| About this Abstract |
| Meeting |
2024 ASC Technical Conference, US-Japan Joint Symposium, D30 Meeting
|
| Symposium
|
2024 ASC Technical Conference, US-Japan Joint Symposium, D30 Meeting
|
| Presentation Title |
Numerical Methodology for Wrinkling Prediction in Automated Fiber Placement |
| Author(s) |
Rowen Burney, Ben Francis, Alex Brasington, Roudy Wehbe, Ramy Harik, Matthew GodBold |
| On-Site Speaker (Planned) |
Ben Francis |
| Abstract Scope |
In Automated Fiber Placement (AFP), a critical factor to ensuring optimal part quality is controlling defect propagation during layup. By implementing defect prediction models into the process planning phase, optimal manufacturing strategies that require little rework can be rapidly identified and executed. Wrinkling is a common AFP defect that, unlike gaps and overlaps, requires extensive knowledge beyond basic geometrical conditions to predict. Previous work has established a methodology to model wrinkles by using geodesic curvature and established mechanics models from literature. However, this work was accomplished using one complex surface with known geometrical conditions. To extend these efforts to any arbitrary manufacturing scenario, a numerical methodology for wrinkle prediction on generic tow surfaces is proposed here. The tow surfaces are first parameterized into discrete points along both the top and bottom edge. By comparing length mismatch between each segment at the top and bottom, areas that violate the critical strain condition are identified. These areas are then passed into a wrinkle shape algorithm, which utilizes Rodrigues’ rotation matrix and an assumed cosine form to predict wrinkle geometry. Demonstration of the model’s efficiency is presented by generating wrinkle predictions for various layup paths on a doubly curved surface. The presented methodology helps pave the way for complete understanding of AFP defect formation in process planning. |
| Proceedings Inclusion? |
Definite: Post-meeting proceedings |