Influence of Film Cooling Unsteadiness on Turbine Blade Leading Edge Heat Flux

Author(s):  
James L. Rutledge ◽  
Paul I. King ◽  
Richard B. Rivir

Film cooling in the hot gas path of a gas turbine engine can protect components from the high temperature main flow, but it generally increases the heat transfer coefficient h partially offsetting the benefits in reduced adiabatic wall temperature. We are thus interested in adiabatic effectiveness η and h which are combined in a formulation called net heat flux reduction (NHFR). Unsteadiness in coolant flow may arise due to inherent unsteadiness in the external flow or be intentionally introduced for flow control. In previous work it has been suggested that pulsed cooling flow may, in fact, offer benefits over steady blowing in either improving NHFR or reducing the mass flow requirements for matched NHFR. In this paper we examine this hypothesis for a range of steady and pulsed blowing conditions. We use a new experimental technique to analyze unsteady film cooling on a semicircular cylinder simulating the leading edge of a turbine blade. The average NHFR with pulsed and steady film cooling is measured and compared for a single coolant hole located 21.5° downstream from the leading edge stagnation line, angled 20° to the surface and 90° to the streamwise direction. We show that for moderate blowing ratios at blade passing frequencies, steady film flow yields better NHFR. At higher coolant flow rates beyond the optimum steady blowing ratio, however, pulsed film cooling can be advantageous. We present and demonstrate a prediction technique for unsteady blowing at frequencies similar to the blade passing frequency that only requires the knowledge of steady flow behavior. With this important result, it is possible to predict when pulsing would be beneficial or detrimental.

Author(s):  
James L. Rutledge

It is necessary to understand how film cooling both reduces the adiabatic wall temperature and influences the heat transfer coefficient in order to predict the net heat flux to a gas turbine hot gas path component. Although a great number of studies have considered steady film cooling flows, the influence of film cooling unsteadiness has only recently been considered. Unsteadiness in the freestream flow or the coolant flow can cause fluctuations in both the adiabatic effectiveness and heat transfer coefficient, the dynamics of which have been difficult to measure. In previous studies, only time averaged effects have been measured. The present study has determined time resolved adiabatic effectiveness and heat transfer coefficient waveforms using a novel inverse heat transfer methodology. Unsteady film cooling was examined on the leading edge region of a circular cylinder simulating the leading edge of a turbine blade. Unsteady interactions between h and η, were examined near a coolant hole located 21.5° downstream from the leading edge stagnation line, angled 20° to the surface and 90° to the streamwise direction. The coolant plume is shown to shift back and forth as the jet’s momentum fluctuates. Increasing freestream turbulence was found to both reduce η, and the amplitude of the η waveforms.


Author(s):  
Mingjie Zhang ◽  
Nian Wang ◽  
Andrew F. Chen ◽  
Je-Chin Han

This paper presents the turbine blade leading edge model film cooling effectiveness with shaped holes, using the pressure sensitive paint (PSP) mass transfer analogy method. The effects of leading edge profile, coolant to mainstream density ratio and blowing ratio are studied. Computational simulations are performed using the realizable k-ε turbulence model. Effectiveness obtained by CFD simulations are compared with experiments. Three leading edge profiles, including one semi-cylinder and two semi-elliptical cylinders with an after body, are investigated. The ratios of major to minor axis of two semi-elliptical cylinders are 1.5 and 2.0, respectively. The leading edge has three rows of shaped holes. For the semi-cylinder model, shaped holes are located at 0 degrees (stagnation line) and ± 30 degrees. Row spacing between cooling holes and the distance between impingement plate and stagnation line are the same for three leading edge models. The coolant to mainstream density ratio varies from 1.0 to 1.5 and 2.0, and the blowing ratio varies from 0.5 to 1.0 and 1.5. Mainstream Reynolds number is about 100,900 based on the diameter of the leading edge cylinder, and the mainstream turbulence intensity is about 7%. The results provide an understanding of the effects of leading edge profile and on turbine blade leading edge region film cooling with shaped-hole designs.


Author(s):  
Ross Johnson ◽  
Jonathan Maikell ◽  
David Bogard ◽  
Justin Piggush ◽  
Atul Kohli ◽  
...  

When a turbine blade passes through wakes from upstream vanes it is subjected to an oscillation of the direction of the approach flow resulting in the oscillation of the position of the stagnation line on the leading edge of the blade. In this study an experimental facility was developed that induced a similar oscillation of the stagnation line position on a simulated turbine blade leading edge. The overall effectiveness was evaluated at various blowing ratios and stagnation line oscillation frequencies. The location of the stagnation line on the leading edge was oscillated to simulate a change in angle of attack between α = ± 5° at a range of frequencies from 2 to 20 Hz. These frequencies were chosen based on matching a range of Strouhal numbers typically seen in an engine due to oscillations caused by passing wakes. The blowing ratio was varied between M = 1, M = 2, and M = 3. These experiments were carried out at a density ratio of DR = 1.5 and mainstream turbulence levels of Tu ≈ 6%. The leading edge model was made of high conductivity epoxy in order to match the Biot number of an actual engine airfoil. Results of these tests showed that the film cooling performance with an oscillating stagnation line was degraded by as much as 25% compared to the performance of a steady flow with the stagnation line aligned with the row of holes at the leading edge.


Author(s):  
M. Salcudean ◽  
I. Gartshore ◽  
K. Zhang ◽  
Y. Barnea

Experiments have been conducted on a large model of a turbine blade. Attention has been focussed on the leading edge region, which has a semi-circular shape and four rows of film cooling holes positioned symmetrically about the stagnation line. The cooling holes were oriented in a spanwise direction with an inclination of 30° to the surface, and had streamwise locations of ±15° and ±44° from the stagnation line. Film cooling effectiveness was measured using a heat/mass analogy. Single row cooling from the holes at 15° and 44° showed similar patterns: spanwise averaged effectiveness which rises from zero at zero coolant mass flow to a maximum value η* at some value of mass flow ratio M*, then drops to low values of η at higher M. The trends can be quantitatively explained from simple momentum considerations for either air or CO2 as the coolant gas. Close to the holes, air provides higher η values for small M. At higher M, particularly farther downstream, the CO2 may be superior. The use of an appropriately defined momentum ratio G collapses the data from both holes using either CO2 or air as coolant onto a single curve. For η*, the value of G for all data is about 0.1. Double row cooling with air as coolant shows that the relative stagger of the two rows is an important parameter. Holes in line with each other in successive rows can provide improvements in spanwise averaged film cooling effectiveness of as much as 100% over the common staggered arrangement. This improvement is due to the interaction between coolant from rows one and two, which tends to provide complete coverage of the downstream surface when the rows are placed correctly with respect to each other.


Author(s):  
Mingjie Zhang ◽  
Nian Wang ◽  
Andrew F. Chen ◽  
Je-Chin Han

This paper presents the turbine blade leading edge model film cooling effectiveness with shaped holes, using the pressure sensitive paint (PSP) mass transfer analogy method. The effects of leading edge profile, coolant to mainstream density ratio, and blowing ratio are studied. Computational simulations are performed using the realizable k–ɛ (RKE) turbulence model. Effectiveness obtained by computational fluid dynamics (CFD) simulations is compared with experiments. Three leading edge profiles, including one semicylinder and two semi-elliptical cylinders with an after body, are investigated. The ratios of major to minor axis of two semi-elliptical cylinders are 1.5 and 2.0, respectively. The leading edge has three rows of shaped holes. For the semicylinder model, shaped holes are located at 0 deg (stagnation line) and ±30 deg. Row spacing between cooling holes and the distance between impingement plate and stagnation line are the same for three leading edge models. The coolant to mainstream density ratio varies from 1.0 to 1.5 and 2.0, and the blowing ratio varies from 0.5 to 1.0 and 1.5. Mainstream Reynolds number is about 100,000 based on the diameter of the leading edge cylinder, and the mainstream turbulence intensity is about 7%. The results provide an understanding of the effects of leading edge profile on turbine blade leading edge region film cooling with shaped hole designs.


2011 ◽  
Vol 133 (4) ◽  
Author(s):  
Stephen P. Lynch ◽  
Karen A. Thole ◽  
Atul Kohli ◽  
Christopher Lehane

Three-dimensional contouring of the compressor and turbine endwalls in a gas turbine engine has been shown to be an effective method of reducing aerodynamic losses by mitigating the strength of the complex vortical structures generated at the endwall. Reductions in endwall heat transfer in the turbine have been also previously measured and reported in literature. In this study, computational fluid dynamics simulations of a turbine blade with and without nonaxisymmetric endwall contouring were compared to experimental measurements of the exit flowfield, endwall heat transfer, and endwall film-cooling. Secondary kinetic energy at the cascade exit was closely predicted with a simulation using the SST k-ω turbulence model. Endwall heat transfer was overpredicted in the passage for both the SST k-ω and realizable k-ε turbulence models, but heat transfer augmentation for a nonaxisymmetric contour relative to a flat endwall showed fair agreement to the experiment. Measured and predicted film-cooling results indicated that the nonaxisymmetric contouring limits the spread of film-cooling flow over the endwall depending on the interaction of the film with the contour geometry.


Author(s):  
Pingfan He ◽  
Martha Salcudean ◽  
Ian S. Gartshore

Computations of film cooling are presented based on the geometry of a UBC experimental turbine blade model. This model has a semi-circlar leading edge with four rows of laterally-inclined film cooling orifices positioned symmetrically about the stagnation line. The computational domain follows the physical domain and includes the curved blade surface as well as the coolant regions in the circular coolant orifices. The injection orifices are inclined spanwise at 30° to the blade surface. A multi-zone curvilinear grid is used to simulate the complex configuration. Grids are generated by a block-structured elliptic grid generation method which represents exactly the curved blade surface as well as the circular injection orifices. Computations over the cooled turbine blade model are carried out for overall mass flow ratios of 0.52 and 0.97. The relative mass flow ratios from each orifice are specified to match experimental values. Density ratios of coolant to free stream were taken to be unity (constant density). Comparison of predicted film cooling effectiveness with experimental data showed reasonable agreement.


Author(s):  
Stephen P. Lynch ◽  
Karen A. Thole ◽  
Atul Kohli ◽  
Christopher Lehane

Three-dimensional contouring of the compressor and turbine endwalls in a gas turbine engine has been shown to be an effective method of reducing aerodynamic losses by mitigating the strength of the complex vortical structures generated at the endwall. Reductions in endwall heat transfer in the turbine have been also previously measured and reported in the literature. In this study, computational fluid dynamics simulations of a turbine blade with and without non-axisymmetric endwall contouring were compared to experimental measurements of the exit flowfield, endwall heat transfer and endwall film-cooling. Secondary kinetic energy at the cascade exit was closely predicted with a simulation using the SST k-ω turbulence model. Endwall heat transfer was overpredicted in the passage for both the SST k-ω and realizable k-ε turbulence models, but heat transfer augmentation for a non-axisymmetric contour relative to a flat endwall showed fair agreement to the experiment. Measured and predicted film-cooling results indicated that the non-axisymmetric contouring limits the spread of film-cooling flow over the endwall depending upon the interaction of the film with the contour geometry.


1990 ◽  
Vol 112 (3) ◽  
pp. 418-427 ◽  
Author(s):  
J. Karni ◽  
R. J. Goldstein

A naphthalene sublimation technique is used to study the effect of surface injection on the mass (heat) transfer from a circular cylinder in crossflow. Using a heat/mass transfer analogy the results can be used to predict film cooling effects in the leading edge region of a turbine blade. Air injection through one row of circular holes is employed in the stagnation region of the cylinder. Streamwise and spanwise injection inclinations are studied separately, and the effects of blowing rate and injection location relative to the cylinder front stagnation line are investigated. Streamwise injection produces significant mass transfer increases downstream of the injection holes, but a relatively small increase is observed between holes, normal to the injection direction. The mass transfer distribution, measured with spanwise injection through holes located near the cylinder front stagnation line, is extremely sensitive to small changes in the injection hole location relative to stagnation. When the centers of the spanwise injection holes are located 5 deg or more from the stagnation line, the holes lie entirely on one side of the stagnation line and the injection affects the mass transfer only on that side of the cylinder, approaching the pattern observed with streamwise injection.


Sign in / Sign up

Export Citation Format

Share Document