TP 6 – Experimental investigation and development of reduced models on the influence of cyclic variations on knocking combustion

Knocking combustion is a major limiting factor for further increases in efficiency in gasoline engines. This phenomenon therefore remains the subject of current research work and a comprehensive, fundamental understanding is indispensable for tapping further potential.

The auto-ignition processes in end gas areas are often located in the immediate vicinity of the combustion chamber wall and are therefore not directly influenced by the primary flame front. Studies have shown that locally increased mixture temperatures by even a few Kelvin compared to the surrounding gas may be sufficient to cause the development of hot spots. Additionally, low temperature kinetics in the range of 800-900K are decisive for the self-ignition behavior and differences in anti-knock properties of different fuels can be attributed to that.

In the recent past, various measures have been investigated in research studies, with which the knock limit can be extended to higher loads. These include water injection and also the use of cooled, external exhaust gas recirculation (EGR) at full load. In addition, numerical models have also been successively further developed, which allow a more precise prediction of the knock limit in engine process calculations. Nevertheless, research projects that investigate the influence of the characteristic cycle-to-cycle variations on the knock limit in detail are still little documented in literature.

Project Goals

There are four main goals for VKAs part in this project. The first goal is the experimental analysis of the relationship between cyclic variations and knocking combustion processes. A combination of novel measurement techniques allows a directional determination of the knocking events in the combustion chamber while the incidence of self-ignition is correlated with the combustion process. Furthermore, the generated measurement data provides valuable validation data for the high-resolution numerical simulations performed in TP7 by our colleagues from TU Darmstadt, so that a jointly performed single cycle based backward analysis of the influence of cyclic variations along the engines multiscale chain of effects is made possible.

The second goal is to investigate and describe the influence of different fuels on the manifestation of cyclic variations and on the resulting development of the first self-ignitions. Here the combustion processes of standard RON95E10, pure Isooctane and two different four-component surrogate fuels are analyzed and evaluated in comparison. A special focus lies on the fuel sensitivity and its effect on the onset of early knock events.

The third goal is the derivation of suitable correlations to extend existing models in 0D/1D motor process calculation. Based on both the statistical and single cycle experimental data, knowledge about local cyclic variations in the form of intensity and direction within the combustion chamber will be used to validate and improve existing models.

The fourth goal is to gain an understanding of the flow and mixture conditions in the area near the spark plug for homogeneous operation with direct injection. A stable and reproducible flow to the spark plug area is essential for early flame propagation and therefore has direct influence on self-ignition. In order to demonstrate this correlation between the local mixture around the spark plug, flame propagation and self-ignition in relation to the influencing parameters, various optical measurement techniques will be used on a non-fired transparent test engine.

Research approach

To determine local differences in pressure in the combustion chamber and localize weak knock events, a novel optical measurement system is being developed. State of the art knock localization systems mainly use fiber-optical spark plugs (FOSP) and photomultiplier arrays to localize knock events. Due to the turbulent flame structures occupying the combustion chamber during later stages of combustion and therefore interfering with the optical measurement, low intensity knock events are very hard to detect. Furthermore, FOSP are highly expensive, very delicate and often in a different heat range than non-optical spark plugs. This influences early flame propagation which in turn affects knock intensity and location.

Due to these limitations a new fiber-optical insert (FOI) is being developed together with the Chair for Technology of Optical Systems TOS at RWTH, which will be inserted through the pressure indication bores in the cylinder head of the used SCRE. Redundant coverage of the complete end gas region with a high number of optical channels in combination with a newly developed detection algorithm based on machine learning will lower the detection limit of conventional FOSP and enable the investigation of the aforementioned influences of cyclic variations on weak knock events.

Currently, the first FOI prototype is in production and due to be tested on the test bench later this year. Thermodynamic evaluation of the behavior of different fuels in the SCRE is being conducted as this report is written and will be used to predetermine both boundary conditions for numerical investigations as well as help plan and optimize the larger optical measurement campaign planned for the beginning of next year.

Dr.-Ing. Marco G√ľnther marco.guenther(at)
Dr.-Ing. (USA), Universitätsprofessor Stefan Pischinger pischinger_s(at)
Jonathan Schneider schneider_jona(at)