A wrong --but simple to understand-- theory for lift states:

  • Air molecules that were close before being separated at the leading edge must be close again after they join back at the trailing edge. For this to happen, molecules must travel the two paths in the same time.

    enter image description here

  • As the path is longer on the upper side, due to the camber of the wing, air is accelerated on the upper side. By Bernoulli's equation, static pressure decreases when velocity increases, so the pressure is lower on the upper side, and higher on the lower side.

  • Lift is the force created by this pressure differential on the wing sides.

(What is wrong is not Bernoulli's equation, but the assumed relationship between path length and speed.)

Such theory is, for instance, presented as correct in this book.

More at Nasa.


We know this theory is wrong and lift created by path length difference would be a very small percentage of the lift actually measured (and even null for symmetrical airfoils).

Let's assume a rectangular wing with a simple cambered airfoil, at the critical angle of attack:

  • What would be the pressure differential according to equal transit time difference? What percentage of actual lift does it represent?
  • 3
    $\begingroup$ Excellent question, though I'm afraid I'm not going to understand a word of the math that this will turn up. :( Someone (ahemPeterKampfcough) will come up with a final number and I'll understand that, though. $\endgroup$ – FreeMan Oct 19 '17 at 12:08
  • 2
    $\begingroup$ That book is an absolute joy to read. I've never seen so many fallacies condensed in so little space. As it turns out, induced drag decreases lift, flaps are used to trim the aircraft, there are slots [sic] in front of the wings to reduce the air 'piling up', airliners are equipped with a steady supply of oxygen and the cabine is only pressurised to keep your blood from boiling. $\endgroup$ – Sanchises Nov 16 '17 at 8:14

The theoretical lift coefficient of a NACA 4412 at the critical angle of attack (about 16 degrees) is approximately 1.7. \begin{equation} c_\mathrm{l}=1.7 \end{equation}

For our purposes, that same airfoil has its upper "surface" about 1.03 times as long as its lower, which can be seen by using the distance formula on the coordinates or integrating the cambered NACA airfoil equation. Here is the distance-formula method applied to the Airfoil Tools numbers (thanks to ROIMaison for making me check this calculation): 4412

On average, therefore, the fallacy would predict the air moving along the upper surface at about 1.03 times the velocity as along the lower in order to join up properly.

Making all the assumptions from the NASA link, Bernoulli's equation says \begin{equation} P_1+\frac{\rho V_1^2}{2}=P_2+\frac{\rho V_2^2}{2} \end{equation} Let's represent the lower surface with subscript 1 and the upper with subscript 2. The respective velocities are averages across the surfaces. We have \begin{align} P_1-P_2&=\frac{\rho (1.03V_1)^2}{2}-\frac{\rho V_1^2}{2}\\ &=\frac{609}{20000}\rho V_1^2\\ \end{align} So the average difference in pressure coefficient (i.e., the lift coefficient) is given by \begin{equation} \mathrm{avg}(c_\mathrm{p_{ETT}})=\frac{609}{10000}\frac{V_1^2}{V^2}=c_\mathrm{l_{ETT}} \end{equation} where $V$ is the freestream velocity and ETT stands for equal transit time.

We have another expression for pressure coefficient that we can apply to get a final value. We end up with \begin{equation} c_\mathrm{l_{ETT}}=\frac{609}{10000}\frac{V_1^2}{V^2}=1-\frac{V_1^2}{V^2} \end{equation} or \begin{equation} \frac{V_1^2}{V^2}=\frac{10000}{10609} \end{equation} and therefore \begin{equation} c_\mathrm{l_{ETT}}=\frac{609}{10609}\approx.0574 \approx .034c_\mathrm{l} \end{equation} which is less than 5% of the actual value. However, it is important to keep in mind that the fallacy predicts no lift change with angle of attack (just one of the many reasons it is silly), and so the relationship between $c_\mathrm{l_{ETT}}$ and $c_\mathrm{l}$ varies with angle of attack for a given airfoil. For the 4412, they meet at just over a -5° AOA. For a symmetric airfoil, they meet at zero AOA. For any other airfoil, the calculation must be redone with the correct path lengths and actual lift coefficients, although magnitude of the result is not likely to change dramatically.

  • 1
    $\begingroup$ Do you have any evidence/support for "that same airfoil has its upper "surface" about twice as long as its lower (this can be seen by using the distance formula on the coordinates or integrating the cambered NACA airfoil equation)"? It seems like a lot to me $\endgroup$ – ROIMaison Nov 16 '17 at 10:21
  • 3
    $\begingroup$ I had an extra term in the distance calculation that made the result way too much. Good catch! I've edited the result (a substantial change) and I appreciate your pointing this out. $\endgroup$ – Peter Schilling Nov 16 '17 at 13:02
  • 1
    $\begingroup$ That's an excellent answer, the kind I was looking for. $\endgroup$ – mins Nov 16 '17 at 13:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.