Direct Numerical Simulations of NOx formation in spatially developing turbulent premixed Bunsen flames with mixture inhomogeneity
Permanent link to this recordhttp://hdl.handle.net/10754/625559
MetadataShow full item record
AbstractDirect Numerical Simulation of three-dimensional spatially developing turbulent methane/air flames are performed. Four flames are simulated; they differ for the level of premixing of the fuel inlet: one has a fully premixed inlet, the other three have a partially premixed inlet that mimic a common injection strategy in stationary gas turbines. The jet consist of a methane/air mixture with global equivalence ratio ɸ = 0.7 and temperature of 800 K. The simulations are performed at 4 atm. The inlet velocity field and the fuel/air fields were extracted from a fully developed turbulent channel simulation. Chemistry is treated with a new skeletal chemical mechanism consisting of 33 species developed specifically for the DNS. The data are analyzed to study possible influences of partial premixing on the flame structure and the combustion efficiency. The results show that increasing the level of partial premixing, the fluctuations of heat release rate increase, due to the richer and leaner pockets of mixture in the flame, while the conditional mean decreases. Increasing the level of partial premixing, the peak of NO and the range of NO values for a given temperature increase. An analysis of NO production is performed categorizing the different initiation steps in the Ndecomposition through four pathways: thermal, prompt, NNH and NO. Different behaviour with respect to laminar flames is found for the NNH pathway suggesting that turbulence influences this pathway of formation of NO.
CitationLuca S, Attili A, Bisetti F (2017) Direct Numerical Simulations of NOx formation in spatially developing turbulent premixed Bunsen flames with mixture inhomogeneity. 55th AIAA Aerospace Sciences Meeting. Available: http://dx.doi.org/10.2514/6.2017-0603.
SponsorsResearch reported in this publication was supported by the King Abdullah University of Science and Technology (KAUST). The authors acknowledge valuable support from KAUST Supercomputing Laboratory (KSL) in the form of computational time on the CRAY XC40 “Shaheen”.
Conference/Event name55th AIAA Aerospace Sciences Meeting