Examination of the Mechanical, Corrosion, and Tribological Behavior of Friction Stir Welded Aluminum Alloy AA8011

e-mail: r.vairavignesh@gmail.com Aluminum alloy AA8011 is emerging as a promising material for modern engineering applications in which improved tensile strength, hardness, corrosion-resistance, and wear-resistance of materials are required. Typically, AA8011 alloys are utilized in air-conditioning ducts and heat exchanger fins in ships, leisure boats, luxury vessels, workboats, fishing vessels, and patrol boats. However, the conventional welding of AA8011 is a challenging procedure. In this context, this paper focuses on the development of an effective solid-state welding methodology for AA8011 alloy welding. The AA8011 alloy was friction stir welded by varying the tool rotation speed, traverse speed, and shoulder diameter. Examination of the Mechanical, Corrosion, and Tribological Behavior of Friction Stir Welded Aluminum Alloy AA8011


INTRODUCTION
Aluminum alloys are widely used in many industries due to their high strength to weight ratio, good ductility, and corrosion resistance (Vasudevan and Doherty, 2012). Typically, AA8011 alloys are utilized in air-conditioning ducts and heat exchanger fins in ships, leisure boats, luxury vessels, workboats, fishing vessels, and patrol boats (Magibalan et al., 2018;Mishra, 2020;Cui et al., 2019;Mandal, 2017). However, welding of aluminum alloys using the conventional welding process is a serious limitation. During conventional welding of aluminum alloys, weldments rapidly oxidize, resulting in porosity, incomplete fusion, and cracks (Palani et al., 2018a;2018b;Nandan et al., 2008;Ghetiya et al., 2016;Davidson and Neelakrishnan, 2018). This leads to poor joint efficiency, as well as poor corrosion, and tribological properties. Hence, aluminum alloys should be joined in solidstate to overcome the rapid oxidation phenomenon and improve the desired properties (Akca and Gürsel, 2016).
In the solid-state welding process, the material being welded is heated to less than the melting temperature of the base material and joined without any filler material or inert atmosphere (Vora and Badheka, 2019). Only a handful solid-state welding methods include friction welding, friction stir welding, explosion welding, forge welding, hot pressure welding, and ultrasonic welding (Akinlabi and Mahamood, 2020). Comparatively lower residual stresses and heat-affected zones in the weldments make friction stir welding a distinct and promising methodology to produce sound joints in aluminum alloy (Nandan et al., 2008;Midling et al., 1998;Mishra and Ma, 2005).
In friction stir welding, a non-consumable rotating tool is traversed along the joint line of the plates being welded by the action of the axial force (Midling et al., 1998;Wahid et al., 2016). The traverse of the rotating tool under normal load generates frictional heat, which plasticizes the material and enables the creation of a sound joint in the plasticized state (Vaira Vignesh et al., 2016;. However, excess or insufficient heat generation and poor or turbulent material flow result in observable defects such as voids, insufficient penetration, lack of fusion, surface lack of fill, root-flow defect, nugget collapse, surface galling, and scalloping in the friction stir weldments (Vaira Vignesh, 2018a;2018b). The defects could be reduced or eliminated through the proper choice of friction stir welding process parameters (Vaira Vignesh and Ramasamy, 2017).
Earlier studies suggest that tool rotation speed, tool traverse speed (welding speed), and tool profile (shoulder diameter, pin profile) are process parameters that have the highest impact on friction stir welding (Palani et al., 2018a;2018b;Nandan et al., 2008;Ghetiya et al., 2016;Davidson and Neelakrishnan, 2018;Wahid et al., 2016;Vaira Vignesh et al., 2018a;2018b;2017). In this study, aluminum alloy AA8011 is friction stir welded by varying the friction stir welding process parameters, namely tool rotation speed, welding speed, and tool shoulder diameter. Joint strength, elongation and efficiency, corrosion rate, and wear rate of the weldments were evaluated. Tensile specimen fractography, the surface morphology of corrosion test specimens and wear test specimens were analyzed using a high-resolution scanning electron microscope. The elemental composition of the corroded and worn region was analyzed using energy dispersive x-ray electron spectroscopy. Models based on a hybrid quadraticradial basis function were developed to correlate the friction stir welding process parameters with weldment properties (strength, elongation, efficiency of the joints, corrosion rate and wear rate). The influence of the process parameters on the evolution of properties was described. In addition, optimum process parameters for friction stir welding of aluminum alloy AA8011 was established.

Materials
The composition of the AA8011 aluminum alloy is given in Table 1. The alloy plate was received in the annealed condition. The plate was cut into 5 mm thick, 50 mm wide, and 150 mm long work pieces. Plate sides were milled and ground to ensure their proper clamp in the fixture.

Element
Al

Design of Experiments
This study emphasizes the role of friction stir welding process parameters, namely, tool rotation speed, welding speed, and tool shoulder diameter, on weldment properties. The chosen level of friction stir welding process parameters is given in Table  2. Process parameters were varied at three levels as per Box-Behnken design and the experimental layout for friction stir welding trials is shown in Table 3.

Friction Stir Welding
The work pieces were degreased with acetone and clamped in the fixture. The friction stir welding trials were performed in a numerically-controlled vertical milling center, as shown in the experimental layout in Table 3. A dwell time of 60 sec was selected for each friction stir welding trial. The friction stir welding tool was made from high-speed steel having the hardness of 65 HRC.  Chosen levels of friction stir welding process parameters.

Microhardness
Specimens were cut from the base material and weldments and polished in compliance with the ASTM E3-11 standard. Vicker's microhardness of the specimens was measured using a microhardness tester (UHL, as prescribed by the ISO 6507 standard). The specimens were indented using a diamond indentor, by applying the normal load of 500-gram force for 15 s.

Tensile Strength
Tensile test was performed to assess the ultimate tensile strength of the friction stir welded specimens. Two specimens were machined from each friction stir welded work piece using abrasive waterjet machining. The tensile test was performed in a computerized tensile test machine (Tinius Olsen) in keeping with the ASTM B557-10 standard. The stress-strain graph was plotted and analyzed in the dedicated software.

Wear Rate
The weldment wear rate was assessed based on the wear test conducted in a pin-on-disc tribometer. In the pin-on-disc tribometer, the pin slides across the disc at a particular velocity, under an axial load. Uniform contact between the pin and the disc was maintained using the suitable lever arrangement. 10 mm × 10 mm specimens were cut from the weldments and mounted in the hollow steel tube using a cold setting compound. The specimens were slid against the hardened steel EN316. The

Sl.
Tool rotation speed ( wear test parameters were chosen as follows: the load of 9.81 N, the sliding velocity of 0.628 m/s, and the sliding distance of 565 m. The mass of the specimen before and after the wear test was measured using a precision balance having the readability of 0.0001 g.

Corrosion
The corrosion test was performed by exposing small sections of the welded material to the test medium. The specimens were polished and degreased with acetone before the corrosion test. The specimens were immersed in an accelerated corrosion test medium, which was prepared by mixing 100 ml of distilled water, 10 g of sodium chloride (NaCl), and 5ml of acetic acid (20:2:1). The pH of the medium was 2.0, which was measured using a pH meter having the sensitivity of 0.1. The temperature of the medium was maintained at 30°C using a temperature-controlled water bath. The mass loss of the material was measured after the immersion period and the corrosion rate was calculated. The mass of the specimen before and after the corrosion test was weighed using a precision weighing balance having the readability of 0.0001 g.

Characterization
Tensile test specimen fractograph, the surface morphology of the corroded specimens, and the surface morphology of the worn specimens were observed under a high-resolution scanning electron microscope. The elemental composition of the corroded and worn out specimens were analyzed using energydispersive X-ray spectroscopy.

Quadratic -Radial Basis Function Model
Conventionally, the relationship between process parameters and response variables is modeling using the mathematical regression equations (linear, quadratic, cubic, or polynomial function). However, the serious limitation is that the regression equation could not model the complex non-linearity characteristics of response variables (Ilangovan et al., 2017;Ashwin et al., 2019;Anil Kumar et al, 2019;Murugan et al., 2018).
Recently, researchers have started working on the development of soft computing techniques like fuzzy-logic and artificial neural networks to associate process parameters with response variables that have non-linearity characteristics (Malekimoghadam and Icardi, 2019;Shehabeledeen et al., 2019;Barath et al., 2018). In this study, the authors have developed hybrid models integrating the quadratic function and radial basis function (mono-layered artificial neural network) to associate process parameters with response variables. The details of model development are discussed elsewhere (Vaira Vignesh et al., 2018c).

Microhardness
The average Vicker's microhardness of the base material and friction stir welded specimens is presented in Figure 1. The average microhardness of base material was 40.78 HV. The plot indicates that the microhardness of the majority of friction stir welded specimens is lower than the microhardness of the base material. Among the friction stir welded specimens, specimen FSW03 had the average microhardness of 40.56 HV, which is closer to that of the base material. The microhardness of specimens FSW08 and FSW06 is closer to the microhardness of the base material. Given that specimen FSW07 had pin-holes in the weldments it had the lowest microhardness of 27.78 HV, which is 32% less than the microhardness of the base material.

Elongation, Tensile Strength, Joint Efficiency
The tensile test was performed on the base material and friction stir welded specimens in a computerized tensile testing machine, and the stress-strain graph was plotted. The percentage elongation of the specimens was calculated using equation (1).
where ∆l is the change in length, l is gauge length.
The average percentage elongation of the base material and friction stir welded specimens in the course of the tensile test is shown in Figure 2 (a). The base material had the highest elongation of 22.88%, whereas specimen FSW12 had the lowest elongation of 5.63%. Among the welded specimens, specimen FSW11exhibited high elongation of 14.48%. The results indicate that the percentage elongation decreases after the welding process.
The average ultimate tensile strength of the base material and friction stir welded specimens is shown in Figure 2  The joint efficiency is the numerical value which is the ratio of the strength of the weld and the strength of the base material. The joint efficiency is one of the crucial aspects to introduce the safety factor in the welding process. The percentage joint efficiency of the joints was calculated using equation (2).
where σ FSW is the strength of the weld, σ BM is the strength of the base material.
The joint efficiency of friction stir welded specimens is shown in Figure 2 (c). The results indicate that the joint efficiency of all friction stir welded specimens exceeds 30%. However, joint efficiency greater than 50% indicates sound weldments. Specimens FSW01, FSW02, FSW03, FSW06, FSW08, and FSW11 had the joint efficiency greater than 50, whereas specimens  FSW04, FSW05, FSW07, FSW09, FSW10, FSW12, FSW13 had joint efficiency lower than 50%. Among friction stir welded specimens, specimen FSW11 had the highest joint efficiency of 55.62% and specimen FSW12 had the lowest joint efficiency of 33.34%.

Immersion Corrosion
The corrosion rate of the base material and friction stir welded specimens was determined based on the immersion corrosion test. The specimens were immersed in the accelerated corrosion environment and their mass loss was calculated. The corrosion rate of the specimens was calculated using equation (3).
where K is the conversion factor = 87.6, ΔM is mass loss, ρ is density, A e is exposure area, T is immersion time.
The graphical representation of the corrosion rate of the base material and the friction stir welded specimens is shown in Figure 4. The corrosion rate of all friction stir welded specimens was lower than the base material corrosion rate of 11.320 μm/ year. However, specimen FSW06 had the corrosion rate of 11.215 μm/year, which is closer to the corrosion rate of the base material. The surface of specimen FSW05 had a scaly appearance, as shown in Figure 5 (a), Figure 5 (b), and Figure 5 (c). The corrosion rate of specimen FSW05 was 1.414 μm/year, which is ~87.5% less than that of the base material. Specimen FSW06 had a scaly appearance with many corrosion pits, as observed in Figure 5 Figure 6 (a) and Figure 6 (b) respectively. The major elements in the corroded surface were Al, Si, O, and Fe. However, a peak corresponding to Na and Cl was observed in minor counts.

Wear Rate and Coefficient of Friction
Wear test was carried out in the pin on disk wear test machine. The wear rate of the specimen was calculated using equation (3). Experimental results are presented as a bar chart in Figure 7 (a).
where ∆M is mass loss, N f is normal force, L is sliding distance, F f is frictional force. The wear rate of the base material and friction stir welded specimens is shown in Figure 7 (a). The wear rate of the base material was measured to be 3.24×10 -7 g/Nm. The experimental results indicate that the wear rate varied between 1×10 -7 g/Nm and 5×10 -7 g/Nm in most of the friction stir welded specimens. Figure 8 (a) shows feeble wear tracks, Figure 8 (b) shows wear debris, and Figure 8 (c) shows the delamination and crater on the worn-out specimen FSW05. In line with the observations, the low wear rate of 1.26 ×10 -7 g/Nm was recorded for specimen FSW05. On the other hand, specimen FSW13 had the highest wear rate of 7.39×10 -7 g/Nm, which is ~56% higher than the wear rate of the base material. Figure 8  The elemental composition of the worn-out region of specimens FSW05 and FSW13 is shown in Figure 9 (a) and Figure  9 (b) respectively. The significant peaks were obtained for Al, Si, O, and Fe elements. Therefore, the wear debris consists of the chipped base material in its original and oxidized state. The coefficient of friction was derived from the ratio of frictional force to normal force, as given by equation (4). The average coefficient of friction for the base material and the friction stir welded specimens is shown in Figure 7 (b).
The average coefficient of friction of the base material and of the friction stir welded specimens is shown in Figure 7 (b). The coefficient of friction of friction stir welded specimens is significantly higher than the coefficient of friction of the base material. The coefficient of friction is similar in all friction stir welded specimens.

Quadratic -Radial Basis Function Model
Equation (6), Equation (7), Equation (8), Equation (9), Equation (10), and Equation (11) are hybrid models for predicting the microhardness, tensile strength, joint efficiency, elongation, corrosion rate and wear rate of the friction stir welded specimens, developed in the Matlab® technical computing environment. In the developed model, coded levels of process parameters were used (lower limit being -1 and the upper limit is +1).     Radial basis functions RBF 1 , RBF 2 , RBF 3 , RBF 4 , RBF 5 , RBF 6 were developed with a multiquadratic kernel and regularization parameter of 0.0001. The parameters for the radial basis functions RBF 1 , RBF 2 , RBF 3 , RBF 4 , RBF 5 , RBF 6 are given in Table 5. The developed models were assessed to determine their prediction efficiency based on statistical parameters, i.e. the coefficient of determination (R 2 ), the adjusted coefficient of determination (adj. R 2 ), and root mean squared error (RMSE). The closeness of RMSE value to 0, R 2 to 1, and the equivalence of R 2 and adj. R 2 indicate that the developed model is efficient in prediction (Cavazutti, 2013;Eriksson et al., 2000;Park, 2007). The statistical parameters for the developed models are given in Table 6.    shows that the R 2 value is close to 1, that R 2 , and adj. R 2 values are equal, and that RMSE values are close to 0. Therefore, the predictions of the developed models are statistically significant. A linear trend was observed between the experimental and predicted values of microhardness, tensile strength, joint efficiency, elongation, corrosion rate, and wear rate, as shown in Figure 10. The linearity of the graphs is indicative of the high prediction efficiency of the developed models. These statistically significant models were utilized to study the influence of friction stir welding process parameters on the microhardness, tensile strength, joint efficiency, elongation, corrosion rate, and wear rate, which are discussed in the following section.

The Influence of Friction Stir Welding Process Parameters on Microhardness
The influence of tool rotation speed, welding speed, and shoulder diameter on the microhardness of the friction stir welded specimens is shown in Figure 11. Figure 11 (a) shows the effect of tool rotation speed and welding speed at shoulder diameter of 18 mm on the microhardness of the welded specimens. Figure 11 (b) shows the effect of tool rotation speed and shoulder diameter at the welding speed of 45 mm/min on the microhardness of the welded specimens. Figure 11 (c) shows the effect of welding speed and shoulder diameter at the tool rotation speed of 1400 rpm on the microhardness of the welded specimens.
The summary of the contour plots is given in Table 7. The following results are deduced from the contour plots. • Friction stir welding of specimens at the tool rotation speed of 1400 rpm, welding speed between 55 and 60 mm/min and shoulder diameter between 15 and 15.5 mm resulted in the lowest microhardness. • Friction stir welding of specimens at tool rotation speed between 1500 rpm and 1600 rpm, welding speed 45 mm/min, and shoulder diameter between 20 mm and 21 mm resulted in the highest microhardness.

Sl.
Model

Influence of Friction Stir Welding Process Parameters on Tensile Strength
The influence of tool rotation speed, welding speed, and shoulder diameter on the tensile strength of the friction stir welded specimens is shown in Figure 12. Figure 12 (a) shows the effect of tool rotation speed and welding speed at shoulder diameter of 18 mm on the tensile strength of the welded specimens. Figure 12 (b) shows the effect of tool rotation speed and shoulder diameter at the welding speed of 45 mm/min on the tensile strength of the welded specimens. Figure 12 (c) shows the effect of welding speed and shoulder diameter at the tool rotation speed of 1400 rpm on the tensile strength of the welded specimens.
The contour plot summary is given in Table 8. The following results have been deduced from contour plots. • Friction stir welding of specimens at the tool rotation speed between 1500 rpm and 1600 rpm, welding speed between 30 mm/min and 35 mm/min, and shoulder diameter of 18 mm resulted in the lowest tensile strength.
• Friction stir welding of specimens at the tool rotation speed between 1200 rpm and 1250 rpm, welding speed of 45 mm/min, and shoulder between 20 mm and 21 mm resulted in the highest tensile strength.

Influence of Friction Stir Welding Process Parameters on Joint Efficiency
The influence of tool rotation speed, welding speed, and shoulder diameter on the joint efficiency of the friction stir welded specimens is shown in Figure 13. Figure 13 (a) shows the effect of tool rotation speed and welding speed at shoulder diameter of 18 mm on the joint efficiency of the welded specimens. Figure 13 (b) shows the effect of tool rotation speed and shoulder diameter at a welding speed of 45 mm/min on the joint efficiency of the welded specimens. Figure 13 (c) shows the effect of welding speed and shoulder diameter at the tool rotation speed of 1400 rpm on the joint efficiency of the welded specimens.
The contour plot summary is given in Table 9. The following results have been deduced from the contour plots.
• Friction stir welding of specimens at the tool rotation speed between 1500 rpm and 1600 rpm, welding speed between 30 mm/min and 40 mm/min, and shoulder diameter of 18 mm resulted in the lowest joint efficiency. • Friction stir welding of specimens at the tool rotation speed between 1200 rpm and 1250 rpm, welding speed of 45 mm/min, and shoulder between 20 mm and 21 mm resulted in the highest joint efficiency.   Table 9.
Summary of joint efficiency contour plots. The influence of tool rotation speed, welding speed, and shoulder diameter on the elongation of the friction stir welded specimens is shown in Figure 14. Figure 14 (a) shows the effect of tool rotation speed and welding speed at shoulder diameter of 18 mm on the elongation of the welded specimens. Figure 14 ( b) shows the effect of tool rotation speed and shoulder diameter at a welding speed of 45 mm/min on the elongation of the welded specimens. Figure 14 (c) shows the effect of welding speed and shoulder diameter at the tool rotation speed of 1400 rpm on the elongation of the welded specimens.

Sl. Reference
The contour plot summary is given in Table 10. The following results have been deduced from the contour plots. • Friction stir welding of specimens at the tool rotation speed between 1500 rpm and 1600 rpm, welding speed between 30 mm/min and 35 mm/min, and shoulder diameter of 18 mm resulted in the lowest elongation. • Friction stir welding of specimens at the tool rotation speed between 1200 rpm and 1250 rpm, welding speed of 45 mm/min, and shoulder between 20 mm and 21 mm resulted in the highest elongation.    The influence of tool rotation speed, welding speed, and shoulder diameter on the wear rate of the friction stir welded specimens is shown in Figure 15. Figure 15 (a) shows the effect of tool rotation speed and welding speed at shoulder diameter of 18 mm on the wear rate of the welded specimens. Figure 15 (b) shows the effect of tool rotation speed and shoulder diameter at the welding speed of 45 mm/min on the wear rate of the welded specimens. Figure 15 (c) shows the effect of welding speed and shoulder diameter at the tool rotation speed of 1400 rpm on the wear rate of the welded specimens.
The contour plot summary is given in Table 11. The following results have been deduced from the contour plots. • Friction stir welding of specimens at the tool rotation speed between 1400 rpm and 1500 rpm, welding speed of 45 mm/min, and shoulder diameter between 18 mm and 20 mm resulted in the lowest wear rate. • Friction stir welding of specimens at the tool rotation speed between 1200 rpm and 1250 rpm, welding speed of 45 mm/min, and shoulder between 15 mm and 16 mm resulted in the highest wear rate.

Sl. Reference
Process parameter range Wear rate (g/Nm) Tool rotation speed (

Influence of Friction Stir Welding Process Parameters on Corrosion Rate
The influence of tool rotation speed, welding speed, and shoulder diameter on the corrosion rate of the friction stir welded specimens is shown in Figure 16. Figure 16 (a) shows the effect of tool rotation speed and welding speed at shoulder diameter of 18 mm on the corrosion rate of the welded specimens. Figure 16 (b) shows the effect of tool rotation speed and shoulder diameter at the welding speed of 45 mm/min on the corrosion rate of the welded specimens. Figure 16 (c) shows the effect of welding speed and shoulder diameter at the tool rotation speed of 1400 rpm on the corrosion rate of the welded specimens.
The contour plot summary is given in Table 12. The following results have been deduced from the contour plots. • Friction stir welding of specimens at the tool rotation speed between 1200 rpm and 1400 rpm, welding speed of 35 mm/min and 45 mm/min, and shoulder diameter between 17.5 and 19.5 resulted in the lowest corrosion rate. • Friction stir welding of specimens at the tool rotation speed between 1550 rpm and 1600 rpm, welding speed between 55 mm/min and 60 mm/min, and 18 mm shoulder resulted in the highest corrosion wear rate.

Sl. Reference
Process parameter range Corrosion rate (μm/year) Tool rotation speed ( Effect of friction stir welding process parameters on corrosion rate (a) tool rotation speed and welding speed; (b) tool rotation speed and shoulder diameter; (c) welding speed and shoulder diameter.

CONCLUSION
Aluminum alloy AA8011 plates were successfully joined using the friction stir welding process. Friction stir welding tests were conducted on AA8011 by varying tool rotation speed, welding speed, and tool shoulder diameter. The joined plates were subjected to tensile test, wear test, and corrosion test. Quadratic -Radial Basis Function models were developed to correlate the friction stir welding process parameters with joint properties. The statistical parameters have verified the high effectiveness of the developed models. The results are as follows: • Friction stir welding is an effective methodology for joining aluminum alloy 8011 and properties are appreciably affected by friction stir welding process parameters. • Friction stir welding of a specimen at 1200 rpm, 45 mm/ min, 21 mm resulted in the maximum ultimate tensile strength of 62.20 MPa, with the high joint efficiency of 55.62%.

•
The microhardness of aluminum alloy 8011 was found to be 41.6 HV. Friction stir welding of a specimen at 1600 rpm, 45 mm/min, 21 mm resulted in the high hardness (40.56 HV) in the nugget zone among the tested specimens. However, joint strength was ~2% lower than measured in specimen FSW11. • The wear rate of aluminum alloy 8011 is 3.24×10 -07 g/ Nm. Specimen FSW05 (welded at 1400 rpm, 45mm/min, and 18mm) had the low wear rate of 1.26×10 -07 g/Nm. The results of the immersion corrosion test show that specimen FSW05 had the lowest corrosion rate of 0.0014 mm/year among the tested specimens.
• Based on joint efficiency as primary constraint, optimum process parameters for friction stir welding of aluminum alloy AA8011 have been established as follows: tool rotation speed of 1200 rpm, tool traverse speed of 45 mm/min, and tool shoulder diameter of 21 mm.