https://drive.google.com/open?id=0B5fyJwRhjar5Zk1wR0l3ekszUWs

 

Remember that if not braked and with the engine at idle, it will weathervane into the wind, probably also because of X-Plane tire model in this phase, but once you start the takeoff roll, it should be fully controllable in yaw. I can takeoff and land with 30 kts direct crosswind. It may not be perfect, but should be much better than the default.

https://drive.google.com/open?id=0B5fyJwRhjar5Zk1wR0l3ekszUWs

 

Remember that if not braked and with the engine at idle, it will weathervane into the wind, probably also because of X-Plane tire model in this phase, but once you start the takeoff roll, it should be fully controllable in yaw. I can takeoff and land with 30 kts direct crosswind. It may not be perfect, but should be much better than the default.

defiantly better the default ! Have you talked to Austin about your semi-fix ?

Feels better now Murmur.

 

I made a little change in the configuration ( PM ).

 

Also, as we can get from Austin's answer to my post at the Developers Blog, he may not be able to correct the behaviour when brakes are released.

 

It's always a compromise in a sim. I don't liek to see the true geometry of solutions like your's, or using Art Stab or other tricks to overcome the limitations of the FDM, but heck, if they produce good effects, than why not use it ...?

https://drive.google.com/open?id=0B5fyJwRhjar5Zk1wR0l3ekszUWs

 

Remember that if not braked and with the engine at idle, it will weathervane into the wind, probably also because of X-Plane tire model in this phase, but once you start the takeoff roll, it should be fully controllable in yaw. I can takeoff and land with 30 kts direct crosswind. It may not be perfect, but should be much better than the default.

Great job!

It's always a compromise in a sim. I don't liek to see the true geometry of solutions like your's, or using Art Stab or other tricks to overcome the limitations of the FDM, but heck, if they produce good effects, than why not use it ...?

 

I agree, I don't like using art stab or physically implausible tricks to tweak the flight model.

 

The modifications I made to the C172, are just very quick and dirty.

 

That is not how I would make them, if I had to make an accurate aircraft and hence if I had a lot of time to tune its flight model.

 

After all, Plane Maker + Airfoil Maker allows a lot of control over every aspect of the flight model. For example, the "right way" to tune the control and stability coefficients, would be to tune the area, taper ratio, aspect ratio of the various surfaces, and then tune the flight controls chords, etc.

 

This way, you could tune the flight dynamics to account for the inevitable inaccuracies of the native flight model, while at the same time keeping the basic geometry accurate and physically correct, and hence without having to expect weird behaviour in some other part of the flight envelope.

 

Of course, the optimal way to do that requires a precise workflow, for example, the first thing to do would be probably to tune the lift and drag (and hence the performance), then to tune the longitudinal and lateral dynamics, and then finally to tune the control derivatives.

I asked Austin what prevents it to work better with the default FDM in X-Plane ...  Let's see if I can understand his answer. I remember both of us were in a cyce of emails where Ben himself also tried to get to understand why there was such a limitation in the core FDm ( weathervaning )...

 

 http://developer.x-plane.com/2016/12/fight-model-improvements-done-for-x-plane-11-00/#comment-16009

 

But I guess that if we had access to the design process in sims like DCS and IL.2, we would find exactly the same subtleties... Just as with the works of Art done by A2A, RealAIr, PMDG and others with FSX...  Only Tom Goodrick always insisted in using only the default tools available in MSFS to build is aircraft models, which, btw, were really good...

 

What sometimes surprises me is how for instance the author of a Bf 109 G-2, and now G-6 too for X-Plane 10 and 11 can do what he does using just Plane-Maker for the tuning of the flight model. Sometimes we get the idea of the true potential behind X-Plane's FDM...

 

And, given the applause it is receiving from the simmers community, let's hope Austin really finds the time to dedicate to the cause this time :-)

I asked Austin what prevents it to work better with the default FDM in X-Plane ...  Let's see if I can understand his answer. I remember both of us were in a cyce of emails where Ben himself also tried to get to understand why there was such a limitation in the core FDm ( weathervaning )...

 

 http://developer.x-plane.com/2016/12/fight-model-improvements-done-for-x-plane-11-00/#comment-16009

 

But I guess that if we had access to the design process in sims like DCS and IL.2, we would find exactly the same subtleties... Just as with the works of Art done by A2A, RealAIr, PMDG and others with FSX...  Only Tom Goodrick always insisted in using only the default tools available in MSFS to build is aircraft models, which, btw, were really good...

 

What sometimes surprises me is how for instance the author of a Bf 109 G-2, and now G-6 too for X-Plane 10 and 11 can do what he does using just Plane-Maker for the tuning of the flight model. Sometimes we get the idea of the true potential behind X-Plane's FDM...

 

And, given the applause it is receiving from the simmers community, let's hope Austin really finds the time to dedicate to the cause this time :-)

 

 

It may be just a buggy code lying somewhere deep in the engine ?

 

It's not possible to take off the G2 / G6 during X-wind in XPlane.

 

 

It's not possible to take off the G2 / G6 during X-wind in XPlane.

 

Even with the tailwheel locked Zulfi ? I confess I never tried with x-winds... The Carenado C185 for sure can't....

Even with the tailwheel locked Zulfi ? I confess I never tried with x-winds... The Carenado C185 for sure can't....

 

Hmm i don'r remember though but i can try again later today with the tail wheel locked. The 185 is also impossible and it don't have any tail lock

 

 

The 185 is also impossible and it don't have any tail lock

 

I believe that if you assign the tail lock function to a button it also works in the C185 ;-)

I believe that if you assign the tail lock function to a button it also works in the C185 ;-)

 

Ok i will check it out. But i don't know if the real 185 has a tail wheel lock.

I asked Austin what prevents it to work better with the default FDM in X-Plane ...

 

I'm convinced that there is not a real problem with the FDM and weathervaning, not in the sense that the FDM produces accurate results using accurate dimensions, but in the sense that it can be tuned to give accurate results.

 

After all, the flight model is simply a sum of forces and moments for every point of the flight envelope. Now, the forces in the linear range (in this case, the unstalled vertical tail) can be easily tuned by modifying stab area, taper ratio, airfoils, etc. The forces in the non-linear range, can be tuned instead by editing the airfoil in the stalled region.

 

Of course, this also requires an iterative process, analyzing cycle dump output, etc., and of course it also requires doing test flights with the real aircraft and/or real world flight test data.

 

So, why most X-Plane aircrafts have this exaggerated weathervaning tendency? My thought is because X-Plane's default flight model tend to overestimate some forces, and most aircraft designers do not have the ability to tune the flight model so that it is really accurate in all its aspects (I mean, in most of its stability & control derivatives), because that requires a very good knowledge of flight dynamics.

 

My impression is that most aircraft designers (including some payware's) just use default X-Plane flight model, tweaking it mostly for accurate performance, but with only slight tweaks for the rest of the flight dynamics. For example, you can notice that there are payware aircrafts that do not use custom values for radii of inertia.

 

All of this is true provided that the tire friction model is correct. But after my latest experiments, it seems to be: the forces the tires can apply are as predicted. There is only a doubt on the transition phase between not moving tires and moving tires, where it could probably be improved, but the exaggerated weathervaning tendency during the takeoff and landing roll do not seem to be due to that.

Murmur, could the weathervane be caused by fuselage drag? I remember that Laminar says that bodies on X-Plane will have lateral drag even if we set their drag coef to 0.0, this may be causing the issue as maybe the fuselage and all bodies are trying to align into the wind?

Well i tried the G2 / G6 at Half Moon Bay , used custom weather. The rwy is 30 and I set the wind direction to 105 @11kt @surface level , locked the tail wheel first try i went completely sliding to the left and then next try it was way better

Murmur, could the weathervane be caused by fuselage drag? I remember that Laminar says that bodies on X-Plane will have lateral drag even if we set their drag coef to 0.0, this may be causing the issue as maybe the fuselage and all bodies are trying to align into the wind?

 

Yes, fuselage drag also plays a role, hence the reason why in this more simplistic approach Murmur clipped part of it, as he will certainly explain further :-)

 

It's good to know such a factor is taken into consideration in the core FDM, but, as Murmur points out, in order to buils a close to real representation of a given aircraft, much more than simple aircraft design ( 3d ) and Plane-Maker configuration may be required, including tweak of airfoils.

 

Unfortunately when real airfoil data is used in Airfoil-Maker, as a friend used in a glider model design, the results aren't the expected :-/