CNG-DI-Engine at λ = 1-Operation with Highload-EGR

Gaseous fuels have a high CO₂-reduction potential and knock resistance due to their chemical and thermal properties, which make them an interesting fuel alternative for the use in modern spark-ignited engines. Within this research project the stoichiometric and homogeneous combustion process with compressed natural gas (CNG) direct injection is investigated in combination with high-load exhaust gas recirculation (EGR).

To study the flow characteristics and mixture formation of CNG injection in a combustion chamber, experimental investigations were performed in a low-pressure injection chamber and on an optically accessible single-cylinder engine. The CNG spray investigations in the low-pressure injection chamber were performed with the aim to macroscopically characterize the natural gas jet using the Schlieren technique and planar laser-induced fluorescence (pLIF). A 3D-CFD model was created and validated with the optical single cylinder investigations and a qualitative PIV method at low-end torque and catalyst heating conditions. Both central and lateral injection positions were investigated using PIV and these results proved a good compatibility with the 3D-CFD simulations. Both the optical investigations in the low-pressure chamber and with the motored, optically accessible single-cylinder engine show that the direct natural gas injection with the side injector position has advantages compared to the central position, since the primary and secondary tumble are increased by an early and late injection timing.

Experimental studies with the thermodynamic single-cylinder research engine show that the potential to control combustion using EGR is limited, however, an optimum combustion phasing could be maintained despite a lower combustion speed.

The dilution capability of CNG was investigated using both EGR and excess air using two compression ratios of CR = 13 and 14.7. The investigations with excess air dilution showed the highest in-crease in indicated efficiency. The CR = 13 configuration allowed an increase of ~2.5%-points and a maximum air/fuel-ratio of λ = 1.5.

A 0D/1D engine model was created using a predictive combustion model (SITurb) for the direct injection of natural gas and validated using measurement results. The investigations showed that low-pressure EGR cannot be used for high-load EGR with single supercharging due to reaching the com-pressor map limits. Therefore, a high-pressure EGR system was implemented.

A Ford Eco Boost Engine was used for the engine tests and modified for CNG operation. The full-load tests at IMEP = 23 bar showed that no reduction in the cylinder peak pressure could be achieved with cooled EGR at constant center of combustion. The indicated efficiency could be slightly in-creased up to n = 4500 1/min due to the reduced wall heat losses. NO_X emissions were significantly reduced over the entire speed range from 25–62% at 5% EGR. However, HC, CH4, and CO emissions increased simultaneously by 6–26%. No knocking occurred throughout the entire full-load test at IMEP = 23 bar. A trend line comparison indicates that the results are transferable to engines with higher peak pressure limits.

A longitudinal dynamics simulation of an RDE cycle indicated a reduction of CO₂ emissions by 22%.