Abstract

The combustion process in a typical can combustor of an industrial gas turbine is determined by the nature of turbulent flow, the chemical reaction and the interaction with each other. Turbulent non-premixed combustion can be divided into different flame regimes in terms of time- and length scales. A typical non-premixed turbulent diffusion flame in a gas turbine combustor covers all regimes. PDF methods are suitable to describe the entire combustion regime without any limitation to a certain regime. In this paper a hybrid pdf/RANS method is presented. The pdf model is based on the transported composition pdf equation, coupled with a commercial three dimensional CFD solver. A stochastic particle system in a Lagrangian framework is used to solve the pdf equation. The chemistry is described by an ILDM approach. The numerical results have been validated with measurements. The test rig consists of an non-premixed gas turbine can combustor with a typical primary and secondary zone. A main air swirler stabilizes the natural gas/air mixture in the primary zone, followed by a burnout and a mixing zone. The setup is investigated using conventional measurement techniques. Field measurements of compositions and mixture fraction as well as temperature are compared with the pdf/RANS calculations. The benefit of this approach is a realistic prediction of all relevant species. The complete one point statistics of the numerical calculations are used to identify the different combustion regimes from the combustor to the exit. The numerical comparison of pdf-, edm- and flamelet-model shows that the pdf approach can be used to describe a realistic gas turbine combustor. In the past, pdf-methods were applied only on simple generic model flames. The purpose of the presented paper is to demonstrate the application of a transported-pdf approach to a realistic gas turbine combustor.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call