| About this Abstract |
| Meeting |
2023 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2023)
|
| Symposium
|
2023 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2023)
|
| Presentation Title |
A Data-driven Reverse Shape Compensation Method to Reduce Large Deformation in Binder Jet Parts |
| Author(s) |
Basil J. Paudel, Hao Deng, Albert C. To |
| On-Site Speaker (Planned) |
Basil J. Paudel |
| Abstract Scope |
Binder jet parts undergo significant deformation during the sintering, a process that facilitates densification. This sintering distortion may result in parts with unacceptable geometric accuracy. The current work proposes an approach to compensate input geometry based on mechanistic simulations using a data-driven method. A multi-step machine learning approach is proposed for the first time to learn the deformation pattern in binder jetted parts and offset for the sintering deformation. Initial geometries with several reverse scaling factors are simulated using a physics-based constitutive model to generate a training database. Once the training dataset is obtained, a dimension reduction technique is applied to extract the training dataset's features effectively. The model is trained and utilized to predict the compensated part. Finally, the proposed approach's efficacy is validated both numerically and experimentally by comparing the deformed sintered shape against the target. |
| Proceedings Inclusion? |
Definite: Post-meeting proceedings |