HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Dec 16, 2010, 08:23 PM
yyz
Registered User
yyz's Avatar
USA, CA, Paso Robles
Joined Dec 2004
2,401 Posts
Help!
XFLR5 Wing Design Help Needed

Greetings All,

I am attempting to design a scale sailplane wing with XFLR5 and am having some issues that I could use some help with.

The wing is for a 1:3 scale Schempp-Hirth Discus 2c. The wing is flapless which, counterintuitively, makes the design a little more challenging.

I am using the tutorial written by Francesco Mescia (http://www.rcsoaringdigest.com/pdfs/...SD-2008-02.pdf), plugging in my values where appropriate.

The problem I'm having is with the spanwise lift distribution. Can someone explain the downward "spikes" I'm getting around the quarter span?

Thanks in advance,

Mike
yyz is offline Find More Posts by yyz
Last edited by yyz; Dec 16, 2010 at 11:17 PM.
Reply With Quote
Sign up now
to remove ads between posts
Old Dec 17, 2010, 12:21 AM
Master of the Wind
G Norsworthy's Avatar
United States, CA, San Jose
Joined Sep 2008
1,168 Posts
I'm not an expert but I will play around with it if I have any time to goof off tomorrow.
G Norsworthy is offline Find More Posts by G Norsworthy
Reply With Quote
Old Dec 17, 2010, 03:40 AM
flying monkey
frichte's Avatar
Europe
Joined Sep 2008
123 Posts
Quote:
Originally Posted by yyz View Post
The problem I'm having is with the spanwise lift distribution. Can someone explain the downward "spikes" I'm getting around the quarter span?
You have 12 y-panels on the inner portion of the wing but only 1 for each of the other sections. Try increasing the number of y-panels for these sections first.
About the airfoil choice: I don't think you will get away with a foil that only has 8% thickness. I'd recommend 11% at the root and between 9 an 10% for the rest of the wing.

Friedmar
frichte is offline Find More Posts by frichte
Reply With Quote
Old Dec 17, 2010, 01:28 PM
Master of the Wind
G Norsworthy's Avatar
United States, CA, San Jose
Joined Sep 2008
1,168 Posts
My XC wing was done with 11 y panels per division.

Norbert Habe has a large collection of thicker airfoils for large wing planes. I looked at them for XC and they did not apply but maybe for this they would be ideal.

http://tracfoil.free.fr/airfoils/h.htm
G Norsworthy is offline Find More Posts by G Norsworthy
Last edited by G Norsworthy; Dec 17, 2010 at 01:32 PM. Reason: add link
Reply With Quote
Old Dec 17, 2010, 02:32 PM
yyz
Registered User
yyz's Avatar
USA, CA, Paso Robles
Joined Dec 2004
2,401 Posts
Quote:
Originally Posted by frichte View Post
You have 12 y-panels on the inner portion of the wing but only 1 for each of the other sections. Try increasing the number of y-panels for these sections first.
About the airfoil choice: I don't think you will get away with a foil that only has 8% thickness. I'd recommend 11% at the root and between 9 an 10% for the rest of the wing.

Friedmar
Excellent find, Friedmar. There is a "Reset Mesh" button in the wing editing dialog box. I clicked that and reran the analysis. See attached image.

Structural concerns aside, what is it specifically about the choice of airfoil that you think won't work? I am looking for a high L/D at cruising speeds, sacrificing some performance in climb.

Thanks,

Mike
yyz is offline Find More Posts by yyz
Reply With Quote
Old Dec 17, 2010, 02:34 PM
yyz
Registered User
yyz's Avatar
USA, CA, Paso Robles
Joined Dec 2004
2,401 Posts
Quote:
Originally Posted by G Norsworthy View Post
My XC wing was done with 11 y panels per division.

Norbert Habe has a large collection of thicker airfoils for large wing planes. I looked at them for XC and they did not apply but maybe for this they would be ideal.

http://tracfoil.free.fr/airfoils/h.htm
Thanks Greg. I'll run a few of these and see what I get.

Mike
yyz is offline Find More Posts by yyz
Reply With Quote
Old Dec 17, 2010, 02:53 PM
flying monkey
frichte's Avatar
Europe
Joined Sep 2008
123 Posts
Quote:
Originally Posted by yyz View Post
Structural concerns aside, what is it specifically about the choice of airfoil that you think won't work? I am looking for a high L/D at cruising speeds, sacrificing some performance in climb.
It's mainly structural concerns but with a glider that big (and with a high wing loading) you won't sacrifice performance significantly with a little more thickness. With 3% camber 10 or 11% thickness might even perform a little better at negative angles of attack (speed, inverted flight). So why make it almost impossible or extremely expensive to build a stiff plane when it's not needed?
While on the tracfoil website take a look at Thierry Platons TP210 and TP202 series. They seem well suited for your application.

Friedmar
frichte is offline Find More Posts by frichte
Reply With Quote
Old Dec 19, 2010, 10:37 PM
yyz
Registered User
yyz's Avatar
USA, CA, Paso Robles
Joined Dec 2004
2,401 Posts
Perpetual motion machine?

I did a little playing around with the TP210 airfoil that Greg and Friedmar suggested and I'm getting some pretty strange results. Perhaps someone can point out the error of my ways.

Using the published planform from the f/s Discus 2c (18m), I created a 1:3 scale wing that used the 11% version of the airfoil at the root and 10% for the remainder of the wing. After a couple of minor tweaks to the twist at several stations, I was able to get what looked -- at least to my untrained eye -- to be a reasonable spanwise lift distribution.

The GPS Triangle racing rules provide a formula for maximum wing loading so I did the math and came up with a maximum weight of 15.8 kg. and guessed at a minimum weight of 9 kg.

Here is where the interesting part comes in. The L/D polar is "better" at the higher wing loading. Not just a little bit better but a whole bunch better (42:1 vs 38:1). That's almost 10% better and just can't be.

Maybe I'm just flailing around without a clue (and please point that out if that's the case) but something is not quite right.

Is my not accounting for the drag of the fuselage skewing the results for the wing and the tail combination? Did you guys recommend a "magic" airfoil?

I have to be missing something really simple,

Mike

ps: Just ran the venerable MH-32 section using the same wing geometry and it yields similar results
yyz is offline Find More Posts by yyz
Last edited by yyz; Dec 19, 2010 at 11:06 PM. Reason: misread min sink polar
Reply With Quote
Old Dec 20, 2010, 12:19 AM
Master of the Wind
G Norsworthy's Avatar
United States, CA, San Jose
Joined Sep 2008
1,168 Posts
I saw the same thing with some other designs I was looking at. Seems strange at first but what I think is happening is you are increasing the Re when you increase the weight and the performance at that point is the strongest function of the Re. What is really weird is in some cases the sink rate actually gets better at the higher wing loading. That is why I am now building a smaller XC model.
G Norsworthy is offline Find More Posts by G Norsworthy
Reply With Quote
Old Dec 20, 2010, 12:51 AM
Registered User
Joined Jan 2009
495 Posts
Highest L/D for a given design is at highest mass (wing loading) authorized by structural design/rules, most rearward CoG with corresponding optimized decalage, cleanest configuration (smooth, all openings, doors, joints taped), lowest altitude and temperature (highest air density). For the models the Re is an important contributing factor, but the fact that by setting the glider that way you reduce the angle of attack (therefore the total drag) is why the L/D improves. On full scale because of the configuration and mass limits the L/D increase is around 1 to 2 points from "more standard" settings. On a model glider, depending on the mass variation, you may get as much as 4 points (or more) improvements. Remember, your calculations are depending on the software used for the analysis and you are only looking at the wing, not the full glider. The gain will be much less in reality due to the fuselage and all the interferences drag. The design of the fuse at this level becomes critical, this includes the design setting for the wing incidence at the optimal speed and the fuselage shape/drag at this AoA.
fnev is online now Find More Posts by fnev
Reply With Quote
Old Dec 20, 2010, 05:26 AM
Registered User
ron gijzen's Avatar
Liempde, Netherlands
Joined Jul 2007
150 Posts
Quote:
Originally Posted by G Norsworthy View Post
That is why I am now building a smaller XC model.
That doesn't makes sense!?
If you reduce the size, than the chord and hence also Reynolds will be reduced, reducing the L/D ratio.
I think you should ballast your model more, not make it smaller for the same weight.
ron gijzen is offline Find More Posts by ron gijzen
Reply With Quote
Old Dec 20, 2010, 06:08 AM
the answer 42 is
Switzerland, AG, Lenzburg
Joined Jan 2006
1,894 Posts
well in fact it makes a lot of sense, a smaller plane is easier and cheaper to build, it is easy to transport and you will need less ballast, in fact you can build it on the heavy side from the beginning, so instead of using lots of materials on a 4m wing you can use the same amount for an extra heavy yet smaller wing

just my two cents

EZ
Edwinzea is offline Find More Posts by Edwinzea
Reply With Quote
Old Dec 20, 2010, 08:19 AM
flying monkey
frichte's Avatar
Europe
Joined Sep 2008
123 Posts
Back on topic:

Quote:
Originally Posted by yyz View Post
Here is where the interesting part comes in. The L/D polar is "better" at the higher wing loading. Not just a little bit better but a whole bunch better (42:1 vs 38:1). That's almost 10% better and just can't be.
A 10% improvement in L/D for increasing the weight by more than 50% sounds feasible to me. Please note that the sink rate (Vz over Vx graph) also increases! It's almost linear to wing loading.
Please keep in mind that it's not just the glide ratio that counts for distance flying. The lower the sink rate the more overall flight time you have for finding thermals and they can also be weaker while still allowing you to climb.

Mike, I am not sure if it's wise to do all the polar calculations with the tailplane included. To make it more realistic you would have to adjust decalage and CG for each operating point on the polar. So it's easier to skip the tail while still optimizing the wing. And if you can live with a semi-scale plane: Increasing the area of the horizontal tailplane by maybe 5-10% will help maneuverability and flying characteristics. Original tails tend to be on the small side of things.

For further optimisation of the wing you might want to have a look at the spanwise plot of local c_d. If the increase in c_d towards the tip is significant at slower flying speeds, then optimizing the tip airfoils for lower Re is a good idea. I think you can reduce thickness to 9% at the tip on the outermost wing segment and maybe move the maximum thickness of the airfoil forward a little (by 2% of chord).

Friedmar
frichte is offline Find More Posts by frichte
Last edited by frichte; Dec 20, 2010 at 08:21 AM. Reason: typo
Reply With Quote
Old Dec 20, 2010, 11:42 AM
yyz
Registered User
yyz's Avatar
USA, CA, Paso Robles
Joined Dec 2004
2,401 Posts
Quote:
Originally Posted by fnev View Post
Highest L/D for a given design is at highest mass (wing loading) authorized by structural design/rules, most rearward CoG with corresponding optimized decalage, cleanest configuration (smooth, all openings, doors, joints taped), lowest altitude and temperature (highest air density). For the models the Re is an important contributing factor, but the fact that by setting the glider that way you reduce the angle of attack (therefore the total drag) is why the L/D improves. On full scale because of the configuration and mass limits the L/D increase is around 1 to 2 points from "more standard" settings. On a model glider, depending on the mass variation, you may get as much as 4 points (or more) improvements. Remember, your calculations are depending on the software used for the analysis and you are only looking at the wing, not the full glider. The gain will be much less in reality due to the fuselage and all the interferences drag. The design of the fuse at this level becomes critical, this includes the design setting for the wing incidence at the optimal speed and the fuselage shape/drag at this AoA.

Hahhhhhhhhhh! I'm not an idiot Thanks for the explanation, fnev.

So I think it is safe to say that flying at maximum wing loading is better, at least in strong conditions, with the caveat that you are going to incur a penalty climbing because of the increased sink rate which jibes with proven, 50+ year-old soaring theory.

What if you turn the problem on its head and design the wing with the thickest airfoil possible -- yielding the most volume for disposable ballast (water, let's say) -- that still performs well at high speed? The trick is then finding a "thick" airfoil that will also perform well at lower wing loading for weaker conditions?

For the wing geometry posted earlier, a back-of-the-napkin calculation suggested that wing could carry ~2.5 kg of water. Intuitively, that seems like an appreciable amount of weight and variability in wing loading.

On the other hand, the current "hot" ships (eg, Baudis Antares, KV 304 Shark) on the GPS Triangle circuit are taking the opposite, more F3B-like approach: thinner, lower camber airfoils.

Wondering if anyone has really done the math,

Mike





M
yyz is offline Find More Posts by yyz
Last edited by yyz; Dec 20, 2010 at 12:54 PM.
Reply With Quote
Old Dec 20, 2010, 11:44 AM
yyz
Registered User
yyz's Avatar
USA, CA, Paso Robles
Joined Dec 2004
2,401 Posts
Quote:
Originally Posted by frichte View Post
Back on topic:



A 10% improvement in L/D for increasing the weight by more than 50% sounds feasible to me. Please note that the sink rate (Vz over Vx graph) also increases! It's almost linear to wing loading.
Please keep in mind that it's not just the glide ratio that counts for distance flying. The lower the sink rate the more overall flight time you have for finding thermals and they can also be weaker while still allowing you to climb.

Mike, I am not sure if it's wise to do all the polar calculations with the tailplane included. To make it more realistic you would have to adjust decalage and CG for each operating point on the polar. So it's easier to skip the tail while still optimizing the wing. And if you can live with a semi-scale plane: Increasing the area of the horizontal tailplane by maybe 5-10% will help maneuverability and flying characteristics. Original tails tend to be on the small side of things.

For further optimisation of the wing you might want to have a look at the spanwise plot of local c_d. If the increase in c_d towards the tip is significant at slower flying speeds, then optimizing the tip airfoils for lower Re is a good idea. I think you can reduce thickness to 9% at the tip on the outermost wing segment and maybe move the maximum thickness of the airfoil forward a little (by 2% of chord).

Friedmar

Thanks again for the help, Friedmar. Very much appreciated.

Mike
yyz is offline Find More Posts by yyz
Last edited by yyz; Dec 20, 2010 at 12:48 PM.
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Help! XFLR5 help needed Lacquerhead Hand Launch 17 Jan 28, 2013 02:32 PM
Discussion LARGE Quadracopter, 400 pound payload, design help needed WarJarrett Multirotor Talk 15 Jan 20, 2012 04:10 PM
Help! XFLR5 help needed GaryO Modeling Science 37 Aug 05, 2011 07:47 AM
Discussion Ligeti Stratos, help with Blucor design needed himeros Foamies (Scratchbuilt) 19 Dec 24, 2010 05:03 AM
Discussion XFLR5 help needed GaryO Hand Launch 4 Oct 14, 2007 10:42 PM