All that is wrong with the MSFS Flight Model (Inertia, Stalling, Pitch Authority, Trim & Sensitivity)

I up voted.

I’m so dissapointed about the current flight model in general (I’m pilot IRL so I know what to expect).

On all aircrafts I have this issue:

  • motionless on the ground: flight controls react immediatly to my control inputs with no lag
  • as soon as speed is alive: control response lags behind the control inputs.

Because of this, I simply don’t use the simulator anymore (I almost didn’t use it at all). I keep installing the updates and reading the dev updates but it looks like the flight model improvement is not part of their priority…

9 Likes

I do like the flight models a lot. :slight_smile:
Unfortunately this problem seems to be user specific.
E.g. I’m experiencing zero delay in the Extra, neither at 0kts on ground nor in flight.
Others are having the delayed response problem as you apparently experiencing it.
And the third group does have this problem even when sitting static on the runway.

I have read this topic with a lot of interest. A lot of sensible things highlighted here.

I am also a 737 / GA pilot and can only agree with what has been said previously.

I really hope they fix this lack of inertia / lack of realism in the implementation of the flight controls before it becomes too late (= too many addons coded on a flawed generic flight model engine, making subsequent modifications too difficult). Until it is fixed I don’t plan to use FS2020. It sure is great for the eye candy but that is all… And it is a shame. Having the airplanes fly properly should be at the core of a flight simulator. Only when this is implemented properly can you think about other aspects of the simulation.

I really hope Asobo is seriously working on those issues, not with a “walkaround based” approach, but with a “physics / engineering based” one instead.

Please upvote this thread, it is really important. We need these things to be fixed for FS2020 to become the great sim platform it could be. Otherwise a lot of simmers will just jump ship…

5 Likes

I think in terms of controller input, Joystick position should = real time 1:1 ratio with the sim yoke/control column.
Magnitude of control surfaces deflection then chosen and applied by the IRL control system type on each aircraft.

  • C152 using cables, the joystick moves 1:1 with the control column and the control column moves 1:1 with the surfaces.
  • Airbus, joystick moves 1:1 with sim control column(sidestick), amount of surface deflection then calculated by the FBW system.

Things I would like to see changed;
Sim control column/Yoke animation - match what the controller inputs are
Control input delay - Remove
Variable Surface deflection limits with airspeed - Remove
Lack of Inertia - The FPV follows changes in pitch roll and yaw too quicky causing over sensitivity.

2 Likes

If the in-sim controls move 1:1 with the joystick its gonna become more twitchy as you would be able to apply full elevator deflection for example at very high speed without feeling the increased stick forces associated with those speeds.

I think considering the joystick input as a force input would be more realistic, full deflection on the joystick is the maximum force an average pilot is able to exert on the flight controls. At zero speed the controls move 1:1, at higher speed this becomes limited by increase in stick force.

ASOBO is doing this already in some way, at higher speed, when pulling the stick full back the in-sim flight controls also move full aft, the elevator deflection however is limited at higher speeds. You can only see this in dev mode when looking at control surface deflections.

The implementation is quite poor and overdone to the point that it is impossible (at least the last time I checked) to pull any plane into an accelerated stall or maintain altitude during a 60 degree AoB steepturn.

2 Likes

I also like the “force-based” approach. Neutral joystick controller position would be the position corresponding to the least amount of force for the current trim state, and the virtual flight controls position would move with the trim setting regardless of the controller position. Which would allow to use the trim kind of “properly” on non force feedback devices.

But otherwise, your approach would be nice as well. I don’t like the kind of hybrid solution we have at the moment, especially with the delays and all…

Or maybe even better, allow the user to select which behavior he wants for his flight controls input behavior, choosing between :

  • position based behavior
  • force based behavior

Position based would be ideal for force feedback controllers, or for users who don’t want to get annoyed with the trim.

And force based solution would be ideal for non force feedback setups if you want to play with the trim setting of the aircraft.

I think it would be the perfect solution to have the choice between the two.

The current implementation is not satisfying.

2 Likes

That would be perfect!

1 Like

Twitchiness would increase but If the lack of inertia was fixed at same time to offset the increased responsiveness I think the sim would fly pretty nicely. It would force you to use a small input as speed increases, unforgiving but more realistic.

I also like the force based approach, the only issue I see with that system is, if i understand correctly, having a variable ratio between joystick displacement and surface displacement.
For example, slowing down on approach I would expect the required displacement of the controls to increase the slower the aircraft got to get the same changes in pitch and roll.

You are right, this can be an issue with this option.

However it could be circumnavigated as follows:

  • define a maximum human force variable (how hard you can physically pull on the yoke or stick)
  • if the force required to move the flight controls is greater than what is allowed by human strength, do not move the aircraft flight controls beyond that point, even if the user controller is being moved further. (There would be a limit to mimick air loads).
  • at lower speeds, air loads are reduced, you could use a greater range of your controller to move the aircraft control surfaces.

It would encourage the proper use of the trim, especially at higher speeds. At least for conventional planes.

For fly by wire aircraft such as the A320, the 1:1 position approach would work best, as the flight control computers would deal with the amount of control surface deflection, and as there is no force feedback on that aircraft.

That is true, the force required would be accurate, the control surface displacement not. If you would have roll out of trim for example, flying level would require a constant roll input on the joystick where in real life the yoke would remain more or less level and it just requires more force keeping it centered.

Its not a perfect solution, but its better than the current hybrid approach. There is no perfect solution unfortunately, unless you have a force feedback yoke.

1 Like

Wow, so I was really right and my theory that this simulator has serious problems with basic principles of laws of physics and aerodynamics. I found the flight model they created very “Arcade”, it’s no wonder the simulator was criticized for it and classified by haters as “Physical GTAV”

The a320 sometimes behaves like a sheet of paper, it seems like there’s no resistance to moving a huge body of tons…

I really hope someone from Asoboo reads your thread, because the coolest thing for me in simulators is landing, I spend hours and hours taking off and landing in the same place, and I always found the physics of this simulator very bad compared to the others.

did you add your suggestion via Zendesk? if not, i recommend it.

Thank you for contributing your vision so deeply.

I think to have a decent flying model In MSFS, asobo will have to develop a completely zero model. with the help of pilots and engineers, people who really know how an airplane should behave.

2 Likes

i love this game but the flight characteristics regarding stalling drives me crazy. the detail included by the author of this thread is remarkable and i really hope they’ve got asobo’s attention. factors like gforce need to have an impact on flight characteristics.

2 Likes

Not so much about airspeed as about AoA. Consider slow flight.

Slow flight and approach are two competely different things.

Energy management is the difference.

A plane in slow flight needs to provide additional thrust to maintain level.

A plane on approach is effectively a plane in slow flight with a negative flight path vector. So there is no need to maintain level.

However, if in full approach configuration, and ATC says stop your descent, those engines are gonna scream.

1 Like

Looks like you misread/misunderstood my initial reply.

Again, except for the Concorde I’m not awary of any civil jet which requires more thrust at e.g. Vref than at Vref +10kts.

1 Like

The elevator trim issue simply needs to be added to the model animation. A bit silly they missed it, but those that do not have an in-depth knowledge of aerodynamics (or understand the workings of these devices) would miss it. The elevator-trim action simply needs to be added to the elevator model so that it reflects the trim tab ‘holding’ it in position.

I don’t think the missed it and I would certainly not call it silly.
Since the elevator position matches the joystick position and the joystick neutral point doesn’t change with trim, it’s simply not possible to change the elevator deflection due to trimming.

It is definitely possible, but there is one problem:
The amount of elevator deflection changes with airspeed. At zero airspeed there is zero elevator deflection due to trim. At speed, it will.

A custom animation can do that. We simply need to multiply the airspeed (as a percentage) to the elevator-trim variable. So at zero airspeed, the custom animation returns zero. at 100% of that aircraft’s speed, it returns the full deflection.

Elevator (and aileron and rudder for that matter), animations can then have that custom animation added to them on a separate NLA track, and that’s something I’ll soon add to my DR250 and the Caravan no pod mods, now that this thread has reminded me about it.

Plus the amount of available elevator travel changes with trim.
If e.g. elevator trim pushes the elevator down by 50% of its travel, there’s only 50% down travel left, but the joystick still has 100% down travel.