Effect of Copper Donor Material Assisted Friction Stir Welding of AA6061-T6 Alloy On Downward Force, Microstructure, And Mechanical Properties

In this research, Copper (Cu) donor material assisted friction stir welding (FSW) of AA6061-T6 alloy was studied. Cu assisted FSW joints of AA6061-T6 alloy were prepared at a constant tool rotational rate of 1400 rpm and various welding speeds at 1 mm/s and 3 mm/s. The Cu donor material of different thickness (i.e., 20%, 40%, and 60%) with respect to the workpiece thickness was selected to assist the FSW joining at the plunge stage. It is observed that the downward force generated in the FSW process was gradually decreased after introducing Cu donor material with incremental thicknesses with respect to workpiece at the plunge stage. Post-weld analysis was characterized in terms of microstructure, and mechanical properties. The results of microstructure analysis at the stir zone (SZ) show the formation of ner grains due to dynamic recrystallization and plastic deformation. Microhardness tests reveal that the hardness decreased from the base metal (BM) to the SZ across the heat affected zone (HAZ) and thermo-mechanically affected zone (TMAZ). The lowest value of hardness appeared in the TMAZ and HAZ where tensile failure occurs. With increasing welding speed, the average hardness in the SZ decreased due to lower heat input and faster cooling rate. Tensile test plots show no signicant change in ultimate tensile strength with or without Cu donor material. Fractography of tensile tested samples shows both ductile and brittle like structure for given welding parameters. This proposed work of FSW with Cu donor material is promising to increase tool life due to the decrement of the downforce during plunge and throughout the welding stage. Meanwhile, the inclusion of donor material did not compromise the weld quality in terms of the mechanical properties and micro-hardness. that are hard to weld by conventional fusion welding. This paper provides the promising idea/approach to them on FSW applications of improving friction stir tool life without deteriorating the welded joints properties.


Introduction
Friction stir welding (FSW) is a novel solid-state joining process invented at The Welding Institute (TWI) of UK in 1991. It has been broadly used in aerospace, aircraft, shipbuilding, automotive, railways, and marine industries [1][2][3][4]. FSW has many advantages including lower heat input and avoidance evasion of phase change during the welding process in comparison to conventional fusion joining processes such as MIG or TIG welding. Since it is a solid state joining, this process does not cause material phase change, resulting in minimum microstructural changes and better mechanical properties than conventional welding processes [5][6][7][8]. NASA and aerospace manufactures are applying FSW to effectively and reliably join high-strength aerospace aluminum alloys such as Al 6061, Al 2024, and Titanium alloys that are hard to weld by conventional fusion welding. In the shipbuilding industry, FSW is used for joining high strength steel for hulls and decks.
A typical FSW process consists of four stages: plunge, dwell, welding, and retraction as shown in Fig. 1 (b). During the plunge stage, a non-consumable rotating tool's pin makes initial contact and penetrates the workpieces being joined. The plunge force and friction between the tool and the workpiece generate heat on the surrounding area of the tool, and result in viscous plastic deformation. The temperature keeps increasing, which softens the material to a paste-like state through thermo-mechanical deformation of the material around the pin. The axial load decreases and then enters the dwell stage. In the dwell stage, the tool shoulder establishes contact with the workpieces.
During the dwell stage, the tool can no longer maintain a vertical displacement. It rotates in place until the surrounding material reaches a stabilized temperature for the welding stage. The length of the dwell time is decided by the properties of the workpiece material and the design of tool bit. The welding stage is the actual joining process of the workpieces. The rotating tool advances along a prede ned welding path. As the tool moves along the welding path, viscous material is transferred from the advancing side (AS) of the pin to the retreating side (RS) and vice versa. As the tool passes along the welding path, the material cools and solidi es, resulting in a solid state joining of materials. The important control parameters in FSW are: rotation speed of the tool in rpm, traverse speed of the welding tool in mm/second, tilt angle, welding path, and tool design, etc. [9].
When applying the FSW in manufacturing of industrial parts, an issue hindering its wider application is that tool wear leads to high tooling cost when joining high strength materials such as steel and titanium alloys. The excessive wear in FSW results in frequent replacement of costly tools after an insu cient length of weld. The research team has proposed the idea of using donor materials in plunge stage to reduce the tool wear and increase tool life [10,11,12]. Due to equipment restriction, the previous research in [10][11][12] was performed on a manual mill; it only studied donor material's effect in the plunge stage of FSW. Through a literature research, there is no reported research that examined the effect of donor material on the welding stage and the joint properties of FSW. Initiated by these, a set of experiments were performed to study the Cu donor material in assisting FSW of Al 6061. The downward force, microstructure, mechanical properties, and micro-hardness in the Cu assisted FSW joining Al 6061 workpieces were examined at a combination of tool rotational rate of 1400 rpm and various welding speeds at 1 mm/s and 3 mm/s respectively. These experiments aim to address the research question whether donor material can assist welding stage in FSW processes and affect post-weld properties.
The rest of this paper is organized as follows. Section 2 reviews related research work on FSW process assisted by auxiliary energy and donor materials. A research gap is summarized at the end of this section. Section 3 discusses the details of the proposed experimentation techniques. Section 4 describes experiment results and discussion to illustrate the physical mechanism of donor materials assisted FSW joints formation. Section 5 concludes the research and outlines future direction.

Literature Review
The related literature was searched on FSW assisted by various types of auxiliary energy and donor materials. Wu et al provided a comprehensive review [ 1 3] on the methods that provides auxiliary energy. These methods majorly are using electricity, induction, laser, and ultrasonic vibration etc., which partially heat and/or soften the alloys before or during the FSW, reduce the load requirement on the tool. Thus, it improves the tool performance and life, optimizes process window, and enhances welding e ciency and quality, etc.
Electrically assisted FSW (EAFSW) softens materials by Joule heating effect and electroplastic effect [ 1 4, 1 5]. Due to both effects, the EAFSW process produced signi cantly higher temperatures in the plunge stage to soften the material, which reduced the plunge force signi cantly [14,15]. This would facilitate welding of thick part, improve the wear resistance of the tool, and increase the tool life and performance [14]. Luo et al applied EAFSW for joining light alloys of AZ31B and Al 7075 [16,17]. Morphologies of the weld line in [16] con rmed smoother surfaces with ner arc shaped features due to more uniform and denser corrugation of the material. It was also reported that the in uence of electric current would reduce the average grain size in the SZ [16]. Although the average grain size was almost unchanged in the TMAZ, the grain distribution was more uniform with more elongated grains due to severe plastic deformation. Also, it formed less narrow and more equiaxed grains in the TMAZ due to more severe dynamic recrystallization caused by electric current [16]. The difference in grain re nement in the SZ and TMAZ of the Al and Mg alloys is induced by the different responses of the alloys to nucleation rate of dynamic recrystallization under similar conditions of heat input and plastic deformation [17]. The higher is the nucleation rate of the dynamic recrystallization, the ner are the grains formed. The grains in the HAZ of EAFSW joints were coarsen with increasing electric current because of the long exposure of the HAZ to high temperature and secondary recrystallization. The hardness pro les of the EAFSW welds were consistent with the grain re nements and microstructure developments. The hardness values varied directly with the grain re nements. The higher heat inputs improved material ow. The viscoplasticity at the root of the weld improved the weld quality by reducing the risk of various defects in welds.
Induction assisted FSW (IAFSW) uses electrical inductance to generate heat to soften the workpiece materials.
IAFSW has been reported for joining medium strength Aluminium alloy [18] and high strength steel [19,20]. Both researchers in [18 and 19] reported that the downward force was reduced, but mechanical properties in the joints still maintained. Further, IAFSW would reduce the residual stress distortion and HAZ softening and increases the plastic deformation due to its low heat input ux and well de ned heating. This would eliminate the heavy clamping systems thus reducing the size of the FSW machine [18]. When joining steel, the IAFSW, comparing to the FSW, process induced more intense grain re nement [19], which induced an increase in the hardness and tensile strength of the SZ. However, the ferrite/austenite ratio of this steel remained unchanged for both processes.
Laser assisted FSW (LAFSW) is the most widely used external energy assisted FSW. The application of laser preheating lowered the resistance of the material to tool penetration and forward motion [21]. Therefore, the need to apply large forces on both the tool and the workpieces is lowered. When applying EAFSW to high strength steel with ferrite pearlite structure, signi cant amount brittle Martensite and Bainite phases were found in the joints, but the LAFSW could prevent the brittleness [21] and retain the original phases. LAFSW was even applied for joining of super alloys such as Inconel [ 2 2]. Laser preheat induced intense grain re nement in the SZ, so it signi cantly improved the overall strength and mechanical properties of Inconel alloy. An improvement was reported in the fatigue behavior of Aluminium lap joints produced by LAFSW [ 2 3]. Alvarez et al. also reported the life of a PCBN tool used in the LAFSW of marine grade steel was increased up to several feet [20].
Ultrasonic energy assisted FSW (UAFSW) uses high frequency vibration to soften material without signi cant heating it. UAFSW reduces the friction heat by complementing the softening. However, at the same time, it would increase the heat generation from the additional deformation. The overall result of these two effects may increase the heat input slightly. UAFSW expanded the plastically deformed region [24] and improved the material ow [25], thus it improved lap shear force and hardness of the joints [25]. UAFSW also suppressed the formation of voids and tunnel defects [24,26].
Numerical [10][11][12] and experimental [11][12] studies of donor materials that assisted FSW suggest that local preheating at the plunge stage is generated by friction between the tool pin and the donor material and also by the plastic deformation in the material. This preheating would generate heat in the copper donor material due to high frictional forces and this preheating is expected to be transferred to the work piece by conduction, which will result in softening the workpiece material and reducing the plunge force. This in turn is expected to reduce tool wear. For the selection of a donor material, it is desired for the donor material to possess high thermal conductivity to allow rapid heat transfer into the workpiece. Additionally, it is desired that the donor material will not advance along the welding path, because this will in uence the weld quality by producing a nonhomogeneous weld line. It has been demonstrated by previous studies that material moves around the tool pin and becomes deposited behind it without being transported forward with the tool's pin [27][28][29]. Hence, donor material from the plunge phase is restricted to the initial weld area and can be sacri ced.
Comparing to the auxiliary energy assisted FSW, the donor material assisted FSW does not require the welding tool to be electrically conductive, and it needs no installation of complex insulation system between the tool and the FSW machine. It also avoids the di culty in IAFSW on controlling the current ow, generating the spark, and heating of conductive materials involved in the current path. It also avoids the complex equipment for laser and ultrasonic energy. This donor material assisted FSW idea is unique. Except the paper [10][11][12] on examining the plunge stage, there is no reported work examining the effect of donor material on the welding stage and the joint properties of FSW. To ll this gap, the proposed methodology is presented in Sect. 3.

Experimental Procedure
The selected base metal in this study was an extruded 6.35 mm thick AA6061-T6 aluminum alloy which is a commonly used light weight metal in aerospace applications. The AA60601 comes in sheet form with dimensions of 317.5 mm (length) × 76.2 mm (width). Cu 110 was selected as the donor material due to its high thermal conductivity. The chemical composition and mechanical properties of the selected materials are listed in Tables 1,   2, and 3 respectively.  mm (width) is created to anchor the stir donor material to the workpiece through set screws. The aperture and the donor material were cut using a Haas CNC mill prior to the welding as shown in Fig. 2 (b-c). The Cu donor material thickness for the present investigation was set at 20%, 40%, and 60% of the depth of the workpiece thickness.
Prior to FSW, the workpiece surfaces and surrounding areas were cleaned using ethanol followed by scouring with an abrasive pad to remove oxides and debris. Finally the workpieces were cleaned again with ethanol. The workpieces were then tightly clamped using a welding xture that was secured to the back anvil/worktable of the welding direction was parallel to the extruded direction of the workpiece ( Fig. 1(a)). All the joints were friction stir welded at a speed of 1 mm/s or 3 mm/s. The tool rotational rate was kept constant at 1400 rpm and a penetration depth of 5.85 into the workpieces. It is noted that relative to the welding and tool rotational directions as indicated in Fig. 1(a), the welds were joined with the counter-clockwise tool rotation. The nal friction stir welded aluminum plates are shown in Fig. 2 The metallographic analyses were accomplished using an Olympus optical microscope equipped with quantitative image analysis software for microstructure and macrostructure studies, respectively. Automated Vickers microhardness measurements were conducted across the welded cross-sections using a load of 100 g, a dwell time of 15 seconds, and an indent interval of 0.3 mm (i.e., at least three times the diagonal length of the indentation to prevent any potential effect of the strain elds caused by adjacent indentations). The tensile tests experiments were carried out by a universal tensile testing machine (INSTRON-5869) with a 50kN load capacity. At least two samples were tested at each welding condition. Tensile samples were prepared according to ASTM E8/E8M-13a [30]. All tensile tests were performed up to failure at room temperature and at a crosshead speed of 1.5 mm/min.
The fractured surface of the frictionally stir welded (FSWed) samples subsequent to the tensile testing was examined using a JEOL 6700 eld effect-scanning electron microscope (FE-SEM) equipped with three-dimensional (3-D) fractographic analysis capacity.
Results And Discussion

Downward force generation
The downward force is a signi cant parameter, which in uences the joint quality and tool life during the plunge and welding stages. A load cell with a National Instrument Data Acquisition (NI-DAQ) device were used to collect the force data in the joining experiment. Figure 3 (a-b) demonstrates the downward force produced at a constant 1400 rpm tool rotational speed and welding speeds of 1 mm/s and 3 mm/s on different donor material thickness settings, respectively. It can be noticed from the plots of Fig. 3(a-b) that the downward force in the aforementioned plunge, dwell, welding and retraction stages shows the following trend. During plunge stage the rotating tool approaches the surface of the workpiece and penetrates into the workpiece generating initial frictional heat. At this point the downward force starts to increase and reaches to the peak position until to the end of plunge stage. Thereafter, in the dwell stage the downward force begins to drop gradually, which indicates that the required operating temperature is achieved to weld the workpiece. In the welding stage, the constant frictional energy generates high temperatures which soften the workpiece, at this point the downward force keeps stable and increased slightly until it reaches to the exit position, then it immediately reduced to zero downward force in the retraction stage. It is understandable that the higher downward forces were taking place during the plunge stage which resulted in producing frictional heat that softened the surrounding material [31][32]. Several authors [33][34] also reported similar ndings on friction stir welding of aluminum alloys. They also observed that the downward force increases during the initial plunge stage and then suddenly drops with further advancement of the tool head. It was observed that with the embedded donor material at the initial plunge stage, the downward force decreases signi cantly as the donor material thickness increases by 20%, 40%, and 60% of the workpiece material. In Fig. 3 (ab), the downward force variation with time is compared for the as-received welded sample and the donor material welded ones.
The downward force obtained during the plunge stage for the welding speed of 1mm/s and tool rotational rate of 1400 rpm was much higher; it reached approximately 8kN for an as-welded (AW) sample, 12 kN for 20%, 8 kN for 40%, and 7.5 kN for 60% Cu donor material assisted sample. As compared to the plunge stage, during the welding stage, the downward force was predominantly reduced by 83%, 66%, and 40.6% when the donor material was 60%, 40%, and 20% of the workpiece samples respectively. Similarly, during the plunge stage at a welding speed of 3mm/s and tool rotational rate of 1400 rpm, the downward force reached approximately 8.75 kN, 7.5 kN, and 6 kN for the 20%, 40%, and 60% Cu donor material respectively.
During the welding stage, the downward force uctuates between 2.5 kN to 3.75 kN for 20%, 40%, and 60% donor material. There was no signi cant difference noticed in the downward force during the welding stage of the aluminum plates with or without the donor material for the 3 mm/s welding speed. This could be due to lack of generating enough heat during the initial plunge stage to soften the surrounding material in addition to a faster cooling rate. Therefore, it is evident that plunging through the Cu donor material generates frictional heat that transfers to the aluminum workpiece that further softens the aluminum workpiece. The use of a donor material concept in preheating the workpiece could further help in selecting proper welding conditions and also in producing optimal welds.

Microstructure
The thermo-mechanical condition change in FSW leads to form different regions in the welding location: namely the SZ or weld nugget around the weld center, the TMAZ and the HAZ on advancing and retreating sides of the weld and the BM. These regions are shown in Fig. 4 (a-e, where the typical microstructures were observed across a weld made at 1 mm/s welding speed and 1400 rpm tool rotational rate. Similar observations have been documented by others [35][36]. It is generally believed that the material in FSW undergoes continuous and discontinuous dynamic recrystallization [6, 28, and 36], grain re nement, and grain coarsening. In the SZ (Fig. 4(b)), a severe plastic deformation and the elevated temperature led to dynamic recrystallization of the AA6061-T6 alloy plates. The grains became predominately equiaxed in both the AS and the RS respectively. The grains in the SZ were smaller and equiaxed as compared to the grains in the BM, suggesting that the current selected FSW assisted donor material conditions resulted in grain re nement. In the TMAZ, the microstructure was highly deformed and grains appeared to be distorted, and elongated. The dynamic recrystallization process did not fully develop in TMAZ during FSW experiment.
The HAZ microstructure which falls between the BM and the TMAZ is mostly experiencing the effect of frictional heating and no interference of mechanical forces. This leads to partial recrystallization of the larger grains in the BM microstructure. Figure 5 shows images of the microstructures of selected regions within the SZ obtained at 20%, 40%, 60% Cu donor material and as-welded condition. At the SZ weld location, there is no signi cant change in grain growth observed both with or without donor material as clearly seen in Fig. 5 (a-d). This could be due to the rapid dissipation of the heat generated during the plunge stage throughout the workpiece surrounding area.

Micro-hardness
Vickers micro-hardness pro les were generated at the top, mid, and bottom layers of the workpiece across the weldments with different welding parameters and different donor material thicknesses as shown in the Fig. 6 (a-b). The indentation hardness is measured starting from the BM advancing side to the BM retreating side. It is observed that the hardness pro les exhibited a W-shape. From the plotted hardness pro les in Fig. 6, it was noticed that the RS of the weld closer to the tool pin had lower hardness compared to the AS, this is due to strains developed in the workpiece by the welding tool. The temperature during FSW on the AS side is higher than on the RS side, which enhances the hardness on the AS side.
The minimum hardness occurred at the HAZ with HV values of about 65-68 and it rose up to HV values of 80-85 at the SZ. The highest hardness values of HV of 100-110 occurred in the BM zone. The average minimum hardness in the SZ was equivalent to 77 % of the base metal. The hardness in the SZ also depends on the precipitation distribution since the AA6061-T6 is a precipitation hardened alloy [37]. The hardness values in the SZ were higher than in the TMAZ and the HAZ, due to dynamic recrystallization and grain re nement as discussed above in the microstructure section. In the HAZ, the material undergoes only softening, the hardness observed were the lowest in each weld. There is no intense plastic deformation and grain re nement in the HAZ, which is evident from the grain coarsening.
The TMAZ undergoes some plastic deformation with addition of heating; this leads to grain re nement compared to the HAZ. Therefore, it was found that higher hardness values occurred in the TMAZ as compared to the HAZ due to strain hardening. With increasing Cu donor material thickness, the heat input and peak temperature increase. Thus, it further accelerates the dissolution, precipitation, and growth of secondary phases. According to [38][39], this grain re nement at SZ and HAZ is mainly attributed to the dissolution and precipitation of secondary phases and change in microstructure of the weld zone. The increase in the hardness in the SZ is due to severe plastic deformation and dynamic recrystallization with the donor material inserted on top of the workpiece during the plunge stage. With the amount of heat generated by friction between the tool head and the donor material, the grain re nement predominantly occurred in the SZ. This led to increased hardness inside the SZ. It is noted that the hardness in the SZ, the TMAZ, and the HAZ with 20% Cu donor material is higher than that of the 60% Cu stir donor material and also of the Al-Al control welded samples with 3 mm/s welding speed. This is due to slower cooling rate in the 20% Cu donor samples, which led to the SZ to retain most of its heat for a longer period which resulted in added hardness of the material. During microstructural analysis, it was observed that the grain re nement has taken place in the order of SZ < TMAZ < HAZ < BM, with average grain size of 4.77 µm in the SZ, 9.6 µm in the TMAZ, 56.73 µm in the HAZ, and 83 µm in the BM. Since the AA6061 is a heat treatable alloy, the precipitations of Mg 2 Si in Al 6061 are responsible to enhance the strength of the alloy [40]. Several investigations have demonstrated that the hardness of the aluminum alloys was mostly affected by precipitate distribution rather than grain size [41]. These precipitates highly dissolve in FSW processes, which eventually reduces the UTS of FSW joints. The changes in hardness distribution at different zones could be due to high dissolution of the precipitate and grain size during FSW [42][43]. The UTS values were 187 and 166 MPa for the 60% and 20% donor assisted samples respectively. It was also observed that the ductility decreased constantly from Al-Al to 20%, 40%, and 60% Cu donor material at 1400 rpm, 3 mm/s welding conditions. This ductility loss is due to secondary cracks generated in the metal matrix during welding. The amount of heat loss is more due to faster cooling rate and lower heat input is less in 3 mm/s welding speed. Fig. 7 (c-d) shows the failure location of joints produced at 1400 rpm with a speed of 1 mm/s and 3 mm/s respectively. All samples were fractured on the RS side near the TMAZ/HAZ in conjunction with the micro-hardness pro les where the lowest hardness occurred in the HAZ location. The HAZ location is the weakest zone of the welded samples. He et al. [41] also reported similar fracture locations subsequent to tensile testing in FSW joint made of thick AA6061-T6 plates. The nature of the fractured surface appeared to be ductile with dimple-like structures. It was noticed that the donor material did not impact the UTS of the joints for the of 1400 rpm and 1 mm/s tested samples. However, the joints of the welded samples at 1400 rpm and a speed of 3 mm/s were found to have some secondary cracks and tear ridges, which led to a combination of brittle and ductile fracture as shown in Figure 9. These preexisting secondary cracks sites are the easiest location where crack initiation starts internally and cause rapid failure during the tensile testing. The tensile fracture starts at the HAZ location in a brittle fashion and propagates at 45° into the TMAZ/SZ in a ductile fashion. This can be seen in the images shown in Fig. 7 (c) and (d).

Fractography
Fractured surfaces of all tensile samples were investigated by eld emission scanning electron microscope (FE-SEM). Figure 8 (a-h) shows the fracture morphologies of the 1400 rpm and 1 mm/s welded samples. These fractographic images consist of crack initiation, spherical and broken dimples, secondary cracks, tear ridges, and particles. It has been observed that crack initiation occurs near the TMAZ/HAZ on the RS. The fractured surface presents large number of small spherical dimples with layered distribution, which often belongs to a ductile fracture. The fractured surface is covered with small deep dimples embedded with secondary particles which indicates good ductility. In higher magni cation, dimples are clearly visible in images as shown in Fig. 8 (b, d, f) and Fig. 9 (a, e, g). These dimples appear in the ductile fractured surface, which are only corresponding to a void. The failure mode was observed to be ductile fracture in AW with 20%, 40%, and 60% donor material. However, there were also some tearing ridges seen on the fractured surface which indicates loss of ductility. Fig. 9 (a-i) shows fracture morphologies of the 1400 rpm with a speed of 3 mm/s welded sample. These fractographic images consist of crack initiation, spherical and broken dimples, secondary cracks, tear ridges, and secondary phase particles. In the AW fractured surface, the secondary phase particles were embedded in the dimples which indicates good ductility. This can be observed in the tensile test plots Fig. 7 (b). The secondary cracks and broken dimples are observed on all fracture surfaces which indicates brittle fracture and lost ductility in comparison with AW condition. This loss of ductility can be clearly seen in Fig. 7 (b). All these features indicate that a combination of brittle and ductile fractures was observed in the 20%, 40%, and 60% Cu donor materials assisted FSW.

Conclusions And Future Work
In this study, Cu donor material assisted FSW was successfully used in the joining of AA6061-T6 alloy plates using different welding speeds and a constant tool rotational speed. The downward force, post-weld microstructure and mechanical properties of the welded plates were studied. The following important conclusions are drawn from this study.
1.The downward force generated in the FSW processes gradually decreased subsequent to embedding the Cu donor material by 20%, 40%, and 60% of the workpieces. The downward force is decreased by a maximum of 83% for a 60% Cu donor material due to excessive heat transfer from the donor material to the workpiece. This proposed work of integrating a Cu donor material at the initial plunge stage is promising to increase the tool life without compromising the mechanical properties for the given welding parameters. The implementation of the donor material increases the lifetime of the FSW tools which is critically useful and economical.
2. The stir zone (SZ) and the thermo-mechanically affected zone (TMAZ), which were predominantly composed of equiaxed grains, experienced full dynamic recrystallization. The grain size in the SZ decreased with an increasing heat input (i.e., increasing tool rotational rate and decreasing welding speed). The average grain size in the SZ is 4.77 µm, the TMAZ is 9.6 µm, the HAZ is 56.73 µm, and the BM is 83 µm. There was no change in grain size in the SZ with embedded Cu donor material for the different thicknesses of the workpiece used in this research (20%, 40%, and 60%).
3. The joints presented a W-shaped hardness pro le the their cross-section. The minimum hardness of HV 65-68 was measured in the HAZ and increased to a HV of 80-85 at the center of SZ. When the welding speed was increased from 1 mm/s to 3 mm/s, the measured hardness of the HAZ gradually decreased whereas the hardness in the SZ increased. The overall measured hardness across the SZ was slightly higher at a lower heat input (i.e., at a constant tool rotational rate and/or a higher welding speed). 4. The maximum tensile strength obtained at 1400 rpm and 1 mm/s was 192 MPa for both with and without Cu donor material, which is 73.2% of the base material. The joints made with Cu donor material had no signi cant effect on the improvement of the tensile strength.
5. Fractography of the tensile testing failure samples for the 1400 rpm, 1 mm/s joints exhibited ductile like fracture with some secondary particles in addition to spherical and broken dimples. The fractured surface for the 1400 rpm, 3 mm/s speed appeared to be a combination of brittle and ductile like fracture. The secondary cracks and broken dimples were also observed for the 20%, and the 60% Cu donor material. In summary, although it is proven that the downward force is drastically decreased with embedded Cu donor material, no signi cant changes occurred in the mechanical properties of the welded joints. In the future, to further understand the temperature pro les and enhance the mechanical properties, the authors will continue to conduct further testing and measurement of the temperature pro les at different welding regions and conduct heat treatment with several solution treatment and aging time limits.
Declarations Advanced Manufacturing Technology.
h. Authors' contributions: Dr. S. Bhukya carried out experimentation and data analysis, and wrote/revised the manuscript. Dr. Z. Wu, as the Principal Investigator, secured funding and resources for the research, designed experiment, wrote and revised the manuscript. Mr. J. Maniscalco carried out FEA simulation of the experiment and proofread the manuscript. Dr. A. Elmustafa conceptualized the experiment, reviewed and revised the manuscript. Steps for making FSW samples.    Optical microstructure images of SZ for different Cu Donor material thicknesses 20% (a), 40% (b), 60% (c), and Aswelded condition (d).

Figure 6
Micro-hardness pro les of a FSW with and without donor material (a) 1400 rpm, 1mm/s, (b) 1400 rpm, 3 mm/s.

Figure 7
Tensile test plots of with and without donor material: 1400 rpm, 1 mm/s (a), and 1400 rpm, 3 mm/s (b), Tensile tested fractured samples of 1400 rpm, 3mm /s (c), and Fractured samples of 1400 rpm, 1 mm/s (d).  FE-SEM Fractography images for 1400 rpm and 3 mm/s welding condition after tensile test, AW samples show secondary phase particles (a-b), 20 % Cu donor specimen shows brittle fracture, secondary cracks(c-e), 40% Cu donor specimen shows brittle and ductile fracture images (f-g), and 60% Cu donor specimen shows secondary cracks with a combination of brittle and ductile fracture with small and elongated dimples (h-i).