HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Dec 18, 2012, 07:11 PM
Registered User
Yak 52's Avatar
UK
Joined Jan 2009
1,248 Posts
Discussion
Winglets for Free Flight Rubber

I was thinking about experimenting with winglets for Free Flight rubber duration models.

From what I have read so far there may be some benefit for span restricted classes and for aircraft flying at lower speeds/higher lift coefficients where induced drag is dominant. FF rubber models are trimmed to fly at min sink (more or less) with no requirement for a wide speed range.

It's even better when a minimum class weight is specified because the structural weight budget is fixed,

Has anyone experimented with winglets for classes like P30 or even Bostonian? Both have a limited span and minimum weight. I'm wondering if the gain in effective span and the improved effect on wingloading would see an overall benefit?
Yak 52 is offline Find More Posts by Yak 52
Reply With Quote
Sign up now
to remove ads between posts
Old Dec 18, 2012, 08:26 PM
B for Bruce
BMatthews's Avatar
The 'Wack, BC, Canada
Joined Oct 2002
11,424 Posts
I know that true winglets, not the little tabs, provide considerable dihedral effect. So with winglets you could consider reducing the center dihedral by a bunch. A simple shallow V with vertical or near vertical winglets would be about the same as a polyhedral wing where the tips were pretty much the same height as the tip of the winglet.

Because the winglets lift inwards they won't do much for the overall wing loading. The other trick would be to figure out how to determine the correct washout needed in the true winglet in order to actually reduce the trailing tip vortices and actually lower the overall drag while maintaining the maximum lift.
BMatthews is offline Find More Posts by BMatthews
Reply With Quote
Old Dec 18, 2012, 09:55 PM
"...certainty is absurd."
kcaldwel's Avatar
Joined Jan 2007
3,632 Posts
Winglet design on models is very challenging, because of the low Re of the winglet. A fixed Cl makes it a bit easier, but indoor models obviously operate at extremely low Re. The intersection shape is very important to limit the intersection drag. You might get a few percent on minimum sink rate, but there isn't any data at that low of an Re I know of.


Winglet contribution to dihedral effect varies with wing Cl. The effect decreases with increasing wing Cl, and isn't that large to start with according to wind tunnel testing. That could be why they don't change the dihedral angle on airplanes they add winglets to, and the effect on aircraft handling is usually small.


http://www.cb-roter-baron.de/winglets.html

Sink rate is proportional to span loading, not wing loading. You can improve the sink rate by reducing the induced drag without lowering the wing loading.

http://ntrs.nasa.gov/archive/nasa/ca...1983017396.pdf

http://ostiv.org/TSgdp_e.doc

Kevin

Edit: I found some quite low Re CFD analysis and wind tunnel testing of winglets. The design range they tested didn't have much effect until they lowered the angle to 30 degrees. That won't work with a fixed span class, unless you use the outer wing panel.

http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA399230
kcaldwel is offline Find More Posts by kcaldwel
Last edited by kcaldwel; Dec 18, 2012 at 11:49 PM.
Reply With Quote
Old Dec 18, 2012, 11:53 PM
internet gadfly
nmasters's Avatar
Colorado
Joined Aug 2006
2,171 Posts
Have a look at Dr Martin Hepperle's site. It uses frames so you have to go through menus to get to the winglet page. Pick <Aerodynamics> from the main menu then <a close look at winglets> on that page
nmasters is offline Find More Posts by nmasters
Reply With Quote
Old Dec 19, 2012, 03:12 AM
I don't like your altitude
Stupot46's Avatar
Joined Sep 2011
3,325 Posts
Interesting stuff.Apologies if I missed a ref.to them,but would Hoerner tips be of any use in this case?
Stuart
Stupot46 is offline Find More Posts by Stupot46
Reply With Quote
Old Dec 19, 2012, 04:05 AM
I bail out, anywhere, anytime
Taurus Flyer's Avatar
The Netherlands, OV, Almelo
Joined Nov 2010
2,959 Posts
Sorry Gents,
Deleted, wrong subject, rubber instead of RC

Success TF
Taurus Flyer is offline Find More Posts by Taurus Flyer
Last edited by Taurus Flyer; Dec 19, 2012 at 07:38 AM.
Reply With Quote
Old Dec 19, 2012, 05:57 AM
Registered User
Yak 52's Avatar
UK
Joined Jan 2009
1,248 Posts
Quote:
Originally Posted by kcaldwel View Post
Winglet design on models is very challenging, because of the low Re of the winglet. A fixed Cl makes it a bit easier, but indoor models obviously operate at extremely low Re. The intersection shape is very important to limit the intersection drag. You might get a few percent on minimum sink rate, but there isn't any data at that low of an Re I know of.


Winglet contribution to dihedral effect varies with wing Cl. The effect decreases with increasing wing Cl, and isn't that large to start with according to wind tunnel testing. That could be why they don't change the dihedral angle on airplanes they add winglets to, and the effect on aircraft handling is usually small.


http://www.cb-roter-baron.de/winglets.html

Sink rate is proportional to span loading, not wing loading. You can improve the sink rate by reducing the induced drag without lowering the wing loading.

http://ntrs.nasa.gov/archive/nasa/ca...1983017396.pdf

http://ostiv.org/TSgdp_e.doc

Kevin

Edit: I found some quite low Re CFD analysis and wind tunnel testing of winglets. The design range they tested didn't have much effect until they lowered the angle to 30 degrees. That won't work with a fixed span class, unless you use the outer wing panel.

http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA399230
Thank you for the references, I will go away and digest...

Yes, the Re is very low: around 15,000-20,000 on a Bostonian. But it seems that there has been some success in F1M (indoor duration) with them so it may be worth a try.

The thinness required for low Re works for me, I was considering tips from 1/32 sheet, maybe with twist and camber streamed or moulded in. And a plug in tip will give the option to experiment and test a bit. We are talking small stick and tissue models here, nothing high tech!

My last Bostonian came out needing 2.5 grams of ballast so I was thinking how I might make use of that dead weight to reduce drag in some way.


Quote:
Originally Posted by BMatthews View Post
Because the winglets lift inwards they won't do much for the overall wing loading.
Bruce, I found this diagram in Peter Masak's work, so I think there will be a modest improvement if you can get them to work.



The problem I have is that in Bostonian the span AND chord are limited, so you can't improve the planform without murdering the wing loading. If winglets let me fly a square wing with less drag then they'll be worth it. In P30 if they permit a lower aspect ratio at the same drag then there is an overall wing loading benefit.
Yak 52 is offline Find More Posts by Yak 52
Reply With Quote
Old Dec 19, 2012, 02:00 PM
Registered User
ShoeDLG's Avatar
Germany, BW, Stuttgart
Joined Mar 2012
806 Posts
I played around with a notional Bostonian wing with winglets using a Vortex Lattice code.

Assumptions:
-Span: 16 in.
-Chord 3 in.
-CL: 0.4
-Vertical location of CG at wing root
-Inviscid analysis

For a baseline, I considered a rectangular platform with 3 degrees of washout (near optimum washout for CL = 0.4) and 3 degrees of dihedral. For this configuration I found:

L/D = 43.4
CL_beta (dihedral effect) = 0.044

For the winglet case, I removed the dihedral from the "main wing" and considered a winglet 2 inches high (90 degrees dihedral) with a root chord of 3 inches and a tip chord of 1.5 inches. I did a coarse "optimization" to arrive at 1.5 degrees of washout for the main wing, and 0.5 degrees of "toe in" for the winglet (untwisted winglet) and found:

L/D = 53.9
CL_beta = 0.156

So the inviscid L/D is better as expected, and the dihedral effect is much greater (more than 3x) than for the no-winglet, 3 degree dihedral baseline.

The difference in span loading correlates well with the plot you provided.
ShoeDLG is offline Find More Posts by ShoeDLG
Reply With Quote
Old Dec 19, 2012, 03:47 PM
Registered User
Yak 52's Avatar
UK
Joined Jan 2009
1,248 Posts
That's brilliant Shoe, thanks for taking the trouble.

I would expect the CL to be a little higher, flying at min sink. I know it will be impaired by the low Re but with a good free flight airfoil (thin, cambered) I estimated about CL0.6-0.7?

That's better for the case for winglets though? ie higher induced drag.
Yak 52 is offline Find More Posts by Yak 52
Reply With Quote
Old Dec 19, 2012, 04:58 PM
Registered User
ShoeDLG's Avatar
Germany, BW, Stuttgart
Joined Mar 2012
806 Posts
If you give a specific geometry and CL I can make this more representative. What this doesn't tell you is the profile drag penalty. I suspect it may end up a wash (or worse).
ShoeDLG is offline Find More Posts by ShoeDLG
Reply With Quote
Old Dec 19, 2012, 07:54 PM
"...certainty is absurd."
kcaldwel's Avatar
Joined Jan 2007
3,632 Posts
I ran Shoe's geometry on XFLR5 VLM with viscosity, using a 3% camber biarc airfoil with the max camber at 40%. The winglets are flat plates. I used a mass of 30g which I have no idea if it is close.

For what it is worth, XFLR5 shows a small sink rate improvement. I have not tried to optimize the cant angle or height of the winglets.

I can easily alter the geometry, and I have quite a few indoor airfoils that run on XFLR5.

Kevin
kcaldwel is offline Find More Posts by kcaldwel
Reply With Quote
Old Dec 20, 2012, 05:45 AM
Registered User
ShoeDLG's Avatar
Germany, BW, Stuttgart
Joined Mar 2012
806 Posts
Kevin,

Can you run the geometry on XFLR5 without viscosity to see how the results compare?
ShoeDLG is offline Find More Posts by ShoeDLG
Reply With Quote
Old Dec 20, 2012, 08:47 AM
Registered User
Yak 52's Avatar
UK
Joined Jan 2009
1,248 Posts
Quote:
Originally Posted by kcaldwel View Post
I ran Shoe's geometry on XFLR5 VLM with viscosity, using a 3% camber biarc airfoil with the max camber at 40%. The winglets are flat plates. I used a mass of 30g which I have no idea if it is close.

For what it is worth, XFLR5 shows a small sink rate improvement. I have not tried to optimize the cant angle or height of the winglets.

I can easily alter the geometry, and I have quite a few indoor airfoils that run on XFLR5.

Kevin
Kevin, again: really appreciate you running this - I have yet to get to grips with XFLR5.

The weight without rubber is set at a minimum of 14g and the motor would be around 3-4g. So an all up weight of 18g is reasonable.

I've attached pics of the just finished model but another is on the drawing board.

The model I have just finished is 16" span with tapered tips. The rib airfoil is Verbitsky BE50 but because of the stick and tissue construction this is not that accurate for the whole wing, especially forward as there is no D box. The turbulator spars give flattish facets so I might try riblets on the next design. The stated max thickness is 23.8% but because of tissue sag I'd guess it's more like 35%. I did calculate the section power factor and best CL^1.5/CD for this airfoil but I'll have to dig it out. As I remember there was not much difference between best CL^1.5/CD and best L/D on this type of airfoil.

The fuselage is 1.5" wide (them's the rules) so the overall L/D is low, I'd guess about 4 or 5. Dihedral is 6 degrees on the outer panels.

I will experiment with some simple plug in tips for this model but I intend to design a new one with a constant 3" chord and winglets. The outer panel could be canted to 30 degrees but there would still need to be a max span of 16" so I'd need to be sure there will still be an improvement. I did wonder about canted tips AND vertical winglets like this?



I worked out the speed as 3.5-4m/s and Re 18-20,000.

So that gives:
Span 16"
Chord 3"
S 48in^2
W 18g
Airfoil approx 4% camber and 7% thick.


The next design will have a few more low drag features, spinner, streamlined nose and maybe wheel spats.

Jon
Yak 52 is offline Find More Posts by Yak 52
Last edited by Yak 52; Dec 20, 2012 at 03:47 PM.
Reply With Quote
Old Dec 20, 2012, 10:24 AM
"...certainty is absurd."
kcaldwel's Avatar
Joined Jan 2007
3,632 Posts
I screwed up. The straight dihedral results were at 30g, and all the others were at 18. That is why the sink rate looked worse, and the L/D better. Higher Re at the higher weight helped the L/D.

Re-running with matching weights shows very little difference with the winglets with viscosity. The inviscid results show an improvement with the winglets.

Sorry about that.

Jon,

I re-ran the sim with an 18g weight, and the BE50 airfoil on XFLR5. I tried a 30 degree outer panel as well. Please take all these sims as very rough approximations only - XFLR5 is too simple to model many of the nuances, and it makes some fundamental errors as well.

The XFOIL results for the Be50 show it badly needs a turbulator. I couldn't get decent results with a turbulator at 25%, but did with it at 15% chord. XFOIL is actually more accurate at Re below 20k, where there aren't as many problems with the transition point, and it is quite good even at the messy transition Re. I'll try to find the aft most point for the turbulator if I have time.

I can't model the fuse - well, I can, but XFLR5 results for anything other than a wing are pretty meaningless so I won't bother. I tried a 4in chord with no improvement in sink rate at least with no fuse drag.

The 30 degree canted tip panels do not show an improvement on XFLR5.

Kevin
kcaldwel is offline Find More Posts by kcaldwel
Last edited by kcaldwel; Dec 20, 2012 at 12:12 PM.
Reply With Quote
Old Dec 20, 2012, 12:13 PM
"...certainty is absurd."
kcaldwel's Avatar
Joined Jan 2007
3,632 Posts
I made a mistake. The straight dihedral results were at 30g, and all the others were at 18.

I have now corrected it above.

Kevin
kcaldwel is offline Find More Posts by kcaldwel
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Free flight scale rubber plans tectonite Free Flight 0 Nov 27, 2012 12:36 AM
Discussion anyone know who to contact for free flight rubber motor props? Unclesteve Free Flight 4 Sep 05, 2012 07:55 PM
Discussion New Hobbyking free flight rubber powered planes. Tartago Scratchbuilt Indoor and Micro Models 4 Jun 22, 2012 02:25 PM
Build Log Rubber Powered Free Flight Conversion to RC mark1968 Parkflyers 0 May 28, 2012 11:41 PM
Discussion Geared Motor For 36" Free Flight Rubber Conversion?? Gemini2 Power Systems 6 Apr 05, 2012 10:21 AM