As a starting point for the project work, multi-scale CFD tools able to simulate a whole engine, i.e. including the combustion chamber (cylinder) as well as the full intake and exhaust ducts, were developed. They were based on coupling 1D-CFD codes, which provide an adequate prediction of the flow in the intake and exhaust lines, with 3D-CFD codes using the innovative LES technique, which is necessary to accurately reproduce the complex interactions inside the cylinder between the turbulent flow and combustion chemistry which is key factor for CCV. The results achieved allow simulating the whole engine, which is necessary to accurately account for the numerous sources of CCV, which are as much related to the flow in the combustion chamber, than yo the flow in the rest of the engine.
The resulting CFD tools were applied to study CCV in 3 types of gasoline engines: A port fuel injection spark-ignition engine, a direct injection homogeneous charge spark-ignition engine and a CAI engine. Multi-cycle simulations covering between 10 and more than 20 consecutive engine cycles were achieved, requiring important computational resources. The predictions were compared with experimental data concerning flow and combustion, validating the developed approaches. These comparisons also allowed confirming that the employed innovative LES approaches were able to reproduce the experimentally observed CCV levels in terms of cylinder pressure. A key element of the overall project work was to undertake detailed analysis of the important data volume gnerated by these complex simulations, in order to explore in detail the main causes and effects of CCV in the 3 different engine types of the project. It was found that for the first 2 engines of the homogeneous charge spark-ignition type, CCV was mainly the result of the turbulent nature of the engine flow and of its tumbling nature, leading to small cyclic variations of flow conditions. They are amplified owing to the non-linear nature of combustion chemistry, the amplitude of CCV being dependent on the relative scales of turbulence and chemistry for a given engine operating condition. The CCV in the CAI engine were found to result from interactions between the fuel spray and intake flow, and to depend on thermal effects resulting from important IGR values and on auto-ignition chemistry. It was shown that the dependency between consecutive cycles was much more important than in the other 2 studied engines.
This was complemented by research aimed at a detailed exploration of local factors playing a role in the appearance of CCV. Interactions between a turbulent flow and early flame kernel growth after spark-ignition and the effects of mean convection and turbulence close to the spark plug on ignition and combustion were studied using Direct Numerical Simulations. This supported the modelling and analysis work throughout the project.
The resulting improved understanding of CCV in gasoline engines was capitalised by formulating and validating different reduced CCV models able to render their effects in 1D CFD and system simulations. They were shown to allow reproducing CCV in multi-cycle system simulations over a whole engine map, with approaches ranging from pure mathematical models as Wiebe functions, up to approaches based on reduced 3D physical combustion models. The latter were found to be able to reproduce CCV for different engines without the necessity to repeat expensive LES studies, owing to their predictive capabilities. At the same time their real-time ability was demonstrated.
Finally, case studies concerning full engine and vehicles allowed exploring how using the developed reduced CCV models could potentially improve the fuel efficiency on typical drive cycles, via a better knowledge of the limits of the engine operation related to CCV. Potential future fuel consumption benefits between 1% and up to 5% were estimated.