Optimization of friction stir welding processes using hybrid-taguchi methods: a comparative analysis

During friction stir welding (FSW) of similar or dissimilar materials, it is observed that the weld quality depends on various input parameters, like tool rotational speed, welding speed, axial load, tool shoulder geometry, tool tilt angle, pin geometry etc. To achieve the best weld quality, it is thus recommended to operate the FSW process while setting its input parameters at their optimal levels. In this direction, Taguchi methodology has evolved out as a potent single-objective optimization tool for deriving the optimal settings of FSW parameters for each of the weld characteristics which are often impractical to maintain during real-time welding operation. To resolve the issue, this paper proposes the applications of different mathematical tools hybridized with Taguchi methodology, like Taguchi-loss function, Taguchi-grey relational analysis (Taguchi-GRA), Taguchi-super ranking, Taguchi-data envelopment analysis-based ranking (Taguchi-DEAR) and Mahalanobis-Taguchi system to act as multi-objective optimization tools for simultaneous achievement of the most desired values of several conflicting weld characteristics. A comparative analysis is also conducted to validate their performance based on two past FSW experimental datasets. It is noticed that the optimization performance of Taguchi-loss function, Taguchi-GRA and Taguchi-DEAR approaches is exactly the same for both the considered examples. For the first example, an optimal combination of tool rotational speed = 900 rpm, welding speed = 60 mm/min, shoulder diameter = 18 mm and pin diameter = 5 mm; and for the second example, an optimal intermix of tool rotational speed = 950 rpm, traverse speed = 28 mm/min and tool tilt angle = 3° would provide the compromised values of all the weld characteristics under consideration.