Rudder issue

[wlbragg]

@dany93 I have always had a small issue with spotty failures of the rudder, where one side is weighted more than the other. I don't recall exactly which side and the details other than I could get way more range out of one side over the other. Sometimes it would be so bad it would not be flyable.

It seems like maybe auto-coordination would improve this bug but that I am not positive about.

I recall a conversation with you while doing hydro tuning in the FDM where it was apparent that the rudder was weighted to one side for some reason I never understood.

Anyway I am piggy backing this with an issue brought up by a user in the forums.

I'm hoping you might be able to shed some light on this issue.

https://forum.flightgear.org/viewtopic.php?f=25&t=34847&sid=015feb752b3e4e81d59cc72d577d5b04

[dany93]

Hi @wlbragg

Two contributions make the aircraft steer leftwards: Spiraling Propwash effect #947 P-Factor #1130

Both are stronger at low or moderate airspeed, and decrease, then almost cancel, at higher airspeed.

• Spiraling Propwash: even with the aircraft stopped on the ground,
• P-Factor: due to asymmetry in the propeller blades velocity when the AoA is different from 0 (which happens at low or moderate airspeed). To understand this, think of helicopters (would be an aircraft with AoA close to 90 deg).

Its sensed effect is stronger with the amphibious stopped on the runway or with floats on the water (because the nose wheel do not counteract it), but the yaw moment is the same.

It was too strong when stopped on the runway or at start of rolling with the Spiralling propwash alone, but the introduction of the P-factor (very weak at very low airspeed) enabled me to decrease the total effect in these conditions.

The necessary rudder correction (right rudder) is not more than 0.1 to 0.2 (except for very brief correction), which never makes it not flyable. This rudder correction is for keeping coordinated flight (slip ball centered).

At low speed, high thrust, yawing leftwards is easy with weak or no rudder, but yawing rightwards needs significant rudder (about 0.2).

This response to try "shedding some light", but do not hesitate to ask questions, that's the best way to get appropriate responses (as long as I can...).

[wlbragg]

@dany93 so sorry to cause you to yet again explain Spiraling Propwash and P-Factor. This bug or by designed effect was neither of those two effects. It was a physical manifestation of the rudder that could be seen in the external view. After some though I am sure it was the rudder would move 100% to the left but only a fraction of that to the right. I recall you looking at the code and verifying there was something about the rudder movement portion of the code that did something slightly different to left rudder over right (or something that led me to believe it was not a programmed as simple left right movement). Unfortunately I just can't recall the details of the discussion. I know I didn't fully understand the explanation and also I do think the p-factor/propwash most certainly came up in that discussion so chalked it up to a bug with my local installation as it was really spotty for me. I'll investigate my installation and see if I can reproduce the physical limits of rudder movement I am describing. Maybe if I dig through the hydro tuning discussions I can find the section where it was discussed.

[dany93]

Can it be connected to the rudder trim value? I recall a discussion about the elevator animation vs the trim value. You did the animations

[dany93]

The rudder trim (like any trim) is aimed at having the aircraft fly straight and leveled with no action (commands can be loosened) despite intrinsic mechanical and aerodynamic asymmetry.

• In reality, if you set some rudder trim value, the rudder physical position with no pilot action is shifted, off-centered. Like if the pilot would exert some pressure on the yoke, but he doesn't need to. However, the rudder surface can reach the min and max positions (say, -1 to +1). Only, one direction is harder to reach, you need more effort. The loosened yoke position is offset.
• In FG, the simulation is a bit wrong: setting a trim value come down to shifting the origin (offset). e.g. trim +0.3 ==> the rudder amplitude becomes (should become) from -0.7 to +1.3. But the rudder amplitude in FG (as in reality) is limited between -1 and +1. Which makes that, with a trim = +0.3, the FG rudder only goes from -0.7 to +1. Asymmetric, not full deflection on one side. Differently from reality. Moreover, the loosened yoke position is always centered (no position feedback), which adds to confusing. With a trim value = +1 or -1, the FG pilot has no rudder action at all in one direction. The joystick will do nothing on half of its movement.

The default rudder trim value for the FG c172p is (and remains) 0.02, which is weak. It can be unnoticed.

Differently, the elevator default value is at 0, but if you set it at 0.2 to trim it (pitch down) at cruise, and let it at this value for landing, in FG you will no longer be able to reach the stall AoA at low airspeed (of course this is not a thing to do). For the real pilot, reaching the full deflection will be harder, for the FG one it will be impossible.