Mixture preparation and combustion in a GDI engine under stoichiometric or lean charge: an experimental and numerical study on an optically accessible engine

In direct injection spark ignition (SI) engines, the spray dynamics and interaction with the surrounding air flow are recognised as strongly affecting mixture quality, combustion development and amount of pollutants released at the exhaust. Spray impact against walls has also to be considered since undesired deposition of liquid as wallfilm causes unburned hydrocarbons and soot formation, hence increased fuel consumption and even augmented cyclic dispersion. Present work aims at clarifying the dynamics of sprays generated by multi-hole high pressure injectors in the combustion chamber of a gasoline direct injection,(GDI) engine, as well as to characterize the combustion development and pollutants formation under various injection modes. Stoichiometric and lean operations are both studied into detail through a combined experimental and numerical approach. Experiments are conducted on an optically accessible engine, whereas numerical simulations are made after the development and validation of a three-dimensional (3D) sub-model reproducing at the best the spray dynamics also in its impact over walls. Early injection is shown to anyway determine at spark timing slight inhomogeneity in the equivalence ratio distribution of the stoichiometric charge, which makes for the flame propagation being anyway not spherical, but such to exhibit a preferential direction towards the richest zone of the combustion chamber. This leads to the formation of a pocket in the end gases where the knocking phenomenon is likely to occur under some circumstances. Lean charge operation is instead discussed as injection pressure and spark timing are varied to highlight mechanisms of formation of the main pollutants and to define routes for the development of proper engine control strategies. A deep insight in the in-cylinder thermo-fluidynamic processes is achieved thanks to the here followed synergic experimental-numerical approach. (C) 2016 Elsevier Ltd. All rights reserved.