Prediction of Ultra-Lean Spark Ignition Engine Performances by Quasi-Dimensional Combustion Model With a Refined Laminar Flame Speed Correlation

The turbulent combustion in gasoline engines is highly dependent on laminar flame speed S_L. A major issue of the quasi-dimensional (QD) combustion model is an accurate prediction of the S_L, which is unstable under low engine speeds and ultra-lean mixture. This work investigates the applicability of the combustion model with a refined S_L correlation for evaluating the combustion characteristics of a high-tumble port gasoline engine operated under ultra-lean mixtures. The S_L correlation is modified and validated for a five-component gasoline surrogate. Predicted S_L values from the conventional and refined functions are compared with measurements taken from a constant-volume chamber under micro-gravity conditions. The S_L data are measured at reference and elevated conditions. The results show that the conventional S_L overpredicts the flame speeds under all conditions. Moreover, the conventional model predicts negative S_L at equivalence ratio ϕ ≤ 0.3 and ϕ ≥ 1.9, while the revised S_L is well validated against the measurements. The improved S_L correlation is incorporated into the QD combustion model by a user-defined function. The engine data are measured at 1000–2000 rpm under engine load net indicated mean effective pressure (IMEP_n) = 0.4–0.8 MPa and ϕ = 0.5. The predicted engine performances and combustions are well validated with the measured data, and the model sensitivity analysis also shows a good agreement with the engine experiments under cycle-by-cycle variations.