Virtual development of a multi-cylinder hydrogen spark ignition engine operating at lean burn conditions

Propulsion systems for automotive applications are facing the issues related to the stringent noxious emission and CO2 regulations which are driving the manufacturer and the researchers to the development of innovative powertrains. Among the different solutions, relevant efforts are being directed towards the decarbonisation of internal combustion engines (ICEs). In this scenario, the hydrogen, particularly the green one obtained through the water electrolysis using renewable energy sources, is emerging as a promising fuel for a future sustainable mobility. In this work, a multi-cylinder hydrogen Spark Ignition (SI) engine conceived for automotive applications is developed and optimized to operate under ultra/lean air/hydrogen mixtures. In particular, a 3-cylinder turbocharged hydrogen SI engine is virtually developed through a 0D/1D model, starting from a naturally aspirated single-cylinder SI unit; the latter is properly validated with experiments and 3D CFD outcomes in a previous authors’ activity. A fractal model is used for combustion modelling, coupled to a user-defined turbulence sub-model. The thermo-diffusive instability sub-model, based on Howarth theory, is integrated into the combustion model to correctly simulate the hydrogen engine operation at ultra/lean mixtures. In addition, the knock onset is evaluated with a Tabulated Kinetic of Ignition (TKI) approach, requiring a pre-computation of the auto-ignition delay times at varying the thermodynamic conditions. The main aim of the work is to forecast the hydrogen engine operation under full load and ultra-lean conditions, analysing performance, efficiency and NOx emissions at the knock limit. Subsequently, an optimization of the engine operation is carried out exploring different compression ratios, air/fuel proportions (lambda values) and water-to-fuel ratios. The objective of the initial optimisation is to optimise efficiency while respecting the safety constraint on the occurrence of shocks. Comparison with the base case, in which the maximum value is 35.6%, reveals an increase of 1.98% with CR equal to 13. However, in the case of the high-speed analysis, there is a decrease compared to the base case. The second analysis shows that the maximum lambda value that meets the imposed BMEP target is 2.1, thus minimising the NOx emissions of a multi-cylinder hydrogen engine under all conditions. A final analysis was conducted to study the impact of water injection on NOx for fixed rpm.