JTrade names or manufacturers' names are used in this report for identification only. This usage does not constitute an official endorsement, either expressed or implied, by the National Aeronautics and Space Administration. SummaryThe purpose of the present work was threefold: (1) to determine if a free-stream turbulence length scale existed that would cause the greatest augmentation in stagnation-region heat transfer over laminar levels; (2) to investigate the effect of velocity gradient on stagnation-region heat transfer augmentation by free-stream turbulence; and (3) to develop a prediction tool for stagnation heat transfer in the presence of free-stream turbulence. Heat transfer was measured in the stagnation region of four models with elliptical leading edges that had ratios of major to minor axes of 1: 1, 1.5: 1, 2.25: 1, and 3: 1. Five turbulence-generating grids were fabricated; four were square mesh, biplane grids made from square bars. The fifth grid was an array of fine parallel wires that were perpendicular to the model spanwise direction .Heat transfer data were taken at Reynolds numbers ranging from 37 000 to 228 000. Turbulence intensities were in the range of 1.1 to 15.9 percent while the ratio of integral length scale to leading-edge diameter ranged from 0.05 to 0.30. Stagnation-point velocity gradient was varied by nearly 50 percent.Stagnation-region heat transfer augmentation was found to increase with decreasing length scale but no optimum length scale was found. Heat transfer augmentation due to turbulence was found to be unaffected by the velocity gradient near the leading edge. A correlation was developed that fit heat transfer data for the square-bar grids to within ±4 percent.

An experiment was conducted to measure the heat transfer from a heated cylinder, three diameters in length, in crossflow in an array of identical circular cylinders. Both in-line and staggered array patterns were studied. The cylinders were spaced on a 2.67 × 2.67 pitch. The Reynolds number range, based on cylinder diameter and average unobstructed channel velocity, was from 5000 to 125,000. Turbulence intensity profiles were measured for each case at a point one-half space upstream of the row containing the heated cylinder. The basis of comparison for all the heat transfer data was the single row with the heated cylinder. For the in-line cases, the addition of a single row of cylinders upstream of the row containing the heated cylinder increased the heat transfer by an average of 50 percent above the base case. Adding up to five more rows caused no increase or decrease in heat transfer. On the other hand, adding rows in the staggered array cases resulted in average increases in heat transfer of 21, 64, 58, 46, and 46 percent for one to five upstream rows, respectively.

No abstract

A study has been conducted at the NASA Lewis Research Center to investigate the mechanism that causes free-stream turbulence to increase heat transfer in the stagnation region of turbine vanes and blades. The work was conducted in a wind tunnel at atmospheric conditions to facilitate measurements of turbulence and heat transfer. The model size was scaled up to simulate Reynolds numbers (based on leading edge diameter) that are to be expected on a turbine blade leading edge. Reynolds numbers from 13,000 to 177,000 were run in the present tests. Spanwise averaged heat transfer measurements with high and low turbulence have been made with “rough” and smooth surface stagnation regions. Results of these measurements show that, at the Reynolds numbers tested, the boundary layer remained laminar in character even in the presence of free-stream turbulence. If roughness was added the boundary layer became transitional as evidenced by the heat transfer increase with increasing distance from the stagnation line. Hot-wire measurements near the stagnation region downstream of an array of parallel wires has shown that vorticity in the form of mean velocity gradients is amplified as flow approaches the stagnation region. Finally smoke wire flow visualization and liquid crystal surface heat transfer visualization were combined to show that, in the wake of an array of parallel wires, heat transfer was a minimum in the wire wakes where the fluctuating component of velocity (local turbulence) was the highest. Heat transfer was found to be the highest between pairs of vortices where the induced velocity was toward the cylinder surface.

No abstract