Fernand Marquis

Abstract:

The need for powerful and reliable thermal management systems has increased exponentially in the last two decades in order to sustain the performance of a very wide range of systems. Conventional heat transfer fluids such as water, ethylene glycol, water/ethylene glycol mixtures and lubricating oils are poor heat transfer fluids due to their low thermal conductivity. To overcome this limitation, nanofluids were first developed in 1995 by suspending spherical nanoparticles of metal and metal oxides in heat transfer fluids. It was later observed that these nanofluids showed significant limitations associated with the sphericity and agglomeration of the nanoparticles and lower values of thermal conductivity. Later on in 2001 a new class of nanofluids consisting of colloidal suspensions of carbon nanotubes was developed. Single wall carbon nanotubes have a thermal conductivity in the axial direction up to 6,000W/mK and can be nano engineered into the appropriate nanofluid. This results in carbon nanotube nanofluids with a much higher thermal conductivity then those based on metal and oxide particles, better stability, increased lubricity, good fluidity, non-clogging properties and low chemical reactiveness. The reported values for the thermal conductivity of carbon nanotube nanofluids cover a wide range, depending on the base fluid, nano chemistry, processing routes and temperatures. Typical top range increments can exceed 175% for a 1 vol% load. Higher increments have been achieved at higher loads but with significant increase in the viscosity. The best fluids have good fluidity, no significant settling in stationary mode over a period of several years and no significant separation in dynamic or flowing mode. This paper presents the thermal conductivity of carbon nanotube nanofluids from the nano to the macro scales.