PhD defense - V. Granet

Is the large eddy simulation a suitable tool to predict cycle-to-cycle variations in spark ignition engines?

Delivered by Institut National Polytechnique de Toulouse
Speciality: Fluid Dynamics

September 20, 2011 - CERFACS

This PhD work, funded by the CIFRE program between Renault and CERFACS, is part of an overall will of the industrials to evaluate the ability of Large Eddy Simulation (LES) to predict cycle-to-cycle variations (CCV) in spark ignition engines. The improvement of spark ignition engines is a major challenge for engineers in order to produce more energy, to minimize fuel consumption and to reduce pollutant emissions. The new technologies, which appeared in the last years, bring the engines closer to their stability limit while increasing various unwanted phenomena. Among these phenomena, cycle-to-cycle variations need to be minimized in order to keep the performances as high as possible and prevent engine damage. In order to study CCV, LES has been used in the present manuscript to simulate a mono-cylinder spark ignition engine. This document highlights a numerical methodology for the prediction of CCV and numerous engine cycles were simulated by LES in order to validate the results in comparison to the experimental findings. LES seems to be able to capture stable and unstable (in terms of CCV) operating points of the engine. In addition, the sources of CCV are also analyzed and a modification of the engine is proposed to reduce CCV

J. Borée	LEA ENSMA
D. Veynante	Ecole Centrale de Paris
T. Baritaud	Ferrari
G. Bruneaux	IFPEn - ANRT
B.-O. Ducreux	ADEME
P. MOIN	Stanford University
V. Dugué	Renault
T. Poinsot	IMFT

Back to PhD Defenses main page