Cerfacs Enter the world of high performance ...

Modeling of the formation of soot in gas turbines taking into account the morphology of the aggregates

https://youtube.com/live/0Rfs-vgnc7o?feature=share

In a context of facts based evidences of climatic disruption boosted by the human activities, some industrial and institutional actors focus on the de- velopment of technological solutions to mitigate the impact of mankind on global warming. Reducing man related pollutant emissions is a major lever and is currently thoroughly studied. While alternatives sources of energy start to arise, those struggle to find their way in different industries such as aviation and fossil fuel combustion is expected to remain the major energy source for aviation in the upcoming decades. To diminish its pollution, increasing combustion efficiency and lowering the related pollutant emission are thus interesting short to medium term levers. Amongst all the pollutants emitted by aviation, soot particles are an interesting topic of study since they play a short term role in climate change because of their short lifespan. While the understanding of these particles is quite poor, the topic has been recently pop- ularized and work began on their modeling, boosted by the growing computational capabilities, improvements in combustion and aerosol modeling as well as new experi- mental methods. In recent years, Large Eddy Simulations (LES) of turbulent sooting flames became more and more common while detailed kinetic modeling of such flames became conjointly computationally affordable. Besides, quantitative measurements of soot population in complex flames gained accuracy which enabled for the compar- ison of the results obtained both experimentally and numerically, aiming at a better modeling hence better understanding of soot particles properties to help reduce their emission. In this context, this work focused on the detailed modeling of sooting flames based on detailed combustion mechanisms coupled with detailed soot chem- istry modelling. A soot solver based on the Discrete Sectional Method (DSM) was implemented in a pre-existing chemical Solver (Cantera) to validate state-of-the-art kinetic mechanisms against experimental data on canonical flames in terms of soot production. Similarly, jet fuel models at varying levels of accuracy are challenged against similar flames. Next, the chosen mechanism is reduced with the selected jet fuel surrogate in an Analytically Reduced Chemistry (ARC) paradigm using the code ARCANE to make it compact enough to be used for the LES of a turbulent aeronau- tical flame while maintaining its accuracy on soot prediction. This newly developed mechanism is challenged against experimental measurements on an aeronautical-like Rich-Quench-Lean (RQL) burner to asses its accuracy as well as the improvements in comparison with previous methods. The new mechanism proved good accuracy in terms of soot population estimation which confirms the viability of the mechanism selection & reduction methods put in place. Especially, results are in much bet- ter agreement with measurements than those obtained with the previously available mechanisms, enabling for a thorough analysis of the levers to reduce soot formation when designing new injection systems or combustion chambers technologies.

Jury

Benedetta FRANZELLI	CentraleSupélec	Rapporteur
Jérôme YON	INSA Rouen	Rapporteur
Frédéric GRISCH	INSA Rouen	Examinateur
Alberto CUOCI	Politecnico di Milano	Examinateur
Nicolas BERTIER	ONERA	Invité
Bénédicte CUENOT	CERFACS	Directrice
Eléonore RIBER	CERFACS	Co-directrice