
Sciences Naturelles et de l'Ingénieur
Global gyrokinetic electrostatic turbulence simulations usin... ›
Sciences Naturelles et de l'Ingénieur
Local and global gyrokinetic simulations of microturbulence ... ›
Sciences Naturelles et de l'Ingénieur
Numerical Simulation of the Cardiovascular System ›
Sciences Naturelles et de l'Ingénieur
Numerical simulation of the fast ion dynamics in thermonucle... ›
Sciences Naturelles et de l'Ingénieur
Atomic-Scale Investigation of the Defect Levels at Ge and II... ›
Sciences Naturelles et de l'Ingénieur
Application of classical, mixed quantum mechanical/molecular... ›
Sciences Naturelles et de l'Ingénieur
Physico-chemical processes at surfaces and in solution ›
Sciences Naturelles et de l'Ingénieur
Large-Eddy Simulations of High Reynolds Number Incompressibl... ›
Sciences Naturelles et de l'Ingénieur
Large Eddy Simulation of Particle Removal inside a Different... ›
Prof. Dr. Michel Deville Professeur ordinaire EPFL
Sciences Naturelles et de l'Ingénieur
Large Eddy Simulation of Particle Removal inside a Differentially Heated Cavity
Professeur ordinaire EPFL
5 September 2011
In nuclear safety most severe accident scenarios lead to the presence of fission products in aerosol form in the closed containment atmosphere where turbulent convection currents are dominant. It is important to understand the particle depletion process to estimate the risk of a release of radioactivity to the environment. Experiments have shown a higher settling rate under turbulence. In this project we examine the problem by means of large eddy simulation (LES). For LES a low-pass filter is applied to all flow variables so that the filtered velocity and temperature fields can be adequately resolved on a relatively coarse grid. The interaction between the filtered scales (“subgrid scale”) and the resolved scales and the influence of the subgrid scales on the aerosols need to be modeled. As a model for the containment we use the differentially heated cavity problem together with the Boussinesq equations. Last Numerically, the LES equations are discretized by the spectral element method. Particle trajectories are computed using the Lagrangian particle tracking methods, including the relevant forces on the particles (drag, gravity, thermophoresis, lift). The goal of this work is to run massively parallel LES simulations on the BlueGene/P at turbulence levels as high as in the real application.









