Multi-objective Optimization Of Injection Molding Process Based On Integrated Multi-platform Environment

Injection molding has the strength of producing good-quality products in less timeand with high efficiency by multi-cavity production, as well as producing lighter weightedand multicolored products compared with metals. Due to the competition of3C(computer，communication and consumer-electronics) market, the requirements onproduct design become more and more complexed, and higher quality and shorterdevelopment cycle become more strict. The effort taken by trial and error or experiencedworker to design new products will not suit the current market requirements. Hence, it hasbeen one of focused research topics regarding how to develop high-quality plastic productsfaster and more effectively.This thesis presents methodology for multi-objective optimization of plastic injectionmolding process based on an integrated multi-plastform environmentThe thesis mainlyfocused on the application of numerical simulation and optimization design method basedon multiple platforms integration environment in the field of the multi-objectiveoptimization design optimization and put forward a set of comprehensives method tooptimize the qualities of plastic parts., and also demonstrates its applicability. Optimizationtechniques applications can be used in manufacturing process to improve efficiency andproduct quality. The optimization techniques methods included direct search (such asSimple Method), gradient search and heuristic methods (such as GA). First, with theknown maximum packing pressure and total packing time, apply the simplex method forminimal shrinkage by optimizing the packing curve for the injection molding processoptimization. The Moldflow and modeFRONTIER software are employed successfully ina co-simulation way for optimization of packing curve. The optimized shrinkage ofinjection molded parts has been obviously improved.Second, in this thesis, used five process parameters (mold temperature，melttemperature, packing pressure, packing time, and cooling time) to determine the optimalinitial process parameter settings for injection-molded plastic parts with a thin shell featureand under two quality characteristics (warpage and clamp force). In this thesis describesthe Ellipsoidal Basis Function Neural Network model and the Non-dominated Sorting Genetic Algorithm were employed for this optimization. The approximate model wasestablished to optimize the process parameters basis on Moldflow and Isight, as well asusing the Orthogonal experiment method and EBF Network. By using NSGA-Ⅱin theapproximate model, the Pareto front between warpage and clamp force was eventuallyobtained, which provides effective reference to multi-objective optimization of injectionmolding process.