Prediction of Polymer Composite Material Products using Neural Networks

Alexander V. Gaganov, Olya A. Karaeva, Mitar Lutovac, Alexey M. Kudrin, Alexander V. Kretinin, Andrey A. Gurtovoy

This article contains results of mathematical modeling of technological processes for manufacture of pre-production polymer composite prototypes having certain performance characteristics conducted with application of engineering analysis software using artificial neural networks. Based on mechanical testing results, the following material strength characteristics mathematical models were defined: ultimate strength (at room temperature); modulus of elasticity (at room temperature); compressive strength (at room temperature); compressive strength (at Т=150°С); ultimate strength (shear in the sheet’s plane); modulus of elasticity (shear in the sheet’s plane); ultimate strength (interlayer shear). The first step was to evaluate the statistic importance of outside parameters which influence on materials' properties. The main task was to obtain the function of dependence of the mechanical properties from the outside factors.

 

Print