An alternative to ribs for internal heat transfer enhancement of gas turbine airfoils is dimpled depressions. Relative to ribs, dimples incur a reduced pressure drop, which can increase the overall thermal performance of the channel. This experimental investigation measures detailed Nusselt number ratio distributions obtained from an array of V-shaped dimples (δ / D = 0.30). Although the V-shaped dimple array is derived from a traditional hemispherical dimple array, the V-shaped dimples are arranged in an in-line pattern. The resulting spacing of the V-shaped dimples is 3.2D in both the streamwise and spanwise directions. A single wide wall of a rectangular channel (AR = 3:1) is lined with V-shaped dimples. The channel Reynolds number ranges from 10,000–40,000. Detailed Nusselt number ratios are obtained using both a transient liquid crystal technique and a newly developed transient temperature sensitive paint (TSP) technique. Therefore, the TSP technique is not only validated against a baseline geometry (smooth channel), but it is also validated against a more established technique. Measurements indicate that the proposed V-shaped dimple design is a promising alternative to traditional ribs or hemispherical dimples. At lower Reynolds numbers, the V-shaped dimples display heat transfer and friction behavior similar to traditional dimples. However as the Reynolds number increases to 30,000 and 40,000, secondary flows developed in the V-shaped concavities further enhance the heat transfer from the dimpled surface (similar to angled and V-shaped rib induced secondary flows). This additional enhancement is obtained with only a marginal increase in the pressure drop. Therefore, as the Reynolds number within the channel increases, the thermal performance also increases. While this trend has been confirmed with both the transient TSP and liquid crystal techniques, TSP is shown to have limited capabilities when acquiring highly resolved detailed heat transfer coefficient distributions.

Jet impingement is often employed within the leading edge of turbine airfoils to combat the heat loads incurred within this region. This experimental investigation employs a transient liquid crystal technique to obtain detailed Nusselt number distributions on a concave, cylindrical surface that models the leading edge of a turbine airfoil. The effect of hole shape and differing hole inlet and exit conditions are investigated. Two hole shapes are studied: cylindrical and racetrack-shaped holes; for each hole shape, the hydraulic diameter and mass flow rate into the array of jets is conserved. As a result, the jet's Reynolds number varies between the two jet arrays. Reynolds numbers of 13,600, 27,200, and 40,700 are investigated for the cylindrical holes, and Reynolds numbers of 11,500, 23,000, and 34,600 are investigated for the racetrack holes. Three inlet and exit conditions are investigated for each hole shape: a square edged, a partially filleted, and a fully filleted hole. The ratio of the fillet radius to hole hydraulic diameter is set at 0.25 and 0.667 for the partially and fully filleted holes, respectively, while all other geometrical features remain constant. Results show the Nusselt number is directly related to the Reynolds number for both cylindrical and racetrack-shaped holes. The racetrack holes are shown to provide enhanced heat transfer compared to the cylindrical holes. The degree of filleting at the inlet and outlet of the holes affects whether the heat transfer on the leading edge model is further enhanced or degraded.

An alternative to ribs for internal heat transfer enhancement of gas turbine airfoils is dimpled depressions. Relative to ribs, dimples incur a reduced pressure drop, which can increase the overall thermal performance of the channel. This experimental investigation measures detailed Nusselt number ratio distributions obtained from an array of V-shaped dimples (δ/D = 0.30). Although the V-shaped dimple array is derived from a traditional hemispherical dimple array, the V-shaped dimples are arranged in an in-line pattern. The resulting spacing of the V-shaped dimples is 3.2D in both the streamwise and spanwise directions. A single wide wall of a rectangular channel (AR = 3:1) is lined with V-shaped dimples. The channel Reynolds number ranges from 10,000–40,000. Detailed Nusselt number ratios are obtained using both a transient liquid crystal technique and a newly developed transient temperature sensitive paint (TSP) technique. Therefore, the TSP technique is not only validated against a baseline geometry (smooth channel), but it is also validated against a more established technique. Measurements indicate that the proposed V-shaped dimple design is a promising alternative to traditional ribs or hemispherical dimples. At lower Reynolds numbers, the V-shaped dimples display heat transfer and friction behavior similar to traditional dimples. However, as the Reynolds number increases to 30,000 and 40,000, secondary flows developed in the V-shaped concavities further enhance the heat transfer from the dimpled surface (similar to angled and V-shaped rib induced secondary flows). This additional enhancement is obtained with only a marginal increase in the pressure drop. Therefore, as the Reynolds number within the channel increases, the thermal performance also increases. While this trend has been confirmed with both the transient TSP and liquid crystal techniques, TSP is shown to have limited capabilities when acquiring highly resolved detailed heat transfer coefficient distributions.

Impinging jets are often employed within the leading edge of turbine blades and vanes to combat the tremendous heat loads incurred as the hot exhaust gases stagnate along the exterior of the airfoil. Relative to traditional cylindrical jets, racetrack shaped impinging jets have been shown to produce favorable cooling characteristics within the turbine airfoil. This investigation experimentally and numerically quantifies the cooling characteristics associated with a row of racetrack shaped jets impinging on a concave, cylindrical surface. Detailed Nusselt number distributions are obtained using both a transient liquid crystal technique and commercially available CFD software (Star CCM+ from CD-Adapco). Three geometrical jet inlet and exit conditions are experimentally investigated: a square edge, a partially filleted edge (r/dH,Jet = 0.25), and a fully filleted edge (r/dH,Jet = 0.667). Additionally, to investigate the effect of high crossflow velocities at the inlet of the jet, a portion of the flow supplied to the test apparatus radially bypasses the impingement section. Thus, the mass flow rate into the test section is varied to achieve the desired inlet crossflow conditions and jet Reynolds numbers. As a result, jet Reynolds numbers (ReJet) of 11500 and 23000 are investigated at supply duct Reynolds numbers (ReDuct) of 20000 and 30000. The results are compared to baseline cases where no mass bypasses the test section. Additionally, the relative jet – to – jet spacing (s/dH,Jet) is maintained at 8, the relative jet – to – target surface spacing (z/dH,Jet) is 4, the target surface curvature – to – jet hydraulic diameter (D/dH,Jet) is 5.33, and the relative thickness of the jet plate (t/dH,Jet) is 1.33. Measurements indicate that the addition of fillets at the edges of the jet orifice and the introduction of significant crossflow velocity at the inlet of the jet can significantly degrade the cooling characteristics on the leading edge of the turbine blade. The magnitude of such degradation generally increases with increasing fillet size and inlet crossflow velocity. The V2F model is adequate for predicting the flow field and target surface heat transfer in the absence of inlet crossflow; however, it is believed the turbulence within the jet is overpredicted by the CFD leading to elevated heat transfer coefficients (compared to the experimental results).

No abstract