A Modeling Study of Cyclic Dispersion Impact on Fuel Economy for a Small Size Turbocharged SI Engine

In this paper, the results of an extensive experimental analysis regarding a twin-cylinder spark-ignition turbocharged engine are employed to build up an advanced 1D model, which includes the effects of cycle-by-cycle variations (CCVs) on the combustion process. Objective of the activity is to numerically estimate the CCV impact primarily on fuel consumption and knock behavior. To this aim, the engine is experimentally characterized in terms of average performance parameters and CCVs at high and low load operation. In particular, both a spark advance and an air-to-fuel ratio (α) sweep are actuated. Acquired pressure signals are processed to estimate the rate of heat release and the main combustion events. Moreover, the Coefficient of Variation of IMEP (CoVIMEP) and of in-cylinder peak pressure (CoVpmax) are evaluated to quantify the cyclic dispersion and identify its dependency on peak pressure position. In parallel, the experimentally tested engine is fully schematized in a 1D framework. The 1D model, developed in the GT-Power™ environment, makes use of user defined sub-models for the description of combustion, turbulence and knock phenomena. 1D analyses are carried out for various engine speeds, load levels, α ratios, and spark timings, without changing any tuning constant. In a first stage, the model is validated in terms of overall engine performance parameters, and ensemble-averaged pressure traces inside the cylinder, and within the intake and exhaust ducts, as well. A more detailed comparison is also performed with reference to the average rate of heat release in different operating conditions. In a subsequent step, the effects of CCVs are introduced in the model in terms of Gaussian distributed modifications of the burning rate, predicted with reference to the ensemble-average operation. Consistently with the experimental data, applied burning rate are modified to simulate a train of pressure cycles statistically equivalent to the measured ones. The influence of the CCVs on the instantaneous peak position, air flow rate, and Indicated Specific Fuel Consumption (ISFC) is consequently investigated on a cycle-by-cycle basis, and compared to the average operation. Numerical analyses show that CCVs cause a reduced ISFC penalty, which can be considered significant only in case of delayed combustions and increased CoVs. Knock limited spark advance is also identified with and without CCV, highlighting some additional fuel economy penalties.