A numerical study of mixing phenomena and reaction front propagation in partially premixed combustion engines

Sammanfattning: This work treats the modelling of PPC engine combustion through 3D computational fluid dynamics. In order to develop a cost-effective CFD model for the design and performance analysis of PPC engines it is important to first understand the multiple modes of combustion in PPC engines, e.g., the onset of the first ignition kernels, the propagation of the ignition/flame fronts, and the effect of charge stratification on the combustion process. To this end the mixing process and incylinder turbulent flow are studied in the first part of the thesis. The results of the study, which are covered in papers I and III, show that the swirling flow structure generated in the intake stroke loses non-axial components in later stages, which supports the use of sector mesh in the combustion stroke CFD simulations. The sector simulations show similar behaviour in terms of spray induced momentum and fuel distribution compared to a full cylinder simulation as long as the mixing occurs far from the cylinder center. In the second part of the thesis, CFD simulations of the combustion process are performed, based on a direct coupling of the finite rate chemistry with the flow transport. Different models for turbulence/chemistry interaction are considered; the model based on well-stirred reactors and partially stirred reactors, and the model based on transported probability density function within the Eulerian Stochastic Fields (PDF-ESF). These models are used to analyse the mixed mode combustion process in an experimental PPC engine. The LES models are shown to be able to resolve sufficient scalar stratification to properly represent the combustion phasing of the ignition front combustion mode. While the LES-ESF shows better predictions of the required inlet temperatures, the model is computational more demanding than the well-stirred/partially stirred reactor models – the computational time increases with the number of stochastic fields.