UAV Glider project

[dude_joel]

Gday everyone,
its fantastic to finally be able to add something to this forum, ive been following it for a while now and have only very recently been able to be an active member.
Ive been planning for a long time now to build a large RC glider, capable of long duration flights, and apart from takeoff and landing, be completely controlled by GPS positioning.
unfortunately, like most of us i am unable to "do it alone" so am wondering if anyone on this forum, be it in Australia or overseas, would like to be part of a development team to get this thing in the air.
as a basic starter, the glider will be able to:
*be guided by GPS location
*accurately measure altitude and airspeed
*be able to remain in stable flight autonimously.
*have a solar charging and battery setup able to sustain overnight (or longer) flights.
*be able to automatically detect and utilize thermal activity.
*be able to send live video back to ground station.
*be able to perform long duration flights.

at the moment i think a 6 meter wingspan or there abouts should be suitable for our weight requirements and it will also be enough wing area to house sufficient numbers of solar cells.

before anyone shoots me down for copying AC Propulsion’s Solong UAV, yes it is very similar, but the idea did not come from this. saying that, i think it would be a fantastic idea to borrow some ideas from AC Propulsion, with their permission of course, or even work with them to develop a new glider. the major difference between AC Propulsion’s glider and this one will be it will (hopefully) find and use thermals for climbing and remaining at altitude.

the project will be COMPLETELY OPEN SOURCE, i don't believe in suppressing technology. i am willing to fund the project and put all the parts together.
i also have very large amounts of land for testing available and know a few people in air traffic control that would be ably to help us along when it comes time for testing (and im not talking about testing in the center of Melbourne either)
Here is what i can do for the project:
Build the structure itself, I'm an aircraft engineer and scratch build world war two aircraft as a job. I'm also well and truly qualified in composites and have access to CNC and any other metal forming machinery you can think of.
I can also have prototype PCB's made and organize just about anything anyone working on the project requires.
unfortunately i don't know enough about electronics engineering or programing to be of great help in this area, which is where I'm hoping others will jump on board.
please let me know if your interested, keeping in mind this wont cost anyone apart from me anything, and it doesn't matter if it takes 3 years to get the bugger flying.
cheers
joel

[zlite]

This is a great project, although as you realize, not an easy one. The solar-powered UAV glider I'm most familiar with the Swiss "sun surfer" project (http://sky-sailor.epfl.ch/docs/Report%20Sun-Surfer.pdf) and the two salient points for you from that are:

--You need super high-performance pv cells to get sufficient power to charge the batteries for something like overnight duration. The swiss team had to work with a company making prototype cells and had them custom made on a substrate formed in the shape of their airfoil. I think the total cost of PV cells for the plane was on the order of $50,000.

--Automatic thermal-seeking is very tough. I don't know of any team that has successfully cracked that one.

[dude_joel]

I absolutely agree, it wont be an easy or relatively cheap task, although with advances in technology, solar cells wont cost $50000. the sun surfer project also had many size and time constraints that we don't. for example, the sun surfer was to be under 1 meter wingspan and was to be completed and flown in less than 12 weeks.
their are also a few other projects similar to this, namely AC Propulsion’s Solong UAV and a few others mentioned here: http://sky-sailor.epfl.ch/docs/Conceptual_Design_of_Solar_Powered_Airplanes_for_c ontinuous_flight.pdf
saying this they have done a huge amount of work that will make life easier for whoever is involved in this project.
automatic thermal seeking is very tough, although I'm sure possible. this forum has an enormous knowledge base and i believe if people work together on this, building a thermal seeking UAV is a reasonable challenge.
Zlite, would you be interested in helping with this project? I have read a lot of your work and it is obvious you know an enormous amount about the programing and electronics development side of UAV's. It would be fantastic to have you on board.

[zlite]

I'm flattered to be asked, but I have to admit that my time at the moment is consumed with a project heading in other direction, developing a cheap platform for an indoor aerial robotics contest for kids with blimps. In many ways it's as challenging as your project, given that we have to develop synthetic GPS for indoor use and we have a weight budget of just a bit over 100 grams, but there's little overlap with gliders I'm afraid.

Along with posting here, you might want to also post at our DIY Drones (http://diydrones.com/), which is built on a social network to be a more project-oriented site than this one.

[jtprouty]

I already have a mold made for a payload carrying pod for a motor glider and hope to pull the first fuse out of it in the coming weeks. It's currently designed for a 120-140" wing and will be of pod and boom construction. The fuselage is very slick and should be low drag as well. Wing mounts on top of a pylon to keep it off of the ground when landing and also allow for full span flaps.

Construction will be fiberglass/carbon/kevlar fuse with a carbon tail boom and bagged glass wings. Plenty of room in the pod for all kinds of payload.

Happy flying,

Jimmy

[danstrider]

Hi Joel,

Great project! I know another guy who just might like to talk with you...
http://www.rcgroups.com/forums/showthread.php?t=458540
http://hem2.passagen.se/skj/home.htm

Autonomous soaring is definitely possible. I'm not sure if you've seen the Cloudswift that NASA did a couple years ago, but they have published (yay for public organization!) their algorithms for how to find and use thermals autonomously:
http://ntrs.nasa.gov/search.jsp?R=311009&id=4&qs=N%3D4294704531

If you're looking for a thermal sensor, I'd recommend the Wstech vario. Wolfgang (the owner/designer) can modify the vario to include a pin of the raw vario signal. This can be run through an ADC. (The Skymelody will work if you have digital IO lines.) Combine with a GPS and you're just about on par with NASA's sensor. Add math and you can have a good and cheap thermal finder. I made a logger that was just this and when plotting the flight path X/Y and making the vario signal a color dependant on the strength, thermals start to pop out.

I ran some numbers on the energy budget of SoLong to give a feel of the scope of the project. They were able to sustain flight on 90W, which is what let them fly through the night on their power reserves. 90W for ~12hr is 1080Wh worth of onboard energy storage. IMHO, I think you'll have to match this low power required in order to beat the night-barrier. If you can design a good clean (since very little lift at night, need a low min sink) and strong airframe with the solar cells and the battery bank in mind, there could be a niche market for it.

Also IMHO, Zephyr forgot about the real-world aspect of flying an aircraft and instead maximized for min-sink. Pity though because the airframe doesn't seem as useful as SoLong even though it has a higher endurance figure.

Definitely keep me posted on this project!
Dan


Power calculations for SoLong...
Energy required to fly overnight:
90W * 12hr = 1080Wh

Cells used for storage:
3.6V * 2.6Ah = 9.36Wh for 46.5g

So for 1080Wh:
1080Wh * 46.5g/9.36Wh = 5365.4g = 5.4kg of batteries

[dude_joel]

I already have a mold made for a payload carrying pod for a motor glider and hope to pull the first fuse out of it in the coming weeks. It's currently designed for a 120-140" wing and will be of pod and boom construction. The fuselage is very slick and should be low drag as well. Wing mounts on top of a pylon to keep it off of the ground when landing and also allow for full span flaps.

Construction will be fiberglass/carbon/kevlar fuse with a carbon tail boom and bagged glass wings. Plenty of room in the pod for all kinds of payload.

Happy flying,

Jimmy

thanks for the post.
are you interested in sharing designs of your fuse? what is your mold made of? my wing molds have to be machined alluminium so as to fit the solar panels and handle the "cooking" process used for pre preg.

danstrider,
thats a fantastic post. those links will be invaluable. ill get hold of Staffan Kjerrström shortly and i have no doubt he'll be interested.
your calculations on weight were very close to mine, at about 7kg of batteries and hopefully a total aircraft weight of 16 to 18 kg.

thanks everyone for your fantastic posts. I'm a little run off my feet today so these have to be short replies. ill get into things more tonight hopefully.

[jtprouty]

Hi Joel,

All of my molds are hand laid and have an epoxy gel coat and then several layers of varying weights of glass cloth. I only use epoxy in my molds and parts, no polyurethane.

The plug was made up from pink insulation foam which was sanded to shape and then glassed. A lot of filling and sanding later and it was ready to mold. I'd like to get into hollow core wings but don't have access to CNC equipment needed to make the molds. (We have lots of aluminum blanks to make the molds with tho.)

What information would you like on the fuselage?

Happy flying,

Jimmy

[danstrider]

Battery weight of Solong, as well as a LOT of really nice number to play with can be found on page 2:
http://www.acpropulsion.com/solong/ACP_SoLong_Solar_UAV_2005-06-05.pdf
They list 5.6kg in battery weight and an AUW of 12.8kg (for 4.75m span).

The real trick seems to be how to carry enough capacity to fly overnight and still have enough time during the day to charge it all. Even if you were to duplicate Solong's airframe and dimensions, I'd be impressed. Surely there is room for improvement even if it looked the same!

Looking at auto-soaring...
Since battery energy can be expressed in Watt-hr easily,

From Allen's results, auto-soaring was able to achieve 600m with an average rate of 2.8m/s (bottom of pg 21, NASA/TM-2007-214611). This translates to:

600m at 2.8m/s means 214.3s, hauling 6.8kg

Energy
E = mgh = 6.8kg * 9.81m/s^2 * 600m = 40024.8 m^2 kg / s^2 = 40024.8 Joules

So gained 40000 J over 214.3s

Power
P = Joules/sec = 40000J / 214s = 186W gained (excess)

Including the ~90W required to cruise at a level altitude*
Pgain = 186W + 90W = 276W

Energy gain
276W * 214s * (1Whr/3600J) = 16.4Whr = energy

So over the course of the 214s in the thermal, soaring gained 16.4Whr of energy.

Obviously thermals don't support 2.8m/s avg climb rates very long, so what is needed is a longer time window, perhaps say 30 minutes. That will give the average energy gain to be expected from auto-soaring.

Dan



* 90W for the SBXC comes from eqn 4.3 in: http://art1.mae.ncsu.edu/twiki/pub/Main/CompetitionPapers/M.pdf

[dude_joel]

Thanks everyone for your huge amount of support. at the moment there are a few people that have expressed a lot of interest in helping with this project.
i have decided (and i was a very big decision) to start with a purpose built airframe, instead of retrofitting a commercially available model with uav gear. some of the reasons are:
Commercially built model airframes are not designed for the stress of having a wing full of solar panels and fuse with ~ 8kg of electronics. Although they may fly alright for a short time, something as little as a strong thermal entry or exit could cause structural failure.
Building a purpose built wing means i can properly mount the solar panels into the wing. Ribs, spars, control surfaces, sensors and panels can all be built in together without interfering with each other in a detrimental way.
Also, its the only part that im 100% confident in (building aircraft) because i do i every day. i dont want to become redundant with my own project... :)

Once i find a suitable airfoil and decide on a final wing shape ill get into the building process. So as too not slow down anyone else who is thinking about this project, the wing will incorporate 140 Sunpower A300 cells, giving just over 400w in good sun. Im budgeting on being able to store 1500wh which is a touch over 7kg of batteries. i would like the plane to be able o remain air born on 100watts, this will give us plenty of power to charge the batteries during daylight hours.

any suggestions on airfoil design or wing/tail shape?
cheers
joel

[RaoulDuke]

Your concept will have a problem on cloudy days with good thermals. Either you stay under the clouds getting good lift but less solar power, or you cruise between clouds charging the batteries but using more energy to stay in the air.
How difficult would it be to use the motor/prop as a generator? This way you could circle in the core of a thermal for a longer time, storing energy in your battery instead of gaining altitude.

[danstrider]

I really wonder how high you have to be working lift in order to be in the cloud's shadow. Maybe a full-scale guy would know?

It'd be interesting to see someone actually manufacture a motor/generator combo and show results how much energy they can get back from the altitude into the battery!

[RaoulDuke]

I do fly full-scale gliders, and on a good cumulus day you fly in the shadow for quite a while. Also, the air that is rising under the cloud obviously has to come down somewhere, so you need even more energy for level flight between clouds. Thats why gliders usually fly much faster than the best glide speed between clouds, to minimize the time spent in a sinking air mass.
I think a variable pitch prop with a brushless motor could easily be used as a generator, but you would need a special ESC for that (or disable the ESC during charging and build a charging circuit, which might be much easier).

[dude_joel]

that is a pretty good sounding idea RaoulDuke, i will definitely have a think about it in the later design stages.
At the moment im looking at using the MH114 airfoil.
now to the drawing stage :)

[RaoulDuke]

BTW: I just saw where you come from... Do you work at precision aerospace? I was there two years ago, loved that museum/workshop. I live in germany, but i will spend the next two weeks gliding at corowa :D

[jtprouty]

I did some preliminary work using a brushless motor as a generator. I used a triple phase bridge rectifier and was able to produce the DC we were looking for. My biggest concernt for this application would be drag produced by the prop/generator.

Happy flying,

Jimmy

[dude_joel]

gday RaoulDuke, yes i work at precision aerospace, is you or anyone else are in the area, give me a yell and ill show you through the place. if you like models you'll love this place. if you like warbirds, you'll like it even more. unfortunately due to public liability BS we had to close to the public, so don't just turn up at the door or you'll be sent home...

i think using the motor as a generator would work well. it'll mean using a variable pitch prop though, instead of a folding prop like i was planing. this will mean more drag even when the prop is completely turned into the wind. as long as the prop can generate more energy during thermaling than it costs in drag in normal flight (by diving at the same rate as lift in thermals etc) it would be worth having.

[dude_joel]

I'm drawing the wing at the moment, well when i remember to bring Winfoil home. Does anyone have exact dimensions on the Sunpower A300 solar cells? they don't list full dimensions on the website.

[kk6yb]

I'm drawing the wing at the moment, well when i remember to bring Winfoil home. Does anyone have exact dimensions on the Sunpower A300 solar cells? they don't list full dimensions on the website.
125mm x 125mm according to http://www.sunpower.de/pdf/A-300.pdf

I haven't been able to find a weight anywhere though.
Gary

[danstrider]

Any updates?

[dude_joel]

Sorry i havn't posted updates. I'm in Melbourne for the next week and have been for the past few. As soon as i get home ill get back into it. While I'm in Melbourne I'll organize the pre preg composites to construct the airframe. I'm also still toying with different wing shapes.
I spoke to a few people regarding the A300 cells. Nobody will sell the unpackaged but they will package them in whatever shape we require to attach to the wing. I'll post more in depth discussion next week, when I have access to a friendlier computer.

Joel

[Wilson84]

hi.. i also looking for unpackage Sunpower A300 solar cells. i spoke to sales rep of Sunpower, they only selling 3rd cut cells. Hence, if any one has contacts for A300, can please share with me? I would love to share with you guys if i have any too.