LAUSR

Abstract In the present work, we use artificial neural network (ANN) approach to develop a tool for prediction of the effective viscoelastic properties - storage and loss moduli - of vinyl ester reinforced with graphite nanoplatelets. Explicit results are obtained in terms of the constituents’ volume fractions, temperature, and loading frequency. The experimental data for ANN training and testing ware obtained using a Dynamic Mechanical Analyzer (DMA) and contains 153 data sets; the training and testing sets consisted of randomly selected 131 and 22 sets, respectively. The good accuracy of the model demonstrates that ANN is efficient for predicting viscoelastic properties in terms of three independent parameters.