The Balanced-Flap Anomaly

As well as the empirical support briefly mentioned by O’Gorman, some fur­ther evidence favorable to Lanchester emerged in the following year. It came from Bairstow’s own laboratory and arose from the attempt to clarify some disconcerting experimental results about control surfaces. With the con­struction of ever-larger aircraft, the forces that pilots had to exert on the con­trols became correspondingly greater. To overcome this problem the controls were “balanced,” that is, part of the area of the control surface was positioned in front of the hinge around which it turned so that some of the aerodynamic forces worked with, rather than against, the pilot. An experimental study of balanced controls was carried out at the National Physical Laboratory in early 1916 by John Robert Pannell and Norman Robert Campbell. Pannell, who had been on the staff since 1906, was the senior assistant in the Aerodynamics Department. He was a familiar figure who bicycled to work every morning, arriving on wet days with an umbrella held aloft in one hand while steering with the other. His main concern was with tests on full-size airship, and he was to die in 1921 when the R38 met with disaster on its trial flight.68 The balancing experiments, however, were conducted using the “flaps” on an air­craft wing. “Flap” was the old name for the lateral control surfaces, today called ailerons, which allow the pilot to bank and roll the aircraft.

Pannell’s co-worker in these experiments was a Cambridge experimental physicist who had played a prominent role in early debates about relativity theory.69 Campbell, who had been seconded to the NPL for war work, was in the process of writing a book on scientific method, Physics: The Elements, which was published after the war and was to prove an influential work in the philosophy of science. Campbell argued for the importance of models in scientific inference and theory construction.70 On this occasion the models that interested Campbell were not models of the atom or the electromagnetic field but scale models of the wings of a 110-foot-span biplane that was under construction at the Royal Aircraft Factory.71

The “flaps” of the projected aircraft ran along the rear edge of the outer portion of the wings but also included the tips of the wings themselves. When the part of the control surface that was on the trailing edge was lowered, then the part of the wingtip that was connected to it, and that was in front of the axel on which it pivoted, went up. The whole tip of the wing was thus part of the flap. This construction was meant to give the desired balance. Pannell and

Campbell wanted to find the proportion of the area that should be in front of the axel. The model wings were placed vertically in the 4 X 4-foot tunnel at a wind speed of 40 feet per second. Different proportions of fore and aft area were tested with the wings set at 0°, +4°, and +12° to the wind and with the flaps (and tips) put at a variety of angles relative to the main wings.

It proved impossible to find a fully satisfactory balance. Frustratingly, there was no ratio of the areas, fore and aft of the pivot, that fully balanced over the desired range of angles. Also, when looked at in detail, the results had some odd features. On occasion, where the experimenters had expected to be able to detect forces at work on the wingtip, there weren’t any: “In particular it was found that when the main planes of a biplane were inclined at +12° to the wind, there was no moment on the portion of the wing flap forward of the hinge, if this flap was inclined at an angle of -5° to the wind.”72 This result suggested that the flow of air near the wingtip was itself at a negative angle to the undisturbed flow. Given the conventions for designating angles positive or negative, this meant that the air near the wingtip was moving upward rela­tive to the wing. The air was going round the tips from what, during normal flight, would count as the lower to the upper surface.

In order to shed light on this, Pannell and Campbell conducted a further, qualitative investigation of the flow near the tips.73 Using a direction and ve­locity meter to plot the velocity components of the moving air, they found what they called “a very simple and obvious explanation” for their “remark­able results.” It became clear that air was indeed flowing round the wingtips from the lower to the upper surface. This was associated with a movement of air along the span of the wing, that is, not just from the leading to the trailing edge of the wing but lengthwise along it. There was a component of outward movement, toward the tips, on the lower surface and an inward movement, away from the tips, on the upper surface. Pannell and Campbell argued that

The presence of this flow round the wing tips affords, in outline at least, an explanation of the result on the balancing of wing flaps. . . . For, if there is a marked flow near the wing tip directed from the lower to the upper surface, a plane parallel to this flow will experience no wind force. Now it was precisely when the balancing flap was inclined at a negative angle to the wind, so that its plane lay along a flow having a component from the under to the upper side of the plane, that the experiments indicated that there was no force on it. (141)

The qualitative study also addressed the component of flow along the chord of the wing, that is, in the direction of flight rather than around the tips or along the span. Measurements were taken with the direction and ve­locity meter to build up a picture of the disturbed flow. Next, the steady, undisturbed flow was subtracted from it. If the resultant flow was made up of two parts, the free stream plus a circulation, this subtraction would expose the circulatory component. This is exactly what it did. In the experimenter’s words, it emerged that “the component of the wind disturbance which is par­allel to the direction of flight is in the direction of flight almost everywhere below the wing, and in the opposite direction everywhere above the wing. There is therefore some indication of the cyclical motion of the air round the wing in the vertical plane of flight which has been assumed by Mr. Lanchester in his discussion of the theory of the aerofoil” (142). Although the full path of the circulation had not been traced, those parts of it above and below the wing had been factored out and, as it were, exposed to view. Here again was evidence for the reality of the circulatory component that was central to Lanchester’s theory.