Delayed Wash... and the woobling thing...

[Guest deleted@50488]

I just wonder if this recent update to X-Plane 11 might be of some interest for the Devs here at 1C / 777 since I believe X-Plane and IL-2 share very much in common regarding the modeling of flight dynamics...

 

The part of the video where Austin starts talking about delayed wash and how it can cause aircraft perturbed in pitch to react less abruptly - wooble less - when the control are returned to neutral is around min 35:00...

 

https://youtu.be/WROEk_Jv33w?t=2129

[19//Moach]

I think this may have hit the nail right on the head.

 

I too have always found x-plane to feel "squirrely" in a very similar way to BoS - something the more "parametric" sims such as FSX, (as opposed to geometrically-driven such as this one), somehow managed to avoid. And in that particular regard, these perhaps less scientific, symptomatically-modeled sims ended up feeling tighter and a bit closer to the real thing.

 

This 10~20th of a second delay he mentions just might be the root cause of the infamous and ever so elusive "wobbles" -- That specific time frame also happens to roughly match the observed period of strange instability (over-rotation, rubber-banding, call it what you will) we experience in BoS.

 

It may well be the root cause and possible solution to what I (and I suppose many others) see as the major single issue regarding flight modeling in this series.

 

 

To sum up what's being mentioned in the video, and how it applies to the "wobbles" issue in X-Plane (and quite possibly here too):

 

 

 

Downwash on the tail caused by increasing lift with AoA imparts a proportional force on the tail, pushing it down.

 

This causes a well known moment of instability, where such a rotation acts to exaggerate itself in a manner similar to a feedback loop.

More AoA = more downwash = more force pushing down on the tailplane = even more AoA; And so on...

 

However, the time required for the downwash air deflected by the wing to reach the tail, is taken as infinitely short (instantaneous) in the sim.  This causes an unnaturally large reaction to changes in AoA. 

 

In reality though, due to the non-infinite airspeed of the plane,  this downwash change takes a fraction of a second to reach the tail.  

 

Such a seemingly insignificant delay is enough to noticeably dampen the resulting pitching motion, forcing the nose to remain closer to the freestream velocity vector throughout the maneuver. The airplane thus feels "tighter" on the controls.

 

This would be especially noticeable at lower airspeeds, when the delay between cause (downwash deflected by wing) and effect (downwash pushing on the tail) is largest.  Curiously indeed, this is exactly the behavior observed in the sim, where the wobbles are most apparent when flying near stalling speeds.

 

 

 

Simply put:   The short delay between an AoA change and the resulting downwash reaching the tail provides a dampening effect on the angular motion of the aircraft. According to Austin Meyer, this was the cause of the very similar wobble issue found in X-Plane. 

 

It is not any stretch of logic to assume it could be the cause of the very similar wobbles in this simulator as well.

 

 

That's the theory, at least. 

 

 

 

 

And here's Austin's simple fix in a nutshell:

 

Subtract the delta-downwash (pitch rate * immediate downwash * (distance from wing to tail / airspeed)) from the instantaneous downwash received by the tail on each simulation step.  This is the formula written on the board as seen on the video.

 

This effectively reduces the immediate forces on the tail by trimming off the excessive downwash (which isn't supposed to have gotten there yet) during AoA changes, without need for tracking wake shedding and all the associated complications of that complex method.

 

 

 

 

 

Edited December 12, 2019 by 19//Moach

[unreasonable]

On 10/6/2019 at 11:12 PM, jcomm-il2 said:

I just wonder if this recent update to X-Plane 11 might be of some interest for the Devs here at 1C / 777 since I believe X-Plane and IL-2 share very much in common regarding the modeling of flight dynamics...

 

The part of the video where Austin starts talking about delayed wash and how it can cause aircraft perturbed in pitch to react less abruptly - wooble less - when the control are returned to neutral is around min 35:00...

 

https://youtu.be/WROEk_Jv33w?t=2129

 

Not sure why I missed this: interesting, thanks for posting the link.