Search Results

Some results of a systematic study of the effects of fuel and chamber gas properties on the transient evaporating spray mixing process are presented. The study uses an existing two-dimensional stochastic thick spray model. The results show that the combustion process in typical heavy duty, quiescent, DI diesel engines can be mixing limited rather than vaporization limited. In addition, the results show that the mixing process of a transient evaporating spray is characterized by the combined effects of fuel evaporation and its turbulent mixing with the surrounding air. In general, increasing the evaporation rate alone does not necessarily increase the fuel-air mixing rate. Furthermore, two dimensionless parameters have been used to quantify the relative effects of fuel and chamber gas properties on the transient spray evaporation process. Finally, through detailed comparisons between spray and gas jet results, the transient evaporating spray mixing process is better understood.

Fundamental aspects of performance and regeneration of a porous ceramic particulate trap are described. Dimensionless correlations are given for pressure drop vs. flow conditions for clean and loaded traps. An empirical relationship between estimated particulate deposits and a loading parameter that distinguishes pressure drop changes due to flow variations from particulate accumulation is presented. Results indicate that trapping efficiencies exceed 90% under most conditions and pressure drop doubles when particulate accumulation occupies only 5% of the available void volume. Regeneration was achieved primarily by throttling the engine intake air. For various combinations of initial loading level, trap inlet temperature and oxygen concentration, it was found that regeneration rate peaked after 45 seconds from initiation.

A semiempirical, mathematical model describing the formation of nitric oxide in direct-injection diesel engines is derived. The model is used in conjunction with injection and thermodynamic cycle simulation programs. This approach enables prediction of nitric oxide emissions from design dimensions and operating parameters only, without the use of experimental data. Predicted results are compared with experiments for typical naturally aspirated and turbocharged engines. The accuracy of prediction is very good except under light-load naturally aspirated conditions. The model is used in an extensive parametric study, together with experimental verification. The agreement between prediction and experiments is excellent, except under conditions of excessive smoke or of high swirl.