Investigation of Particulate Matter Size Distribution and Concentration During Low Temperature Combustion.

摘要

Stringent emissions regulations imposed by US EPA on diesel engine manufacturers have shifted the focus towards development and implementation of in-cylinder emissions reduction strategies. In order to simultaneously achieve low oxides of nitrogen (NOx) and particulate matter (PM) emissions, along with high thermal efficiencies, advanced combustion concepts, such as low temperature combustion (LTC), have been identified as promising approaches. Furthermore, fuel properties, such as cetane number (CN) have been known to critically affect the combustion process and hence exhaust emissions.;The primary objective of this study was to investigate the impacts of LTC on PM size distribution and number concentration for two different fuels - one a low and other a high CN. A 1.9l turbocharged diesel engine equipped with a common rail injection system and cooled high pressure EGR was employed for this work. Additionally, the OEM engine control unit (ECU) was replaced by an independent controller, allowing open access to control a number of ECU parameters. The engine was instrumented with an in-cylinder pressure transducer to calculate the in-cylinder bulk temperature as a function of crank angle position. Gaseous brake specific emissions were measured from a full-flow CVS tunnel system following the procedures outlined in 40 CFR, 1065. The study was aimed at characterizing the size distribution and concentration of solid accumulation mode particles, as these reflect the in-cylinder PM formed during LTC operation. Due to an expected increase in hydrocarbon emissions, the PM sampling setup was designed to remove the volatile fraction by means of a two stage dilution system, with a hot first stage and cold second stage dilution. A scanning mobility particle sizer (SMPS, 3936) was employed to measure number concentrations and size distributions of PM emissions. A design of experiment (DOE) technique, making use of orthogonal arrays and ANOVA, was employed in order to simplify the process of achieving LTC by varying only four engine parameters, namely variable geometry turbocharger (VGT) vane position, exhaust gas recirculation (EGR) rate, start of pilot injection and fuel rail pressure.;Low temperature combustion was achieved at an engine speed of 2100 rpm and a target brake mean effective pressure (BMEP) of 3.5 bar by employing a split fuel injection strategy, increasing the EGR rates and fuel injection rail pressures, and advancing fuel injection timing. A unique combination of all these four parameters was identified to simultaneously produce low NOx, low soot, and low in-cylinder bulk temperatures. The identified optimal parameter setting for both fuels showed a reduction in the NO x-PM trade-off, compared to conventional combustion. PM emissions decreased by over 84% during operation of the low CN fuel, along with a 51% increase in the NOx emissions, when compared to the high CN fuel. Although the low CN fuels exhibited zero soot characteristics, there was a drastic increase in the THC and CO emissions when compared to the high CN fuels. Advancing SOI timing resulted in an increase in nanoparticle emissions for both the fuels supporting the hypothesis. Finally, a comparison of PM size distribution between conventional combustion and LTC during the operation of a low CN fuel revealed a significant shift towards nucleation mode particles thus suggesting a strong impact of CN property on exhaust emissions.