Friction stir spot welding of AA5052 with additional carbon fiber-reinforced polymer composite interlayer

In this study, similar aluminum alloys AA5052 with additional carbon fiber-reinforced polymer composite (CFRP) interlayer were selected to investigate the effect of welding parameters (rotational speed and dwell time) on the mechanical properties, joint efficiency, and microstructure of friction stir spot weld joint. The maximum tensile shear load was 1779.6 N with joint efficiency of 14.6% obtained at rotational speed of 2,000 rpm and 2 s dwell time, which is 39.5% higher than the value at low rotational speed 850 rpm and 2 s dwell time. Meanwhile, themaximummicrohardness 58 HVwas attained in the keyhole region at rotational speed of 2,000 rpm and dwell time of 5 s, which is 22.4% higher compared to low rotational speed. The SEM-EDS results reveal the presence of intermetallic compounds (Al–Mg–C), which enhance the intermetallic bonding between elements.


Introduction
There is an increasing demand for lightweight structures, especially in transportation sector [1]. Recently, applications of lightweight materials such as aluminum, magnesium, and metal foams, as typical porous materials [2], have been increased in automotive and aerospace industries [3]. Also, it is a critical challenge to reduce the weight in these industries to improve the performance of vehicle and airplane [4]. Light weight structure designs can be achieved by using composite materials, which are high cost-intensive and yet not suitable for mass production due to the necessity of manual processing [5]. In recent years, composite materials have triggered worldwide investigations to manufacture improved structures with superior mechanical characteristics [6] same as laminated composite plates [7]. Carbon fiber-reinforced composites (CFRPs) have excellent thermal and mechanical properties and are commonly applied in the fabrication of polymer-matrix composites [8]. Composite and nanocomposite materials are utilized as a lightweight material [9]. CFRP industry is developing by constant growth of demand from defense and aerospace to automotive sectors [10]. In the past decades, nanocomposite and composite materials have attracted great attention because of their exclusive improving and properties' effects on specific mechanical performance which are not easy to achieve with other materials [11]. New microand nano-technologies used in fabrication of advanced materials have been resulted in lightweight structures with a large range of applications [12]. Despite the fact that the extraordinary mechanical characteristics of CFRPs make them an attractive option for designers, they have to outperform existing lightweight structural automotive materials [13], whereas carbon nanotubes are suitable for electrical and thermal application [14]. Moreover, reinforcing core layer with CNTs leads to remarkable drop and increase in thermal and mechanical buckling resistances, respectively [15]. Rezaei et al. [16] proposed the effect of fiber length on thermomechanical properties of carbon fiber. The results show that short carbon fiber-reinforced polypropylene thermal stability increased by increasing the length of carbon fiber. Therefore, extensive research has been performed on the application of conventional pitch-based CFRPs in improving and optimizing the characteristics of a wide variety of structures [17]. The machining of CFRP composites types is carried out mainly on plain-woven carbon fiber and epoxy resin matrix, including directional, unidirectional, and multidirectional CFRP laminates [18]. Aluminum alloys are one of the most promising lightweight materials due to their corrosion resistance, high specific stiffness, strength, high recyclability, and impact resistance [19]. Friction stir spot welding FSSW is a solid state joining technique and has been applied to aluminum alloys such as 2,000 and 5,000 series [20]. These alloys are apt to solidification and liquation cracking, which can be avoided by friction stir spot welding [21]. FSSW is an advanced form of friction welding which is mostly used to weld the similar and dissimilar alloys [22]. FSSW provides reduction in energy consumption up to 90 and 40% in capital cost compared to resistance spot welding (RSW) [23]. Also, the formation of stirring zone in FSSW requires a small amount of heat generation which reduces welding time and energy consumption [24]. Procesov [25] showed the comparison between FSSW and RSW of AA5005 alloy. The results showed that the mechanical performance of FSSW joint was higher than RSW and concluded that FSSW could be an alternative to RSW. The most important parameter for heat input in friction stir spot welding is rotational speed [26] and peak temperature increased with increased rotational speed [27]. The temperature should be less than melting point of metals during the processes [28]. Rana et al. [29] presented the effect of different rotational speed on FSSW; the optimal rotational speed was 1,800 rpm for high mechanical performance. High rotation speed causes high temperature, and as a result, more intensive stirring and mixing of the material and that energy was positively connected to the bonding area with long weld strength [30]. Due to fixed pin length and shoulder, only a single lap joint could be welded with constant thickness during friction stir spot welding [31]. Dwell time is another parameter affecting FSSW and provides the heat required for the formation of stir zone and bonding region between upper and lower sheets [32]. Long dwell times should be applied for metals with high melting points to generate required heat for plasticity flow material [33]. Also, less than one second would be small softening in heat-affected zone for heat-treated aluminum alloys [34]. Rafiei et al. [35] investigated the mechanical properties of dissimilar friction welding of aluminum (AA5052)-magnesium and aluminium (AA2024)copper. The processes parameters were rotational speed 1,250 rpm and feed rate 160 mm/min. The hardness was low in heat-affected zone and thermal-mechanicalaffected zone (TMAZ) due to welding defects such as hook defect. Also, residual stress is considered as a defect on welding quality [36]. Khosa et al. [37] evaluated the effect of thermomechanical during FSSW of AA6082-T6. The result showed that the effect of temperature on the microstructure and mechanical properties was severe due to the deformation and material flow of welded sample. Kubit et al. [38] studied additional sealant interlayer in AA7075-T6 alloy. The obtained results show that the joint properties of weld joint could be improved by optimal thickness of interlayer. Sadoun et al. [39] evaluated the effect of interlayer in dissimilar aluminum AA2024 and AA7075 alloys. The obtained result showed 18% improvement in joint strength due to grain refinement compared to without interlayer joint. Andre et al. [40] improved the adhesion mechanisms and mechanical performance of aluminum AA2024-T3 and CFRP with 100 μm PPS film additional interlayer fiction spot welding. The effect of micro-mechanical interlocking at the interface of interlayer achieved with sandblasted specimens was clear. The optimal tensile shear force was 3,068 N, due to effect on contact surface area. Also, sandblasting was found to be the most effective treatment which maximized joint mechanical performance. Andre et al. [41] investigated the effects of rotating speed and joining pressure on the mechanical strength and microstructure of friction spot welding of carbon fiber and aluminum AA2024-T3 with 100 μm PPS film interlayer. They reported that lap shear force was increased from 2,700 to 3,070 N, which was higher than those for joints without interlayer because of their improved micro-mechanical interlocking, better load distribution, and larger bonding area. Temperature was measured on surface which ranged from 325 to 417°C. In the microstructure of two bonding interfaces, the PPS into the crevices of aluminum and CFRP entrapment by aluminum deformed. However, in joints without interlayer, a transition zone with high air bubbles was created between adhesion zone and plastically deformed zone. Abed et al. [42] proposed the effect of friction stir spot welding of AA6061-T6 with copper interlayer and found that the increase of dwell time and plunge depth was beneficial to the formation of joint and directly increased tensile shear load. Joint formation was improved with cupper interlayer addition and this improvement could be related to the increase of bonding area and intermetallic compound from the reaction of metal and interlayer. High load fracture occurred in cupper interlayer rather than those without interlayer. However, interlayer considerably increases the protection of weld against corrosive environments [43]. The reaction between aluminum and carbon fiber during welding processes caused to form aluminum carbide (Al 4 C 3 ) at temperatures above 500°C, which negatively affected the performance of interfacial layer in composites [44].
In the current study, the effect of welding parameters and CFRP interlayer on the mechanical and microstructural properties of welding joints has been investigated in FSSW joint process. The objective of this study is to evaluate the effect of CFRP interlayer on the performance of mechanical properties (tensile shear load and microhardness) and microstructural properties (SEM-EDS) and joint efficiency of the weld joints.

Experimental work
Aluminum alloy AA5052 plates were chosen in this study. The dimensions of the plate were 100 mm × 25 mm × 4 mm, based on American Welding Society standard (AWS C1.1M/C1.1:2012) [42] as showed in Figure 1(a). CFRP with thickness 0.5 mm was applied as interlayer between similar aluminum sheets. For each combination of process parameters, three replicates were made. The surface of material was cleaned using alcohol to remove contaminants that could generate oxides. The tool was rotated at high speed, then forced into workpiece until the shoulder contacted top metal surface.
The tool was made from steel alloy AISI 4340 (UNS G43400), as showed in Figure 1b. The length of the pin was 6.2 mm and its diameter was 5 mm with cylindrical design. Friction stir spot welding consists of three stages: plunging, stirring, and retracting, as showed in Figure 2.
The experimental full factorial design with two level (rotation speed and dwell time) was chosen for joining processes, as illustrated in Table 1.
In this study, temperature evolution on the top surfaces of aluminum was monitored during friction stir spot welding (Thermometer PCE-T390). The maximum process temperature was considered as the highest temperature measured on the top surfaces of aluminum sheets. The mechanical performance of joints was evaluated through tensile shear load by using universal tensile test machine (INSTRON-5582). Three identical samples were tested for each joining condition. Load rate was adjusted at 3 mm/min for all specimens and were mounted on the jaws of the machine. Joint efficiency (η) of AA5052 with CFRP was obtained by applying the following equation [45].    cross section of the joint by optical microscope. Also, in microstructure evaluation, energy dispersive spectroscopy (SEM-EDS) was applied.
3 Results and discussions 3.1 Temperature evaluation Figure 3 shows peak temperature of aluminum and CFRP interlayer at different rotational speed and dwell time.
The peak temperature was about 298.5°C at rotational speed 2,000 rpm and dwell time 5 s, while the minimum temperature was 216.9°C at rotational speed 850 rpm and dwell time 2 s. Generally, Al 4 C 3 compound could not be formed due to low temperature during the FSSW processes. Also, it was clear that the increase of dwell time from 2 to 5 s at all rotating speeds directly affected temperature, dramatically increasing it by up to 10%. Heat generation at the surface of aluminum was not high enough to melt the interlayer at the interface in the whole overlapped area. This was due to the lower thermal conductivity of CFRP (0.19 W/m K) than aluminum [46].

Tensile shear load
The results of tensile shear load of AA5052 FSSW joints are showed in Figure 4. Maximum tensile of 1779.6 N was produced at rotational speed 2,000 rpm and dwell time 2 s with 14.6% joint efficiency, while minimum tensile occurred at 850 rpm and dwell time 2 s was 1,373 N. In addition, the results showed that increase of rotational speed from 850 to 2,000 rpm at 2 s dwell time with 39.5% improvement in tensile shear load was obtained. It was necessary to decrease the distribution of intermetallic compounds to have high tensile and avoid poor performance of welding [45]. The highest tensile shear load could be related to the large bonding area and good load distribution in FSSW process [47]. On the other hand, increase of dwell time from 2 to 5 s and a 9% decline in tensile shear load at low to high rotational speeds caused to increase heat input that led to overheat the material in weld region, which induced the grain growth and eventually decreased tensile shear load [41,48].
High heat input was developed due to high dwell time, which led to initiate the cracks in the weld interface and reduce the tensile shear load [49]. The maximum elongation of 8% was founded at rotational speed of 850 rpm and 2 s of dwell time. Large elongation could be due to large joint area and increased amount of CFRP attached to aluminum surface [50].

Microhardness
Rotational speed and dwell time are significant parameters in FSSW processes, such that increase of dwell time from 2 to 5 s at all rotational speeds directly increased microhardness. Microhardness profile is illustrated in Figure 5. Maximum microhardness value was 58 HV which was obtained at 2,000 rpm rotating speed and 5 s dwell time in keyhole, whereas minimum value was 24.5 HV obtained at 1,300 rpm rotational speed and 2 s dwell time in stir zone (SZ). The graph showed the highest microhardness in keyhole area rather than SZ and TMAZ. In key hole area, 29% improvement was achieved by increasing rotational speed and dwell time. This was due to mixing and incorporating CFRP and aluminum during FSSW processes. Hardness test results provided further support to the fact that large amounts of CFRP were uniformly dispersed in the matrix over large  region. However, heat generation during FSSW minimized microhardness in TMAZ region. Large amounts of geometrically necessary dislocations (GNDs) [51,52] induced during FSSW are thought to be an important reason of strengthening. There was a large difference in the coefficient of thermal expansion (CTE) between AA5052 (23.8X10-6 K-1) and CFRP (5.5X10-6 K-1) along transverse and longitudinal directions, respectively. The mismatch of CTE led to the generation of GNDs during FSSW [44]. Hence, hardness in SZ and TMAZ could be attributed to the comprehensive effect of variations in grain size and strengthening of precipitates [53].

Microstructure
The microstructure of AA5052 with CFRP interlayer was presented in this study. Weld microstructure depended on the type of the process applied, welding parameters, and workpiece material properties such as thermal conductivity and external cooling conditions [49]. Nub was slightly inserted into thermoplastic composite part, which consequently increased mechanical interlocking between joining partners and contributed to joint mechanical strength under shear loading [54]. The samples of rotational speed 850 rpm and dwell times 2 and 5 s had lower mechanical performance compared with those at 2,000 rpm. The main reason for this was low  heat input during FSSW process, which was not enough to melt interlayer to improve microhardness, as shown in Figure 6(a and b). In the cross section joint of 1,300 rpm and 5 s, crack was observed under tool pin after welding. Cracks on right and left sides rapidly extended along the top surface to pull out the entire joint from the upper plate. In the hook defect, the crack initiation was identified and extended along the hook toward the top surface to pull out stir zone; meanwhile, cracks at hook defect on the left side extended along the hook toward the bottom of stir zone. The joint broke along the interface and only a small part was pulled out. Microcracks in stirring zone were clear and their number was increased by increasing rotational speed and dwell time, as showed in Figure 6(c). In cross section joint at rotation speed 2,000 rpm and dwell time 2 s, interlayer was melted and squeezed out of the center to the edges of the joint. This was due to high temperatures (about 298.5°C) during welding process which led to the melting of interlayer at the center of the joint. This occurred due to lower molten viscosity associated with higher frictional energy generation in this region [46]. The molten polymer could flow into crevices on aluminum surface during joining process. This caused micro-mechanical interlocking between parts, leading to increased tensile shear load, as showed in Figure 6(d) [55].
According to the microstructural characteristics of grain size and precipitates, it was found that weld structure was symmetrical with respect to tool axis. FSSW produced SZ or nugget and TMAZ, which could be identified in sequence from keyhole periphery towards base  material. During tool penetration in friction stir spot welding, rapid strain rates and heating were imposed by rotating tool. In conventional friction stir spot welding method, a stir zone structure containing fine equaled grains fully develops near the end of penetration process when the tool is almost completely penetrated into sheets. Evidence of such cracking was also noted in AA7075 refill friction stir spot welds, as reported by Shen et al. [20]. The formation of a shiny surface on friction stir spot welds combination of material and welding parameters may be prone cracking. Plastic deformation and high temperature induced local melting and it was argued that liquation and solidification occurred repeatedly due to oscillating temperature resulting in nonequilibrium solidus temperature. This paper (SEM-EDS) analyzed the welds produced at rotational speed 2,000 rpm and dwell time 2 s. Similar aluminum alloys made strong bonds with CFRP.
It is necessary to uniformly distribute CFRP in the matrix for fabricating aluminum alloys with CFRP and the results showed that the concentration of aluminum and carbon elements confirmed chemical bonding between aluminum and molten CFRP interlayer. Also, SEM-EDS analysis was performed to the compositions, to take place and form intermetallic compound. Dwell time was the main effect parameter to provide time for the diffusion. Figure 7 presents an SEM-EDS mapping microanalysis to identify the chemical intermetallic composition of the phases present in the welding joint and disruptions of carbon particles into the matrix, which might significantly affect structures and their properties.
Moreover, aluminum carbide (Al 4 C 3 ) was not observed because of low heat input during welding processes (Max. 298.5°C). Contact and bonding between CFRP and AA5052 was high due to thermal-mechanical effect.
The presence of Al, Mg, and C elements was proved, as illustrated in Figure 8, which indicated the presence of strong bonding, and there may be mechanical interlocking leading to high adhesive forces at joint areas which was due to the presence of carbon molecular bonding. [56]. Also, Figure 8 shows ternary intermetallic compound (Al-Si-C) and these elements mixed together due to heat input during friction processes between aluminum and CFRP [57,58]. These compounds might increase microhardness by about 58% in keyhole rather than in SZ and TMAZ [59]. In addition, more carbon fiber increased microhardness. On the other hand, more fiber reduced the distance between aluminum and fibers and increased stress, which decreased the strength of composites [60].

Conclusions
The current study demonstrated that similar aluminum alloys 5052 with additional CFRP interlayer were successfully joined by applying friction stir spot welding. The effect of tool rotational speed and dwell time on mechanical performance and microstructure was investigated. The following conclusions were drawn: • The maximum tensile shear load was obtained at rotational speed 2,000 rpm and 2 s dwell time with joint efficiency of 14.61%. The increase of tensile load due to heat input melted CFRP interlayer and squeezed out of the center, whereas low rotational speed had low value of tensile shear load. Consequently, this decrease in tensile shear load may be due to the formation of microcracks within the stir zone. • The maximum hardness was observed at rotational speed 2,000 rpm and 5 s dwell time, which gradually reduced by decreasing rotational speed and dwell time. The increase in key hole area was due to plastic deformation rather than SZ and TMAZ. Carbon fiber interlayer and welding parameters showed significant effect on microhardness result. This was due to highly different coefficients of thermal expansion, mixing, and incorporating between AA5052 sheets and CFRP interlayer. • In the microstructural, the main discrete regions observed in FSSW joints are SZ and TMAZ. Interlayer was melted and squeezed out of the center of edge joint due to high heat input during the processes. Scanning electron microscope showed welding defects such as cracks which were clearly observed on the left and right sides of keyhole and were increased by increasing welding parameters. Also, SEM-EDS analysis was applied to observe the intermetallic compound and their element distribution. The elements such as Al, Mg, and C increase the mechanical performance.
Funding information: The authors state no funding involved.
Author contributions: All authors have accepted responsibility for the entire content of this manuscript and approved its submission.

Conflict of interest:
The authors state no conflict of interest.