By flat plate I mean this:

enter image description here
Source: Physics.SE

Low performance indeed, but how much low? How would:

  • The angle of attack,
  • The speed range,
  • The turn capability,
  • The fuel consumption,
  • Or other significant characteristic

compare to the ones of some reference foil, e.g. a NACA 00xx?

Link to related benchmarks appreciated.


4 Answers 4


I cannot answer all your questions, but maybe point you to some facts to come closer to an answer.

Most important is the thickness of your parallelepiped - more than a few percent will just increase drag, without benefiting performance. It should be as thin as structurally possible. There are lots of model airplanes using flat plates for lift. The most popular are flying discs like that below:

Flying disc with witch-on-broom motive being launched

(picture source)


The lift curve slope of a flat plate is the same as that of regular airfoils, complete with stall. They have a zero-lift angle of attack of 0° (obviously) and separated flow on the suction side. The maximum lift coefficient of 0.7 to 0.8 is reached at moderate angles of attack - details depend on the aspect ratio.

lift curve slope comparison

Your typical four-digit NACA airfoil will have a maximum lift coefficient of 1.2 to 1.6, depending on camber and Reynolds number (picture source).


Due to the missing nose thrust, the aerodynamic force vector will stand almost perpendicular to the plate's plane. It will be slanted slightly backwards due to friction, but the separated flow on the suction side means that friction will mostly happen on the pressure side. As a rough approximation, the drag coefficient of the flat plate is $$ c_D = c_L \cdot \sin\alpha + \frac{0.43}{log\left(\frac{100}{R}\right)^{2.56}} + 0.3 \cdot\delta\cdot\cos\alpha$$ with $R$ the relative surface roughness, $\delta$ the relative thickness and $\alpha$ the angle of attack. The first part is caused by the direction of the lift vector, the second part is the friction contribution and will be dwarfed by the pressure part, and the third part is an approximation for the suction along the thick trailing edge. At $\alpha$ = 5° and a lift coefficient of 0.5, the drag coefficient will already be around 0.05.

Regular NACA airfoils will have a $c_{D0}$ of 0.004 to 0.01, depending on Reynolds number, and their $c_D$ at stall is normally between 0.02 and 0.025.


From the above it is fair to assume that drag in cruise condition will easily be 4 times higher (induced drag is not affected), and the optimum cruise condition will be at a lower speed due to the higher zero-lift drag. On top, the airplane will need a bigger engine to stay airborne, which will reduce its useful load. Expect fuel consumption to increase in line with the drag increase.

The stall speed of an aircraft with a flat plate wing will be approx. 50% higher than with a proper airfoil on the wing, and drag will increase much more at high lift coefficients than usual, which translates into very low sustained turn rates and load factors.

Only the angle of attack range will stay similar. All other parameters will be markedly affected.

  • $\begingroup$ @ArtOfCode: I used to do multiplication with \cdot, but found that shift-opt-9 produces the same centered dot. Your edit did not change the appearance of the equation on my screen, but is this different in that other OS? Should I better use \cdot again for better compatibility? $\endgroup$ Commented Sep 30, 2015 at 12:02
  • $\begingroup$ The manual dot you add gets put in thanks to Unicode - just not all computers and browsers will support it. MathJax's \cdot is, IIRC, compatible in most if not all places. $\endgroup$
    – ArtOfCode
    Commented Sep 30, 2015 at 12:05
  • $\begingroup$ @PeterKämpf: shouldn‘t the MachNumber be considered as well? Looking at the wing profile of a F16 or Concord they look much more like a flat plate due to the transonic aerodynamics. $\endgroup$
    – rul30
    Commented Jan 16, 2018 at 5:52
  • $\begingroup$ @rul30: You are right; I answered only for subsonic speed. $\endgroup$ Commented Jan 16, 2018 at 9:27

One last note: flat plate wings work extremely well on small scale lengths, as for example in the insect world. Consider that a dragonfly can hover, perform flat turns at speed, fly inverted, and accelerate from zero to 30 MPH in a couple of seconds with two pairs of flat wings, and 250 million years +/- of selection pressure have not improved their design much.

Because a dragonfly's wings twist significantly during operation, they are not strictly speaking without camber, but anyone who has ever watched them pluck gnats out of the air cannot but be amazed at what they can accomplish, aerodynamically speaking.

  • 1
    $\begingroup$ Yes the Square-Cube law does them many favours. $\endgroup$
    – Koyovis
    Commented Jan 16, 2018 at 4:16

It depends of course: laminar or turbulent flow, and at which Reynolds number. $C_D$ for AoA zero looks like this (found here)

enter image description here

Flat plate aerodynamics are relevant for helicopter blades and wind turbines, and measurements have been conducted over 0 - 360 deg of AoA with results that are of interest for fixed wing as well. For instance from this document:

enter image description here

The $C_L$ of NACA 0012 over 180 degrees shows that the real gain in lift is within the first 10 degrees of AoA, after which the airfoil stalls, but $C_L$ starts rising again after its initial stall behaviour due to flat plate aerodynamics. The 0012 profile is quite thin, and the maximum $C_L$ is reached at 45 degrees - however if we look at 10 deg AoA, the 0012 $C_L$ is around 1 while the equivalent flat plate $C_L$ would be around 0.4 (extend the sine wave graph 15 < AoA < 170 to zero).

$C_D$ follows a similar pattern, again showing large gains until stall occurs and then continuing according to flat plate aero. at 10 deg, $C_D$ is barely above zero instead of the 0.12 value obtained when the sine curve is extended.

enter image description here

So the NACA profile has a much higher lift for a much lower drag than an equivalent flat plate. In order to get to a $C_L$ of 1, the AoA of a flat plate needs to be 30 deg instead of the 10 deg of the NACA. At 30 deg, $C_D$ of the wing is 0.6 instead of the 0.02 of the NACA, 30 times higher! So turn rate and other parameters are only of interest if the engines can deliver 30 times more thrust.

  • 2
    $\begingroup$ Your comparison makes the flat plate look worse than it is. It reaches its optimum at a lower lift coefficient, so the comparison at c$_L$=1 is misleading. This is not to say that a flat plate is not a poor choice for subsonic aircraft - it certainly is. $\endgroup$ Commented Jul 17, 2017 at 8:20
  • 1
    $\begingroup$ Yes good point for the cruise, although it would make for much higher landing speeds if we can't get to $C_L$ of 1. $\endgroup$
    – Koyovis
    Commented Jul 17, 2017 at 9:31
  • 1
    $\begingroup$ yes, or you use a much bigger wing. $\endgroup$ Commented Jul 17, 2017 at 14:51

The answers already provided discuss aerodynamics in some detail.

One other significant characteristic not so far discussed is structural integrity. In a structure like a wing, there will be trade-offs between strength, thinness and lightness: the more you have of one, the less you can have of the others.

By the time you're dealing with very thin structures, then even choosing stronger-but-heavier materials won't help enough to be useful in a wing that has to support an aircraft in the air.

Even if a flat plate wing had advantageous aerodynamic properties, its structural weaknesses would rule it out of many applications.

  • 1
    $\begingroup$ The wing of the F104 Starfighter was pretty flat. $\endgroup$
    – Koyovis
    Commented Oct 8, 2017 at 19:32
  • 3
    $\begingroup$ Flat is not the same as thin! A cube has only flat surfaces. $\endgroup$ Commented Oct 8, 2017 at 20:46

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .