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 |
Real-Time Prediction of Working Distance in Laser-Wire DED Via Multi-Modal In-Situ Data Fusion Using a CNN Model |
Author(s) |
Seonghun Ji, Hyub Lee, Seunghwan Lee |
On-Site Speaker (Planned) |
Seonghun Ji |
Abstract Scope |
Laser-Wire Directed Energy Deposition (LW-DED) offers high deposition rates, precise geometric control, and minimal heat-affected zones due to its focused laser beam and melt pool control, making it ideal for complex metal part fabrication. However, during prolonged deposition, heat accumulation disrupts the balance between wire feed rate and heat input, leading to bead geometry deviations and variations in working distance. This alters the laser spot size and beam intensity, reinforcing process instability through cumulative feedback. This study introduces a real-time monitoring and prediction framework for estimating working distance using multi-modal in-situ data, including melt pool imaging via coaxial camera and thermal emission measurements via two-color pyrometer. These synchronized data streams are processed by a convolutional neural network (CNN) model trained to infer trends in working distance under varying conditions. The approach enables predictive compensation strategies to maintain deposition stability amid geometric and thermal variation. |
Proceedings Inclusion? |
Planned: Post-meeting proceedings |