Contribution to manufacturing control of particle-filled composites by RTM process

This paper proposes a complete model for simulating and controlling particle distribution in the final part manufactured by the resin transfer molding (RTM) process. This model combines (i) a numerical model to simulate the injection of a particle-filled resin through the fibrous medium and (ii) the integration of the genetic optimization algorithm to control the final particle distribution in the composite. Firstly, the numerical model is validated by comparing its numerical results with experimental measurements from the literature. Next, global sensitivity analysis methods were applied to identify the parameters (material and process) with the most significant influence on final particle distribution. As a result, it was found that the injected particle concentration is the most influential parameter compared to the injection pressure and the initial volume fraction of the preform, which has negligible effects. Finally, optimization applications were carried out to test the proposed approach’s robustness. Four different and interesting profiles in the final distribution of particles in the composite were targeted: uniform distribution (UD) and functionally graded (FG) reinforcements (FG-O, FG-X, and FG-NI). The last three profiles are based on functionally graded materials (FGMs) with a substantial property gradient. In all cases, an optimal solution for the evolution of the initial concentration is reached, and the final particle distribution obtained after the injection simulation with this optimized initial concentration leads quite precisely to the desired profile.