Date of Graduation


Document Type



Diesel engines constitute a valuable solution to the growing demand for higher fuel efficiencies, especially in relation to the rapid growth of fuel prices in recent years. However, the substantially lean conditions typical of diesel exhaust pose a serious challenge to the traditional exhaust aftertreatment devices. The Selective Catalytic Reduction (SCR) technology has proven to be a viable technology for reducing emissions of oxides of nitrogen (NOx) from heavy-duty diesel engines. The diesel particulate filter (DPF) significantly reduces particulate matter (PM); primarily, the solid carbon component of total PM. However, the absence of carbon, downstream of the DPF introduces new dynamic effects viz. PM formation, growth, and composition. Additionally, changes in the PM characteristics occur as the diesel exhaust flows through the SCR, which is generally located downstream of the DPF. It is hypothesized that the chemical and physical characteristics of the SCR-out PM are related to engine-out PM; hence, different engine calibrations could be engaged to minimize SCR-out PM emissions. The objectives of this study were to reduce mass emissions of NOx and PM, from a modified MY 2004 heavy-duty diesel engine to US 2010 by developing engine calibrations for a given DPF and SCR system. Specifically, urea injection control strategies were developed to optimize SCR operation, and measurements were made of PM concentrations and size distributions, morphology and chemical composition of the particulate matter emitted by the SCR-equipped engine. The response of the exhaust aftertreatment system was investigated over different engine calibrations under steady state (ESC) and transient (FTP) conditions. The particle number and size distribution was sampled with the Scanning Mobility Particle Sizer (SMPS) and with the Differential Mobility Spectrometer (DMS). Analysis of SCR-out particle morphology was conducted with the Scanning Electron Microscope (SEM). This study presents the development of a method to generate count median diameter (CMD)-specific particle size distributions based on only four engine parameters. The method makes use of design of experiments (DOE) and analysis of variance to generate an empirical model capable of CMD predictions. The predictions were found to be accurate within the confidence intervals. A correlation was found in the accumulation mode region between the different combustion strategies for the aftertreatment optimization and the particles detected downstream of the SCR. PM exhibited a morphology entirely different from the agglomerate particles typical of diesel particles. The SEM analysis revealed a bright and shining crystalline shape, and the ion chromatography analysis revealed the presence of sulfates. The PM concentration and size distribution of the engine-out PM produced during acceleration and deceleration event of the FTP cycle were found to be in agreement with those measured over steady state cycles.