Stability in a swept wing with no washout from the elevons alone.

Discussion in 'Aircraft Design / Aerodynamics / New Technology' started by Michealvalentinsmith, Dec 3, 2009.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Dec 3, 2009 #1

    Michealvalentinsmith

    Michealvalentinsmith

    Michealvalentinsmith

    Well-Known Member

    Joined:
    Oct 14, 2009
    Messages:
    361
    Likes Received:
    6
    Location:
    Rainbow beach QLD Australia
    There's been a bit of discussion on fancy twists and sweep in swept wings from bell distribution to elliptical. The consensus is that building in the twist and getting the sweep, washout and taper correct is quite hard with a narrow range of workable solutions.

    But what about untwisted designs that rely on the elevons? Don Mitchel did it successfully with a range of Mitchel wings from the B10 on, using detached slotted elevons. He used a pretty constant foil across the span with no significant twist.

    He got pretty bad adverse yaw by all accounts and the spin could be nasty if the CG was aft. But it definitely worked with a reasonable LD at about 16:1 at the time. And he used sweep platform well under the ideal 20 degrees - that looked to be about 10 or 15.

    Later modification increased the elevon size and set them up further and eliminated the adverse yaw by all reports.

    So rather than designing in a difficult twist - or using the hassles of a detached elevon, needing mass balancing etc, is there any reason why an elevon that's maybe 70% of the tip chord at the tip and tapers to zero at the mid span can't be used to induce the twist at the tips without twisting the foil?

    This way you can tailor in the required "twist" by adjusting the elevon settings and CG. I know of quite a few RC wings that do this (Paiole, Zaggi etc). Is there any compelling reason it won't work full scale?
     
  2. Dec 3, 2009 #2

    Autodidact

    Autodidact

    Autodidact

    Well-Known Member

    Joined:
    Oct 21, 2009
    Messages:
    4,513
    Likes Received:
    799
    Location:
    Oklahoma
    Not exactly what you described - apparently Boeing was interested in flying wings at one time, they may have actually built a fighter prototype with this type of wing before the war:
     

    Attached Files:

  3. Dec 3, 2009 #3

    Autodidact

    Autodidact

    Autodidact

    Well-Known Member

    Joined:
    Oct 21, 2009
    Messages:
    4,513
    Likes Received:
    799
    Location:
    Oklahoma
    Somewhat closer to what you described, but still not exactly. The inverted airfoil at the tip I don't understand at all:
     

    Attached Files:

  4. Dec 3, 2009 #4

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,739
    Likes Received:
    5,420
    Location:
    Orange County, California
    No reason at all that it won't work, but recognize two things:

    1) There will be a rather significant drag penalty at the design point for having to carry deflected control surfaces. Given what you're designing, that may or may not be of any consequence to you.

    2) However, there's no getting around the fact that some aft stick will have to be held to keep the nose up at all times. At the very least you'll want to include some kind of stick-force device (what most pilots are talking about when they say "trim"), either by spring or servo tab. This is the point of the twist in the first place. To set the trim for the design flight point with no control-surface deflection.

    Actual pitch stability is determined solely by the CG position relative to the neutral point of the wing planform and other parts of the airframe exposed to the outside flow. Twist and control deflection don't play a part in that.
     
    ShamJ and plncraze like this.
  5. Dec 3, 2009 #5

    bmcj

    bmcj

    bmcj

    Well-Known Member HBA Supporter

    Joined:
    Apr 10, 2007
    Messages:
    12,918
    Likes Received:
    4,867
    Location:
    Fresno, California
    You know someone is bound to ask if reflexed tips on a swept wing affect stability and NP position. :gig:
     
  6. Dec 3, 2009 #6

    Norman

    Norman

    Norman

    Well-Known Member

    Joined:
    Nov 29, 2003
    Messages:
    2,880
    Likes Received:
    918
    Location:
    Grand Junction, Colorado
    Who says it's "quite hard"? Building it is just a mater of careful jigging and the math to figure it out is available from many sources. Spreadsheets based on the Panknin/Schenk formula for linear twist can be downloaded from half a dozen web pages. You can find more complicated looking math in several books but they all come out within a fraction of a degree. It's what I used for the twist calculation in my spreadsheet. The program "nurflugel" , linked in this thread, will calculate a bell shaped lift distribution but the interface is in German so it's kind of hard to use for a lot of people. The truth is that a swept 'wing gets its directional stability from the static margin not the twist. Many German modelers who use "nurflugel" are putting the CG farther forward than the program suggests.

    I wouldn't call two planes a range.


    The B-10 has one airfoil section but the U-2 has at least two sections and 4 degrees of geometric twist. Since I only have data on one of the airfoils I can't calculate how much aerodynamic washout it has but it's probably not far from the geometric twist. The stabilators are also twisted 10 degrees because the plane doesn't have enough washout.

    Both symptoms of too far aft CG

    12 degrees on the 1/4 chord
     
  7. Dec 3, 2009 #7

    bmcj

    bmcj

    bmcj

    Well-Known Member HBA Supporter

    Joined:
    Apr 10, 2007
    Messages:
    12,918
    Likes Received:
    4,867
    Location:
    Fresno, California
    I think the B-10 also benefits from a little bit of pendulum stability in pitch due to the pilot (a significant percentage of the total weight) hanging below the wing.
     
  8. Dec 3, 2009 #8

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,739
    Likes Received:
    5,420
    Location:
    Orange County, California
    I know. It's becoming my personal crusade to stamp out the "stable airfoil" myth. ;)
     
  9. Dec 4, 2009 #9

    Michealvalentinsmith

    Michealvalentinsmith

    Michealvalentinsmith

    Well-Known Member

    Joined:
    Oct 14, 2009
    Messages:
    361
    Likes Received:
    6
    Location:
    Rainbow beach QLD Australia
    Many thanks guys. I've made some inaccurate assumptions about the Mitchel wing over the years based on my own impressions - and it seems I still am. (Though I almost guessed the sweep).

    The Lippish an other pics (many thanks) are the same ones I got the original idea from - nice to see others with those pictures think the same.

    When I saw the Mitchel wing foil in some construction pics, I assumed it was symmetrical - I was later told it was an NACA foil (I think) and definitely not. The guy did seem to indicate it was the same foil across the span. I know one guy increase the performance by changing to a different foil with a higher CL max but i don't know how that effected stability.

    I asked one guy about the twist and he said it was negligible - I wouldn't call 4 degrees negligible - thanks for that Norm - that changes things quite a bit. As for building in the twist and difficulty, well I have a lot more trouble designing it correctly than in planks - and a lot more trouble building it - but this is RC wings. I almost always need to fiddle with the wing afterwards as well. I was referring to Topaz's comment on another thread about the narrow range of workable options without using additional devices like slots. (But is is very interesting to see Mitchel got such a low sweep angle to work - you could argue the elevons are slotted though).

    Regarding pendulum stability this is not much of a factor in this case. I know from my own experiments (and later confirmation by professionals) this doesn't really come into play until the mass is about 2 chordwidths below the wing, or reaches a specific percentage of the total weight that I forget now - and the pilot mass must be fixed on the pitch axis - which it is to a degree in the B10 Mitchell wing - but not in hang gliders (unless the pilot white knuckles the bar).

    Thanks Topaz for the comment. You hit the main problem on the head. I've been wondering how such elevons can be used to maintain pitch stability by being fixed at a given washout angle, but still work as pitch control devices. This isn't a problme in RC wings with servos but I expected it would be full scale. The solution for me I think is not a trim system - but not to use the elevons as elevons at all.

    If I ever build this pie in the sky idea the concept is to put the pilot inside the wing much like the Horten Hoxa/b and c. To get around the high speed landing with elevons I planned to use weight shift for pitch. This should eliminate the trim problem you mention and the elevon are now really ailerons and only moved for roll - fixed in pitch. (it also gets around throw problems where your rolling and want to pitch up and run out of elevon throw)

    But this now introduces the problem of pitch return with the pilot inside the wing as I've now lost the overhead hang point and swing effect. Simple rollers would have an unstable and divergent effect - pitch down and the pilot wants to roll further forward, and pitch down further etc - so hands off would be a disaster.

    I tried all sorts of solutions like rubber return bungees before coming up with the simplest - curved rollers. So the pilot's attached to the wing with curved rollers in a curved slot either side. This should give the pitch return forces required and can be tailored to give whatever return force needed by adjusting the curvature.

    Nobody has ever build a commercially successful, footlaunched pilot in wing glider since the Hortens and I'm not so vain as to think i can do better than real experts. But the concept is tantalizing and offers so many benefits. The problems in the past all seem surmountable to me.
     
  10. Dec 4, 2009 #10

    Michealvalentinsmith

    Michealvalentinsmith

    Michealvalentinsmith

    Well-Known Member

    Joined:
    Oct 14, 2009
    Messages:
    361
    Likes Received:
    6
    Location:
    Rainbow beach QLD Australia
    Regarding reflex tips there seems to be some evidence that the foils selected for the tip can influence the overall stability quite a bit - but I'm not sure about reflex.

    The article I have is for RC wings but it seems relevant - I can post it here if anybody wants. They claim inverting the tip foil has a substantial effect on overall stability as the pitch recovery force is now in the required direction.

    I find it simpler to think of the tips in flying wings as two smaller and more closely coupled tails (though this doesn't account for spanflow problems with sweep etc) and we discussed the effect of inverted foils in tradition tails in STOL aircraft elsewhere.

    The theory is inverting the tip foils will also direct the resultant lift vector from aileron deflection such that adverse yaw is eliminated as well. I have the diagrams somewhere.

    As Norm pointed out Lippish's wings may have worked ok - but the result of transitioning the foils from the root to the tip, resulted in a less efficient machine than the Hortens wings. Which is interesting when you consider the Horten wings typically started with a somewhat positive Cm root foil (with what looks like reflex in the diagrams) to a symmetrical tip foil that was geometrically washed out out as well.

    I think the analysis I have puts one of their sailplanes at an overall CL of not much more than 1.0. You'd think stating with a higher Cl max root foils and morphing to an inverted tip foil would give a better overall Cl max (and the RC article I have seems to say this). But the evidence seems otherwise. The Hortens wings certainly were more successful than Lippish.
     
  11. Dec 4, 2009 #11

    Michealvalentinsmith

    Michealvalentinsmith

    Michealvalentinsmith

    Well-Known Member

    Joined:
    Oct 14, 2009
    Messages:
    361
    Likes Received:
    6
    Location:
    Rainbow beach QLD Australia
    You'll note on the TWIT site where these pics are posted they report that the early wings he made prior to this design didn't roll well - among other things.

    He the decided on this seed shape (I forget the exact word now) and the roll response was quite good. Interesting the shape pretty much results in what Urden found most effective as well. Most of the elevon as far out toward the tip as possible. Lippish just rounded the end off.
     
  12. Dec 4, 2009 #12

    Autodidact

    Autodidact

    Autodidact

    Well-Known Member

    Joined:
    Oct 21, 2009
    Messages:
    4,513
    Likes Received:
    799
    Location:
    Oklahoma
    Zanonia seed. Apparently the trees have done some "experimenting" as well :grin::
     

    Attached Files:

  13. Dec 4, 2009 #13

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,739
    Likes Received:
    5,420
    Location:
    Orange County, California
    I should clarify - a narrow range of planforms and CG locations where there is both small induced drag, acceptable stall characteristics, and positive static stability. The Lippisch/Northrop style 'fix' was to design for the former and latter, and deal with the stall characteristics issue with slots and such to prevent tip stall, which in a swept wing means massive and usually uncontrollable pitch-up at the stall. Nasty. The Hortens tended to design for small induced drag and stall characteristics, and accepted a relaxed level of stability as the 'price' for those virtues. Getting all three in one airframe tends to constrain the design variables to a rather extreme degree.

    Okay, I'm going to be pedantic here, but it's really critical that you understand the difference between stability and trim, and which one is applicable to twist (washout) and things like reflexed airfoils and such.

    Pitch stability is only a product of the planform shape and the CG location. Period. Twist, control surface) deflection, and airfoil selection have virtually no effect upon the stability of the aircraft. For our discussion here, you can neglect those factors completely. For stability, it's only CG location and the planform, which latter determines the neutral point. The difference in location between the neutral point and the CG location determines the static stability margin of the airplane.

    Now, twist (aerodynamic or geometric), airfoil selection, and control surface deflection all change the lift distribution of the wing. The way the lift is distributed along the span also determines (in a swept wing design) the way lift is distributed fore and aft of the CG. More lift inboard (forward of the CG) and less outboard (aft of the CG) means a moment that will pitch the nose up, the opposite creates a moment that pitches the nose down. The sum total of airfoils, twist, and control surface deflection create a distribution that determines the airspeed (angle of attack) at which the aircraft will develop zero pitching moment and have no forces that want to change its pitch attitude - the airplane is trimmed, in the engineering sense. If it's also stable, it will remain at that angle of attack until disturbed by the pilot or turbulence, and then return to that same trim point once the disturbance is over. So trim (twist, airfoil selection, control surface deflection) sets the trimmed airspeed, and stability maintains it.

    If you put the control surfaces in trail (zero deflection), the amount of twist and airfoils used will be chosen to develop zero pitching moment at the design airspeed, whatever the designer decides that will be. Hopefully the airfoil (or airfoils) is(are) also chosen to be at it's(their) design lift coefficient at that speed, too. And give you the lift and drag characteristics you want. To change the airspeed at which the airplane will be trimmed (faster or slower), you move the pitch control surfaces. On a swept flying wing, this changes the lift distribution and the airplane pitches until the sum moments are again zero. Then it stabilizes at that new airspeed because of the locations of the CG and neutral point. On a flying plank, you're effectively changing the reflex of a portion of the wing, and therefore the pitching moment of that section, and so on to change the pitching moment of the entire wing, same as a swept flying wing does by changing the distribution of lift along the span (and therefore fore and aft of the CG).

    Sorry to run through this yet again - hopefully I've done a better job of describing it this time - but there are simply vast amounts of incorrect information out there (especially on the Web) about stability and control of flying wings. You'll often hear about "stable airfoils", especially in conjunction with flying planks, and that is pure mythology. Airfoils (and twist, etc.) never have, and never will, have a significant effect upon stability on any airplane, and that includes flying wings. The sooner we stamp out these myths, the sooner we're likely to have some more successful flying wing designs soaring around the skies.
     
    Last edited: Dec 4, 2009
  14. Dec 4, 2009 #14

    Norman

    Norman

    Norman

    Well-Known Member

    Joined:
    Nov 29, 2003
    Messages:
    2,880
    Likes Received:
    918
    Location:
    Grand Junction, Colorado
    That's a myth. The Hortens designed their planes with quite large static margins some of the pilots chose to ignore the design CG because they mistakenly believed that they could improve their contest scores by forcing an elliptical lift distribution. Because of the high taper ratio and the funky twist they (the pilots not the designers) just made the plane dangerous.

    I couldn't agree more. And like all tedious pedants you couldn't resist slipping your favorite myth into the lecture
     
    Last edited: Dec 4, 2009
  15. Dec 4, 2009 #15

    Norman

    Norman

    Norman

    Well-Known Member

    Joined:
    Nov 29, 2003
    Messages:
    2,880
    Likes Received:
    918
    Location:
    Grand Junction, Colorado
    It's a vining cucurbit that climbs trees.
     
    Last edited: Dec 4, 2009
  16. Dec 4, 2009 #16

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,739
    Likes Received:
    5,420
    Location:
    Orange County, California
    And you never can resist making personal attacks when someone says anything other than absolutely glowing about your own personal idols. It's like I'm committing sacrilege or something. If you're going to snipe at me every time I don't put the Hortens up on a pedestal, this is indeed going to get tedious very quickly.

    While I do not appreciate the snide way in which you did it, I do appreciate that you got me to go back and check my references, which I should've done in the first place. I'm going to retract my statement about stability in the Horten's gliders, because I've gone back and calculated the static margin on the HIII through HVI from Table 2.7.1 in Tailless Aircraft in Theory and Practice. You're quite correct, those aircraft had quite large static margins with the CG in the designed position.

    As I further re-read the pireps and design discussions about the Horten's work, it seems I had it backwards. The Hortens kept good controlability in pitch, and static margin in their aircraft, and sacrificed increased induced drag to get it. In fact, the H-IV was carrying as much as 13.4% extra, by the calculation in my reference. (Since I know you're going to argue this point, that reference is section 11.1 of Tailless Aircraft..., about the middle of page 442 in my edition. Please spare me the 'applicablity of lifting-line theory' argument and so on, because I'm not going to go there with you again. Even if Dr. Nickel's calculation is off by a factor of two [and it's not] that's still a huge penalty for a sailplane.)

    So. Now that we've got that matter straightened out, I hope we can get beyond this and back to the actual thread at-hand. That other 99% of my post, regarding stability and trim was quite correct and you know it. And relevant to the discussion, too. Why you chose to snipe at me for one sentence completely peripheral to the topic when you could've simply suggested a correction with some measure of civility is beyond my understanding. That you keep doing it is becoming more than annoying.
     
    Last edited: Dec 4, 2009
  17. Dec 4, 2009 #17

    Norman

    Norman

    Norman

    Well-Known Member

    Joined:
    Nov 29, 2003
    Messages:
    2,880
    Likes Received:
    918
    Location:
    Grand Junction, Colorado
  18. Dec 4, 2009 #18

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,739
    Likes Received:
    5,420
    Location:
    Orange County, California
    Good Lord, is that what this is all about? You've been carrying a grudge about that for all this time?

    It makes it all the stranger since that's hardly a good example of whatever point you're trying to make. I've got Orion's opinion to back me up on that one, so I'm guessing it's pretty close to the mark, overall. So I had it backwards about what the Hortens sacrificed to get the performance they did. Did it really justify three years of sniping like this? How about you just point me at a good reference instead, eh? I wish I could find the reference that showed the Hortens using relaxed stability themselves, except for training. I really wish I could. Not to justify my position, but find out where the confusion lies so that we can move beyond this. I remember it. I can see it in my memory. Just can't remember where I saw it, three years ago.

    I just don't get it, Norman. I've been advocating flying wings and tailless aircraft as a design option as hard as you have. I don't think they're a panacea, but they have their place and purpose. But simply because I don't share your ideology about the Hortens, BSLD, and so on, you keep trying to snipe me down. You're so good with airfoil analysis and have so much historical knowledge about flying wings that it absolutely boggles my mind that you can't seem to stand an objective look at the problems of tailless flight. Baffles me completely. The Hortens made major progress for the time, and made some airplanes that flew very competitively by the standard of their day. But that day was nearly seventy years ago, and our understanding has moved on. That their methods and theories may not be "perfect" by modern standards isn't a shot against them. They did the best they could at the time, and had their flashes of genius and conceptual blind spots just like any other designer. We can do better now, using real-world numbers, proven methods, and learning from the few mistakes they did make. We understand more, and the problems better. The designers of the SB-13 proved that.

    Can we please get back on topic? If you want to continue making shots at me, have the decency and courtesy to do it in a PM, will you? I have no problem admitting a mistake publically, even if it's pointed out to me privately.
     
    Last edited: Dec 4, 2009
  19. Dec 4, 2009 #19

    Autodidact

    Autodidact

    Autodidact

    Well-Known Member

    Joined:
    Oct 21, 2009
    Messages:
    4,513
    Likes Received:
    799
    Location:
    Oklahoma
    Cool! Then the vining cucurbits have done some experimenting along with the trees :)! I assumed it was a tree because it is so similar to the seeds I would see twirling down from the trees when I was a little kid. I'm told that those were maple seeds, but I'll have to go and look.
     
  20. Dec 4, 2009 #20

    Autodidact

    Autodidact

    Autodidact

    Well-Known Member

    Joined:
    Oct 21, 2009
    Messages:
    4,513
    Likes Received:
    799
    Location:
    Oklahoma
    Cool stuff:
     

    Attached Files:

Share This Page



arrow_white