Heat transfer simulation of nanofluids heat transfer in a helical coil under isothermal boundary conditions using COMSOL multiphysics

Abstract

In the current paper, a numerical model was designed using COMSOL Multiphysics software based on experimental results to predict Nusselt number of TiO2, ZnO and Ag water-based nanofluids in helical coil under isothermal boundary conditions. All Nusselt number data extracted from the model were compared with experimental data to validate the model and the results showed that the model gives good accuracy and the model could be used to predict non-experimental data. Root Mean Square Error (RMSE) was calculated to evaluate model accuracy and the results were 4.35, 2.40 and 2.53 for TiO2/water, Ag/water and ZnO/water respectively. The model was used to predict non-experimental data and the results were logically predicted compared with experimental data and the model was proven to be sufficient for predicting non-experimental data. Model results were compared with two types of artificial neural networks predicting Nusselt number Feed Forward Neural Network (FFNN) and Generalized Regression Neural Network (GRNN). The deviation between predicted data and experimental data was calculated and the results indicated that GRNN network has the highest accuracy with maximum deviations of +0.02% & −0.3% while the COMSOL model has a maximum deviation of +4.5% & −5.9% and finally FFNN network +8.9% & −5.7%.