Numerical simulation of a DISI engine with a reduced chemical kinetic mechanism for gasoline–ethanol blends

R. M. Braga,F. S. Cota,C. M. Martins,M. G. J. Vaz,G. Q. Hindi,J. G. C. Baeta,R. Huebner
DOI: https://doi.org/10.1007/s40430-024-05344-5
IF: 2.361
2024-12-17
Journal of the Brazilian Society of Mechanical Sciences and Engineering
Abstract:The urgent need for decarbonizing the transportation sector and combating global warming has prompted countries to seek effective solutions considering their energy matrices. In Brazil, ethanol plays a crucial role on achieving this goal with the existing flex-fuel vehicles and the well-established infrastructure. In this context, having accurate models for predicting the engine performance and helping understand complex phenomena is very desired for both industry and academy. In this paper, a reduced chemical kinetic mechanism for gasoline/ethanol blends, having 75 chemical species and 343 reactions, was implemented and validated in 0D/1D simulations for laminar flame speed and ignition delay time. Expanding this study to 3D CFD simulations of a direct injection spark ignition engine, the mechanism was validated for fuel blends varying from 22% in volume of ethanol added in gasoline to neat ethanol (named as E22, E27, E50, E85 and E100) under stoichiometric air–fuel mixture and partial load conditions. The numerical in-cylinder pressure, heat release rate and specific NOx emissions agreed reasonably with the engine test data, although for ethanol fuel rich a slightly delay in combustion phase was predicted and for gasoline-rich fuel a propensity to knock was seen in the simulation results when a hot exhaust valve temperature was used. A sweep in exhaust valve wall temperature was carried out from 525 to 450 K and in the latter condition, the knock was eliminated. The investigated mechanism has proven to deliver reasonable results in a relatively short time interval, which can be suitable for industry applications.
engineering, mechanical
What problem does this paper attempt to address?