Experimental study on the dynamic viscosity of hydraulic oil HLP 68-Fe3O4-TiO2-GO ternary hybrid nanofluid and modeling utilizing machine learning technique

Background: Considering the importance of hydraulic oils in various tasks, such as lubrication and cooling, this study evaluated the feasibility of improving the efficiency of hydraulic systems by modifying the thermophysical properties and rheological behavior of base hydraulic oil. Methods: The rheological behavior of the hydraulic oil HLP 68 as a base fluid in the presence of a novel ternary combination of iron oxide (Fe3O4), titanium dioxide (TiO2), and graphene oxide (GO) as nano-additives were evaluated experimentally in a wide range of solid volume fractions (VFs) (0 to 1%), nanomaterial mixing ratios (MRs) (1:1:1, 2:1:1, 1:2:1 and 1:1:2), and temperatures (15 to 65 degrees C). Significant findings: Analysis of changes in dynamic viscosity versus shear rate for all MRs revealed that the THNFs have a Newtonian behavior. It was found that the highest increase in base fluid viscosity in the presence of a 1% VF of GO: Fe3O4: TiO2 is 345%, 1821%, 1763%, and 1990% for MRs of 1:1:1, 1:1:2, 1:2:1, and 2:1:1, respectively, which occurs at a temperature of 15 degrees C. Also, the maximum increase in viscosity with temperature reduction from 65 degrees C to 15 degrees C for the MRs of 1:1:1, 1:1:2, 1:2:1, and 2:1:1 was found to be 66%, 75%, 60%, and 70%, respectively, which occurs at the highest solid VF. In addition, an algorithm for optimizing the structure/training parameters of the subtractive clustering-based ANFIS system as a leading regression technique in machine learning was developed.