Rudder pedals very sensitive

[hbeni]

The pedals are much more sensitive than with the 172...
Should that be the case?

Especially with taxiing this is sometimes problematic.

[hbeni]

From a first glimpse the nose wheel steering parameters are okay per POH.

• The /fdm/jsbsim/fcs/rudder-pos-deg and -rad values are nearly doubled compared to the C172p (16° vs. 27°).
• The aero/coefficient/CYdr (Side_force_due_to_rudder) is also different (0.150 vs. 0.187)
• The aero/coefficient/Cndr (Yaw_moment_due_to_rudder) is also different (-0.0645 vs. table based on velocities/u-fps ranging from -0.0604 to -0.0805, having the same for the 'climb' value)

@HHS81 Where do these numbers come from?

[hbeni]

@HHS81 aaand: is there a reason, the Pitch_moment_due_to_elevator_deflection does not start at 0 too, but with 107.1 -1.029 <!--Approach--> ?

c182s/c182s.xml

Lines 1890 to 1905 in 560a99c

	<function name="aero/coefficient/Cmde">
	<description>Pitch_moment_due_to_elevator_deflection</description>
	<product>
	<!--<property>aero/qbar-psf</property>-->
	<property>aero/function/qbar-induced-psf</property>
	<property>metrics/Sw-sqft</property>
	<property>metrics/cbarw-ft</property>
	<property>fcs/elevator-pos-rad</property>
	<table>
	<independentVar>velocities/u-fps</independentVar>
	<tableData>
	107.1 -1.029 <!--Approach-->
	133.5 -1.369 <!--climb-->
	220.1 -1.122 <!--cruise-->
	</tableData>
	</table>

[TheFGFSEagle]

Uh, yeah, because, when the speed is very low, the effect of t he elevator will also be very low - it increases quadratically with increasing airspeed. If you set it to 0 at u-fps=0, you would kind of converting this to increasing cubically (not exactly - but it helps to get your head around it !). The reason that there are different values at all is because when you fly very slow or very fast, you have different angles of attack, at which the elevator deflection has different strengths of effect - but making that depend on u-fps doesn't make any sense at all, IMO. But you could ask @ysopflying.

[ysopflying]

Thanks for the kind mentioning.
aero/coefficient/Cndr is ok with literature (-0.658)
aero/coefficient/CYdr Don't worry! The aircraft are not identical.
/fdm/jsbsim/fcs/rudder-pos-deg What is the POH saying? Could be mechanically different as well.

The separation into approach, climb and cruise takes into account different alpha, maybe flap and Reynolds effects. It could be done way more complex than just three values.
Between 0 and 107.1 there are no values available, so better to stick with the lowest known.
So the table looks fine.

The issue is primarily related to the noise wheel steering or the aerodynamic effect of the rudder?

[ysopflying]

Checked trajectory on the ground during taxi. Appears realistic.
As we don't feel accelerations at our desk, it is easy to miss the correct amount of rudder input, which is less of an issue in real life.

[HHS81]

The rudder pedals are quite sensitive on the real airplane as well, as I heard. I thinkt that's one of the first lessons for a student, stearing an aircraft on he ground and keeping it on the center line while taxiing.

[hbeni]

Thank you all for responding so quickly.
For me here in my desk the c172 and 182 react differently; the 172 way more damped and easier to control.

Is the 172 then not realistic enough?
I had assumed, that ground behaviour is at least in the same ballpark…
For takeoff with the 182, i need roughly 1/3 of rudder only.

[hbeni]

/fdm/jsbsim/fcs/rudder-pos-deg What is the POH saying? Could be mechanically different as well.

I searched and could not find numbers. Also I searched the web.
The full rudder deflection angle is more than the 172p Simulation. That may explain the difference in behaviour.

[HHS81]

/fdm/jsbsim/fcs/rudder-pos-deg What is the POH saying? Could be mechanically different as well.

I searched and could not find numbers. Also I searched the web. The full rudder deflection angle is more than the 172p Simulation. That may explain the difference in behaviour.

Yes, the C182 has more rudder deflection angle. It is a bigger, heavier aircraft. You get all numbers from the type certicifation sheets like: https://www.easa.europa.eu/en/document-library/type-certificates/noise/easaima052-cessna-182-series-skylane

Or better:
https://www.google.com/url?sa=t&source=web&rct=j&opi=89978449&url=https://www.easa.europa.eu/en/downloads/7421/en&ved=2ahUKEwi60r3EmpqFAxVyxQIHHY92DBoQFnoECBYQAQ&usg=AOvVaw14apKvozWr-MCa2-hDuAPc

or

https://web.archive.org/web/20100201153302/http://www.airweb.faa.gov/Regulatory_and_Guidance_Library/rgMakeModel.nsf/0/084b5675182d8714862575f6006d1bf4/$FILE/3A13%20Rev%2069.pdf

But there is something which might explain it: the sheet states full rudder deflection 24° parallel to 0.00 W.L line, and 27° perpendicúlar to the hinge line. Since JSBSim doesn't use the hinge line as I think, we might should use the 24° deflection angle

[HHS81]

@ysopflying Or does JSBSim uses the hinge line as reference, or a perpedicular hinge line?

[ysopflying]

As JSBsim control surface effects are not part of the core, but more of the aircraft specific formula set, it is more about industry standard of deflection angles.

Normally a rudder deflection is measured as an angle around the hinge axis. This hinge axis can be oriented at any angle relative to the aircraft reference system.
So max deflection should be measured the same way. Hence in my humble opionion 27°.

[ysopflying]

Crosschecked with windtunnel data for the C310. Cndr is depending on alpha and beta by some percent.
The C182 numbers as they are are in the right ball park.

[hbeni]

Thank you for verification.

so we will close this as wontfix?

[hbeni]

@ysopflying maybe, when you have data already at hand - could you also peek at #517 and #520 ?

[ysopflying]

so we will close this as wontfix?

Open for discussion. We need more real stick time!

I'd would leave this as it is, as behavior appears normal. A bit of mechanical damping is certainly beneficial.

Open to research is, if it is possible, that the cldr (roll moment due to rudder) at full deflection can overpower clda (roll moment due to aileron) at full deflection.
We need more stick time (with nobody watching.)!

[ysopflying]

@ysopflying maybe, when you have data already at hand - could you also peek at #517 and #520 ?

No data for these issues. A bit of thought posted for the pitching up. Will dive into it deeper later.
For the engine issue I am not an expert. We would need a good iteration to solve this.

[TheFGFSEagle]

We need more stick time

There is a C182 based at the nearest airstrip - if I ever get a chance of flying in it, what should I pay attention to, what do you need ?

(with nobody watching.)!

Why ?

[ysopflying]

For sure @hbeni has a long list for this!

Just because, when an airplane owner gives you the yoke of his precious toy, it might not be the best idea say "Hold your beer, I just need some FDM data" and throw it into high speed dives, spins, inverted spins.....

Joking aside, there is a lot, which can be done with a camera and the sensors in the cellphone.

[hbeni]

For sure @hbeni has a long list for this!

ohhhhh sure! We can start at the gauges behaviour when applying power, when the Engine starts and when is operating.
Or actual displayed data at an engine runup :)
Actuaylly a highres video of the instruments at various flight (and preflight) situations would be awesome.

[hbeni]

But for getting data of this, maybe a video of the pedals durint taxi and takeoff should suffice.

[ysopflying]

Suggest to open up an improvement issue "real pilot feedback", so people can tune in on various subjects and the wish list can go there.

[hbeni]

Good idea, i just drafted something.
Will go trough my scattered list and add as i find findings…

[hbeni]

So catching up on this again.

A bit of mechanical damping is certainly beneficial.

OK, just merged ee6e31e and ported it from the c172p.

But there is something which might explain it: the sheet states full rudder deflection 24° parallel to 0.00 W.L line, and 27° perpendicular to the hinge line. Since JSBSim doesn't use the hinge line as I think, we might should use the 24° deflection angle

But wouldn't that mean, that jsbsim uses the tilted 27° for calculations where the formulas currently assume 0° hinge line? So ending up in too big values?
But the comments also explicitely state Perpendicular to hinge line, so 27 seems correct:

c182s/c182s.xml

Lines 955 to 963 in 560a99c

	<aerosurface_scale name="Rudder Control"><!-- Perpendicular to hinge line-->
	<input>fcs/yaw-trim-sum</input>
	<gain>0.01745</gain>
	<range>
	<min>-27</min>
	<max>27</max>
	</range>
	<output>fcs/rudder-pos-rad</output>
	</aerosurface_scale>

I just tried out 24° vs. 27°; and it seems it does not matter much.
With the damping kinematic, it is already better for me. Maybe it was just jitter from the rudder pedals making the FDM confused or so?

[ysopflying]

It would matter 23/27= abt.15%
Kinematic damping is certainly beneficial, considering the rudder needs to be moved mechanically.
Reference axis needs to be same as in the source of the data.

[hbeni]

See other discussions; tldr: port lag filter from PA28 for rudder, aileron and elevator.
#582 (comment)