Stall safety and tapered planforms

[chris-67]

Hello everybody,

we are aware that the elliptic distribution of lift yields best results with regard to the optimum loads for a given wing´s planform.
The problem is, how do we get this specific type of distribution when designing a wing with a given "under-elliptic" planform, i.e. not rectangular, not weakly tapered, not elliptical, but strongly tapered! I need not tell you, that a scale modeller will always hide the aerodynamic and physical necessities "within" the wing, keeping the original wing´s planform. He cannot simply change the wing´s topview due to better flight performance.

Together with this problem, another more specific problem exists:
Strongly tapered wings tend to "tip-stall", before "root-stall". That´s bad, as we know.

Now - in order to get elliptic lift, the cl-coefficients of the wing sections close to the tip need to be increased because of the small chords there. Otherwise this may- as far as I know - undermine stall safety.

Will it be possible to find some combination of wing twist (washout) and reasonable airfoil changes along the span to match the requirements of safe stall characteristics, at all?
What shall be done?

-Christoph-

[Sparky Paul]

What should be done?
Keep the speed up!
Here's a shot of a full-scale replica of your avatar, tip-stalling at landing..

[Ollie]

Design the lift distribution. You just know the airfoil zero angle of attack.

See simple the compute tool:
http://aero.stanford.edu/WingCalc.html
Or complex the compute tool:
http://www.amadistrictii.org/cjrcc/wing2/wing.html

[biber]

The increased cls towards the tip are nothing, you will desire, but are a rather natural consequence of the excessive taper. What is desired to keep the wing docile is a tip section, that is cacapable of the big lift coefficients occuring (or chose a root section that stalls way before the tip section does). If you really want to stick with the elliptical lift distribution, that´s the only way to make that wing safe. Too much of twist will make your plane fly like a parachute and may lead to a poor lift distribution aswell.

If sufficient RE-numbers are provided, a thicker, more cambered tipfoil will help. In case of lower RE-numbers, the points of max thickness and max camber should be shifted slightly towards the LE at the tip, while thickness shouldn´t exceed a certain value(depending of the actual RE).

biber

[chris-67]

Thanks for your responses, first of all!

@ Sparky Paul,
my avatar is flying better than this replica, I guess! It´s from www.airworld.de, fibre, span 2.06m=81.1", weight 4.2kg=9.26lb=148oz. Fantastic model. The reason for this question on the forum is an 'A6 Intruder', span 1.80m (71") span, ducted fan jet. I´m in honor bound to do the wing/airfoil calc´s for a friend.

@ Ollie,
I use some German freeware (www.zanonia.de/nurfluegel.zip). It was originally written for Nurfluegel applications, but can well be used for calculating and depicting distributions of circulation (= proportional to lift), cl, drag, induced angles of attack and more. The latest release 2.25 even gives you movable 3D views of your wing! Quite nice - if you know what you´re doing :D
Thank you for your links - I will try them damn sure.

@ biber,
sounds quite sophisticated. From my point of you - I mean, what I read in some technical publications - it may be reasonable to put a little more camber at the airfoil sections close to the tips and then correct the differing zero angles of attack to common cl-values. The latter is done by some downwash.
Common cl-values? Not really, you may argue. If I really tried to select the downwash such that the cl along the wing span are equal again, then I would just have victimized elliptic distribution on the altar of stall safety.
I feel, some compromise has to be made. But how can this compromise be calculated as accurate as possible and reasonable?

Greetings from Augsburg,
Christoph.

[gunsguns]

Hallo,

I would suggest to sacrifice the elliptical lift distribution. Model airplanes do not have the firmness problems that make the elliptical lift necessary for man carrying aircraft, and normally it is not necessary to outturn any other models :-)) I totally agree with biber concerning the measures to make the root section stall first. You will never miss an elliptical lift distribution (unless you make a dogfight) but you will miss forgiving stall behaviour every time you turn into final too slow.

[Sparky Paul]

Intruders are seen as PSS slopers, I doubt much care was taken for the airfoil.
With a jet, low speed is the enemy with their high wing loadings.
Regardless of the finesse used for the wing, keeping the speed up is the only real safety of flight factor.

[gunsguns]

By the way, the A6 Intruder has a significantly swept wing.
The lift distribution on such a wing cannot be made elliptical, since airflow goes outward the peak lift is approximately at half wingspan, very little lift at the root and on the tips (think of an inverted W )

[biber]

Yes, an Intruder would be a cool model. Not the taper is the main prob though (how much is it? not below 0.6 I guess, wich would still be ok), but the additional sweep maybe. Anyway, it won´t be too hard to make it fly well. For such sort of model a bit more induced drag is no problem, as long as you don´t try to win a EDF pylon race competition with it, wich I guess is not exactly your aim.
To the keep-up-speed-issue I´d admit, that typical EDF´s efficiency will suffer badly of lacking airspeed anyway.

biber

[biber]

Oh, just saw, there already does exist an EDF A6 Intruder :D :

http://www.gadgetstuff.com/product.asp?id=12459&cid=

But, well, seems to be not too scale... :rolleyes:

Sorry, couldn´t resist :rolleyes:

biber

[gunsguns]

Hallo chris-67,

I am curious about your DH-88. Thats the most beautiful
twin ever. Have you information about the airfoils that are used? Any twist?

Servus
Andi

[green66]

Christoph,

You may want to try the free LiftRoll spreadsheet LiftRoll (http://www.geocities.com/jebbushell/COOKBOOK.htm) which uses the VLM to calculate lift & Cl distribution for any wing up to 4 panels per side, with taper, sweep, and twist adjustable for each panel.

The screenshot shows the program, which I've modified slightly to allow input of physical twist separate from aerodynamic twist (airfoil transitions) in order to keep track of things, and also to clarify some variables.