1. Introduction
Recently, with the requirement for carbon dioxide emission reduction and energy saving, high strength aluminum alloys (6xxx-series) have been widely employed in the automotive, shipbuilding and aerospace industries because of their remarkable specific strength, excellent damping capacity and low density [
1,
2,
3,
4]. Consequently, effective joining techniques for these aluminum alloys have attracted growing global interest. When traditional fusion welding is used, the joints often develop numerous defects like lack of fusion, porosity and solidification cracks [
5,
6,
7,
8]. Furthermore, weld joint distortion when welding thin aluminum sheet thicknesses can significantly deteriorate the welding joint’s mechanical properties [
9].
As a solid-state welding process, friction stir spot welding (FSSW) has several benefits over fusion spot welding, including high joint performance, minimal joint distortion, absence of cracks, low power consumption, and no emissions [
10,
11]. As a result, FSSW is widely used in the transportation and electronics industries. Therefore, many efforts have been made to optimize its process parameters for welding different metals and alloys in similar or dissimilar lap joints, such as tool design [
12,
13,
14], rotational speed [
15,
16,
17], dwell time [
5], plunge depth [
18], plunge speed [
19] and downward force [
20]. The flow of materials at the FSSW joint interface is determined by the tool design, plunging rate, dwell time, and downward force; however, the generation of frictional heat is mostly dependent on the tool’s rotational speed [
20,
21,
22]. However, despite its many advantages, FSSW is susceptible to several common weld imperfections that can compromise joint integrity and performance. These include hooking, which characterized by the upward bending and thinning of the upper sheet due to excessive tool penetration. This directly reduces the bonded ligament length and joint strength [
2,
3,
4]. Additionally, flash formation, which resulting from material expulsion during tool plunge, can lead to material loss and negatively affect the joint’s aesthetic quality [
5,
15,
17]. Foremost among these, and a critical focus of current research, is the keyhole defect, which is a result of the tool being removed from the workpiece at the final stage of the welding process [
23,
24]. This defect is a cylindrical cavity that remains at the core of the spot lap welds [
25]. It presents a number of difficulties, especially in applications where corrosion resistance, surface quality, aesthetics, and fatigue resistance are crucial. This is a critical concern in the automotive and aerospace industries, where joint performance is vital. Therefore, techniques for refilling or eliminating the keyhole defect have become a research hotspot [
26,
27,
28,
29]. To address this, various techniques have been developed and explored in the literature, each with distinct advantages and inherent limitations [
25,
30,
31,
32]. Among these, pinless FSSW can provide a smoother surface finish compared to traditional FSSW with a pin, as there is no exit hole that requires post-processing [
32]. In contrast, it has limitations regarding the joint strength, thickness, and complex configurations, making it only suited for simple, thin-sheet welding applications [
25]. As a result, this technique is not well-suited for spot joining high-strength aluminum alloys used in structural applications.
Lunetto et al. [
29] examined two techniques: Active Filling Friction Stir Repair (AF-FSR) and Passive Filling Friction Stir Repair (PF-FSR) for repairing pre-drilled holes in AISI 304 stainless steel. This work highlights the challenges of repairing high-melting-point metals and evaluates the impact of process parameters and filler materials on repair quality. The authors demonstrated that PF-FSR offers superior mechanical strength and defect-filling capabilities. Ji et al. [
23] succeeded in filling the keyhole defect of AZ31-B magnesium friction stir welds through a complex mechanical technique named active-passive filling friction stir repairing. In this technique, three non-consumable pinless tools having the same shoulder features (a six-spiral-flute shoulder) and different diameters (6, 10, and 14 mm) were used to refill the keyhole defect in the presence of an extra filler material. The refilling process was achieved at a constant tool rotational speed of 1300 rpm, a holding time of 5 s and various plunge speeds (1, 2, 4 and 6 mm/min). According to their results, a plunge speed of 1 mm/min and a rotational speed of 1300 rpm produced a defect-free joint. In contrast, Silva et al. [
33] were unable to produce defect-free refilled FSSW joints in 2 mm thick AA6082-T6 alloy using the refilling FSSW technique (RFSSW), which utilizes a multi-part tool system (pin, sleeve, and clamping ring). Their experiments, using the suggested parameters, resulted in three different hook defect configurations in the AA6082-T6 lap joints. They concluded that these hook defects were related to the direction and position of the material flow around the sleeve, as well as the development of the drop zone during the plunging and retracting stages. Tier et al. [
34] also concluded the formation of volumetric defects using the same technique, RFSSW prototype machine: Sleeve plunge variant, during the RFSSW of 1.5 mm thick AA5052-O aluminum alloy to produce lap joints. In contrast, Reimann et al. [
3] based on the concept of utilizing a probe, sleeve and clamping ring tool setup system, succeeded to refill a keyhole of 7.5 mm in diameter of the FSSW AA6061-T6 aluminum sheet using a cylindrical plug of the same alloy material. The plug was joined to the surrounding keyhole cavity through a friction spot welding process, which created a sealed keyhole. The results showed that they achieved flawless welds without any reduction in strength compared to the spot joints in the bead-on-plate specimens. In fact, this technique is time-consuming, complex, and involves several procedures. Furthermore, it is crucial to optimize the process parameter. Zhao et al. [
18] also attempted to FSSW a 7B04-T74 aluminum alloy (1.9 mm thick) sheets in lap-joints with avoiding the formation of a keyhole defect via a refilling method employing a movable tool assembly. Their results demonstrated that the optimal joint strength was achieved at a 3 mm tool plunge depth. However, this approach required additional post-weld treatment to eliminate groove defects. While these aforementioned techniques aim to address the keyhole defect, they often present significant collective limitations. These include the necessity for specialized tools or complex equipment, leading to increased process complexity, prolonged cycle times, and higher overall costs. Furthermore, some methods may not fully eliminate defects or are unsuitable for high-strength structural alloys.
Therefore, this work addresses the need for simpler and less time-consuming FSSW procedures for high-strength Al alloys, where most prior research employs high tool rotational speeds (>1000 rpm) [
35,
36,
37]. This study investigates the feasibility of FSSW of 1 mm to 2 mm thick AA6082-T6 Al alloy sheets at low rotational speeds, aiming to optimize weld quality through a focused investigation into the isolated effects of rotational speed. This work additionally aims to refill the keyhole in the produced high-performance spot-joint. The refilling process of this volume defect was achieved using a newly designed AA6082-T6 consumable rod of the same base alloy. This process employs an additive manufacturing technology called continuous multi-layer friction stir deposition (CMFSD). Unlike the aforementioned refilling approaches by Reimann et al. [
3], Zhao et al. [
18], Ji et al. [
23], Ikumapayi et al. [
25], Lunetto et al. [
29], Muhayat et al. [
32], Silva et al. [
33], and Tier et al. [
34], our proposed CMFSD represents an advanced solid-state additive manufacturing (AM) technique for keyhole refilling. Unlike fusion-based AM methods, CMFSD builds material layer-by-layer without melting, using a combination of frictional heat and severe plastic deformation to consolidate metallic feedstock into fully dense structures [
38,
39]. The process involves a rotating consumable tool that plunges into the substrate or previous layer, generating heat through interfacial friction and mechanically stirring the material to form strong metallurgical bonds [
40]. In this study, we modified our existing FSSW machine to perform the CMFSD process, employing optimized parameters with an AA6082-T6 consumable rod to directly refill the keyhole defects in the AA6082-T6 FSSW joints. This solid-state refilling process with similar materials inherently simplifies material management and ensures metallurgical compatibility. While the current laboratory setup involves distinct steps, this approach holds significant potential for integration into automated multi-tool FSW systems, which would streamline the process, reduce overall cycle time, and potentially lower costs compared to methods requiring complex tool changes, dissimilar fillers, or extensive post-treatments. Finally, a thorough evaluation was conducted on both the conventional AA6082 FSSW and the subsequent refilled lap joints. This assessment covered joint appearance, macroscopic and microscopic structural features, hardness measurements, tensile-shear strength values, and fracture behaviors. Electron backscattered diffraction (EBSD) was also utilized to study the grain structure features.
3. Results
3.1. Surface Appearance of Spot Welds
Figure 7 represents the top surface condition of the FSSW AA6082-T6 joints produced at various rotational speeds of 350, 550, 750, and 950 rpm with a fixed dwell time of 5 s. The results show that the carefully chosen FSSW factors in this work were successful in spot-joining the AA6082-T6 aluminum strips, which had different thicknesses of 1 mm (upper sheet) and 2 mm (lower sheet). The distinctive marks of the FSSW in terms of the shoulder projection (outer circle where the tool shoulder comes in contact with the upper AA6082-T6 sheet) and keyhole (inner circular area, which is the pin penetration area) appear clearly and completely round without distortion in the spot-welding joints. This indicates the efficiency of the fixation system and the good selection of the applied heat input despite the thin thickness of the spot-welded samples. It has been reported that an increase in tool rotational speed at a constant dwell time increases the heat input introduced to the FSSW joints, which increases the plasticity of the material beneath the tool shoulder and may cause a circular flash [
5]. This phenomenon was observed in
Figure 7c,d at the high rotational speeds of 750 and 950 rpm, respectively. The dimensions of the keyhole and the extent of this shoulder projection of the FSSW AA6082-T6 are directly influenced by thermal and mechanical energies during the friction stiring process. The shoulder volume for the joints processed at 350, 550, 750, and 950 rpm was 27.434 mm
3, 94.234 mm
3, 157.034 mm
3, and 237.434 mm
3, respectively. This data shows a clear trend of increasing shoulder volume with higher rotational speeds., In addition, the measured keyhole volumes increase nonlinearly with rotation speed: 80.5 mm
3 at 350 rpm, 147.3 mm
3 at 550 rpm, 210.1 mm
3 at 750 rpm, and 290.5 mm
3 at 950 rpm. The nonlinear growth pattern reflects the complex interplay between thermal softening and material flow dynamics. Insufficient heat input at lower rotational speeds (≤350 rpm) can result in incomplete material plasticization, potentially causing irregularly shaped keyholes due to inadequate material flow upon pin retraction. Conversely, high rotational speeds (≥750 rpm) can generate excessive heat, leading to uncontrolled material extrusion and increased burr formation. The value of 550 rpm provides a favorable balance between the welding conditions, resulting in effective material plasticization and controlled flow, thereby contributing to the formation of a high-quality joint.
3.2. Temperature Profile During the FSSW Process
Temperature begins to generate when the pin is inserted into the upper sheet of the AA6082-T lap joint during the plunge stage. The tool shoulder then contacts the top surface, causing the temperature to rise rapidly. The tool shoulder causes frictional deformation and the pin causes plastic deformation. Finally, the tool is pulled from the FSSW joint, and it is cooled to room temperature. The thermal history profiles of the AA6082-T6 FSSW joints processed at different tool rotational speeds (350–950 rpm) and a constant dwell time of 5 s are given in
Figure 8. The measuring temperatures were plotted against time. It is evident from
Figure 8 that while the peak temperatures at each tool rotational speed vary, the thermal cycle profile for all produced spot-welds shows the same trend. The thermal cycle of the FSSW shows three stages. (1) Heating stage: As the tool rotates and penetrates the aluminum sheets, friction between the tool and material generates heat. Higher rotational speeds (750 and 950 rpm) produce more heat due to increased frictional energy, causing faster heating and higher peak temperatures. Lower rotational speeds (350 and 550 rpm) result in slower heating and lower peak temperatures. It can be concluded that higher rotational speeds led to higher peak temperatures, driven by the increased frictional heat generated throughout the stirring operation. It attains the value of 238, 269, 301, 346 °C at the applied tool rotational speeds of 350, 550, 750 and 950 rpm. (2) Holding stage: The tool remains in contact with the workpiece for 5 s. During this stage, the heat conduction within the material balances with heat dissipation, stabilizing the temperature during this time. (3) Cooling stage: After the tool is retracted, the spot-weld joint cools down. Cooling rate depends on the peak temperature in the SZ; higher peak temperatures (at higher speeds of 750 and 950) typically result in steeper cooling gradients. All samples cooled under identical conditions while secured by the clamping system until they reached room temperature. The cooling process is non-linear, with a faster initial cooling phase that is critical for microstructural development. Consequently, the cooling rates were determined by analyzing the temperature curves (
Figure 8) and calculating the average slope from the peak temperature at different rotational speeds (350–950 rpm) to 100 °C. This approach provides a more representative metric for understanding the influence of thermal history on the final microstructure and properties of the produced joints [
51,
52]. Based on these calculations, the cooling rates were 10 °C/s at 350 rpm, 6.1 °C/s at 550 rpm, 6.7 °C/s at 750 rpm, and 7.2 °C/s at 950 rpm. These thermal histories, characterized by varying peak temperatures and cooling rates within the critical temperature range, lead to measurable differences in mechanical properties and microstructural evolution of the FSSW joints.
3.3. Macrostructure of the FSSW Joints
Figure 9 displays macroscopic cross-sections of AA6082-T6 FSSW joints, produced at tool rotation rates of 350, 550, 750, and 950 rpm, with a fixed welding time of 5 s. A notable observation from these macrographs is the creation of flawless spot-welds at the overlapping interface between the 1 mm and 2 mm AA6082-T6 Al alloys, indicating optimal FSSW conditions. The macrostructure of the friction stir spot welded joints of AA6082 aluminum alloys exhibits different zones caused by the localized heat and plastic deformation. These zones are: Base material (BM), stir zone (SZ), thermo-mechanically affected zone (TMAZ), and heat-affected zone (HAZ). Each zone has distinct features determined by the thermal cycle to which it was subjected [
33,
35]. A defining attribute of the fabricated spot welds is the presence of a unique central opening. This feature, known as a keyhole, is a consequence of the pin’s withdrawal and constitutes the most substantial defect in FSSW, representing localized material absence.
3.4. Results of Hardness and Tensile-Shear Tests of the AA6082-T6 FSSW Joints
Hardness measurement is an important factor in determining the mechanical performance and quality of FSSW joints, especially for lightweight high strength materials like AA6082-T6 aluminum alloy sheets. The FSSW process involves complex thermomechanical interactions that significantly influence the microstructure and, consequently, the hardness values across the weld zones (SZ, TMAZ, and HAZ). Analysis of the hardness profiles for the 1 mm and 2 mm thick AA6082-T6 alloy joints, processed at different rotational speeds (350, 550, 750, and 950 rpm) with a fixed welding time of 5 s (
Figure 10), provides essential insights into how tool rotational speed affects on joint integrity
Such investigations are vital for optimizing welding parameters, ensuring structural reliability, and expanding the potential applications of FSSW in demanding engineering fields. The mean hardness values of the AA6082-T6 alloy sheets are 110 ± 4 and 125 ± 5 HV0.3 for the 2 mm and 1 mm sheets, respectively. From
Figure 10, it is evident that the hardness values in the weld zones of the produced joints decreased significantly under all applied FSSW conditions compared to the base materials (BMs) of the 1 mm and 2 mm AA6082-T6 sheet alloys. This hardness loss is ascribed to the annealing effect on the AA6082-T6 BMs, a phenomenon caused by the heat produced during FSSW [
53]. The hardness of the distinct zones within the FSSW lap-joints (SZ, TMAZ, and HAZ) is governed by the introduced heat-input and the initial state of the BM-T6 condition [
22,
54]. Accordingly, prolonged thermal influence is predicted to cause a reduction in hardness within the weld, stemming from either coarsening or dissolution of the hardening precipitates [
55]. The SZ of all AA6082-T6 spot-lap joints, processed at different rotational speeds (350–950 rpm), consistently displayed peak hardness values. This phenomenon is attributed to the formation of dynamically recrystallized equiaxed fine grains and the reprecipitation of phases during the cooling cycle [
5,
17]. The HAZ, however, registered the lowest hardness, a consequence of both its grain morphology and over-aging phenomena [
15,
17]. The TMAZ displayed intermediate hardness values, being lower than the SZ but higher than the HAZ. Based on the recorded hardness measurements in
Figure 10, it can be concluded that the hardness of the joint processed at the tool rotational speed of 550 rpm shows higher values than the remain joints processed at 350, 750 and 950 rpm. The maximum hardness measurements were 94.6 ± 1.4, 88 ± 2, and 82 ± 1.4 HV0.3 for the SZ, TMAZ, and HAZ, respectively, of the FSSW joint produced at 550 rpm (
Figure 10b).
The tensile-shear test is a fundamental mechanical evaluation used to assess the load-bearing capacity and deformation behavior of the FSSW AA6082-T6 aluminum lap-joints. This test provides critical insights into the joint’s structural integrity and mechanical performance under shear stress.
Figure 11 shows the tensile-shear load of the FSSW joints at different welding conditions versus the elongation value in mm. These values represent the cross-head displacement of the testing machine, which, while an indirect measure, is used to establish a valid comparative trend of the deformation behavior across all samples tested under identical conditions. Among all spot-jointed specimens, the FSSW joints welded at 550 rpm and 5 s had the highest joint quality in terms of both tensile-shear load and elongation compared to the other joints processed at 350, 750, and 950 rpm for the same welding time. The maximum tensile-shear load for the AA6082-T6 FSSW joints varied according to the welding conditions of each joint, showing a close relationship with the tool’s rotational speed, while the welding time remained fixed at 5 s. The maximum tensile-shear loads were as follows: 3.57 ± 0.25, 4.73 ± 0.27, 3.92 ± 0.22, and 2.97 ± 0.20 kN for the joints welded at the rotational speeds of 350, 550, 750 and 950 rpm, respectively, as shown in
Figure 12.
The thermal history during the FSSW process has a direct influence on the final mechanical properties and microstructure of the joint. The highest peak temperatures generated at 750 and 950 rpm led to significant thermal softening, which is reflected in lower hardness values and reduced maximum tensile-shear loads of 3.92 ± 0.22 kN and 2.97 ± 0.20 kN, respectively. Conversely, the moderate peak temperature of 269 °C at the optimal 550 rpm condition created a balanced thermomechanical environment. This condition was sufficient to promote dynamic recrystallization, resulting in a fine-grained microstructure within the SZ, and avoiding the negative effects of excessive heat input. This optimal balance directly contributes to the peak hardness of 94.6 ± 1.4 HV0.3 and the highest tensile-shear load of 4.73 ± 0.27 kN observed in these joints.
The experimental results demonstrate a clear relationship between tool rotational speed, joint geometry, and the resulting tensile-shear load. The initial stiffness, represented by the slope of the elastic region, indicates the quality and size of the bonded area. This optimal performance at 550 rpm can be directly correlated with the measured bonded area and hook height for the joint. As the rotational speed increased from 350 rpm to 950 rpm, the bonded area also continuously increased, from 113.04 mm2 to 214.90 mm2. Simultaneously, the hook height increased from 0.15 mm to 0.38 mm. The observed peak in tensile-shear load at 550 rpm, with a bonded area of 132.63 mm2 and a hook height of 0.32 mm, represents an optimal balance between these two competing factors. At lower speeds (350 rpm), the lower tensile-shear load (3.57 ± 0.25 kN) is primarily attributed to an insufficient bonded area. Conversely, at higher speeds (750 and 950 rpm), while the bonded area continues to increase, the tensile-shear load decreases significantly to 3.92 ± 0.22 kN and 2.97 ± 0.20 kN, respectively. The abrupt load drops observed in these joints are a direct result of stress concentration at the keyhole defect, which acts as a primary fracture initiation point. This decline is likely due to the negative influence of the higher hook height, which can act as a stress concentration point, and other potential defects related to excessive material flow. Therefore, the 550-rpm welding condition provides the most favorable combination of a sufficiently large bonded area and a controlled hook height, leading to superior mechanical properties.
3.5. Fracture Location and Morphology
The macro-fracture locations and features of the tensile-shear test specimens of the FSSW joints provide critical insights into the joint performance of the aluminum alloy sheets at varying welding conditions [
5,
17]. For the AA6082-T6 lap-joints of 1 mm and 2 mm sheet materials, the fracture characteristics differ based on the welding conditions, particularly the tool rotational speed, as the welding time is constant. At a low tool rotational speed of 350 rpm, insufficient heat generation leads to weak bonding between the two sheets. Macro images of the fracture surfaces show uneven plastic deformation and a notable lack of complete material mixing in the SZ (
Figure 13). Fractures tend to initiate at the sheet interface, propagating along the weakly bonded regions, which results in a relatively low load-carrying capacity compared to the moderate heat inputs at the tool rotational speeds of 550 and 750 rpm (
Figure 12). At 550 rpm, the weld joint exhibits the most uniform and consistent bonding, leading to high joint performance. A portion of the spot-joint remains connected after the tensile-shear test (
Figure 13), indicating a high joint load-carrying capacity. Fractures typically initiate at the edge of the SZ, suggesting an optimal balance between heat generation and material deformation. The improvement in tensile-shear load-carrying capacity is attributed to a larger fully bonded area and an optimal hook height [
17,
56]. At the highest rotational speed (950 rpm), the introduced heat input leads to potential softening of the weld zone, which causes decreasing in the maximum tensile shear load (
Figure 12). Macro features include excessive thinning near the upper sheet and flash formation. Fractures are dominated by necking in the upper sheet due to localized softening.
Figure 14 provides insight into the fracture behavior of the AA6082-T6 alloy BMs. Two primary fracture types, ductile and brittle, are observed for the 1 mm (
Figure 14a,b) and 2 mm (
Figure 14c,d) sheet thicknesses. The ductile fracture is confirmed by the dimpled texture of the aluminum matrix, with dimples varying in size on the 1 mm sheet and appearing shallow and elongated on the 2 mm sheet. In contrast, brittle fracture is characterized by the presence of precipitate phases of Mg
2 Si and (Fe, Mn)
3 SiAl
l2 which were identified using Energy-Dispersive X-ray Spectroscopy (EDS) point analysis (
Figure 15, Spots 1 and 2, respectively). The large second phases visible on the fracture surfaces are these brittle intermetallic compounds, specifically (Fe, Mn)
3 SiAl
l2 particles, which act as stress concentration sites. Under tensile loading, fracture initiates either by the cracking of these particles themselves or by decohesion at the particle-matrix interface. Microcracks and microvoids are also detected upon examining the two alloy fracture surfaces. These are due to microvoid coalescence and precipitates pullout, respectively. These findings are in agreement with those identified in other works [
33,
54].
Figure 16 shows SEM images of the fracture surfaces (after tensile-shear testing) of the lower sheets of the FSSW joints. These joints were welded at various rotational speeds: 350 rpm (
Figure 16a,b), 550 rpm (
Figure 16c,d), 750 rpm (
Figure 16e,f) and 950 rpm (
Figure 16g,h), with a constant welding time of 5 s. Failure in all FSSW lap joints occurred due to a tensile-shear mechanism. During the application of tensile-shear loads, microcracks commonly began at the hook’s tip, specifically in areas with incomplete bonding. These cracks then primarily advanced along the horizontal plane at the joint’s interface, shearing the SZ and causing failure. The time to failure and the load-carrying capacity of the spot-joined materials depend mainly on the joining heat input in terms of tool rotational speed when the welding time is constant.
The analyzed fracture surfaces consistently displayed predominant ductile characteristics, evidenced by the presence of small, deep dimples. A partially brittle fracture mode, marked by fragmented precipitates, was also observed. These findings suggest significant grain refinement within the SZ during FSSW when compared to the elongated dimples and large precipitates in the two AA6082-T6 alloy BMs [
57]. For the suggested welding parameters in the present work and based on the results of hardness measurements (
Figure 10), the tensile-shear test findings (
Figure 11 and
Figure 12), and the macro and micro examinations of the fracture surfaces in
Figure 13 and
Figure 16, respectively, it can be concluded that the best condition to join a 1 mm sheet of AA6082-T6 on a 2 mm sheet of AA6082-T6 via the conventional FSSW technique is 550 rpm and 5 s to attain the highest value of the joint performance.
Based on the above findings, the next sections of the current study will concentrate on addressing the keyhole defect in the FSSW AA6082-T6 joint produced under the optimal welding conditions for 1 mm and 2 mm sheet thicknesses (i.e., 550 rpm and 5 s). This will be achieved by implementing a friction stir deposition (FSD) technique to create continuous multi-layers of the same alloy (AA6082-T6), thereby eliminating the defect and enhancing the joint efficiency. A comparative analysis will be conducted between joints with the keyhole defect and those refilled using the FSD technique, focusing on macrostructure, microstructure, hardness and tensile-shear test results.
3.6. Macro, Microstructures and EBSD of the FSSW and the Refilled Joints
Figure 17a,b illustrations the macrostructures of the AA6082-T6 FSSW joint (
Figure 17a) and its refilled joint (
Figure 17b). Both joints were FSSW at the same condition of 500 rpm and 5 s, while the keyhole repair was achieved via refilling process utilizing a FSD technique to build continuous malti-layers of AA6082-T6. Based on some trail experiments and pervious work [
38], the deposition parameters were established as 450 rpm and 1 mm/min. It should be noted that this approach enabled effective keyhole repair using the identical AA6082-T6 alloy and the deposited continuous malti-layers exhibited excellent adhesion to the keyhole’s inner surface. This leads the absence of porosity or microcracks in the final produced joint. The controlled refilling process deposits AA6082-T6 material that exceeds the keyhole cavity height by 0.3–0.5 mm, ensuring complete defect elimination. Quantitative analysis of the refilled zone confirms effective keyhole remediation, with cross-sectional examination revealing a sound weld joint containing minimal residual burr. The automated deposition control system maintains consistent refill quality while preventing excessive material accumulation.
Figure 17c–f represents the microstructure features of SZ, TMAZ, HAZ and base materials (1 mm AA6082-T6 and 2 mm AA6082-T6), While
Figure 17g,j depicts the microstructures observed in different areas of the deposited AA6082-T6 material. The main structural features of the FSSW zones are dynamic recrystallization fine, equiaxed grains in the SZ (
Figure 17c) due to the severe plastic deformation and thermal cycle effects. The grain size in the SZ is smaller than that found in the 1 mm and 2 mm BM sheets, contributing to localized strengthening. The OM-images of the TMAZ (
Figure 17d) shows elongated and distorted grains (
Figure 17c,d) due to mechanical stirring and moderate thermal input, whereas relatively coarser grains compared to the SZ and base material are detected in the HAZ due to thermal cycle exposure during the FSSW process. Both the BM alloys microstructure images (
Figure 17e,f) show elongated grains in the direction of the rolling process; besides, the presence of Mg
2Si precipitates enhance the mechanical properties of the BMs. Totally, these gained microstructure features for the cross-sections of the AA6082-T6 FSSW zones are in good agreement with that mentioned by different authors in different works [
33,
35,
43].
The EBSD IPF coloring maps provide valuable insights into the microstructural evolution in the SZ and FSD layers of AA6082-T6 joints as given in
Figure 18. Both the SZ of FSSW (
Figure 18a) and the friction stir deposited layers (
Figure 18c) in the refilled FSSW joints exhibit fine, equiaxed grains resulting from dynamic recrystallization due to severe plastic deformation and elevated temperatures as a result of stirring action during FSSW [
5,
17,
56] and FSD [
38] processes. The average grain size of the 1 mm AA6082-T6 and 2 mm AA6082-T6 BMs were 8.6 ± 0.7 (
Figure 17e) and 9.76 ± 1.6 (
Figure 17f), respectively. These values decreased to be 5.05 ± 0.09 in the SZ of the AA6982-T joint spot welded at 550 rpm and 5 s (
Figure 18b), demonstrating substantial grain refinement as a result of the spot-welding process. Also, filling the keyhole defect with continuous layers of the same alloy using a technique that gives a smaller grain size of 7.80 ± 0.1 (
Figure 18d) than the base material gives high mechanical properties to the spot welded joint, according to the Hall-Pitch relationship, the smaller the particle size, the higher the strength [
58]. So, it can be said that the refilled FSSW joints, with their improved grain structure demonstrate superior mechanical properties compared to the conventional FSSW joints, highlighting the effectiveness of the FSD process in addressing keyhole defects.
The FSSW and FSD technologies are based on the principles of friction stir welding, and are considered forms of friction stir processing. In these processes, the initial material is subjected to a severe plastic deformation accompanied by the frictional heat. In most cases these techniques promote dynamic recrystallization and formation of equiaxed fine grains at the expense of the original elongated grain structure associated with rolling in the case of sheets or extrusion in the case of rods [
39,
59]. Two areas were selected, one of which was in the SZ of the FSSW joint, which gave the best mechanical properties, and one in the keyhole refilled zone (AA6082-T6 friction stir deposited layers) to be investigated using EBSD. The EBSD obtained results are given in terms of Inverse pole figure (IPF) coloring maps (
Figure 18a,c) and grain size histograms (
Figure 18b,d). It is observed that for both regions the average grain size is somewhat close, between 5.05 µm and 8.80 µm for the SZ and FSD zone, respectively. This slight difference in average grain size may due to the different of initial forming conditions the BM in terms of whether it is a AA6082-T6 sheet or a rod, and/or also due to the difference in the amount of heat input generated by the treatment with friction stir process. However, the main feature is the participation in the occurrence of dynamic recrystallization, which results in a relative improvement in the quality of the joint and the homogeneity of the spot-welding area as a whole. This was clearly shown in the following mechanical measurements.
3.7. Hardness and Tensile-Shear Results
In the FSSW AA6082-T6 lap joints processed at 550 rpm and 5 s dwell time, the hardness map distribution across the cross-section typically exhibits different areas with different colors indicating different hardness values measured as shown in
Figure 19. These colored areas represent the SZ, TMAZ, HAZ and BM. The BM maintains the highest hardness as it retains its original T6 temper microstructure with uniformly distributed precipitates. The hardness in the SZ is generally lower than the base material (BM) due to thermal softening and dissolution of strengthening precipitates (e.g., Mg
2Si) caused by the high heat input [
35,
54]. The TMAZ shows a gradual decrease in hardness as it transitions from the SZ to the HAZ. This is due to partial grain deformation and precipitate coarsening without complete recrystallization. The HAZ exhibits the lowest hardness in the joint depend on the thermal exposure duration.
When the keyhole is refilled using a consumable AA6082-T6 rod at 450 rpm and a feed rate of 1 mm/min, the hardness contour map reveals some features, as give in
Figure 19. The main features of the refilled joint cross-section are different from those of a conventional FSSW cross-section in terms of the disappearance of the keyhole defect, resulting in a sound weld joint at the cross-sectional level. The deposited layers also showed a homogeneous distribution of hardness contour maps, and there is an improvement in the hardness of the adjacent zones (SZ, TMAZ, and HAZ). The AA6082-T6 deposited zone shows an increase in hardness compared to the original SZ with a slight hardness gradient at the interface due to the thermal and mechanical effects of the deposition process. The metallurgical bonding between the deposited material and the original SZ enhances overall joint integrity. For the TMAZ and HAZ, the hardness profile in these regions remains similar to the original FSSW joint, with some minor recovery in the HAZ due to re-heating during the deposition process. In fact, the hardness contour maps highlight the benefits of refilling the AA6082-T6 FSSW joint, particularly in enhancing the hardness of the SZ and eliminating weak zones associated with the keyhole defect. This improvement contributes to higher joint efficiency, making the refilled FSSW approach a viable solution for improving the mechanical performance of the FSSW lap joints in aluminum alloys. This is clearly shown in the tensile-shear curves of both FSSW joints in the presence of the keyhole and when this hole is filled with CMFSD of AA6082-T6 alloy, as illustrated in
Figure 20. The tensile-shear curve for the AA6082-T6 FSSW lap joint welded at 550 rpm and 5 s typically shows an initial elastic region followed by plastic deformation, reaching a peak load (4.73 ± 0.27 kN) that represents the joint’s maximum tensile-shear strength. Beyond the peak load, a sudden drop in load is noticed, indicating failure. The overall elongation is relatively limited due to the keyhole defect, which represents a loss in the connection of the welding area and may act as a stress concentrator, causing a reduction in the joint’s capacity to sustain plastic deformation. While, the tensile-shear curve for the refilled AA6082-T6 lap-joint FSSW at the same condition and refilled via the FSD using AA6082-T6 rod material at 450 rpm and feeding rate of 1 mm/min exhibits a more extended plastic region compared to the original FSSW joints. The maximum tensile-shear load (6.93 ± 0.19 kN) of the refilled FSSW is higher than that of the FSSW joint due to the elimination of the keyhole defect and improved bonding from the friction stir deposition (FSD) process. The elongation to failure is significantly increased, indicating enhanced ductility and energy absorption capacity of the refilled joint. The noticeably reduced stiffness of the refilled joint (
Figure 20) likely stems from residual stresses or microstructural gradients at the deposition interface, which significantly affect initial elastic resistance. However, this effect is counterbalanced by the the refined grain structure and keyhole elimination, which collectively enable both higher peak strength and substantially prolonged plastic deformation.
It is noted that the refilled-FSSW joint fabricated at 550 and for 5 s and keyhole refilled via FSD of AA6082-T6 consumable rod at 450 rpm and feeding rate of 1 mm/min gives superior joint performance over that only FSSW at the same condition. After the sample was subjected to the tensile-shear test, the FSSW and filling zones resisted the applied load and a fracture occurred in the upper sheet of 1 mm (
Figure 21a), which indicates that the welding area after repairing the keyhole area became stronger than the upper 1 mm sheet thickness in load carrying capacity. The refilled joint’s mechanical performance is influenced by the deposited material’s thickness, which exceeds the keyhole depth by 0.3–0.5 mm to ensure complete defect elimination. This overfilling strengthens the joint, as evidenced by fracture occurring in the 1 mm base sheet (not the refilled zone), confirming the deposited material’s superior load-bearing capacity. While quantitative thickness-strength correlations could further validate these findings, the current results demonstrate that the overfilled design mitigates stress concentration and promotes ductile failure in the parent material.
It is also noted that the characteristics of the fracture surface of the AA6082-T6 deposited layers of the refilled FSSW joint, as given in
Figure 21b–d, do not differ in general from the characteristics of the fracture surface of the SZ of the FSSW joint spot welded under the same conditions without filling the keyhole (
Figure 16c,d). This similarity may be a result of the keyhole filling zone, which is the AA6082-T6 deposited layers, being similar in chemical composition to the primary spot-welded zone (SZ of the AA6082-T6 FSSW joint). Both of them resulted from a friction stir processing process, which is exposure to a severe plastic deformation in the presence of heat, which led to similar changes in the microstructure of the BMs in terms of formation of dynamically recrystallized fine grain structures and fragmentation of the stable large size precipitates. Thus, the fracture surface shows similar features with a slight difference due to the difference in heat input in both cases. The difference in the morphology of the dimples may be due to the difference in the amount of heat input in the two cases, but they are completely identical in the existing phases due to the agreement in the chemical composition. The fracture surface of the investigated area shows a ductile fracture mode with some features of a brittle fracture mode for the hard precipitate phases.