Cure Optimization in Pultrusion Process by a Heuristic-Analytical Approach

This paper deals with a hybrid optimization of the degree of cure and temperature field into the work piece during the pultrusion manufacturing process. The procedure is based on the consecutive application of a heuristic technique (Genetic Algorithms), with an analytical method (simplex technique), to minimize an opportunely defined fitness function. The objective function measures the uniformity of the degree of cure, with a satisfactory mean value, in the exit cross section of the forming die, taking into account constraints related to the degradation of the processing resin system, due to an exothermic peak temperature major than the carbonisation temperature of the polymeric matrix. Optimal values of the temperatures of the die heating platens have been evaluated by applying the proposed procedure to a three dimensional thermo-chemical numerical model, solved using a finite difference scheme. The robustness of the developed optimization procedure has been tested using several combinations of process parameters, such as temperature of the resin bath, temperature of the die cooler, and composite material pull speed. A finite element model is used to analyze temperature field and degree of cure profiles after the optimization of the temperatures of the heating platens. The proposed procedure can be used as a good tool to optimize part quality or productivity, for the conventional pultrusion process, as well as, for post die shaping pultrusion, in which the processing part is completely formed out of the heating die and material cure is then completed using U.V. rays or other heating sources.