Details
| Original language | English |
|---|---|
| Title of host publication | ASME 2011 Turbo Expo |
| Subtitle of host publication | Turbine Technical Conference and Exposition, GT2011 |
| Pages | 1551-1561 |
| Number of pages | 11 |
| Edition | PARTS A, B, AND C |
| Publication status | Published - 3 May 2012 |
| Event | ASME 2011 Turbo Expo: Turbine Technical Conference and Exposition, GT2011 - Vancouver, BC, Canada Duration: 6 Jun 2011 → 10 Jun 2011 |
Publication series
| Name | Proceedings of the ASME Turbo Expo |
|---|---|
| Number | PARTS A, B, AND C |
| Volume | 7 |
Abstract
For the design of highly efficient turbine exhaust diffusers, it is important to take into account the unsteady flow field induced by the last turbine stage. A 1/10 scale model of a gas turbine exhaust diffuser consisting of an annular followed by a conical diffuser is used to investigate the influence of the unsteady flow conditions on the performance of the diffuser. To reproduce the outflow of the last turbine stage, a NACA profiled rotor is placed at the inlet of the diffuser. Measurements with 3D hot-wire probes are conducted in order to resolve the unsteady flow mechanisms inside the annular diffuser. Additionally, unsteady pressure transducers are installed at the shroud of the diffuser and on the surface of the NACA blades to detect rotating instabilities generated by the rotor. For operating points with a high flow-coefficient, vortices are generated at the tip of the blades. They support the boundary layer at the shroud with kinetic energy up to the halflength of the annular diffuser, which leads to a high pressure recovery. For operating conditions without generated vortices, the pressure recovery is significantly lower. The analysis of the pressure signals at the shroud and at the rotating blades with auto- and cross-correlations show that the number of generated vortices at the tip of the blades is lower than the number of blades. For the operating point with the highest flow coefficient, it can be shown that fourteen vortices are generated at the tip of the thirty blades. In modern RANS-model based CFD-codes, turbulence is modeled as isotropic flow. By comparing the three Reynolds Stress components behind the rotor it can be shown that the flow field especially in the wake of the blades is non-isotropic. This shows that diffuser flows should be modeled with turbulence models which account for non-isotropy.
Keywords
- Boundary layer, Diffuser, Hot-wire, Isotropy, Rotating instabilities, Vortex
ASJC Scopus subject areas
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
ASME 2011 Turbo Expo: Turbine Technical Conference and Exposition, GT2011. PARTS A, B, AND C. ed. 2012. p. 1551-1561 (Proceedings of the ASME Turbo Expo; Vol. 7, No. PARTS A, B, AND C).
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - Influence of Unsteady Turbine Flow on the Performance of an Exhaust Diffuser
AU - Kuschel, Marcus
AU - Seume, Joerg R.
PY - 2012/5/3
Y1 - 2012/5/3
N2 - For the design of highly efficient turbine exhaust diffusers, it is important to take into account the unsteady flow field induced by the last turbine stage. A 1/10 scale model of a gas turbine exhaust diffuser consisting of an annular followed by a conical diffuser is used to investigate the influence of the unsteady flow conditions on the performance of the diffuser. To reproduce the outflow of the last turbine stage, a NACA profiled rotor is placed at the inlet of the diffuser. Measurements with 3D hot-wire probes are conducted in order to resolve the unsteady flow mechanisms inside the annular diffuser. Additionally, unsteady pressure transducers are installed at the shroud of the diffuser and on the surface of the NACA blades to detect rotating instabilities generated by the rotor. For operating points with a high flow-coefficient, vortices are generated at the tip of the blades. They support the boundary layer at the shroud with kinetic energy up to the halflength of the annular diffuser, which leads to a high pressure recovery. For operating conditions without generated vortices, the pressure recovery is significantly lower. The analysis of the pressure signals at the shroud and at the rotating blades with auto- and cross-correlations show that the number of generated vortices at the tip of the blades is lower than the number of blades. For the operating point with the highest flow coefficient, it can be shown that fourteen vortices are generated at the tip of the thirty blades. In modern RANS-model based CFD-codes, turbulence is modeled as isotropic flow. By comparing the three Reynolds Stress components behind the rotor it can be shown that the flow field especially in the wake of the blades is non-isotropic. This shows that diffuser flows should be modeled with turbulence models which account for non-isotropy.
AB - For the design of highly efficient turbine exhaust diffusers, it is important to take into account the unsteady flow field induced by the last turbine stage. A 1/10 scale model of a gas turbine exhaust diffuser consisting of an annular followed by a conical diffuser is used to investigate the influence of the unsteady flow conditions on the performance of the diffuser. To reproduce the outflow of the last turbine stage, a NACA profiled rotor is placed at the inlet of the diffuser. Measurements with 3D hot-wire probes are conducted in order to resolve the unsteady flow mechanisms inside the annular diffuser. Additionally, unsteady pressure transducers are installed at the shroud of the diffuser and on the surface of the NACA blades to detect rotating instabilities generated by the rotor. For operating points with a high flow-coefficient, vortices are generated at the tip of the blades. They support the boundary layer at the shroud with kinetic energy up to the halflength of the annular diffuser, which leads to a high pressure recovery. For operating conditions without generated vortices, the pressure recovery is significantly lower. The analysis of the pressure signals at the shroud and at the rotating blades with auto- and cross-correlations show that the number of generated vortices at the tip of the blades is lower than the number of blades. For the operating point with the highest flow coefficient, it can be shown that fourteen vortices are generated at the tip of the thirty blades. In modern RANS-model based CFD-codes, turbulence is modeled as isotropic flow. By comparing the three Reynolds Stress components behind the rotor it can be shown that the flow field especially in the wake of the blades is non-isotropic. This shows that diffuser flows should be modeled with turbulence models which account for non-isotropy.
KW - Boundary layer
KW - Diffuser
KW - Hot-wire
KW - Isotropy
KW - Rotating instabilities
KW - Vortex
UR - http://www.scopus.com/inward/record.url?scp=84865489398&partnerID=8YFLogxK
U2 - 10.1115/GT2011-45673
DO - 10.1115/GT2011-45673
M3 - Conference contribution
AN - SCOPUS:84865489398
SN - 9780791854679
T3 - Proceedings of the ASME Turbo Expo
SP - 1551
EP - 1561
BT - ASME 2011 Turbo Expo
T2 - ASME 2011 Turbo Expo: Turbine Technical Conference and Exposition, GT2011
Y2 - 6 June 2011 through 10 June 2011
ER -