Emission standards for diesel engines in vehicles have been steadily reduced in recent years, and a great deal of research and development effort has been focused on reducing particulate and nitrogen oxide emissions. One promising approach to reducing emissions involves the addition of oxygen to the fuel, generally by adding an oxygenated compound to the normal diesel fuel. Miyamoto et al. [1] showed experimentally that particulate levels can be significantly reduced by adding oxygenated species to the fuel. They found the Bosch smoke number (a measure of the particulate or soot levels in diesel exhaust) falls from about 55% for conventional diesel fuel to less than 1% when the oxygen content of the fuel is above about 25% by mass, as shown in Figure 1. It has been well established that addition of oxygenates to automotive fuel, including both diesel fuel as well as gasoline, reduces NOx and CO emissions by reducing flame temperatures. This is the basis for addition of oxygenates to produce reformulated gasoline in selected portions of the country. Of course, this is also accompanied by a slight reduction in fuel economy. A new overall picture of diesel combustion has been developed by Dec [2], in which laser diagnostic studies identified stages in diesel combustion that had not previously been recognized. These stages are summarized in Figure 2. The evolution of the diesel spray is shown, starting as a liquid jet that vaporizes and entrains hot air from the combustion chamber. This relatively steady process continues as long as fuel is being injected. In particular, Dec showed that the fuel spray vaporizes and mixes with air and products of earlier combustion to provide a region in which a gas phase, premixed fuel-rich ignition and burn occurs. The products of this ignition are then observed experimentally to lead rapidly to formation of soot particles, which subsequently are consumed in a diffusion flame. Recently, Flynn et al. [3] used a chemical kinetic and mixing model to study the premixed, rich ignition process. Using n-heptane as a representative diesel fuel, they showed that addition of an oxygenated additive, methanol, to the fuel reduced the concentrations of a number of hydrocarbon species in the products of the rich ignition. Specifically, methanol addition reduced the total concentrations of acetylene, ethylene and 1,3-butadiene, as well as propargyl and vinyl radicals, in the ignition products. These are the same species shown in a number of studies [4-6] to be responsible for formation of aromatic and polycyclic aromatic species in flames, species which lead eventually to production of soot. Flynn et al. did not, however, examine the kinetic processes responsible for the computed reduction in production of soot precursor species. At least two hypotheses have been advanced to explain the role that oxygenated species play in diesel ignition and the reduction in the concentrations of these species. The first is that the additive, methanol in the case of Flynn et al., does not contain any C-C bonds and cannot then produce significant levels of the species such as acetylene, ethylene or the unsaturated radicals which are known to lead to aromatic species. The second hypothesis is that the product distribution changes very naturally as oxygen is added and the overall equivalence ratio is reduced. In the present study, we repeat the ignition calculations of Flynn et al. and include a number of other oxygenated species to determine which of these theories is more applicable to this model.

Qirui Yang develops a model chain for the simulation of combustion and emissions of diesel engine with fully variable valve train (VVT) based on extensive 3D-CFD simulations, and experimental measurements on the engine test bench. The focus of the work is the development of a quasi-dimensional (QDM) flow model, which sets up a series of sub-models to describe phenomenologically the swirl, squish and axial charge motions as well as the shear-related turbulence production and dissipation. The QDM flow model is coupled with a QDM combustion model and a nitrogen oxides (NOx) / soot emission model. With the established model chain, VVT operating strategies of diesel engine can be developed and optimized as part of the simulation for specific engine performance parameters and the lowest NOx and soot emissions.

In order to fulfil future emissions legislations, new combustion systems are to be investigated. One way of improving exhaust emissions is the application of multiple injection strategies and conventional or partially premixed combustion conditions to a Diesel engine. The application of numerical techniques as CFD supports and improves the quality of engine developments. Unfortunately, current spray and combustion models are not accurate enough to simulate multiple injection systems, being in this way a topic of research. The goal of this study was the development of a novel simulation method for the investigation of Diesel engines operated with multiple injection strategies and different combustion modes. The first part of this work focused in improving the spray modelling. The inform ation of 3D CFD simulations of the injector nozzle was introduced in the spray simulation as boundary conditions developing coupling subroutines for this issue. The atomisation modelling was also improved using validated presumed droplet size distributions. Moreover, to avoid the simulation of the injector nozzle for every investigated operating point, a novel interpolating tool was developed in order to create spray boundary conditions based on few 3D CFD simulations of the nozzle under certain initial and boundary conditions. The second part of this thesis dealt with the combustion modelling of Diesel engines. For this issue, a laminar flamelet approach called Representative Interactive Flamelet model (RIF) was selected and implemented. Afterwards, an extended combustion model based on RIF was developed in order to take into account multiple injection strategies. Finally, this new model was validated with a wide range of operating points: applying multiple injection strategies under conventional and partially premixed combustion conditions.

Completely eliminated with the in- cylinder combustion techniques until now, hence, after-treatment is still necessary to meet the present emission legislations. Also with the development of the new engines which have different combustion regimes such as Homogeneous Charge Compression Ignition (HCCI), Modulated Kinetics (MK), Premixed Charge Compression Ignition (PCCI), Low Temperature Combustion (LTC), other emissions such as HC and CO became significant for compression ignition (CI) engines. This study investigates mainly formation/reduction of NOx and soot emissions in diesel engine coinbustion, especially in Heavy Duty Diesel (HDD) engines with the help of CFD engine modeling of the engine. The KIVA-3VR2 and CHEMKIN packages were used for the modeling purposes. CHALMERS diesel oil surrogate (DOS) model represented by a blend of aliphatic (n-heptane, 70%) and aromatic (toluene, 30%) components, turbulence/chemistry interaction approach, Partially Stirred Reactor (PaSR) model, applied with the detailed chemical mechanism and modified spray models were implemented into the KIVA-3VR2 for the modeling tasks. Diesel surrogate oil and detailed chemical mechanism were validated with shock-tube experiments on ignition delays for different pressures, temperatures and air/fuel ratios. Then modeling results for Volvo D12C engine for two compression ratios (18.0 and 14.0) and two different combustion regimes, MK and LTC, were compared with the experimental data. The reaction mechanism is modified in order to improve its NOx-soot emissions behavior which was not accurate enough. Different fuel injection times, loads, and both EGR-free and EGR cases were studied to extend the modeling capabilities. For all cases presented modeling approach is used to predict in-cylinder pressure, temperature, Rate of Heat Release (RoHR), combustion efficiency, NOx and soot emissions. Although tendency ofthe predicted emissions is in a good agreement with the experiments, a quantitative improvement of emission predictions is still required. Accurate modeling based on the detailed chemistry approach requires a proper balance between NOx formation, soot and CO oxidations in the chemical mechanism which is not easy to achieve. Also a new scientific tool, parametric ( )T dynamic map analysis, to evaluate engine combustion and emission formation based on the detailed chemical model of the diesel oil surrogate fuel. Emission formation and combustion efficiency can be predicted with the usage of this new type of analysis. The consistency of the map technique is mature enough to use it as a common tool, to analyze the engine combustion and emission formation processes.