# Aerodynamicists: What is the advantage of negative dihedral on the Harrier?

I guess the title says all that needs saying.

(Or, why do the wings droop?)

It helps the Harrier swim better.

I guess this means that the roll problem is so serious that it is worth paying the penalty of computer stabilization (or, alternatively, never allowing the pilot to relax) to avoid the tendancy to turn upside down in normal, high speed flight.

That probably isn’t too much of an added complication though. As I understand it, modern fighters are generally so unstable in order to get maneuverablity that they have to be computer stabilized anyway.

Hmm. I must have it arse about. I was under the impression it was the opposite: the fighters need to be computer stabilised to get manoeuvreability, so they have to be unstable. I have heard that in the early days of computer stabilisation, they achieved this simply by adding some weights to the back of the aircraft.

Here’s my understanding (and please correct me if I’m wrong): when an aircraft tilts to the side (i.e. roll), it starts to slide down sideways towards that side. Any force that tends to act against this sliding motion gives the aircraft roll stability. There are three in general:
[ol]
[li]Dihereal. If the aircraft slides to the right, the right wing gets more push from the sideways motion than the left wing. So the right wing is pushed up.[/li][li]High wing. If the aircraft slides to the right, air pushes against the right side of the fuselage and creates a pocket of higher pressure air below the right wing. This pushes the right wing up.[/li][li]Wing angle (i.e. swept-back wing). If the aircraft slides to the right, airflow onto the right wing’s leading edge becomes more perpendicular and generates more lift, pushing it up.[/li][/ol]
The Harrier has high-mounted swept-back wings so those two factors alone give it more than enough (i.e. too much) roll stability. Too much stability means not enough maneuverability, so you want to reduce the stability with a negative dihereal.

I don’t think so. WWII fighters like the P-51, Spitfire etc. are marginally stablein roll. In order for a roll-to-turn airframe to turn it must first roll so as to get the wing lift vector pointed in the direction of turn. In order to get into a turn as rapidly as possible you need a low roll moment of inertia and little or no tendency of the airframe to oppose roll, for example, dihedral.

Well, not exactly. Aerodynamicists can clean this up as needed. If a plane rolls the lift vector is pointed in the direction of roll. If the plane starts to slide sideways the vertical sabilizer immediately provides a torque to turn the fuselage into the relative wind.

When a plane with dihedral is in level flight the lift vector of each wing is pointed up and in toward the fuselage. When the plane rolls the lift vector of the down wing becomes more vertical than the lift vector on the up wing. It therefore has the larger vertical component lifting it back to level flight than the up wing and so the plane returns to level flight.

I don’t think so - you need torque to counteract a roll. Torque is created by the component of the lift vector perpendicular to the wing, not the vertical component.

Aren’t Harriers too old to be using computer aided stabilization they seem to have been originally designed in th 60’s (at least the early ones)?
http://www.aerospaceweb.org/aircraft/attack/harrier/

I’d been led to believe that one of the reasons high-wing mounted planes often have droopy wings is to lower the effective center of lift so that less control deflection (and consequent drag) is required to achieve a desired roll rate. Puts the vertical location of the lift center closer to the center of roll rotation (and also changes its position a bit) and/or vertical center of mass (depending on how you want to look at it), in part to compensate for the excess roll stability that a high wing often affords. But that’s just what my aircraft design and flight dynamics professors told me.

scr4: David Simmons is correct that the primary effect of dihedral in roll stability is that the downwards wing has more of its lft vector pointing at the ground, and therefore generates a force opposing the force that tipped the airplane over in the first place. Also, the downwards wing will generate more induced drag, which will tend to pull the nose of the aircraft in that direction.

The reason high-wing airplanes have more roll stability is mostly because of the pendulum effect - a fuselage is a heavy weight hanging under the wing. Roll the airplane, and the weight of the fuselage wants to pull it level again.
As the stability of the design goes up, so do the forces keeping the aircraft in trim. This has two negative consequences - it reduces overall performance of the aircraft, and it makes the aircraft less manoeverable. For example, the tail of an airplane produces a force pointing down. The center of lift of the wing is behind the center of gravity. So without the tail, the aircraft would pitch over on its nose. The downforce of the tail keeps it straight. If the airplane goes into a dive, it will speed up. This increases the negative force on the tail, which pulls the nose back up again. But the price for this is the drag induced by the lift the tail must create to keep the plane stable. And when you want to change pitch rapidly, you have to overcome the natural balance between these forces.

If you make the tail smaller, and move the center of gravity back a bit, you can reduce both the pitching force and the downforce of the tail. The plane is still stable, but when something knocks it out of trim, the smaller forces act less strongly, and the plane takes longer to get back into trim. So it will osclillate back and forth more as it seeks its trimmed condition. Reduce it enough, and you can make the plane neutrally stable. Now you are flying more efficiently, and you can manoever better because those forces are much smaller. But the plane, if knocked out of trim, will stay in the new attitude because there are no forces to oppose the new trim condition.

Go a bit further, and make the plane unstable. Now it’s wildly manoeverable, but impossible to fly. So you need a computer to fly it. So computers are used because neutral or negative stability enhances manoeverability. It’s not the computer creating the manoeverability.

In the case of the Harrier, the proper design solution to whatever its problem may be is not necessarily obvious. The combination of a swept wing, large heavy fuselage, high mounted wing position, and a large operating speed range makes it non-obvious what kind of aerodynamic tweaking it needs.

I still don’t get it. Isn’t “roll” a rotation around the axis parallel to the plane’s path? If so, why is it the vertical component that counts, and not the component perpenducular to the wing? The former changes, but the latter stays the same regardless of roll angle.

Why does drag increase?

I don’t see the analogy. On a pendulum, the torque depends on pendulum angle. If it is offset from the equilibrium point (bottom), torque increases and tends to push the pendulum back towards the equilibrium point. On a plane, the net torque on the fuselage from the wings doesn’t depend on roll angle.

Isn’t the lift vector always perpendicular to the wing? It is a lifting surface afterall.

Yes, that was my point. So the torque (in the roll axis) caused by the wings’ lift shouldn’t depend on roll angle, right?

A)This link shows how dihedral restores the normal flight attitude of an airplane.

B) Because the lower wing generates more sustentation, thus there is an increase on the parasite drag it produces.

C) When there´s assimetric lift (one wing lifts more than the other) as shown on the link, it´s easy to see how the higher lift wing will go up and rotate the plane on it´s longitudinal axis.

I mean, the induced drag… :smack:

I wouldn’t think so, but my experience tells me different. In shallow banks, I need to hold some alieron into the bank to hold the bank angle and prevent the plane from leveling out. For steep banks, opposite alieron is needed to hold the bank angle and prevent the plane’s bank from steepening. Although I think that has something to do with the fact that I am also turning, so the “up” wing is travelling faster than the “down” wing. So for steep banks the higher lift produced by the “up” wing overcomes the natural roll stabilty of the airplane.

But what you’re saying makes sense. The wings are fixed relative to the plane’s roll axis no matter how it’s rolled around the axis.

Ah, that makes sense. Lousy gravity, messing up the symmetry of the situation.

Others have already answered but here is another take on the subject.

Since the down wing has a larger vertical component than the up wing and the wings are aerodynamically matched the net vertical component of lift will be at the same place both wings. As this diagram shows the down wing not only has a larger vertical force but A is a larger lever arm than B as well. The resulting net torque rolls the plane back toward level. Notice also that the horizontal force on the up wing is greater than on the down resulting in an overbanking tendancy that must be counteracted by a slight reversed control stick pressure opposite the bank after the bank is established.

This might come as a shock but we had computers in the 1960’s. A special purpose computer, otherwise known as an autopilot, to provide stability could have been in the Harrier from its beginning.