Aaron R. Byerley

The effect of freestream turbulence on film cooling adiabatic effectiveness

Abstract The film-cooling performance of a flat plate in the presence of low and high freestream turbulence is investigated using liquid crystal thermography. This paper contributes high-resolution color images that clearly show how the freestream turbulence spreads the cooling air around a larger area of the film-cooled surface. Distributions of the adiabatic effectiveness are determined over the film-cooled surface of the flat plate using the hue method and image processing. Three blowing rates are investigated for a model with three straight holes spaced three diameters apart, with density ratio near unity. High freestream turbulence is shown to increase the area-averaged effectiveness at high blowing rates, but decrease it at low blowing rates. At low blowing ratio, freestream turbulence clearly reduces the coverage area of the cooling air due to increased mixing with the main flow. However, at high blowing ratio, when much of the jet has lifted off in the low turbulence case, high freestream turbulence turns its increased mixing into an asset, entraining some of the coolant that penetrates into the main flow and mixing it with the air near the surface.

Heat Transfer Enhancement Within a Turbine Blade Cooling Passage Using Ribs and Combinations of Ribs With Film Cooling Holes

A transient heat transfer method using liquid crystals has been applied to a scale model of a turbine rotor blade passage. Detailed contours of local heat transfer coefficient are presented for the passage in which the heat transfer to one wall was enhanced first by ribs and then with ribs combined with holes. The hole geometry and experimental dimensionless flow rates were representative of those occurring at the entrance to engine film cooling holes. The results for the ribbed passage are compared to established correlations for developed flow. Qualitative surface shear stress distributions were determined with liquid crystals. The complex distributions of heat transfer coefficient are discussed in light of the interpreted flow field.

Measurements in a Turbine Cascade Flow Under Ultra Low Reynolds Number Conditions

With the new generation of gas turbine engines, low Reynolds number flows have become increasingly important. Designers must properly account for transition from laminar to turbulent flow and separation of the flow from the suction surface, which is strongly dependent upon transition. Of interest to industry are Reynolds numbers based upon suction surface length and flow exit velocity below 150,000 and as low as 25,000. In this paper, the extreme low end of this Reynolds number range is documented by way of pressure distributions, loss coefficients, and identification of separation zones. Reynolds numbers of 25,000 and 50,000 and with 1 percent and 8-9 percent turbulence intensity of the approach flow (free-stream turbulence intensity, FSTI) were investigated. At 25,000 Reynolds number and low FSTI, the suction surface displayed a strong and steady separation region. Raising the turbulence intensity resulted in a very unsteady separation region of nearly the same size on the suction surface. Vortex generators were added to the suction surface, but they appeared to do very little at this Reynolds number. At the higher Reynolds number of 50,000, the low-FSTI case was strongly separated on the downstream portion of the suction surface. The separation zone was eliminated when the turbulence level was increased to 8-9 percent. Vortex generators were added to the suction surface of the low-FSTI case. In this instance, the vortices were able to provide the mixing needed to re-establish flow attachment. This paper shows that massive separation at very low Reynolds numbers (25,000) is persistent, in spite of elevated FSTI and added vortices. However, at a higher Reynolds number, there is opportunity for flow reattachment either with elevated free-stream turbulence or with added vortices. This may be the first documentation of flow behavior at such low Reynolds numbers. Although it is undesirable to operate under these conditions, it is important to know what to expect and how performance may be improved if such conditions are unavoidable.

The effect of turbulence intensity and length scale on low-pressure turbine blade aerodynamics

Abstract Unpredicted losses have been observed in low-pressure gas turbine stages during high altitude operation. These losses have been attributed to aerodynamic separation on the turbine blade suction surfaces. To gain insight into boundary layer transition and separation for these low Reynolds number conditions, the heat transfer distribution on a Langston turbine blade shape was measured in a linear cascade wind tunnel for turbulence levels of 0.8% and 10% and Reynolds numbers of 40–80 k . Turbulence levels of 10% were generated using three passive biplanar lattice grids with square-bar widths of 1.27, 2.54 and 6.03 cm to investigate the effect of turbulence length scale. The heat transfer was measured using a uniform heat flux (UHF) liquid crystal technique. As turbulence levels increased, stagnation heat transfer increased and the location of the suction-side boundary layer transition moved upstream toward the blade leading edge. For this turbine blade shape the transition location did not depend on turbulence length scale, the location is more dependent on pressure distribution, Reynolds number and turbulence intensity. For the 10% turbulence cases, the smaller length scales had a larger affect on heat transfer at the stagnation point. A laser tuft method was used to differentiate between boundary layer transition and separation on the suction surface of the blade. Separation was observed for all of the low turbulence (clean tunnel) cases while transition was observed for all of the 10% turbulence cases. Separation and transition locations corresponded to local minimums in heat transfer. Reattachment points did not correspond to local maximums in heat transfer, but instead, the heat transfer coefficient continued to rise downstream of the reattachment point. For the clean tunnel cases, streamwise streaks of varying heat transfer were recorded on the concave pressure side of the turbine blade. These streaks are characteristic of either Gortler vortices or a three-dimensional transition process. For the 10% turbulence cases, these streaks were not present. The results presented in this paper show that turbulence length scale, in addition to intensity have an important contribution to turbine blade aerodynamics and are important to CFD modelers who seek to predict boundary layer behavior in support of turbine blade design optimization efforts.

Using Gurney Flaps to Control Laminar Separation on Linear Cascade Blades

This paper describes an experimental investigation of the use of Gurney flaps to control laminar separation on turbine blades in a linear cascade. Measurements were made at Reynolds numbers (based upon inlet velocity and axial chord) of 28×10 3 , 65×10 3 and 167×10 3 . The freestream turbulence intensity for all three cases was 0.8%. Laminar separation was present on the suction surface of the Langston blade shape for the two lower Reynolds numbers. In an effort to control the laminar separation, Gurney flaps were added to the pressure surface close to the trailing edge. The measurements indicate that the flaps turn and accelerate the flow in the blade passage toward the suction surface of the neighboring blade thereby eliminating the separation bubble. Five different sizes of Gurney flaps, ranging from 0.6 to 2.7% of axial chord, were tested. The laser thermal tuft technique was used to determine the influence of the Gurney flaps on the location and size of the separation bubble. Additionally, measurements of wall static pressure, profile loss, and blade-exit flow angle were made. The blade pressure distribution indicates that the lift generated by the blade is increased. As was expected, the Gurney flap also produced a larger wake. In practice, Gurney flaps might possibly be implemented in a semi-passive manner. They could be deployed for low Reynolds number operation and then retracted at high Reynolds numbers when separation is not present. This work is important because it describes a successful means for eliminating the separation bubble while characterizing both the potential performance improvement and the penalties associated with this semi-passive flow control technique.

Switching Behavior of a Plasma-Fluidic Actuator

This work presents the plasma-fluidic actuator as a new type of flow control actuator. The device combines the best features of plasma actuators and fluidic oscillators to create an actuator that decouples the actuation frequency from the flow rate and overcomes the velocity limitations of the plasma actuator. Plasma actuators are used to effectively switch a high momentum jet between Coanda attachment walls by inducing a flow asymmetry across the nozzle of a fluidic oscillator. In this manner the device effectively serves as a fluid amplifier. Switching time, defined as the delay between the plasma initiation and completion of the switching process between opposing attachment walls, is used as a metric to quantify the performance of these new devices. In these tests the jet issuing from the plasma-fluidic actuator was switched between attachment walls in as little as 19 ms. Results indicate that the switching time decreases as the power of the plasma actuator increases. Furthermore, switching times decrease with decreased velocity, as the plasma has more substantial control authority over the primary jet. This new flow control actuator concept has the potential of being used in many applications such as thrust vectoring and jet flow mixing.

Design Considerations for Using Indented Surface Treatments to Control Boundary Layer Separation

Although several studies have assessed the effectiv eness of indented surface treatments for delaying boundary layer separation, definitive design guidelines have never been published. This study examined the previous literature with an eye towards determining the optimum placement, depth, and length of th e indentations. The best location for the grooves or dimples is just prior to the region of separation. The ratio of indentation depth to local boundary layer thickness should be between 0.6 and 1.0. The ratio of indentation depth to length should be 0. 10 to 0.15. These guidelines were tested on a NACA 0015 airfoil using Computational Fluid Dynamics (CFD) and surface flow visualization. Both the CFD and surface flow visualization confirmed the existence of a laminar separation bubble on the suction sur face of the airfoil at low angle of attack. The addition of a groove following the established design criteria eliminated the separa tion bubble behind the groove, but it resulted in a loss of lift and an increase in drag. This is believed to be due to an inclined impingement on the aft surface of the groove , which may be eliminated through a more optimal groove design .

Flow Characteristics Within and Downstream of Spherical and Cylindrical Dimple on a Flat Plate at Low Reynolds Numbers

A comprehensive experimental study has been performed in the U.S. Air Force Academy water tunnel to obtain a better understanding of the complicated flow patterns in shallow dimple configurations (h/D ≤ 0.1), including single cylindrical and spherical dimples, as well as single spanwise rows of dimples. The flow patterns, in-dimple separation zone extent, and bulk flow oscillation frequencies have been measured at low Reynolds number conditions. Three different single dimples and two single rows of dimples have been tested over a range of Reynolds numbers ReD of 3,170 to 23,590 including laminar and turbulent flow patterns downstream of a dimple. To visualize the fine flow features, five different colors of dye were injected through five cylindrical ports machined at locations upstream and inside the dimples. The measured results revealed unsteady and three-dimensional flow features inside and downstream of the dimple. The Reynolds number, dimple shape and the presence of adjacent dimples all play important roles in determining the nature of the flow pattern formation. Some preliminary conclusions regarding the laminar-turbulent flow transition after a dimple are presented.Copyright

A “Cool” Thermal Tuft for Detecting Surface Flow Direction

An optical flow diagnostic technique was developed to detect surface flow direction in weakly separated flows. The technique makes use of thermochromic liquid crystals applied to a surface that is first coated with flatblack paint. A thin reflective material is affixed to the test surface. The surface is heated uniformly from above with infrared heaters. The reflective surface remains cool relative to the nearby black surface. A thermal tuft of relatively cool air (having just passed over the reflector) is advected in the direction of the surface flow thereby creating a comet tail of liquid crystal color response. The comet tail points in the direction of the surface flow The cool thermal tuft successfully provides a means for detecting surface flow direction. The advantages of this method are the following: it is sensitive enough to detect the presence of weakly separated flows; it is non-obtrusive; it is reversible so does not require clean-up between runs; and it can be applied to vertical surfaces because there is nothing to drip

Detailed Measurements of Local Heat Transfer Coefficient in the Entrance to Normal and Inclined Film Cooling Holes

The local heat transfer inside the entrance to large-scale models of film cooling holes has been measured using the transient heat transfer technique. The method employs temperature-sensitive liquid crystals to measure the surface temperature of large-scale perspex models. Full distributions of local Nusselt number were calculated based on the cooling passage centerline gas temperature ahead of the cooling hole. The circumferentially averaged Nusselt number was also calculated based on the local mixed bulk driving gas temperature to aid interpretation of the results, and to broaden the potential application of the data. Data are presented for a single film cooling hole inclined at 90 and 150 deg to the coolant duct wall. Both holes exhibited entry length heat transfer levels that were significantly lower than those predicted by entry length data in the presence of crossflow. The reasons for the comparative reduction are discussed in terms of the interpreted flow field.