A representative interactive linear-eddy-model (RILEM) for simulating spray combustion

Engine development aims at reducing pollutant emissions (e.g. NOx , soot,<br />UHC) while maintaining high efficiencies. Detailed experimental results<br />in combination with precise numerical predictions are of great importance<br />in order to develop combustion systems for new clean and efficient inter-<br />nal combustion engines. Computational fluid dynamics (CFD) tools must<br />be able to deal with multi-mode (premixed, partially premixed and non-<br />premixed) and multi-regime (from kinetically controlled to mixing con-<br />trolled) turbulent combustion under various conditions (low temperatures,<br />high pressures, high EGR rates). Work on laboratory flames [1, 2] clearly<br />demonstrates the strong need to account for the impact of unresolved tur-<br />bulent fluctuations of temperature and composition on chemical reaction<br />rates in Reynolds-averaged and large-eddy simulations (LES) of turbulent<br />combustion. Combustion models that neglect this so called turbulence-<br />chemistry interaction (TCI) cannot predict fundamental physical phenom-<br />ena like local or global extinction and re-ignition, possibly leading to in-<br />precise predictions of essential quantities including heat release rates, tem-<br />peratures and emissions. In order to maximize volumetric combustion ef-<br />ficiency and minimize pollutant formation, it is desirable to maximize the<br />rate of combustion subject to the limits of overall flame stability. These<br />limits of flame stability are determined by the rate of local extinction. Es-<br />pecially in new combustion concepts for engines, including stratified charged<br />compression ignition (SCCI), lean stratified premixed combustion, and the<br />use of high levels of exhaust gas recirculation (EGR) the impact of the flow<br />field on the chemistry plays an important role.<br />This thesis is divided into two parts. The first part deals with the intro-<br />duction of a new regime independent combustion model for engine applica-<br />tions, which is able to predict turbulent combustion under the beforemen-<br />tioned challenging conditions, while the second part deals with the extension<br />of existing knowledge about extinction in non-premixed combustion