Underwater friction-stir welding of a stir-cast AA6061-SiC metal matrix composite: optimization of the process parameters, microstructural characterization, and mechanical properties

Aluminum alloy was employed as a matrix material. Because of its strong corrosion resistance, great machining qualities, low weight, and ductility, the aluminum alloy was chosen. The alloy Al 6061 is mostly utilized in the aerospace sector. 6061/SiC AMC was employed in this study, as well as UWFSW 6061 alloy enhancement together with SiC particles. The matrix is a standard 6061 aluminum alloy with the specified chemical composition, as provided in Table 1. The stir-casting approach was used to generate SiC enhancement 6061 matrix composites in a top-loading electric resistance muffle furnace, as indicated in the literature [1]. The average size of the SiC particles was 400 mesh (40 m). Castings were reinforced with 5 SiC wt.%, 10 SiC wt.%, and 18 SiC wt.% scale (150 mm × 750 mm × 10 mm) composite specimens that have already been slashed, cleaned, and butt-welded along the joint line for UWFSW.

The geometrical dimensions are shown in Figure 1. The conical pin was designed to facilitate piercing the tool to the specimens as straightforwardly as possible during the welding process. The welding experiments were performed on an automated vertical milling machine with a custom-made fixture to clamp the welding plates, as shown in Figures 1B and 1C.

Fig. 1

Following the standard metallographic procedure, mirror polishing and etching with Keller's reagent (HCL + HF + HNO₃ + distilled water) were performed. The metallographic specimens were made from 6061 AMC and 6061 AMC that had been welded together. According to ASTM E8-04, the standard tensile specimens were sliced perpendicular to the joint line and created from all specimens [18]. TS has been prepared by the length of a gauge of 57.2 mm, the width of a gauge of 12.2 mm, and a thickness of 10 mm. The tensile specimens were done using a Blue Star universal tensile testing machine (SE UTE 200) with an extreme capacity of 200kN, as indicated in Figure 2. A Chinese-made optical scanning microscope was used to examine the microstructure of UWFS-welded specimens (XPZ-830 T) [16]. In the same way, the Al-matrix stir cast as a composite cast aluminum matrix and the welded junction were microstructurally examined. The UWFS-welded and other specimens were MHV using the FIE model VM50 VH tester. VH measurements were taken at various locations on both sides of the weld zone under a constant load of 0.8 kgf for a dwell period of 17 s. The hardness data were then categorized as average microhardness [17, 19].

Fig. 2

The surface finishing for each UWFSW specimen of Al 6061/18% SiC gave a smooth weld surface without lateral flash; meanwhile Al 6061/5% SiC shows an unsmooth weld surface condition. Based on observation, we can infer that this is the relation between rotation speed, travel speed, the ratio of SiC, and flash occurrence. The lower values of rotation, travel speed, and SiC 18% resulted in an absence of lateral flash. In lieu, the decrease in SiC may direct to the presence of lateral flash. Hence, it explains that the welding particulars may influence the character of the external surface. Tensile specimens have been prepared in standard dimensions according to ASTM E8 as shown in Figure 2 and tested using the universal testing machine. A Santam STM-50 machine has been used to evaluate tensile properties. The testing machine had a crosshead speed of 5 mm/min. During the test, stress and strain have been recorded simultaneously, as shown in Figure 4 and Figure 5. The maximum UTS and YS have been obtained for specimens produced by Al 6061/18% SiC. The corresponding combination of tool rotational speed and travel speed has been 1,800 rpm and SiC 18%, respectively. The UTS and the YS have been 228 MPa and 140.3 MPa, respectively, which have been 91.2% and 92.3% of the base material.

Fig. 4

Fig. 5

Table 2 shows how the Box–Behnken model is used to create an experimental design matrix. FSW on hybrid composites was carried out by varying process parameters such as N, S, and SiC following the experimental design matrix provided in Table 3. According to ASTM standards, test specimens for the tensile and VH tests were made using manufactured welded joints. To create the empirical mathematical relationships, UTS and WNH were measured from each test specimen. The duties of N, S, and SiC were the UTS and WNH of the UWFSW of AA6061/SiC's UWFSW. Eqs (2) and (3) can be used to describe the response surface, as can Eq. (4), which is a second-order polynomial (regression) [20].

The quadratic model is created first. The ANOVA method is used to identify significant responses and interaction terms. The F-statistic is used to determine the significance factor. p-value, F-value, degree of freedom, the sum of squares, mean sum of squares, coefficient of variation, determination coefficient, adjusted determination coefficient, and other statistical analyses were utilized. The sum of squares obtained from the ANOVA is used to evaluate the percentage of contributions for each component. The properties of independent and dependent variables are learned using ANOVA in a quadratic polynomial equation with both independent and dependent variables. Eqs (6) and (7) are the final regression relations for calculating the UTS and WNH of a UWFS-welded joint for AA6061/SiC: (6) UTS=161.09418+3.70237xA− 0.010857xB+0.095210xC− 2.81977E−004xAxB− 5.14535E−003xAxC− 4.00000E−004xBxC−0.070669xA∧2+ 1.13889E−005xB∧2+ 1.30556E−003xC∧2 \matrix{ {{\rm{UTS}}} \hfill & { = 161.09418 + 3.70237x{\rm{A}}} \hfill \cr {} \hfill & { - \;0.010857x{\rm{B}} + 0.095210x{\rm{C}}} \hfill \cr {} \hfill & { - \;2.81977{\rm{E}} - 004x{\rm{A}}x{\rm{B}}} \hfill \cr {} \hfill & { - \;5.14535{\rm{E}} - 003x{\rm{A}}x{\rm{C}}} \hfill \cr {} \hfill & { - \;4.00000{\rm{E}} - 004x{\rm{B}}x{\rm{C}} - 0.070669x{{\rm{A}}^ \wedge }2} \hfill \cr {} \hfill & { + \;1.13889{\rm{E}} - 005x{{\rm{B}}^ \wedge }2} \hfill \cr {} \hfill & { + \;1.30556{\rm{E}} - 003x{{\rm{C}}^ \wedge }2} \hfill \cr } (7) WNH=50.31928+0.48867xA− 3.60481E−003xB−0.21980xC+ 1.88808E−004xAxB−0.011438xAxC− 1.15979E−004xBxC− 4.26816E−003xA∧2+ 2.92847E−006xB∧2+ 8.59556E−003xC∧2 \matrix{ {{\rm{WNH}}} \hfill & { = 50.31928 + 0.48867x{\rm{A}}} \hfill \cr {} \hfill & { - \;3.60481{\rm{E}} - 003x{\rm{B}} - 0.21980x{\rm{C}}} \hfill \cr {} \hfill & { + \;1.88808{\rm{E}} - 004x{\rm{A}}x{\rm{B}} - 0.011438x{\rm{A}}x{\rm{C}}} \hfill \cr {} \hfill & { - \;1.15979{\rm{E}} - 004x{\rm{B}}x{\rm{C}}} \hfill \cr {} \hfill & { - \;4.26816{\rm{E}} - 003x{{\rm{A}}^ \wedge }2} \hfill \cr {} \hfill & { + \;2.92847{\rm{E}} - 006x{{\rm{B}}^ \wedge }2} \hfill \cr {} \hfill & { + \;8.59556{\rm{E}} - 003x{{\rm{C}}^ \wedge }2} \hfill \cr } The ANOVA approach confirms the appropriateness of the final established regression relationships. If the calculated value of the developed model's F-ratio is smaller than the standard F-ratio value from the F table at the specified degree of confidence, it verifies the developed model's acceptability (say 95%) [14, 21].

From the interaction plot shown in Table 4, we find that the interaction of the tool rotating speed and the welding speed of the silicon carbide was not significant (p-value > 0.05), while the interaction between the tool rotating speed and the silicon carbide was significant (p-value < 0.05). Table 5 shows the ANOVA result for the microhardness. From the statistical analysis considering three parameters and three levels, it is evident that all the main effects of the parameters and their interaction effects are not significant to the microhardness as their p-values are >0.05 in a 95% confidence interval.

The effect of UWFSW parameters on UTS is represented in the perturbation plot (Figure 6) for an optimized design. From the plot, it is clear that the response changes when each factor moves away from the common or reference points keeping all other factors constant at that reference point. It is also clear from the plot that the tool rotating speed (A) is the most dominating factor on UTS followed by welding speed (B) and SiC (C).

Fig. 6

The interaction influences of N, S, and SiC on the UTS are depicted in Figures 7–9. The 2D and 3D contour plots demonstrate the influence of N and S on UTS arising from retaining AA6061/18 SiC (Figures 2A and 2B).

Fig. 7

Fig. 8

Fig. 9

The ideal value of the UTS (MPa) is reached in the center of the contour plot, which is shown as a concentric circle (Figures 2A and 2B). The optimal UTS of 210 MPa is attained at an N of 1,800 rpm and an S of 10 mm/min, as shown in the plot. When the N and S diverge from the specified values, the UTS will tend to drop or grow.

Figure 9 depicts the interaction impact of SiC and S at a constant N of 1,800 rpm. The approximate value of the optimal UTS of 210 MPa is attained at an S of 10 mm/min and a SiC of 18%, as illustrated in the 2D and 3D contour plots.

The perturbation plot (Figure 10) for an optimized design depicts the effect of UWFSW process parameters on WNH. At any operating level of N, WNH should be higher than that of base metal. The microhardness of base metal AA6061/18% SiC is 60.2 HV and it is usually lower compared to the stir zone (60.2 HV). The interaction impact of N, S, and AL on the WNH is depicted in Figures 11–13. Figures 13A and 14B show the effect of N and S on WNH while keeping the SiC constant at 18%.

Fig. 10

Fig. 11

Fig. 12

Fig. 13

Fig. 14

WNH (HV) is depicted as a concentric ellipse, with the ideal WNH value located in the contour plot's center (Figures 12A and 12B). The ideal WNH of 60.2 HV is attained at an N of 1,800 rpm and an S of 105 mm/min, as illustrated in the plot. Figure 15 depicts the interaction impact of SiC and N at a constant S of 10 mm/min. The optimal WNH is attained around 60.2 HV at an N of 1,800 rpm, as seen in the 2D and 3D graphs. Figure 13 depicts the interaction impact of SiC and S at a constant N of 1,800 rpm. We observe from the 2D and 3D plots that the ideal WNH of 60.2 HV is attained at an S of 10 mm/min and a SiC of 6 kN.

Fig. 15

The desirability function approach is a widely utilized technique in multiple response optimization. Harrington first presented the desirability function technique in 1965 [14, 20, 21]. This approach determines the optimal operating conditions for obtaining the desired response values. All outputs are translated to individual DFs with a scale factor of between 0 and 1. These functions are organized by initializing the values to be the target, minimum, or maximum output obtained during the experiments. The optimal conditions of UWFSW parameters to maximize hardness and ultimate tensile strength of UWFS-welded Al 6061/5%wt. SiC, Al 6061/10%wt. SiC, and Al 6061/18% SiC were determined using response surface methodology. The optimized UWFSW parameters to maximize hardness and ultimate tensile strength of the weld joint were obtained using a ramp chart and bar chart, as shown in Figure 14. The optimized parameters were N, a rotation speed of 1,736.36 rpm, a travel speed of 11.58 mm/min, and a SiC of 16.67%, with a desirability value of 0.922.

The experimental data are used to validate the model created using the desirability technique, and the errors are estimated for all 27 runs, as shown in Table 2. For UTS and WNH, Table 7 shows the actual value, anticipated value, and percentage error. The real values are discovered through trials, and the anticipated values are derived using the Design-Expert software's empirical equations [22,23,24]. For UTS, the proportion of error ranges from −6.133 to +5.19. WNH's percentage of inaccuracy is similar, ranging from −1.550 to +5.372. As a consequence, the newly created model has predicted UTS and WNH values that are quite similar to the experimental data. The findings of the validation trials are presented in Table 7. The model is also tested for the ideal welding circumstances that are anticipated, with an N of 1,736.36 rpm, an S of 11.59 mm/min, and a SiC of 16.96%.

Figure 15A depicts the presence of coarse grains in the base composite material (AA6061/18% SiC), as well as the dendritic structure caused by the stir-casting process and SiC particle dispersion in the metal matrix. The macrostructural investigation of the UWFS-welded connection is shown in Figure 17B. In the cross weld microstructure of UWFS-welded MMC joints, the four separate zones of UWFSW, WNZ, TMAZ, HAZ, and unaffected zone (base material), can be seen in Figures 16A–16F.

Fig. 16

Due to variable heating and cooling circumstances in UWFSW, it also indicates the presence of diverse microstructure and grain sizes in WNZ, TMAZ, and HAZ. In comparison to the TMAZ, HAZ, and unaffected zones, the weld nugget zone features finer granules. The presence of tiny recrystallized structures in the WNZ is also visible in this photomicrograph. Due to the mechanical stirring action of the UWFSW tool, the coarse grain structure visible in the base composite material (Figure 16A) transforms into a fine grain structure, as seen in Figure 17A–17F. The optical microscopic of the base composite material and the UWFS-welded composite is shown in Figures 16A–16F. The grain size and distribution of SiC in the base composite material (AA6061/18%wt. SiC) are shown in Figure 16A. Figure 16 depicts the existence of the four distinct zones (TMAZ, WNZ, HAZ, and unaffected zone). When compared to the TMAZ, HAZ, and unaffected zone, the WN zone possesses a smooth grain structure (Figure 16C–16F). It demonstrates the existence of reinforcing particles (SiC) in the grain borders, which function as a barrier to dendritic development. MMCs’ UTS and microhardness improve when dendritic development is reduced. It also illustrates how the degree of cluster aggregation and its expansion with the addition of SiC particles gives MMCs a lot of strength. Following the confirmation of the results, an optical microscope investigation of a cracked tensile specimen of FS-welded (AA6061/18%wt. SiC) MMC with optimal settings was performed. The optical microscope report of the microstructure of the broken tensile specimen is shown in Figures 16A–16F. It clearly distinguishes between a non-fractured and a fractured surface area.

Fig. 17

The high-resolution optical micrographs and the XRD patterns of the prepared AMCs are shown in Figures 16 and 17, respectively. The diffraction peaks of SiC, which represent the major elements of the SiC particles, are distinctly identified. The intensities of the peaks rise as SiC content within the matrix increases. It is noticed in Figure 17 that the diffraction peaks of aluminum in the composites are slightly shifted to lower 2θ compared to that of the base alloy due to the addition of SiC particles in the aluminum matrix. It is obvious from Figure 17 that there are no other diffraction peaks detected except the peaks for elements Al, SiC. This observation leads to a conclusion that during the casting of AMCs, the integrity of SiC particles was conserved. At the processing temperature of compo casting, SiC particles are thermodynamically stable. The occurrence of any interfacial reaction between fly ash particles and aluminum matrix during casting has not been observed. Such interfacial reactions in the composites would often result in the formation of brittle intermetallic compounds and degrade their significant properties. A wider zone of interfacial reaction and consequently the formation of brittle MgAl₂O₄ spinal were observed at the interface of aluminum–fly ash in A356/fly ash AMC prepared by stir casting, as reported by Rajan et al. [25].