Biodiesel Spray Combustion Modeling Based on a Detailed Chemistry Approach

Doctoral thesis, 2012

Replacing conventional diesel fuels with biodiesel has the potential to drastically reduce engine-out emissions of soot, carbon monoxide, and total unburnt hydrocarbons, at the cost of slightly increased nitrogen oxide emissions. However, to realize the full benefits of using biofuels in this way, there is a need for theoretical models describing their combustion in internal combustion engines. While master models of this kind have been developed recently, they involve very large numbers of reactions and intermediates, making them difficult to study using current computing resources even when using one-dimensional flame modeling. In order to derive a more computationally tractable semi-detailed mechanism (~500 species/2000 reactions) for engine studies, the detailed mechanism has been subjected to a sophisticated reduction strategy based on methods such as Directed Relation Graph and its variants. The aim was to generate a reduced mechanism that would accurately reproduce the key features of the combustion process, including the fuel's auto-ignition behavior and the concentrations of the various intermediate species involved. The reduced mechanism was therefore validated against experimental data and results obtained using the master model for combustion under various simplified conditions, e.g. shock-tube, jet stirred reactor, laminar flame propagation, opposed-diffusion flame. To provide input data for flame modeling, the transport species properties of the fuels studied using the reduced biodiesel mechanism were estimated using semi-empirical methods. The estimated thermo-physical properties and the outputs of the spray combustion models were compared to real spray characteristics measured in a spray chamber. The liquid penetration, flame lift-off and average Sauter Mean Radius (SMR) for sprays of different fuels were calculated and compared. Ignition Quality Tester modeling was conducted to assess the Cetane Number of the fuels studied using the biodiesel oxidation mechanism and spray sub-models, e.g. spray atomization, droplet breakup and evaporation models. The results obtained indicated a potential direction for mechanism optimization since there was found a considerable uncertainty regarding the thermodynamics of the master mechanism. In addition, IQT modeling also provides a way to understand a relative influence of physical spray dynamics and combustion chemistry on the overall ignition delay for biodiesel. To study the impact of bio-diesel fuels on engine performance, 3D CFD simulations for both heavy duty (Volvo D12C) and light duty (Volvo NED5) engines were carried out under different operating conditions. Predictions were validated against measured in-cylinder parameters and exhaust emission concentrations. The semi-detailed mechanism was shown to provide an efficient and accurate representation of actual biodiesel combustion and emissions formation. Moreover, the engine modeling also suggested that using biodiesel in place of conventional diesel fuel increases nitrogen oxide emissions by 10%. Quantitative analyses of the in-cylinder temperature, equivalence ratio and nitrogen oxide distributions provided deep insights into the origin of these elevated NOx emissions. It was found that oxygen-containing fuels such as biodiesel increase the concentration of molecular and atomic oxygen present in the cylinder during combustion, and this is likely to be the cause of the elevated NOx emissions.