Multi-response optimization of friction stir welding using fuzzy-grey system

Abstract

Abstract This study reports that a fuzzy logic-based grey system using Taguchi’s method provides a reliable approach to the prediction of friction stir welding (FSW), which is capable of improving the multi-response performance of butt welds. Using the grey relational database, a fuzzy logic analysis is carried out. Analysis of variance is used to determine the effect of parameters of multi-response behaviors on butt welds. Meanwhile, multiple responses for multivariables are simultaneously optimized. The experimental results show that the fuzzy logic-based grey system using Taguchi’s design reveals the optimal settings of the parameters, which improves the whole properties by more than 28.04%, while the individual properties, such as tensile strength by 13.35%, bending strength by 1.91%, impact strength by 1.05%, and hardness by 0.26%, when compared with the best test in orthogonal arrays. Based on experimental validation tests, excellent agreement between model predictions and experimental results is shown. Notably, the intelligent method proposed is applied to butt welds, and the results of the implementation of fuzzy logic based grey system using Taguchi’s design proved its feasibility and effectiveness with respect to the improvement of the mechanical properties of FSW.