Thermodynamically Consistent Modelling of Gas Turbine Combustion Sprays

Abstract

In order to support the design procedure and increase the reliability and safety of combustion engines fired with liquid fuel at a reasonable cost, numerical prediction tools well validated by comprehensive experimental data are needed. As there is today enough evidence that Large Eddy Simulation (LES) is able to well capture intrinsically time and space dependent phenomena, LES will be employed. However, in most LES based spray modules for predicting spray combustion the interactions between both phases and between evaporating droplets and combustion are either not adequately considered or not incorporated at all. The objective of this work is to develop and validate a thermodynamically consistent spray module for Large Eddy Simulation that allows describing accurately the essential processes featuring spray combustion in gas turbine combustion chambers. These include besides the injection of liquid fuel, the turbulent droplet dispersion, the vaporization of the droplets and mixture formation and the subsequent spray combustion. In particular, (1) a physically consistent SGS-model describing the influence of droplet diameter and interface transport on the gas phase turbulence as well as the effect of the droplet evaporation on the mass and scalar transport processes (turbulence modulation) has been adapted for LES into an Eulerian–Lagrangian framework. (2) Apart from classical evaporation models valid in atmospheric conditions, an advanced evaporation model, the so called non-equilibrium model, appropriate for gas turbine conditions have been integrated and validated. (3) The chemistry-turbulence interaction under droplet evaporating conditions has been considered according to a presumed (filtered) probability density function while the combustion process itself is described following a tabulated detailed chemistry based on FGM (Flamelet Generated Manifold). (4) All the developed sub-models along with the complete model have been implemented in the working package FASTEST/LAG3D and validated in non-reacting and reacting configurations with available experimental data. Comparisons include exhaust gas temperature, droplet velocities and corresponding fluctuations, droplet mean diameters and spray volume flux at different distances from the exit planes. An overall good agreement with experimental data has been achieved. Parts of this contribution has been already reported as mentioned throughout the paper.