| Summary: | The work presented in this thesis is intended to investigate the effects of fuel properties, injection strategy and timing on autoignition and combustion characteristics of a Homogeneous Charge Compression Ignition (HCCI) engine with a negative valve overlap (NVO) strategy. Conventional (pressure-transducer based) measurements and passive optical research have contributed to understanding of the chemical-physical sites of HCCI autoignition and combustion. This experimental work was undertaken on matching thermal and optical single cylinder research engines in configurations derived from a production Jaguar V8 engine. A thermal engine study using a range of fuels including conventional gasoline and primary reference fuels has been performed to gain insight into autoignition and combustion characteristics of various chemically dissimilar blends or components. This was done at different operating conditions by varying the engine speed and the proportion of residuals trapped. These measurements have shown that the autoignition and combustion characteristic of an HCCI operated engine are highly dependent on fuel blend composition and are also affected by engine operating conditions. It was found that the autoignition process type which the mixture undergoes, whether it is one- or two-step, depends very strongly both on fuel blend composition and on engine operating conditions. More specifically the presence and also proportion of particular chemical compounds in a blend could significantly contribute to the alteration of the process type. Similar experiments using the chosen engine operating points were repeated on the optical engine using passive optical diagnostics such as imaging and spectroscopy. Thereby it was possible to gain insight into the chemistry of one-step and two- step ignition processes. The image analysis of the port fuel injected (PFI) HCCI operation have been carried out for stoichiometric and lean conditions. A crank-angle resolved high-speed imaging technique was employed a piston crown window for optical access to the combustion chamber. The spatial repeatability nature of autoignition occurrence and the directions of combustion progress were evaluated using especially developed image processing technique. The insight into the expansion rates of burned areas and of the spreading velocities of reacting structures fronts was also gained by introducing two new image processing techniques. Various direct injection strategies (single and split injection) and timings, including fuel injection prior to and during the negative valve overlap period were optically investigated. The comprehensive study included the application of three diagnostic instruments: the Complementary Metal-Oxide Semiconductor (CMOS) high-speed colour imager, the intensified Charge Couple Device (CCD) and the imaging spectrograph. Among the other observations the applied passive techniques, the imaging and the spectroscopy in conjunction with adequate image processing techniques have shown that the combustion behaviour and also the colour of the burning mixture are dependent on the fuel injection scheme. With the investigated split (double) injection, when some of fuel is injected prior to TDC NVO the combustion behaviour is significantly different than when it is injected during even at TDC (NVO). There is a strong indication that a form of incandescence occurs during the NVO, which probably comes from the glowing soot. This is further supported by a quantification of the emitted luminescence and spectroscopic measurements during this phase.
|
|---|
