Real World Pilots, please state your feedback about the flight model

I’m familiar with adverse yaw. I thought you meant using the rudder to initiate a turn.

2 Likes

Negative.

:wink:

Just a shame you don’t in the sim, with the GA aircraft I’ve flown anyway.

Paul Faulke.

I did not find the CAP 10 fleece
The wing geometry of the CAP 10 is identical to that of Emeraudes and Super-Emeraudes: same shape in plan, same surface and same NACA 23012 profile. I don’t have the polar either!
Aerodynamically, it is slightly different, and because there have been some changes. (wingspan, rigging).

It is true that on the CAP 10 of the simulator for normal flight we do not need the steering, except for certain aerobatic figures: slow rolls, spins, reversals, rolls triggered, slips etc …
In truth, it is very different.
From the first meters of taxiing on take-off, the plane goes to the left, it is necessary to counter the direction to stay on the runway axis: this is mainly due to the helical breath of the propeller (instead the designers of the simulators of general public flight, sometimes prefer to model the P factor and the overturning torque, whereas in my opinion these torques are negligible)
Then when moving to the 2-point position (two wheels) the gyroscopic torque of the propeller and the engine the aircraft tends to swerve to the left which must be countered in the steering.
It should be noted that unlike the tricycle landing gear, the aircraft is unstable in direction (a swerve tends to be amplified if the pilot does not react!)
After takeoff, during the climb, as the speed is low, it is necessary to maintain pressure on the right on the rudder to maintain symmetrical flight, otherwise the ball (slip indicator) goes to the right. Then when the speed increases we can release the rudder. Normally the designer of the aircraft has provided a tail wedge so that the steering is in neutral at cruising speed. Above the cruising speed it is necessary to put the direction to the left to maintain a symmetrical flight (otherwise the ball goes to the left).
It should be understood that in aerobatics there are large variations in speed and the pilot must constantly make corrections with the rudder.
If we analyze the reversal (stall turn) to the left, these are the actions that the pilot must take.

  • Speed ​​250 km / h
  • climb vertically while keeping the wings horizontal
    for that, as the speed decreases it is necessary to put more and more rudder to the right if not while arriving at the vertical the left wing tip will be under the horizon.
  • when the speed reaches 85 (75 solo), engage the steering (rudder) fully to the left. (as we put a lot of direction to the right while going up, there remains a large amplitude to the direction, whereas in the case of a reversal to the right, there would remain little amplitude, this is why the reversal to the right is more difficult).
    the aircraft initiates a yaw movement to the left, it is then necessary to warp to the right to counteract the induced roll and the overturning torque. It is necessary to put depth forward to counter the gyrocopic couple.
    Imagine the couples on legendary planes like the AD4 Skyraider (2800 hp) or the F8F Bearcat (2100 hp)
    How many pilots left the runway on take-off …

Thank you Tacot37 for your detailed info, again. I collect these notes. Even if I can not now put these notes into flight model, I think that we will get more documentation from Asobo or from reverse engineering.
My experience is: As long as there is “weight on the wheels (lift is less then weight)”, the airplane moves left. After take-off the airplane flies straight. The low speed airplane left tendency, cruise speed straight and high speed right tendency is not in the flight model. I have opened Zendesk request #74488.
I tested p_factor_on_yaw, gyro_precession_on_yaw. The first parameter affects 3-point position, the second parameter affects 2-point position. Both only have influence in ground contact and lift is less then weight.
I tested torque_on_roll, gyro_precession_on_roll. These parameters “roll” the airplane depending on engine power.
Test is easy: I look up the documentation and increase the parametes I think influence the behavior two times or even ten times. This should give a strong effect if my hypothesis is correct.
This process is not very demanding, but it takes time. And - for me - it is satisfying. Reverse engineering FS2020 and be able to divide the Asobo CEO Youtube video in the part “correct” and in the part “marketing”. The 1000 surfaces of an airplane MAY be true for the triangles that make up the 3D model of the airplane. The aerodynamic model of the airplane has left wing, right wing, fuselage, htail, vtail. It is possible to combine wing, aileron, flaps, spoiler geometry and the modifieres (aileron_efficiency, …) into the “current wing surface”, as it is possible to combine htail, elevator, … into the “current htail surface”. There is a tiny “appendix” in the flight model for gear. Therefore I see no evidence for the CEO statement “We finally re-worked the way gears and external parts, like external tanks are simulated instead of just hard coding the moments they apply on the aircraft they are actually physical elements in the airflow which are attached to the airframe outside of the aircraft and they now realistically simulated aerodynamically just like the aircraft and apply their own forces and moment where they should how they should”.
For me this CEO statement is partly wishful thinking. What do I see in the configuration files? Simple example: the vtail has 6 geometry parameters, the rudder has 3 geometry parameters. Many more parameters describe the “current vtail surface”, say 30 parameters all together. One fuel tank has 3 geometry parameters and 2 weight parameters. I can believe that using these tank parameters a moment calculation is possible. The gear has 2 parameters: drag_coef_gear and pitch_moment_gear. It is hard for me to believe that only two parameters can describe “forces and moment where they should how they should”.

Yes! There’s a good bug thread here about adverse yaw that can be voted on:

2 Likes

“After take-off the airplane flies straight”. Yes, but in real life if we do not push rudder right, we are in side-slip as long as cruising speed is not reached.
Have you seen the little aerobatic vidéo Aerobatic flight with CAP 10 ?

I have about 2000h in Piper singles /twins and every Cessna Single except the Corvalis and a few hours in rather rare airplanes like Beech-18, T6, and Bellancas.
For the most part I think the flight model is not too far off from the real aircraft in that class.
I think for most the real difficult part is to get their controller set up to “simulate” the input correctly. I have never found this so hard in any other sim MS or otherwise. And then I had it pretty close and Update 5 threw in a nice curve ball.

Without force feedback trimming a sim airplane is also much different than a real airplane. My beloved Saratoga is well known as one of the most stable GA aircraft.
And of course you get to know your aircraft if you fly the same one a lot. Power change and trim change are then almost muscle memory and a little tweak to fine tune one or the other. And unless it is a really rough day the next half hour or so she’ll just fly happily along that path without needing much attention.

In FS you must watch the gauges to judge where your manual inputs and the trim setting is putting your attitude and then react and correct. But since you can not feel the control getting lighter or heavier it is just guessing if you have it right.

One thing that no sim I have used got really right, aside from a Level-D I got to try with a Senior LH Captain, is the interaction between runway and wheels.
And not only friction values that change with surface quality / conditon but sideloading etc.

You will occasionally land with a little bit of crab and in the real airplanes, unless you are really off that will not be a huge issue. There will be same yaw momentum if you have a really sensitive butt and then the aircraft will roll out more or less straight depending on winds etc.
In MSFS it seems that initial sideways vector simply remains attached to the aircraft and keeps tugging it left or right depending on that initial touch down offset.
But as I said X-plane does that too…so maybe that is simply a compromise between air and ground behaviour that goes that way.

2 Likes

Most modern aircraft use a variety of techniques to mitigate the effects of adverse yaw, differential or Frise ailerons being the most common method. Some aircraft need a bootful of rudder to initiate a coordinated turn, some need none. It also depends on speed, AOA, power setting, climb angle, CG, bank angle, etc. My Baby Ace (which has one speed - slow!) needs a fair amount of rudder. The adverse yaw is noticeable, and if not corrected, can actually lead to a sideslip departure. I had a CTSW that required virtually no rudder to initiate a turn.

It would be nice to see adverse yaw (and all flight and ground dynamics) more accurately modeled in MSFS. There can still be a dumbed-down ‘gamer’ setting.

4 Likes

I would have thought the 350i would be the worst of the flight models. Torqe is not correct.

I wrote a bug report to Zendesk because I see no parameter that defines this “real airplane” behavior. This is not the first one. The normal procedure is: After days or weeks I get the state “solved”. But normally nothing I can seen has happened. The definition of “solved” seams to be: We have solved the open bugs number count by clicking “solved” button.

@PaulFalke
What flight model are you using to do your testing/adjusting? Modern or Legacy?

I only use modern flight model. For the reason see https://youtu.be/nWHi2r9zclY
Some features are just not in the classic model.

I have a reasonable number of hours on the 350. Although I find the actual ‘flight’ dynamics to be pretty good, there are some serious deficiencies in the power modelling. In the sim the only control that has any effect is the power. Prop and condition levers have virtually no effect other than feather or cutoff. Otherwise I quite enjoy the aircraft.

Realistic turboprop behavior is on the list, but nothing has been started as of last week. (See below.)

If you haven’t already, find that original thread - Turboprop Engine Logic Issues - and vote on getting the problem fixed.

1 Like

I flew the Kingair myself in real life, the Kingair in FS2020 is complete nonsense. Mainly because propeller drag is missing. It doesn’t slow down, keeps on floating. Its a joke.

Frise / differential and rudder-aileron coupling or a combinations of those are calibrated for cruise speed, anything above or below and adverse yaw will be noticeable. Because of propeller slipstream effect adverse yaw is increased or reduced depending on direction of turn.

But unfortunately FS2020 does not do a great job simulating any of those effects.

Props in FS2020 for some reason do not feather and only let the ITT shoot-up. I assume you mean condition levers on the ground as in the air they serve no purpose.

My biggest problem with the FS2020 Kingair is the lag of propeller drag. It just doesn’t decelerate and keeps floating in ground effect for ever. The real aircraft falls from the sky pretty quickly when pulling the power to flight idle.

2 Likes

Agreed. Did a fair bit of work ferrying crews into unimproved strips. Wasn’t a twin otter but dumping full flap and chopping the power was enough to peg you into the shoulder straps. Like tossing out an anchor. In MSFS I can pull off the power 15 nm out and coast to the threshold. Yup, not quite right.

ADD: yes, ground ops. No Beta. Would burn up brakes pretty quick in MSFS, even doing s.e. taxi.

As most of my time in the 350 was dirt strips in the middle of BF, we didn’t use much flair. Steep, full flap, plant it and brake. That technique works pretty good in MSFS. Fly it onto the ground rather than set it onto the ground and you can hit TD targets even with the silly GE modelling.

2 Likes