Exponential for pitch/roll/rudder axes

Quote

Unless I have not seen it, it appears that only linear relationships between controller and pitch/roll/rudder deflections can be setup. I prefer an exponential setup so the amount of movement near the centering point is much less than near the extremes.

Hey,

Currently Aerofly implements a proportional stick deflection ratio (simulated deflection to real control stick deflection is a constant). Actual centering forces from spring loaded trims, aerodynamic trims, aerodynamic forces and forces caused by acceleration (e.g. mass inertia of the controls), friction and so on are not simulated currently. And because only very few, mostly expensive joysticks actually have a true force feedback with a moving neutral position one can say that the hardware isn't quite there yet. Almost all devices available on the market have a spring loaded centering mechanism that brings the joystick to the exact center. However this is not how real aircraft operate. The force neutral point could be quite far forward or aft depending on the trim setting.

Also, the majority of input devices affordable today have a stick that is much much shorter than the in the real world aircraft. A lot of airliners in the Aerofly FS 2 have a yoke that is hinging near the flight deck floor. Unless you lengthen your joystick you will probably never get the same fine control inputs as in the real world on any desktop flight simulator.

Personally I have removed my joystick spring and also have abandoned any non-linear control behavior since I stared flying in the real world. The gliders that I've flown in the real world all only require very small inputs and using something like expo in the simulator would make me overcontrol the aircraft in real world. So personally for me expo is an absolute no-go. But I understand that you want to control the aircraft realistically without purchasing more expensive hardware, so maybe we can do something like expo.

Quote

I also notice that the Extra has a second order damped response (it rings/oscillates a little with a step input) - is that also for the actual aircraft?

The response of the Extra to a step-input is what I would expect. All aircraft that I have flown IRL have had such a reaction. First pitch input, then angle of attack build up, then stick release, continuation of the rotation and then return to the aerodynamical neutral position in an oscillating fashion.

Any input that you make will add a torque onto the airframe, which over time will start to spin the aircraft faster (pitch rate, roll rate). The speed at which the aircraft increases its pitch/roll/yaw rate depends on its inertia around the corresponding axis. Me model this inertial and the mass distribution physically accurate and in the .tmd file of each aircraft.

Once the input is neutralized the aircraft is still spinning and only aerodynamic effects such as the stabilizing effect of the tail surfaces are creating a stabilizing torque to dampen and counteract the rotation speeds. Depending on the type of airplane, its center of gravity location and the rotation axis an oscillating reaction can be observed, it's very typical for the pitch axis for example. With a short and abrubt input the real world aircraft and the Aerofly FS 2 aircraft (non-computer stabilized) show a brief angle of attack oscillation followed by the much longer phygoid oscillation.

Cheers,
Jan

First of ll you need a Unicode capable text editor like EditPad, Notepad++ or something like that. Then you can open the Aerofly FS 2 installation folder (in steam apps) and make a backup of the current aircraft that you want to modify, just in case something goes wrong.

• Open up the aircraft.tmd file (e.g. a320.tmd or extra330.tmd) in the aircraft's folder. Search for "ServoLeftAileron" until you find a block of code that describes a servoclassic with Name "ServoLeftAileron".
  This object works like a polynomial function with: output = P0 + input * P1 + input^2 * P2 + input^3 * P3.
  P0 would be used for a static aileron trim,
  P1 is the linear part
  P2 can be used for aileron differential
  P3 is the "expo" part
• For correct expo you want to increase P3 whilst decreasing the P1 value by the same amount, so that your center becomes smoother and your maximum deflections (input = 1.0 or -1.0) stay the same. So go ahead and set the P3 to something like 0.15. Now reduce the P1 value by the same amount.
  E.g. P1 of 0.5 would become 0.5 - 0.15 = 0.35.
• Repeat the process for the ServoRightAileron but this time set P3 negative and increase the value of P1. P3 would be -0.15 for example and the original -0.5 of the P1 value would now be increased to -0.35 instead.
• Save the file and reload the aircraft in Aerofly FS 2.
• When you are happy in the way the ailerons react you can then also change the ServoElevator in the same way as the ServoLeftAileron as well as the ServoRudder, the same way as the ServoRightAileron.

I hope this is of some use to you, just try it out. Make sure to backup your work (save the .tmd file somewhere else) so that any automatic updates from Steam don't overwrite your changes.

Cheers,
Jan

If you use 0.2 P3 and I think you are already doing something like 40% expo, I not sure how expo is defined exacly, I'd guess p3 / f(1.0)...
Yes inputs are normalized from -1.0 to +1.0 for all the primary flight controls in Aerofly FS 2. In fact almost all inputs are either from -1.0 to +1.0 or from 0.0 to 1.0.