I am interested in building the Astro Hog Jr. from plans...

...it seems like a good model to build as a low wing trainer. I have a motor and gearbox combo that should work out well for it. My question is that it has a 1 1/2 inch dihedral in each wing panel ( I am assuming 1 1/2 " at wing tips?? ) would this be enough dihedral for simple 3 ch operation...thr, elev, rud........??? Would I need more dihedral. The plans do show ailerons but I would like to build for 3ch function. The plans can be found at http://my.pclink.com/~dfritzke/ in pdf format.

From : Don Stackhouse

The short answer to your question is "No".

Just scaling off the plans, I estimate that dihedral equates to a little more than about 4 1/2 degrees per panel. In my experience, a good rule of thumb for decent roll response from rudder alone is at least 6 degrees per panel for a high wing, and about 8 degrees per side minimum for a low wing. This suggests to me that you need about double the amount of dihedral called for on the plans if you want good, strong roll response from rudder alone on your model.

One other thing that can help is to use a wingtip that's just beveled off on the underside. This acts like extra dihedral in that portion of the tip.

Note, roll control requirements are somewhat subjective; what's good enough for some folks is flat-out unsafe in the eyes of others. For example, the original GWS Lite Stik came with little more dihedral than the flex in the wing in flight caused by the plane's own weight. Obviously someone at GWS thought that the roll authority from that was adequate. However, just about everyone I know had to modify the kit in one way or another to give it something more than the almost nonexistent roll authority it had in the stock form. I notice that the Slow Stik comes with quite a bit more dihedral!

When an airplane is yawed, the pressures on both sides of the fuselage change. This interacts with the inboard portions of the wings to create a dihedral-like effect. In the case of a high wing, it acts like additional dihedral, while on a low wing it has the opposite effect. This is why low wing airplanes typically need about 2 degrees per side more dihedral than a high wing to get the same total dihedral effect.

Note, the effect depends on the shape of the fuselage cross-section. A well rounded fuselage tends to minimize the effect, while a square box with sharp corners accentuates it. On a profile fuselage it can really be noticeable.

On our Roadkill Series Piper "Cub" and Curtiss-Wright "Junior" we offer the builder option of making them with scale dihedral and coupled rudder+ailerons (it is possible to install an extra servo to provide independent rudder and ailerons), or with extra dihedral and no ailerons. There's about a 6 degree per side difference between the two options, with the Junior needing just a little more than the Cub because the Junior's parasol wing doesn't get as much dihedral effect from the fuselage as the Cub does. For sport flying I prefer the extra dihedral and no ailerons versions. It is possible to roll both versions of both kits, it's just that rolls with the no-ailerons versions will be a little bit less axial, requiring a little more pull-up at the beginning of the roll to offset the rudder's tendency to kick the nose down a little in the knife-edge portion near the end of the roll.

Most of the Roadkill Series models use coupled aileron and rudder, using one servo. It's really quite simple to set up. In the typical arrangement we use music wire torque rods to operate the ailerons, with control horns at the fuselage to move them. We run a short pushrod from each side of the servo output to the aileron horns. We have an extra hole on the end of the aileron horns, and we run a pushrod from one of those to the rudder horn. That's all there is to it. On some of the airplanes like the Junior and the WW I airplanes, we mount the servo (with a 3-horned output) in the fuselage, run pushrods upwards from the horizontal arms of the servo to operate the ailerons, and run the rudder pushrod to the vertical arm on the servo. If the tailwheel strut pivot isn't in line with the rudder hinge line, we just strap on a short extra pushrod at the back to operate the tailwheel. With the very simple pushrod couplers we use for the Roadkill models, it's all very simple to connect and to adjust.

We have tested these models with just rudder, and with just aileron, using nearly scale dihedral, and the results were not pretty. They definitely needed both.

One pitfall you do have to watch out for on a coupled rudder+aileron model is the aileron rigging. If your model is out of trim in roll, and you just crank in some aileron trim in the transmitter, you will also be introducing a dose of rudder with that aileron. This can cause some perplexing problems. The model will now track wings-level in roll, but the rudder that came with the aileron trim will now be yawing the plane to one side. Particularly on a profile model, this can cause some massive drag problems. The thing to do is to trim the airplane so it flies level with the stick centered, then do some turns in both directions. For a given bank angle (30 degrees or so is a good angle for these tests), the plane should turn equally well in both directions, with about the same turn radius in both directions. Note in particular if the plane seems to track well in one direction, but flies with the nose yawed towards the outside of the turn in the other direction. If you see that, then you need to readjust the rudder (without changing the ailerons) till it tracks in yaw equally during turns in both directions. If you have coupled nosewheel or tailwheel steering as well, then after setting up the aileron trim and then the rudder trim, do some taxi tests to fine-tune the steering on the ground.

Getting back to the no-ailerons option, one other problem you may have to deal with is the possible need for more vertical tail area. There is a balance required between wing dihedral and vertical tail effects. With too much vertical tail and/or not enough dihedral you get spiral instability; i.e.: in a turn, the plane wants to tighten up by itself into steeper and steeper banks until it winds itself into a "graveyard spiral". It's usually not a major issue on a powered aerobatic model, but it can be a nuisance. On a sailplane where you need to make nice, consistent round circles to fly in a thermal, it can become a major problem.

OTOH, if you have too much dihedral and/or not enough vertical fin effect, the plane can have dutch roll problems, where the plane wants to oscillate side to side in both roll and yaw, something like a falling leaf. Theoretically it's the vertical tail moment arm that's involved. However, in most airplanes the fuselage has some fin effect of its own (particularly on a profile model, which is why our Fokker Triplane has great yaw stability despite that tiny rudder and no fin). Because of this, when you add area to the fin+rudder while keeping the fuselage the same, the effective moment arm of the fuselage+vertical tail does move back closer to the tail.

Generally the design problem boils down to finding that combination of dihedral + tail that hits that "sweet spot" in between the zone of designs with dutch roll problems and the zone of designs with spiral instability. The size of this sweet spot depends on a lot of factors, but the amount of mass in the plane's extremities, particularly the tail and wingtips, is one of the biggest factors. Our profile models tend to be very light in the tips, and those profile tails are also very light (which is also why we don't have too much trouble getting the C/G correct, even on those notoriously short-nosed WW I models), so Roadkill Series models tend to be very tolerant in this regard. However, if you don't do a good job of keeping the entire tail assembly (including the fuselage) light, as well as the wingtips, and also keep all the heavy stuff mounted as close as possible to the C/G, it's entirely possible to end up with a model where that "sweet spot" has a "negative size"; i.e.: there is no combination of dihedral + tail that doesn't have dutch roll, spiral instability, or both!

One more factor to consider on some airplanes is sweep. Yes, an aft-swept wing does act like it has some extra dihedral. Winglets also add some dihedral effect. This is why the Rutan VariEze and LongEze have some anhedral (downward dihedral) in their wings. Because of the sweep and the winglets, they would have dutch roll problems from too much dihedral effect if they didn't counteract some of it with the anhedral.

There's an old rule of thumb that says that three degrees of sweep acts like about one degree of dihedral. Unfortunately, like many rules of thumb, this one can be a gross oversimplification, to the point that it can get you in trouble.

The problem is that the dihedral effect of wing sweep depends on the lift coefficient. It is zero at zero lift, and reaches its max at max lift, near stall. The effect is linearly proportional to both the lift coefficient and to the cosine of the sweep angle (which means it's pretty small for sweep angles less than about 15 degrees or so).

The problem is that if you try to use sweep for some of your dihedral on a no-ailerons model, you might get adequate roll authority near the stall, but all that roll control authority will disappear at high speed. In addition (and this is true whether you have ailerons or not), the model will act like it has varying dihedral. Near the stall it may have dutch roll problems due to too much dihedral effect, which when coupled with a swept wing's tendency to tip stall can result in some pretty wicked snap rolls at exactly the wrong time (such as right in the middle of flaring for touchdown!). At the same time, it will possibly have an insufficiency of dihedral effect at high speed, resulting in spiral instability problems. Getting decent handling qualities from a swept wing can be quite a challenge!

As usual with most things in airplane design, seemingly minor changes in one little parameter can have ripple effects that influence all sorts of other things.

Mike, in your case I'd be inclined to keep the ailerons, either as coupled ailerons+rudder on one servo, or separate with two servos. Going to ailerons alone, even if you include coupled tailwheel steering, is likely to have some adverse yaw problems in flight, and a total lack of steering ability when the mains are still on the ground on takeoff and landing, but the tailwheel is off the ground. Going to rudder alone could be made to work, and work well, but be prepared to do some experimenting with both the wing dihedral and the tail design. If your goal is to build just one model, and you just want to eliminate the ailerons to make the building quicker and easier, I'd say that approach is likely to end up being "penny wise and pound foolish". When you start to alter an already proven design, odds are that like it or not, you will have to become an airplane designer at some point in the project.

Don Stackhouse
DJ Aerotech