Study of a Gas Turbine Combustion Chamber: Influence of the Mixing on the Flame Dynamics

Author(s):  
Christophe Duwig ◽  
Laszlo Fuchs

The new challenge of the Gas Turbine industry is to develop new technologies for meeting electricity demand growth and reducing harmful emissions. Thus a better understanding of the combustion phenomenon and an improvement in simulation capabilities are needed. Large Eddy Simulation tools brought the hope of meeting these two conditions and enabling the design of safe and clean burners. In the present paper, the influence of the unsteady mixing on the flame in a Lean Premixed Pre-vaporized combustor have been investigated. A premixed combustion flamelet model has been extended to non-uniform fuel/air mixtures cases. Extra terms in the equations, their effects and the modeling issues are discussed. Additionally, the effects of mixing on the flow field in an industrial gas turbine combustion chamber have been investigated. The mean flow field has been found to be weakly sensitive to the mixing effects. It is deduced that the modeling of the mixing and the combustion can be decoupled in the RANS framework. Regarding the flame dynamics, all runs show similar characteristic frequencies. However, different details of models lead to differences in the temperature fluctuations. This suggests that a rigorous modeling of the thermo-acoustic sources (e.g. heat-release fluctuations) requires accurate modeling of the mixing/combustion coupling, for handling accurately the dynamics of the flame.

2004 ◽  
Vol 126 (4) ◽  
pp. 899-905 ◽  
Author(s):  
E. Riesmeier ◽  
S. Honnet ◽  
N. Peters

Combustion and pollutant formation in a gas turbine combustion chamber is investigated numerically using the Eulerian particle flamelet model. The code solving the unsteady flamelet equations is coupled to an unstructured computational fluid dynamics (CFD) code providing solutions for the flow and mixture field from which the flamelet parameters can be extracted. Flamelets are initialized in the fuel-rich region close to the fuel injectors of the combustor. They are represented by marker particles that are convected through the flow field. Each flamelet takes a different pathway through the combustor, leading to different histories for the flamelet parameters. Equations for the probability of finding a flamelet at a certain position and time are additionally solved in the CFD code. To model the chemical properties of kerosene, a detailed reaction mechanism for a mixture of n-decane and 1,2,4-trimethylbenzene is used. It includes a detailed NOx submechanism and the buildup of polycyclic aromatic hydrocarbons up to four aromatic rings. The kinetically based soot model describes the formation of soot particles by inception, further growth by coagulation, and condensation as well as surface growth and oxidation. Simulation results are compared to experimental data obtained on a high-pressure rig. The influence of the model on pollutant formation is shown, and the effect of the number of flamelets on the model is investigated.


1998 ◽  
Vol 37 (10) ◽  
pp. 843-852
Author(s):  
Daniele Contini ◽  
Marco Ruggiero ◽  
Giampaolo Manfrida

Author(s):  
Elmar Riesmeier ◽  
Sylvie Honnet ◽  
Norbert Peters

Combustion and pollutant formation in a gas turbine combustion chamber is investigated numerically using the Eulerian Particle Flamelet Model (EPFM). The code solving the unsteady flamelet equations is coupled to an unstructured CFD code providing solutions for the flow and mixture field from which the flamelet parameters can be extracted. Flamelets are initialized in the fuel rich region close to the fuel injectors of the combustor. They are represented by marker particles which are convected through the flow field. Each flamelet takes a different pathway through the combustor leading to different histories for the flamelet parameters. Equations for the probability of finding a flamelet at a certain position and time are additionally solved in the CFD code. To model the chemical properties of kerosene, a detailed reaction mechanism for a mixture of n-decane and 1,2,4-trimethylbenzene is used. It includes a detailed NOx submechanism and the build-up of polycyclic aromatic hydrocarbons (PAHs) up to four aromatic rings. The kinetically based soot model describes the formation of soot particles by inception, further growth by coagulation and condensation as well as surface growth and oxidation. Simulation results are compared to experimental data obtained on a high pressure rig. The influence of the model on pollutant formation is shown, and the effect of the number of flamelets on the model is investigated.


Author(s):  
D. Fiaschi ◽  
G. Manfrida ◽  
L. Bonciani

The results of the investigation of the flow in a transparent (quartz tube) downscaled (≈1:3) model of a lean-premix type gas turbine combustion chamber are presented and discused. The model was tested at atmospheric pressure in reacting conditions; flow measurements were taken by a two-channel fiber-optic laser doppler velocimeter, using Al2O3 seeding of the air flowrate. The measurements cover a wide flow field inside the combustion chamber, including flame development and recirculating regions. Long-time samples (10–20 s) were used in order to achieve a good accuracy in the measurement of average flow conditions over the whole flow field; this involved a limited capability of representation of high-frequency components of turbulence, which could be locally obtained with optimization of the data rate and seeding conditions. Fast measurements were also locally performed where the seeding conditions were favourable. Integral variables and power spectra for reacting conditions show some distinctive aspects for the turbulence structure of reacting turbulent flows in confined spaces. Further measurements cover the outlet throat section of the premix combustor, demonstrating the persistence of a radial flow component on account of wall curvature effects and a certain degree of asimmetry in the inlet velocity distribution.


Author(s):  
Yigang Luan ◽  
Lianfeng Yang ◽  
Bo Wan ◽  
Tao Sun

Gas turbine engines have been widely used in modern industry especially in the aviation, marine and energy fields. The efficiency of gas turbines directly affects the economy and emissions. It’s acknowledged that the higher turbine inlet temperatures contribute to the overall gas turbine engine efficiency. Since the components are subject to the heat load, the internal cooling technology of turbine blades is of vital importance to ensure the safe and normal operation. This paper is focused on exploring the flow and heat transfer mechanism in matrix cooling channels. In order to analyze the internal flow field characteristics of this cooling configuration at a Reynolds number of 30000 accurately, large eddy simulation method is carried out. Methods of vortex identification and field synergy are employed to study its flow field. Cross-sectional views of velocity in three subchannels at different positions have been presented. The results show that the airflow is strongly disturbed by the bending part. It’s concluded that due to the bending structure, the airflow becomes complex and disordered. When the airflow goes from the inlet to the turning, some small-sized and discontinuous vortices are formed. Behind the bending structure, the size of the vortices becomes big and the vortices fill the subchannels. Because of the structure of latticework, the airflow is affected by each other. Airflow in one subchannel can exert a shear force on another airflow in the opposite subchannel. It’s the force whose direction is the same as the vortex that enhances the longitudinal vortices. And the longitudinal vortices contribute to the energy exchange of the internal airflow and the heat transfer between airflow and walls. Besides, a comparison of the CFD results and the experimental data is made to prove that the numerical simulation methods are reasonable and acceptable.


Sign in / Sign up

Export Citation Format

Share Document