About this Abstract |
Meeting |
2022 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2022)
|
Symposium
|
2022 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2022)
|
Presentation Title |
Data-driven Surrogate Model for Laser Powder Bed Fusion Part - Process Design |
Author(s) |
Jannatul Bushra, Hannah D. Budinoff, Md Habibor Rahman, Mohammed Shafae |
On-Site Speaker (Planned) |
Jannatul Bushra |
Abstract Scope |
The quality of parts produced using laser powder bed fusion (LPBF) is affected by numerous input variables such as process parameters and part geometry. Experimentation is frequently used to optimize part quality given a set of parameters and geometry. However, experimentation is time-consuming and expensive, especially early in the design process when part geometry and dimensions are iteratively updated. Physics-based simulations can predict part quality, but these simulations are computationally expensive. We present a data-driven surrogate model to approximate simulation output rapidly. We use calibrated part-scale finite-element LPBF process simulations to generate part distortion training data for a Gaussian process-based surrogate model. The surrogate model enables sensitivity analysis and visualization of relationships between inputs (i.e., volumetric energy density and part geometry) and the simulated output (i.e., maximum part distortion). We quantify time savings from the surrogate model, showing that this approach can speed process planning and the product development cycle. |
Proceedings Inclusion? |
Definite: Post-meeting proceedings |