Numerical Simulation And Sensitivity Analysis Of Reactive Resin Transfer Molding Processes

Recently, Resin Transfer Molding(RTM) is a fast developing technology for manufacturing fiber reinforced composite, which involves that the controlled fabrics are preplaced in the mold cavity, and then the resin with low viscosity is injected under a proper injection pressure, following by a curing stage. As advantages of low cost and high production efficiency, RTM Processes has been applied more and more extensively in many fields, such as aerospace, automotive and architecture industries.The resin filling process is the main step in RTM processes. Many factors, such as mold structure, characteristics of fiber and resin, manner of heating, edge effect, resin curing reaction, the processing parameters and so on, will influence the performance and quality of products. As a result, considering the various influencing factors, the mechanism and the dynamic process of resin infiltration should be investigated deeply, and then the defect control system of the advanced composites is established. Consequently, the performance of advanced composites can be improved and the production cost is reduced.In this paper, the resin flow process is researched, using multi-disciplines, the functional relationships among the materials systems, processing conditions, material microstructure and flow behavior are constructed and analyzed. The equations fulfilled by the different chemical and physical field variables were solved by an uncoupled algorithm. And then, the core programs for the numerical simulation of isothermal/non-isothermal RTM filling process is developed. The sensitivity analysis method is introduced, in the view of quantitatively analyzing the influencing law and degree of the processing parameters and materials properties on the resin flow patterns.The main work and conclusion were as follows:For the case of fiber preform that exhibit characteristics of dual-scale porous medium, the volume averaged method is introduced to deduce the governing equations in the preform, and the momentum equations contain the inertia and viscous terms, and its application is more extensive than that of the Darcy law. Based on the governing equations of fully developed flow in a rectangular duct and the formulation of the equivalent permeability, the modified Navier-Stokes equations are adopted to describe the resin flow in the edge channel, which considering the effect of the mold thickness on the flow patterns. Considering the different curing reaction mechanism, stepwise polymerizations and chain addition polymerizations, the weight-average molecular weight of resin is computed by employing the probability theory and the recursive nature of polymers. The above governing equations are integrated with the curing reaction kinetic equation and the chemorheological equation to model the resin infiltration process.In order to reduce the computational complexity, the single approach is adopted, therefore, the resin flow behavior in preform/edge area can be described by one set of general governing equations, and it can avoid giving the explicit formulation of the boundary conditions at these two areas. Considering the interaction between resin fluid and air fluid, the fluid filling process are treated as two-phase flow, this method can avoid giving the explicit formulation of resin flow front. The finite volume method is applied to deduce the discrete expressions of the related governing equations. The staggered grid and the SIMPLE algorithm are introduced to compute the pressure and velocity coupling. The direct numerical integration and backward difference method are used to discretize the curing kinetic model. The Volume of Fluid/Piecewise Linear Interface Construction approach is implemented to track the resin flow front advancement during the resin flow process. The above method is the critical technique for applying the numerical simulation of the RTM filling process.The core programs for the numerical simulation of RTM filling process is developed. The phenomenological model and a viscosity model based on the branching theory are used to describe the chemorheological behavior of BMI resin and Epoxy resin during the resin flow process, respectively. And then, the evolution of the curing degree, molecular weight, resin viscosity and flow patterns is simulated. And effects of curing reaction, fluid temperature, injection flow rate and permeability on flow patterns and distribution of curing degree and molecular weight in the mold cavity are analyzed. Some significant resluts are obtained.In order to qualitative and quantitative explore the effect of processing parameters and material properties on the resin flow infiltration, the sensitivity analysis method is introduced, the mathematical models of sensitivity equations for the related physical quantities are deduced, the sensitivity equation of the resin flow front shape is established, and the coupling solution method for the sensitivity equations is determined. Numerical simulations of the resin flow process under constant flow rate injection condition and under constant pressure injection condition are carried out separately. And then the influencing law and degree of the resin temperature, flow rate, injection pressure, fiber permeability and the permeability in the edge area on the filling time and resin flow front are investigated. The simulated results provide valuable suggestions for improving the flow efficiency and products quality.The relationship between the resin flow patterns and temperature field are analyzed. Considering the heat conduction in the thickness direction, the coupling resolution method is established for the 2-D flow field/3-D temperature field. The discrete expression of the 3-D energy equations in the time domain and space domain is deduced. And then the program codes for the non-isothermal resin flow process are complied. The evolution of temperature, curing degree and resin viscosity are analyzed, and the effects of edge effect and thermal conductivity on the distribution of related physical quantities are investigated. The studies show that the edge effect will aggravate a heterogeneous distribution of related physical quantities in the mold cavity.