Computational Fluid Dynamics Modelling of Vacuum-Assisted Resin Infusion in Composite Sandwich Panels during Wind Turbine Blade Manufacturing
Topic(s) : Material and Structural Behavior - Simulation & Testing
Co-authors :
Md Tusher MOLLAH (DENMARK), Maksim LARIONOV (DENMARK), Robert PIERCE (DENMARK), Jon SPANGENBERGAbstract :
Vacuum-Assisted Resin Infusion (VARI) is an essential process in the modern composite manufacturing industry, particularly in wind turbine blade manufacturing that requires large-scale production of lightweight and efficient components. Understanding resin flow is a key aspect of this process. Significant areas of wind turbine blades are comprised of composite sandwich panels with a lightweight core having specific grooves and channels, and non-crimp glass fibre skins. The manufacture of these sandwich regions is complex to predict in terms of resin uptake and flow behavior, as there are a large number of parameters to consider. Therefore, this study focuses on developing a Computational Fluid Dynamics (CFD) model to provide a more accurate prediction of resin flow, for the investigation of a wide range of parametric changes. Specifically, the study focuses on small-scale resin infusion in these sandwich panels.
A CFD model is developed in FLOW-3D® to simulate resin flow through the lower reinforcement layers and shallow grooves of composite sandwich panels. The model comprises a non-porous shallow groove in the core material and a porous channel at the top replicating the lower reinforcement layers of non-crimp glass fiber fabric skins, cf. Figure 1-a. The model uses the Volume of Fluid method to track the free surface of the resin, and the Finite Volume Method is used to discretize and solve the continuity and momentum equations. The flow within the porous medium is solved based on the Darcian saturated drag model that adds a drag term to the momentum equations. Time-dependent pressure boundary conditions have been implemented to better replicate experimental conditions. The CFD model has been compared with experiments for different combinations of fabric skins (UD and BIAX) that result in different porosity, permeability, and thickness of the channel. The experimental and simulated results agree relatively well; providing a new venue for investigating the resin infusion performance. Furthermore, the model has been exploited to investigate the influence of material properties (viscosity of resin), glass fabric properties (porosity, permeability, thickness of fabric) and processing properties (groove spacing) on the flow front, volume of fluid, and infusion time. Figure 1-b demonstrates the flow front evaluation and pressure profile for different infusion times.
A CFD model is developed in FLOW-3D® to simulate resin flow through the lower reinforcement layers and shallow grooves of composite sandwich panels. The model comprises a non-porous shallow groove in the core material and a porous channel at the top replicating the lower reinforcement layers of non-crimp glass fiber fabric skins, cf. Figure 1-a. The model uses the Volume of Fluid method to track the free surface of the resin, and the Finite Volume Method is used to discretize and solve the continuity and momentum equations. The flow within the porous medium is solved based on the Darcian saturated drag model that adds a drag term to the momentum equations. Time-dependent pressure boundary conditions have been implemented to better replicate experimental conditions. The CFD model has been compared with experiments for different combinations of fabric skins (UD and BIAX) that result in different porosity, permeability, and thickness of the channel. The experimental and simulated results agree relatively well; providing a new venue for investigating the resin infusion performance. Furthermore, the model has been exploited to investigate the influence of material properties (viscosity of resin), glass fabric properties (porosity, permeability, thickness of fabric) and processing properties (groove spacing) on the flow front, volume of fluid, and infusion time. Figure 1-b demonstrates the flow front evaluation and pressure profile for different infusion times.