Coanda Effect: Understanding Why Wings Work

Archive Notice: This page is part of the Jef Raskin historical archive, preserved for its academic and historical significance.

By Jef Raskin (1994)

Ask a pilot how a wing generates lift. Ask a physics teacher. Ask the author of an aeronautics textbook. You will likely get some version of the same story: air traveling over the curved upper surface of a wing moves faster than air below, creating lower pressure on top, and this pressure difference pushes the wing upward.

That story is, at best, incomplete. At worst, the version taught in most classrooms is demonstrably wrong.

Jef Raskin wrote this paper in 1994 to offer a more satisfying explanation — one grounded in what actually happens when a fluid flows over a surface, and one that doesn’t require hand-waving to explain why symmetrical wings (common on aerobatic aircraft) and inverted flight both work.

The Problem with the Bernoulli Account

The Bernoulli explanation of lift typically goes like this: the upper surface of a wing is more curved than the lower surface, so air traveling over the top has a longer path to cover. To reach the trailing edge at the same time as air traveling along the shorter bottom path, the upper-surface air must move faster. By Bernoulli’s principle — which relates flow speed to pressure in an ideal fluid — faster-moving air exerts less pressure. The net result is lower pressure on top, higher pressure below, and an upward force.

The problem begins immediately with the “equal transit time” assumption. There is no physical reason why air parcels that split at the leading edge must reconvene at the trailing edge. They don’t. This assumption is simply wrong, and can be disproven by measurement: air over the upper surface of a wing reaches the trailing edge before air that took the lower route. The two parcels do not meet again.

Discard the equal-transit-time version, and you might rescue Bernoulli by saying only that the upper-surface flow is faster — which it is — without specifying why. But then the explanation becomes circular: lift is caused by faster upper flow, and faster upper flow is caused by the wing’s shape creating lift. The mechanism is gone.

More concretely: if Bernoulli curvature were the whole story, a flat plate held at an angle would generate no lift. It does. A symmetrical airfoil (curved identically on both surfaces) would generate lift only when level, and not when angled. It generates lift at any angle of attack. A wing flown inverted should create negative lift. It can create positive lift. The Bernoulli-curvature account struggles with all of these.

The Coanda Effect

Henri Coanda was a Romanian aeronautical engineer who, in the early twentieth century, observed and investigated a property of fluid jets that now bears his name. The Coanda effect describes the tendency of a moving fluid — gas or liquid — to follow a curved surface rather than continuing in a straight line.

The phenomenon is easy to observe. Hold a spoon near a stream of water from a faucet. The water bends around the convex back of the spoon and follows its curve rather than splashing off. The stream adheres to the surface. Tilt a pitcher of water: the liquid tends to run down the outside of the pitcher rather than falling freely. These are manifestations of the Coanda effect.

What causes it? Two factors work together. First, the fluid in contact with the surface is slowed by friction (viscosity). This creates a velocity gradient: layers of fluid progressively farther from the surface move progressively faster. Second, the faster-moving outer layers exert a lower pressure perpendicular to their direction of flow — here, Bernoulli is correctly applied, to the difference between adjacent fluid layers, not between top and bottom of a wing. This lower pressure in the outer flow draws the inner, slower-moving fluid outward toward it. Combined with the surface constraint on the innermost layer, the result is that the whole fluid stream curves to follow the surface, attaching to it and bending around it.

Wings and the Coanda Effect

Apply this to a wing in motion. Air approaching the wing separates at the leading edge. The upper-surface air encounters the curved, convex upper surface of the airfoil. Due to the Coanda effect, this air stream adheres to and follows the curved surface rather than continuing in a straight line.

Because the surface curves downward toward the trailing edge, the air is deflected downward. Air exits the trailing edge with a downward velocity component — the wing has redirected a mass of air from a horizontal path to a downward path.

Newton’s third law completes the argument. The wing has exerted a downward force on the air, so the air exerts an equal upward force on the wing. That upward force is lift.

This account explains the cases that trouble the Bernoulli story. A flat plate angled into the wind deflects air downward via the Coanda effect on its upper surface — lift is generated. A symmetrical airfoil at angle of attack deflects air downward on whichever surface is on top — generating lift regardless of which face is up. Inverted flight works for the same reason. The mechanism is the angle of the surface and the fluid’s tendency to adhere to it, not the specific curvature of a particular airfoil shape.

Why the Bernoulli Account Persists

The Bernoulli explanation is not entirely wrong; it is incomplete and often incorrectly applied. Bernoulli’s principle accurately describes the relationship between flow speed and pressure in a moving fluid. The error is in the causal story built on top of it — the equal-transit-time myth, the claim that curvature alone drives differential velocity, and the treatment of Bernoulli as a complete explanation rather than one relationship within a more complex system.

The persistence of the incorrect account is a case study in how simplified explanations, once embedded in textbooks and teacher training, become extremely difficult to dislodge. The simplified Bernoulli story has intuitive appeal. It can be demonstrated with a paper strip and a breath of air. It is taught to children and repeated to adults.

Raskin, characteristically, was interested not just in the correct technical answer but in the problem of incorrect mental models. If people hold a wrong model of how wings work — one that seems to explain the phenomena they observe — they will not notice the model’s failure even when it fails. The equal-transit-time myth survives because those who hold it never encounter a situation that falsifies it in a way they recognize as falsification. The model is wrong, but it is robust against casual evidence.

Significance and Citations

This paper has been cited in multiple Wikipedia articles on aerodynamic lift, the Coanda effect, and the physics of flight. It has also circulated in physics education communities and aeronautics forums as a corrective to the Bernoulli myth.

Raskin wrote the paper in his characteristic style: clear, direct, unwilling to accept conventional wisdom simply because it is conventional. He was not an aeronautical engineer by trade — he was a cognitive scientist and interface designer — but his habit of examining explanations for logical coherence served him as well in physics as in human-computer interaction.

The paper includes quantitative analysis and references to experimental data supporting the Coanda account. It also includes an appendix discussing the historical development of lift theory and where the Bernoulli misapplication originated.

The Raskin Center