Context
In order to increase the efficiency of next-generation solar tower power plants, the operating temperature of the solar receiver must be increased to around 1000°C. Current transfer fluids cannot withstand these temperature levels, so new alternatives must be found. One way of achieving this goal is to use a gas-particle mixture as the transfer fluid within the solar receiver. Controlling gas-particle flows in future solar tower power plants is a major scientific challenge. Couplings between agitation, the two-phase nature of the flow and temperature make the physics particularly complex. This PhD thesis aims to improve the understanding and modeling of parietal heat transfer in these flow configurations.

Objectives
The objectives of the PhD project are as follows (presented in chronological order):
1. Development of the thermal part of the fluid-particle numerical simulation method using the Front-Tracking method of the TrioCFD software and a DEM approach.
2. Realization and analysis of anisothermal numerical simulations of dense gas/particle flows.
3. Development of heat transfer models between the solar receiver wall and the gas/particle mixture. 

Method
In this PhD thesis, we will use the Front-Tracking method of the TrioCFD software to take into account the two-phase nature of the flow. This uses a moving surface mesh that explicitly represents the interfaces. It can therefore accurately describe any particle geometry and its interaction with the surrounding fluid. Recently, this method, originally developed for liquid/gas flows, has been adapted for fluid/particle flows (see [1] and [2]). Particle/particle interactions are modeled by Soft Sphere Collision Laws (SSCL), and the non-deformable nature of the particles is achieved by penalizing them with the viscosity of the solid phase.  
To simulate dense fluid/particle flows, by explicitly representing the particles, we use high-performance computing and several hundred or even thousands of processors. According to the literature, direct numerical simulation (DDS) of fluidized beds requires mesh sizes of at least forty meshes per particle diameter, in order to properly capture the viscous sublayer (friction) and the conductive sublayer (fluid/solid heat transfer). These simulations are therefore extremely costly numerically. However, to take into account the collective effects prevalent in this type of flow - and thus get closer to solar applications - it is essential to carry out simulations with more than 10,000 particles. The aim of this PhD thesis is to find the resolution that strikes the right balance between computational cost, accuracy and representativeness of flows in high-temperature solar receivers. 
To this end, the PhD student will carry out a mesh sensitivity study and a parametric study on the size of the domain and the number of particles simulated. He/she will build up an extensive database of anisothermal fluidized bed simulations resolved on a scale smaller than particle diameter. He/she will physically analyze the results obtained. Particular attention will be paid to parietal heat transfer.  For example, the part of the flow exchanged with the particles and that exchanged with the gas will be evaluated. To do this, he/she will look at averages, standard deviations, Fourier transforms and probability densities of these heat fluxes as a function of flow properties.
These analyses will enable the development of heat transfer models between the wall of the solar receiver and the gas/particle mixture. A first step could be the development of a correlation for the mean wall to bed heat flux. However, the aim of the PhD work is more ambitious, with the development, by upscaling, of a model that can be used in the Euler-Euler approach (two-fluid model - TFM). The challenge is to estimate the instantaneous local wall to bed heat flux from data such as solid volume fraction, gas and dispersed phase velocities and temperatures, and fluid and particle agitation.