Cannard Airliner?

[3holelover]

Many moons ago, when I first realized the horizotal stab on a big bird is in fact an inverted airfoil, I began wondering why it is that none of the Boeing's or Airbus's etc. had bothered to correct that obvious waste of efficiency?
If they flip the thing right side up and stick it in front of the mainplane, it works for you instead of against. (and can still look after "mach tuck" which, I understand, is why it's inverted when it's at the back)
I've asked people about this and the best answer I've got so far is that "people wouldn't fly on something that looked like that"!
Actually, I once spoke to a Boeing engineer who said that they'd had discussions about cannard's, but decided that, because of an inherent instability, they would stick to tradition. Now doesn't that sound just like Boeing!
Any thoughts?

[Squawk 8888]

The biggest concern with canards is their behaviour during the stall- because the centre of lift is in back it'll drop tail-first http://www.pprune.org/ubb/NonCGI/eek.gif The solution for existing canard A/C (such as the Avanti) has been to give the canard a higher angle of incidence so that it stalls and drops the nose while the wing is still flying. So far so good, but deep stalls are still unrecoverable. Perhaps FBW would be the way to go here.

[Old Aero Guy]

You're pretty hard on Boeing for not using a canard design. Actually, there are some very good reasons canards don't work well as transport airplanes. These are:

1)If a canard configuration is to have the same lifting area (canard and wing) in cruise as a conventional design (wing and h. tail), the canard configuration will need to have the same high lift devices fitted on the canard as the wing. In fact, the canard high lift devices will probably need to be more extensive than the wing. Therefore, the canard will be more expensive to build and maintain. The conventional design needs only wing high lift devices.

If you don't want to have high lift devices on the canard configuration, it will need more combined lifting area than a conventional design to have the same low speed performance. More lifting area means higher skin friction drag for the canard configuration in cruise and lower economic performance.

2)Canards are more difficult to load than a conventional design and have higher induced drag.

Consider that on a conventional design, all payload (passengers, cargo and fuel) is distributed about the wing center of lift. Usually, the C.G. is near the center of lift and tail lift up or down is relatively small. This allows the wing to produce most of the required lift and minimizes induced drag. As the fuel in the wing burns off, the C.G. position and tail load remain relatively constant, so induced drag remains relatively constant and relatively low.

For a canard, all the passengers and cargo are ahead of the center of wing lift and the fuel is near the center of wing lift. At the start of cruise, canard and convential designs may have nearly the same induced drags since in both cases the wing is producing most of the lift. As fuel burns off for the canard configuration, the C.G. must shift forward. This will unload the wing and load the canard to maintain trim. Since the canard span is less than the wing, induced drag will increase and be higher than a conventional design.

Economically, a stable canard cannot compete with a convential airplane due its inherently higher skin friction and induced drag. An unstable canard would be more competitive, but the regulators and I suspect most pilots aren't quite ready for that.

[3holelover]

Interesting point re: the high lift devices... Although, they get pretty simple when they get smaller. I wonder if that couldn't be done?
Your fuel explanation seemed to be discounting the possibility of a trim tank in the canard (only one 'n', got it thanks). With an airbus style C of G fuel management system, wouldn't that solve that part of the equation?
And hang on a minute...
"If a canard configuration is to have the same lifting area (canard and wing) in cruise as a conventional design (wing and h. tail), the canard configuration will need to have the same high lift devices fitted on the canard as the wing. "
Why would you need any high lift devices in cruise?
Assuming a same wing area as a conventinal a/c the "lifting area" would in fact be greater, simply because the stab normally cancels some and the canard will in fact add some...
The stability question puzzles me. But I do know there are a lot of birds with computers handling some pretty tricky aspects of flying.
I haven't heard of any deep stall troubles with a canard, I thought because of a higher angle of incidence on the canard it was impossible to stall the mainplane. And if you dropped one in the air at 0 kts wouldn't it drop it's nose anyway? I've never seen a lawn dart fall tail first.
I'm hard on Boeing for their stubborn insistance on continuing to do things as they always have. But I guess I must admit, the last few new ones out of Seattle have shown signs of trying a few things the other guys do.

[Old Aero Guy]

3holelover

In response to your questions:

Why would I need any high lift devices in cruise?

You don't. In cruise, the wing provides plenty of lift. In fact, wing area is often larger than required for optimum lift to drag ratios because it may be sized to perform other tasks such as to carry enough fuel for the mission, provide for airplane growth etc.
Takeoff and landing is another story though. There you need the "clever bits" i.e. the high lift devices. The success of jet transports is due in no small part to being able to minimize wing area for cruise while using efficient high devices to have acceptable takeoff and landing speeds.

This is where a canard has a problem. Look at your comment to Squawk 8888. The canard will stall before the wing on a canard configuration, because the wing is operating in the downwash field of the canard. At stall, the canard angle of attack higher is than the wing because of the canard downwash. You might say the downwash from the canard "stall proofs" the wing. One of the best things about a canard configuration is their stall are usually very gentle.

On a conventional configuration, the wing will stall before the h. tail for the same downwash reason.

When the forward surface stalls on both configurations, a similar aft surface effect is very necessary. When the forward surface stalls, downwash decreases and the aft surface develops positive lift. This pushs the airplane to an attitude necessary for stall recovery, less canard or less wing angle of attack.

Now let's consider how leading edge high lift devices work. Their primary effect is to allow the wing to reach a higher angle of attack before the wing stalls. On a conventional configuration, this works fine. The h.tail drives the wing to a higher angle of attack with the leading edge devices deployed than it would be capable of with no leading edge devices. The tail doesn't stall because its in the wing downwash field and its negative lift is relatively small as its moment arm to the C.G. is relatively large. The wing angle of attack gain increases overall lift and allows the airplane to operate at lower speeds.

If you put leading edge devices only on the wing of a canard, its a wasted effort. Stall is still controlled by the canard. When the canard stalls, the airplane stalls and the wing will not have developed its full lifting potential. If you put leading edge devices only on the canard, then the canard may not have stalled when the outboard wing panels, which aren't protected by the canard downwash, stall. This a very dangerous configuration. It will pitch up when the wing stalls because the canard hasn't reached its stall angle of attack and with the outboard wing stalled, there may be no roll control. To make leading edge dervices work on a canard configuration, they need to be on both wing and canard which lead to my comment in the earlier message about increased cost and maintenance. If you chose to not use leading edge devices on a canard configuration, the lift provided by the canard isn't a good trade when compared to the ability of a conventional configuration with leading edge devices to make better use of the primary lifting surface, the wing. Therefore a no leading edge device canard configuration will have more lifting surface area (and skin friction drag) than a conventinal configuration with a good high lift system assuming both have the same takeoff and landing speeds. (Note that I've left trailing edge flaps out of this discussion to keep things simple. My only comment is that if you put them on the wing, you also need to put them on the canard.)

Why not a trim tank in the canard to reduce induced drag as wing fuel burns off during cruise and the C.G. shifts forward?

For cruise, a trim tank doesn't help. The problem is that as you burn fuel from the wing the C.G. moves forward, increasing the canard loading and increasing induced drag. Having a trim tank forward only makes the problem worse, as now the C.G. is forward to begin with. Likewise, trapping some fuel at the wing tips can't delay the forward shift in C.G. as the wing fuel is burned. If you take fuel out of the wing and distribute it in the body, you lose on two counts. The body fuel reduces cargo capacity and the wing has to get heavier because it has no bending moment relief from acting as a fuel tank.

In summation, the small amount of lift which seems to be gained from having a canard makes you pay a very high price for eliminating a small tail download which often is not there in cruise anyhow. Most cruising flight for a conventional configuration has a C.G. position at or aft of the wing center of lift so the tail either has no load or even a slight upload.

Popular Mechanics is not the best place to learn about airplane design.

[3holelover]

Old Aero Guy:

I'll ignore your last sentence since I can't be sure you intended it as it sounded.

I like your description of the need for high lift devices on both canard and wing. You make a good argument and I'll come back to that in a bit... You're losing something in the fuel vs C of G and center of lift argument though...

As in most beasts of this kind, the bulk of the fuel would be in the wing center section and inboard sections of the wing. In flight the c of g needs to be kept where the center of lift is. The wing will obviously be placed such that, that will work for the canard design. Without the books at hand, I'd guess that to be somewhere pretty close to the main wing root, depending on things like sweep angle and ratio of wing area to canard area. I can't see how it wouldn't be possible to manage the fuel burnt from a trim tank in the canard, the center wing section, the wing root tanks and outer wing tanks so as to maintain the c of g. In fact, doesn't the Concorde do just that? Except that because of it's very thin and unusual wing it has umpteen bloody tanks. And I think fore and aft trim tanks!
So I'm not buying that part of your argument.

Back to the leading edges and trailing edge flaps... One of the smallest wings I've ever seen had one of the simplest designs for leading edge devices, and I think(?) that was an F5. (a small fighter in the colours of the US Thunderbirds) It had a simple spring loaded slat the length of the wing. In flight, when the angle of attack was increased to a point where the aerodynamic forces would allow the spring pressure to overcome the air pressure on the nose of the slat, it'd pop out and then aerodynamic forces kept it out until angle of attack decreased sufficiently again... I mention this only to illustrate the potential for simplicity. Such a device couldn't be a high maintenance item I wouldn't think. As for flaps, yep, I guess if you've got to have them, then you've got to maintain them. But there are some reasonably simple flap designs as well. Look at the DC10 for instance.

How did Rutan deal with these puzzles? How well does the Beech Starship perform in terms of efficiency I wonder?

I think the only airplane design I ever saw in a popular mechanics mag was an "aerocar"... it was pretty funny.

[This message has been edited by 3holelover (edited 10 February 2001).]

[Genghis the Engineer]

A couple of thoughts about Canards: -

In cruise, no high lift devices are used, therefore there are two lifting surfaces, this is more efficient than a wing+tailplane design. If anybody doubts this, I have performance figures for my own single seat canard that knock spots off anything with a tailplane and similar weight and engine.

When using high lift devices for take-off and landing, unless you are prepared for huge pitch changes (that probably need to be automated out) as the devices are selected, it's going to be necessary to fit devices to both wing and canard - a complexity not required in the same way for a tailplane aircraft.

With a tailplane, the stall is (usually*) marked by wing stall - or in other words when there's hardly any lift left available. With a canard, the stall is (usually*) marked by canard stall - so at that point the wing could still potentially generate more lift.

This means that a canard aircraft will have a higher stalling speed, and thus a higher approach speed and thus need more runway.

So, a canard configuration defeats the airline manufacturer's desire to be able to use the shortest possible runway. This, in my opinion is what kills the idea - all the other problems could be solved if there was a good enough reason to.

G

* My "usually" is because in fact a great many aircraft have the stall speed defined by a pitch control limitation, and the wing may well still be lifting.

[Prof2MDA]

A few points to add to the excellent points already mentioned:

1. Modern transports are now have almost neutral static stability, so there is little "downforce" provided from the horizontal stabilizer as it is. Even on convential airliners there is less downforce than you might expect, with the horizontal stab countering the pitching moment on the wing but the CG does not necessarily have to be ahead of the CP.

2. In high speed aircraft the requirement of thin airfoils without high lift augmentation creates the possibility of leading edge stall. The high pitch rate that results from such a stall combined with the canard being well ahead of the CG means that the canard AoA will continue to be high a lot longer due to the pitch rate than would otherwise be the case. In fact, the pitch rate can lead to the AoA actually increasing as the nose descends. It is possible that the nose could then even "tuck under", leading to a roll about the lateral (y) axis. Passengers haven't learned to appreciate that sort of maneuver for some reason.

3. Canards are very useful in reducing pitch transients, so you can get faster reaction to pitch changes. Great for a fighter, not all that useful for transport aircraft.

4. As others have stated, the airflow behind a canard is downwash. However, because the canard will have less span than the wing, the effect is a lower AoA on the root section of the wing and a higher AoA towards the tips. This creates two problems, in that it increases induced drag due to the spanwise distribution of lift coefficient (why beyond the scope of this discussion), and also the obvious adverse effect of increasing the likelyhood of a tip stall on the wings -- not a good thing.

This can be offset with forward sweep, but that also results in new problems in structure and a lot "harder" ride in turbulence, among other things.

In balance, don't hold your breath!

[Old Aero Guy]

3holelover:

Let's make sure we are communicating about some basic canard config. vs conventional config. issues. Assume both airplanes have the same static margin with a typical cruise C.G. What approximate percentage of total airplane lift is carried as a trim upload on the canard or a trim download on the h.tail? No need to be too precise, the nearest 5% will do. To make the job easier, assume the conventional config. is an A330 or 777 and then use a roughly comparable canard config. Its a pretty eye openning exercise.

[Grey Area]

Ref High Lift Devices etc. Not totally true, you could always fit a simple sweep system to move the canard CP instead of high lift devices. Then again beechcraft already did that..... Anyone remember the Starship?

[Old Aero Guy]

Grey Area

Remember my prior notes on high lift devices were about leading edge devices, not trailing edge. The Starship variable sweep canard was useful in trimming the extension of wing trailing edge flaps It wouldn't be of much use in compensating for the lack of wing or canard leading edge devices on a .8M cruise canard airliner.

As to the application of a variable sweep canard on an airliner, my fear is that when I go on to my just reward, Satan will assign me the integration of variable sweep canard, the flight deck, an electronics bay and the first class cabin. Every time a solution is reached, he'll remind me it needs to fold to make sure it isn't hit by the jetway at the gate.

[Hay Ewe]

Ok, this is a pretty old thread tat I found wilst doing some lookingup ont eh Beechcraft starship, but one thing you all seem to have over looked is the fact that the torque box for a Hor-Stab
is quite substantial and to have this near the flight deck is quite impractical and a regualr aircraft.

That is no doubt why they are all placed at the aft end of the fuselage. Of course there are exceptions, avanti, starship and The Wright Brothers Flyer!!! Tail first config.

Hay Ewe

[BOING]

Did not read the details in the other posts but the reason you will not sell a canard airliner is that you cannot park it at a jetway!