| Summary: | The thesis is concerned with the study of the characteristics of the turbulont flow field inside the combustion chambers of spark ignition engines. A comprehensive literature survey has been undertaken of the problem of cyclic variation in spark ignition engines, and the role of turbulence on flame propagation. The experimental methods used involve a detailed I application of hot wire anemometry, signal processing and computer techniques to the statistical analysis of the problem of cyclic variation in gas velocity and turbulence structure (intensity, scale, power spectrum and eddy diffusivity), at the spark plug location of motored spark ignition engines. Two engines have been used in the investigation, a Rolls-Royce V-SL and a Ford 2 litre V4. Extensive test programs were c arried out on both engines at a range of engine speeds and throttle settings, and a range of turbulence promoting devices have been investigated with regard to their suitability for promoting small scale turbulence during the ignition period. The variation of flow field characteristics with depth inside a wedge combustion chamber of the Rolls-Royce engine has been investigated. Cylinder-to-Cylinder variation in turbulence characteristics has also been investigated for the Rolls-Royce engine. The spectral composition of turbulent eddies inside a wedge and a heron combustion chamber was invetstigated over a wide range of engine speeds. It was found that increasing engine speed resulted in increases of both the fluctuating velocity components and the high frequency content of the eddies. Both effects are believed to be responsible for increases in turbulent flame speed in engines with increases in engine speed. Comparisons between the results obtained in the present work and the reported data of other investigations on turbulence measurements in spark ignition engines and closed vessels, show close agreement. A direct correlation has been established between the characteristics of the turbulence field in bn-g-ilnes and other isotropic flow fields. In particular, a correlation is found to exist between the eddy diffusivity as obtained from the anemometer measurements in an engine and the eddy diffusivity data for highly turbulent pipe flow obtained by other workers. These findings establish a relatively straightforward method for obtaining quantitative information on turbulence characteristics from mean velocity measurements on an engine, which require simple equipment. A theoretical model of combustion variation has been developed, based on the assumption that cyclic variation originates during the growth period of the initial flame kernel. This process was related to variations in eddy diffusivities of the small scale turbulence for different cycles. The model equations are chocked by making use of the established correlation between eddy diffusivity and gas mean velocity and using the reported experimental data of other workers, which shows very good agreement between the predictions of the model and the measured values.
|
|---|
