Exercising hybrid statistical tools GA-RSM, GA-ANN and GA-ANFIS to optimize FDM process parameters for tensile strength improvement

The properties of functional parts printed by additive manufacturing are highly dependent on various process parameters of the machine. The process parameters can be optimized by hybrid statistical tools to enhance the objective function. The present study investigates the tensile strength of Dogbone American Society for Testing and Materials (ASTM)-D638-V standardized parts fabricated by FDM 3D printer, using poly lactic acid (PLA) plus material. The test specimens were fabricated by varying three parameters: infill density (20–100%), speed (50–150 mm/s) and temperature (190–210 °C). For the parametric combination, response surface methodology (RSM) based central composite design (CCD) matrix was developed using second order polynomial fitting model. The maximum tensile strength of testing specimens on UNITEK-94100 universal testing machine (UTM) was recorded as 45.27 MPa. Further, hybrid optimization techniques like genetic algorithm-artificial neural network (GA-ANN), genetic algorithm-response surface methodology (GA-RSM) and genetic algorithm-adaptive neuro fuzzy interface system (GA-ANFIS) are deployed to optimize these process parameters. Among these tools, the maximum accuracy of 99.89% obtained with GA-ANN which results in optimized parameters as infill density 100%, temperature 210 °C, and speed 124.778 mm/s to achieve the maximum tensile strength of 47.0212 MPa. The results of this examination will facilitate the added substance producing units to choose the optimize factor value of input factors for FDM parts fabrication with improved mechanical properties. The hybrid developed models could be proposed for precise prediction and optimization of other process parameters and results for any industrial application problems.