Airfoil Design - How much more is possible?

[infopimp]

...an unanswerable question, but being still fairly new to this hobby and being an "aspiring armchair aerodynamics buff" I wonder all the time...

How much more evolution do you thnk we will see in airfoils in the future?

Have we pretty much maxed out on what is possible with our current materials and techniques?


I'd think that scientists can run randomizations (or hybridizations) and then crank out polars by the millions for everything from a brick, to a fractal textured object, to... a thin sheet of paper by now... haven't they done this? Is this going on right now?

Thanks for putting up with my silly inquiry.

Basically, I'm just wondering how active of a field of research this is... is it mature with few new improvements happening, or still moving ahead... especially as it relates to our hobby scale.

-steve

/who thinks maybe we need shark skinned (or feathered?) planes...

[vintage1]

Almost none.

Especially at our Reynolds numbers.

The birds had it sorted millions of years ago.

[macboffin]

After over a hundred years of folks designing and flying various contraptions, some very efficient, some less so, aerodynamics is pretty well understood.
Due especially to great interest sponsored by DARPA and NASA in the last decade or so in very small surveillance UAVs there has been a lot of progress and innovation not only in miniaturisation of electronics, autopilots, servos etc, but also very small aircraft, not only fixed wing and choppers, but also ornithopter types, able to be used for flying around inside buildings by a remote pilot to check for hostiles, explosive devices etc.Being so small, it's relatively quick, cheap and easy to check out new ideas for these craft.
Nothing very startling has emerged from all this activity.Unless you count such things as a Cox .01 with a tuned pipe for powering a six inch span UAV!
Birds are stuck with feathers, and evolution has gone a long way with them, although they are not the ideal surface for low speed flight. The best "natural" flyers are bats, and dragonflies. I have seen two metre span fruit bats soaring in thermals in India, and they do pretty well.
Loops and rolls are out, though!
Perforating wing surfaces with tiny holes connected to a suction device can reduce drag, it's been done, but the problem is stopping the tiny holes getting clogged up with dust, bugs etc; theoretically good, and work in practise, but not for long! Sharks are low drag, but in water. Analogues of shark skin have been tried on aircraft ; manufacturing such surfaces is very expensive, the improvements in performance are very low, of the order of 1 or 2 per cent drag reduction, and as with all rough surfaces, difficult to keep clean in service.Various kinds of stepped,rippled and other surfaces have been tried many times over the years. You see any in service?
Advances are likely to come from lighter/stronger materials rather than the shapes of those surfaces. But feel free to experiment!

[Batmanwpg]

I don't think it has stopped at all. Things can develop rapidly for a while such as in the 80's when people found out how to use composites to build stronger wings. This lead to the development of some thinner low drag airfoils, which where impossible with out composites. Before that period and now in the present designers look for improvements in the order of 1-2% and get real happy if they can achieve it. Right now we just aren't in a boom phase of any big new developments.

[BMatthews]

Well, it's my own impression from seeing various models fly both of my own and other folk's that the differences are enough to win in a contest but not of night and day amounts.

Thanks to a host of recent designers and Xfoil there's a large number of "optimized" airfoils out there to do various jobs. Will you make them any better at the job they are meant for? Perhaps a very few% better but then you need to be able to build them to the required accuracy to see the gain or all is lost. Really if you look at the field of model airfoils it's mostly about selecting a camber and thickness to suit the job and only then do you select a shape to help optimise the performance of that job.

If you're expecting some sort of airfoil equivalent of Perpetual Motion or Cold Fusion then no, it isn't going to happen. Can we expect a 4% gain here and a 7% gain there? Likely in some cases. But then you need to build it that accurately.

What I'd like to see is an airfoil shape along the lines of Dr Drela's AG34 that was designed around simple real world building realities but perhaps go even more so towards reducing the amount of sheeting to achieve a more open frame structure. Something that uses a minimal top sheeting and perhaps some extra turbulator spars to support the covering. Something that is relatively easy to build while providing more of a glow through the wing covering while still being decently efficient at flying a wide speed range.

[ciurpita]

it seems that the research the likes of donovan, selig and drela has contributed greatly to the UAVs of today, besides being supported by and benefiting the modeling community.

but i would think that software tools such as xfoil will become just a sub-components of a more comprehensive aircraft analysis tool that iteratively drives these sub-components to determine the optimal airfoil at each span position depending on the task(s) needed. in other words, imagine specifying the task, and letting a program run overnight to design the aircraft for that task, or the best compromise for a multitude of tasks.

given the blended wing stuff we're beginning to see, what if the overall airfoil shape could be changed by more than just controlling disjoint trailing edges? what if its shape (or at least portions, such as the trailing or leading edge) could be controlled dynamically, using more sophisticated software to adapt the airfoil shape during flight given the current flight conditions? what if instead of providing a control input for a control surface, the input is the desired change in lift on some segment of an airsurface, and software make the appropriate change? (but this software must be reliable). see "computers take flight" and "The Story of Self-Repairing Flight Control Systems" by James E Tomalylko at
http://www.nasa.gov/centers/dryden/history/Publications/index.html

[Ralph Walton]

I'm alyways blown away buy this type of video. W've got along way to go to catch the birds. There's lots more scope for improvement.

http://animal.discovery.com/convergence/spyonthewild/birdtech/birdtech.html

I suspect inflight camber change (not just by drooping the ailerons) and wing area/shape changes to be come common

[macboffin]

it seems that the research the likes of donovan, selig and drela has contributed greatly to the UAVs of today, besides being supported by and benefiting the modeling community.

but i would think that software tools such as xfoil will become just a sub-components of a more comprehensive aircraft analysis tool that iteratively drives these sub-components to determine the optimal airfoil at each span position depending on the task(s) needed. in other words, imagine specifying the task, and letting a program run overnight to design the aircraft for that task, or the best compromise for a multitude of tasks.

given the blended wing stuff we're beginning to see, what if the overall airfoil shape could be changed by more than just controlling disjoint trailing edges? what if its shape (or at least portions, such as the trailing or leading edge) could be controlled dynamically, using more sophisticated software to adapt the airfoil shape during flight given the current flight conditions? what if instead of providing a control input for a control surface, the input is the desired change in lift on some segment of an airsurface, and software make the appropriate change? (but this software must be reliable). see "computers take flight" and "The Story of Self-Repairing Flight Control Systems" by James E Tomalylko at
http://www.nasa.gov/centers/dryden/history/Publications/index.html Darpa has funded some work in this area, with "morphing" wing shapes in plan view as well as flexing wings, rather than seperate control surfaces. The major problem to implementation on production aircraft so far appears to be maintaining function in actual service without excessive maintainance down-time. Improved materials may improve matters in the future. Such systems are unlikely to assist models however ; being even more difficult to miniaturise than just to make at, say fighter size. If they are adopted, it will probably be for UCAVs, since flesh and bone pilots react too slowly to benefit, and in any case have G and size limitations.(Pilots are a liability ; they require sitting space, a good view around,an ejector seat, controls, instuments etc etc,taking up space and adding a lot of weight.)

[mlbco]

Steve,

I don't think the design space for 2-D flow over low Re airfoils has been fully explored. Codes like X-Foil are based on stringent assumptions of how the boundary layer can form and this restricts the design space of feasible airfoils to a small subset of possible shapes. For example, a design with a notch in the upper surface that captures the laminar separation bubble can not be properly analyzed in X-Foil, yet this type of airfoil works very well at low Re. There are many other similar "tricks" that remain unexplored simply because the analysis codes can't handle these cases. This leaves room for a large number of new airfoil design ideas yet to be explored and properly analyzed.

Steve Morris

[infopimp]

Steve... in an interview posted in the Slope forum recently I read the following with interest:

...We reduced wing area in favor of aspect ratio, and I plotted the airfoil, an HQ 0.8/8 with 0.8% camber and 8% thickness.

...

We thought that was ridiculous and it will never fly well. Because of this, we built the plane in a quick and dirty kind of way. Jim used cheap mylars and applied just a tad too much vacuum during the bagging process. As a result, the wing had barely noticeable dimples all over it. But that thing was a screamer!

Read it: Reed Sherman interviewing sailplane personality Dieter Mahlein (aka ShredAir) (http://www.jartworld.com/jartworld_pgs/inter_mahlein.html)
Discuss it: http://www.rcgroups.com/forums/showthread.php?t=765860

To me it is fascinating to think you might be able to add a tweak (e.g., turbulators, winglets) or otherwise innovate on design... and perhaps find something better a quantum leap beyond expectations... living outside the scope of the current design tools.

Thanks for that insights, and please keep them coming to those just finding this thread.

[Tom Harper]

The availability of low cost instrumentation and low cost UAV systems brings serious research within the reach of hobbyists. A fully instrumented model could be flown with various wings under identical conditions. The results could be recorded and compared.

That is now possible on a modest budget.

Tom

[Mike the Snake]

I think that dynamic soaring may be open to new foil shapes.

The aerodynamic environment of DS is very different. There seems to be an aerodynamic "wall" around 300mph.

I think maybe laminar flow foils, or something not yet discovered will allow much greater speeds.

Foils are examined at varying AOA's, but DS foils may not need to operate efficiently at high AOA's, they may just need a new shape?

In DS, great speeds have been obtained both by flying tight circuits with low WL's, and flying huge circuits with high WL's.

Both work, but in the huge circuits, the foils are Not operating at high AOA's, so maybe a high wing loading and a laminar flow foil that doesn't dig well, but allows greater top speed will be the key to speeds well above 300?

The reynolds numbers at 300mph are not low, and compression begins to come into play as well.

[Ralph Walton]

The huge G's in DS must make it difficult for wings to perform well. I ran some theoretical numbers on a large plane at 1G at 300 MPH. Cl (coefficient of lift) was .02. At 20G Cl was .4. At 40G -.8Cl

This throws a whole new set of variable into the "things to optimise on an airfoil"
basket.


I believe the DS feild is wide open for improved foils

[BMatthews]

Not really, The high speed airfoils are operating at a higher Reynolds number during the high speed phases of the flight so there's some other airfoil options that are thicker but since we can achieve our structural needs with thinner and faster options there's not much to be gained on that front.

Really it's all about ensuring that the airfoil's lower drag operating area spans the lift coefficients extremes that you plan on encountering in a typical DS flight.

If we're looking for some form of Holy Grail airfoil that suddenly works twice as well it ain't going to happen. All the design software of the last decade has been available long enough for folks to design legion's of airfoil families. Profili II is full of them. It's amazing that they all seem to end up looking relatively the same. That alone should suggest something.

We've probably got room for another 5% or so in terms of optimizing the airfoils to a particular job. But with that optimization comes the need for extreme tolerances in the fidelity of the shape to control the separation bubbles and pressure ramp shapes. Also with that optimization comes a reduced range of optimum operational lift coefficients.

If someone can develop and actual shape changing wing that alters it's shape for controlling the lift as a flight control then yes, we'd be looking at some big advantages. But it's not just about adding a rubber skin. The shapes that the airfoil morphs into would need to be carefully analyzed and optimised over a very wide range of shapes and camber values so that the whole range of shape changing is fully optimized. Either that or some sort of airflow measuring devices need to be incorporated to provide a closed system of feedback to monitor the airflow and alter the shape to optimize each shape each moment.

[MarkusN]

Either that or some sort of airflow measuring devices need to be incorporated to provide a closed system of feedback to monitor the airflow and alter the shape to optimize each shape each moment.
Sounds like something feathers might be mighty good at.

[Brandano]

Either that or some sort of airflow measuring devices need to be incorporated to provide a closed system of feedback to monitor the airflow and alter the shape to optimize each shape each moment.
That could be achieved with a series of pitot probes, or just a pitot probe, some math modeling and a lookup table. As for the profile altering wing I guess that some of it could be achieved with a clever wing rib design and muscle wires, or pressure driven actuators.

[BMatthews]

That could be achieved with a series of pitot probes, or just a pitot probe, some math modeling and a lookup table. As for the profile altering wing I guess that some of it could be achieved with a clever wing rib design and muscle wires, or pressure driven actuators.

Ah, but remember that a series of probes can introduce their own turbulence. Not to say that it can't be done in some manner.

The wing's shape would need to be controlled far more than possible with a simple arrangement of muscle wires or such. Although with a complex enough network of them all working in harmony and controlled by a computer I supose it may work. The trick being to control the shape to within a few thousandths of an inch at any given location and moment. But how you could introduce spars and wing borne fuel tanks into this sort of morphing wing I'm not sure.

The low speed tunnel research done by Michel Selig and reported in Soartech 8 from some years back made note of the fact that on some of the more critical airfoils that even a very small discontinuity could make all the difference in the airfoil producing the predicted performance and just so-so performance. So any sort of flexing airfoil would require some very carefull control for shaping.

[macboffin]

Saw a TV program two nights ago, (BBC) two skydivers with cams on helmets, one a falconer.Jumped from 10k feet height from hot-air baloon, Peregrine falcon released from baloon same time. They soon reached terminal velocity about 120 moh, falcon didn't just catch them up, it was circling and stunting around them, at points must have been a lot quicker! Feathers and a morphing wing shape sounds like the answer!

[infopimp]

I agree, feathers give you variable camber across the entire airfoil, and the feathers are embedded in a sensor matrix to detect what is going on across the wing. :)

[BMatthews]

Saw a TV program two nights ago, (BBC) two skydivers with cams on helmets, one a falconer.Jumped from 10k feet height from hot-air baloon, Peregrine falcon released from baloon same time. They soon reached terminal velocity about 120 moh, falcon didn't just catch them up, it was circling and stunting around them, at points must have been a lot quicker! Feathers and a morphing wing shape sounds like the answer!

I'll bet that the falcon was having a blast chasing his master... :D

I saw a hawk doing something strange at my power field one day. It was making all sorts of funny calls and then dove. I just figured it was hunting. But then it pulled up and over trying to do a LOOP! ! ! ! But he didn't quite have enough speed the first time and fell out of the top with a helluva squawk. Circled in a thermal to get the height back and did it again but this time with a longer dive. Up and over and down the back side he went and then got back in the thermal with a couple of rather joyous sounding CAW's and away he went. I guess he figured he'd shown us a thing or two.... :D

[Ralph Walton]

I regularly watch seagulls around where I fly. Its very interesting to watch the number of different ways they can fold and tuck their wings for different flight conditions. I'm amazed how fast they can go if they want to. They use varying amounts of sweep combined with span alterations. What a masterpiece of engineering that would be if you could copy it.

They used to be very curious of our slopers and would fly in formation (above and behind) for minutes at a time. We fly so regularly that they got bored with us, but we noticed a curious trend. The younger ones would come over and occasionally started to fly erratically.

One occasion I noticed one was doing what I would describe as "half rolls" rapidly side to side. Left wing would be vertical then right half a dozen times in a row. It was very unbirdlike. They were copying us. Coincidence?

Now I'm waiting for them to try flying inverted. Then I'll Know for shure.

[NX-687]

Evolution has only optimised them for more or less level flight so to speak

Negative G's not required

I guess with fixed wings we are still stuck way back :(

What year do you estimate :) a sensor matrix will be in wings

[BMatthews]

......One occasion I noticed one was doing what I would describe as "half rolls" rapidly side to side. Left wing would be vertical then right half a dozen times in a row. It was very unbirdlike. They were copying us. Coincidence?.....

I've seen some birds do that particular action to act as a slowing down maneuver to avoid colliding with another bird. So I'm not sure they learned it from us.

[BarrelRoll]

Evolution has only optimised them for more or less level flight so to speak

Negative G's not required

I guess with fixed wings we are still stuck way back :(

What year do you estimate :) a sensor matrix will be in wings

I guess to mock the birds we would actually need:
- sensor matrices covering wings and stabilizers
- ultra-fast servo matrices in wings and stabilizers
- very powerful control processor

In 20 years, maybe? :)

[NX-687]

By servo you mean a wing full of muscle wires, connected to a controller responding to sensors , sounds easy :)

[BMatthews]

By servo you mean a wing full of muscle wires, connected to a controller responding to sensors , sounds easy :)

I don't know if I'd call it easy. The complex layup of the muscle wires and the pivoting or flexing mini beams they would need to interplay with would rival one of those 10,000 piece jigsaw puzzles done in one color only for complexity of building.

[infopimp]

Well then, this does seem on topic...

Watch the bird trying to kill my sloper - quite a fun video:

http://www.imageseek.com/TMC/gallery/albums/tropical_storm_edgerc/tropical_storm_bird1.jpg

Video - When Birds Attack - Tropical Storm and Acacia and... Hawk? (http://www.rcgroups.com/forums/showthread.php?t=706665)

Right-click, download to disk... then watch.

Small: http://www.imageseek.com/TMC/20070629-ComHill.mov 11 MB

Medium: http://www.imageseek.com/TMC/20070629-ComHill_medium.mov 40 MB

Large: http://www.imageseek.com/TMC/20070629-ComHill_high.mov 110 MB

[NX-687]

I don't know if I'd call it easy. The complex layup of the muscle wires and the pivoting or flexing mini beams they would need to interplay with would rival one of those 10,000 piece jigsaw puzzles done in one color only for complexity of building.

But you know how modelling can simplify things such as a depron 6mm wing with a muscle wire on one side , responding to whether the model is right way up or not , and then to model attitude
I always think the simple way is overlooked , Eg, first model Harrier flew in 93 , no one can seem to crack it even now in 07

[Ralph Walton]

Great video

"did try taping a bit of bread on the nose once thinking it'd start a chase, nothing [QUOTE]"

We're well off topic here, but the same thought had occurred to me ;) . Though, I was thinking of putting a photograph of a piece of bread on my plane for aerodynamic reasons :D

Getting more on topic; What impresses me about a bird like a common seagull, is they only weigh a few hundred grams, but if required, they can get the pace of a 60" RG15 foamie that weighs 3 or 4 times as much.

They have an undercambered wing, which you wouldn't think is condusive to high speed, but I figure as they swing the outer portion back and the inner portion foreward, they not only change their aspect ratio, but also the % undercamber related to effective chord.

Bird wing profiles have been copied and tried with limited success, but have they been tried using the effective chord of a swept wing?

Has anyone tried real birds wings (obviously dead ones) in a wind tunnel?

[MarkusN]

Has anyone tried real birds wings (obviously dead ones) in a wind tunnel?Early airfoil catalogs contain eagle and vulture profiles with a step on the underside (for the bone), so I assume these have been tested in wind tunnels also.

[NX-687]

To me bird wing is not possible unless held in correct attitude by live bird , also its the adjustments that are made , its not fixed as you know :)

[Mike the Snake]

A little bailing wire, some duct tape, no prob holding a bird steady! LOL

Birds are great and all, (I'm an avid bird watcher) but can't be compared to models (application of materials) as the structures are so different.
It's really all about wing loading. Gulls and hawks (all birds) can go from slow to fast because they can change their wing loadings by pulling their wings in tight. Wing area is halved or quartered, wing loading goes way up. Bird's ability to alter their wings and change their wing loadings is what we try to simulate on full scale jet planes. Look at a 747 wing, for landings and take-offs, the slats and flaps probably more than double the area of the wings. They also seriously change the shape of the foil to undercambered for more lift and drag just as birds do. Models and full scale have camber changing.

Regardless of foil, if you increase the wing loading (by adding weight, or reducing wing area) anything will fly fast/er.

I still wonder though about model airfoils above 300mph to 400mph, where compression and G-forces are so great. For DS, my thinking is there is some panacea foil shape that we haven't found yet that will release the planes from the drag wall that exists now around 300 and above.

I wonder if there's any information to be gained by looking at full scale helicopter rotor blade foils?

[Flyingwingbat1]

I doubt a 747 doubles its wing area with everything hanging, but I understand your point. Laminar-flow airfoils seem like a good idea for the high-speed DS machines, since their REs are approaching full-scale territory. With that said, a full-scale airfoil may still have to be tweaked to account for that RE difference. The wingloadings they experience in the circuit must be tremendous, certainly within the range of some full-scale planes in level flight.

I did my own experiments with serrated (LE) wings; they work great for lowering the stall speed at ultra-low RE, but the L/D suffers. I don't know how serrated TE's would compare.

[Mike the Snake]

For super low RE's the indoor pattern guys flying the flat winged pattern planes are using paddles, serrated LE's and fences, vortex generators, crazy little boomerang wings over the canopy areas, cool stuff.

[Ralph Walton]

I've been looking for Hq (horstman quast) and delft DU foils. in a .dat or.cor etc format to run some numbers thru profilli. Higher speed DS planes may be able to benefit from some of these foils.

Anyone know where I can get hold of them?

[flieslikeabeagle]

Back in 2004 I read about some research results on humpback whale flippers which seemed very promising for some model aircraft: the whale flippers have demonstrated ability to resist stall at high angles of attack and have low drag at these high angles compared to more conventional airfoils, and all this at Re around 100,000, i.e. in the range where many small RC parkfliers fly.

So I started a thread on this, which quickly devolved into arguments about how perfect internal combustion engines are ( :rolleyes: ). Three years later, I still haven't seen anyone try a 3D model aircraft with wings based on the humpback whale flipper design. (It would be a pain to carve from foam, but looks easy enough to mould).

Here's a link to the (short-lived) thread I started on this back in 2004: http://www.rcgroups.com/forums/showthread.php?t=259792

-Flieslikeabeagle

[Flyingwingbat1]

That's an interesting article. From what I can gather, the LE bumps act as turbulators of sorts to minimize large-scale flow separation. Serrated LE's on plane wings act the same way. It seems a delta shape, no matter the RE, will generate attached counterrotating vortices at a sufficiently high AOA. It woks all the way from Concorde down to 3D planes (and even smaller, in my experience)

[flieslikeabeagle]

From the article it sounded as though the bumpy leading edges on the whale's flippers didn't cause the sort of additional drag that serrated leading edges on foam wings do.

Considering the huge difference in drag between a square cross-section fuselage and a circular cross section fuselage of the same cross sectional area, and the general idea that sharp edges and small included angle obstructions are draggy, perhaps this is simply because the bumps on the whale flipper are a sort of circularly symmetrical serrated leading edge?

-Flieslikeabeagle

[BMatthews]

So.. we should all grow barnacles on our leading edges?... our WINGS that is....





Sorry, I couldn't resist.... :D

Oddly enough the warts the humpbacked whales have are being replicated by turbulators and serrations in the 3D wings these days.

But I can't help but think that the bit about the Reynolds numbers is bushwa. Water is incredably thick compared to air and the last time I measured a whale the flippers were a LOT wider than any small model airplane wing. All of that tends to make me think that the Re # is missing at least 2 zeros on the end of that 100K and it should be more like 10M.

[flieslikeabeagle]

But I can't help but think that the bit about the Reynolds numbers is bushwa.

Hmm. Lessee. Kinematic viscosity of water is about 13 times lower than air, and I found this via a Google search:

Humpback whales normally swim 3-9 mph (4.8-14 kph), but can go up to 15-16.5 mph (24-26.5 kph) in bursts when in danger. Feeding speeds are slower, about 1.2-3.5 mph.

The article also mentioned the Humpback uses those flippers to maneuver during feeding, IIRC - so at a speed of 1.2 - 3.5 mph.

Putting those two together, a model aircraft wing the same chord as a whale flipper would need to be moving at roughly 2.5 x 13 mph, or 32.5 mph. If the whale flipper were twice as wide as the RC aircraft wing, the aircraft would need to do around 65 mph. Those numbers don't seem too far off what I see at the park and flying field.

I agree there may be some minor differences, but certainly the RC plane and whale flipper seem to be experiencing Re values of the same order of magnitude (i.e. varying by a factor of no more than the square root of 10, or roughly 3). I really don't see how there could be a two order of magnitude difference (the two missing zeros you suggest).

-Flieslikeabeagle

[kcaldwel]

I've been looking for Hq (horstman quast) and delft DU foils. in a .dat or.cor etc format to run some numbers thru profilli. Higher speed DS planes may be able to benefit from some of these foils.

Anyone know where I can get hold of them?
I think most of the HQ airfoils are here, and a couple of Delft ones:

http://www.ae.uiuc.edu/m-selig/ads/coord_database.html

You may need to try Delft itself for any others.

Kevin

[Ralph Walton]

Thanks Kevin , but those are Helmut Quaback HQ foils, not Horstman Quast HQ foils.

Confusing isn't it?

[kcaldwel]

Thanks Kevin , but those are Helmut Quaback HQ foils, not Horstman Quast HQ foils.

Confusing isn't it?

Oops, sorry Ralph! The Quast airfoil data seems to be kept proprietary. Here is how to get the Delft airfoil data (maybe just the wind turbine stuff?):

http://www.tudelft.nl/live/pagina.jsp?id=dd703747-9751-413a-9747-85d7d7f21ca2&lang=en

Kevin