Ground looping: Why?

[lensrc]

Can someone tell me the root cause of ground looping? I have a micro that I cannot stop from spinning around when trying to take off.

[Ollie]

There are many possible causes that boil down to a lack of balance among the yaw forces and moments. Use of ailerons with adverse yaw can start a ground loop if not compensated by rudder input. Difference in rolling friction between the left and right wheels could be the cause. Again, steering with rudder while on the ground is the solution. It could be P-effect. Again the solution is steering with rudder to keep the heading straight. Taking off directly into the wind will make rudder steering a lot easier. Crosswind take off of planes with high wings or dihedral may require aileron compensation in addition to rudder compensation. Toe-in of the wheels may help.

[mode1]

Not much information to go on but I would venture a guess. High rudder throw/overcontroling? Main landing gear placement (too far forward/not far enough)? Do you have a few degrees of toe in on the mains? I use toe in and negative caster on all my taildraggers with good results.

[tim hooper]

I asked a similar question some time ago, which may have a bearing on the problem;

http://www.rcgroups.com/forums/showthread.php?s=&threadid=107282&highlight=torquesteer

tim:)

[lensrc]

This is an indoor micro that I'm having trouble with. I was hoping that someone knew about some common mistake I could have made that would cause groundlooping. Steering with rudder does not help. I am an acomplished R/C pilot,but have never had a plane with this problem this severe. It will not reach flight speed without swapping ends, no matter how I try to stop it. I will take a pic of it in a while and maybe that will give somebody a clue.

[lensrc]

Tim,after looking at the thread you posted, I dont believe it is the same problem. This plane will loop both directions at will. I may post this Question in indoor and micro,as I have seen a few micros do this,just not as bad.

[lensrc]

Pic one:
http://www.casstel.net/~lensrc/MicroBipe1.jpg

Pic two:
http://www.casstel.net/~lensrc/MicroBipe2.jpg

[Fly]

You can reduce ground looping by putting on a tail wheel (even if it is not steerable), and use lots of initial up elevator (down force on the wheel). You have to make the tail much more willing to go forward than to slide sideways. A dragging tail doesn't care which way it slides, and any sideways slide induces looping by steering the front in the opposite direction.

-- Alan

[Ollie]

The wheels are way too far forward. Move the wheels back to a point where they are just slightlly forward of the CG when the tail is up and the fuselage horizontal. Shorten the main landing gear to give minimum prop clearance to the ground in takeoff attitude. You might lengthen the tail skid or better a tail wheel back there to raise the tail to a more horizontal position. Because of the dihedral, lack of ailerons and high position of the top wing above the ground, you shouldn't try to take off in any sort of cross wind.

Hand launching is a good option.

[lensrc]

This thing wieghs 1.3 oz,wind is NOT an option! I was suspisious of the foward gear, I will see if I can move it.

[me11owman]

Ground loops are caused by the CG being too far behind the landing gear......so either move the gear back. or get faster on ther rudder/tailwheel steering..(thats what the classic planes of the 30's and 40's did...)ie pilot had to be quick on the rudder pedals in tail wheel a/c

Jim

[Gene Bond]

Yes, Len, That gear way up front makes it kind of like trying to hover a pusher plane :) It's a real balancing act, since there is little to keep the back end in check. I suspect why you did it that way, but you may have to give up some of the protection to rog. I have similar problems with a bipe, too far forward, and it loops, too far back, and it noses over :(

[lensrc]

Well,I can change the gear arangment now that I ripped it off in test flights!Wing too! I'm begining to think this micro stuff aint for me.

[Karl B²]

Yet more free advice (worth at least what ya paid for it! :) ):

You may be able to gain control by goosing the throttle quickly to half or more, giving a good propwash over the fin & rudder. This worked well for me when taking off in rough grass or chunky gravel roads with larger models where a slow increase wouldn't allow enough speed before one of the wheels caught on something. At your size, running into dust motes or joints in the tile may be enough to spin it around. :)

Agree that the wheels look far forward. Another rule of thumb is to put the axles at or just behind the leading edge.

Karl

[Tim Green]

Get your wheels behind the prop. Make sure there's toe-in and that they don't bind. Then ROG. :)

[Peter D Rieden]

Assuming that everything is straight, square and true, a tendency to groundloop is caused by excessive longditudenal distance between the centre of gravity and the centre of friction for the undercarriage - ie the wheels are too far in front of (or behind) the CG.

The cures for you problem would be too move the wheels closer to the CG and if the prblem persists try some toe-OUT on the wheels. The reasons are discussed bellow - apologies for the length - I normally explain this sort of thing with diagrams on a whiteboard or similar.

The long distance between the CG and undercarriage centre of friction means that as the model yaws slightly the latteral (spanwise) distance betwen the centre of friction and the centre of gravity gets bigger. This produces a rotating couple as the aircraft inertia acts through the CG but is primarily reacted by the dominant force (undercarriage friction) acting to one side of it. The greater the yaw the bigger the destabilising force, so it "runs away" and around we go. The cure here is to have the wheels as close as possible to the CG. Note that this applies just as much to trike undercarriages as it does to tail-draggers - a trike with the mainwheels too far aft will also be prone to groundlooping.

A further force is significant in a tractor aeroplane; the propeller itself is destabilising in both pitch and yaw. This is because, if the aeroplane is pointing in any direction other than its "direction of motion" the propeller will be taking air in at an angle but accelerating it straight backwards, changing the momentum of the air through an angle. In the process the nose will experience an equal and opposite change of mementum which manifests iteself as a turning force that tends to turn the aeroplane further and thus just gets worse in another "runaway". The cure here is to either move the prop behind the CG(!) or to be quicker on the rudder.

A final point - various references have been made in this thread to using toe-IN to alleviate this problem. This is a common misconception and I'm afraid it's bunkum. On a tail-dragger adding toe-in will destabilise whilst adding toe-OUT will tend to stabilise. The opposite will be the case with trikes. It's to do with roll-centres, weight transfer and rolling friction (as a vector rather than scalar phenomenon) but it would need a bigger whiteboard than we have here!

Hope that wasn't too boring!

Peter D Rieden

[Karl B²]

Diagrams are from p. 120 (top) and p. 118 (bottom), "R/C Model Airplane Design" by Andy Lennon. I usually adjust my main gear caster so that the toe-in relative to the ground remains the same whether the tail is on the ground or at take-off attitude.

Peter, the drag vectors shown make sense to me, and they support my real world experience of an increasing tendancy to groundloop as my toe-in gradually becomes toe-out with rough landings. If this diagram is flawed or there are additional contradicting factors, please explain.

Karl

[jrb]

All except the T-33 are tail draggers; yes the A-10 is as well!

The H-9 Cub ground loops well; gyro helps immensely, though I’m headed cross wind shortly after take-off.

I’ve ground looped the American Flyer a couple of time when the tail wheel some attention.

Never have with the Mustang, FreeStyle, or A-10. See a Mustang take-off here: http://rcgroups.com/gallery/showgallery.php?cat=998&stype=2&si=jrb

The Cub has the wheels closest to the CG, the American Flyer next.

The FreeStyle is furthest away and tracks the best! It also sits most level, the Cub most tilted.

All things being equal, based on my experience I’d put the wheels further forward and have a plane sit nearly level.

[Peter D Rieden]

*sigh*

This is a very well presented example of the common misconception! The illustration serves well in illustrating the cause of the problem, but when it comes to the *solution* it ignores the fact that wheels don't generate "drag"; they generate "rolling resistance" and "turning force" (the latter only being present when the axis of rotation isn't at right-angles to the direction of travel). It compounds this by assuming that the forces on the tyres act through the CG - which they only do in plan view.

I've dashed off a quick diagram in the only editor I have to hand (which isn't brilliant at this sort of thing, but it's that or nothing).



Consider a taildragger aeroplane moving along the ground with weight on the wheels. Now assume the pilot applies full left rudder - what happens? In plan view the tail is pushed to the right, presenting the mainwheels at an angle to the direction of travel such that the tyres generate a turning force to the left. The force on the rudder and the force on the tyres form a torque couple which rotates the aeroplane.

But FRONT elevation (see diagram) we see that the momentum of the aircraft (which acts through the CG) and the turning force on the tyres (which acts at the runway surface) are vertically displaced, so there's a second torque couple which tries to roll the aircraft outwards, increasing the "weight" on the outboard wheel. If the wheels are aligned with zero toe in/out this doesn't affect much, but if the wheels ore toed inwards the increased loading increases the inward-turning force because the outboard wheel has greater force than the (lightly loaded) inboard one. But this increase in turning force increases the weight on the outboard heel, which further increases the turning force and so on. The situation is unstable because it is a positive-feedback system.

On the other hand if the wheels are toed OUT then the increased weight on the outboard wheel REDUCES the turning force, easing the tendency to swing. This is more stable because it is a NEGATIVE feedback system - a disturbance from the mean results in a force which tends to oppose that disturbance.

If the wheels are toed out sufficiently that they track through the CG there will be no net turning force no matter how much rudder is applied, but then you wouldn't be able to turn off the runway!

I hope this makes it clear, but if it doesn't please ask further questions because this is one of those bits of model aeroplane folklore which is repeated so often that many take it as gospel. It's embarassing! Other examples are the "high drag of high pressure air" explanation of adverse yaw or the "shorter wing" explanation of dihederal as roll stabiliser.

Hope it's of interest,

Peter D Riedenh:\toe out.jpg

[jrb]

As mentioned above my A-10 is a tail dragger. Thrust is applied well aft of the CG while the mains are forward (nearly equally as far away) of the CG.

Probably not the best case; most likely worse case, like a wheel borrow?

The A-10 has never ground looped!

[Peter D Rieden]

The A-10 is probably less prone to ground-looping because there is very little vertical seperation between the ground and the CG.
Your comment about tail-down ground angles is relevant for tendencies to *swing* rather than ground-looping per se.

Aircraft with steep tail-down angles and large amounts of power exhibit a powerful swing when the pilot lifts the tail - something which he has a strong urge to do as early as possible in order to see where he's going. The sudden change of angle rotates the prop in the pitch plane which in turn produces a powerful gyroscopic precession force about the yaw axis, and it does it at a time when the airspeed is low and the rudder authority is minimal!

Good examples of this include the "overpowered" fighters from the latter part of WW2; Bearcat, Typhoon & Tempest for instance. The Typhoon was reputed to have a vicious swing that was barely controllable until a ferry pilot tried a "dumb" take off - he simply opened up to full power and left the elevator alone. The aircraft tracked straight as a die and the tail came up smoothly in its own time with no associated swing...

More than you wanted to know, I'm sure!

Peter

[jrb]

Peter,

I agree it’s the plane’s reaction to the gyroscopic force produced by the propeller that’s causing the severe ground looping.

The old high school physics experiment proves this; the one where you sit on a stool that is free to turn left and right. You then hold a bicycle wheel out in front of yourself between both hands. Spin up the wheel, and as you change the angle of the axis of rotation your body reacts to the gyroscopic force by turning on the stool.

In our case, just lifting the tail a large amount, like with a plane that sits at quite an angle, provide a lot of gyroscopic force; hence a tail loop.

As an experiment, change the prop to a pure disk (no air flow issues), power up, and lift the tail. While the tail is being lifted the gyroscopic force produced by the spinning disk will cause the to rotate about the yaw axis.

So there’s no thrashing of the air against the rudder, no change in AoA of the up swinging versus down swinging prop blade, etc.; bye-bye mumbo jumbo, its just gyroscopic force.

Wonder if my gyro would actually be more effective rotated 90 degree; working in the pitch axis versus the yaw???

Also, don’t think you need to be way over powered either.

Turning a large diameter prop, even at moderate power, will cause this effect a great magnitude.

So, a large prop, which normally means a lot of angle, acerbates the issue substantially; which is probably the root of the problem with model of the original poster. Raise the tail with a long tail wheel and there “wheel” be less ground looping.

Back to the H9 Cub, most are over powered; mine as well (happily flies @ 1/3 throttle) and I’m turning a 16” prop, glow is 10”.

Jim

[Majortomski]

Peter, your argument is interesting, but in 26 years of full-scale aviation, including an engineering degree, mechanics license, and pilots license, I've NEVER seen TOE OUT recommended on any aircraft in any configuration. Because it's the easiest way to START a ground loop. Mr. Lennon (a retired Boeing engineer) drawing illustrates how toe in negates the forces that tend to start a ground loop. These forces are predominant at the start of the problem. The forces you identify start to take over after the looping has started. Yet I still fail to see how toe out is going to affect the situation you present.

[Peter D Rieden]

Tom said:

<Peter, your argument is interesting, but in 26 years of full-scale aviation, including an engineering degree, mechanics license, and pilots license, I've NEVER seen TOE OUT recommended on any aircraft in any configuration.>

Well my PPL lapsed soon after our first daughter was born and I've only been an aeronautical engineer for 16 years, so you're ten ahead of me, but as a fair amount of that has been in stability & control roles I'm confident in my analysis. As I said at the start, I've had this conversation before, and it's best discussed with the aid of a whiteboard, several coloured pens and a large crate of beer [real beer, not that 'orrible american stuff :) ]

We have to keep it simple in this forum - debates about s-planes and the root loci tending towards real roots wouldn't really help much around here, so we'll keep it descriptive rather than analytical.

<Because it's the easiest way to START a ground loop.>

Would you like to expand on that? I would suggest that a taildragger aircraft is directionally unstable in the same sense as a broom handle balanced on the palm of the hand is unstable. The CG is displaced from the point of action of the supporting force and so any disturbance results in forces tending to move the CG further in the direction of the disturbance. Tail dragger aeroplanes have damping (the fin) and an active stability system (the pilot controlling the rudder), but the system is still directionally unstable on the ground. When the system is in equilibrium (ie when the velocity vector passes through both the CG and the centre of U/C forces) the presence of toe in *either* direction has no effect (other than to increase rolling resistance), so I can't see how toe-out could START a ground loop.

Having been involved in a number of light-aircraft projects, mostly assisting in debugging homebuilts in my spare time, I know that quite a few of the lower-powered taildraggers often have toe in, but this isn't to promote stability. It's done to increase the rudder authority at the start of the take-off run so that the *pilot* can use the rudder to kill a starting swing. It's done in the knowledge that the aircraft will be LESS stable on the landing roll and it's a trade-off. I've also had three Pitts models (a .40 size S2 and quarter scale S1 and S2) each of which was a handfull to keep straight on take off and each of which was tamed by the addition of toe-out to the mainwheels.

<Mr. Lennon (a retired Boeing engineer) drawing illustrates how toe in negates the forces that tend to start a ground loop.>

Well as I said before, the analysis presented in the diagram is two-dimensional (ie it ignores the roll-centre and weight-transfer issues) and it also seems to assume that the wheels are skids which have the same friction at all angles and generate no turning forces when rolling at an angle of attack to the velocity vector. A brief cross reference to car steering geometry and mechanics of tyre friction will soon illustrate how misleading these simplifications are. The diagram somehow suggests that the only force acting on the tyre is "drag"; that the tyre is locked stationary and is simply dragging along the ground. In reality the wheel is still turning at an angle of attack to the velocity vector of the runway and it experiences rolling resistance parallel to the velocity vector plus a lateral force parallel to the axis of rotation of the wheel - this is actually what makes the aircraft turn.

Remember that we're discussing a GROUND VEHICLE at this point, and so we need to consider how tyre forces act. If you go and have a chat with a circuit-racing car team they'll soon tell you (and demonstrate, if you're persuasive enough) that toe-in on the front wheels makes the steering much more lively, but at the expense of straight-line stability. Take a circuit car and drive it on a long motorway (freeway) and you'll find it very hard work because the car won't settle into a straight-line run. The reason is that it has a tad more toe-in than the road-going version of the same car to liven-up the cornering response.

Finally, and at the risk of labouring a point, please remember that this is the situation for taildraggers. If the mainwheels are *behind* the CG then the opposite is true and it's toe-out that destabilises while toe-in tends to stabilise.

Tell you what - next time you're in the Surrey/Hampshire (that's OLD Hampshire!) area I'll provide the room, whiteboard and pens and you bring the beer!

Regards,

Peter D Rieden

[DeaninMilwaukee]

Interesting discussion!

I had an relevant experience to share here recently.


A buddy of mine made a roughly 40 span cub style plane, ( tail dragger) that he had added landing gear to to give ROG a try.

He had a steerable tailwheel, but was having an awful time going straight at anything close to take off speed.

I had a look at it, and first noticed the LG was way too far forward. That corrected, the plane was only slightly better to take off, a real bear actually.

I recalled the toe in solution, and we tried varying amounts with no improvement at all.

As a last resort, we tried toe out, and it was like a miracle, the plane was able to take off close to normally. Success!


My take on this is that equally toed out wheels will try to " steer" the plane to their own sides, but normally cancel each other out, just like toe in would.

When a ground loops starts, the wheel on the " inside " of the groundloop lifts and loses traction while the outside one gets more weight on it.

This action causes the outside wheel, which being toed out, to actively " steer " the plane away, out of the groundloop.

As soon as the plane straightens out and rudder is released, the wheels have equal weight on them again and the effect is again canceled out. In summary, its sort of a roll controlled self steering action.


Come spring, I'm going to try adding toe out to my 1/12 scale fock wulf ta 152. This plane sits very nose high with its 11" prop, and the ground loop tendancy is vicious untill the planes going fast enough to lift it tail on its own.

Hopefully it'll show the same kind of improvement that my friends cub did.


Dean in Milwaukee

[DeaninMilwaukee]

Heres a pic showing how nose high the ta 152 is!

Excellent flying plaen BTW.

[Tim Green]

Pete,

If the main wheels are *behind* the CG, won't the plane just tip over on it's nose? Or am I not understanding the asterisks?

How about some beer?

[Peter D Rieden]

Pete,

<If the main wheels are *behind* the CG, won't the plane just tip over on it's nose? Or am I not understanding the asterisks?>

Well I was refering to other undercarriage configurations which place the mainwheel behiond the CG (eg a trike layout). In general the best results are achieved by placing the wheels as close as possible to the cg (whether in front or behind) commensurate with the other requirements. The "other requirements" include putting sufficient weight on the nosewheel or tailwheel to give adequate steering authority in crosswinds and (in the case of a taildragger) enough tail-down moment to prevent nosing over.

Putting the mainwheels of a trike too far back will be just as destabilising, like the Ripmax Trainer of the mid 80s whose mainwheels were under the wing training edge. This aeroplane had abysmal ground handling - it would groundloop and was almost uncontrollable on the ground unless there was a strong wind to weathercock into. It would also "bunny hop" on landing and took a great deal of skill to land well. All of this was completely cured by moving the mainwheels forward such that the axles were half an inch behind the CG.

<How about some beer?>

Well if you're in the area drop me a line (peter.rieden@baesystems.com), but in the mean time have this one on me!

Regards,

Peter

[Tim Green]

Thanks.

[Karl B²]

I did a Google search on "taildragger" and "toe-in" a few days ago, then promptly forgot about the whole thing. (Growing old sucks, except for having more money :) )

I've never seen large discussions on toe-in vs. toe-out on this board, but it appears that it's a frequent topic of debate elsewhere, especially in the full-scale world (enter the search keywords and see for yourself.) The majority of opinions expressed were either zero or slight (fractions of a degree) toe-out, though some called for slight toe-in. The toe variation I've seen on models due to sloppy building and rough landings (+/- 5° or more) would be intolerable in a full-scale for sure.

Just as the top view diagram looks only at tire scrub in a two dimensional model, so does the front view showing a lateral tire force resulting in an outward roll. Much of what I read lends as much or more weight to such variables as gear fore-aft position, spread/length ratio, gear flex (both up-down and fore-aft), tail-angle, vertical CG, etc. as is given simple toe adjustments.

The point was made also about trade-offs such as quick or slow commanded turn response at low speed or stability at higher speeds. I can't claim multiple years experience as either a schooled engineer or pilot (ie. zero time on both). The taildragger models I've built & flown (& designed) all tend to have springy, low, wide gear mounted rather close to the CG (usually less than Lennon's guideline), and that's just my aesthetic preference. Slight toe-in and unloaded positive camber (enough to zero out under load) have always helped eliminate groundloop tendancies on taildown acceleration, which is where I like help. Without video analysis, I can't say for sure that my toe-in does or does not become neutral under acceleration. I know that wire gear on rough ground will dribble fore-aft more than a wheel diameter's range.

Again, most of the theory I encountered in my search involved full-scale aircraft carrying stiffer gear and more live humans than my models. :)

End point: I'd try both toe-out and -in before giving up to handlaunching.

[rorywquin]

Originally posted by Kinnic^2
I did a Google search on "taildragger" and "toe-in" a few days ago, then promptly forgot about the whole thing. (Growing old sucks, except for having more money :) )SNIP.

You don't look that old in your picture ;)

[Karl B²]

ahh, the wonders of PhotoChop. :) It's how old ya feel, and I often feel older than I might look.