IN625 고에너지 직접적층 공정에서의 다중목표 최적화 적용 연구

Abstract

The IN625 nickel-based superalloy, known for its excellent thermal resistance, oxidation stability, and corrosion resistance, presents machining challenges that make additive manufacturing a viable alternative for fabricating complex geometries with reduced material waste. This study validated the multi-objective optimization framework to identify the best combination of processing variables of direct energy deposition (DED) for IN625. 48 experimental trials were performed to establish the database of three processing variables: 200–500 W laser power, 550–1,000 mm/min scan speed, and 2–3 g/min feed rate. Their effects on three characteristic variables (i.e., relative density, hardness, and dimensional accuracy) were adequately interpreted using a regression model based on the quadratic form. Combining this model with NSGA-II algorithm constructed the Pareto front composed of 1,000 optimal parameter sets, which significantly reduced the demands for an extensive number of physical experiments in DED parameter optimization. Furthermore, the TOPSIS method was subsequently applied to refine these solutions, resulting in the selection of five candidates that met the predefined criteria. This study has verified that integrating surrogate modeling with multi-objective optimization significantly decreases experimental costs and development time while effectively determining optimal process conditions for the DED fabrication of IN625.