Thread Tools
Aug 23, 2008, 05:51 PM
Deniable plausibility
Shedofdread's Avatar
Thread OP

F3B design - airfoils etc

Hi Everyone,

I have a few questions relating to the design of F3b sailplanes to which I would appreciate your thoughts / wisdom.

What is the best (most competitive) current F3B airfoil? Many current designs appear to use 'secret' airfoils but of the published airfoils, which is the best. I am finding it hard to find published polars for the popular airfoils.

Many airfoils now seem to use what to my mind is quite a low camber (1.5ish %). How well do they do the 10 minutes in marginal conditions?

Which is more important an efficient planform (ie an elliptical lift distribution) or having enough tip chord to keep the Reynolds number at an acceptable level? The current crop of designs have tip chords that should be too small but they clearly perform well - where does the balance lie?

Just by way of explanation I started flying when you had to build high performance not buy it and wishing to re-start competitive flying I just have to use my own design. I know this isn't logical but where's the logic in a sport that involves throwing 1000s of £/$ off a cliff?

Thanks for reading this and hopefully replying.

Sign up now
to remove ads between posts
Aug 23, 2008, 06:28 PM
Detail Freak
target's Avatar
I'm no expert, but here are some of my observations:

You are right that the current crop of profiles seem to have less camber than in days gone by. I think the thought is less drag in section, and you can add camber with a little TE drooping. You would then adjust the section to suit your task.

I like this way of thinking, since I fly F3F only so far. Why reflex an undercambered foil to make it less draggy when that is the (my) primary focus??

The other question is that many of the new crop of design vary either the thickness of the root section as it narrows towards the tips, or they use a different variant of the profile altogether. That allows the tips to work better than you think they should.

As for what's the best section? Ha! Ask ten people, you'll get propably ten answers.
There are lots of good new sections out there. But even some of the older ones are still competitive.
If I were going to scratch build, I would look for a section that has a reasonable thickness percentage, especially at or near the hinge line. That will allow building a stiffer structure more easily.

Good luck with your design.


PS. Here's a link you may be interested in:
Last edited by target; Aug 23, 2008 at 10:50 PM.
Aug 23, 2008, 10:23 PM
Registered User
mlachow's Avatar
Actually you will find the best designs use a series of airfoils so they can be tweaked a little for the Rn differences at the tips as well as for structural reasons since it's easier to build thin tips than thin center sections where the bending loads are greatest.

The idea is to launch as high as you can, and hopefully that's high enough to make the 10 minutes. Actually it's unusual to get really dead air, there is almost always some activity. So while it would be easy to design a model to make the 10 minutes. You might want to trade that off on a little more speed.

You also see the speed influence in fuselage design. If there is any extra room after the battery and servos, it's too big.
Aug 24, 2008, 05:53 AM
Deniable plausibility
Shedofdread's Avatar
Thread OP
Thanks both of you for your thoughts.

Target - I see what you mean with regard to camber changing and I also understand that the thinner the airfoil the lower the Re that can be tollerated. That however raises the problem of access to the coordinates for a transitioning set of airfoils!

mlachlow - Absolutely, parasitic drag is to be eliminated. I am about 3/4 way through making a new fuselage plug that at its biggest is only 50mm deep by 35mm wide.

I've nearly finished a 2m span slope model that should test some of my concerns with regard to planform. This model uses the mg06 airfoilwhich looks quite good (low pitching moment, decent CL max) however at Cl = 0 (or thereabouts) there appears to be drag rise. Any thoughts? Obviously Miraj is quite quick but any experience of this airfoil?

Thanks again for your input,

Aug 24, 2008, 06:17 AM
Registered User
So, when do we expect a wing with 20+ airfoils incorporated?
Just kidding, but the number of foils is no guarantee for success, it might even become a marketing tool. But in the end, I do believe in successive developments.
Aug 24, 2008, 06:39 AM
Deniable plausibility
Shedofdread's Avatar
Thread OP

Thanks for saying that - I was thinking that there was a lot of 'marketing' and not so much science in some of the 'new' designs but I daren't say it!

Aug 24, 2008, 07:24 AM
launch high, go fast
luvF3b's Avatar
F3b design, hah easy (not)

Well, I've been dabbling a bit in both Profili and Xflr5 developing my design ( several hundred hours so far ! )

Firstly you need to consider the launch. Given the limited power available on the modern FAI winch, the model size should be limited to around 60dm². Since the foil should be biased towards fast flight due to F3b scoring implications (see below), the camber will be low. A larger leading edge radius can help to improve the max Cl. Of course the low camber helps in the zoom where Cl is close to zero.

Of all the tasks, speed is the one that gives the pilot the opportunity to win. Thermal and distance are tasks where you can only loose points, so the modern model is orientated towards speed. The low area also helps here, as does selection of a good profile, minimisation of parasitic drag etc..

So, for speed performance, the important factors are 1) Launch performance, 2) straight line speed and 3) turn performance. For launch performance (1) you need a model size matched to the winch power, and a profile that can deliver high Cl. For straight line performance (2), where profile drag is the highest you need a profile that has extensive laminar flow, and for turn performance (3), a profile that can deliver high Cl, with low drag. These are conflicting requirements.

Modern F3b foils are low camber, but to obtain a wide drag bucket they are still around the 8 - 8.5% thickness mark. I would suggest having a look at the website On that site, they describe the development of the Freestyler, a currently popular Feb/F3f model (profiles AH37, AH69). Take note of the use of MH31 bottom surface on the profiles, and the benefits that surface gives with extensive laminar flow. Looking closely at the Radical, Ceres, Shooter; these models appear to have a similar bottom surface which is substantially flat (or slightly convex) from 30% chord rearwards. These foils don't need a large amount of reflex to go fast. Also reflex creates a discontinuity on the bottom surface that is sure to disrupt laminar flow. Choose the profile to drive speed performance, so it should be low camber, and not require much reflex flap.

The shape of the top surface appears to be much more important for launch, thermal and distance, and not so much for speed.

The biggest factors for thermal performance are wing loading and aspect ratio. At slow speeds, induced drag is king, and that is determined by aspect ratio and the operating Cl. Low wing loading reduces the operating Cl for a given (slow) flight speed, and high aspect ratio reduces it further.

This creates another conundrum. Increasing aspect ratio for a given wing area reduces the reynolds number. This is not all bad, at least for speed. As you have commented, as the reynolds number increases many profiles start to actually produce HIGHER drag at Cl=0.05 (Re= 500K - 1000K). This can be verified by looking at type 2 polars in Xflr5 or Profili where the polar turns to higher drag at low Cl. Profiles with the MH31 style bottom surface don't suffer as much.

You would think that sorting out the performance at slow speed and high speed would guarantee performance at mid speed (distance) - not so. At these speeds the effect of reynolds number is most pronounced. A high wingloading helps here, and a profile chosen carefully to minimise the top surface separation bubble that will occur at distance speeds. Choosing a top surface on the profile that allows the flap/aileron hingeline to act as a natural turbulator may be of benefit. The goal here is to try for laminar flow to the 75% chord point (the hingeline). This of course creates compromises at slower speeds.

So, at this stage, I'm reading that the state of the art is to choose a profile that gives extensive laminar flow on the bottom surface at high speed (ie low camber and minimal reflex flap), and uses a top surface that controls the laminar flow to the hingeline where turbulated transition to turbulent flow occurs.

OK. So far we've been talking about the average profile on the wing. What about transition? Some posts have already commented about matching the profile across the wing to the local reynolds number. Eric dahl Christiansen comments on his website about planforms. Basically, an elliptical planform to about the 90% semi span point seems to be most efficient. After that the tip is lost in 3D flow and vortex. For reynolds number effects, see the design study by Frits Donker Duyvis on the fletcher Basically the profile at the 90% span point should be thinner, and have a more forward thickness point. There is however one more consideration, and that is at the root of the wing, where the profile is operating under the turbulent influence of the fuselage (N-Crit is smaller) Profiles at that point tend to be thicker (due to structural constraints) and have a more rearward thickness point (to try to maintain a longer laminar flow in the allready slightly turbulent oncoming flow)

Sorry for the long reply, but thats the story as I see it!
Aug 24, 2008, 08:44 AM
Deniable plausibility
Shedofdread's Avatar
Thread OP
Luvf3b - thanks for taking the time to post in such detail.

Have previously looked at the Freestyler stuff. The airfoils look similar in concept to the HD series. The Trinity F3b uses HD45 - anyone got any experience?

Have used 'liftroll' to optimise planform from an elliptical lift distribution point of view. Planform looks good (99% and not too bad tip stall) but very small tip chord. Even at the 90% span point still only about 120mm. Low Re flow is unreliable (your coments re profile mods notwithstanding).

Winglets? Increased effective aspect ratio, no impact on roll rate. Thoughts?

Thanks again to everyone who has posted. The wisdom of others allows one to crystalise ones thoughts.

Aug 24, 2008, 09:05 AM
Detail Freak
target's Avatar
I like the HD45, but I don't own a plane with one. I built a plane for my brother that uses HD45.
I think the profile is good all around from my limited stick time on that plane, but may be somewhat lacking in the speed department compared to the new crop of profiles in developement/production now.
The Trinity f3B looks pretty decent for F3F, and I watched one turn in a 33 second run once.

Aug 24, 2008, 09:30 AM
working to the closest cm
jirvin_4505's Avatar
Something that may influence your design is the method of construction.

What construction methods do you have available?

IIRC some airfoils are more critical than others.

I'm also interested in your timeframe - when did you want the model to fly? This may affect the depth of research and iterations performed.

Is this a lone design/build or are you in with a group?

I remember when Nic Wright visited Australia he proposed that you would do better at competition by practice flying models rather than designing and building models.

Similarly when Martin Webershock visited Australia he emphasised the benefits of a team flying the one model as this provided much more feedback on what works for set up of a model.

I believe these 2 pieces of advice are factors in the design cycle.

Testing your design.. How would you know you had the ....
the best (most competitive) current F3B airfoil?

>>>Many current designs appear to use 'secret' thinking here is that even if they were available would they make a difference? Could you see the difference?

Probably a better question in helping you with your design is what difference do you want? Which bias in the design appeals to you? Iffy on duration and good at speed and visa versa.

Say you were interested in F3j you could use Mark Drela's published airfoils and build a Supra like model or you could obtain the "secret" Pike perfect airfoils and base you design on that model. In the end the airfoils fit with the thinking that went with the model. Plus would having the secret airfoil produce better results than going with the published airfoil? What impressed me in the details on designing the Pike Perfect is the depth of understanding about what was required in an airfoil to do a particular job in a particular position.

Cheers Jeff
Really enjoying the discussion for the insights into what make the modern F3b such a great alround flying machine.
Aug 24, 2008, 10:08 AM
Deniable plausibility
Shedofdread's Avatar
Thread OP

Through work I have access to negative moulding technology but the first one would be a foam wing version to test and unfortunately it will be a solo build but on the plus side that gives me control (power!!!).

What you say re Nic Wright strikes a chord - he had what appeared to be very pragmatic approach which clearly yeilded success. If you look at the Electra E1 (3 view on it appears to be a very 'reliable' design. Through work I heard a saying - 'to finish first, first you have to finish and finish first'. This applies to flying as to many other areas.

Aug 24, 2008, 06:04 PM
launch high, go fast
luvF3b's Avatar
Winglets can offer advantages at intermediate to high lift coefficients, providing they are designed correctly. At low lift coefficients they give more drag than a conventional tip.

Have a look at where an example is shown for full size sailplanes. Because flying fast in full size operates at lift coefficient around 0.2 - 0.3, a benefit is seen for the whole flight envelope. For F3b, flying fast is Cl around 0.0 - 0.05. Winglets in this regime increase the profile drag and don't help with induced drag, because there is very little at these low Cl's
Aug 24, 2008, 06:36 PM
Detail Freak
target's Avatar
Not to mention just being a real pain for transport and being subject to damages in landing transport.
I think they would have make a real gain in performance to make the rest worth while.
For a scale plane to look scale I could see using them.

What about the slightly curved up tips??
Personally, I don't like them for all the above reasons I've mentioned, but do they help reduce drag?
They don't seem to be in "style" as much as they used to be.

Aug 24, 2008, 07:07 PM
Deniable plausibility
Shedofdread's Avatar
Thread OP
If the last 10% of the wingspan is deflected upwards (but with a suitable radius to reduce separation due to boundary layer stress) to form a winglet would this provide an efective solution? the winglet should be designed such that the projected planform would still be the same.

I agree absolutely no help with speed but with surely the fact that they give increased effective aspect ratio would mean you could build a wing with lower aspect ratio and hence more area and spar depth?

I also agree that transport would be awkward and the potential for damage is great.

Keep the ideas coming,

Aug 24, 2008, 07:37 PM
working to the closest cm
jirvin_4505's Avatar
Wingtips – difference between theory and practice.

Mark Drela has posted his thoughts on wingtips somewhere on rcgroups – he designs without them. IIRC not needed or usefull. More important is ailerons through to the very tip.

The Calypso Cobra held the F3b speed record until recently – no curvy wing tips ailerons to the very tip

The Avionic F5b models - world championship standard – no curvy wing tips ailerons to the very tip.

Always interesting the choice between theory, practice and results.

Cheers jeff
Rather not have the curvy ones myself – catch on the wing bags when packing.

Quick Reply

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Question What is F3B, F5B, F5J etc.??? cwayland Slope 2 Apr 26, 2007 02:26 PM