Diesel Combustion and Transient Emissions Optimization Using Taguchi Methods

The optimization of diesel engine design parameters for low emissions using Taguchi Methods is described in this paper. Monte Carlo simulation was performed to predict production emissions distributions. A multi-cylinder development engine equipped with hydraulically-actuated electronically-controlled unit injector (HEUI) fuel system was used for the EPA HD transient cycle tests.

Effects of the engine design parameters on transient emissions were established. The key parameters affecting emissions were identified through analysis of variance. An optimum parameter setting was determined. Transient cycle emissions at the optimum setting were predicted and confirmed through additional confirmation runs. Taguchi methods were found to be effective in simultaneously optimizing multiple combustion parameters and identifying the effect of each parameter on transient cycle emissions. Monte Carlo simulation proved to be a useful statistical tool for predicting production emissions variability.