Laser powder bed fusion of atomized industrial waste-derived Inconel 725 alloy powders: A machine learning-assisted process optimization

Among nickel-based superalloys, Inconel® 725 (IN725) stands out for its excellent strength and corrosion resistance. Despite this, its application in additive manufacturing remains largely unexplored. This study investigates laser powder bed fusion of metals (PBF-LB/M) applied to IN725 powder derived from recycled industrial waste, addressing sustainability and process optimization goals. Using the design of experiments approach, the laser power–scan speed process parameter space was explored. Gaussian process regression models were developed to predict surface roughness, relative density, and microhardness. Both direct process parameters and volumetric energy density were evaluated as model inputs to assess predictive performance. The findings established a broad optimal process window for manufacturing high-quality IN725 parts using PBF-LB/M. Specifically, an optimal combination of 99.99% relative density, 7.3 μm roughness, and 311 HV microhardness was achieved by processing the powder at 250 W and 1,500 mm/s. By demonstrating the feasibility of using recycled IN725 powder, this study contributes to the development of sustainable manufacturing practices and supports wider adoption of PBF-LB/M in oil and gas, marine, and chemical processing industries, where IN725 is widely employed.