摘要
A methodology to define proportions of compounds in gasoline surrogates dedicated to auto-ignition modelling in engines is described. Based on auto-ignition properties and global atom content of the reference gasoline, this methodology provides surrogates nearing gasoline auto-ignition in engine but also gasoline density or lower heating value. Increasing the number of surrogate compounds induces a convergence of the mixture defined according to the methodology towards the classification into chemical families of the gasoline to emulate, which is a key for pollutant emission representativity. Applied to a European unleaded gasoline with an octane rating of 95, a ternary surrogate made of 13.7%of n-heptane (by mole fraction), 42.8%of isooctane and 43.5%of toluene is obtained. The direct experimental comparison of this surrogate and of the reference gasoline in an engine operated in Controlled Auto-Ignition (CAI) combustion mode confirms the global validity of the approach (auto-ignition timing but also power output, efficiency and pollutant emissions). The selected operating point for this comparison is also modelled with 3D CFD to demonstrate that the final aim of gasoline emulation in engine codes with such a surrogate can be successfully reached.