Readers who remain unconvinced about the thrust generating properties of winglets should do an internet search for winglets and see what they find.
The following is an extract of something I found by this means.
The important bit is in the third paragraph. (I'm afraid I haven't got to grips with the process of inserting links yet but you can find the whole item at airspacemag.com).
QUOTE
In 1976, shortly after an energy crisis sent fuel prices skyward, Richard Whitcomb, a NASA aerodynamicist, published a paper that compared a wing with a winglet and the same wing with a simple extension to increase its span. As a basis for comparing both devices, the extension and the winglet were sized so that both put an equal structural load on the wing. Whitcomb showed that winglets reduced drag by about 20 percent and offered double the improvement in the wing's lift-to-drag ratio, compared with the simple wing extension.
The aspect ratio of a wing is the relationship between its span and its chord-the distance from leading edge to trailing edge. A U-2 has a high aspect ratio; an F-104 has a low one. A wing with high aspect ratio will provide longer range at a given cruise speed than a short, stubby wing because the longer wing is less affected, proportionally, by the energy lost to the wingtip vortex. But long wings are prone to flex and have to be strengthened, which adds weight. Winglets provide the effect of increased aspect ratio without extending the wingspan. One rule of thumb says that for an increase in wing-bending force equal to that of a one-foot increase in span, a wing's structure can support a three-foot winglet that provides the same gain as a two-foot span extension.
The airflow around winglets is complicated, and winglets have to be carefully designed and tested for each aircraft. Cant, the angle to which the winglet is bent from the vertical, and toe, the angle at which the winglets' airfoils diverge from the relative wind direction, determine the magnitude and orientation of the lift force generated by the winglet itself. By adjusting these so that the lift force points slightly forward, a designer can produce the equivalent of thrust. A sailboat tacking sharply upwind creates a similar force with its sail while the keel squeezes the boat forward like a pinched watermelon seed.
If winglets are so great, why don't all airplanes have them? Because winglets are a tradeoff: In the highly visible case of the 777, an airplane with exceptionally long range, the wings grew so long that folding wingtips were offered to get into tight airport gates. Dave Akiyama, manager of aerodynamics engineering in Boeing product development, points out that designing winglets can be tricky-they have a tendency to flutter, for example. "We find that it really doesn't matter what kind of wingtip device you use-they're all like span," he says. "The devil is in the details. Span extensions are the easiest and least risky." In the past, winglets were more likely to be retrofitted to an existing wing than to be designed in from the start, but now that is beginning to change. Unlike those tailfins on cars, winglets really work.
Originally published in Air & Space/Smithsonian, Aug/Sep 2001 . All rights reserved.
UNQUOTE
I'm not sure how much Mr Whitcomb or NASA earned from this research, but I suspect that it was not enough to enable them to design an engine that runs on water.
Last edited by Keith.Williams.; 11th July 2004 at 19:06.