An experimentally based approach for the space-averaged laminar burning velocity used for modeling premixed turbulent combustion

Abstract

Turbulent flame front propagation of methane/, ethane/, and propane/ air mixtures has been investigatedover a wide range of equivalence ratios inside a closed cylindrical vessel. Pressure/time relationships were recorded by a piezoelectric pressure transducer, in order to evaluate turbulent burning velocities using a thermodynamic model. To generate different scales of turbulence, two fans with variable speed of rotation were positioned within the explosion vessel. Turbulent flow velocities were measured under isothermal conditions by a forward scattering laser Doppler anemometry system, while integral length scale data were obtained calculating the cross-correlation function of the signals of two hot-wire probes positioned at variable distances from each other. Turbulent explosion experiments exhibit a strong influence of preferential diffusion/stretch interactionson the turbulent burning velocity. These effects cause a stronger increase in the turbulent burning velocity for lean methane/air mixtures than for rich ones and, in the opposite sense, for ethane/ and propane/air mixtures. Based on the flamelet assumption, an equation was derived to allow for flame straining reduction in the space-averaged laminar burning velocity S L,T . Starting from a theoretical model which is based on the superposition of characteristic timescales, a new combustion model was developed which takes into account preferential diffusion/stretch effects on the turbulent burning velocity. With the assumption that the space-averaged laminar burning velocity is approximately independent of turbulence parameters, comparison of experimentally determined turbulent burning velocities with those determined from the model shows satisfactory agreement. With mixtures characterized by high positive Markstein numbers, quenching phenomena could be observed at high turbulence intensities. In this regime, the reaction model used is no longer valid. A criterion based on the Karlovitz number was derived which was proven to be in close agreement with quenching limits from previous research.