Search Results

The applicability of several popular diesel particulate matter (PM) measurement techniques to low temperature combustion is examined. The instruments' performance in measuring low levels of PM from advanced diesel combustion is evaluated. Preliminary emissions optimization of a high-speed light-duty diesel engine was performed for two conventional and two advanced low temperature combustion engine cases. A low PM (<0.2 g/kg_fuel) and NOx (<0.07 g/kg_fuel) advanced low temperature combustion (LTC) condition with high levels of exhaust gas recirculation (EGR) and early injection timing was chosen as a baseline. The three other cases were selected by varying engine load, injection timing, injection pressure, and EGR mass fraction. All engine conditions were run with ultra-low sulfur diesel fuel. An extensive characterization of PM from these engine operating conditions is presented.

The purpose of this study is to investigate the turbulent mixing in a diesel spray by large eddy simulation (LES). As the first step for the numerical simulation of diesel spray by LES, the LES of transient circular gas jets and particle laden jets were conducted. The simulation of transient circular jets in cylindrical coordinates has numerical instability near the central axis. To reduce the instability of calculation, azimuthal velocity around the central axis is calculated by the linear interpolation and filter width around the axis is modified to the radial or axial grid scale level. A transient circular gas jet was calculated by the modified code and the computational results were compared with experimental results with a Reynolds number of about 13000. The computational results of mean velocity and turbulent intensity agreed with experimental results for z/D>10. Predicted tip penetration of the jet also agreed to experimental data.

In order to investigate the soot formation process in a diesel spray flame, simultaneous imaging of soot precursor and soot particles in a transient spray flame achieved in a rapid compression machine was conducted by laser-induced fluorescence (LIF) and by laser-induced incandescence (LII) techniques. The 3rd harmonic (355nm) and the fundamental (1064nm) laser pulses from an Nd:YAG laser, between which a delay of 44ns was imposed by 13.3m of optical path difference, were used to excite LIF from soot precursor and LII from soot particles in the spray flame. The LIF and the LII were separately imaged by two image-intensified CCD cameras with identical detection wavelength of 400nm and bandwidth of 80nm. The LIF from soot precursor was mainly located in the central region of the spray flame between 40 and 55mm (270 to 370 times nozzle orifice diameter d0) from the nozzle orifice. The LII from soot particles was observed to surround the soot precursor LIF region and to extend downstream.

The accuracy of the injection rate meter based on W. Zeuch's method in the measurement of multiple injection rate and amount was calibrated using a small cam driven piston that is driven by an electric motor. For the pre- or early-injection, a sensor with a high sensitivity can be applied to measure the small pressure increase due to the small injection amount. In case of the multiple injection that has the post and/or late injection, a pressure sensor with a low sensitivity must cover not only the large pressure increase due to the main injection but also the small pressure increase due to the post and/or late injection because the output of the high sensitivity sensor is saturated after the main injection. So the linearity of the low sensitivity pressure sensor was calibrated with the cam driven piston prior to the experiment with the actual injection system.

To investigate the ignition process in a diesel spray, the ignition in a transient fuel spray is analyzed numerically by a discrete droplet spray model (DDM) coupled with the Shell kinetics model at various operating conditions. Predicted results show that the fuel mixture injected at the start of injection, which travels along midway between the spray axis and the spray periphery, contributes heavily to the first ignition in a spray. The equivalence ratio and temperature of the first ignited mixture are kept nearly constant until the start of hot ignition. The temperature of the first ignited mixture is kept at a constant value of higher temperature than the thermodynamic equilibrium temperature of the mixture before the hot ignition starts. The equivalence ratio of the first ignited mixture is around 1.6 at initial gas temperatures between 750 K and 850 K.

The time of, and location where ignition first occurs in a diesel fuel spray were investigated in a rapid compression machine (RCM) using the two–dimensional techniques of silicone oil particle scattering imaging (SSI), and the planar laser induced fluorescence (LIF) of formaldehyde. Formaldehyde has been hypothesized to be one of the stable intermediate species marking the start of oxidation reactions in a transient spray under compression ignition conditions. In this study, the LIF images of the formaldehyde formed in a diesel fuel spray during ignition process have been successfully obtained for the first time by exciting formaldehyde with the 3rd harmonic of the Nd:YAG laser. SSI images of the vaporizing spray, and the LIF images of formaldehyde were obtained together with the corresponding time record of combustion chamber pressures at initial ambient temperatures ranging from 580 K to 790 K.

To investigate the ignition process in a diesel spray, the ignition in a transient fuel spray is analyzed numerically by a simple quasi-steady spray model coupled with the Shell kinetics model at various operating conditions and validity of this model is assessed by a comparison with existing experimental data. The calculated results indicate that the competition between the heat absorption of fuel and the hot air entrainment determines the equivalence ratio of mixtures favorable for the ignition to occur in the shortest time.

A thermodynamic two-zone model which assumes a stoichiornetric burned gas region and unburned air region is presented in an attempt to calculate more precise rate of heat release of diesel combustion. A comparison is made of the rate of heat release obtained by the two-zone model with that obtained by the conventional single-zone model. It shows around 10 % increase in the rate of heat release with the two-zone model. The effect of state equation of gas is also examined with the single-zone model and the use of a real gas law in stead of the perfect gas law is found to yield minor difference in the rate of heat release at a high boost operating condition.

Local inhomogeneity of mixture concentration affects combustion characteristics in the lean burn system and also in the stratified charge combustion system. To investigate such combustion systems, the effects of inhomogeneous mixtures were examined using a carefully controlled experimental system. In this study, a constant-volume chamber, which can simulate an idealized stratified charge by using a removable partition inside the chamber, was developed. Flow and combustion characteristics were examined by indicated pressure analysis, Schlieren photography, ion probe measurements and local equivalence ratios measurements while varying the combination of initial equivalence ratios on each side of the partition. As a result, combustion characteristics of charge stratified, very lean propane-air mixture were clarified.

Three-dimensional computation has been applied to analyze combustion and emission characteristics in direct-injection diesel engines. A computational code called TurboKIVA was used to investigate the effects of the swirl ratio, one of the fundamental factors related to combustion control, on combustion characteristics and NOx and soot emissions. The code was first modified to calculate soot formation and oxidation and the precise behavior of fuel drops on the combustion chamber wall. As a result of improving calculation accuracy, good agreement was obtained between the measured and predicted pressure, heat release rate and NOx and soot emissions. Using this modified version of TurboKIVA, the effects of the swirl ratio on NOx and soot emissions were investigated. The computational results showed that soot emissions were reduced with a higher swirl ratio. However, a further increase in the swirl ratio produced greater soot emissions.

In an attempt to achieve lean combustion in Diesel engines which has a potential for simultaneous reduction in no and soot, the authors developed a micro-hole nozzle which has orifices with a diameter as small as 0.06 mm. Combustion tests were carried out using a rapid compression-expansion machine which has a DI Diesel type combustion chamber equipped with the micro-hole nozzle. A comparison with the result of a conventional nozzle experiment revealed that the ignition delay was shortened by 30 %, and in spite of that, both peaks of initial premixed combustion and diffusion combustion increased significantly. The combustion in the case of the micro-hole nozzle experiment was accompanied with a decrease in soot emission, whereas an increase in NO emission.

A rapid compression-expansion machine was developed, which can simulate intake, compression, expansion and exhaust strokes in a single Diesel cycle by an electrically controlled and hydraulically actuated driving system. The whole system which is composed of a hydraulic actuator, fuel injector and a valve driving device, is sequentially controlled by a micro-computer. The machine features; 1) accurate control of piston position at TDC, 2) no effect of lubricant on HC emission due to the use of dry piston rings; 3) independent control of local wall temperature; and 4) high power output to drive heavy piston at high frequency. The single cycle operation permits Diesel combustion experiments under a wide range of operating conditions and easy access of optical diagnostics with minimized amount of test fuel. The performance test showed that the machine can drive a DI Diesel type piston with a 100 mm bore at a maximum frequency of 16.7 Hz at a maximum compression pressure of 15 MPa.

The turbulent dispersion of particles in an unsteady two dimensional particle-laden jet was simulated by a discrete vortex method coupling with a model of gas/particles interaction. Numerical analysis of a spray yielded the distributions of vorticity, fuel mass concentration and local Sauter mean diameter (SMD) of droplets in a spray. The predicted distribution of local SMD of droplets in a spray demonstrated that the size of droplets in the spray periphery is larger than that of droplets in the center region of spray. This trend of distribution of drop size coincided with that of measured one. The predicted distributions of drop size and vorticity revealed that the larger droplets are easily centrifuged to the periphery of the spray. The effects of the pattern of injection rate on the mixing process in a transient spray were also investigated.

Two kinds of planar soot imaging techniques, laser induced incandescence (LII) and laser induced scattering (LIS) techniques were applied simultaneously to an unsteady free spray flame achieved in a rapid compression machine. An analysis of LII and LIS images yielded three kinds of qualitative images of soot concentration, size of soot particles, and number density of soot in the flame. These images revealed the fact that the soot is formed mainly in the center region of a flame resulting in an appearance of soot cloud with high number density and small particle size in this region, and then the soot size increases and the number density decreases while soot is conveyed downstream.

Air-entrainment characteristics of non-evaporating sprays and flames were measured by means of high-speed photography including ordinary shadowgraphy of sprays, back diffused light illumination photography and laser shadow photography of flames. Effects of injection pressure and nozzle orifice diameter on air-entrainment characteristics were investigated parametrically. The amount of air entrained into a flame was calculated by a two-zone thermodynamic model with data obtained from the photographs and the pressure measurement in the combustion chamber. The air-entrainment characteristics of flames were compared with those of the corresponding sprays. It showed that immediately after the start of ignition, the air entrainment into a flame increased more rapidly as compared with the corresponding spray and then, with the development of diffusion combustion, the air entrainment gradually approached that of the spray.

To measure the fuel vapor concentration in an unsteady evaporating spray injected into nitrogen atmosphere, the exciplex-forming method, which produces spectrally separated fluorescence from the liquid and vapor phase, was applied in this study. Two experiments were conducted to investigate the qualitative and quantitative applicability of the technique in a high temperature and high pressure atmosphere during the fuel injection period. One is to examine the thermal decomposition of TMPD dopant at a high temperature and a high pressure nitrogen atmosphere during a short period of time. The other is to calibrate the relationship between fluorescence intensity and vapor concentration of TMPD at different vapor temperatures. And then, the qualitative measurement of fuel vapor concentration distributions in diesel sprays was made by applying the technique.

The cross-sectional distribution of fuel vapor concentration in an evaporating spray was measured quantitatively by a new scattering imaging technique, silicone particle scattering imaging method, which was proposed in a previous paper[1]. When fuel containing silicone oil injected into a nitrogen environment at high temperature, the volatile base fuel in the droplets vaporized rapidly, leaving behind small droplets of silicone oil suspended in the vapor-gas mixture. The silicone oil droplets were illuminated by a thin laser sheet, and the scattered light was imaged by a CCD camera. The cross-sectional distribution of vapor concentration was estimated from the scattering image of the silicone oil droplets by Mie scattering theory. The results demonstrated clearly the inhomogeneity of the fuel vapor concentration. The distribution of vapor concentration was discontinuous, and islands of rich mixture with a scale of several millimeters existed in the center region of the spray.

Local heat flux from the flame to the combustion chamber wall, q̇, was measured the wall surfaces of a rapid compression-expansion machine which can simulate diesel combustion. Temperature of the flame zone, T1, was calculated by a thermodynamic two-zone model using measured values of cylinder pressure and flame volume. A local heat transfer coefficient was proposed which is defined as q̇/(T1-Tw). Experiments showed that the local heat transfer coefficient depends slightly on the temperature difference, T1-Tw, but depends significantly on the velocity of the flame which contacts the wall surface.

Experiments were conducted with 4-cylinder and single-cylinder direct injection diesel engines to examine the effects of combustion chamber insulation on heat rejection and thermal efficiency. The combustion chamber was insulated by using a silicon nitride piston cavity that was shrink-fitted into a titanium alloy crown. The effect of insulation on heat rejection was examined on the basis of heat release calculations made from cylinder pressure time histories. High-speed photography was used to investigate combustion phenomena. The results showed that heat rejection was influenced by the combustion chamber geometry and swirl ratio and that it was reduced by insulating the combustion chamber. However, because combustion deteriorated, it was not possible to obtain an improvement in thermal efficiency equivalent to the reduction in heat rejection.

An experimental evaluation of the reliability of the Zeuch's method was carried out. The following were derived: 1) cavitation limits the minimum back pressure available; 2) the injection rate measured by the Zeuch's method agrees with that by the W.Bosch's method; 3) the effect of dynamic pressure of the injected fuel jet has a negligible effect on the pressure sensor which is attached to the chamber wall; and 4) the high-frequency noise after the end of injection observed in the Zeuch's measurement can be effectively removed by either a low-pass filter or an inverse Fourier transform processing.