Modeling and Sensitivity Analysis of Thermal Conductivity of Ethylene Glycol-Water Based Nanofluids with Alumina Nanoparticles

Abstract

Nanofluids containing alumina nanoparticles have been used in different thermal devices due to their favorable characteristics including ease of synthesis, relatively high stability and proper thermal features. Nanofluids thermal conductivity could be modeled with high exactness by employing intelligent techniques. In the current paper, thermal conductivity of EG-Water-based nanofluids with alumina particles is modeled by utilizing Multi-Layer Perceptron (MLP) and Group Method of Data Handling (GMDH) as two efficient intelligent approaches. In case of utilizing MLP two transfer functions, tangent sigmoid and radial basis functions, are applied. Results showed that utilizing MLP with radial basis provides the highest precision of the prediction in its optimal architecture. R2 of the models by applying MLP with tansig and radial basis functions and GMDH are 0.9998, 0.9998 and 0.9996, respectively. Furthermore, sensitivity analysis reveals that base fluid thermal conductivity has the most significant role in the thermal conductivity of the considered nanofluids.