otrcman

Hi Sal,

I agree with Burkhard in that the correct CG will turn out to be well aft and that the pitch damping should be very high. That said, you may wind up with an airplane that can be managed in spite of some basic pitch instability. You may find that the airplane wants to slowly diverge into a climb or a dive if you don't keep your eye on it, but that it can be managed as long as you stay alert. This is similar to flying an airplane that has a spiral instability in which it will wind up into a left or right spiral if you don't stay on top of it all the time.

Now, with regard to figuring out ahead of time where to put the lead. Rather than trying to come up with the perfect final solution on the first flight, you might consider some temporary expedient for the first flight or two just to learn more about the final CG. I built a Grumman F3F-2 and wasn't sure about CG placement due to the short, fat fuselage and small empennage. In that case, I didn't dare add more lead to the nose "just to be safe" because the airplane could have been out of limits in the forward direction just as easily as out of limits aft. My solution was to temporarily add area to the horizontal and vertical surfaces. I made "gloves" that slipped over the leading and trailing edges of the surfaces to increase the areas by about 40%. The gloves were held in place with masking tape. On the first flight, the plane flew like it was on rails. It recovered easily from a spin and it slowed to a stall with up elevator. Now I knew I was in a safe range. So I removed half of the gloves from the horizontal stab. Flown again, it wasn't quite as solid, but still safe. Then I reduced the horizontal gloves to 1/4 of original size. It still flew OK, safe spin recovery and could make a 3-point landing. Lastly, I removed the final bit of horizontal glove. Still OK. Then I incrementally removed the vertical gloves. In a total of 6 flights, I had safely determined that my GC calculations were correct. The gloves had been nothing more than cheap insurance.

With that huge stabilizer and long fuselage on the llya, I don't think you need to worry about the CG being too far forward. Aft will be your only concern. So you could just temporarily fit a rod to stick out a foot or so ahead of the nose, and put a lead weight on the end of the rod. As a matter of fact, perhaps two rods would be better; one rod on each side of the fuselage or attached to struts. Maybe make the first flight at a CG of 50% MAC. The rods would be ugly and non-scale, but they would allow you to fly safely and learn more about the airplane. Then incrementally reduce the weights as you learn more about the flight characteristics. At that point, you will know EXACTLY where you want the CG and can install ballast in a more scale like fashion if necessary.

By the way, directional stability may be a concern with this model. I see those two little extra verticals and they look like add-ons to me. Perhaps Sikorsky wasn't happy with the directional stability as originally built. Are those extra verticals fixed or movable ? If they are fixed, he was dealing with a stability issue. If movable, then he needed more control.

Dick

UStik

Thanks a lot Sal, it's really nice to read a Good Morning first. :-)

To be precise, I don't know if the Russian document means the C/G was that far aft or if it just discusses a what-if.
So let's do that ourselves:

I checked that a forward C/G (ahead of the NP) is possible at all. I'd be afraid the undercambered stab would produce an "inverted" stall at negative AoAs. I just don't know what could happen in this case (at least much drag, but maybe also loss of elevator power). But the calculation says there will be some positive AoA even if the C/G is pretty forward. I wouldn't go as far as Dick suggests, though, since a C/G at 50% of chord means only a tad of positive AoA what might even stall the stab's bottom. Besides it would make for 28% stability margin what is really much.

20% is a good stability value, C/G at 59% of chord. Still the stab's AoA is too small for my taste so I would even positively avoid such forward C/G positions in this case. I know that's counterintuitive (for me as well) but you see the reason why. And I wouldn't see it that way if I wouldn't think the IM is manageable even if not pitch stable. Let me again remind of using a gyro on the elevator. In lock mode it would lock a certain pitch which could be adjusted by short stick pushes or pulls (like the heading of a helicopter). If you want to know which elevator throw is needed to balance the model you can switch off the lock mode.

Dick, your method with the gloves is a great idea.You say a forward C/G would be of no concern in this case but an aft one. But to me the IM seems to be so C/G insensitive that there is no concern at all. For instance, a -20% static margin would put the C/G on the trailing edge and let the stab lift 6.7 lbs what it would happily do. This would even settle my paranoia about the stab's AoA. So I would feel best if the C/G would just come out somewhere in the 60% to 100% chord range and no weight would be added to the already heavy model - and if there'd be a gyro to my help. ;-)

And to the point of too small vertical stab: Maybe the calculation can contribute to clear it up a bit. The yawing moment coefficient Cnrv is -0.0012 what seems perfectly good to me. What I would be afraid of with this airplane is a huge yawing moment needed to counteract adverse yaw. I think the reflex in the airfoil camber doesn't help because the ailerons are merely appended to the wing whose airfoil is virtually unchanged. Unfortunately, the calculation doesn't help in this question.

UStik

Another consideration, worth a new post:

I checked my airfoil books (with wind tunnel measurements) and found a few similar airfoils that are used for free-flight competition models. Maybe closest is the Sokolov with 6.4% camber at 43.5% chord and 7.2% thickness at 22.2% chord. There are no pitching moment measurements for this airfoil but for the Kaczanovski, a similar airfoil.

Anyway, the diagrams show that these airfoils go up to a 1.4 lift coefficient Cl even at very small Re numbers, and they are even most efficient at nearly maximum lift.
They lose efficiency (meaning produce drag due to bottom-side stall) at Cl smaller than 0.5 which is reached at zero AoA. This calms me down. ;-)
But this means you have to keep the IM at slow speed and have to avoid high speed.
On the other hand, the drawing you sent seems to suggest 8° wing incidence. That is indeed the most efficient point with 1.3 Cl.
The calculation says maximum speed (Cl=0.5) would be 32 mph and minimum / most efficient speed (Cl=1.3) 19 mph.

I think this is all counterintuitive again. You can't control this speed since you're on the ground and have no instruments (telemetry?). And if the model comes out unstable (but still perfectly manageable), what I suspect, you can't even trim it for a certain speed. You can only control the model by visual clues, mainly horizontal fuselage, what is hard from various viewing angles (or use a speed controller).

Now I checked the pitching moment of the Kaczanovski: Its Cm is -0.275, even more than estimated (-0.16).
The neutral point isn't changed by that (if it's true), stability is unaffected. But the necessary incidence angles are different.
It lessens the load on the stab so requires a more aft C/G to keep the stab lifting (avoid bottom stall).
+15% stability margin (stable) would be still possible with the stab lifting only 2 lbs (Cl=0.16), though.
-15% (unstable) would make the stab Cl=0.5 and the stab lift 6 lbs.
Sal, the drawings you posted seem to show this setup (the corresponding incidence angles).

FWIW.

Sal C.

Gentlemen

Again, thank you for your interest and comprehensive responses. Reading all those figures my head is going to explode. I'm good with balsa and an Xacto blade, however,
math was not my best subject.....
Dick, the outer rudders were not add ons, but used throughout the Grand and Ilya Muromets designs. They are undercambered and supposedly used for engine out conditions. They move in unison with the main rudder, driven by rods from the main rudder.
IN all the reading I have done about the Ilya Muromets, the elevators were extremely heavy and the pilot needed assistance pulling on the Deperdussin control column during landing. Obviously, they were not balanced. Also, due to the extreme incidence angle on the wings, the aircraft flew nose down at about 3 degrees.
To sum up, with all the calculations, the model will not be CG sensitive and anywhere between 60 percent to 100 percent of wing should be okay.
We have an amazing field to fly from, right in the center of a large SOD farm. I can test hop when completed.
It is still a long way off, the devil is in the details.
I will probably start a thread shortly.
CHeers, Sal

otrcman

Sal, I envy you your sod farm flying field. We fly out of a cow pasture and you have to be very careful where you step !

The fact that the outer vertical surfaces are movable definitely says that their intent was to increase control authority. With the reliability (or lack thereof) of period engines, coping with an engine out situation sounds pretty logical. But dealing with adverse yaw may well have been equally important. Those long wings would promote strong yaw when the ailerons were deflected. Which brings up a point for the model: Sal, do you intend to mix some rudder in with the aileron control ? Or are you accustomed to manually mixing by moving both sticks at the same time ?

Looking at the overall planform configuration of this airplane, I think we can safely call it a "tandem wing" design. Another piece of evidence that the CG was at or near the wing trailing edge is the payload location. The passenger or payload section of an airplane is typically centered at the CG so that variations in weight of the load don't change the CG too much. From the limited pictures that I've seen, the passenger area looks like it is centered pretty much at the wing T.E.

Dick

Sal C.

Hi Dick

I have a few WW1 types, some of them do well with just aileron like the Fokker D-VII. What a great upper wing!!! However, I have a 1/3 scale Pup that loves rudder.
So I am always on both sticks. It's like driving a truck..it loves pointing into any type of wind but has so much lift it takes off in a few feet and crawls in for a landing.
Obviously, this Muromets monster will only be flown in slight or no wind conditions. I just had to build it...It was a significant step in aviation history.
Cheers, Sal I just started a thread in RC Warbirds.

UStik

Since I'll be off tomorrow and since Dick mentioned flying the thing, I want to get this off my chest: ;-)

Flying the pitch-unstable airplane is like flying a helicopter in a a sense. I don't mean the heading lock in this case but pitch and, in the case of a helicopter, roll as well. These are stabilized in any case, either by mechanical means (paddle bar) or today by electronic gyro. Still you need good visual clues of the helicopters attitude, by the canopy's shape and/or striped rotor blades. To some extent you can control the helicopter by judging its movements. Mechanical stabilization makes these movements around the pitch and roll axis slow/sluggish so a human being is able to react properly. Electronic stabilization even locks attitude.

I think the IM is at least as sluggish as a helicopter, but it has a limited speed range and can't be stopped mid-air just to test its reactions. It won't stay nearby, either, so you may have a hard time to recognize its pitch attitude from some viewing angles. That's why I would use a lock-mode gyro for pitch so I can at least rely on a certain pitch being maintained. This way I have the time needed to observe if the model climbs or descends (independent of gusts) and apply small corrections (like with a helicopter).

On second thought a variometer could be a good idea as well. You'd have the vario tone in the earphone and would intuitively and immediately realize any climb or descent.

And maybe you should prepare a combi mixer (rudder to aileron) even if you're a proficient "stick and rudder man". ;-)
If the model is not too close to the pilot he may have a hard time finding the proper amount of rudder. After all the visual clues needed for that are lost with distance all too soon.

Naah, I won't suggest FPV now, that's a different department... :-)




I forgot:
Dick, if I got the stability calculator correctly, the IM has no spiral stability either.

otrcman

Not surprised that there wouldn't be any spiral stability. That's no big deal. But it would be nice to have basic yaw and pitch stability.

Burkhard, I had never thought about using a "heading hold" gyro mode for pitch only on a model. Most people think in terms of yaw when talking about heading hold. I can see where pitch hold might be useful on the llya, especially as insurance on the first couple of flights.

When I was working for NASA Flight Research Center back in the 1970's as a flight test engineer, one of the projects that I worked on involved a Piper Twin Comanche in which we installed a fly by wire flight control system. The purpose of the project was to explore a number of advanced pilot control schemes. In addition to doing engineering work, I participated as a pilot test subject in evaluating some of the control modes. One of the pitch control modes was essentially identical to what the model airplane people now call heading hold, or what Futaba calls AVCS. We called it "attitude command", which I still think is a more descriptive term.

In roll attitude command, control was easy and intuitive. You simply turned the yoke to bank the wings. Just like a car steering wheel. Twist the yoke ten degrees and you got a ten degree bank so long as you held the yoke in that position. Twenty degrees of yoke made twenty degrees of bank. As soon as you allowed the yoke to center, the wings returned to level and the turn stopped.

In pitch, the attitude command mode was easy to fly, but had some subtle problems. Basically pitch was just like roll. The airplane stayed level as long as you didn't touch the control yoke. If you pushed the yoke forward one inch, the nose went down ten degrees. Press the yoke forward two inches and the nose went down twenty degrees. Release pressure on the yoke and the nose came back to its original position. So far so good, huh ? OK, so what happened if you were cruising along in level flight and then pulled the throttles back ? Answer: The nose stayed level and the speed reduced. If you throttled back too far without trimming the nose down, the plane would stall. Conversely, if you firewalled the throttles while in cruise flight the nose stayed level and the speed increased. For specific tasks, such as level flight cruise or even an instrument approach, the pitch attitude command worked fine. But you had to understand the system and act accordingly when doing something like an ordinary visual approach and landing.

I think this same behavior would occur in a model if you tried to fly in "pitch heading hold". As long as you remembered to re-trim the elevators as you throttled back for landing approach, all would be well. But if you throttled back without retrimming elevators, the gyro would try to keep the nose level and the plane would eventually stall. If you are savvy enough to use the mixers in your modern transmitter, you would no doubt do a little throttle-to-elevator mixing so that the nose would go down as you retarded the power. But I think I'd play with that mode in a docile trainer before incorporating it into a valuable model.

Don't let me discourage you from considering a gyro. I think they can do great things. And I do have gyros in some of my planes. You just need to understand the ramifications of the particular control mode you have selected.



This gets more interesting all the time.

Dick

Sal C.

Hi Dick

Okay, so what about using the Futaba AVCS in flight and switched off on approach, would that be the best of both worlds? Between both of you, I feel that I have my very own
Flight Test Center!!!

Cheers, Sal

otrcman

Since I've not tried it, I can only speculate on switching the gyro off for landing approach. If the airplane is pitch stable, then there would be no downside to turning it off for landing. But if it is all that stable, then you may not need the gyro anyway. As I said before, the smart plan would be to try it out on some expendable trainer type plane that you can afford to lose. I keep such a model around just to fly newly purchased radios until I am confident that they are reliable.

You may find that you can fly all the way down to landing with pitch AVCS operating. By the same token, you may find that the llya flies much like AVCS with no gyro at all. It's just that the basic airplane might tend to wander nose up or nose down while the AVCS would lock the attitude more positively. When you are close in on landing approach you will likely to see any wandering and correct it easily, whereas at a distance you might not be able to detect a nose up or nose down change as quickly.

Dick

UStik

Sorry, I'm so used to flying with such gyros that I didn't think about problems adopting it. And I had trouble to put into words how it's different.

I think it's not quite your attitude command, Dick. Heading lock on a helicopter means the stick controls the yawing rate, which would be indicated by the turn-and-bank indicator. Likewise, a pitch lock gyro will lock pitch attitude and stick deflection controls pitching rate. That came natural to me because in a full-size airplane I hold attitude myself after throttle is reduced for descent to reduce speed first. The gyro is even more convenient. A roll lock gyro is even more natural since aileron is deflected until a certain bank angle is reached and then returned to neutral. The gyro does the same plus correction for gusts. and holding top aileron.

I use a German brand of gyros which all have a lock mode included, which can be switched off (then it's a rate gyro, a damper). They have special airplane (as opposed to helicopter) gyros, with one channel for rudder and elevator, and with two channels for aileron. Gyro response and stick input are mixed (cross-faded) and you can set a deflection limit.

I have a few micro helis, one self-righting by arrangement of the paddle bar, the other with flybarless controller. The latter is like a lock mode gyro. The heli maintains the attitude I set with the stick and acts accordingly, that is it accelerates if it's not exactly level. This is way more stressful than the self-righting heli. Still I can switch between them effortlessly, it seems just a different mindset, like switching between airplane and helicopter.

The stress comes not from the flybarless (attitude lock) controller but from the small and squirrely heli which I can manage only with help from the controller. The IM will be a calm flying airplane, just not stable, and should no problem even if you're still not used to the lock mode. And Sal should manage it even better than I can.

And reducing throttle might have adverse effects in an unstable plane like IM. The gyro would help in this case.

otrcman

Yes, I think you are correct, Burkhard. Not having flown a model in AVCS in any axis, my understand was limited. So what we're talking about is a "rate command" system, rather than attitude command ? In other words, yoke deflection commands a rotation rate, and then when the yoke is released, the rate stops and the airplane wants to remain fixed in it's current attitude ?

If that's the case, then having AVCS in the roll axis should work pretty well. It's sort of like a perfectly linear roll control system, since ailerons are already a roll-command device. I did fly the Commanche in roll and pitch rate command. As I recall, the roll axis was quite pleasant other than having no apparent spiral stability. In pitch, it seemed to work well for certain tasks and less well for other tasks. I didn't do any landings in pitch rate command.

One of the X-15 rocket airplanes that was flying at our facility concurrent with the Piper had an experimental flight control system that included pitch rate command. I had very little involvement with the X-15, so didn't know much about it. But I do recall the pilots saying that landing in pitch rate command required some adjustment in technique. Here it is only 50 years later and I'm already getting fuzzy on the details.

Do I understand that you have flown an airplane (model or full size) in pitch rate command ? It would be interesting to learn what adjustments you may have found necessary as opposed to flying a conventionally stable airplane.

Our discussion here may be drifting a bit for Sal's tastes. Sal, would you prefer that Burkhard and I move to Private Message or email among ourselves for this gyro talk, or do you like to follow along ?

Dick

Sal C.

Hi Dick

By all means, stay put. I don't understand much of what is being written but I am happy that the two of you are interested in my project and quick to respond.
Plus, you both seem like nice gents!!!

Sal

UStik

Yes Dick, basically you are right as to the lock-mode gyro. I didn't get the other points across, though. I was in a hurry and had not enough time to put it all shortly.

I just remembered full-size flying "by hand". The transition from fast level flight to a slower approach speed and to the glide slope would be easier with a lock-mode gyro, I imagine. Just throttle back, wait till speed is reduced, and with a short push of the stick set the glide slope. I don't have to hold the elevator up to slow down in level flight, and speed in glide is at least roughly adjusted by attitude. I just have to do corrections and otherwise work the throttle. I didn't think about landing so far.

My experiences with gyros are mostly with model helicopters. I figure since there's no problem to switch from a self-righting heli to a flybarless aerobat and back then there should be no problem switching to a pitch-controlled airplane, either.

What I didn't consider so far is the IM's behavior when power is reduced to commence the landing approach. Since the whole stab is blown by the propwash and since the wing must be influenced as well and both have undercambered airfoils, it could be interesting. That's why I would feel better with a gyro.

Flare and touch down seem not special to me in the handling department. Due to the big pitch damping I would switch off the gyro (and an aileron-rudder mixer).

otrcman

Hello, Burkhard,

Did you mean to say that you would switch off both pitch lock mode and aileron-rudder mixer for landing ? I can understand the logic behind switching off the pitch gyro, but why the A-R mixer ? Seems to me the mixer would be beneficial at all times, particularly if the model had extreme adverse yaw (which I would expect to be the case).

I am becoming more interested in the merits of pitch rate command (what you call lock-mode). In the case of an airplane with a wide speed range, power reduction would produce a level flight deceleration until approach speed is reached, then a small nose down push to maintain approach speed as you describe. Then the new nose down attitude would pretty much lock you onto your new airspeed. The only special consideration for the Ilya may be that it is pretty much a "one speed" airplane, where the cruise speed is no higher than the approach speed. It's no doubt going to be a high drag airplane as well. So probably Sal won't ever have to reduce power to idle on landing approach. Probably he will only reduce power a small amount before seeing a sink rate which is acceptable for landing approach. Then carry the power all the way to touchdown and only do a final power reduction at the moment of touchdown. He might even increase power a bit just before touchdown to reduce the sink rate.

One last comment: I don't think you would want lock-mode active in yaw. If you incorporate yaw lock-mode, I think the gyro would not understand your intent to change heading unless you held some rudder all the while you were turning. if you banked the wings and then neutralized rudder (to prevent over banking), the gyro would begin to fight any further heading change that occurred after releasing rudder. You could find yourself in a situation where the airplane was continuing to change heading due to the bank angle, but the rudder gyro interpreted the heading change as an uncommanded event.

Dick

UStik

Yes Dick, in this case I really meant it, but obviously I didn't think about it so it's more a habit of mine. It's all about flare and touch down. I feel disturbed when I want to correct a bank angle and there's a coupling. The mixer just isn't quite correct, it's always too much or too little and that's less predictable than adverse yaw, which is hence corrected intuitively. And I like the sideslip method in crosswind, that is flying with crossed controls, and feel adverse yaw is even helpful for that. But probably you're right that IM has so severe adverse yaw that a mixer should at least reduce it to a normal amount. That would be not disturbing but helpful since there'll be always the "positive" adverse yaw we're used to. And probably the sideslip method is not possible with IM since the long wings are so near to the ground.

And if the IM is a constant-speed airplane you don't even need different amounts of mix like we use with flight modes in our transmitters. And yes, that would mean there's no big problem in the transition to glide slope for approach since speed and power setting stay roughly the same.

And of course you're right as to heading lock as well. I don't even like it on helicopters since I'm annoyed by being compelled to hold the yawing rate throughout a turn. Unfortunately, that's the way it is. In the simulator, I can switch between lock mode and rate mode, meaning the gyro is only a damper what is useful and even required. I even have a rate gyro as yaw damper in one of my airplanes, a Stik-like electric aerobat with a quite small vertical tail. Maybe that could be useful in case of the IM if it has big yaw moment of inertia (the wings and engines) and little yaw damping (the tail).

I wish there would be affordable CAD systems which derive stability characteristics from the 3D model and the specified materials.

UStik

Seems I'm somewhat cursory. Checked the wind tunnel measurements again and it seems I mistook one of the lift-over-AoA curves for the pitching moment curve. Anyway, the value is only about -0.125 what is not that terrible. Since that's the moment around the quarter-chord point, it lessens with increasing AoA (to -0.05 at max. AoA), I think because maximum camber is far aft at 43.5% chord. At least this seems to be standard for such free-flight airfoils so they have the same pitching moment behavior.

Maximum lift can go up to Cl=1.7 if Re number is bigger (than in free-flight, which is below 100,000). They show a diagram for center of lift, which is at 65% chord for Cl=0.3 and goes forward to 30% at Cl=1.6 (maximum). IM's neutral point was calculated to be at 78% chord.

This should mean that flaring the IM was actually helped by the airfoil characteristics. Sal, in that case you should be right that it was hard to flare due to the elevator's weight which was borne by lift in normal flight.

Dick, we puzzled over positively cambered stabs before (Feb. 2014). Only now I remembered that the thin cambered airfoils are efficient, meaning have a steep lift-over-AoA curve and big maximum lift. I think the neutral point concept had not yet been invented in 1913 and they drew a moment plan. What if the wing lift increases so much with AoA that they felt/calculated an undercambered stab is needed to keep the plane balanced, the more so as the wing's center of lift moves forward?

The calculation tool I'm using has at least Cm0 as a parameter but not the variation of Cm with AoA. The simulator has, but it's tedious to build a model...

Sal C.

Guten Tag Burkhard...

Well, It's not that I was right about elevator weight, it is merely what I read. The Elevators are large and deep, without balancing, they will be heavy on the model also.
But that's not my problem, it will be the servo's problem. I chose strong, digital servos for the rudder and both elevator halves. I certainly won't feel anything on the transmitter
sticks. I started a thread in RC Warbirds and Warplanes but there doesn't seem to be much interest. I think the subject is too rare....

Auf Wiedersehen,,, Sal

BMatthews

This is a very interesting thread and a very interesting subject model.

I'd like to add a couple of comments if I may.

First off is the concern over the spiral and yaw stability. The vertical tails are small but again they are mounted on a rather long tail moment arm. So that means that the vertical tail volume coefficient isn't as small as it would first appear. And another factor that we modelers overlook is the effect of the fuselage. There's a LOT of side area of the fuselage that will be providing some additional vertical tail area. On a model of this sort I'd suggest that this needs to be considered and included. Even if it's just to feel a little more secure over the vertical tail area situation.

On the subject of the rearward CG location and the tail lift. Don't get the cart before the horse. With the CG that far back the tail has no option but to lift during level cruise. The amount of lift it needs to generate is set by the wing to tail sizes and the moment arm along with the CG location. The airfoil used on the tail won't change that. What using a lifting section on the tail does is allow the stabilizer to provide that lift more efficiently. It also allows the tail to operate at a higher lift coefficient for a given angle of attack. And that's good because you really don't want the tail stalling and altering the amount of lift it generates before the wings stall.

If there was ever a model that was made to benefit from the stability enhancement features of the newer radio gear this was it. But it might be that it's not as bad as some may think. You know all those flat foam flippy flyers that folks fly around hovering and harriering and generally defying the laws of belief with the stunts they do? Have you flown any of those correctly set up? Turns out that the CG of choice for such models is a trifle behind the NP and that they are mildly UNstable in pitch. But because of the good damping value and the low airspeed they fly at it's not actually at all hard to keep up with the small elevator changes needed. The one I flew for a while wasn't as strongly negative as some of the others around the field. I found that holding a pitch heading was no more work than keeping my car in a straight line with the little nudges this way and that. Yes, it's negative. But it was very easy to provide the slight nudges needed to fly level without creating a noticeable wobbly/drunken flight path through the air. And we all know how small models seem so "flitty". A bigger model like the IM would likely not be bad at all. I'd still plan on using a heading holding setup. But it likely won't be working as hard as some might think.

Best of luck with the rest of the build and I wish you a safe "one piece landing" for your first flights.

otrcman

BMatthews said, " A bigger model like the IM would likely not be bad at all. I'd still plan on using a heading holding setup. But it likely won't be working as hard as some might think."


Glad to have you on board here. I always enjoy reading your insights. Regarding the quote from you. Do you mean that you would plan on using heading hold in the yaw axis or in the pitch axis ?


Dick

UStik

Sal, thank you for the German words!
Unfortunately, you might be right that the subject is too rare. I remember a fellow modeler in the RC Gliders forum who - like you - built fantastic models of historic airplanes which were somehow important for the development of airplane technology. Still very few RCU members took notice. I think these airplanes have too little of what we take for granted today to be interesting. Sad but true.

UStik

Dick, I think he means pitch axis since he said "holding a pitch heading". I could imagine at least one useful application for true (yaw) heading lock, though: If you had to be afraid of losing an engine on a multi-engine airplane, heading lock could be a good idea. Then again IM will be electric and there's no concern. It would be just nice to have like the pitch hold gyro, as we had found and as BMatthews confirmed. Does that settle the matter? ;-)

otrcman

Yes, I thought that BMatthews meant to apply the lock mode to pitch. I just wanted to be sure of what he was saying.

Regarding the use of yaw lock for engine out control, yes that might be useful. The PA-30 that I flew many years ago did incorporate heading hold (heading lock) in the attitude command mode and (I believe) in rate command as well. What the designers of the system did was very innovative. The heading lock automatically engaged any time the wings were level. As soon as a bank was initiated, the heading lock disengaged and allowed for a heading change. Then as soon as the wings came back to level, the heading locked up again on the new heading.

The mechanization on the PA-30 was a little crude in that heading lock/unlock occurred at 3 degrees bank angle. Thus it was not possible to make small, subtle heading changes with slight bank as one might wish on an ILS approach. When a bank was initiated, the unlock caused a minimum heading change of 5 degrees or so. I found that the most precise way to make small heading changes was to press the rudder a bit and cause a wings-level skid. The controls designers said that I was making a most unpilot-like maneuver. I responded that they had created a most unairplane-like control system and my behavior was simply a piloting adaptation to their system.

One of the limitations of the model airplane gyro systems that I have played with is that one must select the same mode for all three axes. If you select RATE, then you have rate in all axes. If you select HOLD, then you get hold in all three axes. But I think a more friendly control system would incorporate different modes in each axis. I like HOLD (rate command) in the roll axis, RATE (simple damping) in pitch, and selectable HOLD (rate command) in yaw. With one of the more complex open source gyros it might be possible to incorporate the heading hold/bank angle algorithm as used on the PA-30. I have not investigated that possibility with a manufacturer.

Dick

My definition of terms:

RATE DAMPING -- Called "RATE" by most modelers. The gyro simply tries to maintain angular rates to zero if there are no pilot commands. Provides for less bouncing around in turbulence and reduces overshoot with large stick movements.

RATE COMMAND -- Called "HOLD" or "AVCS" by most modelers. Airplane angular rates are proportional to stick displacement. The airplane tends to remain in the last commanded attitude when there are no pilot commands (zero stick = zero rate). Rates are proportional to the magnitude of stick displacement.

ATTITUDE COMMAND -- Airplane attitude is proportional to stick position. If the stick is deflected 10 degrees to the left, the airplane banks 10 degrees left. When the stick is neutralized, the airplane returns to level flight. Same with pitch. When the stick is pulled 10 degrees aft, the nose goes up ten degrees. When the stick is neutralized, the nose returns to level flight. Note: The ten degrees stick for ten degrees roll or pitch is arbitrary. It could also be that ten degrees stick resultes in twenty degrees roll. The point is that the airplane attitude is directly linked to stick position.

BMatthews

Pitch IS the primary one I was referring to of course. That was the core of this thread topic after all.

Ya know..... given all the work that you've invested in this giant scale model I would suggest that perhaps a simplified 4 foot version to use as a trainer and investigation tool for the flight manners of the big IM would not be amiss. I'd make it a "stand way back and squint" sort of version. Or perhaps even a "seen from horseback at a fair clip" sort of scale model.

It's also not like it would be a waste of time either. I suspect you don't get a lot of chances to fly the big stuff. So a more casual version to use at the local field on weekends would not go to waste after you've learned what it has to teach about the CG and spiral stability issues.

UStik

BMatthews, I too always enjoy your posts, especially the two in this thread. Thanks!

Dick, I couldn't agree more. Now I see what the options of a control system could be. The gyro designations are a bit confusing, though. I hardly found the meaning of AVCS: Angular Vector Control System. It's just Heading-Hold without drift (newer technology). Still Futaba calls it a rate gyro. Anyway...

I even found another application for heading lock: takeoff. Even though the moderator of another forum discourages us from using it, I would try it. His reason was that rudder is natural for a taildragger pilot and hence changing the stick behavior to Rate Command would be confusing. After viewing many warbird event videos on YouTube I doubt the first part, and from my experience the second part as well. As I said before, for me changing mindset is easy. The gyro is switched off after takeoff, anyway. Maybe it's a matter of training.

I'd rather doubt that heading lock for engine-out control is a good idea. You have to leave it on for the chance of losing an engine while it's actually inconvenient. Fortunately it's not needed for IM.

Now I see how an integrated control system could automatically switch modes or even gyro gain and couple some reactions. Since there is free and open software for some of these control systems you should be able to program all you want. Some systems are even able to control VTOL with transitions. Just that's what I would be interested in most, but I have already too many hobbies/projects. And programming such a system for IM (sort of a robot test pilot) would be a project of its own and is not really (not at all) needed so I would stick to separate gyros.

If not before, I should switch to off-topic now...
[off-topic]
Dick, I really enjoy your explanations of the PA-30 control system (particularly, in addition to all your other posts). First reason is my interest in control systems and the ideas you brought up. Even though I'm an engineer I'm with you and see it completely from the pilot's perspective. A good pilot adapts to the system he happens to have; if he wouldn't/couldn't he wouldn't be a good pilot. He does it the way he thinks is best, what may be different from how it was meant to be by the designer. Now it dawns on me that we may easily switch between different mindsets like pilot/designer (as well as rate/lock mode) and are at any time established in only one of them. I wouldn't deny that switching makes for some workload, though.

Second reason is the PA-30. Even though I have only three and a half hours on it (too expensive), these were very nice and memorable flights (fond memories), all with instructor. He had done military formation aerobatics and found the PA-30 to be a good aerobat, just lacking the necessary structural strength. He even demonstrated gentle aerobatics, but what I remember most are the overhead trim crank and the two three-position tank valves between the seats (and the many levers, it was a turbo). Oh well...

He was a military transport instructor and told me a funny (not necessarily true) story about paradropping (what I did small-scale with a six-seater). When the jump master demanded course correction "five degrees to the right" he just dipped the right wing once (without really changing course) and the jump master was satisfied. Now I'm wondering how that fits your PA-30 control system description. Maybe the designers thought less than 3° is not a significant bank angle. ;-)

On the other hand, I used rudder for course corrections as well, for gentle corrections in the approach to the dropping point. Any banking disturbed the VOR indicators (the CDI was not balanced but a pendulum, or it had much damping), which we needed in pre-GPS times. What else could I have done...
[/off-topic]