Combination of G-Equation and Detailed Chemistry: An application to 3D-CFD hydrogen combustion simulations to predict NOx emissions in reciprocating internal combustion engines

Abstract

In the recent years, the growing pressure by the European Union to phase out the internal combustion engines has raised the quest for alternative solutions for low-environmental-impact mobility. Nevertheless, concerns on the life-cycle emissions of battery electric vehicles and perplexities on the socio-economic sustainability of the ecological transition suggest that maintaining the interest in internal combustion engines can be strategic, provided that carbon-neutral fuels are adopted. On the basis of the technological neutrality principle, relying on already existing and well-established technologies requires less effort and cost to convert the whole road transport. Moreover, the adoption of bio- or e-fuels obtained from renewable sources widely spread across the globe is not of secondary importance. In fact, cost reduction and worldwide diffusion of the resources are both main promoters of socio-economic sustainability.

In this scenario, green hydrogen represents one of the main solutions for the survival of reciprocating engines. Since the production is solely based on renewable energy sources, it is not simply characterized by zero CO₂ emissions at the tailpipe, but it can be considered overall carbon neutral. A technical drawback in the use of hydrogen is represented by emissions of nitrogen oxides (NO_x), due to the ever-present high temperature combustion process. For this reason, an ad-hoc design is mandatory to minimize NO_x production, and CFD can be a valid tool to reduce cost and time to market for the development of hydrogen engines.

In this regard, the current work proposes a 3D-CFD numerical methodology, based on the combination of G-Equation and Detailed Chemistry models, for NO_x prediction in in-cylinder simulations of reciprocating internal combustion engines fueled with hydrogen. Although the combination of level-set method and chemical kinetics is not a novelty in literature, it is the first time that it is applied to evaluate NO_x emissions in H₂ engines. The proposed approach is validated against experimental data on a direct injection, spark ignition, hydrogen engine. The methodology is able to properly predict NO_x emissions at different mixture qualities, revving speeds and spark times. The total number of investigated cases is 17, which is a large set of simulations compared to the existing literature. Considering the best chemical mechanism (i.e. the one providing the best results among the tested ones), the error in the NO_x prediction is always lower than 25% for all the simulations.

Once the methodology is validated, the effect of spark and injection timings on NO_x is discussed. Such a deepening is useful to emphasize the potential of the CFD to investigate phenomena leading to emission formation and, thus, to optimize engine parameters for NO_x reduction.