Influence of scandium interlayer on the mechanical and metallurgical characteristics of friction stir welded AA1200-H14/ Sc/AA6061-T6

The present investigation observed the friction stir welding (FSW) of joints made from the interfacial layered dissimilar AA1200-H14 and AA6061-T6 using scandium about the UTS, frictional heat, and metallurgical properties. The experimental examination is carried out utilising a three-level, three factors, Box-Behnken Design matrix (17 tests) in response surface approach, with the welding speed (WS), axial force (AF), and tool rotation speed (TRS) as the stated input parameters. Analysis of variance (ANOVA) was used to evaluate the model’s dependability. Macroscopic and microscopic joint defects, as well as the alloy intermixture, have been found using optical microscopy and SEM. The SEM microstructural study exhibited that the generated grains are fine and equiaxed. The optimal WS (50 mm min−1), A−1F−1 (10 kN) and TRS (1750 rpm) settings produced the maximum terminal tensile strength (200.182 MPa) and perfect frictional heat observations. The inclusion of scandium interlayer as well as optimised parameters improved the joint’s mechanical characteristics and produced a fine-grained microstructure.


Introduction
The emergence of hybrid structures and intricate patterns employing a range of temperatures has greatly increased the appeal of welding aluminium and its alloys.On account of considerable variations in their physical and mechanical properties, mainly their flash points, some aluminium alloys (AA) are challenging to classic fusion weld.In addition, less ductility and strength are achieved when joining AA1200-H14 to AA6061-T6 because brittle intermetallic compounds (IMC) are deposited in the nugget zone.These restrictions encourage the use of FSW, a technique for solid-state (SS) welding [1], to combine the incompatible alloys AA1200-H14 and AA6061-T6.FSW is an approach of a SS joining that is carried out less than the solidus condition of the materials to be coupled [2].FSW joints were defined by superior mechanical strength, excellent metallurgical characteristics and reliability.The TRS, transverse speeds, tool depth, tilt angle, and design or material are the primary FSW process characteristics that determine both the weld quality and process performance.Additionally, interlayers, such as zinc foil and silver [3], are utilized when joining similar and dissimilar alloys.
When AA1200-H14 and AA6061-T6 are welded together, the extra material might function as an interlayer, reducing the number of IMCs produced.The interlayer material should be chosen that promotes the growth of a ductile phase as opposed to brittle aluminium-magnesium intermetallics phases [3].It has been shown that adding rare earth metals to Al alloys has the desired effects [4][5][6].In this experiment, AA1200-H14 to AA6061-T6 is FSWed using Scandium (Sc) as an interlayer.Sc is supposed to make alloying elements more ductile at high temperatures.The aluminium scandium's solid solubility changes greatly with temperature [7].It also deforms at a high strain rate despite having a higher diffusivity.The melting temperature of Sc is substantially higher at 1541 °C than that of aluminium, which is 660 °C, which is another difference.It is expected that Sc will disperse evenly throughout the nugget zone at welding temperature, reducing the number of IMCs that develop.Sc is one of the greatest features for strengthening Al when utilized in AA [8].Al-Sc alloys have good mechanical characteristics because the Al 3 Sc particles are hard and tightly ordered at both low and high temperatures [9][10][11].Al 3 Sc precipitates are stable in the Al matrix at high temperatures [12].It has been discovered that Al 3 Sc nanoparticles form in response to ageing and thermo-mechanical phenomena [13,14].The Al-Sc alloys' metallurgical and general mechanical characteristics are pretended by frictional heat and mechanical treatments [15][16][17].A variety of industries use heat-treatable (HT) as well as non-heat-treatable (NHT) alloys of aluminium with Sc addition.The aerospace and automotive industries regularly employ the 1200-H14 and 6000-series Al alloys for structural elements [18][19][20][21].Furthermore, table 1 depicts a comparison has been made between the current study and the results of the recent review articles.
According to the literature review, there aren't many studies on FSW on dissimilar aluminium alloys using Sc as the interlayer.Furthermore, very few researches were carried out on the Box-Behnken Design matrix in response surface technique to optimize the welding constraint about the frictional heat generated throughout the FSW.In this work, two dissimilar alloys AA1200-H14 and AA6061-T6 are FS welded together using Sc as the interfacial layer, and the frictional heat provoked in the time of the welding activity is compared to TRS, WS, and axial stress using optimization technique.Through the use of the ANOVA technique, the process parameters are optimized.The clarity of grain deformation during the process is also obtained through microstructural studies and fractography.

Materials and methods
Economically obtainable 5 mm thick sheets of the work-hardened AA1200 alloy were revealed on the advancing side (AS) and the precipitation-hardened AA6061 alloy was positioned on the retreating side (RS).
AA1200-H14 and AA6061-T6 were sandwiched together by a 2000 μm thick Sc strip [22].The mechanical attributes of parent metals and the amount of the principal alloying elements in the combination are depicted in tables 2 and 3, appropriately.The specified amount of the initial components were processed at 650 °C-750 °C in a sealed induced dissolved furnace to produce a moulded Al/Mg/Sc blend along with a theoretical fabric similar to AA with a 2% weight per volume Sc addition as shown in figure 1 (Metmech Laboratory, Chennai) [22,23].An FSW experiment has been conducted at the central workshop laboratory-in Chennai, utilizing a 3-axis machine.The machine has the following specifications: a spindle speed of 3000 rpm, a load capacity of 300 kN a and motor 1.5 hp [24].The fixture provided a solid hold on the specimens, allowing for accurate installation and a strong clench of the work materials.This experiment illustrates the usage of a clamping technique in the figures 2(a)-(c).Depending on the shoulder pressure, the FSW zone is enlarged at the top exterior of the welded section by the circling rotation, as shown in the figure 2(c).The FSW research was carried out with different process parameters given in table 4. A universal testing device LDW-50, Blue star was used for tensile trials to measure the UTS as well as % of prolongation.The microstructural analysis was done using an EVO 18; ZEISS scanning electron microscope and an optical microscope from Wi-Tec called the alpha 300 R. The ASTM standards were followed for investigating mechanical and metallurgical computations.Vickers micro hardness tester (Wilson  Wolpert, Germany) is used to evaluate the stiffness of welded sample in various zones.A digital infrared thermometer (MT 4 Industrial IR Thermometer) was used to detect the frictional heat dissipated during FSW.

Box-behnken matrix-response surface methodology (RSM)
Design Expert 11 is likely the software used to implement and analyze the experimental design.It's a powerful tool commonly used for designing and analyzing experiments, particularly in the fields of engineering and physical sciences.The most essential aspects that influence the strength of the weld joint were thought to be the axial force (AF), welding travel speed (WTS), and tool rotational speed (TRS).The BBD was utilized to determine the correlation between these factors.The Box-Behnken method frequently employs response surface designs, particularly when the F-statistic is calculated when three levels and three factors are present.Using the Box-Behnken design, the current study examines the effects of three independent variables AF, WTS, and TRS on the response tensile strength.Also established were the appropriate design parameters for the strength of FSWed.The welded tensile samples are shown in figure 3.
The weld joints were built to accommodate different input parameter combinations.Utilizing the Box-Behnken design, the current study examines how the axial force welding travel speed, and tool rotation speed affect the response tensile strength.The best design criteria for MMCs that were friction stir welded were also established.Each cross-weld specimen has been described in accordance with the design matrix.
The tensile strength and frictional heat as the output parameters were mathematically correlated with these responses as the input values.It has been used to optimise the process criteria for producing the in-house created MMC and acts as a purpose function for the desirability method.The accuracy of the developed model in terms of facts was confirmed by the verification test and compared with the expected outcomes.An objective function is necessary for the FSW process parameter optimization.Design of experiments was used to expand a correlation between process criteria and a minimal amount of welding investigations.The key process parameters for FSW were identified, along with their ranges, using the Box-Behnken model [22].
These components were smooth and etched with a Keller's reagent.High-contrast pictures may be obtained by making slight modifications to the substance, etching or exposure length, temperature, and difference among  the AA1200-H14 and AA 6061-T6 elements.The 1200 aluminium was frequently the majority responsive material to such an etch.The AA6061 may also be favored for etching as a result of large changes in the concentration of the HF component.The heat dissipated during the FSW process is measured using the thermometer, which is shown in figure 4.
In order to identify the base materials' grain structures and the dissimilar welded joints, optical microscopy and SEM were utilized to study the welded specimens.The longitudinal sections of the tensile samples were measured transverse to the path of the weld by the ASTM E8 criterion.The UTS were accomplished on test samples with fascinating surfaces of 30 mm and decreasing estimate lengths of 25 mm utilizing a pull rate of 10 −3 s −1 .

Development of regression equations
The FSW process was executed using the provided parameters, and the welded joints were used to fabricate cross-weld test specimens.Weld joint tensile strength from each specimen was tested to establish the mathematical correlations.A, B, C, AC, A 2 , B, and C 2 are essential model conditions in this instance.If the rate is higher than 0.1000, model terms are not considerable.If the model has a lot of extraneous terms, the model reduction might get better it (excluding those necessary to support hierarchy).The influence of each input parameter on the expected and experimental findings was verified using ANOVA (table 5).
The resulting model's F-values for the UTS and frictional heat were 64.59 and 623.47, respectively, while the F-values for the lack of fit were 84.36 and 2.21 (non-significant).A 95% confidence interval was used to evaluate the values' significance.The R-Square (R-Sq) and adjusted R-Square (adj.R-Square) values help to identify these factors' importance.R-Sq, adj.R-Sq, and predicted R-Sq all had computed values of 0.9881, 0.9728, and 0.8603, respectively.These figures highlight the consequence of the created observed relationship; the anticipated R-Sq reveals an 86% disparity in the empirical relationship.Both the expected and actual UTS and the predicted versus real frictional heat levels of the FSW specimen are plotted in the perturbation plots in figures (a) and (b), respectively.The findings show that there is little deviation from the straight-line slope of the plot, which indicates a low error proportion, between the actual and anticipated values.
Weld nugget hardness and UTS from each specimen were experienced to establish the mathematical correlations.A, B, C, AC, A 2 , B, and C 2 are essential model terms in this occurrence.(Here A-TRS, B-WS and C-Axial force).Utilize the equation written in terms of coded aspects; it is feasible to forecast the feedback for particular levels of each ingredient.The high values of the factors are by default recorded as +1 and the low levels as −1.The coded equation can be used to evaluate the aspects co-efficient and establish the aspect's comparative weight.The Probability of residual with a standard distribution of UTS and Temperature are shown in figures 5(a) and (b), respectively.The developed final regression models in coded form to predict tensile strength and hardness of FSW composite joints are given in equations (1) and (2).
The model is started with a quadratic model and the interaction terms and significant responses were found by analysis of variance (ANOVA).The frictional heat and UTS values for the welded specimens are shown in tables 6 and 7 together with the actual and expected values.The dissimilarity between the genuine and expected values of the output answers is represented by the fraction of error.The actual, predicted and error % of the UTS with temperature is shown in table 8.The precision of the created response model interaction graph for UTS and temperature is depicted in figures 6(a) and (b), respectively.
Figure 6 the predicted versus actual plot allows for a more accurate assessment of the model's suitability.The normal distribution of output parameters is indicated by the straight line that connects the data points in a  normal probability diagram.Additionally, the plot of the projected versus real data similarly follows a straight line, suggesting that this model can be utilized to make reliable predictions.

Consequence of input constraints on UTS
Figure 7 There is a curvature impact for each input parameter.The result did not change linearly when the input was changed from the smallest limit to the maximum bound.The weld characteristics are more sensitive to a higher curvature.The graph displays the response changes if the reference points differ from the chosen reference point.A factor is sensitive when its slope or curve is steep.The perturbation graph shown in figure 7(a) depicts the effect of the input parametric parameters on the out-turn UTS of the FS welded joints.The perturbation technique employed in this work is based on the desirability approach.The input welding constraints are denoted by plots A, B and C. The curvature of the parabolic curves reveals the amount to which they influence the UTS of the FS welded samples.The parabolic curves A, B and C indicate TRS, WS, and load, respectively.The curve A shows the greater curvature followed by B and C. TRS was the most significant characteristic, followed by WS and load.A mixed flow zones are created due to TRS in the weld stir zone also it generates frictional heat during welding to plasticize the base materials.Also, TRS emerges to be the most necessary process variable because it influences the transfer speed.During stirring, the nugget zone received the most heat and plastic deformation.It caused recrystallization, resulting in a reduction in grain size which improved the UTS.Increasing the WS results in a minor improvement in UTS but a drop in ductility.This could be because of the substantially larger heat input associated with the slowest WS, which leads to less strain hardening behaviour, more recovery, and grain coarsening, all of which improve ductility.The applied axial load is influenced by the flow of materials within the weld area by an extrusion process in which the material has undergone plastic deformation.The impact of the input parametric parameters on the out-turn parameter UTS is shown by the 3D surface plots in figures 8(a)-(c).The contour plots in the figure 8(a) shows the growing TRS and WTS together with the rising UTS trend.The UTS steadily increased when the parameters were increased, culminating at 50 mm min −1 WTS and 1750 rpm TRS, and then gradually declining after that (figure 8(a)).Axial force and TRS both had an effect on the UTS of the FSW, as seen in figure 8(b).When the tool is rotating at 1750 rpm and the load is 10 kN, the specimen's UTS steadily increases with corresponding parametric increases to achieve its maximum value (figure 8(b)).The highest UTS (200.182MPa) was found to be the result of better material joining at the weld interface as an outcome of these parametric values.Axial force and WTS had a parametric impact on the UTS (figure 8

Consequence of input parameters on frictional heat
The contact of the input parametric parameters on the out-turn UTS of the FS welded joints is depicted by the perturbation graph in figure 7(b).The desirability approach serves as the foundation for the perturbation technique used in this work.The input welding parameters are denoted by plots A, B and C. It can be identified that curve A (TRS) has a steep curve, followed by curves B (WS) and C (axial force).
The frictional heat levels of the welded sample are most influenced by TRS.The frictional impacts of the rotating tool on the work material FSW cause heat to be produced.As the TRS increases the temperature in the work material also increases.This is due to increased frictional consequences of the rotating tool on the work material as TRS increases, which causes more heat to be generated and, as an outcome, raises the temperature of the weld.For the superior mechanical qualities of the FSW welds, extremely high frictional heats are not advised.Increasing heat generation may cause precipitates at the weld zone to become coarser or dissolve, which grades in fragile bonding at the WZ.Also, high WS may cause excessively high temperatures, which could lead to flaws and poor mechanical qualities.The flow of materials through the weld region is affected by the applied axial load as a consequence of the plastic deformity of the material during the extrusion process.
Figures 9(a)-(c) three-dimensional surface plots show how the input parametric parameters affect the FSW joint's friction heat output parameter.Increasing TRS, axial force, and WTS showed improvements in the weld joint's friction heat levels.The TRS, axial force, and frictional heat were at their highest, and WTS were 1750 rpm, 10 kN, and 50 mm min −1 , respectively.It was discovered that the TRS and WTS had an impact on the FSW specimen's frictional heat by altering the material flow in the weld region, precipitate coarsening or dissolution, and redistribution of grains.Finally, the bar chart in figure 10 shows that the factors TRS, WS, and axial force all had desirability values equal to 1, demonstrating that they all met the criterion.
The desirability approach is an effective method that determines factor settings that optimize a 'score' that is assigned to a group of responses.One of the approaches for multiple response process optimizations that are most frequently utilized in industry is the desire function technique.In a desirability approach, where parameters have attained the value of 1 without any deviation, it suggests that all the factors or criteria being considered are at their optimal levels or ideal states.Each parameter is meeting the desired conditions or specifications without any variation, and this is reflected in the desirability score of 1.

Microstructural observation
The visual microscopic descriptions of the weld cross-section of SZ, merging of two dissimilar metals with Sc interlayer are shown in figure 11.The aluminium alloy plates are subjected to heat and severe plastic deformity during FSW, and process parameters control the temperature increase and deformation of the alloys.The dissimilar AA1200 and AA6061 welds undergo plastic deformation, which improves material flow and refines the grain structure as well as creates a solid-state link between the sheets.The materials' grains change shape and position during plastic deformation to accommodate the mechanical stress in the weld zone [25].To achieve a solid-state bond and make continuous welds, the material needs to undergo plastic deformation.On addition of Sc interlayer to the welds illustrates the spatial circulation of Sc in the WN of different alloys figures 11(a) and (b).These Sc fragments are also unfairly distributed throughout the WN.Because of the dispersion in the material stream, the stack of fragments in the WN can be destructive, producing holes and tunnels to appear [26,27].The grain size is found to be finer in the WN than in TMAZ and HAZ regions.Equiaxial grains and homogeneous distribution of grains have formed in the weld zone due to the existence of Sc from the interlayer during the FSW process [28].
Figures 11(c) and (d) display cross-sectional SEM descriptions of the interlayer of the Al 1200-H14 /Sc/ Al 6061-T6 joint adjacent to the surface.The SEM microstructural study showed that the generated grains are fine and equiaxed.This is because friction and stirring among the tool and the parent metals originate tremendous heat, which causes severe plastic deformation and dynamic plastic deformation (DRX).Grain refining is aided by DRX, which is typically found in WN [29].The thermally mechanically affected zone (TMAZ), being thin and has an extended grain structure at the boundary between WN and HAZ and is situated closer to WN than HAZ, produces more heat than HAZ.Grain dislocation brought on by plastic deformation produces elongated grains.In comparison to the deformed grains, the recrystallized grains have lower dislocation densities and better crystallographic orientations, which aids in releasing the stress and energy that were trapped during the plastic deformation phase of the welding process.This reduces residual stresses, improves the weld's mechanical properties, and encourages a homogeneous microstructure [30][31][32][33].
Since Sc typically solidifies into solutions with various Al alloys at the weld temperature, this phase is recognized as unique.When compared to Sc particles found in the NZ, it can be seen that the restrictions of the Sc particles near the surface are more spread.Sc accumulation demonstrated a further decrease in grain size in weld nuggets, indicating the capacity of Al-Sc intermetallics to do so [34].The optimal values of input parameters including TRS, WS, and axial force are also in charge of the FSW process's microstructural change [35][36][37].The highest UTS sample shattered near the RS HAZ, as seen in figures 11(e) and (f).For instance, AA1200-H14/Sc/AA6061-T6 exhibits a fully ductile phase with equi-axed dimple patterns predominating across the fracture surface.Fine grains generally imply that the crystal grains in the material are small in size.The grain size in the weld zone is influenced by FSW parameters such as TRS, WS, and tool geometry.Increasing the rotational speed tends to result in finer grains.This is because higher rotational speeds lead to increased heat input and more dynamic recrystallization, promoting the formation of smaller grains.Lower traverse speeds can also contribute to finer grain sizes.Slower traverse speeds provide more time for recrystallization processes to occur during FSW.Material flow during FSW is affected by TRS, WS, and tool design.Both TRS and WS influence the strain rate during FSW.Higher strain rates can result in increased dislocation density along grain boundaries.

Microhardness
The hardness of the base metal was around 43Hv for AA1200 and 110Hv for AA6061, respectively.Although joints have been found to have the highest hardness values, which are around 80 HV. Figure 12 indicate the optimized range of low, high and confirmation test hardness plot.This is due to the development of an extremely tiny equiaxed grain at the WN of the junction prepared by the high frictional heat produced at ideal TRS [38].The layering of 1200-H14 and 6061-T6 alloys has caused the hardness of the stir zone to vary as well.Because 1200-H14 and 6061-T6 occur often in the SZ, there is a variation in hardness in the weld produced by frictional heat.The greatest hardness value shows that 6061 predominated in the nugget, which is indentation at permanent distances.Sc inserts increased the hardness of the nugget area.This is caused by the occurrence of intermetallic Al 3 Sc in the WN as well as inexplicable Sc elements in the matrix of aluminium.Due to differences in microstructure, the TMAZ and HAZ displayed lesser hardness [39][40][41].
This can have several positive effects on material properties.For example, materials with fine grains often exhibit improved mechanical properties, such as higher strength and hardness.The grain size and distribution in the weld zone can impact the mechanical properties of the joint.Finer grains often lead to higher joint strength and improved mechanical performance [42].

Conclusion
This research looked at a scandium interlayer dissimilar aluminium alloy FSW joint using parametric optimization using the Box-Behnken Design in Response Surface Methodology.Researchers looked into how Sc affected the FSW joint.In order to determine which input process parameter was the most important, the effects of the input welding conditions were assessed.The investigation's findings are: 1.The results showed a peak UTS of 200.182MPa and frictional heat of underneath 10kN of axial force, 50 mm min −1 WTS, and 1750 rpm of TRS.
2. It was discovered that the most crucial input parameter, the TRS, increased both the UTS of the FSW sample and the frictional heat in the nugget area.The affecting parametric values were displayed in perturbation graphs and contour plots.
3. The optimized scandium interlayered FSW joint's microstructural analysis revealed superior mechanical properties and smaller grain sizes, which in turn increased the weld's UTS.

Figure 2 .
Figure 2. (a) Friction stir welding machine (b) FSW Process (c) Welded area of the plate (d) 17 trails and confirmation welded samples.

Figure 5 .
Figure 5. Probability of residuals with a standard distribution (a) UTS, (b) Temp.
(c)), which established increasing examination trends that matched those in the figures 8(a), (b).

Figure 6 .
Figure 6.Precision of the created response model (a) Interaction graph for UTS, (b) Temp.

Figure 7 .
Figure 7. Comparability between the experiments and desired results for (a) UTS and (b) Temp.

Figure 8 .
Figure 8. 3D surface plots illustrates the impact of various welding conditions for UTS.

Figure 9 .
Figure 9. 3D surface plots illustrate the impact of various welding conditions for Temp.

Table 1 .
Summary of the different review articles.

Table 4 .
Welding framework and levels.

Table 5 .
Experimental design matrix and results.

Table 6 .
Tensile strength data from an ANOVA.

Table 7 .
Temperature data from an ANOVA.

Table 8 .
Actual, predicted and error % of the UTS with temperature.