What are the advantages of a canard wing aircraft?

What are the advantages of a canard wing aircraft. I've heard about their supposed stall advantages, but is there any weight advantage without sacrifices to strength(assuming materials are kept the same) or stability? Can the canard be sized the same as the stabilizer on a conventional plane?

Mike, I'll let you in on a secret: There is no such thing as a canard!

Before the mob of torch-carrying townspeople surrounds my castle, maybe I'd better explain.

There are single-surface aircraft (flying wings). There are two surface aircraft (tailed monoplanes, tailless biplanes and "canard" aircraft). There are three surface aircraft (biplanes with tails, the Piaggio Avante, etc.). Obviously we can continue this line of reasoning ad infinitum, but I presume you see the point.

Let's look at the two surface category a little closer. Let's ignore the tailless biplanes for the moment and just look at the canards and the so-called "conventional" monoplanes (which in this case means any monoplane with the tail at the back, whether it's a V, T, or "conventional" tail).

The arguement typically given in favor of canards is that they are more efficient (less induced drag) than an aft-mounted tail, because both surfaces are lifting upward, instead of a wing lifting upward and a tail lifting downward. In actual practice there are some iceberg-sized holes in this Titanic bit of logic.

First of all, the horizontal tail on an aft-tailed aircraft doesn't necessarily lift downwards. It is very possible for a fully stable aft-tailed aircraft to have upward lift on the tail surfaces. All that is required for stability is that the lift on the wing in front change less quickly in response to an angle of attack change than the lift on the tail. If the plane noses down (angle of attack decreases by the same amount on both the wing and the tail), the lift at the more sensitive tail sees a bigger decrease in lift than the wing, allowing the tail to drop and the nose to rise until angle of attack is back to its original setting. The lift changes on both surfaces during this maneuver, but what matters is not how much "positive" or "negative" lift there was on either surface to begin with, but instead how much proportionate change in the lift occured on the two surfaces in comparison to each other.

In this sense the aft-tailed aircraft is nothing more than a canard aircraft where the canard is much bigger than the wing. Or, if you prefer, a canard aircraft is nothing more than a "conventional" aircraft where the tail is bigger than the wing. Like I said at the beginning, there really is no such thing as a canard aircraft; it's merely one particular zone of a continuum that covers the entire range of two-surface aircraft. From the aircraft's point of view (the ONLY point of view that really matters in the end), there is no distinct point where it changes from a "canard" to a "conventional tail".

But what about the cases where the tail does lift downwards? Isn't there a so-called "trim drag" associated with this case of the tail apparently fighting against the wing? Doesn't a canard lifting upwards and "helping" the wing eliminate this drag? Well, not necessarily.

To understand what's happening here we need to talk about "downwash". A lifting surface makes lift by grabbing air and accelerating it in one direction, causing a reactive force in the other direction that we call lift. Yup, it's Newton's third law about action and reaction. In this case, because a surface making "positive" lift is accelerating air downwards, there is a downdraft behind the flying surface that we call "downwash". If you have another flying surface behind the first one, it must fly in this downwash. In the case of a typical canard aircraft where the wing and canard both lift upwards, the middle part of the wing (the part that makes most of the lift!) has to fly through the perpetual downdraft created by the canard. This changes the direction of the local "relative wind" at the wing, tilting the lift vector aft and therefore increasing the drag of the wing. In the case of a downward-lifting aft-mounted stabilizer, the downwash of the wing helps the stabilizer, tilting its lift vector forwards and reducing the drag. In effect, the aft-mounted downward-lifting stabilizer is recovering some energy from the wing's downwash. The net result is that the total lift of the aircraft, as represented by the total downwash in the aircraft's wake, will be determined by the aircraft's weight and "G" loading, regardless of which part you officially designate the "tail". There will be some losses due to inefficient interactions between the various surfaces perpetrated in the name of something called "stability", but in the end, the net efficiency of the aircraft in holding itself up in the air and in keeping itself stable will depend more on how well you design and integrate the various parts of the aircraft rather than whether it follows any particular design philosophy. The induced drag penalties to provide stability in a canard aircraft are generally about the same as the penalties for the same amount of stability with a conventional tail.

Now lets talk about stall characteristics.

The other big arguement put forth in favor of canards is that they can be made stall-proof. This is true. It is also true that they can very easily be made stall-divergent. There was a late-WW II era experimental canard fighter called the Curtiss "Ascender" which quickly became known to its test pilots by a slightly different pronunciation of that name, when they discovered that the wing could stall before the canard! When this happens, the aircraft in the blink of an eye attempts to turn itself into an aft-tailed configuration. Next usually comes a few unprintable words from the test pilot about the species of the chief engineer's mother, sometimes followed by a shower of broken control surface parts and then a parachute with the test pilot under it.

Other canard aircraft have occasionally shown this characteristic, as well as some other strange behavior (such as unrecoverable dives when raindrops turbulated a laminar flow canard, causing a substantial loss of lift). Just as with any other design concept, there are limits, and if you exceed those limits you immediately void all warrantees expressed or implied. The same is true of aft-tailed aircraft. Ice on a horizontal tail, or a stall of the stabilizer during landing flare can under the wrong circumstances thoroughly ruin your whole day! No particular concept has a monopoly on good or bad flying characteristics; it's possible to design a really crummy airplane from almost any conceptual starting point!

In the case of the canard layout, the canard is supposed to stall first. This causes the nose to go down, reducing the angle of attack of both the wing and the canard until the canard un-stalls. The wing never stalls in this arrangement, which means that you never lose roll control. Theoretically, that is. In some canard aircraft with very low wing loadings and long wings (such as sailplanes), it's possible in a tight turn to get a big difference in airspeed between the inside wingtip and the rest of the wing. In this case it is possible to stall the inside wingtip while the center section, outside wingtip and canard are all still flying. The loss of lift (BEHIND the C/G), and the increase in drag on the inside wingtip, results in this case in a pitch-up plus a yaw and roll-off towards the center of the turn, possibly followed by a fairly flat spin. The same scenario with a conventional aft-mounted tail would result in the same roll-off and yaw (the classic "tip stall"), but without the pitch-up that worsens the initial stall.

Assuming we don't get into a tip stall problem, the canard still pays a price for its supposed stall resistance. The primary lifting surface (the part that supports most of the aircraft's weight) in both types of layouts is the wing. If the canard stalls before the wing, that means that the wing never can reach its maximum lift. In order to make up for this artificial limit on the wing's lifting ability, we need to use more wing area. This adds weight and whetted area, which adds drag. On the other hand, in an aft-tailed aircraft we typically use some form of geometric and/or aerodynamic washout (i.e.: wingtips that are twisted nose-down, or have a more stall-resistant airfoil than the root), to insure that the root always stalls first. This means that part of the wing (the outer panels) never gets all the way to stall, so it doesn't use its full lifting abilities either. "There ain't no free lunch." Things like stability and good stall behavior always exact a cost of some sort, no matter which layout you choose.

It is easier in some regards to get good stall behavior from a canard than a downward-lifting stabilizer. When a canard stalls, the resulting nose drop naturally reduces the angle of attack and helps to un-stall the canard. If a downward-lifting stabilizer stalls, the resulting nose drop increases its angle of attack, thereby keeping the stabilizer stalled and causing a much more pronounced nose drop. Of course you're supposed to design an aft-tailed aircraft (regardless of whether the stabilizer is upward or downward-lifting) so that the WING stalls first. If you think about it, regardless of what you call the thing in front or the thing in back, you want to make sure that the thing in front stalls first, and that the tips of the lifting surface with the longest span (your primary roll-control surface) never stall. See, there's no difference in the fundamental design rules of a forward tail vs. an aft tail. Like I said, there is no fundamental difference between the two, and therefore there is really no such thing as a canard.

Rules of thumb on sizing, etc.? Traditionally the same rules of thumb for tail sizing of a conventional tail will also work for a canard. C/G is a bit more complicated, mainly because most folks are familiar with specifying c/g by the WRONG method. Typically we think of C/G in terms of percent of the aerodynamic center of the WING, ignoring the tail. The old rules for power models say the c/g should be ahead of the aerodynamic center of the wing's mean aerodynamic chord ("MAC") if we want the model to be stable. With a little though and observation you could see that there's a major flaw in this concept. Although the aerodynamic center is supposedly about 1/4 of the way aft of the leading edge at the MAC, on sailplanes we often put the c/g at more than 50% aft of the leading edge of the MAC, yet we still have positive pitch stability! How can that be?

What we really should be looking at is the aerodynamic center of the entire aircraft, including the tail. If you look at the total aerodynamic center of the wing and tail combined! On those sailplanes with "ridiculously" aft c/g but still positive stability, you will see that the c/g is still AHEAD of the aerodynamic center of the entire aircraft! This is the key. Take a look at some of the models you've flown that behave to your liking. Calculate the aerodynamic center of the total aircraft, then figure what percentage of the MAC the c/g is ahead of the AIRCRAFT's aerodynamic center. Do the same thing for your canard model, you'll be in the ballpark.

So what size does the "canard" or "stabilizer" really need to be? Well, the limit in one direction is the size of the wing. At the point where the "tail" or "canard" becomes bigger than the other lifting surface, it ceases to be the "tail" anymore and becomes the "wing". This has nothing to do with aerodynamics, it's purely a matter of semantics; the biggest lifting surface is called the "wing". If they're both the same size, it's called a "tandem wing aircraft". The air molecules couldn't care less, all they know is that it's something that pushes on them and they push back.

Is there a limit on how small it can be? Yes and no. We're mostly used to designing airplanes where the odd little quirks of airfoils and wings, such as the effects of airfoil shape, twist distribution, sweep and airfoil pitching moments on the pitch trim and stability of the aircraft can be mostly ignored. If it's a bit different than expected, we just change the c/g, or the incidence of the "tail" (whether it's in front or in back) just a little bit to compensate, then we don't worry about it. We're used to taking the lazy way out, using the tail as a crutch to allow us to take some shortcuts in the wing design. Nothing wrong with that, as long as it doesn't cost us too much in performance, weight, etc.. But what if we didn't want to rely on that "crutch"? Well, as we make the tail (or canard) smaller and smaller, we have to start paying more attention to pitch stability, pitch damping, keeping the forces in equilibrium, etc., but it's still possible to do this if we do our homework properly. Is there a limit? Yes! When the area of the "tail surface" becomes ZERO, we can't make it any smaller! Has it ever been done? Why yes, many times, it's called a "flying wing" or a "tailless aircraft"! (I'll bet you knew that one was coming!) So I guess we can conclude that a flying wing is nothing more than a two-surface aircraft where the area of one of the surfaces is zero. We can also say that a canard or conventional tailed aircraft is nothing more than a three-surface aircraft where the area of one of the surfaces is zero! And a three-surface aircraft is nothing more than..... I presume you see the point. Like I said, the air doesn't care. They're ALL just objects that push on the air and the air pushes back.

So what about tailless biplanes? Well, those are nothing more than a "two-surface aircraft" where the moment arm between the two surfaces is approximately zero. The same rules apply. Since the moment arm is a prominent player in most of the stability equations, we have to use some of the same techniques as with flying wings to get the pitch stability and damping to work out, but otherwise they follow pretty much the same methods. About the only other quirk is that you have to be careful with two surfaces in close proximity to each other, they have a tendency to squabble with each other over who has control of the airflow between them, much like a couple of jealous siblings. Keep them far enough apart (usually about one and a half chord lengths or more) and they get along reasonably well.

Is there any advantage to doing this? As a matter of fact, in some cases, yes there is. If you have a limit on span and an excess of weight to carry, or if low speed performance is paramount (such as with short-field or weight-lifting aircraft, or if you have a situation where thermalling performance is all that matters, and you can totally ignore cruising or penetration), a biplane (with sufficient separation between the two wings) has LESS induced drag (the drag resulting from the production of lift) than a monoplane of the same span. If all you care about is hanging in the air at the minimum possible airspeed, by all means make it a biplane!

This is because a well-designed biplane has better "mass flow" than a monoplane of the same span. The induced drag of an aircraft depends a great deal on how big a chunk of air it grabs hold of to produce its lift. In the case of a monoplane it's a cylinder with a length determined by the speed of the airplane. Go faster, the cylinder of air gets longer, with more volume and mass, and the induced drag goes down.

The diameter of that cylinder is determined by the wingspan of the aircraft. It's NOT the aspect ratio like you might have heard, it's strictly a matter of span and how efficiently the wing uses that span; aspect ratio (how skinny the wing chord is compared to its length) has no direct effect on induced drag, that idea is an old wive's tale. What matters is the volume (its cross-sectional area times the length, times the air density) of the chunk of air (a cylinder in the case of a monoplane), which the wing is acting on to produce its lift. And that cross sectional area which determines that volume is determined by wing span.

The cross-sectional area of the biplane's chunk of air includes the semi-cylinder of air abouve the top wing and the semi-cylinder of air under the bottom wing (if the wings are the same size, we've so far just exactly equalled the cylinder processed by the monoplane's wing), PLUS the chunk of air between the wings. Voila! Extra cross-sectional area, extra mass flow, less induced drag. Of course the interference between them exacts a penalty, and there's also extra parasite drag, etc., but if we do a good job of minimizing those, it will be better than a monoplane at very low speeds. Just don't try to go fast.

Well, let's see, we've discussed aerodynamic advantages (and determined that there aren't any fundamentally inherent in a canard), we've talked about mass flow and total wake effects, now what about your questions about structures?

An externally braced biplane has very obvious structural weight advantages (at the expense of parasite drag), which was one of the main reasons they were so popular in the early days of aviation. To take advantage of that structural efficiency you have to do your homework and make sure that nothing is stronger than it needs to be, but properly done the results can be very impressive. For example, on the "Bucker Jungmeister", the 1935 German biplane that literally wrote the book on modern aerobatics, the aircraft is strong enough to withstand 12 G's (about 6 or 7 TONS of load on the wings!) in normal service (and they didn't have carbon fiber in 1935, either), yet each of its four wing panels weighs less than 25 pounds, covered, painted and ready to mount on the aircraft!

For a canard or an aft-tailed aircraft it's a bit more complicated. If you keep the c/g well forward and the tail aft, you have to somehow connect the tail to the rest of the aircraft. You can use a "joined wing" concept, where the forward surface is swept back, and the aft surface is swept forward till the two surfaces meet at their tips. This is actually another variation of external bracing, and now you have to deal with two surfaces that come close to each other, causing interference problems and reducing the effective moment arm (which hurts stability). You also have to deal with the aerodynamic and aeroelastic effects of large sweep angles, which is another whole can of worms.

The traditional approach is to connect the two surfaces with a tail section on the fuselage, which adds weight and whetted area. This is where a forward-mounted tail (i.e.:"canard") can have some advantages by eliminating some of the weight and whetted area of the fuselage. Of course the greater wing area required (because the canard's wing can't work to its full capabilities because of stability requirements) may completely eat up that weight and whetted area advantage, plus maybe a bit more. You might also have to add some structure and whetted area to get the vertical stabilizing surfaces far enough aft of the c/g to do their job effectively. If you're not careful, you could end up with even more weight and whetted area than the conventional approach.

In the end it all comes down to the initial requirements you specify for the design, and how well you get all the features of the aircraft to work together to meet those requirements. If there was a clear-cut obvious answer that always works, all airplanes would look alike!

Don Stackhouse @ DJ Aerotech