1994 Volume 51 Issue 4 Pages 229-235
The neural network method was applied to estimate theoretical Young's moduli of thermotropic copolyesters constituted from p-hydroxybenzoic acid as a main repeating unit. Theoretical Young's moduli of the copolyesters based on p-hydroxybenzoic acid, m-hydroxybenzoic acid and, 2-hydroxy-6-naphthoic acid were calculated by Treloar's method, and used to determin some parameters for the neural network system. After the neural network system was completed, Young's moduli of other samples were estimated by this system. The values obtained agreed well with calculated one's. The completed neural network system is also able to evaluate the effects of the structure parameters for Young's moduli. The neural network is thus a useful method to evaluate the correlation between the structure parameters and physical properties.