Search Results

EGR dilution is a promising way to improve fuel economy of Spark-Ignited (SI) gasoline engines. In particular, at high load, it is very efficient in mitigating knock at low speed and to decrease exhaust temperature at high speed so that fuel enrichment can be avoided. The objective of this paper is to better understand the governing mechanisms implied in EGR-diluted SI combustion at high load. For this purpose, measurements were performed on a modern, single-cylinder GDI engine (high tumble value, multi-hole injector, central position). In addition 0-D and 1-D Chemkin simulations (reactors and flames) were used to complete the engine tests so as to gain a better understanding of the physical mechanisms. EGR benefits were confirmed and characterized at 19 bar IMEP: net ISFC could be reduced by 17% at 1200rpm and by 6% at 5000rpm. At low speed, knock mitigation was the main effect, improving the cycle efficiency by a better combustion phasing.

The 3-Zones Extended Coherent Flame Model (ECFM3Z) and the Tabulated Kinetics for Ignition (TKI) auto-ignition model are widely used for RANS simulations of reactive flows in Diesel engines. ECFM3Z accounts for the turbulent mixing between one zone that contains compressed air and EGR and another zone that contains evaporated fuel. These zones mix to form a reactive zone where combustion occurs. In this mixing zone TKI is applied to predict the auto-ignition event, including the ignition delay time and the heat release rate. Because it is tabulated, TKI can model complex fuels over a wide range of engine thermodynamic conditions. However, the ECFM3Z/TKI combustion modeling approach requires an efficient predictive spray injection calculation. In a Diesel direct injection engine, the turbulent mixing and spray atomization are mainly driven by the liquid/gas coupling phenomenon that occurs at moving liquid/gas interfaces.

In this paper three turbulent combustion models with different underlying hypothesis are compared with measurements from an extensive experimental database. The reference model is ECFM3Z, with the Tabulated Kinetics of Ignition (TKI) model for auto-ignition modeling, together with the CO reduced kinetics (CORK) model and the extended Zeldovich model for the nitrogen oxides. The VVTHC (Variable Volume Tabulated Homogeneous Chemistry) model predicts both the heat release and species evolutions (including CO). The most evolved model proposed is the ADF-PCM (Approximated Diffusion Flame-Presumed Conditional Moment) approach, based on the laminar flamelet equation of the progress variable. ADF-PCM and VVTHC are tabulated models based on a progress variable approach and are then coupled to the tabulated NO model NORA based on relaxation (NO Relaxation Approach). All the present combustion models are coupled to a phenomenological soot kinetics PSK approach.

In this paper, a sectional soot model coupled to a tabulated combustion model is compared with measurements from an experimental engine database. The sectional soot model, based on the work of Vervisch-Klakjic (Ph.D. thesis, Ecole Centrale Paris, Paris, 2011) and Netzell et al. (P. Combust. Inst., 31(1):667-674, 2007), has been implemented into IFPC3D (Bohbot et al., Oil Gas Sci Technol, 64(3):309-335, 2009), a 3D RANS solver. It enables a complex modeling of soot particles evolution, in a 3D Diesel simulation. Five distinct source terms are applied to each soot section at any time and any location of the flow. The inputs of the soot model are provided by a tabulated combustion model derived from the Engine Approximated Diffusion Flame (EADF) one (Michel and Colin, Int. J. Engine Res., 2013) and specifically modified to include the minor species required by the soot model.