Visualization of flame-fluid interactions and influence on kernel growth in a DISI Engine

The ever more demanding of fuel economy and emission standards require innovative control strategies for internal combustion engines (ICE) and new procedures for the use of alternative fuels as well as improved after-treatment systems. This development is intrinsically linked to improved understanding of in-cylinder processes, with complex dependencies between phenomena that are related to flow, thermal and chemical engineering. In this thesis, kernel formation and flame front propagation were investigated in an optically accessible Direct Injection Spark Ignition (DISI) engine. The optical analysis was made possible by a Bowditch design that features an extended piston combined with a quartz window mounted within the piston crown. The results pertaining to this work can be divided into two categories: the first one dedicated to the thermodynamic analysis, i.e. the analysis of parameters derived from in-cylinder pressure measurements (Indicated Mean Effective Pressure IMEP, Covariance of the same COVIMEP, Net Heat Release Rate NHRR, Mass Fraction Burned MFB, Volume Fraction Burned VFB etc..); the second part of the study is focused on the optical analysis of images related to the sequencies obtained through cycle-resolved digital imaging using a high-speed camera. Therefore, the two types of data were combined and compared to investigate how changing operative parameters affect the thermodynamic and optical results. Three different configurations (i.e. Cross-, Counter- and Uniflow with respect to the direction of tumble) in terms of spark plug orientations further emphasized the complex interaction between chemical phenomena and fluid motion. One of the main conclusions was that the spark plug orientation plays a significant role in flame kernel displacement and influences the direction of propagation of the flame front. Thanks to the large amount of experimental data acquired in the laboratory, a further goal of this research is to create a statistical database that contains detailed information on flame kernel formation and evolution. This database will constitute a solid basis for validation and implementation of a 3D CFD model capable of providing valuable information on the in-cylinder phenomena.