| 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 |
AI-Driven Process Control for Resin Transfer Molding |
| Author(s) |
Navid Niknafs, Suresh G. Advani, Pavel Simacek |
| On-Site Speaker (Planned) |
Navid Niknafs |
| Abstract Scope |
Resin Transfer Molding (RTM) is a process in which a thermoset polymer is injected into a mold filled with dry fabric reinforcement. The inherent imperfection in fabric placement causes flow disturbances, especially around corners and edges of a complex mold geometry, by providing high permeability channels. The resin will racetrack along these pathways and arrive at the vent before fully impregnating the entire fabric resulting in dry spots within the part which is scrapped due to this defect. Anticipating and addressing this racetracking phenomenon in an automated fashion to ensure full impregnation of the fabric is critical for maintaining quality standards and efficiency in high-volume production settings.
This study proposes a real-time AI-driven process control to identify flow disturbances and provide flow correction strategies during the RTM process. An AI methodology identifies potential flow perturbations, synchronized with flow recognition sensors within the preform, to determine the strategy for activating auxiliary gates to counterbalance the flow disturbances. Locations of gates, vents, and sensors are strategically designed using a mold filling simulation prior to online process control using an optimization methodology based on filling time, quality, and flow recognition capability. Numerical case studies in this work demonstrate that employing this real-time AI-driven process control significantly improves the reliability of the RTM process for high-volume production of automotive components. |
| Proceedings Inclusion? |
Definite: Post-meeting proceedings |