Ok, this is interesting. The static friction should be identical whether the brakes are on and off, if the wind is exactly lateral. That would be a simple test case for a bug report.

 

True, and I already opened one :-)

Ok, this is interesting. The static friction should be identical whether the brakes are on and off, if the wind is exactly lateral. That would be a simple test case for a bug report.

 

I think it has been that way for ages - there are two different ground models used in X-Plane. One for "brakes on", another one for "brakes off".

 

Pre XP11.0pb3 they simply "glued" the plane to the ground when the parking brake is set - because the "brake off" ground model is so borked that the plane would weathervane into the wind (due to computational creep, iirc) - just like everyone is"finding out" now.

 

What Austin did was adding "dynamic" motion to the "brakes on" model. The "brakes off" model is unchanged (afaik) and therefore still suffers the old problems (too little sideforce friction).

 

Jan

I think it has been that way for ages - there are two different ground models used in X-Plane. One for "brakes on", another one for "brakes off".

 

Pre XP11.0pb3 they simply "glued" the plane to the ground when the parking brake is set - because the "brake off" ground model is so borked that the plane would weathervane into the wind (due to computational creep, iirc) - just like everyone is"finding out" now.

 

What Austin did was adding "dynamic" motion to the "brakes on" model. The "brakes off" model is unchanged (afaik) and therefore still suffers the old problems (too little sideforce friction).

 

Jan

 

And only solutions ( rather ingenious I must confess... ) like those you used for the 727, or Dan Klaue for the Carenados, or simply using the Plane Maker Art Stab for speeds up to Vs or Vmc, can somehow offer alternatives...

 

Will this get fixed AT LAST in the 11 run, or will we have to wait for another enthusiastic moment after the release of XP12 :-) ?

 

I started posting at the .Org about this decades ago...

To be clear, the issue I was talking about was brakes off.

 

You should not have to fight right rudder to taxi at 10mph. It's exceptionally annoying, but it's existed forever. I have no idea why it hasn't been fixed yet.

Don't worry Jose like the Torque thing reported by Murmur it got fixed now this reported by you hmmmm 

 

 

[TECHNICAL POST]

 

Great explanation, thanks for taking the time.  

 

Regarding the weathervaning, If it only involves increasing the lateral force in the ground friction model when brakes are off, it seems like it would be an easy fix.  What am I missing?

Great explanation, thanks for taking the time.  

 

Regarding the weathervaning, If it only involves increasing the lateral force in the ground friction model when brakes are off, it seems like it would be an easy fix.  What am I missing?

 

You are missing that Austin doesn´t hold much interest in spending time on this, thinks its kinda ok and everyone kinda got used to it/has resigned from trying to make them change it.

 

The stance is like "yeah, its complicated, we will get to it...sometime"...

 

Only solution: Massive, continuous, easy to reproduce, crystal clear bug reports with unrefutable evidence and log files by MANY users. If the same people keep writing Austin about this, he simply blends them out as "nerds". So we need many reports - ideally from people with real light-aircraft experience.

 

You know what to do.

 

Jan

Only solution: Massive, continuous, easy to reproduce, crystal clear bug reports with unrefutable evidence and log files by MANY users.

 

Made other experiments, the plot thickens.

 

I made a three-wheeled "cart" in Plane Maker, to study tire friction under controlled conditions.

 

Now, the max side load that the gear is able do apply, is in absolute agreement with the theoretical predictions. This is true both with brakes applied or not (in a perfect side load, of course, the brakes should have no effect).

 

The issue arises when the gear is subject to a rotational moment (torque). In this case, the max moment applied by the gear is significantly smaller than it should be.

 

Also, the cart starts to creep, rotating on itself, a lot before than it should (i.e. for lower values of the applied rotational moment).

 

Apparently, this deficiency in the max moment that the gear can apply, could be at the origin of the weathervaning/ground handling issue.

Ok, after further tests, there are 2 issues when the gear is subject to a rotational moment:

 

1) apparently, the max moment it can apply is the half of what it should be;

 

2) moreover, it starts to creep (rotating on itself) at very low moments, one order of magnitude lower than the predicted ones.

Murmur,

 

does your cart have an asymmetrical surface exposed to the relative wind, something resembling a fin, or a homogeneous ( i.e. rectangular shaped ) area ?

Murmur,

 

does your cart have an asymmetrical surface exposed to the relative wind, something resembling a fin, or a homogeneous ( i.e. rectangular shaped ) area ?

 

No, it does not have any surfaces except for two mini wings with a microscopic area (necessary to give it a Cd greater than zero).

 

But it does not matter anyway, because I'm experimenting with zero wind and only using forces and moments provided by a plugin I wrote.

Or maybe a heavier:

 

How sad is that ... an 'old girl', reaching for the skies - where she belongs. Hmmm....

But it does not matter anyway, because I'm experimenting with zero wind and only using forces and moments provided by a plugin I wrote.

 

 

 

How are you varying the Cfs with "v" in your model, for comparison with X-Plane's internal results ?  I mean the normal and axial components only ( rolling and sideways/sliding  ) ?

 

There are a few formulas available, some depending on the detail / complexity of the wheel/tyre modelling, which we do not exactly know for sure how it's being modelled in XP11...

 

The most basic approaches, not even taking into consideration any lift effects that are present in aircraft, are for example those mentioned here:

 

http://www.school-for-champions.com/science/friction_sliding_coefficient.htm#.WF4_BIXXLb0

 

but much more detailed formulas can be used taking into consideration the variation with increasing speeds, and other factors. Tyre and strut modelling can also interfere since they apply a moment to the whole system, creating differences in the calculations between the upwind and downwind sides, and the center / nose or tail wheels...

 

It can be complex, specially if we do not know exactly what model Austin is using...

 

This one looks cool :-)

How are you varying the Cfs with "v" in your model, for comparison with X-Plane's internal results ?  I mean the normal and axial components only ( rolling and sideways/sliding  ) ?

 

I'm using the simple equation: Max_Static_Side_Friction = mu_Static * Normal_Load, with mu_Static set to 1.00 in Plane-Maker.

 

Rolling friction (set to 0.0035 in Plane Maker) is negligible compared to static friction. Cart is unbraked.

 

I'm pretty sure that X-Plane uses a basic model like the ones you linked. In my experiments, there is only a yawing moment and I made sure that all other moments and deformations are negligible.

 

Basically it's a three-wheeled cart with a wheelbase of 3 metres and a weight of 10000 N (1020 kg). With a total normal force of 5000 N on each of the two axes, it should resist a pure yawing moment of 15000 Nm.

 

But in X-Plane it starts to slide when a moment of less than 1500 Nm is applied.

Murmur, at least you're in the good way - maybe, just as hapened with the torque, you can make Austin sensible to it :-)