Application of RSM and ANN for the prediction and optimization of thermal conductivity ratio of water based Fe2O3 coated SiC hybrid nanofluid

A novel water-based hybrid nanofluid has been synthesized from Iron oxide (Fe2O3) coated Silicon carbide (SiC) at various mixing ratios (Fe2O3: SiC) to investigate the effect of nanoparticle combination pattern on effective thermophysical properties. In this paper, artificial neural network (ANN) and response surface methodology (RSM) modelling methods are used at an optimized nanoparticle mixing ratio to predict the thermal conductivity ratio of the nanofluid. Further, Analysis of Variance (ANOVA) technique and 3D surface plots are used to determine the significance of all possible interaction parameters on the response variable. Based on the results obtained through RSM, a new correlation has been developed by taking nanoparticle concentration (phi: 0.2-1 vol %), nanofluid temperature (T: 25-45 degrees C), particle size (Dp: 20-60 nm) and ultrasonication time (Ut: 0-40 min) as the independent variables to determine the effective thermal conductivity of the nanofluid. The suggested model performance has been measured in terms of correlation coefficient (R2), mean square error (MSE), root mean square error (RMSE), standard error of prediction (SEP) and average absolute deviation (AAD) to acknowledge the best fit. In this study, both ANN and RSM (R2 = 0.969, AAD = 1.165%) methods are able to predict the thermal conductivity ratio of nanofluid without any assumptions. However, ANN (R2 = 0.9999, AAD = 0.397%) has been more accurate.

Automated summary

A novel water-based hybrid nanofluid was synthesized and the thermal conductivity ratio was predicted using artificial neural network (ANN) and response surface methodology (RSM) with optimized nanoparticle mixing ratio, demonstrating the effectiveness of both methods in predicting the thermal conductivity ratio of the nanofluid.