Search Results

In direct-injection gasoline (GDI) engines with charge stratification, minimizing engine-out nitrogen oxide (NOx) emission is crucial since exhaust-gas aftertreatment tolerates only limited amounts of NOx. Reduced NOx production directly lowers the frequency of energy-inefficient catalyst regeneration cycles. In this paper we investigate NO formation in a realistic GDI engine. Quantitative in-cylinder measurements of NO concentrations are carried out via laser-induced fluorescence imaging with excitation of NO (A-X(0,2) band at 248 nm), and subsequent fluorescence detection at 220-240 nm. Engine modifications were kept to a minimum in order to provide results that are representative of practical operating conditions. Optical access via a sapphire ring enabled identical engine geometry as a production line engine. The engine is operated with commercial gasoline (“Super-Plus”, RON 98).

In this work, a methodology is developed to study engine deposit formation mechanisms. It relies on analyzing the deposit with electron microscopy for morphology and infrared absorption spectroscopy for identifying typical chemical functions. Two lab-scale experiments are used to calibrate these measurement techniques by creating deposits through the two main phases: liquid film and soot deposition. To test this methodology, an optical engine is used to create a library of deposits. Two main deposit morphologies are found: a homogeneous underlayer as well as soot-like agglomerates. The underlayer is attributed to a fuel-film mechanism whereas the latter is attributed to particles formed through the combustion process. The influence of engine parameters, such as injection phasing and cooling temperature, on the quantity and morphology of the deposits is studied.

The complexity of the mixture formation in direct injection engines requires - according to a suitable mixture transportation and vaporization of the fuel - detailed knowledge of the in-cylinder processes to reliably place an ignitable mixture at ignition timing near the spark plug for any speed and load. Two different optical measurement techniques were adapted to a single cylinder engine and the spray propagation was observed from the start of injection until ignition. 3-pentanone tracer-LIF signals (laser-induced fluorescence) and CN spark emission signals were detected simultaneously in order to get information about the local equivalence ratio at the spark plug and compare the two methods. While there is a good correlation for homogeneous operating conditions of the engine, the results diverge in the stratified mode.

This paper presents results from a quantitative characterization of the NO distribution in a SI engine fueled with a stoichiometric iso-octane/air mixture. Different engine operating conditions were investigated and accurate results on NO concentrations were obtained from essentially the whole cylinder for crank angle ranges from ignition to the mid expansion stroke. The technique used to measure the two-dimensional NO concentration distributions was laser induced fluorescence utilizing a KrF excimer laser to excite the NO A-X (0,2) bandhead. Results were achieved with high temporal and spatial resolution. The accuracy of the measurements was estimated to be 30% for absolute concentration values and 20% for relative values. Images of NO distributions could also be used to evaluate the flame development. Both the mean and the variance of a combustion progress variable could be deduced.

Multiphase-induced ignition is frequently discussed as a trigger for early ignition in internal combustion engines. In this context, we investigated the ignition process of single lubricant-oil droplets and their interaction with the bulk air/fuel mixture in a high-pressure shock tube, mimicking oil-fuel interaction in turbocharged internal combustion engines at the end of the compression stroke. A fast micro-dispensing injector released single fuel or lubricant oil droplets with a diameter of 200±50 µm into shock-heated fuel/air mixtures consisting of PRF95 and synthetic air. The injector was flush-mounted in the sidewall of the shock tube. The droplets were released into the gas after the passage of the reflected shock waves at post-shock conditions of 2 MPa and 750-950 K. With a high-frame-rate color camera, the entire evolution of droplet injection and ignition was traced in space and time through a large sapphire window in the endwall of the shock tube.