Large Eddy Simulation and Filtered Density Function

Large Eddy Simulation and Filtered Density Function


Large Eddy Simulation

Large Eddy Simulation has a long history dating back to the infancy of the research efforts in computational climate science by Smagorinsky (1963), Lilly (1967) and Deardorff (1964). Large Eddy Simulation (LES) is a technique that allows for computationally feasible modeling of turbulent fluid motion. The main assumption behind LES is that while large scales of turbulence (the large eddies) are specific to the geometry and boundary conditions, the smallest scales are universal. As such, the smallest scales are problem independent and can be modeled, while the large scales need to be resolved. Also, large scales contain most of the kinetic energy, while the smallest scales, despite controlling dissipation, are responsible for a tiny fraction of the overall turbulent fluctuations. The recent increase in affordability and availability of computing power has allowed to apply LES to the simulation of flow in complex geometries of interest relevant to engineering applications.

Filtered Density Function

In the last decade, LES of turbulent combustion has become computationally affordable. Large-eddy simulation coupled with a Filtered Density Function (FDF) approach has been considered as a promising method to capture the interaction between hydrodynamics and chemical kinetics, making it possible to predict the occurrence of slow chemistry effects, extinction, and possibly reignition. Recently, FDF methods have also been applied to the modeling of particulates (soot) formation in turbulent non-premixed combustion.

The Reactive Flow Modeling Laboratory at KAUST is involved in large scale simulations of non-equilibrium turbulent non-premixed flames using hybrid LES / FDF approaches.


F. Bisetti, J.-Y. Chen, J. H. Chen, E. R. Hawkes, Probability density function treatment of turbulence/chemistry interactions during the ignition of a temperature stratified mixture for application to HCCI engines modeling, Combust. Flame 64:571-584 (2008).