Enhancing interfacial integrity in Al/CFRP friction stir welds: Microstructure control and shear strength improvement

Abstract

Lightweight design of structures has been widely used in the transportation, aerospace and electronics industries because of its advantages in reducing energy consumption, reducing emission and improving environmental performance. This study investigates the effects of welding parameters on microstructural evolution and joint performance in friction stir welding lap joints using 6061-T6 aluminum alloy and carbon fiber-reinforced polymer (CFRP). The microstructure of weld surface and the cross section of the joint is systematically examined to analyze surface defects, material softening characteristics, plastic deformation distribution and micropore filling. Experimental results demonstrate that the increased heat input enhances the plastic flow of the material and deepens the shoulder influence zone to 0.54 mm. The optimized welding parameters can inhibit the formation of micropores caused by insufficient material flow, and the defect rate of joints is 8.03%, which increases the resistance to crack propagation and enhances the shear strength of joints. The optimized joints achieve a maximum lap shear strength of 32.9 MPa, representing 63.1% of the CFRP base material strength.

Introduction

The rapid development of science and technology has promoted the unprecedented transformation of manufacturing industry. Lightweight design, as an important technical means to save energy, improve performance and reduce cost, has attracted great attention from industry. ¹ Especially in transportation and aerospace engineering, lightweight is an important index for evaluating transportation capacity and flight performance. ² The realization of lightweight design at the constraint of structural strength and stiffness mainly depends on two key ways: improving material property and optimizing structural configuration. The comprehensive application of composite materials and light alloys conforms to this development trend. ³ For instance, the synergistic combination of carbon/glass fiber-reinforced composites with aluminum/magnesium/titanium alloys in critical components (e.g., car body, driver cab and equipment compartment) achieves complementary material advantages while effectively reducing vehicle weight in railway vehicle manufacturing. ⁴ In aerospace engineering, aircraft structures typically employ aluminum/magnesium alloys combined with carbon/glass fiber-reinforced polymers. These composite materials demonstrate exceptional flexibility to accommodate complex structural designs while maintaining optimal strength-to-weight ratios, thereby providing essential technical support for modern aircraft to realize adaptive and autonomous flight capabilities. ⁵

Carbon fiber-reinforced polymer (CFRP) and aluminum alloys are extensively utilized lightweight materials across multiple engineering domains. The dissimilar joining of the materials can streamline assembly processes while enhancing connection reliability and durability. ⁶ However, the significant physicochemical property disparities between CFRP and aluminum alloys has brought continuous challenges in engineering applications. ⁷ At present, the joining methodologies mainly include adhesive bonding, mechanical fastening and welding techniques. Adhesive bonding is achieved by polymer adhesives, but there are inherent limitations, including low joint strength and easy aging by adhesive. ⁸ Mechanical fastening introduces connectors to establish structural continuity, but inevitably damages the integrity of materials through local structural damage and stress concentration. ⁹ Friction stir welding (FSW) is a solid-state joining process. Friction heat is generated by the high-speed rotating tool, causing local materials to plasticize, and then the dense weld is consolidated under axial forging pressure. The FSW process between CFRP and aluminum alloys takes place under thermally and chemically asymmetric conditions characterized by differential material flow behaviors, cooling rates and stress distributions. The complex thermo-mechanical interaction involves successive stages of thermal softening, plastic deformation and rapid solidification. The inherent chemical incompatibility between these materials results in suboptimal interfacial bonding, frequently manifesting as weak interfacial defects. Furthermore, the significant differences in physical properties including density, thermal conductivity, thermal expansion coefficient and specific heat capacity will produce uneven temperature and stress field distributions during welding. Localized overheating may induce CFRP carbonization, while post-weld cooling exacerbates residual stress accumulation. ¹⁰ These synergistic effects often produce characteristic defects such as surface grooves, internal voids, and microcracks within the joint region. Therefore, research on the formation of FSW joint between CFRP and aluminum alloys is still in the exploratory stage. It is necessary to carry out systematic research to establish a reliable connection agreement between dissimilar material, so as to promote lightweight manufacturing technologies.

In the investigation of FSW between polymer-based composites and metal materials, scholars have carried out extensive research to explore the joining characteristics of different materials. The primary focus lies in achieving effective bonding between polymer matrix composites and lightweight alloys by FSW technology, and to clarify the potential bonding mechanisms at the same time. The results have revealed that plasticized metal infiltrate into the composite material and molten polymer matrix composites penetrate into the metal at the elevated temperatures. The mutual flow and interweaving of dissimilar materials create a continuous and compact interfacial structure, thereby ensuring favorable mechanical properties of the joints.^11,12 Regarding FSW of polymer matrix composites and lightweight alloy dissimilar materials, Khodabakhshi et al. ¹³ conducted experiments on 5059 aluminum alloy and high-density polyethylene (HDPE) dissimilar friction stir lap welding (FSLW) joints. Their findings show that serious plastic deformation occurred in the stir zone and the honeycomb structure of aluminum alloy (average size <100 nm) is embedded in the polymer matrix. Mahmoudi et al. ¹⁴ demonstrated that the high thermal conductivity of 1120 aluminum alloy accelerates heat transfer to the HDPE plate and leads to the surface melting of the polymer. The molten polyethylene subsequently flows into aluminum and its oxide layers by the tool, exhibiting a wetting phenomenon. Huang et al. ¹⁵ revealed that the dynamic material flow caused by the friction stir spot welding (FSSW) tool in the composite of 6061-T6 aluminum alloy and CFRP promotes plastic deformation of metal, forming aluminum anchors that penetrate through lap joint intersections. These anchors are characterized by curved, deformed and elongated particles, which penetrate into the molten and resolidified composite material. Geng et al. ¹⁶ conducted a study to evaluate the feasibility of tool geometry modifications in pinless friction spot joining (FSPJ) between 6061-T6 aluminum alloy and CFRP with the aim of enhancing joint performance. The interfacial microstructure, bonding strength, and fracture surfaces of joints produced using six distinct tools under varying rotational speeds were systematically analyzed and compared. Based on a comprehensive assessment of the thermal history and mechanical properties, the use of concave-shaped tools was recommended for FSPJ of aluminum and CFRP. Nagatsuka et al. ¹⁷ performed friction stir lap welding (FSLW) to join CFRP and A5052 aluminum alloy. It was found that surface grinding of the aluminum alloy promoted the formation of aluminum hydroxide, which contributed to increased tensile and shear strength of the joint. As the welding speed was raised from 100 mm/min to 1600 mm/min, the tensile shear strength was observed to initially increase and subsequently decrease. Meng et al. ¹⁸ proposed a friction self-riveting (FSRW) process, characterized by multi-scale mechanical interlocking and bonding, to achieve reliable joints between metals and CFRP. The composite material was softened and subsequently flowed into preformed holes and porous structures under the combined effects of frictional heat and forging pressure, resulting in enhanced macroscopic and microscopic mechanical interlocking that improved the overall mechanical properties. Ashong et al. ¹⁹ elucidated that frictional heat transfer from 6014 aluminum alloy to CFRP induces material interdiffusion, forming a thin polymer-metal melt layer at the interface that acts as an adhesive between the dissimilar materials. Bolouri et al. ²⁰ identified hooking behavior at the interface of 1050 aluminum alloy and CFRP friction stir lap welding joints, where CFRP embeds into the undulating deformed aluminum surface. This phenomenon promotes molten polymer matrix infiltration into surface crevices of the alloy. Sun et al. ²¹ demonstrated that appropriate tool offset enables effective mechanical interlocking structures in butt joints between 6061-T6 aluminum alloy and polycarbonate (PC). The disordered interface formed by the interpenetration of polymer-metal fragments along the seam line significantly enhances the mechanical properties of the seam.

The joining of polymer matrix composites to non-aluminum metals by FSW represents a critical focus and challenge in the field of mechanical vehicle joining and design, with extensive investigations conducted by researchers. ²² Choi et al. ²³ studied the FSW joint between pure titanium and CFRP and found that the polymer is thermally decomposed when the interface temperature exceeded the thermal decomposition threshold of CFRP (350°C). The fused CFRP and plasticized metal form interlocking or interwoven structures at the joint interface, which realizes the firm combination between the two materials. Pandey et al. ²⁴ employed friction stir spot welding (FSSW) to join copper (Cu) and polymethyl methacrylate (PMMA). The polymer material are driven into the rough metal surface and gather along the bending area of the copper sheet, which promotes the refinement and mixing of materials. Wu et al. ²⁵ successfully bonded oxygen-free copper to CFRP using friction lap joint (FLJ) technology. Continuous metal infiltration into CFRP eliminated void or gap defects, while a thin Cu₂O transition layer formed at the interface, demonstrating direct nanoscale bonding between Cu₂O and the polymer. Moghanian et al. ²⁶ also achieved joining between pure magnesium and polypropylene (PP), with microscale mechanical interlocking structures observed in the stir zone. However, significant differences in thermal expansion coefficients and mechanical properties between dissimilar materials often lead to stress concentration and crack formation at the interface. To address this, Wang et al. ²⁷ developed an innovative hybrid approach combining FSW with mechanical interlocking technology—termed friction stirring interlocking (FSI)—to join AZ31 magnesium alloy and CFRP. This method enhances the strength of the joint by forming interlocking structures, and reduces the stress concentration and crack defects at the interface. Nevertheless, it brings new challenges in the precise control of temperature and strain states in welding. Specifically, higher temperatures and effective strains on the advancing side compared to the retreating side result in greater CFRP deflection on the latter, where polymer matrix extrusion predominantly occurs.

In a word, although the existing research has extensively studied the FSW process between polymer and metals by a single process conditions, which provides valuable references for understanding the structure and properties of different joints between polymer-based composites and metals. There are still key challenges in defect control of CFRP/aluminum alloy joints. The significant thermophysical property differences between CFRP and aluminum alloys combine with CFRP susceptibility to fiber degradation and resin matrix decomposition in high-temperature welding. This combination frequently generates porosity defects, severely weakening interfacial bonding strength. In this study, the dynamic performance evolution of CFRP/aluminum alloy FSW joints under parameter adjustments is systematically studied, and the influence of tool pin geometry on driving and distributing plastic deformed aluminum alloy into CFRP is emphatically analyzed. Furthermore, through controlled variations in welding parameters, the formation mechanism of welding defects (such as tunnels and cavities) in the surface and cross-sectional profile caused by material flow mode is experimentally characterized. The interfacial bonding characteristics under different process configurations are comprehensively elucidated through multiscale analysis.

Material and Experimental Procedure

In this study, high-density carbon fiber reinforced plastics (CFRP) with a polyetherimide (PEI) matrix reinforced by 22 vol% randomly oriented short carbon fibers (type T300, nominal aspect ratio ∼50) were used, along with aluminum alloy 6061-T6. Both materials were supplied as 2 mm thick plates. The physical properties of CFRP are systematically shown in Table 1, while Table 2 delineates the chemical composition and mechanical characteristics of the aluminum alloy. Prior to joining operations, the standardized specimens with dimensions of 150 mm × 100 mm × 2 mm are precision-sectioned from both material sheets by CNC machining protocols.

Table 1.

Chemical compositions and mechanical properties of 6061-T6 aluminum alloy Ref. 27.

Tensile strength (Mpa)	Yield strength (Mpa)	Elongation (%)	Incipient melting temperature (°C)	Thermal conductivity [W m−1 K−1]
260	240	9	652	167

Cu	Si	Fe	Mn	Mg	Zn	Cr	Ti
0.4	0.8	0.7	0.15	1.2	0.5	0.35	0.15

Table 2.

Physical and mechanical properties of the CFRP Ref. 27.

Tensile strength (Inplane)(MPa)	Glass transition temperature (°C)	Melting temperature (°C)	Transversal thermal conductivity (Longitudinal) [W m−1 K−1]	Longitudinal thermal conductivity (Transverse) [W m−1 K−1]	Through-thickness thermal conductivity (Thicknesss) [W m−1 K−1]
53	215	260	0.28	4.42	0.30

The friction stir lap welding (FSLW) experiments are conducted using a 6 kW HT-JM20X8/2 FSW machine (Aerospace Engineering Equipment Co., Ltd, Suzhou). The dissimilar materials are arranged in a lap joint configuration with the aluminum alloy positioned atop the CFRP sheet as shown in Figure 1, securely clamped to a backing plate by clamping tools. Aluminum surfaces are mechanically abraded with SiC paper and ultrasonically cleaned to mitigate oxide layer effects before the welding process. A custom-designed H13 tool steel FSW tool featuring a threaded cylindrical pin (⌀10 mm × 2.3 mm height) and concave shoulder (⌀50 mm) is employed. The experimental parameters are determined through statistical optimization employing a full factorial design combined with analysis of variance to maximize lap shear strength. While comprehensive statistical details will be presented in a subsequent publication, the parameter ranges investigate in this study included: rotational speed (1800 rpm), welding speed (10 mm/min-25 mm/min), plunge depth (2.5 mm), and tool tilt angle (2°). This parametric window is established through preliminary trials as the operational domain for achieving joints.

Figure 1.

Welding schematic diagram.

Transverse cross-sections of the welded joints are extracted from the weld centerline region by the cutting equipment. The samples are prepared by standardized metallography, including continuous grinding with SiC abrasive papers (300-2000 grit) and then suspension polishing with diamond. Microstructural revelation is achieved through 30 s etched in Keller reagent (composition: 95 vol% H₂O, 2.5 vol% HNO₃, 1.5 vol% HCl, 1 vol% HF). Multi-scale morphological analysis is performed using complementary characterization techniques: optical microscopy (OM, Leica DM IRM) for macrostructural observation and scanning electron microscope (SEM, JSM-6360LV) for three-dimensional surface topography. Quasi-static lap shear tests are conducted on a universal testing machine with ASTM D3163-01 specifications (The sampling position is shown in Figure 1). The crosshead displacement rate is maintained constant at 5 mm/min under the environmental laboratory conditions. The ultimate shear strength values are derived from three samples of each parameter group, and the statistical reliability is ensured by the average data report.

Results and Discussion

Surface Morphology

The surface morphology of the weld seam at different welding speeds is shown in Figure 2. In Figure 2(a) and (b), the relative motion between the pin tool and the material inside the weld seam decreases per unit time with lower welding speeds (10 mm/min and 15 mm/min). The stirring pin has more time to contact and act on the aluminum alloy and CFRP, and its shearing force on the material increases accordingly. The increased shear force is transformed into the extrusion force on the welded sheet, which leads to the overflow of CFRP from the surface holes of the incompletely fused aluminum alloy. At the welding thrust of the stirring pin, the holes are elongated to form obvious tunnel defects. At the same time, the residence time of the shoulder of the pin tool is increased on the sheet, and the polymer matrix composites have more opportunities to adhere to the weld surface.

Figure 2.

Surface morphology of weld seam at different welding speeds.

When the welding speed is increased to 20 mm/min, the surface morphology of the weld becomes smoother with obvious fish scales, as shown in Figure 2(c). However, as the welding speed further rises to 25 mm/min, the weld zone is subjected to instantaneous heat input, and the sufficient softening of the aluminum alloy and CFRP is prevented by the shorter dwell time of the pin tool at the joint. Consequently, the amount of plastically deformed metal and molten CFRP is inadequate during weld formation. Moreover, the rapid movement of the pin tool hinders the fusion between the dissimilar materials, resulting in an insufficiently strong bond of the joint. This leads to surface collapse in the weld seam and forms tunnel defects, as shown in the upper right corner of Figure 2(d).

Internal defect of joint

The microstructure of the joint section is shown in Figure 3. There are hole defects at different welding speeds, as highlighted in the magnified views of the marked regions in Figure 3. Compared to Figure 3(b) (c) (d), the voids in Figure 3(a) exhibit a higher density in quantity and a tendency toward larger sizes. The melting and flow behaviors of the metal and polymer composite materials are often asynchronous under thermo-mechanical effects during the friction stir process. Even if the welding speed is adjusted to optimize the surface morphology of the weld seam, it is inevitable that there will still be areas of insufficient melting in the weld seam. These under-melted areas are characterized by void defects in the cross section of the joint.

Figure 3.

Defects in the cross-section of the joint weld at different welding speeds.

With decreasing welding speed under high-heat-input conditions, the frictional heat generated at the interface between the stirring tool and the materials, as well as the heat due to plastic deformation, is significantly increased. Meanwhile, a considerable difference in thermal conductivity between aluminum and CFRP (as provided in Tables 1 and 2) results in a distinct thermal gradient across the Al/CFRP interface. On the aluminum side, heat is rapidly dissipated owing to its high thermal conductivity, whereas in CFRP, heat accumulates within the interfacial region, leading to a localized temperature increase and non-uniform distribution. This thermal environment promotes the degradation of polymers and further entanglement of molecular chains within the CFRP matrix, which increases viscosity and limits its flowability in the molten state, making it significantly lower than the flowability of metals (as shown in Figure 4, where the temperature on the forward and backward sides of the weld seam was detected using a thermal imaging instrument, and temperature data points were continuously collected along the entire length of the weld seam, with a temperature range of 200°C-350°C in the welding direction, higher than the thermal decomposition temperature of PEI (CFRP matrix)).). Meanwhile, the aluminum alloy forms a continuous structure in the weld zone due to its favorable thermal conductivity and plastic flow behavior. In contrast, the CFRP, subjected to localized overheating induced by the thermal gradient and further hindered by its inherently poor flowability, cannot adequately diffuse or fill the weld space. This leads to material deficiency in certain regions of the joint. As a result, the homogeneity of material mixing between the two materials is reduced, ultimately resulting in the formation of porosity defects. Furthermore, the decrease of welding speed promotes the accumulation of heat in the weld area, the CFRP material is ablated and the material loss occurs in some areas. Although the metal has better thermal conductivity and enable rapid heat dissipation, the extended thermal cycling time negatively impacts the density of joint, further contributing to void defects in the cross-sectional region.

Figure 4.

Joint temperature with welding speed of 20 mm/min.

To visually represent the distribution of defects within the joint, the micrograph of the joint is analyzed using ImageJ software to calculate the ratio of the defect-free area to the entire joint region. The original image is first converted into a grayscale format. Each labeled connected region is then traversed, and its pixel count is recorded. Finally, the proportion of pixels in each connected region relative to the total pixel count of the entire area is calculated. A schematic diagram of this procedure is shown in Figure 5. The statistical results of the proportion of defective zones are summarized in Table 3. As the welding speed increases, the proportion of defective zones is measured as 18.82%, 25.1%, 8.03%, and 31.76%, respectively. A higher defect proportion is found to significantly reduce the mechanical properties of the joint, making it more prone to failure.

Figure 5.

Schematic diagram for calculating the proportion of defective zones.

Table 3.

The proportion of defective zones at different welding speeds.

Welding speed (mm/min)	10	15	20	25
The proportion of defective zones	18.82%	25.1%	8.03%	31.76%

Microstructure of Dissimilar Joints

In the initial phase of friction stir welding, the pin tool applies combine rotational and downward forces to the workpieces. The high-speed rotation of pin mechanically cuts materials within the weld zone, while the specially designed thread geometry on the pin induces directional material flow along predetermined paths. Simultaneously, the surface layer of the material in the center of the weld is also deformed by the direct extrusion of the shoulder of the mixing head, resulting in a certain depth in the action zone of the shoulder of the joint. Therefore, the metal near the shoulder is deformed by flow under the direct influence of the shoulder, and this deformation area is defined as the shoulder-affected zone. The depth of the affected zone of the joint shaft shoulder at different welding speeds is measured several times, and the measurement results are shown in Figure 6. The change of welding speed will significantly change the depth of the affected zone of the shoulder. Specifically, Figure 6(a) exhibits a deeper penetration depth of 0.54 mm at the welding speed of 10 mm/min, contrasting with the shallower depth of 0.22 mm observed in Figure 6(d) at reduced welding speeds.

Figure 6.

The size of the area affected by the joint at different welding speeds.

Aluminum alloy and CFRP materials gradually soften under the friction and thermal effects generated by the stir tool during the welding process. The softened materials exhibit enhanced fluidity, which allows material flow within the core zone of the weld not only to penetrate the surface layer, but also to gradually penetrate into the interior of the weld. As the welding speed increases, the contact time between the stir tool and the materials decreases, thereby limiting the duration for bonding and penetration between the aluminum alloy and CFRP. With an increase in welding speed, the contact time between the stir tool and materials diminishes. Consequently, the bonding and penetration process between aluminum alloy and CFRP becomes time-constrained, making it difficult for fragmented metallic particles in the aluminum alloy sheet to infiltrate the polymer matrix of the CFRP sheet. This restriction limits the flow and deformation capacity of material within the shoulder-affected zone, resulting in a reduced depth of shoulder-affected zone. Conversely, the ample contact time allows sufficient bonding between the aluminum alloy and CFRP at lower welding speeds. Additionally, the prolonged compressive force from the shoulder of pin tool causes greater extrusion deformation on the upper layer of materials in the weld center, thereby increasing the depth of the shoulder-affected zone.

The microscopic features of the joint bonding region are shown in Figure 7. In addition to existing as fragmented particles, the aluminum alloy is also observed in the form of “hook-like” structures, appearing as bands or blocks within the joint. The embedding angle of the aluminum alloy on the retreating side into the CFRP is measured as indicated by the angle labels in Figure 7. With increasing welding speed, the embedding angles are measured as 52°, 70°, 44°, and 18°, respectively. Local magnification of these regions reveals imperfectly bonded zones around multiple hook-like structures. The widths of these imperfectly bonded regions are measured at locations marked by double arrows in Figure 7. The average widths of the loose bonding zones are determined to be 18.46 μm, 81.54 μm, 46.15 μm, and 43.08 μm, respectively, as summarized in Table 4.

Figure 7.

Micro-bonding of joint materials at different welding speeds.

Table 4.

The average widths of the loose bonding zones at different welding speeds.

Welding speed (mm/min)	10	15	20	25
The average widths of the loose bonding zones (μm)	18.46	81.54	46.15	43.08

The local friction heat generated in the welding zone causes high temperatures and keeps the aluminum alloy in a plastic state, which is characterized by softened internal structures and enhanced material flow. The screw structure of the pin provides a flow channel and path for the metal, so that the material gradually traverses the whole joint. Through rotational motion of the pin tool, the aluminum alloy undergoes directional flow along the helical path while being subjected to intense shear forces. This dynamic interaction results in the formation of plastically deformed lamellar structures, where the aluminum alloy becomes mechanically stirred into elongated or block-shaped fragments. With sustained welding progression, the plastically deformed aluminum segments undergo further fragmentation into smaller-dimension particles through continuous shear deformation.

When the welding speed ranges from 10 mm/min to 15 mm/min as shown in the analysis of Figure 7, the shear force predominantly transforms into compressive force acting on the material. The stirring tool applies compressive stress to fragmented aluminum alloy particles. The stress triggers remodeling processes that progressively compact the particles into denser blocky configurations. The threaded geometry and rotational motion of the pin tool generate directional flow channels. These channels guide the alignment of blocky structures, forming oriented interconnections along predetermined trajectories. Ultimately, these blocky structures further consolidate into banded configurations by continuous stirring action, as illustrated in Figure 7(a) and (b). At an increased welding speed of 20 mm/min, the joint metal experiences significant plastic deformation. The deformation process generates characteristic hook-like structures accompanied by fine fragmented particles. During the friction stir lap welding process, the rotational and translational motion of the stirring tool typically induces severe plastic deformation in the aluminum alloy. This dynamic interaction facilitates the formation of characteristic “hook-shaped” or “interlocking” microstructures within the weld zone. These architecturally complex features become embedded within the CFRP matrix interface, establishing a robust mechanical interlocking effect that enhances interfacial bonding integrity. When the welding speed reaches 25 mm/min, the reduced heat input and shorter thermal dwell time result in accelerated cooling rates. This thermal condition leads to premature solidification of the aluminum alloy before achieving sufficient microstructure refinement. Therefore, as observed in Figure 7(d), the joint metal is present predominantly in a bulk form. Compared to band-shaped structures, the coarse morphology of the bulk aluminum alloy reduces the effective contact area with the CFRP. Moreover, the embedding angle of the aluminum into the CFRP is measured as only 18°. Under external loading, the joint is more prone to fracture as a result of these morphological characteristics.

Due to the difference in the coefficients of thermal expansion between aluminum alloy and CFRP, deformation is readily induced at the interfacial region between the two materials. As heat accumulates, the discrepancy in thermal expansion becomes more pronounced, leading to the formation of microscopic defects within the materials. After welding, as the heat dissipates, the welded zone begins to cool. At this stage, the aluminum alloy and CFRP exhibit distinct shrinkage behaviors due to their different contraction rates. Consequently, the two materials contract and exert mutual pulling forces at the interface during cooling, which amplifies the deformation and results in the formation of crack defects. As shown in Figure 7(a) and (b), the presence of multiple band-shaped aluminum structures further accentuates this phenomenon. The interfacial bonding between the band-shaped aluminum and CFRP is relatively weak, resulting in a loosely bonded region between the dissimilar materials. These regions are characterized microscopically by gaps, micro-cracks, or incomplete bonding, which collectively reduce the overall toughness of the joint. For instance, when the welding speed is 15 mm/min, the average width of the poorly bonded region reaches 81.54 μm. Under external loading, these poorly bonded regions act as weak zones within the joint.

The SEM and EDS results of the joint produced at a welding speed of 20 mm/min are shown in Figure 8. The EDS spectra and quantitative analysis of points A and B in Figure 8(a) are presented in Figure 8(b) and (c), respectively. The carbon and oxygen contents at point A are measured as 12.85% and 42.3%, respectively, while those at point B are 41.54% and 38.1%. Elemental mapping images are shown in Figure 8(c) to (f).

Figure 8.

EDS analysis and SEM microstructure of CFRP on the joint at 20 mm/min.

The variations in aluminum and oxygen contents among different test points reflect localized differences in the penetration of the aluminum alloy into the CFRP matrix. ²⁸ The high oxygen content observed at point A indicates that the oxide layer from the aluminum alloy base material is fragmented and distributed within the joint, forming oxygen-enriched zones at various locations on the cross-section. The polymer undergoes decomposition reactions by the effects of frictional heat and mechanical stirring, which releases small oxygen-containing molecules that are incorporated into the weld zone and increase the oxygen content across the cross-section. ²⁹ These two factors collectively determine the distribution characteristics of oxygen.

The fracture inclination angle of a joint is defined as the acute angle between the final fracture surface and a reference axis parallel to the direction of the applied load during tensile-shear testing. It serves as a geometric parameter that characterizes the fracture path features of the joint. With the increase in welding speed, the fracture cross-sections of dissimilar joints are shown in Figure 9. At welding speeds of 10 mm/min and 15 mm/min, large-sized void defects are observed within the joints, predominantly concentrated at the center of the dissimilar material stirring zone. Cracks preferentially start from these defective areas and propagate in a specific direction under the external load, which leads to the increase of the inclination angle between the fracture path and the horizontal direction. Furthermore, the analysis of Figure 7 reveals that the mismatch in thermal expansion coefficients and shrinkage rates between aluminum alloy and CFRP during heating and cooling processes induced volumetric incompatibility. The mismatch creates loosely bonded regions at the joint interface, which served as sensitive zones for crack propagation and further amplified the inclination angle relative to the horizontal plane. Consequently, the fracture crack angles relative to the horizontal direction are measured as 40° and 28° for joints welded at 10 mm/min and 15 mm/min, respectively, as shown in Figure 9(a) and (b).

Figure 9.

Fracture crack angle of joint at different welding speeds.

When the welding speed is increased to 20 mm/min and 25 mm/min, the aluminum alloy underwent plastic deformation to form banded structures adjacent to the heat-affected zone (HAZ). In contrast to the low-speed welding scenario, fracture cracks no longer propagate directly through the relatively brittle central stirring zone. Instead, cracks exhibite a tendency to propagate along the interfaces between the banded structures and the surrounding matrix material. The altered propagation mechanism consequently reduces the inclination angles between fracture paths and the horizontal direction to 23° and 18° for welding speeds of 20 mm/min and 25 mm/min, respectively, as shown in Figure 9(c) and (d).

The fracture surface morphology of the joint produced at 20 mm/min is presented in Figure 10(a). Region A indicated in Figure 10(a) is magnified, and the resulting image is shown in Figure 10(b). Pit features are observed on the CFRP surface adhered to the aluminum alloy side. The presence of these pits provides direct evidence of plastic flow experienced by the materials during welding, which reflects both the dynamic response and deformation behavior of the materials in the welding process. This observation is also consistent with the void defects found in the cross-sectional area of the joint, as shown in Figure 3.

Figure 10.

Fracture surface morphology of joint at 20 mm/min.

Mechanical Properties

The hardness values of the dissimilar joint under different welding speeds are presented in Figure 11. The average hardness of the aluminum alloy fragments is lower than that of the aluminum alloy base material, which is attributed to the coarsening and bending of grains in the bonding zone caused by the welding heat, resulting in the formation of a softened region with reduced hardness. As the welding speed increases, the heat input decreases, thereby avoiding element segregation during alloy solidification under high-temperature conditions and enhancing the microhardness of the aluminum alloy. Consequently, the hardness of the aluminum alloy gradually increases. The average hardness in the CFRP region of the joint is higher than that of the base material (23.3 HV). In the joint bonding zone, the embedded aluminum alloy fragments act as barriers within the molten and re-solidified CFRP, restricting its flow deformation and thus improving the local hardness.

Figure 11.

Microhardness of dissimilar joint.

The variation in tensile-shear strength of dissimilar joints with welding speed is shown in Figure 12 and Table 5. At welding speeds of 10 mm/min and 15 mm/min, rough weld surfaces accompanied by carbonized polymer overflow are observed. Large-scale void defects identified in joint cross-sections significantly reduce the load-bearing capacity against tensile-shear forces. When the welding speed increased to 25 mm/min, the range of the shoulder-affected zone contracted, and the shortened dwell time of the stirring tool induced blocky structures in the plastic ally deformed aluminum alloy, thereby constraining its bonding with CFRP. At the welding speed of 20 mm/min, a smooth weld surface is achieved with no macroscopic void defects detected in cross-sectional observations. However, the reduced tool dwell time limit sufficient penetration of metallic particles into the polymer matrix, which impairs material flow deformation capability in the shoulder-affected zone and diminishing its penetration depth. The formation of banded “hooked” structures embedded within CFRP enhanced interfacial mechanical interlocking, ultimately increasing the critical fracture load required for joint failure.

Figure 12.

Tensile-shear results of joints at different welding speeds.

Table 5.

Tensile shear strength of joints at different welding speeds.

Welding speed (mm/min)	Tensile shear strength (MPa)	Ratio of CFRP matrix
10	19.7	37.8%
15	28.5	54.7%
20	32.9	63.1%
25	18.6	35.7%

Conclusions

In this study, 6061-T 6 aluminum alloy and CFRP are taken as typical lightweight materials, and the research goal is to clarify the joint formation mechanisms and quantitatively evaluate the influence of welding parameters on the interface properties. During the FSW of aluminum alloy and CFRP dissimilar materials, precise regulation of welding velocity facilitates controlled plastic deformation and material flow, thereby generating well-consolidated weld seams with optimized surface morphology. It is worth noting that heat energy input causes significant plastic deformation in the aluminum alloy, which enhances the interface contact area with CFRP and promotes the mechanical interlock between different materials. However, dynamic change of stirring seriously affects the microstructure evolution in the welding nugget zone. Aluminum alloys tend to break into coarse fragment structures, and their discontinuous accumulation damages the integrity of joints by forming micropore defects. The area of defects in the joint accounts for up to 31.76% of the total area of the joint, and micropore defects become the priority position of crack nucleation. In addition, post-weld cooling will cause different interface deformation due to the mismatch of thermal expansion coefficient and shrinkage rate between materials, resulting in weak bonding interface area. The damaged areas of these structures show high sensitivity to crack propagation under external load conditions. The optimized joint configuration has achieved a maximum lap shear strength of 32.9 MPa, which represents 63.1% of the inherent strength of CFRP substrate materials.