swirling oscillation at low speeds P-51 1.2.0

PeterP · July 19, 2012

I have the feeling that something isn't right with the FM of the P-51...

It's no "big" issue for me , but I think it is worth mentioning.

(but sure can be a big problem for aerobatic teams)

This happens to me very often in low speed after take-off and while landing since the first beta - and I can still reproduce this in 1.2.0 - so I thought I open a thread about it.

I think it's a flaw of the flight model .

After I counteracted the Propeller Torque and established a "somewhat" stable flightpath - the P-51 starts to move the nose in a spiral.

Is this normal ? or a limitation of the flight-model ?

If normal: please explain me the physics behind this.:)

(please ignore the spiking throttle.. Its not causing this. - I will replace the poti this weekend )

Here are the settings to reproduce it:

(It happens also at different settings - but I found out that this works best to reproduce it 100%)

A Track is attached.

The swirling oscillation starts at 1:30

Edited July 20, 2012 by PeterP

PeterP · July 23, 2012

Shameless Bump!

Am I really the only one that experience this in 1.2.0 ?!

I'm asking because this post indicates that this was corrected before the 1.2.0 release :

>>> http://forums.eagle.ru/showthread.php?p=1458794#post1458794

post date : 05-15-2012

This bug only reproduces in the open beta you have. It was OK in previous beta versions and has been already corrected for open beta.

Edited July 23, 2012 by PeterP

HiJack · July 23, 2012

Checked your eyes latly Peter? :P

OT: I will check if i also can spot this today. I have laced my eyes :)

PeterP · July 23, 2012

Checked your eyes latly Peter? :P

Sorry HiJack - but I don't get what to do with your reply at all .

Serious - would you be so kind to elaborate a little more - as I really don't understand what this have to do with the topic.

There can also be something lost in translation on my end .

HiJack · July 23, 2012

Oh my are we on the wrong side here.

Nate--IRL-- · July 23, 2012

No idea of the true cause but the fact that you are riding the edge of a stall might have a very big part to play.

Nate

MTFDarkEagle · July 23, 2012

No idea of the true cause but the fact that you are riding the edge of a stall might have a very big part to play.

Nate

This I think...

HiJack · July 23, 2012

Well, that was easy to spot and answer. It's the new "shaking" of engine and cockpit that is buildt in to the SIM now. It was not there earlier.

(HJ)

PeterP · July 23, 2012

Oh my are we on the wrong side here.

Sorry ! Here , just for you >> :):):):):)

are this enough smilies to realize that I'm not offended?!

I was just asking for a better explanation- nothing more- nothing less.

If this was a joke : I didn't get it .

Please explain it - so I can laugh also! and again >>>:)

(Gosh! do we really need always emotion icons in all postings ?!)

Edit:

I have checked my eyes lately (I made a complete health check-up with reaching 35) - still 180% ! :) ... so 80% more than what is defined as optimal among doctors. :)

On topic:

No idea of the true cause but the fact that you are riding the edge of a stall might have a very big part to play.

Nate

Thanks Nate,

Yes I know! :)

But even if I'm at the edge of a stall...

There is no Physical law (at last in my universe) that explains this behaviour.

All I want to know :

Is this a limitation of the AFM ?

Further:

Will we have to live with it?

or

can we expect a fix ?

Edit:

I know: This last question is hard to answer for you - and you will probably ride the devil if you do :P

As I said: I can live with it - even if it is very annoying ...

(I can quite imagine how difficult it is to simulate the "edges" of a non-linear flight-model on a PC while still having a stable FPS...)

But some aerobatic teams will maybe have a hard time to fly some "advanced" manoeuvres - Like a high AoA flyby...

:)...Just wanted to report a BUG and getting some feedback if it really is bug.

Edited July 23, 2012 by PeterP

Vibora · July 23, 2012

Hello PeterP.

It has been reported internally.

MTFDarkEagle · July 23, 2012

Hello PeterP.
It has been reported internally.

Cool.

PeterP · July 23, 2012

Thanks Vibora!

Let's hope that Yo-Yo will redefine the receipt a little by using the right blend of spices... - and also having a lucky hand by using his wooden-spoon in his SAFM*-kitchen.. :P :)

*=Super-Advanced-Flight-Model

To be more clear what I want to say with this "SAFM-kitchen" example :

I'm in awe how well the FM is working up to now !

-point-

But If you really want to know how far it can get : you start to "surf" on the edges .

-that's what I'm doing most of the time right now in the P-51D...

and I discovered this "empty space/gap"

right between a stall <-> and just before falling from the sky in a spin.

(I hope you get my point)

It would just be very nice if this "empty gap" can be filled.

And again: I have a little knowledge how hard it can be to in-cooperate a non-linear model into a linear one.:) - so I fully understand why this "empty gap" is existent.

-- let's not talk about the processing power that is needed for something like this.

Edited July 23, 2012 by PeterP

Yo-Yo · July 23, 2012

Thanks Nate,
Yes I know! :)
But even if I'm at the edge of a stall...
There is no Physical law (at least in this universe) that explains this behaviour.

.

What makes you so confident? The prop at AoA produces two extra moments: the first one is a moment of extra transversal force applied as a prop is a wing at the same AoA, i.e. up. THe second is a pure moment because of thrust moment decentration. It is known as a P-factor. At high positive AoA P-51 prop generates left yaw moment. And this was the case then no AoS presents.

If we let the plane yaw left freezing the controls, the prop then is at AoS and AoA in the same time or we can say that its "local vertical" TURNED counter clockwise. What P-factor does the prop generate now? Yes, nose down and still left yaw. As we have now nose down pitching moment and left yaw moment the plane will increase yaw and decrease pitch that causes in its turn further rotation of the "local vertical".

It's a kind of precession but caused not by gyroscopic effect but the effect of non-axial prop inflow.

Generally it's because the P-factor always rotate the plane perpendicularly to the angle between its axis and the aircraft velocity vector.

The second factor that increases this kind of instability is the first moment - as it direction is always to the prop axis deflection it decreases both directional and pitch stability. It's a well known fact - for Mustang the pitch stability change can be up to several %% of MAC when power-on.

The third factor is a pure gyro - it amplifies the P-factor effect: rotating left due to P-factor - nose up that increases it. But regarding the P-factor swirling it acts in counterphase so it can be count as a damping factor.

Of course, stability and damping factors are airframe stabiliy and self-damping.

The strongest factor that reduces yaw stability are the flaps. When the plane yaws, the slipstrim goes to the lagging wing increases its drag (and lift). Increased drag leads to the further yaw and the lift decreases natural dihedral effect.

Overall plane stabiliy depends on all these factors.

And what is your experiment has set is the worst conditions:the strongest P-factor at low speed (high AoA) the strongest downwash and I will be not surprised if you use 100% of fuel load.

And the worst balance, by the way, is a civil plane with 100%.

Anyway, I tried to reproduce your experiment myself and did not encounter this swirling. Not so stable, walking right and left but if you trim it properly there are no steady or even dampen noticable oscillation.

Edited July 23, 2012 by Yo-Yo

PeterP · July 23, 2012

Seems you didn't read my last massage wile you wrote your reply.

-It is no offend against you or the FM as a whole.

What makes you so confident?

Because when I counteracted the torque and establish a "somewhat stable flight path" - this happens all the time .

Also when I only use keyboard controls. - so I can rule out the vibration from the FFB at my stick.

The plane normal has to go in left wing down spin.(as long I understand the physical-laws in our universe right)

But it only stays wing-level while it is spinning the nose via the GC.

And what is your experiment has set is the worst conditions:the strongest P-factor at low speed (high AoA) the strongest downwash and I will be not surprised if you use 100% of fuel load.
And the worst balance, by the way, is a civil plane with 100%.

Anyway, I tried to reproduce your experiment myself and did not encountered this swirling. Not so stable, walking right and left but if you trim it properly there are no steady or even dampen noticable oscillation.

I fully understand ! :)

My image is only a "example" because you get to the point when this happens almost instantly . - but this happens also often in other conditions like posted in this thread:

Pitch and yaw stabilty in Approach config.

Just wanted to point out that P-Factor and Torque can't describe entirely what is happening in this state.

:)*Don't shoot the massager!*:)

EDIT:

Next Thursday I will have the time/concentration to record a "keyboard only" track that shows the same behaviour as the video in post #1 (and I will try to keep the track as short as possible)- please let me know if I have to take the time to reproduce this again. - so you can examine the track/input logs.

Up to now I have only tracks where it happens somewhere after 15min of flight - and many of the Tacks I have recorded are also broken in replay when using "fast-forward"...

Edit again:

Yo-Yo

please don't get me wrong.

I'm really amazed how good the flight model feels - even better now that I have working FFB. .

I really don't want to use a other flight-simulator any-more... !

(and I really cant say much how it has to be in RL - I have only gathered about 17houer as a passenger/somewhat "co-pilot" in Cessna 162/172 with my 35years through my whole life up to now - I have no PPL - and it looks that I never can afford one here in Germany with my current job)

But,...

I hope you agree : A stable flightpath is no "challenge" for a really good flight-model...

Lets improve the "edges".

And I already pointed out that the "causal" pilot won't notice this "swirling oscillation at low speeds" during a normal flight.

.

It's no "big" issue for me , but I think it is worth mentioning.
(but sure can be a big problem for aerobatic teams)

But some few will.

I'm just trying to help you to improve the FM even more.

:)

Edited July 24, 2012 by PeterP

Yo-Yo · July 23, 2012

Ok... but I can make a video where no steady oscillations presents. Where is a difference?

My reply was not defencive-to-offence :) I only want to show that precession nature of P-factor is intrinsic for it. By the way - the precession is very intrinsic either for airframe (roll and yaw free movement makes a nose to draw ellipses because of 90 deg phase shift between roll and yaw) or for prop itself (gyro and P-factor).

PeterP · July 23, 2012

Ok... but I can make a video where no steady oscillations presents.

I can make also a video with "no steady oscillations presents" too.

But that isn't the question...

..I wasn't asking if someone can do a video with the same settings showing none of this symptoms:)...

The best outcome would be if no-one would be able to make ever a video with this steady oscillations .

My reply was not defencive-to-offence :) I only want to show that precession nature of P-factor is intrinsic for it. By the way - the precession is very intrinsic either for airframe (roll and yaw free movement makes a nose to draw ellipses because of 90 deg phase shift between roll and yaw) or for prop itself (gyro and P-factor).

I fully agree again!

Yo-Yo , I try now a different approach for explaining it,

...lets assume that we have a fully working simulation of the weather-engine... with air-holes , turbulence , whatever...(I know:- I'm dreaming:))

We both would never have this conversation...

Let me explain:

- if the moving air layers would be fully simulated:

a P-51D pilot in DCS would be very busy to hold his flightpath - so he would have to change his stick-position much more often as we are used to.

...

So it would probably very-very difficult to get in the same conditions predictable that are pictured in posting #1.

(I hope you get my point)

So lets get back now to your question:

Where is a difference?

The difference is that I now can reproduce predictable something that I would never experience in RL.

And if - only for a very small time-frame until I would have to counteract it - so I wouldn't have my left wing rotating down.

But right now I can reach a state in the sim where I have not to counteract a wing-spin.

- the plane simply reaches a state between stall/spin where it isn't able to decide which way to go...

(sorry - my English isn't good enough to describe it more clearly)

More precise: I can hold my controls "rock-solid" while the plane does this:

This is the difference.

So maybe adding something like a "randomiser" would prevent the plane from getting stuck into this this "loop" - as long we have the possibility to simulate also the "weather" .

Edited July 23, 2012 by PeterP

Yo-Yo · July 23, 2012

The difference is that I now can reproduce predictable something that I would never experience in RL.
And if - only for a very small time-frame until I would have to counteract it - so I wouldn't have my left wing rotating down.

But right now I can reach a state in the sim where I have not to counteract a wing-spin.

- the plane simply reaches a state between stall/spin where it isn't able to decide which way to go...
(sorry - my English isn't good enough to describe it more clearly)

More precise: I can hold my controls "rock-solid" while the plane does this:

This is the difference.

So maybe adding something like a "randomiser" would prevent the plane from getting stuck into this this "loop" - as long we have the possibility to simulate also the "weather" .

I think that you probably right (I could use some scary words like "phase plane", "bifurcation", "saddle', etc) and the ideal athmosphere makes possible to repeat this cycle in very unstable area of the phase plane.

PeterP · July 23, 2012

Thank you for this reply !

- I really appreciate it -

and I know ...

(better: I believe to understand)

...how hard your job is right now!

Edit:

(I could use some scary words like "phase plane", "bifurcation", "saddle', etc)

Haha - I wanted also use wordings like "Saddle" - but I wasn't really sure if the German mathematical description "Sattel" equals the English one...or you understand it when you translate it back to Russian...:) so I just wrote :

(sorry - my English isn't good enough to describe it more clearly)

or just instead writing something like "the plane feels like the "Schroedinger-cat" .. It waits for the pilot to make some input - so it is able to decide if it is dead or not ...

Edited July 23, 2012 by PeterP

Yo-Yo · July 23, 2012

Thank you for this reply !
- I really appreciate it -
and I know ...
(better: I believe to understand)
...how hard your job is right now!

Edit:

Haha - I wanted also use wordings like "Saddle" - but I wasn't really sure if the German mathematical description "Sattel" equals the English one...or you understand it when you translate it back to Russian...:) so I just wrote :
or just instead writing something like "the plane feels like the "Schroedinger-cat" .. It waits for the pilot to make some input - so it is able to decide if it is dead or not ...

Седло, седловина.

I think you will be glad: Nick Grey just watched your video and his feedback on it was that this is typical behavior at low speed. Sic.

Edited July 23, 2012 by Yo-Yo

PeterP · July 23, 2012

I think you will be glad: Nick Grey just watched your video and his feedback on it was that this is typical behavior at low speed. Sic.

Yes!

I'm very glad- not kidding!

Great!

Great job on the SAFM Yo-Yo!

...but now I (we both) have a real problem...:D :

Is this normal ? or a limitation of the flight-model ?

If normal: please explain me the physics behind this.

... I have the feeling that this answer will give me some sleepless nights....

:smilewink:

Again:

Many thanks for your response!

and greetings to Nick and the whole team to answer questions like this at late European times while "normal" persons are since hours in the bed - have a peaceful rest.

edit:

Would be very nice to get some new " DCS Mustang Developer Notes " about this. :)

Edited July 23, 2012 by PeterP

leafer · July 24, 2012

(I could use some scary words like "phase plane", "bifurcation", "saddle', etc) and the ideal athmosphere makes possible to repeat this cycle in very unstable area of the phase plane.

Those words are rather sexy. You free to night, Yo-Yo? :D Joking aside, excellent work, ED.

VIMANAMAN · July 24, 2012

Sounds like this issue has been explored pretty much fully...

...but for the record it existed before 1.2. This is 1.1.2.1 from 6-7-12.

I've only experienced it this once but found it pretty odd then. But equally I was aware I was on the edge of control.

For what it's worth...

Wolf Rider · July 24, 2012

^ This video is private. ?

VIMANAMAN · July 24, 2012

Thanks Wolf - I thought private meant only those with link could view it:doh:

It's public now.

PeterP · July 24, 2012

Thanks !

Do you have still the Track?

Can you upload it?

I hope you know the rules :D :

No track = it didn't happened !

:P

I have still a 1.1.2. on my system - I really want to see it and compare with 1.2.0.

Sounds like this issue has been explored pretty much fully...

...but for the record it existed before 1.2. This is 1.1.2.1 from 6-7-12.

I've only experienced it this once but found it pretty odd then. But equally I was aware I was on the edge of control.

For what it's worth...

Edited July 24, 2012 by PeterP

Sign In

swirling oscillation at low speeds P-51 1.2.0

Recommended Posts

PeterP

PeterP

HiJack

PeterP

HiJack

Nate--IRL--

MTFDarkEagle

HiJack

PeterP

Vibora

MTFDarkEagle

PeterP

Yo-Yo

PeterP

Yo-Yo

PeterP

Yo-Yo

PeterP

Yo-Yo

PeterP

leafer

VIMANAMAN

Wolf Rider

VIMANAMAN

PeterP

Recently Browsing 0 members