About this Abstract |
| Meeting |
2027 TMS Annual Meeting & Exhibition
|
| Symposium
|
Materials and Manufacturing in Space
|
| Presentation Title |
Real-Time Prediction and Inverse Process Design for In-Space Laser Welding |
| Author(s) |
Meysam Faegh, Ehsan Setayesh, Noshin Tasnim Tuli, Fredrick Michael, Andrew O’Connor, Louise Littles, Jeffrey W. Sowards, Tuhin Mukherjee, Tarasankar DebRoy, Azadeh Haghighi |
| On-Site Speaker (Planned) |
Azadeh Haghighi |
| Abstract Scope |
Laser beam welding is a promising technology for the repair and assembly of structures during long-duration space missions, where reliance on replacement components from Earth is logistically constrained. Advancing this capability requires predicting welding behavior under reduced pressure, temperature, and gravity conditions. While experimental studies and high-fidelity numerical simulations provide valuable insight, their cost and computational expense limit their use for rapid exploration and process design. This work presents a framework for real-time prediction and inverse process design of keyhole-mode laser welding under space-relevant conditions. The framework rapidly predicts thermal-fluid behavior and weld geometries across continuous combinations of laser power, travel speed, ambient pressure, and gravitational acceleration. Leveraging this capability, an inverse model identifies process-parameter combinations that achieve target weld dimensions under specified ambient conditions. The proposed framework enables rapid process exploration and design across Earth, Lunar, Martian, and orbital environments, providing a pathway toward digital-twin-enabled in-space manufacturing and repair. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Joining, Machine Learning, Modeling and Simulation |