1. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Analysis of thermo-acoustic properties of combustors and afterburners Guillaume Jourdain Division of Fluid Dynamics Department of Applied Mechanics http://www.chalmers.se/am guillaume.jourdain@chalmers.se lee@chalmers.se

2. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Outline Motivation Project goals Validation Rig I Screech and Buzz modes Combustion model and Results Porous wall modelling Conclusions

3. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Motivation For premixed confined flames, pressure oscillations are often encountered in the desirable operating range. Alleviation of problem is difficult due to a lack of knowledge of mode structure and heat release feedback mechanism. Need improved analysis tools

4. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Mean flow, Unsteady fluctuations Flame (in)stability Mean flow Unsteady CFD: LES, URANS Linearized solver Eigenmode extraction Eigenmodes, frequencies, aerodynamic damping If the flame is stable, how reliable is the answer? Sensitivity? CFD strategies Turbulence model, combustion model Small fluctuations Advantage: Mean flow is validated. Most dangerous modes can be identified => Risk Assessment

5. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Project goals To investigate three different linearization concepts in combination with the spectral transformation and Arnoldi eigenmode extraction technique: –Linearized Euler Equations (LEE). –Linearized Navier-Stokes Equations (LNSE). –(Numerically) Linearized Unsteady RANS approach (L-URANS). To select optimal linearization concept based on accuracy, efficiency, user friendliness and maintainability. To evaluate out-of-loop flame transfer modeling (LEE, LNSE) or in-the-loop flame transfer modeling (L-URANS). To further improve and develop knowledge and modeling capabilities (both flow and acoustics) in the field of combustor liner walls, i.e. porous walls of various types. To further refine and validate the existing Arnoldi eigenmode extraction tool for more challenging test cases. To evaluate the eigenmode extraction tool for complex 3D combustor/afterburner cases

6. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Turbulent premixed flame No pressure oscillation Buzz mode Screech mode High speed Schlieren visualization of flame downstream of flameholder Flame flapping Flame front “waves” Test case: Validation Rig I, (Volvo Aero Corporation), Geometry of the rig

7. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology LEE solverLNSE solver RANS solver Arnoldi procedure Eigenmodes, frequencies, aerodynamic damping Mode stability analysis tool ”Out-of-the loop” stability analysis Implementation of an “improved combustion model” Implementation of the ”improved” porous wall model in URANS, LNSE solvers and Arnoldi extraction method

8. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Test cases Validation Rig I Preheated case: Case 1: T = 600 K,  = 0.61, m = 0.6 kg/s, stable flame Non-preheated cases : Case 2: T = 288 K,  = 0.61, m = 0.6 kg/s, stable flame Case 3: T = 288 K,  = 0.72, m = 1.1 kg/s, buzz and screech present Photo of turbulent premixed propane-air flame

9. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Validation Rig I, RANS solution, Realizable k-eps turb model, PaSR-WD type comb model case 2, Static temperature (T) Experimental and numerical velocity profiles at the centerline X-Velocity at the centerline

10. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Geometry of the rig Val Rig I Improved Combustion Model A new combustion model is chosen, two-step global mechanism: Westbrook and Dryer for propane. Water is a catalyst in the second reaction. Turbulence influence is included in the combustion model (only in the first global reaction) with the help of a Partially Stirred Reactor model. Six species are present in the model. (C3H8, O2, N2, CO, CO2, H2O)

11. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Val Rig I, cold cases: Eigenmode analysis results Longitudinal mixed mode, Candidate for the buzz and screech modes, LNSE – case 2 (Real and imaginary parts) Observations : A narrow flame brush i.e. a slower flame development leads to a lower buzz mode (75 Hz). The corresponding frequency is 120 Hz in the experiments A wider flame brush i.e. a faster flame development leads to higher frequency. The average flame brush thus seems to have a significant influence on the buzz frequency. The screech mode frequency is in agreement with the experiments.

12. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Extended Porous wall model The porous wall model was applied with succes in the past but this model has between improved to take into account inertial effects which are crucial for predicting the resonance frequency. This model is implemented in the URANS code. Mass conservation, Bernouilli’s equation with linear loss, Momentum equation with non linear loss Inertial effects are taken into account With for small pressure gradient With R effective linear resistance, σ wall porosity, C D discharge coefficient

13. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Extended Porous wall model Validation of the extended porous wall model for the Helmholtz resonator. Currently applied on Validation Rig I cases Theoritecal Frequency = 321 Hz - Frequency for maximum damping = 354 Hz - Absorption 28,99 % (1- |Cref|²/|Cinc|²)

14. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Extended Porous wall model - Effective linear resistance R = 10 or 400 - Wall porosity, σ = 0,05 - Discharge coefficient C D = 0,8 A additional domain is added under the bottom wall, 2 cm thick in order to add a porous wall. The porous wall starts 40 cm upstream the flame holder and stops 20 cm downstream the flame holder

15. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Extended Porous wall model Fourier spectrum (R = 10)

16. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Extended Porous wall model Fourier spectrum (R = 400), Absorption coefficient for f = 1125 Hz (Screech mode candidate) => 7.12 %

17. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Conclusions The two-step combustion model combining a Westbrook & Dryer kinetic model and a Partially Stirred Reactor model has been implemented in the code and tested for several cases. The flame brush development could explain the low frequency of the buzz mode candidate. The ”improved” Porous Wall model has been implemented in the URANS solver and studied for the HelmHoltz resonator test case and the Validation Rig I for several cases. The results from the URANS solver, which include the porous wall modeling, shows that the screech mode could be damped (in the studied case, the absorption coefficient is 7%) Publications : - ASME Turbo Expo, Glasgow, 14-18 June 2010 (Conference paper) - AIAA Aeroacoustics Conference, Stockholm, 7-9 June 2010 (Conference paper) - Licenciate thesis: Thermo-Acoustic Properties in Combustion Chambers, Chalmers, Gothenburg, 2010 Two conference papers: - 17th AIAA /CEAS aeroacoustics conference 2011, June 6-8, Portland, USA. - 20th ISABE conference 2011, September 12-16, Gothenburg, Sweden.

18. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Future Work Validate the implementation of the porous wall model for the LNSE solver and Arnoldi extraction method. To further improve and develop knowledge and modeling capabilities (both flow and acoustics) in the field of combustor liner walls, i.e. porous walls of various types for the Validation Rig I test cases. To further refine and to evaluate the eigenmode extraction tool for complex 3D combustor/afterburner cases

19. Turbomachinery & Aero-Acoustics Group Chalmers University of Technology Tack för uppmärksamheten Frågor?