DRIFTING LEFT ON TAKE OFF - OB UPDATE 6

[Chipwich]

@Stacker +1

 

You have to jump on the pedals early. I tested at various T/O weights and wind speeds and directions. If you have 4 m/s and above you really have to stay on the rudder. No crosswind, the aircraft tracks straight. Differential braking helps. The NWS becomes ineffective above around 40 KIAS, but until that point it is helpful, regardless as to what I've read about disabling it once the aircraft is lined up on the runway.

 

So, what works for me is to enable NWS and be prepared to input rudder (and sometime aileron in direction of crosswind) up to a full deflection to stay aligned with the runway. You can't just sit there as a casual observer. I've not flown a Mirage IRL, so not sure about the vertical stab's effectiveness. But all of the general aviation aircraft that I have flown require you to steering them on takeoff. You might have 5 KIAS right on your nose, but even if you take away P-Factor and torque associated with propeller driven aircraft, there are always other factors. For instance, rarely are runways flat.

 

In other words, fly the airplane.

[Wayc00lio]

I've noticed that on TO, the nose pitches up significantly. I pull up to a nose up attitude of 5 degrees until lift off and the FPM pitches up to about 10 degrees very quickly.

[diditopgun]

About drifting (when a bit of cross wind for exemple), I notice that more your payload is heavy more you drift and need to apply rudder.

With a 4 m/s cross wind and a payload of 2x GBU-12 / 2x Magic / 2x fuel tank, I need to apply full rudder to stay on the runway... but I notice too that applying roll with the stick help and work as a second rudder.:joystick:

 

I don't know what is wrong actually... rudder authority ? fuselage wind sensitivity ? plane inertia regarding payload ? Tires ground friction ?

A mix of all ? :dunno:

 

(1.5.3 Open beta)

Edited May 23, 2016 by diditopgun

[diditopgun]

New flight in multiplayer (the host and 1 client).

 

Wind 3 m/s, payload: 2x fuel tank / 4x MK-82S / 2x Magic

 

We tried two time to take off and we crashed two times going outside of the runway... We need to set wind at 0 m/s to take off straight without any problem.

 

It confirm that more your plane is heavy, more the correction on the rudder to counter the wind need to be wide and so it amplify the oscillating path to an unrecoverable point where rudder become ineffective.

 

To illustrate (great drawing isn't it ?! :D):

1. No wind

2. Wind + light payload

3. Wind + heavy payload

 

I attach two tracks (each failed take off) and the mission file.

 

(1.5.3 Open beta)

server-20160522-222944.trk

server-20160522-223730.trk

Training M2000 MP3.miz

Edited May 23, 2016 by diditopgun

[Gliptal]

I really feel there's something wrong here. I understand the concept of PIO, but there's way too much lag behind the rudder on the ground, and I also think there's something amiss with the tires' friction (which would explain why hitting the brakes on landing kicks you on one side of the runway).

[diditopgun]

Thank you very much to the team to listen to us and to be so reactive ! :worthy:

[Kayos]

Thanks CptSmiley!

[gospadin]

But in any event, I was watching him try to takeoff and I see exactly the issue all of you are talking about. Us on the team must be conditioned to counter for it or something but it was enlightening watching him try to takeoff.

 

So today I did a component analysis breaking down every force and moment to build up exactly what is happening. Basically we found a couple issues with the gear collision model complicating the ground reaction behavior as well as some issue with the gear physics configuration.

 

I don't have it solved completely but it's on the right track and I'm excited to get it fixed.

 

Awesome, I'm glad you're on the right track.

 

Fundamentally, it breaks down to this (from what I can tell):

 

1) accelerate for takeoff (full AB)

2) apply light rudder with pedals

3) plane yaws (~1-2 degrees/second was my goal)

4) center rudder by centering pedals

 

Once #4 occurs and the rudder is centered, the yaw of the airplane should cease. Any yaw moment built up from #3 should be almost instantly halted by the straight rudder. The tail on this plane is HUGE.

 

Instead, I see that even after the rudder is centered, the plane continues to yaw for several seconds, causing swerving to start. I can confirm rudder is centered by watching the animation of the rudder.

 

Anyway, i'm psyched you could reproduce it. Thanks for the module!

[T_A]

Thanks for looking at the issue , i also started experiencing hard takeoffs and having to fight with the rudder, i haven't had that much 'fun' since i tried to learn the Dora (which i failed miserably since i couldn't take off :( )

[Whisper]

On top of crosswind making the issue really bigger, I think cold / icy conditions also make it impossible. The planes is gliding like crazy as soon as you look at the brake lever. Anyone seen this too?

[Gliptal]

I cannot stop on landing without the plane drifting off on one side, at a rate impossible to fix with rudder or MWS.

[Hydrox]

The ones who have had plane drifting in landing or takeoff, in that particular mission/flight, did you have any fuel tanks at any time?

[Gliptal]

Central fuel tank, with two Magic IIs.

[Hydrox]

If you check my bug thread, there's a much wrong with having fuel tanks, especially wing tanks, or full fuel load, I can't takeoff properly because of drifting, and I don't even try anymore, cause I know it's not right - don't wanna bang my head on the wall anymore.

I don't know if it helps, but maybe you could try a flight without fuel tanks and see if it still does that.

[Voodoo109]

I was having the same issue with the wing mounted fuel tanks. Its is nearly impossible to get off the ground. With just the center tank the drifting is still there, but it is much more manageable. It may just be something with the weight of the plane making the drifting issue more and more prominent as the weight goes up.

[Whisper]

While issue is looked at, a workaround is to lower the weight on the nosewheel by applying slight stick backpressure very soon during take off.

[Jeepyb]

New flight in multiplayer (the host and 1 client).

 

Wind 3 m/s, payload: 2x fuel tank / 4x MK-82S / 2x Magic

 

We tried two time to take off and we crashed two times going outside of the runway... We need to set wind at 0 m/s to take off straight without any problem.

 

It confirm that more your plane is heavy, more the correction on the rudder to counter the wind need to be wide and so it amplify the oscillating path to an unrecoverable point where rudder become ineffective.

 

To illustrate (great drawing isn't it ?! :D):

1. No wind

2. Wind + light payload

3. Wind + heavy payload

 

I attach two tracks (each failed take off) and the mission file.

 

(1.5.3 Open beta)

 

Just experienced the issue with the 3 fuel tanks loaded. Impossible to get airborn; aircraft pull HARD to the right, try to compensate and ground loop to the left every time. I did a few test after, and I feel like people saying we should ''fly the airplane'' or ''compensate for it'' or that it is just ''the crosswind'' haven't tried taking off with a full payload of 3 fuel tanks. The weird tendency to drift on takeoff is there on lighter payloads, but manageable to an extent. With 3 fuel tanks, the problem is on a whole other level.

 

I have done tests with different payloads after, and I have to say this drawing is right on. Wind or no wind, good luck trying to take off with 3 tanks. Like it was mentioned, your only chance it to put your stick full back for most of your takeoff run and correct the slight tendency to drift off that still can be felt by moving it left or right. Any correction with the rudder pedals will likely result in you losing control.

Edited May 31, 2016 by Jeepyb

[Voodoo109]

Started using toe brakes with my rudder pedals, and got in the air easily. Much better than trying to counter with rudder. Also pulled back on the stick to get the nose up.

[The Legman]

It gets much worse when the central tank is attached and even worse when the under wing tanks are as well. Something weird with the drag?

 

I don't have rudder pedals at the moment I'm using an axis on my hotas and is really hard to take off sometimes.

[gospadin]

I made a quick video to demonstrate the issue:

 

At around 0:14, I apply left rudder. I then re-center the pedals about 1 second later.

 

The plane continues turning sharper and sharper until it crashes, even though the rudder forces should be zero.

 

Airport is Min-Vody, DCS:W stable 1.5.353279, 3 fuel tanks.

 

 

[ame]

[/ame]

 

 

Edit: My (latest) hypothesis is that the thrust force in the flight model is actually pushing on the model behind the rudder, perhaps even at the nozzle, making tail & rudder completely ineffective. In fact, if it's pushing behind the rudder, then any yaw would become magnified, like throwing a paper airplane backwards. I'd expect the engine to be effectively pushing through the airplane's CG, which is well in front of the rudder. The tail & rudder should act like a weathervane and help keep the plane pointing in the direction of travel. The more the plane yaws, the more the airflow deviation from fuselage alignment should force the rudder (and thus the fuselage) back in-line with the movement vector.

 

I found I could simulate this centering force by using the wheel brakes, which act as a pair of drags to balance any asymmetric yaw.

 

The below video is taken with the same conditions above, and you'll notice i'm applying left rudder during the entire takeoff roll. Holding the plane straight with enough additional right brake to overcome the rudder, but a bit of left brake to provide some balance, was relatively easy (with occasional corrections). In all my tests with only a single brake, I crashed very quickly, but by adding the drag/resist force on both sides, takeoff wasn't too hard:

 

[ame]

[/ame]

 

 

Edit #2: If I enable airbrakes during takeoff, the plane has even less yaw stability than without. This would imply that the airbrakes (in the middle of the wing) are similarly pushing backwards against airflow on a point significantly ahead of where the thrust is pushing on the plane, thus amplifying any yaw asymmetry when the nose and the travel vector don't line up.

Edited June 2, 2016 by gospadin

[rrohde]

I also can confirm the issues that people are experiencing here. Esp. a moment ago when landing with a central fuel tank, ending up with the same A/C behavior as the #3 drawing depicted in the drawing above (quite accurately drawn, btw). It really seems to go awry really quick when braking and using rudders... as if these forces are compounding each other, rather than bringing the plane under control.

 

In stark contrast, the braking / using rudder behavior of ED's Su-27 is far more predictable and should serve as a prime example for ground handling during take off and landings - imho. :)

Edited June 2, 2016 by rrohde

[Gliptal]

The thrust point was moved from the CG to the nozzle some time ago.

 

Regarding landings, I found it helped immensely to tap the brakes rather than hold them.

[gospadin]

The thrust point was moved from the CG to the nozzle some time ago.

 

If that's the case, it would explain the behavior on the ground when we have a non-zero yaw deviation from the direction of travel.

[silentconvo]

If you apply some early pitch up with your takeoff it will dampen the drift people are getting. It will work even with a pretty decent crosswind. I have found the same for landings, the longer you can keep your nosewheel off the ground the easier it will be.

[Manuel_108]

Shouldn't you pitch down on initial take off roll, generally speaking?