Large eddy simulation (LES) of flame instabilities in Gas Turbines

Dnr:

SNIC 2016/1-534

Type:

SNAC Medium

Principal Investigator:

Jens Klingmann

Affiliation:

Lunds universitet

Start Date:

2016-12-13

End Date:

2018-01-01

Primary Classification:

20301: Teknisk mekanik

Secondary Classification:

20304: Energiteknik

Tertiary Classification:

20302: Rymd- och flygteknik

Webpage:

Allocation

Abstract

Gas turbines have made substantial gains in performance since their initial demonstration in jet powered aircraft and power turbines. The performance, noise characteristics, and pollutant emissions of gas turbines for propulsive applications continue to improve. Flame instability in Gas turbine are characterized by large-amplitude oscillations of one or more natural acoustic modes of the combustor. Such instabilities have been encountered during the development and operation of propulsion (e.g. afterburners), power generation (e.g., land-based gas turbines), boiler and heating systems, and industrial furnaces. The flame instabilities are reported to be highly related to flame structures and flow fields. In the project, LES investigation together with detailed chemical reactions of CH4-air and H2-CH4-air would be adopted. The LES simulation performs better than RANS turbulence model, especially in predicting the effects of turbulence on flame instabilities. However, LES is much more computational consuming than RANS. Hence more powerful computational resources would be needed. Phenomenon of flame partially quenched and re-ignition would also be studied to further explain the deeper reason for flame instability. Two flame stabilization mechanisms: bluff body and swirl stabilized flames, which are commonly used in industrial, would be investigated and compared with each other. Diffusion flame and premixed flame would be included in the investigation. Since in the large eddy simulation, especially in the flow with high Reynolds number or boundary layer flow, higher resolution mesh is needed which makes the computational consuming more. Hence different turbulence models together with turbulence-chemical reaction interaction models will be tested and evaluated with experimental results. The work will be mainly based on ANSYS FLUENT together with in-home developed user defined functions (UDF). Results will be useful to get a deeper insight of flame instabilities in our previous experiments and industrial Gas Turbine. Moreover, it will be helpful in the design process of burners in the Gas Turbine.