Prediction of CO Emission Index for Aviation Gas Turbine Combustor Using Flamelet Generated Manifold Combustion Model

Abstract

Abstract Carbon monoxide (CO) has been identified as one of the regulated pollutants and gas turbine manufacturers target to reduce the CO emission from their gas turbine engines. CO forms primarily when carbonous fuels are not burnt completely, or products of combustion are quenched before completing the combustion. Numerical simulations are effective tools that allow a better understanding of the mechanisms of CO formation in gas turbine engines and are useful in evaluating the effect of different parameters like swirl, fuel atomization, mixing etc. on the overall CO emission for different engine conditions like idle, cruise, approach and take off. In this paper, a thorough assessment of flamelet generated manifold (FGM) combustion model is carried out to predict the qualitative variation and magnitude of CO emission index with the different configurations of a Honeywell test combustor operating with liquid fuel under idle condition, which is the more critical engine condition for CO emission. The different designs of the test combustor are configured in such a way that they yield different levels of CO and hence are ideal to test the accuracy of the combustion model. Large eddy simulation (LES) method is used for capturing the turbulence accurately along with the FGM combustion model that is computationally economical compared to the detailed/reduced chemistry modeling using finite rate combustion model. Liquid fuel spray breakup is modeled using stochastic secondary droplet (SSD) model. Four different configurations of the aviation gas turbine combustor are studied in this work referring to earlier work by Xu et al. [1]. It is shown that the FGM model can predict CO trends accurately. The other global parameters like exit temperature, NOx emissions, pattern factor also show reasonable agreement with the test data. The sensitivity of the CO prediction to the liquid fuel droplet breakup model parameters is also studied in this work. Although the trend of CO variation is captured for different values of breakup parameters, the absolute magnitude of CO emission index differs significantly with the change in the values of breakup parameters suggesting that the spray has a larger impact on the quantitative prediction of CO emission. An accurate prediction of CO trends at idle conditions using FGM model extends the applicability of FGM model to predict different engine operating conditions for different performance criteria accurately.