Optimizing Wing Aerofoil for UAV long range flight

3689491683?profile=original

Around 10 days ago I started a debate on wing Aerofoil choices to improve flight range rather than endurance. A number of folk joined in with ideas and suggestions, and the choice was eventually made - The Selig S7075 at 9% thickness.

Many Aerofoils were simulated and the S7075 choice seemed a good one,  so I proceeded to make a prototype set of wings to begin some trials.

The 'original' wing is a Clark Z Aerofoil, 316mm Chord and 1.8meter span. Fitted to our SurVoyeur aircraft with an all up weight of 4.8kg ( a heavy camera...) we achieve around 35km flight range at 21m/s airspeed on two 4cell lipo packs, each 5000mAH.

The 'New' wing is the S7075, chord 315mm, and span 2meters. The wings are CNC foam cut cores, vacuum bagged with Carbon cloth and mylars top and bottom. Ailerons are almost full wing length, but split so that the inboard can be used solely as flaps to test the limits for low speed autolanding and autolaunch.

Flight trials will start in the next few days.

For all those who were part of the discussions, here are some photos of the process and the results...

Regards

Joe

The NamPilot...Swakopmund

3689491798?profile=original

3689491739?profile=original

3689491811?profile=original

3689491832?profile=original

3689491853?profile=original

3689491841?profile=original

E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones

Comments

  • I forgot to add a turbulator to S7075. With a turbulator (red curve) it looks a bit better than SA7035.

    3692576212?profile=original

  • Sorry to hear that, Joe. I haven't used the software myself.

    Another thing. Thus far we've only looked at airfoil polars at certain Re numbers. The problem with a, let's say Re=300k polar, is that you only get one point of useful information off that polar. That's because that particular Re number occurs only at one certain speed. So, looking at polars where Re is a constant is a bit deceiving.

    A model airplane is never a point design like an airliner. Luckily, there are type 2 polars in XFLR5 and XFOIL. In a type 2 polar the  Reynolds number is also a variable. So, in result you get a polar that takes Re changes into account. You could also construct a type 2 polar manually by picking points from a type 1 polars.

    For example, here's a type 2 polar comparing SA7035 and S7075:

    3692576036?profile=original

    You can easily see that SA7035 has an edge at the lower Cl's you're flying at.

  • Martin,

    I have been trying hard to get something useful out of the Evolution software....

    I understand they have used it - or a derivative- for the Shark A/C and for a large span glider, but even if I use the setup as in their glider tutorial it gives a result which make little sense - And if you change anything even slightly, you invariably get strange error message , with no explanation, and cannot regress, ie, acknowledging the error message just dumps you out of the program, with nothing saved and you start all over again - sometime this happens after two hours of processing - not a very user friendly, or in fact usable, software...

    They do not respond to any queries - this was 10days ago - maybe all on holiday..

    If you have it and are playing - I gave two simple inputs point:

    1st point - for takeoff and land-

    Clmax = 1.5

    Re=200K

    Mach =0.033

    Constraint = minimize Cd/Cl^1.5

    2nd point for cruise

    Cl = 0.4

    Re=365K

    Mach = 0.06

    Constraint = minimize Cd

    I also chose Very Light Aircraft under Geometry for run-1

    For another attempt I chose Sailplane - RC-Model.

    The results were not believable - you can save the co-ords of ay selected foil to an excel file - I did so , converted to .dat, and read into profilli Pro - the plots are nothing like what is shown in Evolution!

    If you select a constraint such as - 

    1st point - for takeoff and land-

    Clmax = 1.5

    Re=200K

    Mach =0.033

    Constraint = minimize Cd/Cl^1.5

    2nd point for cruise

    Cl = 0.4

    Re=365K

    Mach = 0.06

    Constraint = maximize Cl/Cd - ( the only change from the previous runs)

    Then it dies part way through and gives a useless error message, and you cannot restart/continue, etc.

    Doing anything just exits to windows....

    Seems like a tool that was used by the design engineers, who then thought is a great idea, and then pasted a GUI around it, warts and all..help files are useless...

    No free lunches!

    Joe

  • Hi Vince,

    Sorry, was out in the bush for a few days break..

    Vince, the CNC Router I built from the ground up - so no real off the shelf bits I am afraid. Here are some pics if you want ideas - Bit off topic I suppose..

    I used slides from a German company - the slides are called 'ROLLON'

    Steppers were from an old (BIG) Calcomp drum pen plotter. The spindle is a 1kW Perske 45Krpm 3phase variable speed unit. Stepper drivers are GECKO DRIVE, very good. Table and base frame is aluminium, X and Y gantry are composite and Z axis all aluminium. 

    Working Area is about 900X750mm - I would advise to try go for longer - 1600mm X 700 would be great...

    3692576054?profile=original

    3692576075?profile=original

    3692576116?profile=original

    3692576088?profile=original

    3692576015?profile=original

    3692576136?profile=original

  • Take a look at this program --> http://www.evolutiondesigns.eu/

    Thread in RC Groups -- http://www.rcgroups.com/forums/showthread.php?t=1785953

    The program is free to use, but if you want coordinates for one of the generated airfoils then you're going to have to pay for it.

    It certainly won't find you an airfoil that has 2x lower drag than other comparable foils. But the performance gain might be worthwhile.

  • Joe, can I ask which main bits you used for the cnc cutter. I almost started building one last year but got side tracked as usual.

    I would like to ‘multi task’ the breakout board and stepper drivers so they also do a 3d printer and router.

    I see there are many cheap ones that work from a printer port but think the USB ones would be more future proof.

    How about this setup.

    http://www.ebay.co.uk/itm/4-axis-CNC-USB-Card-Mach3-Breakout-Board-...

     

    Vince

     

     

  • Moderator

    Nice job.

  • Sorry about that. I didn't think about thickness. Nevertheless if you compare S7075 and SA7035, which have similar thickness but different camber, you can that the polar of S7075, which has more camber, is up and to the right in regard to SA7035.

    I'm not sure how much the tip shape is going to help. But I guess any shape is better than just straight cut tips.

  • Your comparing 2 thick foils with 2 thin ones. No surprise the thinner ones have a lower Cd. That does illustrate my point though, that the 4512, for a 12% foil is looking v good at higher Cl (loiter region). Perhaps I should be more concerned about induced drag at these speeds, so am considering the wing tip shape. Not big winglets or anything but a simple Hoerner.

    Sorry Joe, think I just opened another bag of worms on your blog.

  • A simple rule applies to airfoils. When you increase camber (or deflect a camber flap) the Cl/Cd polar is shifted upward and to the right. so, when you decrease camber (or add reflex)  the polar is shifted down and to the left. With a small camber increase there's no drag penalty. The drag might even decrease.

This reply was deleted.