A Physical-Based Approach for Modeling the Influence of Different Operating Parameters on the Dependency of External EGR Rate and Indicated Efficiency 2018-01-1736

External Exhaust Gas Recirculation (EGR) provides an opportunity to increase the efficiency of turbocharged spark-ignition engines. Of the competing technologies and configurations, Low-Pressure EGR (LP-EGR) is the most challenging in terms of its dynamic behavior. Only some of the stationary feasible potential can be used during dynamic engine operation. To guarantee fuel consumption-optimized engine operation with no instabilities, a load point-dependent limitation of the EGR rate or alternatively an adaptation of the operating point to the actual EGR rate is crucial. For this purpose, a precise knowledge of efficiency and combustion variance is necessary. Since the operating state includes the actual EGR rate, it has an additional dimension, which usually results in an immense measuring effort. With the objective of avoiding long measuring periods, the given contribution introduces a physically based approach that models the influence of different engine operating parameters on the dependency between the external EGR rate and indicated efficiency. This also implies the dependency between the EGR rate and combustion variance. Since the model addresses vehicle applications, it only uses input parameters that can be assumed to be known during vehicle operation, such as the valve timing, engine load, engine speed, air fuel ratio, intake manifold pressure or ignition timing. Knowing the correlation of these parameters along with the EGR rate, the indicated efficiency and the EGR tolerance allows the fuel consumption to be optimized while maintaining the stability limits. As part of this publication, the measuring program that is used to set up the model is explained and presented. The mathematical formulation is described and a comparison of the model and measurement data is presented. The model quality is evaluated in terms of specific parameter variations. Finally, the results obtained are discussed under the aspect of how well the model can be used for the virtual calibration of spark-ignition engines.