Flow sensing and control strategies to address race-tracking disturbances in resin transfer molding. Part I: design and algorithm development

Abstract

In the resin transfer molding process for advanced polymer composites manufacturing, the fiber preform is placed in the mold cavity and a thermoset resin is injected into the mold to impregnate the stationary preform. The resin displaces the air in the mold through openings called vents. Once the resin emerges out of the vents, the injection is discontinued. The near net-shaped composite part can be demolded after the resin cures. Ideally, the vents should be placed at the locations where the resin arrives last to ensure the complete saturation of the preform. However, the racetracking phenomenon, in which the resin flows faster along the minuscule channels induced by imperfect fits between the preform edges and the mold walls, can dramatically change the resin infiltration process. The ramifications of racetracking are that the resin may arrive at the vents before completely impregnating the preform and create undesired dry spots, which are fiber regions devoid of resin. The racetracking strength is not repeatable and may vary from one injection trial to next. Hence, the online strategic flow control can be useful in improving the processing reliability and the parts quality by re-directing the flow to arrive last at the vents. In this article, an online strategic flow control system consisting of a flow sensing network and a flow actuation network is proposed. A flow pattern recognition technique, which is based on the dimensionless time vector collected by the flow-sensing network, is developed in order to perform the online flow characterization effectively. Flow simulations are utilized to off-line design the flow control system. An evaluation function is formulated to optimize the flow sensing network design. A multi-tier genetic algorithm is implemented to optimize the locations of vents and gates along with the necessary control actions. A numerical case study for testing the computer-generated flow control solutions is presented. It was found that there was significant improvement in the success rate (fewer dry spot regions) due to the use of the strategic flow control and the automated design approach.