Download files
Abstract
With the concerns growing regarding the need to reduce NOx emissions and improve combustion efficiency, HCCI technology has been developed. In spite of its benefits, it presents challenges which are mainly concerned with controlling the auto-ignition timing and limiting the rate of heat release. Numerical simulations have become useful tools to provide a detailed understanding of the combustion process leading to overcome these challenges. This thesis focuses on developing a model based on the first-order CMC approach for modelling the ignition of lean mixtures under HCCI conditions. The premixed CMC equations are derived with the choice of enthalpy as a conditioning variable. The performance of the CMC model is examined for different fuels under various conditions. Three DNS data-sets are employed. The first DNS data-set modelled the ignition of n-heptane/air mixture with mean temperatures in NTC regime whereas the second and third sets of DNS data modelled the ignition of thermally stratified n-heptane/air and iso-octane/air mixtures, respectively, with mean temperatures outside the NTC regime. The CMC results are in excellent agreement with the DNS for the cases with small-to-medium temperature inhomogeneities whereas the CMC under-predicts the ignition delay in the cases with large temperature inhomogeneities. Further investigation using the DNS data shows that the conditional fluctuations are significant in the cases with large thermal stratifications which in turn cause a breakdown of the first-order closure hypothesis. In these cases where a mixed mode of deflagration and spontaneous ignition exists and the dissipation fluctuations generate conditional fluctuations. An assessment of the conditional variance equations, derived for premixed flames, shows correlations between dissipation and conditional fluctuations and correlations between reaction and conditional fluctuations are the dominant sources of conditional fluctuations. The performance of the CFD-CMC solver is also evaluated for modelling the effects of compression heating and expansion cooling. For this purpose, a DNS of a lean thermally stratified ethanol/air mixture is used. The results show the inclusion of compression heating and expansion cooling suppresses the deflagration mode and hence the CMC model predicts the ignition process very well for all cases even for those with large thermal stratification.