Densities and isothermal compressibilities from perturbed hard-dimer-chain equation of state: application to nanofluids

Abstract Densities and isothermal compressibilities of several nanofluids were modelled using a perturbed hard-chain equation of state (EoS) by an attractive term from Yukawa tail in 273–363 K range and pressure up to 45 MPa. The nanofluids of interest comprise TiO2-Anatase (-A), TiO2-Rutile (-R), SnO2, Co3O4, CuO, ZnO, and Al2O3 as nanoparticles dispersed in ethylene glycol, water, poly ethylene glycol, ethylene glycol + water, and poly ethylene glycol + water as base fluids. The EoS was capable of estimating 1397 density data of 9 nanofluids with the overall average absolute deviations (AAD) of 0.90%. The coefficients of isothermal compressibility of 6 selected nanofluids were also predicted using the EoS with the AAD of 5.74% for 1095 data points examined. The PHDC EoS was not capable of estimating the excess volumes of 3 selected EG-, PEG-, and water-based nanofluids accurately as the relative deviations from the literature data were greater than 34%, even though the trend of results against the nanoparticle concentration was in accord with the literature. To further investigate the density prediction, we have trained a neural network with a single hidden layer and 17 neurons which was able to predict the densities of nanofluids accurately.