Trim, forces and control devices in FS - explained

[robyoung]

LAST EDITED ON Jan-02-02 AT 04:18AM (EDT)[p]I've been following the Capt's interesting thread about solving the control force problem in FS. Whilst the Capt and I have not seen eye to eye in the past, I do warmly welcome this discussion as I think it may bring together many people who otherwise did not agree on some fundamental issues about Fs2002.

At the moment there are two main alternatives when designing a flight model in FS. One is to assign the aircraft absolutely full control potential, thereby allowing theoretically extreme control inputs which in normal circumstances might break off the wing of a real aircraft if attempted at high airspeed.

The other is to compromise on the above by allowing sufficient control to attain normal, and sometimes marginally extreme manouevres outside the "normal" envelope. Without decent force feedback this in my view currently provides for more acceptable and smoother control in most airliner types going about their "normal" flights, but of course this is not ideal.

The other problem is trim. Trim on a real aircraft only has one principle function: to relieve stress to the arms of the pilot, by means of a trim tab which nullifies control forces and keeps the control column in the desired position for a given flight envelope.

FS, and many other sims, are unable to truly emulate the trim tab and instead the trim routines in effect change the Centre of lift, and weight, forward and back according to settings. In effect, trim in most home sims becomes another, extra, powerful elevator.

The problems this causes are obvious. Trim becomes a way of actually directing pitch, rather than relieving arm pressure. There are other unwanted side-effects, the most serious of which is that aircraft tend to become very twitchy if given generous elevator control authority AND a full trim range.

The force-feedback mechanisms provided by some manufacturers of control devices have not yet really solved the problem, since their implimentation is crude and rather un-refined.

If FS sim aircraft are to be given absolute and real-world control surface potential, then we need two vital things:

1. A joystick which has sufficient total movement to allow the user to control with finesse and subtelty.

2. Force feedback which progressively introduces resistance, commensurate with actual forces building as airspeed and other G related forces increase.

The other thing that needs to be fined tuned in flight models, (and many of us involved in this are increasingly finding ways to emulate this), is the vertical inertia of an aircraft when for instance a dive recovery is executed. When an aircraft is pointing down (and here I do not include high performance aircraft with very high G tolerance) and increasing in airspeed, no amount of elevator is going to suddenly get that aircraft to travel upwards. It might succeed in pointing the nose up but the vertical inertia of the aircraft needs *time* to be overcome. In this cases extreme stick movements merely add enormously to the already increasing G force and this is where wings would be in danger of snapping off.

A well-designed piece of force feedback software, custom made for each type of aircraft, in combination with a joystick which had accurate control and correct forces, would go a long way to discourage the PC pilot from heavy handed control and if that was the case I would completely redesign my flight models to fit.

So in essence I am all for what Joe is proposing to achieve, and at the same time I'm posting this in a spirit of new year goodwill!

Regards,

Rob Young

[Captain_Slarty]

Many thanks Rob.. n A Happy one to you n yours.

Excellent post, and I agree fully.

There is a problem at the moment with the 'throw' range of the current sticks and yokes, even the much touted (I actually own one :-)) AETI has a pitch travel of about 4 inches - thats all, as opposed to perhaps 12 to 18 in some light aircraft and sometimes more in heavies.

The only thing that can be done at the mo for the mechanism is to control the forces, centre and 'feel' of the stick. I have actually tested inducing a slight high pitch sinewave (available in ff spec) oscillation into the stick handle that increases with innacurate placing - a sort of 'warning' to the pilot.

I used to like FU2 with its 'creaking' noises as loads increased .. whislt not necessarily realistic, it is a good thing (imho) due to the boundries of the simulated world.

If the FF routine works out ok, and from first tests it seems to, then we need to look at perhaps redesigning yokes and sticks with a force mechanism - not necessarily proprietry FF - that are just for fs fans.

because we dont want to simulate shakes and rattles, explosions etc, a far simpler and more responsive mechanism and electronics can be used.

I personally think this is the next major step in flight simulation. Airfile guys can build a superbly accuarate numbers and inertia file within the constraints of the sim and a separate feel file added for the type of controller in use.

I am currently working on some outline designs for controllers that would be affordable to the masses too, I could do with input from a production engineer as to cost estimates and engineering specs.

Joe.



http://www.captainslarty.com/images/sl0015fps.gif

[Captain_Slarty]

Just pondered that one Rob..

If you saw my post re the flybywire envelope for an airbus, you will see that it SHOULD be possible to control stick input even if the user pushes it all the way. Theoretically, the vertical inertia you talk of could be accomplished in the feel model, ie - the throw is limited for the duration of the desired action.

this does not get over the lack of throw or forces with current sticks / yokes, but certainly offers an option.

with a long throw stick yoke, it should be fully possible to load the force to a very high factor and also limit envelope if the deficiencies exist within fs's model. Interesting...

:-)

Joe

http://www.captainslarty.com/images/sl0015fps.gif

[Pace]

Joe,

Im glad we are all blowing "kisses" around at the moment :-)

I too feel that a "force Simulator" (MicroSoft Force Simulator or MSFS:-)will be a major advancement in flight simulations especially if it is widely available to the ordinary desktop flight sim guy.

Force feedback associates too much with those vibrator things which bounce around in your hands so Force Simulator is maybe a better name.

Yes we do need to have a simple trial aircraft and build a set of graphs to work from....
Just thinking that one out........

Flower power(whoops! force power and kisses to all :-)

Peter

[Captain_Slarty]

May it be with you Luke !

:-)

http://www.captainslarty.com/images/sl0015fps.gif

[robyoung]

LAST EDITED ON Jan-02-02 AT 06:38AM (EDT)[p]OK, with you Slarty.

What we need, time allowing, is an aircraft, probably not a jet but a prop twin, to serve as a test bed. The tuning can be done to accomodate the proposed FFB.

Might be a good idea to have two flight models, one accomodating the Forces and the other not, in order to compare.

Regards,

Rob

[TheFlyingFinn]

I know I'm really talking things I don't understand, but this trim thing. With the FF2 joystick it is possible to move the center the joystick wants to return to. I don't think FS2002 works that way but with a external program, can't you change the "center" of the joystick, somewhat the way real trim works. I apologize if this is what you've been talking about or impossible or a really bad idea... :-hmmm

[Captain_Slarty]

Just spoke to Peter.
He is gonna get loads of readings from the Seneca.

that should form a great start - we can ignore all 2 engine stuff for the time being, just the force table and reactions.

yep, sounds good.

will test it first on a standard ff stick, but as an aside..

I have just ordered some motor/gearbox assemblies for an FF Yoke. I have a simple design that should work well, also, I have a couple of genuine Piper yokes I can use one for the test rig.


Joe
http://www.captainslarty.com/images/sl0015fps.gif

[Captain_Slarty]

yes it is, I do believe that fs attempts this anyway, ?? it is just not very good at it - ie, there should be resistance to return - even though it is returning to a position, it should not just fly back, it needs to be controlled in some cases as loads are removed from the flying surfaces and applied in an opposite direction.

http://www.captainslarty.com/images/sl0015fps.gif

[ytell]

a few comments:

- In order for a joystick to emulate the real thing it is not very important that the physical movement would be the same. Accurate forces, smoothness and accuracy are much more important.

- I don't see the importance of taking into account the inertia of an AC as part of joystick emulation considerations. It is really a matter of what the a/c model will do as a result of a stick input. (BTW it is emulated pretty good in the finest AIR files)

- Trim forces - again, if a simmer uses the trim button as a pitch command rather than stress relief, it is only an indication of flying habits rather than a criteria for a good/bad joystick. I really think that emulating the real forces on the stick is the biggest challenge.

I addition, it would be very useful if those AIR file experts would deal with one subject that has left behind for a long time - drag due to asymmetric flight (slip)

To make the initiative above more of a good engineering practice, I would like to recommend that you first agree on a set of requirements you're going to deal with.

All the best

Yoav

[Captain_Slarty]

Hi Yoav.

>a few comments:
>
> - In order for a
>joystick to emulate the real
>thing it is not very
>important that the physical movement
>would be the same. Accurate
>forces, smoothness and accuracy are
>much more important.

I agree to a certain extent, I believe, for a yoke, the throw (pitch) needs to be far more than is currently available.

>
> - I don't see the
>importance of taking into account
>the inertia of an AC
>as part of joystick emulation
>considerations. It is really a
>matter of what the a/c
>model will do as a
>result of a stick input.
>(BTW it is emulated pretty
>good in the finest AIR
>files)

Rob is far more qualified in this area, and I hope will chip in. I am under the impression that one cannot obtain everthing from the FM and compromises need to be made. Control of throw (range) at certain times can be a useful addition to the overall performance.

>
> - Trim forces - again,
>if a simmer uses the
>trim button as a pitch
>command rather than stress relief,
>it is only an indication
>of flying habits rather than
>a criteria for a good/bad
>joystick. I really think that
>emulating the real forces on
>the stick is the biggest
>challenge.

I cannot control HOW people fly :-)... the trouble at the moment is that yo CAN fly with the aircraft way out of trim due to the lack of semi realistic forces. In the real aircraft this would become very tiring very quickly, the more force and feel you can add, the easier it is to eliminate incorrect habits.
What I would hope for is the feel of an 'incorrect' situation to be passed to the user as it would in the real aircraft, what they then do with it is beyond our control...

>
>I addition, it would be
>very useful if those AIR
>file experts would deal with
>one subject that has left
>behind for a long time
>- drag due to asymmetric
>flight (slip)

Again, Rob is the best man to answer this.

>
>To make the initiative above more
>of a good engineering practice,
>I would like to recommend
>that you first agree on
>a set of requirements you're
>going to deal with.
>
>All the best
>
>Yoav

Thanks Sir.

I definitely know what I want to achieve,ie - as much realism in the feel department as possible. There is a solid barrier with conventinal sticks / yokes, and certainly more compromises would be needed here, but with a custom design, most if not all I want can be accomplished. A rotation of aileron axis of 340 degrees, a pitch throw (yoke) of 12" (As compared to the current 3-4")
A max loading of apx 4-5 Kg (10Lbs) in pitch and 6 pounds in roll. Smooth control of forces and centering, and no violent opposite reactions.

;-)


http://www.captainslarty.com/images/sl0015fps.gif

[Captain_Slarty]

in case ya missed it. this is more up your street Sir.

Joe.

http://www.captainslarty.com/images/sl0015fps.gif

[robyoung]

Yes, the "no violent opposite" reactions is the key. Too many of the FF stick makers are over-concerned with feeding back skirmishes over bumps rather than getting the correct feel. That's understandable when you consider most FF users are probably playing driving type games (sorry "sims"!).

We need a dedicated FLYING FF stick/yoke. Joe, please check you e-mail.

Rob

[LTEtabs]

I like these ideas a lot. It would add tremendously to the realism of the sim. You guys could probably make a whole boatload of $ too if you sold this idea and design to MS and CH or Saitek...

[lnuss]

One thing to keep in mind is that the F-16, for example, has almost no movement of the stick. According to a friend of mine who flies them, they originally had NO movement, but it made it tough for the pilots, so they added a small amount -- maybe a quarter inch or a bit less (or did he say 1/16th inch).

So the actual AMOUNT of travel is not nearly so important as the forces that come back to your hands. A Cessna yoke has a travel of about 90º each direction for aileron control, as do many others, but in a Grumman-American Tiger or Cheetah, you get maybe half that.

In a Cub, you have maybe a foot, or a bit more, of aft travel from a level cruise flight position of the stick to hitting the aft stop (at the top of the stick) and you probably get a similar amount pulling the yoke back in a Cessna. These are guesses from memory, not actual measurements.

All this to suggest that you be mostly concerned with forces, not control travel. I note that you DID specify that there would need to be config not only per aircraft, but per control device type, as well. I fully agree.

If you are able to succeed in this endeavor, and it's sufficiently inexpensive, then I hope you will look at doing the same for feedback on rudder pedals and toe brakes (Hmmm..... could you simulate heel brakes too?), since those forces are (to my mind) equally important.

Another thing to .air file "fixers." While you want to be able to fly the aircraft smoothly with existing devices, you need to have a reasonable amount of control authority available, and people can (if they will) use the SHIFT-Z "G-meter" to avoid overstressing the aircraft.

Larry N.

http://www.flightsim.com/dcforum/User_files/3c22868f655d5d5a.jpg

[Pace]

LAST EDITED ON Jan-02-02 AT 07:47PM (EDT)[p]Larry

BTW thanks for the E mail I might have something just up your street (patch of Sky:-)

I was staggered with some tests I did on the real world Seneca five where at cruise speed, 3000 feet and 155 kts ias I pitched to ten degrees the initial force was quite considerable but what really surprised me was that the yolk movement was a very tiny 1.5 cm (yikes)while at the other end of the scale ie flairing to land just above the stall the yolk came back 10 -12 inches.

I agree with you on the rudder pedals which would need their own force simulation device.

I love that banner of yours its really bright and cheerful and I also had a quick look at your site and history (my kinda guy)
Heres a shot taken from the well travelled seneca over Spain a couple of months back

All the best

Peter

[Pace]

LAST EDITED ON Jan-02-02 AT 07:53PM (EDT)[p]Larry :-)whoops try again with the pics

[lnuss]

Will the third time be the charm? (:^)))

Larry N.

http://www.flightsim.com/dcforum/User_files/3c22868f655d5d5a.jpg

[Pace]

oh well one more go ;-)

http://www.flightsim.com/dcforum/User_files/3c33ac336803293c.jpg

[lnuss]

Yes, it's been that way in most aircraft I've flown. Small motion and large force at higher speeds, lower forces and longer travel at lower speeds. True of all three axes. Which is a lot of the reason we instructors usually talk about control PRESSURES, not control MOVEMENTS.

Pilots fly by control pressures -- that seems to be what the muscles "remember" best.

Larry N.

http://www.flightsim.com/dcforum/User_files/3c22868f655d5d5a.jpg

[lnuss]

Aha, the third time WAS the charm. Great picture. Looks like Spain has the same haze problem as much of the U.S. Pretty, though.

Larry N.

http://www.flightsim.com/dcforum/User_files/3c22868f655d5d5a.jpg

[Pace]

and a nice sunset

http://www.flightsim.com/dcforum/User_files/3c33acca6a29c17a.jpg

[Spongey87]

On thing that I hate about my joystick (Logitech Force 3D) is that the movement is not smooth. It is in small "bumps" which makes precision dificult. For example, on descent, you pull back a little, and you feel a bump. But the you notice you are too nose-high, so you let go of the joystick a little bit. But you are too nose-low, so you pull back, but the exact point you want your joystick at is not atainable, as the "bump" is on it, making your airplane look like a happy, bouncing bunny. I hope I was clear enougn, as it is difficult to describe. I dunno if it is just my joystick, or if it is like that in all the FF joysticks.

FF joysticks: Can't descent with them, can't live without them. Sighh...

Spongey

[flyinggriffin]

Larry,
I put in about 12 hours on a FRASCA simulator this spring and the control system while pretty smooth isn't much better than a home flight sim. Without any control forces it was quite easy to overcontrol.
My pet peeve is with trimming. You can trim out a real plane without looking or having to let go the yoke. With sims, you learn by guess and by golly how much trim to get a rough trim then have to keep letting go the yoke to see if you are level. I get the feeling my plane looks like Flipper going up and down untill I get the right amount of trim dialed in. It's a whole lot easier to trim out an airplane in real life than in the sim.
Funny thing is, that the older radio control model transmitters used to have a trim lever that moved the whole gimbel assembly to trim the model. This had the same effect as on a full size airplane, but went out style with newer transmitters. Perhaps something like this would work on flight sim yokes. You would just need enough throw on the yoke.
HEEL BRAKES!!! Yikes, I did a couple hours in a Cub and it's hard to break habits.

[ladamson]

> My pet
>peeve is with trimming. You
>can trim out a real
>plane without looking or having
>to let go the yoke.
>With sims, you learn by
>guess and by golly how
>much trim to get a
>rough trim then have to
>keep letting go the yoke
>to see if you are
>level.

I use two buttons on my joystick for "simulated" elec. trim.

I suppose my mind just about senses when the trim is very close. I'm holding the stick to my trimmed position, whether a climb, level, or descent, & the planes nose moves up or down very slightly when I've met or exceeded my trimmed position. Neverless, my mind has "faked" it quite well to feel like the real thing.

Ladamson

[lnuss]

On my Saitek X-36 Combo, there are a couple of rotary wheels that can be configured as control axes. Each wheel also has a center detent. One of them I have set to elevator trim. This makes it MUCH easier to fly smoothly in the sim.

Given that I am actually moving something physically, I can get a quick sense of how much trim I have added, and of how much travel is left on the trim, so I use it a LOT more than I did when I was limited to keys on the keyboard, and I fly more precisely because of it.

Joystick buttons are probably the next best thing, but you can't tell anything about how much you have changed it except for the response of the stick/yoke. Worst is hunting for the keys on the keyboard -- a major distraction, especially on short final or flying formation.

As to heel brakes, they are certainly very awkward to learn to use well. The Aeronca Chief (11AC) in which I learned to fly had heel brakes that were actually mounted on the rudder pedals. But all the other aircraft I've flown with heel brakes (Cub, Champ, Great Lakes, etc.) have them floor mounted, making for an awkward contortion of the foot, especially bad in the Great Lakes, since it was built for someone about 5'2" and 120 lbs. and neither the seats nor the pedals are adjustable. I exceed those figures quite a bit, and so after 20-30 minutes I had to land and get the kinks out. Otherwise a delightful aircraft and the most gentle taildragger (on the ground) that I've ever flown, a good thing, since it was twice as awkward to get to the brakes as on anything else, because of seating position.

Larry N.

http://www.flightsim.com/dcforum/User_files/3c22868f655d5d5a.jpg

[flyinggriffin]

I use a yoke with rocker switch set up to trim like the electric trim in a couple of airplanes I fly.
Loved flying the cub. First trick was getting in it, I am 6-1. I suppose that if you do it enough times a graceful entry and exit can be made. After an hour of air work and pattern work....well lets say that I am glad my wife didn't take any pictures of me trying to haul myself out of it.

[lnuss]

Few people can find a graceful way to get in a Cub. Some folks are able to stand just in front of the wing strut, or even on the tire, place their butt on the seat, then swing their feet in. For me, it's right foot on the step (behind the strut), hold on to the overhead crossbar with one hand, and a windshield bracing strut with the other, swing my left foot up and over the stick while pulling myself up with both hands, then swing the right foot over the door sill, then let myself down into the seat.

I've also seen quite a few other variations by veteran Cub pilots. They're fun to fly, but you better know what the rudder is for, and after an hour or so, I need out. The seating is great for a few minutes, then it starts getting worse. Certainly not desirable for x-country work, though it's do-able.

Larry N.

http://www.flightsim.com/dcforum/User_files/3c22868f655d5d5a.jpg