A Chemical-Kinetic Approach to the Definition of the Laminar Flame Speed for the Simulation of the Combustion of Spark-Ignition Engines

The laminar burning speed is an important intrinsic property of an air-fuel mixture determining key combustion characteristics such as turbulent flame propagation. It is a function of the mixture composition (mixture fraction and residual gas mass fraction) and of the thermodynamic conditions. Experimental measurements of Laminar Flame Speeds (LFS) are common in literature, but initial pressure and temperature are limited to low values due to the test conditions: typical pressure values for LFS detection are lower than 25 bar, and temperature rarely exceeds 550 K. Actual trends in spark ignition engines are to increase specific power output by downsizing and supercharging, thus the flame front involves even more higher pressure and temperature since the beginning of combustion. The most widespread models used to extrapolate the experimental data to the engine like conditions are derived from that of Metghalchi and Keck, but they often fail to correctly predict LFS values outside the experimental space. Thanks to the development of accurate chemical kinetic models together with the increase of computer performance, it is possible to numerically predict the laminar flame speed over a wide range of conditions for a range of fuel mixtures, so to overcome some of the limitations of the Metghalchi and Keck model. The aim of the present work is to evaluate the effectiveness of an exploitable open source chemical solver (Cantera) for the evaluation of laminar flame speed. Results are compared against experimental data available in scientific literature and a review of the main analytical correlation for LFS is accomplished. Finally, a new correlation is proposed to better fit the numerical results in the high pressure and temperature range of the real engine desig