The Warpage Optimization Of Injection Molding Process Of A Plastic Bearing Stand Based On CAE And BP Network

The quality of injection plastic molded parts relates to precise geometry, surface, strength,durability, and other indicators that are associated with the mold, materials, injection process, andservice environment. The warpage is one of the main defects of injection products. In order tominimize warpage to ensure the precise shape of molded parts, we need to combine design, serviceconditions, process parameters, material properties and other factors in the design andmanufacturing. Finite element software and material database are used to analyze the occurrence ofwarpage, and the results of analysis will contribute to the improvement and optimization ofinjection molding process of typical parts.The warpage from the CAE analysis is significantly affected by the mesh quality and materialproperties. In this paper, Moldflow is used to analyze the injection process of a bearing stand, andwe tested the convergence of the finite element analysis results by refining mesh and checking themesh quality. We find that the mesh quality is the key for the improvement of computing process.By comparing finite element analysis results and measured data we, can obtain confidence inestimating maximum warpage, and the experiment also validated the importance of accurateanalysis. In this paper, we also try using different viscosity models in warpage analysis toaccumulate experience for analysis of the mold design and machining process optimization basedon the finite element method.Based on warpage analysis with Moldflow, we obtained experimental parameters by theorthogonal experimental design method, and processed the test results data through range analysisand ANOVA for the best process parameters. In order to find the optimal process conditions in thesolution space, we use experimental data to establish backpropagation (BP) network for predictingwarpage of the bearing stand. With a proper transfer function and the BP network architecture,results from the BP network method met the demand of accuracy. Finally we search the optimalsolutions in the BP network by the genetic algorithm. Comparing the optimization results, wefound that the optimization method based on the BP network have higher optimization rate.