Discussion in 'Aircraft Design / Aerodynamics / New Technology' started by Hot Wings, May 18, 2017.
-we dont like the tails for flying!
Waldo Waterman was a pretty interesting guy, and they should make a movie about him for which they would need some cool replicas! The Aerobile was really one of the better flying cars, IMO.
PS, if we can get everyone who wants to invent the flying car to build their requisite 5 copies, then there should be plenty of them!
The SWIFT airfoil is distinctly Liebeck-esque. Somebody here gave coordinates based on the available pdf. I cleaned it up a bit and it looked pretty interesting, XFLR5 did give it a pretty sharp stall, as well as lots of lift. If anyone has accurate data, I'd love it!
View attachment swift-mod-smoothed.txt
XFOIL and XFLR5 aren't terribly difficult to use, are they? Are they useful - especially XFLR5 - for actual aircraft design?
My opinion is, yes they are useful.
XFLR5 has a bit of learning curve, though. And it does have problems depending on what system you have to run it on. Mine keeps crashing. The author can't figure it out. I have been thinking about compiling it myself and see if that works better. I don't
XFLR5 is nice for designing flying wings, particularly pure flying wings. Because you don't have to worry about the fuselage skewing the results. There is a warning about using a fuselage when doing stability analysis.
I used it to examine modifications on my (currently on hold) build, at it agreed quite well with manual calculations. Also agrees with theory if you don't do anything too unusual.
XFOIL is a command line program for airfoil design and analysis. XFLR5 provides a fairly easy GUI to Xfoil plus the ability to perform stability analysis. The documentation is pretty sparse but there are a lot of tutorials on youtube. Here's one of the better playlists.
Thanks for the link! It appears that the "It's for models only" statement is mainly a legal strategy to keep people from saying "Hey man, I designed an airplane with your program and it crashed and killed my brother-in-law and my neighbors prize show cow, and now you're gonna pay" type things.
Drela is a hand launch RC glider nut. So a lot of this stuff was originally developed for optimizing low Reynolds number designs starting with airfoils then leading to whole airframes. X foil goes a long way back to before Windows in the DOS days. Dr. Selig was running the low turbulence wind tunnel at Princeton at the time. It was an interesting time in RC glider advancement. Anyone who thinks that RC gliders are toys built by holding the thumb up don't understand the way people were trying to figure out how to optimize out separation bubbles on tiny airfoil sections. Xfoil was written specifically for that task. So they never really considered that people would try to design anything over 100k Reynolds number with it. Now there are people doing 500MPH dynamic soaring models that can withstand ridiculous G forces and are creeping up in Re values even on tiny sections.
Anyway, just some background. There's always a reason. I don't think it was lawyers unless it was MIT's paranoia due to affiliation. I think it was just one of those we can't really guarantee the results cause we haven't tested and verified in that regime thing.
Ok. The ability to analyze in the higher Re ranges was a concern of mine, but in one of the tutorials I watched the guy punched in an Re of 5,000,000 so I'm assuming that the panel method code used was valid for that Re range - I noticed that there is also a Mach number variable that can be used as well. It looks like a fairly comprehensive virtual wind tunnel to me...
Don't know how accurate it is, but I have run it as high as Re=20,000,000 with no problems. Results seem ok.
There is no difference in the math and I am sure it works fine. What I was trying to say is that a lot of the initial work of Xfoil was actually verified in the wind tunnel by a bunch of grad student enthusiasts. I think most of the stuff they used was at model glider speeds and not over 12" in chord so very low Re. The wind tunnel at Princeton could only be run at night on clear calm nights because if the wind changed during the runs it could throw it off into orbit. So a lot of the tweaking and correlation was sorted at that scale. I think that is why there is no real claim that there has ever been any real world GA correlation studying done. I think in general it is a bit less critical for large Re in this sort of code. Generally larger stuff is better behaved than smaller stuff when looking at turbulent separation. Xfoil has some pretty sophisticated boundary layer analysis built in.
Thanks everyone for the comments, this has been very enlightening.
Well, your bottom-most concept reminds me of something.... (http://www.homebuiltairplanes.com/forums/album.php?albumid=18&attachmentid)
If you fancy flying around holding constant back-pressure on the stick, by all means, use a symmetrical airfoil. Reflex in flying-plank designs is there to make the airplane trim automatically at the design flight speed, whatever the designer has chosen that to be. You then deflect the pitch controls to trim for some other speed. If you use a symmetrical airfoil, you have to deflect the pitch control surfaces to maintain level flight in all flight conditions where the airplane isn't pointed directly nose-down at the ground.
Ideally, you choose a reflexed airfoil that produces the correct Cm for your airplane to be trimmed at the design airspeed, plus has the other characteristics that you want in an airfoil for your particular design - L/D max at the Cl corresponding to you cruise airspeed, stall characteristics, Clmax corresponding to that needed for your stall speed, etc.
I had thought about this some; looking at the section view of a reflexed airfoil, the reflexed part is really a fixed (and upwardly deflected) flap. As soon as you cut it out and hinge it, it will need a trim tab to keep its upward deflection. Some of the reflexed airfoils have very good drag characteristics, but I'm currently favoring the Mitchell approach to flying wing design; the B-10 uses a NACA 23015 airfoil with separate "ailevators" which look an awful lot like Junkers flaps to me - this approach has good control, good stall, and good drag characteristics and the drag and lift numbers for the 23015 and the 0015 airfoils are only slightly different and the 23xxx series have a supposedly sharper stall, but there are many other factors affecting that as well.
Concerning Flea type aircraft, many of them use the 23xxx series sections without difficulty and these have very similar drag characteristics to the 00xx series symmetrical sections.
I've seen no numbers to show that the difference between 23xxx and 00xx airfoils will be anything more than negligible for Mitchell type wings and Fleas. Other wings I think would be better off with a good reflex airfoil - I think the airfoil selection is heavily dependent on the wing's configuration w/respect to planform, control surfaces etc. That's why I was inquiring about XFOIL and XFLR5, other than stall behavior they might be very useful in selecting the airfoil for a specific flying wing configuration.
PS, the little plank I drew earlier will have a relfexed wing, like the Marske XM-1D...
I'm probably the wrong guy to ask about this, or for suggestions. I'm in the camp that says you can't choose the "right" airfoil until you know what the characteristics of that airfoil need to be for your specific airplane design. I wouldn't even be looking at families at this exceptionally early point in the design process.
On a conventional airplane I agree with Topaz, but on a small minimalist flying wing like this I believe that you can pick a known good airfoil that works in the Reynolds numbers and speed ranges you want to be in, and build the "airplane" around that, because the wing is the whole airplane. Everything else gets adjusted to the airfoil, instead of the airfoil being adjusted to suit the other parts.
Mind you I have no formal expertise on which to base that statement. But it passes the common sense test for me.
For a slow LSA or near-ultralight flight envelope, if you took a Fauvel or Hepperle airfoil that is either designed to work well in this environment or has demonstrated to work well in the environment... and built a small man-carrying aircraft around it... I would bet any one of you a steak dinner that the airplane would be "in the ballpark" with a good chance of being reasonably good.
If we are getting into faster speeds, heavier weights, or wider CG ranges because of more seats, then my backwoods redneck simpleton way of looking at the design would become less valid. And there would have to be a more dedicated mathematics session, with smarter people at the table, to select an airfoil.
Separate names with a comma.