A Computational Procedure for Predicting Nitrogen Oxide Emissions from Diesel Engines 2006-01-0240

The calculation of the Nitrogen Oxide (NO) formation emitted from diesel engines usually involve direct integration of a set of nitrogen chemistry elementary reactions that involve formation and destruction of NO. The primary hydrocarbon chemistry is usually simplified as long as the main species and heat release are predicted correctly. The result of the integration is the net NO formation rate evaluated using the local concentrations and thermodynamic parameters.

In the present work a method for calculating NO emissions from diesel engines is proposed that takes into consideration the effect of residence time as a measure of turbulence effects on chemistry. This is based on the assumption that for mixing-limited conditions the turbulent eddy turn-over time can be taken as a characteristic reaction residence time.

The proposed procedure depends on a detailed investigation of the primary hydrocarbon combustion chemistry decoupled from the flow-field prediction. The laminar NO formation rates calculated from the detailed chemistry are linked to the flow-field prediction via the stoichiometric conditional probability density function. This procedure assumes that the state of mixing that is significant to thermal NO formation is centralized around stoichiometric conditions. By this assumption, the convolution integral of the probability density function can be reduced to a single functional evaluation at stoichiometric conditions. The net NO formation rates can still be calculated from detailed or reduced mechanisms and stored in look-up tables or expressed as fitted correlations.

The procedure is implemented in KIVA3V code and used to predict NO emissions from a state-of-the-art medium duty diesel engine. The NO emissions prediction indicated the usefulness of this approach in speeding-up the calculations while retaining all of the necessary information about effect of mixing intensity on the chemistry of NO formation.