Prediction of Thermal Conductivity and Heat Capacity of PAN and PEG Copolymers Composite using Artificial Neural Network

In this investigation, the BIOVIA material studio was used to evaluate the PAN/PEG composite. The composite's thermal conductivity and heat capacity were measured, and the root mean square error was predicted using the ANN model in the Synthia module of the BIOVIA application. The composite's thermal conductivity and heat capacity were evaluated throughout a range of polyethylene glycol mass fractions, from 0% to 1%. Compared to the pure material, the composite's thermal conductivity increased by 46%, and its heat capacity increased by 9.4% when PEG made up 1% of the mass fraction. Thermal conductivity and heat capacity data were taken in a dry laboratory and then analyzed using a bi-layer artificial neural network (ANN). The ANN's thermal conductivity and heat capacity estimates held up throughout testing. The implemented ANN modeled Both attributes well, with a Mean Squared Log Error of 1.29 and 14.46, respectively