Effect of friction-welding parameters on the tensile strength of AA6063 with dissimilar joints

A BSTRACT . In this paper, the effect of welding parameters of rotary friction welding between AA6063 and AISI4130 and AA6063 and Copper are investigated. The major influencing parameters considered are upset pressure, friction time and friction pressure of friction welding are considered for this study. The Taguchi’s design of experiments was conducted for the influencing parameters and their levels. The tensile test experimentation was carried out and the results of the AA6063 and AISI4130 and AA6063 and Copper are compared. The ultimate tensile strength of AA6063-AISI4130 joint and AA6063-Copper joint was improved by increasing upset pressure up to 97MPa with FP of 71 MPa and FT of 4 sec. On the side of AA6063, intermetallic compounds have formed, as seen in SEM micrographs. Microcracks are forming on the side of AA6063 and propagates along the grain boundaries. The effect of the influencing parameters on the tensile strength of the dissimilar joints are studied using the Taguchi’s DOE and ANOVA. From the outcomes it is observed that the friction pressure influence more on the strength of the AA6063 dissimilar joints. highest tensile strength. The increased strength of AA6063-AISI4130 steel is due to the higher strength of the steel than the copper. For all cases of Taguchi’s DOE, AA6063-AISI4130 steel provides the highest strength. The obtained results of the experimentations are the input functions for the Taguchi’s analysis. Commercially available statistical tool is used to analyze the obtained result for AA6063-AISI4130 and AA6063-Cu dissimilar welding joints. The quality attributes were determined by transferring the results into the signal to noise(S/N) ratio. The influence of the process parameters like UP, FP and FT on the tensile strength of the AA6063-AISI4130 and AA6063-Cu dissimilar welding joints are analyzed using the signal to noise ratio.


INTRODUCTION
ne of the most effective method to join the dissimilar metals is rotary friction welding (RFW) which is economic and provides safety to operator and does not emit the IR rays, fume and smoke during the welding [1][2]. The application of RFW, in many industries such as aircraft structures and aero engine components, is required to join dissimilar metals which is inevitable now a days [3][4]. Similar to it, linear friction welding (LFW) and inertia friction welding (IFW) were also used to join the dissimilar metals which gain the more interest now a days [5][6]. In welding of aluminium-stainless steel dissimilar metals using inertia friction welding, the intermetallic compound layer generated at the interface of joining which was seen to be increases with the increment in the rotational speed [7]. These innovative welding techniques to join the dissimilar metals plays the important role in the automobile industry where it requires light weight structures with high strength to weight ratio for automobiles. Thus, the friction welding method becomes the O highly efficient [8] and promising technology to weld the dissimilar metals with narrow heat-affected zone (HAZ) [9]. In friction welding, among the two dissimilar metal rods, one of them is held in the chuck and rotating and another rod is fixed in the pneumatic holder. Th axial load applied by the pneumatic holder generates the frictional heat which necessary to weld both rods together. The friction welding (FW) machine generates the required heat to weld the two dissimilar rods through their surface interaction while rotating. This welding is possible due to the intermolecular diffusion formed between the faces of two dissimilar rods. In the friction welding, the heat generated at the interface is lower than the melting temperature of base metals [10]. Thus, the melting of metals will not happen in it. To get the strong joint, the factors like upset pressure (UP), friction time (FT) and friction pressure (FP) influence significantly [11]. However, when the friction and upset pressure increases the heat generated and extruded metal heat also increases. In case of FW of aluminum and steel, the probability of generation of thin intermetallic compounds, which were brittle in nature [12][13]. The intermetallics, such as FeAl and Fe 2 Al 5 , were formed in the process of FW of the AA6061-AISI1018 steel [3]. The increase in friction time during friction welding while joining of AA6063 with SS304 [14], at low pressure, leads to increment in the temperature which softens the weld. Thus, the increment in the temperature, due to higher pressure and FT, and cooling rate causes the inducement of residual stress in the welded joints. The addition of the nickel interlayer enhances the tensile strength [15] and addition of silver causes the reduction of magnesium in aluminum and reduces the width of intermetallic layer and enhances the tensile strength [16]. While joining of AA5052 and AA6063 with the friction welding, Adrian Lis et al. [17] observed that lower friction pressure causes the reduction in the thickness of the soft HAZ in turn increases the tensile strength of the welded joint. Ajith et al. [18] studied the FW of the S32205 duplex stainless steel and concluded the upset and FP were the majorly influencing process parameters. Many authors [19][20][21][22] used Taguchi's method of optimization and design of experiments (DOE) [23] to study the effect of process parameters. Literatures shows the dissimilar metals like aluminum-steel, steel-copper alloys, carbon-stainless steel, carbon steelaluminium are welded using the friction welding [24][25][26][27][28]. The joint's tensile strength will be affected by changes in pressure and time because the base metals' have different coefficients of thermal expansion. The objective of present work is to find the influence of the process parameters on the tensile strength of AA6063-AISI4130 and AA6063-Cu dissimilar welding joints. In this work, it is intended to use RFW to combine AA6063 with AISI4130 steel and AA6063 with copper specimens. The main purpose of this work is to identify the RFW parameters to weld the dissimilar metals, like, steel and copper with aluminum to reduce the weight of the machine components. Some of the applications where RFW of aluminum with steel are Axle shafts, driveshafts, gears to shafts, hydraulic cylinder rod end, pipe to pipe segments etc. This study is for determining the influence of FW parameters utilized for welding of dissimilar metals on their tensile strength. The Taguchi's DOE and ANOVA has been utilized to analyze the experimental outcomes. The confirmation experiment will also be carried out to validate the experimental results.

MATERIALS AND PREPARATION
he materials chosen for this work are aluminum 6063 alloy, AISI4130 low alloy steel and copper in industrial and domestic applications. The chemical composition of AA6063, AISI4130 are listed in Tab. 1(a-b). AISI4130 steels have enhanced properties like corrosion resistance, ductile, moderate hardness, higher strength and flexibility and able to withstand shock loads. It is having the low carbon content up to 0.03%. Its major alloying elements T are Cr, Mn and Mo. Aluminum 6063 alloy have silicon, iron and magnesium as major alloying elements. Copper is the soft, ductile and malleable metal and its melting point is 1084 o C. Among many excellent properties of copper, electrical conductivity, thermal conductivity are the properties which makes the copper to be used in electrical parts, due to the high corrosion resistance of copper, it can also be find applications in marine, gas and domestic plumbing industries. Some of the material properties of copper are listed in Tab. 1(c).

Welding parameters selection
etailed study on rotary friction welding shows the possibility of effect of its process parameters on the joint and its strength. Parameters influencing tensile strength of RFW joint are UP, FP, FT and spindle speed (SS). In this work, the keeping the spindle speed (SS) constant at 1500 rpm, the other parameters such as FP, FT and UP are varied which influences majorly on tensile strength of welded joint. With the increase of process parameters, to analyze the data, large number of samples must be experimented. Taguchi's design of experiment (DOE) will provide the special design of orthogonal arrays which gives the minimum sets of experiments to analyze the data effectively. Taguchi Method is a statistical method for improving quality & is based on three loss function viz., Larger is better, Smaller is better and Nominal is better and hitting the target with minimum variation. The Taguchi's DOE is carried out using the commercially available statistical tool for the process parameters FP, FT and UP with the three levels as mentioned in the Tab    The workpiece (AA6063, AISI4130 and Copper) dimensions considered are 25mm diameter and length 100mm. The surfaces of the workpieces are prepared well so that welding surface will be free from imperfections. In this welding, joining of dissimilar metals requires influence of pressure and relative motion of the two workpieces.
In comparison with steel and copper, degree of deformation is larger for aluminum. Thus, the weld flash on AA6063 is larger (shown in Fig.2). Due to higher thermal conductivity of AA6063, it is cooled fast then the steel. The Fig. 2 shows the welded samples of AA6063-AISI4130 and AA6063-Cu using the RFW.   The tensile testing experimentation has been carried out for the specimen of dimensions: 6mm gauge diameter, 50mm gauge length as prescribed by the ASTM standard in a computerized universal testing machine. The tensile tests have been carried out for AA6063-AISI4130 and AA6063-Cu welded specimens. The results of the experiment such as load and displacement are recorded and analyzed to evaluate the tensile strength. The three specimens were tested for each cases and average values have been listed in the Tab. 4. The standard deviations are ranging from 0.5 to 2.6.  In comparison with the AA6063-Cu, the tensile strength of the AA6063-AISI4130 steel exhibits the highest tensile strength. The increased strength of AA6063-AISI4130 steel is due to the higher strength of the steel than the copper. For all cases of Taguchi's DOE, AA6063-AISI4130 steel provides the highest strength. The obtained results of the experimentations are the input functions for the Taguchi's analysis. Commercially available statistical tool is used to analyze the obtained result for AA6063-AISI4130 and AA6063-Cu dissimilar welding joints. The quality attributes were determined by transferring the results into the signal to noise(S/N) ratio. The influence of the process parameters like UP, FP and FT on the tensile strength of the AA6063-AISI4130 and AA6063-Cu dissimilar welding joints are analyzed using the signal to noise ratio.  The factors FP, FT and UP are statistically major in the S/N ratio and are also observed that FP significantly influence results of the tensile strength followed by FT and UP. Fig. 4(a) and 4(b) displays the influence of the factors FP, FT and UP on the tensile strength of the said welded joints.

Main Effects Plot for SN ratios (AA6063-AISI41 30) Data Means
Signal-to-noise: Larger is better

Main Effects Plot for SN ratios (AA6063-Cu) Data Means
Signal-to-noise: Larger is better From Fig 4(a) and 4(b) it is seen that as the increment in the process parameters such as FT, FP and UP causes increment in the tensile strength initially for the both AA6063-AISI4130 and AA6063-Cu welded joints. Further increment in FP, FT and UP the decrement in the tensile strength has been observed. It is obvious that increment in the pressure deceases the strength of the welded joints.

ANalysis Of VAriance (ANOVA)
The experimental results, which inputted into the Taguchi's DOE, were analyzed using ANOVA. ANOVA gives the confidence level of each parameter and its percentage contribution on the tensile strength of the welded joints. The confidence level considered here is 95%, thus the significant level is 0.05. The contribution of each parameter on the strength is given in Tab Fig. 4(a) displays the optical microstructure of the AA6063-AISI4130 joint at the R/2 location. On the AA6063 side of the joint, there were primarily two zones: base material (BM) zone and heat-affected zone (HAZ). Between BM and HAZ, the dynamic recrystallized zone (DRZ) and thermal mechanically affected zone (TMAZ) can be found. The thermalmechanical connection effect caused the interface temperature to rise quickly during friction welding, which encouraged the recovery and dynamic recrystallization of the aluminum alloy close to the interface. The microstructure evolved from fine, equiaxed grains to streamline shapes, leading to the formation of the dynamic recrystallized zone. In a transition zone (TMAZ), where most grains were bent, and most partial grains underwent dynamic recrystallization. Streamlines also underwent considerable plastic deformation, changing from their original straight to curved shape. Because the interface temperature, which was lower than that of fusion welding but still reached the dynamic recrystallization temperature, grains in the HAZ were smaller than those of the BM in Fig. 4(a). The joint also primarily had two zones on the AISI4130 side: the BM and HAZ. Because of broken of partial original grains, by increased UP at higher temperature along the interface, the grains in the HAZ on the AISI4130 side were finer. As shown in Fig. 4, the obvious reaction layer between AA6063 and AISI4130 developed with an average thickness of 2.5μm as a result of the shared distribution of alloying materials under the influence of thermo-mechanical coupling. The interface between reaction layer and AISI4130 was smooth, whereas the interface between reaction layer and AA6063 was rough as shown in Fig. 4(a), indicating that the reaction layer advance towards the side of AA6063. The microstructure of the reaction layer in the R/2 site under various welding circumstances is shown in Fig. 4(b). The thickness of the interfacial intermetallic compounds (IMC) initially declined with an upset pressure of 97 MPa, but it subsequently began to stabilize (approaching 2.50μm) as friction pressure increased. It was evident from the examination of the interfacial microstructure that excessive interfacial IMC thickness significantly reduced the UTS of AA6063-AISI4130 joint. In order to thoroughly extrude the oxidized metal and dangerous contaminants in the interface, which was advantageous to metallurgical reaction, it was profitable to increase upset pressure. The UTS of joints was improved by increasing upset pressure up to 97MPa with friction pressure of 71 MPa and friction time of 4 sec. According to the analysis above, the ideal parameters were FP=71MPa, FT=4sec, and UP=97MPa, with a maximum UTS of 301 MPa for the AA6063-AISI4130 joint. After UP of 68MPa, backscattered pictures at the interface show a decline in AA6063-Copper joint quality ( Fig. 5(a)). At the interface center, a flaw in incomplete joining may be detected. Cracks developed in aluminum close to the interface at the R/2 and edge positions. Microcracks may be seen on the side of AA6063 in the backscattered picture at the edge position ( Fig. 5(a)). From the Fig. 5(a), it can be observed that the propagation of irregular microcracks along the grain boundaries of AA6063 at edge position. The tensile strength of edge samples that underwent UP of 97MPa drastically decreased because of the presence of microcracks. The interface of the two layers of the AA6063-Copper joint for different UP is given in Fig. 5(b). The dark IMC layer near the AA6063 appears to be Al2Cu, and the other layer's composition is similar to Al4Cu9. Al2Cu and Al4Cu9 layers can't be separated from one another because IMC layers are relatively flat and parallel in the centre. The thickness of IMC layers grows as the distance from the centre rises, and some minor IMC particles disperse in the AA6063 at the R/2 position. After UP of 38MPa, the IMC layer thickness slightly increases. The UTS of AA6063-Copper joint was improved by increasing upset pressure up to 97MPa with FP of 71 MPa and FT of 4 sec which gets the maximum UTS of 222 MPa.

Confirmation Experiment
The final step in design process is the confirmation experiment. The experimental outcomes attained must validate using the statistical analysis. The statistical analysis was determined using the regression equation carried out using MINITAB tool. The regression model gives the relation between the variable and the response, by fitting the linear equation to the existing data. The regression Eqns. 1 and 2 respectively obtained from the analysis of AA6063-AISI4130 and AA6063-Cu welded joints.  According to the results of the ANOVA, FP is the most important factor that affects tensile strength, followed by FT, and UP is the least important component. The regression Eqns. 1 and 2 shows that an increase in the FP, FT, and UP results in an increase in the tensile strength which can be also seen in Fig 4. Tab. 6 compares the results of the experiment with the statistical conclusions drawn from the regression model. The experimental and statistical tensile strength values were observed to vary by an error percentage ranging from 1.4 to 11.1 percent for welded joints made of AA6063-AISI4130, whereas it varies from 2.2 to 10.3 percent for AA6063-Cu. As a result, except for case 5, the tensile strength estimated by the experimental and regression models agree with one another, with an error rate of less than 10 percent. Hence the validation of experimental and statistical results, agree with each other and are within the permissible error [20].

CONCLUSIONS
he experimental work on RFW of AA6063-AISI4130 and AA6063-Cu dissimilar metal welded joints led to the following conclusions.  The RFW of the dissimilar joints such as AA6063-AISI4130 and AA6063-Cu was successfully performed using different parameters. The maximum tensile strength obtained for the AA6063-AISI4130 is 26% higher than AA6063-Cu welded joints. This is a result of the use of AISI4130, which is stronger than copper and aluminum.  The Taguchi's and ANOVA analysis reveals that the FP influences more on the tensile strength of mentioned welded joints followed by the FT and UP.  The SEM micrographs reveals the formation of intermetallic compounds on the side of AA6063. The formation of microcracks is on the side of AA6063 which propagates along the grain boundaries at R/2 location. Thus, the failure occurs at the AA6063.  The confirmation experiment through the statistical analysis agrees with the experimental results. Thus, the experimental results were validated.

FUNDING
his research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
T T