Modélisation 0D/1D des émissions de particules de suie dans les turbines à gaz aéronautiques

Abstract

Because of more stringent regulations of aircraft particle emissions as well as strong uncertainties about their formation and their effects on the atmosphere, a better understanding of particle microphysical mechanisms and their interactions with the engine components is required. This thesis focuses on the development of a 0D/1D combustion model with soot production in an aeronautical gas turbine. A major objective of this study is to assess the quality of soot particle emission predictions for different flight configurations. The model should eventually allow performing parametric studies on current or future engines with a minimal computation time. The model represents the combustor as well as turbines and nozzle with a chemical reactor network (CRN) that is coupled with a detailed combustion chemistry for kerosene (Jet A-1) and a soot particle dynamics model using the method of moments. The CRN was applied to the CFM56-2C1 engine during flight configurations of the LTO cycle (Landing-Take-Off) as in the APEX-1 study on aircraft particle emissions. The model was mainly validated on gas turbine thermodynamic data and pollutant concentrations (H2O, COX, NOx, SOX) which were measured in the same study. Once the first validation completed, the model was subsequently used for the computation of mass and number-based emissions indices of the soot particulate population and average diameter. Overall, the model is representative of the thermodynamic conditions and succeeds in predicting the emissions of major pollutants, particularly at high power. Concerning soot particulate emissions, the model's ability to predict simultaneously the emission indices as well as mean diameter has been partially validated. Indeed, the mass emission indices have remained higher than experimental results particularly at high power. These differences on particulate emission index may be the result of uncertainties on thermodynamic parameters of the CRN and mass air flow distribution in the combustion chamber. The analysis of the number-based emission index profile along the CRN also highlights the need to review the nucleation model that has been used and to consider in the future the implementation of a particle aggregation mechanism.

Date	13 Jun 2014
Original language	French
Awarding Institution	École de technologie supérieure
Supervisor	François Garnier (Supervisor) & Patrice Seers (Co-supervisor)