Gate Location Optimization Of Plastic Injection Molding Based On Genetic Algorithm With Discrete Variables

Injection molding is the mostly used processing technology for all thermoplastic materials and part of thermosei materials. The process of injection molding is very complicated because of properties of materials. Since long time, mold designing and manufacture and process of injection are dependent on experiences of designers and workers, the molding test is the only way to testify rationality of design, and defects of produce are mainly modified by fixing mold, all of above issues result in long design period , high cost and low level of produce. So the numerical simulation of molding is considered by designers more and more. Now there are many flow simulation software, and by using them structure properties can be forecasted, defects will be found and solved at designing stage, and times of fixing mould can be reduced. However, the numerical simulation only can supply researchers with forecast results of numerical experiment, not reasonable mould structure or technics parameters, the ultimate design plans are completed by lots of computing, comparing, analyzing with experiences and skills of designers. Therefore, combination of numerical simulation and optimization are very necessary, which optimizes relative parameters, it can reduce requirements of experiences of designer and design cost and period.When designing a mould, qualities of produces being good or bad is the ultimate criterion, except for design efficiency and cost. The location of injection gate is one of the most important variables, which effect qualities of produces. The improper gate location can result in overpressure, high shear rate, bad pattern of weld-line and warp, so qualities of produces can be improved by optimizing the location of a gate evidently. This paper introduces an information entropy-based multi-population genetic algorithm with discrete variables, which is combined with the flow numerical simulation program to search the optimum location of the gate. Numerical examples show mat this algorithm has high accuracy and effectiveness for issues of searching gate location.The research work is supported by the Major program (10590354) of NSFC.