In my simulator (flightgear) directional control during take-off of most tail-draggers is extremely difficult. When the tail lifts off, it immediately starts to turn to the left with no crosswind, but with sufficiently strong crosswind from the right it might be to the right. It takes very swift and hard rudder application to arrest and then easing the pressure to avoid overcontrolling. Otherwise the turn rate keeps increasing and ground loop occurs.

It is worse with some models than others. Some models are almost impossible to take off at full or normal take-off power and while most are, it is very difficult to keep them on the runway, because even if I manage to arrest the initial swerve, the aircraft is then turned a bit and attempts to re-align it tend to again provoke ground loop.

Controlling propeller aircraft with tricycle landing gear is on the other hand never difficult, even for the more powerful ones.

I suppose some of the models are wrong. For example the Pitts S1C model is almost impossible to take off. But how difficult should it really be to keep the plane aligned?

And, what would likely be wrong with model that is too difficult to control? Centre of gravity too far aft of the mail landing gear? Or something with the vertical stabilizer? There does not appear to be any parameter for vorticity of the propeller slip stream; it is apparently calculated from the power.

Or is this one of the cases where simulator is much more difficult than the real thing, because in the real thing you have force feedback and get to use your built-in accelerometer (vestibular system) and in the simulator you don't?


6 Answers 6


Your simulator gives you a realistic experience: This is normal. What you experience is the gyro effect of the propeller and engine while you change the pitch attitude. When you lift the tail, the axis of rotation is changed and the engine-propeller combination pushed downwards, causing a precession force which acts orthogonally to the plane defined by the down movement and the axis of rotation.

Since the engine is forward of the center of gravity, this precession force is pushing the nose sideways: Left for a propeller which spins clockwise (when observed from the pilot's viewpoint), and vice versa.

The best way to avoid a change in direction is to keep the tail to the ground with the stick aft until you know you are fast enough to lift the tail, and when you lift the tail up, you apply rudder simultaneously. Don't let the tail come up by itself!

I learned this in a Dimona motorglider. Now think of flying a small, light airplane where the propeller is bolted firmly to the heavy engine crankcase, both rotating. The airplane will dive or climb when you change direction! This is what pilots had to endure in the First World War, when they flew aircraft with rotary engines. When engines became more powerful, this became so bad that some aircraft were almost unflyable. See this answer for more.

That crosswind is turning the nose into the wind after lifting the tail should also be no surprise: While the main gear is still on the ground, the plane will weathervane into the wind once the tail is lifted.

  • $\begingroup$ What about the tendency of the turn to develop into ground loop rather rapidly if it is not corrected very swiftly with rudder? The effect of course also exists. But how bad is it in real, properly balanced aircraft? $\endgroup$
    – Jan Hudec
    Commented Jul 25, 2014 at 4:48
  • 1
    $\begingroup$ @Jan Hudec: My experience is limited to planes with large moments of inertia around the vertical axis (big wingspan!) and small propellers, so they never showed more than maybe 10° deviation from the runway line. On smaller, more powerful engines the side force would be bigger and cause more directional change. And then you are moving sideways with the main gear down, which could very well result in a ground loop. $\endgroup$ Commented Jul 25, 2014 at 7:23
  • 3
    $\begingroup$ @JanHudec My first solo was in a tail-dragger and I did a ground loop on landing. The problem with that type was that the tailwheel was controlled by the rudder pedals for deflections of maximum 28 degrees (the same as the rudder). After that, the tailwheel disconnected from the controls and it behave like the wheels of a shopping cart, usually increasing its deflection. So, if the pilot didn't control the turn early on, the situation developed rapidly to the point where there was no control on the wheel and due to low airspeed, no control on the rudder. $\endgroup$
    – Emil
    Commented Jul 25, 2014 at 10:26
  • $\begingroup$ @Emil: On landing, the propeller spins less quickly, so precession should not be a factor. I am at a loss to explain what caused your ground loop. $\endgroup$ Commented Jul 25, 2014 at 10:54
  • 1
    $\begingroup$ @PeterKämpf: Yes, the deviation was not as important as during take-off, but it was visible. Especially after a touch-down with the tail quite high, when the tail started to drop, there was a significant deviation. If you add here poor visibility on ground and little flying experience, going 30 degrees off-track was a matter of seconds. $\endgroup$
    – Emil
    Commented Jul 25, 2014 at 11:06

What you're experiencing is the natural tendency of an aircraft to turn left under high power.
Several forces combine to cause this as shown in the diagram below:
Cause of left-turning tendencies
(These forces are also discussed in this YouTube video which is worth a watch.)

The left-turning tendencies are more pronounced in higher powered aircraft (more torque). They are even more pronounced in tailwheel aircraft when the tail comes up because you no longer have a third point of contact with the ground resisting the forces, and the pitch change exaggerates the effect of gyroscopic precession. (Similarly when the tail comes up in a strong crosswind the aircraft has an even higher tendency to weathervane into the wind, because there's no third point of contact with the ground fighting its tendency to pivot).

The position of the center of gravity relative to the gear also contributes to some of the differences in handling on the ground (and their tendency to ground-loop), but it's probably not the major factor in the difficulty you're describing.

In regard to your difficulty fighting the aircraft's left-turning tendency, part of this is undoubtedly the fact that a simulator has poor (zero) feedback compared to sitting in an actual aircraft, but some of it probably has to do with your technique: You're describing very swift and hard rudder application, and frankly no aircraft is going to like that (and tailwheel aircraft will make their displeasure known in the swiftest possible ground-loop).

What's called for instead are smooth control inputs - advance the throttle slowly, feeding in enough rudder input to track the centerline as the engine power increases. Remember that ultimately you're balancing forces that are changing: Slipstream and torque reaction are more pronounced as the engine develops more power, P-factor varies depending on the aircraft's pitch angle, and the effects of precession vary as you change the aircraft's pitch angle (applying forces to the rotating propeller).
While these forces are changing the force exerted by the rudder is also changing, becoming more effective as you accelerate.

Because of the way these factors combine the corrective forces you need to apply with the rudder aren't constant, and your feet should be "dancing" on the pedals making small corrections all the time as smoothly as you can (a little wiggle on the centerline is far preferable to stabbing at the pedals and making the aircraft pirouette on the runway).

Since you're using a simulator you can practice all of this in a no-wind situation to get a feel for the aircraft and the sort of control inputs that will be required (and ideally you can start with a trainer aircraft like a J-3 cub, and move up to something with a bigger engine as you get comfortable). You can also do high-speed taxi runs (with the tail up) to get a good feeling for how everything interacts, without having to worry about breaking a perfectly good aircraft if you ground-loop.

Disclaimer: I've got all of 1 hour in a tailwheel aircraft and the best I can say for my advice above is "by following that advice I didn't ground-loop the thing!"
You can probably get far better advice from a CFI with tailwheel experience, or a tailwheel pilot who's done it far more (and probably far better) than me!

  • $\begingroup$ You say apply rudder progressively as increasing power and while the tail wheel is still on the ground. Well, that might be the problem with the simulator. Because while the tail is on the ground, the simulated aircraft tends to go straight with very little effect of the rudder and then it starts turning very suddenly when the tail lifts off. $\endgroup$
    – Jan Hudec
    Commented Jul 25, 2014 at 5:09
  • $\begingroup$ @JanHudec The left-turning as the tail comes up is almost entirely gyroscopic precession like Peter mentioned (and maybe losing the tailwheel, if it's a steerable wheel) - it sounds like your simulator is doing the right thing. Basically as the tail comes up you'll need more rudder pressure, and when you stop changing the pitch attitude of the plane you'll need less - how much more/less depends on the prop size / speed which determines the gyroscopic rigidity the prop experiences. $\endgroup$
    – voretaq7
    Commented Jul 25, 2014 at 15:54
  • $\begingroup$ Thanks. The steerable wheel should be loosing grip progressively and therefore it's effect should cease progressively, but as long as gyroscopic precession is indeed the important effect here, it makes sense. And maybe suggests that applying some up elevator to make the tail come up more slowly should help. $\endgroup$
    – Jan Hudec
    Commented Jul 25, 2014 at 16:04

I have not flown that many taildraggers, but some can be kind of twitchy like you describe.

Like most things in flying the trick is to not get behind. If you anticipate needing some rudder as you feed in power get the control input in before the power. Also, remember as the aircraft picks up speed the control surfaces will be more effective, so you need to moderate your control inputs.

You also have to compensate for the descending prop blade causing the center of thrust to shift off center because of the high angle of attack of the prop blade as you start rolling. This commonly called p-factor.

Each plane, let alone model of plane, tends to be unique in it's handling characteristics. While I have never flown a Pitts I can imagine that since it has a lot power (big p-factor), short fuselage (short moment, lower longitudinal stability), and rather big control surfaces (small input, big effect) it would be a handful. Maybe you should try something easier for a bit like a Piper and get that down. Then work up to a stable plane like a Cessna 185, and a twitchier plane like a Citabria, and then onto the Pitts.

Also, I think the Pitts does not have flaps, so you will probably need to slip to a landing. That is really fun and it also provided good runway visibility since you are coming in at an acute angle to the runway.

Have fun,



I have a few hundred hours of taildrsgger time, both airplanes and gliders. Even though gliders don't have an engine they have to be kept flying straight especially when the tow rope hook is under the CG instead of the nose.

There are two things that matter, practically speaking:

  1. Don't let the tail wheel come up until the rudder is moving fast enough through the air to have full effectiveness. Many TWs have locking TWs to assist with keeping the airplane straight

  2. anticipate the yawning, be aggressive when correcting, and "uncorrect" to keep the nose straight. This means that when you bring up the tail, start feeding in rudder, faster than you need it, and before the nose is straightened out, start feeding in opposite rudder to prevent from overshooting. If you overshoot, you will start to oscillate with larger and larger movements, and you will lose control


I have 5,000+ hours of tail-wheel experience. So, heres my two cents:

I won't elaborate on the physics of what's happening since voretaq7 explained that topic pretty well (thankfully). I will concentrate on a more practical understanding.

How difficult is it to avoid ground loop in practice?

It's not difficult, you just keep the plane straight...at all times, no matter what it takes. I would say your simulator isn't very accurate in portraying real-life tail-dragger characteristics. I have never flown a Pitts, however of the other tail-draggers I have flown, they are not extremely difficult to keep straight, just attention.

what would likely be wrong with model that is too difficult to control?

One where the aircraft does not have enough rudder-authority to compensate for either:

  1. The current wind velocity.

  2. The amount of torque the engine produces.

  3. A combination of 1 and 2.

These characteristics seem to be more per-dominate by the design of an aircraft as well. (short-coupled and narrow-wheelbase aircraft being in that group). I do not know much about physics, so I can't elaborate on the why, other than I know from experience.

I find that once the tail is up and flying, it is EASIER to keep the airplane straight, you have more rudder authority (i.e. control) of the aircraft to main directional control, but that may be just me.

In short, flying a tail-dragger is not all that difficult, it sounds like your simulator does a poor-job giving tail-dragger flight characteristics. With that said, I also find simulators to feel very un-natural.

As far as flying a real tail-dragger. It takes a bit getting used, but it's like anything else. Once you get the hang of it, your good to go.

  • $\begingroup$ Thanks. I suspect a big problem with simulators is that they only give you visual information and no associated feel of motion and not correct feedback on control (unless you buy a special force-feedback joystick that I don't have). And even the visual is not really that good; when I learned how to handle skid in a car, I was told to keep my head aligned with the road facing external reference (and it did help a lot), which does not really work in simulator. $\endgroup$
    – Jan Hudec
    Commented Aug 1, 2014 at 12:03
  • $\begingroup$ You said "just keep the plane straight... at all times, no matter what it takes". However, some tail-wheel airplanes sit on the ground with such an extreme "nose-up" stance, the pilot can't see a damn thing in the forward direction. So how does a pilot keep the airplane pointing in the direction of the runway center line... when he can't see the runway? In some cases the pilot can't see anything at all forward, not even trees in the distance, so he can't replace a view of the center-line with a view of a tree straight ahead. Seems like looking sideways out the window wouldn't help much. $\endgroup$
    – bootstrap
    Commented Jan 27, 2019 at 21:11

How difficult is it to avoid ground loop in practice?

Not difficult. Practice makes perfect. That's why in real world type ratings exist, you get extra training to fly taildragger types.

Desktop PC simulators are a waste of time and not at all realistic, the "full-motion" ones airlines use are lightyears better... but obviously nothing beats the real thing !


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .