Laser-Powder Bed Fusion of AISI 316L-Cu Alloy: AI-Assisted Process Parameter Optimisation, Microstructure and Mechanical Properties Analysis

Metal Additive Manufacturing (AM) has revolutionized the production of complex metal components by enabling the fabrication of intricate geometries with high precision. This technology's potential can be significantly enhanced through the integration of artificial intelligence (AI) methods, particularly Machine Learning (ML), which offers advanced capabilities in establishing complex interrelationships and improving system and product quality control. ML algorithms present a transformative opportunity to address manufacturing challenges, optimize resource consumption, and enhance process efficiency by exploring the intricate linkages between process parameters, material properties, microstructural characteristics, and their resultant properties. This thesis aims to determine the most precise ML algorithm for achieving the process parameters defect detection relationship of AI316L stainless steel alloy components containing 2.5% copper fabricated via the Laser Powder Bed Fusion (L-PBF) method. Recognizing these relationships enables the optimization of process parameters to attain specific objectives. This optimization method facilitates applications that balance productivity and quality, allowing the selection of parameters that satisfy both criteria. By defining these parameter relationships, ML models can be created to predict optimal process parameters based on desired outcomes, such as low defect content, high productivity, and low surface roughness, thus facilitating more efficient and customized AM processes. Following the assessment of multiple models with varying training and testing sizes for the relative density of samples, the Support Vector Regression (SVR) model has been identified as the most effective model. Moreover, the effect of adding Cu has been delineated by equiaxed and columnar grains and cells observed in SEM images. Melt pool dimensions have been analyzed across components with differing process parameters to assess their relationships qualitatively. Finally, printed samples underwent tensile testing to examine the relationship between porosity and mechanical properties. The SEM images of the fractured surfaces reveal both brittle and ductile fractures, with the brittle fracture displaying a quasi-cleavage plane, possibly indicating the melt pool boundary. The microscale analysis reveals ductile fracture characteristics with extensive dimple networks.