\require{AMSmath} \require{eqn-number}

Building The U. Maine Trainer

- Direct Drive Ailerons -

(last update: 27 Jan 2013)

Presumably you have a main wing that looks something like the following:

The plywood end plates on the wing tips and inner trailing edges were not covered in earlier sections.

The wing tip end plates have blind 4-40 nuts so that things can be bolted to the wing tip. In this case another end plate with a servo mounted to it. The servo crown gear will be the outer anchor and pivot point for the aileron. In this way the aileron is a direct drive aileron, there are no intervening links or control horns.

The inner trailing edge end plates, one on each side of the wing, have pins to act as the other pivot points of the two ailerons. The photo below highlights these areas of the wing.

One question comes to mind, with a direct drive aileron.

Will the servo be able to move the aileron, and hold the aileron in position, in a moving air stream?

Most control surface implementations (rudder, elevator, aileron ...etc) have a control horn mounted on the control surface. A control rod then runs to a servo arm or to an intervening bell crank and control rod to the arm mounted on the crown gear of the servo. In each case there are lever arms, which operate around points of rotation, thus consuming or providing torque. Each lever arm represents some level of mechanical advantage or disadvantage, depending on whether one views the system from the driver or the consumer. In either case there is an overall mechanical advantage.

If the device (an aileron) which consumes torque has a 10 mm arm, and the provider (the servo) also has a 10 mm arm the net mechanical advantage is 1 or no gain. The mechanical disadvantage at the servo arm is 10 mm. In order for the torque of the servo's crown gear to move the control rod connection back and forth, it must do so through an arm of 10 mm in length. At the aileron the opposite is required. The control rod connection point can apply torque to the aileron pivot through an arm of 10 mm in length. The ratio of one arm to the other is the mechanical advantage.

MA = \frac{10 \ mm}{10 \ mm}
MA = 1

In a way the two arm lengths cancel each other out. This would not be the case if the two arms were of different lengths. Note that the important length is the distance of the control rod to the center of rotation. The overall physical length of the control horn, and the servo arm is incidental. What matters is how far away the control rod is anchored from its respective arm's pivot point.

What this suggests is that there is seldom much in the way of overall mechanical advantage, in any given implementation. There are always exceptions to this observation. However, in practice this tends to be the case. The "take away" point is that whatever torque is available at the servo crown gear will be, more or less, all that is available to control the aileron. It will not make much difference whether you drive it with control horns or directly.

#### However, there are a few more points...

One point is that, in conventional simple aileron schemes, a single servo drives both ailerons. This means that roughly twice as much torque is required from the servo. Thus only half the servo's available torque is available to work either of the ailerons.

The other point is that the linkage from the servo to the aileron has friction, and therefore also consumes torque. However, this source of inefficiency can be made quite low. Conversely it can also be quite severe, if done poorly.

A direct drive aileron scheme eliminates the linkage.

It also requires two servos, one for each half of the wing. This means that control of each aileron enjoys the full force of a single servo.

Two servos can also be used in conventional schemes but you can't avoid the linkage or linkages.

Below is a picture of a complete, except for some additional shaping, direct drive aileron wing.

#### So... How might a direct drive aileron be constructed...

If one pivot point of a given aileron will be a servo crown gear, that means there will need to be another pivot on the other end of the aileron.

There are several ways to do this. Depending on where one choses to pivot the aileron. In practical terms there are only two. One is the very leading edge of the aileron. This is classically the location of aileron hinges. The other obvious place is to set back the pivot point from the aileron leading edge, one half the width of the aileron at the leading edge. This is where I chose to define the aileron pivot point.

Photos of the construction to follow will make this choice more clear.

Below is a highlight of the location of the left aileron's inner pivot point.

Below is a detail view of the right wing aileron inner pivot point. The brown washer is root beer bottle plastic. No... it isn't Kapton (\$\$\$\$\$\$). I'm silly but not THAT silly.

I had a choice. I could make the inner aileron pivot point the servo crown gear or the outer aileron pivot point the servo crown gear. The outer location more or less implies the wing tip.

I chose to mount the servo out on the wing tip because is seemed easier to implement. I didn't want to do all the surgery required on a small wing. At the wing tip there are more options during construction, and it is easier to get at the assembly for modifications.

The only real problem with the wing tip location is that the mass of the servo and foundation are so far outboard. All that mass will reduce the roll rate. In this case, that is fine. I am favoring stability over aerobatics. Having roll damping mass out on the wing tip should aid stability.

#### Inner Aileron Pivot

Below is a view of where an inner aileron pivot will be affixed.

Next is a picture of the pieces which make up the pivot.

The wood is simple 2.8 mm (110 mil, 7/64 in) aircraft plywood. I have a bunch of it left over from a large scale Super Decathlon kit. I just cut the shapes I need from the die cut left overs. The wire is #2 control rod stock (1.82 mm, 71 mil).

The ailerons were cut and shaped from the wing stock previously. Just trace the outline of aileron on the plywood. Then rough cut the shape from the stock. Sand the rough cut down from there. The leading edge doesn't need to be rounded. It can be square, so that it butts up against the main wing.

The key point is to center the pivot point along the chord, and not let it drift too far forward. If it is too far forward the leading edge of the aileron will bind against the wing. If that happens the wing will have to be sanded back until the aileron rotates freely.

The trick is to get enough binding that only a small amount of sanding is needed. A gap of less than a millimeter is good. More than a millimeter should be avoided.

In the photo below the test fit shows there is too much forward and too little aft. No big deal. The forward end can be sanded back. I decided to live with the shortened aft end.

In the photo above, also note the root beer bottle washer.

Be sure the anchor end of the wire does not protrude from the plane of the opposite face of the plywood. If it is all good, go ahead and glue it to the wing. Make sure to carefully center the pin, and be sure it is not pointing out of the plywood as some odd angle. It should be pointing parallel down the wing toward the tip. Be sure to get that right. It will be tricky to bend the wire later, so get it right before it is glued.

Before the servo, and wing tip assembly is built, an aileron is needed. The aileron is doubly important because its length will dictate the location of the servo.

In principle it is better to start with an aileron blank that is longer than needed. The idea being it is easier to cut or sand something down than it is to make it "grow" longer. Lengthening an aileron is not difficult but it is unnecessary, and tedious.

#### Aileron

A completed direct drive aileron is shown in the next photo.

At each end there is a plywood end-plate/bushing, which looks something like the following:

In principle this type of end plate, at each end, could support direct drive or conventional control. The end plates are simply reinforced places where the pivot point rod holds the aileron against the slip stream. There is no harm in extending the end plate down the full chord.

Direct drive simply affixes a servo control arm to one of the end plates, on center with the pivot point. If a quad or dual arm is used, the other arms are removed. The single arm extends down the chord of the aileron, with the crown socket pointing outward. The arm can be screwed to the end plate for extra security.

In this implementation the arm was epoxied to the end plate, and sandwiched with a piece of foam. Here's what that looks like:

Be careful not to allow epoxy to bleed into the crown gear socket. If epoxy gets in there it will be impossible to press the socket onto the servo crown gear.

Here is another view with a ruler giving an idea of the size.

To avoid flutter (bending oscillation under load), the leading edge was reinforced.

Reinforcement was a length to 3/4 oz fiberglass at least 2 cm wide epoxied to the full leading edge using West System epoxy. This effectively forms a "C" section beam the full length of the aileron. It makes the aileron much stiffer. No harm in using heavier glass or additional layers... within reason.

#### Wing tip mount servo assembly.

Above is an early version of the wing tip servo assembly.

Next is the mounting plate on the wing tip.

The hole near the leading edge is used to set the pitch of whatever gets mounted to this foundation. The load is handled by one or both of the 4-40 blind nuts toward the aft of the wing section. The notch is for clearance of the servo's gear housing. The small hole between this notch and the first of the 4-40 blind nuts is a clearance hole for a wood screw that secured one end of the servo.

The notch is for clearance of the servo's gear housing.

The small hole between this notch and the first of the 4-40 blind nuts is a clearance hole for a wood screw that secured one end of the servo. This hole is no longer needed since this inner servo mounting screw has been abandoned.

The first end plate (left wing tip) did not have its 4-40 blind nuts installed. The wing foam was cut back to install them. The photo below shows the two scares where this was done.

A set of long nose pliers was used to pinch the blind nuts in place. The scares of the pliers can be seen on the plywood in the previous photo.

The foam was cut back all the way through the section. After the blind nuts were installed, pieces of quarter inch foam were then stuffed into the holes, epoxied, and sanded flush.

Below is an early fit test.

The photo shows that too much of the servo crown gear is not engaging with the crown socket.

To correct this one of the three plies of the plywood was carved off...

After splitting off the ply, the surface was sanded down, and the servo re-installed.

Now the fit looks like this,

The crown gear is now deeply set into the crown gear socket. As long as the epoxy holds the servo arm to the aileron, it isn't going anywhere.

If you are worried about the epoxied servo arm coming loose, you can add a screw to one of the servo arm's holes. If you don't have a small enough screw, a small piece of control rod could be used. The idea is to provide some sort of shear (parallel to the slip stream) reinforcement. A screw provides an orthogonal steal post which opposes this shear.

Above the entire right wing assembly is in place.

Next the tip assembly needs some sanding along the wing profile, and a fairing needs to be fashioned for all the stuff hanging outboard of the wing tip. This is the principle short coming of this implementation. All the junk hanging out in the breeze at the end of the wing tip.

Even so... it should fly the way it is. Especially if the wing is lengthened. The intent is to have high lift, docile handling, and excellent low speed characteristics (good camera platform). The airframe can be cleaned up for greater efficiency later.