Forced and natural convection in aggregate-laden nanofluids

Abstract

A number of experimental and theoretical studies of convective heat transfer in nanofluids (liquid suspensions of nanoparticles, typically with features below 100 nm in size) reveal contrasting results; nanoparticles can either enhance or reduce the convective heat transfer coefficient. These disparate conclusions regarding the influence of nanoparticles on convective heat transfer may arise due to the aggregation of nanoparticles, which is often not considered in studies of nanofluids. Here, we examine theoretically forced and natural convective heat transfer of aggregate-laden nanofluids using Monte Carlo-based models to determine how the aggregate morphology influences the convective heat transfer coefficient. Specifically, in this study, it is first shown that standard heat transfer correlations should apply to nanofluids, and the main influence of the nanoparticles is to alter suspension thermal conductivity, dynamic viscosity, density, specific heat, and thermal expansion coefficient. Aggregated particles in suspension are modeled as quasi-fractal aggregates composed of individual primary particles described by the primary particle radius, number of primary particles, fractal (Hausdorff) dimension, pre-exponential factor, and degree of coalescence between primary particles. A sequential algorithm is used to computationally generate aggregates with prescribed morphological descriptors. Four types of aggregates are considered; spanning the range of aggregate morphologies observed in nanofluids. For each morphological type, the influences of aggregates on nanofluid dynamic viscosity and thermal conductivity are determined via first passage-based Brownian dynamics calculations. It is found that depending on both the material properties of the nanoparticles as well as the nanoparticle morphology, the addition of nanoparticles to a suspension can either increase or decrease both the forced and natural convective heat transfer coefficients, with both a 51% increase and a 32% decrease in the heat transfer coefficient achievable at particle volume fractions of 0.05. This study shows clearly that the influence of particle morphology needs to be accounted for in all studies of heat transfer in nanofluids.