The objectives of this study were to 1) determine the effects of cycle-to-cycle variations (ccv) on nitric oxide (NO) emissions, and 2) determine if the consideration of ccv affects the average NO emission as compared to the mean cycle NO emission. To carry out the proposed study, an engine simulation model was used. The simulation determines engine performance and NO emissions as functions of engine operating conditions, engine design parameters, and combustion parameters. An automotive, spark-ignition engine at part load and 1400 rpm was examined in this study. The engine cycle simulation employed three zones for the combustion process: (1) unburned gas, (2) adiabatic core region, and (3) boundary-layer gas. The use of the adiabatic core region has been shown to be especially necessary to capture the production of nitric oxides which are highly temperature dependent. Past research has shown that cyclic variations in combustion cause ccv of burn duration, ignition delay and equivalence ratio. Furthermore, literature has shown that variations of these three input parameters may be approximated by a normal frequency distribution. Using the mean and standard deviation, and a random number generator, input values were tabulated for the ignition delay, burn duration and equivalence ratio. These three input parameters were then used to simulate cyclic variations in the combustion process. Calculated results show that cyclic variations of the input parameters cause the cycle-by-cycle NO emissions to increase and decrease by as much as 59% from the mean cycle NO of 3,247 ppm. The average NO emission resulting from ccv was 4.9% less than the mean cycle NO emission. This result indicates that cyclic variations must be considered when calculating the overall NO emissions.

The currently reported thesis was concerned with visualisation of the charge homogeneity and cyclic variations within the planar fuel field near the spark plug in an optical spark ignition engine fitted with an outwardly opening central direct fuel injector. Specifically, the project examined the effects of fuel type and injection settings, with the overall view to understanding some of the key mechanisms previously identified as leading to particulate formation in such engines. The three fuels studied included a baseline iso-octane, which was directly compared to two gasoline fuels containing 10% (E10) and 85% (E85) volume of ethanol respectively. The engine was a bespoke single cylinder with Bowditch style optical access through a flat piston crown. Charge stratification was studied over a wide spectrum of injection timings using the Planar Laser Induced Fluorescence (PLIF) technique, with additional variation in charge temperature due to injection also estimated when viable using a two-line PLIF approach. Overall, both gasoline-ethanol fuels generally exhibited a higher degree of stratification, albeit at least partly alleviated with elevated rail pressures. Under both warm and cold liner conditions the E10 fuel showed clear evidence of fuel droplets persisting up until ignition. Interestingly, with late injection timing the repeatability of the injection was superior (statistically) with higher ethanol content in the fuel, which may have been associated with the higher charge temperatures aiding control of the evaporation of the main mass of alcohol. The findings were corroborated by undertaking a comprehensive study of the influence of varying fuel type and injection settings on thermodynamic performance and engine-out emissions during firing operation, with additional gas exchange effects also influencing the optimum fuel injection timings.