About this Abstract |
Meeting |
2025 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2025)
|
Symposium
|
2025 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2025)
|
Presentation Title |
Data-Driven Prediction of Geometric Features Influencing Process-Induced Distortion in Laser Powder Bed Fusion |
Author(s) |
Jannatul Bushra, Hannah D. Budinoff, Juan Machado, Pablo Luna Falcon |
On-Site Speaker (Planned) |
Jannatul Bushra |
Abstract Scope |
The complex geometry of additively manufactured parts affects distortion during fabrication, which can lead to build failure. However, predicting distortion across sets of parts with diverse geometric shapes and sizes—such as needed for computational design or early-stage design exploration—through experiments or simulations is expensive. This study proposes a data-driven framework for rapid distortion prediction in laser powder bed fusion (LPBF) process using machine learning techniques. Over 500 parts with varying geometries were utilized, using geometric features (e.g., area, curvature) as model inputs and distortion metrics from thermomechanical finite element analysis simulations as outputs. Our results identify specific geometric features that exhibit strong correlations with high process-induced distortion from LPBF. The resulting model offers a faster and geometry-aware distortion prediction in seconds, compared to hours typically required for simulations. In addition, this interpretable model can provide useful insights to support part design and manufacturing planning. |
Proceedings Inclusion? |
Planned: Post-meeting proceedings |