A System Optimization Approach For Plastic Injection Molding

The plastic injection molding is widely applied to produce discrete plastic of complex shape with the highest precision and a low cost. It's known that, in the injection molding design, there're four factors should be considered synthetically, i.e. plastic part design, mold design, molding process parameters selections, and materials selections and so on. In the past, the plastic injection molding optimization was only simply investigated in a certain aspect. There was no synthetic and systematic investigation in view of feasible mold design, mold machining, and manufacture process.Generally speaking, in the filling phase, in order to satisfy the flow balance, the injection velocity, melt temperature, mold temperature, mold gate-runner system, and plastic part structure should be optimized. In the packing phase, the packing profile (packing pressure versus packing time) need to be optimized. And in the cooling phase, there are two kinds of items should be considered, i.e. geometric variables (i.e. cooling circuits location and geometric dimensions) and cooling process parameters (i.e. coolant inlet temperature, coolant inlet velocity, and cooling time etc). In order to reduce the part shrinkage and warpage, the molding process parameters should also be optimized.Therefore, the optimization of plastic injection molding should be considered as a system optimization problem. In each processing phase of injection molding, there're some corresponding mold structure parameters and processing parameters. In order to solve the complicated and obscure system optimization problem, the multilevel decomposition optimization approach and system engineering notion are introduced and the problem are divided into several subsystems.In this paper, for simplicity, only four subsystems were considered (shown in Fig.1), i.e. gate location optimization, packing profile optimization, cooling system optimization, and molding process parameters optimization. Firstly, a gate location design under a certain set of filling process parameters (mold temperature, melt temperature, and injection time) was developed to minimize the mold filling pressure, over-packing, friction heat, and temperature difference during the mold filling process. An information entropy-based multi-population evolutionary algorithm in conjunction with simulation programs was used to search the optimal gate location. Secondly, based on the gate-runner design and the same set of molding process parameters as the first subsystem, the packing profile is optimized by using surrogate optimization method to minimize the part shrinkage due to plastic cooling. Thirdly, based on the previous gate-runner design, packing profile, and molding process parameters, the cooling system are optimized by using decomposition optimization method. The purpose of the system optimization is to obtain the optimum part quality. So the shrinkage and warpage of the final product, which denote the part quality, can be considered as optimization object of the system layer. Therefore, fourthly, the molding process parameters are optimized by minimizing the shrinkage and warpage of the product using black-box optimization methods. Then, the system optimization can be finished.