| About this Abstract |
| Meeting |
2021 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2021)
|
| Symposium
|
Special Session
|
| Presentation Title |
Detecting Defects Caused by Optical Lens and Focus-related Aberrations in Laser Powder Bed Fusion |
| Author(s) |
Aniruddha Gaikwad, Harry DeWinton, Prahalada K. Rao, Paul Hooper |
| On-Site Speaker (Planned) |
Aniruddha Gaikwad |
| Abstract Scope |
The objective of this work is to predict part defects caused by optics-related aberrations, such as variations in the laser focus height in the laser powder bed fusion process. To realize this objective we monitor images of the meltpool acquired from an in-process high-speed video camera. The key idea is to characterize the meltpool shape and spatter features. Various machine learning approaches, including both supervised and unsupervised machine learning approaches are trained to predict the laser focus height based on these features. As part of this objective, we test the hypothesis that pragmatic meltpool features selected based on physical understanding of the process, leads to higher fidelity (F-score) of detecting deviations in Laser Focus Height, as opposed to using so-called black box deep learning approaches. |
| Proceedings Inclusion? |
Definite: Post-meeting proceedings |