Feb 15, 2013, 04:06 PM
Registered User
Joined Jun 2005
4,891 Posts
Quote:
 Originally Posted by John235 Hi Don, I'd like to ask your opinion about an alternative explanation for this problem of canard turning stability. My suggestion is the problem actually starts as a dutch roll condition that is exacerbated by low dynamic yaw stability in a canard model.
So, too much dihedral and not enough fin...

Quote:
 In other words, a lack of damping allows the dutch roll oscillation to initiate a tip stall of the main wing.
Such an event would result in a sudden, and probably violent departure in roll, not the more steady oscillatory rocking of dutch roll.

Quote:
 Even if a canard model has just about enough static yaw stability, the damping may still be too low due to the short tail moment. Do you think the addition of a forward fin can actually increase the dynamic yaw stability in a helpful way in this scenario?
I doubt it.

Lets go back to what's going on when damping occurs. In "Control Theory 101", "damping" is the force that tries to stop oscillations, and the amount of damping force is proportional to how fast the objects that are oscillating are moving. In this case, the amount of damping is proportional to how fast the nose is swinging side to side.

In these cases, the motion we are looking at is typically something resembling a sine wave, something like a pendulum. This kind of motion has maximum deflection from center at both ends of the swing, and at those locations, as the travel is changing from away from the center to back towards the center, the velocity is zero.

Conversely, when the deflection is zero (as the object is at the middle of travel), the velocity is maximum. Therefore, the traditional "damping" force (which is in most typical cases proportional to the velocity) is also maximum at the center of travel.

In your scenario, yes, a fin on the nose would make additional damping force as it is swinging through the center of the yawing oscillation.

HOWEVER, you describe an airplane with "just about enough static yaw stability", which I interpret as "positive but barely adequate", maybe even mildly inadequate static yaw stability. In that case, the additional fin added to the nose is going to make the marginal static stability worse. The forces it makes at the ends of each oscillation will try to make the oscillations bigger. When the plane oscillates in yaw, it will probably oscillate at a lower frequency (which will tend to reduce the additional damping effects from the added fin), but it will oscillate farther to each side due to the weakened static yaw stability. If the plane has significant dihedral effect, it's also going to see bigger oscillations in roll, due to the larger oscillations in yaw. Altogether, I doubt the results will be quite as desirable as something that deals with the original problem directly.
Feb 16, 2013, 10:40 AM
Registered User
Joined Jun 2005
2,610 Posts

How about Equaduck for the name of my new model? My views on the advantages of the design follow:

The CG is at the center of the fuselage for balanced loading and to provide adequate lever arms for all control surfaces.

The canard lift with a 12 inch distance from the CG equals the wing lift with an 8 inch distance from the CG.

Canard area is 51% of the main wing area which at our Reynolds numbers should produce improved performance when compared to the 30% version.

Since the Equaduck is a 93% low wing replica of the Twin Canard, it should duplicate the natural tendency of the Twin to do nose up three point landings.

Because the lifting moments of both wings are equal about the CG and fore and aft inertias are near equal, the model should be able to recover from a power dive with less probability of a high speed stall than a conventional design.

The prop blast is directed toward the fin and rudder for good ground control with a 20 inch rudder lever to the CG.

The thrust line of the motor which is located on the horizontal CG and is about one inch above the vertical CG passes through tha canard's center of lift and should torque the nose down for better penetration at high power.

Charles

# Images

 Feb 16, 2013, 11:06 AM What could possibly go wrong? Market Harborough Joined Apr 2006 4,354 Posts She flies! Wahoo! And looks fabulous in the afternoon sun... First sun I've seen seen since 6th Jan. The plane is nowhere near to finished as you can see but, when you get a day like today, you've just got to get out there. She's the sweetest plane to fly, first time I've been able to manage consistent greasy smooth canard landings. The wing loading is less than 10 oz /sq ft. I think I noticed a little yaw slip on climbing out, but the stall was nice and straight and she glides very steadily down to a very slow speed. After trimming the canard is scarcely as much as +1 deg relative to the wing. On the first flight I had a fright coming in to land...CG definitely too far back but a blip of power was enough to get out of trouble. CG is now about an inch further forward. Har-di-har-har.
 Feb 16, 2013, 11:11 AM Registered User Hampshire, U.K. Joined Jun 2008 2,264 Posts Well done, Nick! I'm sure the success is all the more satisfying having got at least part of the big starship to fly at last!
 Feb 16, 2013, 11:15 AM What could possibly go wrong? Market Harborough Joined Apr 2006 4,354 Posts Charles, hi! Our posts must have been almost simultaneous. Lovely to see your pictures. Excellent work and original ideas as always. Cheers
 Feb 16, 2013, 11:18 AM What could possibly go wrong? Market Harborough Joined Apr 2006 4,354 Posts And Trevor too! We're all at it today. Hi Trevor PS you know what trouble I have with retracts. That flimsy little front leg works really well this time. I've used one of the old fashioned retracts with a separate servo and it locks down nicely. Last edited by nickchud; Feb 16, 2013 at 11:25 AM.
Feb 17, 2013, 05:45 AM
Registered User
Sydney, Australia
Joined Mar 2006
1,316 Posts
Quote:
 Originally Posted by Don Stackhouse In your scenario, yes, a fin on the nose would make additional damping force as it is swinging through the center of the yawing oscillation. HOWEVER, you describe an airplane with "just about enough static yaw stability", which I interpret as "positive but barely adequate", maybe even mildly inadequate static yaw stability. In that case, the additional fin added to the nose is going to make the marginal static stability worse. The forces it makes at the ends of each oscillation will try to make the oscillations bigger. When the plane oscillates in yaw, it will probably oscillate at a lower frequency (which will tend to reduce the additional damping effects from the added fin), but it will oscillate farther to each side due to the weakened static yaw stability. If the plane has significant dihedral effect, it's also going to see bigger oscillations in roll, due to the larger oscillations in yaw. Altogether, I doubt the results will be quite as desirable as something that deals with the original problem directly.
Don, Thanks for your explanation. I think I read on your website that its possible to create an aircraft that has spiral instability and also suffers from dutch roll oscillations. My 60" canard slope soarer has an apparent problem in turning stability. When I attempt a thermal-style turn at high alpha and bank angle, sometimes the model will suddenly and uncontrollably yaw into the turn and go into a dive. I probably provoked it by using the rudder to assist the turn. At the time I assumed it was a spiral stability problem, but now I'm not really sure.

If its a spiral stability problem I might attempt a fix by reducing the tail fin area, but I worry about ending up with a dutch roll problem. Part of the problem may be due to the sweepback used on the main wing. If I build another version it will have less sweepback. I would also make some changes to the airfoil and panform, because I am not sure if the model has enough safety margin to make the main wing resistant to stalling.

Do you think it reasonable to expect a purpose built forward fin will add more damping than a similar increase in fuselage lateral area ?
Feb 17, 2013, 08:44 AM
Registered User
Joined Jun 2005
4,891 Posts
Quote:
 Originally Posted by John235 ... I think I read on your website that its possible to create an aircraft that has spiral instability and also suffers from dutch roll oscillations.
Yes. Sweep in particular can cause that, because of the change in dihedral effect with change in angle of attack. Too much dihedral effect at high alpha, not enough at low alpha.

The size of the "sweet spot" between the two problems seems to be related to the mass in the extremities, and the resulting effect on roll and yaw inertia. As the inertia increases, the amount of energy that needs to be dissipated by the available damping effects increases, making the dutch roll problem more difficult. With enough mass in the extremities, even a conventional layout can have both problems simultaneously. It's possible to design a truly atrocious airplane from essentially any starting concept!

Quote:
 My 60" canard slope soarer has an apparent problem in turning stability. When I attempt a thermal-style turn at high alpha and bank angle, sometimes the model will suddenly and uncontrollably yaw into the turn and go into a dive. I probably provoked it by using the rudder to assist the turn. At the time I assumed it was a spiral stability problem, but now I'm not really sure.
The fact that it happens suddenly, particularly when the turn is rushed with a little too much rudder input, suggests a tip stall. Spiral instability typically manifests as a steady, gradual tendency to overbank into a spiral.

Quote:
 If its a spiral stability problem I might attempt a fix by reducing the tail fin area, but I worry about ending up with a dutch roll problem.
It can be a delicate balance. However, it sounds to me more like a tip stall. Washout increases can help that in some cases, but if you have a very light wing loading, allowing an extremely tight turn (some of our 1.5 meter RCHLG sailplanes can make thermal turns, with a reasonable bank angle, that results in the inside wing tip drawing out a 3 ft diameter circle), the difference in airspeed between the inside and outside wingtips can be more than 2:1, so the inside wingtip is required to make more than four times the lift coefficient of the outside wingtip! This could also be significantly more than the lift coefficient at the root, and with a small fraction of the Reynolds number. Dealing with that really cannot be done with washout (unless you use massive amounts of it, to the point that the lift distribution is badly distorted), so about the only decent option at that point is differences in airfoils along the span. Even then, an airplane with a light enough wing loading (and therefore a tight enough turning radius compared to its span) can be forced to tip stall if you push it far enough. There is only so much you can do, even with airfoil changes.

Quote:
 Part of the problem may be due to the sweepback used on the main wing. If I build another version it will have less sweepback. I would also make some changes to the airfoil and panform, because I am not sure if the model has enough safety margin to make the main wing resistant to stalling.
It definitely appears that it doesn't!

Sweep also causes a tendency for tip stalling. The sweep causes spanwise flow that has the effect of reducing the loading and lift coefficient at the root (the "lift valley"), and a small increase in the loading at the tips, making the tips prone to stalling first. In addition, that same spanwise flow causes a stall that initiates inboard on the wing to almost immediately propagate outboard. Getting good stall characteristics from a plane with significant sweep can be a tricky proposition.

Quote:
 Do you think it reasonable to expect a purpose built forward fin will add more damping than a similar increase in fuselage lateral area ?
Thin shapes (such as a fin, or a profile fuselage), will have more lateral effects than a square-cornered box, which will have more lateral effects than a round-cornered box, which will have more than a round fuselage, which will have more than one with smooth chines (like the SR-71).

Although they might have more damping, the most powerful effect is likely to be the decrease in static stability. Fins (or things with fin-like effects) on the nose are generally a very bad way to try to increase yaw damping.

In the case of your slope soarer, the amount of sweep is so small that its effects on general handling are likely to be negligible. Until you get up around 10-20 degrees per side, sweep is not likely to play much of a role in overall behavior. Your description sounds much more like a classic tip stall problem, not spiral instability.

One easy thing to try that may help, or at least indicate if tip stalling is a likely culprit, is to add a turbulator at about 12-15% chord on the outer third of the wing. At low Re's, turbulators can significantly improve max lift coefficient and stall resistance. It's easy to try, just a few layers of trim striping tape. If they don't help, peel them off and try again in a different chordwise location, or try something else. Martin Hepperle's website http://www.mh-aerotools.de/airfoils/index.htm has a good article on turbulators, including how to calculate how high they need to be.
Last edited by Don Stackhouse; Feb 17, 2013 at 09:06 AM.
Feb 17, 2013, 10:47 AM
Registered User
Joined Jun 2005
4,891 Posts
Quote:
 Originally Posted by canard addict John, From looking at the top view of your slope glider it appears that the CG will not be far enough from the wing's LE to allow adequate performance of the canard.
Sorry, Charles. I just don't see that as a relevant explanation of his problem. That parameter is not a valid measure of just about anything. It's another one of those rules of thumb that don't actually mean anything.

For maximum performance with a canard, you need to have a reasonable Vht (around 0.5 or so), but with a very small canard (about as small as you can get without getting into too-low-Re problems) on a long moment arm. This makes the wing the primary lifting surface, with the canard only providing pitch stability, same approach as with an aft tail. The C/G with this arrangement will be well aft, probably behind the wing leading edge by a small amount, although that is not what matters. What matters is the distance from the C/G to the entire aircraft's aerodynamic center ("AC") in percent of the wing's mean aerodynamic chord ("MAC"). We call this the "static margin". Where all of that falls relative to the wing's leading edge depends on an number of factors, but the end result relative to that leading edge is pretty much irrelevant with regards to static and dynamic pitch stability.

With that small-canard-long-moment-arm setup, it is very possible to have excellent static and dynamic stability in pitch. I have one that has a canard area less than 16% of the wing area that has outstanding static and dynamic pitch stability, in fact I could get away with making the canard quite a bit smaller, even without increasing the moment arm.

Of course none of that says much of anything about yaw or roll.

Quote:
 I can see why you may want to add a fin up front or more fuselage area to prevent nose drop in a banking turn.
YOU DO NOT NEED "FIN AREA" ON THE NOSE TO "HOLD THE NOSE UP" DURING A TURN!!!!

Ideally, in a properly coordinated turn, the airflow at the fuselage is lined up with the relative wind (there is an exception which I will discuss in a moment). In that case, there is no angle of attack of significance to cause that fin on the nose to make any side forces.

The exception is in an extremely tight turn, such as a max-effort thermal turn with a very lightly loaded airplane, so that the turn radius is small compared to the length of the fuselage. In that situation, the airflow past the plane is curved by the turn, so that if the relative wind is lined up with the fuselage at the wing, it is trying to yaw the nose away from the turn and pitch it down, while the tail is being pushed into the turn and down. In those situations the local airflow is trying to make the airplane pitch down and yaw away from the turn, so that any dihedral effect will then try to roll it back towards level flight.

In those situations you generally need to apply up elevator, and rudder into the turn, just enough to cancel out these effects. Meanwhile, the rolling moments of the left and right wings are equal, and their combined lift vector is perpendicular to the wings. Gravity and centrifugal force are in balance, so the plane (and any occupants inside) "feel" like the plane is in level flight. People and airplanes cannot tell the difference between centrifugal force and gravity, or some combination of the two. There is no need to "hold the nose up". You are not trying to make sideways forces, other than to cancel out the curvature of the flow in extremely tight turns, and those corrections are in the opposite direction, into the turn, not away from it.

Quote:
 You seem to have about 15 to 20 percent canard area to wing area. My inclination would be to raise the canard area to about 40 percent with no less than 4.5" wing tip chord and maybe a 6" center chord. This will move the CG forward but will also decrease the canard's moment arm. The short coupling may cause yaw instability in spite of having increased rudder leverage.
Well, um...NO.

Moving the C/G forward will increase the vertical tail moment arm, and decrease the nose's moment arm, both of which will increase yaw stability. However, moving the C/G forward does not change the moment arm between the AC of the wing and the AC of the canard. The dynamic stability in pitch will increase because of the additional canard area, but not due to the difference in the distances of he canard and the wing from the C/G. In any case, that isn't the problem.

Quote:
 The added canard area should add enough vertical lift in a banking turn to keep the nose from dropping.
The canard, even if it has dihedral, does not make "vertical lift" if the turn is properly coordinated. It makes lift perpendicular to the canard, same direction as the lift from the wing.

It means that less of the plane's weight is carried by the wing, and more of it will be carried by the smaller, less efficient canard, so the plane's L/D will get worse. The plane will still have to satisfy the requirements for equilibrium in pitch, so after retrimming for the effects of the newly sized canard, there will be no change there.

His problem is that in tight turns, particularly when "forced" by an excess of rudder, the plane is suddenly dropping the inside wingtip and rolling steeper into the turn. It's pretty obviously a simple tip stall (that's what the "suddenly" tells us). It has essentially nothing to do with "holding the nose up".

Quote:
 I made a similar change to my original Georgia Goose but to stop the wobble the fuselage had to be lengthened forward of the main wing....
The changes you saw were principally due to the longer moment arm, which increased the dynamic stability. Dynamic stability is proportional to the SQUARE of the moment arm, but only linear with area.

Quote:
 ...I have found that on a top wing conventional design with a long rear fuselage that a small horizontal tail is only adequate at higher speeds. As speed is reduced, the tail will drop, the AOA of the main wing increases and stability is jeopardized. Therefore I feel that large tail areas are favorable. I also feel that large canard areas are better....
Again, NO. What you describe indicates a C/G problem, causing the tail to stall before the wing, because of too much lift being demanded from the tail. You should also be seeing poor static pitch stability in that situation. The exception is that if you have something like very large flaps on the wing, and the tail is not big enough to overcome their large aerodynamic pitching moments before the tail stalls. That's a special case. It is a control authority problem, not a stability problem related to a plane with variable geometry (flaps in this case).

Other than Reynolds number issues, there are no stability-related lower limits on practical tail size. Flying wings (when properly designed) can be stable both statically and dynamically, with a tail area of ZERO.

In any case, none of that will have any meaningful effect on John's tip stall problem.

John, as far as yaw stability, there are a couple factors against you on this model. The struts for your parasol-mounted canard have some fin effect of their own, as do the turned-up tips on your canard. The fat, square forward fuselage also hurts. However, your extended aft fuselage and fairly large fin should offset this. Without running any exact numbers, my gut feel says your yaw stability should not be great, but is probably adequate. The lack of dihedral might set you up for a spiral instability situation, but not a strong one, just a gentle tendency to overbank. Your observation of a sudden wing drop indicates a different problem, tip stall.
Feb 17, 2013, 12:02 PM
Registered User
Joined Jun 2005
2,610 Posts
Quotes:
Quote:
 Quote:Charles The added canard area should add enough vertical lift in a banking turn to keep the nose from dropping. Quote Don The canard, even if it has dihedral, does not make "vertical lift" if the turn is properly coordinated. It makes lift perpendicular to the canard, same direction as the lift from the wing. It means that less of the plane's weight is carried by the wing, and more of it will be carried by the smaller, less efficient canard, so the plane's L/D will get worse. The plane will still have to satisfy the requirements for equilibrium in pitch, so after retrimming for the effects of the newly sized canard, there will be no change there.
Don, With all due respect, I must add to my view of vertical lift of an aircraft. In straight and level flight the wing or wings must create enough lift to match the force of gravity and since the lift is vertical, there are no other components of the lift. If aileron is applied, the perpendicular lift breaks into two other parts which are horizontal and vertical componets. Unless the elevator which now acts more like a rudder is applied to increase the wing's AOA and increases perpendicular lift and vertical lift, the plane will lose altitude. My larger proposed canard area for John's model will, as I see it, do a better job in holding the nose up in a turn with increased up elevator. I doubt that the wing tips are stalling.

Charles
 Feb 17, 2013, 02:18 PM Registered User United States, OH, Bradford Joined Jun 2005 4,891 Posts Charles, with all due respect, that's just plain wrong. Lift is always perpendicular to the surface generating it. When the wing (or canard) is banked, the lift is still perpendicular to the plane of that wing. The lift can be broken down for analysis into a horizontal component and a vertical component. Assuming a steady-state turn (i.e. level flight, or constant rate of ascent or descent) the vertical component of that lift is equal to the plane's weight, and the horizontal component is equal to the centrifugal force (OK, for you purists out there, the centripetal acceleration) of the turn. In a coordinated turn, the vector sum of the two is perpendicular to the wing, and equal and opposite to the lift. The plane "thinks" it is in level flight. If they are not equal and opposite, then the plane is not in a coordinated turn, and the plane is slipping or skidding sideways relative to the plane of the wing, and making enough sideways forces (relative to the wing, not the ground) with the plane's vertical surfaces (fin, struts, fuselage sides, side-lifting airfoils, etc.) so that the vector sum of all the aerodynamic forces on the plane is still equal and opposite to the vector sum of the weight plus centrifugal force. Lift is perpendicular (in the front view) to the plane of the flying surface that generates it, and (in the side view) to the flight path through the air. Stall speed increases as bank angle increases, because only some (and an increasingly smaller portion as the bank angle increases) of the lift is acting to support the weight, and therefore more total lift has to be made in order to keep this decreasing upward component of it still equal to the plane's weight. In the extreme case of a 90 degree bank, the plane's stall speed is theoretically infinite, since the component of the wing's lift that is upwards is exactly zero. In actual practice, what we call "knife-edge flight", we are making sideways lift with the fuselage and other side-lifting surfaces, as well as yawing the plane enough that the thrust of the powerplant has an upward component. In this case the wing is not supporting the plane's weight at all (nor for that matter, the canard). However, whatever lift is being made by the wing and/or canard is still perpendicular to them in the front and rear view. Last edited by Don Stackhouse; Feb 17, 2013 at 05:12 PM.
 Feb 17, 2013, 08:48 PM Registered User Joined Jun 2005 2,610 Posts Great discussion and I agree but still cannot find error in my view. Please look here. http://tommywiklind.com/aviation/TurningFlight.html Note that in a banking turn the vertical component of lift vector is equal to gravity which maintains constant altitude. I agree that the perpendicular lift is the principle component but it is greater because of angular acceleration, I believe, which keeps the vertical component equal to gravity. As you said about the knife edge, the vertical component of thrust helps to hold the nose up although the forward thust in the main component. I am pleased with all of the help from you and others over the last seven years without which my models could not have been changed or improved and my interest in the hobby would have waned. Charles
Feb 17, 2013, 10:07 PM
Registered User
Joined Jun 2005
4,891 Posts
Charls, I re-read your earlier post, and yes, I missed that your initial statement was referring to level flight. Yes, you are correct, to a point.

Quote:
 ...My larger proposed canard area for John's model will, as I see it, do a better job in holding the nose up in a turn with increased up elevator. I doubt that the wing tips are stalling....
does not follow from that, and is essentially unsupported. You do not provide any viable explanation why you think it is not tip stalling, even though the evidence given by John most certainly indicates that.

Furthermore, adding more area to the canard will require moving the C/G forward, which will unload the wing some more, and lower the overall wing loading. This will tighten the turning radius for a given overall lift coefficient, making the discrepancy in airspeed between the inboard and outboard tips even greater, which will make tip stalling even more likely.
Feb 17, 2013, 11:39 PM
Registered User
Sydney, Australia
Joined Mar 2006
1,316 Posts
Quote:
 Originally Posted by Don Stackhouse John, as far as yaw stability, there are a couple factors against you on this model. The struts for your parasol-mounted canard have some fin effect of their own, as do the turned-up tips on your canard. The fat, square forward fuselage also hurts. However, your extended aft fuselage and fairly large fin should offset this. Without running any exact numbers, my gut feel says your yaw stability should not be great, but is probably adequate. The lack of dihedral might set you up for a spiral instability situation, but not a strong one, just a gentle tendency to overbank. Your observation of a sudden wing drop indicates a different problem, tip stall.
Your conclusion about tip stall makes sense and I believe it is most likely the correct one.

My next canard model is probably going to be a light weight 110" soarer for hi-start launch and also for slope on calm days. It will have crow brakes, 940 sq" main wing area and 60oz flying weight. The canard surface will be built without winglets as I dont think they have a benefit with larger dimesions and higher Reynolds numbers. Instead of parasol (V mount), the horizontal stabiliser will be mounted on the top of a thin pylon, and have an aluminium strut on just one side. Its kind of like a front mounted T-tail .. The pylon will act as a fin, but I don't think its an issue if I add adequate rear fin to compensate. The 60" model appearing on my blog has a Vertical tail co-efficient of 0.0644 based on the tail moment relative to the half-span. For the model as built the yaw stability is adequate as you guessed. I have seen a very minimal tail wag on rare occaisions, but overall I am happy with the yaw stability and not seeking any improvement there. If the new model has any yaw stability issue I will probably fit for a rate gyro for active damping.