wing/tail incidence & design decisions [Archive]

Neil Walker

Feb 12, 2007, 01:00 AM

OK, time for another question from Neil :)

I recently watched Paul Naton's (Radio Carbon Arts) "Performance Tuning" DVD. For those not familiar with the RCA videos, they're mostly aimed at glider pilots. The Performance Tuning DVD is more or less about how to set up a full-house sailplane and get it flying as well as possible. One of the things it discussed was setting the incidence of the stab w.r.t. the main wing. In the video he pretty much implies that you should always aim for 0-0 or slightly negative incidence for the stab (w.r.t.) the wing. Now I know Paul is a highly experienced glider guy and the rest of the video was great (I already knew most of it anyway, but it was nice to see it all laid out in detail) but I really wondered about the stab incidence. Soooooo.... the questions:

- Isn't the "correct" wing/stab incidence entirely dependent on the design of the glider? Might not some designs call for more negative (or even positive?) wing/stab incidence?

- What would typically guide such decisions? What benefits do you see from a design that is 0-0 (or close to it)? I'm guessing a glider that could have a wide speed range without requiring major trim changes? Would this also necessitate a very rearward CG - perhaps the RCA video was sort of assuming a high performance glider?

- When would you see designs that weren't 0-0? Positive? Negative incidence?

Thanks!

Neil

JetPlaneFlyer

Feb 12, 2007, 01:39 AM

The difference between wing and stab incidence is often referred to as 'decalage' or 'longitudinal dihedral'. For stable flight the wing should be at a greater angle of incidence than the tail (i.e. positive decalage). The greater the decalage the greater the stability of the model in pitch. The actual angle would vary from one model to the next depending on lots of factors such as stab area, stab moment arm, camber of wing etc.. It is quite possible to have a model that is too stable and the 'right amount' of stability would depend on the use the model was put to and the pilot’s preference.

Having a model set up truly 0-0 would result in neutral stability, it would be incredibly hard to fly such a model. I'm no expert on high performance gliders but I suspect that they are not really set up 0-0. I'd guess that they may be set up with the stab at zero in relation to the chord line of the wing. However for a non symmetrical airfoil this would leave you with an effective decalage of maybe 2 or more degrees (depending on the airfoil used) due to the difference between chord line and zero lift angle of a cambered airfoil.

You would never want to set the stab at an angle of attack greater than the wing. This would result in an unstable and un-flyable model... unless you had some form of computerised stability control like a modern jet fighter ;)

yoyoML

Feb 12, 2007, 04:05 AM

OK, time for another question from Neil :)
- Isn't the "correct" wing/stab incidence entirely dependent on the design of the glider? Might not some designs call for more negative (or even positive?) wing/stab incidence?

Indeed it's dependent on the design, but all designs aren't entirely different, either.

- What would typically guide such decisions? What benefits do you see from a design that is 0-0 (or close to it)?

Premises:
1) A plane must have its CG before the neutral point to be stable, controllable (without computer aid).

2) Preferably, all control surfaces should remain neutral (not deflected) when cruising at your desired cruise speed to minimize drag.

3) On a glider, the horizontal stabilizer is usually far away from main wing wash, so the stab's AoA is not affected by the main wing.

1) means the main wing's wing loading is higher than the stab's wing loading. On a glider, the wings usually have very high aspect ratio, so the slope of CL v.s. alpha is the same for both main wing and stab.

With the same slope but higher wing loading in front, the main wing must operate at a higher AoA when in cruise. The difference between the main wing's and the stab's AoA can come from a deflected elevator or decalage, or both.

From 2) it's obvious the difference in AoA had better come from decalage. 3) says the decalage actually equals the difference in AoA. The main wing operates at a higher AoA so the decalage is positive.

Note that "AoA" here means the AoA w.r.t. the zero-lift AoA, when dealing with non-symmetrical airfoils.

I'm guessing a glider that could have a wide speed range without requiring major trim changes? Would this also necessitate a very rearward CG - perhaps the RCA video was sort of assuming a high performance glider?

A glider, like most planes, requires trim changes with speed.

Only when the CG coincides with the neutral point can a glider cruise with zero decalage and no elevator deflection, but then the glider will have no stability and most likely unflyable without computers.

- When would you see designs that weren't 0-0? Positive? Negative incidence?
Neil

Main wing wash can change the horizontal stab's AoA, so if the stab. sits right under the wash, zero or negative decalage may possibly work.

If a plane spends as much time up side down as right side up, zero decalage should be best. (3D planes)

If the plane is computer/gyro-stablized, premise 1) is gone and the wing loadings can be (more) freely distributed. Then the decalage can be anything, depending on design.

Neil Walker

Feb 12, 2007, 03:33 PM

Having a model set up truly 0-0 would result in neutral stability, it would be incredibly hard to fly such a model. I'm no expert on high performance gliders but I suspect that they are not really set up 0-0. I'd guess that they may be set up with the stab at zero in relation to the chord line of the wing. However for a non symmetrical airfoil this would leave you with an effective decalage of maybe 2 or more degrees (depending on the airfoil used) due to the difference between chord line and zero lift angle of a cambered airfoil.

*bing* (sound of lightbulb going on). Thanks - that fills in a piece of the puzzle that's been bugging me for a bit :) I wasn't thinking about cambered vs. symmetrical foils.

So am I right in thinking that video is over-simplifying things? I should really be setting my incidence by first getting the CG where I'm happy with it, then setting the stab incidence so that I end up with neutral elevator deflection at my cruise airspeed?

A glider, like most planes, requires trim changes with speed.

Only when the CG coincides with the neutral point can a glider cruise with zero decalage and no elevator deflection, but then the glider will have no stability and most likely unflyable without computers.

Yes, I understand this. But is it fair to say that having the CG closer to the neutral point and a small declage will mean that the trim changes required with speed changes won't be as severe?

Thanks for the responses.

yoyoML

Feb 12, 2007, 09:27 PM

So am I right in thinking that video is over-simplifying things? I should really be setting my incidence by first getting the CG where I'm happy with it, then setting the stab incidence so that I end up with neutral elevator deflection at my cruise airspeed?

Yes. The end result is close to what the video says anyway.

Yes, I understand this. But is it fair to say that having the CG closer to the neutral point and a small declage will mean that the trim changes required with speed changes won't be as severe?

Thanks for the responses.

It depends on how you judge the "amount" of trim change.

Indeed, with the CG closer to the neutral point (smaller static margin), less change in elevator deflection is needed trimming to a different speed.

BUT, usually as you move CG backward, you simultaneously reduce max elevator throw so the elevator stick doesn't become too sensitive. Then the lower required trim change combined with a lower max throw give you exactly the same % of stick travel as the required "trim change".

In short, the amount of trim change reduces in "degrees" but stays constant in "%".

BMatthews

Feb 13, 2007, 12:35 AM

Running your design through one of the neutral point calculators that are online would be a good place to start. From there you can work your CG back towards this point in small steps and perform dive tests and generally fly the model until you decide that it's as far back as you can live with it. For 3D aerobatic models this can often be at or even a smidge behind the NP. But for gliders that are often flown at speck ranges a slight but definite bit of positive stability is a good thing even if it means giving up that last 2% of efficiency.

This is intended as a practical side to the good stuff that I'm reading above.

Ollie

Feb 13, 2007, 08:35 AM

Measuring incidence so that the sailplane is trimmed in pitch is one point of view.

My point of view is:

When a symphony orchestra prepares to perform, just before the performance, it tunes up. This is so the various instruments can harmonize with each other to produce beautiful music.

A pilot, plane, launching device and radio are a system where the parts have to be adjusted to each other in a harmonious way to make beautiful flights. When the parts are not adjusted to each other, do we blame the parts or the adjustor? Often two people will have differing opinions about a particular design of plane because they attribute the differences to the design rather than to differences in the way the otherwise identical planes were adjusted to match the systems they were being flown in.

There is a wrong headed point of view that is running rampant. It is that the way a plane flies is entirely due to the design and to the manufacture. I think of this as the "appliance" mentality. It appeals to a lot of people who only buy ARF's and who expect the plane to perform properly when assembled according to the instructions without any further effort on the part of the owner. These people have generally not informed themselves of the affects of various adjustments. This approach can only result in a relatively safe but mediocre performance.

For best results the CG location has to be adjusted to the pilot's ability and flying style. The decalage has to be adjusted to the CG location and to the pilot's skill and flying style. The tow hook location has to be adjusted to the pilots skill, the launch technique and launch equipment. The control throws have to be adjusted to the pilot's flying style. Furthermore the adjustments have to accomodate the flying conditions and the particular purpose of the flight to be made. The adjustments will be different if nonobjective, relaxed flying rather than contest flying is the order of the day.

The CG location is the only adjustment that affects stability. Therefore CG location should be used to establish the desired stability and for no other purpose! To set the trimmed flight speed use the elevator trim setting on the transmitter or by adjusting the decalage (angle between the wing and horizontal tail). If the elevator neutral is too far off geometric center, then elevator offset can be replaced by a decalage change to get the elevator neutral closer to geometric center.

When the CG is too far aft, the plane can't porpoise because the flight path diverges and there is no correction to that divergence when the contols are neutralized. Porpoising results from stability. The more stable the plane the harder it will porpoise. Porpoising results from a gust or abrupt control input that disturbs the trimmed flight path and speed. A stable plane reacts to the disturbance by correcting but because of inertia the correction over shoots and a decaying oscillation results if the disturbance is small and the plane isn't driven to the stall. If the disturbance is large enough that the plane is driven to the stall, then the porpoising does not die out with the controls in neutral and the oscillation can grow without limit if the pilot does not intervene. If the plane is trimmed to fly too close to the stall, then even a small disturbance will drive the plane into a stall and the oscillation will grow if unchecked by the pilot.

If, in the unlikely event that an unstable plane is driven to stall, it will only stall once. It will not recover from the dive that follows the stall without pilot intervention.

If the plane is trimmed to fly too close to the stall, then even a small disturbance will drive the plane into a stall and the oscillation will grow if unchecked by the pilot. Take the extreme example of the LSF eight hour slope task. The ability to fly continuously for eight hours depends on a light pilot work load and adjusting with enough forward CG to do it is essential. On the other hand, it the task is seven minute precision duration that depends on finding and working thermal, then a more aft CG results in a plane that signals lift better and the additional pilot work load can be easily tolerated.

poulsbobill

Feb 13, 2007, 09:17 PM

When the CG is too far aft, the plane can't porpoise because the flight path diverges and there is no correction to that divergence when the contols are neutralized.

Ollie,

Can you explain this ? I thought too far aft cg caused porpoising.

Thanks

Bill

Ollie

Feb 14, 2007, 04:01 AM

Bill,

Divergence (like a gust) causes porpoising with stable flight and requires an under damped frequency. If the over damped divergence then is no porpoising and any CG position.

When the CG is at the neutral point, the porpoising frequency goes to zero and the porpoising period goes to infinity.

I suggest you shift some mass under control of a servo under damped flight condition. As you move the CG aft some before neutral point and then at the neutral point, then you can see in flight and how the porpoise frequency changes with CG position under servo control.

ciurpita

Feb 14, 2007, 07:49 PM

this past summer i experienced something on the field that i've never really read about and addresses another aspect of decalage. i've often wondered why the more advanced guys often said they preferred the CG to be a bit aft of the neutral point. what i've read is that it helps them read the air better.

this past summer, i asked some experts to fly my plane and they suggested i remove some nose weight. i was very surprised when i found that the plane flew more slowly and mushy at times. i was forced to push the stick forward, but could later relax it. it became obvious that this occurred when in sinking air, which because the CG was behind the neutral point, caused the plane to increase it's angle of attack slightly. (of course in lift, the plane picks up speed, and a complementary stick adjustment is needed).

while this is certainly not the hands off type of flying characteristic that is often desired, it clearly helped me see if the plane were flying in lift or sink. this was a real eye opener for me.

to get a feel for how far back the CG can go, i experimented with a small hand launch where i had some lead taped on the outside of the plane. i kept moving the lead back further and further and noticed how the flight characteristics changed.

Ollie

Feb 14, 2007, 08:50 PM

ciurpita! I have not seen you in this hobby for a few years. Your work is useful to me and to poeple in the hobby. Thank you very much.

vintage1

Feb 14, 2007, 09:13 PM

Not sure what the Neutral Point is..but you can sure have the CG aft of the wing center of pressure and have a fully lifting tailplane.

In fact for a high performance sailplane, you get better lift to drag if you do..
Stability only requires that the average center of pressure moves forward as speed increases.

I guess that requires some positive decalage..unless you had a manically reflexed wing maybe..

JetPlaneFlyer

Feb 15, 2007, 01:26 AM

Ciurpita,
I’d be pretty confident that you did not have the CG behind the neutral point. If you did you would have a box of bits by now. What makes you think that the CG was behind the neutral point anyway?... How did you work out where the neutral point was? The very fact that you could fly the model at all indicates to me that the CG was forward of the neutral point. I suspect that there is some confusion with your 'advanced guys' between neutral point and aerodynamic centre (a common mix up). Aerodynamic centre is at 25% chord, it is quite possible to have the CG behind this point, no problem. The neutral point however would usually, for a conventional layout, be further back, typically perhaps 40-50% chord, in some designs however it could be aft of the wing TE... You can work out approx Neutral Point location with this calculator: http://adamone.rchomepage.com/cg_calc.htm

It would be interesting to run your model through this calculator and see where your CG comes out in relation to the calculated Neutral Point.

Steve

ciurpita

Feb 15, 2007, 07:54 AM

Work in Progress

Feb 15, 2007, 10:05 AM

Not sure what the Neutral Point is..but you can sure have the CG aft of the wing center of pressure and have a fully lifting tailplane.

In fact for a high performance sailplane, you get better lift to drag if you do..

I still don't believe the second sentence of these two. The first I buy.

The more you move the CG back, the more tail area you need. Your starting point is, let's say a glider with a 180 sq in wing and a 20 sq in tail. As you move the CG back, for a given pitch stability and horizontal area, you have to enlarge the tail and shrink the wing. Eventually you reach the tandem-wing case, 100 sq in each end. Then if you carry on moving the CG back you make the rear lifting surface bigger still and shrink the front surface, and you end up with the canard case, 20 sq in at the front, 180 sq in at the back. OK, all fine and dandy.

But as the tail gets bigger, as we move along the continuum from the "conventional" to the "tandem wing" case, the L/D efficiency must go down, because you're losing aspect ratio and/or Reynolds number rapidly from the more efficient surface, and making the more inefficient surface do more of the work.

What you want for best L/D is as much of that area as possible in one surface, and as little as possible in the secondary surface, which should be doing as little work as possible. Which is why the world's leading competition sailplanes and the best airliners all have teeny-weeny tails in relation to their wings. (The airliners DO have far bigger tails than the gliders, to give them a wider allowable CG range as the punters thunder up and down the fuselage and the baggage handlers stuff the cases containing the divers' gear in silly places, and to enable them to generate enough pitch moment to cope with the massive out-of-trim forces that come from deploying all the slattery and flappery, but they are still no bigger then they need to be.)

JetPlaneFlyer

Feb 15, 2007, 12:02 PM

you can read about the neutral point in the march 2006 issue of RC soaring digest
http://www.rcsoaringdigest.com./pdfs/RCSD-2006/RCSD-2006-03.pdf

you can choose to disagree with me. i don't think i ever did calculate the actual CG and neutral point, and i'm not saying you are far behind the NP. what i thought was more important was how dramatic an affect there was on the aircraft.

again, i suggest to try playing with the CG on a small hand-launch and get a feel for how it affects the aircraft.

The article you refer to is interesting... if you go to page 38 you will find this statement "Static stability requires that Cma be negative, that the CG be in
front of the neutral point" and later on the same page: "for a conventional tailed aircraft, stability requires that the CG be in front of the neutral point"

A statically unstable model (i.e. one with the CG aft of the neutral point even by a little) will be un-flyable.

Neil Walker

Feb 15, 2007, 01:11 PM

I still don't believe the second sentence of these two. The first I buy.

The more you move the CG back, the more tail area you need. Your starting point is, let's say a glider with a 180 sq in wing and a 20 sq in tail. As you move the CG back, for a given pitch stability and horizontal area, you have to enlarge the tail and shrink the wing. Eventually you reach the tandem-wing case, 100 sq in each end. Then if you carry on moving the CG back you make the rear lifting surface bigger still and shrink the front surface, and you end up with the canard case, 20 sq in at the front, 180 sq in at the back. OK, all fine and dandy.

But as the tail gets bigger, as we move along the continuum from the "conventional" to the "tandem wing" case, the L/D efficiency must go down, because you're losing aspect ratio and/or Reynolds number rapidly from the more efficient surface, and making the more inefficient surface do more of the work.

What you want for best L/D is as much of that area as possible in one surface, and as little as possible in the secondary surface, which should be doing as little work as possible. Which is why the world's leading competition sailplanes and the best airliners all have teeny-weeny tails in relation to their wings. (The airliners DO have far bigger tails than the gliders, to give them a wider allowable CG range as the punters thunder up and down the fuselage and the baggage handlers stuff the cases containing the divers' gear in silly places, and to enable them to generate enough pitch moment to cope with the massive out-of-trim forces that come from deploying all the slattery and flappery, but they are still no bigger then they need to be.)

I'm not quite sure I understand this reasoning (that the further back you move the CG the more tail area you need). If you've got a design (as vintage1 was suggesting) where you can move the CG aft of the centre of pressure and get a lifting tail, surely there is some point where the tail doesn't need to do any lifting at all. You can have: CG ---- CoP ---- NP, where you have a pushing tail, you can move the CG back past the CoP to get: CoP ---- CG ---- NP and now you have a lifting tail. Somewhere in between those two (which is a very aft CG that may not even be reachable for some desings) the tail doesn't have to do any work and can thus be as small as possible. Isn't that the point vintage1 was referring to to minimize drag? What am I missing?

Work in Progress

Feb 15, 2007, 02:02 PM

Hi, Neil,

I'm sure V1 will speak for himself and I don't want to inadvertently misrepresent his thinking, 'cause he's good at thinking...
... but what I THINK he meant is it's more efficient to get the tailplane to consistently contribute to (upward-pushing) lift in a steady-state cruise at best L/D, than to have it doing as little work as possible, trimmed for min drag, and to put as many of your square inches as you can into the main wing.

If I'm wrong about what you meant, Vinto, I take it all back, and please correct me. if that IS what you meant then that's where I am not convinced.

All other things being equal, if you have "just enough" pitch stability and you want to bring the CG back to load the tail to make it contribute lift consistently at best L/D, which is generally quite a high angle of attack for a high aspect ratio wing, then you need to make the tail bigger in order to maintain pitch stability.

A good big wing is always more efficient than a good little wing, and the main wing should spend more of its time doing heavy lifting than a stability / control surface, so my contention is that for best L/D overall you should not try to make the tailplane routinely provide lift at the best L/D cruise.

Neil, I'm not sure whether I'm disagreeing with you or not, my brain hurts after a 12 hour working day and I have to go and find a nice quiet place to lie down :rolleyes:

Neil Walker

Feb 15, 2007, 03:23 PM

A good big wing is always more efficient than a good little wing, and the main wing should spend more of its time doing heavy lifting than a stability / control surface, so my contention is that for best L/D overall you should not try to make the tailplane routinely provide lift at the best L/D cruise.

Neil, I'm not sure whether I'm disagreeing with you or not, my brain hurts after a 12 hour working day and I have to go and find a nice quiet place to lie down :rolleyes:

I think we're agreeing on what you ideally want the tail surface to do, I just wasn't sure about the "further you move the CG back, the bigger the tail has to be" bit. But I see that in the context of how you interpreted V1's comment, that makes sense. It's all good!

I do love these threads. Lots of experienced folks dumping knowledge out into the forums, and every once in a while a little light switches on for me and I end up understanding more about how my gliders fly - sometimes I even learn something that helps me fly them better. Now about my helicopters... ;)

vintage1

Feb 16, 2007, 08:18 AM

I'm not quite sure I understand this reasoning (that the further back you move the CG the more tail area you need). If you've got a design (as vintage1 was suggesting) where you can move the CG aft of the centre of pressure and get a lifting tail, surely there is some point where the tail doesn't need to do any lifting at all. You can have: CG ---- CoP ---- NP, where you have a pushing tail, you can move the CG back past the CoP to get: CoP ---- CG ---- NP and now you have a lifting tail. Somewhere in between those two (which is a very aft CG that may not even be reachable for some desings) the tail doesn't have to do any work and can thus be as small as possible. Isn't that the point vintage1 was referring to to minimize drag? What am I missing?

The fact that the tail still has to do work. Not statically, but dynamically.

Forgetting the neutral pint, which seems to be defined as 'the point beyond which noving the CG aft will result in instability' my thinking goes soemwaht like this.

1/. The condition for stability is that the center of lift moves forward with increasing speed. (or the CG moves back!!) Obviously for level flight it must be at the centre of gravity.

2/. In general that is achieved by running the foreplane at a bigger angle if incidence than the rear plane. This covers all cases - including canards. Flying wings are to be considered as the tow planes being extended till they meet.

3/. For lowest drag, the tail may well be parasitic..at high speeds you probably want the tail to be thin and more or less at zero incidence to the airflow, to minimize induced drag, and as small as possible. At low speeds you may well find that haveing the wing and tail generating lift, is worth the extra drag penalty from the tail. Hence long tails on glider, with large tail areas.

4/. I consider airliners to be slow aircraft..they will be optimised for lift to drag, not lowest overall drag.

5/. warplanes are the opposite..they need manouverability, and high top speed. Hence short tails, small tailplanes and a forward CG work better.
.

6/. I haven't yet worked out why a foreplane at say 2 degrees AofA will generate more lift incrementally than a tail at +1 degree AofA with increasing speed..but obviously a tailplane at zero degrees generates no lift, but the wing lift will increase with increasing speed, generating the required forward movement of the center of lift..and giving a nose up moment with increasing speed.

7/. I haven't either worked out the drag penalty of e.g. a small tailplane at negative angle of attack versus a larger one at say zero angle of attack. I suspect that the former is as I have said better at low speeds and the latter at high,.

vintage1

Feb 16, 2007, 08:32 AM

I think we're agreeing on what you ideally want the tail surface to do, I just wasn't sure about the "further you move the CG back, the bigger the tail has to be" bit. But I see that in the context of how you interpreted V1's comment, that makes sense. It's all good!

I think the simplest way to look at it is, to see that the condition of stability is that in a dive, the model goes faster, and you want to use that speed to generate a pitch up moment on it to bring the nose back up.

Now the shorter the tail the less righting moment a loss of lift or extra downforce will have. So the shorter the tail moment the bigger the tail has to be Or the more negative force is needed on it in the first place. Either will net you an increase incremental downforce on the tail.

THEN with all that extra nose up moment on the tail so that it CAN work over its small size/short distance, the CG must go forward to compensate.

That leads naturally to the idea that a flying wing has a very forward CG (typically 15% MAC) and a very reflexed TE. Those of us who flew combat wings in control line will recognse this. Its a VERY fast turning model layout indeed. You will also find that e.g. WWII planes have this arrangement - tails are small, and CG's well forward - typically 20-22% of MAC ..

Likewise when you have limited total surface area, and want to get the most lift out of it for the least drag, you SEEM - for reason I am still not clear about - to end up with huge tails on long booms..and a CG aft of the trailing edge of the main wing..many contest free flighters ended up like this. And I can attest that they had VERY lazy stability. If they ended the power run in a dive..they tended to stay in it a LONG time..BUT once settled into a stable state, they floated forever..

I think this is also the best arrangement for a low sink glider, but with lifting tails, they have a very low glide speed, and lack penetration in wind.

Again I suspect airliners are almost like this. They are not highly manouverable, but they are fuel efficient in cruise..

I do love these threads. Lots of experienced folks dumping knowledge out into the forums, and every once in a while a little light switches on for me and I end up understanding more about how my gliders fly - sometimes I even learn something that helps me fly them better. Now about my helicopters... ;)

Ollie

Feb 16, 2007, 09:07 AM

A fuel powered model carries fuel mass that is used up. If the fuel tank is not position on the CG, the model's CG position shifts. When the model's CG shifts due to used fuel or surges of fuel fore or aft, then the model's stabilty changes. If the tank is large, then it is a good thing to baffle the tank to prevent fuel mass surges. If the tank is large, then the tank position placed near the model's CG.

yoyoML

Feb 16, 2007, 10:11 AM

I think there are a few different ideas mixed up here.

The neutral point is suposed to be the point around which the total torque (in pitch) remains constant with a small variation in AoA. It is similar to a wing's aerodynamic center, only it speaks for the whole plane. I may be wrong but this is my understanding.

To vintage1:
"Stability" concerns these two things:

Static stability
A small increase in AoA results in a small decrease in pitch up torque, which tries to reduce the AoA back. It is exactly the same as a spring: a small increase in length results in a small increase in tension, that tries to pull the spring back.

Dynamic stability
As a plane oscillates in pitch, like how a mass attached to a spring oscillates, the amplitude decreases over time without pilot intervention. In short, the oscillation is damped.

All this is quite similar to mass-spring-damper systems in textbook examples.

The "spring constant" or "spring rate" for the static stability is (approxmiately):
(total wing area)*(static margin)*(CL slope)*( 1/2 * rho * v^2 )
The first 3 terms are specific to plane configuration.

The damping constant I haven't really worked out, but I think it is positively correlated to (tail length)^2 * (stab area) in something like gliders. Hence the long tails.

To ciurpita:
I don't think the plane had the CG behind the neutral point. Maybe very close, but not on or behind it. Because if the CG were behind the neutral point, any small increase in AoA results in more pitch up torque which in turn results in more increase in AoA... until it diverges. If the neutral point were determined by simple calculations instead of experimentally measured, it could be the real neutral point is further back than calculated.

Indeed what I said is consistent with p.38 of the linked aritcle.

To JetPlaneFlyer:
The "25% point" is not the center of pressure, but the aerodynamic center which remains close to 25% MAC for a range of AoA.

To Work In Progress:
When the plane's wing config is fixed, it is actually more efficient that the stab provides some positive lift, by moving CG backward closer to the neutral point but not increasing stab area. It is because lift is proportional to AoA but drag is proportional to AoA squared.

Imagine the tandem case:
a) Have the CG coincide with the neutral point, so that both wings have equal wingloading, and operate at the same AoA, alpha.

The total lift is 2*(some lift constant)*alpha, and the total drag is 2*(some drag constant)*alpha^2.

b) Have the CG coincide with the center of lift of the first wing, so that the first wing has double wingloading and the second has zero. The first wing then operates at 2*alpha, the second at zero.

The total lift is 1*(some lift constant)*(2alpha), and the total drag is 1*(some drag constant)*(2*alpha)^2.

In both a) and b) the lift is the same.

In b) the total drag is ~ 4*alpha^2, but in a) the total drag is only ~ 2*alpha^2, half as in b).

You can easily estimate this for other cases with arbitrary wing areas or even arbitrary number of wings, but you'll still find the more evenly you distribute the wingloading (with the config fixed, by moving CG), the lower the total drag.

JetPlaneFlyer

Feb 16, 2007, 11:00 AM

To JetPlaneFlyer:
The "25% point" is not the center of pressure, but the aerodynamic center which remains close to 25% MAC for a range of AoA.
Quite correct... that's me getting my terminology mixed up. I'll correct the original post to prevent confusion.

Steve

Ollie

Feb 16, 2007, 12:01 PM

JetPlaneFlyer

Feb 16, 2007, 03:10 PM

A very skilled pilot can fly with the CG about one percent of the MAC behind the neutral piont. The pilot must control the plane instantly before it diverges. The pilot needs his focus constantly on the plane's airspeed and pitch attitude.

A circus performer on the high wire is unstable. The performer has trained his very short reaction time. Some performer uses a weighted pole to make the system rotate slowly. Another performer uses an umbrela in his hand and the system damped in rotation.

Ollie,
I'm not sure anyone could ever say for sure that a CG was 1% behind the neutral point or not because not sure the neutral point can be calculated with any certainty to a tolerance of 1% :rolleyes:. If it can I’d like to know how.

Walking upright on two legs is unstable too but everyone masters it ;) Our balance system has evolved over millions of years to enable us to do this. This same balance system can be used by the pilot of a full size aircraft because the pilot is sitting in the aircraft and ‘feeling’ all it’s subtle movements. He has the benefit of his inner ear balance mechanism plus ‘seat of the pant’ feel. He is also assisted by stick feedback and a host of instrumentation. An expert full size aircraft pilot ‘may’ therefore be able to cope with neutral, or perhaps even very slight negative stability. Having said this all the refernces I’ve seen says it can’t be done without fly by wire computer assistance.

The 'pilot' of a model however has only his vision to deduce airspeed and attitude. His visual references need to be made at considerable distance; possibly in conditions of less than ideal visibility. He has no stick feedback and no instrumentation. His inner ear balance mechanism is of no use and he can’t feel the aircraft through the seat of his pants. It stands to reason therefore that a RC model needs a bigger margin of stability than may be tolerated in a full size aircraft. I’d be amazed if it was possible for anyone to fly an R/C with negative static stability for more than a few seconds, anyone who could would certainly have my undying respect.

Steve

Ollie

Feb 16, 2007, 05:48 PM

Steve,
I can't prove it to you. However, I have a strong opinion on this matter. I should have shut my mouth until I could prove it.

ciurpita

Feb 16, 2007, 08:26 PM

as i wrote this, i realized that my perspective is as a sailplane pilot, and what i consider as stable or neutral may be very different from a powered pilot. with this in mind ...

when discussing stability, there's this sense that if you violate the rule that the CG is aft beyond a certain point, that there is a very dramatic change in the characteristics of the plane that will make it very difficult, some say impossible to fly. and presumably pilots of different skil levels require different amounts of stability. if this is true, is there some way to calculate in some rough way how much stability a novice needs, versus an expert? (does it matter?)

are there flying conditions (wind, tubulence) where you would say a particular plane is not stable enough to fly straight and level without some human correction, but a different plane is stable enough? if you could quantify stability would you need to consider flying conditions to classify a design as stable enough (how quickly does it correct)?

i think that a plane may be technically stable, but difficult or unpleasant to fly because of other things. do you think a well trimmed plane could be challenging for someone with less skill, but enjoyable to fly, and would that person would gain skill quickly with that plane?

i've read that most expert SAILPLANE pilots fly planes that are neutral so that the plane is easily affected by the air allowing them to find lift more easily. but once in lift, they actively fly in circles, constantly correcting the circle as they see the plane being bounced arround by the turbulence to gain altitude. do you think they would have much of a problem flying the plane straight in level looking for lift?

i know that technically a +/- sign determines if a plane is stable or unstable. but i believe there is a range of CG positions equally forward and aft of the neutral point where the plane is more neutral than stable or unstable. forward of this range would be considered stable, and aft of the range unstable. i think that the type of aircraft (sailplane, powered, acrobatic, jet) determines that range and whether pilots prefer stable or neutral aircraft.

vintage1

Feb 16, 2007, 09:23 PM

You can fly a plane that is slightly unstable in pitch..i've had one model trimmed such that if you pushed the nose down..it went into a slightly steepening dive..needles to say it was a very nervy plane, and it flew up and down as I tried to correct it..fortunately it got a bit tamer as power was reduced, and I manged to land it reasonably. Needless also to say that some lead went in the nose before it went up again.

Total neutrality is nice if you want to fly pattern style aerobatics..I had a foamie Alliance as an early plane..it was really more than I could handle though, but that stayed in whatever attitude you put it. Nice design.

Ollie

Feb 16, 2007, 11:28 PM

See:
http://www.charlesriverrc.org/articles/asfwpp/helmutlelke_asfwpp.htm
http://www.charlesriverrc.org/articles/asfwpp/lelke_passivepitch.htm
http://www.charlesriverrc.org/articles/asfwpp/lelke_activepitch.htm
http://www.charlesriverrc.org/articles/asfwpp/lelke_practicalandservos.htm
"Prior to the summer of 2001 the farthest aft CG I achieved was around 28%. I had hit a brick wall. Any attempts to bring the CG farther aft resulted in loop instability. Up to this point I had primarily used the Hitec HS-225BB servos believing that it was fast enough for the application. Then, by accident, I discovered a deep dark secret about RC servos - the so called "high speed" servos in general don't meet advertised speed specifications! Some of them don't even come close!"

And

"Both approaches gave the same disturbing results - "High Speed" servos don't come close to meeting the advertised speeds. All servos were driven with 4 and 5 cell subC Nicad packs. Speed measurement results are listed in Table 1 for the 16 servos tested. The fasted servo proved to be the Airtronics 94145 driven with the 5 cell Nicad pack. The advertised speed for this servo is 0.07 seconds with 6 volt drive. The measured speed was 0.1 seconds with the 5 cell Nicad (6.8V) - significantly below specification."

I have a flying buddy who uses an aft CG position and no active controls. He flies his sailplane model like it is on rails. He flies while he dances with his thumbs on sticks. His reacation time seems to be less than 0.1 second.

My reaction time is about 0.4 second. I couldn't fly with his sailplane at all. His flying skill is amazing.

yoyoML

Feb 17, 2007, 01:00 AM

To ciurpita:
There is not a "dramatic" change in flight characteristics as the CG moves backward through the neutral point. It's just that the "spring constant" goes from positive to zero to negative continuously, with a gradual loss in static stability.

About flying a statically unstable plane, I think it depends on a few things.

The elevator needs extra authority (mainly deflection) to push the nose back after reaching a high AoA, because the plane doesn't want to right itself. And indeed, for an unstable plane to maintain a "positive" AoA, you actually apply a constant "down" elevator, after an instant of "up" elevator to provoke it.

I remember the F-16 can go into a cobra, only it doesn't have the elevator to get out of it.

The total response time of eye->brain->thumb->servo rotation->linkage slop must be significantly lower than a "characteristic time" of the plane. This "characteristic time" may be taken as the oscillatory period on a stable plane, but on an unstable plane I'm not sure if there's a good definition. For the moment, let's just take it as the time to spontaneously go from 1 degree AoA to stall.

This time should increase with pitch moment of inertia, pitch damping; it should decrease with a more negative static margin, higher airspeed. If I ever figure out how to calculate such time, I'll post ;)