Design considerations for 3D aircraft airfoils

[JCaste]

I'm working on a 3d plane design, ~70cm wingspan, aspect ratio around 4 (i've included a small pic at the end). :) I've been reading some foil theory, but that is not much help for this type of flight, since most sources consider the foil to operate at Re numbers a lot higher than this one does, and below that, there's the reigns of darkness. Basically, those planes operate in completely different conditions than most: high angles of attack, very slow or still flow, very low alt... :rolleyes:

I've tried to sum up what things must be considered to make a succesful 3D profile, to help me see the problem with better "perspective".

I see there are two tendencies. In foamies and similar sized planes, flat plates are used, mainly for their simplicity but also because of their nice behaviour in small models.
In balsa planes, mostly the models .40 up, where the wing is a wood structure, quite thick profiles are used. While those thicker profiles should have an overall broader operational enviroment, I'm not sure this is felt when you fly at walking speed. The reason for their thickness, as I see it, is because it's easier to make light and stiff than a narrower wing (good thing when you're making those crazy maneuvers). That could also have a comercial interes; you see a very broad wing, but very skinny ribs and structure, and gives you a sensation of "lightness".

Another thing to consider is the role of the powerplant. By definition, it must be able to - at least - lift the weight of the ensemble. Ideally, we're looking at 1.3-2 power/weight ratio, where the model thus can hover and rocket into the sky. Beyond a certain speed and angle, wings soon yield very little lift, most of it being provided by the powerplant. But even then, the prop is moving a huge mass of air over the central section of the wing. This turbulent flow may attach to the wing and generate some lift, despite the plane is hovering, i. e.

With this in mind, I'd look for a turbulent flow symmetrical airfoil (If I wanted simplicty, I'd go for the flat plate), in a low aspect ratio wing, with moderately thick root and diminishing thickness spanwise. This is MHO, so I'd really like to hear what people with more experience/knowledge can add or change to help me understand the subject. ;)

Thanks in advance!

http://image.rcuniverse.com/gallery/galleryimages/lg-44470.jpg

[Ollie]

Any old airfoil with zero mean camber will do! Thrust to weight ratio makes airfoil's wing lift and drag coefficients unimportant! The 3D aircraft flies on the prop, not on the wing.

[kristianb]

the higher the wing loading, the more critical profile nose radius get
(higher wing loading=>larger profile nose radius=>thicker profile)
profile itself is for this kind of air craft without any major influence, if it looks ok, it will work! ;)
vbr KristianB

[BMatthews]

One reason the flat airfoil works on the foamies is because it stalls quick and deep so the model brakes to a stop right away. For a lot of 3D stuff this is what you want.

A thicker "proper" airfoil tends to delay and try to avoid this quick and deep stall effect so when you pitch the model hard it tries to do a loop and bite its own tail. The thick airfoil fun fly models used this trait to compete in the old fun fly events like the most loops in so many seconds and the most rolls in so many seconds and other similar events where a stalled wing would slow the model and reduce the score.

I've noticed that on some of the more modern TOC style models that they are shifing to symetrical airfoils of lower % thickness these days. Some thickness is good as it provides more depth of spar for strength. But I would say that is you're after good 3D style flying you'll want to keep the thickness to around 8 to 10% so it offers a reasonable depth but not so much thickness that it strongly delays the stall.

Playing with the leading edge radius is also another way to control when the airfoil stalls and how fast and deeply it stalls. A sharper airfoil will stall earlier and more decisively.

If you start with a sharper shape and use a strip of covering or even just packing tape over the leading edge you could then alter the radius until it's what you want by removing the leading edge covering strip and sanding in a more rounded shape until it's where you want it to be. Beyond that just put the high point at around the 27 to 30% mark and use a smooth progresive curve. It'll likely be about as good as any other high priced option that has a real name since with 3D it's all about the prop and size of controls for a large part of it.

[JCaste]

Thanks Bruce, that makes sense. Luckily, with Solidworks is very easy to draw a parabolic curve and adjust its shape to those specs. A whole different thing though, is how accurately that shape translates into the built model...
I know 3D flying is about power, but a good wing design sure helps, despite the wing often is operrating as a wheather vane, or creating lots of drag that are required during some maneuvers.

[Ollie]

Better to use expanding logarithmic curves than parabolic for your airfoil.
http://www.modelresearchlabs.com/pricelist.htm
Expanding Logarithmic Spiral Patterns Makes tapered wings easy, draws any airfoil from 1% to 10% thick and 1" to 20" chord.

[BMatthews]

Ollie, I think you got that link wrong. That's a price list for composite materials.

Another good option for a generic shape that works well but not great is the standard old NACA 00xx series of symetricla airfoils. If you can find some freeware that generates the NACA coordinates then you can make a simple shape.

I've also been known to "cheat" and produce my own TLAR airfoils that use a flat rear portion to help make building the wings easier. The shapes being largely guess work but based on looking at and studying tons of plans over the years. Here's an example.....

[Ollie]

BMatthews,
Go to
http://www.modelresearchlabs.com/pricelist.htm
About 2/3 down the price list, you will find Drawings, Plans, etc. The first item under Drawings, Plans, etc. is Expanding Logarithmic Spiral Patterns Makes tapered wings easy, draws any airfoil from 1% to 10% thick and 1" to 20" chord $5.00.

[coosbaylumber]

Bruce:

On and off, certainly on occasion, I have been CAD drawing up the Pretner Curare' from numerous sources. On one of the plans it calls for main wing airfoil of: NACA0018 and for the Hoz-stab to be NACA0012. However, what is actually drawn on one of the plans looks nothing like a generated one or like yours.

Any idea as to what may be been used which loosely fits the criteria? I generated a 65A018 and it comes closest in appearances to the plan view, but the blunt L.E. like you have shown does not.

Wm.

[BMatthews]

Sorry William, not a clue. Many of those early pattern models probably used their own airfoils based on what worked for the needs of the pattern at the time. High point location and leading edge radius were all understood and used back then to make the model perform as they wanted. Sharper LE and high point back to enhance the snaps and more round and forward to delay the snaps. The thickness being set to add drag and control the downpath speeds. The articles of the day talked about all this stuff but hardly ever mentioned specific airfoil names. When they did it was often as not reffered to as a modified NACA 00'whatever. Modified until it's own mother wouldn't recognise it in most cases.

If you've got the plans and know that it's the same airfoil that Pretner used then I'd scan it it and CAD over it. That's likely as accurate as you're going to get.

And welcome to the Florshiem school of airfoil design.... :D

Ollie, thanks for the correction. I only scanned down for a short ways.

[JCaste]

William, you can either stick to the drawing as Bruce says, or use the newer profiles, if you have the knowledge to know which and how (I don't). :o

Just checked Martin Simons book; there are several symm profiles, some of them resemble yours. As Ollie says, an old airfoil is probably a good candidate since they are made for turbulent flow. I've seen that many of the "structure" wings have thickness bigger than 15%, so I'll pretty much stick to your advice and get mine around 10%. Another common and succesful technique is "cheating" the shape of the trailing edges so they can use flat ailerons instead of shaped ones (as seen in pattern or other performant planes). Another good reason that's making me stick to the TLAR aproach and forget about subtle things for which those pseudoplanes don't care for. :D
In my case though, I may shape the ailerons since they are nearly 1/3 of the chord. Alternatively, I played with the idea of a swing wing mechanism, but several things are taking me that idea. Weight, loss of simplicity and strenght, and sharp flow transitions in the articulation plane are heavy arguments.

BTW Ollie, I can draw spirals with my CAD software, but I don't see how does that look on a profile. :confused:

[Ollie]

Not just any spiral but Expanding Logarithmic Spiral. Read this:
http://www.modelresearchlabs.com/hand_launch_glider_airfoils.htm
Use two spirals, top and bottom, for your airfoil. You can modify the nose radius with a circle tangent to the two logarithmic spirals (top and bottom).

The nice curve size is to simlar to its self. So, one template works for any rib size in the wing taper.

You want the pitch about 1 degree for a flatter curve. If you want the curve fatter, increase the pitch degrees.

http://en.wikipedia.org/wiki/Spira_mira
And
http://www.aiaa.org/content.cfm?pageid=406&gTable=mtgpaper&gID=55408

[coosbaylumber]

A good five to ten years ago, I loaned out one of my magazines. Do not remember what the exact name (Flying Models, Model Airplane News, Model Aviation ?) or date of issue (circa 1980 +/- three years) but it was a huge article containing information about the (then) current trends of competitive international events.

I seem to remember the discussion concerning radio gear and powerplants of which the N.A. (including those from Canada) competitors had an unfair advantage over the Europeans, then how the Europeans (especially the Germans) were altering the spectrum by designing whole new models using flaperons, speed brakes, and new laboratory designed airfoils. It went on and on about many differences between the teams and their tallent, and being critical about how they choose and how they practice.

Does anyone have a recollection of this article? I do remember it told of what the Germans found concerning wings, and airfoils, citing optimum numbers and design criteria then popular.

Wm.

[Ollie]

BMatthews has the main point in this airfoil selection for 3D. It is the stall and its recovering. There are three types of stall:
1. Mushing (partly stalled) with a few degress increase in angle of attack with constant high lift and very incresing high drag.
2. Sudden stalling and sudden recovery at the same angle of attack.
3. Stalling at angle of attack but with recovery at a much lower angle of attack (hysteresis).

Many airfoils are not measured performance beyond the stalling angle of attack and recovery.

[JCaste]

I've tried Profili and that has really helped. using the data Bruce recommends to make a search filter, I still have a lot of profiles that stick to the numbers and dicsard the ones with big LE radius.

The stall we need here, I think, is preferably #2 in Ollie's list, maybe #1. Not too sure, because in some maneuvers you like to stall the wing quick, while in others high lift and drag are also very desirable.

Ease of construction will be the deciding factor, if nothing else helps. :rolleyes:

[ADB2]

Hi,
I am not sure I should revive this thread but since I am experimenting with airfoils for 3D and this thread has some very interesting posts, I thought I could ask:
- Did anybody come to any definitive conclusion about the best kind of stall for 3D flight, as listed by Ollie above?
- Once the choice of type of stall is made, how does that translate into a choice of airfoil?

[BMatthews]

It's pretty hard to guage that since it's a combination of so many variables. The shape for starters, the leading edge sharpness Any sag in the covering if it's not sheeted and even the quality of the gloss in the covering. It all affects how the airflow separates for 3D flying or doesn't separate for normal flying. And because there's more to it than just the shape it's just hard to quantify. The best you can do is pick one that seems to have what you want and be prepared to alter the final model. Or better yet if the final model is a highly complex one that will require many hours to build then start with a Stik style hack ship and work on the airfoil on a model that uses a nice easy to build constant chord wing and see how it works out. If it doesn't don't be afraid to hack into the wing to alter the leading edge curvature, add turbulator strips, play with gloss or flat textures on the wing leading edges, or even build a new wing or two with airfoils using different thickness and high point locations.

The problem as mentioned before is that there's just no way to really accurately predict airflow at the stall point or really near to it other than the pressure curves from something like Xfoil. But what is good and what isn't? Run the curves for various known decent airfoils including flat plates and after studying them use a highly educated guess at what sort of blend or direction you want to try for a first cut.

[ADB2]

Thanks Bruce, on reading your earlier posts in this thread, and checking a few profiles of both full-size aerobatic aircraft and recent 3D models, I noticed there seems to be two schools of thought:
- 3D airfoils derived from aerobatics airfoils: for traditional aerobatics the airfoils chosen are generally thicker (12-15%), with a blunt LE. These have sometimes been modified slightly for 3D by making the leading edge very slightly sharper and moving the maximum thickness a little bit backwards (sometimes not at all). Whatever changes are made, the airfoil is kept rather thick. The reasoning seems to be that some wing drag is desirable, and the general characteristics of aerobatics airfoils are also suitable for 3D flight.
- 3D airfoils derived from flat sheet wings: since flat sheet wings are really where 3D flight began and developed at first, some 3D airfoils seem to have been developed by trying to retain the most desirable characteristics of flat sheets while adapting them to more traditional aerodynamic airfoil shapes and wing building techniques. The results are in general thin airfoils (up to but not exceeding 10% thickness) with a sharper leading edge and the maximum thickness pushed back to around 30% chord.
I am a total beginner when it comes to aerodynamics but I believe the second family of airfoils may provide better 3D model performance, with a clean stall and recovery. Lift seems to be a secondary consideration since wing loading is generally low and thrust-to-weight ratio in recent 3D models is very high. Also enough parasitic drag is provided by the model's fuselage, landing gear and other features that the lower wing drag of a thinner airfoil is not a concern in downpaths.
What do you think? :confused: