Dihedral wings seems to be very common (almost if not all low-wing airliners, many modern gliders,...). Polyhedral seems to be less common and mainly found in relatively old aircraft (F4 Phatom, F4U Corsair, DR400,...). By polyhedral, I mean at least 2 dihedral angle with a sharp discontinuity between both. I don't include wings whose diheadral angle varies smoothly along the wingspan (such as this B787).

But I was surprised to discover that this design (polyhedral with sharp angle between the distinct diheadral sections of the wing) was elected in at least 2 project whose first flight was in the last 10 years (namely the solar impulse and the qinetiq zephyr). I do realize that both examples I gave have similar missions (stay airborne as long as possible, and thus fly slowly; gather as many sun-energy as possible, and thus fly as horizontally as possible) and may thus have similar design.

Given recently-designed aircraft with polyhedral wing exist, what are the pro/cons of such design over well spread dihedral wings and over the smoothly varying dihedral angle?

EDIT: note that the polyhedral does not necesseraly increase the dihedral angle as we go away from the wing root (Be 12, Habicht,...)


3 Answers 3


While dihedral wings offer increased stability in the roll axis, they also decrease the lift that the wings provide. A polyhedral wing provides a compromise: a steep dihedral at the tip of the wing in order to maximize stability, and a shallower dihedral angle closer to the root in order to maximize lift. Poly wing

Therefore, a polyhedral wing allows the best of both worlds for a wing design. Other applications of a polyhedral wing are utilitarian; the gull wing on the Vaught F4U Corsair is designed so that the landing gear is positioned close to the ground while maintaining roll stability and aerodynamics.

Corsair landing gear

  • $\begingroup$ Sorry, I can't help it. I can't look at the wing on a Corsair without feeling, all, funny, ooo err, it's happening again, nurse, nurse! The <center> cannot hold it is too late. The light is blinding, my ears hurt, oh man, there it goes again,,,,for pitys.......sake.......help.......me....... $\endgroup$
    – Simon
    Commented Feb 23, 2016 at 18:44
  • 2
    $\begingroup$ Oh man, @Simon has lost it! Darn, he wrote really good answers! $\endgroup$
    – TomMcW
    Commented Feb 23, 2016 at 19:06
  • 1
    $\begingroup$ @Timpanus I'm quite confused. Your answer says dihedral increase roll stability. But this website, this answer, this one and this one say that it provides slip-roll coupling (and not roll stability). $\endgroup$
    – Manu H
    Commented Feb 23, 2016 at 23:28
  • $\begingroup$ @ManuH: It's both, depending. The increase in roll stability is primarily due to the CG being effectively lowered due to a dihedral effectively raising the wing by raising the wingtips (some may explain roll stability to be due to one wing being straighter in a bank but that explanation is false). On a low wing aircraft (like WW2 fighters) dihedral may not increase roll stability but only neutralise it (from negative roll stability due to low wing). Actually.. now that I've written it, I guess increasing roll stability from negative to zero is technically increasing it. So, yeah. $\endgroup$
    – slebetman
    Commented Feb 24, 2016 at 2:36
  • $\begingroup$ I always thought that the Corsair's distinctive wings were a way to keep the enormous prop away from the ground while using relatively short landing gear. But I don't recall an authoritative source for that. It's just one of those things I always heard. $\endgroup$ Commented Jul 19, 2019 at 19:41

Polyhedral wings are used in aircraft for various reasons, mainly to improve roll stability without significant change in the overall wing design.

  • The original design of the F-4 Phantom (which started as a 'super F3' had a constant swept wing with anhedral.

F4 prototype

By U.S. Navy - U.S. Navy National Museum of Naval Aviation photo No. 1996.253.7320.028, Public Domain, https://commons.wikimedia.org/w/index.php?curid=24564607

However, wind tunnel tests indicated stability problems, which were were overcome with giving the outer wings a 12$^\circ$ dihedral, corresponding to a 3$^\circ$ overall (with changes in wing sweep too). This saved a major redesign of the wing, while also allowing the spar to pass below the fuselage.

  • Another example of this is the Antonov An-12 'Cub'. The early prototypes had a straight high wing. However, test flights showed the aircraft having poor lateral and longitudinal stability, which was remedied by increasing anhedral on the outer wing panels.


By Sefjo - Own work, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=29397593

  • $\begingroup$ Re "The original design of the F-4 Phantom (which started as a 'super F3' had a constant swept wing with anhedral."-can you give a source for the idea that original design wing had anhedral? I have never heard that before. (Have heard that the dihedral angle of the tips was a later addition to fix stability issues.) I hope that you are not basing that "anhedral" idea solely on the included photo, as it is very difficult to assess the 3-d geometry of a wing from a single view at that sort of angle. In my opinion the geometry of the wing in that photo might be completely flat (no dihed or anh.) $\endgroup$ Commented May 31 at 19:28

Lets consider the dihedral angle from three points of view:

  • Stability, the more conventional one, it mainly influences spiral stability, which is the tendency of the aircraft to enter into a spiral due to a perturbation (i.e.: gust): The aircraft should not enter this spiral, it is allowed to enter in a spiral only when the motion is slow enough for the pilot to safely recover.

  • Reduction of induced drag, a more "unconventional" aspect to analyse the dihedral angle, requires higher dihedral angles.

  • Integration, keeping clearance between the ground and the wing tip and/or engines/propellers. This is specially important for the ground phases.

From the stability point of view the dihedral main effect is on the $C_{l_{\beta}}$ (and $C_{l_{r}}$) derivative. This derivative plays a major role (along with other such as ($C_{n_{r}}$ $C_{l_{p}}$, $C_{n_{p}}$, $C_{n_{r}}$) in defining the spiral and dutch-roll modes.

These other derivatives usually depends heavily on the design of the vertical stabilizer design (except $C_{l_{p}}$, which is mainly influenced by aspect ratio). Since the vertical stabiliser design is more complex than trading the dihedral angle (whose main influence is on $C_{l_{\beta}}$) , the latter can be used to "adjust" the desired spiral and dutch-roll modes, to desired values.

The dihedral angle, when exaggerated, has a secondary effect on induced drag as well. This can be seen from the configuration on the bottom right in the figure below. Here it can be noted how a (very) high dihedral angle has a positive effect on the aerodyanmic efficiency factor $\epsilon$. Generalizing, the higher the displacement of the wing (the dihedral angle) the better it is for induced drag (think about the new B787 or A350 designs). Since induced drag is the main component of drag in low speed, and since the aircraft you mentioned (solar impulse, kinetiq zephir), are solar powered and flying slowly I can imagine that they have a stronger use case to minimize drag.

Finally the dihedral angle can be used to mitigate "integration" problems. The other example you mentioned I think fall in this category. On the Beriev Be 12,the propellers would be too close to the water, expecially when the aircraft is floating (see here). And similar the Habicht wihtout the gull wing would have a wing tip too close to the ground. Minimize Indiced Drag

  • 2
    $\begingroup$ Your third bullet point ends rather abru $\endgroup$
    – FreeMan
    Commented Apr 8, 2016 at 12:06
  • $\begingroup$ Also, second to last paragraph: "The secon aspect is to consider the dihedral angle". Should that be "The second aspect..."? There are enough terms I'm not 100% sure on that I'm not comfortable just making the edit myself. /proofreader $\endgroup$
    – FreeMan
    Commented Apr 8, 2016 at 12:10

You must log in to answer this question.

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