Turbulent combustion modelling of fuels using LES

Abstract

Combustion simulations with high fidelity turbulence models and detailed chemistry may suffer from high computational power requirements due to the combined cost of time-scale dissipation and small integration steps. Such a limitation can be avoided by employing a hybrid reaction mechanism reduction method called local self-similarity tabulation (LS2T). LS2T directly solves several dominant species reactions and incorporates the effects of other species on dominant ones by data retrieval from pre-calculated tables. This study is based on the application of the LS2T method to high fidelity 3D combustion simulations with different fuels and combustion physics. The test cases that are selected for demonstration purposes are Sandia Flame-D, the premixed methane combustion, and Sandia Spray-A, the non-premixed n-dodecane combustion. The combustion simulations use large eddy simulation (LES) as turbulence solver and transported probability density function (TPDF) for species transport, to increase the accuracy of the simulation and avoid the use of any additional reaction model. This study is the first demonstrator of LS2T approach application to 3D combustion problem with Sandia Flame-D simulation and it is also the very first Spray-A simulation that is executed using LES and TPDF in a 3D resolved domain. The report consists of wide literature research regarding the LES combustion analyses of both test cases and chemistry reduction techniques applied, a detailed theory behind all the physics solution methods used, the computational methods implemented for the study, results, and discussions of the 0D and 3D simulations. The results show that by the use of the LS2T method, it is possible to have high accuracy and generate results similar to the detailed chemistry while maintaining an acceptable computational effort.