(-4r) 



. = o<° 



Therefore, visually, the specification of a positive velocity gradient up to 



a certain position on the foil surface, and a similar specification of a 



reversed flow condition proceeding in the opposite direction up to a certain 



position, indicates that separation must occ-ur between these respective 



locations. When the interval between such points is very small, the 



prediction of the location of the separation point is accurate. Figure l6 



shows such an interval between positive and negative velocity gradients. 



The two wires in the flow are straight and excited with a d-c voltage . One 



is positioned at the front stagnation point of the foil shape and illustrates 



a positive velocity gradient. The other is smoothly attached perpendicular 



to the surface of the foil and supplies bubbles into a reversed flow region. 



In the thin interval between these groups of bubbles lies the separation 



point on the surface of the body. The separation streamline is contained 



between these two groups of hydrogen bubbles . Motion-picture photography 



of such a separation region obviously enables determination of any cyclic 



motion of this point and the correlation of such motion with wake phenomena. 



One of the more severe limitations of the hydrogen-bubble technique, 



k. 

 which has also been foiond by other investigators, is the low-velocity 



flows to which the technique has been confined. This is primarily due to 



the delicate hardware used to install bubbles in a water flow . To install 



a 0.001- in. diameter platinum wire in a water flow so that the wire remained 



straight, the following was done. A heavy rod or steel wire frame, which 



was adequately insulated from the water, was made into the configuration 



visible in Figures 1 and 2. The thin cathode wire was soldered across the 



two cantilevered portions of the wire holder. Before the two ends of the 



wire were soldered, however, the cantilevered portions of the wire holder 



were bent together very slightly. When the remaining end of the wire was 



soldered, these portions of the holder were released and stretched the 



thin wire taut . As one can easily imagine, the amount of wire-holder 



bending required extensive practice before conipetent wire installations 



21 



