Correlating physical properties of both polymeric and energetic materials and their organic precursors of polymers using Artificial Neural Networks

Artificial Neural Networks (ANN) are used to predict the statistical properties of polymeric materials. The statistical properties examined are the characteristic ratio and the temperature coefficient of the characteristic ratio. We also examined the physical properties of numerous organic molecules which are used in the synthesis of polymers. ANNs are used to correlate the boiling point (B.P.), melting point (M.P.), refractive index (R.I.), density (D), and dipole moment (D.M.) to the molecular weight of a variety of organic compounds used in the synthesis of polymers. The results demonstrate that all 5 of the physical properties are required to make accurate correlations of the molecular weight. Monte Carlo simulations were used to generate the various statistical properties of the polymers studied. Artificial Neural Networks were then used to correlate the statistical properties of these polymeric materials with the conformational potential energy surfaces generated via ab initio SCFMO calculations. Neural network computing has also been successfully applied for the prediction of density, heat of formation, sensitivity, detonation velocity and Chapman-Jouget pressure of energetic materials. The method is based on automatic determination of structure-property relationships thus allowing quick estimation of desired parameters from structural formulas or composition of explosives. The method does not require any prior knowledge of the nature of these properties, and can be successfully performed even by non-specialists.