Dynamic void content prediction during radial injection in liquid composite molding

Abstract

Abstract In liquid composite manufacturing, a resin is injected through a dry porous media previously placed in a closed mold. To reach aeronautic requirements or for short injection cycles in the automotive industry, entrapped air in the final part before curing has to remain as low as possible. Air entrapment will depend on the fibrous structure, such as fiber content or tow and preform architecture, and on the injection parameters, or more precisely on the fluid pressure and the flow front orientation with respect to the fibrous direction. Void content prediction during unidirectional flow in liquid composite molding processes was presented in a previous article [Gourichon B, Binetruy C, Krawczak P. A new numerical procedure to predict dynamic void content in liquid composite molding. Composites Part A 2006;37(11):1961–9]. In that article, it was shown that microvoids creation, evolution and elimination could be numerically predicted. Good agreement was obtained with experimental measurements of entrapped air volume for 1D flow. Usually, during injection of real parts, flow orientation is not constant, according to the part shape or to the injection scheme employed. The aim of this paper is to evaluate the importance of fiber orientation with respect to the flow front direction by taking as a reference a radial flow where any flow direction can be studied. A bidirectional model was thus developed based on the comparison of the direction of the flow front to the fiber orientation during radial injection by defining an entrapment angle. An experimental analysis of radial flow is conducted in order to analyze the voids creation process regarding fiber orientation by comparing experimental and numerical fluid pressure profiles during injection. Good agreement between numerical and experimental pressure profiles were obtained, thus validating the developed numerical code. However, the change in pressure level due to the presence of microvoids is not significant. In our case, pressure is thus not a reliable parameter to testify for microvoids creation.