Experimental analysis and RSM-based optimization of friction stir welding joints made of the alloys AA6101 and C11000

In the present study, the evaluation of FSW input parameters on output response ultimate tensile strength (UTS) of the friction stir welded AA6101-C11000 joint is in agreement. The response surface methodology (RSM) was adapted for generating the mathematical regression equation to predict the UTS and to develop the FSW parameters to attain the highest UTS of the FSW joints. The central composite design (CCD) method from RSM with five levels and three factors, i.e., tool rotational speed, feed rate, and tool offset used to conduct and minimize the number of tests. During FSW, base sheet cu (hard metal) was stationed on the advancing side (+1 mm, +1.68 mm tool offset) and the base sheet Al (soft metal) on the retreating side (−1 mm, −1.68 mm tool offset). The radiography studies were accomplished to inspect the internal flaws of the FSW joints (Al-Cu).The XRD and SEM investigation of the ruptured specimens during the tensile test to evaluate the IMCs phase anatomy and fracture analysis. The maximum UTS value measured during the experimental work was 142.69 MPa at 1000 rpm, 40 mm min−1, and −1.68 mm tool offset. The highest joint efficiency obtained was 82% compared with the AA6101 UTS value. RSM adapted for this work was 92% accurate and satisfactory.


Introduction
Friction stir welding (FSW) is an innovative solid-state process that originated through the welding academy. FSW helps to join the Al-alloys series (7xxx and 2xxx), which is difficult to weld from traditional welding. The FSW performed at the lowest temperature avoids ease of melting the base plates, less protection, and environmental issues, is fault-free, and reveals finer mechanical characteristics [1][2][3][4][5]. In FSW, while welding, the tool slips over the surface of the plates. Due to the stirring and frictional effect develop plastic deformation while impingement within the tool shoulder and plates which develops heat to weld the base plates [6][7][8][9]. Present research in progress of electrical industries requires maximum strength, excellent electrical conductivity, and wear resistance. The metal alloys (aluminium and copper) were chosen for such application, although welding They used radiography to detect the defects from the FSW joints. Suresh Kumar et al, [31], used pure aluminium (1100) of size 100 × 50 × 6 mm 3 with tool material high-chromium high-carbon steel. They found that from the radiographic testing there were no defects in the joints. Chetan Patil et al, [32] used base metals AA7075 and AA6061 of size 150 × 70 × 6.35 mm 3 .At various welding speeds and feed rates, a square trapezoidal pin was made of H13 tool steel. Using the x-ray radiography technique, it was discovered that other FSW welds had void flaws and no penetrations in any of the weld samples. Qasim M Doos and Bashar Abdul Wahab [33], used AA6061-T6 at four welding rates (0.5, 1, 2, 3 mm sec −1 ) and three tool rotational speeds (500, 630, 800 rpm). They found that all the welds were defect-free at maximum tool rotational speed. Suresh [34] used AA5083 of 6.5 mm thickness plate at welding speed from 1000 rpm to 2000 rpm. They found that all the FSW joints were defect-free which has been revealed from radiography testing. Ismail Ibrahim Marhoon et al, [35], used AA5086 plates of 150 × 75 × 3 mm 3 . They found that defect-free from radiography test and which gives high strength of the mechanical properties.

Experiment
2.1. Decision on working materials (Al, Cu sheets, and FSW tool) for executing FSW In this study, the aluminium sheet (series-AA6101) and copper sheet (series-C11000) of identical measures 100 × 50 × 5 mm 3 (length X width X thickness). The properties of these plates were relay on the electrical properties with excellent corrosion resistance. The OHNS (Oil Hardened Non-Shrinkable) steel was hard and durable engages as the main joining tool and was favoured to accomplish the FSW action on the Al and Cu sheets. Influence the precise blending of the base sheets. Base sheets Al have a UTS value of 174.2 MPa and Cu has 248 MPa retrieved from the ultimate tensile strength test machine. Tables 1, and 2 shows the chemical composition of the base sheets and weld tool in percentages. Figures 1(a), and (b) show the sizes of Al, Cu sheets, and OHNS steel weld tools.

Design of experiments
Response Surface Methodology (RSM) is an analytical technique that makes use of a number of mathematical and statistical tools to model the output responses that are obtained and to track the effective response. It helps to investigate the correlation between different input parameters and the influence of these parameters on one or more output response variables. This analysis helps in predicting the different combinations of input parameters and optimizing the output response from the FSW experiments [22]. The RSM model applied was an easy estimation method to notice the best solution. In these experiments, the three input factors (tool rotational speed (rpm), feed rate (mm/min), and tool offset (mm)), and the output response was ultimate tensile strength (UTS) MPa were chosen. In RSM, to develop and perform the FSW of the Al-Cu joint, a central composite design was used. Further, CCD develops a coefficient of quadratic model for the output response variable. In this work, the CCD method has two-level factorial: full factorial design with 20 runs of experiment ,blocks 1 and point type selected was (8 cube points, 6 axial points, 6 cube centre points, and 0 axial centre points),five levels and three factors were chosen was shown in table 3. The design matrix was obtained from Minitab software where tool rotational speed (S) was fixed among 663.64 rpm and 1336.36 rpm, the feed rate (F) was placed among 23.18 mm min −1 ,and 56.82 mm min −1 , and the tool offset (O) varied among −1.68 mm to +1.68 mm was shown in table 4. Finally, the regression equation obtained was predicted and optimized for output response (UTS).
The 20-run welding was conducted corresponding to the design matrix from table 4 generated from the RSM. The base Cu sheet (hard metal) was stationed on the advancing side and the base Al sheets (soft metal) on the retreating side considering the regular blending of the sheets and laminar slide of the metal at the joints. Tool offset faces the required task in the desired sliding of the metals from the Al to the Cu matrix and Cu to Al matrix while joining in which the tool stirs at the joints. When tool offset (0 mm) signifies weld carried at the Al-Cu   figure 2(a). The Al and Cu sheets were joined by a functioning modelled FSW machine shown in figure 2

Evaluating the surface, internal structure, and mechanical features of the FS welded Al-Cu joints
Following the 20-run from the experimental welding, the pictorial review supports finding the weakness in the surface structure extremely at the joints. In this study, the digital radiography system was used to perform a radiography test (Model: XXQ-2505, voltage-250 Kv, radiation source: x-rays, IQI (Image quality indicator)-2A06 type: Wire, focal spot: 1.5 × 1.5 mm 2 , density range 2-4 g cm −3 ). The source of the x-ray started with a voltage of 150 Kv, the exposure time and processing time used for obtaining the image were 0.4 min and 5 min as shown in figure 3.
The wire EDM (Model-DK7735, Molybdenum wire of diameter-0.18 mm) process to cut the mean of two specimens from every trial run to conduct the UTS (ultimate tensile strength) test was shown in figure 4(a). Every specimen set following the system ASTM E8−16a was revealed in figure 4   The anatomic transformations at the Al-Cu joints were revealed from the scanning electron microscope (SEM), Model: TESCAN-VEGA3 LMU. The presence of distinct intermetallic compounds (IMCs) was evaluated using an XRD (generator 30 mA, 40 kV). The percentage of every IMCs generated was confirmed by energy dispersive spectroscopy (EDS).

Results and discussion
3.1. Interpretation of Al-Cu welded surface anatomy Following the welding, the weld surface anatomy was examined. The welding runs of every surface feature were revealed in table 5.
To control the FSW of Al-Cu joints, independent parameters (S) tool rotational speed (rpm), (F) feed rate (mm/min), and (O) tool offset (mm) progress instantly and undertakes the part of clear welding without any flaws. At low speed, due to insufficient frictional heat, as the flashes expand, the shortage of melting between the Al and Cu metal, induces openings at the joint surface. At medium speed 1000 rpm with sufficient friction between the tool and base sheets, enough heat was generated to melt Al and Cu reason for clearing the flaws. At a larger speed of 1336.36 rpm, the heat generation rises, increased tool friction stirring at the joints, excessive melting of Al and Cu sheets induces flaming of the sheets, and surface openings expand with the flashes. Tool offset near the Al sheet side and at the center gives the satisfied welding, flaws-free joint better than the tool offset (Cu sheet side), massive openings [37][38][39]. Table 6 shows the cross-section of the various welding runs. At low tool rotational speed (800 rpm), fewer mixing at the nugget zone or stir zone (SZ) of FSW of Al-Cu joints due to less friction and heat. It gives a medium UTS value of about 118.5 MPa. At medium tool rotational speed (1000 rpm), proper mixing at the SZ due to enough friction and heat developed between the tool and base sheets.Sound dynamic recrystallization at a sufficient temperature, intense plastic deformation at the thermomechanically affected zone (TMAZ), equiaxed refined grains at heat affected zone (HAZ), and non-affected zone (base sheet). It gives a high UTS value of about 142.69 MPa with excellent joint quality. At a very high tool rotational speed (1336.36 rpm), improper mixing at the SZ due to more friction and heat was generated. It gives a low UTS value of about 51.3 MPa. Table 7 shows the outcomes of the x-ray radiography analysis conducted on all trials of different FSW Al-Cu joints generated at tool rotational speed (800-1336.36 rpm), feed rate (30-56.82 mm min −1 ), and tool offset (0, +1, −1, −1.68, +1.68 mm).

Interpretation of internal structure from radiography test
At trial 13, the tool offset towards softer material AA6101, medium tool rotational speed 1000 rpm, 40 mm min −1 , and no internal blemish were examined in the weld area. This is due to its regular fusing of base metals,    abundant heat, sufficient plastic deformation and temperature in the joint area. In trials with low to medium tool rotational speed of (800 to 1000 rpm), constant feed rate of (30 to 50 mm min −1 ), and pin offsets (at weld joints and towards the Al side), the faults found were porosity. Also observed satisfactory mixing of base metals at Al-Cu joints. Trial 5 and trial 9 with low tool rotational speeds (800, 663.64 rpm), cause a lack of fusion due to the tool not being stirred regularly, insufficient heat, poor plastic deformation and temperature, irregular mixing, and high strength portion of Al metals struggle to move towards Cu matrix. Also, observed that some portion appeared as satisfactory mixing at weld joints. Trial 6, trial 8, and trial 10 with very high tool rotational speeds (1200, 1336.36 rpm), feed rate (30 to 40 mm min −1 ), and tool offset towards the Cu side (hard metal), the faults found were burned through. This is because the tool stirs at a high rate, and excessive heat, irregular flow of the metal, and temperature at weld joints. The high-strength portion of copper metal stumbles to move toward the Al matrix. Also, observed that some portion appeared as satisfactory mixing at weld joints. Hence it is clearly observed that at medium tool rotational speed, medium feed rate and tool offset towards the softer metal side will have the internal fault free from the x-ray radiography analysis and increases the load-bearing ability of the FSW Al-Cu joints.

Evaluation of ultimate tensile strength (UTS) from the experimental work
The figures 5-11 present the stress versus strain curve and load versus displacement curve for low, medium, and high tool rotational speed (up to 1336.36 rpm) with different feed rates and tool offset. From the experimental work, the ruptured specimens from the UTS test and interaction of UTS with % Elongation were shown in figures 12(a), (b). Toward medium tool rotational speed 1000 rpm, 40 mm min −1 feed rate, and −1.68 mm tool offset (at Al side), has a significant effect on UTS value about 142.69 MPa and 3% elongation. Sufficient frictional heat generated, plastic deformation at the joints converting to the pure blending of Al-Cu materials. Lamellar movement of the materials and the development of thin grains evenly spread at the joints were the cause of the highest UTS value. Toward high tool rotational speed 1336.36 rpm, 40 mm min −1 , and 0 mm tool offset (At center), has a very low effect on UTS value about 51.3 MPa and 0.6% elongation. The stirring action of the tool engages a high frictional effect leading to the turbulent slide of the materials, randomly distributed thick grains formed at the joints were the basis of decrement in UTS values. With an increase in tool rotational speed and the tool offset at the center and AA6101 side, the UTS value increases. At higher speeds, the tool offset towards the C11000 side, and the UTS value decreases [40][41][42][43].
In this experiment, the apparent intermetallic compounds (IMCs) were generated through the diffusion of the Al and Cu materials, while friction resulted from the tool's contact with the base materials. The generation of these IMCs confides on tool rotational speed (rpm). As the speed expands, the IMCs with an extensive unit of massive thickness develop, influencing the Al-Cu joints. Figures 13(a)-(c) illustrates the XRD investigation of the ruptured specimens during the tensile test to evaluate the IMCs phase anatomy. At low UTS value of 51.3 MPa (1336.36 rpm, 40 mm min −1 , 0 mm tool offset), the IMCs initiated were Al 4 Cu 9 and Al 3 Cu 2 will implement a weak Al-Cu metallic compaction prime to weak strength at joints. At an average UTS value of 118.5 MPa (800 rpm, 30 mm min −1 , −1 mm tool offset), the IMCs initiated were Al 4 Cu 9 and Al 1 Cu 3. At a high UTS value of 142.69 MPa (1000 rpm, 40 mm min −1 , −1.68 mm tool offset), the IMCs initiated were Al 4 Cu 9 and Al 1 Cu 3 .At these input factors range will perform solid Al-Cu Metallic compaction prime to supreme joint Where UTS-Ultimate tensile strength. S-Tool rotational speed, F-Feed rate, O-Tool offset. The regression coefficient values obtained from the regression analysis with interactions of different factors were displayed in table 8, where 'T' defines the ratio of coefficient and standard error coefficients, and their values with predicted input factors. The P-value indicates the significance of the coefficients of each input parameter. If P-value is lower or equal to 0.05, it is normally regarded as statistically significant (P > 0.05, nonsignificant).In this work, the maximum P-value (less than 0.05) obtained for each coefficient of parameters was statistically significant. P-values larger than 0.05 are present for the input factor combinations of feed rate and feed rate (F X F) in table 8. As a result, equation (1) will no longer contain the (F 2 ) term. The confirmation test took into account the shortened equation (2).    Another measure widely used to show the indecency of a fitted regression mathematical model was the regression coefficient of measurement (R 2 ), which should be nearly unity for a quality model and obtained was 99.48%. The R 2 adjusted obtained for this model was 99.01% showing the significance of the regression model obtained.R 2 predicted obtained was 95.80% which indicates the experimental values were suitable with the predicted values. Figure 14(a) shows the normal probability plot, the experimental values (blue colour dots) of UTS obtained were sliding along the diagonal line (red colour) designated the residuals were shared normally for the generated regression model. Figure 14(b) shows the match type of predicted and experimental values. Nearly all the entire experimental values (blue colour dots) were distributed to the horizontal line (dotted line) that designates the residuals was a perfect match for the generated regression model.

The model confirmation
The model verification was done by randomly selecting trials 3 and 6 with (R), (F), and (O) combinations from the design matrix in order to get validation of the experimental results.
Input factor values were replaced in the shortened equation (2). Table 9 displays the results of the confirmation test. The experimental values and the UTS predicted values from this table were reasonably close. The RSM adapted for this study has a 92% accuracy rate.

Effect of independent factors on UTS for optimization
The 3D response surface graphs and 2D contour graphs were obtained from the Design expert software. Figures 15(a)-(c) show the 3D response surface graph and figures 16(a)-(c) shows the 2D contour plots for the output response UTS values. Both the figure demonstrates the relationship outcome of two random independent factors (X1 and X2 axis) and another output factor in the middle. The UTS value of the FSW Al-cu joints increases from low to medium tool rotational speed and starts dropping at higher rpm. At low-medium tool rotational speed, the fine-sized refinement of grains was caused due to sufficient friction, and laminar plastic movement of materials. When expanding at the tool rotational speed, the coarsening size of grains appears due to insufficient friction, and turbulence movement of soft materials at the Al-Cu joints. The 2D contour model ( figure 16) shows tool rotational speed versus feed rate, in this case, the UTS value starts increasing from low to medium (speed and feed rate). At higher rpm, UTS values start decreasing at higher speeds and the feed rate. The model shows tool rotational speed versus tool offset, in this case, the UTS value increases from low to medium speed with tool offset (−1 mm) at the softer material and center (0 mm) of the base materials. Clear progress of Al material approaches the Cu matrix. Later UTS value starts dropping when the tool offset (1 mm) toward the harder material and high rpm. The hard Cu materials become impotent to join      figure 17(c). The IMCs generated were Al 4 Cu 9 , Al 1 Cu 3, and Al 3 Cu 2, which affect the UTS value of the specimens. The size of the IMCs increases from thin to thicker when the tool rotational speed increases from low to high. At low speed, the IMCs brittleness was less, leading to the specimens having sufficient ductility. The UTS values expand and suddenly drop at a higher speed. The brittleness of the IMCs expands at the weld joint area leading to moderate ductile specimens [44,45]. The EDS investigation shows the percentage of the IMCs developed from figures 18(a)-(c), and table 10. Typically, the maximum Al 4 Cu 9 IMCs phase developed in all the specimens. In some specimens, the IMCs Al 1 Cu 3 were seen (trials 800 rpm, 30 mm min −1 , −1 mm and 1000 rpm, 40 mm min −1 ,−1.68 mm).The IMCs Al 3 Cu 2 were observed in the low UTS value ( 1336.36 rpm, 40 mm min −1 , 0 mm).

Conclusion
In this research, the RSM method was used to generate the mathematical regression equation, and to compare the experimental and predicted values of output response ultimate tensile strength (UTS). The central composite  design used from RSM consists of three factors tool rotational speed (S), feed rate (F), and tool offset (O) to conduct the experimental work for friction stir joining the AA6101 and C11000 plates. The following results obtained are as follows: 1. The maximum UTS value measured during the experimental work was 142.69 MPa at tool rotational speed 1000 rpm, feed rate 40 mm min −1 , and −1.68 mm tool offset. The UTS values at joints rises as tool rotational speed rises, and at high speed, the UTS value falls. The highest Al-Cu joint efficiency obtained was 82% compared with the AA6101 UTS value. The RSM adapted for this work was 92% accurate and satisfactory.
3. From the SEM study, at standard tool rotational speed of 1000 rpm, feed rate of 40 mm min −1 , and −1.68 mm tool offset, the fracture analysis was ductile-brittle mode. At overhead tool rotational speed 1336.36 rpm, feed rate 40 mm min −1 , and 0 mm tool offset, the brittleness fracture of the specimen was found.

Data availability statement
All data that support the findings of this study are included within the article (and any supplementary files).