The post-flame oxidation of unburned hydrocarbons (HC) released from the ring-pack crevice in spark-ignition engines was investigated. Experimental data was collected in two engines running at steady-state, fully warmed-up condition, fixed load and fixed speed. Nitrogen was injected into the intake manifold to dilute the charge in order to maintain the average burned gas temperature as the equivalence ratio varied. For a given operating condition, similar in-cylinder pressure and burned gas temperature profiles were obtained at different equivalence ratio.
Two models for post-flame oxidation were proposed: 1) a mixing-controlled model, based on an empirical correlation of two parameters, and 2) a kinetically controlled model, based on a zero-dimensional model using detailed chemical kinetic.
The mixing-controlled model consisted of a linear correlation between the peak mass flow rate of crevice gas returning to the combustion chamber and a post-oxidation metric defined as Global HC Consumption Rate (GCR). Using this correlation, the engine-out HC emissions were estimated. In general, the results were satisfactory, but the main shortcoming of the model was that the constant of the linear correlation needed to be derived empirically for each engine and each air-fuel ratio.
A kinetically controlled model was developed to study the effect of the mixture composition and temperature in the post-oxidation process. To account for the mixing between the unburned crevice gas and the burned gas in the cylinder, two approaches were taken: 1) an optimized mixing rate based on a constant (with time) value for the mass of burned gas entrained into an unburned gas parcel, and 2) a time-varying mixing rate, in which the crevice outgassing process was modeled using a self-similarity solution for both a turbulent round jet and a wall jet. Using both approaches, the engine-out HC emissions were calculated. The results were compared to the measured HC emission. In general, satisfactory (or expected) results were found for cases in which the engine was run rich or near stoichiometric. For the lean combustion cases, the results were unsatisfactory (or unexpected) with both mixing approaches. A hypothesis that explain the discrepancy in the lean combustion side was suggested.
For a given operating condition, the potential to reduce actual engine-out HC emission was identified. The results showed that during the crevice outgassing process, there was a relatively narrow period of time in which it is feasible to increase the overall rate of post-oxidation. It is suggested that improving the mixing rate during this window of time can boost the level of post-oxidation within the cylinder and subsequently reduce the engine-out HC. This insight can be used to develop strategies for engine-out HC emission reduction by focusing on the period of the engine cycle that matter the most.