ABSTRACT
In Resin Transfer Molding (RTM), the preform is placed inside a mold and resin is injected to fill the empty spaces between the fibers. For many advanced resin systems, the mold and preform are heated and the cold resin is injected into the mold. Heat is conducted by the hot mold and preform to the resin and also convected through the preform due to the resin movement during the mold filling. The temperature of the resin will change the viscosity and may also effect the cure-kinetics, hence we need to predict the temperature field during mold filling. Practice has converged on modeling the convection of heat by using the average Darcy velocity and ignoring microscale movement of the resin between the spaces in the preform which is known as heat dispersion. Our work evaluates the importance of heat dispersion within fibrous preforms with an experimental configuration in which we maintained a constant surface temperature of a heated rectangular mold. The cold resin was injected into the heated preform inside the hot mold, and the temperatures were measured along the surface of the mold and along the center line of the preform. A controlled set of experiments were performed to explore the influence of heat dispersion in porous media. The first set of experiments was conducted by injecting cold resin into random preforms, the second set with uni-directional fibers with the flow of resin along the fibers and the final set with the flow across the uni-directional fibers. The measured temperatures within the random fiber preform along the flow direction at the midplane were compared with analytical and numerical results with and without dispersion. In addition, temperature histories of experiments performed with uni-directional preforms oriented along and across 88the fiber direction were compared. Both comparisons clearly show that heat dispersion is very significant and a transverse dispersion coefficient needs to be included into the model to match the experimental steady-state temperatures. A correlation between dispersion in the transverse direction to the flow and the Peclet (Pe) number is proposed for the case of random preforms.
