18 ème Congrès Français de Mécanique Grenoble, 27-31 août 2007 1 Thermal Conductivity of Fluids Containing Quasi-Spherical Nanoinclusions Kliment Hadjov 1 , Mihai Chirtoc 2 & Jean-François Henry 2 1 University of Chemical Technologies and Metallurgy of Sofia, blv.Kl.Ohridski 8, 1756 Sofia, Bulgaria 2 Lab. Thermophysics, UTAP (EA3802), University of Reims, BP 1039, 51687 Reims Cedex 2, France E-mail of the corresponding author: klm@uctm.edu Abstract: The aims of this study is to extend an existing homogenization model in order to predict the thermal conductivity of nanofluids containing non-spherical (ellipsoidal) inclusions, taking into account the interfacial layers effect, the inclusions anisotropy and size distribution. The model predictions have been compared to the Yu-Choi’s model, which is able to consider the above mentioned effects. Comparisons between these models and a model concerning spherical and isotropic inclusions are made. The influence of the interfacial layers conductivity and thickness on the effective one concerning the three models is given. The theoretical predictions are compared with experimental data. Résumé: Le but de cette investigation est de faire une extension d’un modèle existant d’homogénéisation, afin de prédire la conductivité thermique des nanofluides contenants des inclusions d’une sphéricité imparfaite (ellipsoïdale), en tenant compte de l’effet de l’interface, de l’anisotropie et de la répartition des dimensions des inclusions. On a comparé les prédictions de ce modèle au celles du modèle de Yu-Choi qui est en état de tenir compte des effets décrits plus haut. On a comparé les résultats de ces deux mo- dèles à un modèle concernant des inclusions isotropes et parfaitement sphériques. On a montré l’influ- ence de la conductivité et de l’épaisseur de l’interface sur la conductivité effective du nanofluide calculée à la base de ces modèles. Les prédictions théoriques ont été comparées aux données expérimentales. Key-words: nanofluids; thermal conductivity; homogenization 1. Introduction Nanofluids are solid-liquid composite materials consisting of solid nanoinclusions with different shapes suspended in liquid (Choi and al. 2004).Thermal management is becoming one of the key enabling technologies leading to mass production of fuel-cell, electric and hybrid vehicles, which must operate at temperatures significantly lower than conventional internal combustion engines. Therefore, there is a strong need to develop advanced heat transfer fluids with signi- ficantly higher thermal conductivities and improved heat transfer than are presently available and to transfer this technology to the automotive industry. Many homogenization theories have been proposed to obtain the effective thermal conductivity of nanofluids taking into account the thermal behaviours of the phases and the inclusions shape and volume fraction. The aims of this study is to extend an existing homogenization model (Hadjov 2002) in order to predict the thermal conductivity of fluids containing non-spherical (ellipsoidal) inclusions, taking into account the interfacial layers effect, the inclusions anisot- ropy and size distribution. Comparisons between other models are made. The influence of the interfacial layers conductivity and thickness on the effective one concerning different models is given. The theoretical predictions are compared with experimental data. 2. General framework