Numerische Untersuchung unkontrollierter vorzeitiger Selbstzündprozesse unter motorrelevanten Bedingungen

In the present work premature and uncontrolled auto-ignition phenomena in boosted spark ignition direct injection (DISI) engines also referred to as „pre-ignition“ or more commonly as „low-speed pre-ignition (LSPI)“ have been investigated. The work focuses on the identification of the fundamental processes triggering LSPI in order to provide a deeper understanding of this irregular combustion mode and enabling the development of mitigation strategies. LSPI represents an irregular combustion phenomenon that is associated with modern high efficiency engine concepts, i. e. downsized DISI engines, posing a serious challenge in SI engine development. The premature onset of combustion, occurring before the spark plug fires, generally provokes severe engine knock. The most interesting feature of LSPI is the fact that it may occur in bursts consisting of several pre-ignition events alternating with regular cycles. In recent years LSPI has become a prominent issue in SI engine development. A considerable research effort has been spent to unfold the underlying mechanisms of this combustion phenomenon driven by the following reasons: First of all LSPI has a fatal impact on engine durability. Moreover, a common hypothesis assumes that the occurrence of LSPI indicates that a critical limit in engine efficiency has been approached. A combined research approach is applied covering theoretical considerations, numerical simulations of auto-ignition processes, three dimensional CFD computations as well as experimental investigations in a production engine. The research project has been carried out in collaboration with the department of engine research (IFKM) at KIT. The present thesis focuses on the numerical investigations of auto-ignition processes under engine relevant conditions based on a detailed description of the chemical kinetics and the molecular transport processes. The CFD simulations as well as the experimental part have been performed at the IFKM. In a first step potential triggers of LSPI have been identified taking into account the fundamental aspects of auto-ignition chemistry and the governing processes of mixture formation inside the combustion chamber. The conditions that enable a premature auto-ignition in the different potential development scenarios have been determined using numerical simulations. In addition, the sensitivity of auto-ignition with respect to the governing parameters has been analyzed. The comparison of the required conditions predicted by the simulation and the evidence obtained from experimental investigations and CFD computations enables an assessment of the potential of the considered mechanism to trigger LSPI. Considering the results obtained by applying the described approach the majority of the potential sources of LSPI were found to be ineffective. Auto-ignition provoked by incandescent solid particles was identified as the most likely agent of LSPI. Such particles are either formed during the incomplete combustion of droplets potentially originating from a liquid reservoir inside the piston top land crevice or due to the flaking of combustion chamber deposits. In addition the hypothesis of a particle driven LSPI occurrence provides a probable explanation for the observed bursts of alternating pre-ignition and regular combustion events. On the basis of the suggested triggering mechanism of LSPI potential mitigation strategies have been derived. Crucial prerequisites for the occurrence of LSPI are the build up of a liquid reservoir in the piston top land gap as well as a rather inefficient gas exchange. The formation of the liquid reservoir may be hindered by adjusting the spray targeting properly and hence reducing spray impingement on the liner. Improving the scavenging reduces the amount of residual gas and decreases the probability of incandescent particles in the compression stroke. The present thesis provides valuable insights in the fundamental mechanisms provoking LSPI in boosted DISI engines and represents a promising basis for further studies in the field of LSPI. The suggested further experimental investigations may provide some of the missing links to promote a fundamental knowledge of LSPI. Moreover, the proposed mitigation strategies may reduce the issues associated with LSPI in production engines