Chapter 4: The High-Speed Propeller Program
[127] When we embarked on the project to measure the pressure distribution on the rotating blade of an axial-flow compressor in 1944 (refs. 96, 97), the ultimate application of the technique in the back of our minds was propeller pressure distributions at high speeds. If pressure data could somehow be obtained they could be analyzed to yield the blade section characteristics throughout the regions of the propeller over which the flows were supercritical and transonic. Not only were such airfoil data nonexistent in 1944, but also no method existed to apply airfoil data with confidence to the conditions existing over the outer region of the blade-conditions characterized by three-dimensional effects and a strong radial velocity gradient. The action of centrifugal force on the blade boundary layers was an additional uncertainty. Clearly the full-scale propeller program at 16-foot would be importantly enhanced if a technique for pressure measurement could be evolved.
[128] Assuming the pressure transfer device could be made to work under high-speed conditions, we recognized that the next most difficult problem was how to install hundreds of pressure taps in the highly stressed blades without losing their structural integrity. I brought up this question at lunch with C. S. MacNeil, Chief Engineer of the Aeroproducts Division of General Motors, during his visit to Langley on September 1, 1944. Aeroproducts was producing hollow propellers fabricated from steel sheet and it had occurred to me that perhaps pressure tubes could be installed internally during fabrication. MacNeil thought they could and he promised to study the problem. About a week later he called to say the scheme was feasible and that he would like to build four test blades for us, each containing two sections with 24 pressure taps per section. I started the procurement with a memorandum to Mr. Miller describing our plan in detail (ref. 147). Some time after the work had been started at Aeroproducts, MacNeil, in his mid-thirties, suffered a fatal heart attack. The project continued but never recovered from the loss of MacNeil's zealous interest. When the test blades were delivered, many of the tubes were found to be blocked, and many others were leaking. None of the blades was ever used in research.
Corson took up the problem during the summer of 1946. By that time, the compressor-blade pressure measurements had been obtained successfully, and Corson's idea was to apply a similar method of tube installation in solid Duralumin propeller blades. In a sketch dated September 19, 1946, he suggested locating pressure tubes near the surface in radial slots on the test blade and covering them with a suitable filler. Langley shop supervisors improved on this scheme. They retained the tubing by peening the edges of the grooves and then filling them with a metal spray and refinishing the blade to its original contours. Holes were then drilled at the outermost station at the tip for the first tests. After completion of the test run, this row of holes was filled with a low-melting point alloy, and a second row of holes was then drilled at the adjacent inboard radial station. In this way, a total of 264 pressure taps were eventually installed in each blade, and only 24 radial tubes were needed. The first successful results with this technique were achieved in the fall of 1947 on the standard NACA test propeller, using the mercury-seal transfer device (ref. 148). By the end of 1949, five additional propellers had completed [129] pressure-distribution testing (ref. 149) using the improved mechanically sealed transfer device developed by R. S. Davy (ref. 98). A total of 47 blade sections was investigated and 6554 individual pressure distributions were measured, in addition to wake surveys, force tests, and blade deflection measurements.
One of the first important uses of the high-speed pressure data was in the derivation of supercritical and transonic blade-section force coefficients for use in a general method for predicting propeller performance at high flight speeds involving transonic conditions on the propeller. Significant departures from two-dimensional airfoil data are evident in the outboard regions, chargeable to the combined effects of tip relief, Mach number gradients, radial flow of the boundary layers, and possibly to an induced-camber effect. The method successfully predicted the performance of the 4-foot propellers tested at airspeeds up to Mach 0.93 in the repowered 8-foot tunnel program (ref. 150).