Predicting thermal conductivity and dynamic viscosity of nanofluid by employment of Support Vector Machines: A review

Abstract

Nanofluids are employed in variety of thermal energy systems owing to their modified properties and ability of enhancement in heat transfer. Among the properties of nanofluids, thermal conductivity and dynamic viscosity have significant role in the heat transfer characteristics and fluid flow specifications. Among the intelligent techniques, Support Vector Machines (SVMs) are attractive choices regarding their significant exactness in prediction and modeling. In this article, research works on the forecasting and modeling of the mentioned properties of nanofluids with SVMs are reviewed and their outcomes are presented. In accordance with the results of the reviewed works it can be stated that the developed approaches based on the SVM technique have shown significant performance in accurate modeling of both thermal conductivity and dynamic viscosity and can have superior performance in comparison with other intelligent methods in some cases. Furthermore, it can be pointed out that the performance of these systems is dependent on some elements like the employed optimization algorithms. Some recommendations are represented for future studies to reach models with more favorable performance.