Effects of Process Parameters on the Quality of Metal Additive Manufacturing in Laser Powder Bed Fusion Systems

Abstract

Additive Manufacturing (AM) has revolutionized the manufacturing industry since its emergence in the 1980s. Initially used for rapid prototyping, AM now enables the direct fabrication of complex geometries based on computer-aided design models. Its ability to transition from mass production to mass customization has revolutionized high-tech industries. Despite its smaller market size compared to conventional manufacturing, AM has experienced rapid growth in recent years, with research indicating a significant increase in the number of parts produced, highlighting a growing trend in the global AM market. Among metal-based AM techniques, Laser Powder Bed Fusion (L-PBF) has appeared as a revolutionary technology in metal AM, offering unprecedented capabilities in producing complex geometries and customized components. Despite advancements in L-PBF, challenges persist in achieving high-quality parts that meet stringent industry standards. Critical quality determinants such as surface roughness, porosity, hardness, tensile strength, and ductility significantly impact the operational performance of LPBF-produced parts. The complex interactions between process parameters, and the Associated quality characteristics necessitate systematic studies to optimize the L-PBF process for enhanced part quality. The primary objective of this research is to explore the effects of process parameters on the quality characteristics of AlSi10Mg parts manufactured using L-PBF technique. Explicit objectives include analyzing the impact of process parameters on surface characteristics, evaluating porosity levels and their correlation with mechanical properties, and assessing relationships between process parameters and hardness, microstructural attributes, tensile strength, and ductility of LPBF-produced parts. The research methodologies were designed to achieve the objectives of the study, i.e., examining the effects of process parameters on the surface characteristics, porosity, hardness, tensile strength, and ductility of L-PBF-produced components. For the systematic design of experiments, both the Taguchi method and Response Surface Methodology (RSM) were employed. Statistical analysis and Machine Learning (ML) enabled establishing correlations between process parameters and mechanical properties and enhancing predictive capabilities, respectively. The surface characteristics were studied using quantitative measurements of roughness supported by qualitative analysis via Scanning Electron Microscopy (SEM) images. The research identified common surface defects specific to L-PBF, highlighting the influence of parameters like laser power, layer thickness, scanning speed, and part orientation on defect formation. Advanced SEM image analysis and standardized hardness tests provided insights into how process parameters affect material properties that are crucial for assessing material integrity and mechanical performance in L-PBF components. Tensile strength and ductility were investigated through systematic sample preparation, adhering to ASTM standards, and employing advanced testing methodologies. The research highlights the significant impact of process parameters on parts manufactured using L-PBF. Key factors like laser power, layer thickness, scanning speed and part orientation were found to strongly influence surface roughness. Additionally, scanning patterns such as bidirectional, island, and spiral, along with variations in hatching distance, played a crucial role in affecting the hardness and porosity of the produced parts. These results underscore the importance of precise control over these parameters to optimize hardness, tensile strength, and ductility, ensuring the high quality and reliability of L-PBF components. Machine learning models, including Random Forest Regression (RFR), Support Vector Regression (SVR), and Artificial Neural Networks (ANN), were employed, with RFR performing best in predicting surface roughness and tensile properties, while SVR followed closely. The ANN's predictive performance was limited by the small data set available. In summary, L-PBF has transformed metal AM, enabling the production of complex and customized components. This study underscores the essential role of process parameters in shaping surface quality, porosity, and mechanical properties like hardness and tensile strength. The use of ML enhances process optimization, offering the potential for improved quality outcomes.