The filling stage in injection/infusion moulding processes plays a key role in composite manufacturing that can be influenced by the inlet and vent ports. This will affect the production of void-free parts and the desirable process time. Flow control is usually required in experiments to optimise such a stage; however, numerical simulations can be alternatively used to predict manufacturing-induced deficiencies and potentially remove them in the actual experiments. This study uses ANSYS Fluent software to model flow-front advancement during the impregnation of woven fabrics. A developed technique is applied by creating tracking points (e.g., on-line monitor) in the direction of the flow to report/collect data for flow-front positions as a function of time. The study adopts the FVM-VOF-based two-phase flow model together with an implicit time-stepping scheme, i.e., a dual-time formulation solution method with a preconditioned pseudo-time derivative. Initially, three time-step sizes, 5 s (small), 25 s, and 50 s (large), are evaluated to examine their impact on numerical saturation lines at various fabric porosities, 40%, 50%, and 60%, for a two-dimensional (2D) rectangular mould, and predictions are then compared with the well-known analytical Darcy. This is followed by a three-dimensional (3D) curved mould for a fillet L-shaped structure, wherein the degree-of-curvature of fibre preforms is incorporated using a User-Defined Function (UDF) to tailor the impregnation process. The developed approach shows its validation (1–5.7%) with theoretical calculations and experimental data for 2D and 3D cases, respectively. The results also stress that a shorter computational time can be achieved with a large time-step size while maintaining the same level of accuracy.
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)