Nanofluids are engineered by suspending nanoparticles with average sizes below 100 nm in traditional. The ever increasing of thermal loads in such applications requires advanced operational fluid characteristics, for example, high thermal conductivity dielectric oils in transformers and car radiators. These fluids require high thermal conduction, as long as electrical insulation. In the present work the thermophysical and rheological properties of the nanofluids such as thermal conductivity, viscosity and density are obtained from molecular dynamics simulations. These results served as initial data for computational fluid dynamics simulations to calculate heat transfer coefficient. The results show that, adding titanium oxide nanosheet in the base fluid enhanced the thermal conductivity and increased the viscosity and density of the base fluid. The theoretical calculations are confirmed the molecular dynamics simulation results and the simulation methods accuracy. The computational fluid dynamics results show that increasing the amount of titanium oxide nanosheet in the base fluid increases the heat transfer coefficient and increasing ethylene glycol ratio in base fluid leads to lower heat transfer coefficient. Also non-equilibirium molecular dynamics method can use as a effective and accurate method for nanofluids investigation. The coding which used to obtaine the thermal conductivity of nanofluid is a novel and modified type of non-equlibiruim molecular dynamics method. With using this coding the eror persentages of simulations is decreases. The other advantage of this code is reducing the simulation process, becous the molecular dynamics simulations need a long time for processing.