The applications of nanofluids are widely beneficial in heat transmission and cooling systems. Nanofluid viscosity and thermal conductivity have a substantial effect on heat transfer applications and on devices such as solar and geothermal systems. Machine learning models enable faster, less expensive modeling of nanofluid thermophysical properties. These models are secure for future studies and in the development of nanotechnology. In this review, shape, size, temperature, and volume concentration are considered as inputs to develop several machine learning methods, such as artificial neural networks, support vector regression, decision trees, and random forests. These models were analyzed by comparing their R2 values, and the results indicated that machine learning-based models generally exhibited more reliable performance than the other approaches. The observation in this review was that thermal conductivity increases with temperature and volume fractions, whereas viscosity decreases with size, temperature, and volume fractions. To determine the optimal nanoparticle type, size, and concentration for specific applications such as data center cooling and high-heat-flux electronics, future research may employ ML-based optimization techniques.
Jadidi et al. (Fri,) studied this question.