Processing-microstructure-property relationship of metallic components fabricated by electron beam melting

• Main Author: Ngiam, Chao Xiang
• Other Authors: Zhou Kun
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Engineering; 3D printing; Powder bed fusion; Additive manufacturing; Titanium alloy; Ti-6Al-4V
• Online Access: https://hdl.handle.net/10356/177841
• Institution: Nanyang Technological University

Since its invention, additive manufacturing (AM) has grown to be a versatile and flexible way to manufacture parts, making it a great option for low to mid volume manufacturing, especially in a move towards sustainability. Aside from the commonly used laser powder bed fusion (L-PBF) process, electron beam powder bed fusion (EB-PBF) has also emerged as an alternative, suitable for higher productivity printing. With the expansion of artificial intelligence (AI) in the fourth industrial revolution, AM will also have to adopt the use of AI as well. Incorporating AI with AM will transform manufacturing processes, opening opportunities for higher levels of automation, enhancing customization, and improving productivity levels. Inching towards this goal, huge amounts of data will be required for training purposes in machine learning (ML) algorithms. Hence, this project aims to gather and evaluate the process-property-microstructure relationship of Ti-6Al-4V manufactured via EB-PBF. Speed function, line offset, and layer thickness were varied, generating a total of 36 distinct parameters in this study. This encompassed a wide array of energy density, that enable a comprehensive outlook on the print quality. Subsequently, high porosity samples were further evaluated on their hardness value, which resulted in a range of hardness from 322HV to 391.4HV. Lastly, microstructure analysis was carried out to complete the understanding of the hardness results and also to correlate the relationship between process-property-microstructural relationship of EB-PBF printed Ti-6Al-4V. Data gathered in this project hopes to be used for ML training purposes in the future.