About this Abstract |
| Meeting |
MS&T26: Materials Science & Technology
|
| Symposium
|
Additive Manufacturing Modeling, Simulation, and Machine Learning: Microstructure, Mechanics, and Process
|
| Presentation Title |
Statistics-Based Modeling of Spatter-Related Defects Using High-Throughput CT and Operando Mechanistic Insights |
| Author(s) |
Kyle Mumm, Tao Sun |
| On-Site Speaker (Planned) |
Kyle Mumm |
| Abstract Scope |
In Laser Powder Bed Fusion, the complex interactions between the laser and the powder bed lead to a wide range of defect formation mechanisms. In particular, the spattering process introduces a near-stochastic distribution of defects, leaving components at risk for unexpected mechanical failure. Despite a well-developed understanding of the physics governing the spattering process, spatter-related defect formation cannot yet be effectively described by first-principles equations. Spatter sizes and distributions are stochastic, and the resulting defects are not clearly distinguishable from keyhole or lack-of-fusion defects, especially after laser remelting from subsequent print layers. Therefore, this work discusses a statistics-based approach to quantifying the detrimental additions of spatter-induced defects to the standard defect population. Using high-throughput computed tomography on samples across the build plate, we achieve quantification of spattering-induced defect distribution anisotropy. We guide and bound the model using insights into basic spattering physics from multi-track operando x-ray imaging experiments. |