Drag rudders and nurflügels

[surfimp]

Anyone tried these frise type of drag rudders on all-wing models, or care to comment on expected performance? At some point I would like to make a finless flying wing and was thinking that this structure would be easier to create in a smaller sized model than the more "conventional" split surface drag rudder design typical of the Horten and Northrop planes.

Please ignore the extraneous features of this drawing (rubber band & eyelet on lower surface, slot hinge, reflexed profile), I use it only because it provides a useful depiction of a "dragplate" style control surface.

Thanks!

Steve

[JMP_blackfoot]

We have made quite a few finless wings according to the Horten design, and haven't yet found that a drag rudder was necessary. The Northrop wings did (and still do) use the split drag rudders, but then they used a linear twist design, straight from root to tip, which the Horten don't. A rule of thumb is to keep the center 40% - 45% of span untwisted, and gradually twist more and more to the tip. 8 -12 degrees washout at the tip works for us. C.G. is fairly critical. A bit forward is safe, too much forward gives a dive, a bit rearwards is dangerous. To find the right C.G., move it rearwards carefully until the model can be made to actually stall and drop a wing (sometimes going shortly into a falling leaf type of spin), then move it back forward a bit. You won't see any adverse yaw, and your model will be a pleasure to fly.

I would advise to make a simple one without drag rudders, then decide if you really need them. If you need them, you might as well make them a separate control of its own.

A successful design is shown as an example.

[surfimp]

Thank you! Beautiful plane there. Sounds like some good tips; have you tried them on glider versions as well? If I do the drag rudders, I am definitely planning to have them as separate controls from the elevons. 8-12º of washout--my! That ought to do it :)

Steve

[JMP_blackfoot]

have you tried them on glider versions as well?

Yes, a 30 gram glider version has been extensively flown by my friend Gerard Jumelin, from dunes by the seaside, in mild to average winds. This design has been made in several sizes, from 50 cm span, 28 gram weight, up to 110 cm, 200 gram with EDF and 125 cm, 600 gram with Speed 480. No reason it would not behave the same in a larger size.


If I do the drag rudders, I am definitely planning to have them as separate controls from the elevons. 8-12º of washout--my! That ought to do it :)

Steve

It also depends on the intended speed of the model, of course. A fast model requires less washout than a slow one. It is better to have to raise the elevons at slow speed than drop them at high speed. Otherwise said : it is better to compensate for a lack of twist by raising the elevons than dropping them to compensate for too much twist.

[Ollie]

Swept back wings have yaw to roll coupling. When the wing is operating at zero lift (going vertical) there is zero yaw to roll coupling but as the wing starts to produce lift, adverse yaw sets in and increases as the square of the angle of attack or lift coefficient. Adverse yaw can result in adverse roll because of the yaw to roll coupling and this can result in not being able to turn or even turning in the wrong direction if the angle of attack is too high.

The idea of drag rudders is to control yaw. Application of enough drag on the inside of the turn can compensate for adverse yaw and result in a coordinated turn at the expense of an increase in drag.

The genius of the Horton twist distribution is that it involves enough twist so that the wing tips are lifting down instead of up. Under these conditions the adverse yaw effect is reversed and, with enough twist, even proverse yaw can result. With just the right twist, resulting in just the right lift distribution and in combination with just the right sweepback and taper, the turns are automatically coordinated and the handling becomes simple and sweet. The Horton twist distribution produces these desirable results at the cost of increased induced drag even when the plane isn't turning.

Another twist distribution is the Culver twist distribution which minimizes the induced drag but results in a lot of adverse yaw so that some form of drag rudders become very important.

Yet a third twist distribution is the linear twist distribution which is easier to design and manufacture. It is a compromise between the two previously mentioned twist distributions. The Panknin twist formula establishes the twist, center of gravity and trim for a linear twist distribution, in conjunction with a given aspect ratio and taper. It requires a milder application of drag rudders than the Culver twist distribution.

All of the above design approaches are based on a linear taper and constant sweepback angle. If the taper isn't linear or the sweep back line is curved, then the design process becomes even more complex requiring vortex panel methods that are only practical to calcualte with a computer.

Flying wings are extremely difficult to design because of the complex interactions between design decisions involving twist, taper, aspect ratio and airfoil selection. A change in any one of the afore mentioned design decisions can affect all of stability, trim, efficiency and handling qualities.

[JMP_blackfoot]

The genius of the Horten twist distribution is that it involves enough twist so that the wing tips are lifting down instead of up. Under these conditions the adverse yaw effect is reversed and, with enough twist, even proverse yaw can result.

Actually, the tips should be neither lifting up nor down, they are supposed to operate at zero lift when flying in a straight line. But, due to the lift distribution, the local airflow is upwards, so the tips are at a negative angle of attack RELATIVE TO THE FLIGHT PATH, and at zero angle of attack relative to the (local) airflow.

If for example the right elevon moves up for a right turn, it creates downward lift. This lift is at right angle to the airflow, and since the airflow is at a negative angle to the flight path, the lift is down and towards the rear. It creates a force down and backwards, which helps pull the right wing tip down and rearwards.
In the meantime, the left elevon moves down the same amount and, due again to the negative angle of the local airflow, creates an up and forward force which pulls the left wingtip up and forwards.
Drag forces are not accounted for here, because the tip airfoils being symetrical, the drag is also symetrical, and is equal on both sides, creating no yaw.
This is the theory of the Horten proverse yaw effect, also called "negative induced drag" by Reimar Horten. In practice, models made along the lines above are easy, safe and a lot of fun to fly, plus they are extremely pleasing to the eye :cool:

[JMP_blackfoot]

All of the above design approaches are based on a linear taper and constant sweepback angle. If the taper isn't linear or the sweep back line is curved, then the design process becomes even more complex requiring vortex panel methods that are only practical to calcualte with a computer.

A design rule for flying wings is that for a given local lift, as dictated by the chosen lift distribution, one can trade lift coefficient for local wing chord.
This means that if the chord is locally X% less than the ideal straight taper, then the local angle of attack can be increased so the local lift coefficient be X% higher. What matters is the distribution of (Lift coefficient x Chord). Now the problem is in knowing the local slope of the lift curve...
The constant sweep back angle matter can be strayed from, but then you'll have to find the correct C.G. by trial and error.

[surfimp]

Yes, a 30 gram glider version has been extensively flown by my friend Gerard Jumelin, from dunes by the seaside, in mild to average winds.

30g! Wow! Is/was it made from EPS? I don't even know if I have any radio gear that is that light. The smallest and lightest glider that I fly is 80cm span and 100g AUW! But it is made from EPP and is very durable ;) That glider must have used magnetic actuators?

To both of you who have responded, the advice and summary of the various twist methods has been most helpful. Thank you!

Steve

[JMP_blackfoot]

That glider must have used magnetic actuators?

Check this : http://perso.wanadoo.fr/jmquetin/1indoor/Variante.htm
and this : http://perso.wanadoo.fr/jmquetin/1indoor/Intex03e.htm
and this : http://www.inter-ex.com/english/interex18/bild703e.htm#25

For details of how these little things fly outdoors (outfly doors ?).

To both of you who have responded, the advice and summary of the various twist methods has been most helpful. Thank you!

Steve

Always a pleasure to spread the word on Horten wings :p

[Ollie]

I can believe that the wing area near the tip can have zero lift at some design coefficient of lift, angle of attack and trimmed airspeed. Then if the angle of attack is reduced and the airspeed increased, (say by moving the CG forward so that the twist isn't affected by control deflection) I do believe that the wing lift distribution near the tip will go negativeof course the lift distribution at the extreme wing tip is always zero. Having said all that, I do agree that the design lift distribution should be bell shaped and only go to zero at the extreme tip.

[JMP_blackfoot]

Ollie,

I just tried to stay near the theoretical ideal basic idea, and explain proverse yaw in the simplest terms I could (When I explained it to myself like this the first time, I was surprised how simple the principle really is :p ). If you're interested, I have done the basic theoretical calculations for the Bell-shape distribution (Sin3) at a given lift coefficient, as an exercise, and come up with a simple accurate formula to calculate the necessary twist of a straight taper wing (ignoring the sweepback and "middle" effects). Will post it when I'm back home tonight.

[surfimp]

Will post it when I'm back home tonight.

Post away, my friend, post away! I look forward to reading it! :)

Steve

[JMP_blackfoot]

Here is the results of my calculations. They check ok with the curves given by Reimar Horten in the book Nurflügel by Horten/Selinger.
The first part in the bracket defines the induced angle, i.e. the angle of the local airflow.
The second part defines the angle of attack of the airfoil relative to the local airflow.
You will note that all angles are proportional to the wing lift coefficient, hence the incentive for our other design, the Twisteuse (look for this in the Indoor and micromodel forum).
The position of the CG given graphically is actually too much forward. Test over the proverbial long soft grass. ;)

Hope it all makes sense.

[surfimp]

Thanks much! :)

Steve

[surfimp]

So I have another question: do you think that the ultra-low wingloading and small size (i.e. low inertia) of these planes help their finless performance? At 125cm and 600g (the size of the largest model you mentioned), that's still a relatively small model. Wouldn't a larger and heavier model with higher wingloading and inertia tend to be less forgiving, more prone to entering unrecoverable stalls, etc.?

I mean, to take this to its "logical" conclusion, if this combination of airfoil, twist, and etc. is sufficient for yaw stability, then why do fullsize flying wings all have drag rudders of some sort? Not meaning to be a pain but just trying to get this sorted out, as my applications will tend towards the larger/heavier end of the scale. I guess the 1/4 scale Ho 229 is a bit of a handful to fly, is it not?

Steve

[JMP_blackfoot]

I think the stall is not something a Horten wing can do, when balanced properly.
At the Inter-ex, I was fortunate to witness the flights of the model seen here :
http://www.inter-ex.com/english/interex18/bild203e.htm
and it definetly did not appear to be a handful. Just beautiful smooth flying.
I couls see not drag brakes on it.
Neither are there any on this full-size :
http://www.nurflugel.com/Nurflugel/Horten_Nurflugels/PUL-10/body_pul-10.html

Now, as pointed out by Ollie, the Horten lift distribution is not the most efficient manner to design a flying wing. It incurs an increase of at least 33% induced drag compared to an elliptical lift distribution. Whether this is a concern is a designer's choice. This loss can be compensated by increasing the aspect ratio accordingly. But the stability is there indeed.

[surfimp]

Sounds good. From your earlier links, I had found those pictures of the large Horten. This time I found the video...wow, what an amazing airplane! It does not appear that they have any major (if any) yaw stability problems!!!

It's funny that in the Electric > Jets forum people debate the viability of their little EDF Ho 229s, but here we have a LARGE one flying no problem on turbine power. Obivously there are differences in RE that play into this as well, but then on the smaller scale we have your examples. No built-in engine nacelles that I've seen so far but still very nice little planes. Any hope of videos for your little Hortens?

And my last question for now: any general suggestions on how to effect the twist exponentionally from the middle of the wing towards the tip? I assume your planes are built with solid core foam construction, and cut with a hotwire, or is this inaccurate?

Thank you!

Steve

[DaveGherardini]

looking at the cockpit pic of the pul-10 i see it has pedals. i assume for nose gear only? how could you handle a 90* crosswind landing without some kind of yaw control. im use to crabbing to keep heading before wheels down on windy days when wind is crossing the runway. very interesting lil aircraft there. Obviously not needed on rc wings without gear, But could still be useful in certain situations..

[DaveGherardini]

well on the spec pages it says its is equipped with drag rudder controled via cables. that makes alot more sense now. I think ill just read the whole thing next time...hehee