Optimization of friction stir welding process of AA7075 aluminum alloy to achieve desirable mechanical properties using ANFIS models and simulated annealing algorithm

Current work deals with experimental investigation, modeling, and optimization of friction stir welding process (FSW) to reach desirable mechanical properties of aluminum 7075 plates. Main factors of process were tool pin profile, tool rotary speed, welding speed, and welding axial force. Also, main responses were tensile strength, yield strength, and hardness of welded zone. Four factors and five levels of central composite design have been utilized to minimize the number of experimental observations. Then, adaptive neuro-fuzzy inference systems (ANFIS) have been used to generate mapping relationship between process factors and main response using experimental observations. Afterward, the developed models were applied as objective function to select optimal parameters, in which the process reaches to its desirable mechanical properties by using the simulated annealing algorithm. Results indicated that the tool with square pin profile, rotary speed of 1,400 RPM, welding speed of 1.75 mm/s, and axial force of 7.5 KN resulted in desirable mechanical properties in both cases of single response and multi-response optimization. Also, these solutions have been verified by confirmation tests and FSW process physical behavior. These verifications indicated that both ANFIS model and simulated annealing algorithm are appropriate tools for modeling and optimization of process.