Utilization of Machine Learning Methods in Modeling Specific Heat Capacity of Nanofluids

Abstract

Nanofluids are extensively applied in various heat transfer mediums for improving their heat transfer characteristics and hence their performance. Specific heat capacity of nanofluids, as one of the thermophysical properties, performs principal role in heat transfer of thermal mediums utilizing nanofluids. In this regard, different studies have been carried out to investigate the influential factors on nanofluids specific heat. Moreover, several regression models based on correlations or artificial intelligence have been developed for forecasting this property of nanofluids. In the current review paper, influential parameters on the specific heat capacity of nanofluids are introduced. Afterwards, the proposed models for their forecasting and modeling are proposed. According to the reviewed works, concentration and properties of solid structures in addition to temperature affect specific heat capacity to large extent and must be considered as inputs for the models. Moreover, by using other effective factors, the accuracy and comprehensive of the models can be modified. Finally, some suggestions are offered for the upcoming works in the relevant topics.

Specific Heat Capacity Extraction of Soybean Oil/MXene Nanofluids Using Optimized Long Short-Term Memory

“…Their proposed GA-based SVR prediction model beat the MLP in calculating cpNF with higher accuracy than the MLP prediction model. Assad et al [33] reviewed the influence of different parameters in cpNF. They found that properties of solid structures, concentration (wt.%) and temperature have the considerable influence in cpNF [33].…”

“…Assad et al [33] reviewed the influence of different parameters in cpNF. They found that properties of solid structures, concentration (wt.%) and temperature have the considerable influence in cpNF [33]. Zobeiry et al [34] used MLP to solve the heat transfer equation in engineering applications and advanced manufacturing.…”

“…In order to eliminate the time-consuming and expensive practical extraction of the thermophysical properties of the soybean oil/MXene NFs, machine learning methods are used to calculate the requested outcomes. However, the abovereviewed literature [23][24][25][26][27][28][29][30][31][32][33][34][35] shows a lack of comprehensive investigation to extract the most efficient method for calculating the thermophysical properties of soybean oil/MXene nanofluids. This study aims to develop a methodology to extract the efficient machine learning method to calculate the cpNF based on the temperature and nanoflakes concentration.…”