Shear-dependent thermal conductivity of alumina nanofluids

The thermal conductivities (k) of aqueous alumina nanofluids of various particle shapes (rods, bricks, blades) were measured at the dynamic state for the first time. The dynamic k was measured under torsional flows by using a homemade parallel-plate system. The homemade system was validated by numerical simulations and experiments with homogeneous liquids. All the nanofluids tested here showed decreasing k with increasing shear rate. This newly observed phenomenon was named 'shear-reducing thermal conductivity.' The dispersion characteristics were characterized by the dynamic light scattering (DLS) and rheological techniques. From the rheological properties of nanofluids it was inferred that the alumina nanofluids should have network structures and these microstructures should be destroyed or deformed by shearing. But not all the networks were destroyed by shearing. The DLS data revealed that some nanoparticles in nanofluids should exist as individual particles. The effective medium theory cannot explain the shearreducing characteristics of nanofluids at the dynamic state. The rheological data imply that the heat percolation through the network may not be the sole reason for heat transfer enhancement in nanofluids. It is suggested that the Brownian motion of the primary particles cannot be excluded in heat conduction through nanofluids.