Prediction of thermal stability, crystallinity and thermomechanical properties of poly(ethylene oxide)/clay nanocomposites with artificial neural networks

• We model the effect of clay % on the properties of PEO nanocomposites with ANNs.
• Thermal stability increases with decreasing crystallinity, and increasing modulus.
• The prediction ability of TGA and DSC modeling is much lower compared to DMA.
• The modeling success is affected by the complex relation between input and output.
• The ANN technique is a useful mathematical tool in the thermal analysis of polymers.

The artificial neural network (ANN) technique with a feed-forward back propagation algorithm was used to examine the effect of clay composition and temperature on thermal stability, crystallinity and thermomechanical properties of poly(ethylene oxide)/clay nanocomposites. Based on dynamic mechanical analysis (DMA), differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) experiments, values of decomposition temperature, char yield, enthalpy of melting, storage modulus (E′) and tan δ were successfully calculated by well-trained ANNs. The simulated data is in very good agreement with the experimental data. ANN results confirm that thermal stability of PEO nanocomposites increases with the decrease of enthalpy of melting and relative crystallinity, and there is a directly proportional relationship between the modulus (stiffness) and thermal stability. The ANN technique is confirmed to be a useful mathematical tool in the thermal analysis of polymer/clay nanocomposites.

Figure optionsDownload as PowerPoint slide