The trend of implementing tighter emission regulations has led to increased efforts in Diesel engine emission control development. On one hand, alternative combustion mechanisms such as LTC, PCCI and HCCI have been introduced to reduce NOx and PM emissions. On the other hand, different aftertreatment devices are developed to control Diesel engine emissions. Even though these emission control techniques are very effective in engine-out emission reduction, there are inevitably part-to-part and cylinder-to-cylinder variations in terms of combustion and emissions. And this variability has the potential to adversely affect combustion, and hence torque and emissions production. The present work numerically investigates the sources of variability in a Diesel engine and performs injection optimization to balance cylinder-to-cylinder variations and reduce NOx.
A commercial engine simulation software GT-Power is used to conduct the numerical investigation. This software employs high-fidelity engine models and has the ability of predicting engine performance and emissions. Three sources of variability in Diesel engines, i.e. design-driven, component-driven and environment-driven sources of variability, are studied. In the design-driven variability analysis, cylinder-to-cylinder variations of air, residual, temperature, NOx and ISFC are investigated. It is found that there are non-negligible variations between cylinders. And maximum variations exist between cylinder 1 and cylinder 6. In the component-driven variability analysis, a design of experiments is created by varying six component related factors. Sensitivity analysis is performed to study the effect of these factors on NOx emission and ISFC. It is observed that these factors have different impacts on NOx and ISFC at different operating conditions. Overall, EGR fraction and boost pressure are more significant than injection pressure and swirl ratio on NOx. Boost pressure, SOI and fuel quantity are three most significant factors for ISFC. In the environment-driven variability analysis, four environmental factors are selected for analysis. The results show that DPF back pressure has dominant effect on NOx. At low speed and low load condition, ambient pressure has same level of effect as DPF back pressure on NOx.
Injection optimization for cylinder-to-cylinder variations elimination and NOx reduction is also investigated in this work. Fuel bank control and cylinder-by-cylinder control are considered for the optimization. Three types of design of experiments are created to extensively study the benefits of these two control strategies. A fault injection scenario is also included. It is found that fuel bank and cylinder-by-cylinder control are effective in NOx reduction. Cylinder-by-cylinder control shows advantages of balancing variations across cylinders and maintaining benefits in reducing NOx at fault injection mode as compared to the fuel bank control.