While we're talking about stagger bees.

I've got one and once I got used to how "tippy" it is I really enjoy flying it, my big trouble is that with the first motor set up... it would ground loop often. It flew a bit fast and heavy but it was fun. I've since changed the motor set up..., the thrust is up the weight is down but the ground looping has gotten so bad that its become a major struggle to get it into the air. Any thoughts on what to do about the ground loop trouble? I really like the little tumble bug but 8 or 9 tries to get airborne is silly.

From : Don Stackhouse

The design of tailwheel aircraft landing gear geometry is a compromise. If the main wheels are too far forward it will tend to groundloop, and also tend to bounce at touchdown. With the mains too far aft it will tend to nose-over.

The nose-over problem should be obvious. For the mains to hold the nose up, they have to be sufficiently ahead of the C/G

The bouncing problem on touchdown is because the upward force of the ground against the wheels at touchdown, acting ahead of the C/G, tends to push the nose up. There are several ways to mitigate this. When making a "wheel" landing (i.e.: touch on the mains first, with the fuselage level), it's common practice to jab some down elevator at the moment of touchdown. This cancels out the nose-up effect of the upward force of the ground against the main wheels. In the case of a full stall, or "three point" landing, you can judge the flair so that the tailwheel touches just a hair before the mains.

BTW, it's worthwhile to be proficient at both three-point landings and "wheel" landings. The three-point landing gives you the minimum speed at touchdown, and the shortest landing roll, as well as making sure that as soon as the plane is on the ground, it's done flying. However, a three point landing means that you're taking the airplane right to the edge of a stall (and therefore very soggy control effectiveness and even a a potential loss of control) while still in the air, which in the wrong airplane or under the wrong circumstances could be a very bad idea. In large, heavy aircraft with lots of mass in the extremities (such as big, heavy engines mounted out on the wings) such as a Beech 18 or a DC-3, it's normal to do wheel landings in almost all cases. Airplanes that can get squirrely at stall, such as the Pitts Special and other aerobatic types (which must have sharp stall characteristics in order to do clean snap rolls) are also often best wheel landed with some airspeed and control. Airplanes that have problems with the fin and rudder getting blanked by the wings and fuselage when in three-point attitude also are often best "wheel" landed. Even in airplanes that are well behaved in three point landings might need to be wheel landed if it's a windy, gusty day and you need to carry some airspeed right through touchdown so you have the control authority to fight the turbulence.

It's common practice to design the gear geometry so that the wing is at or slightly above the stall angle of attack relative to the ground when the plane is sitting on all three wheels. It's pretty tough to bounce if the wing is stalled! However, if you have an airplane designed with this type of landing gear geometry but you don't flair it sufficiently to actually touch down in three-point attitude, the wing will not be stalled at the moment of touchdown and the plane could still bounce. Good design can't always make up for bad piloting!

Two other design-related factors can be important here. At the moment of touchdown, there is an upward force against the mains that tries to push the nose up, but there is also a friction force on the main wheels pulling aft on them, which tries to pull the nose down. With just the right combination of landing gear geometry, power setting, airspeed and pitch attitude, it's possible for the nose down effect of friction to almost exactly cancel out the nose up effect. With most of the tailwheel airplanes I've flown, I've been able to find a setup for final approach and flair that gives this result.

Tail design is a big factor in this as well. Some of you may remember some past discussions of the difference between static stability (the tendency of an airplane to try to return to its original attitude after it's been disturbed) and dynamic stability (the ability to damp out oscillations). If you missed that or need a refresher, you can read up on the subject in the "Ask Joe and Don" section of our website, there's quite a few articles there that discuss this subject. In the case of bouncing tendencies at touchdown, dynamic stability in pitch is particularly important since it resists the pitch-up and stops the bounce from getting started in the first place. Since dynamic stability is proportional to the tail area, and to the square of the tail moment arm, planes with large horizontal tails, and especially ones with very long tail moments will have strong resistance to bouncing.

For example, one of the full-scale WW II warbirds that's noted for being especially easy to land (contrary to what you might expect) is the Spitfire. Despite it's relatively small tail surfaces, it has an unusually long tail moment arm, which gives it extremely good dynamic stability. It also has the weight concentrated close to the C/G, with very little mass in the extremities, which also helps the dynamic stability. Another example is the Peck Polymers "Prairie Bird". This 42" R/C model is a scaled up version of a rubber free flight. It has the landing gear mounted very far ahead of the C/G, which should make it an incorrigible bouncing and ground-looping monster. However, it has enormous tail surfaces set on a long moment arm. This combination gives it outstanding dynamic stability and exemplary ground handling, among the best I've ever seen in both taildraggers and tricycle gear model aircraft! The gear is so far forward that it is nearly immune to nose-over, but that huge tail kills any thought of a bounce or ground-loop before it has a chance to get started.

So much for bouncing, now what about Rob's original question, ground looping? This one gets a little more complicated.

If the airplane is rolling along the ground and it is yawed to one side, the wheels generate a sideways force. Since the main wheels of a taildragger are ahead of the C/G, that sideways force tries to make the yaw worse. If left uncorrected, that increases the yaw, which increases the sideways force, which increases the yaw, which increases the sideways force, which increases the yaw, which increases the sideways force, which increases the yaw... you get the picture. If not caught soon enough, the yawing effect generated by the wheels can exceed the yaw authority of the rudder, and at that point the now helpless pilot is just along for the ride. The massive sideways force can also tip the airplane over onto the outside wingtip, and if the gear is tall enough the sideways force can also generate enough rolling friction to cause a nose-over.

The further ahead of the C/G the main wheels are located, the worse this ground-looping effect becomes. At this point we need to introduce another concept, the "vertical C/G location". We're all familiar with the idea of the fore-and-aft location of the C/G, and most of us are familiar with the idea of lateral balance, which simply means that the side-to-side location of the C/G is on the centerline of the aircraft. However, the C/G also has a vertical location. Since the wing is usually one of the biggest structural weights in the airplane, it tends to be fairly close to the wing. In our electric models the battery and motor locations are also big factors. In general, typical vertical C/G locations for low wing models tend to be around 1/3 of the fuselage height up from the bottom of the fuselage, and for high wing models it tends to be about 1/3 of the way down from the top of the fuselage. You can turn your model on its side and balance along the side of the fuselage to find the vertical C/G location, just like we balance along the underside of the wing to find the fore-and-aft C/G location.

Note that the main wheels' contact with the ground is well below the vertical C/G location. This means that when the plane is sitting in three-point attitude, the wheels are further ahead of the C/G than they are when the plane is in a fuselage level "wheel landing" attitude. This means that the plane's ground-looping tendencies will be worse when in three-point attitude. Add some blanking of the fin and rudder by the fuselage, wing and horizontal tail and things could get very dicey indeed.

The general rule of thumb for locating the main wheels is to draw the plane in fuselage-level attitude, plot the location of the C/G in both the fore-and-aft and vertical senses, then draw a line downward from the C/G and angled forward about 10 to 15 degrees forward of vertical, depending on how much nose-over resistance you want vs. how concerned you are with bouncing and ground-looping. The main wheels should be located somewhere along this line, and low enough to provide adequate prop clearance, including when the fuselage is maybe 10 degrees or so nose-down. The tail wheel or tail skid should be long enough to put the wing at the stall angle of attack relative to the ground when the plane is in three-point attitude. Note that this last requirement could also influence the incidence angle between the wing and the fuselage datum line.

So we should keep the airplane's fuselage level as much as possible when on the ground? Well, not necessarily. Whether the plane has a tailwheel or a tail skid comes into play at this point. A tail skid generates a friction force that's parallel to the tail's direction of motion, in other words pretty much straight aft. If the plane is yawed, this friction force is shifted to one side and does help try to stop the yaw, but this effect is usually not very strong. OTOH, a tailwheel can generate an very significant sideways force, just like the mains, and it usually acts through a much longer lever arm than the mains, so within limits it can definitely curb the yawing tendency of the mains. However, the tailwheel is generally much smaller than the mains, and has less traction than the mains, so there are limits to just how much it can do. Once again, if you let things get too far out of hand, you're just along for the ride after that.

The bottom line is that if you have a tailwheel, hold full up elevator when on the ground at low speeds (i.e.: speeds where there isn't enough airflow over the rudder for good rudder effectiveness) to keep as much load on the tailwheel to maximize its traction and steering authority. Once you have enough speed for good rudder authority, get the tail up and keep it up.

If you have a tail skid, try to either go as fast as possible with the tail up (to get the tail up out of the wing and fuselage blanking and get as much clean airflow over the rudder as possible), or keep the tail on the ground and the speed very low. The danger zone is when accelerating through the speeds in between these two extremes. On takeoff use as much power as you can handle without the torque and P-factor overcoming the rudder (for models that usually means giving it FULL throttle) to get as much speed as possible as quickly as possible. On landing it means keeping the tail up as long as possible with down elevator, then lowering the tail quickly when the speed has decayed to a slow walk. This is the technique required on our Roadkill Series Fokker Triplane, and I suspect it might apply to planes like the Bee as well. The rudder gets blanked by the rest of the airplane (including a windmilling prop if you pull the power all the way off), so if you don't have a tailwheel, start the takeoff run with full down elevator before you open the throttle, use full throttle, center the elevator as soon as the tail comes up. On landing, wheel it on, keep some power on to maintain prop blast over the rudder, and let it decelerate to a slow walk before lowering the tail. Slowly reduce power at that point until the plane stops, so you can maintain some prop blast over the tail as long as possible. I typically use about half throttle on the Triplane on final approach, leave it at half throttle till I get the tail down on landing roll (the Triplane has lots of drag from all those wings, so it likes lots of power), then reduce to 1/4 throttle and keep it there till the plane stops. I then taxi it very slowly.

The surface you're operating on is also a major factor. If the surface is relatively slippery, then the mains can't get enough traction to support a really good ground loop. When I was flying full scale J-3's, Taylorcrafts, Aeroncas, etc., I much preferred to land on grass runways if possible. The airplane that is a handful on a paved runway is often a total pussycat on a grass surface. For models, this also means that the tire tread can be a factor. A slick tire with no treads and a fairly hard rubber formulation that gets poor traction will tend to reduce the ground-looping tendency.

Another factor that often comes up in these discussions is wheel alignment, specifically the use of toe-in. By bending the ends of the axles forward so that the forward edges of the tires are slightly closer together than the aft edges, we create another correcting force. When a plane with toe-in starts to yaw into a ground loop, the wheel on the outside of the turn starts to scrub on the ground and create more friction, while the wheel on the inside of the turn rolls more easily. This difference does tend to fight the ground loop in its early stages. However, once the ground loop gets started and the plane's weight is shifted to the outside wheel, it can actually make the ground loop worse. Overall, though, since the best way to fight ground loops is to not let them get started in the first place, I've found that about 1 to 2 degrees of toe-in per wheel is usually beneficial. It's also usually a good idea to tilt the tops of the wheels outward by about the same amount. This reduces the scrubbing of the wheels a little when rolling straight ahead, and also acts like toe-in when the plane is in three-point attitude. Without it, you could end up with toe-out when the tailwheel is on the ground!

Toe-in is most effective on planes with a very wide gear stance, such as the P-51 Mustang. On something with tall, narrow gear like the Me109, it has less effect, although it's still probably worth doing, even if it does aggravate any tendencies to tip to one side and drag a wingtip. It also doesn't help that much on the Spitfire, but because of the Spitfire's very long tail moment arm and its high dynamic stability, it doesn't really need a lot of extra help from things like wheel alignment.

Another factor that can help in some airplanes is the use of a solid axle that turns with the wheels, so both main wheels are forced to turn at exactly the same speed. This fights any turning tendency as long as both wheels have some traction. However, once the plane starts to tilt a little from the centrifugal force of a developing ground loop, the inside wheel will lose traction and this beneficial effect will be lost. The design of your shock absorbing system could be important here; if the wheels are sprung so that they can both maintain firm contact with the ground even though the plane is starting to tilt a little, this could maximize the beneficial effects of a solid axle.

So Rob, in summary, try some toe-in, make sure the mains aren't too far forward, add a tailwheel (if it's steerable, make sure the linkage to it is sufficiently stiff, a soggy steering linkage can make a tail wheel act more like a tail skid), make sure the rudder is big enough, use full power on takeoff and get good rudder authority ASAP. A solid axle might help, but it's a long shot.

All of this may make tailwheel landing gears seem like some fearsome monster to be avoided at all costs. This is definitely not the case. Properly designed, a tailwheel landing gear can be better behaved than most tricycle-geared models, and is generally lighter, stronger, and has less drag, lower vibration and better prop efficiency. Besides, they look cool!

Don Stackhouse
DJ Aerotech