Numerical Simulation Of Injection Molding And Multi-objective Optimization

As one of the most important polymer processes, the procedure of injection molding involves heating the thermoplastic into viscous flow state, injecting the melt into the mold cavity under high pressure, packing and cooling to make the plastic into various products with desirable size and shape,in which the process conditions play a key role in determination of the flow state of polymer melt and the final product quality. Since the products from injection molding are widely used in many industries such as automobiles,aviation,consumer electronics and so on,inceasingly higher qualities are required.The thesis mainly focused on the application of numerical simulation and optimization design method in the field of the multi-objective optimization and put forward a set of comprehensive method to optimise the qualities of plastic parts. Orthogonal experiments were carried out with mold temperature,melt temperature,packing pressure,packing time and injecting pressure were selected as the experimental variable,injection time,package and volumetric shrinkage the quality objectives.The simulation results of the experiment were executed with the CAE software Moldflow. The optimal factor level combination and the importance of single process parameter to the target quality variable were obtained. To achieve the target to get a comprehensive optimizational result, Grey Correlation Analysis was introduced.The analysis of mean degree of grey correlation with optimal factor combination was conducted to find the optimal parameters. In comparison with qualities from process parameters recommended, the package and volumetric shrinkage was cut down from 0.2731mm to 0.1915mm and 15.65% to 10.66% respectively. In the end,an artificial network model based BP algorithm with desirable performance was trained,which can predict the quality objectives as the CAE method.The model can execute the calculation rapidly and save a host of time,especially when there were a great deal of orthogonal experiments to be simulated. The set of hybrid optimization were demonstrated to be effective and can be applied in the workshops.