Mould compensation for thermoset composites parts

Abstract

Samtech, a company whose main activity is the development of a FE software named Samcef, is interested in gaining know-how about mold compensation techniques for thermoset composites parts. For the latter purpose, an internship stage of 4 months has been dedicated to study the subject with Siemens's software NX, which includes Samcef capabilities.

A bibliographical research on the mold compensation techniques has been performed. Several aspects have been studied: the necessity of these compensation methods on the modern industry, the manufacturing processes of thermoset composite materials, the change in properties they experience (curing process), the origin of the geometrical deviations of the produced part from the desired geometry, the available numerical solutions to model the latter phenomena, and the already employed mold compensation techniques available in the literature: CAD-based and mesh-based strategies.

After that, a first numerical model has been developed in order for the identified mold compensation techniques to be implemented. An easy geometry with a predictable deformation has been selected \textbf{--} the deformation mode has been controlled with the composition of the employed laminate on the part. After refinement of the model, the mold compensation techniques have been carried out for the first time and a methodology, i.e. a set of steps and tasks, has been defined for each of them.

With the first simple geometry and laminate, the compensated mold has been found with the different compensation strategies. With these results, a first comparison of the methods has been done. Following that, different geometries with different deformation modes have been studied: the first geometry and a different one have been used with 4 different laminates. The compensation methods have suffered some modifications to adapt the new geometries and deformation modes, and success has not always been achieved. Besides, some restrictive factors have limited the application of some methods to some cases.

The results obtained have been valuable and varied enough to extract thorough conclusions on the strengths and weaknesses of the methods: their robustness, universality, accuracy, total time necessary or complexity have been analyzed. Finally, some proposals on the future continuation of the thesis have been discussed.