About this Abstract |
Meeting |
2019 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2019)
|
Symposium
|
2019 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2019)
|
Presentation Title |
The Digital Twin in Metal Additive Manufacturing – A Paradigm Integrating Modeling, Sensing, and Machine Learning for Defect Prediction. |
Author(s) |
Reza Yavari, Aniruddha Gaikwad, Mohammad Montazeri, Prahalad K. Rao, Kevin Cole, Linkan Bian |
On-Site Speaker (Planned) |
Reza Yavari |
Abstract Scope |
The goal of this work is to prevent occurrence of defects in parts made using metal additive manufacturing (AM) though closed-loop control and in-situ process correction. As a step towards this goal, the objective of this work is to develop and apply a theoretical model-based approach to track the thermal profile of parts made using the directed energy deposition (DED) metal AM process, and subsequently, use the model-derived thermal predictions in conjunction with in-process temperature measurements to detect occurrence of lack-of-fusion flaws. In other words, instantiate the digital twin concept to identify and isolate defects in the DED metal AM process. The central hypothesis of this work is that, a deviation in the observed melt pool temperature from its model-derived counterpart signifies a process drift, indicative of an impending fault. This approach presented herein is demonstrated to predict the occurrence of lack-of-fusion flaws with statistical fidelity approaching 90% F-score. |
Proceedings Inclusion? |
Definite: Post-meeting proceedings |