This article reports on the International Nanofluid Property Benchmark Exercise, or INPBE, in which the thermal conductivity of identical samples of colloidally stable dispersions of nanoparticles or "nanofluids," was measured by over 30 organizations worldwide, using a variety of experimental approaches, including the transient hot wire method, steady-state methods, and optical methods. The nanofluids tested in the exercise were comprised of aqueous and nonaqueous basefluids, metal and metal oxide particles, near-spherical and elongated particles, at low and high particle concentrations. The data analysis reveals that the data from most organizations lie within a relatively narrow band ͑Ϯ10% or less͒ about the sample average with only few outliers. The thermal conductivity of the nanofluids was found to increase with particle concentration and aspect ratio, as expected from classical theory. There are ͑small͒ systematic differences in the absolute values of the nanofluid thermal conductivity among the various experimental approaches; however, such differences tend to disappear when the data are normalized to the measured thermal conductivity of the basefluid. The effective medium theory developed for dispersed particles by Maxwell in 1881 and recently generalized by Nan et al. ͓J. Appl. Phys. 81, 6692 ͑1997͔͒, was found to be in good agreement with the experimental data, suggesting that no anomalous enhancement of thermal conductivity was achieved in the nanofluids tested in this exercise.

Thermal conductivities of two kinds of Au nanoparticles were measured in water and toluene media. The water soluble particles, 10-20 nm in mean diameter, made with citrate stabilization showed thermal conductivity enhancement of 5%-21% in the temperature range of 30-60°C at a loading of 0.000 26 ͑by volume͒. The effect was 7%-14% for Au particles stabilized with a monolayer of octadecanethiol even for a loading of 0.011%. Comparatively lower thermal conductivity enhancement was observed for larger diameter Ag particles for significantly higher loading. Effective enhancement of 9%, even at vanishing concentrations, points to additional factors in the thermal conductivity mechanism in nanofluids. Results also point to important chemical factors such as the need for direct contact of the metal surface with the solvent medium to improve enhancement.

A comprehensive model has been proposed to account for the large enhancement of thermal conductivity in nanofluids and its strong temperature dependence, which the classical Maxwellian theory has been unable to explain. The dependence of thermal conductivity on particle size, concentration, and temperature has been taken care of simultaneously in our treatment. While the geometrical effect of an increase in surface area with a decrease in particle size, rationalized using a stationary particle model, accounts for the conductivity enhancement, a moving particle model developed from the Stokes-Einstein formula explains the temperature effect. Predictions from the combined model agree with the experimentally observed values of conductivity enhancement of nanofluids.

No abstract

No abstract