DCS P-51 Propwash Effects Video

[flightace37]

I did a little testing to see how propwash was modeled. This should help answer some of our questions about ground handling. Enjoy the video!

 

pUgDqhPnekw

 

Working on embedding it, but the youtube tags don't seem to be working.

EDIT: Thanks Viper! Now I see how the format works.

 

EDIT2:

. Edited May 9, 2012 by flightace37

[Anatoli-Kagari9]

Even if it was not for the rest, and although I never really appreciated air combat, I bought LockOn on 2005 just because of the excellent physics model. Back than my rig wasn't up to the requirements, but now I have an i5 2500 and a Geeforce GT 450 1GB, plus 8 GB RAM on windows 7 Home Premium 64 bit. I am sure DCS-World will run on it :-)

 

And... birthday comming along ... will be 48 in a few days so... Guess what :-)

 

Great video, great simulation platform... How much I, and many, would love an incursion into civil aircraft, even if some hybrid like the C90 or Kingair also used for military operations, a Cessna 337, etc...

[ErichVon]

I did a little testing to see how propwash was modeled. This should help answer some of our questions about ground handling. Enjoy the video!

 

pUgDqhPnekw

 

Working on embedding it, but the youtube tags don't seem to be working.

EDIT: Thanks Viper! Now I see how the format works.

 

Well, if you get low enough with the A.I. guy on your tail you just might blind him with the prop's dust storm it creates. LOL!

[flightace37]

You'll notice the rudder doesn't function too well. I imagine ED will fix this before release though. It'd be interesting to see if they model the effect between aircraft. It would certainly add a whole new dimension to air combat.

Edited May 9, 2012 by flightace37

[xxJohnxx]

I don't think that the rudder can do that much, as long as you stand on your weelbrakes. If the propeller thrust itself can't push the plane forward, because it has a good grip and strong brakes, a "little" (compared to the propeller) bit more force against one wheel won't do that much I think, but I like to be corrected.

[effte]

John,

absolutely correct.

 

OTOH, I see little reaction to the movement of the elevator.

[leafer]

Still pretty impressive, though.

[flightace37]

Yeah, I think you guys are right about the brakes. I noted as much in the youtube description of the video when I first posted it. Does anyone see a tiny little bit of yaw motion when I kick the rudder over? It's hard to tell from that angle.

 

The way I see it, in a fully realistic model, the aircraft obviously wouldn't move around on the ground because of rudder action with the brakes held. It should, however, twist side to side a little bit because of the flexibility of both the tires and the frame. Asking for that to be modeled would be too much of course, but it would definitely be interesting if it was.

 

If you look closely at your aircraft on the ground, you'll actually see that your tires are squished down a little bit where they make contact.

[effte]

Try full rudder deflection from a standstill with brakes off and then a burst of power though. On the road now so I can't test it again, but I'm fairly certain the effect of rudder in that situation was just about none at all.

[sobek]

Try full rudder deflection from a standstill with brakes off and then a burst of power though. On the road now so I can't test it again, but I'm fairly certain the effect of rudder in that situation was just about none at all.

 

Not true, the effect is relatively small compared to the self tightening cornering behaviour, though. However, you can determine the direction that it will start turning into.

[sobek]

OTOH, I see little reaction to the movement of the elevator.

 

Are we watching the same vid?

[effte]

Yes. I see raising and lowering of the tail due to power changes, but it doesn't really seem to be in sync with the elevator movements.

 

Edit: Interesting test though - perhaps you or the OP can confirm that it can be/was done without changing the power setting?

Edited May 9, 2012 by effte

[159th_Viper]

It should, however, twist side to side a little bit because of the flexibility of both the tires and the frame....

 

It does.

[flightace37]

Regarding power changes:

 

Power was as constant as my physical throttle would allow (which is pretty darn good, being a CH Pro Throttle). The reason that the elevator wasn't having a very sudden effect is because I only applied enough power to barely get the tail off the ground at full deflection. If you watch carefully, you'll see that less deflection was required with the tail further up. That's something we would expect as the CG moves further towards the wheels.

 

 

Regarding the rudder:

 

Actually, I was doing some checks on ground rudder authority earlier yesterday. I turned the aircraft into a hockey puck by holding some directional brake and kicking the power up pretty high, then releasing brakes. Once spinning, I tried using the rudder to adjust the rate of rotation and it didn't do a thing. Only the brakes had any effect. It's true that the CG is aft of the wheels, but the rudder is still well on the outside, so it should have had some effect over the long term.

 

Same results for both left and right spins, so it's not just a matter of engine torque completely overcoming the rudder. I think that it's either not modeled, or modeled very weakly. Whether or not weak modeling is realistic, I'll leave up to the people with tail dragger experience.

 

@Viper: Cool. I'm not crazy then.

 

 

On another topic:

 

The tail wheel has a severe tendency to flip around when unlocked. Even with the brakes held, it flips around the wrong way under the weight of the aircraft, and stabilizes in that position. I think that the caster behavior might actually be modeled backwards.

 

EDIT: Thinking about that a little more, caster effect does require forward motion of the airframe. The real question is how the weight of the aircraft will affect the wheel, given the design. Does the tail sit lower to the ground with the wheel forward, or backwards? Does the wheel produce enough static friction to prevent it from moving?

Edited May 9, 2012 by flightace37

[sobek]

Yes. I see raising and lowering of the tail due to power changes, but it doesn't really seem to be in sync with the elevator movements.

 

Are you watching this on a cellphone? This is not what i am seeing.

[Weta43]

It should, however, twist side to side a little bit because of the flexibility of both the tires and the frame.

If you bang that video up to 1080 & fullscreen it, then watch the wheels as the rudder moves, don't you think it does squirm a little bit ?

 

 

(Edit : beaten to the punch badly while reading... )

[flightace37]

If you bang that video up to 1080 & fullscreen it, then watch the wheels as the rudder moves, don't you think it does squirm a little bit ?

 

 

(Edit : beaten to the punch badly while reading... )

 

Indeed. I wasn't sure if it was a trick of my imagination or not, but that's at least two other people who have confirmed my observation so far. With regards to your edit, "Don'tcha love forums?"

 

I continue to be impressed with the detail of the flight model in this sim. What we need now to alleviate all the concerns about rudder authority is a video of a real Mustang wagging its tail like I did in this video. No need to risk the prop by lifting the tail off the ground. We just need to see rudder deflection in the prop wash. Even a relatively low-power civilian bird ought to be enough to demonstrate how pronounced the effect might be.

[effte]

The tail wheel has a severe tendency to flip around when unlocked. Even with the brakes held, it flips around the wrong way under the weight of the aircraft, and stabilizes in that position. I think that the caster behavior might actually be modeled backwards.

 

EDIT: Thinking about that a little more, caster effect does require forward motion of the airframe. The real question is how the weight of the aircraft will affect the wheel, given the design. Does the tail sit lower to the ground with the wheel forward, or backwards? Does the wheel produce enough static friction to prevent it from moving?

 

I know I said I didn't have any complaints regarding the explanations given, but there was actually a minor niggle. The friction isn't what does it. Forces and geometry is. If the pivot axis is tilted forward, the tail sits lower with the wheel aligned. Any deflection will raise the tail and generate a centering torque. It's the same as on a pedal bike, where the rake of the steering column makes sure the front wheel aligns with the frame. Part of the weight reaction force will be perpendicular to the pivot axis and will have a moment arm when the wheel is not aligned with the fuselage centerline.

 

There are a couple of pictures in the maintenance manual which I think, even if they are not completely unambigious, shows a definite caster to the tail wheel pivot axis. I think I can have a look at The Real Thing in a bit, unless someone happens to be sitting on better data, e g blueprints.

 

In other words, if there is caster the centering force is there as soon as there is weight on the tail wheel, regardless of whether it is moving or not.

[flightace37]

Effte, that is actually not true. The effect of caster angle (whether in an arrangement like a bike, or a shopping cart caster) is to cause the contact patch of the wheel to trail behind the intersection of the pivot axis and the ground. Sideslip friction (for lack of a better term) on the wheel while the object is in forward motion is what causes it to return to center, not weight.

 

I just went outside and ran an experiment on my bicycle to test this. I set the wheel at a moderate angle from center (yaw), and pushed down hard, with the bike sitting vertical. There was no centering tendency at all. Moving the bike forward (at moderate speed) keeps the wheel aligned due to sideslip forces. This is what allows some (crazy) people to ride hands-off without having the wheel flip all over the place on them.

 

If that's not enough, think about a shopping cart wheel. With the cart sitting on flat ground, rotate the wheel with your foot. It doesn't try to return to the original alignment; it just sits there. The same is true with rake/negative caster on a bike. The contact patch offset is just created using a different method.

 

 

Now let's move on to the specific arrangement on the P-51. If you examine the tail wheel assembly closely, you'll find it looks something like this:

 

*
*
*
 *
   *
     *
     wheel

 

This is the exact design of a shopping cart caster, with all of the same properties. With the pivot axis perfectly perpendicular to the ground, there will be no torque applied unless the vehicle is in motion. The trouble comes when the pivot axis is not aligned perpendicular to the ground. This introduces a compounded problem. (You can call it caster of something that is castered, if you like; I don't know if there's a specific term for it, or if that description is even valid.)

 

With this additional tilt, the wheel will tend to self-center when weight is applied because CG height is now a function of wheel yaw angle. The things you describe in your most recent post apply.

 

If you rotate that diagram a few degrees counter-clockwise and then imagine applying weight straight down, the wheel will tend to flip towards the right, placing the weight closest to the ground. Imagine tilting it clockwise now. The wheel will tend to flip to the left for the same reasons.

 

We seem to agree on the basic concept here, but it is not the angle of the actual pivot axis alone that is causing the tail wheel to flip. It's a combination of the displacement of the wheel from the pivot axis's contact point with the ground (which creates a caster angle), and the angle of the pivot axis itself (which would be caster angle if the wheel was directly in line with the pivot axis). I'm sure you'll agree with that assessment.

 

So, given what we are seeing in the simulator, I think it's safe to assume that the pivot axis is modeled to be vertical when the aircraft is strictly horizontal. The problem is the aircraft is not horizontal while sitting on the ground. If we think about the aircraft in the same "diagram" that I made above, with the propeller on the left side, and the tail on the right, we can see that the pivot axis would be tilted clockwise. This causes the tail wheel to flip around to the left.

 

The question that needs answering then, is whether this design is correct or not. I can't imagine that to be the case, because it would have caused endless ground handling pains. The optimum design would ensure that the pivot axis is perfectly vertical when the tail wheel is in contact with the ground. When the aircraft moves, the wheel would swing to trail behind in the direction of motion, without having to fight a centering force caused by weight and this compounded caster mess.

 

I think I derailed my own topic, and it looks like we may be fighting over confusion of terminology, Effte. Unless you see a flaw in this analysis, what's say we call it quits?

Edited May 10, 2012 by flightace37

[Smokin Hole]

In the real plane the nose would go right into the dirt without the tail tied down. And I still think that the propwash around the vertical stab and rudder is either un- or undermodelled.

 

[EDIT] OK I'll eat a little crow on that one. The nose will go right into the dirt unless you keep the MP below 40 inches (just like the book says). Propwash is definately modelled as can be seen by the rocking side to side when moving the rudder with the brakes set. But there is still something odd about the takeoff roll. It just seems too unstable. I've talked to people who have flown the P-51. I once had a friend who owned one briefly (too expensive). None of these people ever mentioned any yaw instability when the tail comes up. A big tailwheel airplane like this one should be fairly easy for anyone with time in smaller, squirrelier types.

Edited May 10, 2012 by Smokin Hole

[Kankkis]

Aerofly FS propwash

 

Eg7CKCKlJC8

Edited May 10, 2012 by 159th_Viper
Vid embedded

[Smokin Hole]

Many moons ago I pissed off the owner of the glider operation I flew for by taking off in his Maule M7 from a tail-up start. Mr. Maule himself got airborne before passing outside through the doors of his hangar by doing this. So of course I had to try it. Always best done with someone elses plane :-)

[Tailgate]

The dummy is more interested in filming a tree, but you can see some a/c and rudder movement.

 

tDppwPbmbrY

[flightace37]

The dummy is more interested in filming a tree, but you can see some a/c and rudder movement.

 

Couldn't see a thing with all the bouncing and panning that guy was doing with the camera. All I want to do now is take a torch to that tree. Oh well.

 

We did get a nice look at the tail wheel. I think it moved around a little bit with all the vibrations, but never actually flipped around towards the front of the aircraft like it does consistently in our sim, nor does it auto-center without the aircraft in motion. There's a little bit of work to be done in that case, but we need to be patient with the devs.

 

Thanks for the video!

[Yo-Yo]

In the real plane the nose would go right into the dirt without the tail tied down. And I still think that the propwash around the vertical stab and rudder is either un- or undermodelled.

 

[EDIT] OK I'll eat a little crow on that one. The nose will go right into the dirt unless you keep the MP below 40 inches (just like the book says). Propwash is definately modelled as can be seen by the rocking side to side when moving the rudder with the brakes set. But there is still something odd about the takeoff roll. It just seems too unstable. I've talked to people who have flown the P-51. I once had a friend who owned one briefly (too expensive). None of these people ever mentioned any yaw instability when the tail comes up. A big tailwheel airplane like this one should be fairly easy for anyone with time in smaller, squirrelier types.

 

If you set TO assistant to100% you will see how easy is to hold the plane straight. It's not a problem of the model - it's a problem of in-sim acceleration percepetion. If you get the most accurate model of a bycicle you never ride it in sim or it will take a long time to train. The reason is lack of acceleration info. If you ride a bike you react instantly as you feel tendency to go aside. In sim you have to wait a little till your brain calculates the acceleration and right amount of control input and the more you wait the more is deviation - the worse is situation.

 

I am mentioning it so often that it deserves to be in FAQ.