Hope this makes sense, and someone actually reads that in an already long thread and neverending debate.

 

I read the whole thing, and I think you make some great points. Pilots can instinctively keep a plane level because, among other sensations, they can feel the yoke wanting to deflect one way or another. They're flying level, the yoke starts tugging to the right, so they instinctively compensate by turning to the left. Having never flown a real plane before, I can only guess, but I wonder just how many of these minor, instinctive corrections happen subconsciously? Then when they fly a desktop sim, they no longer have their subconscious instincts working for them, and suddenly the plane is "unrealistically" rolling left and right.

 

The tricky part is hitting that magic middle ground where everybody is happy and you don't have one person complaining that the aircraft is too stable and another complaining that it's not stable enough. Maybe Laminar could consider adding a simple slider for adjusting aircraft stability instead of asking us to muck around in Plane Maker -- of course the problem there is that one setting won't necessarily work for every aircraft, so maybe it could be defined on a per plane basis?

 

Just throwing some ideas out there.

Good thoughts / suggestions here by all.

 

It really helps when people that know the inner workings of Xplane chime in, because lets face it, 99% (?) of us haven't got a clue what were doing or adjusting in Plane Maker. It's a fantastic tool to have available, but a little daunting to say the least.

 

When guys like Tom get involved with Plane Maker suggestions, it allows the average user to (hopefully) configure a given aircraft to their own liking, and then everyone is happy. Like Goran has always said, the tools are there, you just need to know what your doing. Now I do see the other side of it, where they don't want you messing around with a Payware aircraft. I suppose that would be fine as long as the end user knows that you do so at your own risk and don't expect support for your mistakes. With the default aircraft though, this can be a great learning tool for the average user.

 

 

Glen

I read the whole thing, and I think you make some great points. Pilots can instinctively keep a plan level because, among other sensations, they can feel the yoke wanting to deflect one way or another. They're flying level, the yoke starts tugging to the right, so they instinctively compensate by turning to the left. Having never flown a real plane before, I can only guess, but I wonder just how many of these minor, instinctive corrections happen subconsciously? Then when they fly a desktop sim, they no longer have their subconscious instincts working for them, and suddenly the plane is "unrealistically" rolling left and right.

 

That's not the way it is. Even a helicopter that's more like controlling a marble on an inverted plate, will settle into a trimmed state during cruise flight. It's not a case of subconsciously moving the stick in several directions to maintain a level flight. Airplanes as a general rule won't fly as on an auto-pilot. You will be making some corrections, as they'll drift off heading or altitude, especially if upset from air currents above dead calm. However, it's nothing like that marble & inverted plate sense of feel & reactions.

 

Geoff nailed it with the "hunting" label. It's basically the reason I don't fly X-Plane for long flights. I have no wish to constantly try to maintain spacing with the monitor's horizon just to get the simulated airplane to remain within so many feet of desired altitude. I also believe it's the reason that some coming to X-Plane from that other sim............... seem to think that XP is more of a challenge & possibly the more correct flight dynamics.

 

L.Adamson

That's not the way it is.

 

I think that it is Larry. What we are dealing with here is equally cognitive and physiological as it is the flight model. I have flown a lot of long flights without autpilot in real life and it is mentally tiring...because you are always making sure you're on course and altitude. Flying is busy business. You say you have no wish to continully monitor the horizon...yet my flight instructor tells me to continually monitor the horizon (or the instruments). I think Mountain Man is also exactly right. Folks are not aware of how instinctive they react to physical phenomenon, you don't think about it in a real plane.....you have peripheral vision, g-forces, things that just make you do the right thing while you talk on the radio...courtesy of decades of experience in balance, interpretation and vision. If I was to put stock into the greater "effect" as to XPs flight behavior, I'd place more weight on human perception than the flight model.

 

Geoff. From my perspective, I can control the XP Baron in a trimmed condition beautifully so I am not sure what you are seeing (that I can measure) that makes you say different. I'm not saying you're not seeing it, just that I don't know what it is. If it's just a 'feel', I unfortunately cannot work with that beyond a certain level....close? yes I believe so, but if your cognitive interpretation of your sim experience is the final gap, then I just won't ever be able to reach that on my end and you'd have to tweak from your end. I have made a few more Baron adjustments yet that you haven't tried based on your feedback and I would like you to give that a go soon. I haven't put it up yet, but will put a link to it when I do. I find your descriptions of "flight behavior" the most helpful though. I can't work with "it's not right" as much as I can "it rolls too much with quick yoke inputs". I'm happy to make changes and describe them though if it helps.

 

Tom K

Well, I think I have learned a lot about trim stability using the DCS p51d :-) I have to accept it is realistic, not only because it has been explained by the lead designer but also thanks to the comments of some RW p51 pilots participating at both the DCS and A2A forums. Being that way, sometimes I think that all airplanes in X-Plane are now too easy to trim.../ control...

 

Regarding X-Planes, I still believe (a few have told it to me...) that probably when an aircraft is designed with correct, as realistic as possible inertias (translated into radii of gyration in PM) and aero data, it should behave rather acceptably. Of course we can use control phase-out and even stability augmentation, but while the first is more acceptable to me in the absence of proper force feedbak for the hinge forces, the later is something I did use but regard as "cheating" at least in as far as a vanilla GA aircraft goes.

 

I use the LES dc-3 as a good example of an aircraft where I beileve the author didn't use any of the two (control phase out or stability augmentation) and yet it felt and reacted to control inputs, power settings, etc... in a very plausible way, as opposed to most of the initially released default aircraft (at least the C90 feels a lot better on the latest betas...).

 

A good design, with the closest to real data will probably work out as very nice / realistic handlig characteristics, I believe.

Mountain man is wrong . From the iphone

Mountain man is wrong .

I think Mountain Man is also exactly right.

 

:Thinking:

I think that it is Larry. What we are dealing with here is equally cognitive and physiological as it is the flight model. I have flown a lot of long flights without autpilot in real life and it is mentally tiring...because you are always making sure you're on course and altitude. Flying is busy business. You say you have no wish to continully monitor the horizon...yet my flight instructor tells me to continually monitor the horizon (or the instruments). I think Mountain Man is also exactly right. Folks are not aware of how instinctive they react to physical phenomenon, you don't think about it in a real plane.....you have peripheral vision, g-forces, things that just make you do the right thing while you talk on the radio...courtesy of decades of experience in balance, interpretation and vision. If I was to put stock into the greater "effect" as to XPs flight behavior, I'd place more weight on human perception than the flight model.

 

Geoff. From my perspective, I can control the XP Baron in a trimmed condition beautifully so I am not sure what you are seeing (that I can measure) that makes you say different. I'm not saying you're not seeing it, just that I don't know what it is. If it's just a 'feel', I unfortunately cannot work with that beyond a certain level....close? yes I believe so, but if your cognitive interpretation of your sim experience is the final gap, then I just won't ever be able to reach that on my end and you'd have to tweak from your end. I have made a few more Baron adjustments yet that you haven't tried based on your feedback and I would like you to give that a go soon. I haven't put it up yet, but will put a link to it when I do. I find your descriptions of "flight behavior" the most helpful though. I can't work with "it's not right" as much as I can "it rolls too much with quick yoke inputs". I'm happy to make changes and describe them though if it helps.

 

Tom K

 

Tom -this is a great conversation. I used to fly 12 hour round trips in the Baron from Michigan to Baton

Rouge to pick up my son from college-and oddly when I owned the Baron I never, ever (except in practice) used the autopilot. It was just to fun to hand fly, and I have always personally considered autopilot time not really flying. I was tired after these flights-but with flight sim I can feel that way after a few minutes (and I am talking all flight sims presently). For the drivers out there-check the next time you are on a highway how many adjustments to the steering wheel you are doing rapidly without thinking. You have a few more axis's on the plane-but this also becomes instinctive after a while.

 

Considering fm-from my standpoint there are several components:

 

1) Numbers. When vfr pilots make the transition to ifr, at least my first lesson was to take the plane out, put it in various configurations and power settings, copy the results down and memorize the numbers. e.g. in my Baron clean 15" 2400 rpm in the winter and 17" 2400 rpm in the summer takes airspeed right down to flap range-drop the gear and with small trim adjustment 500 fpm descent-makes approaches quite easy.

 

2) Control feel -Yes perhaps difficult to do but it seems doable. The improvements you made on the Baron do capture its control "heaviness". A feeling of momentum. The feel of the controls from a trimmed state, that slight fighting and delay to a change of control. Again-your default Baron is capturing this very nicely now. It was and still is one of my strongest critics of some xplane models-a feeling of the control being attached like a rubber band, instant response and way to rapid, and no momentum. When xplane 10 came out and I tried the carenado bonanza this was the first time I had felt this in xplane and is mostly what kept me with this version

 

3) Stability: For those unfamiliar here is a great succinct article:

www.allstar.fiu.edu/aero/axes33.htm

 

4) instrument effects sounds etc. After all-with no seat of the pants feeling-the only way we can judge if our little spring loaded joystick did something is how the instruments, sounds, and out the window view changes.

 

#3 is the big one for me that still needs work. When we had a thread about stability a year or so ago I remember posting the faa requirements-cant seem to find them now...however...

 

With a light GA aircraft with no turbulence/shear etc. if you trim it-it should stay pretty much where you left it. If a light gust of wind comes along-thru a series of subtle oscillations it should find its way back to its initial trimmed condition-called Dynamic and Static stability.

 

In a typical Cessna or Piper for instance in a trimmed condition-you should be able to put it into a shallow bank, let go-and the plane will come back gently to level. Pull up or down on the stick, let go-and the plane will thru some gentle gyrations come back to where it was. Try these tests on any sim aircraft-does it show this Longitudinal,Lateral, and directional stability?

 

There is a famous story of a non ifr priest who got into ifr conditions. Instead of trying to fly the plane-he started praying. When he came out of the clouds some time later he was fine-the plane was trimmed and did a fine job of flying itself-most who are not ifr rated that get in this situation would be better off to follow the same.

 

I can do some testing, and even (shock) compare for instance the fsx Milvitz b55, the xplane Carenado Bonanza, and the default xplane Baron, and their relative stability.

 

In fsx-it seemed quite stable in the roll dept, the yaw department was pretty much non existent (yay xplane for dutch rolls which is much more realistic) but the pitch was always way to sensitive.

 

In xplane I find all axis need work-and like Larry I get frustrated after a short amount of time because a real aircraft is 10 fold easier to work with. I feel I am always fighting to maintain a heading and altitude in even calm conditions-add some wind/turbulence and I can lose control of the aircraft.

 

A quick try on the default xplane Baron I just did-with no wind and no turbulence.

Level off at 2000 ft., set my power to 24/24. Trim the pitch -hands off. In 3 seconds it is in a 45 degree bank getting worse. Now I could use rudder/aileron trim-but in the real bird you don't-and I assume we are chasing reality?

 

Pitch-within 2 seconds of getting trimmed and letting go of the control I am either climbing +_1000 ft. a minute-even going to 2000. Again-this would not happen rw-even with turbulence.

 

The plane in no way, shape, or form tries to get back to its initial trimmed state.

 

So unlike the video I posted above-I find I am fighting the controls in an unrealistic fashion all the time and precision flying which I pride myself on is pretty much impossible. Add some wind/turbulence and sometimes I can't even control the thing.

 

Again, I flew about 600 ifr right seat hours with a friend who had an Arrow with the Hershey Bar wing-probably one of the most stable wings ever made.

 

He would get to our assigned altitude of say 4000 ft. trim it-and then fly hands off the rest of the flight except in extremely turbulent conditions which was rare (who wants to fly in that?).

 

Yes it used to bug me to see his altitude rise to 4050-4075 and then slowly self correct-fall back below 4000 and then come back up because I am of the Ernest

Gann school and don't like to see my altitude even 20 ft. off. But it did work, and he always bragged he did not need an autopilot as the plane was his autopilot-and he was right. The heading was not quite as stable-but depending on conditions it would take quite a while before the heading was more than 10 degrees off.

 

Now once again-the Carenado aircraft were the first I experienced in xplane that had a better #2 and #3.

 

...and I'd like to state I was always unsatisfied with the general pitch axis on fsx-but I was/am able to control a typical fsx aircaft in a more precise realistic fashion than I am an xplane right now.

 

I hope improvements can be made in this area-because I do think in landing/takeoff phases (other than the screaming brakes-can I cast a vote to get rid of that pronto!) in xplane is much more credible. I also think the "feel" of airflow in cruise flight simulates more of a "feel" of actually flying than fsx does.

 

I hope this clarifies at least what I am after. I have used all sims to practice for whatever rating I was going for at the time in the past. Now-I primarily just like to use it to keep ifr skills cobwebs off. I'd sure like it to be more stable so I could start taking those 12 hour flights on the sim B)

they can feel the yoke wanting to deflect one way or another

 

Mountain Man, from my experience in gliders, where a seat of the pants experience and sensibility (having a sensible &@($*....) helps a lot specially on those mild blue thermal days, I'll say that those column feedbacks happen only under moderate / severe turb!

 

Most of the time you do "anticipate" or rather react to the change in the horizon or humps and bumps in your &@($* :-) On a glider you should primarily use the horizon / panel relation to keep the desired speed / AoA on a turn, and the string for the coordination.

 

There is not much feedback, if any, from the controls playing an important role in your inputs, although certainly on a tight turn you have to pull using your harm force a lot, and trimming is mandatory...

 

Then, on an helicopter it is even more sensible, and I am sure the control feedback is really minimal, even on a low end machine like a simple R-22, and helos are certainly tricky to fly ;-)

Imagine putting a webcam (transmitting on the monitor of your PC) inside the cockpit of a Bonanza or Baron, and hydraulic servo-controls (remotely operated by your joystick) acting with _INFINITELY LARGE FORCES_ on yoke and rudder pedals.

 

Now, would remotely piloting the aircraft in those conditions yield the same "feeling" of maneuverability, stability, etc. than piloting in the cockpit? I don't have a definite answer, what do you think? I imagine that there are a lot of factors, namely lack of "seat of the pants", lack of peripheral vision, lack of flight controls feedback, that all together could make the pilot perceive different handling qualities.

 

On 1984 there was a Controlled Impact Demonstration ( http://en.wikipedia....t_Demonstration ) where a Boeing 707 was remotely piloted by a NASA test pilot. Here are some excerpts I took from the NASA report ( http://ntrs.nasa.gov..._1989006539.pdf ):

 

.The ground pilot’s controls consisted of conventional stick and rudder pedals for aircraft three-axis proportional control. A stick computer allowed stick and rudder characteristics, such as breakout force, force gradient, limits, and trim rates, to be selected by the project pilot to obtain desirable handling qualities. (Page 6)

 

.Early in the program, the RPV pilot chose an attitude command system, rather than rate command, for primary control in the roll axis. The RPV pilot commanded roll attitude of the aircraft proportional to the lateral displacement of the control stick in the ground-based cockpit. (Page 9)

 

.The design of the guidance systems was continually being modified to accommodate the comments of the RPV pilot. (Page 14)

 

.All RPV takeoffs and landings were generally satisfactory, although the landings proved to be a difficult piloting

task. The difficulty was a result of poor depth perception and lack of peripheral vision through the outthe-

nose video monitor. (Page 15).

 

So, looks like it's not a simple matter of connecting servo controls to a joystick and remotely flying the aircraft (what basically virtually happens inside a flight simulator). The excerpts I highlited seem to suggest that the feeling and handling when remotely piloting, could be significantly different and hence require extensive "tuning" of the flight control system.

 

Now, all of this is fine and dandy, but at the end of the day the important thing is: how can be achieved the max "realism" in the simulated aircraft? Of course, the real world pilot doesn't care that the simulated aircraft has a 100% correct flight model, if the final handling of the aircraft appear to be incorrect. As has been said before (by tkyler I think), probably it's a matter of a compromise on part of the designer, to give the simulated aircraft a "realistic" feeling even at the cost of sacrificing a 100% correct (from an engineering standpoint) flight model.

 

For example, I think that all FSX aircrafts (not only default ones) model some sort of flight controls travel reduction in any phase of flight: in real life, in theory, an aircraft without a movable stablizer should have roughly the same maneuverability irrespective of the elevator trim position (provided the pilot can overcome the flight forces, hence the word in "theory"!). In FSX, try putting full nose down trim and you'll probably see that the aircraft cannot be stalled. Why is that "tweak" necessary? Well, probably because, contrary to what LAdamson says, remotely piloting an aircraft with a light loaded, 6'' stick connected to uber-powerful servo controls (what effectively happens in a non-tweaked flight simulator) is much different that piloting it in the cockpit.

 

Marco

Murmur-perhaps I confused but I am talking about stability which can be very easily demonstrated and not simulated control feel-though at the present in sims it seems to be easier to model control feel than stability. :wacko:

 

Interesting though-if you zoom out to be able to see the yoke in cockpit response-in the Carenado Bonanza you are able to utilize control input like my video shows and get a similarly expected response. Trying other xplane models you don't.

There is not much feedback, if any, from the controls playing an important role in your inputs

 

Well, yes, I imagine it's like a car, and the only time you really feel drag on the steering wheel is when your tire is in a rut or something, but to my credit I did say that yoke pressure was only one of several sensations one would use when flying a plane. Your brain processes countless sensory input, and you just naturally do the right thing without thinking about it. Limit the sensory input and things become more twitchy.

Now, would remotely piloting the aircraft in those conditions yield the same "feeling" of maneuverability, stability, etc. than piloting in the cockpit?

 

I don't think so.... Unless you add to your scenario a 6dof platform fed by the accelerometers feedback from the aircraft, unlles we're talking about an heavy airliner and/or a very steady day...

Geofa...definitely clarifies a lot. A few points:

 

I was tired after these flights-but with flight sim I can feel that way after a few minutes (and I am talking all flight sims presently)

I propose something here that is very difficult to prove....and that is the lack of peripheral vision in a desktop sim. What we generally deal with is akin to putting on blinders....and the lack of peripheral input, IMO, makes our perception and anticipation of response much more difficult. The rods in our eyes are the super-sentive ones of course and in reality, we have a "heads up" to take corrective action that we dont' get in a sim until it' too late...therefore we have to focus even harder with a desktop sim. We don't have "early warning rods" and with brain processing in billionths of a second (for most anyhow ) that's enough time lost to make us feel like we're always behind. Anyone flying IFR "in the soup" knows how mentally draining it is without that extra visual input, its exhausting. I think anything short of a full surround screen will keep this problem at the forefront....so then, how to deal with it is the question?

 

 

Interesting though-if you zoom out to be able to see the yoke in cockpit response-in the Carenado Bonanza you are able to utilize control input like my video shows and get a similarly expected response. Trying other xplane models you don't.

 

As far as the Carenado aircraft...I do not know what they do, but I would not yet go so far as to assume they have it right "across the board"....this is because I've seen some developers use some techniques that I myself do not use because it causes unrealistic behavior in other flight regimes....and depending on the importance of the regime, I might not go there. I can think of a 100 ways to 'fool' a user and attain any one performance parameter to convince anybody of anything using any method as long as they don't dig around under the hood and complain about what I put there. Now if countering the aforementioned problem due to lack of peripheral vision means "going there" in some cases and stretching the model to make it behave well in a flight regime deemed most important....then I can't argue that...perhaps that what Carenado has pursued.

 

In the engineering world....in finite elements at least, we may do several models depending on what we're after and the level of detail. A model sufficient for one regime isn't sufficient for another and I WILL contend that to me, I find it perfectly acceptable to have multiple flight models of aircraft optimized for certain "simulation scenarios". Thinking a single model can work across all scenarios is not a valid assumption in the world of numerical simulation, which is what XP is....and I think we all (myself included) can fall victim trying to make a single representation universal...like some challenging puzzle with no solution but because we never reach it, we believe we just haven't solved it yet and fight a never ending battle. Unless a model is tested across all regimes, it can't be said that is is accurate across all. My engineering experience says this doesn't really exist, something is always sacrificed, approximated or "ignored" as not relevant for the solution sought.

 

Now as to why the real baron doesn't roll with no trim as you say...that is intriguing to me....I'd ask "at what speed" though? I see a tab on the right trailing edge of the main wing....what is that for? An aileron trim tab? If XP doesn't model the effect of propwash countering the torque enough...then chasing reality is impossible except with alterations from reality. I could use a trim tab, but that would be "speed dependent". A baron might not roll in cruise as the trim tab is set to counter it...but at lesser speeds it would, etc. It's mathematically impossible (balance of forces) for a Baron to exhibit no roll from standstill to cruise with a non-altering surface / power configuration....so the question is, "what is the configuration at any given speed". (us engineer types would translate that into a force diagram where forces must balance if the plane is non-rolling/pitching/yawing) If you contend that the Baron never needs roll input (yoke, trim or trim tab) ever in any combination of power...be it takeoff, steep climb, etc....then I'd have to contend that just doesn't jibe with my aircraft design education and there's something missing from the equation we have yet to account for....it might be as simple as XP's ineffective propwash model and if so, we'd simply have to depart from reality a little bit to adjust. For many, the simple application of some aileron trim suffices....but I consider this a deeper "mental journey" into the XP flight model toolbox for our daily dose of cerebral exercise.

 

Currently, I have no setting to counter torque on the Baron currently...it will roll from straight and level until I put something in to counter it, or you do (like trim). I could set both engines to counter-rotate...but animate the props in the same direction to fool you, but that is an example of an area I won't go...I'd rather understand it all and make it work. I agree with jcomm that I think the real effect from propwash is stronger than that in XP....and possibly in combination with an aileron trim tab is what balances forces in cruise...that is my hypothesis as of today. With no trim tab on the XP Baron...and a 'weak' prop-wash model, I would expect it to roll. I can "up" the aileron tab to make up for XP's weak propwash effect on the wings (for a twin)....but it would then overcompensate and roll the plane in certain instances (like when power is removed at cruise) and the aileron tab overpowers the yoke.

 

I would say that probably the one thing I do that most do not that makes me go one way or the other is I test across a broader regime as the developer...and if I get good behavior in one area, but not in another, I might decide to make concessions so no one area is perfect. I test with slow speed or engine out or engine out with flaps, hard turns, full deflection with each engine out, both out, at high altitude, low altitude, low fuel levels, high fuel levels, etc and try to find middle ground. I can probably be convinced though to say, " those regimes are fringe enough to be sacrificed Tom...get with it and focus on normal operating range", the assumption being that for a sub-100.00 flight simulator...you won't get perfection in all flight attitudes.

 

Tom K

 

P.S. I am SUPER guilty of rewording my posts AFTER posting...be wary as I may clarify or include new info to better make my point...especially on my longer-winded posts...usually I get to a "final form" within 5-10 minutes of posting....best to wait till my post time + 10mins to be sure it's in a final form....sorry

Murmur-perhaps I confused but I am talking about stability which can be very easily demonstrated and not simulated control feel-though at the present in sims it seems to be easier to model control feel than stability. :wacko:

 

I wrote my long post before your reply, I wasn't referring to stability. So yes, I completely agree with you on the stability issue. :smile: