Turbulence-Premixed Flame Interactions

Abstract

The interaction between premixed flames and turbulence is an inherently non-linear phenomenon. Understanding such interactions has profound practical implications towards the development of better combustion devices and turbulent combustion models. To this end, three statistically planar, freely propagating, turbulent premixed lean H2-air flames with varying turbulence intensities are generated using Direct Numerical Simulations (DNS). A newly developed backward tracking technique is applied to identify the source locations of iso-scalar surfaces of the turbulent premixed flames. In this technique, flame particles embedded on the iso-scalar surface are tracked backwards in time.

Using the available flame particle trajectories, finite-sized Lagrangian triangles are created and tracked forward in time to investigate changes in their shape and size. These changes approximate corresponding modifications of the underlying flame surface. Based on the inferences obtained, a phenomenological model proposed for the evolution of geometric structures in non-reacting flows is modified and validated for the present cases. The evolution of probability density function (pdf) of Lagrangian triangle area is then studied to understand the conditional stretch rate of the triangles, as they disperse out to generate the complete flame surface. An optimization problem is posed to obtain the conditional stretch rate, and it is found the stretch rate is dependent on the instantaneous triangle size.

Based on the outcomes of the above-mentioned exercises, the expressions for turbulent flame speed and hence the burning rate of the flame are found to be implicitly dependent upon the statistics of the leading portions of the flame surface, but at an earlier time. This signifies the importance of these surface generating locations that have been identified as the “leading points”, a concept used in turbulent combustion modelling.

In summary, Lagrangian methods have been utilized in this work to investigate the generation mechanism of turbulent premixed flames in their statistically stationary state