Small scale asymptotic models for large scale simulations of turbulent premixed flames

Replacing a finite thickness premixed flame by its asymptotic limit of a zero thickness interface (flamelet) separating burnt and unburnt gas is a rigorous procedure for laminar flows. However, most flows of interest are turbulent, in which case the assumptions for the flamelet asymptotic model are not quite satisfied. Still, it is computationally convenient to retain the laminar flamelet representation to compute the large scales of a turbulent premixed flame and to model the effects of the small scales (unresolved in any practical computation). In this talk, I will review the basic asymptotic laminar flamelet formulation and also recent theoretical results for the turbulent case. Some of the results have somewhat surprising and counter-intuitive consequences with regards to practical large scale flamelet codes. In particular, it turns out that turbulence models based on the laminar flamelet assumption can be systematically biased towards over-predicting the turbulence effects. As an alternative, I will present a new numerical strategy based on a rigorous asymptotic model to parameterize the effective subgrid burning speed for a variety of small-scale turbulent-like flow fields.

Refreshments will be served in Math Annex Room 1115, 3:15 p.m.