Search Results

The effects of reduced fuel injection pressure on the cold start performance of a GDI engine have been studied in a single-cylinder, optically-accessible research engine. Two Delphi Automotive Systems DI-G injectors, with included spray cone angles of 60° and 80° respectively, were studied. Both injectors are designed to operate at a nominal fuel line pressure of 10 MPa. For the study they were operated at several fuel feed pressures between 10 MPa and 2 MPa. Two start of injection timings (50° and 100° ATDC) were examined. Cold start performance was characterized by measurements of the GIMEP, COV of GIMEP, and total engine out UHCs. Simultaneous Planar Laser Induced Fluorescence (PLIF) and Mie Scattering images of the fuel spray were used to observe spray penetration, mixing, and in-cylinder fuel distribution throughout the intake and compression strokes. Ultimately these images were used to explain observed performance differences.

Experiments have been conducted in a laminar flow system and in a research engine to investigate the effect of additives on the combustion of gasoline-like fuels. The purpose of the laminar system is to enable rapid screening of additives to determine which, if any, have an enhancing effect on the early stages of combustion, especially under conditions of poor fuel vaporization which exist during cold-start in a spark ignited engine and which make flame propagation difficult to start and sustain. The base fuel used in the laminar and engine systems was a 9 component mixture formulated to simulate those components of gasoline expected to be present in the vapor phase in the intake system of an engine under cold-start conditions. In the laminar system, the pre-mixed, pre-vaporized fuel-air mixture is ignited and a time history of the combustion generated, hydroxyl radical chemiluminescence is recorded.

Results are presented from an experimental study of the effects of engine speed and injection pressure transients on the cold start performance of a gasoline direct injection engine operating on iso-octane. The experiments are performed in an optically-accessible single-cylinder research engine modified for gasoline direct injection operation. In order to isolate the effects of the engine speed and injection pressure transients, three different cold start simulations are used. In the first cold start simulation the engine speed and injection pressure are constant. In the second cold start simulation the injection pressure is constant while the engine speed transient of an actual cold start is simulated. In the third cold start simulation both the engine speed and the injection pressure transients of an actual cold start are simulated.

The objective of this research was to characterize the effects of fuel component volatility on gasoline direct injection (GDI) engine cold start. Three different fuel components, representing gasoline light end, mid-point and heavy end components, were used to form three fuel blends of different volatility. Performance tests and in-cylinder fuel distribution imaging tests using these fuel blends were carried out in a firing single-cylinder optically-accessible engine following a simulated cold start test schedule. Performance results, based on in-cylinder pressure and engine-out hydrocarbon measurements, during the initial transient phase of GDI cold start showed significantly degraded performance with the low volatility fuel blend, while the high volatility blend showed slightly improved performance. Neither the low nor high volatility fuel, however, showed a discernable effect on the quasi-steady state cold start performance.

The effects of intake-generated swirl and tumble on cold start performance have been investigated in a firing single-cylinder Gasoline Direct-Injection (GDI) engine. The engine utilizes a Ford Zetec cylinder head modified for GDI operation and a fused silica cylinder which provides extensive optical access to the combustion chamber. Uniquely designed port-inserts were positioned in the intake ports to generate enhanced swirling or tumbling motion inside the cylinder. Experiments were conducted using a constant speed (∼ 900 rpm) simulated cold start procedure, where the engine is motored for approximately 40 cycles, after which fuel injection and spark ignition commence and continue for 190 cycles and then the engine is stopped. Measurements were made of the various engine temperatures, engine-out total hydrocarbon emissions, and in-cylinder pressure throughout the test period.

Experiments have been conducted in a firing single-cylinder spark-ignition engine employing a Ford Zetec cylinder head that has been modified to operate with either standard port-fuel-injection, air-forced port-fuel-injection or direct-injection. The engine utilizes a fused silica cylinder and therefore provides extensive optical access to the combustion chamber. Tests were conducted using a constant speed simulated cold start procedure, which is composed of an initial start-up transient and a quasi-steady-state idle period. In this procedure, the engine is briefly motored at 889 rpm and then combustion commences shortly after the start of fuel injection. Measurements which were performed include in-cylinder pressure as well as intake valve, exhaust valve, piston, cylinder, head, and intake air temperature throughout each cycle of the test period. The engine-out total hydrocarbon emissions were also measured.

The application of planar laser-induced Rayleigh scattering for quantitative 2-D measurements of vapor-phase fuel concentration in the main combustion zone of a direct-injection Diesel engine has been explored, developed and demonstrated. All studies were conducted in an optically accessible direct-injection Diesel engine of the “heavy-duty” size class at 1200 rpm and motored TDC conditions which were typical of the production version of this engine. First, this study verifies that beyond 27 mm from the injector all the fuel is vapor phase. This was done by investigating the Diesel jet under high magnification using 2-D elastic scatter imaging and subsequently evaluating the signal intensities from the droplets and other interfering particles (Mie scattering) and the vapor (Rayleigh scattering).

An experimental study has been conducted to characterize NO and soot evolution in an optically-accessible D.I. diesel engine with a square combustion chamber. Two-dimensional laser-induced fluorescence was used to characterize NO evolution. Soot evolution was characterized by two-dimensional laser-induced incandescence (LII) and Mie scattering techniques as well as direct photography of the flame luminosity. The engine operating parameters were set to provide optimum conditions for NO imaging. Attenuation of the UV beam proved to be the major obstacle in obtaining NO images. Therefore, oxygen was added to the intake air charge in order to reduce the optical density of the combustion medium. The NO images showed that the NO formation started almost immediately after ignition and ceased no later than 40 degrees ATDC. No soot images could be obtained by the laser-induced incandescence or Mie scattering methods before 20 degrees ATDC since the soot concentration was very low.

Flame kernel growth and cylinder pressure data were simultaneously obtained from an optically-accessible, square piston, SI engine. Flame kernel growth was measured using simultaneous, orthogonal, Schlieren photography, while cylinder pressure was measured using a piezoelectric pressure transducer. The data were analyzed on a cycle-resolved basis to determine the correlation between cyclic fluctuations in flame kernel growth and cylinder pressure. The engine was operated at 875 RPM with premixed, prevaporized, stoichiometric isooctane in air. The engine, designed with ported intake and exhaust, was fired every tenth cycle to ensure complete scavenging. Tests were conducted with and without nitrogen dilution, while ignition timing was fixed at 25° BTDC. With 0% dilution the percent variation in the maximum cylinder pressure was 8.5%, while with 10% dilution the percent variation increased to 14%.

Enhancement of methanol combustion in a direct injected Diesel engine using catalytically coated glow plugs was examined for platinum and palladium catalysts and compared to a non-catalytic baseline case. Experiments were performed for 6 and 10 brake Kilowatts (bKW) at 2500 rpm. Comparisons were made based on combustion, performance, and emissions including carbon monoxide (CO), oxides of nitrogen (NOx), unburned hydrocarbons (UHC), unburned methanol (UBM), and aldehydes. Results show a decrease in glow plug temperature of 100 K is achievable using platinum catalysts, and 150 K for palladium. Furthermore, the palladium catalyst was found to provide better combustion characteristics than the platinum catalyst. Also, the use of both catalysts produced lower aldehyde emissions, and the palladium reduced NOx emissions as well. However, unburned methanol increased for both catalytic glow plugs with respect to the non-catalytic case.

The effect of chromium oxide (Cr2O3) and partially stabilized zirconia (PSZ) surfaces on the gas-phase ignition of hydrocarbon fuels has been studied in a high pressure chemical flow reactor under low heat-loss Diesel engine simulated conditions. Results are presented for propane and n-hexane fuels at 6 and 10 atmospheres and are compared to previously reported results obtained under similar conditions with platinum surfaces. Thermal enhancement of the gas-phase reactions, as evidenced by the catalytic production of CO and CO2, was found to be negligible for both Cr2O3 and PSZ, where as significant thermal enhancement was observed with platinum. Chemical enhancement of the gas-phase ignition reactions, as evidenced by an increase in the production of intermediate hydrocarbons, was observed under fuel rich conditions with both Cr2O3 and PSZ. Oxygen was shown to be essential in that no change in the intermediate hydrocarbon concentration was observed under pyrolysis conditions.

A model of turbulent flame kernel growth has been developed based on the fractal nature of premixed turbulent flame structure. The flame kernel model uses a recently developed heuristic model for the fractal dimension of premixed turbulent flame surfaces in order to predict the flame kernel's structure. The fractal flame kernel model requires independent knowledge of the turbulence intensity, Kolmogorov scale, integral scale and energy spectrum, as well as, the laminar flame speed. Comparisons with measurements from turbulent flow tunnels and engines show good agreement with the fractal model's predictions.

The potential for catalytically enhanced ignition in low-heat rejection Diesel engines has been experimentally studied under engine simulated conditions in a high pressure chemical flow reactor. Results are presented for propane oxidation on platinum at 6 and 10 atmospheres, at temperatures from 800K to 1050K, and at equivalence ratios from 0.5 to 4.0. For turbulent transport rates which are typical of those in an engine, as much as 20% of the fuel was found to react on the catalyst before the onset of the gas-phase ignition reactions. Depending on the adiabaticity of the combustion chamber walls, this could lead to significant thermal enhancement of the gas-phase ignition process. Evidence of chemical enhancement was also observed, at 10 atm under very fuel rich conditions, in terms of a change in the concentration and distribution of the hydrocarbon intermediate species. Possible mechanisms for the observed chemical enhancement due to surface generated species are discussed.