Optimization Study On Injection Product Warpage Based On ANN (Artificial Neural Network)

With the development of the industry and the wide use of injection molding technology, Plastic products are being used in almost every trade of life. Thus manufacturers are paying more and more attention to the quality of plastic products. Since warpage is one of the most important factors which affect the quality of plastic products, warpage gets more and more attentions from the designers, quality control and customers to guarantee plastic product quality. Due to the fairly complicated factors that result in warpage, in the paper, after some properties and other requests are put forward, according to the principle of rationality in economy and feasibility in technique, then the reasonable processing parameters are adjusted and determined by the way of the combination of orthogonal design and CAE, Artificial Neural Network and the warpage of plastic part is decreased.The research object of this paper is FANBOX, based on analyzing of Moldflow, experimental data and numerical simulation methods, to study affection of gating system and molding process parameters for warpage, aimed to research on the improvement of the quality of plastic products.First, according to the principle of establishing the gating system, different gating systems of frame part are established. Moldflow is then used to do optimum analysis on different molding schemes with different gating systems. Simulation results obtained from different molding scheme is compared in order to get the optimal molding scheme. Consequently the gating system of the part is optimized.Second,molding process parameters which affect the warpage of plastic part are studied. The molding process parameters which need to be studied are established according to the Orthogonal Experiment Method and Variance Analysis. The optimal combination of molding process parameters from the arranged experiments is obtained according to analyze experiment data. The affection on warpage is described. Third, in order to find the optimal combination of molding process parameters from the global range, combine orthogonal design and BP neural network are used in this study. BP neural network is trained and tested by the training sample and testing sample from the experiment data. BP neural network is proved reliable from the testing result. The predicting capability of BP neural network is used to predict the warpage of frame part in other different molding process parameters. The predicting results tell the affection on warpage, caused by interaction of different molding process parameters and single molding process parameters.Lastly, get optimal project based on Moldflow, orthogonal design and BP neural network. And get the summary of topic and further direction.This study shows that the Optimization of gating system and molding process parameters do affect the warpage of frame part. The warpage of the part is minimized according to the optimal results. The quality of frame part is improved by study.