Aug 31, 2012, 06:19 AM
Registered User
Thanks Charles, I'll post a pic but it's battered and ugly, I'm just about to start a new one.

Check out this new canard
Sign up now
to remove ads between posts
Aug 31, 2012, 09:37 AM
Registered User
Thanks for showing Velocities' V Twin, Captarmour, it looks like a low coefficient of drag type with a well loaded canard at 3.5 degrees and same airfoils which, to me seems like a great combination. There is a hybrid electric ship now which is similar.

Sep 01, 2012, 01:15 PM
Registered User
Don Stackhouse's Avatar
As we have discussed here a number of times before, there is nothing special about the 3.5 degrees incidence, outside of the context of the plane's other parameters. C/G location is more important, but it too is only one parameter out of a fairly long list, all of which have their influence.

Likewise, having the same airfoil on the canard and the wing will not guarantee that the canard will stall first. If the wing has a high enough aspect ratio that its chord is less than the canard, the lower Reynolds number(s) for the wing could make it stall before the canard, depending on how the other parameters interact.

Getting back to the 3.5 degrees incidence, if you had a C/G location that put the plane out of equilibrium at the intended design flying speed, you would need to apply elevator correction to establish the necessary equilibrium. What matters then is not the canard incidence alone, but the combination of incidence PLUS elevator deflection, in compbination with things like wing and canard planforms and areas, moment arms, C/G location, airfoils, fuselage effects, etc..

If two different airplanes happen to fly well at the intended flight conditions, using the same incidence settings, it is a coincidence, nothing more. Furthermore, within a reasonable range of values, it would probably be possible to set either of them up with slightly different settings, and still have acceptable behavior and performance. It would also be possible to set up either plane with that specific set of incidences, and by "tweaking" some of the other parameters, get totally unacceptable behavior and/or performance.

You have to look at the WHOLE picture, not just one particular parameter.
Sep 01, 2012, 06:34 PM
who has rabbit ears down
Captain Canardly's Avatar
Last weekend I had another bad launch- after 5- the 6th, was half way up tow and no response- ended up cat ate wire again, so fine I didn't see it before flight- she went into a series of 'deep' stalls, and missed pancaking by about 15 degrees radially- got a great shot- gotta get it though
Latest blog entry: My new factory
Sep 01, 2012, 10:43 PM
Registered User
Don, Thank you for the discussion. Lady Luck must have been with me on the Twin. The canard data is: Symmetrical airfoil, three degrees positive incidence, 8.25 inch center chord, 16 degree LE sweep back and 50% of the main wing's area.

The main wing's data is: Semi-symmetrical airfoil, zero incidence, 9 inch center chord, 12 degree LE sweep back.

At low cruising speed and full UP elevator it gently drops the nose to prevent main wing stall.

It seems that my Delta Duck had two degrees incidence on the canard with front motor and large delta main wing which would mush rather than stall.

Sep 02, 2012, 09:00 AM
Registered User
Don Stackhouse's Avatar
Charles, there is a problem with your numbers. The actual incidence (from an aerodynamic standpoint) of at least one of your examples is not what you think it is.

There are a number of different ways to measure the angle of an airfoil, but the three main ones are:
1. from a tangent to the underside of the airfoil.
2. from the airfoil's chord line.
3. from the airfoil's zero-lift line, the line where if that line is set to a zero angle of attack, the airfoil's lift coefficient is zero.

The most common reference in the model world is the chord line, and I'm making the assumption that's where you are measuring your incidences from.

That third reference, the zero-lift line, is the most aerodynamically significant, but the hardest to determine, since it depends on the airfoil shape itself, particularly the amount of camber.

For a symmetrical airfoil (assuming no funny side effects that make it look unsymmetrical to the airflow, such as a turbulator or a laminar separation bubble on one side that does not match the other side), the zero-lift line is the same as the chord line.

For an unsymmetrical airfoil, the zero-lift line will be different from the chord line, and the specific amount will depend in particular on the camber.

CompuFoil (among others) does calculate the difference between the chord line and the zero-lift line. If I load in the popular (but NOT particularly good at model Reynolds numbers!) Clark Y airfoil, it shows a camber of 3.55%, and a zero-lift angle of attack of -3.52 degrees (in other words, when the angle of attack is zero relative to the zero-lift line and the lift coefficient is therefore zero, the chord line is at an angle of attack of -3.52 degrees).

If I cut the camber down to 2%, a reasonable number for a typical sport model at typical model Reynolds numbers ("Re"), the airfoil is now very clearly semi-symmetrical, and the zero-lift angle of attack relative to the chord line is now -1.98 degrees.

If I look at a popular modern low-Re airfoil, Martin Hepperle's MH-32, it shows a camber of 2.37% and a zero-lift angle of -2.42 degrees. If I reset the camber to 2 degrees, the zero-lift angle becomes -2.05 degrees. Note, the camber is now the same as the modified Clark Y we looked at above, but although the zero-lift line is close to that of the modified Clark Y, the exact numbers are NOT the same.

You can make rules of thumb about so-much-camber results in such-and-such a zero-lift line location, but they will only be approximations at best. For example, if I look at an MH60 airfoil ( a semi-symmetrical section used for flying wings, with a slightly reflexed trailing edge, so that its aerodynamic pitching moment coefficient is zero) and set its camber to 2%, the zero lift angle relative ot the chord line is at -0.63 degrees. Likewise, if I look at a Selig 4083, which has a fair amount of "aft loading" (a lot of positive camber near the trailing edge) and set its overall camber to 2%, the zero-lift angle is -2.28 degrees relative to the camber line. The shape of the camber as well as the amount is important.

In any case, the exact amount of camber is important. Just specifying "flat-bottomed", "symmetrical", or "semi-symmetrical" is nowhere near close enough! You MUST look at the actual airfoil, and its exact shape, particularly the exact amount of camber!

So, in your "Twin" example, we know the canard airfoil is symmetrical, and we will assume that to the airflow it does look symmetrical (a significant assumption that might not always be valid, as I mentioned above). In that case, the zero-lift line should be the same as the chord line, and the incidence relative to the zero-lift line is 3 degrees relative to the aircraft datum.

Let's assume your "semi-symmetrical" wing airfoil is our modified Clark Y with 2% camber. In that case, your true aerodynamic incidence is the geometric incidence relative to the chord line, PLUS the difference of 1.98 degrees between the chord line and the zero-lift line, so its aerodynamic incidence is +1.98 degrees. That means the true aerodynamic incidence between your wing and canard is only 1.02 degrees!

If you made another one using the same airfoil for both wing and canard (so the zero-lift lines were the same), the true incidences would be different from those of your present Twin.

As far as stalling of the wing vs. canard, we need to look at more than just the root chord. The Mean Aerodynamic Chord ("MAC") would be a better reference, but the fact is that stall can begin anywhere along the planform, so you have to look at the entire flying surface. For example, if the taper ratio of your wing was greater than that of the canard, the Re's at the wing tips would be lower than for the canard tips, and therefore they could stall at a lower angle of attack, assuming all other things being equal.

As I've said before (and many times), you MUST look at all the parameters, and how they interact together!
Last edited by Don Stackhouse; Sep 02, 2012 at 09:44 AM.
Sep 02, 2012, 10:00 AM
Registered User
Don, A great discussion! I trust that all viewers enjoyed it as much as I did. As shown in the pictures, I used your example #2 with a zero degree chord line which has worked well in the past. The decalage is probably very close to the one degree as you said. The Twin, just as my Long EZ does, breaks loose from the runway without UP elevator so that it is hard to determine when it happens. The Twin's wings will be scaled down by 7% to a lighter version for a single motor placed at the CG in the middle of the fuselage. It becomes addictive to me when things evolve into a good combination.


As far as stalling of the wing vs. canard, we need to look at more than just the root chord. The Mean Aerodynamic Chord ("MAC") would be a better reference, but the fact is that stall can begin anywhere along the planform, so you have to look at the entire flying surface. For example, if the taper ratio of your wing was greater than that of the canard, the Re's at the wing tips would be lower than for the canard tips, and therefore they could stall at a lower angle of attack, assuming all other things being equal.
I was fortunate here to have ended up with the lower chord up front and the lower taper at the rear. The area of the wings was being changed without much thought to put the CG at exactly mid fuselage.

Sep 02, 2012, 10:25 AM
Registered User
Don Stackhouse's Avatar
The other thing in your favor may be the tip fins. Although it's complicated, and they don't always work this way, tip fins/winglets can in at least some cases delay tip stall.
Sep 02, 2012, 05:27 PM
Registered User
Don Stackhouse's Avatar
Another thing we need to be aware of: although we need to determine the design incidence at the root in order to build the fuselage, the initial incidence determination needs to be at the MAC, the chord on the wing where all of the forces are collected together, and seem to act as one. The flying surface's Aerodynamic Center ("AC") is typically on its MAC about 25-28% aft of the MAC's leading edge.

The MAC is NOT the same as the average chord, unless the wing has no taper (in which case it's the same as all of the other chords!). For a single-taper (trapezoidal) wing, it's at the spanwise location where the area outboard of it is the same as the area inboard of it, so it's a little inboard of the mid-span if the tip is narrower than the root.

I've attached a program written by Chuck Anderson, that calculates the MAC and AC of an arbitrary flying surface by breaking it up into a series of trapezoids.

Where this is important for incidence is if the flying surface has any twist ("washout" or "wash-in"). If a typical trapezoidal wing has two degrees of washout, then a little less than one degree of that will occur between the MAC and the root. Once you have calculated the necessary incidence at the MAC, you need to add that difference to it to find the necessary incidence at the root.

Just another example of why you have to look at the whole picture, not just one parameter by itself!
Sep 03, 2012, 08:37 AM
Registered User
Hey guys just want you to know I love this forum, especially the wisdom of Don. Keep it coming.
Sep 03, 2012, 10:49 AM
Registered User
Sorry for this outofbox question, but I'm looking for an airfoil for a straight wing, with strong pitch up at stall. Preferably as 'flat bottomed' as possible. I'm supposing something with a lot of aft camber.
Sep 03, 2012, 11:15 AM
Registered User
Don Stackhouse's Avatar
"Flat bottomed" and "aft camber" is somewhat contradictory. Those requirements push you into a lot of thickness, which at our Re's is bad.

The "pitch up at stall" criterion essentially requires a trailing-edge type stall that begins with a fairly large area of initial separation, followed by more gradual separation over the remaining forward portions of the airfoil. Keeping the aft 20-30% of the upper surface flat or even slightly concave can produce this effect. The flat portion will let go more or less all at once, causing the lift vector to move forwards.

We did exactly this on our '94 version of our "Monarch" 1.5 meter RCHLG. The original '93 version had such utterly benign stall characteristics that it wa possible to get into a "mush" in a thermal turn and lose too much altitude before realizing it. We wanted to maintain gentle stall behavior with good roll control, but make the stall more well-defined. I flattened the last 20% of the upper surface of the root airfoil, while keeping the airfoils at the poly break and tip the same as before. The flow near the root would separate suddenly, with the pitch-up due to the forward shift in the lift distribution making the effect more obvious, while the rest of the wing kept on flying, with good stability and control. The stall was still gentle, but much more obvious.
Sep 03, 2012, 11:39 AM
Registered User
Im trying to picture the 'last 20 to 30% flat. Ok r u saying that at about 70% of the top surface it would flatten out so that lower surface would come up to forma a point or the lower surface remains 'flat', leaving TE blunt?
Or concave so as to get an 'S' camber line?

I was looking at lippisch patent 3830179 where he describes a 5% flat wing with round LE and the upper surface streamlined from about 75% (last 25%) down to meet flat lower surface making a sharp TE.

Im thinking something vaguely similar may start flow separation at the TE and pitch up at or near the stall.

(Interestingly I tried an AR 0.6 flying wing at roughly 5% thickness with round LE and blunt TE, symmetrical, and when hand tossed hard it would pitch up hard before settling into glide. With a 45 degree cambered LE it would do the opposite, ie, tuck hard but with round LE and top aft 60 degree bevel it went straight as an arrow. I can't quite figure out what is going on there!)
Sep 03, 2012, 12:00 PM
Registered User
Don Stackhouse's Avatar
You're overthinking it. "Flat" means just that, flat. As in "not curved". If you hold a ruler's edge chordwise against the upper surface back there, it will touch along the entire surface of the last 20 or 30% of the upper surface, with the surface of the airfoil forward of there being convex by various amounts. What is going on with the lower surface or the trailing edge is a totally separate matter, we're just talking about the shape of the aft part of the upper surface.

This means that the adverse pressure gradients in the flat region will be nearly the same over that entire portion of the surface, so when the angle of attack gets too high, and the adverse pressure gradients get too high for the flow to overcome, that entire 20-30% reaches that condition at the same time, and the flow over the entire rear 20-30% separates all at once. The forward 70-80% of the airfoil continues to lift, but the sudden loss of lift in the aft portion of the airfoil causes a sudden pitch-up.
Sep 03, 2012, 01:04 PM
Registered User
Ok gr8! Im with you. Thanks a mill.

Any comments on the last part of my last post?

How is your flying wing project coming along?