About this Abstract |
Meeting |
2024 TMS Annual Meeting & Exhibition
|
Symposium
|
Additive Manufacturing: Advanced Characterization with Synchrotron, Neutron, and In Situ Laboratory-scale Techniques III
|
Presentation Title |
AI-powered In-situ Pore Generation and Evolution Dynamics for Laser Powder Bed Fusion Process |
Author(s) |
Sen Liu, Vivek Thampy, Peiyu Quan, Nick Calta, Christopher Tassone |
On-Site Speaker (Planned) |
Sen Liu |
Abstract Scope |
In this talk, I will present the results of real-time X-ray radiography and diffraction of Aluminum alloy during the fast solidification laser powder bed fusion (LPBF) process at the Stanford synchrotron radiation light source (SSRL). With newly developed artificial intelligence (AI) powered tracking algorithms, it found that the porosity generation sources from the powder bed, substrate gas pore nucleation, keyhole tip pore and keyhole stimulated vapor pore. The pore nucleation and movement in the melt pool are quantitatively described with AI-based image segmentation and quantification workflow. The fluid dynamics of the melt pool are reconstructed experimentally with the object tracking of small gas pores. The mechanisms of vapor depression periodic oscillation process for keyhole pore generation process at high laser power are studied. The theoretical physics model to describe the porosity defects nucleation and evolution dynamics is developed, to provide practical guidance for the development of AM printable alloy materials. |
Proceedings Inclusion? |
Planned: |
Keywords |
Additive Manufacturing, Aluminum, Machine Learning |