Large Eddy simulation of premixed and stratified turbulent combustion

Abstract

The work presented in this thesis focuses on the modelling of turbulent fully premixed and stratified flames, where stratification refers to the equivalence ratio of the unburned mixture varying in space.

The modelling approach is based on a transported probability density function formulation in the context of Large Eddy Simulation, in conjunction with the fully Eulerian stochastic fields solution method. This method provides detailed statistical information about the flow field and reactive scalars without making assumptions about the combustion regime, maintaining generality.

In the probability density function formulation the chemical reaction rate term appears in closed form and a reduced yet still detailed reaction mechanism is employed to model the methane-air combustion over a range of stratification ratios.

Amendments to the formulation and implementation are proposed. These account properly for the limit case of no sub-grid viscosity, which is important in modelling flames in well resolved and low Reynolds number flow regions as found in the vicinity of flame holders used in many premixed burner configurations.

Two laboratory scale burner configurations are investigated numerically, under various operating conditions and mixture stratification ratios, and results are compared against existing detailed experimental measurements. Theses configurations are a low Reynolds number slot burner that uses a rod to stabilise a V-shaped flame and a moderate Reynolds number co-annular burner with a central bluff-body to stabilise swirling and non-swirling flames. The equivalence ratio of the mixtures fed in by the slots or annuli were varied to produce different stratification ratios.

Improvements in ability to model low Reynolds number flames and capture stratified flame behaviour are demonstrated. The methods is able to predict the flow field as well as the mean and rms values of the reactive scalars to a good degree of accuracy.