About this Abstract |
Meeting |
2024 TMS Annual Meeting & Exhibition
|
Symposium
|
AI/Data Informatics: Computational Model Development, Verification, Validation, and Uncertainty Quantification
|
Presentation Title |
Learning Incremental Forging Policies for Robotic Blacksmithing |
Author(s) |
Michael Groeber, Stephen Niezgoda, Josh Groves, Anahita Khojandi, Glenn Daehn |
On-Site Speaker (Planned) |
Michael Groeber |
Abstract Scope |
The use of advanced incremental forming has been validated by blacksmiths and parts can be made that are much larger than a given available press. Systems with large robots and modestly-sized presses can develop these large forgings and in a fraction of the current time as dies do not need to be designed or built. The goal of this work is to produce components where location-specific material properties/performance metrics are met in addition to the geometry requirements. We will present an initial robotic system - both its cyber and physical components. We will also highlight initial results in training a machine learning system to develop a policy used to control the system to operate in a semi-autonomous manner. |
Proceedings Inclusion? |
Planned: |
Keywords |
Machine Learning, Shaping and Forming, Computational Materials Science & Engineering |