Agus Widyianto, Herminarto Sofyan, Gunadi Gunadi, Aan Yudianto, Semuel Boron Membala
Micro Friction Stir Welding (μFSSW) is a derivative of Friction Stir Welding (FSW), where in μFSSW, it is used to join materials with a thickness of less than 1 mm. Several parameters of the μFSSW process were used in this study, such as dwell time (ms) and plunge depth (micron). At the same time, the quality of the weld is a parameter of the success of a welding process. Weld quality in FSSW includes pin diameter (mm), shoulder diameter (mm), plunge depth (micron), TMAZ area (mm2), shear tensile (N), and cross tensile (N). The fuzzy logic model was developed from parameter data FSSW and weld quality FSSW. There are two membership function (MF) inputs, and six membership function (MF) outputs with a triangular membership function type. The number of fuzzy rules refers to the MF input used so that there are nine rules. Furthermore, this fuzzy logic model is optimized for MF input and output with the Genetic Algorithm (GA) method to predict close to the experimental results. The results of the fuzzy logic model show a decrease in the error percentage after the fuzzy logic model is optimized by the GA method. The error percentages for pin diameter, TMAZ area, and cross-tensile parameters were 7.20%, 16.14%, and 20.21%, respectively. The response surface model gives good enough results to predict the input parameters to the output parameters. © 2024 Author(s).
Department of Automotive Engineering Education, Faculty of Engineering, Universitas Negeri Yogyakarta, Yogyakarta, Indonesia; Graduate Student in Mechanial Enginering Departement, Universitas Hasanuddin, Makassar, Indonesia
