Yokes that work better (brain storm thread)

[Jeroen Hoppenbrouwers]

Maybe this is slightly off-topic, but:

Aerowinx PSX is the pinnacle of desktop/home built flight simulation. Basically Level-D software, the only one that is accessible not just by professionals, but enthusiasts as well.

But when are we going to see a USB column yoke that simulates the forces and feel of the real yoke on the 747? For years I've been waiting for something like this to come out, but we are still stuck with the spring loaded/gas damper ones.

Although perhaps the forces would be so strong that mounting this kind of yoke would be problematic unless you can bolt it to some concrete floor...

Moved from another thread.

[Jeroen Hoppenbrouwers]

I think the only reasonable way to create such a yoke would be to bolt it down on a ground plate that you can put a seat on. No need to bolt it into the concrete or wooden floor, but you need something to counteract the pitch forces and with an arm of a meter those are considerable.

Real yokes also pivot below the floor, so a realistic feel requires serious work.

Can the heavyweight constructors under us chime in? What did you learn about building your own yoke system or adapting a real yoke to your sim pit?

Hoppie

[emerydc8]

+1.

I think the same applies for the rudders. Not only is there no reasonably affordable simulator rudder set that duplicates the throw of the real Boeing (even FDS falls short here), the control  input required on the real airplane is much different than on a set like the Cirrus PFC rudders and yoke I'm using. It's a bit misleading because anyone who thinks they've got the feel of the real plane because they can do it on their PFC set at home will be in for a rude awakening if they ever find themselves in the actual airplane. It puzzles me that the manufacturers don't strive to come closer here, even if they had to charge two or three times the price for a premium set. I would pay it.

Jon

[cavaricooper]

Jon-

I thought you purchased some other home brew set of pedals you liked better? I use PFC Boeing Pedals and agree they are far from adequate. For controlling a GA aircraft they are close to passable (notice, not great) but are hopelessly dissimilar to anything Boeing. I was hoping that the other set had merit, but reading between the lines it would appear they were no better...?

Best- C

[skelsey]

Real yokes also pivot below the floor, so a realistic feel requires serious work.

I think Gary still has nightmares about his experience digging a trench in the world's most over-engineered garage floor for his yokes...

[Panos Bilios]

Hi

I have a question to our experts

How the pilot in manual flight knows from the yoke
that the plane is out of trim, and how the yoke moves
when trim is applied

Thanks
Panos

[Jeroen Hoppenbrouwers]

There's a difference in trim technology between axes.

The aileron and rudder trims are done by making the neutral center point move, which means the yoke and pedals actually move. This is done "upstream" by ingenious mechanisms; the actual flight control surfaces are moved (on the 744) by hydraulics and there are no mechanical links all the way to the surfaces. If you feel your yoke/pedals are not centered when you take your hands/feet off them, you know it's trim.


The elevators are not trimmed. Instead, the whole tail plane (horizontal stabilizer) is pitched. This gives much more range and effect, while not significantly altering the drag, and not putting constant strain on the elevators. The yoke also does not move.

To give the pilot some feedback of the trim situation, the counterforce of the yoke is changed. I believe the yoke becomes heavier to operate when the trim changes more to tail-heavy, to avoid pilots accidentally overcontrolling the aircraft in pitch. I may be wrong here, careful.

An out of trim stabilizer is directly noticeable because you need to keep a force on the yoke to remain at the same pitch.

[Hardy Heinlin]

To give the pilot some feedback of the trim situation, the counterforce of the yoke is changed. I believe the yoke becomes heavier to operate when the trim changes more to tail-heavy, to avoid pilots accidentally overcontrolling the aircraft in pitch. I may be wrong here, careful.

Almost correct :-) A tail-heavy aircraft has the CG more forward aft. So the tail needs to generate more lift to keep the tail up. This is nice for the overall performance as the total lift increases and you can fly at higher altitudes. The pitch control is less also more effective. Imagine a triangle of force vectors and apply the sine function: The effective pitch angle is alpha. The vertical opposite leg is the tail's lift vector, and the horizontal hypotenuse is the distance from the CG to the tail.

sine of alpha = opposite leg / hypotenuse

The longer the hypotenuse, the smaller is alpha.

The elevator feel computer doesn't know the CG position. But it knows the stab trim position. When the pilot has set the stab trim very much nose down, the elevator feel computer can conclude the CG is very much forward aft.

When the CG is more aft, the elevator feel computer will command an increased yoke centering force, as the elevator is very effective (short hypotenuse, greater alpha).

When the CG is more forward, the elevator feel computer will command less yoke centering force, as the elevator is less effective (long hypotenuse, smaller alpha).


Cheers,

|-|ardy

[emerydc8]

Jon-

I thought you purchased some other home brew set of pedals you liked better? I use PFC Boeing Pedals and agree they are far from adequate. For controlling a GA aircraft they are close to passable (notice, not great) but are hopelessly dissimilar to anything Boeing. I was hoping that the other set had merit, but reading between the lines it would appear they were no better...?

Best- C

I was going to buy a home-made set from a forum member, but he never shipped them, so the search continues. The PFC controls will do the job if there's no other option, but they fall woefully short on both throw and pressure required, compared to the real plane. They are not a true representation of the real plane -- not even close.

[ahaka]

So I think it's clear that an ultra realistic yoke would have to be motorised for force feedback. And then a software would need to control the motor according to the simulation state. The motor and mechanics would have to be extremely smooth, most likely a direct-drive solution and not a belt-driven one. Perhaps easier said than done, but all this technology is available so it should be possible.

But whoever would create such product, needs to have inside information on how the real yoke feels and behaves under different conditions. Maybe this is one of the reasons these kind of products dont exist. Everyone knows how a car steering wheel behaves, but only few people know anything about a 747.

I'm sure if a team of skilled individuals from this forum would join forces, they could pull it off.

[Jeroen Hoppenbrouwers]

Hardy, I don't get it -- I think just upside down of your description?!



Shorter arm = more force required?


Hoppie

[Hardy Heinlin]

Does this picture explain it better?

Edit: New picture; the previous one was misleading ("CG" should've read "CP").

[skelsey]

Almost correct :-)

<snip>

I'm sure you have better information on this than I do, Hardy, but I didn't think it was even as complex as that - I thought the elevator feel computer worked purely on dynamic pressure (ie airspeed) - so higher airspeed = more pressure to the elevator feel = stiffer controls.

Aft CG is typically associated with reduced longitudinal stability, increased manoeuvrability and a reduction in fuel burn (because the stab is normally designed to produce a downward force to balance the aircraft, does not have to produce such a large down force and therefore induced drag is reduced).

The opposite is true for a forward CG - increased longitudinal stability, reduced manoeuvrability and increased drag and fuel burn.

[Hardy Heinlin]

Oops, you're right, of course: The tail needs to generate more lift when the CG is aft, not forward.

The rest of what you wrote (re manoeuvrability) agrees with what I posted above.

Re elevator feel computer: The stab trim data is just one parameter. It also takes the EAS into account (equivalent airspeed). The higher the EAS, the higher the centering force. There are certain curves of how the parameters are mixed for the final elevator feel command. This is modeled in PSX as well and can be monitored on the Analysis / Misc page. The data can also be read from the PSX network and can be used for external yoke hardware.

As the aft CG provides increased manoeuvrability, the centering force is increased even more, in addition to the EAS factor.


|-|ardy

[farrokh747]

hi

Using all available data i could find everywhere (not a lot), I made a chart of what the stick force should be vs the psi supplied to the elevator feel unit/actuator

http://farrokhchothia.com/slists/elevfeel/FEEL%20TEST.pdf

I have the oem 744 feel unit below the floor, with a variable hyd press valve working on values from psx -

there is a slight difference in my stick values from the known values due to rigging,moments, etc - but im working on biasing it as close as possible

Ive used the Elev UP values to calibrate, assuming if this is correct, Down will be too....

The Aileron and Rudder units are OEM as well, but since they're static press, easier to setup

The rudder trim unit also has the trim actuator, which feeds the trim gauge directly - so it works well for both - nice to see the pedals move off center as you trim.. 

http://farrokhchothia.com/slists/elevfeel/IMG_5173.MOV   rudder moving

http://farrokhchothia.com/slists/elevfeel/pics/   images

cheers

fc

[skelsey]

Oops, you're right, of course: The tail needs to generate more lift when the CG is aft, not forward.

Not quite; the tail is generating less lift with an aft CG.

The stab is always producing a downward force on the tail ('negative' lift, if you like). This is necessary to balance out the nose-down couple generated because the CofP is (in all conventionally-designed aircraft) located behind the CofG.





The further foward the CG is, the greater the nose-down couple and therefore the nose up (tail down) force exerted by the stab must increase = more 'negative lift' from the tail = more induced drag.

With an aft CG, however, the nose-down couple is smaller and therefore the stab only needs to produce small downward force:



The fact that the stab is always producing a downward force is what gives the aeroplane its longitudinal stability; if a gust disturbs the aircraft in a nose up sense, as the nose moves up the stabiliser moves down, meets the air at a greater AoA and produces a corrective force, and vice versa.

Hoppie added width=500 to the image code

[Jeroen Hoppenbrouwers]

Once you draw stuff with both weight, lift, and tailplane negative lift, as you did, you see that the arm of the tailplane to the center of gravity is basically constant compared to the arm of the lift vector to the center of gravity.

As you said, the CG is always forward of the lift vector (for traditional airplanes). "Aft" and "forward" are correct, but need to be seen in a quite limited context.

It's that relatively short arm that determines the outcome. If very close together, the aircraft is nearly in balance by itself and the tailplane does not need to work hard to change pitch -- so better increase the yoke force else the pilot may overcontrol. If a bit further apart, the tailplane needs to keep more downward pressure to balance the plane and work harder to change pitch, so more elevator is needed for a change and yoke forces can be decreased to assist the pilot.

Correct now?

Hoppie

(I believe one of the special aspects of the MD-11 is that it is designed with the CG more aft than usual, as this improves flight economics and maneuverability, at the expense of subjecting pilots to more chance for overcontrol?)

[Hardy Heinlin]

Skelsey, I'm simply trying to agree with you by saying that a tail-heavy aircraft needs more nose-down counter-force.

When the tail goes down, compensate that by pushing the nose down.

When the tail goes up, compensate that by pushing the nose up.

By up/down I mean the angular movement at the outer tip of the respective arm. It's a rotational movement.

Lift and gravity relationships are everywhere. In this sense, "more lift" is an increment, and, in my opinion, this increment can be described with words like "more or less upward force" or "more or less downward force". It doesn't change the mathematical meaning of the vertical force vector at the tail. Do you see what I mean? :-)

if a gust disturbs the aircraft in a nose up sense, as the nose moves up the stabiliser moves down, meets the air at a greater AoA and produces a corrective force, and vice versa.

That's right. But this effect also works when the stab has a neutral installation angle, or even a slight lift (like in more modern aircraft). It's approximately the same stabilization effect that a dart tail provides, or that an aircraft's vertical stabilizer provides when yawing.

[Hardy Heinlin]

It's that relatively short arm that determines the outcome. If very close together, the aircraft is nearly in balance by itself and the tailplane does not need to work hard to change pitch -- so better increase the yoke force else the pilot may overcontrol. If a bit further apart, the tailplane needs to keep more downward pressure to balance the plane and work harder to change pitch, so more elevator is needed for a change and yoke forces can be decreased to assist the pilot.

Correct now?

Yep.

[Hardy Heinlin]

hi

Using all available data i could find everywhere (not a lot), I made a chart of what the stick force should be vs the psi supplied to the elevator feel unit/actuator

http://farrokhchothia.com/slists/elevfeel/FEEL%20TEST.pdf

I have the oem 744 feel unit below the floor, with a variable hyd press valve working on values from psx -

Wow! So you're now really controlling your valve by the elevator feel data output stream from PSX? I guess you're the first :-)


Cheers,

|-|ardy