Modeling thermal conductivity and collective effects in a simple nanofluid

Abstract

Molecular dynamics simulations are used to model the thermal properties of a fluid containing solid nanoparticles (nanofluid). The flexibility of molecular simulation allows us to consider the effects of particle mass, particle-particle and particle-fluid interaction and that of the spatial distribution of the particles on the thermal conductivity. We show that the heat conductivity of a well dispersed nanofluid is well described by the classical Maxwell Garnett equation model. In the case of particle clustering and strong inter particle interactions the conductivity can be again described by effective medium calculation taking into account the aspect ratio of the cluster. Heat transfer is increased when particles are aligned in the direction of the temperature gradient. This kind of collective effects could be a first step to understand the substantial increase in the conductivity observed in some experiments.