. They observed that a microstructure of intercalated layers in the stir zone was formed, and the maximum tensile strength was 93 MPa. Zhao et al. [15] welded the AZ31 Mg and Al6063 Al alloys underwater and investigated the mechani-cal properties and formation of intermetallic constituents in the mixing zone. They achieved a joint strength of up to 152.3 MPa and observed an intermetallic formation between the interface of magnesium and aluminum. Ji et al. [16] joined the AZ31 Mg and AA6061 Al alloys using the ultra-sonic technique and achieved a joint strength of 120 MPa. Hernández-García et al. [17] studied the FSW of AA7075-T6 and AZ31 B-H24 alloys. The measured tensile strength and microhardness values were 61.35 MPa and 122 HV, respectively. They noticed that intermetallic compounds, cavities, and tunnel defects were observed in the weld line. Farahani et al. [18] investigated the influence of the embed-ded silicon carbide nanoparticles (SiC) in the stir zone on the FSW of AZ31 Mg and AA7075 Al alloys. Authors reported a maximum measured joint strength of ~ 106 MPa, and hard and brittle intermetallic phases were observed in the mixing zone. Lambiase et al. [19] joined the AA6082-T6 aluminum alloys using FSW and investigated the effect of the forces and temperature distributions. They observed that the tem-peratures increased and fluctuations occurred in the main forces due to the preheating process during the tests. Yan et al. [20] fabricated a composite plate using AA7075 alu-minum and AZ31B magnesium alloys by explosive welding technique and investigated the microstructural properties and interface bonding effects of the obtained specimens. They reported maximum bending and shear strengths of 670 MPa and 70 MPa, respectively. In addition, a number of research-ers have studied the dissimilar FSW properties of AZ31B

and different aluminum alloy series and investigated the effect of welding parameters on the joint strength, interme-tallic formations, microhardness values, grain refinements, and corrosion resistance [21–28].

A limited number of the experimental studies have been conducted in FSW of AZ31B Mg and AA7075-T6 Al alloys. Therefore, in the present study, the joining behaviors of AA7075-T6 aluminum and AZ31B magnesium alloys were studied using a friction stir welding (FSW) method. A novel special apparatus and an open-flame heating system were used during the FSW process to improve the welding quality. The forces and temperatures were continuously monitored during the tests and determined the relationship between the forces, temperatures, process parameters, TRS, tensile strength, microhardness, microstructure and joint quality of the obtained specimens.

2 Experimental Procedures

2.1 Fixture Design

Figure 1a, b shows the special apparatus designed and fab-ricated to combine the AA7075-T6 and AZ31B magnesium alloys. The axial and feed forces were measured by load cells and monitored using the developed fixture apparatus during the welding process. Two pieces of S-type load cell and one flat-type load cell were used to obtain the axial and feed forces, respectively (Fig. 1). The top and bottom plates of the fixture apparatus were made of aluminum material to ensure equal heat dissipation on the welding area and to prevent hot crack formation on the joint line. As shown in Fig. 1a, the top plate of the apparatus (label 13) fixed by a linear bearing (label 12) and a shaft (10) transmits the force directly to the S-type load cells (label 6). The stiffness of the fixture is maintained by linear sliders (label 3) positioned on the linear guides (label 2), shafts (label 10), and additional plates (label 4). The upper plate is released from the y-axis, and the force from the feed direction is directly transferred to the flat-type load cell (label 9). The compression springs (label 11) ensure a stable axial force minimizing the changes in the force arising during the FSW operation. During the welding tests, the force data obtained from the load cells were transferred to a computer and monitored in real time by a weight indicator.

2.2 FSW Parameters and Experimental Design

For the FSW study, AA7075-T6 Al and AZ31B Mg speci-mens with dimensions of 300 mm × 100 mm × 5 mm were used as the base materials. The chemical compositions and mechanical properties of these base materials are presented in Tables 1 and 2, respectively. K10 grade carbide was

1083Arabian Journal for Science and Engineering (2020) 45:1081–1095

1 3

chosen for the tool material, and two different types of the tools with triangular and square pin profiles with a shoulder diameter of 20 mm were fabricated for the tests. Every FSW test was repeated three times, and average values were cal-culated for the experimental results.

The experimental setup for FSW and the welding tool specifications are shown in Fig. 2. The FSW tests were per-formed on a robust universal vertical milling machine. The FSW parameters were determined based on the intensive preliminary experiments and previous research results [29]. Selected welding variables and their levels are presented in Table 3. A Raytek MI3 2M digital infrared pyrometer was mounted on a milling machine table at a distance of 30 cm from the welding point for the noncontact temperature measurements during FSW. The beam of the MI3 sensor was focused on the contact point, and the surface tempera-ture of the specimen and tool was continuously monitored

during the tests. A Dremel Versaflame 2200 soldering iron with 75 min operating time and open-flame temperature of 1200 °C was used in the experiments. Before starting the friction mixing welding process, heating was performed on the AA7075-T6 side because of the higher mechanical val-ues of the aluminum alloy. The surface of the base materials was cleaned using absolute alcohol and then firmly clamped by fishplates on the top plate of the fixture apparatus. The AA7075 aluminum and AZ31B magnesium alloys were fixed to the fixture apparatus on the AS and on the retreat-ing side (RS), respectively. Firstly, the milling machine was adjusted according to the test parameters at a constant spin-dle speed of 500 rpm. Then, the mixing tool was positioned in a plunge depth of 4.8 mm and kept in the same position for 4 min to generate sufficient heat. The open-flame heating system was activated 1 min before the tool was moved. The waiting and activating times for the heating system of the mixing tool were determined during the pretesting according to the generated axial force and temperature.

2.3 Evaluation of Microstructural and Mechanical Properties

The radiographic measurements were taken using a Yxlon MGC41 X-ray system in accordance with EN 17636-2 to determine the invisible weld defects in the mixing zone of

Fig. 1 Technical drawings and images of the fixture apparatus used in the FSW tests

Table 1 Chemical composition of AA7075-T6 and AZ31B (wt%)

Elements Si Fe Cu Mn Ni Zn Ti Cr Ca Al Mg

AA7075-T6 0.07 0.12 1.39 0.059 0.0039 5.63 0.026 0.18 – Bal. 2.45AZ31B 0.01 0.004 < 0.0005 0.29 0.0013 0.75 – – < 0.005 2.9 Bal.

Table 2 Mechanical properties of AA7075-T6 and AZ31B (wt %)

Elements Ultimate tensile strength (MPa)

Elongation(%)

Yield strength (MPa)

Hardness(HV)

AA7075-T6 580 13 510 195AZ31B 240 7 140 68

1084 Arabian Journal for Science and Engineering (2020) 45:1081–1095

1 3

the combined samples. After the radiographic analyses, the welded samples were subjected to a penetration test to deter-mine the surface defects. The surfaces of the specimens were cleaned by BT-69 cleaning spray, and BT-68 liquid penetrant and BT-70 developer sprays were applied to the welded zone. For the microstructural investigation, the specimens were etched in three stages using the method by Firouzdor and Kou [30]. The macro- and microstructural studies were performed using a Leica DM4000 and JEOL JSM-6060 LW scanning electron microscope (SEM).

The cutting layout of mechanical test samples and the dimension of the specimens for TRS, tensile strength, and hardness test are shown in Fig. 3. Three mechanical tests samples for each experiment were cut using a Mitsubishi MV1200S wire electrical discharge machining according to the ASTM E8/E8M-09 standard. The tensile and TRS val-ues of the welded specimens were determined utilizing an Instron 3369 universal testing machine. The Vickers micro-hardness tests were conducted along the joint line using a DuraScan-G5 microhardness measurement device with an interval of 0.5 mm between the 10-mm AS and the 10-mm RS.

3 Experimental Results and Discussion

In preliminary experiments, the dissimilar FSW behavior of Mg and Al was investigated based on 27 experiments. However, a successful weld joint between the Mg and Al alloys could not be achieved. Large tunnel defects, porosity errors, and cracks along the weld joint were observed, and lower mechanical values were measured. The radiographic and microstructural analyses indicated that the desired mate-rial flow was sufficient for heat generation during the tests. Therefore, specimens were subjected to a heating process using an open-flame device to increase the material flow during the welding process. In the present study, the 16 FSW experiments were conducted based on the Taguchi L16 experimental design. The forces and temperature generated depending on the FSW parameters were monitored real-time and visually inspected their effect on the surface appearance quality of the joint line during FSW.

3.1 Effect of Axial and Feed Forces

The mean values of three different axial force, x-axis feed force, and surface temperature of the specimens are provided in Table 4. Images of the surfaces and nondestructive results and functions of the forces and temperatures for selected welded specimens are shown in Figs. 4 and 5. In these fig-ures, the temperature graphs show the surface temperature as a function of time during the welding process, and the axial and x-axis feed force graphs show the real-time monitored force values as a function of time measured from the begin-ning to the end of the FSW. A high or a deep point on the force graphs clearly corresponds to a welding defect on the

Fig. 2 Experimental setup for FSW (a) and specification of the welding tool (b)

Table 3 FSW parameters and test levels

Variables Unit L1 L2 L3 L4

Tool spindle speed rpm 500Welding speed mm/min 20 22 24 26Tool offset mm 0 0.25 0.5 0.75Tool tilt angle degree 0° 3°Tip profile Triangle Square

1085Arabian Journal for Science and Engineering (2020) 45:1081–1095

1 3

macroimage of the specimen; penetration and radiographic tests were performed at the same point. The macroscopic images and nondestructive examination results of the welded specimens indicate that there is a relationship among the welding quality, the axial force, the x-axis feed force, and the welding temperature. The forces and material flow in the welding zone were affected by the welding variables, and the plastic deformation was influenced by the welding tempera-ture. The axial forces varied in the range of 6650–11,500 N, and the acceptable welding surface quality was obtained

under an axial force of 8000 N. The differences in the weld-ing forces are due to the increase in the material flow stress at lower process temperatures. It was observed that instanta-neous changes occurred in the forces and temperatures when the mixing tool was located in the middle of the weld joint during the experiments. Based on visual inspection, smooth surfaces were observed during the tests. However, the radio-graphic analysis showed that the force and temperature fluc-tuations were due to the change in the microstructure of the stir zone, which prevented the formation of a homogeneous

Fig. 3 Dimensions of specimens for a tensile, b TRS, and c hardness tests

Table 4 FSW parameters and x-axis feed force, axial force, and welding temperature values

Test number Tool travel speed (mm/min)

Tool offset (mm)

Tip profile Tool tilt angle(°)

x-axis feed force(N)

Axial force (N) Welding temperature (°C)

T1 20 0.00 Square 0 2136.77 10,044.37 315.12T2 20 0.25 Square 0 2193.43 9945.59 316.45T3 20 0.50 Triangle 3 1594.68 7100.00 319.56T4 20 0.75 Triangle 3 1478.42 6649.93 321.24T5 22 0.00 Triangle 0 2028.59 8540.00 310.64T6 22 0.25 Triangle 0 1961.00 8260.00 309.69T7 22 0.50 Square 3 2562.62 9093.13 319.86T8 22 0.75 Square 3 2870.27 9088.57 320.91T9 24 0.00 Square 3 3151.124 10,075.94 312.43T10 24 0.25 Square 3 3106.23 10,100.00 314.47T11 24 0.50 Triangle 0 2570.65 8989.58 308.13T12 24 0.75 Triangle 0 2550.70 8778.00 309.45T13 26 0.00 Triangle 3 2720.04 9652.40 309.47T14 26 0.25 Triangle 3 2941.25 8823.48 310.37T15 26 0.50 Square 0 4002.76 11,500.00 304.30T16 26 0.75 Square 0 3949.40 11,460.00 303.55

1086 Arabian Journal for Science and Engineering (2020) 45:1081–1095

1 3

mixture. Tunnel defects, porosities cracks, and micro-voids were seen in the stir zone along the weld line. When the axial and feed force graphs, temperature graph, macroscopic image of the welded region, and the penetrant test result of the specimens are evaluated together, the defects on the welded zone can be attributed to sudden changes in the axial force, in the feed force, and in the temperature.

An onion-ring structure was formed and the appear-ance of the welded surface was completely smooth when lower travel speeds, higher tool offset values, and triangle pin tool geometry were employed as process parameters. In such cases, more stable welding forces were obtained, resulting in an optimal heat generation in the stir zone. The optimal heat generation facilitated the material flow, and the defects on the welding surface, such as porosities, cracks, and micro-voids, were eliminated (Fig. 4). Although a bet-ter surface appearance was achieved on the welded area, the radiographic analysis showed that tunnel defects appeared under certain FSW conditions due to the unfavorable heat generation. A tunnel formation was observed at a tool offset of 0.50 mm, welding speed of 20 mm/min, tilt angle of 3°

when triangle tool pin profile was used. However, the size of the resulting tunnel defect decreased, which improved the plastic deformation depending on the improvement of the heat generation. It was also observed that the tunnel defects were located on the magnesium side (RS) owing to the atomic lattice of the Mg alloy. The plastic deforma-tion behavior of Mg alloys is strongly limited to aluminum alloys because of the hexagonal close packed atomic lat-tice. However, the ductility of magnesium alloys increases with the increase in the temperature values and the restricted slip systems can be activated [31]. This suggests that when certain test parameters were applied, the desired welding temperatures did not generate a plasticized flow of the Mg alloy and resulted in tunnel defects.

The experiment results showed that when the mixing tool was positioned in the middle of the joint line without any given tool offset value, higher fluctuations occurred in the axial force, feed force, and temperature and consequently resulted in irregularities in the welding zone such as the formation zigzags, burrs, and voids. The welding tempera-ture decreased with the increase in the welding speed, and

Fig. 4 Surface appearance, penetration test result, and radiographic image of specimens for experiments T3

1087Arabian Journal for Science and Engineering (2020) 45:1081–1095

1 3

the lower heat generation resulted in an excessive built-up edge (BUE) formation on the mixing tool during the welding process [32] The BUE was formed on the mixing tool at a welding speed of 26 mm/min, and the surface appearance of the weld joint and the mechanical properties deteriorated. Therefore, these specimens and welding parameters are not suitable to be considered as weld joints of dissimilar FSW.

3.2 Microstructural Properties

Figure 6 depicts microstructure and SEM images of the welded zone of test specimen T4. This specimen was used for the measurement of the maximum joint strength, and the welding process was performed at a tool travel speed of 20 mm/min, tool offset value of 0.75 mm, tool tilt angle of 3°, and using a triangle pin tool geometry. At the top of the mixing zone, the formation of a lamellar shear bands inter-calated structure due to the heavy plastic deformation dur-ing the FSW was observed (Fig. 6a). The flow of the mate-rial was generated in the direction of the tool rotation and formed a complex intercalated structure and vortex struc-ture in the stir zone. The resulting flow structures, such as lamellar shear bands, contributed to the improvement of the TRS and joint strength of specimens. It could be observed

that AZ31B Mg and AA7075-T6 Al alloys in the stir zone were mixed and achieved a perfect metallurgical interface bonding.

No welding defects were seen on the welding area, and a better surface appearance was obtained. However, an inner cavity was seen at the bottom side of the mixing zone near the weld root, which was not formed along the weld joint such as a tunnel defect. The formation of the cavity defect could be considered to the decreasing heat input at the weld root and to the insufficient heat generation required at the bottom side of the welded zone. The plastic deformation of magnesium is lower than that of aluminum under lower heat generation; thus, the magnesium alloy exhibits a slow move-ment around the mixing tool, resulting in a cavity defect in the mixing zone [11, 15]. On the EDS image shown in Fig. 6b, a formation of Al12Mg17 intermetallic phases can be observed between the stirring zone and base material. These hard and brittle intermetallic phases are also very sig-nificant deteriorating factors of the mechanical properties of the welded specimens in FSW of Mg and Al alloys. It can be considered that the cavity defect and intermetallic formation decrease the tensile strength and TRS values.

Figure 7 shows cross-sectional images of the stir zone at a tool welding speed of 22 mm/min, offset value of 0.5 mm,

Fig. 5 Surface appearance, penetration test result, and radiographic image of specimens for experiment T4

1088 Arabian Journal for Science and Engineering (2020) 45:1081–1095

1 3

Fig. 6 Optical and scanning electron microscope (SEM) micrographs of the stir zone for experiment T4 and energy-dispersive spectroscopy (EDS) results

1089Arabian Journal for Science and Engineering (2020) 45:1081–1095

1 3

Fig. 7 The micrographs of the stir zone for experiment T7 and EDS results

1090 Arabian Journal for Science and Engineering (2020) 45:1081–1095

1 3

tool tilt angle of 3°, and using a square tool tip. A uniform and better mixing over the middle of the stir zone and the formation of onion-ring structures in the banded zone can be seen in Fig. 7. It was observed that the flow of the material was more effective under the shoulder diameter of the tool and the grain boundaries moved in the direction of rotation of the tool (Fig. 7a). Compared to the magnesium side, a more intensive plastic deformation was obtained on the alu-minum side due to the much higher heat generation on the AS (AA7075) than on the RS (AZ31B). Tunnel and cracks defects were observed to be formed in the bottom side of the welded joint, which resulted in a decrease in the joint strength of specimens. The tunnel defect occurred along the joint weld due to the insufficient heat input during FSW. Figure 7d, e shows the interface structure of the stir zone and the magnesium side of the welded specimens. It can be clearly seen that the microstructures of the TMAZ on the Mg side consist of different grain sizes. The excessive plastic deformation and high temperature during dissimilar FSW resulted in grain refinement in the TMAZ near the stir zone on the Mg side, resulting in dynamic recrystallization [33]. Thus, a resistance can arise in this zone against the plastic deformation. As a result, poor interface bonding between the mixing zone and base material occurred, resulting in decreased tensile strength and increased microhardness.

As can be seen in Fig. 8, an unsuccessful joint weld was obtained due to the reduction in the welding temperature resulting from the increase in the welding speed. Although the appearance of the welded surface was smooth in certain experiments, it was observed that a tunnel defect formed through the weld joint due to the inadequate material flow and heat input at the weld root during FSW (Fig. 8(c)). A thin open crack from the top surface of the weld joint to weld root formed along the joining interface and the Al12Mg17 continuous intermetallic layer could be seen a liquated struc-ture resulting from the lack of interlocking during the pro-cess [34]. It should be noted that this formation results in a weak interface bonding between the mixing zone and the base material and decreases the joint tensile strength [35].

3.3 Mechanical Properties

3.3.1 Tensile Strength and TRS

The TRS and tensile strength measurements were repeated three times at each level of the tests, and the mean values were calculated. All tensile specimens were fractured from the magnesium side in the TMAZ, and the joint strength values were measured in the range of 20–122.06 MPa. As shown in Fig. 9, the measured maximum tensile strength, elongation, and TRS were 122.06  MPa, 1.28%, and 712.1 MPa, respectively. The measured joint strength and TRS values were approximately 50.85% and 57.62% of the

base material of AZ31B magnesium alloy, respectively. The maximum tensile strength of the welded specimen improved by 15% rate compared to studies in the existing literature [17, 18, 20] and this improvement can be attributed to the open-flame heating system during the dissimilar FSW of AZ31B magnesium and AA7075-T6 aluminum alloys. As shown in Fig. 6, intercalated structures formed in the mixing zone, which positively affected the tensile strength of the joint. The decrease in the TRS and tensile strength values for the dissimilar FSW specimens is due to the lower duc-tility of the welded specimen compared to that of the base alloys. The hardness values increased at the welded line, which decreased the toughness of the specimens due to the intermetallic formations. Al3Mg2 and Al12Mg17 intermetallic compounds were observed in the transition region, which affected the mechanical properties of the joints due to their brittle and hard structures. The intermetallic compounds, micro-cracks, porosities, and tunnel failures on the joint line resulted in the decrease in the TRS and joint strength. How-ever, the mechanical interlocking in the mixing zone could result in the improvement of the joint strength [31].

The liquefaction in the mixing zone increased with the increase in the generated heat, and the higher heat input indicated a negative effect on the material flow during the FSW process. The formation of liquid films along the Mg/Al interface and grain boundaries results in a decrease in the weld strength of Al–Mg interface due to crack formation. Similar findings have been seen in previous studies in the literature [30, 36–38]. The rise in the heat input results in a reduction in the mechanical locking effect in the mixing zone due to the increase in the thickness of the intermetallic constituents and cracks (Fig. 8). Thus, the lack of interlock-ing between the Al and Mg results in insufficient material flow and low tensile strength of weld joint. Depending on the FSW parameters, axial force, x-axis feed force, and welding temperature, the TRS of the welded specimens exhibited different behaviors. The welded specimens exhibited a brittle behavior due to the Al3Mg2 and Al12Mg17 intermetallic for-mations under the three-point bending test, and the measure-ment results were obtained in the range of 82.2–712.1 MPa.

3.3.2 Microhardness

Figure 10 shows the maximum hardness values of all stud-ied welded specimens. The measured maximum microhard-ness values were 256 HV and 240 HV on the Al and the Mg side at distances of 1.5 and − 2.0 mm from the weld center, respectively. The hardness measurements were taken at the joint line of the welded specimens between − 10 and 10 mm on each side of the weld joint. It was found that the transition of the microhardness values from the aluminum to the magnesium side depended on the tool pin diameter and the revolution direction from Al to Mg, in addition to

1091Arabian Journal for Science and Engineering (2020) 45:1081–1095

1 3

the occurrence of intermetallic phases and grain refine-ments. Figure 11 depicts the microhardness distribution on the weld joint for FSW experiments T1–T4. The measured

microhardness values on the AZ31B side (BM) of the weld joint were in the range of 50–100 HV and on the AA7075 side (BM) were in the range of 100–155 HV.

Fig. 8 The micrographs of the stir zone for experiment T13 and EDS results

1092 Arabian Journal for Science and Engineering (2020) 45:1081–1095

1 3

The tensile strength and TRS test results of the welded specimens indicated the occurrence of a brittle fracture. Thus, the higher hardness values confirmed that the num-ber of brittle phases in the mixing zone was high and con-sequently resulted in brittle fracture [24]. Similar brittle fracture behavior has also been observed at the welding of superalloy materials [39]. Moreover, AA7075 and AZ31B alloys were sufficiently mixed especially at lower tool trav-erse speeds and exhibited a lamellar shear bands intercalated structure. This lamellar structure resulted in the variations in microhardness values in the stir zone. Irregular hardness fluctuations were observed in the mixing zone and higher hardness values were measured on the AS of the weld joint compared with those on the RS because the larger plastic deformation resulted in grain refinement and equaled grain structure. The microhardness values are strongly decreased in the TMAZ and in the heat-affected zone (HAZ) on both the aluminum and the magnesium sides due to the increasing

grain size combined with the effects of the frictional heat input and thermal softening [13].

4 Conclusions

Open-flame heat-assisted dissimilar FSW process of Mg and Al alloys was presented in this study and investigated the effects of the FSW process parameters on the welding qual-ity. The conclusions can be summarized as follows:

• The sudden changes in the axial force, x-axis feed force, and temperatures resulted in porosities and prevented the homogenous mixture in the weld joint. The x-axis feed force values were measured in the range of 1478.42–4002.76 N, the axial force values were in the range of 6303.55–11500 N, and the temperature values in the range of 321.24 °C and 649.93 °C.

• The increase in the tool travel speed from 20 mm/min to 26 mm/min resulted in an increase in the axial and feed forces and decreased the temperature during the weld-ing process, resulting in lower mechanical properties and welding quality of the specimens.

• The stable axial forces generated optimal heat input sup-ported by the open-flame heating system. The plastic deformation was improved, and the welding defects were considerably decreased under suitable welding tempera-tures.

• All tensile specimen exhibited brittle fracture behav-ior, and the fractures occurred at the TMAZ along the welded joint of the Mg side. The tensile strength and TRS test values were obtained between the range of 20–122.06 MPa and 82.2–712.1 MPa, respectively. Among the investigated FSW process parameters, the optimal tensile strength and TRS values were achieved

Fig. 9 Tensile strength, TRS, and elongation values for the welded specimens

Fig. 10 Maximum hardness values of the weld joint for all specimens

1093Arabian Journal for Science and Engineering (2020) 45:1081–1095

1 3

at a tool travel speed of 20 mm/min, tool offset value of 0.75 mm, tool tilt angle of 3°, using a triangle pin tool geometry.

• The highest tensile strength, elongation, and TRS values were obtained as 122.06  MPa, 1.28%, and 712.1 MPa, respectively. The joint strength efficiency and TRS values were approximately 50.85% and 57.62% of the base material of AZ31B magnesium alloy, respectively, while the microhardness value was 164 HV.

• The variation in the microhardness in the stir zone was mainly affected by the microstructural formations, and the hardness values were in the range of 131–256 HV on the Al and the Mg side at distances of 1.5 and − 2.0 mm from the weld center, respectively.

Acknowledgements The author wishes to thank Gazi University Sci-entific Research Projects Coordination Unit for the financial support of this study [Grant Number 59/2016-01].

References

1. Baghdadi, A.H.; Rajabi, A.; Mohamad Selamat, N.F.; Sajuri, Z.; Omar, M.Z.: Effect of post-weld heat treatment on mechanical behaviour and dislocation density of FSWed Al 6061. Mater. Sci. Eng. A 754, 728–734 (2019). https ://doi.org/10.1016/j.msea.2019.03.017

2. Dorbane, A.; Mansoor, B.; Ayoub, G.; Shunmugasamy, V.C.; Imad, A.: Mechanical, microstructural and fracture properties of dissimilar welds produced by friction stir welding of AZ31B and Al6061. Mater. Sci. Eng. A 651, 720–733 (2016). https ://doi.org/10.1016/j.msea.2015.11.019

Fig. 11 Microhardness graphs obtained from the experiments T1, T2, T3, and T4

1094 Arabian Journal for Science and Engineering (2020) 45:1081–1095

1 3

3. Liu, L.; Ren, D.; Liu, F.: A review of dissimilar welding tech-niques for magnesium alloys to aluminum alloys. Materials (Basel) 7, 3735–3757 (2014). https ://doi.org/10.3390/ma705 3735

4. Karam, A.; Mahmoud, T.S.; Zakaria, H.M.; Khalifa, T.A.: Friction stir welding of dissimilar A319 and A413 cast aluminum alloys. Arab. J. Sci. Eng. 39, 6363–6373 (2014). https ://doi.org/10.1007/s1336 9-014-1220-6

5. Liu, P.; Li, Y.; Geng, H.; Wang, J.: Microstructure characteristics in TIG welded joint of Mg/Al dissimilar materials. Mater. Lett. 61, 1288–1291 (2007). https ://doi.org/10.1016/j.matle t.2006.07.010

6. Ben-Artzy, A.; Munitz, A.; Kohn, G.; Bronfin, B.; Shtechman, A.: Joining of light hybrid constructions made of magnesium and aluminum alloys. In: TMS Annual Meeting, pp. 295–302 (2002)

7. Chi, C.T.; Chao, C.G.; Liu, T.F.; Lee, C.H.: Aluminum element effect for electron beam welding of similar and dissimilar magne-sium-aluminum-zinc alloys. Scr. Mater. 56, 733–736 (2007). https ://doi.org/10.1016/j.scrip tamat .2007.01.028

8. Bannour, S.; Abderrazak, K.; Mattei, S.; Masse, J.E.; Autric, M.; Mhiri, H.: The influence of position in overlap joints of Mg and Al alloys on microstructure and hardness of laser welds. J. Laser Appl. 10, 10 (2013). https ://doi.org/10.2351/1.47926 15

9. Gao, M.; Mei, S.; Li, X.; Zeng, X.: Characterization and formation mechanism of laser-welded Mg and Al alloys using Ti interlayer. Scr. Mater. 67, 193–196 (2012). https ://doi.org/10.1016/j.scrip tamat .2012.04.015

10. Gao, Y.; Morisada, Y.; Fujii, H.; Liao, J.: Dissimilar friction stir lap welding of magnesium to aluminum using plasma electrolytic oxidation interlayer. Mater. Sci. Eng. A 711, 109–118 (2018). https ://doi.org/10.1016/j.msea.2017.11.034

11. Asl, N.S.; Mirsalehi, S.E.; Dehghani, K.: Effect of TiO2 nano-particles addition on microstructure and mechanical properties of dissimilar friction stir welded AA6063-T4 aluminum alloy and AZ31B-O magnesium alloy. J. Manuf. Process. 38, 338–354 (2019). https ://doi.org/10.1016/j.jmapr o.2019.01.023

12. Niu, P.L.; Li, W.Y.; Vairis, A.; Chen, D.L.: Cyclic deformation behavior of friction-stir-welded dissimilar AA5083-to-AA2024 joints: effect of microstructure and loading history. Mater. Sci. Eng. A 744, 145–153 (2019). https ://doi.org/10.1016/j.msea.2018.12.014

13. Li, W.; Niu, P.L.; Yan, S.R.; Patel, V.; Wen, Q.: Improving micro-structural and tensile properties of AZ31B magnesium alloy joints by stationary shoulder friction stir welding. J. Manuf. Process. 37, 159–167 (2019). https ://doi.org/10.1016/j.jmapr o.2018.11.014

14. Li, P.; You, G.; Wen, H.; Guo, W.; Tong, X.; Li, S.: Friction stir welding between the high-pressure die casting of AZ91 magne-sium alloy and A383 aluminum alloy. J. Mater. Process. Tech-nol. 264, 55–63 (2019). https ://doi.org/10.1016/j.jmatp rotec .2018.08.044

15. Zhao, Y.; Lu, Z.; Yan, K.; Huang, L.: Microstructural characteriza-tions and mechanical properties in underwater friction stir welding of aluminum and magnesium dissimilar alloys. Mater. Des. 65, 675–681 (2015). https ://doi.org/10.1016/j.matde s.2014.09.046

16. Ji, S.; Meng, X.; Liu, Z.; Huang, R.; Li, Z.: Dissimilar friction stir welding of 6061 aluminum alloy and AZ31 magnesium alloy assisted with ultrasonic. Mater. Lett. 201, 173–176 (2017). https ://doi.org/10.1016/j.matle t.2017.05.011

17. Hernández-García, D.; Saldaña-Garcés, R.; García-Vázquez, F.; Gutiérrez-Castañeda, E.J.; Deaquino-Lara, R.; Verdera, D.: Fric-tion stir welding of dissimilar AA7075-T6 to AZ31B-H24 alloys. MRS Adv. 2(64), 4055–4063 (2017). https ://doi.org/10.1557/adv.2017

18. Tabasi, M.; Farahani, M.; Givi, M.K.B.; Farzami, M.; Mohar-ami, A.: Dissimilar friction stir welding of 7075 aluminum alloy to AZ31 magnesium alloy using SiC nanoparticles. Int. J. Adv. Manuf. Technol. 86, 705–715 (2016). https ://doi.org/10.1007/s0017 0-015-8211-y

19. Lambiase, F.; Paoletti, A.; Di Ilio, A.: Forces and tempera-ture variation during friction stir welding of aluminum alloy AA6082-T6. Int. J. Adv. Manuf. Technol. 99, 337–346 (2018). https ://doi.org/10.1007/s0017 0-018-2524-6

20. Yan, Y.B.; Zhang, Z.W.; Shen, W.; Wang, J.H.; Zhang, L.K.; Chin, B.A.: Microstructure and properties of magne-sium AZ31B-aluminum 7075 explosively welded composite plate. Mater. Sci. Eng. A 527, 2241–2245 (2010). https ://doi.org/10.1016/j.msea.2009.12.007

21. McLean, A.A.; Powell, G.L.F.; Brown, I.H.; Linton, V.M.: Fric-tion stir welding of magnesium alloy AZ31B to aluminium alloy 5083. Sci. Technol. Weld. Join. 8, 462–464 (2003). https ://doi.org/10.1179/13621 71032 25009 134

22. Somasekharan, A.C.; Murr, L.E.: Microstructures in friction-stir welded dissimilar magnesium alloys and magnesium alloys to 6061-T6 aluminum alloy. Mater. Charact. 52, 49–64 (2004). https ://doi.org/10.1016/j.match ar.2004.03.005

23. Verma, J.; Taiwade, R.V.; Reddy, C.; Khatirkar, R.K.: Effect of friction stir welding process parameters on Mg-AZ31B/Al-AA6061 joints. Mater. Manuf. Process. 33, 308–314 (2018). https ://doi.org/10.1080/10426 914.2017.12919 57

24. Yan, J.; Xu, Z.; Li, Z.; Li, L.; Yang, S.: Microstructure char-acteristics and performance of dissimilar welds between mag-nesium alloy and aluminum formed by friction stirring. Scr. Mater. 53, 585–589 (2005). https ://doi.org/10.1016/j.scrip tamat .2005.04.022

25. Lee, K.J.; Kwon, E.P.: Microstructure of stir zone in dissimilar friction stir welds of AA6061-T6 and AZ31 alloy sheets. Trans. Nonferrous Metals Soc. China (Engl. Ed.) 24, 2374–2379 (2014). https ://doi.org/10.1016/s1003 -6326(14)63359 -8

26. Fu, B.; Qin, G.; Li, F.; Meng, X.; Zhang, J.; Wu, C.: Friction stir welding process of dissimilar metals of 6061-T6 aluminum alloy to AZ31B magnesium alloy. J. Mater. Process. Technol. 218, 38–47 (2015). https ://doi.org/10.1016/j.jmatp rotec .2014.11.039

27. Malarvizhi, S.; Balasubramanian, V.: Influences of tool shoul-der diameter to plate thickness ratio (D/T) on stir zone formation and tensile properties of friction stir welded dissimilar joints of AA6061 aluminum-AZ31B magnesium alloys. Mater. Des. 40, 453–460 (2012). https ://doi.org/10.1016/j.matde s.2012.04.008

28. Kwon, Y.J.; Shigematsu, I.; Saito, N.: Dissimilar friction stir weld-ing between magnesium and aluminum alloys. Mater. Lett. 62, 3827–3829 (2008). https ://doi.org/10.1016/j.matle t.2008.04.080

29. Bilgin, M.; Karabulut, Ş.; Özdemir, A.: Study on the mechani-cal properties of dissimilar friction stir welding of AA 7075 T6 and AZ31B Alloys. In: Proceedings of the 2018 9th Interna-tional Conference on Mechanical and Aerospacce Engineering ICMAE 2018, pp. 467–471 (2018). https ://doi.org/10.1109/icmae .2018.84676 79

30. Firouzdor, V.; Kou, S.: Formation of liquid and intermetallics in Al-to-Mg friction stir welding. Metall. Mater. Trans. A Phys. Metall. Mater. Sci. 41, 3238–3251 (2010). https ://doi.org/10.1007/s1166 1-010-0366-4

31. Mohammadi, J.; Behnamian, Y.; Mostafaei, A.; Gerlich, A.P.: Tool geometry, rotation and travel speeds effects on the properties of dissimilar magnesium/aluminum friction stir welded lap joints. Mater. Des. 75, 95–112 (2015). https ://doi.org/10.1016/j.matde s.2015.03.017

32. Elangovan, K.; Balasubramanian, V.; Valliappan, M.: Influences of tool pin profile and axial force on the formation of friction stir processing zone in AA6061 aluminium alloy. Int. J. Adv. Manuf. Technol. 38, 285–295 (2008). https ://doi.org/10.1007/s0017 0-007-1100-2

33. Venkateswaran, P.; Reynolds, A.P.: Factors affecting the properties of Friction Stir Welds between aluminum and magnesium alloys. Mater. Sci. Eng. A 545, 26–37 (2012). https ://doi.org/10.1016/j.msea.2012.02.069

1095Arabian Journal for Science and Engineering (2020) 45:1081–1095

1 3

34. Mofid, M.A.; Abdollah-Zadeh, A.; Gür, C.H.: Investigating the formation of intermetallic compounds during friction stir weld-ing of magnesium alloy to aluminum alloy in air and under liquid nitrogen. Int. J. Adv. Manuf. Technol. 71, 1493–1499 (2014). https ://doi.org/10.1007/s0017 0-013-5565-x

35. Sato, Y.S.; Park, S.H.C.; Michiuchi, M.; Kokawa, H.: Consti-tutional liquation during dissimilar friction stir welding of Al and Mg alloys. Scr. Mater. 50, 1233–1236 (2004). https ://doi.org/10.1016/j.scrip tamat .2004.02.002

36. Hou, Z.; Sheikh-Ahmad, J.; Jarrar, F.; Ozturk, F.: Residual stresses in dissimilar friction stir welding of AA2024 and AZ31: experi-mental and numerical study. J. Manuf. Sci. Eng. 140, 051015 (2018). https ://doi.org/10.1115/1.40390 74

37. Shah, L.H.; Othman, N.H.; Gerlich, A.: Review of research pro-gress on aluminium–magnesium dissimilar friction stir weld-ing. Sci. Technol. Weld. Join. 23, 256–270 (2018). https ://doi.org/10.1080/13621 718.2017.13701 93

38. Kou, S.; Firouzdor, V.: Al-to-Mg friction stir welding: effect of positions of Al and Mg with respect to the welding tool. Weld. J. 88, 213–224 (2009)

39. Ramkumar, T.; Selvakumar, M.; Narayanasamy, P.; Ayisha Begam, A.; Mathavan, P.; Arunraj, A.: Studies on the structural property, mechanical relationships and corrosion behaviour of Inconel 718 and SS 316L dissimilar joints by TIG welding with-out using activated flux. J. Mater. Process. Technol. 30, 290–298 (2017). https ://doi.org/10.1016/j.jmapr o.2017.09.028