Neural network modeling of thermo-hydraulic attributes and entropy generation of an ecofriendly nanofluid flow inside tubes equipped with novel rotary coaxial double-twisted tape

Abstract

The current paper numerically predicts the convective heat transfer coefficient, pumping power, and total entropy generation of an ecofriendly-functionalized graphene nanoplatelets nanofluid inside the tubes enhanced with a novel rotary coaxial double-twisted tape, which rotates at various rotational speeds. The impacts of the nanoparticle concentration and twisted ratio are considered. The mathematical models of the heat transfer coefficient, pumping power, and total entropy generation rate are obtained as a function of the weight fraction, angular velocity, and twisted ratio based on the numerical outcomes using Artificial Neural Network (ANN). Several configurations of the neural network are examined, and finally, the best model is found having one hidden layer with 7 neurons in this layer, which can provide excellent precision for measuring the outputs. Moreover, the relevant correlations are presented for all three outputs based on the weights and biases of the ANN.