About this Abstract |
Meeting |
2021 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2021)
|
Symposium
|
Physical Modeling
|
Presentation Title |
Non-orthogonal Adiabatic Boundaries in Semi-analytical Laser Powder Bed Fusion Simulations Using Machine Learning |
Author(s) |
Christian Gobert, Evan Diewald, Nicholas Jones, Jack Beuth |
On-Site Speaker (Planned) |
Christian Gobert |
Abstract Scope |
Simulating temperature fields for laser powder bed fusion (L-PBF) additive manufacturing (AM) processes through Finite Element Analysis (FEA) is computationally expensive. Semi-analytical approaches for simulating L-PBF can achieve drastic decreases in computation time compared to FEA, however lack the ability to incorporate complex adiabatic boundaries. Identical simulations were run in semi-analytical and FEA environments across multiple process conditions including variation in laser power, laser speed, scan strategy and geometry. A convolutional neural network was then trained to perform heat diffusion on adiabatic boundaries in the semi-analytical environment, using FEA simulations as ground truth. |
Proceedings Inclusion? |
Definite: Post-meeting proceedings |