Search

A numerical method is presented for the estimation of experimental resin transfer molding (RTM) process parameters from in-mold temperature data. The approach is built around the simulation of either radial or axial flow of reactive resins through a fiber preform inside a mold cavity. The simulation considers a non-isothermal environment, with one-dimensional flow and one-dimensional heat transfer, and allows for either pressure or flow rate controlled inlet conditions. Based on the observed temperature history at a selected point in the mold the best fit values of the unknown process parameters are obtained by minimizing the difference between the simulated and experimental temperature histories. This minimization is accomplished by integrating the simulation model with a MINPACK nonlinear least-squares routine (1). As an example, the chemical kinetic parameters for a step-wise polyurethane reaction are estimated based on temperature data reported by Gonzalez (2). Simulated temperature histories using the parameter values obtained agree very well with experimental results.