The Community for Technology Leaders


Pages: pp. 5-6


To the Editor:

Bob Colwell's discussion of lift ("Leave Bernoulli Out of This," At Random, May 2003, pp. 10-12) is better informed than many elementary texts on the subject, but it leaves out a more interesting question about how wings fly.

Colwell points out the errors in the usual explanation—the one that depends on a wing's greater curvature on top. His excellent counterexamples are that flat wings generate lift and that planes can fly upside down. It was making those very arguments in grade school—and having my teacher tell me to shut up—that helped turn me into a scientist. Logic told me I was right even though the teacher said I was wrong.

Although, like Colwell, I prefer a Newtonian approach, it is possible to correctly explain lift in terms of Bernoullian pressure differences. In Life in Moving Fluids (Princeton University Press, 1994), an authoritative yet entertaining book, Steven Vogel explains flight from a Bernoullian standpoint and applies it to a wide variety of physical and biological phenomena.

Colwell says that Richard von Mises "obtains his formulas for lift by applying Newton's second law." In fact, von Mises uses circulation and Bernoulli's law to explain lift in the traditional way ( Theory of Flight, Dover Publications, 1959, p. 174)—although he does it properly, and not in the naive way that Colwell correctly criticizes.

Colwell notes that "what holds up a plane" is that its wings "deflect a large mass of air downward; the reaction is to push the wing upward." It can be shown that most of the air that is deflected downward is acted on by the top of the wing. It is easy to say the plane rises due to the wing's reaction to this action, but we need to ask how it does this and to physically link the motion of the air to the resultant force on the wing.

Jef Raskin, Pacifica, Calif.;

To the Editor:

As usual, Bob Colwell's "Leave Bernoulli Out of This" was a highlight of Computer's May issue. I would add a postscript by noting that Newton's laws nicely explain the seemingly arcane rotary-wing aerodynamics: The rotary wing is a propeller that pushes air downward. In addition, the bird analogy is the hummingbird, which is the official symbol of the American Helicopter Society—now called the Vertical Flight Society.

William R. Dunn, Solvang, Calif.;

To the Editor:

As a sailor, I have had a practical interest in the same question that Bob Colwell asks in "Leave Bernoulli Out of This." After all, a boat sail is nothing more nor less than a vertical wing—with the peculiarity that it is (usually) extremely thin, a bent sheet, in fact.

The first basic rule of sailing is to let the sail out as far as possible without the wind puckering it in the direction opposite to the desired curve—in other words, minimize the angle of attack and let the curve do the work. A skilled helmsman may increase the angle of attack slightly under certain circumstances.

Bernoulli does come into the question to some extent. The air passing over a sail covers the same distance on both sides. As it passes over the sail, the air tends to continue in a straight line (Newton), leading to a drop in pressure around the back (outside) of the sail. Similarly, an increase in pressure occurs on the front (inside) of the sail. The net change in the direction of the air movement—shoving a mass of air sideways—drives the boat.

For a sail, there is no differential distance over the curved surface from front to back to make any difference. Without a pressure drop over the outside of the bulge, however, the air will not follow the curve. Pressure drops, so velocity increases. Similarly, inside the bulge, a concomitant rise in pressure occurs with deflection of the airflow. Pressure rises, so velocity falls.

The major implications of Bernoulli's principle would therefore seem to be in the dynamic behavior of the air after leaving the trailing edge of the sail or wing: fast, low-pressure air adjacent to slow, high-pressure air.

Peter Bissmire, Deri, South Wales, UK;


To the Editor:

Although I agree with many of the points that Venkat N. Gudivada presents in "The Computing Profession at a Crossroads"(The Profession, May 2003, pp. 92, 90-91), I would like to offer a few additional comments based on my observations and understanding.

First, there is a core in computer science studies. This core encompasses both theories and designs for fast, smart, reliable computation machines (system software included) and algorithms that solve a broad scope of problems. Finding people with different academic backgrounds in computer science studies is not surprising; the development of interdisciplinary areas is not surprising, either. However, computer science is not a simple mixture.

Second, a major problem with software engineering is that it lacks discipline. Perhaps software engineering should focus more on problems at the application-domain level. On one hand, thinking that software engineering is computer science is dangerous; on the other hand, thinking that programming is software engineering is wrong. The lack of a clear vision about what software engineering does or should do has contributed to the gap between software engineering education and industry's expectations.

Third, the IT industry might have been wrong in overlooking the importance of formal education. Some software engineering practices are so primitive that anyone who has certain certification can work on the projects. Consequently, predicting the quality and performance of those projects is difficult. That has been one reason for the failures and delays that we see all too often.

Yibing Wang, Birmingham, Ala.;

66 ms
(Ver 3.x)