| Particulate and NOx are most important exhaust emission from diesel engine. The main components of emissions from diesel engine are soot, organic HC and sulphate, etc. The repeatability and reproducibility of measuring devices for particulate become worse and worse with decreasing particulates emission from diesel engine. The organic ingredients of diesel particulate take physical changes and chemical reactions such as nucleate, accumulation and condensation with changed size, shape and mass. This makes the stability of particulate measurement become much worse. Therefore it is necessary to study the factors affecting particulate measurement. This can be used for the device design of particulate measurement, particulate measurement and engine emission improvement.The main works around diesel engine particulate measurement and emission prediction are described in the following sections.1. Study on diesel engine emission prediction modelsPartil least squares regression has the main function of multiple linear regression, canonical correlation analysis and principal component analysis, which is widely used in the field of economic statistics, chemical and analysis and bioscince. Independent variables were optimized to improve reliability and prediction accuracy of PLS model by deleting insignificant variables. Independent variables were tested whether they were significant to dependent variable with full cross validation method by calculating their regression coefficients and variances. First, the most significant independent variable were selected by stepwise PLS regression. Second, the optimized PLS regression model were obtained by increasing interactive and quadratic term of the main independent variables optimized.Prediction models of NOx emission and in-cylinder peak pressure for HP common rail TCI diesel engine were established using PLS. This model has clear physical meaning and high prediction accuracy, and reducing the calibration word load significantly. The repeatability of NOx and Rb emission from a HP common rail TCI diesel engine in ESC cycle were studied in detail, and the prediction models of NOx and Rb emission were established using PLS. NOx are more sensitive to the engine load and air flow, and the fluctuate is more sensitive to air temperature at intercooler before and after, and back pressure, etc. Rb is more sensitive to specific fuel consumption and exhaust temperature and it’s fluctuate is more sensitive to fuel consumption.The particulate measurement accuracy is affected by many parameters, in which the most important parameters are dilution ratio, PM weight on filter and exhaust flow. For partial flow dilution PM measuring device, dilution ratio is especially important. The accuracy calculated by CO2trace gas is better than that calculated by mass flow measured with mass flowmeter. The accuracy of CO2is determined by the linearity of CO2analyzer.Linearization of a CO2analyzer has been studied by partial least squares regression (PLS). The prediction accuracies with PLS model were35.1%and39.3%higher than that with least squares regression and the model with Chebyshev polynomial regression, respectively, and the accuracy of dilution ratio was improved greatly.2. Study on the measurement uncertainty of particulate matter from diesel engineThe diameter of particulate from diesel engine ranges from3to300nm, occasionally1～10μm for few very large particulates deposited on the surface of transfer tube reinter into the flow gas. The main particulate mass loss is thermophoresis in the progress gas flow from exhaust pipe of diesel engine to the filter collecting particulate, and little loss in Brown diffusion, eddey diffusion, inertial and static, etc.The Influences of dilution ratio, exhaust temperature at the inlet of dilution tunnel, dilution air and mixed gas temperature on saturation ratio of C25were studied using a simple saturation ratio model. The concept of critical dilution ratio was firstly presented. Saturation ratio, SOF and PM increased with increasing dilution ratio when dilution ratio was lower than the critical dilution ratio. They decreased with increasing dilution ratio when dilution ratio was larger than the critical dilution ratio, and they were insensitive to dilution ratio when dilution ratio was next to the critical dilution ratio. Critical dilution ratio grew up with increasing exhaust temperature. So the Influence of dilution ratio to PM measurement was dependent on the engine load and exhaust temperature.Firstly, the main reason affecting repeatability of PM measurement was the soot and SOF particulates which were deposited on the surface of exhaust pipe and transfer tube. They were related to the diesel engine particulate level before PM measurement. It varied with engine modes such as the temperatures of exhaust gas and tube surface. If they are higher, the PM measured will increased in that the SOF particulates deposited on the surface of transfer tube will be volatilized, reentered into flow gas and then condensed into SOF particulates when diluted and cooled. In the contrast, the PM measured will decreased in that the organic HC in the flow gas will be deposited on the surface of transfer tube.The major countermeasure to minimum PM measured fluctuant is that engine and whole PM measuring system should be pre-conditioning sufficiently before PM measurement. For instance, the exhaust temperature should be high enough and the gas temperature at the outlet of transfer tube should be higher than425℃and last for enough time to ensure the SOF particulates deposited on the transfer tube surface to vaporize out.Secondly, for low emission engine, the ratio of PM in surroundings to PM filtered increased greatly. It would be27～60%in an ESC cycle for a State IV engine. The PM in surroundings varied with dilution air and the PM deposited on the transfer tube surface. PM in the secondary dilution air was higher than that in primary dilution air. So it was very important to reduce and hold steady the PM in secondary dilution air. Since the filtered compressed air was used as dilution air in some partial flow dilution system like the secondary dilution air in CVS system, it was very important to reduce the PM. However, the filtered fresh air from outside of the test bed, which was more stable, was used as dilution air in the SDPM partial flow dilution system developed. Another major factor affecting the PM measuring stability was absorption/vaporization of gas organic ingredient filtered, which was dependent on the temperature of filter surface. It would be more stability than that the PM filtered in the primary filter minused the PM filtered in the secondary filter, which was near pure gas.The influences of some parameters on PM measurement were studied in the experiment. Firstly, the SOF and PM matter measured would be reduced with increasing filter sampling time and face speed. Secondly, the ISOF matter measured would increase with increasing the humidity filter conditioned for filter easy absorbing water in the gas mixture. Moreover, the influence of CVS total flow on PM measured was dependent on the comparison of dilution ratio with the critical dilution ratio. Finally, SOF matter filtered was reduced decreased with increasing mixture gas temperature on filter.The Influences of some controls parameters on gaseous emission and PM measured were studied in the experiment. Firstly, with increasing intake air temperature, the air flow rate and AFR decreased, while NOx emission decreased and CO, smoke and PM measured increased. Secondly, with increasing intake air temperature after intercooler, the maximum combustion temperature and NOx emission increased, but the HC, CO and smoke emission decreased as combustion speeded up and burned completely. Moreover, with increasing flow resistance of intercooler, the NOx emission from engine would be reduced, but smoke and PM emission would be increased for oxygen deficit. Moreover, with the back pressure of engine increasing, the NOx emission would be reduced, while CO, smoke and PM emission would be increased, as turbining charger speed was slowed down, then AFR decreased, waste gas in cylinder increased, and the combustion speed slowed down and burned incompletely. Finally, the comparison experiment between fuels for state Ⅲ and state Ⅳ showed that the PM emission from diesel engine with state Ⅳ fuel decreased greatly and NOx emission only with slight effect, because the ratio of sulphate to PM is much high and the sulphur content in fuel has great influence for low emission engine.3. Development research of a partial flow dilution system A partial flow dilution system named SDPM was developed for PM measurement at steady mode based on literatures home and abroad. This system has the advantage of small, low cost, convenient to use, high accuracy, and very high correlation with CVS full flow dilution system. Firstly, two NDIR CO2analyzers with high and low range were used to measure the CO2in raw exhaust, dilution gas and dilution air, and the accuracy of dilution ratio and PM measured was much higher. Secondly, two ways for sampling units were used to reduce the PM measurement error caused by extreme PM being reentered into flow gas. Moreover, it has flexible framework and easy to use.A series of experiments were tested for PM measurement with SDPM, CVS and AVL SPC472at the same time. The COV of PM measured for state0to state Ⅲ diesel engines were about1.2to5.8%and that for state Ⅳ were about10%. The repeatability of PM measurement with SDPM was the same as that with CVS and was superior to that with472. The PM measured with SDPM system optimized was very closely linear correlation with that measured with CVS and the R-Square of the regression model was up to0.9988.4. Development of a smoke sampling system with variable sampling volumeA smoke sampling system with variable sampling volume was developed, which had higher accuracy than the traditional smoke meter, especially suitable for low emission diesel engine smoke measurement. As smoke fluctuation measured decreased, the smoke measurement accuracy was improved with increasing sampling volume. A method for large volume calibration was proposed in this study. This method was simple, practical, reliable and could be used to calibrate many volume. |
PDF Full Text Request