Concept for the evaluation of an ANN-based modeling approach for the Sheet Molding Compound (SMC) compression molding process
Program Director UROP
- +49 241 80-90299
- Send Email
- Project Offer-Number:
- UROP International
- Mechanical Engineering
- Organisation unit:
- Institut für Kunststoffverarbeitung (IKV) in Industrie und Handwerk an der RWTH Aachen
- Language Skills:
- Fluent in English or German
- Computer Skills:
- Not necessary but helpful
- Prof. Dr.-Ing. Christian Hopmann
SMC is a glass fiber reinforced polyester material used mainly in compression molding. During the molding process, different flow fronts occur depending on the insertion position and geometry of the SMC in the mold. The glass fibers in the material align themselves orthogonally to the flow front. Consequently, different insertion positions and geometries result in different fiber orientations in the final part. The relationship can be simulated numerically, but this process is time-consuming. Therefore, optimization based on numerical simulation cannot be implemented in an economical time frame. To solve this problem, a matamodel based on simulation data using artificial neural networks (ANN) was developed. Based on the metamodel, optimization is possible.
The metamodel can be evaluated in several ways. Your task is first to create a literature-based collection on existing procedures. Then you evaluate their potential for the case at hand and select which or which combination of methods should be evaluated in practice. After interpreting the results, you will develop approaches to improve the metamodels.
Study in a natural science department. Basic knowledge of programming is helpful, but not necessary.