| About this Abstract |
| Meeting |
MS&T23: Materials Science & Technology
|
| Symposium
|
Additive Manufacturing Modeling, Simulation, and Machine Learning: Microstructure, Mechanics, and Process
|
| Presentation Title |
Quantification of Spatter Counts and Trajectories in
Laser Powder Bed Fusion using Machine Learning Analysis of High Speed Imaging
|
| Author(s) |
Christian Gobert, Jack L Beuth |
| On-Site Speaker (Planned) |
Jack L Beuth |
| Abstract Scope |
In laser powder bed fusion additive (L-PBF) manufacturing, key process variables such as laser power, scan speed and hatch distance are varied and the respective impact on porosity content for coupon specimens is observed and reported in process maps. The resulting process maps help end-users identify key process regimes such as keyholing, fully dense, and lack of fusion in parameter space. One potential cause of porosity in L-PBF is spatter, which encompasses ejected material from laser-material interaction. In this work, spatter generation quantified through high-speed video observations of the printing process is correlated to the porosity content in cross-sectioned coupons across process space. Trained machine learning models are used to identify and track spatter in high-speed observations of the printing process. Spatter generation related to key process variables and materials is studied in the aim of identifying optimal process settings to minimize spatter emissions. |