Combustion and soot formation modelling in modern gasoline direct injection engines using computational fluid dynamics

This thesis investigates the use of Computational Fluid Dynamics (CFD) to model the formation of soot precursors in modern direct injection gasoline engines, with a particular emphasis on statistically efficient calibration of key sub-models. The research addresses the challenges associated with accurately representing combustion processes, liquid fuel films on cylinder surfaces, and surrogate fuel formulations within CFD frameworks. A novel evaluation and calibration approach is developed for the spray-wall interaction “Bai-Onera” model, based on published experimental results, which delivers accurate fluid film formation from fuel injections to the end of the engine cycle. A novel “mapping” approach is proposed to allow dual-fuel simulations within the same simulated engine cycle to capture all relevant mixture formation and combustion related phenomena. Statistical methods are employed to systematically calibrate and validate the “G-Equation” combustion and ignition model parameters against experimental data while minimizing the computational cost typically associated with conventional trial-and-error approaches. The developed methodologies are validated in an investigation of two single engine variable sweeps to highlight the capability and advantages of the new approach. Additionally, simple correlations are built harvesting simulations data to offer easier combustion calibrations and soot precursors prediction capabilities. The work presented in this thesis advances the reliability of CFD as a diagnostic and predictive tool in engine development, contributing to the optimization of clean and efficient combustion technologies.

Grantor dc:publisher

Oxford Brookes University

Author dc:creator

• Biagiotti, Federico

Contributors dc:contributor

• Bonatesta, Fabrizio
• Morrey, Denise
• Yang Changho

Statement dc:rights

• All rights reserved

Language dc:language

en

DOI dc:identifier

https://doi.org/10.24384/fr2e-xe08

OAI identifier oai:identifier

tle:acd30d9a-8c56-45e3-8467-9928133e2d05:d6bd9758-527a-46cd-bfe2-c433766e8fca:1