Numerical Study of Mixing, Combustion and NOx Emissions in a DI High Performance Hydrogen Engine Operated at Stoichiometry

The roadmap towards carbon neutrality by 2050 makes necessary drastic reduction of road vehicle tailpipe carbon emissions. One viable approach to reach the abatement of carbon monoxide and dioxide is to fuel internal combustion engines (ICEs) with hydrogen. The burning of a hydrogen-air mixture inside the combustion chamber reduces to minimal amount the production of carbon emissions and particulate matter that are only produced by the presence of lubricant oil. However, the high temperatures reached by the end-gases promote the formation of nitrogen oxides. In high-performance ICEs, the pursuit for high-specific power by means of the adoption of stoichiometric mixtures is hindered by the need to reduce NOx - as this pollutant drastically drops when moving towards ultra-lean mixtures.

The paper aims to present a CFD-3D framework to simulate the full engine-cycle of a high-performance Spark-Ignited (SI) Direct-Injection (DI) ICE fuelled at stoichiometric conditions. The methodology is validated thanks to experimental data collected at part-load condition (2000 rpm and 4.5 IMEPH) changing the injection timing and at full-load operation (6000 rpm and 24 bar IMEPH). The experimental measurements of nitrogen oxides at the part load conditions are compared with the outcome of the CFD. In particular, the most delayed injection case is compared with the most advanced one to give an overview of the effect of the stratification of the mixture in the formation of the analysed pollutants. Furthermore, the validated model is employed to predict amount of NOx generated in the most demanding condition (6000 rpm) to better understand the influence of temperature, pressure and mixture composition on the NOx production pathways.