Optimization of gate and vent locations for resin infusion processes using genetic algorithms

In composite manufacturing processes, such as RTM and VARTM, resin is infused into a mold containing a preform. The location where the resin is injected is called a gate and the location through which the air is extracted is called a vent. The Liquid Injection Molding Simulation (LIMS), which was developed at the University of Delaware, is capable of simulating the movement of the resin flow fronts through a fiber preform, when the gate and vent locations are specified. When two flow fronts meet and trap air between them, a resin-starved area, called a 'dry spot', is formed, causing a structural weak-ness in the final composite part. This problem can be solved by the optimal placement of gates and vents. In order to determine optimal gate and vent locations, a genetic algorithm has been used here as a search tool. The objective function for the optimization is expressed as a Process Performance Index and incorporates the fill time and dry spot formation. The PPI is at a minimum at the optimum gate and vent(s) configuration. We demonstrate the method through a case study, Pr here less than 1% of the possible LIMS simulations were required to find a near-global optimum.