STUDY OF NATURAL CONVECTION HEAT TRANSFER IN Al2O3/WATER NANOFLUIDS

Abstract

In the forced convection heat transfer, the nanofluids have been proved efficient, theoretically and experimentally. But, a conflict exists between the theoretical and experimental studies on the natural convection heat transfer in nanofluids. The theoretical investigations evidenced an enhancement in heat transfer capabilities of nanofluids in natural convection, whereas, a deterioration was reported in the experimental findings, which are few in number.The above conflict between the theoretical and experimental studies on natural convection is the inspiration of the current study. A square enclosure, heated from the bottom and cooled from the top, identical to Rayleigh-Bénard cavity is filled by the water based Al2O3 nanofluids. First, the water is used as the working fluid to validate the test section, followed by the Al2O3/water nanofluids of different particle concentrations and sizes. The particle concentrations of 0.01 and 0.1 vol.%, and particle diameters of 20 and 40 nm are selected to investigate their effects on the heat transfer. In addition tothe concentrations and sizes of nanoparticles, the aspect ratio (height/width) and heat flux at the bottom plate are other parameters, which are examined in the present study. After the preparation of nanofluids, a stability analysis is conducted to mark the optimum sonication time and to estimate the stability period of nanofluids using the visual inspection method and zeta potential measurement. Further, based on the experimental results reported in the literature, the empirical correlations are formulated to estimate the viscosity and thermal conductivity of nanofluids as a function of concentration and size of nanoparticles and temperature. These thermo-physical properties of nanofluids are used to calculate the non-dimensional parameters, such as Nusselt number, Rayleigh number, Prandtl number, and the thickness of the thermal boundary layer.Additionally, the statistical analysis, consisting of temperature histograms, distribution of root mean square values and profiles of skewness and flatness is supervised on the experimental results to understand the responsible mechanisms behind the heat transfer in nanofluids. II Through the analysis of experimental values, various outcomes related to the alteration of heat transfer in nanofluids are witnessed. The heat transfer in the presence of nanoparticles in the base fluid is enhanced at a particle concentration of 0.01 vol.%, while there is deterioration in the heat transfer at the particle concentration of 0.1 vol.%. The heat transfer is also increased with the Rayleigh number, function of height (aspect ratio) and temperature gradient (heat flux). At a constant temperature gradient, the Nusselt number is increased with the aspect ratio. A generalized correlation is also developed to demonstrate the essence of aspect ratio, particle size and concentration and Rayleigh number on the Nusselt number.It is found that the Rayleigh number is contributed maximum to improve the Nusselt number, while particle concentration is responsible for the deterioration in Nusselt number due to increase in viscosity.The dependency of Nusselt number on the Rayleigh number and Prandtl number for the nanofluids is also formulated, which is a major contribution of the present study, as it is not reported in literature. Above results are furthermoreexperienced by estimating the thermal boundary layer thickness and statistical analysis. A reduction in the boundary layer thickness is observed with the increment in Rayleigh number.On the basis of Rayleigh number, the difference between the soft and hard turbulence is also explained using the root mean square distribution of temperature data. The profiles of skewness and flatness are applied to prove the availability of hot and cold plumes and inhomogeneity in the flow for both distilled water and nanofluids. The various correlations are formulated byrelating the different variables with the non-dimensional parameters and are used to quantify and validate the results of the present study.