Pumping power and heat transfer rate of converging microchannel heat sinks: errors associated with the temperature dependency of nanofluids

Abstract

To find the sensitivity and dependence degree of the numerical simulation predictions on the property variations arising from the temperature gradients, a 3D conjugate heat transfer of Al2O3-water nanofluid convecting through rectangular microchannel heat sinks (MCHS) is considered in the present study. The Koo-Kleinstreuer-Li model is adopted to capture the temperature-dependent nature of thermophysical properties of the working nanofluid compared to the pure fluid (i.e., water). Both straight and width-tapered flow passages are studied using finite volume method within the laminar flow regime to see how sensitive are the predictions to the temperature dependency of the thermophysical properties for both the pure base fluid and nanofluid. Results show that the constant property assumption obtains unrealistic results up to 140% for the Reynolds number, which may mislead in predicting the flow regime (laminar/turbulent). The constant property approach predicts the convection heat transfer coefficient and the pumping power, respectively, 31% lower and 33% higher than those of the temperature-dependent property approach. In addition, the present study concludes that the MCHS should be simulated based on the temperature-dependent thermophysical property approach to be more realistic, especially for converging flow passages due to high-temperature gradients and for nanofluids for their induced temperature-dependent properties. The last two issues induced each other and increase the deviation of the predictions based on the constant property assumption. Finally, because of underestimating the heat transfer rate and overestimating the pumping power, the MCHS would be over-designed if one adopts the constant property assumption for conceptual design and the MCHS would perform under inefficient and off-design conditions.