Cross Flow Effects on Endwall Heat Transfer in Film-Cooled Turbine Nozzle Guide Vanes

Film cooling effects on endwalls in the stagnation point region are of special interest since the heat transfer is influenced drastically by secondary flows. Additionally, a complex vortex structure exists along the stagnation streamline which influences heat transfer on the endwall. The flow phenomenon is described and discussed in the open literature but it is still difficult to predict the heat transfer on the endwall and the turbine profile by CFD methods with sufficient accuracy. In this paper it is examined how the flow field in the stagnation region should be simulated using CFD. The effect of meshes with various grid resolutions and turbulence models as k-ε-, k-ω-SST- and DES-turbulence models have been investigated. The CFD-data are compared with the experimental results obtained by Naphthalene Sublimation Method, Pressure Sensitive Paint, Laser Induced Fluorescence and Particle Image Velocimetry. Three cases, namely film cooling on a flat plate, the endwall flow near a symmetrical airfoil and the symmetrical airfoil with endwall film cooling, are examined in detail.

Effect of Coolant Density on Leading Edge Showerhead Film Cooling Using PSP Measurement Technique

The density ratio effect on leading edge showerhead film cooling has been studied experimentally using the pressure sensitive paint (PSP) mass transfer analogy method. Leading edge model is a blunt body with a semi-cylinder and an after body. There are two designs: seven-row and three-row of film cooling holes for simulating vane and blade, respectively. The film holes are located at 0 (stagnation row), ±15, ±30, and ±45 deg for seven-row design, and at 0 and ±30 for three-row design. Four film holes configurations are used for both test designs: radial angle cylindrical holes, compound angle cylindrical holes, radial angle shaped holes, and compound angle shaped holes. Coolant to mainstream density ratio varies from DR = 1.0, 1.5, to 2.0 while blowing ratio varies from M = 0.5 to 2.1. Experiments were conducted in a low speed wind tunnel with Reynolds number 100,900 based on mainstream velocity and diameter of the cylinder. The mainstream turbulence intensity near leading edge model is about 7%. The results show the shaped holes have overall higher film cooling effectiveness than cylindrical holes, and radial angle holes are better than compound angle holes, particularly at higher blowing ratio. Larger density ratio makes more coolant attach to the surface and increases film protection for all cases. Radial angle shaped holes provides best film cooling at higher density ratio and blowing ratio for both designs.

Comparison of Predictions From Conjugate Heat Transfer Analysis of a Film-Cooled Turbine Vane to Experimental Data

Conjugate heat transfer analysis was conducted on a 648 hole film cooled turbine vane using Code Leo and compared to experimental results obtained at the Air Force Research Laboratory Turbine Research Facility. An unstructured mesh with fully resolved film holes for both fluid and solid domains was used to conduct the conjugate heat transfer simulation on a desktop PC with eight cores. Initial heat flux and surface metal temperature predictions showed reasonable agreement with heat flux measurements but under prediction of surface metal temperature values. Root cause analysis was performed, leading to two refinements. First, a thermal barrier coating layer was introduced into the analysis to account for the insulating properties of the Kapton layer used for the heat flux gauges. Second, inlet boundary conditions were updated to more accurately reflect rig measurement conditions. The resulting surface metal temperature predictions showed excellent agreement relative to measured results (+/− 5 degrees K).

Mean and Turbulent Velocity Profile Measurements on the Suction Side of a Film Cooled Turbine Vane

Boundary layer velocity and turbulence profiles were measured on the suction side of a scaled up, film-cooled turbine vane airfoil. There have been a number of previous studies of the velocity profile on a turbine vane, but few have taken velocity profile data with film cooling, and none have taken such data on the suction side of the vane. Velocity and turbulence profile data were taken at two locations on the suction side of the vane — one at a high curvature region and one further downstream in a low curvature region. Data were collected for high (20%) and low (0.5%) mainstream turbulence conditions. For the upstream, high curvature location, velocity and turbulence profiles were found with and without the showerhead blowing and within and outside of the merged showerhead coolant jet. The data for the low curvature, downstream location was taken with injection from the showerhead alone, a second upstream row of holes alone, and the combination of the two cases. It was found that the presence of an active upstream row of holes thickens the boundary layer and increases urms both within and beyond the extent of the boundary layer. Span-wise variations showed that these effects are strongest within the core of the coolant jets. At the downstream location, the boundary layer velocity profile was most strongly influenced by the row of holes immediately upstream of that location. Finally, turbulence integral length scale data showed the effect of large scale mainstream turbulence penetrating the boundary layer. The increase in turbulence, thickening of the boundary layer, and large scale turbulence all play important roles in row to row coolant interactions and affect the film cooling effectiveness.

Convex Curvature Effects on Film Cooling Adiabatic Effectiveness

Surface curvature is known to have significant effects on film cooling performance, with convex curvature inducing increased film effectiveness and concave curvature causing decreased film effectiveness. Generally, these curvature effects have been presumed to scale with 2r/d at the film cooling hole location, where r is the radius of curvature and d is coolant hole diameter. In this study, the validity of this scaling of curvature effects are examined by performing experiments in regions of large and low curvature on a model vane. Single rows of cylindrical holes were placed at various locations along the high curvature section of the suction side of the vane. For the first series of experiments, a single row of holes was placed at two locations with different local surface curvature. The coolant hole diameters were then adjusted to match 2r/d values. Results from these experiments showed that there was better correspondence of film performance when using the 2r/d scaling, but there was not an exact matching of performance. A second series of experiments focused on evaluating the effects of curvature downstream of the coolant holes. One row of holes was placed at a position upstream of the highest curvature, while another row was placed at a downstream position such that the radius of curvature was equivalent for the two rows of holes. Results indicated that the local radius of curvature is not sufficient in understanding the performance of film cooling. Instead, the curvature envelope downstream of the coolant holes plays a significant role on the performance of film cooling for cylindrical holes.

Effect of Upstream Purge Slot on a Transonic Turbine Blade Passage: Part 1 — Aerodynamic Performance

In this paper, detailed experimental results of total pressure loss and secondary flow field are presented for a high turning (127°) airfoil passage in presence of an upstream purge slot (with and without coolant injection). The experiments were performed at Virginia Tech’s quasi 2D linear turbine cascade operating at transonic conditions. Measurements were made at design exit Mach number 0.88 and design incidence angle. The selected coolant to mainstream mass flow ratio (MFR) was set at 1.0%. In order to match engine representative inlet/exit blade loading, a diverging endwall was utilized where the span increased from the inlet to the exit at a 13 degree angle. A 5-hole probe traverse was used to measure exit total pressure. Pressure loss coefficients were calculated both along pitchwise and spanwise directions at 0.1 axial chord downstream of the blade trailing edge. CFD studies were conducted to compliment the experimental results. The backward facing step present with the upstream slot affects the approaching boundary layer and influences the passage horse-shoe vortex strength. The addition of coolant from the purge slot further increased the aerodynamic losses. However, the backward facing step of the upstream slot seems to be the predominant factor in affecting pressure losses when compared to with or without blowing cases. These results provide further understanding of the passage secondary flow characteristics and aid towards improved design of endwall passages. The heat transfer experiments, designed to find the heat transfer coefficient and the film cooling effectiveness are described in detail in part II of this paper [1].

Assessment of a Double Hole Film Cooling Geometry Using S-PIV and PSP

A novel, double hole film cooling configuration is investigated as an alternative to traditional cylindrical and fanshaped, laidback holes. This experimental investigation utilizes a Stereo-Particle Image Velocimetry (S-PIV) to quantitatively assess the ability of the proposed, double hole geometry to weaken or mitigate the counter-rotating vortices formed within the jet structure. The three-dimensional flow field measurements are combined with surface film cooling effectiveness measurements obtained using Pressure Sensitive Paint (PSP). The double hole geometry consists of two compound angle holes. The inclination of each hole is θ = 35°, and the compound angle of the holes is β = ± 45° (with the holes angled toward one another). The simple angle cylindrical and shaped holes both have an inclination angle of θ = 35°. The blowing ratio is varied from M = 0.5 to 1.5 for all three film cooling geometries while the density ratio is maintained at DR = 1.0. Time averaged velocity distributions are obtained for both the mainstream and coolant flows at five streamwise planes across the fluid domain (x/d = −4, 0, 1, 5, and 10). These transverse velocity distributions are combined with the detailed film cooling effectiveness distributions on the surface to evaluate the proposed double hole configuration (compared to the traditional hole designs). The fanshaped, laidback geometry effectively reduces the strength of the kidney-shaped vortices within the structure of the jet (over the entire range of blowing ratios considered). The three-dimensional velocity field measurements indicate the secondary flows formed from the double hole geometry strengthen in the plane perpendicular to the mainstream flow. At the exit of the double hole geometry, the streamwise momentum of the jets is reduced (compared to the single, cylindrical hole), and the geometry offers improved film cooling coverage. However, moving downstream in the steamwise direction, the two jets form a single jet, and the counter-rotating vortices are comparable to those formed within the jet from a single, cylindrical hole. These strong secondary flows lift the coolant off the surface, and the film cooling coverage offered by the double hole geometry is reduced.

Enhanced Film Cooling Effectiveness With Surface Trenches

In this study, cylindrical and fan shaped film cooling holes are evaluated on the blade surface numerically, using the Computational Fluid Dynamics (CFD) tool ANSYS-CFX, with the objective of improving cooling effectiveness by understanding the flow pattern at the cooling hole exit. The coolant flow rates are adjusted for blowing ratios of 0.5, 1.0 & 1.5 (momentum flux ratios of 0.125, 0.5 & 1.125 respectively). The density ratio is maintained at 2.0. New shaped holes viz. straight, concave and convex trench holes are introduced and are evaluated under similar operating conditions. Results are presented in terms of surface temperatures and adiabatic effectiveness at three different blowing ratios for the different film cooling hole shapes analyzed. Comparison is made with reference to the fan shaped film cooling hole to bring out relative merits of different shapes. The new trench holes improved the film cooling effectiveness by allowing more residence time for coolant to spread laterally while directing smoothly onto the airfoil surface. While convex trench improved the centre-line effectiveness, straight trench improved the laterally-averaged and overall effectiveness at all blowing ratios. Concave trench improved the effectiveness at blowing ratios 0.5 and 1.0.

Heat-Flux Measurements for a Realistic Cooling Hole Pattern With Multiple Flow Conditions

Predicting cooling flow migration and its impact on surface heat flux for a turbine operating at design-corrected conditions is a challenging task. While recent data sets have provided a baseline for comparison, they have also raised many questions about comparison methods and the proper implementation of boundary conditions. Simplified experiments are helpful for bridging the gap between the experimental and computational worlds to develop the best procedures for generating predictions and correctly comparing them to experiments. To this end, a flat-plate configuration has been developed that replicates the cooling hole pattern of the pressure side of a high-pressure turbine blade. The heat transfer for this configuration is investigated for a range of flow properties of current interest to the industry using a medium-duration blowdown facility. Heat-flux measurements are obtained using double-sided Kapton heat-flux gauges arrayed in two rows in the axial direction along the centerline of the hole pattern. Gauges are located upstream of the holes, in between rows of holes, and extending far downstream of the last row of holes. New parameters are proposed for analyzing the data including a corrected Stanton number and the length-corrected heat flux reduction parameter. These parameters are used for exploring the influence of Reynolds number and blowing ratio on local heat transfer. In addition, the temperatures of the main flow and the test section walls were varied to determine the effect of cooling on the local adiabatic wall temperature and to enable comparisons using the adiabatic cooling effectiveness.