Modeling resin flow, preform deformations and residual stresses in RTM manufacturing

Paper i proceeding, 2011

In the present contribution a generic algorithm is developed to simulate resin infusion in a wide range of popular composites manufacturing technologies, such as Liquid Resin Infusion (LRI) and Resin Transfer Molding (RTM). The ultimate goal is to model a complete manufacturing chain, allowing us to predict the final product properties of the composite material. The major challenges to be addressed by in this modelling vary between different processing steps. One of the most important one concerns the migration of the free surface due to resin infiltration into the highly deformable fibrous preform. Considering the LRI process, the modeling challenge is to predict the final shape of a highly deformable preform due to interaction between external loading and the intrinsic fluid pressure. Moreover, in the RTM process the constant compaction load due to rigid top part of the mold needs to be modeled. To resolve both these processes in one single framework, a compressible two-phase porous media formulation is put forward. The developed model involves a fluid compressibility and permeability dependence on the saturation degree. This is to account for the coupled response of partially saturated solid-fluid media, typical for the transition zone at the free surface between full- and non-saturation. The approach automatically monitors the free surface, whereby the monitoring of the resin front migration using e.g. level set or front tracking control is completely avoided. The proposed formulation of the manufacturing infusion has been implemented and used for both LRI and RTM simulations, and numerical results are provided for a hat stringer problem. To complete the manufacturing process chain, the curing step is considered in the present contribution as a simple thermal shrinkage problem, whereby the fiber content obtained from the LRI or RTM–simulations is used for the evaluation of residual stresses and component distortion.