Three-Lifting Surface 3

[Sparweb]

A lot of discussion ensues when somebody brings up the topic of "canards", which invariably involves lousy stall characteristics, Mr. Rutan's designs, and landing speeds about 20 knots below Vne.
The solution seems to exist in one very unique Italian aircraft called the Piaggio Avanti which has 3-count-em-three lifting surfaces. The main wing has flaps. The forward wing has flaps, too. The tail wing has some kind of elevator-stabilator mechanism that I don't quite understand that controls pitch with and without flaps.
As far as I can tell, this is the only 3LS aircraft certified by the FAA.
Are there any others? Whenever I search the internet, all I get are wannabe designers' pipe dreams and scaled-down knock-off's.
Apart from its aerodynamic complexity, what hidden drawbacks might be in the design?

 

[JohnFortier]

Miper,

First, thanks forthe extra information, especially regarding the possibility of a fuel tank in the aft baggage compartment. I guess I'll just have to get familiar with Keflavik, not that there's much there to get familiar with.

"Detailed analyses of canard configurations by Ilan Kroo and others show that, from a minimum interference drag standpoint, the best vertical location for a canard is well above the main wing. This is similar to the result from biplane theory, that shows positive stagger to be better than negative stagger. Canard test results confirm this conclusion, even if it isn’t intuitive. Like you, I had thought a low canard was better before I saw these results. In any case, few configuration lend themselves to a high canard, and the proposed Boeing Sonic Cruiser is one of the few I’ve seen."

As I recall from my reading, the reason for positive stagger in biplanes was to preventthe high pressure region below the upper wing from interfering with the low pressure region above the lower wing, rather than anything to do with wake interference. Surely, if the foreplane on the Avanti, which is separated from the mainplane more than sufficiently for pressure region interference to be a non event, is producing lift, then the wash will be deflected downward, passing under the mainplane, this wash position being first determined by the low position of the cannard, which means that a downward tending airflow starts out in a low position. I can understand the analysis results if the cannard is close to the mainplane, and pressure region interference is a possibility, but with the separation on the Avanti, I just don't see how a high cannard would be superior to the position chosen.

If analyses and tests indicate otherwise, all I can say is "I'm gobsmacked!" I also find it hard to envision how the foreplane wake can interfere with the mainplane on the Avanti. Still, I've just managed to make a circuit, which was perfect on paper and in simulation, stop oscillating when realized as a prototype. Nothing to do with the design, but the printed circuit board added some "phantom" components which managed to add TWO positive feedback loops to the circuit, so I shouldn't be surprised if, in aerodynamics, a subject about which I know far less than I know of electronics, "Things are not always what they seem, skimmed milk masquerades as cream!" (G&S).

One aircraft which has a cannard above the mainwing is the EF2000 Typhoon, which uses a cannard/delta configuration for extreme maneuverability. But comparing the configuration of a supersonic, high maneuverability fighter with a nine place executive turbo-prop is probably an exercise in frustration. I do notice, however, that the cannard on the Typhoon is far closer to the mainplane than it is onthe Avanti. Perhaps there is a preffered vertical position for the foreplane for every separation, for which a simple formula/graph could be derived. Just a thought.

John



John

 

[Sparweb]

An acquaintance just pointed me to a website,

http://www.aceair.ch

showing a flying prototype 3-wing aircraft. I realized that I had seen this thing on the web a few years ago when it was just a mock-up and moulds doing some airshow tours. Looks like the paper airplane has become a reality in this case. It literally looks like a mini-Avanti!

STF

 

[JohnFortier]

It does look like a mini Avanti, down to the anhedral on the cannard amd the swept back vertical stabilizer and stabilator. I also notice that it uses a rotary, presumably wankel, engine, which will make it almost as vibration free as a turbo. Performance is pretty fair for just over 100 HP.

I would say that Mr. rutan has some very viable competition.

Anybody noticing a common thread about these aircraft - that thread being Europe?

John

 

[Miper]

John,

Your aerodynamic insights are quite good, actually; much better than my electrical insights, of which there are few to none.

I looked around for something to clarify the placement issue for biplanes and canards, but haven’t found the right box of reports yet. (Remodeling a house can make many things disappear under stacks of boxes.) I will admit to being a little sloppy in saying that the canard and biplane interference characteristics are similar. I meant that the result was similar – it’s best to put the forward wing up high (or put the high wing forward, depending on the point of view).

You have given the usual explanation for the advantage of positive stagger, an explanation based on near-field pressures. It’s interesting that Munk’s biplane theory doesn’t show any difference for positive vs. negative stagger; but that’s because it’s based on Prandtl’s lifting line theory. Munk’s result, therefore, wouldn’t have anything to say about the interference of local airfoil pressures. Anyway, canard interference effects would be more of a far-field thing.

It is important to keep in mind that, in subsonic flight, the effects of a wing are felt at quite a distance from the wing, even forward of it. So far-field effects can be extensive. (Wing vortex and source-sink methods use mathematical forms for the aerodynamic force fields that are identical to the form for the field of a charged particle.) Interestingly, what makes supersonic flow different is that the air ahead of the shock wave is unaware of the airplane coming because pressure effects propagate at the speed of sound, which is slower than a supersonic airplane.

A typical value for main wing downwash angle for a business turboprop in cruise is about 1.5 –3 degrees, depending on cruise speed, altitude, and wing loading. At 20 ft behind a wing, then, the center of the wing vortex sheet wake would be only about 6 – 13 in below the wing. A highly loaded canard might have roughly twice as much downwash angle.

The interference between a canard and the main wing is a far-field effect, meaning that we should consider the vortex wake patterns. Tip vortices tend to be on and above the plane of the wing as they are shed off the tips. Something like this.

Tip Tip
Vortex Vortex
_ _ _ _
/ \ Vortex Sheet / \___/______________________\___/

What we see is that for a low canard, while the center of the canard vortex sheet (trailing edge wake) does not impinge on the main wing in cruise, the canard tip vortices almost always do. So it is entirely possible that the reason that it is better to put the canard high is that the tip vortices of a high canard are less likely to impinge on the wing.

 

[JohnFortier]

Miper,

Thanks for the explanation. I can see how tip vortices could impinge on the mainplane, especially at higher angles of attack. It would appear that this is one explanation for the anhedral on the foreplane, since, as I understand it, this will reduce tip vortices.

If tip vortices are a problem, however, I would have expected Piaggio to have attached some cool looking, swept back tip fences to the foreplane! Maybe I'm totally wrong here, but I was underthe impressionthat these significantly reduce tip vortices (Not the cool sweep back, just the fences)

A rather wierd thought, but is it possible that Piaggio are actually using the foreplane tip vortices to increase lift on the mainplane. As I understand it, the tip vortex spirals off the wing tim and spreads as it is left behind. If the vortex is actually striking the lower surface of the maniplane, increasing local pressure, then this might contribute to lift. Such an effect would depend on airspeed but, at cruise speeds, might be reliable enough to be useful.

 

[Miper]

The strength of a tip vortex depends on a wing's total lift and its spanwise lift distribution. I don't know how much anhedral affects it, but it wouldn't be a lot (2% ?). Tip plates or winglets effectively add span, slightly reducing (by ~4-5%)the vortex at the tip, itself.

I've mentioned before that the Avanti canard tips line up with the nacelles. This may be merely an artifact of the original layout, when it may have been judged that vortex impingement on the nacelle would be better than on the wing. I doubt that canard anhedral was in the early layouts.

Wakes on an airplane with high wing loadings are formidable. Sometimes quite a bit of work is required to eliminate wakes that hit the tail as the airplane approaches stall. Pilots call this buffeting; but in flight test, until we get it controlled, the wake effects might be better called "bone jarring rattling and shaking". This effect is exacerbated by wake that enters an aft propeller disk. The turbulence is then multiplied by the horsepower. Pilots report nothing pleasant about the experience, especially in go-around tests. Nor is it safe. Buffet loads are among the worst that the tail of an airplane experiences; and prolonged buffet loads on an aft turboprop installation would do severe damage. Even if you could make it structurally safe, customers won't accept it. The airplane shakes so much that everything in it is a blur -- not the kind of thing you want to do each time you flare for landing.

So I suspect that the canard anhedral may very well be there to lower the vertical position of its tip vortex. There are other possible reasons, of course; but you would have to talk to their wind tunnel and flight test people to get more details.

 

[chasbo]

Guys:

I want to thank you for this very interesting discussion. Based on the discussion, for my jyrodyne, I'm lowering the aspect ratio of the canard to help keep it from stalling before the main lower forward biplane wing, and plan to add drooped tapers to the ends a la Steve Wittman's on his Tailwind. This should reduce and lower the tip vortices. It's real low below the fuselage anyway, so they should be below the lower wing far enough to miss it during regular cruise. I'm going to be looking at the vortex interactions at high angles of attack to estimate the onset of buffeting of the canard vortices on the lower wing.

Chasbo

 

[i278]

This is a really interesting thread. To add a little, I just came across this Ilan Kroo paper; it's somewhat germane to this topic, but far too general to be of real use in an Avanti discussion. Good fodder for configuration trade studies though:

http://aero.stanford.edu/Reports/MultOp/multop.html

Anyway, back to that foreplane. I'd tend to agree with you, Miper--anhedral in the canard looks suspiciously like a late-stage change. I'm sure the original 'napkin drawing' was perfectly straight.

What puzzles me is that if the tips were dropped to eliminate some undesirable wake/vortex impingment, why such a small amount? I can only guess (so take this for what it's worth), but if Piaggio found that, say, the canard tip vortices were impinging on the propeller disk, the vertical distance they've dropped probably wouldn't be enough to solve the problem consistently. Or am I full of it?

Of course, it could well be that the canard tips were dropped as far as possible allowing for tip over clearance, and the change was just enough to sort things out. Who knows?

Something I can see happening though is that perhaps due to structural/packaging issues, the canard mounting had to be raised. The aero guys, however, weren't confident in raising the tip vortices, for the same reasons Miper and John have listed previously. So, the compromise--tips stay where they are, and the roots move up. Just being the devil's advocate

After all this great discussion, it sure would be nice to get a hold of a Piaggio engineer for a few hours!

 

[Miper]

1278,
Thanks for the reference. The discussions about various configurations are very useful. The author does note that his aero model has some simplifications, so there will be some flight conditions and configurations not covered by these results. Still, they are generally applicable.

Your thoughts about anhedral on the Avanti are good ones. If they used the anhedral to avoid wake interference, it would only have been a useful fix if the wake impingement had occured only at extreme high angles of attack. Again, we would have to hear from the designers to know exactly why they did it.

I know of one business jet with a large amount of tail dihedral that was added after flight tests that showed the airplane needed just a little more dihedral. (Dihedral is dihedral. It matters little where you put it, but you will need more degrees of dihedral on the tail than on the wing to get the same overall effect, because the tail is smaller.) It was cheaper to put it in the tail than in the wing, because the tooling was already built and the tail tooling was much less complex and easier to modify. I have always thought the tail was awkward looking, but it's the best selling jet in the world, so looks aren't everything. There are two reasons for relating this story --1. Many design tweaks can happen in order to obtain desired flying and handling qualities, and 2. It isn't always obvious why some tweaks were done.

 

[JohnFortier]

I have always been mystified by the old F4 phantom layout; Dihedral on the outer wings and anhedral on the tailplane.

I supposse one possibility is that the tail has anhedral to avoid the wing downwash and the wings have dihedral to compensate.

Again, sometimes its not the "normal" reason why things are done, it's some reason you won't think of unless you can read the designers mind.

The Aceair Aericks also has anhedral on the foreplane and, from the front elevation supplied in the web site, the tip vortices should pass well outside the propellor arc or anything else, except it it's so nose high as to be stalled. The anhedral, in this case, is unlikely to be to increase maneuvarability, since the large ventral fin will ensure adequate roll rate.

Anyway, I've broken down sufficiently to contact Aceair and ask them why their design has foreplane anhedral. If they answer, I'll share it with you guys. Whether their reason will also apply to the Avanti is another moot point.

Regards

John

 

[i278]

John...

I'm not sure, but I've heard that the anhedral was added to the Phantom tail to eliminate high-mach-number instability--although what specifically was the problem and how the fix worked I don't know. The dihedral was added to the wingtips, I assume, to maintain the plane's level of roll stability with the minimum tooling impact. (Existing tooling drives a shocking number of late-stage design changes, as Miper aptly pointed out)

The same anhedral on the Harrier, however, is there for an entirely different reason: it, along with anhedral on the wing, reduces unwanted roll due to sideslip. So, reinforcing Miper's comments, looking simply at the fix needn't give any insight into what the fix is fixing

Thanks for asking Aceair; it'll be interesting to see if they actually give you an answer. And, if they do, I hope it's a little more than 'because the Avanti does it that way!'

Regards

 

[JohnFortier]

Hi all,

I receioved this answer from Aceair:

"Dear Mr.Fortier,
Thank for your interest for our Aeriks 200.
First let me answer you to your second question:
The Foreplane anhedral is for improve pilot outside visibility and increased gap between the foreplane wake and the main wing surface."

Which sems simple enough and just about what the discussion here had surmised.

Regards

John

 

[eewings]

I'm a bit late to this thread, which seems to have stagnated...
but my understanding is that anhedral is applied to the
forward wing to reduce dutch roll tendencies. Anhedral
can also be applied to the rear stab for the same reason,
and if you notice on the Avanti both the forward wing and
rear stab have anhedral.

Given all the CFD and wind tunnel optimization that the
Avanti went through, the vortex control aspect of anhedral
that this thread has pursued was likely noticed and
exploited, but I'm wondering if anyone could comment on
this stability aspect of 3SLC anhedral.

 

[Sparweb]

I wouldn't expect that the lateral stability contribution of the nose stab would be very great. Its span is small compared to the main wing: about 1/4.


STF

 

[JohnFortier]

The high horizontral stabilizer, mounted on the tip of the vertical stabilizer, will provide a great deal of lateral stability, for the same reason that a high wing monoplane is stable. Any, undesired, roll will cause side slip which in turn builds preasure under the stabilizer (or wing) on the side toward which the aircraft is slipping. This in turn tends to reduce and straighten out the roll.

the Avanti also has a pair of rather large strakes under the rear fuselage. I initially thought that these were to protect the propellors turning excessive take off rotation, but they may also ensure adequate roll rate, given the aircraft's inherent stability due to the high horizontal stabilizer.

Anhedral on the horizontal stabilizer is probably there to reduce the stabilizing effect of the design, again allowing adequate roll rate.

Anhedral on the forward surface is probably there to ensure that foreplane wake misses the wing and, as mentioned above, to improve pilot visibility. Its affect on stability, compared to the other forces acting on the airframe, will probably be minimal.


Regards

John

 

[eewings]

The delta shaped ventral strakes on the Avanti are most likely stall prevention devices, though they do provide some needed yaw stability. Such strakes were first used on the Lear Jet. At low angles of attack, the drooped strakes provide no lift; but at high angles of attack, they develop lift and raise the tail. The delta shape is key to this behavior.

Why does the Avanti need this?
3LS aircraft are supposedly stall resistant. True, but
this depends upon the balance between the fore and aft wing moments.

An inherent problem with 3LS pusher aircraft is that the loaded CG is farther forward of the landing gear than on conventional aircraft, which are placed to avoid tail sitting when empty. At rotation, the aircraft
rotates around its landing gear; to keep rotation speed low,
the elevator must have a lot of power. More than it needs in cruising flight, enough to bring the main and
fore wing very close to stall during approach,
when they are working hard. The ventral strakes prevent
the nose from lifting too high while extracting the maximum lift coefficient out of the main wing.

The aerodynamic advantage of 3LS (and canard) aircraft is that they need less wing area in cruise than conventional designs since there's less trim drag (to overcome the additional intersection and wing interaction drag of the 3LS configuration, 3LS designers must minimize total wing area). This advantage disappears
at takeoff and landing. More so than canards, 3LS aircraft can have nasty unrecoverable stall behavior, and can't be allowed to stall. They tend to have longer takeoff and landing distances than conventional aircraft of similar
wing loading because of this. The game 3LS designs play is "how close can we get to
the edge", and the Avanti uses every trick in the book.

That's my 2 cents. Please correct me if I'm wrong.

 

[JohnFortier]

I wuld disagree about the stall behaviour of 3LS aircraft, since the stalled surface, assuming this to be the foreplane, is not the control surface, which is the big problem with cannards during stall.

It's probable that the strakes perform more than one function on the design - prop protection, roll rate assist and stall prevention. Certainly, as the angle of attack increases, the affect of the strakes on lonitudinal stability will increase, but I haven't heard any reports of the Avanti needing a long and careful glide path. The takeoff run, at under 900 meters, is pretty typical for aircraft of this size and performance and doesn't indicate any difficulty in getting airborne.

It's going to take a lot of 3LS aircraft in the air before everyone is convinced that the design is the next logical step for subsonic aircraft, now that materials, control systems and manufacturing methods can actually produce an aircraft which can demonstrate the real advantages of this design.

I remain convinced that this is a brilliant design and still intend to buy one.

Regards

John

 

[eewings]

Good point about the separation of trim and control surfaces. The Avanti doesn't have a bad takeoff run, but
it also has gobs of power...

The Avanti is a brilliant design. If I had 4.3mil I
might buy one as well. To me it serves to illustrate that
3LS designs must be very carefully engineered to realize
the benefits of the configuration. This translates into
increased development costs that may be borne better by
the executive jet/prop market than the light/sport aircraft
market. If the configuration proves generally advantageous,
the development premium will shrivel.

It appears that 3LS + single engine tractor propellors are
not a good combination, if the Eagle 150 is any example.
This design tried to marry the practical operating and maintainance advantages of a tractor engine/prop with the
aero advantages of 3LS, but didn't wind up with anything
better than a simpler, cheaper conventional design.

The A-200 has heeded the lessons of the Avanti, and has
carefully optimized the aerodynamic advantages of 3LS in
an unusually roomy single-engine configuration.
The price is an unconventional drivetrain,
a rather long takeoff relative to other designs with similar wing and power loading owing to the tradeoffs I posted
previously, and an arrangement that may not be scalable to
a 4-place cabin ( but could be scalable , curiously enough, to a 6-place ).

Any 3LS designers out there?

 

[Sparweb]

There are many "paper airplane" 3LS designs that I've seen on the 'net. If you want to see a site full of bizarre, and some not so crazy, airplane designs:

http://www.the-wildings.com/aircraft/

That should get you guys going...

STF

 

[RitchieLinford]

Have just completed a project at uni, the brief being to desgin a 1000 seater aircraft still within the JAR regs.

We used 3 LS on our desgin, prooved very helpful in the CofG calcs as was able to put more weight up front which helped with the CofG margin, and we planned to use fuel triming durning flight for stablilty

Quite a good project but all a bit of a pipe dream, 1000 people in one plane, maybe someday

Ritchie