Abo Bibliothek: Guest
TSFP DL Home Archive Vorstand


Inga Mahle
Fachgebiet Stroemungsmechanik, Technische Universitaet Muenchen Boltzmannstr. 15, 85748 Garching, Germany

Joern L. Sesterhenn
Fachgebiet Stromungsmechanik, Technische Universitat Munchen, Boltzmannstr. 15, D-85748 Garching; Department of Numerical Mathematics (LRT1), Universitat der Bundeswehr (UniBw) Munchen D-85577 Munich, Germany

Rainer Friedrich
Lehrstuhl für Aerodynamik, Technische Universität München Boltzmannstrasse 15, D-85748 Garching, Germany

Alexandre Ern
CERMICS, ENPC 6 et 8, av. Biaise Pascal, 77455 Marne la Vall&3233;e cedex 2, France


Besides a simplified treatment of chemistry, many existing turbulent combustion models also use simplified diffusion processes (Hilbert et al., 2004). Differential and thermal diffusion effects are only included in recent DNS of turbulent combustion, for example in de Charentenay and Ern (2002) and show mainly local influence on the flame structure. Since for non-premixed test cases exact modeling of the mixing process is a prerequisite of correct combustion predictions, the goal of our work is to further investigate the effects of detailed diffusion, modelled at different levels of precision, on turbulent mixing in the non-reacting case.
Motivated by this, DNS of temporally evolving, turbulent compressible shear layers with gradients of species and temperature have been performed.
The species are called active scalars because they influence the flow via the density due to their different molecular weights and via the transport coefficients like heat conductivity and diffusion coefficients.
Two different levels of approximation for the species diffusion fluxes and the heat flux are used and their effects are investigated. A quantity of special interest in this paper is the scalar dissipation rate as it is directly related to the reaction rate in combustion and therefore important for combustion modeling, for example in LES.
The first section of the paper describes the configurations and the initial parameters of the simulations. Then, the Navier-Stokes and species transport equations including detailed or simplified diffusion are presented. A self-similar state from which mean profiles and statistics are taken is defined in the following section. Next, mean profiles of diffusion flux and heat flux are analyzed. With the help of the diffusion flux, a mean Schmidt number is defined and its profile is evaluated which allows to assess the approximation of a spatially constant Schmidt number for each species. After this, the influence of the diffusion description on mean profiles, instantaneous fields and pdfs of the scalar dissipation rate or related quantities is investigated. Finally, conclusions are drawn.