UAV Wing: Scratch build log

[DaveAtDigital]

What’s up all? Thought I’d start keeping a journal of a project my buddy and I are working on:

The Objective:

To create an UAV Wing capable of autonomous takeoff, landing and flight paths.

Other goals:
Maximum height,
Maximum distance
Maximum flight time.


I spose it’s worth mentioning at this point, that I have ZERO RC or fabrication experience. None. Seriously, it takes me like 2 hours to put together a 15 minute ikea shelf

But my amigo and I have a decent programming background, and that’s where we’ll start.

So far, we have the following hardware:
Servo Controller & servos
FM modems (Xbee Pro 9600bps 15mile range)
GPS receiver
On board CPU
Sparkfuns 6DOF IMU
Car battery charger, wire & lots of foam.

He’s working on the FM modem protocol, & IMU; I’m working on the servo controller, and the Joystick interface.

This + an aircraft should get us flying manually...we’re hoping to launch our first flight sometime this month.

So far, its starting to look like a cool project, and we may as well document it. Maybe you guys can learn from our mistakes, and share a little knowledge as well.

Here’s an example of my servo controller /joystick code running:

Servo Controller] (http://www.youtube.com/watch?v=hnXG9wlB68M)

Any input is appreciated,


Dave
PS: More coming soon!

[jetblackaircra]

looks cool guys. Any reason you're starting with a flying wing? Not the most forgiving aircraft to start with. Keep up the good work!

[Idriveaboxster]

You should get one of those 3axis joysticks where the stick turns for the rudder control. I dont know if u need that, but it is definitely helpful on flying helis on simulators. nice project, but i wouldnt start with a flying wing. pitch control can be difficult to master and cg is more difficlut to find. awesome ur using the computer for control.

JPP

[DaveAtDigital]

Why the wing? I like the amount of surface area, but mostly because of looks. Guess i've always liked them.

Yeah, i've seen some pretty slick joysticks out there...and they all have triggers! i need to make the trigger do something for sure.


Tried cutting a wing for the first time!

Screwed it up...but it still looks kinda like a wing, so i dig it. The edges are a little thin because my pivot point was too low, i lost tension on the wire once, so you can see there's a dip in the wing...and i rushed one of the leading edges. I figure i'll screw up a few more before i get it perfect

[dmgoedde]

I like the wing! I made a similar hot wire cutter and have had lots of fun making small wings. I get airfoils from UIUC airfoil data base, print them to laser paper at 100% size, then face-down the paper onto 1/16" ply and rub back of paper with acetone soaked rag to transfer toner image to the wood. It is useful to increase the size of template to overcome kerf of the hotwire by about 1/16" (just leave the template big all around by 1/16")

What happened to your leading edge is a constant problem for me, never having gotten over it completely. Your trailing edge being thin issue will be helped greatly by inflating the template (overcoming kerf comment above) by 1/16" or so, and I find it helpful to wet sand the foam afterwards to smooth it out. The rough surface is not a problem with CNC hot wire cutters, but hand-done it is solved by the sanding.

[jetblackaircra]

You can help to get rid of the problem of the wire not being at the same spot in each template by the following:

On your template, have a set of marks an equal distance around the template (10% increments or so). Have one person on each end of the wire. Use a metronome (time keeping clicking device) at a tempo which is good for your hotwire and make sure you hit each mark on a tick. If you get good at it you can make sure your wire is always at the right place on each side.

Or get a CNC hot wire :-)

Flying wings can definately be a very efficient aerodynamic shape. This has been understood by aerodynamicists for a while. Large aspect ratio, every surface is contributing to the lift, so all drag is induced or skin friction. But, there are many issues, like incorporation of a propulsion device always causes issues, props especially. And, as said in previous posts, CG is not easy to locate, and pitch stability in general is tough. If you use enough sweep angle and enough wash-out the airframe will be very stable and pitch control is as easy as a conventional aircraft, but this will be non-optimal as far as drag goes.

Anyway, good luck. If you need any help with CG location or airfoil selection, let me know.

[Idriveaboxster]

i use the method jetblackaircra uses, just without the metronome. My friend and i do it in pretty fast now. i usually hook in the templates and do the bottom because its less likely to have a mistake. then i move the template up just slightly on the TE so that it does not get too thin. we then go from back to front again and finish it by trimming off the tabs. Works well. After that i set them in the beds and sand for a while just to get a even surface all across.

Are u planning on glassing it or using tape?

JPP

[aeroanalysis]

Just eyeballing your airfoil it looks like you're using a conventional, aft loaded airfoil. For your flying wing to be of decent efficiency it needs a zero or slightly positive pitching moment airfoil. With a conventional airfoil you'll need a LOT of twist to generate lift. The Zagi airfoil is hard to beat for overall flying quality. The coordinates are available on the UIUC airfoil site (Tsagi 8%).

I commend you on your ambition. Here is some unsolicited advice that will get you to your goal a lot quicker - Learn to walk before you run. Get your software and hardware sorted out on an R/C car. If it works there, move on to a very stable rudder/elevator airplane. Something like a polyhedral glider with an electric motor in the nose. If you are serious about getting results, this is the shortest path to success. Why make things hard for yourself? Once you have the rudder-elevator airplane under control, move on to the flying wing. Good luck with your project!

[jetblackaircra]

Sound advice.

[dmgoedde]

DITTO DITTO DITTO!!! Sort out the very basics of navigation with an RC car or truck. Not much can go wrong that will destroy your equipment. You can swear your code should work and is perfect, then testing in the real world flushes out 20 things you never thought of, plus wrong assumptions you started out with. Then with navigation and WP sequencing under you belt a plane like the Miss2 is perfect. It keeps evolving.

I have been a freaking lunatic for autopilots for 2.5 years (and RC flight for 30 years), starting out with RC truck in early 2006, and only now am I able to design and code an autopilot that will stabilize and navigate a flying wing. Good luck.

[CenTexFlyer]

Stay with that wing! We have used flying wings for fully autonomous and SAR platforms for years now in glass, full carbon, and carbon/kevlar. We've worked out all the bad habits and have some very smooth flyers that will fly when others are grounded. Look up "Center of Pressure" and how it works with the center of gravity. Once you work that out, you'll have a pussycat that will do EVERYTHING you ask of it!

[jetblackaircra]

There's a bit more to it than that.... but that's a start. I must say it's a bit more wise to learn to crawl before you try to run.

[DaveAtDigital]

Wow, lots of advice, thanks!

Yeah, I will work on slowing the cutting down, maybe play a song with a constant slow beat and use it to pace myself, good call jetblackairca.

Dmgoedde, do you have a build log or pics of your autopilot? It seems many autopilots don’t try to tackle the flying wing, because they use separate systems for navigation/stabilization.

CenTexFlyer, do you have pics of these projects? A carbon fiber wing would be siccccc.

Idriveaboxter, I have no idea what I’ll use to cover the foam as of now…suggestions ? What is glassing?

I fully expect to crash this thing into the side of a mountain for the first couple flights. (hence the desire to cheaply scratch build the aircraft)The hardware, though somewhat expensive, can be replaced. The software will no doubt survive, and that’s what has value to us. I wonder if NASA had the same optimism with space chimps?

An RC car would be neat, but I think it’s a totally different environment. The first few flights will be 100% manual, but we’ll throw a 6DOF IMU on there to record what’s going on. This way, we can see what it takes to make a turn, to climb etc.

I think we’re almost there! The protocol for servo commands over TCPIP/serial is just about finished, the next vid I’ll post is the same joystick on the laptop, communicating with the aircraft cpu via FM modem (15 mile range). After that, it’s mounting and flying/crashing.

Aeroanalysis, I’m really lost when it comes to airfoils. I looked at the Tsagi 8%. I was hoping to use something with a flat bottom, can you, or anyone else suggest a good flat bottomed airfoil ?

More coming next week!

Attached: Still messing w/ the 6dof. The kalman filter code isn’t in place yet, so the response isn’t super fast yet, but you get the idea.

-Dave

6DOF (http://www.youtube.com/watch?v=M2N42IuxmPI)

[Idriveaboxster]

I use airfoils with some reflex because stability at the cost of some drag usually ends up being worth it, especially on a first wing. What's the wingspan gonna be?

Glassing-fiberglass layup on top to provide a stronger less breakable wing. I like tape because its easily applied, extremely light, and repaired by sticking another piece on.Just use enough carbon reinforcing depending on your max flying speed. Fiberglass is kinda heavy on small planes like under 48in. Above that, the peace of mind of having a stronger plane containing expensive electronics outweighs the extra weight and necessary extra power. Just do tape on the first one and see how it turns out. have u gone to yourzagi.com? i dont think its running anymore, but if u find cached version it shows u everything u need to know that might come up in many tries in building homemade styrofoam wings. awesome site.

Ur doing well and i look forward to more progress.

JPP

[CenTexFlyer]

CenTexFlyer, do you have pics of these projects? A carbon fiber wing would be siccccc.


Here ya go.... a double shot.

Here is a full carbon two piece wing - supports a swappable center section pod. Strong like you would not believe!

Below that is a fully autonomous bird that uses carbon/kevlar hybrid D box leading edge and full glass for the rest of the wing. It has fully embedded servos and control links. By the time it's all said and done, you can remove some internal bracing and let the skin carry the load. So it's just about a wash in weight comparison between film and glass. The big advantage is there is a magnitude order gain in durability!

Bear in mind this is practical application from heavy field use and lots of flying hours. This isn't theory, supposition, or someone's opinion based on what they read that someone else wrote. The real deal!

Gene

[Idriveaboxster]

Hi CenTexFlyer,

Do u have any build pics of the all carbon wing? i started work on one and was doing it with a foam core, but it was very heavy after just one layer, so i wasnt sure about continuing. how much does that weigh? The white one underneath is similar to my 66 inch wing (http://www.rcgroups.com/forums/showthread.php?p=9824489#post9824489) I have never done a carbon leading edge because getting a straight edge on it was a bit difficult and the added weight wasnt really worth it.

Nice planes

JPP

[CenTexFlyer]

Sorry JP, we sure didn't take any pics while we were building that one. Actually we build so many we don't take pictures until they're done:)

I will weigh that wing tomorrow but I want to say it's something like 14 oz with servos, controls, and an aluminum spar.

Thanks for the kudos..... we've been working on and promoting the wings for about 7 years now.

Gene

[Myron]

We do have a couple of other pics of the hybrid wing right after it came out of the vacuum bag.. It actually weighs 23 oz with servos and aluminum and CF spar systems. The wing is super rigid and you cant twist the panels with your bare hands.. We also set a 25lb sack of lead shot on one wing panel and it doesnt bow either..

Myron
www.rpflightsystems.com

[clolson]

We do have a couple of other pics of the hybrid wing right after it came out of the vacuum bag.. It actually weighs 23 oz with servos and aluminum and CF spar systems. The wing is super rigid and you cant twist the panels with your bare hands.. We also set a 25lb sack of lead shot on one wing panel and it doesnt bow either..


I'm involved in a project where we are developing a 7.5' span flying wing. The wing panels are made of foam core with carbon fiber/kevlar/glass hybrid. We had one wing that came out bad, so I suspended it between two 2x4's in my driveway and stood in the middle. As I slowly added my weight I could see the weave pattern in the carbon fiber start to pop out.

Yes, my full weight crunched it, but afterwards, the wing still had plenty of structure and was still way stronger than it needed to be. I could have easily flown with it and probably not noticed any difference between that and a brand new wing ... except for the big foot print depression in the middle.

We've put our test wing through a lot of abuse ... testing different launch and recovery methods in some pretty challenging conditions. We've broken the wing tips off, we've broken the wing root rib, we've even shredded a carbon fiber spare (tube), but as far as I know, the actual wing panel has never taken more than a scratch to the top layer of paint! To be fair, our test center section is showing a lot of wear and tear ... it's difficult to build something that is truly indestructible, yet still light enough to fly. :-)

Curt.

[Idriveaboxster]

Those look awesome. is it 23oz per panel or for the entire wing?

JPP

[CenTexFlyer]

It actually weighs 23 oz with servos and aluminum and CF spar systems.www.rpflightsystems.com

Oh, a mind is such a terrible thing to waste! This CRS disease is killin' me.... Both wing panels together are 23 oz with the goodies in 'em, I watched him weigh them.

[DaveAtDigital]

Well, i wish i had skills like Centex...or some other crafty people on this board...but i don't...so the idea of the scratch wing is on hold. (still REALLY want to learn to carbon fiber a wing though!)

I found a 63'' Katana on sale and threw it together. Not exactly what i picture when i think of computer controlled drone...but, hey, it's a stepping stone.

Bonzai 1] (http://www.youtube.com/watch?v=mqP288cguhY)

We'll be transmitting servo and sensor data via a proprietary transport protocol my buddy made...it works on both serial (for the FM modems) or TCPIP (for 3g wireless cards). And naturally, the first few flights will not be autonomous.

We've got a few things to test out, and some code to clean up, but we should try our first flight within a week. I'll be sure to get some video of our first flight/crash. ;)

[airmcn_3]

Your doing a very good job! Keep up the good work!

[tychoc]

That looks really cool. Those are some wild throws though! Even the "low sensitivity" ones...

Definitely needs missiles :D

-tychoc

[Medve]

CTF, Myron,
Those are some seriously wicked looking planes. Amazing work.