Optimization of parameters of micro-plasma transferred arc additive manufacturing process using real coded genetic algorithm

Micro-plasma transferred arc additive manufacturing (μ-PTAAM) process developed at IIT Indore has proven be an energy and material efficient additive manufacturing process for various meso-scale ALM applications of high melting point metallic materials. This paper reports on optimization of three most important parameters (i.e. micro-plasma power, worktable travel rate and wire feed rate) of μ-PTAAM process by real coded genetic algorithms so as to minimize the aspect ratio (i.e. ratio of deposition width to deposition height) with an overall objective to increase productivity of this process. Objective function for aspect ratio was formulated using generic theoretical thermal developed in terms of μ-PTAAM process parameters and properties of the substrate and deposition material and models developed using regression analysis and artificial neural networks (ANN). It gave optimized values of micro-plasma power as 370, 355 and 360 W, respectively, by the thermal model, regression model and ANN model, and that of travel speed of worktable and wire feed rate as 100 mm/min and as 1700 mm/min by all three models. The optimized results were validated experimentally by depositing 0.3-mm diameter wire of P20 on 5-mm-thick substrate of the same material. The optimized values of the aspect ratio using objective function based generic thermal model, regression model and ANN model are 1.15, 1.31 and 1.36, respectively, with corresponding experimental values being 1.48, 1.5 and 1.48, respectively. Use of optimum process parameters resulted in very good quality and accuracy of the deposition which has excellent bonding with the substrate material and no internal defects.