Friction Stir Welding of Thin Sheets of the AA2024-T3 Alloy with a Ceramic Tool: RSM and ANOVA Study

In this study a new ceramics tools with different groove distributions were designed and manufactured in order to enrich technological storage of joining thin-wall structures and obtain sound joint with high quality of alclad AA2024-T3 alloy of 0.5 mm in thickness. Four types of tools were tested, without grooves, with 1, 2 and 6 grooves. The tools are made of two materials. The straight shank is made from tungsten carbide and tool body made from ceramics strengthened with whiskers. The influence of technological parameters on the strength of FSW joints was tested by the RSM (Response Surface Methodology) and ANOVA (analysis of variance) method. The least durable weld is produced by a tool without grooves. The 1 and 2-flute tool produces a good quality weld over a wide range of tool speeds. It has been shown that the grooves on the tool shoulder significantly affect the quality of the obtained FSW joint.


Introduction
Friction stir welding (FSW) is a solid-state welding technique that has evolved as a solution for joining different metal sheets especially dissimilar materials that are difficult to weld. It is currently used as an alternative to riveting for e.g. the assembly of airplane fuselages [1]. FSW uses a rotary pin to locally mix the materials of the two sides of the joint below the melting point temperature [2]. Thus, the formation of welding defects such as hot cracking is prevented [3]. Friction stir welding is a continuous, hot shear, autogenous process involving non-consumable rotating tool of harder material than the substrate material [4]. Fig. 1 explains the working principle of FSW process.

Fig. 1. Schematic representation of FSW principle
In order to increase the efficiency of FSW welding, many different tool variants are being developed. Both the tool material and the tool geometry are subjected to modification. Tool material should be wear-resistant under high thermal stress [5]. Tool shoulder plays a key role in generation of surface friction and rise of workpiece temperature. Tool pin generates stirring and therefore materials in processed zone experience sever plastic deformation. Flowing and recrystallization in processed zone occurs with tool rotation and linear movement. This results in creation of a fine equiaxed microstructure [6,7].
In this paper, the mechanical properties (tensile strength) of the samples of welded 2024-T3 aluminum alloy sheets were investigated to determine the best rotational speed and welding speeds. The work determines the optimal parameters using one of the optimized methods to ensure proper welding performance based on tensile strength, study the obtained optimal parameters of mechanical properties and performance using the design of experiments (DOE) technique and development of models. It can help designers and engineers to achieve perfect welding. ANOVA techniques were used to identify the relevant factors influencing the ultimate tensile strength (UTS).

Material and Experimental Procedure
The initial material used in this work is a cold-rolled commercial AA2024-T3 aluminum alloy sheet with the 0.5 mm in thickness Alclad covered. In this investigation, the joining region are carefully cleaned prior to welding. After polished by abrasive paper and cleaning with acetone, several weld plates were subjected to FSW along the rolling direction. The blank sheet dimensions were 180×100 [mm] (Fig. 2). The FSW experiments were carried out using special prepared CNC milling machine MAKINO PS95 and the welding tools (Fig. 3). Cylindrical tool made from whiskerreinforced ceramic with geometrical features and process inputs are reported in Table 1.  Tool worked without a tilt angle, perpendicular to the surface of the welded material. The butt joint configuration was prepared to produce the joints. Welding has been done on the 180 [mm] long section.
A research plan was prepared for the experiment. Design-Expert software by Stat-Ease, Inc. was used. The range for technological parameters was determined for the rotational speed from 1000 to 2000 [rpm] and the welding speed from 200 to 1000 [mm/min] (Fig. 4). The last factor was the number of spiral grooves on the tool face. On this basis, a matrix of technological parameters was created. The set of parameters is shown in Fig. 5. Several regression models have been tested. The best results were obtained for the CUBIC model (Table 2).  Backward elimination regression model were developed. As an input factors tool rotation per minute, feedrate and number of grooves were selected. Factors coded -1 for lower and +1 for high limit, no. of grooves set as categoric 4 levels: 0, 1, 2, 6 which presents Table 3. Ultimate tensile strength test results have been taken into account as the response. Performed analysis of variation (ANOVA) and cubic model were selected with regard to fit summary of different models. Then applied backward eliminator algorithm which removes insignificant input factors with p-value less than 0.1 but with hierarchical agreement. Table 4 presents ANOVA for Reduced Cubic Model of UTS. The Model F-value of 19.24 means that model is significant and only a 0.01% chance that this large values could result due to noise. P-value for factors is less then 0.05 which means that model is significant.
The modified CUBIC model was subjected to a convergence analysis. The following analyzes were performed: Externally Studentized Residuals, Predicted vs. Actual analysis and Cook's Distance analysis [8]. The obtained results are shown in Fig 6.  The RMS (Response Surface Methodology) [9] analysis shows the influence of the welding parameters on the strength of the FSW joint (UTS) depending on the used tool (Fig.7).

Results
The tool made of tool whiskers reinforcement ceramics fully meets the expectations regarding the implementation of the FSW process. Tool ceramics are characterized by low thermal conductivity which allows to shorten the dwell time significantly. Another advantage of this approach is high resistance to abrasion and mechanical loads. The manufacturing process of an FSW tool is typical for the manufacture of cutting tools. No problems were encountered in this regard. Moreover, ceramic tools do not shown any signs of wear. The geometry of the tool shoulders significantly improves the quality of the weld. A single concentric helix has been shown to help produce the best joint.
The applied technological parameters of FSW welding allowed for the preparation of regression models with a relatively high convergence of nearly 80% as evidenced by the performed diagnostics of the CUBIC model. Application advanced statistical analyzes to select the optimal technological parameters for welding AA2024-T3 alloy sheets of 0.5 mm in thickness. The results of the experiment confirmed these assumptions.

Conclusion
• The use of a whisker-reinforcement ceramic tool enables the production of high-quality FSW joints with a strength exceeding 95% compared to the parent material. • The geometry of the helical grooves on the face of the tool greatly affects the quality of the FSW joint.