Optimization and prediction of shrinkage of thin-wall injection molded parts by CAE and artificial neural networks.

摘要

Recent trends in the injection molding industry have been moving toward the molding of thinner parts. In addition, modern injection molding process requires the operation of thin-walling without the loss of production or the reduction of the physical properties of the end products. Resins with high flow rates should be carefully chosen for thin-wall injection molding. Among all the possible part defects, the shrinkage problem is one of the major considerations in the post-processing of the product. There are several parameters which affect the shrinkage of the polymer product processing, such as melt temperature, mold temperature, holding pressure, holding time, and injection speed.; In this research shrinkage was analyzed based on part thickness as well as processing parameters, and the most important parameters affecting shrinkage were investigated. For the calculation of shrinkage of the molded parts, the CAE Software called C-MOLD was used. The optimization and prediction of the shrinkage of thin-wall parts that could not be determined experimentally, were achieved by using Artificial Neural Networks whose source program was based on the Backpropagation Network.