Abstract
This work presents a systematic approach proceeding from the engine process simulation (one-dimensional (1D)-computational fluid dynamics (CFD)) and design of experiments (DoE) up to the experimental validation to build a dedicated exhaust-gas recirculation (EGR) system for a stationary four-cylinder naturally aspirated gas engine. This system should ensure an equal distribution of the recirculated exhaust gas, coming entirely from the rich-operated dedicated cylinder. The rich combustion enables an in-cylinder production of highly reactive species (mainly H2 and CO), resulting in increased EGR reactivity, which improves the dilution tolerance, leading to reduced wall heat losses and lower knock tendency in the EGR receiving cylinders. However, the EGR system design represents a challenge due to the pulsating exhaust gas flow from the dedicated cylinder, which leads to a considerable EGR maldistribution in the receiving cylinders. A numerical analysis of this effect demonstrated that the EGR distribution uniformity depends on the design and dimensions of the EGR path. Considering the numerous design parameters and taking into account that the optimum design of the EGR path is not necessarily the sum of optima from the one-factor-at-a-time variations, efficient DoE methodologies were applied. They enabled identifying the optimum set of the EGR path design-parameters, with an EGR rate maldistribution of about 1% points. To evaluate the quality of the numerical results, the dedicated EGR path with the optimum parameters set was built on the engine test bench. The experimental results confirm the simulative prediction accuracy, demonstrating the reliability of the pursued approach.
