I made a plane like a litestick and the wing is straight without dihedral. While steering to the left the plane would drift to the right (with controls setup properly)

Thinking it may be dihedral I added winglets to the front wing. The plane works better but is still hard to turn. Someone I know has a plane without dihedral and it turns just fine. The rudder was extended 3 times its size and works better but it is very hard to fly the plane. The length of the fuselage is pretty long. Why would elevator work so good and rudder not?

From : Don Stackhouse

Your rudder is working fine. However, the rudder is not what steers an airplane, at least not directly. The most fundamental function of a rudder is to control yaw, not roll.

Except for some exceptions involving side-lifting airfoils (like what the Air Force tried some years ago on an F-16), an airplane turns in response to banking the wing. The lift of a wing is perpendicular to that wing's surface. When you roll the wing to some bank angle, the lift of the wing is directed partially to that side. It's this sideways component of the wings lift that opposes the centrifugal force from the turn, and pulls the airplane around in a circle.

So the trick is to get the wing to roll to the desired bank angle in the first place. Ailerons and/or wing warping do this directly, by changing the chord line (and therefore the angle of incidence) on one wing relative to the other. The result is that the wing now effectively has a twist in it from one side to the other, something like the pitch of a free-wheeling propeller, and like a free-wheeling propeller it wants to corkscrew as it passes through the air, resulting in a roll in that direction. The airplane will continue to roll as long as the controls are deflected, which is how an aileron roll is done.

For a normal turn, we just deflect the controls until the plane has rolled to the desired bank angle, then center the controls (other than small corrections if necessary to hold the desired bank angle during the turn) during the turn, and finally we give it opposite aileron to roll the plane back to level flight at the end of the turn.

A rudder does NOT directly control roll, it controls yaw. If you deflect rudder, the first thing that happens is that the nose slews to one side (yaws), with the wings initially remaining level. To steer an airplane with rudder instead of ailerons, we need to have an adequate amount of what's called "yaw-roll coupling". This simply means that if the plane is yawed, something about its aerodynamics makes it want to roll in response to that yaw. Without any yaw-roll coupling, a yaw from the rudder causes no change in bank angle, and therefore you don't have any roll control from the rudder alone. This is precisely what's wrong with your 'Arlene'.

You noticed that when steering to the left the plane would drift to the right. That's very observant on your part. Yes, that's precisely what we should expect. When you deflect the rudder for a left turn, the rudder pushes the tail to the right (and therefore the nose to the left). This means that there is now a sideways force to the right acting against the tail, and since the tail is indeed connected to the rest of the airplane, this force tends to push the whole airplane to the right. We see the same thing in helicopters. The torque in the main rotor shaft is a purely twisting effect, but we counter that with a sideways force (the thrust from the tail rotor) acting on the end of a lever arm (the tail boom). This sideways thrust tends to push the entire helicopter sideways, so for a stationary hover we normally have to bank the main rotor the opposite direction a little bit to balance out the thrust from the tail rotor. This is why if you look closely at a hovering helicopter, you will see that it's banked a couple degrees to one side.

Another factor that has a minor effect is the adverse roll effect from the vertical tail. The sideways lift of the deflected rudder is usually above the center of gravity of the airplane, so it has a weak rolling effect that tries to roll the airplane the wrong way. Normally this is insignificant in comparison to the rolling effects of ailerons or wing dihedral, but if your airplane does not have any of these, it could be a problem. With a very tall rudder and a perfectly flat, short-span wing, you could actually have a plane that rolls to the right when you apply "left" rudder.

The most common way to cause some yaw-roll coupling is to add some dihedral. When the plane yaws, the relative wind will tend to hit more under the forward-yawed wing (tending to lift it), and hit more on top of the aft-yawed wing (tending to push it down). This acts just like ailerons, and causes the wing to roll in the same direction as the yaw. Normally, to get adequate stability in turns we only need a few degrees of dihedral, but to get decent roll control you normally need at least about eight degrees of dihedral per side. You can see this in our Roadkill Series Piper J-3 Cub or Curtiss-Wright Junior kits. The kits come with laser-cut dihedral gages that have two different possible settings. If you are building the version with ailerons, you use the low setting for just the scale amount of dihedral. However, for the no-ailerons version (which uses rudder for roll control) you use the higher setting that provides enough extra dihedral to get enough yaw-roll coupling for decent roll control from rudder alone.

Wing sweep can act like dihedral, although its dangerous to rely on that alone. The old rule of thumb is that three degrees of sweep is about equivalent to one degree of dihedral, but that is only true at fairly high lift coefficients. As you go faster and start flying at lower lift coefficients, the dihedral effect decreases and therefore so does the amount of yaw-roll coupling. At zero "G", such as during a push-over form a steep climb, the yaw-roll coupling from sweep becomes zero. Not some thing you want to rely on for your primary means of steering the airplane!

Winglets do have a small amount of dihedral effect of their own, as do wing tips that are beveled up on the underside. This is what you observed when you added winglets to your model. You had some dihedral effect, but still not enough to get sufficient roll control.

I suspect that your friend's model that had roll control from rudder alone and without any geometric dihedral probably had a high wing on a fairly slab-sided fuselage, such as a Piper Cub's. When you yaw an airplane with that arrangement, there are some interactions between the fuselage and the wing roots that act something like dihedral. The Wittman "Tailwind" and other planes with flat wings and very boxy fuselages are known to behave this way. The effect can be very pronounced on a profile fuselage such as one of our Roadkill Series models. For example, the Curtiss-Wright Junior (which has a "parasol" style wing, i.e.: mounted above the fuselage on struts) needs a couple degrees more dihedral than the Piper Cub (high wing mounted directly on the top of a high, slab-sided fuselage) to get the same roll control authority from the rudder.

Your model has essentially no aerodynamically significant fuselage with regard to this particular effect, so it gets no yaw-roll coupling from this.

The same effect in reverse occurs on low-wing airplanes. For this reason a low wing airplane often needs a couple degrees more dihedral than a high wing airplane to get the same amount of roll stability and yaw-roll coupling.

There is more on all of this in the "Ask Joe and Don" section of our website.

Don Stackhouse
DJ Aerotech