Originally Posted by Don Stackhouse

Yes, in that case they are distributing a lot of the plane's mass, in particular the highly variable payload mass, along the length of the plane. That calls for a longer moment arm for the "tail" surfaces. However, other than some sort of electronic active yaw stabilization, the same approach is not viable for the yaw stability problem, so the plane would need larger vertical fins to compensate.

When I posted the C-wing info, I wasn't sure if there would be much efficiency to be gained by the addition of the canard to the C-wing. Perhaps we can assume the canard surface has more to do with practical benefits than efficiency gains. Dr Kroo has authored a number of older papers investigating Canard layout at subsonic speeds. I have been long intending to track down some of the papers, although I don't think they are available online for free.

You wrote above that active stabilisation is not viable for yaw control. I previously advised a friend building a large canard sailplane that maybe a rate gyro could be used on the rudder to improve yaw stability. Do you see the same issues affecting viability of this on a model aircraft?

Originally Posted by canard addict

The D. Duck 2 will use the same canard wing and being only 11% of the main, may hold down the climb rate at full throttle. The big delta wing seems to be highly sensitive to CG changes. A Mark 3 version may have a flat bottom canard at zero incidence relative to the bottom and a vertical tail assembly with it's leading edge at ailerons TE. What could go wrong with those changes on a proven good flier?

My big question is: will the new wing experience a tip stall in high speed turns?

Charles, Can I ask why you are concerned about tip stall in the turns? Is it just because of the taper of the wing? I thought there may be a potential issue with the prop wash over the canard preventing it from stalling at high Alpha flight, but wasn't that the point of this design?

Apart from the prop wash issue, I wouldn't be too concerned about a main wing stall. The aspect ratio of the main wing is only around 3.0 which is slightly lower than the canard AR. I did a simulation of the 36in Delta duck in XFLR5 and it dosen't predict a main wing stall until Alpha = 15.5 degrees. The previous Delta Duck 30" wing was predicted to stall at Alpha = 20.0 degrees. My simulation with a high-lift canard airfoil with a lot of down canard flap stalls at 13.5 degrees. However the prop wash over the wing will definitely change that.

I didn't realise you would build a new fuselage for the delta duck II. I suppose you made the fuselage a bit wider to allow the battery to be placed further back. Are you making any other changes to the design? How about the landing gear, will it be the same as the first version?

Originally Posted by John235

...You wrote above that active stabilisation is not viable for yaw control.

That is NOT what I said. I said that "OTHER THAN some sort of active yaw stabilization..." You can hang a canard on the front, adjust the C/G forward to compensate for the new Neutral Point for the complete aircraft, and end up with acceptable static and dynamic pitch stability. The same is not true for hanging a dumb, fixed, un-controlled vertical surface on the nose. Any non-actively-controlled side area added to the nose means additional area is needed in the fins to maintain static yaw stability.

Active yaw stabilization through a vertical fin/rudder on the nose will work, although the failure modes get very tricky. Back in the 50's the need for a yaw damper was enough of an issue to be one of the main excuses for killing Northrop's YB-49 flying wing project. However, many airliner-class airplanes today have yaw dampers.

The problem in this case is a matter of degree of criticality. If the yaw damper on a typical modern airliner goes out, the main result is that the passengers start filling up most of the air-sick bags, but the plane would probably still be flyable to a safe landing. If they used a nose-mounted fin and rudder on that C-wing airliner in the sketch (as a means to not need such large fins on the wing tips), and the active control system on it went in-op, in all probability the plane would immediately swap ends about the yaw axis, followed by a shower of shredded control surface parts, followed shortly after that by a shower of shredded and mangled major primary structure parts. The yaw damper becomes a "Catastrophic" class safety-of-flight-critical item.

I previously advised a friend building a large canard sailplane that maybe a rate gyro could be used on the rudder to improve yaw stability. Do you see the same issues affecting viability of this on a model aircraft?

It's a judgement call. If any one item fails in the list including the gyro, the servo, the connection to the system from the receiver, the power supply to the servo and gyro, the mounting for the servo, the mounting for the gyro, the hinging for the rudder, the rudder itself, and the control linkage from the servo to the rudder, will that result in loss of the aircraft?

Note, in that case we're talking about a list of roughly a dozen components, the failure of any single one of which could result in the plane and its owner having a very bad day, and a potential sighting of the endangered species "aircraftus precipitous". The odds of that happening are relatively large.

So, is that acceptable? If not, is there a way to make any single one, or reasonable (i.e.: likely to occurr in the same flight) combination, of all of those failures survivable? In other words, with the yaw damping system out of comission, is the plane still flyable to a safe landing?

If the answer to either of those questions is a solid, unequivocal "Yes", then yes, it's viable.

As far as whether adding a canard to the C-wing in the sketch makes it more efficient, I strongly doubt it. However, properly done, it could make the plane more versatile by providing a wider allowable C/G range, and therefore the ability to carry a wider variety of payloads and payload arrangements.

John, Thanks, I am flattered that you are checking out my Duck.

John 235

Charles, Can I ask why you are concerned about tip stall in the turns? Is it just because of the taper of the wing? I thought there may be a potential issue with the prop wash over the canard preventing it from stalling at high Alpha flight, but wasn't that the point of this design?

Back in the summer when the Duck was under test, it was put into a high speed full throttle turn with full UP on both elevons and wings nearly vertical. It was flying into it's own turbulence and after about three revolutions, the highest wing tip caused what I call a whip stall making the model leave the circle. I was hoping that the new smaller tip chord would have less tendency to do the same. I have put all of my planes, conventional and canards through the same test and have had only one other, a conventional do the same. High speed tight turns are generally smooth in all of my models except for the D Squared which had the rear wing do a high speed stall in a power dive with sudden full UP elevons.

John 235

I didn't realise you would build a new fuselage for the delta duck II. I suppose you made the fuselage a bit wider to allow the battery to be placed further back. Are you making any other changes to the design? How about the landing gear, will it be the same as the first version?

The fuselage is unchanged. The smaller 2200 battery easily fits in and will need to be further forward because of the larger tail. The wing and tail are the only changes. The 3/32"wire LG will remain the same. The old one took a beating when a forward CG was tried to hold down the climb rate. That did not work but the 2 degree down thrust did. Moving the battery back about 1.5 inches turned it back into a floater at low speeds and fixed the LG problem. It still drops like a brick on landing which makes things predictable and with a little rotation the LG is preserved. I appreciate your keen observations.
Charles

http://www.eaa.org/news/2010/2010-11-04_rutan.asp

Originally Posted by Don Stackhouse

That is NOT what I said. I said that "OTHER THAN some sort of active yaw stabilization..." You can hang a canard on the front, adjust the C/G forward to compensate for the new Neutral Point for the complete aircraft, and end up with acceptable static and dynamic pitch stability. The same is not true for hanging a dumb, fixed, un-controlled vertical surface on the nose. Any non-actively-controlled side area added to the nose means additional area is needed in the fins to maintain static yaw stability.

Don, Thanks for your comments. I think I now understand you were referring to the lateral area of the fuselage shown in the concept for the C-Wing canard airliner. What I don't understand is whether you are actually advocating using a forward mounted canard vertical stabiliser with active yaw damping. If an active yaw damper is required, wouldn't it work just as well when applied to the vertical surfaces on the winglets?

Originally Posted by Don Stackhouse

So, is that acceptable? If not, is there a way to make any single one, or reasonable (i.e.: likely to occurr in the same flight) combination, of all of those failures survivable? In other words, with the yaw damping system out of comission, is the plane still flyable to a safe landing?

If the answer to either of those questions is a solid, unequivocal "Yes", then yes, it's viable.

I think that a good quality rate gyro is fairly reliable. I never had a problem when flying model helicopters. To me it seems the risk is quite low compared to what already exists with the battery, connectors, radio reception and servos etc. There's a good possiblity of being able to land the model safely without an active yaw stabiliser. In fact I have seen many people successfully flying canard models with poor yaw stability. With the rudder servo forced into full deflection, maybe it is possible, or maybe not.. At worst, there is probably some chance to choose where it crashes.

Originally Posted by John235

... I think I now understand you were referring to the lateral area of the fuselage shown in the concept for the C-Wing canard airliner. What I don't understand is whether you are actually advocating using a forward mounted canard vertical stabiliser with active yaw damping. If an active yaw damper is required, wouldn't it work just as well when applied to the vertical surfaces on the winglets?

Well, we're almost on the same page, but not quite.

The original comment was that adding a canard to the nose allowed for a wider C/G range, which is true, at least as far as pitch stability is concerned. However, that does not fix the problem of yaw stability variations due to that same wider C/G range. What I was pointing out is that the same fix that worked for pitch does not work for yaw. If you stick a "vertical canard" on the front of the airplane, it will reduce static yaw stability, not help it (although it will help the dynamic yaw stability), unless the "vertical canard" is actively controlled to make it artificially provide static stability.

Of course, if you provide stability artificially, you then have the problem of what happens WHEN (not if) the artificial system fails.

I think that a good quality rate gyro is fairly reliable.

For model purposes I'd tend to agree. However, the key here is "fairly". That's a relative term.

If it's a sport model that will probably be crashed many times in its short life due to mainly pilot error, and if the costs of that are just time and materials that someone would spend anyway because they enjoy building and experimenting, then the failure rate of the gyro would probably be an insignificant contributor to the overall failure rate. Properly shock-mounted, and if it's something like a piezo gyro that might be expected to survive a crash better than the old mechanical gyros, the gyro would also probably not add significantly to repair costs.

However, if the failure of the gyro would significantly increase the likelihood of a subsequent crash, and the model was very expensive and/or had the potential to create a lot of mayhem (such as a high-performance turbine model), and the model was essentially unflyable if the gyro failed, then some sort of redundant systems or backup plan might be wise.

If this was an "airplane with seats", or a UAS that was expected to be able to fly safely over populated areas, then the stakes get exponentially higher. What was "fairly reliable" for a sport model might be "not nearly good enough!" for this more critical application.

I never had a problem when flying model helicopters. To me it seems the risk is quite low compared to what already exists with the battery, connectors, radio reception and servos etc. There's a good possibility of being able to land the model safely without an active yaw stabilizer....

That last sentence is the game-changer. It changes the acceptable failure rate requirements by changing the hazard class.

For example, in full-scale aircraft FMECA's (Failure Modes, Effects and Criticality Analysis), we generally recognize four hazard classes: Minor (nuisance items), Major (significant loss of performance, some possible injuries), Hazardous (serious emergency, some limited aircraft damage, limited number of injuries or deaths), and Catastrophic (unrecoverable, loss of aircraft, multiple deaths).

The acceptable failure rates for those classes vary from whatever is commercially acceptable (maintenance costs, unscheduled cancellations of flight, etc.) for Minor class, once per 100,000 flight hours for Major, once per 10,000,000 flight hours for Hazardous, and once per 1,000,000,000 flight hours for Catastrophic category failures. The idea is that the entire fleet of a given aircraft type will never accumulate even close to a billion flight hours, so the "Catastrophic" failures will probably never happen.

Of course that ignores the fundamental failing of statistical analysis. If the probability of something happening is once per billion flight hours, that just means that somewhere during each billion flight hours we can expect one of those failures to occur. However, that does not say anything about WHERE in that billion hours the failure will occur. The probability is the same for it to happen in the billionth, the thousandth, the five-hundredth, or even the first flight hour. It's like having a deck of a billion cards, and just one of those cards is the "Ace of Catastrophic Failure". If the deck is shuffled, we have no idea where that card is located in the deck. It might be right on top.

Life, contrary to what the lawyers on the TV would like you to believe, is an uncertain thing. We can make a safe outcome more likely, but there is no way within the entire capabilities of our species to make anything ABSOLUTELY safe.

However, getting back to the original question, if the plane can be flown to a safe landing after the gyro fails, then that puts the failure in the "Minor" or at most the "Major" class, instead of being potentially catastrophic. That means the acceptable failure rate can be four orders of magnitude (ten thousand times!) greater than for a catastrophic failure (where the failure makes the plane literally unflyable), and still be acceptable.

Here's my progress with flying the Polar Duck. Still lots to learn. I can't get a very high alpha though. The rudder steering is improving. what do you think?

PolarDuck_0001.wmv (5 min 6 sec)

cheers

Nick
PS
Sorry about the irritating dog noise, You can hear sheep too, he doesn't bother them at all, they're used to him.

And thanks to George and Alice for the camera work.

Great video, Nick!! You have there a land and water plane which will do it all.
I don't know how you manage all that wind but when you need power, it really cooks! My Duck seldom sees over 5 mph wind. Finally got to the field yesterday at temperature of 52F but today it is now at 27 and heading for the teens. Thanks for the personal touch, it's good to know you a little better.

PS, Does RC Groups still have spelling check or did I just lose it on switching to Windows 7?
Charles

Nick, Thanks for the lovely video. You are flying it very very well and keeping plenty of power in the turns. The landing looked a treat. The model looks ready for water takeoff and landing, but I guess its far too cold for that.

If it's cold enough ( << 0 deg. C, 32 deg. F), doesn't that make it a lot easier to get the model up "on plane"? ;-)

Originally Posted by Don Stackhouse

The original comment was that adding a canard to the nose allowed for a wider C/G range, which is true, at least as far as pitch stability is concerned. However, that does not fix the problem of yaw stability variations due to that same wider C/G range. What I was pointing out is that the same fix that worked for pitch does not work for yaw. If you stick a "vertical canard" on the front of the airplane, it will reduce static yaw stability, not help it (although it will help the dynamic yaw stability), unless the "vertical canard" is actively controlled to make it artificially provide static stability.

That make sense. I hadn't really thought in terms of optimising yaw stability unless it was for a thermal sailplane. The simplist approach is to design the aircraft with excess yaw stability, so it is adequate for all CG locations. I understand it will cause some excess drag, but I think the penalty is less than having excess pitch stability. For an airliner I realise that every bit of drag has to be paid for with fuel consumption.

Hearing that dog barking would be a huge distraction for me. Doesn't it ruin your concentration?

Now I've got the plane, perhaps I should try and describe the mission statement:

Small enough to fit in my car assembled and to use 25 amp, 2200 MaH, 9-12g servos etc.
Tough enough for rough terrain, water is a bonus.
Short take-off and landing - (I might enlarge the size of that rear elevator)
Maneuverable at low speeds - (I might increase the travel in the rudder, and move the CG back)
Fast but not screaming - (I might buy a 40 amp ESC and 7x5 prop, the motor can handle it)
Mean time before catastrophy: 100 hours, hopefully.

The existing prop is APC 5x5 (maybe even 4.7x4.7). for water, I tried the 7x5, but the ESC cut out on me when the hatch was closed. So 40amp ESC could be essential for ROW.

Thanks guys!

Nick

Steve

Hearing that dog barking would be a huge distraction for me. Doesn't it ruin your concentration?

I'm used to it and I don't always take him with me. I frequently ask my neighbours if it bothers them and they always say no. The only thing worse than a yapping dog is the sound of an angry owner shouting at him.

How many watts are you asking that little 5" prop to absorb?