Influence of pulsed TIG welding process parameters on the mechanical characteristics of AA5083 with AA6082 weldments

Aluminium alloys have been widely accepted in manufacturing lightweight materials with high strength. Therefore, welding aluminium alloys is essential in industrial applications to attain complex shapes. In the present work, AA5083 and AA6082 dissimilar alloys were welded using pulsed tungsten inert gas (PTIG) welding since PTIG reduces welding defects more than TIG welding. But to get better mechanical strength on the weld joints, PTIG welding process parameters must be optimized. During PTIG welding, peak current, pulse frequency, and welding speed were chosen as the input parameters, and ultimate tensile strength (UTS) and microhardness were measured as output responses in the current investigation. The UTS of the welded AA5083-AA6082 alloys was predicted using an empirical relationship. For the design of experimental trials, a three-variable, five-stage central composite design is adopted. The findings indicate that the welding speed impacts tensile strength the most, followed by the peak current and the pulse frequency has the least impact. Therefore, the peak current of 197 A, pulse frequency of 4.9 Hz, and welding speed of 181 mm min−1 was identified as the optimal welding parameters to weld AA5083 and AA6082 alloys with high UTS values. The hardness analysis on the optimized welded samples showed that the lowest hardness values of 40 to 50 Hv0.5 and the highest value of 90 to 100 Hv0.5 were observed on the HAZ of the AA6082 side of the weldment.


Introduction
The marine, automotive, and aerospace sectors paid close attention to the aluminium alloys AA5083 and AA6082 because of their beneficial qualities, including outstanding formability, good strength, exceptional corrosion resistance, and excellent weldability. Production of fuel-efficient, lightweight structures is under demand, and integration of preferred joining techniques for aluminium alloys with zero micro-or macrodefects is still necessary [1,2]. This necessity has sparked innovation in real-time production system joint design considerations. The AA5083 is a non-heat treatable alloy mainly used in shipbuilding, armour vehicle, etc [3]. Hence work hardening is used to increase strength even further. The availability of alloying elements (Mg, Mn, Fe, and Si), which cause the dispersion of phases and solid solution strengthening inside the matrix, determines the intrinsic hardening effect. Because of its deep drawing and high stretch forming capabilities, the AA5083 alloy is frequently used as chilled rolled sheets in lightweight vehicles' exterior and inner structural panels [4,5]. AA6082 is a heat-treatable aluminium alloy used in trusses, cranes, ore skips, etc. The worst mechanical property degradation occurs in heat-treatable alloys because the fusion process makes it impossible to prevent the development of softening zones in the weldment. The axial grain structure sometimes referred to as the typical fusion welding circumstance is characterized by coarse and long columnar grains appearing in the weld zone after solidification in the direction of the heat supply. Compared to heat-treatable aluminium alloys, the overall drop in mechanical characteristics is less severe for non-heat-treatable aluminium alloys. So, research is required to reduce the drop in mechanical properties of heat-treatable aluminium alloys during welding processes.
Various aluminium alloys have been joined using constant current TIG welding and studied by numerous researchers [6][7][8][9][10][11][12]. But the main issue when welding thin parts using TIG welding is distortion. However, the pulsed TIG (PTIG) technique controls distortion. In PTIG, the welding current is cycled at a predetermined regular frequency from a high level to a low level. The high peak current level is often chosen to provide good penetration and bead contour. In contrast, the low background current level is set high enough to maintain a steady arc. PTIG welds have metallurgical benefits commonly mentioned in the literature, such as increased weld depth/width ratio, grain refinement in the fusion zone, reduced HAZ width, less distortion, control of segregation, reduced hot cracking sensitivity, and lower residual stresses [6]. All of these benefits contribute to better mechanical characteristics compared with TIG welding. During the welding investigations of the Al-5083 alloy, Zhu. C. et al [7] employed narrow-gap gas metal arc welding (GMAW). They revealed that the welding current (WC) and welding speed (WS) substantially influence the formation of pores. It is claimed that with sufficient arc oscillation and preheating at 250°C, weld strength up to 90% could be achieved. Singh, L., et al [8] improved the strength of TIG-welded Al-5083 by optimizing the welding process parameters (WPP) using the Taguchi technique. It was claimed that the tensile strength decreased as the welding current was raised over its optimal level. Mustafa, U., et al [9] discovered small precipitates of Al 6 (Fe, Mn) at the grain boundaries that led to a high hardness value in their investigation of the corrosion and mechanical properties of TIG-joined Al-5083. Guo, Y., et al [10] investigated the effect of WS, current, and plasma gas flow rate (GFR) on Al-5083. It has been discovered that different heat cycles significantly altered the joint's microstructure and bead width. Additionally, it has been noted that penetration increases with both higher WC and slower WS. Liang, Y. et al [11] used a hybrid welding process of TIG-Cold metal transfer TIG-CMT to examine the impact of WC on the microstructural and mechanical characteristics of 6061 alloy. A high welding current decreased the strength of the joint because of the prolonged solidification time that formed large grain in the HAZ, and, as a result, a reduction in hardness was witnessed. Yao Liu et al [12] evaluated two fusion welding methods: the tendency for pore development between TIG and GMAW. The results indicated that compared to the GMAW, TIG welding created fewer pores.
The authors of the abovementioned experiments have utilized constant welding parameters to investigate how TIG-welded different aluminium alloys' mechanical characteristics were affected. But it is crucial to optimize the appropriate welding parameters. While in the optimization of welding parameters, only few researchers have studied PTIG welding of aluminium alloys. The detailed review of the literature makes it abundantly clear that there has been less study on the optimization of WPP to improve the microstructural and mechanical characteristics of this dissimilar AA5083-AA6082 during PTIG welding [6][7][8][9][10][11][12]. Therefore, in this study, the welding process parameters are optimized using the response surface method (RSM) to improve the mechanical and microstructural characteristics of the PTIG-joined AA5083-AA6082 alloys. The output response parameter taken for optimization is the UTS value. To learn more about the grain size, orientation, precipitate form and size, precipitate distribution, and precipitate volume percentage, a thorough microstructural examination of the sample welded using optimal welding conditions was conducted. To investigate the changes in joint strength and hardness in the various zones, microhardness tests results are shown on the sample that was welded using optimal WPP.

Weld joint fabrication
The parent metals AA5083-H111 and AA6082-T6 are purchased as 5 mm thick sheets and cut into required dimensions of 100 mm×55 mm. The elemental analysis of the parent metals was analysed using optical emission spectroscopy (Spectro make, model Arcspark) and is presented in table 1 and their mechanical properties are listed in table 2. Peak current (PC), pulse frequency (PF), and welding speed (WS) are the three main WPP varied in the pulsed TIG welding process. The base current of 50 A and an average arc voltage of 30 V were maintained during the 20 different welding experiments. ER5356 filler rod of 2.4 mm diameter at tip angle of 60°with argon shielding gas at the gas flow rate of 6 l min −1 is used during the welding process. The PTIG welding process is conducted by maintaining a constant gap of 2 mm between the parent metal and the tungsten electrode.
The limits and their levels for the welding parameters are obtained by performing a detailed literature survey and are listed in table 3 [8,13,14]. A V-groove of 30°is taken on the joining edge of the specimen using a milling machine. The base metal is cleaned with acetone and is clamped tightly. Welding was conducted using the various WPP combinations framed using a central composite design (CCD).

Metallurgical characterization
The metallurgical characterization of the weldments is investigated utilizing a scanning electron microscope (SEM) (Carl Zeiss make, model Evo 18 research). Using a wire electrical discharge machining (WEDM), samples for the metallurgical characterization were taken along the weldment's transverse direction. The extracted samples are mounted using a hot mounting press to simplify polishing. SiC abrasive sheets with grit sizes between 80 to 2000 are used to polish the mounted specimens. The final polishing is carried out using a silk cloth and diamond paste. After polishing, the samples were thoroughly etched by applying Keller's reagent (1% HF, 1.5% HCl, 2.5% HNO 3 , and 95% H 2 O) to disclose the grains and precipitates. The microstructural images were obtained from the polished samples using SEM images. The elements present at the weld zone are found using SEM-EDAX analysis.

Mechanical characterization
Tensile testing and microhardness studies were used to characterize the weldment mechanically. According to ASTM-E8 standards, samples for tensile testing were taken using an EDM machine along the weldment's transverse direction. Utilizing a UTM (Bluestar make, model LDW50), tensile tests were performed at a crosshead rate of 1 mm min −1 . With the aid of a Vickers microhardness tester (Esewy make, model EW-423DAP), the hardness of the weldment was analyzed along its various zones. A diamond indenter was used to deliver a load of 0.5 kg for a dwell duration of 20 s during the hardness test. The weldments' top, middle, and bottom rows were analyzed to measure hardness. Forty-one hardness values were taken at each row along different zones of the weldment with equal spacing of 0.5 mm.

Results and discussion
3.1. Analysis of weld morphology Figure 1 presents 20 welded samples using the different welding combinations obtained as per RSM. Figure 2(a) shows the extracted tensile samples as per ASTM-E8 standards and figure 2(b) shows the typical stress-strain curve obtained during the tensile test experiment of the 11th sample, which shows UTS of 128 MPa. Table 4 shows the 20 experimental runs along with the output of the UTS value. The visual inspection shows that all the welds are intact and defect-free.

Framing empirical relationship using RSM
RSM combines mathematical and statistical methodologies to relate and maximize a response of interest affected by several input factors. For example, PC, PF, and WS all influence the ultimate tensile strength (UTS). Table 4 presents the UTS value produced by the weldment and the design matrix for the trials framed using the RSM approach.
The CCD was chosen for this work out of a wide range of experimental design approaches that may be used to analyze the regression coefficient. To get every coefficient, the Design Expert tool was employed. Using the

Validation of optimized results
Analysis of variance (ANOVA) was used to assess how effectively the established relationship has behaved and is given in  illustrates the relevance of the lack of fit. Noise has a 0.04% possibility of causing a substantial lack of Fit F-value. The discrepancy between the predicted R 2 of 0.8497 and the adjusted R 2 of 0.9616 is less than 0.2, indicating that they are reasonably in agreement. Adequate precision measures the signal-to-noise ratio. The ideal ratio is at least 4, and an obtained ratio of 22.063 exhibits a significant signal. A perturbation plot was drawn representing the effect of each input parameter with respect to the output response, UTS and is shown in figure 3. The perturbation plot is plotted for each input parameter by keeping the other two input parameters at the centric points (i.e. coded value = 0). From the plot, it was observed that an increase in peak current (A) causes a rise in the heat input, which increases the weld area, weld penetration, and weld width. The increase in peak current enhances the weld's heat input, which causes a significant amount of the material around the welding face to melt, expanding the weld zone, according to similar observations by Bagha et al [15]. Vikash Chaudhary [16] has stated that an increase in peak current improves the weld area, but high heat input causes stress to concentrate on the weld pool, lowering the mechanical characteristics. Vibration amplitude increases with pulse frequency, resulting in finer grains. Additionally, it can be shown that increasing the WS decreases the fusion zone area by reducing the heat input. The heat input is high when the welding speed (C) is too low, which causes the fusion zone (FZ)'s breadth and depth to rise. As a result, poor base metal fusion will occur because of high WS. Additionally, it was shown that substantial porosity results from rapid cooling and vice versa. While cooling rates are sluggish, gas bubbles have plenty of time to float, mix, and exit the molten weld. As a result, fewer pores were created in samples that received medium and moderate heat input. In addition, the increase in the fluidity of the molten pool increases the likelihood that trapped gases will escape and give birth to improved mechanical characteristics because of the low porosity [17,18]. Figure 3(a) displays the perturbation plots of the output response, and tensile strength, to the input parameters, PC, PF, and WS (a). The perturbation plot consistently demonstrates that the UTS rises as the peak current increases from 140 A to 210 A. However, a dip in tensile value was noticed when it crossed 210 A. The UTS value increases as the PF rise from 2 to 5 Hz. The UTS value decreases as the pulse frequency is raised over 5 Hz. The UTS value increased as the WS increased from 140 to 170 mm min −1 , while tensile strength decreased as the welding speed grew. The projected versus actual plot is displayed in figure 3(b). The square-coloured dots in the graph represent the values acquired, while the linear line on the plot represents the anticipated value line. The experimental values coincide with the anticipated value line, as seen in the plot. This demonstrates that the model is successful as well.

Study on the interaction between input and output response
To efficiently determine the optimal WPP, the interaction plots between input and output parameters for the response surface analysis were extracted using the design expert 13 software. The interaction plots show the interaction effect of two input parameters concerning the output parameter, tensile strength, by keeping the third input parameter in the centric value (i.e. coded value = 0). When the contour is circular, it indicates the  Figure 4(a), shows the interaction effect of PC and PF with respect to the output response UTS by keeping the WS at the centric point value. It can be seen that the UTS increases as the PC increase from 140 A to 200 A, where it achieves its highest value. Tensile strength decreases as the peak current rises above 200 A. The primary cause is grain coarsening at the weld pool when the peak current increases because of high heat and slower cooling during solidification. When the peak current is low, the heat input is also low and insufficient to form strong joints, but when the peak current is high, the heat input is much increased, resulting in coarse grains. Thus, an optimum peak current value exists, resulting in a better tensile strength rating. The optimum peak current value for the present investigation was identified as 197 A. Figure 4(b), shows the interaction effect of PC and WS with respect to the output response UTS by keeping the PF at the centric point value. When the pulse frequency is 2 Hz, it is shown that the tensile strength is relatively low. Tensile strength increases as the pulse frequency rise from 2 to 5 Hz. The vibration and temperature oscillations formed on the molten metal are decreased to a higher range as the pulse frequency increases above the optimal value, which has a lower impact on the weld pool. However, when the pulse frequency value is in its optimal range, the molten weld pool is aggressively agitated, which causes the weld zone to have finer grains. As a result, an ideal pulse frequency value yields finer grains. The ideal pulse frequency for the current investigation is 4.9 Hz. Figure 4(c), shows the interaction effect of WS and PF with respect to the output response UTS by keeping the PC at the centric point value. As observed, the tensile strength increases as the WS increases from 140 to 180 mm min −1 , reaching its maximum value at 170 mm min −1 . Tensile strength values decreased as WS increased over 180 mm min −1 . The primary cause is that the heat input is particularly high at lower WS, which causes grain coarsening in the FZ. A fast WS resulted in insufficient heat input and an inadequate heat supply.

Optimization of welding parameters
The optimal welding parameters are determined using RSM, and empirical relationships are established. After being employed for optimization, the optimized coded numbers were changed back into actual values. Finally, the processed values are optimised using design expert analysis software. Tensile strength is considered to be the maximizing problem in the optimization. The experiment's findings indicated that the optimum welding parameters: PC 197 A, PF 4.9 Hz, and WS 181 mm min −1 resulted in the UTS value of 146.8 MPa. Table 6 shows the experimental and predicted optimal welding process parameters and their corresponding results.

Microstructural and elemental analysis
The microstructural characteristics of the base metal and the weld zone has taken at 2000X magnification using an SEM and are shown in figure 5. The SEM micrograph of the parent metals AA5083 and AA6082 are presented in figures 5(a) and (b) sequentially. The SEM images taken at the HAZ of AA5083 and AA6082 are shown sequentially in figures 5(c) and (d). Finally, the SEM image of FZ of the PTIG welded sample joined using the optimized welding parameters is given in figure 5(e).
It is known that the aluminium alloy AA5083 attains its strength from solid solution and precipitate strengthening. In a substitutional solid solution, the Mg atoms replace the Al atoms. The major intrmetallic particles that give strength to the aluminium alloy AA5083 are Al 3 Mg 2 , Al 6 (Fe, Mn), and Mg 2 Si. It can be seen in figure 5(a) that the base metal zone is covered fully with the intermetallic particles and only a few areas show the α-Al matrix. On the other hand, the aluminium alloy AA6082 is a heat-treatable aluminium alloy which attains its strength only through precipitate strengthening. The AA6082 has magnesium and silicon as the major alloying element, so the only strengthening precipitate present in this alloy is Mg 2 Si. It is evident from figure 5(b) that the base metal zone has a lot of Mg 2 Si precipitates, which are spread all over the α-Al matrix.
On comparing the microstructures of BM and HAZ of the AA5083 side, a less reduction in the quantity of the intermetallic particles is observed. It can also be seen that the sizes of the intermetallic particles are reduced to a greater extent. On comparing the BMZ and HAZ of the AA6082 side, it can be seen that a lot of strengthening precipitates (Mg 2 Si) have been lost during the welding process. However, it can be seen in the HAZ of AA6082 that only a few small precipitates are left out. The main reason behind this is the evaporation of magnesium alloys during the PTIG welding process. Similar observations have been made by Vasu et al [19]. It has been reported that there will be around a 39% reduction in the Mg alloy due to evaporation. Since Mg 2 Si is the only strengthening precipitate of the AA6082 aluminium alloy, the dissolution of the precipitates at the HAZ of the AA6082 side makes it the weakest zone of the weldment. As a result, the fusion zone's SEM image contains many elongated intermetallic particles spread all over the FZ. This has happened because of the agglomeration of undissolved intermetallic particles during solidification.
The SEM-Energy Dispersive x-ray Analysis (EDAX) was carried out on the fusion zone of the sample welded using the optimized welding parameter given in figure 6. The elemental mapping method is done to know the distribution of the particles in the fusion zone. It can be seen that there is a marginal drop in the percentage of Fe, Zn, Cr and Cu in FZ compared to the BMZ. But a high drop in the percentage of Mg particles is seen. The % of Mg in the base metal is 4.31% and 1.46% for AA5083 and AA6082, respectively, but it can be seen that only 2.7% Mg can be seen in the fusion zone. This proves that a significant quantity of Mg elements evaporated during the PTIG welding process [20]. It can also be seen that there is a minimum loss in the quantity of Si and Mn particles too at the fusion zone of the weldment. The dissolution of the Mg 2 Si precipitate during the welding process is confirmed by the observations made through EDAX analysis. Figure 7 shows the results of the SEM analysis of the fracture surface of the tensile specimen. Figures 7(a) and 7(b) show the fractography images of the base metals AA5083 and AA6082 sequentially. Figure 7(c) displays the fractography of the joint PTIG-weldment using the optimum welding parameters. It is obvious from figure 7(a) that there are many dimples of varying sizes on the fractured surface. The presence of the dimples is a crucial aspect of a ductile fracture because the size of the dimples reveals the material's ductility. High-strength materials are indicated by fine dimples, whereas low-strength materials are indicated by large dimples [17]. Since figure 7(a) has many small dimples, it shows that the AA5083 base metal is strong and ductile. In figure 7(b), many fine dimples and quasi-cleavage fractures are visible. It can also be seen that there are few intermetallic particles on the fractured surface. The presence of excessive intermetallic and other precipitates in FZ supports the claim that it serves as void coalescence and crack expansion nucleating locations [21]. Because these precipitates weaken the grain boundary, grain sliding is more likely to occur [22,23].  surface show that the samples have undergone brittle failure. The mixed failure mode clearly shows that the TIG welding of AA5083-AA6082 alloys has reduced the ductility.

Hardness analysis
To analyze the hardness variation of the welded joints, the hardness survey was carried out at different zones of the weldment at the lines taken at the top, a middle and bottom surface of the weldment and is shown in figure 8. It is evident from figure 8 that the lowest hardness value is obtained at the HAZ of AA6082 side ranging between 45 to 50 Hv 0.5 . A decrease in hardness value is attained because of 'Mg' evaporation and the dissolving of second-