Molecular Dynamic Simulation of Thermal Conductivity of Electrically Insulating Thermal Nano-Oil

Abstract

In applications such as coolants in electrical devices, in addition to high heat transfer capabilities, the cooling fluids are required to have low electrical conductivity also. As nanoparticle suspensions (nanofluids) show excellent thermal performance due to enhanced thermal conductivity, it would be advantageous to evolve nanofluid-coolants, which are electrically insulating also, for such applications. A theoretical analysis of one such suspension is performed in the present work, to evaluate the thermal conductivity enhancement due to the presence of nanoparticles in the base fluid. The nanofluid analyzed is a suspension of hexagonal boron nitride (h-BN) in mineral oil, for application as a cooling fluid in electrical transformers. The thermal conductivity of the boron nitride suspension is computed using equilibrium Molecular Dynamics (MD) simulations followed by the application of the Green-Kubo auto correlation function. The Lennard–Jones potentials and simple harmonic oscillation potentials are used as the intermolecular potentials to appropriately describe the various atomic and molecular interactions in the boron nitride suspension. The molecular dynamics simulations are performed using LAMMPS software. The computational results are benchmarked with experimental findings on the thermal conductivity enhancement in the suspension at various temperatures and concentrations of nanoparticles, obtained using a transient measurement technique.