Validation of a multidimensional CFD approach for ethanol-fueled spark ignition engines at knock-limited conditions

• G-Equation model overpredicts combustion phasing for neat ethanol SI operation. • Detailed chemistry solver accurately predicts ethanol SI combustion. • Lower Mach CFL number recommended to capture end-gas knock. • Detailed chemistry models necessary for developing diagnostic tools for SI engines. • Demonstration of CFD approach to capture knock at advanced spark timings. The alcohols (ethanol and methanol) are promising fuel substitutes for high efficiency, low emissions spark ignition (SI) engines due to their favorable fuel properties. The fundamental feature of knock imposes limitations on the compression ratio and/or combustion phasing in SI engines. Knock is a complicated process that presents challenges for both experimental and simulation studies. Current research is aimed at exploring the gap between ethanol combustion and knock prediction in SI engines. Multidimensional computational fluid dynamics (CFD) simulations were conducted to explore the choice of the combustion model and Mach Courant-Friedrichs-Lewy (CFL) number to predict SI engine operation at a knock-limited spark advance (KLSA) condition fueled with neat ethanol (E100) and hydrous or “wet” ethanol (WE92) containing 8% water by mass. The validation of the G-Equation model and a detailed chemistry solver was informed by the comparison of bulk cylinder pressure and heat release rates against experimental data. The G-Equation model predicts advanced combustion phasing whereas cylinder pressures based on the detailed chemistry results were found to be in good agreement with the experimental data for the neat ethanol operation. The recommended choice for the detailed chemistry solver is further highlighted by the excellent validation observed for the hydrous ethanol operation. The necessity of a lower Mach CFL number to qualitatively capture knocking trends with reliable accuracy is demonstrated on the basis of knock indices developed for local points in the end gas. The simulation validation characteristics highlight the success of the detailed chemistry solver in reliably predicting SI engine performance with a recommendation for a lower Mach CFL number to successfully capture pressure oscillations associated with end-gas knock. A spark timing sweep is performed to further demonstrate the ability of the CFD model to represent different modes of SI engine operation.