Vacuum- assisted resin transfer molding (VaRTM) is emerging as one of the most robust alternatives for the autoclave process. VaRTM applies the resin infusion method in a vacuum environment. It generally uses reinforcement and a polymer matrix separately in the process. VaRTM is influenced mainly by the characteristics of the constituent materials, such as preform permeability and resin viscosity. Among these, the process design involving the arrangement of the resin inlet/outlet line significantly influences the process performance. Furthermore, an inappropriate inlet/outlet layout cause voids, porosities, or resin-starved areas, which has a decisive effect on quality degradation. In a typical VaRTM process, the inlet and outlet lines are used as a runner for the resin to flow. Furthermore, an equal length of line is used for uniform infiltration and outflow rate during infusion. In addition, excess resin is infused into the preform to reduce the risk of unsaturated zone after the infusion process. This prolongs the infusion process and results in excessive consumption and wastage of resin. In this study, a highly curved and twisted winglet spar structure was used as the test specimen. Its complex and asymmetrical shape caused non-uniform impregnation, which results in local unsaturated zone during infusion. Therefore, the adjustment of the inlet/outlet line length was attempted for flow simulation and experiment. The permeability, which is the most critical parameter for flow simulation, was predicted theoretically by incorporating an experimental constant. Furthermore, it was verified through a comparison with a reference experimental measurement and was used for flow simulation. In addition, the geometrical ply wrinkle on the outer flange of the spar preform and cross-sectional ply wrinkle on the corner area caused by the local compression load over the preform material were investigated. Local slitting on the 0° layer and the rotated rosette method were experimentally verified and proposed. Moreover, the effect of the draping membrane material on single diaphragm forming was studied experimentally. It was verified that during single diaphragm forming, the draping membrane should have sufficient stiffness and plasticity to maintain a uniform draping pressure and constant rate. The infusion process for an asymmetrical complex-shaped winglet spar structure was simulated using PAM-RTM software. Nine cases with the adjustment of the inlet/outlet line length proportionally selected were attempted for simulation. The most optimized simulation case of the winglet spar was determined based on the impregnation behavior, injected volume, and lost volume calculated by flow simulation. Furthermore, the simulation results were compared and verified with the results of three actual test part experiments. There was a good agreement between the simulation and experimental results. The simulation result helped predict the resin impregnation behavior and defect during infusion. It was verified that the unsaturated zone was altered and improved by controlling the resin infiltration rate by adjusting the inlet/outlet line length. It was verified that the flow simulation could be used effectively to reduce the need for the trial-and-error method in infusion process design and to optimize the infusion process window for full impregnation of preform.