Download files
Abstract
Homogeneous charge compression-ignition (HCCI) engines have been considered to hold potential for next generation internal combustion engines with low emissions and low fuel-consumption. However, some technical hurdles, such as low combustion-efficiency at low load and excessive pressure-rise rate (PRR) at high load, significantly challenge its practical application.
In this thesis, fundamental studies of fuel ignition response to thermal stratification were first conducted by using direct numerical simulations coupled with a detailed chemistry mechanism. For a two-stage ignition fuel with negative temperature coefficient (NTC) behaviour, dimethyl ether, the auto-ignition regime was found to depend strongly on the initial temperature. Molecular diffusion was found to be negligible in comparison to chemical reaction when the initial temperature fell inside NTC regime; however, once the initial temperature was outside NTC regime, diffusion became relatively more significant. Diffusion was also observed to decrease with an increase of the length-scale. PRR was found to be reduced with thermal stratification, but this was also dependent on the mean temperature.
Then, non-reacting multi-dimensional engine modelling was conducted to investigate the effects of fuel direct injection on the resulting mixture distribution. It was found that as the start of injection was retarded, more fuel was concentrated in the central areas of the cylinder, leading to a potential increase of combustion efficiency and potential reduction of carbon monoxide and unburned hydrocarbons, but a potential increase of excessive nitrogen oxides. Droplet-wall interaction and spray-to-spray interaction were observed to play essential roles in fuel distribution. Furthermore, the use of high injection pressure can enhance the mixing, while the use of high swirl ratio and low injection pressure showed negative effects on the global mixing.
Finally, reacting engine simulations were carried on to study the effects of thermal stratification on a fully premixed HCCI engine fuelled by ethanol. These studies pointed out many challenges with attempts to model HCCI predictively, owing to strong sensitivities to initial charge temperature and pressure, wall temperatures, residual gas composition, initial turbulence intensity and models for its evolution and wall models. These sensitivities were analysed and used to construct an optimised model that agreed quite well with experimental pressure traces and associated quantities such as the PRR, the indicated mean effective pressure, and the thermal efficiency. Analysis of the optimised model results was used to determine that enhanced thermal stratification demonstrated a significant reduction of the PRR. The degree of the reduction was found to depend on the penetration of thermal stratification into the bulk-gas regions. In addition, turbulence played an important role in the control combustion phasing primarily by altering the distributions of thermal stratification.