A Method for Selecting the Structure of a Polymer Composite Material Using Neural Network Modeling
摘要
Abstract
A method for determining radial stresses in polymer composite materials based on continuous fibrous fillers and an epoxy matrix is proposed. Using a specific composite as an example, the optimal distance between elementary fibers has been determined based on the criterion of minimizing radial stresses. To solve the problem of selecting a rational structure for a polymer composite material, a neural network architecture has been developed, and the content of each of its stages has been examined.