Final Report Summary - NEWSMILE (Near-Wall Simulations and Measurements in Lean-Burn Engines)
Executive Summary:
An effusion cooled liner segment was installed into a pressurized combustor (Fig. 1). A piloted lean premixed swirling flame was operated such that the impinging region at the effusion cooling plate was accessible to laser diagnostic investigations. Using coherent anti-Stokes Raman spectroscopy, phosphor thermometry, and stereoscopic particle image velocimetry the interaction between flame and cooling jets was examined in detail. Large Eddy Simulations and Reynolds Averaged Navier Stokes computations of the test rig have been conducted. Chemistry look-up tables were generated and used from the flamelet generated manifold (FGM) model and integrated with the ß-function as presumed probability density function. RANS computations also tested the implemented models for radiative heat fluxes. These were supported by radiative spectral measurements of the test rig tile and the combustor tile of the ITD owner. All CFD investigations were coupled to Finite-Element Methods. The near-wall temperature and heat transfer coefficients were provided by CFD to FEM, so that a thermal analysis would return a more appropriate metal temperature for CFD in return. The numerical investigations include grid studies, the application of different parameters and the analysis of differences and deviations from the experimental results. The made experiences resulted in recommendations and suggestions of methodology for the analysis of the real aero-engine combustor.
Project Context and Objectives:
Within this project the objective was to experimentally investigate the interaction of film cooling and swirling flames using advanced laser diagnostic methods. The boundary conditions of the combustor and the effusion plate geometry have been chosen to mimic important features of real aero-engine combustors. Unique data on wall temperatures, gas temperatures and turbulent flow field including the area close to the effusion cooling plate have been recorded. The measurements also include radiative spectral properties of the combustor tile with estimation of averaged emissivity and absorptivity at different temperatures.
The data was then used to analyse and validate the development of computational methods including LES and RANS, coupled to FEM for heat flux analysis.
Project Results:
The results of this work are classified into wall temperatures, gas temperatures in the bulk and within the close proximity of the effusion cooling liner, and comprehensive flow field data in the bulk as well as close to the wall. High spatial and temporal resolution enabled to disclose the interaction of single cooling jets with the swirling flame. Results are exemplified in Fig.2 and Fig. 3 showing mean wall temperatures of the
effusion cooling plate and the ensemble-averaged axial velocity component along the central plane in the close proximity of the effusion liner. For the present operation conditions the jets issuing from the effusion liner detach from the wall. In the wake of the jets wall temperatures are reduced remarkably. The mean axial flow is shown in the region where the swirling flame impinges at the wall. Although the cross velocities are high, the effusion jets penetrate a couple of millimetres into the bulk flow. Mixing appears rapidly such that the cooling action is restricted in downstream direction to a few hole diameters. Good agreements with the computational results were achieved. The comparison of the flow field (Fig.4) and the scalar fields in the core of the combustion chamber were well matched by RANS and LES. However, the chosen models and methods in LES under-predict the near-wall temperature intolerably (Fig.5). On the other hand, using the heat-transfer coefficients from CFD and the near –wall temperatures outside of the boundary layer, to perform FE simulations of the wall, gives a good estimate of the wall temperature. Numerical error because of insufficient resolution was excluded due to grid studies. Radiative fluxes do not play a major roll for the generic combustor, but for the real aero-engine combustor. Here, radiation has a major impact. CFD seems to under-predict the radiative heat-flux. The error of the predicted wall temperatures when using conventional estimates of the radiation is in the order of the magnitude of the uncertainty in the applied thermal paint data.
Potential Impact:
Based on this unique experimental data models and prediction capability of numerical simulation tools can be benchmarked. Thereby predictive simulation skills are enhanced. Integrated into the design
process of next generation aero-engine gas turbines this project supports substantially green aviation.
List of Websites:
http://cordis.europa.eu/project/rcn/192562_en.html
An effusion cooled liner segment was installed into a pressurized combustor (Fig. 1). A piloted lean premixed swirling flame was operated such that the impinging region at the effusion cooling plate was accessible to laser diagnostic investigations. Using coherent anti-Stokes Raman spectroscopy, phosphor thermometry, and stereoscopic particle image velocimetry the interaction between flame and cooling jets was examined in detail. Large Eddy Simulations and Reynolds Averaged Navier Stokes computations of the test rig have been conducted. Chemistry look-up tables were generated and used from the flamelet generated manifold (FGM) model and integrated with the ß-function as presumed probability density function. RANS computations also tested the implemented models for radiative heat fluxes. These were supported by radiative spectral measurements of the test rig tile and the combustor tile of the ITD owner. All CFD investigations were coupled to Finite-Element Methods. The near-wall temperature and heat transfer coefficients were provided by CFD to FEM, so that a thermal analysis would return a more appropriate metal temperature for CFD in return. The numerical investigations include grid studies, the application of different parameters and the analysis of differences and deviations from the experimental results. The made experiences resulted in recommendations and suggestions of methodology for the analysis of the real aero-engine combustor.
Project Context and Objectives:
Within this project the objective was to experimentally investigate the interaction of film cooling and swirling flames using advanced laser diagnostic methods. The boundary conditions of the combustor and the effusion plate geometry have been chosen to mimic important features of real aero-engine combustors. Unique data on wall temperatures, gas temperatures and turbulent flow field including the area close to the effusion cooling plate have been recorded. The measurements also include radiative spectral properties of the combustor tile with estimation of averaged emissivity and absorptivity at different temperatures.
The data was then used to analyse and validate the development of computational methods including LES and RANS, coupled to FEM for heat flux analysis.
Project Results:
The results of this work are classified into wall temperatures, gas temperatures in the bulk and within the close proximity of the effusion cooling liner, and comprehensive flow field data in the bulk as well as close to the wall. High spatial and temporal resolution enabled to disclose the interaction of single cooling jets with the swirling flame. Results are exemplified in Fig.2 and Fig. 3 showing mean wall temperatures of the
effusion cooling plate and the ensemble-averaged axial velocity component along the central plane in the close proximity of the effusion liner. For the present operation conditions the jets issuing from the effusion liner detach from the wall. In the wake of the jets wall temperatures are reduced remarkably. The mean axial flow is shown in the region where the swirling flame impinges at the wall. Although the cross velocities are high, the effusion jets penetrate a couple of millimetres into the bulk flow. Mixing appears rapidly such that the cooling action is restricted in downstream direction to a few hole diameters. Good agreements with the computational results were achieved. The comparison of the flow field (Fig.4) and the scalar fields in the core of the combustion chamber were well matched by RANS and LES. However, the chosen models and methods in LES under-predict the near-wall temperature intolerably (Fig.5). On the other hand, using the heat-transfer coefficients from CFD and the near –wall temperatures outside of the boundary layer, to perform FE simulations of the wall, gives a good estimate of the wall temperature. Numerical error because of insufficient resolution was excluded due to grid studies. Radiative fluxes do not play a major roll for the generic combustor, but for the real aero-engine combustor. Here, radiation has a major impact. CFD seems to under-predict the radiative heat-flux. The error of the predicted wall temperatures when using conventional estimates of the radiation is in the order of the magnitude of the uncertainty in the applied thermal paint data.
Potential Impact:
Based on this unique experimental data models and prediction capability of numerical simulation tools can be benchmarked. Thereby predictive simulation skills are enhanced. Integrated into the design
process of next generation aero-engine gas turbines this project supports substantially green aviation.
List of Websites:
http://cordis.europa.eu/project/rcn/192562_en.html
