The specific aim is to validate an engineering model for direct injection (DI) Diesel engine emissions. Characteristic times describing the controlling fluid mechanics and chemical kinetics will be employed in the model to correlate both NOx and particulate emissions. Because the model equations are algebraic, they are suitable for implementation in a phenomenological cycle simulation program, or as an emissions model option in a computational fluid dynamics code.
An original premise was that earlier work on global NO chemistry based on pollutant emissions dominated by diffusion flame contributions had adequately elucidated the kinetic aspects of the model. It is shown here that this approach is not valid for modern engines. Rather, an improved two-zone flame model for NO formation/decomposition is required. Mellor et al. [1] propose such a model, but include only qualitative preliminary model validation. Here we further test the model using measurements from a high speed direct injection (HSDI) Diesel engine with a common rail fuel injection system. Exhaust gas recirculation variations from zero to maximum level possible are reported and correlated for parametric variations of engine speed, injection pressure, and load.
The two-zone NOx HSDI Diesel characteristic time model (CTM) involves a skeletal mechanism consisting of seven elementary reactions [1]. A surrogate stoichiometric flame temperature around a spray plume at start of combustion conditions represents the environment for NO-forming eddies in the first zone. The second zone surrounds the stoichiometric region, and NO decomposition may occur depending on a surrogate flame temperature at the end of combustion computed from fuel/air dual (i.e., limited-pressure) cycle analysis. Both Zeldovich and nitrous oxide mechanisms are considered in each zone.
Two objectives are accomplished in the present paper. First, in Mellor et al. [1] the N2O mechanism was postulated to be important to the NO formation kinetics under Diesel engine in-cylinder conditions. In addition, it was suggested that NO decomposition reactions may occur at high engine loads through the reverse N2O and Zeldovich mechanisms. These hypotheses are validated using the measured engine data. Second, we extend the previous work through preliminary development of the fluid mechanic aspects of the model. A characteristic mixing time that is a function of Reynolds number, Weber number, and engine geometry is developed. When used in combination with the kinetic time for NO formation as a Damköhler number, correlations of NOx emissions for the above parametric variations are obtained for conditions where NO decomposition does not occur.