Prediction of Physical Properties for Plastic Automotive Parts Using Polymer Analysis

摘要

In this paper, tensile, flexural and impact strength properties of polyolefin with respect to talc filler content were predicted using neural network model. Talc content, tensile test speed, thermal properties and rheology data were used as modeling input factors. The models were compared quantitatively by average error rate. The neural network model results were determined as the most meaningful with a high reliability. And, we can be expanded and applied to other materials.