For several applications, granular media deals with complex physical phenomena. The multi-scale aspect of their microstructure they contains makes the prediction and modeling of heat transfer and its properties non-trivial to compute. In the case of ceramic powders, highly polydisperse powders with grain size extending over several orders of magnitude, are commonly employed. The different pore sizes and the large amount of heat transfer surfaces make the evaluation and simulation of the thermal properties of powders complex to compute.

Homogenized empirical laws are commonly used for this purpose. They allow fast and cheap computation of equivalent properties but rely on a number of empirical parameters which limit their application domain. Explicit simulation tools like DEM/FFT method [1], provide a more detailed description of the microstructure of the granular shape and packing, with higher computational costs. We use this method to tackle the models and better understand the competition between the different modes of heat transfer in the packed beds (heat conduction in gas, heat conduction in grains, conductance at contact between grains, radiation, etc.).

In order to test and improve the simulations and the models, a previous phD thesis was carried out on the effect of grain size and atmosphere on the equivalent thermal conductivity [2]. It gave a better understanding of the competition between heat transfers of different porosity scales and the effect of gas penetration in the microstructure, and to the proposal of new equivalent thermal conductivity models.

The aim of this thesis consists in pursuing this work to study the influence of grain size distribution on the conductivity of the packed bed. It will involve experimental measurements to acquire reliable and controlled data, and a simulation/modeling work to better understand and model the thermal properties of these media. The thesis will involve a collaboration between the Departement of Nuclear Fuels (IRESNE institute, CEA Cadarache) and IUSTI institute in Marseille. The experimental part will take place at IUSTI in Marseille, and the simulations in CEA Cadarache. Specific attention will concern the analysis of measurement and simulation uncertainties.

This subject has applications to many industrial fields such as energy production and transformation, heat exchangers and process engineering. It will gives to the candidate skills at the end of the thesis which might be of a great interest whether in industry or in academic research.

[1] Calvet, T., Vanson, J. M., & Masson, R. (2022). A DEM/FFT approach to simulate the effective thermal conductivity of granular media. International Journal of Thermal Sciences, 172, 107339.

[2] Letessier, J., Gheribi, A. E., Vanson, J. M., Duguay, C., Rigollet, F., Ehret, N., ... & Gardarein, J. L. (2023). Thermal transport-porosity-microstructural characteristics: unpicking the relationship in ultra-porous a-Al2O3 powder. International Journal of Heat and Mass Transfer, 205, 123898.